Water Transport in Periodic Mesoporous Organosilica Materials - The

3 hours ago - E; Energy & Fuels · Environmental Science & Technology · Environmental Science & Technology Letters; I; Industrial .... Tools & Sharing...
0 downloads 0 Views 590KB Size
Subscriber access provided by University | of Minnesota Libraries

C: Surfaces, Interfaces, Porous Materials, and Catalysis

Water Transport in Periodic Mesoporous Organosilica Materials J. Benedikt Mietner, Michael Fröba, and Rustem Valiullin J. Phys. Chem. C, Just Accepted Manuscript • DOI: 10.1021/acs.jpcc.8b00046 • Publication Date (Web): 25 May 2018 Downloaded from http://pubs.acs.org on May 25, 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 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 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.

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 27 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

Water Transport in Periodic Mesoporous Organosilica Materials Benedikt J. Mietner,

†,‡

Michael Fröba,

∗,†,‡

and Rustem Valiullin

∗,¶

†Institute of Inorganic and Applied Chemistry, University of Hamburg, Hamburg, Germany ‡The Hamburg Centre for Ultrafast Imaging, Hamburg, Germany ¶Felix Bloch Institute for Solid State Physics, University of Leipzig, Leipzig, Germany E-mail: [email protected]; [email protected]

1

ACS Paragon Plus Environment

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

Abstract Water transport in periodic mesoporous organosilicas (PMO) was studied using pulsed eld gradient (PFG) NMR. A series of isogeometric PMO materials with dierent chemical compositions of the pore walls were investigated and compared to a purely siliceous MCM-41 material with an identical pore size. The long-range water diusivities measured were found to be largely controlled by the macroscopic textural properties of the materials, namely by the particle geometry and a degree of the particle agglomeration, and by thermodynamic conditions under which the experiments were performed. It is shown that their combined eect caused water molecules either to propagate predominantly along the capillary-condensed water domains or to frequently alternate their trajectories between these domains and the water phase in the inter-particle space. Because the transport rates in these two regimes dier substantially, it is suggested that, by a purposeful choice of the PMO composition, both the long-range transport rate and the chemical functionality can deliberately be tuned.

Introduction Diusive mass transfer is the most ubiquitous phenomenon in nature and is often a decisive step in various biological and physico-chemical processes. 1,2 Among them, water transport in nanoporous solids has long been in the focus of scientic research with the main intentions being to understand and to control the rates of water propagation under dierent thermodynamic conditions. 312 Concerning dierent strategies serving for these goals, purposeful designs of the pore system and of the pore surface chemistry are most widely explored. Along the former route, the correlations between the pore space morphology and molecular dynamics are exploited to manipulate the diusive transport. The structural details of the pore space in this case are eectively combined into a geometric tortuosity factor reecting slowing down of the diusion rate as compared to the bulk liquids. On decreasing the pore dimensions to the meso- or micropore ranges, in addition to 2

ACS Paragon Plus Environment

Page 2 of 27

Page 3 of 27 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

purely steric or geometric eects 13 strong connements give rise to additional mechanisms aecting the molecular mobilities. Thus, in the extreme cases, with the most prominent example being zeolitic materials, 14 tight connements may elevate the role of the intermolecular interactions so that they become the key factor determining diusive transport. As an example, in the pore channels with the pore diameters comparable to the molecular diameters, such that the molecules cannot bypass each other, the emergence of the singlele diusion regime severely aects the transport rates. On the other hand, in mesoporous materials with the pore sizes in the range between 2 and 50 nm, molecular ordering can alter the local mobilities. 15 The presence of the pore walls can eect the rates of water transport in two ways. First of all, the surface eld exerted by the pore walls can result in the position-dependent mobilities across the pore. 12 In particular, it is widely discussed that water mobilities in the rst hydration shells dier notably from those in the layers further apart from the pore walls. 1618 Alternatively, surface interactions may alter the phase state of water in the intra-pore spaces. It has found many experimental evidences that the phase state and transport under these circumstances are strongly coupled. 10,1921 This coupling is of a particular interest for an ecient manipulation of the water transport. Thus, it is speculated that the phase state oscillations in hydrophobic pores, as induced by the external pressure changes, may be responsible for directed water transport through biological membranes. 22 Hence, experimental studies of water dynamics in the pore channels with tunable surface chemistry and under dierent thermodynamic conditions is of immediate interest for improving our understanding of water behavior in nano-channels. Among them, a quantitative dierentiation between the eects of polarity and of pure steric eects is the most challenging one. 23 On considering transport in complex materials, the textural complexity present over different length scales is respectively ngerprinted in molecular transport rendering it length scale-dependent or heterogeneous. 24,25 Because the majority of the mesoporous materials are typically found in a powder form, two important transport modes need to be distinguished. 3

