1

Electronic Supplementary Information

Insight into the Charge Transfer in Particulate Ta3N5 Photoanode with

High Photoelectrochemical Performance

Zhiliang Wang,‡ab Yu Qi,‡ab Chunmei Ding,a Dayong Fan,ab Guiji Liu,ab Yongle Zhaoab and Can Lia*

aState Key Laboratory of Catalysis, Dalian Institute of Chemical Physics, Chinese Academy of Sciences,

Dalian National Laboratory for Clean Energy, The Collaborative Innovation Center of Chemistry for

Energy Materials (iChEM), Zhongshan Road 457, Dalian, 116023, China. E-mail: canli@dicp.ac.cn

bUniversity of Chinese Academy of Sciences, China.

Electronic Supplementary Material (ESI) for Chemical Science.This journal is © The Royal Society of Chemistry 2016

2

Preparation of different kinds of metal based substratesThe Pt and Ta based substrates were prepared by sputtering coating. Generally, metal layer with thickness of 3~5 μm was deposited onto a Ti foil (1 cm×2 cm) with the target (Pt or Ta) on a home-built sputtering coater.Characterization of Photocatalytic water oxidationThe powder (13 mg) of necked-Ta3N5 was peeled off from a large piece of electrode (2.7 cm ×2.8 cm). For comparison, similar amount of raw Ta3N5 powder (16 mg) was used for the PC reaction.The photocatalytic reactions were carried out in a Pyrex top-irradiation-type reaction vessel connected to a closed gas circulation system.1 100 mL aqueous solution containing a certain amount of photocatalyst was adopted, and 100 mg AgNO3 (Alfa Aesar, ≥ 99.9 %) was used as sacrificial electron acceptor. 0.10 g La2O3 (Sinopharm Chemical Reagent, ≥ 99.95 %) was added to keep the pH value of the reaction solution. Prior to irradiation, the reaction mixture was evacuated to make sure that the air was completely removed. Then a 300 W xenon lamp equipped with an optical filter (Hoya, L-42; λ≥ 420 nm) to cut off the ultraviolet light irradiated the system from the top side. A flow of cooling water was used to maintain the system at room temperature. The evolved gases were analyzed by gas chromatography (Agilent; GC-7890A, MS-5A column, TCD, Ar carrier).Fabrication of Ta3N5 device for TPV testTypically, the Ta3N5 device was fabricated by covering the electrode with a piece of FTO (Be careful in case of short circuit). During the experiment, it is found the Ti foil is not flat enough, so we used Ti coated glass as substrate. About 10 μm Ti was sputtered onto the glass (2 cm ×1 cm) in a home-built sputtering coater at the power of 50 W. Then the Ta3N5 powder was deposited onto it by EPD. The device was fabricated by clipping a piece of FTO (2.5 cm ×0.7 cm) on the Ta3N5 membrane. And for the device illuminated from frontward and backward direction (Figure S7 a), it is fabricated by depositing raw-Ta3N5 on FTO and further covering another pieces of FTO on the electrode.TPV test is conducted with 355 nm pulsed laser (5 ns). The power is 122 mW unless otherwise mentioned. The signals are connected by an oscilloscope (Tektronix TDS 3012C).Necking treatment of Ta3N5 electrode prepared by oxidation-nitridationTa3N5 electrode was prepared according to the literature. Typically, Ta foil (2 cm×1 cm) was oxidized at 580 oC for 7 min to form a layer of tantalum oxide. The raw Ta3N5 electrode was prepared by nitriding of the oxidized Ta foil in NH3 flow (250 sccm) at 950 oC (2 oC/min) for 5 h. After that, 20 mM TaCl5 methanol solution was used for necking treatment with 10 μL per time (five times in total). Then the electrode was heated at 600 oC (5 oC/min) for 1 h in NH3 flow (100 sccm).Estimation of Cbulk and Ccrystal

Cbulk here is defined as capacitive charging to the porous Ta3N5 matrix. It can be treated as a parallel capacitor, and the capacitance is determined as:

(S1)𝐶𝑏𝑢𝑙𝑘=

𝜀𝜀𝑜𝑑

The is the permittivity constant of Ta3N5, estimated to be 17 2; is the vacuum permittivity constant; 𝜀 𝜀𝑜d is the distance of two substrates, estimated to be 5 μm from SEM image below. And Cbulk is estimated to be 3×10-5 F/cm2. Cbulk is mainly determined by the intrinsic property of Ta3N5 (ε of Ta3N5). Thus, we assume that the values of Cbulk are at the same order of magnitude for raw and necked Ta3N5 devices.Ccrystal can be estimated from the τ1 (τ=RC). Ta3N5 has the electric conductivity of 1.3×10-4 (Ω cm)-1 3. Thus, the Ccrystal is estimated to be at the order of 2×10-8 F/cm2.

