Kinetically Controlled Formation of Octahedral trans-Dicarbonyl Iron(II

Apr 29, 2010 - Ligand field influence on the electronic and magnetic properties of quasi-linear two-coordinate iron( ii ) complexes. Nicholas F. Chilt...
0 downloads 0 Views 2MB Size
4932

Organometallics 2010, 29, 4932–4942 DOI: 10.1021/om1001638

Kinetically Controlled Formation of Octahedral trans-Dicarbonyl Iron(II) PNP Pincer Complexes: The Decisive Role of Spin-State Changes§ David Benito-Garagorri,† Luis Gonc-alo Alves,† Luis F. Veiros,^ Christina M. Standfest-Hauser,† Shinji Tanaka,† Kurt Mereiter,‡ and Karl Kirchner*,† †

Institute of Applied Synthetic Chemistry and ‡Institute of Chemical Technologies and Analytics, Vienna University of Technology, Getreidemarkt 9, A-1060 Vienna, Austria, ^Centro de Quı´mica Estrutural, Instituto Superior T ecnico, 1049-001 Lisboa, Portugal Received February 28, 2010

Treatment of either cis-[Fe(PNP)(X2)(CO)], trans-[Fe(PNP)(X2)(CO)], or [Fe(PNP)X2] (X = Cl, Br; PNP are tridentate pincer-type ligands based on 2,6-diaminopyridine and 2,6-diaminopyrimidine) with 1 equiv of AgBF4 in the presence of CO afforded selectively octahedral iron(II) complexes of the type trans-[Fe(PNP)(CO)2X]þ. The same reaction carried out with trans-[Fe(PNPiPr)(Cl)2(CO)] in the absence of CO affords also trans-[Fe(PNP-iPr)(CO)2Cl]þ together with unidentified paramagnetic species. This reaction involves an intermolecular CO transfer between coordinately unsaturated [Fe(PNP-iPr)(CO)(Cl)]þ intermediates. In all reactions studied, there was no evidence for the formation of cis dicarbonyl complexes. X-ray structures of representative complexes are presented. A detailed mechanism, based on DFT/B3LYP calculations, is presented, suggesting that upon irreversible removal of X- transient cationic intermediates [Fe(PNP)(CO)(X)]þ of two conformations, one with the CO in the apical and the halide in the basal position (A) and vice versa (B), are formed. These adopt a singlet ground state in the case of A and a triplet ground state in the case of B. The formation of trans-[Fe(PNP)(CO)2X]þ is kinetically controlled, with A in the singlet ground state being the key intermediate. Pathways originating from complexes with a triplet ground state are “spin-blocked” (spin forbidden) or thermodynamically disfavored.

Introduction Carbon monoxide gives rise to some of the largest splittings of ligand field energy levels of transition metal complexes. Accordingly, addition of CO to coordinatively unsaturated high-spin inorganic or organometallic iron(II) compounds is typically accompanied by a high-spin/lowspin crossover from S = 2 or S = 1 to S = 0.1,2 Spin-state changes are of particular relevance in iron(II)-heme systems, where the spin state changes from high-spin Fe(II) (S = 2) to low-spin Fe(II) (S = 0) on CO binding.3 However, whether § Part of the Dietmar Seyferth Festschrift. In honor of Prof. Dietmar Seyferth for his outstanding contributions as editor of Organometallics. *Corresponding author. E-mail: [email protected]. (1) For a review of open-shell complexes see: Poli, R. Chem. Rev. 1996, 96, 2135. (2) (a) Keogh, D. W.; Poli, R. J. Am. Chem. Soc. 1997, 119, 2516. (b) Carreon-Macedo, J.-L.; Harvey, J. N. J. Am. Chem. Soc. 2004, 126, 5789. (c) Strickland, N.; Harvey, J. N. J. Phys. Chem. B 2007, 111, 841. (3) (a) Alberding, N.; Austin, R. H.; Chan, S. S.; Eisenstein, L.; Frauenfelder, H.; Good, D.; Kaufmann, K.; Marden, M.; Nordlund, T. M.; Reinisch, L.; Reynolds, A. H.; Sorensen, L. B.; Wagner, G. C.; Yue, K. T. Biophys. J. 1978, 24, 319. (b) McMahon, B. H.; Stojkovic, B. P.; Hay, P. J.; Martin, R. L.; Garcia, A. E. J. Chem. Phys. 2000, 113, 6831. (c) Thompson, D. W.; Kretzer, R. M.; Lebeau, E. L.; Scaltrito, D. V.; Ghilardi, R. A.; Lam, K.-C.; Rheingold, A. L.; Karlin, K. D.; Meyer, G. J. Inorg. Chem. 2003, 42, 5211. (d) Spin Crossover in Transition Metal Compounds I; Topics in Current Chemistry 235; G€ utlich, P., Goodwin, H. A., Eds.; Springer-Verlag: Berlin, 2004. (e) Harvey, J. N. J. Am. Chem. Soc. 2000, 122, 12401.

