Galvanic contact deposition of CdTe layers using ammoniacal basic aqueous solution

Article abstract of DOI:10.1149/1.1870756

The galvanic contact deposition of CdTe layers from ammoniacal basic solutions was carried out, and their deposition behaviors were investigated. The structural and electrical properties of deposits were examined and then compared with those of deposits prepared by normal electrodeposition from the same solutions. The cathode potential was always spontaneously kept at around -0.7 V vs. standard hydrogen electrode during the contact deposition although CdTe also deposited on the Cd sheet, that is, the anode. The current density gradually decreased with time just like that in the normal potentiostatic electrodeposition. The current efficiency was approximately 100% under illumination, whereas it was less than 50% in the dark. These behaviors were the same as those observed in normal electrodeposition from the same electrolytes. The resulting deposits both under illumination and in the dark were polycrystalline CdTe layers with almost stoichiometric composition. The as-deposited CdTe layer had a p-type conduction with resistivity of the order of 10⁷ Ω cm.

Full text of DOI:10.1149/1.1870756

Journal of The Electrochemical Society, 152 ͑4͒ C237-C242 ͑2005͒
C237
0013-4651/2005/152͑4͒/C237/6/$7.00 © The Electrochemical Society, Inc.
Galvanic Contact Deposition of CdTe Layers Using
Ammoniacal Basic Aqueous Solution
*
*
Kentaro Arai, Souichi Hagiwara, Kuniaki Murase, Tetsuji Hirato, and
Yasuhiro Awakuraz
Department of Materials Science and Engineering, Kyoto University, Kyoto 606-8501, Japan
The galvanic contact deposition of CdTe layers from ammoniacal basic solutions was carried out, and their deposition behaviors
were investigated. The structural and electrical properties of deposits were examined and then compared with those of deposits
prepared by normal electrodeposition from the same solutions. The cathode potential was always spontaneously kept at around
−0.7 V vs. standard hydrogen electrode during the contact deposition although CdTe also deposited on the Cd sheet, that is, the
anode. The current density gradually decreased with time just like that in the normal potentiostatic electrodeposition. The current
efficiency was approximately 100% under illumination, whereas it was less than 50% in the dark. These behaviors were the same
as those observed in normal electrodeposition from the same electrolytes. The resulting deposits both under illumination and in the
dark were polycrystalline CdTe layers with almost stoichiometric composition. The as-deposited CdTe layer had a p-type con-
duction with resistivity of the order of 107 ⍀ cm.
© 2005 The Electrochemical Society. ͓DOI: 10.1149/1.1870756͔ All rights reserved.
Manuscript submitted August 23, 2004; revised manuscript received November 9, 2004. Available electronically March 9, 2005.
Cathodic electrodeposition is one of the key techniques for the
preparation of thin-layered compound semiconductors ͑e.g., CdTe
and CuInSe2͒ for solar cell applications, and the cell made up of an
n-CdS/p-CdTe heterojunction has been put into production on an
industrial scale. Since the pioneering work of Kröger’s group in the
late 1970s, aqueous acidic sulfate solutions have historically and
almost exclusively been employed as the bath for CdTe electrodepo-
sition. In contrast, we have proposed that aqueous basic, or alkaline,
solutions containing ammonia or amine ligands are also suitable for
the electrodeposition of a uniform CdTe layer, because these basic
solutions have a relatively high solubility of Te͑IV͒ species as
TeO23− ions.
CdTe deposition under “semipotentiostatic” conditions.
