S. CHABEREK, R. J. ALLEN,AND G. GOLDBERG
2834
Dye-Sensitized Photopolymerization Processes.la 111. The Photoreducing Activity of Some Dicarbonyl Compounds
by S. Chaberek,IbR. J. Allen, and G. Goldberg Technical Operations Research, Burlington, Massachusetts (Received December 7 , 1964)
0-Diketones such as 2,4-pentanedione and dimedon were found to be photoinitiators in conjunction with thionine and methylene blue for the photopolymerization of acrylamide. Studies into the relation between photoactivity and molecular structure showed that the active compounds are those capable of undergoing keto-enol tautomerism and that the enol forms are responsible for reactivity. Photoactivity was also observed for a variety of metal chelates of 2,4-pentanedione. On the basis of limited studies on Cu(I1)- and Al(III)-Pdiketone chelates, it appears that the photoinitiating activity must be at least partly due to the hydrolysis of the complex to produce photoactive ligand. Studies of the Cu(I1)dimedon system have shown an accelerating effect on the photopolymerization process.
Introduction A variety of dye-sensitized photopolymerization processes have been reported in which the polymerization-initiating free radical is produced by an oxidationreduction reaction between the light-excited dye and a mild reducing agent.2-9 Substances used as reducing agents include secondary and tertiary amines and amino acids, thiourea and its derivatives, ascorbic acid, and thiocyanate ion. During the course of our investigation into dye-sensitized photopolymerization processes we have found that many dicarbonyl compounds, especially 0-diketones, reduced light-excited thionine (and methylene blue) in the presence of monomers and brought about polymerization. To our knowledge there has been only one report on the photoreducing action of dicarbonyl compounds in the visible light range, and none on photopolymerization. Mauzeralllo described the photoreduction of some porphyrins in the presence of ethyl acetoacetate. This paper summarizes, then, some of our results on the photopolymerization of acrylamide initiated by free radicals formed by reaction of light-excited thionine and various dicarbonyl compounds under anaerobic conditions. Experimental Work Materials. A pure sample of acrylamide was obtained from the American Cyanamid Co. Preliminary screening of this sample for anaerobic thionine-sensiThe Journal of Physical Chemistry
tized polymerization showed no polymer formation in the absence of added activators and only a trace of dye bleaching over a period of 10 min. Thionine was purified by three recrystallizations from water. The absorption coefficient was found to be 5.8 X lo41. cm.-l mole-1 at 5980 8. Biacetyl, 2,5-hexanedione, 2,4-pentanedione (acetyl(diacetone), and 5,5-dimethyl-l,3-~yclohexanedione medon) were obtained from the Aldrich Chemical Co. The metal chelates of 2,4-pentanedione were obtained from MacKenzie Chemical Works, Inc. (1) (a) This study was performed under Contract No. AF33(657)-8754 and AF33(647)-11553, Photographic Branch, Reconnaissance Division, Air Force Avionics Laboratory, Wright-Patterson Air Force Base, Ohio; J. R. Pecqueux, project engineer; (b) t o whom inquiries should be sent a t Cowles Chemical Co., Skaneateles Falls, N. Y . (2) (a) G. Oster, Phot. Eng., 4, 173 (1953); (b) G. Oster, Nature, 173, 300 (1954). (3) G. K. Oster, G. Oster, and G. Prati, J . Am. Chem. Soc., 79,595 (1957). (4) N. Uri, ibid., 74, 5808 (1952). ( 5 ) G. Delzenne, W. Dewinter, S.Toppet, and G. Smets, J . Polymer Sci., ZA, 1069 (1964). (6) A. Watanabe and M. Koizumi, Bull. Chem. SOC.Japan, 34,1086 (1961). (7) G.Delzenne, S.Toppet, and G. Smets, J . Polymer Sci., 48, 347 (1960). ( 8 ) S. Chaberek, A. Shepp, and R. J. Allen, J . Phys. Chem., 69, 641 (1965). (9) S. Chaberek and R. J. Allen, ibid., 69, 647 (1965). (10) D. Mauzerall, J. Am. Chem. Soc., 82, 1832 (1960).
