U.S. patent number RE28,610 [Application Number 05/525,121] was granted by the patent office on 1975-11-11 for fine geometry solar cell.
This patent grant is currently assigned to Communications Satellite Corporation. Invention is credited to Joseph Lindmayer.
United States Patent |
RE28,610 |
Lindmayer |
November 11, 1975 |
**Please see images for:
( Certificate of Correction ) ** |
Fine Geometry Solar Cell
Abstract
A fine geometry solar cell having a top surface contact
comprising substantially more and finer metallic fingers spaced
close together for collecting photocurrent. Junction depth and/or
impurity concentration may be reduced significantly. The method for
making the fine geometry solar cell, comprises in ordered steps,
the processes of diffusion, oxidation, photolithography,
metallization and plating.
Inventors: |
Lindmayer; Joseph (Bethesda,
MD) |
Assignee: |
Communications Satellite
Corporation (Washington, DC)
|
Family
ID: |
22676700 |
Appl.
No.: |
05/525,121 |
Filed: |
November 19, 1974 |
Related U.S. Patent Documents
|
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
Issue Date |
|
Reissue of: |
184393 |
Sep 28, 1971 |
03811954 |
May 21, 1974 |
|
|
Current U.S.
Class: |
136/256;
148/DIG.33 |
Current CPC
Class: |
H01L
31/022433 (20130101); Y10S 148/033 (20130101); Y02E
10/50 (20130101) |
Current International
Class: |
H01L
31/0224 (20060101); H01L 031/02 (); H01L
031/04 () |
Field of
Search: |
;136/89,572 |
References Cited
[Referenced By]
U.S. Patent Documents
Other References
M Wolf Proceedings of the IRE July 1960 pp. 1246, 1,254-1,266.
.
Technical Report AFAPL-TR-65-8 2/1965 "Research on Thin Film
Polycrepballire Solar Cells," by Aldrich et al. FF, 125 126 pp.
129, 133..
|
Primary Examiner: Curtis; Allen B.
Attorney, Agent or Firm: Sughrue, Rothwell, Mion, Zinn &
Macpeak
Claims
What is claimed is:
1. A solar cell comprising a semiconductor material having top and
bottom surfaces and having a p-n junction at a distance of between
500 A and 2,000 A from the top semiconductor surface thereof, said
top surface being adapted to receive incident light radiation, an
electrode on said bottom semiconductor surface, and a patterned
electrode on said top semiconductor surface, said patterned surface
comprising a plurality of thin metallic fingers electrically
connected together, said thin metallic fingers being separated by
.Iadd.a.Iaddend. .[.distances.]. .Iadd.distance.Iaddend. .[.on the
order.]. of n .times. 10.sup.-.sup.2 centimeters where n is .[.a
non-zero integer.]. .Iadd.any number from one to nine..Iaddend.
2. A solar cell as claimed in claim 1 wherein said semiconductor
material is silicon.
3. A solar cell as claimed in claim 2 wherein said p-n junction is
at a depth of approximately 1,500 A and divides said semiconductor
material into a top surface layer and a bottom layer.
4. A solar cell as claimed in claim 2 wherein said thin metallic
fingers are separated by distances of approximately 0.03
centimeters.
5. A solar cell as claimed in claim 2 wherein said p-n junction
divides said semiconductor material into a top n-type layer and a
bottom p-type layer.
6. A solar cell as claimed in claim 2 wherein said material between
said top surface and said p-n junction has an impurity
concentration of about 10.sup.19 - 10.sup.20 impurity
atoms/cm.sup.3.
7. A solar cell as claimed in claim 2 wherein said thin metallic
fingers are spread substantially evenly over the surface of said
solar cell at a density in one dimension of approximately 30
metallic fingers per centimeter.
8. A solar cell as claimed in claim 2 wherein the width of said
fingers is between about 1-20 microns.
9. A solar cell as claimed in claim 3 wherein said thin metallic
fingers are separated by distances of approximately 0.03
centimeters. 10. A solar cell as claimed in claim 3 wherein said
p-n junction divides said semiconductor material into a top n-type
layer and a bottom p-type layer.
1. A solar cell as claimed in claim 3 wherein said material between
said top surface and said p-n junction has an impurity
concentration of about
10.sup.19 - 10.sup.20 impurity atoms/cm.sup.3. 12. A solar cell as
claimed in claim 3 wherein said thin metallic fingers are spread
substantially evenly over the surface of said solar cell at a
density in the dimension
of approximately 30 metallic fingers per centimeter. 13. A solar
cell as claimed in claim 3 wherein the width of said fingers is
between about 1-20
microns. 14. A solar cell as claimed in claim 9 wherein said p-n
junction divides said semiconductor material into a top n-type
layer and a bottom
p-type layer. 15. A solar cell as claimed in claim 14 wherein said
material between said top surface and said p-n junction has an
impurity
concentration of about 10.sup.19 - 10.sup.20 impurity
atoms/cm.sup.3. 16. A solar cell as claimed in claim 15 wherein
said impurity atoms are atoms selected from a group consisting of
phosphorous, arsenic and
antimony. 17. A solar cell as claimed in claim 16 wherein said thin
metallic fingers are spread substantially evenly over the surface
of said solar cell at a density in one dimension of approximately
30 metallic
fingers per centimeter. 18. A solar cell as claimed in claim 17
wherein the width of said fingers is between about 1-20 microns.
