Fluid Composition and Kinetics of the in Situ Replacement in CH4

N2 as coolant) and a pressure of about 0.1 Pa. ..... support in framework of the SUGAR I and II (SUbmarine Gashydrat-Lagerstätten: Erkundung, Abbau u...
0 downloads 0 Views 5MB Size
Subscriber access provided by University of Otago Library

Article

Fluid Composition and Kinetics of the InSitu Replacement in CH-CO Hydrate System 4

2

Andrzej Falenty, Junfeng Qin, Andrey N Salamatin, Lei Yang, and Werner F Kuhs J. Phys. Chem. C, Just Accepted Manuscript • DOI: 10.1021/acs.jpcc.6b09460 • Publication Date (Web): 08 Nov 2016 Downloaded from http://pubs.acs.org on November 12, 2016

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

The Journal of Physical Chemistry C is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 38

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

Fluid Composition and Kinetics of the in-situ Replacement in CH4-CO2 Hydrate System A. Falenty1*, J. Qin1, A.N. Salamatin2, L.Yang1,3 and W.F. Kuhs1 1

GZG Abt. Kristallographie, Georg-August-Universität Göttingen, 37077 Göttingen, Germany 2

Dept. of Applied Mathematics, Kazan (Volga Region) Federal University, 420008 Kazan, Russia 3

Key Laboratory of Ocean Energy Utilization and Energy Conservation of Ministry of Education, Dalian University of Technology, Dalian, China

* Andrzej Falenty, [email protected], tel. +49 551-39-12546, fax. +49 551-39-95-21

ACS Paragon Plus Environment

1

The Journal of Physical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 38

Abstract

The exchange process between CO2 and methane hydrate has been observed in numerous laboratory experiments, computer simulations and recently confirmed in a field test. Yet, up to date there is no kinetic model capable of accurately predicting the swapping process at given fluid composition and p-T conditions. Major obstacles on the way to an adequate mathematical description are caused by the insufficient characterization of experimental environments and a nearly complete lack of information on the time-resolved composition of the two-phase fluid at the gas hydrate interface. Here we show that all necessary data can be provided by a combination of cryo-SEM, Raman and neutron diffraction measurements that deliver accurate spaceaveraged, time-resolved in-situ data on the CH4-CO2 exchange reactions at conditions relevant to sedimentary matrixes of continental margins. Results from diffraction are cross-correlated with ex-situ Raman spectroscopy to provide reliable information on the preferential sites for CO2 and CH4 in the (partially) exchanged hydrate. We also show a novel approach based on scattering of neutrons to probe the fluid composition during the in-situ replacement in a time-resolved, noninvasive manner. The replacement process is seen as a two-step process including 1) a fast surface reaction parallel to a fast enrichment of the surrounding fluid phase with CH4 2) followed by a much slower permeation-limited gas swapping between the gas hydrate and mixed ambient CH4-CO2 fluid. The main part of the replacement reaction takes place in the second stage. Based on our earlier experimental studies and existing literature we work towards a quantitative gas exchange model which elaborates the hole-in-cage-wall diffusion mechanism to describe the two-component gas replacement.

ACS Paragon Plus Environment

2

Page 3 of 38

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

INTRODUCTION Methane-dominated natural gas hydrates (NGH) are fairly common in nature, growing in the presence of a surplus of gas and water within a well-defined p-T zone extending globally along continental margins and in (and below) permafrost e.g.

1-4

. The estimates of the total carbon

content locked in NGH is still under a debate but generally is quoted to be at least as large as the total inventory of coal, oil and conventional gas sources altogether 1-3. Given their abundance and high gas concentration (ca. 165 STP m3 per 1 m3 of solid hydrate), gas hydrates (GH) are considered as a very potent, alternative but still not commercialized source of energy. Intensive scientific and engineering efforts over the last decade (see summary in

5-7

) nevertheless have

succeeded bringing NGH deposits to the stage of field trials. Existing exploitation ideas are currently focused around development of highly permeable gas hydrate bearing sandy sediments either via 1) destruction of clathrate crystals through depressurization, warming up or combination of both methods e.g.

5, 8-10

or 2) spontaneous exchange of guest gases between

existing gas (mainly methane) hydrate and a new clathrate forming phase e.g. 6, 11-15. The dissociation scenarios were tested first during production trials from below permafrost

16

and later from marine sediments 17 proving the large scale applicability of the technology. Even if considered successful, the trials showed that a considerable input of energy is needed not only to sustain the reaction but also to counterbalance the endothermic character of the dissociation. Moreover, a prolonged exploitation revealed a substantial sand production17 due to a loss of the

ACS Paragon Plus Environment

3

The Journal of Physical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 38

integrity of otherwise unconsolidated host rock. A continuous extraction in such case could lead to a subsidence around the borehole, destabilization of overlaying sediments, uncontrolled leakages or even trigger slope failures 5. The above-mentioned issues are largely circumvented in gas exchange scenarios that in laboratory trials showed only marginal alteration of clathrate microstructures and demonstrated a more efficient energy balance 6, 18. Safety-wise injections of pressurized CO2 have initially raised a number of concerns because of possible leak problems, similarly to Carbon Capture and Storage (CCS) technologies, but carbon dioxide can actually provide a self-sealing mechanism via formation of secondary CO2 hydrates along potential fractures

19, 20

; hence gas replacement

can be considered as a nearly emission-free technology at reduced hazards to the environment. The viability of this method has been recently confirmed in a field test where natural hydrates were exposed to CO2-N2 gas

21

. With growing concerns about climate warming due to

anthropogenic emission of CO2 and a worldwide pressing demand of energy, the gas exchange creates particularly interesting solution to both issues. The final success of the above approaches depends on the ability to correctly predict production rates from various classes of deposits

22

. While the dissociation is fairly

straightforward and can be described in terms of heat transfer

23

and permeability of the host

rock, there is no quantitative description of the exchange process. The thermodynamic basis of the gas replacement reactions is considerably well understood e.g.

6, 11, 24, 25

but identification of

kinetics-controlling parameters from available laboratory data turned out to be very difficult and only some general factors are firmly established: 1) The kinetics seems to follow a classic, two stage, shrinking-core type pattern that to some extent resembles the formation of gas hydrates from ice 26: A rapid surface reaction is followed

ACS Paragon Plus Environment

4

Page 5 of 38

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

by much slower stage in which parent- and ambient-gas molecules must be diffusively transported across the mixed gas hydrate structure 18, 25, 27-30. 2) The initial rapid exchange at the interface typically involves some partial decomposition 25, 31-34

35-37

6,

, leading to highly heterogeneous cage occupancies in the initially replaced hydrate layer

.

3) The reaction front between the original and replaced hydrate may be either sharp or diffuse. The sharp transition is proposed for the exchange with a phase transition e.g. (sI) - (sII) or (sH) - (sI)

39

35, 37, 38

. Due to a topologic incompatibility between different clathrate structures

31

the

former phase needs to be completely destroyed and reformed to the new one. The character of the reaction front for the isostructural exchange (e.g. sI CH4 - sI CO2) still remains unclear and both sharp 6, 28, 29, and diffuse 18 variants have been proposed to be validated in future. 4) Kinetics at the later stage is assumed to be governed by solid state diffusion 25, 28, 37, 40 but no details about the migration mechanisms are given. Among the various possible scenarios

26, 41

a

good candidate for this role may be a water-vacancy assisted hopping of gas molecules between cages (hole-in-cage-wall model)

42-44

recently applied to the formation of GH from ice

26, 41

.

Because of the limited mobility of guest gasses in this stage the exchange process is incomplete in most cases and can be reversed only to a certain degree 36, 37. 5) The overall fluid composition determines the concentration of guest molecules at the fluidGH interface and, thus, is directly related to the driving force for the exchange defined as the difference in chemical potentials of guest gases in the solid and fluid phase. Therefore open, flow-through systems that maintain very high concentrations of the newly applied guest show a faster conversion than samples with low permeability or closed setups where the exchanged parent gas dilutes the replacing fluid

36, 37, 45

. Moreover, injection of liquid

ACS Paragon Plus Environment

45-47

or supercritical

5

The Journal of Physical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

CO2 48, CO2 emulsions with gaseous CO2

40

14, 49

and CO2-N2 gas mixtures

50

Page 6 of 38

results in higher recovery as compared

. Nevertheless, it remains unclear if the gain in these cases is linked to the

replacement itself or is at least partly caused by surface destruction and reformation of GH. Here we extend our earlier work

18

on the most common scenario, the (sI) > (sI) transition in

the CH4-CO2 system where we use neutron diffraction to follow the replacement in-situ in a series of isobaric and isothermal experiments on large, representative samples in well-defined systems. We also present a novel noninvasive, in-situ method based on the scattering of neutrons to measure the composition of pore fluid during the exchange reaction. This information is combined with local probing by Raman and cryo scanning electron microscopy (cryo-SEM) as well as bulk structural information from neutron diffraction to provide a solid base for developing a quantitative model of the exchange process on a µm-to-mm scale. Effective guest diffusion coefficients deduced in this way can then be used in the large-scale reservoir simulations to establish strategies for a later field application.

