pubs.acs.org/Langmuir © 2009 American Chemical Society
Influence of the Structure and Composition of Mono- and Dialkyl Phosphate Mixtures on Aluminum Complex Organogels† Miles G. Page‡ and Gregory G. Warr* School of Chemistry F11, The University of Sydney, NSW 2006, Australia. ‡ Current address: Weizmann Institute of Science, Rehovot 76100, Israel. Received February 8, 2009. Revised Manuscript Received April 2, 2009 The rheological properties and structure of organogels formed by the in situ complexation and self-assembly of aluminum isopropoxide and didodecyl phosphate surfactant in decane are investigated as mono-n-dodecyl phosphoric acid and bis(2-ethylhexyl) phosphoric acid complexing agents are added. At low loadings, the bulky bis(2-ethylhexyl) additive disrupts the physical gel structure by changing the packing around the aluminum centers, weakening the transition from viscoelastic fluid to physical network of branched cylinders, and completely suppresses gelation at high loadings. Monododecyl phosphate affects coordination at the Al center. At low substitution, it shifts the composition at which the transition to a physical gel occurs while simultaneously improving long-term stability. Structures deduced from the rheological response are confirmed by small-angle neutron scattering, which shows that the aggregates are locally cylindrical and molecularly thin at all compositions studied, although the cross section of the cylinders depends on the alkyl chain structure and composition of the organic phosphate mixtures.
Introduction Recent advances in the understanding of molecular organogels1 have allowed the range of potential gelators and their applications to be considerably widened. The rational design both of molecules that can act as gelators2-5 and useful properties of the resulting gels is becoming increasingly widespread.6 Gelating species with interesting properties such as chirality,7 luminescence,8 electro-optical9 and antioxidant10 properties, and tuning by reversible oxidation-reduction11 have recently been demonstrated. One aspect that still offers significant room for improvement is the ability to control sensitively the rheological properties and long-term stability of molecular organogels. As pointed out by Huang et al.,12 in many organogels these are strongly dependent on the stress history and/or methods of preparation. These authors therefore exploited the gel incubation temperature, while carefully controlling other preparation parameters, to influence strongly the resulting structure of the self-assembled fibrillar network (SAFIN) and therefore the physical properties of the †
Part of the Molecular and Polymer Gels; Materials with Self-Assembled Fibrillar Networks special issue. *Corresponding author. E-mail:
[email protected] (1) Molecular Gels: Materials with Self-Assembled Fibrillar Networks; Weiss, R. G., Terech, , P., Eds.; Springer: Dordrecht, The Netherlands, 2005. (2) van Esch, J. H.; Feringa, B. L. Angew. Chem., Int. Ed. 2000, 39, 2263. (3) Tamaru, S.-i.; Nakamura, M.; Takeuchi, M.; Shinkai, S. Org. Lett. 2001, 3, 3631. (4) Mieden-Gundert, G.; Klein, L.; Fischer, M.; Vogtle, F.; Heuze, K.; Pozzo, J.-L.; Vallier, M.; Fages, F. Angew. Chem., Int. Ed. 2001, 40, 3164. (5) Pozzo, J.-L.; Clavier, G. M.; Desvergne, J.-P. J. Mater. Chem. 1998, 8, 2575. (6) George, M.; Weiss, R. G. Acc. Chem. Res. 2006, 39, 489. (7) Maffezzoni, R.; Zanda, M. Tetrahedron Lett. 2008, 49, 5129. (8) De Paoli, G.; Dzolic, Z.; Rizzo, F.; De Cola, L.; Vogtle, F.; Muller, W. M.; Richardt, G.; Zinic, M. Adv. Funct. Mater. 2007, 17, 821. (9) He, J.; Yan, B.; Yu, B. Y.; Bao, R. Y.; Wang, X. O.; Wang, Y. H. J. Colloid Interface Sci. 2007, 316, 825. (10) Lo Nostro, P.; Ramsch, R.; Fratini, E.; Lagi, M.; Ridi, F.; Carretti, E.; Arnbrosi, M.; Ninham, B. W.; Baglioni, P. J. Phys. Chem. B 2007, 111, 11714. (11) Liu, J.; Yan, J. L.; Yuan, X. W.; Liu, K. Q.; Peng, J. X.; Fang, Y. J. Colloid Interface Sci. 2008, 318, 397. (12) Huang, X.; Raghavan, S. R.; Terech, P.; Weiss, R. G. J. Am. Chem. Soc. 2006, 128, 15341.
