Langmuir 2009, 25, 761-768
761
Self-Association of Short-Chain Nonionic Amphiphiles in Binary and Ternary Systems: Comparison between the Cleavable Ethylene Glycol Monobutyrate and Its Ether Counterparts Ying Zhu,† Anne-Gae¨lle Fournial,‡ Vale´rie Molinier,† Nathalie Azaroual,‡ Gaston Vermeersch,‡ and Jean-Marie Aubry*,† LCOM, Equipe “Oxydation et Formulation”, UMR CNRS 8009, ENSCL BP 90108, F-59652 VilleneuVe d’Ascq Cedex, France, and LCOM, Equipe “RMN et Photochimie”, UMR CNRS 8009, Faculté de Pharmacie, UniVersité de Lille 2, BP 83, 59006 Lille Cedex, France ReceiVed September 9, 2008. ReVised Manuscript ReceiVed October 30, 2008 In the context of environmental concerns for the production of surface active species, the introduction of a carbonyl function into the skeleton of ethyleneglycol-derived solvo-surfactants is a way to access cleavable compounds with presumed enhanced biodegradability. Ethylene glycol monobutyrate (C3COE1) was synthesized and compared to its ether counterparts, ethylene glycol monopropyl (C3E1) and monobutyl ethers (C4E1), to assess the effect of the insertion of a carbonyl function in the skeleton of short-chain ethoxylated amphiphilic compounds. In aqueous solutions, the ester has intermediate behavior between that of the two ethers with regard to surface tension, solubilization of Menaphtalene in water, and self-diffusion by PGSE NMR. In ternary systems, C3COE1 and C3E1 have the same optimal oil (EACN ) 2.8), which is much more polar than that of C4E1 (EACN ) 8.5). With regard to the ability to form structured systems, the behavior in water does not differ significantly for the three compounds, and the transition between nonassociating solvents and amphiphilic solvents, sometimes called solvo-surfactants, is gradual. In ternary systems, however, only C4E1 and C3COE1 form a third phase near the optimal formulation, which tends to show that C3COE1 possesses the minimum amphiphilicity to get a structuration. Self-diffusion NMR studies of the one-phase domains do not, however, allow us to distinguish between different degrees of organization in the three systems.
1. Introduction The search for environmentally safer surface-active species is a key issue in research programs related to formulation in both the academic and industrial communities. In this context, esters of polyethylene glycol are cleavable compounds that may represent more environmentally friendly alternatives to the widely used ethers. Tehrani-Bagha and Holmberg have shown that the octyl, 1-methylhexyl, and 1-ethylhexyl esters of tetraethylene glycol reach 60% biodegradation after 28 days in a closed-bottle test and that the central scission between the lipophilic and the hydrophilic parts is the rate-determining step.1 In a recent publication, the well-defined decanoyl esters of tri- and tetraethylene glycol (C9COEj, j ) 3, 4) were synthesized and their physicochemical properties in aqueous solutions investigated in comparison with their ether counterparts (C10Ej).2 The precise phase boundaries of the liquid-crystal phases, especially the liquidcrystal-isotropic-liquid coexisting regions, were determined by NMR spectroscopy.3 In aqueous solutions, it was found that the ester surfactants exhibit similar, although somewhat more polar, behaviors compared to the ether counterparts. The ester surfactants form only a lamellar liquid-crystal phase, the domain of which is smaller than that of the ether with regard to both temperature and concentration. 2H NMR studies of the liquid-crystal phases show a reduced order parameter compared to that of the ether homologues. With regard to the stability of such compounds over time, it has been shown that the hydrolysis rate is higher * Corresponding author. E-mail:
[email protected]. † ENSCL. ‡ Universite´ de Lille 2. (1) Tehrani-Bagha, A.; Holmberg, K. Curr. Opin. Colloid Interface Sci. 2007, 12, 81–91. (2) Zhu, Y.; Molinier, V.; Queste, S.; Aubry, J. M. J. Colloid Interface Sci. 2007, 312, 397–404. (3) Fournial, A. G.; Zhu, Y.; Molinier, V.; Vermeersch, G.; Aubry, J. M.; Azaroual, N. Langmuir 2007, 23, 11443–11450.
