Infrared Photodissociation Spectroscopy of Vanadium Oxide

Article pubs.acs.org/JPCA

Infrared Photodissociation Spectroscopy of Vanadium Oxide− Carbonyl Cations A. D. Brathwaite, A. M. Ricks, and M. A. Duncan* Department of Chemistry, University of Georgia, Athens, Georgia 30602, United States S Supporting Information *

ABSTRACT: Mass selected vanadium oxide−carbonyl cations of the form VOm(CO)n+ (m = 0−3 and n = 3−6) are studied via infrared laser photodissociation spectroscopy in the 600−2300 cm−1 region. Insight into the structure and bonding of these complexes is obtained from the number of infrared active bands, their relative intensities and their frequency positions. Density functional theory calculations are carried out in support of the experimental data. The effect of oxidation on the carbonyl stretching frequencies of VO(CO)n+, VO2(CO)n+, and VO3(CO)n+ complexes is investigated. All of these oxide−carbonyl species have C− O stretch vibrations blue-shifted from those of the pure vanadium ion carbonyls. The V−O stretches of these complexes are also investigated, revealing the effects of CO coordination on these vibrations. The oxide−carbonyls all have a hexacoordinate core analogous to that of V(CO)6+. The fully coordinated vanadium monoxide−carbonyl species is VO(CO)5+, and those of the dioxide and trioxide are VO2(CO)4+ and VO3(CO)3+, respectively.

■

INTRODUCTION Transition metal oxides have unique structural, electrical and magnetic properties, with widespread applications in catalysis, fuel cells, chemical sensors, and magnetic materials.1−9 Additionally, metal oxides are important throughout surface science, and their interactions with organic and inorganic molecules are the subject of extensive investigation.10,11 Small oxide complexes produced in the gas phase provide model systems with which to investigate fundamental bonding interactions and their dependence on electronic structure without the complications of solvation or other environmental perturbations. Furthermore, gas phase ions can be size-selected, allowing for the systematic investigation of complexes with well-defined composition. In this study, we employ these concepts with infrared photodissociation spectroscopy to probe both the oxide stretches and carbonyl stretches of vanadium oxide, dioxide and trioxide carbonyl cations. Carbon monoxide is the most widely studied molecular adsorbate in surface science,10,11 where it is a sensitive probe of bonding sites for both pure metal and oxide surfaces. The C−O stretching vibrations have characteristic frequencies depending on the type of binding (atop, bridging, 3-fold).10,11 The CO ligand is also well-known in inorganic and organometallic chemistry, where the C−O stretch is diagnostic of structure and bonding.12−15 The bonding in transition metal carbonyl systems can be understood using the Dewar−Chatt− Duncanson complexation model.12−15 The details of the electronic structure and bonding in these systems, with special attention to the C−O stretch vibration, have been investigated extensively by theory.16−26 In so-called “classical” metal carbonyls, π-back-bonding is the dominant interaction, which involves charge donation from occupied metal d orbitals into © 2013 American Chemical Society

the antibonding LUMO on CO. This results in C−O stretches that are lower in frequency (i.e., red-shifted) than that of the free CO molecule (fundamental = 2143 cm−1).27 Oxidation of the metal is likely to reduce its charge donating ability, as electron density is used to form metal−oxygen bonds. Consequently, reduced red shifts, or even blue shifts, may be observed for metal oxide−carbonyl complexes. Experimental vibrational measurements on neutral metal carbonyls are widespread, employing both infrared and Raman spectroscopy.12−15 However, studies on ionized carbonyls in which the role of charge can be investigated are more recent. Both saturated and unsaturated carbonyl ions have been produced in cryogenic rare gas matrices and examined using infrared absorption spectroscopy.20,28−34 In the gas phase, metal ion complexes have been investigated primarily with mass-selected infrared photodissociation spectroscopy. Using this method, Fielicke and co-workers investigated several metal cluster carbonyl systems, Mm(CO)n±, using an infrared free electron laser.35−39 More recent work has employed new infrared optical parametric oscillator (OPO) laser systems.41−52 In one particular study, our research group measured spectra for V(CO)n+ complexes, finding slightly red-shifted C−O stretches and a coordination of six ligands, even though a sevencoordinate complex was predicted by the 18-electron rule.43,47 Small vanadium oxides and their carbonyls have also been investigated. Infrared spectra for the vanadium + CO2 insertion products, OVCO and OVCO+, were measured by Andrews et Special Issue: Terry A. Miller Festschrift Received: July 11, 2013 Revised: August 7, 2013 Published: August 8, 2013 13435

dx.doi.org/10.1021/jp4068697 | J. Phys. Chem. A 2013, 117, 13435−13442

The Journal of Physical Chemistry A

Article

al. via matrix isolation.53 Asmis and co-workers used a free electron laser for infrared photodissociation spectra of the VO+, VO2+, and VO3+ ions tagged with rare gas atoms.54−56 In the present study on VO(CO)n+, VO2(CO)n+, and VO3(CO)n+ ions, we examine the effect of oxidation on the carbonyl stretches as well as that of the CO binding on the oxide stretches.