pubs.acs.org/Organometallics

Published on Web 04/29/2010

or not carbonylation actually takes place depends strongly on the number and the nature of co-ligands, the complex geometry, and the spin state of the metal center. It is thus not surprising that coordinatively unsaturated high-spin iron(II) complexes exhibit a large range in reactivity toward CO. Binding of CO can be very strong and irreversible, weak and reversible, or completely rejected if spin-state changes are kinetically or thermodynamically disfavored.4-7 In this case the terms “spin-forbidden” or “spin-blocked” are commonly used. In this context it is interesting to note that high-spin iron(II) CO complexes, according to our knowledge, are unknown in the literature. We have recently reported8,9 that the 16e- high-spin complex Fe(PNP)Cl2 readily adds CO to afford, depending (4) Hardman, N. J.; Fang, X.; Scott, B. L.; Wright, R. J.; Martin, R. L.; J. Kubas, G. J. Inorg. Chem. 2005, 44, 8306. (5) Stynes, D. V.; Hui, Y. S.; Chew, V. Inorg. Chem. 1982, 21, 1222. (6) Ellison, J. J.; Nienstedt, A.; Shoner, S. C.; Barnhart, D.; Cowen, J. A.; Kovacs, J. A. J. Am. Chem. Soc. 1998, 120, 5691. (7) (a) Danopoulus, A. A.; Pugh, D.; Smith, H.; Sassmansshausen, J. Chem.;Eur. J. 2009, 15, 5491. (b) Henry, R. M.; Shoemaker, R. K.; Newell, R. H.; Jacobsen, G. M.; DuBois, D. L.; Rakowski DuBois, M. Organometallics 2005, 24, 2481. (c) Breuer, J.; Fruhauf, H.-W.; Smeets, W. J. J.; Spek, A. L. Inorg. Chim. Acta 1999, 291, 438. (8) Benito-Garagorri, D.; Puchberger, M.; Mereiter, K.; Kirchner, K. Angew. Chem., Int. Ed. 2008, 47, 9142. (9) Benito-Garagorri, D.; Alves, L. G.; Puchberger, M.; Mereiter, K.; Veiros, L. F.; Calhorda, M. J.; Carvalho, M. D.; Ferreira, L. P.; Godinho, M.; Kirchner, K. Organometallics. 2009, 28, 6902. r 2010 American Chemical Society

Article

Organometallics, Vol. 29, No. 21, 2010 Scheme 1

on the reaction conditions, either cis-[Fe(PNP)(Cl2)CO] or trans-[Fe(PNP)(Cl2)CO] (PNP are pincer-type ligands based on 2,6-diaminopyridine and 2,6-diaminopyrimidine) (Scheme 1). These reactions are accompanied by a spin-state change from S = 2 to S = 0 and are fully reversible. In the course of this investigation we have also observed that in solution the cis-chloro complexes isomerize to form the thermodynamically more stable trans complexes. This reaction involves chloride dissociation and the formation of transient cationic intermediates [Fe(PNP)(CO)(Cl)]þ, which exist in two conformations, one with the CO in the apical and the Cl in the basal position (A) and vice versa (B), and two different spin ground states (Scheme 2). In the case of A the singlet ground state 1A is energetically favored over the triplet state 3A, while in the case of B the stability order is reversed, with the triplet state 3B being more stable (“1” and “3” superscripts denote the spin multiplicity (2Sþ1)). On the basis of recent DFT calculations,9 the energetically most favorable pathway proceeds via 1A, 3A, and 3B and involves spin crossover steps (vide infra). There was, however, no indication for the formation of singlet or triplet (paramagnetic) [Fe(PNP)(CO)(Cl)]þ species when the isomerization process was monitored by 1H and 31P{1H} NMR spectroscopy. In order to obtain further evidence for the intermediacy of the coordinatively unsaturated intermediates [Fe(PNP)(CO)X]þ (A and/or B), in particular for the involvement of high-spin species 3A and/or 3B, we describe in this article the reactions of cis- and trans-[Fe(PNP)(CO)X2] (X = Cl, Br) complexes with halide scavengers (Agþ, Naþ) in both the presence and absence of CO, with the surprising result that selectively trans-dicarbonyl complexes of the type trans-[Fe(PNP)(CO)2X]þ were formed (Scheme 2). Based on DFT calculations, a detailed mechanistic proposal is also presented identifying 1A as a key intermediate.

