Surface analysis of coprecipitated binary Ni(II)-Fe(II1) hydroxide solid solutions using Auger electron spectroscopy and depth profiling has revealed ...
Journal ofElectron Spectroscopyand RelatedPhenomena, 68 (1994) 597-604 0368-2048/94/$07.00 @ 1994 - Elsevier Science B.V. All rights reserved
597
AUGER ELECTRON SPECTROSCOPIC STUDY OF BINARY NICKEL(II)-IRON(II1) HYDROXIDES Alexander A. Kamnev *a and Alexander A. Smekhnovb &Laboratory of Structural Research Methods, Institute of Biochemistry and Physiology of Plants and Microorganisms, Russian Academy of Sciences, 410015 Saratov, Russia bInstitute of Superhard Materials, Academy of Sciences of Ukraine, 252153 Kiev, Ukraine Surface analysis of coprecipitated binary Ni(II)-Fe(II1) hydroxide solid solutions using Auger electron spectroscopy and depth profiling has revealed redistribution of the components (nickel and iron) within the surface layers showing significantly increased iron-to-nickel ratios, as compared to the corresponding gross averaged values in the bulk. The results are compared with transmission Mossbauer spectroscopic data obtained earlier for these systems, which are interpreted terms of the presence of surface-related iron(III)in containing forms featured by essentially higher quadrupole splitting values. The surface analysis data presented are also discussed in connection with the high electrocatalytic activity of composite nickel-iron hydroxide systems towards anodic oxygen evolution in alkaline electrolytes revealed previously. 1.
INTRODUCTION
In materials analysis, a significant role is played by the characterization of surfaces and interfaces [1,2], the structure and composition of which are responsible for the properties and physicochemical behaviour of adsorbents, catalysts, electrode materials, etc. Various spectroscopic techniques have
been widely applied to surface analysis [3;4] owing to their sensitivity and convenience. Composite oxide/ hydroxide systems based on transition metals are of particular importance for the above applications [5-lo] because of a wide-range variability of their properties. Recently, we have reported the results of transmission MUssbauer spectroscopic stu-
* Author to whom correspondence should be addressed. SSDI 0368-2048(94)02163-T
598
dies of various composite nickel-iron hydroxide systems concerning their behaviour 'in alkaline media [ll], adsorbability of ferric hydroxo complexes at nickel hydroxides [12] and their strong electrocatalytic effect towards anodic oxygen evolution in alkali [13]-. These investigations have revealed- a clearly exhibited tlsurfacett ferric hydroxo effect of species on the anodic oxygen evolution kinetics at the nickel hydroxide electrode, in full agreement with some electrochemical data reported earlier [5,14]. Note that binary coprecipitated nickelalso iron hydroxides have been shown to be excellent oxygen evolution electrocatalysts [5,6,15]. Our preliminary Mbssbauer study of such phases [ll] showed binary to their resonant spectra consist, most probably, of a superposition of quadrupolefeaturing doublets split different Fe(III)-containing forms, ’ which has recently been confirmed in the course of a more detailed Mbssbauer investigation [16]. In the present conununication, we report the results of Auger electron spectroscopic (AES) analyses of the surface layers of coprecipitated Ni(II)-Fe(II1) binary hydroxides with average bulk nickel-to-iron ratios corresponding to the highest electactivity [63. rocatalytical
The results obtained are also compared with our previous transmission MUssbauer data [11,16] on the'binary nickeliron hydroxide systems.
