Vous êtes sur la page 1sur 7

Solid State Ionics 252 (2013) 19–25

Contents lists available at ScienceDirect

Solid State Ionics


journal homepage: www.elsevier.com/locate/ssi

Hydronium dynamics in the perchloric acid clathrate hydrate


Arnaud Desmedt a,⁎, Ruep E. Lechner b, Jean-Claude Lassegues a, François Guillaume a,
Dominique Cavagnat a, Joseph Grondin a
a
Groupe Spectroscopie Moléculaire, Institut des Sciences Moléculaires, UMR 5255 CNRS — Université Bordeaux 1, 351 Cours de la Libération, 33405 Talence Cedex, France
b
Helmholtz Zentrum Berlin, BENSC, Hahn-Meitner-Platz 1, 14109 Berlin, Germany

a r t i c l e i n f o a b s t r a c t

Article history: Clathrate hydrates are nanoporous crystalline materials made of a network of hydrogen-bonded water
Received 15 March 2013 molecules (forming host cages) stabilized by the presence of foreign (generally hydrophobic) guest
Received in revised form 10 June 2013 molecules. The encapsulation of strong acids such as perchloric acid (with an acid to water ratio of 5.5)
Accepted 12 June 2013
leads to an ionic clathrate structure formed with perchlorate anions within cationic cages. The excess
Available online 9 July 2013
protons, at the origin of the large protonic conductivity of the compound, are delocalized over the cage
Keywords:
sub-structure through a Grotthus mechanism. Such ionic crystal constitutes a good model system for
Quasielastic neutron scattering investigating the spatial and time characteristics of the elementary steps of the proton conduction. In this
Proton conduction paper, the dynamics of hydronium ions are investigated by means of incoherent quasi-elastic neutron
Mechanism scattering (QENS) experiments performed by tuning the observation time to the timescale of the probed
Dynamics dynamical process met in the perchloric acid clathrate hydrate at 220 K. Thanks to the specific versatility
Hydronium of time-of-flight spectrometer, two elementary processes have been identified for modeling the hydronium
Ionic clathrate hydrates ion dynamics: hydronium reorientations coupled to intermolecular proton transfer. The hydronium ions
undergo reorientations with a characteristic time of 42 ± 8 ps while the proton transfer within the
hydrogen bond between hydronium ion and neighboring water molecule occurs with a characteristic
time of 1.4 ± 0.3 ps. A previous high-resolution QENS investigation has shown that water molecules
surrounding the hydronium ions undergo reorientations with a characteristic time of 0.7 ± 0.1 ns at
220 K. Such experimental results outline the key role played by the relaxation of the water molecules
surrounding the hydronium ions prior to any proton transfer.
© 2013 Published by Elsevier B.V.

1. Introduction at investigating the dynamic properties of ionic clathrate hydrates


exhibiting high protonic conductivity.
Clathrate hydrates constitute an important class of nanoporous Numerous arrangements of the water cages exist to form a
crystalline materials, formed by water cages encapsulating foreign three-dimensional crystalline structure, depending on the nature of
(generally hydrophobic) molecules [1]. The existence of large the encapsulated molecules [5]. The nanometer scale cages are
quantities of natural hydrocarbon hydrates on hearth has probably formed from polygonal rings of water molecules connected at their
motivated numerous researches and developments relevant to edges by means of hydrogen bonds. Each oxygen atom of the host
energy and to environmental sciences [2–5]. Over the last decade, a sub-structure is surrounded by four equiprobable (half occupied)
particular attention has been directed towards the new technological hydrogen sites [4,16], so that the hydrogen atoms of the aqueous
opportunity that clathrate hydrates offer for storage and transportation sub-matrix are dynamically disordered, under the constraint of the
of gases such as hydrogen [6–11] in relatively “soft” pressure and so-called ice rule (i.e. each oxygen atom accepts and donates two
temperature conditions. It comes out that the physico-chemical H-bonds). When encapsulating strong acids such as perchloric acid
properties of gas hydrates have been extensively investigated these (with an acid to water ratio of 5.5), the clathrate structure formed
past years [5,12,13]. However, ionic species may also form clathrate is the type I [14]: a cubic unit cell contains two types of cages filled
hydrates, the counterions being incorporated into the water framework with guest molecules, as represented in Fig. 1. In the HClO4·5.5H2O
[14]. Because ionic clathrate hydrates display specific properties, these clathrate hydrate (melting point at 228 K), the acidic molecules are
systems are particularly promising for potential applications in various deprotonated [14,17] so that the cages are filled with ClO−4 anions and
fields [15]. The present paper falls in the frame of this issue and aims the excess protons participate to the cage structure, thus being made
of water molecules and hydronium cations. Leading to the violation of
the ice rule (there are more than two protons per oxygen atoms),
⁎ Corresponding author. Tel.: +33 5 4000 2937; fax: +33 5 4000 8402. such ionic defects are dynamically delocalized and are then at the or-
E-mail address: a.desmedt@ism.u-bordeaux1.fr (A. Desmedt). igin of a large protonic conductivity (of the order of 10− 2 S·cm− 1 at

