s. c. ABRAHAMS, x.P. GINSBERG,A N D K . KNOX
Iiaorga nic Claernistr y
For Re:! with site symmetry C3h-', the triangle of K + ions a t the prism faces is not required to be rotated exactly 60' with respect to the triangles of the prism top and bottom, but, in fact, i t is only slightly less than 4' from this position. Furthermore, the averages of the nine Re-K distances in the two different prisms, 3.78 and 3.88 A., respectively, are nearly equal. Thus, the nonequivalence of the Re atoms appears to be a matter of crystallographic description rather than a
fundamental physical or chemical difference, in agreement with previous studies.Is2 The potassium ions are far enough from the Re for there to be plenty of space around the Re to accommodate the hydrogen atoms, even though the latter may occupy an appreciable amount of space.
558
Acknowledgments.-IVe thank hlr. C. R. Sprinkle for assistance in growing the crystals and Miss N. 1'. Vaughan for help with the refinement of the structure.
COXTRIBUTION FROM
THE
BELLTELEPHOXE LABORATORIES, I s c o RPORATED, MURRAY HILL,SEW JERSEY
Transition Metal-Hydrogen Compounds. 11. The Crystal and Molecular Structure of Potassium Rhenium Hydride, K2ReH, BY S. C. ABRA4HAMS,1A. P. GINSBERG,
ASD
K. KNOX
Receiced Nooenzbei 6 , 1963 X neutron diffraction analysis of K&Hy has established the composition of this unusual potassium rhenium hydride, Further chemical analysis has verified the stoichiometry. The hydrogen atom positions have been determined within space group P62m as: three H in (3f) (x00, etc.) with x = 0.1789 i 0.0024; six H in 6(i) (xOz, etc.) with x = 0.8789 =t0.0014, z = 0.2171 i 0.0032; six H in 6(k) (xy'l2, etc.) with z = 0.1483 f 0.0018, y = 0.6250 =k 0.0024; and twelve H in 12(1) ( x y z , etc.) with x = 0.2240 i 0.0010, y = 0.5254 i 0.0013, z = 0.7125 i 0.0020. The two crystallographically independent R e atoms are essentially equivalent. Both are a t the centers of trigonal prisms of atoms, with three additional H atoms beyond the centers of the prism faces. The average Re-H distance is 1.68 zk 0.01 8 . ; the average H-Re-H angle between hydrogens in the same vertical mirror plane is 93.6 i 0.6". An elementary 1,CAO-MO description of the ReHsp2 ion is given. The optical spectrum of K2ReHg in alkaline aqueous solution is reported. The band observed at 46,080 cm:-' is assigned as a transition between bonding and antibonding molecular orbitals. The absorption intensity is related t o the polarity of the &Io's. I t is concluded t h a t the bondirig M0's involving the Re 5d orbitals are most likely polarized toward the H atoms, The n.m.r. and exchange equivalence of all the hydrogens in KeHS-2 is explained by their similar environment and by the existence of deformation modes which, with small displacements, can interchange the prism and equatorial hydrogens
H
Introduction
(1954).
This formula is in disagreement with the conclusions of Ginsberg, Miller, and Koubek,i who found the hydride to have the stoichiometry K:!ReHs. That the two hydrides are the same is shown by their identical infrared spectrai and the siniilarity of their methods of preparation. The relation between the hydride formed by reduction of perrhenate with alkali metals and Lundell and Knowles' rhenide has also been discussed recently.8 I t was concluded, on the basis of spectroscopic and polarographic evidence, that these must be different species, although rhenide is also probably a rhenium hydride. The formula K2ReH8for the hydride gave rise to two objections, since the rhenium is formally in the + B oxidation state with a 5d1electron configuration. Such a material might be expected to show both color and temperature-dependent paramagnetism, neither of which is found.7r9 I n an effort to find an explanation for these facts, X-ray and neutron diffraction investigations of the structure of potassium rhenium hydride were initiated. The X-ray results have already been re-
(3) G. E. F. Lundell and H. B. Knowles, J . Res. N u l l . Buv. Std., 18,629 (1937). ( 4 ) J. G Floss and A . V. Grosse, J. l i z o v g . Nucl. C h e m . , 9, 318 (1959). ( 5 ) A . P. Ginsberg, J. >I, Miller, J. R. Cavanaugh, and B. P. Dailey, Y o t i w e , 185, 628 (1960). (6) J. G. Floss and A . V. Grosse, J . I z o v g . S u c l . Ckewz., 16, 36 (1960).