EXPERIMENTAL BACKGROUND Starting material Deuterated CH4 hydrates used in the following experiments were grown from ice spheres formed by spraying D2O water into liquid N2 (see

51

). Recovered CH4 clathrates were crushed

under liq. N2 and sieved through a set of 200 and 300 µm meshes. In order to prevent possible contamination of deuterated hydrates with H2O frost, spraying and later crushing have been performed in a sealed glove-box under inert N2 atmosphere. The GH has been transferred to Alvials, ~0.6 g each, and divided in two groups:

ACS Paragon Plus Environment

6

Page 7 of 38

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

1) Clathrate hydrate powders with irregular particles, typical for crushed materials with some small fraction of electrostatically charged finer detritus attached to larger grains (Figure 1a). The porosity of this loosely packed powder has been established on ~55-60%. Following our earlier experience

52

, samples were preconditioned for a few hours in the pressure cell directly before

the experiment to reform a minor fraction of GH clathrate surface dissociated during the crushing 2) Powders annealed for 10 days at the melting point of D2O ice prior to the experiment to recover original GH surfaces after crushing (Figure 1b). After the treatment, powder particles reveal signs of advanced sintering with considerable smoothening of GH particles and fairly well recognizable facets of individual crystals. The annealing also reduced porosity down to 25-30%.

In-situ neutron diffraction Several hours long in-situ experiments (Table 1) were performed during two separate campaigns on the high-flux 2-axis neutron diffractometer D20 at the Institut Laue-Langevin (ILL), Grenoble, France

53

. In the first round of experiments the diffractometer for these

experiments was set to λ=2.148 Å (Ge 45° takeoff) for a fast data acquisition on the expense of some resolution. The last points from the powder diffraction series for the run 1 and 2 were measured at λ=1.536 Å (Cu). Based on this experience, during the second campaign the diffractometer was tuned to λ=2.147 Å (Ge 90° takeoff), trading some of the flux for an improved resolution at higher 2-Theta. The detailed description of the sample insertion procedure and experimental setup can be found elsewhere

35

. Temperature control has been

provided by a PID loop temperature controller attached to an “orange” He-flow cryostat with the operation range of 1.7 - 300 K. In both cases data were collected with a time step of 60 or 300 s depending on the reaction rate. Samples loaded in the pressure cell were kept for 10-20 min at

ACS Paragon Plus Environment

7

The Journal of Physical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 38

the pure CH4 gas pressure until the chosen experimental temperature became stable. Next, the fluid in pore spaces was exchanged via a series of flushings with CO2 via rapid pressure releases and recompressions. During this time p-T conditions were kept above the dissociation limit for pure CH4 hydrate. After this procedure was completed, the replacement kinetic was followed in a closed setup. Because of the usual time restrictions on neutron sources, after several hours of the reaction samples were taken out of the beam under pressure, immediately immersed in the cooling bath and reacted further until later reinsertion into a cryostat located in the beam. This approach allowed us to follow two parallel exchange reactions during each campaign in a semicontinuous manner. The ongoing exchange between the CH4 hydrate and injected CO2-fluid was followed, albeit only qualitatively, using the LAMP software 54 through changes in the intensity ratio of 321/210 Bragg reflections of the sI type hydrate that are particularly sensitive to the swapping process 18. The same software was later used to retrieve and calibrate background levels from diffraction patterns that were used in the determination of the in-situ pore fluid composition. Quantitative information on the exchange kinetics was obtained with a full pattern Rietveld refinement package – GSAS

55

– for every time frame. Good quality of the collected data

allowed for simultaneous refinement of the mixed hydrate weighted fractions α, overall scale factor, lattice constants, eight to ten Chebychev type background parameters and concentrations of CO2 and CH4 in large cages (LC-s) as functions of time. The simultaneous refinement of occupancies in large and small cages (SC-s) was generally hindered by a partial cancellation of scattering signal from both guest gases and overall low concentration of CO2 in small cavities. Consequently a stable solution to this approach, albeit at substantially decreased accuracy for SC-s, was found only for the run 2 (Fig. S2). Due to this fact the occupancy of SC-s for all

ACS Paragon Plus Environment

8

Page 9 of 38

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

experiments were fixed for all refinements at 5% guided by more accurate, later Raman analysis of recovered samples. The zero-shift was established by tuning the direct beam signal and kept constant for each insertion. Instrumental profile parameters were fitted to Na2Ca3Al2F14 standard, refined for a pure CH4 hydrate sample and kept fixed. Displacement parameters were set to values from the expansivity data on pure CH4 and CO2 hydrates 56. Strongly textured aluminum reflections and other undesired features from the pressure cell were excluded from the patterns in the least-squares refinements.

Raman spectroscopy All Raman spectra were acquired using a LabRAM HR800 (Horiba Jobin Yvon) Raman spectrometer equipped with a Peltier-cooled CCD detector and a 600 grooves/mm grating. An Ar+ laser provided a wavelength of 488 nm at the output power of 20.5 mW. The beam was focused on the sample via a 50×long-working distance objective (Olympus) and 100 µm confocal hole. Raman signals were collected in the back-scattering orientation (180°) during two accumulations of 30 s exposure time. The spectral resolution was 2.2 cm-1. Powdered hydrate samples were measured in a cooling stage (Linkam THMS600) at 113 K under the ambient pressure of nitrogen. Peaks in the regions of interest were calibrated by the Gaussian-Lorentzian mixed function, in accordance with procedures in 57.

Cryo-SEM The starting material and selected samples recovered at the end of the neutron diffraction runs were investigated with an ex-situ cryo-SEM, using a FE-SEM (FEI Quanta 200FEG) equipped with a Polaron cryo-stage. Images of surface features were taken at ~ 90 K (with liq. N2 as

ACS Paragon Plus Environment

9

The Journal of Physical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 38

coolant) and a pressure of about 0.1 Pa. The destructive power of the electron beam on uncoated samples was minimized through a low acceleration voltage of 2.5-4 keV and measurement time not exceeding 60-90 min per loading.

RESULTS AND DISCUSSION All experimental runs in this work (Table 1) represent the initial pressurization with CO2 fluid, in the “huff-and-puff” approach in which methane hydrate is exchanged at constant temperature and pressure in a closed, non-fluid-flow system with periodically injected batches of new guest gas.

Initial CH4 hydrate Size distribution and morphology (shape, surface roughness) of the starting material are two major microstructural factors controlling the replacement kinetics in pure GH systems like those investigated in our work. These two characteristics define the surface area accessible to the exchanging fluid and the total volume of particles that needs to be penetrated by diffusing molecules. Surprisingly, although these two parameters are of fundamental importance, this information is typically not well known, and, thus, numerous experimental results remain purely qualitative and cannot be compared. It is particularly relevant to larger scale reactors where any form of imaging or sample recovery is difficult. Smaller setups allow for higher flexibility and micro structural information can be readily accessed before loading via standard cryo-SEM imaging

18, 58, 59

that provides qualitative and quantitative link to observed kinetics. As we

showed above in Figure 1, depending on the preparation procedure alone, GH can exhibit vastly different morphologies. Freshly crushed, powder samples exhibit numerous sharp edges and

ACS Paragon Plus Environment

10

Page 11 of 38

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

contain small fraction of smaller detritus particles that could potently enhance the apparent specific surface area (SSA) and essentially interfere with the initial exchange kinetics. The annealed samples, on the other hand, show only connected polycrystalline particles with welldeveloped crystals that represent clear signs of an advanced coarsening 60. It should also be noted that both materials were prepared from the same spherical ice particles but the initial ice morphology, particle size and porosity have been all erased through the volume increase during the formation

26, 61

and sintering

60, 61

. For this reason properties of initial ice particles cannot be

used to elucidate any characteristics of the formed hydrate. Although not a subject of this study, it should be kept in mind that the exchange systems involving synthetic or natural mineral matrices (e.g. sand, sand and clays etc.) are even more complex as GH spatial distribution in pores, free water content and character of the GH bearing interface become also important. e.g.62, 63

. While the microscopic processes of gas exchange on a crystallite-scale proceed as described

here, the local chemical activities of the constituents at the hydrate crystallites’ surfaces will depend on mass and heat transfer processes in the porous sedimentary matrix. In such cases a realistic view of local microstructures can be provided only by in-situ methods like MRI 45, 64-66 or high resolution X-ray tomography techniques

62, 67, 68

. Depending on the chosen method, the

resolved details can be as fine as a few µm down to several hundreds of nm size.

