5904
Langmuir 2006, 22, 5904-5910
Luminescence Techniques and Characterization of the Morphology of Polymer Latices. 3. An Investigation of the Microenvironments within Stabilized Aqueous Latex Dispersions of Poly(n-butyl methacrylate) and Polyurethane I. Soutar† and L. Swanson* The Polymer Centre, School of Physics and Chemistry, Lancaster UniVersity, Lancaster LA1 4YA, U.K.
T. Annable,‡ J. C. Padget, and R. Satgurunathan§ Zeneca Resins, P.O. Box 8, Runcorn, Cheshire WA7 4QD, U.K. ReceiVed February 8, 2006. In Final Form: April 4, 2006 Fluorescence techniques (including time-resolved anisotropy measurements, TRAMS) have been used to probe differences in morphology between two stabilized aqueous latex dispersions (poly(n-butyl methacrylate), PBMA, and polyurethane, PU). Use of the emission characteristics of probes such as pyrene and phenanthrene dispersed within particles reveals that the PU latices are more heterogeneous in nature: evidence exists, particularly from quenching measurements and TRAMS, that voids and channels of water permeate the PU structure, resulting in a relatively soft, open particle, swollen by ingress of the bulk aqueous phase. Fluorescence measurements indicate that PBMA colloids, however, are composed of relatively hard, hydrophobic particles. In addition, TRAMS are considered to be a valuable tool both for probing the morphological characteristics of such dispersions and in estimating the average particle size.
Introduction Over a number of years, luminescence techniques have proven invaluable as a tool with which to probe the domains of watersoluble polymers. (See, for example, references 1 and 2 and references therein.) Such investigations have yielded information regarding macromolecular behavior in aqueous media and upon the effects of interactions with various species such as salts,3,4 colloids,5 and other macromolecules.2,6-10 The success achieved in such fundamental studies has prompted the use of fluorescence spectroscopy in ever more complex and technologically important11-16 systems, resulting in exposure of the technique to more diverse scientific interests. * Corresponding author. E-mail:
[email protected]. Present address and address for correspondence: Department of Chemistry, University of Sheffield, Brook Hill, Sheffield S3 7HF, U.K. † Deceased. ‡ Present address: Avecia Specialities, Hexagon Ho, Old Market Street, Blackley M9 8DE, U.K. § Present address: DSM NeoResins, Sluisweg 12, P.O. Box 123, Waalwijk, Tge, The Netherlands. (1) Ghiggino, K. P.; Tan, K. L. In Polymer Photophysics; Phillips, D., Ed.; Chapman and Hall: London, 1985; Chapter 7. (2) Soutar, I; Swanson, L. ACS Symp. Ser. 1995, 598, 388. (3) Branham, K. D.; McCormick, C. L. ACS Symp. Ser. 1995, 598, 551. (4) Chee, C. K.; Rimmer, S.; Rutkaite, R.; Soutar, I.; Swanson, L. J. Photochem. Photobiol., A, in press. (5) Soutar, I.; Swanson, L.; Wallace, S. J. L.; Ghiggino, K. P.; Haines, D. J.; Smith, T. A. ACS Symp. Ser. 1995, 598, 363. (6) Morawetz, H. In Photophysical Tools in Polymer Science; Winnik, M. A., Ed.; NATO Advanced Study Institute Series; Reidel: Dordrecht, The Netherlands, 1986; p 85. (7) Arora, K. S.; Turro, N. J. J. Polym. Sci., Polym. Phys. Ed. 1987, 25, 243. (8) Iliopoulos, I.; Hilary, J.; Audebert, R. J. Polym. Sci., Polym. Chem. Ed. 1988, 26, 275. (9) Heyward, J. J.; Ghiggino, K. P. Macromolecules 1989, 22, 1159. (10) Soutar, I.; Swanson, L.; Thorpe, F. G.; Zhu, C. Macromolecules 1996, 29, 918. (11) Winnik, M. A. In Photophysical Tools in Polymer Science; Winnik, M. A., Ed.; NATO Advanced Study Institute Series; Reidel: Dordrecht, The Netherlands, 1986; p. 611. (12) Winnik, M. A.; Xu, H.; Satguru, R. Makromol. Chem., Macromol. Symp. 1993, 70-71, 107.
