Subscriber access provided by READING UNIV
Article
Diffusion of water through the dual porosity swelling clay mineral vermiculite Emmanuel Tertre, Sébastien SAVOYE, Fabien Hubert, Dimitri Prêt, Thomas Dabat, and Eric Ferrage Environ. Sci. Technol., Just Accepted Manuscript • DOI: 10.1021/acs.est.7b05343 • Publication Date (Web): 23 Jan 2018 Downloaded from http://pubs.acs.org on January 23, 2018
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.
Environmental Science & Technology 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 29
Environmental Science & Technology
1
Diffusion of water through the dual porosity swelling clay mineral vermiculite
2 3
Emmanuel Tertre*,a, Sebastien Savoyeb, Fabien Huberta, Dimitri Prêta, Thomas Dabata and
4
Eric Ferragea.
5 6
a
7
Bât. B8, TSA - 51106, 86073 Poitiers cedex 9, France.
8
b
9
Radionucléides, F-91191 Gif-sur-Yvette, France.
Université de Poitiers/CNRS, UMR 7285 IC2MP, Equipe HydrASA, 5 rue Albert Turpain,
CEA, DEN/DANS/DPC/SECR/Laboratoire de Mesures et Modélisation de la Migration des
10 11
*Email of the corresponding author:
12
Emmanuel Tertre (
[email protected])
13 14 15 16 17 18 19 20 21 22 23 24 25 1 ACS Paragon Plus Environment
Environmental Science & Technology
Page 2 of 29
26
Abstract
27
Prediction of water and solute migration in natural clay-based materials requires a detailed
28
understanding of the roles played by different porosity types (around or inside clay particles)
29
on the overall transfer process. For smectite, a reference material for the design of migration
30
models, this discrimination is complex due to osmotic swelling of the structure in water-
31
saturated conditions. Diffusion experiments with water tracer (HDO) were conducted on 0.1-
32
0.2, 1-2 and 10-20 µm size fractions of Na-vermiculite, a swelling clay mineral with no
33
osmotic swelling. Results obtained for the two finest fractions suggest that osmotic swelling
34
and associated impact in pore structure is responsible of the low De values reported in
35
literature for smectite compared to vermiculite. When considering only interparticle porosity,
36
De values for vermiculite are similar to those reported for non-porous grains (Na-kaolinite,
37
Na-illite). This indicates that interparticle porosity has a primary effect on the overall water
38
diffusion process, whereas interlayer porosity is shown to imply a small proportion of HDO
39
adsorption. This study provides evidence that vermiculite is a promising reference mineral for
40
the understanding of the roles played by pore structure and mineral-water interaction on
41
transport properties of water in claystones and for associated refinement of dual-porosity
42
diffusion models.
43 44 45 46 47 48 49 50 51 52 2 ACS Paragon Plus Environment
Page 3 of 29
Environmental Science & Technology
53 54
Introduction
55 56
Chemical and transport properties of natural clayey rocks, such as argillites (i.e., claystones),
57
have been widely studied in the context of the disposal of waste materials and sequestration of
58
energy resources. These media are known to have high confinement properties with respect to
59
water and solutes due to their low hydraulic conductivity1, making diffusion the main
60
transport process occurring in these systems.2,3
61
disposal, many studies have focused on obtaining macroscopic diffusion data (i.e., the
62
effective diffusion coefficient) from diffusion experiments (e.g., through-diffusion
63
experiments) performed on compacted clayey systems made either of pure minerals
64
(kaolinite, illite, montmorillonite) or complex natural clayey rocks (bentonite, argillite).
65
Swelling clay-based materials, such as argillite, characterized by the presence of smectite
66
minerals feature two types of porosity. The first type of porosity results from the crystal
67
structure of smectite particles, which are composed of stacks of solid layers separated by an
68
interlayer space. This interlayer porosity is filled with water molecules exposed to 2-
69
dimensional confinement and cations compensating the negative layer charge. The second
70
type of porosity results from the mutual arrangement of smectite particles in the porous
71
media, leading to the development of an interparticle pore network, which most often includes
72
the external diffuse double layer of particles.4 Due to the small size of clay particles, both
73
interlayer and interparticle porosities are subject to water and cationic diffusion, and in some
74
cases the interlayer porosity can be the dominant pathway at low porosity values in very dense
75
media composed of smectite particles.5 Diffusion experiment data obtained from smectite-
76
based materials have been widely used in the literature to constrain transport models based on
77
these two types of porosity (e.g., interparticle and interlayer).6-8
among others
In the context of nuclear waste
3 ACS Paragon Plus Environment
Environmental Science & Technology
Page 4 of 29
78
In addition to well-known crystal swelling due to hydration of interlayer cation9,
79
smectite minerals exhibit interlayer and interparticle osmotic swelling under water-saturated
80
conditions.10-15 The extent of swelling depends on the characteristics of the clay particles
81
(particle size and nature of the cation compensating the layer charge) and pore water
82
chemistry (especially ionic strength).16
83
exfoliation of individual layers from smectite particles, which in turn leads to an ill-defined
84
interlayer volume, and a reorganization of the medium. The presence of osmotic swelling in
85
compacted swelling clay-based materials then hampers the discrimination between interlayer
86
and interparticle porosities in the sample although diffusivities of water and solutes are
87
extremely contrasted in both types of porosity. As an illustration, several authors17,18
88
demonstrated that the mobility of water molecules in the interlayer space of smectite could be
89
at least 3 times lower than the mobility in interparticle porosity. Moreover, some authors19
90
observed a significant variation in water diffusion in smectite when varying the nature of the
91
cation compensating the layer charge and concluded that this phenomenon was related to the
92
development of a colloidal gel phase for some specific compensating cations.
and references therein
Osmotic swelling provokes the
93
In this study, diffusion experiments were carried out with vermiculite, a swelling clay
94
mineral having a structure similar to smectite, but having a higher layer charge allowing to
95
prevent osmotic swelling under water-saturated conditions.20 This lack of swelling implies
96
that the size and shape of the particles are preserved once in suspension, resulting in a well-
97
constrained interlayer porosity. In addition, this mineral forms centimeter-sized monocrystals,
98
from which different particle size fractions can be obtained.21-25 Water (HDO) diffusion
99
experiments were performed on three size fractions of vermiculite particles using a through-
100
diffusion set-up adapted from previous studies.26,27 The through-diffusion experiments were
101
performed at different degrees of compaction (i.e., porosity) in order to vary the proportion of
102
interparticle porosity to total porosity (interlayer + interparticle). The obtained experimental
4 ACS Paragon Plus Environment
Page 5 of 29
Environmental Science & Technology
103
data allow a direct evaluation of the influence of particle size on the effective diffusion
104
coefficient of water, and a comparison with values found in the literature for water diffusion
105
in smectite-rich samples is used to determine the effects of the modification of pore structure
106
by osmotic swelling. Finally, the implications of the results obtained in this study are
107
discussed in terms of predicting the diffusion of water in polymineral and natural clayey
108
samples.
