J. Chem. SOC., Faraday Trans. I , 1982, 78, 3213-3222 Kinetics of the Photopolymerization of 2,5-Distyrylpyrazine in Solution BY EL-ZEINY M. EBEID, MAHMOUD H. ABCEL-KADER A N D SALAH E. MORSI* Centre for Graduate Studies and Research, University of Alexandria, P.O. Box 832, Alexandria and Chemistry Department, Faculty of Science, University of Tanta, Tanta, Egypt Received 22nd December, 198 1 The photoreactivity of 2,5-distyrylpyrazine (DSP) in methanolic solutions has been investigated using various spectroscopic and kinetic methods. The effect of the excitation wavelength has also been studied : light of wavelength i = 403 nm induces a photo-oligomerization process whereas light of wavelength A = 365 nm causes both photo-oligomerization and photopolymerization processes. These two processes are manifested as two stages in the reaction.The photo-oligomerization stage is characterized by higher values of quantum yield for chemical reaction (4 = 1.8 at 30 "C) and a higher energy of activation compared with the photopolymerization stage. The pattern of change in the absorption spectra of DSP solutions as a result of excitation with light of I = 365 nm depends upon the exciting light intensity. A sharp isosbestic point at 287 nm is obtained when high light fluxes (ca. 7.57 x lo-' ein min-') are applied, whereas three isosbestic points are obtained at 350, 325 and 287 nm on applying low light fluxes (ca. 5 x lo-* ein min-I) or by applying 403 nm light regardless of its intensity. The photo-oligomerization stage undergoes a power kinetic law and has an activation energy E, = 38.4k 1.5 kJ mol-I, an activation enthalpy AH* = 36.9k 1.5 kJ mol-' and an activation entropy AS* = - 184.3k0.3 J K-' mol-I.The photopolymerization stage obeys second-order kinetics with activation parameters : E, = 16.0k 1.0 kJ mol-', AH* = 13.6-t 1.0 kJ mol-' and AS* = -240.2k0.3 J K-' mol-I. The energy of activation in both cases accounts for a diffusion-controlled process that brings the monomer molecules to a spatial configuration conductive for the photocycloaddition of the olefinic double bonds. The highly negative values of the entropies of activation in both stages indicate that the activated states are highly ordered compared with the ground states. A mechanism for the reaction has been postulated that is based on a two-centre attack by an excited DSP monomer molecule on two ground-state molecules during the photo-oligomerization stage.A stepwise mechanism prevails in the photopolymerization stage. Both the fluorescence and excitation spectra of DSP in methanolic solution decrease in intensity as the U.V. irradiation time increases, indicating that the photoproducts have no detectable fluorescence in this spectral region. 2,5-Distyrylpyrazine (DSP) is a prototype compound in the field of lattice-controlled photopolymerization reactions. Its photoreactivity has been observed previously' but a comprehensive study of this photoreaction has recently been initiated by Hasegawa et al.,'. who showed via various analytical techniques that a photopolymerization process takes place both in solution and the solid state upon U.V.irradiation. The photochromism observed4 in DSP single crystals has prompted us to study the effect of light on DSP solutions in order to investigate this reaction at the isolated molecular level. This effect has been studied qualitatively,j? and a photo-oligomeriza- tion process has been found to occur as a result of light irradiation (3, 3 380 nm) of DSP-tetrahydrofuran solutions. The resulting photo-oligomer has been reported:' to consist mainly of pentamers. Apart from the kinetic studies previously made on the solid-state photoreactivity 32133214 PHOTOPOLYMERIZATION OF 2,5-DISTYRYLPYRAZINE of DSP,6*7 very little information is presently available on the kinetics of photopoly- merization in solution.In this study various spectroscopic techniques