Figure 1 shows a set of potential-pH diagrams of the Cd-NH3
-H2O and Cd-Te-NH3-H2O systems for the basic region.2,3 As men-
tioned, the deposition of stoichiometric CdTe takes place at poten-
tials positive to the Nernst potential ͑ECd͒ for bulk Cd deposition
2+
4
Cd͑NH3͒ + 2e ꢀ Cd + 4NH3
͓1͔
According to the diagram of the Cd-NH3-H2O system, the potential
ECd at pH 10-11, where CdTe electrodeposition from ammoniacal
basic baths is usually conducted by the authors’ group, is about
−0.73 V vs. standard hydrogen electrode ͑SHE͒. In the same pH
range, the positive limit for the CdTe electrodeposition is sectioned
by the reaction
The deposition of stoichiometric CdTe takes place at potentials
positive to the Nernst potential for elemental Cd deposition. Regard-
ing the acidic baths, Kampmann et al. found that a deposition po-
tential of only +5 mV vs. the elemental Cd deposition led to the
formation of CdTe with relatively high crystallinity.1 Although the
basic solution gave CdTe layers having a lower crystallinity com-
pared to the layers from the acidic baths, the nearer the deposition
potential approaches the potential for elemental Cd, the higher the
CdTe deposition current becomes. Hence, a positive control of cath-
ode potential during the electrodeposition of CdTe layer is of impor-
tance for both acidic and basic baths. On a laboratory scale, elec-
trodepositions under controlled potentials, i.e., potentiostatic
electrodepositions, are easily conducted by a three-electrode setup
using a potentiostat and nonpolarizable reference electrode together
with a Luggin capillary. However, when upscaled to an industrial
level, most electrodeposition processes are reorganized to be current
controlled, i.e., galvanostatic electrodepositions, to avoid the com-
plexity arising from the three-electrode configurations, while such
galvanostatic conditions may result in a large fluctuation in deposi-
tion potential.
2+
Cd͑NH3͒ + Te + 2e ꢀ CdTe + 4NH3
͓2͔
4
of which the Nernst potential ͑ECdTe͒ is around −0.22 V. These
imply that the galvanic contact plating of CdTe deposition can be
realized using a Cd metal electrode as the auxiliary electrode, i.e.,
anode. Under this concept, the galvanic contact plating of CdTe
from conventional acidic media using the Cd auxiliary electrode was
examined by Bhattacharya et al. in the early 1980s. However, in this
case, the use of Cd electrode as the anode seems to be a drawback to
the process, because the use of a soluble Te anode for continuous
feeding of Te͑IV͒ ions is practical, owing to the low solubility of
Te͑IV͒ ions to the baths. In the field of semiconductor electrodepo-
sition, CdS,4 CdSe,5 ZnTe,6 and CdTe7 layers have also been pre-
pared by the galvanic contact plating using acidic media.
In the present study, we tried using basic ammoniacal baths for
the galvanic contact deposition of CdTe. Here, a photo-assisted
growth technique of CdTe2 was employed. The deposition phenom-
ena and the properties of the resulting CdTe layer, i.e., stoichiom-
etry, morphology, and electrical properties, were compared with
those electrodeposited using an external current source.
Without the potentiostat and reference electrode, the potential of
the working electrode, i.e., the deposition substrate, for cathodic
electrodeposition can be held almost constant using the galvanic
contact plating technique. In this technique, the substrate is short-
circuited to an auxiliary electrode, of which the potential when im-
mersed into the electrodeposition bath, or into another solution con-
nected electrically with the electrodeposition bath, is negative to the
Nernst potential for the deposition of the desired substance. In other
words, the substrate and the auxiliary electrode constitute a short-
circuited galvanic cell, where the former acts as cathode and the
latter as anode. Such a galvanic contact plating may make possible
Experimental
Deposition of CdTe.—Basic aqueous electrolytes were prepared
by dissolving 40-60 mM CdSO4·8/3H2O and 10 mM TeO2 in an
ammoniacal buffer solution containing 4.0 M NH3͑aq͒ and 0.5 M
͑NH4͒2SO4.8 All chemicals ͑Nacalai Tesque, Inc.͒ were of reagent
grade and were used without pretreatment. The deionized ͑DI͒ water
used to prepare the ammoniacal buffer had a specific resistance
larger than 5 ϫ 106 ⍀ cm. The pH of the pre-prepared ammoniacal
buffer, where ͓NH3͔ + ͓NH+4͔ = 5.0 M, was 10.7 at 25°C and did
not change by addition of CdSO4·8/3H2O and TeO2; Cd͑II͒ and
Te͑IV͒ species in the electrolytes were thus dissolved to form
2+
Cd͑NH3͒ and TeO23− ions, respectively. An Ag/AgCl electrode
*
4
Electrochemical Society Active Member.
z E-mail: yasuhiro.awakura@materials.mbox.media.kyoto-u.ac.jp
͑Horiba 2080A-06T͒ immersed in 3.33 M KCl was used as a refer-
Downloaded on 2015-01-13 to IP 66.254.251.28 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

C238
Journal of The Electrochemical Society, 152 ͑4͒ C237-C242 ͑2005͒
Figure 2. Typical picture of a sample for measurements of electrical prop-
erties. The sample has four small Au contacts in four vertices of square CdTe
on nonconductive epoxy resin.