109 110
Materials and methods
111 112
Sample preparation and characterization. Compacted vermiculite samples were prepared
113
from three different size fractions of particles, i.e., 0.1-0.2, 1-2 and 10-20 µm. These particles
114
were obtained using a previously established protocol based on the sonication of vermiculite
115
macro crystals (1-4 mm) immersed in water.21 The crystal chemistry and morphology of each
116
size fraction were obtained previously.21 The crystal chemistry and aspect ratio (i.e., particle
117
thickness over equivalent disk diameter) were identical for the three size fractions, while
118
specific surface areas of the external surfaces vary from ~1 m2·g-1 for the 10-20 µm size
119
fraction to ~100 m2·g-1 for the 0.1-0.2 µm size fraction.
120
Prior to the HDO diffusion experiments, the size fractions were Na-saturated using
121
five saturation cycles in a 1 mol·L-1 NaCl solution. A dialysis procedure was used to remove
122
chloride from the samples until the silver nitrate test for Cl- was negative. Then, each size
123
fraction was air-dried, sieved through a 150 µm mesh to avoid coarse aggregates and stored at
124
25°C. Drying at a higher temperature was not performed to avoid collapse of the vermiculite
125
interlayers. In its Na+-saturated form, vermiculite features interlayers displaying a bihydrated
126
(2W) state, i.e., two water layers with a layer-to-layer distance (d001) of 15 Å, at both high
127
relative humidity (>70%) and water-saturated conditions.28,29
5 ACS Paragon Plus Environment
Environmental Science & Technology
Page 6 of 29
128 129
Through-diffusion experiments. Diffusion of the water tracer (HDO) through the
130
vermiculite samples was studied using the through-diffusion technique intensively used in the
131
literature30,31 and
132
clayey samples. The set-up used in this study was adapted from the one proposed by Van
133
Loon et al.27 and is described below. According to this set-up, diffusion was studied
134
perpendicularly to the compaction plane of the particles.
references therein
to study the diffusion of water and solutes through compacted
135 136
Experimental set-up. The set-up consisted of a PEEK diffusion cell characterized by an
137
inner diameter of 9.49 mm and two aqueous reservoirs (i.e., an upstream reservoir and a
138
downstream reservoir) containing 50 mL of a 10-2 M NaCl solution. The sample was directly
139
compacted in the diffusion cell at a given bulk dry density (i.e., porosity) based on the dry
140
mass of the solid put in the cell and the volume of the cell. The typical thickness (L) used in
141
this study for the compacted samples varied from 3.4 to 8.4 mm (Table 1). The compacted
142
sample was mechanically maintained in the cell by the following components on each side of
143
the sample, listed in order from the sample to the exterior: (i) a VCWP cellulose membrane
144
with a pore size of 0
Equation 5
214
where Co is the HDO concentration in the upstream reservoir (mol·m-3) corrected from the
215
natural concentration of HDO present in ultra-pure water (see details in the Aqueous analysis
216
section) and L is the thickness of the sample. By considering these conditions, Equation 2 was
217
solved by using Equation 6, which describes the diffusive flux (i.e., J(x=L,t) in mol·s-1) in the
218
downstream reservoir, as reported by Crank35:
219
= , ) =
220
where S is the cross-sectional area perpendicular to the diffusive direction (m2). All other
221
parameters have been defined for Equations 2, 4 and 5.
222
The corresponding total amount of cumulative tracer in the downstream reservoir (i.e., n(x=L,
223
t) in mol) is as follows:
224
0 = , ) = 123 4 − − ∑&./ 5 ,
& 1 + 2 ∑&./−1) exp
6
+ & ,
+/)7 &
))
exp −
Equation 6
& ,
8
Equation 7
9 ACS Paragon Plus Environment
Environmental Science & Technology
Page 10 of 29
225
Characteristic diffusion parameters (i.e., De and α) were obtained by least-square
226
fitting of the experimental results for diffusive flux incoming in the downstream reservoir.
227
Fully analytical solutions were obtained in Laplace space and were then subsequently
228
numerically inverted to provide the solution in time, as performed by Savoye et al.36 based on
229
a previous method.37 Several authors demonstrated that the effective diffusion coefficients of
230
clayey porous media can be underestimated if diffusion through the stainless-steel filter is not
231
taken into account in the modelling procedure.38 For this reason, interpretation of the
232
experimental data obtained in this present study was performed by considering this effect.
233
Consequently, a porosity equal to 0.28 and a De of HDO equal to 2.3×10-10 m2·s-1 were used
234
for the stainless-steel filters, as proposed by several authors38,39 for these materials.
235
The uncertainties in both De and α were calculated by taking into account uncertainties
236
in the measured tracer fluxes (corresponding to the uncertainty in the concentration
237
measurements) and the thickness of the sample (an error of 0.1 mm). Finally, the uncertainty
238
in the total porosity of the samples was assessed to be ± 0.025 and was assumed to be mainly
239
due to the uncertainty associated with the sample volume (see Table 1).
240 241
Validation of the through-diffusion set-up. Although the experimental through-diffusion
242
set-up used in the present study is based on the one developed by Van Loon et al.27, it differs
243
from the latter in terms of sample size and volumes of the reservoirs. A preliminary
244
experiment was performed to validate the set-up used through the comparison of the water
245
diffusion parameters (i.e., De and α) with literature data obtained with a reference material.
246
For that purpose, the same experiment as that proposed by some authors39 using a porous
247
medium made of Na-saturated kaolinite particles (KGa-2) and compacted to a total porosity
248
(ε) of 0.26 (i.e., a dry bulk density of 1.9 g.cm-3) was performed (Table 1). Experimental data
249
(flux and cumulative amount) measured for this experiment are reported in Figure S1. 10 ACS Paragon Plus Environment
Page 11 of 29
Environmental Science & Technology
250
Obtained De and α values used to reproduce experimental data with Equation 2 were found to
251
be 1.8±0.5×10-10 m2·s-1 and 0.25±0.025, respectively (Table 1). Given the associated
252
uncertainties, the diffusion parameters obtained in the present study for a porous medium
253
composed of Na-kaolinite particles were in good agreement with those reported by Gonzalez
254
Sanchez et al.39 (De=1.34±0.15×10-10 m2·s-1 and α=0.28±0.16) and Faurel29 (De=1.4×10-10
255
m2·s-1 and α=0.25) for the same medium. Therefore, these results validate the approach used
256
in this study to acquire the diffusion parameters of vermiculite samples. Finally, the α value
257
calculated from the data is equal to the total porosity of the sample, confirming well that HDO
258
can be considered as an inert tracer (i.e., =0) for diffusion in kaolinite compacted samples.
