A Multifrequency Transient Nutation EPR Study - ACS Publications

Feb 4, 2016 - ... Activation of CH4: A Multifrequency Transient Nutation EPR Study ... in the catalytic reaction network of the oxidative coupling of ...
0 downloads 0 Views 1MB Size
Subscriber access provided by NEW YORK UNIV

Article

Evidence for Exchange Coupled Electrons and Holes in MgO after Oxidative Activation of CH4: A Multi-Frequency Transient Nutation EPR Study Pierre Schwach, Maik Eichelbaum, Robert Schlögl, Thomas Risse, and Klaus-Peter Dinse J. Phys. Chem. C, Just Accepted Manuscript • DOI: 10.1021/acs.jpcc.5b11057 • Publication Date (Web): 04 Feb 2016 Downloaded from http://pubs.acs.org on February 8, 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 32

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

Evidence for Exchange Coupled Electrons and Holes in MgO after Oxidative Activation of CH4: A Multi-Frequency Transient Nutation EPR Study

P. Schwach, M. Eichelbaum, R. Schlögl Department of Inorganic Chemistry Fritz-Haber-Institut der Max-Planck-Gesellschaft Faradayweg 4-6, D-14195 Berlin, Germany

T. Risse Institute of Chemistry and Biochemistry, Freie Universität Berlin Takustrasse 3, D-14195 Berlin (Germany)

K.-P. Dinse Department of Physics, Freie Universität Berlin Arnimallee 14, D-14195 Berlin (Germany) Email: [email protected]

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 The effective spin of paramagnetic centers generated during the reaction of a mixture of CH4 and O2 on activated MgO was determined with Transient-Nutation EPR (TN-EPR). For the first time it could be shown that apart from the generation of surface-trapped superoxide radicals O2-, a significant amount of paramagnetic centers are formed with an effective spin larger than 1/2. Results from HYSCORE and other pulsed EPR experiments further give evidence that these centers are localized at the surface of the MgO catalyst. These centers are proposed to be exchange coupled clusters of holes and electrons, respectively, some of them formed at specific exposed sites of the highly structured MgO surface. Judging from the relative high concentration of the centers ascribed to hole clusters, it is tempting to assume that they are involved in the catalytic reaction network of the oxidative coupling of methane.

ACS Paragon Plus Environment

Page 2 of 32

Page 3 of 32

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

The Journal of Physical Chemistry

Introduction The quest to identify active sites for a better understanding of catalytic reactions has stimulated research over nearly a century. In the particular case of metal and metal oxide catalysts, the important role of low-coordinated atoms at step edges or corners has recently been identified1. With the renewed interest of the chemical industry in processing natural gas to more valuable compounds such as higher hydrocarbons (Fischer-Tropsch) or alkenes (oxidative coupling of

methane (OCM)), the simplest catalysts for the latter reaction, MgO, has obtained new attention. Quite some time ago, Lunsford proposed a model for the active site, particular of Li-doped 2-4

MgO

. In this model suboxide ions (O-) stabilized by the formation of charge compensating Li+-O-

pairs substituting for Mg2+-O2+ ions, play an important role for the primary hydrogen abstraction process in methane5-6. In particular, O- species were shown to be active in hydrogen abstraction in

hydrocarbons and molecular hydrogen7. The Lunsford model was later challenged as it is inconsistent with a variety of experimental observations showing that under standard reaction

conditions no such Li+-O- pairs can be observed8,9. Various experimental studies clearly show, however, that oxygen surface sites are part of the active center. In addition a wealth of data has been collected by EPR about paramagnetic sites in the bulk and at the surface of MgO10-18. This included in particular paramagnetic superoxide O2- radicals, which can be generated via different routes e.g. photo-chemically or by reaction of H2/O2 mixtures19-22. In a recent study it was

observed that superoxide O2- can be formed by addition of CH4 and O2 over MgO catalyst already at 300 K23-24. A detailed study of the local structure of these radicals using one- and twodimensional EPR will be described in a forthcoming publication. In the course of these pulsed EPR experiments two more paramagnetic sites were identified, which did not exhibit the characteristic behavior of S = 1/2 radicals. It was noted quite early25

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

that focusing on the idea of an “isolated” site might be misleading, because clusters or arrays of defect stabilized electrons or holes at the surface might also be involved, providing a convenient reservoir for charge transfer in a catalytic reaction. This aspect was also later noted by Lunsford in his review article3. Based upon high resolution electron energy loss spectroscopy (HREELS) of model Li doped MgO catalysts, Goodman and coworker26 concluded that F-centers (an oxygen ion vacancy with two trapped electrons) or F-center aggregates function as the active centers in LiMgO. Therefore, it is fair and important to emphasize that these centers are part of the reaction network and hence should be considered as another aspect which can play an important role for the atomistic understanding of the OCM process. Performing an experimental proof for the existence of such clusters is difficult, because weak magnetic dipolar or strong electronic exchange couplings of such sites would evade detection by optical or photo electron spectroscopy. The distinction between weakly coupled S=1/2 centers (for instance by magnetic dipolar interaction) and strongly coupled sites (by electronic exchange coupling) can be made by the magnitude of interaction compared with the Zeeman energy of the system. If the Zeeman energy is dominant, the response of the interacting spins is still determined by the spin of the individual sites, however, modified for instance by dipolar broadening. Only if the coupling terms are much larger than the Zeeman energy, the response is determined by the total effective (coupled) spins. An interaction of such magnitude can only be reached by electronic “exchange” coupling. By determination of their total spin, EPR is not only capable to prove the presence of exchange coupled clusters but may also allow for a determination of the number of constituting sites27. Given the large number of EPR studies of MgO

based catalysts it is astonishing that the possibility of identifying exchange coupled (VO.)+ sites (trapped electron at oxygen vacancy, F+ centers in “old” notation) or clusters of (VMg´´-OO.)(trapped hole at oxygen ion close to Mg vacancy) by their total spin has not yet been discussed in literature. Kröger-Vink notation is used to identify the charge state of the various species discussed here.

