U.S. patent application number 11/220826 was filed with the patent office on 2006-05-04 for surface-emission laser diode operable in the wavelength band of 1.1-1.7 um and optical telecommunication system using such a laser diode.
Invention is credited to Teruyuki Furuta, Shuuichi Hikichi, Akihiro Itoh, Naoto Jikutani, Ken Kanai, Masayoshi Katoh, Kazuya Miyagaki, Akira Sakurai, Shunichi Sato, Shinji Satoh, Takuro Sekiya, Satoru Sugawara, Koei Suzuki, Takashi Takahashi, Atsuyuki Watada.
Application Number | 20060093010 11/220826 |
Document ID | / |
Family ID | 27585757 |
Filed Date | 2006-05-04 |
United States Patent
Application |
20060093010 |
Kind Code |
A1 |
Sekiya; Takuro ; et
al. |
May 4, 2006 |
Surface-emission laser diode operable in the wavelength band of
1.1-1.7 um and optical telecommunication system using such a laser
diode
Abstract
A surface-emission laser diode includes a distributed Bragg
reflector tuned to a wavelength of 1.1 .mu.m or longer, wherein the
distributed Bragg reflector includes an alternate repetition of a
low-refractive index layer and a high-refractive index layer, with
a heterospike buffer layer having an intermediate refractive index
interposed therebetween with a thickness in the range of 5-50
nm
Inventors: |
Sekiya; Takuro; (Kanagawa,
JP) ; Sakurai; Akira; (Kanagawa, JP) ; Katoh;
Masayoshi; (Kanagawa, JP) ; Furuta; Teruyuki;
(Kanagawa, JP) ; Miyagaki; Kazuya; (Kanagawa,
JP) ; Kanai; Ken; (Tokyo, JP) ; Watada;
Atsuyuki; (Shizuoka, JP) ; Sato; Shunichi;
(Miyagi, JP) ; Suzuki; Koei; (Miyagi, JP) ;
Sugawara; Satoru; (Miyagi, JP) ; Satoh; Shinji;
(Kanagawa, JP) ; Hikichi; Shuuichi; (Miyagi,
JP) ; Jikutani; Naoto; (Kanagawa, JP) ;
Takahashi; Takashi; (Miyagi, JP) ; Itoh; Akihiro;
(Miyagi, JP) |
Correspondence
Address: |
OBLON, SPIVAK, MCCLELLAND, MAIER & NEUSTADT, P.C.
1940 DUKE STREET
ALEXANDRIA
VA
22314
US
|
Family ID: |
27585757 |
Appl. No.: |
11/220826 |
Filed: |
September 8, 2005 |
Related U.S. Patent Documents
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
|
|
10085204 |
Feb 26, 2002 |
6975663 |
|
|
11220826 |
Sep 8, 2005 |
|
|
|
Current U.S.
Class: |
372/99 ;
372/50.1 |
Current CPC
Class: |
H01S 5/18313 20130101;
H01S 5/32366 20130101; H01S 5/18361 20130101; H01S 5/0071 20130101;
H01S 5/02251 20210101; H01S 2301/173 20130101; H01S 5/18358
20130101; H01S 5/0683 20130101; H01S 5/0264 20130101; H01S 5/0262
20130101; H01L 2224/48091 20130101; H01S 5/423 20130101; H01L
2224/48091 20130101; H01L 2924/00014 20130101 |
Class at
Publication: |
372/099 ;
372/050.1 |
International
Class: |
H01S 3/08 20060101
H01S003/08; H01S 5/00 20060101 H01S005/00 |
Foreign Application Data
Date |
Code |
Application Number |
Feb 26, 2001 |
JP |
2001-050083 |
Feb 26, 2001 |
JP |
2001-050145 |
Feb 26, 2001 |
JP |
2001-050171 |
Feb 26, 2001 |
JP |
2001-051253 |
Feb 26, 2001 |
JP |
2001-051256 |
Feb 26, 2001 |
JP |
2001-051266 |
Feb 27, 2001 |
JP |
2001-053190 |
Feb 27, 2001 |
JP |
2001-053200 |
Feb 27, 2001 |
JP |
2001-053213 |
Feb 27, 2001 |
JP |
2001-053218 |
Feb 27, 2001 |
JP |
2001-053225 |
Mar 15, 2001 |
JP |
2001-073767 |
Mar 27, 2001 |
JP |
2001-090711 |
Jan 29, 2002 |
JP |
2002-019748 |
Feb 12, 2002 |
JP |
2002-033590 |
Feb 26, 2002 |
JP |
2002-050548 |
Claims
1-8. (canceled)
9. A distributed Bragg reflector, comprising: a first semiconductor
layer having a first bandgap; a second semiconductor layer having a
second bandgap, said first bandgap smaller than said second
bandgap, said first and second semiconductor layers being stacked
alternately; and a material layer, having a third bandgap
intermediate between said first and second bandgaps, provided
between said first and second semiconductor layer, said material
layer changing a valence band energy thereof in a thickness
direction from said first semiconductor layer to said second
semiconductor layer, said material layer comprising a first layer
adjacent to said first semiconductor layer and a second layer
adjacent to said second semiconductor layer, and said first layer
and second layer having first and second rates of compositional
change such that said first rate being larger than said second
rate.
10. A distributed Bragg reflector as claimed in claim 9, wherein
said intermediate layer changes said valence band energy
continuously and gradually from said first semiconductor layer to
said second semiconductor layer.
11. A distributed Bragg reflector as claimed in claim 9, wherein
said intermediate layer changes said valence band energy stepwise
from said first semiconductor layer to said second semiconductor
layer.
12. A distributed Bragg reflector as claimed in claim 9, wherein
said intermediate layer comprises a layer in which said valence
band energy changes continuously and a layer in which said valence
band energy changes stepwise.
13. A distributed Bragg reflector as claimed in claim 9, wherein
said first and second layers have respective first and second
thicknesses, such that said first thickness is smaller than said
second thickness.
14. A distributed Bragg reflector as claimed in claim 9, wherein
there is a stepped change of valence band energy at an interface
between said first semiconductor layer and said material layer.
15. A distributed Bragg reflector as claimed in claim 9, wherein
said first and second semiconductor layers comprise a material of
AlGaAs system.
16. A distributed Bragg reflector as claimed in claim 9, wherein
said first and second semiconductor layers comprise a material of
AlGaInP system.
17. A distributed Bragg reflector as claimed in claim 9, wherein
said first and second semiconductor layers and said intermediate
layer have a carrier density of 5.times.10.sup.17
cm.sup.-3-2.times.10.sup.18 cm.sup.-3, said intermediate layer has
a thickness in the rage of 5-40 nm, and said intermediate layer is
characterized by an average change rate of Al content in the range
of 0.02-0.15 nm.sup.-1.
18-62. (canceled)
Description
BACKGROUND OF THE INVENTION
[0001] This invention generally relates to laser diodes and further
to the art of optical telecommunication that uses a laser diode.
Especially this invention is related to a so-called
surface-emission laser diode that emits a laser beam in a generally
vertical direction to a substrate surface. Also, the present
invention is related to an optical transmission/reception system
and optical-fiber telecommunication system that uses such a
surface-emission laser diode.
[0002] A surface-emission laser diode is a laser diode that emits a
laser beam in a generally vertical direction from a surface of a
substrate. By using surface-emission laser diodes, two-dimensional
array integration of laser diode is achieved easily. Further, the
laser diode has an advantageous feature of relatively narrow
divergent angle of the output optical beam (about 10 degrees),
which is particularly suitable for coupling with optical fibers.
Furthermore, inspection of the laser diode device is made easily in
a surface-emission laser diode.
[0003] Thus, surface-emission laser diodes are suited to construct
an optical transmission module (optical interconnection apparatus)
of parallel-transmission type, and research and development are
conducted prosperously. The immediate application of the optical
interconnection apparatus would be the parallel connection between
computers or circuit boards in a computer, including short-range
optical fiber telecommunication. In future, application to a
large-scale computer network and trunk line system of long-range,
large-capacity telecommunication is expected.
[0004] Generally, a surface-emission laser diode includes an active
layer of a group III-V semiconductor material such as GaAs or
GaInAs, and an optical resonator is formed by disposing an upper
semiconductor Bragg reflector and a lower semiconductor Bragg
reflector arranged respectively above and below the active
layer.
[0005] In such a construction, the length of the optical resonator
is remarkably short as compared with the case of an edge-emission
laser diode. Therefore, it is necessary to increase the reflectance
of the reflector to a high value (99% or more) for facilitating
laser oscillation. Because of this, it is practiced to use a
semiconductor Bragg reflector in a surface-emission laser diode as
a reflector, wherein a semiconductor Bragg reflector is formed of
an alternate and repetitive stacking of a high-refractive index
material such as GaAs and a low refractive index material such as
GaAs with a period of 1/4 wavelength.
[0006] However, in the conventional semiconductor Bragg reflector
that has the structure mentioned above, there arises a spike
structure in the energy band as a result of band discontinuity at
the hetero interface, at which the materials of different bandgaps
are jointed, and the spike structure thus formed tends to function
as a barrier against carriers. Thereby, there arises a problem in
that the semiconductor multilayer part increases the resistance of
the laser diode. Because of this, conventional surface-emission
laser diodes constructed on a GaAs substrate have suffered from the
problem of comparatively high operating voltage of about 2.5 volts
or more. Because of this, it has been difficult to use the
surface-emission laser diode with a CMOS driver integrated circuit,
which produces a laser driving voltage of 2 volts at best. The
itemization of this operating voltage of 2.5 volts is: 1.5V for the
diode part; and 1V for the device resistance. In order to reduce
the operational voltage below 2 V, it is necessary to reduce the
device resistance by one-half, while it is extremely difficult to
meet for this requirement at the present stage of technology.
[0007] In the case of a laser diode of long-wavelength band for use
in optical telecommunication, such as the laser diode of 1.3 .mu.m
band or 1.55 .mu.m, a low voltage operation is expected in view of
the fact that only a voltage of 1 volt or less is applied to the
diode part of the laser diode. Unfortunately, the desired low
voltage operation is not materialized in such a long wavelength
laser diode. In conventional long-wavelength laser diode, InP is
used for the substrate and InGaAsP is used for the active layer. In
such a system, the lattice constant of InP constituting the
substrate is large, and it is difficult to achieve a large
refractive-index difference in the reflector when a material that
achieves lattice matching with the InP substrate is used for the
reflector. Consequently, it has been necessary to stack 40 or more
pairs in the reflector for realizing sufficient reflectance. In
such a construction, however, the resistance of the reflector
increases again as a result of increased stacking number of the
reflector. Thus, it has been difficult to drive the laser diode
driver by a CMOS integrated circuit.
[0008] In a surface-emission laser diode formed on an InP
substrate, there is another problem of change of laser
characteristic caused by the temperature. Because of this, it has
been necessary to add an apparatus for stabilizing the temperature
in the laser diode constructed on such an InP substrate. However,
the use of such a temperature regulator is difficult in the
apparatus for home use, which is subjected to a severe demand of
cost reduction. Because of these problems of increased number of
stacking in the reflector and the poor temperature characteristics,
practical long-wavelength surface-emission laser diode has not yet
commercialized.
[0009] In order to deal with the foregoing problems, there is a
proposal to construct a surface-emission laser diode on a GaAs
substrate by using an AlInP layer, which achieves a lattice
matching with the GaAs substrate, in at least one of the upper and
lower semiconductor Bragg reflectors as the low refractive index
layer, and further by using a GaInNAs layer in at least one of the
upper and lower semiconductor Bragg reflectors, as disclosed in
Japanese Laid-Open Patent Application 9-237942, such that a large
refractive index difference is realized in the reflector and the
number of stacking therein is reduced while maintaining high
reflectance.
[0010] In the foregoing conventional art, the bandgap of the active
layer is reduced by 1.4 eV by using GaInNAs, in which N is
introduced into the III-group V semiconductor material system of
GaInAs. As a result, the laser diode can produce an optical beam
with a wavelength longer than 0.85 .mu.m. In the aforementioned
prior art, it should be noted that the material system of GaInNAs
can achieve a lattice matching with the GaAs substrate. Further,
the prior art describes the semiconductor layer of GaInNAs can be a
promising material for the long-wavelength surface-emission laser
diode operable in the 1.3 micron band and 1.55 micron band.
[0011] In spite of such a description in the prior art with regard
to the possibility of surface emission laser diode operable in the
wavelength band longer than 0.85 .mu.m, there has no such a laser
diode actually materialized. The present situation would be
something like that the theoretical construction is already
established but the actual construction for materializing the laser
diode is not discovered yet.
[0012] In one example, there is a laser diode that uses a
semiconductor Bragg reflector formed by stacking high-refractive
index material layers of GaAs and low refractive index material
layers of AlAs alternately as noted above with the periods of 1/4
wavelength. However, the laser diode structure thus formed does not
provide optical emission at all, or operates but only with low
power, indicating that the efficiency of optical emission is
extremely small.
[0013] Similarly, there is a laser diode disclosed in the Japanese
Laid Open Patent Application 9-237942 in which an AlInP layer is
used for the low refractive index layer of the semiconductor Bragg
reflector. In this case, too, the luminous efficacy of the laser
diode is far from the level of practical use.
[0014] The reason of this unsatisfactory result is attributed to
the chemical activity of the material including Al. More
specifically, it is thought that the use of a material containing
Al easily invites formation of crystal defects originating from Al.
Thus, there have been proposals, as in the Japanese Laid-Open
Patent Application 8-340146 and Japanese Laid-Open Patent
Application 7-307525, to construct the semiconductor Bragg
reflector with materials free from Al such as GaInNP and GaAs.
However, the material system of GaInNP and GaAs can provide a
refractive-index difference of about half as compared with the
material system of AlAs and GaAs. Thus, the stacking number in the
reflector has to be increased, and the object of reducing the
resistance of the surface-emission laser diode is not attained.
[0015] Thus, at present, the surface-emission laser diode operable
at the long-wavelength of 1.1-1.7 .mu.m does not exist, and because
of this, it is not possible to construct a computer network or
optical-fiber telecommunication system that uses such a laser
diode.
[0016] As explained before, in a conventional surface-emission
laser diode, it was also not possible to use a CMOS circuit for the
laser diode driver, and it has been necessary to use an expensive
special driver circuit. On the other hand, if a mass-produced CMOS
driver integrated circuit could be used, the cost of the optical
telecommunication system that uses such a surface-emission laser
diode would be reduced significantly.
[0017] Furthermore the use of a CMOS circuit can reduce the power
supply voltage of the driver integrated circuit as well from 5V to
3.3V. With this, it is possible to reduce the power consumption of
the system to about one-half, and a very large effect of electric
power saving is obtained.
[0018] As noted before, there is a widespread expectation of
optical-fiber telecommunication in relation to computer networks,
and the like. Especially, there is a need of realizing a low cost
system in order that the public accepts such an optical
telecommunication system. Unfortunately, the surface-emission laser
diode that can be used for this purpose and can be used with a
low-cost CMOS driver integrated circuit, and oscillates at the
long-wavelength band of 1.1-1.7 .mu.m does not exist. Hence, the
telecommunication system that uses such a surface emission laser
diode does not exist.
[0019] Meanwhile, in the abovementioned semiconductor Bragg
reflector, in which semiconductor layers of different bandgaps are
grown alternately, there arises the problem of spike formation in
the band structure thereof at the hetero interface as a result of
the band discontinuity. When such a spike structure is formed, the
spike structure functions as a barrier with regard to the carriers.
Thus, there arises a problem in that the electric resistance
becomes very high in the semiconductor multilayer part of the
surface-emission laser diode. This effect also contributes to the
large drive voltage of 2.5 V for the surface-emission laser diode
constructed on a GaAs substrate. As noted previously, it has been
difficult to drive a laser diode having such a large driving
voltage by the driver integrated circuit formed of a CMOS circuit
(driving voltage is below 2 volts).
[0020] As noted previously, the itemization of this operating
voltage of 2.5 volts is: 1.5V for the diode part; and 1V for the
device resistance, and it is necessary to reduce the device
resistance by one-half in order to drive the laser diode with a
drive voltage below 2 volts. However, this is a very difficult
subject.
[0021] Recently, the optical systems are used also for peripheral
transmission/reception systems, and there is a widespread
expectation about the computer networks using the optical-fiber
telecommunication technology including such a peripheral
transmission/reception system. Especially, there is a keen interest
about a low cost optical system required for spreading of the
optical fiber technology to the general public. However, the
surface-emission laser diode that can be used for this purpose and
can be used with a low-cost CMOS driver integrated circuit, and
oscillates at the long-wavelength band of 1.1-1.7 .mu.m does not
exist yet. Hence, the telecommunication system that uses such a
surface emission laser diode does not exist at the moment.
[0022] In such an optical-fiber telecommunication system that uses
the long-wavelength surface-emission laser diode operating at the
wavelength band of 1.1-1.7 .mu.m, the photodetection device
constructed on a Si substrate cannot be used, as such a
photodetection device cannot detect the wavelength of 1.1-1.7
.mu.m. In such a system, it is necessary to use a photodetection
device that has a sensitivity to the wavelength of 1.1-1.7 .mu.m.
However, the photodetection device that has sensitivity to the
desired wavelength band of 1.1-1.7 .mu.m is expensive as compared
with the low cost Si photodetection device. Thus, simple
replacement of a conventional Si photodetection device with the
photodetection device having the sensitivity to the wavelength of
1.1-1.7 .mu.m causes an increase of cost of the whole optical-fiber
telecommunication system. Thus, in order to realize an optical
telecommunication system that uses the long-wavelength
surface-emission laser diode of 1.1-1.7 .mu.m band, an approach
other than replacing the conventional Si photodetection device with
an expensive photodetection device is needed.
[0023] Furthermore, a GaInNAs active layer having a high strain is
used in the long-wavelength surface-emission laser diode, as will
be explained below. In such a laser diode, deterioration of device
characteristic may be caused as a result of the thermal stress
caused by the difference of linear thermal expansion coefficient
with regard to the mounting substrate.
[0024] Meanwhile, in the optical-fiber telecommunication system
that uses a surface-emission laser diode, it is possible to arrange
a number of laser diode elements, each formed of a surface-emission
laser diode, with high integration density. Thus, the distance
between the optical fibers can be reduced as compared with the case
in which a conventional edge-emission laser diode is used for the
laser diode array. Generally, optical fibers accommodated in an
optical cable is provided with a marker band or a plastic ring in
the form of a coloring layer or identification code (ID mark), in
order to allow identification of the transmission line. When the
distance between the optical fibers is reduced, the space available
for these protection layers or rings is reduced.
[0025] In the production of an optical module that accommodates
therein an array of surface-emission laser diodes, it should be
noted that the produced optical module would becomes a detective
product unless a necessary quality is secured for a predetermined
number of laser diode elements in the array. Otherwise, the product
loses the value thereof.
[0026] This issue is related to the yield of the laser diode
production process in the production of the module product that
uses an array arrangement of the laser diode elements, there is an
acute demand of establishing the production process in which the
modules that function normally are utilized efficiently and the
yield of production of the module is improved.
[0027] Summarizing above, there is no available long-wavelength
surface-emission laser diode operable at the wavelength band of
1.1-1.7 .mu.m and that there is no available optical
transmission/reception system that uses such a laser diode.
SUMMARY OF THE INVENTION
[0028] Accordingly, it is a general object of the present invention
to provide a novel and useful surface-emission laser diode operable
in a long wavelength band and an optical transmission/reception
system or optical telecommunication system that uses such a
surface-emission laser diode.
[0029] Another and more specific object of the present invention is
to provide a surface-emission laser diode or laser diode array
having a distributed Bragg reflector tuned to a wavelength of 1.1
.mu.m or longer wherein the electric resistance of the distributed
Bragg reflector is minimized while maintaining high
reflectance.
[0030] Another object of the present invention is to provide a
surface-emission laser diode or laser diode array in which an
intermediate layer is interposed between a high refractive index
layer and a low refractive index layer constituting a distributed
Bragg reflector with a refractive index intermediate between the
high refractive index layer and the low refractive index layer,
wherein the thickness of the intermediate layer is optimized for
minimizing the resistance of the distributed Bragg reflector while
maintaining a high reflectance.
[0031] Another object of the present invention is to provide a
surface-emission laser diode or laser diode array in which an
intermediate layer or heterospike buffer layer is interposed
between a low refractive index layer having a wide bandgap and a
high refractive index layer having a narrow bandgap with an
intermediate bandgap, wherein the compositional profile of Al in
the heterospike buffer layer is optimized so as to minimize the
resistance of the distributed Bragg reflector while maintaining a
high reflectance.
[0032] Another object of the present invention is to provide an
optical interconnection system or optical telecommunication system
using such a surface-emission laser diode or surface-emission laser
diode array.
[0033] Another object of the present invention is to provide an
optical transmission/reception system that uses a long-wavelength
surface-emission laser diode operable at the laser oscillation
wavelength of 1.1-1.7 .mu.m with low operating voltage and small
oscillation threshold current.
[0034] Another object of the present invention is to provide an
optical transmission/reception system suitable for construction
inside a building by using a surface-emission laser diode chip in
which the operating voltage is reduced and the threshold current
for laser oscillation is reduced.
[0035] Another object of the present invention is to provide a
stabilized optical transmission/reception system by using a
long-wavelength surface-emission laser diode chip operating stably
at the wavelength of 1.1-1.7 .mu.m for the optical source.
[0036] Another object of the present invention is to eliminate
various problems that arise when such an optical
transmission/reception system is actually incorporated in an
electronic apparatus.
[0037] Another object of the present invention is to provide a
low-cost and energy-saving optical transmission/reception system by
using a surface emission laser diode chip operable at low voltage
with a small threshold current of laser oscillation.
[0038] Another object of the present invention is to facilitate the
construction of an optical-fiber telecommunication system that uses
a surface-emission laser diode operating at low voltage with low
oscillation threshold current as an optical source, by increasing
the length of the optical fiber cable extending from a module
package beyond a certain length, and hence by improving the
productivity of assembling the module package.
[0039] Another object of the present invention is to provide an
optical-fiber telecommunication system that enables optical
transmission of large capacity with low cost, by using a
surface-emission laser diode having a reduced operational voltage
and reduced oscillation threshold, as an optical source.
[0040] Another object of the present invention is to provide a
reliable optical-fiber telecommunication system by using a
surface-emission laser diode having a low operational voltage and
low oscillation threshold as an optical source, such that the
change of operational characteristic of the laser diode is
suppressed and the lifetime of the laser diode is increased.
[0041] Another object of the present invention is to provide an
optical telecommunication system realizing excellent optical
coupling between a laser diode and an optical fiber, by using a
surface-emission laser diode operable at a low operational voltage
with low oscillation threshold, for an optical source.
[0042] Another object of the present invention is to provide an
optical telecommunication system having a simple construction
characterized by reduced number of parts and is simultaneously
capable of realizing excellent optical coupling as a result of use
of a surface-emission laser diode operable at a low operational
voltage with low oscillation threshold, for an optical source.
[0043] Another object of the present invention is to provide an
optical telecommunication system realizing excellent optical
coupling between a laser diode and an optical fiber, by using a
surface-emission laser diode operable at a low operational voltage
with low oscillation threshold, for an optical source.
[0044] Another object of the present invention is to provide an
optical telecommunication system capable of using a laser diode
without causing damage therein, by using a surface-emission laser
diode operable at a low voltage with low threshold current of laser
oscillation.
[0045] According to the present invention, the surface-emission
laser diode oscillates at the wavelength band of 1.1-1.7 microns
suitable for use in an optical-fiber telecommunications in computer
networks or long-range, large-capacity telecommunication trunks.
The surface-emission laser diode of the present invention
oscillates stably at this wavelength band with low operating
voltage and low oscillation threshold. Conventionally, such a low
surface-emission laser diode did not exist. The laser of the
present invention oscillates at the aforementioned wavelength
region with low operational voltage and low threshold current as a
result of use of an improved semiconductor Bragg reflector. As a
result of low power consumption, the surface-emission laser diode
of the present invention successfully eliminates the heating
problem. Thereby, the surface-emission laser diode of the present
invention oscillates stably. By using such a surface-emission laser
diode, it became possible to construct a practical point-to-point
optical transmission/reception system with low cost.
[0046] In constructing such a point-to-point optical
transmission/reception system, the present invention avoids
localized bending of the transmission path. As a result, the
optical transmission/reception system connects two points easily
and with low cost, without damaging the optical fiber.
BRIEF DESCRIPTION OF THE DRAWINGS
[0047] FIG. 1 is a diagram showing a cross-sectional view of a
long-wavelength surface-emission laser diode according to a first
embodiment of this invention;
[0048] FIG. 2 is a diagram showing an example of reflection
spectrum of a distributed Bragg reflector used in the laser diode
of FIG. 1;
[0049] FIG. 3 is a cross-sectional view showing the constitution of
a semiconductor Bragg reflector used in the long-wavelength
surface-emission laser diode of the first embodiment;
[0050] FIG. 4 is a diagram showing a compositional gradation of a
heterospike buffer layer used in the semiconductor Bragg reflector
constituting a part of the laser diode of the first embodiment;
[0051] FIG. 5 is a diagram showing an example of changing an Al
composition in the heterospike buffer layer;
[0052] FIG. 6 is a diagram showing a result of evaluating a
differentiation sheet resistance of the heterospike buffer layer of
FIG. 4;
[0053] FIG. 7 is a diagram showing the band structure of a
distributed Bragg reflector near a heterojunction interface in a
thermal equilibrium state;
[0054] FIG. 8 is a diagram showing the band structure of the
heterospike buffer layer of FIG. 3 in a thermal equilibrium
state;
[0055] FIG. 9 is a diagram showing an example of the band structure
of the heterospike buffer layer used in the present invention;
[0056] FIG. 10 is a diagram showing an example of the band
structure of the heterospike buffer layer used in the present
invention;
[0057] FIG. 11 is a diagram showing an example of the band
structure of the heterospike buffer layer used in the present
invention;
[0058] FIG. 12 is a diagram showing an example of the band
structure of the heterospike buffer layer used in the present
invention;
[0059] FIG. 13 is a diagram showing the relationship between the
differential sheet resistance of the distributed Bragg reflector
and the Al compositional profile in the heterospike buffer
layer;
[0060] FIG. 14 is another diagram showing the relationship between
the differential sheet resistance of the distributed Bragg
reflector and the Al compositional profile in the heterospike
buffer layer;
[0061] FIG. 15 is a diagram showing another band structure of the
heterospike buffer layer;
[0062] FIG. 16 is a diagram showing the relationship between the
reflectance of the distributed Bragg reflector and the thickness of
the heterospike buffer layer;
[0063] FIG. 17 is a diagram showing the relationship between the
resistance of the distributed Bragg reflector and the thickness of
the heterospike buffer layer;
[0064] FIG. 18 is another diagram showing the relationship between
the resistivity of the distributed Bragg reflector and the
thickness of the heterospike buffer layer;
[0065] FIG. 19 is a further diagram showing the relationship
between the resistivity of the distributed Bragg reflector and the
thickness of the heterospike buffer layer;
[0066] FIG. 20 is a further diagram showing another example of the
band structure of the heterospike buffer layer;
[0067] FIG. 21 is a diagram showing the relationship between the
resistivity of the distributed Bragg reflector having the
heterospike buffer layer of FIG. 20 and the thickness of the
heterospike buffer layer;
[0068] FIG. 22 is another diagram showing the relationship between
the resistivity of the distributed Bragg reflector having the
heterospike buffer layer of FIG. 20 and the thickness of the
heterospike buffer layer;
[0069] FIG. 23 is a diagram showing the relationship between the
resistance of various distributed Bragg reflectors and the
thickness of the heterospike buffer layer;
[0070] FIG. 24 is another diagram showing the relationship of the
resistance of various distributed Bragg reflectors and the
thickness of the heterospike buffer layer;
[0071] FIG. 25 is a diagram showing the relationship between the
reflectance of various distributed Bragg reflectors and the
thickness of the heterospike buffer layer;
[0072] FIG. 26 is another diagram showing the relationship between
the reflectance of various distributed Bragg reflectors and the
thickness of the heterospike buffer layer;
[0073] FIG. 27 is a sectional view showing the constitution of the
long-wavelength surface-emission laser diode according to a second
embodiment of the present invention;
[0074] FIG. 28 is a diagram showing a room temperature
photoluminescence spectrum of the active layer formed of a
GaInNAs/GaAs double quantum well structure;
[0075] FIG. 29 is a diagram showing a sample structure;
[0076] FIG. 30 is a diagram showing a depth profile of nitrogen and
oxygen;
[0077] FIG. 31 is a diagram showing a depth profile of Al;
[0078] FIG. 32 is a diagram showing the structure obtained for a
case in which growth interruption process is used in the form of a
carrier gas purging process;
[0079] FIG. 33 is a diagram showing a depth profile of the Al for
the case in which a growth interruption process is provided and
purging is conducted by using a hydrogen gas;
[0080] FIG. 34 is a diagram showing a depth profile of nitrogen and
oxygen for the case in which a growth interruption process is
provided and purging is made with a hydrogen gas;
[0081] FIG. 35 is a plane view showing a wafer and a laser diode
chip in which the long-wavelength surface-emission laser diode of
this invention is formed;
[0082] FIG. 36 is a diagram showing an example of optical
transmission/reception system that connects an optical source and a
photodetection unit by a straight the transmission path;
[0083] FIG. 37 is a diagram showing an overall view of the above
optical transmission/reception system;
[0084] FIG. 38 is a diagram showing an example of bending the
transmission path connecting an optical source and a photodetection
unit at a right angle for avoiding an obstacle;
[0085] FIG. 39 is a diagram showing an example of the optical
transmission/reception system of this invention that avoids an
obstacle by bending the transmission path between an optical source
and a photodetection unit;
[0086] FIG. 40 is a diagram showing another example of optical
transmission/reception system of this invention that avoids an
obstacle between the optical source and the photodetection unit by
bending the transmission path;
[0087] FIG. 41 is a diagram showing an example of an optical
transmission/reception system that uses a long-wavelength
surface-emission laser diode of this invention as the optical
source;
[0088] FIG. 42 is a plane view showing an example of a room in
which the optical transmission/reception system of this invention
is provided;
[0089] FIG. 43 is a plane view showing an example of a room in
which a conventional optical transmission/reception system is
provided;
[0090] FIG. 44 is a plane view showing a further example of a room
in which a conventional optical transmission/reception system is
provided;
[0091] FIG. 45 is a schematic diagram showing an example of the
optical transmission/reception system of this invention;
[0092] FIG. 46 is a schematic diagram showing another example of
the optical transmission/reception system of this invention;
[0093] FIG. 47 is a diagram showing an example of the
electrophotographic copying machine to which the present invention
is applied;
[0094] FIG. 48 is a diagram showing an example of an
electrophotography copying machine that uses an optical
transmission/reception system of this invention;
[0095] FIG. 49 is a diagram showing an example of an ink jet record
apparatus to which the present invention is applied;
[0096] FIG. 50 is a diagram showing an example of the ink jet
record apparatus that includes an optical transmission/reception
system of this invention;
[0097] FIG. 51 is a diagram showing further example of the optical
transmission/reception system of the present invention;
[0098] FIG. 52 is a diagram showing an example of optical-fiber
telecommunication system to which a long-wavelength
surface-emission diode according to an embodiment of this invention
is connected;
[0099] FIG. 53 is a diagram showing a bi-directional system
constructed by using the long-wavelength surface-emission laser
diode of an embodiment and an optical-fiber telecommunication
system connected thereto;
[0100] FIG. 54 is a diagram showing an example of large capacity
optical-fiber telecommunication system that uses plural optical
fiber groups in which the long-wavelength surface-emission laser
diode of this invention is used;
[0101] FIG. 55 is a diagram showing a long-wavelength
surface-emission laser diode and an optical-fiber telecommunication
system connected thereto according to an embodiment of this
invention;
[0102] FIG. 56 is a diagram showing the constitution of an optical
connection module used in an optical-fiber telecommunication system
together with the long-wavelength surface-emission laser diode
according to an embodiment of this invention;
[0103] FIG. 57 is another diagram showing the constitution of an
optical connection module used in an optical-fiber
telecommunication system together with a long-wavelength
surface-emission laser diode according to an embodiment of this
invention;
[0104] FIG. 58 is a diagram showing another constitution of an
optical-fiber telecommunication system that uses a long-wavelength
surface-emission laser diode according to an embodiment of this
invention;
[0105] FIG. 59 is a diagram showing the constitution of another
optical-fiber telecommunication system that uses a long-wavelength
surface-emission laser diode according to an embodiment of this
invention;
[0106] FIG. 60 is a diagram explaining the constitution of an
optical connection module used in an optical-fiber
telecommunication system together with a long-wavelength
surface-emission laser diode according to an embodiment of this
invention;
[0107] FIG. 61 is a diagram showing the constitution of an
optical-fiber telecommunication system that uses a long-wavelength
surface-emission laser diode according to an embodiment of this
invention;
[0108] FIG. 62 is a diagram showing length setting of optical fiber
for guiding in an optical-fiber telecommunication system that uses
a long-wavelength surface-emission laser diode according to a first
embodiment of this invention;
[0109] FIG. 63 is a diagram showing the constitution of an
optical-fiber telecommunication system that uses a long-wavelength
surface-emission laser diode and plural optical fibers according to
an embodiment of this invention;
[0110] FIG. 64 is a diagram showing an emission angle of an optical
beam emitted from a long-wavelength surface-emission laser diode
according to an embodiment of the present invention;
[0111] FIGS. 65A-65C are diagrams showing the process of fixing
plural optical fibers with a resin according to an embodiment of
this invention;
[0112] FIGS. 66A-66C are diagrams showing a the connect mode of a
long-wavelength surface emission laser diode to plural optical
fibers according to an embodiment of this invention;
[0113] FIG. 67 is a diagram showing possible locations of optical
fibers in a closely packed state according to an embodiment of this
invention;
[0114] FIG. 68 is a diagram showing an example of arranging plural
optical fibers with a closest packed state according to an
embodiment of this invention;
[0115] FIG. 69 is a crass-sectional view showing the constitution
of an optical fiber connected to a long-wavelength surface-emission
laser diode according to an embodiment of this invention;
[0116] FIG. 70 is a plane view showing the constitution of the
optical-fiber telecommunication system that uses a long-wavelength
surface-emission laser diode and an optical fiber connected thereto
according to an embodiment of this invention;
[0117] FIG. 71 is plane view showing the constitution of a
bi-directional optical-fiber telecommunication system that uses a
long-wavelength surface-emission laser diode and an optical fiber
connected thereto according to an embodiment of this invention;
[0118] FIG. 72 is a plane view showing the constitution of LAN in a
building in which the function of a long-wavelength
surface-emission laser diode and the function of an optical fiber
connected thereto are separated;
[0119] FIGS. 73A and 73B are respectively a plan view showing a
long-wavelength surface-emission laser diode and an integral
optical fiber telecommunication apparatus and a cross-sectional
view of an optical waveguide according to an embodiment of this
invention;
[0120] FIGS. 74A and 74B are diagrams showing the constitution of a
long-wavelength surface-emission laser diode chip according to an
embodiment of this invention;
[0121] FIG. 75 is a diagram explaining the operation of a
semiconductor laser diode chip of FIGS. 74A and 74B;
[0122] FIGS. 76A and 76B are diagrams showing the constitution of a
long-wavelength surface-emission laser diode chip according to an
embodiment of this invention;
[0123] FIG. 77 is a diagram explaining the operation of a laser
diode chip shown in FIGS. 59A and 59B;
[0124] FIG. 78 is a diagram showing an optical transmission part of
a telecommunication system in which the long-wavelength
surface-emission laser diode according to an embodiment of this
invention is used;
[0125] FIG. 79 is a diagram showing an optical transmission part of
a telecommunication system in which the long-wavelength
surface-emission laser diode of an embodiment of this invention is
used;
[0126] FIG. 80 is a cross-sectional view showing an optical
coupling device used in a long-wavelength surface-emission laser
diode according to an embodiment of this invention;
[0127] FIG. 81 is a cross-sectional view showing the optical
coupling device of FIG. 63;
[0128] FIG. 82 is a cross-sectional view showing an optical
coupling device that couples with a long-wavelength
surface-emission laser diode according to an embodiment of this
invention;
[0129] FIG. 83 is a cross-sectional view showing an optical
coupling device that couples optically with a long-wavelength
surface-emission laser diode according to an embodiment of this
invention;
[0130] FIG. 84 is a cross-sectional view showing an optical fiber
fixation apparatus used with a long-wavelength surface-emission
laser diode according to an embodiment of this invention;
[0131] FIG. 85 is a cross-sectional view showing the structure of
an optical coupling device;
[0132] FIG. 86 is a diagram showing a positional relationship
between a monitor photodetection device and a surface-emission
laser diode in an optical-fiber telecommunication system that uses
the long-wavelength surface-emission laser diode according to an
embodiment of this invention;
[0133] FIG. 87 is a block diagram showing the constitution of a
control circuit that controls output of a surface-emission laser
diode according to an embodiment of this invention;
[0134] FIG. 88 is a diagram showing the constitution of an
optical-fiber telecommunication system that uses a long-wavelength
surface-emission laser diode according to an embodiment of this
invention;
[0135] FIG. 89 is a schematic diagram showing the constitution of
an optical-fiber telecommunication system that uses a
long-wavelength surface-emission laser diode according to an
embodiment of this invention;
[0136] FIG. 90 is a diagram showing a monitor photodetection
device, a reflection surface and a surface-emission laser diode of
an optical-fiber telecommunication system that uses the
long-wavelength surface-emission laser diode according to an
embodiment of this invention;
[0137] FIG. 91 is a diagram showing then constitution of an
optical-fiber telecommunication system formed of a long-wavelength
surface-emission laser diode, an optical fiber and an optical
connector according to an embodiment of this invention;
[0138] FIG. 92 is a diagram showing the positional relationship
between an optical fiber, a ferule and divided sleeves according to
an embodiment of this invention;
[0139] FIG. 93 is a diagram showing the relationship between a
light emission angle and beam diameter in a long-wavelength
surface-emission laser diode according to an embodiment of this
invention;
[0140] FIG. 94 is a diagram showing the relationship between beam
spreading and a core diameter and an optical path length of a
long-wavelength surface-emission laser diode according to an
embodiment of this invention;
[0141] FIG. 95 is a diagram showing a calculation example of a
relationship between a beam diameter and an optical path length of
a long-wavelength surface-emission laser diode according to an
embodiment of this invention;
[0142] FIG. 96 is a diagram showing the constitution of a
connection part of a laser diode and an optical fiber of an
optical-fiber telecommunication system that uses a long-wavelength
surface-emission laser diode according to an embodiment of this
invention;
[0143] FIGS. 97A and 97B are diagrams showing the constitution of a
connection part of a laser diode and an optical waveguide of an
optical-fiber telecommunication system that uses a long-wavelength
surface-emission laser diode according to an embodiment of this
invention;
[0144] FIG. 98 is a diagram showing the constitution of a
connection part of a laser diode and an optical waveguide in an
optical-fiber telecommunication system that uses a long-wavelength
surface-emission laser diode according to an embodiment of this
invention;
[0145] FIG. 99 is a diagram showing the constitution of an
optical-fiber telecommunication system that uses a long-wavelength
surface-emission laser diode of an embodiment of this invention in
which the laser diode and the optical fiber are coupled
directly;
[0146] FIG. 100 is a diagram showing the constitution of an
optical-fiber telecommunication system that uses a long-wavelength
surface-emission laser diode according to an embodiment of this
invention;
[0147] FIG. 101 is a diagram showing the constitution of an
optical-fiber telecommunication system that uses a long-wavelength
surface-emission laser diode according to an embodiment of this
invention;
[0148] FIG. 102 is a diagram showing the constitution of an
optical-fiber telecommunication system that uses a long-wavelength
surface-emission laser diode according to an embodiment of this
invention;
[0149] FIG. 103 is a diagram showing the constitution of an
optical-fiber telecommunication system that uses a long-wavelength
surface-emission laser diode according to an embodiment of this
invention;
[0150] FIGS. 104A and 104B are diagrams showing an example of
mounting a long-wavelength surface-emission laser diode of an
embodiment of this invention;
[0151] FIG. 105 is a diagram showing the constitution of an
optical-fiber telecommunication system that uses a long-wavelength
surface-emission laser diode according to an embodiment of this
invention;
[0152] FIG. 106 is a block diagram showing the constitution of a
control device used with an optical-fiber telecommunication system
according to an embodiment of this invention in which a
long-wavelength surface-emission laser diode is used;
[0153] FIG. 107 is a diagram showing the constitution of an
optical-fiber telecommunication system that uses a long-wavelength
surface-emission laser diode according to an embodiment of this
invention;
[0154] FIG. 108 is a diagram showing the constitution of an
optical-fiber telecommunication system that uses a long-wavelength
surface-emission laser diode according to an embodiment of this
invention;
[0155] FIG. 109 is a diagram showing the constitution of an
optical-fiber telecommunication system that uses a long-wavelength
surface-emission laser diode according to an embodiment of this
invention;
[0156] FIGS. 110A and 110B are diagrams showing an emission angle
of a laser diode according to an embodiment of this invention;
[0157] FIGS. 111A-111C are diagrams showing an optical fiber bundle
according to an embodiment of this invention;
[0158] FIG. 112 is a diagram showing an optical fiber bundle
according to an embodiment of this invention;
[0159] FIGS. 113A and 113B are diagrams showing the cross-section
of an optical fiber for multimode transmission according to an
embodiment of this invention;
[0160] FIG. 114 is a diagram showing the cross-section of an
optical fiber for single mode transmission according to an
embodiment of this invention;
[0161] FIG. 115 is a diagram showing the electric current-optical
output characteristic of a long-wavelength surface-emission laser
diode of an embodiment of this invention for each temperature;
[0162] FIG. 116 is a diagram explaining the electric current
control of a long-wavelength surface-emission laser diode of an
embodiment of this invention;
[0163] FIG. 117 is a diagram showing the constitution of a
long-wavelength surface-emission laser diode according to an
embodiment of this invention that uses an electric current
control;
[0164] FIG. 118 is a diagram showing the interior of an optical
fiber telecommunication apparatus in which a long-wavelength
surface-emission laser diode of an embodiment of this invention is
provided;
[0165] FIG. 119 is a diagram showing the constitution of a
long-wavelength surface-emission laser diode module according to an
embodiment of this invention;
[0166] FIG. 120 is a diagram showing the constitution of a
long-wavelength surface-emission laser diode according to an
embodiment of this invention;
[0167] FIG. 121 is a diagram showing the constitution of a
long-wavelength surface-emission laser diode module according to an
embodiment of this invention;
[0168] FIG. 122 is a diagram showing an example of laser chip that
is used in this invention;
[0169] FIG. 123 is a diagram showing a different example of laser
chip that is used in this invention;
[0170] FIG. 124 is a diagram showing an example of system of this
invention;
[0171] FIG. 125 is a diagram showing the constitution of such a
laser chip that explains the system of this invention;
[0172] FIG. 126 is a diagram showing an example of control system
of this invention;
[0173] FIG. 127 is a diagram showing a different example of the
control system of this invention;
[0174] FIG. 128 is a diagram explaining the system of this
invention;
[0175] FIG. 129 is another diagram explaining the system of this
invention;
[0176] FIG. 130 is a diagram showing the production process of a
laser array module that uses the long-wavelength surface-emission
laser diode according to an embodiment of this invention;
[0177] FIG. 131 is a diagram explaining the quality control process
used in the production process of FIG. 130.