Averaged exchange model The exchange reaction is systematically found to be locally very heterogeneous

35, 37, 65

and

therefore the determination of representative composition requires an appropriate averaging procedure over larger sample volumes. This issue is particularly related to Raman spectroscopy that remains inevitably less precise unless a large number of individual points are measured for a

ACS Paragon Plus Environment

11

The Journal of Physical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 38

better representativity. Furthermore, until recently 57, this method did not provide a reliable way to quantify the amount of exchanged gas in GH. Better averaging can be reached with NMR 69, 70 but even in this case the quantification of replaced CO2 is also problematic. The most suitable for this task are X-ray or neutron diffraction 71 that are capable of averaging in-situ over hundreds of thousands of individual crystals on a minute time scale. This was successfully demonstrated on numerous occasions in case of clathrate formation, involving structural changes

35, 38

26

decomposition,

52

and gas replacement

. Yet, in contrast to these simple cases, data treatment in

isostructural systems is more challenging. The metrics and topology of the host lattice remains nearly unchanged during the replacement 56 and the reaction can be essentially followed only by the evolution of the intensity ratio of "guest-sensitive" reflections 18. Consequently, quantitative information can be obtained only via a full pattern structure refinement. Here we employ the hole-in-cage-wall diffusion concept

42, 43

for which the replacement creates a continuous

compositional profile across the GH particle with initially high CO2 concentration near the surface and a steep gradient flattening with time (Figure 2). A direct implementation of gradient crystals with such a profile in the Rietveld refinement is very challenging but, fortunately, can be well approximated with a step function by a two-phase GH reference particle, composed of (1) the outer shell of CH4-CO2 hydrate with refined average composition of guests in LC-s and (2) the inner part to be pure CH4 hydrate; the CO2-content in this inner part is below the detection limit of our diffraction data and can be estimated from our least-squares refinements to be smaller than about 0.6%. Subsequently the overall molar fraction of the exchanged gas as a function of time can be calculated by summing up a content of gas components in both phases.

Time resolved composition of the fluid phase

ACS Paragon Plus Environment

12

Page 13 of 38

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

Exchange in porous media inevitably leads to a feedback loop in which increasing enrichment of the ambient atmosphere with the released parent guest gas results in decreasing driving force caused by the change in fluid composition. Modeling such non-equilibrium phenomenon requires time resolved information on the distribution of guest species not only in the solid but also in the fluid phase, preferably close to the exchange interface. This should not be confused with a special case of the exchange where a constant flux of fresh fluid is pumped through the sample to maintain the environmental composition at a quasi-constant level e.g.

37, 45, 48

. Either way, in all

cases the fluid composition or its time evolution must be established (measured and modelled) for correct interpretation of the observed kinetics. Although the enrichment in the parent gas phase has been frequently measured with gas chromatography (GC) e.g. 28-30, 32, 69, 72, the data are typically limited to final replacement products; a continuous sub sampling with GC without introducing perturbations to the system is possible only in flow-through systems. A non-invasive Raman spectroscopy also is not well suited for this task because of the averaging issues discussed above and a loss of resolution with penetration depth

35

. Until now the only time

resolved average fluid composition has been provided in recent MRI experiments

65

. Thus, in

spite of a comprehensive number of experiments there is actually very little known about this important kinetic factor. In our experiments we address this issue of changing driving force with a novel, non-invasive approach that allows establishing the fluid composition from the same neutron data that are taken for the structure refinement. For this purpose we make use another type of nuclear interaction with the neutron beam known as incoherent scattering that contributes in an important manner to the isotropic background level. The magnitude of this effect depends on the spin and isotope state of the elements present in the sample 73 making CO2 and CH4, respectively, very weak and

ACS Paragon Plus Environment

13

The Journal of Physical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 38

particularly strong sources of incoherent scattering. To this source of background other components must be added; an important source of background is also the coherent elastic (and inelastic) scattering form the gas molecules74 as well as multiple scattering contributions in the forward direction75. Using this unique contrast - only available in neutron scattering - it is possible to trace changes in the fluid composition during the replacement by following relative changes in the isotropic background level. The quantification of this effect requires a number of calibration steps (Supplementary Information, Appendix 1) which can be divided in two groups: 1) the separation of contributions specific to the fluid phase from the measured background and 2) scaling of the remaining background signal originating from the fluid to the number of molecules in the illuminated volume. As a result, this calibration procedure predicts the spatially averaged molar fraction of CO2 and CH4 in the illuminated part of the fluid phase for each diffraction pattern. The overall precision of this method is somewhat decreased at low molar concentrations (e.g. low porosity, low pressure) but improves at higher pressures and larger volumes. It should also be noted that, in spite of careful calibration, piling up of systematic errors may also diminish the quantification accuracy. This turns out to be particularly problematic for runs 3 and 4 where the low porosity and scatter in the structural data (low degree of replacement) significantly increases the overall measurement errors thus making the fluid analysis considerably less reliable. In our experimental setup only the pressure cell is seen by the neutron beam, which means that the averaged fluid composition at the gas-fluid interface (relevant for the driving force of exchange) and in the pore space between GH particles is deduced from the background data, and not the total gas volume of the set-up.

Exchange kinetics

ACS Paragon Plus Environment

14

Page 15 of 38

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

The experimental runs presented here in Table 1 can be divided in two separate groups aimed at different aspects of the replacement process. The first two (run 1 and 2) explore the exchange reaction at 277 K and two selected constant pressures of 3.8 and 6 MPa when pure CO2 is in gaseous and liquid state, respectively. Inversely, in the two other experiments (run 3 and 4) the samples were reacted at the same constant pressure of 5.5 MPa but at two different temperatures, 275 and 280 K when pure CO2 remains liquid. In spite of the different p-T conditions the overall exchange kinetics in all investigated cases (Figure 3) reveals very similar patterns of the two-stage reaction that was also established in earlier investigations

18, 27-30

. The initial stage is very rapid but only lasts for a short period of

time (~2-7 min.). Capturing these first moments of the exchange in our neutron diffraction data turned out to be quite challenging and could be partially seen only in experiments with the freshly crushed GH powder (runs 1, 2) (Figure 3A). Runs on annealed samples (runs 3, 4) captured only the diffusion controlled regime (Figure 3B). It is likely that the initial sampling rate of one diffraction pattern/min is too low to reliably distinguish this rapid initial exchange and can be observed only due to a fortuitous enhancement effect caused by the broad particle size distribution and surface morphology, hence the higher SSA of crushed GH. The vigorous initial exchange, possibly also combined with a partial dissolution of GH, has a twofold consequence in all studied cases: 1) rapid dilution of the pore fluid with the replacement products and 2) formation of binary hydrate at the surface of the parent CH4-hydrate. The impact on the initial dilution of the fluid is well visible on our time resolved data where the starting fluid composition falls far from pure CO2 (Figure 4). As the CO2 containing mixed-hydrate shell expands into the original GH particles the inbound and outbound transport of guest species becomes increasingly difficult until the replacement enters the purely diffusion controlled stage.

ACS Paragon Plus Environment

15

The Journal of Physical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 16 of 38

The CH4 concentration in the fluid diminishes over time as exchanged molecules are gradually carried away from the observed volume into the rest of the experimental system. The removal of the excess methane out of the experimental volume and influx of fresh CO2 from the rest of the setup is nevertheless sluggish. For runs 1 and 2 this transition is marked by a sharp slowdown of the exchange after reaching ~18-20 wt.% of the mixed phase (Figure 3A). At this point about 25-30 % of all LC-s in the mixed hydrate are occupied by CO2 (Figure 3C). The fluid composition in the pore spaces readily reacts to the reduced flux of CH4 and becomes gradually richer in CO2 (Figure 4) as the liberated CH4 is diluted into the reservoir. This process was found to be particularly slow in gaseous environment of the run 1 where the removal of the excess CH4 is likely to be hindered by a slow permeation through the network of pores. Effectively, the ongoing replacement becomes very slow. The rate of enrichment is considerably higher for run 2 at 6.0 MPa where the CO2 concentration quickly reaches ~70% and remains stable during the experiment. A continuous increase in mix phase during the permeation controlled stage for this suggests also an efficient removal of the excess CH4 from the volume illuminated by the neutron beam. This conclusion is additionally supported by the monotonous enrichment of CO2 molecules in the large cages (Figure 3C) that could not be maintained otherwise. Although, we have no direct visual observations, we suspect that once the solubility limit of CH4 in CO2 is reached the surplus of CH4 may be removed via the nucleation of CH4-bubbles that would buoyantly migrate up through the pore network, thus, preserving a good contact of liquid CO2 with clathrate particles. Some supporting argument for this scenario can be found upon a close inspection of diffraction patterns which indeed confirms a presence of a small, broad liquid CO2 diffraction peak (Figure S1) underlying the Bragg reflections from the GH phases.

