Thin-film

Thin -film CdS/CulnSe2

CdS /CuInSe2 heterojunction solar
solar cell
cell

Wen
Wen S.S. Chen,
Chen, Reid
Reid A.A. Mickelsen
Boeing
Boeing Aerospace
Aerospace Company,
Company, Mail
Mail Stop
Stop 88-43,
88 -43, P.O.
P.O.Box
Box3999,
3999,Seattle,
Seattle,Washington
Washington 98124
98124

Abstract
Abstract

The
The development
development of of aa polycrystalline,
polycrystalline, thin
thin-film
-film solar
solar cell
cell utilizing
utilizing aa heterojunction
heterojunction structure
structure based
based
upon N-type
upon N -type CdS
CdS and
and P-type
P -typeCuInSe2
CuInSe2 semiconductor
semiconductor materials
materials isis described. TheThe cell,
cell, prepared
prepared entirely
entirely byby
vacuum
vacuum deposition
deposition andand sputtering
sputtering techniques
techniques onto
onto inexpensive
inexpensive substrates,
substrates, has
haspotential
potentialapplications
applicationsasasa alowlow-
-
cost
cost mass
mass produced
produced device for photovoltaic
device for photovoltaic power
power generation
generation systems.
systems. AA device efficiency
efficiency of of 7.5%
7.5% under
under
simulated AM-1-1 illumination
simulated AM illumination is
is reported.
reported. Material
Material andand device properties pertinent
device properties pertinent toto the
the development
development of of the
the
high
high efficiency cell are
efficiency cell are reviewed.
reviewed. The electrical, optical,
The electrical, optical, and
and structural
structural properties
properties ofofthe
thedeposited
depositedthinthin-
-
film materials
film materials areare described.
described. Results
Results of
of detailed
detailed cell
cell characterization using aa variety
characterization using variety ofof electrical,
electrical,
optical, and
optical, and thermal
thermal measurements are presented
measurements are presented and
and analyzed
analyzed inin terms
terms of
of aa photovoltaic
photovoltaic cell
cell model
model dominated
dominated
by
by interface
interface state recombinations. Finally,
state recombinations. Finally, the projected, realistically
the projected, realistically achievable
achievable performance
performance of of this
this
thin-film
thin -film cell
cell is
is discussed.
discussed.

Introduction
Introduction

Ternary chalcopyrite
Ternary crystals have
chalcopyrite crystals have considerable
considerable technological
technological interest
interest because
because of
of their
their promise
promise for
for
application in
application in aa variety
variety of
of opto-
opto-electronic
electronic device such as
device such as light
light emitting
emitting diodes,
diodes, infrared
infrared defectors,
defectors, optical
optical
parametric oscillators,
oscillators, upconverters
upconverters andand for
for infrared
infrared generation.'
generation. 1 Recently,
Recently, the
the potential
potential utilization
utilization of
of
I-III-VI2
I- III -VI2semiconductors
semiconductors as asalternatives
alternativestotopresent
presentphotovoltaic
photovoltaicmaterials
materials has
has been
been demonstrated.
demonstrated. TheThe
report
report of
of aa 12%
12% efficient,
efficient, epitaxially-grown
epitaxially -grownCdS CdSonon single
singlecrystal
crystal CuInSep
CuInSe2 device 2 is
device2 is particularly
particularly encouraging
encouraging
since
since this
this is
is the
the only
only case
case ofof aa photovoltaic
photovoltaic device, other than
device, other than the
the Si,
Si, GaAs,
GaAs, and
and InP
InP/CdS
/CdS semiconductor
semiconductor
systems, having solar
systems, having solar energy
energy conversion
conversion greater
greaterthan
than1010%.
%.

Properties of potential
Properties of potential solar
solar cell
cell I-I-III-VI2
III -VI2 materials
materials are
are listed in Table
listed in Table I.
I. These
These materials
materials all
all have
have
band
band gaps
gaps near
near the
the optimum
optimum value terrestial solar
value for terrestial solar energy
energy conversion.3
conversion. 3 They
They are
are direct
direct band
band gap
gap semi-
semi-
conductors
conductors which
which minimizes
minimizes the requirement for
the requirement for long
long minority
minority carrier
carrier diffusion
diffusion lengths.
lengths. Except
Except for
for CuGaSe2,
CuGaSe2,
which has
which has only
only exhibited
exhibited PP-type
-type behavior,
behavior, the
the other compounds have
other three ternary compounds have been
been grown as both
grown as both N-
N- and
and
P-type
P -type crystals.
crystals. Thus construction
Thus construction ofof homojunction
homojunction devices
devices isis possible.
possible. These
These chalcopyrite
chalcopyrite compounds
compounds can
can
also be
also be paired
paired with CdS to
with CdS to potentially make efficient
efficient pp-n-n heterojunction
heterojunction solar
solar cells
cells because
because they
they have
have
compatible lattice
compatible structures with
lattice structures with acceptable
acceptable lattice
lattice mismatches,
mismatches, andand favorable
favorable differences
differences of
of electron
electron
affinities.
affinities.

Table 1.
Table 1. Properties of
Properties of Potential
Potential Solar
Solar Cell
Cell I-
I-III-IV2
III -IV2 Materials
Materials
Mobility
Mobility
Lattice
Lattice 0 Mismatch (cm 2 V-
Mismatch (cm- V^S' 1)
-S-1)
Eg(ev)
Eg(ev) Constant
Constant (Á)(A) With CdS
With CdS (300°K)
(300 °K) Electron
Material
Material (300°K)
(300 °K) Transition
Transition a
a cc (%)
( %) nn p
p Affinity (ev)
Affinity (ev)
CuGaSe
CuGaSe2 2 1.68
1.68 direct
direct 5.618
5.618 11.01
11.01 3.8
3.8 -- 20
20
CuInS
CuInS22 1.55
1.55 direct 5.523
5.523 11.12
11.12 5.56 200 15
15
CuInSe2 1.04
1.04 direct
direct 5.782
5.782 11.62
11.62 1.16
1.16 320 10
10 4.58
CuInTe
CuInTe22 0.96
0.96 direct 6.179 12.36
12.36 5.62
5.62 200 20
20
CdS
CdS 2.42
2.42 direct 4.136
4.136 6.716 --
-- 250
250 -- 4.5
4.5

Reported
Reported performances
performances of solar cells
of solar cells based
based on
on the
the ternary
ternary compounds
compounds are
are listed
listed in
in Table
Table 2.
2.
Table 2.
Table 2. Reported
Reported Photovoltaic
Photovoltaic Efficiency

Efficiency (%)
Efficiency ( %)
Material Single Crystal
Single Crystal Thin-Film
Thin -Film

