In Situ Electron Paramagnetic Resonance and X-ray Diffraction

Subscriber access provided by the University of Exeter

Article

In situ Electron Paramagnetic Resonance and X-Ray Diffraction Monitoring of Temperature-Induced Breathing and Related Structural Transformations in Activated V-doped MIL-53(Al) Irena Nevjestic, Hannes Depauw, Karen Leus, Geert Rampelberg, Claire A. Murray, Christophe Detavernier, Pascal Van Der Voort, Freddy Callens, and Henk Vrielinck J. Phys. Chem. C, Just Accepted Manuscript • DOI: 10.1021/acs.jpcc.6b04402 • Publication Date (Web): 08 Jul 2016 Downloaded from http://pubs.acs.org on July 10, 2016

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

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

Page 1 of 35

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

In Situ Electron Paramagnetic Resonance and X-Ray Diffraction Monitoring of Temperature-Induced Breathing and Related Structural Transformations in Activated V-doped MIL-53(Al) Irena Nevjestić,a Hannes Depauw,b Karen Leus,b Geert Rampelberg,a Claire A. Murray,c Christophe Detavernier,a Pascal Van Der Voort,b Freddy Callens,a Henk Vrielincka* a) Ghent University, Dept. Solid State Sciences, Krijgslaan 281-S1, B-9000 Gent (Belgium) b) Ghent University, Dept. of Inorganic and Physical Chemistry, Krijgslaan 281-S3, B-9000 Gent (Belgium) c) Diamond Light Source, Harwell Science and Innovation Campus, Didcot, OX11 0DE, United Kingdom * Phone: 0032 9 264 43 56, Fax: 0032 9 264 49 96, [email protected]

ACS Paragon Plus Environment

1


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

Page 2 of 35

Abstract

The Metal-Organic Framework MIL-53(Al) is characterized by a distinct reversible structural transition between a narrow pore (NP) and a large pore (LP) state, resulting in expansion or contraction of this three-dimensional porous framework also called breathing. This transition is studied for vanadium-doped MIL-53(Al), induced by temperature (T) using in situ Electron Paramagnetic Resonance (EPR) and X-ray Diffraction (XRD) in air and in vacuum. The EPR active VIV=O molecular ions are used as local probes to detect the NP to LP transitions. The EPR spectra of VIV=O embedded in the NP and LP MIL-53(Al) states are clearly distinguishable. The temperature dependent EPR and XRD data can consistently be interpreted in terms of T-ranges in the experiments where one of the states is predominantly present and a narrow T-range in which the two states coexist. In addition the XRD data indicate that the NP state undergoes a transition to a metastable state characterized by different lattice parameters than the NP state at room temperature, before the transition to the LP state occurs. The EPR spectra, on the other hand show that only in the LP state the VIV=O ions can exhibit an interaction with paramagnetic O2 molecules from air.


2

Page 3 of 35

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


Introduction Metal-Organic Frameworks (MOFs) are microporous materials in which metal clusters are connected by organic linkers. This class of materials has great potential for a wide range of applications from (photo)catalysis, gas storage and separation to sensing.1-3 Selecting the combination of metal complex and organic ligand allows to tune the specific surface area and the size, volume and shape of the pores. In comparison with rigid zeolite frameworks or activated carbon, MOFs offer a broader range of potential applications and better performance which is especially interesting for industry where adsorptive separation and catalytic activities are very important and where need for more efficient, selective and stable materials is omnipresent. MIL-53(Al)4 [Al(OH)(BDC); BDC = terephthalate or 1,4-benzenedicarboxylate, MIL = Materials of the Institute Lavoisier] is a MOF constructed of chains of trans corner-shared aluminum octahedra (AlO4(OH)2) linked by BDC in the four directions perpendicular to the chains. In MIL-53(Al)AS (AS = as-synthesized), the pores are mainly occupied with terephthalic acid molecules. The MOF is activated by removing all molecules trapped in the pores after synthesis via heat treatment in an oxidizing atmosphere (calcination) or solvent extraction methods.5 The flexibility and mechanisms for breathing of MOFs is intriguing but still not completely elucidated. Activated MIL-53(Al) exhibits strong framework flexibility. The structure can reversibly change between a narrow pore (NP) and a large pore (LP) state by changing temperature, pressure and by exposure to certain gas molecules, in particular CO26 and H2O4. First principles calculations predict that at room temperature (RT) the LP state is most stable.7 However, under ambient air conditions, uptake of H2O leads to pore contraction. For this reason, very often the NP state (or a mixture of NP and LP states) is found back at RT when storing


