Investigations of surfactant effects on gas hydrate formation via infrared spectroscopy

Journal of Colloid and Interface Science 376 (2012) 173–176

Contents lists available at SciVerse ScienceDirect

Journal of Colloid and Interface Science www.elsevier.com/locate/jcis

Investigations of surfactant effects on gas hydrate formation via infrared spectroscopy Chi Lo a,b, Junshe Zhang a, Ponisseril Somasundaran b,⇑, Jae W. Lee a,⇑ a b

Department of Chemical Engineering, The City College of New York, New York, NY 10031, United States Department of Earth and Environmental Engineering, Columbia University, New York, NY 10027, United States

a r t i c l e

i n f o

Article history: Received 21 January 2012 Accepted 3 March 2012 Available online 12 March 2012 Keywords: Infrared spectroscopy Gas hydrates Surfactant SDS Hydrate formation

a b s t r a c t This infrared (IR) spectroscopic study addresses surfactant effects on cyclopentane (CP) hydrate–water interfaces by observing both ice-like (3100 cm 1) and water-like (3400 cm 1) bands in the bonded OH region together with free OH bands. IR spectroscopy of hydrates has not been actively employed due to the overwhelming signal saturation of the OH bonding. However, this work is able to utilize this large signal of the OH bonding to understand the water structure changes upon adding sodium dodecyl sulfate (SDS) to CP hydrate–water interfaces. The spectral data suggest a change to more ice like (3100 cm 1) features starting from 100 ppm to 750 ppm SDS, indicating favorable nucleation. At the same instance, water like (3400 cm 1) features are also shown in this range of SDS concentration, which suggests looser hydrogen bonding that is an indicator for facilitating hydrate growth. Additionally, this ATR-IR study firstly identifies both symmetric and anti-symmetric free OH bands of the hydrogen bond (HB) acceptors in the clathrate hydrate system. Relative area ratios of free and bonded OH bands provide important information about spatial arrangements of adsorbed SDS monomers. Ó 2012 Elsevier Inc. All rights reserved.

1. Introduction Gas hydrates usually contain a guest molecule of typically low molecular weight, i.e. methane, ethane, cyclopentane (CP), etcetera, being encaged by a hydrogen bonded network of water molecules. These hydrates have three known structures of sI, sII, and sH and the structure type results in different hydration numbers [1]. Gas hydrate formation favors low temperatures and high pressures. The formation requires the stabilization of the water host network. This stabilization is achieved by enclathration of guest molecules and is detectable by near-infrared spectroscopy in terms of observing the OH stretching of tetrahedrally coordinated water [2]. The OH stretching indicates an ordered or unordered hydrogen bonding [3]. However, there are only a few gas hydrate studies using mid-IR [2,4–5] since OH broadening is quite extensive and covers a lot of the spectra, such that other chemical vibrations are covered by the absorption of OH. Thus, the technical problem of using IR is that water has a high infrared absorptivity. On the contrary, this problem turns out to be an advantage for the study of hydrogen bonding at hydrate–water interfaces because OH signal bands are ⇑ Corresponding authors. Address: Chemical Engineering Department, The City College of New York, 140th St., and Convent Ave., New York, NY 10031, United States. Fax: +1 212 650 6660. E-mail addresses: [email protected] (P. Somasundaran), [email protected]. edu (J.W. Lee). 0021-9797/$ - see front matter Ó 2012 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.jcis.2012.03.012

relatively large, leading to greater resolution. The other problem for using IR for hydrate studies is the cooler temperature. IR is sensitive to temperature changes and absorption signal intensity can be lowered due to the low temperatures, typical of hydrate forming conditions (less than 10 °C). However, if any reduction in intensity due to the lower temperatures, the intensity of this system is still higher than the rest of the spectra, as mentioned before, because the HOH stretching is the interested band area. Surface studies of hydrates are limited due to the difficulties in detecting scattered particles and uneven hydrate formations. Many have used intricate methods of hydrate growth on ATR surfaces like vapor deposition and epitaxial growth [2,4–5]. These methods tried to increase the contacting surface area to amplify signal strength. However, it is observed in this work that the signal strength at the hydrate interface is high due to the inherent spectrum affinity for OH absorption. Most of previous studies on free OH bonds have employed the technique of Vibrational Sum Frequency Generation (VSFG) at various interfaces. This powerful method has been able to probe the anisotropy of the liquid–air and liquid–liquid interface [6]. However, the weaker hydrogen bond (HB) acceptor species have not had well defined structures from this method. Conrad and Strauss [7] showed that ATR-IR is able to have well defined spectra for the HB acceptor of water in oil (alkanes). Shin et al. [8] determined water cluster structures by identifying the HB acceptor and its hydrogen bonded cage structure using size selected infrared spectroscopy.

