[M(CO)4PPh3]·- radicals (M = Cr, Mo, W): DFT and single crystal EPR investigations

(1)

Article

Reference

[M(CO)4PPh3]·- radicals (M = Cr, Mo, W): DFT and single crystal EPR investigations

BERCLAZ, Théo, et al.

Abstract

[M(CO)4PPh3]− (M = Mo, W) were trapped at 77 K in X-irradiated single crystals of M(CO)5PPh3 and studied by EPR. Structures of [M(CO)4PPh3]− (M = Cr, Mo, W) were optimized by DFT; predicted g and 31P-hyperfine tensors agree with experiments for M = Mo, W. The anions adopt a slightly distorted pyramidal structure with PPh3 in basal position and the spin mostly delocalized in a metal-dz2 orbital and carbon-pz orbitals of carbonyls. The EPR tensors are slightly modified by annealing, they suggest that new constraints in the matrix distort the structure of [M(CO)4PPh3]− (M = Cr, Mo, W).

BERCLAZ, Théo, et al . [M(CO)4PPh3]·- radicals (M = Cr, Mo, W): DFT and single crystal EPR investigations. Chemical Physics Letters , 2007, vol. 440, no. 4-6, p. 224-228

DOI : 10.1016/j.cplett.2007.04.052

Available at:

http://archive-ouverte.unige.ch/unige:3595

Disclaimer: layout of this document may differ from the published version.

1 / 1

(2)

[M(CO)

₄

PPh

₃

]

radicals (M = Cr, Mo, W): DFT and single crystal EPR investigations

The´o Berclaz, Bassirou Ndiaye, Shrinivasa Bhat, Abdelaziz Jouaiti, Michel Geoﬀroy

^*

Department of Physical Chemistry, University of Geneva, 30 Quai Ernest Ansermet, 1211 Geneva, Switzerland Received 8 March 2007

Available online 19 April 2007

Abstract

[M(CO)4PPh3](M = Mo, W) were trapped at 77 K in X-irradiated single crystals of M(CO)5PPh3and studied by EPR. Structures of [M(CO)4PPh3](M = Cr, Mo, W) were optimized by DFT; predictedgand³¹P-hyperﬁne tensors agree with experiments for M = Mo, W. The anions adopt a slightly distorted pyramidal structure with PPh₃in basal position and the spin mostly delocalized in a metal-d_z2

orbital and carbon-pzorbitals of carbonyls. The EPR tensors are slightly modiﬁed by annealing, they suggest that new constraints in the matrix distort the structure of [M(CO)₄PPh₃](M = Cr, Mo, W).

1. Introduction

In the last decade considerable eﬀorts have been made to understand the role played by paramagnetic transition metal complexes as reaction intermediates[1,2]. It is gener- ally admitted that these intermediates can be divided into two main classes[3,4]: (i) organic radicals coordinated to a transition metal complex[5], (ii) paramagnetic complexes with a considerable localisation of the unpaired electron on the metal. We have recently shown that (CO)₅M–PR₂, a species of the former class, can be formed from M(CO)5PR3

(with M = Cr, Mo, W)[6]. Here, we show that, by changing the experimental conditions, the same precursor can lead to a metal-centred radical anion [M(CO)4PR3]. Such open- shell ﬁve-coordinate mononuclear complexes have been proposed as intermediates when passing from four- to six- coordinate complexes[7]. We will give a description of the structure of these species as obtained from DFT calculations and compare the predicted EPR parameters with the experimental ones. Due to their short lifetime, it is often dif- ﬁcult to measure the tensors of reaction intermediates; in

the present study this was performed by trapping the paramagnetic species in the crystal matrix of the irradiated precursor.

2. Results 2.1. EPR spectra

Single crystals of M(CO)5PPh3(M = Mo, W, Cr) were X-irradiated in liquid nitrogen and studied by EPR at 77 K. In a ﬁrst set of experiments, any increase of temperature between irradiation and EPR measurements was carefully avoided.

