Article Cite This: J. Phys. Chem. C XXXX, XXX, XXX−XXX
pubs.acs.org/JPCC
Guest Partitioning and Metastability of the Nitrogen Gas Hydrate Claire Petuya,† Françoise Damay,‡ Bertrand Chazallon,§ Jean-Luc Bruneel,† and Arnaud Desmedt*,† †
Groupe Spectroscopie Moléculaire, Institut des Sciences Moléculaires, UMR 5255 CNRS, Université de Bordeaux, 351 Cours de la Libération, F-33404 Cedex Talence, France ‡ Laboratoire Léon Brillouin, UMR 12 CEA-CNRS, CEA Saclay, Bât. 563, 91191 Cedex Gif-sur-Yvette, France § Physique des Lasers Atomes et Molécules (PhLAM), UMR 8523 CNRS, Université Lille, F-59000 Lille, France S Supporting Information *
ABSTRACT: Gas clathrate hydrates or gas hydrates are made of H-bonded water molecules forming cages, within which gaseous (guest) molecules are encapsulated. The formed clathrate structures, which may be metastable, depend on the nature and on the partitioning of the guest molecules in the water cage. This work focuses on the structural and vibrational properties of nitrogen hydrate in its two clathrate forms (namely, SI and SII) in the thermodynamic ranges 50−200 bar and 150−270 K, together with a comprehensive analysis of the transformation from SI to SII of this gas hydrate. The thermal expansion of both structures has been measured at 1 bar, and the melting of the nitrogen hydrate has been measured at ca. 210 K at 1 bar. Moreover, the SI structure is metastable in the studied pressure region: from time-dependent neutron powder diffraction analysis, it is shown that the SI structure transforms over time to the SII structure with a rate of (1.37 ± 0.17) × 105 s−1 at 100 K and at 1 bar. The transformation is also characterized by an induction time (i.e., the lifetime of the pure SI structure) of 0.49 day. We have also investigated the guest partitioning of the nitrogen hydrate using highresolution Raman scattering. Vibrational bands of nitrogen molecules encapsulated in large cages are measured at lower wavenumbers than the one associated with encapsulation in small cages (by 1.1 cm−1 in SI and 0.8 cm−1 in SII). In the case of the thermodynamically stable SII phase, the dependence of the guest partitioning has been characterized as a function of the pressure−temperature conditions. Variation of the relative cage filling is demonstrated. While the small cages remain singly occupied according to previous neutron diffraction analysis, this variation is attributed to large cages of the nitrogen hydrate that easily catch or release nitrogen guest molecules. This study thus provides new opportunities for preparing nitrogen gas hydrates with a “targeted” structure and relative cage filling not only by varying the pressure and temperature but also by playing with the structural metastability.
■
INTRODUCTION Gas hydrates have been discovered in pipeline plugs and then on the ocean floor in the natural state, after having been first synthesized in the early 20th century.1 Since these discoveries, research interests in gas hydrates have continuously expanded not only for their applications in broad areas of flow assurance, energy, technology, or the environment, but also for their fundamental interests ranging from physical chemistry to astrophysics.2 In this respect, the nitrogen hydrate is of particular relevance for many aspects. Nitrogen plays a role as a promoter in the technology of methane replacement by carbon dioxide in natural gas hydrates.3 Nitrogen hydrate is also particularly relevant in the field of atmospheric chemistry and in astrophysics due its natural occurrence on earth and its hypothetical existence elsewhere in the solar system.4 Air hydrates, formed with the help of entrapped nitrogen and oxygen in sedimentary ices, have been widely investigated to evaluate their role in the reconstruction of the earth’s atmospheric composition through the measurements of gas concentrations. 5−10 Gas hydrates are also essential in astrophysics because of their hypothetical implication in the © XXXX American Chemical Society
formation of planetesimals, comets, and other planets of our solar system (e.g., Saturn’s moons, such as Titan11 or Enceladus12,13). N2 gas is the most abundant form of nitrogen in the protosolar nebular,14 and the plausible existence of nitrogen hydrate may explain the nitrogen depletion in the atmosphere of comets such as 67P/Churyumov-Gerasimenko.15 Nitrogen hydrate crystallizes into the so-called structures SI (cubic with a ≈ 12 Å) or SII (cubic with a ≈ 17 Å).1 Such nanoporous structures contain two classes of water cages: small cages and large cages (denoted SC and LC, respectively). The SI structure contains 2 SCs (denoted 512) and 6 LCs (51262), and the SII structure contains 16 SCs (512) and 6 LCs (51264), slightly bigger than the SI LCs. The work of Davidson et al.16 has shown that small molecules such as Ar or Kr may form the SII structure. The ability of such small molecules to be included in small cages will promote the formation of the SII structure, Received: October 13, 2017 Revised: November 17, 2017 Published: December 12, 2017 A
DOI: 10.1021/acs.jpcc.7b10151 J. Phys. Chem. C XXXX, XXX, XXX−XXX
Article
The Journal of Physical Chemistry C
combining experimental techniques particularly appropriated to the investigation of gas hydrates: high-resolution Raman scattering (HRRS) 31 and powder neutron diffraction (NPD).32 Thus, after the quantitative analysis of the kinetics associated with the SI−SII transformation by means of NPD, guest partitioning and structural expansion are comprehensively detailed for the pure SI ad SII structures. Finally, details about the temperature and pressure dependence of the guest partitioning are provided for the SII structure.
