- Email: [email protected]
Defoaming effect of calcium soap
Journal of Colloid and Interface Science 279 (2004) 539–547 www.elsevier.com/locate/jcis
Defoaming effect of calcium soap Hui Zhang a , Clarence A. Miller a,∗ , Peter R. Garrett b , Kirk H. Raney c a Department of Chemical Engineering, Rice University, Houston, TX 77251-1892, USA b Department of Chemical Engineering, UMIST, PO Box 88, Manchester, M60 1QD, UK c Shell Chemical L.P., P.O. Box 1380, Houston, TX 77251-1380, USA
Received 27 January 2004; accepted 30 June 2004 Available online 28 July 2004
Abstract The effect of calcium oleate on foam stability was studied for aqueous solutions of two commonly used surfactants (anionic and nonionic) under alkaline conditions in the absence of oil. For the anionic surfactant, defoaming by calcium oleate appears to involve two mechanisms. One is that oleate and calcium ions are presumably incorporated into the surfactant monolayers with a resulting decrease in the maximum of the disjoining pressure curve and therefore produces less stable thin films. The other is bridging of the films by calcium oleate particles. The latter mechanism was especially important in freshly made solutions where precipitation in the aqueous phase was still occurring when the foam was generated. Foams generated after aging (hours) when precipitation was nearly complete were more stable even though solution turbidities were greater. Foams of the nonionic surfactant were less stable than those of the anionic surfactant but were also destabilized by sufficient amounts of calcium oleate and exhibited a similar aging effect. A simplified model was developed for estimating the sodium oleate concentration at which precipitation commences in solutions of the anionic surfactant containing dissolved calcium. It includes enhancement of calcium content in the electrical double layers of the surfactant micelles. Predictions of the model were in agreement with experiment. 2004 Elsevier Inc. All rights reserved. Keywords: Foam; Thin film; Sodium soap; Calcium soap; Defoaming; Aging effect
1. Introduction Foam stability in the presence of soluble/insoluble soaps is of practical importance in applications such as laundry, personal and home cleaning, the potential use of foam for mobility control in alkaline/surfactant processes for improved oil recovery, etc. In hard water the precipitation of calcium and magnesium soaps of long chain fatty acids may occur and destabilize the foam. Soap is sometimes added to laundry products for defoaming action and can also form in situ during the detergency process due to the presence of fatty acids in sebum-like soils. We have shown in a previous study [1] that the combination of oil and calcium soap produces a synergistic effect facilitating the bridging instability * Corresponding author. Fax: +1-713-348-5478.
E-mail address: [email protected] (C.A. Miller). 0021-9797/$ – see front matter 2004 Elsevier Inc. All rights reserved. doi:10.1016/j.jcis.2004.06.103
of foam films or Plateau borders and producing a substantial antifoam effect. However, the defoaming effect of calcium and magnesium in the absence of oils is not yet well understood. Peper [2] proposed that rapid defoaming of detergent solutions by soap or fatty acid occurs when conditions are favorable for the formation of a solid monolayer by the action of calcium ion. He advanced the hypothesis that the surfaces of the foam bubbles are heterogeneous and consist of a continuous film of adsorbed detergent in which there are islands of solid calcium soap film. He suggested that these islands make the film unstable because of their inflexible, brittle nature. In this paper we describe a systematic study of oil-free solutions containing calcium soap, which was formed by adding sodium oleate to surfactant solutions with dissolved calcium chloride and pH adjusted to 9 by sodium hydroxide. Alkalinity was applied to limit the hydrolysis of oleate to oleic acid, which would otherwise be significant due to easy incorporation of oleic acid into micelles. Measurements of
540
H. Zhang et al. / Journal of Colloid and Interface Science 279 (2004) 539–547
the rate of collapse of a foam column were supplemented by microscopic observations of individual foam films and by measurement of turbidity and surface tensions.
2. Materials Two commercial surfactants supplied by Shell Chemical L.P. (Houston, TX) were used as foaming agents. One is an anionic surfactant, an alkyl ethoxy sulfate sodium salt with a straight C12 –C15 hydrocarbon chain and an average of three ethylene oxide (EO) groups, denoted as N25-3S later on. It is typical of surfactants used in hand dishwashing and shampoo products. The other is a nonionic surfactant, a linear alcohol ethoxylate also with a C12 –C15 hydrocarbon chain but with an average of seven EO groups, denoted as N25-7 later on. It is representative of nonionic surfactants which are used in combination with anionic surfactants in household laundry products. Sodium oleate with a purity 99% was purchased from Fluka Chemie. Calcium chloride (from Alfa Products, reagent grade) was dissolved in deionized water to prepare hard water. The concentration of hardness was calculated as CaCO3 in accordance with the usual convention, i.e., the concentration that would exist if the same amount of calcium had been added as CaCO3 . Alkalinity was provided by sodium hydroxide purchased from Fisher Scientific with assay of NaOH 98.5%. Water used for experiments was distilled and deionized. Mixtures of triolein (with a purity of 99%, from Fluka Chemie) or n-hexadecane (with a purity of 99%, from the Humphrey Chemical Company) with a small amount of oleic acid (with a purity of 95%, from Fisher Scientific Company) were sometimes used as defoaming agents in experiments conducted for comparison with the oil-free systems. The mixture of hexadecane and oleic acid at weight ratio of 9 to 1 is denoted as C16/HOl, and the mixture of triolein and oleic acid at weight ratio of 9 to 1 is denoted as TO/HOl.
3. Methods 3.1. Foam stability tests Foam stability was determined by measuring the rate of collapse of a vertical foam column formed by mixing air with a surfactant solution at the base of the column [1]. The procedure consisted of filling the apparatus with approximately 440 ml of solution to the 0-cm mark of the graduated cylinder. Foam was generated while the solution was pumped through a recycle loop and air was allowed into the recycle line. The pump was shut off when the foam reached the 20-cm mark, and foam decay was recorded as a function of time.
3.2. Optical observation of horizontal foam film Millimeter sized foam films were observed in reflected monochromatic light with wavelength of 546 nm by using the method of Sheludko and Exerowa [3,4]. The film was formed from a biconcave drop placed in a capillary (i.d. 3 mm, height 1.2 mm) by sucking out liquid from a side orifice. The film was illuminated and observed in a direction perpendicular to its surfaces with a videomicroscopy system. 3.3. Surface tension measurement A University of Texas Model 300 spinning drop tensiometer was utilized to measure surface tensions. Measurements were made by introducing a small air bubble to the sample tube which was filled with surfactant solution. Data were collected after running the sample for at least 20 min. 3.4. Turbidity measurement Turbidity was measured by a Brinkmann PC 800 colorimeter. The original measured datum was the percentage of transparency, T%, of the solution. Turbidity was then calculated from 100 T%. The colorimeter was calibrated before each measurement with deionized water which gives 100 − T%. 3.5. Ultrafiltration and titration Ultrafiltration (UF) was used to separate free electrolyte solution from surfactant micelles. Hutchinson [5] showed that this could be achieved at sufficiently high rates of filtration through UF membrane. UF membranes used in this study were Amicon YM3 with 3000 nominal molecular weight limit (NMWL) from Millipore Corp. The material was regenerated cellulose. Surfactant and electrolyte solutions were charged to a stirred filtration cell of 50 to 55 mL capacity and forced through UF membrane with 330 kPa [48 psi] nitrogen pressure. Filtration rates were 0.48 to 0.53 g/min. During each filtration, successive samples of filtrate were collected and analyzed for surfactant and calcium. Concentration of anionic surfactant and calcium ion were analyzed by potentiometric titration. The apparatus was an automatic titrator, Metrohm Titrino Model 716, purchased from Brinkman Instruments. For analysis of anionic surfactant, the titrant solution is 0.004 M benzothonium chloride (Hyamine 1622), reagent, ordered from Gallord– Schlesinger, Inc. The indicator is a combination anionic surfactant specific electrode purchased from Phoenix Electrode Co., Cat. No. SUR1502R. For analysis of calcium ion, the titrant is reagent EDTA, a product of Fisher Chemicals. The indicator is a calcium ion-specific electrode purchased from Phoenix Electrode Co., Cat. No. CAL1502R.