ACS Paragon Plus Environment

Page 4 of 32

Page 5 of 32

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 the present study we will present clear evidence that exchange coupled sites are present at the surface of polycrystalline MgO catalysts, formed upon exposure to a gas mixture of CH4 and O2. The determination of the effective spin is in principle possible by observing nuclear spin transitions in the mS dependent hyperfine field28, or by observing higher order spin operators using high frequency EPR29-30. A more direct method, however, is based on observing the Rabi nutation frequency within the multitude of electron spin levels. For this purpose a specific twodimensional EPR technique (Transient Nutation EPR, TN-EPR) is used, mapping the Rabi nutation frequency of the paramagnetic centers along one axis, as function of the EPR absorption pattern (second axis), scanned by the magnetic field31-32. The principle of the TN-EPR spectroscopy is based on the fact that the nutation frequency of spin magnetization in the rotating frame of an oscillating microwave (mw) field is proportional to the transition moment between spin sublevels on resonance. Within specific limits33, depending on the relative size of spin Hamiltonian parameters compared to the irradiating mw amplitude, the nutation frequency between |S, mS> → |S, mS+1> in terms of the reference nutation frequency νref of a S =1/2 spin state, is given as 1/ 2

ν = ( S ( S + 1) − mS (mS + 1) ) ν ref

(1)

With this method the effective spin of different species can unambiguously be determined and assigned to their respective EPR spectra, which appear separated in the resulting 2D plot. In addition to the spin state the local environment of the spin bearing sites was investigated exploring hyperfine interactions with nuclear spins of the surrounding nuclei. In particular another 2D EPR

technique (HYSCORE)34 was used, probing for nuclear spins in the neighborhood of the paramagnetic centers35.

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

Experimental 1) Sample preparation As starting material ultra pure MgO (Alfa Aesar, Puratronic (R), 99.998% (m.b) Ref. 10800, lot:24686) was used. In a first step the compound was calcined in synthetic air at 1123 K for 6 h. The specific surface area after calcination at 1123 K is relatively small and was determined by N2 physisorption as 54 m2/g. The side length of the crystallites observed by TEM is approximately 20-30 nm, which is in agreement with the mean crystallite size of 23 nm as deduced from an analysis of the XRD peak width. The primary particles are of regular cubic morphology exposing (100) faces with sharp edges and corners. The height of the steps that often appear close to the edges of the cubes is in the nanometer range. Smaller steps, in the range of just a few nm and even in the dimension of one Mg-O distance (mono-atomic steps) can regularly be observed. The primary particles are agglomerated to form clusters up to 5 µm in size36. MgO with modified surface properties was prepared by heating MgO in a 4 mm ID quartz tube to 1073 K and exposing it to a flow of 90 ml/min of 20% H2O, 2% O2 in N2 for 90 hours. In the following text this MgO modification is labelled as “sintered”. After this treatment in water vapor, the specific surface area is reduced to 18 m2/g and the average mean crystallite size (determine by XRD) increases to 66 nm. For EPR spectroscopy, 50 – 100 mg MgO was placed in a quartz tube (3.9 or 2.9 mm o.d.for Xand Q-band experiments, respectively) closed at one end and connected to a vacuum system through high vacuum valves allowing evacuation of the cell as well as applying the reactant gases. So called “activated” samples were prepared by heating the MgO sample to 1073 K under dynamic vacuum for 6 hours (end pressure ca. 10-6 mbar), defining the "activation" phase. After activation, the samples are exposed to a mixture of 100 mbar CH4 (or 13CH4, CD4, respectively) ACS Paragon Plus Environment

Page 6 of 32

Page 7 of 32

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

and 50 mbar O2 at the desired reaction temperature for 30 min, followed by evacuation of the cell to a pressure of 5.10-6 mbar or lower. After cooling to room temperature in dynamic vacuum, the cell was disconnected from the gas feed and HV lines using Young HV valves and transferred to a pump for finally sealing off under HV conditions. To avoid possible contamination with oxygen, samples could also be prepared avoiding the transfer to a separate HV pump.

2) EPR spectroscopy Field swept EPR spectra were recorded by echo detection (FSE-EPR), typically using a 2-pulse mw sequence (20-200-40 ns), resulting in an absorptive type spectrum. FSE recording is in particular suited for measuring broad, unstructured EPR spectra. Experiments at X-band frequencies (9.7 GHz) were performed using a commercial Bruker ElexSys 680 spectrometer equipped with an Oxford Helium cryostat. Sample tubes of 3.9 mm o.d. were inserted in a commercial Bruker Flexline ENDOR probe head. 34 GHz experiments were performed with a Bruker ElexSys E580 spectrometer. For Q-band experiments sample tubes of 2.9 mm o.d. were used, which were studied with a home-built cavity (F. Lendtzian, TU Berlin). Pulsed experiments were typically performed in the range of 5 – 30 K. Determination of the EPR susceptibility was performed using a sample prepared according to the “standard” protocol (i.e., activation at 1073 K in dynamic vacuum, then at 300 K addition of 50 mbar O2 and 100 mbar CH4, then evacuation) compared to a Cr3+/MgO dilute solid solution standard on a cw Bruker ESR 300 E spectrometer at 77 K. The exact concentration of Cr3+ was determined by PGAA (Prompt Gamma activation analysis). Even starting with ultra-pure educts the samples show trace amounts of impurities after the multistep treatment performed to prepare the catalyst and run the reaction. Previous ICP

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

measurements provide a concentration of 0.2 ppm of Cr and 3 ppm of Mn. Such concentrations are well within the detection limit of EPR spectroscopy. The TN-EPR measurements were performed using the PEANUT pulse sequence (Phase inverted Echo Amplitude NUTation) as introduced by Stoll et al.32 This has the advantage that all echoes are detected at the same time, thereby minimizing loss of resolution because of relaxation broadening, as well as minimizing distortions arising from spectrometer dead time. Microwave (mw) field amplitude variations across the sample, which would lead to nutation frequency variations, are also refocused. The mw pulse sequence used is shown in Scheme 1. The mw amplitudes for the preparation pulse and the nutation pulses were chosen as equal. The width of the π/2 pulse was 20 ns in X and Q band, the initial High Turning Angle (HTA)x pulse was 32 ns, the total length T set to 512x4 ns, the initial delay τ was 300 or 100 ns, respectively

Scheme 1: The PEANUT pulse sequence. The first pulse selects spins of spectral width determined by the pulse width. The High Turning Angle (HTA) nutation pulse of total length T is split into two parts with opposite phase. Incrementing the phase boundary of the nutation pulse leads to the formation of a rotary echo, which is maximized for t = T/2, and can be detected at T + 2τ. The second part of the composite nutation pulse is 180° out of phase with respect to the other pulses.

