DOI: 10.1002/chem.201304631

Full Paper

& Iron Complexes

Iron Dicarbonyl Complexes Featuring Bipyridine-Based PNN Pincer Ligands with Short Interpyridine CC Bond Lengths: Innocent or Non-Innocent Ligand? Thomas Zell,[a] Petr Milko,[b] Kathlyn L. Fillman,[c] Yael Diskin-Posner,[b] Tatyana Bendikov,[b] Mark A. Iron,[b] Gregory Leitus,[b] Yehoshoa Ben-David,[a] Michael L. Neidig,*[c] and David Milstein*[a]

Abstract: A series of iron dicarbonyl complexes with bipyridine-based PNN pincer ligands were synthesized and characterized by multinuclear NMR spectroscopy (1H, 13C, 15N, 31P), IR spectroscopy, cyclic voltammetry, 57Fe Mçssbauer spectroscopy, XPS spectroscopy, and single-crystal X-ray diffraction. The complexes with the general formula [(R-PNN)Fe(CO)2] (5: R-PNN = tBu-PNN = 6-[(di-tert-butylphosphino)methyl]-2,2’-bipyridine, 6: R-PNN = iPr-PNN = 6-[(diisopropylphosphino)methyl]-2,2’-bipyridine, and 7: R-PNN = PhPNN = 6-[(diphenylphosphino)methyl]-2,2’-bipyridine) feature differently P-substituted PNN pincer ligands. Complexes 5 and 6 were obtained by reduction of the corresponding dihalide complexes [(R-PNN)Fe(X)2] (1: R = tBu, X = Cl; 2: R = tBu, X = Br; 3: R = iPr, X = Cl; 4: R = iPr, X = Br) in the presence of CO. The analogous Ph-substituted complex 7 was synthesized by a reaction of the free ligand with iron pentacarbonyl. The low-spin complexes 5–7 (S = 0) are diamagnet-

Introduction The properties and the reactivity of transition-metal complexes are influenced by the electronic and steric properties of the surrounding ligands.[1] In most cases, however, ligands have a spectator function in reactions; they maintain their structure

[a] Dr. T. Zell, Y. Ben-David, Prof. D. Milstein Department of Organic Chemistry Weizmann Institute of Science, Rehovot, 76100 (Israel) Fax: (+ 972) 89344142 E-mail: [email protected] [b] Dr. P. Milko, Dr. Y. Diskin-Posner, Dr. T. Bendikov, Dr. M. A. Iron, Dr. G. Leitus Department of Chemical Research Support Weizmann Institute of Science Rehovot, 76100 (Israel) [c] K. L. Fillman, Prof. M. L. Neidig Department of Chemistry University of Rochester Rochester, New York, 14627 (USA) Fax: (+ 1) 585-2760205 E-mail: [email protected] Supporting information for this article is available on the WWW under http://dx.doi.org/10.1002/chem.201304631. Chem. Eur. J. 2014, 20, 4403 – 4413

ic and have distorted trigonal bipyramidal structures in solution, whereas in the solid state the geometries around the iron are best described as distorted square pyramidal. Compared to other structurally characterized complexes with these PNN ligands, shortened interpyridine CC bonds of about 1.43  were measured. A comparison with known examples, theoretically described as metal complexes bearing bipyridine p-radical anions (bpyC), suggests that the complexes can be described as FeI complexes with one electron antiferromagnetically coupled to the ligandbased radical anions. However, computational studies, at the NEVPT2/CASSCF level of theory, reveal that the shortening of the CC bond is a result of extensive p-backbonding of the iron center into the antibonding orbital of the bpy unit. Hence, the description of the complexes as Fe0 complexes with neutral bipyridine units is the favorable one.

and do not actively participate in reactions and the reactions take place at the metal center. Redox-active (or “redox non-innocent”) ligands are of significant relevance in catalytic transformations and have attracted much interest in the recent literature, especially in the development of catalysts based on earth-abundant 3d metals.[2] These ligands have energetically accessible orbitals, allowing them to change their charge states during redox reactions. Therefore they can actively participate in the redox reactions of metal complexes and can serve in catalytic reactions as electron reservoirs or even form reactive ligand-based radicals that actively take part in bond-formation or bond-cleavage steps in the catalytic cycle. Additionally, the substrates of catalytic reactions can also be redox active. The combination of iron with redox non-innocent ligands has led to the discovery of remarkable, highly active, and versatile catalysts.[2i] Among the most prominent iron catalysts featuring redox non-innocent ligands are bis(imino)pyridine and iminopyridine complexes. Pioneering studies in the late 1990s, performed independently by the groups of Brookhart[3] and Gibson,[4] revealed the superior activity of bis(imino)pyridine iron complexes in olefin polymerization reactions. Chirik and

