Composition of the first coordination sphere of Ni2 + in concentrated aqueous NiBr2 solutions by xray diffraction M. Magini, M. de Moraes, G. Licheri, and G. Piccaluga Citation: The Journal of Chemical Physics 83, 5797 (1985); doi: 10.1063/1.449659 View online: http://dx.doi.org/10.1063/1.449659 View Table of Contents: http://scitation.aip.org/content/aip/journal/jcp/83/11?ver=pdfcov Published by the AIP Publishing Articles you may be interested in The coordination of Ni2+ in aqueous solution at elevated temperature and pressure J. Chem. Phys. 104, 2036 (1996); 10.1063/1.470960 Ni–Cl bonding in concentrated Ni(II) aqueous solutions at high Cl /Ni2+ ratios. An xray diffraction investigation J. Chem. Phys. 76, 1116 (1982); 10.1063/1.443079 Ni2+ coordination in aqueous NiCl2 solutions: Study of the extended xray absorption fine structure J. Chem. Phys. 71, 2381 (1979); 10.1063/1.438643 Xray diffraction study of the average solute species in CaCl2 aqueous solutions J. Chem. Phys. 64, 2437 (1976); 10.1063/1.432534 Xray diffraction study of CaBr2 aqueous solutions J. Chem. Phys. 63, 4412 (1975); 10.1063/1.431159 This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP: 200.145.174.189 On: Wed, 26 Mar 2014 19:41:11 http://scitation.aip.org/content/aip/journal/jcp?ver=pdfcov http://oasc12039.247realmedia.com/RealMedia/ads/click_lx.ads/www.aip.org/pt/adcenter/pdfcover_test/L-37/586982248/x01/AIP-PT/JCP_CoverPg_101613/aipToCAlerts_Large.png/5532386d4f314a53757a6b4144615953?x http://scitation.aip.org/search?value1=M.+Magini&option1=author http://scitation.aip.org/search?value1=M.+de+Moraes&option1=author http://scitation.aip.org/search?value1=G.+Licheri&option1=author http://scitation.aip.org/search?value1=G.+Piccaluga&option1=author http://scitation.aip.org/content/aip/journal/jcp?ver=pdfcov http://dx.doi.org/10.1063/1.449659 http://scitation.aip.org/content/aip/journal/jcp/83/11?ver=pdfcov http://scitation.aip.org/content/aip?ver=pdfcov http://scitation.aip.org/content/aip/journal/jcp/104/5/10.1063/1.470960?ver=pdfcov http://scitation.aip.org/content/aip/journal/jcp/76/2/10.1063/1.443079?ver=pdfcov http://scitation.aip.org/content/aip/journal/jcp/76/2/10.1063/1.443079?ver=pdfcov http://scitation.aip.org/content/aip/journal/jcp/71/6/10.1063/1.438643?ver=pdfcov http://scitation.aip.org/content/aip/journal/jcp/64/6/10.1063/1.432534?ver=pdfcov http://scitation.aip.org/content/aip/journal/jcp/63/10/10.1063/1.431159?ver=pdfcov Composition of the first coordination sphere of Ni2+ in concentrated aqueous NiBr2 solutions by x-ray diffraction M. Magini and M. de Moraesa) Divisione Chimica/TIB, ENEA, CRE-Casaccia, Rome, Italy G. Licheri and G. Piccaluga Dipartimento di Scienze Chimiche, Universita, Via Ospedale 72,09100 Cagliari, Italy (Received 17 May 1985; accepted 15 July 1985) Two concentrated solutions of NiBr2 have been examined by x-ray diffraction. The Fourier transformed scattering data indicate inner complex formation between Ne + and Be ions. Average numbers of bonded bromide ions per nickel atom have been determined for each solution and the reliability of the complexation numbers as well as of the other structural parameters has been critically examined. INTRODUCTION The study of the interactions between NiH and halide ions has received considerable attention in the past few years. Solutions of NiCl2 in particular have been investigated by several workers, who, among other things, have studied by various techniques the composition of the first cationic coordination shell in different conditions of concentration and composition. In concentrated solutions of NiCI2, where CI-/Ne+ atomic ratios greater than 2 were obtained through the addi­ tion ofLiCI or HCI, x-ray diffraction 1 (XRD) and EXAFS spectroscopy2 