C. Lo et al. / Journal of Colloid and Interface Science 376 (2012) 173–176

Raman has revealed spectroscopic evidences for hydrate-like sensitivity related to nucleation, initial growth, gas hydrate structure, and hydrate stability [1,9]. However the role of surfactants has not been understood throughout the extended growth phase of hydrates when a small amount of sodium dodecyl sulfate (SDS) is involved in hydrate formation. Thus, this work attempts to elucidate why SDS acts not only as a nucleation promoter but also as a growth promoter and to tie these two roles by employing IR spectroscopy. In addition, we unprecedentedly report the determination of free, both symmetric and asymmetric, OH bands of the HB acceptor vibrational signal on hydrate–water interfaces from ATR-IR spectroscopy. The spectra retain very clear peaks and the relative strength of free OH bands to bonded OH bands provides important insights into the configuration of SDS molecules adsorbed onto the hydrate–water interface.

3100cm

-1

3400cm

-1

2

1000ppm SDS 750ppm SDS 500ppm SDS 250ppm SDS

Absorbance

174

100ppm SDS

50ppm SDS

0

0ppm SDS

2800

2. Experimental section

3200

3600

4000

4400

-1

Wavenumber (cm )

Cyclopentane (CP) and sodium dodecyl sulfate (SDS) were purchased from Sigma–Aldrich with a purity of 99+%. All chemicals were used as received without additional purification. Deionized (D.I.) water was produced in our lab with a resistivity of 18 MX cm 1. 2.2. Synthesis of CP hydrate slurries Three hundred grams of CP and water mixture (the mass ratio of CP to water was 1:9) were charged to a 1 L glass bottle and then hermetically sealed. The bottle was placed in a freezer at 263 K until all water converted to ice, and then the bottle was shaken vigorously at room temperatures. As the ice melted, CP was enclathrated at the same time. The occurrence of CP hydrates was confirmed by the appearance of fine white particles. After visually observing that large portions of ice had melted, the bottle was placed in a chiller at 277 K for 1 week and it was shaken daily to enhance the enclathration. The concentration of CP hydrates in slurries was 51 wt.% by calorimetric measurements. 2.3. Attenuated total reflectance infrared spectroscopy A Perkin Elmer Spectrum 100 is used to take measurements using a cooled MCT sensor. The ZnSe Horizontal Attenuated Total Reflectance (HATR) trough is cooled to 1–5 °C during the experiment with a mixture of dry ice and water ice. Hydrate slurry mixture is ladled onto the crystal surface. A spectrum from 500 cm 1 to 3800 cm 1 was taken. The cooled HATR surface is taken as the background. 100 measurements were taken for each sample. Water/solution subtraction (difference) was done after measurements via the Perkin Elmer software. 3. Results and discussion Both bonded OH stretching and free OH stretching of cyclopentane (CP) hydrates are the focus of this study. CP hydrates are a good model hydrate system under spectroscopic measurements [9] since their formation requires ambient pressure and at the same time they possess the same structure as natural gas hydrates (sII). The bonded OH stretching can represent ordered ‘‘ice-like’’ (ca 3100 cm 1) and unordered ‘‘water-like’’ environments (ca 3400 cm 1) [3,10]. Red/blue shifts in the bands are inferred as more ordered/unordered tetrahedrally coordinated hydrogen bonded water. Free OH stretching or the HB acceptor bands are used to determine the penetration depth of a water molecule into an alkyl phase and the free OH stretching can also determine the

Fig. 1. ATR-IR measurements of CP hydrates with varying SDS concentrations with a background subtraction for aqueous solutions.

