A new pathway for the formation of radial nematic droplets within a lipid-laden aqueous-liquid crystal interface

Abstract

A new pathway for the formation of liquid crystal (LC) droplets with radial LC ordering in the presence of surfactants and lipids is reported. This study also shows that the response of these droplets due to the interactions between an enzyme and the topological defects in the LC may be exploited in applications such as sensing.

---

Recently, liquid crystal (LC) droplets have been widely appreciated as a new class of functional materials due to their large surface areas, rich phases, well-defined director configurations and unique tunable optical properties.^1–10 In particular, they offer routes to design simple, economic and convenient passive sensing devices that provide a high spatial resolution of micrometers with a very high sensitivity. Therefore, it is very important to explore new pathways of preparing stable and uniform LC droplets that will ultimately provide a simplified and robust LC sensing platform. Past reports established the feasibility of forming LC droplets via sequential ultra-sonication and vortexing of LCs in water with emulsifying agents.^1,10–15 However, droplets prepared using these methods have lower stability and a broad size distribution, which limits their widespread use in real applications. To address these issues, much progress has been made to stabilize LC droplets with uniform sizes. For example, uniform silica particles coated with polyelectrolyte multilayers (PEMs), microfluidic devices, etc. have been used as templates for stable and uniform LC droplets.^5,12,16,17 These polyelectrolytes and other surface active agents (e.g., surfactants) can assemble at the interface of the LC droplet giving rise to a stable director configuration that is governed by surface anchoring and bulk elastic energies of the LC. However, most of these techniques are either tedious or not suitable for large scale production. For instance, in the case where silica particles are coated with PEMs for the preparation of LC droplets, silica cores need to be first etched with hydrofluoric acid (HF) to achieve hollow polymeric capsules followed by filling the capsules with LCs.^5,12,16 Because of the highly corrosive nature of HF, the preparation of LC droplets becomes tedious and time-consuming. Similarly, in the case of microfluidic devices, a stream of co-flowing liquids was used to squeeze and break off LCs into spherical droplets.¹⁷ Therefore, it is difficult to customize the size of the LC droplets as they depend on the dimensions of the channel. Herein, we report a new pathway for the easy formation of spontaneous uniform LC droplets. While the techniques reported in the past have resulted in the preparation of LC droplets with a life span of a few hours,¹⁸ our approach has provided LC droplets with a stability of days to months. We observed the spontaneous formation of well-developed LC droplets with radial defects in the presence of phosphatidylcholine (PC) within the confined boundary created by a grid system, suggesting new principles for the design of LC-based chemical and biological sensors.

Although observations reported in the past established the feasibility of the formation of LC droplets with diameters in the micrometer-to-sub-micrometer range and their size dependent ordering,¹⁶ direct observations of the spontaneous evolution of LC droplets with radial LC ordering in the presence of surfactants and lipids have not yet been investigated. The study reported in this paper reveals the first breakthrough that shows characteristic micrometer-scale LC droplet patterns in the presence of lipids and their control over LC droplet sizes. We have also demonstrated that interactions between an enzyme and the topological defects in the LC mediate the response of these droplets and thus provide new designs for stimuli-responsive soft materials.

We first performed an experiment to determine if the nematic ordering of 5CB (4′-pentyl-4-cyanobiphenyl) influences the organization and assembly of PC adsorbed to an aqueous-5CB interface. We hosted 5CB in the pores of 20 μm thick electron microscopy grids supported on N,N-dimethyl-n-octadecyl-3-aminopropyltrimethoxysilyl chloride (DMOAP) coated glass slides which induced homeotropic anchoring of the LC.¹⁹ After subsequent immersion in an aqueous solution, the optical appearance of the LC became bright, consistent with an orientation ordering transition induced by contact with water.²⁰ The results shown in Fig. 1A were obtained by adsorbing PC (∼0.5 mg mL⁻¹) to the aqueous-5CB interface, which leads to a change in the optical appearance of the LC from bright to dark within a few minutes. After incubation for 3 h or more, we observed well-developed droplets which were characterized by a single point defect located at the centre of the droplets shown in the bright field optical micrography image (Fig. 1B). We hypothesized that the monolayer of PC rearranges itself around the LC, resulting in the formation of LC droplets with defects (see below for details). The schematic representation and the process of the time-dependent formation of droplets are shown in Fig. 1C and S1 (ESI†).