DETAILED DESCRIPTION OF PREFERRED EMBODIMENTS
First Embodiment
[0178] First, a light-emitting device used in an optical-fiber
telecommunication system of this invention will be explained with
reference to FIG. 1.
[0179] FIG. 1 shows an example of a long-wavelength
surface-emission laser diode that oscillates at the wavelength of
1.1-1.7 .mu.m in which the transmission loss becomes minimum.
[0180] As explained before, while there have been some suggestions
about the possibility of long-wavelength surface-emission laser
diode that oscillating at the wavelength of 1.1-1.7 .mu.m, there
have been no knowledge available with regard to the material and
constitution for the realization such a laser diode.
[0181] This invention provides concrete constitution of such a
long-wavelength surface emission laser diode that uses a material
of GaInNAs system for the active layer.
[0182] In this invention, a high refractive index layer and a low
refractive index layer of n-type AlGaAs respectively having a
composition represented by AlxGa1-xAs (x=1.0) and AlyGa1-yAs (y=0)
are stacked on an n-type GaAs substrate 11 having a (100) surface
orientation alternately and repeatedly for 35 periods with a
thickness .lamda./4 for each layer, wherein the .lamda./4 thickness
is a thickness of 1/4 times the oscillation wavelength .lamda. of
the laser diode. As a result, an n-type semiconductor Bragg
reflector (AlAs/GaAs lower semiconductor Bragg reflectors) 12 is
formed on the GaAs substrate 11.
[0183] Next, an n-type GaInPAs layer 13 having a composition
represented as GaxIn1-xPyAs1-y (x=0.5, y=1), is provided on the
Bragg reflector 12 with a thickness of .lamda./4. In this example,
this n-type GaxIn1-xPyAs1-y (x-0.5, y=1) layer 13 also constitutes
one of the low refractive index layers that forms a part of the
lower part reflector 12.
[0184] Further, a lower part spacer layer 14 of undoped GaAs is
formed on the GaInPAs layer 13, and an active layer 15 having a
multiple quantum well structure, in which three quantum well layers
15a each having a composition represented as GaxIn1-xAs are stacked
on the lower part spacer layer 14 with a GaAs barrier layer 15b
having a thickness of 20 nm interposed therebetween. Further, an
upper part spacer layer 16 of undoped GaAs is provided on the
active layer 15. The active layer 15 forms a resonator 15R having a
thickness .lamda. corresponding to one full wavelength of
oscillation wavelength of the laser diode together with the upper
and lower spacer layers 14 and 16. It should be noted that the
resonator 15R constitutes the active region of the surface-emission
laser diode.
[0185] In the constitution of FIG. 1, a p-type GaInPAs layer 17
doped with C (carbon) and having a composition of GaxIn1-xPyAs1-y
(x-0.5, y=1) is formed further on the spacer layer 16. Further, a
Zn-doped GaAs layer having a composition of AlxGa1-xAs (x=0) and a
Zn-doped AlAs layer having a composition represented as AlxGa1-xAs
(x=1.0) are formed on the p-type GaInPAs layer 17 alternately with
a thickness of 1/4 times the oscillation wavelength .lamda. in each
medium, to form a periodical structure (one period).
[0186] Further, a semiconductor Bragg reflector 18 is formed
thereon by stacking a C-doped, p-type AlGaAs layer having a
composition represented as AlxGa1-xAs (x=0.9) and a Zn-doped,
p-type GaAs having a composition represented as AlxGa1-xAs (x=0)
alternately each with a thickness of 1/4 times the oscillation
wavelength .lamda. of the laser diode in each medium, to form a
periodic structure (25 periods). In this example, the p-type
GaInPAs layer 17 also forms one of the low refractive index layers
constituting a part of the upper part reflector.
[0187] In this embodiment, each of the upper part reflector 18 and
the lower part reflector 12 has the constitution of stacking a low
refractive index layer and a high refractive index layer
alternately, wherein it should be noted that a heterospike buffer
layer having a composition represented as AlzGa1-zAs
(0.gtoreq.y<z<x.gtoreq.1) and a refractive index intermediate
between the low refractive index layer and the high refractive
index layer is interposed in this invention between the low
refractive index layer and the high refractive index layer as shown
in FIG. 2.
[0188] Hereinafter, description will be made about the constitution
of the reflector of the present invention that reflects the
wavelength of 1.1 .mu.m or more in detail with reference to FIG.
2.
[0189] FIG. 2 shows a part of the semiconductor Bragg reflector 18.
A similar constitution is formed also for the semiconductor Bragg
reflector 12. In the description below, explanation for the Bragg
reflector 12 will be omitted in view of the essentially same
constitution of the reflector 18 and the reflector 12.
[0190] FIG. 2 shows an example of the reflection spectrum of the
reflector in which a structural unit of AlAs/GaAs is repeated 24
times (24 pairs). In the example of FIG. 2, each of the AlAs layers
is formed to have a thickness of 93.8 nm and each of the GaAs
layers is formed to have a thickness of 79.3 nm, wherein it should
be noted that these thicknesses are chosen so as to be equal to
1/4n wavelength of the optical radiation that has a wavelength of
1.1 .mu.m in the vacuum environment. Here, n represents the
refractive index of each of the AlAs layer and the GaAs layers. By
setting the thickness of the layers constituting the distributed
Bragg reflector to be equal to 1/4n times of a given wavelength
.lamda., it is possible to tune the reflector to this wavelength
.lamda., and the distributed Bragg reflector thus tuned shows a
large reflectance in the wide wavelength band including the
foregoing tuned wavelength .lamda. This wavelength .lamda. will be
referred to as designed wavelength.
[0191] It should be noted that the material of the AlGaAs system
has various advantages for the material of the distributed Bragg
reflector. For example, the AlGaAs material can be grown on a
commonly available low-cost GaAs substrate with lattice matching.
Further, the material has excellent heat radiation capability as
compared with other semiconductor materials. Furthermore, by using
the material system of AlGaAs, it becomes possible to secure a
large diffractive index as compared with the case of using other
material systems. For example, a refractive index difference of 0.5
is realized between the end member compositions AlAs and GaAs that
constitute the AlGaAs system at the wavelength of 1.3 .mu.m. Thus,
it becomes possible to achieve a large reflectance with reduced
number of stacked pairs as compared with the case of using other
material system.
[0192] FIG. 3 shows the construction of the distributed Bragg
reflector constituting the upper reflector 18 or lower reflector 12
of the laser diode of FIG. 1.
[0193] FIG. 3 is referred to.
[0194] In the present embodiment, each of the upper and lower
reflectors 18 and 12 is formed of a distributed Bragg reflector
having a reflection wavelength of 1.1 .mu.m or more and has a
construction of stacking a lower refractive index layer 18a and an
upper refractive index layer 18b, wherein it can be seen that there
is provided a heterospike buffer layer 18c of AlxGa1-xAs
(0.ltoreq.y<z<z.ltoreq.1) having a refractive index
intermediate between the refractive index of the low refractive
index layer 18a and the high refractive index layer 18b.
[0195] Such a heterospike buffer layer has been studied in relation
to the laser diode of 0.85 .mu.m band. However, it is still in the
stage of feasibility study and no detailed study has been made with
regard to the material, thickness, and like of the heterospike
buffer layer. Further, there has been no proposal at all about such
a heterospike buffer layer in relation to the long-wavelength
surface-emission laser diode of 1.1-1.7 .mu.m band as in the case
of this invention. This is because the long-wavelength
surface-emission laser diode of 1.1-1.7 .mu.m band itself is a new
field and few researches have been made so far.
[0196] The inventor of this invention noticed the usefulness of
optical telecommunication technology that uses a long-wavelength
surface-emission laser diode of 1.1-1.7 .mu.m band and devotedly
conducted a study so as to realize such a laser diode.
[0197] Such a heterospike buffer layer 18c is formed at the time of
formation of the semiconductor Bragg reflector 12 or 18 by an MOCVD
process, by controlling the source gas flow rate, such that the Al
content in the AlGaAs film forming the heterospike buffer layer
changes continuously or stepwise. With this, the refractive index
of a film changes also continuously or stepwise.
[0198] In more detail, the supply rate of Ga and Al is changed such
that the value of the compositional parameter z is changed in the
AlzGa1zAs (0.ltoreq.y<z<x.ltoreq.1) layer from 0 to 1.0, at
the time of the formation of the AlGaAs film 18c. With this, the
film composition of the heterospike buffer layer changes gradually
from GaAs to AlGaAs to AlAs. Such a change of the supply rate is
caused by controlling the gas flow rate at the time of the
formation of film 12c as noted before. A similar effect is obtained
when the ratio of Al and Ga is changed stepwise or
continuously.
[0199] The reason that such a heterospike buffer layer is provided
is to eliminate the problem of increased resistance, which appears
in a semiconductor Bragg reflector, particularly in a p-type
semiconductor Bragg reflector such as the Bragg reflector 18. Such
increase of electric resistance is caused as a result of hetero
barrier formed at the hetero interface where two different
semiconductor layers of the semiconductor Bragg reflector make a
contact. By changing the Al composition gradually at the hetero
interface from the low refractive index layer to the high
refractive index layer as in the case of this invention, there is
also caused a corresponding gradual change of refractive index at
the hetero interface. In this way, the occurrence of the hetero
barrier at the hetero interface is successfully suppressed.
[0200] Hereinafter, more detailed explanation will be made about
such a heterospike buffer layer with reference to FIG. 4.
[0201] FIG. 4 shows an example of the semiconductor Bragg reflector
18 provided with the heterospike buffer layer 18c between two
semiconductor layers 18a and 18b constituting the semiconductor
Bragg reflector. It should be noted that FIG. 4 shows the case of
using the semiconductor material AlzGa1-zAs
(0.ltoreq.y<z<x.ltoreq.1) of the AlGaAs system for the
material of the semiconductor Bragg reflector.
[0202] The two semiconductor layers 18a and 18b constituting the
semiconductor Bragg reflector 18 of FIG. 4 are AlAs and GaAs, and a
compositional gradation layer changing the Al content gradually
therein is provided between the layers 18a and 18b as the
heterospike buffer layer. The heterospike buffer layer thus formed
has valence band energy intermediate between the valence band
energy of AlAs and the valence band energy of GaAs. Thus, the ratio
of Al to Ga is changed from GaAs to AlGaAs to AlAs in the
heterospike buffer layer, and the value of z of the AlzGa1-zAs
(0.ltoreq.y<z<x.ltoreq.1) heterospike buffer layer is changed
gradually from 0 to 1.0.
[0203] In the semiconductor material of the AlGaAs system, the
bandgap energy increases with the Al content and the refractive
index falls off with the Al content. In the conduction band, there
occurs an increase of band energy until the Al content x reaches
0.43, and the band energy starts to decrease thereafter. In the
valence band, on the other hand, the valence band energy falls off
monotonously with the increment of the Al content x. In total, the
bandgap energy increases with the Al content x.
[0204] In the case of the quaternary system of AlGaInP, a similar
trend as in the case of increasing the Al content in the AlGaAs
system appears with increase of the AlInP composition, and the
conduction band energy increases up to the point in which the AlInP
composition has reached 0.7. Thereafter, the conduction band energy
starts to decrease. On the other hand, the valence band energy
decreases monotonously with the increase of AlInP composition.
[0205] In the example of FIG. 4, it should be noted that the rate
of the compositional gradation (rate of increase of bandgap energy)
is set larger in the region near the GaAs layer (region I in FIG.
4) as compared with the region near the AlAs layer (region II in
FIG. 4). For the purpose of comparison, a structure having a linear
compositional gradation layer, in which the Al content therein is
changed only linearly, for the heterospike buffer layer 18c is
shown in FIG. 5.
[0206] FIG. 6 shows the result of evaluation of the electric
resistance of the p-type distributed Bragg reflector 18 tuned to
the reflection wavelength of 1.3 .mu.m, wherein the Bragg reflector
18 is formed by repeating the AlAs/GaAs unit structure, including
therein the heterospike buffer layer of 20 nm thickness between the
AlAs layer 18a and the GaAs layer 18b, four times (four pair
stacking).
[0207] In FIG. 6, each of the layers 18a-18c of the distributed
Bragg reflector 18 including the heterospike buffer layer 18c is
formed of a p-type layer having a career density of
1.times.10.sup.18 cm.sup.-3, and the vertical axis of FIG. 6
represents the value of differential sheet resistance at near zero
bias state. On the other hand, the horizontal axis of FIG. 6
represents the Al compositional gradation in the region I. It
should be noted that the "Al compositional gradation" is defined as
the change of the Al content in the region I divided by the
thickness of the region I. Thus, FIG. 6 represents the case in
which the thickness of the region I is changed variously, while it
should be noted that the total thickness of the region I and the
region II is maintained constant at 20 nm. Thus, the thickness and
the compositional gradation of the region II are determined by the
thickness and the compositional gradation of the region I. The Al
compositional gradation for the case a simple linear compositional
gradation layer is provided between the GaAs layer and the AlAs
layer becomes 0.05 nm.sup.-1. This hits the A point of the
drawing.
[0208] From FIG. 6, it can be seen that the resistance is reduced
further by increasing the Al compositional gradient in the
heterospike buffer layer 18c from the region II to the region I as
compared with the case of FIG. 5 in which the compositional
gradation is linear throughout the heterospike buffer layer 18c.
Also, it can be seen that there exists an optimum Al compositional
gradation in which the resistance becomes minimum. For example, in
the case the thickness of the region I is 10 nm (the same thickness
as the region II), it can be seen that the resistance is reduced to
about 80% of the conventional resistance value obtained for the
case in which the Al compositional gradation is set to 0.09
nm.sup.-1. This trend does not change depending on the applied
voltage.
[0209] Next, the reason of this will be explained.
[0210] FIG. 7 shows the valence band structure of the distributed
Bragg reflector having the AlAs/GaAs structure for the part near
the hetero surface in thermal equilibrium state.
[0211] FIG. 7 is referred to.
[0212] It can be seen that the heterospike originating from the
band discontinuity appear predominantly at the side of the widegap
AlAs layer. In the side of the GaAs layer, the occurrence of the
notch is trifling. Thus, the notch at the side of the GaAs layer
does not become the cause of increase of resistance. Therefore it
is concluded that, for the reduction of resistance of a distributed
Bragg reflector, it is important to eliminate or reduce the
heterospike appearing at the side of the AlAs layer within a
limited thickness of the heterospike buffer layer.
[0213] In the structure of FIG. 7, it can be seen that the Al
content of the heterospike buffer layer 18c is increased sharply at
the side of the GaAs layer 18b in which there occurs the notch
formation, while this corresponds to a gentle compositional
gradient at the side of the AlAs layer, in side which the
remarkable heterospike formation takes place. By doing so, it
becomes possible to reduce the spike formation as compared with the
case of changing the composition of the heterospike buffer layer
linearly as in the case of FIG. 5. When the Al compositional
gradation is set smaller in the region I with regard to the region
II, on the other hand, there occurs an unwanted increase of
resistance.
[0214] FIG. 8 shows the valence band diagram of the structures of
FIGS. 3 and 4 in a thermal equilibrium state, wherein the
continuous line represents the band structure of FIG. 3 while the
dotted line represents the band structure of FIG. 4.
[0215] FIG. 8 is referred to. By using the compositional gradation
profile of FIG. 4, it becomes possible to reduce the compositional
gradation of the heterospike buffer layer 18c in the region II at
the side of the AlAs layer 18a with regard to the region I at the
side of the GaAs layer 18b as compared with case of using the
simple compositional gradation of FIG. 5 with the same thickness.
Thus, it is concluded that the resistance of the semiconductor
distributed Bragg reflector is reduced by increasing the
compositional gradation in the region I of the heterospike buffer
layer I.
[0216] In the example of FIG. 4, it can be seen that the
heterospike buffer layer 16c is formed of two regions I and II each
having a linear Al compositional profile. On the other hand, it is
also possible to change the Al content non-linearly as represented
in FIG. 9. Even in such a case, it is possible to define the
boundary between the regions I and II as an intersection of the
intercept drawn to the valence band at the boundary between the
GaAs layer 18b and the heterospike buffer layer 18c and the
intercept drawn also to the valence band at the boundary between
the AlAs layer 18a and the heterospike buffer layer 18c, as
represented in FIG. 9.
[0217] In the heterospike buffer layer 18c, it is not necessary
that the Al composition changes continuously but another layer may
exist between the regions I and II as represented in FIG. 10.
[0218] According to FIG. 6, the differential sheet resistance of
the Bragg reflector is reduced with decreasing thickness of the
region I. Further, it can be seen that the minimum of the
differential sheet resistance is achieved when there is no region I
provided as represented in FIG. 11 or when the thickness of the
region I is small enough that it can be regarded that there is a
discontinuity in the conduction band between the GaAs layer 18b
(narrow gap layer) and the heterospike buffer layer 18c (widegap
layer) as represented in FIG. 12.
[0219] FIG. 13 shows the differential sheet resistance of the
distributed Bragg reflector similar to the one shown in FIG. 6 as a
function of the Al compositional gradation in the region II. It
should be noted that the Al compositional gradation in the region
II is determined uniquely from the Al compositional gradation in
the region I and the thickness of the region I.
[0220] From FIG. 13, it can be seen that there appears a minimum of
the differential sheet resistance at a particular value of the
compositional gradation in the layer 11, wherein this particular
value of the compositional gradation is irrelevant to the
thicknesses of the regions I and II.
[0221] FIG. 14, on the other hand, shows a similar relationship as
the case of FIG. 13 for the case the total thickens of the regions
I and II is 40 nm.
[0222] Referring to FIG. 14, it can be seen that a similar minimum
of resistance appears at a particular value of the Al compositional
gradation.
[0223] From FIGS. 13 and 14, it is possible to evaluate the optimum
discontinuity of valence band energy between the narrow gap layer
(GaAs layer 18b of FIG. 3) and the heterospike buffer layer 18c for
minimizing the resistance of the distributed Brag reflector having
a heterospike buffer layer.
[0224] As noted above, in the semiconductor distributed Bragg
reflector having a heterospike buffer layer, it is possible to
minimize the resistivity of the reflector by changing the Al
composition of the heterospike buffer layer sharply in the region I
and optimizing the Al compositional gradation in the region II
correspondingly, such that there appears a discontinuity or
substantial discontinuity at the boundary between the narrow gap
layer 18b and the heterospike buffer layer 18c.
[0225] For example, it is possible to reduce the resistance of the
distributed Bragg reflector by about 75% by reducing the thickness
of the region I to 1 nm in the heterospike buffer layer 18c having
a thickness of 20 nm, as compared with the distributed Bragg
reflector having a heterospike buffer layer of the same thickness
except that the Al composition is changed linearly throughout the
heterospike buffer layer.
[0226] In this case, too, it is not necessary that the Al content
changes linearly in the regions I and II of the heterospike buffer
layer 18c but the Al content may be changed non-linearly as
represented in FIG. 15.
[0227] Further, it should be noted that the foregoing consideration
is not limited to the distributed Bragg reflector of the AlGaAs
system but is applicable also to the distributed Bragg reflector of
other material systems such as AlGaInP system. In the case of the
distributed Bragg reflector of the AlGaInP system, it is possible
to achieve a similar effect by changing the AlInP composition in
the heterospike buffer layer.
[0228] The present invention has a feature in that the main layer
18a or 18b and the heterospike relaxation layer 18c constituting
the semiconductor distributed Bragg reflector of FIG. 3, have a
carrier density in the range from 5.times.10.sup.17 cm.sup.-3 to
2.times.10.sup.18 cm.sup.-3, the heterospike relaxation layer 18c
has a thickness in the range from 5 nm to 40 nm, and the average
change rate of Al composition in the region II (see FIG. 4) is in
the range from 0.02 nm.sup.-1 to 0.15 nm.sup.-1. Here, the Al
compositional gradient in the region I is defined as "Al
compositional gradient={variation (0-1) of Al content in the region
I)/dI". By choosing each parameter of the heterospike buffer layer
18c that includes the distributed Bragg reflector to the foregoing
range, reduction of resistance is achieved easily and
effectively.
[0229] Table 12 below shows the Al composition gradient that
provides minimum of the electric resistance and the corresponding
sheet differential resistance for the case of changing the carrier
density of the distributed Bragg reflector and the heterospike
buffer layer 18c (5.times.10.sup.17 cm.sup.-3, 2.times.10.sup.18
cm.sup.-3) and the thickness of the heterospike buffer layer 12 in
the structure of FIG. 9, together with the percentage of the
electric resistance decrease in comparison with the case in which a
simple linear compositional gradient is used for the heterospike
buffer layer (the structure of FIG. 5). TABLE-US-00001 TABLE 12
Heterospike buffer layer 5 .times. 10.sup.17 [cm.sup.-3] 2 .times.
10.sup.18 [cm.sup.-3] thickness carrier density carrier density 5
[nm] 0.16 [nm.sup.-1]/8.4 .times. 10.sup.-6 [.OMEGA.cm.sup.2]/
0.16/4.5 .times. 10.sup.-8 [.OMEGA.cm.sup.2]/ 83% 90% 40 [nm] 0.02
[nm.sup.-1]/2.1 .times. 10.sup.-9[.OMEGA.cn.sup.2]/ . . . 91%
[0230] In view of increase of electrical resistance with decrease
of the carrier density, the value of 5.times.10.sup.17 cm.sup.-3 is
chosen as the actually allowable lower limit. Further, a value of
2.times.10.sup.18 cm.sup.-3 is chosen as the allowable upper limit
in view of conspicuous optical absorption particularly in the case
of a p-type semiconductor.
[0231] In the case the thickness of the heterospike buffer layer 12
is increased, a remarkable decrease of resistance is achieved. On
the other hand, such a decrease of the thickness of the heterospike
buffer layer 12 is not preferable in view of decrease of
reflectance of the distributed Bragg reflector. From the viewpoint
of reflectance, it is believed that the value of 40 nm or less is
important for the practical thickness of the heterospike buffer
layer.
[0232] In the case the thickness is too small, the desired
resistance decrease is not attained. Thus, it is believed that the
value of 5 nm or more is important for the thickness of the
heterospike buffer layer 12 for realizing sufficient resistance
reduction effect.
[0233] As compared with the case of the simple compositional
gradation layer in which the Al content is changed linearly from
the small-bandgap layer to the large bandgap layer constituting the
main layers of the distributed Bragg reflector, the foregoing
construction can achieve further reduction of the resistance.
Within the foregoing range, the differential sheet resistance is
decreased to about 75% (1.2.times.10.sup.-9.OMEGA. cm.sup.2 in
terms of differential sheet resistance) in the embodiment of claim
3, and thus a significant effect is achieved.
[0234] Thus, the present embodiment can reduce the resistance
further as compared with the case of using the linear compositional
gradient in the heterospike buffer layer 18c of the same thickness.
In the case of achieving the same resistance value, on the other
hand, the present embodiment allows the use of reduced thickness
for the necessary heterospike buffer layer 18c. Thus, adversary
effect on the optical properties such as reflectance is
minimized.
[0235] Thus, it becomes possible to obtain a distributed Bragg
reflector excellent in terms of electric properties and optical
properties, by choosing the structure of the distributed Bragg
reflector and heterospike buffer layer as set forth in the
claims.
[0236] Next, the optimum thickness of such a heterospike buffer
layer AlzGa1-zAs (0.ltoreq.y<z<x.ltoreq.1) will be
described.
[0237] FIG. 16 shows the relationship between the reflectance and
thickness of the heterospike buffer layer for a distributed Bragg
reflector tuned to the 0.88 .mu.m band and 1.3 .mu.m band for the
case the heterospike buffer layer 18c has a linear compositional
profile of Al as represented in FIG. 5. The reason that the
reflector tuned to the conventional wavelength of 0.88 .mu.m was
prepared in this experiment is to make a comparison with a
conventional reflector tuned to this wavelength. It should be noted
that GaAs absorbs optical radiation shorter than 0.87 .mu.m. In
FIG. 16, GaAs is used for the high refractive index layer and AlAs
is used for the low refractive index layer.
[0238] The distributed Bragg reflector thus produced needed 18
pairs or more of the low refractive index layer and the high
refractive index layer in order to achieve the reflectance
exceeding 99.9% at the wavelength band of 0.88 .mu.m, while 23
pairs or more were needed for achieving the same reflectance at the
wavelength band of 1.3 .mu.m. It should be noted that FIG. 16
represents the relationship between the reflectance and thickness
of the heterospike buffer layer of the distributed Bragg reflector
tuned to the respective wavelengths.
[0239] Table 1 summarizes the reflectance of FIG. 16.
TABLE-US-00002 TABLE 1 0 nm 5 nm 10 nm 0.88 .mu.m band 99.914
99.912 99.905 1.3 .mu.m band 99.923 99.923 99.920
[0240] Referring to Table 1, it can be seen that there is no
decrease of reflectance in the 1.3 .mu.m band until the thickness
of the heterospike buffer layer 18c reaches 5 nm. In the 0.88 .mu.m
band, on the other hand, there starts a decrease of reflectance at
the thickness of 5 nm of the heterospike buffer layer 18c. In the
case of a surface emission laser diode, the cavity length is
characteristically small and the effect of reflection loss by the
mirror appears conspicuously. Thus, even in the case the decrease
of the reflectance is very small, a large effect is caused in the
threshold current of the laser diode.
[0241] FIGS. 17 and 18 show the resistivity of the AlAs/GaAs
distributed Bragg reflector, having the heterospike buffer layer
18c similar to FIG. 5 and tuned to the reflection wavelength of 1.3
.mu.m, near the zero bias point. In the drawings, the resistivity
is represented in terms of .OMEGA. cm.sup.2 and is defined as dV/dJ
(V: voltage represented in terms of volts; J: current density
represented in terms of A/cm.sup.2). In the illustrated example,
the number of pairs is four. It should be noted that FIG. 17 and
FIG. 18 shows essentially the same diagram except that FIG. 17
shows the diagram in the logarithmic representation while FIG. 18
shows the diagram in the linear scale representation. In FIGS. 17
and 18, the broken line represents the resistivity evaluated from
the bulk crystal resistivity, neglecting the effect of the band
discontinuity. Further, each layer of the distributed Bragg
reflector is doped to the p-type with the carrier density of
1.times.10.sup.18 cm.sup.-3.
[0242] Referring to FIG. 17, it can be sent that the resistivity of
the distributed Bragg reflector decreases with the thickness of the
heterospike buffer layer 18c. In the case of the reflector tuned to
the wavelength of 1.3 .mu.m, the proportion of the compositional
gradation layer 18c with regard to the rest of the layers 18a and
18b is relatively small. Thus, it becomes possible to reduce the
resistivity by two orders by providing the heterospike buffer layer
18c with the thickness of 5 nm, without causing influence to the
reflectance. Longer the reflection wavelength, the heterospike
buffer layer 18c can be formed with increased thickness. When the
thickness is smaller than 5 nm, on the other hand, the desired
reduction of the resistance is not obtained as can be seen from
FIG. 17.
[0243] As represented in FIG. 17, the distributed Bragg reflector
shows a very high resistivity value of 1.OMEGA. cm.sup.2 when the
heterospike buffer layer 18c is omitted. Thus, in the matter of
practical problems, it is extremely difficult to drive a laser
diode having a distributed Bragg reflector in which 20 or more
stacks are provided. In order to drive such a laser diode, a very
high voltage would be needed. Thus, it is difficult to apply such a
distributed Bragg reflector to a current driven device such as a
surface-emission laser diode.
[0244] In the case of providing the heterospike buffer layer 18c
with the thickness of 5 nm as noted above, the resistance of the
reflector is reduced by two orders as compared with the case not
providing such a heterospike buffer layer. As a result, supply of
the drive current to the laser diode is facilitated and laser
oscillation becomes possible. Further, the necessary drive voltage
is reduced, and the problems such as failure or malfunctioning of
the laser diode or reliability are resolved. As represented in
Table 1, there is no decrease of reflectance. Thus, the laser
oscillation can be achieved at a low threshold current density.
[0245] Thus, the foregoing thickness of 5 nm is thought as being
the lower limit of the heterospike buffer layer that enables
decrease of resistance without causing problems to the reflection
performance at long wavelengths. Therefore, the heterospike buffer
layer 18c should be formed with a thickness of 5 nm or more.
[0246] By increasing the thickness of the heterospike buffer layer
18c, the resistivity falls off sharply, and with this, the drive
voltage and the heat generation of the device are reduced.
Associated with this, the operational temperature range of the
laser diode is extended and the laser output is increased.
[0247] In the case of allowing 99.8% reflectance for the
distributed Bragg reflector, the maximum thickness of the
heterospike buffer layer that can be provided to the distributed
Bragg reflector tuned to the 0.88 .mu.m wavelength is limited to 20
nm or less. In the case of the distributed Bragg reflector tuned to
the 1.3 .mu.m, it, is possible to provide the heterospike buffer
layer 18c up to the thickness of 50 nm.
[0248] As represented in FIG. 6, the resistance of the distributed
Bragg reflector decreases with the thickness of the heterospike
buffer layer 18c up to the thickness of 50 nm for the heterospike
buffer layer 18c, and a resistivity of 1.05 times the bulk
resistivity is achieved at the thickness of 50 nm for the layer
18c. When the thickness of the heterospike buffer layer 18c is
increased further, on the otherhand, there appears a saturation in
the decrease of the resistivity.
[0249] It should be noted that the reflectance of the distributed
Bragg reflector starts to decrease sharply with increase of
thickness of the heterospike buffer layer 18c, and the reflectance
is reduced to the level of 99.8% or less when the thickness of the
layer 18c has exceeded 50 nm. Thus, it is necessary to limit the
thickness of the heterospike buffer layer 18c to be 50 nm or less
in order to satisfy the requirement of high reflectance and low
resistivity simultaneously.
[0250] FIG. 19 shows the reduction of the reflectance in detail,
wherein FIG. 19 shows the change rate of the reflectance R (dR/dt)
with regard to the thickness t of the heterospike buffer layer
18c.
[0251] Referring to FIG. 19, it can be seen that there occurs a
sharp rise of the reflectance when the thickness of the heterospike
buffer layer has exceeded 50 nm. With this, there occurs an
increase of the threshold current of laser oscillation.
[0252] Thus, in the distributed Bragg reflector tuned to the
wavelength of 1.3 .mu.m and having a heterospike buffer layer
having a thickness of 5 nm or more but 50 nm or less, it is
possible to reduce the resistance caused by the heterointerface
while maintaining a high reflectance. By using such a distributed
Bragg reflector in a surface-emission laser diode, it is possible
to achieve a laser oscillation under a practical drive
condition.
[0253] In the case of a surface-emission laser diode, it is
necessary to design the reflector such that the distributed Bragg
reflector 18 at the exit side has a relatively smaller reflectance
for facilitating optical output in order to increase the output
power of the laser diode. Further, in order to obtain a stable
laser oscillation up to the high output region (high level
injection region), it is necessary to set the output saturation
point as high as possible by suppressing the device heat
generation. The distributed Bragg reflector having a relatively
thick (50 nm) heterospike buffer layer 18c of the present invention
satisfies these conditions and is suited for use in a high-power
laser diode.
[0254] Thus, according to the long-wavelength surface-emission
laser diode of the present invention, it is possible to optimize
the reflectance and electric properties of the distributed Bragg
reflector used therein by optimizing the thickness of the
heterospike buffer layer 18c within the range of 5-50 nm.
[0255] In the example of FIG. 3, the low refractive index layer 18a
is formed of an AlAs layer and the high-refractive index layer 18b
is formed of a GaAs layer. On the other hand, it is also possible
to use two AlGaAs layers for the low-refractive index layer 18a and
the high refractive index layer 18b, by changing the Al content in
the AlGaAs layers. In this case, it is preferable to increase the
difference of the Al content between the low-refractive index layer
18a and the high refractive index layer 18b for decreasing the
number of stacks in the reflector, as the reflectance of the
reflector is increased with increasing degree of difference of the
refractive index between the low refractive index layer and the
high refractive index layer. Thus, it will be noted that the
construction of FIG. 3 provides the maximum refractive index
difference between the low refractive index layer and the high
refractive index layer by using GaAs and AlAs.
[0256] As noted already, in the heteroepitaxial structure in which
there is a large difference of Al content between the low
refractive index layer and the high refractive index layer, the
band discontinuity at the valence band is also increased and there
arises a problem of increased device resistance in trade off to the
increase of the reflectance. Thus, in such a case, it is necessary
to provide the heterospike buffer layer of sufficiently large
thickness between the high refractive index layer 18b and the low
refractive index layer 18a for reducing the resistivity of the
reflector. However, use of such a heterospike buffer layer has been
difficult in the conventional distributed Bragg reflector tuned to
the wavelength band of 0.85 .mu.m. In the present invention, on the
contrary, it is possible to achieve a high reflectance and
simultaneously a low resistance while using the material system of
GaAs/AlAs for the distributed Bragg reflector.
[0257] In the distributed Bragg reflector of FIG. 3, it is possible
to form the heterospike buffer layer 18c in the range of 20-50 nm
in the case the reflector is tuned to the reflection wavelength of
1.1 .mu.m or more.
[0258] Referring to FIG. 17 again in more detail, it can be seen
that the resistivity of the distributed Bragg reflector decreases
sharply with increase of the thickness of the heterospike buffer
layer 18c and then approaches gradually to the bulk resistivity. In
the example of FIG. 17, the threshold thickness of the heterospike
buffer layer 18c above which the saturation of resistivity takes
place is about 20 nm. At this thickness of 20 nm, it can be seen
that the resistivity is reduced to about twice the bulk
resistivity. Thus, by using the thickness of 20 nm or more but 50
nm or less, it is possible to obtain a distributed Bragg reflector
having a resistivity comparable to the bulk resistivity.
[0259] FIG. 20 shows a band structure for another distributed Bragg
reflector.
[0260] In the example of FIG. 20, GaAs is used for the
high-refractive index layer 18b in the structure of FIG. 3 while
the low refractive index layer lea formed of
Al.sub.0.8Ga.sub.0.2As. Further, the example of FIG. 20 uses a
compositional gradation layer having a thickness of 30 nm for the
heterospike buffer layer 18c in the example of FIG. 20, the
compositional gradation layer is formed such that the valence band
energy changes parabolic with the thickness. In such a parabolic
compositional gradation layer, the valence band energy has a
downwardly convex shape.
[0261] In the construction of FIG. 20, it should be noted that the
design wavelength (.lamda.) of the distributed Bragg reflector is
1.5 .mu.m, and the thickness of the Al.sub.0.8Ga.sub.0.2As layer
18a and the thickness of the GaAs layer 18b are respectively chosen
to 110.8 nm and 125.5 nm so as to be equal to .lamda./4n, wherein n
is a refractive index in the respective layers.
[0262] In the distributed Bragg reflector of FIG. 20, each of the
semiconductor layers is formed to have a thickness subtracted with
the optical thickness of the parabolic compositional gradation
layer 18c from the above noted thickness. In the case an
Al.sub.0.6Ga.sub.0.4As layer is used for the low-refractive index
layer 18a as in the case of the embodiment to be described later,
the .lamda./4n thickness of the Al.sub.0.6Ga.sub.0.4As layer for
the distributed Bragg reflector tuned to the 1.5 .mu.m wavelength
becomes 121.7 nm.
