BENZYLCHLOROCARBENE: UV ABSORPTION SPECTRUM, KINETICS FOR 1,2-H MIGRATION AND MECHANISM FOR REACTION WITH ACETIC ACID AS DETERMINED BY COMBINED CONTINUOUS AND LASER-FLASH PHOTOLYSIS

The UV absorption spectrum of benzylchlorocarbene, generated by laser flash photolysis of 3-chloro-3-benzyldiazirine, has been observed in the 290-330 nm range. The lifetime of this species, 18 ns at 25C, is determined by the rate ofthe 1,2-H migration to produce chlorostyrenes. Quenching rate constants of this carbene by acetic acid and tetramethylethylene have been measured. Comparison of this kinetic data with the quantitative analysis of the products obtained under continuous irradiation gives further insight into the mechanisms of carbene-acetic acid reactions.


INTRODUCTION
The most common rearrangement reaction of alkyl substituted carbenes is the migration of a hydrogen atom to the carbene centre which affords an alkene. RCHCX -RCHCHX H When X is a chlorine atom and in the presence of reactants such as olefins, alcohols or acids, this reaction is in competition with reactions of the carbene with either the ethylenic double bond to give cyclopropanes or the OnH bond to give ethers or esters. The relative amount of the products obtained under continuous irradiation of precursors for this type of carbene (such as 3-chloro-3-benzyldiazirine, 1, which gives the benzylchlorocarbene, 2) has been measured as a function of the concentration of several olefins. 2 From these studies, it was concluded that the mechanism of the addition reaction of carbenes to olefins involves an intermediate complex from which both rearrangement and cyclopropanation reactions are possible. The formation of this complex could be reversible as suggested by Turro and Moss 3 to account for the negative activation energies observed for the addition of Ph--C--CI to trimethyl-and tetramethyl-ethylene However, theoretical calculations by Houk and Rondan 4 provided no evidence for a carbene-alkene complex. We have recently reported a laser flash photolysis study 5 of the system "Benzylchlorocarbene + tetramethylethylene" and confirmed the existence of this complex but we could not give definitive conclusion concerning the question of the reversibility of its formation.
In the present paper, we wish to examine the photolysis of I in the presence of acetic acid under both continuous irradiation and laser-flash-photolysis conditions in order to get information about the mechanism for the reaction of this acid with carbene 2 and, more precisely, about the existence (or the non-existence) of an intermediate such as the ion-pair previously suggested. 6 2. EXPERIMENTAL SECTION 3-chloro-3-benzyldiazirine, 1, was prepared by the Graham's method 7 and purified by chromatography on 60-200 mesh silica gel. Pyridine was distilled over Call2 which has been shown to be superior to other desiccants, including sodium. 8 Absence of chlorostyrene prior to photolysis was confirmed by NMR spectroscopy. Continuous irradiation at 25C was carried out with 3500-UV lamps in a Rayonet photoreactor until all the diazirine (0.03 M in isooctane with various amounts of acetic acid) was destroyed. The reaction mixture was worked-up by diluting with water and extracting repeatedly with carbon tetrachloride. The combined extracts were washed with NaHCO3 solution, water and then dried. The products were analyzed on a Varian Vista 6000 gas chromatography fitted with a 6 ft x 0.125 in. stainless steel column packed with CSP-20M. The G.C. traces were calibrated by using authentic samples of the reaction products. Peak area were integrated with a HP-3390A recorder. The laser-flash-photolysis set-up uses a crossed-beams arrangement. The sample in a 10 10 mm cell was excited at 355 nm by single light pulses (200 ps; 5-30 mJ) provided by a frequency tripled mode-locked Nd-YAG laser (Quantel). The detection system (pulsed Xe-arc, monochromator, photomultiplier and a Tektronix 7912 transient recorder) has a response time around 5 ns.

Continuous Irradiation
Photolysis of 1 in isooctane yielded Zand E-/3-chlorostyrenes in a 3"13 ratio. Addition of acetic acid to the photolysis medium gave 1-chloro-2-phenylacetate 6 as well as the Zand E-fl-chlorostyrenes. The acetate formed was found to increase with increasing the acetic acid concentration whereas the amount of fl-chlorostyrenes decreases, indicating that these products are formed by two competitive processes such as" Carbene -Styrenes (1) However, a plot of the ratio [Acetate]/[Styrenes] (see Table 1) as a function of [AcOH] shows a pronounced curvature indicating that the reaction mechanism is not a simple competition between reactions (1) and (2)  Furthermore, as the acetic acid concentration increases, the Z/E ratio increases up to values larger than one and when deuterated acetic acid is used as solvent, about 20% of the chlorostyrenes contained deuterium, indicating that there are at least two routes for the formation of the chlorostyrenes.