Fig. 1 (A) Crossed polars (CP) and (B) bright field (BF) images of aqueous-5CB interfaces within the TEM grids supported on DMOAP coated glass slides upon exposure to 0.5 mg mL⁻¹ phosphatidyl choline (PC) after 10 min, 6 h and 24 h. The insets (in (A) and (B)) show the corresponding high-magnification images that reveal the formation of stable and well-developed LC droplets exhibiting radial configuration. (C) Top to bottom: schematic illustration of the time-dependent formation of LC droplets with radial LC ordering. Scale bar = 40 μm.

To provide an insight into whether PC is responsible for the formation of LC droplets with radial LC ordering, we first incubated a LC film (supported on DMOAP coated glass slides) in an aqueous phase for three days followed by an exchange of the aqueous phase with the PC solution. Past investigations reported that a micrometer-thick film of the LC supported on OTS (octadecyltrichlorosilane) treated glass substrates led to the formation of water droplets when immersed under water.²¹ On immersion of the LC film under water, we observed a spontaneous evolution of well-developed water droplets formed at the LC–DMOAP interface within 72 h (Fig. 2A). Interestingly, when exchanging the aqueous phase with the PC solution we observed a topological defect formation within the droplet after 19 h. We note two observations from Fig. 2B. First, the regions of LCs around the droplets have transformed into a homeotropic orientation, as evidenced by the black areas. Second, LC droplets with topological defects are formed in the same location where water droplets were previously present. We do not yet fully understand the reason for the formation of LC droplets from water droplets after adding PC. It may be hypothesized that in the presence of PC, there could be a gradual replacement of water by LCs which is stabilized through topological defects. Fig. S2 shows the time lapse images of the organization of water droplets to well-defined LC droplets (see ESI†).

Fig. 2 (A) Polarized optical micrograph (crossed polars) of nematic 5CB hosted in a gold grid supported on a DMOAP treated glass substrate after contact with water for 3 days. Water droplets formed spontaneously on the LC film. (B) Polarized optical micrograph of the aqueous-5CB system (as in (A)) on exchanging the aqueous phase with the PC solution after 19 h. The well-developed water droplets reorganize to LC droplets with radial LC ordering. The insets in (A) and (B) show the images of a water droplet and the corresponding LC droplet at high magnification, respectively. Scale bar = 40 μm.

