AQUEOUS FOAMS IN THE PRESENCE OF SURFACTANT CRYSTALS

Aqueous foams are used extensively in many fields and anionic surfactants are commonly used foaming agents. However, potential trouble may arise when they are utilized in hard water areas and/or at low temperatures. Anionic surfactants, like sodium dodecyl sulfate (SDS), may precipitate in the form of crystals when the concentration of divalent counterions such as Mg 2+ exceeds a certain limit. In an attempt to prepare ultra-stable foams containing precipitated crystals, the behaviour of SDS in water was systematically investigated as a function of surfactant concentration at different concentrations of Mg(NO 3 ) 2 prior to a study of their foam properties. We quantitatively study the conversion of surfactant micelles to crystals and the redissolution of crystals into micelles. It was found that the presence of surfactant crystals reduced the initial foam volume and foam half-life but greatly improved the long-term stability of foams. Foam studies were also conducted for the supernatant and sediment isolated from crystal dispersions so that the importance of surfactant crystals to foam stability could be established. Despite the foamability of a sediment being low, an order of magnitude increase in foam half-life was related to the coverage of bubble surfaces by surfactant crystals. Both rapid cooling and ultrasonication were shown to influence the surfactant crystal shape and size with an impact on foam properties.


INTRODUCTION
An aqueous foam is a dispersion of gas bubbles in water and anionic surfactants such as sodium dodecyl sulfate (SDS) are the most commonly used foaming agents. [1][2][3][4] Surfactantstabilized aqueous foams are common and products are extensively used in many fields like food, pharmaceuticals, cleaning and cosmetics. [5][6][7][8][9][10][11] As is well-known, foams stabilized by surfactants are thermodynamically unstable and destabilization occurs through a combination of drainage, coalescence and disproportionation. [12][13][14][15][16] Unlike surfactants, solid particles with appropriate wettability can possess much higher adsorption energy to the air-water interface and give rise to foams of long-term stability. 17,18 More and more studies have focused on particle-stabilised foams in the last decade or so. [19][20][21][22] Binks and co-workers carried out a series of studies with silane-grafted fumed silica particles of varying hydrophobicity and aqueous foams generated were completely stable against coalescence. [23][24][25] Alargova et al. 26 and Liu et al. 27 produced foams stabilised by synthetic polymer microrods or plate-like layered double hydroxide particles respectively. Foam bubbles were sterically separated by a dense layer(s) of interfacial particles endowing the foam with high stability.
In the literature, examples exist in which particles are elaborately synthesized or modified ex situ. 28,29 Only a few articles report aqueous foams stabilised by solid surfactant particles. surfaces. Ionic surfactants are likely to precipitate in the form of crystals due to a decreased solubility below the Krafft point. 34,35 For anionic surfactants in particular, conditions leading to surfactant precipitation are easy to meet in conditions of low temperature or hard water.
However, the stabilization of aqueous foams in the presence of surfactant crystals has not received much attention.
We have carried out a systematic study to link the process of surfactant precipitation to foaming performance. According to the precipitation domain of magnesium dodecyl sulfate (in excess of either of the precipitation components Mg(NO3)2 or SDS), 36 quantitative studies 4 concerning the conversion of surfactant molecules from monomers to micelles and/or crystals have been conducted along with simultaneous foam studies to assess their impact on foaming ability and foam stability. Compared to the work of Zhang et al., 32,33 we have investigated over forty different concentrations for surfactant and divalent salt which cut through different regions in the phase diagram and where surfactant crystals are formed before foaming. In addition to conventional characterization of the foams, foam studies of the supernatant and sediment isolated from crystal dispersions also helped us understand the separate roles of surfactant crystals and free surfactant molecules to long-term foam stability. Finally, rapid cooling and ultrasonication of aqueous phases were found to greatly influence the shape and size of surfactant crystals formed which further impacts on foam properties.