Stabilized aqueous latex dispersions based on acrylic polymers such as poly(n-butyl methacrylate), PBMA, are industrially significant14 because of their extensive use in paint formulation. Coalescence of the particles, following drying at room temperature, results in the formation of a clear film (due in part to the low glass transition temperature, Tg, of PBMA). The minimum film-forming temperature can be decreased through the addition of polyurethane, PU, particles to the PBMA dispersion.14 (This is attractive because it produces water-based paints that require a less volatile organic component, VOC, in the formulation.) In attempts to understand the film-forming process further, we have been engaged in a research program that has utilized luminescence techniques to probe coalescence phenomena at the molecular level. Because particulate morphology is considered to be an important part of the film-forming process, luminescence spectroscopy has also been used to provide more information regarding the latex structures in aqueous solution, prior to drying.16,17 In earlier publications,16,17 the emphasis was placed on investigation of the morphology of PBMA latices through the use of labeled dispersions. The findings from this work16,17 imply that the fluorescent label may fashion its own distinctive microenvironment during polymerization. This can lead to a label dependence on the fluorescence observed in certain experiments. In this article, we report upon studies of the use of fluorescence techniques to compare and contrast the morphologies of PBMA and PU via luminescent probes dispersed within the stabilized particles. It was hoped that use of dispersed probes would (13) Pankasem, S.; Thomas, J. K.; Snowden, M. J.; Vincent, B. Langmuir 1994, 10, 3023. (14) Satguru, R.; McMahon, J.; Padget, J. C.; Coogan, R. G. J. Coat. Technol. 1994, 66, 47. (15) Flint, N. J.; Gardebrecht, S.; Swanson, L. J. Fluoresc. 1998, 8, 343. (16) Soutar, I.; Swanson, L.; Annable, T.; Padget, J. C.; Satgurunathan, R. Polym. Int., in press. (17) Soutar, I.; Swanson, L.; Annable, T.; Padget, J. C.; Satgurunathan, R. Langmuir, submitted for publication.
10.1021/la060376b CCC: $33.50 © 2006 American Chemical Society Published on Web 05/18/2006
Morphology of Polymer Latices
Langmuir, Vol. 22, No. 13, 2006 5905
minimize any potential for perturbation of the colloid by the fluorescent reporter and ultimately provide information on the particle morphology. This in turn might result in a broader insight into the coalescence process involved in film formation from aqueous colloidal dispersions of PU and PBMA and their mixtures.
Results and Discussion (i) Steady-State Fluorescence Spectra and Excited-State Lifetime Measurements. The fluorescence emission spectrum of pyrene is sensitive to the polarity of the environment in which it is dispersed.18,19 In particular, the ratio of the fluorescence intensities of two vibronic bands, termed III and I, can serve as a sensor of hydrophobicity: a ratio of 1.7, for example, would be expected when pyrene is dispersed in n-pentane20 whereas in aqueous solution a value of 0.55 would be observed.21 Spectroscopists have used this property, as a probe of microenvironmental polarity, in a variety of media such as micelles,22 microemulsions,23 and microgels13,15 to detect the ingress of water, for example. The excited-state fluorescence lifetime (τf) of pyrene is also sensitive24 to the polarity of its microenvironment. Indeed, τf data, coupled with information gathered from fine structure measurements, have been used by polymer scientists in the investigation of the conformational behavior of pH-responsive, water-soluble systems.2,21,25-28 Poly(methacrylic acid), PMAA, for example, forms a hypercoiled structure at low pH that is capable of solubilizing low molar mass material.29 A III/I intensity ratio of 1.0 has been reported2,27 for pyrene solubilized within the hypercoil. The ratio decreases to 0.55 at high pH. (PMAA forms an uncoiled flexible structure at pH values greater than 6 because of repulsive interactions between carboxylate functions that serve to expand the chain. Consequently, the probe is ejected into the aqueous phase, resulting in a reduction in the III/I intensity ratio.) The III/I ratio is therefore capable of sensing the “conformational switch” of PMAA when plotted as a function of pH.2,27 This approach has proven invaluable in probing the conformational