- Email: [email protected]
Study of postradiation graft polymerization of acrylic and methacrylic acids to the copolymer of tetrafluoroethylene with vinylidene fluoride
Polymer Science U.S.S.R. Vol. 32, No. 3, pp. 435--441, 1990 Printed in Great Britain.
0032-3950/90 $10.00 + .00 © 1991 Pergamon Press plc
STUDY OF POSTRADIATION GRAFT POLYMERIZATION OF ACRYLIC AND METHACRYLIC ACIDS TO THE COPOLYMER OF TETRAFLUOROETHYLENE WITH VINYLIDENE FLUORIDE* YE. A . RAGOL'SKAYA, S. G . PRUTCHENKO, S. A . PAVLOV a n d E . N. TELESHOV Karpov PhysicochemicalResearch Institute (Received 14 October 1988)
Postradiation graft polymerizationof acrylicand methacrylicacids from solution and the gas phase onto the copolymer of tetrafluoroethylene with vinylidenefluoride has been studied. The dependences of the degree of grafting on the thickness of the films, temperature and grafting time have been obtained. Scanning electron microscopyhas been used to study the distribution of the grafted polymer over the section of the samples. From the results some conclusionsare drawn on the aspects of the effect of diffusionfactors on the gross kinetics of the process. A significantdifference in the polymerizationrates of acrylicand methacrylic acids was established in identical experimental conditions. The fall observed in the polymerization rate of methacrylic acid is related to the possibilityof "occlusion" of the terminal growing radicals established by EST spectroscopy. THE BRISK development of membrane technology in the last few years has revived the interest of investigators in the processes of modifying polymeric film materials by radiation graft polymerization. Of considerable interest for the requirements of membrane technology are materials based on fluorine-containing polymers with high chemical stability and good physical and thermomechanical properties. Grafting polymeric acids to these materials helps to impart to them a whole gamut of practically useful properties such as hydrophilicity, ion-exchange properties, raised adhesion, etc. In this work we study the postradiation graft polymerization of acrylic (AA) and methacrylic ( M A A ) acids on the copolymer of tetrafluoroethylene and vinylidene fluoride ( C O P O L ) which is now finding wide application in the production of film materials, asymmetrical microfiltration membranes and hollow fibres. We used C O P O L containing according to elemental analysis 70% vinylidene fluoride and 30% tetrafluoroethylene units. Prior to use C O P O L was purified by reprecipitation from solution in acetone into ethanol or water. The films were produced by casting from 2.5% C O P O L solution in acetone onto a cellophane support followed by drying at room temperature in air and adjustment to constant weight in v a c u o at 50°C. The degree of crystallinity of the films thus obtained determined by radiographic methods was 45%. A A and M A A were purified by standard techniques [1]. Preliminary irradiation of the films was carried out in v a c u o in sealed ampoules with 6°Co y-radiation with K-300,000 unit (Karpov Physicochemical Research Institute) at room temperature. The total absorbed dose was 10-50 k G r at the dose rate 2.4-8.0 Gr/s.
* Vysokomol. soyed. A32: No. 3,495-501, 1990.
435
436
YEA, RAGOL'SKAYAet al.
Postradiation graft polymerization was conducted both from an aqueous solution of the monomers and the vapour phase. For polymerization from solution to suppress possible homopolymerization to the solution was added an inhibitor--Mohr salt in a concentration 8 x 10 -5. After arrest of the reaction the grafted films were washed with hot water and dried to constant weight in vacuo at 50°C. The yield of the graft polymer (degree of grafting) was determined gravimetrically
P-P0 x 100%, P0
Q= - -
where P0 and P are the mass of the film before and after the reaction. The ESR spectra were recorded with the PA-100 spectrometer at the temperature of liquid nitrogen on carrying out polymerization in solution and at room temperature in the case of vapour phase grafting. The distribution of the graft polymer over the section of the modified samples was studied by scanning electron microscopy (SEM) and X-ray microanalysis (XMA) with the Jeol ISM-35CF instrument and also optical microscopy. To employ the SEM and XMA methods the samples were contrasted by introducing by ion exchange Pb 2÷ ions into the grafted PAA amd PMAA. Typical kinetic curves of the polymerization of AA from aqueous solution on a film of COPOL are given in Fig. 1. It will be seen that the reaction proceeds at a rate of dropping in time (for practically constant concentration of monomer in solution), a characteristic of the curves being the presence of an initial portion with auto-acceleration. The time taken to reach the limiting polymer yield alim significantly depended on the thickness of the film d and usually amounted to 1-3 h. Both the initial rate* and Qlim rose with increase in the radiation dose D practically similarly with the accumulation in the initial C O P O L matrix of R ° radicals determined by EST spectroscopy. At doses - 5 0 kGr there was simultaneous limitation of the rise in the concentration of radicals [R'] and the initial rate and Qlim.
