Supporting information

Hydrothermal growth of highly oriented single crystalline Ta2O5 nanorod arrays

and their conversion to Ta3N5 for efficient solar driven water splitting

Zixue Su,[a] Lei Wang,[a] Sabina Grigorescu,[a] Kiyoung Lee,[a] Patrik Schmuki [a,b] *

[a] Department of Materials Science and Engineering, WW4-LKO, University of Erlangen-Nuremburg,

Martensstr. 7, D-91058 Erlangen, Germany.

* Corresponding Author: schmuki@ww.uni-erlangen.de (P. Schmuki)

[b] Department of Chemistry, King Abdulaziz University, Jeddah, Saudi Arabia.

Electronic Supplementary Material (ESI) for ChemComm.This journal is © The Royal Society of Chemistry 2014

Experimental Section

Two different sizes of Ta foils with a thickness of 0.10 mm (99.9% purity, Advent,

England) were used including the standard size of 1.25 cm x 1.25 cm, and the small size of

0.6 cm x 0.6 cm. In a typical synthesis of Ta2O5 nanorod arrays, a piece of Ta foil was first

degreased by sonication successively in ethanol, acetone, and deionized water. After drying in

N2, the pretreated Ta foil was put into a 250 ml Teflon-lined autoclave filled with an aqueous

solution of 0.1 M HF with or without H2O2. The filling factor of the solution (volume of

liquid to total autoclave volume) varied from 10% to 40%. The autoclave was then heated to

the target temperature of 240 °C for 6 ~ 48 h to enable the growth of Ta2O5 nanorod arrays on

the surface of Ta foil. To obtain Ta3N5 nanorod arrays, the hydrothermally grown Ta2O5

nanorod arrays on the Ta foils were then heated in a quartz tube furnace in a gaseous

atmosphere of NH3 with a flow rate of about 200 mL min-1 at 1000 °C for 2 h.

Before the water splitting measurement, Co(OH)x or Co-Pi were deposited on the Ta3N5

nanorod surface as oxygen evolution catalysts. (resulting in both cases in approx. 0.3 at.% Co

content (EDX in Figure S4)) In a typical Co(OH)x treatment, the nanorod photoanodes were

immersed in a mixed solution of 0.1 M CoSO4 and 0.1 M NaOH with a ratio of 1:1 for 25 min,

then washed and dried in N2. The Co-Pi catalyst was loaded by electrodeposition in a solution

of 0.5 mM Co(NO3)2 in 0.1 M potassium phosphate buffer at pH=7 at 1 V versus Ag/AgCl for

8 min. The photoelectrochemical water splitting was carried out in a standard three-electrode

system, where the Ta3N5 nanorod arrays, Ag/AgCl electrode and Pt electrode acted as

working electrode, reference electrode and counter electrode, respectively. The electrolyte

used was aqueous KOH solution with a pH value of 13.7. The working electrode was

illuminated under AM 1.5G irradiation (100 mW cm-2). According to the Nernst equation

(ERHE = EAg/AgCl+0.059pH+0.196), potentials vs. Ag/AgCl can be converted to potentials vs.

the RHE.

Morphological characterizations of the grown Ta2O5 and Ta3N5 nanorod arrays were

carried out on a field emission scanning electron microscopy (Hitachi FE-SEM S4800, Japan).

X-ray diffraction (X’pert Philips MPD with a Panalytical X’celerator detector, Germany) was

carried out using graphite monochromized Cu Kα radiation (Wavelength 0.154056 nm).

Transmission electron microscopy (TEM) was performed by using a Philips CM300

UltraTWIN, equipped with a LaB6 filament and operated at 300 kV. TEM images and

diffraction patterns were recorded with a fast scan (type F214) charge-coupled device camera

from TVIPS (Tietz Video and Image Processing Systems), with an image size of 2048 x 2048

pixels. The SAED patterns were evaluated by using the software JEMS1 and incorporating

crystal data information from the inorganic crystal structure data base (ICSD). For TEM

investigations the samples were first mechanically scratched from the Ta substrate and the

resulting sample powder was prepared on commonly used copper TEM grids coated with

lacey carbon film.

1. A. Stadelmann, Jems Electron Microscopy Software (1999–2012), java version 3.7624U2012,

CIME-EPFL, Switzerland.

Figure S1. TEM image showing the micropores present in the base oxide layer underneath

the nanorod arrays.