The combination of the results described above leads to the following overall understanding of the LC droplet formation. During the formation of the macro emulsion, the reduction of the interfacial tension reduces the amount of mechanical work required to break the inner phase into dispersed particles. So, the use of surfactants decreases the surface tension as well as the rate of coalescence of the dispersed LC particles by forming mechanical, steric or electrical barriers around them. Theoretical approaches also reported the formation of an equilibrium state of a LC droplet which could be described as the minimum of the free energy functional (F), composed of both a volume (F_v) and a surface part (F_s).²² The F_s can again be modeled as surface tension which has an isotropic part σ and an anisotropic part W_A (i.e., F_s ∼ 4π(σr² + W_Ar²), r being the radius of the droplet). In the case of cyanobiphenyl, when the anchoring part W_A (∼10⁻⁵ to 10⁻⁶ J m⁻²) is much smaller than σ (∼10⁻³ to 10⁻² J m⁻²), F_s (∼4πσr²) is dominated by the surface tension and thus proportional to r². It has been demonstrated that PC could decrease σ from 10⁻³ to 10⁻⁴ J m⁻² which in turn decreases F_s.²³ In such cases, F is reduced to a greater extent which results in the formation of LC droplets, stabilized by topological defects. Our understanding of the formation of LC droplets is also supported by prior experiments. For example, the measured surface tension of a free interface (such as the LC–argon and LC–glycerin interface) with no surfactant is in the order of 15–20 mJ m⁻².²² In contrast, the presence of a small amount of lecithin in glycerin decreases the surface tension between the LC and the glycerin to 10 mJ m⁻² for the cholesteric (Ch) and smectic A (SmA) phase but drops to 3 mJ m⁻² in SmA*.²⁴ Kim et al. reported that the interface between 5CB and water + cetyl trimethylammonium bromide (CTAB) has a surface tension in the order of (1–6) mJ m⁻².²⁵ Similarly, a plot of surface tension vs. temperature for 5CB–glycerin (no surfactant) showed that the values of surface tension are higher, again between 20 and 15 mJ m⁻².²² Of course, glycerin and water are different matrices but adding a surfactant would generally decrease the interfacial tension between a LC and the surrounding isotropic fluid. The value inevitably depends on the concrete materials, as in the case of SmA*, it was decreased by a factor of 5–7, i.e., an order of magnitude.

In addition to the reduction of the interfacial tension, PC as reported imposes the anchoring boundary conditions for the nematic director.²⁶ PC, due to its hydrophobic part, interacts with the alkyl chain of 5CB and results in the homeotropic ordering of the LC. It then rearranges itself around the LC molecules and creates a radial arrangement around the LC molecules where each droplet has a single point defect at the centre (called hedgehog). Previous studies show the radial arrangement of the LC droplets with lecithin.²⁶ Past reports by Terentjev demonstrated that the stability of the nematic macro emulsions is greatly enhanced compared to that of the isotropic counterparts and thus the elastic constant of the LC and surface tension creates an energy barrier for coalescence.²⁷ This energy barrier of coalescence is ∼K²/W. With K ∼ 10⁻¹¹ N and W ∼ 10⁻⁵ J m⁻², the energy barrier is very high (∼10⁻¹⁷ J) compared to the typical thermal energy at room temperature.²⁷ Hence, it is very likely that in the presence of PC, there is an additional contribution of the elastic energy to the formation of the topological defects in the confined nematic and radial LC droplets, which prevents coagulation compared to their isotropic counterparts.

In order to provide a further insight into the microstructure of the LC droplets in the presence of PC, fluorescently tagged PC was used to examine the location of the surfactant (PC) in the co-assembly. Fig. 3A shows the optical micrograph of 5CB hosted within gold grids supported on DMOAP coated glass slides after contact with PC mixed with fluorescent PC (NBD PC, see the Experimental section in the ESI† for details) for 3 days. We observed that the fluorescence was mainly present at the boundary of (and also in between) the droplets, with no fluorescence at the centre of the droplets. This led us to conclude that the formation of the PC monolayer stabilizes these droplets by rearranging itself around the droplets. A bright field optical image of the same is shown in Fig. 3B. To understand that we have LCs outside the droplet (not only inside), we performed confocal and fluorescence microscopy with a fluorescent N,N′-bis(2,5-di-tert-butylphenyl)-3,4,9,10-perylenedicarboximide (BTBP) compound which is known to align in the presence of LCs.²⁸ As expected from the fluorescence (Fig. 3C) and from the confocal image (Fig. 3E), we observed more BTBP outside the droplet. This study suggests that the background between the droplets is birefringent. Bright field fluorescence and confocal images are shown in Fig. 3D and F, respectively.

Fig. 3 (A) Fluorescence micrograph and (B) the corresponding bright field image of nematic 5CB hosted within a gold grid supported on a DMOAP treated glass substrate after contact with PC doped with 1 μM fluorescent PC (NBD PC) for 3 days. (C) and (E) represent fluorescent and confocal microscopy images of BTBP doped 5CB hosted within TEM grids after contact with PC vesicles for 5 days, respectively. (D) and (F) correspond to the fluorescentce and confocal bright field images, respectively. Scale bar = 55 μm.