Figure 1. Potential-pH diagrams of ͑a͒ Cd-NH3-H2O and ͑b͒ Cd-Te-NH3
-H2O systems for basic region at 298 K calculated assuming activities of
dissolved Cd͑II͒ and Te͑IV͒ species and total ammonia concentration are
0.06, 0.01, and 5.0 M, respectively. Two arrows correspond to the process
expected for ͑a͒ anode and ͑b͒ cathode reactions during galvanic contact
deposition.
deposits were determined by electron probe microanalysis ͑EPMA͒.
A single crystal of CdTe having a composition of 50.00 atom % Cd
was used as the standard for EPMA. The reflectivity spectra of de-
posits were measured by a spectrophotometer ͑Hitachi U-3500͒ with
an integrating sphere to determine the bandgap energy.
Hall effect measurement.—The CdTe layer was transferred from
the Au/Cu substrate onto nonconductive epoxy resin ͑Torr Seal;
Varian Associates͒ for electrical characterization.8 After transferring
the CdTe layer, four Au electrodes ͑70-100 nm thick͒ were deposited
by vacuum evaporation through a cross-shaped mask using two
strips of aluminum foil; the width of the strips, that is, the distance
between each pair of adjacent Au electrodes, was 2 mm. Then, four
parts of the CdTe layer exposed in between each adjacent pair of Au
electrodes were scraped off with a cutter knife. In this way, a sample
for electrical characterization, having four ideal small Au contacts in
four vertices of square CdTe, could be obtained. Four Au wires were
fixed onto the side of the underlying epoxy resin with quick-drying
glue, and one end of each of the four Au wires was attached to each
of four Au electrodes with silver paste. Figure 2 shows a typical
picture of the sample.
The resistivity and Hall effect were measured at room tempera-
ture with a resistivity/Hall measurement system ͑Toyo Corp.
ResiTest 8310͒ using the van der Pauw method. The resistivity of
the CdTe layer was so high ͑e.g., 108 ⍀ cm͒ that Hall voltage mea-
surement by the conventional dc mode was impossible for our CdTe
deposits. Therefore, the ac magnetic field method was employed for
the measurement. The magnitude and frequency of the magnetic
field were 0.6 T at the maximum and 50 mHz, respectively. The
conduction type was distinguished by a phase shift between the
magnetic field and the Hall voltage; for this system, the type was
determined to be p- or n-type when the shift was in the range −60 to
30° or 120-210°, respectively. The density and mobility of the ma-
jority carrier were calculated from the absolute value of the Hall
voltage.
ence to measure the cathode potential; the potentials were recalcu-
lated for the SHE. A Cd sheet ͑99.99%; Nilaco Corp.͒ 100
ϫ 15 mm in area and 1.5 mm thick was used as the anode. The Cd
sheet was pretreated with 3 M HCl solution for 3 min and then
rinsed with DI water before each experimental run. The electro-
chemically active area of the Cd sheet immersed in the electrolyte
for CdTe deposition was 6 cm2 in total. A gold-plated copper sheet
polished with 0.3 µm alumina abrasive was used for the cathode
substrate; the thickness of the Au plating was 1-3 µm. A part of the
Au/Cu sheet surface was covered with Teflon adhesive tape so that
a set of two squared areas ͑about 10 ϫ 10 mm͒ was exposed to the
electrolytes as the surface of the substrate. The deposit on one
square was transferred onto an epoxy resin to examine the electrical
properties, and that on the other was used for characterization of the
crystallinity, composition, and thickness. The current was measured
by using a zero-shunt ammeter ͑Hokuto Denko HM-101͒ connected
to a coulometer ͑Hokuto Denko HF-201͒, and the cathode potential
was monitored with a conventional digital voltmeter. The solution
was thermostatted at 343 K ͑70°C͒ and stirred at approximately 500
rpm. CdTe deposition was performed under white light irradiation or
in the dark. A 500 W xenon arc lamp ͑Wacom KXL-500F͒ installed
in a lamp housing was used for the irradiation. The approximate
integrated irradiance of the white light just before the electrolytic
vessel was 300 mW cm−2. When deposited in the dark, the electro-
lytic cell was placed in a light-resistant box to eliminate any photo
effects on the growing CdTe. The concentration of Cd͑II͒ ions was
set at 40 mM under illumination and 60 mM in the dark, as opti-
mized for electrodeposition from ammoniacal basic electrolytes.2
Analysis of deposits.—The morphology and crystallinity of the
resulting deposits were examined with an X-ray diffractometer
͑Rigaku RINT 2200͒ fitted with a molybdenum X-ray tube. All the
X-ray diffraction ͑XRD͒ measurements were scanned in 2␪-␪ mode.