CdS/CnGaSe
CdS /CnGaSe22 54 --
CdS/CuInS 2
CdS /CuInS2 -- 3.25
3.2555
CdS/CuInSe
CdS /CuInSe22 12
1222 6.6,
6.6,55 5.76
5.7 6
CdS/CuInTe
CdS /CuInTe22 No
No significant
significant photovoltaic
photovoltaic effect
n, pp CnInS2
n, CnInS 2 -- 3.6
3.655
n, pp CuInSe2
n, CuInSe 2 3.0
3.055

62
62 / SPIE Vol
/ SPIE Vol. 248
248Role
RoleofofElectro-
Electro-Optics in Photovoltaic
Optics in Photovoltaic Energy
Energy Conversion
Conversion (1980)
(1980)

Downloaded From: http://proceedings.spiedigitallibrary.org/ on 05/19/2015 Terms of Use: http://spiedl.org/terms

 THIN-FILM CdS/CulnSe2
THIN -FILM CdS /CuInSe2 HETEROJUNCTION CELL
SOLAR CELL
HETEROJUNCTION SOLAR

The high efficiency of

The of the single crystal
the single device as
crystal device well asas the
as well thin-film
polycrystallinethin
the polycrystalline -film device makes
device makes
the CdS/CuInSe2
the CdS cell aa particularly
/CuInSe2 cell interesting approach
particularly interesting approach for future application
for future to achieve
application to low cost
achieve low power
cost power
generation.
In this
In paper, the
this paper, methods and
the methods properties of the
and properties CuInSe 2 thin
deposited CuInSe2
the deposited film are
thin film described.
are described. Cell fabri-
Cell fabri-
cation and
cation results of detailed cell
and results characterization are
cell characterization presented.
are presented.
Preparation and
Preparation and properties films
CuInSe? films
properties of CuInSe2
The ternary material
The ternary material CuInSe2 solar cell
film solar
CuInSe 2 for thin film development has
cell development been prepared
has been simultaneous
by aa simultaneous
prepared by
elemental evaporation
elemental evaporation technique onto substrate. By controlling the
heated substrate.
onto aa heated rates of
individual rates
the individual three
the three
of the
elements, single phase
elements, aa single CuInSe2 film
phase chalcopyrite CuInSe2 can be
film can obtained. While
be obtained. either NN-type
While either or PP-type
-type or semi-
-type semi-
conducting thin film can be
film can deposited, the
be deposited, effort has
the effort concentrated only
has concentrated theP P-type
onlyononthe variety in
-type variety context
in context
planned heterojunction
with the planned involvingN N-CdS.
cell involving
heterojunction cell -CdS.

source design
The source
The is shown
design is Figure 1.
in Figure
shown in key feature
1. A key boat configuration
is a crossed boat
feature is for the
configuration for Cu and
the Cu In
and In
vaporization. Both boats are
Both boats tungsten, with
are made of tungsten, an alumina
with an barrier. A carbon
alumina barrier. carbon block was inserted
block was the
in the
inserted in
center of the
center the In boat in
In boat to maintain
order to
in order two vapor
maintain two on each
sources on
vapor sources side of
each side Cu. Spacing between
the Cu.
of the the
between the
boats was
boats inch. AA dual
1/8 inch.
was approximately 1/8 dual channel co-evaporation
channel co- (Inficon Model
controller (Inficon
evaporation controller Model Sentinel 200)
Sentinel 200)
based upon
based impact emission
electron impact
upon electron (EIES) was
emission spectroscopy (EIES) was used to simultaneously
used to monitor and
simultaneously monitor the
control the
and control
individual
individual rates Cu and
of Cu
rates of In. The
and In. EIES sensor
The EIES was located
sensor was it could
so that it
located so both of
"see" both
could "see" In sources
the In
of the as
sources as
well as the central
well as boat.
Cu boat.
central Cu
For the
For Se evaporation,
the Se two tungsten
evaporation, two tungsten boats were installed
boats were opposite ends
on opposite
installed on of the
ends of substrate for
the substrate deposition
for deposition
uniformity.
uniformity. Selenium vaporization raterate was with aa quartz
controlled with
was controlled deposition controller
crystal deposition
quartz crystal shields
and shields
controller and
provided to
were provided
were avoid directly exposing
to avoid EIES sensor
the EIES
exposing the to Se,
sensor to the sensor
while the
Se, while would still
sensor would function in
still function the
in the
presence of
presence Se, there
of Se, was the
there was the possibility of this
possibility of high vapor
this high pressure material
vapor pressure transported through the
being transported
material being the
attached light
attached coating the
pipe, coating
light pipe, feedthru window,
the feedthru window, and the EIES
and thereby attenuating the signals.
optical signals.
EIES optical
CuInSe^ films were
CuInSe2 films the equipment
deposited with the
were deposited onto polycrystalline
equipment onto alumina substrates
polycrystalline alumina to 350-450^.
heated to
substrates heated 350-450%.
The relative elemental
The evaporation rates
elemental evaporation were adjusted
rates were produce P P-type
adjusted toto produce -type films Kq/n
of 55 I<
films of to 800
/ to KQ/Q
800 K2 /0
resistivity
resistivity for -3 ym
for 22-3 thickness. According to
um thickness. reflection and
to reflection and tranmission diffraction, the
electron diffraction,
tranmission electron deposits
the deposits
were phase chalcopyrite
single phase
were single chalcopyrite CuInSe2. revealed aa grain size
analysis revealed
CuInSe 2 . SEM analysis size of approximately one
of approximately in
micron in
one micron
dimension. The
dimension. The absorption spectrum of these these films have been
films have measured in
been measured the wavelength
in the range 0.7
wavelength range to 2.5
0.7 to ym.
2.5 pm.
Results indicate
Results the material
that the
indicate that material has gap near
band gap
has aa direct band eV.
1.04 eV.
near 1.04
For cell development,
solar cell
For solar development, low low resistivity material is
CuInSe? material
resistivity CuInSe2 is desired. However, we found
desired. However, the low
that the
found that
resistivity (< -~ 50
resistivity (< 50 KSZ P -typeselenide
Kft/Q)/) P-type filmexhibited
selenidefilm large number
exhibited a a large of nodules
number of nodules when exposed to
when exposed CdS. SEM
to CdS.
and analysis
photomicrographs and
photomicrographs analysis by by electron beam microprobe
electron beam showed these
microprobe showed nodules to
these nodules be pure
to be copper. All
pure copper. cells
All cells
nodules were
containing nodules
containing found to
were found have low
to have efficiency.
low efficiency.
In order
In to avoid
order to nodule formation,
the nodule
avoid the the deposition
formation, the two-layer
of two
deposition of -layer selenide films has
selenide films been applied in
has been the
in the
fabrication of
fabrication the CdS
of the cells. In
/CuInSe22 cells.
CdS/CuInSe In practice, films have
these films
practice, these have been deposited with fixed
been deposited In and
fixed In Se
and Se
rates and
deposition rates
deposition the Cu
and the rate adjusted
Cu rate to achieve
adjusted to the desired
achieve the composition or
desired composition resistivity. The
or resistivity. The exact
nature of
nature the selenide
of the composition gradient
selenide composition gradient resulting from this
resulting from approach is
this approach determined by
is determined the particular
by the Cu
particular Cu
rate, the relative
rate, the relative layer temperature since
substrate temperature
layer thickness and the substrate processes are
diffusion processes
since diffusion at
active at
are active
the high substrate
the high temperatures. Accordingly,
substrate temperatures. Accordingly, there has been
there has high degree
been aa high flexibility in
of flexibility
degree of in the selenide
the selenide
deposits. upon cell
Based upon
deposits. Based performance, the
cell performance, most successful
the most procedure has
successful procedure been to
has been deposit high
to deposit resistivity
high resistivity
material over
material resistivity layer.
over aa moderately low resistivity low resistivity
The low
layer. The film favors
resistivity film formation of
the formation
favors the an
of an
ohmic contact
ohmic the base metallization, is
to the
contact to highly adherent,
is highly possesses aa larger
adherent, possesses grain size,
larger grain creates
and creates
size, and
the possibility
the back surface
of back
possibility of field (BSF)
surface field effects in
(BSF) effects the cell
in the structure. By
cell structure. then including
By then high
including aa high
resistivity layer
resistivity on top
layer on of this
top of deposits which do not exhibit the
layer, deposits
this layer, the Cu nodule formation
Cu nodule been
have been
formation have
Providing that the deposition parameters
obtained. Providing
obtained. been adequately
have been
parameters have the selenide
controlled, the
adequately controlled, films
selenide films
formed by
formed layered technique
the layered
by the have appeared to
technique have be quite
to be reprodicible. The
quite reprodicible. limitation on
major limitation
The major on the process
the process
is believed
is to be
believed to with the
associated with
be associated sensing of the mixed Cu and In
proper sensing
the proper In vapor emitted by
beams emitted
vapor beams by the large
the large
sources.
area sources.
area

