铬酸钇YCrO3固体氧化物燃料电池阳极材料的合成和性能初研究_图文

1、NOTICE WARNING CONCERNING COPYRIGHT RESTRICTIONSThe copyright law of the United States Title 17, United StatesCode governs the making of photocopies or other reproductions of copyrighted material. Under certain conditions specified in the law, libraries and archives are authorized to furnish a photo

2、copy or other reproduction. One of these specified conditions is that the reproduction is not to be used for any purpose other than private study, scholarship, or research. If a user makes a request for, or later uses, a photocopy or reproduction for purposes in excess of "fair use," that

3、use may be liable for copyright infringement. This institution reserves the right to refuse to accept a copying order if, in its judgement, fullfillment of the order would involve violation of copyright law. No further reproduction and distribution of this copy is permitted by transmission or any ot

4、her means.Rapid #: -9166028CROSS REF ID:LENDER:BORROWER:TYPE:JOURNAL TITLE:638236NED : Snell LibraryWVU : Evansdale LibraryArticle CC:CCGECS transactionsUSER JOURNAL TITLE:ECS transactionsARTICLE TITLE:ARTICLE AUTHOR:VOLUME:ISSUE:MONTH:YEAR:PAGES:ISSN:OCLC #:20131479-14891938-5862H2 Oxidation on Dop

5、ed Yttrium Chromites Anode of Solid Oxide Fuel CellLi, W.571Processed by RapidX:4/8/2015 8:19:21 AMThis material may be protected by copyright law (Title 17 U.S. CodeECS Transactions, 57 (1 1479-1489 (201310.1149/05701.1479ecst ©The Electrochemical SocietyH 2 Oxidation on Doped Yttrium Chromite

6、s Anode of Solid Oxide Fuel CellWenyuan Li, Mingyang Gong, and Xingbo LiuMechanical & Aerospace Engineering Department, West Virginia University,Morgantown, WV 26505, USADoped yttrium chromites as potential anodes for SOFC are studiedwith respect to the electrode performance and anode reactionme

7、chanisms. Both electrical conductivity and electrodeperformance of doped yttrium chromites have been enhanced afterNi doping. Electrochemical impedance spectra results indicate thatcharge transfer process at high frequency and surfaceadsorption/diffusion processes at low frequency domain can be thed

8、ominant anode reaction steps. Ni doping accelerates the surfaceprocesses and improves the charge transfer process probably byincreasing the amount of adsorbed H on the electrode surface. Amodel of H2 oxidation reaction is proposed, revealing thisdependence stems from the reaction between adsorbed H

9、and thelattice oxygen.IntroductionIn the state-of-the-art SOFC, Ni/YSZ cermet is the most commonly used anode. However, it suffers performance degradation when feeding hydrocarbon and impurity containing syngas due to coking and poisoning by contaminants such as S (1. To overcome this drawback, cera

10、mic anodes including oxides with perovskite structure such as doped lanthanum chromites (2-5 and lanthanum doped strontium titanate have been proposed as potential anode alternatives (6-9. YCrO3 based materials have been evaluated by Yoon et al. as a potential anode after doping with Ca and Co at th

11、e A and B sites separately. Compared with other ceramic anodes, YCrO3 shows several advantages: 1 its coefficient of thermal expansion (CTE is close to YSZ, which is more compatible with the Ni/YSZ system; 2 the electro-catalytic properties of YCrO3 can be improved by B-site doping; and 3 unlike som

12、e anodes such as (La, SrVO3, it can be manufactured in the air. Preliminary results have shown promising performance and good S-tolerance of YCrO3-type anodes (10, but to the best of our knowledge, there have not been any systematic investigations on anode reaction mechanisms reported so far.Explora

13、tion of the electrode reaction mechanism is crucial to guiding the design of materials to improve electrode performance. Although mechanisms and kinetics research based on Pt and Ni anodes have been extensively investigated, the same for perovskite based anodes is scarce (11-19. To this end, Co and

14、Ni doped YCrO3-YSZ composite anodes in this work were developed and characterized in H2-containing atmospheres by electrochemical impedance spectra (EIS. The effect of doping on catalytic activity and anode performance was evaluated. The rate limiting steps and H2 dependence of polarization resistan

15、ce associated with different dopants were determined. Finally, a model concerning anode reaction mechanism was proposed based on these results.ExperimentalY 0.8Ca 0.2CrO 3 (YCC, Y0.8Ca 0.2Cr 0.8Co 0.2O 3 (YCCC, and Y0.8Ca 0.2Cr 0.9Ni 0.1O 3(YCCN were synthesized by EDTA-citric sol-gel method detaile

