Experimental Evidence for Proposed Transformation Pathway from the

Jul 6, 2015 - Ordered soft nanomaterials form by the self-assembly of block copolymers1 or amphiphilic molecules in lyotropic liquid crystalline phase...
0 downloads 0 Views 4MB Size
Letter pubs.acs.org/Langmuir

Experimental Evidence for Proposed Transformation Pathway from the Inverse Hexagonal to Inverse Diamond Cubic Phase from Oriented Lipid Samples Adam M. Squires,*,† Samina Akbar,† Marissa E. Tousley,‡ Yekaterina Rokhlenko,‡ Jonathan P. Singer,‡ and Chinedum O. Osuji‡ †

Department of Chemistry, University of Reading, Whiteknights Campus, Reading, U.K. RG6 6AD Department of Chemical and Environmental Engineering, Yale University, New Haven, Connecticut 06520, United States



S Supporting Information *

ABSTRACT: A macroscopically oriented inverse hexagonal phase (HII) of the lipid phytantriol in water is converted to an oriented inverse double diamond bicontinuous cubic phase (QIID). The initial HII phase is uniaxially oriented about the long axis of a capillary with the cylinders parallel to the capillary axis. The HII phase is converted by cooling to a QIID phase which is also highly oriented, where the cylindrical axis of the former phase has been converted to a ⟨110⟩ axis in the latter, as demonstrated by small-angle X-ray scattering. This epitaxial relationship allows us to discriminate between two competing proposed geometric pathways to convert HII to QIID. Our findings also suggest a new route to highly oriented cubic phase coatings, with applications as nanomaterial templates.



INTRODUCTION Ordered soft nanomaterials form by the self-assembly of block copolymers1 or amphiphilic molecules in lyotropic liquid crystalline phases containing, for example, industrial surfactants2 or biological lipid molecules.3 They adopt a range of morphologies showing periodicity in one, two, or three dimensions (Figure 1). One structure consists of a 2-D hexagonal lattice of cylinders that may be surfactant micelles surrounded by water (categorized as a type I hexagonal lyotropic phase, often denoted HI), water channels surrounded by lipid (HII), or one type of polymer surrounded by another (block copolymer cylindrical phase). Another class of structure, showing periodicity in three dimensions, comprises bicontinuous morphologies with cubic symmetry (Figure 1). In the case of (type II) lipid cubic phases, these contain branching 3D networks of water channels separated by a continuous lipid bilayer and are denoted QII. In block copolymer and type I lyotropic systems, the networks are instead formed not from water but instead from a specific polymer block or from branching tubular surfactant micelles, respectively. The most widely studied of these phases is the double gyroid (space group Ia3d); other bicontinuous cubic structures are the double diamond (Pn3m) and primitive (Im3m). In lipid systems these are respectively denoted QIIG, QIID, and QIIP. The double diamond is commonly found in lipid systems, especially under excess-water conditions; it has been observed in block copolymers, although it is much less common than the double gyroid.4 © 2015 American Chemical Society

Figure 1. Top row: bicontinuous cubic phases. Left to right: double gyroid, double diamond, and primitive. Bottom: hexagonal phase. In the case of type II lipid phases, these are denoted QIIG, QIID, QIIP, and HII, respectively, and the red/green rods represent the water channel centers whereas the yellow/blue surfaces show the center of the lipid bilayer.

These nanomaterials are of interest for a variety of scientific and commercial reasons. Lipid mesophases themselves are naturally occurring biological membrane structures and Received: May 12, 2015 Revised: July 6, 2015 Published: July 6, 2015 7707

