摘要:
1978 93Reactions of Co-ordinated Molecules. Part 9.t Calculation of Carbonyl-stretching Force Constants of Metallo-acetylacetonate Molecules : Evi-dence supporting Intramolecular BondingBy Charles M. Lukehart and G. Paul1 Torrence, Department of Chemistry, Vanderbilt University, Nashville,The carbonyl-stretching force constants of three complexes containing the metallo-acetylacetonate moiety, cis-[AI{(OCMe),M(CO)a),] and cis-[Re(MeCO . . . H . . . OCMe)(CO),] ( M = M n or Re), have been calculatedusing the Cotton-Kraihanzel approximation. The results permit the assignment of the normal modes of the carbonylligands of the metallo-acetylacetonate group to the experimentally observed stretching frequencies. The values ofthe force constants, k,, are related to the magnitude of the intramolecular bonding between the two axial-carbonylTennessee 37235, U.S.A.ligands and the metallo-acetylacetonate ligand.IN earlier papers of this series we reported the prepar-ation and characterization of a number of complexes ofthe type shown below which contain the metallo-p-diketonate group, cis-(OC),M(RCO) (R'CO), where M =Mn or Re.1-3 The metallo-p-diketonate ligand isisoelectronic and isostructural with a p-diketonateI nligand, and it represents the formal substitution of themethine group of a p-diketonate anion by the metalcomplex, ck-[M(CO),].When M' = A1 or Ga, n = 3,and a tris-chelate complex is formed. If M' = H+, thenn = 1 and the neutral metallo-p-diketone molecule isisolated as the enol tautomer.The i.r.spectra of these complexes in the terminalcarbonyl region have four bands indicating that thelocal-oscillator approximation holds. Each metallo-p-diketonate ligand acts as an isolated cis-[M(CO),L,]complex. With this assumption, the four observedfrequencies were assigned tentatively to the four i.r.-active normal modes using qualitative reasoning basedon the relative intensities of the bands.3 However, theassignment of two of the four bands was particularlyuncertain because of their closeness in frequency and anangular distortion of two of the carbonyl ligands.The present paper utilizes the Cotton-Kraihanzelapproximation to calculate the stretching force constantsof the carbonyl ligands of three metallo-acetylacetonatecomplexes (R, R' = Me).For each complex, only oneof the six possible assignments is correct and the sameassignment holds for all the three complexes. LaserRaman data support this assignment. A comparisonof the values of the force constants indicates a possiblecorrelation of the value of the axial-carbonyl forceconstant with the degree of bonding between the n-electron system of the metallo-acetylacetonate ligand andthese two carbonyl ligands.,t Part 8; C. M. Lukehart and G. P. Torrence, Inorg. Chem.,1978, 17, in the press.1 C. M. Lukehart, G. P. Torrence, and J. V. Zeile, J . Amer.Chem. SOL, 1975, 97, 6903.Laser Raman data are presented to support the band assignment.EXPERIMENTALThe complexes cis-[Al{ (OCMe),Mn(CO),},] (l), cis-[Al-{OCMe),Re(CO),},j(2), and cis-[Re(MeCO - H 0CMe)-(CO),] (3) were prepared by literature Thei.r. spectra were recorded on a Perkin-Elmer 727 spectro-meter of ca.0.01 mol dmP3 solutions in cyclohexane in 0.1-mm sodium chloride cavity cells using the pure solvent as areference. Band wavenumbers are reported in cm-l andwere calibrated with a polystyrene film. The wavenumbersof complex (1) obtained in this way were within 2 cm-l ofthe values obtained via calibration with gaseous CO andDCI.Laser Raman spectra were recorded on a Beckman model700 laser Raman spectrometer using a 2-W argon laserhaving a green line a t 5 145 A. The Raman spectra andpolarization measurements were recorded in benzene-methylene chloride (9 : 1) and benzene solutions for com-plexes (2) and (3), respectively.Solutions of complex (1)decomposed within 30 s when exposed to the laser beam.The polarization ratio, p, was calculated by normal pro-cedures.5RESULTS AND DISCUSSIONThe three complexes studied are shown below:M\" ,c -I - - - - %-= /Me0 \ I%:s 0( 1 ) M=Mn( 2 ) M= ReThe carbonyl regior. of the i.r. spectrum of (1) has beendisplayed previously and consists of four bands,A-D, where A is the band of highest