DYE-SENSITIZED PHOTOPOLYMERIZATION PROCESSES
2835
THIONINE=I x I O - ~ M pH = 8 . 5 5 - A A = ~ X IO-~M 16% ---AAa4X
17
I
/
v)
y
6.00
h Ql I 2 0 0
r
5.0%MONOMER y 5 . O X M O N O M E R
2 0.IO w>
5.00 4.00
'5.0%
i
e 0.08
MONOMER
W
5
3.00 2 .oo
I .oo
3 0.06
MONOMER= 2.0,3.5,
c 0.5
2
0.04 IDENTICAL BLEACHOUT FOR ALL MONOMER LEVELS
1.0
1.5 2.0 2.5 3.0 TIME, MINUTES
3.5
0.02 ~~
4.0
1
2
3 4 5 TI M E, M IN UTES
6
7
8
Figure 1. Effect of monomer concentration on the anaerobic thionine-AA-acrylamide system.
Triacetylmethane and the 3,3-dimethyl-, 3-isowere prepared nitroso-, and 3-amin0-2~4-pentanediones by methods described in the literat~re.l'-'~ 3-Diethylamin0-2~4-pentanedione was prepared by the and diethylamine, reaction of 3-chlor0-2~4-pentanedione in the following manner. A mixture of 27 g. of 3chloro-2,4-pentanedione and 60 ml. of diethylamine was refluxed for 17 hr. The reaction mixture was poured into 300 ml. of diethyl ether, and the precipitated diethylamine hydrochloride was filtered. The ether solution was evaporated under reduced pressure, leaving a dark that on This crude product was recrystallized from ethyl acetatepetrol and final purification W&S made by Vacuum sublbation. The purified product was a white solid having a point of 84,4370. ~ ~ C&d. ~ forl c ~ H ~ , N c, ~ ~63.1; : H, 10.0; N, 8.2. Found: c, 63.2; H, 9.9; N, 8.1. Bis(acety1acetone)ethylenediimine was prepared by the method of McCarthy, Hovey, Ueno, and Martell.16 procedure* The ex??erimentd and Procedures used in this study were the Same as those described previously.16 The light from a 500-w. projection lamp was passed through a Kodak Wratten 23A filter the reaction Glter prior to cuts out radiation having wave lengths lower than 5600
A.,
transmits 11% at 5700 A. and 82.7% at 6000 A. The rate of dye bleaching, Rr, was calculated directly from spectrophotometric measurements. The polymerization rate, R,, was determined by taking aliquots of the reaction solution at several time intervals, precipitating the polyacrylamide in methanol, filtering it, and drying the residue to constant weight. Polymer molecular weights were determined viscometrically.
Results and Discussion 2,4-Pentanedione ( A A ) . 2,4-Pentanedione was the
first substance for which photoreducing activity was found. Figure the effectsOf monomer and AA concentrations on the rates of photobleaching and polymerization. The solid lines show the monomer variation , at an AA level of 8 X 10" M , while the dotted lines show the effect of a lower AA concentra-
(11) J. u. Nefg Ann., 277971 (1893). (12) R. G . Pearson, E.A. Mwerle, and J. M. M a s , J . Am. C h a . SOC.,7 3 , 927 (1951). (13) A. Wolff, Ann., 325, 139 (1902). (14) W. D. Cash, F. T. Semeniuk, and W. H. Hadung, J. Org. cha.,21,999 (1956). (15) P. J. McCafihy, R. J. Hovey, K. Ueno, and A. E. Msrtell, J . Am. C h a . SOC.,7 7 , 5820 (1955). (16) A. Shepp, S. Chaberek, and R. MacNeil, J. phy8. Chem., 66, 2563 (1962).