4403

 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper co-workers later showed that these complexes are capable of catalyzing several other reactions, including hydrogenation,[5] hydrosilylation,[5a, 6] hydroboration,[7] and cyclization[8] reactions. Ritter et al. recently established the use of iron iminopyridine complexes as catalysts for the 1,4-hydrosilylation of 1,3dienes,[9] the 1,4-hydroboration of 1,3-dienes,[10] the 1,4-addition of terminal olefins to 1,3-dienes,[11] and the polymerization of 1,3-dienes.[12] Metrical parameters of ligand systems, derived from X-ray crystallography, have been shown to be, in many cases, a useful tool for determining the oxidation state of redoxactive ligands.[2b,j, 13] However, the models of metrical oxidation states have limitations for systems with extensively delocalized bonding situations and in cases in which steric effects compete with the electronically favored structures.[2b] A reliable differentiation between the descriptions of ligands in different oxidation states usually requires a combination of appropriate theory and various experimental methods.[14] Over the past few years our group has studied a different type of non-innocent ligand behavior in bifunctional substrate activation reactions and has applied it in catalytic transformations. These activation processes employ metal–ligand cooperation based on dearomatization/aromatization of pyridine- and bipyridine-based ligands caused by deprotonation/protonation of the benzylic arm of the ligand.[1b, 15] In this context, we applied the dearomatized, bipyridine-based pincer complex [(tBu-PNN*)Ru(H)(CO)] (A, tBu-PNN = 6-[(di-tert-butylphosphino)methyl]-2,2’-bipyridine, Scheme 1, the asterisk denotes a dearomatized pincer ligand) as a catalyst for hydrogenation reactions[16] and for acceptorless dehydrogenation reactions.[17] Furthermore, we have reported the oxidation of alcohols to

Scheme 1. Ruthenium and iron pincer complexes. Chem. Eur. J. 2014, 20, 4403 – 4413

www.chemeurj.org

carboxylic acids and dihydrogen using water as an oxygen source[18] and the selective deuteration reactions of alcohols employing D2O as deuterium source[19] catalyzed by A. There has been remarkable progress in the application of iron-based catalysts for (transfer) hydrogenation and dehydrogenation reactions in the recent literature.[20] We have reported the hydrogenation of carbon dioxide to sodium formate in aqueous NaOH solutions[21] and the formal reverse reaction, the dehydrogenation of formic acid in the presence of trialkylamines.[22] Both reactions are efficiently catalyzed by the pyridine-based PNP iron pincer complex trans[(tBu-PNP)Fe(H)2(CO)] (B, Scheme 1, tBu-PNP = 2,6-bis(di-tertbutylphosphinomethyl)pyridine). Furthermore, we demonstrated that the related complexes [(iPr-PNP)Fe(H)(CO)(Br)][23] (C, Scheme 1, iPr-PNP = 2,6-bis(diisopropylphosphinomethyl)pyridine) and [(iPr-PNP)Fe(H)(CO)(h2-BH4)][24] (D, Scheme 1) serve as catalysts for the hydrogenation of ketones under mild conditions. Very recently we reported on the E-selective semihydrogenation of alkynes to (E)-alkenes catalyzed by the acridinebased pincer complex [(HACRPNP)Fe(CH3CN)(k2-MeCHNB(H2)H)] HACR (E, Scheme 1, PNP = 4,5-bis(diphenylphosphino)-9H[25] acridine-10-ide). Prompted by the remarkable activity of A in catalytic reactions, we are currently investigating analogous iron complexes. In this context, we recently reported on ligation reactions of FeBr2 and FeCl2 with the bipyridine-based pincer ligands tBu-PNN, iPr-PNN (6-[(diisopropylphosphino)methyl]-2,2’-bipyridine), and Ph-PNN (6-[(diphenylphosphino)methyl]-2,2’-bipyridine).[26] These reactions give, depending on the initial ratios between ligand and metal salts and depending on the PR2 substituents, mono-chelated high-spin complexes of the type [(R-PNN)Fe(X)2] (1: R = tBu, X = Cl; 2: R = tBu, X = Br; 3: R = iPr, X = Cl; 4: R = iPr, X = Br; Scheme 1) and bis-chelated dicationic low-spin complexes of the type [(R-PNN)2Fe]2 + . Furthermore, we have shown that stepwise deprotonation of the complexes [(R-PNN)2Fe]2 + results in the dearomatization of the central pyridine of the PNN ligands with formation of monocationic complexes of the type [(R-PNN)(R-PNN*)Fe] + and subsequently the neutral complexes of the type [(R-PNN*)2Fe]. Shortly before our publication, a report on the application of the dichloro complex 1 as a highly active precatalyst for the hydroboration of alkenes with pinacolborane was published by Huang and co-workers.[27] In a typical reaction, complex 1 was activated by the addition of three equivalents of NaBHEt3. No catalytic activity being observed in the absence of NaBHEt3. The authors proposed on the basis of “preliminary data and precedent regarding related iron-catalyzed alkene addition processes” a mechanism that is initiated by the reduction of the complex by NaBHEt3 to give a four-coordinated iron(0) alkene complex of the type [(tBu-PNN)Fe(h2-H2C=CHR)]. However, no experimental evidence for the formation of such complexes was provided. Herein, we describe the synthesis, characterization, and crystal structures of three differently P-substituted iron dicarbonyl pincer complexes of the type [(R-PNN)Fe(CO)2], which are obtained by reduction reactions of the corresponding iron(II) complexes in the presence of CO or by reaction of [Fe(CO)5] with the free pincer ligand. A NEVPT2/CASSCF study