have clearly demonstrated the formation of chloro complexes. Coordination parameters estimated in these two studies were in excellent agreement as far as Ni2 + - H20 and NiH -CI- direct distances are concerned, and in good agreement as regards the average number ofNe+ -CI­ contacts, nCI-' Going to stoichiometric solutions of NiCl2 (CI- /Ne+ ratio = 2), the situation is less clear, at least apparently. In fact: (a) EXAFS investigations at the metal K edge3.4 seem to indicate absence of CI- ions in the first coordination sphere of NiH; (b) studies by neutron diffraction with isotopic sub­ stitutions (NDIS)s.6 support the full hydration of NF+ ions; (c) XRD data analysis7 suggests the existence of a consider­ able percentage (- 50%) of the complex Ni(H20)sCI + at concentrations> 3MB ; (d) studies of proton, deuteron, and chlorine relaxation rates (NMR)9.1O support the presence of NiH -CI- direct bonds and propose a linear dependence of nCI- on the concentration, according to which the number of CI- nearest neighbor to the cation is about 0.5 at the satura­ tion concentration; (e) Raman spectroscopy investiga­ tions ll• 12 show evidence of Ni2 + -CI- contacts and put forth the hypothesis that these contacts take place in NiCI~­ units. Actually the situation is less chaotic than it appears. In fact, if we neglect the Raman results, as the existence of NiCI~ - units is a mere speculation in complete disagreement with XRD, EXAFS, NMR, and NDIS studies, the discre- ')Permanent address: Istituto de Quimica, UNESP, Araraquara, Sao Paulo, Brasil. pancies between results from XRD and NMR and those from EXAFS and NDIS narrow dramatically so that they are likely reflecting the different sensitivity of these tech­ niques to minority interactions. As discussed by one of the authors,7 in a 50% mixture ofNi(H20)~ + and Ni(H20)sCI + , the average number of NiH -CI- contacts is only 8.3%. It is probable that such small percentage goes unnoticed or does not reveal itself in some observations, even if it involves a great amount of halo complexes. 13 On the other hand, this fact draws a serious limit to the ability of the mentioned techniques to reveal halo complexes in solution. To clear up this matter, the investigation of NiBr2 solu­ tions should be a very good choice. In fact, there are several reasons why NiH -Br- contacts should be better identifi­ able than the NiH -CI- ones in a structural investigation. First, the NiH -Br- distance is longer than the NiH -Cl­ one; therefore, whichever be the technique used, it should come easier to separate Ne+ -H20 from NiH -Br- interac­ tions. In EXAFS the advantages should come from the very large backscattering amplitudes and from the different be­ havior of the phase function for a Br scatterer, as well as from the easier accessibility ofthe energy range of the anion K edge to the experiments. Finally, in XRD, since the Be scattering factor is much larger than the CI- one, NiH - halogen pairs should give a heavier contribution to the total scattering, so that even a small number of contacts might be revealed and characterized. As extensively discussed in the Discussion, CI- and Br- ions show similar tendency to bind NiH ions. Results on NiBr2 solutions should be therefore a valuable help also to clarify the ambiguities mentioned about NiCl2 solutions. In the light of these considerations we thought it convenient to extend our x-ray diffraction investigations to two NiBr2 so­ lutions (2 and 4 M) at room temperature (T = 20 ± 1°C). Looking at the literature, with surprise we took notice that, in two recent XRD studies 14.IS on NiBr 2 solutions, very different evaluations of the bromo complexation were given. This seems to wipe out the hope that clearer results can be obtained from the study of the chosen system. However, a comparison of the reduced intensity data published in the papers quoted above clearly shows that important differ­ ences exist between the two sets of experimental data. There- J. Chem. Phys. 83 (11). 