Magnetic Circular Dichroism of Se;' and Te;' BY P. J. STEPHENS* Dept. of Chemistry University of Southern California Los Angeles, California U.S.A. Received 8th September 1969 Magnetic circular dichroism is used to investigate the species formed in sulphuric acid solutions of selenium and tellurium. The data strongly support the identification of absorption bands in these solutions with square-planar Sei+ and Te:+. These bands are shown to be in-plane polarized, consistent with assignment to x +x transitions. We present an application of magnetic circular dichroism (MCD) to the charac-terization of new chemical species. Magnetic optical activity has not previously been employed for this purpose unless we include its use in the identification of colour centres.l'* This can be attributed both to the lack of adequately sensitive instrumentation and to the nature (and complexity) of the phenomenon.While it is unlikely that magnetic optical activity will become widely used in detective chemistry even to the extent of visible-u.-v. spectroscopy we wish to demonstrate that in special cases it can be of considerable value. Gillespie and co-workers have shown that solutions of selenium and tellurium in oxidizing solvents such as oleum and HS03F + S206F2 contain a variety of polymeric cations. The species Sei+ Sei+ and Tei+ have been identified from cryoscopic, conductometric and magnetic susceptibility meas~rements.~* O Solid salts of Sez+ have been obtained l 1 and X-ray analysis of Se4(HS20,)2 has shown this ion to be square-p1anar.l2 Raman and i.-r.studies are consistent with this structure l3 and the Raman spectra of solutions containing Te:+ have been interpreted in terms of an analogous geometry.1° Visible-u.-v. spectra have been reported for Sei+, Sei+ Tei+ and for solutions obtained on oxidation of Tei+.9* lo Both Sei+ and Tez+ show one low-lying reasonably intense band. That in Set+ has been attri-buted l2 to a transition between n levels deriving from Sep,-orbitals. We have studied the MCD of selenium and tellurium in sulphuric acid solutions, with a view to further characterization of the species obtained. The data provide strong support for square-planar Seq+ and Te:+. Historical precedence for our experiments can actually be claimed by Bizette and S ~ h k r e r ~ ~ who in 1937 studied the magnetic optical rotation of solutions of tellurium in concentrated sulphuric acid in the region of the visible absorption band.However despite a field of 50 kG their data were not good and no firm conclusions were drawn. EXPERIMENTS Selenium dissolves in 30 % fuming HzS04 to give a green solution. On standing or on Further oxidation causes all addition of oxidizing agents the solution becomes yellow. * Alfred P. Sloan Research Fellow. 4 P . J . STEPHENS 41 $ I 1 I I L I 250 350 450 5 50 650 7 50 1 (nm) ‘1.0 A (nm) FIG. 1 .+u) Absorption spectrum and (b) MCD of Se/30 % fuming sulphuric acid (- 3.212 g/l. ; -- 1.036 g/l.). colour to disappear. The absorption spectra and MCD of fresh green solutions and a yellow solution obtained using (NH4)2S208 as oxidizing agent are shown in fig.1 and 2. Tellurium dissolves slowly in concentrated H2SO4 and rapidly in 30 % fuming H2SO4, giving purplish-red and wine-red solutions respectively. On addition of oxidizing agents the solutions turn orange yellow and then colourless. The absorption spectra and MCD of a solution in concentrated H2S04 and solutions in 30 % fuming HzSO4 with and without added (NH4)2S208 are shown in fig. 3 and 4. MCD was measured in Prof. C. Djerassi’s laboratory at Stanford University with the assistance of Dr. E. Bunnenberg and Nrs. R. Records on a Durrum-Jasco ORD/CD/UV-42 MCD O F Se;+ AND Tei+-n -0.5 1 - 1.0 \ I V b) FIG. 2 . 4 ~ 2 ) Absorption spectrum and (b) MCD of Se (NH4)2S208/30 % fuming sulphuric acid (0.980 g/l. and 14.880 g/l.respectively). equipped with a Lockheed superconducting magnet. In all cases the field was 49.5 kG. Absorption spectra were taken on a Cary 14. The noise in all spectra was negligible and is not indicated in the figures. We report absorption data in terms of optical density O.D., and MCD as A(0.D.) = (O.D.)L-(O.D.)R the magnetic field being in the same direction as the light beam. The optical path-length in all experiments was 1 mm. All solutions used were freshly prepared excepting the yellow selenium and concentrated H2S04+ tellurium solutions which were 12 h and 23 days old respectively. Note added in proof The MCD data in fig. 26 and the right-hand part of fig. 36 are in error and should be multiplied by 1.26 P . J. STEPHENS 43 r' n e .- g -1.0 c.l 2 2 % I I I I I 2 50 350 450 550 650 A .- 1.0 A (nm) FIG.3.