The thickness of the deposits was measured by surface texture mea-
surements ͑Tokyo Seimitsu 1400D-12M͒. The compositions of the
Results and Discussion
Deposition behavior of CdTe.—The open-circuit potentials of
the Au/Cu substrate and Cd sheet before CdTe deposition were
around 0.0 and −0.71 V, respectively. The two electrode potentials
became almost equal when the Au/Cu substrate was short-circuited
Downloaded on 2015-01-13 to IP 66.254.251.28 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

Journal of The Electrochemical Society, 152 ͑4͒ C237-C242 ͑2005͒
C239
Figure 4. Electrode potentials of Au/Cu substrate and Cd sheet vs. logarithm
of current in galvanic contact deposition. It is assumed that Reactions 3 and
1 occur on Au/Cu substrate and Cd sheet, respectively.
the exchange current density of Reaction 1, the redox of Cd, i.e., a
normal metal, is expected to be higher than that of CdTe deposition
͑Reaction 3͒ involving the redox of irreversible oxyanion TeO32−
.
Furthermore, the CdTe deposition is suppressed by the adsorption of
Cd͑II͒-ions, resulting in a smaller Tafel slope of the polarization
curve for Reaction 3.9 As a result of the difference in polarization
curves for the reactions, the potential Ec͑=Ea͒ during the contact
deposition is expected to be close to the Nernst potential of Reaction
1. If a Cd anode with a smaller surface area is employed, the depo-
sition proceeds at a more positive potential.
The initial current density under illumination was about
1.0 mA cm−2, while that in the dark was lower, 0.30 mA cm−2. In
both cases, the currents gradually decreased as the time elapsed.
These behaviors were observed in the case of normal electrodeposi-
tion of CdTe from the same electrolytes.2 In the previous work,2 we
conclude that the photo effect on the growth of CdTe is caused by a
photoconduction, where an extremely low carrier density, i.e., elec-
tron density, of the resulting CdTe layer in the dark is increased by
the irradiation of visible light, which is pumping up electrons to the
conduction band. By this photo-assisted deposition, it took 60 min to
flow the charge of 1.5 C cm−2, vs. 1694 min in the dark condition.
After the CdTe deposition, it was revealed by XRD and EPMA
that CdTe was also deposited on the Cd sheet that is an anode. The
formation of CdTe on the Cd sheet is explained by a simple dis-
placement reaction
Figure 3. Variation of ͑a͒ cathode potentials and ͑b͒ current densities during
galvanic contact deposition. Dashed lines in ͑b͒ are current during normal
electrodeposition under illumination and in the dark.
