3137
PEIOTOHYDRATIOIV OF ?YRIMIDINES
the free energy of the radical ions (AG2 > 0.4 eV). Similarly, the ash spectrum of pyrene-dimethylaniline in acetonitrile12shows the presence of the pyrene radical anion, the pyrene triplet, and possibly the dimethylaniline radical cation, once again in accordance with the fact that thc radical ions lie about 0.8 eV above the pyrene triplet The observation of Nakato, et aZ.,I0 that the naphthalene triplet and the TillPD radical cation are formed on flashing the system TMPD-naphthalene in ethanol (where the free ions certainly lie above the naphthalene triplet) is also to be expected. Examples where tjlie energy of the free ions formed in the reactions J S lower than that of the lowest triplet are provided by the system p-dimethoxybenzene-chlorani120 (the chloranii triplet lying more than 1 eV above the ions) and ,ai,N.-dimethyl-2-naphthylamine dimethylisoDhthalateZ3(where the dimethvlnaththvlamine triulet " & . lies about 0.2 eV above the ions). In these two only ions are seen in the gash spectrum, and in the
second case it can be concluded that no triplet (the spectrum of which was known) was formed in the reaction. These observations generally confirm the experimental findings reported here. Acknowledgments. The authors wish to express their gratitude to Mrs. S. Reiche for her assistance with much of the experimental work, and one of us (A. R. W.) wishes to thank the Alexander-von-Humboldt-Stiftung for the award of a fellowship, during the tenure of which part of this work was carried out. Thanks are also due t o Dr. V(ra Kiihnle, who purified the solvent used in this work. (19) Y. Nakato, N. Yamamoto, and H. Tfiubomura, Bull. Chem. sot. JaTap., 40, 2480 (1967). (20) K. Kawai, N. Yamamoto, and H. Tsubomura, {bid., 42, 369 (1969). (21) M. Koizumi and H. Yamashita, Z. Phys. Chem. (Frankfurt a m Main), 57, 103 (1968); 13.Yamashita, H. Kokubun, and M. Koizumi, ~ ~ iC ih m . . SOC. Jap., 41, 2312 (1968).
cts on the Photohydration of Pyrimidines
y Wm. A. Summers, Jr., and John G. Burr* Chemistry Department, Oklahoma University, Norman, Oklahoma
78069 (Received February 26, 1972)
Publication costs assisted by the National Cancer Institute
A study of the effects of viscosity change on the photohydration quantum yield of uracil and dimethyluracil has been carried out to help evaluate possible mechanisms for these reactions. The results indicate that the rate of photoaddition of water to the 5,6 double bond of these pyrimidines changes only slightly with increasing viscosity and thus is not diffusion controlled. I n light of known effects of pH, deuterium isotope effects, enhanced reactivity with stronger nucleophiles, and energy transfer data, these results suggest that the precursor for photohydration is longer lived than the r , ~ fluorescent * singlet.
Intrnduction The change in quantum yield with viscosity change for the photohydration of uracil and dimethyluracil was measured in the following two types of solutions: (1) aqueous solutions of polyethylene oxide (Union Carbide WSR-301 and WSR-205), and ( 2 ) aqueous solutions of glycerine. The former are characterized by large changes in viscosity and small changes in water activity, while the latter by small changes in viscosity and large changes in water activity. These studies were performed at a uniform temperature of 16'. The relationsKip of diffusion constant to viscosity was measured for each type of solution,
(Eastman, spectroquality) and glycerine (Baker) were used without further purification. Uracil, dimethyluracil, and 2-hydroxypyrimidine were obtained from Cyclo Chemical Co. The polymers were a gift from Union Carbide Gorp. Viscosities were measured with an Ostx-aXd viscosimeter. Light intensities were determined with the ferrioxalate' and dimethyluraci12a3 actinometers. The light source was an array of six symmetrically arranged GE 15-W germicidal lamps. A jacketed cell (1) C . 6. Hatchard and C. A. Parker, Proc. Roy. Soc., Ser. A , 235, 518 (1956). (2) J. G. Burr, B. R. Gordon, and E. K . Park, Photochem. Photobiol., 8 , 73 (1968); c j . the discussion in ref 3 regarding this actinometer.
Water mas doubly distilled from glass; acetonitrilc
(3) J. G. Burr, E. €1. Park, and A . Chsn, S. Amer. Chem Soc., 94, 5866 (1972).