behavior of various other types of water-soluble polymer/pyrene dispersions.2,13,15,21,38,30 Figure 1a shows the fluorescence emission spectra of pyrene dispersed in water, PU, and PBMA. Clearly, changes in the vibrational fine structure are apparent upon comparison of the (18) Nakajima, A. Bull. Chem. Soc. Jpn. 1970, 43, 967. (19) Kalyanasundaram, K.; Thomas, J. K. J. Am. Chem. Soc. 1977, 99, 2039. (20) Lucas, D. M. Ph.D. Thesis, Lancaster University, 1993. (21) Ebdon, J. R.; Hunt, B. J.; Lucas, D. M.; Soutar, I.; Swanson, L.; Lane, A. R. Can. J. Chem. 1995, 73, 1982. (22) Jay, J.; Johnston, L. J.; Scaiano, J. C. Chem. Phys. Lett. 1988, 148, 517. (23) Lianos, P.; Lang, J.; Zana, R. J. Phys. Chem. 1982, 86, 4809. (24) Birks, J. B. In Photophysics of Aromatic Molecules; Wiley-Interscience: London, 1970. (25) Chu, D. Y.; Thomas, J. K. Macromolecules 1984, 17, 2142. (26) Chu, D. Y.; Thomas, J. K. J. Am. Chem. Soc. 1986, 108, 6270. (27) Olea, A. F.; Thomas, J. K. Macromolecules 1989, 22, 1165. (28) Chee, C. K.; Rimmer, S.; Soutar, I.; Swanson, L. Polymer 2001, 42, 5079. (29) Barone, G.; Crescenzi, V.; Liquori, A. M.; Quadrifoglio, F. J. Phys. Chem. 1967, 71, 2341. (30) Chee, C. K.; Rimmer, S.; Soutar, I.; Swanson, L. React. Funct. Polym. 2006, 66, 1. (31) Soutar, I.; Swanson, L.; Annable, T.; Padget, J. C.; Satgurunathan, R.; Wilkinson, T. S. To be submitted for publication. (32) Winnik, F. M. Chem. ReV. 1993, 93, 587. (33) Soutar, I.; Swanson, L. Eur. Polym. J. 1993, 29, 371. (34) Arora, K. S.; Turro, N. J. J. Polym. Sci., Polym. Chem. Ed. 1987, 25, 259. (35) Delaire, J. A.; Rodgers, M. A. J.; Webber, S. E. J. Phys. Chem. 1984, 88, 6219. (36) Muncey, R. J. Ph.D. Thesis, University of Sheffield, 2005. (37) Chee, C. K.; Rimmer, S.; Soutar, I.; Swanson, L.; Ghiggino, K. P.; Smith, T. A. Polymer 2001, 42, 2235. (38) Ludescher, R. D.; Peting, L.; Hudson, S.; Hudson, B. Biophys. Chem. 1987, 28, 59.
Figure 1. (a) Steady-state fluorescence emission spectra of pyrene dispersed in water (10-6 M) (-), PBMA (0.35 wt %; 10-5 M) (- -), and PU (0.20 wt %; 10-5 M) (- -) at 298 K. (b) Steady-state fluorescence excitation spectra of pyrene dispersed in water (10-6 M) (-), PBMA (0.35 wt %; 10-5 M) (- -), and PU (0.20 wt %; 10-5M) (- -) at 298 K.
fluorescence from the probe in an aqueous solvent to that in the colloidal dispersions: the III/I ratio of the fluorescence from pyrene increases upon progression from water to a PU dispersion to a PBMA latex. These observations imply that the latex particles are capable of solubilizing the probe into regions that are relatively hydrophobic in nature. Indeed, the ratio in PBMA (0.82) is close to that observed when pyrene is dissolved in toluene (0.9), which indicates that the probe experiences a relatively hydrophobic environment under these conditions. When dispersed in PU, a III/I value of 0.68 results. This implies that pyrene exists in a more hydrophilic microenvironment than that of PBMA. This, in turn, suggests that the occluded probe cannot access an environment in the PU dispersion that is as hydrophobic as that afforded by the polymeric components of the PBMA latex. This finding reinforces the belief, based upon fluorescence quenching data involving fluorescently labeled latices,16 that PBMA-based colloids are composed of highly viscous, hydrophobic domains. Furthermore, it can be inferred that the PU dispersions employed in the current work contain particles that are much more “open” than those of the PBMA latices and thereby are much more permeated by water from the bulk aqueous phase. This conclusion also finds support from fluorescence quenching measurements that compared the accessibilities of fluorescent labels of the PU and PBMA media to an aqueous-borne quencher (nitromethane).31 [The comparison of pyrene III/I intensity ratios is complicated by several effects. The ratio is influenced by the extent to which the probe is partitioned between the colloid (including surfactantrich regions) and the bulk aqueous phase. In a previous publication,16 it was argued that a given aromatic probe will diffuse into a colloid to “seek out” domains of maximum hydrophobicity (subject to the constraints imposed by the microviscosity of the colloid’s interior). The III/I ratio derived from the fluorescence spectrum of pyrene represents an “average environment” sensed by molecules solubilized within the colloid
5906 Langmuir, Vol. 22, No. 13, 2006
Soutar et al.