16、d elsewhere (20-21. After dissolving standard nitrates (Alfa Aesar in stoichiometric percentage along with citric acid (Alfa Aesar and dissolving EDTA powders (Fisher Scientific into two beakers of distilled water, they were blended together followed by adjusting the pH to 8 via ammonia water (Alfa

17、Aesar, then held at 80 o C and stirred until gelation. The gel was heated to 400 o C and the resultant powders were calcined at 1200 o C for several cycles with intermediate grinding between cycles. X-ray diffraction (XRD, Bruker AXS, Cu K radiation tests were conducted to examine the purity of the

18、calcined powder. Commercial software Jade 5 was used to analyze the XRD spectra. Powders were also pressed into pellets for scanning electron microscopy (SEM, JEOL JSM7600F observation.The anode slurry was made by grinding sintered powders with YSZ in ink vehicle. The weight ratio of the anode powde

19、rs: YSZ: vehicle was 4:6:11. Electrodes were made by screen printing the anode slurry onto both sides of the YSZ electrolyte symmetrically. YSZ pellets (Nextech Co., 28 mm in diameter and 350 m in thickness, were used as the electrolyte. The active electrode area was 0.7 cm2 at each side. The as-mad

20、e symmetric cells were sintered at 1000 o C for 2 h in air. Pt mesh was used as current collector.EIS was employed to characterize these symmetric cells in various H2-containing atmospheres by Solartron 1287 electrochemical interface and 1260 impedance analyzer at open circuit condition (OCV over th

21、e frequency range from 0.1 Hz to 1MHz. The AC signal applied was 20 mV. H2 content in the H2/N2 mixture was controlled by mass flow controllers (Alicat Scientific. A water bubbler at room temperature was used to humidify mixed gases before feeding them to the samples. The resultant spectra were deco

22、nvoluted using Z-view software.Results and Discussion It has been reported by K. J. Yoon etc. that a pure phase of Ca and Co co-doped YCrO 3 was obtained by glycine-nitrate method at 1200 o C (10. Fig. 1 shows the XRD spectra from the powder samples of pristine and doped YCCs made by sol-gel method

23、in this paper. All peaks correspond to a single phase, showing orthorhombic perovskite structure (PDF#48-0474. The indexes of main peaks have been identified.( 2 11(002(120 (022 (122 (111I n t e n s i t y (a . u . (311 (112 (131 (213 ( 2 (331 (422 YCCN YCCCYCC203040506070802T (degFigure 1. XRD spect

24、ra of doped YCCsFigure 2. SEM cross-sectional view of YCCN electrode sintered at 1000 o C (a; morphology of YCCC (b, YCC (c, and YCCN (d sintered at 1200 o C.ECS Transactions, 57 (1 1479-1489 (2013The microstructure and morphology of these powders and the YCCC electrode were examined by SEM. It is i

25、ndicated in Fig. 2a that the microstructure is uniform and there is enough porosity for gas transportation in the YCCC electrode sintered at 1000 o C. The adhesion between electrode and YSZ electrolyte is good and no delamination at the interface was observed. Shown in Figs. 2b to 2d, the grain size

26、 sintered at 1200 o C is 0.20.5 m for YCC, 11.5 m for YCCC and 0.20.5 m for YCCN respectively. (a-Z ' ' (: c m 2 2Z' (: cm Figure 3. Impedance spectra of YCCs at OCV tested in wet 5% H2N2 at 850 o C (a and the equivalent circuit used to fit EIS (b.Fig. 3a shows the EIS for YCCs tested at

27、 850 o C in wet 5% H2-N 2 (5% H2 and 95% N 2. The scattered symbols are measured data and the solid lines are the corresponding fitting arcs from Z-view program. Fig. 3b is the equivalent circuit used to fit the spectra, where L represents the overall inductance from the lead wires and the Solartron

28、 system, R 0 the total ohmic resistance composed of contributions from electrolyte, electrodes and lead wires. A constant phase element (CPE is adopted due to the frequency dispersion phenomenon in the electrode process (22. As can be seen in Fig. 3a, the chosen equivalent circuit demonstrated a goo

29、d representation of the observed results. Two main arcs are distinguishable in these spectra, which are denoted as high frequency (HF and low frequency (LF arcs separately in the frequency domain. All of these spectra possess roughly similar shapes, indicating that the main electrode processes remai

30、n the same after doping even if the performance of YCC has been enhanced significantly. (al n (T /R 1000/T (1/K (bl n (T /R 1000/T (1/KFigure 4. Arrhenius curves for HF arc (a and LF arc (b derived from EIS of YCCs in wet 5% H2N2 from 650850 o CFig. 4 demonstrates the Arrhenius curves for HF and LF

