bis(ketopyrrolyl) complexes of co(ii) stabilised by trimethylphosphine ligands

PAPER www.rsc.org/dalton | Dalton Transactions

Bis(ketopyrrolyl) complexes of Co(II) stabilised by trimethylphosphineligands†‡§¶Sonia A. Carabineiro,a Pedro T. Gomes,*a Luıs F. Veiros,a Cristina Freire,b Laura C. J. Pereira,c

Rui T. Henriques,d John E. Warrene and Sofia I. Pascu f

Received 4th July 2007, Accepted 19th September 2007First published as an Advance Article on the web 5th October 2007DOI: 10.1039/b710162g

2-Formylpyrrole and 2-acetylpyrrole were deprotonated with NaH to give the corresponding Na salts 1aand 1b, respectively. The reactivity of these salts towards cobalt chloride compounds was studied. Theresulting new bis(ketopyrrolyl) Co(II) 19-electron complexes [Co(j2N,O-2-NC4H3C(R)=O)2(PMe3)2](R = H 2a, and Me 2b) were characterised by single crystal X-ray diffraction, to show an octahedralgeometry with the PMe3 ligands in trans positions to each other, and two bidentate ketopyrrolyl ligandsoccupying the remaining coordination positions in a transoid conformation. Powder and solutionmagnetic susceptibility measurements together with EPR and UV/Vis/NIR spectra revealed a low-spinground state (dz2 , S = 1/2) for Co(II) in these compounds. Analysis of the EPR superhyperfinecouplings suggested that the longer distances (z axis) of the hexacoordinate Co coordination sphere areoccupied by the keto-O atoms of the bidentate ligand, leaving the pyrrolyl N and the phosphine Patoms within the equatorial plane. This is confirmed by means of DFT calculations, which also indicatethat the most thermodynamically stable isomers are low-spin (S = 1/2) complexes with coordinationgeometries corresponding to the molecular structures obtained by X-ray crystallography.

Introduction

Bidentate 2-iminopyrrolyl ligands (Chart 1, A) have been pre-viously used in the synthesis of transition-metal compounds.1

However, in recent years, there has been a renewed interestedin ligands of this family incorporating bulky aryl substituents.These have been employed as frameworks for olefin polymerisationmetal catalysts and also in structural studies of other transitionmetal complexes.2 In this latter perspective, we have reportedon the synthesis and characterisation of a series of homolepticbis(2-formiminopyrrolyl) and bis(2-acetiminopyrrolyl) complexesof Co(II). Increased ligand bulkiness gave rise to a rare tetrahedral

aCentro de Quımica Estrutural, Departamento de Engenharia Quımica eBiologica, Instituto Superior Tecnico, Av. Rovisco Pais, 1049-001 Lis-boa, Portugal. E-mail: [email protected]; Fax: +351 218419612;Tel: +351 218419612bREQUIMTE/Departamento de Quımica, Faculdade de Ciencias, Univer-sidade do Porto, R. Campo Alegre, 4169-007 Porto, PortugalcDepartamento de Quımica, Instituto Tecnologico e Nuclear, CFMCUL,Estrada Nacional 10, 2686-953 Sacavem, PortugaldInstituto de Telecomunicacoes, Instituto Superior Tecnico, Av. Rovisco Pais,1049-001 Lisboa, PortugaleScience & Technology Facilities Council Daresbury Laboratory, Warring-ton, Cheshire, UK WA4 4ADfChemistry Research Laboratory, University of Oxford, Oxford, UK OX12TA† Dedicated to the memory of Prof. Alberto Romao Dias, who died on15th July 2007.‡ The HTML version of this article has been enhanced with colour images.§ CCDC reference numbers 642094 and 642095. For crystallographic datain CIF or other electronic format see DOI: 10.1039/b710162g¶ Electronic supplementary information (ESI) available: Figures withrepresentations of all the optimised structures and the corresponding tablesof atomic coordinates. Figures with frontier orbitals and spin density forthe two spin isomers of complexes. See DOI: 10.1039/b710162g

to square planar geometrical transition.3 Iminopyrrolyl ligandprecursors are commonly prepared by the condensation of 2-formylpyrrole or 2-acetylpyrrole with primary amines. However,the use of their bidentate N,O 2-ketopyrrolyl analogues (Chart 1,B), which are simultaneously less sterically demanding and weakerelectronic donor ligands than the corresponding isoelectronic N,Ncounterparts, has scarcely been investigated.

Chart 1 2-iminopyrrolyl (A), and 2-ketopyrrolyl (B) ligands (R = H, Me;R′ = alkyl, aryl).

To date, the published examples are mainly concerned withcomplexes of Cu,4 Ni,5 Co,4a,6 Ru,7 and V,8 the latter beingpotent olefin copolymerisation catalysts. In particular, the onlycobalt complexes reported so far supported by the bidentate2-ketopyrrolyl ligand are the Co(III) complexes [Co(j2N,O-2-NC4H3C(R)=O)3],4a [Co(g5-C5R′

5)(j2N,O-2-NC4H3C(R)=O)I]and [Co(g5-C5R′

5)(j2N,O-2-NC4H3C(R)=O)(L)]I (R = H, Me;R′ = H, Me; L = phosphines, phosphites).6 Interestingly, severalrelated cobalt compounds such as Co(II) porphyrinates andCo(III) corrolates have been synthesised via a one-pot synthesis,by the cyclisation of 2-formylpyrrole derivatives catalysed byCoCl2/PPh3.9

In light of our previous studies on [Co(j2N,N-2-iminopy-rrolyl)2] complexes,3 the present work describes the reactivity ofsodium 2-ketopyrrolyl salts towards cobalt chloride compounds.The synthesis of 19-electron stable Co(II) complexes containing

5460 | Dalton Trans., 2007, 5460–5470 This journal is © The Royal Society of Chemistry 2007

bidentate 2-ketopyrrolyl ligands is reported herein. These arethe first Co(II) species containing this type of bidentate N,Oligands and have been characterised by elemental analysis, X-ray diffraction, magnetic susceptibility measurements, EPR andUV/Vis/NIR spectroscopies, and DFT calculations.

Results and discussion

Synthesis of complexes

The Na salts of 2-formylpyrrole and 2-acetylpyrrole (1a and1b, respectively) were generated in situ by deprotonation oftheir neutral precursors using a stoichiometric amount of NaHin THF. Addition of 2 equivalents of 1a or 1b to a THFsuspension of anhydrous CoCl2 gave rise to amorphous powders.The products did not correspond to the desired bis(chelating)[Co(j2N,O-2-ketopyrrolyl)2] complexes, which were expected toform by analogy to the iminopyrrolyl derivatives [Co(j2N,N-2-iminopyrrolyl)2] previously obtained.3 Elemental analysis in-dicated that the unidentified products obtained by this methoddid not even correspond to the tris(chelating) Co(III) complexes[Co(j2N,O-2-ketopyrrolyl)3] (by analogy to the iminopyrrolyl Cocomplexes obtained following the reaction of CoCl2 with stericallyunhindered iminopyrrolyl ligands).10

These facts led us to attempt the stabilisation of the unsaturatedand possibly very reactive “[Co(j2N,O-2-ketopyrrolyl)2]” species,probably generated via the previous routes, with a donor ligandsuch as trimethylphosphine (PMe3). Two general methods havebeen employed (Scheme 1): a) reaction of anhydrous CoCl2

with 2 equivalents of the 2-ketopyrrolyl sodium salt in THF,in the presence of 2 equivalents of PMe3, and b) reaction ofCoCl2(PMe3)2 with 2 equivalents of the corresponding ligandprecursor salt also in THF. Both methods afforded the 19-electron complexes [Co(j2N,O-2-NC4H3C(R)=O)2(PMe3)2] (R =

Scheme 1 Synthesis of complexes 2a and 2b using routes a)–c).

