Polym Int 52:542–547 (2003) DOI: 10.1002/pi.1029
Polymer International
Encapsulation of an organic phthalocyanine blue pigment into polystyrene latex particles using a miniemulsion polymerization process† S Lelu, C Novat, C Graillat, A Guyot and E Bourgeat-Lami* LCPP, CNRS-CPE, Baˆt. 308F, 43, Bd. du 11 Nov. 1918, 69616 Villeurbanne, France
Abstract: Aqueous dispersions of polystyrene latexes encapsulating a copper phthalocyanine blue pigment were formulated using the miniemulsion polymerization technique. The organic pigment was first suspended into the monomer phase, and the resulting oily suspension was subsequently converted into stable miniemulsion droplets using various types and concentrations of hydrophobe (costabilizer). The pigmented monomer emulsions were finally polymerized using potassium persulfate as the initiator. It was shown that the organic pigment could stabilize the miniemulsion droplets, and be thus satisfactorily encapsulated without introducing any other compound in the formulation. In a subsequent approach, the stability of the miniemulsion droplets was improved by using either hexadecane, hexadecanol or a polystyrene prepolymer as the hydrophobe. Dynamic light scattering and transmission electron microscopy measurements showed that the size and the morphology of the resulting pigmented polymer particles were greatly influenced by the presence of the costabilizer. # 2003 Society of Chemical Industry
Keywords: encapsulation; phthalocyanine; organic pigment; miniemulsion; polymerization
INTRODUCTION
In the last ten years, a variety of encapsulation techniques have been developed to synthesize nanocomposite particles for applications in catalysis, optics and coatings.1,2 One of the most common methods is emulsion polymerization.2–6 Successful encapsulations have been reported, for instance, with titanium dioxide pigments3–5 and colloidal silica6,7 using various polymeric materials. Efficient encapsulation, however, requires modification of the surface of the pigment in order to afford convenient interactions with the polymer. In addition, the particles must be correctly dispersed in the continuous phase to avoid the presence of large aggregates in the encapsulated materials. For these reasons, and also because of the complexity of the particle nucleation mechanism in emulsion polymerization, it appears that it is often difficult to achieve high encapsulation efficiencies by this technique. In recent literature, miniemulsion polymerization was found to be particularly attractive to obtain polymeric nanoparticles which cannot be achieved by current procedures8,9 Indeed, the characteristic features of miniemulsion polymerization offer several advantages in comparison with emulsion polymerization. Miniemulsions consist of a liquid/liquid dispersion of a monomer phase in water with diameters in
the range of approximately 50–500 nm. The size of the monomer droplets is usually controlled by shearing the system in the presence of a surfactant and an hydrophobe (costabilizer), whose role is to stabilize the emulsion against diffusion degradation (Ostwald ripening). Contrary to conventional emulsion polymerization, the monomer droplets are sufficiently small and numerous so that the polymerization predominantly occurs by radical entry into the preexisting miniemulsion droplets without formation of new particles. Miniemulsion polymerization is therefore particularly attractive for the encapsulation reaction of any compound that can be satisfactorily suspended into the monomer phase10,11 as schematically represented in Scheme 1. In the present work, we want to report our preliminary investigations along this line on the encapsulation of an organic pigment, copper phthalocyanine blue pigment. Organic pigments have a wide range of commercial applications in coatings, printing and paint industries. They are of particular interest because of their photosensitivity, colour strength and overall stability. However, organic pigments are insoluble in water, and thus difficult to disperse in aqueous solution. The coating of organic pigments by polymers may be of great benefit to improve their processing. The coating could prevent pigment ag-
* Correspondence to: E Bourgeat-Lami, LCPP, CNRS-CPE, Baˆt 308F, 43, Bd du 11 Nov 1918, 69616 Villeurbanne, France E-mail:
[email protected] † Poster presentation – Paper presented at the Formula III Conference: New Concepts and Strategies in Formularies, from Laboratory to Industry, 13–16 October, 2001, Hrault, France (Received 21 January 2002; revised version received 9 April 2002; accepted 30 April 2002)
# 2003 Society of Chemical Industry. Polym Int 0959–8103/2003/$30.00
542
Encapsulation of phthalocyanine blue pigment into polystyrene
Scheme 1. Schematic representation of the encapsulation reaction of organic pigments through miniemulsion polymerization.
glomeration and protect it from environmental aggressions (UV radiation, pH). Better storage stability, colour stability and durability are expected to be obtained after coating.12 Polymer encapsulated organic pigments could therefore find applications in water based paints into which they could be readily incorporated.