have been used to study the photoreaction of DSP in methanolic solutions. EXPERIMENTAL DSP was prepared using the method of Hasegawa et a1.8 The monomer was recrystallized from xylene, chromatographed on basic silica gel 60-80 mesh using AnalaR methylene chloride as eluent followed by vacuum sublimation of the resulting batch. The purity of the yield was verified by matching both the absorption and excitation spectra of 1.02 x mol dm-3 methanolic solution. The U.V. spectra were recorded using a Unicam SP 8000 spectrophotometer. Samples were thermostatted using apparatus similar to that described in ref. (3), consisting of a rotor, distilled water as a thermoregulating medium, a Pyrex beaker, a magnetic stirrer, a thermoregulator and a metallic cell-holder fixed vertically such that it is covered by the thermoregulating medium. The thermoregulator used was a copper coil in which thermostatted circulating water is pumped by an ultrathermostat (Haake model FE).The temperature fluctuations were found to be d 0.5 "C. The reaction vessel was a well stoppered silica cell of 10 mm optical path-length which was immersed in the thermoregulating medium by mounting in the cell holder. Irradiation did not begin until the prescribed temperature was reached. The solutions were deoxygenated by flushing with pure nitrogen gas for ca. 30 min before kinetic runs. To obtain the spectral runs the cell was removed, dried well and put into the spectrophotometer. The light source used was a mercury high-pressure lamp (HBO 200 W/2, Osram) combined with interference filters (maximum transmission at 365 and 403 nm with spectral width at half transmission of 8 nm).The incident light intensity was measured using ferrioxalate actinometry as described by Hatchard and P a r k e ~ . ~ Both fluorescence and excitation spectra were recorded on a Shimadzu RF 510 spectrofluorophotometer. The kinetic studies were performed within the range of concentrations where Beer's law holds, (i.e. absorbance changing linearly with concentration). RESULTS A N D DISCUSSION EFFECT OF EXCITATION WAVELENGTH Fig. 1 shows the effect of 403 nm irradiation on a 1.01 x mol dm-3 DSP methanolic solution with increasing time. Three isosbestic points are obtained at 3 5 0 , 325 and 287 nm.The reaction product is believed to be a lower chemical aggregation of DSP molecules and it exhibits absorption (besides a band at ca. 333 nm) in the form of two bands at ca. 225 and 260 nm. A linear correlation has been found between the absorbances at the two wavelengths ?L = 378 and 1 = 260 nm, indicatinglO that the observed spectral change is due to a single photoreaction. The remaining bands at ca. 333 and 225 nm overlap too much to be monitored. Since the irradiation conditions in this case compare favourably with those of Hasegawa et a1.5*6 we expect this photoproduct to be a DSP oligomer. The pattern of change in the spectrum was independent of the light intensity; however, increasing the light intensity leads to an increase in the rate of photoreaction.Upon irradiation with light of 3, = 365 nm (rather than 3, = 403 nm) the reaction was found to occur via two clearly distinguishable stages with different rates, as indicated from the analysis of the absorption spectra (see fig. 7). The earlier stage involves the formation of an oligomer and can be produced by applying low light intensities (ca. 0.5 x lop7 ein min-l). In this case, spectra similar to those shown in fig. 1 are obtained as the period of irradiation is increased, but a 'stationary absorption'E-2. M. EBEID, M. H. ABDEL-KADER A N D S. E. MORSI 3215 1 0 v) * 2 0 8 f b) 0 5 0 6 e 0, - 0 4 13 x c * v) - ; 0 2 200 225 250 275 300 325 350 400 h/nm FIG. 1 -Effect of irradiation with light of wavelength A = 403 nm on the absorption spectrum of a 1.01 x 10-5 mol dm-3 DSP methanolic solution at 30 O C : (-) fresh sample; (-.