H. Zhang et al. / Journal of Colloid and Interface Science 279 (2004) 539–547
Table 1 Turbidity measurements with 0.01 wt% N25-3S, 300 ppm (0.03%) Ca, pH 9
Table 3 Turbidity measurements with 0.1 wt% N25-7, 300 ppm (0.03%) Ca, pH 9
Transmittance, T% Fresh 4h 23 h 29 h
541
Transmittance, T%
0.0001% NaOl
0.0004% NaOl
0.001% NaOl
NaOl 0.001% 0.002%
99.9 99.9 99.8 99.9
98.8 97.7 96.8 96.6
96.3 92.6 89.6 89.6
Fresh 4h 23 h 29 h
99.9 99.9 99.9 99.9
99.6 99.7 99.3 99.5
0.003% 0.004% 99.4 2 h 95.0 25 h 92.2 34 h 93.1
91.4 83.0 77.3 79.7
0.005% 0.01% 88.7 7 h 59.3
69.5 34.6
Table 2 Turbidity measurements with 0.05 wt% N25-3S, 300 ppm (0.03%) Ca, pH 9 Transmittance, T% Fresh 2h 25 h 50 h
0.0005% NaOl
0.001% NaOl
99.8 99.9 99.7 99.6
97.9 96.5 95.8 96.1
4. Results and discussion 4.1. Turbidity As indicated previously, the main objective of this study was to determine the effect of calcium soaps on foam stability. For this purpose it is important to know whether solid soap is present. Most calcium soaps have low solubility in water, e.g., the solubility product of calcium oleate is ten to the power of −15.4 at 295 K and zero ionic strength [6]. So when sodium oleate (NaOl) is added to an alkaline aqueous phase containing hardness, turbidity of the solution is expected to increase. Accordingly, a series of turbidity studies was carried out by measuring the transmittance as a function of the concentration of NaOl and aging time. Samples were blanketed by nitrogen during aging. Transmittances of solutions containing N25-3S at two concentrations, 0.01 wt% (cmc) and 0.05 wt%, are given in Tables 1 and 2, respectively. Water hardness was 300 ppm. Regarding the stoichiometry of calcium oleate, CaOl2 , calcium is in excess for all compositions of NaOl studied. Here, fresh solution means that the measurements were conducted within 30 min after the samples were prepared. Clearly, turbidity increased with aging time over periods of hours except for the first solutions of Tables 1 and 2. This increase shows that solid calcium oleate particles are growing. Turbidity increased as the concentration of NaOl increased, indicating more precipitate formed in the solution. As concentration of N25-3S increased, turbidity decreased, indicating fewer particles were formed because more oleate and calcium were incorporated in micelles. For solutions of the nonionic surfactant N25-7 at 0.1 wt%, transmittances were measured at 300 ppm hardness and different concentrations of sodium oleate, as shown in Table 3. The solutions are clear up to 0.002% NaOl. Beyond that point, transmittance decreased rapidly with increasing con-
Fig. 1. Foam stability with 0.01% N25-3S at different concentrations of sodium oleate.
centration. Again aging effects over periods of hours were observed. 4.2. Foam stability Foam stability was tested by the method described in Section 3.1. Reproducibility was confirmed by running each system twice or more. 4.2.1. Experiments with freshly mixed solutions A series of fresh solutions containing 0.01 wt% N25-3S and different amounts of sodium oleate and hardness were tested at pH 9 for foam stability. Fig. 1 illustrates the results at 300 ppm hardness. At the lowest studied concentration of NaOl, 0.0001 wt%, foam stability was not influenced by the addition of NaOl. Nor was turbidity, according to Table 1, which indicates that calcium oleate particles were absent. As the concentration of sodium oleate increased, foam stability decreased. The lower the concentration of sodium oleate, the longer the delay before the initiation of foam decay, and the larger the final stable foam height. Substantial decrease in foam stability was found for NaOl concentration up to 0.005% with little additional effect beyond that point. Solutions containing 0.0004 and 0.0008 wt% sodium oleate were visually transparent, although transmittance was less than 100% as shown in Table 1. With 0.001% NaOl, the solution was bluish. When the concentration of sodium oleate increased to 0.005% and 0.01%, solutions were quite turbid. Foam stability was tested for a hardness range of 20– 300 ppm, the range of interest for laundry, dishwashing, and other cleaning applications. Trends similar to those of Fig. 1 were observed for other calcium concentrations, as shown in Table 4, which gives foam heights after 20 min when foam stability curves reached equilibrium heights. Since calcium
542
H. Zhang et al. / Journal of Colloid and Interface Science 279 (2004) 539–547
Table 4 Foam height (cm) after 20 min NaOl (wt%) 0.0004 0.001
Hardness 20 ppm
50 ppm
100 ppm
300 ppm
8.0 6.2
7.7 6.1
7.4 6.1
7.9 6.0
Fig. 3. Foam stability with N25-7 at different concentrations of sodium oleate.
Fig. 2. Effect of pH on foam stability with solution containing 0.01% N25-3S and 300 ppm hardness.
is in excess, foam stability is relatively insensitive to hardness. Without any sodium oleate, it was observed that alkaline pH influences the stability of foam of 0.01% N25-3S containing hardness (CaCl2 ). As shown in Fig. 2, foam is less stable at pH 9 than at neutral pH for a freshly mixed solution. The difference in stability occurs after 10 min when the dry foam consists of very thin film lamellae. It is found that pH of the initially alkaline solution decreased as the solution aged. When the solution was tested again after about 43 h, pH was around 7.5 and foam stability was similar to that of the freshly mixed solution at neutral pH. On the contrary, pH did not change in the absence of N25-3S, implying that some component in this commercial surfactant seems to slowly consume the hydroxide ions. Behavior similar to that of Fig. 2 was observed at other hardness levels besides 300 ppm, e.g., 50 ppm. A possible explanation is suggested by the work of Rutland and Pugh [7]. They made measurements with surface forces apparatus of forces between negatively charged mica plates separated by alkaline aqueous solutions. When the only counterion in solution was Na+ , there was a strong shortrange repulsion, i.e., hydration force. But when calcium was present at pH 9, they reported that substantial Ca(OH)+ was specifically adsorbed at the negatively charged surfaces and hydration forces were minimal. This is because the water of hydration of Ca(OH)+ was less strongly bound than that of Na+ and Ca2+ . The relevance of our case to their study is (1) the two surfaces of our foam films are negatively charged due to presence of anionic surfactants; (2) Na+ , Ca2+ and Ca(OH)+ are the only cations present; (3) a strong shortrange repulsion helps stabilize thin film; (4) instability of foam at alkaline pH occurs only when foam lamellae become very thin. Thus, the decrease of foam stability as pH changes from 7 to 9 may be related to Rutland and Pugh’s finding. As the solution ages, pH decreases with the result that concen-
tration of Ca(OH)+ also decreases. This may be why foam stability returns to that for solutions formed at neutral pH. For studies with nonionic surfactant N25-7, surfactant concentration was kept at 0.1 wt%. Foam stability of pH 9 solution containing 300 ppm hardness and different amounts of sodium oleate was tested. The results are shown in Fig. 3. No significant decrease in foam stability was seen up to 0.002 wt% NaOl. Indeed, foam stability seemed to increase slightly, probably owing to the surface charge introduced by adsorbed oleate ions. At this concentration the solution has virtually 100% transmittance (per Table 3). Decrease in foam stability was noticeable with 0.003% NaOl, which made the solution a little bit bluish. At higher concentrations of sodium oleate, which made the solutions turbid, foam stability was further decreased. Substantial decrease in foam stability was found up to 0.005% NaOl with little additional effect for further increases. 4.2.2. Experiments with aged solutions—aging effect in foam stability We saw previously that turbidity changed with time for solutions containing calcium soap particles. Now, we look at how aging affects foam stability. When the solutions were mixed, foam stability was not tested until after overnight or longer. During the aging process, the solutions were blanketed with nitrogen. Results are compared with those of the preceding section for fresh solutions. With concentration of the anionic surfactant N25-3S being fixed at 0.01 wt% and hardness at 300 ppm, foam stability was tested at several concentrations of sodium oleate. Fig. 4 shows the results at 0.0004% and 0.01% NaOl. Fig. 5 shows the results at 0.001% NaOl. Compared with the fresh solutions, foam stability increased after aging. The change was significant with 0.0004% and 0.01% NaOl, while even larger at the intermediate concentration of 0.001% NaOl. The solution with 0.0004% NaOl looked transparent even after aging (refer to Table 1). The solution with 0.01% NaOl was quite turbid from the beginning, obviously with substantial amount of insoluble soap particles. The solution with 0.001% NaOl looked bluish, indicating the existence of solid particles, but they must have been small. The pH of these solutions decreased from 9 to about 7.5 after aging.
H. Zhang et al. / Journal of Colloid and Interface Science 279 (2004) 539–547
543
Fig. 4. Foam stability of fresh and aged solutions of N25-3S containing 0.0004% and 0.01% NaOl. Fig. 7. Foam stability of fresh and aged solutions of N25-7 containing NaOl.
4.3. Horizontal foam film
Fig. 5. Foam stability of fresh and aged solutions of N25-3S containing 0.001% NaOl.