Results and Discussion Treatment of cis- or trans-[Fe(PNP-iPr)(CO)Cl2] (1a/2a), cis- or trans-[Fe(PNPClpym-iPr)(CO)Cl2] (1b/2b), and cisor trans-[Fe(PNPEtOpym-iPr)(CO)Cl2] (1c/2c), respectively, with 1 equiv of AgBF4 in nitromethane (or acetone) in the presence of CO at room temperature afforded selectively the cationic complexes trans-[Fe(PNP-iPr)(CO)2Cl]BF4 (5a), trans-[Fe(PNPClpym-iPr)(CO)2Cl]BF4 (5b), and trans-[Fe(PNPEtOpym-iPr)(CO)2Cl]BF4 (5c), respectively, in 82 to 95% isolated yields (Scheme 3). The outcome of these reactions is independent of whether cis- or trans-[Fe(PNP)-

4933

(CO)Cl2] is utilized. Alternatively, complexes 5a-c were also obtained by reacting directly Fe(PNP-iPr)Cl2 (3a), Fe(PNPClpym-iPr)Cl2 (3b), and Fe(PNPEtOpym-iPr)Cl2 (3c), respectively, with 1 equiv of AgBF4 in the presence of CO (Scheme 4). The analogous bromide complexes trans-[Fe(PNP-iPr)(CO)2Br]BF4 (6a) and trans-[Fe(PNPClpym-iPr)(CO)2Br]BF4 (6b) were obtained in a similar fashion from Fe(PNP-iPr)Br2 (4a) and Fe(PNPClpym-iPr)Br2 (4b) according to Scheme 4. Complexes 5 and 6 are thermally robust red solids that are air stable both in the solid state and in solution for several days. Characterization was accomplished by elemental analysis and 1H, 13C{1H}, and 31P{1H} NMR and IR spectroscopy. In addition, the solid-state structures of 5a and 6a were determined by single-crystal X-ray diffraction. The 31P{1H} NMR spectra of complexes 5a and 6a show a singlet resonance at 119.2 and 118.5 ppm, respectively, while complexes 5b, 5c, and 6b give rise to two doublets centered at 116.8/102.3, 114.2/101.8, and 116.9/101.5 ppm, respectively, with JPP coupling constants of 63-82 Hz, which is consistent with a trans-P,P configuration. In the 13C{1H} NMR spectra of 5 and 6 the two CO ligands exhibit a single low-intensity triplet resonance in the range 208 to 211 ppm, thus clearly revealing that the two CO ligands are trans to one another. Complexes 5 and 6 give rise to only one band between 2011 and 2020 cm-1 in the IR spectrum (cf. 2143 cm-1 in free CO) for the mutually trans CO ligands, which can be assigned to the asymmetric CO stretching frequency. The symmetric CO stretching band is IR inactive and thus not observed. The selective formation of octahedral trans dicarbonyl iron(II) complexes is surprising, since simple bonding considerations would suggest the cis isomers to be the more stable ones. Only a few octahedral iron(II) trans-dicarbonyl complexes are reported in the literature, e.g., trans-Fe(OEP)(CO)2 (OEP = octaethylporphyrin),10 trans-[Fe(CN)4(CO)2]2-,11 and cis-Fe(CO)4X2 (X = Cl, Br, I).12 DFT calculations with the model complex [Fe(PNP-iPr)(CO)2Cl]þ indeed confirm that the trans-dicarbonyl complex is thermodynamically disfavored by 5.9 kcal/mol (solvent corrected, solvent = nitromethane) over the respective cis-dicarbonyl compound (vide infra). It has to be noted that the synthesis of a cis isomer, viz., cis-[Fe(PNP-Ph)(CO)2Br]BPh4, has been described recently by following a different methodology.13 Attempts to isolate or at least spectroscopically detect intermediates [Fe(PNP-iPr)(CO)(Cl)]þ (A and/or B) failed. In fact, treatment of 1a or 2a with 1 equiv of Agþ (SbF6-, PF6-) or Naþ (BAr’4- with Ar0 = 3,5-C6H3(CF3)2)) salts in nitromethane afforded trans-[Fe(PNP-iPr)(CO)2(Cl)]þ (5a) and one (or more) as yet unidentified paramagnetic species (Scheme 5). The identity of 5a was unequivocally established by 1H and 31P{1H} NMR as well as IR spectroscopy (vide supra). This reaction obviously involves an intermolecular CO transfer between coordinately unsaturated [Fe(PNPiPr)(CO)(Cl)]þ intermediates that are formed in situ upon (10) Silvernail, N. J.; Noll, B. C.; Schulz, C. E.; Scheidt, W. R. Inorg. Chem. 2006, 45, 7050. (11) Jiang, J.; Koch, S. A. Angew. Chem., Int. Ed. 2001, 40, 2629. (12) (a) Hieber, W.; Bader, G. Chem. Ber. 1928, 61, 1717. (b) Robertson, E. W.; Wilkin, O. M.; Young, N. A. Polyhedron 2000, 19, 1493, and references therein. (13) (a) Benito-Garagorri, D.; Wiedermann, J.; Pollak, M.; Mereiter, K.; Kirchner, K. Organometallics 2007, 26, 217. (b) Benito-Garagorri, D.; Becker, E.; Wiedermann, J.; Lackner, W.; Pollak, M.; Mereiter, K.; Kisala, J.; Kirchner, K. Organometallics 2006, 25, 1900. (c) Benito-Garagorri, D.; Kirchner, K. Acc. Chem. Res. 2008, 41, 201.

