Thin Solid Films 615 (2016) 271–280

Contents lists available at ScienceDirect

Thin Solid Films

journal homepage: www.elsevier.com/locate/tsf

Iron pyrite thin films grown through a one-step annealing of iron oxide
using sulfur sources, tert-butyl disulfide and H2S
Siva P. Adusumilli a,b, Jeremiah M. Dederick d, In-Tae Bae c, Sean M. Garner e, Anju Sharma c,
Charles R. Westgate a,b, Tara P. Dhakal a,b,⁎
a
Department of Electrical and Computer Engineering, Binghamton University, Binghamton, NY 13902, United States
b
Center for Autonomous Solar Power (CASP), Binghamton University, Binghamton, NY 13902, United States
c
The Small Scale Systems Integration and Packaging Center (S3IP), Binghamton University, Binghamton, NY 13902, United States
d
Department of Physics, Applied Physics and Astronomy, Binghamton University, Binghamton, NY 13902, United States
e
Corning Incorporated, Corning, NY 14831, United States

a r t i c l e i n f o a b s t r a c t

Article history: In this work, we report synthesis of pyrite thin films using tert-butyl disulfide (TBDS) and hydrogen sulfide (H2S)
Received 7 January 2016 in one-step atmospheric pressure sulfurization of iron oxide films at 400 °C on a soda-lime glass, molybdenum
Received in revised form 5 July 2016 coated soda-lime glass and sodium-free glass substrates. The iron pyrite thin films grown using TBDS did not re-
Accepted 7 July 2016
quire the presence of sodium to form the pyrite phase, whereas H2S grown pyrite thin films did. It was observed
Available online 8 July 2016
that the pyrite formation and thus the sulfur diffusion into the oxide film was slower in TBDS compared to H2S.
Keywords:
The synthesized films were characterized for their surface morphology and phase identification using scanning
Earth abundant absorbers electron microscopy, transmission electron microscopy (TEM), X-ray diffraction (XRD) and X-ray photoelectron
Photovoltaics spectroscopy (XPS) techniques. The S:Fe atomic ratio as well as their chemical bonding states were monitored to
Hydrogen sulfide obtain and maintain a stoichiometric 2:1 ratio through the entire film thickness as a function of the sulfurization
tert-Butyl disulfide time by performing an XPS depth profile. Transmittance measurements confirmed the pyrite phase with an op-
Pyrite tical bandgap of 1.15 eV. The TEM electron-beam diffraction spots were used to verify the impurity phases ob-
Flexible glass served in XRD patterns. Hall Effect measurements showed p-type carriers for the pyrite films.
© 2016 Elsevier B.V. All rights reserved.

1. Introduction Since the 1980s, various research groups have proposed and synthe-
sized iron pyrite thin films using a variety of experimental techniques
Earth abundant materials have great potential in producing low cost which include sputtering [10], spray pyrolysis [11–13], chemical vapor
and high efficiency solar cells [1]. Iron disulfide a.k.a. pyrite (FeS2) is one transport [5], metal-organic chemical vapor deposition [14,15], direct
such earth-abundant semiconductor that exhibits useful photovoltaic chemical vapor deposition [16], electrodeposition [17], nanocrystal for-
behavior with a direct bandgap of 0.95 eV, an optical absorption coeffi- mation [9,18–21], hydrothermal methods [2,22], evaporation [23,24].
cient of the order of 105 cm−1 [2] and long minority carrier diffusion Some of the sulfur precursors/sources used for pyrite are di-tert butyl di-
lengths [3]. In spite of these advantages, the best efficiency observed is sulfide (TBDS) [25–27], hydrogen disulfide (H2S) [14,28,29] and ele-
2.8% using a photochemical cell with an open circuit voltage of mental sulfur (S) [3,30]. Iron sources used are Fe [30,31], FexOy [3,9,
187 mV [4]. This value of the open circuit voltage is well below the the- 32], iron pentacarbonyl (Fe(CO)5) [25,27,33], and iron acetylacetonate
oretical prediction of 500 mV [5]. This modest performance is usually at- (Fe(acac)3) [18,34,35]. Compound semiconductor pyrite was formed
tributed to sulfur deficiency and the resulting bulk deep donor states [6, by cobalt doping for n-type pyrite [15]. Regarding sulfur precursors, sin-
7], high density of thermodynamically unstable surface states [8,9], and gle or combination of sulfur precursors [18,35,36] are used.
the presence of trace amounts of orthorhombic marcasite and hexago- In this work Fe2O3 was chosen as the iron source with TBDS and H2S
nal troilite phases which have very small bandgaps of 0.34 eV and as the sulfur sources. The choice of Fe2O3 is preferred because of the for-
0.04 eV respectively [2]. More investigation is needed for the synthesis mation of the pyrite without going through the intermediate phase such
of a quality pyrite material with regards to its purity, stoichiometry, as FeS as depicted in Fig. 1(Fe\\O\\S Gibbs free energy phase diagram)
crystallinity and grain size for the solar cell applications. [4,37]. When only Fe is used, the pyrite formation has to go through the
monosulfide phase to get into disulfide region. The impetus of this work
⁎ Corresponding author. is to synthesize phase pure and stoichiometric pyrite film for application
E-mail address: tdhakal@binghamton.edu (T.P. Dhakal). as a photovoltaic absorber.