80 :
:L
J * o
6' J
#0
Jo
$o
.Jog
Time, rain FIG. 1. Kinetic curves of polymerization of A A from 40% aqueous solution on films of COPOL of thickness 5 (1), 10 (2), 20 (3), 40 (4), 60 (5) and 96/zm (6). Dose of preliminary radiation 10 kGr, 25°C.
Similar behaviour was also characteristic of the polymerization of MAA on COPOL films although with the difference that grafting of M A A proceeded at far lower initial rates and Qlim (Fig. 2). Study of the kinetics of the polymerization of A A on films of different thickness d showed that the * Since the initial portion of the kinetic curves is of an S-shaped character, by the initial rate we shall later understand the mean tangent of the angle of slope of the straight line drawn through the origin of coordinates and the point of inflection in the curve.
Postradiation graft polymerization
437
a,'/, 200
80
-i '~°
6
100
2 I Time, h FIG. 2
- .
,
5O
.
I00 ~m
Fio. 3
FI~. 2.
Kinetic curves of polymerization of AA (1) and MAA (2) from 40% aqueous solution on COPOL films 20/.~m thick. Dose of preliminary radiation 50 kGr, 25°C. FIG. 3. Dependence of polymer yield on film thickness d at different stages of the reaction on postradiation polymerization of A A on COPOL. Time of reaction 5 (1), 10 (2), 15 (3), 30 (4) and 60 min (5); (6) Qlim- Dose of preliminary irradiation 10 kGr; concentration of AA solution 40%; 25°C.
initial rate and Olim essentially depend on d (Fig. 3). It will be seen that a l i m r o s e monotonically with increase in d while Q at the initial stages of the reaction passed through a maximum. The position of the maximum shifted to the region of large thicknesses with increase in the reaction time. The initial rate also passed through a maximum but for somewhat lower d values. It should be noted that the dependence of the kinetic parameters of the reaction on the geometric dimensions of the sample is a serious argument in favour of diffusion control of the reaction rate. Another consequence of the diffusion nature of the process is the marked non-uniformity of the distribution of the grafted phase over the section of the sample. The corresponding distribution profiles obtained by the XMA method are given in Fig. 4. It will be seen that the distribution is of a practically rectangular character. Such a type of distribution is quite characteristic of the processes of graft polymerization of A A on different polymeric supports (see, for example, Refs [2--6]). A distribution of a similar type was obtained by us both for polymerization of AA from solution and from the gas phase. Figure 4 shows that as the process deepens the fronts of the grafted phase move to the centre of the film although the concentration of the polymer in the grafted regions rises insignificantly. The degree of grafting for which the fronts meet rises somewhat with increase in the concentration of the monomer in solution and the thickness of the film. After meeting of the fronts the polymerization rate, as a rule, falls sharply and the kinetic curve is translimited. It is also interesting to note that when polymerization was conducted from the gas phase greater non-uniformity of the distribution was observed than from solution (Fig. 4, curves 6, 7) and in some cases translimitation of the polymer yield in time occurred before the boundaries of the grafted phase met. It is important to note that the patterns observed experimentally in some respects do not agree with the conclusions of the elementary theory of diffusion-controlled graft polymerization [7, 8]. Thus, for example, from the theory it follows that the rate of the process not complicated by secondary factors such as non-isothermicity, dependence of the diffusion coefficient on concentration, etc. decreases in porportion to d 2. The limiting degree of polymerization naturally must not depend on d. The advent of strongly convex (practically rectangular) fronts of the distribution is also not a consequence of the elementary diffusion theory which, on the contrary, predicts the concave character of this distribution [7, 8].
438
YEA. RA6OL'SKAYAet al.
i
l Ol~.m
!
101~m
Fic. 4. Distribution profiles of grafted PAA over the section of the film obtained by the XMA method: (1-5) polymerization from 40% aqueous solution on films 55/~m thick; (6, 7) polymerization from the gas phase on films 20 p.m thick. Radiation doses: 10 (1-5) and 50 kGr (6, 7). Yield of grafted polymer 13 (1), 27 (2), 42 (3), 54 (4), 73 (5), 28 (6) and 55% (7); 25°C.
