CC BY
17
UDC 542.9:628.34
M. D. Goldfein, N. V. Kozhevnikov, N. I. Kozhevnikova, G. E. Zaikov
KINETICS AND MECHANISM OF POLYMER DISPERSION FORMATION ON BASED OF (METH)ACRYLATES. PART 3
Keywords: polymerization, emulsifying agent, ammonium persulfate, comonomers, latex.
The feasibility of synthesis of stable polymeric dispersions by polymerization in the absence of emulsifying agent under specially selected conditions by temperature, the concentration of the initiator (ammonium persulfate), the presence and nature of comonomers is shown. The mechanisms offormation and stabilization of latex particles in the emulsifier-free conditions are discussed.
Ключевые слова: полимеризация, эмульгатор, персульфат аммония, сомономеры, латекс.
Показана возможность синтеза полимеризацией без эмульгатора стабильных полимерных дисперсий в специально подобранных условиях температуры, концентрации инициатора (персульфат аммония), наличия и характера сомономеров. Обсуждаются механизмы формирования и стабилизации латексных частиц в условиях отсутствия эмульгаторов.
1 Introduction
Polymer dispersions are widely used in industry, agriculture, building, and home. They are typically prepared by emulsion polymerization in the presence of stabilizers for polymer particles (emulsifiers) [1]. However, there is a possibility of producing some synthetic latexes by emulsion polymerization without specially added surfactants, which allows one to obtain an environmentally friendly product [2]. The synthesis of polymer dispersions is complicated by the need to ensure their colloidal stability, which is achieved by introducing an emulsifier in conventional emulsion polymerization [3].
This paper presents the results of our study of the emulsifier-free polymerization of the alkyl esters of acrylic acid and their copolymerization with methacrylic acid and acrylonitrile.
2 Experimental
Since polymerization is accompanied by volume effects, dilatometry fully conforms to the specified requirements to rate estimation [4]. There exists proportionality between the volumetric changes of the polymerizate during the reaction and the weight monomer-to-polymer conversion degree, determined by the densities of the monomer and polymer: AK _Pn~ P m AP , K Pn Po
where AV and AP are the changes in volume of the polymerizate and the weight of the monomer during polymerization, Vo and Po the initial volume and
weight of the monomer, pm and pp the densities of the
monomer and polymer at the polymerization temperature.
Polymerization was carried out in special glass dilatometers of original design. The main elements of such a dilatometer are a measuring capillary, a reaction vessel, and a filling system providing the opportunity of releasing the monomer and dispersion medium from dissolved gases before polymerization. To avoid the
influence of oxygen on the processes studied, the water-soluble and oil-soluble components of the reaction system were separately freed from dissolved air by multiple freezing (liquid nitrogen), high vacuum pumping, and thawing in vacuo with subsequent transfusions through the measuring capillary into the reaction vessel of the dilatometer. The tool filled with inert gas was then placed into a water bath over an electrical magnetic stirrer. The desired temperature was maintained via an ultrathermostat.
The estimation of the number and size of the resulting latex particles formed in the reaction was made by classical theory of emulsion polymerization [5] and turbidity spectrum method.
The classical theory establishes the nature of the influence of the concentration of initiator In and emulsifier E just on the number of particles in the emulsion N, which, starting from the second (fixed) stage of the reaction, was considered to be independent of the conversion degree and, for micellar nucleation, is described by: N = K[Inf 4[Ef 6. The emulsion polymerization rate is determined by the number of
particles, the average number n of radicals, and the monomer concentration [M]p therein:
W = kp n [M]pN/NA.
The turbidity spectrum method [6] based on the determination of the wavelength exponent in Angstrom's equation z = const describing the spectral dependence of the turbidity .zof colloidal solutions within a relatively narrow wavelength range. Turbidity is the coefficient in the exponent of Bouguer's
law I _ I e~z l, where l is the scattering layer
thickness (the cuvette length) characterizing the ability of the disperse system to reduce the intensity of the incident light due to scattering. Turbidity depends on the number N of scattering centers, their sizes and optical properties: z= Nmr2K(a,m),
where K(a,m) is the scattering efficiency factor (or scattering coefficient), a = 2m-nc/l the relative size of a particle, m = n/no their relative refractive index (r the
radius of a particle, 2the wavelength of light, n and no the refractive indices of a particle and the dispersion medium). The wavelength exponent u is a function of a and m and can be found by measuring the slope of the straight portion of the graphic dependence of turbidity (or absorbance) D = lg (I0/I) = T-1/2.303 on 2
in the log-log coordinates.