http://dx.doi.org/10.1016/j.tsf.2016.07.016
0040-6090/© 2016 Elsevier B.V. All rights reserved.
272 S.P. Adusumilli et al. / Thin Solid Films 615 (2016) 271–280

to study the effect of sodium (Na) in pyrite phase formation. The iron
oxide films were sulfurized in TBDS and H2S environments using a
one-zone quartz tube furnace. Prior to the sulfurization of the Fe2O3
films, the quartz tube was baked for 2 h at 400 °C and then naturally
cooled to room temperature to outgas any chamber contamination.
The flow of nitrogen (N2) gas was controlled through mass flow control-
lers (MFC) as shown in Fig. 2. In case of TBDS, N2 was bubbled through
the precursor via MFC-1 to achieve an appreciable amount of TBDS
vapor, which then combined with an additional N2 flow through MFC-
2 before entering the quartz tube. The annealing temperature was
kept at 400 °C, while the time was varied from 6 h to 4 days. The anneal-
ing was performed in Argon (Ar) gas flow as well, but the results were
the same as those for N2. After annealing, the precursor Fe2O3 film con-
verted into pyrite (FeS2) phase.
The setup for H2S sulfurization was similar to the TBDS as shown in
Fig. 2. In this case, H2S and N2 gases were sent through mass flow con-
trollers, MFC-3 and MFC-4 respectively. The temperature of annealing
was same as that of the TBDS case.
Fig. 1. Gibbs free energy phase diagram of Fe\
\O\
\S [reproduced with permission from ref.
37].
2.2. Characterization techniques

2. Experimental details The film surface morphology of the synthesized FeS2 films was im-
aged using a Supra 55 VP high-resolution scanning electron microscope
2.1. Iron oxide film deposition (HR-SEM). The crystallinity of the films and their phase identification
were investigated using PanAnalytical X'Pert PRO X-ray diffraction sys-
Prior to iron oxide (Fe2O3) film growth, the glass substrates were tem, which used Cu Kα X-rays. The Bragg peak identification was car-
cleaned with soapy water followed by ultra-sonication in dilute HCl ried out with the help of International Centre for Diffraction Data
(25% in DI water), acetone and ethanol in succession. Then the sub- software. Compositional analysis of the surface and bulk of the sulfu-
strates were rinsed in DI water and dried using the clean room grade rized films was performed on a PHI 5000 Versaprobe XPS system from
dry nitrogen. A 600 nm thick Mo film was then sputtered onto the Physical Electronics, Inc. that employed monochromatic Al Kα X-rays
glass substrate using a target with 99.95% purity purchased from Kurt of energy 1486.6 eV. The X-ray spot size used was 200 μm with power
J. Lesker Co. The sputtered molybdenum (Mo) had a sheet resistance 50 W. Pass energies used were 117 eV for survey scans (0–1400 eV)
of about 5 Ω/sq. For iron oxide film, a Fe2O3 sputtering target of 99.9% and 23.5 eV for high resolution spectra. The takeoff angle used was
purity (from Kurt J. Lesker) was used. Thin films of Fe2O3 of approxi- 45°. The survey scans were taken after a two-minute Ar sputter cleaning
mately 300 nm thickness were sputtered at room temperature on of the surface at 0.5 kV. During depth profiling, Ar sputtering at 2 kV was
bare as well as Mo-coated glass substrates, allowing us to observe the used. High-resolution transmission electron microscopy (HR-TEM) im-
effect of Mo acting as a blocking layer for Na leaching. The sputter- ages were obtained from JEM 2100F from JOEL. Transmittance and re-
coated films were confirmed as Fe2O3 by X-ray photoelectron spectros- flectance spectra were collected using Perkin Elmer Lambda 950 UV–
copy (XPS) and X-ray diffraction (XRD) measurements. vis-NIR spectrophotometer equipped with a 150 mm integrating
Two types of glasses were used; soda-lime glass from Cardinal Glass sphere. Hall voltages and mobilities were measured using the van der
Inc. and sodium-free Corning® EAGLE XG® Glass [38]. This allowed us Pauw method at 10 μA current in a magnetic field of 0.6 T. For contacts,

Fig. 2. TBDS and H2S based experimental setups for sulfur annealing of iron oxide films.
 S.P. Adusumilli et al. / Thin Solid Films 615 (2016) 271–280 273

Fig. 3. Sputtered oxide films grown on Mo-coated glass. a) Surface view and b) cross-sectional view.

