Oct. 20, 1961
4161
ELECTRONIC SPECTRA OF SOME TETRAHEDRAL NICKEL(II)COMPLEXES
netic moment of Co(I1) in [(CH,)aN]2-[Co(NCS)4] was meaaured in a tube carefully calibrated with HgCo(NCS)4 using the susceptibility reported by Figgis and for the latter. Thus the difference in the moments of Co(I1) in the two compounds is measured with maximal accuracy. It is seen that both the spectral data and the magnetic moments support the conclusion that in these compounds, as in those studied by Schaffer, attachment of the sulfur atoms enhances the ligand field a t the ion to which the nitrogen atoms are coordinated. Finally, we have prepared the complex Co [ (C6Hb)3PO 12 (NCS)2, that is, the triphenylphosphine oxide analog of the triphenylphosphine complex discussed above, and studied its spectrum and magnetic moment. We have also measured the spectra and treated the spectral data along with previously reported magnetic data4J in the manner outlined in ref. 5 . All of these data and derived quantities are reported in Table IV. It is evident from the comparisons exhibited there that in the phosphine oxide complex the thiocyanate ion is coordinated (20) See the Exprrimental section for more detailed discussion of the accurscy of calibrations and comparisons of moments.
TABLE IV SOMEPARAMETERS OF THE SPECTRA AND ELECTRONIC STRUCTURE OF Co(I1) IN [Co((C&&POI2X2], X = -XCS-, Co-, Br-NCS CI Br 5560 It 400 6900 3= 400 5710 f 200 "1. cm. - 1 15 300 V I , cm. -1 16,000 15 BOO 3180 r 300 1030 =t300 3270 f 150 A 8 Em -1 760 754 B', cm -1 725 0 78 0 75 0 79 B ( - B'/B)' 4 63C 4 hSC Magnetic moment, B.M. 4 4Gb A', cm. -1
156
148
B(free ion) = 967 cm authors. e Cf.ref. 4b and 6. 0
164
Measured by present
-I.
through nitrogen since both the spectra and the magnetic moments show that it is here providing a contribution to the ligand field which is far greater than that given by either CI- or Br- rather than one intermediate between those due to these two ions. Acknowledgments.--We are grateful to the Ilnited States Atomic Energy Commission for financial support under contact No. AT(30- I ) 1965. We also thank Claus Schaffer and C. Klixbull- Jorgensen for private discussions.
[CONTRIBUTION FROM THE DEPARTMENT OF CHEMISTRY, MASSACHUSETTS INSTITUTE O F TECHNOLOGY, CAMBRIDGE 39, L f A S S 1
Electronic Spectra of Some Tetrahedral Nickel(I1) Complexes BY D. M. L. GOODGAME, M. GOODGAME AND F. A. COTTON' RECEIVEDAPRIL34, 1961 The electronic spectra of eight tetrahedral complexes of nickel(II), oiz., [NiCla]z-, [XiBr,]z-, [NiIJZ-, [Ni(Ph?P0)2C1Z], [Ni(PhaPO)rBr,], [Ni(Ph3PO)rIz], [Ni(PhaAsO)~Clzland [Ni(PhaAsO)nBr~l,have been carefully studied in the reeion of the two highest energy, spin-allowed bands, YJ and YS. I t is shown that the tetrahalo ions are very sensitive to solv6iysis, even by nitromethane and acetonitrile, and that the spectra of the solvolyzed species, most probably [ KiXasolvent] -, especially in the region of Y Z , are quite different from the true spectra of the [Nixdl 2- ions. The true spectra can be obtained by measuring solid compounds either by reflectance (Y, only) or using mulls. It is then found that if excess X-, in the form of soluble salts of R4P+ or R,N+ cations, is added to the solutions the bands of the solvolyzed species can be completely, or almost completely, suppressed and the true [ N i x 4 ] spectra obtained. From these spectra the values of A and B have been calculated using Liehr and Ballhausen's complete theory. The order of A values is [NiI4lz-- [NiBr4]?- < [NiC14]2-. T h e five mixed ligand complexes are not so readily subject to solvolysis. Their spectra in the va and v2 regions are also reported and analyzed t o yield A and B values. The oscillator strengths of the bands are reported and discussed. Some remarks on the relatively low magnetic moment of [Ni14]? - are also given.
