Biacetyl, photolysis

An example of the difficulties we are citing is found with biacetyl photolysis at 4358 A at room temperature. The following figures have been obtained by various authors... 

Experimental results of the iodine-inhibited reaction show that biacetyl photolysis takes place almost exclusively through free radicals . Since, among the radicals occurring in the system, it is the acetyl radical that is present in highest concentration at low temperatures , it seems reasonable to suggest that the primary process is... 

The products of biacetyl photolysis are CO, CH4, CjHg, CH3COCH3, CH2CO and CH3C0C0C2H5 . Acetaldehyde could also be detected in small a-mounts at room temperature . The formation of these products, with the exception of acetaldehyde, can be interpreted by the mechanism suggested by Blacet and Bell which is supported, to some extent, also by the flash photolytic ex-... 

WAVELENGTH DEPENDENCE OF THE PRIMARY QUANTUM YIELD OF BIACETYL PHOTOLYSIS, CALCULATED ON THE BASIS OF RELATIONS (33) AND (34)... 

Biacetyl. Photolysis of biacetyl was first reported in 1923 by Porter, Ramsperger, and Steel 121> who initiated the vapor at 100° and identified the products as ethane and carbon monoxide. Subsequently methane, acetone, ketene, and 2,3-pentanedione were also recognized as products. Considerable effort has been expended to achieve a full understanding of the details of the primary processes and progress was reviewed by Noyes, Porter and Jolley 113> in 1956. Irradiations have been performed over a wide range of pressure, temperature and wavelength (including the far ultraviolet 68> and mercury photosensitization 69>). Overall and individual quantum yields varied widely. 

The present discussion is by no means exhaustive. It is designed to provide a summary of the most significant and reliable kinetic data, at least those that appear so to the author. There is a variety of methods for producing alkyl radicals, and, naturally, there will be certain restrictions on experimental conditions depending on the method chosen. Some of the common methods for generation of methyl radicals, for example, include photolysis of acetone, pyrolysis of di-ferf-butyl peroxide, photolysis of biacetyl, photolysis of azomethane and decarbonylation of acetaldehyde. In the majority of cases discussed here, the reactions were followed by product determinations, employing gas chromatography. 

The photolysis of tertiary alkyl azides has been utilized for the synthesis of mono- and di-N-derivatives of biacetyl (2.26) 243a b)... 

Photolysis of dicyclopentadienyltin results in formation of the Cp- radical (again detected by ESR), along with the precipitation of some unidentified yellow solid54. In contrast, photolysis of dicyclopentadienyllead produces no Cp-, unless di-f-butyl peroxide or biacetyl are added to the reaction mixture. The trimethylstannylcyclopentadienyl radical was produced by photolysis of bis(trimethylstannyl)cyclopentadiene (reaction 35), and was detected using ESR spectroscopy57. 

The phosphorescence of a 5 X lO" M solution of biacetyl in de-aerated 2-propanol at room temperature could be quenched completely by 1 a,d,e (10 8 M) 84). In all three cases, the corresponding photoreduction products 2a,d,e emerge from analogous preparative scale biacetyl sensitized runs. Since 2e is also formed, steric hindrance to hydrogen abstraction from solvent cannot be too effective when a (probably longer-lived) triplet is populated, whereas it might be effective in the direct photolysis ot 1 e 88) where isomerisation competes with reduction probably in the (short-lived) singlet state. 

Faust, B. C., K. Powell, C. J. Rao, and C. Anastasio, Aqueous-Phase Photolysis of Biacetyl (an a-Dicarbonyl Compound) A Sink for Biacetyl and a Source of Acetic Acid, Peroxyacetic Acid, Hydrogen Peroxide, and the Highly Oxidizing Acetylperoxyl Radical in Aqueous Aerosols, Fogs, and Clouds, Atmos. Environ., 31, 497-510 (1997). 

In studies of the vapor phase photolysis of biacetyl it was observed that a new product, which quenched both the phosphorescence and primary dissociation of biacetyl, was formed. A strong absorption at 275 nm, which was associated with the quenching activity, was also observed. Since no new product which displayed significant quenching activity could be isolated, it was concluded that the quenching was due to the enol of biacetyl (8).57 This assignment was supported by the disappearance of the absorption at 275 nm when IC1 was added to an irradiated aqueous solution of biacetyl.58... 

Photoreduction was quenched by high concentrations of biacetyl, slightly retarded by iodonaphthalene, but not affected by azulene or anthracene.113 These observations led to the unsatisfying conclusion that reduction proceeded via a triplet state which could be only selectively quenched. However, later work114 using flash photolysis showed that the benzophenone ketyl radical was generated upon irradiation of solutions of benzophenone and acridine, and that its predominant mode of disappearance was by reaction with... 

In their original paper, Sieger and Calvert proposed the second of these routes as the predominant primary step. This is unlikely to be the case because no trace of products which could arise by the reactions of the trifluoroacetyl radical have been observed in the photolysis of either trifluoro- or hexafluoroacetone and it is assumed that this radical is unstable. In support of this is the evidence that the yield of ethane rises steadily with increase in temperature whereas that of hexafluoroethane remains approximately constant. Finally, in a recent re-investigation, Dawidowicz and Patrick49 have identified biacetyl in the products of the photolysis of trifluoroacetone. The most probable primary step is, therefore, the former, which provided the acetyl radical and a trifluoro-methyl radical. 