[0263] In the present embodiment, each of the
Al.sub.0.8Ga.sub.0.2As layer 18a, the GaAs layer 18b and the
parabolic compositional gradation layer 18c is doped to the p-type
such that each of the foregoing layers have a uniform carrier
density of 5.times.10.sup.17 cm.sup.-3.
[0264] Thus, in the present embodiment, the resistivity of the
distributed Bragg reflector is reduced to the value generally equal
to the bulk resistivity in spite of the fact that the distributed
Bragg reflector is doped to a relatively small doping density of
5.times.10.sup.17 cm.sup.-3 due to the use of the relatively thick
compositional gradation layer 18c (50 nm). The use of such a low
doping density is also advantageous in view of reduced optical
absorption caused by the transition between the valence bands. It
should be noted that, in such a distributed Bragg reflector tuned
to the longer wavelength of 1.5 .mu.m, it is easy to provide a
thick parabolic compositional gradation layer 18c while maintaining
high reflectance.
[0265] While a compositional gradation layer having a parabolic
gradation is used for the heterospike buffer layer 18c in the
embodiment of FIG. 20, the heterospike buffer layer 18c may be
formed of another layer. Further, the distributed Bragg reflector
may be tuned to the wavelength other than 1.5 .mu.m. Similarly, the
Al composition of the low-refractive index layer may be different
from the one explained above. Further, the doping density may be
changed in each of the layers in the distributed Bragg
reflector.
[0266] As noted above, the resistance of the distributed Bragg
reflector tuned to the long wavelength band of 1.1 .mu.m or more is
reduced sharply when the thickness of the heterospike buffer layer
18c is increased as is represented in FIG. 17, until the thickness
of the heterospike buffer layer 18c exceeds a predetermined
thickness. This predetermined thickness of the heterospike buffer
layer 18c within which the sharp change of the resistivity takes
place is related to the doping density used in the distributed
Bragg reflector.
[0267] FIG. 21 shows the case the doping density of
7.times.10.sup.17 cm.sup.-3 is used in the AlAs/GaAs distributed
Bragg reflector of FIG. 20 for each of the layers therein.
[0268] Referring to FIG. 21 in which the doping density is reduced,
it can be seen that the thickness range of the heterospike buffer
layer 18c in which there occurs a sharp drop of the resistivity in
the distributed Bragg reflector has been increased to 30 nm. When
the foregoing limit is exceeded, it can be seen that the
resistivity is changed linearly to the value of the bulk
resistivity. Particularly, in the case of a p-type distributed
Bragg reflector, there occurs an increase of optical absorption
when the hole density (doping density) is increased as a result of
optical absorption between the valence bands in addition to the
free carrier absorption. Thus, when such a distributed Bragg
reflector is applied to a laser diode, there is caused an increase
of the laser threshold current.
[0269] Thus, from the viewpoint of optical absorption, it is
preferable to use a low carrier density. As the effect of the
transition between the valence bands appears conspicuous for
optical radiation of long wavelengths, it is important to suppress
the absorption loss in the long wavelength laser diode operating at
the wavelength of 1.1 .mu.m or more. In the case of a conventional
distributed Bragg reflector having a doping density exceeding
1.times.10.sup.18 cm.sup.-3, it is difficult to reduce the
absorption loss.
[0270] Because of this reason, there are cases in which the doping
density is reduced to 1.times.10.sup.18 cm.sup.-3 of less in one of
the low refractive index layer 18a, the high refractive index layer
18b, the heterospike buffer layer 18c, or all of the layers
18a-18c. In such a distributed Bragg reflector in which the doping
density is reduced, however, there occurs an increase of extension
of the depletion layer, and the problem of increase of resistance
is inevitable. When the doping density is thus reduced, therefore,
it is necessary to increase the thickness of the heterospike buffer
layer, while it is noted that such a compensating effect of the
heterospike buffer layer appears when the thickness thereof has
reached 30 nm or more. In the case the doping density is reduced
further, a more thick heterospike buffer layer 18c is needed. In
such a case, the effect of the hetero interface is compensated for
by using a larger thickness for the heteroepitaxial buffer layer
within the foregoing range, such as 40 nm or 50 nm. A similar
argument applies also to the case when AlGaAs is used in place of
AlAs in the distributed Bragg reflector.
[0271] FIG. 22 shows the result of calculation of resistivity of
the distributed Bragg reflector including 4 pairs of the
Al.sub.0.8Ga.sub.0.2As/GaAs structure. In the example of FIG. 22,
the doping density of each layer is reduced further to the level of
5.times.10.sup.17 cm.sup.-3.
[0272] In this case, too, the heterospike buffer layer 18c has a
thickness of 30 nm or more and it can be seen that the resistivity
of the distributed Bragg reflector is comparable to the resistivity
of the bulk resistivity. In the case of a distributed Bragg
reflector in which any or all of the high refractive index layer
18a, low refractive index layer 18b and the heterospike buffer
layer 18c are set smaller than the conventional doping density of
1.times.10.sup.18 cm.sup.-3, for example, there appears a
saturation of resistivity when the thickness of the heterospike
buffer layer 18c has reached 30 nm. Thus, in the case at least one
of the layers constituting the distributed Bragg reflector has the
doping density of 1.times.10.sup.18 cm.sup.-3 or less, it is
possible to decrease the resistance effectively by using the
heterospike buffer layer having a thickness in the range of 30-50
nm.
[0273] Of course, the foregoing thickness of the heterospike buffer
layer 18c is effective also in the distributed Bragg reflector
having a larger doping density and may be used in the case all of
the layers constituting the distributed Bragg reflector has a
doping density of 1.times.10.sup.18 cm.sup.-3 or more. On the other
hand, it is especially advantageous to use the doping density of
1.times.10.sup.18 cm.sup.-3 or less in combination with the
heterospike buffer layer having a thickness appropriately chosen
from the range of 30-50 nm in view of the fact that both the
absorption loss and the resistance are reduced.
[0274] By using the distributed Bragg reflector of the embodiment
of FIG. 22 for the reflector in the surface-emission laser diode,
it is possible to obtain a surface-emission laser diode having a
superior characteristic. As noted before, it has been difficult to
realize the thickness of 30-50 nm for the heterospike buffer layer
in the conventional distributed Bragg reflector tuned to the 0.85
.mu.m wavelength. The heterospike buffer layer of this thickness
became possible for the first time in the distributed Bragg
reflector tuned to the wavelength of 1.1 .mu.m.
[0275] In the distributed Bragg reflector of the present invention,
it is also possible to set the difference of Al content between the
low refractive index layer 18a and the high refractive index layer
18b of the distributed Bragg reflector to be less than 0.8 by
constituting the layers 18a and 18b by any of AlAs, GaAs and AlGaAs
mixed crystal.
[0276] When the difference of Al content between the high
refractive index layer 18b and the low refractive index layer 18a
is less than 0.8 in the distributed Bragg reflector constituted of
a semiconductor material of the AlGaAs system and tuned to the
design wavelength of 1.1 .mu.m or longer, it is possible to reduce
the electric resistance while maintaining a high reflectance.
[0277] In a mixed crystal of AlGaAs, it should be noted that the
valence band energy decreases monotonously with the increase of Al
content therein, and the band discontinuity of the valence band at
the interface to a GaAs crystal increases with increasing Al
content in the AlGaAs mixed crystal. Thereby, there is formed it
large potential barrier at the hetero interface, and this potential
barrier has been the cause of the high resistance of the
distributed Bragg reflector. Further, it should be noted that the
decrease of the valence band energy is generally proportional to
the Al content, and the band discontinuity between the
semiconductor layers having different Al compositions correspond to
the Al compositional difference.
[0278] FIGS. 23 and 24 show the resistivity of the p-type
distributed Bragg reflector tuned to the design wavelength of 1.3
.mu.m and including 4 pairs of high refractive index layer 18b and
low refractive index layer 18a for various thicknesses of the
heterospike buffer layer. In FIGS. 23 and 24, it should be noted
that the distributed Bragg reflector uses GaAs for the high
refractive index layer 18b and AlAs, Al.sub.0.8Ga.sub.0.2As and
Al.sub.0.6Ga.sub.0.4As have been used for the low refractive index
layer. Further, each of the layers has been formed to have a
thickness tuned to 1/4 the design wavelength while taking into
consideration of the refractive index of the respective layers.
Further, the doping density is set to 5.times.10.sup.-17 cm.sup.-3
throughout the layers in FIG. 23 and 1.times.10.sup.-18 cm.sup.-3
throughout the layers in FIG. 24. It should be noted that the
doping density of 1.times.10.sup.-18 cm.sup.-3 is used commonly in
the doping of the conventional p-type distributed Bragg
reflector.
[0279] From FIGS. 23 and 24, it can be seen that the resistivity of
the distributed Bragg reflector increases with increasing Al
content in the AlGaAs layer and decreasing thickness of the
heterospike buffer layer, and that it is necessary to provide a
thick heterospike buffer layer in order to reduce the resistivity
of the distributed Bragg reflector to the bulk resistivity level.
In the case of the distributed Bragg reflector that uses
Al.sub.0.6Ga.sub.0.4As and GaAs as the low refractive index layer
18a and the high refractive index layer 18b, for example, there
appears a band discontinuity of about 300 meV, while the band
discontinuity increases to about 400 meV when
Al.sub.0.8Ga.sub.0.2As is used for the low refractive index layer
18a and GaAs is used for the high refractive index layer 18b.
[0280] From the results of FIGS. 23 and 24, it is concluded that
the thickness of 20 nm or more is preferable for the heterospike
buffer layer 18c in order to reduce the resistivity of a
distributed Bragg reflector formed of AlAs layers and GaAs layers
in order to reduce the resistivity of the distributed Bragg
reflector effectively, although this would also depend on the
doping density. Thus, in the case the band discontinuity is 400 meV
or less, in other words the Al compositional difference is less
than 0.8, it can be seen that the resistivity of the heterospike
buffer layer can be reduced effectively by increasing the thickness
of the heterospike buffer layer to 20 nm or more.
[0281] Actually, the increase of resistance of the distributed
Bragg reflector caused by the band discontinuity depends not only
on the barrier height and barrier thickness but also on the
effective mass of holes used for the carriers. However, there is no
large difference in the effective mass for heavy holes between
AlGaAs, AlGaInP and GaInAsP, and the band discontinuity can be
regarded as the index of the heterointerface resistance. Thus, in
the case the valence band discontinuity is less than 400 meV, and
hence the Al compositional difference is less than 0.8, it is
possible to reduce the resistance of the distributed Bragg
reflector by using the heterospike buffer layer having a thickness
of 20 nm or more.
[0282] With regard to the upper limit of the heterospike buffer
layer 18c, the thickness of the heterospike buffer layer 18c should
be chosen, in view of the tuned wavelength of the distributed Bragg
reflector, such that no remarkable decrease of reflectance occurs.
By doing so, a distributed Bragg reflector having excellent
electrical and optical characteristics is obtained.
[0283] In the distributed Bragg reflector having a structure shown
in FIG. 3, it is possible to use any of AlAs, GaAs and AlGaAs mixed
crystal for the low refractive index buffer layer 18a and the high
refractive index layer 18b such that the Al compositional
difference between the layers 18a and 18b is 0.8 or more.
[0284] In this embodiment, it is also possible to reduce the
resistance of the distributed Bragg reflector effectively while
maintaining a high reflectance.
[0285] When AlAs and GaAs are combined for the low refractive index
layer 18a and the high refractive index layer 18b, it should be
noted that there appears a band discontinuity of about 500 meV in
the valence band, and thus, it is necessary to provide a thick
heterospike buffer layer for reducing the resistivity of the
distributed Bragg reflector, although the thickness of the
heterospike buffer layer may depend on the doping density. In such
a case, it is concluded, in view of the results of FIGS. 23 and 24,
that the heterospike buffer layer 18c has a thickness of 30 nm or
more. Thus, from the view point of band discontinuity of the
semiconductor materials constituting the distributed Bragg
reflector, it becomes possible to reduce the resistance of the
distributed Bragg reflector by setting the thickness of the
heterospike buffer layer to be 30 nm or more in the case there
exists a band discontinuity is 400 meV or more.
[0286] Thus, there holds a relationship explained with reference to
FIGS. 23 and 24 between the Al compositional difference and the
valence band discontinuity, and the valence band discontinuity of
400 meV corresponds to the Al compositional difference of 0.8 or
more. Thus, in the case the Al compositional difference is 0.8 or
more, the thickness of 30 nm or more is needed for the heterospike
buffer layer 18c so as to reduce the resistance effectively. By
providing such a heterospike buffer layer as such, it is possible
to reduce the resistance of the distributed Bragg reflector
effectively.
[0287] With regard to the upper limit of the heterospike buffer
layer, the thickness of the heterospike buffer layer 18c is chosen
in view of the tuned wavelength of the distributed Bragg reflector
such that there occurs little decrease of reflectance. Thereby, it
becomes possible to realize a distributed Bragg reflector having
excellent characteristics both in terms of electric properties and
optical properties.
[0288] In the distributed Bragg reflector of the present invention
tuned to the design wavelength of 1.1 .mu.m, it is also possible to
set the thickness of the heterospike buffer layer 18c with respect
to the tuned wavelength .lamda. [nm] to be equal to or smaller than
(50.lamda.-15) [nm]. In such a case, too, it is possible to reduce
the resistance of the distributed Bragg reflector while maintaining
high reflectance. FIG. 25 shows the relationship between the
thickness of the heterospike buffer layer and reflectance of the
distributed Bragg reflector tuned to the wavelength of 1.1-1.7
.mu.m. It should be noted that the distributed Bragg reflector has
a structure explained with reference to FIG. 3 and uses GaAs for
the high refractive index layer 18b and AlAs for the low refractive
index layer 18a. The layers 18a and 18b are repeated with the
number determined such that the reflectance exceeds 99.9% at each
of the tuned wavelengths. Thus, in the case of the reflector is
tuned to 0.88 .mu.m, the layers 18a and 18b are repeated 18 times,
while in the case of the reflector is tuned to 1.1 .mu.m, the
layers 18a and 18b are repeated 22 times. Further, in the case the
reflector is tuned 1.3 micron, the layers 18a and 18b are repeated
23 times while in the case the reflector is tuned to 1.5 .mu.m, the
layers 18a and 18b are repeated 23 times. Further, in the case the
reflector is tuned to 1.7 .mu.m, the layers 18a and 18b are tuned
to 24 times.
[0289] FIG. 26 shows a change rate of the resistivity of the
distributed Bragg reflector shown in FIG. 25 (dR/dt) with the
thickness of the heterospike buffer layer.
[0290] From FIG. 25, it can be seen that reflectance of the
reflector decreases with increasing thickness of the heterospike
buffer layer. Further, FIG. 26 shows that the decrease of the
reflectance increases suddenly at a certain thickness of the
heterospike buffer layer. For the sake of facilitating
understanding of this situation, FIG. 26 shows lines drawn as a
tangential for each of the curves as the reference reflectance
change rate. For example, in the distributed Bragg reflector tuned
to 1.3 .mu.m, it can be seen from FIG. 26 that the change rate of
the reflectance increases sharply when the thickness of the
heterospike buffer layer 18c has exceeded 50 nm. In correspondence
thereto, there occurs a sharp drop in the reflectance of the
distributed Bragg reflector. Thus, in the surface-emission laser
diode having such a distributed Bragg reflector for the reflector,
the threshold current of laser oscillation increases sharply at the
foregoing thickness of the heterospike buffer layer 18c. Further,
it should be noted that the thickness of the heterospike buffer
layer 18c corresponding to such a sharp change of reflectance
changes depending on the design wavelength of the laser diode.
Thus, longer the design wavelength of the laser diode, the
thickness of the semiconductor layers constituting the reflector is
increased, and the effect of the heterospike buffer layer is
reduced.
[0291] Thus, the thickness of the heterospike buffer layer
corresponding to such a sudden increase of change rate of the
reflectance changed depending on the design wavelength, while the
threshold of the change rate corresponding to the onset of the
sudden growth does not change with the design wavelength and
maintains the value of about 0.9 as can be seen from FIG. 26.
[0292] Table 2 below summarizes the threshold thickness for each
wavelength shown in FIG. 25. TABLE-US-00003 TABLE 2 design
wavelength 1.1 .mu.m 1.3 .mu.m 1.5 .mu.m 1.7 .mu.m threshold
thickness 40 nm 50 nm 60 nm 70 nm
[0293] From Table 2, it can be seen that the design wavelength and
the threshold thickness are in a generally linear relationship and
there holds a relationship between the threshold thickness t [nm]
and the design wavelength .lamda. [nm] of the distributed Bragg
reflector as follows: t=0.05.lamda.-15. (1)
[0294] Thus, by providing the heterospike buffer layer 18c in the
distributed Bragg reflector tuned to the wavelength of 1.1 .mu.m or
more with the thickness not exceeding the thickness t given by
Equation (1), it is possible to realize a low-resistance
distributed Bragg reflector having a high reflectance. In the
example above, it was assumed that the heterospike buffer layer has
a linear compositional graduation profile. However, it is also
possible to use non-linear profile. In this case, too, similar
results and effects are achieved.
[0295] In the present invention, it is possible to increase the
thickness of the heterospike buffer layer to be 20 nm or more in
the distributed Bragg reflector having a thickness of 1.1 .mu.m. In
this case, too, it is possible to reduce the resistance of the
mirror while maintaining a high reflectance.
[0296] As explained before, there holds a relationship of Equation
(1) between the thickness of the heterospike buffer layer and the
design wavelength of the distributed Bragg reflector that can
maintain a high reflectance.
[0297] With regard to the electric properties of the distributed
Bragg reflector, it is possible to reduce the effect of the hetero
interface by increasing the thickness of the heterospike buffer
layer as noted before, and a distributed Bragg reflector having a
reduced resistance is obtained. Further, the effect of the
resistance reduction by the heterospike buffer layer id determined
by the materials used for the distributed Bragg reflector and the
doping density and further by the compositional profile.
Essentially, it does not depend on the reflection wavelength. Thus,
it is concluded that there exists a lower limit in the thickness of
the heterospike buffer layer in which the resistance of the
distributed Bragg reflector is decreased sufficiently. In order to
provide a distributed Bragg reflector having a sufficiently low
resistance, it is necessary to provide the heterospike buffer layer
with a thickness exceeding a predetermined thickness.
[0298] In the example of the distributed Bragg reflector of FIG. 24
in which the semiconductor layers are doped uniformly to the doping
density of 1.times.10.sup.18 cm.sup.-3, it can be seen that the
resistivity of the distributed Bragg reflector increases in terms
of orders when the thickness of the distributed Bragg reflector is
les than 20 nm. When the thickness of the heterospike buffer layer
has exceeded the value of 20 nm, on the other hand, it is
preferable to use the thickness of 20 nm or more for the
heterospike buffer layer, provided that the semiconductor layers
constituting the distributed Bragg reflector has the foregoing
doing density.
[0299] From the description noted above, it is concludes that a
distributed Bragg reflector having a sufficiently low resistance
and simultaneously maintaining a high optical reflectance is
obtained by choosing the thickness t [nm] of the heterospike buffer
layer with regard to the design wavelength .lamda. of the
distributed Bragg reflector so as to satisfy the relationship
20.ltoreq.t.ltoreq.50.lamda.-15.
[0300] In the distributed Bragg reflector of this embodiment, it is
also possible to set the thickness of the heterospike buffer layer
to be 30 nm or more. Even in such a case, it is possible to
decrease the resistance of the reflector effectively while
maintaining a high reflectance at the tuned wavelength or design
wavelength of 1 .mu.m or longer.
[0301] In semiconductor materials, there is a tendency that optical
absorption increases also for the photons having energy smaller
than the bandgap with increase of free carriers. Further, in the
case of the p-type semiconductor materials, there arises
conspicuous optical absorption caused as a result of the valence
band to valence band absorption with the increase holes acting as
the carriers. As the problem of the valence band to valence band
absorption becomes conspicuous with increasing optical wavelength,
this problem becomes a serious problem in the distributed Bragg
reflector having a long wavelength of 1.1 .mu.m or more. It should
be noted that such optical absorption becomes the cause of
decreasing the reflectance of the distributed Bragg reflector. In
the case of the laser diode using such a reflector, there arises a
problem of increase of threshold current caused by optical
absorption and decrease of efficiency. Thus, from the view point of
decreasing the optical absorption, it is preferable that the doping
density of the semiconductor layers is set as small as possible.
However, with the decrease of the doping density, there arises a
problem of increase of thickness of the heterointerface, and the
effect of the interface potential is increased, thus leasing to the
problem of increase of resistance of the distributed Bragg
reflector.
[0302] Thus, in the case of the semiconductor distributed Bragg
reflector having the reduced doping density, it is necessary to use
a thicker heterospike buffer layer for reducing the resistivity. In
the case of doping the semiconductor layers constituting the
distributed Bragg reflector to the doping density of
5.times.10.sup.17 cm.sup.-3, it can be seen from FIG. 23 that the
resistivity is reduced to the level of the bulk material when the
thickness of the heterospike buffer layer is increased to 30 nm or
more.
[0303] With regard to the doping density and profile, there are
numerous combinations, while there arises a similar tendency when
the doping density of one of the semiconductor layers is less than
1.times.10.sup.18 cm.sup.-3. This is because the potential barrier
formed at the hetero interface has the height and thickness
depending on the doping density of the semiconductor layer adjacent
to the hetero interface. Lower the doping density, as in the case
of the doping density having a value of 1.times.10.sup.18
cm.sup.-3, larger the influence of the hetero interface. Further,
the electric property of the distributed Bragg reflector is mainly
determined by the hetero interface where the doping density is
small. Thus, the present invention is effective also in the case in
which at least one of the semiconductor layers constituting the
distributed Bragg reflector has a doping density of less than
1.times.10.sup.18 cm.sup.-3.
[0304] It should be noted that a similar tendency appears also in
the case of the distributed Bragg reflector doped to the order of
1.times.10.sup.17 cm.sup.-3. Of course, it is possible to reduce
the doping density further, provided that the thickness of the
heterospike buffer layer is increased to 30 nm or more within the
upper limit noted before.
[0305] Thus, according to the present embodiment, it is possible to
obtain a distributed Bragg reflector having a sufficiently low
resistance and high optical reflectance by choosing the thickness t
[nm] of the heterospike buffer layer with regard to the design
wavelength .lamda. of the laser diode so as to fall in the range of
30.ltoreq.t.ltoreq.50.lamda.-15.
[0306] In the conventional laser diode operable at the wavelength
band of 0.85 .mu.m, there has been a study to provide a heterospike
buffer layer as noted above. On the other hand, such a heterospike
buffer layer is most effectively used in the long-wavelength
surface-emission laser diode operable at the wavelength of 1.1-1.7
.mu.m. In the 1.1-1.7 .mu.m band, for example, it is possible to
set the thickness of the material layer constituting the
heterospike buffer layer about twice the thickness for the case of
the 0.85 .mu.m band, in order to obtain the same reflectance (99.5%
or more, for example). Thereby, the resistance of the semiconductor
Bragg reflector is reduced, and the operational voltage,
oscillation threshold current, and the like, are likewise reduced.
Thereby, advantageous features such as suppression of heating,
stable laser oscillation and low energy drive is obtained for the
laser diode.
[0307] Thus, the provision of such a heterospike buffer layer to
the semiconductor Bragg reflector according to the present
invention is deemed an advantageous improvement especially in the
case of the long-wavelength surface-emission laser diode operable
in the laser oscillation wavelength of 1.1-1.7 .mu.m.
[0308] For example, in the case of the surface-emission laser diode
operable at the wavelength band of 1.3 .mu.m and having a
semiconductor Bragg reflector in which a low refractive index layer
of AlxGa1-xAs (x=1.0) and a high refractive index layer of
AlyGa1-yAs (y=0) are stacked for 20 periods, a reflectance of 99.7%
or less is obtained, provided that the thickness of the hetexospike
buffer layer AlzGa1-zAs (0.ltoreq.y<z<x.ltoreq.1) is set to
30 nm. Further, a reflectance of 99.5% or more is obtained when the
heterospike buffer layer has a thickness of 53 nm. Thus, a film
thickness control of .+-.2% is sufficient when the distributed
Bragg reflector is designed to have a reflectance of 99.5% or more.
In the experiments, distributed Bragg reflectors were produced with
the thickness of 10 nm, 20 nm and 30 nm for the heterospike buffer
layer. It turned out that a reflectance sufficient for practical
use was achieved in any of these experiments. In this way, a 1.3
.mu.m band surface-emission laser diode having a reduced resistance
for the semiconductor Bragg reflector was realized and laser
oscillation was achieved. The remaining structural feature of the
laser diode thus produced will be explained later.
[0309] In a multilayer reflector, there exists a band of high
reflectance in the designed wavelength band (assuming that complete
film thickness control is achieved). This is called high
reflectance band, wherein the high reflection band may include also
the wavelength band in which a reflectance exceeding a target value
is achieved for a target wavelength. In the high reflectance band,
the reflectance becomes largest at the designed wavelength, while
the reflectance falls off but only slightly as the deviates from
the designed wavelength. When the deviation exceeds a certain
limit, the reflectance falls off sharply.
[0310] Thus, in a multilayer reflector, it is necessary to control
the thickness of the multilayer reflector completely with atomic
layer level so that a reflectance exceeding the necessary
reflectance is that achieved at the target wavelength. In practice,
a deviation of about .+-.1% is inevitable for the film thickness,
and thus, it is a common place that the target wavelength and the
wavelength in which the reflectance is maximized are different. In
the case the target wavelength is 1.3 .mu.m, for example, the
wavelength providing the maximum reflectance is deviated by 13 nm
when there is an error of .+-.1% in the film thickness control.
Therefore, it is desirable that this high reflectance band has a
wide bandwidth. Here the high reflectance band is defined as the
wavelength band in which a reflectance exceeding a necessary
reflectance is obtained for a target wavelength.
[0311] Thus, in the long-wavelength surface-emission laser diode
oscillating at the wavelength of 1.1-1.7 .mu.m, it is possible to
reduce the resistance value of the semiconductor Bragg reflector
while maintaining high reflectance, by optimizing the constitution
of the reflector. Thereby, the operating voltage, oscillation
threshold current, and the like, of the laser diode are
successfully reduced, and heat generation is suppressed
effectively. As a result, stable laser oscillation is realized and
low energy driving of the laser diode becomes possible.
[0312] Once again FIG. 1 is referred to.
[0313] It can be seen that a p-type GaAs layer 19 having a
composition represented as AlxGa1-xAs (x=0) is provided on the
upper semiconductor Bragg reflector 18 as a contact layer
(p-contact layer), so as to achieve a contact with the p-side
electrode 20.
[0314] In the constitution of FIG. 1, it should be noted that the
content x of the quantum well active layer is set to 39% (Ga
0.61In0.39As). Further, the thickness of the quantum well active
layer is set to 7 nm. The quantum well active layer thus formed
accumulates a compressional strain of about 2.8% with respect to
the GaAs substrate.
[0315] In the surface-emission laser diode of FIG. 1, the
deposition of the semiconductor layers is conducted by an MOCVD
process. In this case, no lattice-relaxation phenomenon was
observed. Each of the semiconductor layers of the laser diode may
be formed by using TMA (trimethyl aluminum), TMG (trimethyl
gallium), TMI (trimethyl indium), AsH.sub.3 (arsine), PH.sub.3
(phosphine) as source materials, together with a carrier gas of
H.sub.2. In the case of growing the active layer (the quantum well
active layer) in the form of highly strained layer as in the case
of the device of FIG. 1, it is preferable to use a low temperature
growth process that realizes a non-equilibrium growth. In the
present case, the GaInAs layer 15a (quantum well active layer) is
grown at 550.degree. C. It should be noted that the MOCVD process
used herein is characterized by high degree of supersaturation and
is suited for the crystal growth of highly strained active layer.
Further, the MOCVD process does not require high vacuum environment
as in the case of MBE process. Further, the MOCVD process is
suitable for mass production, as the process control is achieved by
merely controlling the supply rate or supply duration of the source
gases.
[0316] In the illustrated laser diode, there is formed a high
resistance region 15F outside the current path by means of ion
implantation of protons (H+) as a current confinement
structure.
[0317] In the constitution of FIG. 1, it should be noted that there
is formed a p-side electrode 20 on the p-type contact layer formed
in turn on the uppermost part of upper part reflector 18 and
constituting a part thereof, except for an optical exit part 20A.
Also there is formed an n-side electrode 21 at the rear surface of
the substrate.
[0318] In the present embodiment, it should be noted that the
active region including the upper part and lower part spacer layers
14 and 16 in addition to the multiple quantum well active layer 15
and forming the resonator between the upper and lower reflectors 12
and 18, in which the career recombination is caused upon injection
of carriers, is formed of a material not containing Al (the
proportion to the group III element(s) is 1% or more). Furthermore
the low refractive index layers constituting the lower part
reflector 12 and also the upper part reflector 18 closest to the
active layer is formed of the non-optical recombination elimination
layers 13 and 17 each having a composition of GaxIn1-xPyAs1-y
(0<x.ltoreq.1,0<y.ltoreq.1), in which the compositional
parameters x and y are chosen appropriately. More specifically,
GaInP or GaInPAs or GaPAs is used for the non-optical recombination
elimination layer. The material constituting the non-optical
recombination elimination layer thus has a composition of
GaxIn1-xPyAs1-y (0<x.ltoreq.1,0<y.ltoreq.1), wherein the
non-optical recombination layer may be added further with a trace
amount of element(s) other than Al.
[0319] In such a structure, the careers are confined between the
low refractive index layer of the upper part reflector and the low
refractive index layer of the lower part reflector that are located
nearest to the active layer and formed of a wide gap material. In
such a structure, there occurs non-optical recombination of
carriers even when the active region is formed of the layer not
containing Al (the proportion of Al with regard to group III
elements is 1% or less) upon injection of the careers, as there
occurs non-optical recombination of carriers at the interface
between the active layer and the low refractive index layer of the
upper or lower reflector, as long as the foregoing low refractive
index layer (wide gap layer) contains Al. When the non-optical
recombination of carriers takes place, the efficiency of optical
emission is deteriorated. Therefore it is desirable to form not
only the active region but also the low refractive index layers
adjoining thereto by a material not containing Al.
[0320] Further, the non-optical recombination elimination layer
having the primary composition of GaxIn1-xPyAs1-y
(0<x.ltoreq.1,0<y.ltoreq.1) has a lattice constant smaller
than the lattice constant of the GaAs substrate. Thus, the layer
accumulates a tensile strain.
[0321] In an epitaxial growth process, the growth is made while
reflecting the information of the foundation layer on which the
growth is made. Thus, when there are defects on surface of the
foundation layer, the same defects crawl up to the growth layer. On
the other hand, it is known that such a crawling up of the defects
can be suppressed by interposing a strained layer on the path of
the defects.
[0322] In the case an active layer accumulates therein a strain,
there sometimes occurs a problem in that growth of the active layer
with desired thickness is difficult because of the reduced critical
thickness. Particularly, there arises a problem in that the film
growth does not take place at all even if a low temperature growth
process or non-equilibrium growth process is employed, due to the
existence of defects. This problem occurs particularly when the
active layer accumulates a compressive strain of 2% or more, or
when growing the active layer beyond the critical thickness
thereof.
[0323] On the other hand, when there exists a strained layer
adjacent to the active layer, such a crawling up of the defects is
intercepted and the efficiency of optical emission is improved.
Further, it becomes possible to grow the active layer even in the
case the active layer accumulates a compressive strain of 2% or
store. Further, it is possible to grow the strained layer beyond
the critical thickness.
[0324] It should be noted that the GaxIn1-xPyAs1-y
(0<x.ltoreq.1,0<y.ltoreq.1) layers 13 and 17 adjoin the
active region and function so as to confine the careers into the
active region. On the other hand, the GaxIn1-xPyAs1-y
(0<x.ltoreq.1,0<y.ltoreq.1) layers 13 and 17 also have the
feature in that the bandgap energy can be increased by decreasing
the lattice constant. In the case of GaxIn1-xP (y=1), for example,
there occurs an increase of the lattice constant when the
compositional parameter x has increased and the film composition
has approached the composition GaP. Associated therewith, there
occurs an increase of bandgap. It should be noted that the bandgap
energy Eg is given as Eg (.left
brkt-top.)=1.351+0.643x+0.786x.sup.2 in the case of direct
transition and Eg (X)=2.24+0.02x in the case of indirect
transition. Therefore, the hetero barrier height between the active
region and the GaxIn1-xPyAs1-y (0<x.ltoreq.1,0<y.ltoreq.1)
layer 13 or 17 is increased and the degree of career confinement is
improved. Thereby, the threshold current is reduced and the
temperature characteristics are improved.
[0325] Furthermore, it should be noted that the non-optical
recombination elimination layer 13 or 17 having the composition
GaxIn1-xPyAs1-y (0<x.ltoreq.1,0<y.ltoreq.1) has a lattice
constant larger than the lattice constant of the GaAs substrate.
Also the lattice constant of the active layer is larger than the
lattice constant of the GaxIn1-xPyAs1-y
(0<x.ltoreq.1,0<y.ltoreq.1) layer 13 or 17. Thus, these
layers accumulate compressional strain therein. As the direction of
the strain accumulated in the GaxIn1-xPyAs1-y
(0<x.ltoreq.1,0<y.ltoreq.1) layer is the same as the
direction of the strain accumulated in the active layer, the
substantial amount of the compressional strain that the active
layer senses is reduced. Larger the strain, the influence of
external factor is also large. Thus, the construction of the
present invention is especially effective in the case that the
active layer accumulates a large compressional strain of 2% or
more, or the critical film thickness has been exceeded.
[0326] It is preferable to form a surface-emission laser diode of
the 1.3 .mu.m band on a GaAs substrate. Further, there are many
cases in which a semiconductor multilayer reflector is used for the
resonator. In such a case, it is necessary to grow 50 to 80
semiconductor layers with a total thickness of 5-8 .mu.m. (In the
case of an edge-emission type laser diode, on the other hand, the
total thickness before growth of the active layer is about 2 .mu.m,
and it is sufficient to grow just about three layers.)
[0327] Thus, even if a high quality GaAs substrate is used,
increase of defect density of the surface on which the growth of
the active layer is made, in the state immediately before the
growth of the active layer growth is inevitable. Once formed, the
defects thus formed crawl up in the direction of the crystal
growth. Further, there may be additional defect formation at the
hetero surface. On the other hand, measures such as provision of
strain layer before the growth of the active layer or reduction of
the substantial strain which the active layer senses, are effective
for reducing the influence of the defects on the surface on which
the growth of the active layer is made in the state immediately
before the growth of the active layer.
[0328] In this embodiment, Al is expelled from the active layer and
further from the interface region between the active region and the
reflector. Thus, the problem of non-optical recombination of
carriers originating from the crystal defects and caused by Al at
the time of career injection is successfully removed.
[0329] As noted before, while it is preferable to provide the
non-optical recombination elimination layer not containing Al at
the interface of both of the reflectors 12 and 18, beneficial
effect can be obtained also in the case the non-optical
recombination elimination layer is provided to only one of the
reflectors. In the illustrated example, both of the upper and lower
reflectors 12 and 18 are formed of the semiconductor Bragg
reflector. However, it is possible to form one of the reflectors by
the semiconductor Bragg reflector and form the other reflector by a
multilayer dielectric reflector.
[0330] In the foregoing example, only the low refractive index
layer closet to the active layer forms the GaxIn1-xPyAs1-y
(0<x.ltoreq.1,0<y.ltoreq.1) non-optical recombination
elimination layer 13 or 17 in any of the reflectors 12 and 18.
However, it is possible to form the non-optical recombination
elimination layer 13 or 17 by a plurality of GaxIn1-xPyAs1-y
(0<x.ltoreq.1,0<y.ltoreq.1) layers.
[0331] In this embodiment, the above concept is applied
particularly to the lower reflector 12 located between the GaAs
substrate and the active layer, and the problem of crawling up of
crystal defects originating from Al at the time of the growth of
the active layer and the adversary influence thereof, are
suppressed successfully. As a result, the active layer can be grown
with high quality, and a highly reliable surface-emission laser
diode oscillating with high efficiency and sufficient for practical
use is obtained. In the present embodiment, the Al-free,
GaxIn1-xPyAs1-y (0<x.ltoreq.1,0<y.ltoreq.1) layer is used
only for the low refractive index layer located closest to the
active region in the semiconductor Bragg reflector. Therefore, it
is possible to achieve the foregoing effect without increasing the
number of stacks of the reflector.
[0332] The surface-emission laser diode thus produced oscillated
successfully at the wavelength of about 1.2 .mu.m. While the
wavelength of GaInAs formed on a GaAs substrate increases with
increase of In content therein, such an increase of the In content
is accompanied with an increase of strain. Thus, it has been
thought that the wavelength of 1.1 .mu.m would be the limit of the
increase of laser oscillation wavelength in the laser diode that
uses GaInAs. See IEEE Photonics. Technol. Lett. Vol. 9 (1997) pp.
1319-1321.
[0333] In the present invention, the inventor could successfully
achieve the laser oscillation at 1.2 .mu.m by using highly
non-equilibrium growth processes conducted at low temperature of
600.degree. C. or less. By using such a process, it became possible
to achieve a coherent growth of the highly strained GaInAs quantum
well active layer with large thickness not possible before.
Meanwhile, this wavelength is transparent with respect to a Si
semiconductor substrate. Thus, the laser diode of the present
invention can be used to construct a circuit chip that uses optical
transmission through the Si substrate, by integrating an electronic
device and an optical device commonly on a Si substrate.
[0334] From the explanation noted above, it was discovered that a
long-wavelength surface emission laser diode can be constructed
successfully on a GaAs substrate by using a highly strained GaInAs
layer containing a large amount of In and hence a large amount of
compressive strain for the active layer.
[0335] As noted before, such a surface-emission laser diode can be
formed by using an MOCVD process. However, it is also possible to
use an MBE process or other growth process for this purpose. In the
embodiment described heretofore, a triple quantum well structure
(TQW) was used for the active layer. However, it is also possible
to use a structure having different number of quantum well layers
(SQW, MQW).
[0336] Further, it should be noted that the laser diode may also
have a different structure. Further, the resonator length is not
limited to X but may have a length of integer multiple of .lamda./2
or preferably .lamda..
[0337] In the embodiment described heretofore, the laser diode was
constructed on the GaAs substrate. However, it is also possible to
use an InP substrate. Further, the period of repetition of in the
reflectors may be changed.
[0338] In the embodiment described heretofore, the active layer of
GaxIn1-xAs (GaInAs active layer) containing Ga, In and As as the
major element is used in this example. On the other hand, it is
possible to add N to the active layer of the laser diode. In this
case, the active layer contains Ga, In, N and As as major elements
(GaInNAs active layer), and the laser diode can oscillate at
further longer wavelengths.
[0339] By changing the composition of the GaInNAs active layer, it
is possible to achieve laser oscillation at any of the 1.3 .mu.m
band and 1.55 .mu.m band. By choosing the composition of the active
layer appropriately, it is also possible to realize a
surface-emission laser diode laser oscillating at further longer
wavelengths such as 1.7 .mu.m.
[0340] It is also construct a surface-emission laser diode operable
at the wavelength of 1.3 .mu.m band on a GaAs substrate by using
GaAsSb for the active layer.
