Insights into the Geometric and Electronic Structure of Transition Metal

Article pubs.acs.org/accounts

Insights into the Geometric and Electronic Structure of Transition Metal Centers from Valence-to-Core X‑ray Emission Spectroscopy Published as part of the Accounts of Chemical Research focus issue “Understanding Heterogeneous Chemical Processes Using X-ray Techniques”. Christopher J. Pollock† and Serena DeBeer*,‡,§ †

Department of Chemistry, The Pennsylvania State University, University Park, Pennsylvania 16802, United States Max-Planck-Institute for Chemical Energy Conversion, Stiftstrasse 34-36, D-45470 Mülheim an der Ruhr, Germany § Department of Chemistry and Chemical Biology, Cornell University, Ithaca, New York 14853, United States ‡

CONSPECTUS: A long-standing goal of inorganic chemists is the ability to decipher the geometric and electronic structures of chemical species. This is particularly true for the study of small molecule and biological catalysts, where this knowledge is critical for understanding how these molecules effect chemical transformations. Numerous techniques are available for this task, and collectively they have enabled detailed understanding of many complex chemical systems. Despite this battery of probes, however, challenges still remain, particularly when the structural question involves subtle perturbations of the ligands bound to a metal center, as is often the case during chemical reactions. It is here that, as an emerging probe of chemical structure, valence-to-core (VtC) X-ray emission spectroscopy (XES) holds promise. VtC XES begins with ionization of a 1s electron from a metal ion by high energy X-ray photons. Electrons residing in ligand-localized valence orbitals decay to fill the 1s hole, emitting fluorescent photons in the process; in this manner, VtC XES primarily probes the filled, ligand-based orbitals of a metal complex. This is in contrast to other X-ray based techniques, such as K-edge X-ray absorption and EXAFS, which probe the unoccupied d-manifold orbitals and atomic scatterers surrounding the metal, respectively. As a hard X-ray technique, VtC XES experiments can be performed on a variety of sample states and environments, enabling application to demanding systems, such as high pressure cells and dilute biological samples. VtC XES thus can offer unique insights into the geometric and electronic structures of inorganic complexes. In recent years, we have sought to use VtC XES in the study of inorganic and bioinorganic complexes; doing so, however, first required a thorough and detailed understanding of the information content of these spectra. Extensive experimental surveys of model compounds coupled to the insights provided by DFT calculated spectra of real and hypothetical compounds allowed the development of a framework whereby VtC XES spectra may be understood in terms of a molecular orbital picture. Specifically, VtC spectra may be interpreted as a probe of electronic structure for the ligands bound to a metal center, enabling access to chemical information that can be difficult to obtain with other methods. Examples of this include the ability to (1) assess the identity and number of atomic/small molecule ligands bound to a metal center, (2) quantify the degree of bond activation of a small molecule substrate, and (3) establish the protonation state of donor atoms. With this foundation established, VtC has been meaningfully applied to long-standing questions in bioinorganic chemistry, with the potential for numerous future applications in all areas of metal-mediated catalysis.

■

INTRODUCTION Over the past several years, metal Kβ X-ray emission spectroscopy (XES) has undergone a period of rapid development. During this time, the technique has evolved from one used principally by the condensed matter physics community to one with quantitative applications to a wide variety of chemical systems. Some of the most striking advances have occurred with respect to the analysis and interpretation of the valence-to-core (VtC) region of these spectra. This high energy region was observed at least as early as the 1930s,1,2 during which time authors noted its origins in the valence, ligand-localized orbitals.3−8 However, its applications remained © XXXX American Chemical Society

largely outside the realm of chemical investigation until relatively recently. The advent of synchrotron lights ourceswhich provide an intense, tunable source of X-raysallowed the practical measurement of these weak VtC features from molecular systems on reasonable time scales. Coupled to high resolution analyzers, synchrotron beamlines have enabled the expansion of VtC XES into the chemical sciences. Making use of these practical developments, pioneering work by Bergmann, Glatzel, Received: June 21, 2015

A

DOI: 10.1021/acs.accounts.5b00309 Acc. Chem. Res. XXXX, XXX, XXX−XXX

Article

Accounts of Chemical Research

Figure 1. Schematic of the instrumentation for a typical XES measurement. The Rowland circles are depicted at left; the value of θ is varied to scan in energy. Adapted with permission from ref 12. Copyright 2005 Elsevier.

As the transitions of interest in the VtC region are of relatively low intensity (∼1000× weaker than the 2p to 1s Kα lines), most modern VtC experiments are performed using the high intensity beams provided by synchrotron lightsources. The instrumentation for a VtC XES experiment is similar to that used for synchrotron XAS measurementsincluding maintaining the sample at cryogenic temperatureexcept, because the incident X-ray beam does not need to be monochromatic, XES experiments can be performed with the higher intensities provided by “pink” beam; this also enables these experiments to be performed at X-ray free electron laser (XFEL) sources. Unlike in XAS, however, standard solid state detectors lack the energy resolution needed for XES measurements (>100 eV versus ∼1 eV required), and so selection of the fluorescent photon energy is accomplished using a multicrystal array spectrometer employing Bragg optics.12,17−20 The sample, analyzer, and detector are most commonly arranged in a Rowland circle geometry (Figure 1), wherein the energy is determined by the angle θ between the sample and analyzers;12 thus, the selected energy can be modified simply by translating the analyzers and detector along the circle to change θ. Spectrometers in such an arrangement can achieve resolutions below 1 eV21a value that compares well with the 1s core hole lifetimes of first row transition metalswith even better resolutions possible at the expense of count rate.22 Dispersive arrangements, whereby an entire spectrum is collected at once using a position sensitive detector, are also becoming increasingly common.23−27 As a hard X-ray technique, XES has the benefit of being applicable to a wide range of sample environments, including solids, single crystals,11 liquids, dilute solutions,28 and high pressure cells.29 Sample temperature can also be dynamically controlled, allowing for temperature dependent phenomena to be explored. The inherently low intensity of VtC features and relatively small solid angle collected by the analyzers impose some limitations on how dilute the species of interest can be, though the high flux available at third generation synchrotron sources and free electron lasers has enabled spectra to be measured on protein solutions with metal concentrations as low as 800 μM.28

