ELECTRIC DIPOLEM O M E N T O F ALUMINUMTRIS(2,d-PENTANEDIONATE)
3439
The Electric Dipole Moment of Aluminum Tris(2,P-pentanedionate) l by Ralph D. Nelson, Jr., and Charles E. White2 Chemistry Department, Middlebury College, Middlebury, Vermont 06763 (Received April 7 , 1969)
The molar refractions of benzene solutions of aluminum tris(2,4-pentanedionate) are reported. A far-infrared refractometer using an HCN laser (337-pm wavelength) is described. The electric dipole moment for esu cm, confirming X-ray structural aluminum tris(2,4-pentanedionate) is determined to be 1.1 x analyses which indicate a polar structure. Previously observed unusually low values of dielectric relaxation time for these chelates are explained in terms of torsional jumps from one polar conformation t o another with reversal of direction of dipole moment.
Introduction In 1938 Finn, Hampson, and Sutton measured the total molar polarization PT for several metal 2,4pentanedionates and stated that if there were any contributions to PT“due to fully chelated molecules being polar, then modern structural theory would be proved seriously in error.”3 Since then structural theory has changed and considerable data have been presented on these chelates, but the question of their polarity has not been clearly resolved. X-Ray diffraction studies have established the molecular structures of numerous 2,4-pentanedionate (Ptdn) chelates, commonly known as acetylacetonate~.~It will be shown below that most such structures are necessarily polar by virtue of their symmetries. Dielectric permittivity and loss measurements in the microwave region have indicated polarity, but uncertainty regarding the frequency dependence of dielectric loss precluded accurate estimates of dipole Now that apparatus is available for determining directly the far-infrared refractive index and consequently the far-infrared molar refraction RFIR, the dipolar orientation polarization PO can be established with only two measurements and the relationship PO= PT - RFIR. The dipole moment 1.1 is then computed from p2 =
9kTPo/(4~No)
(1)
where T is absolute temperature, k is the Boltzmann constant, and No is the Avogadro number.* The assumptions involved are (1) that the measurement frequency lies below all the absorption bands due to electronic and vibrational transitions, (2) that the electronic, vibrational, and dipole reorientation processes are the only ones involved in PT,and (3) that the relation between PO and 1.1 holds at the concentration in question, with no significant orientation correlation between the polar molecules. The spectra of numerous metal Ptdn chelates indicate that the first assumption is met.g Recent work has indicated a process assigned to collisionally induced dipoles.1° Our model for these chelates consists of
collisionally reoriented dipoles, satisfying the second assumption. The third assumption can be accepted if there is a linear relation between solution polarization and mole fraction of solute, as was found for the system studied here.
Experimental Section The experimental apparatus is shown in Figure 1. Radiation from the HCN laser enters the solution at an angle of incidence of zero, so the direction of propagation is changed only at the mirror on the bottom of the cell and at the exit interface, where Snell’s law applies. The goniometer mirror is turned until a maximum intensity is received at the detector, indicating proper optical allignment. The refractive index is related to the goniometer readings by the relation niiq
=
2(OP - &is) 2($ - &ir)
n a i r COS
COS
(2)
where n is refractive index, 8 is the goniometer reading, and p is for the apparatus set up with the 45” mirror (1) Presented at the First Northeast Regional ACS Meeting, Boston, Mass., Sept 1968. (2) This paper represents part of the work submitted by C. E. W. to Middlebury College in partial fulfillment of the requirements for the degree of Master of Science. (3) A.E.Finn, G. C. Hampson, and L. E. Sutton, J . Chem. Soc., 1254 (1938). (4) (a) E. C. Lingafelter, Coord. Chem. Rev., 1, 151 (1966); (b) R. B. Roof, Jr., Acta Crystallogr., 9, 781 (1956); (0) J. V. Silverton and J. L. Hoard, Inorg. Chem., 2, 243 (1963); (d) V. Amirthalingam, V. M. Padmandhan, and J. Shankar, Acta Crystallogr., 13,201 (1960); (e) E.A. Shugam and L. M. Shkol’nikova, Dokl. Akad. N a u k SSSR, 133,386(1960). (5) R. D. Nelson, Jr., Ph.D. Thesis, Princeton University, Princeton, N. J., 1963. (6) E. Dasgupta and C. P. Smyth, J. Amer. Chem. SOC.,89, 5532 (1967). (7) E.N.DiCarlo and R. E. Stronski, Nature, 216,679 (1967). (8) C. P.Smyth, “Dielectric Behavior and Structure,” McGraw-Hill Book Co., Inc., New York, N. Y., 1956. (9) R. D. Nelson, Jr., Paper 05, Ohio State University Symposium on Molecular Structure and Spectroscopy, June 1965. (10) M. Davies, G. W. F. Pardoe, J. E. Chamberlain, and H. A. Gebbie, Trans. Faraday SOC.,64,847 (1968).