2. EXPERIMENTAL 2.1. Preparation of binary hydroxides Binary Ni(II)-Fe(III) hydroxide systems with gross Ni-to-Fe molar ratios 4:l and 9:l (i.e. 20 and 10 at.% Fe/ (Ni+Fe), respectively) were synthesized as described in Ref.[l7] by coprecipitation from mixed nickel(I1) and iron(II1) sulphate solutions of appropriate compositions (1 M total Ni+Fe concentration) in a slight excess of hot NaOH solution. The precipreliminarily pitates were washed, partially dried, washed again with doubly distilled water until negative test for sulphate and hydroxide (pH 7 to 8) ions and completely dried at 393K. All reagents used were of not less than analytical reagent grade. 2.2. AES analysis The resulting samples of binary hydroxide powders were pressed into In carefully metal platelets and attached to the Auger spectrometer were spectra holder. AES recorded using an LAS-3000 AES-2000 an complex with (Riber, Auger spectrometer France). The primary electron beam energy for AES analysis was 3 keV (1=0.0005 mA). The starting pressure of residual gases in the AES analysis chamber was 5 nPa. The elements content was using elemental calculated sensitivity factors [18].
599
2.3. Depth profiling
3. RESULTS AND DISCUSSION
Depth profiles of elements distribution were obtained by means of surface sputtering with an argon ion beam (3 keV; 1=0.05 mA), after which an Auger spectrum was recorded, the cycles of surface sputtering and spectrum registration being alternated.
In Auger spectra of the samples studied, the following four lines were detected corresponding to the Auger transitions: nickel LMM (E=848 eV); iron LMM (E=703 eV); oxygen KLL (E=510 eV); carbon KLL (E=281 eV). Typical Auger spectra for the samples with 20 and 10 at.% Fe/(Ni+Fe) after 2.5 and 2 min of sputtering, respectively, are shown in Fig.1.
except All experiments, where otherwise indicated, were performed at ambient temperature.
dN/dE, &
b’ f r
a
r Y
u
n i t
s
E, eV Figure 1. Typical Auger electron spectra of binary Ni(II)-. Fe(II1) hydroxides with 20 (1) and 10 at.% Fe/(Ni+Fe) (2) after 2.5 and 2 min of sputtering, respectively.
600
Table 1 AES analysis data for coprecipitated Ni(II)-Fe(II1) hydroxides Gross sputNi:Fe tering atomic time, ratio S 4:1 4:l 4:l 9:l
o* 50 150 120
Elements content, at.% ____________-__-____--_-Iron Nickel Oxygen Carbon
Fe/(Ni+Fe), at.% -__-_---___---__-_ Bulk Surface &v-S)
20 19 26 15
30 22 21 30
35 43 49 35
15 16 4 20
20 20 20 10
40 46 55 33
* The original surface (without sputtering). The percentage values for the elements calculated from the spectra are presented in Table 1; .Since, for the maximum of 2.5 min of sputtering time, the sputtering rate for metal hydroxide materials [l] enables one to reach the maximal sputtering depth of 2 to 3 nm, the data in Table 1 represent the elements content at the surface and in the near-surface region (i.e. within several atomic layers). It should also be noted that very close values of sputtering coefficients for pure Fe- and Ni-containing materials [1,19] allow the process of surface sputtering to be regarded as proceeding uniformly, which is corroborated by relatively constant concentration sums of the cations. (Fe+Ni) both at the original surface and after different sputtering periods (see Table 1). Carbon detected in the surface and near-
surface layers of both samples (see Fig.l, Table 1) may be due both to cracked residual gas components in the AES analysis chamber and to traces of carbonates included in the samples during precipitation in alkaline solutions containing a small carbonate impurity. The main feature which attracts attention when considering the AES data presented in Table 1 is the significantly increased surface Fe-toNi ratio values for both samples, as compared to the coraveraged responding gross bulk values. It may also be noted that in going from the original surface to near-surface regions reached after sputtering cycles, the Fe/ (Ni+Fe) values tend to increase (see Table 1). Thus the AES data clearly reveal redistribution of the components (nickel and iron) the surface layers within with a marked enrichment of
601
the latter with iron as compared to the bulk, which is illustrated by Fig. 2. The
60
/
P
/' /
40
20
,’
0 0
40
20
60
[Fd(Ni+Fe]bUlk, a-L%
Figure 2. Correlation between Fe percentage in the surface layers of coprecipitated Ni(II)-Fe(II1) hydroxides and that in the bulk (gross avevalues). raged Sputtering time 2 min. redistribution process should evidently be regarded as occurring in the course of precipitation of the binary systems owing, most probably, to differences in the hydrolytic precipitation kinetics of these cations, as well as to the well-known capability of ferric hydroxo complexes to
