Investigation of the Si/TiO
2
/Electrolyte Interface using
Operando
Tender X-ray Photoelectron Spectroscopy
M. F. Lichterman
a,b
*, M. H. Richter
b
*, S. Hu
a,b
*, E. J. Crumlin
c
*, S. Axnanda
c
,
M. Favaro
b,c
, W. Drisdell
b,c
, Z. Hussain
c
, T. Mayer
d
, B. S. Brunschwig
b,e
, N. S. Lewis
a,b,f
,
Z. Liu
c,g,h
, and H.-J. Lewerenz
b‡
a
Division of Chemistry and Chemical Engineering, California Institute of Technology,
Pasadena, CA 91125, USA.
b
Joint Center for Artificial Photosynthesis, California Institute of Technology, Pasadena,
CA 91125, USA.
c
Advanced Light Source, Lawrence Berkeley National Laboratory, Berkeley, CA 94720
d
Surface Science Division, Materials Science Department, Darmstadt University of
Technology, 64287 Darmstadt, Germany.
e
Beckman Institute, California Institute of Technology, Pasadena, CA 91125, USA
f
Kavli Nanoscience Institute, California Institute of Technology, Pasadena, CA 91125,
USA.
g
State Key Laboratory of Functional Materials for Informatics, Shanghai Institute of
Microsystem and Information Technology, Chinese Academy of Sciences, Shanghai
200050, China.
h
School of Physical Science and Technology, ShanghaiTech University, Shanghai
200031, China.
‡
Corresponding author E-mail: lewerenz@caltech.edu
Semiconductor-electrolyte interfaces allow for the creation of
photoactive semiconductor systems that have band bending and
other characteristics analogous to semiconductor-metal junctions
(Schottky junctions). We demonstrate herein that XPS
measurements can be obtained on a full three-electrode
electrochemical system under potentiostatic control by use of
tender X-rays to provide photoelectrons with sufficient kinetic
energy to penetrate through a thin electrolyte overlayer on a
portion of the working electrode. The response of the
photoelectron binding energies to variations in applied voltage
demonstrates that the XPS investigation works in an
operando
manner to elucidate the energetics of such interfaces.
Introduction
Photoelectrochemical cells have garnered much attention for potential applications in
solar-energy conversion and storage (1). Numerous experimental techniques have been
deployed to further the understanding of the semiconductor/liquid junctions that are at the
heart of such systems (2-11). Semiconductor-metal (Schottky) junctions have been
thoroughly analyzed by surface analyses and optoelectronic methods, whereas the
energetic relations at semiconductor/liquid junctions remain elusive. The inherent
difficulty results from the requirement that a conductive electrolyte must be present on
10.1149/06606.0097ecst ©The Electrochemical Society
ECS Transactions, 66 (6) 97-103 (2015)
97
) unless CC License in place (see abstract).
ecsdl.org/site/terms_use
address. Redistribution subject to ECS terms of use (see
131.215.70.231
Downloaded on 2016-11-21 to IP
the surface of the semiconductor, precluding the use of conventional surface-science
methods to probe the semiconductor/liquid interface of an operating photoelectrode.
Ambient-Pressure X-ray Photoelectron Spectroscopy (AP-XPS) has been used to
investigate gas-solid and gas-liquid interfaces as well as related processes such as oxygen
reduction on fuel cell cathodes (12), mechanisms of heterogeneous catalysis (13) and
formation of metal oxide overlayers (14). Herein, we describe the use of AP-XPS, with
real-time control
of the electrochemical potential, through a conductive aqueous
electrolyte on a semiconductor/liquid junction. A ~13 nm thick hanging meniscus on the
(semiconductor) working electrode in a three-electrode setup permitted electrochemical
control of the solid/liquid interface. Differential pumping allowed maintaining the XPS
analysis chamber at ~ 15 Torr while the analyzer was kept at a much lower pressure (15).
TiO
2
has recently been reported to provide a protective layer that permits the use of
small band-gap semiconductors (such as Si, GaAs, CdTe, etc.) as photoanodes with
extended stability relative to unprotected surfaces of such semiconductors. In this
application, the photogenerated holes transport through the TiO
2
during the
photoelectrochemical oxidation of H
2
O to O
2
(16, 17). Herein we describe the use of AP-
XPS to further characterize these TiO
2
/electrolyte and TiO
2
/Ni/electrolyte interfaces.