Volume 69, Number 9 September 1066
S. CHABEREK, R. J. ALLEN,AND G. GOLDBERQ
2836
tion in the presence of 5% acrylamide. These data indicate both a similarity and a difference in the behavior of this activator when compared with the amineactivated systems represented by triethanolamine (TEA).g As with the TEA-containingsystems, both the rates of dye bleaching and polymerization increase with an increase in activator level. Also, an increase in acrylamide concentration results in an increase in the polymerization rate for AA and TEA systems. However, the dye-bleaching rate dependences for the systems differ in two respects. First, in the absence of monomer, no significant thionine bleaching occurs in the presence of AA; in the amine systems, extensive reduction of the dye is obtained in the absence of monomer. Second, although these rates decrease with an increase in monomer level in the amine systems, Figure 1 shows that for the AA system the rate is independent of monomer levels between 2 and 7.5%. In addition, both fading and polymerization rates are pH sensitive. Figure 2 shows the variation in Rf and R, as a function of solution pH. The shapes of the curves are strikingly similar to those obtained with TEA. Both rates increase from pH 7 to maxima at 8.4(for R P )and 8.8 (for Rf), after which the values decrease rapidly. The strong pH dependence of this P-diketone initiator in the pH interval of 7 to the pH of maximum activity must be related t o the acidic properties of the enol form of AA arising from keto-enol tautomerism. It appears that structural changes in the molecule that decrease its tendency toward tautomerism result in a loss of photoinitiating activity. Thus, biacetyl (I) and 2,5-hexanedione (11) are inactive as photopolymerization initiators. In these
0 0
0
0
/I /I
II
II
CH&--C--CH8
CH+34H&H&-CHa
I1
I
cases, both the shortening and the lengthening of the carbon chain between the carbonyl groups shifts the keto-enol equilibrium far toward the keto form. To obtain more information concerning the relation between tautomerism and photoactivity, studies were extended to dimedon (DM) and some 2-dkyl (acyl) 2,4-pentanediones. Dimedon. Interest in this substance stemmed from its high degree of enolization. Schwarzenbach and Felder" showed that in dilute aqueous solutions AA exists in the enol form to the extent of about 15.5%, whereas DM is 95.3% enolized. Measurements of Rr and R, for the thionineDM-acrylamide combination showed this system to be similar qualitatively to the corresponding AA-containing one in that (1) Rf and R, The Journal of Ph&al
Chemistry
THIONINE- I X 1 6 5 M ACRYLAMIDE'0.704 M ( 5 % ) DIMEDON=sX I 6 ' M ACETYLACETONE; exio+u IO
w
-?
I
I
I
I 6
I
I
I
I 8
9
7
I
W
0
I
P
2 X P
U
/
I
0
4
5
7
I 10
OH
Figure 2. Effect of pH on the thioninedimedon- and thionine-AA-acrylamide system.
increase with an increase in the activator concentre tion; (2) Rr and R, are sensitive to solution pH; (3) Rf and R, are essentially independent of the dye concentration in the range 0.5 to 2 X 10" M; (4) Rf and R, increase with an increase in monomer concentration in the range 1 to 7.5%. The most striking aspect of the DM system, however, is its susceptibility to solution pH. A comparison of this system with the AA-containing one is shown in Figure 2. It is seen that the rates of thionine bleaching for both systems have a maximum value, but the pH at which it occurs varies markedly. For DM, it occurs at a pH of about 5.0 to 5.5; for AA, it occurs at a pH of (17) G. Schwarzenbach and E. Felder, Helv. Chim. Acta, 27, 1701 (1944).
DYESENSITIZED PHOTOPOLYMERIZATION PROCESSES
2837
Table I : Variation of Enolate Concentratiom for Acetylacetme and Dimedon with pH 7
Compd.
4.0
AA
0.00
DM
5.3
4.5
0.00
15.0
5% enolate at pH 7.0
6.0
6.5
6.0
0.01 35.3
0.04 62.0
0.11 81.5
about 9. The polymerization data show clearly the greater range of activity for the DM system. The rate of polymerization increases in the pH range of about 3 to 6 , remains essentially constant at pH values of 6 to about 8.5, and finally decreases at higher pH values. In contrast, for AA, polymerization is restricted to a pH range starting at about 7, with maximum activity observed at a pH of about 9. In the pH range between 8 and 9, AA is a superior initiator; that is, higher polymer yields are obtained with this system. However, its efficiency rapidly falls off with increasing solution acidity, so that at pH 7.5, both systems have comparable activity, but at pH 7, AA is essentially inactive while DM shows its maximum activity. These data on the pH dependence of the thionineDM system indicate that the enolate ion may be the most reactive form. On the basis of this assumption, let us consider briefly the data shown in Figure 2. Schwarzenbach and Felderl' have measured the PKE and KT values for AA and DM for the reactions KT
+ [enolate-] [enol] 1_ [H+] + [enolate-]
[keto]
[enol]
1.1 93.7
7.6
8.0
8.6
9.0
10.0
3.0 94.8
6.6 95.1
10.9 95.2
13.7 95.3
15.3 95.3
and a constant rate at, values exceeding this range. The experimental data of Figure 2 shows that these conclusions account for part of the observed results; they confirm the predicted increase in Rf and R, in the pH range where the enolate ion concentration is increasing to its maximum value. However, the postulations do not account for the subsequent decrease in these rates at the higher pH levels where the enolate ion concentrations remain essentially constant. In the thionine-AA combination, the .decrease in activity at high pH values may be correlated with changes in the light-absorbing properties of the dye. This cannot be the reason for the decrease in DM activity since it occurs at much lower pH values. Further work is required to determine the reason for this falloff. 3-Alkyl (Acyl) Substituted Acetylacetones. Table I1 summarizes photobleaching and photopolymerization data on systems containing 3-methyl-2,4-pentanedione (MAA), 3,3dimethyl-2,4-pentanedione(MMAA), and triacetylmethane (TAM). Reference data for AA are included.
