B. STEVENS AND B. E. ALGAR
3794
The Photoperoxidation of Unsaturated Organic Molecules. 111. Autoperoxidation in Polymer Films by B. Stevens1 and B. E. Algar Department of Chemistry, Shefield University, Shefield, England
(Received M a y 7, 1968)
The autoperoxidation yield of 9,10-dimethyl-l,2-benzanthracene(DMDA) in nylon 6.6 and in polyethylene films has been measured as a function of the partial pressure of environmental oxygcn. The observed behavior is similar to that exhibited by the same hydrocarbon in benzene solution under conditions where oxygen quenching of the singlet state of DMBA is restricted by the much lower diffusion coefficient of molecular oxygen in the polymers.
S + hv
Introduction Although the presence of 0 2 1 A a (lo2*) has yet to be established directly in systems undergoing photosensitized peroxidation, the available evidence2 strongly supports the intermediary role of this species in the sequential processes
lS* +S
lS*
IS*
+
302
-/+
3S*
+ lo2*
even when this is exothermic. Oxygen quenching of the sensitizer singlet state does, however, appear necessary to promote the formation of sensitizer triplet state by collision-induced intersystem crossing represented by lS* 302 +3S* 302
+
+
for those sensitizers with very high (100%) fluorescence quantum Processes 1-7 therefore constitute the simplest kinetic scheme consistent with the observed dependence of yhro2on such experimental parameters as incident light intensity, substrate concentration, and concentration of dissolved oxygen.3a T h e Journal of Physical Chemistry
+ 302+3S* + 302 3S*+S(+ hv)
3S*
+ n1 +3302
where the asterisk denotes electronic excitation of the sensitizer S and 11 is the unsaturated molecule (substrate) from which the peroxide ;\IOz is produced. An analysis of the dependence of the quantum yield of the over-all reaction yh102 on concentration of dissolved oxygen, in terms of lo2*participation, has shown that at least for (a) the autoperoxidation ( S r M ) of 9,10-dimethyl-l,2-benzanthracene3~ (DJIBA) and (b) the anthanthrene-sensitized photoperoxidation of DR'IBA and 9,lO-dimethylanthrac~ne~~ (DJIA), the singlet oxygen molecule is produced entirely by oxygen quenching of the sensitizer triplet state 3S* and not in the spinallowed oxygen quenching of the sensitizer singlet state, i.e.
+ hv
IS* +3s*
S* + 302+-S + lo2* '02*
+ lS*
+ 302+S + lo2* lo2*+302
102*
+ ;\I +1 1 0 2
(1)
(2) (3) (4)
(5) (6)
(7)
The asterisk denotes electronic excitation, internal conversion of IS* does not contribute significantly to the over-all relaxation rate of those aromatic hydrocarbons which undergo autoper~xidation~a (with the possible exception of naphthalene4), and process 4 includes both radiative and nonradiative relaxation of the triplet state 3S*. Under photostatioriary conditions, processes 1-7 lead to the following expression for ~ M O ~
- 1= {
l+k?[114] lC6}X
YMOZ
where k , is the rate constants of the ith process. Equation I provides a quantitative description of the variation in yhro2 with concentration of dissolved oxygen over the range to 3 X ill for the autoperoxidation of DMBA in b e n ~ e n e , ~where a at low oxygen concentrations such that k3 [Oz]
> k4
the nonlinear dependence of 1 / y ~ on 0 ~1/ [OZ]reflects a competition between the intramolecular and collisioninduced intersystem crossing of the sensitizer (processes 2 and 3); under the same conditions oxygen quenching of the sensitizer to the ground state would effectively inhibit the over-all reaction. On the other hand, k3[OZ]>> IC2 0 for sensitizers of very high (-100%) fluorescence quantum yield (ruo ~ linearly with 1/ [OZ] brene and DMA) and 1 / y ~ varies in the (high) experimentally accessible concentration range. 3 a The over-all quantum yield of peroxide formation is essentially determined (eq I) by the magnitudes of bikb [OZ],IG7 [AS 1) molecular quenching frequencies (k3 [OZ], relative to the unimolecular relaxation constants (kl, IC4, k 6 ) of the electronically excited species involved (IS*,%*, lo2*); it is therefore of interest to examine the oxygen concentration dependence of YXO, under conditions of limited molecular diffusion to establish the generality of the proposed mechanism and to obtain limiting values for the lifetimes of the metastable intermediates. This communication describes an investigation of the autoperoxidation of DAIBA in polymer films at various partial pressures of environmental oxygen.
-
Experimental Section Nylon 6.6 and polyethylene films of 0.003-in. thickness were refluxed in a concentrated ethanolic solution of D N B A until the concentration of solute in the film, monitored by absorption spectrophotometry, reached the value required (dictated by the optical density of the films at the actinic wavelength of 365 mp). At concentrations of the order of 114, obtained after a period of some 30 min, no evidence of solute aggregation was discernible in either absorption or emission, and films containing a similar concentration of pyrene exhibit no trace of the excimer fluorescence band characteristic of this aggregated solute.6 The washed and dried films were placed between two 0.25 in. thick plywood squares (2 X 2 in.) bolted together so that a 1 cm diameter circle of the film was visible through 1 ern diameter holes aligned in the plywood plates; the edges of these holes were covered with quartz plates and apertures bored through the assembly in the plane of the film permitted the circulation of a gas stream over the film. I n effect the films were suspended in a cylindrical plywood cell parallel to its quartz end windows and subjected to the passage of a Nz-Oz gas mixture of variable composition. After establishing the absence of photochemical reaction in the absence of environmental oxygen, the latter gas was admitted to the flowstream and the quan-
3795 tum yields of photoperoxidation computed from the time dependence of the solute optical density at the wavelength of incident radiation as described previously. The prevailing temperature was 23 j= 2". Absorption spectra recorded as a function of exposure time were found to exhibit an isosbestic point at 252 mp, the final spectrum being almost identical with that of the corresponding peroxide in cyclohexane and showing no absorption at wavelengths >350 mp. Concentrations of dissolved oxygen in polyethylene were estimated from its solubility (3.44 X M at 760 mm) in the completely amorphous polymer with a volume fraction of 0.65, and the partial pressure of oxygen PO, (atm) in the flowstream, i.e., from the rela tionship6
x
[02]= 2.23
10-3Po,M
Solubility data for oxygen in nylon 6.6 could not be found. The bimolecular encounter constant of oxygen in polyethylene was estimated from measurements of the relative fluorescence intensities F o / F of pyrene in this polymer in nitrogen and oxygen environments, and the relationship
Fo/F
=
1
+
k 0 , 7 ~ [ 0 2 ]=
1.7
This yields a value for ko, = 6.5 X lo8 M-' sec-l at 23"
with a fluorescence lifetime for molecular pyrene of 480 sec and the value for [Oz J quoted; in view of the absence of evidence for photoassociation (molecular self-quenching) of this solute in the polymer, it is concluded that this high value for ko, is due entirely to the rapid diff usion of molecular oxygen in polyethylene. Results and Discussion The quantum yields of DMBA autoperoxidation ([MI = 4 X M ) in nylon 6.6 and polyethylene films are plotted against the partial pressure of environmental oxygen on a reciprocal basis in Figure 1. The linear dependence of ~ / Y M O , on 1 / [ 0 2 ] is provided by eq I which reduces to
under the condition kz >> k3[0~]established by the much lower encounter constant k3 in the polymer and confirmed by the very small oxygen quenching effect (