125 nm thick gold over 20 nm of chromium was deposited using a cryo- The SEM images of sulfurized iron oxide thin films using TBDS setup
pump based thermal evaporator from Varian. at various annealing periods of 2 h, 6 h, and 4 days (96 h) are shown in
Fig. 4a–c. As the time of annealing increased, the grains grew larger with
some grains protruding higher. The cross-sectional image of the 4-day
3. Results and discussions annealed sample showed compact pyrite layer (Fig. 4d). The thickness
of the converted iron pyrite layer was 630 nm. A grainier thin layer
3.1. Scanning electron microscopy (SEM) (around 86 nm) between iron pyrite and Mo named “Layer 1a” was
also observed. This layer was sodium rich due to sodium leaching
Scanning electron microscopy images of iron oxide films sputtered from the glass through the Mo layer. This layer could be avoided either
on Mo sputtered glass substrate are shown in Fig. 3. Fig. 3a shows the by increasing the Mo-thickness or using a Na-blocking layer between
surface of the sputtered Fe2O3 film on a Mo-coated glass substrate and the Mo and the oxide precursor layer.
Fig. 3b shows the corresponding cross-sectional image. The thicknesses Fig. 5 shows the images of the pyrite films obtained with H2S anneal-
of Mo and Fe2O3 are 600 nm ± 50 nm and 300 nm ± 50 nm respective- ing. Fig. 5a and b show the surface and cross-sectional images of the
ly. The Mo layer was used because it is a potential back contact metal for annealed pyrite film (confirmed from XRD) grown on a bare glass sub-
solar cell devices. strate. The film thickness was 625 nm. The film surface and the cross-

Fig. 4. Sulfurized iron oxide films on Mo-coated glass substrates using the TBDS setup. Surface view of the films at various annealing time periods; a) 2 h b) 6 h and c) 4 days. d. Cross-
sectional view of the pyrite film on Mo.
274 S.P. Adusumilli et al. / Thin Solid Films 615 (2016) 271–280

Fig. 5. Sulfurized iron oxide on glass using the H2S setup a) surface and b) cross-sectional view; sulfurized iron oxide on Mo-coated glass using H2S setup c) surface and d) cross-sectional
view.

section show distinct granular texture. When the same growth condi- 3.2. X-ray diffraction
tions were used on Mo-coated glass substrates, the surface and cross-
sectional images showed compact films (Fig. 5c–d). Similar to TBDS an- XRD patterns of the thin films annealed in a) TBDS and b) H2S setups
nealing, an intermediate layer just above the Mo-layer (termed “Layer using Fe2O3 thin films grown on bare glass substrates containing sodi-
2” in Fig. 5d) was observed. However, these sulfurized films on Mo- um are shown in Fig. 6. The sulfur annealing time periods were varied
coated glass showed several non-pyrite phases as confirmed by the from 6 h to 48 h to study the rate of sulfur diffusion in both setups.
XRD (see Fig. 7). When H2S was used as the sulfur source, the presence The TBDS sulfurized films were pyrite phase although a few Fe2O3
of Na was necessary to form the pyrite phase. peaks were present, which disappeared slowly as the time of annealing

Fig. 6. XRD pattern of iron sulfide thin films grown for different time periods ranging from 2 h to 48 h using TBDS setup (orange) and H2S setup (blue).
 S.P. Adusumilli et al. / Thin Solid Films 615 (2016) 271–280 275

Fig. 7. XRD pattern of iron sulfide thin films grown using TBDS and H2S for 2 days and 4 days on Mo-coated glass and Corning® EAGLE XG® glass substrates.

Fig. 8. XPS depth profiles of a) 2 days and b) 4 days sulfur annealed iron oxide thin films on Mo glass using TBDS setup.

increased. However, the conversion of oxide films into the pyrite phase
was faster in case of H2S annealing (Fig. 6b). Even for 6 h annealing, no
oxide phases were present although some peaks of monosulfide troilite
and pyrrhotite phases with nominal intensity were observed. The
phases disappeared with increases in annealing time. This implies that
the conversion of Fe2O3 to FeS2 is slower in TBDS than in H2S.
The most important finding of this work is shown in Fig. 7. It has
been reported previously that Na is required to form pyrite [36].
When H2S was used, sulfurized films on Mo-coated soda-lime glass or
sodium free Corning® EAGLE XG® glass showed phases other than py-
rite (Fig. 7a); however, when a Na containing bare glass substrate was
used, only the pyrite phase was observed (Fig. 6b). When TBDS was
used during annealing, only pyrite phases were observed whether it
was on a Na-free substrate such as Corning® EAGLE XG® glass
(Fig. 7b) or a Na-containing bare glass (Fig. 6a). Thus the presence of
Na is not necessary for pyrite phase formation when TBDS is used for
sulfurization. This is important for solar cell applications because the py-
rite film has to be formed on a metallic layer, such as Mo. Berry et al. re-
ported marcasite phase when Mo was coated on the soda lime glass
Fig. 9. Graph indicating the FeS2 stoichiometry with thin film depth (bulk composition) of because it blocked most of the sodium required for pyrite phase forma-
various samples. tion [36]. This marcasite phase had to be converted back to the pyrite
276 S.P. Adusumilli et al. / Thin Solid Films 615 (2016) 271–280

phase by annealing in sulfur at 550 °C. However, in our work we showed

that if TBDS is used as a sulfur precursor, pyrite films can be formed at
lower temperature (400 °C) without the need of sodium.

