pubs.acs.org/Langmuir © 2009 American Chemical Society
Characterization of Adsorbate-Induced Ordering Transitions of Liquid Crystals within Monodisperse Droplets Jugal K. Gupta,† Jacob S. Zimmerman,† Juan J. de Pablo,† Frank Caruso,*,‡ and Nicholas L. Abbott*,† † Department of Chemical and Biological Engineering, University of Wisconsin-Madison, 1415 Engineering Drive, Madison, Wisconsin 53706, and ‡Centre for Nanoscience and Nanotechnology, Department of Chemical and Biomolecular Engineering, The University of Melbourne, Victoria 3010, Australia
Received March 4, 2009. Revised Manuscript Received May 13, 2009 The ordering of liquid crystals (LCs) within micrometer-sized droplets is known to depend strongly on the presence of interfacial adsorbates, although the exact sequence of ordered equilibrium states that accompany a change in interfacial anchoring from tangential to perpendicular has not been established. In this paper, we report use of a methodology that permits the preparation of monodisperse LC droplets in aqueous phases to investigate ordering transitions in the LC droplets that accompany the adsorption of amphiphiles at the aqueous-LC droplet interface. By using an amphiphile that undergoes reversible adsorption at the aqueous-LC interface (sodium dodecylsulfate, SDS), we identified six distinct topologically ordered states of the LC droplets as a function of increasing concentration of SDS. We exploited the reversible adsorption of the SDS to LC droplets with diameters of 8.0 ( 0.2 μm to confirm that these topological states are equilibrium ones. We also exposed LC droplets to a continuous gradient in concentration of SDS to document the continuous transitions between topological states and to confirm the absence of additional, intermediate topological states. The formation of the LC droplets as aqueous dispersions also enabled an investigation of ordering transitions in LC droplets driven by biomolecular interactions. Surprisingly, enzymatic hydrolysis of the phospholipid L-dipalmitoyl phosphatidylcholine (L-DLPC) by phospholipase A2 at the interfaces of the LC droplets was observed to trigger the same progression of topologically ordered states of the LC as was observed with SDS. Overall, the results presented in this paper resolve prior conflicting data in the literature by providing an unambiguous set of observations regarding topologically ordered states encountered in LC droplets. This paper provides a data set against which future theories and simulations of LCs can be compared to develop a fundamental understanding of the competition between volumetric and interfacial effects in droplets. The results also suggest that topological ordering transitions in LC droplets can be exploited to report interfacial enzymatic reactions.
Introduction A large number of factors have been identified in past studies to impact the topological ordering of liquid crystals (LCs) within confined systems. These factors include properties intrinsic to the LCs (such as elastic moduli),1 temperature,2,3 anchoring at the LC interface,4,5 the presence of external fields,3,6 the geometry and scale of the confinement,7-10 and the hydrodynamic flow of liquid around the LCs.11 In the absence of external fields (including external hydrodynamic fields), the ordering of LCs confined *Correspondence should be addressed to Nicholas L. Abbott (E-mail
[email protected], Fax 1 + 608-262-5434) or Frank Caruso (E-mail
[email protected], Fax +613 8344 4153). (1) Drzaic, P. S. Liquid Crystal Dispersions; World Scientific Publishing Company: Singapore, 1995. (2) Golemme, A.; Zumer, S.; Allender, D. W.; Doane, J. W. Phys. Rev. Lett. 1988, 61, 2937–2940. (3) Erdmann, J. H.; Zumer, S.; Doane, J. W. Phys. Rev. Lett. 1990, 64, 1907– 1910. (4) Goyal, R. K.; Denn, M. M. Phys. Rev. E 2007, 75, 021704–10. (5) Prishchepa, O. O.; Shabanov, A. V.; Zyryanov, V. Y. Phys. Rev. E 2005, 72, 031712–11. (6) Doane, J. W.; Vaz, N. A.; Wu, B.-G.; Zumer, S. Appl. Phys. Lett. 1986, 48, 269–271. (7) Gupta, J. K.; Sivakumar, S.; Caruso, F.; Abbott, N. L. Angew. Chem., Int. Ed. 2009, 48, 1652–1655. (8) Fernandez-Nieves, A.; et al. Phys. Rev. Lett. 2007, 99, 157801–4. (9) Lavrentovich, O. D. Liq. Cryst. 1998, 24, 117–125. (10) Tixier, T.; Heppenstall-Butler, M.; Terentjev, E. M. Langmuir 2006, 22, 2365–2370. (11) Fernandez-Nieves, A.; Link, D. R.; Marquez, M.; Weitz, D. A. Phys. Rev. Lett. 2007, 98, 087801–4.