■

EXPERIMENTAL SECTION VOm(CO)n+ ions are produced in a pulsed nozzle laser vaporization source using the third harmonic of a Nd:YAG laser (355 nm; Spectra-Physics INDI). The laser is focused onto a rotating and translating 1/4 in. diameter vanadium rod mounted on the front of a pulsed nozzle (General Valve Series 9) in the so-called cutaway configuration.57 Ions are produced using a mixture of 5% O2 in carbon monoxide, at a backing pressure of 150 psig. The expansion is skimmed into a differentially pumped chamber where positive ions are extracted into a homemade reflectron time-of-flight mass spectrometer.58 Ions of a specific mass are selected by their flight time using pulsed deflection plates located at the end of the first flight tube. Selected ions are excited in the turning region of the reflectron with the tunable output of an infrared OPO laser system (LaserVision) pumped by a Nd:YAG laser (Spectra Physics Pro 230). This laser provides infrared light in the region 2000−4000 cm−1 with a line width of about 1 cm−1. With difference frequency generation in an added AgGaSe2 crystal, the tuning range 600−2000 cm−1 is accessed. Infrared spectra are recorded by monitoring the appearance of one or more fragment ions as a function of the laser wavelength. DFT calculations were carried out to determine the structure and bonding of these complexes using the B3LYP functional59,60 as implemented in the Gaussian 2003 computational package.61 The Def2-TZVP basis set62 was used for vanadium atoms, and the DZP basis set63 was used for carbon and oxygen. The computed carbonyl frequencies were scaled by a factor of 0.971, as discussed previously,47 and given an 8 cm−1 fwhm Lorentzian line shape for comparison to the experimental spectra. The computed metal oxide frequencies were scaled using the factor of 0.9167 derived by Asmis and co-workers for vanadium oxide cations.54−56

Figure 1. Infrared photodissociation breakdown spectra of VO(CO)n+ (n = 5−8) complexes. These spectra were obtained by subtracting a mass spectrum of a mass selected cluster with the laser off from one with it on.

tion laser “on” versus that with it “off”. The negative peak indicates depletion of the parent ion via dissociation, whereas the positive ones indicate the fragment ions produced. Ideally, the integrated area of the parent ion depletion peak should be equivalent to the sum of the areas of the fragment peaks (i.e., charge should be conserved). However, this is not always the case because the mass spectrometer cannot be focused on the parent and fragment ions simultaneously. As shown in the bottom trace, VO(CO)n+ ions smaller than n = 5 do not fragment efficiently. This is evident from the weak parent depletion and fragment production. We calculate the binding energy of CO in VO(CO)5+ to be approximately 13.5 kcal/mol (4720 cm−1). Because this is significantly more than the energy of an infrared photon in the CO stretching region (∼2100 cm−1), the small amount of fragmentation for this complex is attributed to multiphoton absorption. Much more efficient fragmentation is observed for complexes larger than n = 5. As shown in Figure 1, these undergo efficient CO elimination terminating at n = 5. This indicates that the first five carbonyl ligands are strongly bound and the VO(CO)5+ species is the fully coordinated ion. Similar fragmentation spectra were measured for VO2(CO)n+ and VO3(CO)n+ and are presented in the Supporting Information. All VOm(CO)n+ species were determined to have a total coordination number of m + n = 6, analogous to the fully coordinated V(CO)6+ ion observed previously.43,47 The oxygen atoms apparently each occupy a single binding site similar to those of the CO ligands. To examine the vibrational spectroscopy of these systems, we measure the wavelength dependence of these fragmentation processes. Figure 2 compares the photodissociation spectra in the region of the carbonyl stretch for V(CO)6+, VO(CO)5+, and VO3(CO)3+ ions, each measured by monitoring the loss of a CO ligand. According to the breakdown data above, these are fully coordinated ions without any excess ligands. Their ligands are therefore relatively strongly bound and CO elimination is only possible via a multiphoton process. This fragmentation is relatively inefficient, giving rise to the somewhat low signal

■

RESULTS AND DISCUSSION Laser vaporization of a vanadium rod in an expansion of pure CO produces complexes of the form V(CO)n+, as described previously.43,47 The VO(CO)n+ species were first observed as minor peaks in that mass spectrum, due to residual oxide on the surface of the metal rod. The abundance of the desired oxide− carbonyl ions was enhanced by seeding oxygen into the expansion; a mass spectrum obtained in this way is shown in the Supporting Information. VOm(CO)n+ complexes with up to 13 CO ligands are produced. The larger species consist of a strongly bound metal−ligand core ion with additional molecules attached to its exterior via electrostatic and/or van der Waals interactions, as discussed previously.26 Insight into the stability and coordination of the VOm(CO)n+ complexes can be obtained from their photofragmentation behavior. Figure 1 shows the infrared induced fragmentation of VO(CO)n+ ions (n = 5−8). These data are collected by adjusting the IR wavelength to the most intense carbonyl stretch resonance for each cation, and taking the difference between the mass spectrum obtained with the photodissocia13436



Article

Figure 2. Infrared photodissociation spectra of the V(CO)6+, VO(CO)5+, and VO3(CO)3+ ions in the carbonyl stretching region.

