754
J . Phys. Chem. 1984,88, 754-756
0.005 M NaLS with k, = 1.1 X lo7 but is very rapid beyond that. Once micelle formation ensues, the positively charged uranyl ion is drawn to the micelle where it rapidly reacts with the sulfate head groups. Similar behavior is seen if various amounts of KH2P04/ Na2HP04are added to form 1 X lo4 M uranyl solutions in water. The phosphate mixture is in fact a pH 7.4 buffer solution. Initially the steady-state emission intensity rises and the lifetime increased from 11 to 54 p s . Further addition of phosphate leads to static quenching only; the lifetime remains the same. Binding. As noted earlier, above the cmc NaLS is a highly effective quencher for the excited-state uranyl ion. Negatively charged quenchers, on the other hand, are electrostatically repelled from the silica particles and consequently show very slow reaction rates. If uranyl acetate is allowed to equilibrate with NaLS, followed by addition of silica to form 20% Si No. 1 115/ 1 X 1O4 M uranyl acetate/0.05 M NaLS solution, the uranyl ions are easily
seen, by means of steady-state fluorescence, to migrate from the NaLS to the silica particles until essentially all of the uranyl is on the silica. Migration is significant within a few minutes and complete within hours. Conclusion Studies on the uranyl ion indicate that it enjoys significantly different photochemical properties when it is bound to a colloidal silica particle as compared to free aqueous solution. This is reflected in a huge increase in the emission lifetime and other spectral features. These properties could not be obtained in free solution or with the group of surfactants studied.
Acknowledgment. We thank the Army Research Office via Grant No. DAAG29-80-K-007 POO1, for support of this research, and NALCO Chemicalf Co. for the generous gift of colloidal silica. Registry No. Uranyl ion, 16637-16-4; silica, 7631-86-9.
Homogeneity and Magnetic Susceptibility in Some Substituted Cadmium Spinels Mark Tellefsen, Louis Carreiro, Robert Kershaw, Kirby Dwight, and Aaron Wold* Department of Chemistry, Brown University, Providence, Rhode Island 0291 2 (Received: May 3, 1983)
The normal spinels CdFe204,CdGal,,Feo,204,and CdRhl,8Feo,204 were prepared, and their magnetic susceptibilities were measured. The low perfvalue of 4.72 ( 5 ) p B found for CdFe204can be attributed to hybridization of the 6Sand 4G wave functions resulting from spin-orbit interaction coupled with strong crystal fields having trigonal components at the spinel B sites. Magnetic susceptibility was sensitive to the homogeneity of samples of CdGal,8Feo,204 and CdRhl,xFeo,204, which was dependent upon the method of preparation. For homogeneous samples, the magetic susceptibility approaches the theoretical value for high-spin Fe3+(3ds). The slight remaining discrepancy from spin-only moment is due to the statistical existence of a few small Fe3+clusters.
Introduction The spin-only moment observed for the Fe3+(3d5) ion in a number of oxide systems (Fe2O3, Fe203/Rh203,ZnFe204) has been reported to be unusually low compared to the theoretical In an earlier study, Selwood' indicated that iron-iron interactions between adjacent cations were probably responsible for the low moment observed in a-Fe203. Similarly, low moments were observed by K r h 2 in a study of the solid solution Fe2O3Rh203. However, this system presents a further problem in that complete homogeneous solid solution is difficult to attain.4 Lotgering3 has also reported on the low moment observed for Fe3+ in the normal spinel ZnFe204,which he attributed to the presence of a small number of Fe3+ ions at tetrahedral sites. However, the reported temperature dependence of the magnetic susceptibility for ZnFe204was linear and hence the lowering of the moment could not be attributed to the properties of A-O-Bclusters. Another possible explanation might be interaction of t,, orbitals on neighboring B sites resulting in a trigonal distortion of the cubic field, permitting admixture of the 6S and 4G states. In order to study the role of such t2*-t2, interactions, the systems CdGa2-,Fe,04 and CdRh2-,Fe,04 were chosen. The magnetic susceptibility of the normal spinel CdFe204has not been reported. However, a lowering of the spin-only moment of Fe3+ would be anticipated. In addition, members of the systems CdGa2-xFe,04 and CdRh2-,Fe,04 are expected to crystallize with the normal (1) P. W. Selwood, L. Lyon, and M. Ellis, J . Am. Chem. SOC.,73,2310 (1951). (2) E. KrCn, P. Szabb, and G.Konczos, Phys. Lett., 19 (2), 103 (1965). (3) F. K. Lotgering, J . Phys. Chem. Solids, 27, 139 (1966). (4) L. Carriero, unpublished research.