Fig. 6. Comparison of foam stability between fresh and aged solutions of N23-3S in the presence of oil.
As a comparison, solutions containing both oils and calcium soap were also tested after aging. The systems studied were pH 9 solutions of 0.01 wt% N25-3S plus 300 ppm hardness and 0.01 wt% oil mixtures containing n-hexadecane or triolein with 10 wt% oleic acid, C16/HOl or TO/HOl. As demonstrated in our previous work [1], calcium soaps form in situ at the oil–water interface in these systems. Fig. 6 shows that foams were still quite unstable after aging in the presence of these oils. Results with N25-7 are given in Fig. 7. Foam stability of fresh and aged solutions were compared at two concentrations of NaOl, 0.005 and 0.01 wt%. Clearly, stability of the foams increased after the solutions aged. In contrast to the system of anionic surfactant N25-3S, alkalinity of these solutions did not change as the solutions aged. When an oil mixture was present, that is for pH 9 solutions containing 0.05 wt% C16/HOl and 300 ppm hardness, foam was even less stable than without oil, as indicated by the curve marked by ‘+’ for a fresh solution, and the antifoam effect actually increased with aging, as indicated by the curve marked by dots in Fig. 7.
In order to obtain a better understanding of the defoaming mechanism of calcium soap and of the intriguing aging effect on foam stability, we studied individual foam films with diameters between 0.6 and 1 mm in Sheludko cell. A film produced from the pH 9 solution containing 0.01 wt% N25-3S and 300 ppm hardness but no NaOl was taken as a reference. It went through the thinning process illustrated in Fig. 8. Asymmetric drainage of the dimple took place within 1–2 s after the film was formed (Fig. 8a). Channels disappeared and a white planar film formed at around 1 min (Fig. 8b). After further continuous drainage, a uniform stable black film formed at around 3 min (Fig. 8c). Film thinning behavior was not influenced by addition of NaOl at concentrations lower than 0.001 wt%. Ten films produced from a freshly mixed pH 9 solution containing 0.01% N25-3S, 300 ppm hardness, and 0.001% NaOl were studied. Eight of them behaved similarly to the film with no NaOl additive. Two of them showed a different thinning pattern. After the initial asymmetric dimple drainage (within 2 s), a new dimple-like pattern formed, as shown in Fig. 9. A particle was observed in the middle of the newly formed dimple. Film drainage was somewhat slowed down due to the dimple. No white planar film was observed. A uniform black film finally formed after approximately 3 min. At a higher concentration of sodium oleate, 0.005%, rapid initial asymmetric dimple drainage was again observed. However, at around 1 min the white film was not uniform (Fig. 10a); at around 3 min, the black film was not uniform either (Fig. 10b). At an even higher concentration of sodium oleate, 0.01%, we saw stable films and less stable films with similar frequency. After the initial quick asymmetric dimple drainage, a new dimple formed, and it was still there after 1 min. Some films broke after several minutes. Some films were still stable after 8 min with effect of particles visible, as seen in Fig. 11. Many particles were seen in these films. When the solutions containing the above three concentrations of NaOl were tested again after aging 40 h, the phenomena described above were not observed. The thinning process was the same as for solutions without any oleate (Fig. 8). Although the presence of solid soap particles in
544
H. Zhang et al. / Journal of Colloid and Interface Science 279 (2004) 539–547
(a) (a)
(b)
(b) Fig. 10. Foam film with pH 9 solution containing 0.01% N25-3S, 300 ppm hardness, and 0.005% NaOl. (a) Nonuniform white film formed at around 59 s; (b) nonuniform black film formed at around 3 min.
(c) Fig. 8. Thinning process of a film produced from solution containing 0.01% N25-3S and 300 ppm hardness. (a) Asymmetric drainage of the dimple within 1–2 s after the film was formed; (b) white planar film formed at around 48 s; (c) a uniform stable black film formed at around 3 min.
Fig. 11. Foam film with pH 9 solution containing 0.01% N25-3S, 300 ppm hardness, and 0.01% NaOl.
4.4. Mechanism of defoaming by calcium soap
Fig. 9. Foam film with pH 9 solution containing 0.01% N25-3S, 300 ppm hardness, and 0.001% NaOl.
the solution was obvious, particles were all expelled during drainage and unable to stay in the film.
For foams made from alkaline solutions containing 0.01% N25-3S and 300 ppm Ca2+ , foam stability decreased, according to the results presented above, whenever enough NaOl was added to decrease light transmittance through the solution, i.e., whenever some calcium oleate particles formed. In contrast, at the lowest NaOl content investigated, 0.0001%, both transmittance and foam stability were the same as in the absence of NaOl. Visual observations in this system revealed that at low concentrations of NaOl (0.0004–0.0008%) where turbidity was low and relatively few soap particles were formed, the foams had quite uniform bubble size and decayed from the top of the foam column. Moreover, the height of the residual relatively stable foam decreased as NaOl concen-
H. Zhang et al. / Journal of Colloid and Interface Science 279 (2004) 539–547
545
be less easily swept from the film during drainage, thereby providing more time for them to enter the opposite surface and bridge the film. Whether or not such nucleation occurs, the higher supersaturation of oleate in the fresh solutions should facilitate particle entry. For instance, entry might be favored by a lower surface charge of the surfactant monolayers (see above). For the fresh solutions bridging clearly occurs more rapidly at higher NaOl concentrations, where the greater turbidity suggests that there are enough large particles to destabilize the films even during foam generation. It is noteworthy that the capillary pressures reached in the foam column (some 1000 Pa in a column 10 cm high) are considerably higher than those in our Sheludko cell (40 Pa). This difference is one possible reason that films in the Sheludko cell with NaOl concentrations up to 0.005% were stable, whereas the corresponding foams were not. Moreover, Monin et al. found that foam films in bulk foams break at lower capillary pressure than do isolated films due to the very different perturbations in the two situations [10]. Higher capillary pressures also favor particle entry into the film surfaces. Thus, even though we did not observe particle entry for the aged solutions in the Sheludko cell, we cannot exclude the possibility of entry for the same solutions in the foam column. What we can conclude, as indicated above, is that entry is facilitated in the fresh solutions. For solutions containing the nonionic surfactant N25-7 no significant adverse effect on foam stability was seen except at concentrations of NaOl of at least 0.003% (Fig. 3). Indeed, at lower concentrations of NaOl the turbidity results indicate that no calcium soap particles formed. Moreover, adsorption of oleate ions provided a surface charge, which actually increased foam stability slightly. Particle bridging is also an important defoaming mechanism, as indicated by observations of coalescence during foam generation for fresh solutions with NaOl concentrations of at least 0.004%. The increase in foam stability after aging is similar to that for the anionic surfactant and presumably occurs for the same reason. When oils were present, foam stability did not increase after aging. As discussed previously [1], the defoaming mechanism when both oil and insoluble soap are present is a synergistic effect. The small soap particles formed at the surfaces of the much larger oil drops help destabilize the pseudoemulsion films between the drops and the film surfaces, thereby facilitating entry of the drops into the surfaces. Then bridging by the drops destabilizes the films. This mechanism is not expected to change after aging.
Fig. 12. Foam stability of 0.01% NaOl at pH 9, 0.01% N25-3S, and their mixture.