The Journal of Phgsical Chemistry, VoE. 7 8 , A'o. 22, 1972
WM. A. SUMMERS, JR.,AND JOHN G. BURR
3135 holder (totally of Vycor) of special design served to mount the cuvet for irradiation, maintain constant temperature, and filter the light. The formation of hydrate was determined by the "heat recovered uracil" method2 and by chromatographic separation on cellulose using [2-14C]uracil in butanol-water at) 0". Chromatographic separation of photohydrate was only effective with the glycerinewater systems. A Sargent Model XVI polarograph was used t o determine id fclr 2-hydroxypyrimidine in both glycerine -water solutions and polymer-water solutions. Thcse polarographic reductions were carried out in borate buffercd (pH '7) solutions with loF3M 2-HP. The E,,, was observed to decrease with added polymer (-1.35 to -1.67 V) but no change was observed with glycerine, indicating some phenomenon occurring at the solvent-electrode intcrfacc which inhibits electron transfer. The diffusion current determined is related to the diffusion cot4Kcient) by the Ilkovic equation, ll = (idihnz""t""n(:) 2 . Tracer labeling was used to determine the diffusion coefficients of glucose. (standard) and uracil. The devicr used consisted of a thermostated 200-ml flask with a side arm for periodic sampling. The labeled material, in a solution of known viscosity, was allowed to diffuse into the large volume of solution of the same viscosity, v hich \\.as stirred mechanically, through a large porosity sint)rred glass disk. The change in concentration with timc was determined by liquid scintillation and related to the diffusion coefficient by Fick's sccond law dc/dt = d(D dc/dx)/dc.
k
Thc .Journal of Physical Chcmislry, Vol. '76,.To. 28, 1972
Y
I
I
Y
9
?(CW
IO
100
lp00
I0,000
Figure 1. Quantum yields, for photohydration us. viscosity in water-polymer solutions; 0 is normalized to unit water activity: C , uracil, pI-1 2; A , 1,3-dimethgluracil, pH 7; 0, uracil, ~ € 1 7 .
25
20
e
Results The quanturn yield normalized to unit water activity for photohydration of uracil and dimcthyluracil as a function of viscosity, in the polymer solutions, increases slightly (Figure 1). The polymer solutions were very dilute, 1.3% (w/w)for the lo4 cP solution, and the water activity was found to decrease by about 870 over the range. Quantum yields for solutions of water-polymer -acetonitrile (Figure 2 ) also revealed a slight increase at high viscosity. These data suggest that the photohydration process is relatively insensitive to viscosity changes. The slight increases in quantum yield at high viscosity will be discussed later. I'or the glyccrine-water mixturrs, the variation of the quantum yields (based on heat, recovery) as a function of viscosity are shown in Figurcs 3 and 4 for uracil and dimethyluracil, respectivdy. When data, based on the chromatographic separation of the uracil photohydrattx, are normalized t o unit water activity and plot t c d against viscosity (I'igure 5) the quantum yield is found to increaw about 3Oc/o. This increase implies greatcr efficiency in more viscous media. Thr usual S-shapcd curvm (Figures 3 and 4) are bp1ievc.d to result from another reaction of photoex-
T
n
0
rr
U
I .5
I.o
3 /
SkP)
IO
100
1.000
IODOO
Figure 2. The quantum yields, 0,for photohydration of uracil viscosity in mixtures of water, acetonitrile, and polymer WSIl301; pII = 7: 0, water and polymer only; A , water-acetonitrile ( I : I, v/v) and polymer; water -acetonitrile (1:3,v/v) and polymer (lowest line).
us.
cited pyrimidine with glycerine. This product might) ,~ is also heat be similar to the methanol a d d ~ c twhich labile. A second spot was observed on the tlc plates using the labeled uracil but it could not be separated from unrcactcd uracil. It was felt that this might be the glycerine-watm adduct. (4) S. Y . Warig, Nature (London), 184, 184 (1959).
values compared to calculated values using the StokesEinstein equation, D = RTICSrrNq. The data indicate that polymer-water solutions are not classical with respect to the Stokes-Einstein equation but glycerine-water solutions are. I n any case a change in diffusion coefficient does occur with polymer-water solutions which should be reflected by %L decrease in the reaction rate, if the reaction were diffusion controlled. Little change in quantum yield was observed for these systems and a definite increase was observed in the glycerine mixtures. 1.0
2.C
3.0
4.0
5.0
6.0
7.0
8.0
?pP)
Table I : Polarographic Determination of Diffusion Coefficient for Uracil Based on the Ilkovic Equation
Figure 3. Quantum yield for photohydration (uracil) us. viscosity, @ based on heat reversible product.
Solvent
Water Glycerinewater 0.2% WSR-301 water 0.6y0 WSR-301 water
Temp,
Viscosity,
O C
CP
16 16
Dobsda
Doaloda
QC
1.19 3.1
6.2 1.85
5.8 1.8
1.0
16
7.0
4.9
0.97
1.56
16
122.0
2.25
0.06
1.57
1.5
a All values multiplied by om2sec-’. Based on StokesEinstein equation, all values mukiplied by cm‘ see-’. c Quantum yield “observed” for photohydration of uracil; all values multiplied by 10‘.