Figure 1. Compounds under study: monobutyrate of ethylene glycol (C3COE1) and its ether counterparts, ethylene glycol butylether (C4E1), and ethylene glycol propylether (C3E1).
for concentrations below the critical micellar concentration and that it is highly pH-dependent.1,2,4 The monodecanoyl ester of tetraethylene glycol C9COE4 has a half-life of 906 min at pH 11.0 and 90 min at pH 12.1,2 whereas ester-containing surfactants are known to be stable for more than 1 month at neutral or slightly alkaline pH.1,4 In the present work, the same substitution of an ether bond by an ester linkage was performed on a shorter homologue, with the aim of finding substitutes for the short-chain ethylene glycol ethers because several members of this family of compounds are blamed for reprotoxicity.5,6 It is expected that the insertion of a cleavable function will also increase the biodegradability in the case of short-chain amphiphiles. The pure monobutyrate of ethylene glycol (C3COE1) was synthesized and compared to its ether counterparts, the butylether of ethylene glycol (C4E1) and the propylether of ethylene glycol (C3E1) (Figure 1). The influence of the insertion of an ester bond on the physicochemical properties is expected to be greater in this case than for the already-studied longer homologues. With a low molecular weight and an amphiphilic structure, C3E1, C4E1, and C3COE1 belong to a family of nonionic short(4) Nardello, V.; Chailloux, N.; Joly, G.; Aubry, J. M. Colloids Surf., A 2006, 288, 86–95. (5) The Toxicology of Glycol Ethers and Its ReleVance to Man; Technical Report No. 64; European Center for Ecotoxicology and Toxicology of Chemicals: Brussels, 1995. (6) Laudet-Hesbert, A. Toxicol. Lett. 2005, 156, 51–58.
10.1021/la802963j CCC: $40.75 2009 American Chemical Society Published on Web 12/15/2008
762 Langmuir, Vol. 25, No. 2, 2009
chain amphiphilic compounds that are sometimes called solvosurfactants. These compounds are of great interest both in academic research and industrial applications because they accumulate some typical properties of solvents (volatility, solubilization · ) and of surfactants (surface activity, self-aggregation).7 The short-chain (poly)ethylene glycol ethers, sometimes simply called glycol ethers (CiEj, j < 3), are the most commonly used solvo-surfactants in industry. The behavior of solvo-surfactants in binary and ternary systems, which is significantly different from that of “true” surfactants, can be studied in a similar manner. In an aqueous environment, surfactant molecules, because of their distinct and bulky hydrophilic and hydrophobic groups, have the ability to self-associate and exhibit complex phase behavior at high concentrations with the formation of various liquid-crystal structures. The autoassociation is a cooperative phenomenon, the onset taking place at a precise concentration that is the critical micellar concentration (cmc). Solvo-surfactants, which also possess distinct but smaller hydrophilic and lipophilic regions, also adsorb at the interface and are believed to undergo some kind of aggregation. The exact nature of this aggregation phenomenon, in particular, whether it takes place in a cooperative manner, is still a matter of debate. The aggregation behavior of short-chain amphiphilic species belonging to short-chain ethylene glycols8-19 or alcohols20-23 has been extensively studied by various techniques. In our work, the aggregation behavior of C3COE1 and its ether counterparts in water was investigated by surface tension and PGSE NMR experiments. In equilibrium ternary water/oil/surfactant systems, the type of system formed (Winsor I, II, or III) depends on many formulation variables, such as the chemical structures of oil and surfactant, the temperature, and the salinity.24 Near the optimal formulation, where the surfactant has a balanced affinity for oil and water, a triphasic system with a bicontinuous microemulsion in equilibrium with excess water and oil (Winsor III system) can be formed. Solvo-surfactants show phase behavior similar to that of “true” surfactants except that a much higher concentration of amphiphile is required to obtain a Winsor III-type system. Moreover, the minimal interfacial tension between the excess oil