Bilayer structure of ester-amide-type cationic surfactants in a dilute aqueous solution

Colloids and Surfaces A: Physicochem. Eng. Aspects 441 (2014) 140–148

Contents lists available at ScienceDirect

Colloids and Surfaces A: Physicochemical and Engineering Aspects journal homepage: www.elsevier.com/locate/colsurfa

Bilayer structure of ester-amide-type cationic surfactants in a dilute aqueous solution Youhei Kawabata a,∗ , Kento Hayashi a , Tomoki Kanao b , Akira Ishikawa b,∗∗ a b

Department of Chemistry, Tokyo Metropolitan University, Hachioji, Tokyo 192-0397, Japan Kao Corporation, 1334 Minato, Wakayama, Wakayama Prefecture 640-8580, Japan

h i g h l i g h t s

g r a p h i c a l

a b s t r a c t

• Lamellar structures of ester amide hydrochloride salt solutions are investigated. • Lamellar repeat distance is swollen up to d ∼ 28 nm by the dilution. • Transit ion to unilamellar vesicles is induced below the Krafft temperature.

a r t i c l e

i n f o

Article history: Received 10 June 2013 Received in revised form 26 August 2013 Accepted 26 August 2013 Available online 12 September 2013 Keywords: Cationic surfactant Lamellar structure Unilamellar vesicle

a b s t r a c t The ester-amide-type surfactant (2-[N-[3-alkanoyl(C16-18)aminopropyl]-N-methylamino]ethyl alkano(C16-18)ate, EA) is a dichain cationic amphiphile that has been used as the main base of a fabric softener. We observed the structure of a diluted aqueous solution of EA below the melting temperature (Krafft temperature) using a polarised or fluorescence microscope, small/wide-angle X-ray scattering (SAXS/WAXS), dynamic light scattering (DLS) and cryogenic-temperature transmission electron microscopy (cryo-TEM). From the optical microscope observations, we found that the lamellar domains disappear below the EA concentration (EA ) of ≈3 wt%. Below EA ∼ 1 wt%, in the cryo-TEM images, we found round dispersions with 35 nm diameters. For the lamellar structure, it was found from the SAXS profiles that the repeat distance of the lamellar gel phase increased up to 28 nm with EA decreasing to 5 wt%. Furthermore, from the SAXS and DLS analysis, it was also found that unilamellar vesicles, whose diameter is 30–46 nm, are formed below EA ≈ 1 wt%. All of these experimental results point to the conclusion that the transformations from planar lamellar structures to unilamellar vesicles are induced by dilution of the EA aqueous solution. © 2013 Elsevier B.V. All rights reserved.

1. Introduction Surfactants are widely used in various fields, such as medical supplies and the industrial materials. While typical examples of surfactant application include washing and emulsification, they are

∗ Corresponding author. Tel.: +81 42 677 1111; fax: +81 42 677 2525. ∗∗ Corresponding author. E-mail addresses: [email protected], [email protected] (Y. Kawabata), [email protected] (A. Ishikawa). 0927-7757/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.colsurfa.2013.08.052

also used as protective materials on the surfaces of solids, lubricants, etc. Dichain cationic surfactants are often used as the main base of fabric softeners because they adsorb onto textiles due to their negative-charged surfaces in water [1]. The adsorbed surfactants are formed in layers on the surface of the textiles; these layers also protect the textiles. Surfactant concentration in the fabric softener is about ∼20 wt% depending on the product. The self-assemble structure in the fabric softener is a lamellar gel phase (Lˇ ), and multilamellar vesicles can sometimes form [2,3]. There are some reports where the solution structures composed of dichain cationic surfactants have been investigated [2–11].

Y. Kawabata et al. / Colloids and Surfaces A: Physicochem. Eng. Aspects 441 (2014) 140–148

141

Fig. 1. (a) Structural formula of ester-amide-type cationic surfactant (EA) where R is alkyl (C16-18). (b) Polarised microscope image of EA solution whose EA concentration EA is 17 wt%. (c)–(f) The polarised microscope images of the dilute EA solutions at each EA concentration. Birefringence due to the planar lamellar structures decreases with the EA concentration decreasing. (Scale bar is 100 ␮m.)