-*-.) after 17 min irradiation; ( * .. . . - ) after 52 min irradiation (intensity = 5.8 x lo-* ein min-l). 0.0 8 - I c .* E 0.06 I m vl + .d c 3 w 2 0.04 e 0, -2 W . Y 0.02 0 50 100 relative intensity (arb. units) FIG. 2.-Effect of light intensity on the rate constants of the photo-oligomerization process. stage is obtained after reaching ca. 10% reaction. At this ‘stationary absorption’ the reaction rates fall dramatically and the reaction becomes too slow to be followed. This early stage exhibits a relatively high initial quantum yield (4 = 1.8 at 30 “C). We attribute this ‘stationary absorption’ to the formation of a stationary concen- tration of the oligomer, since the rate of its generation equals the rate of its consumption either by photodissociation to the monomer or by photopolymerization to higher molecular aggregates.The reason why the photo-oligomer reacts is apparent from its considerable absorption component in the 365 nm region of the spectrum.3216 PHOTO POL Y M ERI Z A TI ON OF 2 3 - D ISTY RY L PY R A Z INE .g 0.8 .- 2 ' Y 0.21 .r: I 0 225 250 275 300 325 350 LOO X/nm FIG. 3.-Effect of irradiation on a 1.25 x mol dm-3 DSP methanolic solution with light of wavelength A = 365 nm (intensity = 7.57 x lo-' ein min-I): (-) fresh; (---.-) after 45 min irradiation; (. . . . . .) after 105 min irradiation ; (- - - -) after 195 min irradiation. U SCHEME 1 .-(Q) Photopolymerization of DSP; (6) activated state during the photo-oligomerization stage.E-Z.M. EBEID, M. H. ABDEL-KADER AND S. E. MORSI 3217 Fig. 2 shows the effect of 365 nm light intensities (in arbitrary units) on the rate constants for this stage at 30 O C . The curvature in the plot is attributed to the multiphoton processes occurring during this stage. The method of extracting the rate constants for this stage is mentioned in the kinetic part of this work. On applying higher intensities of the 365 nm light (ca. 7.6 x ein min-') only one sharp isosbestic point at 207 nm is obtained. Fig. 3 shows the effect of intense 365 nm light on the absorption spectrum of a 1.25 x 10-5 mol dm-3 DSP methanolic solution. The sharpness of the isosbestic point at 287 nm indicates that the polymer being formed has a very low polydispersity. The initial quantum yield for chemical reaction during this stage is relatively low (4 = 0.1 1 at 30 "C).The irradiation product in this case has bands at ca. 230 and 287 nm, the band at 378 nm being due to the original species. There is also a linear correlation between the absorbances at the two wavelengths ;C = 378 and R = 230 nm, indicating', that the observed spectral change is due to a single photoreaction. The decrease in wavelength of the absorption maximum is attributed to the loss of extended conjugation along the DSP molecule owing to the consumption of the double bonds in forming cyclobutane rings according to scheme (1 ;I). The percentage of these consumed double bonds increases as the number of the polymerized monomeric units ( n ) increases. This is accompanied by a greater blue shift of the absorption maxima.The same argument has been proved3? l1 for the photoproducts of DSP in the solid state. EMISSION AND EXCITATION SPECTRA Fig. 4 shows the effect of 365 nm light on the excitation spectrum of a 0.430 x lop5 mol dm-3 DSP solution. The uniform decrease in intensity of the excitation bands with no change in the spectral pattern indicates that the product has no detectable fluorescence in this range of the ,pectrum. This is also confirmed by the uniform decrease in intensity of the fluorescence band at 450 nm as a result of excitation with light of R = 365 nm. The fluorescence spectrum of DSP in solution is shown in fig. 5 (&,, = 365 nm, concentration 1.24 x lop6 mol dm-"). KINETIC ANALYSIS OF THE FIRST STAGE The absorbance change (x) at the A = 378 nm is monitored as a function of irradiation time with 365 nm light of very low intensity (ca.0.7 x ein min-l). The