ACS Paragon Plus Environment

Page 8 of 32

Page 9 of 32

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

. For each field setting the intensity of the echo at T+2τ is detected as function of t. For suppression of the spin locked responses maximized at t = 0 and t = T, sine-bell windowing is used before Fourier transformation. In the two-dimensional HYSCORE experiment correlations are probed between nuclear frequencies in different electron spin states (usually mS = +/- 1/2), which are generated by the first two pulses of the sequence separated by τ (fixed). Choice of τ is critical, because “blind spots” in the spectral pattern are generated at odd multiples of ν = 1/2τ. In HYSCORE, the time between the second (π/2) and third (π) pulse is varied in one dimension and the time between the π and next π/2 pulse is varied in a second dimension, the effect of the π pulse being inter-converting nuclear spin coherences between the two electron spin levels. Measurement protocol and data processing have been described in detail in a previous publication35.

Results 1) FSE-EPR spectra The red trace in Figure 1 shows the FSE spectrum of the MgO reacted with a mixture of CD4/O2 at 370 K. A rather strong signal dominates the spectrum whose shape and position is characteristic of O2- radicals. The signal (red trace) is obtained, when optimizing the mw pulse power for the maximal echo response of the radical ion, known to have a spin S = 1/2. The observed signal shows the expected powder pattern of centers with dominant g matrix anisotropy. Using data recorded with higher spectral resolution it could be shown by spectral simulation that at least 4 different superoxide radicals contribute to the spectrum. The set of gzz fit values for these differently positioned radicals are found as (2.093; 2.088; 2.080; 2.075), respectively. A

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 10 of 32

more detailed description of the observed signals will be presented in a forthcoming publication, however it is important to mention at this point that these O2- radicals are strongly coupled to either a proton or deuteron abstracted from the methane as deduced from experiments using CH4 or CD4, respectively. Using CH4 as substrate, a similar response is obtained at room temperature. The observed isotope effect proves that O2- binds directly to the hydrogen abstracted from the methane and is hence intimately connected to the catalytic reaction pathway. In the following we will use the term O2- radicals as an abbreviation of these hydrogen-coupled species. It was noticed, however, that a narrow spike at 1215 mT could be optimized in amplitude (lower trace, blue) relative to the O2- signal by reducing the mw power by 10 dB (keeping the pulse length settings unchanged). Under this condition the response of the S = 1/2 spin system nearly vanishes, which is expected because of the reduction of the corresponding rotation angles in the π/2 − τ − π pulse sequence by a factor of 101/2. It was also noted that the weak signal at 1229 mT could also be optimized at lower mw power settings. This observation indicated that the narrow signals originate from centers with an effective spin higher than 1/2. For further confirmation of the presence of “high spin” signals, samples were prepared at different reaction temperatures (300 K, 370 K, 470 K, and 1020 K) and using isotopically labeled reactants (CD4, 13CH4). In all cases the unexpected narrow signal (1215 mT at 34 GHz) could be detected. Its detection was not only possible using the high sensitivity 34 GHz spectrometer, but could also conveniently be performed at standard X-band frequencies, as is shown in Fig. 2. In this figure a sample is studied, which was exposed to the reaction gas mixture at 1020 K. The red trace shows the EPR absorption with mw power settings optimized for the O2- radical (S = 1/2). The narrow signal close the high field edge of the superoxide radical signal is clearly visible even under “unmatched” excitation conditions. Using a microwave power reduced by 101/2, the

ACS Paragon Plus Environment

Page 11 of 32

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

response of the O2- signal nearly vanishes, leaving a narrow peak as observed at 34 GHz. A sextet of weak signals with a splitting of 8.6 mT is also detected, which allows unambiguous assignment of these lines to Mn2+ impurities. In Fig. 3 EPR spectra are compiled showing the effect of reacting MgO with a mixture of CH4 and O2 at different temperatures. For reference the EPR spectra of MgO before and after activation are also shown. It is documented that the formation of O2- radicals is already observed when performing the reaction at room temperature. Performing the reaction at 1020 K the superoxide signal nearly vanishes. The signal shown in Fig. 3 is due to a reaction with trace amounts of oxygen during EPR sample preparation. The generation of the narrow signal is strongly favored at elevated temperatures, although it can also be detected as weak signal in samples reacted at room temperature. No trace of this signal is found in the “activated” and “as received” MgO, proving that that this species is formed during the reaction with CH4/O2.

ACS Paragon Plus Environment

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 32

electron spin echo amplitude (arb. units)

The Journal of Physical Chemistry

1150

1170

1190

1210

1230

1250

magnetic field (mT)

Fig. 1: 34 GHz FSE-EPR spectra of MgO reacted with CD4/O2 at 370 K. For this particular sample a somewhat elevated reaction temperature was chosen in order to improve the concentration of trapped O2-. Microwave pulse sequence used was 24 – 300 – 48 ns, detection at 30 K. Intensity of both traces can be directly compared. The red trace is recorded using a mw power setting optimized for S = 1/2 spin, the blue trace is taken with 10 dB less mw power.

ACS Paragon Plus Environment

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

electron spin echo amplitude (arb. units)

Page 13 of 32

-

O2

Mn

310

320

330

340

2+

350

360

370

380

magnetic field (mT) Fig. 2: 9.7 GHz FSE-EPR spectra of MgO reacted with CH4/O2 at 1020 K. Microwave pulse sequence used was 24 – 300 – 48 ns, detection at 30 K. The red trace (shifted for clarity) is recorded using the mw power setting optimized for the S = 1/2 spin, the blue trace is taken with 10 dB less mw power.

ACS Paragon Plus Environment

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

electron spin echo amplitude (arb. units)

The Journal of Physical Chemistry

Page 14 of 32

CH4/O2 reacted at 1020 K

CH4/O2 reacted at 300 K

"activated" "as received"

280

320

360

400

magnetic field (mT) Figure 3: 9.7 GHz FSE-EPR spectra of MgO catalysts prepared using different preparation protocols (see text). The red traces are recorded using the mw power setting optimized for S = 1/2 spins, the blue trace is taken with 10 dB less mw power. Spectra were taken at 30 K with mw pulse setting 20 – 300 – 40 ns. The relative intensity of all spectra can directly be compared because being recorded with identical amplifier settings.