(4
KE
(b>
Using the following values for KT and KE, we have calculated the per cent enolate ion existing at various pH values for these activators. The values of KT and KE for AA and DM are AA DM
KT
PKE
0.184 20.3
8.13 5.23
These data are listed in Table I. For AA, essentially no enolate is present at pH values lower than 7; a rapid increase in its concentration occurs in the pH interval 7 to 10. We would expect, therefore, that no photopolymerization would occur at pH values less than 7 and that an increase in pH in the 7 to 10 range would bring about increasing rates of polymer formation that would level off at values around 10. For DM, appreciable enolate ion is present at pH 4; its concentration increases in the range 4 to 8 and becomes essentially constant at values higher than 8. Thus, we would expect photopolymerization to start at pH 4,with increasing rates in the range 4 to 8,
Table II : Anaerobic Photopolymerization of Acrylamide
in the Presence of Some 3-Substituted Acetylacetones" Enolate, Compd.
KT
+
Enol enolate,
PKE Bf X 101 R, X 10' MI X 106 M
AA 0.184 8.13 MAA 0.029 9.50 MMAA ... ... TAM ... ...
1.21 0.20 None 0.76
9.2 2.0 None 5.4
" All systems contain 1 0 4 M thionine, 8 0.704 M acrylamide at pH 8.55.
9.0 0.23 None
...
x
X 106
1.243 0.22 None
...
10-6 M activator,
From the data of Table 11,some interesting comparisons among the four systems can be made. First, progressive substitution of methyl groups for the hydrogen on the methylene group rapidly decreases the enolizability of the resulting pdiketone. The substitution of one CHa group decreases the degree of enolization by a factor of about 6, while complete substitution destroys enolization. Both the Rt and R, values for these compounds parallel the decrease in the KT values, the Volunte 69,Number 9 September 1966
S. CHABEREK, R. J. ALLEN,AND G. GOLDBERG
2838
completely substituted compound being inactive. Second, consider the relative amounts of total enol (enol enolate) between AA and MAA. This amount is about 5.6 times greater for AA than for MAA. Therefore, if photoactivity depends upon the enol forms and all other factors are equal, we can expect the relative Rf and R, magnitudes for AA to be 5.6 times greater than for the M U . The experimental values of Table I1 show that Rfis six times greater and R,,4.6 times greater-an excellent agreement with the value of 5.6. Thus, these data confirm further that photoactivity resides in the enol forms and that the keto tautomer is not reactive. Substantial activity was also observed with the triketone, TAM. Unfortunately, enolization data are unavailable for this substance, and therefore it is not possible to correlate activity with enol content. However, its appreciable efficiency indicates that substitution in the 3-position need not destroy the photoactivity of pdilretone derivatives. Other Substituted p-Dilcetones. In addition to the 3alkyl-substituted dilretones, a number of 3-substituted nitrogen-containing derivatives were checked for photopolymerization activity. The screening was done qualitatively, with thionine and calcium acrylate. The criterion for activity was the onset of dye bleaching and formation of the water-insoluble calcium polyacrylate. 3-Amino-2,4-pentanedione was found to be a sufficiently strong reducing agent to bleach thionine in the dark. The corresponding 3-isonitroso deriv& tive was inactive. However, 3diethylamin0-2~4-pentanedione showed some activity although it was inferior to AA. In addition to the above compounds, bis(acetylacetone)ethylenediimine (SB) was screened for activity. This substance has the structure
Table III: R, Values for SB, TEA, and AA
+
I
N
0
II
II
C
C
/\ /\ CHa
Of keto-eno1 tautomerism. The with those Of polymerization rate for SB is
Compd.