0167-2738/$ – see front matter © 2013 Published by Elsevier B.V.


http://dx.doi.org/10.1016/j.ssi.2013.06.004
20 A. Desmedt et al. / Solid State Ionics 252 (2013) 19–25

512 the dynamics of clathrate hydrates [25] – is used to provide insights


into the hydronium dynamics for timescales ranging from a tenth to
a hundred of picoseconds thanks to the versatility of time-of-flight
techniques [26].

2. Experimental details

The HClO4·5.5H2O sample was obtained from an aqueous solution


of 50.37 wt.% HClO4 prepared under nitrogen atmosphere by adding
Milli-Q water to 70 wt.% HClO4 commercially available solu-
tion (99.999% Aldrich). Due to the strong acid character of the
sample, the HClO4·5.5H2O solution was contained between two
poly(tetrafluoroethylene) foils (10 μm thickness, Goodfellow)
in an Indium-sealed aluminum flat sample holder. In order to
minimize the effects of multiple scattering, a sample thickness of
51262 less than 0.2 mm was used, such that the transmitted intensity
was more than 90% of the intensity of the incident beam. QENS
spectra were recorded using the time-of-flight (ToF) spectrometer
12Å NEAT [27] at the Berlin Neutron Scattering Centre of the Hahn–Meitner
Institute (Berlin, Germany). The scattering angles (2θ) on this instru-
Fig. 1. Representation of the type I structure (12 Å cubic unit cell) adopted by the ment are in the range of 13° to 136° and the angle between the
perchloric acid clathrate hydrates. The 512 cage (7.8 Å diameter) is formed with 12
pentagons and the 51262 (8.7 Å diameter) is formed with 12 pentagons and 2
incoming beam and the plane of the sample holder was 45°. Various
hexagons. Both types of cage are filled with perchloric anions (not represented). energy resolutions have been used (see Table 1) in order to investigate
the energy resolution dependence of the QENS signal. The sample has
been cooled to 180 K and then warmed up and stabilized at 220 K for
ca. 220 K in perchloric acid clathrate hydrate [18,19]). Analysis of subsequent data collection. QENS broadening has been observed at all
impedance measurements [20] suggests that the proton delocaliza- used energy resolutions except at 1.3 meV energy resolution. The raw
tion follows the Grotthus mechanism (see references [21–23] for data were corrected for detector efficiency and sample-geometry
details about the various existing mechanisms): the long-range dependent attenuation, and the background was subtracted using
proton diffusion involves successive proton transfers between local- the measurement for the empty sample container (including the 2
ized charge carriers and reorientations of these carriers. Despite poly(tetrafluoroethylene) foils and the Indium sealing). The data were
such a mechanism is known since two centuries, there is very little normalized to the elastic scattering of vanadium and transformed to
experimental information about the spatial and time characteristics an energy scale. Several detectors were grouped together in order to
of the elementary steps responsible for the proton conduction improve the signal on noise ratio and coherent scattering peaks were
in ice-like systems. removed from the experimental data. All these procedures were carried
The protonic transport in ionic clathrate hydrates involves out using the INX software [28]. Scattering law analysis was carried
dynamical processes occurring on a broad timescale typically out using the program NEMO [29].
ranging from nanosecond to picosecond. By combining high-
resolution quasi-elastic neutron scattering (QENS) and 1H pulse- 3. Results
field-gradient nuclear magnetic resonance experiments [24], the
long-range translational diffusion of protons has been modeled as a 3.1. General considerations about the neutron scattering law of ionic
jump-diffusion between oxygen sites of the clathrate hydrate (with a clathrate hydrates
diffusion coefficient of 3.5 × 10−8 cm2/s at 220 K and an activation
energy of 29.2 ± 1.4 kJ/mol). Such a model implies that the protons In the present QENS experiment, the reduced spectra, Sexp(Q,ω),
remain located around the oxygen sites with a mean residence time of provide information about the proton dynamics (hydrogen atoms
3.7 ns at 220 K (the mean oxygen–oxygen distance being 2.8 Å). The have the largest existing incoherent neutron scattering cross sec-
oxygen sites are then occupied either by water molecules or by tion; all other nuclei have negligible contribution to the incoherent
hydronium ions and both molecular species may undergo localized mo- scattering signal) through the access to the incoherent scattering
tions on a timescale shorter than this residence time. According to the law of the clathrate hydrate, Sclathrate(Q,ω), recorded as a function
high-resolution QENS experiments [24], water molecule reorientations of the momentum transfer ℏQ and of the energy transfer ℏω [30]:
occur with a characteristic time of 0.7 ns at 220 K (activation energy
of 17.4 ± 1.5 kJ/mol [24]). However, the hydronium ion dynamics −ℏω =2 kT
Sexp ðQ ; ωÞ ¼ F ðQ Þ⋅e ½Sclathrate ðQ ; ωÞ⊗RðQ ; ωÞ þ BðQ Þ ð1Þ
could not be properly observed because of the limited energy window
accessible in these high-resolution QENS experiments. In the present
paper, a detailed analysis of the dynamic properties of hydronium ions where F(Q) is a scaling factor (including the Debye–Waller factor
exp(− Q2〈u2〉), 〈u2〉 being the mean square displacement), e− =2 kT
ℏω
incorporated into the host cages of the perchloric acid clathrate hydrate
is presented. An experimental approach by means of quasi-elastic is the detailed balance factor, B(Q) is a flat background and R(Q,ω)
neutron scattering – technique particularly appropriated to investigate is the instrumental resolution function. In the case of a protonic

