U.S. patent application number 09/334146 was filed with the patent office on 2002-03-21 for optoelectronic assembly and method of making the same.
Invention is credited to VERDIELL, JEAN-MARC.
Application Number | 20020034834 09/334146 |
Document ID | / |
Family ID | 21704232 |
Filed Date | 2002-03-21 |
United States Patent
Application |
20020034834 |
Kind Code |
A1 |
VERDIELL, JEAN-MARC |
March 21, 2002 |
OPTOELECTRONIC ASSEMBLY AND METHOD OF MAKING THE SAME
Abstract
An optoelectronic assembly having an insulating substrate with a
planar surface and a metal layer bonded to the planar surface such
that selected regions of the substrate are exposed and a step is
produced between the substrate and a top surface of the metal
layer. An active optical device is mounted on the metal layer and a
passive optical device is aligned with the active device using the
step as a fiduciary for positioning the former. The metal layer
provides an electrical path to the active device. The thickness of
the metal layer is selected such that the heat generated by the
active device is dissipated, the substrate does not interfere with
the propagation of light along the first optical axis, and such
that the in-plane coefficient of thermal expansion (CTE) of the
metal layer is constrained by the substrate. The optoelectronic
assembly is also suitable for mounting active devices provided with
submounts or without.
Inventors: |
VERDIELL, JEAN-MARC; (PALO
ALTO, CA) |
Correspondence
Address: |
BLAKELY, SOKOLOFF, TAYLOR & ZAFMAN LLP
MICHAEL J. MALLIE
12400 WILSHIRE BOULEVARD
SEVENTH FLOOR
LOS ANGELES
CA
90025
US
|
Family ID: |
21704232 |
Appl. No.: |
09/334146 |
Filed: |
June 15, 1999 |
Related U.S. Patent Documents
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
|
|
09334146 |
Jun 15, 1999 |
|
|
|
09003114 |
Jan 6, 1998 |
|
|
|
5977567 |
|
|
|
|
Current U.S.
Class: |
438/22 ;
438/29 |
Current CPC
Class: |
G02B 6/4244 20130101;
G02B 6/4204 20130101; G02B 6/4245 20130101; G02B 6/4208 20130101;
H01L 2224/48091 20130101; H01L 2224/48091 20130101; G02B 6/4272
20130101; H01L 2924/00014 20130101; G02B 6/4273 20130101; G02B
6/4274 20130101; G02B 6/4267 20130101; G02B 6/4268 20130101; G02B
6/423 20130101 |
Class at
Publication: |
438/22 ;
438/29 |
International
Class: |
H01L 021/00; H01L
021/44; H01L 021/48 |
Claims
What is claimed is:
1. An optoelectronic assembly comprising: a) an active optical
device having a first optical axis; b) an insulating substrate with
a planar surface; c) a metal layer bonded to said planar surface
such that selected regions of said insulating substrate are exposed
and a step is produced between said insulating substrate and a top
surface of said metal layer, said active optical device being
mounted on said metal layer such that said first optical axis is
parallel to said insulating substrate, said metal layer further
providing an electrical path to said active optical device and
having a predetermined thickness such that: i) heat generated by
said active optical device is dissipated; ii) said insulating
substrate does not interfere with the propagation of light along
said first optical axis; and iii) the in-plane coefficient of
thermal expansion (CTE) of said metal layer is constrained by said
insulating substrate.
2. The optoelectronic assembly of claim 1 wherein said active
optical device is selected from the group consisting of diode
lasers, optical amplifiers, optical modulators, optical switches,
optical fibers, and light detectors.
3. The optoelectronic assembly of claim 1 further comprising a
first passive optical element having a second optical axis and
being positioned in one of said selected regions such that said
second optical axis is aligned with said first optical axis.
4. The optoelectronic assembly of claim 3 wherein said passive
optical element is selected from the group consisting of optical
fibers, lenses, filters, diffraction gratings, beam splitters,
isolators, and mirrors.
5. The optoelectronic assembly of claim 3 wherein said passive
optical element is positioned on said insulating substrate.
6. The optoelectronic assembly of claim 5 wherein said passive
optical element has a flat bottom.
7. The optoelectronic assembly of claim 3 wherein said metal layer
is patterned such that said step is a fiduciary for aligning said
passive optical element.
8. The optoelectronic assembly of claim 7 wherein said passive
optical element has a flat side wall fitting against said
fiduciary.
9. The optoelectronic assembly of claim 1 wherein said metal layer
is patterned as an electrical circuit.
10. The optoelectronic assembly of claim 9 further comprising
electrical components mounted on said metal layer.
11. The optoelectronic assembly of claim 1 wherein said
predetermined thickness ranges between 50 .mu.m and 1,000
.mu.m.
12. The optoelectronic assembly of claim 1 wherein said metal layer
is made of a material selected from the group consisting of copper,
copper-tungsten, copper alloys, copper composites, copper plated
materials, copper clad materials and copper laminates.
13. The optoelectronic assembly of claim 1 wherein said insulating
substrate is a ceramic chosen from the group consisting of alumina,
BeO and AlN.
14. The optoelectronic assembly of claim 1 further comprising a
bottom metal layer bonded to said insulating substrate opposite
said planar surface.
15. The optoelectronic assembly of claim 14 wherein said bottom
metal layer is patterned as an electrical circuit.
16. The optoelectronic assembly of claim 14 further comprising a
cooling means.
17. The optoelectronic assembly of claim 16 wherein said bottom
metal layer is patterned and covered with a cover to provide
channels for a liquid coolant.
18. A method of making an optoelectronic assembly comprising the
following steps: a) selecting an insulating substrate with a planar
surface; b) bonding a metal layer to said planar surface such that
selected regions of said insulating substrate are exposed and a
step is produced between said insulating substrate and a top
surface of said metal layer; c) mounting an active optical device
having a first optical axis on said metal layer such that said
first optical axis is parallel to said insulating substrate, and
such that said metal layer provides an electrical path to said
active optical device; d) selecting a predetermined thickness for
said metal layer such that: i) heat generated by said active
optical device is dissipated; ii) said insulating substrate does
not interfere with the propagation of light along said first
optical axis; and iii) the in-plane coefficient of thermal
expansion (CTE) of said metal layer is constrained by said
insulating substrate.