R, X 104,Maeo.-1
TEA SB
3.05 7.04
AA
10.0
for metal chelates such as ferrioxalate and uranyl oxalate upon irradiation with ultraviolet or blue light, we are not aware of any reports for a similar red-lightsensitized process for nonbiological metal chelate systems. Table IV summarizes R, and dye bleaching data for nine metal acetylacetonate chelates.I8 Table IV : Anaerobic Photopolymerization of Acrylamide in the Presence of Some Metal Acetylacetonate Chelates" Approx.
Approx.
bleaching
bleaching
RP X 104,
time,
Compd.
M seo.-1
min.
Compd.
M mo.-1
min.
CU(AA)~ TiO(AA)t Cr(AA)s Co(AA)* VO(AA)2
5.6 8.9 5.4 None 9.4
None 3.5 None None 3.5
AI(AA)a Fe(AA)t Mn(AA)* Ni(AA)*
11.0 11.0
2.5 3.5 3.5 4.0
RP X 104,
time,
10.0 10.0
a All systems contain lo* M thionine, 10-4 M metal acetylacetonate, and 5% acrylamide at pH 8.6.
With the exception of the Co(II1) chelate, all the r e maining metal chelates showed polymerization activity. Not all activity was accompanied by bleaching of the dye, however. In the presence of the Cu(I1) and Cr(111) chelates, polymerization occurred without sign& cant dye bleaching, and the initial polymerization rates in the presence of these substances were approximately half those obtained with the other compounds. Since the polymerization rates of the remaining compounds were comparable, there appeared to be little correlation between this activity and the following: (1) stability of the metal chelates, since the magnitudes for this series vary widely; (2) visiblelight absorption by the metal chelates, since some are colored and some are colorless; (3) valence state of the coordinated metal ions, since some have only one state and others have
TEA and in 'I1' system contained thionine, x and 5% acvlamide at a pH of 8.55. The activity of the diimine is seen to be substantial. To obtain a better insight into the function of these &Diketone Initiator Systems Containing Metal Ions. of the be systems, we pmed a limited One of the surprising discoveries during this research program was the photopolymerization-initiating ac(18) In View of their low water solubilities, these chelatea were d b tivity of the thiOninemetal acetylacetonate combin* solved in a small amount of methanol and diluted to final volumes tions. Although such photoactivity has been reported with water. The Journal of Physkxd Chemistry
DYESENSITIZED PHOTOPOLYMERIZATION PROCESSES
THIONINE= IXlG5 M p H * 8.55
2839
-X-
6 x 1 0- 5 M C U ( A A l 2
----A----
4X l f j S M AI (An), - 4 M AA
--&-lI.2X10
MONOMER 3 5 %
10.00 9.00 8 .oo
% 7.00
X
I 6.00 V
t
8
5.00
W
z
4.00
P
f
3.00 2 .oo
.5
1.0
1.5 2.0 2.5 3.0 TIME, MINUTES
35
4.0
1
2
3
4
5
6
7
8
TIME, MINUTES
Figure 3. Comparison of acetylacetone, cupric acetylacetonate, and aluminum acetylacetonate at constant conditions.