tration increased (Fig. 1). At high concentrations of NaOl— at least 0.005%—where solutions were much more turbid indicating that many solid particles formed, some coalescence of air bubbles was observed during foam generation, and film breakage was observed throughout the foam during drainage. It is suggested that two mechanisms contribute to the decrease in foam stability produced by added NaOl. One is incorporation of oleate and calcium ions into the surfactant monolayers with a resulting decrease in foam film stability. Even in the absence of calcium, NaOl foams are less stable than those of N25-3S (Fig. 12). In the present experiments the oleate which adsorbs with N25-3S likely promotes additional calcium adsorption with a resulting decrease in surface charge. Whatever the precise monolayer composition this mechanism is supported by the observation that for the aged solutions where the Sheludko cell results indicate that few particles enter the film surfaces, foam stability is less than in solutions containing N25-3S alone. Also providing support is that the nearly constant heights of foams for the fresh solutions after 20–30 min, when most particles have left the films, are only slightly less than the corresponding heights for the aged solutions (Figs. 4 and 5). The small difference is due to the higher supersaturation of oleate in the fresh solutions, as also indicated by their slightly higher surface tensions (Table 5). The second mechanism is bridging by calcium oleate particles. That hydrophobic particles can destabilize foam films is well known [8]. The Sheludko cell results show that many more particles are present in the film surfaces in the fresh than in the aged solutions. This behavior suggests that particle bridging is likely responsible for much of the defoaming action observed at short times for fresh solutions. Perhaps some particles actually nucleate in the film surfaces when foam is generated during the early stages of precipitation, as suggested by Raghavachari et al. [9]. If so, they would Table 5 Surface tension (mN/m) of solutions containing soap and hardness
0.01 wt% N25-3S, 300 ppm hardness, pH 9 plus Freshly mixed solution Aged 40 h Aged 140 h
0.0% NaOl
0.001% NaOl
0.005% NaOl
0.05% NaOl
27.5 27.5 27.4
27.8 27.6 27.7
28.8 27.9 27.5
30.1
546
H. Zhang et al. / Journal of Colloid and Interface Science 279 (2004) 539–547
4.5. Precipitation of calcium oleate in micellar solutions of N25-3S In micellar solutions, precipitation of calcium soap is a more complex process than that in the absence of micelles. Since some calcium and oleate are incorporated in the micelles, concentrations of molecularly dissolved calcium and oleate are less than the total amounts added. This effect should be taken into account in the analysis to determine the conditions for precipitation. In this section, a simplified approach is applied to model such behavior. The objective is to estimate the precipitation boundary of calcium oleate, i.e., the amount of sodium oleate which must be added to a solution with given concentrations of N25-3S and CaCl2 before calcium oleate starts to precipitate. Our interest here lies in systems where the molar concentration of oleate is much less than that of N25-3S and calcium. Owing to the low solubility product of calcium oleate, the monomer concentration of oleate ion will be very low for significant amounts of calcium in bulk solution. Therefore, there should be very little oleate in the micelles, and their behavior should be similar to N25-3S micelles. So we assume that binding of calcium with micelles in the presence of a small amount sodium oleate and alkalinity can be estimated by the same model as for a solution without sodium oleate. Hirasaki and Lawson [11] have studied the association of calcium with micelles of N25-3S and justified their electrostatic model, which does not consider the fine points of ion binding interaction, by data obtained with an ultrafiltration (UF) membrane. We performed UF when NaOl was also present, determined the concentration of free calcium ions, C 0 2+ , in the filtrate, and used this value to approxiCa mate free calcium concentration in the micellar solution. If the measured concentration of free calcium ions is not available, it can be estimated from the model of Hirasaki and Lawson [11]. Anionic surfactants have been found to mix ideally in micelles. Accordingly, we assume here ideal mixing of N253S and oleate, even though the former is ethoxylated. Thus, with a solubility product Ksp of calcium oleate of 10−15 [6], 0 , is calculated by the monomer concentration of oleate, COl −
0 cmc = Ksp/C 0 2+ . With cmc of NaOl at pH 9 (COl ) COl − Ca as 0.052 mM [12], the composition of oleate in the mixed 0 /C cmc . The concentration micelle is calculated by x = COl − Ol of N25-3S in micelles is the difference between total surfactant concentration, CST , and monomer concentration, CS0 . The latter is based on measured surfactant concentration in the filtrate at the same total calcium concentration. These micelles are expected to contain oleate in the amount of m = x(C T − C 0 ). Therefore, the total oleate added at the COl − S S m + C0 . precipitation boundary is COl − Ol− The precipitation boundary calculated this way was compared with that found experimentally, which is the boundary between the isotropic region (100% transmittance) and the region with turbidity (with transmittance less than 99.8%). Results for surfactant concentrations between 0.01 and 0.05 wt% are given in Fig. 13. The calculated and experimental boundaries match well in spite of the limitations of our simplified model. As Fig. 13 shows, the effect of the micelles must certainly be included. Fig. 13 is based on experiments and calculations summarized in Table 6. It is clear from the fifth column in Table 6 that the percentage of oleate in mixed micelles was as low as 1%. This justifies our assumption that there was very little oleate in the micelles, and that binding of calcium should be similar to that of N253S micelles. In fact, experimental values of calcium binding were compared for solutions containing 0.05% N25-3S and
Fig. 13. Precipitation boundary of calcium oleate in micellar solution of N25-3S.
Table 6 Calculation of precipitation boundary at 300 ppm hardness (CaCl2 ) Conc. of N25-3S (wt%)
Conc. of N25-3S CST
Conc. of free Ca2+ C 0 2+ a
Conc. of mono. Ol− C 0 − × 104
mol% of Ol− in micelles
Conc. of Ol− in micelle C m − × 104
Total oleate Cm − + C0
Ol− (mM × 104 )
Calculated boundary (wt% × 105 )
Experiment boundary (wt% ×105 )
0.0 0.01 0.02 0.03 0.04 0.05
0.0 0.22 0.44 0.66 0.88 1.10
3.00 2.87 2.79 2.73 2.59 2.47
5.77 5.90 5.99 6.05 6.21 6.36
0.0 1.13 1.15 1.16 1.19 1.22
0.0 7.9 33.4 59.4 87.2 128.5
5.77 13.8 39.4 65.4 93.4 134.9
1.76 4.22 12.0 19.9 28.5 41.1
– 9 15 22 30 42
Ca
Ol
Ol
Ol
Note. Unit of concentration is mM if not otherwise specified; conc. of mono. is monomeric concentration. a Obtained by direct measurement of filtrate composition; C 0 (cmc of N25-3S at 300 ppm hardness) used is 0.15 mM based on the measured surfactant S concentration in the filtrate.
H. Zhang et al. / Journal of Colloid and Interface Science 279 (2004) 539–547
300 ppm hardness with and without 0.0003% NaOl. No difference was found. Although precipitation boundaries for hardness less than 300 ppm were not actually determined experimentally, theoretical calculation shows that the assumptions made for this model are reasonable for calcium concentrations as low as 50 ppm for N25-3S concentrations up to 0.05 wt%. Thus, for other situations such as low concentration of NaOl at hardness less than 300 ppm, a frequently occurring situation in laundry applications, this model can be used to predict the amount of N25-3S needed to prevent calcium oleate from precipitating. Also, this same method could be used to estimate the precipitation boundary (or at least turbidity boundary) for other anionic surfactants.
547
foam stability after aging is similar to that for the anionic surfactant and presumably occurs for the same reason. A simplified model for estimating the precipitation boundary including the enhancement of calcium content in the electrical double layers of anionic surfactant (N25-3S) micelles yielded results in agreement with experiment. This model assumes that mixing of N25-3S and oleate in the micelle is ideal; that micelles with a small amount of sodium oleate behave similarly to N25-3S micelles; and that calcium binding with micelle can be described by an electrostatic approach. It is demonstrated that predictions of this model gave reasonable agreement with the measured precipitation boundary.
Acknowledgment 5. Summary Calcium soap, without dispersed oil, can decrease foam stability for both anionic (N25-3S) and nonionic (N25-7) surfactants. For the anionic surfactant, defoaming by calcium oleate appears to involve two mechanisms. One is incorporation of oleate and calcium ions into the surfactant monolayers, which decreases the maximum of the disjoining pressure curve and therefore produces less stable thin films. The other is bridging of the films by calcium oleate particles. Foams generated from freshly mixed solutions were less stable than those generated hours later when precipitation and growth of calcium oleate particles were nearly complete. Since observations of individual foam films showed that particle entry into the film surfaces was facilitated for the fresh solutions, we conclude that particle bridging was mainly responsible for the reduced stability of their foams. Foams of the nonionic surfactant were also destabilized by sufficient amounts of calcium oleate. Particle bridging is again an important defoaming mechanism. The increase in
The authors acknowledge the National Science Foundation for supporting this work under grant CTS-9911954.
References [1] H. Zhang, C.A. Miller, P.R. Garrett, K.H. Raney, J. Colloid Interface Sci. 263 (2003) 633. [2] H. Peper, J. Colloid Sci. 13 (1958) 199. [3] A. Sheludko, D. Exerowa, Kolloid-Z. 155 (1957) 39. [4] A. Sheludko, Adv. Colloid Interface Sci. 1 (1967) 391. [5] E. Hutchinson, Z. Phys. Chem. Neue Folge 21 (1959) 38. [6] R.R. Irani, C.F. Callis, J. Phys. Chem. 64 (1960) 1741. [7] M. Rutland, R.J. Pugh, Colloids Surf. A Physicochem. Eng. Aspects 125 (1997) 33. [8] P.R. Garrett, in: P.R. Garrett (Ed.), Defoaming: Theory and Industrial Application, Dekker, New York, 1993, chap. 1. [9] R. Raghavachari, K.S. Narayan, N. Nayaar, in: P.R. Garrett (Ed.), Abstracts of Eurofoam Conference, 2002. [10] D. Monin, A. Espert, A. Colin, Langmuir 16 (2000) 3873. [11] G.J. Hirasaki, J.B. Lawson, SPE Reservoir Eng. (March 1986) 119. [12] K. Theander, R.J. Pugh, J. Colloid Interface Sci. 239 (2001) 209.