19. The method of claim 18 wherein said active optical device is
selected from the group consisting of diode lasers, optical
amplifiers, optical modulators, optical switches, optical fibers,
and light detectors.
20. The method of claim 18 further comprising the step of
positioning a first passive optical element having a second optical
axis in one of said selected regions such that said second optical
axis is aligned with said first optical axis.
21. The method of claim 18 further comprising the step of
patterning said metal layer.
22. The method of claim 18 wherein said metal layer is made of a
material selected from the group consisting of copper,
copper-tungsten, copper alloys, copper composites, copper plated
materials, copper clad materials, and copper laminates.
23. The method of claim 18 wherein said insulating substrate is
selected from ceramics from the group consisting of alumina, BeO
and AlN.
24. The method of claim 18 further comprising the step of forming
said metal layer by a process selected from the group consisting of
photolithography, etching, stamping, brazing, electrodeposition,
and electro-discharge machining (EDM) before said bonding step.
25. The method of claim 18 further comprising the step of forming
said metal layer by a process selected from the group consisting of
photolithography and diamond sawing after said bonding step.
26. The method of claim 18 further comprising the step of polishing
said metal layer to said predetermined thickness.
27. The method of claim 18 further comprising the step of etching
said metal layer to said predetermined thickness.
28. The method of claim 18 further comprising the step of placing a
passive optical element having a second optical axis in one of said
selected regions such that said second optical axis is aligned with
said first optical axis.
29. The method of claim 28 wherein said placing step includes
actively displacing said passive optical element in one of said
selected regions.
30. The method of claim 28 wherein said placing step includes
displacing said passive optical element with said step serving as a
fiduciary along which said passive optical element is
displaced.
31. A method of making at least two optoelectronic assemblies
comprising the following steps: a) selecting an insulating
substrate with a planar surface; b) bonding at least a first and
second metal layers to said planar surface such that selected
regions of said insulating substrate are exposed and steps are
produced between said insulating substrate and top surfaces of said
metal layers; c) mounting active optical devices each having first
optical axes on said metal layers such that said first optical axes
are parallel to said insulating substrate, and such that said metal
layers provide electrical paths to said active optical devices; d)
selecting a predetermined thickness for said metal layers such
that: i) heat generated by said active optical devices is
dissipated; ii) said insulating substrate does not interfere with
the propagation of light along said first optical axes; iii) the
in-plane coefficient of thermal expansion (CTE) of said metal
layers is constrained by said insulating substrate; and e) dividing
said insulating substrate into at least a first and second parts,
each of said parts including one of said metal layers.
32. The method of claim 31 further comprising the step of forming
said metal layers by a process selected from the group consisting
of photolithography, etching, stamping, brazing, electrodeposition
and electro-discharge machining before said bonding step.
33. The method of claim 31 further comprising the step of forming
said metal layers by a process selected from the group consisting
of photolithography, etching and diamond sawing after said bonding
step.
34. The method of claim 31 wherein said bonding step is performed
after said dividing step.
35. The method of claim 31 further comprising the step of placing
passive optical elements in said selected regions.
36. The method of claim 35 wherein said placing step is performed
before said dividing step.
37. The method of claim 35 wherein said placing step is performed
after said dividing step.
38. An optoelectronic assembly comprising: a) an insulating
substrate with a planar surface; b) a metal layer bonded to said
planar surface such that selected regions of said insulating
substrate are exposed and a step is produced between said
insulating substrate and a top surface of said metal layer; c) a
first active optical device on a submount, said first active
optical device having a first optical axis, said submount being
positioned on said insulating substrate where said step is a
fiduciary for aligning said first active optical device such that
said first optical axis is parallel to said insulating substrate,
said submount further providing an electrical path to said first
active optical device and having a predetermined thickness such
that: i) heat generated by said first active optical device is
dissipated through said submount; and ii) said insulating substrate
does not interfere with the propagation of light along said first
optical axis.
39. The optoelectronic assembly of claim 38 wherein said first
active optical device is selected from the group consisting of
diode lasers, optical amplifiers, optical modulators, optical
switches, optical fibers, and light detectors.
40. The optoelectronic assembly of claim 38 further comprising a
first passive optical element having a second optical axis and
being positioned in one of said selected regions such that said
second optical axis is aligned with said first optical axis.
41. The optoelectronic assembly of claim 40 wherein said passive
optical element is positioned on said insulating substrate.
42. The optoelectronic assembly of claim 38 further comprising a
second active optical device having a third optical axis and being
mounted on said metal layer such that said third optical axis is
aligned with said first optical axis.
43. The optoelectronic assembly of claim 38 wherein said
predetermined thickness ranges between 50 .mu.m and 1,000
.mu.m.
44. A method of making an optoelectronic assembly comprising the
following steps: a) selecting an insulating substrate with a planar
surface; b) bonding a metal layer to said planar surface such that
selected regions of said insulating substrate are exposed and a
step is produced between said insulating substrate and a top
surface of said metal layer; c) placing a first active optical
device having a first optical axis on a submount; d) positioning
said submount on said insulating substrate where said step is a
fiduciary for aligning said first active optical device such that
said first optical axis is parallel to said insulating substrate,
said submount further providing an electrical path to said first
active optical device and having a predetermined thickness such
that: i) heat generated by said first active optical device is
dissipated through said submount; and ii) said insulating substrate
does not interfere with the propagation of light along said first
optical axis.
45. The optoelectronic assembly of claim 44 wherein said first
active optical device is selected from the group consisting of
diode lasers, optical amplifiers, optical modulators, optical
switches, optical fibers, and light detectors.
46. The optoelectronic assembly of claim 44 further comprising a
first passive optical element having a second optical axis and
being positioned in one of said selected regions such that said
second optical axis is aligned with said first optical axis.
47. The optoelectronic assembly of claim 46 wherein said passive
optical element is positioned on said insulating substrate.
48. The optoelectronic assembly of claim 44 further comprising a
second active optical device having a third optical axis and being
mounted on said metal layer such that said third optical axis is
aligned with said first optical axis.
