of 15
1
Supplementary Information For:
Stable Solar-Driven Water Oxidation to O
2
(g) by Ni Oxide
Coated Silicon Photoanodes
Ke Sun
1,2
, Matthew T. McDowell
1,2
, Adam C. Nielander
1
, Shu Hu
1,2
, Matthew R. Shaner
1,2
, Fan
Yang
1,2
, Bruce S. Brunschwig
3
, Nathan Lewis*
1-4
1
Division of Chemistry and Chemical Engineering, Ca
lifornia Institute of Technology, Pasadena,
CA 91125, USA.
2
Joint Center for Artificial Photosynthesis, Califo
rnia Institute of Technology, Pasadena, CA
91125, USA.
3
Beckman Institute and Molecular Materials Research
Center, California Institute of Technology,
Pasadena, CA 91125, USA.
4
Kavli Nanoscience Institute, California Institute
of Technology, Pasadena, CA 91125, USA.
*Correspondence to: nslewis@caltech.edu
Experimental Procedures
Materials and Chemicals
All materials were used as received, except where o
therwise noted. The chemicals used were:
sulfuric acid (H
2
SO
4
, J. T. Baker, ACS reagent, 95%-98%), concentrated
hydrochloric acid (HCl,
Sigma Aldrich, ACS Reagent 37%), hydrogen peroxide
(H
2
O
2
, Macron Chemicals, ACS grade
2
30%), concentrated ammonium hydroxide (NH
4
OH, Sigma Aldrich, ACS reagent 28%-30%),
Buffered HF Improved (Transene Company Inc.), potas
sium hydroxide pellets (KOH, Macron
Chemicals, ACS 88%), potassium ferrocyanide trihydr
ate (K
4
Fe(CN)
6
· 3H
2
O, Acros, >99%),
potassium ferricyanide (K
3
Fe(CN)
6
, Fisher Chemicals, certified ACS 99.4%), iron(II)
sulfate
heptahydrate (FeSO
4
· 7H
2
O, ReagentPlus eagentPluslusFeSO, certified AC
3
, Sigma Aldrich,
reagent grade 97%), zinc powder (Zn, Sigma Aldrich,
dust <10

m,
98%), and potassium
chloride (KCl, Macron Chemicals, Granular ACS 99.6%
). Water with a resistivity of 18.2
M

cm was obtained from a Millipore deionized water sy
stem.
Preparation of Substrates and Fabrication of Emitte
r
An np
+
-Si sample coated by FTO glass, with the p
+
emitter layer fabricated using ion
implantation, was obtained from the Joint Center fo
r Artificial Photosynthesis at Lawrence
Berkeley National Laboratory. The detailed fabricat
ion of the emitter onto a single-side polished
Czochralski-grown n-type (P-doped) prime-grade, (10
0)-oriented Si wafer has been described
elsewhere.
1
The FTO glass was removed from these samples as d
escribed previously
2
. Briefly,
the FTO-coated np
+
-Si was immersed in a mixture of 0.5 M HCl, 0.4 M F
eSO
4
, and 0.1 M FeCl
3
and then the mixture was agitated while ~3.5 g Zn w
as slowly poured into the mixture. The
solution turned clear after ~2 min and the samples
were then removed and rinsed with deionized
water. To remove any surface metal contaminants, t
he samples were immersed for ~30 s in an
oxidizing solution that contained 1.0 M HCl and 0.5
M FeCl
3
, and the samples were then
removed from the solution and rinsed. Samples were
then further cleaned using an RCA clean, as
described below.
3
Three types of Si wafers, all with diameters of 3”,
were used: p
+
-Si(111) and p
+
-Si(100) (B-
doped with a resistivity,
ρ
< 0.002

cm, 380±25 μ m thick, prime grade, Addison Engineeri
ng,
Inc.); and n-Si(111) (P-doped with
ρ
= 0.1–1