Cramer, and co-workers recognized the potential that VtC spectra had to answer questions presented by chemical systems and laid much of the groundwork for the application of this technique to inorganic chemistry.9−11 The ultimate goal of our work has been to apply the insights available from VtC XES to complex systems in inorganic and bioinorganic chemistry. To do so, however, required an understanding of the chemical information content of these spectra and appropriate techniques to analyze and interpret the data. We have built upon the existing foundation of understanding and, from the perspective of an inorganic chemist, have attempted to fully elucidate their information content. In addition to experimental studies, a key tool in exploring VtC spectra has been the application of density functional theory (DFT). DFT calculated spectra have been shown to have excellent agreement with experimental data,13−15 allowing the insights from chemical theory to be applied to the analysis of VtC XES. In these relatively simple calculations, the ground state electronic structure is used to compute transitions between the valence orbitals and the metal 1s orbital; inclusion of a metal 1s core hole was found not to improve agreement with experiment.15 The intensities of the VtC transitions are calculated by summing the electric dipole (>95% of the total intensity),13,14 magnetic dipole, and electric quadrupole oscillator strengths, while the energies are computed using the difference between the ground state 1s and valence orbital energies.13 The relative intensities of the VtC features generally match well with experiment, while the absolute energies obtained from these calculations are systematically shifted from the experimental values by a significant amountover 100 eV is not uncommonthat is strongly dependent on the level of theory used (e.g., functional, basis set, inclusion of relativistics). The relative energies of the transitions, however, are very well produced and thus the energy scale can be effectively calibrated to the experiment by a scalar shift. Such calibration studies exist for a number of transition metals, including Cr,16 Mn,14 and Fe.13

■

■

EXPERIMENTAL OVERVIEW Valence-to-core XES uses transitions of electrons between occupied orbitals and a metal 1s core hole to probe the valence of a metal complex. Experimentally, high energy incident X-ray photonswith energies well above the metal K absorption edgeare used to ionize a 1s electron from the metal atom, leaving the absorber with the required 1s core hole. When the decaying electrons originate from donor orbitals with dominantly ligand character, the result is a VtC transition.

MOLECULAR ORBITAL FRAMEWORK For some time, VtC XES spectra of inorganic complexes have been known to exhibit great variability in terms of the number of features present and their respective intensities, though the source of these differences remained largely unexplored. As we sought to apply this technique to complexes of chemical and biological interest, we first needed an understanding of the information contained within these spectra. B


Article


Figure 2. Qualitative MO diagram for an MX6 complex (left) and the corresponding calculated VtC XES spectrum for FeCl63− (right). Representative orbitals giving rise to each of the transitions are also shown to demonstrate the t1u σ nature of the interactions; the small sticks at high energy reveal the negligible intensity contributed by the Fe−Cl π interaction and the Fe 3d orbitals. The small splitting observed in the transitions within each peak is due to spin polarization of the α and β orbitals.

rise to more intense VtC features,9,13 since shorter bonds inherently allow for greater interaction with the metal p orbitals. This simple example also highlights the sensitivity of VtC XES to the identity of atoms bound to a metal center. Using the MO diagram in Figure 2, if the ionization energies of the ligand orbitals are shifted to higher absolute energy, for example, then the resulting MOs of the complex will also be correspondingly shifted to higher energy. Thus, the transitions observed in VtC XES will be sensitive to the ionization energies and hence the identities of the atoms bound to the metal.9 The case is most favorable for ligand s energiesthose of N, O, and F are separated over a range of more than 8 eV9enabling straightforward discrimination between these donors in many cases. This sensitivity is in contrast to EXAFS, where light atoms generally cannot be distinguished within Z ± 1. This ability is retained even in unusual chemical environments, such as multimetallic clusters.31 So robust is this discriminatory power that VtC XES has already been applied to complex bioinorganic systems, where bridging oxo ligands were identified in the S1 state of photosystem II28 and the central atom in the MoFe cofactor of nitrogenase was identified as a carbon.32 The simple Oh MO diagram in Figure 2 can of course be extended to more complex and interesting systems. Perhaps the most beautiful example of these is provided by ferrocene (Figure 3).33 In this conventionally D5d compound, the cyclopentadienyl ligand MOs combine to produce six ligand SALCs that can interact with the metal np orbitals (a2u and e1u), giving rise to a large number of allowed VtC transitions. The richly featured experimental spectrum matches this prediction exactly, with every expected transition visible (Figure 3). It is, then, clear that VtC XES is capable of providing detailed information about the electronic structures of ligands bound to a metal center.

Transition intensity in XES spectra has been shownboth experimentally and computationallyto be governed by the dipole selection rule,11,13−15 whereby transitions from the donor orbitals gain intensity via mixing with small amounts of metal np character. The dipolar nature of these transitions has important implications for spectral interpretationindeed, it is one of the overarching themes governing the interpretation of VtC spectrabecause it determines which ligand-localized orbitals give rise to VtC features. Perhaps the simplest case with which to understand these implications is the molecular orbital (MO) structure of the perfect octahedral MX6, as shown in Figure 2. In Oh symmetry, the metal p orbitals transform as a triply degenerate t1u set; thus, for any ligand orbitals to gain significant intensity in a VtC spectrum, they too must transform as t1u. In the case where the ligands are simple halide ions, the ligand s orbitals transform as a1g, eg, and t1u, while the p orbitals become a σ interacting set of a1g, eg, and t1u and a π interacting set of t1g, t2g, t1u, and t2u. These symmetry adapted linear combinations (SALCs) provide three opportunities for significant interaction with the metal np orbitals: One for the ligand s orbitals and one each for the σ and π sets of p orbitals. The energetics of these orbitals are largely dictated by the electronic structure of the ligands themselves, with smaller perturbations introduced by (1) the electrostatic charge on the metal and (2) direct orbital mixing with the metal-based orbitals (particularly in the d manifold).30 From the calculated spectrum, however, only two peaks are observedone in the low energy Kβ” region and another in the higher energy Kβ2,5 region (Figure 2); corresponding peaks can be seen in the experimental spectrum.13 Analysis of the character of the donor MOs reveals that only those that can interact with the metal in a σ fashion have appreciable metal np mixing, hence only two peaks are observed. Study of a variety of compounds has established this observation to be quite general: If a ligand-based MO is expected to have significant VtC intensity, it must interact with the metal via a σ interaction.30 Furthermore, it is generally true that the more metal p character a ligand orbital has, the greater the VtC transition intensity,13,30 though further discussion on this point is warranted (vide infra). Intuitively, it makes sense that the ligand s orbitals would be more localized on the ligands themselves and thus less available to interact with the metal, while the p orbitals project outward and can interact much more strongly. This also nicely rationalizes the observation that shorter bondssuch as those present in low spin and high valent complexestend to give