A set of anisotropic lines (signals A hereafter) was recorded with an irradiated crystal of Mo(CO)5PPh3. The angular variation of these signals A, in the three reference planes, clearly reveal hyperﬁne interaction with ³¹P; the corresponding coupling constants have been calculated assuming three negative eigenvalues; they are given in Table 1together with theg-tensor.

Intense signals (marked B onFig. 1) together with small satellite lines (B1 and B2) were detected with W(CO)5PPh3. The angular variations of the signals B, B1 and B2 were analyzed in the three reference planes. The signals B exhibit an hyperﬁne splitting attributed to ³¹P while the satellite

doi:10.1016/j.cplett.2007.04.052

* Corresponding author. Fax: +41 22 379 61 03.

E-mail address:michel.geoﬀroy@chiphy.unige.ch(M. Geoﬀroy).

www.elsevier.com/locate/cplett

(3)

lines B1 and B2 are assigned to ¹⁸³W (natural abun- dance = 14.3%, I= 1/2). The g-tensor as well as the ³¹P coupling constants (calculated in the hypothesis of three negative eigenvalues) are reported inTable 1. After diago- nalization, the absolute values of the¹⁸³W are found to be equal to 27, 30 and 33 MHz. However, the lack of precision on the position of these small satellite lines prevented us from discussing their anisotropy.

The spectrum obtained at 77 K, immediately after X- irradiation of a crystal of Cr(CO)5PPh3, is very complex.

Due to the frequent overlap of the numerous lines it was not possible to follow their angular dependence.

As reported in Ref.[6], at room temperature, single crystals of W(CO)5PPh3, Mo(CO)5PPh3and Cr(CO)5PPh3pre- viously irradiated at 300 K, showed only the signals due to the trapping of the phosphinyl radical linked to the penta- carbonylmetal: (CO)5M–PPh2. However, these spectra drastically changed when the crystals, previously irradiated at 300 K, were studied at 77 K. Probably due to saturation, the lines assigned to (CO)5M–PPh2considerably decreased in intensity, while new signals clearly appeared on the spectra; moreover, the above described signals A and B, gener- ated by irradiation at 77 K, were not detected. The spectral modiﬁcations observed on crystals irradiated at 300 K were reversible.

At 77 K, two sets of anisotropic lines (signals K and L hereafter), characterized by a coupling with a spin 1/2 nucleus, were observed with Mo(CO)5PPh3 irradiated at room temperature. Similar anisotropic lines (signals M and N) were also detected, at the same temperature with a crystal of W(CO)5PPh3irradiated at 300 K. In the same conditions, the spectra obtained with Cr(CO)5PPh3clearly show the trapping of two species (T and U) exhibiting a coupling assigned to ³¹P. The angular variations of these six doublets could be followed in the three reference planes;

the resulting g and hyperﬁne tensors are reported in Table 2.

It is worthwhile remarking that for all species mentioned in Table 1 (irradiation at 77 K or at 300 K) the ³¹P coupling unambiguously reveals the participation of the phos- phine group to the structure of the paramagnetic species; it also shows, however, that the corresponding phosphorus

Table 1

EPR tensors measured at 77 K with single crystals previously X-irradiated either at 77 K or at 300 K

Crystal Irradiation temperature Species g ³¹P-couplings^a(MHz)

Aiso saniso

Mo(CO)5PPh3 77 K A 2.001 2.022 2.024 43 7 2 9

300 K K 2.004 2.018 2.044 16 6 2 4

L 2.000 2.025 2.038 33 10 4 6

W(CO)5PPh3 77 K B 1.991 2.059 2.070 44 7 2 9

300 K M 1.995 2.053 2.122 24 6 1 5

N 2.017 2.025 2.103 47 14 4 10

Cr(CO)5PPh3 300 K T 1.998 2.016 2.039 21 2 1 3

U 1.999 2.020 2.041 26 5 0 5

a Only the absolute values of the principal values are determined experimentally, the isotropic (Aiso) and anisotropic (saniso) coupling constants have been calculated by assuming the combination of signs which agrees with the DFT values.

Fig. 1. Signals assigned to [W(CO)4PPh3] obtained at 77 K with a crystal of W(CO)5PPh3irradiated at 77 K.