8810 DOI: 10.1021/la900477v
gelled liquid.12 Additionally, in most systems the gelator concentration and the operational temperature are the only other parameters available for tuning gel performance. Surfactants with phosphorus-based headgroups have attracted recent attention13-15 as one component of binary low-molecularweight organic gelators (LMOGs), which are systems whereby in situ gelation is achieved by adding, in this case, a metal ion complexing agent such as ions of iron,14,15 aluminum,13,14 and boron,14 resulting in SAFINs of a one-dimensionally polymerized organometallic complex. Forming the complex by an in situ reaction of the surfactant (e.g., a dialkyl phosphate, as shown in eq 1) with the metal alkoxide, the ratios of the two components (i.e., the surfactant and its complexing agent) can be adjusted, providing additional control over the physical properties of the gels.13
The microstructure of the aggregates that form the SAFIN in these gelating systems is represented schematically in Figure 1. There are two primary factors that control the formation of the aggregates. First, the chemical coordination between the complexing agent and the surfactant headgroup determines the structure of the core. Second, the structure of the surfactant tails determines the steric profile of the alkyl shell, which may influence properties of the aggregates, how they interact with the carrier liquid, and even whether they will form.12 To understand the “bridging” coordination structure16-18 satisfactorily, both of these factors must be considered. Resultant gel properties of the (13) Page, M. G.; Warr, G. G. J. Phys. Chem. B 2004, 108, 16983. (14) George, M.; Funkhouser, G. P.; Weiss, R. G. Langmuir 2008, 24, 3537. (15) George, M.; Funkhouser, G. P.; Terech, P.; Weiss, R. G. Langmuir 2006, 22, 7885. (16) Fukasawa, J.-I.; Tsutsumi, H. J. Colloid Interface Sci. 1991, 143, 69. (17) Rose, S. H.; Block, B. P. J. Am. Chem. Soc. 1965, 87, 2076. (18) Crescenzi, V.; Giancotti, V.; Ripamonti, A. J. Am. Chem. Soc. 1965, 87, 391.
Published on Web 04/29/2009
Langmuir 2009, 25(15), 8810–8816
Page and Warr
Figure 1. Schematic representation of the structure of aluminum ion-phosphate surfactant SAFIN sheathed by didodecyl tails arranged into cylinders, showing phosphate group bridging coordination to adjacent octahedral aluminum centers (Al, gray; P, green; O, red).
SAFIN gel in a given liquid are then a function of aggregate formation and, if they do form, the particle-particle and particle-liquid interactions between the aggregates in the SAFIN and the liquid in question (which are also influenced by the composition and structure of the shell). George et al.14 have examined a series of phosphonate surfactants as gelators with aluminum-, boron-, and zirconium-based complexing agents. They found, for example, that boron-based SAFINS were highly crystalline, aluminum gave amorphous, nonbirefringent gels (as we also observed13), and Zr failed to form a gelating network. They also experimented with different phosphonate surfactants, finding that hexadecylphosphonic acid and a bolaform C10 biphosphonate were far less effective gelators than the methyl ester of hexadecylphosphonic acid.14 The methyl ester, relative to the nonesterified alkylphosphonic acid, is probably a more effective gelator because of the chemistry of the headgroup, where only one available proton, not two, seems to be ideal for the formation of the aggregate core (see below). The microstructure and some properties of these self-assembled organogels show a number of similarities with inverse wormlike and bicontinuous microemulsions, including the extensively studied lecithin/water/oil19 and dialkyldimethylammonium salt/ water/oil systems.20-23 The stability, structure, and properties of these systems are extremely sensitive to the molecular structure of the amphiphile,24 which is a consequence of their packing at the water-oil interface. By analogy, changing the structure of the alkyl phosphate should alter the packing constraints around the metal center and hence influence gel formation. By forming the gels in situ, further refinement of the SAFIN composition should thus be possible by mixing different surfactant (or complexing) species rather than selecting a single gelator. George et al. found, for example, that although biphosphonate is not generally a good gelator in itself it could be added to the monophosphonate ester, acting as a cross linker and resulting in a similar gel with lower thermal stability.14 (19) Shchipunov, Y. A. Colloids Surf., A 2001, 183, 541. (20) Chen, S. J.; Evans, D. F.; Ninham, B. W.; Mitchell, D. J.; Blum, F. D.; Pickup, S. J. Phys. Chem. 1986, 90, 842. (21) Chen, V.; Evans, D. F.; Ninham, B. W. J. Phys. Chem. 1987, 91, 1823. (22) Chen, V.; Warr, G. G.; Evans, D. F.; Prendergast, F. J. Phys. Chem. 1988, 92, 768. (23) Chen, C.-M; Warr, G. G. J. Phys. Chem. 1992, 96, 9492. (24) Warr, G. G.; Sen, R.; Evans, D. F.; E.Trend, J. J. Phys. Chem. 1988, 92, 774.