3.3. X-ray photoelectron spectroscopy (XPS)

The XPS spectra obtained for the sulfur S 2p (2p1/2 and 2p3/2) and
iron Fe 2p3/2 peaks confirmed the presence of the pyrite phase [39].
The depth profile data was used to estimate the ratio of the iron to sulfur
and to evaluate the conversion of oxide film into pyrite phase over the
entire thickness of the film.

3.3.1. TBDS-XPS depth profiling

The XPS depth profiles of 2 days (48 h) and 4 days (96 h) TBDS-
annealed Fe2O3 thin films on Mo-glass substrates are shown in Fig. 8a
and b respectively. For 48 h of annealing, the entire 300 nm of Fe2O3
film appeared sulfurized. No presence of oxygen (O1s peak) was seen
before the rise of Mo3p3 indicating the beginning of the Mo-layer
(Fig. 8a). However, the sulfur to iron atomic ratio was b2.0 (Fig. 9) indi-
cating the presence of some amorphous FeS phases within the film
which wasn't possible to detect by XRD (see Fig. 7b). A small quantity
of Mo (b5%) in the bulk of the film was observed. The rise of O1s at
the interface of Fe2O3 and Mo and the presence of the sulfur peak indi-
Fig. 10. Raman Spectra of iron oxide, pyrite mineral and pyrite films (on soda lime glass cate the presence of MoS2 and/or some mixed phase such as MoOxS2-x
substrate) annealed in TBDS and H2S respectively for 48 h. as suggested in previously reported works [18,40].

Fig. 11. HR-TEM images of a) TBDS 48 h on bare glass b) H2S 48 h on Mo-coated glass c) TBDS 48 h on Mo-coated glass and d) TBDS 4 days on Mo-coated glass films.
 S.P. Adusumilli et al. / Thin Solid Films 615 (2016) 271–280 277

growth methods with respect to annealing time periods. The ratios are
depicted in Fig. 9.
Increasing the sulfur annealing time increased the atomic ratio clos-
er to 2. In case of TBDS annealing, the film annealed for 4 days main-
tained the average S:Fe ratio at 2. As mentioned earlier, the ratio was
above 2 on the surface. The ratio was slightly b2 for the 2 days TBDS
annealed film.
When H2S was used as the sulfur source, the sulfurized films on Mo-
coated glass did not show the pyrite phase (Fig. 7). Although the S:Fe
ratio was fairly close to 2 for 2-days H2S annealed film, the phase was
not pyrite. On the other hand, the film grown on a Na-containing bare
glass annealed for same time in H2S environment showed pyrite
phase, but was clearly sulfur deficient (black curve in Fig. 9).

3.4. Raman spectroscopy

Raman spectra were obtained for the as-deposited Fe2O3 and

annealed pyrite films using a 532 nm laser. Fig. 10 shows the Raman
peaks for both TBDS and H2S annealed pyrite films on soda lime glass
substrates (c.f. XRD graph of Fig. 6), which matched nearly perfectly
Fig. 12. TEM e-beam diffraction of H2S grown film on Mo-coated glass indicating [131]
with the peaks at 336 cm−1, 372 cm−1, and 424 cm−1 of the pyrite min-
zone axis of marcasite and the Bragg's reflections.
eral. These peaks also matched with the reported values for pyrite films
[9,29,36,42]. In addition, the annealed pyrite films did not show any
When the annealing time period was increased to 4 days (96 h) peaks corresponding to the precursor Fe2O3 film or the impurity phase
under the same growth conditions, the S/Fe ratio in the bulk was close such as marcasite [42].
to the desired value of 2.0 (see Figs. 8b and 9). The S/Fe ratio was slightly
N2.0 at the surface of the film. This could be due to the presence of ex- 3.5. High-resolution transmission electron microscopy (HRTEM)
cess elemental sulfur resulting from the decomposition of the organic
precursor TBDS. Another interesting fact is the absence of the Mo3p3 Fig. 11 shows the TEM cross-sectional images of the pyrite film
peak in the bulk of the pyrite film for the 4 days annealed case grown on bare glass and Mo-coated glass using TBDS and H2S for 48 h
(Fig. 8b). There was some presence of Mo in the bulk of the film for and 96 h. The film grown on the bare glass (Fig. 11a) is highly porous,
2 days annealed sample as shown in Fig. 8a. The presence of Mo in the but with the Mo-layer the films were fairly compact (Fig. 11b, c, and
2 days-annealed film could be due to a slightly more porous intermedi- d). The TBDS annealed film became more compact when the time of an-
ate layer compared to the 4 days-annealed film (compare Fig. 11c and nealing increased from 48 h to 96 h (4 days) (Fig. 11c and d). The 48 h
d). In addition, sodium presence is significant in the 4 days-annealed annealed film in both H2S and TBDS environments had similar appear-
film between the pyrite and Mo layer. This indicates the migration of so- ances in terms of the compactness of the layers, but the film annealed
dium from the glass through the Mo-layer when the temperature is in H2S showed multiple phases like marcasite and troilite instead of py-
maintained for a prolonged period [36,41]. Such a high content of Na rite (Fig. 7). The presence of marcasite phase was also confirmed by the
was not detected in 2 days annealed film. TEM electron beam (e-beam) diffraction image as shown in Fig. 12.
On the other hand, TEM e-beam diffraction pattern obtained for 4-
3.3.2. S:Fe ratio with thin film depth days TBDS annealed film confirmed pyrite structure as shown in
The XPS depth profiling was also performed on the H2S annealed Fig. 13a and b. The pyrite phase of the films on Mo-coated glass sulfu-
film on both bare glass and Mo-coated glass substrates. The S:Fe stoichi- rized in TBDS sulfur source was also confirmed by XRD measurement
ometry was monitored to observe the impact of the TBDS and H2S (Fig. 7b).