havior of metal complexes of ,f?diketones as photopolymerization initiators. The Cu(I1)-AA and Al(II1)-AA systems were chosen for further study because their initiating properties were substantially different. In addition, the activity of the Al(II1) chelate was representative of the behavior of most of the metal chelates tested. A comparison of the activities of Cu(I1)-AA and Al(II1)-AA is shown in Figure 3. All systems contain the same amount of AA. (The molar concentr% tions of the chelates are of course different because more than one AA molecule is bound to the metal ions.) In the Cu(I1)--AA system, the addition of Cu(I1) ions to AA drastically changes the polymerization and dye-bleaching reactions. Dye bleaching is almost completely eliminated. In addition, the shapes of the polymerization rate curves for the two systems are different. Polymer formation for AA fa& off at longer times, whereas in the presence of this metal chelate, polymerization begins at a rate lower than that for AA but continues at about this same initial rate for longer times. In contrast to this behavior, the addition of Al(II1) alters neither tbe bleaching nor the polymerization rate. In this case, and for all systems comparable in activity to Al(I1I)-AA, the metal chelate is probably
extensively hydrolyzed; thus, we may be observing only AA activity. The validity of this conclusion is substantiated by the following considerations. As a rough approximation let us define the hydrolysis of the metal chelate by
+
+
MAs 30H- 3 M(OH)ai 3A (e) where MA3 denotes the metal chelate and A, the displaced ligand. It follows that
where K M Athe , chelate stability constant, and KsO,the solubility product, are defined by the equations
Kh = [M][OH-]*
(3) Let us now estimate the per cent dissociation for the alu”(II1) and iron(II1) acetylacetonate chelates in solutions containing lo-‘ M MA8 at a solution pH of 8.55 using the following values for K M A and ’~ K S , , . ~ (19) J. Bjerrum, G. Schwarzenbach, and L. G. Sillbn, “Stability Constants, Part 1, Organic Ligands,”The Chemical Society, London,1957. (20) J. Bjerrum, “Stability Constants, Part 11,Inorganic Ligands,” The Chemical Society, London, 1958.
Volume 68,Number 9 September 1966
S. CHABEREK, R. J. ALLEN,AND G. GOLDBERG
2840
Using the approximation method we find that the Al(II1) chelate is about 90% hydrolyzed, while the Fe(II1) chelate is about 97% hydrolyzed. InsufEcient KMa and Ks,, data prevent our testing other metal chelates listed in Table IV in this way. In view of the difference in the behavior of Cu(I1)AA, it was considered desirable to explore further the effect of metal chelate stability on photoinitiating activity. Two substances were chosen for this purpose: SB and DM. Since the Cu(II)-SB chelate is substantially more stable than Cu(I1)-AA, its degree of hydrolysis would be less under the experimental conditions employed and consequently less free ligand would be available. In contrast, DM cannot form stable Cu(I1) chelates because of steric hindrance. Consequently, in this case, solutions would contain the uncoordinated ligand and Cu(I1) ions. Figure 4 summarizes bleaching and polymerization data for AA, SB, and DM in the absence and in the presence of the Cu(11) ion. Consider first the dye bleaching reaction. In the presence of Cu(II), dye bleaching is drastically
decreased compared to the systems containing the ligands alone. In fact, for SB and AA, dye fading is almost completely eliminated. Several differences are apparent between the ligand- and Cu(I1)-containing plots. First, consider the shapes of the curves. In the absence of the Cu(I1) ion, the plots for all three ligands start at a high rate of conversion, and this rate decreases with time, so that at times greater than about 4 min. the percentage of polymer formed remains essentially constant. In the presence of Cu(II), the polymer formation proceeds at essentially the initial rate for time intervals exceeding those at which the conversion falls off in the absence of the metal ion. Second, consider the initial polymerization rates between the corresponding ligand- and Cu(I1)-containing curves. For SB and AA, the initial polymerization rates of the Cu(I1)-containing systems are less than those for the ligands alone, and the difference is greater for SB. If, as it is for Al(II1)-AA, photoactivity is the result of the free ligand concentration, these data show that the polymerization efficiency decreases with an increase in the stability of the Cu(I1) chelates. Here, the concentration of free ligand should be substantially less in the presence of AA, and we can expect a greater decrease in the photoactivity of the Cu(I1)-SB combination.
-4
1.6 X IOP4M SE ATpH 8.55 --o 1.6 X IO M AA AT pH 8.55 -X-SXIO-~M CU(AA),AT p~ 8.55; 1 . 6 xI ~ ~ C U - SAT B pH 8 . 5 5 ----A---I.6X10-4M DIMEDON AT pH 7.0 --.-+-.-8X10-5M C u " t I.6X164M DIMEDON AT pH 7.0
-0-
-
.5
1.0
1.5 2.0 2 5 TIME, MINUTES
3.0 3.5
4.0
1
2
THIONINE; I x I O ~ M MO N OM E R = 5 %
3
4
5
6
7
8
TIME,MINUTES
Figure 4. Comparison of acetylacetone and dimedon and diacetylacetoneethylenediiminein the presence of Cu(I1).
