Supporting Information

Nonmetal P-doped Hematite Photoanode with Enhanced

Electron Mobility and High Water Oxidation Activity

Yuchao Zhang, a Shiqi Jiang, b Wenjing song, a Peng Zhou, a Hongwei Ji, a Wanhong Ma, a Weichang Hao, b

Chuncheng Chen,* a Jincai Zhao a

a Key Laboratory of Photochemistry, Beijing National Laboratory for Molecular Sciences, Institute of

Chemistry, Chinese Academy of Sciences, Beijing 100190, China

b Center of Materials Physics and Chemistry, Beihang University, Beijing 100191, China

Electronic Supplementary Material (ESI) for Energy & Environmental Science.This journal is © The Royal Society of Chemistry 2015

EXPERIMENTAL SECTION

Photoanode preparation. Hematite nanowires were prepared on a fluorine-doped tin oxide (FTO, TCO-15,

Nippon Sheet Glass, Japan, 14 ohm/sq) glass substrate by a modified procedure reported by Yat Li and

coworkers.1 Briefly, 100 ml aqueous solution containing 2.43 g of ferric chloride (FeCl3.6H2O, Alfa Aesar, 98%)

and 0.85 g of sodium nitrate (NaNO3, J&K, 99%) at pH = 1.4 (adjusted by HCl) were prepared in a Teflon-lined

stainless steel autoclave. Several FTO glass slides (2 4 cm), washed with acetone, ethanol, and then deionized

water, were placed in the autoclave and heated at 95 °C for 4 hours. After the hydrothermal treatment, a uniform

layer of iron oxyhydroxides (FeOOH) was coated on the FTO glass, which was washed completely with deionized

water to remove any residual salts. The obtained film was sintered in air at 550°C for 2 hours to convert the

FeOOH nanowires into hematite nanowires. The hematite nanowires were further annealed at 650°C for 20 min, or

at 750°C for 15 min.

P-doped hematite nanowires were prepared by the same procedure as hematite nanowires, except that FeOOH

films were soaked in 0.05M of Na2HPO4 solution with different pH for 1min and then immediately underwent the

sintering and annealing procedure without other modifications. The pH of the soaking solution was adjusted by

HCl and NaOH solutions and measured by a pH-meter (Thermo Scientific, 3-Star).

“Co-Pi” modification was fabricated according to the photo-assisted electrodeposition methods.2 The hematite

film was submerged in a solution of 0.5 mM cobalt nitrate in 0.1 M potassium phosphate (KPi) buffer (pH 7). A

150 W Xenon lamp coupled to a filter (AM 1.5G) was used as the white light source. The light power density of

100 mW/cm2 was measured with a radiometer (CEAULIGHT, CEL-NP2000). “Co-Pi” was electrodeposited at

+0.1 V vs. Ag/AgCl for 15 min.

Photoelectrochemical characterization PEC performances of the films were measured in 1M NaOH electrolyte

solution (pH 13.6) in a three-electrode electrochemical cell. The electrolyte solution was deaerated by purging

argon for 30 min before J-V scan. The measured potentials vs Ag/AgCl were converted to the reversible hydrogen

electrode (RHE) scale according to the Nernst equation: ERHE = EAg/AgCl + 0.059pH + 0.1976. A 150 W Xenon

lamp coupled to a filter (AM 1.5G) was used as the white light source. The light power density of 100 mW/cm2

was measured with a radiometer (CEAULIGHT, CEL-NP2000). The exposed area for all the photoanodes were

controlled by a Teflon mold with a standard 1 cm2 hole.

Electrochemical impedance spectra were measured under visible light illumination (λ>420 nm, I0=60 mW/cm2),

recorded by an electrochemical workstation (PGSTAT302N autolab, Metrohm). A sinusoidal voltage pulse of 10

mV amplitude was applied on a bias voltage, with a frequency that ranged from 10 kHz to 1Hz. Mott-Schottky

plots were collected with bias voltage scanned from -0.4 to 0.1 V vs. Ag/AgCl with a 50 mV interval, in the dark

at a frequency of 1000Hz. The raw data were fitted using Nova 1.8 software from Metrohm Inc.

Oxygen detection Oxygen was detected by a Thermo Scientific Orion 3-Star dissolved oxygen meter. The

photoelectrochemical cell was a two-compartment cell divided by a porous ceramics for cutting off the gas flow

but keeping the conduction of the circuit. The hematite photoanode together with Ag/AgCl reference electrode and

the sensor of the dissolved oxygen meter were placed in one cell, and the Pt counter electrode was placed into the

other one. Both cells were filled with 1M NaOH electrolyte solution (pH 13.6). The electrolyte in the working

compartment was 30 mL. A 150 W Xenon lamp coupled to a filter (AM 1.5G) was used as the white light source.