fabrication and
Cell fabrication
Cell performance
and performance
thin-film
The polycrystalline thin cells have
CdS /CuInSe2cells
-film CdS/CuInSeo prepared on
been prepared
have been metallized alumina
on metallized alumina substrates
(2" xx 2"
(2" x 0.025
2" x MRC Superstrate).
0.025",", MRC Superstrate). Molybdenum films deposited by
films deposited RF sputtering
by RF were used
sputtering were low-cost
used as aa low -cost
metallization for the
metalli ation for The Mo
cells. The
the cells. Mo layers found to
were found
layers were be stable,
to be low resistivity
stable, low (0.2 ^/D
resistivity (0.2 Q/ for films
for films
3000A thickness),
of 3000A
of thickness), highly adherent, and
highly adherent, an ohmic
formed an
and formed contact.
selenide contact.
ohmic selenide
The cell structure
thin -film cell
The thin-film which was
structure which deposited on
was deposited top of
on top these metallized
of these substrates is
metallized substrates is depicted
depicted in
Figure 2.
Figure 2. As shown,
shown, the
the first
first layer
layer was
was the
the low-resistivity,
low- resistivity, selenide
selenide film
film and
and was
was deposited
deposited at aa
at
350°C to 450°C
350 °C to substrate temperature.
450 °Csubstrate temperature. After depositing approximately one micron of
one micron this material,
of this Cu
the Cu
material, the
was adjusted
rate was
rate for deposition
adjusted for of the
deposition of selenide layer.
final selenide
the final interrupting
This step change was made without interrupting
layer. This
deposition. Finally,
the deposition.
the substrate temperature
Finally, the substrate was increased
temperature was 450°C
to 450
increased to for the
°C for remainder of
the remainder the deposit
of the

SPIE Vol.
SPIE Vol.248
248 Role
Roleof
ofElectro-Optics
Electro- OpticsininPhotovoltaic
PhotovoltaicEnergy
EnergyConversion
Conversion (1980)
(1980) / / 63
63

Downloaded From: http://proceedings.spiedigitallibrary.org/ on 05/19/2015 Terms of Use: http://spiedl.org/terms

 CHEN, MICKELSEN
CHEN, MICKELSEN

when the initial

when the was deposited at
layer was
initial layer lower temperature.
at aa lower temperature. No cells with
No cells with efficiencies greater than
efficiencies greater 3%
than 3%
were obtained
were unless the
obtained unless 450°C
the 450 temperature was
°C temperature included and
was included higher efficiencies
and higher were measured
efficiencies were for the
measured for two--
the two
temperature procedure over aa constant
procedure over 450°C
constant 450 °C deposit. pressure during all
Chamber pressure
deposit. Chamber all selenide depositions
selenide depositions
been 33-8
has been -8 x 6 torr.
10~-6
x 10 torr.