4404

 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper shows that these complexes are best described as Fe0 complexes, with “innocent” ligands, despite the presence of a short CC bond connecting the rings of the bpy ligand.

Table 1. Selected spectroscopic data of complexes 5–7. The 15N chemical shifts were obtained by 1H-15N-HMQC NMR measurements.

Results and Discussion Treatment of the dihalide complexes 1–4 with two equivalents of NaBHEt3 or with an excess of NaBH4 in the presence of CO, leads to the formation of the corresponding neutral dicarbonyl complexes [(tBu-PNN)Fe(CO)2] (5) and [(iPr-PNN)Fe(CO)2] (6) (Scheme 2). These complexes are, in contrast to the starting materials, highly soluble in apolar solvents, such as toluene, benzene, and pentane. No differences in the reactivites of the chloro and bromo complexes were observed in these reactions. However, the reduction of 2 or 4 with elemental zinc under CO atmosphere in toluene at 105 8C is a more convenient way to synthesize 5 and 6, as filtration and evaporation of the solvent gives spectroscopically clean products. The analogue Ph-substituted complex [(Ph-PNN)Fe(CO)2] (7) was synthesized by heating the free ligand with a slight excess of [Fe(CO)5] (1.8 equiv) in dioxane to 95 8C (Scheme 2).

Scheme 2. Synthesis of the neutral dicarbonyl complexes 5–7.

The intensely blue-purple, air-sensitive complexes 5–7 are diamagnetic. Magnetic measurements in solution at room temperature (Evans’ method[28]) were performed on the isolated complexes. No paramagnetic shifting was observed, which is consistent with S = 0 spin ground states. Selected spectroscopic data of complexes 5–7 are given in Table 1. The complexes 5 and 7 have a plane of symmetry through the iron pincer moiety in solution at room temperature, suggesting a distorted trigonal bipyramidal structure. The 1H NMR spectra of the complexes show one resonance for the benzylic protons, and one set of resonances for the substituents on the phosphine moieties in the 1H and 13C{1H} NMR spectra. The two carbonyl ligands result in one doublet in 13C{1H} NMR spectra with coupling constants of 2JPC = 13.6 and 16.9 Hz for complex 5 and 7, respectively (Table 1). The NMR spectra of complex 6 show at room temperature extremely broad resonances for the atoms that are directly coordinated to the iron atom (P and CO resoChem. Eur. J. 2014, 20, 4403 – 4413

www.chemeurj.org

[(tBu-PNN)Fe(CO)2] [(iPr-PNN)Fe(CO)2] [(Ph-PNN)Fe(CO)2] (5) (6) (7) d(P) [ppm] d(N1) [ppm] d(N2) [ppm] d(CO) [ppm] n(CO)[a] [cm1] n(CO)[b] [cm1]

137.9[c] 257.1[c] 261.2[c] 221.8[c] 1860, 1920 1838, 1895

120.7[d,e] 262.5[d,f] 265.6[d,f] 219.3[d,e] 1856, 1913 1847, 1907

101.4[c] 259.9[d] 264.3[d] 218.2[c] 1863, 1925 1857, 1917

[a] In DCM solution. [b] Thin film on NaCl. [c] In C6D6. [d] In [D8]toluene. [e] Measured at 80 8C. [f] Measured at 40 8C.

