Langmuir 2007, 23, 7963-7971
7963
Crystallization of Silver Carboxylates from Sodium Carboxylate Mixtures Jingshan Dong,*,† David R. Whitcomb,‡ Alon V. McCormick,† and H. Ted Davis† Department of Chemical Engineering and Materials Science, UniVersity of Minnesota, 421 Washington AVenue SE, Minneapolis, Minnesota 55455, and Eastman Kodak Company, 1 Imation Way, Oakdale, Minnesota 55128 ReceiVed NoVember 13, 2006. In Final Form: February 28, 2007 Silver carboxylates can be made by the reaction of silver nitrate and the corresponding sodium carboxylates. The length of the alkyl chain has a significant impact on the product behavior. In this study, 18, 20, and 22 carbon chains (stearate, arachidate, and behenate, respectively) have been selected. All three sodium carboxylates are very insoluble in water at room temperature. Solutions are obtained above the Krafft temperature, which precipitates lamellar crystals if cooled at the proper cooling rate. Depending on the chain length, metastable morphologies, such as vesicles and tiny fibers, can be seen consecutively before hexagonal plates form. The carboxylate with the shorter chain length reaches equilibrium more quickly. All three silver carboxylates also take on a lamellar structure. Small-angle X-ray scattering (SAXS) shows that the d spacing of the crystals increases as the chain length increases. Cryo-TEM illustrates that the crystallites are the result of micelle nucleation and micelle aggregation. In addition, the crystallization process in the presence of silver bromide nanocrystals has been investigated. In the initial stage, an epitaxial interface is formed between the silver carboxylate crystallites and the cubic silver bromide grains. Budlike and strandlike structures grow because of it. The consequent strand enclosure restrains the crystal growth, which reduces the size and changes the morphology of the crystals.
Introduction Previously, we studied the crystallization process of silver stearate (AgST) from a sodium stearate (NaST) dispersion.1 The reaction stages from 30 s to 10 min were probed by using cryogenic transmission electron microscopy (Cryo-TEM). Results show that the reaction of the NaST dispersion with silver nitrate (AgNO3) solution is a diffusion-controlled process. In addition, the AgST crystals nucleate from spherical micelles, and micellar aggregates grow into micrometer-sized platelike lamellar structures within minutes. In other words, colloidal particles, in this case, AgST micelles and micelle aggregates, can be the intermediate structures for the formation of nano- or microcrystallites. Compared to the present system that forms crystals from the organic colloids, the existence of inorganic colloidal intermediates during the growth of inorganic nanocrystals has been more widely reported. For example, Davis et al. studied the aggregative growth of zeolite nanocrystals from precursor nanoparticles experimentally as well as by computer simulation.2 Penn summarized the mechanism of oriented aggregation and pointed out that the self-assembly of primary nanoparticles can result in new single crystals, twins, and intergrowths during the growth of oxide, oxyhydroxide, hydroxide, selenide, and sulfide nanocrystals.3 Our findings not only add to the fundamental understanding of the kinetics of crystallization processes in some aqueous systems but also are valuable in the photographic film industry, which has used silver carboxylates (also called soaps) in * Corresponding author. E-mail:
[email protected]. † University of Minnesota. ‡ Eastman Kodak Company. (1) Lin, B.; Dong, J.; Whitcomb, D. R.; McCormick, A. V.; Davis, H. T. Langmuir 2004, 20, 9069-9074. (2) Davis, T. M.; Drews, T. O.; Ramanan, H.; He, C.; Dong, J.; Schnablegger, H.; Katsoulakis, M. A.; Kokkoli, E.; Mccormick, A. V.; Penn, R. L.; Tsapatsis, M. Nat. Mater. 2006, 5, 400-408. (3) Penn, R. L. J. Phys. Chem. B 2004, 108, 12709-12712.