Experimental
Films of TiO
2
deposited by atomic-layer deposition (16,17), were prepared on
degenerately doped p-type silicon (“p
+
-Si”) substrates. Si (100) wafers with a resistivity,
ρ
< .005
*cm, were first cleaned with an oxidizing etch by so
aking in a 3:1 (v:v)
“piranha” solution of concentrated H
2
SO
4
(98%) to 30% H
2
O
2
for 2 min, and then
etching for 10 s in a 10% (by volume) solution of HF(aq). The wafers were then etched in
a 5:1:1 (by volume) solution of H
2
O, 36% hydrochloric acid, and 30% hydrogen peroxide
for 10 min at 75 °C. The TiO
2
was deposited from a tetrakis(dimethylamido)titanium
(TDMAT) precursor. A 0.1 s pulse of TDMAT was followed by 15 s purge of N
2
at 20
sccm, following by a 0.015 s pulse of H
2
O before another 15 s purge with N
2
. This
process was repeated for 1500 cycles to provide films ~ 70 nm in thickness. Where
desired, Ni was deposited at a RF sputtering power of 150 W for 80 s. Atomic-force
microscopy (AFM) images attested to the smoothness of these films on the nanometer
scale.
To prepare electrodes for
operando
AP-XPS, strips of the p
+
-Si/TiO
2
/(Ni) wafers
were cut into 1 cm x 3.5 cm rectangles. Highly doped p-type silicon (p
+
) was used
simultaneously as a support material as well as to provide an effective back contact to the
ALD-TiO
2
(Figure 1). The ohmic contact at the back of the semiconductor was connected
to the photoelectron analyzer to provide a high conductivity ground for the sample.
To
make ohmic contact, an In/Ga eutectic was scribed into the back of the Si wafer, and Ag
paint was then used to contact the electrode to a strip of Cu tape that was supported on a
0.8 cm x 3 cm glass slide.
ECS Transactions, 66 (6) 97-103 (2015)
98
) unless CC License in place (see abstract).
ecsdl.org/site/terms_use
address. Redistribution subject to ECS terms of use (see
131.215.70.231
Downloaded on 2016-11-21 to IP
Figure 1. Schematic of the layer stack investigated by
operando
XPS. The thicknesses of
the various layers are not drawn to scale (see text). On bare electrodes, no nickel was
present between the ALD TiO
2
and the electrolyte.
To perform the
operando
AP-XPS measurements, the p
+
-Si/TiO
2
/(Ni) electrode was
mounted and contacted by a three-axis manipulator onto which the Ag/AgCl reference
electrode (eDAQ) and a platinum foil counter electrode were also mounted and separately
contacted. To collect data, the sample was “dipped” into the electrolyte before a potential
was set using the potentiostat. The electrode was then raised to location that produced a
thin film of the electrolyte (~13 nm) at the XPS sampling location, as determined by
IMFP calculations based on the intensity of the signal observed on the detector relative to
the X-ray intensity incident onto the system.
AP-XPS data were collected on a Scienta R4000 HiPP-2 system in which the
photoelectron collection cone was aligned to the beamline X-Ray spot at a distance of
~300 μ m. Beamline 9.3.1 at the Advanced Light Source was used to provide Tender
X-Rays of 4 keV photon energy out of a possible photon energy range of 2.3-5.2 keV,
with a resolving power of E/
E = 3000-7200. Photoelectrons for the Ti 2p level we
re
collected in kinetic energy ranges of 3538
±
8 eV. Multiple scans [10 – 15] were required
t
o achieve a satisfactory signal-to-noise ratio.
Results and Discussion
Mott-Schottky measurements of the inverse square of the differential capacitance vs
potential showed that the p
+
-Si/TiO
2
system adopted a flat-band position at a potential of
~ -0.7 V vs the reversible hydrogen electrode, RHE, in 1.0 M KOH(aq) (Figure 2). Other
measurements have indicated that an ohmic contact was formed between p
+
-Si and the
ALD-TiO
2
used here (16). As will be demonstrated below, a stack consisting of p
+
-
Si/TiO
2
/Ni, when cycled electrochemically in a fast redox couple solution, displayed
ohmic behavior, indicating effective ohmic metallurgical contacts throughout the
structure.
The measured XPS data and the related energy-band alignments can be influenced by
surface roughness effects, particularly if the semiconductor is highly doped. The
electrolyte makes a conformal contact to even a very rough semiconductor surface, so a
larger volume fraction of the semiconductor surface can contain a space-charge region on
a roughened semiconductor surface than on a flat surface. For a semiconductor with a
space-charge region width on the order of nanometers, this behavior may influence the
position and width of an XPS peak. Moreover, the band energetics directly at the surface
Ni
ALD TiO
2
p
+
-Si
Ohmic contact
SiTiO
x
Electrolyte
ECS Transactions, 66 (6) 97-103 (2015)
99
) unless CC License in place (see abstract).