Volume 73, Number 10 October 1969
RALPHD. NELSONJR.,AND CHARLES E. WHITE
3440
Polarizations were calculated according to the Clausius-Mosotti equationj8and eq 1 was used to compute the dipole moment from PO. Table I gives the data for the test liquids. Table I1 gives the results Figure 1. The laser refractometer: S, HCN laser (pulsed, from G and E Bradley, Ltd.); F, 45' mirror (front surfaced); C, thermostated sample cell (with front-surfaced mirror a t the bottom of the sample wedge); G, goniometer mirror (mounted on an Ealing divided circle); D, golay detector; 31, screw-driven, semikinematic slide.
removed and the goniometer mirror turned to reflect light directly back into the laser. Equation 2 is valid if path SFCGD lies in one plane and path SFGD lies on a single line perpendicular to the gravitational force. The exit beam from the cell will strike the goniometer mirror at a nonaxial point unless the distance FG is correct for the refractive index of the sample.
+
FG = t tan 20, h tan
(nli,
sin 20,)
- tan (nairsin ze,)
Liquids Compared to the Square Root of the Static Dielectric Permittivity Compd
nFIR
Cyclohexane Pentane Hexane Octane
1.422 1.359 1.384 1.399
%G 1.422 1.358 1.376 1.408
a All values from A. A. Maryott and E. R. Smith, " Table of Dielectric Constants of Pure Liquids," Xational Bureau of Standards, Circular 514, U. S. Government Printing Office, Washington, D. C., 1951.
Table 11: Density and Refraction Data for Benzene
(3)
where t is the distance from sample surface to mirror C along FC, h is the distance from F to the sample surface = 270" - 2(0, - Bair) is the wedge along BC, and angle between the liquid surface and mirror C. To provide for the necessary adjustment, the cell (with the 45" mirror mounted on it) is fastened to a screw-driven semikinematic slide. An error analysis predicts that goniometer uncertainties of 1' of arc will result in uncertainties in refractive index of 0.004 for a cell wedge angle of 13" and nlia = 1.4. This agrees well with the scatter of results observed for test liquids. The HCN laser produces several lines, the strongest one at 337-pm wavelength." The laser and Golay cell could be replaced by another collimated monochromatic source and the appropriate detector to provide absolute values of refractive index at other wavelengths, providing there is no significant absorption by the sample at those points in the spectrum. Densities good to 0.0003 g ~ m were - ~ determined with 25-m1 capacity pycnometers. All measurements were thermostated to 25.0 i 0.1", except for the test liquids, which were run at 20". Chemicals. AlPtdn3 from J. T. Baker Chemical Co. was dissolved in benzene and shaken with an acidic lOyo KSCN aqueous solution to extract an iron impurity which had made the initial benzene solution amber in color. After rinsing the benzene layer and drying it with CaC12, it was reduced in volume by evaporation. The addition of cyclohexane induced precipitation. The vacuum-dried crystals melted at 191.5-193' (lit.' 193-193.5'). Spectrophotometric grade benzene from Baker was used to make up the solutions. The Journal of Physical Chemistry
Table I : Far-Infrared Refractive Index of Several Nonpolar
Solutions of AIPtdnr a t 25' dn4,
?l$,
Xa
doma
om* mol-'
0 0 03186 0.04287 0,05652
0.8734 0.9021 0.9109 0.9230
89.43 9.5.28 97 34 99.74
XZ
nmRIP
RFIR'I
0 0.02851 0.03628 0.04668 0.05621 0.05652
1.5076 1.5232 1.5280 1.5328 1,6376 1.5376
26.65 28.93 29.59 30.40 31.17 31.17
I
I
for the chelate solutions. The best straight line passing through the data was determined by a least-squares fit. The results gave
+ 89.446 0.02 cm3mol-' R12 = 80.396X2 + 26.651 i 0.01 cm3 mo1-I TI^
=
182.91X2
(4)
(5)
for the molar volume and the far-infrared molar refraction of the solutions as functions of the mole fraction of solute. Combining the results of the far-infrared work RFIR = 107.05 cm3 mol-' with low-frequency resultsI2 PT = 131.7 cm3 mol-' produces a value of 1 = 1.15 D for AIPtdns (1D = 10-ls esu cm). (11) A. G. Maki and D. R. Lide, Jr., J. Chem. Phys., 47, 3206 (1967). (12) R. D. Nelson, Jr., and C. P. Smyth, J . Phys. Chem., 69. 1006 (1965).