2) TN-EPR spectra From the analysis of the FSE-EPR response to mw amplitude settings it was concluded that the FSE spectra arise from a superposition of spectra of paramagnetic sites with differing effective spin were present. To identify the effective spin and to disentangle the corresponding EPR pattern, a series of TN-EPR spectra were recorded. In a first example, a 9.7 GHz TN-EPR spectrum of an MgO sample reacted at room temperature is shown in Fig. 4. In the 2D plot the signal of the O2- radical is clearly identified at a nutation frequency of 10(1) MHz. The narrow

ACS Paragon Plus Environment

Page 15 of 32

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

spike observed previously by FSE-EPR is now detected as separated peak at 346 mT/32(2) MHz. In the 2D TN-EPR plot an additional broad pattern extending from 320 mT up to 345 mT is detected, with a nutation frequency different from both the S=1/2 (O2- radical) as well as from the narrow signal around 346 mT. This absorption previously evaded detection with standard continuous wave (cw) and even with FSE techniques, probably because being weak and lacking narrow structures. Its full range is shown in Fig. 5. The contribution of paramagnetic centers with effective spin different from 1/2 is impressive, considering the apparent relative intensity in particular of the broad component nutating at 16(2) MHz. For further characterization, TN-EPR spectra were recorded at 34 GHz (see Fig. 6). In this sample, which was reacted with CH4/O2 at 1020 K, the broad absorption at lower field values is even more prominent as compared to the result measured with a sample reacted at room temperature. When comparing with results measured at 9.7 GHz, it should be noted that the nutation frequencies differ because of limited microwave power. The “reference” nutation frequency of the O2- (S = 1/2) radical in the 34 GHz experiment is 6.5(5) MHz. As already seen in the 34 GHz FSE-EPR spectrum shown in Fig. 1, further peak is detected in the 2D plot at 1230 mT/13(1) MHz. (Slightly differing field positions quoted for the Cr3+ line are caused by differing mw frequencies used in the various experiments) Using higher spectral resolution for FSE detection (Fig. 7), this signal, which is centered at g = 1.980 as calibrated by the weak 55Mn(II) lines, can unambiguously be assigned to trace amounts of Cr3+ in its S = 3/2 spin state because of resolved 53Cr hyperfine interaction (hfi).37 Its presence can thus be used for an additional nutation frequency reference in the 2D plot. Five weak signals of the expected hfi sextet of 55Mn(II) are also detected as indicated in Fig. 9. These lines not only serve as convenient field marker, allowing determining g values of the “high spin” and Cr3+ lines as 2.003 and 1.980, respectively,

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 16 of 32

but are also providing a nutation frequency reference, originating from the center mS =(-1/2, 1/2) transition of the S = 5/2 spin state of Mn(II).

Fig. 4: 9.7 GHz FSE and TN-EPR spectra of CH4/O2 reacted MgO (Tr = 300 K). The EPR spectral range of O2- is indicated in both spectra. The vertical black line correlates the peak in the 2D plot with the narrow signal assigned to a species with effective spin larger than 1/2.

ACS Paragon Plus Environment

Page 17 of 32

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

Fig. 5: Full range 9.7 GHz TN-EPR spectrum of 300 K reacted sample depicted in Fig. 3.

Fig. 6: 34 GHz TN-EPR spectrum of CH4/O2 reacted MgO (Tr = 1020 K). ACS Paragon Plus Environment

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

electron spin echo signal (arb. units)

The Journal of Physical Chemistry

Page 18 of 32

3+

Cr

1.8 mT 50.4 MHz 8.8 mT 246 MHz

2+

Mn

1200

1210

1220

1230

1240

1250

magnetic field (mT) Fig. 7: 34 GHz FSE-EPR spectrum of CH4/O2 reacted “sintered” MgO (Tr = 300 K). The weak hfi satellites of 53Cr (9.5% nat. abundance) used to identify the Cr3+ signal are indicated. The g value 1.980 is in agreement with literature values37-38. The spectrum was recorded with mw power setting optimized for the “high spin” signal (-13 dB). A full range spectrum using different mw power settings is depicted in Fig. S4.

3) HYSCORE spectra In order to obtain information about the local environment of the paramagnetic sites in the samples, 2D HYSCORE spectra were measured. In particular we focused on the question if sites could be discriminated being positioned in the “bulk” or at the “surface”. For this purpose MgO was reacted with

13

C labeled CH4, anticipating that some carbon will be deposited on the surface

during the reaction providing a possible source for hyperfine interaction with paramagnetic species. In

Fig. 8 the results of HYSCORE and Echo Spin Echo Envelope Modulation (ESEEM) experiments are shown. Spectra were taken by observing at the peak of the narrow signal, adjusting the pulse settings for an optimal response of the high spin signal. As indicated, apart

ACS Paragon Plus Environment

Page 19 of 32

from a signal of

25

Mg, strong signals of the free nuclear frequency of

13

C and its multiples are

observed at the spectrum diagonal. Weak cross peaks are also detected, indicating coherence transfer between multi quantum excitations. The striking multi-quantum coherence of 13C nuclei observed indicates coupling of the “high spin” center to many equivalent carbon sites, which requires a spatially extended spin distribution. It should be noted that there is no indication of dipolar 25

hyperfine coupling to the

Mg nuclei in the HYSCORE or ESEEM plot, which is perfectly

consistent with a delocalized spin distribution.

ν(

(arb. units)

Ft of 3pESEEM signal

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

13

C)

2ν(

13

C)

3ν( ν(

25

M g)

13

C) 4ν(

13

C) 5ν(

0

5

10

13

C)

6ν(

15 20 fre q u e n c y (M H z )

13

C)

7ν(

25

13

C)

8ν(

13

30

C)

35

Fig. 8: HYSCORE and Fourier-transformed 3-pulse ESEEM signal measured at 9.7 GHz. MgO catalyst reacted at 1020 K with 13CH4/O2.

For characterization of the broad signal centered at B = 1100 mT at 34 GHz, HYSCORE spectra were also recorded although the signal intensity compared to that of the narrow peak at 1236 mT 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 20 of 32

was quite low. A peak at the free 25Mg frequency was observed, confirming our assumption that these paramagnetic centers are formed in MgO (see Fig. S3, supporting information).