1 December 1985 0021-9606/85/235797-05$02.10 @ 1985 American Institute of Physics 5797 This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP: 200.145.174.189 On: Wed, 26 Mar 2014 19:41:11 5798 Maglni 6t at : Coordination sphere of NP+ TABLE I. Compositions in moV t. atomic ratios and densities in glcm3 are given for the solutions investigated. Solutions NiBr22M NiBr24M Ni2+ 2.102 4.071 H20INi2+ 24.7 11.9 d 1.396 1.761 fore, the differences in the final results are unlikely to be a consequence of the ambiguity of the analysis, but they must follow from some inaccuracy in one of the two experiments. A further objective of our work is therefore also the under­ standing of these controversial results. EXPERIMENTAL AND DATA TREATMENT The solutions of NiBr2 were prepared from Carlo Erba reagent grade NiBr2 • 6H20. Nickel and bromide ion con­ centrations in the solutions were determined by EDT A and argentometric titrations, respectively. The two methods agreed within less than 1%. The densities of the samples were obtained by a digital precision densimeter. Density val­ ues obtained by us are in good agreement with those estimat­ ed by Wakita et al., 15 while the value proposed by Caminiti et al.14 for the 2M solution they examined is slightly smaller. The compositions of the solutions and their reference sym­ bols are given in Table 1. X-ray apparatus and data normalization procedures have been described elsewhere. 16,17 Diffraction intensities (at least 100 000 counts per point) were recorded in the angular range e = 1. - 7(1', using a Mo x-ray tube (Ii = 0.7107 A.), corresponding to an s range from Smin 0.3 to Smax 16.6 A. - I, wheres is 41T sin e / Ii. The observed intensities were correct­ ed for background, absorption, and polarization. Most of the incoherent scattered radiation was eliminated by using a quartz monochromator on the diffracted beam. The data normalization was carried out using standard methods. 16-18 A correction for spurious ripples below 1.0 A. was also ap­ plied. 19 From the normalized intensities I e.u. the structure func­ tions were obtained according to m its) = I e.u. - L nJf(s) (1 ) i =t 1 and the radial distribution functions D (r) were then evaluat­ ed by a Fourier transformation: D (r) = 4tr,zpo + 2r/tr i:~ si(s)M (s) sin rs ds. (2) FIG. 1. Experimental (dots) and calculated structure functions (solid strong lines) are shown for the solutions investigated. Contributions of the contin­ uum are also shown by weak solid lines. RESULTS Inspection of the radial curves The functions its), after multiplication by r, are shown in Fig. 1. In Fig. 2 the radial distribution functions D (r) are given, while in Fig. 3 radial curves are reported as difference functions D (r) - 4tr,zpo, as they better display the medium range structure. In Fig. 2, the D (r) of the 4 M solution exhib­ its three peaks at about 2.05, 2.60, and 3.35 A.. On the basis of the ionic radii of the species and of previous diffractometric works,I.7.20-24 these distances can be ascribed without ambi­ guity to the pairs NiH -H20, NiH -Br-, and Br- - H20, respectively. In the case of the 2 M solution, while the peaks at 2.05 and 3.35 A. are still present, the intermediate one is replaced by a shoulder at about 2.80 A.. In this case the greater abundance of water makes direct H2O-H20 interac­ tions (usually falling at about 2.80 A.) more important than the NiH -Br- ones; the NiH -Br- distances are probably masked in the large envelope of the peak, as, on the other hand, are H2O-H20 distances under the peak NiH -Br- in 30 OCr) el" A-I. 10-" 24 18 12 In Eqs. (1) and (2), nl's are the stoichiometric coefficients of S the assumed unit, containing m kinds of atoms; /; 's are the scattering factors of the species; Po is the average electronic 0 t--====----L-1- bulk density; M (s) is a modification function of the form with k = 0.005. 