第二離子質譜在SOFCs上的應用.pdf

Application of secondary ion mass spectrometry (SIMS) technique on the durability of solid oxide fuel cell (SOFC) materials K.Yamaji,N.Sakai,H.Kishimoto,T.Horita,M.E.Brito and H. Yokokawa National Institute of Advanced Industrial Science and Technology (AIST), Ibaraki, Japan 1 INTRODUCTION Solid oxide fuel cells (SOFCs) consist of a complicated layered structure of cell components, i.e., anode, electrolyte, cathode, interconnect, and surrounding components such as the current collector, gas distributor, gas sealant, etc. Each component is in contact with the neighboring components that consist of completely different elements. Furthermore, the cell and the surrounding components are continuously exposed to air and fuel flow at high temperatures. The electrolyte and interconnect are also exposed to a large gradient of oxygen potential from air (p(O 2 ) = 21.3kPa) to fuel (p(O 2 ) ～ 10 ?13 Pa in 3% humidified hydrogen at T = 1273K). Hence, during the research and development of the SOFC, the chemical stability and compatibility among the cell components (electrolytes, anodes, cathodes, and inter- connects) have been the most important issue to establish the long-term stability of performance of the cell. The possible degradation phenomena in SOFC compo- nent materials are roughly categorized into the following three aspects. 1. Degradation of a component caused by the inter- action with ambient gaseous air and fuels: Gener- ally, the candidate materials for SOFC components are chemically stable under the operating conditions. However, alloy interconnect is one of the exceptions. The alloy surface is oxidized in both air and fuel atmospheres during the SOFC operations. Hence the characteristic features of oxide scale formation should be always taken into account when one evaluates the durability and compatibility of an alloy as SOFC interconnects. 2. Degradation caused by reactions or interdiffusion of metal elements (cations in oxides) at interfaces: The electrode/electrolyte interface in SOFC is the most important part in power generation, since the reduc- tion of oxygen molecule to oxide ion and the oxidation of fuel proceed at the gas/electrode/electrolyte three- phase interface. Hence the electrode/electrolyte interface should be always clean to keep satisfactory catalytic activity for redox reaction and high electrical conductiv- ity. However, there are always steep gradients of metal (cation) concentration at the interfaces of the cell com- ponents. Hence, if an element (cation) in one cell com- ponent has a sufficiently high diffusivity, it may diffuse into the neighboring cell component across the inter- face according to the concentration difference (chemical potential difference). Furthermore, if recombination of elements between both cell components leads to a more stable compound, it may result in the formation of a secondary phase at the interface. Such metal interdif- fusion or secondary-phase formation at the interfaces may cause a decrease in the electronic conductivity or Handbook of Fuel Cells – Fundamentals, Technology and Applications. Edited by Wolf Vielstich, Hubert A. Gasteiger, Arnold Lamm and Harumi Yokokawa. ? 2010 John Wiley hence one can obtain information on elemental dis- tribution in the vicinity of the surface as a function of depth. If a well-focused beam (diameter ～0.5μm) is scanned on the surface, one can obtain the two-dimensional image of secondary ions, which may represent the lateral distribu- tion of the elements. By combining the depth profile and the images, it is possible to obtain the three-dimensional distribution of the elements. The features of SIMS analysis are shown and compared with those of other analyses in Table 1. SIMS has the following advantages: O 2 ? (for detection of metal) (for detection of nonmetal) Energy slits Secondary ion optics (magnetic fields and slits) Mass detection unit (electron multiplier Faraday cup, etc.) Elements liberated from etched surface Electrical field Sample Lenses Primary ion optics (lens and apertures) Primary ion beam source Cs ? Figure 1. Schematic view of dynamic secondary ion mass spectrometry (D-SIMS). Application of SIMS Technique on Durability of SOFC Materials 3 Ta b l e 1 . Comparison of analysis techniques a ssociated with the e leme ntal distribution across t he dissim ilar m aterials in SOFCs. Analysis technique SEM/EDX E PMA a (WDX) b TEM c /EDX d XPS e AES f GD-OES g D-SIMS Detectable elements B ( or Be) – B – B – Li – U B – U H – U H – U Detection lim it 0.3 – 1 % 0 .1% 1 atom ～ 0 . 