frequency. Thespectra of the other two complexes have nearly identicalpatterns with some slight differences in band frequencies.Each M(CO), group of (1) and (2) is independent of theother two identical groups since coupling between thecarbonyl oscillators of different M(CO), groups is not2 C.M. Lukehart and J. V. Zeile, J . Amer. Chem. Soc., 1976,98, 2365.C. M. Lukehart, G. P. Torrence, and J. V. Zeile, Inorg.Chem., 1976, 15, 2393.C. M. Lukehart and G. P. Torrence, Inorg. Chim. Acta, 1977,22, 131..S. K. Freeman, ' Applications of Laser Raman Spectroscopy,'Wiley, New York, 1974, p. 29J.C.S. Daltonobserved. The local-oscillator approximation appliesand each M(CO), moiety can be treated as a cis-disub-stituted octahedral complex of the type, cis-[M(CO),L,],possessing Czt, symmetry. The observed frequencies foreach complex are shown in Table 1.TABLE 1Observed carbonyl-s tretching bands of the complexes(1)-(3) in cyclohexane solutionBand (cm-I)Complex A R c D2 065m 1 985s(sh) 1 978s 1 960m2 048m 1989s(sh) 1982s 1959m2 099m 2 000s(sh) 1 993s 1 962m(1)(2)(3)The cis[M(CO),L,] spectrum is characterized by fouri.r. active carbonyl-stretching modes.Following thequalitative work of Orge16 and the quantitative studyof Cotton,' the four bands can be assigned tentatively.The two axial-carbonyl ligands, which lie out of the ML,al) .8 Secondly, the X-ray structural determinations of(1) and (3) reveal that the two axial-carbonyl ligands aretilted toward the plane of the metallo-acetylacetonateligand thereby defining a co-ordination axis whichdeviates significantly from 180".1*2In complex (1) , the C (axial) -Mn-C (axial) angle of156.3" is 9.1" more bent than the other two axes of theco-ordination octahedron, whilst in complex (3) theC(axia1)-Re-C(axia1) angle of 170.6" is 7.7" more bentthan the other two co-ordination axes.This pronouncedtilting should cause the A,(axial) band to gain intensityat the expense of the B, (axial) band, which contributesmore uncertainty to the correct assignment of bands Band C. We undertook a quantitative study of thesespectra to resolve these questions.The secular equations as derived from the Cotton-Kraihanzel approximation for the cis-[M(CO),L,] complexwere taken from the literat~re.~ The three carbonyl-stretching force constants which must be determined areTABLE 2The calculated force constants and B-band frequency for each assignnient of the complexes (1)-(3)Observed bandsand assignments Force constants /N ni?Calculated wavenumberComplex ,41(2) A,(') B, B, k l k , k17 of band B (cm-l)7 r A rc B D *€3 D c 1611 1613 31 1987 2 A3 A D B c 1621 1531 41 18954 A c D B 1 589 1616 32 19645 A D c B 1 546 1639 30 19386 A €3 c L) 1580 1637 28 1978C I3 D *B D c 1612 1601 25 19863 A D B c 1619 1527 33 19024 A c D B 1600 1602 26 19745 A D C E3 1 543 1633 23 19406 A B C D 1572 1631 22 19752 A R D C 1 644 1634 40 2 0013 A D B c 1 656 1521 52 18734 A c D B 1622 1637 41 19795 A D c B 1 545 1679 37 I 9336 A B c D 1 590 1675 36 1986* Imaginary roots obtained.* * (1) 1 A* * (2) ; 2(3) 1 A c B D * * *plane, give two bands (A,(,) and B, symmetry) where theAJ2) band is of higher frequency and much weakerintensity than the B, band.The two equatorial-carbonyl ligands, which lie in the ML, plane, also givetwo bands ( A i l ) and B, symmetry) where both peakshave nearly equal intensity.Unfortunately, the relativepositions of these two sets of bands cannot be predictedfrom symmetry considerations alone. The tentativeassignment of the bands observed for complex (1) is:peak A (A,(,), axial); peak B ( A i l ) , equatorial); peakC (B,, axial); and peak D (B,, equatorial).This assignment requires a quantitative confirmationfor two reasons. First, the spectra of most, if not all,of the previously reported cis-[M(CO),L,] complexesshow the C band as the weaker shoulder of the moreintense B band giving an assignment: A (A,(2), axial) ;B (B,, axial) ; C (A,(,), equatorial) ; and D (B,, equatori-L. E.Orgel, Inorg. Chem., 1962, 1, 25.F. A. Cotton, Inorg. Chem., 1964, 3, 702.k, (equatorial CO), k , (axial CO), and ki (the inter-action force constant).The correct assignment is obtained by the followingproced~re.