Supplementary informationToward highly efficient in-situ dry reforming of H2S contaminated

methane in solid oxide fuel cells via incorporating coke/sulfur resistant bimetallic catalyst layer

Bin Hua a, Ning Yan b, Meng Li c, Yi-Fei Sun a, Jian Chen d, Ya-Qian Zhang a, Jian Li

c, Thomas Etsell a, Partha Sarkar e, Jing-Li Luo a, *

* Corresponding author

a Department of Chemical and Materials Engineering, University of Alberta, Edmonton, Alberta, T6G 1H9, Canada, Tel.: +1 780 492 2232; fax: +1 780 492 2881. E-mail address: jingli.luo@ualberta.ca

b Van’t Hoff Institute for Molecular Sciences (HIMS), University of Amsterdam, Amsterdam, 1098XH, The Netherlands, Tel.: +31 020 525 6468; E-mail address: n.yan@uva.nl

c Center for Fuel Cell Innovation, School of Materials Science and Engineering, State Key Laboratory of Material Processing and Die & Mould Technology, Huazhong University of Science and Technology, Wuhan, Hubei, 430074, China

d National Institute for Nanotechnology, Edmonton, Alberta T6G 2M9, Canada

e Environment & Carbon Management Division, Alberta Innovates-Technology Futures, Edmonton, Alberta, T6N 1E4, Canada

Electronic Supplementary Material (ESI) for Journal of Materials Chemistry A.This journal is © The Royal Society of Chemistry 2016

Methods

Preparation of the catalysts

NiCu-Ce0.8Zr0.2O2 (ZDC) catalyst was prepared by a glycine-nitrate process

(GNP). Ni(NO3)2·6H2O, Cu(NO3)2·3H2O, Zr(NO3)4·5H2O and Ce(NO3)3·6H2O were

dissolved into distilled water at a molar ratio of 6:1:2:8 with glycine added. This

solution was then combusted at 200 °C and calcined at 800 °C for 2 h to form NiCuO-

Ce0.8Zr0.2O2 powder, which was in-situ reduced in H2 to form NiCu-Ce0.8Zr0.2O2

catalyst. The volume percentage of metal phase (Ni-Cu alloy) in NiCu-ZDC cermet is

around 23.3%, ensuring adequate electronic conductivity for SOFC　operation. In a

control group, we also prepared the Ni-ZDC (molar ratio of Ni:Zr:Ce is 7:2:8), NiCo-

ZDC (molar ratio of Ni:Co:Zr:Ce is 6:1:2:8) and NiFe-ZDC (molar ratio of

Ni:Fe:Zr:Ce is 6:1:2:8) using the same method.

Preparation of the cathode

To prepare La0.6Sr0.4Co0.2Fe0.8O3-δ (LSCF) cathode material,

La(NO3)3·6H2O, Sr(NO3)2, Co(NO3)2·6H2O and Fe(NO3)3·9H2O were dissolved in

distilled water with EDTA/citric acid added as the chelating agent. The aqueous

solution was heated at 80 °C under agitation to a viscous gel, and then dried at 180 °C

to form a black foamy intermediate product, which was ground into fine powder and

calcined at 800 °C for 2 h in air to obtain perovskite structure LSCF.

Cell fabrication

Tape-casting/screen-printing/sintering processes were used for fabrication of the

anode supported cells. NiO (Type A Standard, Inco) and YSZ (TZ8YS, Tosoh) in

57:43 weight ratio were ball milled for 24 h in toluene/ethanol solvent with fish oil as

the dispersant and corn starch as the pore former. It was further milled for another 24

h after adding polyvinyl butyral (Richard E. Mistler Inc., USA) as the binder and

polyethylene glycol (Richard E. Mistler Inc., USA) as the plasticizer. Such prepared

slurry was cast into sheet (anode support layer) by using a tape casting machine, then

dried in air to obtain the anode support (ϕ16×1.4 mm), to which functional NiO-YSZ

(60:40 wt. %, Type F Standard, Inco-TZ8Y, Tosoh) and electrolyte (TZ8Y, Tosoh)

were then screen printed in sequence, prior to sintering at 1390 °C for 3 h. To prepare

the baffle and cathode, Ce0.9Gd0.1O1.9 (GDC, NIMTE, CAS) and LSCF-GDC (70:30)

pastes were then screen-printed successively on the surface of the sintered YSZ

electrolyte, followed by sintering separately in air at 1300 °C and 950 °C for 2 h to

complete the fabrication of the anode-supported cells. The size of the obtained SOFC

button cell was about ϕ13×1 mm with an active area of 0.5 cm2 (cathode). The

NiCuO-Ce0.8Zr0.2O2 slurry was painted on the outer surface of anode and sintered at

900 ºC for 2 h in air.