■

PROBE OF LIGAND ELECTRONIC STRUCTURE It is in this role as a probe of ligand electronic structure that VtC XES becomes immensely useful for answering difficult questions in (bio)inorganic chemistry. After all, the reaction of an inorganic center with a bound substrate molecule involves oftentimes subtlestructural modifications made to the ligand(s), precisely the attributes to which VtC XES is most sensitive. C


Article


Figure 4. Overlay of the background subtracted VtC spectra for the PDI iron complexes where the growth of the N2-derived features can clearly be seen.34 The VtC areas have been set equal to unity to facilitate comparison of the relative spectral changes.

what extentthese small molecule substrates are activated before being subsequently converted to products. This task has traditionally been the realm of vibrational spectroscopiesIR, Raman, NRVSthough many systems, particularly those in proteins, remain inaccessible with these methods. With its sensitivity to the electronic structure of the ligands, VtC XES has been shown to offer a unique window into questions concerning small molecule bond activation. During the course of activating a chemical bond, the distance between the atoms involved becomes increasingly large until no interaction remains. Such a process is reflected in the electronics of the ligand in question; as the distance between atoms increases, the overlap of the respective orbitals decreases until the MOs collapse to isolated atomic orbitals. The energetic changes involved are largeon the order of 10 eVsuggesting that they could be quantified by VtC XES if sufficient interaction with the metal np orbitals exists. A case study to this effect was performed on a series of N2bridged Fe dimers,35 each with a distinct and well-characterized degree of N2 bond activation. The spectra revealed a peak in the VtC region that varied in energy as the N−N bond was lengthened (Figure 5). This peak at around 7100 eV, which was assigned based on the MO argument presented above as deriving from the 2s2s σ* orbital of the N2 unit, shifted to lower energy as the N−N bond was lengthened and eventually cleaved to become two isolated N 2s atomic orbitals. Moreover, the energetic shift was found to vary quantitatively with the length of the N−N bond and also correlated extremely well with the bond’s vibrational frequency, validating VtC XES as a probe of bond activation. While this system is not catalytic, the underlying methodology holds for possible in situ studies. This sensitivity to ligand bond length is of course dependent on proper interaction of a suitable ligand MO with the metal np orbitals. The linear Fe−N−N−Fe unit of the complexes mentioned above resulted in optimal overlap of the N−N 2s2s σ* orbital with the Fe np, giving rise to the large observed intensity. Rotation of the diatomic into a bent configuration, however, would change the dominant donor orbitals on the ligand from the 2s2s σ to the 2p2p π orbitals, which would be expected to result in lower sensitivity to bond length changes and also less favorable energetic positioning of the resultant VtC features. Such effects were observed for a Mn-peroxo complex, where the bent coordination mode of the bridging peroxo unit failed to give rise to intense VtC features with

Figure 3. Experimental background subtracted VtC data and fit for ferrocene (top) and the DFT-calculated MO diagram showing the six orbitals that contribute to the spectrum (bottom). Adapted with permission from ref 33. Copyright 2011 American Chemical Society.

One straightforward example uses this ligand sensitivity to simply count the number of small molecules bound to a metal center. Given how critical a step the binding of a substrate and the order of substrate bindingcan be for a chemical transformation, identifying such an event is often desirable when attempting to elucidate a mechanism. Demonstrating such a binding can, however, be challenging, particularly if similar ligands are already bound to the metal center. Because the binding of a substrate to a metal should generally result in additional opportunities for metal np orbital mixing, it thus should be visible in the VtC spectrum. To test this hypothesis, an aryl-substituted bis(imino)pyridine (iPrPDI) iron complex, (iPrPDI)FeCl2, was converted to (iPrPDI)FeN2 and (iPrPDI)Fe(N2)2 via displacement of the chloride ligands with N2. These changes were clearly observed in the VtC spectrum, first by the appearance of a feature at 7101.3 eV for (iPrPDI)FeN2 that was not present for (iPrPDI)FeCl2 (Figure 4) and by this feature’s subsequent growth in intensity upon addition of the second N2. These observations confirmed the sensitivity of VtC XES to the number of ligands bound to a metal center.34 Once a substrate has bound to a metal, the next step is commonly to activate an intraligand bond in order to effect reactivity. An often critical piece of this puzzle is whenand to D


Article


oxo/μ-hydroxo, and finally to a bis-μ-hydroxo core while maintaining the same supporting ligands and metal oxidation state. The O 2s-derived peak, which is clearly visible in the bisμ-oxo case, decreases in intensity in the μ-oxo/μ-hydroxo and is completely absent for the bis-μ-hydroxo core (Figure 6).

Figure 6. Spectra of manganese dimers bridged by a bis-μ-oxo (black), μ-oxo/μ-hydroxo (red), and bis-μ-hydroxo (blue) core are shown. The O 2s peak at ∼6519 eV becomes progressively weaker as the bridges are protonated until it is no longer visible in the bis-μ-hydroxo case. Reproduced with permission from ref 38. Copyright 2013 American Chemical Society.