H 2a, and Me 2b) in good yields (>80%) as brown crystalswith paramagnetic properties. Their solutions were particularlyair sensitive.

In the case of 2a, no identifiable products were obtained whenthe reaction was carried out for only 2 h and a longer reactiontime (ca. 12 h) was required. However, the reactivity of 1b towardsCo halides seems much higher since compound 2b formed afteronly 2 h in moderate to high yield (40% and 69%, respectively, formethods a) and b)).

A third method, i.e. the addition of an excess of the 2-acetylpyrrolyl ligand salt to CoCl(PMe3)3 in THF, gave rise toformation of compound 2b. This route involves oxidation of Co(I)to Co(II) (Scheme 1). Although less rational, this route was foundto be very effective since, in 2 h, a yield of 85% was obtainedfor compound 2b. However, this method was not successful forcompound 2a when the same reaction time was used.

X-Ray diffraction

Crystals of 2a and 2b suitable for X-ray diffraction analysis wereobtained, enabling the determination of their molecular structures.Selected bond distances (A) and angles (◦) are listed in Table 1.Both complexes show asymmetric units formed by a half moleculeof [Co(2-ketopyrrolyl)2(PMe3)2], with the Co atom occupyinga special position. The corresponding molecular structures aregenerated by symmetry (Fig. 1). The geometries of compounds 2aand 2b are practically octahedral around the Co atoms, since bondangles P1–Co1–O1 and P1–Co1–N1 are, respectively, ca. 88.5 andca. 91.5◦. The two PMe3 ligands are placed trans to each other,while the two bidentate ketopyrrolyl ligands occupy the remaining

Table 1 Selected bond distances and angles for compounds 2a and 2b

2a 2b

Distances (A)Co(1)–P(1) 2.2513(5) 2.2528(9)Co(1)–O(1) 2.3854(15) 2.320(3)Co(1)–N(1) 1.9231(18) 1.921(3)N(1)–C(4) 1.385(3) 1.381(4)N(1)–C(1) 1.342(3) 1.344(4)C(2)–C(1) 1.406(3) 1.394(5)C(3)–C(2) 1.381(3) 1.385(6)C(4)–C(3) 1.408(3) 1.406(5)C(4)–C(5) 1.407(3) 1.422(4)C(5)–C(6) — 1.502(5)C(5)–O(1) 1.251(3) 1.253(4)

Angles (◦)P(1)–Co(1)–O(1) 88.58(4) 88.45(7)P(1)–Co(1)–N(1) 91.77(5) 91.57(9)N(1)–Co(1)–O(1) 78.23(6) 77.95(9)Co(1)–O(1)–C(5) 103.78(12) 107.7(2)Co(1)–N(1)–C(4) 117.04(14) 117.2(2)Co(1)–N(1)–C(1) 136.74(15) 136.4(2)N(1)–C(4)–C(3) 109.90(18) 109.3(3)C(4)–C(3)–C(2) 106.11(19) 106.7(3)C(3)–C(2)–C(1) 106.96(19) 106.4(3)C(2)–C(1)–N(1) 110.81(19) 111.2(3)C(1)–N(1)–C(4) 106.21(17) 106.4(3)C(5)–C(4)–C(3) 132.6(2) 133.1(3)C(5)–C(4)–N(1) 117.45(18) 117.6(3)C(4)–C(5)–O(1) 123.44(19) 119.6(3)C(4)–C(5)–C(6) — 120.5(3)C(6)–C(5)–O(1) — 119.8(3)

This journal is © The Royal Society of Chemistry 2007 Dalton Trans., 2007, 5460–5470 | 5461

Fig. 1 ORTEP III diagrams of complexes (a) 2a and (b) 2b using 50% probability level ellipsoids. Hydrogens are omitted for clarity.

coordination positions in a transoid conformation in relation toeach other, likely reflecting a better stereochemical arrangement.

All the bond distances and angles are similar for both complexes,except those involving the oxygen or the methyl atoms. The mostprominent difference (D = ∼0.065 A) is observed for the bonddistance Co1–O1 (2.3854(15) for 2a, and 2.320(3) A for 2b), whichcould be taken as the result of a better donor capacity of theacetylpyrrolyl oxygen due to a higher electron induction of themethyl substituent in relation to that of the hydrogen. However,this conclusion was not confirmed by DFT calculations, whichsuggested that stereochemical reasons account for the differencein the Co1–O1 bond distances of the two complexes (see discussionbelow). In addition, in the only structure found in the literaturereporting a Co complex supported by a bidentate acetylpyrrolylligand, i.e. the three-legged piano stool [Co(g5-C5Me5)(j2N,O-2-NC4H3C(Me)=O){P(=O)(OMe)2}],6 a Co–O distance of1.970(2) A was observed, which is significantly lower than thatof 2b, while the corresponding Co–N distances are similar (ca.1.92 A). These relatively large differences (0.35–0.41 A) observedbetween the Co–O bond distance of the CoCp* species and thoseof 2b are certainly related to the fact that the first is an 18-electronCo(III) complex whereas 2b is a 19-electron Co(II) compound,thus containing one electron in an antibonding orbital (see DFTcalculations). In fact, the Co1–O1 distances found in 2a and 2b arethe longest ever reported for metal-ketopyrrolyl complexes (1.906–2.212 A, for metals such as Ni(II),5 Cu(II),4 V(II),8 Ru(II),7 all ofthem being unsaturated species). On the contrary, the Co1–N1distances are among the shortest reported for metal-ketopyrrolylcompounds (1.880–2.155 A, for metals such as Ni(II),5 Cu(II),4

V(II),8 Ru(II)7). These chelating ligands are planar which points toa partial extension of the pyrrolyl ring electronic delocalisation to-wards the keto part of the ligand. The chelate bite angles N1–Co1–O1 in both 2-ketopyrrolyl ligands (78.23(6) for 2a, and 77.95(9)◦

for 2b), are slightly lower than the value found in the literature forthe referred Co(III) 2-acetylpyrrolyl complex,6 but within the rangereported for other bidentate 2-acetylpyrrolyl ligands coordinated

to other metals (76–78◦, for Ru(II)7 and V(II),8 82–84◦, for Cu(II)4

and Ni(II)5). The C5–O1 bond distances (ca. 1.25 A) are the lowestvalues reported for bidentate 2-acetylpyrrolyl metal complexes(1.26–1.28 A, for Cu(II),4 V(II),8 Ni(II),5 Co(III),6 Ru(II)7) and aresimilar to those found, for instance, for delocalised double bondsin organic carboxylate anions.11 The distances Co1–P1 (ca. 2.25 A)are in the range of the values found in literature for other cobaltcomplexes containing PMe3 ligands (2.22(4) A).12

Magnetic measurements

The effective magnetic moments (leff = 2.83(vT)1/2 lB) of com-plexes 2a and 2b, obtained from the magnetic susceptibility (v)measurements, recorded both in powder (Faraday system) andin solution (Evans method), are shown in Table 2. The temper-ature dependence of the effective magnetic moment obtained forpolycrystalline powder samples of complexes 2a and 2b, usingthe Faraday system was found to be almost constant from 25 to300 K (Fig. 2). The decrease of the magnetic moment at lowertemperatures (T < 20 K) is a consequence of the high fieldemployed in the measurements (5 T). At room temperature, leff

values for both complexes are very similar, and are very close to theexpected spin-only value for low-spin S = 1/2 complexes (1.73 lB),in agreement with the EPR experiments and DFT calculationspresented below. In solution measurements, at 300 K, the valuesof leff found for complexes 2a and 2b are slightly above thoseobtained in the solid state, but still lying in the range typical oflow-spin S = 1/2 Co(II) complexes (2.1–2.9 lB

13).