and water phases attained at the optimal formulation is much higher (∼0.1 mN · m-1) for solvo-surfactants than in the case of real surfactants (∼10-3-10-4 mN · m-1). The organization of microemulsions formed by short-chain amphiphiles appears to (7) Lunkenheimer, K.; Schroedle, S.; Kunz, W. Prog. Colloid Polym. Sci. 2004, 126, 14–20. (8) Ambrosone, L.; Costantino, L.; D’Errico, G.; Vitagliano, V. J. Colloid Interface Sci. 1997, 190, 286–293. (9) D’Arrigo, G.; Mallamace, F.; Micali, N.; Paparelli, A.; Vasi, C. Phys. ReV. A 1991, 44, 2578–2587. (10) D’Arrigo, G.; Teixeira, J.; Giordano, R.; Mallamace, F. J. Chem. Phys. 1991, 95, 2732–2737. (11) Kato, T. J. Phys. Chem. 1985, 89, 5750–5755. (12) Kilpatrick, P. K.; Davis, H. T.; Scriven, L. E.; Miller, W. G. J. Colloid Interface Sci. 1987, 118, 270–285. (13) Koehler, R. D.; Schubert, K. V.; Strey, R.; Kaler, E. W. J. Chem. Phys. 1994, 101, 10843–10849. (14) Mallamace, F.; Micali, N.; D’Arrigo, G. Phys. ReV. A 1991, 44, 6652– 6658. (15) Nishikawa, S.; Tanaka; Mashima, M. J. Phys. Chem. 1981, 85, 686–689. (16) Onori, G.; Santucci, A. J. Phys. Chem. B 1997, 101, 4662–4666. (17) Ortona, O.; Vitagliano, V.; Paduano, L.; Costantino, L. J. Colloid Interface Sci. 1998, 203, 477–484. (18) Quirion, F.; Magid, L. J.; Drifford, M. Langmuir 1990, 6, 244–249. (19) Elizalde, F.; Gracia, J.; Costas, M. J. Phys. Chem. 1988, 92, 3565–3568. (20) D’Arrigo, G.; Giordano, R.; Teixeira, J. Langmuir 2000, 16, 1553–1556. (21) D’Arrigo, G.; Giordano, R.; Teixeira, J. Eur. Phys. J. E 2003, 10, 135– 142. (22) Frindi, M.; Michels, B.; Zana, R. J. Phys. Chem. 1991, 95, 4832–4837. (23) Hajji, S. M.; Errahmani, M. B.; Coudert, R.; Durand, R. R.; Cao, A.; Taillandier, E. J. Phys. Chem. 1989, 93, 4819–4824. (24) Salager, J. L.; Ranto´n, R.; And´; erez, J. M.; Aubry, J. M. Tech. Ing., Ge´nie Proce´de´s 2001, J2157/1–J2157/20.
Zhu et al.
be very different from the well-structured ones of the true surfactants. Amphiphiles are usually classified as strong and weak according to their amphiphilicity factor,13 with the shortchain polyglycol ethers being part of the latter group.25,26 Weak amphiphiles lead to weakly structured mixtures in opposition to the strongly structured microemulsions formed by real amphiphiles. In this work, the phase behavior of ethylene glycol monobutyrate and its ether counterparts in the one-phase domain of their optimal ternary systems was examined by PGSE NMR in order to put forward differences in the organizational state in each case.
2. Experimental Section 2.1. Chemicals. Butyric acid (99%), 2-chloroethanol (99%), sodium hydroxide (97+%), ethylene glycol monopropyl ether (99.4%), and ethylene glycol butyl ether (99.5%) were purchased from Sigma-Aldrich. The alkanes and alkylbenzenes used for the construction of “fish” diagrams were obtained from Sigma-Aldrich or Fluka with the highest available grades. Toluene (99%) used for extractions was from Acros. All chemicals were used as received. 2.2. General Methods. Analytical NMR spectra were recorded on Brucker AC spectrometers at 300.13 MHz for 1H and 75.47 MHz for 13C. Gas chromatography analyses were performed on an Agilent 6890N apparatus, equipped with a HP-1 cross-linked methyl silicone gum column (60 m × 0.32 mm × 0.25 µm), with N2 as the gas vector and an FID detector. 2.3. Synthesis. Butyric acid (44 g, 0.5 mol) was dissolved in 40% sodium hydroxide solution (50 g, 0.5 mol, 1 equiv) at 95 °C with stirring. 2-Chloroethanol (48 g, 0.6 mol, 1.2 equiv) was slowly added over 45 min at the same temperature. The reaction was maintained at 95 °C for 8 h under reflux and agitation. After cooling to room temperature, the medium was filtered and extracted with toluene (3 × 100 mL). The organic phase was washed with a saturated solution of sodium hydrogenocarbonate until reaching neutral pH and was then dried over magnesium sulfate. After solvent evaporation under reduced pressure, a colorless liquid (28.5 g, 0.22 mol, 43% yield) was obtained. The crude product is a mixture of monoester, diester, and residual butyric acid. The pure monoester was obtained by distillation under reduced pressure (T ) 58 °C, P ) 5 Pa). The purity of the compound was ascertained by 1H and 13C NMR spectroscopy and GC. The product was kept in the refrigerator under an argon atmosphere and was used within 15 days to avoid slow hydrolysis taking place in moist air. Monobutyrate of ethylene glycol (C3COE1): 1H NMR (300 MHz, CDCl3-1% TMS): δ 0.96 (3H, t, J ) 7.5 Hz, CH3), 1.67 (2H, sx, J ) 7.5 Hz, CH2β), 2.34 (2H, t, J ) 7.5 Hz, CH2R), 2.70 (1H, t, J ) 5.8 Hz, -OH), 3.81-3.86 (2H, m, CH2RO), 4.20-4.23 (2H, m, CH2βO). 