49. The optoelectronic assembly of claim 44 wherein said
predetermined thickness ranges between 50 .mu.m and 1,000
.mu.m.
50. A method of making an optoelectronic assembly comprising the
following steps: a) selecting an insulating substrate with a planar
surface; b) bonding a metal layer to said planar surface such that
selected regions of said insulating substrate are exposed and a
step is produced between said insulating substrate and a top
surface of said metal layer; c) providing said planar surface with
a metalization; d) positioning an active optical device having a
first optical axis on said metalization of said insulating
substrate where said step is a fiduciary for aligning said active
optical device such that said first optical axis is parallel to
said insulating substrate, said metalization further providing an
electrical path to said active optical device.
51. An optoelectronic assembly comprising: a) an active optical
device with an edge; b) an insulating substrate with a planar
surface and with an edge; c) a metal layer with an edge: wherein
said metal layer is bonded to said planar surface such that said
metal layer edge is flush with said insulating substrate edge, and
said active optical device edge is flush with said metal layer
edge, said metal layer further providing an electrical path to said
active optical device and having a predetermined thickness such
that: 1) heat generated by said active optical device is
dissipated; 2) said insulating substrate does not interfere with
the propagation of light along said first optical axis; and 3) the
in-plane coefficient of thermal expansion (CTE) of said metal layer
is constrained by said insulating substrate.
52. The optoelectronic assembly of claim 51 wherein said active
optical device is selected from the group consisting of diode
lasers, optical amplifiers, optical modulators, optical switches,
optical fibers, and light detectors.
53. The optoelectronic assembly of claim 51 wherein said
predetermined thickness ranges between 50 .mu.m and 1,000
.mu.m.
54. The optoelectronic assembly of claim 51 wherein said metal
layer is made of a material selected from the group consisting of
copper, copper-tungsten, copper alloys, copper composites, copper
plated materials, copper clad materials and copper laminates.
55. The optoelectronic assembly of claim 51 wherein said insulating
substrate is selected from ceramics from the group consisting of
alumina, BeO and AlN.
56. The optoelectronic assembly of claim 5,1 further comprising a
bottom metal layer bonded to said insulating substrate opposite
said planar surface.
Description
FIELD OF THE INVENTION
[0001] This invention relates to an optoelectronic assembly and,
more specifically, to a method and assembly for mounting active and
passive optical devices and elements.
BACKGROUND OF THE INVENTION
[0002] Optoelectronic components or active optical devices such as
diode lasers, light-emitting diodes (LEDs) and photodiode detectors
are used for printing, data storage, optical data transmission and
reception, as well as pumping of high power lasers and a multitude
of other applications. Optoelectronic packages are intended to
provide a way for mounting passive and active optical elements and
devices, as well as electrical components, in a robust structure
which preserves proper alignment. Typically, an optoelectronic
package includes an assembly upon which the optoelectronic
components are mounted. The requirements of a package depend upon
the application. In most cases, a package should provide precision
alignment for the internal components, enable high speed electrical
operation, provide for heat dissipation, match the coefficient of
thermal expansion (CTE) between the mount and the device, and
provide for simple external electrical connections and hermetic
sealing. In addition, the package should be mechanically robust and
be highly reliable. Clearly, satisfying all of these requirements
calls for a judicious choice of materials and mounting techniques.
In cases where numerous optical parts including other active
devices and passive optical elements, e.g. lenses, gratings,
fibers, mirrors and the like are intended to cooperate with each
other, alignment of these parts with respect to each other is
crucial.
[0003] These requirements have resulted in packages that are an
order of magnitude larger, costlier, and more difficult to
manufacture than purely electronic packages. In fact, the cost of
most optoelectronic devices is dominated by the package rather than
the optical devices themselves. New optoelectronic technologies
will not succeed in the marketplace if the cost of packaging
remains as high as it is now.
[0004] In a laser-fiber coupler, for example, the relative
positions of fibers, lenses, mirrors and lasers must be precisely
adjusted and permanently fixed to maintain beam coupling
efficiency. Single-mode optical fibers, for example, require
optical alignment tolerances of less than 1 .mu.m while multi-mode
optical fibers require optical alignment tolerances of less than 10
.mu.m. To achieve precision alignment, the optoelectronic device is
operated and monitored while the optical components are moved. The
components are typically secured in place once a coarse alignment
is achieved and then fine-tuned for optimal performance.
[0005] Precision alignment is complicated by the expansion and
contraction of package materials during temperature fluctuations
brought about primarily by the heat required to attach (solder) the
individual active optical devices and variations in ambient
temperature. Many prior art techniques use materials with varied
coefficients of thermal expansion (CTEs). During thermal cycling
the components of an optoelectronic device can drift out of
alignment causing poor performance or even complete malfunction.
Also, the mechanical stress produced can damage the components.
[0006] Existing packaging techniques often require that packages be
manufactured individually. For example, individual constituent
mechanical components of a package may be assembled into a finished
device one at a time in an assembly line process. Batch processing
techniques have been developed which can fabricate large numbers of
optoelectronic assemblies. These techniques, however, are usually
limited in their ability to manufacture a wide variety of
optoelectronic devices and result in performance sacrifices. This
is chiefly due to a high number of parts and a reliance on 3-D
alignment. For example, conventional TO cans and high performance
butterfly packages are not planar and thus their production cannot
be easily automated.
[0007] The cost of present devices is further increased by the fact
that optoelectronic assemblies frequently reside on substrates
which need to be mounted on other substrates to produce the final
package. For example, optoelectronic components mounted on silicon,
a frequently-used material, must be remounted when placed in an
optoelectronic package. Silicon is problematic for high-speed
optoelectronic packages because it is a rather lossy dielectric.
Progress in the fields of optics and electronics yields ever faster
optoelectronic devices and therefore, this characteristic of
silicon is limiting. Also, silicon is a brittle material
susceptible to cracking and chipping, a liability in mechanically
demanding applications. Further, its thermal conductivity is far
lower than that of conventional heatsinks such as copper.