,-
Introduction It has been established conclusively during the last few years that tetrahedral complexes of nickel(I1) do exist. It has been shown by Nyholm and Gill2 and o t h e r ~ ~that v ~ , the ~ tetrahedral tetrahalonickel(I1) complex anions can exist in crystalline compounds provided the cations are large and also in solvents of only moderate dielectric constant and coordinating power. It is now also known that certain mixed ligand complexes are tetrahedral. The most numerous class of these have the general formula [NiL2X*]; they include three types: (1) those in which L is triphenylphosphine and X is C1, Br or 16s7;(2) those in which L is triphenylphosphine oxide and X is C1, Br or (1) Alfred P. Sloan Foundation Fellow. (2) N. S. Gill and R . S. Nyholm, J. Chcnr. SOC.,3997 (1959). (3) D. M . Gruen and R . L. McBeth, J. Phys. Chcm., 63, 393 (1959). (4) F . Cotton and R . Francis, J . A m . Chcm. SOC.,82, 2986 (1960). ( 5 ) F . Cotton and K. Francis, J . Inor;. N u d c a r Chem., 17,62 (1961). (6) L. M. Venanzi, J . Chem. Soc., 719 (1958). (7) F . A. Cotton, 0 . D.Faut and D. M. L. Goodgame, J , A m . Chcm. SOC.,83, 344 (1961).
-
I*; and (3) those in which L is triphenylarsine oxide and X is C1 or Br.9 The second established class of mixed ligand complexes are those of the stoichiometry [NiLX8]-, and the known members of this class include some well-characterized compounds in which L is triphenylphosphine and X is Br or I7 as well as, probably, some similar ones which are not so well characterized.' A study of the spectra of the mixed ligand complexes containing triphenylphosphine was published recently. It was shown that the ligands present in these species differ so much in their positions in the spectrochemical series that there are strong low symmetry components in the ligand fields which have quite pronounced effects on the spectra. For the complexes containing triphenylphosphine oxide and triphenylarsine oxide the low symmetry components are evidently rather small since no effects clearly attributable to them were found in the spectra*J' nor did the magnetic moments of the (8) F. A Cotton and D. M . L. Goodgamr, ihrd.. 82, 5771 (1960). (9) D. M. L. Goodgame and F. A. Cotton, sbrd , 82, 5774 (1960).
4162
D. A I . L. I
GOODGAME,
L!f. GOODGAME AND F. A. COTTON
Vol. 83
1
r\
J
Fig. 1.-spectra of ((C6H5)3MeP)2[XiCld] in the Y: region: (1) mull in hexachlorobutadiene; ( 2 ) 0,0104 dl solution in acetonitrile; (3) 0.01 iZI solution plus 0.1 M ((C6Hj)&4eP)Clin acetonitrile; (4) 0.01 JI solution in nitromethane; (5) 0.01 solution plus 0.1 &' ((C6&)3MeP)Cl in nitromethane.
compounds, which are rather sensitive in this respect, show any evidence of the existence of strong deviations from tetrahedral ~ y m m e t r y . ~ , ~ Before proceeding further with chemical studies of tetrahedral nickel(I1) complexes, it seemed desirable to carry out a fairly thorough study of the spectra of those complexes which are truly tetrahedral, that is! the [h'iX4I2- species, and those which appeared, as noted above, to approximate very closely to being truly tetrahedral, that is, the [Tc'i(PhaP0)2X2]and [ N ~ ( P ~ ~ A S Ocompounds. )~X~] For truly tetrahedral complexes, theoretical treatment is feasible and, indeed, the necessary calculations for the analysis of the spectra have been made by Liehr and Ballhausen.'O
Experimental
I 802
! 1000
1200 *nvetePglh
1400
1600
1300
mu
Fig. 2.-Spectra of ((CSH5)4P)2[NiBr4] in the up region: (1) mull in hexachlorobutadiene; (2) 0.01 i19 solution in acetonitrile; (3) 0.01 A9 solution in acetonitrile plus 0.1 it1 ((CsHj)a-n-BuP)Br; (4) 0.01 M solution in nitromethane; ( 5 ) 0.01 JI solution in nitromethane plus 0 1 d l ( ( C ~ H S ) ~ - W BuP)Br. yellow-green solutions, and was insoluble in benzene, chlorobenzene, ligroin, cyclohexane, dioxane and ethyl acetate. Electrolytic Conductance Measurements.-Electrolytic conductance measurements w r e carried out using a Serfass bridge and a conventional cell, previously calibrated n ith an aqueous solution of potassium chloride.