With the completion

With the completion of the selenide
of the deposition, the
selenide deposition, temperature was
substrate temperature
the substrate lowered to
was lowered 150° -- 200
to 150° °C for
200°C for
the in CdS deposit.
-situ CdS
in-situ The
The CdS vaporization source
CdS vaporization was aa partitioned
source was tantalum foil
partitioned tantalum box with
foil box reusable
with aa reusable
quartz wool
quartz wool pad. crystal monitor
quartz crystal
pad. A quartz system was
monitor system used to
was used control the
to control rate at
deposition rate
the deposition at typically
typically
120 ÁA/s.
120 /s. At aa 150 deposition temperature,
150°C°C deposition sheet resistivity for
the sheet
temperature, the -5 ym
for 33-5 thick films
pm thick in the
was in
films was range
the range
60 -200 Kft/d.
60-200 K2/3. For cells, after deposition of approximately 0.8
For cells, ym of
0.8 pm pure CdS,
of pure the CdS
CdS, the films were
CdS films doped
were doped
with In
with 1.5%)
(~ 1.5
In (= co-evaporation.
by co-
%) by evaporation. doping formed
This doping low-resistivity
verylow-
formed aa very resistivity (30 /D) region
-100 Qft/Q)
(30-100 region inin
contact the deposited
with the
contact with deposited grid structure.
grid structure.
An aluminum film (1.5
An ym) grid
(1.5 pm) was deposited
grid was on top
deposited on the CdS
of the
top of using thru
CdS using -metal mask
thru-metal techniques. The
mask techniques. The
grid structure
grid consisted of 10 equally spaced lines
structure consisted of 50
lines of ym widths
50 pm covering aa 0.8
widths covering 0.8 cm
0.8 xx 0.8 Four
area. Four
cm area.
these cell
of these structures were formed on
cell structures substrate.
each substrate.
on each
a 1 cm2areas
into al
CdS into 2
and to
cm areas and edge
eliminate edge
to eliminate
Photolithographic techniques
Photolithographic then used
techniques were then to define
used to the CdS
define the
between the
shorting between
shorting CdS and
the CdS metallization. No
Mo metallization.
and Mo No detrimental
detrimental effect performance has
cell performance
effect on cell has been observed
been observed
as result of
the result
as the exposure to
of exposure to the based solutions
the solvents and aqueous based used in
solutions used photolithography. The
the photolithography.
in the
property of these cells
photovoltaic property
photovoltaic characterized as
generally characterized
be generally
cells can be having short
as having current
circuit current
short circuit
(25 -34 mA/cm
(25-34 2 ) and
mA /cm2) circuit voltage
open circuit
lowopen
andlow voltage (150 -390 mV)
(150-390 with efficiency
mV) with in the
efficiency in 7.48%.
range 2%2%toto7.48
the range %. The later
The later
value is
value highest reported
the highest
is the reported efficiency for thin film
this thin
for this device.
film device.
In order to
In order the open
improve the
to improve circuit voltage,
open circuit methods for
voltage, methods depositing lower
for depositing resistivity CuInSe2
lower resistivity films
CuInSe2 films
without Cu
without nodules have
Cu nodules been attempted.
have been attempted. The first cell with an
cell with efficiency greater
an efficiency than 6%
greater than produced by
was produced
6% was by
adjusting thickness of the
the thickness
adjusting the the two layers as
selenide layers
two selenide as well putting 10%
as putting
well as CdSe in
10% CdSe layer. The
CdS layer.
the CdS
in the as
The as
deposited cell
deposited cell had 0.34 VV and
VQC == 0.34
had Vo J $c =- 29
and Jsc mA/cm
29 mA /cm22 under simulated AMI
under simulated AM1 illumination. After aa series
illumination. After series of of
heat treatments
heat treatments from I50 150-200°C airororforming
-200 °Cininair gas,the
forminggas, performa ce was
cell performance
thecell was improved Voc == 0.39 V,
to Voc
improved to V,
27.5 mA/cm
sc == 27.5
JJsc mA/cm2, efficiency equal
2 , efficiency equal to 6.5% for
to 6.5% area of
cell area
total cell
for total of 11 cm' active area
for active
(7.05% for
cm 2 (7.05% area ofof
0.92 cm
0.92 2 ), and
cm2), and fill factor equal
fill factor 0.61. The
to 0.61.
equal to characteristics for
The II-V-V characteristics for the high efficiency cell
the high under
cell under
simulated AMI mW/cnr)
(100 mW
AMI (100 /cm illumination is
) shown in
is shown Figure 3.
in Figure From the
3. From the dark
dark J characteristics of
J-V-Vcharacteristics of the cell,
to cell,
the diode
the factor, A,
diode factor, and the
A, and current, Jo,
the reverse saturated current, found to
Jo, were found to be 1.37 and
be 1.37 1.3 xx 10-
and 1.3 A /cm 2 ,
10"° A/cm ,

respectively. TheseThese values with the

agreed with
values agreed measurement of
the measurement VQ C characteristics.
J sc -- Voc
the Jsc
of the characteristics. The selenide
The selenide
the junction
near the
near junction had high resistivity
had aa high resistivity which exceeded the
which exceeded measurementrange
the measurement four-point
thefour
rangeofofthe probe.
-point probe.

characteristics of
The characteristics
The of the fill factor
the fill efficiency as
and efficiency
factor and function of
as aa function of light intensity in
light intensity terms of
in terms of
SC /J0
JJsc is shown
/Jo is in Figure 4.
shown in The illumination
4. The intensityisislimited
illumination intensity /cm2 2(AM1)
limitedtoto100100mW mW/cm 3.87
(AMI)andand3.87 /cm2. .
mW mW/cm2
The data points
The data value. Both
points indicate the measured value. Both fill factor and
fill factor increase as
efficiency increase
and efficiency as the intensity is
the intensity is
reduced. At the lowest intensity
reduced. mW/cm
(3.87 mW
intensity (3.87 2 ), fill
/cm2), factor and
fill factor efficiency as
and efficiency high as
as high 0.64 and
as 0.64 10.9%,
and 10.9 %,
respectively, are
respectively, reached. The
are reached. The top smooth lines
top two smooth are the
lines are fill factor
ideal fill
the ideal and fill
factor and factor with
fill factor series
with series
resistance (1.64
resistance Q) and
(1.64 ft) parallel resistance (10
and parallel (10 5 s2) function of
ft) asas aa function SC /J0
of JJsc from Mitchell's
calculated from
/Jo calculated
analysis. 7 The
analysis.? The measured
measured fill factor may be
fill factor lower than
be lower the calculated
than the value due
calculated value to the
due to combination of
the combination of
Rs , R.
Rs, and the
Rp, and voltage dependence
the voltage of the
dependence of efficiency as
collection efficiency
the collection indicated by
as indicated Mitchell. 7
by Mitchell.?
The original
The of adding
intent of
original intent 10% CdSe
the 10%
adding the into the
CdSe into layer was
CdS layer
the CdS to prevent
was to Cu nodules
the Cu
prevent the on low
nodules on low
resistivity CuInSe2
resistivity films. It
CuInSe2 films. It was, however, later determined that
was, however, nodule formation
the nodule
that the solely
depended solely
formation depended
upon the selenide
upon the composition. The
selenide composition. Cu rich
The Cu selenide film
rich selenide (low resistivity)
film (low formed Cu
resistivity) formed nodules during
Cu nodules during
CdS deposition
CdS whether or
deposition whether CdSe was
not CdSe
or not was present. Therefore, we
present. Therefore, the CdSe
we discontinued the additions and
CdSe additions concen-
and concen-
trated on
trated on improving selenide composition.
the selenide
improving the
date, our
To date,
To best cell
our best demonstrated aa total
has demonstrated
cell has area (1
total area cm 2 ) efficiency
(1 cm2) efficiency of 7.48% (7.87%
of 7.48% of active
(7.87% of area
active area
0.95 cm 2 ). The
0.95 cm'). deposited cell
as deposited
The as had Voc
cell had /cm , and
J == 2828 mAmA/cm
0.27 V,V, Jsc
VQC == 0.27 fill factor
and poor fill
, under AM1
factor under illumina-
AMI illumina-
tion. After aa series
tion. of heat
series of 175°C
at 175
treatments at
heat treatments °C in forming gas for
in forming total time
for aa total of 90
time of minutes, the
90 minutes, cell
the cell
performance was
performance greatly improved.
was greatly The light JJ-V
improved. The characteristics of
-V characteristics the 7.48%
of the cell as
7.48% cell as shown in Figure
shown in 3.
Figure 3.
The performance under
cell performance
The cell condition (99.4
AMI condition
simulated AM1
under simulated /cm 2 ) were:
(99.4 mWmW/cm were: )