(b) Methods
All experiments were conducted at room temperature (T = 20 ± 3 ℃) except for those involving rapid cooling.
[SDS] and [Mg(NO3)2] represent the initial concentration of SDS and Mg(NO3)2 in the system respectively (mM).  Table S1). 5 Determination of actual surfactant concentration in supernatant and sediment. Each crystal dispersion separated into a supernatant and sediment after standing for 2 days. The volume fraction of supernatant (sup) and sediment (sed) was determined from height measurements (see Figure S1). The equilibrium concentration of free surfactant in the supernatant was determined using the two-phase Epton titration involving Hyamine 1622 as a cationic titrant. 37 1 mL of supernatant was pipetted into a 100 mL stoppered mixing cylinder. 2 mL of aqueous indicator solution containing 0.14 mM disulphine blue, 0.21 mM dimidium bromide and 0.1 mM sulphuric acid was then added followed by 10 mL of Milli-Q water and 15 mL of chloroform. An aqueous solution of 1.009 mM Hyamine 1622 was then added as titrant from a burette. As titration proceeded, there was a colour change until it reached the end-point when the pink colour in the chloroform layer turned grey-blue and the yellow colour in the water layer turned green (see Figure S2). Each titration was repeated three times to determine the average. The concentration of surfactant within a sediment was calculated applying a mass balance (see ESI) and three samples were also titrated to verify these calculations.
Crystal size measurement. The particle size of surfactant crystals was measured using a  Microscopy. Surfactant crystals and foam bubbles were observed immediately after preparation using an Olympus BX51 optical microscope. Micrographs were taken with a 16-bit Olympus 6 DP70 camera with Image Pro Plus software. By using a U-POT polarizer and a reflected light analyser (U-AN, Olympus T2 Japan), crossed-polarized light was also applied for these observations. Additionally, cryo-scanning electron microscopy (SEM) was carried out on a fresh foam formed from a sediment separated from a crystal dispersion of 30 mM SDS in 10 mM Mg(NO3)2 (see ESI.) Rapid cooling and ultrasonication. Crystal dispersions of 2-30 mM SDS in 10 mM Mg(NO3)2 were held in a 60 ℃ water bath (IKA RCT basic) for 2 h and then moved to a 0 ℃ water bath (Grant Optima™ TC120) directly. They cooled down from 60 ℃ to 11 ℃ within 4 min while stirring with a glass rod. Surfactant precipitated as crystals during this rapid cooling process (14 ℃ min -1 ). These crystal dispersions showed no sign of subsequent precipitation or redissolution after storing at room temperature (T = 20 ± 3 ℃) for 5 h and foam studies were conducted using 500 mL (max. 650 mL) glass graduated cylinders. Additionally, four samples of 30 mM SDS in 10 mM Mg(NO3)2 crystal dispersions were freshly prepared at room temperature and treated using a high-intensity ultrasonic processor (100 Watt, Sonic & Materials VC100 Vibra-cell) equipped with a CV18 ultrasonic probe (tip diameter: 3 mm).
Crystal dispersions were sonicated at an amplitude of 100% for different times (1-10 min) in a 20 ℃ water bath. Foams were then formed in 500 mL (max. 650 mL) glass graduated cylinders at room temperature (T = 20 ± 3 ℃).

(a) Precipitation of SDS upon adding Mg(NO3)2
It is known that the presence of Mg(NO3)2 brings different effects to aqueous SDS solutions depending on the concentration of both SDS and Mg(NO3)2. The most common phenomenon is that an increase in salt concentration decreases the critical micelle concentration (cmc). This is a result of the charge neutralization of surfactant head groups by counterions facilitating micellization. 38 The

(b) Conversion of surfactant monomers to micelles and crystals
The appearance of crystal dispersions of 2-30 mM SDS in 10 mM Mg(NO3)2 after standing for two days is given in Figure 3(a-1) and that of crystal dispersions of 15 mM SDS in 3-100 mM Mg(NO3)2 is given in Figure 3

(c) Foam behaviour of surfactant solutions and crystal dispersions
We quantify the foamability of a liquid by its initial foam volume (Vf0) and the foam stability is discussed in terms of foam half-life (t1/2, the time taken for a foam column to collapse to half its initial volume) and relative foam volume after 2 days. Foam studies were conducted for the four series of surfactant solutions and crystal dispersions shown in Figure 1 Figure 1). Before the majority of surfactant precipitates into crystals, the foamability is not significantly affected below 2 mM Mg(NO3)2 (see Figure S8). Once surfactant crystals are formed, Vf0 decreases sharply and reaches a low plateau value above ca. 20 mM salt when no micelles remain. This demonstrates that the presence of micelles which deliver surfactant monomers is more favourable for foaming compared with that of surfactant crystals as the latter diffuse slower to bubble surfaces. As mentioned earlier, particle-stabilised foams generally possess superior stability to surfactant-stabilised ones. 26,33 However, this advantage is not apparent here in the data for foams containing surfactant crystals if we only use t1/2 to describe foam stability. Instead, we quantify the long-term stability of foams by determining the ratio of the residual foam volume after two days to the initial foam volume (Vf2d/Vf0). Figure 8 shows the variation of Vf2d/Vf0 for series 2 and 4 which involve surfactant crystal formation. Foams formed from dispersions containing an adequate amount of crystals are more stable in the long-term than micellar solutions despite the higher surfactant concentration required for the latter. We monitored the appearance of bubbles at the top and towards the bottom of the foam column as it drained for a surfactant crystal+micelle dispersion ( Figure S12). Bubbles remain spherical in the lower layer but become non-spherical and polyhedral in the upper layer. Even after 1 min the surfaces of these bubbles appear structured and it is likely that surfactant crystals are adsorbed at least partially. A proportion may also remain in the thin aqueous films between bubbles as the foam becomes dry.