Table 1
Experimental parameters of the NEAT time-of-flight spectrometer at HZB-Berlin. The measured HWHM (averaged values over the whole Q-range) of the QENS broadenings of the
perchloric acid clathrate hydrate at 220 K are provided.

Energy resolution [μeV] 30 50 90 150 300 1300


Wavelength [Å] 8.1 6.3 5.1 5.1 5.1 5.1
Q-range [Å−1] 0.2–1.4 0.3–1.8 0.4–2.2 0.4–2.2 0.4–2.2 0.4–2.2
HWHM [meV] 66 ± 12 123 ± 15 181 ± 24 373 ± 28 523 ± 74 –
A. Desmedt et al. / Solid State Ionics 252 (2013) 19–25 21

conductor, the classical incoherent neutron scattering function, ∙ the water molecule reorientations are at the origin of a QENS broad-
which is of main interest in the following, can be written as [24]: ening characterized by a HWHM of 1.8 μeV. With respect to the
ToF-QENS energy resolution, the scattering law S2LM(Q,ω) reduces to
Sclathrate ðQ; ωÞ ¼ STD ðQ ; ωÞ⊗SLM ðQ; ωÞ ð2Þ a Dirac function. Combining Eqs. (3) and (5), the ToF scattering law
reads:
where STD(Q,ω) is the classical incoherent neutron scattering function
resulting from the long-range translational diffusion of the protons   
−ℏω =2 kT 1 3 3
and SLM(Q,ω) is the classical incoherent neutron scattering function SToF ðQ; ωÞ ¼ F ðQ Þ⋅e SLM ðQ ; ωÞ þ δðωÞ ⊗RðQ; ωÞ þ BðQ Þ :
4 4
due to the localized motions of the ionic species (reorientations for
ð6Þ
instance).
Due to the full dissociation of the perchloric acid molecules in the Considering the localized feature of the hydronium ion dynamics
HClO4·5.5H2O clathrate hydrate, water molecules and hydronium (Grotthus mechanism), the combination of Eqs. (4) and (6) leads to
ions form the host cages trapping the perchlorate anions. These two describe the ToF spectra as:
chemical species are involved in the localized diffusive processes, so (" #
3
that the scattering law SLM(Q,ω) reads: 1 ΔLM ðQ Þ
SToF ðQ ; ωÞ ¼ F ðQ Þ⋅e =2
ℏω
− kT
A0 ðQ ÞδðωÞ þ ½1−A0 ðQ Þ þ ⊗RðQ ; ωÞ
) π Δ3LM ðQ Þ2 þ ω2
1 3 3 2 þBðQ Þ ð7Þ
SLM ðQ ; ωÞ ¼ S ðQ; ωÞ þ SLM ðQ ; ωÞ ð3Þ
4 LM 4