Figure 3. Infrared photodissociation spectra of the V(CO)7+, VO(CO)6+, and VO3(CO)4+ ions in the carbonyl stretching region.

levels in these spectra. The V(CO)6+ spectrum (top trace) has an intense, broad feature at 2103 cm−1, which is red-shifted from the free CO stretch at 2143 cm−1. As described before,43,47 this single feature arises from overlapping bands for the six-coordinate complex which is slightly distorted from octahedral symmetry to a D3d structure. The spectrum of VO(CO)5+ in the middle trace contains an intense band at nearly the position of the free-CO stretch, along with a minor peak at 2204 cm−1. The bottom trace shows the spectrum of VO3(CO)3+, with a single intense band at 2228 cm−1 that is blue-shifted from the free-CO frequency. It is evident from this comparison that progressive oxidation of the metal center leads to a reduced red shift and then a blue shift of the carbonyl stretch. Figure 3 shows spectra for the same complexes studied in Figure 2, but with each having an additional CO. Because of the fragmentation behavior discussed above, this extra CO is expected to be in the second coordination sphere with a much lower binding energy. These ions can therefore be indicated as “core+1” species. CO is therefore eliminated from these ions efficiently by infrared excitation, and these spectra show a noticeable improvement in the signal levels compared to the data in Figure 2. The V(CO)7+ spectrum in the top trace has a broad, intense red-shifted feature at 2103 cm−1, as well as a weaker band at 2166 cm−1. The latter blue-shifted band is consistent with the presence of a second sphere CO ligand; its position is highlighted with blue shading in the figure to identify this feature in each of these complexes.26,43,47 As discussed previously, electrostatic polarization of the external CO ligand by the core ion leads to an increase in its stretching frequency.26,43,47 The width of the main peak in the V(CO)7+ spectrum is likely due to multiple isomeric structures with external CO ligands in different positions around the core ion. These isomers are close in energy but may have slightly different spectra. The VO(CO)6+ spectrum in the middle trace, like that of VO(CO)5+, has an intense band at 2143 cm−1. Additionally, two weaker blue-shifted features are observed. The peak at 2169 cm−1 is near the position of the vibrations of

second-sphere CO ligands, whereas the assignment of the band at 2213 cm−1 is not immediately obvious. The bottom trace of this figure shows the spectrum of VO3(CO)4+, which has two blue-shifted peaks at 2232 and 2175 cm−1, and a weaker feature at 2205 cm−1. The two main bands are representative of core and second-sphere carbonyl ligands, respectively. These spectra are all consistent with structures having the same core ions seen above in Figure 2, but each has a weakly bound external CO ligand. Again, therefore, progressive oxidation leads to reduced red shifts and eventually to blue-shifted carbonyl bands. There is no evidence in these larger complexes for additional multiplet structure that might be expected if these complexes have core ions with higher coordination numbers.43,47 Metal−oxygen stretches for these complexes are expected at frequencies near 1000 cm−1. In this range, both the photon energy and the laser power are quite low and dissociation cannot be measured unless complexes have weakly bound CO molecules that can be eliminated with a single photon. We therefore focus these studies on complexes having the “core +1” CO stoichiometry, like those in Figure 3. Infrared spectra in the 600−1200 cm−1 range (black traces) are presented in Figure 4 for the VO(CO)6+, VO2(CO)5+, and VO3(CO)4+, all measured via the elimination of a single CO. Again, the low binding energies of the external molecules produce spectra with good signal levels. In previous work, Asmis and co-workers investigated the corresponding bare oxide ions via helium tagging and infrared excitation with a free electron laser.54−56 We have simulated spectra with their line width and band positions in red for comparison to our measurements for oxide carbonyls. Our spectrum for VO(CO)6+ has a single band at 1000 cm−1 with a fwhm of approximately 8 cm−1, attributed to the single V−O stretch. The bare metal oxide cation studied by Asmis with helium tagging also had a single peak in this region, but its position was at 1053 cm−1.54−56 The photoelectron spectrum of VO produced a band at 1060 ± 40 cm−1 attributed to the VO+ stretch.64 Both of these previous studies agreed on the frequency for VO+ and its assignment to the 3Σ− ground state. The differences between our carbonyl species and the 13437



Article

Figure 5. Infrared spectrum of VO(CO)6+ compared to the spectra predicted by DFT theory.

Figure 4. Low-frequency infrared photodissociation spectra of VO(CO)6+, VO2(CO)5+, and VO3(CO)4+ compared to simulated spectra (red) for VO+, VO2+, and VO3+ ions using frequencies reported by Asmis and co-workers.54−56

agreement or disagreement with the experiment depends critically on the scaling factor chosen for the theory. Because the singlet is predicted to lie lowest in energy and because its C−O stretch bands match the experiment, we conclude that the observed spectrum of the VO(CO)6+ species is that of the singlet ground state. We can also then conclude that the scaling factor employed here, which was derived from the work of Asmis,54−56 is reliable for these kinds of V−O stretch vibrations. DFT calculations also predict a singlet ground state for VO2(CO)5+, with a triplet state lying 33.9 kcal/mol higher in energy. The predicted spectrum shown in Figure 6 has three