photothermographic films for X-ray diagnostic applications.4 In such thermally developable imaging films, silver carboxylate crystals such as AgST are reduced to Ag nanoparticles, which are catalyzed at the latent image sites of the exposed silver bromide (AgBr) nanocrystals.4 Motivated by this application, the influence of AgBr on the crystallization process of AgST was also investigated in our previous study.5 The addition of cubic-shaped 50 nm AgBr single-crystalline grains causes the AgST spherical micelles to deposit onto AgBr surfaces rather than form aggregates of their own. Budlike and strandlike structures then grow from the expitaxial interface that is created. The subsequent encasing of the long strands produces lamellar crystals, similar to the growth of the silver carboxylates in the absence of AgBr, but they are significantly smaller in size. In the present article, the influence of chain length variation and the consequent morphology changes of sodium carboxylates on their reaction with AgNO3 were studied. Sodium carboxylates with 18, 20, and 22 carbons, namely, stearate, arachidate, and behenate, respectively, as well as myristate (14C), were selected as model molecules (Figure 1). The same experimental approaches as in our previous work were applied, and it was found that although arachidate and behenate sodium salts have different morphologies than sodium stearate their reactions with AgNO3 proceed similarly. A nearly equal molar mixture of stearate, arachidate, and behenate was also investigated, and similar results were found as well, which means that the small differences in the chain length and morphology of the reactants may not be important factors in the crystallization of silver carboxylates. The only significant consequence of the chain-length mixture seems to be the degradation of crystallinity. (4) Whitcomb, D. R. Kirk Othmer Encyclopedia of Chemical Technology; http://www3.interscience.wiley.com/cgi-bin/mrwhome/104554789/HOME. (5) Dong, J.; Whitcomb, D. R.; McCormick, A. V.; Davis, H. T. Nanotechnology 2005, 16, S592-S600.
10.1021/la063321i CCC: $37.00 © 2007 American Chemical Society Published on Web 06/14/2007
7964 Langmuir, Vol. 23, No. 15, 2007
Dong et al.
Figure 1. Schematic diagram of the molecular structures of four metallic carboxylates.
Experimental Section A. Materials. Myristic, stearic, arachidic, and behenic acid powders were mixed with sodium hydroxide (HMY, HST, HAR, HBE, NaOH, Fisher Scientific Co., Fair Lawn, NJ) and water, in which NaOH is 1% lower in molar amount than the acids in order to avoid the reaction of excess alkali with AgNO3 that will be added later to produce silver carboxylate. The carboxylate acids, sodium hydroxide, and the water mixture were heated in a water bath at about 95 °C (above the Krafft temperature of the sodium carboxylates) and cooled to ambient temperature by turning off the heat so that dispersions of sodium carboxylate crystallites could be obtained. The corresponding silver carboxylates were produced by adding an excess molar amount of 1 wt % AgNO3 (Allied Chemical Co., NJ) solution. Cubic-shaped AgBr grains were provided by Eastman Kodak Company as a gel and diluted to a 0.2 wt % dispersion. All operations involving AgBr were done in a red-light-illuminated room as quickly as possible to minimize any light exposure. B. Characterization Methodologies. Digital light microscopy (DLM) and cryogenic transmission electron microscopy (CryoTEM)6-8 experiments follow similar procedures described previously.1,5 SAXS was performed by using an Antar SAXcess smalland wide-angle X-ray diffractometer (Anton Paar Co., Graz, Austria). The X-ray source is Cu KR radiation (wavelength λ ) 0.1542 nm). The scattering patterns were analyzed by a PDS software package (Anton Paar Co., Graz, Austria).