Discussion 1) Superoxide radical O2In the present publication this signal, consisting of a superposition of at least 4 superoxide radicals being localized at different sites, mainly serves as reference for the nutation frequency νref of an effective S = 1/2 spin system. The integrated intensity can also be used to estimate the

concentration of the various paramagnetic centers. In a sample reacted with CH4/O2 at room temperature under standard conditions (100 mbar CH4, 50 mbar O2), the concentration of O2radicals was determined as 1.2.1017 spins/g by comparing with a Cr3+ standard. With approximately 50 m2/g specific surface area, this results in 2.5.1015 spins/m2. Assuming statistical distribution, an average distance of 20 nm between radicals is predicted. Here it should be noted that this estimate neglects complications which might arise from the affinity of the O2- radicals to specific corner or edge sites39, leading to much higher local concentrations. It should also be noted that superoxide radicals are not observed for samples reacted at 1020 K and subsequently permanently kept under HV conditions. This is in agreement with the prediction that these radicals are not stable at high temperatures. As apparent from the TN-EPR plots, the nutation frequency of all O2- radicals is identical and constant over their full field range. Such a behavior is expected if the g value anisotropy ∆g/g (determining the relative width of the spectrum) is only of the order of a few percent. A single horizontal ridge further indicates that isolated radicals are observed with dipolar or exchange ACS Paragon Plus Environment

Page 21 of 32

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

couplings smaller than the pulse excitation width40, in agreement with the estimated rather low surface density.

2) Cr3+ impurity The peak detected at 1230 mT/13(1) MHz, unambiguously assigned to arise from Cr3+, is well separated from all other signals in the 2D plot. The signal therefore can serve as a convenient reference value in the 2D plot of the TN-EPR spectra. The 2D TN spectrum only shows a single peak at twice the reference νref, in perfect agreement with the predicted nutation frequency of the mS (-1/2, 1/2) transition of the quartet spin state. This indicates that the “extreme low mw condition” is met, i.e., not all transitions of the S = 3/2 spin system are excited simultaneously with the mw pulse excitation width of ca. 15 MHz. The powder pattern caused by non-vanishing ZFS must be large enough to spread the signals over a wide field range, preventing detection of an expected nutation frequency (3)1/2νref, which is not surprising considering the low concentration of Cr3+ centers. As has been demonstrated in a study invoking the quartet spin state of N@C60 as local symmetry sensor, even small effects by a phase transition in a regular crystal will lift the degeneracy of the allowed EPR transitions28. The assumption of a finite ZFS is in agreement with the results of Codling and Henderson37, who determined a zero field splitting value D in the order of 9 GHz with a single crystal study. The signal was modeled assuming a tetragonal distortion caused by charge compensation by a next neighbor Mg vacancy. The estimated experimental value was later reproduced by a DFT/superposition model41. The important conclusion of the present 2D TN analysis is that no undistorted octahedral Cr3+ sites are present in our MgO sample, because for such high symmetry sites with vanishing zero-fieldsplitting37 all mS (3/2,1/2) and mS (-1/2, 1/2) transitions would be superimposed. In such a case 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 22 of 32

all signal components are predicted nutating at a common frequency31-32, equal to the S = 1/2 nutation reference frequency (1227 mT, 6.5 MHz in our case). Such a signal is completely absent. Cr3+ impurities are apparently situated at distorted lattice sites, which points to a local nature of charge compensation. As stated above, the Cr3+ signal in combination with the O2- pattern create references in the 2D landscape and give an impression about the accuracy by with nutation frequencies can be determined. 3) Narrow “High Spin” signal (g = 2.003) At the field setting of 346 mT (Fig. 4, 5) and 1217 mT (Fig. 6), a peak at 3νref is observed in the 9.7 GHz and 34 GHz TN-EPR spectra, indicating nutation of a ∆mS (-1/2, 1/2) transition of a sextet spin state (S = 5/2) (see Fig. 6). Because of partial spectral overlap with the superoxide signal (S = 1/2), it cannot safely be excluded that additional, smaller nutation frequencies are also present at this field setting. It is possible that the 2D TN spectrum for this center is obtained under “intermediate” conditions, meaning that the mw excitation width is comparable to the powder pattern width, originating from ZFS of this high spin center. The assignment of this signal to an effective spin S = 5/2 is supported by the simultaneous observation of hyperfine (hf) lines of trace amounts of Mn2+ in the 34 GHz TN-EPR spectrum shown in Fig. 9. Probing a narrow field range, five weak narrow, equally spaced signals can be detected at a nutation frequency of 21(1) MHz. The observed field difference of 8.6 mT between these signals gives clear evidence that they arise from 55Mn(II) in its S = 5/2 state. The predicted increase by a factor three of the observed nutation frequency of the ∆mS (-1/2 → 1/2) transition compared to the S = 1/2 reference of O2- is well reproduced. The coincidence of the position of the strong narrow peak at 1213 mT (Fig. 9) with the horizontal line defined by the 55Mn signals confirms its assignment to an effective S = 5/2 spin center. It should be noted that different field ACS Paragon Plus Environment

Page 23 of 32

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

positions quoted in Figs. 4-6,9 are caused by slightly differing mw frequencies used for the respective experiments. In all cases an invariant g value was determined.

Fig. 9: 34 GHz TN-EPR spectrum of MgO reacted at 1020 K with 13CH4/O2. The sextet of narrow peaks at a nutation frequency of 20(1) MHz can unambiguously be assigned to Mn2+. Their position in the 2D plot serves as further reference for the determination of the effective spin of the strong peak at 1213 mT.