Results:
[Me(CeH6)3P]2 [ NiClJ [n-Bu&.;l~[P\'iI,]
158
168
26.3 25.6
The molar conductance of a AI solution of tetraphenylphosphonium tetrabromonickelate in nitrobenzene at 26.9" previously has been reported to be 61.7 Magnetic Measurements.-Measurements of the magnetic susceptibility of tetra-n-butylammonium tetraiodonickelate(I1) were made using the Gouy method as previously described .I1
Preparation of Compounds. Cesium Tetrachloronickelate(II).--il solid solution of cesium tetracliloronickelate (11) in cesium tetrachlorozincate(II), (-24 mole yo NiDia(11)), was prepared by the method of Gruen and M ~ B e t h . ~ xmc o x . magnetic X 106 cox. We are indebted to Dr. Antony Blake for this sample. Results: c.g.s.u. X 106 p e s , B.M. Methyltriphenylphosphonium Tetrachloronickelate(II).-On addition of ethyl acetate (20 ml.) to a hot solution of [n-Bu4TY]2[XiI4] 300.0 4970 612 3.47 zk 0.04 nickel chloride hexahydrate (1.58 g., 0.0067 mole) and 73.7 17820 612 3 . 2 5 f 0 . 0 4 methyltriphenylphosphonium chloride (4.18 g., 0.0132 mole) in anhydrous ethanol (12 ml.) blue crystals began to Spectral Measurements.-The reflectance spectrum of separate. After the mixture had cooled, these blue crystals tetra-n-butylammonium tetraiodonickelate(I1) was measwere filtered o f f ,washed with a 1:1 mixture of ethyl acetate ured with a Beckman DU spectrophotometer, using a standand anhydrous ethanol ( 9 ml.) and dried in vacuo over sul- ard reflectance attachment and magnesium carbonate as the furic acid. The yield vas 1.19 g. (24y0). The compound reference. A Cary Model 14 recording spectrophotometer melted a t 198". was used for all other spectra. The results are shown in Table I and Figs. 1-3. A?iuZL Calcd. for C38H36ChKiP2: C, 60.44; H, 4.81; C1, 18.18; P, 5.20. Found: C, 60.55; H, 5.16; C1, 15.76; Discussion P, 8.10. In this discussion, the reader's familiarity with Tetraphenylphosphonium Tetrabromonickelate(II).--The preparation of this compound has been reported previously.T or access to the theoretical treatment of tetraTetra-n-butylammonium Tetraiodonickelate(I1) .--A solu- hedrally coordinated Ni(I1) given by Liehr and tion of nickel iodide (2.08 g., 0.0067 mole) and tetra-n- Ballhausenlowill be assumed. We have found quite butylammonium iodide (4.92 g., 0.0133 mole) in anhydrous ethanol (20 nil.) mas evaporated to very small volume on a generally that the data reported here cannot be water bath. -1deep red solid separated, which was dried interpreted satisfactorily with any less rigorous in vacuo over sulfuric acid. The compound melted a t 115". treatment. In particular, the use of the relatively I t was hygroscopic. simple Tanabe and Suganolz matrices, in which .InnZ. Calcd. for Cj?H7dnXiAT2:C, 36.56; 13, 6.90; N, spin-orbit coupling is ignored, leads to internal 2.fi7. Found: C, 36.67; H, 6.96; N,2.91. inconsistencies and unlikely A values. I n these The compound was soluble in the cold to give red solutions in acetone, methylene chloride, chloroform and nitrometh- complexes, the X values are larger relative to the aiic. It was decomposed by alcohols, with the formation of orbital level separations than for any other metal ( 1 0 ) A. 1). Liehr a n d C. J. Ballhausen, A n a . Phys. (A'eu, Y o l k ) , 6, 134 (l03Oj.
(11) R. H. Holm and I?, A. Cotton, J . Chem. Phys., 31, 788 (1959). (12) P.Tanabe and S. Sugano, J. P h y s . SOL.J a p a n , 9, 753 (1954).
