pubs.acs.org/Langmuir © 2010 American Chemical Society
Influence of Chain Lengths and Loading Levels on Interlayer Configurations of Intercalated Alkylammonium and Their Transitions in Rectorite Zhaohui Li,*,†,‡,§ Wei-Teh Jiang,‡ Chun-Jung Chen,‡ and Hanlie Hong§ Geosciences Department, University of Wisconsin - Parkside, Kenosha, Wisconsin 53141-2000, ‡Department of Earth Sciences, National Cheng Kung University, 1 University Road, Tainan 70101, Taiwan, and §Faculty of Earth Science, China University of Geosciences, Wuhan, Hubei, China 430074
†
Received December 14, 2009. Revised Manuscript Received March 12, 2010 There have been extensive studies on intercalation of alkylammonium into swelling clay minerals for the purpose of surface charge determination of the clay minerals, as well as their interlayer configurations in the clay minerals. The most accepted findings are that the intercalated alkylammonium molecules adopted horizontal monolayer, bilayer, pseudotrilayer, and vertical paraffin-like configurations in the interlayer space of the swelling clay minerals depending on the chain length and loading level of the alkylammonium used. This study examined the interlayer configurations of intercalated alkyltrimethylammonium and their transition structure as a function of alkylammonium inputs and chain lengths. As the amount of alkylammonium intercalation increased, the bilayer configuration of the intercalated alkylammonium was absent during a transition from a horizontal monolayer to a pseudotrilayer intercalation. Instead, the transition structure involved a mixed layer made of rectorite intercalated with one layer and rectorite intercalated with pseudotrilayer of alkylammonium molecules. When intercalated in horizontal monolayer, the alkylammonium molecules took a random, more gauche-like arrangement. On the other hand, as alkylammonium molecules adopted a pseudotrilayer, particularly the vertical paraffin-like arrangement, a more ordered all-trans configuration was achieved. As layer charge is one of the most important properties of swelling clay minerals, commonly determined by intercalation of n-alkylammonium ions, the identification of mixed-layer transition structure in this study may suggest a need for further investigations on principles of layer charge determination of swelling clays by the alkylammonium method.
Introduction Layer charge is one of the most important properties in studying the physiochemical behaviors of clay minerals in industrial or environmental processes. There have been many studies focusing on the intercalation of alkylammonium into smectite for the purpose of layer charge determination.1-4 The basal spacing of n-alkylammonium montmorillonites changed in a characteristic way depending on the length of the n-alkylarnmonium ion.1 When the number of carbon on the tail group was seven or less, a flat-lying monolayer at a constant d-spacing of 1.34-1.36 nm was observed. With a chain length between 8 and 14 carbons, a bilayer with a d-spacing of 1.77 nm were formed, while a pseudotrimolecular structure was proposed based on the d-spacing of 2.15 nm when the chain length was 15-20 carbons.1 Recent studies focused on the interlayer configurations of intercalated alkylammonium in swelling clay minerals with initial inputs greater or much greater than the cation exchange capacity *Corresponding author. Tel: 1-262-595-2487, Fax: 1-262-595-2056, E-mail:
[email protected]. (1) Lagaly, G.; Gonzalez, M. F.; Weiss, A. Clay Miner. 1976, 11, 173–187. (2) Yoshida, T.; Suito, E. J. Appl. Crystallogr. 1972, 5, 119–124. (3) Maes, A.; Stul, M. S.; Cremers, A. Clays Clay Miner. 1979, 27, 387–392. (4) Lagaly, G. Clay Miner. 1981, 16, 1–21. (5) Zhou, Q.; Frost, R. L.; He, H.; Xi, Y. J. Colloid Interface Sci. 2007, 307, 50–55. (6) Xi, Y.; Ding, Z.; He, H.; Frost, R. L. J. Colloid Interface Sci. 2004, 277, 116– 120. (7) Xi, Y.; Frost, R. L.; He, H. J. Colloid Interface Sci. 2007, 305, 150–158. (8) Xi, Y.; Mallavarapu, M.; Naidu, R. Appl. Clay Sci. 2010, in press. doi:10.1016/j.clay.2009.11.047. (9) Xu, L.; Zhu, L. J. Colloid Interface Sci. 2009, 331, 8–14. (10) Xi, Y.; Martens, W.; He, H; Frost, R. L. J. Therm. Anal. Calorim. 2005, 81, 91–97. (11) Feng, Y.; Hu, G.; Meng, X.; Ding, Y.; Zhang, S.; Yang, M. Appl. Clay Sci. 2009, 45, 239–243.
