Bimolecular cyclization of di-p-tolylamine to 2,7-dimethylacridine in the presence of bromoform: microsecond flash photolysis study

J. Photochem. Photobiol. A: Chem.. 66 (1992) 205-214 205 Bimolecular cyclization of di-p-tolylamine to 2,7_dimethylacridine in the presence of bromoform: microsecond flash photolysis study M. F. Budyka+, G. V. Zakh arova, 0. D. Laukhina and M. V. Alfimov Photochemistry Department, Institute of ChemicaI Physics, Russian Academy of Sciences, 142432, Chernogolovka, Moscow (Russia) (Received November 18, 1991; accepted January 30, 1992) Abstract The W irradiation of di-p-tolylamine and bromoform results in the formation of 2,7- dimethylacridine (with a quantum yield of 0.20 kO.04). The primary transient spectrum (L%x = 700 nm), recorded following microsecond flash lamp excitation, can be attributed to the amine cation radical and, probably, the neutral diarylnitrogen radical, which arises from the former on deprotonation. Both of these radicals can recombine with the dibro- momethyl radical to give the intermediate IMl. This is transparent in the visible region of the spectrum and is transformed into the second intermediate IM2 (A,,=540 nm), which finally gives rise to acridine. The observed first-order rate constant of IM2 formation is dependent on amine concentration, indicating the participation of an additional amine molecule in the transformation of KM1 to IM2. A new reaction scheme is proposed which assumes that the attack of the alkyl radical is predominantly directed towards the nitrogen atom of the amine radical. 1. Introduction The photochemical reactions of aromatic amines with halogenomethanes have been investigated by a number of workers [l-lo]. These species have been shown to form charge-transfer complexes (CTCs), which dissociate on UV irradiation producing triphenylmethane dyes [l, 5, 71. Recently, it has been observed that diphenylamine photocyclizes to acridine compounds in the presence of CHX3 or CX,, (X=Cl, Br) [7, 8, 111. In comparison with the well-known intramolecular photocyclization of diarylamines to carbazoles [12], the formation of the acridine nucleus involves at least two molecules - one amine and one halogenomethane; therefore this reaction involves bimolecular photocyclization. The reaction has been proposed to proceed via photoinduced electron transfer from amine to halogenomethane, followed by recombination of the amine cation radical and alkyl radical and intramolecular cyclization of the intermediate formed [S, 113; The intermediate with an absorption band at about 540 nm, which has been observed during the low-temperature photolysis of di-p-tolylamine (DTA) and bromoform, has been shown to transform quantitatively to 2,7_dimethylacridine in a first-order process 1131. ‘Author to whom correspondence should be addressed. lOlO-6030/92/$5.00 0 1992 - Elsevier Sequoia. All rights reserved 206 This paper deals with the results of microsecond-to-second flash photolysis of DTA and bromoform in toluene solution at room temperature to give the formation of 2,7_dimethylacridine. Particular attention was given to the kinetic behaviour of the intermediate observed at 540 nm. 2. Experimental details Amine was purified by crystallization and bromoform by vacuum distillation prior to use. Toluene (spectrograde, Reachim) was used as received. All experiments were carried out at room temperature in quartz cuvettes (path length, 1 cm). For quantum yield determinations, the solutions were degassed by repeated freeze-pump-thaw cycles. Irradiations were performed with a high-pressure mercury lamp; the line at 365 nm was selected using