0022-3654/84/2088-0754$01.50/0
spinel structure, and it may be possible to minimize the trigonal distortion and hence obtain a measured moment per Fe3+ approaching the theoretical value. Experimental Section Preparation. Polycrystalline samples of the systems CdRh2-xFe,04 and CdGa2-,Fe,04 were prepared by the solid-state reaction of the binary oxides. Samples of CdRh,,8Feo,204 were also prepared by a precursor method starting from ammonium hexachlororhodate(II1) and iron(I1) sulfate h e ~ t a h y d r a t e . ~ Solid-State Reaction of the Oxides. Starting materials consisted of Fe203 (Mapico Red, Columbian Carbon Co.), GazO3 (Gallard and Schlesinger, 99.999%), RhzO3, and CdO. Rhodium(II1) oxide was prepared in the high-temperature, ambient pressure form (space group Pbca) by heating finely divided rhodium metal (Engelhard Inc., 99.99%) under flowing oxygen at 800 O C 5 Cadmium oxide was obtained from the decomposition of C d C 0 3 at 450 O C in air.6 Members of the system CdRh2-,Fe,04 were prepared by thoroughly grinding together stoichiometric quantities of the oxides (with a 5% by weight excess of CdO) and heating in a silica boat open to air at 800 OC. Fast-scan X-ray diffraction analysis indicated completion of the reaction by the absence of Fe203 or Rh203after two 24-h heating intervals with intermittent grkding. After complete reaction, excess CdO was removed by washing the product with 50 mL of hot 1 M (aqueous) NH,Cl, followed (5) H. Leiva, R. Kershaw, K. Dwight, and A. Wold, Mater. Res. Bull., 17, 1539 (1982). ( 6 ) V. S. Nguyen, R. Kershaw, K. Dwight, and A. Wold J . Solid State Chem., 36,241 (1981).
0 1984 American Chemical Societv
The Journal of Physical Chemistry, Vol. 88, No. 4, 1984 755
Magnetic Susceptibility in Some Substituted Cd Spinels
TABLE I: Magnetic Properties of CdGa,~,Fe,,,O, and
CdFe,O,
-7r
CdRh L.8Fe0.20,
_ _ L L A - I . _ . ~ L - L _ L - I _ I
120
1 130 80
ri 4
composition CdFe,O, CdCa,~,Fe,,,0,
ji
CdRh,,,Feo.,O,
0 \
Fe3+(3d5)d
E
"
-
starting materials oxides oxides oxides precursors
prepn temp, "C
800 800 900 1000 800
x293K a 0.0096 (1) 0.0122 ( l ) b 0.0131 (1) 0.0137 (1) 0.01 14 (1)' 0.0137 (1) 0.0150
Units: emu/mol equiv of Fe. Corrected for core diamagnetism." Uncorrected. ptheo = 5.92 p ~ . a
X
20
4
/'
/'' 1
obeys Curie law with peN = 4.72 (5) pB. This value of p e f fis less
than the value of 5.50pB reported for ZnFez0,3 and is significantly -----------4
0 0
100 Temperature
200
300
(K>
Figure 1. Inverse magnetic susceptibility vs. temperature for CdFe204.