It is possible to determine the parameters of the dispersion particles from the found wavelength exponent with the known dependences u(a,m) and K(a,m) on the relative size and the relative refractive index. These relationships for a discrete set of m according to the formulae of G. Mie's light scattering theory [7] were calculated [6]. However, these dependencies have an oscillating character for monodisperse systems. The allowance for the actual polydispersity of colloidal systems is typically provided by a graphical smoothing on the principle of symmetry of the oscillating curves plotted for specific values of the relative refractive index m. It is more convenient to use approximate analytic expressions for the characteristic functions of light scattering [8]. This approach allows computer processing of experimental data. Besides, due to the fact that the relative refractive index of the particles may continuously vary during the experiment (e.g., as the polymerization depth increases, the monomer-polymer ratio in the particles varies and, therefore, their spectral characteristics do as well), the table data calculated for a discrete set of m turn out to be inapplicable.
The mean radius and number concentration of particles in dispersed systems were calculated by the formulae [6]:
r = a2„/2m0 , N =4n'T(Am)n02/Am2K(a,m) a2, where 2m is the middle of the wavelength range in the
tog^h^ scale (!m = ■yj2max2min ).
The allowance for refractive index dispersion was made in accordance with the approaches to this problem developed in Ref. [9]. In this paper it is shown that the theoretical value of the wavelength exponent uo, which is calculated by the approximate equations with
r and N, and which corresponds to systems with negligible dispersion is related to the experimental parameter u by a relation like: uo = u + Au, where Au = kou + 2m(k - ko)/(m - 1) when u > 2 and Au = kou + um(k - ko)/(m - 1) when u < 2.
The quantities k and ko describe the refractive index dispersion of the dispersed phase n(2) and the dispersion medium no(2) and represent their logarithmic derivatives with respect to wavelength. The values of the parameter k are proposed [10] to be evaluated from the inverse relative dispersion S and the refractive index at the wavelength corresponding to the yellow line of sodium nD. Approximating n( 2) by Cauchy's
binomial formula, we obtain:
t(I) = (nD -1)( V + ¿c2) / nDS(AF2 - V) =
,
where 2 = [ 2f~2 + 2c~2 /2~1/2 = 552.4 nm is the middle of the spectral interval in the 1/2 coordinates between the F and C lines.
The relation for k(2) was used for the final
polymer dispersion when the conversion q approached 100%, and the latex particles were composed of polymer entirely. However, this approach is inapplicable in analysis of the properties of latex particles depending on the polymerization depth, as the particles consist of the polymer and monomer with their varying ratio depending on q. In the absence of monomer droplets, when almost all the monomer and polymer are in the latex particles, the refractive index of the scattering centers can be estimated through the corresponding values for the monomer nm and polymer np, in view of the volume ratio 9 of the monomer and polymer in a particle, calculated, in turn, from the conversion degree: n = (pnm + np)/(1 + 9).
But the values of S for such particles are unknown.
It has been shown [9] that the parameter k( 2)
can also be estimated based only on nD, since for more than 100 samples of polymers it is described by a general dependence approximated as: - k( 2) = Bo +
B1(nD - 1.4) + B2(nD - 1.4)2, where Bo = 0.01675, B1 = -0.026858, and B2 = 0.780829. The use of this approximation requiring the knowledge of nD, within the range nm ^ np (in the case of MA polymerization it is 1.4040-1.4725) to assess the value of k, apparently gives wrong results, since when nD = 1.4 - Bo/2B2 = 1.4172, this function has a minimum, which contradicts our experimental data. The dependence of k on the refractive index in this range of its values can be more accurately represented by linear interpolation between the values calculated for the monomer and polymer.
The parameter ko for the medium in the case of aqueous dispersions can be obtained from direct spectral measurements of no(2). Approximation of the data from Ref. [11] for no(2) of water at 20°C by Cauchy's tripartite formula gives ko = -0.0155 (2 = 552.4 nm).
Turbidity spectra were recorded on a SF-26
spectrophotometer. The value 2cp=2 = 552.4 nm was selected as the wavelength mid-range, and measurements were performed with a constant logarithmic step Alg2= 0.02.
The polymer dispersion formed by the emulsion polymerization of acrylates (the monomer concentration is 10-20%, the initiator is ammonium persulfate, APS) have relatively high optical density; therefore, they were diluted with water before measuring turbidity. However, it turned out that the results of measurements, and, in particular, the wavelength exponential in Angstrom's equation and the reduced turbidity (the turbidity multiplied by the dilution R) depended on the dilution (Fig. 1), especially at its relatively small values. These dependencies are not observed at high dilutions. These data indicate that in sufficiently concentrated dispersions there is multiple secondary scattering due to which more (as compared
with single scattering) light passes through the cuvette and reaches the receiver as if to reduce the turbidity of the system. The theory underlying the turbidity spectrum method does not account for multiple light scattering. Therefore, to estimate the parameters of the dispersion it is necessary to dilute it to such an extent that the reduced turbidity becomes independent of R and remains constant at any dilution, which serves as the criterion of no multiple scattering [10].
