Properties of New Glucamide Surfactants - Langmuir (ACS Publications)

---

Langmuir 1996, 12, 2701-2705

2701

Properties of New Glucamide Surfactants Julian Eastoe* and Philippe Rogueda School of Chemistry, University of Bristol, Cantock’s Close, Bristol BS8 1TS, U.K.

Andrew M. Howe and Alan R. Pitt Kodak European Research, Headstone Drive, Harrow HA1 4TY, U.K.

Richard K. Heenan ISIS, Rutherford Appleton Laboratory, Chilton OX11 0QZ, U.K. Received October 3, 1995. In Final Form: March 13, 1996X Properties of new nonionic surfactants are reported. The amphiphiles possess two n-alkyl chains and two glucamide head-groups and have the formula (CnH2n+1)2C[CH2NHCO(CHOH)4CH2OH]2 with n ) 5-9 (abbreviated here as di-(Cn-Glu)). For di-(C7-Glu) in the concentration range 0.4-8%, an upper critical solution temperature (UCST) is observed; this is unusual behavior for a nonionic surfactant. Nonetheless these materials behave like classic surfactants with regard to changes in liquid crystal phases, surface tensions, and critical micelle concentrations (cmc’s) as a function of chain length n. Small-angle neutron scattering (SANS) data indicate the presence of anisotropic micelles, cylindrical for di-(C5-, 6-, and 7-Glu) but discoidal for di-(C8-Glu). There is apparently little effect of temperature (25-70 °C) on cmc or micelle structures.

Introduction Alkyl saccharide surfactants may be wholly or partly derived from renewable resources, and these materials are currently of interest from both academic and commercial viewpoints.1,2 Here we report on the solution behavior of a new class of such nonionic ‘sugar surfactants’ possessing a novel molecular architecture: two identical opened glucamide groups linked to two equal n-alkyl chains. The materials form an homologous series, and the generic formula is (CnH2n+1)2C[CH2NHCO(CHOH)4CH2OH]2 where n ) 5-9 (Figure 1).3 The abbreviation di(Cn-Glu) is used to reflect this structure. The most common nonionic surfactants are n-alkyl polyoxyethylene glycol ethers (CiEj’s), which display a characteristic ‘cloud point’ or lower critical solution temperature (LCST).4 For example, with C14E7 at a concentration of approximately 1.8%, phase separation occurs at ∼57 °C.5 The critical micelle concentrations (cmc’s) show a small decrease with temperature. Depending on the CiEj surfactant the micelle size may, or may not, vary with temperature; however, attractive interactions between micelles always increase sharply as the LCST is approached.4,6 The properties of analogous single-chain, single-head, sugar-based surfactants have been investigated previously.7-16 In some cases, unusual behavior is observed, such as gel formation at low concentrations or insolubility; X

Abstract published in Advance ACS Abstracts, May 1, 1996.

(1) Salka, B. Cosmet. Toiletries 1993, 108, 89. (2) Balzer, D. Tenside, Surfurfacts Deterg. 1991, 28, 419. (3) (a) Eastoe, J.; Rogueda, P.; Harrison, W. J.; Howe, A. M.; Pitt, A. R. Langmuir 1994, 10, 4429. (b) Briggs, C. B. A.; Newington, I. M.; Pitt, A. R. J Chem. Soc., Chem. Commun. 1995, 3, 379. (c) U.S. Patent No. 4, 892, 806, 1990. (4) Zulauf, Z.; Weckstrom, K.; Hayter, J. B.; Degiorgio, V.; Corti, M. J. Phys. Chem. 1985, 89, 3411. (5) Corti M.; Mineiro C.; Degiorgio V. J. Phys. Chem. 1984, 88, 309. (6) Lesemann, M.; Belkoura, L.; Woermann, D. Ber. Bunsen-Ges. Phys. Chem. 1995, 99, 695. (7) Pfannemu¨ller, B.; Welte, W. Chem. Phys. Lipids 1985, 37, 227. (8) Mu¨ller-Fahrnow, A.; Hilgenfeld, R.; Hesse, H.; Saenger, W. Carbohydr. Res. 1988, 176, 165. (9) Kida, T.; Morishima, A.; Nakatsuji, Y. J. Am. Oil. Chem. Soc. 1994, 71, 705.