Langmuir 2009, 25(15), 8810–8816
Article
Figure 2. Chemical structures of (mono)dodecyl phosphoric acid (m-C12, top), didodecyl phosphoric acid (center), and bis-(2-ethylhexyl) phosphoric acid (bis-C2C6, bottom).
In this work, we further investigate the effect of varying different components of the aggregate structure within the framework of in situ gel formation via the reaction described above (eq 1). This also has important industrial implications because the stability25 and viscosity26 of some basic gel formulations were found to be affected by the presence of other alkylphosphoric acids, such as monoalkyl phosphate and trace amounts of a trialkyl phosphate, which is a common byproduct found in industrial formulations. Using the previously investigated13 didodecyl phosphate (diC12) surfactant and aluminum isopropoxide complexing agent as a starting point, we examine the systematic substitution of two alternate molecules, (mono)dodecyl phosphate, m-C12, and bis-(2-ethylhexyl) phosphate, bis-C2C6, in the reaction mixture. The chemical structures of these molecules as well as that of the di-C12 molecule that is the basis for SAFIN formation are indicated in Figure 2. These compounds were chosen to provide microstructural control over the two factors governing SAFIN formation: the coordination chemistry of the SAFIN core is accessed by mixing the m-C12 diacid with the di-C12 monoacid, and the steric profile of the alkyl shell is controlled, without obvious change to the core chemistry, by blending the bulky, double-chained bis-C2C6 with the straight-chained di-C12 molecules. Using oscillatory shear rheology, we illustrate improved control over the performance of the aluminum alkyl phosphate organogel that can be achieved via these additional compositional parameters. Furthermore, this control can be explained in terms of predictable alterations to the microstructure of the aggregates in the SAFIN, as deduced by combined rheological and small-angle neutron scattering measurements, illustrating a route to the rational design of in situ metal alkoxide-surfactant molecular gels.
Materials and Methods Both didodecyl phosphoric acid (di-C12) and dodecyl phosphoric acid (m-C12) were gifts from Rhodia Inc. They were synthesized by Rhodia as described elsewhere.27 They were (25) Huddleston, D. A. Liquid Aluminium Phosphate Salt Gelling Agent; U.S. Patent 5,202,035, 1993. (26) Kim, V.; Bazhensov, A. V.; Kienskaya, K. I. Colloid J. 1997, 59, 455. (27) Reierson, R. L. Phosphation Reagent; U.S. Patent 6,136,221, 2000.