and those dispersed in more polar environments both within the colloid and the bulk aqueous phase. This does not detract from the general observations expressed above: the PBMA particles present a less polar environment for their pyrene guests than is afforded by the PU species]. Further evidence for the solubilization of the probe in both the PU and the PBMA matrixes is evident from a comparison of the leading edge of the conventional emission spectral profiles of pyrene dispersed in the various media (cf. Figure 1a). A red shift (ca. 2 nm) in the fluorescence is discernible in both of the polymer colloid samples when compared to that observed in an aqueous solution of pyrene. More dramatic red shifts are apparent, however, following an examination of the corresponding steadystate excitation spectra for each of the dispersions (cf. Figure 1b). Such behavior has been observed previously in systems in which excimer formation involving the “preassociation” of pyrene occurs32 and also when probes are solubilized within the hydrophobic cavities of certain water-soluble polymers.21 In the latter instance, spectral shifts were considered21 to be indicative of guest species residing in a solvating, hydrophobe-rich medium. Indeed, previous experience21 suggests that whereas such red shifts may reflect the occurrence of “preformed” sites in systems that exhibit pyrene excimer formation32 the phenomenon is by no means restricted to such systems. Because there is no evidence of excimer formation in the current work, it can only be construed that the spectral shifts reflect the fact that both the PU and the PBMA particles are capable of sequestering the probes into relatively hydrophobic regions. Furthermore, if these red shifts were to be taken as indicators of the polarity of the probe’s microenvironment, then it would then have to be concluded that the PU particles present the pyrene with a distribution of solubilizing domains that is, on average, (marginally) more hydrophobic than that fashioned within the PBMA colloids. This contrasts with the relative order of polarity that the III/I emission intensity ratios appeared to indicate (as discussed earlier). The singlet excited-state lifetime (τf) of pyrene depends on the hydrophobicity of the medium in which it is dispersed.24 Consequently, it is capable of acting as a polarity sensor. This property has been used to study the conformational behavior of certain water-soluble polymers.2,15,21,25-28,30 (In the compact form of PMAA, for example, the excited state of the probe is longlived,21 reflecting the occlusion of pyrene into a hydrophoberich domain, where it is protected from the deactivating effects of the aqueous phase. In a similar manner, the fluorescence lifetime of pyrene should be capable of gauging the solubilizing capability and hydrophobicity of the two latex dispersions currently under investigation). Fluorescence lifetime measurements were made on pyrene dispersed in water and within the PU and PBMA particles. (In addition, τf data were also recorded when phenanthrene was dispersed in water and in each of the latex dispersions to obtain further information regarding the morphological characteristics of the stabilized particles.) Perhaps somewhat surprisingly, considering the diverse range of environments that might be imagined to exist in these systems, the transient emission from each probe/dispersion combination could be adequately described, on statistical grounds, through use of a single-exponential function of the form
( )
I(t) ) Io exp -
t τf
(1)
where τf is the excited-state lifetime. Figure 2 shows an example of a fluorescence decay curve for pyrene dispersed in PBMA,
Figure 2. Fluorescence decay of pyrene dispersed in PBMA (0.35 wt %; 10-5 M), “best fit” to a single-expeontial model and resultant residuals. (λex ) 335 nm; λem ) 380 nm). Table 1. Bimolecular Quenching Constants and Excited-State Lifetime Data Obtained for the Various Dispersions probe
solvent
τf/n
kq/mol-1 dm3 s-1
pyrene phenanthrene pyrene phenanthrene pyrene phenanthrene pyrene
water water PBMA PBMA PU PU PMAA (10-3 wt %) pH 3
134 ( 2 36 ( 1 306 ( 2 53 ( 1 311 ( 2 50 ( 1 257 ( 2a
8.0 × 109 7.0 × 109 1.3 × 107 4.2 × 107 5.9 × 107 1.0 × 108 2.2 × 107
a 〈τ〉 value calculated from eq 5 following dual exponential analysis to the fluorescence decay curve.
with the associated single-exponential fit and statistics. The low value of χ2 (χ2 should be close to unity for a good fit), and the randomness of the residuals provide statistical confidence in the quality of the fit. The τf data, consequent upon such analyses, are listed in Table 1 for each of the various dispersions. By reference to Table 1, it is evident that the excited-state fluorescence lifetime of pyrene increases from ca. 130 ns when dissolved in water to ca. 310 ns when dispersed in either of the stabilized latex particles. Such behavior is consistent with the uptake of the probe into more protective (than that of the aqueous phase) hydrophobic regions within the polymeric colloids. Interestingly, the τf data are not capable of distinguishing a PU particle (τ ) 311 ( 2 ns) from a PBMA particle (τ ) 306 ( 2 ns). This behavior contrasts with both the spectral shift data and with the vibrational fine structure observations discussed earlier. Further examination of Table 1 reveals that the excitedstate lifetime of phenanthrene also increases on going from an aqueous environment (ca. 36 ns) to that extant within each of the latex dispersions (ca. 50 ns). Clearly, the probe enjoys a greater degree of protection from the quenching effects of the aqueous phase within both PBMA and PU. This implies that the latex particles are capable of solubilizing and sequestering phenanthrene into a medium that is less polar than water. To provide further information regarding the microenvironment as sensed by the probe, pyrene was dispersed (10-6 M) in PMAA (10-3 wt %) in the current work. (It was hoped that the pyrene/ PMAA dispersion would act as a benchmark to aid in the comparison of the two polymer colloid samples). In contrast to the behavior observed in the latex particles, the decay of fluorescence from pyrene/PMAA could be adequately modeled only through the use of a double-exponential equation of the form