(d) Comparison of supernatants, sediments and crystal dispersions
To further understand the roles of free surfactant and surfactant crystals in foam behaviour, clear supernatants and crystal-concentrated sediments were separated from crystal dispersions after standing for two days. Foam studies were carried out for supernatants and sediments in the same way as above and compared to those of intact crystal dispersions. Figures 9 and 10 show the variations of Vf0 and t1/2 as a function of actual surfactant concentration for the same volume of the three kinds of aqueous phase (series 2). Regarding the foamability, generating the same volume of foam requires the highest surfactant concentration for sediments followed by crystal dispersions and lowest for supernatants. The sequence of foamability is thus supernatant > crystal dispersion > sediment. Furthermore, we used the initial foam volume of supernatants (Vf0,sup) and sediments (Vf0,sed) and their volume fractions in crystal dispersions (ϕsup and ϕsed) to derive the initial foam volume of intact crystal dispersions using 12 The derived initial foam volume (Vf0,calculated) is plotted as a dashed line in Figure 9 where it can be seen to be very close to the experimental data. This demonstrates that free surfactant (mainly in supernatants) and surfactant crystals (mainly in sediments) both contribute in a superimposed way to the foamability of intact crystal dispersions.
By comparing surfactant concentrations needed to achieve a certain t1/2 for foams formed from supernatants, sediments and crystal dispersions, the order of foam stability is sediment > supernatant > crystal dispersion. In addition, t1/2 of foams from sediments is over an order of magnitude higher than that for both supernatants and intact crystal dispersions and the former display excellent long-term stability (see Table 1). In series 2, values of Vf2d/Vf0 for supernatant foams are either zero (complete foam collapse) or very low as a function of SDS concentration.
Those for sediment foams however increase from 0.2 to between 0.5 and 0.6 in the same range. Since a sediment is concentrated in surfactant crystals, it is predicted that the enhanced stability of foams prepared from them is due to the increased coverage of bubbles by surfaceactive surfactant crystals. To prove this, foams formed from crystal dispersions and sediments were observed using crossed-polarizers as shown in Figure 11(a-c). Foams produced from intact crystal dispersions give perfectly spherical bubbles with smooth surfaces with few crystals adsorbed. However, bubbles in foams from sediments are rougher and brighter indicating higher surfactant crystal coverage. Cryo-SEM images given in Figure 11(d) confirm the plate-like morphology of crystals and compactly covered bubble surfaces. A high surface density of adsorbed crystals can retard bubble coalescence and inter-bubble gas transfer leading to the outstanding stability of foams from sediments. This is in line with reports on ultra-stable foams in which an armoured layer of solid particles imparts excellent stability to both spherical and non-spherical bubbles. 25,52,53 The same conclusions were drawn from foam studies in series 4 for 15 mM SDS as a function of salt concentration (Figures S13-S16).

(e) Ultrasonication and rapid cooling
Since the precipitation of surfactant crystals at room temperature has limited control of both the size and shape of crystals, ultrasonication and rapid cooling were used separately to 13 effect changes in the morphology of crystals. By applying energy to break surfactant crystals prepared at room temperature, ultrasonication was employed to reduce the crystal size. The average diameter of surfactant crystals within 30 mM SDS in 10 mM Mg(NO3)2 crystal dispersions decreases with increasing time of ultrasonication, as shown in Figure 12(a). The decrease is from around 150 µm without sonication to 45 µm, 29 µm, 21 µm and 17 µm after sonicating for 1 min, 2 min, 5 min and 10 min respectively ( Figure S17). The impact of crystal size reduction on foam properties was investigated. As shown in Figure 12 Rapid cooling is an alternative way to control crystal size/morphology. While preparing crystal dispersions of SDS in 10 mM Mg(NO3)2, rapid cooling renders crystals more uniform and square in shape, as shown in microscope images of Figure 13(a). The average crystal diameter decreases to between 85 µm and 110 µm compared with those prepared at room temperature (145 µm-175 µm, Figure S18). Such a change however has virtually no impact on foam properties of crystal dispersions, as shown in Figure 13(b).