Table 2

Geometrical parameters^aobtained for the optimized structures of [M(CO)4PPh3]

M–P M–C6 M–C4 M–C3 M–C5 P–M–C6 C3–M–C4 C4–M–C5^b

½CrðCOÞ₄PPh3 2.361 1.864 1.889 1.884 1.832 173.7 156.05 101.14

½MoðCOÞ₄PPh3 2.525 2.015 2.048 2.049 1.960 172.4 163.4 98.95

½WðCOÞ₄PPh3 2.523 2.007 2.042 2.041 1.957 171.2 163.6 98.1

a Distances in A˚ , angles in degree.

b For the three anions:R1 = C3MC5 + C3MC4 + C4MC5 lies in the range 359.4–359.7andR2 = PMC4 + C4MC6 + C6MC3 + C3MP in the range 357.4–357.8.

T. Berclaz et al. / Chemical Physics Letters 440 (2007) 224–228 225

(4)

spin densities are very small (for comparison a spin density of 0.01 in a 3s orbital of atomic phosphorus leads to Aiso130 MHz, and a density of 0.01 in a 3p orbital to saniso7 MHz)[8].

2.2. DFT calculations

Our attempts to optimize the structure of the excess- electron centre [Mo(CO)5PPh3] remained unsuccessful and led to a dramatic increase of the molybdenum–phosphorus distance. On the other hand, an energy minimum was found for the anions having lost a carbonyl group:

[W(CO)4PPh3], [Mo(CO)4PPh3] and [Cr(CO)4PPh3]. Some of the resulting geometrical parameters are given in Table 2; a representation of [Mo(CO)4PPh3] is given in Fig. 2.

The g and ³¹P-hyperﬁne tensors for the three anions have been calculated, they are reported inTable 3, together with the anglenformed bygminand³¹P-smax.

3. Discussion

The experimentalg-tensors reported inTable 1 for the species (A, B) trapped immediately after X-irradiation at 77 K are considerably more anisotropic than those previ-

ously measured for (CO)5M–PPh2 at room temperature.

Moreover, hyperfine interaction with the metal is clearly detected for the tungsten complex. This is consistent with a greater localization of the unpaired electron on the transition metal and suggests that the species trapped at 77 K result from the capture of electron by the diamagnetic precursor. As mentioned above, [M(CO)5PPh3]tends to dis- sociate. In the crystal matrix, the cage effect being considerably more efficient for PPh3than for CO, it is reasonable to expect the trapping of [M(CO)4PPh3]. More- over, this agrees with previous reports which mention that one-electron reduction of polycarbonylcomplexes (e.g. Cr(CO)4(bpy)) labilises a transition metal–CO bond [9]. Accordingly, as shown inTable 2, a minimum energy structure was found for each of the three anions [M(CO)4PPh3]. In their equilibrium geometry the complexes adopt a structure close to the square pyramid with the PPh₃ group in the basal position. As shown by the C3–M–C4 angle values, less than 180, some distortion towards the trigonal bipyramid occurs, this departure from the idealized symmetry is particularly marked for the chro- mium complex (C3CrC4 = 156). For the three anions, the shortest metal–carbon distance is associated with the carbonyl in axial position.

As indicated by the gross orbital spin populations, for [M(CO)4PPh3] the unpaired electron mainly lies in pz

and d²_z orbitals of the transition metal and in pz orbitals of the carbon atoms of the basal carbonyl groups (10%

on C3,10% on C4, 3% on C6). For the three radical anions the spin density on the phosphorus is considerably smaller than that on the metal (Mulliken atomic spin densities: [Cr(CO)4PPh3]: Cr: 0.65, P: 0.01, [Mo(CO)₄PPh₃]: Mo: 0.58, P: 0.03, [W(CO)₄PPh₃]: W:

0.57, P: 0.03. The SOMO of [Mo(CO)4PPh3]is illustrated inFig. 3.