Smaller changes are also seen in the Kβ2,5, where a low-energy shoulder gains intensity with increasing protonation. Given the ubiquity of proton transfer chemistry in biology, these results strongly suggest that VtC XES could be used to probe these often difficult-to-capture events. These events are visible in VtC XES because protonation of a coordinating atom serves to both stabilize the ns and npz orbitals and also delocalize their density (Figure 7). The Figure 5. For a series of compounds possessing a bridging N2 unit, the 2s2s σ* feature at ∼7100 eV (solid blue peak) moves to progressively lower energy as the N−N bond is elongated (top and middle panels). For the compound in the bottom panel, the N−N bond is completely cleaved to give two nitrides. Reproduced with permission from ref 35. Copyright 2013 American Chemical Society.

positions that could be easily quantified.36 Sufficiently large electronic changes are also required for quantification: Attempts to distinguish between subtly different oxidation states of NO in a series of related {Fe−NO}7 complexes did not lead to unambiguous assignment.37 Another widespread theme in reactivityparticularly in bioinorganic chemistryis the (de)protonation of ligands during a reaction. Indeed, protonation and proton-coupledelectron-transfer steps are proposed to occur in many, if not most, enzymatic transformations, though identifying precisely when and where these events have occurred is generally not straightforward. The sensitivity of VtC XES to ligand electronic structure, however, offers a new tool to investigate the protonation states of ligands bound to metals. A recent experimental study of (hydr)oxo bridged manganese dimers beautifully demonstrated the practical application of XES to identifying ligand protonation state.38 The three dimeric compounds investigated possessed bridging cores that varied systematically from having a bis-μ-oxo, to a μ-

Figure 7. Upon protonation, the atomic s and p orbitals split into bonding and antibonding pairs, perturbing the energetics of the system. The electron density is also delocalized over the added proton, reducing the density available for interaction with a metal. Hence, protonation tends to lower both the energy and intensity of VtC features. Similar changes would occur for subsequent protonation steps.

increased stabilization shifts the Kβ″ (and the pz contribution to the Kβ2,5) to lower energy while the delocalization reduces its intensity due to a decreased ability to interact with the metal np orbitals. Numerous computational studies have addressed this sensitivity, both for simple inorganic complexes15 and also for larger, more complex systems.13,39,40 For example, in a series of hypothetical four-coordinate Fe(II) complexes, it was shown E


Article


any geometry that allows p−d hybridization. In this sense, VtC spectra also allow direct access to the metal electronic structure, even though the 3d orbitals are generally still extensively mixed with the ligands. Throughout the discussion above, the origin of the metal p characterfrom the 3p or the 4pwas left ambiguous. While it has been noted that the intensity of VtC transitions correlates with the total amount of metal p character present in the donor orbital, it is also clear that this is not a perfect correlation, with the largest deviations often occurring for Kβ″ features.30 Given that no significant quadrupole intensity is observed,13 this naturally engenders the question of whether the origin of this p character is behind these deviations: Is it coming from the metal 3p, 4p, or a combination of the two? This question is critical because the metal 3p orbitals have a much more contracted radial distribution and thus a much higher intrinsic transition dipole moment integral, so even a small amount of metal 3p character could have a dramatic influence on the spectra. On one hand, based on the total amount of metal p character calculated to be present in the ligand orbitals, we can conclude that some amount of metal 4p must be present.30 On the other, additional computational work has suggested that both the 3p and the 4p contribute.13,14 These observations were explored more fully in a study of iron carbonyl complexes.42 Therein, similar to previous observations, a simple linear correlation between intensity and metal p character could not be obtained. To investigate this, the transition dipole moment integrals were calculated for transitions from both Fe 3p and 4p orbitals; the values for the 3p integrals remained constant across oxidation states while those for the 4p varied strongly with both oxidation state and chemical environment. With these values in hand, the relative contributions of the 3p and 4p to each transition could be approximately deconvoluted (Figure 9). It was found that, while the majority of the calculated metal p character derived from the 4p orbitals, most of the intensity was a result of the small amount of 3p character present. Furthermore, the orbitals contributing to the Kβ″ were found to possess almost exclusively metal 4p character, consistent with the ligand 2s

that sequential protonation of a bound nitride resulted in dramatic intensity reduction and shifts to lower energy (∼2 eV per protonation) for the Kβ″ feature.39 This study also showed that the Fe XAS pre-edge was largely insensitive to these changes, supporting the notion that VtC XES has access to unique structural information. Similar computational observations were also made for a series of hypothetical Mn−NHx compounds.40

■

SUBTLETIES Until this point, the structural attributes to which VtC XES is sensitive have been those accompanied by relatively large energetic changes, such as addition of atoms/molecules to a metal or sizable changes in bonding. During our systematic studies of model compounds, however, we have identified more subtle geometric and electronic effects that can also be probed. For example, during the course of investigating a set of oxygen-bridged dimeric iron compounds, it was observed that the Kβ″ peak intensity did not correlate well with the Fe−O bond length.41 This was an unusual observation because the Kβ″ intensity is modulated by the amount of metal p character in the donor MO, which is principally dependent on the metal−ligand bond length. Careful study of these spectra by DFT revealed that there were, in fact, appreciable differences in metal np character that were independent of bond length. By invoking a Walsh-type picture, this difference was rationalized in terms of the Fe−O−Fe bond angle (Figure 8), whereby

Figure 8. By decreasing the Fe−O−Fe bond angle from 180°, the total amount of Fe np character mixed into the O 2s orbital increases, thus resulting in a more intense Kβ″ feature in compounds with smaller bond angles. Reproduced with permission from ref 41. Copyright 2014 American Chemical Society.