[0008] The teachings of U.S. Pat. Nos. 4,357,072, 4,119,363, and
4,233,619 hinge on proper placement and alignment of the optical
and electronic components in three dimensions. Under these
circumstances, the alignment of a fiber to a laser or the alignment
of a lens to a detector is very difficult. Each component must be
actively positioned, incurring all the expense associated with
active alignment. Due to their design, these packages preclude the
use of batch process manufacturing techniques. Also, high speed
operation is problematic since these packages use numerous
electrical connections and have complex geometries. Because the
packages include dissimilar materials, great care is required to
ensure that the differences in the CTEs of the materials do not
cause misalignment during temperature fluctuations. This situation
often leads package designers to compromise between heat
dissipation and mechanical stability requirements.
[0009] Another disadvantage of these assemblies is the fact that
they require a relatively large number of steps to fabricate. Each
assembly has numerous subassemblies and, in general, one step is
required for fabricating each subassembly. In turn, multiple steps
and large numbers of subassemblies increase the cost, complexity
and size of the package.
[0010] A technique for providing passive alignment between a
plurality of diode lasers and optical fibers is described in U.S.
Pat. No. 5,163,108. By forming alignment pedestals on the substrate
for holding the laser chip and grooves in the substrate for holding
optical fibers, simple alignment is accomplished. Unfortunately,
this technique is limited because it does not provide an adaptable
method for including other optical components such as lenses or
mirrors. Further, the technique does not provide for enhanced heat
dissipation for active elements generating large amounts of
heat.
[0011] U.S. Pat. No. 5,123,074 describes a substrate for mounting
active and passive optical elements and optical components. The
substrate is made of an insulating block with metal regions formed
for electrical connections and for mounting components. A reversed
structure with a metal block and insulating regions is also
described. Electrical circuits are formed on the insulating regions
with the metal regions serving as bonding pads for the optical
hardware. In the embodiment with an underlying metal block, the
metal block acts as a ground plane, enhancing the high speed
electrical characteristics of the substrate. The surfaces of the
metal and insulating regions of the substrate lie in one plane, so
that the substrate provides no mechanical alignment. Thus, this
invention has the disadvantage of requiring the components to be
actively aligned. Also, since the surface of the substrate is a
single plane, different components cannot be mounted at different
heights to allow for optical alignment.
[0012] U.S. Pat. No. 4,926,545 discloses a batch process for
manufacturing optoelectronic assemblies. The assemblies can provide
passive alignment or simplified active alignment for optoelectronic
and optical components. The device includes metalization patterns
for aiding alignment. The metalization patterns act as visual
alignment aids for the placement of components. A problem with this
device is that it uses silicon and therefore has all the
disadvantages associated with silicon optoelectronic substrates. In
addition, the process only provides for device positioning in which
the optical axes are perpendicular to the substrate. This limits
the ability of an optoelectronic circuit designer since orienting
the optical axes parallel to the substrate allows one to design
more complex optoelectronic circuits in a smaller space.
[0013] U.S. Pat. No. 5,119,448 discloses a method of making a
substrate for mounting optical components by forming relief
structures into the surface of the substrate. The relief structures
provide mechanical alignment for the components. This method is
primarily concerned with mounting fiber arrays and using such fiber
arrays as sensors. No provisions are made for incorporating active
optoelectronic components such as lasers and using the substrate as
an optoelectronic assembly. Further, this method does not yield an
optoelectronic package, or assembly for a package, and requires at
least two patterning steps.
[0014] In the prior art, no single optoelectronic packaging
technique offers the simultaneous advantages of high speed
electrical operation, adaptability to produce various
optoelectronic devices, adaptability to batch processing, effective
heat dissipation and resistance to misalignment caused by changes
in temperature. Yet a combination of these characteristics is very
desirable for further progress in the field of optoelectronic
packaging and circuit design.
OBJECTS AND ADVANTAGES OF THE INVENTION
[0015] In view of the above, it is an object of the present
invention to provide an optoelectronic assembly which allows one to
mount active optical devices, passive optical devices, and
electronic components. The optoelectronic assembly may be used in
optoelectronic packages and is compatible with any combination of
active and passive optical devices and electronic components.
Another object of the invention is to provide a method for making
the optoelectronic assembly. The method is particularly well-suited
for batch processing fabrication. Additionally, the method is
compatible with pick and place manufacturing techniques to achieve
simple, precise, and reproducible active alignment of the active
and passive devices and elements.
[0016] A further object of the invention is to provide an
optoelectronic assembly which is compatible with high-speed
electrical operation and which exhibits superior heat dissipation
capability and mechanical stability over a wide temperature
range.
[0017] Another object of the invention is to provide CTE matching
between the optoelectronic components and the assembly. This
increases device life and allows for a wider range of operating
temperatures.
[0018] These and other objects and advantages will become apparent
upon consideration of the ensuing description and the accompanying
drawings.
SUMMARY
[0019] These objects and advantages are achieved by an
optoelectronic assembly having an insulating substrate with a
planar surface and a metal layer bonded to the planar surface such
that selected regions of the substrate are exposed and a step is
produced between the substrate and a top surface of the metal
layer. An active optical device such as a diode laser, optical
amplifier, optical modulator, light detector or any other device is
mounted on the metal layer. The active device has a first optical
axis and in this arrangement this axis is parallel to the
substrate. In addition, the metal layer provides an electrical path
to the active device. It is essential for the optoelectronic
assembly that the metal layer have a well-defined thickness. This
thickness is selected such that the heat generated by the active
device is dissipated, the substrate does not interfere with the
propagation of light along the first optical axis, and such that
the in-plane coefficient of thermal expansion (CTE) of the metal
layer is constrained by the substrate.
[0020] Suitable thicknesses of the metal layer are between 50 .mu.m
and 1,000 .mu.m. The metal layer is preferably made of a material
selected from the group consisting of copper, copper-tungsten,
copper alloys, copper composites, copper plated materials, copper
clad materials, and copper laminates. The actual material selected
will depend upon the purpose of the assembly and design
specifications. The insulating substrate is thick enough to lend
mechanical stability to the assembly, is a good dielectric for high
speed electronics, and bonds well with metal. Alumina, BeO and AlN
are good examples.