S0743-7463(95)00827-4 CCC: $12.00

Figure 1. Molecular structure of the glucamide surfactants, referred to as di-(Cn-Glu) in the text.

however, such problems do not arise with most of the new glucamide surfactants described here. Experimental Section Glucamide surfactants were supplied by Kodak.3 From initial surface tension measurements it was found that only di-(C5Glu) contained any organic impurity (amine), which was removed using a cation-exchange resin (Amberlite IRC-50 BDH). All other surfactants were used as received but stored in a dessicator over refreshed P2O5 (Lancaster). Elemental microanalysis, as well as 1H and 13C NMR (JEOL GX400), was carried out in the School of Chemistry (University of Bristol). Water was purified using a Purite reverse-osmosis/ion-exchange system. D2O was obtained from CDN Isotopes (99.5% D atom). Binary surfactant-water phase behavior was determined using accepted procedures: visual inspection of thermostated systems (Grant LTD 6G); polarizing optical microscopy (Nikon Optiphot II and Linkam hot stage), and 2H NMR (JEOL JM-FX 100 or GX400). Airsolution surface tensions were measured using a Kru¨ss K12 tensiometer and Pt-Ir du Nouy ring, calibrated against a range of standard liquids and thermostated to within (0.05 °C. Small-angle neutron scattering (SANS) measurements were performed using the LOQ spectrometer on the ISIS pulsed (10) Fuhrhop, J. H.; Svenson, S.; Boettcher, C.; Ro¨ssler, E.; Vieth, H. M. J. Am. Chem. Soc. 1990, 112, 4307. (11) Zabel, V.; Mu¨ller-Fahrnow, A.; Hilgenfeld, R.; Saenger, W.; Pfannemu¨ller, B.; Enkelmann, V.; Welte, W. Chem. Phys. Lipids 1986, 39, 313. (12) La Mesa, C.; Bonincontro, A.; Sesta, B. Colloid Polym. Sci. 1993, 271, 1165. (13) Denkinger, P.; Kunz, M.; Burchard, W. Colloid Polym. Sci. 1990, 268, 513. (14) Frindi, M.; Michels, B.; Zana, R. J. Phys. Chem. 1992, 96, 8137. (15) Platz, G.; Thuning, C.; Po¨licke, J.; Kirchhoff, W.; Nickel, D. Colloids Surf., A 1994, 88, 113. (16) Auvray, X.; Petipas, C.; Anthore, R. Langmuir 1995, 11, 433.

© 1996 American Chemical Society

2702

Langmuir, Vol. 12, No. 11, 1996

Eastoe et al.

Table 1. Elemental CHN Analysis of di-(Cn-Glu) SurfactantssExpected (e) and Found (f) C/%

di-(C5-Glu) di-(C6-Glu) di-(C7-Glu) di-(C8-Glu) di-(C9-Glu)

H/%

formula

Mr

e

f

C25H50N2O12 C27H54N2O12 C29H58N2O12 C31H62N2O12 C33H66N2O12

570.89 598.95 627.01 655.07 683.13

52.6 54.1 55.5 56.8 58.0

51.5 54.0 52.0 55.0 56.5

e

a

N/% f

8.9 9.0 9.1 9.6 9.3 9.7 9.6 9.9 9.8 10.0

e

f

4.9 4.7 4.4 4.3 4.1

4.8 4.7 4.2 4.1 3.9

neutron source at the EPSRC Rutherford Appleton Laboratory, U.K. The measurements determine the absolute scattering probability I(Q) (cm-1) as a function of momentum transfer Q (Å-1) ) (4π/λ) sin(θ/2) with λ the incident neutron wavelength (2.2-10 Å) and θ the scattering angle ( 0.04 Å-1 21 ) is also included in the modeling routines.