bare oxide can be attributed to the effects of the carbonyl binding and differences in the oxide electronic structure, as discussed below. VO2(CO)5+ has two bands in the low frequency region at 944 and 971 cm−1, likely due to asymmetric and symmetric O−V−O stretches. Again, these can be compared to a similar doublet for the bare oxide. However, our data are again red-shifted, with each band about 46 cm−1 lower than the corresponding peaks for the bare oxide. The VO3(CO)4+ complex yields a spectrum with two main peaks at 991 and 1084 cm−1, and a minor one at 1045 cm−1, whereas the bare oxide complex had only two bands at 1037 and 1069 cm−1. Our main doublet is more widely spaced than that of the isolated oxide ion, with an extra weak band. It is possible that the weak feature observed in our data were not resolved in the Asmis experiment because of the greater line width when the free electron laser was used. To aid in the interpretation of these spectra, we employ DFT calculations on these vanadium oxide cations and also on the corresponding bare oxide cations. Complete details on the calculated structures, energetics, and spin states are provided in the Supporting Information for this paper. We investigated various isomers and spin states for each system. Predicted spectra for the two lowest energy spin states along with their computed structures are presented together with the corresponding experimental spectrum for each complex. Our computations find that the singlet spin state for VO(CO)6+ is lower in energy than the triplet by 29.7 kcal/ mol. Although their geometries are quite similar, the singlet and triplet spin states have very different spectra. Figure 5 shows the experimental spectrum of VO(CO)6+ along with those predicted for the two low-lying spin states. The experiment finds a single intense band in the carbonyl stretch region along with two much weaker features. The spectrum predicted for the singlet state matches the experiment almost perfectly, whereas the pattern of three peaks in the spectrum for the triplet state is quite different. In the low frequency region, both experiment and theory yield a single intense band. Again, the spectrum for the singlet state matches the experiment, but the triplet also has a single band at lower frequency. Because both predicted spectra have only a single band here, it is clear that the

Figure 6. Infrared spectrum of VO2(CO)5+ compared to the spectra predicted by DFT theory.

bands in the carbonyl stretching region and two in the low frequency region, agreeing beautifully with the experiment. This includes both the band positions and relative intensities for the triplet pattern in the C−O stretching region and the doublet in the V−O stretching region. The spectrum predicted for the triplet state has a distinctively different pattern not consistent with the experiment in either region. We therefore assign the experimental spectrum to be that of the singlet ground state. Theory allows assignment of the more intense feature (944 13438



Article

cm−1) at low frequency to the asymmetric O−V−O stretch, whereas the higher frequency band with lower intensity (971 cm−1) is the symmetric stretch. DFT computations for the VO3(CO)4+ ion produce lowest energy singlet and triplet spin states that differ in energy only by 5.3 kcal/mol, with the singlet again predicted to lie at lower energy. The structure is not a symmetric trioxide but instead has individual VO (vanadyl) and VO2 (superoxide) moieties. This is consistent with the findings of previous computational work on the VO3+ ion without ligands.65 Figure 7 shows the experimental spectrum for this ion compared to

in both the oxide and carbonyl stretching regions, together with the close energies for the singlet and triplet structures, lead us to conclude that we likely also have a small amount of the singlet species present. The positions and intensities of the CO bands for the vanadium oxide ions in this experiment are noticeably different from those of the pure vanadium carbonyls. Whereas the pure metal complexes have slightly red-shifted bands, the oxides have either no shift or blue shifts. This indicates that oxidation of the metal atom does indeed have an effect on the carbonyl bonding. In pure metal carbonyls, π-back-bonding is the most important interaction and red-shifted CO frequencies are common for these systems. On the other hand, in the oxide carbonyls, the d electrons on the metal atom are used to form strong oxide bonds, and their availability for back-donation is reduced.19 This trend is shown nicely in the data of Figures 2 and 3, in which progressive oxidation leads to a reduced red shift and then a blue shift in the CO stretches. Because π-backbonding is reduced in these systems, the other well-known effects in carbonyl bonding of σ donation and electrostatic polarization become more important.18−26 As discussed extensively in the literature, and demonstrated in recent experiments on transition metals with filled d shells (gold, copper),40,46 these effects lead to blue-shifted carbonyl stretches, consistent with results here for the oxides. A similar effect occurs for the uranium oxide carbonyl ions, UO2(CO)n+, compared to the pure metal carbonyls, U(CO)n+.45 Similar blue-shifted frequencies are also well-known for CO adsorbates on metal oxide surfaces.10,11,69 It is therefore not surprising that progressive oxidation leads to the blue shifts seen here for the dioxide and trioxide species. Our data also show that the addition of carbonyl ligands induces vibrational band shifts on the corresponding bare vanadium oxide ions. For the monoxide and dioxide, our carbonyl−oxide vibrations are red-shifted by 53 and 46 cm−1, respectively, compared to those of the corresponding oxides without ligands, whereas the trioxide vibrations take on a more complex pattern. To explain this, we consider the orbital occupation in the oxides. It is well-known that the bonding in even simple metal oxides is complicated because the 3d−3d exchange energy is greater than the orbital separations.70 Because of this, it can be more favorable energetically to add electron density to a π* orbital than to a nonbonding orbital. If CO ligands donate electron density into such an orbital, the order of the V−O bond is reduced, its length should increase, and its stretching frequency should be lower, as seen here for the mono- and dioxides. A similar red shift of the metal− oxygen vibrations upon carbonyl binding was seen previously for the UO2(CO)n+ system.71 However, another important consideration is the spin state on the metal. The diatomic VO+ species has a triplet ground state (3Σ−), as noted above, whereas our VO(CO)6+ complex has a singlet ground state. As shown in our computations for VO+ (see the Supporting Information) and those for VO(CO6+ (Figure 5), the triplet state has a lower frequency than the corresponding singlet. The red shift from the carbonyl binding in the singlet state is therefore even more significant. VO2+ and VO3+ were concluded by Asmis to have a singlet and triplet ground states, respectively, like those seen here for VO2(CO)5+ and VO3(CO)4+, and therefore these ions are compared for the same ground spin states. The trioxide species here with its separate vanadyl and superoxide vibrations illustrates the complexity that may be expected in future studies of other