ACS Paragon Plus Environment

16

Page 17 of 38

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

Samples in the run 3 and 4 enter the diffusion stage earlier and show also lower degrees of replacement. This is not entirely surprising as the annealed GH particles are likely to have a reduced contact area with the exchanging fluid. Both samples indicate also a noticeable temperature effect on the filling of large cages in the mixed phase with run 4 (280 K) showing nearly twice higher enrichment in CO2 than for run 3 (275 K) performed at somewhat lower temperature (Figure 3D). Unfortunately, the decreased porosity in these samples do not allow for a reliable deduction of the fluid composition.

Raman Spectroscopy Confocal Raman microscopy is able to detect molecular distribution within the focal spot of approximately 1 µm diameter in the lateral direction and tens of µm along the beam

76

. To

provide information on the overall-replacement reactions, thirty-five Raman spectra were recorded from CH4-CO2-exchanged hydrates recovered at the end of the experiments. Figure 5 shows three representative spectra of the CH4-CO2-exchanged hydrates. Two bands at 1275.5 cm-1 (ν-) and 1379.4 cm-1 (ν+) correspond to the lower and upper levels of Fermi resonance of CO2 molecules trapped in hydrate cavities. As can be seen, the Raman spectra significantly change, depending on the degree of the gas replacement: the intensities of CO2 bands grow up, while the intensities of CH4 bands drop down, at nearly invariable intensities of the Raman peaks of D2O host lattices (Figure 5). Cage occupancies of CH4 in LC-s (θML) and SC-s (θMS), as well as the total mean cage occupancies of CO2 molecules, θCT = (θCL×6 + θCS×2)/8, in the mixed CH4-CO2 hydrates sI are calculated as

ACS Paragon Plus Environment

17

The Journal of Physical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

θ ML

AML × 46 × FMH = H , A × 6 × FML

(1)

AMS × 46 × FMH , AH × 2 × FMS

(2)

θ MS =

θCT

AC × 46 × FCH = H , A × 8 × FC

Page 18 of 38

(3)

where A is the integrated intensity of the specific peak, and F/FH represents the relative Raman quantification factor of guest to host cavities given in 57. The contribution of residual D2O ice Ih to the O-D intensities is close to negligible as its weight percentage in the starting CH4 hydrates and CH4-CO2 hydrates after replacement estimated from Synchrotron X-ray diffraction patterns is less than 3 %. Figure 6 shows that cage occupancies of CH4 and CO2 in the mixed hydrates are scattered over a wide range; the relative cage occupancies of CH4 in LC-s and SC-s decrease with the increasing total cage occupancies of CO2. This suggests that the degree of the CO2-to-CH4 replacement reaction in different observed particles is remarkably inhomogeneous in accordance with earlier studies

13, 37

. It should be noted that the measured variability may originate not only

from a spatially inhomogeneous replacement but also from various measurement depths that assuming the “shrinking core” exchange model will produce a similar effect. CH4 molecules in LC-s are more easily displaced by CO2. It can also be seen that after the total cage occupancies of CO2 in the mixed hydrates reach ~15%, the absolute fractional filling of CH4 molecules in LC-s becomes smaller than that in SC-s (see Figure 6). In accordance to the neutron diffraction, the overall recovery of methane measured by Raman (Table 2) is found to be the highest for the

ACS Paragon Plus Environment

18

Page 19 of 38

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

Run 2 where after reaction with liq-CO2 0.17±0.12 and 0.08±0.06 of large and small cages have been exchanged. The replacement on the same type of material but with volatile CO2 (run 1) yields somewhat lower degree of replacement of 0.10±0.08 and 0.06±0.04 for LC-s and SC-s, respectively. In contrast to these two runs, annealed particles, both exposed to liq-CO2, in runs 3 and 4 show very low degree of replacement below 10% for both types of cages. These results are generally consistent with neutron diffraction data (see Figure 3E, F) but also visualize noticeable differences between the methods; characteristically for a shrinking core process spectroscopy measurements done near the surface will tend to somewhat overestimate the degree of replacement as in comparison to the volume averaged values from the neutron diffraction.

Surface morphology of partially exchanged hydrates Cryo-SEM images taken on all exchanged samples (Figure 7) reveal alterations as compared to starting powders (Figure 1). Unaltered surfaces known from the starting material are observed only in the areas mechanically fractured during the sample handling after exchange (Figure 7A). We can also clearly distinguish surfaces of samples exposed to gaseous and liquid CO2 environment. However, there are no visible volume changes that are commonly seen during the formation 26 and dissociation processes 52. Clathrate particles exchanged with CO2 gas (Figure 7A) show a strong relief that might be a response to the short period of instability triggered by the injection of CO2 gas at the beginning of the reaction. This is also in accordance with a spike in the molar fraction of CH4 in the gas atmosphere at the beginning of the reaction (Figure 4). Further dissociation would have been halted by the formation of a mixed phase hydrate. Fairly well recognizable faces of numerous individual GH crystals suggest also some limited coarsening on the particle surface during the

ACS Paragon Plus Environment

19

The Journal of Physical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 20 of 38

replacement reaction. The typical size of observed crystals is 2-5 µm in diameter. Surfaces in contact with CO2 fluid at 6 MPa (Figure 7B) reveal smoother faces covered with well distinguishable crystals reaching 5-25 µm. Although the procedures in both exchange experiments were seemingly identical we do not observe any dissociation-like features in this case. We suspect that any initial surface “etching” upon introducing liq. CO2 could be obscured by a much more advanced coarsening. This still ongoing competition between these processes might be visible on the sample reacted at 280 K 5.5 MPa (Figure 7C). Original particles annealed prior to the replacement (Figure 1B) were found to be covered with a pattern of rounded crystals of 2-5 µm in size. The appearance of microstructures of another sample, recovered after the replacement run 4 (Figure 7D) resemble those in sample 2. Also here fairly large, well distinguishable crystals reaching 5-25 µm can be observed at exposed surfaces.

Microstructural mechanism of the replacement The gas replacement is found to follow a two stage process where after a relatively short surface reaction the swapping moves to a diffusion controlled regime where most of the reaction takes place. In the first, initial stage the exchange is rapid and strongly correlated with the SSA of the parent hydrate samples as can be inferred from our experiments. The original CH4hydrate-fluid interface is likely to undergo a partial destruction due to a sudden change in the chemical potentials at the moment of a CO2 injection. Some surface melting can be also expected due to temperature perturbations and re-equilibration of water vapor with a typically dry fluid 6. Our cryo-SEM images of etched clathrate surfaces are in principle in accordance with MD simulations of the very first moments of the isostructural replacement33, 34 that suggest that the destruction is limited to the surface before a new, mixed hydrate is reformed, most likely

ACS Paragon Plus Environment

20

Page 21 of 38

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

epitaxially on the underlying crystal. The affected thickness of nanometric range reported by these calculations seems to be nevertheless underestimated; most likely a consequence of assumed boundary conditions and simulation box size. In the following step the reaction switches to much slower bulk transport of gas species through the already exchanged clathrate cages 26. Some evidences suggest that in case of (sII) > (sI) 31, (sI) > (sII) 35 as well as (sH) > (sI) 39

type the exchange is supported by the destruction of the initial structure; thus a mobilization of

gas molecules from all types of cages is possible – at least at the internal exchange / recrystallization front. For the isostructural exchange when both structures are nearly identical this may be no longer the case, and there is actually no unequivocal evidence supporting a nucleation of new crystals in the bulk. Only some surface coarsening (requiring destruction and reconstruction of cages) is supported by our SEM observations. In fact, various natural phenomena, laboratory observations e.g. 35, 41, 51, 77 and computer simulations 42-44 suggest that the guest-gas molecules can penetrate ("diffuse") through bulk hydrate without major breaking of the host lattice. Previously it was often assumed that the gas permeation mechanism was linked to polydispersity / polycrystallinity of GH (particle/grain boundaries and triple junctions), linear crystallographic defects, and other imperfections of a crystalline structure. In this case a certain, though small, amount of mobile gas molecules was envisaged to be present in the bulk of clathrate hydrates. This pseudo-solute component could be taken up directly from the environment. The above approach was phenomenologically explored in series of our publications 26, 51, 61, 78

. Although we cannot completely exclude the above scenario involving linear defects

and grain boundaries, we turn the attention to another, more physical option to describe the gas transport in clathrates as cage-to-cage diffusion of the guest molecules

26, 41

(Figure 1). This

mechanism undoubtedly needs point defects in the water framework to proceed, but leaves the

ACS Paragon Plus Environment

21

The Journal of Physical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 22 of 38

overlying crystalline structure intact. It is supported by molecular modeling techniques as reviewed and discussed in