ELECTRIC DIPOLEhlO.1IENT
OF
ALUMINUM TRIS(2,4-PENTANEDIONA17E)
344 1
Discussion X-Ray diffraction studies of many Ptdn chelates have shown that although the ligand portion of the ring is planar, the metal atom is bonded so that the oxygenmetal-oxygen plane is about 20" out of coplanarity with the ligand.4 An exception is FePtdn3 which is reported to have completely planar rings.4b Elementary vector analysis shows that for AIPtdns, where the oxygen atoms have octahedral symmetry about the Al, the metal-oxygen bond moments cancel each others' effects, but the carbon-oxygen and other ligand bond moments will not completely cancel out, producing inX stead a resultant molecular dipole moment of pL sin 4. Here 4 is the angle between the ligand and the 0-AI-0 planes, and p L is the ligand (in-plane) moment, 1.940 D, for Ptdn here. Similar calculations for other Ptdn chelates give over-all dipole moments of sin 4 for tetrahedral metal-oxygen symmetry found for BePtdnz, 2 p sin ~ 4 for the cis-square-planar case found for CoPtdnz, and zero for the trans-square-planar case. In both the tetrahedral and octahedral cases, flipping any ring from rp to - 4 will change the direction but not the magnitude of the over-all dipole moment. For the square-planar case, such a flip changes the conformation from cis to trans. For cases in which the metal is surrounded by four ligands as in ThPtdna, the ligand in-plane moments may be resolved into components in the oxygen-metaloxygen planes (which cancel) and components perpendicular to these planes. The latter components point in the tetrahedral directions and cancel to zero only if all four point out. If three point in and one points out (or vice versa), the over-all dipole moment is 2 X p~ sin 4. If two point out and two in, the result is 2.61 X p L sin rp. Reversal of 4 for a ring in these tetraligand cases results in a change in both the magnitude and direction of the over-all dipole moment. Published measurements of the dielectric loss factors of benzene solutions for several Ptdn chelates have indicated nonzero dipole m ~ m e n t s . j - ~The dielectric relaxation data indicate that only very short times are required for changes in the direction of the molecular dipoles. These times are much too short to be consistent with over-all molecular r ~ t a t i o n . ~Far-infrared studies have shown strong gbsorption bands as low as 100 cm-', which have been assigned to the metalligand stretching modes.g Modes attributed to bending have not been observed (down to 30 cm-I), although they should be strong enough to be noticed. Such modes might well be strongly perturbed by asymmetric solvent cages, becoming hindered inversion motions which would produce short dielectric relaxation times without requiring over-all molecular rotation. This sort of motion of a central atom in a potential well with several minima or in a ring-shaped equipotential minimum has been observed in the dielectric relaxation of diphenyl ether12 and in the spectrum of HCN.I3
,-t I N VERSION
di
Figure 2. Free energy us. inversion coordinate compared with free energy for rigid rotation (dotted line). The position of the metal atom with respect to the ligand plane is shown for the initial, activated, and final states, also the bonding in these states. Note the torsion about the MOCC fragment during the inversion.
The temperature dependence of the dielectric relaxation time' for AlPtdn3 indicates an activation free energy at 300°K of 850 cal mol-' for the observed relaxation process compared to an expected value of 3600 cal mol-' based on over-all rotation of the chelate in benzene.* Figure 2 compares the energy requirements for the two mechanisms and illustrates a simple classical model for electron localization and molecular torsion in the activated state, which would result in the inversion of the angle of nonplanarity. Numerous X-ray studies4 have reported large values for the thermal broadening parameter of the metal atom, confirming the considerable freedom of motion indicated by this model. The model proposed here provides results in good agreement with the dipole moments derived from dielectric relaxation work5--? and from polarization differences between the solid state and s o l ~ t i o n s . ~ ~ ~ ' ~ The temperature dependence expected for the total polarization is within the errors quoted for recent solution work which had concluded that the molecules were nonpolar. l6 Early gas-phase work" suff ered uncertainty due to decomposition of the vapors, and the results were based on extrapolations to zero time. These values were always higher than the values observed and it is possible that the polarization values were overcorrected. (13) J. Hougen, Paper P1,Ohio State University Symposium on Molecular Structure and Spectroscopy, Sept 1967. (14) C. C.Meredith, L. Westland, and G. F. Wright, J . Amer. Chem. Soc., 79,2385 (1957). (15) P. Podleschka, L. Westland, and G. F. Wright, Can. J . Chem., 36,574 (1958). (16) E.N.DiCarlo, T. P. Logan, and R. E. Stronski, J . Phys. Chem., 72,1517 (1968). (17) I. E.Coop and L. E. Sutton, J. Chem. Soc., 1269 (1938).