Results and Discussions A. Morphology of Sodium Carboxylate. As reported by Davis and McCormick’s groups,9,10 NaST lamellar crystals in water have aging effects in which not only the d spacing between layers but also their morphology change with time. As shown in Figure 2, the fresh 1 wt % NaST dispersion, which is cooled to room temperature from above its Krafft temperature, mainly has fiberlike crystals as the solid phase. After 2 weeks, roughly half of the solid phase changes into ribbonlike crystals. If the aging time is long enough (e.g., 6 months), then the fiberlike crystals will completely disappear, and the platelike shape is the only morphology that can be observed by a light microscope. (6) Bellare, J. R.; Davis, H. T.; Scriven, L. E.; Talmon, Y. J. Electron Microsc. Tech. 1988, 10, 87-111. (7) Danino, D.; Talmon, Y. Cryo-Transmission Electron Microscopy. In Physical Chemistry of Biological Interfaces; Baszkin, A., Norde, W., Eds. Marcel Dekker: New York, 2000. (8) Talmon, Y. Cryogenic Temperature Transmission Electron Microscopy in the Study of Surfactant Systems. In Modern Characterization Methods of Surfactant Systems; Binks, B. R., Ed. Marcel Dekker: New York, 1999. (9) Dong, J. Cryo-TEM of Morphology and Kinetics of Self-assembled Nanostructures. University of Minnesota, Twin City, 2006. (10) Liang, J. Phase Behavior, Morphology and Polymorphism of Surfactant Systems. Ph.D. Dissertation, University of Minnesota, Twin City, MN 2001.
Figure 2. DLM of 1 wt % NaST; scale bar 25 µm: (a) freshly prepared, (b) after 2 weeks, and (c) after 6 months.
A similar aging effect is also observed in the dispersions of the other sodium carboxylates studied currently (myristate, arachidate, and behenate). Figure 3 gives some light micrographs of a 0.2 wt % sodium myristate (NaMY) dispersion. A freshly prepared sample is full of fiberlike crystals (Figure 3a), which look identical to that of the fresh NaST dispersion. Nearly all of the fibers turn into platelike structures within a week (Figure 3b). They age faster than those in the NaST dispersion where many fibers still exist after 2 weeks (Figure 2). In other words, the morphology change of NaMY crystals with time is faster than that of NaST. The aging phenomenon of a sodium arachidate (NaAR) dispersion is shown in Figure 4. A fresh 0.2 wt % NaAR dispersion has two kinds of condensed-phase morphologies (Figure 4a). The first kind of structure is the circular particles that are suspended in water. Particle drifting is due to water convection accelerated by heat that comes from the light microscope, which also makes imaging of the particles difficult. The particle size varies greatly, from a few micrometers in diameter (black arrow in Figure 4a) to more than 10 µm (white arrow). The second morphology is short fibers (upper part of Figure 4a), which are highly entangled with each other and appear as white flocculi to the naked eye. The same sample was imaged by the light
Crystallization of SilVer Carboxylates
Figure 3. DLM of 0.2 wt % NaMY; scale bar 25 µm: (a) freshly prepared and (b) 1 week later.
Figure 4. Morphology of 0.2 wt % NaAR, where all of the arrows point to typical vesicles; scale bar 25 µm: (a) freshly prepared and (b) 1 month later.
microscope 1 month later (Figure 4b). Short fibers originally less than 50 µm (Figure 4a) had grown to at least 500 µm long (with the latter length exceeding the size of the viewing region). It is noteworthy that they are very close to NaST fibers in size by comparison with Figure 2. At this time, some circular particles can occasionally still be found (black arrow in Figure 4b). More
Langmuir, Vol. 23, No. 15, 2007 7965
Figure 5. DLM of 0.2 wt % NaBE, where the dashed circle and the arrow indicate vesicles; scale bar 25 µm: (a) freshly prepared and (b) 1 month later.