Zemb et al. [4] and Dubois and Zemb [5] have reported the phase behaviour of DDAB (didodececyldimethylammonium bromide) or DHDA (dihexadecyldimethylammonium acetate) aqueous solutions over a wide concentration range. They have showed that the lamellar phase is highly swollen up to ∼1.4 wt% and that a critical behaviour near the lamellar-lamellar phase separation region exists. Caboi and Monduzzi [6] have also reported this phase separation of the DDAB system. They found that the two coexisting lamellar phases correspond to vesicles and planar lamellae in the lower and higher concentrations, respectively, and that the vesicles in the lower concentration transformed into the planar lamellae with increasing temperature at a fixed concentration. Haas et al. [7] have investigated the phase structures of dichain cationic surfactants with symmetric or asymmetric alkyl chains and showed that planar bilayers are transformed into multilamellar vesicles by decreasing the surfactant concentration. Tucker et al. [8–11] reported the phase behaviour of DHDAB (dihexadecyldimethylammonium bromide)/water and DHDAB/polyoxyethylene type surfactant (Ci Ej )/water systems. They performed a curve analysis of the small-angle neutron scattering profiles and explained the phase behaviour by using the obtained structural parameters that characterise lamellar or vesicular phases.

On the other hand, the ester-amide-type surfactant 2propyl]-N-methylamino]ethyl [N-[3-alkanoyl(C16-18)amino alkano(C16-18)ate, hydrochloride (EA) is also a dichain cationic amphiphile and has been used as the main base for fabric softeners (Fig. 1(a)). An aqueous solution of EA is believed to have high product stability because of its characteristic feature of planar lamellar structures that form at higher concentrations by using polyoxyethylene-type nonionic surfactant as the co-surfactant. Fig. 1(b) shows a polarised microscope image of the EA aqueous solution. It is clear that any Maltese crosses, which show the existence of vesicles, cannot be found. However, until now, structures of EA solutions at dilute concentrations have not been clarified. These are important for understanding the function of EA as a fabric softener in a laundry setting. In this research, we observed the structure of an EA dilute solution by using polarised or fluorescence microscopy, small/wide-angle X-ray scattering (SAXS/WAXS), dynamic light scattering (DLS) and cryogenic-temperature transmission electron microscopy (cryo-TEM). We found that upon decreasing the EA concentration, the repeat distance of the lamellar gel phase increased up to 28 nm. Below the EA concentration of 3 wt%, the planar

142

Y. Kawabata et al. / Colloids and Surfaces A: Physicochem. Eng. Aspects 441 (2014) 140–148

lamellar structures were transformed into unilamellar vesicles of about 30–46 nm in diameter. 2. Experimental 2.1. Sample preparation We used an ester-amide hydrochloride salt (Kao Corp.) as the main surfactant without any purification. Polyoxyethylene(21) lauryl ether (C12 E21 : C12 H25 (OC2 H4 )n OH, n = 21 on average, Kao Corp.)

was used as the co-surfactant. EA was mixed with nonionic surfactant aqueous solution at 60 ◦ C (above the melting temperature of EA, 45 ◦ C). After neutralisation by hydrochloric acid, the EA solution was cooled by iced water stirring up the solution. The concentration of EA, defined as the weight percent of EA (EA ) in the solution, was prepared to be 17 wt%. The nonionic surfactant concentration nonionic was 2.5 wt%. The prepared EA solution was diluted by mixing with ion-exchanged water so that the EA concentration EA ranged from 0.003 to 15 wt% with nonionic /EA fixed at 0.147. For the fluorescence microscope observations, we used the lipophilic fluorescence tracer DPH (1,6 diphenyl-1,3,5-hexatriene, Invitrogen Inc., ex = 360 nm, em = 428 nm), which was mixed with the EA solution by the ratio of 0.1 wt% to EA. 2.2. Experimental method For polarised and fluorescence microscopy observations, we used a BX-51 microscope with a Hg lamp (Olympus Inc.). A vitro tube (VitroCom Inc.) was used as a sample cell, whose inner size was 2 mm (W) × 0.2 mm (H). The optical microscope images were taken using a digital video camera, HDR-XR500V (SONY Inc.). SAXS/WAXS experiments were performed using the synchrotron radiation SAXS spectrometer installed in BL6A at the Photon Factory (PF) of the High Energy Accelerator Research Organization (KEK) and BL45XU at SPring-8. The scattering X-rays for the SAXS and WAXS were simultaneously detected with a PILATUS (300 K at BL6A and BL45XU, Dectris Inc.) and flat panel (C9730DK10, Hamamatsu Photonics Inc.), respectively. The X-ray wavelength is 0.09 nm at BL45XU and 0.15 nm at BL6A, respectively. The scattering vector q (= 4 sin /, where 2 is the scattering angle and  is the X-ray wavelength) was 0.3–2.0 nm−1 for the SAXS and 10–40 nm−1 for the WAXS. The sample cell was made of copper with Kapton windows. The thickness of the sample was 1.0 mm. The DLS experiments were performed with a mono modefibre compact goniometer system (ALV-5000) using a diode-laser pumped Nd-YAG laser (wavelength in vacuum was 532 nm). The scattering angle was set at 40–120◦ . The cylindrical cells made of quartz were used with the samples and were put into the cell housing. All the scattering or microscope experiments were conducted at room temperature (20–25 ◦ C). The cryo-TEM measurements were made using a Hitachi H-7650 at an accelerating voltage of 120 kV under a low electron dose. As described elsewhere [12], thin vitrified films of the suspensions were prepared for cryo-TEM using a Leica EM CPC fast freezing device. A drop of an EA suspension was deposited on a carbon film, and the excess liquid was blotted with filter paper. Before evaporation of the remaining fluid, the grids were quench-frozen in liquefied ethane cooled to −171 ◦ C with liquid nitrogen. The specimens were then transferred into a cryo-holder kept at −180 ◦ C with liquid nitrogen. 3. Results and discussion 3.1. Microscope measurement