ecsdl.org/site/terms_use
address. Redistribution subject to ECS terms of use (see
131.215.70.231
Downloaded on 2016-11-21 to IP
will be overestimated in an XPS spectrum for semiconductors with surface roughness on
the same scale as the depth of the space-charge region, relative to the situation for a flat
surface. Specifically, in situations where surface roughness allows a much larger fraction
of the semiconductor volume to be described as in the space-charge region, the effect can
be large, because of the binding energy change
E
B
(x) near the surface, where x~0
(equation 1), will dominate:
Δ
|
= −
−
∙
∙ −
[1]
H
ere
U
0
is the applied potential to the back of the semiconductor,
d
is the thickness of the
semiconductor space-charge region, and
x
is the distance to the surface.
However, for surface roughness on a length scale that is substantially smaller than the
extension of the semiconductor space-charge region, or with a space-charge region that is
considerably larger than the penetration depth of the technique, this effect is essentially
negligible.
Figure 2. 3D AFM images of (a) bare silicon, (b) ~70 nm thick TiO
2
on silicon, and (c)
the same as (b) but with ~2 nm Ni.
The TiO
2
used in this study had a space-charge region on the order of tens of nanometers
at 1 V of applied bias, as calculated from the dopant densities determined by the Mott-
Schottky analysis. Thus, sample roughness effects on the collected data should be
minimal in the absence of a macroscopically rough surface. As shown in Figure 2, the
roughness of the TiO
2
samples is on the order of ~1 nm across tens of nanometers, hence
surface topography should not significantly influence the results observed herein by AP-
XPS.
(a)
(b)
(c)
ECS Transactions, 66 (6) 97-103 (2015)
100
) unless CC License in place (see abstract).
ecsdl.org/site/terms_use
address. Redistribution subject to ECS terms of use (see
131.215.70.231
Downloaded on 2016-11-21 to IP
Figure 3. Dark
J-E
measurements for p
+
-Si/TiO
2
(red) and p
+
-Si/TiO
2
/Ni (black)
electrodes in (a) 1M KOH and (b) a solution containing 50mM potassium ferricyanide
and 350mM potassium ferrocyanide. The noisy cathodic and anodic currents at ~ -20
mA
cm
-2
and 100 mA
cm
-2
are limited by ferri/ferrocyanide concentration and transport.
Electrochemical experiments were undertaken to examine the nature of the
TiO
2
/electrolyte and TiO
2
/Ni/electrolyte junction before
operando
measurements were
performed. As demonstrated in Figure 3, electrodes consisting of p
+
-Si/TiO
2
in an
aqueous ferri/ferrocyanide solution showed anodically rectifying behavior for the
kinetically slow process of water oxidation as well as for the kinetically rapid process of
ferrocyanide oxidation. However, with Ni present on the electrode surface, both
processes exhibited little charge-transfer resistance; while the enhanced water oxidation
current density might be argued to originate from enhanced catalysis facilitated by the
presence of Ni, the change in the observed electrochemical behavior in contact with the
ferri/ferrocyanide redox system strongly indicates that the addition of Ni drastically
improved the ohmic conductivity of TiO
2
/electrolyte interface and allowed for efficient
hole-charge transport.
(a)
(b)
ECS Transactions, 66 (6) 97-103 (2015)
101
) unless CC License in place (see abstract).
ecsdl.org/site/terms_use
address. Redistribution subject to ECS terms of use (see
131.215.70.231
Downloaded on 2016-11-21 to IP
Figure 4. (a) Ti 2p XPS spectra of the bare TiO
2
electrode in 1.0 M KOH(aq) at -0.2 V
and +0.6 V vs Ag/AgCl. The inset shows the change in band bending as the potential was
made more positive. (b) Ti 2p XPS spectra of a TiO
2
/Ni electrode under the same
conditions. The inset shows the manner in which the TiO
2
band bending was invariant
with respect to the applied potential.
As shown in Figure 4a, for a bare TiO
2
surface, the observed binding energy shifted
in accord with a change in the applied potential. Thus, this system followed the ideal
semiconductor/liquid junction behavior in this potential range, as described in the inset.
Specifically, as the Fermi level of the semiconductor shifted to a more positive potential,
the fixed positions of the band edges in the semiconductor/liquid junction caused the
observed binding energies to decrease. In contrast, the metallized TiO
2
film (Figure 4b)
exhibited very different behavior. Specifically, effectively little or no shift was observed
in the TiO
2
core-level emission peak positions as the applied potential was varied. This
behavior is consistent with the observation that metallization (using Ni) of the TiO
2
surface allows for charge-carrier transport (16). These results therefore demonstrate that
the Ni on the surface pins the semiconductor energetics to that of the metal instead of the
electrolyte, and plays a key role in allowing for charge conduction to take place.