Table 2 Effective magnetic moments leff (lB) for complexes 2a and2b, measured in powder (Faraday system) and toluene solution (Evansmethod), at 300 K

Complex Faraday system leff/lB Evans method leff/lB

2a 1.88 2.432b 1.83 2.21


Fig. 2 Temperature dependence of the effective magnetic moment, leff,of complexes 2a and 2b, in the range 2–300 K (Faraday system).

EPR spectra

The experimental EPR spectra of complexes 2a and 2b obtained at77 K, and their corresponding simulations are depicted in Fig. 3.Both complexes show almost identical EPR spectra, indicatingthat the two different substituents R = H and CH3 in the bidentateligand do not have a significant influence in the electronic structureof the cobalt centre. The spectra of complexes 2a and 2b are ofaxial type with large tensor anisotropy and gav = 2.17 and 2.16,respectively ([gav = (2g1,2 + g3)/3] where g1,2 and g3 refer to thelowest and highest magnetic field g values), are typical of low-spincobalt(II) complexes (d7, S = 1/2).14–19 The spectra also exhibit, inthe higher magnetic field region, well-resolved hyperfine splittingdue to cobalt nuclei (ICo = 7/2) and superhyperfine splitting dueto two equivalent phosphorus nuclei (IP = 1/2; relative intensity1 : 2 : 1).

Fig. 3 Experimental (Exp) EPR spectra of complexes 2a and 2b obtainedin frozen toluene at 77 K and their corresponding simulated spectra (Sim).

Table 3 EPR parameters for cobalt(II) complexes

Complex g⊥ g‖ |A|Co⊥ |A|Co

‖ |a|P⊥ |a|P

‖

2a 2.25 2.01 4 96 3 102b 2.24 2.01 6 90 3 10

In the absence of EPR crystal data for complexes 2a and 2b,the observed similarity between their g features and those ofhexacoordinate cobalt(II) complexes14–19 can be further extendedto support the following orientation scheme for the tensor axis inthese cobalt complexes: g1,2 = g⊥ and g3 = g‖, ACo

1,2 = ACo⊥ and

ACo3 = ACo

‖ and aP1,2 = aP

⊥ and aP3 = aP

‖. The EPR parametersobtained from the simulated spectra are summarised in Table 3.The g and ACo patterns observed in the spectra of both complexes,namely g⊥ < g‖ and ACo

⊥ < ACo‖ and their values are typical

of hexacoordinate cobalt(II) complexes with an axial elongationalong the z axis and an 2A1(dz2 ) ground state,14–19 confirming theN2O2P2 coordination sphere for the cobalt centre. Analysis of thesuperhyperfine splittings in the parallel region, due to the electronspin interaction with the nuclear spin of the two phosphorus nucleibelonging to the coordinated phosphines, can give further insightsinto the P(CH3)3 groups location within the cobalt coordinationsphere. There are two possible positions: in the equatorial planeor in the axial axis (z axis). If the phosphines were in the axialpositions there would be a direct interaction between the cobaltunpaired spin (dz2 ground state) and the nuclear spin of the twoequivalent phosphorus nuclei, which would lead to aP

‖ values inthe range of 100–270 Gauss.14 On the other hand, if the PMe3

groups were in the equatorial plane (xy plane), there would be anindirect interaction between the cobalt unpaired spin (dz2 groundstate) and the nuclear spin of the two equivalent phosphorus atomsin the xy plane and, consequently, a large decrease in the aP

‖ valuesis expected to occur when compared with the direct interaction.The aP

‖ values obtained for complexes 2a and 2b are 10 Gauss,clearly not falling in the range typical of direct interaction betweenmetal unpaired spin and phosphorus atoms, thus suggesting thephosphines are coordinated to cobalt in the equatorial plane andnot axially. Furthermore, as no other superhyperfine splitting wasdetected in the parallel region, in particular no splitting fromnitrogen atoms was observed (IN = 1 and typical values of aN

‖for direct interaction are in the range 13–25 Gauss14–19), we canconclude that the elongated axial positions (z axis) are occupiedby the oxygen atoms (I = 0 for the isotope with higher naturalabundance) from the bidentate ligand, leaving the nitrogen atomsalso in the equatorial plane. Thus, the EPR data is compatiblewith the X-ray data for both complexes since it suggests thatlonger distances (z axis) in the hexacoordinate Co coordinationsphere are occupied by the O atoms from the bidentate ligand,leaving the nitrogen and phosphorus within the equatorial planeregion. Further insights into the different available positions ofthe coordinating atoms within the cobalt coordination sphere willbe analysed by DFT calculations (see below).

Electronic spectra

The spectra of the ligand precursors formyl- and acetylpyrrole andof complexes 2a and 2b were recorded in the range 200–1500 nm,in toluene solutions, under nitrogen atmosphere. The values


Table 4 Electronic bands (kmax) and corresponding absorption coefficients (e) for ketopyrrole ligand precursors and their Co(II) complexes 2a and 2b intoluene

Compound kmax/nm (e/dm3 mol−1 cm−1)

Formylpyrrole 287 (6710), 296 (5870)2a 287 (47 300), 290 (45 500), 322 (18 000), 350(10 500), 432 (260), 602 (50)Acetylpyrrole 287 (12 100), 305 (1060)2b 284 (41 500), 290 (35 300), 320 (14 600), 350 (8200), 604 (20)

of electronic bands (kmax) and their corresponding absorptioncoefficients (e) are summarised in Table 4. The ligand precursorsshow several electronic bands in the range 270–350 nm thatcan be assigned to p←n and p*←p transitions. The electronicspectra of both complexes are quite similar, which parallelsthe resemblance of their EPR spectra, showing very intensebands (e > 35 000 dm3 mol−1 cm−1) in the range 270–300 nm,medium intensity bands in the interval 300–400 nm (20 000 <

e < 8000 dm3 mol−1 cm−1) and much less intense bands for k >

400 nm (e < 500 dm3 mol−1 cm−1). The comparison between theligand precursor spectra and those of the corresponding Co(II)complexes enabled the tentative assignment of the bands observedin the range 280–300 nm to intra-ligand transitions. Furthermore,the bands in the range 300–400 nm can be assigned to metal →ligand charge transfer transitions (MLCT) by considering theirextinction coefficient values, which are significantly lower thanthose attributed to intra-ligand transitions, and the low thirdionization energy of cobalt, combined with the likely presenceof low-energy ligand p* orbitals. However, their e values aresomewhat higher than expected, which can be attributed to theproximity and consequent superimposition of the highly intenseintra-ligand bands at k < 300 nm. Finally, the bands with k >

400 nm can be attributed to spin allowed d–d transitions, on thebasis of their extinction coefficient values.20

The assignment of the d–d bands to the corresponding d–dtransitions is not straightforward since no quantitative informa-tion is known about the magnitude of the complex axial distortionrelative to octahedral geometry (in this case the ground state is2E and 2T1, 2T2 are the excited states20,21). The electronic spectraof axial elongated Co(II) low-spin six coordinate complexes havebeen interpreted using the same models used for five- (squarepyramidal) and four- (square planar) coordinated complexes inparallel with the analysis of EPR data.14,17–20,22–26 However, theband assignment is not simple due not only to the several possibleenergy sequences of the d orbitals, but also due to the high numberof possible transitions for each energy sequence.