Fig. 2. (a)–(h) Fluorescence microscope images of the EA solutions at each EA concentration. (Scale bar is 100 ␮m.) The fluorescence of each image is due to the lamellar domain involved lipophilic tracer DPH. At EA = 3 wt%, fluorescence can be scarcely found, as indicated by the arrows in (g).

3.1.1. Polarised and fluorescence microscope Fig. 1 shows the polarised microscope images at each EA concentration EA . With EA decreasing, birefringence due to the planar lamellar structure is clearly observed from EA = 7–15 wt%. At 5 wt%, birefringence was not clearly observed and the existence of lamellar domains could not be confirmed. Therefore, to confirm the existence of lamellar domains, we observed the solution mixed with the lipophilic fluorescence tracer DPH, which can enhance the contrast between the lamellar structures and the solvent. Fig. 2 represents the fluorescence microscope images at each EA . The

Y. Kawabata et al. / Colloids and Surfaces A: Physicochem. Eng. Aspects 441 (2014) 140–148

143

Fig. 3. Cryo-TEM images (a) and (b) were taken at EA = 0.05 and 0.006 wt%. (Scale bar is 200 nm.) Round dispersions can be found, as indicated by the arrows.

fluorescence of each image is due to the lamellar domain involved DPH. As EA decreases, the number of lamellar domains and their size decrease. At EA = 5 wt%, fluorescence remains to some extent while the birefringence could not be observed. At 3 wt%, fluorescence can be barely observed, as indicated by the arrows in Fig. 2(g). All these results indicate that large planar lamellar domains disappear at EA at least less than 3 wt%.

analysis using Image J [13]. The average size Dave and the standard deviation ıD in diameter are 34.6 ± 6.02 nm at EA = 0.05 wt% and 34.5 ± 5.35 nm at EA = 0.006 wt%, respectively. Considering that the unit structures of the molecular assemblies are bilayers below the Krafft temperature, the particles in the TEM images should be vesicles or bicelles.

3.2. X-ray diffraction 3.1.2. Cryo-TEM Fig. 3 shows the cryo-TEM images at EA = 0.05 and 0.006 wt% and spherical or disc like dispersions can be observed as indicated by the arrows. The particle size distributions were obtained from an

Fig. 4(a) shows the SAXS profiles observed at each EA . The first-order Bragg peak of the SAXS profile at EA = 17 wt%, in a stock solution, is at around 0.4 nm−1 , and the Bragg peaks

6 5

Fitting result

4

EA17% EA12% EA8% EA5% EA2%

3

2

8 7 6

×10 2

5

Intensity / a.u.

Intensity / a.u.

3

10

4 3

×4

2

×2.5

10

2

10

×0.8

8 7 6

φEA=5% 2 9 8

φEA=4%

7

5

6

4

0.2

0.4

0.6

0.8

1.0

1.2 -1

q / nm

1.4

1.6

1.8

0.2

0.4

0.6

0.8

1.0

1.2 -1

1.4

1.6

1.8

q / nm

Fig. 4. (a) The obtained SAXS profiles at each EA concentration. The first-order Bragg peak of the SAXS profile at EA = 17 wt%, in a stock solution, is at around 0.4 nm−1 , and the Bragg peaks appear up to fourth-order. The blue curves are the fitted results using the scattering model proposed by Nallet et al. [14]. With the EA concentration EA decreasing, the Bragg peaks shift to lower q and the intensities of the first-order peaks become weaker. (b) The SAXS profiles at EA = 4, 5 wt%, where the Bragg peaks disappear, are magnified. The arrows in the SAXS profile of EA = 5 wt% are the second- and third-order peaks. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of the article.)