Figure 7. Infrared spectrum of VO3(CO)4+ compared to the spectra predicted by DFT theory.

those predicted for these singlet and triplet states. The experiment has two intense bands in the C−O stretching region, with the more intense of these having a noticeable shoulder on the low frequency side of the peak. A similar pattern is repeated in the metal oxide stretch region. Unfortunately, neither of the theoretical spectra agrees well with the experiment. In the CO stretching region, the relative intensity and splitting of the two main bands are consistent with the structure predicted for the triplet, but the singlet is predicted to be lower energy structure. In the VO stretch region, both spin states have individual vanadyl and superoxide vibrations, but neither of the patterns predicted for either spin state match the experiment. Asmis and co-workers found a similar problem for the isolated VO3+ species.54−56 Apparently, DFT at this level does not predict the oxide stretch modes of VO3+ reliably, perhaps because of mixed character between the vanadyl and superoxide units. The superoxide unit has significant negative charge on it, and DFT is well-known to have trouble with charge transfer systems.66 To make an assignment of this spectrum, we also note that DFT is wellknown to have trouble with the relative energies of transition metal spin states.67,68 We therefore do not assign high significance to the predicted energies of these states. Because the spectrum predicted for the triplet state matches the experimental pattern in the CO stretch almost exactly, and it also more closely matches the spacing and relative intensities of the oxide stretches, we assign our experimental spectrum to be that of the triplet state. This accounts for the stronger doublets in both regions of the spectrum. The minor features observed 13439



Article

MmOn+ species. The complex electronic structure of oxygen and the charge transfer likely to be found in these systems represent significant challenges for density functional theory. This is an important area because of the emergence of many new metal oxide nanocluster materials. For example, small neutral vanadium oxide clusters coated with ligands have recently been isolated by our group, and these species have an intense band near 1000 cm−1 like that seen here for the vanadyl stretches.72

(3) Campbell, C. T. Ultrathin Metal Films and Particles on Oxide Surfaces: Structural, Electronic and Chemisorptive Properties. Surf. Sci. Rep. 1997, 27, 1−111. (4) Kodama, R. H. Magnetic Nanoparticles. J. Magn. Magn. Matter 1999, 200, 359−372. (5) Hill, C. L. Progress and Challenges in Polyoxometalate-Based Catalysis and Catalytic Materials Chemistry. J. Mol. Catal. A: Chem. 2007, 262, 2−6. (6) Luo, X. L.; Morrin, A.; Killard, A. J.; Smyth, M. R. Application of Nanoparticles in Electrochemical Sensors and Biosensors. Eletroanalysis 2006, 18, 319−326. (7) Franke, M. E.; Koplin, T. J.; Simon, U. Metal and Metal Oxide Nanoparticles in Chemiresistors: Does the Nanoscale Matter? Small 2006, 2, 36−50. (8) Dorcioman, G.; Ebrasu, D.; Enculescu, I.; Serban, N.; Axente, E.; Sima, F.; Ristoscu, C.; Mihailescu, I. N. Metal Oxide Nanoparticles Synthesized by Pulsed Laser Ablation for Proton Exchange Membrane Fuel Cells. J. Power Sources 2010, 195, 7776−7780. (9) Joshi, U.; Palasyuk, A.; Arney, D.; Maggard, P. A. Semiconducting Oxides to Facilitate the Conversion of Solar Energy to Chemical Fuels. J. Phys. Chem. Lett. 2010, 1, 2719−2726. (10) Yates, J. T., Jr.; Madey, T. E. Vibrational Spectroscopy of Molecules at Surfaces; Plenum: New York, 1987. (11) Somorjai, G. A. Introduction to Surface Chemistry and Catalysis; John Wiley and Sons, Inc.: New York, 1994. (12) Cotton, F. A. Advanced Inorganic Chemistry, 6th ed.; John Wiley and Sons, Inc.: New York, 1999. (13) Huheey, J. E.; Keiter, E. A.; Keiter, R. L. Inorganic Chemistry Principles of Structure and Reactivity; Harper Collins: New York, 1993. (14) Hartwig, J. F. Organotransition Metal Chemistry; University Science Books: Sausalito, CA, 2010. (15) Nakamoto, K. Infrared and Raman Spectra of Inorganic and Coordination Compounds; John Wiley: New York, 1997. (16) Bauschlicher, C. W., Jr. Transition Metal-Ligand Bonding, II. J. Chem. Phys. 1986, 84, 260−267. (17) Bauschlicher, C. W., Jr.; Bagus, P. S.; Nelin, C. J.; Roos, B. J. The Nature of Bonding in XCO for X=Fe, Ni and Cu. J. Chem. Phys. 1986, 85, 354−364. (18) Barnes, L. A.; Rosi, M.; Bauschlicher, C. W., Jr. Theoretical Studies of the First- and Second-Row Mono- and Di-Carbonyl Positive Ions. J. Chem. Phys. 1990, 93, 609−624. (19) Sodupe, M.; Branchadell, V.; Rosi, M.; Bauschlicher, C. W., Jr. Theoretical Study of M+-CO2 and OM+CO Systems for First Transition Row Metal Atoms. J. Phys. Chem. A 1997, 101, 7854−7859. (20) Zhou, M.; Andrews, L.; Bauschlicher, C. W., Jr. Spectroscopic and Theoretical Investigations of Vibrational Frequencies in Binary Unsaturated Transition-Metal Carbonyl Cations, Neutrals, and Anions. Chem. Rev. 2001, 101, 1931−1962. (21) Goldman, A. S.; Krogh-Jespersen, K. Why Do Cationic Carbon Monoxide Complexes Have High CO Stretching Force Constants and Short CO Bonds? Electrostatic Effects, Not σ Bonding. J. Am. Chem. Soc. 1996, 118, 12159−12166. (22) Lupinetti, A. J.; Fau, S.; Frenking, G.; Strauss, S. H. Theoretical Analysis of the Bonding Between CO and Positively Charged Atoms. J. Phys. Chem. A 1997, 101, 9551−9559. (23) Lupinetti, A. J.; Frenking, G.; Strauss, S. H. Nonclassical Metal Carbonyls. Angew. Chem., Int. Ed. 1998, 37, 2113−2116. (24) Frenking, G.; Fröhlich, N. The Nature of the Bonding in Transition-Metal Compounds. Chem. Rev. 2000, 100, 717−774. (25) Lupinetti, A. J.; Strauss, S. H.; Frenking, G. Non-Classical Metal Carbonyls. Prog. Inorg. Chem. 2001, 49, 1−112. (26) Ricks, A. M.; Reed, Z. E.; Duncan, M. A. Infrared Spectroscopy of Metal Carbonyl Cations. J. Mol. Spectrosc. 2011, 266, 63−74. (27) Huber, K. P.; Herzberg, G. Molecular Spectra and Molecular Structure IV. Constants of Diatomic Molecules; Van Nostrand Reinhold Co.: New York, 1979. (28) Zhou, M.; Andrews, L. Infrared spectra of RhCO+, RhCO, and RhCO− in solid neon: A Scale for Charge Support in Catalyst Systems. J. Am. Chem. Soc. 1999, 121, 9171−9175.