Effectively, metallization allows the semiconductor band edges to shift with potential
across a wide potential range.
Conclusion
Operando
XPS technique can be used to determine the nature of band energetics in a
semiconductor-liquid junction. The nature of the junction formed can be tuned by
appropriate modification of the semiconductor surface, which has crucial effects on the
nature of space-charge region formation or charge conduction therein. Further
experiments are underway to elaborate in more detail the diverse influences of
semiconductor doping, electrolyte choice, and other relevant variables on the system.
(a)
(b)
E
F
CBM
VBM
+U
E
F
CBM
VBM
+U
ECS Transactions, 66 (6) 97-103 (2015)
102
) unless CC License in place (see abstract).
ecsdl.org/site/terms_use
address. Redistribution subject to ECS terms of use (see
131.215.70.231
Downloaded on 2016-11-21 to IP
Acknowledgments
This work was supported through the Office of Science of the U.S. Department of
Energy (DOE) under award no. DE-SC0004993 to the Joint Center for Artificial
Photosynthesis, a DOE Energy Innovation Hub. The Advanced Light Source is supported
by the Director, Office of Science, Office of Basic Energy Sciences, of the U.S.
Department of Energy under Contract No. DE-AC02-05CH11231. We thank Dr. Philip
Ross for contributions to the conceptual development of the AP-XPS end-station and
experimental design.
References
1. H. Gerischer,
J. Electroanal. Chem. and Interfacial Electrochem.
,
58
, 263 (1975).
2. A. J. Bard,
J. Phys. Chem.
,
86
, 172 (1982).
3. A. J. Bard, A. B. Bocarsly, F. R. F. Fan, E. G. Walton, and M. S. Wrighton,
J. Am.
Chem. Soc.
,
102
, 3671 (1980).
4. A. Heller, K. C. Chang, and B. Miller,
J. Electrochem. Soc.
,
124
, 697 (1977).
5. A. Heller, B. Miller, and F. A. Thiel,
Appl. Phys. Lett.
,
38
, 282 (1981).
6. N. S. Lewis,
J. Electrochem. Soc.
,
131
, 2496 (1984).
7. B. Miller, A. Heller, S. Menezes, and H. J. Lewerenz, Faraday Discuss.,
70
, 223
(1980).
8. H. J. Lewerenz,
Chem. Soc. Rev.
,
26
, 239 (1997).
9. K. Jacobi, M. Gruyters, P. Geng, T. Bitzer, M. Aggour, S. Rauscher, and H. J.
Lewerenz,
Physical Review B
,
51
, 5437 (1995).
10. H. J. Lewerenz, M. Aggour, C. Murrell, M. Kanis, H. Jungblut, J. Jakubowicz, P.
A. Cox, S. A. Campbell, P. Hoffmann, and D. Schmeißer,
J. Electrochem. Soc.
,
150
, E185 (2003).
11. M. Letilly, K. Skorupska, and H.-J. Lewerenz,
J. Phys. Chem. C
,
117
, 16381
(2013).
12. Y.-C. Lu, E. J. Crumlin, G. M. Veith, J. R. Harding, E. Mutoro, L. Baggetto, N. J.
Dudney, Z. Liu, and Y. Shao-Horn,
Sci. Rep.
,
2
, 715 (2012).
13. V. V. Kaichev, I. P. Prosvirin, and V. I. Bukhtiyarov, J Struct Chem, 52, 90
(2011).
14. A. Y. Klyushin, T. C. R. Rocha, M. Havecker, A. Knop-Gericke, and R. Schlögl,
Phys. Chem. Chem. Phys.
,
16
, 7881 (2014).
15. S. Axnanda, E. J. Crumlin, B. Mao, S. Rani, R. Chang, W. C. Stolte, P. G.
Karlsson, M. O. M. Edwards, M. Lundqvist, R. Moberg, P. Ross, Z. Hussain, and
Z. Liu, Submitted (2015).
16. S. Hu, M. R. Shaner, J. A. Beardslee, M. F. Lichterman, B. S. Brunschwig, and N.
S. Lewis,
Science
,
344
, 1005 (2014).
17. M. F. Lichterman, A. I. Carim, M. T. McDowell, S. Hu, H. B. Gray, B. S.
Brunschwig, and N. S. Lewis,
Energy Environ. Sci.
,
7
, 3334 (2014).
ECS Transactions, 66 (6) 97-103 (2015)
103
) unless CC License in place (see abstract).
ecsdl.org/site/terms_use
address. Redistribution subject to ECS terms of use (see
131.215.70.231
Downloaded on 2016-11-21 to IP