Langmuir 2010, 26(11), 8289–8294
(CEC) of the minerals.5-13 At the octadecyltrimethylammonium (ODTMA) loading levels of 2.0 and 4.0 times the CEC, a reflection at 2.03 nm was observed for montmorillonite after intercalation.6,7 A more recent study showed reflections at 3.57 and 1.85 nm after a bentonite was treated with hexadecyltrimethylammonium (HDTMA) at 4.0 CEC.8 An even larger reflection at 4.16 nm was found for a bentonite intercalated with ODTMA at a loading level of 1.4 CEC.9 However, the d-spacing determined in these studies was based on only one or two loworder reflections.6-9 In addition, a somehow continuous increase in d-spacing from 1.45 to 2.3 nm was noticed when a Namontmorillonite SWy-2 was modified by ODTMA from 0.2 to 4.0 CEC and was attributed to monolayer, bilayer, and pseudotrilayer intercalation.10 With limited numbers of basal reflections, it would be impossible to determine whether the d-spacing calculated from the (001) reflection was indeed rational (an integral multiple) of the d-values derived from other (00l) reflections. Reflections at 4.24, 2.01, 1.34, and 0.99 nm were observed when a bentonite was modified by HDTMA with an initial amount of 2.5 CEC, and these reflections were indexed as (001), (002), (003), and (004) reflections.11 On the contrary, another bentonite modified by HDTMA with an initial amount of 3.0 CEC only resulted in a d-spacing of 2.0 nm with all other highorder reflections missing.12 The discrepancies in d-spacing of intercalated montmorillonite in the literature indicate that montmorillonite might not be a good choice as the substrate mineral for the investigation of interlayer configurations of intercalated alkylammonium. (12) Messabeb-Ouali, A. EL; Benna-Zayani, M.; Kbir-Ariguib, N.; TrabelsiAyadi, M. Phys. Procedia 2009, 2, 1031–1037. (13) Zhou, Q.; Frost, R. L.; He, H.; Xi, Y.; Zbik, M. J. Colloid Interface Sci. 2007, 311, 24–37.
Published on Web 03/23/2010
DOI: 10.1021/la904677s
8289
Article
Li et al.
Rectorite is a regularly interstratified clay mineral made of an illite component and a smectite component. In studying the layer sequence of rectorite, strong (00l) reflections with l = 1, 2, 5, 8, and 13 were observed in an early study.14 Higher-order reflections with l up to 21 were also found, except 0,0,10 and 0,0,15 for airdried sample and 0,0,11; 0,0,16; 0,0,18; and 0,0,21 for ethylene glycol (EG) saturated sample.15 Separately, (00l) reflections with l = 1-16, except l = 3 and 6 for untreated sample, and l = 1-22, except l = 11, 18, and 21 for EG-saturated sample, were also observed for rectorite.16 As basal reflections are necessary to study the interlayer configuration of intercalated organic compounds in swelling clay minerals, rectorite offers a unique opportunity compared to smectite. Intercalation of alkylammonium at certain loading levels into the swelling smectite component of rectorite resulted in an irrational (002) reflection to the (00l) sequence,17-19 which was speculated as made of mixed layers made of monolayer intercalated rectorite (MIR) and pseudotrilayer intercalated rectorite (TIR). At higher alkylammonium intercalation levels, the reflections became rational by and large and the peak width became narrower, suggesting a change from mixed-layer clays to regular alkylammonium intercalated rectorite.17-19 However, detailed X-ray diffraction (XRD) analyses as well as numerical simulations in regard to the proposed mixed-layer stage during alkylammonium intercalation into rectorite were not carried out. In this study, the influence of alkylammonium chain lengths and loading levels on rectorite was investigated by XRD and Fourier transform infrared (FTIR) analyses together with NEWMOD simulation in order to elucidate the interlayer configurations and their transition structure of the intercalated alkylammonium in rectorite.