Morphology of Polymer Latices
( )
I(t) ) Io1 exp -
Langmuir, Vol. 22, No. 13, 2006 5907
( )
t t + Io2 exp τf1 τf2
(2)
Two lifetimes of ca. 113 and ca. 356 ns were derived as a consequence of this form of analysis. The short lifetime is close to that observed in water (cf. Table 1) and presumably reflects a group of pyrene molecules that populate the aqueous phase. The longer-lived component will, most likely, originate from a distribution of probes that are buried deep within the hypercoil of PMAA. Interestingly, this lifetime (ca. 356 ns) is longer than that observed in both the PBMA (ca. 306 ns) and PU (ca. 311 ns) dispersions. It is not clear at this time why the lifetime of the probe in the compact form of PMAA is so much longer than that in each of the particles: it may merely be an effect of parametrization of the data because the lifetime from the singleexponential analysis of pyrene/PMAA (and that of the average τ, 〈τ〉, see later) is ca. 275 ns. Fitting the decay data with a dual-exponential function recovers, obviously, one longer and one shorter component. Whether these lifetimes can be associated with distinct populations of pyrene in the aqueous phase and solubilized in PMAA is open to debate. There are arguments from other water-soluble polymer/probe/label investigations that suggest28,30 that the heterogeneous nature of the systems may be better described by a distribution of label/probe sites requiring a more complex function than that of a dual-exponential model. Provided that we are confident of the validity of the longer-lived component in the pyrene/PMAA decay (ca. 356 ns), it could be that any surfactant that is present within the latices might either have the effect of quenching the fluorescence of the probe or of plasticizing the medium to produce a more fluid environment. (Indeed, anisotropy evidence exists that indicates that residual surfactant renders the cilia less viscous than other areas within colloidal PBMA.17) Both of these effects would serve to reduce the τf of the probe. However, fluorescence quenching studies of labeled polymer systems indicate that the fluidities of the PU interior and PMAA hypercoil are similar.31 The interiors of PBMA latex particles, however, seem to be much more viscous.17 A final comment upon the observation that the time-resolved fluorescences from the probe/colloid dispersions can be modeled using a single-exponential decay function seems warranted. This implies that the solubilization of pyrene by the colloidal particles is virtually complete. (If “free” unsolubilized probe remained in the aqueous phase, then a second shorter-lived component (of ca. 120 ns) might be expected in the fluorescence decay.) The fact that two distinct populations of pyrene excited states are evident in the transient fluorescence from aqueous solutions of PMAA is not directly related to the solubilization capacities of PU, PBMA, and PMAA. The amount of polymeric host present in the colloidal dispersions employed in these spectroscopic studies is 200-300 times greater than that in the PMAA solutions. In the latter, high dilutions were employed to ensure that the hydrophobic domains under investigation were those formed intramolecularly within isolated polymer chains. Consequently, the τf data are likely to reflect the fact that there is simply more colloid present (whether PMBA or PU) to uptake the organic material from the aqueous phase. However, it can be remarked upon that the population of probes that are solubilized in PMAA with a τf of ca. 356 ns reside in an extremely protective environment: indeed, the hypercoil conformation can sustain a longer lifetime of pyrene than in either of the colloidal samples studied in the current work. (ii) Fluorescence Quenching Experiments. The accessibility of a fluorescent molecule, F, to a low molar mass species, Q, (which is capable of accepting energy from an excited state) can be estimated from quenching experiments. The general form of
the process is outlined in eq 3:
F* + Q f F + Q*
(3)
The rate at which a dynamic quencher accesses the excited state is governed by the Stern-Volmer equation
τ f° ) 1 + kqτf°[Q] τf
(4)
where τf is the fluorescence lifetime in the presence of some concentration of quencher [Q], τf° is that in the absence of Q, and kq is the bimolecular quenching constant. (Hence kq is a measure of the ease of access of Q to F*.) If a fluorescently labeled polymer is used, then information regarding the “openness” or compactness of the chain can result from such experiments.1,2,7,25,28,30 Quenching measurements on probes dispersed in various media1,13,15,27 can provide, in principle, information regarding the fluidity or microviscosity of the resulting environment. In an attempt to gain further insight into to the morphologies of the current latex samples, quenching experiments were carried out using two different fluorescent probes: pyrene and phenanthrene. Fluorescence decay data from each probe, dissolved in water and dispersed within each of the latex systems, were acquired in the presence of various concentrations of nitromethane. (Nitromethane has been extensively