Nonetheless, the similarity between EPR data of complexes 2aand 2b with those of adducts of Co(II) complexes containing salentype ligands14,17–20,22–26 enable the use of such models to interpretthe electronic spectra of our complexes. In this context, the mostlikely energy sequence of the d orbital for complexes 2a and 2bis dx2−y2 (b1) > dz2 (a1) > dyz, dxz (e) > dxy (b2). The key point inthe interpretation of the electronic spectra is the position of thedz2 orbital in the referred energy sequence. In fact, its energy ishighly dependent on the axial ligand field, being closer to thatof dx2−y2 , for strong axial fields (small distortion from octahedralgeometry) and nearer to those of dyz, dxz orbitals, for weak axialligand fields (typically square planar or highly tetragonal distortedcomplexes), leading to different electronic band assignments. In

the former case, the lowest energy transition in the electronicspectrum may correspond to 2B1←2A1 and the 2E ←2A1 and2B2←2A1 to the subsequent transitions; in the latter case thelowest energy transitions may be attributed to 2E ← 2A1 and2B2←2A1.17–20,22–26 Taking into account that, in complexes 2a and2b, the axial oxygen atoms belong to the bidentate ligands thatcontain the other atom (nitrogen) coordinated in the equatorialplane, which induce somehow a considerable axial ligand field, wepropose that the electronic band at k ≈ 600 nm may correspondto the 2B1←2A1 transition and the one at k ≈ 430 nm (only seenin complex 2a) to the 2E←2A1 transition. The other allowed d–dbands may occur at lower k values and, consequently, will bemasked by the more intense metal → ligand charge transferbands.

DFT calculations

The electronic structure of Co(II) complexes of the type [Co(2-ketopyrrolyl)2(PMe3)2] was investigated by DFT calculations27 inorder to understand the spin state and coordination geometrypreferences in those species. All possible coordination isomers(Chart 2) were optimised for each ketopyrrolyl ligand, 2-formyl-pyrrolyl (R = H) and 2-acetylpyrrolyl (R = Me), and for two spinstates, that is, low spin molecules, S = 1/2, and high spin speciesS = 3/2. The atomic coordinates of all optimised complexes(in a total of twenty: 5 coordination and 2 spin isomers for 2ketopyrrolyl ligands) are presented as ESI¶, as well as figures withthe representations of the calculated structures (Fig. S1 and S2¶).

The coordination isomers considered are represented schemat-ically in Chart 2. The first two, I and II, have a trans arrangementof the two trimethylphospine ligands. While in I there is atransoid conformation of the ketopyrrolyl ligands, in II the relativeconformation of these ligands is cisoid. In the remaining threeisomers (III–V) the two phosphine ligands occupy cis coordinationpositions, the differences being dictated by the relative orientationof the two bidentate ketopyrrolyl ligands. In III one nitrogen atomfaces an oxygen atom of the opposite ligand, in IV the two nitrogenatoms are trans to each other, and in V the two oxygen atomsoccupy opposite positions.

In all cases, that is, for each spin state (S = 1/2 and S = 3/2)and for each ketopyrrolyl ligand (2-formylpyrrolyl, R = H; and2-acetylpyrrolyl, R = Me), the most stable coordination isomercorresponds to I (by 1–10 kcal mol−1). This is the geometry foundin the X-ray structure of complexes 2a and 2b, and its stabilityresults mostly from the best stereochemical arrangement aroundthe metal coordination sphere, with the bulkier phosphines onopposite sides of the molecule.

The geometry optimised for complexes 2a and 2b can becompared with the corresponding X-ray structures in order to


Chart 2 Coordination isomers of the [Co(2-ketopyrrolyl)2(PMe3)2]complexes.

test the performance of the theoretical method employed (seeComputational details). The mean (d) and maximum (D) absolute

deviations between the experimental Co–X distances and theones calculated for the corresponding low spin complexes ared = 0.07 and D = 0.15 A for 2a, and d = 0.06 and D =0.15 A for 2b. These values indicate a reasonable performanceof the theoretical method in the structural description of thesystem studied, but it should be noticed that the larger deviationsobserved (D) correspond, for both 2a and 2b, to the Co–Pdistances. In fact, a better agreement between experimental andcalculated Co–P bond lengths would require a bigger basis set inthe geometry optimisations. However, the number of optimisedstructures and our computational limitations preclude the use ofan improved basis set. The agreement observed between calculatedand experimental structures is sufficient for the semi-quantitativediscussion intended here, but the conclusions should be taken withsome caution.

Once the preferred coordination geometry for the complexeswith the two ketopyrrolyl ligands was established, the relativestability of the two spin states was addressed. The optimisedstructures obtained for complexes 2a and 2b provide a firstindication. In fact, the agreement between the geometry calculatedfor the low spin isomers of complexes 2a and 2b and theexperimental structures (see above) is much better than that foundfor the high spin molecules: d= 0.25 (2a) and 0.24 A (2b), and D =0.41 (2a) and 0.42 A (2b). The optimised geometries obtained forthe two spin isomers of complexes 2a and 2b are represented inFig. 4, with the Co–X distances.

The low spin species is clearly more stable for both 2a (7.7 kcalmol−1) and 2b (4.7 kcal mol−1), in good agreement with theexperimental magnetic moments (see above). One major differencebetween spin isomers of the two complexes is the coordinationdistances of the ligands. There is an enlargement of the Co–P

Fig. 4 Optimised geometries (B3LYP/b1) of the high (top) and the low spin (bottom) molecules of complexes 2a (left) and 2b (right). The calculatedand the experimental (italics) Co–X distances (A) are presented, as well as the relative energy (B3LYP*/b2//B3LYP/b1).


and the Co–N bond distances going from the low spin to thehigh spin species, more marked in the case of Co–P (0.27 A) thanin the case of Co–N (0.13 A). There is an electronic reason forthose differences that results from the occupancy of the frontierorbitals of the complexes, as illustrated by the highest energysingle occupied molecular orbital (SOMO) of each species. TheSOMO of the two spin isomers of complex 2a are representedin Fig. 5.

Fig. 5 Highest energy single occupied molecular orbital of the low (top)and the high spin (bottom) of complex 2a.

In the case of the low spin molecule, the SOMO corresponds to aCo–O antibonding orbital (r*) resulting from Co z2 and the oxygenlone pairs, while for the high spin molecule the correspondingorbital is Co–P and Co–N antibonding (r*) involving Co x2 −y2 and no oxygen participation.28 Thus, the occupancy of anantibonding orbital in the S = 1/2 molecules justifies the large Co–O distances observed in both the calculated and the experimentalstructures of the complexes (see above). Moreover, moving fromlow to high spin corresponds to the addition of an electron to aCo–N and Co–P antibonding orbital and, consequently, to weakenthese bonds. The electronic reason behind the coordinationdifferences observed in the spin isomers of 2a is further supportedby the Wiberg indices (WI)29 of the Co–X bonds. The Co–P areconsiderably weaker in the high spin molecules (WI = 0.193) thanin the low spin species (WI = 0.385), the same happening withthe Co–N bonds (WI = 0.192 and 0.363, respectively). Equivalent

conclusions can be drawn for complex 2b (see ESI¶ for the relevantorbitals).

