Infrared Photodissociation Spectroscopy of TM+(N2)n (TM=V,Nb) Clusters E. D. Pillai, T. D. Jaeger, M. A. Duncan Department of Chemistry, University of Georgia Athens, GA 30602-2556 www.arches.uga.edu/~maduncan/ U.S. Department of Energy Why Study TM-Nitrogen? Biological systems require N 2 as components of proteins, nucleic acids, etc. But N 2 is highly inert (IP = 15.08 eV, BE = 225 kcal/mol). Nitrogenases catalyze N2 reduction and carry metal centers such as Fe, Mo, V. Large scale ammonia synthesis uses Fe as catalyst. N2 is isoelectronic to CO, C2H2 which are prevalent throughout inorganic and organometallic chemistry N2 activation gauged by change in N-N bond distance or N-N vibrational frequency N2 + H2 Fe catalyst 350 - 1000 atm 300 - 500 oC 2NH3 Previous Work Electronic spectroscopy of M+(N2) (M = Mg, Ca) by

Duncan and coworkers. CID studies by Armentrout and coworkers for Fe and Ni with N2 FT-ICR studies by H.Schwarz and coworkers, and electronic spectroscopy by Brucat and coworkers on Co+ (N2) Theoretical studies on TM-N2 carried out by Bauschlicher ESR spectra for V(N2)6 and Nb(N2)6 done by Weltner. IR studies using matrix isolation on M(N2) (M = V, Cr, Mn, Nb, Ta, Re) done by Andrews and coworkers Experimental Bond Energies* Direct absorption in our experiments is not possible due low ion densities. Solution is photodissociation. Ni+(N2)n IR photon 2359 cm-1 ~ 7 kcal/mol Small clusters may fragment via multiphoton process. Large clusters will be easier to fragment Fe+(N2)n n= 1 2 3 4 5

Bond Energy (kcal/mol) 13 19 10 13 15 n= 1 2 3 4 V+(CO)n n= 1 2 3 4 5 * Armentrout and coworkers 6 Bond Energy (kcal/mol) 27 27 14 2

Bond Energy (kcal/mol) 27 22 17 21 22 24 LaserVision OPO/OPA 2000-4500 cm-1 Production of cold metal ion complexes with laser vaporization/ supersonic expansion. Mass selection of cations by time-of-flight. Tunable infrared laser photodissociation spectroscopy. Nb+(N2)n Nb+ 6 4 2 10

5 n= 1 16 200 Mass 400 600 6 Fragmentation of Nb+ (N2)n 7 6 9 6

5 Fragmentation ends at n = 6 suggesting that this cluster is more stable. 8 7 n=6 100 200 300 mass 400 500 Infrared Photodissociation Spectra for Nb+(N2)n Free N2 mode 2359 cm-1 n=4

2265 Fragmentation is inefficient for the n = 1-3 clusters. n=3 The n=4 cluster shows fragmentation 95 cm-1 red of the free N2 stretch n=2 2100 2200 2300 -1 cm 2400 Dewar-Chatt-Duncanson Model of -bonding N N

TM -donation from occupied 1u or 3g N2 orbital into empty d-orbitals of the metal N TM N N TM N - type back donation from filled dxy, dyz, dxz orbitals to g* orbitals of N2 N TM

Both factors weaken the N-N bonding in nitrogen. The N-N stretching frequencies shift to the red. N 2212 n=7 Spectra show a red shift of 95 cm-1 for n=4 as compared to free N2 stretch 2214 n=6 An additional red shift of 60 cm-1 is observed for n>4 cluster sizes 2204 n=5 n=4 The spectra of n=6 has a lower S/N ratio suggesting the complex is harder to dissociate owing to unusual stability

2265 n=3 2100 2200 -1 cm 2300 2400 B3LYP/ DGDZVP Nb+ 6-311+G* N De= 33.8 kcal/mol Freq = 2291 cm-1 Osc. Strength = 55 km/mol De= 18.6 kcal/mol Freq = 2160 cm-1 Osc. Strength = 169 km/mol De= 8.3 kcal/mol Freq = 2209 cm-1

Osc. Strength = 376 km/mol De= 19.7 kcal/mol Freq = 2262 cm-1 Osc. Strength = 354 km/mol 1. DFT calculations favor linear over T-shaped structures ( De ~ 15 20 kcal/mol 2. T-shaped complexes red-shift N-N stretch by 150-200 cm-1 whereas linear complexes red shift by 50-100 cm-1. Nb+ + Nb (N2)3 Grnd state: 4d4 5D 1st state: 4d35s 5F 6.7 kcal/mol 2nd state: 4d4 3P 15.9 kcal/mol 5 A1 3 B2

2100 2200 2300 2400 Spectrum has two modes because there are only two equivalent N2 + Nb (N2)4 2265 Single peak spectrum points to a high symmetry structure. DFT (B3LYP) calculations for the n = 4 complex for the 5D spin state show good correspondence to the IR spectra. 5

B2g 3 A1g 2100 2200 2300 -1 cm 2400 What is causing the additional red shifts for the n>4 clusters ? Nb+(N2)5 1. 3 A2 In addition all spectra are single peak

signifying that no isomers are present. 5 B1 2. 2100 2200 Other structures such as T-shaped or inserted complexes? DFT studies consistently predict linear structures over T-shaped structures. Energy differences ~ 15 kcal/mol and 20 kcal/mol. 2300 2400 A change in spin state? DFT (B3LYP) calculations for the n = 5 for triplet spin state shows better correspondence to IR spectrum than the quintet state. Also triplet state is found to be lower in energy by ~ 15 kcal/mol

Comparison of Nb+(N2)n and V+(N2)n Nb+(N2)n Greater red-shifts for Nb+(N2)n than V+(N2)n 2271 2212 2258 n=7 n=6 V+(N2)n n=7 2214 n=6 2271 2204 2258 n=5 n=5

2288 n=4 2265 n=4 n=3 2100 n=3 2200 2300 2400 2500 2100 2200 2300 2400

2500 N N TM N N TM 1. N2 and CO are -accepting ligands and so dback donation is expected to dominate the bonding interaction. 2. d orbitals more diffuse for second row TM leading to better s-d hybridization.

3. Frequency shifts for V+(N2)n and Nb+(N2)n seems to justify this reasoning. Conclusions IR spectroscopy coupled with DFT calculations of Nb+(N2)n reveals the structures of these clusters. The spectra show that N2 binds in an end on configuration to Nb+. The results also reveal possible evidence for a change in multiplicity in the metal cation due to solvation effects. The N-N stretch in Nb+(N2)n red shifts further than in V+(N2)n consistent with the previous conclusions based on various TM(CO)n systems that -back donation is the more significant interaction in these TM-ligand systems. + Nb (N2)n n=10 n=9 n=8 2212 n=7

2214 n=6 2100 2200 2300 2400