and the thermodynamically stable structure of nitrogen hydrate is indeed the SII type. The SI structure is predicted from thermodynamic modeling to be stable at high pressure (typically at pressures of 1000 bar or higher), while the SII structure is expected at moderate to low pressures (typically 100 bar).17 Pure SI nitrogen hydrate has been synthesized and analyzed by means of X-ray diffraction analysis3 and Raman scattering.3,18 At a moderate pressure of ca. 500 bar, the SI structure has been found to also be a metastable phase of the nitrogen hydrate, existing at the initial stage of the hydrate formation.8 By analogy with the isosteric carbon monoxide hydrate,19,20 the SI phase is kinetically favored while the SII phase is thermodynamically stable when considering moderate pressure on the order of hundreds of bar. Unfortunately, only qualitative analysis of this transformation of the nitrogen hydrate is available in the literature to the best of our knowledge. Such studies reveal that the size of the guest molecule is not the unique parameter triggering the type of formed hydrate structures (SI or SII); the cage occupancy is also involved. The cage occupancy has been investigated by means of neutron and X-ray diffraction8,10,21 and by means of modeling and simulations22−24 of the SII structure: at 200 bar and 273 K, the occupancies are 82.2% and 99.6% in the SCs and LCs, respectively, according to Rietveld refinement of the structure.10 In the SI phase at 1093 bar and 268 K, the occupancies are 98% and 112% in the SCs and LCs, respectively.3 Whatever the clathrate structure is, the SCs are almost fully occupied and the LCs become partly doubly occupied above a certain pressure. Spectroscopic signatures of the cage occupancy are difficult to determine, due to the weak interactions between the nitrogen molecules and the water cages.25,26 Previous Raman analysis has revealed that the N2 vibron is observed at lower frequency in the hydrate phase than in the gas phase: respectively, 2322 and 2329 cm−1 at 40 K in the 200−400 bar pressure range.27 Several Raman analyses3,5−7,9,18,20,28,29 have been performed without identifying the vibrational signatures of nitrogen partitioning in the SCs and in the LCs at moderate pressure (i.e., on the order of 100 bar). The Raman observation of this partitioning has been tentatively done either with the help of mixed gas hydrates30 or by performing experiments at pressures significantly higher than 1000 bar.29 The co-inclusion of methane and N2 molecules leads to the formation of an SI structure with the Raman signatures measured at 2322.98 and 2324.17 cm−1 (at 84 K and atmospheric pressure) for N2 in LCs and SCs, respectively.30 By forming mixed gas hydrate with propane and N2, the SII Raman signatures of N2 in LCs and in SCs are measured at 2323.50 and 2323.88 cm−1, respectively (at 84 K and atmospheric pressure).30 Sasaki et al. tentatively attributed the vibrational signatures of N2 molecules encapsulated in SCs and in LCs, with a splitting of 3.1 cm−1 at 2900 bar and room temperature:29 the N2 vibrons are observed at 2322 and 2325.1 cm−1 and assigned to the N2 stretching in the LC and SC, respectively. Moreover, they propose that doubly occupied LCs give a Raman signature contained within this 3.1 cm−1 range,29 in agreement with simulated Raman spectra.23 The present paper is dedicated to the investigation of the nitrogen hydrate at pressures up to 200 bar, aiming at determining the guest partitioning in the SI and SII structures and at quantitatively analyzing the impact of the thermodynamic conditions on the metastability of the hydrate structure and its guest partitioning. Such properties can be unraveled by
■
EXPERIMENTAL DETAILS Neutron Powder Diffraction. The experiments were performed at the Laboratoire Léon Brillouin (CEA-Saclay) using the cold neutron two-axis powder diffractometer G 4-1. Measurements were done with a wavelength of 2.428 Å. Due to a better instrumental resolution at low scattering angle 2θ, the angular range 2θ < 60° was considered. The samples (sealed in a cylindrical vanadium holder) were cold transferred into the cryostat at 80 K. Diffractograms were recorded at 40, 60, 80, and 100 K with increasing temperature, with an acquisition time of 40−60 min. For the melting point and thermal expansion studies, the sample was cooled to 20 K and warmed from 20 to 220 K for SII and to 250 K for SI, with a heating rate of 0.66 K/min between 20 and 120 K and of 0.33 K/min between 120 and 220 K (or 250 K). At each temperature (every 20 K below 120 K and every 10 K above 120 K), diffractograms were recorded with a 30 min acquisition. The pattern-matching of the powder diffractogram was performed with the Le Bail algorithm33 and the FULLPROF program from the FULLPROF suite.34 Raman Scattering. The spectra were recorded with a Labram microspectrometer (Horiba Jobin Yvon, Villeneuve d’Ascq, France) using a 514 nm wavelength laser as the excitation source. A 50× objective (Olympus) allows the incident laser beam to be focused on a micrometer-sized part of the sample. The Raman signal is dispersed by a holographic grating of 2400 lines/mm (high resolution) and analyzed with a Peltier-cooled charge-coupled device (CCD) detector (Andor, Belfast, U.K.). Thus, the intensity of different wavelengths of the spectrum can be measured with a spectral resolution of 0.8 cm−1 (full width at half-maximum) and a CCD pixel size of 0.2 cm−1. The wavenumber calibration of the spectrometer is done using the 520.7 cm−1 mode of a silicon sample and the neon source excitation at 2348.4 cm−1. The data are collected in spectral regions between 150 and 3800 cm−1. For the in situ experiments, the sample temperatures are maintained at the desired value (±0.1 K) during the acquisition by using a laboratory-modified temperature-controlled stage (based on the THMS600, Linkam Scientific Instruments Ltd., United Kingdom), including a homemade high-pressure optical cell equipped with a 2 mm thick sapphire optical window. The sample pressures are controlled by the PM high-pressure pump. All spectra are collected over typically 25 min with increasing temperature (every 20 K between 150 and 270 K) along the isobar at 200 bar (line B−A in Figure 1). Then by staying in the equilibrium thermodynamic region of the nitrogen hydrate, spectra are recorded with increasing pressure (at 60, 80, 100, 120, 150, 170, and 200 bar) along the isotherm at 150 K (line E−B in Figure 1). Examples of spectra recorded in the full spectral range are provided in the Supporting Information (Figure S2). Sample Preparation. For the neutron diffraction experiment, the sample was formed using a powder (grain size B
DOI: 10.1021/acs.jpcc.7b10151 J. Phys. Chem. C XXXX, XXX, XXX−XXX
Article
The Journal of Physical Chemistry C
Figure 1. Phase diagram of nitrogen clathrate hydrate. The square point corresponds to the melting point as estimated by means of neutron powder diffraction. After refs 35−38.