u z R
4,0
3,6
3,4
1
ff*»
0--0--0.....6.....o------o-
100 80 60 40 20
20
40
60
R
Fig. 1 - Dependence of the wavelength exponent (24) and the reduced turbidity of the dispersions (1, 5, 6) obtained by MA emulsion polymerization (50°C) with the monomer concentration of 10 (2, 4-6) and 20% (1, 3) on the dilution prior to measuring turbidity. The emulsifier (1%) is LS (1-3, 6) and Neonol (4, 5); [APS] x 103 = 2 (2, 4-6) and 36 mol/l (1, 3)
The obtained data have allowed us to estimate the contribution of multiple light scattering at different dilution degrees, since the proportion of multiply scattered light at the exit of the cell T2 can be expressed [10] through the reduced turbidity at the given dilution (zR) and infinite dilution (zR)m = lim (zR) :
R—
T2 = I2/I = 1 - 10°-°1, where D - D1 =
I (zR) - (zR)M 2.303 R '
(I) being the intensity of light transmitted through the dispersion, which consists of the intensity I1 weakened due to single scattering, and the additional intensity I2 arising due to multiple scattering of radiation in the direction of the receiver).
The effect of multiple light scattering quickly decreases with increasing dilution (Fig. 2).
The necessary dilution degree depends on the dispersion properties, in particular, it is higher when sulfated oxyethylated alkylphenols (C-10, nonoxynol-9-12) are used as an emulsifier in comparison with sodium lauryl sulfate (LS), and it also depends on the monomer content in the initial emulsion.
Dilution simultaneously facilitates washing of the latex particle surface from the emulsifier molecules stabilizing them, whose refractive index differs from the corresponding quantities of the monomer or polymer, which, in the absence of dilution, introduces an additional uncertainty into the results. Therefore,
polymer dispersions are pre-diluted with water by 100200 times and measurements are usually performed in 0.3-cm cuvettes with the value of transmittance maintaining within the range of 0.2-0.8.
Thus, our studies have shown that the turbidity spectrum method can be applied not only to the final polymer dispersions resulting from the emulsion polymerization of acrylic monomers but also to the emulsion systems arising in the course of the reaction at various conversions. It is necessary to take into account multiple (secondary) light scattering, and the dependences of the scattering properties of particles on the polymerization depth, those of the refractive indices of the particle and medium on wavelength. Under these conditions, the turbidity spectrum method can be used in kinetic studies of this reaction.
Fig. 2 - Dependence of the fraction of multiple light scattering at X = 552 nm at the cell outlet with I = 0.1 cm on the dilution degree of the dispersion obtained by MA polymerization of 20% (1) and 10% (2, 3) with LS (1, 3) and Neonol (2) as the emulsifier. [APS] x 103 = 2 (2, 3) and 36 mol/l (1); 50 0C
3 Results and Discussion
Some Aspects of Kinetics and Mechanism of Processes of Formation Polymeric Latexes in Absence Emulsifier
When emulsifier-free polymerization with persulfate-type initiators, the polymer particles are stabilized by the initiator's charged end groups. The
water-dissolved oligomeric radicals * MnSO4- grow
until some critical chain length is reached, whereby they lose their solubility and are extracted from solution to form nucleus PMP, leading to homogeneous nucleation [12]. Stabilization of the particles can be enhanced by copolymerization of monomers with ionizing or highly hydrophilic monomers.
Moreover, the "micellar" nucleation mechanism can be realized in the emulsifier-free conditions, which, with visible similarity, substantially differs from the particle formation occurring during conventional emulsion polymerization. In the absence of emulsifier micelles, the primary charged radicals resulting in the aqueous phase, when decomposition of a water-soluble persulfate-type initiator, after several acts of growth (interacting with the water-dissolved
0
monomer) react with each other to form oligomeric molecules with surface activity and being able to create micelle-like structures, which play the role of an "own" emulsifier. Subsequently, the monomer and oligomer radicals are absorbed by these "micelles", where chains can grow [3].
Therefore, preconditions are created for PMP stabilizing and the implementation of polymerization by the emulsion mechanism and without the participation of a specially introduced emulsifier.
In the absence of emulsifier, the polymerization of the monomer systems discussed proceeds by the emulsion mechanism, primarily in monomer-polymer particles. The basic regularities typical for the reaction in the presence of emulsifier are observed. This is indicated by the formation of an emulsion just at the initial stage of polymerization and relatively stable polymer dispersion upon its completion, the similarity of the kinetic curves (Fig. 3) and the kinetic regularities of both emulsifier-free and traditional emulsion polymerization in the presence of specially added emulsifiers. The most important differences between these two processes are the ways of particle generation and stabilization.