13C NMR (75 MHz, CDCl31%TMS): δ 13.7 (CH3), 18.4-36.1 (2CH2), 61.3-65.9 (2CH2O), 174.1 (COO). 2.4. Surface and Interfacial Tension Measurements. Isothermal surface and interfacial tensions were measured with an ITConcept tensiometer in rising drop mode. The temperature was maintained at 25.0 ( 0.2 °C by using a thermostatted cell. The data reported are equilibration values typically obtained after a time interval of 5 to 60 min for surface tension and 5 min for interfacial tension. Six measurements were taken for each point. The data reported are the mean values. 2.5. Solubilization of 1-Methylnaphthalene. Solutions (2 g each) of the hydrotrope in water were prepared at different concentrations, and 1-methylnaphthalene was added carefully until reaching saturation (cloudy solution). The solutions were stirred at room temperature for 24 h. After this period, the solutions were centrifuged to accelerate the phase separation, and the aqueous phase was removed after complete decantation (clear solution). The amount of 1-methylnaphthalene solubilized in the aqueous phase was determined by UV absorption of the solutions at 282 nm. Prior to the measurement, (25) Gradzielski, M.; Langevin, D.; Sottmann, T.; Strey, R. J. Chem. Phys. 1997, 106, 8232–8238. (26) Kahlweit, M.; Strey, R.; Busse, G. Phys. ReV. E 1993, 47, 4197–4209.
Self-Association of Short-Chain Amphiphiles
Langmuir, Vol. 25, No. 2, 2009 763
a calibration curve was constructed at this wavelength. The solutions were diluted in ethanol before measurement. 2.6. Diffusion Coefficients by PGSE NMR. The self-diffusion coefficients were determined by the PGSE-NMR technique by monitoring the 1H signal on a Bruker Avance 500 spectrometer equipped with a field gradient probe unit. This method was first introduced by Stejskal and Tanner,27 and in our experiments, the BPP-STE-LED sequence28 combining constant time, stimulated echo, bipolar pulse, and the longitudinal eddy current delay method was used. The 2D DOSY treatment was performed with NMRnotebook software by NMRtec. All of the NMR signals give rise to monoexponential decays following eq 1
[
(
A ) A0 exp -γH2δ2G2 ∆ -
δ D 3
)]
(1)
where A is the echo amplitude in the presence of the gradient pulse, A0 is the echo amplitude in the absence of the gradient pulse, γH is the proton gyromagnetic ratio, δ is the gradient pulse length, G is the strength of the applied field gradient, ∆ is the interval between two field gradient pulses, and D is the diffusion coefficient. The gradient strengths were varied from 1 to 35 G cm-1, and parameters δ and ∆ were adjusted for each sample to obtain the full decrease in the echo signal. By varying the field gradient amplitude G, a series of experiments were collected and the diffusion coefficient was extracted using a simple fit to eq 1 for well-separated resonances. The main methylene peak was chosen to fit eq 1. The temperature control of the probe was within (0.5 K. The field gradient was calibrated with the diffusion of H2O in H2O/D2O mixtures. In binary systems, measurements were performed at 10 °C to be far below the cloud temperature of C4E1 (43 °C) and for comparison with the data previously obtained with the longest homologues.3 In ternary systems, measurements were performed at 17 °C to be in the optimal temperature range of the systems. 2.7. Fish Diagrams. Ternary S/O/W systems were prepared in test tubes with the studied solvo-surfactant and chosen oil. Equal weights of oil and water were first introduced, and increasing amounts of solvo-surfactant were added. After each addition, the test tubes were shaken vigorously and placed in a thermostatted bath at 25.0 ( 0.1 °C until the attainment of equilibrium. The type of Winsor system (I, II, III, and IV) was determined by visual observation and in some cases by using a laser pointer to locate the microemulsion phase.