[0341] Conventionally, there has been no material suitable for
realizing a laser diode operable at the wavelength of 1.1-1.7
.mu.m. By using a highly strained layer of GaInAs, GaInNAs or
GaAsSb for the active layer, and by using a non-optical
recombination elimination layer, the present invention successfully
realized a highly efficient surface-emission laser diode operable
in the long wavelength region of a/the 1.1-1.7 .mu.m band.
Second Embodiment
[0342] Next, the constitution of another long-wavelength
surface-emission laser diode applicable to an optical
transmission/reception system of this invention as a light-emitting
device that is will be described by using FIG. 27. In the drawing,
those parts corresponding to the parts described previously are
designated by the same reference numerals and the description
thereof will be omitted.
[0343] FIG. 27 is referred to.
[0344] Similarly to the previous embodiment of FIG. 1, this
embodiment also uses an n-type GaAs substrate 11 having a surface
orientation of (100).
[0345] On the GaAs substrate 11, it can be seen that the n-type
semiconductor Bragg reflector (Al.sub.0.9Ga.sub.0.1As/GaAs lower
part reflectors) 12 is formed by alternately depositing the n-type
AlGaAs having a composition of AlxGa1-xAs (x=0.9) and an n-type
GaAs having a composition of AlxGa1-xAs (x-0) alternately for 35
periods, each with a thickness of 1/4 times the oscillation
wavelength .lamda. in each medium (thickness of .lamda./4).
Further, the n-type InGaP layer 13 is formed on the lower Bragg
reflector 12 with a composition of GaxIn1-xPyAs1-y (x=0.5, y=1) and
a thickness of .lamda./4. Thereby, the n-type GaxIn1-xPyAs1-y
(x=0.5, y=1) layer 13 forms one of the low refractive index layers
constituting a part of the lower part reflector 12.
[0346] On the InGaP layer 13, the multiple quantum well active
layer 15 is formed by stacking the undoped lower GaAs spacer layer
14 and further the GaxIn1-xNyAs1-y quantum well active layer 15a
three times on the spacer layer 14, with the GaAs barrier layer 15b
of 15 nm thickness interposed therebetween to form a triple quantum
wells (TQW) structure. Further, the undoped upper GaAs spacer layer
16 is formed on the multiple quantum well active layer 15 thus
formed, wherein the lower spacer layer 14, the multiple quantum
well active layer 15 and the upper GaAs spacer layer 15 constitute
together a resonator 15R having a thickness of one wavelength
(.lamda.) of the oscillation wavelength in the medium. The
resonator 15R constitutes the active region of the surface-emission
laser diode.
[0347] Further, the p-type semiconductor Bragg reflector (the upper
part reflector) 18 is formed on the resonator 15R.
[0348] It should be noted that the upper part reflector 18 includes
a low refractive index layer having a thickness of 3.lamda./4
formed of an AlAs layer 18.sub.1, used for a selective oxidizing
layer, wherein the foregoing low refractive index layer is
sandwiched by the GaInP layer 17 and an AlGaAs layer. The GaInP
layer 17 is doped with C has a composition represented as
GaxIn1-xPyAs1-y (x=0.5, y=1), wherein the GaInP layer 17 is formed
to have a thickness of .lamda./4-15 nm. On the other hand, the AlAs
layer constituting the selective oxidizing layer is doped with C
and has a composition represented as AlzGa1-zAs (z=1) and a
thickness of 30 nm. Further, the foregoing AlGaAs layer is a
C-doped AlGaAs layer having a composition represented by AlxGa1-xAs
(x=0.9) and a thickness of 2.lamda./4-15 nm.
[0349] On the foregoing low refractive index layer, a GaAs layer
having a thickness of .lamda./4 is formed for one period, and a
p-type AlGaAs layer, doped with C and having a composition of
expressed with AlxGa1-xAs (x=0.9) and a p-type GaAs layer doped
with C and having a composition represented by AlxGa1-xAs (x=0),
are stacked 22 periods each with a thickness of 1/4 times the laser
oscillation wavelength in each medium. Thereby a periodic stack
structure constituting the essential part of the upper part
reflector 18 is formed.
[0350] In this embodiment, too, the heterospike buffer layer 18c of
intermediate refractive index is provided in the semiconductor
Bragg reflector 18 by an AlzGa1-zAs (0.ltoreq.y<z<x.ltoreq.1)
layer between the low refractive index layer 18a and the high
refractive index layer 18b as already explained with reference to
FIG. 2. In FIG. 27, illustration of the heterospike buffer layer
18c will be omitted for the sake of simplicity.
[0351] In this embodiment, the p-type GaAs layer having the
composition of AlxGa1-xAs (x=0) and constituting the uppermost part
of the semiconductor Bragg reflector 18 functions as a contact
layer (p-contact layer) that secures electrical contact to the
electrode.
[0352] In the surface-emission laser diode of this embodiment, the
in content x of the quantum well active layer 15a is set to 37% and
the N (nitrogen) content is set to 0.5%. The thickness of the
quantum well active layer 15a is set to 7 nm.
[0353] In this embodiment, the growth of the semiconductor layers
constituting the said surface-emission laser diode was conducting
by using an MOCVD process. More specifically, TMA (trimethyl
aluminum), TMG (trimethyl gallium), TMI (trimethyl indium), AsH3
(arsine) and PH3 (phosphine) are used respectively as the source
materials of Al, Ga, In, As and P. Further, DMHy (dimethyl
hydrazine) was used as the source material of nitrogen. DMHy
decomposes at a low temperature and is suited for a low temperature
growth process conducted at 600.degree. C. or less. Particularly,
it is suited to grow a quantum well layer of large strain, which
needs a low temperature growth process. In the MOCVD process,
H.sub.2 is used for the carrier gas. Further, the growth of the
GaInNAs layer (quantum well active layer) was conducted at
540.degree. C.
[0354] It should be noted that MOCVD process is characterized by
large degree of supersaturation and is suited for the crystal
growth of a material that contains N and simultaneously other group
V elements. Further, the MOCVD process does not require the high
vacuum environment, contrary to the MBE process. Further, it is
suitable for mass production, as it is only sufficient to control
the supply rate and supply time of the source gases.
[0355] In this embodiment, a predetermined part of the stacked
structure thus formed is etched until it reaches the p-type
GaxIn1-xPyAs1-y (x=0.5, y-1) layer 17. Thereby, there is formed a
mesa structure that exposes the p-AlzGa1-zAs (z=1) selective
oxidizing layer 181 at the sidewall thereof. Furthermore, said
AlzGa1-zAs (z=1) layer 18.sub.1 thus exposed is oxidized from the
mesa sidewall by water vapor, and there is formed a current
confinement layer 182 having a composition represented as
AlxOy.
[0356] Finally, the part removed with the mesa etching process
previously is filled by polyimide to form a planarized structure,
and the polyimide film covering the upper part reflector is
removed. With this, there is formed a polyimide region 18.sub.3.
Furthermore the p-side electrode 20 is formed on-the p-type contact
layer except for the optical exit par and an n-side electrode 21 is
formed on the rear side of the GaAs substrate 11.
[0357] In this embodiment, the GaxIn1-xPyAs1-y
(0<x.ltoreq.1,0<y.ltoreq.1) layer 17 is inserted below the
selective oxidizing layer 18.sub.1 as a part of the upper part
reflector 18. In the case a wet etching process by using a sulfuric
acid etchant is employed in the formation of the mesa structure,
the etching stops spontaneously at the GaxIn1-xPyAs1-y
(0<x.ltoreq.1,0<y.ltoreq.1) layer 17, as a material of the
GaInPAs system functions as an etching stopper layer to the etching
process of a material of the AlGaAs system. Thus, by using a wet
etching process by a sulfuric etchant for the formation of the mesa
structure, it is possible to control the height of the mesa
structure rigorously.
[0358] Because of this, the homogeneity and reproducibility are
improved substantially for the surface-emission laser diodes that
are formed simultaneously on a substrate and the production cost is
reduced. This feature is particularly advantageous when producing a
laser diode array in which a number of surface-emission laser
diodes are integrated in one-dimensional or two-dimensional
array.
[0359] In the embodiment of FIG. 27, it is noted that the
GaxIn1-xPyAs1-y (0<x.ltoreq.1,0<y.ltoreq.1) layer 17 that
acts also as an etching stopper layer is provided on the side of
upper part reflector 18. Further, a similar GaInP layer 13 is
provided on the lower part reflector 12.
[0360] In this embodiment, too, the active region 15 sandwiched
between the upper and lower reflectors 12 and 18 and cause
recombination upon injection of carriers is formed of a material
free from Al. Further, the low refractive index layer of the lower
and upper reflectors 12 and 18 closest to the active layer 15 is
formed of the non-optical recombination elimination layer 13 or 17
having a composition represented by GaxIn1-xPyAs1-y
(0<x.ltoreq.1, 0<y.ltoreq.1). Thus, Al is not contained at
the interface between the active region 15 and the reflector 12 or
18, and the problem of non-optical recombination caused by the
crystal defects, which in turn are caused by Al, is effectively
eliminated.
[0361] While such a construction of using an Al-free material at
the interface between the reflector and the active region is
preferably provided to both of the upper and lower reflectors 12
and 18, the advantageous effect of the present invention is
obtained also in the case such a construction is used on only one
of the upper and lower reflectors 18 and 12. Further, it is also
possible to use a semiconductor Bragg reflector in only one of the
reflectors and form the other reflector by a dielectric
reflector.
[0362] As the lower part reflector 12 provided between the GaAs
substrate 11 and the active layer 15 is constructed similarly to
the case of FIG. 1 in the present embodiment, the adversary effects
of crawling up of the crystal defects, originating from Al, into
the active layer is effectively suppressed. Thereby, the active
layer 15 can be formed with high quality.
[0363] It should be noted that such a non-optical recombination
elimination layer 13 or 17 constitutes a part of the semiconductor
Bragg reflector 12 or 18 in any of the constitution of FIG. 1 or
27, and thus, the thickness thereof is set to 1/4 the oscillation
wavelength X as measured in the medium .lamda./4). It is also
possible to provide such a non-optical recombination elimination
layer in plural numbers.
[0364] In the example above, the non-optical recombination
elimination layer 13 or 17 is provided to a part of the
semiconductor Bragg reflector 12 or 18. On the other hand, it is
also possible to provide such a non-optical recombination
elimination layer inside the resonator 15. In the case the
resonator 15 is formed of the active layer 15 consisting of the
GaInNAs quantum well layer 15a and the GaAs barrier layer 15b, for
example, it is possible to use the GaAs layer as the first barrier
layer and the non-optical recombination elimination layer of
GaInPAs, GaAsP or GaInP as the second barrier layer. Thereby, it is
possible to set the thickness of the resonator to one wavelength.
In this case, because of the large bandgap of the non-optical
recombination elimination layer as compared with the first GaAs
barrier layer, the active region in which carrier injection takes
place is substantially limited up to the region of the GaAs barrier
layer.
[0365] In the case that the process of removing the residual Al
source material or Al reactant, Al compound or Al during the
fabrication process of the laser diode, it is possible to provide
such a process during the step of forming the non-optical
recombination elimination layer. Alternatively, the removal of
residual Al may be conducted in a process of growing a GaAs layer
interposed between the process of forming the Al-containing layer
and the process of forming the non-optical recombination
elimination layer.
[0366] According to the present embodiment, it becomes possible to
form a highly efficient, reliable surface-emission laser diode.
Thereby, it is possible to achieve the foregoing advantageous
effect of the present invention without increasing the number of
stacks in the reflector, as the non-optical recombination
elimination layer formed of the Al-free GaxIn1-xPyAs1-y
(0<x.ltoreq.1,0<y.ltoreq.1) layer is used in only the low
refractive index layer located closest to the active region in the
semiconductor Bragg reflector 12 or 18.
[0367] Further, it should be noted that filling of polyimide is
easily made, and thus, it is possible to provide interconnection
wiring forming the p-side electrode, without risking the
disconnection of the wiring at the stepped part. Thereby, the
reliability of the device is improved further.
[0368] It was confirmed that the surface-emission laser diode thus
produced oscillates at the laser oscillation wavelength of about
1.3 .mu.m.
[0369] In the present embodiment, it became possible to construct a
surface-emission laser diode of long wavelength band on the GaAs
substrate 11 by using a GaInNAs layer containing Ga, In, N and As
for the major element. By forming a current confinement structure
by applying a selective oxidation process to the AlAs layer
18.sub.1 to form the oxide current confinement layer 18.sub.2, the
threshold current of laser oscillation is reduced effectively. In
such a current confinement structure, formed of the AlAs layer
18.sub.1 in which the selective oxidation layer 18.sub.2 is formed,
it is possible to provide the current confinement structure close
to the active layer, and lateral diffusion of the injected electric
current is suppressed effectively. As a result of use of such a
current confinement structure, it becomes possible to confine the
careers effectively into a minute region not exposed to the
atmosphere. The foregoing current confinement structure has another
advantageous feature in that the Al oxide constituting the layer
182 has a low refractive index and the efficiency of carrier
confinement is increased further. Thereby, the efficiency of laser
oscillation is improved furthermore and the threshold electric
current is reduced. Further, the current confinement structure has
an advantageous feature in that formation thereof is made easily
and the production cost is reduced.
[0370] From the explanation noted above, the laser diode of FIG. 27
oscillates efficiently at the wavelength of 1.3 .mu.m similarly to
the case of FIG. 1. The laser diode of FIG. 27 also has
advantageous feature of small power consumption and low production
cost.
[0371] The surface-emission laser diode of FIG. 27 can be formed
also by an MOCVD process, similarly to the case of FIG. 1. However,
it is also possible to use an MBE process or other growth process.
Further, it is possible to use nitrogen or a nitrogen compound such
as NH3 in an activated state.
[0372] Furthermore, it is possible to replace the triple quantum
well structure (TQW) in the active layer 15 with another structure
including different number of quantum wells such as an SQW
structure, DQW structure or an MQW structure. Further, it is
possible to use a laser diode of different structure.
[0373] By adjusting the composition of the GaInNAs active layer 15a
in the surface-emission laser diode of FIG. 1, it becomes possible
to realize a surface-emission laser diode of the 1.55 .mu.m band
and further the 1.7 .mu.m band. In the present invention, the
GaInNAs active layer may contain other III-V elements such as TI,
Sb, P, and the like. Further, it is also possible to construct a
surface-emission laser diode of 1.3 .mu.m band on a GaAs substrate
by using GaAsSb for the active layer.
[0374] In the description heretofore, it was assumed that the
active layer contains Ga, In and Ss as the major elements (GaInAs
active layer) or Ga, In, N and As as the active layer (GaInNAs
active layer). Further, the active layer may be formed of any of
GaNAs, GaPN, GaNPAs, GaInNP, GaNAsSb, and GaInNAsSb. The present
invention is particularly effective also in these cases in which
the active layer contains N.
[0375] FIG. 28 shows the room temperature photoluminescence
spectrum obtained for the active layer that includes a GaInNAs/GaAs
double quantum well structure formed by a GaInNAs quantum well
layer and a GaAs barrier produced by an MOCVD process. Further,
FIG. 29 shows the structure of the sample used for the
measurement.
[0376] FIG. 29 is referred to.
[0377] It can be seen that there are formed a lower part cladding
layer 202, an intermediate layer 203A, an active layer 204
containing therein nitrogen, another intermediate layer 203B, and
an upper part cladding layer 205 consecutively on the GaAs
substrate 201.
[0378] Next, FIG. 28 is referred to.
[0379] The curve A shows the result with regard to the sample that
uses an AlGaAs layer as the cladding layer 202 and the double
quantum well structure is formed while interposing a GaAs layer
similar to the intermediate layer 203A or 203B between the quantum
wells. The curve B, on the other hand, shows the measurement result
with regard to the sample that uses a GaInP layer as the cladding
layer 202 and the double quantum well structure is formed
continuously while interposing a GaAs layer similar to the
intermediate layer 203A or 203B.
[0380] As can be seen in FIG. 28, the photoluminescence intensity
of the sample A has fallen off by one-half or more with respect to
the photoluminescence intensity of the sample B. This means that
there arises a problem of decrease of photoemission efficiency of
the active layer when the active layer containing nitrogen therein
such as GaInNAs is grown continuously on a semiconductor layer that
contains Al as constituent element such as AlGaAs while using a
single MOCVD apparatus. Because of this, the threshold current
density of the GaInNAs laser diode formed on an AlGaAs cladding
layer is increased by twice or more as compared with the case in
which the same laser diode is formed on an GaInP cladding
layer.
[0381] The inventor of this invention conducted through
investigation of this problem and obtained a knowledge explained
below.
[0382] FIG. 30 shows the depth profile of nitrogen and oxygen in
the device for the case the cladding layers 202 and 205 are formed
by an AlGaAs layer, the intermediate layers 203A and 203B are
formed of GaAs, and the active layer is formed of the GaInNAs/GaAs
double quantum well structure and the device is formed by using a
single MOCVD apparatus. It should be noted that the measurement of
FIG. 30 was conducted under the measurement condition summarized in
Table 1 by secondary ion mass spectrometry. TABLE-US-00004 TABLE 3
first ion specie Cs+ first acceleration voltage 3.0 kV sputter rate
0.5 nm/s measurement area 160 .times. 256 .mu.m.sup.2 degree of
vacuum <3E-7 Pa polarity of measured ion --
[0383] FIG. 30 is referred to.
[0384] It can be seen that there appear two nitrogen peaks in the
active layer 204 in correspondence to the GaInNAs/GaAs double
quantum well structure. Further, it is also noted that an oxygen
peak is detected in the active layer 204. On the other hand, the
oxygen concentration in the intermediate layers 203A and 203, which
are free from N and Al, is lower in terms of one order as compared
with the oxygen concentration in the active layer 204.
[0385] Further, the depth profile of oxygen measured about the
sample having a constitution in which the active layer 204
including therein the GaInNAs/GaAs double quantum well structure is
formed on the cladding layer of GaInP together with the GaAs
intermediate layers 203A and 203B, has revealed that the oxygen
concentration in the active layer 204 is in the background
level.
[0386] Thus, it was confirmed experimentally that the problem of
oxygen incorporation into the active layer 204 occurs singularly in
the case in which the laser diode has the nitrogen-containing
active layer 204 in the state that the active layer 204 is
separated from the substrate by the semiconductor layer 202
containing Al and the growth of the layers constituting the device
is conducted by a single MOCVD apparatus while using a nitrogen
compound source material and an organic Al source material. The
oxygen atoms thus incorporated into the active layer form a
non-optical recombination state therein and cause a decrease of
luminous efficacy of the active layer.
[0387] Thus, from the result of FIG. 30, it is concluded that
oxygen incorporated into the active layer is the cause the problem
of decease of optical emission efficiency in the device in which
the Al-containing layer is provided between the substrate and the
N-containing active layer. It is conceivable that the origin of
this oxygen contamination would be the oxygen-containing material
that has captured residual oxygen in the deposition apparatus or
the impurity contained in the nitrogen compound source
material.
[0388] Next, the mechanism of oxygen incorporation into the active
layer 204 will be examined.
[0389] FIG. 31 shows the depth profile of Al obtained for the same
sample of FIG. 30. The measurement was conducted by the secondary
ion mass spectroscopy similarly to the case of FIG. 30 under the
measurement condition shown in Table 2. TABLE-US-00005 TABLE 4
first ion specie O2+ first acceleration voltage 5.5 kV sputter rate
0.3 nm/s measurement area 60 .mu.mO degree of vacuum <3E-7 OPa
polarity of measured ion +
[0390] FIG. 31 is referred to.
[0391] It can be seen that Al is detected in the active layer 204,
which was formed without introducing an Al source material.
Further, it is noted that the Al concentration in the GaAs
intermediate layer 203A or 203B formed adjacent to the
Al-containing cladding layer 202 or 205 is lower than the Al
concentration in the active layer 204 by the order of one. This
indicates that Al in the containing cladding layer 202 or 205 and
has substituted active layer 204 has caused diffusion from the
Al-containing cladding layer 202 or 205 and has substituted Ga in
the active layer 204.
[0392] On the other hand, Al was not detected in the active layer
in the case that the N-containing active layer was grown on a
semiconductor layer free from Al such as GaInP.
[0393] From this, it is concluded that Al detected in the active
layer 204 originates from residual Al such as residual Al source
material or residual Al reactant or residual Al compound or
residual Al remaining in the growth chamber or gas supply line and
has been incorporated into the active layer as a result of coupling
with the nitrogen source compound or the impurity such as water
contained in the nitrogen source compound. Further, it is concluded
that the Al incorporation into the N-containing active layer is
inevitable when a semiconductor light emitting device having an
Al-containing layer between the substrate and the N-containing
active layer is grown continuously by a single epitaxial
apparatus.
[0394] By comparing FIG. 31 with FIG. 30, it is noted further that
the two oxygen peaks in FIG. 30 corresponding to the double quantum
well do not coincide with the two nitrogen peaks but coincide with
the peak profile of Al shown in FIG. 31. This results indicates
that oxygen in the GaInNAs quantum well layer has been incorporated
into the active layer not with the nitrogen source but with Al in
the form coupled with residual Al that has been incorporated into
the quantum well layer.
[0395] More specifically, the residual Al source material or
residual Al reactant or residual Al compound or residual Al in the
MOCVD chamber is thought to cause a coupling with water contained
in the nitrogen source compound or oxygen-containing material
remaining in the gas line or reaction chamber and is incorporated
into the active layer in the form coupled with oxygen or water.
[0396] Thus, it was clarified by the inventor of the present
invention that, in the surface-emission laser diode of a
conventional GaAs system, there occurs oxygen incorporation into
the active layer and oxygen thus incorporated has caused the
problem of decrease of efficiency of optical emission.
[0397] The foregoing discovery provides a clue to the improvement
of efficacy of the surface-emission laser diode of the GaAs system
in that it is necessary to remove the impurity at least from the
part of the deposition apparatus used for the production of the
laser diode and may have a chance to make a contact with the
nitrogen compound source material in the deposition chamber or the
impurity contained in the nitrogen compound source material.
[0398] Thus, by providing a step of removing the residual Al after
the growth of the Al-containing layer but before the growth of the
N-containing active layer, it becomes possible to reduce the
concentration of Al and oxygen impurity incorporated into the
active layer. By doing so, the concentration of residual Al source
material or Al reactant or Al compound or Al concentration is
already reduced when the nitrogen compound source material is
supplied for the growth of the N-containing active layer, and the
reaction between the residual Al and the nitrogen source compound
or the impurity contained in the nitrogen source compound is
successfully suppressed.
[0399] It is further advantageous to remove the residual Al before
the end of the growth of the non-optical recombination layer. By
doing so, it is possible to suppress the non-optical recombination
of carriers in the active layer at the time carriers are injected
into the active layer by way of current injection.
[0400] For example, it became possible to carry out a continuous
oscillation of the surface-emission laser diode of the GaAs system
at room temperature environment by reducing the Al concentration in
the N-containing active layer to 1.times.10.sup.19 cm.sup.-3 or
less. Further, by reducing the Al concentration in the N-containing
active layer to 2.times.10.sup.18 cm.sup.-3 or less, optical
emission characteristics substantially equal to the case when the
active layer is grown on an Al-free semiconductor layer is
obtained.
[0401] The removal of the residual Al source material or residual
Al reactant or residual Al compound or residual Al from the part of
the deposition chamber that may have a chance to make a contact
with the nitrogen compound source material or the impurity
contained in the nitrogen compound source material, can be
advantageously conducted by a purging process that uses a carrier
gas as the purge gas. Here, the time of the purging process is
defined as the time after the supply of the Al source material into
the deposition chamber is stopped in correspondence to completion
of the growth of the Al-containing layer but before the start of
supply of the nitrogen compound source material to the deposition
chamber for the start of growth of the N-containing semiconductor
layer. It is also possible to interrupt the growth of the
intermediate layer, which is free from any of Al and nitrogen, and
conduct a purging process by using the carrier gas as a purging
gas. In this case, the interruption of the growth process of the
intermediate layer may be achieved after the start of growth of the
Al-containing layer until the midway of the growth of the
non-optical recombination layer.
[0402] FIG. 32 shows an example of the semiconductor light-emitting
device formed by providing such a carrier purging process. In FIG.
32, those parts corresponding to the parts described previously are
designated by the same reference numerals and the description
thereof will be omitted.
[0403] Thus, it can be seen that the Al-containing first
semiconductor layer 202, the first lower part intermediate layer
601, the second lower part intermediate layer 602, the N-containing
active layer 204, the upper part intermediate layer 203, and the
second semiconductor layer 205 are stacked consecutively the on
substrate 201 in FIG. 32.
[0404] In forming the structure of FIG. 32, the crystal growth is
carried out by an epitaxial growth apparatus by using an
organic-metal Al source and an organic nitrogen source material. In
the case, the growth interruption process is provided before the
start of growth of the second lower part intermediate layer 602 but
after the growth of the first lower part intermediate layer 601.
During the growth interruption process, the part of the growth
chamber that may make a contact with the nitrogen compound source
material or the impurity in the nitrogen compound source material
is purged by a hydrogen gas used as the carrier gas, such that the
residual Al source material or residual Al reactant or residual Al
compound or residual Al is removed.
[0405] FIG. 33 shows the result of measurement of the depth profile
of Al concentration on a semiconductor light-emitting device in
which there has been provided a growth interruption between the
first lower part intermediate layer 601 and the second lower part
intermediate layer 602 and a purging process was conducted for 60
seconds.
[0406] FIG. 33 is referred to. It can be seen that the Al
concentration in the active layer 204 is reduced to
3.times.10.sup.17 cm.sup.-3 or less as a result of such a growth
interruption process and purging process. This value of Al
concentration is the same degree as the Al concentration in the
intermediate layers 601 and 602.
[0407] FIG. 34 shows the depth profile of nitrogen and oxygen for
the device of FIG. 32.
[0408] FIG. 34 is referred to. It can be seen that the oxygen
concentration in the active layer 204 is reduced to
1.times.10.sup.17 cm.sup.-3, which is a background level. The
oxygen peak appearing in the lower part intermediate layer 601 or
602 in FIG. 34 is interpreted as showing oxygen segregation to the
growth interruption interface as a result of interruption of
growth. Thus, it is preferable to conduct the growth interruption
period after the termination of the growth of the semiconductor
layer containing Al but before the growth termination of the
non-optical recombination elimination layer, in the case there is
provided a growth interruption and purging process. In the
non-optical recombination elimination layer, it is possible to
increase the bandgap energy as compared with the quantum well
active layer or barrier layer, and the adversary effect of the
non-optical recombination caused by the oxygen segregated to the
growth interruption interface is effectively suppressed when career
injection is made into the active layer. The constitution that uses
the non-optical recombination elimination layer provides a
particularly advantageous effect in the case of using an active
layer containing nitrogen.
[0409] In the semiconductor light-emitting device of FIG. 32, the
impurity concentration level of Al and oxygen in the
nitrogen-containing active layer 204 is reduced successfully by
interrupting the growth process between the first lower part
intermediate layer 601 and the second lower part intermediate layer
602 and by conducting a purging process for 60 minutes. In this
way, the optical efficacy of the active layer 204 was effectively
improved.
[0410] Further, it is also possible to carry out the Al removal
process while heating the susceptor during the purging process
conducted with a carrier gas. By doing so, the Al source material
or reaction product adsorbed on the susceptor or the region near
the susceptor is effectively decoupled and removed. In the case of
conducting such a heating of the substrate during the purging
process, it is necessary to supply the group V source gas such as
AsH3 or PH3 so as to avoid thermal decomposition of the uppermost
semiconductor layer.
[0411] It is also possible to transport the substrate to a chamber
different from the growth chamber when conducting the purging
process with a carrier gas in an MOCVD apparatus. In the case that
the substrate is transported to another chamber from the growth
chamber, it is not necessary to keep supplying the group V source
gas such as AsH.sub.3 or PH.sub.3 to the growth chamber. Thereby,
the efficiency of thermal decomposition of the Al reactant
deposited on the susceptor or the region surrounding the susceptor
is substantially facilitated and the efficiency of removal of Al is
improved.
[0412] Further, it is also possible to carry out the purging
process while continuing the growth of the intermediate layer. In
the constitution of FIG. 27, for example, the non-optical
recombination elimination layer 13 is provided between the
N-containing active layer 15 and the reflector 12 formed of the
AlGaAs material containing Al. Thereby, the distance between the
N-containing active layer 15 and the Al-containing layer is
increased, and it is possible to increase the duration of the
purging process in the case the purging process is conducted
simultaneously to the growth. In such a case, it is advantageous to
decrease the growth rate and increase the purging time.
[0413] Further, it is also possible to carry out the growth of the
Al-containing reflector 12 of AlGaAs material and the N-containing
active layer 15 in respective, separate apparatuses. In this case,
too, the impurity concentration level of Al or oxygen in the active
layer 15 can be reduced, by setting the regrowth interface below
the non-optical recombination elimination layer 13.
[0414] In the case that the surface emission laser diode is
produced with the crystal growth process that does not use an
organic-metal Al source or nitrogen compound source material such
as an MBE process, there is no report about the degradation of the
optical efficiency in the semiconductor light-emitting device in
which-N-containing active layer is provided on the Al-containing
semiconductor layer. In case of an MOCVD process, on the other
hand, degradation of optical efficacy of the GaInNAs active layer
formed on the Al-containing semiconductor layer is reported.
[0415] In Electron. Lett., 2000, 36 (21), pp. 1776-1777, for
example, there is a report that the photoluminescence intensity
deteriorates remarkably when a GaInNAs quantum well layer is grown
continuously on an AlGaAs cladding layer in the same MOCVD growth
chamber, even when an intermediate layer of GaAs is provided on the
AlGaAs cladding layer. In the foregoing report, the GaInNAs active
layer is grown in a MOCVD chamber different from one used for
growing the AlGaAs cladding layer so as to improve the
photoluminescence intensity.
[0416] Thus, it is believed that the aforementioned problem arises
more or less when growing a crystal layer by an organic-metal Al
source and a nitrogen compound source, as in the case of an MOCVD
process.
[0417] In the MBE process, the crystal growth is carried out under
a ultra low-pressure environment (high vacuum state). On the other
hand, an MOCVD process is conducted under a process pressure of
several ten Torr to the atmospheric pressure. Thus, the mean free
path of the gaseous molecules is overwhelmingly short in an MOCVD
process as compared with the MBE process. Because of this, it is
conceivable that the source gas molecules or the carrier gas
molecules make a contact with various parts of the gas line,
reaction chamber, and the like.
[0418] Thus, in the case of the growth process that uses a
relatively high pressure for the reaction chamber or gas line as in
the case of the MOCVD process, it becomes possible to eliminate the
incorporation of oxygen into the N-containing active layer, by
providing a removal step for removing the residual Al source
material or Al reactant or Al compound or Al from the location of
the reaction chamber, in which the reaction chamber may make a
contact with the nitrogen compound source material or the impurity
contained in the nitrogen compound source material. Such a removal
step should be conducted after the growth of the Al-containing
semiconductor layer but before the start of growth of the
N-containing active layer. More preferably, the removal process
should be carried out after the growth of the Al-containing
semiconductor layer but before the end of the growth of the
non-optical recombination elimination layer.
[0419] For instance, it is possible to evacuate the gas line or the
growth chamber with a vacuum evacuation process after the growth of
the Al-containing semiconductor layer but before the growth of the
N-containing active layer. In this case, it is preferable to carry
out the evacuation process in the state that the substrate is
heated.
[0420] Also, it is possible to remove the residual Al after the
growth of the Al-containing semiconductor layer but before the
growth of the N-containing active layer by supplying an etching
gas. For example, an organic compound gas may be used as the
etching gas that reacts with the Al-containing residue.
[0421] For example, it is possible to cause a reaction in the
Al-containing residue by supplying a DMHy gas, which is one of the
organic compound gases, at the time of the growth of the active
layer from a DMHy cylinder. Therefore it is possible to remove
these Al-containing residue by causing a-reaction between the
Al-containing residue remaining on the reaction chamber surface,
heating zone, or jigs used for supporting the substrate, and the
organic compound gas by supplying the organic compound gas from a
gas cylinder, after the growth of the Al-containing semiconductor
layer. By a method, too, it is possible to suppress the
incorporation of oxygen into the active layer. Particularly, by
using the same gas as the gas used for the nitrogen source it is
possible to avoid the problem of providing a special gas line. This
process can be carried out while interrupting the growth or growing
a dummy layer containing nitrogen such as GaNAs, GaInNAs, GaInNP
separately from the growth of the active layer. By doing conducting
the Al removal process simultaneously to the crystal growth
process, it is possible to reduce the time loss as compared with
the case of interrupting the growth process, and the throughput of
the semiconductor device production is improved.
[0422] By using GaInAs in the active layer of laser diode, it
became possible to realize a laser diode oscillating at the
wavelength of 1.2 .mu.m by growing a highly strained GaInAs quantum
well active layer by using a low temperature growth process of
600.degree. C. This wavelength value is substantially longer than
the wavelength of 1.1 .mu.m, which has been conceived as the upper
limit of laser oscillation wavelength. Conventionally, there was no
material, suitable for the laser diode of 1.1-1.7 .mu.m. On the
other hand, it became possible to realize a highly efficient
surface-emission laser diode operable in the wavelength band of
1.1-1.7 .mu.m, by using a highly strained active layer of GaInAs,
GaInNAs or GaAsSb together with the use of the non-optical
recombination elimination layer. With this, the possibility of
applications to the optical-fiber telecommunication systems was
opened.
Third Embodiment
Third Embodiment
[0423] FIG. 35 shows a laser diode chip 32 including the
long-wavelength surface-emission laser diode of FIG. 1 or FIG. 27
explained previously. The example that formed on an n-type GaAs
wafer 31 of a surface orientation (100) is shown.
[0424] FIG. 35 is referred to.
[0425] It can be seen that a number of laser diode chips 32 are
formed on the GaAs wafer, and each of the laser diode chips 32
carries thereon n laser diodes each having a construction explained
before. The number n and arrangement thereof are decided according
to the use of laser diode chip 32.
[0426] FIG. 36 shows the example of the optical
transmission/reception system using the long-wavelength
surface-emission laser diode having the laser oscillation
wavelength of 1.1-1.7 .mu.m band. Those parts explained previously
are designated by the same numerals and explanation is omitted.
[0427] FIG. 36 is referred to.
[0428] The surface-emission laser diode chip 32 is provided at the
A point as an optical source such that the emission part 32A of the
laser diode chip 32 makes an optical coupling with the optical
fiber 33. The optical signal emitted from the emission part 32A of
the laser diode chip 32 is injected into the optical fiber 33. The
optical signal is transferred in the direction shown with the thick
arrow in FIG. 36. The end point of the optical fiber 33 is located
at the B point. The photodiode or other suitable photodetection
device 34, which constitutes the photodetection unit is provided in
the B point, and the optical detection part 34A thereof is coupled
optically to and the optical fiber 33. Thus, the optical
transmission/reception system is composed.
[0429] In this example, the location B for the photodetection unit
and the location A for the optical source are tied with a straight
line by the optical fiber 33.
[0430] FIG. 37 shows the construction of FIG. 36 schematically.
[0431] FIG. 37 is referred to.
[0432] The black circles A, B in the figure show the location of
the optical source 32 and the photodetection unit 34 respectively.
A black thick line shows the optical fiber 33.
[0433] Conventionally, optical source 32 and photodetection unit 34
are connected optically with optical fiber 33 that provides the
transmission path and constitute an optical transmission/reception
system. The transmission path is formed extending from several tens
of meters to several tens of kilometers in the optical
transmission/reception system with the long-wavelength
surface-emission laser diode of laser oscillation wavelength
1.1-1.7 .mu.m band of this invention. Thereby, there may be
obstacles between the points A and B.
[0434] For example as for FIG. 38, obstacles 35A and 35B exist
between the points A and B. Thus, the point B for the location of
the photodetection unit and the point A for the location of the
optical source cannot be connected with the transmission path 33 of
a straight line. Thus, FIG. 38 shows the example of bending the
transmission path by a right angle.
[0435] When the optical fiber 33 is bent at the point where the
transmission path is bent by a right angle with the same angle as
in the case of FIG. 38, the optical fiber 33 is damaged and stops
functioning as a transmission path. It is possible to provide a
reflection part in the bending part of the transmission path so as
to deflect the optical beam. The use of the reflection part in such
a bending part of the transmission path increases the cost.
[0436] Thus, the present invention proposes an approach to bend the
optical transmission path without providing a reflection part and
without spoiling the functioning of the optical fiber 33 as a
transmission path.
[0437] FIG. 39 shows a third embodiment of this invention. Those
parts explained previously are designated by the same reference
numerals and explanation is omitted.
[0438] FIG. 39 is referred to. Obstacle 35C exists between the A
point and B point. The location B and the location A for the
photodetection unit and the optical source 32, respectively, can
not be connected by a transmission path of straight line. However,
as shown in this embodiment, the location B for the photodetection
unit 34 and the location A for the optical source 32 are connected
by bending optical fiber 33, which composes a transmission path
without causing a local angle. Thus, the optical fiber 33 composing
the transmission path is not damaged by the bending, and it is
possible to avoid the obstacle 35C. According to the present
invention it is possible to construct a fine optical
transmission/reception system even when obstacles must be
avoided.
[0439] FIG. 40 is another example according to this invention. In
this case obstacles 35D, 35E exist between the points A and B.
Location B and location A of photodetection unit 34 and optical
source 32, respectively, can not be connected with a straight line
transmission path.
[0440] However, by bending the optical fiber 33 of the transmission
path, without causing a local angle, as shown in this example of
this invention it is possible to connect the location B for the
photodetection unit and the location A for the optical source. As a
result, the transmission path is bent continuously in this example
of the present invention. It is not formed step-like and a local
angle is not required.
[0441] Furthermore the above explanation pertains to the optical
transmission/reception system using the long-wavelength
surface-emission laser diode with laser oscillation wavelength of
1.1-1.7 .mu.m band to suitably provide middle to long distance
telecommunication systems. According to this invention it is
possible to deploy this system inside of a building, for example,
even in the case of constructing an optical transmission/reception
system with a length of several centimeters to several meters. Even
in this case, the transmission direction of the optical beam can be
changed in all directions without providing reflection part, by
bending the optical fiber along the transmission path, without
causing a local angle therein. According to evaluations conducted
by the inventor of the present invention, there was not any case
wherein the function of a transmission path failed due to the
optical fiber 33 being bent, even in the case that the transmission
path was bent continuously as indicated in FIG. 40. There was no
damage to the transmission paths such as those shown in FIG. 39 and
FIG. 40, as long as the diameter of the curvature of the
transmission path does not exceed 20 cm. To put it differently, the
transmission path can be arranged to avoid obstacles such as shown
in FIG. 40, even if reflection part or the like, that changes the
direction is not provided, if the diameter of the transmission path
curvature does not exceed 20 cm. Thus, the present invention
provides a fine optical transmission/reception system that
functions with high reliability at a low cost.
Fourth Embodiment
[0442] Next the fourth embodiment of this invention is
explained.
[0443] FIG. 41 shows another example of an optical
transmission/reception system using the long-wavelength
surface-emission laser diode with laser oscillation wavelength of
1.1-1.7 .mu.m band. Those parts explained previously are designated
by the same reference numerals and explanation is omitted.