2 riA 3 4 FIG. 2. Radial distribution functions D (r) are shown; the parabolic curves represent the 41rj2po functions. J. Chern. Phys., Vol. 83, No. 11, 1 December 1985 This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP: 200.145.174.189 On: Wed, 26 Mar 2014 19:41:11 Magini st a/. : Coordination sphere of Ni2+ 5799 9 6 3 Cal -3 3 -3 -6 1 2 3 4 r /A 5 6 7 15 12 Cbl 9 6 3 -3 , , ' \ I , I -6 \ I \ ....... I \/ 2 3 6 7 9 6 Ni-Br 2nd 3 -3 3 -3 -6 2 3 4 r /A5 6 7 FIG. 3. (a) Difference radial distribution curves, D (r) - 4trrpo, (b) Ditrer­ ence radial curve of the 4 M solution (solid line), radial contributions com­ ing from Br(H20)i complexes (dotted solid line) and radial difference curve after subtraction of the dotted curve (dashed line, (c) Difference radial distri­ bution curves of the two solutions, after subtraction of the contributions coming from Br(H20)6- complexes. The 4 M curve is the same (dashed) curve in (b). the 4 M solution. On the whole, these observations confirm that NiH -Br- interactions are more evident than the NF+ -CI- ones, thus encouraging the quantitative analysis. In Fig. 3(a), besides the peaks just discussed, a meaning­ ful multicomponent peak appears in the 4-5 A range. Impor­ tant contributions to peaks in this distance region are usually provided by interactions among cations and water molecules set in second hydration spheres, H 20 U molecules (the sub­ script I denoting ions or molecules in the first hydration shell, when specification is necessary for clarity). The com­ parison of the difference curves of the two solutions shows that the peaks at issue are different both in shape and in height; oddly, the peak is higher in the more concentrated solution, where less water is available for second ordered hydration shells. This seems to suggest that, in the 4 M solu­ tion, some Br- ions, Brii, set themselves in the second ca­ tionic coordination shells. To check this hypothesis, we tried to subtract from the difference radial curves the contribu­ tions coming from Br- - H20 and H 2O-H20 interactions in the anionic complexes Br(H20)6- , whose existence will be discussed shortly: these contributions are sketched as a dot­ ted line in Fig. 3(b). The results of the subtraction are shown in Fig. 3(c). The residual peaks at 4-5 A, while retaining their complex nature, shift their maximum going from 4.1 A in the 2 M sample (possible NiH -H20 U distance) to 4.6 A in the 4 M sample (possible NiH -Brii distance), in agreement with the explanation proposed above. Analysis of the structure functions In order to evaluate quantitatively the average number of bonded bromide ions per Ni atom, nBr-' as well as t,o obtain a set of structural parameters for the species in solu­ tion, model structure functions may be compared, through a least squares procedure,I,2,16,17 with the experimental i(s) functions. This implies the choice of a plausible model for the solution, in which each atomic species is surrounded by a region with discrete structure, followed by a uniform distri­ bution of distances (continuum). Direct interactions inside the ordered regions dominate the high s values of the struc­ ture functions. Therefore, in order to have indications about the average ionic coordinations and the type of complexes that must be considered for a complete simulation of the i(s)'s, an analysis of the high s range of the 4 