to the Cd sheet. Figure 3 illustrates the variation of cathode poten-
tials and current densities during the deposition under illumination
or in the dark. The cathode potentials were spontaneously kept at
around −0.7 V throughout the deposition both under illumination
and in the dark, indicating that a potentiostatic situation was more or
less realized without an external potentiostat with a reference elec-
trode. Note that the potential −0.7 V is almost the same as that of
the potential optimized for CdTe normal electrodeposition from am-
moniacal basic electrolytes using a conventional potentiostat. Figure
4 illustrates a set of schematic polarization curves for redox reac-
tions
2+
3Cd + TeO23− + 8NH3 + 6H+ → 2Cd͑NH3͒ + CdTe + 3H2O
4
͓4͔
2+
Cd͑NH3͒ + TeO23− + 4e + 6H+ ꢀ CdTe + 4NH3 + 3H2O
4
which proceeds spontaneously due to the negative Gibbs free energy
change ͑⌬G͒ of the reaction: −529 kJ mol−1 at 298 K, pH 10.7,
͓3͔
2−
2+
aTeO = 0.01, and aCd͑NH ͒ = 0.06; here a set of thermochemical
and Reaction 1. The curves are given as a relationship between the
electrode potentials and the logarithm of the current. During gal-
vanic contact deposition, the electrode potentials of the Au/Cu sub-
strate ͑Ec͒ and the Cd sheet ͑Ea͒ become almost equal, when the
IR-drop in the solution and lead wire is negligible. In this case, the
forward process of the electrochemically nobler reaction ͑Reaction
3͒, i.e., cathodic deposition of CdTe, takes place on the Au/Cu sub-
strate, while the backward process of the less noble reaction ͑Reac-
tion 1͒, i.e., anodic dissolution of Cd, occurs on the Cd sheet. Here,
3
data3in Table I of Ref. 3 was used for the calculation. Once the Cd
sheet is thoroughly covered with the CdTe layer, the CdTe growth
on the Au/Cu substrate is expected to cease, as the potential control
by Cd/Cd2+ redox is lost. However, the deposition of CdTe on the
Au/Cu substrate took place continuously even after the deposition of
CdTe on the Cd sheet. It seems that the continuous potential control
and the deposition of CdTe on the Au/Cu substrate was achieved by
the presence of cracks in CdTe on the anode. Through the cracks, Cd
4
Downloaded on 2015-01-13 to IP 66.254.251.28 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

C240
Journal of The Electrochemical Society, 152 ͑4͒ C237-C242 ͑2005͒
intensity. We have found that CdTe layers prepared from the ammo-
niacal electrolyte using normal potentiostatic electrodeposition have
a random orientation irrespective of their thickness.10 However, each
of the first and second peaks of the substrate has a weak shoulder
corresponding to 220 and 331 reflections. The XRD pattern of the
thicker deposit ͑c͒ gave other reflections such as 220 ͑2␪ = 17.8°͒,
331 ͑20.9°͒, 400 ͑25.3°͒, 331 ͑27.6°͒, and 422 ͑31.1°͒ due to the
stronger intensity.11 The relative intensities of the reflections were
similar to those of powder data, suggesting that the contact depos-
ited CdTe layers also have a random orientation. The mean crystal-
lite sizes of the CdTe deposits estimated from the half-width of the
111 diffraction peak using Scherrer’s equation were about 10 nm,
independent of thickness.
Composition and current efficiency.—The composition, thick-
ness, and current efficiency of the resulting deposits are summarized
in Table I. The current efficiency was calculated from the total
charge passed, the density of CdTe ͑5.9 g cm−3͒, and the thickness
of the deposits. Both under illumination and in the dark, the com-
positions of the deposits were nearly stoichiometric ͑49.4-49.6 atom
% Cd͒, although they were slightly richer in Te, as is often the case
with normal electrodeposition. Figure 6 depicts typical surface and
cross-sectional scanning electron microscopy ͑SEM͒ images of a
CdTe layer deposited under illumination. The sample for the cross-
sectional SEM observation was prepared by cutting the substrate
with the CdTe layer using a diamond wheel saw followed by me-
chanical polishing of the resulting cross section with a set of emery
papers and then with alumina abrasives ͑1.0 and 0.3 µm͒; a hand-
made stainless holder was employed for the mechanical polishing.
When assessed from the SEM image, it was found that the CdTe
layer was closely and uniformly grown on the substrate. By EPMA
line analysis along the white line in Fig. 6, it was confirmed that the
ratio of Cd and Te was almost constant in the direction of film
thickness. The thickness of the CdTe layer determined from the
image was about 13.5 µm, which is in agreement with that measured
by the surface texture measurements.
The current efficiency of the deposition was less than 50% in
dark conditions, and there was a tendency for the efficiency to de-
crease with increasing thickness of the deposits. However, under
illumination, the current efficiency of the deposition was approxi-
mately 100% independent of the thickness. This tendency of current
efficiency was also observed in normal electrodeposition from the
basic electrolytes.2 One of the plausible side reactions which lowers
the current efficiency is the reduction of dissolved oxygen and/or
hexavalent Te͑VI͒ ions as
Figure 5. XRD patterns of deposits by galvanic contact deposition ͑a͒ in the
dark and ͑b, c͒ under illumination. Thickness of deposits: ͑a͒ 0.6 µm, ͑b͒ 1.3
µm, ͑c͒ 13.5 µm.
can make contact with the solution, and the electrode potential was
kept negative. Such a displacement reaction on the Cd sheet can be
avoided by using a two-conpartment cell where an electrolyte with-
out Te͑IV͒ ions is used as an anolyte.