Thus, the experimentally observed anomalous dependences of the polymer yield and the initial polymerization rate on d in the first place must be linked with the dependence of the diffusion coefficient and also the boundary conditions of the diffusion-kinetic equation on the concentration of the monomer and chiefly the grafted polymer. From the physical point of view change in the boundary conditions of the problem in the course of the process is due to change in the values of equilibrium sorption of the monomer with change in the initial matrix during grafting. A large volume of calculations is required to solve the diffusion-kinetic equation having regard to the factors mentioned. It is planned to look at this question in future publications. The result of the quite complex interaction of the diffusion and kinetic factors is also the appearance of a limiting dependence of the polymerization rate on temperature (Fig. 5). Here, evidently, it is necessary to take into account not only the temperature dependence of the rates of propagation and termination of the kinetic chains but also the rise with temperature in the rate of the process of decay of the starting radicals competing with initiation and also (and possibly chiefly) the temperature dependence of the coefficient of diffusion and sorption of the monomer. We used the ESR method to investigate the behaviour of the radicals in the system directly in the
Postradiation graft polymerization FIG. 5. Temperaturedependence of the degree of graftingof AA on COPOL. Radiationdose 10 kGr; monomerconcentration40%. Thicknessof initialfilm20/~m. Fio. 6. Changein the ESR spectrum of radicalsstabilizedin the initial matrix of COPOL in the courseof postradiation graft polymerizationof MAA: (a) spectrum before contact with monomer (magnificationx 5); (b) spectrumafter contactwith MAA vapours (magnificationx 1).
439
~7= L 023
1
o
'
'
~- n
~
(a)
_/ l \
~)
a.%1
e0 FIG. 5
Jo
T,°C Flo. 6
course of graft postpolymerization. It was established that on irradiation of the starting COPOL films with doses up to 50 kGr radicals accumulate with a concentration (1-2)x 10TM spin/g. On contact of the irradiated samples with solution or the vapours of MAA the starting radicals passed practically quantitatively into the terminal radicals of propagation of the PMAA chains with the characteristic time ~ 1 min. The signal from the radicals accumulated in the matrix on preliminary irradiation disappeared and in its place in the ESR spectrum a highly characteristic nine-component signal from the terminal PMAA-radicals appeared (Fig. 6). This important result indicates that practically all the starting radicals irrespective of the region of their localization in the matrix (in crystallites or amorphous regions) are available for interaction with the monomer molecules. The possibility of the initiation of polymerization by the radicals localized in the crystalline regions is evidently connected with the existence of effective mechanisms of migration of radicals from the crystallites to their boundaries and accordingly to the monomer sorbed in the amorphous phase. The practically quantitative correspondence of the concentration of the terminal PMAA-radicals to the concentration of the starting ones and also its high value (~10 TM spin/g) indicates the small length of the kinetic and polymer chains of grafted PMAA. Thus, for a degree of grafting of PMAA 20% in the usual experimental conditions and at the concentration indicated of the terminal radicals the length of the chains will, on average, be ~5 x 102 monomer units. It should be noted that the process of stabilization found in the matrix of the terminal PMAA macroradicals may be due to the proneness of the growing PMAA centres to so-called self-occlusion characteristic of polymerization both of MAA itself and practically all its derivatives in heterogeneous conditions and also at the deep stages of the reaction on polymerization in bulk. By the "occlusion" effect is usually understood the phenomenon of heavy loss in activity of the growing macroradical not only in the reaction of quadratic termination but chiefly in the reaction of PS 3 2 : 3 - C
440
YE A. RAGOL'SKAYAet al.