Oct. 20, 1961
ELECTRONIC SPECTRA OF SOME TETRAHEDRAL NICKEL(II)COMPLEXES
ion of the first transition series excepting only Cu(I1). Table I summarizes numerically all of our experimental observations. The strong, multicomponent bands observed in the visible spectra are designated v3 since they are assigned to the highest of the three expected spin-allowed bands. The distinctly weaker but still relatively strong bands found between 6500 and 10,000 cm.-l are assigned to the second highest of the three expected spinallowed transitions and are therefore listed in the column headed v2. The position of v l for all of the complexes can be calculated to be below our range of observation (-4500-25,000 cm.-l), namely, in the region of 3000 t o 4000 em-'. Unambiguous recognition of these v1 bands would be a difficult matter in any event, since this same region will include vibrational bands due to C-H fundamentals in the complex organic cations and solvents, as well as 0-H bands due to any moisture present. Therefore, no attempts were made to study the region of v1 for these complexes. I n the columns of Table I headed 'G and ID are listed weak bands which we assign as spin-forbidden transitions from the 31'1 ground state to certain states arising from the lG or lD states of the free ion. Spectra of the Tetrahalo Ions.--An important qualitative observation to be made concerning the spectra of these complex ions is their marked sensitivity, especially for the v2 bands, to the environment of the complex species. I n Figs. 1-3 the spectra of the three [NiX4I2- species in the region of the v2 bands are plotted so as to show the effects of changes in the environment of the ions. The same general features are found in all three cases, but the details differ from one to another. Our general interpretation is as follows. The spectra of the crystalline compounds in the form of hexachlorobutadiene mulls are assumed to be spectra of the true [XX4l2- anions. Since the upper state for v2 is not orbitally degenerate, the additional complexity of the absorption in the v2 region for the solutions cannot be attributed to splitting of the upper state in solvated species. Instead, i t must be ascribed to the presence of tetrahedral species other than the undistorted [XiX4I2ones in the solutions. It seems likely that these other species would be [NiX3(solvent)1- or similar ones in which halide ions have been displaced by solvent molecules. Since the halide ions lie a t the weak end of the spectrochemical series, it might then also be expected that the v2 bands due to solvent-containing species would be a t higher energies than those due to the [NiX4I2- ions. The [NiC14]2-Ion.-The mull spectrum of the methyltriphenylphosphonium salt of this ion shows only one rather symmetrical band a t 7272 ern.-'. I n acetonitrile solution the band maximum is displaced somewhat, to 7550 ern.-', and shows a definite asymmetry indicative of an unresolved weak band on the high energy side. In nitromethane there is a band a t 7520 an.-' and a second one of about the same intensity a t 8620 cm.-l. If, as suggested above, the higher frequency absorptions are properly attributed to solvolyzed
4163
D
Q
L
800
1000
1200
1400
1600
I800
Wovelength, m p .
Fig. 3.-Spectra of (n-BuJ)n[XI1] in the YZ region: (1) mull in hexachlorobutadiene; (2) 0.002 M solution in nitromethane; (3) 0.002 ill solution in nitromethane plus 0.1 M (n-Bu&)I; (4) 0.01 Msolution in nitromethane plus 0.5 111 (n-Bu4N)I.
species, it should be possible to suppress them by introducing excess halide ion into the solutions.13 In each of these cases, use of solvent containing 0.1 mole/l. of ((CsHb)&eP)Cl instead of pure solvent completely suppresses the high energy absorption as seen in Fig. 1. The addition of the excess phosphonium chloride also modifies the v3 absorption, though less dramatically. Now that the "true" v2 bands of [NiC14I2- are identified in the various media, it is of some interest to note the variability in their frequencies, viz., from 7272 cm.-l for crystalline ((C&,)&feP)2[NiCL] to 7750 cm.-' for CSZ[(Zn0.76h'i0.24)C14], with the frequencies for the solutions lying between these extremes. It should first be noted3 that the energy of the 3A2 state (the upper state for v2) is a very steep function of A ; indeed, the entire range of v2 values, from 7270 to 7760 cm.-l, covers a range in A of only 190 cm.-l The much smaller variability in the energy of v3 is expected theoreti200 cm.-' in A alters the cally, since a change of energy of the 3T1(P)levels by only 125 cm.-l, a magnitude barely outside the range of experi-mental uncertainties. We believe that these variations in the position of v2 must be due mainly to variations in the compressive force to which the [NiC14I2- complex ion is subjected in the various environments and perhaps also to varying degrees of polarization of the coordinated anions by the cations present. In particular, it seems quite reasonable that in the host Cs2ZnC14lattice the compression would be greater than in the ((C&)3MeP)ZNiC14 lattice because of the smaller size of c s + compared to ((CsH5)shfeP)+ and especially because of the smaller size of Zn2+which will tend to cause a contraction of the whole array compared to the dimensions i t would assume if these were dictated entirely by Ni2+ions. In this connection the interesting high pressure studies on ((C6H5)4As)ZNiC14 by Stephens and D r i ~ k a m e r ' may ~ be noted. They find that the frequency of v2 changes
-
-
(13) This technique of adding excess halide salt already has been used by Gill and Nyholmzin studying the Y O bands of several complexes. (14) D. R . Stephens and H. G . Drickamer, J . Chem. Phys., 36, 429 (1961).