Experimental Methods The rectorite was obtained from Zhongxiang, Hubei, China. Its clay fraction was isolated by sedimentation and used throughout the study. It has a CEC of 410 mmolc/kg20 and a surface area of 10 m2/g.19 The alkylammoniums used were HDTMA (from Aldrich), dodecyltrimethylammonium (DDTMA, from Acros), and octyltrimethylammonium (OTMA, from Fluka), all in the bromide form. Intercalation of alkylammonium into rectorite was performed in batch experiments in duplicate. To each 50 mL centrifuge tube, 0.50 or 1.00 g of rectorite and 20 mL of alkylammonium solution at initial concentrations of 5 to 50 mmol/L with a 5 mmol/L increment were combined. The initial amounts corresponded approximately to 0.25, 0.50, 0.75, 1.00, 1.25, 1.50, 1.75, 2.00, 2.25, 2.50, 3.00, 3.50, 4.00, 4.50, and 5.00 CEC of the mineral. The mixtures were shaken at 150 rpm for 24 h and centrifuged at 3000 rpm for 30 min. The supernatant was filtered through 0.45 μm filter discs before being analyzed for equilibrium alkylammonium and counterion bromide solution concentrations by HPLC methods.17-19 The amounts of alkylammonium and bromide adsorbed were determined by the difference between the initial and equilibrium concentrations. Powder XRD patterns were recorded using a Rigaku D/MaxIIIa diffractometer with Ni-filtered Cu KR radiation at 30 kV and 20 mA. Oriented samples were prepared by depositing rectorite suspension onto glass slides and drying naturally. Samples were scanned from 2° to 15° 2θ at 2°/min and 0.01° per step. The slit (14) Bradley, W. F. Am. Mineral. 1950, 35, 590–595. (15) Brindley, G. W. Am. Mineral. 1956, 41, 91–103. (16) Nishiyama, T.; Shimoda, S. Clays Clay Miner. 1981, 29, 236–240. (17) Li, Z.; Jiang, W.-T.; Hong, H. Spectrochim. Acta, Part A 2008, 71, 1525– 1534. (18) Li, Z.; Jiang, W.-T. Thermochim. Acta 2009, 483, 58–65. (19) Li, Z.; Jiang, W.-T. Clays Clay Miner. 2009, 57, 139–149. (20) Hong, H. L.; Jiang, W.-T.; Zhang, X.; Tie, L.; Li, Z. Appl. Clay Sci. 2008, 42, 292–299.
8290 DOI: 10.1021/la904677s
Figure 1. Adsorption isotherms of OTMA (a), DDTMA (b), and HDTMA (c) and counterion bromide on rectorite. The lines are Langmuir fit to the observed data. conditions were 1° for divergent and antiscattering slits and 0.3 mm for detector slit. For samples whose d001 is greater than 4.4 nm, a scan from 0.75° to 8° 2θ at 1°/min and 0.01° per step was performed on a Bruker X-ray diffractometer, utilizing Cu KR radiation at 40 kV and 40 mA with a divergent slit of 0.6 mm, an antiscattering slit of 0.6 mm, and a detector slit of 0.1 mm. FTIR spectra were acquired on a Perkin-Elmer Spectra One Spectrometer equipped with a diamond-press attenuated total reflection (ATR) accessory. The spectra were obtained by accumulating 256 scans at a resolution of 4 cm-1.
Results and Discussion Adsorption Isotherm. The adsorption of alkylammonium on rectorite was well described by the Langmuir adsorption isotherm with adsorption capacities of 290, 600, and 1200 mmol/kg for OTMA, DDTMA, and HDTMA, respectively, corresponding approximately to 0.71, 1.76, and 3.20 CEC of the mineral (Figure 1). The results are comparable to those of HDTMA adsorption on SAz-1.21 Counterion bromide adsorption was minimal throughout the initial concentration range up to 100 mmol/L. The ratio of bromide to OTMA adsorbed was 1:12, suggesting that OTMA was in a monomer form intercalated in the interlayer space of the montmorillonite component and cation exchange was the major mechanism for OTMA adsorption and intercalation, which resulted in minimal counterion balance accompanying OTMA intercalation. A similar trend was found for OMTA adsorption on illite, even though the adsorption was on the external surfaces, resulting in no interlayer expansion, thus no intercalation.22 As the intercalated OTMA was in a monomer (21) Xu, S.; Boyd, S. A. Environ. Sci. Technol. 1995, 29, 312–320. (22) Li, Z.; Alessi, D.; Zhang, P.; Bowman, R. S. J. Environ. Eng. -ASCE 2002, 128, 583–587.