1% 0.1% – ppb – p pm Sensitivity Light elem ents Insensitive I nsensitive I nsensitive S ensitive S ensitive S ensitive V ery s ensitive Heavy e lem e nts S ensitive S ensitive S ensitive S ensitive S ensitive S ensitive S ensitive Quantitative a nalysis 0 .5 – 5 % 0 .1 – 0 .2% 0 .5 – 5 % – ～ 10% Needs standard mat e ri al s ～ 10%; needs standard mat e ri al s Spatial resolution Line analysis 1 – 3 μ m 1 0–3 0 μ m ～ 1 n m – 30 nm (FE-AES) Impossible 1 – 3 μ m Depth p rofiling I mpossible I mpossible I mpossible S everal nanometers 3n m ～ 10 nm 1 – 1 0 n m I m a g e a n a l y s i s 1–3 μ m 1 0–3 0 μ m 1 nm Impossible I mpossible I mpossible 1 μ m Dif fi culty of sample preparation E asy E asy D if ficult E asy E asy E asy E asy Operation S im ple S im ple D if ficult D if ficult D if ficult S im ple C om plicated or dif fi cult a EPMA, e lectron p robe microanalyzer b WDX, wave disper sive X- r a y s pectr o m e ter c T E M , t r a nsm i ssion electr o n m icr o scope d EDX, ener gy dispersive X-ray s pectrometer e XPS, X- r a y photoelectr o n s pectr o scopy f AES, auger e lectron s pectroscopy g GD- OE S, glow dischar g e optical em ission spectr o m e tr y 4 Materials for high temperature fuel cells 1. High sensitivity for light elements: Mass spectrometry is very effective in detecting the light elements. It is in striking contrast with conventional X-ray analyses such as energy dispersive X-ray spectrometer (EDX), which has high sensitivity for heavy elements but is insensitive for light elements. 2. Very low (precise) detection limit: D-SIMS has high sensitivity for most elements. For example, the detection limits of boron or chromium are below the parts per million level (i.e., in parts per billion level). 3. Wide detection range: It is possible to detect secondary ion intensity from 1 to 10 6 counts/s with a secondary- electron multiplier, which indicates that an element can be detected from 1ppm to 100%, for instance. 4. Isotope analysis: Isotopes can be clearly separated and analyzed by SIMS, which makes it possible to measure the self-diffusion coefficient using stable isotopes. On the other hand, the following points should be noted as disadvantages of the SIMS analysis: 1. Mass interference effect: The secondary ions are filtered by mass/charge (m/e), and the mass resolution depends on the type of mass spectrometer. The sector-magnet- type SIMS has a mass resolution (M/Delta1M) of around 300–20000; however, the sensitivity is lowered if one sets higher mass resolution. The quadrupole-type SIMS has a low mass resolution of around 1. For example, a mass resolution of more than 10000 will be required to separate 40 Ca + and 40 Ar + ,soitisvery difficult using mass spectrometry. As well as single ions, there are more combinations if one considers molecular ions such as 2 D + and 1 H 2 + , 56 Fe + and 40 Ca 16 O + ,etc. SIMS is not a suitable method for identification of unknown materials because many different combinations can appear in one particular mass line. This mass interference becomes more significant with increasing mass number. This is the reason why SIMS is suitable for analysis of light elements. 2. Problem in quantitative analyses: The relationship between secondary ion intensity and concentration of an element i can be described by the following equation: N i = V × [M i ] × T × Y i (1) where N i is the number of the counted ion M i , V is the volume etched by the primary ion, T is the transmission coefficient of the mass spectrometer, Y i is the yield of the ion M i ,and[M i ] is the atomic concentration of the element M i . It should be noted that the parameters V,T,andY i change considerably with the analysis condition, and so it is very difficult to reproduce the same values in a series of analyses. V depends on the primary ion intensity, etched area, and the binding energy of the elements on the sample surface; T depends on the optical conditions including apertures and slits; and Y i depends mainly on the stability of secondary ions. Although the quantitative analysis of an element is possible by using a standard material that has a known, constant concentration as the element i,the analysis condition should be exactly the same between the standard material and the measured sample. Even if such a careful treatment is done, the uncertainty may reach up to 10%. 3. Complexity in operation: The optics of primary and secondary ions consists of many lens, apertures, and slits. It is difficult to keep the primary ion and secondary images well focused for a long time; therefore, operators must focus them every time before they start analyses. This causes poor reproducibility in quantitative analysis. Although D-SIMS has many disadvantages as an analysis technique, its advantages are very unique. By combining the conventional scanning electron microscopy and energy dispersive X-ray spectrometer (SEM/EDX) analyses, one can get more precise and new results for small concentration changes in the vicinity of surfaces and interfaces. The application of such a combined analysis technique will be very important to investigate the degradation mechanism of SOFCs. In the following sections, some examples of D-SIMS application are introduced. 