26

. The principal conclusion is that, most probably, the gas diffusion

needs the presence of empty cages and is assisted by water vacancies in the cage walls, resulting in the so-called "hole-in-cage-wall" permeation 42, 44, 79. In CO2-hydrates (sI), around 30% of the small cages and a few percent of the large cages are empty at the typical conditions of laboratory experiments. However, small cavities in sI hydrate are not directly linked to each other. Moreover, the activation energy to hop from SC to LC was found to be on average 1.5 times higher than that of hops between LC-s 43. Therefore, it is very likely that the long-range transport in sI hydrate mainly proceeds through LC-s, while SC-s work as sinks or sources of guest-gas molecules and are not directly involved in the pathways of the diffusion process. Indeed, attempts to refine simultaneously occupancies of both LC-s and SC-s in the mixed hydrate phase have proven to be largely unsuccessful mainly due to a marginal replacement in the SC-s as shown in the only stable free refinement of run 2 (Figure S2). This has also been independently verified via ex-situ Raman spectroscopy on recovered samples. It is likely that these few percent of isolated cages are exchanged primarily during the initial surface reaction. In the latter stage small cages play only a lesser role in the diffusive long range transport that proceeds predominantly via large cavities (Figure 3 C, D). The creation of water vacancy-interstitial pairs is thought to be the rate-limiting process that controls the gas transport rates at high temperatures 41

. A model for the gas replacement in hydrates based on the hole-in-cage-wall mechanism of the

inward and outward migration of guest molecules is presently under construction.

Overall efficiency of the sI-sI exchange

ACS Paragon Plus Environment

22

Page 23 of 38

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

A comparison between various experimental groups, methods, starting materials and measurement techniques is not easy80, in particular an important information on the sample characteristics and conditions of the experiments are not fully available. The accumulated data still do not leave doubt that the replacement in sI-sI systems is generally slow on laboratory time scales (Figure 8) with a strong indication for a solid state diffusion as the main acting mechanism. To our knowledge, currently the only viable explanation of such transfer is the intrinsically slow two-directional hoping of guests through cages in the "hole-in-cage-wall" permeation scenario. Notably, these reported low exchange rates are in a stark contrast to earlier more optimistic reports15, 81 with a degree of CH4-recovery degree never reproduced by any later, comparable studies. It is likely that these cases actually do not represent a pure replacement but involve also a significant dissociation of the initial gas hydrate volume thus artificially increasing the exchange rate. The intrinsically sluggish replacement creates a serious issue to potential applications leaving only a handful of ways to accelerate the process. These aim on the maximization of the driving force via a proper choice of p-T conditions and injected fluid composition/phase. From the thermodynamic standpoint this requires an ability to maintain a very high chemical activity of replacing guest molecule at GH interface; a condition met in principle only in flow through systems with permeable samples. In industrial applications such system can be realized as a net of injection and recovery wells. Closed systems representing a one-well, huff-and-puff approach, similar to laboratory conditions considered in this paper, are not seen as competitive because of the accumulation of replaced products in the reservoir matrix that will gradually suppress the replacement process. However, in case of liq. CO2, this effect could be controlled to some extent due to a low miscibility with CH4 and bubbling-out of excess gas. An additional problem in

ACS Paragon Plus Environment

23

The Journal of Physical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 24 of 38

pressure-confined samples (inside sedimentary matrices) may be an increase hindrance to the buoyant rise of bubbles as compared to our unconfined samples. As shown in runs 3 and 4, the decreased porosity (and most likely permeability) substantially reduces the overall replacement rate even if the initial exchange fluid was liq-CO2. Once again, it should be emphasized that the lowest exchange rates are predicted in system exposed to CO2 gas.

CONCLUSIONS The isostructural gas-exchange process in clathrates is a challenging subject of fundamental and applied research. Experiments need to be performed under well-controlled conditions considering, the microstructural aspects of investigated samples and fluid composition. Up to now no quantitative model exists for predicting the exchange kinetics. Based on our experimental work and review of existing literature, we provide a solid ground under a realistic gas exchange model implementing the hole-in-cage-wall diffusion picture emerging from molecular dynamics simulations

43, 79

. Such an approach has been currently developed

41

for the

clathration reaction of gas with ice spheres as the next iteration of POWDER models 26, 41, 61, 77, 82 and, with further extension to binary diffusion, seems to be straightforwardly applicable to modeling gas exchange-reactions in sI-hydrates. This description should also account for the changing gas composition as revealed in background incoherent scattering in our neutron diffraction data.

SUPPORTING INFORMATION Figure S1 shows diffraction patterns of CH4-hydrate and CH4-CO2-hydrate plotted together with a broad diffraction peak from pure liquid CO2. Figure S2 gives the time evolution of CO2

ACS Paragon Plus Environment

24

Page 25 of 38

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

fraction in large and small cavities during the exchange at 277 K and 6 MPa. Appendix A provides the calibration procedure for the time-resolved average fluid composition retrieval from neutron scattering data.

ACKNOWLEDGEMENTS We thank the Bundesministeriums für Bildung und Forschung (BMBF) for financial support in framework of the SUGAR I and II (SUbmarine Gashydrat-Lagerstätten: Erkundung, Abbau und TRansport) projects. Furthermore, we thank Dr. Kirsten Techmer (Göttingen) for her help in the SEM sessions, the Institut Laue-Langevin (ILL) at Grenoble for beam time and support, Thomas Hansen (ILL) for very valuable experimental help and Heiner Bartels as well as Ulf Kahmann (Göttingen) for technical assistance. Neutron diffraction data for run 3 and 4 can be access though the ILL data portal under: http://doi.ill.fr/10.5291/ILL-DATA.5-25-218. Diffraction data for run 1 and 2 as well as Raman spectra for all runs are freely available upon request.

REFERENCES 1. 2. 3.

4.

5.

6. 7.

Klauda, J. B.; Sandler, S. I. Global Distribution of Methane Hydrate in Ocean Sediment. Energy Fuels 2005, 19, 459-470. Milkov, A. V. Global Estimates of Hydrate-Bound Gas in Marine Sediments: How Much Is Really out There? Earth-Sci. Rev. 2004, 66, 183-197. Pinero, E.; Marquardt, M.; Hensen, C.; Haeckel, M.; Wallmann, K. Estimation of the Global Inventory of Methane Hydrates in Marine Sediments Using Transfer Functions. Biogeosciences 2013, 10, 959-975. Kvenvolden, K. A.; Lorenson, T. D., The Global Occurrence of Natural Gas Hydrate. In Natural Gas Hydrates: Occurrence, Distribution, and Detection, Paull, C. K.; Dillon, W. P. Eds.; AGU: Washington, D. C.; 2001; Vol. 124, pp 3-18. Moridis, G. J.; Collett, T. S.; Pooladi-Darvish, M.; Hancock, S.; Santamarina, C.; Boswell, R.; Kneafsey, T.; Rutqvist, J.; Kowalsky, M. B.; Reagan, M. T., et al. Challenges, Uncertainties, and Issues Facing Gas Production from Gas-Hydrate Deposits. SPE Reserv. Eval. Eng. 2011, 14, 76112. Jung, J. W.; Espinoza, D. N.; Santamarina, J. C. Properties and Phenomena Relevant to CH4-CO2 Replacement in Hydrate-Bearing Sediments. J. Geophys. Res.-Solid Earth 2010, 115, B10102s. Collett, T.; Bahk, J. J.; Baker, R.; Boswell, R.; Divins, D.; Frye, M.; Goldberg, D.; Husebo, J.; Koh, C.; Malone, M., et al. Methane Hydrates in Nature-Current Knowledge and Challenges. J. Chem. Eng. Data 2015, 60, 319-329.

ACS Paragon Plus Environment

25

The Journal of Physical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

8. 9.

10.

11.

12.

13.

14. 15. 16.

17.

18.

19.

20.

21. 22.

23. 24. 25.