cm, UniversityWafer, Inc.). The wafers, and the
np
+
-Si samples from which the FTO coating had been rem
oved, were cleaned using an RCA SC-1
procedure of soaking in a solution of NH
4
OH:H
2
O
2
:H
2
O (6:1:1, by volume) for 20 min at 70 °C.
The wafers were then rinsed, dried, and placed in B
uffer HF Improved for ~ 10 s. The samples
were then rinsed and etched using the RCA SC-2 proc
edure of soaking in a solution of
HCl:H
2
O
2
:H
2
O (6:1:1, by volume) for 20 min at 70 °C. Samples
were thoroughly rinsed with
deionized water and were dried using N
2
(g).
Deposition of NiO
x
and Ni Metal Coatings
Reactive RF sputtering was performed using an AJA h
igh-vacuum magnetron sputtering system
(AJA International Inc.) with a chamber having a ma
ximum base pressure of 8×10
-8
Torr. The
Ar/O
2
ratio was 20, while the Ar flow was held constant
at 20 sccm and the working pressure was
held at 5 mTorr. The substrates were held at 300 °C
. The RF power on the Ni target (Kurt Lesker,
2” diameter × 0.125” thickness, 99.95%) was maintai
ned at 150 W. The deposition rate was kept
low, in the range of 0.42-0.62 Å/s, and therefore t
he deposition time was normally ~20-30 min,
depending on the aging of the sputtering target. Th
e same conditions were used to deposit
metallic Ni, except that the stage was not heated a
nd O
2
(g) was not supplied. The deposition
times under these conditions were normally ~ 2-3 mi
n, depending on the aging of the sputtering
target.
Preparation of Electrodes
4
Ohmic contacts to the back sides of Si samples were
formed by scribing an In-Ga eutectic alloy
(Alfa Aesar, 99.99%
)
onto the unpolished surfaces. High-purity Ag pain
t (SPI supplies) was then
used to mechanically attach the ohmic contact to a
coiled, tin-plated Cu wire (McMaster-Carr)
which was then threaded through a glass tube (Corni
ng Incorporation, Pyrex tubing, 7740 glass).
The sample was then encapsulated and sealed to the
glass tube using a mixture of 2:1 grey epoxy
(Hysol 9460F) and white epoxy (Hysol 1C). The epoxy
was typically allowed to set under
ambient laboratory conditions for mechanically atta
ch the ohmic contused. A high-resolution
optical scanner (Epson Perfection V370 with a resol
ution of 2400 psi) was used to image the
exposed surface area of each electrode, and the geo
metric areas were determined by analyzing the
images using ImageJ software. All of the electrode
s used in this study were ~ 0.1 cm
2
in area
unless otherwise specified.
Electrochemical Measurements
A Mercury/Mercury oxide (Hg/HgO in 1.0 M KOH, CH In
struments, CH152) electrode was used
as the reference electrode, and a carbon cloth plac
ed within a fritted glass tube (gas dispersion
tube Pro-D, Aceglass, Inc.) was used as the counter
electrode for all electrochemical
measurements performed in 1.0 M KOH(aq) electrolyte
, including photoelectrochemical, spectral
response, and faradaic efficiency measurements. Th
e equilibrium potential for the oxygen
evolution reaction (OER) was determined to be 0.32
V versus the Hg/HgO reference based on the
measured pH of the solution (pH 14). A custom elec
trochemical cell with a flat glass (Pyrex)
bottom was used for all of the electrochemical meas
urements. During measurements, the
electrolyte was vigorously agitated with a magnetic
stir bar driven by a model-train motor
(Pittman) with a Railpower 1370 speed controller (M
odel Rectifier Corporation). The data
5
presented for electrochemical measurements in aqueo
us solutions do not include compensation
for the series resistance of the solution. ELH-typ
e (Sylvania/Osram) and ENH-type (EIKO)
tungsten-halogen lamps with a custom housing and po
wered by a transformer (Staco Energy
Products Co.) were used for long-term photoelectroc
hemical stability measurements. A Xenon
arc lamp (Newport 67005 and 69911) equipped with an
infrared filter (Newport 61945) and an
AM 1.5G filter (Newport 81094 and 71260) was used a
s the light source for
J
-
E
and spectral
response measurements. The illumination intensity
at the position of the working electrode in the
electrochemical cell was determined by placing a ca
librated Si photodiode (FDS100-Cal, Thor
Labs) into the cell at the same position occupied b
y the exposed area of a photoelectrode. To
illuminate bottom-facing photoelectrodes, a quartz
diffuser (Newport 15Diff-Vis) together with a
broad-band reflection mirror (Newport dielectric mi
rror, 10Q20PR-HR) was used to bend the
uniform light beam from the horizontal to the verti
cal direction.
Cyclic voltammetry as well as quantum efficiency me
asurements were performed using a
Biologic SP-200 potentiostat (Bio-Logic Science Ins
truments). Cyclic voltammetric data were
obtained at a constant scan rate of 40 mV s
-1
and the scan range varied depending on the
photovoltage of the sample. The quantum yield was
measured using the potentiostat with the
current output connected to a lock-in amplifier, an
d the incident light was chopped at a frequency
of 20 Hz.
Electrochemical Impedance Spectroscopy and Mott-Sch
ottky Analysis
Electrochemical impedance spectroscopy was used on
n-Si|NiO
x
, n
+
-Si|NiO
x
, and p
+
-Si|NiO
x
to
determine the barrier height of the n-Si|NiO
x
interface and the doping level of NiO
x
. The
measurement was conducted in solution of 50 mM K
3
Fe(CN)
6
, 350 mM K
4
Fe(CN)
6
, and 1.0 M
6
KCl in 200 mL of H
2
O. The reverse bias dependence of the area-normali
zed semiconductor
depletion-region capacitance is given by the Mott-S
chottky relation:
1