Langmuir 2010, 26(11), 8289–8294
Li et al.
Article
Figure 2. XRD patterns of rectorite intercalated with different amounts of OTMA.
form, further intercalation of naphthalene or anthracene into the OTMA modified interlayer space of montmorillonite was not successful.23 Due to the lack of counterions, the OTMA modified clays were proven not good candidates to adsorb and remove anionic contaminants from water.22 At the DDTMA and HDTMA adsorption capacities, the ratio of bromide to DDTMA or HDTMA adsorbed was 1:2, suggesting a micellar form of intercalated alkylammonium.17,19 Thus, in addition to cation exchange, hydrophobic interaction also played an important role in their intercalation into rectorite.21 Due to the presence of micelles, further intercalation of naphthalene or anthracene into the interlayer space of DDTMA and HTDMA modified montmorillonite was permissible23 and retention of anionic compounds by modified clays via anion exchange between the solution anions and bromide was achievable.22,24 XRD Analyses. The d001 for OTMA intercalated rectorite was about 2.45 nm in contrast to 2.55 nm for raw rectorite (Figure 2). As rectorite is made of one illite layer and one montmorillonite layer in a regular interstratification, the intercalation of OTMA will predominantly be located in the interlayer space of the montmorillonite component. The slight decrease in d001 resulted from the substitution of calcium with two layers of water by OTMA. Peak broadening was seen only at the OTMA intercalation level of 0.23 CEC. At 0.43 CEC intercalation and beyond, the peak became narrower and more symmetric. Meanwhile, the peak location and peak intensity remained about the same regardless of the amount of OTMA intercalation. The results suggested that, when the amount of OTMA intercalated was about half of the CEC of the mineral, the fundamental interlayer structure would be stable and the interlayer was made by a horizontal monolayer of OTMA, in contrast to a predicted bilayer intercalation when the number of carbons on the tail group of n-alkylammonium was (23) Ogawa, M.; Shirai, H.; Kuroda, K.; Kato, C. Clays Clay Miner. 1992, 40, 485–490. (24) Li, Z.; Bowman, R. S. Environ. Eng. Sci. 1998, 15, 237–245.
Langmuir 2010, 26(11), 8289–8294
Figure 3. XRD patterns of rectorite intercalated with different amounts of DDTMA (solid lines) and NEWMOD simulation of the experiment data (dashed lines). Dark solid line indicates the change and split of (002) reflection with increase in DDTMA intercalation.