O2 + 4H+ + 4e → 2H2O
͓5͔
TeO24− + 2H+ + 2e → TeO23− + H2O
Although tellurium ions are initially dissolved as tetravalent Te͑IV͒
͓6͔
Morphology of deposits.—Figure 5 shows typical XRD patterns
of deposits made under illumination or in the dark. Every diffraction
peak could be assigned to that of CdTe and Au/Cu substrate; no
peaks attributable to elemental Cd or Te were observed. This indi-
cates that both deposits under illumination and in the dark consist of
a single phase of polycrystalline CdTe. CdTe layers electrodeposited
from ammoniacal basic electrolytes have random orientations and
mean crystallite sizes of approximately 10 nm.10 CdTe deposits by
contact deposition appeared to have a ͗111͘ preferential orientation
for the thinner deposits ͑a and b͒, because the reflections other than
ions, they can be spontaneously oxidized by dissolved oxygen as
2TeO23− + O2 → 2TeO42−
͓7͔
because E°Te͑IV͒/Te͑VI͒ at pH 10.7 is 0.26 V,12 which is lower than the
E°H O/O of 0.60 V. It is assumed that the difference of current effi-
ciency 2under illumination and in the dark is attributable to the dif-
ference of each deposition rate. Given that half of the current is
caused by the side reaction in dark conditions and that the side
2
11
111 index ͑2␪ = 10.9°͒ were not clearly seen due to their weak
Table I. Total charge passed and time during deposition, thickness and composition of deposits, and current efficiency. Current efficiencies
calculated from total charge and volume of deposits.
Total charge ͑C cm−2
͒
Depositiontime͑min͒
Thickness ͑µm͒
Composition ͑atom % Cd͒
Currentefficiency͑%͒
Under illumination
In the dark
1.9
18.8
1.9
60
1980
1694
8220
1.3
13.5
0.6
49.6
49.5
49.4
49.6
97
100
46
8.3
1.7
29
Downloaded on 2015-01-13 to IP 66.254.251.28 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

Journal of The Electrochemical Society, 152 ͑4͒ C237-C242 ͑2005͒
C241
Figure 7. Reflectivity spectra of ͑a͒ single-crystal CdTe plate and CdTe
layers prepared by ͑b͒ galvanic contact deposition and ͑c͒ electrodeposition
from ammoniacal basic electrolyte. Single-crystal CdTe plate was 1 mm
thick. Both contact deposited and electrodeposited CdTe were 5 µm thick.
Evaluation of the bandgap energy.—Figure 7 shows the reflec-
tivity spectra of CdTe prepared by galvanic contact deposition and
by normal electrodeposition compared with that of the single-crystal
CdTe. The thickness of the CdTe layers prepared by contact depo-
sition and by normal electrodeposition was approximately 5 µm. The
single-crystal CdTe was a plate 1 mm thick. In all the spectra, a
decrease in the reflectivity was observed at almost the same energy,
which was attributed to the bandgap energy. The bandgap energy of
the contact deposited CdTe evaluated from the spectra was 1.45 eV,
the same as that of both the electrodeposited and single-crystal
CdTe. However, the reflectivity gradually decreased from the energy
below 1 eV in the contact deposited and normal electrodeposited
CdTe, whereas the decrease in the reflectivity was very steep in the
single-crystal CdTe. In other words, the absorptions at the bandgap
energy of both the contact deposited and normal electrodeposited
CdTe were less clear than that of the single-crystal CdTe. This indi-
cates that both the contact deposited and normal electrodeposited
CdTe have some located impurity energy levels which were attrib-
uted to several defects such as Cd or Te vacancies and grain bound-
aries.