Preparation of the LSCF-GDC cathode and NiCuO-ZDC catalyst slurry

The cathode and the catalyst pastes were prepared by ball milling the powder and

a self-made binder with a weight ratio of 50:50 for 1 h. The self-made binder was

prepared by adding 4 wt. % cellulose into 96 wt. % terpilenol, followed by stirring

and heating at 80 °C to completely dissolve the cellulose.

Catalytic activity evaluation

The catalytic activity and sulfur tolerance of Ni-YSZ (57:43, sintered at 1390 ºC)

and NiM (M= none, Co, Cu, Fe)-ZDC were compared. Catalytic activity

measurements for dry reforming of methane (DRM) reaction were performed at

atmospheric pressure using a compound of 0.2 g catalysts and 0.4 g

catalytically inactive quartz powder, which was sieved into the particle size ranging

from 30 to 60 mesh and packed on a bed of quartz tube. Prior to the catalytic

evaluation, the samples were heated up to 850 ºC and reduced in H2 for 5 h. As regard

the assessments of the sulfur tolerance, the reduced catalysts were exposed to H2-500

ppm H2S for 5 h prior to the test. The gas mixtures of sweet CH4-CO2 (mole ratio=1:1)

or sour CH4-CO2 (mole ratio=1:1, and balanced with 50 ppm H2S) were fed into the

reactor at the flow rate of 20 ml min-1. Compositional analysis of the effluent gases

was performed with a gas chromatography (GC, Hewlett Packard Series two). The

catalytic reactions were performed at the temperatures ranging from 550 to 800 ºC up

to 48 h. The percentages of CH4 conversion and CO selectivity were calculated

according to Eqs.1 and 2, separately.

Carbon deposition resistance evaluation

The carbon deposition resistance of the catalysts was evaluated by analyzing the

nature of the carbon deposited on the catalyst through Raman spectroscopy. To

accelerate the rate of carbon formation on the catalyst, we exposed the as-

reduced/treated catalysts to pure CH4 at 800 °C for 30 min and cooled them down to

room temperature in H2, followed by carrying out the Raman tests.

Other Characterizations

The X-ray diffraction (XRD) of as-synthesized and reduced powder was

identified by using Cu Kα radiation at a tube voltage of 40 kV and a tube current of

44 mA, within a 2θ range between 20° and 80° at 1 deg. min-1. The microstructures of

the samples were examined by using scanning electron microscopy (SEM, JEOL

6301F). Raman spectrometry (Thermo Nicolet Almega XR Raman Microscope) was

employed to detect the graphitization degree of deposited carbon on the catalyst. The

NiCu-ZDC was also analyzed by transmission electron microscopy (TEM, JEOL

2200 FS). A SDT-Q600 (TA instrument, USA) machine was used to carry out the

thermogravimetry analysis (TGA) experiments of the catalyst reduction process. The

carbon depositions on the catalysts were quantitatively investigated by using

temperature programmed oxidation [TPO via coupled TGA-Mass Spectrometer (MS,

Pfeiffer Vacuum GmbH)].

Results and discussion

Table S1 Crystallite sizes of NiM bimetallic alloys calculated by Scherrer

equation.

Ni NiCo NiCu NiFeCrystallite sizes (nm) 17.8 19.2 20.3 15.9

Calculated by using the (111) metallic Ni plane. The Scherrer equation size is similar to the TEM results.

Table S2. Comparison of the fuel cell performance

Anode Electrolyte/Cathod

e

Temperature (°C) Fuel composition PPD (W cm-2) Ref. No.

Ni-YSZ YSZ/LSCF-GDC 800 CH4-CO2 (1:1, 50 ppm H2S) 0.96 Present study

SDC-Ni-YSZ YSZ/LSM-SDC 800 Dry H2 0.525 1

(Ni0.75Fe0.25-MgO)-YSZ YSZ/LSM-YSZ 800 Wet CH4 (3% H2O) 0.648 2

3 wt % Ru-Al2O3-Ni-YSZ YSZ/LSM-YSZ 800 CH4-CO2 (2:1) 0.705 3

LiLaNi-Al2O3-Ni-ScSZ ScSZ/LSM-ScSZ 850 CH4-H2O (2:1) 0.532 4

GdNi-Al2O3-Ni-YSZ YSZ/LSM-YSZ 800 CH4-CO2 (2:1) ~0.8 5

4 wt % Au-Ni-GDC YSZ/LSM-YSZ 850 CH4-H2O (2.07:1) 0.41 6

Figure S1

Figure S1. XRD patterns of as-synthesized La0.6Sr0.4Co0.2Fe0.8O3-δ perovskite.

Compared to the JCPDS file <01-082-1961>, it is confirmed as-prepared powder

formed a simple perovskite.

Figure S2

Figure S2. XRD patterns of as-synthesized and reduced NiMO-Ce0.8Zr0.2O2.

The XRD results of the as-prepared and reduced catalysts show that 20 mol. % ZrO2

doped CeO2 (ZDC) was identical in both reducing and oxidizing atmospheres while

NiM(M=Cu, Co, Fe)O solid solution was fully reduced to Ni or Ni alloys in H2.