4934

Organometallics, Vol. 29, No. 21, 2010

Benito-Garagorri et al. Scheme 2

Scheme 3

Scheme 4

irreversible removal of a chloride ligand from 1a or 2a. On the basis of these experimental findings it is reasonable to assume that isomer A, i.e., where CO is in the apical and chloride in the basal position, is the key intermediate in solution. ORTEP diagrams of 5a0 (with CF3SO3- instead of BF4- as counterion) and 6a 3 (C2H5)2O are depicted in Figures 1 and 2, with selected bond distances and angles reported in the captions (structural information for 5a is provided in the Supporting Information). Both Fe complexes adopt a distorted octahedral geometry around the metal center with the CO ligands being in trans position to one another. The PNP ligand is coordinated to the iron center in a typical tridentate meridional mode, with P-Fe-P angles of 167.80(3) and 167.89(3). The bond lengths around Fe in 5a0 are Fe-P(1) 2.2557(9) A˚, Fe-P(2) 2.2558(9) A˚, Fe-N(1) 1.987(3) A˚, Fe-C(18) 1.831(3) A˚, Fe-C(19) 1.808(3) A˚, and Fe-Cl 2.302(1) A˚. The corresponding bond lengths in 5a0 (with BF4-) and 6a are very similar, except for Fe-Br 2.4461(5) A˚. Despite having significantly different crystal structure architectures, all these complexes have analogous conformations with regard to the orientations of the iPr groups relative to the FePNP core, i.e., two of the iPr groups trans to each other have the CH group exo-oriented (torsion angles Fe-P-C-H 172 to 179), whereas the other two have it endo-oriented (torsion angle Fe-P-C-H 35 to 58; C-H inward oriented to pyridine N). In context with this feature there are small systematic distortions of the FeNP2C2(Cl,Br)

coordination octahedron with P-Fe-C bond angles at about 88 for angles open to the endo-oriented iPr groups and P-Fe-C bond angles at about 93 for the complementary ones. All three compounds show as a typical feature hydrogen bonds between the NH groups of the cationic Fe(PNP) complexes and the counterions BF4- and CF3SO3(here with O as acceptors). There, N 3 3 3 F,O distances are about 2.90 A˚ for F as well as for O. These hydrogen bonds lead to zigzag chains of hydrogen-bonded FePNP complexes and counterions in all three compounds. Thus, in the absence of detectable intermediates on the way from either [Fe(PNP)(Cl)2], cis-[Fe(PNP)(Cl)2CO], or trans-[Fe(PNP)(Cl)2CO] to the dicarbonyl complexes trans-[Fe(PNP)(CO)2Cl]þ we performed DFT (B3LYP) calculations using GAUSSIAN 03. The reliability of the computational method (details in the Experimental Section) was tested before in the reaction of 3a with CO as well as in the isomerization reaction between 1a and 2a.9 In addition, the performance of the theoretical model is further supported by the good agreement between the calculated geometries of 2a and trans-[Fe(PNP-iPr)(CO)2(Cl)]þ (5a0 ) and their X-ray structures. This is shown by the maximum (Δ) and the mean (δ) absolute deviations for all coordination distances, Fe-X: Δ = 0.07 A˚ (2a and 5a0 ); δ = 0.05 (2a) and 0.04 (5a0 ) A˚. Upon irreversible removal of chloride by halide scavengers from cis-[Fe(PNP-iPr)(Cl)2CO] (1a) and trans-[Fe(PNPiPr)(Cl)2CO] (2a), respectively, the pentacoordinated intermediates A and B are formed (Figure 3). The singlet ground state 1A is energetically favored by 4.9 kcal mol-1 over the triplet state 3A, whereas in the case of B the stability order is reversed. The triplet state 3B is more stable than the singlet state 1B by 11.4 kcal mol-1, but lying 2.0 kcal mol-1 lower than 1A. This process is accompanied by a spin-state change (spin crossover) from S = 0 to S = 1. The minimum energy

Article

Organometallics, Vol. 29, No. 21, 2010

Figure 1. Structural view of trans-[Fe(PNP-iPr)(CO)2Cl]CF3SO3 (5a0 ) showing 50% thermal ellipsoids (CF3SO3- anion and C-bound H atoms omitted for clarity). Selected bond lengths (A˚) and angles (deg): Fe1-P1 2.2557(9), Fe1-P2 2.2558(9), Fe1-N1 1.987(3), Fe1-C18 1.831(3), Fe1-C19 1.808(3), Fe1Cl1 2.302(1), P1-Fe1-P2 167.80(3), C18-Fe1-C19 171.3(2).