value of x corresponds to ( A , - A , ) where A , is the initial absorbance and A , is the absorbance at time t. The A, was kept the same for all kinetic runs. Simple kinetics based on orders of reactionl29l3 was not applicable to this stage, but a power law has been applied. Fig. 6 shows a plot of x' against t. The applicability of this power law is due to the complexity of this stage, which does not account for a simple process. The complexity of this stage is also demonstrated by the relation between the rate constants and light intensity as shown in fig. 2. The slopes of the lines in fig. 6 give the rate constants at various temperatures.Thermodynamic data were obtained from a least-squares fit14+ l5 to four data points, giving an activation enthalpy for the photochemical process of AH* = 36.9_+ 1.5 kJ mo1-l. The entropy of activation, AS*, is - 184.3 k0.3 J K-I mol-I. The energy of activation E, = 38.4+ 1.5 kJ mo1-l. The highly negative value of the entropy of activation indicates that the activated state is highly ordered compared with the ground state.3218 PHOTO POLYMER I Z A T I ON OF 2 3 - D ISTY RY L PY R A Z INE J. i ' I I I I 250 300 3 50 4 00 h/nm FIG. 4.-Changes in the excitation spectrum of a 0.43 x 1 0-5 mol dmP3 DSP methanolic solution as a result of irradiation with exciting light of wavelength E. = 365 nm: (---) fresh sample; ( . . .. .) after 1 min irradiation; (-.-.-.-) after 20 min irradiation. Note that the spectrum was obtained by following the 450 nm fluorescence band. KINETIC ANALYSIS OF THE SECOND STAGE Fig. 7 shows a plot of the fractional change a, where a = (A, - A , ) / A , , against the time ofirradiation with intense 365 nm light (ca. 7.0 x ein min-l). The dotted parts of the curves correspond to the first stage, which is not well resolved under conditions of intense irradiation. A second-order rate equation applies to the second stage, and is given by16 where a is the number of moles of the reactant at the beginning of the reaction and x is the number of moles which have reacted at time t. Taking the initial absorbance of the DSP monomer, A,, to be proportional to a, A, to be proportional to (a-x) and (A,-A,) to be proportional to x, we can write the second-order rate equation in the form 1 A , - A , A , - A , A, ( A, )=!=)* k t = - ~E-Z.M. E B E I D , M. H. ABDEL-KADER A N D S. E. MORSI 3219 h/nm FIG. 5.--Changes in the fluorescence spectrum of a 1.24 x lop6 mol dmT3 DSP methanolic solution as a result of irradiation with 365 nm light (A = 365 nm): (-) fresh sample; (-.-.-.-) after 10 min irradiation; (-----) after 25 min irradiation; ( . . . . .) after 58 min irradiation. 0 5 10 15 tlmin FIG. 6.-Plot of x2 values as a function of time t: 0, 25; 6, 34; 0, 43, and 0, 49 OC.3220 PHOTO POLY MER I Z A TI ON OF 2 3 - D ISTY RY L PY R A Z I NE 0. 0 T . h j- 0. 0 s 0 . 0 10 20 30 40 50 60 t/min FIG. 7.-Piot of a against t showing the fractional change in the absorption maximum as a result of irradiation with intense (ca.7 x lop7 ein min-I) light of wavelength 1 = 365 nm: 0 = 13; 0 , 21 ; 0, 30 and 0, 40.5 "C. Fig. 8 shows a plot of the values ( A , - A , ) / A , A , against t after correction by subtracting values of the ordinates corresponding to the first stage. The slopes of the straight lines give the rate constants at the cited temperatures. A statistical analysis gives the energy of activation Ea = 16.0 1 kJ mol-l, the enthalpy of activation AH* = 13.6+ 1 kJ mol-l and the entropy of activation AS* = -240.2f0.3 J K-l mol-l. This highly negative value of AS* indicates that the activated state is highly ordered compared with the ground state. GENERAL DISCUSSION AND CONCLUSIONS In the light of the kinetic data we can propose a mechanism for the photoreaction of DSP in solution.DSP monomer molecules are electronically excited by both 365 and 403 nm light. The excited monomer molecule then makes a simultaneous two-centre attack on two monomer molecules, forming an activated state that might be represented by scheme (1 6). This activated state then leads to chemically bonded oligomers. The high monomer concentrations at the beginning of the reaction may allow this process to occur. The highly negative value of entropy of activation (AS* = - 184.3 0.3 J K-l mol-l) is an indication of this ordered activated state. ThisE-Z. M. EBEID, M. H. ABDEL-KADER A N D S. E. MORSI 322 1 0 I 0 10 20 30 40 50 60 70 80 tlmin FIG. 8.