form metastable polynuclear species [ll]. Note that for Cu-Co binary oxide catalysts obtained by calcination of coprecipitated hydroxonitrates at 723 K for 24 h, the redistribution process resulting in the surface enrichment with Co has been shown to proceed during calcination in air, while calcination of the samples in nitrogen resulted in a uniform distribution of the cations both at the surface and in the bulk [201= As for the Ni-Fe hydroxide systems studied here, the non-uniform distribution of the cations between the surface and the bulk might, in our opinion, facilitate the formation of heterogeneous oxide mixtures during their calcination [21]. A similar redistribution of metal cations with the minor component (3% and less) accumulating at the crystallites surface was assumed in particular cases for Ni-Fe and Co-Fe coprecipitated hydroxide systems [22] on the basis of their adsorptional properties and comparison with the corresponding behaviour of individual hydroxides, as studied by IR spectroscopy and gas chromatography techniques. According to X-ray and IR spectroscopic analyses El71 r the coprecipitated binary systems studied in the present work are homogeneous solid solutions having the alpha-type structure [23], which is characteristic of binary Ni( II)-Fe(II1) hydroxides within a relatively wide
602
range of Ni-to-Fe ratios (from ca. 10 up to at least 40 at.% Fe/(Ni+Fe) [171) This implies that surface redistribution of the components occurs within the same structure in a relatively near-surface region, thin which may be correlated with the appearance of additional resonant doublets in their Mdssbauer spectra featured by markedly' higher quadrupole splitting values [16]. In this case, the latter M&asbauer spectral components should generally be expected to diminish in intensity if the binary system is treated in, e.g., a highly alkaline medium owing to dissolution the (at least partial) of ferric component primarily from the surface and nearsurface layers of crystallites, in contrast to nickel hydroxides which are practically insoluble in alkaline media [24, 251. Our recent which data on the changes were detected in the MCSssbauer spectrum of binary Ni-Fe hydroxide after its anodic treatment in 8 M KOH, as compared to, that of the initial sample (see Fig.4 in Ref. [111)1 seem to corroborate the above conclusion. It is noteworthy that the results recently reported by Grabke et al. [26] showed Fe impurities in nickel metal to concentrate at the surface of the nickel oxide layer, forming in the course of thermal oxidation, owing to the lloutiron diffusion of ward" cations. l
Considering the AES analysis data presented in Table 1, it should also be noted that the total Fe+Ni concentrations, which were shown above to be relatively conthe stant in going from initial surface to somewhat deeper near-surface regions, slightly exceed those expected for the lVbulk-likelVhydroxides and approach the nomivalues for the nal bulk This corresponding oxides. might probably be ascribed to some partial surface dehydration under the conditions of the AES analyses and surface sputtering. As was indicated above, hydroxide composite Ni-Fe systems are among the best anodic oxygen evolution ele[5,6,15], the ctrocatalysts ferric component being responsible for this surfacerelated electrocatalytic effect [10,13-15). For example, the latter effect for layerby-layer precipitated "sandwich" electrodes was found [5] to disappear when an iron hydroxide layer was covered by a nickel hydroxide one, whereas it was clearly pronounced if the upper layer consisted of either iron or nickel-iron coprecipitated hydroxides. It has recently been shown that anodic oxygen for evolution overpotential Fe-containing nickel hydroxide electrodes decreases with the increasing fraction of the electrode surface covered (i.e. species with iron surface coverage) [6,13,14]. In view of this, the marked
603
surface enrichment of the layers of binary Ni-Fe the iron hydroxides with component revealed in the present work may well contribute to the electrocatalytic effect owing to a correspondiminished anodic dingly oxygen evolution overpotential. This would obviously be more essential for electrolyte solutions with a lower alkalinity (e.g., with pH