259 260
Results and discussion
261
Figures 1A, 1B and 1C show experimental data (i.e., instantaneous fluxes and cumulative
262
amount of diffused HDO) obtained for the 0.1-0.2, 1-2 and 10-20 size fractions of Na-
263
vermiculite, respectively. The respective De and α values interpreting these data are reported
264
in Table 1.
265 266
Effect of particle size and porosity on the rock capacity factor (α α). By analyzing the α
267
values interpreting the data (Table 1), the sole experiment for which HDO can be considered
268
as a conservative tracer (inert tracer) is the one performed with the sample composed of the
269
0.1-0.2 size fraction compacted to a total porosity ε of 0.5. Indeed, this is the sole case for
270
which the data yield a α value equal to ε, indicating the absence of HDO adsorption in the
271
sample. In all other cases, the α values interpreting data are systematically higher than the
272
total porosity of the samples, indicating that HDO is adsorbed in the samples. The distribution
273
ratios () corresponding to these α values are reported in Table 1 and are between 0.1 and
274
0.5 mL·g-1. Note that uncertainties on calculated values also plead for potential HDO 11 ACS Paragon Plus Environment
Environmental Science & Technology
Page 12 of 29
275
adsorption for sample with 0.1-0.2 size fraction compacted at ε=0.5 (Table 1). All attempts to
276
derive HDO adsorption from batch experiments proved unsuccessful owing to (i) the lower
277
solid/solution ratio in batch experiments (typically 10 g·L-1) than in diffusion experiments
278
(typically 1000 g·L-1) and (ii) the high associated uncertainties in the HDO aqueous
279
concentrations measured before and after adsorption (data not shown). Although significant,
280
the overall proportion of adsorbed HDO does not exceed 0.3-2% of the total amount of
281
interlayer water molecules in the vermiculite interlayers, indicating an extremely limited
282
adsorption effect for all size fractions and ε values tested. Interestingly, although most
283
experimental studies devoted to diffusion of water tracers (mainly HTO) in compacted
284
swelling clay minerals showed that water tracers can be considered inert, a few studies have
285
previously reported a slight adsorption of HDO or HTO during diffusion. For example, some
286
authors40 reported a non-zero distribution ratio for HTO in argillite samples ( equal to 0.01
287
mL·g-1).
288
In the case of diffusion in vermiculite-based materials, the adsorption effect can be
289
tentatively assigned to the high layer charge in this swelling clay mineral and associated
290
enhancement of water confinement compared to low-charge montmorillonite. Indeed,
291
different studies using molecular simulations have evidenced the coexistence of different
292
populations of water molecules in high-charge swelling clay interlayers with contrasting
293
structural and dynamical properties.17,25,41 Note that the range of (HDO) parameters (i.e.,
294
0.1-0.5 mL/g, Table 1) used to interpret our experimental data is in agreement with the
295
(HDO) value of 0.11 mL/g which can be calculated according to HDO/H2O fractionations
296
between bulk water and water confined in mesoporous silica deduced from experimental data
297
reported in literature.42,43
298
materials made of high charge swelling clays, and likely linked to the nature of the
Such retardation of HDO (≠0) during diffusion in clayey
12 ACS Paragon Plus Environment
Page 13 of 29
Environmental Science & Technology
299
confinement, should thus be considering when interpreting mineral-water interactions on the
300
basis of isotopic analyses.
301 302
Effect of particle size and porosity on effective water diffusion coefficients (De). For the
303
two finest size fractions (i.e., 0.1-0.2 and 1-2 µm), De values are reduced when decreasing the
304
total porosity of the samples (Table 1, Figure 2A). The same correlation is also observed by
305
plotting pore diffusion coefficient values (Dp; see data analysis section for definition) as a
306
function of the total porosity (not shown). Such behavior is in agreement with literature data44
307
and references therein
308
interpreted by an increase in the tortuosity of the pathways (i.e., change of pore structure)
309
with decreasing sample porosity. This behavior can be also attributed to an increase in the
310
relative proportion of interlayer porosity when total porosity ε decreases as previously
311
mentioned for cations.5 Indeed, diffusion in the interlayers of the swelling clays is reduced
312
compared to diffusion in the interparticle porosity. Furthermore, for the same ε value between
313
~0.3 and ~0.5, the De values obtained for the size fractions 0.1-0.2 µm and 1-2 µm follow
314
similar trends. For this range of particle sizes, this behavior indicates that there is no effect of
315
the particle size on the diffusion of the water tracer in compacted swelling clays.
on diffusion in bentonites or other related montmorillonite-rich media and is
316
The De value obtained for the 10-20 µm size fraction at ε ≈ 0.35 is approximately two
317
times lower than those obtained for the finest size fractions at the same total porosity. Note
318
that a replication of the experiment led to similar value whereas the high permeability for this
319
size fraction did not allow exploring porosity value higher than 0.35 without significant
320
contribution from advective flow (results not shown). A broader distribution in particle size
321
for the 10-20 µm fraction compared to finer ones could be responsible of the observed lower
322
De values. According to a previous study23, the morphological analysis revealed a narrower
323
distribution for the 10-20 µm fraction compared to finer fractions, thus discarding a potential 13 ACS Paragon Plus Environment
Environmental Science & Technology
Page 14 of 29
324
effect from geometrical properties of particles to explain contrasted De values between the
325
different samples for a given porosity. The contrasted behavior for this sample made with the
326
coarsest size fraction could possibly be explained by variation in the particle organization.