Page 26 of 38

Schicks, J. M.; Spangenberg, E.; Giese, R.; Steinhauer, B.; Klump, J.; Luzi, M. New Approaches for the Production of Hydrocarbons from Hydrate Bearing Sediments. Energies 2011, 4, 151-172. Falser, S.; Uchida, T.; Palmer, A. C.; Soga, K.; Tan, T. S. Increased Gas Production from Hydrates by Combining Depressurization with Heating of the Wellbore. Energy Fuels 2012, 26, 6259-6267. Wang, Y.; Feng, J.-C.; Li, X.-S.; Zhang, Y.; Li, G. Large Scale Experimental Evaluation to Methane Hydrate Dissociation Below Quadruple Point in Sandy Sediment. Appl. Energy 2016, 162, 372-381. Ohgaki, K.; Takano, K.; Sangawa, H.; Matsubara, T.; Nakano, S. Methane Exploitation by Carbon Dioxide from Gas Hydrates - Phase Equilibria for CO2-CH4 Mixed Hydrate System. J. Chem. Eng. Jpn. 1996, 29, 478-483. Ota, M.; Saito, T.; Aida, T.; Watanabe, M.; Sato, Y.; Smith, R. L.; Inomata, H. Macro and Microscopic CH4-CO2 Replacement in CH4 Hydrate under Pressurized CO2. AlChE J. 2007, 53, 2715-2721. Beeskow-Strauch, B.; Schicks, J. M. The Driving Forces of Guest Substitution in Gas Hydrates— a Laser Raman Study on CH4-CO2 Exchange in the Presence of Impurities. Energies 2012, 5, 420-437. Zhou, X.; Fan, S.; Liang, D.; Du, J. Replacement of Methane from Quartz Sand-Bearing Hydrate with Carbon Dioxide-in-Water Emulsion. Energy Fuels 2008, 22, 1759-1764. Lee, H.; Seo, Y.; Seo, Y. T.; Moudrakovski, I. L.; Ripmeester, J. A. Recovering Methane from Solid Methane Hydrate with Carbon Dioxide. Angew. Chem.-Int. Edit. 2003, 42, 5048-5051. Dallimore, S. R.; Collett, T. S.; Taylor, A. E.; Uchida, T.; Weber, M.; Chandra, A.; Mroz, T. H.; Caddel, E. M.; Inoue, T. Scientific Results from Japex/Jnoc/Gsc et al. Mallik 5l-38 Gas Hydrate Production Research Well, Mackenzie Delta, Northwest Territories, Canada; 2005; pp 1-140. Yamamoto, K.; Terao, Y.; Fujii, T.; Ikawa, T.; Seki, M.; Matsuzawa, M.; Kanno, T. Operational Overview of the First Offshore Production Test of Methane Hydrates in the Eastern Nankai Trough, In Offshore Technology Conference, Houston, TX, USA 5-8 May 2014; OTC25243. Falenty, A.; Murshed, M. M.; Salamatin, A. N.; Kuhs, W. F. Gas Replacement in Clathrate Hydrates During CO2 Injection - Kinetics and Micro-Structural Mechanism, In Tenth ISOPE Ocean Mining and Gas Hydrates Symposium, The International Society of Offshore and Polar Engineers (ISOPE): Szczecin, Poland, 22-26, September, 2013; pp 109-115. Kvamme, B.; Graue, A.; Buanes, T.; Kuznetsova, T.; Ersland, G. Storage of CO2 in Natural Gas Hydrate Reservoirs and the Effect of Hydrate as an Extra Sealing in Cold Aquifers. Int. J. Greenhouse Gas Control 2007, 1, 236-246. Tohidi, B.; Yang, J.; Salehabadi, M.; Anderson, R.; Chapoy, A. CO2 Hydrates Could Provide Secondary Safety Factor in Subsurface Sequestration of CO2. Environ. Sci. Technol. 2010, 44, 1509-1514. Schoderbek, D.; Boswell, R., Ignik Sikumi #1, Gas Hydrate Test Well, Successfully Installed on the Alaska North Slope. Fire ice 2011, pp 1-5. Moridis, G. J.; Collett, T. S.; Dallimore, S. R.; Satoh, T.; Hancock, S.; Weatherill, B. Numerical Studies of Gas Production from Several CH4 Hydrate Zones at the Mallik Site, Mackenzie Delta, Canada. J. Pet. Sci. Eng. 2004, 43, 219-238. Clarke, M.; Bishnoi, P. R. Determination of the Activation Energy and Intrinsic Rate Constant of Methane Gas Hydrate Decomposition. Can. J. Chem. Eng. 2001, 79, 143-147. Dornan, P.; Alavi, S.; Woo, T. K. Free Energies of Carbon Dioxide Sequestration and Methane Recovery in Clathrate Hydrates. J. Chem. Phys. 2007, 127, 124510. Kvamme, B.; Qasim, M.; Baig, K.; Kivela, P. H.; Bauman, J. Hydrate Phase Transition Kinetics from Phase Field Theory with Implicit Hydrodynamics and Heat Transport. Int. J. Greenhouse Gas Control 2014, 29, 263-278.

ACS Paragon Plus Environment

26

Page 27 of 38

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

26.

27.

28.

29. 30.

31. 32.

33. 34.

35.

36. 37.

38.

39. 40. 41. 42.

43.

44.

Falenty, A.; Salamatin, A. N.; Kuhs, W. F. Kinetics of CO2-Hydrate Formation from Ice Powders: Data Summary and Modeling Extended to Low Temperatures. J. Phys. Chem. C 2013, 117, 8443-8457. Yoon, J. H.; Kawamura, T.; Yamamoto, Y.; Komai, T. Transformation of Methane Hydrate to Carbon Dioxide Hydrate: In Situ Raman Spectroscopic Observations. J. Phys. Chem. A 2004, 108, 5057-5059. Yuan, Q.; Sun, C. Y.; Yang, X.; Ma, P. C.; Ma, Z. W.; Liu, B.; Ma, Q. L.; Yang, L. Y.; Chen, G. J. Recovery of Methane from Hydrate Reservoir with Gaseous Carbon Dioxide Using a ThreeDimensional Middle-Size Reactor. Energy 2012, 40, 47-58. Zhao, J.; Zhang, L.; Chen, X.; Fu, Z.; Liu, Y.; Song, Y. Experimental Study of Conditions for Methane Hydrate Productivity by the CO2 Swap Method. Energy Fuels 2015, 29, 6887-6895. Lee, B. R.; Koh, C. A.; Sum, A. K. Quantitative Measurement and Mechanisms for CH4 Production from Hydrates with the Injection of Liquid CO2. Phys. Chem. Chem. Phys. 2014, 16, 14922-14927. Dec, S. F. Surface Transformation of Methane-Ethane sI and sII Clathrate Hydrates. J. Phys. Chem. C 2012, 116, 9660-9665. Liu, B.; Pan, H.; Wang, X.; Li, F.; Sun, C.; Chen, G. Evaluation of Different CH4-CO2 Replacement Processes in Hydrate-Bearing Sediments by Measuring P-Wave Velocity. Energies 2013, 6, 6242-6254. Iwai, Y.; Nakamura, H.; Hirata, M. Molecular Dynamics Simulation of Replacement of Methane Hydrate with Carbon Dioxide. Mol. Simul. 2012, 38, 481-490. Bai, D. S.; Zhang, X. R.; Chen, G. J.; Wang, W. C. Replacement Mechanism of Methane Hydrate with Carbon Dioxide from Microsecond Molecular Dynamics Simulations. Energy. Environ. Sci. 2012, 5, 7033-7041. Murshed, M. M.; Schmidt, B. C.; Kuhs, W. F. Kinetics of Methane-Ethane Gas Replacement in Clathrate-Hydrates Studied by Time-Resolved Neutron Diffraction and Raman Spectroscopy. J. Phys. Chem. A 2010, 114, 247-255. Beeskow-Strauch, B.; Schicks, J. M.; Spangenberg, E.; Erzinger, J. The Influence of SO2 and NO2 Impurities on CO2 Gas Hydrate Formation and Stability. Chem. Eur. J. 2011, 17, 4376-4384. Schicks, J. M.; Luzi, M.; Beeskow-Strauch, B. The Conversion Process of Hydrocarbon Hydrates into CO2 Hydrates and Vice Versa: Thermodynamic Considerations. J. Phys. Chem. A 2011, 115, 13324-13331. Halpern, Y.; Thieu, V.; Henning, R. W.; Wang, X. P.; Schultz, A. J. Time-Resolved in Situ Neutron Diffraction Studies of Gas Hydrate: Transformation of Structure II (sII) to Structure I (sI). J. Am. Chem. Soc. 2001, 123, 12826-12831. Lee, Y.; Kim, Y.; Seo, Y. Enhanced CH4 Recovery Induced via Structural Transformation in the CH4/CO2 Replacement That Occurs in sH Hydrates. Environ. Sci. Technol. 2015, 49, 8899-8906. Ota, M.; Abe, Y.; Watanabe, M.; Smith, R. L.; Inomata, H. Methane Recovery from Methane Hydrate Using Pressurized CO2. Fluid Phase Equilib. 2005, 228, 553-559. Salamatin, A. N.; Falenty, A.; Hansen, T. C.; Kuhs, W. F. Guest Migration Revealed in CO2 Clathrate Hydrates. Energy Fuels 2015, 29, 5681-5691. Peters, B.; Zimmermann, N. E. R.; Beckham, G. T.; Tester, J. W.; Trout, B. L. Path Sampling Calculation of Methane Diffusivity in Natural Gas Hydrates from a Water-Vacancy Assisted Mechanism. J. Am. Chem. Soc. 2008, 130, 17342-17350. Demurov, A.; Radhakrishnan, R.; Trout, B. L. Computations of Diffusivities in Ice and CO2 Clathrate Hydrates via Molecular Dynamics and Monte Carlo Simulations. J. Chem. Phys. 2002, 116, 702-709. Liang, S.; Liang, D.; Wu, N.; Yi, L.; Hu, G. Molecular Mechanisms of Gas Diffusion in CO2 Hydrates. J. Phys. Chem. C 2016, 120, 16298-16304.