Fig. 13. TEM e-beam diffraction of TBDS grown film on Mo-coated glass substrate indicating (a) [103] and (b) [132] zone axis of pyrite and their corresponding Bragg's reflections.
278 S.P. Adusumilli et al. / Thin Solid Films 615 (2016) 271–280

Fig. 14. Graphs indicating the optical characteristics of iron sulfide films grown at various time periods in a) TBDS and b) H2S atmosphere. The insets show plots for indirect band gaps and
the transmittance.

3.6. Optical properties decreased from 1.27 eV to 1.18 eV with the increase in annealing time
period from 6 h to 48 h. In both cases, the initial presence of oxides
The optical properties of TBDS and H2S growth models are shown in led to high bandgap values but with longer sulfur annealing, the entire
Fig. 14. For TBDS grown films shown in Fig. 14a, the direct bandgap of film converted to pyrite and the bandgaps (1.19 eV for TBDS and
the thin films, as obtained from the Tauc plot [43] decreased from 1.18 eV for H2S) corresponded to the pyrite phase [42]. As shown in
1.46 eV to 1.19 eV as the annealing time period increased from 6 h to the Fig. 14b inset, the indirect band gaps obtained for H2S annealed
4 days (96 h). The band gap of 1.19 eV for 96 h annealed film matches films were slightly lower (0.89 eV for 6 h, 0.79 for 24 h, and 0.78 eV
with the direct band gap of pyrite confirming the complete conversion for 48 h) than those for the TBDS films.
of Fe2O3 precursor film into the pyrite phase [9,42]. The indirect band
gaps were assessed by (αhυ)1/2 vs. hυ plot. The indirect band gaps for 3.7. Optical images of the grown pyrite films
6 h, 24 h, 48 h, and 4 days annealed films were 0.9 eV, 0.89 eV,
0.87 eV, and 0.85 eV respectively (see inset of Fig. 14a). These values The films grown on a rigid Corning® EAGLE XG® glass and a flexible
are in the same range as the pyrite indirect bandgaps reported earlier Corning® Willow® Glass substrates can be seen in Fig. 15. Both films
[3,42]. As shown in Fig. 14b, the bandgap of the films grown with H2S show characteristic gold-like luster of pyrite crystals. As most of the
 S.P. Adusumilli et al. / Thin Solid Films 615 (2016) 271–280 279

Fig. 15. (a) Pyrite grown on Corning® EAGLE XG® glass. (b) Pyrite thin film grown on flexible Corning® Willow® Glass.

Table 1
Electrical characterization of thin films grown using TBDS and H2S setups.

Sample Hall mobility (cm2/V-sec) Sheet resistance (Ω/sq) Career concentration (cm−3) Conductivity (S/cm)

TBDS_2 days 0.09 10,053.16 2.13 × 1020 3.06

TBDS_4 days 0.27 18,530.87 3.89 × 1019 1.68
H2S_2 days 0.05 21,366.64 1.67 × 1020 1.44

solar industry is moving towards flexible substrates, we wanted to ob- References