Defoaming effect of calcium soap Hui Zhang a , Clarence A. Miller a,∗ , Peter R. Garrett b , Kirk H. Raney c a Department of Chemical Engineering, Rice University, Houston, TX 77251-1892, USA b Department of Chemical Engineering, UMIST, PO Box 88, Manchester, M60 1QD, UK c Shell Chemical L.P., P.O. Box 1380, Houston, TX 77251-1380, USA
Received 27 January 2004; accepted 30 June 2004 Available online 28 July 2004
Abstract The effect of calcium oleate on foam stability was studied for aqueous solutions of two commonly used surfactants (anionic and nonionic) under alkaline conditions in the absence of oil. For the anionic surfactant, defoaming by calcium oleate appears to involve two mechanisms. One is that oleate and calcium ions are presumably incorporated into the surfactant monolayers with a resulting decrease in the maximum of the disjoining pressure curve and therefore produces less stable thin films. The other is bridging of the films by calcium oleate particles. The latter mechanism was especially important in freshly made solutions where precipitation in the aqueous phase was still occurring when the foam was generated. Foams generated after aging (hours) when precipitation was nearly complete were more stable even though solution turbidities were greater. Foams of the nonionic surfactant were less stable than those of the anionic surfactant but were also destabilized by sufficient amounts of calcium oleate and exhibited a similar aging effect. A simplified model was developed for estimating the sodium oleate concentration at which precipitation commences in solutions of the anionic surfactant containing dissolved calcium. It includes enhancement of calcium content in the electrical double layers of the surfactant micelles. Predictions of the model were in agreement with experiment. 2004 Elsevier Inc. All rights reserved. Keywords: Foam; Thin film; Sodium soap; Calcium soap; Defoaming; Aging effect
1. Introduction Foam stability in the presence of soluble/insoluble soaps is of practical importance in applications such as laundry, personal and home cleaning, the potential use of foam for mobility control in alkaline/surfactant processes for improved oil recovery, etc. In hard water the precipitation of calcium and magnesium soaps of long chain fatty acids may occur and destabilize the foam. Soap is sometimes added to laundry products for defoaming action and can also form in situ during the detergency process due to the presence of fatty acids in sebum-like soils. We have shown in a previous study [1] that the combination of oil and calcium soap produces a synergistic effect facilitating the bridging instability * Corresponding author. Fax: +1-713-348-5478.
E-mail address: [email protected] (C.A. Miller). 0021-9797/$ – see front matter 2004 Elsevier Inc. All rights reserved. doi:10.1016/j.jcis.2004.06.103
of foam films or Plateau borders and producing a substantial antifoam effect. However, the defoaming effect of calcium and magnesium in the absence of oils is not yet well understood. Peper [2] proposed that rapid defoaming of detergent solutions by soap or fatty acid occurs when conditions are favorable for the formation of a solid monolayer by the action of calcium ion. He advanced the hypothesis that the surfaces of the foam bubbles are heterogeneous and consist of a continuous film of adsorbed detergent in which there are islands of solid calcium soap film. He suggested that these islands make the film unstable because of their inflexible, brittle nature. In this paper we describe a systematic study of oil-free solutions containing calcium soap, which was formed by adding sodium oleate to surfactant solutions with dissolved calcium chloride and pH adjusted to 9 by sodium hydroxide. Alkalinity was applied to limit the hydrolysis of oleate to oleic acid, which would otherwise be significant due to easy incorporation of oleic acid into micelles. Measurements of
540
H. Zhang et al. / Journal of Colloid and Interface Science 279 (2004) 539–547
the rate of collapse of a foam column were supplemented by microscopic observations of individual foam films and by measurement of turbidity and surface tensions.
2. Materials Two commercial surfactants supplied by Shell Chemical L.P. (Houston, TX) were used as foaming agents. One is an anionic surfactant, an alkyl ethoxy sulfate sodium salt with a straight C12 –C15 hydrocarbon chain and an average of three ethylene oxide (EO) groups, denoted as N25-3S later on. It is typical of surfactants used in hand dishwashing and shampoo products. The other is a nonionic surfactant, a linear alcohol ethoxylate also with a C12 –C15 hydrocarbon chain but with an average of seven EO groups, denoted as N25-7 later on. It is representative of nonionic surfactants which are used in combination with anionic surfactants in household laundry products. Sodium oleate with a purity 99% was purchased from Fluka Chemie. Calcium chloride (from Alfa Products, reagent grade) was dissolved in deionized water to prepare hard water. The concentration of hardness was calculated as CaCO3 in accordance with the usual convention, i.e., the concentration that would exist if the same amount of calcium had been added as CaCO3 . Alkalinity was provided by sodium hydroxide purchased from Fisher Scientific with assay of NaOH 98.5%. Water used for experiments was distilled and deionized. Mixtures of triolein (with a purity of 99%, from Fluka Chemie) or n-hexadecane (with a purity of 99%, from the Humphrey Chemical Company) with a small amount of oleic acid (with a purity of 95%, from Fisher Scientific Company) were sometimes used as defoaming agents in experiments conducted for comparison with the oil-free systems. The mixture of hexadecane and oleic acid at weight ratio of 9 to 1 is denoted as C16/HOl, and the mixture of triolein and oleic acid at weight ratio of 9 to 1 is denoted as TO/HOl.
3. Methods 3.1. Foam stability tests Foam stability was determined by measuring the rate of collapse of a vertical foam column formed by mixing air with a surfactant solution at the base of the column [1]. The procedure consisted of filling the apparatus with approximately 440 ml of solution to the 0-cm mark of the graduated cylinder. Foam was generated while the solution was pumped through a recycle loop and air was allowed into the recycle line. The pump was shut off when the foam reached the 20-cm mark, and foam decay was recorded as a function of time.
3.2. Optical observation of horizontal foam film Millimeter sized foam films were observed in reflected monochromatic light with wavelength of 546 nm by using the method of Sheludko and Exerowa [3,4]. The film was formed from a biconcave drop placed in a capillary (i.d. 3 mm, height 1.2 mm) by sucking out liquid from a side orifice. The film was illuminated and observed in a direction perpendicular to its surfaces with a videomicroscopy system. 3.3. Surface tension measurement A University of Texas Model 300 spinning drop tensiometer was utilized to measure surface tensions. Measurements were made by introducing a small air bubble to the sample tube which was filled with surfactant solution. Data were collected after running the sample for at least 20 min. 3.4. Turbidity measurement Turbidity was measured by a Brinkmann PC 800 colorimeter. The original measured datum was the percentage of transparency, T%, of the solution. Turbidity was then calculated from 100 T%. The colorimeter was calibrated before each measurement with deionized water which gives 100 − T%. 3.5. Ultrafiltration and titration Ultrafiltration (UF) was used to separate free electrolyte solution from surfactant micelles. Hutchinson [5] showed that this could be achieved at sufficiently high rates of filtration through UF membrane. UF membranes used in this study were Amicon YM3 with 3000 nominal molecular weight limit (NMWL) from Millipore Corp. The material was regenerated cellulose. Surfactant and electrolyte solutions were charged to a stirred filtration cell of 50 to 55 mL capacity and forced through UF membrane with 330 kPa [48 psi] nitrogen pressure. Filtration rates were 0.48 to 0.53 g/min. During each filtration, successive samples of filtrate were collected and analyzed for surfactant and calcium. Concentration of anionic surfactant and calcium ion were analyzed by potentiometric titration. The apparatus was an automatic titrator, Metrohm Titrino Model 716, purchased from Brinkman Instruments. For analysis of anionic surfactant, the titrant solution is 0.004 M benzothonium chloride (Hyamine 1622), reagent, ordered from Gallord– Schlesinger, Inc. The indicator is a combination anionic surfactant specific electrode purchased from Phoenix Electrode Co., Cat. No. SUR1502R. For analysis of calcium ion, the titrant is reagent EDTA, a product of Fisher Chemicals. The indicator is a calcium ion-specific electrode purchased from Phoenix Electrode Co., Cat. No. CAL1502R.