EXPERIMENTAL Materials and methods
The monomer (styrene, Aldrich) was distilled under vacuum and kept refrigerated until use. The surfactant sodium dodecyl sulfate (SDS, Aldrich), the initiators: potassium persulfate (KPS) and azobis isobutyronitrile (AIBN) from Aldrich, and the hydrophobes, hexadecanol (Janssen) and hexadecane (Acros Organics), were used as received. The organic copper phthalocyanine blue pigment, a gift from Tioxide (blue Magistral), was used as supplied. An ultrasonic bath was used to help disperse the blue pigment in the monomer phase. Polymerizations
The recipes used for the syntheses of the polymer
latexes are summarized in Table 1. In a typical experiment, a known amount of the organic pigment (typically 0.3 g) was suspended into 10 g of styrene with the aid of ultrasound or under magnetic stirring. The dispersion was subsequently introduced into 100 g of an aqueous solution containing the surfactant (SDS, 0.33 g) and known amounts of the hydrophobe (hexadecane or hexadecanol, from 0 to 0.8 g), stirred for 1 h, and sonified with a 600 W ultrasonic Branson sonifier for 1 min at 90% output power, to create the minimulsion droplets. As an alternative, a polystyrene prepolymer was synthesized in presence of the pigment, and used as the hydrophobe. Prepolymerization was performed in bulk using a capped glass vessel fitted with a condenser at 70 °C for 3 h in a nitrogen atmosphere using 80 g of styrene, 0.8 g of AIBN and 4 g of pigment. The polymerization reached 40% conversion and afforded a low molecular weight polymer with narrow molecular weight distributions (Mn = 35 200 g mol 1 and Mw /Mn = 1.86). A known amount of the resulting suspension (6.3 g) containing residual monomer (2.5 g), polystyrene prepolymer (3.5 g) and the blue pigment (0.3 g) was then diluted into the corresponding amount of styrene (8.5 g), introduced into water, and emulsified as described
Table 1. Experimental recipes, diameters and polymer conversions in the synthesis of crude (runs 1–4) and pigmented (runs 5–8) polystyrene latex particles through miniemulsion polymerization
Runs 1 2 3 4 5 6 7 8 9
Hydrophobe (g)
Pigment (g)
0 0.07a 0.77a 0.8b 0* 0.77a 0.77b 0.77b,* 3.5c,*
0 0 0 0 0.3 0.3 0.3 0.3 0.3
Droplet size (nm)
Particle size (nm)
Ddroplets
PI
Dparticles
PI
Conversion (%)
Ndroplets/Nparticles
800 218.5 174 169.5 220 192 181 189 176
1.0 0.22 0.11 0.17 0.45 0.17 0.23 0.2 0.189
73.5 87.8 124.6 97.8 94 117 117 113 130
0.07 0.05 0.03 0.08 0.30 0.15 0.22 0.21 0.22
88.1 96 96.5 88.8 97 83 55.3 69.2 82.8
0.001 0.07 0.42 0.22 0.1 0.25 0.29 0.23 0.43
Nature of the hydrophobe: a hexadecanol. b hexadecane. c polystyrene prepolymer (Mn = 35 220 g mol 1 and Mw = 65 600 g mol 1). Styrene: 10 g, SDS: 0.33 g, KPS: 0.1 g, Water: 100 g. PI: polydispersity index. * Pigment dispersion performed under magnetic stirring.
Polym Int 52:542–547 (2003)
543
S Lelu et al
above. Polymerization of the miniemulsion droplets was carried out in a double wall thermostatted reactor at 70 °C for 3 h under permanent stirring (150 rpm). KPS was used as a water-soluble initiator, and was introduced at once to start polymerization. Pigment solubility in the monomer phase was determined gravimetrically after separation of the insoluble part by centrifugation. Characterizations
The particle and droplet sizes were determined by dynamic light scattering (DLS) using a LoC Malvern QELS instrument working at a fixed angle of 90°. Droplet size was recorded immediately after sonication by diluting the sample with a saturated SDS aqueous solution. The autosizer provided average diameters and polydispersity indexes (PI). The monomer conversions were determined by gravimetry. Transmission electron microscopy (TEM) was used to characterize the morphology of the polymerization products and to give evidence of encapsulation. TEM measurements were performed on a Philips 120 microscope. One drop of the suspension, diluted into water, was placed on a copper grid and allowed to dry in air before observation. The polymer molecular weights were determined by gel permeation chromatography (GPC) analysis with a Waters 600 apparatus equipped with Shodex columns and a Waters R410 refractometer detector using polystyrene standards.