ACS Paragon Plus Environment

27

The Journal of Physical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

45.

46. 47. 48. 49.

50.

51.

52. 53.

54. 55. 56. 57. 58.

59.

60.

61.

62.

63.

Page 28 of 38

Graue, A.; Kvamme, B.; Baldwin, B. A.; Stevens, J.; Howard, J.; Aspenes, E.; Ersland, G.; Husebo, J.; Zornes, D. MRI Visualization of Spontaneous Methane Production from Hydrates in Sandstone Core Plugs When Exposed to CO2. SPE J. 2008, 13, 146-152. Hirohama, S.; Shimoyama, Y.; Wakabayashi, A.; Tatsuta, S.; Nishida, N. Conversion of CH4Hydrate to CO2-Hydrate in Liquid CO2. J. Chem. Eng. Jpn. 1996, 29, 1014-1020. Ota, M.; Morohashi, K.; Abe, Y.; Watanabe, M.; Smith, R. L.; Inomata, H. Replacement of CH4 in the Hydrate by Use of Liquid CO2. Energy Conv. Manag. 2005, 46, 1680-1691. Deusner, C.; Bigalke, N.; Kossel, E.; Haeckel, M. Methane Production from Gas Hydrate Deposits through Injection of Supercritical CO2. Energies 2012, 5, 2112-2140. McGrail, B. P.; Zhu, T.; Hunter, R. B.; White, M. D.; Patil, S. L.; Kulkarni, A. S. A New Method for Enhanced Production of Gas Hydrates with CO2, In AAPG Hedberg Conference: “Gas Hydrates: Energy Resource Potential and Associated Geologic Hazards”, Collett, T. S.; Johnson, A.; Knapp, C.; Boswell, R. Eds.; Vancouver, BC, Canada, September 12-16, 2004. Park, Y.; Kim, D.-Y.; Lee, J.-W.; Huh, D.-G.; Park, K.-P.; Lee, J.; Lee, H. Sequestering Carbon Dioxide into Complex Structures of Naturally Occurring Gas Hydrates. Proc. Natl. Acad. Sci. U. S. A. 2006, 103, 12690-12694. Falenty, A.; Genov, G.; Hansen, T. C.; Kuhs, W. F.; Salamatin, A. N. Kinetics of CO2 Hydrate Formation from Water Frost at Low Temperatures: Experimental Results and Theoretical Model. J. Phys. Chem. C 2011, 115, 4022-4032. Falenty, A.; Kuhs, W. F. "Self-Preservation" of CO2 Gas Hydrates-Surface Microstructure and Ice Perfection. J. Phys. Chem. B 2009, 113, 15975-15988. Hansen, T. C.; Henry, P. F.; Fischer, H. E.; Torregrossa, J.; Convert, P. The D20 Instrument at the ILL: A Versatile High-Intensity Two-Axis Neutron Diffractometer. Meas. Sci. Technol. 2008, 19, 034001s. Richard. D; Ferrand, M.; Kearley, G. J. LAMP-the Large Array Manipulation Program, Institut Laue-Langevin: Grenoble, France. Larson, A. C.; Dreele, R. B. General Structure Analysis System (GSAS); Los Alamos National Laboratory: 2000; pp 86-748. Hansen, T. C.; Falenty, A.; Kuhs, W. F. Lattice Constants and Expansivities of Gas Hydrates from 10 K up to the Stability Limit. J. Chem. Phys. 2016, 144, 054301. Qin, J. F.; Kuhs, W. F. Quantitative Analysis of Gas Hydrates Using Raman Spectroscopy. AlChE J. 2013, 59, 2155-2167. Stern, L. A.; Kirby, S. H.; Circone, S.; Durham, W. B. Scanning Electron Microscopy Investigations of Laboratory-Grown Gas Clathrate Hydrates Formed from Melting Ice, and Comparison to Natural Hydrates. Am. Mineral. 2004, 89, 1162-1175. Techmer, K. S.; Heinrichs, T.; Kuhs, W. F. Cryo-Electron Microscopic Studies of Structures and Compositions of Mallik Gas-Hydrate-Bearing Samples; Bulletin 585; Geological Survey of Canada: 2005; pp 1-12. Uchida, T.; Kishi, D.; Shiga, T.; Nagayama, M.; Gohara, K. Sintering Process Observations on Gas Hydrates under Hydrate-Stable and Self-Preservation Conditions. J. Chem. Eng. Data 2015, 60, 284-292. Staykova, D. K.; Kuhs, W. F.; Salamatin, A. N.; Hansen, T. Formation of Porous Gas Hydrates from Ice Powders: Diffraction Experiments and Multistage Model. J. Phys. Chem. B 2003, 107, 10299-10311. Chaouachi, M.; Falenty, A.; Sell, K.; Enzmann, F.; Kersten, M.; Haberthür, D.; Kuhs, W. F. Microstructural Evolution of Gas Hydrates in Sedimentary Matrices Observed with Synchrotron X-Ray Computed Tomographic Microscopy. Geochem. Geophys. 2015, 16, 1711-1722. Hauge, L. P.; Gauteplass, J.; Høyland, M. D.; Ersland, G.; Kovscek, A.; Fernø, M. A. Pore-Level Hydrate Formation Mechanisms Using Realistic Rock Structures in High-Pressure Silicon Micromodels. Int. J. Greenhouse Gas Control 2016, 53, 178-186.

ACS Paragon Plus Environment

28

Page 29 of 38

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

64. 65. 66.

67.

68.

69.

70.

71.

72. 73. 74. 75. 76. 77. 78. 79.

80.

81.

82.

Kossel, E.; Deusner, C.; Bigalke, N.; Haeckel, M. Magnetic Resonance Imaging of Gas Hydrate Formation and Conversion at Sub-Seafloor Conditions. Diffusion Fundamentals 2013, 18, 1-4. Birkedal, K. A.; Hauge, L. P.; Graue, A.; Ersland, G. Transport Mechanisms for CO2-CH4 Exchange and Safe CO2 Storage in Hydrate-Bearing Sandstone. Energies 2015, 8, 4073-4095. Ersland, G.; Husebø, J.; Graue, A.; Baldwin, B. A.; Howard, J.; Stevens, J. Measuring Gas Hydrate Formation and Exchange with CO2 in Bentheim Sandstone Using MRI Tomography. Chem. Eng. J. 2010, 158, 25-31. Kerkar, P. B.; Horvat, K.; Jones, K. W.; Mahajan, D. Imaging Methane Hydrates Growth Dynamics in Porous Media Using Synchrotron X-Ray Computed Microtomography. Geochem. Geophys. 2014, 15, 4759-4768. Yang, L.; Falenty, A.; Chaouachi, M.; Haberthür, D.; Kuhs, W. F. Synchrotron X-Ray Computed Microtomography Study on Gas Hydrate Decomposition in a Sedimentary Matrix. Geochem. Geophys. 2016, 17, 3717-3732. Cha, M.; Shin, K.; Lee, H.; Moudrakovski, I. L.; Ripmeester, J. A.; Seo, Y. Kinetics of Methane Hydrate Replacement with Carbon Dioxide and Nitrogen Gas Mixture Using in Situ NMR Spectroscopy. Environ. Sci. Technol. 2015, 49, 1964-1971. Lee, S.; Park, S.; Lee, Y.; Seo, Y. Thermodynamic and 13C NMR Spectroscopic Verification of Methane–Carbon Dioxide Replacement in Natural Gas Hydrates. Chem. Eng. J. 2013, 225, 636640. Kuhs, W. F.; Hansen, T. C., Time-Resolved Neutron Diffraction Studies with Emphasis on Water Ices and Gas Hydrates. In Neutron Scattering in Earth Sciences, Wenk, H. R. Ed.; 2006; Vol. 63, pp 171-204. Ors, O.; Sinayuc, C. An Experimental Study on the CO2–CH4 Swap Process between Gaseous CO2 and CH4 Hydrate in Porous Media. J. Pet. Sci. Eng. 2014, 119, 156-162. Sears, V. F. Neutron Scattering Lengths and Cross Sections. Neutron News 1992, 3, 26-37. Barker, J. G.; Mildner, D. F. R. Survey of Background Scattering from Materials Found in SmallAngle Neutron Scattering. J. Appl. Crystallogr. 2015, 48, 1055-1071. Tiita, A.; Tunkelo, E. Correction for Multiple Scattering in Cold Neutron Experiments. Nucl. Instrum. Methods 1972, 103, 575-580. Everall, N. J. Confocal Raman Microscopy: Common Errors and Artefacts. Analyst 2010, 135, 2512-2522. Salamatin, A. N.; Kuhs, W. F. Formation of Porous Gas Hydrates, In Proceedings of the Fourth International Conference on Gas Hydrate, Yokohama, 2002; pp 766-770. Genov, G.; Kuhs, W. F.; Staykova, D. K.; Goreshnik, E.; Salamatin, A. N. Experimental Studies on the Formation of Porous Gas Hydrates. Am. Mineral. 2004, 89, 1228-1239. Buch, V.; Devlin, J. P.; Monreal, I. A.; Jagoda-Cwiklik, B.; Uras-Aytemiz, N.; Cwiklik, L. Clathrate Hydrates with Hydrogen-Bonding Guests. Phys. Chem. Chem. Phys. 2009, 11, 1024510265. Qin, J. F.; 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. Komai, T.; Kawamura, T.; Kang, S.; Nagashima, K.; Yamamoto, Y. In-situ Observation of Gas Hydrate Behaviour under High Pressure by Raman Spectroscopy. J. Phys.-Condes. Matter 2002, 14, 11395-11400. Kuhs, W. F.; Staykova, D. K.; Salamatin, A. N. Formation of Methane Hydrate from Polydisperse Ice Powders. J. Phys. Chem. B 2006, 110, 13283-13295.