[0021] The optoelectronic assembly of the invention further admits
a passive optical element with a second optical axis positioned in
one of the selected regions. The positioning of this passive
optical element should be such that the second optical axis is
aligned with the first optical axis of the active device. Any
passive elements such as optical fibers, lenses, filters,
diffraction gratings, beam splitters, isolators and mirrors may be
placed on the substrate.
[0022] In the preferred embodiment, the optoelectronic assembly
incorporates square lenses and square isolators. An example of
square lenses are graded index of refraction (GRIN) lenses, such as
Gradium lenses. The metal layer is patterned as an electrical
circuit and electrical components are mounted on the metal layer.
This is particularly advantageous for highly integrated packages
which include all three types of components, i.e. active and
passive optical elements and electrical parts.
[0023] A bottom metal layer can be bonded to the insulating
substrate opposite the planar surface. This second layer can also
be patterned as an electrical circuit. In one embodiment the
assembly with such a second layer has cooling elements mounted on
the opposite side of the substrate. These cooling elements may
include Peltier coolers. Micro-channel coolers using fluid coolants
can also be patterned in the second layer of metal.
[0024] The method of making the optoelectronic assembly includes
the steps of selecting the metal layer and substrate and bonding
the metal layer to the substrate. In the patterning step, the metal
layer is formed by photolithography, etching, stamping,
electrodeposition or electro-discharge machining before bonding it
to the substrate. Alternatively, photolithography or diamond sawing
is used to pattern the metal layer after the bonding step. The
desired thickness is achieved by plating the metal layer on a thin
metal film on the insulating substrate, or by a polish-back step.
The active device may then be mounted on the metal layer.
[0025] During placing, the passive optical element is actively
displaced in the selected region or is allowed to come to rest at
the intended location under the force of gravity. In either case,
the passive element moves along the step which serves as a
mechanical fiduciary.
[0026] The method of the invention is particularly well-suited for
batch processing. The assemblies are formed by patterning the metal
layer on one large, flat insulating substrate which is then divided
into at least two parts. Preferably, many assemblies are
simultaneously produced. Bonding of the active devices on the metal
layers can be performed before or after the dividing step.
Likewise, the placement of passive elements can be performed before
or after the dividing step, depending on the production
process.
[0027] In still another embodiment, the active optical device is
mounted on a submount and the step serves as a mechanical fiduciary
to align the active optical device on this submount with any second
active device or passive elements. The thickness of the submount is
selected such that the optical axes of the components are aligned
and the heat generated by the first active device is dissipated
through the submount. A corresponding method is also a part of the
invention.
[0028] The details of the invention will be better appreciated upon
reading the detailed description below and referring to the drawing
figures.
BRIEF DESCRIPTION OF THE FIGURES
[0029] FIG. 1 is an isometric view of an optoelectronic assembly
according to the invention.
[0030] FIG. 2 is an isometric view of active and passive optical
elements in the optoelectronic assembly of FIG. 1.
[0031] FIG. 3 is an isometric view of feed through lines of one
embodiment.
[0032] FIG. 4 is a cross sectional view along a first optical axis
of the view in FIG. 2.
[0033] FIG. 5 is a three dimensional view showing the placement of
passive elements according to the invention.
[0034] FIG. 6 is a cross-sectional view of an assembly with a
substrate made of a metal base layer and a thin dielectric layer on
top of the base layer. The patterned metal layer is bonded on top
of the dielectric layer.
[0035] FIG. 7 illustrates a batch manufacturing method according to
the invention.
[0036] FIGS. 8A-B are cross sectional views showing the fabrication
steps for producing patterned metal layers.
[0037] FIG. 9 is a top plan view showing a metal layer patterned as
a circuit.
[0038] FIG. 10A is a cross sectional view of an embodiment with
cooling elements.
[0039] FIG. 10B is a cross sectional view of an embodiment with
channels for cooling fluid.
[0040] FIG. 11 is an isometric view of a portion of an assembly in
which the active device is a submount.
[0041] FIG. 12 is an isometric view illustrating the positioning of
the active device of FIG. 11.
[0042] FIG. 13 is an isometric view illustrating the positioning of
a free-standing active device in an assembly portion similar to
that of FIG. 11.
[0043] FIG. 14 is a side view of an embodiment wherein the active
optical element (laser) is edge mounted. This embodiment can be
used as a subassembly.
DETAILED DESCRIPTION
[0044] A preferred embodiment of an optoelectronic assembly 10 is
illustrated in FIG. 1. Assembly 10 is built on an insulating
substrate 12 which provides mechanical stability to entire assembly
10. The particular assembly 10 shown is a laser-fiber coupler as is
commonly used to send optical signals. Suitable materials for
substrate 12 are ceramics such as alumina, BeO and AlN. The overall
stability and size restrictions on assembly 10 will dictate the
thickness and material of choice of substrate 12 in specific
instances.
[0045] A metal layer 14 is bonded on a planar face 11 of substrate
12. Suitable metals include copper, copper-tungsten, various copper
alloys, copper composites, copper plated materials, copper clad
materials, and copper laminates. Copper composites may be
copper-molybdenum laminates or copper materials made from mixed
copper powders. Metal layer 14 is preferably made of copper because
of this metal's high electrical and thermal conductivity, low cost,
and because it is easily direct bonded to alumina. The thickness T
of layer 14 (see FIG. 3) ranges from 50 .mu.m to 1,000 .mu.m and
most preferably is equal to about 250 .mu.m. The actual thickness
is determined according to rules described below.
[0046] Metal layer 14 is patterned to accommodate an active optical
device 16 in the middle of assembly 10. In this case, optical
device 16 is an edge-emitting diode laser with one terminal. In
general, however, device 16 can be any optically active device such
as an optical amplifier, an optical modulator or a light detector.
Further, optical device 16 may have two, three or more input
terminals. Passive optical elements 22, 24 are positioned in region
28 where substrate 12 is exposed. In this case, optical element 22
is a square isolator and optical element 24 is a spherical lens or
square lens, such as a Gradium lens.