Figure S3

Figure S3. Temperature dependent weight changes of NiM-Ce0.8Zr0.2O2 catalysts

in 10 % H2.

The results show the effect of the alloying elements on the reduction temperatures and

rates of the catalysts. The NiCuO solid solution exhibits the lowest reduction

temperature and highest reduction rate.

Figure S4

Figure S4. (a) TEM bright field (BF) image; (b) EDX elemental mappings of Ni;

(c) and (d) High resolution TEM (HRTEM) images of NiCu bimetallic

nanoparticles at the corresponding sites labeled in BF image.

Figure S5

Figure S5. TEM BF image and EDX mappings of Ni, Co, Ce, Zr and O.

This EDX mapping confirms that the elements well disperse in the catalyst from a

relatively macro perspective comparing to Figs. S4 and 2c.

Figure S6

Figure S6. Raman spectra of fresh and H2S treated catalysts after exposing to

dry CH4 at 800 °C for 30 min.

The nature of the deposited carbon in the samples was analyzed by Raman

spectroscopy and the results are shown in Fig. S8. Two intense bands related to the

deposited carbon appeared in the test range, i.e., the D (defect) band associated with

disorder structure of carbon and the G (graphite) band featuring the graphitic layers

and the tangential vibration of carbon atoms. The intensity ratio of these two bands R

(= ID/IG) has been widely used to characterize the graphitization degree of carbon. A

higher value of the R corresponds to a lower degree of graphitization and an active

carbonaceous species that is readily removable. The results showed that NiCu alloy

had the best carbon deposition resistance, and H2S treatment improved the carbon

deposition resistance of all the samples.

Figure S7

Figure 7. Cross-sectional microstructures of the as-prepared cell and cell

components: (a) reduced cell; (c) LSCF-GDC cathode/GDC baffle/YSZ

electrolyte; (b) and (d) NiCuO-Ce0.8Zr0.2O2/Ni-YSZ supported layer.

As the microstructure of the reduced cell shown, porous anode is well adhered to

the dense YSZ electrolyte, which is approximately ~12 μm in thickness. The

reduction of NiO in the anode support and functional anode generates a significant

amount of pores to guarantee fuel gas permeability and three phase boundaries (TPBs)

for electrochemical reaction. The LSCF-GDC cathode, forming a well-sintered porous

structure, is also showed to establish adequate coherence to the anode/electrolyte

substrate. After sintering at 900 °C for 2 h, porous NiCuO-Ce0.8Zr0.2O2 catalyst layer

is integrated to NiO-YSZ supported layer, ensuring efficient electronic transportation.

Figure S8

Figure S8. I-V and I-P curves of C-SOFC and TA-SOFC in various fuels at 800

°C before H2S treatment.

Figure S9

Figure S9. I-V and I-P curves of C-SOFC and TA-SOFC in various fuels at 800

°C after H2S treatment.

Figure S10

Figure S10. Electrochemical impedance spectra of H2S treated TA-SOFC in H2

and CO under -0.5 V over potential.

The EIS are featured by flatten arcs, intersected by the real axis at high and low

frequencies. The intercept of high-frequency is estimated as the ohmic resistance (RΩ),

including electronic resistance of electrode, ionic resistance of the electrolyte and

interfacial contact resistance. The low frequency intercept corresponds to the overall

resistance (RO), while the difference between the RΩ and RO represents the electrode

interfacial polarization resistance (RP).

Figure S11

Figure S11. The current dependent exhaust gas composition of C-SOFC after

H2S treatment fed with sweet CH4-CO2.

Figure S12

Figure S12. I-V and I-P curves of C-SOFC and TA-SOFC in CH4-CO2 at

temperatures between 650 °C and 800 °C before H2S treatment.

Figure S13

Figure S13. I-V and I-P curves of H2S treated C-SOFC and TA-SOFC in CH4-

CO2 at temperatures between 650 °C and 800 °C.

Figure S14

Figure S14. The mass spectroscopic data from the effluent that obtained during

the temperature-programmed oxidation (TPO) of the catalyst after DRM test in

sour CH4-CO2.

A typical mass spectrogram, obtained from the TPO test, demonstrated that the

removal of carbon deposition is accompanied by the consumption of O2 and

generation of CO2. It is also known from the spectrogram that the process of carbon

oxidation started at the temperature around 400 °C.

References of the electronic supplementary information

1. Zhang, L. et al., Journal of Alloys and Compound. 2009, (482), 168.

2. Liu, Y. et al., Journal of Power Sources. 2011, (196), 9965.

3. Wang, W. et al., International Journal of Hydrogen Energy. 2011, (36), 755.

4. Wang, W. et al., Journal of Power Sources. 2011, (196), 90.

5. Wang, W. et al., Journal of Power Sources. 2011, (196), 3855.

6. Niakolas, D. K. et al., International Journal of Hydrogen Energy. 2010, (35), 7898.