serve the compatibility of our pyrite growth on 100 μm thick flexible
[1] C. Wadia, A.P. Alivisatos, D.M. Kammen, Materials availability expands the opportu-
Corning® Willow® Glass. As shown in Fig. 15b, we confirmed the viabil- nity for large-scale photovoltaics deployment, Environ. Sci. Technol. 43 (2009)
ity of the pyrite growth process on the flexible glass substrate as well. 2072–2077, http://dx.doi.org/10.1021/es8019534.
[2] C. Wadia, Y. Wu, S. Gul, S.K. Volkman, J. Guo, A.P. Alivisatos, Surfactant-assisted hy-
drothermal synthesis of single phase pyrite FeS2 nanocrystals, Chem. Mater. 21
3.8. Electrical properties (2009) 2568–2570, http://dx.doi.org/10.1021/cm901273v.
[3] G. Smestad, A. Ennaoui, S. Fiechter, H. Tributsch, W.K. Hofmann, M. Birkholz, W.
The electrical properties of the grown pyrite samples were assessed Kautel, Photoactive thin film semiconducting iron pyrite prepared by sulfurization
of iron oxides, Sol. Energy Mater. 20 (1990) 149–165, http://dx.doi.org/10.1016/
by using Hall Effect measurement technique (Table 1). Three samples, 0165-1633(90)90001-H.
two TBDS and one H2S grown, were used for this measurement. All [4] A. Ennaoui, S. Fiechter, C. Pettenkofer, N. Alonso-Vante, K. Büker, M. Bronold, C.
these three samples were grown on bare glass substrate and all were Höpfner, H. Tributsch, Iron disulfide for solar energy conversion, Sol. Energy
Mater. Sol. Cells 29 (1993) 289–370, http://dx.doi.org/10.1016/0927-
pyrite phase confirmed by the XRD measurements. Our measurements
0248(93)90095-K.
showed that all the samples were p-type as expected. [5] S. Fiechter, J. Mai, A. Ennaoui, W. Szacki, Chemical vapour transport of pyrite (FeS2)
Our champion sample (TBDS_4 days) had the highest mobility value with halogen (Cl, Br, I), J. Cryst. Growth 78 (1986) 438–444, http://dx.doi.org/10.
of 0.27 cm2/V-sec compared to 0.05 cm2/V-sec for H2S_2 days and 1016/0022-0248(86)90144-2.
[6] S. Fiechter, M. Birkholz, A. Hartmann, P. Dulski, M. Giersig, H. Tributsch, R.J.D. Tilley,
0.08 cm2/V-sec for TBDS_2 days films. The carrier densities were in The microstructure and stoichiometry of pyrite FeS2–x, J. Mater. Res. 7 (1992)
the order of 1019 cm−3 to 1020 cm−3. 1829–1838, http://dx.doi.org/10.1557/JMR.1992.1829.
[7] M. Cabáan-Acevedo, N.S. Kaiser, C.R. English, D. Liang, B.J. Thompson, H.E. Chen, K.J.
Czech, J.C. Wright, R.J. Hamers, S. Jin, Ionization of high-density deep donor defect
4. Conclusions states explains the low photovoltage of iron pyrite single crystals, J. Am. Chem.
Soc. 136 (2014) 17163–17179, http://dx.doi.org/10.1021/ja509142w.
To conclude, we have synthesized high quality iron pyrite thin films [8] A. Ennaoui, S. Fiechter, G. Smestad, H. Tributsch, Preparation of iron disulfide and its
use for solar energy conversion, World Renew. Energy Congr. 1 (1990) 458–464.
in a one-step sulfur annealing process using a low cost organic sulfur [9] Y. Bi, Y. Yuan, C.L. Exstrom, S.A. Darveau, J. Huang, Air stable, photosensitive, phase
source, TBDS. Pyrite phase purity has been obtained with stoichiometric pure iron pyrite nanocrystal thin films for photovoltaic application, Nano Lett. 11
stability by having Se:F atomic percentage ratio maintained at 2.0. Com- (2011) 4953–4957, http://dx.doi.org/10.1021/nl202902z.
[10] G. Willeke, R. Dasbach, B. Sailer, E. Bucher, Thin pyrite (FeS2) films prepared by
paring the TBDS with H2S, TBDS is the better choice for sulfurization be- magnetron sputtering, Thin Solid Films 213 (1992) 271–276, http://dx.doi.org/10.
cause with TBDS the presence of sodium is not necessary for pyrite 1016/0040-6090(92)90293-K.
formation. H2S, while more reactive, does not yield pure pyrite phase [11] A. Yamamoto, M. Nakamura, A. Seki, E.L. Li, A. Hashimoto, S. Nakamura, Pyrite
(FeS2) thin films prepared by spray method using FeSO4 and (NH4)2Sx, Sol. Energy
when grown on metal-coated substrates. Importantly, for the pyrite
Mater. Sol. Cells 75 (2003) 451–456, http://dx.doi.org/10.1016/S0927-
phase formation, H2S is dependent on sodium presence whereas TBDS 0248(02)00205-2.
is independent of the availability of sodium. The optical bandgap mea- [12] G. Smestad, A. Da Silva, H. Tributsch, S. Fiechter, M. Kunst, N. Meziani, M. Birkholz,
surement confirmed pyrite direct band gap values; 1.19 eV for TBDS Formation of semiconducting iron pyrite by spray pyrolysis, Sol. Energy Mater. 18
(1989) 299–313, http://dx.doi.org/10.1016/0165-1633(89)90044-0.
and 1.18 eV H2S annealed films. [13] A.K. Raturi, L. Ndjeli, K. Rabah, FeS2 thin films prepared by spray pyrolysis, Renew.
Energy 11 (1997) 191–195, http://dx.doi.org/10.1016/S0960-1481(96)00123-1.
Acknowledgements [14] G. Chatzitheodorou, S. Fiechter, R. Könenkamp, M. Kunst, W. Jaegermann, H.
Tributsch, Thin photoactive FeS2 (pyrite) films, Mater. Res. Bull. 21 (1986)
1481–1487, http://dx.doi.org/10.1016/0025-5408(86)90088-7.
This study was supported by the Office of Naval Research; Grant No. [15] J. Oertel, K. Ellmer, W. Bohne, J. Röhrich, H. Tributsch, Growth of n-type polycrystal-
N00014-11-1-0658. line pyrite (FeS2) films by metalorganic chemical vapour deposition and their
280 S.P. Adusumilli et al. / Thin Solid Films 615 (2016) 271–280