3


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

Page 4 of 35

MIL-53(Al) at ambient conditions. When heating MIL-53 to above 375 K water is released from the structure and the LP state is obtained.4,8 The sensitivity and selectivity of the MIL-53 framework towards organic molecules may find applications in gas sorption, with the selective adsorption of CO2 from CO2 /CH4 mixtures in MIL-53(Cr) as a notable example.9 Such detailed information on the breathing phenomenon has been obtained by monitoring the physical and chemical properties of the MOF10,11 with a variety of analytical techniques. In situ X-ray diffraction (XRD) has played a central role in many of the studies: e.g. X-ray thermodiffractometry and CO2 sorption12 measurements were used to study thermal behavior of MIL-53(Cr)8, host-guest and guest-guest interactions in the MIL-53(Fe) framework were explored with in situ XRD under exposure to a variety of liquid-phase organic compounds.13 The lattice parameters for the AS, NP and LP states of MIL-53(Al) are summarized in Table 1. We adopt the parameters reported by Loiseau et al.4 from powder XRD analysis. In Table 1 the space groups and lattice parameters have been converted14 in such way that for each structure a corresponds to the longest lattice parameter and c to the shortest. In this way for all three structures the c axis lies along the Al-OH chains in the framework and indicates the direction of the channel-like pores. The Miller index labeling of XRD peaks in this paper follows this definition.


4

Page 5 of 35

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


Table 1. Crystal structure with lattice parameters4 of MIL-53(Al) in the as-synthesized, narrow pore and large pore state. As-synthesized (AS) state

Narrow pore (NP) state

Large pore (LP) state

Orthorhombic Pnam (no. 62)

Monoclinic Cc (no. 9)

Orthorhombic Imcm (no. 74)

a = 17.129 Å

a = 19.513 Å

a = 16.675 Å

b = 12.182 Å

b = 7.612 Å

b = 12.813 Å

c = 6.628 Å

c = 6.576 Å

c = 6.608 Å

β = 104.24° Blue: aluminum, red: oxygen, grey: carbon, black: unreacted linkers and/or solvent molecules

Electron paramagnetic resonance (EPR) and electron–nuclear double resonance (ENDOR) also provide detailed information on molecular structure, but have so far not been extensively used in the study of MOFs. A study of the breathing in Cr-doped MIL-53(Al) upon CO2 adsorption showed that EPR is very sensitive in distinguishing the LP and NP states of the MOF, but is also sensitive to different amounts and configurations of CO2 molecules trapped into the pores.15 The hysteresis in the temperature-induced breathing of Cr-doped MIL-53(Al) was also studied with the aid of EPR. It was shown that, depending on the thermal history, the LP and NP states can coexist at RT. In addition, it was found that even at 8 K, a considerable fraction (~20%) of the sample still is in the LP state.16 Sheveleva et al.17 studied the breathing of MIL-53(Al) by EPR


5


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 35

using 2,2,6,6-tetramethyl-1-piperidinyloxyl (TEMPO) radicals as paramagnetic probes. Kozachuk et al.18 studied breathing with EPR on V-doped MIL-53(Al). The conclusion of this study was that in samples with low vanadium concentration, the majority of the dopant ions end up on sites that are not affected by the NP-LP transition of the MIL-53(Al) host. Our group recently reported a detailed EPR and ENDOR study of the mixed metal system of MIL-53(Al/V)AS, where we clearly demonstrated the substitutional incorporation of the dopant.19 In the present study we focus on the breathing effect in activated V-doped MIL-53(Al) triggered by temperature and monitored with in situ EPR and XRD in air and in vacuum. The EPR active VIV=O molecular ions prove to be excellent local probes to detect the NP to LP transitions in the framework.