importantly, several layers inside the particle are resolved, which means that it is a multiwalled vesicle (also called a liposome), which will be discussed in detail later in this article. The same experimental procedures were applied to sodium behenate (NaBE). A fresh 0.2 wt % NaBE dispersion has only vesicles or liposomes as the heterogeneous phase (Figure 5a). Some of the liposomes are larger than 30 µm, which is large enough to give the sample some degree of turbidity. Because of the turbidity, this sample should be classified as a dispersion rather than a solution. Unlike the NaMY, NaST, and NaAR dispersions under similar conditions, no fiberlike crystals are found in this NaBE dispersion. However, a considerable number of short fibers appear after 1 month of aging at room temperature, and they are highly entangled (Figure 5b), which selectively gives a region dominated by fibers. Care must be exercised in the interpretation of this image because large portions of the sample are still full of vesicles and liposomes. The liposome indicated by the black arrow is merely to show the coexistence of the two morphologies. Inside the dashed circle of Figure 5a, there is a liposome that is broken, which may provide some insight into the mechanism of how the morphology changes. Fibers probably can grow from vesicles by breaking the cell walls and repacking them parallel to each other to form small lamellae, whose appearance under a light microscope can be fiberous. More insight into this process will be discussed later. By comparing the aging effect of the NaAR and NaBE dispersion, it can be found that the NaBE dispersion ages more slowly than the NaAR dispersion during the morphology change of liposomes to fibers because 1 month is sufficient for NaAR to convert almost all of its liposomes (Figure 4), whereas for NaBE many liposomes still remain after 1 month (Figure 5). As noted above, for the fiber-to-plate aging sequence of the NaMY and NaST dispersion (Figures 2 and 3), NaST is slower than NaMY. Therefore, we can conclude that the shorter the
7966 Langmuir, Vol. 23, No. 15, 2007
Dong et al.
Figure 7. SAXS of three 1 wt % sodium carboxylates and their mixture.
Figure 6. DLM of a 3NaST/3NaAR/4NaBE 0.2 wt % dispersion; scale bar 25 µm: (a) freshly prepared sample and (b) 1 month later.
hydrocarbon chain, the faster its crystals in water will change its morphology under aging. Therefore, the aging sequence for the morphology of the sodium carboxylate crystals in water is liposome to fiber to plate. The ribbonlike structure appears to be an intermediate state between the fiber and plate. A 0.2 wt % dispersion of a NaST, NaAR, and NaBE mixture in about a 3:3:4 molar ratio was prepared and observed with DLM as well. It behaves very similarly to the pure NaAR 0.2 wt % dispersion except that fibers of the mixture seem to be shorter than those of the NaAR dispersion (Figure 6). A freshly prepared dispersion of the mixture has both vesicles and fibers (Figure 6a), and most of the crystals take the shape of fibers after a month (Figure 6b). de Mul et al.11 studied each of three possible two-soap mixtures from the combination of the three sodium carboxylates and found no new structures other than lamellar with slightly different d spacing. McBain noticed that the coexistence of homologs in commercial soap might not change the phase behavior qualitatively compared to that of a pure soap in water.12,13 Regardless of their morphologies, all of these sodium carboxylate dispersions give lamellar peaks in a small-angle X-ray diffractometer (Figure 7). The presence of a large amount of water causes the patterns to have a poor signal-to-noise (S/N) ratio, even when the concentration of the dispersions is raised from 0.2 to 1 wt % in order to enhance diffraction. The (001) peak for the individual sodium carboxylate and their mixture can be identified, although it is weak. The d spacings are calculated accordingly and are estimated to be 4.5, 4.9, 5.9, and 5.7 nm, respectively, for 1 wt % NaST, NaAR, NaBE, and 3NaST/3NaAR/ (11) de Mul, M. N. G.; Davis, H. T.; Evans, D. F.; Bhave, A. V.; Wagner, J. R. Langmuir 2000, 16, 8276-8284. (12) Laughlin, R. G. The Aqueous Phase BehaVior of Surfactants; Academic Press: London, 1994. (13) McBain, J. W. Colloid Chemistry; The Chemical Catalog Co.: New York 1926; Vol. I.