oc == 0.37
VVoc V,
0.37 V, Jsc
bu 32.7 mA/cm2
J, r =- 32.7 mA/cm 2y
V>
°' 26 V,
Vmp == 0.26 Jmp == 28
Jmp ' 6 mA/cm
28.6 mA/cm2
nn == 7.48% (total
7.48% (total area), 7.87% (active
area), 7.87% area)
(active area)
F.F. == 0.61
F.F. 0.61
had aa series
cell had
The cell
The resistance ofof1.4
series resistance and shunt
1.4szft and resistance of
shunt resistance 10 Q.
of 66 xx 104 efficiency and
ft. The efficiency fill
and fill
factor as
factor function of
as aa function light intensity had
of light characteristics similar
had characteristics to those
similar to shown in
those shown 4. They again
Figure 4.
in Figure
indicated that the
indicated that series resistance
the series resistance was dominant factor
was aa dominant in degrading
factor in the fill
degrading the The source of
factor. The
fill factor.
high series
high resistance has
series resistance been found
has been found toto be partly due
be partly resistance of
the resistance
to the
due to of the base metal
thin base
the thin layer.
metal layer.
The dark JJ-V
The dark characteristics at
-V characteristics at both forward and
both forward reverse bias
and rever5se bias have been measured.
have been The diode factor
measured. The was
factor was
1.37 and
1.37 the dark
and the reverse saturation
dark reverse saturation current was 88 xx 10-
current was A/cm'..
10 A/cm For high efficiency cells,
For these high no
cells, no
crossover of light
crossover and dark
light and characteristics has
-V characteristics
dark JJ-V been observed.
has been observed.

64 / SPIE
/ SPIE Vol.
Vol. 248 Role ofElectro- Optics in
of Electro-Optics in Photovoltaic
Photovoltaic Energy
Energy Conversion
Conversion (1980)
(1980)

Downloaded From: http://proceedings.spiedigitallibrary.org/ on 05/19/2015 Terms of Use: http://spiedl.org/terms

 THIN-FILM
THIN CdS/CulnSe2
-FILM CdS HETEROJUNCTION SOLAR
/CuInSe2 HETEROJUNCTION SOLAR CELL
CELL

The
The mechanisms responsible for
mechanisms responsible for the
the low
low open
open circuit
circuit voltage
voltage and
and high
high short
short circuit
circuit current
current in
in our
our high
high
efficiency cells
efficiency cells will be discussed
will be discussed in
in the
the next
next Section.
Section.
A preliminary
preliminary analysis pertaining to
analysis pertaining to photon
photon economy
economy has
has been
been made.
made. TheThe total
total reflectance
reflectance ofof aa cell
cell
sample (CdS
sample (CdS/CuInSe2/Mo/alumina
/CuInSe2 /Mo /alumina without
withoutthe thetop
topgrids)
grids)was
wasmeasured
measuredasas a a function
function of
of wavelength
wavelength byby using
using
an integrating
integrating sphere
sphere attached
attached toto aa Beckman
Beckman DKDK-2A
-2A spectrophotometer.
spectrophotometer. The technique used
The technique used was
was similar
similar to the
method reported by Bragagnolo and Fagen.8
method reported Fagen. 8 The result
result is
is shown
shown in
in Figure
Figure 5. 5. The average reflectance
reflectance in in
the wavelength
the wavelength range
range of
of 500
500-1000 nm was
-1000 nm was approximately
approximately1414%.%. From absorbance
From absorbance measurements
measurements ofof aa CdS
CdS layer
layer
which had
which had approximately thethe same
same thickness
thickness andand In
In doping
doping as
as deposited
deposited onon the
the cell,
cell, itit appeared
appeared that
that the
the
CdS
CdS layer
layer absorbed
absorbed about
about 2%2% of the
the light in in this
this wavelength range.
range. The
The top grid contact
top grid contact now
now being
being used
used
had aa nominal
had nominal shading
shading area
area ofof 5%
5% of
of the
the total
total area.
area. The total
total photon
photon loss
loss inin the existing
existing cell
cell structure
structure
was, therefore,
was, therefore, approximately
approximately2121%.%. The remaining
The remaining 79%79% of
of total
total photons
photons has
has been
been assumed
assumed to
to be
be absorbed
absorbed
in
in the
the CuInSe
CuInSe22 layer for
for carrier
carrier generation.
generation.
Except
Except for
for the
the CdS
CdS absorption loss, the
absorption loss, the reflectance
reflectance and grid shading
shading could be
be minimized. With
With aa more
more
advance
advance design and an
design and an antireflection
antireflection coating,
coating, the
the reflectance should be
reflectance should be reduced
reduced below
below 5%
5% as
as shown
shown in
in the
the
experience of
of the
the Cu2S
Cu2S/CdS
/CdS cell
cell development.
development.99 Redesign
Redesign of
of the
the grid
grid contact structure
structure may
may not
not only
only decrease
decrease
the
the shading loss to
shading loss to 11-2%
-2% but
but may
may also
also reduce
reduce the
the series
series resistance.
resistance.

If
If all
all the
the improvements
improvements to
to reduce the photon
reduce the photon losses
losses can
can be
be applied
applied to
to out
out 7.5%
7.5% cell,
cell, aa cell
cell of
of 9%
9% effi-
effi-
ciency could
ciency could be
be obtained
obtained without any
any other
other material
material optimization.
optimization. The
The estimation
estimation is
is summarized
summarized in
in Table
Table 3.
3.
Table 3.
3. Estimated Improvement of the
Estimated Improvement the 7.5% Cell
Cell by
by Minimizing
Minimizing Photon
Photon Loss
Loss

Reflectance
Reflectance Grid Shading
Grid Shading CdS
CdS Voc
Voc JJsc
sc 2
Loss
Loss Loss Absorption (V)
(V) (mA/cm
(mA /cm ) F.F.
F.F. n (%)
n ( %)

