J. Am. Chem. zyxwvut SOC. zyxwvut 1981, 103, 7515-7520 Synthesis and Characterization of the Pentacarbonylmanganese( 0) Radical, Mn( CO)5, in Low- Temperature Matrices 7515 Stephen P. Church, Martyn Poliakoff,* John A. Timney, and James J. Turner* Contribution from the Department of Chemistry, University zyxw of Nottingham, University Park, Nottingham NG7 2RD, England. Received April zyxwvu 17, 1981. Revised Manuscript Received June 1 I, 1981 Abstract: Mn(CO)5 has been synthesized by UV photolysis of HMn(CO)5 in solid CO matrices at 10-20 K. The HCO radical is also produced. A combination of I3COenrichment and IR spectroscopy shows that Mn(CO), has a square-pyramidal, C , structure with an axial-equatorial bond angle of 96 zyxwvut f 3O. Mn(CO), has a weak visible absorption (A, zyxwvu = 798 nm), close to the absorption reported for Mn(CO)5 generated in ethanol solution by pulse radiolysis. Photolysis with plane-polarized UV light has been used to unravel some of the complexities in the IR spectrum and to show that the UV absorption of Mn(C0)s in the region 300-340 nm has a transition moment of e symmetry. There has long been interest in the radical species MII(CO)~, partly because of its importance as an intermediate in photo- chemical reactions’ and partly because of its structural signifi- MII(CO)~ is expectedZto have a C4, structure with a low-spin d7 electronic distribution, giving an ESR signal near the free-spin value. There has been a considerable effort to obtain such ESR data: Fischer and co-workers3 claimed that sublimation of Mnz(CO)lo onto a cold finger gave signals due to Mn(CO)5 but Symons4 et al. showed that these signals were actually due to 02Mn(CO),. UV photolysis of in solution might seem the obvious way’ to generate Mn(CO)5. Unfortunately early claims to have observed Mn(CO)5 by this route, using ESR de- tection: were later shown to be misinterpreted because of the formation of Mn(I1) species6 ?-irradiation of solid Mn2(CO)lo and MII(CO)~B~ at 77 K gives rise to ESR signals assigned to uptake of an electron withour bond fission,’ i.e., formation of [Mn,(CO),,]- and [M~I(CO)~B~]-. At the moment there is no convincing ESR data for M~I(CO)~ although CIDNP experimentsS suggest the presence of a radical with g = 2 (MII(CO)~?) on hydrogenation of a-methylstyrene by HM~I(CO)~. Hence there is no structural information via this route. Flash photolysis of MnZ(CO),,, in cyclohexane and T H F so- lutions shows the instantaneous disappearance of Mn2(CO)lo and the reformation of this compound by a fast second-order process? believed to involve 2Mr1(C0)~ - Mnz(CO)lo. However, no UV/visible bands38 could be assigned to MII(CO)~. More recently, in elegant pulse radiolysis studies of Mnz(CO)lo and Mn(CO)sBr, Wojcicki, Dorfman, and colleagues1° have generated a transient speceis with an absorption maximum at 830 nm. The reaction kinetics of this species are consistent with it being Mn(CO)5, and, when we allow for solvent differences, the rates agree with the flash photolysis data. Unfortunately, the whole near-UV region of the spectrum (A < 440 nm), was masked by absorption bands ~ (1) Wrighton, M. Chem. Rev. 1974, 74, 401-430. (2) Burdett, J. K. “Molecular Shapes”; Wiley-Interscience: New York; (3) Fischer, E. 0.; Offhaus, 0.; Muller, J.; Nothe, D. Chem. Ber. 1972, (4) Fieldhouse, S. A.; Fulham, B. W.; Neilson, G. W.; Symons, M. C. R. (5) Hallock, S. A.; Wojcicki, A. J. Organomet. Chem. 1973,54, C27-C29. (6) Hudson, A.; Lappert, M. F.; Nicholson, B. K. J. Organomet. Chem. (7) Anderson, 0. P.; Symons, M. C. R. J. Chem. SOC., Chem. Commun. 