9016 DOI: 10.1021/la900786b
within micrometer-scale systems is governed by a subtle balance of surface and volumetric effects, which is defined by minimization of the total free energy (F)1 Z F ¼
Z fb dV þ
V
fS dA
ð1Þ
S
where fS and fb are, respectively, the orientation-dependent part of the surface energy and volumetric elastic energy density. Recently, we reported on this interplay of surface and volumetric energetic effects in the context of a study of size-dependent ordering of LCs within droplets with diameters on the micrometer scale and smaller.7 The study was enabled by a new method12 that permits the synthesis of large populations of LCs droplets with precise control over size and interfacial chemistry. In brief, the method involves the layer-by-layer deposition of polymers (e.g., poly (sodium-4-styrene sulfonate) (PSS) and poly(allylamine hydrochloride) (PAH)) onto the surfaces of silica templates dispersed in aqueous solution (Scheme 1). After deposition of the polymeric layers, the silica cores are etched, and the resulting capsules are filled with low molecular weight LC. Selection of the size of the silica template and the chemistry of the polymeric shell provides the basis of the independent control over size and interfacial chemistry of the LC droplets. Our past studies7 based on this experimental system yielded two important insights into the balance (12) Sivakumar, S.; Gupta, J. K.; Abbott, N. L.; Caruso, F. Chem. Mater. 2008, 20, 2063–2065.
Published on Web 07/06/2009
Langmuir 2009, 25(16), 9016–9024
Gupta et al.
Article
Scheme 1. Schematic Illustration of the Procedure Used to Prepare LC Droplets of Predetermined Sizes within Polymeric Multilayer Shells
Polymeric shells were prepared by sequential deposition of poly(styrene sulfonate) (PSS) and poly(allylamine hydrochloride) (PAH) onto silica templates and subsequent etching of the silica. The resulting polymeric shells were filled with LC.
of surface and volumetric energetic effects in LC droplets. First, by systematically decreasing the diameters of the LC droplets from 10 μm to 700 nm (without variation in composition of the polymeric shell), we established that nematic 40 -pentyl-4-cyanobiphenyl (5CB) exhibits size-dependent ordering within the LC droplets with large droplets (diameters >3 μm) assuming a bipolar configuration, and small droplets (30 μm) nematic droplets, and changes in temperature were used to triggered the ordering transition within the droplet. Their results revealed a continuous anchoring transition from a planar to radial configuration, involving the formation of an intermediate state comprising an equatorial disclination loop. We note, however, that changes in temperature can potentially impact both the interfacial and bulk properties of LCs,1 and thus both effects may underlie these experimental observations. In a second study by Prischepa et al.,14 LC droplets were dispersed in a polymeric matrix containing lecithin (polymer-dispersed LC films), the concentration of which was varied in the polymer/LC mixture (before generation of LC droplets) to achieve a change in anchoring of the LC at the droplet surface. Interpretation of the experimental observations, however, is difficult because the distribution of lecithin within the polymer matrix was inhomogeneous. Comparisons between experimental observations and simulations of director configurations were made by Prischepa et al.5 and, contrary to the observations reported by Volovik and Lavrentovich, Prischepa et al. did not find evidence of disclination loops during the transition from the bipolar configuration to the radial configuration In contrast to these past studies,13,14 the experiments reported in our manuscript are (1) performed in a system where the LC droplets are dispersed in an aqueous phase, 9022 DOI: 10.1021/la900786b
thus allowing dynamic equilibrium between the adsorbates present in the bulk and at the interface, and (2) performed at constant droplet size and temperature, thus allowing us to assign the origins of the various topological states to changes in surface anchoring (and not changes in bulk properties of the LCs). In addition, we have established that the topological states of the LC droplets seen in our experiments correspond to equilibrium states of the system and thus can be connected by reversible pathways (as a function of concentration of SDS). Specifically, we have found that it is possible to access six stable topological states of LC droplets as a function of the surface concentration of the SDS. As discussed in the Introduction, in the second set of experiments described in this paper, we investigated the topological states encountered in LC droplets during enzymatic degradation of the phospholipid L-DLPC that was irreversibly adsorbed at the aqueous-LC interface. The choice of this experimental system was influenced in part by the results of our past studies,20,28 which established that the composition of an interface between a thin film of LC saturated with L-DLPC and an aqueous phase can be changed via enzymatic degradation of L-DLPC upon contact with the enzyme phospholipase A2 (PLA2) in the presence 5 mM CaCl2. The enzyme catalyzes the hydrolysis of L-DLPC into laurate and the corresponding lyso-phospholipid, both of which desorb from the LC-aqueous interface (into the bulk aqueous solution). Accompanying this enzymatic process is an anchoring transition of the LC from an initially homeotropic orientation (in presence of L-DLPC) to a planar orientation.20,28 To prepare the L-DLPC-coated LC droplets, we first prepared LC droplets (diameter of 8 μm) that were encapsulated by hydrogen-bonded multilayers formed from poly(methacrylic acid) (PMA) and poly (vinylpyrrollidone) (PVPON) (see Methods section for details). As reported previously,12 the PMA-PVPON shells can be disassembled from the surface of the LC droplets via a change of pH. (28) Brake, J. M.; Abbott, N. L. Langmuir 2007, 23, 8497–8507.