144

Y. Kawabata et al. / Colloids and Surfaces A: Physicochem. Eng. Aspects 441 (2014) 140–148

30

12

(a)

(b)

10

26

Bilayer numbers

Repeat distance / nm

28

24 22 20

8 6 4

18 2 4

5

6

7

8

9

6

10

8

10

12

14

16

φEA / wt%

φEA / wt%

Fig. 5. (a) Lamellar repeat distance as a function of the EA concentration on a double-logarithmic scale. The line corresponds to d ∝ −0.39 . (b) The number of bilayers in a lamellar domain obtained from the fitting to the SAXS profiles using the scattering model proposed by Nallet et al. [14].

appear up to fourth-order. This indicates that there is a long-range order between the lamellar structures. With the EA concentration decreasing, the Bragg peaks shift to the lower q and the intensities of the Bragg peaks become weaker. In Fig. 4(b), the SAXS profiles below EA = 4, 5 wt% are magnified and it is clear that the Bragg peaks disappear at around 4 wt%. The broad peak at 4 wt% may be due to the form factor of the bilayers and unilamellar structures should form. This fact is consistent with the results obtained by the fluorescent and polarised microscopy observations that the little lamellar domains can be seen below EA = 5 wt%. To analyse these SAXS profiles, we used the scattering model proposed by Nallet et al. [14]. The X-ray scattering intensity from a lamellar structure is given by [14,15],

I(q) =

2 PM (q)S(q) d q2

(1)

where d is the lamellar repeat distance and S(q) is the static structure factor between bilayers,

I/I0

1.0

φEA=17 wt% φEA=12 wt% φEA=9 wt% φEA=7 wt% φEA=5 wt% φEA=3 wt% φEA=1 wt%

 × exp

˛(i) =

1−

i N

 cos

i=1

qdi 1 + 2q2 d2 ˛(i)

2q2 d2 ˛(i) + q2 d2 i2 − 2(1 + 2q2 d2 ˛(i))



1 1/2 (1 + 2q2 d2 ˛(i))

,

 {ln(i) + }. 22

(2)

(3)

and where N is the number of bilayers in a domain and q is the width of the resolution function, which is estimated to be 0.007 nm−1 [16].  is Euler’s constant, q0 is the position of the firstorder Bragg peak. The form factor of bilayers is given as follows [14,15,17], PM (q) =

4 2 [( h − w ) sin (q(ıc + ıh )) + ( c − h ) sin (qıc )] . q2

(4)

Here, w , h and c are the scattering length densities of water, hydrophilic and hydrophobic parts of surfactants, respectively. ıc the hydrophobic tail length, and ıh the hydrophilic segment length. By using the molar volume of EA, 654.4 cm3 mol−1 obtained from

(a)

(b)

Intensity / a.u.

1.2

N−1  

S(q) = 1 + 2

0.8 0.6

10

4 9 8 7 6 5

12

13

14

15 -1

q / nm

16

17 18

15

20

25

30

35

-1

q / nm

Fig. 6. (a) WAXS profiles normalised by the intensities of the Bragg peaks at q ≈ 15 nm−1 , which corresponds to the hexagonal order of surfactant molecules in plane. (b) The obtained WAXS profile at EA = 0.006 wt%.

Y. Kawabata et al. / Colloids and Surfaces A: Physicochem. Eng. Aspects 441 (2014) 140–148

145

φEA=1% (a) φEA=0.1% φEA=0.05% φEA=0.01% φEA=0.003% fitting results (vesicle) fitting results (bicelle)

×100 5

10

×30

4

Intensity / a.u.

10

×5

(b)

45 40 35 30 0.25

Polydispersity

×20

Diameter / nm

50

3

(c)

0.20 0.15 0.10 0.05

10

(d)

ρ / nm

-2

0.4

2

10

0.3 0.2 0.1 0.0

6 7 8 9

0.1

2

3

4

4 6

5 6 7 8 9

-1

1

q / nm

2

0.01

4 6

2

0.1

4 6

1

φEA / wt%

Fig. 7. (a) SAXS profiles with the fitting results obtained using the vesicle model. At EA = 0.05 wt%, the fitting curve calculated by the bicelle model is also represented to compare those fitting results. (b)–(d) are the EA concentration dependence of the fitting parameters obtained from the analysis using the vesicle model. (b) The diameter of vesicles, (c) the polydispersity of the vesicle size and (d) the charge density on the vesicle surface.