= 
2





− 





− 


where
A
is the device area,
N
d
is the donor impurity concentration in the semicon
ductor,

0
is the
vacuum permittivity, and

r
is the relative permittivity,
q
is electron charge,

b
is the
C
-
V
barrier
height,
V
n
is the voltage difference between the potential of
the Fermi level and the potential off
the conduction-band edge of the n-type semiconducto
r in the bulk,
k
b
is the Boltzmann constant,
T
is the temperature in K, and
V
app
is the difference between the applied potential an
d the redox
potential of the solution. A Mott-Schottky (M-S) p
lot of
C
-2
versus
V
app
should be linear with a
slope related to
N
d
and an x-intercept related to

b
. M-S plots obtained from the impedance data
for the devices were linear (R
2
>0.9999).
The doping level was calculated based on the equati
on:
=
2






1 

⁄ 



where (d(1/
C
2
)/d
V
app
) is the slope of the M-S plot. The solution was q
uiescent and was kept in
the dark during the measurements. The electrochemi
cal impedance data were fit to a model that
consisted of a parallel resistor and a capacitor wi
th a fixed constant-phase element at the Si|NiO
x
interface arranged electrically in series with anot
her resistor and capacitor in parallel at the
NiO
x
/electrolyte interface.
The dielectric constant of NiO
x
was determined from the measured refractive index
using the
equation:
= 

− 

7
where
ε
’ is the real part of the complex dielectric consta
nt and
n
is the refractive index and
k
is
the extinction coefficient. The values of
n
and
k
were determined using spectroscopic
ellipsometry (vide infra).
The average value of
N
d
extracted from the M-S slope (2×10
16
cm
-3
) of n-Si|NiO
x
was in excellent
agreement with the range specified by the manufactu
rer (5×10
16
~5×10
15
cm
-3
), confirming that
the measured capacitance was indeed the depletion-r
egion capacitance at the Si|NiO
x
interface.
The M-S slope for p
+
-Si|NiO
x
was expected to be ~ 2.4×10
13
– 2.4×10
12
F
-2
m
4
V
-1
, which was
consistent with the measured slopes. The average va
lue of
N
d
extracted for p
+
-Si|NiO
x
was ~ 10
19
cm
-3
.
Measurements of the Faradaic Efficiency of Oxygen E
volution
A Neofox fluorescence probe (Foxy probe, Ocean Opti
cs) was used to monitor the concentration
of oxygen throughout the experiment. The fluoresce
nce response was calibrated against the
standard concentration of oxygen in water (7700