decreasing the angle from 180° results in an increase in metal np character, even at constant bond length; it was found that the Kβ″ intensity increased by 15% on going from a 180° core angle to a 90° angle. This was rationalized by the ability of Fe px/py mixing to occur once the Fe−O−Fe angle deviated from 180°. From these data it became apparent that, while VtC intensity generally increases with decreasing metal−ligand bond length, the intensity cannot be used to predict bond length in the absence of other structural information. This observation is especially important when attempting to use VtC data to deduce the structures of metalloenzyme active sites, since bond length and bond angle may introduce competing effects; this was found to be the case for the proposed cores of methane monooxygenase intermediate Q.41 Geometric effects can also lead to instances where the metal 3d manifold can be probed directly via VtC spectra. Namely, in geometries that deviate from centrosymmetry, direct mixing of the metal 3d and np orbitals becomes allowed, providing a mechanism by which the 3d orbitals can gain significant intensity. Sizable contributions from the 3d manifold has been observed in several cases42 and would in fact be expected for

Figure 9. The amount of Fe 3p and 4p character present in a donor orbital can be estimated based on the calculated transition oscillator strength compared to the transition dipole moment integrals for pure Fe 3p and 4p orbitals.42 Reproduced by permission of the Royal Society of Chemistry. F



■

Article

OUTLOOK AND FUTURE DIRECTIONS Given the VtC region’s sensitivity to chemical environment, one can begin to envision exploiting this for developing more complex experiments. One potential such use of the VtC is the collection of “ligand selective” XAS spectra, whereby an X-ray absorption spectrum is collected via detection of fluorescence from a VtC feature.43 Such an experiment is analogous to Kα or Kβ high energy resolution fluorescence detected (HERFD) XAS, wherein the crystal analyzer is set to the energy of a given emission feature and then the incident beam is scanned to collect an XAS spectrum. Similar experiments have already been demonstrated at the Kβ mainline for mixed manganese compounds44 and Prussian blue.45 VtC-detected XAS spectra have also been reported,46−48 though only recently was a systematic pilot study conducted with the intent of determining the level of sensitivity of these measurements.49 Preliminary data on this series of manganese compounds revealed that dramatic changes in pre-edge intensity, ranging from −33% to +62%, may be observed depending on the VtC feature chosen for detection (Figure 11). These effects are not yet well-

orbitals being spatially localized on the ligand and thus unable to interact with the contracted metal 3p. This observation rationalizes why, in the example presented in Figure 2, there is an average of 12.6% Fe p character calculated for the Kβ″ transitions and only 10.2% for the more intense Kβ2,5 transitions. The differential 3p/4p mixing may also explain why some experimental spectra fail to show any visible Kβ″ while others with similar calculated intensities do have this feature. Moreover, differences in VtC intensity are also observed between different metals, even when controlling for bond lengths. This was observed most quantitatively for Mn and Fe compounds, where Mn appeared to have inherently greater VtC areas.14 In this study, it was found that, while Mn and Fe have very similar atomic transition dipole moments, these dipole moments distort in a molecular environment. In the Kβ″ region, Mn was found to have significantly more metal np character mixed into the ligand orbitals relative to its Fe congener, despite having a somewhat reduced transition dipole moment. In the Kβ2,5 region the situation was reversed: The amount of metal np character was similar, but Mn possessed a much larger transition dipole moment. Both of these effects give rise to greater intensity in the Mn VtC spectra; this subtle interplay of factorsthe amount of metal np character mixed into the ligand and the distortion of those orbitalswill be at play for every transition metal and thus will impact the experimentally observed intensity. The description presented above, with VtC XES serving as a means of mapping the ligand MO electronic structure, captures most of the important features of these spectra. It would imply a false sense of completeness, however, to claim that all observed peaks can be fully explained within an MOframework; some spectra do indeed present features that cannot be rationalized with this methodology. Examples of this phenomenon include very low energy features observed in some metal oxides and fluorides as well as “shoulder” features that are sometimes observed on the sides of the Kβ2,5 peak, particularly in lower valent high spin compounds (Figure 10). It

Figure 11. An overlay of the valence-to-core detected and total fluorescence yield XAS spectra for KMnO4 reveals dramatic pre-edge intensity modulations. The large peak at lowest energy results from scatter of the incident beam. Reproduced with permission from ref 49. Copyright 2014 American Chemical Society.

understood and require further experimental and theoretical work, but they clearly present a possibility for exploiting the chemical sensitivity of the VtC region. These uncertainties notwithstanding, the picture that has emerged from our investigation of VtC XES is one of a technique that offers unique insights into the geometric and electronic structures of inorganic complexes, many of which would be difficult or impossible to obtain with standard analytical methods. The elucidation of these sensitivities should further broaden the range of systems to which VtC XES can be applied. Naturally, advances in instrumentation will also serve to expand the applications of VtC XES. Perhaps most exciting among these is the development of dispersive spectrometers both at traditional synchrotrons and also free electron lasers that can translate XES into the time domain.23−27 Some of these conceptual designs subtend large solid angles, allowing for greater signal collection without the need for increases in incident X-ray flux. Furthermore, using an XFEL source, these

Figure 10. Fit to the spectrum of MnO with peaks that cannot be rationalized by MO theory denoted by arrows.

is likely that these features arise from many-electron transitions, though this hypothesis has not been definitively established or explored. Room thus remainsboth experimentally as well as theoreticallyfor refinement and expansion of the interpretation of these spectra; efforts along these directions are currently ongoing. G