propagation and transfer to the monomer of the kinetic chains on reaching a certain length of the material chains. The result is that in the system there is accumulation of terminal propagation radicals in sufficiently large numbers for their direct observation by the ESR method. The nature of this effect has not been finally established and calls for special study. It is also important to note that occlusion in the sense indicated is characteristic to a far lesser degree of monomers of other types (in particular, acrylic) not containing an a-methyl group at the double bond. The fact that the concentration of the occluded radicals practically does not change over a lengthy period suggests that the process of chain transfer to low molecular mass or other possible transfer agents through which quadratic termination may be realized is also heavily retarded. We observed a different situation on polymerization of A A on COPOL. The introduction of the monomer into the ampoule with irradiated film very insignificantly (5-10%) reduced the initial concentration of the radicals [R°]. No signals associated with the radicals of propagation of the PAA chains were found. It should be noted that there was no appreciable drop in the concentration of the starting radicals up to very high degrees of grafting (100% and more). A similar result was obtained when AA was polymerized both from aqueous solution and the vapour phase. It is important to note that some drop in the concentration of the starting radicals was observed on sorption by the irradiated sample of organic solvents, not monomers, and in the case of a good solvent for COPOL, for example, acetone the drop in [R °] was substantially larger than in the case of a precipitant, benzene, ethanol. This effect must be linked with rise in the rate of recombination of radicals in the matrix through the plasticizing action of the solvent raising the mobility of the chains. Thus, the observed drop in JR'] in the course of grafting in all likelihood is not connected with the occurrence of polymerization in the system. It should be noted that the A A and M A A molecules quite weakly differ both in diffusional properties and reactivity in polymerization and, therefore, it is hard to assume that the bulk of the starting radicals is unavailable for AA while practically all the starting radicals are available for the M A A molecules. Next it should be noted that A A on polymerization in heterogeneous conditions is not inclined to the process indicated above of occlusion of the growing radicals and, consequently, together with the chain propagation reaction the process of chain transfer to the polymer in the system is possible. The transfer radicals thereby formed are stabilized in the polymer matrix and are recorded by the ESR method. Since a very wide ESR spectrum ( - 2 0 0 gauss) with an unresolved structure (see Fig. 6) is characteristic of radicals stabilized in the COPOL matrix it is practically impossible to distinguish the starting radicals and those formed as a result of chain transfer to the polymer. The fact that polymerization effectively occurs in the system indicates the non-degenerate nature of this process. In fact, for the propagation radicals not to be observed in the ESR spectrum but only the transfer radicals, only a fundamental difference in their lifetimes is required and having regard to the real noise level on recording the spectra this difference must be more than 100 times. Thus, occlusion of the growing P M A A radicals may be regarded as a specific mechanism of termination playing the main role in systems with limited mobility of the chains. Occlusion of the growing centres is evidently also the main reason for the translimitation of the kinetic curves in time and also the fall in the polymerization rate of M A A as compared with AA in the system considered (Fig. 2). In the light of the ideas developed the cause of translimitation of the kinetic curves on polymerization of AA on C O P O L remains not quite clear. In all likelihood it is connected not with the expenditure of the initiating radicals but with irreversible changes in the matrix occurring in the course of polymerization. In particular, these may include change in the mechanical properties of the material in the course of grafting (with increase in the degree of grafting the rigidity of the
Nitrates-nitric acid system
441
matrix heavily rises) which may lead to fall in the availability of the transfer radicals or to hampering of the chain propagation reaction for steric reasons.
Translated by A. CRozv REFERENCES 1. N. I. GAL'PERINA, Dissert. Cand. Chem. Sci. (in Russian) p. 44, Karpov Physicochemical Research Inst. Moscow, 1976. 2. E. A. HEGAZY, I. ISHIGAKI, A. RABIE, A. M. DESSOUKI and J. OKAMOTO, J. Appl. Polymer Sci. 26: 3871, 1981. 3. Idem, Ibid. 28: 1465, 1983. 4. Idem, Ibid. 22: 3673, 1984. 5. K. KAJI, Ibid. 32: 4405, 1986. 6. Idem, Ibid. 28: 3767, 1983. 7. K. IMRE and G. ODIAN, J. Polymer Sci. Polymer Chem. Ed. 17: 2601, 1979. 8. L. P. K R U L ' , Geterogennaya struktura i svoistva privitykh polimernykh materialov (Heterogeneous Structure and Properties of Grafted Polymeric Materials) p. 16. Minsk, 1986.
PolymerScienceU.S.S.R. Vol. 32, No. 3, pp. 441-446, 1990 Printed in Great Britain.
0032-3950/90 $10.00+ .00 © 1991PergamonPressplc
EQUILIBRIUM IN THE CELLULOSE NITRATES--NITRIC ACID SYSTEM* A. A.
CHICHIROV,A. V. KUZNETSOV,Yu. M. KARGIN,V. V. KLOCHKOV, G. N. MARCHENKO and G. G. GARIFZYANOV Ulyanov-Lenin Kazan State University (Received 14 October 1988)
Equilibrium has been studied in the cellulose nitrates-nitric acid-water system. Samples relating to equilibrium conditions at 293 K have been synthesized and their monomer composition determined by high resolution 13C NMR. The equilibrium constants of the substitution reactions in the elementary cellulose unit have been calculated. The diagram of the distribution of the monomeric composition relative to the concentration of nitric acid in equilibrium conditions has been plotted.
THE DEVELOPMENT of high resolution 13CNMR for analysing the monomeric composition of partially substituted cellulose nitrates (CNs) allows more detailed consideration of a number of aspects of their synthesis and structure. The authors of Refs [1-3] relate the reaction of nitration of cellulose (C) to an equilibrium reaction although they note that the attainment of equilibrium is made very difficult by the *Vysokomol. soyed. A32: No. 3,502-506, 1990.