c

Results and Discussion 1. Surfactant Purity. The CHN elemental analysis for di-(Cn-Glu) surfactants, and a comparison between expected and found mass percentages, are given in Table 1. 1H and 13C NMR spectra (in D2O solvent) were consistent with the expected molecular structures but indicated D-H exchange of the hydrophilic OH groups. 2. Surfactant-Water Phase Behavior. Figure 2 shows a schematic phase progression for di-(C5-Glu), and phase diagrams for the di-C6, di-C7, and di-C8 materials (boundaries accurate to within (1%). The nomenclature follows that suggested by Tiddy,22 with L1, H1, V1, and LR used to designate normal micellar and hexagonal phases and viscous “bicontinuous” cubic and lamellar phases, (17) Heenan, R. K.; King, S. M. In Proceedings of the International Seminar on Structural Investigations at Pulsed Neutron Sources, Dubna, Russia, Sept 1992; Publication E3-93-96; JINR: Dubna. (18) Heenan, R. K. FISH Data Analysis Program. Rutherford Appleton Laboratory Report RAL-89-129; Chilton OXON, U.K., 1989. (19) Cabane, B. In Surfactant Solutions: New Methods of Investigation; Zana, R., Ed.; Marcel Dekker: New York. (20) Kotlarchyk, M.; Chen, S.-H.; Huang, J. S.; Kim, M. W. Phys. Rev. A 1984, 29, 2054. (21) Cummins, P. G.; Staples, E.; Penfold, J. J. Phys. Chem. 1990, 94, 3740. (22) Tiddy, G. J. T. Phys. Rep. 1980, 57, 1.

d

Figure 2. Binary water-surfactant phase diagrams for di(Cn-Glu) surfactants. Phase boundaries determined by polarizing microscopy and 2H-NMR are accurate to (1%. Krafft points are indicated by horizontal lines. (a) di-(C5-Glu) (schematic phase progressions); (b) di-(C6-Glu); (c) di-(C7-Glu); (d) di-(C8Glu).

respectively. Some of the biphasic regions were not determined accurately and so are not shown. For both the di-C5 and di-C6 surfactants, no Krafft point Tk was

Properties of New Glucamide Surfactants

detected, at least down to 5 °C, whereas for di-C7 and di-C8 Tk was found to be 34 ( 0.5 °C. The main features of the phase behavior are as follows: (a) di-(C5-Glu). Between 25 and 80 °C, a progression from LR (oily streaks) to V1 (non-birefringent and viscous) to H1 (fans) to L1 (fluid isotropic) is evident on contacting pure surfactant with water. Above 80 °C, the H1 and V1 phases melt, leaving LR and micellar L1. Owing to the relatively high cmc, and the limited availability of this compound, phase sequences were only determined microscopically. (b) di-(C6-Glu). Our initial work on this system referred to a viscous, birefringent phase sandwiched between H1 and L1;3a this has now been identified as a biphasic region. Although it is optically anisotropic, the 2H NMR quadrupolar spectra show a singlet typical of an isotropic phase. Melting of the H1 and V1 phases occurs between 70 and 85 °C. (c) di-(C7-Glu). A two-phase loop occurs at low surfactant concentrations and temperatures (inset), where a lower viscous (surfactant-rich) phase separates from a low-viscosity (water-rich) phase. In this region the tielines are parallel to the concentration axis. This behavior is consistent with an upper critical solution temperature (UCST) or Tc. As Tc is approached, by decreasing temperature, the system becomes turbid. Between 8% and 62%, solutions are once again monophasic and isotropic (by both microscopy and 2H NMR). At 62 ( 2%, a biphasic L1/H1 region forms. Above 70 ( 1%, the hexagonal phase apparently coexists with a lamellar phase LR. It is expected that a narrow V1 region is also present, but this is difficult to detect. (d) di-(C8-Glu). The phase behavior of di-(C8-Glu) is much simpler and the boundaries are less temperaturesensitive than those for the lower homologues. Up to 0.8%, a single isotropic L1 phase is present. Above this concentration, turbidity develops and eventually phase separation occurs. The region marked X has a slightly different optical texture to LR and may be the precursor to a biphasic region. An obvious change in optical texture and viscosity occurs at ∼85%. (e) di-(C9-Glu). This compound is insoluble in water up to 100 °C. However, above 65% an optical texture characteristic of a lamellar phase is observed, and temperature (5-100 °C) does not change the phase boundary. Owing to its insolubility, this material was not studied further. 3. Air-solution Surface Tension. Air-solution surface tensions γ were determined as a function of both surfactant concentration c and temperature (20.0-46.0 °C). Example γ vs c curves, measured at 35 °C, are shown in Figure 3. After purification of di-(C5-Glu) all the surfactants gave clean breaks in γ vs c characteristic of well defined cmc’s; the values are given in Table 2. At 35 °C, log(cmc) displays a linear dependence on the number of carbon atoms in the aliphatic chain, n: log(cmc) ) 2.7 ((0.3) - 0.42 ((0.03)n. This compares favorably with the case for other nonionics, for example CnE8,23 Cn-glucoside,24 and Cn-glucamide,25 where typical values for the intercept are on the order of 2.0-3.0 and gradient -0.50. The changes in γ for pre-cmc concentrations can be described using the Gibbs isotherm, and fits to the data are shown in Figure 3. Table 2 gives estimates for as at (23) Shinoda, K.; Yamanaka, T.; Kinoshita, K. J. Phys. Chem. 1958, 63, 648. (24) Shinoda, K.; Yamaguchi, T.; Hori, R. B. Chem. Soc. Jpn. 1961, 34, 237. (25) Okawauchi, M.; Hagio, M.; Ikawa, Y.; Sugihara, G.; Murata, Y.; Tanaka, M. B. Chem. Soc. Jpn. 1987, 60, 2718.