the specific gravity measurement, h and c were estimated at 17.4 × 1010 and 7.92 × 1010 cm−2 , respectively. The w was calculated by reference to the literature [15]. The solid curves in Fig. 4 are the fitted results with Eqs. (1)–(4), which are explained well by the scattering function. The fitting parameters for the bilayer structures, such as ıc and ıh , could not be obtained with sufficient accuracy at EA = 5–17 wt%. Therefore, the ıc and ıh were estimated from the fitting to the SAXS profile at EA = 2 wt% where the Bragg peak does not exist by assuming S(q) to be unity. The obtained ıc and ıh are 2.15 ± 0.01 nm and 0.46 ± 0.01 nm, which seem to be reasonable values that can be calculated with the extended length ıc = 2.30 ± 0.13 nm (C16-18) [18]. Below 1 wt%, we could not analyse the SAXS profiles with the above equation because the broad peaks between q = 0.4–1.6 nm−1 become very small. Fig. 5 shows the EA concentration dependencies of d and N, which were obtained from the analysis. The initial repeat distance at EA = 17 wt% is about 16 nm. According to the lamellar swelling limit, expressed dlim = ı/˚ (ı = bilayer thickness, ˚ = volume fraction of surfactant), dlim could be estimated to be 33 nm at EA = 17 wt% assuming one phase of lamellae with ı = 5.22 nm. Therefore, the lamellar phase of this system coexists with excess water. Generally, lamellar repeat distance of a surfactant/water binary system that coexists with an excess water is constant

independent of concentration. In the present case, the repeat distance increases exponentially up to ∼28 nm (d ∝ −0.39 ) and N decreases to 1 with the EA concentration decreasing, although the system has excess water. This is because the bilayer comprises two components of surfactants, EA and C12 E21 , and the repeat distance might change due to the surfactant composition in the bilayers. This means that the lamellar structure is swollen due to repulsive forces, such as Helfrich repulsion (steric repulsive force due to membrane undulation [19]) or electrostatic forces with the bilayer composition changing. Thus, the charge density of the surface or the flexibility of bilayers should be changed by dilution, so that this lamellar phase swells. To confirm this, we checked the WAXS profiles at each EA . Fig. 6 shows the WAXS profiles normalised by the intensities of the Bragg peaks at q ≈ 15 nm−1 . These peaks correspond to the hexagonal order of surfactant molecules in plane and indicate that the bilayers form in Lˇ phase (gel phase) [20]. It is clear that the peak width increases with EA decreasing. Nagai et al. have reported that bilayers below the Krafft temperature in a nonionic surfactant system become flexible and vesicles form spontaneously due to the bulky hydrophilic part of the nonionic surfactant and that the Bragg peak becomes broad in such a situation [21]. The same holds true for this EA system, and then packing of the surfactant molecules in plane becomes loose and the flexibility of the bilayers increases

Y. Kawabata et al. / Colloids and Surfaces A: Physicochem. Eng. Aspects 441 (2014) 140–148

as EA decreases. Therefore, the suppression of the peak intensities in the SAXS profiles and the lamellar swelling may be due to the enhancement of the fluctuation of the bilayers. Fig. 6(b) shows the WAXS profile obtained at EA = 0.006 wt%. As indicated by the arrows in Fig. 6(b), the Bragg peak corresponding to the hexagonal packing of surfactants in plane appears although the peak intensity is very weak. At 0.003 wt%, we cannot find the Bragg peak but can conclude that the bilayer exists at more than 0.006 wt%. Therefore, the bilayer formations at lower concentrations may be vesicles or bicelles, when taking the SAXS/WAXS results into account.

2.0 1.8

g2(t)

146

1.6 1.4 1.2 1.0 2.0 1.8

To determine whether vesicles or bicelles are formed at the dilute regime, we analysed the SAXS profiles by using the form factor of vesicles and bicelles. The X-ray scattering intensity is given by

1.6

 ×

1−

(1 + 4q2 /s2 )

−(z+1)/2 2 s

cos[(3 + z) arctan(2q/s)] + s2

4q2

2.0 1.8

φEA=1 wt%

1.6 1.4 1.2



1.0 .

(6)

10

Here, z is the Schulz distribution width, s = (z + 1)/R0 , R0 is the average size of the vesicles. For the bicelle form factor, we used the function of core shell discoidal model [22]. We used S(q), which describes inter-particle correlations for a repulsive screened Coulomb interaction potential [23]. Fig. 7(a) shows the SAXS profiles obtained at the dilute regime. Curves are the fitted results of Eqs. (5) and (6). For a comparison between the analysis of the vesicle and bicelle functions, we analysed only the profile at EA = 0.05 wt% using both form factors. The dotted line is the fitted results of bicelle function with the radius obtained from the cryo-TEM image analysis. It is clear that the SAXS profiles can be explained by the vesicle form factor, while the fitted results using the bicelle function deviate from the experimental profiles. Furthermore, Fig. 7(a) shows the concentration dependence on the vesicle diameter obtained from the above analysis. The absolute values of diameter, which are about 30–46 nm with a polydispersity of p = 1/(z + 1) = 0.1 ∼ 0.2 (Fig. 7(b)), are in good agreement with those estimated from the cryo-TEM images. Fig. 7(c) indicates the concentration dependence of the charge density obtained from the fitting parameter of S(q). This increases with decreasing EA concentration, which indicates that the composition of nonionic surfactant and the EA in the bilayers might vary with dilution. This result is consistent with the fact that the WAXS peak width increases with EA decreasing as shown in Fig. 6. Thus, all these SAXS and WAXS results clearly indicate that the nonionic surfactant concentration in bilayers decreases with decreasing EA . 3.4. DLS measurement Fig. 8 shows the typical autocorrelation functions g2 (t) at EA = 0.006, 0.1, 1 wt%. g2 (t) is related to the normalised field correlation function g1 (t) as g2 (t) = 1 + ˇ|g1 (t)|2 .