used as a quencher in investigations of water-soluble polymer and dispersible systems.1,2,13,15,33-35) For all Stern-Volmer experiments, the transient behavior of each of the probes studied in the current work was modeled in terms of a single-exponential function of the form of eq 1. The resultant kq values are listed in Table 1. Reference to Table 1 reveals that the efficiency of quenching is very high for both pyrene and phenanthrene, when dispersed in water. (A bimolecular quenching constant of ca. 7.5 × 109 mol-1 dm3 s-1 results.) This value is close to that expected for a diffusion-controlled process in aqueous media and implies that the probe is (i) in a very fluid environment and (ii) within a very “exposed” or accessible phase. A dramatic decrease in the quenching efficiency is apparent, when compared to that in water, following the dispersion of both pyrene and phenanthrene in PBMA. (The respective kq values are also listed in Table 1.) The relative magnitudes of the quenching constants imply that nitromethane experiences greater difficulty in accessing the probe when occluded in the acrylic latex than in water. This implies that the fluidity of the medium, as sensed by each of the fluorophores, decreases markedly upon solubilization within PBMA. These data are consistent with a model that envisages the interiors of the acrylic particles as viscous, protective, hydrophobic domains. The dispersion of both pyrene and phenanthrene in a PU latex results in kq values on the order of (0.6-6) × 108 mol-1 dm3 s-1 (cf. Table 1). This implies that the quenching of the probe’s excited-state lifetime is more efficient in PU than in a PBMA particle. Presumably, the dominant effect leading to this enhancement of the kq value is that the particle interiors of the PU dispersion are much less viscous than those of the acrylic latex. (In a previous publication,16 it was argued that nitromethane is equipartitioned between the acrylic colloid and bulk aqueous phase. This is also likely to be the case for the PU particles.) The picture that evolves for the PU colloid is that of a relatively soft, water-permeated, irrigated structure of enhanced microfluidity compared to the harder and more homogeneous PBMA cores.
5908 Langmuir, Vol. 22, No. 13, 2006
Soutar et al.
An interesting comparison can be made upon further examination of the data in Table 1: when pyrene is dispersed in PMAA, at pH 3, a kq value on the order of 2 × 107 mol-1 dm3 s-1 results, which is very close to that in the PBMA latex. This implies that the quenching experiment reflects a similar degree of access in the acrylic particle to that in the hypercoiled conformation of PMAA. [It should be noted that with the current pyrene/PMAA quenching data the fluorescence decays, recorded upon the addition of various concentrations of nitromethane, were modeled in terms of a double-exponential function of the form of eq 2. To treat the τ data in the Stern-Volmer analysis, an average lifetime, 〈τ〉, was estimated according to
〈τ〉 )
∑ ∑Biτi
Biτ2i
(5)
for each nominal concentration of quencher. (〈τ〉 in the absence of nitromethane for pyrene/PMAA is listed in Table 1)]. A caveat should be noted here: the rates at which fluorescence from the probe is quenched by nitromethane, in these media, are significantly reduced compared to those of fluorescently labeled dispersions of the same type. The probe data are reduced relative to those of the labeled system by a factor of between 4 (for PBMA) and 20 (for PMAA). (Naphthyl labels of PBMA17 or PU31 colloids and PMAA33 in its hypercoiled state exhibit kq values on the order of 5 × 107 mol-1 dm3 s-1.) Clearly, the label is located at a site within a particular polymer environment that is created during polymerization or during coil collapse for PMAA. The probe, however, enters the solubilizing domains of the various water-borne polymers and will naturally diffuse into the most hydrophobic sectors that it can access. In so doing, it is likely that, in hydrophobically interacting with its surroundings, the fluorescent reporter will, in turn, alter the morphology of its immediate vicinity. (There is some photophysical evidence that suggests that organic probes can effect changes in the dynamic, hydrophobic domains created in certain water-soluble polymers and their modifications,36 for example.) Whether the differences between the quenching of fluorescence from labels and probes originate solely from the probe’s propensity to seek out the most hydrophobic and therefore the most inaccessible regions of a matrix or result from the fluorescent reporter altering the nature of its microenvironment is not clear at present. (Spectroscopic data suggest that the label, too, may have some potential17 for the partial creation of its local habitat.) What is certain is that it cannot be assumed that the spectroscopic and quenching characteristics of the probe necessarily furnish information immediately relevant to considerations of the particle interiors of unperturbed latices. Time-Resolved Anisotropy Measurements (TRAMS). The observed anisotropy r(t) can be constructed from its fluorescence components via the relationship
r(t) )
I|(t) - I⊥(t) I|(t) + 2I⊥(t)
(6)
where I|(t) and I⊥(t) are the time-dependent intensities of emission detected in planes parallel and perpendicular, respectively, to that of vertically polarized excitation. If the decay of anisotropy is due to a simple, single relaxation process, then r(t) can be described by eq 7
( )
r(t) ) ro exp -
t τc
(7)
Figure 3. Decay of anisotropy, [r(t)], of pyrene dispersed in PBMA (0.35 wt %; 10-5 M) and the associated single-exponential fit with the distribution of residuals. (λex ) 335 nm; λem ) 380 nm).