[0444] FIG. 41 is referred to. In this example the laser diode
emission part 32A is coupled optically to an optical fiber F1. The
laser beam that is emitted from the laser diode emission part 32A
is transferred in the direction of the thick arrow in FIG. 41. The
laser beam that was transmitted through the optical fiber F1 is
then bent 90 degrees by a reflection part R. The laser beam then
enters optical fiber F2. Optical fiber F2 at the end point couples
optically to the light detector part 34A of the photodetection
device 34 such as a photodiode. Thus the optical
transmission/reception system is composed.
[0445] It should be noted that the foregoing description is merely
for the exemplary purpose of the optical transmission/reception
system of this invention. Thus, only one laser diode emission part
32A is shown in the present example. However, it also is possible
to construct, by using the advantageous feature of the
long-wavelength surface-emission laser diode of the present
invention, to construct a large capacity optical
transmission/reception system of the multiple laser array type, in
which a number of laser diodes 32A are formed on one laser diode
chip 32A and the optical transmission/reception system is
constructed by using a number of optical fibers F1 and
photodetection devices 34 coupled optically to the laser diodes
32A.
[0446] FIG. 42 shows another arrangement of the optical
transmission/reception system of FIG. 41. The optical
transmission/reception system of FIG. 41 may be a yard, whereas in
FIG. 42 it is arranged inside a wall of a building. Those parts
explained previously are designated by the same reference numeral
and explanation is omitted.
[0447] FIG. 42 is referred to.
[0448] A room 42 is defined with a wall 41 by which the building is
formed. Within a space 41A inside the wall 41, a long-wavelength
surface emission laser diode chip 32 having laser oscillation
wavelength of the 1.1-1.7 .mu.m band is provided at the A point as
an optical source. Further, The photodiode that constitutes the
photodetection unit 34 is provided in the B point. Also the
reflection part R is provided in the space 41A so as to deflect the
transmission path between the A point and the B point. FIG. 42
shows the plan view of the room and the optical
transmission/reception system merely for the purpose of
explanation. Because of this, the room or the wall and the optical
transmission/reception system are not drawn to scale.
[0449] Furthermore, though not indicated by FIG. 42, in each
optical transmission/reception unit 32 and 34 at the points A and
B, there maybe other connect instruments or connectors. These may
be in the internal space 41A of the wall 41, or may be in the room
42.
[0450] FIG. 43 shows the plan view of an optical
transmission/reception system in the room 42. Those parts explained
previously are designated by the same reference numerals and
explanation is omitted.
[0451] FIG. 43 is referred to. The laser diode 32 and the
photodiode 34 are tied with one straight line by the optical fiber
F12 crossing the inside of room 42. Therefore the optical
transmission/reception system is arranged on the floor, under the
floor, or in the ceiling of the room.
[0452] FIG. 44 shows another example of the optical
transmission/reception system in the room 42. Those parts explained
previously are designated by the same reference numerals and
explanation is omitted.
[0453] FIG. 44 is referred to.
[0454] In this case the laser diode 32 and the photodiode 34 are
tied by the optical fiber F12 by curving the optical fiber F12
while utilizing the flexibility of the optical fiber.
[0455] However, the optical fiber F12 has to be arranged so as to
cross the room 42, but it can change the direction of the
transmission path only with big curvature even if it uses the
flexibility of optical fiber. However, the optical fiber has to be
bent with a large radius of curvature, and the optical
transmission/reception system has to be constructed to or under the
floor or in the ceiling of room 42 in the state that the optical
fiber 12 crosses the room 42. Thus, it is necessary to dispose the
optical fiber under the floor or on the floor or on the
ceiling.
[0456] In the arrangement shown in FIGS. 26 and 27 the optical
fiber has to be arranged so as to cross the room 42. Therefore, the
optical fiber on the floor, under the floor or in the ceiling of
the room is arranged in a very complicated and disorderly manner.
It gets more complicated when a plural optical
transmission/reception system is arranged. Then there are the
problems of intertwisting, and maintenance after installation
become difficult. Especially, such an arrangement is very dangerous
in the case the optical fiber F12 is arranged on the floor where a
pedestrian can hook the foot.
[0457] The optical fiber transmission path of this embodiment of
the present invention is well-suited to the wall 41 or pillar or
other parts of the building, because the optical fiber transmission
path can be bent 90 degrees by using the reflection part R as shown
in FIG. 42 and is otherwise arranged beautifully in appearance,
even if it is arranged in a place where it is not inside of wall 41
and can be seen by the eye. It gets complicated if a plural optical
transmission/reception system is arranged, but there is not a
problem of the optical fibers intertwisting. The maintenance after
installation can be done easily.
[0458] The reflection part R for the deflection of the transmission
path is provided, in constructing such a peripheral (out of sight)
optical transmission/reception system in this embodiment. By using
the reflection part R, the transmission path is able to be arranged
effectively. Instead of exposing the transmission path
unnecessarily, the reflection part material R puts the transmission
path inside the ceiling, under the floor or in the wall of a
building. The transmission path does not occupy area needlessly.
The building design comes to be produced effectively. The degree of
freedom for an aesthetic design increases.
[0459] Furthermore the reflection part R is not necessarily
restricted to 90 degrees, in the case that the direction of the
transmission path is bent by reflection part material R. However,
in the normal case, a building is executed/designed, unless there
is a demand for a special design, for pillars, walls, floors, and
ceilings to provide many cases where one straight line crosses
another straight line at 90 degrees. In such normal cases it is
desirable for reasons of esthetic beauty and of efficiency, when
the optical fiber transmission path changes 90 degrees, for it to
be arranged along a pillar, wall, floor, ceiling and the
others.
[0460] Also, exposing the optical fiber transmission path to the
eye unnecessarily be avoided in constructing such an optical
transmission/reception system, by using the reflection part
material R for changes in direction of the transmission path so
that the transmission path does not occupy area needlessly inside
the ceiling, under the floor or in the wall of a building. Thereby
the building design comes to be produced effectively and the degree
of freedom for aesthetic design increases.
[0461] FIG. 45 shows an example of another optical
transmission/reception system of this embodiment. The optical
signal emitting from the emission part 32A of the laser diode 32,
in this example, is transferred in space. It goes straight along
the arrow of FIG. 45. It is received with the light detector part
34A comprised of a photodiode and other photodetection devices
34.
[0462] As explained according to the embodiments of the present
invention previously, it is preferable to use a surface-emission
laser diode equipped with the improved semiconductor Bragg
reflectors 12 and 18 further with the non-optical recombination
elimination layer and having the laser oscillation wavelength band
of 1.1-1.7 .mu.m for the laser diode, in view point of reliability,
low power driving, and low production cost. The foregoing
wavelength was not possible before in a surface emission laser
diode.
[0463] The following description will be made for the construction
in which there is only one laser diode. However, it should be noted
that, in the case of the long-wavelength surface-emission laser
diode of the 1.1-1.7 .mu.m band, a number of laser diodes are
formed on a single chip with low cost, and it is easy to construct
a multiple laser array system. Thereby, a large capacity
telecommunication system is realized.
[0464] FIG. 46 is another example of the optical
transmission/reception system of this embodiment. Those parts
explained previously are designated by the same reference numerals
and explanation is omitted.
[0465] FIG. 46 is referred to.
[0466] The optical signal emitted from the emission part 32A of
laser diode 32 is transferred through space. It goes straight in
the direction of the arrow in FIG. 46. The progress direction of
the optical path is bent by a midway, for example by the reflection
part R. The optical signal is received with the light detector part
34A comprising the photodetection devices 34 such as a photodiode,
and the like.
Fifth Embodiment
[0467] Conventionally the signal transmission of inside an
electronic apparatus has been achieved by using a conductor cable
in the form of electrical signals. Therefore numerous conductor
cables are connected inside various electronic apparatuses. The
arrangement/processing of the conductor cable was complicated,
presenting problems in the design stage and with the assembling
stage at the factory.
[0468] Thereupon, instead of exchanging of the signals by such
conductor cable, in this embodiment of this invention the
electrical signals are substituted for with optical signals by the
optical transmission/reception system as shown in FIG. 45 or FIG.
46. The optical signal that is substituted is transferred through
space. Such substitution omits or decreases conductor cable. The
conductor cable inside the instrument is decreased as much as
possible. Internal wiring is simplified. By simplifying the
internal wiring of the electronic apparatus the complexity of the
layout and processing of conductor cable is reduced. Reducing or
eliminating the complexity can increase the degree of freedom of
each part/unit and other layout items inside the apparatus.
[0469] Electronic apparatuses far which the optical
transmission/reception system of this embodiment of this invention
is incorporated include, for example, recording apparatuses such as
a copying machine or laser printer that uses electrophotography,
ink-jet recording apparatus, or recording apparatus using a silver
salt photography process. In addition, the present invention can be
used in computers, video apparatuses, television sets.
[0470] As for FIG. 47, the figure shows an electrophotographic
copying machine 541 to which this embodiment is applied. FIG. 48
enlarges FIG. 47.
[0471] FIG. 48 shows the internal structure of the
electrophotographic copying machine 541 schematically. Those parts
explained previously are designated by the same reference numerals
and explanation is omitted.
[0472] FIGS. 47 and 48 are referred to.
[0473] The electrophotography copying machine 541 has a sheet feed
cassette 542 and a sheet recovery tray 543. Furthermore wiring
board 550 and image formation mechanism 549 including sheet feed
mechanism 548 and photosensitive drum 546 are connected by
electrical circuit and included in the main body case that is
closed with cover 547a and fitted to the case 547.
[0474] And also the photodiode 34 and the laser diode 32 are
located in the case 547 respectively at the point A and the point
B, all previously explained. The laser diode 32 and photodiode 34
are connected optically, by the construction explained with FIG. 41
previously and use optical fiber F1, the mirror R and optical fiber
F2.
[0475] FIG. 49 shows an ink-jet recording apparatus 551, as another
example to which the present invention is suitably applied. FIG. 50
is the figure that enlarges FIG. 49 and schematically shows the
internal structure.
[0476] FIGS. 49 and 50 are referred to.
[0477] The ink-jet recording apparatus 551 has a case 554
comprising a lower part case 554b and an upper part case 554a. A
sheet feed mechanism 555 and an ink-jet recording head 556 are
provided on the upper part case 554a.
[0478] FIG. 50 is referred to. The ink-jet recording head 556 is
provided so as to move right and left on a carriage 556A. Image
formation is performed on the recording sheet that is pressed on
platen roller 558 by recording part 557.
[0479] Wiring board 559 provides an electronic circuit in the lower
part case 554b. Optical fiber F12 similar to the embodiment
explained above is provided between the A point and B point in the
wiring board 559. Optical fiber 12 transfers the optical signal to
the photodetection device 34 at the B point from the optical source
of the laser diode chip 32 provided at the A point.
[0480] According to this embodiment of the present invention,
conductor cable in the electronic apparatuses is eliminated or
reduced. Conductor cable is replaced inside the instrument by an
optical transmission/reception system. The examples of an
electrophotography copying machine and ink-jet recording apparatus
were shown in FIGS. 47-50, as the electronic apparatuses to which
this embodiment is applied. This invention is not restricted to
these apparatuses.
Sixth Embodiment
[0481] In the recording apparatuses such as an ink-jet recording
apparatuses or recording apparatus of silver salt photography or an
electrophotographic recording apparatus, there arises a problem in
that toners or ink or liquid developer form a mist. Further, dust
particles such as paper dust may float inside the apparatus. Thus,
the inside of such apparatuses is not an ideal environment for an
optical transmission/reception system of this invention.
[0482] In consideration this problem, the sixth embodiment of this
invention shown in FIG. 51 provides a cover member 32B covering the
laser diode chip 32 or a cover member 34B covering the photodiode
34. For example, these cover members 32B and 34B are formed by a
transparent glass-like member. In FIG. 51, those parts explained
previously are designated by the same reference numbers, and
explanation thereof will be omitted.
[0483] It is possible to use a high-precision plastic member, from
which strain is removed, in addition to a glass, for the cover
members 32B and 34B. Such a cover member provides physical and
chemical protection of the laser diode and the photodetection
device used in this invention that are produced by a sophisticated
process.
[0484] Furthermore the toner, ink or paper dust and other foreign
matter cannot be allowed to adhere to the laser diode chip of this
invention and spoil the function of the optical
transmission/reception system, such as shading of the laser beam.
Therefore it is desirable to form these cover members detachable.
By making the cover members detachable, the cover members may be
removed right away anytime that foreign matter adheres. Furthermore
a cover is not provided to reflection member R in FIG. 51. As
occasion demands a cover can be provided also to the reflection
member R.
[0485] Considering the contamination originating from toner and
paper dusts, it is thought that it would be more effective to
provide such a cover member 32B or 34B in the apparatuses such as a
printer or copying machine than a computer or a video apparatus or
a television set.
Seventh Embodiment
[0486] FIG. 52 shows an example another optical
transmission/reception system that uses the long-wavelength
surface-emission laser diode having the laser oscillation
wavelength of 1.1-1.7 .mu.m band. Those parts explained previously
are designated with the same reference numerals and explanation is
omitted.
[0487] FIG. 52 is referred to. This embodiment concatenates and
connects plural optical fibers along the optical transmission path
in the construction of FIG. 36. This embodiment enables an optical
transmission path of long distance. There are three optical fibers
in this example.
[0488] The optical telecommunication system was previously studied
with the wavelength of 0.85 .mu.m. However, the transmission loss
of the optical fiber was large at this wavelength and it was not
practical. On the other hand, it has been difficult previously to
compose the surface-emission laser diode oscillating stably at a
practical long wavelength band in which the transmission loss
becomes minimum in the optical fiber.
[0489] In accordance with this invention, the surface-emission
laser diode that oscillating at the wavelength band of 1.1-1.7
.mu.m is realized by improving the semiconductor Bragg reflector 12
or 18 and providing the non-optical recombination elimination
layers 13 and 17. Thus a practical long-wavelength band optical
telecommunication system is now possible.
[0490] In the illustrated example, the laser beam that emitted from
the laser diode emission part 32A of the long-wavelength
surface-emission laser diode chip 32 is injected into a first
optical fiber FG1 for transmission, and the laser beam exited the
first optical fiber FG1 is injected to a second optical fiber FG2
for further transmission. Furthermore, the laser beam that exited
this second optical fiber FG2 is injected into a third optical
fiber FG3 for transmission, wherein a the photodiode chip 34 having
the photodetector part 34A is coupled to the third optical fiber
FG3 for detection. Thus, the optical fibers FG1, FG2 and FG3 are
disposed along the transmission path of the laser beam.
[0491] Between the laser diode chip 32 and the first optical fiber
FG1, there is provided an optical connection module MG1 for
connecting the laser diode and optical fiber. Similarly, optical
connection modules MG2, MG3 and MG4 are provided between the
optical fibers or between the optical fiber and the photodiode chip
for optical coupling.
[0492] FIG. 53 shows the bi-directional optical
transmission/reception system that has the construction that
provides transmission system FGA corresponding to the system of the
above FIG. 52 and transmission system FGB inverted to the system of
the FIG. 52. For the parts in the FIG. 53 that are the same as
those explained above with the same reference numerals, explanation
is omitted.
[0493] FIG. 53 is referred to.
[0494] The transmission system FGB includes, from the right to the
left, a third optical fiber FR3 transmitting the laser beam emitted
from the laser diode part 32A of the surface-emission laser diode
chip 32, a second first optical fiber FR2 transmitting the laser
beam emitted from the optical fiber FR3, a first optical fiber FR1
transmitting the laser beam emitted from the optical fiber FR2, and
the light detector part 34A of the photodiode chip 34 is coupled
optically to the first the optical fiber FR1.
[0495] A connect module MR4 is provided between the laser diode
chip 32 and the third optical fiber FR3 in the transmission system
FGB, wherein the connect module MR4 couples optically the laser
diode chip 32 and the optical fiber FR3. Similarly, there are
provided connect modules MR1, MR2 and MR3 are for optical couplings
respectively between the optical fibers FR3 and FR2, between the
optical fibers FR2 and FR1, and between the optical fiber FR2 and
the photodiode chip 34.
[0496] FIG. 54 shows an example of parallel optical
telecommunication system in which n optical transmission/reception
systems each having the construction of FIG. 52 are arranged in
parallel.
[0497] FIG. 54 is referred to.
[0498] The plural laser emission part 32A is provided on a laser
diode chip 32. Further, first, second and third optical fiber
groups MFG1, MFG2 and MFG3 are constructed by a number of optical
fibers each corresponding to one of the plural optical emission
parts 32A. Further, there are provided a number of photodetector
parts 34A in the photodiode chip 34 in correspondence to the plural
laser emission parts 32A.
[0499] In the present invention, in which the surface-emission
laser diode chip is used, it is easy to provide a number of laser
emission parts 32A on the laser diode chip 32. By providing a
number of laser emission parts 32A on the laser diode chip, it
becomes possible to realize a large capacity telecommunication
system easily.
[0500] Furthermore it is possible to construct a large capacity
bi-directional optical transmission/reception system that expands
the optical transmission/reception system of FIG. 54, in accordance
with the construction of the bi-directional optical
transmission/reception system of FIG. 53 and using the optical
fibers of plural groups in parallel, although FIG. 57 does not
indicate a bi-directional system.
Eighth Embodiment
[0501] Next another embodiment of this invention will be
explained.
[0502] FIG. 55 shows the construction of an optical connection
module MG1 that is used to optically couple the surface-emission
laser diode chip 32 and the first optical fiber group MFG1 in the
left part of FIG. 54. FIG. 55 shows optical connection module MG1
schematically with a rectangular dotted line.
[0503] Detailed construction of the optical connection module MG1
will be explained hereinafter with reference to FIGS. 39-59. Those
parts explained previously are designated by the same reference
numerals and explanation is omitted.
[0504] FIGS. 56 and 57 show the construction of the
surface-emission laser diode chip 32 and the optical fiber group
MFG1 respectively in the state before they are coupled and the
state after they are coupled.
[0505] FIG. 56 is referred to.
[0506] Optical connection module MG1 shown by a rectangular dotted
line part schematically in FIG. 55 is formed of a fiber holder 61
and a fiber holder 62 that holds each of the optical fibers fg1
that constitutes the optical fiber group FG1. The Fiber holder 62
is inserted into chip holder 61 as shown in FIG. 57 with arrows so
as to face each other. Thus, the optical coupling is achieved
between each of the laser diode emission parts 32A and the edge
surface of the corresponding optical fiber fg1. Thereby the desired
optical coupling is achieved.
[0507] In the case of this invention, the holders 61 and 62 are
provided with discrimination means such that each of the laser
diode emission part 32s A and each of the optical fiber fg1 face
with each other in one-to-one relationship, without ambiguity in
the lateral or vertical directions. For this purpose, an arrow mark
is provided to each of the laser diode chip holder 61 and the fiber
holder 62 in the example of the drawing. Thereby, optical
connection can be performed very efficiently, because of the
ability to discriminate the direction of holders 61 and 62
correctly and instantaneously at the time of the connection between
laser diode chip 32 and optical fiber group FG1, at the time the
optical transmission/reception system of the present invention is
constructed.
[0508] Furthermore, the discrimination means is not restricted to
the arrow mark but it is also possible to use a different color.
Furthermore, the discrimination means is not necessarily restricted
to the ones that can be visually recognized. It is possible to have
a discrimination means using the tactile sense by utilizing the
unevenness and other features of form. It is possible to
discriminate the direction of holders 61, 62 briefly even in the
case of construction work in darkness or nighttime, in the case
that discrimination means able to discriminate by the tactile sense
is used.
[0509] FIGS. 58 and 59 show a different example of the optical
connection module MG1 that consists of the combination of laser
diode chip holder 61 and fiber holder 62.
[0510] FIGS. 58 and 59 are referred to.
[0511] With the case as shown in FIGS. 39 and 40 this embodiment
provides discrimination means to each of laser diode chip holder 61
and fiber holder 62 similarly. So that it is possible to connect
and locate both holders accurately, in this embodiment a flange
surface is additionally formed so as to allow the edge part of
laser diode chip holder 61 engage the fiber holder 62. The flange
surfaces 61A and 62B are shown in FIG. 58.
[0512] Next is explained another feature of this invention with
regard to the discrimination means, the location/connection means
shown in FIGS. 56-59, and the optical connection module MG1 that
causes coupled light between the surface-emission laser diode chip
and the 1st optical fiber group. Yet the concept of this invention
is not only applied to optical connection module MG1 that couples
optically the surface-emission laser diode chip and the first
optical fiber group MFG1 but it is also applied to the optical
connection module MG2 so that connects the first optical fiber
group MFG1 and the second optical fiber group MFG2 optically.
[0513] FIG. 60 shows an example of optical connection module MG2
that couples optically such an optical fiber group to a different
optical fiber group.
[0514] FIG. 60 is referred to.
[0515] In this case, the means for discriminating the direction is
provided on the fiber holders 64 and 65 in the form of arrow mark
such that that each optical fiber fg1 of the first optical fiber
group MFG1 couples properly to the corresponding optical fiber fg2
in the second optical fiber group MFG2. By providing such
discrimination means it becomes possible to discriminate the mutual
direction instantaneously at the time of the connection between the
first optical fiber group MFG1 and the second optical fiber group
MFG2, as the optical transmission/reception system is constructed
in accordance with the present invention.
[0516] Furthermore, FIGS. 39-59 are examples of the discrimination
means and the present invention is not limited to such an arrow
mark. The discrimination means might well use the difference of
color. The discrimination means might well be able to use the
tactile sense that senses the unevenness and others of a certain
form. By using the discrimination means that is able to use the
tactile sense, the advantage is being able to feel the direction of
the fiber holder in the case of working during darkness, or even in
the case of performing construction work at night.
[0517] In the construction of FIG. 60, not only the discrimination
means is provided to each of the first fiber group holder 64 and
the second fiber group holder 65, but there are formed the flange
surfaces on the edge part of the first fiber group holder 64 (A
part of FIG. 60) and on the edge of the second fiber group holder
65 (B part of FIG. 60) such that the flange surfaces engage with
each other in the state that there is an optimum optical
coupling.
[0518] Furthermore, such discrimination means and the
location/connection means can be used also in the optical
connection module MG3 that couples the second optical fiber group
MFG2 and the third optical fiber group MFG3 optically or in the
optical connection module MG4 that couples optically the third
optical fiber group MFG3 and the photodiode chip 34. When the
optical transmission/reception system is constructed according to
this invention, the mutual orientation can be discriminated between
the optical fiber groups or between the optical fiber group and the
photodiode chip instantaneously and exact positioning and optical
coupling is achieved.
Ninth Embodiment
[0519] Furthermore, the optical connection module according to
another embodiment of the present invention is accomplished by
using plural optical fibers in parallel. It should be noted that
the reason it has become possible to construct the large-capacity
optical transmission/reception system for a distance from several
centimeters to several hundred kilometers is that stable laser
oscillation of surface-emission laser diode at the wavelength of
1.1-1.7 .mu.m wavelength has become possible as a result of the
present invention as noted before. By using the surface-emission
laser diode of the 1.1-1.7 .mu.m oscillation wavelength, the
inspection of laser elements in the laser diode array is
substantially facilitated. Further, the productivity of the optical
transmission/reception system is improved remarkably. Construction
of such an optical transmission/reception system was not possible
in the conventional surface emission laser-diode of 0.85 .mu.m
wavelength band. By using the surface-emission laser diode the
present invention, it became for the first time possible to realize
a commercial base optical transmission/reception system.
[0520] FIG. 61 shows another example of the telecommunication
system that uses a long-wavelength surface-emission laser diode. It
should be noted that FIG. 61 shows an optical fiber 72 drawn out
from a module package 71 in which a surface-emission laser diode
chip is accommodated and an optical fiber 73 for telecommunication
that is connected to the optical fiber 72. In FIG. 61, those parts
explained previously are designated by the sane reference numerals
and explanation is omitted.
[0521] FIG. 61 is referred to.
[0522] The optical fiber 72 drawn out from module package 71 via
the connector 71A is connected by welding or other methods to the
optical fiber 73 for telecommunication at fiber connection part 74.
In this case, a predetermined connection margin is required. With
regard to the optical fiber cable 72 drawn out from the module
package 71, when the optical fiber length Lg shown in FIG. 61 is
too short, a very fine work is needed to assemble the module
package 71. Thereby, the production cost is increased. Also it
becomes difficult to produce a high-precision module.
[0523] The optical fiber used in the telecommunication systems of
this invention is a very minute thing. The diameter is only a 50
.mu.m or 62.5 .mu.m in the typical case. When assembling the module
package 71 by using such an optical fiber, for example, it is
unnecessary to hold the optical fiber 72 extending from the module
package 71 by using a pair of tweezers. Thereby, it is difficult to
hold the optical fiber 72 unless the length Lg of the optical fiber
72 is less than a certain length.
[0524] Much of the assembling work can be done by an automation
apparatus. Even in such a case, a tool is necessary for holding
such a minute optical fiber. In consideration of easiness of
holding the distal end part of the optical fiber by the holding
tool, it is necessary to secure a certain length Lg for the optical
fiber 72.
[0525] The inventor of the present invention, recognizing the
importance of doing the assembly work smoothly, conducted and
evaluated extensive assembly of optical fiber systems. Through
various kinds of experimental production operations, it was
determined that length Lg shown in FIG. 61 must be at least 1 mm.
In other words, the length Lg (the guide optical fiber length Lg
shown in FIG. 61) of the optical fiber 72 drawn out from the module
package 71, in which the surface-emission laser diode chip is
accommodated, has to be set to one or more millimeters.
[0526] In the case the length of this part is reduced to the order
of microns comparable to the optical fiber diameter, assembling
work has to be done under a microscope, using equipment that is
expensive and high-precision and highly mechanized. Such work is
unrealistic from industrial view point.
[0527] It should be noted that such a guide optical fiber 72 is
connected to the optical fiber constituting a part of the optical
telecommunication system shown in FIG. 61 by welding. Thereby,
there is a need to secure a welding margin (margin Gm shown in FIG.
62) in view of melting of the fiber edge of the guide optical fiber
72. The welding margin would be the order of millimeters.
[0528] The inventor of the present invention conducted welding
experiments and evaluated the necessary length of the welding
margin Gm of FIG. 62.
[0529] In the case that guide optical fiber length Lg is in the
order of microns, it was discovered that the fiber edge of the
optical fiber 71 melts excessively. For example, in the case the
length Lg is 200 .mu.m, 500 .mu.m or 900 .mu.m, the edge surface of
optical fiber 71 melted excessively and a fine joint was not
obtained. On the other hand, it was confirmed that a fine joint can
be obtained in the case the length Lg is made to be one millimeter
or more, such as 1 mm or 3 mm.
[0530] In summary, it is concluded that assembling of then module
package 71 requires that the length Lg be one or more millimeters
in order to connect with the optical fiber of the optical
telecommunication system to the guide optical fiber.
[0531] With regard to the upper limit of the guide optical fiber
72, it is sufficient for module assembly work to secure the length
of 20 mm for the guiding purpose. In the case the optical fiber 7
is too long, the excessive optical fiber can be cut after the
assembling of the optical module 71 to a suitable length.
[0532] Next is an explanation about another aspect of this
embodiment.
[0533] The above explanation discussed the relation with the
surface-emission laser diode 32 acting as the optical source and
the guide optical fiber 72, while the same relation holds also at
the detecting side.
[0534] While illustration is omitted, it is possible to realize a
photodetection unit by replacing the surface emission-laser diode
32 of FIGS. 61 and 62 with the photodetection device 34 such as
photodiode. Thereby, there is an optical fiber corresponding to the
guide optical fiber 72 of FIGS. 61 and 62 in direct contact with
the photodetection device 34.
[0535] The same thought is necessary with regard to the module
assembling and connection to the optical telecommunication system
in the side of the photodetection device.
[0536] In the present invention, a highly reliable and practical
assembling of the photodetection unit is achieved by setting the
length of the guide optical fiber connected directly with the
photodetection device 34 (corresponding to the guide optical fiber
72 of FIGS. 61 and 62) to one millimeter or more. Further, a
reliable connection is achieved with the optical fiber of the
optical telecommunication system.
[0537] With regard to the upper limit, a length of 20 mm is
sufficient similarly to the module package for the optical source
side.
Tenth Embodiment
[0538] FIG. 63 shows the example of a telecommunication system that
uses the long-wavelength surface-emission laser diode chip 32 with
the laser oscillation wavelength 1.1-1.7 .mu.m band with plural
optical fibers. Those parts explained previously in the FIG. 63 are
designated with the same reference numerals and explanation is
omitted.
[0539] In the construction of FIG. 63, the laser diode 32A in the
laser diode chip 32 is driven by a communication control unit 81
via a laser diode driver circuit 82. The optical signal is supplied
to the optical fibers 72 and 73 in the form of an optical beam.
[0540] Conventionally, an optical telecommunication system was
studied by using the laser oscillation wavelength of 0.85 .mu.m.
However, because of the large transmission loss in the optical
fiber, it was not practical for long distance telecommunication.
Further, in the case of the telecommunication system that uses the
plural optical fibers in combination of conventional edge-emission
type laser diodes, there is a need of adjusting the optical
coupling for each the optical fibers with each of the laser diodes.
Thus, the adjustment process was complicated. Further, it was
difficult to connect the laser diodes directly to form a two
dimensional array. Further, in an edge-emission laser diode, the
emission angle of the laser beam is large and the laser beam
generally has a large aspect ratio. Thus, it has been provide a
coupling lens between the emission part and the optical fiber for
better optical coupling efficiency.
[0541] Contrary to the prior art, a stable drive is realized with
low energy and low heating by using the long-wavelength
surface-emission laser diode of 1.1-1.7 .mu.m band according to
this invention.
[0542] FIG. 64 shows the construction of the surface-emission laser
diode of the 1.3 .mu.m band according to this invention. The
emission angle of the laser beam in such a surface-emission laser
diode is smaller as compared with an edge-emission laser diode and
takes the value of about 15 degrees in both the vertical and
horizontal directions. The laser beam form is circular, and there
is no need of shaping the laser beam.
[0543] Thus, individual laser diodes 32A can be connected with
corresponding optical fibers 72 without using a coupling lens, as
long as the irradiation diameter of the laser diode is smaller than
the core diameter of the optical fiber 72. The same is true also in
other long-wavelength bands.
[0544] Thus, by using a surface-emission laser diode according to
the present invention, the difference between the irradiation
diameter of the laser diode 32A and the optical fiber core diameter
provides a tolerable margin, and it becomes possible to adjust
plural optical fibers altogether.
[0545] As shown in FIG. 64, the laser beam that is emitted from an
emission part, in other words, the laser diode 32A, is injected
with high efficiency into the core at the edge surface of the
optical fiber. The optical signal thus injected into the optical
fiber is transmitted over a long distance with little transmission
loss due to the long the wavelength of the optical signal. By using
plural optical fibers, a very high quality optical
telecommunication system of practical value is materialized. The
laser diodes 32A can form in arbitrary two-dimensional arrangement
in the surface-emission laser diode chip 32 as long as they do not
cause interference.
[0546] The plural optical fibers 72 are fixed with each other by
using a jig 91 shown in FIGS. 65A-65C, in which each of the optical
fibers is provisionally fixed by the jig 91 and pouring a resin 92
into the jig as represented in FIG. 65C. Thereby, the positional
adjustment can be omitted at the time of system construction, by
providing a positioning guide in the connector connection part 71A.
Thus, it becomes possible to easily construct an optical
telecommunication system while using plural optical fibers.
[0547] Meanwhile, the number of the optical fibers used in an
optical telecommunication system changes depending on the system.
Thus, it is advantageous to design the system such that the number
of optical fibers can be changed flexibly.
[0548] Thus, as shown in FIG. 66B, a number of openings having the
size of the optical fiber diameter are formed on the connector
connection part 71a, and the connector connection part 71a is
attached on the surface-emission laser diode chip 32 of FIG. 49C
with an adjustment. Further, several kinds of connectors 71b
carrying plural optical fibers as shown in FIG. 66A are provided,
and one of the connectors 71b is selected and inserted into the
connector connection part 71a. By using such a construction, waste
of optical fibers in the optical connector 71A is eliminated, and
it becomes also possible to add optical fibers later according to
the needs. Thereby, it becomes possible to increase the degree of
freedom of design of the optical telecommunication system.
[0549] Furthermore, in long-wavelength surface-emission laser diode
chip 32 of 1.1-1.7 .mu.m band of this invention, the laser diode
elements constituting the emission part 32A can be arranged in
two-dimensional array in view of this, each optical fiber 95 is
provided with a resin coating 96 having a hexagonal cross section
as shown in FIG. 67 such that the two-dimensionally arranged
optical fibers do not contact directly with each other. Thereby, an
optical cable having a closest packing structure for the optical
fibers is realized as represented in FIG. 68. In the optical fiber
cable of FIG. 68, the cross-sectional area of the optical fibers is
minimized because of the closest packing arrangement of the optical
fibers.
[0550] In the construction of FIG. 67, it is also possible to use a
fixing member of a resin having a number of holes of hexagonal
arrangement for holding the optical fibers, in place of providing a
resin coating 96 outside the optical fiber 95.
Eleventh Embodiment
[0551] FIG. 69 shows an example of the optical fiber 101 used in a
telecommunication system that uses a long-wavelength
surface-emission laser diode. Those parts explained previously are
designated by the same reference numerals and the description
thereof will be omitted.
[0552] FIG. 69 is referred to. The optical fiber 101 is injected
with the laser beam emitted from the laser diode emission part 32A.
The optical fiber 101 consists of cladding 101B enclosing a core
101A, and the core 101A transmits the laser beam thus injected. The
diameter D of the core 101A is determined with regard to the length
L so as to satisfy the condition 105.ltoreq.L/D.ltoreq.109.
[0553] Conventionally, an optical telecommunication system was
studied by using the laser oscillation wavelength of the 0.85 .mu.m
band. However, the transmission loss of the optical fiber was too
large and there was no practical value. Also there was no laser
diode that can oscillate stably in the long-wavelength band, in
which the transmission loss is minimum, was not realized. In
accordance with the present invention, a surface-emission laser
diode oscillating stably at the laser oscillation wavelength of
1.1-1.7 .mu.m band is materialized as a result of improvement of
the semiconductor Bragg reflectors 12 and 18 and by the provision
of the non-optical recombination elimination layers 13 and 17. As a
result, a practical long wavelength band optical telecommunication
system became possible.
[0554] FIG. 70 shows the construction of a long distance optical
telecommunication system using the above long-wavelength
surface-emission laser diode chip 32 and the optical fiber 101.
[0555] FIG. 70 is referred to. The long distance optical
telecommunication system uses the long-wavelength surface-emission
laser diode chip 32. The optical fiber is injected with the laser
beam that is emitted from the laser diode emission part 32A of the
laser diode chip 32. Further, the optical telecommunication system
includes a first optical fiber 101A having the core diameter D of
50 .mu.m and the length L of 5 km and second optical fiber having
the core diameter D of 50 .mu.m and the length L of 5 km injected
with the laser beam emitted from the optical fiber 101A, wherein
the optical telecommunication system further includes the
photodiode chip 34 including the photodetector part 34A injected
with the laser beam emitted from the second optical fiber 101B.
[0556] Between the laser diode chip 32 and the first optical fiber
101A, the connection module 71 is provided, and a similar
connection module 75 is provided between the second optical fiber
101B and the photodiode chip 34. Furthermore a repeater 101C is
provided between the first optical fiber 101A and the second
optical fiber 101B for amplification and the repeating of the
optical signal.
[0557] In the constitution of FIG. 70, it should be noted that the
first and second optical fibers 101A and 101B are produced usually
with several thousand kilometers as a unit. However, a length of
several hundred kilometers becomes the unit of the optical fiber
product due to existence of pores or other defects that occur
during the production process. In consideration of transmission
loss, the practical length of the optical fiber to the repeater
101C is 5-50 km in the case the optical fiber has a core diameter
of 50 .mu.m. The optical fiber of this length is used for the
optical fiber telecommunication.
[0558] In the case the transmission length is longer than 50 km, it
is necessary to provide a relay point for signal amplification in
view of the transmission loss. When the distance is shorter than 5
m, various telecommunication means are available other than the
optical telecommunication system.
[0559] For optical fiber 101 used with this embodiment, a quartz
glass optical fiber having a core diameter D of several microns, or
a plastics optical fiber having a core diameter D of several
hundreds microns may be used. These optical fibers can be used
alone or in the form of a bundle.
[0560] FIG. 71 shows an optical telecommunication system for
bi-directional optical telecommunication by using a bundle of
quartz glass optical fibers 101-A and 101-B each having a diameter
D of 50 .mu.m and a transmission length L of 500 m, wherein these
optical fibers are separated from each other at the photodetection
end and the optical source end.
[0561] FIG. 71 is referred to.
[0562] There can be seen that optical transmission/reception parts
102A and 102B are provided in the present invention in which the
long-wavelength surface-emission laser diode chip 32 and the
photodiode chip 34 form a pair. Thus, the first optical fiber 101A
is coupled to the emission part 32A of laser diode chip 32 in the
optical transmission/reception part 102A, and the optical fiber
101A is further coupled optically to the optical detection part 34A
of photodiode chip 34 in the optical transmission/reception part
102B. Similarly the second optical fiber 101B is coupled optically
to the emission part 32A of the laser diode chip 32 in the optical
transmission/reception part 102B, and the optical fiber 101B is
coupled optically to the optical detection part 34A of photodiode
chip 34 in the optical transmission/reception part 102A. Thereby,
the laser diode chip 32 and the photodiode chip 34 constituting the
optical transmission/reception part 102A form the connection module
71. Also, the laser diode chip 32 and the photodiode chip 34
constituting the optical transmission/reception part 102B form the
connection module 75.
[0563] By expanding the foregoing construction of using the
long-wavelength surface-emission laser diode chip 32 and the
photodetector chip 34 as a pair for the case in which a number of
laser diode elements 32A are formed on the laser diode chip 32 and
a number of photodiode elements are formed on the photodetector
chip 34, it is possible to realize a large capacity optical
telecommunication system.
[0564] FIG. 72 shows the case of applying the present invention to
a high speed multimedia network that uses a fluorinated plastic
optical fiber having a core diameter of 100 .mu.m and the
transmission length of less than 100 m.
[0565] FIG. 72 is referred to.
[0566] A first optical fiber 111 having a core diameter D of 50
.mu.m extends from a station-side apparatus 110 and a second
optical fiber 112, formed by bundling 50 optical fibers each having
a core diameter D of 50 .mu.m and a length of 1 km, is connected to
the optical fiber 111. As a result, there is formed a the high
speed optical transmission system capable of transmitting
information with the speed of the order of Gbps with a total
distance of 100 km. In the example of FIG. 72, the optical repeater
111R is provided between the optical fiber 111 and optical fiber
112, to compensate for the damping of the signal by the propagation
loss of quartz glass optical fiber.
[0567] The optical signal transmitted through the second the
optical fiber 112 enters into a network terminator 115 at first and
converted to an electric signal. Further, the electric signal is
distributed to necessary lines 115a-115c, and the electric signal
thus supplied to each line is transmitted further in the form of
the optical signal by an optical telecommunication system 116 that
includes the optical transmission/reception parts 102A and 102B
explained to with reference to FIG. 71. The optical signal thus
transmitted is supplied to a corresponding optical output port. In
the optical telecommunication system 116, an optical fiber 117a
having a core diameter of 100 .mu.m and a length of 10 m is used
for the optical transmission to the optical output port
corresponding to the line 115a, while an optical fiber 117b having
a core diameter of 100 .mu.m and a length of 50 m is used for the
optical transmission to the optical output port corresponding to
the line 115b. Further, an optical fiber 117c having a core
diameter of 100 .mu.m and a length of 100 m is used for optical
transmission to the optical output port corresponding to the line
115c. The optical signals thus transmitted are converted to
electric signals by the optical transmission/reception part 102B in
each optical output port and is supplied to corresponding terminal
devices 102C.