Volume 78, Number 10 October 1960
TAPAN K. MUKHERJEE
3442
The polarity of the metal Ptdn chelates had been qualitatively proven by structural studies and is now quantitatively defined by far-infrared refraction work. The results are consistent with dielectric studies at microwave frequencies and provide a model which clarifies the molecular dynamics of the chelates in electric fields. Extensive discussions of the most recent modern structural theories18 are compatible with this model and now stand more fully confirmed.
Acknowledgment. The authors are indebted to the National Science Foundation for a research grant in support of this work.
(18) K. Nakamoto and P. J. Mecarthy, “Spectroscopy and Structure of Metal Chelate Compounds,” John Wiley & Sons, Inc., New York, N. Y., 1968.
Charge-Transfer Donor Abilities of o,o’-Bridged Biphenyls
by Tapan K. Mukherjee Energetics Branch, Air Force Cambridge Research Laboratories, Bedford, Massachusetts 01730
(Received March IO,1969)
From the charge-transfer bands of their complexes with
T acceptors, the order of the donor strengths of o,o’bridged biphenyls has been determined to be carbazole > fluorene > dibenzothiophene > phenanthrene > dibenzofuran. Dibenzothiophene acts as a s donor. Carbazole, apart from being a possible n donor, undergoes reaction with strong electron acceptors. The behavior of dibenzofuran as a x donor is uncertain. The most important finding is that fluorene is a better donor than phenanthrene. Similarly, 1,2-benzofluoreneis a better donor than chrysene.
Introduction The charge-transfer theory predicts a nonlinear relationship between the energy of the charge-transfer band (hv,t) and the ionization potential (I,) of the donor mo1ecule.l Experimentally, however, a straightline relationship has been repeatedly obtained.2 Besides alternant hydrocarbons, the donors include alkyl ,~ and aryl halide^,^ aza-aromatic c o r n p o ~ n d saIcohols,j etc. The empirical linearity is so common that it has been extensively used to determine the I , values of donors for which direct measurements are not available.6 However, for reliable I, values, strict conditions of (a) identical experimental environments, (b) comparable electronic structures of the components without any sterically hindering factors, (c) similar types of complexes, and (d) clean charge-transfer bands, must be preserved. Although I p is a good measure of the donor strength, in those complexes where substantial changes in charge distribution with very little variation in I , are observed, the positions of the C-T band serve as a better tool for comparison. In view of the renewed interest in the excitation’ and emission8 energy levels of the donors belonging to the o,o’-bridged biphenyl system (I), a study of the complexing properties of the donors of this group seemed to be desirable. The Journal of Physical Chemistry
The report concerning the resonance energy transfer processesg from these donors (I) to 9-phenylanthracene (acceptor), and the observation of photoconduction10in dibenzothiophene (IC)and several of the (1) 5. H. Hastings, J. L. Franklin, J. C. Schiller, and F. A. Matsen, J . Amer. Chem. Soc., 75,2900 (1953). (2) G. Briegleb, “Elektronen-Donator-Komplexe,” Springer-Verlag, Berlin, 1961, pp 74-88. (3) J. Walkley, D. N. Glew and J. H. Hildebrand, J . Chem. Phys., 33, 621 (1960). (4) S. K. Chakravarti, Spectroshim. Acta, 24A, 790 (1968). (6) M. J. Kurylo and N. B. Jurinski, Tetrahedron Lett., 1083 (1967). (6) G. Briegleb and J. Caekalla, 2. Elektrochem., 63, 6 (1959), is a somewhat outdated review. (7) (a) E. Merkel, Ber. Bunsenges. Phys. Chem., 69, 716 (19653; (b) R. Gerdil and E. A. C. Lucken, J . Amer. Chem. SOC.,88, 733 (1966); (c) S. Siege1 and J. S. Judeikis, J . Phys. Chem., 70, 2205 (1966); (d) C. A. Pinkham and S. A. Wait, Jr., J . Mol. Spectrosc., 27, 326 (1968). (8) (a) R. N. Nurmukhametov and B. V. Gopov, Opt. Spectrosc., 18, 126 (1965); (b) K. B. Eisenthal, W. L. Peticolas, and K. E. Rieckhoff, J . Chem. Phys., 44,4492 (1966). (9) D, W, Wllis and B. 9.Solomon, {bid., 46, 3497 (1967).