[0047] FIG. 2 affords a more detailed look at the positioning of
these elements. Laser 16 is mounted on portion 30 such that an
active edge 32 of laser 16 is at the height of a top surface 34 of
metal layer portion 30. Thus, first optical axis 36 along which
light propagates from laser 16 is elevated above planar face 11 by
a distance equal to the thickness of the metal layer portion 30
plus the depth of the laser portion in the laser die. In this way,
the height of first optical axis 36 is selected by the thickness of
metal layer portion 30. In many applications this is convenient
because it is relatively easy to determine the precise thickness of
metal layer portion 30. Furthermore, optical axis 36 is parallel to
planar face 11. The laser cavity need not be at the surface of the
die, however.
[0048] A metal layer portion 54 bears an integrated circuit 56 for
controlling the laser. Signals are sent to the laser 16 via a lead
58. Specifically, lead 58 is connected to the first terminal of
laser 16. Another lead 60 is connected to metal layer portion 30.
Because portion 30 is made of a conductive metal, it provides an
electrical path to laser 16 and in particular to the second
terminal of laser 16. Thus, the two electrical connections required
to control laser 16 are easily achieved. Since a person of average
skill in the art will know how to properly connect the terminals of
laser 16, they are not shown in the drawings.
[0049] Referring again to FIG. 1, an optical output fiber 72 has an
optical axes aligned with optical axis 36. Fiber 72 can be
single-mode or multi-mode depending upon the application. It is
commonly known that more precise alignment tolerances
(approximately 1 .mu.m) are required for single-mode fibers than
for multi-mode fibers (up to 10 .mu.m). Fiber 72 is mounted on
metal layer 14 by any suitable means. In the preferred embodiment,
fiber 72 is placed in a groove and bonded to the groove. This is a
standard technique well-known in the art.
[0050] A set of leads 80, 82 and 84 provide electrical connections
to integrated circuit 56. To ensure operation at very high speeds,
portions 90, 92 and 94 are preferably electrical transmission
lines. Such transmission lines are necessary when laser diode 16
requires modulation or control signals in the GHz range.
Furthermore, portions 90, 92 and 94 terminate at the right edge of
substrate 12 in pads 100, 102 and 104 to which are soldered contact
straps or pins 110, 112, and 114, respectively. Similarly, lead 60
is connected via metal layer portion 96, which is preferably also a
transmission line, to pad 106 and pin 116. Two additional pins 117
and 118 are provided on pads 107 and 108 of metal layer portions 97
and 98.
[0051] In another embodiment, as shown in FIG. 3, feedthrough leads
81, 83 and 85 are used instead of leads 80, 82 and 84. Feedthrough
leads 81, 83 and 85 connect integrated circuit 56 to an electrical
transmission substrate 87 located beneath integrated circuit 56,
removing the need for electrical transmission lines 90, 92 and 94;
pads 100, 102 and 104; and straps or pins 110, 112 and 114.
Likewise, feedthrough lead 61 replaces lead 60 and removes the need
for electrical transmission line 96, pad 106 and pin 116.
[0052] Referring again to FIG. 1, non-resident or external
circuitry (not shown) communicates with integrated circuit 56 of
assembly via pins 110, 112 and 114. Direct communication with laser
16 is obtained via pin 116. In the present case pins 117 and 118
are not required for operating assembly 10.
[0053] It is essential for proper operation of assembly 10 that
metal layer 14 and especially metal layer portion 30, have a
certain thickness T in the range from 50 to 1,000 .mu.m. In
particular, thickness T is set to optimize three different
parameters. The cross sectional view of FIG. 4 best explains the
rationale in selecting thickness T in a specific case.
[0054] First, the heat generated by active device 16 must be
dissipated by conducting it through metal layer portion 30 to
substrate 12. In general, the effectiveness of metal layer portion
30 in its function as -a heat sink will improve with increasing
thickness T. A list of thermal conductivity and thermal expansion
values for materials used as metal layer 14 (specifically, portion
30), active device 16, and substrate 12 is provided in Table 1.
1TABLE 1 Coefficient of Thermal Expansion, Thermal Conductivity
Material CTE, x 10.sup.-6/.degree. C. W/m - .degree. C. Alumina 7.0
18.0 BeO 6.4 280 Copper 17.5 392 Copper/Tungsten 7.0 248 GaAs 6.5
53 Silicon 4.2 130
[0055] Second, thickness T of metal layer 14 has to be large enough
such that substrate 12 does not interfere with the propagation of
light along first optical axis 36. Thus, when a light beam 120 is
emitted by laser 16 along first optical axis 36 at a divergence
angle a the light should not scatter off substrate 12 before
reaching isolator 22. This restriction is purely geometrical and
dictates a minimum thickness T dependent upon angle .alpha. and the
distance from the emitting edge of laser 16 to isolator 22.
Preferably, the thickness of the metal layer 14 (portion 30) is
such that the laser 16, isolator 22 and lens 24 share a common
optic axis.
[0056] Third, thickness T should be sufficiently thin for substrate
12 to constrain the thermal expansion of metal layer portion 30
during changes in temperature. In particular, the in-plane
coefficient of thermal expansion (CTE) of portion 30 should be
constrained to the CTE of substrate 12. Generally, the metal layer
14 will have a CTE greater than the CTE of the substrate.
[0057] The above three conditions are satisfied, for example, when
metal layer 14 is made of copper, thickness T of metal layer
portion 30 is about 250 .mu.m, the distance to isolator 22 is 300
.mu.m, and substrate 12 is made of alumina. From Table 1 it can be
seen that the CTE of alumina is much less than the CTE of copper.
Since thickness T of metal layer portion 30 is considerably smaller
than the thickness of substrate 12, and the insulating substrate 12
is stiffer than the metal layer 14, metal layer portion 30 will be
prevented from expanding in- plane, i.e. parallel to substrate 12,
at its natural rate during temperature cycling. In other words, the
copper making up metal layer portion 30 will inherit the in-plane
CTE of the underlying alumina making up substrate 12. Thus, by
selecting a particular material for substrate 12, metal layer 14
(portion 30) can be forced to have a desired in-plane CTE.