where S2LM(Q,ω) and S3LM(Q,ω) are the scattering functions arising 1 3 3


from the water molecules and the hydronium ions, respectively. The with A0 ðQ Þ ¼ A ðQ Þ þ : ð8Þ
4 0 4
weighting factors of each scattering law are calculated by considering
that only the protons of the clathrate hydrate contribute to the In other words, the ToF-QENS spectra should be the superimposition
incoherent scattering law with a 2:9 ratio of H3O+ to H2O species. of an elastic term and of a QENS broadening, both observables being sig-
Due to the spatial localization of these dynamical processes, each of natures of the hydronium ion dynamics. Each ToF-QENS spectrum (that
these molecular scattering laws can be represented by the superim- depends on the momentum transfer, the energy transfer and the energy
position of an elastic term and a quasielastic term: resolution) has been fitted by means of Eq. (7) with free parameters
F(Q), A0(Q) and Δ3LM(Q). In this phenomenological fitting procedure,
h i1 i
ΔLM ðQ Þ
i i i
SLM ðQ ; ωÞ ¼ A0 ðQ ÞδðωÞ þ 1−A0 ðQ Þ ð4Þ Δ3LM(Q) did not depend significantly on the momentum transfer what-
π ΔLM ðQ Þ2 þ ω2
i
ever the energy resolution was and their values averaged over the
whole Q range are given in Table 1. The measured phenomenological
where i refers to the considered chemical species (2 for water and 3 EISF (Q dependent) of the hydronium ions (Fig. 2) is systematically
for hydronium). The Elastic Incoherent Structure Factor (noted EISF), greater than 3/4, as expected according to expression (8). This observa-
Ai0(Q), is the amplitude of the elastic term represented by a Dirac tion confirms the fact that the water molecules undergo reorientations
function δ(ω) and provides information about the geometry of the on a timescale significantly longer than the ToF observation time. More-
localized diffusive motion. The Lorentzian function represents the over, the measured HWHM and EISF are not constant as a function of the
averaged quasielastic contributions for which the half-width at energy resolution. This observation suggests the existence of several hy-
half-maxima (denoted HWHM), ΔiLM(Q), provide information about dronium dynamical processes at the origin of the QENS broadening. In
the characteristic times of the localized motions. order to estimate the observation time on which these various dynam-
ical processes occur, the energy resolution dependence of the EISF has
3.2. Phenomenological analysis of the QENS spectra been fitted by means of the following phenomenological expression:

In QENS experiments, the scattering law is folded with the resolu- 0


A0 ðQ; ΔEÞ ¼ A0 ðQ ; 0Þ
tion function (see the Eq. (1)), for which the energy width, denoted X N   X
N
k k k
ΔE, provides a timescale (or observation time) on which any dynam- þ A0 ðQ ; 0Þexp −ΔEobs =ΔE with A0 ðQ ; 0Þ ¼ 1:
ical process will be observed in the QENS experiment [26]. In the k¼1 k¼0

present case, the energy resolution, ΔE, is greater than ca. 30 μeV. ð9Þ
Any dynamical process occurring on a timescale longer than ca.
100 ps will give rise to a QENS broadening for which the HWHM is of
a few microelectron volts, i.e. significantly narrower than ΔE. In other
words, such a dynamical process will not be observed on the ToF spectra
due to the time-filtering property of the instrumental resolution. In this
issue, it has been shown that two dynamical processes are occurring on
a timescale of the order of the nanosecond – i.e. not resolved with the
ToF spectrometer – by combining 1H-PFG NMR and high-resolution
QENS experiments [24]:
∙ the long-range translational diffusion of protons in the perchloric
clathrate hydrate has been modeled as a jump-diffusion of protons
between oxygen sites of the clathrate hydrate and gives rise to a
QENS broadening smaller than 1 μeV. In the present QENS experi-
ment, such component cannot be resolved with respect to the ToF
energy resolution and the component STD(Q,ω) in Eq. (2) can then
be approximated by means of a Dirac function. Thus the experimental Fig. 2. Observation time [26] (inverse of energy resolution) dependence of the hydronium
ToF scattering law SToF(Q,ω) deduced from Eq. (1) writes as: EISF measured (symbols) with the NEAT time-of-flight spectrometer at HZB-Berlin at
three selected momentum transfers on the perchloric acid clathrate hydrate at 220 K.
−ℏω =2 kT The lines represent the phenomenological fits of the experimental points with the help
SToF ðQ ; ωÞ ¼ F ðQ Þ⋅e ½SLM ðQ; ωÞ⊗RðQ; ωÞ þ BðQ Þ ð5Þ of expression (9).
22 A. Desmedt et al. / Solid State Ionics 252 (2013) 19–25