1.4 1.0

g2(t)

82 (z + 2)(z + 3) PM (q) s2 q2

φEA=0.1 wt%

1.2

(5)

where S(q) is the static structure factor and P(q) is the form factor of the vesicle or bicelle. The vesicle form factor proposed by Pencer et al. has been used [17]. This form factor can be obtained by integrating the monodisperse form factor of a vesicle with the form factor of a bilayer, over the Schulz distribution,

P(q) =

g2(t)

3.3. Analysis of SAXS profiles at dilute regime

I(q) ∝ S(q)P(q)

φEA=0.006 wt%

(7)

-3

10

-1

10

t / ms

1

10

3

Fig. 8. Typical examples of the time correlation function g2 (t) obtained from the DLS experiments. All functions were obtained at a scattering angle of 90◦ , and at EA concentrations EA = 1, 0.006, 0.003 wt%. The lines are fitted results based on Eq. (8).

Here, ˇ is the coherence factor, whose theoretical maximum is unity. In the case of an ideal solution of monodisperse particles, the correlation function g1 (t) is given by a single exponential function as g1 (t) = exp(−D0 q2 t).

(8)

The curves in Fig. 8 are the fitted results with Eqs. (7) and (8). The observed data can be explained by the single exponential correlation function. In practice, because the particles in the solution should be polydispersed, we estimated the relaxation rate by using the methods of cumulants [24,25]. Fig. 9 represents the relaxation rate K1 as a function of q2 . The lines are the fitted results according to the relationship of K1 = Dz q2 , where

Ni m2 Di Dz = i Ni mi

(9)

Here Dz is the z-averaged diffusion coefficient. Ni and mi are the number and the molecular weight of the particles with the diffusion coefficient Di , respectively. The second cumulant of K2 is given by K2 = [D2 z − D2z ]q4 . From the Einstein-Stokes relationship and the slope of the fitted lines in Fig. 9, we calculated the z-averaged hydrodynamic radius Rh z , which was about 30 nm independent of EA . We can estimate the number-averaged size if the particles are vesicle. Assuming that the form factor of a vesicle P(R) = 1 and higher

Y. Kawabata et al. / Colloids and Surfaces A: Physicochem. Eng. Aspects 441 (2014) 140–148

-1

6 5

K1 / ms

the diameters, which were estimated by the SAXS and DLS, assuming the particles are vesicles, are summarised. It is obvious that the absolute values of the diameter and their concentration dependence, which the particle size increases with EA decreasing, are in good agreement with those obtained from the SAXS observation. All this evidence leads us to conclude that unilamellar vesicles are formed by the dilution of the EA solution.

φEA =1 wt% φEA =0.1 wt% φEA =0.006 wt%

7

4 4. Conclusion

3 2 1 0 0

1

2

3

4

5

6

2

q / nm-2

-4

7x10

Fig. 9. Relaxation rate K1 obtained by using the methods of cumulants as a function of q2 . The lines are fitted results based on the relation, K1 = Dz q2 .

55

2N / nm

50 45 40 35 30 25 20 4 6 8

0.01

2

4 6 8

0.1

φEA / wt%

2

4 6 8

1

Fig. 10. EA concentration dependence of number-averaged diameter obtained using the Einstein–Stokes relationship and Eq. (10). Error bars correspond to the polydispersity.

moments of the size distribution were negligible, Selser and Yeh presented the following the relationship [26–28], RN

147

A Dz (1 + 3ız )

(10)

ıN ız where RN and ıN are the number-averaged radius and size dispersion, respectively. A = kB T/6, ız is the relative dispersion of Dz . Fig. 10 shows the concentration dependence of the numberaveraged diameter obtained by the above calculation. In Table 1, Table 1 Diameters estimated by SAXS and DLS, assuming the particles as vesicles, at the typical EA . R ± ıR nm