H. Zhang et al. / Journal of Colloid and Interface Science 279 (2004) 539–547
Table 1 Turbidity measurements with 0.01 wt% N25-3S, 300 ppm (0.03%) Ca, pH 9
Table 3 Turbidity measurements with 0.1 wt% N25-7, 300 ppm (0.03%) Ca, pH 9
Transmittance, T% Fresh 4h 23 h 29 h
541
Transmittance, T%
0.0001% NaOl
0.0004% NaOl
0.001% NaOl
NaOl 0.001% 0.002%
99.9 99.9 99.8 99.9
98.8 97.7 96.8 96.6
96.3 92.6 89.6 89.6
Fresh 4h 23 h 29 h
99.9 99.9 99.9 99.9
99.6 99.7 99.3 99.5
0.003% 0.004% 99.4 2 h 95.0 25 h 92.2 34 h 93.1
91.4 83.0 77.3 79.7
0.005% 0.01% 88.7 7 h 59.3
69.5 34.6
Table 2 Turbidity measurements with 0.05 wt% N25-3S, 300 ppm (0.03%) Ca, pH 9 Transmittance, T% Fresh 2h 25 h 50 h
0.0005% NaOl
0.001% NaOl
99.8 99.9 99.7 99.6
97.9 96.5 95.8 96.1
4. Results and discussion 4.1. Turbidity As indicated previously, the main objective of this study was to determine the effect of calcium soaps on foam stability. For this purpose it is important to know whether solid soap is present. Most calcium soaps have low solubility in water, e.g., the solubility product of calcium oleate is ten to the power of −15.4 at 295 K and zero ionic strength [6]. So when sodium oleate (NaOl) is added to an alkaline aqueous phase containing hardness, turbidity of the solution is expected to increase. Accordingly, a series of turbidity studies was carried out by measuring the transmittance as a function of the concentration of NaOl and aging time. Samples were blanketed by nitrogen during aging. Transmittances of solutions containing N25-3S at two concentrations, 0.01 wt% (cmc) and 0.05 wt%, are given in Tables 1 and 2, respectively. Water hardness was 300 ppm. Regarding the stoichiometry of calcium oleate, CaOl2 , calcium is in excess for all compositions of NaOl studied. Here, fresh solution means that the measurements were conducted within 30 min after the samples were prepared. Clearly, turbidity increased with aging time over periods of hours except for the first solutions of Tables 1 and 2. This increase shows that solid calcium oleate particles are growing. Turbidity increased as the concentration of NaOl increased, indicating more precipitate formed in the solution. As concentration of N25-3S increased, turbidity decreased, indicating fewer particles were formed because more oleate and calcium were incorporated in micelles. For solutions of the nonionic surfactant N25-7 at 0.1 wt%, transmittances were measured at 300 ppm hardness and different concentrations of sodium oleate, as shown in Table 3. The solutions are clear up to 0.002% NaOl. Beyond that point, transmittance decreased rapidly with increasing con-
Fig. 1. Foam stability with 0.01% N25-3S at different concentrations of sodium oleate.
centration. Again aging effects over periods of hours were observed. 4.2. Foam stability Foam stability was tested by the method described in Section 3.1. Reproducibility was confirmed by running each system twice or more. 4.2.1. Experiments with freshly mixed solutions A series of fresh solutions containing 0.01 wt% N25-3S and different amounts of sodium oleate and hardness were tested at pH 9 for foam stability. Fig. 1 illustrates the results at 300 ppm hardness. At the lowest studied concentration of NaOl, 0.0001 wt%, foam stability was not influenced by the addition of NaOl. Nor was turbidity, according to Table 1, which indicates that calcium oleate particles were absent. As the concentration of sodium oleate increased, foam stability decreased. The lower the concentration of sodium oleate, the longer the delay before the initiation of foam decay, and the larger the final stable foam height. Substantial decrease in foam stability was found for NaOl concentration up to 0.005% with little additional effect beyond that point. Solutions containing 0.0004 and 0.0008 wt% sodium oleate were visually transparent, although transmittance was less than 100% as shown in Table 1. With 0.001% NaOl, the solution was bluish. When the concentration of sodium oleate increased to 0.005% and 0.01%, solutions were quite turbid. Foam stability was tested for a hardness range of 20– 300 ppm, the range of interest for laundry, dishwashing, and other cleaning applications. Trends similar to those of Fig. 1 were observed for other calcium concentrations, as shown in Table 4, which gives foam heights after 20 min when foam stability curves reached equilibrium heights. Since calcium
542
H. Zhang et al. / Journal of Colloid and Interface Science 279 (2004) 539–547
Table 4 Foam height (cm) after 20 min NaOl (wt%) 0.0004 0.001
Hardness 20 ppm
50 ppm
100 ppm
300 ppm
8.0 6.2
7.7 6.1
7.4 6.1
7.9 6.0
Fig. 3. Foam stability with N25-7 at different concentrations of sodium oleate.
Fig. 2. Effect of pH on foam stability with solution containing 0.01% N25-3S and 300 ppm hardness.
is in excess, foam stability is relatively insensitive to hardness. Without any sodium oleate, it was observed that alkaline pH influences the stability of foam of 0.01% N25-3S containing hardness (CaCl2 ). As shown in Fig. 2, foam is less stable at pH 9 than at neutral pH for a freshly mixed solution. The difference in stability occurs after 10 min when the dry foam consists of very thin film lamellae. It is found that pH of the initially alkaline solution decreased as the solution aged. When the solution was tested again after about 43 h, pH was around 7.5 and foam stability was similar to that of the freshly mixed solution at neutral pH. On the contrary, pH did not change in the absence of N25-3S, implying that some component in this commercial surfactant seems to slowly consume the hydroxide ions. Behavior similar to that of Fig. 2 was observed at other hardness levels besides 300 ppm, e.g., 50 ppm. A possible explanation is suggested by the work of Rutland and Pugh [7]. They made measurements with surface forces apparatus of forces between negatively charged mica plates separated by alkaline aqueous solutions. When the only counterion in solution was Na+ , there was a strong shortrange repulsion, i.e., hydration force. But when calcium was present at pH 9, they reported that substantial Ca(OH)+ was specifically adsorbed at the negatively charged surfaces and hydration forces were minimal. This is because the water of hydration of Ca(OH)+ was less strongly bound than that of Na+ and Ca2+ . The relevance of our case to their study is (1) the two surfaces of our foam films are negatively charged due to presence of anionic surfactants; (2) Na+ , Ca2+ and Ca(OH)+ are the only cations present; (3) a strong shortrange repulsion helps stabilize thin film; (4) instability of foam at alkaline pH occurs only when foam lamellae become very thin. Thus, the decrease of foam stability as pH changes from 7 to 9 may be related to Rutland and Pugh’s finding. As the solution ages, pH decreases with the result that concen-
tration of Ca(OH)+ also decreases. This may be why foam stability returns to that for solutions formed at neutral pH. For studies with nonionic surfactant N25-7, surfactant concentration was kept at 0.1 wt%. Foam stability of pH 9 solution containing 300 ppm hardness and different amounts of sodium oleate was tested. The results are shown in Fig. 3. No significant decrease in foam stability was seen up to 0.002 wt% NaOl. Indeed, foam stability seemed to increase slightly, probably owing to the surface charge introduced by adsorbed oleate ions. At this concentration the solution has virtually 100% transmittance (per Table 3). Decrease in foam stability was noticeable with 0.003% NaOl, which made the solution a little bit bluish. At higher concentrations of sodium oleate, which made the solutions turbid, foam stability was further decreased. Substantial decrease in foam stability was found up to 0.005% NaOl with little additional effect for further increases. 4.2.2. Experiments with aged solutions—aging effect in foam stability We saw previously that turbidity changed with time for solutions containing calcium soap particles. Now, we look at how aging affects foam stability. When the solutions were mixed, foam stability was not tested until after overnight or longer. During the aging process, the solutions were blanketed with nitrogen. Results are compared with those of the preceding section for fresh solutions. With concentration of the anionic surfactant N25-3S being fixed at 0.01 wt% and hardness at 300 ppm, foam stability was tested at several concentrations of sodium oleate. Fig. 4 shows the results at 0.0004% and 0.01% NaOl. Fig. 5 shows the results at 0.001% NaOl. Compared with the fresh solutions, foam stability increased after aging. The change was significant with 0.0004% and 0.01% NaOl, while even larger at the intermediate concentration of 0.001% NaOl. The solution with 0.0004% NaOl looked transparent even after aging (refer to Table 1). The solution with 0.01% NaOl was quite turbid from the beginning, obviously with substantial amount of insoluble soap particles. The solution with 0.001% NaOl looked bluish, indicating the existence of solid particles, but they must have been small. The pH of these solutions decreased from 9 to about 7.5 after aging.
H. Zhang et al. / Journal of Colloid and Interface Science 279 (2004) 539–547
543
Fig. 4. Foam stability of fresh and aged solutions of N25-3S containing 0.0004% and 0.01% NaOl. Fig. 7. Foam stability of fresh and aged solutions of N25-7 containing NaOl.
4.3. Horizontal foam film
Fig. 5. Foam stability of fresh and aged solutions of N25-3S containing 0.001% NaOl.
Fig. 6. Comparison of foam stability between fresh and aged solutions of N23-3S in the presence of oil.