Figure 2. Time evolution of neutron diffraction of the nitrogen hydrate (D2O) at 100 K and at atmospheric pressure. On each pattern, the duration of the reaction time (pressurization at 200 bar between N2 and powdered ice at 260 K) is indicated as well as the storage duration (in parentheses) at atmospheric pressure and liquid nitrogen temperature. Bragg peaks marked with diamonds, asterisks, and arrows correspond to the SI structure, to the SII structure, and to the hexagonal ice, respectively.
typically on the order of 100−200 μm prepared at 190 K under an inert atmosphere) of deuterium oxide ice (99.9% D) and by applying a N2 gas pressure of 200 bar (purity >99.997%). Deuterated samples are used due to the large incoherent scattering cross section of proton. Using protonated or deuterated water cages does not impact the structural properties of the nitrogen hydrates as recently shown by means of X-ray diffraction.21 The cell temperature was maintained at 260 K (±5 K) with a thermal bath over various duration times (from a few hours to several days) to ensure the formation within the thermodynamic stability region (Figure 1). Such thermodynamic conditions of formation correspond to region A in Figure 1. The sample pressures were controlled by a PM high-pressure pump (Top Industrie, Vaux-le-Penil, France) which can contain 100 cm3 of gas up to 500 bar. The nitrogen hydrate is recovered by depressurizing to atmospheric pressure and simultaneously cooling to liquid nitrogen temperature (77 K) for storage (along lines A−C and C−D in Figure 1). For in situ Raman spectroscopy, Milli-Q water was injected inside the optical cell and the temperature was controlled by using a modified temperature stage (Linkam Scientific Instruments Ltd., United Kingdom). First, the hexagonal ice was formed at ca. 240 K, and then the temperature was raised to 270 K (±0.1 K). At this temperature, a 200 bar N2 pressure was applied with the high-pressure pump on the formed hexagonal ice over various duration times (from a few hours to several days) to allow the formation of the nitrogen hydrate in the SI or in the SII clathrate structure in region A in Figure 1.
ySI =
ISI ISI + ISII
(1)
Figure 3 shows the time evolution of the fraction of SI structures in the nitrogen hydrate. The transformation from SI
Figure 3. SI and SII fraction in the nitrogen hydrate at 100 K (atmospheric pressure) as a function of the duration of the reaction time (pressurization at 200 bar between N2 and powdered ice at 260 K). The dashed line represents the induction time for the SI to SII transformation. The continuous line is the fit to an exponential decaying function (see the text for details).
■
RESULTS Following the SI to SII Transformation. The time evolution of the recovered nitrogen hydrate structure is shown in Figure 2 for various durations of the reaction time at 200 bar between N2 and powdered ice at 260 K. Each neutron powder diffractogram is recorded at 100 K (atmospheric pressure) and has been indexed considering the SI (diamonds) and SII (asterisks) structures. The diffractogram recorded after 4 h exhibits only the Bragg peaks characteristic of the SI structure, while the diffractogram recorded after 3.5 days exhibits only the Bragg peaks of the SII structure. To quantify the time evolution of the fraction of each structure, the integrated intensities of the SI Bragg peak (321) at 45.02° (denoted ISI) and of the SII Bragg peak (111) at 14.1° (denoted ISII) have been extracted. The SI fraction has been determined by normalizing the diffracted intensities:
to SII exhibits an induction time of half a day, during which the SI structure remains, followed by a decaying evolution of the SI fraction. The SI fraction, ySI, can then be fitted with the help of a simple decaying exponential function: ySI = exp[−k(t − t0)]
(2)
In this equation, t0 is the induction time for the SI to SII transformation and could be assimilated to the lifetime of the SI structure. The rate associated with the SI to SII transformation is denoted k. The fit to eq 2 of the experimental data gives an excellent agreement. The induction time is measured as 0.49 ± 0.04 day (represented by a dotted line in Figure 3), and the SI− SII transformation rate is (1.37 ± 0.17) × 105 s−1 at 100 K. Such a quantitative value (corresponding to 0.63 ± 0.08 day) is C
DOI: 10.1021/acs.jpcc.7b10151 J. Phys. Chem. C XXXX, XXX, XXX−XXX
Article
The Journal of Physical Chemistry C
Figure 4. Neutron powder diffractogram of the SI (left) and SII (right) nitrogen hydrate recorded at 100 K and atmospheric pressure. The top and bottom bars represent the Bragg peak positions of ice Ih and of nitrogen hydrate (SI on the left and SII on the right), respectively. The SI and SII hydrates are obtained by recovering powdered ice pressurized at 260 K and 200 bar for 4 h and 3.5 days, respectively.