Langmuir 2009, 25(16), 9016–9024
Gupta et al.
These so-called “naked” LC droplets were subsequently saturated with the L-DLPC by equilibrating the LC droplets in an aqueous dispersion of L-DLPC vesicles (concentration of L-DLPC of 700 μM, doped with 0.1 mol % Texas Red-DHPE) for 1 h. The Texas Red-DHPE was added to enable fluorescence measurements. After the adsorption of lipid for 1 h, the droplets were rinsed three times in TBS buffer to remove the excess lipid from the bulk aqueous phase. The droplets were then observed under a microscope to obtain bright-field, polarized, and fluorescent micrographs. The first column of images in Figure 8 (t=0 min) correspond to bright-field, polarized light, and fluorescent images of LC droplets saturated with L-DLPC. Inspection of these images reveals that the L-DLPC-saturated LC droplet exhibited a radial LC configuration, as characterized by the single point defect located at the center of the droplet in the bright-field optical micrograph and the characteristic cross pattern in the polarized light micrograph (crossed-polars). Interestingly, in addition to possessing a bright near-surface region, the fluorescent micrograph also showed a bright spot located at the center of the droplets. The radial configuration of the LC may serve as a lightguide, thus leading to the bright spot. Next, we mixed 250 μL of the dispersion of L-DLPC-saturated LC droplets with an equal volume containing 2 nM PLA2 (and 10 mM CaCl2) in a glass vial. The scintillation vial was pretreated with BSA to avoid loss of enzyme from the bulk solution to adsorption on glass wall of the vial (see Methods section for more details). The mixture was gently vortexed to mix LC droplets and enzyme, and then, at various time intervals following the mixing, a few microliters of this dispersion was removed from the vial and observed under an optical microscope. Figure 8 shows the corresponding brightfield, polarized, and fluorescent light micrographs of the LC droplets as a function of time (up to 60 min). Inspection of these images reveals that (i) the PLA2 triggers a transformation of the LC from a radial configuration to a bipolar configuration, and (ii) intermediate topological states are observed during this transition, including the presence of an escaped-radial configuration (10 min), a preradial configuration (30 min), and a disclination ring (50 min) topological defects. Interestingly, for each of these topological states, we observe the defects within the LCs to give rise to enhanced levels of fluorescence intensity within the fluorescent images (indicated by white arrows in the bottom row of Figure 8). In the absence of 5 mM CaCl2, which is required for the enzyme activity, no change is ordering of LC droplets was observed upon addition of PLA2. A key conclusion that emerges from the above results obtained with L-DLPC and PLA2 is that the sequence of topological defects encountered during the enzymatic degradation of PLA2 is the same as that encountered during the adsorption of SDS. This result is, at first sight, somewhat surprising, as our past studies17,18 have established that the elastic energy stored inside strained states of LCs can induce phase separation of L-DLPC at interfaces of the LCs to aqueous phases. In the presence of an inhomogeneous distribution of lipid (and associated inhomogeneous anchoring of the LC), the topological states assumed by the LC would be expected to be perturbed relative to those induced by SDS. We comment, however, that the above discussion of phase separation of phospholipids at the aqueous-LC interface requires that sufficient time exists during the transition from the radial to the bipolar configuration for lateral transport of the lipid. In past studies,19 we have quantified the mobility of phospholipids at aqueous-LC interfaces by performing fluorescence recovery after photobleaching (FRAP), and based on these measurements, we calculate the time required for lipid to be Langmuir 2009, 25(16), 9016–9024
Article
transported across a LC droplet with a diameter of 8 μm to be approximately ∼2 h. In our experiments, the surface-driven transition driven by PLA2 was observed to be completed within 1 h. Thus, we conclude that insufficient time exists in our experiments to permit the lateral transport of lipid that would be required to achieve phase separation. This conclusion is supported by the observation of largely uniform fluorescence intensities in the fluorescent micrograph of droplets (other than in the regions of the droplets close to the defects in the LC). Finally, we comment here that we did attempt to perform experiments at lower concentrations of PLA2. In such experiments, however, we observed the LC droplets to become highly unstable and to coalesce when in contact with the PLA2 for more than 1 h. It is possible that this instability is caused by inhomogeneous interfacial distributions of lipid in these experiments. The results described above, in which PLA2 triggers a change in the optical appearance of LC droplets decorated with L-DLPC, are interesting to consider in light of past studies20,28 by us and others in which the enzymatic activity of PLA2 has been reported at DLPC-decorated interfaces of