[0568] Further, the transmission from each terminal device 102C is
transmitted to the optical transmission/reception part 102A in the
form of an optical signal through the optical fibers 117a-117c and
reaches the station-side apparatus 110.
[0569] In this embodiment, the core diameter of the optical fibers
117a-117c has a large value of 100 .mu.m, and high-precision
alignment by using a high precision lens system is not necessary.
Thus, it becomes possible to achieve optical connection to various
apparatuses in a home or office quite extremely.
[0570] In the present embodiment, too, the relationship of
105.ltoreq.L/D.ltoreq.109 is maintained between the length L and
core diameter D of the optical fibers. When the transmission
distance is longer than 100 km, no substantial transmission is
possible due to the transmission loss, and thus, it is necessary to
provide a relay point for amplification and repeating. In the case
the transmission distance is shorter than 10 m, such an optical
telecommunication system is not always necessary and may be
replaced with other means.
[0571] FIG. 73A and 73B show the application of the long-wavelength
surface-emission laser diode of the present invention to a hybrid
optical integrated circuit device, respectively in a plane view and
a cross-sectional view.
[0572] FIGS. 73A and 73B are referred to. The long-wavelength
surface-emission laser diode chip 32 and the photodiode chip 34 are
carried on a ceramic substrate 51 in this embodiment. Further, the
ceramic substrate 51 carries thereon a first optical waveguide 52
having a rectangular cross-section of 7.times.7 .mu.m and a length
of 1 cm for guiding the laser beam emitted from the laser diode
32A. Further, the ceramic substrate 51 carries a second optical
waveguide 53 having a rectangular cross-section of 7.times.7 .mu.m
and a length of 1 cm similar to the optical waveguide 52 in optical
coupling with the photodetector part 34A of the photodiode chip 34
for supplying the optical signal to the photodetector part 34B.
Further, a preamplifier 51 for amplifying the output electric
signal of the photodetector chip 34 is provided on the ceramic
substrate 51.
[0573] Further, an optical fiber 54 having a core diameter D of 100
.mu.m and a length of 100 m is provided in optical coupling with
the first optical waveguide 52 via a connection module 56, and a
second optical fiber 55 having a core diameter D of 100 .mu.m and a
length of 100 m is provided in optical coupling with the second
optical waveguide 53.
[0574] Further, a connection module 57 equipped with the
surface-emission laser diode chip 32 and the photodiode chip 34 is
provided at the other end of the optical fibers 54 and 55 such that
the optical fiber 54 is coupled with the photodetector part 34A of
the photodiode chip 34 and the optical fiber 55 is coupled with the
laser diode 32A of the laser diode chip 32.
[0575] The optical waveguides 53 and 54 are formed on a silicon
substrate 58 by the steps of forming the respective cores and the
claddings that cover the cores by a photolithographic process, and
the silicon substrate 58 is provided on the ceramic substrate 51
commonly with the preamplifier 51A.
Twelfth Embodiment
[0576] FIGS. 74A and 74B show an example of a laser diode chip 120
that is used with the optical telecommunication system in which the
long-wavelength surface-emission laser diode of the present
invention is used, wherein FIG. 74A shows a plane view while FIG.
74B shows a cross-sectional view taken along a line A-A'. It should
be noted that the scale FIG. 74A is not identical with the scale of
FIG. 74B. In the drawings, those parts explained previously are
designated by the same reference numerals and the description
thereof will be omitted.
[0577] FIGS. 74A and 74B are referred to.
[0578] It can be seen that the long-wavelength surface-emission
laser diode 32A and the photodetection device 34A are formed to
monolithically on a laser diode chip 120.
[0579] The photodetection device 34A has a stacked structure
similar to the stacked structure of the long-wavelength
surface-emission laser diode 32A and is formed simultaneously with
the process of forming the surface-emission laser diode 32A. The
structure thus formed can be used as a photodetection device, by
applying with a reverse bias or using without a bias. This
photodetection device 34A has sensitivity to the oscillation
wavelength of the surface-emission laser diode 32A and can detect
the optical emission of the laser diode 32A.
[0580] As will be understood from the plane view of FIG. 74A, the
photodetection device 34A is formed in such a form to surround the
long-wavelength surface-emission laser diode 32A. On the top
surface of the laser diode 32A, an upper electrode 121 is formed
and a lower electrode 123 is formed on the bottom surface of the
laser diode chip 120. Also another upper part electrode 122 is
formed on the top surface of the photodetection device 34A. An
optical window is formed in the upper part electrode 121 of the
laser diode 32A for taking out the optical output. Such a window is
not formed in upper part electrode 122 of the photodetection device
34A.
[0581] FIG. 75 is the figure that explains the operation of the
laser diode chip 120 shown in FIG. 74.
[0582] FIG. 75 is referred to.
[0583] The laser diode 32A in the surface-emission laser diode chip
120 is positioned so as to face the end surface of the optical
fiber 125, and the optical beam emitted from the surface-emission
laser diode 32A is injected into the core of the optical fiber.
Thereby, there occurs a leakage light in the lateral direction from
the sidewall of the surface-emission laser diode 32A, and the
leakage light is detected by the photodetection device 34A provided
adjacent to the laser diode. While the amount of the leakage light
from the sidewall of the surface-emission laser diode 32A is not
much, nevertheless the leakage light is detected by the
photodetection device 34A, as the photodetection device 34A is
formed so as to surround the surface-emission laser diode 32A in
the construction of FIG. 75.
[0584] Furthermore, it should be noted that the semiconductor Bragg
reflector 18 and the upper part electrode 122 are formed on the
photodetection device 34A in FIG. 75. Therefore, the optical signal
and the like transmitted from another party via the optical fiber
125 is reflected at the surface of the photodetection device 34A as
shown with the arrow in the drawing. Thereby, the photodetector 34A
does not detect such an optical signal.
[0585] As it will be clear from the aforementioned explanation, the
present invention forms the photodetection device for detecting the
output detection of the surface-emission laser diode integrally on
the surface-emission laser diode chip. By integrating such a laser
diode chip on a ceramic substrate 111, it becomes possible to
construct the optical telecommunication system having a hybrid
constitution.
[0586] FIGS. 76A and 76B show an example of the semiconductor laser
diode chip used with the optical telecommunication system in which
the long-wavelength surface-emission laser diode of this invention
is used, wherein FIG. 76A shows a plane view while FIG. 76B shows a
cross-sectional view that taken along a line A-A'. The scale of
FIG. 76A and FIG. 76B are not the same.
[0587] FIGS. 76A and 76a are referred to.
[0588] The long-wavelength surface-emission laser diode 32A and the
photodetection device 34A are integrated monolithically on a laser
diode chip, wherein the photodetection device 34A has a stacked
structure similar to the stacked structure of the long-wavelength
surface-emission laser diode 32A.
[0589] Thus, the photodetection device 34A is formed simultaneously
with the same process as the one used for forming the
surface-emission laser diode 32A. The photodetection device 34A
thus formed has sensitivity to the wavelength of the
surface-emission laser diode 32A and can detect the optical
radiation produced by the laser diode 32A.
[0590] In the present embodiment, the upper semiconductor Bragg
reflector 18 is removed from the photodetection device 34A by an
etching process. As can be seen from the plane view of FIG. 76A,
the photodetection device 34A is formed so as to surround the
long-wavelength surface-emission laser diode 32A, and the devices
32A and 34A have respective top surfaces provided with the upper
electrode 121 and the upper electrode 122 formed with the window
for optical input/output. Further, the bottom electrode 123 is
formed on the bottom surface of laser diode chip 120.
[0591] FIG. 77 explains the operation of the laser diode chip of
FIG. 76.
[0592] FIG. 77 is referred to.
[0593] In the surface-emission laser diode chip 120, the laser
diode 32A is positioned so as to face the end surface of the
optical fiber 125 such that the optical beam emitted from the
surface-emission laser diode 32A enters into the core of optical
fiber 125.
[0594] In this embodiment, the sidewall of the surface-emission
laser diode 32A and the sidewall of the photodetection device 34A
are covered with the upper part electrode 121 or 122, and the
leakage light is not formed at the side wall of the laser
diode.
[0595] Thereupon, when an optical beam carrying an optical signal
comes in from another party on the line, the optical beam emitted
from the end surface of the optical fiber 125 enters into the top
surface of the photodetection device 34A while causing spreading as
represented by an arrow in the drawing. The optical beam is then
detected by the photodetection device 34A formed so as to surround
the surface-emission laser diode 32A. On the other hand, the top
surface of surface-emission laser diode 32A is covered with the
semiconductor Bragg reflector 18, and because of this, no
substantial optical beam invades inside the surface-emission laser
diode 32A from outside.
[0596] Thus, as will be clear from the aforementioned explanation,
it becomes possible to construct an optical telecommunication
system having an optical transmission/reception unit of hybrid
construction in the present embodiment, by forming the
photodetection device 34A for monitoring the output of the
surface-emission laser diode 32A integrally on the surface-emission
laser diode chip 120. Such a laser diode chip 120 can be integrated
on the ceramic substrate 111.
[0597] Furthermore the combination of the surface emission laser
diode and the photodetection device that explained above is merely
for the purpose showing an example, and the present invention also
includes the case of forming an array of these devices or the case
of combining the monitoring photodetector 34A for monitoring the
output of the surface-emission laser diode 32A and the
photodetector 34A for detecting the incoming signals are combined.
Naturally, the present invention is not limited to the specific
mutual positional relationship or shape of the surface-emission
laser diode and the photodetector.
Thirteenth Embodiment
[0598] Next, this invention will be explained about other
examples.
[0599] It is important to use a highly strained layer of GaInAs,
GaInNAs or GaAsSb for the active layer in order to realize the long
wavelength surface emission laser diode of 1.1-1.7 .mu.m. In order
to achieve this, it is necessary to minimize the mechanical stress
applied to the laser diode. Such a mechanical stress may result
from the thermal stress caused between the laser diode and the
mount substrate as a result of the operational temperature
environment or due to the heating of the laser diode or driver
circuits. Such a thermal stress is caused in a structure in which
materials of different linear thermal expansion coefficients are
fixed, due to the tendency of the structure that tends to maintain
original shape. Thus, the thermal stress depends on the temperature
change, the coefficient of linear thermal expansion, Young modulus,
and the like. It is possible to eliminate the thermal stress by
controlling the temperature of the whole module including the laser
diode. However, such a measure of providing a temperature
regulating mechanism is difficult in view of the increased cost.
Further, it is difficult to control the temperature completely
constant.
[0600] Thus, it is desirable, in order to provide a highly reliable
system with low cost, to use a material having a coefficient of
linear thermal expansion close to that of the laser diode for the
mount substrate so as to reduce the influence of the thermal stress
on the laser diode as small as possible.
[0601] In this embodiment, various mount substrates were produced
by using various materials having different coefficients of linear
thermal expansion and evaluation was made about the influence of
the thermal stress on the output characteristic of the laser at the
time of the laser oscillation. The surface-emission laser diode
used in the experiment has the structure explained with reference
to FIG. 1 and the laser oscillation wavelength was 1.3 .mu.m.
Further, a chip size of 5 mm.times.10 mm (thickness 0.6 mm) was
used in which 20 laser diodes were.aligned with a pitch of 300
.mu.m. The size of the mount substrate made was 10 mm.times.20 mm
(thickness 2 mm).
[0602] The result is shown in Table 5. In Table 5, O represents the
sample in which a stable output was obtained in the operational
environment of 0-70.degree. C., while .times. represents the sample
in which no stable output was obtained and hence not suitable for
practical use. TABLE-US-00006 TABLE 5 linear thermal expansion
material coefficient laser output quartz glass 0.3 .times.
10.sup.-6/K X smicrystal 2 .times. 10.sup.-6/K X CVD diamond 2
.times. 10.sup.-6/K X Si 4 .times. 10.sup.-6/K .largecircle. SiC 4
.times. 10.sup.-6/K .largecircle. AlN 5 .times. 10.sup.-6/K
.largecircle. GaAs 6 .times. 10.sup.-6/K .largecircle.
Al--Si.sub.(60Al-40Si) 15 .times. 10.sup.-6/K X Cu 17 .times.
10.sup.-6/K X
[0603] The coefficient of linear thermal expansion of the laser
diode of this invention is 6.times.10.sup.-6/K. From the above
results, it can be seen that the thermal stress is reduced and a
stable laser output is obtained when the difference of the linear
thermal expansion coefficient between the laser diode and the mount
substrate within about 2.times.10.sup.-6/K. Particularly, the
materials such as Si, SiC, GaAs, AlN are easily available and are
easily processed and handled for the mount substrate.
[0604] Further, by choosing the heat radiating member on which the
mount substrate is mounted such that the material of the heat
radiating member has a linear thermal expansion coefficient close
to that of the laser diode, the stress to the mount substrate is
minimized and the mechanical stress applied to laser diode is
reduced as a result. Further, the material constituting the heat
radiation member is required to have a high thermal
conductivity.
[0605] In the present invention, various heat radiating members are
formed by using various materials having different linear thermal
expansion coefficients and the effect of heat generated at the time
of the laser oscillation on the laser output characteristics is
evaluated.
[0606] The laser diode used for the evaluation was the same as the
one used in the evaluation of Table 5. In this experiment, an SiC
substrate having the same size as in the previous experiments was
used for the mount substrate.
[0607] Table 6 shows the results, wherein O in Table 4 represents
the sample in which a stable output was obtained in the operational
environment of 0-70.degree. C., while .times. represents the case
in which no stable output was obtained. TABLE-US-00007 TABLE 6
material thermal conductivity laser output SiO.sub.2 about 8 W/mK X
Al.sub.2O.sub.3 about 17 W/mK X Cobar about 17 W/mK X AlN about 200
W/mK .largecircle. Cu/W 180-200 W/mK .largecircle. W about 170 W/mK
.largecircle. Mo about 160 W/mK .largecircle. Cu about 390 W/mK
.largecircle.
[0608] The thermal conductivity of the laser diode of this
invention is 55 W/mK. Thus, from the results noted above, it can be
seen that a preferable result is obtained when the thermal
conductivity of the heat radiating member is larger than the
thermal conductivity of the laser diode. Thus, the heat generated
at the time of the laser oscillation is transferred to the mount
substrate and then to the heat radiation member without returning
to the laser diode when the thermal conductivity of heat radiating
member is larger than the thermal conductivity of the laser diode
of this invention. Thus, the output fluctuation of the laser
accompanied by the effect of heat accumulation does not result and
a stable characteristic suitable for practical use is obtained.
Particularly, the materials such as AlN, Cu/W, W, Mo, Cu are easily
available and can be processed easily to form a heat radiating
member.
[0609] Especially the use of Cu/W is advantageous as it allows
control of the thermal conductivity by controlling the
compositional ratio and can be used as a he package substrate to be
explained later with reference to FIG. 79.
[0610] Hereinafter, an optical telecommunication system of the
present embodiment that uses such a member will be explained.
[0611] The telecommunication system is composed of the optical
fiber or optical waveguide that operates as the transmission path
between the photodetection part including a surface photodetection
device and a driver circuit thereof and the optical transmission
part including a surface-emission laser diode and a driver
circuit.
[0612] The driver circuit of the laser diode or the photodetector
is formed on the same mount substrate commonly with the laser diode
of the photodetector to which the drive circuit cooperates,
Alternatively, the driver circuit is formed on the laser diode
substrate by a wafer process simultaneously to the process of
forming the respective devices to which the driver circuit
cooperates. Further, the optical telecommunication system performs
the bi-direction telecommunication by providing the optical
transmission part and the photodetection part at both ends of the
optical transmission path.
[0613] FIG. 78 shows an example of the optical transmission part of
such an optical telecommunication system. In the drawing, those
parts explained previously are designated by the same reference
numerals and the description thereof will be omitted.
[0614] FIG. 78 is referred to.
[0615] The optical transmission part includes the surface-emission
laser diode chip 32 and a driver circuit 32DR driving the laser
diode chip 32, a mount substrate 131 on which the laser diode chip
32 and the driver circuit 32DR are mounted, a heat radiation member
132 adjustably supporting the mount substrate 131 and for radiation
of the heat, and a metal package 133 holding the heat radiation
member 132 and acting as the heat sink or heat radiation fin, and
an optical transmission 134 acting as the optical path. The metal
package 133 and heat radiation member 132 and also mount substrate
131 are connected mechanically and thermally with each other by a
solder or a resin. Also, the laser diode 32 and the driver circuit
32DR are connected electrically by wire bonding, and the like.
[0616] For example, the laser diode chip 32 includes the laser
diode of FIG. 1, wherein the laser diode may be the one that
oscillates at the wavelength of 1.2 .mu.m. As for the mount
substrate 131, on the other hand, a Si substrate having the
coefficient of linear thermal, expansion of 4.times.10.sup.-6/K can
be used. In this case, the laser diode 32 is die-bonded with an
AuSn solder on the mount substrate 131. Thereby, electrode and
electric mechanical connection are achieved.
[0617] It should be noted that an SiO.sub.2 film 200 nm thickness
is formed on the surface of the Si substrate 131. This SiO.sub.2
film may be formed by a thermal oxidation process or a CVD process
or an SOG process. The foregoing oxide film is used for insulation,
but the heat radiation characteristic of such a film is inferior to
Si. Therefore, it is preferable that the insulation film has as
small thickness as possible within the range of providing
sufficient electric insulation. The oxide film may be omitted if it
is possible. The driver circuit 32DR that drives the laser diode
chip 32 is mounted on the same mount substrate 131 together with
the laser diode chip 32. A thermal conductive AlN substrate
(coefficient of linear thermal expansion 5.times.10.sup.-6/K,
thermal conductivity 200 W/mK) may be used as the heat radiating
substrate that carries the mount substrate. The AlN substrate has
excellent compatibility with regard to the coefficient of linear
thermal expansion. Further, a powder mold product of Cu/W may be
used for the metal package 133. For example, the molded of Cu/W
product forming the metal package 133 may have the composition of
89W-11Cu and a linear thermal expansion coefficient of
6.5.times.10.sup.-6/K and a thermal conductivity of 180 W/mK. By
using such a powder mold product, a highly precision product is
obtained with low cost. Further, it is possible to form heat
radiation fins easily and efficient heat radiation becomes
possible.
[0618] In the example of FIG. 78, a multiple mode optical fiber
having a core diameter of 50 .mu.m is connected optically as the
optical fiber 134 with the laser diode 32A in the laser diode chip
32. The stability of such an optical telecommunication system was
evaluated in the operational environment of 0-70.degree. C. from
and it was confirmed that the laser output is stable. No change of
output characteristics was observed. Further, there was no lifetime
degradation. Thus, it was found that an excellent optical
telecommunication system can be constructed.
[0619] Furthermore, it is possible to use a GaAs (coefficient of
linear thermal expansion 6.times.10.sup.-6/K) or AlN (coefficient
of linear thermal expansion 5.times.10.sup.-6/K) or SiC
(coefficient of linear thermal expansion 4.times.10.sup.-6/K) for
the substrate as the mount substrate 133 instead of the Si
substrate. Because these substrates are an insulating substrate,
formation of oxide film is not necessary when these substrates are
used. Otherwise, the same structure may be used. In view of the
cost and easiness of handling, the use of AlN is most
preferable.
[0620] Also, good results were obtained when Cu/W (coefficient of
linear thermal expansion is 6-8.times.10.sup.-6/K, thermal
conductivity is 180-200 W/mK) such as 89W-11Cu, 85W-15Cu or
80W-20Cu, or the metals of W, Mo, Cu are used for the heat
radiation member in place of AlN.
[0621] In the example of FIG. 78, it was explained based on the
construction that there is only one laser diode in the laser diode
chip 32 forming the transmission part. However, it is possible to
form an array of plural laser diodes also in the laser diode chip
32 or, by forming the array of the laser diode chip 32. Thereby, a
large capacity optical telecommunication system is constructed.
[0622] For example, in the case of the multiple laser array chip in
which plural laser diodes 32A are formed on a single chip 32, the
laser diodes 32A come close with each other and the fluctuation of
the laser output characteristic may be influenced by the thermal
stress caused by the heat generation and accumulation of the heat
thus generated. Furthermore the laser diode driver circuit 32DR is
often formed also on such a chip. Thus, the heat from such a driver
circuit 32DR is superimposed. In this, embodiment, the laser output
that stabilized without bringing about such a problem by choosing
the material of the mount substrate 131 or the heat generation part
132 appropriately.
[0623] Formerly, the surface-emission laser diode having the
oscillation wavelength of 1.1-1.7 microns did not exist. Therefore,
the technological problems with regard to the optical
telecommunication system or mounting of the surface-emission laser
diode chip of such a long-wavelength band have not been announced.
This time, the inventor of this invention recognized the concrete
technological problems for the first time by this invention and
provided the resolution.
[0624] In the explanation above, a multimode optical fiber was used
for the optical transmission path coupled optically to the laser
diode chip 32. However, the optical transmission path may also be a
single mode optical fiber or a plastic optical fiber. The laser
diode of the present invention does not need expensive cooling
device such as a Peltier device. But the present invention does not
exclude the use do such a cooling device.
[0625] Next, another embodiment of the present invention will be
explained with reference to FIG. 79, wherein those parts
corresponding to the parts described previously are designated by
the same reference numerals and the description thereof will be
omitted.
[0626] Referring to FIG. 62, the present embodiment also shows the
optical transmission part of the optical telecommunication system.
In the present embodiment, the heat radiating member 132 functions
also as the metal package 133.
[0627] It should be noted that the illustrated optical transmission
part includes the laser diode chip 32 in which a number of laser
diodes 32A are arranged in the form of laser array, the driver
circuit 32DR driving the laser diodes 32A, a mount substrate on
which the laser diode chip 32 and the driver circuit 32DR are
mounted, a metal package 133 supporting the mount substrate 131 and
functioning as the heat sink and also a heat radiating fin, an
optical fiber 134 constituting the optical transmission path, and a
ferule fixing the optical fiber 134. A solder or a resin connects
the metal package 133 and the mount substrate 131 mechanically and
thermally. Further, the laser diode chip 32 and the driver circuit
(not shown) are connected electrically by using a bonding wire.
[0628] In the illustrated example, the laser diode of FIG. 27
having the oscillation wavelength of 1.3 .mu.m is used, and four
devices are arranged with a pitch of 250 .mu.m, which is the pitch
identical with the pitch of the corresponding optical fibers.
Similarly to the previous embodiments, a Si substrate was used for
the mount substrate and the laser diode was die bonded on the mount
substrate by using an AuSn solder. Further, the electrodes are
connected mechanically and electrically.
[0629] It should be noted that the surface of the Si substrate is
formed with an SiO.sub.2 film having a thickness of 200 nm by a
thermal oxidation process. However, such an oxide film may be
formed by a CVD process or an SOG process. Further, in view of the
heat radiating performance of the SiO.sub.2 film, it is preferable
to form the SiO.sub.2 film with a thickness as small as possible.
The SiO.sub.2 film may be omitted if it is possible. Similarly, the
driver circuit (not shown) that drives the laser diode is formed
also on the common mount substrate.
[0630] As the metal package supporting the mount substrate and
acting also as the heat radiating member, a powder mold product of
Cu/W is used. For example, the molded of Cu/W product forming the
metal package may have the composition of 89W-11Cu and a linear
thermal expansion coefficient of 6.5.times.10.sup.-6/K, which is
close to the linear thermal expansion coefficient of the laser
diode. Further, the metal package has a thermal conductivity of 180
W/mK. By using such a powder mold product, a highly precision
product is obtained with low cost. Further, it is possible to form
heat radiation fins easily and efficient heat radiation becomes
possible.
[0631] In the present embodiment, an array of multimode optical
fibers in which the optical fibers each having a core diameter of
50 .mu.m are arranged with a pitch of 250 .mu.m is used. In the
temperature range of 0-70.degree. C., there occurs no change of
laser output and no lifetime degradation. Thus, excellent optical
telecommunication becomes possible. As the metal package functions
also as the heat radiating member, the number of parts can be
reduced and the efficiency of heat radiation is maximized.
[0632] A similar result is obtained also when a GaAs substrate or
AlN substrate is used in place of the Si substrate. In the case of
the AlN substrate or SiC substrate, they are insulating substrates
and the formation of the oxide film is not necessary. Otherwise,
the foregoing description applies and further explanation will be
omitted.
[0633] In the present embodiment, four laser diodes and four
optical fibers are used. However, the present invention is by no
means limited to this specific construction but the present
invention is applicable also to the case in which the number of the
laser diode is one and the number of the optical fiber is one, or
the number of the laser diodes is 8, 12, 16 and there are 8, 12, 16
optical fibers in correspondence thereto.
[0634] In the case of the multi-laser diode array chip in which a
plurality of laser diodes are formed on a common chip by exploiting
the feature of the surface-emission laser diode of the present
invention, a large number of laser diodes can be formed easily and
is ideal for large capacity telecommunication. On the other hand,
in the case of such a multi-laser array chip, heat dissipation of
the laser diodes becomes an important issue. By using the material
of the mount substrate and the heat radiating member as set forth
in the present embodiment, it becomes possible to operate the laser
diode array without causing problems. Thereby, a stable laser
output can be obtained.
[0635] In the present embodiment, a multimode optical fiber was
used also for the optical transmission path. However, this is not a
mandatory condition and it is possible to use an optical waveguide,
single mode optical fiber, plastic optical fiber, and the like.
Fourteenth Embodiment
[0636] Next a telecommunication system according to another
embodiment of the present invention that uses the long-wavelength
surface-emission laser diode of the present invention will be
explained with reference to FIG. 80. In FIG. 80, those parts
corresponding to the parts described previously are designated by
the same reference numerals and the description thereof will be
omitted.
[0637] Referring to FIG. 80, the telecommunication system that uses
the long-wavelength surface-emission layer diode includes a laser
diode module 141 carrying thereon the laser diode chip 32, a ferule
142A fixing the optical fiber 142, an adapter housing 143
accommodating the ferule 142A together with the optical fiber 142,
a bush 144 fixing the optical fiber 142, a housing 145
accommodating the bush 144, and the base 146 that holds the laser
module 141 and the housing 143 as an integral body. In the
construction of FIG. 63, the length of the optical fiber 142 from
the point A to the point B is set longer than the distance L (fixed
distance) between the end surface of the adapter housing 143 at the
side of the laser diode chip 32 and the end surface of the housing
145.
[0638] More particularly, in the case the fixed length L is set to
1 in FIG. 80, the length A-B of the optical fiber 142 is set to
1.05. Thereby, the optical fiber 142 is provided with a bend, and
the resilience associated with the bend causes the optical fiber
142 to be urged in the axial direction toward the optical emission
part 32A (left direction in the drawing). Thereby, the mechanical
connection is stabilized and an excellent optical coupling can be
realized. Here, a plastic optical fiber is used as the optical
fiber 142 but the present invention is by no means limited to such
a specific case and it is possible to use a quartz glass fiber
having a plastic coating.
[0639] FIG. 81 shows the case in which the length L of FIG. 80 is
set to 1 and the length A-B of the optical fiber is set to 2. Other
conditions are identical with the case of FIG. 80. In the present
embodiment, a larger resilient force is obtained as compared with
the case of FIG. 80 and the connection between the laser diode and
the optical fiber becomes more reliable.
[0640] FIG. 82 shows the cross-sectional view of the adapter
housing 143 used in the optical telecommunication system of FIG. 80
or 81.
[0641] Referring to FIG. 82, the adapter housing 143 includes a
tapered split sleeve 1431, a ferule 142A fixing the optical fiber
142, a coil spring 1432 of a shape-memory alloy exerting a force in
the axial direction, a collar 1433 cooperating with the coil spring
1432 and a plug housing 1434 accommodating therein the split sleeve
1431, the coil spring 1432 and the collar 1433, and a connecting
member 1435 for connecting the plug housing 1434 to the adapter
housing 143.
[0642] The coil spring 1432 exerts a resilient force at the
ordinary temperature or heated state determined by the initial
shape. Thus, with cooling, the coil spring may extend and urge the
ferule 142A in the axial direction thereof. Depending on the design
of the connecting member 1435, it is possible to memorize the shape
such the coil spring 1432 extends at the high temperature side.
[0643] FIG. 83 shows a modification of FIG. 82, wherein those parts
corresponding to the parts described previously are designated by
the same reference numerals and the description thereof will be
omitted.
[0644] Referring to FIG. 83, the coil spring 1432 of the
shape-memory alloy is replaced with a leaf spring 1432A of a
shape-memory alloy and the leaf spring exerts a resilient force at
the ordinary temperature or at the heated state determined by the
initial shape of the spring. Thus, the leaf spring 1432A may extend
with cooling and urge the optical fiber 142 in the axial direction
thereof. Depending on the design of the connecting member 1435, the
leaf spring 1432A may extend at the high temperature side.
[0645] It should be noted that the spring 1432 or 1432A is not
limited to a shape-memory alloy but may be formed of a shape-memory
plastic. In such a connection module, there exists a temperature
difference between the temperature in which the optical
telecommunication system is assembled (generally higher than the
room temperature) and the temperature (room temperature) in which
the system is used. Because of this temperature difference, the
shape-memory material changes the shape thereof and generates a
resilient force in the state that the member is assembled in a
structural body. It should be noted that the sense of temperature
change (between the state in which the system is assembled and the
state in which the system is used) may be reversed.
[0646] FIG. 84 shows another example of the telecommunication
system that uses the long-wavelength laser diode according to the
present invention, wherein those parts corresponding to the parts
described previously are designated by the same reference numerals
and the description thereof will be omitted.
[0647] Referring to FIG. 84, the telecommunication system of the
present embodiment is formed of a laser module 141 carrying the
laser diode chip 32, a ferule 142A for fixing the optical fiber
142, an adapter housing 143 accommodating the ferule 142A therein
together with the optical fiber 142, a fixing apparatus 147 holding
and fixing a part of the optical fiber 142, and abase 146 fixing
the adapter housing 143 and the fixing apparatus 147 and the laser
module 141 as an integral body. Thereby, the telecommunication
system of FIG. 67 provides a bend to the optical fiber 142 between
the adapter housing 143 and the optical fiber fixing apparatus
147.
[0648] The optical fiber fixing apparatus 147 is a device that uses
a frictional grip compression and includes a shape-memory alloy
compressional spring 1471, a collar 1472 cooperating with the
spring 1471, another collar 1473, a ball 1474 disposed between the
collars 1472 and 1473, and a housing 1475 having a conical surface
in contact with the ball 1474. In the housing 1475, the ferule 142B
holding the optical fiber 142 and the split sleeve 1476 are
accommodated in the state that the shape-memory alloy spring 1471
intervenes between the collar 1472 and the split sleeve 1476.
[0649] In the optical telecommunication system of such a
construction, the fiber 142 urges a force acting in the left and
right directions upon thermal deformation, while such an urging
force pushes the ball 1474 in the optical fiber fixing apparatus
147 in the right direction by the action of the compressional
spring 1471. Thus, the optical fiber 142 cannot move in the right
direction and exerts a force to the side of the adapter housing at
the opposite side.
[0650] In the example of FIG. 63, the optical fiber 142 is provided
with a bend and the optical fiber 142 urges the optical fiber 142
in the direction of the laser diode chip 32 as a result of the
resilient force associated with the bend of the optical fiber 142.
In the example of FIG. 84, the resilient force of the optical fiber
142 is boosted with the superimposed resilient force of the spring
147 of the shape memory alloy, and the optical fiber 142 is urged
more effectively to the direction of the laser diode chip 32.
[0651] FIG. 85 shows the optical connector formed of the tapered
split sleeve 1431, the ferule 142A holding the optical fiber 142,
the coil spring 1432 applying an axial urging force to the ferule
142A, the collar 1433, a plug housing 1434 accommodating therein
the split sleeve 1431, the coil spring 1432 and the collar 1433,
and the adapter housing accommodating the ferule 142B and the plug
housing 147, wherein the optical connector having such a
construction is connected with another optical connector by
connecting the respective connector housings 143 with each other by
way of a connecting screw 1436.
[0652] In the construction of FIG. 65, it should be noted that the
coil spring 1432 exerts an urging force at the normal temperature
as a result of the initial shape thereof. Thus, the coil spring
1432 is extended at the normal temperature change, and the ferule
142A is urged in the direction of the axis of the optical fiber
142. Thereby, two optical fibers are connected effectively. In this
example, too, it is possible to use various shape memory alloys and
shape memory plastics. In the example noted above, the present
invention was explained with reference to the example that uses a
shape-memory alloy for the spring. However, it is possible to use
an ordinary resilient material such as copper bronze or urethane
foam. By using such well-known art, it is possible to reduce the
cost of the module.
Fifteenth Embodiment
[0653] Next, a further embodiment of the present invention will be
explained.
[0654] FIG. 86 shows an example of the construction of the optical
transmission part that is provided with the optical
telecommunication system that uses the surface-emission laser diode
of the present invention. In FIG. 86, those parts explained
previously are designated by the same reference numerals and the
description thereof will be omitted.
[0655] FIG. 86 is referred to.
[0656] In the present embodiment, a connector substrate 151 holding
plural optical fibers 142 on the mount substrate 131 is provided
such that each of the plural surface-emission laser diodes 32
forming a part of the array opposes with a-corresponding one the
plural optical fibers 142. Further, plural photodetection devices
142P are provided on the connector substrate 151 so as to monitor
the optical power of the surface-emission laser diodes 32, such
that the photodetection devices 142P correspond respectively to the
plural optical fibers 142. Each of the photodetection device 142P
is provided as it face the corresponding surface-emission laser
diode 32 for detecting the optical beam emitted therefrom.
[0657] In FIG. 86, it can be seen that the laser diode chips 32,
the optical fibers 142 and the monitoring photodetectors 142P are
arranged to form a one-dimensional array. However, a similar effect
is achieved also when the photodetectors 142P are disposed in a
two-dimensional array. In this drawing, each of the laser diodes is
represented as an independent chip 32. However, the laser diodes
may be the one formed on a common chip 32.
[0658] In FIG. 86, illustration of the driver circuit of the
surface-emission laser diode 32 or the laser output control circuit
for feedback control of the laser output based on the signals
acquired from the driver circuit or monitoring photodetectorn
device is omitted for the sake of simplicity.
[0659] In the example of FIG. 86, the connector substrate 151 holds
the optical fiber 142 and is mounted on the mount substrate 131 by
using a connector guide 151A.
[0660] Each of the optical fibers is held in a penetrating hole
provided in the connector substrate surface 151 so that the end
surface of the optical fiber faces a corresponding laser diode
element 32A in the laser diode chip 32. Further, each laser diode
elements 32A and the corresponding optical fiber 142 are aligned
optically. Further, there is provided a monitoring photodetector
142P in the vicinity of the optical fiber insertion hole of the
connector substrate 151.
[0661] Here, the majority of the laser beam emitted from the
surface-emission laser-diode is injected into the optical fiber 142
aligned to the laser diode. On the other hand, the part of the
laser beam failed to enter the optical fiber 142 is detected by the
photodetection device 142P.
[0662] FIG. 87 shows the block diagram of the control system that
stabilizes the output of the surface-emission laser diode 32 by
using the monitoring photodetector 142P shown in FIG. 86.
[0663] Referring to FIG. 87, the surface-emission laser diode 32 is
driven by a laser drive circuit 161 to which the data signal
(electrical signal) is supplied, and there is formed a laser beam
as a result. A part of the laser beam thus formed is detected as
the leakage light, and the output electrical signal of the
photodetector 142P is supplied to a feedback control circuit 162.
The feedback control circuit 162 in turn controls the laser driver
circuit 161 so as to compensate for the change of the output
electrical signal of the photodetection device 142.
[0664] It should be noted that no strong optical power is needed
for the laser beam detected by the monitoring photodetector 142P,
as the monitoring photodetector merely detects the change of the
laser output for feedback control. From the viewpoint of long range
optical telecommunication, it is preferable that most of the
optical power is injected into the optical fiber, while the system
of using the feeble leakage light at the incident end of the
optical fiber 142 is sufficiently practical.
[0665] FIG. 88 shows another embodiment of the present
invention.
[0666] Referring to FIG. 88, the monitoring photodetector 142P is
formed between the surface emission laser diode chip 32 and the
connector substrate 151 in the vicinity of the incident end of the
optical fiber 142 facing the laser diode chip 32. Similar to FIG.
86, the laser diodes 32, the optical fibers 142 and the
photodetection devices 142P are arranged in a one-dimensional
array, while the present invention is not limited to such a
specific case but is applicable also to the case in which there is
formed a two-dimensional array on the mount substrate 131.
Similarly to the explanation above, each of the laser diodes is
represented as a chip, while it is also possible to form a number
of laser diode emission parts 32A on a single chip 32
monolithically as a laser diode element. In view of simplicity of
illustration, the driver circuit and the laser output control
circuit are not illustrated.
[0667] In the example of FIG. 88, the monitoring photodetector 142P
is provided between the substrate 131 and the connector substrate
151 in the form supported on a support member 142W, contrary to the
case of FIG. 86.
[0668] FIG. 89 shows a further embodiment of the present
invention.
[0669] Referring to FIG. 89, the monitoring photodetectors 142P are
disposed on the mount substrate 131 together with the
surface-emission laser diode chip 32 and detects a part of the
optical beam reflected by the connector substrate 131. In this
embodiment, too, it is possible to form a number of laser diode
elements 32A in a laser diode chip 32 and form the photodetectors
142P also on the chap 32 monolithically adjacent to the laser diode
elements 142P.
[0670] According to the latter construction, the laser diode
elements 32A and the photodetectors 142P are formed simultaneously
and with high precision. Further, the cost of assembling is
reduced. In FIG. 89, too, the illustration of the driver circuit of
the laser diode 32 or the output control circuit is omitted for the
sake of simplicity.
[0671] In the present embodiment, most of the laser beam emitted
from the surface-emission laser diode 32 is injected into the
optical fiber 142 coupled optically to the laser diode 32. Thereby,
the part of the laser beam failed to enter the optical fiber 142 is
reflected at the bottom surface of the connector substrate and
enters the photodetection device 142P. Thus, in the present
embodiment, too, it is possible to control the output of the laser
diode 32 by using the leakage light at the incident end of the
optical fiber 142.
[0672] Next, another embodiment of the present invention will be
described with reference to FIG. 90.
[0673] Referring to FIG. 90, it can be seen that the relationship
between the laser diode 32 and the monitoring photodetector 142P is
the same as in the case of FIG. 89 except that there is provided a
reflector 151R at the bottom surface of the connector substrate 151
for improving the efficiency of reflection.
[0674] Such a reflector 151R can be formed by providing a metal
film of Al, Ag or Au on the bottom surface of the connector
substrate 151 that faces the laser diode 32. In order to reflect
the optical beam of long wavelength, it is preferable to use a
metal film of Au.
[0675] In the present embodiment, too, it is possible to control
the laser output of the surface-emission laser diode by using a
leakage light which has failed to enter the optical fiber in the
vicinity of the optical fiber, similarly to the previous
embodiments. As a result of the use of reflector, the optical
output control is achieved with improved precision with weaker
power of the reflection light.