r
,1...2
8 58 0 N @I
25:
-
?
i W v
0 0
2 El
0 0
8 El
? .
h .C
sg
*o c c ? c1 N
Oct. 20, 1901
ELECTRONIC SPECTR,~ OF SOME TETRAIIEDRAI, NICKEI,(II ) C o w I , w E s
4165
+
by about 20 cm.-' per kilobar, while that of about 7100 cni,-' which makes the ratio A t / & v3 changes by only about 4 cm-' per kilobar. about 0.51, while for the octahedral biomo species Our result for vz compares well with that of they give A, values of about 6900 em.-', making &/Ao about 0.49. These ratios are in reasonable Stephens and Drickamer a t 1 bar, eriz., 7500 cni.-', and with the recent report'j of a broad band with a agreement with the theoretical ratio of 4/0 = maximum between 7500 and 7600 cm.-' for a mull 0.414for pure electrostatic complexes." The p values show that the order of the halide of an unspecified [NiC1*I2-salt. The [NiBr4I2Ion.-For the ((C&,)dP)- salt ions in the nephelauxetic series is C1 < Br < I, of this ion, the mull spectrum shows the V J band in order of increasing cloud-expanding effect. Our results are in satisfactory agreement with to be a t 7040 cm.-'. I n pure acetonitrile, there is only one band in this region which has its tnaxi- those of Weakliem,l8 who finds for Nil+ in Cs2inuni a t SO45 cni.-' with a rather long low-energy ZnC14 A = 3550, B = 723 cni.-' and for Ni2+ in tail, which might be due to the presence of a weak CszZnBr? A = 3300 and B = 603 ctn-'. The Magnetic Moment of [1JiIIj2-.-Iri their broad band around 7000 cm.-l. When acetooriginal nitrile containing 0.1 mole/l. of ( ( C B H ~ ) ~ - ~ - B U P ) B ~ report of the preparation of the [NiI4l2-ion, is used instead of the pure solvent, a remarkable as the methyltriphenylarsoniutn salt, Gill and tratiformation (Fig. 2) takes place, the spectrum Nyholm2 gave for p e e a t room temperature 3.49 B. now consisting of only one symmetrical band with M., a value well below those found for other a maxirnum at 6095 ern.-'. In pure nitromethane, tetrahedral nickel(I1) complexes in which the the strongest absorption is a band centered at four ligand atoms are identical or differ little i n 5475 cm.-', but there is a distinct band at 6993 their positions in the spectrochemical series cm.-'. On using a 0.1 111 solution of ((C6&).1- (3.7-4.0 B.M.2J~9).They suggested that this ground state n-BuP)Br in nitromethane instead of the pure might be due to mixing of the solvent, the high-energy band is reduced to a inere with the 3A2 excited state, thus diminishing the shoulder and the major absorption has a niaxiinuni orbital angular momentum in the former. Our at 6990 an.-'. We therefore conclude that v2 for analysis19 indicates that there are no matrix elements between the TI and A2 states; moreover, [NiBr4I2has a frequency of 6090-7040 cni.-'. The [NiI4I2- Lon.-The mull spectrum of the results given in Table I1 show that there is 110 [n-Ru4N]2[NiI4] shows a symmetrical 1 2 band a t sharp decrease in A for [NiI4lZ- as compared to its 6054 ern.-'. In nitromethane solution this band value in [NiClrl2- and [NiBr4I2- such as would is not discernible at all, being replaced by one a t be necessary to explain the appreciably lower 8300 cm. -l. However, with progressive increase value of pee by this or any other mechanism dein the concentration of (n-RudN) ]I the true v 2 band pending on a mixing of states inversely proporat -7000 cni.-l grows in with concomitant de- tional to the first power of A. Two other explanacrease in the intensity of the spurious band. It tions for the low value of /.iefi in [NiI4I2- h a w been was found that acetonitrile attacked this compound considered here. so severely as to give a turbid solution, so no atFirst, if it be assumed that the effective value of tempt was made to study it in this solvent. A, the spin-orbit coupling constant, is 20-25% Electronic Structure Parameters.