(7) A. P. Ginsberg, J. M. Miller, and E. Koubek, J . A m . C h e w S o c , 83, 4909 (1961). (8) A. P. Ginsberg and E. Koubek, Z . uxovg. allge?n. C h e m . , 316, 278 (1962). (9) K . Knox and A . P. Ginsberg, I n o v g . Chenz., 1, 946 (1962).
The nature of the reactive rhenium compound obtained by reduction of perrhenate, in aqueous ethylenediamine solutions, with potassium metal has been the subject of some controversy. Bravo, Griswold, and Kleinberg,2the first to study the reaction, obtained a product which they believed to be K R e . 4 H 2 0 in impure form. They identified this with Lundell and Knowles' rhenide,3 prepared by reduction of acid perrhenate solutions with zinc amalgam and known only in dilute solution. In 1959 Floss and Grosse4 claimed to have isolated the rhenide, K R e ' 4 H 2 0 , in pure form. However, in 1960, Ginsberg, et aZ.,5 reported chemical and nuclear magnetic resonance evidence showing the product described by Bravo, Griswold, and Kleinberg to be a potassium rhenium hydride with hydrogen directly bonded to the rhenium. Floss and Grosse thereupon revised their formula t o KReH4 * 2H20.6 (1) Guest Scientist; Brookhaven National Laboratory, Upton, N . Y . (2) J. Bravo, E. Griswold, and J. Kleinberg, J . P h y s . Cheiiz., 58, 18
Vol. 3, No. 4,April, 1964
CRYSTAL AND
MOLECULAR STRUCTURE OF KzReHg 559
on either side and clear of the "tail" of the peak. Maximum to minimum integrated intensities in the two zones were in the ratio of 245: 1 for hkO and 136: 1 for hOZ. These relatively small ratios, and the large number of unobserved reflections, are caused by the large spin-incoherent cross section for hydrogen of 81 barns. The bound coherent cross section for hydrogen, by comparison, is 1.8 barns. The incoherent scattering results in high background counts and hence greatly reduces the usual signal to noise ratio. The standard deviation in each intensity was obtained from the variation within the form of each reflection and the mean intensity value. The standard deviation in unobserved intensity terms was taken as oneahalf the estimated upper limit of the term. The largest differences in absorption correction are about 5% and were regarded as sufficiently small to be negligible. The mean intensity of each reflection was corrected by the Lorentz factor. The resulting structure factors, scaled to the final set of calculated structure factors, are given in Table I under Fmeaad. In this table, mean values are given for the conimon F(h00) terms. The (h01) layer was also carefully examined at 4.2'K. for evi. dence of possible magnetic scattering. No magnetic reflections were observed. The ratio of I(nuc1ear) a t 4.2'K. to I(nuc1ear) a t 298'K. was generally as large as 1.4: 1, with many previously unobserved I(h01) clearly measurable above background. Chemical.Severa1 single crystals of the hydride were washed with methanol and then with ether. After being blown dry (N2) they were crushed to a powder in an argon-filled drybox. The powder was dried overnight in vacuo a t 82'. All of the following experiments were performed on the dried powder. Infrared spectra (Nujol mull and KBr disk) showed the presence of small amounts of perrhenate. I n order to determine the Experimental amount of KReOa in the sample a series of KBr disks was prepared containing known concentrations of KReOa over the range Potassium rhenium hydride was prepared as described pre0.004 to 0.125%. For each concentration three disks were viously.7 The single crystals of anhydrous KzReHg were grown produced using the same weight of KBr-KRe04 mixture and by slow evaporation of an aqueous KOH solution of the comapplying the same pressure for the same length of time. As a pound. check on the reproducibility of the pressing conditions, the Crystal Data.-Potassium rhenium hydride, KzReHg; formula thickness of each disk was measured with a micrometer, and Hexagonal, weight273.49; dm3.07g.c r n . ~ d, ~ ; 3.094 g; corrections were made for the small variations. The peak with a = 9.607 f.5 and c = 5.508 f 5 A.lO'll Three formula height of the band a t 915 em.-', characteristic of perrhenate,' weights per unit cell. Most probable space group, D3ah-PG2m.l0 was determined on a Perkin-Elmer Model 137 Infracord spectroAbsorption coefficient for X = 1.032 A. neutrons, 0.364 em.-'. photometer. A plot of mean peak height vs. concentration of Volume of the unit cell, 440.24 A.3. Nuclear scattering lengths, KRe04 was linear over the range of 0.004 to 0.030% ( 4 points); b R e = 0.92, b~ = 0.35, b~ = -0.378 cm.). a t the higher concentrations the plot deviated from linearity. Neutron Diffraction.