g L
-1
or 2.4×10
-4
M) under a standard 20.9%
(by volume) oxygen atmosphere. The fluorescence pr
obe, the Hg/HgO/1.0 M KOH reference
electrode, a fritted Pt mesh counter electrode (Alf
a-Aesar, 100 mesh, 99.9% trace metal basis),
and the np
+
-Si|NiO
x
working electrode with a geometric surface area of
0.79 cm
2
were loaded into
an air-tight glass cell that had a volume of 43.6 m
L with no headspace, and that was equipped
with four ports and a side-facing quartz window. Th
e cell and the 1.0 M KOH electrolyte used in
the cell were purged with a stream of ultra-high pu
rity Ar(g) for ~2 h prior to the measurement.
The oxygen-concentration data were converted into m
icrograms of O
2
by first correcting for the
O
2
leak rate measured during the first 10 min of the
experiment, followed by calculating the mass
of O
2
in micrograms at each time point. The calculation
was performed by multiplying the
reported percentage of oxygen by the number of micr
ograms of O
2
dissolved in water at room
8
temperature under 1 atm of pressure, 7700

g L
-1
(assuming the value is the same as for 1.0 M
aqueous acidic solutions), and by the cell volume (
43.6 mL), and dividing by the concentration of
O
2
in air under 1 atm as reported by the florescence
probe (20.9%). To compare the charge
versus time data from the potentiostat with the amo
unt of oxygen generated versus time for a
system operating at 100% faradaic efficiency, the c
harge passed (in mA

h) was multiplied by 3.6
to convert the data into coulombs, and the result w
as then multiplied by 83 (the factor for
conversion of 1 C of electrons to 1

g of O
2
) to convert the value into micrograms of O
2
.
Therefore, both the cumulative oxygen generated and
charge passed during the measurement can
be shown in one plot with two comparable y-axes, wh
ere 0.32 mA

h of charge passed
corresponded to 100

g O
2
generated.
Scanning-Electron Microscopy and Atomic-Force Micro
scopy
Scanning-electron microscopy (SEM) imaging was perf
ormed using a Nova NanoSEM 450 (FEI)
with an accelerating voltage of 5 kV. Atomic-force
microscopy (AFM) images were collected
using a Bruker Dimension Icon microscope operating
in ScanAsyst mode and using Bruker
ScanAsyst-Air probes (silicon tip on a silicon nitr
ide cantilever, spring constant: 0.4 N m
−1
,
frequency: 50-90 kHz, Al back coating) for p
+
-Si|NiO
x
with a thickness of 35~160 nm. The scan
size was typically 1