As a further possibility the ac electrolysis may lead to other products than those of the photolysis. In this case an excited state mechanism is, of course, excluded. Although there is a certain similarity between the electronic structure of an excited state and the reduced or oxidized form of a molecule, they are not identical. Consequently, it is not surprising when photolysis and electrolysis do not yield the same product. Another reason for such an observation may be the different lifetimes. An excited state can be extremely short-lived. Non-reactive deactivation could then compete successfully with a photoreaction. The compound is not light-sensitive. On the contrary, the reduced and oxidized intermediates generated by ac electrolysis should have comparably long life times which may permit a reaction. The ac electrolysis of Ni(II)(BABA)(MNT) (BABA = biacetyl-bis(anil) and MNT - = disulfidomaleonitrile) is an example of this reaction type (63). 

It is clear that if we are to use photolysis to study the reaction of radicals with oxygen it would be preferable to use conditions under which decomposition of the parent molecule occurs primarily from the singlet excited state. However, there is good evidence that methyl isopropyl ketone129 and 2-pentanone141 can be deactivated from the singlet excited state by sufficient pressures of biacetyl. 

It was hoped that the photooxidation of biacetyl would enable one to study the experimentally difficult reaction between oxygen and acetyl radicals without the complication of other radicals such as occur in acetone photooxidation. The primary photolysis of biacetyl gives mainly acetyl radicals18 14... 

The first quantitative photochemical study of a Rh111 amine was reported by Moggi,8 who noted that both 254 nm (LMCT) and 365 nm (ligand field) excitation of [Rh(NH3)5Cl]2+ caused chloride labilization (equation 131). Other early reports include Basolo s study of the photoinduced stereo-retentive halide aquation from [M(en)2X2]+ (M = Rh, Ir X = Cl, Br, I), and Broomhead s observation of chloride aquation from [RhCl2(phen)2]+.726 While halide labilization dominates upon photolysis of [Rh(NH3)5Cl]2+, both bromo and ammine loss occur upon photolysis of the bromo analog (equation 132)685,707 and ammine is labilized from the iodo analog (equation 133).70 Biacetyl sensitization of the bromo complex quenches the biacetyl phosphorescence, but not the fluorescence,707 consistent with a photoreactive triplet state. 

The formation of the triplet state can be directly followed in time through the observation of the transient triplet-triplet (T-T) absorption in flash or modulated photolysis or by the observation of the phosphorescence emission. A typical radiative lifetime of phosphorescence for simple carbonyls is 10 3 s. Therefore, it is extremely difficult to observe the "unrelaxed" phosphorescence emission without collisional relaxation, unless the triplet lifetime is significantly shortened by a competing radiationless process. Under these conditions, the correspondingly low quantum yield of phosphorescence makes such measurement rather difficult. Usually, "relaxed" phosphorescence from a molecule such as biacetyl is observed. Therefore, only the transient T-T absorption can provide useful data in the gas phase, although a determination of the absolute yield is rather involved and difficult. 

Bennett and coworkers (138) have prepared the acetyl radical CH CO by the reaction between Na and acetyl chloride and have obtained an isotropic value of 5.1 gauss for the proton splitting. An ESR spectrum attributed to CH CO with 16 gauss was earlier reported during photolysis of biacetyl (276) in a solid matrix. Recently a series of benzoyl 6-radicals (277) have been observed in the liquid photolysis of the corresponding benzaldehyde in a cyclopropane solution containing di-t-butyl peroxide. For the first time, it is possible to study in these radicals the delocalization of the unpaired electron from the acyl 6-system into the adjacent phenyl ir-system. The significant conclusion from this study is the... 

In the presence of an alcohol or ether the excited acetone abstracts an H atom from the carbon alpha to the 0 atom of the alcohol or ether. Similar photoreduction of perinaph-thenone (365) has been observed. However, the nature of the excited state of acetone for liquid photoreduction in these studies has not been established. Free radicals produced from photoreduction of acetaldehyde, biacetyl and acetoin in the presence of good H-atom donors have been observed by Zeldes and Livingston (216). These authors also studied the photolysis of oxalic acid and esters (366). 

Further work on the mechanism of dione reactions with oxygen includes a detailed study 165> of reactions of biacetyl and benzil with emphasis on the latter. The earlier conclusion2 that photolysis of benzil does not produce benzoyl radicals... 

Oxalyl chloride yields COCl radicals on photolysis and this reaction has been used to directly substitute alkanes with the chloroacyl group (equation 30). Similarly, biacetyl reacts with alkanes in a benzoyl peroxide-initiated chain reaction to give ketones in ca. 60-70% yield (equation Cyanogen... 

There have been attempts to explain this low activation energy at low temperatures in terms of diffusion of radicals to the wall and heterogeneous reaction of CHg -4- acetone, the subseejuent photolysis of accumulated biacetyl (which Noyes has calculated to be insufficient), the reaction 6, and possible hot radical effects. None of these has yielded to quantitative analysis, although A. J. Nicholson, J. Avi. Chem. Soc.y 73, 3981 (1951), has shown that diffusion of CH3 radicals to the walls may become important at low intensities, low acetone pressures, or low temperatures. The fa(5t that I2 does not completely quench formation of CH4 is an indication that hot radical effects may be important. 

---