-(a) Absorption spectrum and (b) MCD of Te+ conc. sulphuric acid (concentration unknown but < 3.8 g/l.). DISCUSSION The spectrum of the yellow selenium solution (bands at 410 and 320nm) is essentially identical to that given by Barr et aL9 for a 4 1 Se S206F2 solution in HS03F and attributed to Se:+. A lower limit for E,, of the 410 nm band obtained assuming all the selenium present as Sez+ is 5900. This compares with 7400 given by Barr et al. for H2S207 solutions. The green selenium solutions contain the yellow species (and this cannot be eliminated by reducing the SO concentration). However if its contribution is subtracted the spectrum remaining is the same (bands at 685 470 and 295 nm) up to -280 nm as Barr et al.for a 10.25 1 Se S206F2 solution in HS03F,9 which was ascribed to Seg+. At shorter wavelengths than used by Barr et al. we observe an additional band (-250 n d . Since neither the wor 44 of Barr et aL9 nor OUT studies found any evidence for other species this band can also be attributed to Sei+. All tellurium solutions show bands around 510 420 360 300 and 250nm, of relative intensities varying with age and with SO3 and (NHJ2S208 concentration. The spectra of solutions of Te and 2 1 Te S206F2 in HS03F given by Barr et a2.1° show bands at 510 and 430nm and 520 420 360 300 and 250 nm respectively. MCD OF Se$+ AND Teil-(m) FIG. 4.-(a) Absorption spectra and (b) MCD of Te/30 % fuming sulphuric acid (- 2.244 g/1; --.0.304 g/l.) and Te (NH4)2S20~ + 30 % fuming sulphuric acid (- - - 2.244 g Te/l.). The former is attributed to Tei+ and the latter to a more highly oxidized species. Our spectra show no additional bands and appear interpretable as a superposition of these two spectra in varying proportions. However our data do not rule out the presence of more than two species ; in particular no isosbestic points were observed. From data on concentrated H2S04 solutions we find a lower limit for E,, of the 510 nm band of 5000 assuming all tellurium present as Tei+ (Barr et a2.1° give 2150). MCD is observed for all bands of the selenium solutions. Indeed for the green solutions MCD indicates the bands more clearly than the absorption spectrum an P . J . STEPHENS 45 shows there to be a band at 580 nni which is not resolved in the spectrum.The most noticeable feature of the MCD is the large effect in the 410 nm band changing sign through the band; this will be shown below to be strong support for assignment of this band to square-planar Se:+. The MCD of the tellurium + concentrated H2S04 solution contains bands at 460 400 360 and 300 nm which are not clearly resolved in the absorption spectrum. The data for the fuming H2S04 solutions are more complicated and require extension to a wider range of solutions for their correlation with the absorption spectra to be unambiguously established. The distinctive feature of the MCD data is the behaviour of the 510nm band which is identical in form to that of the 410nm Se:+ band. This correspondence leads to the conclusion that these bands arise from similar transitions and species.The MCD thus provides strong support for the identification of the 510 nm tellurium band with square-planar Te:+ as assigned by Barr et a1.I' Brown et aZ.12 have proposed that the highest occupied and the lowest unoccupied orbitals of square-planar D4h Sei+ are 71-levels arising from the Se 4p orbitals as shown in fig. 5. On this basis in the absence of spin-orbit coupling the ground state is 'Alg (in agreement with the observed diamagnetism 9* 1 5 ) and the lowest eg-+bzu excitation leads to 'E and 3E excited states. Brown et al. assign the 410 nm Se:+ band to this transition. Since the Se 4p spin-orbit coupling constant is 2000-3000 cin-l excited-state spin-orbit interactions would at first sight be expected to be visible.However spin-orbit interactions within n-electron states of aromatic hydrocarbons are very small due to the form of the n-orbitals and the spin-orbit operator.16 Energy - eg aeu FIG. 5.-x-orbitals of Se:+. Crosses indicate electrons in ground state of Se:+. Identical arguments hold for Se:+ and lead to the expectation that the 3E splitting and 1E,-3E mixing are also small. In this case since other off-diagonal spin-orbit interactions can be ignored only the lAlS-+lE, transition is strongly allowed and only one absorption peak is expected (excluding Jahn-Teller effects). The observed intensity is consistent with the assignment to an electric-dipole-allowed transition. This picture does not account for the weak 320 nm band.If it is due to the same species an additional orbital close to either the e,(n) or b,,(n) orbitals, must be present. However this orbital does not give rise to strong absorption below 40,000 cm-' and without a better theoretical calculation we shall not speculate further on its nature. The important feature of the MCD of the 410 nm band is its large first moment about the mean absorption frequency. General expressions for the zeroth and first moments of the absorption and MCD of a transition are given elsewhere and here we quote only those results relevant to the present study. For a dilute liquid solution of diamagnetic molecular species the zeroth moment of the absorption band and the zeroth and first moments of the MCD of an allowed transition A+J are given b 46 MCD OF Sea-1- AND Teit O.D.a dv = -9O(A+J), (O.D.) = J;- 2 (A(O.D.)