RESULTS AND DISCUSSION Synthesis of bare polystyrene latexes
It is known from previous works on miniemulsion polymerization mechanisms that the presence of small amounts of an appropriate costabilizer is a prerequisite condition to the obtention of stable miniemulsion droplets.13 The substances usually reported in the literature are cetyl alcohol,14 hexadecane,15 polymers,16 alkyl mercaptan17 and alkyl methacrylates,18 among others. Although different mechanisms may be involved in the stability of miniemulsions, the main action of the above compounds is to build up an osmotic pressure in the monomer droplets which counterbalances the Laplace pressure inside the original emulsion, and retards the Ostwald ripening effect characterized by diffusion of monomer from small droplets to larger ones.13 The choice of a convenient costabilizer is therefore determining in the formulation of miniemulsions. To this goal, screening experiments were performed in absence of organic pigment using various types and concentrations of hydrophobe while keeping all other experimental parameters constant (runs 1–4, Table 1). The different hydrophobes used were hexadecanol, hexadecane and a low molecular weight polystyrene prepolymer (Mn = 35 220 g mol 1 and Mw = 65 600 g mol 1). In order to determine the experimental conditions giving the smallest droplet size and the best miniemulsion stability, the formulations were 544
compared on the basis of the size and polydispersity of the miniemulsion before and after polymerization. Indeed, in an ‘ideal’ miniemulsion process, the monomer droplets and the latex particles are expected to have about the same size since the former are converted into the latter. In addition, the droplets must also have small diameters and preferentially (but not necessarily) display a narrow size distribution. The results are summarized in Table 1. It is initially obvious that the addition of small amounts of an hydrophobe significantly decreased both the size and size distribution of the droplets (runs 1 and 2). It also clearly appears that the larger the amount of costabilizer (runs 1–3), the better the control of the miniemulsion stability. While large diameters and size distributions were measured for low hydrophobe concentrations, smaller and more regular droplet sizes were obtained on increasing the amount of cetyl alcohol. None of the formulations listed in Table 1, however, gave rise to the idealized situation of a 1:1 copy of the monomer droplets into polymer particles (the particle number corresponds to at least twice the number of droplets). An attempt to replace the alcohol by hexadecane (run 4) did not yield a better control of nucleation. Similar observations have been reported in the literature in case of styrene monomer and hexadecanol as the hydrophobe.19 It was suspected that the surfactant and the costabilizer formed a complex at the monomer droplet/water interface thus providing a barrier to radical entry. In a recent work, however, Landfester and co-workers gave experimental evidence for the preservation of particle identity in miniemulsion polymerization by a combination of small-angle neutron scattering, conductivity and surface tension measurements under experimental conditions very similar to ours.20 It is questionable, therefore, whether the DLS technique is appropriate for accurate determination of droplets size in the present system. It might be argued that dilution can affect the miniemulsion stability. We think, nevertheless, that if stable miniemulsions are really obtained, it must be possible to determine their size provided that dilution is performed with a saturated aqueous surfactant solution (which was indeed the case) in order to overcome diffusion of surfactant and monomer molecules from the droplets to the water phase. In other words, stable miniemulsions should theoretically be able to withstand dilution at least in the short period of time necessary to measure their size. According to this, and since we are confident in the DLS technique to estimate both droplet and particle sizes, the poor control in particle nucleation addressed above, is presumably more to be related to the miniemulsion preparation method used in the present work. One minute sonication time is perhaps not sufficient to split all the emulsion droplets and reach an osmotically stable steady state. Although it is clear that this point could easily be checked and Polym Int 52:542–547 (2003)
Encapsulation of phthalocyanine blue pigment into polystyrene
improved, in the following work all the miniemulsions were prepared according to this procedure in order to compare them to the preliminary experiments. Synthesis of pigmented polystyrene latexes