2 .o
1.0
4.0
3.0
Discussion
.? (CP)
Figure 4. Normalized quantum yield of photohydration (DMU) us. viscosity, 0 based on heat reversible product.
Solvent:
glycerine-water
I
1.0
2.0
3.”
7j(cP)
4.0
5.0
Figure 5. Quantum yield for photohydration (uracil) (normalized to unit water activity) us. viscosity, @ based on chromatographic separation.
The data presented in this paper lead to the conclusion that the photohydration of uracil and dimethyluracil is not a diffusion-controlled process. The reaction must be either very fast (a rearrangement) or slower than diffusion controlled (not occurring at every collision). The observed effects of pH on the photohydration rate,3 deuterium isotope eff enhanced reactivity with stronger nucleophiles,71s and the kinetic dependence of photohydration rate on water eoncentration,G all suggest that the excited state precursor is long enough lived to become “aware” of its molecular environment in solution. This would correspond to a “slow” bimolecular process, indicated as step 4 in Scheme I. The possibility of a very fast (rearrangement) reaction is suggested by the lifetime see) of the uracil (and thymine) fluorescent ( r , r * ) singlet, estimated from observation of the very faint fluorescence (a = to of these substance^.^ Such a very fast (6) (a) W. R. Moore, Advan. Polgmer Sci., in press; (b) J. G. Burr, B. R. Gordon, and E. H. Park, Advan. Chem. Ser., No. 81, 418 (1968).
I n order to clarify the relation of viscosity to diffusion coefficient in these solvent systcms, a determination of thc diffusion coefficicnts in each type of solution was carried out. Tables I and I1 list the observed
(6) (a) J. G. Burr, Advan. Photochem., in press: (b) J. G . Burr and E. H. Park, Advan. C h m . Ser., No.81,421 (1968). (7) A. M. Moore, Can. J. Chem., 36, 281 (1958). (8) J. G. Burr, R. L. Letsinger, W. A. Summers, and C. Enwall,
manuscript in preparation. (9) M. Daniels and W. Hauswirth, Sciewe, 171, 675 (1971).
The Journal o j Phgsical chemistry, Val. Y 6 , N o . M ,19Yd
WM.A. SUMMERS, JR.,AND JOHN G. BURR
3148 Table I1 : Direct Determination of Diffusion Coefficient Using Tracer Labeling Temp,
Viscosity,
OC
CP
Dobsd
Water
16
1.19
0.2% WSR-301 water
16
7.0
6.73a 10.7 4.70 7.50
0.6% WSR-301 water
16
0.4% WSR-205 water 0.9ycWSE-205 water
16
Solvent
122
0.3 1.77
4.7 13.5
16
Standard for apparatus, 2.2 X losa M glucose in water, D
9.0 2.2 =
Scheme I
I.
T,T*
singlet
11, n , n * singlet ( ? )
reaction might, be the rearrangement of an excited state water-pyrimidine complex, shown by the reactions in Scheme IT. Work is currently underway in our Scheme I1
HH
laboratory to identify such a complex, measure its format>ion constants, and evaluate its contribution to uracil photochemistry. We do have evidence for deuterium exchange between the H on N-3 of l-cyclohexyhracil and DzO at low concentrations in carbon tetrachloride; the thermodynamics of such exchanges aire being determined.’O Tn view of the first-order The Journal of Phuaical Chemistry,Vol. ‘76, hro. 22, 1972
6.73 X
Solute
[“C] Glucose [ W]Uracil [14C]Glucose [ 14C]Uracil. [‘*C]Glucose [ W ]Uracil [14C]Uracil [ W ]Uracil
Doalcd
5.8 0.97
0.06 1.4 0.5
om2 sec-l.
dependence of uracil photohydration yield with water activity (in acetonitrile solution),6 the kinetic importance of such a complex in the photohydration reaction may be limited. There are several indications that, the fluorescent singlet state and the excited state precursor to the photohydrate may not be the same state. Whitten and coworkers no correlation between the fluorescence quenching of DMU by water axid the photohydration rate of DMU. We have reported13 that europium(II1) ion emission can be sensitized by uridine monophosphate, but that, the photohydration of this nucleotide is not quenched by europium ion. I n conjunction with the other characteristics of uracil photohydration, we feel that the relative independencc of photohydration yield on viscosity provides justification for suggesting that the excited state precursor of the photohydrate of uracil and DhlU is a statc which is longer lived than the fluorescent singlet, and which is not the triplet state (since that state is known to lead only to photodimer). This state could be an n,a* singlet, using one of the oxygen lone pair electrons (a “hidden” n,T* ~ i n g l e t ) . ’ ~ ,The ’ ~ lone pair electrons of the nitrogens arc in a orbitals perpendicular to the plane of the ring, making these diketopyrimidines eight-electron, antiaromatic systems. and are thus not “n” electrons. Another possibility could be a tautomeric form of the fluorescent singlct formed in sort of a nonvertical excitation. The n , r * singlet would likely be populated by intersystem crossing from the fluorescent state and would be longcr lived. A precedent for this suggestion is the observation that the excited &ate of azaanthracene which is photore(10) W. A. Summers, J. G. Burr, and S. D. Christian, manuscript in preparation. (11) 0. G. Whitten, J. W. Happ, G. L. B. Carlsen and M.T. McCall, J . Amer. Chem. Soc., 92, 3499 (1970). (12) D. G. Whitten and Y. J. Lee, ibid., 93, 961 (1971). (13) J. G. Burr and A . Sarpotdar, 27th Southwest Regional Meeting of the American Chemical Society, San Antonio, Tex, Dec 1-3, 1971. (14) D. G. Whitten and Y . J. Lee, J. Amer. Chem. Soc., 92, 415 (1970). (15) Y. J. Lee, D. G. Whitten, and L. Pedersen, ibid., 93, 6330 (1971).