nances), while the hydrogen and carbon resonances of the pincer ligand appear as sharp signals in the 1H and 13C{1H} NMR spectra. Variable-temperature NMR measurements (80 to + 80 8C) revealed a significant sharpening of the phosphorous resonance in the 31P{1H} NMR spectra at lower and higher temperatures (Figure S3 and S4 in the Supporting Information). The 1H and 13C{1H} NMR spectra suggest that complex 6 exhibits a plane of symmetry through the iron pincer moiety, similar to 5 and 7. The complex shows one resonance for the protons on the benzylic arms and one set of resonances for the iPr groups throughout the whole temperature range. The carbonyl resonances were not detected at room temperature in the 13 1 C{ H} NMR spectrum. However, one broadened resonance was observed at a chemical shift of d = 219.3 ppm at 80 8C, and the spectrum at + 80 8C exhibits one sharp doublet at d = 219.3 ppm with a coupling constant of 2JPC = 16.9 Hz. The solution and solid-state infrared (IR) spectra of complexes 5–7 show for each complex two strong CO stretches (Table 1). The relative intensities of the two bands are in a range of 1:0.9 to 1:1 for all complexes, indicating OC-Fe-CO angles of about 908 in solution and in the solid state. Within the series 5, 6, and 7, the absorption with the lowest frequency (representing highest p-back-bonding) is found for the tBusubstituted complex 5 and the stretches with the highest frequency (representing weakest p-back-bonding) are observed for the Ph-substituted complex 7 in the solid-state IR spectra (Table 1), as expected. This phenomenon is accompanied by a slight shift of the carbonyl resonances to lower ppm in the 13 1 C{ H} NMR spectra (Table 1). X-ray diffraction analyses were performed on single crystals of complexes 5–7 (Figure 1, for selected bond lengths and angles see Tables 2–4). The molecular structures of the complexes in the solid state are similar and the geometries about the iron atoms are closer to square pyramidal than to trigonal bipyramidal. The pincer ligands coordinate in a meridional fashion, one of the CO ligands (C30) completes the fourth coordination site of the basal plane, while the second (C31) is coordinated at the apical position. The bite angles of the pincer ligands (N1-Fe-P), which range from 145.95(4) to

4405

 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper

Figure 1. ORTEP diagram of the molecular structure of [(tBu-PNN)Fe(CO)2] (5, left), [(iPr-PNN)Fe(CO)2] (6, middle), and [(Ph-PNN)Fe(CO)2] (7, right) (ellipsoids set at the 50 % probability level). The hydrogen atoms are omitted for clarity. For color pictures, see Figure S17 in the Supporting Information. Selected bond lengths and angles are given in Tables 2–4.

Table 2. Selected bond lengths [] of [(tBu-PNN)Fe(CO)2] [(iPr-PNN)Fe(CO)2] (6), and [(Ph-PNN)Fe(CO)2] (7).

FeN1 FeN2 FeP FeC30 FeC31 C30O1 C31O2

(5),

5[a]

5[b]

6[a]

6[b]

7[a]

7[b]

1.9412(12) 1.9281(17) 2.2106(12) 1.7447(19) 1.771(2) 1.1586(18) 1.154(2)

1.944 1.941 2.217 1.743 1.761 1.171 1.167

1.9369(10) 1.9396(10) 2.1910(4) 1.7571(14) 1.7663(12) 1.1580(17) 1.1615(15)

1.943 1.952 2.186 1.749 1.762 1.171 1.170

1.9469(12) 1.9521(12) 2.1758(4) 1.7560(15) 1.7841(16) 1.1605(18) 1.153(2)

1.943 1.946 2.174 1.750 1.771 1.170 1.166

Table 4. Comparison of bond lengths [] of the bipyridine-based PNN pincer ligands of [(tBu-PNN)Fe(CO)2] (5), [(iPr-PNN)Fe(CO)2] (6), and [(Ph-PNN)Fe(CO)2] (7).

C1C2 C2C3 C3C4 C4C5 C5C6 C6C7 C7C8 C8C9 C9C10 C10C11 N1C1 N1C5 N2C6 N2C10 PC11

[a] Determined by X-ray single crystal diffraction. [b] The structure was optimized as a closed-shell singlet at the RM06/SDD(d) level of theory.

Table 3. Selected bond angles [8] of [(tBu-PNN)Fe(CO)2] [(iPr-PNN)Fe(CO)2] (6), and [(Ph-PNN)Fe(CO)2] (7).

N1-Fe-N2 N1-Fe-P N1-Fe-C30 N1-Fe-C31 N2-Fe-P N2-Fe-C30 N2-Fe-C31 P-Fe-C30 P-Fe-C31 C30-Fe-C31 Fe-C30-O1 Fe-C31-O2

(5),

5[a]

5[b]

6[a]

6[b]

7[a]

7[b]

80.09(6) 145.95(4) 95.61(7) 106.65(6) 82.24(6) 160.58(6) 101.13(8) 91.50(6) 105.18(7) 98.25(8) 178.80(13) 176.12(12)

80.30 143.40 96.24 110.50 82.05 164.39 98.78 92.28 103.71 96.69 179.20 178.19

80.38(4) 161.31(3) 95.03(5) 96.16(5) 84.03(3) 141.06(5) 115.80(5) 90.62(4) 99.88(4) 103.13(6) 178.25(12) 174.06(11)

80.34 160.41 94.92 98.31 83.25 137.37 120.32 89.92 99.18 102.30 176.92 172.85

79.98(5) 156.86(4) 92.37(6) 102.05(6) 81.72(4) 146.64(6) 110.82(6) 95.17(5) 97.61(5) 102.52(7) 176.27(13) 175.14(13)