Description
BACKGROUND OF THE INVENTION
This invention relates to solar cells, and more particularly, to a
fine geometry solar wherein the surface through which light enters
comprises a substantial number of very fine metallic lines (or
pattern) which collect current.
The use of photovoltaic devices, commonly known as solar cells,
which convert light energy to useful electrical energy is well
known. Light entering these solar cells is absorbed, thereby
generating electron-hole pairs which are then spacially separated
by the electric field produced by the solar cell junction and are
collected at respective top and bottom surfaces of the solar cell.
For example, in an n-p type solar cell electrons will travel to the
top surface where they will then be collected by a metallic grid
positioned thereon. The metallic grid may typically comprise six
metallic fingers separated along the top surface by a relatively
large distance and connected to each other by a common bus bar. The
electrons will travel either directly to the metallic fingers or
approach the top surface between the fingers and then travel along
the surface of the solar cell until they can be collected by one of
the fingers. Holes, on the other hand, will travel to the bottom
surface of the solar cell where they may be collected by a metallic
sheet covering the entire bottom surface.
The six-fingered metallic grid is necessary at the top surface of
the solar cells in order to enable light to enter the solar cell.
However, one problem associated with the six-fingered construction
relates to the relatively large separation between the fingers.
Electrons which must travel along the surface to the metallic
fingers encounter a high surface resistance. Therefore, due to the
relatively long distance the electrons must travel before
collection, and due to the problem of surface resistance, a series
resistance may develop, thereby limiting the efficiency (electrical
power output/solar power input) of solar cells by limiting the
electrical power output.
The prior art has sought to obviate the above problem by diffusing
an impurity into the surface of the solar cell in a higher order
concentration, on the average of about 10.sup.15 atoms per square
centimeter or higher. Higher order concentration (i.e., heavier
diffusion) lowers the surface resistance but introduces other
problems. Higher order concentration of impurities is obtained by a
process known as "solid solubility diffusion," i.e., the solar cell
is allowed to assume as many impurities as it can on the surface,
e.g., approaching 10.sup.21 atoms per cubic centimeter. However, in
such a solid solubility process there is crystal lattice damage to
the solar cell which propagates deep into the solar cell substrate.
The efficiency of the solar cell is thereby reduced in two ways.
First, the damage to the crystal lattice causes a reduction in the
diffusion length or lifetime of minority carriers. This means that
holes, for example, in an n-type diffused region will recombine
with available electrons before they can be separated by the
junction. Secondly, as discussed below, damage to the crystal
structure affects the power output of the solar cell (which is
basically a diode) by "softening" the current(i)-voltage(v)
characteristics of the diode. In addition, the diffusion of such
higher order concentration of impurities creates a relatively deep
junction of about 4,000 A. This relatively deep junction means that
light of relatively short wavelengths (where solar energy peaks)
cannot penetrate beyond the junction, but is absorbed in the
diffused region (i.e., between the top surface and the junction).
Electron-hole pairs generated in the diffused region have a
relatively short diffusion length (even if there were no crystal
lattice damage) and therefore will largely recombine before
separation by the junction.
The present invention has the advantage of improving the efficiency
of solar cells in the short wavelength, i.e., blue-violet portion
of the spectrum corresponding to 0.3-0.5 microns thereby sharply
increasing output power. The present invention also has the
advantage of enabling a degree of freedom in the design of solar
cells by reducing the junction depth and/or reducing the impurity
concentration while improving solar cell efficiency. In addition,
the effect of radiation damage to the solar cell is decreased with
improvement in efficiency in the short wavelength region. Also, the
use of specified metals for the metallic contact of the present
invention provides a moisture resistant contact.
An advantage of extremely shallow junctions wherein there is no
crystal lattice damage is that diodes become much more ideal. In
the simple junction theory or so-called "diffusion theory" the
following relation applies:
I=I.sub.o (e.sup.qV/kT - 1)
where I is the diode current, I.sub.o is the reverse diode current,
V is the applied voltage and kT/q is the thermal voltage. Actual
solar cells do not follow this relationship, but rather the
following:
I=I.sub.o (e.sup.qV/nkT -1)
where n is a quantity greater than unity. In conventional solar
cells, n .congruent. 2 while of course in the ideal case, n = 1.