4NaBE. Such d-spacing values can come only from lamellar structures for this type of material regardless of whether the layers are in liposomes or fibers or plates. Errors may exist for the d-spacing values because of the low S/N ratio, but the relative position of the (001) peaks of the four systems shows that longer chain length causes larger d spacing. Most importantly, the nearly equal molar mixture of the three sodium carboxylates has just one set of d spacing, which means that the three chains coexist in a single type of lamellar structure rather than form crystals of their own lattice sizes. In addition, the d spacing between the adjacent layers of the mixed crystal is closer to that of NaBE, presumably because the longer species (i.e., NaBE) occupies more of the space of the hydrocarbon chain layers and everything else fits within it. B. Vesicles in the NaST Dispersion. According to the vesiclefiber-plate morphology aging effect for the sodium carboxylates noted above, the NaST dispersion should demonstrate vesicular morphology under some circumstances. However, a fibrous or ribbonlike structure is the commonly reported morphology at room temperature.14,15 Our previous work showed that it has fibrous as well as platelike morphologies.1,5,9 Then the question is, can we obtain the vesicle morphology? The answer comes from a thermodynamic consideration of the morphology change upon aging. Because vesicles age into fibers, vesicular structures should have a high-energy state compared to that of a fiber. By analogy to the quenching of metals, where some structures at high temperature can be preserved if they are cooled down quickly,16 “quenching” the NaST dispersion may also be able to capture its high-temperature structure, which could be a metastable structure such as a vesicle.17 According to the sample preparation procedure of Liang et al.,14,15 they dissolved NaST raw material at 80 °C in water and then cooled it down to 25 °C at a rate of 0.5 °C/min. In our case,1,5 the cooling rate was not controlled precisely, but it was about 0.4 °C/min near the NaST Krafft temperature (68 °C).11,18 Such a cooling rate can be easily achieved by shutting down the heater of the water bath, whereas cooling the dissolved solution in air is 4 to 5 times faster. Indeed, the (14) Liang, J.; Ma, Y.; Chen, B.; Munson, E. J.; Davis, H. T.; Binder, D.; Chang, H.; Abbas, S.; Hsu, F. J. Phys. Chem. B 2001, 105, 96539662. (15) Liang, J.; Ma, Y.; Zheng, Y.; Davis, H. T. Langmuir 2001, 17, 64476454. (16) Callister, W. D. Materials Science and Engineering: An Introduction, 5th ed.; John Wiley & Sons: New York, 2000. (17) Heppenstall-Butler, M.; Butler, M. F. Langmuir 2003, 19, 10061-10072. (18) McBain, J. W.; Vold, R. D.; Frick, M. J. Phys. Chem. 1940, 44, 10131024.
Crystallization of SilVer Carboxylates
Figure 8. Cryo-TEM of a 0.2 wt % NaST dispersion cooled from 80 °C to room temperature in air. The arrow points out a hemisphere of a vesicle.
vesicle metastable structure is prevented from becoming the more stable fiber structure by cooling the 0.2 wt % NaST dispersion in air. Figure 8a, a cryo-TEM image of such an air-cooled sample, shows some single-walled vesicles (that resemble circles) and some double-walled vesicles (two concentric circles). Each wall is a bilayer in which the hydrophobic carbon chains are covered inside and the hydrophilic carboxylate head groups pack side to side, forming an interface that contacts with water. From the upper left corner down to the right of Figure 8a, there is a ribbonlike structure that represents one of the small fiberlike crystals demonstrated in the previous light micrographs. Figure 8b is a region from the same specimen, but it has only vesicles. The white arrow points out a vesicle, half of which is missing. On closer inspection, the “cup” edge is resolved, which demonstrates the 3D nature of vesicles, even though they usually resemble circles because TEM micrographs provide only a 2D projection. In both images of Figure 8, most of the vesicles are slightly deformed because they are soft “balloons” that are either
Langmuir, Vol. 23, No. 15, 2007 7967
Figure 9. Cryo-TEM of 0.2 wt % NaBE dispersion, freshly prepared. The two arrows mark the boundary of two contact vesicles.