In a second set of experiments, the organic pigment was introduced in the miniemulsion formulation. The pigment was first conveniently dispersed in the monomer phase upon agitation (runs 5, 8 and 9) or with the aid of ultrasounds (runs 6 and 7). Since the success of encapsulation was expected to be strongly dependent on the quality of pigment dispersion, we paid special attention to this operation. A fine organic dispersion of pigment in styrene was achieved with ultrasound as judged by the homogeneous aspect of the suspension and the absence of aggregates on the flask wall. In contrast, some agglomerates were identified when the pigment was incorporated under magnetic stirring. The suspension stability was estimated by following pigment sedimentation. The more stable the suspension, the less important the height of the settled phase. Although, in both cases, small amounts of sediments slowly accumulated upon standing for several hours, they instantaneously redispersed with agitation. It is worthwhile to notice here, that attempts to suspend the copper phthalocyanine pigment in other monomers (eg methyl methacrylate, butyl acrylate and vinyl acetate) were unsuccessful. It was presumed that the aromatic structure of the phthalocyanine metal complex displayed favourable p–p interactions with the monomer molecules. Styrene was consequently the only monomer investigated in the current work. In a preliminary experiment, we checked whether a small amount of the organic pigment could play the role of an hydrophobe and conveniently stabilize the miniemulsion. It can be clearly seen from Table 1 that the droplet diameter significantly decreased in the presence of pigment (Ddroplets = 220 nm in run 5) in comparison to the miniemulsion prepared without pigment and costabilizer (Ddroplets 800 nm in run 1). In this work, therefore, the possibility of generating stable miniemulsions with the organic pigment alone is important. In order to account for this result, we determined the pigment solubility in the monomeric phase (see Experimental). The blue dye concentration in styrene was found to be equal to 10 g l 1 at saturation. We can assume, therefore, that the solubilized blue pigment contributes to retardation of Ostwald ripening by increasing the osmotic pressure in the monomer droplets. The broad size distribution of the monomer droplets should, however, be noticed. This might be related to the pigment dispersion quality mentioned above. Indeed, the better the quality of the pigment dispersion, the narrower the miniemulsion droplet size distribution. It can be assumed that pigment agglomerates are too large to accommodate the small monomer droplets, and thus contribute to an increase in droplet size and droplet size distribution. Unfortunately, an attempt to determine the size of the Polym Int 52:542–547 (2003)
pigment once dispersed in the monomer phase by DLS was unsuccessful because of strong light absorption under these conditions. As a direct consequence of the osmotic effect of the solubilized pigment in the monomer droplets, the miniemulsion could be converted further into a stable pigmented polymer suspension which displayed an intense blue coloration. TEM analysis indicates that the pigment particles have been successfully entrapped inside the latex beads giving experimental evidence of monomer droplets nucleation mechanism. However, DLS measurements indicate a number of polymer particles larger than the initial number of droplets (Ndroplets/Nparticles 0.1, see Table 1). As shown before, in the case of unpigmented formulations, this result suggests the occurrence of a renucleation mechanism. If we assume that the miniemulsion droplets are not sufficiently stabilized against diffusion degradation, only a fraction of them (the smallest ones) is expected to be nucleated while, concurrently, some monomer molecules can diffuse out of the unstable droplets through the water phase and generate new polymer particles via homogeneous nucleation. It is worth pointing out here that the decrease in the overall surface area of the destabilized droplets promotes surfactant desorption, and thus increases the probability of particle renucleation since more surfactant is available for stabilization of the newly formed polymer precursors. The TEM image in Fig 1 shows the morphology and the particle size distribution of the so-obtained blue-pigmented latex particles. One can clearly identify spherical pure polystyrene particles with a mean diameter of around 70 nm, and particles with a distorted shape composed of small and dark spots of pigment surrounded by polymer. Estimation of the respective population number gave around one pigmented bead over 10 particles. In other words we measured one pigmented bead for 10 beads (pigmented and unpigmented). This result is consistent with DLS measurements and confirms the above assumption of a mixed mode of particle nucleation giving rise to both empty and pigmented polystyrene latexes. In order to improve the miniemulsion stability, it was decided to introduce some hydrophobe in the
Figure 1. TEM images of the pigmented polystyrene latex particles. (a) without costabilizer (run 5), and (b) in presence of hexadecane (run 8). Scale bar: 200nm.