Langmuir, Vol. 12, No. 11, 1996 2703

Figure 3. Air-water interfacial tension γ vs concentration for di-(Cn-Glu) surfactants at 35 °C: (×) di-(C5-Glu) nonpurified (25 °C) and (9) pure; ([) di-(C6-Glu); (b) di-(C7-Glu); (2) di(C8-Glu). Fitted lines to pre-cmc data are for the Gibbs adsorption isotherm. The parameters derived from this analysis are listed in Table 2. Table 2. Values Obtained from Air-Solution Interfacial Tension Measurements for di-(Cn-Glu) Surfactantsa

di-(C5-Glu) di-(C6-Glu) di-(C7-Glu) di-(C8-Glu)

cmc/mol dm-3

as/Å2

γcmc/mN m-1

∆Gmic/kJ mol-1

(1.05 ( 0.14) × 10-2 (1.35 ( 0.05) × 10-3 (1.60 ( 0.10) × 10-4 (4.2 ( 0.20) × 10-5

81 ( 6 76 ( 2 68 ( 4 72 ( 1

36.8 32.6 28.6 26.2

-21.9 -27.1 -33.3 -36.0

a The mean area per molecule in the pre-cmc monolayer a is s calculated using the Gibbs isotherm. γcmc is the interfacial tension ((1 mN m-1) at the cmc. ∆Gmic is the Gibbs free energy of micellization (0.80 kJ mol-1. Temperature ) 35.0 °C.