DOI: 10.1021/acs.langmuir.5b01676 Langmuir 2015, 31, 7707−7711

Letter

Langmuir intermediates 5 and are used as matrices for protein encapsulation,6 in particular, for protein crystallography.7 They are potential vehicles for drug8 or flavor9 delivery, with phase transitions induced by irradiation or pH changes, suggesting mechanisms for triggered release. 10 Similar bicontinuous structures based on synthetic analogues show increased ion conductivity11 for batteries and fuel cells.12,13 Furthermore, bicontinuous cubic structures based on lipids14 and block copolymers15 have been used as templates for the production of mesoporous metal with gyroid and diamond structures, with enhanced catalytic activity16,14 and unusual optical behavior.17 Understanding the structural properties and phase transitions of lipid mesophases is essential to the targeted development of materials for the applications described in the previous paragraph. The mechanisms of transformation between different QII phases have recently been elucidated using oriented samples to investigate the pathways from the diamond to the gyroid (QIID to QIIG)18 and from the diamond to the primitive (QIID to QIIP),19 confirming previous theoretical predictions.20 Meanwhile, transformations between hexagonal and gyroid structures have been carried out on oriented samples of block copolymer21 and type I lyotropic liquid crystals,22 in both cases suggesting that the hexagonal cylinders are converted into a ⟨111⟩ axis in the gyroid, although recent findings in block copolymer systems suggest that other epitaxial (orientational) relationships are possible.23 Our work here is the first such study of the HII to QII transformations in oriented lipid systems, with any cubic phase morphology; the kinetics of HII to QII transformations have been studied on unoriented systems,24 and this particular transformation has been investigated as the basis for stimuli-responsive drug release.10 Moreover, ours is the first study on the epitaxy of transformations between the hexagonal cylindrical and diamond bicontinuous cubic topologies in any material; the diamond phase is very rare in type I lyotropic liquid crystals or block copolymers,4,25 so type II lipidic liquid crystals represent an attractive target system.



Figure 2. Azimuthally averaged small-angle X-ray scattering patterns of phytantriol coating in a water-filled capillary (representing excess water conditions) with increasing temperature. The intensity axis is plotted on a log scale.

around 65−70 °C (Figure 2). This relatively disordered fluid isotropic phase consists of spherical droplets of water each surrounded by a monolayer of lipid (an inverse micelle). The transformation to the LII phase can be thought of as “melting” the liquid-crystalline phase. This sequence has been reported elsewhere26 with exact phase boundary temperatures dependent on the sample purity and commercial source.27 Slow cooling from the LII to the HII phase in the capillary tube produces an oriented HII phase with cylinders aligned parallel to the capillary axis, which is horizontal, as shown in Figure 3. The vertical (equatorial) reflections correspond to the (100) planes of the HII phase, with lattice parameter 45.8 ± 0.3 Å (d spacing 39.7 ± 0.3 Å). This value is consistent with values quoted elsewhere.26,27 The procedure reproducibly forms highly oriented HII phases with this orientation, through a mechanism likely to involve a combination of surface effects and temperature gradients that we investigate in a separate study, beyond the scope of this work. On further slow cooling, the sample undergoes a phase transition to the QIID phase as confirmed by two reflections at relative 1/d values of 21/2:31/2. The QIID phase formed is also highly oriented (Figure 3). The intense horizontal (axial) reflections in the first ring, which has Miller indices of (110), indicate sample orientation with the [110] axis aligned along the capillary axis and therefore in the same direction as the cylinders of the preceding HII phase. (We have included full simulated spot patterns in Supporting Information for this orientation, which matches the observed 2D SAXS pattern very well, and the alternative [111] orientation, which does not.) Upon continued cooling to room temperature, the QIID phase maintains its orientation, although its lattice parameter increases from 60.7 Å initially to approximately 69.7 ± 0.4 Å. This increase in the lattice parameter on cooling is typical of QII

MATERIALS AND METHODS

Samples were prepared as coatings approximately 30−50 μm thick of the lipid phytantriol inside epoxy-sealed 1.5-mm-diameter Kapton tubes filled with water. SAXS data were obtained using a Rigaku SMAX300 pinhole-collimated laboratory instrument, with a beam of wavelength of 1.542 Å and size 1 mm at the sample plane, which was 80 cm from the detector. SAXS data extending to a slightly wider angle, to confirm phase identities (Figure 2), were obtained on a Bruker Nanostar instrument with camera length 67 cm, with phytantriol coated inside 1.5-mm-diameter glass capillary tubes filled with water. Oriented HII samples were prepared by slow cooling from the inverse micellar (LII) phase at 67 °C to the HII phase at 56 °C over approximately 2.5 h. Transformations to QIID were then induced by cooling the HII phase to 36 °C (below the QIID−HII phase boundary) over approximately 3 h. Further details are given in Supporting Information. For the control experiment, the sample was heated as above, but only to 52 °C, to produce an unoriented HII phase. This was then cooled to room temperature as above.