eight.1,4 Further increases in OTMA content only resulted in rearrangement of the intercalated OTMA into more condensed configurations. The d-spacing of basal reflections of DDTMA and HDTMA intercalated rectorite is listed in Supporting Information Table S1. At DDTMA intercalation level of 0.25 CEC, the d001 was 2.41 nm and the peak was symmetric, indicating that the DDTMA adopted a flat-lying monolayer configuration (Figure 3). In contrast to OTMA intercalation, DDTMA intercalated rectorite showed a transition at an intercalation level corresponding to 0.74 CEC, which was manifested as a broad plateau for the (002) reflection that could be considered as a composite peak made of peaks at 1.31 and 1.19 nm (Figure 3). A similar broad plateau was observed between 2.7 and 1.7 nm when a montmorillonite was modified by HDTMA to 1.5 CEC5 and between 1.83 and 1.47 nm when a montmorillonite was modified by dimethyldioctadecylammonium (DDODA) to 0.8 CEC.13 As the amount of DDTMA intercalated increased further to 1.09 CEC, the broad plateau separated into two peaks at 1.42 and 1.08 nm with an irrational distance. These two peaks further separated to 1.58 and 1.07 nm at DDTMA intercalation of 1.67 CEC. With a d001 of 3.16 nm, these two peaks became part of a nearly rational (00l) sequence and could be considered as the (002) and (003) reflections. A similar observation was found when the amount of DDODA intercalation increased from 0.7 to 1.5 CEC.13 The d-spacing of 3.16 nm could be attributed to a pseudotrilayer configuration of DDTMA in the interlayer of the montmorillonite component. Previous results also showed an irrational feature of the (00l)series, indicating a random succession of interlayer spaces with different spacings.1 The irrational feature was considered as DOI: 10.1021/la904677s
8291
Article
mixed layers of interlayer spaces made of monolayers and interlayer spaces made of bilayers of alkylammonium ions.25 However, the larger interlayer expansion at the DDTMA intercalation of 1.67 CEC in this study was closer to a pseudotrilayer than a bilayer of alkylammonium. Thus, the irrational reflections could be generated by mixed layers of two components with the first one having a flat-lying monolayer configuration with a gallery height of 0.39 nm and the second one having a flat-lying pseudotrilayer configuration with a gallery height of 1.17 nm (Supporting Information Figure S1). NEWMOD Simulation. To further support this hypothesis, NEWMOD26 was used to simulate the experimental results. The NEWMOD program takes consideration of the probabilities and proportions of two structurally or compositionally different types of layers and uses a mathematical-physical approach to simulate XRD patterns for mixed-layer clay minerals. Detailed theoretical and mathematical treatments can be found in the literature.27,28 The method has been demonstrated to have many successful practical applications, particularly for the identification of the layer types, proportions, and ordering state of mixed layers in the last 40 years.29 Peak positions are crucial for confirmation and determination of the ordering state, and proportions of endmember layers within mixed layers and peak intensities are related to the precise configuration of atoms in the unit cell. In this study, a version of NEWMOD for Windows26 was used and the simulation was based on a goniometer radius of 18.5 cm, two slits with 5.0 and 2.5 cm widths, a sample length of 3.0 cm, crystal thicknesses of 3-14 layers, and experimentally measured unit layer thicknesses of intercalated rectorite, but detailed consideration of the molecule positions of intercalated compounds was excluded as no reliable data and references could be made available. The simulated relative intensity variations appear to be reasonably consistent with the measured results as described below, however. The thickness of MIR, TIR, and paraffin-like intercalated rectorite (PIR) was set at 2.42, 3.17, and 4.95 nm, respectively (Supporting Information Figures S1 and S2). The index of ordering R used was set at 0, 0.5, and 1.0. Only the results with R = 0.5 were plotted. At the DDTMA intercalation level of 0.25 CEC, a simulation using 90% of MIR and 10% of TIR fits the experimental data well (Figure 3). Similarly, at 0.74 CEC intercalation of DDTMA, the mixed layer could be considered as that made of 70% of MIR and 30% of TIR (Figure 3). At DDTMA intercalation level of 1.09 CEC, mixed layers made of 40% MIR and 60% TIR matched the experimental data. Finally, At DDTMA intercalation level of 1.67 CEC, the mixed layers could be considered made of 10% MIR and 90% TIR. Considering that the CEC of the mineral was 410 mmolc/kg, while the DDTMA adsorption plateau was 600 mmol/kg, at pseudotrilayer intercalation, only the upper and lower layers could be retained in the interlayer via cation exchange while the middle layer was retained by hydrophobic interaction. The XRD patterns of HDTMA intercalated rectorite up to 1.75 CEC are plotted in Figure 4 and those with HDTMA intercalation from 2.05 to 3.20 CEC, as well as those determined (25) McAtee, J. L. Clays Clay Miner. 1956, 5, 308–317. (26) Reynolds, R. C., Jr.; Reynolds, R. C., III NEWMOD for WindowsTM. The Calculation of One-Dimensional X-ray Diffraction Patterns of Mixed-Layered Clay Minerals; Hanover, New Hampshire, 1996. (27) Reynolds, R. C. In Crystal Structures of Clay Minerals and Their X-Ray Identification, Brindley, G. W., Brown, G., Eds.; Mineralogical Society: London, 1980; pp 249-303, Monograph no. 5. (28) Reynolds, R. C. Clays Clay Miner. 1983, 31, 233–234. (29) Moore, D. M.; Reynolds, R. C., Jr. X-ray Diffraction and the Identification and Analysis of Clay Minerals; Oxford University Press: New York, 1997; p 378.