m × 1

m. The images were analyzed using NanoScope softwa
re version
1.5. Flattening was performed to remove curvature
and slope from images.
UV-Visible Absorption Measurements
The optical absorption of the NiO
x
- or Ni metal-coated Si was determined by using an
integrating
sphere at normal incidence (Agilent Cary 5000 UV-Vi
s spectrometer). The absorption of NiO
x
-
coated Si was determined by subtracting the measure
d reflection and transmission from unity.
9
The optical transmission of NiO
x
and ultrathin Ni metal on quartz substrates were a
lso measured
using the integrating sphere and were normalized ag
ainst the transmission of bare quartz
substrates. Tauc plot was generated by plotting
h
ν
versus (
ah
ν
)
2
, where
α
is the absorption
coefficient determined based on the Beer-Lambert la
w. The band gap of the NiO
x
at the onset of
significant absorption, as well as interband absorp
tion, were determined by a linear extrapolation
of the optical absorption edge to the y axis of the
Tauc plot.
Spectroscopic Ellipsometry
The complex refractive index (
n
,
k
) data for the NiO
x
coatings were measured on p
+
-Si substrates
with a thickness of 35~160 nm and were determined u
sing spectroscopic ellipsometry by fitting
the data according to a general oscillator model th
at assumed that an intermediate 2 nm native
oxide (SiO
2
) formed during sputtering. Moreover, a rough surf
ace with a thickness of 5 nm was
considered based on the AFM measurements of roughne
ss, to include the air/NiO
x
mixture. The
measurement was conducted using a J.A. Woolam V-VAS
E system. The
n
and
k
values were
extracted from a normal fit, using a fixed film thi
ckness based on the calibrated cross-sectional
SEM measurement. The absorption coefficient can be
also calculated using equation:
α
= 4
π
κ
/
λ
.
κ
is the extinction coefficient, which is the imagi
nary part of the complex refractive index
obtained from fitting of the ellipsometric data. T
he position of the absorption peak in the visible
region was consistent with the absorption features
extracted from the Tauc plot using the UV-vis
measurement.
X-ray diffraction spectroscopy
XRD analysis was conducted using Bruker D8 Discover
system equipped with 2-dimentional
Vantec-500 detector. Cu-Ka radiation (1.54 Å) was g
enerated at a tube voltage of 1 kV and a tube
10
current of 50 mA. The incident beam was focused usi
ng a mono-capillary collimator. A laser
beam marks the focal spot on the specimen fixed on
a xyz stage. The scattered diffraction was
registered by a 2-dementional detector with an angu
lar resolution of the detector smaller than
0.04 degree, and enables the simultaneous detection
of the diffraction data in a 2theta range of 20
degree. The detected radiation was counted for 2000
sec to obtain an appropriate XRD profile.
Data were analyzed using Bruker EVA software.
Load-line analysis and definition of photovoltaic p
roperties
The details of a load-line analysis can be found in
references 3,4. Briefly, the photoelectrode
performance is treated in a simplified equivalent c
ircuit approach as a photovoltaic cell connected
in series electrically with a dark electrolysis cel
l. The photovoltaic parameters including the open-
circuit voltage (
V
oc
), the short-circuit current density (
J
sc
), the fill factor (FF) and the energy-
conversion efficiency (
η
) are then evaluated by subtracting the dark electr
olysis J-E data obtained
using a p
+
-Si|NiO
x
dark electrode from the J-E data exhibited by the
illuminated photoelectrode
under evaluation.
Estimation of the lower limit of total Turnover num
ber, percentage of active Ni atoms, charge
required for dissolving the entire Si photoelectrod
e, and branching ratio of water oxidation
relative to Si oxidation
To estimate the lower limit on the turnover frequen
cy per active site, a lattice constant of 0.4195
nm was used for a cubic NiO crystal with a NaCl str
ucture. The unit cell contained 4 Ni atoms
and the density of the Ni atoms in the NiO was calc
ulated using the number of atoms divided by
the cell volume, which is 5.42×10
22
cm
-3
. The turnover frequency
5,6
in the unit of number of O
2
11
molecule per second per electrochemical active Ni a
tom can be calculated assuming a constant
current density of 32 mA cm
-2
(
j
) using equation
 ! =
" ∙ 6.02 × 10
(
1000 ∙ 96485 ∙ 4
1
-.
/
-.0
= 0.18

1
234 5 6/78
Therefore, the total turnover number (TON) is TOF*4
320000 s which is around 0.8×10
6
.
The percentage of active Ni atoms in the film based
on the reductive peak area in the Ni redox
peak region was calculated as follows:
5% =
:
;<
∙ 6.24 × 10
=>
-.
/
-.0
where
Q
act
is the charge density (based on the peak area) in
mC cm
-2
associated with the quantity
of electroactive Ni atoms on the surface.
The total charge passed during the 1200 h of chrono
amperometry was ~144000 C cm
-2
. The
charge needed to dissolve the Si photoelectrode was
calculated by first calculating the atom
density following the same approach used for calcul
ating the density of Ni atoms in NiO. The Si
diamond lattice constant of 0.543 nm was used with
8 Si atoms in each unit cell. The Si atom
density was ~5×10
22
cm
-3
. To break the Si–Si bond during etching in KOH in
volves 4 electrons.
Therefore, the total charge needed can be calculate
d by multiplying the atom density, surface area,
the Si wafer thickness of 400 μ m, and the elementar
y charge of 1.6×10
-19
coulombs:
:
?.
= 4 /
?.