by 50 mL of hot distilled water. It was necessary to pretreat samples of the CdGaz,Fe,04 system by "nitrating" the starting mixtures, because of the poor reactivity of Gaz03. Mixtures were ground until homogeneous and transferred to a porcelain crucible. Concentrated (16 M) nitric acid was added dropwise to sufficiently wet the powder, and the crucible was heated gently until the decomposition of all nitrates was complete. The powder was ground thoroughly and transferred to a silica tube. The tube was sealed with an internal pressure of 1 atm of air. Samples heated at 800 "C required three 24-h intervals with intermittent nitration and grindings. Samples heated at 900 or 1000 "C required two 24-h heating intervals. Cadmium oxide is volatile above 825 "C and must be heated in sealed tubes above this temperature.6 Preparation of CdRh,,8Feo,20,by the Precursor Method. Ammonium hexachlororhodate(II1) (Engelhard Inc., 99.999%) and iron(I1) sulfate heptahydrate were taken in stoichiometric quantities with respect to metal assay, and ground together. After the mixture was transferred to a porcelain crucible, concentrated (1 8 M) sulfuric acid was added dropwise, and heating proceeded as in the nitration method above. CdO was added to the resulting black powder, and the mixture was ground thoroughly and heated at 800 "C. After complete reaction was achieved, the excess cadmium was removed. Analysis of the rhodium complex precursor was performed by reduction under flowing 85% argon/l5% hydrogen at 500 "C, employing a thermogravimetric (TGA) balance. Iron(I1) sulfate heptahydrate was ignited in air at 800 "C to give Fe203.
-
Sample Characterization X-ray Diffraction. Powder diffraction patterns were obtained with a Norelco diffractometer with monochromatic high-intensity Cu Ka, radiation (A = 1.5405 A). Fast scans were taken at a rate of 1" 28/min in the range 12" < 28 < 72". Lattice parameters were determined by least-squares analysis of slow scans at 0.25" 28/min in the range 26" < 28 C 74". Magnetic Measurements. Magnetic susceptibility measurements were performed from 77 to 300 K with a Faraday balance described elsewhere.' Field-dependent measurements were made with field strengths between 6.22 and 10.4 kOe, and the balance was calibrated with platinum wire (x, = 0.991 X 10" emu/g at 275 K). Results and Discussion The normal spinel CdFe204(space group Fd3m) is a red powder when prepared at 800 "C in air. The observed cell parameter, a, = 8.708 (1) A, agrees with the value reported in the literature.s As shown in Figure 1, the magnetic susceptibility data for CdFe204 (7) B. L. Morris and A. Wold, Rev. Sci. Instrum., 89, 1937 (1968). (8) E. J. W. Verwey and E. L. Heilmann, J . Chem. Phys., 15, 174 (1947).
lower than the theoretical value of 5.92 pB for spin-only S = 5 / 2 . In fact, the high-spin 6Sconfiguration is the rigorous ground state only for a free 3d5 ion. Its energy is not affected by crystal fields or by spin-orbit interaction, whereas that of the immediately adjacent 9-fold degenerate 4G manifold is influenced strongly by these factors. Significant hybridization of 6S and 4G wave functions belonging to the same irreducible representation can occur if the crystal fields are sufficiently strong and of less than cubic symmetry so as to permit spin-orbit coupling. Neighboring cations contribute a trigonal component to the crystal field at a B site in the spinel structure. Hence a lowering of the Fe3+ moment by admixture of 4G into 6S is possible when the crystal fields are large enough. The observed reduction from 5.92 to 4.72 p B shows this to be the case for CdFez04. The observation of a smaller reduction for ZnFe204can be attributed to ZnZ+being less covalent than Cd*+, since covalent bonding contributes appreciably to the strength of crystal field^.