NCPV preprints for the 2. world conference on photovoltaic solar energy conversion Page: 61 of 144
This article is part of the collection entitled: Office of Scientific & Technical Information Technical Reports and was provided to Digital Library by the UNT Libraries Government Documents Department.
The following text was automatically extracted from the image on this page using optical character recognition software:
to sort out the contributions of intergrain and intragrain
effects. Single crystals were chosen for this purpose.
The SIMS profile of a Cd PE treated single crystal
sample is shown in Fig. 5. Here, the difference in Cd
concentrations between the baseline (as grown crystal), and
the Cd treated crystal is very clear, and the Cd signal
reaches the baseline value. The in-diffusion of Cd
coincides with a slight depletion of Cu level in the front
region (not shown). We conclude from the above data that
Cd is indeed incorporated in the absorber films. The
mechanism of the reaction through which this occurs must
be clarified. To explore this further, we soaked several
large pieces of single crystals in a solution of 1.0 M CdSO4
in 30% ammonium hydroxide at room temperature. The
concentrations in this mixture are much greater than one
would use in a CBD reaction, and the purpose here is to
exaggerate the effects in a reasonable time. After 4 days,
the solution was decanted and analyzed by inductively
coupled plasma spectroscopy. The results were compared
with a blank solution. The atomic concentrations, in ppm,
were as follows, where the blank values are given in
parentheses: Cu: 0.90 (0.02); In: 0.08 (0.01); and Se: 0.47
(0.09). We find that Cu is preferentially leached out, and
some Se is also carried with it The crystals were cleaned
thoroughly in ultrasonic bath and then examined by XPS.
The depth profile of Cd treated crystal shows the presence
of Cd (see Fig. 6).
Further analysis of the shapes of the Cd 3d and Se 3d
Auger lines demonstrates that the Cd is bound to Se, and
the peak positions can be assigned to CdSe. The line
shapes of the Cd and Se d levels are both altered by the Cd
diffusion. These details will be published separately.
To summarize, our results suggest that the interaction
of the absorber with the CBD solution results in an
interfacial reaction. This might be the most important
aspect of the CBD CdS process. The evidence for this is
the fact that excellent devices are obtained simply by
treating the absorbers in a combination of a Cd salt and
ammonium hydroxide. The implication here is that the
introduction of Cd can dope the absorber n-type and create
a buried junction. This is consistent with the early
observations on single crystals where Cd and Zn were
found to dope the crystals n-type . In studying the time
dependence of this reaction, we found that the Cd profiles
are established at low temperatures (400C) and short times
(1 min). One might envisage an ion-exchange reaction
where the out-diffusion of Cu is driven by the in-diffusion
of Cd; both facilitated by the ability of ammonia to
complex the ions. The creation of a thin, heavily doped n-
region can explain much of the observed phenomena. In a
CdS device, the electric field will be supported entirely, or
mostly, by this n-region, depending on the doping level. It
is also quite possible that the Cu-deficient nature of the
surfaces present natural sites for the occupation by Cd. In
any case, we find that an enhanced n-doped region is more
likely the heart of the device, rather than the CdS layer
itself. The latter undoubtedly offers the benefits of a
lattice-matched window layer, and it shields the absorber
from ZnO and sputter damage. The foregoing discussion
applies equally well to Zn as we have demonstrated with
the ZnPE treatments.
32. Alternative Junction Strategies
The solutions to finding an alternative method for
forming the junction follow from the previous section. Our
approach is to develop methods for forming an n- type,
emitter region by extrinsic doping. The most obvious
candidates appear to be the IrB elements substituting for
Cu; IV elements for the In or Ga; and the halogens for the
Se atoms. The efficiency of the Cd "doping" suggests that
Zn is the logical choice. Similarly, Cl or F could be
effective on the Se sites. We have experimented with a
variety of Zn sources, including elemental Zn. ZnC2 was
chosen as the best candidate because of its low melting
point and high vapor pressure at 2000C. The method of
delivering the ZnC12 to the absorber surface has been
varied. In one approach, the ZnC2 is dispersed in
methanol and applied to the CIGS film. Following an
annealing step at 2000C for 1 h, the reaction products are
chemically removed by etching in HC. The standard
bilayer ZnO is sputtered to complete the device. The best
result obtained by this method is a 13.5% efficiency device,
and it did not require a light soak or heat treatment to
realize the efficiency. The parameters of this device are:
V.e= 0.527 V, J= 36.01 mA.cm-2, and FF = 0.71. On
0 2 4 6 8 10
FIg 6: XPS profile of Cd in CuInSe2 single crystal after
soaking for4 days in 1.0 M CdSO4, 30%NH4OH.
400 600 800 1000 1200
FIg. 7: Spectral response curves for CIGS devices with and
-9- Cd-treated -
T 2 2 iE I i
Here’s what’s next.
This article can be searched. Note: Results may vary based on the legibility of text within the document.
Tools / Downloads
Get a copy of this page or view the extracted text.
Citing and Sharing
Basic information for referencing this web page. We also provide extended guidance on usage rights, references, copying or embedding.
Reference the current page of this Article.
NCPV preprints for the 2. world conference on photovoltaic solar energy conversion, article, September 1, 1998; Golden, Colorado. (https://digital.library.unt.edu/ark:/67531/metadc707815/m1/61/: accessed March 20, 2019), University of North Texas Libraries, Digital Library, https://digital.library.unt.edu; crediting UNT Libraries Government Documents Department.