-These com- higher in [Ni14l2- than in the other complexes putations were made using the results of Liehr and (which is equivalent to assuming that the iodide Ballhausen'o in which X is assigned the value ions cause less of a decrease in X from the free - 273 cm.-' and F:! is taken as 14Fl (equivalent ion value than do C1- or Rr-) the value of peff to C/B = 3.9). The parameters A ( = 10 Dg) and a t room temperature would be 3.5 B.M. acB were adjusted to give an exact fit for vJ (the cording to the theoretical results given by FiggkZ0 energy being estimated as the center of gravity There is, fortunately, a direct test of this hypothof all components of the absorption in the visible) esis, which consists in measuring . u e ~a t a much and v l Table I1 gives the values of A and B lower temperature. The theoryz0 shows that at along with a comparison of the observed and cal- liquid nitrogen temperature the moment should culated positions of some small bands which may be -1.8 B.M. We find for the tetra-n-butylbe assigned as spin-forbidden transitions to upper ammonium salt of [NiI4I2-, however, a drop to states arising from the lD and 'G states of the free only 3 25 B.M. a t 74'K. SO that this explanation is ion These acquire intensity by mixing, win spin- also untenable. Sitice the [NiI.,]2- ion seems the orbit coupling of the upper "singlet" states with least stable one, and also the most deformable, nearby "triplet" states of the same symmetry. we believe that the most likely explanation of the The A values lend to the following order of the low moment for Ni(I1) in [NiI4l2- is that distorhalide ions in the spectrochemical series: Ition of the complex in the crystal causes considerRr- < C1-. The closeness of I- and Br- here is able splitting and separation of the ground state noteworthy. They are close (e.g., in [Co13r4l2- multiplet (32'1, T4and 31'5 of 3T1(F)). It has and [CoI4I2-) but more often with I- < Br-. already been shown for the [Ni((CeHs)3P)2X2j It is interesting to compare the values of A in compounds that the introduction of low-symthese tetrahedral nickel(I1) complexes ( A t ) with metry components into the tetrahedral fields about those reported by Asmusseii and Bostrup16 for (17) C J Ballhausea, Kg Dmnke V,denPkob, Set knb Mat -fys. octahedrally codrdinated nickel(II), (Ao). For M e d d , 29, No 4 (1954). (18) H A Weakliem, R C A Laboratorles, prlvate communication octahedral chloro species, they report A, to be (19) In the group T d , L.S transforms os T I and the direct product
-
-
(15) S . Buffagni and T M Ilunn. Nntxre, 188,937 (1960) (16) R 1%' Asrnus*en and 0 Bostriip, .\cla Chrm 9cnnd
(1957).
11, 715
+
AI X Ti X Ti = AI 4- E -!- TI TI Hence, matri7 elements of the sort (A, L.S TI) are idcntlcaiiy zero (20) D. N. Figgis, Nulure. 183, 1565 (1058)
I
1
D. M. L. GOODGAME, 11. GOODGABSE AND I?.
41GG
LIGAND FIELD STRENGTHS, RACAHPARAMETERS
TABLE I1 CBLCGLATED
AXD S O M E
Vol. s3
COTTOX
S P I S - F O R B I D D E N TRANSITIOXS FOR THE
[Sixdl'-
IOSS
I-pper states of spin-for hidden transtitions Bb ,-.----iD(r6)----..---+-I--(=R'/B) Obsd. Calcd. Obsd. Calcd.
hIedium
cm.-1
4-
B',5 cm.-1
Cs*Z11C14 Mull 0 . 1 A4 [MePhsPICl in CH3CN 0.1 M [MePhsPlCl in CH3S02 Mull 0 . 1 M [Ph3-n-BuP]Br in CHICN 0 . 1 M (Ph3-n-BuP)Brin CHBXO~ hlull
3770 3580 3610 3610 3380
750 74 6 765 765 691
0.728 ,724 ,742 ,742 ,071
3380
713
692
3380 3350
713
,692
2 72
. .JJJ
11400 11600 11630 11630
-lozoo 10680
.--Not10720 obsd.
11340 11270 11490 11490 10600
-20620 -19800 Kot obsd. Not obsd.
20220(r3] i9i80(r3)
moo
i83o0(r3)
.. ..