structure of the water cluster, like the magic number clathrate [8]. The ATR-IR measurements for the CP hydrate interface with varying SDS concentrations are depicted in Fig. 1. The definitions of ‘‘ice-like’’ and ‘‘water-like’’ are in regard to the 3100 cm 1 and 3400 cm 1 bands in the bonded OH region [3,10]. From 0 to 50 ppm SDS, the peak remains at or above the ice-like demarcation (3100 cm 1). Above 100 ppm of SDS, the spectra show red shifts and these indicate ice-like tendencies. This trend extends up to 1000 ppm SDS, where the peak position is the same as for zero ppm SDS. The addition of SDS also changes the water-like band (3400 cm 1) and it results in more blue shift. The OH stretch at 3000–3500 cm 1 represents the spectra for colligative properties of water. The blue shift is indicative of water in a loose bonding structure, therefore a higher energy state. As the SDS concentration increases, the appearance of the water-like band increases in absorbance. At 50 ppm SDS, the water-like band appears at 3400 cm 1. Above 100 ppm SDS, the band begins to show the blue shift indicating a more water-like structure. This blue shift is also indicative of the displacement of water from the surface and also an indication of increased hydrophobicity [10]. The water-like bands possess more blue shifts with the increased SDS concentration, indicating

0.8

Absorbance

2.1. Materials

1000ppm SDS 750 ppm SDS 500ppm SDS 250ppm SDS 100ppm SDS 50ppm SDS 0ppm SDS

0.4

0.0 3500

3600

3700

3800 -1

Wavenumber (cm ) Fig. 2. The free OH bands with varying SDS concentrations on CP hydrate slurry; 3690–3760 cm 1 is the asymmetric OH stretch while 3580–3650 cm 1 is the symmetric OH stretch.

C. Lo et al. / Journal of Colloid and Interface Science 376 (2012) 173–176

Band Area Ratio Absorbance

0.115 0.110 0.105 0.100 0.095 0.090 0.085 0.080 0.075 0.070 0

200

400

600

800

1000

SDS Concentration (ppm) Fig. 3. Band area ratios of free OH to hydrogen bonded OH.

looser hydrogen bonds and a maximum shift seems to occur around 750 ppm SDS as shown in Fig. 1. The free OH bands are depicted in Fig. 2 from 3500 to 3800 cm 1. These free OH bands come from non-colligative interactions of the individual molecules. The higher frequency band is due to the asymmetric OH stretch, whereas the lower frequency band is due to the symmetric OH stretch [3]. These two bands are indicative of the alkyl moiety environment and the penetration of free OH into the alkyl moieties [3,6,8]. As the bands indicate the hydration number, it was shown that n = 6–27 water clusters have different vibrational associations for each hydration cluster [8]. From the data shown in Fig. 2, an n = 11 water cluster is shown, meaning this is a half clathrate or the surface of the clathrate. These bands are present in this system because the clathrate

175

surface is an open surface of pendant hydrogens. This interpretation is similar to a dehydrating surface where the surface concentrations of OH vary under different dehydration stages [11]. We take band ratios of the area of free OH to the area of bonded OH stretch to compare the depth of penetration into the alkyl moiety in Fig. 3. A band ratio peaks around 100 ppm SDS, where we note from Lo et al. [12] that this is where the surfactant is lying flat on the hydrate surface and SDS coverage of the hydrate surface begins. The penetration of water molecules into the alkyl moiety phase is greater at this SDS concentration since the alkyl chain is contacting the surface of the clathrate and it may come along with hydrate formers. Above 100 ppm SDS the absorption intensity is a little lowered. Therefore the half clathrate may not contact with the alkyl layer, inferring that there is a separation of the two layers. Such a separation could be indicative of a monolayer formation where water is squeezed out of the packed chains [10]. There is another maximum at 750 ppm SDS, which is at the concentration of a secondary layer identified from the SDS adsorption isotherm [12]. As more SDS accumulates onto the surface, monolayer and bi-layer can form and induce an increased diffusion of hydrate formers to the bulk and onto these interfaces [12,13]. With SDS concentrations of 100 ppm and above, the bonded OH starts to become mobile (water like) as shown in Fig. 1. Thus, the water-like OH bond may contribute to reducing the diffusion limitation of hydrate former to the hydrate growth front and at the same time, small hydrate clusters (ice-like feature) can easily be accumulated to the growth front to proceed to the extended growth phase. As the growth front expands by producing large crystals, SDS molecules may be precipitated from the solid phase or dissolved in available aqueous phases. The ice-like structure at the interface may be favorable to only nucleation because an armoring effect is present from the shell

At or below 100ppm SDS, surfactant lies flat and in contact with surface

From 100ppm SDS, surfactants form a monolayer and the hydrate surface is stretched to a liquid like state with tails providing the ice -like features necessary for nucleation

Fig. 4. A proposed depiction of the hydrate interface.