M solution structure function was carried out. The terms introduced in the calculations were those describing NF+ -H20, NiH - Br-, H 2O-H20 and Br--H20 direct interactions. By this procedure it was found that the average number of NiH - Br- contacts does not exceed 0.5; the average coordination number of Br- ions came out a little less than 6 (the value proposed in previous diffractometric investigations2o,21), as expected if part of the bromide ions is bound to the NiH ions. Thus, the following complexes were introduced in the calculation of the entire structure functions: Ni(H20)~ + , Ni(H20)sBr+, Br(H20); for the free bromide, Br(H20); for the bound bromide (n < 6). Pair distances, root mean square deviations of the distances, percentage of the complexes Ni(H20)~ + and Ni(H20]sBr+ (which too sets the amounts of the free and bound bromide complexes) were indepen- J. Chem. Phys., Vol. 83, No. 11, 1 December 1985 This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP: 200.145.174.189 On: Wed, 26 Mar 2014 19:41:11 5800 Maginl et al : Coordination sphere of Ni2+ T ADLE II. Parameter values (r = distances, A; q = root mean square de­ viations, A; n = frequency factors) obtained from least squares refinements are given together with their standard errors (in parentheses). Ni2+ -H20 parameters are for both Ni(H20)~ + and Ni(H20)sDr+ complexes; Dr-­ H20 parameters are for both Dr(H20)6- and Dr(H20).- complexes of free and Ni-bonded bromide ions. NiDr22M NiDr24M Parameters Present Work Ref. 14 Ref. IS Present Work Ref. 15 'Nih _H2O 2.066(2) 2.065 2.04 2.079(3) 2.05 O"Ni2+_H :lO 0.083(2) 0.112 0.089 0.096(4) 0.089 'Ni2 +_Br- 2.610(9) 2.62 2.58 2.615(9) 2.53 O"Ni2+_Br- 0.13(1) 0.09 0.13 0.13(1) 0.13 nNP+_Br- 0.29(3) 0.85 0.18 0.44(6) 0.47 '8r--H)0 3.351(3) 3.12 -3.4 3.329(3) -3.4 (18r--H,0 0.210(3) 0.196 0.195(3) 'H2OrHzOil 2.804(5) 2.79 2.75(1) qH2orH~1I 0.05(1) 0.086 0.110(9) n . H 20rHPil 10.2(1) 10.9 4.0(1) (per Ni atom) dently refined. Moreover, interactions due to molecules in cationic second shells were .introduced; obviously this in­ volves a H20} -H20 u term describing the bond between nearest-neighbor and second-neighbor water molecules (fol­ lowing the indication of the previous discussion, for the 4 M solution NiH -Bril interactions had to be accounted for). The transition to a continuous distribution of distances was performed as USual.I.2.16.17 The agreement between experimental and calculated structure functions is shown in Fig. 1. The relevant structure parameters obtained from least square refinements are given in Table II, together with the values proposed for the same parameters in the mentioned studiesl4 • IS of NiBr2 solutions. DISCUSSION AND CONCLUSIONS Preliminary to a discussion of the results, the confidence limits of the parameters given in Table II must be assessed, especially those of nBr- value which is a crucial quantity in the study. It has to be emphasized that, in spite of the com­ plexity of the model, the parameters describing direct inter- 52. i(s).10- 2 32 el·};.2 3 6 9 5/$.'12 15 FIG. 4. Experimental structure function of the 4 M solution (dots) and sum of the contributions from the shorter interactions (continuous line). actions are highly reliable. In fact, these parameters are not much affected by how second shells or continuum are de­ scribed, as the shorter interactions give dominant contribu­ tions at high s. As a proof of this, the simulation of the high s structure function, obtained using direct interactions only, is reported in Fig. 4 for the case of the 4 M solution. However, as we have already discussed, I the parameter errors given in parentheses in Table II are the standard deviations based on the goodness of fit in a nonlinear regression and do not give