6

Page 7 of 35

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


Experimental Methods Synthesis and Material Characterization of V-Doped MIL-53(Al) The synthesis was performed under autogenous pressure for 72 h at 475 K in a 23 mL Teflonlined stainless-steel autoclave by using a mixture of VCl3 and AlCl3, terephthalic acid (H2BDC) and deionized water. In general, 0.232 g (1.4 mmol) of H2BDC was used. The AS powder was white colored and stable in air. After filtration, the powder was washed several times with deionized water and dried under vacuum at RT. We applied a solvent extraction method5 for activation by which the unreacted linker and metal salts were removed. This procedure consisted of three steps. Firstly, the solid was heated at 425 K with dimethylformamide (DMF) in a Teflonlined autoclave for 24 h. Secondly, the solid was treated at 395 K in methanol for 72 h. Finally, the powder was dried at 475 K for 2 h under vacuum to complete activation. Powder XRD patterns were recorded on AS and activated samples by using a Thermo Scientific ARL X’TRA diffractometer, operated at 40 kV, 30 mA, with Cu Kα radiation (λ=1.5406 Å) (see supporting information, Figure S1). N2 sorption measurements on activated samples were performed by using a Belsorp-mini II setup (Bell Japan, Inc) and yielded a Langmuir surface area of 1590 m2/g, the same as for pure MIL-53(Al)4 (see supporting information, Figure S2). A Varian FS220 F-AAS instrument with a nitrous oxide–acetylene flame was used to determine the V/Al atomic concentration ratios (V/Al = 1.5 mol%) in the samples after activation. EPR Measurements EPR spectra were recorded in X-band (9.51 GHz) in continuous-wave (CW) mode, using a Varian E-line spectrometer equipped with a Bruker ER 4114GT high-temperature resonator, a HP 5342A microwave frequency counter and Bruker NMR ER 035M Gaussmeter. The microwave power was set at 5 mW, and field modulation at an amplitude of 0.08 mT and a


7


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

Page 8 of 35

frequency of 10 kHz was applied. The magnetic fields were calibrated using the spectrum of diphenyl picrylhydrazyl (DPPH ; g = 2.0036). Powder samples were placed in one-end closed (under no circumstances sealed) quartz capillaries (3 mm inner diameter) which were during the EPR (and XRD) experiments either evacuated by continuous pumping (~1 Pa with a rotary pump or ~10-2 Pa with a turbomolecular pump), or in contact with air at ambient conditions. The sample temperature was controlled by a continuous flow of heated N2 over the bottom part of these tubes, which was centered in the EPR resonator. Since there was no direct contact between the heating gas and the sample, EPR spectra were recorded 20-30 minutes after a stable temperature in the heating gas was reached. XRD in vacuum In situ powder XRD data in vacuum were collected on the I11 powder diffraction beamline at the Diamond Light Source (λ = 0.825086 Å).20 A borosilicate glass capillary of 0.5 mm internal diameter was filled with powder to a height of approximately 40 mm, and the capillary was stoppered with glass wool. The capillary was continuously pumped to 10-3 Pa with a turbomolecular pump. An Oxford Cryostream (80 – 500 K) system was used for heating the sample from RT to 450 K and cooling back to RT at a heating/cooling rate of 5 K/min. A wideangle position sensitive detector based on Mythen-2 Si strip modules was used to collect the XRD data. The detector was moved at constant angular speed with 10 seconds scan time at each temperature, and 10 seconds waiting time in order to let the temperature stabilize. The data were analyzed using Lebail refinements in TOPAS21 with batch mode with a fixed zero-point error, wavelength and instrument profile. Starting values for the lattice parameters for the LP and NP states were taken from the literature.4 Analysis revealed that the LP state started


8

Page 9 of 35

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


to grow in at 344 K, with the NP state disappearing by 424 K. Trace impurities of terephthalic acid were identified throughout the temperature range studied for this compound. XRD in air In situ powder XRD in air was carried out on a Bruker D8 Discover XRD system equipped with a Cu X-ray source (λ = 1.5406 Å) and a linear X-ray detector. The powder sample was spread on a silicon wafer and placed on the sample heating stage. θ-2θ measurements were carried out in air at atmospheric pressure. The sample was heated from RT to 450 K and cooled back to RT at a heating/cooling rate of 5 K/min. The temperature was measured with a K-type thermocouple. Data were collected in ramping mode with a scanning time of 10 s.