To estimate the branching ratio of water oxidation
relative to Si oxidation, it was assumed that 10
nm of uniform SiO
2
would fully passivate the Si surface. This was co
nservative because if
thinner SiO
2
(3-4 nm) were fully passivating, then the branchin
g ratio estimate would increase by
a factor of 2-3. The total charge needed to form 10
nm of SiO
2
was then calculated by multiplying
12
the Si atom density, thickness of the oxide, expans
ion coefficient assuming that the volume
expansion occurred in one direction normal to the s
urface (1/2.2), and the elementary charge.
:
@A
= 4 /
@A
/2.2
The branching ratio was then given by
Q
ox
/
Q
total
.
Supplementary Figures
Figure S1
. Tafel plots for the oxygen-evolution reaction at
p
+
-Si|NiO
x
electrodes in 1.0 M
KOH(aq) before (black line) and after (red) ten cyc
lic voltammetric scans.
13
Figure S2.
Transmission of metallic Ni coated quartz substrat
es and catalytic activity measured
on metallic Ni coated p
+
-Si substrates in 1.0 M KOH(aq)
Figure S3.
Wavelength-dependent electrochromic properties of
NiO
x
and ultrathin Ni metal
coated FTO glass substrates under anodic operation
(1.6 V vs. RHE) and under cathodic (bleach)
conditions (0.7 V. vs. RHE) in 1.0 M KOH, respectiv
ely.
Supplementary Table:
Table S1. Comparative performances of Si photoanode
s for water oxidation
14
Year/Month
Photoanode
η
(%)/condition*
Stability
Ref
2011/06
n-Si|SiO
x
|TiO
2
|Ir
0.37 / 1 M NaOH (aq)
8 h
7
2011/06
npp
+
-Si|ITO|Co-Pi
0.04 / 0.1 M K-Pi(aq)
12 h
8
2011/11
n-Si|Fe
2
O
3
0.04 / 1 M NaOH(aq)
1 h
9
2012/04
n-Si|SiO
x
|NiO
x
0.006 / 0.25 M PBS Na
2
SO
4
(aq)
30 m
10
2013/02
n-Si|MnO
0.23 / 1 M KOH(aq)
30 m
11
2014/01
np
+
-Si|ITO:Au|NiO
x
0.07 / 52 mW cm
-
2
, 1 M NaOH(aq) 2.5 h
12
2014/01
n-Si|SiO
x
|Ni
0.14 / 200 mW cm
-
2
, 1 M KOH (aq) 24 h
13
2014/04
np+-Si|SiO
x
|CoO
x
0.75 / 1 M NaOH (aq)
24 h
1
2014/05
np
+
-Si|IrO
x
- / 38.6 mW cm
-
2
λ
> 635 nm, 1 M
H
2
SO
4
(aq)
18 h
14
2014/05
np
+
-Si|TiO
2
|Ni islands 0.32 / 125 mW cm
-
2
, 1 M KOH (aq) 100 h
15
2014/09
np
+
-Si|Ni|NiO
x
- / 38.6 mW cm
-
2
λ
> 635 nm, 1 M
KOH (aq)
300 h
16
2014/12
np
+
-Si|NiO
x
2.1 / 1 M KOH (aq)
1200 h
This work
*
η
(%) is the solar-to-O
2
(g) figure-of-merit defined by (
E
-
E
O’
H2O/O2
)
m
*
J
m
/
P
, where the product
(
E
-
E
O’
H2O/O2
)
m
and
J
m
defines the maximum power point relative to the Ner
nst potential for water
oxidation, data point were extracted from the cited
original work. Light intensities are 100 mW
cm
-2
, unless specified. “-” indicates that the figure-o
f-merit cannot be calculated due to the
different light source used.
References
1
Yang, J.
et al.
Efficient and Sustained Photoelectrochemical Water
Oxidation by Cobalt
Oxide/Silicon Photoanodes with Nanotextured Interfa
ces.
J. Am. Chem. Soc.
136
, 6191-
6194, doi:10.1021/ja501513t (2014).
2
McLean, D. & Feldman, B. Method for etching. Unit