Such a phenomenological expression could be likened to an


intermediate scattering function (the time Fourier transform of the
scattering law [30]) in which the Fourier time is replaced by the
experimental observation time [26]; the kth relaxation process is
described by an exponential function characterized by an amplitude,
Ak0(Q,0), analogous to a structure factor and by an observation time,
1/ΔEkobs. The energy ΔEkobs then represents the energy resolution on
which the kth dynamical process will be observed. In the fitting
procedure, these energies do not depend on the momentum transfer;
as previously mentioned, the HWHMs do not depend on the
momentum transfer, whatever the energy resolution was. While the
measured EISFs could not be satisfactory fitted with a single
exponential function (i.e. N = 1 in expression (9)), two exponential
functions in the expression (9) were sufficient to reproduce the Fig. 3. Representation (projection on a plane) of two neighboring oxygen sites (opened
resolution dependence of the EISF with a good agreement as shown circles) of a clathrate hydrate; the one on the left hand houses hydronium ions and the
in Fig. 2 (the fitted parameters are given in Table 2). This analysis right hand site houses water molecules. The hydronium protons can visit the sites
suggests the existence of, at least, two dynamical processes numbered from 1 to 8. According to the expression (9), reorientational motion (jump
rate denoted τ−1reo on the figure) involves sites 1–4 while proton transfer between oxygen
characterizing the hydronium dynamics. The fastest one is observed sites (jump rate denoted τ−1 H on the figure) involves the sites numbered 1 and 5 for
with an energy resolution of ca. 300 μeV while the slower one instance.
gives rise to a QENS broadening observable with an energy resolu-
tion of ca. 50 μeV. This latter dynamical process is characterized
with non-negligible error bars. This is due to observation energy acid clathrate hydrate (〈O ⋯ O〉 ≈ 2.8 Å according to [14]), the pro-
(ΔE2obs = 50 μeV) that is intermediate between the one related to ton transfer within the H-bond may be approximated by means of
the fastest process (ΔE1obs = 292 μeV) and the one related to the a 2-site jump model. Assuming a single characteristic time for each
water molecules (of few microelectron volts [24]). To get a more plausible localized motion, the associated scattering law [24,30] is
quantitative analysis, models of the hydronium dynamics are the superimposition of an elastic term for which the EISF reads
required — issue of the next section.
1
A0;N ðQ Þ ¼ ½1 þ ðN−1Þj0 ðQdÞ ð11Þ
3.3. Modeling the hydronium dynamics N

and a single Lorentzian function for which the HWHM reads


In a clathrate hydrate, one host molecule is surrounded by four
other molecules according to a slightly distorted tetrahedral con- N −1
figuration [14]. Each water molecule thus accepts and donates ΔN ¼ τ : ð12Þ
N−1
two hydrogen bonds so that each oxygen site is surrounded by
four proton sites (having an occupation probability of a half) as In these expressions, N refers to the number of sites (2, 3 or 4), τ−1
represented in Fig. 3. In the case of the perchloric acid clathrate is the jump rate (denoted τ−1 reo in the case of hydronium reorientation
hydrate, each oxygen site might be occupied by a water molecule and τ−1
H in the case of proton transfer in H bond), j0(x) is the spherical
or a hydronium ion [14]. Let us consider a hydronium ion located Bessel function of first kind and of zero order and d corresponds to
at one oxygen site. The various localized dynamical processes that the jump distance. For the hydronium reorientations, the averaged
a hydronium proton could undergo may be summarized with two distance between two hydrogen atoms is d = 1.6 Å (i.e., assuming
main contributions: reorientations of the hydronium ions (proton a bond angle H\O\H of 106.7° and a O\H bond length of 1.0 Å
jumps between sites labeled 1 to 4 in Fig. 3) and proton transfer adopted by a solvated hydronium ion [31]). For the proton transfer
within the H-bond between the hydronium ion and neighboring within the H-bond, the averaged jump distance is d = 1 Å, i.e. the
water molecules (proton transfer between sites labeled 1 and 5 in averaged distance between two proton sites contained in the
Fig. 3 for instance). Assuming that these two dynamical processes H-bond network (sites labeled 1 and 5 in Fig. 3) of the perchloric
are occurring on different timescales, the hydronium scattering acid clathrate hydrate according to structural determinations [14].
law results from the convolution of a reorientational scattering In order to discriminate these components, it is worthwhile to note
law, Sreo(Q,ω), and a H-bond transfer scattering law, SH(Q,ω): that according to the phenomenological analysis previously discussed,
the slower localized motion is observed with an energy resolution
3
SLM ðQ; ωÞ ¼ SH ðQ; ωÞ⊗Sreo ðQ; ωÞ: ð10Þ ΔE2obs = 50 μeV. In a first approximation, the corresponding QENS
broadening will not be disentangled from the resolution function of the
Considering the pseudo-tetrahedral symmetry of the site occupied data recorded with an energy resolution close to ΔE1obs = 292 μeV.
by a hydronium ion, reorientations about C2 axes, about C3 axes or Analyzing the EISF measured with an energy resolution of 300 μeV yields
about both types of axis may occur. Such dynamical processes may the spatial signature of the fastest localized motion by time-filtering
be approximated with the help of instantaneous jumps between 2 the slowest localized motion. In Fig. 4, the EISF measured at 300 μeV
sites, 3 sites or 4 tetrahedrally distributed sites. Moreover, in view resolutions is compared to the EISF calculated by considering the four
of the averaged distance between two oxygen sites in the perchloric possible models given by Eqs. (8) and (11). None of the reorientational
models (i.e. d = 1.6 Å in Fig. 4) are in agreement with the experimental
EISF. Only the 2-site jump model with d = 1.0 Å is close to the experi-
Table 2
mental data. This model has then been fitted to the experimental EISF
Best fit parameters for the phenomenological expression (9) reproducing the energy
resolution dependence of the hydronium EISF. with d as a free parameter: a jump distance of 1.1 ± 0.1 Å has been
obtained. Such value falls in the range of the one expected from the struc-
Q [Å−1] A00(Q,0) A10(Q,0) ΔE1obs [μeV] A20(Q,0) ΔE2obs [μeV]
tural determination [14], i.e. d = 1 Å. Thus, the broader component
1.0 0.93 ± 0.02 0.04 ± 0.04 292 ± 99 0.03 ± 0.02 50 ± 30 associated with the observation energy ΔE1obs = 292 μeV may be
1.4 0.91 ± 0.02 0.06 ± 0.04 292 ± 99 0.03 ± 0.02 50 ± 30 assigned to the proton transfer within the H-bonds while the narrower
1.8 0.88 ± 0.02 0.10 ± 0.06 292 ± 99 0.02 ± 0.02 50 ± 25
component associated with the observation energy ΔE2obs = 50 μeV
A. Desmedt et al. / Solid State Ionics 252 (2013) 19–25 23