ACS Paragon Plus Environment

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

On the length scales below the particle dimensions, the intra-particle diusion rates are mostly determined by the geometry and pore wall chemistry of the material under study. 26 This transport mode is of importance for the processes in which the local mass transfer controls the surface-facilitated reactions or contributes to the connement-facilitated molecular separation. On the larger length scales exceeding dimensions of the individual particles, the overall transport rates are dierent from the intra-particle one. These rates are often referred to as the long-range diusion rates and result as a combined eect of the diusive resistances in the intra- and inter-particle spaces and of the physical mechanisms responsible for the mass transfer between the two spaces. 27 Notably, all these phenomena depend on thermodynamic conditions. 10,28 Often the long-range transport is a rate-determining step and needs to be appropriately tuned in order to yield high eciency or performance of the overall process. Hence, in addition to understanding diusion on a single pore level, explorations of the long-range transport turns out to be challenging and warrants thorough experimental studies. Periodic mesoporous organosilicas (PMOs) are ideal model systems because they combine the highly ordered pore structure of the well-known M41S-phases 29 with the variety of dierent surface chemistry within the pores. PMOs are synthesized using bis-silylated precursors of the form (RO)3 Si−R−Si(OR)3 where R is an organic bridging group which can be altered according to the desired surface properties. Recently, water molecules conned within the pores of a benzene-bridged PMO were shown to diuse and rotate with a smaller activation energy compared to water in the pores of pure siliceous MCM-41 because of the dierent interfacial interaction. 30 Furthermore, PMOs with an aromatic bridging group may exhibit a molecular-scale periodicity or "quasi-crystalline arrangement" within the pore walls. 31 This allows a periodically alternating surface chemistry along the pore channel, caused by arrays of silica and aromatic organic groups. Because of the endless possibilities for the organic bridging function, the surface chemistry of PMOs can be ne-tuned. For example, when going from a benzene bridging group to a biphenyl bridge, the organic hydrocarbon part 4

ACS Paragon Plus Environment

Page 4 of 27

Page 5 of 27 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

becomes larger and thus the overall pore wall gets more hydrophobic. 32 When going from a benzene bridge to a divinylaniline one, the amino function oers the possibility of forming hydrogen bonds with water, thus making the material considerably more hydrophilic again. In this work, we employed MCM-41 silica and benzene- (B), biphenyl- (BP) and divinylanilinebridged (A) PMO materials with cylindrical pore morphology and with the pore diameters of 3.3 nm as ideal model systems to explore the impact of chemical composition on the properties of spatially conned water. The synthesis procedure as well as the structural properties as probed by a number of conventional characterization techniques are thoroughly described in the Supporting Information. We performed the PFG NMR studies of the water transport in these materials at two distinct thermodynamic conditions corresponding with (i) the inter-particle space containing the water vapor phase and with (ii) the inter-particle space blocked by the ice phase.

Methods Pulsed eld gradient NMR The 1 H pulsed eld gradient NMR diusion experiments 33,34 were performed using a homebuild NMR spectrometer operating at 100 MHz resonance frequency for protons. The high values of the magnetic eld gradients up to 35 T/m were applied. 35 This allowed (i) to keep the time intervals in the NMR pulse sequences during which the nuclear magnetic magnetization evolved in the transverse plane relatively short (several ms) and (ii) to assess very small diusivities down to 10−12 m2 /s. The 13-interval pulse sequence, 36 with the pairs of the magnetic eld gradient pulses of dierent polarity applied in the encoding and decoding time intervals, was used with the primary goal to minimize the perturbing eects caused by the strong gradient pulses. Diusion experiments with dierent diusion times t between 10 ms and 160 ms were performed in order to check whether the diusion process probed was normal (the mean-square-displacements linearly growing with t) or some deviations were 5

ACS Paragon Plus Environment

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 6 of 27

caused due to the complexity of the material organization (leading, e.g., to the observation of restricted diusion with the mean-square-displacements growing as tα with α < 1 37 ). The spin-echo signal intensities S measured using this method are directly related to the so-called diusion propagator P (z, t). The latter is the probability density function to nd a displacement z , namely the projection of the displacement vector onto the z -axis along which the magnetic eld gradient is applied, during a given diusion time interval t. The relationship between S and P (z, t) is given by

Z



S(q, t) = S(0, t)

P (z, t)e−iqz dz,

(1)

0

where q ≡ γδg is the wave number, γ is the gyromagnetic ratio for protons, δ and g are the eective length and strength of the magnetic eld gradient pulses, respectively. This simple functional dependency allows to probe directly the characteristic of the diusion process by measuring S as a function of q at xed t. For normal diusion, i.e. when the propagator is of Gaussian form, Eq. 1 transforms to

S(q, t) = S(0, t)e−q

2 tD

(2)

,

with D = hz 2 i/(2t) being the diusivity. Any deviation of S(q, t) from the exponential form signals about either non-Gaussian statistics of the molecular displacements (anomalous diusion 38 ) or distribution of the diusivities over the sample volume (multi-phase or multicompartment diusion). In the latter case, there are dierent molecular ensembles (often referred to as 'phases', which does not necessarily imply thermodynamic phases) in the sample with dierent translational mobilities. In this case, S(q, t) can be described by a weighted sum of Eqs. 2,