The applied potential was 1.23 VRHE. Before every experiment, the whole system was deaerated by purging

high-purity argon for one hour and then all of the connectors were sealed by silicone rubber. This procedure was

repeated for several times until the reading of the dissolved oxygen meter kept below 0.1 mg/L for 30 min. The

quantity of electric charge passing through the circuit was recorded by an electrochemical workstation (CHI 760D).

The charge passed during the experiment was then converted to moles by dividing by 4F (F = 96485 C/mol).

Structural characterization X-ray diffraction (XRD) measurements were performed on a Regaku D/Max-2500

diffractometer with the Cu KR radiation (1.5406 Å). Diffraction patterns were recorded from 20 to 80° 2θ with a

step size of 0.04° at 4°/min. X-ray photoelectron spectroscopy (XPS) data were obtained with an

ESCALab220i-XL electron spectrometer from VG Scientific using 300W Al Kα radiation. The base pressure was

about 3×10-9 mbar. The binding energies were referenced to the C1s line at 284.8 eV from adventitious carbon. Ar

ion etching was conducted at an Ar partial pressure of 2 × 10-7 mbar at 3 kV and 10 mA resulting in 0.07 nm/sec

etching rate. SEM data were obtained with a Hitachi S4800 scanning electron microscope (HITACHI Ltd) with the

acceleration voltage of 15 KV. TEM data were performed on a Tecnai G2 F30 S-TWIN 300kV field-emission

microscope. XAFS experiments were carried out at the XAFS station of the Beijing Synchrotron Radiation Facility

(BSRF). The beamline provided a focused X-ray beam from 5 to 20 keV with a photon flux on the order of 1011

phs/s (@ 10 keV) and a beam size of 0.9（H）× 0.3（V）mm2. The energy was calibrated by a standard Fe metal foil

before experiments started. Fe K-edge X-ray absorption fine structure (XAFS) spectra of the samples were

collected by a Lyttle fluorescence detector at room temperature. The X-ray absorption fine structure (XAFS) was

collected from -20 to +60 eV relative to the Fe K-edge. P K-edge XAFS spectra were collected at beamline 4B7A

of BSRF, with energy range from 2.05keV to 6.0keV, and energy bandpass about 0.3eV at P K-edge. The sample

was measured in PFY model with a SDD detector. The data analysis of XANES was carried out using the Athena

and Artemis interfaces to the IFEFFIT program package.

DFT calculations The first-principle calculations were carried out using the Vienna ab initio simulation package

(VASP) based on DFT+U with projector-augmented wave (PAW) pseudo-potential method. The exchange and

correlation energy was treated via the generalized gradient approximation (GGA) using the Perdew-Burke-Emzerh

of formulation. A conjugate-gradient algorithm was used to relax the ions into their ground states, and the energies

and the forces on each ion were converged within 1.0 × 10-4 eV/atom and 0.01 eV/Å, respectively. The Koln-Sham

orbitals were expanded by a plane wave basis set and an energy cutoff of 400 eV was used throughout. O 2p

orbitals were regarded as valence band orbitals while Fe 3d and 4s were regarded as conduction band orbitals. The

Brillouin-zone integration is per-formed by using the Gamma-centered Monkhorst-Pack scheme with 5×5×1

k-points, together with a Gaussian smearing broadening of 0.1 eV.

The electron effective mass can be calculated as:

1. Y. Ling, G. Wang, D. A. Wheeler, J. Z. Zhang and Y. Li, Nano. Lett., 2011, 11, 2119-2125.

2. D. K. Zhong, M. Cornuz, K. Sivula, M. Graetzel and D. R. Gamelin, Energy. Environ. Sci., 2011, 4,

1759-1764.

0.6 0.8 1.0 1.2 1.4 1.6 1.80.0

0.5

1.0

1.5

2.0

2.5

3.0

J (

mA

/cm

2)

Potential (V vs. RHE)

3

5

6

7

8

9

10

11

12

13

A

2 4 6 8 10 12 14

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

1.8

PO4

3-

HPO4

2-

Ph

oto

cu

ren

t d

en

sit

y (

mA

/cm

-2)

B

0.0

0.2

0.4

0.6

0.8

1.0

H2PO

4

-

Mo

le F

rac

tio

n

pH

H3PO

4

Figure S1. (A) J-V scans collected for P-doped hematite synthesized by phosphate solutions with different pH and

annealed at 650C. (B) Photocurrent densities obtained at 1.23 VRHE.