DOI: 10.1021/la900477v
8811
Article
Page and Warr
received separated and twice recrystallized and were used without further purification. Aluminum isopropoxide (99+%), bis-(2ethylhexyl) phosphate (bis-C2C6, 97%), decane (99+%), and perdeuterated decane (99 atom %) were all purchased from Sigma-Aldrich and used as received. Samples for rheological studies were prepared as described in ref 13. We refer to the molar ratio of aluminum to phosphate in the system as the Al/P ratio, where 1:3 Al/P is the stoichiometric equivalence point (eq 1). Herein we address only the roomtemperature behavior of the system. Liquid decane is heated to ∼40 °C to ensure dissolution of the surfactant component before addition of the aluminum source; however, because SAFIN formation is the result of the reaction between components and not a “curing” mechanism, we do not observe a dependence of the gel properties on this reaction temperature. The composition of the alkyl phosphate component was controlled by mixing stock solutions with a known proportion of substituent (m-C12 or bis-C2C6) and concentration (by mass) with a solution of unsubstituted didodecyl phosphate to the desired proportion of substituent. The total phosphate concentration was then adjusted by the addition of pure solvent, and aluminum isopropoxide was added to the desired Al/P ratio. Rheological measurements were performed at 25 ( 0.1 °C on a Reologica Stresstech controlled-stress rheometer in either coneand-plate or parallel plate geometry. Linear viscoelasticity at the selected strain amplitudes was found for amplitudes of e0.2 and was verified by strain-sweep experiments. In most cases, the rheological data are well described by a single relaxation time Maxwell model as described previously.13,28 Abbreviations m-C12 and bis-C2C6 are used interchangeably for both the surfactants themselves and the mole fraction of additive to total phosphate in the system, where the remainder consists of di-C12. SANS experiments were performed on the NG-3 30 m smallangle neutron scattering spectrometer at the National Center for Neutron Research at the National Institute of Standards and Technology (NIST), Gaithersburg, MD,29 using experimental conditions, calibration to absolute scattering intensity, and background subtraction as described previously.13 Data reduction and fitting was performed using the Igor routines provided by NIST.30
Figure 3. Schematic of aluminum didodecyl phosphate behavior as a function of temperature and Al/P ratio. Reproduced from The Journal of Physical Chemistry B.13
Results and Discussion In our previous study13 of the structure and gelation behavior of pure di-C12/Al organogels, we concluded that increasing Al/P caused a progression from an entangled network of locally cylindrical aggregates (Figure 1) into a branched (or cross-linked) physical gel network near the stoichiometric equivalence point, which proceeded to a saturated network that excluded excess solvent. This is summarized in the basic compositional phase diagram13 of aluminum isopropoxide-didodecyl phosphoric acid in n-decane reproduced in Figure 3. Neither m-C12 nor bis-C2C6 phosphates form gels alone mixed with aluminum isopropoxide in any proportions, although both have pKa’s31 such that they are expected to react with aluminum isopropoxide in the same manner as does di-C12. In the mixed systems, monododecyl phosphate and bis-(2ethylhexyl) phosphate are substituted into the Al-di-C12 system for a portion of the di-C12. We define the molar amount of either m-C12 or bis-C2C6 as a fraction of the total amount of phosphate in the system. This is done by dissolving the surfactants in known molar ratios in the decane before adding the aluminum isoprop(28) Ferry, J. D. Viscoelastic Properties of Polymers, 3rd ed.; Wiley: New York, 1980. (29) Glinka, C. J.; Barker, J. G.; Hammouda, B.; Krueger, S.; Moyer, J. J.; Orts, W. J. J. Appl. Crystallogr. 1998, 31, 430. (30) Kline, S. R. J. Appl. Crystallogr. 2006, 39, 895. (31) Ritcey, G. M.; Ashbrook, A. W. Solvent Extraction: Principles and Applications to Process Metallurgy; Elsevier: Amsterdam, 1984.
8812 DOI: 10.1021/la900477v
Figure 4. Storage (O) and loss (0) moduli for organogels with an added mole fraction of m-C12 = 0.1 as a function of Al/P, showing the transition from viscoelastic liquid to solid: Al/P = (a) 1.05:3, (b) 1.12:3, and (c) 1.16:3. Solid lines are single-relaxation Maxwell model fits.