-Two crystals were used in the neutron The perrhenate content of the hydride was now determined by diffraction investigation. Both exhibited irregular hexagonal preparing KBr disks of the standard thickness containing known prism development. For measurement of I(hkO), the crystal had amounts of KzReHg. They gave a 915 em.-' peak whose height idealized dimensions 2.9 X 3.4 X 5.7 mm. with a volume of 26.8 fell on the linear portion of the calibration curve. The total mm.8; for I(hOZ), the crystal had idealized dimensions 2.2 X rhenium content of the hydride was determined as described 2.4 x 3.4 mm. with a volume of about 10.1 mm.3. The intensibefore.' ties of the diffracted neutron beams in both zones were measured To determine the amount of hydrogen in the hydride, -30-mg. with our single crystal automatic neutron diffractometer,12 a t samples were weighed into a quartz tube (28 cm. long, 12 mm. Brookhaven National Lahoratory. K2ReH9 crystals are un0.d.) whose top was the inner part of a 12/30 ground glass joint. stable in moist air, and they were preserved for the long measurement times within an evacuated goniometer-mounted ~ r y o s t a t ' ~ The tube was connected by a ground glass cap through a threeway stopcock to a closed end manometer reading from 0 to 260 maintained a t room temperature. mm. The entire system was evacuated; then the hydride was Each accessible form of every reflection was measured a t least heated to redness, cooled to room temperature, and the manomonce, many being repeatedly measured, in order to establish good eter reading taken. This procedure was repeated until there mean intensity values and corresponding standard deviations. was no change in the final pressure reading. The amount of A total of 46 out of 68 possible independent hkO reflections and 47 hydrogen evolved was calculated from the previously determined out of 111 possible h01 reflections were measured, within volume of the entire system. At the conclusion of the experithe limit of (sin 8)/X = 0.84 A.-l for X = 1.032 fi. The ment the gas evolved was transferred to an evacuated sample integrated intensities were evaluated by stepping along the tube and analyzed mass spectrometrically. The volume of the central lattice line through each reciprocal lattice point and system (-44 ml.) with the sample size used resulted in final counting for a predetermined monitor count of the monochromapressure readings of -190 mm. tized beam a t each 0.1" 28 interval. The background was deterThe optical spectrum of solutions of KzReHg in 2 M COa-2-free mined by taking the average neutron count over a 1' 28 interval KOH was measured from 1400 t o 210 mfi on a Cary Model 14 (10) K. Knox and A. P. Ginsberg, Inorg. Chcm., S, 555 (1964). spectrophotometer. The final spectra were obtained from a 3.44 (11) Throughout this paper, the value of the error corresponds to the X 10-3 M solution of KzReHs in a I-mm. path length cell with last significant digit in the function value. pure solvent in the reference beam. T o confirm that the absorp(12) E. Prince and S. C. Abrahams, Rev. Sci. Instr., SO, 581 (1959). tion observed was not due to impurity, the spectra of all impuri(13) S. C. Abrahams, ibid., 31, 174 (1960).
ported.gv10 The X-ray work, which was preliminary to the neutron diffraction study, located the potassium and rhenium atoms and confirmed the ratio K : R e = 2. It also showed t h a t the Re-Re distances were too great to allow the existence of Re-Re bonds, which with the chemical evidence demonstrates that the solid contains discrete rhenium hydride anions. This result eliminated one of the possible explanations for the lack of temperature-dependent paramagnetism. The present paper reports a neutron diffraction study of single crystals of potassium rhenium hydride. This completely resolves the problem of its structure and composition. The results show unequivocally that the compound contains nine hydrogens per rhenium : six H a t the corners of a trigonal prism with the Re a t its center and three H beyond the centers of the rectangular prism faces (point group symmetry D s h ) . The revised stoichiometry, K2KeHg, has been confirmed by thermal decomposition and combustion analyses. The magnetic properties previously reported, as well as the optical spectrum of the hydride which we present in this paper, are now easily understood since an unpaired spin is no longer required by the formula. A discussion of the bonding in K2ReH9is also given from the point of view of LCAO-MO theory.
360
S.C. ABRAHAMS,A.P.GINSBERG,AND K, Kxox
Inorganic Chemistry
TABLE I K2ReHgSTRUCTURE FACTORS AT 298°K
MEASCRBD A'XJ CALCULATED VALVES O F THE
hkl
FrLUasd
Fcaloda
100 200 300 400 500 600 700 800