[0058] The advantage of constraining the in-plane CTE of metal
layer 14 is that misalignments between optical components during
operation are prevented. In particular, laser 16 will not undergo
any relative position changes with respect to lens 22 positioned on
substrate 12. Analogously, the alignment of laser 16 with isolator
22, lens 24 and fiber 72 (see FIG. 1) will be preserved. Another
advantage is that the GaAs chip is not stressed, improving
reliability. The mechanical stability of optoelectronic assembly 10
is also improved.
[0059] Most optoelectronic components, such as laser 16, are stress
sensitive and are made from compound semiconductors that have CTEs
that are close to the CTE of alumina. The CTE for GaAs, for
example, is given in Table 1. Since metal layer 14 inherits the
in-plane CTE of alumina, optoelectronic components made from
compound semiconductors can be safely mounted on layer 14. Thus, in
general, the in-plane CTE matching technique allows for the
relatively stress free bonding of large area electronic and
optoelectronic components to metal layer 14. In fact, referring
back to FIG. 2, integrated circuits 56 made from silicon or GaAs
are safely mounted on metal layer portion 54 and experience reduced
thermal expansion-induced mechanical stress.
[0060] Integrated circuit 56 residing on metal layer portion 54 is
driven by external circuitry (not shown) which sends appropriate
control signals through pins 110, .112, 114, electrical micro-wave
waveguides 90, 92, 94 and leads 80, 82 and 84. An electrical path
is also provided via pin 116, transmission line 96, lead 60 and
metal layer portion 30 to the second terminal of laser 16. Thanks
to this construction electrical, control signals in the GHz
frequency range can be delivered to control the gain and other
working parameters of laser 16.
[0061] During operation laser 16 generates heat which is dissipated
through metal layer portion 30, as shown in FIG. 4. The light beam
120 propagates along first optical axis 36 at divergence angle
.alpha. and is focused by isolator 22 and lens 24 into fiber 72.
Due to CTE matching between metal layer 14 (specifically metal
layer portion 30) and substrate 12, the in-plane thermal expansion
of portion 30 is constrained by substrate 12. Thus, misalignment is
prevented between laser 16, isolator 22 and lens 24 as well as
fiber 72.
[0062] In sum, optoelectronic assembly 10 is compatible with
high-speed electrical operation and exhibits superior heat
dissipation capability together with mechanical stability over a
wide temperature range. Furthermore, assembly 10 is simple in
construction.
[0063] The following embodiments serve to better illustrate in a
general manner several important aspects of the invention. It will
be appreciated by those skilled in the art that these aspects can
be applied in any particular optoelectronic assembly.
[0064] An arrangement for aligning multiple laser beams 320 with
multiple lenses is shown in FIG. 5. A single-piece lens 318 with
multiple lens facets 316 is used. The lens 318 can slide on the
surface 304 of the insulating substrate until it is abutted against
the step 310 formed by the patterned metal layer 314. The patterned
metal layer 314 has a cut-in section to accommodate the lens
318.
[0065] This method of placing, guiding, and aligning passive
optical elements is convenient and compatible with pick and place
techniques. A pick and place machine operates by manipulating a tip
that can. grasp the optoelectronic and optical components. The tip
moves rapidly to place the components in their proper locations on
the optoelectronic assemblies. The pick and place machine slides
the components until they abut against the side walls of the metal
layer. In this way, at least coarse alignment is provided. Coarse
alignment reduces the time and expense of active alignment.
[0066] The components can be attached to the assembly with a thin
layer of hard solder. Preferably, the components are soldered
individually using localized heating techniques. Focused laser
beams, for example, may be used to heat small regions of the
assembly to solder the components individually. Also, resistive
heating elements may be used. Resistive heating elements are thin
film resistors that heat up when a current is passed through them.
If the heating elements are placed in locations on the assembly
where components are to be soldered, then they can be independently
controlled to solder individual components. Both laser heating and
resistive element heating are well known in the art. Alternatively,
the components can be attached with a glass solder or a glue such
as epoxy. Such attaching techniques are well known in the art.
[0067] According to a preferred method of the invention,
optoelectronic assemblies are produced in a batch process. This
method is similar in many respects to that used in the manufacture
of integrated circuits, where a large wafer containing many
complete circuits is fabricated and then cut into individual chips.
In batch production according to the invention, an insulating
substrate 450, as shown in FIG. 7, is selected from the group of
suitable materials listed above. Metal layers 452 are then bonded
to substrate 450. The thicknesses of metal layers 452 ranges
between 50 and 1,000 .mu.m. Precise thickness values are fixed
according to the criteria required for heat dissipation,
non-interference with light propagation and in- plane CTE
constraint by substrate 450.
[0068] In fact, metal layers 452 include at least a first metal
layer 452A and a second metal layer 452B. For practical
applications, first and second metal layers 452A and 452B are
identical, since the batch process is designed to fabricate a large
number of identical optoelectronic circuits. Metal layers 452A and
452B will generally consist of numerous metal layer portions to
accommodate the desired active optical devices, passive optical
elements and electronic components.
[0069] Metal layers 452 can be produced and patterned by well-known
techniques such as: photolithography, etching, stamping, and
electro-discharge machining (EDM). Any of these processing
techniques can be used to pattern metal layers 452 prior to bonding
to substrate 450. Alternatively, metal layers 452 can be produced
and patterned by photolithography or diamond sawing after bonding
to substrate 450. The individual pre-cut parts can also be plated
and brazed.
[0070] To produce individual optoelectronic assemblies 454,
insulating substrate 450 is divided. The dividing step is performed
with any suitable dicing apparatus. In FIG. 7 a cutting beam 456
delivered from a high power laser (not shown) is used to divide
substrate 450 along a line 458. Cut assemblies 454A and 454B
correspond to portions of substrate 450 bearing metal layers 452A
and 452B.