where ro is the intrinsic anisotropy of the fluorescent species under study and τc is the correlation time. If a fluorophore is dispersed within a relatively fluid environment, τc is then a measure of the speed of rotation of the probe, which can subsequently be related to the microviscosity of the medium. In attempts to gain further insight into the morphologies of the current stabilized latex dispersions, TRAMS were carried out on pyrene dispersed in both PBMA- and PU-stabilized particles, respectively. Figure 3 shows the decay of fluorescence anisotropy from a pyrene/PBMA dispersion and the associated single-exponential “best” fit. The randomness of the residuals and the low value of χ2 provide statistical confidence in the quality of the model. A value for τc on the order of 6 µs results from this form of analysis. The fact that r(t) follows a simple first-order decay law implies that it is likely that the probe is located in a single type of environment. (If the PBMA particle was markedly heterogeneous in nature, then it would be expected that pyrene would be partitioned into regions of differing fluidity, which would result, in turn, in differing extents of motion of the probe and resultant complexity in anisotropy decay behavior). These observations provide further evidence (in support of that from the spectroscopic and quenching information, discussed earlier) that the interiors of PBMA colloids are relatively homogeneous in their constitution. For a sphere rotating in a medium of viscosity η, τc is related to its molar volume, V, by eq 8
τc )
ηV RT
(8)
where T is the temperature and R is the gas constant. Given a viscosity of 1.002 × 10-3 kg m-1 s-1 for water at 298 K, a τc of 6 µs would characterize the rotational relaxation of a spherical particle with a diameter of 36 nm. This estimate is just under half that (84 nm) derived from light scattering measurements (cf. Table 2). This is perhaps a reasonable estimate given the limitations of the experiment: the τf of pyrene is ca. 300 ns when dispersed in the PBMA latex and so is not well matched to the rate of motion under study. Time-resolved phosphorescence anisotropy measurements of pyrene would move the time scale of the experiment into the millisecond regime and consequently may be more appropriate for such experiments. However, it could be that limited motion of the probe occurs over and above that of whole particle rotation. This would have the effect of superimposing the two to give a composite motion that would be reflected in a reduction in the τc value observed and the resultant particle size derived via eq 8. (Indeed, anisotropy evidence exists from labeled acrylic latices17 that suggests that
Morphology of Polymer Latices
Langmuir, Vol. 22, No. 13, 2006 5909
Figure 4. (a) Decay of anisotropy, [r(t)], of pyrene dispersed in PU (0.20 wt %; 10-5 M) and the associated single-exponential fit with the distribution of residuals. (λex ) 335 nm; λem ) 380 nm). (b) As in part a but fit to a double-exponential decay model. Table 2. Polymer Characterization Data system PBMA PU PMAA
particle size solids (( 5 nm) (wt %) Mn × 103 g mol-1 Mw × 103 g mol-1 84 62
35 20
109 12 150
375 3.6 285
plasticization of the particles via residual surfactant may occur). Suffice it to say that under the experimental conditions adopted TRAMS could contain a contribution from motion from a probe that resides in a relatively viscous medium over and above that from whole latex particle rotation. The dispersal of pyrene in a PU particle results in more complex anisotropy behavior: Figure 4a shows an example of a singleexponential fit, of the form of eq 7, in an effort to describe r(t). The nonrandom distribution of the residuals and the high χ2 value indicate that such a function is an inappropriate model for the decay of the anisotropy of pyrene under these conditions. In the current work, the statistical description of r(t) was adequate only when a function of the form
( )
r(t) ) ro1 exp -
( )
t t + ro2 exp τc1 τc2
(9)
was adopted as the model. Figure 4b shows the improvement in the quality of fitting statistics achieved by the use of eq 9: a random distribution of residuals around zero occurs coupled with a χ2 value of close to unity results from such an analysis. Two correlation times (ca. 68 ns and ca. 2 µs, respectively) are derived as a consequence of modeling r(t) to eq 9. Presumably, the complexity of the decay of anisotropy reflects the fact that the PU particles exhibit a more heterogeneous morphology compared to that associated with a PBMA latex. This would be consistent with the ingress of water into the PU colloid creating channels and voids of a more polar nature within a relatively hydrophobic matrix. This would have the effect of creating a water-swollen particle with “spongelike” characteristics. The longer τc is