appropriate glass filters. The intensity of the incident light, measured using a PP-107 actinometer, was 1 x 10m6 W cm-‘. Electronic absorption spectra were recorded on a Specord M40 spectrophotometer. Flash photolysis was performed with a flash energy of 90 J and a duration half- time of 15 ps; aerated solutions were used. The details of the equipment employed have been described previously [14]. The rate constants of the thermal stages were calculated provided that the kinetic order observed was within the time region of two or more half-lives (error limits are +20%). 3. Results 3.1. Quantum yield The irradiation of a toluene solution of the ClC DTA-CHBrs with 365 nm light from a mercury lamp resulted in an increase in the optical density in the UV and visible region (Fig. 1); the final difference spectrum correspond to that of the 2,7- dimethylacridinium cation [13]. The quantum yield # of 2,7_dimethylacridine formation was defined as the initial linear slope of the AD-l,t curve at increasing but short illumination times (broken Fig. 1. Absorption spectra of a degassed toluene solution, containing [DTA]=6.25 X 10m3 M and [CHBr,] =0.25 M, before (1) and after (2) 3 min of UV irradiation; (3) difference spectrum ((3) = (2) - (1)); (4) 2,7_dimethylacridinium chloride (5.6 X lo-’ M). 207 0.6 AD 0.6 . I& J crn-2 Fig. 2. Absorbance vs. exposure plots on UV irradiation of the CTC DTA-CHBr, in aerated (1, and degassed (2) solukons in- toluene: line in Fig. 2) according to UI = &AI& 1 - 10-D,,) x 8.36 x lo- 6 (1) where AD is the optical density of the acridinium formed, E (M-’ cm-‘) is its molar extinction coefficient, I0 (W cmm2) is the intensity of the incident light, h (nm) is its waveletigth, t (s) is the irradiation time and Do is the optical density of the CTC at the irradiation wavelength. A factor of 8.36 X 10S6 is present because the extinction coefficient, intensity and wavelength of the light are expressed in terms of different spatial units. Figure 2 shows some representative plots based on eqn. (1) obtained for aerated and degassed solutions. The initial slopes of the AD-&r curves were the same in the absence and presence of oxygen; the quantum yield was calculated to be 0.20&0.04. 3.2. Dilute sohhons Figure 3 shows the transient difference spectra of 2.5 X 10m3 M DTA and 0.1 M CHBr, in toluene solution for different delay times after the initiating flash. The decay of the optical density at 700 nm, within a time scale of up to hundreds of milliseconds, obeyed the kinetics of a second-order equal concentration reaction (Figs. 4 and 5). The inset of Fig. 4 shows the plot based on the expression 1 1 =- AD-ALI_ A& C-4 where AD, A& and AD, are the optical density changes at time f, zero and infinity (extrapolated value) respectively, k is the second-order rate constant, I is the path length of the photolysis cuvette (1 cm) and /cobs= k/J is the observed rate constant. The kobS values were dependent on the time region of observation. The tail of the kinetic curve decayed exponentially (Fig. 5). The rate constants determined at different wavelengths and for different time regions are given in Table 1. 3.3. Concentrated solutions Figure 6 shows the transient difference spectra of 0.025 M DTA and 1 M CHBr3 in toluene solution. The primary transient absorption at 700 nm was identical with 208 0.04 4?0 500 600 nm 700 A, , AD 0 26 22 18 14 10 Y, 1ooocm-1 ' Fig. 3. Transient difference spectra of 2.5 X 10e3 M DTA and 0.1 M CHBrs in toluene solution. Delay times after the flash are: 200 us (1); 2 ms (2); 20 ms (3); 200 ms (4); 1 s (5). 