Next we sought to determine the stability of these droplets with respect to time. It was found that the droplets produced in the presence of PC are quite stable for a period of 20 days or more. As shown in Fig. S3A (ESI†), after incubation of PC at the 5CB-aqueous interface for 3 days, the average droplet size reached ∼21.34 ± 4.67 μm and remained for at least 5 days. A slight variation of the average size was observed within these days (i.e., from 21.34 ± 4.67 μm to 21.70 ± 4.70 μm). After 8 days of incubation (Fig. S3B†) the sizes increased to 32.63 ± 11.16 μm and changed continuously with time (Fig. S3C and D†). The time dependent growth of the droplet formation in the presence of PC at the 5CB-aqueous interface is shown in Fig. S3E.†

We investigated and compared the stability and size distribution of the LC droplets to those of existing techniques. Past reports quantified the size distribution (with an average diameter of ∼6 μm) of 5CB droplets formed by sequential sonication and vortex mixing.¹⁸ In our case, we observed an increase in the average size of the diameter of the LC droplets by approximately three times (∼21 μm). Also, the techniques reported for the preparation of LC droplets in the past have a life span of only a few hours (Fig. S4 and S5 ESI†).¹⁸ On the other hand, the droplets prepared by our method are quite stable for a period of several days. This is because in our case, the droplets formed in a confined boundary created by the grid system and thus stabilized (less mobile). These experiments also confirm the suitability of our approach towards a simple and robust platform using LC droplets for further applications.

Next we examined the stability of the obtained LC droplets (with defects) below and above the nematic (N)–isotropic (I) phase transition temperature (T_N–I) of 5CB (∼35 °C). Below T_N–I, the LC droplets with radial LC ordering showed a bright optical appearance under crossed polars (Fig. S6, see ESI†). However, the optical response of the LC droplets exhibited a bright–dark optical appearance after they were heated above the T_N–I of 5CB. After cooling from the isotropic temperature to the N phase, we observed a bright optical appearance of the LC droplets followed by the appearance of the topological defects within these droplets. These results indicate that the LC droplets (with radial configuration) were quite stable and reversible i.e., the optical responses remain unchanged by the temperature-induced phase transition of the LCs.

In addition to PC, which induces the formation of stable droplets with radial LC configuration, we investigated the role of other surfactants and lipids and explored their behaviour on incubation at aqueous-5CB interfaces. We have chosen sodium dodecyl sulfate (SDS), CTAB, lysophosphatidic acid (LPA), 1,2-dilauroyl-sn-glycero-3-phosphocholine (DLPC) and lipopolysaccharide (LPS) and studied their behaviour at aqueous-5CB interfaces. The goal of this experiment is mainly twofold. First, we sought to find out whether other lipids are also suitable for the spontaneous formation of LC droplets in addition to PC as described above. Second, we sought to demonstrate any additional parameter which can advance our understanding in the formation of droplets. All the surfactants we have chosen are ionic and, therefore, can stabilize the droplets electrostatically. Fig. S7† shows the optical images of nematic 5CB hosted within gold grids supported on DMOAP coated glass slides in contact with CTAB, DLPC, SDS, LPS and LPA with respect to time. We monitored these systems carefully for 15 days and observed the following two key points. First, within 3 days of incubation, both CTAB and DLPC resulted in the formation of nice and well-developed droplets with radial LC ordering. Second, LPA does not lead to the formation of stable droplets whereas no droplet formation was observed in the presence of LPS and SDS. These results indicate that (i) CTAB and DLPC are more surface active (in comparison to SDS and the others) and thus resulted in a more efficient reduction in surface tension.^5,23,29 (ii) Past reports also demonstrated that the LC droplets formed in water remained stable due to the adsorption of hydroxide ions at the LC–water interface.³⁰ This hydroxide ion adsorption makes the LC droplet surface negatively charged. So, it can be concluded that in addition to an efficient reduction in surface tension, an electrostatic attraction in the presence of DLPC and CTAB (DLPC is zwitterionic and CTAB is positively charged) at the LC interface (negatively charged) stabilized the LC droplets compared to SDS, LPA and LPS (being negatively charged). Also a negative surface charge density on DMOAP coated surfaces stabilizes the positive and zwitterionic droplets. These results reveal for the first time that, in addition to inducing homeotropic alignment of the LC, the ionic charge in the lipids can play an important role for the spontaneous formation of LC droplets.