-Application of a second-order kinetic law to the change in intensity of the absorption maximum as a function of irradiation time ( t ) using intense (ca.7 x ein min-I) light of wavelength I. = 365 nm. Note that the parts of the ordinate corresponding to the first stage are subtracted. 0, 13, 0, 21, 0, 30 and 0, 40.5 "C. model explains the relatively high rates and high energy of activation characterizing this stage. On exciting with light of a wavelength that these oligomers can absorb (e.g. 365 nm) the oligomer molecules continue incorporating other monomer molecules at both ends in a stepwise mechanism forming a relatively high molecular-weight polymer. This latter type of reaction constitutes the second stage described above. Because of the pronounced difference in resonance stabilization energies of the olefinic double bonds, the resulting photo-oligomer exhibits photoreactivity which is different from that of the monomer. This explains the low quantum yield and the low rates of reaction during the second stage compared with the first stage.The relatively low values characterizing the energies of activation for this photo- reaction account for a diffusion-controlled process and correspond to the energy necessary to bring the reacting molecules to a spatial configuration suitable for the addition of the olefinic double bonds. For a two-centre attack to occur (e.g. the first stage) molecular diffusion involves two monomeric molecules, whereas a stepwise mechanism (e.g. the second stage) involves the diffusion of only one molecule. This explains why the energy of activation for the first stage is nearly twice that of the second stage.The study also shows that the photoproducts in both the first and second stages are non-fluorescent. This is attributed to the loss of extended conjugation and coplanarity in case of oligomer and polymer molecules. We acknowledge with gratitude the financial support of the Egyptian Academy of Scientific Research and Technology.3222 PHOTOPOLYMERIZATION OF 2,5-DISTYRYLPYRAZINE C. F. Koelsch and W. H. Gumprecht, J. Org. Chem., 1958, 23, 1603. M. Hasegawa and Y. Suzuki, J . Polym. Sci., Part B, 1967, 5, 813. M. Hasegawa, Y. Suzuki, M. Nakanishi and F. Nakanishi, Prog. Polym. Sci. Jpn, 1973, 5, 143. E. M. Ebeid, S . E. Morsi and J. 0. Williams. unpublished results. M. Hasegawa, Y. Suzuki and T. Tamaki, Bull. Chem. Soc. Jpn, 1970, 43, 3020. H. Nakanishi, Y. Suzuki, F. Suzuki and M. Hasegawa, J. Polym. Sci., Part A-I, 1969, 7, 753. M. Hasegawa, Y. Suzuki, F. Suzuki and N. Nakanishi, J. Polym. Sci. Part A - I , 1969, 7, 743. J. G. Hatchard and C. A. Parker, Proc. R . SOC. London, Ser. A , 1956, 235, 518. T. Tamaki, Y. Suzuki and M. Hasegawa, Bull. Chem. Soc. Jpn, 1972, 45, 1988. G. M. Harris, Chemical Kinetics (D. C. Heath and Co., Chicago, 1966). ’ C. H. Bamford, G. C. Eastmond and J. C. Ward, Proc. R . Soc. London, 1963, 271, 357. lo H. Mauser, Z . Naturforsch., Teil B, 1968, 23, 1025. l 3 C. R. Metz, Theory and Problems of Physical Chemistry (McGraw Hill, New York, 1976), chap. 10. l4 E. M. Ebeid, S. E. Morsi and J. 0. Williams, J. Chem. Soc., Faraday Trans. I , 1979, 75, 1111. l5 C. I. Chase, Elementary Statistical Procedures (McGraw Hill, New York, 1967), p. 91. l6 P. J. F. Griffiths and J. D. R. Thomas, Calculations in Adtunced Physical Chemistry (Edward Arnold, London, 1962), p. 140. (PAPER 1 /1974)