0.6 0.8 1.0 1.2 1.4 1.60.00

0.05

0.10

0.15

0.20

Potential (V vs. RHE)

J (

mA

/cm

2)

pH3

pH8

pH12

Figure S2. J-V scans collected for undoped hematite photoanodes synthesized by solutions with different pH and

annealed at 650C.

0 5 10 15 20 25 30

0

2

4

6

8

10

Time (min)

O2 p

rod

uc

ed

(m

ol)

NW650

PNW650

NW750

PNW750

A

5 10 15 20 25 300

10

20

30

40

50

60

70

80

Time (min)

Fara

da

ic e

ffic

ien

cy

PNW650

NW750

PNW750

B

Figure S3. (A) O2 evolution and (B) the corresponding faradic efficiency detected by a dissolved oxygen meter

during the PEC water oxidation at 1.23 VRHE for the pristine and P-doped hematite photoanodes annealed at 650C

and 750C. The faradaic efficiencies for all the PEC oxygen evolution were around 70%. Since no other species

except H2O can be oxidized in our PEC systems, the deviation of the faradaic efficiencies from unity should be

caused by some systematic errors.

0.0 0.2 0.4 0.6 0.8 1.0

2.6

2.7

2.8

2.9

3.0

3.1

3.2

3.3

3.4

J (

mA

/cm

2)

Time (h)

A

0 5 10 15 20 25 30 35 402.5

2.6

2.7

2.8

2.9

3.0

J (

mA

/cm

2)

Time (h)

B

Figure S4. Long-term PEC experiments for (A) PNW750+CoPi and (B) PNW750 in 1 M NaOH at 1.23 VRHE.

400 500 600 700 800-0.2

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

Ab

so

rban

ce (

a.u

.)

Wavelength (nm)

NW650

PNW650A

400 500 600 700 800-0.2

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4 B

Ab

so

rba

nc

e (

a.u

.)

Wavelength (nm)

NW750

PNW750

350 400 450 500 550 600 6500

10

20

30

40

Wavelength (nm)

IPC

E (

%)

PNW750

NW750

PNW650

NW650

C

Figure S5. UV-vis data collected for P-doping hematite photoanodes annealed at (A) 650C and (B) 750C. (C)

IPCE values collected for undoped and P-doped hematite photoanodes.

0.60 0.65 0.70 0.75 0.80 0.85 0.900.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

4.5

5.0

Potential (V vs. RHE)C

-2 (

10

10 F

-2 c

m4) NW650

PNW650

Figure S6. Mott-Schottky plots collected in the dark for undoped and P-doped hematite photoanodes annealed at

650C.

0 2 4 6 8 10 12 140

2

4

6

8

10

12

Z' / (102)

-Z'' (

10

2

)

NW650-exp

PNW650-exp

NW650-fit

PNW650-fit

CPE

CPE

Rs

Cbulk

Rtrapping

Rct,trap

Ctrap

Figure S7. Nyquit plot collected for NW650 and PNW650 at 1.2VRHE. Inset is the equivalent circuit used for the fit

and simulation. In this model, Rs stands for all the series resistances in the electrochemical cell, Cbulk represents the

capacitance of the space charge depletion region at the surface of the electrode, Rtrapping represents the resistance in

the process of trapping holes by surface states, Ctrap represents the amount of active sites in surface states, and

Rct,trap represents the resistance in the process of charge transfer across the interface.

pH3 pH8 pH12

-90

-60

-30

0

30

60

90

Ph

oto

cu

rren

t (m

A)

Rc

t,tr

ap

(

3

cm

2) A

0

1

2

3

4

pH3 pH8 pH120

30

60

90

120

150

Rs (c

m2)

B

pH3 pH8 pH120.0

0.5

1.0

1.5

2.0

2.5

Cb

ulk(1

0-5 F

cm

-2) C

Figure S8. Fitted values of (A) Rct,trap, (B) Rs, (C) Cbulk from EIS spectra collected for P-doped hematite anodes

impregnated by phosphate solutions with different pH and annealed at 650C.

-20 -15 -10 -5 0 5 10

-40

-20

0

20

40

DO

S (

sta

tes

.eV

-1.a

tom

-1)

Energy (eV)

Fe2O

3

2.0 eV

Figure S9. DOS of pure hematite.

Figure S10. DOS of P-doped hematite structures.