327
Indeed, the organization of porous media composed of clayey particles can be highly
328
contrasted (ranging from isotropic to anisotropic22,23,45) due to the low aspect ratios of the
329
particles (0.1 for vermiculite particles used in this study).21 In addition, as discussed in the
330
case of the 10-20 µm size fraction used in the present study22,23, the preparation of the sample
331
can also lead to the segregation of domains with contrasting degrees of particle orientation
332
anisotropy. This anisotropic characteristic in particle orientation is known to influence the De
333
value obtained from a given sample.46-48 For instance, in clay-rich samples, previous authors47
334
have reported De values for water molecules that are from 3 to 6 times lower in the direction
335
perpendicular to the sedimentation plane than in the parallel direction. This effect is
336
interpreted by a higher tortuosity of the diffusion pathways in the direction perpendicular to
337
the sedimentation plane than in the parallel direction. However, to the best of our knowledge,
338
no experimental data has been published regarding the degree of anisotropy in clayey media
339
compacted to the same total porosity but made with different particle sizes. Van Loon et al.47
340
mentioned that the degree of anisotropy of a porous clayey media is directly related to the
341
aspect ratio of the particles and that the length of the particle governs the ability to form a
342
perfectly layered media. Although the samples of the different size fractions were prepared
343
differently in the present study and in the work of Hubert et al.22, the 10-20 µm fraction was
344
the sole sample exhibiting the coexistence of domains with isotropic and highly anisotropic
345
particle orientations.22,23 Hence, the possible contribution of a fraction of very oriented
346
particles in this sample could in turn explain the low De values. This interpretation clearly
347
highlights the need for in situ organization measurements of particle orientations as well as
348
for more in-depth theoretical analysis of the roles played by particle orientation anisotropy
14 ACS Paragon Plus Environment
Page 15 of 29
Environmental Science & Technology
349
and heterogeneity in the distribution of anisotropy within samples on the diffusional
350
properties of water.
351 352
Effect of osmotic swelling on water diffusion. For the same ε value, experimental De values
353
for water tracers reported in literature for smectitic porous media19,49,50 are systematically
354
lower than values reported for media composed of non-porous clayey particles, such as Na-
355
saturated kaolinite and illite.39, 50, this study Several of these data are reported as a function of ε
356
value in Figure 2A. This behavior is commonly attributed to the lower mobility of water
357
molecules in the vicinity of charged particle surfaces, including interlayer surfaces (which are
358
accessible to diffusion), than in the interparticle porosity.19,44,51-54
359
Comparison of the experimental data obtained from the two finest vermiculite size
360
fractions (i.e., 0.1-0.2 and 1-2 µm) and the data reported for montmorillonite with similar
361
particle sizes reveals that the De values are systematically higher in vermiculite-rich samples
362
than in smectite-rich samples for a given value of ε. This difference highlights the singular
363
behavior of vermiculite compared to smectite and likely also the role played by osmotic
364
swelling in smectitic porous media on the overall water diffusion. Indeed, the higher layer
365
charge of vermiculite should in principle slow down diffusion of water molecules in the
366
interlayer space, in agreement with molecular simulations17. For instance, several authors17,41
367
showed that the self-diffusion coefficient of water molecules in bihydrated Na+-saturated
368
smectite decreases with tetrahedral layer charge (from ~7×10-10 for a layer charge of
369
0.8/O20(OH)4 (approximately the layer charge of montmorillonite55) to ~3×10-10 m2·s-1 for a
370
layer charge of 1.8/O20(OH)4 (approximately the layer charge of vermiculite25)). Accordingly,
371
a lower water mobility in macroscopic diffusion experiments should be expected for
372
vermiculite samples compared to smectite samples (Figure 2A). Observation of the inverse
373
case likely pleads for an influence of osmotic swelling on water diffusion in smectite-based 15 ACS Paragon Plus Environment
Environmental Science & Technology
Page 16 of 29
374
materials. On the basis of HTO diffusion experiments in smectite saturated with different
375
cations, Melkior et al.19 indeed noted that the presence of colloidal gels in Na- and Ca-
376
smectite was likely responsible for the decrease in the obtained HTO diffusion coefficients
377
compared to Cs-smectite for which large aggregates of layers were observed. Such variation
378
in the same smectite sample is thus related to layer aggregation, which depends on the nature
379
of the interlayer cation and induces a change in the relative proportion of interlayer and
380
interparticle porosity. Beneficiating for the stable 2W hydrate of vermiculite once immersed
381
in water, experimental data obtained in the present study can be used to obtain additional
382
insights into the role played by different types of porosity (interlayer and interparticle) on the
383
overall water diffusion process. In Figure 2B, the obtained De values for the two finest size
384
fractions of vermiculite (i.e., 0.1-0.2 and 1-2 µm) are plotted as a function on the interparticle
385
porosity. For a given total porosity ε, the interparticle porosity εinterp. can be obtained with the
386
following equation:
387
9:&;.??
Equation 8
>.5@
390
Interestingly, the higher porosity vermiculite data points in Figure 2B are close to the
391
values reported in the literature for media composed of Na-saturated kaolinite and illite
392
particles, which are non-porous grains that exhibit only interparticle porosity. This can be
393
seen as a reminiscence of the findings obtained from numerical random-walk simulations by
394
Churakov and Gimmi.56 These authors indeed showed a limited influence of the change in the
395
amount of interlayers for a given interparticle porosity value on the overall diffusion
396
coefficient of water in very compacted clay media. In the same fashion, the results obtained in
397
the present study for vermiculite in less dense porous media thus represent experimental
398
evidence indicating that interparticle porosity is a first order parameter in the overall water 16 ACS Paragon Plus Environment
Page 17 of 29
Environmental Science & Technology
399
diffusion process compared to interlayer diffusion. Note that this situation would differ in the
400
case of cation or anion diffusion due to the contribution of the diffuse double layer at particle
401
surface that would significantly modify the dynamics depending on the charge of the clay
402
material.8,57
403 404
Implications for water and solute diffusion in natural and complex clayey environments.
405
Clayey rocks, such as argillites58, considered for the storage of nuclear waste are composed of
406
complex mineralogical assemblages. The clay mineralogy of these rocks is indeed composed
407
of non-porous grains, such as kaolinite or illite, and partially porous grains, mainly smectite or
408
illite-smectite mixed-layered minerals.59,60 Understanding the roles played by the different
409
type of minerals and porosities is thus crucial for predicting the long-term behavior of water
410
and solutes in these geological media.4,58,61 and references therein In that regard, the results obtained
411
in this study allow deriving two major conclusions.
412
First, the vermiculite used here exhibits a clear dual-porosity behavior and represents a
413
promising reference swelling clay mineral for the refinement of migration models in the
414
context of nuclear waste storage in argillite rocks including diffusive and/or advective
415
processes.62 Indeed, the smectite in these rocks is largely present as smectitic layers in illite-
416
smectite mixed-layer minerals.59,60 In water-saturated conditions, these minerals most often
417
preserve the particle morphology due to little or no layer exfoliation due to osmotic swelling.
418
Accordingly, the results obtained for vermiculite can likely be considered representative of
419
water diffusing through the smectite layers in illite-smectite minerals in argillites and could
420
help in the development of dual-porosity models for the interpretation of diffusion
421
experiments. Owing to this well-constrained porosity distribution, the diffusion data reported
422
here also represent an experimental dataset for designing simulations based on Brownian
423
dynamics or random walk displacements45,56,61,63-64 and for considering the dual-porosity
17 ACS Paragon Plus Environment
Environmental Science & Technology
Page 18 of 29
424
nature of swelling clay-based materials. Once constrained by experimental data, these
425
simulations could help to decipher the role played by different parameters (e.g., mineralogy,
426
particle size, anisotropy and porosity distribution) on the overall diffusional process of water
427
and solutes (cations and anions) occurring in compacted swelling clays.