80.38 157.74 94.18 103.18 82.50 143.88 113.49 91.28 96.64 102.56 178.36 174.44

161.31(3)8, are significantly smaller than 1808, and the CH2P arms are bent out of the plane of the bipyridine-iron unit (N1, N2, Fe) and are twisted away from the apical CO ligand. The distances of the P atoms to this plane are 1.0696(6), 0.3916(3), and 0.5295(4)  for 5–7, respectively. A comparison of the three structures shows that 6 features the strongest distortion towards trigonal bipyramidal coordination. However, the basal www.chemeurj.org

5[b]

6[a]

6[b]

7[a]

7[b]

1.3698(18) 1.397(2) 1.364(2) 1.4084(18) 1.426(2) 1.399(2) 1.369(2) 1.404(2) 1.370(2) 1.4985(19) 1.3655(18) 1.3690(19) 1.3716(17) 1.3677(18) 1.8383(19)

1.377 1.413 1.379 1.408 1.433 1.408 1.380 1.413 1.380 1.497 1.361 1.375 1.367 1.362 1.867

1.3703(17) 1.4046(18) 1.3703(18) 1.4071(16) 1.4328(17) 1.4005(16) 1.3741(18) 1.4056(17) 1.3713(16) 1.5012(17) 1.3608(16) 1.3694(15) 1.3775(14) 1.3724(15) 1.8409(13)

1.380 1.410 1.381 1.407 1.437 1.405 1.381 1.411 1.381 1.499 1.355 1.367 1.370 1.367 1.859

1.367(2) 1.397(2) 1.374(2) 1.401(2) 1.4376(19) 1.400(2) 1.373(2) 1.401(2) 1.371(2) 1.495(2) 1.3635(18) 1.3693(18) 1.3670(19) 1.3710(18) 1.8401(14)

1.379 1.411 1.380 1.407 1.436 1.405 1.381 1.412 1.380 1.497 1.356 1.369 1.369 1.368 1.860

[a] Determined by X-ray single crystal diffraction. [b] The structure was optimized as a closed-shell singlet at the RM06/SDD(d) level of theory.

[a] Determined by X-ray single crystal diffraction. [b] The structure was optimized as a closed-shell singlet at the RM06/SDD(d) level of theory.

Chem. Eur. J. 2014, 20, 4403 – 4413

5[a]

carbonyl ligands are twisted away from the carbonyl in the apical positions in all structures. The distances of the planes fitted through the iron pincer unit (N1, N2, P, Fe) to the carbonyl carbon atom (C30) coordinated in the basal plane are 0.4809(14), 1.1326(13), and 1.0411(15)  for 5–7, respectively. The angles between the two carbonyl carbon atoms (C30-FeC31), as well as the angles between the apical carbonyl carbon atom and the central nitrogen atom of the pincer ligand (N2Fe-C31) are significantly larger than the ideal 908, while the nitrogen–iron–carbon angles between apical carbonyl carbon atom and the central nitrogen are significantly smaller than 1808 (Table 3). The shorter FeC and longer CO distances for the basal carbonyl ligands (Table 2) indicate stronger p-backdonation to these carbonyls in the position trans to the central pyridine nitrogen (N2). This effect is less pronounced for com-

4406

 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper plex 6, which is likely to be a result of the strong distortion towards trigonal bipyramidal geometry. Although there are reports on square pyramidal five-coordinated iron(0) complexes, such as [(iPr-PNP)Fe(CO)2] recently described by Krogh-Jespersen, Goldman, and co-workers,[29] Fe0 complexes with trigonal bipyramidal structures are more expected.[30] The interpyridine C5C6 bond lengths increase within the series 5–7 and are 1.426(2), 1.4328(17), and 1.4376(19) , respectively. These values are significantly shorter than for other structurally characterized complexes bearing the same ligands.[16c, 18, 26] The iron(II) high-spin complexes 1–4, for example, feature C5C6 bond lengths between 1.476(3)  (for 4) and 1.492(3)  (for 1). Wieghardt and others have shown that bipyridine (bpy) ligands are redox active and can exist in three different oxidation states: the neutral ligand (bpy0), the p-radical monoanion (bpyC), and the diamagnetic dianion (bpy2).[14, 31] These three different types have different structures. The lowest-energy unoccupied molecular orbital (LUMO) of the neutral bpy0 has a p-bonding contribution between carbon atoms of the interpyridine CC bond. Thus, reduction of the bpy ligands leads to a distinct stepwise shortening of this bond. Experimental CC bond lengths, determined by X-ray diffraction, for bpy0, bpyC , and bpy2, are approximately 1.49,[32] 1.43,[33] and 1.40-1.36 ,[34] respectively. This suggests that the bpy-based ligands in 5–7 could be considered as radical anions, with the missing spin attributable to antiferromagnetic coupling of the unpaired electron on a low-spin FeI center. Examples of biradical complexes featuring antiferromagnetically coupled bpy radicals have been reported.[31] However, our NEVPT2/CASSCF studies (vide infra) indicate that they are better described as Fe0 complexes with charge-neutral bpy ligands. In order to investigate the electrochemical properties of the dicarbonyl complexes 5-7, cyclic voltammograms (CV) were recorded in THF containing 0.1 m [N(nBu)4][PF6] supporting electrolyte at ambient temperature (Figure 2). The redox potentials, referenced versus the Fc/Fc + couple (Fc = ferrocene), are summarized in Table 5. The complexes undergo three redox pro-