Article


(8) Mukoyama, T.; Taniguchi, K.; Adachi, H. Chemical effect on Kbeta: K-alpha x-ray intensity ratios. Phys. Rev. B: Condens. Matter Mater. Phys. 1986, 34, 3710−3716. (9) Bergmann, U.; Horne, C. R.; Collins, T. J.; Workman, J. M.; Cramer, S. P. Chemical dependence of interatomic X-ray transition energies and intensities − a study of Mn Kβ″ and Kβ2, 5 spectra. Chem. Phys. Lett. 1999, 302, 119−124. (10) Bergmann, U.; Glatzel, P.; Robblee, J. H.; Messinger, J.; Fernandez, C.; Cinco, R.; Visser, H.; McFarlane, K.; Bellacchio, E.; Pizarro, S.; Sauer, K.; Yachandra, V. K.; Klein, M. P.; Cox, B. L.; Nealson, K. H.; Cramer, S. P. High-resolution X-ray spectroscopy of rare events: a different look at local structure and chemistry. J. Synchrotron Radiat. 2001, 8, 199−203. (11) Bergmann, U.; Bendix, J.; Glatzel, P.; Gray, H. B.; Cramer, S. P. Anisotropic valence -> core x-ray fluorescence from a [Rh(en) (3)][Mn(N) (CN)(5)] H2O single crystal: Experimental results and density functional calculations. J. Chem. Phys. 2002, 116, 2011−2015. (12) Glatzel, P.; Bergmann, U. High resolution 1s core hole X-ray spectroscopy in 3d transition metal complexeselectronic and structural information. Coord. Chem. Rev. 2005, 249, 65−95. (13) Lee, N.; Petrenko, T.; Bergmann, U.; Neese, F.; DeBeer, S. Probing Valence Orbital Composition with Iron K beta X-ray Emission Spectroscopy. J. Am. Chem. Soc. 2010, 132, 9715−9727. (14) Beckwith, M. A.; Roemelt, M.; Collomb, M.-N. l.; DuBoc, C.; Weng, T.-C.; Bergmann, U.; Glatzel, P.; Neese, F.; DeBeer, S. Manganese Kβ X-ray Emission Spectroscopy As a Probe of Metal− Ligand Interactions. Inorg. Chem. 2011, 50, 8397−8409. (15) Smolentsev, G.; Soldatov, A. V.; Messinger, J.; Merz, K.; Weyhermuller, T.; Bergmann, U.; Pushkar, Y.; Yano, J.; Yachandra, V. K.; Glatzel, P. X-ray Emission Spectroscopy To Study Ligand Valence Orbitals in Mn Coordination Complexes. J. Am. Chem. Soc. 2009, 131, 13161−13167. (16) MacMillan, S. N.; Walroth, R. C.; Perry, D. M.; Morsing, T. J.; Lancaster, K. M. Ligand-Sensitive But Not Ligand-Diagnostic: Evaluating Cr Valence-to-Core X-ray Emission Spectroscopy as a Probe of Inner-Sphere Coordination. Inorg. Chem. 2015, 54, 205−214. (17) Stojanoff, V.; Hämäläinen, K.; Siddons, D. P.; Hastings, J. B.; Berman, L. E.; Cramer, S.; Smith, G. A high-resolution x-ray fluorescence spectrometer for near-edge absorption studies. Rev. Sci. Instrum. 1992, 63, 1125. (18) Schulke, W.; Kaprolat, A.; Fischer, T.; Hoppner, K.; Wohlert, F. Spectrometer for high resolution resonant inelastic x-ray scattering. Rev. Sci. Instrum. 1995, 66, 2446. (19) Bergmann, U.; Cramer, S. P. Proc. SPIE 1998, 198−209. (20) Sokaras, D.; Weng, T.-C.; Nordlund, D.; Alonso-Mori, R.; Velikov, P.; Wenger, D.; Garachtchenko, A.; George, M.; Borzenets, V.; Johnson, B.; Rabedeau, T.; Bergmann, U. A seven-crystal Johanntype hard x-ray spectrometer at the Stanford Synchrotron Radiation Lightsource. Rev. Sci. Instrum. 2013, 84, 053102. (21) Lundberg, M.; Kroll, T.; DeBeer, S.; Bergmann, U.; Wilson, S. A.; Glatzel, P.; Nordlund, D.; Hedman, B.; Hodgson, K. O.; Solomon, E. I. Metal−Ligand Covalency of Iron Complexes from HighResolution Resonant Inelastic X-ray Scattering. J. Am. Chem. Soc. 2013, 135, 17121−17134. (22) Ament, L. J. P.; van Veenendaal, M.; Devereaux, T. P.; Hill, J. P.; van den Brink, J. Resonant inelastic x-ray scattering studies of elementary excitations. Rev. Mod. Phys. 2011, 83, 705−767. (23) Szlachetko, J.; Nachtegaal, M.; de Boni, E.; Willimann, M.; Safonova, O.; Sa, J.; Smolentsev, G.; Szlachetko, M.; van Bokhoven, J. A.; Dousse, J.-C.; Hoszowska, J.; Kayser, Y.; Jagodzinski, P.; Bergamaschi, A.; Schmitt, B.; David, C.; Lucke, A. A von Hamos xray spectrometer based on a segmented-type diffraction crystal for single-shot x-ray emission spectroscopy and time-resolved resonant inelastic x-ray scattering studies. Rev. Sci. Instrum. 2012, 83, 103105. (24) Alonso-Mori, R.; Kern, J.; Gildea, R. J.; Sokaras, D.; Weng, T.C.; Lassalle-Kaiser, B.; Tran, R.; Hattne, J.; Laksmono, H.; Hellmich, J.; Glöckner, C.; Echols, N.; Sierra, R. G.; Schafer, D. W.; Sellberg, J.; Kenney, C.; Herbst, R.; Pines, J.; Hart, P.; Herrmann, S.; GrosseKunstleve, R. W.; Latimer, M. J.; Fry, A. R.; Messerschmidt, M. M.;

spectrometers enable the combination of XES experiments with scattering and diffraction data.50 While the ability to observe structural and electronic changes occurring on the time scales of most chemical reactions under in operando conditions remains a challenge, developments in this direction are certainly well underway.

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Notes

The authors declare no competing financial interest. Biographies Christopher Pollock received his B.S. in chemistry from Clemson University (2009) and his M.S. in chemistry from Cornell University (2011) before moving to Mülheim, Germany to complete his Ph.D. in inorganic chemistry at the Max-Planck-Institute for Chemical Energy Conversion (2014). He is currently an NIH Postdoctoral Fellow at Penn State University. Serena DeBeer is a Professor and Research Group Leader at the MaxPlanck-Institute for Chemical Energy Conversion in Mülheim an der Ruhr, Germany. She also holds an appointment as an Adjunct Associate Professor in the Department of Chemistry and Chemical Biology at Cornell University in Ithaca, NY. Serena received her B.S. in Chemistry at Southwestern University in 1995 and her Ph.D. from Stanford University in 2002. From 2002 to 2009, she was a staff scientist at the Stanford Synchrotron Radiation Laboratory before moving to her faculty position at Cornell. She is the recipient of a Sloan Fellowship (2011), a Kavli Fellowship (2012), a European Research Council Consolidator Award (2013), and the Society of Biological Inorganic Chemistry Early Career Award (2015). Her primary research interests include the development and application of X-ray based spectroscopy to examine small molecule activation in both chemical and biological catalysis.