428
Second, the small adsorption of HDO evidenced in the transient state for most of the
429
samples is probably due to the high structural layer charge of vermiculite and differs from the
430
general assumption that HDO is an inert water tracer in swelling clay minerals. Further
431
investigation regarding on the role played by layer charge and confinement on the adsorption
432
of the HDO tracer is needed in order to assess the potential implications of this adsorption on
433
the interpretation of experimental diffusion in natural clay-rich media. Indeed, clay minerals
434
in natural polyphasic rocks such as argillites commonly display noticeable chemical
435
heterogeneity, leading to a wide range of layer charges. Additional studies regarding HDO
436
adsorption in high-charge clay mineral as vermiculite could thus highlight potential mineral-
437
water interaction effects on the isotopic measurements commonly used to investigate water
438
transport in clayey matrices.
439 440
Acknowledgements
441
The French national program EC2CO “Biohefect” (Project DIFFMATARG) and the CNRS
442
interdisciplinary “défi Needs” through its “MiPor” program (Project TRANSREAC) are
443
acknowledged for providing financial support for this study. The authors would like to thank
444
Alfred Delville (ICMN, Orléans) for discussion of the dynamics of water in clayey materials
445
and Fabien Baron and Céline Boissard (IC2MP, Poitiers) for sampling assistance during the
446
diffusion experiments. The manuscript was much improved by the constructive comments of
447
four anonymous reviewers and by the associate editor Daniel Giammar.
448
18 ACS Paragon Plus Environment
Page 19 of 29
Environmental Science & Technology
449
Supporting Information
450
Data (Figure S1) obtained for kaolinite (validation of the through-diffusion set-up).
451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508
References: (1) Pusch, R. Highly compacted sodium bentonite for isolating rock-deposited radioactive-waste products. Nuclear Technology 1979, 45, 153-157. (2) Bourg, I.A.; Bourg, A.C.M.; Sposito, G. Modeling diffusion and adsorption in compacted bentonite: a critical review. J. Contam. Hydrol. 2003, 61, 293-302. (3) Charlet, L; Alt-Epping, P.; Wersin, P.; Gilbert., B. Diffusion transport and reaction in clay rocks : a storage (nuclear waste, CO2, H2), energy (shale gas) and water quality issue. Advances Water Research 2017, 106, 39-59. (4) Tournassat, C; Appelo, C.A.J. Modeling approaches for anion-exclusion in compacted Na-bentonite. Geochim. Cosmochim. Acta 2011, 75, 3698-3710. (5) Glaus, M. A.; Baeyens, B.; Bradbury, B.; Jakob, A.; Van Loon, L.R.; Yaroshchuk, A. Diffusion of 22Na and 85Sr in montmorillonite: Evidence of interlayer diffusion being the dominant pathway at high compaction. Environ. Sci. Technol. 2007, 41, 478–485. (6) Appelo, C. A. J.; Wersin, P. Multicomponent diffusion modeling in clay systems with application to the diffusion of tritium, iodide, and sodium in Opalinus clay. Environ. Sci. Technol. 2007, 41, 5002–5007. (7) Bourg, I. C.; Sposito, G.; Bourg, A. C. M. Modeling cation diffusion in compacted water-saturated sodium bentonite at low ionic strength. Environ. Sci. Technol. 2007, 41, 8118–8122. (8) Tinnacher, R.M.; Holmboe, M.; Tournassat, C.; Bourg, I.A.; Davis, J.A. Ion adsorption and diffusion in smectite: molecular, pore and continuum scale views. Geochim. Cosmochim. Acta 2016, 177, 130-149. (9) Ferrage, E. Investigation of the interlayer organization of water and ions in smectite from the combined use of diffraction experiments and molecular simulations. A review of methodology, applications, and perspectives. Clays Clay Miner. 2016, 64, 348-373. (10) Norrish, K. The swelling of montmorillonite. Discuss. Faraday Soc. 1954, 18 (0), 120–134. (11) Madsen, F.T., Müller-Vonmoos, M. The swelling behaviour of clays. Appl. Clay Sci. 1989, 4, 143–156. (12) Morvan, M.; Espinat, D.; Lambard, J.; Zemb, T. Ultrasmall-and small-angle X-ray scattering of smectite clay suspensions. Colloids Surf. A 1994, 82, 193-203. (13) Segad, M., Jonsson, B., Åkesson, T., Cabane, B. Ca/Na montmorillonite: structure, forces and swelling properties. Langmuir 2010, 26, 5782–5790. (14) Abend, S.; Lagaly, G. Sol-gel transitions of sodium montmorillonite dispersions. Appl. Clay Sci. 2000, 16, 201-227. (15) Liu, L. Prediction of swelling pressures of different types of bentonite in dilute solutions. Colloids Surf. A Physicochem. Eng. Asp. 2013, 434, 303–318. (16) Michot, L.J.; Bihannic, I.; Porsch, K.; Maddi, S.; Baravian, C.; Mougel, J.; Levitz, P. Phase diagrams of Wyoming Namontmorillonite clay. Influence of particle anisotropy. Langmuir 2004, 20, 10829-37. (17) Michot, L.J.; Ferrage, E.; Jimenez-Ruiz, M.; Boehm, M.; Delville, A. Anisotropic features of water and ion dynamics in synthetic Na- and Ca-smectites with tetrahedral layer charge. A combined quasi-elastic neutron-scattering and molecular dynamics simulations study. J. Phys. Chem. C 2012, 116, 16619-16633. (18) Porion, P.; Faugère, A.M.; Delville, A. Structural and dynamical properties of water molecules confined within clay sediments probed by deuterium NMR spectroscopy, multiquanta relaxometry and two-time stimulated echo attenuation. J. Phys. Chem. C 2014, 118, 20429-20444.