Existing
Existing 14% 5%
5% 2% 0.37
0.37 33
33 0.61
0.61 7.5%
7.5%
Cell
Cell

Improved
Improved 5% 2% 2% 0.375 38
38 0.63
0.63 9.0%
Cell
Cell

Cell
Cell analysis and modeling
analysis and modeling
In this
In section, the
this section, the performance
performance ofof our
our CdS
CdS/CuInSe2
/CuInSe2 solar cell is
solar cell is analyzed
analyzed in
in detail.
detail. The main objective
rain objective
of the analysis is
is to
to determine
determine why
why our
our cells
cells give
give such
such high
high short
shortcircuit
circuitcurrent
current(>(>3030mAmA/cm
/cm )) and
and low
low
open
open circuit
circuit voltage (< 0.4
voltage (< 0.4 V)
V) and,
and, subsequently,
subsequently, to to suggest
suggest ways toto improve
improve the
the cell
cell performance.
performance.
Under simulated AM
Under simulated AM-1-1 solar
solar illumination,
illumination, the
the experimental
experimental device being
being investigated
investigated had:
had:
Area -- 11 cm2
Area cm
VQC
Voc -- 0.305 VV
0.305
J
Jsc -- 33 mA
mA/cm
/cm22
n
n -
- 5.25%
5.25%
F.F.
F.F. -
- 0.52%
0.52%

The
The diode factor, A,
diode factor, A, and
and the
the reverse
reverse saturated
saturated current,
current, J,,
J0 , were
were found
found toto bebe1.15 and3 3x x1010"
1.15and 6 A/cm 2 ,
-6 A /cm2,
respectively, from
from the
the dark
dark II-V
-V characteristic. These va values
?ues agreed
agreed with
with the the measurements
measurements of of the
the Voc
Voc -- vJ'lsc
sc
characteristics
characteristics as as aa function
function of light intensity
intensity and
and were
were independent
independent of of aa bias
bias light
light of
of different
different wavelengths
wavelengths
from 450 nm
nm to
to 650
650 nm.
nm. The
The Jo
J0 as
as aa function
function of
of temperature
temperature were
were determined
determined by by measuring
measuring the the Voc
Voc -- Jsc
J sc
characteristic
characteristic atat various
various temperatures.
temperatures. The The results
results are shown in
are shown in Figure
Figure 66 which
which gives
gives the
the temperature
temperature
dependence of
dependence of Jo
J0 as:
as:
Jo = Joo
Jo = Joo exp (-AkT)
AE
where AE
where AE =- 0.77
0.77 eV
eV and
and Joo
JQO == 1.93
1.93 x x 105
10 5A A/cm2
/cm2..

The variation
The variation of
of open
open circuit
circuit voltage
voltage with
with temperature
temperature and
and AMAM-1 illumination was
-1 illumination was also
also measured
measured as
as shown
shown inin
Figure 7.
Figure 7. The VVgc varies
varies linearly
linearly with
with temperature
temperature for
for high
high temperature
temperature (T(T> > 260
260°K).
°K). TheThe decrease of slope
slope
below 260
below 260°K
°K may
may be
be due
due to the Mo
to the Mo/CuInSe2
/CuInSe2 or
or Al/CdS
Al /CdS contacts
contacts becoming non-ohmic
becoming non -ohmic andand high
high resistivity at low
resistivity at low
temperatures. The The intercept of the
the linear
linear portion
portion at
at TT == 00 gives:
gives:
V
oc
oc
-= 0.73
0.73 VV
(T=0)
The cell
The cell performance,
performance, high
high Jsc
J sc and
and low
low V,,,
VQC , can
can be
be explained
explained byby the
the model
model developed
developed by
by Rothwarf10
Rothwarf10 in
in
connection with our
our measured
measured results.
results. InIn h°is
his model,
model, the
the characteristics
characteristics of the cell
cell can
can be
be represented
represented by:
by:

SPIE Vol.
SP/E Vol 248 RoleRo
248 of Electro-Optics
/e of Electro- in Photovoltaic
Optics in Energy Conversion (1980)
Photovoltaic / 65
Energy Conversion (19

Downloaded From: http://proceedings.spiedigitallibrary.org/ on 05/19/2015 Terms of Use: http://spiedl.org/terms

 CHEN, MICKELSEN
CHEN, MICKELSEN

-qV
gV oc
J sc == Jo
Jsc JQ (exp
(exp - 1)
sc ° AkT
AkT

where Jsc is
where Jsc is the
the short
short circuit current
current or
or the
the light
light generated
generated current
current and
and AA is
is the
the diode
diode factor.
factor. If
If the cell
cell
transport
transport property is dominated by
property is by interface
interface recombination,
recombination, Jo
JQ should
should be:
be:
oE
Jo = gNcSl exp (AkT)

where
where Nc
Nc is
is the
the density
density of states for
of states for the
the CdS
CdS conduction
conduction band,
band, SI
Sj is
is the
the effective
effective interface
interface recombination
recombination
velocity and
and
AE = Ea1
oE= E gl --oX
A X -SI
- 6X
where EgI
where Egi is
is the
the CuInSe2
CuInSe2 band
band gap
gap energy,
energy, ox
AX the
the difference
difference in
in electron
electron affinity
affinity between
between CdS
CdS and
and CuInSe2,
CuInSe^,
and S1
and 6]_ the
the position
position of
of the
the Fermi
Fermi level
level in
in CuInSe2
CuInSe2 relative
relative to
to the
the valence
valence band.
band. For
For this
this model,
model, the
the diode
diode
factor
factor AA is
is one
one and
and the
the open
open circuit voltage is
is then given by:
by:

\i t~ o v
Voc
V == EEglgl -
- SX
«x - S1
- «! +i'*'T
AqT ln ^S7
gNcsl

The
The diode
diode factor
factor of
of our experimental cell
our experimental cell is
is 1.15
1.15 which
which indicates
indicates the
the transport
transport mechanism
mechanism may
may be
be dominated
dominated
by the
by interface recombincation as the
the interface the model
model predicts.
predicts.

The
The slope of the Jo
J0 vs.
vs. (kT)
(kT)'-1
1 curve
curve (Figure
(Figure 6)
6) gives
gives

oE=Egl
AE = E , --OX-S1
AX - 6j =0.77eV
= 0.77 eV

and
and agrees
agrees very
very well
well with Voc vs.
with the Voc vs. kT
kT measurement (0.73
(0.73 eV).
eV). Since E Q i == 1.04
Since EgI 1.04 eV
eV and
and AX
AX == 0.08
0.08 eV,12
eV, 12
it follows
it follows that:
that: y

6S1
1 =
= 0.19 - 0.23
0.19 - 0.23 eV
eV
The
The measured Joo
JOQ == 1.93
1.93 xx 105 5 2
10 AA/cm
/cm2 yields
yields the
the product
product of NcSI,
N S i.e.,
i.e.,
Nc Sj == 1.21
NcSI 1.21 xx 1024
10 24 cmcm"-22 sec
sec -1
~l
10 _O
For an
For an estimated, typical
typical value
value of
of NcNC = = 2 2 x x1018
10 cmcm-3,,