Figure 2. Cyclic voltammograms of [(tBu-PNN)Fe(CO)2] (5, top), [(iPr-PNN)Fe(CO)2] (6, middle), and [(Ph-PNN)Fe(CO)2] (7, bottom) in THF (0.1 m [N(nBu)4][PF6] supporting electrolyte) at room temperature with scan rates of 100 mV s1 (glassy carbon working electrode). Potentials are referenced versus the Fc + /Fc couple.

sample was measured in a potential range from 2.5 to 1.0 V (Figure S9 in the Supporting Information). The 57Fe Mçssbauer spectra of complexes 5 and 7 were obtained to further evaluate the electronic structure of the [(R-PNN)Fe(CO)2] complexes (Figure 3). The 80 K Mçssbauer spectrum of 5 (Figure 3, top) is well-fit to a major species (ca. 93 % of iron) with parameters d = 0.02 mm s1 and DEQ = 1.49 mm s1. There is also a minor species in this sample with

Table 5. Redox potentials [V] for [(tBu-PNN)Fe(CO)2] (5), [(iPr-PNN)Fe(CO)2] (6), and [(Ph-PNN)Fe(CO)2] (7).[a]

5 6 7

E 11=2

E 21=2

E 31=2

2.43 2.39 2.37

0.73 0.64 0.61

0.18[b] 0.03 0.06

[a] Measured in 0.1 m [N(nBu)4][PF6] in THF, referenced versus Fc + /Fc. [b] Irreversible.

cesses in the range from 3.0 to + 1.5 V. Complex 7 shows an additional oxidation wave at 1.21 V, which presumably stems from a species formed in a nonreversible process. It is likely that this species stems from the nonreversible reduction at 2.37 V, since this oxidation was not observed when the same Chem. Eur. J. 2014, 20, 4403 – 4413

www.chemeurj.org

Figure 3. The 80 K 57Fe Mçssbauer spectra of [(tBu-PNN)Fe(CO)2] (5, top) and [(Ph-PNN)Fe(CO)2] (7, bottom). The data (black dots), overall fit (black lines) and individual components are given for each spectrum. For Mçssbauer parameters see text and for colored graphs, see Figure S18 in the Supporting Information.

4407

 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper d = 1.30 mm s1 and DEQ = 2.67 mm s1, consistent with an oxidized high-spin iron(I) or iron(II) impurity. The 80 K Mçssbauer spectrum of 7 (Figure 3, bottom) is well-fit to a major species (ca. 94 % of iron) with d = 0.04 mm s1 and DEQ = 1.93 mm s1. There is also a minor species in this sample with d = 0.15 mm s1 and DEQ = 0.61 mm s1, potentially a high-spin iron(III) impurity due to the low isomer shift and small quadrupole splitting. The rigorous assignment of iron oxidation state in low-isomer shift iron species with pincer ligation is challenging without theoretical calculations or additional experimental measurements as iron(0), low-spin iron(I), and low-spin iron(II) can all exhibit low isomer shifts (e.g., previous studies of [(PNP)Fe(CO)(Cl)2] low-spin iron(II) complexes by Kirchner and co-workers yielded isomer shifts of ca. 0.13–0.15 mm s1).[35] In this study, the observed isomer shifts of the major species for both complexes are assigned to iron(0) complexes based on the correlation with the electronic structure descriptions of 5 and 7 determined from NEVPT2/CASSCF calculations (vide infra). The differences in the quadrupole splitting observed between complexes 5 and 7 reflect the variations in structural distortions of the two iron(0) complexes due to the different R group substitutions (i.e., tBu and Ph), as evidenced in their solid-state structures. X-ray photoelectron spectroscopy (XPS) was used to probe the composition and the electronic structure of the paramagnetic high-spin formal FeII complex [(tBu-PNN)Fe(Br)2] (2) and the diamagnetic low-spin formal Fe0 complex [(tBu-PNN)Fe(CO)2] (5). The XPS spectrum of 2 shows the main peak of the Fe 2p3/2 binding energy centered at 709.5 eV (see Supporting Information). As expected for paramagnetic complexes,[36] an additional a shake-up satellite is observed centered at 714.3 eV. The Fe 2p3/2 binding energy of complex 5 is found at 708.0 eV and the spectrum does not show a shake-up satellite, in agreement with this being a diamagnetic complex.[36] Due to the high sensitivity of complex 5 towards air, surface oxidation was observed in the measurements (presumably stemming from the sample preparation). Note that XPS is a surface sensitive method that probes approximately the top 10 nm of the material.[37] In the course of our studies several samples of 5 were measured and the spectra show mixtures of complex 5 and a single additional unidentified oxidized species in different ratios (see Supporting Information). The oxidized complex has, similarly to 2, a shake-up satellite, suggesting that it is paramagnetic. The binding energy of the Fe 2p3/2 electron of the oxidized species was observed at 709.9 eV, which is fairly close to 709.5 eV, observed for 2. This suggests that the oxidized species might be a formal FeII complex. By comparison of the core electron binding energies of different metal complexes in the literature two different trends become apparent: The binding energy increases 1) with the oxidation state and 2) with the electron-withdrawing character of the ligands.[36, 38] These effects are, however, not quantitatively predictable and depend strongly on the coordination sphere of the metal center and on the electronic nature of the ligands. The comparison of the spectra of complexes 2 and 5 clearly confirms that the iron center of 5 is more electron-rich and has a lower effective nuclear charge than the iron center of 2. Chem. Eur. J. 2014, 20, 4403 – 4413