This fact "softens" the I-V characteristics of solar cells. As
expressed in terms of the fill factor
F=actual power to load/short circuit current .times. open circuit
voltage
Defined in this fashion, conventional solar cells show an F of
about 72 percent. With the extremely shallow diffusion and reduced
impurity surface concentration practiced by the present invention
as described below, an n value of about 1.1 and F approaching 80
percent may be obtained. These numbers represent an almost ideal
junction.
SUMMARY OF THE INVENTION
The invention comprises any type of solar cell, e.g., silicon or
gallium arsenide, having a top surface current collector comprising
a significantly greater number of fine metallic fingers (or other
fine geometric pattern) wherein the physical separation between the
fingers and the width of each finger is substantially reduced. The
junction depth and/or impurity concentration is reduced in
accordance with the degree of freedom provided by the use of the
fine geometry cell. The solar cell is made by first introducing
impurities into, for example, a silicon slice, and then oxidizing
the solar cell. Then, the fine metallic pattern is placed on the
top surface of the solar cell using a photolithography technique.
Finally, a plating process is used to build up the fingers of the
metallic pattern to a proper thickness.
BRIEF DESCRIPTION OF THE DRAWINGS
FIG. 1 is a general block diagram of a side view of a solar cell
having metallic fingers located on the top surface.
FIG. 2 is a diagram on the standard geometry six-fingered contact
used on the top surface of the solar cell of FIG. 1.
FIG. 3 is a diagram of the fine geometry metallic contact of the
present invention used on the top surface of the solar cell of FIG.
1.
FIG. 4 is a graph of efficiency vs. surface concentration of
diffsued layer for a silicon solar cell comparing the prior art
six-fingered geometry with the fine geometry of the present
invention.
DETAILED DESCRIPTION OF THE DRAWINGS
Referring to FIG. 1 there is shown a side view of a typical solar
cell. There will be described a single crystal, n-p silicon solar
cell though the invention has applicability to all types of single
crystal solar cells including, for example, GaAs solar cells. The
term "single crystal" is well known in the art and refers to
lattices having absolute perfect crystallographic order, but as
described herein, also includes nearly single crystal cells which
are almost perfectly crystallographic. In addition the inventive
concepts are not limited to single crystal solar cells but may be
applied to thin film solar cells.
The single crystal silicon solar cell comprises a silicon substrate
1 of p-type material and a silicon layer 2 of n-type material with
an n-p junction 3 positioned a predetermined depth below the top
surface of silicon layer 2. In an n-p silicon solar cell the
junction 3 will produce an electric field directed towards the
substrate 1 thereby resulting in generated electrons flowing to the
top of surface 2 with holes flowing to the bottom of substrate 1
wherein the holes may be collected by a contact 4 covering the
entire back of the bottom surface of layer 1.
The metallic grid pattern 5 used for collection of the electrons
flowing to the surface through which light enters is positioned on
top of silicon layer 2. In the prior art solar cells the grid
pattern 5 may comprise a six-fingered metallic contact of a type
shown in FIG. 2. For a solar cell having dimensions 2 .times. 2 cm.
each metallic finger is approximately 0.30 centimeters apart with
each finger having a width of about 300 microns. The entire
metallic grid would block between 8-10 percent of the light falling
on the solar cell. The metallic grid of the present invention,
however, as seen in FIG. 3, comprises for a 2 .times. 2 cm solar
cell approximately 60 metallic fingers wherein the separation
between each finger is approximately 0.03 centimeters with each
metallic finger being between 1-20 microns in width. The fine
geometry configuration of the present invention would block less
than 10 percent of the solar light. The fine metallic fingers may
lie parallel to the main busbar and be connected thereto by
tapered, intermediate buses as shown in FIG. 3, or alternately the
fine metallic fingers may all lie perpendicular to the main busbar
and be directly connected thereto, in the manner shown in FIG.
2.
With the fine geometry solar cell it is now possible to lower the
impurity concentration and/or decrease junction depth. If the
junction depth is decreased (e.g., by shortening diffusion time
and/or lowering diffusion temperature) the efficiency will be
improved in three ways. First, more short wavelength light will
penetrate beyond the junction 3 to the p-type silicon substrate 1
to generate electron-hole pairs in substrate 1. Electronhole pairs
generated in substrate 1 have a longer lifetime than electron-hole
pairs generated in n-type layer 2. Secondly, though all the
electrons generated in the solar cell will encounter greater
surface resistance at the top of layer 2 with a lowering of surface
impurity concentration, the distance along the surface needed to be
traveled by the electrons prior to collection will be greatly
reduced. In the alternative, if the surface concentration is
lowered without decreasing junction depth, the resistance
encountered by the electrons will again be offset by the fine
geometry contact of FIG. 3. Also, in the latter case, though most
of the short wavelength light will generate electron-hole pairs in
diffused layer 2, due to a lowering of the impurity concentration
in layer 2 the lifetime of holes will be increased.