> = -dv = -a9Y0(A-+J)H IA( O. DJ (v- v O ) dv - a I1 = - {d 1 ( A + J ) + [S 1 ( A + J ) - h Y O&?O( A+ J ) ] ) H . A11 integrals are over the band. a is a constant depending on the concentration of the species the optical path-length the " effective-field '' correction and fundamental constants. v" is the frequency about which the first moment is taken. H is the magnetic field strength. go and d1 are given by g O ( A - 4 = c I ( A I m I Jn> I * 1 where m and p are the electronic electric and magnetic dipole moment operators, Jn runs over the excited electronic states contributing to the band and the wave-functions are appropriate to the free molecule at a single chosen geometry.go and &ll are parameters depending on off-diagonal matrix elements of p between A and J and other molecular states. Their detailed form will not be needed here. Of the assumptions involved in eqn (1) and (2) we note only that the effects of the solvent are taken into account to first order-i.e. the molecule-solvent interaction is included explicitly but assumed to be weak-and Jahn-Teller complications are fully catered for. The states J1 need not be exactly degenerate at any molecular geometry. The matrix (JA I p I J1.) is zero when there is just one J state. Hence, dl is non-zero only when the J manifold contains more than one absorbing state. We now assume that the 410 nm band is due entirely to a single species. Values for the moments defined in eqn (1) are given in table 1.The first moment is calculated withv" as the mean absorption frequency making J(O.D./v) (v - v")dv = 0. Corrections for the overlap with the 320 nm transition have been included. TABLE 1 Se a Te 0.25 0.28 p+v -0.48 10-4 -0.32 10-4 V" 24 080 19 290 cm-~ ~ ) ( v - v " ) d v -0.38 - 0.33 cm-1 r c 3000 2700 cm-I I 0.66 0.50 cz data for the 410 nm band ; b data for the 510 nni band ; C width at half-height. An outside limit on the contribution of the 99 terms to the MCD first moment can be placed at BoT where r is the band width and this is iikely to be a considerable over-estimate. From table 1 it can then be seen that the major contribution must come from d, P. J. S l E P H E N S 47 To calculate dl we write where all wave-functions refer to the D4, ground state equilibrium geometry; we ignore off-diagonal spin-orbit mixing other than of IE and 3Eu and the (A 1 SM,a) matrix diagonalizes the spin-orbit interaction within the E, 3Eu manifold.But dl is invariant to a unitary transformation on I Ja). Hence the I lEua) I 3EUMsa) basis can be used instead of I JJ to evaluate dl and it is unnecessary to find the (A I SM,a} matrix. Now (lAl 1 m I 3E,Mscx) = 0 (all M,,a). Only the lE, functions thus contribute to dl-i.e. d is unaffected by the presence of the triplet states. With (lAlg I in I ' E J ) = ('Al I my I ' ~ y ) = in, (4) (lE*,x I P I l E Y ) = -BM('Eux I LZ I l&Y) = - iZ&f, where Ex and Ey states transform like x and y and /IM is the Bohr magneton we then obtain dl = m21pM go = 2m2 dlpQ = 31PM.(5) If the a terms of the MCD first moment are neglected eqn (1) and (5) permit I to be evaluated with the result given in table 1. This calculation does not require any knowledge of the concentration optical path-length " effective-field " correction or the value of m. In particular it does not require the assumption that all selenium is present as Se:+. Such calculations have been carried out for several types of system.18-22 For 7c-electron states of planar molecules expressed in terms of p z atomic orbitals matrix elements of L, reduce to two-centre integrals and all one-centre terms vanish.l8. The two-centre integrals are sensitive to the form of the radial function of the orbitals. In view of the crudity of the n-electron model and the uncertainty in the quantitative form of orbitals occurring therein it does not appear worthwhile to consider calcula-tion of I in more detail.From the previous experience the magnitude obtained from the data is very reasonable. We can now see to what extent the MCD actually provides proof of the Sei+ structure and the absorption band assignment. Recognizing