~ The frequencies of A, C, and D are used tocalculate k,, k,, and ki. Band B is not used because it isa shoulder and the observed frequency is probably themost inaccurate (too high in this case). Since thehighest-energy mode must be the A,<,) band, there aresix possible combinations of assignments for the otherthree bands.9 The force constants are calculated foreach assignment. The correct assignment is the onehaving real positive values for the force constants wherek, is significantly greater than k , and where the calculatedfrequency for band B is closest to the experimental value.Table 2 shows the calculated force constants and bandC.S. Kraihanzel and F. A. Cotton, Inorg. Chem., 1963, 2,F. A. Cotton and C . S. Kraihanzel, J . Amer. Chem. SOC.,53.3.1962, 04, 44321978 95B frequency for all six possible assignments of eachcomplex.Assignment (1) of each complex gives imaginary rootsand can be neglected, (2) and (3) have k, either lessthan or nearly equal to k, and can be eliminated, and(4) and (5) have either k , being very nearly equal to k,or a very low calculated value for band B. Assignment(6) of each complex meets all the above criteria, and istherefore the correct assignment. The calculated valueof the wavenumber of band B is lower than the observedvalue for each complex [by 7 cm-I for complex (1) and by14 cm-l for complexes (2) and (3)].This trend isexpected and it should be stressed that all the com-putational and theoretical error is absorbed by thisfreq~ency.~ The quantitative assignment is in agree-ment with the previous qualitative assignment .3The laser Raman data for complexes (2) and (3) sup-port this assignment. Complex (3) gave the betterRaman spectrum due to its excellent solubility in ben-zene. In the Raman spectrum of (3), peak A was themost intense peak in the carbonyl region, and peak Bwas more intense than C or D although B and C were notcompletely resolved. This is strong empirical evidencethat peaks A and B are the totally symmetric vibr-ations.1° Also peak A was a polarized band, p = 0.41,thereby confirming it as an A, vibrational mode.Even though complex (2) was not very soluble, peak Awas visible in the Raman spectrum, and this peak waspolarized also, p = 0.57, thus providing strong evidencethat the peak of highest frequency is a totally symmetricvibration. Solutions of this complex showed slightdecomposition within 5 min of exposure to the laserlo J.Tang and A. C. Albrecht, ‘ Raman Spectroscopy,’ vol. 2,ed. H. A. Szymanski, Plenum, New York, 1970, p. 58.beam. The rapid decomposition of complex (1) in thelaser beam prevented its analysis by laser Ramanspectroscopy,Since complexes (1)-(3) are quite similar one wouldnot expect a large spread in the values of the individualforce constants.The values of k, and ki fall within theranges 1580 &- 10 and 29 7 Nm-l respectively, theuncertainty being that usually considered as normal inthe Cotton-Kraihanzel method. The substitution of aproton for an aluminium ion or a rhenium atom formanganese does not sighificantly affect the value ofk,. However, the value of k, for the rnetallo-enolcomplex, (31, is 41 Nm-l larger than the average value ofk, (1 634 & 3 Nm-l) for the two aluminium complexes,(1) and (2). This is a significant difference. If thevalue of a force constant reflects bond order, then itappears that the C-0 bonds of the axial-carbonyiligands of complex (3) are of higher bond order than theC-0 bonds of those ligands of complexes (1) and (2).This result may support our explanation of the tiltingof these axial-carbonyl ligands in complexes (1) and (3),4According to this bonding mechanism, the C-0 bondorder of the axial-carbonyl ligands of complex (1) shouldbe reduced relative to complex (3). This is the trendobserved in the values of k, for these complexes. Pre-sumably, complex (2) would show a tilting of the axial-carbonyl ligand of the same degree as found in complex(1).VC’e thank the Research Corporation and the n’ationalScience Foundation for support, and Dr. J. Springer of theDepartment of Physics of Fisk University for recording thelaser Raman data and for helpful discussions.[7/095 Received, 19th Janunvy, 1977
ISSN:1477-9226
DOI:10.1039/DT9780000093
出版商:RSC
年代:1978
数据来源: RSC