3. Results and Discussion 3.1. Autoassociation in Water. 3.1.1. Onset of Aggregation by Surface Tension Data and Solubilization Experiments. The evolution of surface tension with the solvo-surfactant molar fraction was measured for C3E1, C3COE1, and C4E1 at 25.0 °C and compared to the evolutions obtained for ethylene glycol (E1), ethylene glycol methyl ether (C1E1), and ethylene glycol ethyl ether (C2E1). The data are presented in Figure 2 on a semilogarithmic scale. The curves obtained in the case of an ideal mixture between water and a solvent having surface tensions of 47.8 and 22.4 mN/m are also presented for comparison. A slight change in the slope starts to be observed with C1E1 and becomes more pronounced when the amphiphilic character is increased (C2E1, C3E1, C3COE1, and C4E1). For ethylene glycol (E1) that does not possess any amphiphilic character, no break point is observed even if the behavior is significantly different from that of an ideal mixture. A clear break point can be set only in the case of C4E1, with the evolution becoming smoother for C3COE1 and even more so for C3E1, C2E1, and C1E1. In these (27) Stejskal, E. O.; Tanner, J. E. J. Chem. Phys. 1965, 42, 288–292. (28) Wu, D.; Chen, A.; Johnson, C. S., Jr. J. Magn. Reson., Ser. A 1995, 115, 260–264.
Figure 2. Surface tension as a function of the molar fraction of E1 (]), C1E1 (*), C2E1 (+), C3E1 (0), C3COE1 (2), and C4E1 (∆) at 25.0 °C. Comparison with the curves obtained in the case of an ideal mixture between water and a solvent having surface tensions of 47.8 mN/m (---) and 31.0 mN/m (...).
latter cases, it should also be pointed out that the surface tension does not reach a plateau at high concentrations as for C4E1. A change in the surface tension versus concentration curve slope is usually observed for short-chain amphiphiles19 and coincides with the concentration at which other physicochemical properties, such as the density, molar volume, and heat capacity, are modified.12 This point is thus generally taken as the onset of aggregation and is called the minimum aggregation concentration (mac). In our case, if we try to work out a breakpoint in the case of the three compounds of interest, the mac’s are found at xc ) 0.017 for C4E1, xc ) 0.020 for C3COE1, and xc ) 0.080 for C3E1. The critical aggregation concentration found for C4E1 is in accordance with the data already published that were obtained by different techniques (cf. Table 1). C4E1 and C3COE1 are found to start aggregating almost at the same concentration, which is much lower than for C3E1. As with their longer homologues, the association phenomenon in the case of short-chain amphiphiles is believed to be driven by hydrophobic effects. This indicates that both compounds have similar hydrophobic parts. Consequently, to assess the hydrophilic/lipophilic balance of the compounds, the carbonyl function should be considered to belong to the hydrophobic chain and not be included in the hydrophilic head. This was also observed for the longest homologues because the behavior of C9COE4 compared well with that of its ether counterpart, C10E4.2,3 Short-chain amphiphilic molecules such as the ones under study can be classified as hydrotropes, a term first introduced to define water-soluble organic molecules having an amphiphilic structure and able to increase the water solubility of hydrophobic compounds greatly.29 Historically referring to anionic shortchain aromatic salts, this term was later extended to cationic and nonionic aromatic compounds and also to short-chain aliphatic compounds exhibiting hydrotropic behavior. In the latter case, the term solvo-surfactant may be preferred for nonionic shortchain amphiphilic molecules.7 The ability of hydrotropes to increase the solubility of organics in water significantly is believed to be linked to their ability to self-associate at high concentrations, above the so-called minimum hydrotropic concentration (mhc).30 Above this concentration, some kinds of aggregates, the structures of which are not yet fully understood, are formed, thus creating concentrated (29) Matero, A. In Handbook of Applied Surface and Colloid Chemistry; Holmberg, K., Shah, D. O., Schwuger, M. J., Eds.; John Wiley & Sons: Chichester, U.K., 2002; Vol. 1, Chapter 18. (30) Hodgdon, T. K.; Kaler, E. W. Curr. Opin. Colloid Interface Sci. 2007, 12, 121–128.