oxide reactant (Materials and Methods). Because the Al/P ratio is expressed as a property of the aluminum phosphate core, it is unaffected by substitution with another phosphate surfactant, although, of course, the ratio corresponding to stoichiometric equivalence will vary when m-C12 is added to the formulation. Figure 4 shows oscillatory shear rheology measurements of systems in which 10 mol % di-C12 organogelator has been replaced with m-C12 at several Al/P ratios. In the absence of m-C12, the stoichiometric 1:3 ratio delineates a transition from liquidlike to solidlike viscoelastic properties (Figure 3). This mixed system shows a similar evolution; however, the solidlike behavior (that is, G0 > G00 at the lowest-accessed frequency of 0.01 Hz) occurs at a much higher Al/P (= 1.16:3) than in the pure diC12 gels. Significantly, at Al/P = 1.12:3, the gel is still rather liquidlike, something not observed even at just under Al/P = 1.0:3 in the absence of m-C12. Pure di-C12 gels exhibit a near-step transition between a metastable, viscoelastic fluid and a viscoelastic solid that exhibited long-term stability (>1 year) at Al/P = 1.0:3 (Figure 3). This is clearly illustrated by a steep increase in the relaxation time τ Langmuir 2009, 25(15), 8810–8816
Page and Warr
(determined from the frequency at which G0 = G00 ) with increasing Al/P ratio shown in Figure 5a for both the pure di-C12 gel and for m-C12 = 0.1. Substitution of 10 mol % m-C12 effectively shifts the transition to a larger Al/P ratio. Also shown in Figure 5a is the change in relaxation time as a function of mole fraction of m-C12 at a constant Al/P ratio of 1.2:3. As the dichain phosphate is increasingly substituted by its monochain analogue up to 50 mol %, the relaxation time continues to decrease. This corresponds to shifting the transition composition to higher and higher Al content. Figure 5b shows that the plateau modulus, G0, of gels containing m-C12 (and hence the average distance between entanglement points in the network, ξ, obtained from G0 = kBT/ξ3 32) increases monotonically as Al/P is increased. This contrasts with the behavior of the pure di-C12 system, for which G0 is constant across the gel transition. Taken in isolation, the effect of m-C12 on the liquid-solid transition composition is unremarkable: sequestration of some of the aluminum into nongelating Al/m-C12 complexes, leaving a lower effective Al/P ratio for the Al/di-C12 SAFIN gelating component, is sufficient to explain the composition shift semiquantitatively. However, the viscoelastic liquid phase containing 10 mol % m-C12 has long-term stability, just as for the fully gelled di-C12, whereas the di-C12-only system is only metastable at Al/P < ∼1.1:3.13 That is, the incorporation of m-C12 in all proportions studied, at constant Al/P, leads to long-term stabilization of viscoelastic liquid gels at room temperature by lowering the temperature of the boundary (Figure 3) between metastable and stable gels. The inclusion of m-C12 allows the relaxation time, indicating gel flow properties, to be adjusted independently as desired without loss of stability. This is a very important point from a rheological perspective, but it also indicates that m-C12 plays its own role in primary SAFIN formation and does not just form a separate, nongelling component. This will be discussed further below. Figure 6 shows the average relaxation time, τ, and plateau modulus, G0, of mixed Al/P organogels as a function of phosphate component composition as m-C12 and bis-C2C6 are added at a fixed Al/P near 1:3. As m-C12 is incorporated into the formulation, the relaxation time and plateau moduli both decrease, although at the Al/P of 1.2:3 shown here, the system maintains a solid gel-like structure up to around m-C12 = 0.3 (τ . 10 s) and remains viscoelastic until around m-C12 = 0.5 (G0 > 10 Pa). The relaxation time of the gel decreases almost exponentially, whereas the plateau modulus is almost unaffected up to 30 mol % replacement. This also argues against the formation of separate Al/m-C12 complexes, which would dilute the gel and lower G0 with increasing substitution. The chemical structure of the m-C12 headgroup may make it unfavorable for the formation of the aluminum-phosphatealuminum bridges found in the di-C12 gel. This is in good agreement with observations on aluminum alkylphosphonate gels, for which no gelation was observed in a range of liquids, even at the stoichiometric Al/P ratio of 2:3.14 A reduction in the number of bridging phosphates may decrease the breaking time of the threads, which would decrease the overall structural relaxation time without altering the network density. The incorporation of bis-C2C6 leads to somewhat different behavior. The gel maintains constant τ and G0 up to about bis-C2C6 = 0.2, beyond which gelation is lost over the range of (32) De Gennes, P.-G. Scaling Concepts in Polymer Physics; Cornell University Press: New York, 1979.