the cmc obtained in this way, and the uncertainties are propagated from the least squares algorithm. As found for CiEj’s the cmc values are not very temperature dependent.26,27 The values of as decrease with increasing chain length, and this may be due to increasing chainchain interactions in the monolayer. For single-chain C12E3 at the cmc, neutron reflectometry measurements show as to be 36 Å2 and in good agreement with interfacial tension measurements.28 Therefore the values reported here are reasonable for a dichained nonionic surfactant. The surface tension at the cmc (γcmc) is a measure of the effectiveness of a surfactant. As expected, increasing the hydrocarbon chain decreases γcmc (Table 2). The values are comparable with, for example, those of a single-chain n-C10 glucoside where γcmc ≈ 28 mN m-1.24 The cmc’s of di-(C5-, di-(C6-, and di-(C7-Glu) remain essentially constant over the temperature range studied, whereas for di-(C8-Glu) there is a shallow minimum (Figure 4). A minimum is typical of some ionic surfactants,29 and recent work by Okawauchi et al.30 shows that single-chain glucamides also show this feature. Using the phase-separation model, the cmc can be related to the Gibbs free energy change on micellization of 1 mol of monomer; ∆Gmic ) RT ln(cmc) (with cmc in (26) Van Os, N. M.; Haak, J. R.; Rupert, L. A. M. Physico-Chemical Properties of Selected Anionic, Catanionic and Nonionic Surfactants; Elsevier: New York, 1993. (27) (a) Meguro, K.; Ueno, M.; Esumi, K. In Nonionic Surfactants; Schick, M. J., Ed.; Marcel Dekker: New York, 1987. (b) Meguro, K.; Takasawa, Y.; Kawahashi, N.; Tabata, Y.; Uneo, M. J. Colloid Interface Sci. 1981, 83, 50. (28) Lu, J. R.; Lee, E. M.; Thomas, R. K.; Penfold, J.; Flitsch, S. L. Langmuir 1993, 9, 1352. (29) Schick, M. J. Phys. Chem. 1963, 67, 1796. (30) Okawauchi, M.; Hagio, M.; Ikawa, Y.; Sugihara, G.; Murata, Y.; Tanaka, M. B. Chem. Soc. Jpn. 1987, 60, 2718.

2704

Langmuir, Vol. 12, No. 11, 1996

Eastoe et al.

Figure 4. Variation of cmc with temperature (lines are guides to the eye): ([) di-(C5-Glu); (0) di-(C6-Glu); (2) di-(C7-Glu); (×) di-(C8-Glu).

mole fraction units31 ). Values for ∆Gmic, determined at 35 °C, are given in Table 2. It is instructive to compare ∆Gmic for di-(C5-Glu) with those for other surfactants that have a similar number of hydrophilic oxygens, such as C10E8 and C10E9. The free energy changes are -27.0 and -26.4 kJ mol-1 for C10E8 and C10E9, respectively (albeit at 25 °C). The temperature dependence of the cmc for di-(C8-Glu) results in a minimum ∆Gmic at ∼40 °C. The contribution from each aliphatic carbon ∆G(CH2) may be evaluated (half of the difference between each homologue, since there is a change of two CH2 units per surfactant). From di-(C5-Glu) to di-(C6-Glu) ∆G(CH2) is -2.6 kJ mol-1, whilst from di-(C6-Glu) to di-(C7-Glu) it is -3.1 kJ mol-1. Homologous single-chain surfactants usually give ∆G(CH2) ∼ -3 kJ mol-1.31 An approximate value for the enthalpy of micellization ∆Hmic can be obtained from the temperature dependence of the cmc, since ∆Hmic ) -RT2[δ ln(cmc)/δT]P.32 For di-(C5-, -C6-, and -C7-Glu) ∆Hmic must be essentially negligible. This compares with, say, C10E9, where ∆Hmic is small and approximately 7 kJ mol-1 at 25 °C.26 For di-(C8-Glu) a sign change in the slope [δ ln(cmc)/ δT]P occurs and the ∆Hmic values range from ∼+40 kJ mol -1 at 35 °C to ∼-40 kJ mol -1 at 43 °C. The entropy ∆S mic can also be evaluated: at 35 °C ∆Smic increases with chain length from ∼70 J K-1 mol-1 for di-(C5-Glu) to ∼250 J K-1 mol-1 for di-(C8-Glu). The positive sign is as expected, and the magnitude for di-(C8-Glu) compares favorably with that for C15E8, where ∆Smic ) ∼180 J K-1 mol-1 at 25 °C.26 4. Small-Angle Neutron Scattering. Example SANS data, and model fits, for D2O solutions of di-(C5-Glu) and di-(C8-Glu) at various micellar volume fractions φmic (4 × 10-2 to 6 × 10-4) are shown in Figure 5. The results obtained from the data analysis are given in Table 3. The experiments were performed above the Krafft point and away from any phase boundaries. Within the single-phase region I(Q) essentially scales with volume fraction φmic, and no significant effects of temperature were observed. Similar SANS data and analysis for the di-(C6-Glu) micelles may be found in ref 3. Given the large excess of D2O solvent and the strong H-bonding capability of the hydrophilic ‘sugar’ OH groups, complete D-H exchange in this region was assumed. The scattering length densities of the surfactant Fsurf ()Fp in eq 1) range from 1.81 × 1010 for di-(C7-Glu) to 1.54 × 1010 (31) Clint, J. Surfactant Aggregation; Blackie Academic: Glasgow, 1992. (32) Kreshek, G. C. In Water; Franks, F., Ed.; Plenum Press: New York & London, 1975; Vol. 4.