squeezed because of the tiny thickness of the vitrified solvent film or pulled by the stress during cryo-TEM specimen preparation. C. NaBE and Its Reaction with AgNO3. Vesicles of NaBE are more stable or have longer lifetime than those of NaST. If the NaBE isotropic solution above its Krafft temperature is cooled to room temperature in the water bath by shutting off the heat, then nearly 100% of the separated phase, the particles, are vesicles (Figure 9a). Too many vesicles make them collide with each other easily, and sometimes the intervesicle attraction is so large that their walls are fused, as are the two in Figure 9b. The boundary where the fusion occurs has three layers, and the distances between two neighboring layers are exactly the same as the d spacing resolved by the previous X-ray diffraction pattern in Figure 7. In fact, such an appearance is what a lamellar structure looks like under cryo-TEM if its layers happen to be parallel to the electron beam.15 Therefore, cryo-TEM gives us not only a direct way to “see” the structure but also an alternative method to characterize its crystallographic parameters. For instance, Liang et al. captured
7968 Langmuir, Vol. 23, No. 15, 2007
Figure 10. Cryo-TEM of NaBE + AgNO3 after about 30 s of reaction time. The dashed ellipsoid marks a nonvitrified region (ice) of the water film. The arrow points out a AgBE micelle.
the multilayers of a NaMY lamellar crystal by using cryo-TEM,15 and claimed that the distance between the adjacent layers does not change if the crystal is tilted. If the total distance of several layers is measured from the cryo-TEM image and then averaged by the layer number, then the lamellar d spacing can be estimated. More importantly, this can be done for very dilute liquid systems whose SAXS patterns are usually very noisy (for example, 1 wt % NaST in Figure 7). To extrapolate this vesicle-vesicle fusion phenomenon to the extreme, if more and more vesicles touch each other and they are extremely compressed so that most of the water inside each vesicle is squeezed out, then the final formation would be a multilayer ribbon (i.e., a lamellar crystal). Perhaps this is what is happening during the vesicle-to-fiber transition noted above for the aging of the sodium carboxylates. It is a simultaneous process during which the system energy, such as interfacial energy, is reduced. Consequently, for these sodium carboxylate systems, the vesicle is not a phase because
Dong et al.
it is not thermodynamically stable, similar to some other surfactant systems reported by other researchers (reviews in refs 19 and 20). When the 0.2 wt % NaBE dispersion is mixed with a 1 wt % AgNO3 solution, then silver behenate (AgBE) precipitates and crystallizes in a short time. Ag+ is present in excess to ensure that all of the sodium behenate is converted. The experimental procedure is similar to that reported previously for the NaST + AgNO3 reaction.1,5 Figure 10 comes from a cryo-TEM observation of a specimen after 30 s of on-the-grid mixing. Similar to the initial stage of silver stearate (AgST) crystallization, AgBE crystallization also starts from spherical micelles that form micellar aggregates. The AgBE spherical micelles have faint contrast in Figure 10a, possibly because this film is not completely amorphous, which is illustrated by the nonuniform background caused by ice formation of the water film. The polycrystalline ice from failed vitrification generates a diffraction contrast so that some regions, such as the one inside the dashed ellipsoid, appear darker than other areas. However, four AgBE micelle aggregates still can be easily identified, and a number of micelles (about 5 nm diameter, e.g., the white arrow) are distributed nearly everywhere. The reaction can be simply viewed as a process where the Na+ of NaBE is replaced by Ag+. It is likely that it is also a reaction whose rate is governed by the Ag+ diffusion, the same as the reaction of NaST and Ag+.1 The morphology of the NaBE reactant is that of vesicles, which concentrates the