The manner in which the n-p, silicon solar cell of the present
invention having a reduced junction depth is made will now be
described. First a p-type silicon piece is cut and polished into a
slab, for example, 2 .times. 2 cm. Then, n-type impurities, e.g.,
any of the elements from Group VA of the table of elements, such as
phosphorus, arsenic or antimony, are diffused into the p-type
substrate forming an n-p junction. Whereas the prior art has a
diffused junction depth of 4,000 A, the junction depth of the
present invention may be as shallow as 1,500 A. To acquire this
junction depth of 1,500 A with phosphorus, the phosphorus is
diffused into the p-type substrate at about 750.degree. to
825.degree. C for about 5-10 minutes. The diffusion gas having the
impurities comprises O.sub.2, N.sub.2 and PH.sub.3 (source of
phosphorus), and is fed into the diffusion furnace at a rate of
1,000 cc/min. for N.sub.2 ; 500 cc/min. of 99% Argon, 1% PH.sub.3 ;
and 75 cc/min. of O.sub.2. The volume concentration of phosphorus
in the surface layer would be of the order of magnitude of
10.sup.19 or 10.sup.20 atoms/cubic centimeter. If arsenic or
antimony were used then the time and temperature of diffusion would
be changed as would be known, to acquire a junction depth of 1,500
A.
After diffusion, the n-p silicon material is exposed to steam for
about 2 minutes at 800.degree.C. This results in the formation of
1,000 A of SiO.sub.2 (glass) extending from the top surface of the
n-type material. In the oxidation process approximately several
hundred (400-500) A of silicon are removed from the top of the
diffused layer which results in several advantages. First, during
the diffusion process approximately 450 A of the top surface is
damaged which results in a shortening of the lifetime of electrons
approaching the surface. Removal of the 400-500 A thereby improves
the lifetime of these electrons. Secondly, removal of the 400-500 A
of silicon further reduces the junction depth which means more
short wavelength light will propagate beyond the junction to
generate more carriers therein.
As an additional step in the process of making n-p solar cells of
the present invention, all or part of the 1,000 A of the SiO.sub.2
may be removed in a conventional manner. Full or partial removal of
the SiO.sub.2 would be desirable since the glass has an index of
refraction of only about 1.46 which means too much light will be
reflected from the surface of the solar cell.
The silicon slab is now ready to have the fine geometry pattern
placed on the top surface of the diffused layer 2. First, the top
surface is coated completely with a photoresist of any known type,
e.g., A-Z resist. Then the photoresist is exposed to light or an
electron beam through any desired mask having a fine pattern such
as the fine geometry pattern of FIG. 3. The method of masking a
fine lined mask is well known. The top surface is then developed
with any known developer used with the A-Z resist thereby forming
the pattern areas (i.e., the fingers) on the bare diffused silicon
layer.
Next, using a known vapor deposition technique, about 300 A of
chromium is evaporated on the entire top surface followed by the
evaporation of 2,000 A of Ag. The photoresist is then dissolved in
any known solvent used with the A-Z resist. The solvent "lifts off"
the photoresist, and, consequently, the metal on top of the
photoresist in areas (between the fingers) where photoresist is in
contact with the silicon. This lift off process is knownw as
"lift-off photolithography" and results in the fine metallic
geometry pattern, i.e., the finger pattern shown in FIG. 3,
positioned on the top of the diffused layer 2. Finally, to build up
the thickness of each of the metallic fingers to about 20 microns
for purposes of good conductivity, silver is plated, in a known
manner, onto the metallic fingers.
An alternative photolithography technique for forming the fine
geometry pattern would comprise the ordered steps of (1)
evaporating a metal, e.g., chromium, over the top surface; (2)
covering the metal with photoresist; (3) exposing the photoresist
to light or an electron beam through a mask; (4) developing the
photoresist; (5) etching off the metal in the area between the
fingers; (6) using a solvent to dissolve the residual
photoresist.
The effect of radiation damage to solar cells is reduced with the
fine geometry solar cell of the present invention. Radiation damage
to a solar cell affects the response of the cell to longer
wavelengths. The present invention, by obtaining more energy output
from the short wavelength region than prior art solar cells, has
therefore reduced the overall effect of radiation damage.
Referring to FIG. 4 there is shown a graph of the efficiency of a
solar cell with respect to the surface concentration of diffused
layer 2. As can be seen, the efficiency obtainable with the fine
geometry solar cell is approximately 50 percent greater than the
efficiency obtainable with the standard six finger geometry solar
cells.
* * * * *