Figure 5. Example SANS profiles for di-(Cn-Glu)-D2O solutions at various micellar volume fractions φmic (i.e. φsurf - φcmc). Solid lines are the model fits as described in the text. Fitted parameters are given in Table 3. Error bars are shown. (a) di-(C5-Glu) at 25 °C: (+) φmic ) 4.34 × 10-2; (b) 3.30 × 10-2; (O) 2.43 × 10-2; ([)1.37 × 10-2. (b) di-(C8-Glu) at 35 °C: (b) φmic ) 1.2 × 10-3; (O) 6 × 10-4.

cm-2 for di-(C8-Glu) (assuming a bulk density of 1 g cm-3). These values were used to obtain φcalc from the fitted scale factor A. Good quantitative agreement between the fitted model and the observed scattering was obtained, and φcalc agrees with φmic to better than (10% in all cases. Allowing for some D2O penetration of the polyalcohol heads, and given the D-H exchange of the OH groups, one might expect to see a micellar “core” of alkane tails, with a “shell” of head-groups contrasted against D2O. Attempts to use a core/shell model, rather than one for homogeneous particles, did not improve the fitting, and this is probably due to poor contrast in the outer headgroup region. For di-(C5-Glu), the micelles were best modeled as cylinders, with a small S(Q)hs contribution, as described before (fits shown as solid lines in Figure 5). The scattering is consistent with short cylinders of radius R ∼ 10 Å and of axial ratio approximately 2. For di-(C6-Glu), much stronger scattering was evident at low Q.3 In this case power law scattering Q-1, characteristic of a unidimensional structure, was observed. Furthermore the data fit well to the cylinder model. As can be seen in Table 3 the micelle radius increases slightly to R ) 12 Å, whereas the length L increases by about a factor of four to approximately 160 Å. For di-(C7-Glu) at φmic ∼0.01, the UCST is approximately 47 °C (Figure 2c). At 54, 63, and 70 °C, the scattering was essentially identical for all the values of φmic studied. The proximity of a two-phase loop

Properties of New Glucamide Surfactants

Langmuir, Vol. 12, No. 11, 1996 2705

Table 3. Parameters Obtained from Analysis of SANS Data from di-(Cn-Glu) D2O Solutions at Various Micellar Volume Fractions Omica Cylinder Micelles di-(C5-Glu)†

di-(C6-Glu) di-(C7-Glu)*

102φmic

R/Å

L/Å

4.34 3.30 2.43 1.37 5.02 2.27 0.90 5.70 1.10 0.60 0.43

9.2 8.7 8.8 11.7 11.2 12.3 12.0 12.3 12.9 12.7 13.0

43 44 44 33 178 196 144 140 208 197 257

Disk Micelles di-(C8-Glu)

103φmic

t/Å

D/Å

1.20 0.60

19.0 17.6

180 188

a The structural models are described in the text. For cylinder micelles R is the cross-sectional radius ((1 Å) and L the length ((5 Å). For disk micelles t is the thickness ((1 Å) and D the diameter ((5 Å). Temperatures: di-(C5-Glu) and di-(C6-Glu), 25 °C; di(C7-Glu), 70 °C; di-(C8-Glu), 35 °C. Structure factors included for ) (†) di-(C5-Glu) hard-sphere and (*) di-(C7-Glu) attractive Ornstein-Zernike. For all samples the volume fraction φcalc derived from the fitted scale factor agrees to within (10% with the known value φmic.