Sj = 6.05 x 105
SI 10 5 cm/sec

This is
This is consistent
consistent with the calculated
with the calculated value
value of
of 6.9
6.9 xx 104
10 -- 6.9
6.9 xx 105
10 cm
cm/sec
/secll for
for Sj
SI based
based upon
upon the
the lattice
lattice
mismatch between CdS
mismatch between CdS and CuInSe2 at
and CuInSe2 at the
the interface.
interface.
The
The hole
hole concentration
concentration of the
the CuIne2
CuInSe2 laXXer
layer close
close to
to the
the junction
junction may
may be
be calculated
calculated from
from the
the S6-j value.
value.
With an estimated value
value of
of 4.18
4.18 xx 1010118 cm"'3 for Nv
cm-i for Nv (the
(the CuInSe2
CuInSe2 valence
valence band
band effective
effective density
density ofof states)
states)
the
the hole
hole concentration
concentration of the
the CuInSe2
CuInSe2 layer
layer close
close to
to the
the junction
junction should
should be:
be:

p =- 2.36
p 2.36 xx 1015
10 15 -- 4.88
4.88 xx 1014
10 14 cmcm"-33
Since the
Since the In
In-doped
-doped CdS
CdS is
is very
very low
low resistivity,
resistivity, the
the depletion
depletion region
region width
width calculated
calculated from
from the
the above
above hole
hole
concentration and
concentration and the
the measured
measured VDVp(0.73---
(0.73---0.77V) is:
0.77V) is:

W == 6.0
6.0 xx 10-5
10" 5 -- 1.29
1.29 xx 10-4
10" 4 cm
cm
p
which
which agrees
agrees with
with the
the value
value obtained from the
obtained from the dark
dark capacitance
capacitance measurement at
at zero
zero bias
bias (c
(c == 7.6 nf/cm 2 ) or
7.6 nf/cm or )

1.16
1.16 xx 10" 4 cm.
10-cm.

From
From the
the above
above analysis,
analysis, the
the model is seen
model is seen to
to be
be consistent with allall cell
cell measurements.
measurements. The low open
The low open
circuit
circuit voltage
voltage is
is mainly
mainly due
due to
to the high 61
the high 6j value,
value, i.e.,
i.e., too high
high resistivity
resistivity ofof the
the CuInSe2
CuInSe2 layer.
layer. The
The
high short
high short circuit current of
circuit current of the
the cell,
cell, 3333 mA/cm^,
mA /cm2, which
which approaches
approaches the
the single
single crystal
crystal cell be explained
cell can be explained
by
by the
the 1.16
1.16 ym
pm wholly depleted CuInSe2
CuInSe2 region.
region. Since the CuInSe2
Since the CuInSe? absorption
absorption coefficient
coefficient for
for photon
photon energies
energies
greater
greater than EEg is
is over 104 cm
over 104 cm'l,
-1, nearly
nearly all
all of
of the
the absorption
absorption of
of the
the solar
solar spectrum
spectrum will
will be
be in
in the
the 1.16
1.16 ym
pm
depleted region.
region. The electric field
The electric field in
in this
this depleted
depleted region
region can
can be
be estimated
estimated toto be:
be:

66 / SP
/ SPIE
/E Vol
Vol.248
248Role
Roleofof
Electro- Optics ininPhotovoltaic
Electro-Optics PhotovoltaicEnergy
EnergyConversion
Conversion (1980)
(1980)

Downloaded From: http://proceedings.spiedigitallibrary.org/ on 05/19/2015 Terms of Use: http://spiedl.org/terms

 THIN-FILM CdS/CulnSe2
THIN -FILM CdS /CuInSe2 HETEROJUNCTION CELL
SOLAR CELL
HETEROJUNCTION SOLAR

2V
2VD
ee == —- = 1.33 xx 10
- 1.33 104 V/cm
4 V/cm
WW

The light generated electron

The light -hole pairs
electron-hole be separated
will be
pairs will by such
separated by high fields
such high collected.
and collected.
fields and The maximum
the holes
transit time of the 10 cm
(assuming uy == 10
holes (assuming /V -sec) under
crrr/V-sec) this high
under this is:
field is:
high field

ICf-10
8.9 xx 10
t == 8.9 sec
10 sec
Hence, the light
all the
Hence, all light generated before they
generated carriers are collected before can recombine
they can in the
recombine in boundaries.
grain boundaries.
the grain
Hence, in
Hence, order to
in order to improve voltage, the
improve the open circuit voltage, 6-^ (or
the 61 the resistivity
(or the of the
resistivity of CuInSe2) should
the CuInSe2) be
should be
Short circuit current may be
reduced. Short
reduced. be reduced from the
reduced from present high
the present values but
high values the overall
but the efficiency
cell efficiency
overall cell
improved.
be improved.
could be
55
AA comparison of our cell
of our (n == 7.5
cell (n cell (n
with Kazmerski's cell
7.5%)%) with = 6.6
(n = %), shows
6.6%), selenide resistivity
our selenide
shows our to
resistivity to
be least one order higher
be at least This would
his. This
than his.
higher than indicate that the
would indicate selenide resistivity
the optimum selenide be
should be
resistivity should
somewhere between
somewhere these values
between these and, more
values and, likely, closer
more likely, to our
closer to In conclusion,
value. In
our value. by further
conclusion, by optimizing
further optimizing
selenide resistivity and by
the selenide
the by minimizing photon losses,
the photon
minimizing the cells with
losses, cells efficiency over
with efficiency should
10% should
over 10%
realistically be
realistically feasible.
be feasible.
Acknowledgements
Acknowledgements