■

ACKNOWLEDGMENTS Julian Rees is thanked for his helpful comments during preparation of the manuscript. Financial support was provided by the Max-Planck-Gesellschaft. C.J.P thanks the National Institutes of Health for a National Research Service Award (GM 113389-01).

■

REFERENCES

(1) Bearden, J. A.; Shaw, C. H. Shapes and Wavelengths of K Series Lines of Elements Ti 22 to Ge 32. Phys. Rev. 1935, 48, 18−30. (2) Beeman, W. W.; Friedman, H. The X-Ray K Absorption Edges of the Elements Fe (26) to Ge (32). Phys. Rev. 1939, 56, 392−405. (3) Ern, V.; Switendick, A. C. Electronic Band Structure of TiC, TiN, and TiO. Phys. Rev. 1965, 137, A1927−A1936. (4) Best, P. E. Electronic Structure of the MnO4−, CrO42, and VO43 Ions from the Metal K X-Ray Spectra. J. Chem. Phys. 1966, 44, 3248−3253. (5) Koster, A. S.; Mendel, H. X-ray kβ emission spectra and energy levels of compounds of 3D- transition metalsI: Oxides. J. Phys. Chem. Solids 1970, 31, 2511−2522. (6) Tsutsumi, K.; Nakamori, H.; Ichikawa, K. X-ray Mn K-beta emission of manganese oxides and manganates. Phys. Rev. B 1976, 13, 929−933. (7) Wood, P. R.; Urch, D. S. Valence-band X-ray emission spectra of manganese in various oxidation states. J. Chem. Soc., Dalton Trans. 1976, 2472−2476. H


Article

Accounts of Chemical Research Miahnahri, A.; Seibert, M. M.; Zwart, P. H.; White, W. E.; Adams, P. D.; Bogan, M. J.; Boutet, S.; Williams, G. J.; Zouni, A.; Messinger, J.; Glatzel, P.; Sauter, N. K.; Yachandra, V. K.; Yano, J.; Bergmann, U. Energy-dispersive X-ray emission spectroscopy using an X-ray freeelectron laser in a shot-by-shot mode. Proc. Natl. Acad. Sci. U. S. A. 2012, 109, 19103−19107. (25) Finkelstein, K. D.; Agyeman-Budu, D.; Lyndaker, A.; Pollock, C.; Krawczyk, T. X-ray Emission Collected in a Novel Energy Dispersive Approach. J. Phys. Conf Ser. 2014, 493, 012033. (26) Pacold, J. I.; Bradley, J. A.; Mattern, B. A.; Lipp, M. J.; Seidler, G. T.; Chow, P.; Xiao, Y.; Rod, E.; Rusthoven, B.; Quintana, J. A miniature X-ray emission spectrometer (miniXES) for high-pressure studies in a diamond anvil cell. J. Synchrotron Radiat. 2012, 19, 245− 251. (27) Anklamm, L.; Schlesiger, C.; Malzer, W.; Grötzsch, D.; Neitzel, M.; Kanngießer, B. A novel von Hamos spectrometer for efficient Xray emission spectroscopy in the laboratory. Rev. Sci. Instrum. 2014, 85, 053110. (28) Pushkar, Y.; Long, X.; Glatzel, P.; Brudvig, G. W.; Dismukes, G. C.; Collins, T. J.; Yachandra, V. K.; Yano, J.; Bergmann, U. Direct detection of oxygen ligation to the Mn4Ca Cluster of Photosystem II by XES. Angew. Chem., Int. Ed. 2010, 49, 800−803. (29) Badro, J.; Rueff, J.-P.; Vankó, G.; Monaco, G.; Fiquet, G.; Guyot, F. Electronic Transitions in Perovskite: Possible Nonconvecting Layers in the Lower Mantle. Science 2004, 305, 383−386. (30) Pollock, C. J.; DeBeer, S. Valence-to-Core X-ray Emission Spectroscopy: A Sensitive Probe of the Nature of a Bound Ligand. J. Am. Chem. Soc. 2011, 133, 5594−5601. (31) Delgado-Jaime, M. U.; Dible, B. R.; Chiang, K. P.; Brennessel, W. W.; Bergmann, U.; Holland, P. L.; DeBeer, S. Identification of a Single Light Atom within a Multinuclear Metal Cluster Using Valenceto-Core X-ray Emission Spectroscopy. Inorg. Chem. 2011, 50, 10709− 10717. (32) Lancaster, K. M.; Roemelt, M.; Ettenhuber, P.; Hu, Y.; Ribbe, M. W.; Neese, F.; Bergmann, U.; DeBeer, S. X-ray Emission Spectroscopy Evidences a Central Carbon in the Nitrogenase IronMolybdenum Cofactor. Science 2011, 334, 974−977. (33) Lancaster, K. M.; Finkelstein, K. D.; DeBeer, S. K-beta X-ray Emission Spectroscopy Offers Unique Chemical Bonding Insights: Revisiting the Electronic Structure of Ferrocene. Inorg. Chem. 2011, 50, 6767−6774. (34) Stieber, S. C. E.; Milsmann, C.; Hoyt, J. M.; Turner, Z. R.; Finkelstein, K. D.; Wieghardt, K.; DeBeer, S.; Chirik, P. J. Bis(imino)pyridine Iron Dinitrogen Compounds Revisited: Differences in Electronic Structure Between Four- and Five-Coordinate Derivatives. Inorg. Chem. 2012, 51, 3770−3785. (35) Pollock, C. J.; Grubel, K.; Holland, P. L.; DeBeer, S. Experimentally Quantifying Small-Molecule Bond Activation Using Valence-to-Core X-ray Emission Spectroscopy. J. Am. Chem. Soc. 2013, 135, 11803−11808. (36) Rees, J. A.; Martin-Diaconescu, V.; Kovacs, J. A.; DeBeer, S. Xray Absorption and Emission Study of Dioxygen Activation by a SmallMolecule Manganese Complex. Inorg. Chem. 2015, 54, 6410−6422. (37) Lu, T.-T.; Weng, T.-C.; Liaw, W.-F. X-Ray Emission Spectroscopy: A Spectroscopic Measure for the Determination of NO Oxidation States in Fe−NO Complexes. Angew. Chem. 2014, 126, 11746−11750. (38) Lassalle-Kaiser, B.; Boron, T. T.; Krewald, V.; Kern, J.; Beckwith, M. A.; Delgado-Jaime, M. U.; Schroeder, H.; Alonso-Mori, R.; Nordlund, D.; Weng, T.-C.; Sokaras, D.; Neese, F.; Bergmann, U.; Yachandra, V. K.; DeBeer, S.; Pecoraro, V. L.; Yano, J. Experimental and Computational X-ray Emission Spectroscopy as a Direct Probe of Protonation States in Oxo-Bridged MnIV Dimers Relevant to RedoxActive Metalloproteins. Inorg. Chem. 2013, 52, 12915−12922. (39) Chandrasekaran, P.; Chiang, K. P.; Nordlund, D.; Bergmann, U.; Holland, P. L.; DeBeer, S. Sensitivity of X-ray Core Spectroscopy to Changes in Metal Ligation: A Systematic Study of Low-Coordinate, High-Spin Ferrous Complexes. Inorg. Chem. 2013, 52, 6286−6298.