0032-3950/90 $10.00 + .00 © 1991 Pergamon Press plc
STUDY OF POSTRADIATION GRAFT POLYMERIZATION OF ACRYLIC AND METHACRYLIC ACIDS TO THE COPOLYMER OF TETRAFLUOROETHYLENE WITH VINYLIDENE FLUORIDE* YE. A . RAGOL'SKAYA, S. G . PRUTCHENKO, S. A . PAVLOV a n d E . N. TELESHOV Karpov PhysicochemicalResearch Institute (Received 14 October 1988)
Postradiation graft polymerizationof acrylicand methacrylicacids from solution and the gas phase onto the copolymer of tetrafluoroethylene with vinylidenefluoride has been studied. The dependences of the degree of grafting on the thickness of the films, temperature and grafting time have been obtained. Scanning electron microscopyhas been used to study the distribution of the grafted polymer over the section of the samples. From the results some conclusionsare drawn on the aspects of the effect of diffusionfactors on the gross kinetics of the process. A significantdifference in the polymerizationrates of acrylicand methacrylic acids was established in identical experimental conditions. The fall observed in the polymerization rate of methacrylic acid is related to the possibilityof "occlusion" of the terminal growing radicals established by EST spectroscopy. THE BRISK development of membrane technology in the last few years has revived the interest of investigators in the processes of modifying polymeric film materials by radiation graft polymerization. Of considerable interest for the requirements of membrane technology are materials based on fluorine-containing polymers with high chemical stability and good physical and thermomechanical properties. Grafting polymeric acids to these materials helps to impart to them a whole gamut of practically useful properties such as hydrophilicity, ion-exchange properties, raised adhesion, etc. In this work we study the postradiation graft polymerization of acrylic (AA) and methacrylic ( M A A ) acids on the copolymer of tetrafluoroethylene and vinylidene fluoride ( C O P O L ) which is now finding wide application in the production of film materials, asymmetrical microfiltration membranes and hollow fibres. We used C O P O L containing according to elemental analysis 70% vinylidene fluoride and 30% tetrafluoroethylene units. Prior to use C O P O L was purified by reprecipitation from solution in acetone into ethanol or water. The films were produced by casting from 2.5% C O P O L solution in acetone onto a cellophane support followed by drying at room temperature in air and adjustment to constant weight in v a c u o at 50°C. The degree of crystallinity of the films thus obtained determined by radiographic methods was 45%. A A and M A A were purified by standard techniques [1]. Preliminary irradiation of the films was carried out in v a c u o in sealed ampoules with 6°Co y-radiation with K-300,000 unit (Karpov Physicochemical Research Institute) at room temperature. The total absorbed dose was 10-50 k G r at the dose rate 2.4-8.0 Gr/s.
* Vysokomol. soyed. A32: No. 3,495-501, 1990.
435
436
YEA, RAGOL'SKAYAet al.
Postradiation graft polymerization was conducted both from an aqueous solution of the monomers and the vapour phase. For polymerization from solution to suppress possible homopolymerization to the solution was added an inhibitor--Mohr salt in a concentration 8 x 10 -5. After arrest of the reaction the grafted films were washed with hot water and dried to constant weight in vacuo at 50°C. The yield of the graft polymer (degree of grafting) was determined gravimetrically
P-P0 x 100%, P0
Q= - -
where P0 and P are the mass of the film before and after the reaction. The ESR spectra were recorded with the PA-100 spectrometer at the temperature of liquid nitrogen on carrying out polymerization in solution and at room temperature in the case of vapour phase grafting. The distribution of the graft polymer over the section of the modified samples was studied by scanning electron microscopy (SEM) and X-ray microanalysis (XMA) with the Jeol ISM-35CF instrument and also optical microscopy. To employ the SEM and XMA methods the samples were contrasted by introducing by ion exchange Pb 2÷ ions into the grafted PAA amd PMAA. Typical kinetic curves of the polymerization of AA from aqueous solution on a film of COPOL are given in Fig. 1. It will be seen that the reaction proceeds at a rate of dropping in time (for practically constant concentration of monomer in solution), a characteristic of the curves being the presence of an initial portion with auto-acceleration. The time taken to reach the limiting polymer yield alim significantly depended on the thickness of the film d and usually amounted to 1-3 h. Both the initial rate* and Qlim rose with increase in the radiation dose D practically similarly with the accumulation in the initial C O P O L matrix of R ° radicals determined by EST spectroscopy. At doses - 5 0 kGr there was simultaneous limitation of the rise in the concentration of radicals [R'] and the initial rate and Qlim.