19 ACS Paragon Plus Environment
Environmental Science & Technology
509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572
Page 20 of 29
(19) Melkior, T.; Gaucher, E. C.; Brouard, C.; Yahiaoui, S.; Thoby, D.; Clinard, Ch.; Ferrage, E.; Guyonnet, D.; Tournassat, C.; Coelho, D. Na+ and HTO diffusion in compacted bentonite: Effect of surface chemistry and related texture. J. Hydrol. 2009, 370, 9-20. (20) Lagaly, G.; Dékány, I. Chapter 8 – Colloid Clay Science. In Handbook of Clay Science; Bergaya, F., Lagaly, G. Eds; Elsevier: Amsterdam 2013, pp 102. (21) Reinholdt, M.X.; Hubert, F.; Faurel, M.; Tertre, E.; Razafitianamaharavo, A.; Francius, G.; Prêt, D.; Petit, S.; Béré, E.; Pelletier, M.; Ferrage, E. Morphological properties of vermiculite particles in size-selected fractions obtained by sonication. Appl. Clay Sci. 2013, 77-78, 18-32. (22) Hubert, F.; Bihannic, I.; Prêt, D.; Tertre, E.; Nauleau, B.; Pelletier, M.; Demé, B.; Ferrage, E. Investigating the anisotropic features of particle orientation in synthetic swelling clay porous media. Clays Clay Miner. 2013, 61, 397-415. (23) Ferrage, E.; Hubert, F.; Tertre, E.; Delville, A.; Michot, L.J.; Levitz, P. Modeling the mesoscopic arrangement of particles in natural clay porous media using 3D packing of elliptic disks. Phys. Rev. E. 2015, 91, 062210-1 062210-18. (24) Dzene, L.; Tertre, E.; Hubert, F.; Ferrage, E. Nature of the sites involved in the process of cesium desorption from vermiculite. J. Colloid Interf. Sci. 2015, 455, 254–260. (25) Dzene, L.; Ferrage, E.; Hubert, F.; Delville, A.; Tertre, E. Experimental evidence of the contrasting reactivity of external vs. interlayer adsorption sites on swelling clay minerals: the Case of Sr2+-for-Ca2+ exchange in vermiculite. Appl. Clay Sci. 2016, 132-133, 205-215. (26) Oscarson, D.W. Surface diffusion: is it an important transport mechanism in compacted clays? Clays Clay Min. 1994, 42 (5), 534-543. (27) Van Loon, L.R.; Glaus, M.A.; Müller, W. Effect of confining pressure on the diffusion of HTO, 36Cl- and layered argillaceous rock (Opalinus Clay): diffusion perpendicular to the fabric. Appl. Geochem. 2003, 18, 1653.
125 -
I in a
(28) de la Calle, C. ; Suquet, H. ; Dubernat, J. ; Pezerat, H. Mode d’Empilement des feuillets dans les vermiculites hydratées à « deux couches ». Clay Min. 1978, 13, 275-297. (29) Faurel, M. Conception et mise en place d’expériences de diffusion de l’eau et de solutés dans des milieux poreux modèles d’argiles gonflantes. Ph. D. Dissertation, Poitiers, France, 2012. (30) Benning, J.L.; Barnes, D.L. Comparison of modeling methods for the determination of effective porosities and diffusion coefficients in through-diffusion tests. Water Resour. Res. 2009, 45, W09419, doi:10.1029/2008WR007236. (31) Savoye, S.; Frasca, B.; Grenut, B.; Fayette, A. How mobile is iodide in the Callovo-Oxfordian claystones under experimental conditions close to the in situ ones? J. Contam. Hydrol. 2012, 142-143, 82-92. (32) Melkior, T. Etude méthodologique de la diffusion de cations interagissants dans les argiles. Application ; mise en œuvre expérimentale et modélisation du couplage chimie-diffusion d'alcalins dans une bentonite synthétique. Ph. D. Dissertation, Ecole Centrale Paris, France, 2000. (33) Kakiuchi, M. Distribution of isotopic water molecules, H2O, HDO, and D2O, in vapor and liquid phases in pure water and aqueous solution systems. Geochim. Cosmochim. Acta 2000, 64, 1485-1492. (34) Sturm, P.; Knohl, A. Water vapor δ2H and δ18O measurements using off-axis integrated cavity output spectroscopy. Atmosph. Measurem. Tech. 2010, 3, 67-77. (35) Crank, J. The Mathematics of Diffusion, 2nd, ed.; Oxford University Press: London, 1975. (36) Savoye, S.; Beaucaire, C.; Grenut, B.; Fayette, A. Impact of the solution ionic strength on strontium diffusion through the Callovo-Oxfordian clayrocks: an experimental and modelling study. Appl. Geochim. 2015, 61, 41-52. (37) Moridis, G.J. A set of semi-analytical solution for parameter estimation in diffusion cell experiments. Report LBNL41857. Lawrence Berkeley National Laboratory, Berkeley, California, 1998. (38) Glaus, M.A.; Rossé, R.; Van Loon, L.R.; Yaroshchuk, A.E. Tracer diffusion in sintered stainless steel filters: measurement of effective diffusion coefficients and implications for diffusion studies with compacted clays. Clays Clay Min. 2008, 56, 677-685.
20 ACS Paragon Plus Environment
Page 21 of 29
573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639
Environmental Science & Technology
(39) González Sánchez, F.; Van Loon, L. R.; Gimmi, T.; Jakob, A.; Glaus, M. A.; Diamond, L. W. Self-diffusion of water and its dependence on temperature and ionic strength in highly compacted montmorillonite, illite and kaolinite. Appl. Geochem. 2008, 23, 3840-3851. (40) Van Loon, L.R.; Jakob, A. Evidence for a second transport porosity for the diffusion of tritiated water (HTO) in a sedimentary rock (Opalinus clay – OPA): application of through- and out-diffusion techniques. Transport Porous Media 2005, 61, 193-214. (41) Tertre, E.; Delville, A.; Prêt, D.; Hubert, F.; Ferrage, E. Cation diffusion in the interlayer space of swelling clay minerals - a combined macroscopic and microscopic study. Geochim. Cosmochim. Acta 2015, 149, 251–267. http://dx.doi.org/10.1016/j.gca.2014.10.011. (42) Richard, T.; Mercury, L.; Massault, M., Michelot, J.L. Experimental study of D/H isotopic fractionation factor of water adsorbed on porous silica tubes. Geochim. Cosmochim. Acta 2007, 71, 1159-1169. (43) Lin, Y; Horita, J. An experimental study on isotope fractionation in a mesoporous silica-water system with implications for vadose-zone hydrology. Geochim. Cosmochim. Acta 2016, 184, 257-271. (44) Bourg, I.C.; Tournassat, T. Self-Diffusion of water and ions in clay barriers. In Natural and Engineered Clay Barriers; Tournassat, C., Steefel, C.I., Bourg, I.C., Bergaya, F., Eds.; Elsevier: Amsterdam 2015; pp 432. (45) Tyagi, M; Gimmi, T., Churakov, S.V. Multi-scale micro-structure generation strategy for up-scaling transport in clays. Advances Water Resources 2013, 59, 181-195. (46) Sato, H.; Suzuki, S. Fundamental study on the effect of an orientation of clay particles on diffusion pathway in compacted bentonite. Applied Clay Science Clay Microstructure. Proceedings of a Workshop held in Lund 2003, Sweden, 15-17 October 2002, 23, 51-60. (47) Van Loon, L.R.; Soler, J.M., Müller, W.; Bradbury, M.H. Anisotropic diffusion in layered argillaceous rocks: a case study with Opalinus clay. Environ. Sci. Technol. 2004, 38, 5721-5728. (48) Sato, H. Effects of the orientation of smectite particles and ionic strength on diffusion and activation enthalpies of I- and Cs+ ions in compacted smectite. Appl. Clay Sci. 2005, 29, 267-281. (49) García-Gutiérrez, M.; Cormenzana, J. L.; Missana, T.; Mingarro, M. Diffusion coefficient and accessible porosity for HTO and 36Cl, compacted FEBEX bentonite. Appl. Clay Sci. 2004, 26, 65-73. (50) Glaus M. A.; Frick, S.; Rossé, R.; Van Loon, L. R. Comparative study of tracer diffusion of HTO, compacted kaolinite, illite and montmorillonite. Geochim. Cosmochim. Acta 2010, 74, 1999-2010.