is expected to introduce an attractive S(Q)oz contribution into the scattering at low Q, and as stated before, this was needed in the fitting. A clear change in the scattering is evident for the di-(C8-Glu) samples, and over an intermediate range of Q, the power law is I(Q) ∝ Q-2, suggesting bilayer or discoid structures. A disk form factor model described the data well, with thickness t approximately 18 Å and diameter D ) 190 Å. The radius R of the cylindrical micelles, and for the di-(C8-Glu) thickness t of the disk micelles, is slightly larger than an all-trans n-Cn chain,31 suggesting either that the chains and heads have more conformational order or that the micellar interface is not sharp as assumed in the analysis. These different micellar structures are to be expected given the changes in mesophase behavior (Figure 2), since cylindrical micelles may form from a hexagonal phase (di-C5, -C6, and C7), and discoidal micelles from an adjacent lamellar phase (di-(C8-Glu)). Conclusions In their phase behavior, temperature dependence of the cmc’s and micellar shapes, these di-(Cn-Glu) amphiphiles behave slightly differently from comparable CiEj nonionics. The more soluble di-C5 and di-C6 materials display a classic progression LR f V1 f H1 f L1. However, for di-(C7-Glu), an UCST is observed, and we believe that

this is the first observation of this transition for a nonionic surfactant. Such behavior is known for zwitterionics such as 3-(decyldimethylammonio)propyl sulphate.33 With di(C8-Glu), only a narrow micellar phase region is observed, whilst di-(C9-Glu) is essentially insoluble. The phase behavior of CiEj nonionics is well-documented, and the existence of a characteristic LCST cloud point (or Tc) has been understood in terms of the dehydration of ethylene oxide residues. The higher solubility of saccharide surfactants is expected to increase Tc, so that for most practical purposes this phase separation is not observed. However, the UCST cannot be rationalized in this way; rather, the transition is driven by attractive micelle-micelle interactions that may be overcome by temperature. In essence the large biphasic L1/LR region for di-(C8-Glu) may be an extension of this critical transition, whereby the two-phase loop has now become significantly enlarged. For the dilute phases there is apparently little effect of temperature on the interfacial adsorption, cmc’s, or micelle structures, and this may be due to stronger H-bonding with water than for the CiEj’s. However, the chain length dependence of the cmc’s and thermodynamic parameters relating to micellization compare well with those found for CiEj’s. The SANS measurements suggest that anisotropic micelles are present for low concentrations. The surfactant packing parameter concept is often used to understand shape transitions in micelles: P ) v/al with v the hydrophobic tail volume, l the length, and a the area per polar head. The changes sphere f cylinder f disk occur at a critical values of P ) 1/3 and 1/2. Although we were not able to obtain an accurate value for a in micelles from the SANS (owing to a limited range at high Q), it may be assumed that a is approximately constant for these di-(Cn-Glu) surfactants. Since v and l both have a near-linear dependence on the alkyl carbon number n, the packing parameter theory does not predict a shape change per se as n increases. Therefore the observed transition short rods di-C5 f long cylinders di-C6 and di-C7 f ‘disks’ di-(C8-Glu) cannot be explained in this way, unless a decreases with n. Although a small decrease in the values of as at the air-solution interface has been observed (Table 2), conformational changes of the glucamide head-group may also play a role in the micelles. Acknowledgment. P.R. thanks Kodak European Research and the University of Bristol for provision of a jointly funded Ph.D. studentship. A Royal Society grant was awarded for microscopy equipment. We extend our thanks to the staff of the analytical and NMR services in the School of Chemistry. The EPSRC Neutron Beam Committee is thanked for an allocation of beam time and financial assistance with chemicals and travel. LA950827K (33) Nilsson, P.-G.; Lindman, B.; Laughlin, R. G. J. Phys. Chem. 1984, 88, 6357.