..... ... .,
10820 K o t obsd.
..... .,
10820 S o t obsd. . . . -13640
13280(r6) i4890(r4)
14810
0 . 5 M [n-Bu4S]Iin CH3N02 3380 h-ot obsd. . , , Kot obsd. ... .. ., a A and B' were calculated from the energies of y 2 and v j and the calculated energies of these transitions therefore match the experimental values exactly. b B for the free Ni(11) ion equals 1030 cm c B' was not determined since the v3 band is too strong to measure under these conditions.
TABLE 111 LIGANDFIELDSTRENGTHS, RACAHPARAMETERS A N D SOME CALCULATED SPIS-FORBIDDES TUXSITIONS FOR (hf = p,&; X = C1, Br, I ) ((CsH5)&fO)2X2]COMPLEXES 3b
Compound
Medi urn
h-i ((C6H5)3P0)2CIz Mull K ~ ( ( C ~ H S ) ~ P O ) ~ BMull ~Z Acetone Ni((C6Hs)aPO)nIs hlull
2s
B'a
3580 3460 3500 3440
800 780 790 736
(=B'/Rj
0.777 ,757
.767 .715
-----1D(r5)-
Upper states of spin-forbidden transitions
Obsd.
12120 11930 S o t obsd. X o t obsd.
G---
I----,
Calcd.
11860
11640 I1790 11120
Obsd.
22420 20410 20490 18690
THE
[Si-
-
Calcd.
20900(r3) 20370(r3)
20600(r3j 18430(I'd ) 193n(rj) 18880(rl)
770 .747 S o t obsd. 11570 18620 C6HjCl 3420 Acetone 3500 696 ,676 Not obsd. 10G30 Kot obsd. ....., . Ni((C6Hj)3AsO)2C12 hlull 3920 790 ,767 12170 11870 21370 21270(r3) C&Cl 3860 847 ,822 12350 sh 12430 S o t obsd. .. .. ... Ni( (CGHj)3A~0)2Br2 hfull 3750 778 ,755 11920 11680 -20660 sh 20730(r3) CaHhCl 3730 810 ,786 -11960 sh 120*50 Not obsd. ..... .. 5 A and B' were calculated from the energies of V? and ya and the calculated energies of these transitions therefore match t h e experimental values exactly. b B for the free Ni(I1) ion equals 1030 ern.-'.
Ni(I1) causes marked reductions in the magnetic spectrochemical series the order of the halide ions moments.' I n regard to the susceptibility of the is I 7 Br < C1 and, a little less neatly, due to the [NiI4I2- ion to deformation by its surroundings, small variations from one medium to another, that results obtained in This Laboratory for [CoI4]?-, in the nephelauxetic series the order is C1 < Br < I. which must surely be rather similar in this respect, These results are in agreement with the accepted show that it also exhibits marked sensitivity to its orders. Furthermore, comparison of the results environment. for the two analogous pairs of phosphine oxide The Neutral Complexes.-In none of these and arsine oxide complexes shows that the order compounds were any major medium effects noted, of these two ligands in the spectrochemical series the media used being the pure crystalline com- is, decidedly, (C&,)aPO < PhaAsO, while in the pounds, acetone and chlorobenzene. Hence the nephelauxetic series there is no difference outside interpretation of the data proceeds straightfor- experimental uncertainties. Finally, comparison wardly. The A and B' values, estimated from of the data for the [Si((C6H6)aP0)2X2] compounds the energies of the v2 and v a bands, using Liehr with data for the [NiX4]?- complexes shows that and Ballhausen's equations,1° are given in Table (CeH5)sPO is a t about the same position as C1- in I11 along with a comparison of the calculated and the spectrochemical series, and weaker than C1- (;.e., observed band positions for some spin-forbidden (C6HJsP0< C1- < Br- < I-) in the nephelauxetic series. Insofar as comparisons can be made with bands. ~ This series of compounds is the most extensil-e analogous or related complexes of C O ( I I ) , *these one of its type so far studied in this way and conclusions are substantiated by the Co(I1) work. several definite conclusions about the ligands can Band Intensities.-For most of the complexes be drawn. First, intercomparison of the A and B' studied we have made graphical estimates, which values for the three triphenylphosphine oxide should be accurate to within =k lUyG, and probably complexes and for the two triphenylarsine oxide less, of the oscillator strengths of the observed complexes demonstrates quite clearly that in the bands. This was not done for [ N ~ ( ( C G H S ) ~ P O ) ? C ~ ~ ] , since only reflectance and mull spectra are avail(21) F. A. Cotton, hl. Goodgame and D. h l L. Goodgame, J . A m able for this complex, nor for [NiI4I2- because Chem. SOL, 83, in press
Oct. 20, 1961
4167
HETERO-BINUCLEAR CHELaTES O F COPPER(II)AND SILVER(1) TABLE IV STREXCTHS OF THE ABSORPTION BAI~DS' OSCILLATOR
Compound
Solrent
CH3CN CH3Iion 0 . 1 ill ((c6Hj)~MeP)Clin CHzNOz ((CfiHd4Ph [NiBrd CH3Xo2 CH3CN 0 . 1 111 ( ( C 6 H a ) S - ? ~ - B ~ in P ) CH3CPi Br in'i ((CsH.&P0)2Br?I Acetone [Ni((CsH5)3PO)zI21 CsHjCl C~H~CI [%((C~H&ASO)ZC~~] [Ni((CsH&AsO)&'d C&,CI a The estimates include all absorption in the region of the v i concerned. sorption known to be due to solvolyzed species is included.