The Journal of Physical Chemistry
9
IO
DYE-SENSITIZED PHOTOPOLYMERIZATION PRQCESSES
However, the Cu(I1)-DM data of Figure 4 show conclusively that these differences cannot be ascribed solely to a reactant concentration factor. In this case Cu(I1) functions as a polymerization accelerator. Contrary to its behavior with AA and SB, the metal ion increases the initial polymerization rate relative to DM. This increase is appreciable; in 30 sec., the Cu(I1)-DM systems produce 6y0 polyacrylamide, compared to 4.2% for DM alone. One may indeed inquire whether Cu(I1) ions or cupric acetylacetonate are photoactive and initiate photopolymerization. Cupric ions have a weak though broad absorption band $ the visible range with a maximum at about 7000 A. However, experiments on solutions containing 8 X M Cu(I1) and 5% acrylamide at pH 8.55 show that no detectable polymer was formed during periods of 12 min. Similarly, the possibility of a thionine-sensitized reaction involving Cu(I1) but no pdiketone was ruled out by similar experiments. Only a trace to 1% polymer was obtained in 10 min. This amount of polymer is obtained by irradiating anaerobic solutions containing only thionine and acrylamide, and possibly arises from the presence of trace impurities in the reagents. Finally, although copper(I1) acetylacetonate has an absorption band at about 8200 8.,irradiation of solutions containing only the chelate and acrylamide gave no polymer in 18 min. A detailed explanation of reaction mechanism for these Cu(I1)-containing systems must await the results of further studies now in progress. However, even these limited data allow some speculation regarding their mode of action. We believe that the observed effects of Cu(I1) are a result of its oxidizing action on the reduced forms of thionine, semithionine (ST), and leucothionine (LT) acccording to the reactions
+ CU'+ 4*ST+ CU+
(d)
.ST+ CU'+ 4T + CU+
(e>
LT
Oster and c o - ~ o r k e r s ~have ~ - ~ shown ~ that various leuco dyes (including leucothionine and leucomethylene blue) produced by reacting the light-excited dyes with secondary and tertiary amines and amino acids are sufficiently powerful reducing agents to reduce a variety of metal ions including Cu(I1). The regeneration of thionine by reaction e accounts for the decreased rate of dye bleaching and also, at least in part, for the differences in the shapes of polymeriz+ tion curves. In the absence of Cu(II), the rapid dye bleaching depletes the concentration of thionine to levels too low to sustain radical generation, and the polymerization rate consequently levels off.
2841
In the presence of Cu(II), the dye concentration remains almost constant (at least for AA and SB), and consequently radical formation can continue as long as the solution is illuminated. However, the Cu(I1)DM data indicate that other factors must be involved. One of these probably involves a decrease in leucothionine concentration as a result of reaction f. Studies have shown that the leuco dye functions as a chain terminator--possibly through reaction with a growing polymer chain, M,., according to (e).
+
+
LT M,. 4MH *ST* (f) An increase in LT concentration results in lower rates of fade and of polymerization, m well as drastically reduced polymer molecular weights. Activity of Other Carbonyl Compounds. Our studies with fldiketones described above show that keto-enol tautomerism is an important prerequisite for dyesensitized photopolymerization activity. Hence, a variety of pdicarbonyl compounds were screened qualitatively for photopolymerization of calcium acrylate. Substantial polymerization was obtained with the following substances: disodium malonate, malonic ester, barbituric acid, 2-acetylcyclohexananone,acetoacetanilide, and 3-ketoglutaric acid. However, factors other than keto-enol tautomerism must be involved. Preliminary studies with 1,Zcyclohexanedione (111) and 2-hydroxy-1-methyl-1-cyclopenten-3-one (IV) showed that in presence of thionine and acrylamide, dye bleaching was obtained in the pH range of 8 to 9.5 without polymer formation. 0
III
0
Iv
Compound I11 is almost completely enolized under our experimental conditions, while compound IV is a stable enol. Thus, a better understanding of the nature of these photochemical processes must await the results of detailed studies now in progress.
Acknowledgments. The authors axe indebted to Mr. J. Panella for preparing the various 2,4-pentanedione derivatives and to Miss P. Rmke for conducting some of the screening experiments described above. (21) G. Oster and N. Wotherspoon, J. Am. C h . Soc., 79, 4836 (1957). (22) G.Oster and G . Oster, ibid., 81, 5543 (1959). (23) J. Joussot-Dubien and G . Oster, Bull. soc. chim. France, 343 (1960).
Volums 69,Number 0 September 1966