Consistent with ligand field theory, such a structure leads to ag-tensor characterized by an axial symmetry with g//2.0023 andg_^>g//. The accord between theg values calculated by DFT and the g values measured from the spectra recorded at 77 K immediately after irradiation in liquid nitrogen for the Mo and W species (species A and B) is quite satisfactory. Moreover, for the ³¹P-hyperfine tensors of these two complexes, it is possible to find a combination of signs of the experimental eigenvalues (all signs negative) which leads to isotropic coupling constants that are in reasonable accord with the constants calculated by DFT. Experimental and calculated anisotropic coupling constants are small. Furthermore, the relative orientations

Fig. 2. Geometry of optimized radical anion [Mo(CO)4PPh3].

Table 3

Calculatedgand³¹P-hyperﬁne tensors for [M(CO)4PPh3]

Anion g ³¹P^a Angle (gmin,³¹P-smax)

Aiso saniso

½CrðCOÞ₄PPh3 2.001 2.015 2.017 44.2 4.5 3.3 7.8 12

½MoðCOÞ₄PPh3 2.002 2.023 2.025 27.4 5.2 3.1 8.3 8

½WðCOÞ₄PPh3 2.003 2.079 2.086 36.4 6.5 4.2 10.8 7

a Hyperﬁne coupling in MHz.

(5)

of gmin and ³¹P-smax predicted by DFT are in very good agreement with the experimental results: for the Mo complex (species A) and the tungsten complex (species B), the experimental anglesn formed by³¹P-smaxandgmin

are equal to 15 and 11, respectively, while the corresponding DFT predictions for [Mo(CO)4PPh3] and [W(CO)4PPh3]are equal to 8and 7, respectively.

Annealing the crystals to room temperature not only provokes an appreciable modification of the EPR parameters measured immediately after irradiation at 77 K but also causes an increase of the number of paramagnetic sites (signals A replaced by signals K and L for [Mo(CO)₄- PPh3], signals B replaced by signals M and N for [W(CO)4PPh3]). Nevertheless, the modifications of the g and³¹P tensors are not very drastic; the main feature con- sists in an increase of thegmax value and in a departure of the axiality of theg-tensor. A single point calculation for [Mo(CO)4PPh3], after a small change in the C4MoC5 (102) and C3MoC4 (160.6) angles led to similar modifications for the g eigenvalues: 2.007, 2.026, 2.035. This suggests that reorganisation in the environment of the radical anion caused by the formation of various radio- genic defects (migration of phenyl groups after formation of (CO)5M-PPh2, migration of CO) causes a reorientation of [M(CO)4PPh3]. In this process, the resulting structure of the radical anion is slightly modified; probably the distortion towards a trigonal bipyramid increases (diminution of the C3MC4 angle) and, finally, two distinct sites are observed at the equilibrium. In this context, it is worthwhile mentioning that the EPR spectrum of the radical anion [W(CO)4P(OMe)3] was reported by Preston et al.

[7], this species which resulted from hydrogen atom loss of the diamagnetic stable host anion [W(CO)4- HP(OMe)3)] could adopt a square-pyramidal structure with almost no alteration of the geometry of the host complex. Accordingly, both the symmetry of the g-tensor

(1.9914, 2.0708, 2.0740) and the³¹P-hyperﬁne tensors were rather similar to those found for our species B (observed immediately after irradiation of W(CO)₅PPh₃ at 77 K) and for the DFT values calculated for[W(CO)4PPh3]. 4. Conclusion

EPR spectra show that in crystalline matrices, at 77 K, the electron capture by M(CO)5PPh3 (M = Mo, W) leads to the loss of a CO group and to the trapping of [Mo(CO)4PPh3] and [W(CO)4PPh3]. In good accor- dance with DFT calculations on [M(CO)4PPh3] (M = Cr, Mo, W), these radical anions adopt a structure close to a square pyramidal structure with PPh3in basal position. Warming the irradiated crystal to room temperature, however, slightly alters this structure as revealed by the EPR tensors measured for [Cr(CO)4PPh3], [Mo(CO)4PPh3] and [W(CO)4PPh3] after annealing at 300 K.

5. Experimental

Cr(CO)5PPh3, Mo(CO)5PPh3, W(CO)5PPh3 were syn- thesized by following already described methods [6,10].