Figure 5. Temperature dependence of the SI (left) and SII (right) nitrogen hydrate cell parameter as measured by powder neutron diffraction at atmospheric pressure. The continuous lines represent the fit of the thermal expansion by using a polynomial description (see the text for details).
pressurized for 12 h at 200 bar and 260 K) between 20 and 220 K. In this case, a small fraction of SII hydrate is observed together with the SI structure (see Figure S1, Supporting Information). By considering only the SI Bragg peaks, the thermal expansion of the SI structure cell parameter has been measured as shown in Figure 4. A similar approach has been used to measure the SII expansion (Figure 4). The temperature (T) dependence has been fitted by using the polynomial expression proposed by Hester et al.:38
in full agreement with the trend previously observed by neutron diffraction experiments: the type I structure formed at 578 bar and stored at temperatures between 80 and 273 K transforms to the type II structure after a couple of days.8 Moreover, it was indicated in this study that “the lower the pressure, the higher the transformation rate”, in view of the SI−SII transformation taking a shorter time at moderate pressure (578 bar) than at a higher pressure (2500 bar). SI and SII Structural Properties. Following the previous analysis, pure SI and SII hydrates are obtained by pressurizing powdered ice (at 260 K and 200 bar) over 4 h and 3.5 days, respectively. In addition, the SII hydrate was stored at liquid nitrogen temperature and atmospheric pressure for 8 weeks to ensure a complete conversion of the initially formed SI into the SII structure. The SI hydrate was stored in the same conditions for 2 days. The measured neutron powder diffractograms are shown in Figure 3. They clearly exhibit the Bragg peaks characteristic of the SI structure (4 h of pressurization) or SII structure (3.5 days of pressurization and 8 weeks of storage) together with those of the hexagonal ice. Due to the limited pressurization time in the SI formation conditions, only a small fraction of ice is converted into the SI hydrate (4.7%). By pressurizing smaller ice particles (typically 10 μm), it is also possible to form the pure SI structure with a smaller fraction of ice.3,10 Nevertheless, such an approach requires the use of 1000 bar of nitrogen pressure, in agreement with the calculated phase diagram.17 In the present study, only 200 bar is sufficient to form the SI hydrate. The thermal expansion of the SI structure has been measured by analyzing a short reaction time sample (powdered ice
⎤ ⎡ a a a = a0 exp⎢a1(T − T0) + 2 (T − T0)2 + 3 (T − T0)3 ⎥ ⎦ ⎣ 2 3 (3)
The fitted expressions lead to a good agreement as shown in Figure 5. The cell parameter a0 has been fixed to the value measured at T0 = 20 K, i.e., a0 = 11.833 ± 0.001 Å for the SI structure and a0 = 17.068 ± 0.005 Å for the SII structure. The fitted parameters of eq 3 are a1 = (2.0 ± 0.03) × 10−5, a2 = (5.0 ± 8.0) × 10−8, and a3 = (1.1 ± 0.5) × 10−9 for the SI structure. Such a variation is in full agreement with the thermal expansion measured in various SI hydrates.21,39 In the case of the SII structure, one obtains a1 = (1.0 ± 0.1) × 10−5, a2 = (2.5 ± 0.4) × 10−7, and a3 = (1.6 ± 1.6) × 10−10, also in agreement with previous measurements.21 Finally, this experiment (in situ neutron diffraction recorded during the heating cycle at an average rate of 0.5 K/min) allows determination of the melting point of the nitrogen hydrate structure at atmospheric pressure: the melting of the SI and SII structures is observed in the ranges of 200−210 and 210−220 K, respectively, as shown by D
DOI: 10.1021/acs.jpcc.7b10151 J. Phys. Chem. C XXXX, XXX, XXX−XXX
Article
The Journal of Physical Chemistry C
Figure 6. Raman spectra of the N−N stretching modes of the nitrogen molecule confined in the SI (left) and in the SII (right) hydrates at 77 K and 1 bar. The SI and SII hydrates are obtained by recovering powdered ice pressurized at 255 K and 200 bar for 4 h and 3.5 days, respectively. The samples were stored at 77 K for 2 days (SI) and 8 weeks (SII).