Comparison between the optimised geometries of the two lowspin complexes, that is, the ones corresponding to the X-raystructures, is also of interest as it could provide a means todistinguish the coordination ability of the two ketopyrrolyl ligands.The most striking structural difference between the two species isthe Co–O distance, which is considerably longer in 2a, by 0.034 A(calc.) or 0.065 A (exp.).

One would expect that the shorter bond length observed in2b would be the result of a stronger donor ability of the 2-acetylpyrrolyl ligand, due to presence of the methyl group (R =Me), when compared to 2-formylpyrrolyl in 2a (R = H). However,the Wiberg indices for Co–O in the two complexes are within 0.001,showing that, despite the differences in bond length, these bondsare equally strong in both complexes, from the electronic point ofview. The same happens with the Co–N bonds of both complexes,with equal Wiberg indices (0.363). These conclusions are furtherconfirmed by the charge distribution on both complexes, obtainedby means of a Natural Population Analysis (NPA).30 Metal chargesof 0.738 (2a) and 0.757 (2b) indicate a more positive Co atom in 2bshowing that, in this complex, the metal is receiving less electrondensity than in 2a. In addition, the charge of the ketopyrrolylligands is approximately the same in both species: −0.682 for2-formylpyrrolyl in 2a, and −0.684 for 2-acetylpyrrolyl in 2b.These results point out two major conclusions. Firstly, there isan electron poorer metal centre in 2b, when compared with 2a,in obvious contradiction with the presence of a stronger electrondonor: 2-acetylpyrrolyl in 2b vs. 2-formylpyrrolyl in 2a. Secondly,the difference in the Co electron density calculated for the twocomplexes is not due to the ketopyrrolyl ligands, since the chargeof these ligands is similar in both species, and, thus, must beoriginated by the phosphines. In fact, the trimethylphosphineligands are more positive in 2a than in 2b, the calculated chargesbeing 0.313 and 0.306, respectively. This indicates that electrondonation from the phosphine to the metal is more efficient in2a than in 2b. The Co–P Wiberg indices further confirm thisconclusion, indicating that these bonds are slightly stronger in2a (WI = 0.388) than in 2b (WI = 0.385).

The above results suggest that the differences in the coordinationgeometry observed in the two complexes are not due mainly toelectronic factors, such as better donor ability of 2-acetylpyrrolylwith respect to 2-formylpyrrolyl, and, thus, must be caused bystereochemical reasons. In other words, interligand repulsion isprobably determining the best overall arrangement around themetal coordination sphere. In addition, it is perhaps not toosurprising that the Co–O bonds are the ones that differ mostwhen both complexes are compared, since these are by far theweakest of all Co–X bonds (WI = 0.096 in 2a and 0.097 in2b) and, thus, they are easier to adjust upon the effect of stericconstraints.

Interestingly, the spin density calculated for 2a (represented inFig. 6) is located along the O–Co–O axis, being centred in themetal and having smaller contributions on the oxygen atoms. Thisis in excellent agreement with the EPR studies (see above), namelythe weak coupling with the P atoms and the presence of the oxygenatoms along the main axis of the metal z2 orbital. The spin densitycalculated for 2b is equivalent (see Fig. S5, ESI¶) to that obtainedfor 2a.


Fig. 6 Calculated spin density of complex 2a.

Conclusions

The first 2-ketopyrrolyl complexes of Co(II), [Co(j2N,O-2-NC4H3C(R)=O)2(PMe3)2] (R = H 2a, and Me 2b), were prepared.As shown by X-ray diffraction analysis, these 19-electron com-pounds show an octahedral geometry with the PMe3 ligandsarranged trans to each other, while the two bidentate ketopy-rrolyl ligands occupy the remaining coordination positions ina transoid conformation. Magnetic susceptibility measurementsas well as EPR and UV/Vis/NIR spectroscopies show thatin these compounds the Co(II) is in a low-spin ground state(dz2 , S = 1/2). Analysis of the EPR superhyperfine couplingssuggest that longer distances (z axis) in the hexacoordinate Cocoordination sphere are occupied by the keto O atoms fromthe bidentate ligand, leaving the pyrrolyls N and the phosphinesP atoms within the equatorial plane region. This is confirmedby DFT calculations, which also indicate that the molecularstructures obtained by X-ray crystallography correspond to thelow-spin (S = 1/2) molecules and to the most stable coordinationisomers.

Experimental

General

All experiments dealing with air and/or moisture sensitive ma-terials were carried out under inert atmosphere using a dual vac-uum/nitrogen line and standard Schlenk techniques. Nitrogen gaswas supplied in cylinders by Air Liquide and purified by passagethrough 4 A molecular sieves. Unless otherwise stated, all reagentswere purchased from commercial suppliers (e.g. Acros, Aldrich,Fluka) and used without further purification. All solvents wereused under inert atmosphere and thoroughly deoxygenated, driedand purified by refluxing over a suitable drying agent followed bydistillation under nitrogen. The following drying agents were used:sodium (for toluene, diethyl ether and tetrahydrofuran), calciumhydride (for hexane). Deuterated solvents were dried by storageover 4 A molecular sieves and degassed by the freeze–pump–thawmethod. Solvents and solutions were transferred using a positivepressure of nitrogen through stainless steel cannulae and mixtureswere filtered in a similar way using modified cannulae that couldbe fitted with glass fibre filter disks.

The ligand precursors, 2-formylpyrrole and 2-acetylpyrrole,were synthesised using a method similar to the described in theliterature.31 The synthesis of CoCl2(PMe3)2

32 and CoCl(PMe3)33

were adapted from the literature.Nuclear magnetic resonance (NMR) spectra were recorded on

a Varian Unity 300 MHz spectrometer at 299.995 MHz (1H)and 75.4296 MHz (13C), and were referenced internally usingthe residual protio solvent resonance relative to tetramethylsilane(d = 0). For air and/or moisture stable compounds, samples weredissolved in CDCl3 and prepared in common NMR tubes. Forair and/or moisture sensitive materials, samples were prepared ina glovebox, in J. Young NMR tubes, using toluene-d6 (magneticsusceptibilities measurements by the Evans method). Elementalanalyses were obtained from the IST or ITN elemental analysisservices.

EPR spectra were obtained using an X-band Bruker ESP300Espectrometer equipped with a microwave bridge ER033, a rectan-gular cavity operating in the T102 mode and a NMR gaussmeterER = 35 M. The modulation amplitude was kept well below theline width and the microwave power well below saturation. Afinger-quartz cryostat with liquid nitrogen was used for acquiringthe spectra at 77 K. Samples were prepared in a glovebox, bydissolving the complexes in freshly distilled toluene and filteringthe solutions into EPR tubes. The reported EPR parameters wereobtained by simulation using the program Win EPR Simfonia(Bruker) assuming axial spin Hamiltonians. The parameters inperpendicular region (g⊥, ACo

⊥ and aP⊥) are less accurate because

of their dependence on the line widths used in the simulation andthe low spectral resolution in this region.