■

CONCLUSION VOm(CO)n+ (m = 0−3 and n = 3−6) complexes, representing “core” and “core+1” ions, are produced in a molecular beam via laser vaporization in a pulsed nozzle source. The ions were mass-selected and studied using infrared photodissociation spectroscopy in both the carbonyl and oxide stretching regions. DFT calculations were conducted to facilitate the interpretation of the experimental spectra. Insight into the structure and bonding of these complexes is obtained from the number of infrared active bands, their relative intensities, and their frequency positions. The effect of oxidation on the carbonyl stretching frequencies of VO(CO)n +, VO 2(CO) n+, and VO3(CO)n+ is investigated. All of these ions have C−O vibrations that are blue-shifted from those of pure vanadium carbonyl cations due to a reduction in the availability of d electron density for back-bonding. The metal−oxygen stretching modes of these complexes are also investigated, and the effects of CO coordination on the V−O and O−V−O stretches are revealed. CO coordination causes these modes to red shift due to the introduction of electron density into antibonding modes. The metal oxide−carbonyl complexes investigated all have a hexacoordinate core, analogous to that of pure vanadium carbonyl. The fully coordinated vanadium monoxide−carbonyl species is of the form VO(CO)5+, and those of the dioxide and trioxide are VO2(CO)4+ and VO3(CO)3+.

■

ASSOCIATED CONTENT

* Supporting Information S

The full citation for ref 61 and the full details of the DFT computations done in support of the spectroscopy presented here, including mass spectra, the structures, energetics, coordinates, and vibrational frequencies for each of the complexes considered. This material is available free of charge via the Internet at http://pubs.acs.org.

■

AUTHOR INFORMATION

Corresponding Author

*M. A. Duncan: e-mail, [email protected]. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS We acknowledge generous support for this work from the U.S. Department of Energy (grant no. DE-FG02-96ER14658) and the Air Force Office of Scientific Research (grant no. FA955012-1-0166).

■

REFERENCES

(1) Cox, P. A. Transition Metal Oxides; Clarendon: Oxford, U.K., 1992. (2) Rao, C. N.; Raveau, B. Transition Metal Oxides; Wiley: New York, 1998. 13440