31、arcs derived from EIS of YCCs in 5% wet H2. It is obvious that resistances from different samples decreased in the order R YCC > R YCCC > R YCCN for both HF and LF arcs. The apparent activation energy (E a derived from Fig. 4a for the HF arc is almost the same, with values of 1.2, 1.3 and 1.2

32、eV for YCC, YCCC and YCCN, respectively. In contrast, for LF arcs, E a varies significantly for different materials, with 1.2, 1.0 and 0.5 eV for YCC, YCCC and YCCN, respectively. Table I summarizes the resistance (R , capacitance (C , activation energy (E a and characteristic frequency (f 0 from EI

33、S collected in wet 5% H2 at 850 o C. After comparing the f 0 and C identified in this study to those reported in the literature for SOFC electrodes (23, 24, it is safe to assign the HF arc and LF arc for YCCs in wet 5% H2 to charge transfer and hydrogen adsorption/diffusion processes, respectively.T

34、able I. Summary of resistance (R , capacitance (C , activation energy (E a and characteristic frequency (f 0 for YCCs tested in wet 5% H2-N 2 at 850 o C.In wet 5% H2-N 2at 850 o CR (cm 2C (F Ea (eV f o (HzHF arcLF arcYCC 157 2.0h 10-7YCCC 44 1.4h 10-7YCCN 24 2.3h 10-7YCC 43 0.7h 10-3YCCC 26 1.6h 10-

35、3YCCN 18 3.8h 10-31.2 48001.2 25000 1.3 27000 1.2 1 1.0 4 0.5 2The performance of YCC was significantly enhanced upon Ni and Co doping as shown in Fig. 3a. For this charge transfer process at HF arc region, the reaction rate is controlled by E a and the amount of available reactants. Given that all

36、three materials show similar E a at HF arc, the amount of available reactants on the YCCN and YCCC electrode can be expected to be higher for the improved charge transfer reaction. Although the ionic conductivity of YCCs is not available so far, it must possess oxygen ion conductivity in some extent

37、 especially in a reducing atmosphere (25. The oxygen non-stoichiometry of YCC was reported to be 0.06 under 10-18 atm P O2 at 1000 o C (26. The doping of Ni and Co should increase this value due to the fact that perovskite oxides with Ni or Co as B site anion are much easier to reduce than those wit

38、h Cr at B site (27. Therefore, YCCs can be viewed as mixed ionic and electronic conductor (MIEC in the anode atmosphere. In this case, the electrode reaction can take place not only at three phase boundaries (3PBs but also at the surface of YCCs grains. The reducibility introduced by Ni and Co dopin

39、g make it easier to activate lattice O in YCCs to react with H adsorbed on the surface. The higher ionic conductivity of YCCC and YCCN would also make the supply of O from YSZ faster. Both factors above would accelerate the charge transfer reaction in 2 phase boundaries (2PBs where the gas channel a

40、nd MIEC meet. From the viewpoint of reactant H, the coverage of adsorbed H may be increased by doping. In the LF arc, YCCN displayed the best performance. The lowest E a of LF arc implies the adsorption/diffusion process is thermodynamically more favorable on the YCCN sample. High stick coefficient

41、of metal Ni for H adsorption has been reported by Morgensen et al. (12. Good H adsorption/diffusion capability of the Ni-based anode in SOFC is also well-known. Based on this similarity, there might be some correlation between YCCN and metal Ni at this point.EIS testing was carried out in a H2 parti

42、al pressure range from 10-1.5 to 1 atm on YCCN and YCCC electrodes (not the same samples as used in Fig. 3. The double logarithm plots in Fig. 5 show the P H2 dependence of HF arc in different temperatures.ECS Transactions, 57 (1 1479-1489 (2013The reaction order n in the relationship of 1/R P n H2

43、is the slope of each curve, ranging from 1/3 to 1/2 as marked in Fig. 5a for YCCN and around 1/4 for YCCC. (a-2.0-2.4 -2.8-3.2-3.6-4.0l o g (1/R log (PH2/atm(bl o g (1/R log(P H2/atmFigure 5. H2 dependence of HF arc for YCCN (a and YCCC (b in different temperaturesECS Transactions, 57 (1 1479-1489 (

44、2013Partial pressure dependence of anode reaction has been extensively researched in the Ni-YSZ system (23-24, 28-31. Most of the results in these references show that the polarization resistance of Ni-YSZ electrode is insensitive to P H2, but strongly dependent on P H2O . According to observations