0.04 AD 0.03 0.02 0.01 0 - I 1 I I I 0 4 6 12 16 20 t, ms Fig. 4. Time dependence of the absorbance at 700 nm, resulting from the flash photolysis of 2.5 x lop3 M DTA and 0.1 M CHBr, in toluene solution (time region, 0.2-20 ms). The inset shows a plot of the data according to eqn. (2). that observed in dilute solution. The decay kinetics of the optical density were of second order in the short-time region and first order in the long-time region. The observed rate constants are given in Table 2. However, in contrast with the dilute solutions, after the disappearance of the primary spectrum, a new transient absorption at 540 nm was observed (Fig. 6, spectrum 3). The appearance and disappearance of this new intermediate (Fig. 7) were first- or pseudo-first-order processes. The rate constant of intermediate formation was kf = 0.27 s-l and that of intermediate decay was kd =0.022 s-l. In previous work, the constant kd was shown to be independent of amine or bromoform concentration [13]. In order to determine the effect of the solution composition on the kr value, the flash photolysis was carried out using solutions of varying DTA concentration with the CIIBr, concentration held constant, and vice versa. 209 0.012 AD 0.009 0.008 0.003 0 0 260 400 600 800 t, ms Fig. 5. Time dependence (time region, l&800 ms) of the transient absorbance at 700 nm in toluene solution containing 2.5 ~10~~ M DTA and 0.1 M CHBr3 (points); curves: (1) second- order approximation; (2) first-order approximation. TABLE 1 The observed rate constants of the transient absorption decay in the red region of the spectrum after flash photolysis of 2.5~ 10M3 M DTA and 0.1 M CHBr, in toluene solution Wavelength (nm) 660 700 800 Time region Reaction (ms) order 1-14 2 4-40 2 l-20 2 Z-90 2 20-200 2 125-700 1 2-90 2 la-700 2 k ob: 6-‘) 3.3 x lo4 3.8X1@ 1.3 x 104 4.0 x l@ 1.8X ld 4.8 9.6x l@ 4.0x ld aFor the second-order reaction kobs=k/d (see eqn. (2)). In Fig. 8 the measured first-order rate constant for intermediate formation &) is plotted as a function of the DTA concentration (region, 8.3 x 10-3-0.125 M). From the slope of the straight line the second-order rate constant k2 was obtained as 10.9 M-1 s-1 . A change in CHBr3 concentration (region, 0.15-l M) did not affect the kf value, provided that the DTA concentration was constant. 4. Discussion In a previous publication [13], the following scheme of photoinduced bimolecular cyclization of DTA to a&dine was proposed, including the formation of two intermediates (IMl and IM2) (AZ-= 4-C&C&Z,- for DTA). 210 Ar$II + cElEw3 = hv Ar2NH*cHBr3 - Ar$H+ - + CHBr; + Br- -----+ -HBr CTC kf 'd - -Brl. - -HBr IM1 H nf2 H - v AH+ R: H, Alkyl Scheme 1. Proposed mechanism of photoinduced acridine formation according to ref. 13. The intermediate IMl, which is transparent in the visible region of the spectrum, has been proposed to transform by a unimolecular process to the second intermediate IMZ, which absorbs at A,,,,= 540 nm. IM2 gives rise to acridine, which is protonated in our reaction conditions to the acridinium cation AH+ by the HBr evolved, In the present paper, the observed rate constant kr of IM2 formation has been found to depend on amine concentration (Fig. S), indicating the participation of an additional amine molecule in this reaction. Therefore Scheme 1 must be corrected. The transient absorption spectra recorded immediately following flash lamp ex- citation (Figs. 3 and 6) correspond to that of the amine cation radical Ar,NH’+ (&la =705 nm, acetonitrile [15]). The spectrum of the neutral ditolylnitrogen radical Ar2N’ (h,,= 735 nm, benzene [16]) is close to that of Ar2NH’+ _ It is worth noting in this context