heavy ultraviolet absorption trailing off into the region of v3 makes accurate intensity assessment difficult. The most striking aspect of the intensity data is the relatively high intensity of the bands of the tetrahedral nickel(I1) complexes as compared to the band intensities for octahedral complexes of nickel(I1). The ratio is about l o 2 . This effect is quite general, having been found for Co(II),21s22Mn(II)23,24and Cu(II),17 and Ballhausen and Liehr have discussed the possible reasons for it.*& The intensity ratios show that (22) F. A. Cotton and M. Goodgame, J . A m . Chem. SOC.,83, 1777 (1961). (23) D. M . L. Goodgame and F. A. Cotton, J . Chem. SOC., Aug.(l961). (24) F. A. Cotton, D. M. L. Goodgame and M. Goodgame, J . A m . Chem. SOC.,in press. ( 2 5 ) C. J. Ballhausen and A. D. Liehr, J . Mol. Spect., 2, 342 (1958); 4 , 190 (1960).
[COXTRIBUTION FROM
THE
f Y?
fva
((CsHd3MeP)~[NiCl4l
fu3/fV2
2 30 x 10-3 2 . 0 1 x 10-4 2.21 x 10-3 2.35 X 2 . 2 4 x 10-3 1.42 X lo-' 3.12 x 10-3 4.85 x 10-4 2.97 X 4.28 X lo-' 3.15 X 2.61 X 1.59 x 10-3 1.63 x 10-4 2 06 x 10-3 1 08 X 1.52 x 10-3 1 21 x 10-4 1.79 x lo-' 2.92 x 10-3 Thus, for the v2 bands, the intensity
11.4 9.40 15 8 6.44 6.94 12.1 9.75 19.1 12.6 16.3 of the ab-
v3 is generally 10 to 20 times stronger than v2. This is in agreement with the fact that in the strong field limit v3 remains a one-electron transition while v2 becomes a two-electron transition. It is also noteworthy that the intensity of v 3 is relatively insensitive to solvolysis effects, whereas for v2 the solvolyzed species are more strongly absorbing than are the [NiX4I2-ions.
Acknowledgments.-The generous financial support of the United States Atomic Energy Commission (Contract No. AT(30-1)-1965) is gratefully acknowledged. \Ire thank Professor H. G. Drickamer and Dr. Herbert TVeakliem for informing us of some of their results in advance of publication and Dr. A. I). Liehr for sending us some numerical results of the Liehr and Ballhausen calculations.
POLYTECHNIC INSTITUTE OF BROOKLYN, BROOKLYS,~ ' E WYORK,A K D MICHIGAN,ANNARBOR,MICHIGAS]
THE
UXIVERSITY OF
Hetero-binuclear Chelates of Copper(I1) and Silver(1) BY
c. H. LIUA4ND CHUIFANLIU RECEIVED MAY4, 1961
Interaction between silver(1) ion and bis-(pyridine-2-aldoxime)-copper(II)ion has been investigated by potentiometric measurements. The hetero-binuclear chelate involved has also been isolated as the perchlorate and examined by infrared spectroscopy.
In a previous investigation,l it has been shown groups. The present work is concerned with the that monohydrogen bis-(pyridine-2-aldoxime)-cop- interaction between the complex ion and silver(1) per(I1) ion has the structure ion. This interaction has been studied by measuring the two competing equilibria r1HC+
I