As a comparison, solutions containing both oils and calcium soap were also tested after aging. The systems studied were pH 9 solutions of 0.01 wt% N25-3S plus 300 ppm hardness and 0.01 wt% oil mixtures containing n-hexadecane or triolein with 10 wt% oleic acid, C16/HOl or TO/HOl. As demonstrated in our previous work [1], calcium soaps form in situ at the oil–water interface in these systems. Fig. 6 shows that foams were still quite unstable after aging in the presence of these oils. Results with N25-7 are given in Fig. 7. Foam stability of fresh and aged solutions were compared at two concentrations of NaOl, 0.005 and 0.01 wt%. Clearly, stability of the foams increased after the solutions aged. In contrast to the system of anionic surfactant N25-3S, alkalinity of these solutions did not change as the solutions aged. When an oil mixture was present, that is for pH 9 solutions containing 0.05 wt% C16/HOl and 300 ppm hardness, foam was even less stable than without oil, as indicated by the curve marked by ‘+’ for a fresh solution, and the antifoam effect actually increased with aging, as indicated by the curve marked by dots in Fig. 7.
In order to obtain a better understanding of the defoaming mechanism of calcium soap and of the intriguing aging effect on foam stability, we studied individual foam films with diameters between 0.6 and 1 mm in Sheludko cell. A film produced from the pH 9 solution containing 0.01 wt% N25-3S and 300 ppm hardness but no NaOl was taken as a reference. It went through the thinning process illustrated in Fig. 8. Asymmetric drainage of the dimple took place within 1–2 s after the film was formed (Fig. 8a). Channels disappeared and a white planar film formed at around 1 min (Fig. 8b). After further continuous drainage, a uniform stable black film formed at around 3 min (Fig. 8c). Film thinning behavior was not influenced by addition of NaOl at concentrations lower than 0.001 wt%. Ten films produced from a freshly mixed pH 9 solution containing 0.01% N25-3S, 300 ppm hardness, and 0.001% NaOl were studied. Eight of them behaved similarly to the film with no NaOl additive. Two of them showed a different thinning pattern. After the initial asymmetric dimple drainage (within 2 s), a new dimple-like pattern formed, as shown in Fig. 9. A particle was observed in the middle of the newly formed dimple. Film drainage was somewhat slowed down due to the dimple. No white planar film was observed. A uniform black film finally formed after approximately 3 min. At a higher concentration of sodium oleate, 0.005%, rapid initial asymmetric dimple drainage was again observed. However, at around 1 min the white film was not uniform (Fig. 10a); at around 3 min, the black film was not uniform either (Fig. 10b). At an even higher concentration of sodium oleate, 0.01%, we saw stable films and less stable films with similar frequency. After the initial quick asymmetric dimple drainage, a new dimple formed, and it was still there after 1 min. Some films broke after several minutes. Some films were still stable after 8 min with effect of particles visible, as seen in Fig. 11. Many particles were seen in these films. When the solutions containing the above three concentrations of NaOl were tested again after aging 40 h, the phenomena described above were not observed. The thinning process was the same as for solutions without any oleate (Fig. 8). Although the presence of solid soap particles in
544
H. Zhang et al. / Journal of Colloid and Interface Science 279 (2004) 539–547
(a) (a)
(b)
(b) Fig. 10. Foam film with pH 9 solution containing 0.01% N25-3S, 300 ppm hardness, and 0.005% NaOl. (a) Nonuniform white film formed at around 59 s; (b) nonuniform black film formed at around 3 min.
(c) Fig. 8. Thinning process of a film produced from solution containing 0.01% N25-3S and 300 ppm hardness. (a) Asymmetric drainage of the dimple within 1–2 s after the film was formed; (b) white planar film formed at around 48 s; (c) a uniform stable black film formed at around 3 min.
Fig. 11. Foam film with pH 9 solution containing 0.01% N25-3S, 300 ppm hardness, and 0.01% NaOl.
4.4. Mechanism of defoaming by calcium soap
Fig. 9. Foam film with pH 9 solution containing 0.01% N25-3S, 300 ppm hardness, and 0.001% NaOl.
the solution was obvious, particles were all expelled during drainage and unable to stay in the film.
For foams made from alkaline solutions containing 0.01% N25-3S and 300 ppm Ca2+ , foam stability decreased, according to the results presented above, whenever enough NaOl was added to decrease light transmittance through the solution, i.e., whenever some calcium oleate particles formed. In contrast, at the lowest NaOl content investigated, 0.0001%, both transmittance and foam stability were the same as in the absence of NaOl. Visual observations in this system revealed that at low concentrations of NaOl (0.0004–0.0008%) where turbidity was low and relatively few soap particles were formed, the foams had quite uniform bubble size and decayed from the top of the foam column. Moreover, the height of the residual relatively stable foam decreased as NaOl concen-
H. Zhang et al. / Journal of Colloid and Interface Science 279 (2004) 539–547
545
be less easily swept from the film during drainage, thereby providing more time for them to enter the opposite surface and bridge the film. Whether or not such nucleation occurs, the higher supersaturation of oleate in the fresh solutions should facilitate particle entry. For instance, entry might be favored by a lower surface charge of the surfactant monolayers (see above). For the fresh solutions bridging clearly occurs more rapidly at higher NaOl concentrations, where the greater turbidity suggests that there are enough large particles to destabilize the films even during foam generation. It is noteworthy that the capillary pressures reached in the foam column (some 1000 Pa in a column 10 cm high) are considerably higher than those in our Sheludko cell (40 Pa). This difference is one possible reason that films in the Sheludko cell with NaOl concentrations up to 0.005% were stable, whereas the corresponding foams were not. Moreover, Monin et al. found that foam films in bulk foams break at lower capillary pressure than do isolated films due to the very different perturbations in the two situations [10]. Higher capillary pressures also favor particle entry into the film surfaces. Thus, even though we did not observe particle entry for the aged solutions in the Sheludko cell, we cannot exclude the possibility of entry for the same solutions in the foam column. What we can conclude, as indicated above, is that entry is facilitated in the fresh solutions. For solutions containing the nonionic surfactant N25-7 no significant adverse effect on foam stability was seen except at concentrations of NaOl of at least 0.003% (Fig. 3). Indeed, at lower concentrations of NaOl the turbidity results indicate that no calcium soap particles formed. Moreover, adsorption of oleate ions provided a surface charge, which actually increased foam stability slightly. Particle bridging is also an important defoaming mechanism, as indicated by observations of coalescence during foam generation for fresh solutions with NaOl concentrations of at least 0.004%. The increase in foam stability after aging is similar to that for the anionic surfactant and presumably occurs for the same reason. When oils were present, foam stability did not increase after aging. As discussed previously [1], the defoaming mechanism when both oil and insoluble soap are present is a synergistic effect. The small soap particles formed at the surfaces of the much larger oil drops help destabilize the pseudoemulsion films between the drops and the film surfaces, thereby facilitating entry of the drops into the surfaces. Then bridging by the drops destabilizes the films. This mechanism is not expected to change after aging.
Fig. 12. Foam stability of 0.01% NaOl at pH 9, 0.01% N25-3S, and their mixture.