Electrical properties.—The usual thickness of CdTe layers em-
ployed for a CdTe/CdS solar cell is 1-2 µm. Therefore, the electrical
properties of CdTe in such a thickness range may be of interest in
view of practical applications. However, the electrical properties of
thin CdTe layers ͑below 3 µm thick͒ could not be determined due to
the high resistivity of CdTe. Therefore, the electrical properties were
measured for thicker layers ͑more than 10 µm͒. In this case, the
layers were prepared under illumination; it was difficult to prepare
such thick layers in the dark due to a slow deposition rate. The
resistivity of the as-deposited CdTe layer ͑13.5 µm thick͒ deter-
mined by the van der Pauw method was 5.5 ϫ 107 ⍀ cm, which is
Figure 6. Typical ͑a͒ surface and ͑b͒ cross-sectional SEM images of CdTe
deposited by galvanic contact deposition, and ͑c͒ line analysis along the
white line in ͑b͒ by EPMA. Thickness of deposits was determined to be 13.5
µm.
reactions are limited by mass transfer from the solution bulk to the
cathode, only the current for CdTe deposition can increase due to
illumination. In this case, the current efficiency under illumination is
calculated to be approximately 90%. If that is the case, it is expected
in dark conditions that the Te͑VI͒ species accumulate in the solution
with time elapsed, and, hence, the current efficiency decreases. As a
result, the current efficiency for the deposition for 8820 min was
lower than that for 1694 min ͑see Table I͒. Recently, we have ex-
amined in situ the change in the current efficiency during normal
electrodeposition using an electrochemical quartz crystal microbal-
ance. Here, the current efficiency was increased up to 90-95% in the
earlier stage of the electrodeposition, and afterwards it gradually
decreased to 60-80%.13
the same level as that of CdTe electrodeposited from acidic
14-16
electrolytes ͑104−108 ⍀ cm͒
and from basic electrolytes ͑107
−109 ⍀ cm͒.8 The conduction type of the CdTe was p-type in anal-
ogy with CdTe layers electrodeposited from the same basic electro-
lytes, whereas CdTe layers electrodeposited from acidic electrolytes
are usually n-type when the deposition potential is close to the
Nernst potential for bulk Cd deposition.
The carrier density of CdTe deposited by galvanic contact depo-
sition was 4.4 ϫ 1011 cm−3, which was five orders of magnitude
larger than that of intrinsic CdTe ͑ca. 106 cm−3͒.8 The Fermi level
EF for the CdTe layer was calculated from the value of the carrier
density, which was equal to the hole density in this case, by the
Downloaded on 2015-01-13 to IP 66.254.251.28 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

C242
Journal of The Electrochemical Society, 152 ͑4͒ C237-C242 ͑2005͒
following equation to be located at 0.42 eV above the top of the
valence band
of Excellence Program “United Approach to New Materials Sci-
ence” from the Ministry of Education, Culture, Sports, Science and
Technology of Japan.
p = NV exp͕−͑EF − EV͒/kT͖
NV = 2͑2␲m*hkT/h2͒3/2
͓8͔
Kyoto University assisted in meeting the publication costs of this article.
͓9͔
References
where p is the hole density, NV is the effective density of states in
the valence band, EV is the energy of the top of the valence band, k
is the Boltzmann constant, T is the absolute temperature, m*h is the
effective mass of the hole; mh* = 0.35m ͑m: mass of free electron͒,16
and h is the Planck constant. The value of the Fermi level 0.42 eV of
this CdTe film is the same level as that of the electrodeposited CdTe
films with p-type conduction.8,17,18
1. A. Kampmann, P. Cowache, J. Vedel, and D. Lincot, J. Electroanal. Chem., 387, 53
͑1995͒.
2. K. Murase, M. Matsui, M. Miyake, T. Hirato, and Y. Awakura, J. Electrochem. Soc.,
150, C44 ͑2003͒.
3. K. Murase, H. Watanabe, T. Hirato, and Y. Awakura, J. Electrochem. Soc., 146
1798 ͑1999͒.
4. B. E. McCandless, A. Mondal, and R. W. Birkmire, Sol. Energy Mater. Sol. Cells,
36, 369 ͑1995͒.
The hole mobility for the CdTe layer was 0.24 cm2 V−1 s−1
,
which lay in the same level as that for electrodeposited CdTe and
was much smaller than those reported for single-crystal CdTe ͑50
-100 cm2 V−1 s−1͒.19-22 Considering that the grain size of the CdTe
is as small as 10 nm, the origin of the small mobility of the CdTe
may be the trapping of carriers at grain boundaries.23,24
5. K. R. Murali, I. Radhakrishna, K. Nagaraja Rao, and V. K. Venkatesan, J. Mater.
Sci., 25, 3521 ͑1990͒.