and simple SC occupancy, respectively. This attribution is supported by molecular dyanamics (MD)-derived Raman spectra.23 However, in the present study, the double-occupancy Raman signature cannot be disentangled. This is probably due to the measurement performed at atmospheric pressure (i.e., with a lower cage shrinkage than at high pressure). Furthermore, according to earlier diffraction data,10 a negligible fraction of doubly occupied cavities is expected when nitrogen hydrate is formed under moderate pressure (200 bar). Thus, the vibrational contribution of double occupancy may not be observed in the present conditions. The occupancy of the cage can be estimated on the basis of the ratio I(ν1)/I(ν2) = ILC/ISC, which can be expressed as
the square displayed in Figure 1 corresponding to the average value at 210 K and 1 bar. Guest Partitioning in the SI and SII Structures. The SI and SII nitrogen hydrates have been analyzed by means of highresolution Raman scattering for identifying the spectral signatures of nitrogen encapsulations in the large cage and in the small cage. The SI and SII nitrogen hydrates were formed by following the same procedure as the one described for the neutron diffraction analysis, i.e., by pressurizing powdered ice at 255 K and 200 bar for 4 h and 3.5 days for SI and SII, respectively. The SI and SII samples were then stored at atmospheric pressure and at liquid nitrogen temperature for 2 days and 8 weeks, respectively. Figure 6 shows the N−N stretching band measured at 77 K and 1 bar. These bands are observed at a frequency (at about 2324 cm−1) significantly smaller than the nitrogen gas frequency (i.e., 2328.4 cm−1 at 200 bar and 270 K) as shown in Figure S2. Such a shift is characteristic of the confinement of nitrogen within the cage.27 Moreover, the nitrogen stretching band exhibits a profile which could not be satisfactory fitted by considering a single peak. Two peaks are required to fit the experimental spectra. In the fitting procedure, the disentangling of these two peaks has been possible by constraining the full width at half-maxima of the two peaks to be at least the spectral resolution (i.e., 0.8 cm−1). No additional fitting constraint has been applied. In such a fitting procedure, an excellent agreement is obtained as shown in Figure 6 for both SI and SII structures. In the case of the SI hydrate, the two peaks are separated by 1.1 cm−1 (ν1 = 2323.3 cm−1 and ν2 = 2324.4 cm−1) and their intensity ratio is measured as I(ν1)/I(ν2) = 3.06 ± 0.25. For the SII hydrate, the two peaks are separated by 0.8 cm−1 (ν1 = 2323.6 cm−1 and ν2 = 2324.4 cm−1) and their intensity ratio is I(ν1)/I(ν2) = 0.30 ± 0.05. Knowing that there are 3 times more LCs than SCs in the SI structure and 2 times more SCs than LCs in the SII structure, one expects the Raman band intensity due to encapsulation in LCs to be greater than that in SCs for the SI structure and the reverse for the SII structure. It follows that the lowest frequency peak is attributed to nitrogen molecules encapsulated in the LCs and the highest frequency peak to the encapsulation in the SCs. Moreover, such an attribution is similar to that made recently in the case of the isosteric carbon monoxide hydrate20 and is also consistent with previous Raman analysis made at higher pressure.29 In the latter study, performed at 2900 bar and 296 K, three hydrate-confined nitrogen Raman bands were decomposed at 2322, 2323.4, and 2325.1 cm−1 for simple LC occupancy, double SC occupancy,
2 ILC nLCθLC ⎛ αLC ⎞ = ⎜ ⎟ ISC nSCθSC ⎝ αSC ⎠
(4)
where θLC and θSC are the occupancies in LCs and in SCs, respectively. nLC and nSC are the numbers of LCs and SCs in the hydrate structure, respectively. The scattering cross sections (i.e., the polarizability variation due to the stretching mode) of nitrogen in LCs and in SCs (αLC and αSC, respectively) are assumed to be similar. Such an assumption has been validated in the isosteric SI carbon monoxide hydrate20 or in the SI methane hydrate.40,41 Thus, in the SI structure, the ratio is simply given by ILC 3θ = LC ISC θSC
(5)
and in the SII structure, one obtains ILC θ = LC ISC 2θSC
(6)
Considering the measured ratio (ILC/ISC = 3.06 ± 0.25), one obtains θLC/θSC = 1.02 ± 0.25 in the SI structure at atmospheric pressure. For the SII structure, the measured ratio ILC/ISC = 0.30 ± 0.05 leads to θLC/θSC = 0.6 ± 0.05 at atmospheric pressure. According to Rietveld refinement of in situ neutron diffraction data, the SC occupancy is 0.98 at 268 K and 1093 bar for the SI structure3 and 0.82 at 273 K and 200 bar in the SII structure.10 These results indicate that, whatever the P−T conditions are, the SC occupancy is close to 1. It follows that assuming a full occupancy of the SCs (i.e., θSC = 1), the measured θLC/θSC ratios indicate variation of the LC occupancy between the SI and SII structures: especially, the LC E
DOI: 10.1021/acs.jpcc.7b10151 J. Phys. Chem. C XXXX, XXX, XXX−XXX
Article
The Journal of Physical Chemistry C
Figure 7. Raman spectra of the N−N stretching modes of the nitrogen molecule confined in the SII along the isotherm at 150 K (left) and along the isobar at 200 bar (right). The SII hydrate was formed at 200 bar and 270 K over 3 days.