80 :
:L
J * o
6' J
#0
Jo
$o
.Jog
Time, rain FIG. 1. Kinetic curves of polymerization of A A from 40% aqueous solution on films of COPOL of thickness 5 (1), 10 (2), 20 (3), 40 (4), 60 (5) and 96/zm (6). Dose of preliminary radiation 10 kGr, 25°C.
Similar behaviour was also characteristic of the polymerization of MAA on COPOL films although with the difference that grafting of M A A proceeded at far lower initial rates and Qlim (Fig. 2). Study of the kinetics of the polymerization of A A on films of different thickness d showed that the * Since the initial portion of the kinetic curves is of an S-shaped character, by the initial rate we shall later understand the mean tangent of the angle of slope of the straight line drawn through the origin of coordinates and the point of inflection in the curve.
Postradiation graft polymerization
437
a,'/, 200
80
-i '~°
6
100
2 I Time, h FIG. 2
- .
,
5O
.
I00 ~m
Fio. 3
FI~. 2.
Kinetic curves of polymerization of AA (1) and MAA (2) from 40% aqueous solution on COPOL films 20/.~m thick. Dose of preliminary radiation 50 kGr, 25°C. FIG. 3. Dependence of polymer yield on film thickness d at different stages of the reaction on postradiation polymerization of A A on COPOL. Time of reaction 5 (1), 10 (2), 15 (3), 30 (4) and 60 min (5); (6) Qlim- Dose of preliminary irradiation 10 kGr; concentration of AA solution 40%; 25°C.
initial rate and Olim essentially depend on d (Fig. 3). It will be seen that a l i m r o s e monotonically with increase in d while Q at the initial stages of the reaction passed through a maximum. The position of the maximum shifted to the region of large thicknesses with increase in the reaction time. The initial rate also passed through a maximum but for somewhat lower d values. It should be noted that the dependence of the kinetic parameters of the reaction on the geometric dimensions of the sample is a serious argument in favour of diffusion control of the reaction rate. Another consequence of the diffusion nature of the process is the marked non-uniformity of the distribution of the grafted phase over the section of the sample. The corresponding distribution profiles obtained by the XMA method are given in Fig. 4. It will be seen that the distribution is of a practically rectangular character. Such a type of distribution is quite characteristic of the processes of graft polymerization of A A on different polymeric supports (see, for example, Refs [2--6]). A distribution of a similar type was obtained by us both for polymerization of AA from solution and from the gas phase. Figure 4 shows that as the process deepens the fronts of the grafted phase move to the centre of the film although the concentration of the polymer in the grafted regions rises insignificantly. The degree of grafting for which the fronts meet rises somewhat with increase in the concentration of the monomer in solution and the thickness of the film. After meeting of the fronts the polymerization rate, as a rule, falls sharply and the kinetic curve is translimited. It is also interesting to note that when polymerization was conducted from the gas phase greater non-uniformity of the distribution was observed than from solution (Fig. 4, curves 6, 7) and in some cases translimitation of the polymer yield in time occurred before the boundaries of the grafted phase met. It is important to note that the patterns observed experimentally in some respects do not agree with the conclusions of the elementary theory of diffusion-controlled graft polymerization [7, 8]. Thus, for example, from the theory it follows that the rate of the process not complicated by secondary factors such as non-isothermicity, dependence of the diffusion coefficient on concentration, etc. decreases in porportion to d 2. The limiting degree of polymerization naturally must not depend on d. The advent of strongly convex (practically rectangular) fronts of the distribution is also not a consequence of the elementary diffusion theory which, on the contrary, predicts the concave character of this distribution [7, 8].
438
YEA. RA6OL'SKAYAet al.
i
l Ol~.m
!
101~m
Fic. 4. Distribution profiles of grafted PAA over the section of the film obtained by the XMA method: (1-5) polymerization from 40% aqueous solution on films 55/~m thick; (6, 7) polymerization from the gas phase on films 20 p.m thick. Radiation doses: 10 (1-5) and 50 kGr (6, 7). Yield of grafted polymer 13 (1), 27 (2), 42 (3), 54 (4), 73 (5), 28 (6) and 55% (7); 25°C.