764 Langmuir, Vol. 25, No. 2, 2009
Zhu et al.
Table 1. Literature Data Concerning the Aggregation Behavior of Ethyleneglycol Monobutyl Ether (C4E1) in Water at 25.0 °C ref
technique
9
ultrasonic and brillouin lightscattering experiments SANS PGSE NMR densitometry, refractometry, tensiometry, 13C NMR microcalorimetry, tensiometry SANS, light scattering tensiometry, foam stability SANS
10 11 12 16 18 19 21
XC
type of aggregate
radius (Å)
micelle-like structures 0.015-0.02 0.018 0.02 0.018 0.017 0.018
hydrophobic regions that aid the solubilization of organics in water, as do micelles in surfactant solutions. In fact, the point at which a compound stops being a hydrotrope and becomes a real surfactant is a matter of debate. Srinivas et al.31 studied the behavior of a series of alkyl benzene sulfonates of increasing chain length and concluded that a clear distinction between hydrotropic and a micellar behavior is not possible because the transition is gradual. Bauduin et al.32 based their work on the shapes of the solubilizing curves of hydrophobic dyes by hydrotropes to determine the mhc. They suggested the use of “co-surfactant”, “solvo-surfactant”, and “hydrotrope” as synonyms, with only the mhc values providing data reliable enough to distinguish the species. Hydrotropic solubilizations of 1-methylnaphthalene by the three compounds under study have been performed and are presented in Figure 3. 1-Methylnaphthalene was chosen because it can be considered to be a good mimic of organic soil and has the advantage of absorbing in the UV region, which makes quantification easier. The concentrations of hydrotrope from which the solubilization of 1-methylnaphthalene starts to be effective, the so-called minimum hydrotropic concentrations, are in good agreement with the minimum aggregation concentrations obtained from surface tension measurements for C4E1 and C3COE1 (ca. 0.9 mol/L/0.016 molar fraction for the former and 1.2 mol/L/0.021 molar fraction for the latter). However, for C3E1, the onset of solubilization increase is not clear, and if we try to work out an mhc, then it is not in agreement with the surface tension data because it seems to occur at lower concentrations. This tends to hint at aggregation taking place at a given concentration for C3COE1 and C4E1, whereas the behavior of C3E1 seems to deviate more from truly associative behavior. 3.1.2. NMR Self-Diffusion. Compared to real surfactants, hydrotropes start aggregating at higher concentrations, and the onset is not as well-defined, which calls into question the exact nature of the aggregates formed. Srinivas et al.33 studied the X-ray crystal structure of conventional hydrotropes (sodium benzenesulfonate derivatives) and concluded that open-layer small assemblies with a small aggregation number of hydrotropes were formed. This layered structure would explain the solubilizing action of hydrotropes that are able to include hydrophobic guest molecules within the layers. Numerous studies on the types of aggregates have been performed in the ethyleneglycol butyl ether (C4E1)-water system using various physicochemical techniques. All results tend to prove that C4E1 aggregates from a critical molar fraction of x ) 0.018 in the form of spherical aggregates of the micelle type (Table 1). (31) Srinivas, V.; Balasubramanian, D. Langmuir 1998, 14, 6658–6661. (32) Bauduin, P.; Renoncourt, A.; Kopf, A.; Touraud, D.; Kunz, W. Langmuir 2005, 21, 6769–6775. (33) Srinivas, V.; Rodley, G. A.; Ravikumar, K.; Robinson, W. T.; Turnbull, M. M.; Balasubramanian, D. Langmuir 1997, 13, 3235–3239.
micelle-like structures short-lived micelles weakly cooperative aggregates with a short lifetime (