the large separation of the ground and excited states from other states (neglecting the 320 nm band excited levels) their spin-orbit mixing with other states can always be ignored to a good approximation. Then the first moment can only be explained if in the zero-spin-orbit-coupling limit the excited-state manifold contains more than one singlet state. This can arise either from true or accidental degeneracy.The latter is reasonably uncommon. Excluding this it then follows that the species has sufficient symmetry to allow electronic degeneracy. This limits the allowed molecular point groups. Accepting the molecular formula Se, such groups as D4h (square-planar) Td (tetra-hedral) and Dmh (linear) are possible. The existence of the first moment does not in itself allow a choice between them. In summary then as long as accidental degeneracy can be excluded the MCD is strong evidence for a highly symmetrical I can be calculated a priuri given explicit lEU wave-functions 48 Se;-' geometry and is consistent with a square-planar structure. Accepting the latter and a IA, ground state the symmetry of the excited singlet state of the 410 nm band must be E, since this is the only degenerate u representation in D4,,.The transition is therefore in-plane polarized. This is strongly suggestive of a 71-71 type of transition though not conclusive proof. The absorption spectrum and MCD of the 510 nm Te:+ band are essentially identical to those for the Se2f 410 nm band and we interpret them analogously. Spin-orbit coupling is larger in Te but spin-orbit energies should still be much smaller than the separations of other states from the ground and excited states of the band. Values for moments and parameters derived therefrom are given in table 1. They are close to those obtained for Sea+. We therefore draw the same conclusions as for Sei+. The observation of degeneracy in both Sei+ and Teif strengthens the assumption that it is caused by molecular symmetry and is not accidental.The data for both species are thus together strong support for their proposed square-planar geometries. CONCLUSION The existence and structure of Sei+ appear to be reasonably established by the work of Gillespie et al. Our studies put this latter species on a firmer basis. We have added little to the existing knowledge of Seg+ or the oxidized tellurium species since the MCD of the former is highly unsingular and that of the latter is both hard to disentangle from the Tei+ contribution and rather complex. Further work on these systems is required. However Tei+ is not so well characterized as yet. I am deeply grateful to Dr. Bunnenberg and Mrs. Records for their generous assistance in the experimental part of this work and to Prof.Djerassi for his hospitality. I also owe thanks to Dr. David Parker for help in the initial stages and to Prof. Anton Burg for the use of his dry-box. The work was supported in part by grants from the National Institutes of Health and the Alfred P. Sloan Foundation. Y . M. d'AubignC and J. Gareyte Compt. rend. 1965 261,689. F. C. Brown B. C. Cavenett and W. Hayes Phys. Letters 1965 19 167. J. C. Kemp W. M. Ziniker and J. A. Glaze Phys. Letters 1966 22,37. J. C. Kemp W. M. Ziniker and J. A. Glaze Proc. Brit. Ceramic SOC. 1967 109. 6. Gehrer and H. Langer Phys. Letters A 1968 26,232. H. Paus and F. Luty Phys. Rev. Letters 1968 20,57. ' J. A. Glaze and J. C. Kemp Phys. Rev. 1969,178,1502 1507. * B. C. Cavenett W. Hayes I. C.Hunter and A. M. Stoneham Proc. Roy. SOC. A 1969,309 53. J. Barr R. J. Gillespie R. Kapoor and K. C. Malhotra Can. J. Chem. 1968 46 149. l o J. Barr R. J. Gillespie R. Kapoor and G. P. Pez J. Amer. Chem. SOC. 1968 90,6855. l 1 J. Barr D. B. Crump R. J. Gillespie R. Kapoor and P. K. Ummat Can. J. Chem. 1968 46, l 2 I. D. Brown D. B. Crump R. J. GiIlespie and D. P. Santry Chem. Comnz. 1968 853. l3 R. J. Gillespie and G. P. Pez Inorg. Chem. 1969 8 1229. l4 H. Bizette and M. SchCrer Compt. rend. 1937 204 1931. l 5 We have also looked at the e.s.r. of all solutions studied here and found no evidence for radicals. l 6 D. S. McClure J. Chem. Phys. 1952,20,682. l7 P. J. Stephens J. Chem. Phys. 1970,52,3489. A preliminary account of this work appeared l 8 P. J. Stephens W. Suetaka and P. N. Schatz J. Chem. Phys. 1966,44,4592. l 8 P. N. Schatz A. J. McCaffery W. Suetaka G. N. Henning A. B. Ritchie and P. J. Stephens, ** P. N. Schatz P. J. Stephens G. N. Henning and A. J. McCaffery Inorg. Chem. 1968,7 1246. 21 P. J. Stephens P. N. Schatz A. B. Ritchie and A. J. McCaffery J. Chem. Phys. 1968 48 132. 22 P. J. Stephens A. J. McCaffery and P. N. Schatz Inorg. Chem. 1968,7 1923. 3607. in Chem. Phys. Letters 1968 2,241. J. Chem. Phys. 1966 45,722