6. A. Mondal, B. E. McCandless, and R. W. Birkmire, Sol. Energy Mater. Sol. Cells,
26, 181 ͑1992͒.
7. R. N. Bhattacharya and K. Rajeshwar, J. Electrochem. Soc., 131, 939 ͑1984͒.
8. M. Miyake, K. Murase, T. Hirato, and Y. Awakura, J. Electrochem. Soc., 150, C413
͑2003͒.
Conclusion
9. K. Murase, H. Watanabe, S. Mori, T. Hirato, and Y. Awakura, J. Electrochem. Soc.,
146, 4477 ͑1999͒.
Flat, smooth, polycrystalline CdTe layers were deposited from
ammoniacal basic aqueous solution by a galvanic contact deposition
technique. All the features and the deposition behavior of the CdTe
layer were the same as those observed for normally electrodeposited
CdTe using an external current source. The current density could be
increased by irradiating white light to the cathode surface during
deposition. All the CdTe deposits were nearly stoichiometric ͑49.4-
49.6 atom % Cd͒. In dark conditions, the current efficiency of depo-
sition was less than 50% and the thicker the deposits, the more the
efficiency decreased. However, under illumination, the current effi-
ciency of deposition was approximately 100% independent of thick-
ness. The bandgap energy of the contact deposited CdTe was evalu-
ated as 1.45 eV, which was almost the same as that of the single-
crystal CdTe. The conduction type of the CdTe deposited by
galvanic contact deposition was p-type, and the resistivity, carrier
density, and mobility were 5.5 ϫ 107 ⍀ cm, 4.4 ϫ 1011 cm−3, and
0.24 cm2 V−1 s−1, respectively.
10. M. Miyake, H. Inui, K. Murase, T. Hirato, and Y. Awakura, J. Electrochem. Soc.,
151, C168 ͑2004͒.
11. Card no. 15-0770, X-Ray Powder Diffraction Standards, ASTM, Philadelphia, PA.
12. E. Deltombe, N. de Zoubov, and M. Pourbaix, Atlas of Electrochemical Equilibria
in Aqueous Solutions, J. W. Arrowsmith, Bristol, U.K. ͑1966͒.
13. K. Murase, Y. Tanaka, T. Hirato, and Y. Awakura, Abstract 672, The
Electrochemical Society Meeting Abstracts, Vol. 2003-1, Paris, France, April
27-May 2, 2003.
14. B. M. Basol and O. M. Stafsudd, Solid-State Electron., 24, 121 ͑1981͒.
15. B. M. Basol, Sol. Cells, 23, 69 ͑1988͒.
16. A. E. Rakhshani, Y. Makdisi, X. Mathew, and N. R. Mathews, Phys. Status Solidi
A, 168, 177 ͑1998͒.
17. J. A. von Windheim and M. Cocivera, J. Phys. Chem. Solids, 53, 31 ͑1992͒.
18. M. Miyake, K. Murase, T. Hirato, and Y. Awakura, J. Electroanal. Chem., 562, 247
͑2004͒.
19. D. de Nobel, Philips Res. Rep., 14, 430 ͑1959͒.
20. O. Mandelung, Grundlagen der Hulbesterphysik, Springer-Verlag, Berlin ͑1970͒.
21. S. M. Sze, Physics of Semiconductor Device, 2nd ed., John Wiley & Sons, New
York ͑1981͒.
Acknowledgments
The authors thank Dr. M. Funato from Kyoto University for his
help with the Hall effect measurements. This work was supported by
Grants-in-Aid for Scientific Research ͑no. 14350399 and 15760548͒
from the Japan Society for the Promotion of Science. The work was
also partly supported by the Kyoto University 21st Century Center
22. Handbook of Chemistry and Physics, 63rd ed., CRC Press, Boca Raton, FL ͑1982͒.
23. J. Y. W. Seto, J. Appl. Phys., 46, 5247 ͑1975͒.
24. W. Huber, A. L. Fahrenbruch, C. Fortmann, and R. H. Bube, J. Appl. Phys., 54,
4038 ͑1983͒.
Downloaded on 2015-01-13 to IP 66.254.251.28 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).