Article

7) Revealed with IR Spectroscopy. J. Phys. Chem. A 2013, 117, 1001− 1010. (48) Brathwaite, A. D.; M., A. Infrared Photodissociation Spectroscopy of Saturated Group IV (Ti, Zr, Hf) Metal Carbonyl Cations. J. Phys. Chem. A 2013, in press. (49) Wang, G.; Chi, C.; Cui, J.; Xing, X.; Zhou, M. Infrared Photodissociation Spectroscopy of Mononuclear Iron Carbonyl Anions. J. Chem. Phys. A 2012, 116, 2484−2489. (50) Chi, C.; Cui, J.; Xing, X.; Wang, G.; Liu, Z.; Zhou, M. Infrared Photodissociation Spectroscopy of Trigonal Bipyramidal 19-electron Ni(CO)5+ Cation. Chem. Phys. Lett. 2012, 542, 33−36. (51) Chi, C.; Cui, J.; Hua, Li, Z.; Xing, X.; Wang, G.; Zhou, M. Infrared Photodissociation Spectra of Mass Selected Homoleptic Dinuclear Iron Carbonyl Cluster Anions in the Gas Phase. Chem. Sci. 2012, 3, 1698−1706. (52) Cui, J.; Xing, X.; Chi, C.; Wang, G.; Liu, Z.; Zhou, M. Infrared Photodissociation Spectra of Mass-Selected Homoleptic Dinuclear Palladium Carbonyl Cluster Cations in the Gas Phase. Chin. J. Chem. 2012, 30, 2131−2137. (53) Zhou, M.; Andrews, L. Infrared Spectra and Density Functional Calculations for OMCO, OM(η2-CO), OMCO+, and OMOC+ (M = V, Ti) in Solid Argon. J. Phys. Chem. A 1999, 103, 2066−2075. (54) Brummer, M.; Kaposta, C.; Santambrogio, G.; Asmis, K. R. Formation and Photodepletion of Cluster Ion-Messenger Atom Complexes in a Cold Ion Trap: Infrared Spectroscopy of VO+, VO2+, and VO3+. J. Chem. Phys. 2003, 119, 12700−12703. (55) Asmis, K. R.; Meijer, G.; Brummer, M.; Kaposta, C.; Santambrogio, G.; Wöste, L.; Sauer, J. Gas Phase Infrared Spectroscopy of Mono- and Divanadium Oxide Cluster Cations. J. Chem. Phys. 2004, 120, 6461−6470. (56) Asmis, K. R.; Sauer, J. Mass-Selective Vibrational Spectroscopy of Vanadium Oxide Cluster Ions. Mass Spectrom. Rev. 2007, 26, 542− 562. (57) Duncan, M. A. Laser Vaporization Cluster Sources. Rev. Sci. Instrum. 2012, 83, 041101/1−19. (58) Duncan, M. A. Reflectron Time-Of-Flight Mass Spectrometer for Laser Photodissociation. Rev. Sci. Instrum. 1992, 63, 2177−2186. (59) Becke, A. D. 3 Term Correlation Functional. J. Chem. Phys. 1993, 98, 5648−5652. (60) Lee, C.; Yang, W.; Parr, R. G. Correlation Functional. Phys. Rev. B. 1998, 37, 785−789. (61) Frisch, M. J.; et al. Gaussian 03, Revision B.02; Gaussian, Inc.: Pittsburgh, PA, 2003. (62) Weigend, F.; Ahlrichs, R. Balanced Basis Sets of Split Valence, Triple Zeta Valence and Quadruple Zeta Valence Quality for H to Rn: Design and Assessment of Accuracy. Phys. Chem. Chem. Phys. 2005, 7, 3297−3305. (63) Dunning, T. H. Gaussian Basis Functions for Use in Molecular Calculations. I. Contraction of (9s5p) Atomic Basis Sets for the FirstRow Atoms. J. Chem. Phys. 1970, 53, 2823−2833. (64) Dyke, J. M.; Gravenor, B. W. J.; Hastings, M. P.; Morris, A. High Temperature Photoelectron Spectroscopy: The Vanadium Monoxide Molecule. J. Phys. Chem. 1985, 89, 4613−4617. (65) Calatayud, M.; Silvi, B.; Andrés, J.; Beltrán, A. A Theoretical Study on the Structure, Energetics and Bonding of VOx+ and VOx (x=1−4) Systems. Chem. Phys. Lett. 2001, 333, 493−503. (66) Cohen, A. J.; Mori-Sánchez, P.; Yang, W. Challenges for Density Functional Theory. Chem. Rev. 2012, 112, 289−320. (67) Harvey, J. N. DFT Computation of Relative Spin-State Energetics of Transition Metal Compounds. Struct. Bonding (Berlin) 2004, 112, 151−183. (68) Cramer, C. J.; Truhlar, D. G. Density Functional Theory for Transition Metals and Transition Metal Chemistry. Phys. Chem. Chem. Phys. 2009, 11, 10757−10816. (69) Eischens, R. P. Infrared Spectroscopy and Catalysis Research. Science 1964, 146, 486−493. (70) Bauschlicher, C. W., Jr.; Maître, P. Theoretical Study of the First Transition Row Metal Oxides and Sulfides. Theor. Chim. Acta 1995, 90, 189−203.