Figure 8. Temperature and pressure dependence of the ILC/ISC Raman intensity ratio (see the text) of the SII nitrogen hydrate along the isotherm at 150 K (left) and along the isobar at 200 bar (right).
to be filled and emptied by simply varying the P−T conditions, as recently observed in the carbon monoxide SI hydrate.20
occupancy in the SII structure is lower than that in the SI structure, while both SI and SII hydrates have been formed in identical pressure−temperature conditions. Such a behavior suggests a time evolution of the cage occupancy, since the Raman measurements have been performed with different reaction times (i.e., pressurization of ice) and storage times for SI and for SII. Temperature and Pressure Dependence of the Guest Partitioning in the SII Structure. The guest partitioning in the SII nitrogen hydrate has been investigated as a function of the pressure and temperature in the stability region of the hydrate as shown in Figure 7. Whatever the pressure and temperature, the Raman profiles exhibit the two characteristic modes of the SII structure. Using the fitting procedure previously described, the LC and SC Raman bands have been successfully disentangled. The frequencies of these bands do not significantly depend on the P−T conditions (see Figure S3, Supporting Information): the N2 molecule confined in LCs and in SCs are measured at 2323.9 ± 0.1 and 2324.5 ± 0.1 cm−1, respectively. The striking information resides in the ratio of these two bands as shown in Figure 8. The cage occupancy ratio given by eq 6 increases with the temperature and pressure. The isotherm measurement is in agreement with the Langmuir behavior as previously observed in the case of the SII nitrogen hydrate.8,10 The ratio rises from 0.34 ± 0.15 (i.e., θLC/θSC = 0.68) at 60 bar to 0.54 ± 0.15 (i.e., θLC/θSC = 1.08) at 200 bar along the isotherm at 150 K and from 0.54 ± 0.15 (i.e., θLC/θSC = 1.08) at 150 K to 0.95 ± 0.15 (i.e., θLC/θSC = 1.90) at 270 K along the isobar at 200 bar. Since neutron diffraction analysis10 indicates that the SC occupancy is 0.82 at 273 K and 200 bar in the SII structure, such a behavior reflects the ability of the LC
■
CONCLUDING REMARKS Raman scattering and neutron diffraction experiments were used to better understand the guest partitioning in the nitrogen gas hydrate, at investigating the impact of the pressure and temperature conditions on this partitioning, and at quantitatively analyzing the encountered structural metastability. Nitrogen hydrates were formed in the SI and SII structures by pressurizing ice at 200 bar and 270 K, and their thermal expansions have been measured. Moreover, thanks to timedependent neutron powder diffraction experiments, the metastability of the SI phase has been quantitatively analyzed: at 100 K and 1 bar, the SI structure transforms into the SII structure with a rate of (1.37 ± 0.17) × 105 s−1 and an induction time (i.e., the lifetime of pure SI structure) of 0.49 day, in agreement with the estimation made from neutron diffraction experiments performed by other groups.8 In addition, it is interesting to note that, in the case of the carbon monoxide hydrate (isosteric system), the SI to SII transformation occurs over several weeks.19,20 The dipolar moment existing in the CO caseand thus the guest−cage interactionwould strongly influence the SI to SII transformation. For both SI and SII hydrate structures, the Raman signature of nitrogen molecules encapsulated within the LCs has been measured at smaller wavenumber than that associated with the nitrogen encapsulation within the SCs. In a pressure range between 50 and 200 bar, this behavior has been observed for the first time thanks to high-resolution Raman scattering. F
DOI: 10.1021/acs.jpcc.7b10151 J. Phys. Chem. C XXXX, XXX, XXX−XXX
Article
The Journal of Physical Chemistry C
Clathrates and their Decomposition to Air Bubbles. Geophys. Res. Lett. 1996, 23, 177−180. (7) Champagnon, B.; Panczer, G.; Chazallon, B.; Arnaud, L.; Duval, P.; Lipenkov, V. Nitrogen and Oxygen Guest Molecules in Clathrate Hydrates: Different Sites Revealed by Raman Spectroscopy. J. Raman Spectrosc. 1997, 28, 711−715. (8) Kuhs, W. F.; Chazallon, B.; Radaelli, P. G.; Pauer, F. Cage Occupancy and Compressibility of Deuterated N2 - Clathrate Hydrate by Neutron Diffraction. J. Inclusion Phenom. Mol. Recognit. Chem. 1997, 29, 65−77. (9) Chazallon, B.; Champagnon, B.; Panczer, G.; Pauer, F.; Klapproth, A.; Kuhs, W. F. Micro-Raman Analysis of Synthetic Air Clathrates. Eur. J. Mineral. 