545
S Lelu et al
Figure 2. TEM image of blue pigmented polystyrene latex particles obtained in presence of a polystyrene prepolymer (run 9). Scale bar: 200nm.
original formulation (runs 6–9). As expected, the size of the droplets slightly decreased while that of the polymer particles increased as a consequence of the improved stability of the miniemulsion (the ratio Ndroplets to Nparticles increased). Comparison between stirred and sonicated pigment suspensions (eg runs 7 and 8) did not indicate significantly different results. In every case, pigment encapsulation was confirmed by TEM (as illustrated on Fig 1(b) for run 8) which showed large pigmented particles whose diameter (eg 200 nm) was close to that of the primary emulsion droplets (eg 189 nm, determined by DLS, see Table 1). Empty pure polystyrene particles were still present next to the pigmented latexes but they were in proportion less numerous than when the pigment was introduced alone in agreement with the DLS data (we measured around one pigmented bead over seven particles). Although the organic pigment can provide sufficient stability to pigmented miniemulsions and be encapsulated satisfactorily by this technique, the above results clearly show that the stability of the droplets can be further improved by addition of a small amount of a more efficient hydrophobe. In a last experiment, we checked whether we could achieve a better control of both miniemulsion stability and polymer particles nucleation by using a polystyrene prepolymer as the hydrophobe (run 9). Low molecular weight polymeric compounds are known to provide increased stability to miniemulsions.21 Another clear advantage of the prepolymerization reaction is the possibility of modifying the surface of the pigment and improving its dispersion in the monomer phase. The results in Table 1 indicate a slight decrease in droplet diameter while the particle size increases concurrently, which suggests a better control of nucleation in comparison to hexadecane or hexadecanol costabilizers. Figure 2 shows the resulting morphology of the particles. We again confirm the formation of pigment-containing polymer particles but 546
the morphology of the particles is in that case significantly different from the one described previously. As shown in Fig 2, the pigment particles are contained into polymer pockets partly recovering smaller empty polystyrene latexes (which can be identified by their low contrast and their spherical shape). In order to explain this morphology, one must think about a possible phase separation occurring in course of polymerization between the polystyrene prepolymer surrounding the blue pigment and the newly formed polystyrene chains. Phase separation may be induced, for instance, by differences in molecular weights or polymer characteristics. Some polymer chains might have been grafted or physically anchored on the surface of the pigment during the prepolymerization reaction. It was our goal in this prepolymerization step to provide physically entangled polystyrene molecules with strong interactions with the surface of the pigment. A last comment must be addressed at this point. The low conversions observed in some of the pigmented formulations listed in Table 1, along with the narrow molecular weight distribution of the polystyrene prepolymer obtained in presence of pigment, suggest that the phthalocyanine metallic complex (see Scheme 1) can exert a certain degree of control in analogy with transition-metal catalysts employed in living radical polymerizations, and contribute first to a decrease in the overall conversion, and second to a control in molecular weights and molecular weight distributions. Although further work is needed to attest the above assumptions, our data gave several indications that the organic pigment plays some active role in the radical polymerization process.
CONCLUSION
We showed by TEM that encapsulation of an organic phthalocyanine blue pigment could be achieved under very simple experimental conditions. Not only did the pigment allow stabilization of the miniemulsion, but it could also be embedded satisfactorily into the polystyrene latex particles. In order to check the preservation of the identity of the particles during polymerization, and to improve the miniemulsion formulation, the droplet and particle sizes were determined by DLS and their respective numbers were compared. Despite the uncertainty of the DLS technique in providing accurate droplet size measurements, the method appeared to give reliable data with good correlation with TEM observations. In a given series, the lower the droplet size and size distribution, the better was the control of nucleation. We showed by DLS and we confirmed by TEM, that addition of a small amount of an hydrophobe was helpful in increasing the droplet stability in comparison with polymerizations performed without costabilizer. This allowed the formation of pigmented latexes with different sizes and morphologies. Despite the success of encapsulation whatever the recipe used for polymerization, it is clear that additional work should be Polym Int 52:542–547 (2003)
Encapsulation of phthalocyanine blue pigment into polystyrene
done to obtain better control and a better understanding of the nucleation mechanism in the present system. The organic pigment obviously plays a determinant role in the control of interfacial properties and an effort could be made for instance to characterize the surfactant partition between the pigmented monomer droplets and the continuous phase. The nature and concentration of the surfactant is of major importance and could be perhaps better adapted to the present formulation. Pigment dispersion in the monomer phase is also a critical step. Although an acceptable level of dispersion has been achieved in the present work, better results could be surely obtained if the surface of the pigment were modified prior to polymerization in order to increase stability and pigment loading. The contribution of the pigment particles to monomer conversions and polymer chains growth also needs to be clarified. Work is currently underway to investigate more in depth the fundamental aspects associated with the miniemulsion formulation and to understand the role of the organic pigment in the control of the miniemulsion stability and polymerization kinetics.