tration increased (Fig. 1). At high concentrations of NaOl— at least 0.005%—where solutions were much more turbid indicating that many solid particles formed, some coalescence of air bubbles was observed during foam generation, and film breakage was observed throughout the foam during drainage. It is suggested that two mechanisms contribute to the decrease in foam stability produced by added NaOl. One is incorporation of oleate and calcium ions into the surfactant monolayers with a resulting decrease in foam film stability. Even in the absence of calcium, NaOl foams are less stable than those of N25-3S (Fig. 12). In the present experiments the oleate which adsorbs with N25-3S likely promotes additional calcium adsorption with a resulting decrease in surface charge. Whatever the precise monolayer composition this mechanism is supported by the observation that for the aged solutions where the Sheludko cell results indicate that few particles enter the film surfaces, foam stability is less than in solutions containing N25-3S alone. Also providing support is that the nearly constant heights of foams for the fresh solutions after 20–30 min, when most particles have left the films, are only slightly less than the corresponding heights for the aged solutions (Figs. 4 and 5). The small difference is due to the higher supersaturation of oleate in the fresh solutions, as also indicated by their slightly higher surface tensions (Table 5). The second mechanism is bridging by calcium oleate particles. That hydrophobic particles can destabilize foam films is well known [8]. The Sheludko cell results show that many more particles are present in the film surfaces in the fresh than in the aged solutions. This behavior suggests that particle bridging is likely responsible for much of the defoaming action observed at short times for fresh solutions. Perhaps some particles actually nucleate in the film surfaces when foam is generated during the early stages of precipitation, as suggested by Raghavachari et al. [9]. If so, they would Table 5 Surface tension (mN/m) of solutions containing soap and hardness
0.01 wt% N25-3S, 300 ppm hardness, pH 9 plus Freshly mixed solution Aged 40 h Aged 140 h
0.0% NaOl
0.001% NaOl
0.005% NaOl
0.05% NaOl
27.5 27.5 27.4
27.8 27.6 27.7
28.8 27.9 27.5
30.1
546
H. Zhang et al. / Journal of Colloid and Interface Science 279 (2004) 539–547
4.5. Precipitation of calcium oleate in micellar solutions of N25-3S In micellar solutions, precipitation of calcium soap is a more complex process than that in the absence of micelles. Since some calcium and oleate are incorporated in the micelles, concentrations of molecularly dissolved calcium and oleate are less than the total amounts added. This effect should be taken into account in the analysis to determine the conditions for precipitation. In this section, a simplified approach is applied to model such behavior. The objective is to estimate the precipitation boundary of calcium oleate, i.e., the amount of sodium oleate which must be added to a solution with given concentrations of N25-3S and CaCl2 before calcium oleate starts to precipitate. Our interest here lies in systems where the molar concentration of oleate is much less than that of N25-3S and calcium. Owing to the low solubility product of calcium oleate, the monomer concentration of oleate ion will be very low for significant amounts of calcium in bulk solution. Therefore, there should be very little oleate in the micelles, and their behavior should be similar to N25-3S micelles. So we assume that binding of calcium with micelles in the presence of a small amount sodium oleate and alkalinity can be estimated by the same model as for a solution without sodium oleate. Hirasaki and Lawson [11] have studied the association of calcium with micelles of N25-3S and justified their electrostatic model, which does not consider the fine points of ion binding interaction, by data obtained with an ultrafiltration (UF) membrane. We performed UF when NaOl was also present, determined the concentration of free calcium ions, C 0 2+ , in the filtrate, and used this value to approxiCa mate free calcium concentration in the micellar solution. If the measured concentration of free calcium ions is not available, it can be estimated from the model of Hirasaki and Lawson [11]. Anionic surfactants have been found to mix ideally in micelles. Accordingly, we assume here ideal mixing of N253S and oleate, even though the former is ethoxylated. Thus, with a solubility product Ksp of calcium oleate of 10−15 [6], 0 , is calculated by the monomer concentration of oleate, COl −
0 cmc = Ksp/C 0 2+ . With cmc of NaOl at pH 9 (COl ) COl − Ca as 0.052 mM [12], the composition of oleate in the mixed 0 /C cmc . The concentration micelle is calculated by x = COl − Ol of N25-3S in micelles is the difference between total surfactant concentration, CST , and monomer concentration, CS0 . The latter is based on measured surfactant concentration in the filtrate at the same total calcium concentration. These micelles are expected to contain oleate in the amount of m = x(C T − C 0 ). Therefore, the total oleate added at the COl − S S m + C0 . precipitation boundary is COl − Ol− The precipitation boundary calculated this way was compared with that found experimentally, which is the boundary between the isotropic region (100% transmittance) and the region with turbidity (with transmittance less than 99.8%). Results for surfactant concentrations between 0.01 and 0.05 wt% are given in Fig. 13. The calculated and experimental boundaries match well in spite of the limitations of our simplified model. As Fig. 13 shows, the effect of the micelles must certainly be included. Fig. 13 is based on experiments and calculations summarized in Table 6. It is clear from the fifth column in Table 6 that the percentage of oleate in mixed micelles was as low as 1%. This justifies our assumption that there was very little oleate in the micelles, and that binding of calcium should be similar to that of N253S micelles. In fact, experimental values of calcium binding were compared for solutions containing 0.05% N25-3S and
Fig. 13. Precipitation boundary of calcium oleate in micellar solution of N25-3S.
Table 6 Calculation of precipitation boundary at 300 ppm hardness (CaCl2 ) Conc. of N25-3S (wt%)
Conc. of N25-3S CST
Conc. of free Ca2+ C 0 2+ a
Conc. of mono. Ol− C 0 − × 104
mol% of Ol− in micelles
Conc. of Ol− in micelle C m − × 104
Total oleate Cm − + C0
Ol− (mM × 104 )
Calculated boundary (wt% × 105 )
Experiment boundary (wt% ×105 )
0.0 0.01 0.02 0.03 0.04 0.05
0.0 0.22 0.44 0.66 0.88 1.10
3.00 2.87 2.79 2.73 2.59 2.47
5.77 5.90 5.99 6.05 6.21 6.36
0.0 1.13 1.15 1.16 1.19 1.22
0.0 7.9 33.4 59.4 87.2 128.5
5.77 13.8 39.4 65.4 93.4 134.9
1.76 4.22 12.0 19.9 28.5 41.1
– 9 15 22 30 42
Ca
Ol
Ol
Ol
Note. Unit of concentration is mM if not otherwise specified; conc. of mono. is monomeric concentration. a Obtained by direct measurement of filtrate composition; C 0 (cmc of N25-3S at 300 ppm hardness) used is 0.15 mM based on the measured surfactant S concentration in the filtrate.
H. Zhang et al. / Journal of Colloid and Interface Science 279 (2004) 539–547
300 ppm hardness with and without 0.0003% NaOl. No difference was found. Although precipitation boundaries for hardness less than 300 ppm were not actually determined experimentally, theoretical calculation shows that the assumptions made for this model are reasonable for calcium concentrations as low as 50 ppm for N25-3S concentrations up to 0.05 wt%. Thus, for other situations such as low concentration of NaOl at hardness less than 300 ppm, a frequently occurring situation in laundry applications, this model can be used to predict the amount of N25-3S needed to prevent calcium oleate from precipitating. Also, this same method could be used to estimate the precipitation boundary (or at least turbidity boundary) for other anionic surfactants.
547
foam stability after aging is similar to that for the anionic surfactant and presumably occurs for the same reason. A simplified model for estimating the precipitation boundary including the enhancement of calcium content in the electrical double layers of anionic surfactant (N25-3S) micelles yielded results in agreement with experiment. This model assumes that mixing of N25-3S and oleate in the micelle is ideal; that micelles with a small amount of sodium oleate behave similarly to N25-3S micelles; and that calcium binding with micelle can be described by an electrostatic approach. It is demonstrated that predictions of this model gave reasonable agreement with the measured precipitation boundary.
Acknowledgment 5. Summary Calcium soap, without dispersed oil, can decrease foam stability for both anionic (N25-3S) and nonionic (N25-7) surfactants. For the anionic surfactant, defoaming by calcium oleate appears to involve two mechanisms. One is incorporation of oleate and calcium ions into the surfactant monolayers, which decreases the maximum of the disjoining pressure curve and therefore produces less stable thin films. The other is bridging of the films by calcium oleate particles. Foams generated from freshly mixed solutions were less stable than those generated hours later when precipitation and growth of calcium oleate particles were nearly complete. Since observations of individual foam films showed that particle entry into the film surfaces was facilitated for the fresh solutions, we conclude that particle bridging was mainly responsible for the reduced stability of their foams. Foams of the nonionic surfactant were also destabilized by sufficient amounts of calcium oleate. Particle bridging is again an important defoaming mechanism. The increase in
The authors acknowledge the National Science Foundation for supporting this work under grant CTS-9911954.
References [1] H. Zhang, C.A. Miller, P.R. Garrett, K.H. Raney, J. Colloid Interface Sci. 263 (2003) 633. [2] H. Peper, J. Colloid Sci. 13 (1958) 199. [3] A. Sheludko, D. Exerowa, Kolloid-Z. 155 (1957) 39. [4] A. Sheludko, Adv. Colloid Interface Sci. 1 (1967) 391. [5] E. Hutchinson, Z. Phys. Chem. Neue Folge 21 (1959) 38. [6] R.R. Irani, C.F. Callis, J. Phys. Chem. 64 (1960) 1741. [7] M. Rutland, R.J. Pugh, Colloids Surf. A Physicochem. Eng. Aspects 125 (1997) 33. [8] P.R. Garrett, in: P.R. Garrett (Ed.), Defoaming: Theory and Industrial Application, Dekker, New York, 1993, chap. 1. [9] R. Raghavachari, K.S. Narayan, N. Nayaar, in: P.R. Garrett (Ed.), Abstracts of Eurofoam Conference, 2002. [10] D. Monin, A. Espert, A. Colin, Langmuir 16 (2000) 3873. [11] G.J. Hirasaki, J.B. Lawson, SPE Reservoir Eng. (March 1986) 119. [12] K. Theander, R.J. Pugh, J. Colloid Interface Sci. 239 (2001) 209.