Neutron activation data gave no hint for the presence of transition metal impurities which can be associated with this signal. One is left either with assuming the presence of carbonaceous structures, clusters of trapped holes (O- centers, OO.), or trapped electrons (VO.) created under reaction conditions. Strong 13C signals at the free nuclear frequency were observed after reaction with 13C labeled CH4 at elevated reaction temperature (1020 K). At this reaction temperature the color of the sample changes from white to slightly greyish. At first sight the strong 13C hyperfine 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 24 of 32

coupling renders it likely that this species is associated with the formation of an extended carbon

network forming the high spin state at the surface of the catalyst. However, no characteristic Raman signals of an aromatic carbon network could be detected even for the samples prepared at 1020 K and showing a greyish color. The complete lack of isotropic and/or anisotropic hfi with 25

Mg and

13

C (in case of

13

CH4 reactants) and in particular lack of coupling with 1H rather

indicates that the high spin state is created by exchange coupled surface (S=1/2) sites. The lack of g anisotropy indicates the absence of spin-orbit coupling, favoring the model of electrons trapped at oxygen vacancies or at morphological defects such as kinks or reversed corners. The coupling with many

13

C nuclei as demonstrated by multi-quantum coherences in the HYSCORE pattern

leading to the observation of multiples of free

13

C NMR frequencies also indicates that the

unknown high spin state is formed at the surface of MgO. The spectacular involvement of many carbon sites suggests co-localization of the trapped electron cluster with carbon deposits formed under reaction conditions. The lack of detectable 13C hfi is also in agreement with the suggested significant delocalization. The absence of noticeable dipolar hfi with 25Mg as documented in Fig. 8 is in contrast to the observation of Chiesa et al.42, who described such effects for various centers ascribed to localized electrons. In turn, the observed coupling to 25Mg lacking hfi can be taken as further evidence for the delocalized nature of the spin center.

4) Broad “High Spin” signal (g = 2.20) The broad peak centered at 1100 mT at 34 GHz (g = 2.20) observed in the TN spectrum is extending from 1050 to 1150 mT, well separated from the absorption pattern of the other paramagnetic centers (see Fig. 6). Its nutation frequency is larger by a factor 1.5(1) compared to the S = 1/2 reference. This value does not seem to match to any factors predicted for effective

ACS Paragon Plus Environment

Page 25 of 32

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

spins in the range of S = 1 to 5/2. Considering the significantly increased g value, which is larger by 10 % with respect to our S = 1/2 reference value, the observed value might comply with 21/2, if considering this g value difference40. However, lacking further information about the spin Hamilton parameters of this center, an effective spin of 3/2 cannot be excluded. In this context it should be mentioned that a narrow structure was observed at the center of the broad pattern, indicating the presence of still a different nutation frequency. Thus the pattern is tentatively assigned to be caused by an effective quartet spin. It should be noted that we could not detect “half field” or “third field” transitions in 34 GHz FSE-EPR spectra, indicating a relatively small zero-field-splitting parameter (see Fig. S1, supporting information). This is in agreement with the observation that the absorption pattern is mainly determined by g anisotropy as is seen by comparing 9.7 and 34 GHz spectra (see Fig. 10). In Figure 10 FSE-EPR spectra of the MgO sample reacted at 1020 K with

13

CH4/O2, measured at 9.7 and 34 GHz are depicted, using a g

value scale. In such a representation, signals with dominant g matrix parameters show up at identical positons. This is clearly seen for the superoxide absorption, as well for the “high spin” signal at g = 2.003 and the Cr3+ resonance. The broad pattern, well defined in the 34 GHz data, centered at g = 2.2, can still be seen in the 9.7 GHz data set, but its structure is broadened. This can be caused in principle by significant hfi, or by ZFS interaction. Hfi can be excluded because of missing off-diagonal signals in the HYSCORE plot. We can therefore attribute the broadening to arise from ZFS, which is expected to be non-vanishing for S>1/2. It should be noted that the integrated intensity of this signal is of the same order of magnitude as the integrated intensity of the “reference” O2- signal (see Figs. S1 and S4), suggesting that it is playing a major role in the catalytic process. In contrast, the high spin signal at g = 2.003 has a much lower abundance by approximately two orders of magnitude.

ACS Paragon Plus Environment

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 26 of 32

electron spin echo signal (arb. units)

The Journal of Physical Chemistry

2.6

2.4

2.2

2.0

1.8

g value Fig. 10: 9.7 GHz (blue trace) and 34 GHz (red trace) FSE spectra of MgO reacted at 1020 K with 13

CH4/O2 depicted using a g value scale. The full range 34 GHz spectrum is also shown in Fig. S1.

The large positive g shift indicates that spin orbit coupling is effective, which is consistent with the assumption that oxygen sites are involved. Thus the signal might be interpreted as arising from holes trapped close to an Mg vacancy. In this context it is interesting to find that a model assuming a Mg vacancy at a corner of a mono-atomically high (001) island of MgO is consistent with the localization of positive charge at nearest neighbor oxygen sites, when calculating the spin and charge distribution under the restriction of a total spin S = 3/2 as shown in Fig. S5. The anticipated large shift of the g matrix elements with respect to the free electron value (g = (2.100 2.186 2.284)) was also reproduced, using the GIAO routine of Gaussian. We are well aware of the fact that the single point calculation will not provide numbers directly comparable with the experiment, nor is this model calculation meant to prove the existence of the particular species in the system. The

ACS Paragon Plus Environment

Page 27 of 32

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

structure model shown is only meant as first attempt towards a realistic description of the observed clusters and should initiate further modelling invoking advanced quantum chemical methods.

Conclusion Invoking a specific two-dimensional EPR method, the effective spin of paramagnetic centers generated during the reaction of a mixture of CH4 and O2 on activated MgO was determined. For the first time it could be shown that apart from the generation of surface-trapped superoxide radicals O2-, additional paramagnetic centers are formed with effective spin larger than 1/2. A very broad and a narrow absorption pattern is observed, centered at g = 2.20 and 2.003, respectively. Results from HYSCORE and pulsed EPR experiments give further evidence that these centers are localized at or close to the surface of the MgO catalyst. They probably have to be described as exchange coupled clusters of electrons and of holes, trapped at oxygen vacancies or morphological defects, formed at specific exposed sites of the highly structured MgO surface. Assuming the validity of this hypothesis, the large positive g shift observed for the broad structure, probably caused by spin-orbit coupling of holes trapped at oxygen sites, suggests an assignment to exchange coupled holes. These “positive charges” are probably found around Mg vacancies. The observed g shift is qualitatively reproduced for a fictitious quartet spin state formed around an exposed Mg vacancy at a mono-step corner, thus supporting the presented model. The narrow signal at g = 2.003 close to the free electron value instead is ascribed to trapped electrons at oxygen vacancies, lacking the contribution of spin-orbit coupling.

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 28 of 32

Judging from the relative high concentration of the centers ascribed to hole clusters, it is tempting to assume that they are involved in the catalytic reaction network, providing a convenient charge reservoir. At present no model for the catalytic cycle can be presented. The probable impact of such clusters originates from the formation of a dense electronic level scheme, allowing facile transfer and storage of electrons like in bio-catalytic centers, quite often consisting of exchange-coupled transition metal ions. Further experiments using

17

O labeled reactants are required to unravel details of the

catalytic reaction steps.

