Matrix isolation experiments

An extensive series of low-temperature matrix isolation experiments has failed to provide any evidence of oxirene formation, either by diazo ketone photolysis (82CB2192) or by attempted photo-retro-cycloaddition (82CB2202). 

The boron trihalides are volatile, highly reactive, monomeric molecular compounds which show no detectable tendency to dimerize (except perhaps in Kr matrix-isolation experiments at 20K). In... 

By ab initio MO and density functional theoretical (DPT) calculations it has been shown that the branched isomers of the sulfanes are local minima on the particular potential energy hypersurface. In the case of disulfane the thiosulfoxide isomer H2S=S of Cg symmetry is by 138 kj mol less stable than the chain-like molecule of C2 symmetry at the QCISD(T)/6-31+G // MP2/6-31G level of theory at 0 K [49]. At the MP2/6-311G //MP2/6-3110 level the energy difference is 143 kJ mol" and the activation energy for the isomerization is 210 kJ mol at 0 K [50]. Somewhat smaller values (117/195 kJ mor ) have been calculated with the more elaborate CCSD(T)/ ANO-L method [50]. The high barrier of ca. 80 kJ mol" for the isomerization of the pyramidal H2S=S back to the screw-like disulfane structure means that the thiosulfoxide, once it has been formed, will not decompose in an unimolecular reaction at low temperature, e.g., in a matrix-isolation experiment. The transition state structure is characterized by a hydrogen atom bridging the two sulfur atoms. 

Thus, overall, it is clear that flash photolysis with uv-visible detection is effective in establishing the broad outlines of the photochemistry of a particular metal carbonyl. Intermediates can be identified from their reaction kinetics, and sometimes, with the help of uv-vis data from matrix isolation experiments. Structural information from uv-vis flash photolysis is at best sketchy. Many questions remain unanswered. Time-resolved IR measurements can fill in some of these answers. 

Relatively little work has been done on the flash photolysis of gas phase metal carbonyls, partly because of the low volatility of many of the compounds. Early work by Callear (41,42) provided some evidence for Ni(CO)3 generated from Ni(CO)4 in the gas phase (41) and Fe atoms produced from Fe(CO)5 (42). This latter process has even been used as the basis of an Fe atom laser (43). More recently Breckenridge and Sinai (44) studied the flash photolysis of Cr(CO)6. Their results, interpreted largely on the basis of data from matrix isolation experiments, were in broad agreement with Kelly and Bonneau s solution work (JJ). In particular, they found no evidence for loss of more than one CO group [Eqs. (4) and (5)]. 

The bands due to Fe(CO)4 are shown in Fig. 8. This spectrum (68) was particularly important because it showed that in the gas phase Fe(CO)4 had at least two vq—o vibrations. Although metal carbonyls have broad vC—o absorptions in the gas phase, much more overlapped than in solution or in a matrix, the presence of the two Vc—o bands of Fe(CO)4 was clear. These two bands show that in the gas phase Fe(CO)4 has a distorted non-tetrahedral structure. The frequencies of these bands were close to those of Fe(CO)4 isolated in a Ne matrix at 4 K (86). Previous matrix, isolation experiments (15) (see Section I,A) has shown that Fe(CO)4 in the matrix had a distorted C2v structure (Scheme 1) and a paramagnetic ground state. This conclusion has since been supported by both approximate (17,18) and ab initio (19) molecular orbital calculations for Fe(CO)4 with a 3B2 ground state. The observation of a distorted structure for Fe(CO)4 in the gas phase proved that the distortion of matrix-isolated Fe(CO)4 was not an artifact introduced by the solid state. 

The complex photochemistry of CH MCCO) (M = Mn, Re) has already been unravelled by matrix isolation experiments, utilising both IR (V(C-O) and V(N N) vibrations) and UV/visible spectroscopy (12). 

In this section, we outline a number of studies by various workers in which matrix isolation experiments have contributed to the understanding of some photochemical pathways and where the objective was to relate the experiments to room-temperature photochemistry. We then consider briefly some current problems involving dinuclear and polynuclear carbonyls. 