The observations reported above are interesting to consider based on prior studies on the topological ordering of LCs within droplets to report interfacial enzymatic reactions.⁶ Past reports established that PLA₂ triggers a transformation of the LC droplet from a radial configuration to a bipolar configuration decorated with L-DLPC.⁶ Motivated by this, we investigated the topological states encountered in the LC droplets during enzymatic degradation of PC using PLA₂. Fig. 4 shows the corresponding polarized light micrograph of the LC droplets in response to the adsorption of PLA₂ (150 nM). Interestingly, upon introduction of PLA₂, an anchoring transition of the LC was observed from an initially homeotropic orientation (radial configuration in the presence of PC) to a planar orientation with a bipolar topological defect. These bipolar droplets were then further explored in detecting various bio-molecules such as bacterial phospholipids (LPS). Past reports demonstrated that LPS induced an ordering transition in LC droplets from a bipolar to radial configuration.⁸ This ordering transition is not mediated by surface anchoring energy but rather consistent with the association of LPS with defects. We investigated this phenomenon with LC droplets formed through enzymatic degradation of PLA₂ (bipolar) and observed a structural transformation to radial configuration. The optical response of these LC droplets with a bipolar configuration to radial configuration after the adsorption of LPS is shown in Fig. S8.† This result, we believe in principle, enables new pathways to exploit interfacial adsorbate-induced properties of LC droplets.

Fig. 4 Top row: (A–D) optical images (crossed polars) of 5CB droplets covered with DLPC upon exposure of 150 nM PLA₂ into an aqueous solution of TBS (pH = 8.9) containing 10 mM CaCl₂. These droplets with a radial defect were formed by contacting the 5CB interface laden with DLPC (0.1 mg mL⁻¹) at an incubation period of 2 days. The second row depicts the change in the anchoring transition of the representative region of the four selected droplets (a–d) from radial to bipolar upon adsorption of PLA₂. Scale bar = 40 μm.

In conclusion, the study establishes first to reveal direct observations of the spontaneous evolution of LC droplets with radial LC ordering in the presence of surfactants and lipids. The formation of stable LC droplets are not only due to the reduction of the interfacial tension between the LC and the surrounding isotropic fluid but also have an additional stability mechanism (against coagulation) associated with the internal elasticity. Our observations also affirm that ionic charge can play an important role in the spontaneous formation of LC droplets with topological defects. Finally, we have shown that interactions of an enzyme with the topological defects in the LC droplets can provide means for developing new responsive soft materials.

Acknowledgements

The authors thank Professor Oleg D. Lavrentovich (Kent State University) for insightful comments and helpful suggestions. This work was carried out with the financial support from IISER Mohali and Department of Atomic Energy (DAE-BRNS). S. Sidiq acknowledges the receipt of a graduate fellowship from UGC.