2 DETERMINATION OF DIFFUSIVITY BY D-SIMS DEPTH PROFILING The formation of secondary phases or the interdiffusion of metallic elements at the interface of SOFC components is the most common degradation phenomena. They cause undesirable property changes, for example, the lowering of electrical conductivity or catalytic activity of electrodes. The transport flux (j) of a metallic element i can be roughly described by the following equation: j i =?D i dC i dz =?C i B i parenleftbigg 1 N dμ i dz parenrightbigg (2) where D i , C i , B i , μ i are diffusivity, concentration, mobil- ity, and chemical potential of the element i, respectively, z is the distance, and N is the Avogadro’s number. As shown in the equation, the element transport flux can be determined by the diffusivity and the concentration gradient. Normally, if the element i is contained in one cell component and not in others, the concentration gradient (dC i /dz) becomes extremely large. Hence the determination of the diffusivity (D i ) is very important to estimate to what extent interdiffusion has occurred across the interface. The diffusion-couple method is effective in investigat- ing both the reactivity and interdiffusivity of elements at Application of SIMS Technique on Durability of SOFC Materials 5 the interface of two different cell components. A diffu- sion couple is made by attaching the smooth surface of dense, polycrystalline bodies of the corresponding mate- rials. After annealing them in a certain combination of temperature and time, some reactions causing the forma- tion of secondary phase or interdiffusion occur, leading to changes in concentration profiles of the constituent elements in the vicinity of the interface. However, the determina- tion of the interdiffusion coefficient (D app ) may contain a large uncertainty because the boundary condition is changed by the formation of the secondary phase. Nevertheless, the apparent interdiffusion coefficients may be determined by fitting the concentration profile to the Fick’s diffu- sion equation solved under the condition of semi-infinite media. [15] C ? C 0 C s ? C 0 = erf ? ? ? z 2 radicalBig D app t ? ? ? (3) where C is the concentration, C s is the concentration at the surface, C 0 is the concentration of the background level (nearly zero), z is the distance from the interface, D app is the apparent diffusivity, and t is the annealing time. Hence it is important to obtain precise information on the concentration change as a function of the distance from the interface. In conventional SEM/EDX analyses, the concentration profile can be obtained to analyze the cross section of the interface so that the resolution of the analyses is in the submicrometer range. Under such analysis conditions, the concentration profile in more than several ten micrometers is required to evaluate the dif- fusivity. However, it should be noted that most SOFC components are metal oxides in which the diffusivity of metal components is quite small. For example, if one wants to obtain the concentration profile of an element that decreases to 10% of that at the interface at a point 50μm from the interface, an annealing time of more than 340000h is required for the case of D app = 10 ?18 m 2 s ?1 . It is a very time-consuming task to determine the diffusiv- ity of metal components in oxides by using conventional analyses techniques; this is the reason why this type of investigation has not been extensively carried out, despite the fact that the importance of the diffusivity data has been well recognized in high-temperature materials science. The extremely high resolution in D-SIMS depth profiling therefore becomes very attractive. The resolution of depth profiling is typically around 10nm when the primary ion accelerating energy is 10kV; hence, this makes it possi- ble to obtain enough data in the concentration profile even with 1μm. If one wants to obtain the concentration profile that decreases to 10% within 1μm from the interface, the required annealing time can be shortened to only 130h at D app = 10 ?18 m 2 s ?1 . Hence the practical use of D-SIMS depth profiling will greatly contribute to rapid data acquisition for the determination of diffusivity. 3 APPLICATION OF THE DEPTH PROFILING TECHNIQUE 3.1 Diffusion at the cathode/electrolyte interface Interdiffusion of metallic components has not been very seriously considered for the combination of yttria stabilized zirconia (YSZ) electrolyte and lanthanum strontium man- ganese oxide (LSM) cathode, which has been widely used in SOFCs operating at high temperatures. The formation of La 2 Zr 2 O 7 was more serious and many attempts were made to suppress it. [16] Application of manganese-rich LSM was effective in suppressing the reaction, and it also accom- panies the manganese diffusion via the grain boundary of YSZ. [17] Manganese diffusion may enhance the electronic conductivity in YSZ and lower the ionic transport number; however, the effect on degradation was less important. [18, 19] However, for the intermediate-operating SOFCs using LSGM electrolytes, significant diffusion of transition-metal ions from cathode materials to LSGM electrolytes