resolution range and the whole momentum transfer range. Examples


of fitted spectra are reported in Fig. 5 and the fitted parameters
are reported in Table 3. The proton transfer within the H bond is
the fastest dynamical process performed by the hydronium ion
in agreement with the phenomenological analysis and with the
expected timescale for such a process [21–23]. Moreover, the jump
distance, dH = 0.96 ± 0.10 Å, is in full agreement with the structure
determination [14]. Concerning the hydronium reorientations, the
jump distance (i.e., dreo = 1.31 ± 0.15 Å) is shorter than the one
expected by assuming a “standard” configuration of the solvated
hydronium ion [31], i.e. 1.6 Å. This observation suggests that the
oxygen atom does not coincide with the center of rotation when
performing a tetrahedral jump. Such a result is in agreement with
the X-ray diffraction analysis [14]: the isotropic thermal parameters
Fig. 4. Experimental EISF (symbols) measured with the NEAT time-of-flight spectrometer of cage atoms are quite large (values between 2.05 Å2 and 2.95 Å2 for
at HZB-Berlin (energy resolution ΔE = 300 μeV) on the perchloric acid clathrate hydrate
oxygen atoms and of 4 Å2 for the protons) and the distribution of
at 220 K. The dashed lines represent the EISF calculated according to the various models
considered to reproduce the localized motion of hydronium ion (see text and proton site around the oxygen atom is slightly distorted from perfect
expression (11) for details). The continuous line is the fit of a two-site jump model tetrahedra.
reproducing the proton transfer between two neighboring oxygen atoms.
4. Concluding remarks
may be assigned to the reorientational motion of hydronium ions. A com-
plete model has to be considered in order to confirm this hypothesis and The present QENS investigations carried out on the perchloric acid
to determine the model for hydronium reorientations. clathrate hydrates yield the development of a model reproducing the
The complete model of the perchloric acid clathrate hydrate has dynamics of hydronium ions inserted in a crystalline framework of
been fitted to the whole set of QENS data i.e., as a function of the hydrogen bonded water molecules. The QENS experiments have
momentum transfer, of the energy transfer and of the energy resolution. been performed by using several energy resolutions (from 30 μeV to
The models for the long-range proton transfer (STD(Q,ω) in Eq. (2)) and 1.3 meV), allowing to tune the observation time to the timescale of
for the water molecule reorientations (S2LM(Q,ω) in Eq. (3)) correspond the probed dynamical process. Thanks to this specific versatility of
to the ones determined by combining 1H-PFG NMR and high-resolution ToF spectrometer, two elementary processes have been identified
QENS experiments [24]. The models and the parameters associated for modeling the hydronium ion dynamics: hydronium reorientations
with these two components are reported in Table 3. The model for coupled to intermolecular proton transfer. The hydronium reorientations
the hydronium scattering law given by Eq. (10) has been considered undergo over four different orientations tetrahedrally distributed
by using a 2-site jump model for the proton transfer within the and occur with a characteristic time of 42 ± 8 ps at 220 K. In this
−1
H-bond (EISF denoted AH H
0,2(Q) and HWHM, Δ2 = 2τH ) and any of reorientational motion, the hydronium O\H bonds are pointing towards
the three dynamical processes (i.e. 2-sites, 3-sites or 4-sites jumps) neighboring oxygen atoms of water molecules. The second dynamical
for the hydronium reorientations (EISF denoted Areo 0,N(Q) and HWHM, process has been assigned to the transfer of a hydronium proton within
−1
Δreo
N = Nτreo /(N − 1)). The hydronium scattering law then writes as the hydrogen bridge formed between a “hydronium” oxygen site and a
the sum of three QENS components: neighboring “water” oxygen site. This process, described by means of a
jump model between two proton sites distant by 0.96 ± 0.10 Å, occurs
S3LM ðQ ; ωÞ ¼ AH reo
0;2 ðQ ÞA0;N ðQ ÞδðωÞ with a characteristic time of 1.4 ± 0.3 ps at 220 K.
h i1 Δ2
H QENS experimental investigations lead to a quantitative de-
þAreo H
0;N ðQ Þ 1−A0;2 ðQ Þ  H 2 scription of the underlying elementary steps responsible for the
π Δ þ ω2
2
h i1 Δ
reo long range proton diffusion [24], related to the proton conductivity.
0;N
þAH reo
0;2 ðQ Þ 1−A0;N ðQ Þ   ð13Þ Following the Grotthus mechanism [21–23], three elementary steps are
π Δreo 2 þ ω2
0;N involved: water molecules reorientations, hydronium reorientations
h ih i1 Δreo H and intermolecular proton transfer. A hierarchical distribution of time-
reo H 0;N þ Δ2
þ 1−A0;N ðQ Þ 1−A0;2 ðQ Þ   : scales is clearly evidenced. The present QENS study reveals that hydro-
π Δreo þ ΔH 2 þ ω2
0;N 2 nium ions undergo dynamical processes on a picosecond timescale.
According to a previous high-resolution QENS investigation prob-
It follows that in the fitting procedure, there were five free param- ing nanosecond timescales [24], water molecules surrounding the
eters: the number of sites (N = 2, 3 or 4) visited during hydronium hydronium ions undergo reorientations with a characteristic time
reorientations, the jump distance characterizing these reorientations of 0.7 ± 0.1 ns at 220 K in the perchloric acid clathrate hydrates.
(denoted dreo), the jump distance characterizing the proton transfer The water molecule reorientations can then be considered as the
within the H-bond (denoted dH), the reorientational characteristic limiting elementary step in the mechanism of the proton conduc-
time (denoted τreo) and the H-bond diffusion characteristic time tivity met in the perchloric acid clathrate hydrate. In view of the
(denoted τH). Within the quasielastic region (i.e. |ℏω| ≤ 3 meV), concentration of hydronium ions in the system (i.e., 1 hydronium
an excellent agreement has been obtained over the whole energy per 4.5 water molecules), such experimental result outlines the