22
Na+ and
36
Cl− in
(51) Kemper, W.D.; Maasland, D.E.L.; Porter, L.K. Mobility of water adjacent to mineral surfaces. Soil Sci. Society America Proceedings 1964, 28, 164-167. (52) Duval, F.P., Porion, P.; Van Damme, H. Microscale and macroscale diffusion of water in colloidal gels. A pulsed field gradient and NMR imaging investigation. J. Phys. Chem. B 1999, 103, 5730-5735. (53) Nakashima, Y.; Mitsumori, F. H2O self-diffusion restricted by clay platelets with immobilized bound H2O layers: PGSE NMR study of water-rich saponite gels. Appl. Clay Sci. 2005, 28, 209-221. (54) Tournassat, C.; Steefel, C.I. Ionic transport in nano-porous clays with consideration of electrostatics effects. Rev. Mineral. Geochem. 2015, 80, 287-329. (55) Mermut, A.R.; Lagaly, G. Baseline Studies of the Clay Minerals Society Source Clays: Layer-Charge Determination and Characteristics of those Minerals Containing 2:1 Layers. Clays Clay Min. 2001, 49, 393-397. (56) Churakov, S. V.; Gimmi, T. Up-scaling of molecular diffusion coefficients in clays: A two-step approach. J. Phys. Chem. C 2011, 115, 6703–6714. (57) Bourg, I.C; Sposito, G. Molecular dynamic simulations of the electrical double layer on smectite surfaces contacting concentrated mixed electrolyte (NaCl-CaCl2) solutions. J. Colloid Interf. Sci. 2011, 360, 701-715. (58) Altmann, S.; Tournassat, C.; Goutelard, F.; Parneix, J.C.; Gimmi, T.; Maes, N. Diffusion-driven transport in clayrock formations. Appl. Geochem. 2012, 27, 463-478. (59) Claret, F.; Sakharov, B.A.; Drits, V.A.; Velde, B.; Meunier, A.; Griffault, L.; Lanson, B. Clay minerals in the MeuseHaute Marne underground laboratory (France): possible influence of organic matter on clay mineral evolution. Clays Clay Min. 2004, 52 (5), 515-532.
21 ACS Paragon Plus Environment
Environmental Science & Technology
640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660
Page 22 of 29
(60) Beaucaire, C.; Tertre, E.; Ferrage, E.; Grenut, B.; Pronier, S.; Madé, B. A thermodynamic model for the prediction of pore water composition of clayey rock at 25 and 80°C – Comparison with results from hydrothermal alteration experiments. Chem. Geol. 2012, 334, 62-76. (61) Robinet, J.-C.; Sardini, P.; Coelho, D.; Parneix, J.-C.; Prêt, D.; Sammartino, S.; Boller, E.; Altmann, S. Effects of mineral distribution at mesoscopic scale on solute diffusion in a clay-rich rock: Example of the Callovo-Oxfordian mudstone (Bure, France). Water Resour. Res. 2012, 48, 5554-5571. (62) Bourg, I.C.; Ajo-Franklin, J.B. Clay, water and salt: controls on the permeability of fine-grained sedimentary rocks. Accounts Chem. Res. 2017, 50, 2067-2074. (63) Churakov, S. V.; Gimmi, T.; Unruh, T.; Van Loon, L. R.; Juranyi, F. Resolving diffusion in clay minerals at different timescales: Combination of experimental and modeling approaches. Appl. Clay Sci. 2014, 96, 36–44. (64) Bacle, P.; Dufrêche, J.F.; Rotenberg, B.; Bourg, I.C.; Marry, V. Modeling the transport of water and ionic tracers in a micrometric clay sample. Appl. Clay Sci. 2016, 123, 18-28. (65) Hassan, M.S.; Villieras, F.; Gaboriaud, F.; Razafitianamaharavo, A. AFM and low-pressure argon adsorption analysis of geometrical properties of phyllosilicates. J. Colloid Interf. Sci. 2006, 296, 614-623.
22 ACS Paragon Plus Environment
Page 23 of 29
Environmental Science & Technology
Figure captions:
Figure 1: Instantaneous flux and cumulative total amount of HDO diffusing through porous media composed of the (A) 0.1-0.2 µm, (B) 1-2 µm, and (C) 10-20 µm size fractions of Navermiculite. Data are plotted as a function of time and for different total porosities of the media. Symbols represent experimental data, while full lines correspond to the fit of the experimental flux. Dotted lines are simulated flux curves calculated by considering the error range of diffusion parameters (see Table 1). Simulated cumulative curves calculated by considering the diffusion parameters interpreting average experimental flux are reported in dashed lines.
Figure 2: Comparison between effective diffusion coefficients (De) reported in the literature and those obtained in this present study for clayey porous media composed of porous (i.e., Na-montmorillonite, Na-smectite-rich materials and Na-vermiculite) and non-porous grains (i.e., Na-kaolinite and Na-illite). (A) As a function of the total porosity ε. (B) As a function of the interparticle porosity εinterp. only (see Table 1 for details).