Electronic spectra were recorded in a Perkin Elmer Lambda 9UV/Vis/NIR spectrophotometer, between 200 and 1500 nm, atroom temperature. Solution samples were prepared in a glovebox,by dissolving the complexes (or the ligand precursors) in freshlydistilled toluene, to give rigorous solutions (in most cases 10−3 and10−5 mol dm−3), and the corresponding spectra were obtained inquartz cells with optical path of 1 cm, under inert atmosphere;the electronic spectra of the complexes were also obtained in nujolmulls, which were deposited on the inner surfaces of quartz cells,under inert atmosphere.

Syntheses

In situ preparation of 2-ketopyrrolyl sodium salts (1a, 1b). In atypical experiment, NaH (2 mmol) was suspended in tetrahydro-furan (THF) and 2 mmol of 2-ketopyrrole (2-formylpyrrole or2-acetylpyrrole) was slowly added as a solid, under a counterflowof nitrogen. An immediate evolution of hydrogen occurred and,after some minutes, a solution of the sodium 2-formylpyrrolide(1a), or sodium 2-acetylpyrrolide (1b), was obtained and stirredfor 90 min.

Attempted preparation of [Co(j2N ,O-2-NC4H3C(R)=O)2] (R =H, Me). Anhydrous CoCl2 (1 mmol) was suspended in THFand cooled to −80 ◦C. The corresponding solution of ligandsodium salt (2 mmol) in THF was filtered and directly addeddropwise to the CoCl2 suspension, and stirred for one hour. Themixture was allowed to warm to room temperature and stirredovernight. The volatiles were evaporated in vacuum to dryness andthe residue washed with n-hexane, and extracted with diethyl ether.The solution (brown for R = H, and deep orange for R = Me)


was concentrated and cooled to −20 ◦C to yield brown (R = H) ororange–brown powders. The characterisation of these powdersby elemental analysis, EPR and NMR did not lead to clearformulations.

Preparation of complex [Co(j2N ,O-2-NC4H3C(H)=O)2-(PMe3)2] (2a). Method a) Anhydrous CoCl2 (1 mmol) wassuspended in THF and 2 ml (2 mmol) of a P(CH3)3 solution (1 Min toluene) was added. The colour of the solution changed fromdeep turquoise blue to dark purple, and the mixture was furtherstirred for 90 min. After cooling to −80 ◦C, a solution of theligand Na salt (2 mmol) in THF was filtered and added dropwise.The resulting reaction mixture was allowed to warm to roomtemperature and stirred for ca. 12 h. The volatiles were evaporatedin vacuum to dryness and the residue extracted with n-hexaneuntil extracts were colourless. The resulting dark brown solutionwas concentrated and cooled to −20 ◦C to yield 0.339 g (85%)of dark brown crystals. Crystals suitable for X-ray diffractionwere obtained. Method b) CoCl2(PMe3)2 (251 mg, 1 mmol) wasdissolved in THF and cooled to −80 ◦C. The solution of ligandNa salt in THF was then filtered dropwise to this dark greensolution, and the resulting mixture was allowed to warm to roomtemperature and stirred for ca. 12 h. The volatiles were evaporatedin vacuum to dryness and the residue was extracted with n-hexaneuntil extracts were colourless. The resulting brown solution wasconcentrated and cooled to −20 ◦C to yield 0.333 g (80%) ofdark brown crystals. Found: C 47.87, H 6.44, N 6.96. Calc. forC16H26CoN2O2P2: C 48.11, H 6.57, N 7.02%.

Preparation of complex [Co(j2N ,O-2-NC4H3C(Me)=O)2-(PMe3)2] (2b). Methods a) and b) The same procedures describedfor 2a were followed. Yields of 0.379 g (85%) and 0.365 g (82%)of deep brown crystals were obtained, respectively, for methods a)and b). The reaction was also carried out using a reaction timeof 2 h. Yields of 0.350 g (40%) and 0.296 g (69%) were obtained,respectively, for methods a) and b).

Method c) CoCl(PMe3)3 (323 mg, 1 mmol) was dissolved inTHF and cooled to −80 ◦C. An excess of ligand Na salt (3 mmol),in THF solution, was then filtered dropwise to this dark greensolution, and the resulting solution was allowed to warm to roomtemperature and stirred for ca. 2 h. The volatiles were evaporatedin vacuum to dryness and the residue was extracted with n-hexane until extracts were colourless. The resulting reddish brownsolution was concentrated and cooled to −20 ◦C to yield 0.363 g(85%) of dark brown crystals of 2b. Crystals suitable for X-raydiffraction were obtained. Found: C 50.40, H 7.05, N 6.66. Calc.for C18H30CoN2O2P2: C 50.59, H 7.08, N 6.56%.

X-Ray experimental data§

Crystallographic and experimental details of crystal structuredeterminations are given in Table 5. Crystals of 2a and 2b wereisolated by filtration, and in each case, a specimen crystal selectedunder an inert atmosphere, covered with polyfluoroether, andmounted on the end of a nylon loop. A synchrotron radiationsource was used to collect diffraction data for 2a, at 120 K. Datawas collected at Station 9.8, Daresbury SRS, UK, using aBruker SMART CCD diffractometer. Data for compound 2bwere collected at 180 K on a Nonius KappaCCD with graphitemonochromated Mo-Ka radiation (k = 0.71073 A). The images

Table 5 Crystal data and structure refinement for compounds 2a and 2b

Complex 2a 2b

Formula C16H26CoN2O2P2 C18H30CoN2O2P2

M 399.28 427.32k/A 0.6923 0.71073T/K 120 180Crystal system Monoclinic MonoclinicSpace group P21/n P21/na/A 8.9836(10) 8.0704(5)b/A 11.3840(12) 10.0296(5)c/A 9.3690(10) 13.2877(8)a/◦ 90 90b/◦ 94.4080(10) 92.070(3)c /◦ 90 90V/A3 955.33(18) 1074.84(11)Z 2 2qcalc/g cm−3 1.388 1.221l/mm−1 1.075 0.948hmax/

◦ 30 26Total data 7238 5493Unique data 1988 2102Rint 0.059 0.070R [I > 3r(I)] 0.0323 0.0350Reflections (R [I > 3r(I)]) 1612 1305wR 0.0389 0.0382Goodness of fit 1.0896 1.1132q min, q max −0.81 −0.26

0.82 0.28

were processed with the DENZO and SCALEPACK programs.34

The structures were solved by direct methods using the programSIR92.35 The refinement (on F) and graphical calculations wereperformed for both 2a and 2b using the CRYSTALS programsuite.36 For both compounds the non-hydrogen atoms were refinedwith anisotropic displacement parameters. Hydrogen atoms werelocated in Fourier maps and their positions adjusted geometrically(after each cycle of refinement) with isotropic thermal parameters.Chebychev weighting schemes and empirical absorption correc-tions were applied in each case.37 Figures were generated usingORTEP3.38

CCDC reference numbers 642094 and 642095.For crystallographic data in CIF or other electronic format see

DOI: 10.1039/b710162g

Magnetic measurements

Magnetic susceptibility measurements in solution were carried outby the Evans method39,40 using a 3% solution of hexamethyldis-iloxane (reference) in toluene-d6.