supported, micrometer-thick films of LC. Those studies documented that the catalytic hydrolysis of the L-DLPC resulted in an ordering transition in the supported film of LC. In particular, when using 1 nM of PLA2, it was reported that the ordering transition induced by PLA2 within the supported film of LC occurred over 300 min.28 In comparison, inspection of Figure 8 reveals that the ordering transition induced in the L-DLPC-decorated LC droplet is complete within 60 min. We note that a number of differences between the droplet and film geometry may account for the differences observed in the PLA2 activity: (i) past studies29 have established that the rate of hydrolysis of L-DLPC by PLA2 is sensitive to the ordering of the lipid molecules within a monolayer, thus leaving open the possibility that the differences in the rates of hydrolysis in the film versus droplet geometry may be related to the ordering of the lipids at the two interfaces; (ii) mass transport of the enzyme to the surface of the droplets will be more rapid for the droplets than for the films, as the droplets are free to diffuse in solution; and (iii) the elastic energy density of the LC within the droplets will likely differ from the film geometry,7,17 thus leading to different distributions of lipids and enzymes at the surface. Although additional experiments are needed to elucidate the origins of the accelerated response of the LC droplet, the experiments shown in Figure 8 do demonstrate that the LC droplet system offers the basis of a generally useful approach for studying biomolecular interactions20,30 occurring at lipid-laden interfaces.
Conclusions We end this paper by emphasizing the advances presented in this paper relative to past studies of adsorbate-induced ordering transitions in LC droplets.13,14 First, we note that the experimental system that underlies our observations is based on monodisperse droplets of LC dispersed in an aqueous phase. Prior studies of adsorbate-induced ordering transitions have utilized LC-droplets dispersed in a polymeric phase, or LC droplets dispsered in glycerol, to which lecithin was added. In the former system, the distribution of lethicin was inhomogeneous, and in the latter system the change in anchoring of the LC was induced by changes in temperature. These two experimental systems lead to conflicting observations regarding the series of topological defects encountered in LC droplets. By using monodisperse droplets of LC in water, and by using the (29) Deems, R. A. Anal. Biochem. 2000, 287, 1–16. (30) Park, J. S.; Abbott, N. L. Adv. Mater. 2008, 20, 1185–1190.
DOI: 10.1021/la900786b
9023
Article
reversible adsorption of SDS from the aqueous phase to the LC-aqueous interface, we have been able to document changes in topologically ordered states of LC droplets at constant temperature. We have also been able to confirm that the states observed are a function only of the concentration of SDS, and not the history of the system, thus confirming that the topological states are equilibrium states. These characteristics of the experimental system used in our study allow the unambiguous identification of a sequence of six distinct and stable topological states. Although we do not yet fully understand the balance of surface and volumetric effects within the LC droplets that give rise to this spectrum of ordered states, the well-defined thermodynamic state of the systems on which our observations are based should enable precise comparisons to simulations and theories of LCs in confined geometries. These simulations are underway and will be reported in the future. A second important difference between our experiments and those reported previously is the use of the aqueous continuous phase. This characteristic of the system allowed us to introduce biological molecules into the continuous phase, and thus explore the impact of their interfacial interactions on the ordering of the LC within the droplets. The results reported in this paper demonstrate that changes in the ordering of LCs within droplets
9024 DOI: 10.1021/la900786b
Gupta et al.
can be used to report interfacial interactions involving biomolecules, and in particular the enzymatic hydrolysis of phospholipids. The level of control over the temperature, interfacial chemistry, and sizes of LC droplets that is made possible by using the methods described in this paper will, we believe, enable new ways to exploit interfacial adsorbate-induced properties of LC droplets. The findings reported in this study are significant as they suggest that such transitions between distinct topological states can provide a means to quantify surface events. Acknowledgment. This work was supported by the ARC Linkage International Materials World Network Grant (F.C and N. L. A.), the ARC Federation Fellowship Scheme (F.C.) and the NSF (DMR-0520527 and DMR-0602570). Partial financial support from the ARO (W911NF-06-1-0314 and W911NF07-1-0446) is acknowledged. Supporting Information Available: Schematic illustration of possible director configurations corresponding to the disclination loop observed in LC droplets (Figure S1). This material is available free of charge via the Internet http://pubs.acs.org.
Langmuir 2009, 25(16), 9016–9024