Table caption:
Table 1: Summary of the diffusion experiments, including characteristics of the compacted sample and diffusion parameters (effective diffusion coefficient and rock capacity factor) interpreting experimental data. The uncertainty ranges for the diffusion parameters are written in brackets (see text for details). 23 ACS Paragon Plus Environment
Environmental Science & Technology
Page 24 of 29
24 ACS Paragon Plus Environment
Page 25 of 29
Environmental Science & Technology
Table 1 De -11 (10 m2.s-1)d
d
Rdd (mL.g-1)
amount of adsorbed HDO -5 (10 mol.g-1)
18 (14-24) 26 (19-42)
0.25 (0.2-0.3) 0.5 (0.5-0.8)
0 (0-0.08) 0 (0-0.22)
0 (0-2.1) 0 (0-6.0)
0.557
15.2 (10.5-19)
0.89 (0.6-0.9)
0.30 (0.12-0.31)
8.1 (3.2-8.3)
0
0.487
9.7 (7-14.5)
0.25 (0.13-0.32)
5.8 (3.0-7.4)
0.50
0.25
0.539
38.5 (25-55)
0.77 (0.550.9) 1.1 (1.0-1.2)
0.44 (0.37-0.52)
11.6 (9.7-13.5)
5.1
0.31
0
0.472
10.2 (6-17)
0.75 (0.5-0.9)
0.24 (0.10-0.32)
5.4 (2.3-7.2)
6.8
0.35
0.03
0.555
3.58 (2.5-4.5)
0.86 (0.60.86)
0.29 (0.14-0.29)
7.8 (3.8-7.8)
Nature of the clay
size fraction (µm)
thickness of compacted sample L (mm) ± 0.1
total porosity a ± 0.025
Nakaolinite
≈ 0.1-1e
6.0
0.26
0.26
0.555
0.1-0.2
8.4
0.50
0.25
0.547
0.1-0.2
7.0
0.41
0.12
0.1-0.2
3.4
0.31
1-2
6.1
1-2 10-20
Navermiculite
interparticle [HDO] porosity concentration in the upinterp.b stream ± 0.025 reservoir (M)c
1 ACS Paragon Plus Environment
Environmental Science & Technology
a
calculated as 𝜀 = 1 −
𝜌𝑎𝑝𝑝. 𝜌𝑠
Page 26 of 29
where s is the real grain density (2.6 for kaolinite according to Hassan et al.65 and 2.7 g.cm-3 for vermiculite as calculated by Reinholdt et al.21 on
the basis of the structural formula). app. is the bulk dry density calculated considering the volume of diffusion cell and the mass of samples measured at 25°C and corrected from mass of interlayer water of Na-vermiculite at room humidity (approximatively 10%) as performed in literature.24 b calculated by considering Equation 8 (see text). c note that the maximum decrease of the HDO concentration in the upstream reservoir due to diffusion is around 3% all along an experiment allowing to assume a constant gradient between two aqueous reservoirs for modelling. d see equation 2 in the text. e from 65
2 ACS Paragon Plus Environment
Page 27 of 29 Figure 1
Environmental Science & Technology
A. 0.1-0.2µm size fraction
6.E-05
6.E-04
4.E-05
4.E-04
flux cumulative amount
2.E-04
0.E+00 4
6
8
6.E-04 4.E-05 4.E-04
2.E-05
flux cumulative amount
2
4
porosity e = 0.50
12
2.0E-04
flux cumulative amount
0.0E+00
0.0E+00
4
6
instantaneous flux (mol/day)
5.0E-05
4.0E-05 4.0E-04
flux cumulative amount
2.0E-04
0.0E+00
0.0E+00 8
12
ACS Paragon Environment timePlus (days)
4.E-04
2.0E-05
flux
2.E-04
cumulative amount
0.E+00 0
2
4
6
8
10
12
16
20
porosity e = 0.35
2.0E-05
1.0E-03
6.0E-04
4
4.0E-05
C. 10-20µm size fraction 8.0E-04
0
6.E-04
time (days)
6.0E-05
2.0E-05
6.0E-05
1.6E-04
1.5E-05
1.2E-04
1.0E-05
8.0E-05
5.0E-06
4.0E-05
flux
cumulative amount
0.0E+00
0.0E+00 0
2
4
6
time (days)
8
10
cumulative total amount (mol)
4.0E-04
8.E-04
14
cumulative total amount (mol)
1.0E-04
time (days)
10
porosity e = 0.31
8.0E-05
cumulative total amount (mol)
instantaneous flux (mol/day)
8.0E-04
6.0E-04
2
8
time (days)
B. 1-2µm size fraction
0
6
8.0E-05
0.0E+00
0.E+00
0
10
time (days)
1.5E-04
2.E-04
instantaneous flux (mol/day)
2
8.E-04 6.E-05
0.E+00
0.E+00 0
1.E-03
instantaneous flux (mol/day)
8.E-04
instantaneous flux (mol/day)
8.E-05
8.E-05
cumulative total amount (mol)
1.E-03
cumulative total amount (mol)
1.E-04
2.E-05
porosity e = 0.31
porosity e = 0.41 cumulative total amount (mol)
instantaneous flux (mol/day)
porosity e = 0.50
Environmental Science & Technology
Page 28 of 29
Figure 2 A
B Na-kaolinite – Na-illite – 0.1/0.2 and 1/2 size fractions of Na-vermiculite
0.1/0.2 and 1/2 size fractions of Na-vermiculite 1.0E-09
1.0E-09
Na-montmorillonite ; Glaus et al., 2010 (50)
Na-kaolinite; Gonzalez Sanchez et al., 2008 (39)
Na-kaolinite – Na-illite
Febex bentonite; Garcia Gutirerez et al., 2004 (49) Na-bentonite ; Melkior et al., 2009 (19)
Na-kaolinite; this study
Na-illite ; Glaus et al., 2010 (50)
Na-kaolinite; this study
De (m²/s)
De (m²/s)
Na-kaolinite; Gonzalez Sanchez et al., 2008 (39)
1.0E-10 Na-illite ; Glaus et al., 2010 (50)
1.0E-10 Na-illite ; Gonsalez Sanchez et al., 2008 (39)
Na-illite ; Gonsalez Sanchez et al., 2008 (39)
Na-smectite 10/20 size fraction of Na-vermiculite
0.1 0.2 Na-vermiculite
0.1 0.2 Na-vermiculite
1 2 Na-vermiculite 1 2 Na-vermiculite
10 20 Na-vermiculite
1.0E-11
1.0E-11 0
0.2
0.4 0.6 total porosity (e)
0.8
0
ACS Paragon Plus Environment
0.05
0.1
0.15 0.2 0.25 0.3 interparticle porosity (einterp.)
0.35
Page 29 of 29
Environmental Science & Technology
TOC
1.0E-09
high charge swelling clays (vermiculite)
=
De (m²/s)
non-swelling clays (kaolinite - illite)
Na-vermiculite 1-2 µm Na-vermiculite 0.1-0.2 µm
1.0E-10
low charge swelling clays (Na-smectite)
1.0E-11 0.1
0.2
0.3
0.4 0.5 total porosity
0.6
0.7
0.8
ACS Paragon Plus Environment