4935

Figure 2. Structural view of trans-[Fe(PNP-iPr)(CO)2Br]BF4 3 (C2H5)2O (6a 3 (C2H5)2O)) showing 50% thermal ellipsoids (BF4- anion, C-bound H atoms, and solvent omitted for clarity). The complex has C2 symmetry with Br, Fe, N1, and C3 located on the 2-fold axis. Selected bond lengths (A˚) and angles (deg): Fe1-P1 and Fe1-P1A 2.2548(5), Fe1-N1 1.987(2), Fe1-C10 and Fe1-C10A 1.824(2), Fe1-Br1 2.4461(5), P1-Fe1-P1A 167.89(3), C10-Fe1-C10A 172.2(1).

Scheme 5

crossing point (MECP)14 between the potential energy surfaces (PES) of the two spin states (CPA) is easily accessible, lying 5.1 kcal mol-1 above 1A and merely 0.2 kcal mol-1 above 3A. The isomerization reaction takes place via transition state 3TSAB, which is readily reached from both 3A and 3 B with energy barriers of 0.4 and 7.3 kcal mol-1, respectively. The alternative isomerization pathway via intermediate 1B, following the S = 0 surface, is both kinetically and thermodynamically unfavorable. The energy barrier associated with this isomerization pathway is 14.1 kcal mol-1 and thus 8.8 kcal mol-1 higher than the S = 1 barrier corresponding to 3TSAB. In addition, the isomer 1B is the least stable of all [Fe(PNP)(Cl)CO]þ isomers. In principle, addition of CO to the pentacoordinated [Fe(PNP)(Cl)CO]þ intermediates can proceed from either of the two isomers (A or B) in either spin state (S = 0 or S = 1). All possible pathways were investigated. (14) (a) In the MECP both the energy and the geometry of the molecule are the same in the two spin-state surfaces. Once that point (MECP) is reached, following the reaction coordinate, there is a given probability for the system to change spin state and hop from one PES to the other, giving rise to the “spin-forbidden” reaction. (b) For more information about MECP and the kinetics of spin-forbidden reactions see, for example: Harvey, J. N. Phys. Chem. Chem. Phys. 2007, 9, 331.

Considering the isomers with an apical chloride ligand (isomers B) as the starting point, the relevant energy profile is depicted in Figure 4. CO addition to 1B has proven to be a barrierless process at the theory level employed (see Computational Details). In fact, 1B readily adds CO to yield 1C, an octahedral cis-dicarbonyl complex, with an overall stabilization of 32.3 kcal mol-1 (right side of Figure 4). Since 1C is not the product experimentally observed, 1B is obviously not an intermediate present in solution, corroborating that isomerization between 1a and 2a occurring in solution follows the S = 1 surface from 1A via 3A to 3B as already discussed above (Figure 3). Another possibility is a direct CO addition to the triplet intermediate 3B, resulting in the formation of 3C with a triplet spin state that subsequently relaxes to the final singlet product 1C. However, 3C is not stable and spontaneously loses CO to regenerate 3B. Interestingly, it was possible to calculate a path for CO addition to 3B leading to the observed product trans-[Fe(PNP-iPr)(CO)2(Cl)]þ (1D/5a). The corresponding energy profile is represented on the left side of Figure 4. Starting from 3B carbonylation takes place along the S = 1 surface via transition state 3TSBD, where the incoming CO ligand is only 2.58 A˚ away from the metal (dFe-C(CO) = 1.87 A˚, in 3D).

4936

Organometallics, Vol. 29, No. 21, 2010

Benito-Garagorri et al.

Figure 3. Energy profile (B3LYP) for the formation of pentacoordinated intermediates [Fe(PNP-iPr)(CO)Cl]þ (A and B) from cis-[Fe(PNP-iPr)(CO)Cl2] (1a) and trans-[Fe(PNP-iPr)(CO)Cl2] (2a). The energy values (kcal mol-1, solvent corrected, solvent = CH3NO2) are referred to the cationic intermediate [Fe(PNP-iPr)(CO)Cl]þ (1A).