Sixteenth Embodiment
[0676] Next, a further embodiment will be described.
[0677] FIG. 91 shows an example of the optical telecommunication
system that uses a long-wavelength laser diode of the-present
invention. More specifically, FIG. 91 shows the construction in
which the optical fiber 72 extending out from the module package
71, in which the laser diode chip 32 is accommodated, is connected
to the optical fiber 73 of the optical telecommunication system. In
FIG. 91, those parts corresponding to the parts described
previously are designated by the same reference numerals and the
description thereof will be omitted.
[0678] Referring to FIG. 74, the optical fiber 72 drawn out from
the module package 71 is connected to the optical fiber 73 used for
optical telecommunication at the optical connector 171. When there
exists air at the connection part, there arises a reflection of the
optical beam in view of the difference of refractive index of the
air with respect to the refractive index of the optical fiber 72 or
73. In order to suppress such a fluctuation of refractive index a
the optical fiber connection part, there are proposals to inject a
refractive index matcher in the form of a gel so as to suppress the
fluctuation of the refractive index or to achieve a complete
attachment of the optical fibers (PC: physical contact). The
present invention uses this PC technology for the connection of the
optical fibers.
[0679] As represented in FIG. 91, the optical fibers 72 and the
optical fibers 73 are held in the connector 171 respectively by a
ferule 72F and a ferule 73F for preventing damaging of the optical
fibers. The ferules 72F and 73F are formed of a cylindrical body of
a zirconia ceramic that has a large mechanical strength. Thereby,
the optical fibers are inserted into the central hollow part, and
fixed therein by using an adhesive. Further, the ferules 72F and
73F are held in the optical connector 171 by a pair of split
sleeves 172.
[0680] FIG. 92 shows the relationship between the optical fibers 72
and 73, the ferules 72F and 73F and the split sleeve 172.
[0681] Referring to FIG. 92, the split sleeve 172 consists of a
pair of members formed of a copper bronze and having the shape of
being split from a hollow cylinder.
[0682] As represented in FIG. 92, a pair of split sleeve members
172 holds the ferules 72F and 73F, and the ferules 72F and 73F are
urged with each other by respective springs 173A and 173B with a
predetermined force. Thereby, there is formed the desired PC
connection. In the present embodiment, it is possible to construct
a rigid and highly reliable optical connection.
[0683] Generally, a laser diode is provided with a safety standard
for preventing damages to human body. In the case of the
surface-emission laser diode of 1.1-1.7 .mu.m, it is possible to
satisfy the safety standard while operating the laser diode with a
large output power as compared with the laser diode having the
oscillation wavelength in the visible wavelength band (0.4-0.7
.mu.m).
Seventeenth Embodiment
[0684] Next, another feature of the present invention will be
described of the coupling between the laser diode and optical fiber
or laser diode and optical waveguide.
[0685] As represented in FIG. 93, the laser beam of the surface
emission laser diode emitted perpendicularly to the substrate has a
optical intensity distribution generally close the Gauss
distribution profile in the plane perpendicular to the optical
axis.
[0686] Assuming that the beam diameter is equal to the FWHM
(full-width half--height) value of the beam profile, it is possible
to evaluate the emission angle .theta. of the optical beam from the
beam diameter and the distance between the laser diode and the
detection surface.
[0687] As a result, it turned out that the emission angle .theta.
of the surface-emission laser diode of the present invention is
generally symmetric about the optical axis and takes a value of
5-10 degrees.
[0688] In contrast to this, conventional laser diode of the
edge-emission type having the same oscillation wavelength of
1.1-1.7 .mu.m has a much larger emission angle .theta. and a very
large aspect ratio. Typically, the optical emission angle
.theta..perp. has a value of about 35 degrees in the direction
vertical to the plane of the substrate and the optical emission
angle .theta.//parallel to the place of the substrate takes the
value of about 25 degrees. Thus, it has been necessary to use an
optical element such as a microlens for realizing high efficiency
optical coupling.
[0689] In contrast to this, in the case of the surface-emission
laser diode of the present invention, the angle of optical emission
.theta. is very small as noted above, and the spreading of the
optical beam is minimized even in the case there is a large
distance between the laser diode and the optical fiber or between
the laser diode and the optical waveguide. Thereby, it becomes
possible to eliminate the microlens, while the removal of the
microlens provides an beneficial effect in that the distance
between the laser diode and the optical fiber is reduced. Further,
even in the case the distance between the laser diode and the
optical fiber or the laser diode and the optical waveguide is
large, the spreading of the optical beam is small and an efficient
optical coupling becomes possible without providing an intervening
lens.
[0690] FIG. 94 shows a schematic diagram with regard to spreading
of the laser bean in the laser diode 32 of the present invention
and FIG. 95 shows an example of calculation of beam spreading.
[0691] Referring to FIG. 94, the laser emission part 32A of the
laser diode 32 has an aperture size d [.mu.m] and the laser beam is
emitted from the laser emission part 32A with an emission angle
.theta.. Further, there is provided an edge surface of the optical
beam 180 at the position separated from the laser diode 32 with an
optical length 1 [.mu.m]. The optical fiber 180 includes a core 181
having a diameter X [.mu.m] and a cladding 182 surrounding the
core, and the optical fiber 180 is provided so as to align the
laser diode optically such that the optical axis of the optical
fiber 180 and the optical axis of the laser diode coincide with
each other. Here, it should be noted that the diameter d of the
aperture represents the diameter itself when the diameter has a
circular shape, while in the case the diameter has a square shape,
the diameter d represents the length of one edge. Further, when the
opening has other shape such as a rectangle, the diameter d
represents the diameter of a circle having an area of an ellipse
inscribing the rectangular opening.
[0692] The reason that the optical path length between the laser
diode and the optical fiber edge or optical waveguide edge is to
include the case of using a mirror for deflecting the optical beam
to be described later with reference to FIG. 97. When there is no
such a deflection of the optical path and the optical beam
propagates straight, the optical path length from the laser diode
to the optical fiber edge or the optical waveguide edge becomes
identical with the distance from the laser diode to the optical
fiber edge or the optical waveguide edge. In the case of an optical
waveguide, it is general that the core cross-section is not a
circle but a square or rectangle. Thus, in the case of a square
core, the edge size is set as X [.mu.m], while in the case of a
rectangular core, the size of the shorter edge is defined as X
[.mu.m].
[0693] In FIG. 94, it can be seen that the beam size of the optical
beam emitted from the laser diode perpendicularly to the epitaxial
layers therein, the beam diameter of the laser beam increases with
the optical path length L and the optical emission angle .theta.
until it reaches the edge surface of the optical fiber or optical
waveguide. Thereby, the beam size at the optical fiber edge or the
optical waveguide edge is given as d+21 tan(.theta./2).
[0694] Thus, when this beam size falls in the core diameter X of
the optical fiber or the optical waveguide, excellent optical
coupling is achieved.
[0695] Based on the above equation, FIG. 95 shows the relationship
between the optical path length 1 and the beam diameter for the
case of the aperture diameter of 5 .mu.m and the optical emission
angle of 10 degrees, in comparison with the conventional case of an
edge-emission type laser diode having an optical emission angle of
35 degrees.
[0696] Referring to FIG. 95, it can be seen that the spreading of
the beam diameter is substantially small as compared with the
conventional case and that it is possible to coincide the core
diameter X and the beam diameter in the case a commonly used
multimode optical fiber having a core diameter of 50 .mu.m
(cladding diameter of 125 .mu.m) is used for the optical fiber,
provided that the optical path length 1 is set to 260 .mu.m.
[0697] As long as the optical path length 1 is smaller than the
foregoing length in which the core diameter X coincide with the
beam diameter, it is possible to achieve an excellent optical
coupling with little optical loss. From the result of FIG. 95, it
is concluded that the optical alignment between the laser diode and
the optical fiber or optical waveguide is not critical in the axial
direction in the case of the laser diode of the present invention,
and a very coarse alignment is sufficient for achieving the
necessary optical coupling. In the case of a plastic optical fiber
having a core diameter of 100 .mu.m is used, the optical path
length 1 in which the laser beam diameter coincides with the core
diameter increases further to 550 .mu.m. Thereby, it is possible to
provide the optical fiber with a separation from the laser diode
package.
[0698] FIG. 96 shows an example of coupling an optical fiber of an
optical telecommunication system with the surface-emission laser
diode of the present invention. In FIG. 96, those parts
corresponding to the parts described previously are designated by
the same reference numerals and the description thereof will be
omitted.
[0699] Referring to FIG. 96, the telecommunication system includes
the mount substrate 131 carrying thereon the surface-emission laser
diode chip 32 in which the laser diodes 32A are formed as a
two-dimensional array, and the mount substrate 131 further carriers
a driver circuit of the laser diode not illustrated. Further, there
are provided a two-dimensional array of photodetectors (not shown
as well as the driver circuit not illustrated cooperating with the
photodetectors, wherein an optical fiber array including optical
fibers 191 is provided on the laser diode chip 32 such that each of
the optical fibers 191 face a corresponding laser diode 32A. It is
also possible to use an optical waveguide array in place of the
optical fiber array in the construction of FIG. 96. Further, the
optical telecommunication system of FIG. 96 may be a bi-directional
optical telecommunication system. In this case, the optical
transmission part and photodetection part are provided at both ends
of the optical fibers.
[0700] Referring to FIG. 96, the mount substrate 131 is formed of a
thermally conductive Si substrate and the laser array chip 32 is
mounted on the mount substrate 131. The laser array chip 32 may
contain the laser diode of FIG. 1 as a laser diode element
constituting the array. In the illustrated example, the laser
diodes having an oscillation wavelength of 1.3 .mu.m is used,
wherein each of the laser diodes 32A has an aperture having the
diameter of 10 .mu.m for optical output. Twelve laser diodes 32A
are formed on the mount substrate 131 with a pitch of 200
.mu.m.
[0701] On the mount substrate 131, there is provided a hole array
member 191 having a number of penetrating holes with a pitch of 200
.mu.m in correspondence to the pitch of the laser diodes 32A on the
chip 32, wherein each of the holes has a diameter of 125 .mu.m in
correspondence to the diameter of the multimode optical fibers. The
hole array member 191 may be formed of AlN and carries a guide 191X
in correspondence to a marker 32X provided on the laser diode chip
32. Thereby, the hole array member 191 is mounted on the laser
diode chip 32 such that the guide 191X coincides with the marker
32x. The hole array member 191 has an excellent thermal
conductivity of about 300 Wm.sup.-1K.sup.-1 and is used as an
effective heat radiation part when it is provide close to the laser
diode chip 32. Thereby, a stable optical telecommunication system
is realized by dissipating the heat associated with the laser
oscillation efficiently.
[0702] In each of the penetrating holes of the hole array member
191, a multimode optical fiber 192 is inserted and the optical
fiber is fixed therein by an adhesive in the state that the optical
fiber engages the laser diode chip 32. With such a construction, it
is possible to achieve an exact alignment between the laser diode
element 32A and the optical fiber 1691 in the direction
perpendicular to the optical axis direction.
[0703] As noted above, the optical fiber 191 is merely contacted
with the chip 32 and the accuracy of optical alignment in the axial
direction is poor. There can be a case in which a space of 0-20
.mu.m may be formed between the laser diode and the optical fiber.
However, because of the very small emission angle .theta. of the
laser diode of 8 degrees in the present example, and in view of the
large diameter d of 10 .mu.m for the laser diode 32A, the optical
beam diameter is only 17 .mu.m at the distance of 50 .mu.m from the
surface of the laser diode chip 32. Thus, the beam diameter is
still much smaller than the core diameter of 50 .mu.m for the
multimode optical fiber 192, and excellent optical coupling is
maintained. Further, there is no need of providing a coupling lens
between the laser diode 32A and the optical fiber 191.
[0704] In the experiment conducted on the construction of FIG. 96
in which the optical fiber 191 is pulled out gradually after being
contacted with the laser diode chip 32 by using a micrometer, it
was confirmed that the laser diameter becomes 66 .mu.m when the
distance between the laser diode and the optical fiber is 66 .mu.m.
In this case, the optical beam diameter is larger than the core
diameter of the multimode optical fiber 191 and the foregoing
relationship is failed. Thus, no sufficient optical coupling would
be achieved in this case.
[0705] Table 7 below represents the experiments conducted by the
inventor. TABLE-US-00008 TABLE 7 d + 21 tan (.theta./2) [.mu.m] X
(.mu.m) remarks 10 50 .largecircle. 20 50 .largecircle. 30 50
.largecircle. 40 50 .largecircle. 50 50 .largecircle. 60 50 .DELTA.
70 50 X 80 50 X 10 62.5 .largecircle. 20 62.5 .largecircle. 30 62.5
.largecircle. 40 62.5 .largecircle. 50 62.5 .largecircle. 60 62.5
.largecircle. 70 62.5 X 80 62.5 X .largecircle. good optical
coupling, sufficient for practical use .DELTA. slightly degraded
optical coupling, difficult for practical use X poor optical
coupling, cannot be used for practical applications
[0706] From the result of above, it can be seen that a practical
optical coupling cannot be achieved unless the core diameter X is
larger than the term given as d+21 tan(.theta./2), where .theta. is
the emission angle of the optical beam and 1 represents the optical
path length from the laser diode to the optical fiber, d represents
the diameter of the laser diode opening.
[0707] In the explanation above, it should be noted that the number
of the optical fibers is not limited to 12 but there may be only
one optical fiber or there may be 4, 8, 16 optical fibers. Further,
the present invention is not limited to a multimode fiber 191b but
a single mode optical fiber may also be used for long distance
transmission. Further, it is possible to use a plastic optical
fiber (POF), which is advantageous to construct a short distance,
low const optical telecommunication system. In such a system, it is
preferable to provide an anti-reflection coating on the edge
surface of the optical fiber. In the present embodiment, it is also
possible to use Si or C or alumina ceramics for the hole array
member 191. As the hole array member 191 functions also as a heat
radiation part, it is preferable to use a material having a high
thermal conductivity.
Eighteenth Embodiment
[0708] Next, another feature of the present invention will be
described of the coupling between the laser diode and optical fiber
or laser diode and optical waveguide.
[0709] The construction of the present invention is similar to that
of FIG. 96 except that the laser diode chip 32 and the optical
fiber are contacted with each other and the distance between the
laser diode chip 32 and the optical fiber 191 or optical waveguide
is set to almost zero. Here, the representation "contact" also
includes a separation of 0-10 .mu.m by taking into consideration of
the error at the time of the assembling process.
[0710] In the present embodiment, a Si substrate having good
thermal conductivity is used for the mount substrate 131 and the
laser array chip 32 including an array of surface-emission laser
diodes 32A is mounted on the mount substrate 131. The laser diode
32A used in the present embodiment has a laser oscillation length
of 1.3 .mu.m.
[0711] In the present embodiment, the diameter of the laser diode
is set to 10 .mu.m and 12 laser diodes 32A are formed with a pitch
of 200 .mu.m.
[0712] Next, the multimode optical fibers 192 are inserted into the
holes in the hole array member 191 having the penetrating holes of
125 .mu.m diameter with the corresponding pitch of 200 .mu.m.
Thereby, the tip ends of the optical fibers 192 are projected
slightly from the hole array member 191 such that the tip end of
the optical fiber makes an engagement with the laser array chip
32.
[0713] Next, the tip end part of the optical fiber is polished, and
the tip end of the optical fibers is aligned at the side of the
hole array member 191 that makes a contact with the laser diode
chip 32. After this, the marker X and the guide 191X are matched
and the optical alignment is achieved between each of the laser
diodes 32A and the corresponding optical fiber 192. The laser array
chip 32 and the hole array member 191 are then fixed in this state,
and the laser diode chip 32A and the optical fiber 192 are coupled
with each other.
[0714] According to the present embodiment, an optical coupling
better than the previous embodiment is achieved. Thus, in the
present embodiment, there occurs little spreading of the optical
beam and the laser beam diameter is more or less coincident with
the diameter d of the aperture forming the laser diode 32A. Thus,
the diameter of the laser beam is sufficiently smaller than the
core diameter of the optical fiber 191 and the alignment margin in
the axial direction is increased further. Thereby, it becomes
possible to construct an optical system using a 1.3 .mu.m surface
emission laser diode with low cost.
Nineteenth Embodiment
[0715] FIGS. 97A and 97B are diagrams showing the coupling of the
laser diode and an optical wave guide according to another
embodiment of the present invention, wherein FIG. 97A shows an
oblique view while FIG. 97B shows a cross-sectional view.
[0716] In the present embodiment, the laser beam emitted from a
laser diode 32A is deflected by a mirror 301. In FIGS. 80A and 80B,
those parts corresponding to the parts described previously are
designated by the same reference numerals and the description
thereof will be omitted.
[0717] Referring to FIGS. 97A and 97B, a Si substrate having an
excellent thermal conductivity is used for the mount substrate 131,
and the laser array chip 32 in which the laser diodes 32A are
arranged in the form of an array is mounted on the mount substrate
131. In the present embodiment, each of the laser diodes 32A has an
aperture with a diameter of 15 .mu.m, and four laser diodes 32A are
formed.
[0718] In this embodiment, another mount substrate 131A is provided
in which a mirror 301 is formed by mounting a reflecting member or
forming a reflecting layer monolithically. Further, an optical
waveguide 302 is formed on the mount substrate 131A. For example,
the mirror 301 may be formed by applying an anisotropic etching to
the Si substrate by using a KOH etchant and by providing an Ag film
on the crystal surface formed as a result of the anisotropic
etching.
[0719] In the present embodiment, an optical waveguide 302 is
formed on the mount substrate 131A. The optical waveguide 302 is
formed by depositing a lower cladding layer 302A, followed by
depositing a polymethacrylate (PMMA), followed further with a
patterning process to form a core pattern 302B. Typically, the core
pattern 302B has a cross sectional form of 50.times.50 .mu.m. Other
than PMMA, the optical waveguide layer may be formed also by
polyimide or epoxy resin or polyurethane or polyethylene. Further,
it is also possible to form an inorganic film such as silicon oxide
film. Further, these organic films may be formed by a spin-coating
process of dip-coating process in combination with a patterning
process. Further, it is possible to form the optical waveguide by a
resin molding process or molding process.
[0720] In the present embodiment, the mount substrate 131 and the
mount substrate 131A carrying the optical waveguide 302 are fixed
such that the optical axis of the laser diode 32A and the optical
axis of the optical waveguide 302 become coincident.
[0721] In the experiments conducted by the inventor, a laser diode
having an aperture of 15 .mu.m diameter and the optical emission
angle .theta. of 10 degrees is used, and the optical path length 1
between the laser diode 32A and the optical waveguide 302 is
changed by setting the position of the end surface of the optical
waveguide 302 to 50 .mu.m, 100 .mu.m and 250 .mu.m. As a result, it
was confirmed that an excellent result is obtained in the case the
optical path length is 50 .mu.m and 100 .mu.m, while no
satisfactory optical coupling was achieved in the case the optical
path length is set to 250 .mu.m.
[0722] Anyway, the present embodiment can eliminate the lens and
only a coarse optical alignment is sufficient for the axial
direction. Thus, by using a surface-emission laser diode of 1.2
.mu.m, it becomes possible to construct an optical alignment having
a tolerance with regard to the optical alignment in the axial
direction or the direction of the optical axis.
[0723] In the present embodiment, the mirror 301 is formed
separately to the optical waveguide 302. On the other hand, it is
possible to form a 45 degree surface at the end of the optical
waveguide 302 by a dicing blade, and the like, and by depositing a
metal film such as Ag on such an oblique surface. Thereby, a mirror
integral with the optical waveguide 302 is formed.
[0724] In the present embodiment, the diameter of the optical
waveguide is set to 50 .mu.m in correspondence to core diameter of
the multimode optical fiber. In the case a single mode optical
fiber is used, the diameter of the optical waveguide has to be set
to about 10 .mu.m. In this case, too, the same fundamental equation
holds. Further, it should be noted that the cross sectional form of
the core pattern 302B in the optical waveguide 302 is not limited
to the square pattern but a rectangular pattern may also be used.
Further, it is possible to transmit plural optical signals in a
single optical waveguide as in the case of an optical sheet.
Twentieth Embodiments
[0725] FIG. 98 shows the construction of coupling the laser diode
chip 32 with the optical waveguide 302 according to another
embodiment of the present invention.
[0726] Referring to FIG. 98, a Si substrate having excellent
thermal conductivity is used for the mount substrate 310 and the
laser diode chip 32 carrying the laser diodes 32A of the structure
of FIG. 1 is mounted on the substrate 310. In the present
embodiment, too, a surface-emission laser diode having an
oscillation wavelength of 1.3 .mu.m is used for the laser diode
32A. The laser diode used in the preset embodiment has the aperture
diameter d of 7 .mu.m and the optical emission angle .theta. of 8
degrees. In the present embodiment, the laser diode 32A is used as
a single device, not in the form of array.
[0727] The mount substrate 310 carrying thereon the laser diode
chip 32 is fixed on a package body 311 with positioning. Further,
the optical axis of the laser diode 32 is aligned with the optical
axis of the single mode optical fiber 313 held on an optical fiber
guide 312, by aligning the optical fiber guide 312 and the package
311 in advance by using a guide 311A on the package body 311 and a
guide pin 312A on the optical guide 312.
[0728] The single mode optical fiber 313 includes a core 313A
having a diameter of 10 .mu.m and a clad having a diameter of 125
.mu.m and surrounding the core 313A, wherein the diameter of the
optical fiber guide 312 is set so as to coincide with the outer
diameter of the cladding of the optical fiber 313.
[0729] In the construction of the present embodiment, an excellent
optical coupling is achieved by contacting the single mode optical
fiber 313 to the laser diode 32. Further, the present invention
enables a broadband optical signal transmission by using the single
mode optical fiber 313 over a long distance. Thus, according to the
present embodiment, it became possible to construct an optical
telecommunication system having excellent optical coupling by using
the surface-emission laser diode of the 1.3 .mu.m band.
Twenty-First Embodiment
[0730] Next, another embodiment of the present invention will be
described.
[0731] FIG. 99 shows an example of the optical telecommunication
system that uses a long-wavelength laser diode according to another
embodiment of the present invention.
[0732] Referring to FIG. 99, the optical transmission system
includes a surface-emission laser diode 32 having the emission part
32A and an optical fiber 352, wherein the laser diode 351 and the
optical fiber 352 are coupled directly at the laser emission part
32A. Of course, it is possible to achieve the optical coupling by
using a lens.
[0733] Assuming that the laser emission part 32A of the
surface-emission laser diode 32 is d and the core diameter of the
optical fiber 312 as F, it can be seen that the laser beam from the
laser diode 32 diverges as represented in FIG. 99 by a broken line
351B. In FIG. 99, it should be noted that d represents the diameter
of the circle inscribing the polygonal laser emission part 32A. In
this case, the emission angle of the laser beam is much smaller
than the case of a conventional edge-emission laser diode. Thus, in
the case the laser diode 32 is provided near the optical fiber, it
should be noted that a high optical coupling can be achieved when
the optical axis of the laser diode is perpendicular to the optical
fiber end surface and is coincident with the optical axis of the
laser diode, and the aperture diameter of the laser diode is equal
to or smaller than the core diameter.
[0734] As noted before, the surface-emission laser diode used in
the present embodiment operates in the wavelength band of 1.1-1.7
.mu.m, and a long distance transmission becomes possible by using
the wavelength band of 1.3-1.55 .mu.m due to the small optical loss
of the quartz optical fiber in this wavelength band.
[0735] Assuming that the laser emission part 32A of the laser diode
32 has the diameter of d (in the case the laser emission part 32A
has a polygonal shape, the diameter d is the diameter of a circle
inscribing the polygonal laser emission part 32A, it is possible to
increase the optical coupling efficiency by setting the foregoing
parameters F and d so as to satisfy the relationship of
F/d.ltoreq.2 (2)
[0736] In the embodiment of FIG. 99, it should be noted that the
laser emission part 32A and the optical fiber 352 are coupled
directly and no lens is interposed.
[0737] In the conventional optical system that uses an edge
emission type laser diode of the 1.3 .mu.m band, there is been a
poor optical coupling efficiency between the laser diode and the
optical fiber, and it has been difficult to couple these directly.
Further, the return beam incident to the laser dude from the
optical beam in the backward direction has caused a change
oscillation state of the laser diode.
[0738] In the case of the present invention, the laser diode is a
surface-emission laser diode and operates in the long-wavelength
band. Thus, the laser diode has a generally circular beam shape
characterized by excellent aspect ratio close to 1. Thereby, the
efficiency of the laser diode is improved substantially as compared
with the case of the edge-emission type laser diode.
[0739] Further, the laser diode of the present invention has the
reflector of very high reflectance and the effect of the return
beam is successfully suppressed. Thereby, it becomes possible to
couple the laser diode and optical fiber directly, and the optical
isolator used in the conventional optical telecommunication system
can be eliminated.
[0740] Designating the diameter of the laser emission part 32A as d
and the optical fiber core diameter as F, it is possible to
increase the optical coupling efficiency by setting the parameters
d and F so as to satisfy the relationship 0.5.ltoreq.F/d.ltoreq.2
(3)
[0741] Hereinafter, the reason of this will be explained with
reference to Table 8, which summarizes the result of the study of
the inventor of the present invention on the optical coupling loss.
TABLE-US-00009 TABLE 8 F/d coupling loss (dB) performance 0.4 >5
X 0.5 3-5 .largecircle. 0.65 1 .largecircle. 2.0 3-5 .largecircle.
2.2 >5 X .largecircle. good X poor
[0742] Referring to Table 8, it can be seen that the efficiency of
optical coupling is decreased when the diameter d of the beam
emission part 32A has exceeded the core diameter F. On the other
hand, it can be seen also that as long as there is a relationship
d.ltoreq.2F (in other words 0.5.ltoreq.F/d), the optical coupling
loss can be suppressed within 3-5 dB. In the case the diameter d of
the beam emission part 32A is smaller than the core diameter, a
highly efficient optical coupling is achieved in the case the
emission angle of the laser diode, determined by the diameter of
the circle inscribing the beam emission part 32A and the
wavelength, is smaller than the numerical aperture NA for allowing
a single mode optical coupling to a single mode optical fiber.
[0743] For example, in the case the core diameter is set to 10
.mu.m and the core refractive index is 1.4469 and the clad
refractive index if 1.4435, the numerical aperture NA that enables
single mode optical coupling becomes 0.0995 at the wavelength of
1.3 .mu.m. The diameter d of the beam emission part 32A
corresponding to this is about 6.5 .mu.m.
[0744] On the other hand, even in the case the diameter of the beam
emission part 32A is less than 6.5 .mu.m, it is possible to
suppress the optical coupling loss to the level of 3-5 dB when the
diameter of the beam emission part 32A is about 5 .mu.m. From this,
it is concluded that it is preferable to choose the parameters F
and d to satisfy the relationship 0.5.ltoreq.F/d.ltoreq.2.
[0745] While the foregoing arguments have been for the case of
using a single mode optical fiber, the foregoing condition is
applicable also to the case of using a multimode optical fiber
coupling a single mode optical fiber via a tapered optical
waveguide.
[0746] Next, another example of the optical telecommunication
system of the present invention will be described with reference to
FIG. 100.
[0747] Referring to FIG. 100, the present embodiment uses a
coupling lens 353 between the surface-emission laser diode 32 and
the single-mode optical fiber. The lens 353 may be formed of a
single lens or a lens system in which a plurality of lenses are
combined. In the case of constructing the lens 353 by a single
lens, it is preferable to dispose the lens 353 close to the beam
emission part 32A of the laser diode chip 32.
[0748] In the construction of FIG. 100, it becomes possible to
suppress the optical coupling loss effectively even in the case the
diameter d of the beam emission part 32A is set larger than the
core diameter of the optical fiber 353, by appropriately choosing
the lens power of the coupling lens 353.
[0749] In the case the core diameter is 10 .mu.m and the diameter d
of the emission part 32A is 20 .mu.m, for example, the beam size is
reduced by 1/2 by the lens 353. Thus, designating the focal
distance of the lens 353 as f, the laser wavelength as .lamda., the
radius of the beam emission part 32A as .omega.0 (=10 .mu.m) and
the refractive index as n, the focal distance f is obtained by the
relationship .lamda. .times. .times. f .pi..omega. 0 2 .times. n 1
+ [ 1 + .lamda. .times. .times. f .pi..omega. 0 2 .times. n ] 2 = 5
10 ##EQU1##
[0750] In this case, the focal distance f becomes about 140
.mu.m.
[0751] In the case of using a coupling lens, it is possible to
achieve an efficient optical coupling when there holds a
relationship between the parameters d and f noted before as
d.gtoreq.f, or F/d.ltoreq.1. (4)
[0752] Further reference should be made to Table 9 below.
TABLE-US-00010 TABLE 9 F/d coupling loss (dB) performance 1.2 >5
X 1.0 3-5 .largecircle. 0.8 1 .largecircle.
[0753] In the embodiment of FIG. 100, the coupling lens 353 was
provided by a single lens. On the other hand, it is possible to
construct the lens 353 by a plurality of lenses. For example, as
represented in FIG. 101, the lens 353 may be formed of two lenses
353A and 353B.
[0754] Referring to FIG. 101, the diverging beam from the beam
emission part 32A is focused by the first lens 353A to the second
lens 353B, and there is formed a beam waist at the plane of the
second lens 353B. The lens 353B functions as the lens 353 of FIG.
100. According to the present embodiment, it becomes possible to
maintain a high coupling efficiency in the case a single lens
cannot be disposed near the beam emission part of the laser diode
or in the case the divergence of the beam is excessive and the
optical coupling efficiency is decreased, by constructing the lens
353 by plural lenses.
[0755] In the description heretofore, it was assumed that a single
mode optical fiber is used for the optical fiber 352. On the other
hand a similar effect is achieved also in the case of using a
multimode optical fiber or a tapered optical waveguide, as long as
the relationship of Equation (4) is satisfied.
[0756] In the construction that uses a conventional edge-emission
laser diode, there is a possibility that the laser oscillation is
influenced by the return beam from the coupling lens, and it has
been necessary to provide an optical isolator. In the case of the
surface-emission laser diode, there is provided a reflector
structure of high reflectance and the effect of the return beam is
successfully eliminated. Thus, it becomes possible to eliminate the
optical isolator.
[0757] Next, another embodiment constructed by a surface-emission
laser diode and the optical fiber will be explained with reference
to FIG. 102.
[0758] Referring to FIG. 102, the laser diode chip 32 includes a
plurality of beam emission parts 32A arranged in the form of array.
Of course, it is possible to dispose the laser diode chips 32
themselves in the form of array. Further, a plurality of laser
diode chips may be arranged in the form of array.
[0759] In FIG. 102, the optical fiber is coupled directly to the
corresponding optical emission part 32A, and thus, it is possible
to satisfy the Equation (3) explained before with reference to FIG.
99. Thereby, it becomes possible to construct an optical
telecommunication system of large capacity.
[0760] FIG. 103 is an embodiment in which a coupling lens is
provided further to the construction of FIG. 102. In the
construction of FIG. 102, too, the relationship between the beam
emission part 32A, the lens array and the optical fiber 352 is the
same as in the case of FIG. 102. Thus, by satisfying the Equation
(4) by choosing the parameters d and F, an efficient optical
coupling becomes possible. As the beam emission part 32A is
arranged in the form of array, an optical telecommunication system
of large capacity is realized. In the present embodiment, it is
possible to combine a plurality of lenses in each of the coupling
lens 355.
Twenty-Second Embodiment
[0761] Next, another embodiment of the present invention will be
described.
[0762] FIGS. 104A and 104B show an example of the optical
telecommunication system that uses a long-wavelength laser diode of
the present invention as an optical source and shows the state in
which the laser diode chip 32 carrying the beam emission part 32A
is connected with a laser driver IC 32D including a CMOS circuit,
wherein FIG. 104A shows a side view while FIG. 104B shows a plan
view.
[0763] Referring to FIGS. 104Aa and 104N, the laser diode chip 32
and the laser driver IC 32D are mounted on a conductive submount
401 by a conductive adhesive. Further, the laser diode chip and the
laser driver IC 32D are connected by a high-frequency transmission
line 402 (microstrip line in the present embodiment), and each chip
32 is connected to this microstrip line 402 by way of a bonding
wire 403.
[0764] In the case of the surface-emission laser diode for optical
telecommunication, it is necessary to carry out a very fast
modulation in the order of several hundred MHz to several GHz.
Thus, it is preferable to decrease the distance between the laser
diode chip 32 and the laser driver IC 32D. On the other hand, there
can be a case in which such a construction is not possible because
of the layout of the optical system. When an ordinary wiring is
used in such a case for connecting the laser diode chip 32 and the
laser driver IC 32D, however, there arises a problem of
electromagnetic emission from such wiring. In the present
embodiment, the laser diode chip 32 and the laser driver IC 32D are
connected such that there occurs no such electromagnetic
radiation.
[0765] By using such a construction, no electromagnetic emission
occurs from the wiring part and the need of providing special
shielding means is eliminated. Thereby, the cost of the optical
telecommunication system is reduced.
[0766] In the present embodiment, a microstrip line was used for
the high-frequency transmission line 402. However, any of
unbalanced transmission line such as coplanar transmission line or
triplate line may be used. Further, the connection between the
laser diode chip 32 and the laser driver IC 32D is not limited to
the bonding wired but flip-chip wiring or TAB bonding or microbump
bondin may be used.
Twenty-Third Embodiment
[0767] In the embodiment of FIG. 52, a system that uses a
surface-emission laser diode of long-wavelength band of 1.1-1.7
.mu.m has been explained. Conventionally, an optical
telecommunication using the wavelength band of 0.85 .mu.m band is
studied. However, in this wavelength band, the optical transmission
loss is large in the optical fiber and practical use of such a
system was not successful. Further, there has been no stable laser
diode device in the long wavelength band in which the optical
transmission loss is small. In the present invention, on the other
hand, it became possible to construct a surface-emission laser
diode operable in the wavelength band of 1.1-1.7 .mu.m by improving
the semiconductor distributed-Bragg reflectors 12 and 18 and by
providing the non-optical recombination elimination layers 13 and
17, such that the laser diode operates stably at low power.
Thereby, a practical optical telecommunication system has become
the matter of reality.
[0768] In the example of FIG. 52, the optical telecommunication
system is formed of the long-wavelength surface-emission laser
diode chip, the first optical fiber FG1 injected with the laser
beam emitted from the beam emission part 32A in the chip 32 and
acting as an optical transmission path, a second optical fiber FG2
injected with the laser beam transmitted through the first optical
fiber FG1 and functioning as an optical transmission path, a third
optical fiber FG3 injected with the laser beam transmitted through
the optical fiber FG2 and functioning as an optical waveguide, and
a photodetector chip 34 including a photodetection part 34A coupled
optically to the optical fiber FG3 for detecting the laser beam
transmitted through the optical fiber FG3.
[0769] Further, the optical connection module MG1 is provided
between the laser diode chip 32 and the first optical fiber FG1,
while the optical connection module MG2 is provided between the
optical fiber FG1 and FG2. Similarly, the optical connection module
MG3 and FG4 are provided between the optical fibers FG2 and FG3 and
between the optical fiber FG3 and the photodetector 34.
[0770] In the example of FIG. 52, the first optical fiber FG1 is
coupled directly to the beam emission part 32A of the laser diode
chip 34 without intervening an optical system.
[0771] With reference to FIG. 94, a description was made on the
relationship that is required between the laser diode of the
present invention and the optical fiber.
[0772] In the long-wavelength surface-emission laser diode of the
present invention, the optical emission angle of the laser beam is
about 10 degrees, and the laser diode is used in combination with
an optical fiber which may have a core diameter of 50 .mu.m (clad
diameter of 125 .mu.m).
[0773] In such a combination of the laser diode and the optical
fiber, the beam emission part 32A may have a size of 0.005
mm.times.0.005 mm-0.002 mm.times.0.002 mm. In terms of area S,
these are represented as 0.000025 mm.sup.2-0.0001 mm.sup.2. In this
case, it is possible to achieve excellent optical coupling without
the need of providing an intervening optical system. Further, the
same result holds also in the case an optical waveguide is used in
place of an optical fiber.
[0774] In the present invention, the beam emission part 32 is tuned
to the wavelength of 1.1-1.7 .mu.m, while the foregoing wavelength
band is related to the operational voltage of the laser diode.
[0775] Thus, in the present embodiment, a detailed analysis was
conducted on the operational voltage of the laser diode 32 suitable
for realizing continues oscillation, not just a pulse
oscillation.
[0776] Table 10 shows the experimental results of the present
embodiments. TABLE-US-00011 TABLE 10 emission size area S No. (mm
.times. mm) (mm2) voltage V V/s remarks 1 0.005 .times. 0.005
0.000025 0.2 8000 X 2 0.005 .times. 0.005 0.000025 0.3 12000 X 3
0.005 .times. 0.005 0.000025 0.375 15000 .largecircle. 4 0.005
.times. 0.005 0.000025 0.4 16000 .largecircle. 5 0.005 .times.
0.005 0.000025 0.5 20000 .largecircle. 6 0.005 .times. 0.005
0.000025 0.6 24000 .largecircle. 7 0.005 .times. 0.005 0.000025
0.75 30000 .largecircle. 8 0.005 .times. 0.005 0.000025 0.9 36000 X
9 0.005 .times. 0.005 0.000025 1.2 48000 X 10 0.01 .times. 0.01
0.0001 1.2 12000 X 11 0.01 .times. 0.01 0.0001 1.3 13000 X 12 0.01
.times. 0.01 0.0001 1.5 15000 .largecircle. 13 0.01 .times. 0.01
0.0001 1.7 17000 .largecircle. 14 0.01 .times. 0.01 0.0001 1.9
19000 .largecircle. 15 0.01 .times. 0.01 0.0001 2.1 21000
.largecircle. 16 0.01 .times. 0.01 0.0001 2.3 23000 .largecircle.
17 0.01 .times. 0.01 0.0001 2.5 25000 .largecircle. 18 0.01 .times.
0.01 0.0001 3 30000 .largecircle. 19 0.01 .times. 0.01 0.0001 4
40000 X 20 0.01 .times. 0.01 0.0005 5 50000 X 21 0.02 .times. 0.02
0.0004 2 5000 X 22 0.02 .times. 0.02 0.0004 4 10000 X 23 0.02
.times. 0.02 0.0004 6 15000 .largecircle. 24 0.02 .times. 0.02
0.00041 8 20000 .largecircle. 25 0.02 .times. 0.02 0.0004 10 25000
.largecircle. 26 0.02 .times. 0.02 0.0000 12 30000 .largecircle. 27
0.02 .times. 0.02 0.00004 15 37500 X 28 0.02 .times. 0.02 0.00004
20 50000 X .largecircle. good oscillation X no oscillation or
damage
[0777] From the results above, it can be seen that, by choosing the
operational voltage S such that the ratio of the operational
voltage V to the area S of the beam emission part 32A of the laser
diode (V/S) falls in the range of 15000-3000, it can be possible to
drive the laser diode without causing a damage in the laser diode
and the laser diode can be operated continuously. By choosing the
drive conduction as such, it is possible to realize a stable
optical transmission system having along lifetime.
Twenty-Fourth Embodiment
[0778] Next, a further embodiment of the present invention will be
explained.