The research work was

research work performed under
was performed Contract(XJ
undera aContract -9- 8021 -1)from
(XJ-9-8021-1) the Solar
from the Energy Research
Solar Energy Institute.
Research Institute.
The authors would
The authors would like to thank
like to G. Wolff
thank G. and G.
Wol f f and Henderson for
G. Henderson their technical
for their assistance.
technical assistance.
Reference
1.
1. J. L.
J. Shey and
L. Shey J. H.
and J. Wernick, Ternary
H. Wernick, Chaleopyrite Semiconductors:
Ternary Chaleopyrite Growth, Electronic
Semiconductors: Growth, Properties
Electronic Properties
and Applications,
and Pergamon Press,
Applications, Pergamon New York
Press, New (1975).
York (1975).
2.
2. J. L.
J. Shay, S.
L. Shay, Wagner and
S. Wagner H. M.
and H. Kasper, Appl.
M. Kasper, Phys. Lett.
Appl. Phys. 27, 89
Lett. 27, (1975).
89 (1975).
3.
3. J. Loferski, J.
J. Loferski,
J. J. Phys. 27,
Appl. Phys.
J. Appl, (1956).
777 (1956).
27, 777
4.
4. Romeo, G.
N. Romeo,
N. Sherregl ieri, and
G. Sherreglieri, and L. Tarricone, Appl.
L. Tarricone, Phys. Lett.
Appl. Phys. 310, 108
Lett. 30, (1977).
108 (1977).
5.
5. L. L. Kazmerski,
L. L. Compounds 1977,
Kazmerski, Ternary Compounds ed. G.
1977, ed. Holah, p.
D. Holah,
G. D. 217, Institute
p. 217, of Physics
Institute of Conference
Physics Conference
No. 35
Series No.
Series (1977).
35 (1977).
6.
6. R. A.
R. Nickel sen and W.
A. Mickelsen W. S. Chen, Appl. Phys.
S. Chen, Lett. 36,
Phys. Lett. (A80).
371 (A80).
36>, 371
7.
7. K. W. Mitchell,
K. W. Evaluation of
Mitchell, Evaluation of the Heterojunction Solar
/CdTe Heterojunction
CdS/CdTe
the CdS Cell, Garland
Solar Cell, Publishing Inc.,
Garland Publishing (1979).
Inc., (1979).
8. J. A. Bragagnolo and E.
J. A. A. Fagen,
E. A. Photovoltaic Material
Fagen, Photovoltaic and Device
Material and Measurements Workshop,
Device Measurements 169,
p. 169,
Workshop, p.
SERI /TP -49 -185, (1979).
SERI/TP-49-185, (1979).
9.
9. J. A.
J. Bragagnolo, Conf.
A. Bragagnolo, Record 13th
Conf. Record IEEE Photovoltaic
13th IEEE Specialist Conf.,
Photovoltaic Specialist p. 412
Conf., p. (1978).
412 (1978).
10. A.A. Rothwarf, Workshop in
International Workshop
Rothwarf, International in CdS Solar Cells
CdS Solar and Other
Cells and Heterojunctions, p.
Abrupt Heterojunctions,
Other Abrupt 9,
p. 9,
Univ. of
Univ. of Delaware,
Delaware, Newark, (1975).
Delaware (1975).
Newark, Delaware
11. R.
11. A. Mickelsen, W.
R. A. S. Chen
W. S. and B.
Chen and Selikson, "CdS
B. Selikson, /Cu Ternary Heterojunction
"CdS/Cu Cell Research,"
Heterojunction Cell Final
Research," Final
Report, Contract
Report, 77 -C -03 -1458, Oct.
ContractEG-EG-77-C-03-1458, 1978.
Oct. 1978.

SPIE Vol.
SPIE Vol.248
248Role
Roleof
ofElectro-Optics
Electro- OpticsininPhotovoltaic
PhotovoltaicEnergy
EnergyConversion
Conversion (1980)
(1980) / / 67
67

Downloaded From: http://proceedings.spiedigitallibrary.org/ on 05/19/2015 Terms of Use: http://spiedl.org/terms

 CHEN, MICKELSEN
CHEN, MICKELSEN

Carbon Cloth
Carbon Cloth
Substrate
Substrate
1 / Heater
r Heater (B)
Substrate
Substrate 30
30
Shutter
=3 Shutter
Se S ensor
Se Sensor (A)
(A)

20
20 - ¡ i

i
i
EIES
EIES
i

Se Source
Se Source Sensor
Sensor N
1

1 1
E

C =
U 1 1

I
E
E i i
10 1

Source
Cu Source
Cu

"071O o2.,,
Oo2 w ,., 03 o O"44
source
In Source 0.
V(Volt)

r
I

Carbon Block
Carbon Block
Figure Li ht J-V
3. Light
Figure 3,
Ce
J-V Characteristics
ils. (A)
Cells. TI - 6,59'
(A) n
for 11 cm2
Characteristics )for CdS/CuInSe,,
cm CdS/CuInSe2
7 0 48%
7.48&S 7.48%
(B) nn - 7.48&%
6,5% (B) ^

Figure 1.
Figure 1.
System configura
System tion for
configuration preparing CuInSe2
for preparing films
CuInSe2 films
using EIES depositio
using EIES deposition controller
n controller
70

68

66

esi i Al grid contact

Al grid contact
2-4um In doped
In CdS
doped CdS
0.5-1 .514m Cds
Cds ZE

High Resistivity
High Resistivity LL
1-3um LL
CuInSe2 w
Resistivity
Low Resistivity
Low
1-3Km
p O.
.! ! í9f i . .í p9
.. d.
CuInSe2
Metal
Metal back
Glass
contact £
back contact
or Alumina
Glass or Alumina
58
Substrate
Substrate

Figure 2.
Figure 2. CdS/CuInS
CdS /CuInSe2 cell structure
e 2 cell structure 56

104 IL/I6 105

Figure 4.
Figure 4. Experimental and theoretical
Experimental and fill
theoretical fill
as a
factor as
factor a function of IL/40
function of for.the
Ir/In for2the
Vcm"
cell,
6.5^ cell.
6.5% 1.3x x117- A /cm
Jo - 1.3

(1980)
Conversion 11980)
/ SP/£Vol.
68 / SPIE Vol.248 Roleof
248Role Electro-Opti
ofElectro- Optics in Photovoltaic Energy Conversion
cs in

Downloaded From: http://proceedings.spiedigitallibrary.org/ on 05/19/2015 Terms of Use: http://spiedl.org/terms

 THIN -FILMCdS/CulnSe7
THIN-FILM CdS /CuInSe2 HETEROJUNCTION SOLAR
HETEROJUNCTION CELL
SOLAR CELL

20
20

10
10
500
500 X(nm)
a(nm) 1000
1000 1500
1500
Figure 5.
Figure 5. Total reflectance of
Total reflectance /CuInSe2? /Mo/Alumina
CdS/CuInSe
of aa CdS as aa
sample as
/Mo /Alumina sample
function of
function wavelength
of wavelength

10-2 0.4

§ 0.3
o
NE 10-3 o

0.2, 2.5 3.0

1.2
KT(1Q~ 2 eV)

Figure 7.
Figure 7. Open voltage as
Open circuit voltage function
as aa function
10 -4 of temperature
of temperature
39 41 43
37
1 /KT (eV"
1/KT 1)
(eV -1)

Figure 6.
Figure 6. JJo as function of
as aa function of inverse temperature
inverse temperature

SPIE Vol.
SPIE 248 Role
Vol. 248 Electro- Opticsin
of Electro-Optics
Role of Energy Conversion
Photovoltaic Energy
in Photovoltaic (1980)/ / 69
Conversion (1980) 69

Downloaded From: http://proceedings.spiedigitallibrary.org/ on 05/19/2015 Terms of Use: http://spiedl.org/terms