Results Two VIV=O states in activated V-doped MIL-53(Al) The top two traces in Figure 1 show the powder EPR spectra of activated 1.5% V-doped MIL53(Al) in two situations: (a) where the majority of the powder sample is in the NP state, and (b) with a predominance of the LP state. Situation (a) was obtained by immersing a tube containing the powder in air overnight in liquid N2. For the other extreme (b), the sample was heated to 445 K under vacuum (1×10-2 Pa) during 4h. In order to exclude possible influences of the temperature and atmosphere, both EPR spectra were recorded at RT and 1×10-2 Pa continuously pumping the sample tube. The third spectrum in Figure 1 (trace c) is that of the AS powder. All three spectra exhibit the typical characteristics for vanadyl (VIV=O, VIV with a 3d1 configuration, S=1/2, I(51V)=7/2) molecular ions. They have a close-to-axial g tensor with 2 > gx ≈ gy > gz and a

51

V HF tensor with Ax ≈ Ay ≈ 165 MHz < Az ≈ 500 MHz. In spite of these similarities, the

three spectra can be clearly distinguished. In the spectra of AS and LP, peaks appear at


9


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

Page 10 of 35

approximately the same field positions. The difference between them seems to be mainly resulting from a change in line width. Spectrum a is more substantially different: in comparison with the other two spectra, resonance peaks appear at clearly different positions. In addition in the central part of the spectrum corresponding with the NP state, more resolved features appear than in the corresponding parts of the spectra for LP and AS. This points to a larger deviation from axial symmetry for the VIV=O sites in the NP state, in comparison with the LP and AS states. This observation seems well correlated with the differences in the local environment (symmetry) at the Al-OH nodes in MIL-53(Al) in these three states (see Supporting information, Table S1). Indeed, the AlO4 ‘planes’ (approximately perpendicular to the Al-OH chains) in the NP state deviate more substantially from a square planar arrangement than for other two states.

Figure 1. Powder EPR spectra of activated 1.5% V-doped MIL-53(Al) in vacuum at RT. a – NP state, b – LP state, c – AS state. The insets show a magnification of the high-field part of the spectra.


10

Page 11 of 35

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


One might thus suspect that the VIV=O ions can be used as paramagnetic probes for observing the NP to LP transitions. This is further explored in the following subsections by in situ heating EPR and XRD experiments. By heating the samples, Kozachuk et al.18 succeeded in isolating one of the spectral components which they labeled II. Judging on the shape of the spectrum and supported by simulations in the supplementary information (see Figure S3 and Table S2), we may conclude that this component corresponds with the spectrum we associate with the LP state. No spectra of the isolated second component were reported by Kozachuk et al. They interpreted the spectra for samples containing a mixture of the two states as arising from an additional component with axial spin-Hamiltonian parameters labeled b. Simulations in the supplementary information suggest that spectra from samples containing a mixture of NP and LP V-doped MIL-53(Al) can indeed be (mis)interpreted in this way. The isolated EPR component associated with the NP state, however, does not match this second axial component b at all. This EPR spectrum thus appears to be due to a center with lower than axial symmetry. Our recent analysis of the EPR spectrum of VIV=O centers in MIL-53(Al/V)AS has shown that obtaining reliable spin-Hamiltonian parameters involves recording EPR spectra in several microwave bands, and may even involve ENDOR and/or pulsed EPR techniques.19 In particular this holds true for the spectra corresponding to the NP and LP states. However, these spectra are sufficiently different to distinguish them, even without such detailed spin-Hamiltonian analysis. Transformation between NP and LP states in vacuum The correlation between the two types of VIV=O EPR spectra and the two states of activated MIL-53(Al) is further explored by recording EPR spectra at low pressure during in situ heating in a high-temperature EPR resonator. For starting the temperature (T) experiment, the sample


11


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

Page 12 of 35

was brought in the NP state as described above (see Figure 2 left bottom trace). The EPR spectra were recorded while continuously pumping the quartz sample tube to ~1 Pa. The temperature was raised from RT (300 K) to 450 K and back to RT, in steps of 25 K, after keeping the sample for ~0.5 h at each temperature in order to ensure stabilization. In Figure 2 right, the results of in situ T-dependent synchrotron XRD experiments on the same sample in vacuum (continuous pumping) are shown. In these experiments XRD patterns were monitored every 60s while continuously ramping up (and afterwards down) the temperature. The experiment was completed in 1 h. In view of the considerable difference in heating speed, sample mass and sample tube dimensions, all influencing the heat transfer, a perfect correspondence between the T axes of the two experiments is not expected. The similarities between the results in Figure 2 are, nonetheless, obvious: 1. In the early stages of the heating experiment the EPR spectrum remains (practically) unaltered and the XRD pattern is that of the NP state. 2. There is a transition T-range where EPR spectra of the two VIV=O states are simultaneously present, and the XRD patterns contain contributions of both the NP and the LP states. 3. From ~400 K onwards (when raising the temperature), only the VIV=O EPR spectrum associated with the LP state is further observed. The EPR spectrum remains basically unaltered when slowly bringing the temperature back to RT. Meanwhile the XRD pattern of the LP state of MIL-53(Al) is observed.