ed States patent (2006).
3
Shaner, M. R., Fountaine, K. T. & Lewerenz, H. J.
Current-voltage characteristics of
coupled photodiode-electrocatalyst devices.
Applied Physics Letters
103
,
doi:10.1063/1.4822179 (2013).
4
Mills, T. J., Lin, F. & Boettcher, S. W.
Physical Review Letters
112
, 148304 (2014).
15
5
Kibsgaard, J., Jaramillo, T. F. & Besenbacher, F.
Building an appropriate active-site
motif into a hydrogen-evolution catalyst with thiom
olybdate [Mo3S13]2− clusters.
Nat.
Chem.
6
, 248-253, doi:10.1038/nchem.1853
http://www.nature.com/nchem/journal/v6/n3/abs/nchem
.1853.html#supplementary-information
(2014).
6
Benck, J. D., Chen, Z., Kuritzky, L. Y., Forman,
A. J. & Jaramillo, T. F. Amorphous
Molybdenum Sulfide Catalysts for Electrochemical Hy
drogen Production: Insights into
the Origin of their Catalytic Activity.
ACS Catalysis
2
, 1916-1923,
doi:10.1021/cs300451q (2012).
7
Chen, Y. W.
et al.
Atomic layer-deposited tunnel oxide stabilizes sil
icon photoanodes for
water oxidation.
Nat. Mater.
10
, 539-544, doi:10.1038/nmat3047 (2011).
8
Pijpers, J. J. H., Winkler, M. T., Surendranath,
Y., Buonassisi, T. & Nocera, D. G. Light-
induced water oxidation at silicon electrodes funct
ionalized with a cobalt oxygen-
evolving catalyst.
Proc. Natl. Acad. Sci. USA
108
, 10056-10061,
doi:10.1073/pnas.1106545108 (2011).
9
Jun, K., Lee, Y. S., Buonassisi, T. & Jacobson, J
. M. High Photocurrent in Silicon
Photoanodes Catalyzed by Iron Oxide Thin Films for
Water Oxidation.
Angew. Chem. Int.
Ed.
51
, 423-427, doi:10.1002/anie.201104367 (2012).
10
Sun, K.
et al.
Nickel oxide functionalized silicon for efficient
photo-oxidation of water.
Energy Environ. Sci.
5
, 7872-7877, doi:10.1039/C2EE21708B (2012).
11
Strandwitz, N. C.
et al.
Photoelectrochemical Behavior of n-type Si(100) El
ectrodes
Coated with Thin Films of Manganese Oxide Grown by
Atomic Layer Deposition.
J.
Phys. Chem. C
117
, 4931-4936, doi:10.1021/jp311207x (2013).
12
Sun, K.
et al.
Si photoanode protected by a metal modified ITO la
yer with ultrathin NiOx
for solar water oxidation.
Phys. Chem. Chem. Phys.
16
, 4612-4625,
doi:10.1039/C4CP00033A (2014).
13
Kenney, M. J.
et al.
High-Performance Silicon Photoanodes Passivated wi
th Ultrathin
Nickel Films for Water Oxidation.
Science
342
, 836-840, doi:10.1126/science.1241327
(2013).
14
Mei, B.
et al.
Protection of p+-n-Si Photoanodes by Sputter-Depos
ited Ir/IrOx Thin Films.
J. Phys. Chem. Lett.
5
, 1948-1952, doi:10.1021/jz500865g (2014).
15
Hu, S.
et al.
Amorphous TiO2 coatings stabilize Si, GaAs, and Ga
P photoanodes for
efficient water oxidation.
Science
344
, 1005-1009, doi:10.1126/science.1251428 (2014).
16
Mei, B.
et al.
Iron-Treated NiO as a Highly Transparent p-Type Pr
otection Layer for
Efficient Si-Based Photoanodes.
J. Phys. Chem. Lett.
5
, 3456-3461,
doi:10.1021/jz501872k (2014).