Table 3
Characteristic parameters of the dynamics of hydronium ions and water molecules in the perchloric clathrate hydrate at 220 K (see text for details about the model).

Long-range proton diffusiona Water reorientationsa Hydronium reorientations Proton transfer

Model Chudley–Elliot 2-Site jump Tetrahedral jump 2-Site jump


Jump distances [Å] 2.8 1.45 1.31 ± 0.15 0.96 ± 0.10
Residence time [ps] 3700 700 42 ± 8 1.4 ± 0.3
a
From [24].
24 A. Desmedt et al. / Solid State Ionics 252 (2013) 19–25

Fig. 5. Experimental (points) and fitted (thick continuous line) QENS spectra of the perchloric acid clathrate hydrate at 220 K (NEAT time-of-flight spectrometer at HZB-Berlin). The QENS
contributions due to water molecule reorientations and to the long-range proton diffusion are shown as thin continuous lines. The three QENS components due to hydronium ions are
shown (see Eq. (10)); the dotted line represents the contribution related to reorientations (HWHM = 4τ−1 reo /3); the dashed line represents the contribution related to proton transfer
within hydrogen bond (HWHM = 2τ−1 −1
H ); the dash–dotted line represents the contribution related to H-bond proton transfer coupled to reorientation (HWHM = 4τreo /3 + τH ).
−1