majority of all of the available NaBE molecules. The ion exchange in the early stage of ion exchange may still retain the overall shape of the original vesicles, as seen in Figure 10b. This phenomenon also means that the vesicle surfaces are good nucleation sites for the NaBE micelle aggregation. Through molecular rearrangement and Oswald ripening, the aggregated spherical NaBE micelles will eventually grow into micrometer-sized, platelike, lamellar crystals; images are not shown here but are very similar to that of NaST reported previously.1 D. AgBE Crystallization in the Presence of AgBr. Cubicshaped AgBr single-crystalline grains that are dispersed by trace amounts of gelatin are added to the NaBE + AgNO3 reaction system. The results are very similar to what occurs in the NaST + AgNO3 system with AgBr nanocrystals present.5 With time, a transition from bud to strand to “encased ring” is also observed (Figure 11). After 1 min of reaction, nearly all of the AgBr grains have a bud projecting from them (Figure 11a). Although it is difficult to tell whether the budlike structure is growing from the corner of AgBr cubes because of their etching, the buds are believed to grow on the edges of the AgBr grains, which are {111}-type lattice planes.5 The buds grow longer with time (i.e., become strandlike structures), and multiple buds associated with a single AgBr grain appear when more and more AgBE molecules are adhered (Figure 11b). After about 5 min, many long buds are available, and some of them connect with each other to enclose many other AgBE molecules. The crystallinity of the final lamellar structures is developing, especially in some strands (dashed circle in Figure 11c), but also can be found elsewhere (dashed square), where the layers of the lamellar structure are visible. E. NaST/NaAR/NaBE Mixture and Its Reaction with AgNO3. In any general mixture of chain lengths, it is not likely that the carboxylates demonstrate only one type of chain length. To investigate the influence of possible polydispersity, a mixture of NaST, NaAR, and NaBE with about a 3:3:4 molar ratio21 was (19) Chernik, G. G. Curr. Opin. Colloid Interface Sci. 2000, 4, 381-390. (20) Gradzielski, M. Curr. Opin. Colloid Interface Sci. 2003, 8, 337-345. (21) Yoshioka, Y.; Ishizuka, T.; Yanagi, T. EP 1283440; 2003.
Crystallization of SilVer Carboxylates
Langmuir, Vol. 23, No. 15, 2007 7969
Figure 12. dispersion.
Cryo-TEM of a 3NaST/3NaAR/4NaBE mixed
Figure 13. Cryo-TEM of the reaction between 3NaST/3NaAR/ 4NaBE and AgNO3. The dashed circle may be an aggregate in the process of forming.
Figure 11. Cryo-TEM of the NaBE + AgNO3 reaction in the presence of AgBr grains, where the dashed circle and square show AgBE lamellar crystals. Reaction times of (a) ∼1, (b) ∼3, and (c) ∼5 min.
prepared, and the reaction between AgNO3 with the mixedchain-length sample was performed. According to Figure 12, the possible morphologies of the sodium carboxylate mixture have fibers and vesicles occurring simultaneously. When the mixed dispersion reacts with AgNO3, spherical micelles and micellar aggregates form at about 30 s of reaction time (Figure 13), which once again is similar to the case of pure NaST participating in the reaction. The soap dispersion in Figure 13 is deliberately diluted to 0.05 wt %, unlike the usual concentration of 0.2 wt % in the previous study. The purpose of dilution is to reveal the blank regions that would be covered by overlapping in the case of too many silver carboxylate micelles being generated. The blank region most likely is a depletion zone that results from the micelle aggregation process. However,
7970 Langmuir, Vol. 23, No. 15, 2007
Dong et al.
Figure 15. SAXS of silver carboxylates filtered from the reaction products of sodium carboxylate dispersions plus AgNO3 solutions.
Figure 14. Cryo-TEM of the 3NaST/3NaAR/4NaBE + AgNO3 reaction in the presence of AgBr, with the arrows pointing out two silver carboxylate crystals. Reaction times of (a) ∼30 s and (b) 10 min and (c) an overview of the 10 min reaction.