Fig. 14). This is related to the more favorable thermodynamics of conversion in the case of 17d compared with 17a, due to the slight stabilization of ketenimine 18d by the fluorine adjacent to nitrogen. The barrier for 17d->18d is still predicted to be ca. 2.5 kcal/mol higher than the barrier for BZ -> K at the same level of theory, which helps to explain why Morawietz and Sander successfully detected benzazirine 17d in their matrix isolation experiments. 

In spite of all this research, systematic criteria for trapping of organic radical cations produced by positive hole transfer in solid argon are not yet fully understood. It would be useful to establish such criteria to evaluate the prospects of characterization of chemically important aliphatic radical cations under the conditions of classic matrix isolation experiments. 

The use of computed spectra to bolster structural assignments has seen heavy use in matrix isolation experiments. This is a slightly atypical example, insofar as the species involved actually require some careful attention to non-dynamical correlation, but represents an excellent example of how theory can aid experiment in the identification of short-lived reactive species. 

Another example of ring-chain tautomerism with participation of a substituent is the interconversion of furoxan-2- and furoxan-5-oxides (104a 104b). The equilibration is presumed to proceed via an intermediate cis- 1,2-dinitrosoalkene (105). In the benzofuroxan series the involvement of the corresponding 1,2-dinitrosoarenes as intermediates has been firmly established by matrix isolation experiments (91CC1178, 91JOC5216,92CL57). 

If this interconversion were being measured in a matrix isolation experiment, one might simply use band intensities to measure the equilibrium constant. With an oriented crystal, however, correction must be made for the... 

As stated above, an alkoxycarbonyl(disilanyl)carbene rearranges to an acylsilene, but not to a silylketene (Scheme 5). For diazoacetate 62, the result of a matrix-isolation experiment (Table 3) can also be reproduced when the nitrogen extrusion is induced by vacuum pyrolysis at 360 °C, by thermolysis of the neat compound at 180 °C or by photolysis in... 

For earlier results from matrix isolation experiments, see H. Schnockel, T. Mehner, H. S. Plitt and S. Schnuck, J. Am. Chem. Soc., Ill, 4578 (1989) and references cited therein. 

The rearrangement of l-silacyclopent-3-en-l-ylidene to unsubstituted silole, first suggested in 1981s, has been confirmed by the characterization of the silole dimer196 and by the recording of UV-visible and IR spectra of the silylene, silole and 1-silacyclopenta-1,3-diene in matrix isolation experiments in which the isomeric silacyclopentadienes were... 

The reaction of CH2 with C02 was first postulated by Kistiakowsky and Sauer (13) as taking place via an a-lactone intermediate. The occurrence of this reaction was subsequently demonstrated by Milligan and Jacox (14) in low temperature matrices. These low temperature matrix isolation experiments, however, could not determine definitely the structure of the CH2 C02 intermediate. The result of our laser absorption experiment shows that the CO is vibrationally excited up to v — 4 with a distribution close to the one predicted by a statistical model assuming the existence of a long-lived CH2 C02 complex. This calculation, however, is insensitive to the structure of the complex assumed. Since the ground state triplet CH2 is known to be less reactive and kinetically behaves like CHj (15,16), which does not react readily with C02, the singlet A CH2 is assumed to be involved in the reaction. 

The HBO molecule is a further example of an unstable intermediate which has only recently been detected in matrix-isolation experiments.413 SCF calculations by Thomson and Wishart414 have yielded reliable values for the bond lengths, and the authors have discussed the bonding and one-electron properties. It is hoped that such calculations will stimulate further experimental work on this and similar unstable triatomic molecules. 

G. CCO and TiCO.—These two species have little in common, but it is convenient to examine them together. OCC423 has been the subject of a near-HF study.lt is an unstable species, as yet characterized only in matrix-isolation experiments.394 The computed bond lengths were R(0—Q = 2.121 bohr, R(C—C) = 2.58 bohr, and several one-electron properties were predicted. The computed force constants were in fair agreement with experiment. 

---