Their crystal structures have been reported [6]. The EPR reference axes were oriented with respect to the crystallo- graphic axes: for Cr(CO)5PPh3: x//a, y// b, z//c*; for Mo(CO)5PPh3:x//a*,y//c*; for W(CO)5PPh3:y//b*,z//a.

EPR spectra were recorded on a Bruker 300 spectrome- ter and crystals were irradiated with a Philips X-ray tube equipped with a tungsten anticathode. The EPR tensors were obtained by using an optimization program which adjusts the terms of the spin Hamiltonian in order to obtain the best ﬁtting with the experimental angular variation of the signals in three perpendicular planes [11].

DFT optimizations of the structures were performed with the Turbomole program[12]by using the B-P86 functional[13,14]and the SV (P) basis set[12]for the transition metal and TZVP [15] for H, C, O and P. Minima were characterized with harmonic frequency calculations (no imaginary frequencies). Then, properties were calculated at these optimized geometries, with the GAUSSIAN03 pack- age[16](B3LYP functional[17], SBKJ basis set[18,19]for the transition metal and IGLO-III basis set [20] for the other atoms).

Acknowledgement

We gratefully acknowledge support from the Swiss na- tional Science Foundation.

References

[1] K. Auclair, Z. Hu, D.M. Little, P.R. Ortiz de Montellano, J.T.

Groves, J. Am. Chem. Soc. 124 (2002) 6020.

[2] C. Limberg, Angew. Chem., Int. Ed. 42 (2003) 5932.

[3] R. Poli, Chem. Rev. 96 (1996) 2135.

Fig. 3. Representation of the SOMO for [Mo(CO)4PPh3].

T. Berclaz et al. / Chemical Physics Letters 440 (2007) 224–228 227

(6)

[4] D.G.H. Hetterscheid, J. Kaiser, E. Reijerse, T.P.J. Peters, S.

Thewissen, A.J. Blok, J.M.M. Smits, R. de Gelder, B. de Bruin, J.

Am. Chem. Soc. 127 (2005) 1895.

[5] T. Bu¨ttner, J. Geier, G. Frison, J. Harmer, C. Calle, A. Schweiger, H.

Scho¨nberg, H. Gru¨tzmacher, Science 307 (2005) 235.

[6] B. Ndiaye, S. Bhat, A. Jouaiti, T. Berclaz, G. Bernardinelli, M.

Geoﬀroy, J. Phys. Chem. A 110 (2006) 9736.

[7] R.C. Hynes, K.F. Preston, J.J. Springs, A.J. Williams, Organomet- allics 9 (1990) 2298.

[8] J.R. Morton, K.F. Preston, J. Magn. Reson. 30 (1978) 577.

[9] S. Zalis, C. Daniel, A. Vleek Jr., J. Chem. Soc., Dalton Trans. 3081 (1999).

[10] C.N. Matthew, T.A. Magee, J.H. Wotiz, J. Am. Chem. Soc. 81 (1959) 2273.

[11] E.J. Soulie´, T. Berclaz, Appl. Magn. Reson. 29 (2005) 401.

[12] The Program System Turbomole R. Ahrlichs, M. Ba¨r, M. Ha¨ser, H.

Horn, C. Ko¨lmel, Chem. Phys. Lett. 162 (1989) 165.

[13] R. Ahlrichs, F. Furche, S. Grimme, Chem. Phys. Lett. 325 (2000) 317.

[14] J.P. Perdew, Phys. Rev. B 33 (1986) 8822.

[15] A. Schaefer, H. Horn, R. Ahlrichs, J. Chem. Phys. 100 (1994) 5829.

[16] M.J. Frisch et al., GAUSSIAN-03, Revision B.03, Gaussian Inc., Pittsburg, PA, 2003.

[17] A.D. Becke, J. Chem. Phys. 98 (1993) 5648.

[18] W.J. Stevens, M. Krauss, H. Basch, P.G. Jasien, Can. J. Chem. 70 (1992) 612.

[19] T.R. Cundari, W.J. Stevens, J. Chem. Phys. 98 (1993) 5555.

[20] W. Kutzelnigg, U. Fleischer, M. SchindlerNMR-Basic Principles and Progress, Vol. 23, Springer, Germany, Heidelberg, 1990.