1998, 10, 1125−1134. (10) Chazallon, B.; Kuhs, W. F. Situ Structural Properties of N2-, O2-, and Air-Clathrates by Neutron Diffraction. J. Chem. Phys. 2002, 117, 308−320. (11) Lunine, J. I.; Stevenson, D. J. Clathrate and Ammonia Hydrates at High Pressure: Application to the Origin of Methane on Titan. Icarus 1987, 70, 61−77. (12) Kieffer, S. W.; Lu, X.; Bethke, C. M.; Spencer, J. R.; Marshak, S.; Navrotsky, A. A Clathrate Reservoir Hypothesis for Enceladus South Polar Plume. Science 2006, 314, 1764−1766. (13) Waite, J. H., Jr.; Lewis, W. S.; Magee, B. A.; Lunine, J. I.; McKinnon, W. B.; Glein, C. R.; Mousis, O.; Young, D. T.; Brockwell, T.; Westlake, J.; et al. Liquid Water on Enceladus from Observation of Ammonia and 40Ar in the Plume. Nature 2009, 460, 487−490. (14) Rubin, M.; Altwegg, K.; Balsiger, H.; Bar-Nun, A.; Berthelier, J. J.; Bieler, A.; Bochsler, A.; Briois, C.; Calmonte, U.; Combi, M.; et al. Molecular Nitrogen in Comet 67p/Churyumov-Gerasimenko Indicates a Low Formation Temperature. Science 2015, 348, 232−235. (15) Lectez, S.; Simon, J. M.; Mousis, O.; Picaud, S.; Altwegg, K.; Rubin, M.; Salazar, J. M. A ∼ 32−70 K Formation Temperature Range for the Ice Grains Agglomerated by Comet 67 P/Churyumov− Gerasimenko. Astrophys. J., Lett. 2015, 805, L1. (16) Davidson, D. W.; Handa, Y. P.; Ratcliffe, C. I.; Tse, J. S.; Powell, B. M. The Ability of Small Molecules to Form Clathrate Hydrates of Structure II. Nature 1984, 311, 142−143. (17) Lundgaard, L.; Mollerup, J. Calculation of Phase Diagrams of Gas Hydrates. Fluid Phase Equilib. 1992, 76, 141−149. (18) van Hinsberg, M. G. E.; Scheerboom, M. I. M.; Schouten, J. A. The Vibrational Spectra of N2 in Clathrate Hydrates: a New High Pressure Phase Transition. J. Chem. Phys. 1993, 99, 752−754. (19) Zhu, J.; Du, S.; Yu, X.; Zhang, J.; Xu, H.; Vogel, S. C.; Germann, T. C.; Francisco, J. S.; Izumi, F.; Momma, K.; et al. Encapsulation Kinetics and Dynamics of Carbon Monoxide in Clathrate Hydrate. Nat. Commun. 2014, 5, 4128. (20) Petuya, C.; Damay, F.; Talaga, D.; Desmedt, A. Guest Partitioning in Carbon Monoxide Hydrate by Raman Spectroscopy. J. Phys. Chem. C 2017, 121 (25), 13798−13802. (21) Hansen, T. C.; Falenty, A.; Kuhs, W. F. Lattice Constants and Expansivity of Gas Hydrates from 10 K up to the Stability Limit. J. Chem. Phys. 2016, 144, 054301. (22) Van Klaveren, E. P.; Michels, P. J.; Schouten, J. A.; Klug, D. D.; Tse, J. S. Stability of Doubly Occupied N2 Clathrate Hydrates Investigates by Molecular Dynamics Simulations. J. Chem. Phys. 2001, 114, 5745−5754. (23) Van Klaveren, E. P.; Michels, P. J.; Schouten, J. A.; Klug, D. D.; Tse, J. S. Molecular Dynamics Simulation Study of the Properties of Doubly Occupied N2 Clathrate Hydrates. J. Chem. Phys. 2001, 115, 10500−10508. (24) Klauda, J. B.; Sandler, S. I. Phase Behavior of Clathrate Hydrates: A Model for Single and Multiple Gas Component Hydrates. Chem. Eng. Sci. 2003, 58, 27−41. (25) Schober, H.; Itoh, H.; Klapproth, A.; Chihaia, V.; Kuhs, W. F. Guest-Host Coupling and Anharmonocity in Clathrate Hydrates. Eur. Phys. J. E: Soft Matter Biol. Phys. 2003, 12, 41−49. (26) Chazallon, B.; Itoh, H.; Koza, M.; Kuhs, W. F.; Schober, H. Anharmonicity and Guest-Host Coupling in Clathrate Hydrates. Phys. Chem. Chem. Phys. 2002, 4, 4809−4816.
Moreover, the investigation of the pressure and temperature dependence of these Raman signatures in the SII structure evidences the dependence of the cage filling on the pressure− temperature conditions. As observed in the case of carbon monoxide hydrate,20 the LCs of the SII hydrate easily catch or release the nitrogen molecules, since the SCs remain singly occupied according to previous neutron diffraction analysis.10,21 As a conclusion, the present study is based on an analysis combining neutron diffraction and Raman scatteringa powerful combination for investigating the relationship between the structural properties and cage occupancy of the nitrogen gas hydrate. To the best of our knowledge, it represents the first quantitative analysis of the transformation from the kinetically favored SI structure to the thermodynamically stable SII structure. Moreover, thanks to high spectral resolution, the Raman scattering allows the variability of the cage occupancy over pressure and temperature to be clearly evidenced, due to the ability of the LCs to easily catch and release small guest molecules. It thus offers the opportunity to prepare gas hydrate with a “targeted” cage occupancy.
■
ASSOCIATED CONTENT
S Supporting Information *
The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.jpcc.7b10151. Additional neutron powder diffractogram and Raman spectra (PDF)
■
AUTHOR INFORMATION
Corresponding Author
*E-mail:
[email protected]. Phone: ++33 5 4000 2937. ORCID
Arnaud Desmedt: 0000-0002-4737-7732 Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS This paper falls in the frame of MI2C Project ANR-15CE290016 funded by the French ANR (Agence Nationale de la Recherche). All Raman experiments were performed at the platform SIV (Spectroscopie et Imagerie Vibrationnelle) at the University of Bordeaux, funded by Europe (FEDER program) and by the Region Nouvelle Aquitaine.