the true uncertainty, which, instead, can be better evaluated through a comparison of the results obtained using different refinement strategies. In this connection, the major problem in the present case was the interference of the NiH -Br­ (distance of about 2.60 A.) and H2O-H20 (distance of about 2.80 A.) terms. Thus, if the percentages ofNi bromo complex in the 4 M sample was kept in the range 30%-55%, fits almost as good could be obtained by adjusting the H2O-H20 term without any constraint. So, the choice of the most likely nBr- value has been made using additional criteria, that is, odd parameer values must be rejected (e.g., rH O-H 0 <2.70 A.) and a good description of direct interactions' mu~t leave a residual curve [D (r)exp - D (r)calc] smooth. IS As an example, in Fig. 5, experimentalD (r), pair distributions at low r, Do (r), and residual curves D (r) -l:Do(r) are given. Following these criteria, the uncertainty of nBr- turned out within ± 10% in the 4 M solution and a little higher in the 2 M solution, in which the term Ni2+ -Be, though im­ portant for a good fit, has a smaller weight. All these facts considering, we can affirm that undoubtedly our results are consistent with those reported by Wakita et 01. IS for both 24 18 12 6 ---_ ... O~-----------~~~-- 24 18 12 6 I , , I ~ , 4M \ / ,.- X,·· / . ~/// \ ...... / i ,I\. ,I .. / I \ ,-· '/ · .. · '. · , . .~# ,/ • / , , I , f , f , " 2M .,-..". , , .... / " . 2 /A 3 4 FIG. 5. Radial distribution functions D (r) (solid line), sum of synthetic pair distribution functions from the model, I.DI/(r) (dotted line) and residual curves D (r) - I.Dij (r) (dashed line). J. Chern. Phys., Vol. 83, No. 11,1 December 1985 This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP: 200.145.174.189 On: Wed, 26 Mar 2014 19:41:11 Magini et a/. : Coordination sphere of NiH 5801 solutions, while they are absolutely inconsistent with those proposed by Caminiti et al. 14 for the 2 M solution. Referring to the latter, we must observe that the disagreement extends also to the important parameter r Br--H 2 0' In fact, while the value we obtained is similar to the ones given in many other bromide solutions, 15.20-24 the distance proposed by Caminiti et al. (3.12 A) is unlikely short, even shorter than the dis­ tances of the CI--H20 pairs,I,7 in spite of the greater ionic radius ofthe Br- ions. As observed by Wakita et al., the stability constants giv­ en in literature for the formation of the species Ni(H20)5Br+ cover too a wide range of values to allow a judgement of the diffractometric results on these grounds to be made. How­ ever, stability constants proposed by various authors,25 al­ though different according to the different determination method used, concordantly indicate that the NiH ions tend to bind CI- and Br- ions almost to the same extent. It is also for this reason that the results of Caminiti et al. do not look much convincing, in that they entail a complexation degree much larger for NiH -Br- than for NiH -Cl- pair. The present study shows that in stoichiometric concen­ trated solutions of NiBr2 bromo complexation ofthe cation takes place. Even in the 2 M solution, the complete omission of NiH -Br- interactions makes the simulation of the ex­ perimental structure function much worse. Because of the discussed similarity ofBr- and CI- ions in the formation of halo complexes with NiH ions, the clearer evidence of NiH -Br- interaction at a given concentration must be as­ cribed to the greater scattering factor of the bromide that makes the pair distribution function D Ni'+ -Br- heavier than the D NiH -CI- one. From this point of view the present result is an indirect confirmation of the existence of halo complexes also in concentrated solutions