12

Page 13 of 35

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


Figure 2. Left: Powder EPR spectra of activated 1.5% V-doped MIL-53(Al) in vacuum recorded at four temperatures. The complete T-dependent EPR experiment is reported in supporting information, Figure S4. Insets show a magnification of the high-field part of the spectra. The arrow indicates start and end EPR spectrum together with associated states. Right: Temperature dependence of synchrotron XRD patterns of activated 1.5% V-doped MIL-53(Al) in vacuum, λ = 0.825086 Å. 2θ is converted to d-spacing in Å via Bragg’s law. Miller indices for all highintensity peaks are indicated, both for the NP and LP states of MIL-53(Al). The sample was at RT in the NP state, after the heating experiment the XRD pattern of the LP state of MIL-53(Al) is observed. The arrows indicate temperatures where EPR spectra presented on the left were recorded.


13


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

Page 14 of 35

Detailed analysis of the XRD patterns shows that the LP state appears at 344 K and the NP state disappears at 424 K. The bulk of the transition occurs between 350-400 K which is in agreement with reported values for MIL-53(Al).16,22 In addition to this essentially identical information that can be extracted from both experiments, the XRD T-experiment reveals an effect that is not obvious in EPR. Between 320 K and 350 K, the diffraction peaks (110), (111 ) and (220) of NP MIL-53(Al) exhibit a clear shift to smaller d-spacings (higher 2θ values, Bragg’s Law) while the NP peaks (200) and (400) shift to slightly larger d-spacings (lower 2θ values). This points to a gradual change in the lattice parameters, resulting in a decrease of the unit cell volume, confirmed by Lebail refinements on the experimental data (see Figure 3 and supporting information, Figure S5 and Table S3). For the LP state none of the diffraction peaks exhibits similarly clear temperature dependence.

Figure 3. Unit cell volume of the NP state as a function of temperature during the heating experiment. Error bars represent 1σ. Transformation between NP and LP states in air A similar in situ T-experiment of EPR and XRD measurements was performed while keeping the sample in contact with air at ambient pressure and humidity. The results are summarized in


14

Page 15 of 35

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


Figure 4. In the first half of the experiment (RT → 450 K) the results are nearly identical to those with the sample in vacuum. The line width of the EPR spectrum associated with the LP state is, however, slightly larger than when the sample is kept in vacuum. This change in line width is most apparent in the center of the spectrum, which exhibits slightly higher resolution when it is recorded in vacuum. In the XRD patterns, the shift of the NP state (110), (111 ), (220) peaks to lower d-spacings, observed in the vacuum experiment, is nicely reproduced. The XRD Texperiment for pristine MIL-53(Al) showed similar gradual change to smaller d-spacings (Supporting information, Figure S6) pointing to the fact that this effect is not induced by vanadium doping.

Figure 4. Left: Powder EPR spectra of activated 1.5% V-doped MIL-53(Al) in air recorded at four temperatures. The complete T-dependent EPR experiment is reported in supporting information, Figure S7. Insets show a magnification of the high-field part of the spectra. The


15


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

Page 16 of 35

arrow indicates start and end EPR spectrum together with associated states. Right: Temperature dependence of XRD patterns of activated 1.5% V-doped MIL-53(Al) in air, λ = 1.54056 Å. 2θ is converted to d-spacing in Å via Bragg’s law. Miller indices for all high-intensity peaks are indicated for both the NP and LP state of MIL-53(Al). The sample was at RT in the NP state, at approximately 425 K full transition to the LP state occurred. Cooling down the sample to RT, the XRD pattern of the NP state of MIL-53(Al) starts appearing again. The arrows indicate temperatures where EPR spectra presented on the left were recorded. Also in the second half of the experiment (450 K → RT) the XRD experiments in air and in vacuum exhibit no important differences, except that under air conditions, close to RT, part of the sample returns to the NP state most probably as a result of uptake of water from the air. The NP state XRD pattern obtained is the same as in the starting conditions of the experiment. On the other hand, the EPR spectra reveal an interesting additional feature. From about 375 K downwards, the line width of the EPR spectra gets broader as temperature lowers, while the spectrum (peak positions) remains that associated to the LP state. This broadening cannot be related to changes in the MIL-53(Al) lattice parameters, since no evidence is found in the XRD patterns for such changes. When the air atmosphere is removed by pumping, the narrow-line EPR spectrum as obtained in the vacuum experiment is recovered. Finally, in the top EPR spectrum of Figure 4 left no clear evidence of the NP state is seen, in contrast with the XRD result. In view of the differences in geometry between the experiments, one may expect that longer exposures to air are necessary in the EPR experiments in comparison with XRD. When monitoring the EPR spectrum at RT for a longer time after ending the T-experiment, the NP component clearly grows in (see Supporting information, Figure S8).