RESULTS AND DISCUSSIONS The phase behavior of phytantriol in excess water has been well characterized.26 As expected, our coatings form a QIID phase in excess water at room temperature by taking on approximately 28% w/w water.26 On heating, they transform to the HII phase at 45−50 °C and then to the inverse micellar (LII) phase at 7708

DOI: 10.1021/acs.langmuir.5b01676 Langmuir 2015, 31, 7707−7711

Letter

Langmuir

QIIG and QIID18 and between QIID and QIIP.19 For the transformation from the HII phase analyzed in this letter, it might be expected that the transformation from HII to QIID would involve the conversion of the water cylinder axis in the former to the [111] axis in the latter. This is superficially attractive because the water channel “rods” making up the skeletal networks in the QIID phase do indeed point in different directions in the ⟨111⟩ family and form a hexagonal array when viewed down that axis (Figure 4). This would also be consistent with epitaxial relationships observed in transformations from hexagonal to gyroid bicontinuous cubic phases in block copolymers21 and type I lyotropic liquid crystals.22

Figure 3. 2D SAXS patterns of the phytantriol coating in a water-filled capillary. (a−f) Heating/cooling from the LII phase to produce oriented samples; (g−i) controlled heating/cooling from the HII phase. (a) Initial QIID phase, room temperature, before heating; (b) heated to the HII phase, 59.5 °C; (c) heated to the LII phase, 67 °C; (d) cooled to the oriented HII phase, 56 °C; (e) cooled to the oriented QIID phase, 36 °C; (f) cooled to the oriented QIID phase, room temperature. (g) Initial QIID phase, room temperature, before heating; (h) heated to the HII phase, 52 °C; (i) cooled to the QIID phase, room temperature (approximately 20 °C).

phases and is associated with a decrease in monolayer curvature, which is favored as the thermal energy of the hydrophobic chains decreases; the lattice parameters are consistent with published values for phytantriol.26 A control experiment demonstrated that the orientation of QIID reflects the transformation pathway from the oriented preceding HII phase rather than some other orienting effect during the HII to QIID transition. For this, the sample was not heated to the LII phase but only to an (unoriented) HII phase, from which it was cooled to the QIID phase under the same conditions. This produced a QIID phase that also showed no orientation (Figure 3), demonstrating that the QIID orientation in Figure 2 does indeed arise from the orientation of the preceding HII phase. The relative orientation of initial and final structures can be used to test hypothesized pathways for lyotropic phase transitions, as shown previously in transformations between

Figure 4. Representation of the QIID phase viewed approximately along the [111] axis (a) and the [110] axis (b). Viewing directions have been slightly offset to aid in visualization. Schematic showing the transformation of half of the water channels into straight channels in the HII phase, viewed along the same [110] direction, i.e., the capillary axis (c−e), and along an orthogonal [110] axis corresponding to the beam direction (f−h). 7709

DOI: 10.1021/acs.langmuir.5b01676 Langmuir 2015, 31, 7707−7711

Langmuir



However, from the data we present here, this hypothesis can evidently be rejected; instead it appears that the HII cylindrical axis is converted to a ⟨110⟩ axis. The QIID phase viewed along such a direction is shown in Figure 4. Although its symmetry does not correspond to a hexagonal array, the projection shows continuous zigzag water channels running in the [110] direction. The preservation of such continuity in HII−QIID transformations evidently makes this epitaxial relationship more attractive than the alternate pathway linking to the [111] axis; channels passing in the [111] direction are regularly intersected by the bilayer, so we would expect the transformation of the [111] QIID axis to or from the HII channel direction to involve a greater degree of membrane disruption. We have tentatively included a schematic description of how such a transformation could occur in Figure 4, which should be seen as complementary to the pathway proposed by Clerc et al.28 for the hexagonal to gyroid transformation in type I surfactant systems. In addition to suggesting a proposed pathway converting HII to QIID, our experiments also demonstrate a new method of producing QIID coatings with an unprecedented degree of inplane alignment. Other routes to oriented lipid cubic phases include the use of much thinner (