1972, 102C-21 Anderson, 0. P.; Fieidhouse, S. A,; Forbes, C. E.; Symons, M. C. R. J. Chem. SOC., Dalton Trans. 1976, 1329-1336. (8) Sweany, R. L.; Halpern, 3. J. Am. Chem. SOC. 1977,99, 8335-8337. (9) Hughey, I. V. J. L.; Anderson, C. P.; Meyer, T. J. J. Organomet. Chem. 1977, 125, C49-C52. (IO) Waltz, W. L.; Hackelberg, 0.; Dorfman, L. M.; Wojcicki, A. J. Am. Chem. SOC. 1978, 100, 7259-7264. 1980. 105, 3027-3035. J. Chem. Soc., Dalton Trans. 1974, 567-569. 1975, 92, Cll-C14. of Mnz(CO)lo and the a-ethanol radical. This meant that Wo- jcicki was unable to discover whether the transient species also had absorptions in this region. Although the evidence is strong that Wojcicki et al. did detect M~I(CO)~, there is only sparse spectroscopic information and no structural data. There is clearly scope for a conclusive structural and spectro- scopic characterization of M~I(CO)~. Matrix isolation is partic- ularly well suited to the study of transition-metal carbonyls.” There is indeed some published work which claims to identify MII(CO)~ by IR spectroscopy following cocondensation of Mn atoms and CO/Ar mixtures.I2 However, it has been suggested that this work is inc~rrect’~,’~ (although similar studiesf5with Re/Co mixtures certainly produce Re(CO)5), and moreover this workI2 gave no UV/visible data for Mn(C0)s. This paper presents what we believe to be a definitive spec- troscopic study of Mn(CO),. The radical is generated by UV photolysis of HMII(CO)~ in solid CO; it is characterized by IR spectroscopy by using I3CO isotope enrichment and is shown to have a C,, structure; there is a visible band at 798 nm and a UV absorption, assigned to a transition of e symmetry by photolysis and spectroscopy by using plane-polarized light. Experimental Section The low-temperatureapparatus, Air Products CS-202 Displex, has been previously described16 but has since been permanently transported to the University of Nottingham. IR spectra were recorded on a modified Grubb-Parsonsspectrometeri7 or a Perkin-Elmer Model 580 spectrom- eter fitted with a wire grid polarizer and interfaced to a Digico Micro 16-V computer for multiple scanning.l* High-resolution IR spectra were obtained by using the Nicolet 7199A Fourier Transform interferometer in the laboratory of Professor I. M. Mills at Reading University and our own portable matrix isolation apparatus, essentially similar to our fixed (1 1) Burdett, J. K.; Turner, J. J. “Cryochemistry”; Moskovits, M., Ozin, G. A,, Eds.; Wiley: Toronto, 1976; Chapter 11, pp 493-523. Burdett, J. K. Coord. Chem. Rev. 1978, 27, 1-58. (12) Hiiber, H.; Kundig, E. P.; Ozin, G. A.; Poe, A. J. J. Am. Chem. SOC. 1975, 97, 308-314. (13) Timnev. J. A. Inora. Chem. 1979, 18, 2502-2506; Ph.D. Thesis, University of Newcastle UGn Tyne, 1979. (14) The frequencies of the YCI) vibrations of a series of transition-metal carbonyls usually decrease smoothly as the number of d electrons decreases. Thus, for the equivalent vibrations of M(CO)4 species the vc-0 frequencies are in the order Ni > Co > Fe > Cr. For M(CO)5, however, the vc-0 bands of “Mn(CO)5”12 are substantially lower in frequency than those of Cr(C0)5. Similar1 the uc-0 frequencies of Cr(CO), and W(C0)s are separated by only 9 cm-’ Y ’ while those of ‘‘Mn(C0)5” and Re(CO)5 are nearly 60 cm-’ apart-ee Table I. 5585-5586. (15) Huber, H.; Kundig, E. P.; Ozin, G. A. J. Am. Chem. SOC. 1974, 96, ____ (16) See, for example: Perutz, R. N. Ph.D. Thesis, Cambridge University, (17) Poliakoff, M.; Turner, J. J. J. Chem. Soc., Dalton Trans. 1974, (18) Smith, K. P. Ph.D. Thesis, University of Newcastle upon Tyne, 1981. 1974. 2276-2285. 0002-7863/81/l503-7515$01.25/0 0 1981 American Chemical Society