16

Page 17 of 35

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


Figure 5 left shows RT EPR spectra of a sample with the MIL-53(Al) lattice predominantly in the LP state, as a function of air pressure. Increasing the air pressure clearly increases the line width of the component associated with the LP state. Meanwhile the NP component, present as a trace, hardly experiences any broadening. In Figure 5 right the EPR spectra as a function of N2 pressure are shown. The spectra for both LP and NP states experience virtually no broadening by exposure to N2. We checked that humid and dry air both lead to a strong broadening of the LP spectral component, while dry and humid N2 do not. The line broadening should thus mainly be due to an exchange interaction between the VIV=O centers and paramagnetic O2 molecules.

Figure 5. Left: RT EPR spectra of activated V-doped MIL-53(Al) in the LP state, as a function of air pressure. Right: RT EPR spectra of activated V-doped MIL-53(Al) in the LP state, as a function of N2 pressure.


17


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 18 of 35

Discussion We have shown that the breathing effect in V-doped MIL-53(Al), induced by varying the temperature above RT, can be monitored by in situ EPR spectroscopy. This yields complementary information to the XRD analysis. The EPR spectra reflect the local environment of the VIV=O molecular ion. Apparently, the EPR spectrum of paramagnetic VIV=O complexes substituting for Al-OH follows the global, reversible LP to NP transitions of the activated MIL53(Al) crystals. In a first approximation, one may expect this to correspond to the geometry of the distorted (O-V=O)O4 octahedron, which apparently follows that of the (OH-Al-OH)O4 octahedra in the activated MIL-53(Al), rather than remaining fixed in itself, as in the case of the fully V=O substituted rigid lattice MIL-47. The hypothesis that EPR probes the environment of the VIV=O nodes very locally is confirmed by the absence of changes in the NP EPR spectra in the heating part of the T-experiments under vacuum and air conditions. XRD demonstrates that the host lattice may undergo global structural changes, corresponding to shifts in the d-spacing in the order of a few percent (in the region from approximately 320 K to 350 K) without notable influence on the NP X-band EPR spectrum of the VIV=O dopant molecular ions. A Lebail refinement of the XRD peak positions shows that these changes mainly correspond with an expansion of the a parameter and a contraction of b, together leading to a slight decrease of the unit cell volume (Figure 3). A metastable intermediate NP state is reached before the transition to the LP state occurs. Because such strong changes in the lattice parameters are absent for the LP state, they are very probably related to interactions with the H2O molecules remaining in the pores that induce the contraction to the NP state. In situ XRD studies of MIL-47(VIII) and various MIL-53(M) variants (M=Al, Ga, Cr, Fe, Sc)8,23-25 have revealed either step-wise or continuous changes in the diffraction patterns when


18

Page 19 of 35

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


following NP to LP transitions. In case of MIL-47(VIII)26, MIL-53(Fe)24 and MIL-53(Sc)25 the NP to LP transition happens via a dehydrated and an intermediate state where half of the pores are completely open and half are completely closed. In MIL-53(Ga)23 a contraction of the pores is observed before the NP to LP transition occurs and has been associated with a dehydrated NP state. Neimark et al.12 reported slight shifts of the XRD peaks during CO2 sorption in MIL53(Cr), that have also been associated with changes in the configuration of the guest molecules. The particular continuous change in lattice parameters to a metastable intermediate NP state observed here for MIL-53(Al) has so far, to the best of our knowledge, not been reported. Cooling the sample down from 450 K to RT, the situations in air and in vacuum are not completely the same. In vacuum the sample stays in the LP state since water cannot reenter the pores to contract them. In air a fraction of the sample contracts to the NP state close to RT within a couple of minutes. In addition the EPR results reveal an exchange interaction between VIV=O and O2 molecules in the LP state. This interaction is already observed at T>375 K but grows spectacularly for T

In Situ Electron Paramagnetic Resonance and X-ray Diffraction

Recommend Documents