of NiCI2• It will be certainly interesting to compare the present results with those coming from other structural techniques. Actually, few studies similar to those described in the Intro­ duction have been published. The studies of proton, deu­ teron, and halide relaxation rates9 • 10 estimate an amount of Br - in the first coordination sphere of the Ni2 + almost coin­ cident with that proposed for Cl- ions and in good agree­ ment with the present evaluation. Besides, only one EXAFS investigation26 exists, where a small percentage of NiH _ Br- contacts is neither confirmed nor excluded. 1M. Magini, G. Paschina, and G. Piccaluga, J. Chern. Phys. 76, 1116 (1982). 2G. Licheri, G. Paschina, G. Piccaluga, and G. Pinna, J. Chern. Phys. 79, 2168 (1983). 3D. R. Sandstrom, J. Chem. Phys. 71, 2381 (1979). 'G. Licheri, G. Paschina, G. Piccaluga, G. Pinna, and G. Vlaic, Chem. Phys. Lett. 83, 384 (1981). sG. W. Neilson and J. E. Enderby, J. Phys. C 11, L625 (1978). 6G. W. Neilson and J. E. Enderby, Proc. R. Soc. London Ser. A 390, 353 (1983). 7M. Magini, J. Chern. Phys. 74, 2523 (1981). SIn this connection we may recall that the Cagliari group, after having sug­ gested full hydration ofNF+ ions in concentrated solutions of NiCl2, has reexamined (Ref. 1) its own data in the case of a 4 M NiCl2 solution, verify­ ing that a model, in which equal quantities of the complexes Ni(H20)~ + and Ni(H20)sCI+ are introduced, agrees with the experimental data even better than a model based only on Ni(H20)~ + ions. 9JI. Weingartner and H. G. Hertz, J. Chern. Soc. Faraday Trans. 175,2700 (1979). wH. Weingartner, C. Muller, and H. G. Hertz, J. Chem. Soc. Faraday, Trans. 1 75,2712 (1979). "M. P. Fontana, G. Maisano, P. Migliardo, and F. Wanderlingh, Solid State Commun. 23,489 (1977). 12M. P. Fontana, G. Maisano, P. Migliardo, and F. Wanderlingh, J. Chern. Phys. 69, 676 (1978). 13The conclusive role of the sensitivity of the techniques used is clearly dem­ onstrated by recent EXAFS and XANES investigations performed by Sandstrom [EXAFS and Near Edge Structure III, edited by K. 0. Hodg­ son, B. Hedman, and J. E. Penner-Hahn (Springer, Berlin, 1985), p. 409] at the CI K edge, in which the existence of C1-_Ni2+ contact at approxi­ mately the same level as detected by x-ray di.1fraction is evidenced. NiH ions in the first sphere of CI- ions are at distances shorter than the water molecules, and, as a consequence, they can be observed more easily than CI- ions in the first shell ofNF+ ions, where H20 is the nearest-neighbor species. I4R. Caminiti and P. Cucca, Chem. Phys. Lett. 89, 110 (1982). ISH. Wakita, M. Ichihashi, T. Mibuchi, and I. Masuda, Bull. Chem. Soc. Jpn. 55, 817 (1982). 16R. Caminiti, G. Licheri, G. Piccaluga, G. Pinna, and M. Magini, Rev. Inorg. Chern. 1, 333 (1979). 17M. Magini, J. Chem. Phys. 70, 317 (1979). ISG. Licheri, G. Piccaluga, and G. Pinna, J. Chem. Phys. 64, 2437 (1976). t9H. A. Levy, M. D. Danford, and A. H. Narten, ORNL Report No. 3960 (1966). 20G. Licheri, G. Piccaiuga, and G. Pinna, Chem. Phys. Lett. 35,119 (1975). 2tG. Licheri, G. Piccaluga, and G. Pinna, J. Chem. Phys. 63, 4412 (1975). 22J. Glaser and G. Johansson, Acta Chem. Scand. Sect. A 36,125 (1982). 23M. Ichihashi, H. Wakita, T. Mibuchi, and I. Masuda, Bull. Chern. Soc. Jpn. 55, 3160 (1982). 24M. Ichihashi, H. Wakita, and I. Masuda, J. Solut. Chem. 13, 505 (1984). 2SL. G. Sillen and A. E. Martell, Stability Constants of Metal Ion Complexes (The Chernical Society, London, 1964, 1971) Vols. 17 and 25. 26p. Lagarde, A. Fontaine, D. Raoux, A. Sadoc, and P. Migliardo, J. Chem. Phys. 72,3061 (1980). J. Chern. Phys., Vol. 83, No. 11. 1 December 1985 This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP: 200.145.174.189 On: Wed, 26 Mar 2014 19:41:11