0

En

erg

y (

eV

)

Fe2O

3

-2.5

2.5

HA K M L H

A)

0

B)

En

erg

y (

eV

)

Fe1

F Q Z-2.5

2.5

0

C)

En

erg

y (

eV

)

Fe2

F Q Z-2.5

2.5

0

D)

En

erg

y (

eV

)

Fe3

F Q Z-2.5

2.5

0

E)

En

erg

y (

eV

)

Fe4

F Q Z-2.5

2.5

Figure S11. Band structures of (A) pristine hematite and (B)-(E) P-doped hematite.

A) B)

Figure S12. Electron density distribution plots for (A) pristine and (B)-(E) P-doped hematite (110) surface.

0 200 400 600 800 1000 12000

100000

200000

300000

400000

500000

600000

700000

800000

Binding Energy (ev)

Co

un

ts

NW650

PNW650A

P 2p

125 130 135 140 145

6000

6500

7000

7500

8000

8500

9000

9500

Binding Energy (ev)

Co

un

ts

B

Figure S13. (A) The XPS survey data collected for NW650 and PNW650. (B) The P 2p detail spectrum collected

for PNW650. The XPS survey data confirm the presence of P on the surface of PNW650, while no P signal is

detected on NW650. The high resolution P 2p spectrum shows that the P 2p binding energy of 133.3 eV which is

C) D)

E)

much higher than the reported for phosphate (132.1-132.9 eV), but is nearly consistent with values reported for

FePO4 (133.4 eV).

Figure S14. SEM images of undoped and P-doped hematite (cross-section view).

Figure S15. SEM images for P-doped (right) and undoped (left) hematite photoanodes annealed at 750C. Similar

to the situation in 650C, the P-doped hematite nanowires annealed at 750C became thinner than the undoped

ones.

10 20 30 40 50 60 70 80

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

1.8

2.0

(30

0)

(10

4) (1

10

)

***

*

2 Theta (degree)

No

rma

lize

d In

ten

sit

y (

a.u

.)

NW650

PNW650

*

* FTO A

10 20 30 40 50 60 70 80

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

1.8

2.0

*(300

)

**

*

(11

0)

(10

4)*

* FTO

2 Theta (degree)

No

rma

lize

d In

ten

sit

y (

a.u

.)

NW750

PNW750B

Figure S16. XRD patterns of undoped and P-doped hematite for (A) annealed at 650C and (B) annealed at 750C.

PNW650 NW650

NW750 PNW750

7114 7116 7118 7120 7122 7124

0.2

0.4

0.6

0.8

E (eV)

No

rmo

lized

x

pH3

pH8

pH12

2150 2151 2152 21530

2

4

6

8

E (eV)

No

rmo

lized

x

pH3

pH8

pH12

Figure S17. A comparison of Fe K-edge X-Ray absorption near edge structure (XANES) spectra (left) and P

K-edge XANES spectra (right). For Fe K-edge XANES spectra, the pre-edge peak at 7114 eV exhibited a gradual

increase with the decrease of pH. This pre-edge peak is a signature of the 1s-3d transition of tetrahedrally

coordinated Fe, which is commonly found in Fe3O4 and quite weak for the α-Fe2O3. The larger pre-edge peak

suggests that the content of tetrahedrally coordinated Fe increased with the decrease of pH.

20 30 40 50 60 70 80

0

1

2

3

4

5

(30

0)

(11

0)

(10

4)

****

*

* FTO

2 Theta (degree)

No

rma

lize

d In

ten

sit

y (

a.u

.)

NW650

pH3

pH8

pH12

A

560 580 600 620 6400.25

0.30

0.35

0.40

0.45

0.50

0.55

0.60

Re

lati

ve

In

ten

sit

y o

f (1

10

) D

iffr

ac

tio

n

pH3

pH8

pH12

Temperature(0C)

B

Figure S18. (A) XRD patterns collected for undoped and P-doped hematite anodes impregnated by phosphate

solution with different pH and annealed at 650C. (B) The relative (110) diffraction from temperature programmed

XRD.

Table S1. Fitted values of the EIS data in Figure 2A

Rs Cbulk Rtrapping Ctrap Rct,trap

NW650 142.5 Ωcm2 5.9 × 10-6 F

cm-2 3194.2 Ωcm2

4.6 × 10-5 F

cm-2 25441.2 Ωcm2

PNW650 119.4 Ωcm2 2.2 × 10-5 F

cm-2 230.6 Ωcm2

4.9 × 10-4 F

cm-2 81.5 Ωcm2

Table S2. The content of P from XPS experiments for P-doped hematite anodes impregnated by phosphate solution

with different pH and annealed at 650C.

pH3 pH8 pH12

P% 4.14 4.87 2.92