www.chemeurj.org

Although these results do not allow for an unambiguous assignment of the oxidation state of the iron in complex 5, this is in line with a lower oxidation state. The electronic structures of complexes 5–7 were investigated computationally at the NEVPT2/CASSCF(12,12) level of theory (see Computational Methods section for full details). A multireference method was chosen to describe the electronic structure of molecules with a potential biradical character.[39] In agreement with the experimental results, the ground states for all complexes are found to be singlet states (S = 0), and the triplet states (S = 1) lie considerably higher in energy. The differences in energy between the singlet and triplet states decrease within the series 5–7 (28.3, 26.6, and 23.3 kcal mol1, respectively). The calculated structural parameters for the singlet ground states of all complexes are in reasonable agreement with the corresponding crystallographic data (Table 2-4). The active space consists of twelve molecular orbitals (MOs) for all complexes (see Supporting Information). A comparison of the different MOs shows that they are very similar for the complexes 5–7. The highest-energy occupied molecular orbital (HOMO = f(6)) and LUMO (= f(7)) of complex 5 are depicted in Figure 4. The HOMO mainly consists of contributions of the dz2 orbital of the iron atom and an orbital of the bpy moiety, which shows significant similarities to the LUMO of the neutral bpy.[14, 29] It features a p-bonding character between the interpyridine carbon atoms (C5C6) and has a nodal plane between this carbon atoms and the corresponding nitrogen atoms (i.e., between C5N1 and C6N2). The LUMO of complex 5 can be described as the antibonding orbital to the HOMO, with additional minor contributions of antibonding of the Fe center to the apical CO. This situation corresponds to p-back-bonding from the iron to the antibonding orbital of the bpy moiety.

Figure 4. HOMO (f(6), left) and LUMO (f(7), right) of [(tBu-PNN)Fe(CO)2] (5). For colored pictures, see Figure S19 in the Supporting Information.

The CASSCF studies for the singlet ground states of the complexes show that only two configurations, labeled as y1 and y2, have major contributions to the wavefunction yCASSCF (Table 6, Table S1). As shown in Table 6, the y1 configuration has all bonding MOs (f(1)–f(6)) occupied by two electrons and the y2 configuration represents the situation of double excitation from the HOMO (f(6)) to the LUMO (f(7)). The y1 configuration is strongly dominant, and the y2 configuration has a very minor contribution to the wavefunction. The single excitation from the HOMO (f(6)) to the LUMO (f(7)), labeled as y3, reflects the situation of an open-shell singlet state. This biradical configuration, however, does not have a significant

4408

 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper fraction, coincide exactly with those reported for the p-radical anionic bipyridine (bpyC). This is a result of extensive p-backbonding from the Fe0 center into the antibonding orbital of the bpy moiety and does not stem from a one-electron transfer from the iron to the ligand under formation of a FeI–bpyC couple.[39] A very recent report by Huang and co-workers on the application of [(tBu-PNN)Fe(Cl)2] (1) as precatalyst for hydroboration reactions[27] demonstrates the high catalytic potential of these complexes. The present study suggests that Fe0 species are formed under the reported conditions, as suggested by the authors, and may be involved in the catalysis. Studies of the scope of complexes [(R-PNN)Fe(L)n] in different catalytic reactions are currently underway in our laboratories.