Wx 104 (mm"1)
Fig. 3 - Dependence of the rate of the emulsifier-free emulsion copolymerization of MA with AN (4%) and MAA (14%) on the monomer conversion degree. [APS] x 103 = 1 (1), 4 (2), 10 (3), and 20 mol/l (4), 700C
The kinetics of classical emulsion polymerization is characterized by three main stages, namely: a rapid increase in the rate (due to the formation of primary latex particles), the stationary reaction (while monomer droplets are present in the aqueous phase and its equilibrium concentration is set in the particles), and the completion of the process (as the monomer in the particles is exhausted)]. The same stages are also peculiar to emulsifier-free polymerization, but their course and duration depend on individual monomers and reaction conditions.
The relatively low stability of the polymer dispersions synthesized in the absence of a special emulsifier is expressed in the fact that, under certain conditions, a coagulum could be formed during
synthesis, as well as a precipitate and a clarified upper layer in the dispersion during its storage [13]. This indicates the existence of particles of different sizes: the larger ones form a precipitate while the smaller ones remain in a dispersed state. The considerable variation of the latex particles by size is possible if the step of their formation has a long duration. In the presence of emulsifier, more uniform dispersions are formed since PMP are formed with a high rate when radicals entering from the aqueous phase into the monomer-filled emulsifier micelles, where polymerization then proceeds with an equal probability. (The polymerization in the presence of certain emulsifiers soluble in the monomer is an exception, which the formation of new particles during the reaction is also characteristic of [14]). The appearance of a clarified layer without precipitate formation indicates that in the absence of a specially added emulsifier, though large but size-uniform particles may occur which, under insufficiently effective stabilization, gradually settle to form two layers with a high and low content of latex particles.
Emulsifier-free polymerization proceeds at a slower rate than in the presence of emulsifier, and an increase in the rate, longer by time and deeper by conversion degree, is observed, i.e. the first stage of the process (latex particle formation) lasts longer. The further variation of the reaction rate with increasing the degree of conversion, as well as in the case of conventional emulsion polymerization, is caused by changes in the number of particles in the emulsion because of their flocculation at high conversions or additional nucleation in the course of the reaction, and peculiarities of the reaction proceeding inside PMP: the formation of coarse particles during emulsifier-free polymerization contributes to the gel effect and an increased rate at high conversions, despite of the lower monomer concentration therein.
It is shown that an increased temperature leads to an increased polymerization rate and an increased number of latex particles in the final product (N), their reduced size (r), and a lower amount of coagulum formed (P) with the dispersion stability improved. Increasing temperature accelerates the decomposition of the initiator and leads to more frequent penetration of oligomeric radicals into particles to increase their charge and stabilization. The amount of radicals in the aqueous phase increases and the probability of their bimolecular interaction to form associates of water-soluble surfactant oligomers having micelle-forming properties and promoting nucleation rises. The probability of oligomeric radicals to reach a critical chain length at which they would lose their solubility and form new particles increases as well. Therefore, changing temperature affects not only the rate of generation of primary radicals and their growth in PMP but also other processes involved in nucleation. Besides, increasing temperature influences the value of gel effect, which also affects the effective activation energy of polymerization. The ratio of the maximum rate to that at 10% conversion (corresponding to the second stationary stage of the reaction) can be regarded as a measure of the gel effect. In the copolymerization of butyl acrylate with AN and MAA in the absence of
emulsifier, it increases with temperature up to 70-75°C, and decreases at higher values (Fig. 4), reflecting opposite tendencies in the effect of the reaction temperature on the probability of coexistence of several radicals in PMP. This probability increases with the initiation rate but decreases when the particle size reduces, which occurs under these conditions. As a result, the maximum rate of polymerization (Wmax)
does not obey the Arrhenius dependence, and the experimental data in the lgWmax vs. l/T coordinates
do not fit a straight line (Fig. 2), indicating a changed reaction mechanism in various temperature ranges and the role of individual factors which determine the rate of the process.
The initiator concentration also renders significant impact on emulsifier-free polymerization. With its increasing, the amount of latex particles in the final dispersion increases but decreases at higher concentrations, passing through a maximum (Fig. 5). When BA is copolymerized with AN and MAA, this maximum is observed at [APS] = 15 x l0-3 mol/l (85°C). The reaction rate also changes, to various degrees, depending on the monomer-to-polymer conversion degree. At low conversions (10%) the rate, like the number of particles, has an extreme dependence on the APS concentration.
the studied monomers are low-boiling (80°C for MA, 77°C for AN). However, the synthesis of dispersions without registering rate (in sealed ampoules) was performed at higher temperatures as well.