[0779] FIG. 105 shows the construction of an optical
telecommunication system using the long-wavelength-laser diode
according to the present invention, wherein those parts
corresponding to the parts described previously are designated by
the same reference numerals and the description thereof will be
omitted.
[0780] Referring to FIG. 105, the optical telecommunication system
of the present invention has a construction to divide the optical
output of the laser diode 32A provided on a mount substrate 301
into an optical beam for optical telecommunication and a monitoring
beam by using the mirror 301 provided on the mount substrate 301A,
wherein the monitoring beam thus divided is detected by the
monitoring photodetector 142P provided on the mount substrate 301A.
The monitoring photodetector 142P corresponds to the monitoring
photodetector 142P explained previously with reference to FIG. 86.
In FIG. 105, as well as in FIGS. 107 and 108 to be described laser,
there are formed a laser diode array by the laser diodes arranged
in the direction perpendicular to the plane of the drawing. In
correspondence to this, there are provided a plurality of optical
waveguides.
[0781] The optical telecommunication system of the present
embodiment is formed of an optical transmission part including the
surface-emission laser diode 32 and the driver circuit thereof, the
photodetection part including the planar photodetection unit and a
drive circuit thereof, and an optical fiber or optical waveguide
providing an optical transmission path extending between the
optical transmission part and the optical reception part. Although
the driver circuit of the laser diode 32 or photodetector 34 is not
shown, these can be formed on the substrate of the respective
devices. Alternatively, these can be formed integrally with the
laser diode device in the device fabrication process. By providing
such an optical transmission part and optical reception part at
both ends of the optical transmission path, it becomes possible to
realize a bi-directional optical telecommunication system.
[0782] As represented in FIG. 105, the laser beam emitted from a
surface of the long-wavelength laser diode is divided by the mirror
301 and one of the optical beams thus divided is then directed to
the optical waveguide 302 provided with an optical alignment. This
optical waveguide 302 may be an optical fiber.
[0783] On the other hand, the other optical beam divided by the
mirror 301 is directed to the monitoring photodetection device 34
provided on the mount substrate 301A. Thus, the mirror 301 is used
to divide out the laser beam to be supplied to the monitoring
photodetector 34 and it is preferable that such a mentoring optical
beam is as weak as possible within the range that a proper control
of the laser diode is possible by using such a monitoring device.
In view of the power consumption of the optical telecommunication
system, it is preferable to provide as much energy as possible to
the optical waveguide 302 as an optical signal.
[0784] Thus, in the present invention, the thickness of the metal
film of Au or Ag or Al used for the mirror 301 such that the
transmittance of the mirror 301 is controlled for the wavelength of
1.1-1.7 .mu.m. Further, it is also possible to provide openings of
various shaped in the form of grooves, circles, squares, and the
like, so as to control the transmittance of the mirror 301.
Thereby, in order to avoid the effect of unwanted optical
interference, it is preferable to change the pitch or size or
location of the openings at random. Further, it is possible control
the transmittance of the mirror 301 by using a dielectric
multilayer mirror or semiconductor layered mirror.
[0785] As compared with the conventional laser diode of
edge-emission type, the surface-emission laser diode used in the
optical telecommunication system of the present invention has an
advantageous feature of small output dependence on temperature and
little degradation with time. While the laser diode of the present
invention has such advantageous features, it is nevertheless
preferable to control the output thereof by way of feedback
control.
[0786] In the case of a conventional edge-emission laser diode,
such a monitoring of the optical power of the laser diode was made
easily by monitoring the laser beam emitted from the rear edge
surface of the laser diode. In the case of the surface-emission
laser diode, the laser output is obtained only at one side of the
laser diode. Thus, the construction of monitoring the laser beam
emitted in the backward direction cannot be used in the present
invention. Further, the laser diode ahs a relatively narrow angle
of optical emission of about 10 degrees. Thus, while there is a
possibility of provide the laser diode close to the optical fiber
or optical waveguide, there is no room at all for inserting a
monitoring photodetector between the laser diode and the optical
fiber.
[0787] On the other hand, the present invention provides an
effective an simple means of obtaining a monitoring laser diode by
providing the mirror 301 as noted above. Thereby, the output of the
laser diode is monitored with out increasing the optical path, and
the output of the laser diode is controlled positively and exactly.
As the present invention uses the mirror 301 for deflecting the
output optical beam of the laser diode, the optical axis of the
reflected signal optical beam becomes parallel to the optical axis
of the optical fiber, and thus, it becomes possible to fix the
optical fiber or optical waveguide on a plane parallel to the plane
of the optical module and the optical waveguide 302 or optical
fiber is fixed easily. Thereby, the obtained structure has an
improved rigidity. Further, it is also possible to provide a lens
before or after the mirror 301.
[0788] Although not illustrated, it is possible to use a plurality
of laser diodes and corresponding optical waveguides or optical
fibers. In this case, the laser diodes are arranged in a direction
perpendicular to the sheet of the drawing, and the mirror 301
extending also in the direction perpendicular to the sheet of the
drawing is used commonly by the laser diodes forming the array.
[0789] FIG. 106 shows the block diagram of the feedback circuit
that controls the output of the laser diode 32 by using the divided
laser beam.
[0790] Referring to FIG. 106, the laser diode 32 is driven by a
driver circuit 161 corresponding the driver circuit 161 explained
previously with reference to FIG. 87, and a part of the laser beam
produced by the laser diode 32 is provided to the monitoring
photodetector 142P. The monitoring photodetector 142P detects the
output of the laser beam thus supplied and produces an electric
output current indicative of the output of the laser diode 32.
Further, the laser control part 162 controls the laser diode 32
such that the output of the laser diode is maintained constant.
[0791] In the construction of FIG. 106, it should be noted that the
laser diode 32 can be driven with a low drive voltage. Thereby,
ordinary CMOS circuit can be used for the driver circuit 161.
Thereby, power consumption of the laser diode is reduced. As the
wavelength band to be detected by the monitoring photodetector 142P
is in the range of 1.1-1.7 .mu.m, it is possible to use a
photodiode of the InGaAs system. Further, in view of the gradual
change of the output of the laser diode, it is not necessary that
the photodetector 142P has a high response speed. Thus, it is
possible to use a high-sensitivity photodetector while sacrificing
the response speed.
[0792] Referring back to FIG. 106, there is formed a laser diode
array including four laser diodes 32A each having the construction
of FIG. 1, and the laser diode array is mounted on the mount
substrate 301 of Si together with the drive circuit and the laser
control circuit not illustrated. The laser diodes 32A has the
wavelength of 1.3 .mu.m and may be disposed on a single chip 32
with the pitch of 200 .mu.m.
[0793] Next, the mirror 301 is formed by using a heat conductive Si
substrate, which is transparent to the wavelength of 1.3 .mu.m. By
using the Si substrate for the mount substrate 301A, it is possible
to form the mirror easily by conducting an anisotropic etching
process. Thereby, a crystal surface determined with respect to the
principal surface of the substrate 301A is formed with exact angle.
The etching process may be conducted by using an etchant such as
KOH. With this, a mirror surface having a 45 degrees is formed.
Further, an Au deposition is conducted on the mirror surface 301,
and the optical waveguide 302 is formed. In such a process, it is
possible to control the transmittance of the mirror 301 by
controlling the thickens of the Au film.
[0794] In the formation of the optical waveguide 302, it should be
noted that a PMMA film is formed as the core layer after the
cladding layer 302A is formed. By patterning the core layer thus
formed, the core pattern 302B is formed. Further, the upper
cladding layer 302C is formed so as to cover the core layer
302B.
[0795] In the present embodiment, the core pattern 302B is
patterned so as to form a cross section of 50.times.50 .mu.m. The
optical waveguide layer 302 thus formed is coupled with the optical
fiber not illustrated, and long distance optical telecommunication
system is constructed.
[0796] For the optical waveguide layer 302, it is possible to use
various resins such as polyimide, epoxy resin, polyurethane or
polyethylene, in addition to PMMA. Further, it is possible to
provide an inorganic film such as silicon oxide film. Further, the
formation of the optical waveguide layer 302 can be made by
combining the spin coating process or dip coating process and a
patterning process. Alternatively, it is possible to form the
optical waveguide by a resin molding process or molding
process.
[0797] Further, the optical axis of the laser diode is set
coincident with the optical axis of the optical waveguide, and the
mount substrates 301 and 301A are fixed. Further, a planar
photodiode is mounted on the mount substrate 301A in optical
alignment with the laser beam divided by the mirror 301 as the
monitoring photodetector 142P. A photodiode having an optical
absorption layer of InGaAs on the InP substrate may be used of this
purpose.
[0798] Further, the output of the photodetection device 142P is
connected to the laser control unit 162 by a wire bonding process.
With this, the feedback control circuit explained with reference to
FIG. 106 is obtained.
[0799] Table 11 below shows the result of evaluation of the optical
transmission unit conducted by changing the external temperature.
TABLE-US-00012 TABLE 11 mirror transmissivity(%) Remarks 0.1 x 0.2
x 0.3 x 0.5 .DELTA. 0.7 .DELTA. 1.0 .smallcircle. 2.0 .smallcircle.
5.0 .smallcircle. 10 .smallcircle. 20 .smallcircle. 30
.smallcircle. 40 .smallcircle. 50 .smallcircle. 60 .DELTA. 70 x
.smallcircle. good .DELTA. fair x poor
[0800] In the experiment, the temperature is changed from
0-70.degree. C. with the step of 10.degree. C., wherein Table 11
shows only the result of 20.degree. C., as the results for other
temperatures were more or less the same as the result of 20.degree.
C.
[0801] Referring to Table 11, it can be sent that an optical power
of 10 .mu.W is detected by the monitoring photodetector 142P when
the mirror transmissivity is less than 1% for the optical power of
mW level, which is used commonly for optical telecommunication. As
the change of the optical power of the laser diode is smaller than
this, no sufficient optical energy needed for the control of the
laser diode is supplied to the photodetection device, and as a
result, there is a fluctuation of optical output in the laser
diode.
[0802] When the transmissivity exceeds 50%, on the other hand, too
much energy of the laser output is used for controlling the output
power of the laser diode. Thus, there occurs decrease of efficiency
of the optical telecommunication system.
[0803] From the result of Table 11, it can se seen that the
transmissivity of 2% or more but 30% or less is preferable for the
mirror 301. By using such a construction, it is possible to realize
an optical transmission unit capable of controlling the laser
output stably. Thereby, it is preferable that the transmissivity of
the mirror 301 is in the range of 1% or more but 50% or less, more
preferably 2% or more but 30% or less.
[0804] In the present embodiment, 4 laser diodes were used in the
form of array. Of course, the number of the laser diode may be one,
or the laser diode may be used in the form of array including 8,
12, 16 or more laser elements.
[0805] It is also possible to use an optical fiber for transmitting
the optical signals in place of the optical waveguide 301. When
transmitting large amount of information over a long distance, the
use of single mode optical fiber is suited. In the case of
transmitting information for short distance with low cost, the use
of plastic optical fiber (POF) is suitable. Further, there may be a
possibility of using a multimode optical fiber in the intermediate
applications.
[0806] FIG. 90 shows another embodiment of the present invention in
which the electrode of the photodetection device is used for the
mirror.
[0807] In the present embodiment, the long-wavelength laser diode
32 of FIG. 1 is mounted on the Si mount substrate together with a
driver circuit and a laser control circuit not illustrated. In the
present embodiment, a laser diode having a laser oscillation
wavelength of 1.2 .mu.m is used.
[0808] Similarly to the previous embodiments, a photodetector using
a GaAsP material is used for the monitoring photodetector 142. In
the present embodiment, the p-type electrode of the optical
detection surface is used for the mirror 301.
[0809] More specifically, an Au film having a thickness of 300 nm
and not allowing passage of the 1.2 .mu.m wavelength radiation is
provided on the photodiode as the electrode, and circular openings
having various diameters in the range of 0.7-5 .mu.m are formed.
With this, a transmissivity of 5% is realized.
[0810] The photodiode 142P thus formed is mounted with an angle of
45 degree with regard to the laser diode 32, and the output of the
photodiode 142P is connected to the laser control unit 162
electrically. Thereby, there is formed a feedback control system of
the laser diode similar to the one shown in FIG. 106.
[0811] By providing a multimode optical fiber 302F having a core
302f of 50 .mu.m diameter and a clad of 125 .mu.m diameter in
optical alignment with the laser beam reelected by the mirror 301,
an optical telecommunication system is constructed. Such an optical
telecommunication system is simple in construction with reduced
number of parts, and it is possible to provide a compact optical
transmission module. Such an optical transmission module can
control the laser output stably and a reliable optical
telecommunication is realized. In the present embodiment, it is
also possible to form the mirror 301 on the surface of the
monitoring photodetection device separately to the electrode.
[0812] Next, another embodiment of the present invention will be
explained with reference to FIG. 108.
[0813] Referring to FIG. 108, the present embodiment provides the
long-wavelength surface-emission laser diode 32 mounted on the Si
mount substrate 301 together with the driver circuit and laser
control circuit not illustrated. In the illustrated example, four
laser diodes 32A are arranged with the pitch of 200 .mu.m, wherein
each of the laser diodes has a laser oscillation wavelength of 1.3
.mu.m.
[0814] In the present embodiment, the edge of the optical waveguide
is cut to form a 45 degree angle by a diamond blade, and the mirror
301 is formed by covering the oblique surface thus formed by an Au
film. Thereby, the thickness of the Au film is controlled such that
the Au film has a transmissivity of 3%. The optical waveguide 302
thus processed is coupled optically with the laser diode 32 by
achieving optical alignment, and the monitoring photodetector 142P
is mounted on the optical path of the optical beam divided by the
mirror 301. According to such a construction, it is necessary to
control the output of the laser diode 32 via the control unit 162
in response to the output of the photodetector 142. By doing so, it
becomes possible to construct an optical module of simple
construction and reduced number of parts.
Twenty-Fifth Embodiment
[0815] Next, another embodiment of the present invention will be
described.
[0816] Conventionally, edge-emission laser diodes are used
extensively in optical telecommunication systems. When using such
an edge emission laser diode in combination with plural optical
fibers, there is a need of achieving optical coupling for each of
the laser diodes one by one. In the case of an edge-emission laser
diode, there is another problem of large beam divergence in that
the laser beam spreads rapidly due to the large emission angle.
Further, the emission angle is different in the lateral direction
and vertical direction, and the laser beam shows an elongated beam
spot characterized by poor aspect ratio. Thus, in the case of
conventional edge-emission laser diode, ft has been necessary to
provide a coupling lens for each of the laser diodes.
[0817] Because of these reasons, it was not possible to form the
laser diode in the form of high density array in the case of using
conventional edge-emission type laser diode.
[0818] In contrast, the present invention has enabled the
construction of a high-density laser array by using a number of
surface-emission laser diodes commonly on a single chip
monolithically. By combining optical fibers with the laser diode
array thus formed, a large capacity optical transmission system is
realized.
[0819] FIGS. 110A and 110B show an example of an optical
telecommunication system in which the long-wavelength laser diodes
of 1.1-1.7 .mu.m are used in combination with optical fibers.
[0820] As represented in FIG. 110A, the surface-emission laser
diode of the present invention is characterized by a narrow
divergence of the optical beam in any of the horizontal and lateral
directions of the beam spot. In fact, the laser beam produced by
such a surface-emission laser diode is characterized by a circular
beam cross-section.
[0821] In contrast, the laser diode of the edge emission type laser
diode produces an optical beam diverging rapidly with large
emission angle, wherein the laser beam produced by such an
edge-emission layer diode shows a beam shape characterized by an
eclipse. Thus, the size of the laser beam spot is different in the
horizontal direction vertical direction.
[0822] Thus, by using the surface-emission laser diode of the
present invention, it becomes possible to construct a large
capacity optical transmission system by arranging a number of
optical fibers with high density.
[0823] In the case of such a high-density telecommunication system,
it is difficult to use coloring layer or identification ring used
commonly in ordinary optical fiber cables for identifying reach of
the optical fibers in the cable.
[0824] In the case of the surface-emission laser diode, it is
possible to form each beam emission part 32A constituting the array
simultaneously and in high density by using the lithographic
process. In the case of handling a number of optical fibers in such
a laser diode array, it is more preferable to handle the optical
fibers in the form of bundle, rather than handing each of the
optical fibers. Particularly, it is preferable to make a
correlation between the optical fibers bundled in the form of a
cable and the laser diodes in the array.
[0825] On the other hand, in the case the optical fibers 101 are
assembled to form a bundle, it is not easy to identify the center
of the bundle or individual optical fibers in the bundle.
[0826] For example, in the case the cladding 101b of the optical
fiber 101X at the center of the optical fiber bundle is colored as
represented in FIGS. 11A-11C, the recognition of the optical fiber
is easy.
[0827] Thus, it is possible to achieve the correlation between the
laser diodes in the array and the optical fibers in the bundle by
first turning on the laser diode 32AZ at the center of the array.
Thereby this laser diode is correlated with the colored optical
fiber at the center of the bundle.
[0828] In the example of FIGS. 112, it can be seen that four fibers
are colored at cladding layer. By correlating these four optical
fibers with four laser diodes, it is possible to easily establish
the correspondence between the optical fiber and the laser diode
for other optical fibers.
[0829] It should be noted that the cladding layer of an optical,
fiber is used merely for confining the light in the core. Thus,
coloring of the cladding layer 101b as shown in FIGS. 113A and 113B
does not cause a problem at all with regard to the information
transmission, which takes place through the core 101a.
[0830] FIGS. 111B and 111C show a modification of FIG. 111A. In the
example of FIG. 111A, the optical fibers 101 are arranged in a
closest packing state. In the case of FIG. 111B, on the other hand,
optical fibers are arranged in a square grid. Further, FIG. 111C
shows an example of uses a ferule 101F for marinating the square
grid arrangement of the optical fibers.
[0831] FIG. 114 shows an example of a single mode optical fiber. As
compared with the optical fibers of multimode represented in FIGS.
113A and 113C, the core diameter of the single mode optical fiber
is very small with regard to the diameter of the cladding layer of
125 .mu.m, and the area of the cladding layer 101b with respect to
the core 101a can reach 172 times as large as the core area of the
multimode optical fiber. In the case of the single mode fiber,
recognition of the optical fiber by using the color layer is
facilitated further.
Twenty-Sixth Embodiment
[0832] Next, another embodiment of the present invention will be
explained. In the conventional laser diode, there has been a
problem that the threshold current is changed depending on the
temperature. In a telecommunication system, the use of the laser
diode at strictly constant temperature is difficult. Thus, in the
case of the edge-emission laser diode, it has been practiced to
detect the leakage light emitted in the backward direction by a
photodetector and the output off the photodetector has been used
for a feed back control of the laser outputs such that the output
of the laser diode is maintained constant.
[0833] In the case of the surface-emission laser diode, however,
such detection of the leakage light is not possible. Thus, in the
conventional surface-emission laser diode, it has been practiced to
provide a photodetector between the laser diode and the optical
fiber or at the downstream side of the optical fiber, and the feed
back control has been applied by using such a photodetector such
that the output of the laser diode is maintained constant.
[0834] FIG. 115 shows an example of the I-L (current-optical
output) characteristic of the long-wavelength laser diode of the
present invention.
[0835] Referring to FIG. 115, the laser diode of the present
invention has a feature, contrary to the conventional laser diode,
in that the change of the threshold current with temperature is
very small. Only the slope of the curve is influenced with the
temperature.
[0836] Thus, as long as the laser diode is driven at a constant
drive voltage, the change of the optical output is small even in
the case there is a temperature change.
[0837] For example, in the case the laser diode is driven at the
drive current of 6 mA in FIG. 115, the change of optical output
associated with the temperature change from 10.degree. C. to
70.degree. C. is only 0.1 mw. Even when the temperature is changed
from 10.degree. C. to 100.degree. C., the change of the optical
output is only 0.25 mW. In terms of S/N ratio, this is respectively
26 dB and 18 dB. Thus, the laser diode of the present invention can
provide a sufficient signal quality in the commonly used
temperature environment of 20-70.degree. C.
[0838] It should be noted that the drift of current of a constant
current source is in the order of .+-.2-3%. Thus, it is easy to
achieve a constant current control.
[0839] Thus, by setting the upper limit and lower limit for the
optical output and further setting the upper limit and lower limit
for the temperature, the present embodiment achieves the desires
control of the optical output within the foregoing upper and lower
limits by controlling the drive current at a constant value x
between a first reference drive current a corresponding to the
upper limit optical output at the lower limit temperature and a
second reference drive current b corresponding to the lower limit
optical output at the upper limit temperature.
[0840] According to the present invention, the laser diode is
driven at a constant current determined with respect to the target
optical output.
Twenty-Seventh Embodiment
[0841] The threshold current of the laser diode increases gradually
with time and when the threshold current has exceeded a
predetermined value, it is the lifetime of the laser diode.
[0842] Thus, there is a demand to avoid deterioration of signal
quality even when the aging is in progress.
[0843] As shown in FIG. 117, for example, there is provided a half
mirror 411 in the transmission path 410 coupled to the laser diode
32, and the optical beam thus divided is detected by the
photodetection device 412. In this case, the intensity of the laser
beam detected by the photodetection device, in other words the
monitor optical strength, is fed back to the control unit 415 and
the constant current source 416.
[0844] In the illustrated example, the output of the photodetection
device 412 is supplied to the communication control unit 414, and
the communication control unit 414 obtains the correction of the
drive current by referring to a conversion table 414A representing
the relationship between the monitor optical output and the
correction of the drive current. According to such a construction,
it is possible to eliminate the effect of aging by removing the
change of optical output caused by aging. Thereby, a practical
telecommunication system is realized.
[0845] In order to monitor for the aging or anomaly related to the
aging, it is sufficient that the output of the photodetection
device at the reception end can be obtained. Thus, it is also
possible to transmit the reading of the photodetection device at
the reception side in the form of data to the transmission side.
For example, it is possible to transmit the reading of the
photodetection device periodically or at any time to the optical
transmission side, separately to the optical telecommunication
data. The data thus transmitted may be forwarded in the
transmission side from the communication control unit to the laser
control unit for correcting the drive current. By doing so, it is
possible to eliminate the fluctuation of optical output caused by
aging.
Twenty-Eighth Embodiment
[0846] Next, a further embodiment of the present invention will be
described.
[0847] FIG. 118 shows an example of an optical telecommunication
system that uses a long-wavelength laser diode 32 of the present
invention.
[0848] Referring to FIG. 118, it can be seen that the system
includes a laser diode module connected with an optical fiber 421,
an optical circuit substrate 423 carrying the laser diode module
422, and electronic substrates 423B and 423C provided above and
below the optical circuit substrate 423A. The circuit substrates
423A-423C are accommodated in a case 420, wherein the substrates
423A-423C form a air flow path in the case 420. Further, the case
420 is provided with a fan 424A, and an air outlet is formed on the
case 420 at the side opposite to the fan 424A. The electronic
substrates 423B and 423C carry electronic components 423a
thereon.
[0849] The fan 424A may be a compulsory air-feeding fan such as
sirocco fan, and supplies a cooling air inside the case 420. The
cooling air flows along the space between the substrates 423B and
423C and there takes place heat exchange between the air and the
laser diode module 422 as the cooling air is caused flow along the
substrates. The cooling air thus exchanged heat is expelled from
the air outlet 424B.
[0850] In the illustrated example, the substrate 423B of the
smallest heat generation is provided at the lower part and the
substrate 423C causing a larger heating is provided thereabove.
[0851] It should be noted that the surface of the optical substrate
423A facing the electronic substrate 423C is reduced with
projections or depressions so as to minimize disturbance of air
flow. Meanwhile, a similar effect is obtained also in the case in
which a flat board carrying no electronic components is used for
the electronic substrates 423B or 423C. While the present
embodiment uses only one fan 424A, it is possible to increase the
number of fans with the number of the laser diodes provided in the
case 420. It should be noted that the heating is increased when the
number of the laser diodes is increased. Alternatively, it is
possible to increase the flow rate of the air. In this case, the
area of the air outlet has to be increased.
[0852] It is also possible to use the fan 424A to pull the air
inside the case 420. Further, it is possible to provide a fan in
the air intake side and another fan in the air outlet side.
Twenty-Ninth Embodiment
[0853] FIG. 119 shows an example of the laser diode module that
uses the long-wavelength laser diode 32 of the present
invention.
[0854] Referring to FIG. 119, it can be seen that two laser diodes
32A are formed on the GaAs substrate 32 monolithically, and the
GaAs substrate 32 is carried by a Si mount substrate 131 having a
large thermal conductivity. Further, the Si substrate 131 is
carried on a ceramic substrate 1363 having a still larger thermal
conductivity. The ceramic substrate 136 is mounted on an optical
circuit board such as the substrate 434A shown in FIG. 101. The
ceramic substrate 136 is cooled.
[0855] Further, electric interconnection to the laser diode 32A is
provided by way of bonding wires connecting an electrode 136A on
the ceramic substrate 136 and the electrode on the laser diode
32A.
[0856] With such a construction, the heat generated by the laser
diode 32A is transferred to the GaAs substrate by thermal
conduction and then to the Si substrate 131 of lower temperature.
Further, the heat is transferred to the ceramic substrate 136 of
lower temperature, and efficient cooling of the laser diode 32A
becomes possible. It should be noted that the surface area of the
ceramic substrate 136 is the largest among the constituent parts,
and thus, the ceramic substrate 136 is cooled efficiently by
radiation and contact with the air.
[0857] The GaAs substrate constituting the laser diode chip of the
present invention has a thermal conductivity of 0.54 W/cmK, while
it is noted that the thermal conductivity of Si is 1.48 W/cmK.
Further, a material having a larger thermal conductivity than Si
such as BeO (2.72 W/cmK) or diamond (9.0 W/cmK) can also be used,
wherein the value of the thermal conductivity is the value at
300K.
[0858] Meanwhile, thermal transfer across two substrates contacting
with each other can be increased by reducing the surface roughness
and increasing the intimateness. On the other hand, excessive
processing of the surface merely invites increase of cost and the
improvement of heat transfer is little. For example, the precision
of surface flatness beyond 10 nm merely increases the cost and is
not practical. Thus, the lower limit of the surface roughness
should be around 10 nm.
[0859] With regard to the upper limit of surface roughness,
excessively rough surface is disadvantageous for contacting two
substrates. Thus, the inventor of the present invention conducted
extensive study about this matter and discovered that the surface
roughness of 1000 nm or less is sufficient for guaranteeing
intimate contact and excellent heat transfer. In the experiments
conducted by causing the laser diode to oscillate, it was confirmed
that the heat generated by the laser diode 32A is transferred from
the laser diode chip 32 to the first substrate 131, and to the
second substrate 136, without causing accumulation of heat. With
this, it became possible to reduce the fluctuation of threshold
current of the laser diode caused by heating. Thereby, it became
possible to cause the laser diode to oscillate stably.
[0860] FIG. 120 show an example of the long-wavelength laser diode
module of the present embodiment, wherein the GaAs substrate 32
carrying the laser diode 32A is fixed upon the Si mount substrate
131 via a thermal conduction layer 137, and the Si mount substrate
131 is mounted on the ceramic substrate 136 via another thermal
conduction layer 138.
[0861] In the present embodiment, the contact surfaces of the
substrates 32, 131 and 136 are polished by using an alumina
abrasive to a surface roughness Ra of 10-1000 nm, and any gap
formed therebetween is filled with the thermal conductive layer 137
or 138.
[0862] Here, it should be noted that the thermal conductive layer
137 or 138 may be formed of an organic polymer material such as
epoxy resin or silicone resin or acryl resin dispersed with metal
powders such as aluminum or gold or silver or copper. The metal
powders have a diameter of several nanometers to 100 nm and the
resin layer is applied with a thickness of 3-100 .mu.m. The
proportion of the metal powders was set to 0.1-1 part with regard
to 1 part of resin.
[0863] FIG. 121 is an example of a long-wavelength laser diode of
the present embodiment and is formed of a laser module package 431
including the laser diode module 422 of FIG. 118 and an optical
fiber 421.
[0864] At the bottom part of the package 431, there is provided a
cooling fin structure 431A having eight cooling fins each having a
rectangular cross section, with a base length of 1 mm and a height
of 3 mm, and the cooling air is caused to flow along the cooling
fin structure 431A.
[0865] It should be noted that the outer surface of the laser
module package 431 performs also as a heat-radiating member similar
to the ceramic substrate 136, in addition to the function of the
packaging of the laser diode. It should be noted that the surface
on which the cooling fins are provided ins not limited to the
bottom surface but may be other surface along with the cooling air
is caused to flow. Further, the number and construction of the
cooling fins are by no means limited to the one illustrated.
[0866] The point of the present embodiment is to provide, in a heat
radiating structure in which a first substrate including a heat
source and a second substrate not including a heat source are
stacked, such that the surface area of the part of the second
substrate not contacting with the first substrate is set larger
than the surface area of the second substrate contacting with the
first substrate. Any of the structure may be used in place of the
cooing fin as long the structure increases the surface area of the
second substrate.
[0867] As noted previously, in the surface-emission laser diode
oscillating at the wavelength of 1.1-1.7 .mu.m, it is possible to
form a number of laser diodes on a common chip and is suitable for
large capacity optical telecommunication system. However, design of
such a construction including a large number of laser diode has to
be made carefully in view of severe heating. The present embodiment
provides a solution to this problem. Larger the number of the laser
diodes, more effective the construction of the present
embodiment.
Thirtieth Embodiment
[0868] FIGS. 122 and 123 show another embodiment of the present
invention, wherein FIG. 122 shows the case a number of
long-wavelength surface-emission laser diodes 32A of the present
invention are used to form a one-dimensional array, while FIG. 133
show the case the laser diodes form a two-dimensional array.
[0869] In the drawings, the laser diode elements 32A provided with
shading have corresponding photodetection devices, while other
elements have no photodetection device and the output thereof is
taken to the outside.
[0870] FIG. 124 shows the construction of FIG. 122 from a different
angle. It can be seen that the laser diode elements 32A provided
with the shading is provided with the photodetection device 142P
and the photodetection device 142P interrupts the laser beam. On
the other hand, the laser beam of the laser diode 32A not provided
with the photodetection device 142P is injected into the optical
fiber 352 by the lens 353.
[0871] The optical output of the laser diode 32A not provided with
the photodetector 142P is calculated based on the output of the
photodetector(s) 142P located in the vicinity thereof. Based on the
output of the photodetector 142P thus obtained, output of each of
the laser diodes 32A is controlled. The variation of output between
the laser diodes 32A is easily corrected by using a correction
coefficient obtained in advance.
[0872] FIGS. 125-127 show an example of output control of the laser
diode using the construction of FIG. 124.
[0873] Referring to FIG. 125, the laser diode chip 32 includes a
one-dimensional array of laser diodes 32A similar to FIG. 122. For
the sake of convenience of explanation, the laser diodes 32A in
FIG. 125 are provided with numbers (0)-(6).
EXAMPLE 1
[0874] In Example 1, the out put of the photodetection device 142P
that monitors the output of the first laser diode 32A (1) is used
for controlling the drive currents of the laser diodes 32A (0)-(6)
of FIG. 125 by using the driver circuit 106 of FIG. 106 such
that
[0875] I.sub.0 is controlled based on the value of
a.sub.0O.sub.0;
[0876] I.sub.1 is controlled based on the value of
a.sub.1O.sub.0;
[0877] I.sub.2 is controlled based on the value of
a.sub.2O.sub.0;
[0878] I.sub.3 is controlled based on the value of
a.sub.3O.sub.0;
[0879] I.sub.4 is controlled based on the value of
a.sub.4O.sub.0;
[0880] I.sub.5 is controlled based on the value of a.sub.5O.sub.0;
and
[0881] I.sub.6 is controlled based on the value of a.sub.6O.sub.0,
wherein Ii represents the drive current of the i-th (i=0-6) laser
diode 32A and O1 represents the optical output of the i-th laser
diode 32A. Further, a.sub.i represents a correction coefficient and
is determined in advance by measuring the output of respective
laser diodes such that the output calculated based on the drive
current is closest to the actual laser output.
EXAMPLE 2
[0882] In Example 2, the out put of the photodetection device 142P
that monitors the output of the first laser diode 32A (1) is used
for controlling the drive currents of the laser diodes 32A (0)-(6)
of FIG. 125 by using the driver circuit 106 of FIG. 106 such
that
[0883] I.sub.0 is controlled based on the value of
a.sub.0O.sub.0;
[0884] I.sub.1 is controlled based on the value of
a.sub.1O.sub.0+b.sub.1O.sub.3;
[0885] I.sub.2 is controlled based on the value of
a.sub.2O.sub.0+b.sub.2O.sub.3;
[0886] I.sub.3 is controlled based on the value of
a.sub.3O.sub.3;
[0887] I.sub.4 is controlled based on the value of
a.sub.4O.sub.0+b.sub.4O.sub.6;
[0888] I.sub.5 is controlled based on the value of
a.sub.5O.sub.0+b.sub.5O.sub.6; and
[0889] I.sub.6 is controlled based on the value of
a.sub.6O.sub.6.
EXAMPLE 3
[0890] In Example 3, the out put of the photodetection device 142P
that monitors the output of the first laser diode 32A (1) is used
for controlling the drive currents of the laser diodes 32A
(00)-(44) of FIG. 128 by using the driver circuit 106 of FIG. 106
such that
[0891] I.sub.0 is controlled based on the value of
a.sub.00O.sub.00;
[0892] I.sub.01 is controlled based on the value of
a.sub.01O.sub.00;
[0893] I.sub.10 is controlled based on the value of
a.sub.2O.sub.00;
[0894] I.sub.11 is controlled based on the value of
a.sub.10O.sub.00;
[0895] I.sub.02 is controlled based on the value of
a.sub.02O.sub.03;
[0896] I.sub.03 is controlled based on the value of
a.sub.03O.sub.03;
[0897] I.sub.04 is controlled based on the value of
a.sub.04O.sub.03,
[0898] I.sub.12 is controlled based on the value of
a.sub.12O.sub.03;
[0899] I.sub.13 is controlled based on the value of
a.sub.13O.sub.03;
[0900] I.sub.14 is controlled based on the value of
a.sub.14O.sub.03,
[0901] wherein a.sub.ij is a correction coefficient.
EXAMPLE 4
[0902] In the example of FIG. 129, the ij-th laser diode 32A of
FIG. 129 is controlled such that
[0903] I.sub.00 is controlled based on the value of
a.sub.00O.sub.00;
[0904] I.sub.01 is controlled based on the value of
a.sub.01O.sub.00+b.sub.01O.sub.04+c.sub.01O.sub.22;
[0905] I.sub.11 is controlled based on the value of
a.sub.11O.sub.00+b.sub.11O.sub.22;
[0906] I.sub.21 is controlled based on the value of
a.sub.21O.sub.00+b.sub.21O.sub.22+c.sub.21O.sub.10;
[0907] In Example, 4, a.sub.ij, b.sub.ij and c.sub.ij are
correction coefficients.
Thirty-First Embodiment
[0908] Next a further embodiment of the present invention will be
explained.
[0909] The present embodiment is related to the production control
of the laser array module that uses the laser diode of the present
invention and used for the long-wavelength optical
telecommunication.
[0910] Here, it should be noted that the laser array module means
that a plurality of surface-emission laser diodes are provided in
the form of a module and includes the case in which a number of
laser diodes are formed on a single chip or the case in which a
number of such chips are arranged. Further, the case of plural
chips, each carrying a single laser diode, are arranged is also
included.
[0911] In the production process of such a laser diode module,
laser chip modules carrying a laser diode array are produced.
[0912] In such a production process, a predetermined number of
laser diode devices or laser diode chips are prepared and these
elements are arranged to form the desired laser array module.
[0913] Thereafter, an inspection process is conducted with regard
to the laser diode modules, wherein the inspection is conducted for
each of the laser diode chips and each of the laser diodes. Upon
confirmation that a desired quality is achieved for each of the
devices and each of the chips, the laser array module becomes the
state ready for shipping.
[0914] In the case there is detected a single device out of n
devices of laser diode chips arranged to form the laser array
module is defective, the laser chip module is treated as a
defective product.
[0915] On the other hand, if it is possible to treat a laser chip
module, in which the number c of the defective laser diodes or
chips is within a predetermined number n of good products, as being
a good product, the productivity of the laser array module is
increased significantly.
[0916] The inventor of the present invention realized the
importance of such a product control process and diligently studied
the production control process applicable to the case of producing
a product that uses a large number of elements each performing a
single and same function as other elements.
[0917] Hereinafter, the present embodiment will be described with
reference to FIGS. 130 and 131.
[0918] Referring to FIG. 130, the production of the laser diode or
chip is conducted according to the step S1 of wafer process step,
step S2 for producing a laser diode array step, and S3 for
producing a laser array module. Further, there is provided an
inspection step S4 for inspecting the array or module produced by
the steps S2 and S3.
[0919] In the present embodiment, the laser array module that
turned out to include a defective device or defective chip is not
immediately treated as a defective product. In the event the number
c of the defects is within a certain threshold (n-c) (n is the
total number of the devices or chips in the module), the product is
shipped as the product having the quality of (n-c). In the process
of FIG. 113, this examination step is conducted at the step S6. In
the case the entire chips-n are defective, the laser diode module
is treated as a defective device.
[0920] Thus, in the step S4, of FIG. 131, an inspection step S41 is
conducted for evaluating the product quality for each channel CH by
using a high-frequency probe or analyzer.
[0921] The data of the product quality for each channel thus
obtained is then examined in the product examination step S42,
wherein the step S42 judges whether or not the required quality is
satisfied for all of the laser diodes or chips. The judgment is
made with regard to the items such as I-L characteristic, I-V
characteristic, fiber coupling loss, pulse modulation
characteristic, temperature dependence, and the like. Thereby, the
laser array modules are shipped according to the grades
corresponding to the number of the properly functioning laser diode
chips.
[0922] From the explanation noted above, it will be apparent that
the present embodiment enables the use of most of the laser array
modules, used for the optical telecommunication system by using the
surface-emission laser diode of 1.1-1.7 .mu.m, according to the
condition of the module.
[0923] According to the present embodiment, therefore, the laser
array modules are used effectively even in the case only n-c of the
laser diodes operate properly, and the loss of the laser array
module during the production process is minimized.
[0924] The present invention is not limited to the embodiments
described heretofore, but various variations and modifications may
be made without departing from the scope of the invention.
[0925] The present application is based on Japanese priority
applications No. 2001-050145 filed on Feb. 26, 2001, No-2001-050171
filed on Feb. 26, 2001, No. 2001-050083 filed on Feb. 26, 2001, No.
2001-051253 filed on Feb. 26, 2001, No. 2001-051256 filed on Feb.
26, 2001, No. 2001-051266 filed on Feb. 26, 2001, No. 2001-053213
filed on Feb. 27, 2001, No. 2001-053218 filed on Feb. 27, 2001, No.
2001-053200 filed on Feb. 27, 2001, No. 2002-50548 filed on Feb.
26, 2002, No. 2001-053190 filed on Feb. 27, 2001, No. 2001-053225
filed on Feb. 27, 2001, No. 2001-073767 filed on Mar. 15, 2001, No.
2001-090711 filed on Mar. 27, 2001, No. 2002-019748 filed on Jan.
29, 2002, No. 2002-033590 filed on Feb. 12, 2002, the entire
contents of which are hereby incorporated by reference.
* * * * *