(40) Kropp, H.; King, A. E.; Khusniyarov, M. M.; Heinemann, F. W.; Lancaster, K. M.; DeBeer, S.; Bill, E.; Meyer, K. Manganese Nitride Complexes in Oxidation States III, IV, and V: Synthesis and Electronic Structure. J. Am. Chem. Soc. 2012, 134, 15538−15544. (41) Pollock, C. J.; Lancaster, K. M.; Finkelstein, K. D.; DeBeer, S. Study of Iron Dimers Reveals Angular Dependence of Valence-toCore X-ray Emission Spectra. Inorg. Chem. 2014, 53, 10378−10385. (42) Delgado-Jaime, M. U.; DeBeer, S.; Bauer, M. Valence-to-Core X-Ray Emission Spectroscopy of Iron−Carbonyl Complexes: Implications for the Examination of Catalytic Intermediates. Chem. Eur. J. 2013, 19, 15888−15897. (43) de Groot, F. High resolution X-ray emission and X-ray absorption spectroscopy. Chem. Rev. 2001, 101, 1779−1808. (44) Grush, M. M.; Christou, G.; Haemaelaeinen, K.; Cramer, S. P. Site-Selective XANES and EXAFS: A Demonstration with Manganese Mixtures and Mixed-Valence Complexes. J. Am. Chem. Soc. 1995, 117, 5895−5896. (45) Glatzel, P.; Jacquamet, L.; Bergmann, U.; de Groot, F. M. F.; Cramer, S. P. Site-selective EXAFS in mixed-valence compounds using high-resolution fluorescence detection: A study of iron in Prussian Blue. Inorg. Chem. 2002, 41, 3121−3127. (46) Meyer, D. A.; Zhang, X.; Bergmann, U.; Gaffney, K. J. Characterization of charge transfer excitations in hexacyanomanganate(III) with Mn K-edge resonant inelastic x-ray scattering. J. Chem. Phys. 2010, 132, 134502. (47) Leidel, N.; Chernev, P.; Havelius, K. G. V.; Ezzaher, S.; Ott, S.; Haumann, M. Site-Selective X-ray Spectroscopy on an Asymmetric Model Complex of the [FeFe] Hydrogenase Active Site. Inorg. Chem. 2012, 51, 4546−4559. (48) Gallo, E.; Bonino, F.; Swarbrick, J. C.; Petrenko, T.; Piovano, A.; Bordiga, S.; Gianolio, D.; Groppo, E.; Neese, F.; Lamberti, C.; Glatzel, P. Preference towards Five-Coordination in Ti Silicalite-1 upon Molecular Adsorption. ChemPhysChem 2013, 14, 79−83. (49) Hall, E. R.; Pollock, C. J.; Bendix, J.; Collins, T. J.; Glatzel, P.; DeBeer, S. Valence-to-Core-Detected X-ray Absorption Spectroscopy: Targeting Ligand Selectivity. J. Am. Chem. Soc. 2014, 136, 10076− 10084. (50) Kern, J.; Alonso-Mori, R.; Tran, R.; Hattne, J.; Gildea, R. J.; Echols, N.; Glöckner, C.; Hellmich, J.; Laksmono, H.; Sierra, R. G.; Lassalle-Kaiser, B.; Koroidov, S.; Lampe, A.; Han, G.; Gul, S.; DiFiore, D.; Milathianaki, D.; Fry, A. R.; Miahnahri, A.; Schafer, D. W.; Messerschmidt, M.; Seibert, M. M.; Koglin, J. E.; Sokaras, D.; Weng, T.-C.; Sellberg, J.; Latimer, M. J.; Grosse-Kunstleve, R. W.; Zwart, P. H.; White, W. E.; Glatzel, P.; Adams, P. D.; Bogan, M. J.; Williams, G. J.; Boutet, S.; Messinger, J.; Zouni, A.; Sauter, N. K.; Yachandra, V. K.; Bergmann, U.; Yano, J. Simultaneous Femtosecond X-ray Spectroscopy and Diffraction of Photosystem II at Room Temperature. Science 2013, 340, 491−495.

I


Insights into the Geometric and Electronic Structure of Transition Metal

Recommend Documents