W x 104 (min1)
СЦх (%)
r (nm) N x 10"13 (mi3)
30 40 [APS] * 10" (mol/l)
Fig. 5 - Dependence of the maximum rate (1) of the emulsifier-free emulsion copolymerization of BA with AN (4%) and MAA (14%), the corresponding degree of conversion (3), the number of particles in the resulting polymer dispersion (2), their average radius (5), and the rate at a 10% conversion (4) on the initiator concentration. 85 0C
Fig. 4 - Temperature dependence of the size of latex particles (2) and the maximum rate-steady rate differences (1) of the emulsion copolymerization of BA with AN (4%) and MAA (14%). 3 - Arrhenius' dependence for Wmax. [APS] = 40 x 10-3 mol/l
When the initiator concentration increases, the depth of polymerization, at which the maximum rate is achieved for given conditions of the process, shifts towards higher values, and the rate itself rises. But its logarithmic dependence on the initiator concentration is nonlinear. In a BA-based monomer system the order by initiator (n,) decreases from 0.6 when [APS] < 10-2 mol/l down to 0.3 at higher concentrations. (According to the classical theory, n,- = 0.4 [15]). In the case of the copolymerization of methyl acrylate (MA) at a sufficiently high content of the initiator (> 10-2 mol/l) Wmax almost no longer depends on the initiation rate as
well. Dilatometric measurements in these systems were carried out at temperatures not exceeding 70°C, because
The initiation rate also affects the stability of the resulting product. BA-based dispersions which form neither precipitate nor clarified layer during storage were obtained at high temperatures (80-85°C) only in a narrow range of APS concentration (10-20) x 10-3 mol/l and simultaneous copolymerization with MAA and AN. In the presence of only one of the said comonomers within a concentration range of 0-14% we have failed to avoid the appearance of a clarified layer, but its volume and rate of formation decrease when the comonomer concentration increases [13].
Delamination to form a clarified layer (even more pronounced) occurs during storage of MA-based dispersions as well. This process is more probable in the case of the synthesis of dispersions with low initiation rates, and the amount of the coagulum produced has a minimum value in the range of APS concentrations (110) x 10-3 mol/l (90°C) (Fig. 6).
Therefore, stable polymer dispersions were obtained only at a high temperature of synthesis, a relatively high content of hydrophilic comonomers, within a narrow concentration range of the initiator.
It should be noted that in the emulsifier-free copolymerization of alkyl acrylates the stability of emulsions and final dispersions decreases in the row butylacrylate > ethylacrylate > methylacrylate, i.e. the stability deteriorates with an increased water solubility (polarity) of the primary monomer. BA-based dispersions consist of a much larger number of smaller particles than those on the basis of MA. Their better stabilization, with surfactant oligomers involved, may be associated with the fact that the surface activity of the "own" surfactants increases with an increased hydrophobicity of the particle surface. This is confirmed by the increase in the interaction energy of the
emulsifier with the organic phase as the polarity of alkyl acrylate decreases [1].
NxlO"1' (sm"3)
Г (nm) P (%)
20 30
[APSl X 10j (mo 1/1) Fig. 6 - Dependence of the number of particles (1) in the polymer dispersion obtained by the emulsifier-free polymerization of MA with AN (4%) and MAA (14%), their average radius (2), and the amount of the coagulum formed (3) on the initiator concentration; 90 оС
The extreme dependences on the APS concentration observed for the number of particles in the final dispersion and the reaction rate at low conversions (Figs 5 and 6), just at polymerization in the presence of emulsifier, are associated with insufficient stabilization of latex particles and their flocculation which occurs at various degrees of conversion, and not only at the initial stage of polymerization, as follows from the classical theory [15]. The reduction of the number of particles in the polymer emulsion when increasing the conversion at the third stage of emulsifier-free polymerization has been found experimentally. In the range of APS concentration corresponding to the maximum number of particles in the dispersion and their minimum size, the lowest number of the coagulum formed and the highest stability of the dispersion during storage are observed. Moreover, the content of the initiator affects the balance among various nucleation mechanisms.
At polymerization with low initiation rates, the aqueous phase of the emulsion system contains a low concentration of oligomeric radicals and their recombination probability is low. Under these conditions, many of them would have time to reach a critical size and to form primary particles by the homogeneous mechanism. Their stabilization occurs due to the polar groups of the monomers and the surface charge arising due to the presence of end groups in the persulfate initiator. To create a sufficient charge density, the primary particles flocculate at the stage of their formation. During subsequent polymerization, together with an increase in the volume (and surface) of the particles, their charge increases due to the periodic
introduction of charged oligomeric radicals * MnSO4~
to PMP. An increased initiation rate leads to an increased number of radicals turning into primary particles, and the total number of PMP increases. More frequent penetration of oligomer radicals to the particles
occurs as well, which improves their stabilization and prevents flocculation at later stages of the reaction.