^ The reduction of Fe3+ moment by this mechanism requires a sufficiently large trigonal distortion from cubic symmetry. In the case of a dilute solution of CdFe204in a diamagnetic host spinel, the distortion at each Fe3+ site depends upon the number of adjacent Fe3+ neighbors, Le., upon the size of its cluster. Consequently, the magnetic moment of each Fe3+ion, its contribution to the Curie constant, and hence pLeN will depend upon the cluster size. The distribution of clusters characteristic of a homogeneous solution can be obtained by statistical analysis; inhomogeneous clustering has a pronounced effect upon the magnetic properties of these solid solutions. This effect is illustrated by the magnetic properties of the solid solutions CdGa2,Fe,04 and CdRh2,Fe,0, reported in this study. The composition CdGal 8F%,204was chosen for study since the effects of both phase homogeneity and the intrinsic magnetic behavior of Fe3+ should be observed at this level. Preparations of the mixed composition CdGa,,8Feo,204were performed at temperatures of 800,900, and 1000 "C. X-ray analysis indicated the formation of a single-phase spinel at each temperature. The color was yellow; unsubstituted CdGa204is white. The cell parameter for CdGal 8 F ~ , 2 remained 04 constant at a value of 8.614 (2) A for all temperatures of prepartion. This value represents an increase from the observed cell parameter for CdGaz04at 8.604 (1) A, which is consistent with the substitution of a cation of larger ionic radius, Fe3+ (r = 0.645 A), for one of smaller ionic radius, Ga3+ ( r = 0.620 A).1o The room temperature magnetic susceptibility of C d G a l , 8 F ~ , 2 0 4 is given as a function of the temperature of preparation in Table I. The values reported represent those attained after repeated heatings produced unchanged magnetic susceptibilities. Clearly, there is an increase of the magnetic susceptibility for CdGa,,8Fe0,,04with increasing temperature of preparation, which is in(9) J. B. Gocdenough, "Magnetism and the Chemical Bond", Wiley-Interscience, New York, 1963. (10) R. D. Shannon and C. T. Prewitt, Acia Crysiallogr., Sect. B, 25, 925 (1969). ( 1 1) P. W. Selwood, 'Magnetochemistry", 2nd ed,Interscience, New York, 1956, p 78.
756
J. Phys. Chem. 1984,88, 756-759
dicative of an increase in phase homogeneity. Preparations of the mixed composition CdRhl,sFe0,204 were performed by solid-state reaction of the oxides and by a precursor method. X-ray analysis of the products in both cases indicated the formation of a single-phase spinel. The measurement of cell parameters was not attempted, since unsubstituted CdRh204and CdFe204have similar cell parameters at 8.765 (1) and 8.708 (1) A, respectively. The magnetic susceptibility values given in Table I indicate that the precursor method for preparing solid solutions achieves maximum homogeneity. Since the effects of phase homogeneity on the magnetic susceptibilities of both solid solution systems have been established, a comparison with unsubstituted CdFe204is appropriate. Within the limits of the synthetic technique, the samples of C d G a l , s F ~ , 2 0 4 prepared at lo00 "C and C ~ E U I ~ , ~ Fprepared Q ~ O ~ by the precursor method are the most homgeneous for the respective compositions. emu/mol equiv The magnetic susceptibility value of 1.37 X of Fe observed for these samples shows a large increase from the value of 0.96 X emu/mol equiv of Fe for unsubstituted CdFe204. The remaining small deviation from spin-only moment is due to the clustering inherent in a random distribution of 10% iron over the spinel B sites. Statistical analysis for this composition predicts that 53% of the Fe3+ions will be isolated and that 28% will exist in nearest-neighbor pairs, 11% in clusters of three, and a total of 8% in larger clusters. The Curie constant and magnetic susceptibility are the sums of contributions from each Fe3+ion; the amount contributed per ion will depend upon the cluster size. For example, the measured value of 1.37 X emu/mol equiv of Fe could be explained by assigning the spin-only value of 1.50 X to isolated and paired Fe3+ions, and the bulk value of 0.96
X
to clusters of three or more.