that the molar extinction coefficient of the diphenylamine cation radical Ph,NI-Y+ at A,,,== 670 nm (E= 1.8 x lo4 M-l cm-‘) is an order of magnitude larger than the extinction coefficient of the diphenylnitrogen radical Ph& at the same wavelength (e=2.2X103 M-l cm-‘) [17]. Assuming that the same ratio exists between the E values of Ar,NH’+ and Ar2N generated from DTA, and taking into account that Ar,NH” is the only primary product of photoinduced electron transfer from DTA to CHBr3, the short-time 0.06 AD 0.04 0.02 0 *-~ 500 600 700 A, nm 26 22 18 14 10 v, 1000cm1 Fig. 6. Transient difference spectra of 0.025 M DTA and 1 M CHBr, in toluene solution. Delay times after the flash are: 20 ms (1); 200 ms (2); 1 s (3). 211 TABLE 2 The observed rate constants of the transient absorption decay in the visible region of the spectrum after flash photolysis of 0.025 M DTA and 1 M CHBr, in toluene solution Wavelength Time region WO (ms) Reaction order k ohs= +-‘I 540 2-90 2 7.6 x lo3 650 2-90 2 3.0 x 103 20-350 2 3.2 x 10’ 200-1000 1 2.7 700 l-20 2 2.6 x 103 2-90 2 5.3 x 102 Z&ZOO 2 1.2x lo2 20~1000 1 2.4 800 2-100 2 1.1 x lo4 l&150 2 1.9 x 103 15@600 1 5.7 aFor the second-order reaction kabJ=k/d (see eqn. (2)). 0.06 AD 0.05 0.02 0.01 t 0 J 0 30 50 90 120 150 t, set Fig. 7. Time dependence of the intermediate (IM2) absorbance at 540 nm in toluene solution containing 0.025 M DTA and 1 M CHBr,. The inset shows plots of the data according to first- order reactions. bimolecular kinetics of the transient decay at 700 nm can be attributed, at least partially, to the recombination reaction of the CHBrz’ radical and the amine cation radical. The neutral ArzN radical, which can arise from deprotonation of AraM+, can also take part in the reaction with CHBr2’ and makes a contribution to the subsequent kinetic behaviour of the 700 nm band in the middle-time region. The dependence of the kobS value on the time region of observation indicates the occurrence of other recombination reactions of primary and secondary radicals. These reactions give rise to the hydrazine, semidine and phenazine derivatives [lg, 193 and are not discussed in this paper. The exponential tail of the kinetic curve (Fig. 5, curve 2) may be assigned to the reactions of radicals with sohrent molecules. 212 1.6 , kf’ s-1 1.4 1.2 1.0 0.8 0.6 0.4 0.2 0 0 0.02 0.04 0.06 0.06 0.10 0.12 0.14 [DTA]. M Fig. 8. First-order dependence of the observed rate constant of intermediate (IMZ) formation (at A,,,= 540 nm) on DTA concentration. The kinetics of primary radical disappearance (wavelength, 700 nm) are not coincident with the kinetics of intermediate formation (wavelength, 540 nm); a delay time exists between these two processes. Therefore, it is confirmed that the intermediate observed at 540 nm is IM2, and the primary radicals give rise to the intermediate IMl, which does not absorb light in the visible region of the spectrum and could not be observed in our experiments. However, the structures of both IMl and IM2 are different from those previously assumed. The corrected reactions following the photoinduced electron transfer from diar- ylamine to bromoform, resulting in acridine formation, are given in Scheme 2. (The first stages of the reaction - the formation of the charge transfer complex and its photodissociation on W irradiation - are similar to those depicted in Scheme 1, and are therefore not shown in the corrected Scheme 2.) ArpH’- + cHBr; - L Ar&xIBr~ J-H+ j-H+ + +Ar.pH Ar2N' + cHBr> - Ar*N-CHEW2 - A+N=CHBr ___+ -Br-- -Hf IMl H II62 H Scheme 2. Corrected scheme of photoinduced acridine formation. 