At higher initiator concentrations, the role of the bimolecular chain breakage increases due to recombination of oligomeric radicals in the aqueous phase. As a result, water-soluble surfactant oligomers appear to play the role of an emulsifier. If their concentration in water exceeds the critical micelle concentration (which is possible at a sufficiently high initiation rate) then the associates of these oligomers form micelles which the monomer diffuses into and wherein polymerization can proceed. Under these conditions, the contribution of the micellar mechanism of PMP nucleation rises. Their surface is protected from the very beginning with the "own" emulsifier, which inhibits flocculation at the first stage of the reaction and promotes increasing N. However, if the "own" emulsifier is formed and adsorbed more slowly than the surface requiring stabilization grows (large rate constants of chain propagation and the concentration of the monomer in particles but a relatively small initiation rate), flocculation takes place at the second stage of polymerization.
At the third step of the reaction corresponding to the disappearance of monomer droplets and the reduction of the monomer concentration in PMP, the protective effect of the "own" emulsifier deteriorates due to an increased surface charge of the particles due to the introduction of new charged oligomeric radicals therein. As a result, flocculation of the particles occurs until the effect of another stabilization mechanism (due to the surface charge as in the case of homogeneous nucleation) gets sufficiently effective. The transition from stabilization by surfactant oligomers to stabilization due to electrostatic forces occurring at the third stage of polymerization is accompanied by reduction in the PMP number. As a result, N passes through a maximum with increasing APS concentration.
In the case of MA-based dispersions, the appearance of another peak on the N vs. [APS] dependence was detected at high initiator concentrations (Fig. 6, curve 1). Under these conditions the rate of formation of the "own" emulsifier by recombination of oligomeric radicals in water may turn out to be so high that flocculation at the second step of polymerization begins to play a less important role, which leads to an increased number of particles. However, flocculation at the third stage of the reaction enhances with increasing the APS concentration, because oligomeric radicals enter the particle more often and the surface charge grows faster to worsen the operating conditions for the emulsifier. Therefore, the flocculation increase with an increased content of the initiator reduces N again.
The surface activity of the "own" emulsifier depends on the structure and composition of the oligomers, and, consequently, on the properties of the comonomers. In the case of BA no second peak has been detected, but it may, like the first maximum, occur at higher initiation rates than that of methyl acrylate latexes, outside of the range investigated. It can be assumed that the transition to the micellar nucleation mechanism as well as the termination of flocculation at the second stage of the reaction, at BA
copolymerization occurs at higher rate initiation than in the case of MA. I.e. the surfactant oligomers resulting from radical recombination in the aqueous phase are produced slowly. These oligomers must have a different composition at copolymerization of MA or BA due to their different reactivity and the solubility in water (5 and 0.2 %, respectively [1]).
Taking into account that the MAA distribution coefficient between the aqueous and organic phases is less than unity, we can assume that within the investigated concentration range of MAA in the monomer system (< 14%) the content of MA in water is higher than for MAA. However, MAA is more actively involved in the copolymerization reaction [16]. Therefore, the oligomeric radicals formed from MA and MAA copolymerization apparently consist of units of both monomers. In contrast, the concentration of BA in water is less than that of MAA, and the activity of MAA in copolymerization with BA is also high (rMAA = 1.3, rBA = 0.3 [17]). In this case, the oligomeric radicals are mainly composed of MAA. It is known that at copolymerization of monomers of different polarity a tendency to cross breakage is observed due to the effects of electron transfer [18]. The rate constant of cross break is higher than at homopolymerization of the corresponding monomers. In this connection, breakage in the aqueous phase to form surfactant oligomers at copolymerization of MA is more probable, which explains the lower (than for BA) APS concentrations, which the maximum number of particles in the dispersion is achieved at.
At copolymerization in the MA-based ternary system, all the monomers are (partially) dissolved in the aqueous phase, namely: MA, MAA, and AN. The copolymerization of AN with MA is characterized by the constants: rAN = 1.5, rMA = 0.84 [19]. Involving AN to oligomeric radicals leads to acceleration of their
bimolecular interaction since the rate constant k of
0
chain termination of AN is almost two orders of magnitude higher than that of MA [18]. This promotes strengthening of the role of the micellar mechanism of particle formation.
In the case of ternary BA-based systems, copolymerization of MAA and AN should mainly occur in the aqueous phase. And AN less actively participates in the reaction, as evidenced by its much slower consumption when the copolymerization reaction (by chromatographic data) and the resulting oligomeric radicals are mainly composed of MAA units. The breakage constant at polymerization of this monomer in water [20] is close to k0 of acrylic esters.
Flocculation of particles at different stages of the polymerization process and its dependence on the concentration of the initiator influence not only the number of particles in the final dispersion but, together with the gel effect, also the value of the reaction rate at different conversions and, consequently, the shape of the kinetic curves. Apparently, just the flocculation at the second stage, growing within some range of the APS concentration, causes an end of the rate growth with the polymerization depth (or even its slight decrease) at
relatively small depths of conversion observed under these conditions (Fig. 3, curve 4).