Laser-Flash-Photolysis
Laser-flash-photolysis of 1 in isooctane in the presence of pyridine produces a transient species with an absorption band peaking around 380 nm. This transient is not present in the absence of pyridine and is attributed to pyridinium ylide 3. Plot of the observed pseudo-first-order rate constants for the growth of the absorption at 370 nm vs [pyridine] is linear. The slope gives the rate constant for reaction of the carbene with pyridine, ky (7.5 + 0.5) x 109 M-is-1, and the intercept (extrapolation to [pyridine] 0) yields the reciprocal lifetime of the carbene 2 which is determined by the rate constant for 1,2-H shift, k (6.0 _+ 0.5) x 107 sat 25C.
The UV spectrum of a solution of diazirine 1 is isooctane consists of a structured absorption band in the 310 to 370 nm region, with low extinction coefficients, and a second one in the 220-290 nm range, very poorly structured, broad and intense. Transient absorption measurements are possible in the window 290-330 nm between these two bands. In the 305-330 nm region, a transient absorption appears within the response time of our analytical system (i.e. within the 3 ns following the 200 ps excitation pulse) and decays to zero with a lifetime around 18 ns. In the 295-305 nm range, a permanent (or very long-lived) absorption remains after the decay of the species with a 18 ns lifetime (see the first species (i.e. 100 ns after the excitation) since the correction which should be made to account for the diazirine bleaching is considered to be negligible. The second species is most probably the fl-chlorostyrene, the first absorption band of which lies around 295 nm. The first species, with a broad absorption over the 290-320 nm region, is assigned to the benzylchlorocarbene, 2, on the basis of the agreement between the rate constants for its decay measured at 300 nm (5.4 + 0.2) x 10 7 s -1 and for the growth of the pyridinium ylide (6.0 + 0.5) x 107 s-1, measured at 370 nm and extrapolated to [pyridine] 0.
The rate constant for the quenching of the carbene by acetic acid was determined from the rate of the carbene decay, kd ki + kq (AcOH], measured at 310 nm as a function of [AcOH]. Least squares analysis of six measurements for [AcOH] ranging from 1 to 15 mM gives: kq (9.61 + 1.0) x 108 Ms-.

DISCUSSION
We have mentioned above that a simple competition between reactions (1) and (2) cannot account for the product distribution and that a fraction of the chlorostyrenes is produced after interaction of the carbene with AcOD. 6 This leads us to consider the following mechanism where, for kinetic treatments, it is not necessary to define precisely the nature of the complex.
The changes of the Z/E ratio as a function of [AcOH] can be analyzed as follows. and, secondly, the incorporation of deuterium in 20% of the styrene products when the reaction is carried out in deuterated acetic acid indicates that (at least) one other mechanism must be effective to produce chlorostyrenes. The yield of chlorostyrenes formed from the complex is q) kq R"[AcOH]/(ki 4-kq[AcOH]), with R" k'i/(k2 + k'i), and the amounts of Z and E isomers are respectively q'//(/ + 1) and (p/(/ + 1). Then 2) x 10 s -1 is well defined because it has been measured many times as a function of many parameters. The error on kq is much larger because the range of (AcOH] which can be used to study the decrease of the carbene lifetime upon addition of acetic acid is very limited since we cannot use "high" [AcOH] yielding carbene lifetime shorter than 10 ns. The uncertainty is estimated to be +25% in this case. Therefore the maximum ranges of value for ki/kq are" 0.043-0.077 from laser flash photolysis measurements, 0.028-0.048 from the Z/E ratio and 0.034-0.046 from the [Styrenes]/[ Acetate] ratio. The fact that a common value for ki/kq determined by these three independent methods can be found between 0.043 and 0.046 strongly supports the mechanisms presented in Scheme 2. We can now examine the possible nature(s) of the "complex." For a better visualisation of the rearrangements leading to the products, deuterated acetic acid is used in the schemes.
i) Protonation of the carbene by the acid would result in the formation of an ion-pair (I) as intermediate complex. 6 Such an ion-pair can collapse to give the acetate as product or, in a reverse acid-base reaction, the AcOcan abstract a proton on the benzylic carbon to give PhCH CDC1 and AcOH. This simple mechanism cannot account for the fact that, when the reaction is conducted in AcOD,6  ii) An alternative mechanism involves the attachment of the carbene to the carboxyl group of the acid to give an ylide type complex (II) followed either by 1,2-H migration and decomposition of the ylide into AcOD and PhCH CHC1 or by a rearrangement to the acetate. This hypothesis is substantiated by the fact that 1,2-H migration has been demonstrated to occur within the ylide formed from the reaction of carbene 2 and pyridine. 5 However, 1,2-H migration to a negatively charged carbon is usually not a fast and easy process: the pyridinium ylide obtained from carbene 2 has a long life-time and reaction products resulting from the rearrangement of PhCH2CH2 in Phm-CHmCH3 by 1,2-H shift have not been observed whereas they were identified for the Ph--C3H6mCH2 and for Ph--C4Hs--CH2 carbanions which rearrange respectively into Phm'CHmC3H7 and Phm-CHCaH9 by 1,4 and 1,5-H shift.
Furthermore, the reaction in AcOD gives significantly more chlorostyrenes under photolytic conditions than in thermolysis and. the ratio [PhCH CHC1]/[PhCH CDC1], equal to 4 under photolysis, is only 1 in thermolysis. 6 Thus, it seems that, Ac with high acetic acid concentration at least, a significant route to non-deuterated chlorostyrenes exists. A mechanism involving the protonation of the excited diazirine (see Scheme 5) has been proposed but the quite short lifetime of the excited diazirine, indicated by the failure to detect its fluorescence and by its unit quantum yield of decomposition, seems to rule out this mechanism at low acetic acid concentrations.
Finally, the rate constant for the quenching of the carbene 2 by tetramethyl ethylene, TME, to form cyclopropane has been determined by measuring the rate of decay of the carbene absorption at 310 nm as a function of [TME]. Least squares analysis of measurements at 22C, for [TME] ranging from 4 to 25 mM, gave kq (7.9 + 0.7) x 10 8 M -1 s-. It has been previously believed that the reaction of carbene and AcOH was so rapid that other trapping reactions were not competitive. It is clear now that this is not true. For example, when I was irradiated in a mixture of TME and AcOH, the acetate and cyclopropane were produced in nearly equal amounts as well as traces of chlorostyrenes. 6