key role played by the relaxation of the molecules surrounding the [3] I. Chatti, A. Delahaye, L. Fournaison, J.P. Petitet, Energy Convers. Manag. 46 (2005)
hydronium ions prior to any proton transfer. 1333.
[4] E.D. Sloan, Nature 426 (2003) 353.
[5] E.D. Sloan, C.A. Koh, Clathrate Hydrates of Natural Gases, CRC Press, Boca Raton,
FL, 2008.
Acknowledgments [6] Y.A. Dyadin, E.G. Larionov, A.Y. Manakov, F.V. Zhurko, E.Y. Aladko, T.V. Mikina,
V.Y. Komarov, Mendeleev Commun. 9 (2009) 209.
A.D. wishes to thank the HZB neutron facilities in Berlin (Germany) [7] Y.A. Dyadin, E.G. Larionov, E.Y. Aladko, A.Y. Manakov, F.V. Zhurko, T.V. Mikina,
V.Y. Komarov, E.V. Grachev, J. Struct. Chem. 40 (1999) 790.
and the staff members for the beam time allocation and for the support [8] H. Lee, J.-W. Lee, D.Y. Kim, J. Park, Y.T. Seo, H. Zeng, I.L. Moudrakovski, C.I. Ratcliffe,
allowing numerous neutron scattering experiments presented in this J. Ripmeester, Nature 434 (2005) 743.
paper to be performed. Finally, the scope of this paper falls in the [9] L.J. Florusse, C.J. Peters, J. Schoonman, K.C. Hester, C.A. Koh, S.F. Dec, K.N. Marsh,
E.D. Sloan, Science 306 (2004) 469.
frame of the project ANR 2011-JS08-002-01, funded by the French [10] W.L. Mao, H.K. Mao, A.F. Goncharov, V.V. Struzhkin, Q.Z. Guo, J.Z. Hu, J.F. Shu,
“Agence Nationale de la Recherche”. R.J. Hemley, M. Somayazulu, Y.S. Zhao, Science 297 (2002) 2247.
[11] E. Pefoute, E. Kemner, J.C. Soetens, M. Russina, A. Desmedt, J. Phys. Chem. C 116
(2012) 16823.
References [12] M.R. Walsh, C.A. Koh, E.D. Sloan, A.K. Sum, D.T. Wu, Science 326 (2009) 1095,
(and references therein).
[1] G.A. Jeffrey, in: J.L. Atwood, J.E.D. Davies, D.D. Mac-Nicol, F. Vögtle (Eds.), [13] J.S. Tse, D.D. Klug, J.Y. Zhao, W. Sturhahn, E.E. Alp, J. Baumert, C. Gutt, M.R. Johnson,
Comprehensive Supramolecular Chemistry, Hydrate Inclusion Compounds, W. Press, Nat. Mater. 4 (2005) 917.
vol. 6, Pergamon, Oxford, 1996, p. 757. [14] D. Mootz, E.-J. Oellers, M. Wiebke, J. Am. Chem. Soc. 109 (1987) 1200.
[2] C.A. Koh, E.D. Sloan, AICHE J. 53 (7) (2007) 1636. [15] K. Shin, J. Cha, Y. Seo, H. Lee, Chem. Asian J. 5 (2010) 22, (and references therein).
A. Desmedt et al. / Solid State Ionics 252 (2013) 19–25 25

[16] F. Hollander, G.A. Jeffrey, J. Chem. Phys. 66 (1977) 4699. [25] A. Desmedt, L. Bedouret, E. Pefoute, M. Pouvreau, S. Say-Liang-Fat, M. Alvarez, Eur.
[17] A.I. Karelin, J. Struct. Chem. 32 (1991) 199. Phys. J. 213 (2012) 103, (Special Topics).
[18] S. Aschrafi-Mahabadi, M. Cappadonia, U. Stimming, Solid State Ionics 70 (71) [26] R.E. Lechner, Physica B 301 (2001) 83.
(1994) 311. [27] B. Ruffle, J. Ollivier, S. Longeville, R.E. Lechner, Nucl. Inst. Methods Phys. Res. A 449
[19] T.-H. Huang, R.A. Davis, U. Frese, U. Stimming, J. Phys. Chem. 92 (1988) 6874. (2000) 322.
[20] M. Cappadonia, A.A. Kornyshev, S. Krause, A.M. Kuznetsov, U. Stimming, J. Chem. [28] F. Rieutord, Institut Laue–Langevin, Technical Report 890RI17, 1990.
Phys. 101 (1994) 7672. [29] Software available on request. Contact A. Desmedt for further details.
[21] K.-D. Kreuer, Chem. Mater. 8 (1996) 610. [30] M. Bée, Quasielastic Neutron Scattering, Adam Hilger, Bristol, 1988.
[22] K.-D. Kreuer, S.J. Paddison, E. Spohr, M. Schuster, Chem. Rev. 104 (2004) 4637. [31] J.M. Hermida-Ramon, G. Karlström, J. Mol. Struct. (THEOCHEM) 712 (2004) 167.
[23] D. Marx, A. Chandra, M.E. Tuckerman, Chem. Rev. 110 (2010) 2174.
[24] A. Desmedt, F. Stallmach, R.E. Lechner, D. Cavagnat, J.C. Lassègues, F. Guillaume,
J. Grondin, M.A. Gonzalez, J. Chem. Phys. 121 (2004) 11916.

Vous aimerez peut-être aussi