1 wt%

0.05 wt%

0.003 wt%

SAXS DLS

30.7 ± 2.23 29.7 ± 6.59

38.7 ± 4.39 44.1 ± 4.48

46.2 ± 9.45 46.7 ± 6.61

We observed the structure of the dilute aqueous solution of an ester-amide-type cationic surfactant (2-[N-[3-alkanoyl(C1618)aminopropyl]-N-methylamino]ethyl alkano(C16-18)ate, EA) below its melting temperature using polarised or fluorescence microscopy, small/wide-angle X-ray scattering (SAXS/WAXS), dynamic light scattering (DLS) and cryogenic-temperature transmission electron microscopy (cryo-TEM). From the microscopic observations, we found that the lamellar domains disappear below the EA concentration of EA ∼ 3 wt%. For the lamellar structure with EA concentration decreasing, Xray scattering profiles indicate that the lamellar repeat distance increases up to 28 nm and the number of bilayers N decreases to unity below the EA concentration of 4 wt%. In the diluted solution below EA ∼ 1 wt%, dispersions whose hydrodynamic diameters are about 30–46 nm could be observed from the DLS results and the round dispersions, whose average diameters are 35 nm, were found in the cryo-TEM observation. We analysed the SAXS profiles in the dilute regime using the unilamellar vesicle scattering function with the inter-particle structure factor and found that the absolute values of diameter, 30 ∼ 46 nm, are consistent with those obtained from the DLS measurements and the cryo-TEM images. We conclude from all these results that the transformations from the planar lamellar structure to unilamellar vesicles are in fact induced by the dilution of the EA solution. There are not many reports regarding the transformation of planar lamellae into unilamellar vesicles with high dilution and swelling of the lamellar repeat distance [29], multilamellar vesicles induced by dilution have been reported in the other dichain cationic surfactant systems [4,5,30,8–11]. To clarify mechanisms of the structural transition, it is important to verify the transition kinetics experimentally and theoretically [31]. Furthermore, from the WAXS and SAXS analysis, it is evident that the bilayer composition might vary and the nonionic surfactant concentration in the bilayers decreases with decreasing EA . This variance in composition with respect to concentration would play an important role in the lamellar swelling and the formation of unilamellar vesicles. These results and considerations may contribute to the development of EA solutions as functional fabric softeners. Acknowledgements The authors thank Dr. K. Torigoe of Tokyo University of Science for his support during the cryo-TEM observations, and Prof. T. Kato of Tokyo Metropolitan University for his fruitful discussions. Our SAXS experiments at KEK and SPring-8 were performed under approval of the Photon Factory and SPring-8 Advisory Committees (Proposal Nos. 2012G164 and 2011B1052), respectively. References [1] R. Puchta, Cationic surfactants in laundry detergents and laundry after treatment aids, J. Am. Chem. Soc. 61 (1984) 367–376. [2] M.L. Lynch, T. Kodger, M.R. Weaver, Anticipating colloidal instabilities in cationic vesicle dispersions by measuring collective motions with dynamic light scattering, J. Colloid Interface Sci. 296 (2006) 599–607. [3] G.C. Shearman, S. Ugazio, L. Soubiran, J. Hubbard, O. Ces, J.M. Seddon, The lyotropic phase behaviour of ester quaternary surfactants, J. Colloid Interface Sci. 331 (2009) 463–469.