electrical characterization, J. Cryst. Growth 198-199 (1999) 1205–1210, http://dx. [29] R. Morrish, R. Silverstein, C.a. Wolden, Synthesis of stoichiometric FeS2 through
doi.org/10.1016/S0022-0248(98)01074-4. plasma-assisted sulfurization of Fe2O3 nanorods, J. Am. Chem. Soc. 134 (2012)
[16] L. Samad, M. Cabán-Acevedo, M.J. Shearer, K. Park, R.J. Hamers, S. Jin, Direct 17854–17857, http://dx.doi.org/10.1021/ja307412e.
chemical vapor deposition synthesis of phase-pure iron pyrite (FeS2) thin [30] N. Hamdadou, A. Khelil, J.C. Bernède, Pyrite FeS2 films obtained by sulphuration of
films, Chem. Mater. 27 (2015) 3108–3114, http://dx.doi.org/10.1021/acs. iron pre-deposited films, Mater. Chem. Phys. 78 (2003) 591–601, http://dx.doi.
chemmater.5b00664. org/10.1016/S0254-0584(01)00577-6.
[17] S. Nakamura, A. Yamamoto, Electrodeposition of pyrite(FeS2) thin films for photo- [31] X. Zhang, M. Manno, A. Baruth, M. Johnson, E.S. Aydil, C. Leighton, Crossover from
voltaic cells, Sol. Energy Mater. Sol. Cells 65 (2001) 79–85, http://dx.doi.org/10. nanoscopic intergranular hopping to conventional charge transport in pyrite thin
1016/S0927-0248(00)00080-5. films, ACS Nano 7 (2013) 2781–2789, http://dx.doi.org/10.1021/nn4003264.
[18] S. Seefeld, M. Limpinsel, Y. Liu, N. Farhi, A. Weber, Y. Zhang, N. Berry, Y.J. Kwon, C.L. [32] J. Hu, Y. Zhang, M. Law, R. Wu, Increasing the band gap of iron pyrite by alloying
Perkins, J.C. Hemminger, R. Wu, M. Law, Iron pyrite thin films synthesized from an with oxygen, J. Am. Chem. Soc. 134 (2012) 13216–13219, http://dx.doi.org/10.
Fe(acac)3 ink, J. Am. Chem. Soc. 135 (2013) 4412–4424, http://dx.doi.org/10.1021/ 1021/ja3053464.
ja311974n. [33] B. Thomas, K. Ellmer, M. Müller, C. Höpfner, S. Fiechter, H. Tributsch, Structural and
[19] J. Joo, H. Bin Na, T. Yu, J.H. Yu, Y.W. Kim, F. Wu, J.Z. Zhang, T. Hyeon, Generalized and photoelectrical properties of FeS2 (pyrite) thin films grown by MOCVD, J. Cryst.
facile synthesis of semiconducting metal sulfide nanocrystals, J. Am. Chem. Soc. 125 Growth 170 (1997) 808–812, http://dx.doi.org/10.1016/S0022-0248(96)00600-8.
(2003) 11100–11105, http://dx.doi.org/10.1021/ja0357902. [34] C. Steinhagen, T.B. Harvey, C.J. Stolle, J. Harris, B.A. Korgel, Pyrite nanocrystal solar
[20] Y.-Y. Lin, D.-Y. Wang, H.-C. Yen, H.-L. Chen, C.-C. Chen, C.-M. Chen, C.-Y. Tang, C.-W. cells: promising, or fool's gold? J. Phys. Chem. Lett. 3 (2012) 2352–2356, http://dx.
Chen, Extended red light harvesting in a poly(3-hexylthiophene)/iron disulfide doi.org/10.1021/jz301023c.
nanocrystal hybrid solar cell, Nanotechnology 20 (2009) 405207, http://dx.doi. [35] L. Reijnen, B. Meester, A. Goossens, J. Schoonman, In situ mass spectrometric study
org/10.1088/0957-4484/20/40/405207. of pyrite (FeS[sub 2]) thin film deposition with Metallorganic chemical vapor depo-
[21] J. Puthussery, S. Seefeld, N. Berry, M. Gibbs, M. Law, Colloidal iron pyrite (FeS2) sition, J. Electrochem. Soc. 147 (2000) 1803, http://dx.doi.org/10.1149/1.1393437.
nanocrystal inks for thin-film photovoltaics, J. Am. Chem. Soc. 133 (2011) [36] N. Berry, M. Cheng, C.L. Perkins, M. Limpinsel, J.C. Hemminger, M. Law, Atmospheric-