213 As stated above, we propose two main pathways of IMl formation: from recom- bination reactions of the CHBr; radical with Ar,NI-P and ArzN radicals. In accordance with the distribution of the unpaired electron in these species, where the spin density on the nitrogen atom predominates over that on the benzene ring carbon atoms [20], the attack of the CHBr; radical should be predominantly directed towards this atom. The debromination reaction follows recombination, producing the methyleneimine derivative, i.e. IMI (Scheme 2). This is known to be electrophilic and reacts with aromatic amines [21] to give the intermediate IM2. The intramolecular cyclization of IM2 results in the formation of acridine. In concentrated solutions ([DTA] > 10m2 M) the ratio between the observed rate constant of IM2 formation (kr=/Q[DTA]) and that of IMZ decay (kd), where kf>kd, favours the observation of this intermediate. In dilute solution ([DTA] = 2.5 X 10e3 M) both constants become of the same order (approximately 10e2 s-l), and IM2 cannot be observed in the spectrum. 5. Conclusions The photoinduced formation of 2,7-dimethylacridine from DTA and bromoform was studied using microsecond flash photolysis. The primary photoproduct with an absorption at 700 nm and the secondary intermediate with an absorption at 540 nm were registered. The study of the kinetic behaviour of this intermediate, i.e. the dependence of the observed rate constant of formation on amine concentration, led us to revise the previously proposed reaction scheme [13] and assume a new scheme, including new structures for the intermediates IMl and IM2. The new scheme can be generally applied to the bimolecular photocyclization of diarylamines to acridines in the presence of halogenomethanes. References 1 R. H. Sprague, H. L. Fichter and E. Wainer, Photogr. Sci. Eng., 5 (1961) 98. 2 W. C. Meyer, J. Phys. Chem., 74 (1970) 2118, 2122. 3 E. A. Fitzgerald, P. Wuelfing and H. H. Richtol, I: Phyr. Chem., 75 (1971) 2737. 4 T. Iwasaky, T. Sawada, M. Okuyama and H. Kamada, J. Phys Chem., 82 (1978) 371. 5 K. Wyrykowska, M. Grodowski, K. Weiss and T. Latowski, Photochem. Photobiol., 28 (1978) 311. 6 T. Miwa, M. Koizumi and H. Genda, Nippon Shashin Gakkaishi, 42 (1979) 174, 246. 7 Y. Wada, S. Ito, M. Yamamoto and Y. Nishijima, Ann. Meeting of SPST, Japan, 1984, p. 124. 8 V. A. Sajnikov, A. G. Strukov, M. G. Stunjas, S. P_ Efimov, 0. M. Andreev and M. V. Alfimov, Dokl. Akad. Nauk. SSSZ?, 288 (1986) 172_ 9 H. Isaka, S. Suzuki, T. Ohzeki, Y. Sakaino and H. Takahashi, J. Phofochem.. 38 (1987) 167. 10 E. I. Mal’tsev, V. A. Tkachev and A. V. Vannikov, 1. Photochem. Photobiol. A: Chem., 43 (1988) 73. 11 M. F. Budyka, 0. D. Lauhina, V. A. Sajnikov, M. G. Stunjas and M. V. Alfimov, DokZ. Akad. Nauk SSS& 309 (1989) 1126. 12 E. W. Forster and K. H. Grellmann, .Z. Am. Chem. Sot., 94 (1982) 634. 13 M. F. Budyka, 0. D. Lauhina, A. A. Korkin and M. V. Alfimov, Zzv Akad. Nauk SSSR, Ser. l&m., (1991) 564. 14 V. M. Savin, P. M. Astafev. B. I, Pogonin and A. K. Chibisov, IN Timitjazevskaja SeMo Hoz~ajsfvennaja Academia, ( 1986) 177 (USSR). 214 15 G. Cauquis, H. Delhomme and D. Serve, Tetrahedron Lett., (1971) 4113. 16 F. A. Neugenbauer and H. H. Fischer, Chem. Ber., 98 (1965) 844. 17 K. H. Schmidt, A. Bromberg and D. Meisel, J. Phys. Chem., 89 (1985) 4352. 18 G. Cauquis, H. Delhomme and D. Serve, Electrochim. Actu, 21 (1976) 557. 19 S. L. Patai (ed.), The Chetnisty of Hydrazo, Am and A.zuxy Group, Wiley, New York, 1975. 20 K. Yoshida, Elecrrooxidation in Oganic Chemkty, Wiley, New York, 1984. 21 H. Bohme and H. G. Viehe (eds.), Iminium Salts in Organic Chemistry, Wiley, New York, 1976, Part 1.