Thus, the experimentally observed anomalous dependences of the polymer yield and the initial polymerization rate on d in the first place must be linked with the dependence of the diffusion coefficient and also the boundary conditions of the diffusion-kinetic equation on the concentration of the monomer and chiefly the grafted polymer. From the physical point of view change in the boundary conditions of the problem in the course of the process is due to change in the values of equilibrium sorption of the monomer with change in the initial matrix during grafting. A large volume of calculations is required to solve the diffusion-kinetic equation having regard to the factors mentioned. It is planned to look at this question in future publications. The result of the quite complex interaction of the diffusion and kinetic factors is also the appearance of a limiting dependence of the polymerization rate on temperature (Fig. 5). Here, evidently, it is necessary to take into account not only the temperature dependence of the rates of propagation and termination of the kinetic chains but also the rise with temperature in the rate of the process of decay of the starting radicals competing with initiation and also (and possibly chiefly) the temperature dependence of the coefficient of diffusion and sorption of the monomer. We used the ESR method to investigate the behaviour of the radicals in the system directly in the
Postradiation graft polymerization FIG. 5. Temperaturedependence of the degree of graftingof AA on COPOL. Radiationdose 10 kGr; monomerconcentration40%. Thicknessof initialfilm20/~m. Fio. 6. Changein the ESR spectrum of radicalsstabilizedin the initial matrix of COPOL in the courseof postradiation graft polymerizationof MAA: (a) spectrum before contact with monomer (magnificationx 5); (b) spectrumafter contactwith MAA vapours (magnificationx 1).
439
~7= L 023
1
o
'
'
~- n
~
(a)
_/ l \
~)
a.%1
e0 FIG. 5
Jo
T,°C Flo. 6
course of graft postpolymerization. It was established that on irradiation of the starting COPOL films with doses up to 50 kGr radicals accumulate with a concentration (1-2)x 10TM spin/g. On contact of the irradiated samples with solution or the vapours of MAA the starting radicals passed practically quantitatively into the terminal radicals of propagation of the PMAA chains with the characteristic time ~ 1 min. The signal from the radicals accumulated in the matrix on preliminary irradiation disappeared and in its place in the ESR spectrum a highly characteristic nine-component signal from the terminal PMAA-radicals appeared (Fig. 6). This important result indicates that practically all the starting radicals irrespective of the region of their localization in the matrix (in crystallites or amorphous regions) are available for interaction with the monomer molecules. The possibility of the initiation of polymerization by the radicals localized in the crystalline regions is evidently connected with the existence of effective mechanisms of migration of radicals from the crystallites to their boundaries and accordingly to the monomer sorbed in the amorphous phase. The practically quantitative correspondence of the concentration of the terminal PMAA-radicals to the concentration of the starting ones and also its high value (~10 TM spin/g) indicates the small length of the kinetic and polymer chains of grafted PMAA. Thus, for a degree of grafting of PMAA 20% in the usual experimental conditions and at the concentration indicated of the terminal radicals the length of the chains will, on average, be ~5 x 102 monomer units. It should be noted that the process of stabilization found in the matrix of the terminal PMAA macroradicals may be due to the proneness of the growing PMAA centres to so-called self-occlusion characteristic of polymerization both of MAA itself and practically all its derivatives in heterogeneous conditions and also at the deep stages of the reaction on polymerization in bulk. By the "occlusion" effect is usually understood the phenomenon of heavy loss in activity of the growing macroradical not only in the reaction of quadratic termination but chiefly in the reaction of PS 3 2 : 3 - C
440
YE A. RAGOL'SKAYAet al.