ACKNOWLEDGEMENT
The authors are very grateful to the referees for their useful recommendations and constructive remarks.
REFERENCES 1 Caruso F, Adv Mater 13:11–22 (2001). 2 (a) Bourgeat-Lami E, in Dendrimers, Assemblies and Nanocomposites, ed by Guyot A and Arshady R, MML Series, Vol 5, Chapter 5, p 150, Citus Books, London (2002); (b) BourgeatLami E, J Nanosci Nanotechnol 2:1 (2002). 3 (a) Templeton-Knight RL, Chem Ind 16:512 (1990); (b) Templeton-Knight RL, J Oil Colour Chem Assoc 731:459 (1990).
Polym Int 52:542–547 (2003)
4 (a) Caris CHM, van Elven LPM, van Herk AM and German AL, Br Polym J 21:133 (1989); (b) Caris CHM, van Herk AM and German AL, in 20th FATIPEC Conference Proceedings, Nice, France, p 325 (1990). 5 (a) Janssen RQF, van Herk AM and German AL, J Oil Colour Chem Assoc 11:455 (1993); (b) Janssen RQF, van Herk AM and German AL, in 22th FATIPEC Conference Proceedings, Vol 1, Budapest, Hungary, p 104 (1994). 6 (a) Bourgeat-Lami E and Lang J, J Colloid Interf Sci 197:293 (1998); (b) Bourgeat-Lami E and Lang J, J Colloid Interf Sci 210:281 (1999); (c) Corcos F, Bourgeat-Lami E, Novat C and Lang J, Colloid Polym Sci 277:1142 (1999). 7 Sondi I, Fedynyshyn TH, Sinta R and Matijevic E, Langmuir 16:9031 (2000). 8 Landfester K, Macromol Rapid Commun 22:896–936 (2001). 9 Soula R, Saillard B, Spitz R, Claverie J, Llauro MF and Monnet C, Macromolecules 35:1513–1523. 10 (a) Tiarks F, Landfester K and Antonietti M, Macromol Chem Phys 202:51 (2001); (b) Bechthold N, Tiarks F, Willert M, Landfester K and Antonietti M, Macromol Symp 151:549 (2000); (c) Tiarks F, Landfester K and Antonietti M, Langmuir 17:5775 (2001). 11 (a) Erdem B, Sudol ED, Dimonie VL and El-Aasser M, J Polym Sci, Polym Chem 38:4419 (2000); (b) J Polym Sci Polym Chem, pp 4431 (2000); (c) J Polym Sci Polym Chem, pp 4441 (2000). 12 Hoy KL, Glancy CW and Lewis JMO, Eur Pat Appl 0 392 065 A1 (1989). 13 Landfester K, Bechthold N, Tiarks F and Antonietti M, Macromolecules 32:5222–5228 (1999). 14 Choi YT, El-Aasser MS, Sudol ED and Vanderhoff JW, J Polym Sci Polym Chem 23:2973 (1985). 15 Delgado J, El-Aasser MS and Vanderhoff JW, J Polym Sci, Polym Ed 24:861 (1986). 16 Miller CM, Sudol ED, Silebi CA and El-Aasser MS, Macromolecules 28:2754 (1995). 17 Mouran D, Reimers J and Schork JF, J Polym Sci, Polym Chem 34:1073 (1996). 18 Chern CS and Chen TJ, J Colloid Polym Sci 275:546 (1997). 19 Miller CM, Blythe PJ, Sudol ED and El-Aasser MS, J Polym Sci, Polym Chem 32:2365 (1994). 20 Lanfester K, Bechthold N, Fo¨rster S and Antonietti M, Macromol Rapid Commun 20:81 (1999). 21 Blythe PJ, Morrison BR, Mathauer KA, Sudol ED and El-Aasser MS, Langmuir 16:898 (2000).
547