176

C. Lo et al. / Journal of Colloid and Interface Science 376 (2012) 173–176

hydrate layer preventing more hydrate formations into the ice melt core [14]. This armoring effect was conjectured to be due to longer growth periods or sufficient grain barriers preventing growth. There is a specific spectroscopic evidence for the armoring effect from this work (no water-like band for 0–50 ppm SDS in Fig. 1). The absence of the water-like band for CP hydrates is an indication of this armoring effect for further hydrate growth. This is consistent with recent studies [15–17], indicating that for further hydrate growth a melting time or slight dissociation of the solidus crystal is required. This continual growth process is due to loose hydrogen bonding at the hydrate interface, enabling hydrate surfaces to fulfill the matrix with hydrate formers. Fig. 4 is an animated schematic of the proposed adsorption process. These loosened hydrogen bonds (disarming hydrogen bonds) at the hydrate interface can facilitate the growth process by preventing the armoring effect. Solidus ice or hydrate can prevent further growth into the water phase and slow down the growth process. SDS adsorbs onto the hydrate surface and is able to loosen the hydrogen bonds upon the surface thereby seemingly melting the ice like layer upon hydrate interfaces. 4. Conclusions The ATR-IR spectroscopic investigation of CP hydrate–water interface reveals that above 100 ppm SDS, the hydrate interface possesses the water-like state facilitating further hydrate growth as well as the ice-like state for facile nucleation. This measurement affirms the mechanism about where the growth process starts once there is a nucleating ice-like feature available and continues once there is loosening of hydrogen bonds by adding SDS. ATR-IR also confirms the lying-down configuration of SDS at 100 ppm SDS,

where the free to bonded OH band ratio is higher than at other concentrations. The spectroscopic data presents a direct measurement of the hydrate interface and supplies a way of identifying interfacial properties upon hydrate interfaces. Acknowledgments The authors are grateful for the support of the National Science Foundation for this work (under Grant Numbers CBET-0854210 and HRD-0833180). References [1] E.D. Sloan, C. Koh, Clathrate Hydrates of Natural Gas, third ed., CRC, Boca Raton, 2008. [2] G.T. Dobbs, Y. Luzinova, B. Mizaikoff, Infrared spectroscopy for monitoring gas hydrates in aqueous solution, in: Proceedings of the 6th International Conference on Gas Hydrates (ICGH 2008). [3] C.S. Tian, Y.R. Shen, J. Am. Chem. Soc. 131 (2009) 2790. [4] F. Fleyfel, J.P. Devlin, J. Phys. Chem. 92 (1988) 631. [5] F. Fleyfel, J.P. Devlin, J. Phys. Chem. 95 (1991) 3811. [6] L.F. Scatena, M.G. Brown, G.L. Richmond, Science 292 (2001) 908. [7] M. Conrad, H.L. Strauss, Biophys. J. 48 (1985) 117. [8] J.-W. Shin, N.I. Hammer, E.G. Diken, M.A. Johnson, R.S. Walters, T.D. Jaeger, M.A. Duncan, R.A. Christie, K.D. Jordan, Science 304 (2004) 1137. [9] C. Lo, J.S. Zhang, P. Somasundaran, J.W. Lee, J. Phys. Chem. Lett. 1 (2010) 2676. [10] N. Kumar, C. Maldarelli, A. Couzis, Colloids Surf., A 277 (2006) 98. [11] C. Moterra, J. Chem. Soc. Faraday Trans. I (84) (1988) 1617. [12] C. Lo, J.S. Zhang, P. Somasundaran, A. Couzis, J.W. Lee, J. Phys. Chem. C. 114 (2010) 13385. [13] P. Somasundaran, B.M. Moudgil, J. Colloid Interface Sci. 47 (1974) 290. [14] L.A. Stern, S.H. Kirby, W.B. Durham, Science 273 (1996) 1843. [15] N. Liu, W. Chen, D. Liu, Y. Xie, Energy Convers. Manage. 52 (2011) 2351. [16] Y. Qi, M. Ota, H. Zhang, Energy Convers. Manage. 52 (2011) 2682. [17] P. Karanjkar, J.W. Lee, J.F. Morris, Chem. Eng. Sci. 68 (2011) 481.