Acknowledgements This work was performed within the Cluster of Excellence “Unifying Concepts in Catalysis”. We thank Dr. A. Trunschke and Dr. J. Noak for providing UV/Vis and Raman spectra. We gratefully acknowledge use of the EPR equipment of the research group of Prof. R. Bittl. DFT Modelling was performed using the High Performance Computing facility of the FU Berlin.

Supporting Information FSE-EPR and HYSCORE spectra of various samples are compiled, which are used for additional characterization of various paramagnetic centers. A figure of the calculated spin density distribution ascribed to the broad signal centered at g = 2.2 is also shown. This information can be found free of charge at htto://pubs.acs.org.

ACS Paragon Plus Environment

Page 29 of 32

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

References 1. Zambelli, T.; Wintterlin, J.; Trost, J.; Ertl, G., Identification of the "Active Sites" of a SurfaceCatalyzed Reaction. Science 1996, 273, 1688-1690. 2. Wang, J.-X.; Lunsford, J. H., Characterizatlon of [Li+O-] Centers Ln Lithium-Doped Mgo Catalysts. J. Phys. Chem. 1986, 90, 5883-5887. 3. Lunsford, J. H., The Catalytic Oxidative Coupling of Methane. Angewandte Chemie 1995, 34, 970980. 4. Driscoll, D. J.; Martir, W.; Wang, J.-X.; Lunsford, J. H., Formation of Gas-Phase Methyl Radicals over Mgo. J. Am. Chem. Soc. 1985, 107, 58-63. 5. Ito, T.; Lunsford, J. H., Synthesis of Ethylene and Ethane by Partial Oxidation of Methane over Lithium-Doped Magnesium Oxide. Nature 1985, 314, 721-722. 6. Ito, T.; Wang, J.-X.; Lin, C.-H.; Lunsford, J. H., Oxidative Dimerization of Methane over a LithiumPromoted Magnesium Oxide Catalyst. J. Am. Chem. Soc. 1985, 107, 5062-6068. 7. Aika, K.-l.; Lunsford, J. H., Surface Reactions of Oxygen Ions. 1. Dehydrogenation of Alkanes by 0on Mgo. J. Phys. Chem. 1977, 81, 1393-1398. 8. Arndt, S.; Laugel, G.; Levchenko, S.; Horn, R.; Baerns, M.; Scheffler, M.; Schlögl, R.; Schomäcker, R., A Critical Assessment of Li/Mgo-Based Catalysts for the Oxidative Coupling of Methane. Catalysis Reviews 2011, 53, 424-514. 9. Myrach, P., et al., Temperature-Dependent Morphology, Magnetic and Optical Properties of LiDoped Mgo. ChemCatChem 2010, 2, 854-862. 10. Chiesa, M.; Giamello, E.; Annino, G.; Massa, C. A.; Murphy, D. M., Multi-Frequency High-Field Epr Study of (H+)(E-) Pairs Localized at the Surface of Polycrystalline Mgo. Chem. Phys. Lett. 2007, 438, 285289. 11. Sterrer, M.; Berger, T.; Diwald, O.; Knözinger, E.; Allouche, A., Ozonide Ions on the Surface of Mgo Nanocrystals. Topics in Catalysis 2007, 46, 111-119. 12. Chiesa, M.; Paganini, M. C.; Spoto, G.; Giamello, E.; Valentin, C. D.; Vitto, A. D.; Pacchioni, G., Single Electron Traps at the Surface of Polycrystalline Mgo: Assignment of the Main Trapping Sites. J. Phys. Chem. B 2005, 109, 1314-1322. 13. Valentin, C. D.; Ricci, D.; Pacchioni, G.; Chiesa, M.; Paganini, M. C.; Giamello, E., O- Radical Anions on Polycrystalline Mgo. Surface Science 2002, 521, 104-116. 14. Murphy, D. M.; Farley, R. D.; Purnell, I. J.; Rowlands, C. C.; Yacob, A. R.; Paganini, M. C.; Giamello, E., Surface Defect Sites Formed on Partially and Fully Dehydrated Mgo: An Epr/Endor Study. J. Phys. Chem. B 1999, 103, 1944-1953. 15. Giamello, E.; Paganini, M. C., A Combined Epr and Quantum Chemical Approach to the Structure of Surface Fs+(H) Centers on Mgo. J. Phys. Chem. B 1997, 101, 971-982. 16. Chen, Y.; Abraham, M. M., Trapped-Hole Centers in Alkaline-Earth Oxides. Journal of Physics and Chemistry of Solids 1990, 51, 747-764. 17. Simon, U.; Arndt, S.; Otremba, T.; Schlingmann, T.; Görke, O.; Dinse, K.-P.; Schomäcker, R.; Schubert, H., Li/Mgo with Spin Sensors as Catalyst for the Oxidative Coupling of Methane. Catal Commun 2012, 18, 132-136. 18. Zavyalova, U.; Weinberg, G.; Frandsen, W.; Girgsdies, F.; Risse, T.; Dinse, K.-P.; Schloegl, R.; Horn, R., Lithium as a Modifier for Morphology and Defect Structure of Porous Magnesium Oxide Materials Prepared by Gel Combustion Synthesis. Chemcatchem 2011, 3, 1779-1788. 19. Gonchar, A.; Risse, T.; Freund, H. J.; Giordano, L.; Di Valentin, C.; Pacchioni, G., Activation of Oxygen on Mgo: O2.- Radical Ion Formation on Thin, Metal-Supported Mgo(001) Films. Angew Chem Int Edit 2011, 50, 2635-2638.