[0071] Of course, metal layers 452 expose selected regions where
substrate 450 is free for placing passive optical elements and
other locations for mounting active optical devices and electrical
components on the metal. According to the invention, metal layers
452A and 452B of assemblies 454A and 454B are fitted with the
required active optical devices, passive optical elements and
electrical components either before or after the dividing step. For
example, active optical devices and electrical components can be
mounted on the metal layers before dicing, and passive optical
elements can be placed on cut assemblies 454A and 454B. The choice
of when the individual devices and elements are mounted is up to
the designer and can be optimized to ensure efficient batch
processing.
[0072] A simple way of preparing a substrate 508 is illustrated in
FIGS. 8A and 8B. First, substrate 508 is provided with a top metal
layer 530 and a bottom metal layer 540, as shown in FIG. 8A. In a
processing step, e.g., etching, top layer 530 is patterned to yield
metal layer portions 510A and 512A, and bottom layer 540 is
patterned to leave block 520A. This method is advantageous because
the patterning of portions 510A and 512A can be carried out
independently and simultaneously with the making of block 520A. In
another embodiment of the invention intended for high-frequency
operation, bottom metal layer 540 can be left intact or merely
etched to a desired thickness to provide a ground plane. The actual
pattern of the ground plane will be selected based on
application.
[0073] The substrate used in this invention may not be entirely
insulating. FIG. 6 shows a cross sectional side view of an assembly
with a substrate composed of a thin layer of dielectric material
710 bonded to an underlying thicker base layer of metal 712. The
metal base layer 712 provides the necessary mechanical rigidity.
The patterned metal layer 714 is bonded to the dielectric layer.
Such substrates are commercially available from The Bergquist
Company in Minneapolis, Minn. The metal base layer 712 can be
copper, aluminum, invar, or other metals and the dielectric layer
710 is preferably ceramic. In this embodiment, the dielectric layer
710 is substantially thinner than the metal base layer 712. The
standard metal base layer thickness for these composites is 1-3 mm.
The patterned metal layer 714 is the same thickness as in the other
embodiments, 50-1000 .mu.m.
[0074] In one advantageous embodiment illustrated in the plan view
of FIG. 9, a bottom metal layer 556 is patterned as an electrical
circuit 554 on a bottom surface 552 of an insulating substrate 550.
Electrical components required for the operation of the
corresponding optoelectronic assembly can thus be mounted on bottom
surface 552. Clearly, it is also possible to pattern bottom surface
552 for mounting additional optoelectronic devices.
[0075] FIG. 10A illustrates how an optoelectronic assembly 600
according to the invention is cooled. Cooling elements 601 are
mounted between a bottom metal layer 602 on the underside of a
substrate 608 and a metal layer 604 on a separate base 606. For
example, cooling elements 601 can be Peltier elements or any other
electronic devices capable of dissipating heat from substrate 608.
Contacts 610, 612 are provided for applying suitable voltages to
elements 601 and result in the cooling of assembly 600 during
operation. The details of mounting and driving cooling elements are
well-known in the art.
[0076] FIG. 10B illustrates how the optoelectronic assembly 600 can
also be cooled by liquid coolants flowing through channels 605
etched in the bottom metal layer 602. Of course, a cover 610 for
the channels 605 is required in this embodiment. One advantage of
optoelectronic assembly 600 is that is has a high hermiticity.
[0077] As shown in FIGS. 11 and 12, the present invention is also
well-suited for positioning an active optical device 622 on an
optoelectronic assembly 618 (only portion shown). In this
embodiment, active optical device 622 is delivered on a submount
620. This adaptation is very useful, since most equipment
manufacturers ship active devices, e.g., diode lasers, optical
amplifiers, optical modulators and light detectors, which are
premounted on submounts.
[0078] Assembly 618 has a metal layer 624 bonded on a planar
surface 626 of an insulating substrate 628. A step 630 is formed
between substrate 628 and the top surface of layer 624. A fiber 632
is mounted in a notch 634 machined in metal layer 624. In fact,
fiber 632 sits at a slight off-set from the edge of layer 624 for
better in-coupling of light from device 622.
[0079] During placement of optical device 622 in assembly 618, a
second optical axis 636 of fiber 632 is aligned with a first
optical axis 638 of device 622. Device 622 is oriented such that
axis 638 is parallel to surface 626 of substrate 628. For this
purpose step 630 is used as a fiduciary to slide submount 620 into
the aligned position illustrated in FIG. 11. The arrows in
[0080] FIG. 11 show how submount 620 is displaced on surface 626
using step 630. Of course, in this embodiment submount 620 is used
as part of the electrical path to device 622. Additional passive
elements having their own optical axes--i.e., a third optical
axis--can be positioned and aligned on substrate 628 with optical
axis 638 of device 622.
[0081] FIG. 13 illustrates how an active optical device 642 having
a first optical axis 646 and no submount can be positioned on a
metalized planar surface 640 of an optoelectronic assembly 650. The
metalization of surface 640 provides the necessary electrical path
to device 642. Otherwise, assembly 650 is identical to assembly 618
and the positioning steps are analogous to those illustrated in
FIGS. 11 and 12.
[0082] FIG. 14 illustrates a particularly useful assembly wherein
an optoelectronic component such as a laser diode 801 is edge-
mounted on the assembly. The edge 810 of the laser 801, edge 816 of
the metal layer 806, and edge 820 of the insulating substrate are
all flush. This arrangement can function as a sub-assembly and be
incorporated into a larger optoelectronic package. Preferably, this
embodiment incorporates the advantageous features described above.
Specifically, the metal layer 806 provides heat dissipation and has
a constrained CTE. Also, the metal layer 806 is patterned to
provide electrical contacts 808 for the laser 801. A wire 818
provides an electrical connection for the laser 801. Direct bond
copper-on-ceramic is a preferred material for this application. Of
course, these assemblies can be manufactured using batch processing
techniques.
[0083] It will be clear to one skilled in the art that the above
embodiments may be altered in many ways to produce an
optoelectronic assembly performing any desired function by
selecting the appropriate components. This will be done without
departing from the scope of the invention. Additional
modifications, such as providing multiple metal layers formed by
bonding additional metal layers on top of existing metal layers, or
patterning the additional metal layers before or after bonding, are
also envisioned by the invention. Accordingly, the scope of the
invention should be determined not by the examples given but by the
following claims and their legal equivalents.
* * * * *