(29) Zhou, M.; Andrews, L. Matrix Infrared Spectra and Density Functional Calculations of ScCO, ScCO−, and ScCO+. J. Phys. Chem. A 1999, 103, 2964−2971. (30) Zhou, M.; Andrews, L. Infrared Spectra and Density Functional Calculations of Small Vanadium and Titanium Carbonyl Molecules and Anions in Solid Neon. J. Phys. Chem. A 1999, 103, 5259−5268. (31) Liang, B.; Zhou, M.; Andrews, L. Reactions of Laser-Ablated Ni, Pd, And Pt Atoms with Carbon Monoxide: Matrix Infrared Spectra and Density Functional Calculations on M(CO)n (n=1−4), M(CO)− (n=1−3), and M(CO)+ (n=1−2), (M=Ni, Pd, Pt). J. Phys. Chem. A 2000, 104, 3905−3914. (32) Liang, B.; Andrews, L. Reactions of Laser-Ablated Ag and Au atoms with Carbon Monoxide: Matrix Infrared Spectra and Density Functional Calculations on Au(CO)n (n=2,3), Au(CO)− (n=1,2), and M(CO)+ (n=1−4), (M=Ag, Au). J. Phys. Chem. A 2000, 104, 9156− 9164. (33) Zhou, M.; Andrews, L. Reactions of Zirconium and Hafnium Atoms with CO: Infrared Spectra and Density Functional Calculations of M(CO)x, OMCCO, and M(CO)2− (M = Zr, Hf; x = 1−4). J. Am. Chem. Soc. 2000, 122, 1531−1539. (34) Gutsev, G. L.; Andrews, L.; Bauschlicher, C. W., Jr. 3d-Metal Monocarbonyls MCO, MCO + and MCO − (M=Sc to Cu): Comparative Bond Strengths and Catalytic Ability to Produce CO2 in Reactions with CO. Chem. Phys. 2003, 290, 47−58. (35) Fielicke, A.; von Helden, G.; Meijer, G.; Pedersen, D. B.; Simard, B.; Rayner, D. M. Size and Charge Effects on Binding of CO to Small Isolated Rhodium Clusters. J. Phys. Chem. B. 2004, 108, 14591−14598. (36) Moore, D. T.; Oomens, J.; Eyler, J. R.; Meijer, G.; von Helden, G.; Ridge, D. P. Gas-Phase IR Spectroscopy of Anionic Iron Carbonyl Clusters. J. Am. Chem. Soc. 2004, 126, 14726−14727. (37) Fielicke, A.; von Helden, G.; Meijer, G.; Simard, B.; Rayner, D. M. Gold Cluster Carbonyls: Vibrational Spectroscopy of the Anions and the Effects of Cluster Size, Charge, and Coverage on the CO Stretching Frequency. J. Phys. Chem. B. 2005, 109, 23935−23940. (38) Fielicke, A.; von Helden, G.; Meijer, G.; Pedersen, D. B.; Simard, B.; Rayner, D. M. Gold Cluster Carbonyls: Saturated Adsorption of CO on Gold Cluster Cations, Vibrational Spectroscopy, and Implications for Their Structures. J. Am. Chem. Soc. 2005, 127, 8416−8423. (39) Fielicke, A.; von Helden, G.; Meijer, G.; Pedersen, D. B.; Simard, B.; Rayner, D. M. Size and Charge Effects on the Binding of CO to Late Transition Metal Clusters. J. Chem. Phys. 2006, 124, 194305−194312. (40) Velasquez, J., III; Njegic, B.; Gordon, M. S.; Duncan, M. A. IR Photodissociation Spectroscopy and Theory of Au+(CO)n Complexes: Nonclassical Carbonyls in the Gas Phase. J. Phys. Chem. A 2008, 112, 1907−1913. (41) Velasquez, J., III; Duncan, M. A. IR Photodissociation Spectroscopy of Gas Phase Pt+(CO)n (n=4−6). Chem. Phys. Lett. 2008, 461, 28−32. (42) Ricks, A. M.; Bakker, J. M.; Douberly, G. E.; Duncan, M. A. Infrared Spectroscopy and Structures of Cobalt Carbonyl Cations. J. Phys. Chem. A 2009, 113, 4701−4708. (43) Ricks, A. M.; Reed, Z. D.; Duncan, M. A. Seven-Coordinate Homoleptic Metal Carbonyls in the Gas Phase. J. Am. Chem. Soc. 2009, 131, 9176−9177. (44) Reed, Z. D.; Duncan, M. A. Infrared Spectroscopy and Structures of Manganese Carbonyl Cations, Mn(CO)n+ (n=1−9). J. Am. Soc. Mass Spectrom. 2010, 21, 739−749. (45) Ricks, A. M.; Gagliardi, L.; Duncan, M. A. Infrared Spectroscopy of Extreme Coordination: The Carbonyls of U+ and UO2+. J. Am. Chem. Soc. 2010, 132, 15905−15907. (46) Brathwaite, A. D.; Reed, Z. D.; Duncan, M. A. Infrared Photodissociation Spectroscopy of Copper Carbonyl Cations. J. Phys. Chem. A 2011, 115, 10461−10469. (47) Ricks, A. M.; Brathwaite, A. D.; Duncan, M. A. Coordination and Spin States in Vanadium Carbonyl Complexes (V(CO)n+, n = 1− 13441



Article

(71) Ricks, A. M.; Gagliardi, L.; Duncan, M. A. Uranium Oxo and Superoxo Cations Revealed Using Infrared Spectroscopy in the Gas Phase. J. Phys. Chem. Lett. 2011, 2, 1662−1666. (72) Ard, S.; Dibble, C. J.; Akin, S. T.; Duncan, M. A. Ligand-Coated Vanadium Oxide Clusters: Capturing Gas-Phase Magic Numbers in Solution. J. Phys. Chem. C 2011, 115, 6438−6447.

13442


Infrared Photodissociation Spectroscopy of Vanadium Oxide

Recommend Documents