Summary and Conclusions The normal spinels CdFe204,CdGal,8Feo,204, and CdRhl,,Feo.204 were prepared, and their magnetic susceptibilities were measured. Unsubstituted CdFe204showed a gcffof 4.72 (5) pB, which is lower than the value 5.92 pB expected for high-spin Fe3+(3d5).The deviation from spin s / 2 behavior is attributed to a hybridization of the 6Sand 4G states resulting from spin-orbit interaction coupled with strong crystal fields possessing trigonal components at the spinel B sites. The magnetic susceptibility of CdGal.sFeo,204 samples was found to increase with the temperature of preparation, which was attributed to an increase in phase homogeneity. Similarly, samples of CdRhl,sFeo,204 prepared by a precursor method were found to be more homogeneous than those prepared directly from the oxides, as determined by magnetic susceptibility. For homogeneous samples of both CdGal,8Feo,204 and CdRh1,sFe0,204, observed magnetic susceptibilities of 0.0137 emu/mol equiv of Fe approach that expected for Fe3+(3ds). The small deviation from ideal spin-only moment is caused by the statistical existence of a few remaining Fe3+ clusters. Acknowledgment. Acknowledgment is made to the Office of Naval Research, Arlington, VA, for the support of Mark Tellefsen and to the National Science Foundation (Grant DMR-82-03667) for the support of Louis Carreiro and Kirby Dwight. Acknowledgment is also made to Brown University's Materials Research Laboratory program which is funded through the National Science Foundation. Registry No. CdFe204,12013-98-8.
Raman Spectra and Force Constants for the Nitric Oxide Dimer and Its Isotopic Species E. M. Nom,+L.-H. Chen, M. M. Strube, and J. Laane* Department of Chemistry, Texas A & M University, College Station, Texas 77843 (Received: May 25, 1983)
The Raman spectra of the solid nitric oxide dimer (16014N14N160) and two of its isotopically substituted species (16015NisNi60 and 18015N15N'80) at 12 K have been recorded and analyzed. The observed isotopic shifts demonstrate that fundamental vibrational frequencies for the normal isotopic species are at 1866 (ui), 1762 ( u s ) , 268 ( u 2 ) , 215 ( u 6 ) , and 189 ( u 3 ) cm-'. A band at 97 cm-' may be due to the sixth vibration (v4). The isotopic frequency shifts observed for the 495-cm-I band demonstrate that this is the combination band u2 + Y6 and is not a fundamental. The data for the three isotopic forms of cis-N202made it possible for the first time to calculate a meaningful set of force constants for this molecule. The stretching force constant for the weak N-N bond was calculated to be 0.32 mdyn/A.
Introduction The vibrational spectra of the nitric oxide dimer, cis-ONNO, which is formed upon condensation of nitric oxide, have been studied at least eight times Nonetheless, general agreement on the vibrational assignments has not been reached. Table I summarizes the previous spectroscopic data reported for cis-N202.There is general agreement that bands near 1860,260, and 170 cm-' are the AI modes, but whether 260 or 170 cm-' results from the N-N stretching has remained an open question. The assignment of the antisymmetric stretching v s at 1760 cm-I is secure, but the frequency of the bending motion v6 has not been firmly established. Durig and Griffins assigned a Raman band at 489 cm-I to this mode, but more recent work'sS places it near 'Present address: Department of Chemistry, Zagazig University, Zagazig, Egypt.
0022-3654/84/2088-0756$01.50/0
200 cm-'. However, the band near 490 cm-I has been observed in other studies1.' and its origin is not obvious. Isotopic substitution with ISN has in one case been previously usedS in an attempt to clarify the assignments. The reported frequency shifts for the vI N = O stretching bands were reasonable (-30 cm-I), but those for the 264- and 176-cm-I bands (-2 and (1) A. L. Smith, W. 189 (1951).
E. Keller, and H. L. Johnston, J . Chem. Phys., 18,
( 2 ) W. G. Fateley, H. A. Bent, and B. Crawford, Jr., J . Chem. Phys., 31, 204 (1959).
( 3 ) W. A. Guillory and C. E. Hunter, J . Chem. Phys., 50, 3516 (1969). (4) C. E. Dinerman and G. E. Ewing, J . Chem. Phys., 53, 626 (1970). (5) J. R. Durig and M. G. Griffin, J . Roman Spccfmsc.,5 , 273 (1976). (6) G. R. Smith and W. A. Guillory, J. Mol. Specfrosc.,68, 223 (1977). (7) A. Anderson and 9. Lassier-Govers,Chem. Phys. Left., 50, 124 (1977). (8) J. R. Ohlsen and J. Laane, J . Am. Chem. SOC.,100, 6498 (1978).
0 1984 American Chemical Society