every zone of a micelle “cloud” is a potential place from which an aggregate can be produced, for example, in the upper corner of Figure 13, where such an aggregate is growing. Interestingly, it is attached to the carbon grid, a possible nucleation site in the on-the-grid reaction condition.1 Compared to the size of the aggregates in Figure 10, the size of these micellar aggregates is dramatically smaller, which might be attributed to the dilution as well. F. AgST/AgAR/AgBE Mixture Crystallization in the Presence of AgBr. In the presence of AgBr single-crystalline grains, the formation of micelle aggregates is bypassed because many preferred nucleation sites are provided by these grains. In the initial stage of reaction, the silver carboxylate micelles coat the surfaces of the AgBr grains. Consequently, these grains wear a dark shell and have some sporadic tiny “buds” (Figure 14a). In less than 3 min, these tiny molecular clusters concentrate on the corners of each AgBr grain (not shown here), just as in the case of AgBE micelles aggregating and growing into budlike structures on the AgBr cubic corners (Figure 11a), which is also similar to the case of pure NaST.5 Following the formation of the budlike structures is a period in which strandlike structures grow and encase them. A finished crystal is shown in Figure 14b, where the strands are nearly closed and the interior is almost filled with silver carboxylate molecules. Cryo-TEM of the 30 min reaction time was done (images not shown here), but statistically there is no detectable change in the crystal size, which means that such a closed structure is the end of the crystal growth process. Figure 14c illustrates an overview of the final silver carboxylate crystals, most of which are on the order of 1 µm. Compared to the reaction of pure NaST with AgNO3 without AgBr, those final crystals can easily grow to 10 µm.1 Therefore, the variation of chain length and the presence of AgBr grains greatly reduce the crystal size of final silver carboxylates. In addition, it should be noted that the silver carboxylate crystals no longer have a hexagonal shape, unlike the AgST produced from pure NaST.1 The impact of chain-length polydispersity and the presence of AgBr on the perfection of the silver carboxylate crystals is also demonstrated by X-ray diffraction. Figure 15 compares the SAXS patterns of the three pure silver carboxylate crystals, their mixture, and their mixture in the presence of AgBr grains that are made from the respective sodium counterparts. To avoid the poor S/N ratio caused by dilution (such as in Figure 7), the current materials were filtered and dried in air. Three of the four distinct lamellar peaks, namely, (001), (002), (003), and (004), can be resolved for the pure silver carboxylates. The mixture shows the first three peaks, and (004) peak is missing, which means that the mixed crystal is not as perfect as any of the pure
Crystallization of SilVer Carboxylates
crystals. The diffraction pattern from the mixture produced in the presence of the AgBr grain barely has one weak peak, which indicates the worst crystallinity. In addition, d spacings of all of the lamellar crystals are calculated on the basis of the q values obtained from the Figure, which are 4.90 nm for AgST, 5.39 nm for AgAR, and 5.89 nm for AgBE (similar to that reported by others,22 4.854, 5.348, and 5.843 nm, respectively). The presence of AgBr does not change the structure except to cause poor crystallinity.
Conclusions NaST, NaAR, and NaBE dispersions, and their nearly equal molar mixture are examined by light microscopy cryo-TEM and X-ray diffraction. Vesicular and fibrous morphologies can be obtained from all of the dispersions when cooled from above the Krafft temperature at the right cooling rates. The vesicle is not a thermodynamically stable structure and can turn into a lamellar fiber after aging at ambient temperature. In general, the longer the hydrocarbon chain length, the more slowly this transition occurs. As a kinetic model for this transition, the fiber could be considered to form from the breaking and repacking of the walls of the vesicle or liposome. (22) Binnemans, K.; Deun, R. V.; Thijs, B.; Vanwelkenhuysen, I.; Geuens, I. Chem. Mater. 2004, 16, 2021-2027.
Langmuir, Vol. 23, No. 15, 2007 7971
Reactions of these sodium carboxylates with Ag+ were also studied by cryo-TEM and X-ray diffraction. The production of silver carboxylates is a crystallization process that starts from silver soap micelle and micellar aggregation. AgBr nanocrystals added to the reaction system changes the crystallization process by bypassing the formation of micelle aggregates. Instead, silver carboxylate budlike structures form, which is followed by their growth and connection, resulting in the final silver soap crystals. Using mixtures of chain lengths and including the presence of AgBr in the reactions not only reduce the silver carboxylate crystal size but also change the crystal shape. The SAXS data from the silver soaps prepared from a mixture of chain lengths also showed a d-layer spacing distance in between the spacings of the individual silver carboxylates. Acknowledgment. We acknowledge financial support from Eastman Kodak Company and the Industrial Partnership for Research in Interfacial and Materials Engineering (IPRIME) at the University of Minnesota. The instrument support from the characterization facility center at the University of Minnesota is greatly appreciated, and we thank Dr. N. C. Howlader (Eastman Kodak Company) for providing the AgBr samples. LA063321I