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 30 of 32

20. Siedl, N.; Elser, M. J.; Halwax, E.; Bernardi, J.; Diwald, O., When Fewer Photons Do More: A Comparative O2 Photoadsorption Study on Vapor-Deposited Tio2 and Zro2 Nanocrystal Ensembles. J. Phys. Chem. C 2009, 113, 9175–9181. 21. Diwald, O.; Knözinger, E., Intermolecular Electron Transfer on the Surface of Mgo Nanoparticles. J. Phys. Chem. B 2002, 106, 3495-3502. 22. Anpo, M.; Che, M.; Fubini, B.; Garrone, E.; Giamello, E.; Paganini, M. C., Generation of Superoxide Ions at Oxide Surfaces. Top Catal 1999, 8, 189-198. 23. Tashiro, T.; Watanabe, T.; Kawasaki, M.; Toi, K.; Ito, T., Partial Oxidation of Methane with Oxygen over Magnesium-Oxide at Low-Temperatures. J Chem Soc Faraday T 1993, 89, 1263-1269. 24. Schwach, P.; Hamilton, N.; Eichelbaum, M.; Thum, L.; Lunkenbein, T.; Schlögl, R.; Trunschke, A., Structure Sensitivity of the Oxidative Activation of Methane over Mgo Model Catalysts: Ii. Nature of Active Sites and Reaction Mechanism. Journal of Catalysis 2015, 329, 574-587. 25. Boudart, M.; Delbouille, A.; Derouane, E. G.; Indovina, V.; Walters, A. B., Activation of Hydrogen at 78°K on Paramagnetic Centers of Magnesium Oxide. J. Amer. Chem. Soc. 1972, 94, 6622-6630. 26. Wu, M. C.; Truong, C. M.; Coulter, K.; Goodman, D. W., Investigation of Active Sites for Methane Activation in the Oxidative Coupling Reaction over Pure and Li-Promoted Mgo Catalysts. J. Catalysis 1993, 140, 344-352. 27. Dinse, A.; Carrero, C.; Ozarowski, A.; Schomäcker, R.; Schlögl, R.; Dinse, K.-P., Characterization and Quantification of Reduced Sites on Supported Vanadium Oxide Catalysts by Using High-Frequency Electron Paramagnetic Resonance. Chemcatchem 2012, 4, 641-652. 28. Weiden, N.; Käss, H.; Dinse, K.-P., Pulse Electron Paramagnetic Resonance (Epr) and ElectronNuclear Double Resonance (Endor) Investigation of N@C60 in Polycrystalline C60. J Phys Chem B 1999, 103, 9826-9830. 29. Dinse, A.; Ozarowski, A.; Hess, C.; Schomäcker, R.; Dinse, K.-P., Potential of High-Frequency Epr for Investigation of Supported Vanadium Oxide Catalysts. J Phys Chem C 2008, 112, 17664-17671. 30. Eichel, R. A.; Meštrić, H.; Dinse, K.-P.; Ozarowski, A.; van Tol, J.; Brunel, L. C.; Kungl, H.; Hoffmann, M. J., High-Field/High-Frequency Epr of Paramagnetic Functional Centers in Cu2+- and Fe3+-Modified Polycrystalline Pb[Zrxti1-X]O3 Ferroelectrics. Magn Reson Chem 2005, 43, S166-S173. 31. Takui, T.; Sato, K.; Shiomi, D.; Itoh, K.; Kaneko, T.; Tsuchida, E.; Nishide, H., Ft Pulsed Esr/Electron Spin Transient Nutation (Estn) Spectroscopy Applied to High-Spin Systems in Solids; Direct Evidence of a Topologically Controlled High-Spin Polymer as Models for Quasi Id Organic Ferro- and Superparamagnets. Mol. Cryst. Liq. Cryst. 1996, 279, 155-176. 32. Stoll, S.; Jeschke, G.; Willer, M.; Schweiger, A., Nutation-Frequency Correlated Epr Spectroscopy:The Peanut Experiment. J. Magn. Resonance 1998, 130, 86-96. 33. Ayabe, K., et al., Pulsed Electron Spin Nutation Spectroscopy of Weakly Exchange-Coupled Biradicals: A General Theoretical Approach and Determination of the Spin Dipolar Interaction. Physical chemistry chemical physics : PCCP 2012, 14, 9137-48. 34. Höfer, P.; Grupp, A.; Nebenführ, H.; Mehring, M., Hyperfine Sublevel Correlation (Hyscore) Spectroscopy: A 2d Esr Investigation of the Squaric Acid Radical. Chemical Physics Letters 1986, 132, 279282. 35. Dinse, A.; Wolfram, T.; Carrero, C.; Schlögl, R.; Schomäcker, R.; Dinse, K.-P., Exploring the Structure of Paramagnetic Centers in Sba-15 Supported Vanadia Catalysts with Pulsed One- and TwoDimensional Electron Paramagnetic Resonance (Epr) and Electron Nuclear Double Resonance (Endor). J Phys Chem C 2013, 117, 16921-16932. 36. Schwach, P.; Frandsen, W.; Willinger, M.-G.; Schlögl, R.; Trunschke, A., Structure Sensitivity of the Oxidative Activation of Methane over Mgo Model Catalysts: I. Kinetic Study. Journal of Catalysis 2015, 329, 560-573. 37. Codling, A. J. B.; Henderson, B., Paramagnetic Resonance Studies of V2+ and Cr3+ in Orthorhombic Symmetry Sites in Magnesium Oxide. J. Phys. C: Solid St. Phys. 1971, 1, 1242-1250. ACS Paragon Plus Environment

Page 31 of 32

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

38. Low, W., Paramagnetic Resonance and Optical Absorption Spectra of Cr3+ in Mgo. Phys. Rev. 1957, 105, 801-806. 39. Giamello, E.; Ugliengo, P.; Garrone, E., Superoxide Ions Formed on Mgo through the Agency of Presorbed Molecules. J. Chem. SOC., Faraday Trans. I 1989, 85, 1373-1382. 40. Ayabe, K., et al., Pulsed Electron Spin Nutation Spectroscopy for Weakly Exchange-Coupled Multi-Spin Molecular Systems with Nuclear Hyperfine Couplings: A General Approach to Bi- and Triradicals and Determination of Their Spin Dipolar and Exchange Interactions. Molecular Physics 2013, 111, 2767-2787. 41. Wan-Lun, Y., Local Distortion of the Orthorhombic Charge-Compensation Defect Sites in Cr3+:Mgo. J. Phys. Chem. Solids 1998, 59, 261-263. 42. Chiesa, M.; Paganini, M. C.; Giamello, E.; Murphy, D. M.; Valentin, C. D.; Pacchioni, G., Excess Electrons Stabilized on Ionic Oxide Surfaces. Acc. Chem. Res. 2006, 39, 861-867.

ACS Paragon Plus Environment

The Journal of Physical Chemistry

toc graphic

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

ACS Paragon Plus Environment

Page 32 of 32