consistent with a pyrene population that is encapsulated within hydrophobic, viscous microdomains. Motion sensed by the decay of anisotropy of the probe under these circumstances is dominated by τc2, which reflects whole PU particle rotation. By adopting eq 8, it is again possible to estimate an apparent size of the PU particle from the current anisotropy data: a τc of 2 µs would correspond to a sphere with a diameter of 25 nm. Again, as was observed with the PBMA latex, this value is smaller than that from light scattering measurements (cf. Table 2) and presumably reflects the mobility of the probe itself within the PU matrix. (Again, this motion would be superimposed onto that from whole particle rotation, which would bias the resulting τc to shorter times. τc1, however, could have its origins in a pyrene population that is occluded within the aqueous channels of the PU particles: because water would have a plasticising effect upon the colloid interior, the local environment would appear to be more fluid to the probe, resulting in significant rotation (on a nanosecond time scale) of the pyrene itself. In summary, the increased complexity in the fluorescence anisotropy decay kinetics of the pyrene/PU dispersion compared to that of pyrene/PBMA provides further evidence for the existence of an open, water-swollen, spongelike morphology: the TRAMS data would indicate that the PBMA latex has a more homogeous, viscous nature than the PU particle. This has important implications as far as paint technology is concerned for the film-forming process of mixtures of PBMA and PU. In a fluorescence energy-transfer study31 using PU labeled with a donor molecule and PBMA with an acceptor species, it was found that the “softer”, more open urethane phase coalesces rapidly at early times following drying. This is ensued by interdiffusion with the “hard” PBMA particles to form small acrylic-rich regions in a continuous phase of urethane. The coalescence and subsequent film morphology can consequently be viewed in terms of the PU acting31 as a softer “glue” for the PBMA particles. Finally, as an interesting comparison, the anisotropy decay was recorded from pyrene dispersed within the hypercoiled form of PMAA at pH 3 and compared to and contrasted with data from the same probe dispersed within the PBMA and PU latices, respectively. As in the case of the PU particle, r(t) from the pyrene/PMAA dispersion could be adequately described statistically only through the use of a function of the form of eq 9. Two correlation times of ca. 24 ns and ca. 1 µs were derived from such an analysis. The shorter component could result from probes that reside in a relatively fluid environment (perhaps in more polar regions of the PMAA coil). Presumably, the longer τc is associated with a distribution of pyrene molecules that are sequestered deep within the relatively viscous hypercoil of PMAA. Under these circumstances, the probes are likely to be “frozen” within the macromolecule, hence TRAMS will monitor whole coil rotation. If this is indeed the case, then following eq 8, a τc of 1 µs would yield a coil dimension of ca. 18 nm. This is not unreasonable for a polymer of the current molecular weight (cf. Table 2). [It is noteworthy that although r(t) for pyrene dispersed in PMAA at pH 3 does appear to follow a dual-exponential decay law it cannot be simply assumed that such a function is entirely appropriate and a physically meaningful model for fluorescence anisotropy behavior. For example, when pyrene is dispersed in poly(N-isopropylacrylamide), PNIPAM, above its lower critical solution temperature, evidence for more complex anisotropy behavior is apparent.37 In the case of PNIPAM, more exotic mathematical functions38 are required to describe r(t).] Because the complexities in the fluorescence anisotropy decays of the pyrene/PNIPAM dispersions have their origins in the range
5910 Langmuir, Vol. 22, No. 13, 2006
of environments throughout which the probe is sequestered,37 the current data (certainly those of PMAA) are also liable to warrant analysis using a more complex model function38 than that represented by eq 9. However, it is likely that even the more sophisticated model, adopted in studies of the PNIPAM solutions, would prove too simple to describe properly the reorientational behavior of the current complex systems. Suffice it to say that the anisotropy data and the analyses adopted are adequate for our purposes. They underline the differences between the acrylicand urethane-based dispersions in terms of the degrees of heterogeneity and ingress of water apparent in the particles’ interiors.
Conclusions (i) Fluorescence techniques (including quenching and TRAMS) are capable of sensing differences in the microenvironment between stabilized, aqueous PBMA and PU latex dispersions. The current data suggest that PU is a relatively water-swollen spongelike particle, whereas PBMA is a viscous impervious colloid. The latter observation reinforces findings from a previous study of a labeled PBMA latex.16 (ii) Simple steady-state spectra and excited-state lifetime measurements of pyrene dispersed within each of the latices imply that both PBMA and PU are more hydrophobic in nature
Soutar et al.
than water. Indeed, the fluorescence data are consistent with the solubilization of the probe into the particles. (iii) Quenching experiments on pyrene and phenanthrene probes dispersed within each of the colloids imply that the PU latices have a structure that is relatively open and water-swollen. (These measurements suggest that the ingress of water occurs readily into the PU particles, creating channels and voids that permeate the whole particle structure.) (iv) TRAMS on pyrene dispersed in PBMA and PU, respectively, provide a reasonable estimate of the particle size of each of the latex particles when compared to that from light scattering measurements. The more complex anisotropy decay behavior observed in the pyrene/PU system is considered to reflect a heterogeneous morphology that is consistent with a water-swollen spongelike structure. Acknowledgment. We thank ICI/Zeneca for providing an SRF award in support of this work. Support by the Leverhulme Trust is also gratefully acknowledged, as is technical assistance in the synthesis of the latices from E. N. J. Russell. Supporting Information Available: Experimental details. This material is available free of charge via the Internet at http://pubs.acs.org. LA060376B