Magnetic measurements of polycrystalline samples were per-formed using a longitudinal Faraday system (Oxford Instruments)with a 7 T superconducting magnet for static measurements in therange 2–300 K. The temperature dependence of the magnetisationwas measured at 5 T. The polycrystalline samples (10–15 mg) wereplaced inside a previously calibrated thin wall Teflon bucket. Theforce was measured with a microbalance (Sartorius S3D-V). Thediamagnetism correction for the experimental data was estimatedusing the Pascal constants.41

Computational details

All calculations were performed with the Gaussian 98 softwarepackage,42 using the B3LYP hybrid functional for geometry


optimisations. That functional includes a mixture of Hartree–Fock43 exchange with DFT27 exchange–correlation, given byBecke’s three parameter functional44 with the Lee, Yang andParr correlation functional, which includes both local and non-local terms.45,46 Geometry optimisations were performed for fivedifferent coordination and two spin isomers of complexes [Co(2-ketopyrrolyl)2(PMe3)2] containing either 2-formylpyrrolyl or 2-acetylpyrrolyl ligands (Chart 2). These calculations were per-formed without symmetry constraints, using a standard LanL2DZbasis set47 for all atoms (b1). Spin contamination was carefullymonitored for all calculations and the values of <S2> indicateminor spin contamination.

The relative energy of spin isomers was obtained through singlepoint energy calculations using the B3LYP/b1 geometries and abasis set (b2) that consisted of Stuttgart/Dresden Effective CorePotentials with valence triple zeta (SDD)48 augmented with an f -polarisation function49 for Co, and a standard 6–311G(d,p)50 forthe remaining elements. Given the known dependence of relativespin-state energies of transition metal complexes on the amountof exact exchange included in the functional,51 energy calculationswere performed with three functionals including different amountsof exact exchange (Hartree–Fock): B3LYP that includes 20%,a modified B3LYP functional (B3LYP*) with only 15%, andBP86, a “pure” DFT functional (without Hartree–Fock exchange)that combines Becke’s 1988 exchange functional with Perdew1981 local correlation functional and gradient corrections.52 Highspin species are increasingly favoured with the amount of theexact exchange included in the functional, as expected.51 BP86grossly overestimates the stability of the S = 1/2 molecules, whilestandard B3LYP yields low energy differences. For complex 2a,the high spin isomer is only 4.5 kcal mol−1 less stable than thelow spin complex, at the B3LYP/b2//B3LYP/b1 level, while for2b, at this same level, the order is reversed and the high spincomplex becomes 1.0 kcal mol−1 more stable than its low spinisomer. The modified B3LYP functional (B3LYP*) provides thebest agreement with experimental results, as it is often the case forfirst row transition metal complexes,51b,51c and, thus, these values(B3LYP*/b2//B3LYP/b1) are presented in Fig. 4 and discussedin the text.

A natural population analysis (NPA)30 was performed with theB3LYP*/b2//B3LYP/b1 density in order to evaluate the chargedistribution on the complexes, and the Wiberg indexes29 obtainedare used as a measure of bond strength. Orbital and spin densityrepresentations were obtained using the program MOLEKEL4.0.53

Acknowledgements

We wish to thank Fundacao para a Ciencia e Tecnolo-gia for financial support (Project POCI/QUI/59025/2004, co-financed by FEDER) and for a fellowship to S. A. C.(SFRH/BPD/14902/2004). We also thank the Royal Society forfunding to S. I. P.

References

1 R. H. Holm, A. Chakravorty and L. J. Theriot, Inorg. Chem., 1966, 5,625–635 and references cited therein.

2 K. Mashima and H. Tsurugi, J. Organomet. Chem., 2005, 690, 4414–4423.

3 S. A. Carabineiro, L. C. Silva, P. T. Gomes, L. C. J. Pereira, L. F. Veiros,S. I. Pascu, M. T. Duarte, S. Namorado and R. T. Henriques, Inorg.Chem., 2007, 46, 6880–6890.

4 For example: (a) C. L. Perry and J. H. Weber, J. Inorg. Nucl. Chem.,1971, 33, 1031–1039; (b) W. J. Birdsall, B. A. Weber and M. Parvez,Inorg. Chim. Acta, 1992, 196, 213–220.

5 For example:W. J. Birdsall, D. P. Long, S. P. E. Smith, M. E. Kastner,K. Tang and C. Kirk, Polyhedron, 1994, 13, 2055–2060.

6 C. Jablonski, Z. Zhou and J. Bridson, Inorg. Chim. Acta, 1997, 254,315–328.

7 (a) J. D. E. T. Wilton-Ely, P. J. Pogorzelec, S. J. Honarkhah, D. H. Reidand D. A. Tocher, Organometallics, 2005, 24, 2862–2874; (b) R. F. R.Jazzar, M. Varrone, A. D. Burrows, S. A. Macgregor, M. F. Mahon andM. K. Whittlesey, Inorg. Chim. Acta, 2006, 359, 815–820.

8 D. Reardon, J. Guan, S. Gambarotta, G. P. A. Yap and D. R. Wilson,Organometallics, 2002, 21, 4390–4397.

9 R. Paolesse, E. Tassoni, S. Licoccia, M. Paci and T. Boschi, Inorg. Chim.Acta, 1996, 241, 55–60.

10 A. Chakravorty and R. H. Holm, Inorg. Chem., 1964, 3, 1521–1524.11 F. H. Allen, O. Kennard, D. G. Watson, L. Brammer, A. G. Orpen and

R. Taylor, J. Chem. Soc., Perkin Trans. 2, 1987, S1–S19.12 F. H. Allen, Acta Crystallogr., Sect. B, 2002, B58, 380–388.13 N. N. Greenwood and A. Earnshaw, in Chemistry of the Elements,

Pergamon Press Ltd, Oxford, UK, 1984.14 C. Daul, C. E. Schiiffer and A. von Zelewsky, Struct. Bonding, 1979,

36, 130.15 B. R. McGarvey, Can. J. Chem., 1975, 53, 2498–2511.16 V. Malatesta and B. R. McGarvey, Can. J. Chem., 1975, 53, 3791–3800.17 M. A. Hitchman, Inorg. Chem., 1977, 16, 1985–1993.18 N. Nishida and S. Kida, Bull. Chem. Soc. Jpn., 1972, 45, 461.19 N. Nishida, K. Kida and S. Kida, Inorg. Chim. Acta, 1980, 38, 213–

219.20 A. B. P. Lever, in Inorganic Electronic Spectroscopy, Elsevier, New York,

USA, 2nd edn, 1984.21 I. Krivokapic, M. Zerara, M. L. Dakua, A. Vargas, C. Enachescu, C.

Ambrusc, P. Tregenna-Piggott, N. Amstutz, E. Krausz and A. Hauser,Coord. Chem. Rev., 2007, 251, 364–378.

22 B. K. Santra and G. K. Lahiri, J. Chem. Soc., Dalton Trans., 1998,139–145.

23 W. C. Lin, Inorg. Chem., 1976, 15, 1114–1118.24 K. G. Caulton, Inorg. Chem., 1968, 7, 392–394.25 A. M. Tait and D. H. Busch, Inorg. Chem., 1976, 15, 197–203.26 C. J. Hipp and W. A. Baker, Jr., J. Am. Chem. Soc., 1970, 92, 792–798.27 R. G. Parr and W. Yang, Density Functional Theory of Atoms and

Molecules, Oxford University Press, New York, USA, 1989.28 It should be noticed that the second and the third single occupied

molecular orbitals of the S = 3/2 (high spin) complexes (SOMO-1and SOMO-2) are centered in the ketopyrrolyl p system with negligiblemetal participation. All frontier orbitals of complexes 2a and 2b in thetwo spin states are depicted in the ESI¶ (Fig. S3 and S4).

29 (a) K. B. Wiberg, Tetrahedron, 1968, 24, 1083–1096; (b) Wiberg indicesare electronic parameters related to the electron density between atoms.They can be obtained from a Natural Population Analysis and providean indication of the bond strength.