Unlike conventional emulsion polymerization with a specially introduced emulsifier which is spent for the formation and stabilization of the resulting latex particles, at emulsifier-free polymerization during the reaction an "own" emulsifier is formed, which participates in the origination of new particles. The gradual emergence of new surfactant oligomers increases the nucleation stage duration. Therefore, with the growth of the initiation rate, the instant of reaching the maximum rate of polymerization shifts towards higher conversion degrees (Fig. 5). Besides, this increases the heterogeneity degree of particles by size, promotes enhancing flocculation, coagulum and precipitate formation during storage of the dispersion. On the other hand, an increased APS concentration promotes a better stabilization of the particles and the clarified layer decreases.
The formation of latex particles during emulsifier-free polymerization occurs by various mechanisms whose relative contributions depend on the reaction conditions. It has turned out that changes in some of them can influence the nature of the effects of other factors. E.g. an increased MAA concentration at relatively low initiation rates ([APS] = 5 x 10-3 mol/l, 70°C) leads to a decreased rate of the emulsifier-free polymerization of methyl acrylate, ethyl acrylate, and butyl acrylate and the number of latex particles, an increased conversion corresponding to the maximum rate. The particle size and the amount of coagulum formed increase as well (Fig. 7). On the contrary, at high initiation rates (85-90°C) MAA improves the stability of the dispersion, increases the number of particles and the reaction rate (Fig. 8).
Fig. 7 - Dependence of the maximum rate of the emulsifier-free emulsion copolymerization of MA with MAA (4), the corresponding depth of polymerization (1), the number of particles in the polymer dispersion (5) and their average radius (2), and the amount of the coagulum formed (3) on the MAA concentration. [APS] = 5 x 10-3 mol/l; 70 0C
W^x lO^rain"1)
Nx 10"12(sm"3)
(70°C) as well, although in the absence of MAA or at its
low concentrations, a decrease in N is observed (Fig. 9).
N/K
[MAA] (% of [M]) Fig. 8 - Effect of the MAA concentration on the maximum rate of reaction (1) and the number of particles in the dispersion (2, 3) formed by the emulsifier-free emulsion copolymerization of BA (1, 2) and MA (3). [APS] х 103 = 5 (3) and 15 mol/l (1, 2); Т = 85 (1, 2) and 90 оС (3)
In the presence of MAA, the probability of recombination of oligomeric radicals in the aqueous phase increases due to their hindered entry into PMP due to the poor solubility in the monomers. In the presence of a specifically introduced emulsifier this leads to reduction in the number of PMP and the polymerization rate and to strengthening of the gel effect in the resulting larger particles. The same phenomena are also observed while emulsifier-free polymerization at relatively low initiation rates (low temperatures) when the homogeneous nucleation mechanism predominates. However, at higher temperatures (higher initiation rates), PMP are mainly formed by the micelle mechanism as a result of the formation of their "own" emulsifier. Strengthening of bimolecular breakage in water under the influence of MAA promotes the appearance of water-soluble surfactant oligomers, playing the role of an emulsifier, the improved particle stabilization, an increase in their number and polymerization rate. Therefore, some of the processes proceeding in the reaction system may result in various effects at conventional and emulsifier-free emulsion polymerizations.
Similar differences in the effect on emulsifier-free emulsion polymerization at different initiation rates have been detected for AN as well, which retards the polymerization of MA and reduces N at 70°C (the predominance of the homogeneous nucleation mechanism) but increases it at 90°C, i.e. under those conditions in which the micellar nucleation mechanism involving surfactant oligomers is better expressed. Moreover, it turns out that the effect of one of the comonomers on the nucleation process may depend on the availability of other comonomers. E.g., an increased AN concentration in the ternary monomer system with a high MAA content (14%), when the micellar nucleation mechanism seems more probable, leads to an increased number of particles at relatively low temperatures
[AN] (% of [M])
Fig. 9 - Effect of AN on the number of particles in the polymer dispersions produced during the emulsifier-free emulsion copolymerization of MA with AN (3) with MAA (4%) and AN (2), and with MAA (14%) and AN (1). [APS] = 4 х 10-3 mol/l; 70оС
An increased AN content in BA-based systems leads to an increased number of particles in the final polymer dispersion but the polymerization rate lowers. Increasing N in emulsifier-free conditions is associated with the improved particle stabilization during copolymerization with a hydrophilic monomer having polar CN-groups. However, just as in the presence of a specific emulsifier, AN weakens the gel effect, thereby reducing the reaction rate in the particle. Moreover, it suppresses flocculation at the third polymerization stage, whereby the number of particles reduces by the end of the reaction to a lesser extent than in the absence of AN. This leads to the observed differences in the effect of comonomer on the number of particles and the rate of emulsion polymerization.