Figure S2. Current-potential curves of PEC water splitting cell with a photoanode of

Co(OH)x treated Ta3N5 nanorod arrays obtained from 2 h nitridation at 1000 °C of 12

h, 24 h, and 48 h grown Ta2O5 nanorod arrays. All curves were measured in aqueous

KOH solution (pH=13.7) under chopped AM 1.5G simulated sunlight, with a scan rate

of 10 mV s-1 from negative to positive.

Figure S3. TEM analysis of the Ta3N5 sample: a) and b) are BF TEM images showing

agglomerates of nanoparticles. The BF TEM image b) shows a magnified region from a). The

SAED pattern in c) was recorded at the position marked with the dashed circle in a). In d-f)

the same experimental diffraction pattern (which is shown in c) is shown with the

corresponding simulations obtained by software JEMS.

Figure S3 shows the TEM analysis of the sample Ta3N5. The bright-field (BF) TEM images

in a) and b) show a representative agglomerate of particles found on the TEM grid after the

preparation procedure. The black dashed circle in a) marks the area where the selected area

electron diffraction (SAED) pattern, which is shown in c), is acquired. The SAED patern in c)

exhibits dotted circles, which indicates the presence of nanoparticles, being in good

agreement with observations from BF TEM imaging. The evaluation of the diffraction

patterns was performed by simulating diffraction patterns with the software JEMS and

comparison with the experimental SAED pattern. Representative results are shown in Figure

S3 d-f), which confirm the presence of Ta3N5 and TaN, having orthorhombic and hexagonal

crystal structure, respectively. For a better overview, in d) the simulated and experimental

diffraction pattern of Ta3N5 (ICSD 66533) is shown, while e) shows the simulated and

experimental pattern of TaN (ICSD 25659). In f) the experimental and simulated patterns of

both, Ta3N5 and TaN, are shown.

Figure S4. EDS analyses of (a) Co(OH)x and (b) Co-Pi treated Ta3N5 nanorods.

Element at.%

Ta 51.7

N 47.95

Co 0.35

Element at.%

Ta 55.56

N 44.12

Co 0.32

(a)

(b)

Figure S5. Typical SEM images of (a) Ta2O5 films hydrothermally grown on pure Ta foils in

0.1 M HF at 240 °C for 24 h with a filling factor of 40%, (b) Ta2O5 films grown on Ta foils

by 40 min oxidation at 500 °C in air, and Ta2O5 films grown on the pre-oxidized Ta foil under

the same condition of (a) for 6 h (c) and 24 (h) respectively.

Figure S6. SEM images of Ta2O5 films grown on a Ta foil by 48 h hydrothermal reaction in

0.1 M HF with 40% filling factor of solution under 240 °C. The Ta foils used in (a, c) are of

standard size, while Ta foils used in (b, d) are of small size.

Figure S7. SEM images of Ta2O5 films grown on a Ta foil by 24 h hydrothermal reaction in

0.1 M HF with 10% filling factor of solution under 240 °C.

Influence of various experimental parameters such as pre-oxidation of the Ta, Ta foil size,

and solution filling factor on the morphologies of Ta2O5 film were studied. Figure S5 shows

that a Ta2O5 film with a thickness of about 1.6 µm was produced after 40 min oxidation at

500 °C in air. Due to the dissolution of the pre-formed Ta2O5 films, the H2TaF7 concentration

in the solution would be higher compared to that when standard pure Ta foils were used after

the same reaction time. This led to an earlier growth of the nanorod arrays (Figure S5c) for 6

h and thicker base oxide films for 24 h reaction (Figure S5c). Since Ta foils of small size

have smaller contact areas with Ta foils of standard size, the H2TaF7 concentration in the

solution will also be lower after the same condition. Figure S6 shows that after 48 h

hydrothermal reaction, the thickness of the base oxide layer is much thinner when a smaller

Ta foil was used. The concentration of H2TaF7 could also be increased by decreasing the

solution filling factor after the same time reaction due to smaller amount of water in the

solution. Figure S7 shows that when the solution filling factor decreased to 10%, 24 h

hydrothermal reaction in 0.1 M under 240 °C led to the growth of a thin nanorod layer and a

thick base oxide layer. Moreover, a thick layer of dispersed flower-like Ta2O5 nanorods was

also found to be deposited on top of the nanorod layer, which probably formed directly in the

bulk solution due to very high concentration of H2TaF7.