3

Figure S1. The SEM image of cross-section of necked-Ta3N5 electrode.

Inte

nsity

(a.u

.)

10 20 30 40 50 60

necked-Ta3N5

heated-Ta3N5

TaCl5-Ta3N5

raw-Ta3N5

2 (degree)

Ta3N5 powderTi

Figure S2. XRD patterns of as-prepared Ta3N5 powder and the fabricated electrodes

4

0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.80

2

4

6

j (

mA/

cm2 )

E (V vs RHE)

15 mM TaCl5-600 oC 20 mM TaCl5-600 oC 25 mM TaCl5-600 oC 10 mM TaCl5-600 oC 5 mM TaCl5-600 oC 20 mM TaCl5 raw Ta3N5-600 oC raw Ta3N5

dark

0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.80

2

4

6

j (m

A/cm

2 )E (V vs. RHE)

450 oC 550 oC 600 oC 650 oC

necked with 20 mM TaCl5

(a) (b)

Figure S3. The influences of (a) TaCl5 concentration and (b) necking treatment temperature on the PEC performance of Ta3N5 electrodes.

0.4 0.6 0.8 1.0 1.2 1.4 1.6

0

2

4

6

j (m

A/cm

2 )

E (V vs. RHE)

Ti

Ti|Ta

Ti|Pt

Figure S4. The influence of substrates on the activity. All the Ta3N5 electrodes were necked with 20 mM TaCl5 and heated at 600 oC in 100 sccm NH3 flow.

5

400 500 600 700 8000.0

0.5

1.0

1.5

Abso

rptio

n (a

.u.)

Wavelength (nm)

raw-Ta3N5 necked-Ta3N5neck-Ta3N5

300 400 500 600 700 8000.0

0.2

0.4

0.6

Wavelength (nm)

Curre

nt F

lux

(A·m

-2·nm

-1)

0

5

10

15

20

25 Integrated Current (mA/cm

2)

12.6mA/cm2

400 600 800

0.6

0.8

1.0

Wavelength (nm)

LHE

0

10

20

30

40

Absorbed Photon Flux

400 600 800 1000 1200

0.0

0.5

1.0

1.5

2.0

Wavelength (nm)

AM 1.

5G (

W·

m-2·

nm-1)

0

10

20

30

40

50

Photon Flux (10

17·m

-2·s

-1·nm

-1)

(a) (b)

(c) (d)

raw-Ta3N5

Figure S5. The calculation (blue) of integrated current of Ta3N5 electrode. (a) The light absorption spectra of raw-Ta3N5 (black) and necked-Ta3N5 (orange) electrode. (b) The spectrum of simulated sunlight (AM 1.5 G). (c) The light harvest efficiency (LHE) and absorbed photons. LHE can be got from the absorption with this expression4: LHE (λ)=1-10-A(λ). (d) The calculated integrated current. All the dates are for raw-Ta3N5 except that in (a).

6

Figure S6. Current-bias response of the Ta3N5 device under dark (red) and illumination (black). The light source used here is simulated AM 1.5G sunlight (1 sun).