45、by Mogensen et al. (12, the sticking coefficient of H on Ni increases with temperature, and was about 0.3 at SOFC operating temperature even for the most unfavorable crystal face, rendering a high H coverage on Ni surface. If the coverage of adsorbed H on the electrode surface is high enough, such a

46、 process would not be affected by H2 molecule content, especially when running EIS at OCV condition. Nonetheless, with lower H coverage, the performance of the YCCs electrode would be harmed by limitation of the H supply when decreasing P H2. In addition, both the ionic conductivity and activation o

47、f lattice oxygen are closely related to the content of H2, which play important roles in the charge transfer step as discussed above. Hence, it is not surprising that such electrodes showed strong H2 dependence.The overall H2 oxidation reaction at the anode can be represented in Eq. 1,&H 2 O o l

48、H 2O+VO 2e 1Base on the observed results, one possible model for the reaction path on YCCsanodes in this study can be presented as following:Step 1 H 2l2H ads, YCCN 2 Step 2 H ads, YCCN lH ads, 3PB 3 -Step 3 H ads, 3PB O O lV O OH e 4Step 4 H ads, 3PB OH lH 2O e 5After taking the following assumptio

49、ns: 1 these reactions are microscopicallyreversible; 2 adsorption of any species is governed by Langmuir type isotherm; 3 the surface coverage for all adsorbed species is low, the corresponding rate equations can be written as:2r 1 k 1P H2 k 1 a Hads 6-r 2 k 2a Hads k 2a H3PB 7r 3 k 3a H3PB exp(EfE

50、k 3 a OH3PB exp( DfE 8r 4 k 4a H3PB a OH3PB exp(EfE k 4P H2O exp( DfE 9where k i and k i - are the rate constants for the corresponding anodic and cathodic reaction;f =F /RT (F , R and T have their usual physical meaning and E the electrode potential. and are symmetric coefficients. The activity of

51、lattice oxygen and oxygen vacancy is taken as constant regardless of the change of P H2.ECS Transactions, 57 (1 1479-1489 (2013 We first assume that the charge transfer step (step 3 is the rate-determination-step (RDS and the other steps are in near equilibrium. It follows: a Hads k1 PH2 k1 10 a H3P

52、B k 2 a Hads k 2 k4 PH2O k 4 a H3PB exp( fE 11 a OH3PB 12 Substituting of Eq. 10 - 12 to Eq. 8, it follows: r3 k 4 k 2 k3 k 4 k 2 k1 PH2 k 4 k2 P H2O exp( E fE k3 k4 k2 k1 k1 exp( (1 D fE k1 PH2 13 Eq. 13 indicates the steady state reaction rate of step 3, which can be converted to current density u

53、sing the relationship of I = nFr. In equilibrium condition there is no net current, i.e. r3=0, it reveals the Nernst potential: E const RT pH2O ln 2F pH2 14 Inserting Eq. 14 back into Eq. 13 yields the relationship of io,3 PH21/4 take =0.5. PH2O at OCV is almost fixed, 0.03 atm, in the present exper

54、iment. Similarly, if assuming the second charge transfer step (step 4 the rds, io,4 PH21/4 can be drawn also. Thus, a 1/4 reaction order is predicted at high frequency domain. As discussed above, doped YCCs possess some level of ionic conductivity at very low PH2 condition. They are MIEC when workin

55、g as anodes, and the charge transfer process (step 3 and 4 not only take place at 3PB area but also occur at 2PB area as O ions can be transported through the bulk of YCCs. The activity of oxygen vacancy in step 3 for 2PB reaction is no longer constant, but dependent to PH2. Furthermore, besides H a

56、dsorption, O and H2O can be adsorbed on electrode surface as well, leading to a complex competitive adsorption model as studies by Ihara and Mizusaki (18, 32. Given all these issues above, the reaction path in the anode may be more complicated than that presented in this model, and the H2 dependence

57、 would be more intricate than derived in this study as well. However, to further justify this model and verify the dominant reaction mechanism, more experiments are needed, such as EIS study under polarization conditions, the effect of H2O on polarization resistance and etc., which are the subjects

58、of upcoming investigations. Downloaded on 2015-04-08 to IP 24 address. Redistribution subject to ECS terms of use (see /site/terms_use unless CC License in place (see abstract. 1487 ECS Transactions, 57 (1 1479-1489 (2013 Conclusions In this study, doped YCCs were made and invest

59、igated by EIS. The effect of Ni and Co doping on electrode performance was discussed. A possible anode reaction model was proposed based on H2 dependence testing. It was found that charge transfer and surface adsorption/diffusion processes could dominate the spectra at high and low frequency domain, respectively. The electrode resistance of