8292 DOI: 10.1021/la904677s
Li et al.
Figure 4. XRD patterns of rectorite intercalated with different amounts of HDTMA. A vertical enlargement was made between 10° and 15°. Dashed lines are the NEWMOD simulation of experimental data.
with wide angle XRD, are plotted in Supporting Information Figures S3 and S4. Similar to DDTMA intercalation, HDTMA intercalated rectorite also showed irrational (00l) reflections, broad plateau, and split of the plateau into two peaks as HDTMA loading increases (Figure 4 and Supporting Information Figures S3 and S4). At 0.25 CEC, the d00l-spacing suggested a mixed layer made of 90% of MIR and 10% of TIR (Figure 4). At 0.50 CEC intercalation of HDTMA, a similar broad peak at (002) reflection seen for DDTMA intercalation also occurred. This composite peak could be considered made of peaks at 1.34 and 1.20 nm. Simulation with NEWMOD produced a mixed layer made of 80% MIR and 20% TIR. The broad peak separated into two peaks of 1.42 and 1.05 nm at the 0.75 CEC intercalation level with a d001-spacing of 3.09 nm. NEWMOD simulation produced a mixed layer made of 30% MIR and 70% TIR. Meanwhile, the newly formed (002) reflection was also a broad peak, which further split into two at 1.25 CEC. The d-values of the three (00l) peaks displayed an irrational multiple relationship. As the amount of HDTMA intercalation increased further, the newly split peaks sequentially approached a rational series. A similar observation was found for DDODA intercalated montmorillonite at 2.5 CEC, in which four reflections at 3.81, 1.98, 1.33, and 1.00 nm were recorded13 and for HDTMA intercalated montmorillonite at 2.5 CEC, in which four reflections at 3.84, 1.94, 1.21, and 0.94 nm were recorded.5 A simulation with 20% TIR and 80% PIR matched the experimental data well for HDTMA intercalation at the 1.75 CEC level (Figure 4). Finally, at the 2.54 CEC intercalation, the XRD pattern could be fitted with mixed layers made of 10% TIR and 90% PIR (Supporting Information Figure S3). At the same time, the (00l) reflections when l = 2 to 8 were all visible and rational, even Langmuir 2010, 26(11), 8289–8294
Li et al.
though the (001) reflection was missing. The calculated d-spacing based on the (002) through (008) reflection was 4.95 nm. At 2θ = 2° the d-value from Bragg’s equation is 4.4 nm, less than the d001 value of 4.95 nm. Thus, only the right side of the (001) reflection were visible in these XRD patterns (Supporting Information Figure S3). To verify that the (001) reflection was indeed missing, selected samples were scanned from 0.5° to 8° 2θ using a Bruker wide angle diffractometer. As expected, the missing (001) reflection showed up at the anticipated 2θ with the symmetry and intensity much higher than those of the (002) reflection (Supporting Information Figure S4). A comparison of XRD patterns of rectorite intercalated with 0.74 CEC of DDTMA to that intercalated with 0.5 CEC of HDTMA showed broadening of the (002) reflection before it split into two. The similarity between the XRD patterns of rectorite intercalated with 1.09 CEC of DDTMA and rectorite intercalated with 0.75 CEC of HDTMA was also found. These similar XRD patterns again strongly suggested that a transition from MIR to TIR was via a mixed layer intermediate made of these two components without a bilayer as an intermediate. A similar plateau was observed when SWy-2 was modified by ODTMA to 0.6 CEC30 and a montmorillonite modified by HDTMA to 1.5 CEC5 and by DDODA to 0.8 CEC.13 However, the peaks at 1.7-1.8 and 1.3-1.5 nm were explained due to bilayer and monolayer arrangements of the intercalated organic molecules, respectively,30 as a bilayer intercalation in montmorillonite would result in a d-spacing of 1.76 nm.1 After adding another 1.0 nm of the illite component to the intercalated montmorillonite, the d001 would be around 2.7 nm, as seen in the XRD pattern of rectorite intercalated with 0.74 CEC of DDTMA (Figure 3). Thus, without considering the (002) plateau, one would assume that the intercalated DDTMA adopted a horizontal bilayer interlayer configuration. However, the broader plateau of the XRD pattern suggested that it was made of a mixed layer of MIR and TIR as confirmed by NEWMOD simulation. Although a bilayer intercalation was indeed observed in TEM fringe image,30 it may represent a selected section of a crystal grain microscopically, while the XRD patterns of the mixed layer feature in this study reflect the interlayer configuration of the intercalated alkylammonium at a macroscopic scale. Thus, the most stable configurations would be horizontal monolayer and horizontal pseudotrilayer, as an ideal bilayer would involve in a much higher ordering, and thus, the entropy would not be favored. Due to the presence of higher orders of reflection after HDTMA intercalation, the basal spacing could be more accurately determined and the interlayer configuration deduced. In contrast, when montmorillonite was used, only (001) reflection was observed.6,7 Reflections at 3.57 and 1.85 nm was observed after a bentonite was treated with HDTMA at an initial amount corresponding to 4.0 CEC, but the first one was attributed to vertical arrangement of the HDTMA molecules with respect to bentonite surfaces, while the second one to horizontal bilayer,8 even though these two should have been treated as (001) and (002) reflection of bentonite intercalated with HDTMA in a vertical or near vertical direction. Due to the absence of higher-order reflections in these XRD patterns,6-9 montmorillonite renders its unique properties to rectorite for the study of intercalation of organic molecules. FTIR Analyses. The bands near 2917 and 2850 cm-1 are CH2 antisymmetric (νas) and symmetric (νs) stretching bands, respectively, and are sensitive to the gauche/trans conformer ratio of the hydrocarbon chains.7,9,11,13,17-19 At the OTMA intercalation (30) Xi, Y.; Frost, R. L.; He, H.; Kloprogge, T.; Bostrom, T. Langmuir 2005, 21, 8675–8680.
Langmuir 2010, 26(11), 8289–8294
Article
maximum, they are located at 2931 and 2860 cm-1 (Supporting Information Figure S5), compared to 2918 and 2849 cm-1 for HDTMA intercalation at the 3.2 CEC loading level17 and 2925 and 2851 cm-1 for DDTMA intercalation at the 1.67 CEC loading level.19 A shift from low wavenumbers, characteristic of highly ordered, all-trans conformations, to higher wavenumbers and increased width was accompanied by an increase in the number of gauche conformers (the “disorder” of the chain).31 Thus, a random gauche conformation would be the major configuration of the monolayer OTMA in the interlayer and mixed layer made of MIR and TIR at an HDTMA loading level of 0.74 CEC in contrast to more orderly all-trans conformation for HDTMA and DDTMA intercalation at higher alkylammonium loading levels (Supporting Information Figure S5).17,19 Due to intercalation, vibrations corresponding to antisymmetric (νas) and symmetric (νs) stretching bands of CH3-N in the headgroup at 3032 and 3019 cm-1 disappeared (Supporting Information Figure S5), suggesting a restricted mobility of the trimethylammonium headgroup. Dynamic molecular simulations of HDTMA intercalated smectite showed that, without cointercalation of counterion acetate, all trimethylammonium headgroups of HDTMA were fixed firmly above the center of the surface sixmembered rings through electrostatic attraction interactions.32 The peak at 1634-1650 cm-1 (Supporting Information Figure S6) was due to H-O-H bending vibration and the intensity reflected the amount of water in the interlayer.17 Its intensity for raw rectorite was 0.053, and it decreased to 0.017 at 0.71 CEC intercalation of OTMA, agreeing well with HDTMA and DDTMA intercalation.17,19 The FTIR results showed that intercalation of OTMA reduced the interlayer water content on one hand, which is due to substitution of hydrated calcium by OTMA, and the state of presence of OTMA in the interlayer was in the formation of monomer or a monolayer on