716–719, http://dx.doi.org/10.1021/ja1096368. Pressure Chemical Vapor Deposition of Iron Pyrite Thin Films, 2012 1–12, http://dx.
[22] R. Wu, Y.F. Zheng, X.G. Zhang, Y.F. Sun, J.B. Xu, J.K. Jian, Hydrothermal synthesis and doi.org/10.1002/aenm.201200043.
crystal structure of pyrite, 266 (2004) 523–527, http://dx.doi.org/10.1016/j. [37] G. Kullerud, Phase relations in the Fe\ \SO system, Carnegie Inst. Wash. Yearb., 56,
jcrysgro.2004.02.020. 1957, pp. 198–200.
[23] C. de las Heras, I.J. Ferrer, C. Sanchez, Temperature dependence of the optical ab- [38] S. Garner, S. Glaesemann, X. Li, Ultra-slim flexible glass for roll-to-roll electronic de-
sorption edge of pyrite FeS2 thin films, J. Phys. Condens. Matter 6 (10177) (1994), vice fabrication, Appl. Phys. A Mater. Sci. Process. 116 (2014) 403–407, http://dx.doi.
http://dx.doi.org/10.1088/0953-8984/6/46/033. org/10.1007/s00339-014-8468-2.
[24] C. De las Heras, I.J. Ferrer, C. Sánchez, Comparison of pyrite thin films obtained from [39] S.P. Adusumilli, T.P. Dhakal, C.R. Westgate, Synthesis of Iron Pyrite Film Through
Fe and natural pyrite powder, Appl. Surf. Sci. 50 (1991) 505–509, http://dx.doi.org/ Low Temperature Atmospheric Pressure Chemical Vapor Deposition, 1447, MRS On-
10.1016/0169-4332(91)90227-B. line Proc. Libr., 2012, http://dx.doi.org/10.1557/opl.2012.1461 null–null.
[25] B. Thomas, C. Höpfner, K. Ellmer, S. Fiechter, H. Tributsch, Growth of FeS2 (pyrite) [40] T.P. Dhakal, S. Harvey, M. van Hest, G. Teeter, Back Contact Band Offset Study of Mo-
thin films on single crystalline substrates by low pressure metalorganic chemical va- CZTS Based Solar Cell Structure by Using XPS/UPS Techniques, Photovolt. Spec. Conf.
pour deposition, J. Cryst. Growth 146 (1995) 630–635, http://dx.doi.org/10.1016/ (PVSC), 2015 IEEE 42nd, 2015 1–4, http://dx.doi.org/10.1109/PVSC.2015.7355623.
0022-0248(94)00528-1. [41] D. Rudmann, D. Brémaud, H. Zogg, a.N. Tiwari, Na incorporation into
[26] C. Höpfner, K. Ellmer, A. Ennaoui, C. Pettenkofer, S. Fiechter, H. Tributsch, Stoichiom- Cu(In,Ga)Se[sub 2] for high-efficiency flexible solar cells on polymer foils, J. Appl.
etry-, phase- and orientation-controlled growth of polycrystalline pyrite (FeS2) thin Phys. 97 (2005) 084903, http://dx.doi.org/10.1063/1.1857059.
films by MOCVD, J. Cryst. Growth 151 (1995) 325–334, http://dx.doi.org/10.1016/ [42] Y. Li, Z. Han, L. Jiang, Z. Su, F. Liu, Y. Lai, Y. Liu, Template-directed synthesis of ordered
0022-0248(95)00066-6. iron pyrite (FeS2) nanowires and nanotubes arrays, J. Sol-Gel Sci. Technol. 72 (2014)
[27] D.M. Schleich, H.S.W. Chang, Iron pyrite and iron marcasite thin films prepared by 100–105, http://dx.doi.org/10.1007/s10971-014-3425-2.
low pressure chemical vapor deposition, J. Cryst. Growth 112 (1991) 737–744, [43] T. Dhakal, D. Vanhart, R. Christian, A. Nandur, A. Sharma, C.R. Westgate, Growth
http://dx.doi.org/10.1016/0022-0248(91)90130-W. morphology and electrical/optical properties of Al-doped ZnO thin films grown by
[28] R.A. Berner, Iron sulfides formed from aqueous solution at low temperatures and at- atomic layer deposition, J. Vac. Sci. Technol. A 30 (2012), http://dx.doi.org/10.
mospheric pressure, J. Geol. 72 (1964) 293–306. 1116/1.3687939.