3141
. ~ ~ o T ~ I NPOLYMERIZATION I T ~ ~ ~ E ~ OF ALKYLMETHACRYLATES
duced by hydrogen abstraction from the solvent is not the ?T,?T* fluorescent singlet but a lower lying, n , r * 6nglet state;l? LPC,el al., have summarized the importance of ?i,?r* singlet states in N hetero~ycles.l~*'~ Formation of this other longer-lived singlet is shown as step 3 in Scheme I, a step which competes with fluorescence (step 21, formation of the triplet (@ = and with radiationless decay processes which are the predominant reactions depopulating the excited state. Only about 0.3% of the excited molecules are engaged in step 3 and 0.01% of them in step 2 so that steps 3 and 2 are not ~ o ~ ~ p e ~(for i ~ ai limited ve supply of the iiluorescent singlet). In conclusion, the slight increase in photohydration quantum yield x%b increasing viscosity may be construed as increasing the probability that the photohydration process is a relatively slow bimolecular re-
action, since such an increase could be the result, of a solvent cage which increases in effectiveness with increasing viscosity.
Acknowledgments. The authors are indebted to Dr. S. D. Christian for stimulating discussion during the preparation of this manuscript. Also, the gift of polyethylene oxide, WSR-301 and ~ S R - 2 0 from ~, Union Carbide Corp. was greatly appreciated. This work was supported by Grant CA 11418 from the National Cancer Institute, whose assistancc is gratefully acknowdedged. (16) A referee has suggested that this photohydrate precursor might be a ground state rearrangement product. Such a possibility is not excluded by the available facts but we have difficulty in visualizing a suitable candidate, at least for DMU, since the enol is ihe most likely candidate.
Electron Paramagnetic Resonance Spectroscopic Study of the Photoinitiate Polymerization of Alkyl Methacrylates1
.Smith" and R. D. Stevens Paul ill. Gross Chemical Laboratory, Department of Chemistry, Duke University, Durham, North Carolina 67706 (Xeceived April i d , 1972) Puhlicntwn costs asehted by the National Science Foundation
Within a continuous-flow system, dilute oxygen-free solutions of 2,Y-azobis [isobutyronitrile], RN : NR, have been irradiated with near-ultraviolet mercury radiation, this occurring as the liquid passed through the cavity of an X-band electron paramagnetic resonance, epr, spectrometer. By this means transient radicals have been observed. The solvents used were toluene, methyl methacrylate, ethyl methacrylate, isobutyl methacrylate, n-butyl methacrylate, and separate mixtures of toluene with each of these esters. The flow rate was varied 30-fold and the temperature from 25 to 39' with little appreciable effect, except as noted. With toluene as the solvent, the only epr signal observed was that from the R. radical. As the toluene mas replaced by an increasing proportion of methyl methacrylate keeping the other reaction conditions unchanged, the signal from the R . radical fell in intensity arid a new absorption took its place, growing progressively more intense so that when the solvent was undiluted methyl methacrylate there was only a small signal from the R. radicaI. The new absorption is consistent with addition radicals of structure R- (-CHz-C(COOC13,)CHa--)&Hz- C(COOCHs)CH3 with the size of n unknown. Kinetic arguments are given which suggest that the average of n may be a t least about ten. The results obtained with the use of the other alkyl. methacrylates were very similar to those found for methyl methacrylate and are interpretable in a like fashion.
Introduction There have been many2-20 electron paramagnetic resonance (epr> studies of liquid systems containing short-lived free radicals of the sort which occur in
n-ith the use of a thermal-redos reaction, and others have relied on the photochemical generation of primary (1) This work was supported by National Science Foundation Grants No. GP-7534 and GP-17579.
The J o w n a l of Physical Chemistry, Vol. 76, N o . 22, 1972