Thus, this study shows the possibility of the synthesis of polymer dispersions in the absence of any emulsifying agent, but only under specially selected conditions (temperature, initiator concentration, the presence and nature of comonomers).
4 Conclusions
The feasibility of synthesis of stable polymeric dispersions by polymerization in the absence of emulsifying agent under specially selected conditions by temperature, the concentration of the initiator (ammonium persulfate), the presence and nature of comonomers is shown. The mechanisms of formation and stabilization of latex particles in the emulsifier-free conditions are discussed [21].
References
1. Eliseeva, V.I. Polymeric Dispersions (Moscow, 1980). 296 pp. (in Russian).
2. Aslamazova, T.R., Bogdanova, S.V. and Movchan,T.G., Ross. Chemical Journal, 37 (4), 112-114 (1993).
3. Eliseeva, V.I. and Aslamazova, T.R., Uspekhi Khimii, 60 (2), 398-429 (1991).
4. Gladyshev, G.P. and Popov, V.A. Radical Polymerization at High Degrees of Conversion (Moscow, 1974), 244 (in Russian).
5. Smith, W.V. and Ewart, R.H. J. Chem. Phys., 16 (6), 592599 (1948).
6. Klenin, V.I., Shchegolev, S.Yu. and Lavrushin, V.I. Characteristic functions of light scattering from disperse systems. Saratov: Saratov Univ. Press, 1977. 177 p.
7. Bohren, C.F. and Huffman, D.R. Absorption and Scattering of Light by Small Particles (Wiley-Interscience, 1983).11. Ivanchev, S.S. Radical Polymerization (Leningrad, 1985). 280 pp. (in Russian).
8. Ramazanov, K.R., Khlebtsov, N.G., Shchegolev, S.Yu. and Klenin, V.I. Colloid Journal, 45 (3), 473-479 (1983).
9. Khlebtsov, N.G., Melnikov, A.G. and Shchegolev, S.Yu. Colloid Journal, 53 (5), 928-933 (1991).
10. Klenin, V.I., Shchegolev, S.Yu. and Lebedeva, L.G/ Optics and Spectroscopy, 35 (6), 1161-1166 (1973).
11. Joffe, B.V. Refractometric methods in chemistry. Leningrad: Khimiya. 1983. 352 p.
12. Gardon, J.L. Rubber Chem. and Techn0l., 43 (1), 74 (1970).
13. Kozhevnikov, N.V., Goldfein, M.D. and Trubnikov, A.V., Intern. Journal Polymeric Mater., 46, 95-105 (2000).
14. Kabanov, V.A., Kurilova, A.I. and Topchiev, D.A. Vys0k0m0lek. S0ed., B-15 (9), 712-717 (1973).
15. Smith, W.V. and Ewart, R.H. J. Chem. Phys., 16 (6), 592599 (1948).
16. Nazarova, I.V. and Eliseeva, V.I., Mendeleyev S0C. J0urnal, 12 (4), 587-588 (1967).
17. Kulikov, S.A.,Yablokova, N.V. and Nikolaeva, T.V. Vys0k0m0l. S0ed. A-31 (11), 2322-2326 (1989).
18. Bagdasariyan, H.S. Theory of Radical Polymerization (Moscow, 1966). 300.
19. Encyclopaedia of Polymers. Vol. 1. Moscow, Savetskaya Encychpaedia, 1972.
20. Kabanov, V.A., Zubov, V.P. and Semchikov Yu.D. Complex-Radical Polymerization (Moscow, 1987). 256 pp.
21. Kozhevnikov, N.V., Goldfein, M.D. and Kozhevnikova N.I., Proceeding of Saratov University. New Series. Chemistry, Bi0l0gy, Ec0l0gy, 14 (1), 38-47 (2014).
© M. D. Goldfein - Doctor of Chemistry, Full Professor, Saratov State University, Saratov, Russia, N. V. Kozhevnikov - Doctor of Chemistry, Full Professor, Saratov State University, Saratov, Russia, N. I. Kozhevnikova - Ph.D., Associate Professor, Saratov State University, Saratov, Russia, G. E. Zaikov - Doctor of Chemistry, Full Professor, Plastics Technology Department, Kazan National Research Technological University, Kazan, Russia, abzaldinov@mail.ru.
© М. Д. Гольдфейн - доктор химических наук, профессор, Саратовский государственный университет, Саратов, Россия, Н. В. Кожевников - доктор химических наук, профессор, Саратовский государственный университет, Саратов, Россия, Н. И. Кожевникова - кандидат химических наук, доцент, Саратовский государственный университет, Саратов, Россия, Г. Е. Заиков - доктор химических наук, профессор, кафедра Технологии пластических масс, Казанский национальный исследовательский технологический университет, Казань, Россия. abzaldinov@mail.ru.