Figure 4. Energy profile (B3LYP) for the addition of CO to the cationic intermediate [Fe(PNP-iPr)(CO)Cl]þ (B) via alternative pathways involving spin-state changes (energies in kcal mol-1, referred to 1A, solvent corrected, solvent = CH3NO2). The left and right curves correspond to the spin-singlet PES (S = 0), and the central curve to the spin-triplet PES (S = 1).

Moreover, in 3TSBD the geometry of the complex changed considerably from the reagent 3B as the CO ligand is moving from the equatorial position in 3B to the apical position in 3D with a concomitant shift of the chloride ligand from the apical position to the equatorial one. This geometry rearrangement is well advanced once 3TSBD is reached, as shown by the relevant angles: N-Fe-CO = 171 (3B) and 123 (3TSBD); N-Fe-Cl = 97 (3B) and 148 (3TSBD). The

energy barrier associated with this step is accessible (10.8 kcal mol-1), and, most importantly, 3TSBD is still more stable than 1B by 0.6 kcal mol-1, indicating that if the reaction of CO addition starts from B, then the preferred path should proceed along the S = 1 spin surface from 3B to 3 D via 3TSBD rather than following the spin singlet surface from 1B to 1C. The final product 1D/5a is formed from 3D after a change from the S = 1 to the S = 0 spin surfaces. The

Article

Organometallics, Vol. 29, No. 21, 2010

4937

Figure 5. Energy profile (B3LYP) for the self-exchange reaction of [Fe(PNP-iPr)(CO)Cl]þ (3A) with CO (energy values in kcal mol-1, referred to 1A, solvent corrected, solvent = CH3NO2).

Figure 6. Energy profile (B3LYP) for the addition of CO to the cationic intermediate [Fe(PNP-iPr)(CO)Cl]þ (1A) via two different pathways on the S = 0 PES (energy values in kcal mol-1, referred to 1A, solvent corrected, solvent = CH3NO2).

crossover occurs through a crossing point CPD, easily reached from 3D going over a negligible barrier of 0.3 kcal mol-1. Alternatively, the participation of the pentacoordinated intermediates A with apical CO as the starting points was also investigated. In the case of the spin triplet intermediate 3 A we were unable to obtain any path leading to the formation of 3D following the S = 1 surface despite numerous attempts (Figure 5). Instead, the only transition state involving CO addition to 3A that could be located was 3TSAA, which corresponds to a CO self-exchange. In this process one CO ligand is expelled from the metal coordination sphere (dFe-C(CO) = 2.25 A˚, in 3TSAA) as a new CO molecule

coordinates to the Fe atom in the opposite apical position (dFe-C(CO) = 2. 52 A˚, in 3TSAA). The calculated energy barrier of 8.4 kcal mol-1 denotes a facile process. Finally, the energy profiles corresponding to CO addition to 1A along the spin singlet surface are represented in Figure 6. Two products can be obtained from CO addition to 1A. One is the thermodynamic product cis-[Fe(PNP-iPr)(CO)2(Cl)]þ (1C), which is experimentally not observed. The reaction from 1A to 1C follows a single-step mechanism through transition state 1TSAC starting from 1AC 3 CO (a pair of reactants 1A and CO). 1AC 3 CO is essentially isoenergetic with respect to the two separated reactants (within 0.4 kcal mol-1). In 1TSAC the new Fe-C(CO) bond is only incipient