ACS Paragon Plus Environment

29

The Journal of Physical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 30 of 38

TOC Graphics

CO2 CH4

ACS Paragon Plus Environment

30

Page 31 of 38

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

A

50 µm

B

50 µm Figure 1. Cryo-SEM images of CH4-D2O clathrate starting materials: A) freshly crushed and sieved powder (200-300 µm), B) crushed clathrate powder annealed for 10 days at 276 K.

ACS Paragon Plus Environment

31

The Journal of Physical Chemistry

0

CO2 average LCCO 22

CH4

1

CO2 concentration profile

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 32 of 38

CH4 hydrate

CH4-CO2 hydrate

Figure 2. On the molecular scale (l.h.s) the exchange is envisaged to proceed via cage-to-cage transport of CO2 (in grey) from the particle surface to the core concomitant with the migration of CH4 (in green) in the opposite direction. An implementation of this approach into a crystallographic Rietveld refinement requires a division of hydrate particles in two phases with constant composition each approximating the continuous concentration changes (r.h.s). The outer shell consists of a binary CH4-CO2-hydrate. The dashed line indicates the sensitivity limit for identifying CO2 in the mixed phase (which is about 0.6 % - as obtained from the standard deviation of the least-squares estimated CO2-content of the mixed phase). The clathrate core inside this limit is assumed to be pure CH4 hydrate.

ACS Paragon Plus Environment

32

Page 33 of 38

Wt. Fraction (Mix Hydrate) [α]

0.5

0.5

Run 1: 277 K 3.8 MPa Run 2: 277 K 6 MPa

0.4

0.4

0.3

0.3

0.2

0.2

0.1

0.1

B

0.0 0

5

10

15

20

25

30

C

0.7

0

35

0.8

(LC) CO2 fraction (Mix Hydrate)

Run 3: 275 K 5.5 MPa Run 4: 280 K 5.5 MPa

A

0.0

40

60

80

100

0.5

0.5

0.4

0.4

0.3

0.3

0.2

0.2

0.1

0.1

120

D

0.7 0.6

0.0

20

0.8

0.6

0.0

0

5

10

15

20

25

30

1.00

Molar fraction of CH4 in hydrate

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

35

E

0

40

60

80

100

1.00

0.95

0.95

0.90

0.90

0.85

0.85

0.80

20

120

F

0.80 0

5

10

15

20

25

30

35

0

20

Time [h]

40

60

80

100

120

Time [h]

Figure 3. Exchange kinetics, composition of large cages (LC-s) and total degree of exchange (shown as a molar fraction of CH4 in hydrate) as a function of time. Dashed lines are guides to the eye only.

ACS Paragon Plus Environment

33

The Journal of Physical Chemistry

1.0

CO2 mole fraciton (Fluid phase)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 34 of 38

0.8 0.6 0.4 0.2 0.0 0

5

10

15

20

25

30

35

Time [h]

Figure 4. CO2 mole fractions in the pore fluid for runs 1 and 2 as a function of time

Raman Shift cm-1

Raman Shift cm-1

Figure 5. Representative Raman spectra of the CH4- CO2-exchanged hydrates (recovered at the end of the neutron experiment) in the wave-number range of (a) 1230-2730, and (b) 2830-3000 for increasing (from black to blue) in the extent of gas replacement.

ACS Paragon Plus Environment

34

Page 35 of 38

Cage occupancy of CH4, θΜ

1.0

1.0 Run 1: 277 K 3.8 MPa

0.9

0.8

0.7

0.7

0.6

0.6

0.5

LC SC

0.4 0.0

0.1

0.2

0.3

0.4

0.5

Run 2: 277 K 6 MPa

0.9

0.8

1.0 Cage occupancy of CH4, θΜ

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

0.5

LC SC

0.4 0.0

0.1

0.2

0.3

0.4

0.5

1.0 Run 3: 275 K 5.5 MPa

0.9 0.8

0.8

0.7

0.7

0.6

0.6

0.5

LC SC

0.4 0.0

0.1

0.2

0.3

0.4

0.5

Total cage occupancy of CO2, θCT

Run 4: 280 K 5.5 MPa

0.9

0.5

LC SC

0.4 0.0

0.1

0.2

0.3

0.4

0.5

Total cage occupancy of CO2, θCT

Figure 6. Cage occupancies of CH4 in LC-s (θML) and SC-s (θMS) as a function of the total mean cage occupancies of CO2 measured on partially exchanged samples recovered at the end of the neutron runs. A linear decreasing trend for both, small and large cage fillings with CH4, can be discerned as a function of increasing total CO2 content; clearly the CO2 molecules replace more efficiently CH4 molecules located in the large cages.

ACS Paragon Plus Environment

35

The Journal of Physical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

A

Page 36 of 38

C

10 µm

B

10 µm

D

10 µm

10 µm

Figure 7. SEM images of surface microstructures of partially replaced gas hydrates recovered at the end of exchange experiments of: (A) run 1, B) run 2, C) run 3, D) run 4. White arrows point to unaltered, inner surfaces freshly exposed during the sample handling/ transfer to the electron microscope.

ACS Paragon Plus Environment

36

Page 37 of 38

1.0

Molar fraction of CH4 in hydrate

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

0.8

0.6

273.2 K/3.26 MPa (v-CO2) [47]

0.4

273.2 K/3.6 MPa (Liq-CO2) [12] 281.2 K/5 MPa (CO2 emulsion) [14] 274 K/3.2 MPa (v-CO2) [8]

0.2

270 K/6 MPa (Liq-CO2) [15] 275 K/2 MPa (v-CO2flow) [81]

0

50

100

Reaction time [h]

Figure 8. Time dependent molar fraction of methane in gas hydrate during replacement for runs 1 to 4 (color coding as on Figure 3 E, F) over-plotted with results of earlier experiments in CH4 – CO2 systems.

ACS Paragon Plus Environment

37

The Journal of Physical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Temperature No. [K] 1 2 3 4

277 277 275 280

Pressure [MPa]

Duration [h]

3.8 6 5.5 5.5

24.31 30.75 108.22 94.34

Final CO2 occupancy in the CO2-CH4 hydrate Frac. LCCO2 Frac. SCCO2 0.505(53) 0.05* 0.639(62) 0.05* 0.236(58) 0.05* 0.392(49) 0.05*

Page 38 of 38

Total replacement [mol %] 7.58 11.4 2.66 5.47

Table 1. List of p-T conditions of the replacement experiments with crushed (No 1 and 2) and annealed (No 3 and 4) GH-s. The final replacement degree in cages is given for the mixed, CO2CH4 hydrate phase. The total moll % of the exchange is recalculated for the whole sample. * CO2 fraction in SC-s has been fixed at 5% with help of Raman spectroscopy.

No 1 2 3 4

Final CO2 occupancy in the CO2-CH4 hydrate Frac. LCCO2 Frac. SCCO2 0.11±0.08 0.06±0.04 0.18±0.12 0.08±0.06 0.05±0.05 0.05±0.05 0.07±0.09 0.06±0.05

Total replacement [mol %] 16±8 24±13 10±7 13±10

Table 2. Final CO2 occupancy in large and small cages and total replacement degree established with Raman spectroscopy on samples recovered at the end of the exchange experiment. The error bounds correspond to the standard deviation for all measurements.

ACS Paragon Plus Environment

38