propagation and transfer to the monomer of the kinetic chains on reaching a certain length of the material chains. The result is that in the system there is accumulation of terminal propagation radicals in sufficiently large numbers for their direct observation by the ESR method. The nature of this effect has not been finally established and calls for special study. It is also important to note that occlusion in the sense indicated is characteristic to a far lesser degree of monomers of other types (in particular, acrylic) not containing an a-methyl group at the double bond. The fact that the concentration of the occluded radicals practically does not change over a lengthy period suggests that the process of chain transfer to low molecular mass or other possible transfer agents through which quadratic termination may be realized is also heavily retarded. We observed a different situation on polymerization of A A on COPOL. The introduction of the monomer into the ampoule with irradiated film very insignificantly (5-10%) reduced the initial concentration of the radicals [R°]. No signals associated with the radicals of propagation of the PAA chains were found. It should be noted that there was no appreciable drop in the concentration of the starting radicals up to very high degrees of grafting (100% and more). A similar result was obtained when AA was polymerized both from aqueous solution and the vapour phase. It is important to note that some drop in the concentration of the starting radicals was observed on sorption by the irradiated sample of organic solvents, not monomers, and in the case of a good solvent for COPOL, for example, acetone the drop in [R °] was substantially larger than in the case of a precipitant, benzene, ethanol. This effect must be linked with rise in the rate of recombination of radicals in the matrix through the plasticizing action of the solvent raising the mobility of the chains. Thus, the observed drop in JR'] in the course of grafting in all likelihood is not connected with the occurrence of polymerization in the system. It should be noted that the A A and M A A molecules quite weakly differ both in diffusional properties and reactivity in polymerization and, therefore, it is hard to assume that the bulk of the starting radicals is unavailable for AA while practically all the starting radicals are available for the M A A molecules. Next it should be noted that A A on polymerization in heterogeneous conditions is not inclined to the process indicated above of occlusion of the growing radicals and, consequently, together with the chain propagation reaction the process of chain transfer to the polymer in the system is possible. The transfer radicals thereby formed are stabilized in the polymer matrix and are recorded by the ESR method. Since a very wide ESR spectrum ( - 2 0 0 gauss) with an unresolved structure (see Fig. 6) is characteristic of radicals stabilized in the COPOL matrix it is practically impossible to distinguish the starting radicals and those formed as a result of chain transfer to the polymer. The fact that polymerization effectively occurs in the system indicates the non-degenerate nature of this process. In fact, for the propagation radicals not to be observed in the ESR spectrum but only the transfer radicals, only a fundamental difference in their lifetimes is required and having regard to the real noise level on recording the spectra this difference must be more than 100 times. Thus, occlusion of the growing P M A A radicals may be regarded as a specific mechanism of termination playing the main role in systems with limited mobility of the chains. Occlusion of the growing centres is evidently also the main reason for the translimitation of the kinetic curves in time and also the fall in the polymerization rate of M A A as compared with AA in the system considered (Fig. 2). In the light of the ideas developed the cause of translimitation of the kinetic curves on polymerization of AA on C O P O L remains not quite clear. In all likelihood it is connected not with the expenditure of the initiating radicals but with irreversible changes in the matrix occurring in the course of polymerization. In particular, these may include change in the mechanical properties of the material in the course of grafting (with increase in the degree of grafting the rigidity of the
Nitrates-nitric acid system
441
matrix heavily rises) which may lead to fall in the availability of the transfer radicals or to hampering of the chain propagation reaction for steric reasons.
Translated by A. CRozv REFERENCES 1. N. I. GAL'PERINA, Dissert. Cand. Chem. Sci. (in Russian) p. 44, Karpov Physicochemical Research Inst. Moscow, 1976. 2. E. A. HEGAZY, I. ISHIGAKI, A. RABIE, A. M. DESSOUKI and J. OKAMOTO, J. Appl. Polymer Sci. 26: 3871, 1981. 3. Idem, Ibid. 28: 1465, 1983. 4. Idem, Ibid. 22: 3673, 1984. 5. K. KAJI, Ibid. 32: 4405, 1986. 6. Idem, Ibid. 28: 3767, 1983. 7. K. IMRE and G. ODIAN, J. Polymer Sci. Polymer Chem. Ed. 17: 2601, 1979. 8. L. P. K R U L ' , Geterogennaya struktura i svoistva privitykh polimernykh materialov (Heterogeneous Structure and Properties of Grafted Polymeric Materials) p. 16. Minsk, 1986.
PolymerScienceU.S.S.R. Vol. 32, No. 3, pp. 441-446, 1990 Printed in Great Britain.
0032-3950/90 $10.00+ .00 © 1991PergamonPressplc
EQUILIBRIUM IN THE CELLULOSE NITRATES--NITRIC ACID SYSTEM* A. A.
CHICHIROV,A. V. KUZNETSOV,Yu. M. KARGIN,V. V. KLOCHKOV, G. N. MARCHENKO and G. G. GARIFZYANOV Ulyanov-Lenin Kazan State University (Received 14 October 1988)
Equilibrium has been studied in the cellulose nitrates-nitric acid-water system. Samples relating to equilibrium conditions at 293 K have been synthesized and their monomer composition determined by high resolution 13C NMR. The equilibrium constants of the substitution reactions in the elementary cellulose unit have been calculated. The diagram of the distribution of the monomeric composition relative to the concentration of nitric acid in equilibrium conditions has been plotted.
THE DEVELOPMENT of high resolution 13CNMR for analysing the monomeric composition of partially substituted cellulose nitrates (CNs) allows more detailed consideration of a number of aspects of their synthesis and structure. The authors of Refs [1-3] relate the reaction of nitration of cellulose (C) to an equilibrium reaction although they note that the attainment of equilibrium is made very difficult by the *Vysokomol. soyed. A32: No. 3,502-506, 1990.