0.0 0.5 1.0 1.5

0

20

40

60

80

Phot

ovol

tage

(mV)

Time (s)

122 mW 22 mW

0 10 20 30

Phot

ovol

tage

(mV)

Time (s)

frontward illumination

backward illumination

0

(a) (b)

Figure S7. (a) The responses of photovoltage upon frontward illumination (black, solid) and backward illumination (black, dashed line). The red dashed line indicates the 0 mV level. (b) The responses of the photovoltage with different laser power: 122 mW (black) and 22 mW (red)

7

-0.4 0.0 0.4 0.8 1.2 1.6-30

-20

-10

0

102nd

Curr

ent (

mA/

cm2 )

Potential (V vs RHE)

1st

Co2+/3+

Fe(CN)4-/3-6

3rd4th/5th

-0.4 0.0 0.4 0.8 1.2

-25

-20

-15

-10

-5

0

5

Curr

ent (

mA/

cm2 )

Potential (V vs RHE)

necked Ta3N5substrate

in 1 M NaOH+0.25 M K3Fe(CN)6

-0.2 0.2 0.6 1.0 1.4 1.8

-40

-20

0

20

40

60

O2 evolution

Cu

rren

t (m

A/cm

2 )

Potential (V vs RHE)

on Pt in 1 M NaOH+0.25 M K3Fe(CN)6

H2 evolution

Fe(CN)4-6 /Fe(CN)3-

6

(d)(c)

-1.0 -0.5 0.0 0.5 1.0 1.5 2.0-60

-40

-20

0

necked Ta3N5

Curr

ent (

mA/

cm2 )

Potential (V vs RHE)

substrate

in 1 M NaOH

H2 evolution

(a) (b)

CoPi/necked-Ta3N5

in 1 M NaOH+0.25 M K3Fe(CN)6

Figure S8. (a) The CV on Pt foil in NaOH (1 M)+K3Fe(CN)6 (0.25 M) aqueous solution to show the redox peaks of ferri/ferrocyanide. (b) and (c) The CV of necked Ta3N5 electrode (black) and its substrate (nitridized Ti foil, red) in NaOH (1 M) and NaOH (1 M)+K3Fe(CN)6 (0.25 M) aqueous solution, respectively. (d) The CV of CoPi/necked-Ta3N5 electrode in NaOH (1 M)+K3Fe(CN)6 (0.25 M) aqueous solution. The scan is repeated for 5 circles and the first 2 circles is found to be influenced by the reduction of CoPi. At the last circle of scan, the CV curve is similar to that of necked-Ta3N5 electrode in the current and potential of the reduction peak.

8

raw-Ta3N5 poor contacte-

e-

e-

e-

OR

R RR

O

necked-Ta3N5 good contacte- e-

e-

e-

e-

R

RR

R

R

O

O

O

O

necking

O: oxidation state, Fe(CN)6

3-

R: reduction state, Fe(CN)6

4-

Figure S9. The schematic illustration of electrochemical reduction of K3Fe(CN)6 on raw-Ta3N5 and necked-Ta3N5 electrode.

-0.1 0.0 0.1 0.2

-150

-100

-50

0

50

100

150500 400300200100

Capa

city

cur

rent

(A/

cm2 )

Potential (V vs SCE)

20

Scan rate: mV/sin NaOH

Figure S10. The CV of necked-Ta3N5 electrode in NaOH (1 M) at different scan rates (20-500 mV/s) in NaOH (1 M) aqueous solution.

9

0.6 0.8 1.0 1.2 1.4 1.6 1.8

0

2

4

6 necking treatment

j (m

A/cm

2 )

E (V vs RHE)

thermal oxide Ta3N5

Figure S11. The j-E curves for Ta3N5 electrode made from oxidation-nitridation (black) with further necking treatment (red). Electrolyte: 1 M NaOH aqueous solution. Illumination: 100 mW/cm2, AM 1.5G.

0.5 0.7 0.9 1.1 1.30

20

40

60

80 CoPi/necked-Ta3N5

inje

ctio

n ef

ficie

ncy

(%)

E (V vs RHE)

necked-Ta3N5

Figrue S12. The injection efficiency of necked-Ta3N5 (black) and CoPi/necked-Ta3N5 (red) electrodes.

References1. S. Chen, S. Shen, G. Liu, Y. Qi, F. Zhang and C. Li, Angewandte Chemie International Edition, 2015, 54,

3047-3051.2. A. Ziani, E. Nurlaela, D. S. Dhawale, D. A. Silva, E. Alarousu, O. F. Mohammed and K. Takanabe, Physical

Chemistry Chemical Physics, 2015, 17, 2670-2677.3. J. Swisher and M. Read, Metallurgical Transactions, 1972, 3, 493-498.4. T. W. Kim and K.-S. Choi, Science, 2014, 343, 990-994.