30 (a) J. E. Carpenter and F. Weinhold, J. Mol. Struct. (THEOCHEM),1988, 169, 41–62; (b) J. E. Carpenter, PhD thesis, University ofWisconsin (Madison, WI), 1987; (c) J. P. Foster and F. Weinhold, J. Am.Chem. Soc., 1980, 102, 7211–7218; (d) A. E. Reed and F. Weinhold,J. Chem. Phys., 1983, 78, 4066–4073; (e) A. E. Reed and F. Weinhold,J. Chem. Phys., 1983, 78, 1736; (f) A. E. Reed, R. B. Weinstock andF. Weinhold, J. Chem. Phys., 1985, 83, 735–746; (g) A. E. Reed, L. A.Curtiss and F. Weinhold, Chem. Rev., 1988, 88, 899–926; (h) F. Weinholdand J. E. Carpenter, in The Structure of Small Molecules and Ions,Plenum, New York, USA, 1988, p. 227.

31 D. O. A. Garrido, G. Buldain and B. Frydman, J. Org. Chem., 1984,49, 2619–2622.

32 K. A. Jensen, P. H. Nielsen and C. T. Pedersen, Acta Chem. Scand.,1963, 17, 1115–1125.

33 (a) H.-F. Klein and H. H. Karsch, Chem. Ber., 1975, 108, 944–955;(b) W. A. Herrmann, in Synthetic Methods of Organometallic andInorganic Chemistry, ed. W. A. Herrmann and G. Brauer, ThiemeVerlag, Stuttgart, Germany, 1997, vol. 7, part 1, pp. 70–71.

34 Z. Otwinowski and W. Minor, in Methods in Enzymology, ed.C. N. Carter, Jr. and R. M. Sweet, Academic Press, New York, USA,1996, p. 276.


35 A. Altomare, G. L. Cascarano, C. Giacovazzo and A. Guagliardi,SIR92, J. Appl. Crystallogr., 1993, 26, 343–350.

36 (a) D. J. Watkin, C. K. Prout, J. R. Carruthers and P. W. Betteridge, inCRYSTALS, Oxford, UK, 1996; (b) P. W. Betteridge, J. R. Carruthers,R. I. Cooper, K. Prout and D. J. Watkin, J. Appl. Crystallogr., 2003,36, 1487; (c) D. J. Watkin, C. K. Prout, L. J. Pearce, in CAMERON,Oxford, UK, 1996.

37 N. Walker and D. Stuart, Acta Crystallogr., Sect. A, 1983, A39, 158–166.

38 L. J. Farrugia, ORTEP3 for Windows, J. Appl. Crystallogr., 1997, 30,565–566.

39 D. F. Evans, J. Chem. Soc., 1959, 2003–2005.40 S. K. Sur, J. Magn. Reson., 1989, 82, 169–173.41 R. L. Carlin, in Magnetochemistry, Springer-Verlag, Berlin, Germany,

1986.42 M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria, M. A.

Rob, J. R. Cheeseman, V. G. Zakrzewski, J. A. Montgomery, Jr., R. E.Stratmann, J. C. Burant, S. Dapprich, J. M. Millam, A. D. Daniels,K. N. Kudin, M. C. Strain, O. Farkas, J. Tomasi, V. Barone, M. Cossi, R.Cammi, B. Mennucci, C. Pomelli, C. Adamo, S. Clifford, J. Ochterski,G. A. Petersson, P. Y. Ayala, Q. Cui, K. Morokuma, D. K. Malick,A. D. Rabuck, K. Raghavachari, J. B. Foresman, J. Cioslowski, J. V.Ortiz, A. G. Baboul, B. B. Stefanov, G. Liu, A. Liashenko, P. Piskorz, I.Komaromi, R. Gomperts, R. L. Martin, D. J. Fox, T. Keith, M. A. Al-Laham, C. Y. Peng, A. Nanayakkara, C. Gonzalez, M. Challacombe,P. M. W. Gill, B. Johnson, W. Chen, M. W. Wong, J. L. Andres, C.Gonzalez, M. Head-Gordon, E. S. Replogle and J. A. Pople, Gaussian98, Revision A.7, Gaussian, Inc., Pittsburgh, PA, USA, 1998.

43 W. J. Hehre, L. Radom, P. v. R. Schleyer and J. A. Pople, in Ab InitioMolecular Orbital Theory, John Wiley & Sons, New York, USA, 1986.

44 A. D. Becke, J. Chem. Phys., 1993, 98, 5648–5652.

45 C. Lee, W. Yang and R. G. Parr, Phys. Rev. B, 1988, 37, 785–789.46 B. Miehlich, A. Savin, H. Stoll and H. Preuss, Chem. Phys. Lett., 1989,

157, 200–206.47 (a) T. H. Dunning, Jr. and P. J. Hay, in Modern Theoretical Chemistry,

ed. H. F. Schaefer III, Plenum, New York, USA, 1976, vol. 3, p. 1; (b) P. J.Hay and W. R. Wadt, J. Chem. Phys., 1985, 82, 270–283; (c) W. R. Wadtand P. J. Hay, J. Chem. Phys., 1985, 82, 284–298; (d) P. J. Hay and W. R.Wadt, J. Chem. Phys., 1985, 82, 299–310.

48 (a) U. Haeusermann, M. Dolg, H. Stoll and H. Preuss, Mol. Phys.,1993, 78, 1211–1224; (b) W. Kuechle, M. Dolg, H. Stoll and H. Preuss,J. Chem. Phys., 1994, 100, 7535–7542; (c) T. Leininger, A. Nicklass,H. Stoll, M. Dolg and P. Schwerdtfeger, J. Chem. Phys., 1996, 105,1052–1059.

49 A. W. Ehlers, M. Bohme, S. Dapprich, A. Gobbi, A. Hollwarth, V.Jonas, K. F. Kohler, R. Stegmann, A. Veldkamp and G. Frenking,Chem. Phys. Lett., 1993, 208, 111–114.

50 (a) A. D. McClean and G. S. Chandler, J. Chem. Phys., 1980, 72, 5639–5648; (b) R. Krishnan, J. S. Binkley, R. Seeger and J. A. Pople, J. Chem.Phys., 1980, 72, 650–654; (c) A. J. H. Wachters, J. Chem. Phys., 1970,52, 1033–1036; (d) P. J. Hay, J. Chem. Phys., 1977, 66, 4377–4384; (e) K.Raghavachari and G. W. Trucks, J. Chem. Phys., 1989, 91, 1062–1065;(f) R. C. Binning and L. A. Curtiss, J. Comput. Chem., 1995, 103,6104–6113; (g) M. P. McGrath and L. Radom, J. Chem. Phys., 1991,94, 511–516.

51 (a) J. N. Harvey, Annu. Rep. Prog. Chem., Sect. C, 2006, 102, 203;(b) J. N. Harvey, Struct. Bonding, 2004, 112, 151–184; (c) J. N. Harveyand M. Aschi, Faraday Discuss., 2003, 124, 129–143.

52 (a) A. D. Becke, Phys. Rev. A, 1988, 38, 3098–3100; (b) J. P. Perdew,Phys. Rev. B, 1986, 33, 8822–8824.

53 P. Flukiger, H. P. Luthi, S. Portmann and J. Weber, MOLEKEL, SwissCenter for Scientific Computing, Manno, Switzerland, 2000.


bis(ketopyrrolyl) complexes of co(ii) stabilised by trimethylphosphine ligands

Documents