■
REFERENCES
(1) Sloan, E. D.; Koh, C. A. Clathrate Hydrates of Natural Gases; CRC Press: Boca Raton, FL, 2008. (2) Broseta, D., Ruffine, L., Desmedt, A., Eds; Gas Hydrates: Fundamentals, Characterization and Modeling; Wiley-ISTE: London, 2017; Vol. 1. (3) Qin, J.; Kuhs, W. F. Calibration of Raman Quantification Factors of Guest Molecules in Gas Hydrates and their Application to Gas Exchange Processes Involving N2. J. Chem. Eng. Data 2015, 60, 369− 375. (4) Miller, S. L. The Clathrate HydratesTheir Nature and Occurrence. In Physics and Chemistry of Ice; Whalley, E., Jones, S. J., Gold, L. W., Eds.; Royal Society of Canada: Ottawa, 1974; pp 42−50. (5) Pauer, F.; Kipfstuhl, J.; Kuhs, W. F. Raman Spectroscopic Study on the Nitrogen Ratio in Natural Ice Clathrates in the GRIP Ice Core. Geophys. Res. Lett. 1995, 22, 969−071. (6) Pauer, F.; Kipfstuhl, J.; Kuhs, W. F. Raman Spectroscopic Study on the Spatial Distribution of Nitrogen and Oxygen in Natural Ice G
DOI: 10.1021/acs.jpcc.7b10151 J. Phys. Chem. C XXXX, XXX, XXX−XXX
Article
The Journal of Physical Chemistry C (27) Nakahara, J.; Shigesato, Y.; Higashi, A.; Hondoh, T.; Langway, C. C., Jr Raman Spectra of Natural Clathrates in Deep Ice Cores. Philos. Mag. B 1988, 57, 421−430. (28) Van Hinsberg, M. G. E.; Schouten, J. A. The Phase Diagram of Nitrogen Clathrate Hydrate. AIP Proceedings 1993, 309, 271−274. (29) Sasaki, S.; Hori, S.; Kume, T.; Shimizu, H. Microscopic Observation and In Situ Raman Scattering Studies on High-Pressure Phase Transformations of a Synthetic Nitrogen Hydrate. J. Chem. Phys. 2003, 118, 7892−7897. (30) Jin, Y.; Kida, K.; Nagao, J. Structural Characterization of Structure H (Sh) Clathrate Hydrates Enclosing Nitrogen and 2,2Dimethylbutane. J. Phys. Chem. C 2015, 119 (17), 9069−9075. (31) Chazallon, B.; Noble, J. A.; Desmedt, A. Spectroscopy of Gas Hydrates: From Fundamental Aspects to Chemical Engineering, Geophysical and Astrophysical Applications. In Gas Hydrates: Fundamentals, Characterization and Modeling; Broseta, D., Ruffine, L., Desmedt, A., Eds.; Wiley-ISTE: London, 2017; Vol. 1, pp 63−112. (32) Desmedt, A.; Bedouret, L.; Ollivier, J.; Petuya, C. Neutron Scattering of Clathrate and Semiclathrate Hydrates. In Gas Hydrates: Fundamentals, Characterization and Modeling; Broseta, D., Ruffine, L., Desmedt, A., Eds.; Wiley-ISTE: London, 2017; Vol. 1, pp 1−62. (33) Le Bail, A. In Accuracy in Powder Diffraction II: Proceedings of the International Conference May 26−29; Prince, E., Stalick, J. K., Eds.; NIST Special Publication; National Institute of Standards and Technology (NIST): Gaithersburg, MD, 1992; Vol. 846, pp 142−153. (34) Rodriguez-Carvajal, J. FULLPROF: A Program for Rietveld Refinement and Pattern Matching Analysis. Satellite Meeting on Powder Diffraction of the XV Congress of the IUCr, Toulouse, France; International Union of Crystallography (IUCr): Chester, England, 1990; p 127. (35) Van Cleeff, A.; Diepen, G. A. M. Gas Hydrates of Nitrogen and Oxygen. Recueil 1960, 79, 582−586. (36) Marshall, D. R.; Saito, S.; Kobayashi, R. Hydrates at high pressures: Part I. Methane-Water, Argon-Water, and Nitrogen-Water Systems. AIChE J. 1964, 10, 202−205. (37) Jhaveri, J.; Robinson, D. B. Hydrates in the Methane-Nitrogen System. Can. J. Chem. Eng. 1965, 43, 75−78. (38) Yoon, J.-H.; Kawamura, T.; Ohtake, M.; Takeya, S.; Komai, T.; Yamamoto, Y.; Emi, H.; Kohara, M.; Tanaka, S.; Takano, O.; et al. Highly Selective Encaging of Carbon Dioxide Meolecules in the Mixed Carbon Dioxide and Nitrogen Hydrate at Low Temperatures. J. Phys. Chem. B 2006, 110, 17595−17599. (39) Hester, K. C.; Huo, Z.; Ballard, A. L.; Koh, C. A.; Miller, K. T.; Sloan, E. D. Thermal Expansivity for sI and sII Clathrate Hydrates. J. Phys. Chem. B 2007, 111, 8830−8835. (40) Sum, A. K.; Burruss, R. C.; Sloan, E. D. Measurement of Clathrate Hydrates via Raman Spectroscopy. J. Phys. Chem. B 1997, 101, 7371−7377. (41) Qin, J.; Kuhs, W. F. Quantitative Analysis of Gas Hydrates Using Raman Spectroscopy. AIChE J. 2013, 59, 2155−2167.
H
DOI: 10.1021/acs.jpcc.7b10151 J. Phys. Chem. C XXXX, XXX, XXX−XXX