Notes and references

1. V. J. Aliño, J. Pang and K.-L. Yang, Langmuir, 2011, 27, 11784–11789.
2. V. J. Aliño, K. X. Tay, S. A. Khan and K.-L. Yang, Langmuir, 2012, 28, 14540–14546.
3. T. Bera and J. Fang, J. Mater. Chem., 2012, 22, 6807–6812.
4. T. Bera and J. Fang, Langmuir, 2013, 29, 387–392.
5. J. K. Gupta, J. S. Zimmerman, J. J. de Pablo, F. Caruso and N. L. Abbott, Langmuir, 2009, 25, 9016–9024.
6. W. Khan, J. H. Choi, G. M. Kim and S.-Y. Park, Lab Chip, 2011, 11, 3493–3498.
7. M. I. Kinsinger, M. E. Buck, N. L. Abbott and D. M. Lynn, Langmuir, 2010, 26, 10234–10242.
8. I.-H. Lin, D. S. Miller, P. J. Bertics, C. J. Murphy, J. J. de Pablo and N. L. Abbott, Science, 2011, 332, 1297–1300.
9. S. Sivakumar, K. L. Wark, J. K. Gupta, N. L. Abbott and F. Caruso, Adv. Funct. Mater., 2009, 19, 2260–2265.
10. J. Zou, T. Bera, A. A. Davis, W. Liang and J. Fang, J. Phys. Chem. B, 2011, 115, 8970–8974.
11. K. A. Simon, P. Sejwal, R. B. Gerecht and Y.-Y. Luk, Langmuir, 2007, 23, 1453–1458.
12. S. Sivakumar, J. K. Gupta, N. L. Abbott and F. Caruso, Chem. Mater., 2008, 20, 2063–2065.
13. E. Tjipto, K. D. Cadwell, J. F. Quinn, A. P. R. Johnston, N. L. Abbott and F. Caruso, Nano Lett., 2006, 6, 2243–2248.
14. O. Tongcher, R. Sigel and K. Landfester, Langmuir, 2006, 22, 4504–4511.
15. J. Zou and J. Fang, Langmuir, 2010, 26, 7025–7028.
16. J. K. Gupta, S. Sivakumar, F. Caruso and N. L. Abbott, Angew. Chem., Int. Ed., 2009, 48, 1652–1655.
17. P. Umbanhowar, V. Prasad and D. Weitz, Langmuir, 2000, 16, 347–351.
18. D. S. Miller and N. L. Abbott, Soft Matter, 2013, 9, 374–382.
19. X. Bi, S. Huang, D. Hartono and K.-L. Yang, Sens. Actuators, B, 2007, 127, 406–413.
20. J. M. Brake, M. K. Daschner, Y.-Y. Luk and N. L. Abbott, Science, 2003, 302, 2094–2097.
21. Z. Yang and N. L. Abbott, Langmuir, 2010, 26, 13797–13804.
22. O. D. Lavrentovich, Liq. Cryst., 1998, 24, 117–126.
23. D. Harrison and M. R. Fisch, Liq. Cryst., 2000, 27, 737–742.
24. O. D. Lavrentovich, Y. A. Nastishin, V. I. Kulishov, Y. S. Narkevich, A. S. Tolochko and S. V. Shiyanovskii, Europhys. Lett., 1990, 13, 313–318.
25. J.-W. Kim, H. Kim, M. Lee and J. J. Magda, Langmuir, 2004, 20, 8110–8113.
26. O. O. Prischepa, A. V. Shabanov and V. Y. Zyryanov, Mol. Cryst. Liq. Cryst., 2005, 438, 141/[1705]–1150/[1714].
27. E. M. Terentjev, Europhys. Lett., 1995, 32, 607–612.
28. O. D. Lavrentovich, Contemp. Math., 2012, 577, 25–46.
29. B. Sohrabi, H. Gharibi, B. Tajik, S. Javadian and M. Hashemianzadeh, J. Phys. Chem. B, 2008, 112, 14869–14876.
30. K. G. Marinova, R. G. Alargova, N. D. Denkov, O. D. Velev, D. N. Petsev, I. B. Ivanov and R. P. Borwankar, Langmuir, 1996, 12, 2045–2051.

---

Footnote

† Electronic supplementary information (ESI) available: Detailed procedure for the preparation of LC films, vesicles, LC droplets and their optical characterization and fluorescent imaging are described in the ESI. See DOI: 10.1039/c3ra48044e

---