Langmuir 2009, 25(15), 8810–8816
Article
Figure 5. Rheological characteristics of mixed aluminum mono/ didodecyl phosphate/decane organogels at a mass fraction of φ = 0.85% w/w. (a) Relaxation time, τ, versus the Al/P ratio for pure diC12 (O, as reported in ref 13) and with m-C12 = 0.1 (b). The solid lines are guides to the eye, indicating the approximate range of the transition in relaxation time at m-C12 = 0, 0.1, and 0.5. Also shown is the effect on τ of increasing the m-C12 mole fraction up to 0.5 at a constant Al/P ratio of 1.2:3 (b). (b) Plateau modulus, G0, versus the Al/P ratio for pure di-C12 (0, as reported in ref 13) and with m-C12 = 0.1 (9).
Figure 6. Relaxation time (τ, top, ), m-C12; (, bis-C2C6) and plateau modulus (G0, bottom, 0, mC12; 9, bis-C2C6) as a function of the mole fraction of added m-C12 (Al/P = 1.2:3) and bis-C2C6 (Al/P = 1.0:3). The total gelator concentration is 1.5% w/w. Open circles show the relaxation times and plateau moduli of pure di-C12 gels over a range of Al/P ratios near 1.0:3.
0.2 < bis-C2C6 < 0.3. However bis-C2C6 differs from di-C12 only in the surfactant tails, which are not directly involved in the structure of the aggregate core. Thus, the absence of gelling in this case is most probably due to an unfavorable shell structure resulting from the bulkier tails. A similar effect was observed in the case of so-called aromatic linker steroid organogels, where a conformational change induced in the steroid by CdC drastically alters the gelating ability of the molecule.12 Up to this point, we have shown how the inclusion of m-C12 immediately provides improved control over gel performance: at low (several months). Aluminum and di-C12, that is, the double-tailed surfactant and single acid group, in the stoichiometric 1:3 ratio (or less)
8816 DOI: 10.1021/la900477v
form independent rodlike aggregates. From this and previous work,14,16 it appears that the di-C12 structure is ideal for bridging between neighboring surfactant ligands, which is the driving force for SAFIN formation. The two primary factors that determine the SAFIN structure are the coordination chemistry in the core of the aggregates and packing effects in the alkyl shell. Incorporation of diacid m-C12 into di-C12 organogels modifies the core structure. The coordination chemistry of m-C12 prevents SAFIN formation by suppressing ligand bridging. In small quantities mixed with di-C12, however, it instead results in an equivalent manner to reducing the Al/P ratio, with the diacid reducing the amount of Al relative to stoichiometric equivalence. Simultaneously, the single-tailed surfactant reduces the critical stability temperature, preventing phase separation that is observed below 40 °C in pure di-C12 gels with Al deficiency (Figure 3). This therefore allows a stable, unbranched, Aldeficient SAFIN, yielding low relaxation time gels. The bulky tail of bis-C2C6 also prevents SAFIN formation by sterically hindering ligand bridging along the 1D growth axis. Incorporating bis-C2C6 in small quantities with di-C12 thus provides a way of modifying the aggregates’ shell structure. Its effect is first to increase the chain packing density around the aluminum ions. At increasing concentrations, it reduces the gelling efficiency by acting as an end-capping agent, shortening the average length of the aggregates, and also by controlling the degree of branching (as observed by the measuring τ) in the network. Acknowledgment. We acknowledge funding from the Australian Research Council SPIRT scheme, the Australian Nuclear Science and Technology Organisation, and Rhodia Inc. as well as the support of the Rheology Group in the Department of Mechanical Engineering, University of Sydney, in providing facilities used in this work. This work utilized facilities supported in part by the National Science Foundation under agreement no. DMR-9986442. We acknowledge the support of the National Institute of Standards and Technology, U.S. Department of Commerce, in providing the neutron research facilities used in this work.
Langmuir 2009, 25(15), 8810–8816