148

Y. Kawabata et al. / Colloids and Surfaces A: Physicochem. Eng. Aspects 441 (2014) 140–148

[4] T. Zemb, D. Gazeau, M. Dubois, T. Gulik-Krzywicki, Critical behaviour of lyotropic liquid crystals, Europhys. Lett. 21 (1993) 759–766. [5] M. Dubois, T. Zemb, Phase behavior and scattering of double-chain surfactants in diluted aqueous solutions, Langmuir 7 (1991) 1352–1360. [6] F. Caboi, M. Monduzzi, Didodecyldimethylammonium bromide vesicles and lamellar liquid crystals. A multinuclear NMR and optical microscopy study, Langmuir 12 (1996) 3458–3556. [7] S. Haas, H. Hoffmann, C. Thunig, E. Hoinkis, Phase and aggregation behaviour of double-chain cationic surfactants from the class of n-alkyl-n-alkyl -n,ndimethylammonium bromide surfactants, Colloid Polym. Sci. 277 (1999) 856–867. [8] J. Penfold, E. Staples, I. Tucker, R.K. Thomas, Surface and solution behavior of the mixed dialkyl chain cationic and nonionic surfactants, Langmuir 20 (2004) 1269–1283. [9] I. Tucker, J. Penfold, R.K. Thomas, I. Grillo, J. Barker, D.F.R. Mildner, Selfassembly in mixed dialkyl chain cationic-nonionic surfactant mixtures: Dihexadecyldimethyl ammonium bromide-monododecyl hexaethylene glycol (monododecyl dodecaethylene glycol) mixtures, Langmuir 24 (2008) 7674–7687. [10] I. Tucker, J. Penfold, R.K. Thomas, I. Grillo, J. Barker, D.F.R. Mildner, The surface and solution properties of di-hexadecyldi-methylammonium bromide, Langmuir 24 (2008) 6509–6520. [11] I. Tucker, J. Penfold, R.K. Thomas, I. Grillo, D.F.R. Mildner, J. Barker, Self-assembly in complex mixed surfactant solutions: the impact of dodecyl triethylene glycol on dihexadecyl dimethyl ammonium bromide, Langmuir 24 (2008) 10089–10098. [12] M. Abe, T. Koike, H. Nishiyama, S.C. Sharma, K. Tsubone, K. Tsuchiya, K. Sakai, H. Sakai, Y.A. Shchipunov, J. Schmidt, Y. Talmon, Polymerized assemblies of cationic Gemini surfactants in aqueous solution, J. Colloid Interface Sci. 330 (2009) 250–253. [13] M.D. Abramoff, P.J. Magelhaes, S.J. Ram, Image processing with ImageJ, Biophoton. Int. 11 (2004) 36–42. [14] F. Nallet, R. Laversanne, D. Roux, Modelling X-ray or neutron scattering spectra of lyotropic lamellar phases: interplay between form and structure factors, J. Phys. II 3 (1993) 487–502. [15] K. Minewaki, T. Kato, H. Yoshida, M. Imai, K. Ito, Small-angle X-ray scattering from the lamellar phase formed in a nonionic surfactant (C16E7)/water system. Analysis of peak position and line shape, Langmuir 17 (2001) 1864.

[16] C. Glinka, J.M. Rowe, J.G. LaRock, The small-angle neutron scattering spectrometer at the national bureau of standards, J. Appl. Crystallogr. 19 (1986) 427–439. [17] J. Pencer, S. Kruger, C.P. Adams, J. Katsaras, Method of separated form factors for polydisperse vesicles, J. Appl. Crystallogr. 39 (2006) 293–303. [18] C. Tanford, The Hydrophobic Effect, Wiley, New York, 1980. [19] W. Helfrich, Elastic properties of lipid bilayers theory and possible experiments, Naturforsch 28c (1973) 693–703. [20] C.E. Fairhurst, S. Fuller, J. Gray, M.C. Holmes, G.J. Tiddy, Lyotropic surfactant liquid crystals, Wiley-VCH, Weinheim, 1998. [21] Y. Nagai, Y. Kawabata, T. Kato, Microscopic investigation on morphologies of bilayer gel structure in the mixed polyoxyethylene-type nonionic surfactant systems, J. Phys. Chem. B 116 (2012) 12558–12566. [22] M.P. Nieh, V.A. Raghunathan, S.R. Kline, T.A. Harroun, C.Y. Huanga, J. Pencer, J. Katsaras, Spontaneously formed unilamellar vesicles with path-dependent size distribution, Langmuir 21 (2005) 6656–6661. [23] J.B. Hayter, J. Penfold, An analytic structure factor for macroion solutions, Mol. Phys. 42 (1981) 109–118. [24] D.E. Koppel, Analysis of macromolecular polydispersity in intensity correlation spectroscopy: The method of cumulants, J. Chem. Phys. 57 (1972) 4814–4820. [25] B.J. Berne, R. Pecora, Dynamic Light Scattering, Wiley, New York, 1976. [26] J.C. Selser, Y. Yeh, R.J. Baskin, A light-scattering characterization of membrane vesicles, Biophys. J. 16 (1976) 337–356. [27] J.C. Selser, Y. Yeh, Letter: a light scattering method of measuring membrane vesicle number-averaged size and size dispersion, Biophys. J. 16 (1976) 847–848. [28] F.R. Hallett, J. Watton, P. Krygsman, Vesicle sizing number distributions by dynamic light scattering, Biophys. J. 59 (1991) 357–362. [29] M. Hishida, H. Seto, N. Yamada, K. Yoshikawa, Hydration process of multistacked phospholipid bilayers to form giant vesicles, Chem. Phys. Lett. 455 (2008) 297–302. [30] M.P. Nieh, C. Glinka, S. Krueger, R. Prosser, J. Katsaras, Sans study on the effect of lanthanide ions and charged lipids on the morphology of phospholipid mixtures. small-angle neutron scattering, Biophys. J. 82 (2002) 2487–2498. [31] M. Dubois, T. Zemb, Swelling limits for bilayer microstructures: the implosion of lamellar structure versus disordered lamellae, Curr. Opin. Colloid Interface Sci. 5 (2000) 27–37.