the other hand. The FTIR absorption band at 1484 cm-1 (Supporting Information Figure S6) was due to the antisymmetric bending mode of the head [(CH3)3Nþ-] methyl group and was sensitive to the extent of disorder and packing of the headgroup.32,33 For micellar systems in the interlayer space of rectorite, it was made of a doublet of 1480 and 1487 cm-1, similar to that of solid HDTMA and DDTMA, reflecting a more orderly arrangement of the intercalated alkylammonium, while for HDTMA and DDTMA monolayer coverage, a single peak was located at 1488 cm-1,.17,19 The absorption band location for OTMA intercalation matched with that for HDTMA17 and DDTMA19 at low loading level, further confirming a monolayer intercalation of OTMA in the interlayer space of the montmorillonite component in rectorite. The doublet of the asymmetric bending mode (δas) of CH3-N at 1480 and 1487 cm-1, reflecting a more orderly arrangement of the alkylammonium, was reduced into a singlet at all OTMA loading levels. Meanwhile, the symmetric bending mode (δs) of CH3-N at 1408 cm-1 shifted to a much higher wavenumber at 1419 cm-1 (Supporting Information Figure S6), again suggesting a stronger interaction between the trimethylammonium headgroup and the surface. In addition, the band at 1431 cm-1 is the scissoring band of methylene groups adjacent to a positively charged nitrogen (δ, R-CH2).31 Its absence also indicates restricted move of the monolayer OTMA in the interlayer. (31) Weers, J. G.; Scheuing, D. R. In Fourier Transform Infrared Spectroscopy in Colloidal and Interface Science, Scheuing, D. R., Ed.; ACS Symposium Series 447; American Chemical Society: Washington, DC, 1990; pp 87-122. (32) Liu, X.; Lu, X.; Wang, R.; Zhou, H.; Xu, S. Am. Mineral. 2009, 94, 143–150. (33) Wong, T. C.; Wong, N. B.; Tanner, P. A. J. Colloid Interface Sci. 1997, 186, 325–331.
DOI: 10.1021/la904677s
8293
Article
Li et al.
Table 1. Particle Thickness (in nm) Calculated from Mudmaster for Samples Loaded with 3.20 CEC of HDTMA diffraction peak
raw rectoriteb
0.71 CEC OTMAb
1.67 CEC DDTMAb
3.20 CEC HDTMA
a b (001) 36.2 23.4 18.5 (002) 21.5 16.6 20.1 27.4 (003) 14.0 19.0 25.7 (004) 16.3 26.5 (005) 19.1 18.5 (006) 21.6 21.0 (007) 17.1 20.0 (008) 22.6 22.9 Average 22.4 16.4 19.4 23.2 a Calculated using Mudmaster.35 b Calculated using method of Cullity and Stock.34
Particle Size Determination. The full width at half-maximum (fwhm) of a XRD peak can be used to estimate the vertical thickness of the layered silicates using of the Scherrer equation34 t ¼
0:9λ B cos θB
ð1Þ
where B is the fwhm in radius, θB is the angle of the peak in degree, λ is the wavelength of X-ray radiation (1.541 78 A˚ for Cu KR). Alternatively, it can be calculated using the Mudmaster program.35,36 A vertical thickness of 36.2 nm was calculated for raw rectorite. Using the d001 spacing of 2.58 nm, the vertical thickness was about 14 layers thick (Table 1). The vertical thickness was reduced to 22.4 nm, or 9 layers, after maximum intercalation of OTMA and further reduced to 16.4 nm, or 5 layers, at DDTMA intercalation maximum. At HDTMA intercalation maximum of 3.2 CEC, the average vertical thickness was between 19.4 and 23.2 nm, suggesting that the particles were made of 4 repeating units. These results suggest that different chain lengths could result in different degrees of delamination, with a longer hydrocarbon chain resulted in a higher degree of dispersion (more reduction in average vertical thickness), when alkylammonium was used as a dispersing agent. In order to reinvestigate the “fundamental particles” and “inter-particle diffraction” proposed by Nadeau et al.,37 physical mixtures of smectite and rectorite were made and examined by X-ray diffraction analysis and transmission electron microscopy.38 It was found that the mixtures of smectite and rectorite of 0.5-2 μm fractions were merely physical mixtures and did not show interparticle diffraction. In contrast, the mixtures of smectite and rectorite in the