S(q, t) = S(0, t)

X

pi e−q

2 tD

i

,

(3)

i

where pi and Di are the relative fractions and diusivities of the nuclear spins in the 6

ACS Paragon Plus Environment

Page 7 of 27 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

phase i. Eq. 3 is also valid if the dierent phases refer to the spatially-extended regions in which molecules have dierent diusivities and there is no molecular exchange between these regions. The occurrence of molecular exchange between dierent ensembles can be unveiled by performing the diusion experiments using dierent diusion times. If during the diusion times t molecules can explore dierent regions with dierent diusivities, pi in Eq. 3 become functions of the diusion time. 21 Note that in Eq. 3 we have implied no nuclear magnetic relaxation weighting for dierent phases. In our experiments the applicability of this assumption is ensured by keeping the transverse relaxation time intervals in the NMR pulse sequences substantially short. For multi-phase systems (in the sense of Eq. 3) the average diusivity

Dav =

X

pi Di

(4)

i

is a useful concept to characterize the overall transport. It can be easily obtained by approaching Eq. 2 to the low-q part of S(q, t). 21

Results and Discussion Vapor phase in the inter-particle space The experimental studies of the water transport have been performed under two distinct conditions. In the rst series of the diusion experiments, only the mesopore space has been lled with the capillary-condensed water. This has been achieved by equilibrating the samples with the water vapor at 80% relative humidity. The water sorption isotherms (see Figure S3 of Supporting Information) reveal that, at this vapor pressure, (i) all cylindrical mesopores are completely lled by water and (ii) some small amount of water (less than 10%) may be found between the particles. Even though the mesopore structure in all materials studied is identical (except the pore

7

ACS Paragon Plus Environment

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

wall chemical composition), the primarily measured NMR spin-echo diusion attenuation functions S(q, t) revealed dramatic dierences in the water transport behaviors. As an example, Figure 1 shows S(q, t) measured for dierent samples at the temperature of −20◦ C. The rst distinct feature observed is that while S(q, t) for MCM-41 and A-PMO are of an exponential form as given by Eq. 2, those for B-PMO and BP-PMO follow the pattern of Eq. 3. In the latter samples the diusion process exhibits features typical of multiphase systems, namely the occurrence of molecular ensembles with dierent diusivities. Secondly, the water diusivities in, e.g., MCM-41 and A-PMO, dier by more than one order of magnitude. In the B-PMO and BP-PMO samples the average diusivities as well are substantially higher than in the A-PMO sample. In order to unveil these dierences obtained, the transport mechanisms need to be established. To do this, we have performed the diusion experiments at dierent temperatures. The spin-echo diusion attenuations obtained were analyzed in the same way as indicated in Figure 1 and the nal results are shown in Figure 2. It turns out that the temperature dependencies of the diusivities for the A-PMO sample and for all other samples exhibit dierent activation energies EA for diusion. While in A-PMO EA ≈ 18 kJ/mol, which is typical for diusion in bulk liquid water, diusion in three other samples exhibits a notably higher activation energy EA ≈ 42 kJ/mol. This activation energy is close to the heat of water vaporisation and this nding is a hint to understand dierent transport patterns observed in the experiments. All samples under study are found in a powder form composed of small particles traversed by the tubular mesopores. The particle sizes are typically in the range of several hundreds of nanometers. The particles may further agglomerate. Because the samples were saturated with water at 80% humidity, the space formed between the particles contains mostly the water vapor phase. The outer surface of the particles is covered by a thin, few monolayer thick water lm and the contact points between the particles may also contain water lms. During the diusion time intervals used in the experiments, the molecular trajectories may alternate 8

ACS Paragon Plus Environment

Page 8 of 27

Page 9 of 27 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

between the vapor and condensed phases. Indeed, the water molecules approaching the water lm on the external particle surface may desorb, perform either Knudsen ights or molecular diusion in the vapor phase, and then be adsorbed again. The alternation frequency depends on the thermodynamic conditions, on the water diusivities in the dierent phases, and the spatial extensions of the domains of dierent phases. As a result, the diusion process, as seen by PFG NMR, becomes strongly dependent on the relationship between the diusion time t and the life-times τi of molecules in the dierent phases. 39 Under the conditions of the slow exchange (τi >> t) S(q, t) is well approached by Eq.3. However, because the fraction pv of water molecules in the vapor phase is substantially lower than pl in the capillary-condensed phase, S(q, t) apparently exhibits a mono-exponential shape. Thus, the thus measured S(q, t) reects the diusion behavior in the capillarycondensed phase only. Hence, by tting Eq. 2 to the experimental data one obtains the diusivity in the condensed phase. In another limit of the fast molecular exchange (τi