4938

Organometallics, Vol. 29, No. 21, 2010

Benito-Garagorri et al. Scheme 6

barrier associated with the formation of 1D going through TSAD is 5.4 kcal mol-1 and is thus the lowest among all carbonylation processes considered. Thus, in agreement with our experimental findings, trans-[Fe(PNP-iPr)(CO)2(Cl)]þ (1D/5a) is the sole reaction product that is formed on kinetic rather than thermodynamic grounds. The fact that CO addition at singlet intermediates is more favorable than at triplet intermediates is readily explained by the frontier orbitals of the relevant species (Figure 7). The LUMOs of the pentacoordinated intermediates with a singlet spin state (1A and 1B) are formed mainly by z2-type orbitals centered in the Fe atom and pointing toward the empty coordination position (Figure 7a). Therefore, these orbitals are ready to receive a pair of electrons from a ligand that would occupy the sixth coordination position (CO in this case) and establish the corresponding σ-bond. On the other hand, in the case of spin triplet intermediates (3A and 3 B), the corresponding orbitals are occupied, being, in fact, the second highest single occupied molecular orbitals (SOMO-1) of these species (Figure 7b). These are not readily available to receive the electron pair from a putative incoming CO, making addition of this ligand a difficult process. In fact, the first empty orbitals (LUMO) in the case of the triplet intermediates correspond to x2-y2-type orbitals centered on the metal and antibonding (σ*) with respect to the four ligands in the equatorial plane. Despite being coordinatively unsaturated, complexes 3 and 4 do not react with simple σ-donor ligands such as THF, MeOH, acetone, DMSO, pyridine, or tertiary phosphines (e.g., PPh3 and PMe3), but react readily with the strong π-acceptor ligand CO. We therefore were interested to see whether CN-tBu reacts similarly to CO to afford complexes of the type [Fe(PNP-iPr)(Cl)(CN-tBu)2)]þ. In the absence of a silver salt, treatment of 3a with g3 equiv of CN-tBu in acetone at room temperature yielded [Fe(PNPiPr)(CN-tBu)3]2þ (7) with chloride as counterion (Scheme 6). If 3a is reacted with 1 or 2 equiv of CN-tBu, complex 7 is also formed together with a minor as yet unidentified species, presumably [Fe(PNP-iPr)(Cl)(CN-tBu)2]þ (31P{1H} NMR: 125.4 ppm, cf. 134.5 ppm in the case of 7), while substantial amounts of 3a remain unreacted. If 3a is treated with g3 equiv of CN-tBu in the presence of 2 equiv of AgBF4, 7 is obtained with BF4- as counterion. A structural view of [Fe(PNP-iPr)(CN-tBu)3](BF4)2 (7) in the form of the nitromethane solvate [Fe(PNP-iPr)(CNtBu)3](BF4)2 3 (CH3NO2)3 (7 3 (CH3NO2)3) is depicted in Figure 8, with selected bond distances and angles reported in the caption. Noteworthy, a similar reaction takes place if 3a is dissolved in neat CH3CN in the presence of 2 equiv of AgBF4, affording [Fe(PNP-iPr)(CH3CN)3]2þ (8) (Scheme 6). The synthesis of this complex has been reported elsewhere.13a In the absence of silver salts even after 5 days at room temperature only small amounts ( 2σ(I) 9004 6476 3876 13 169 7204 no. of params 596 336 193 582 234 a 0.0496 0.0457 0.0423 0.0691 0.0337 R1 (I > 2σ(I)) 0.0674 0.0626 0.0493 0.0742 0.0388 R1 (all data) 0.1238 0.1186 0.1099 0.1606 0.0840 wR2 (all data)a -0.31/0.59 -0.62/0.98 -0.45/0.87 -0.73/0.88 -0.23/0.82 diff Fourier peaks min./max., e A˚-3 P P P P a R1 = Fo| - |Fc / Fo|; wR2 = { [w(Fo2 - Fc2)2]/ [(w(Fo2)2]}1/2. b Nonstoichiometric solvent content disordered and not included in chemical formula and quantities derived thereof.

basal and 2.300 A˚ apical) and blocked against any coordination by a solvent through two ipso-isopropyl groups, which shield the basal plane of the FeNP2Cl2 pyramid (see Supporting Information CIF). Computational Details. Calculations were performed using the GAUSSIAN 03 software package,19 and the B3LYP functional20 without symmetry constraints. That functional includes a mixture of Hartree-Fock21 exchange with DFT22 exchange-correlation, given by Becke’s three-parameter functional with the Lee, Yang, and Parr correlation functional, which includes both local and nonlocal terms. The optimized geometries were obtained with the Stuttgart/Dresden ECP (SDD) basis set23 to describe the electrons of the iron atom. (19) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.; Montgomery, J. A., Jr.; Vreven, T.; Kudin, K. N.; Burant, J. C.; Millam, J. M.; Iyengar, S. S.; Tomasi, J.; Barone, V.; Mennucci, B.; Cossi, M.; Scalmani, G.; Rega, N.; Petersson, G. A.; Nakatsuji, H.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.; Nakai, H.; Klene, M.; Li, X.; Knox, J. E.; Hratchian, H. P.; Cross, J. B.; Bakken, V.; Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev, O.; Austin, A. J.; Cammi, R.; Pomelli, C.; Ochterski, J. W.; Ayala, P. Y.; Morokuma, K.; Voth, G. A.; Salvador, P.; Dannenberg, J. J.; Zakrzewski, V. G.; Dapprich, S.; Daniels, A. D.; Strain, M. C.; Farkas, O.; Malick, D. K.; Rabuck, A. D.; Raghavachari, K.; Foresman, J. B.; Ortiz, J. V.; Cui, Q.; Baboul, A. G.; Clifford, S.; Cioslowski, J.; Stefanov, B. B.; Liu, G.; Liashenko, A.; Piskorz, P.; Komaromi, I.; Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham, M. A.; Peng, C. Y.; Nanayakkara, A.; Challacombe, M.; Gill, P. M. W.; Johnson, B.; Chen, W.; Wong, M. W.; Gonzalez, C.; Pople, J. A. Gaussian 03, Revision C.02; Gaussian, Inc.: Wallingford, CT, 2004. (20) (a) Becke, A. D. J. Chem. Phys. 1993, 98, 5648. (b) Miehlich, B.; Savin, A.; Stoll, H.; Preuss, H. Chem. Phys. Lett. 1989, 157, 200. (c) Lee, C.; Yang, W.; Parr, G. Phys. Rev. B 1988, 37, 785. (21) Hehre, W. J.; Radom, L.; Schleyer, P. v.R.; Pople, J. A. Ab Initio Molecular Orbital Theory; John Wiley & Sons: New York, 1986. (22) Parr, R. G.; Yang, W. Density Functional Theory of Atoms and Molecules; Oxford University Press: New York, 1989. (23) (a) Haeussermann, U.; Dolg, M.; Stoll, H.; Preuss, H.; Schwertfeger, P. J.; Pitzer, R. M. Mol. Phys. 1993, 78, 1211. (b) Kuechle, W.; Dolg, M.; Stoll, H.; Preuss, H. J. Chem. Phys. 1994, 100, 7535. (c) Leininger, T.; Nicklass, A.; Stoll, H.; Dolg, M.; Schwerdtfeger, P. J. Chem. Phys. 1996, 105, 1052.

For all other atoms the 6-31G** basis set was employed.24 Frequency calculations were performed to confirm the nature of the stationary points, yielding one imaginary frequency for the transition states and none for the minima. Each transition state was further confirmed by following its vibrational mode downhill on both sides and obtaining the minima presented on the energy profiles. Solvent effects (CH3NO2) were considered through single-point energy calculations with the optimized geometries using the polarizable continuum model (PCM) initially devised by Tomasi and co-workers25 as implemented in GAUSSIAN 03.26,27 The molecular cavity was based on the united atom topological model applied on UAHF radii, optimized for the HF/6-31G(d) level. In the case of the MECP single-point PCM calculations were performed for both spin states, yielding energy values differing less than 0.7 kcal mol-1. The MECPs between the spin singlet (S = 0) and the spin triplet (S = 1) potential energy surfaces were determined using a code developed by Harvey et al.28 This code consists of a set of shell scripts and Fortran programs that uses the Gaussian results of energies and gradients of both spin states to produce an effective gradient pointing toward the MECP. Three-dimensional representations of the orbitals were obtained with Molekel.29 (24) (a) McLean, A. D.; Chandler, G. S. J. Chem. Phys. 1980, 72, 5639. (b) Krishnan, R.; Binkley, J. S.; Seeger, R.; Pople, J. A. J. Chem. Phys. 1980, 72, 650. (c) Wachters, A. J. H. Chem. Phys. 1970, 52, 1033. (d) Hay, P. J. J. Chem. Phys. 1977, 66, 4377. (e) Raghavachari, K.; Trucks, G. W. J. Chem. Phys. 1989, 91, 1062. (f) Binning, R. C.; Curtiss, L. A. J. Comput. Chem. 1990, 11, 1206. (g) McGrath, M. P.; Radom, L. J. Chem. Phys. 1991, 94, 511. (25) (a) Cances, M. T.; Mennucci, B.; Tomasi, J. J. Chem. Phys. 1997, 107, 3032. (b) Cossi, M.; Barone, V.; Mennucci, B.; Tomasi, J. Chem. Phys. Lett. 1998, 286, 253. (c) Mennucci, B.; Tomasi, J. J. Chem. Phys. 1997, 106, 5151. (26) Tomasi, J.; Mennucci, B.; Cammi, R. Chem. Rev. 2005, 105, 2999. (27) Cossi, M.; Scalmani, G.; Rega, N.; Barone, V. J. Chem. Phys. 2002, 117, 43. (28) Harvey, J. N.; Aschi, M.; Schwarz, H.; Koch, W. Theor. Chem. Acc. 1998, 99, 95. (29) Portmann, S.; L€ uthi, H. P. Chimia 2000, 54, 766.

4942

Organometallics, Vol. 29, No. 21, 2010

Acknowledgment. D.B.-G. thanks the Basque Government (Eusko Jaurlaritza/Gobierno Vasco) for a doctoral fellowship. Supporting Information Available: Complete crystallographic data and technical details in CIF format for

Benito-Garagorri et al. trans-[Fe(PNP-iPr)(CO)2Cl]BF4 (5a), trans-[Fe(PNP-iPr)(CO)2Cl]CF3SO3 (5a0 ), trans-[Fe(PNP-iPr)(CO)2Br]BF4 3 (C2H5)2O (6a 3 (C2H5)2O)), [Fe(PNP-iPr)(CN-tBu)3](BF4)2 3 (CH3NO2)3 (7 3 (CH3NO2)3), and [Fe(PNP-iPr)Cl2] 3 (CH3CN)x (3a 3 (CH3CN)x). This material is available free of charge via the Internet at http://pubs.acs.org.