Table 6. Weights of the contributions (j ci j 2) to the wavefunction yCASSCF for the singlet ground state.

configuration

j ci j 2 of complex 5

j ci j 2 of complex 6

j ci j 2 of complex 7

y1 y2 y3

0.7820 0.0575 < 0.01[a]

0.7929 0.0460 < 0.01[a]

0.7920 0.0494 < 0.01[a]

Experimental Section

[a] Below print threshold of the electronic structure program (Orca).

General considerations

Table 7. NBO group charges (Q) for the singlet ground states of [(tBu-PNN)Fe(CO)2] (5), [(iPr-PNN)Fe(CO)2] (6), and [(Ph-PNN)Fe(CO)2] (7).

Fe C30O1 C31O2 R-PNN

5

6

7

0.09 0.01 0.06 0.03

0.05 0.08 0.09 0.12

0.05 0.06 0.07 0.08

contribution to the wavefunction, as the weights of y3 for all complexes are smaller than 0.1 %. Natural bond analysis (NBO) was performed to investigate the charge distribution within the complexes. The results for the closed-shell ground state are summarized in Table 7 and show only slight positive charges on the iron atom in all complexes (for NBO charges (Q) of excited states and for Lçwdin spin densities, see Table S3 in the Supporting Information). Consequently, complexes 5–7 are best described as Fe0 complexes with a neutral R-PNN ligands, that have low-spin ground states (i.e., S = 0), and the Fe centers are involved in p-backbonding to the bpy LUMO.

All reactions were performed under a nitrogen atmosphere in a glovebox or using standard Schlenk techniques. All solvents were reagent grade or better. Tetrahydrofuran (THF), 1,4-dioxane, benzene, toluene, diethyl ether, and pentane were refluxed over sodium and distilled under an argon atmosphere. Methylene chloride (DCM), methanol (MeOH), and acetonitrile (CH3CN) were degassed by freeze-pump thaw cycles and stored in the glovebox over the appropriate molecular sieves. Deuterated solvents were purged with argon and stored in the glovebox over the appropriate molecular sieves. Other commercially available reagents were used as received. NMR spectra were recorded using Bruker AMX-300, AMX-400, and AMX-500 NMR spectrometers. 1H and 13C{1H} NMR chemical shifts are reported in ppm downfield from tetramethylsilane. 31P{1H} NMR chemical shifts are reported in ppm downfield from H3PO4 (0.0 ppm) and are referenced to an external 85 % solution of phosphoric acid in D2O. NMR assignments (Scheme 3) were assisted by 1 H-1H-COSY, 1H-31P-HMQC, 1H-13C-HSQC, 1H-13C-HMBC and 13CDEPTQ NMR spectroscopy, as required. 15N chemical shifts were identified by 1H-15N-HMQC NMR measurements and are reported downfield from liquid ammonia (0.0 ppm). The effective magnetic moments in solution were measured by the Evans’ method[28] at ambient temperature. IR spectra were recorded on a Nicolet FT-IR spectrophotometer. Elemental analyses and ESI-MS spectroscopy were performed by the Department of Chemical Research Support, Weizmann Institute of Science. 57

Conclusion In this contribution, we have reported on the synthesis and full characterization of new iron dicarbonyl complexes [(R-PNN)Fe(CO)2] (5: R = tBu, 6: R = iPr, and 7: R = Ph) featuring bipyridine-based PNN pincer ligands. These neutral, diamagnetic, low-spin complexes were prepared by the reduction of the corresponding high-spin iron(II) dihalide complexes [(R-PNN)Fe(X)2] under a CO atmosphere or by reaction of [Fe(CO)5] with the free pincer ligand. Although metrical parameters for the assignment of oxidation states of bipyridine ligands are well established and for many examples have been demonstrated to be useful, the reported complexes do not follow this trend. The bond lengths, established by X-ray difChem. Eur. J. 2014, 20, 4403 – 4413

www.chemeurj.org

Fe Mçssbauer spectroscopic data was collected on non-enriched samples of the as-isolated complexes. All samples were prepared in an inert atmosphere glove box equipped with a liquid nitrogen fill port to enable sample freezing to 77 K within the glove box. Each sample was loaded into a Delrin Mçssbauer sample cup for measurements and loaded under liquid nitrogen. Low-temperature 57 Fe Mçssbauer measurements were performed using a SeeCo MS4 Mçssbauer spectrometer integrated with a Janis SVT-400T He/N2 cryostat for measurements at 80 K with a 0.07 T applied magnetic field. Isomer shifts were determined relative to a-Fe at 298 K. All Mçssbauer spectra were fit using the program WMoss (SeeCo). Powder samples were loaded to the XPS instrument via glove-box purged for several hours with N2 or Ar. XPS measurements were carried out with Kratos AXIS ULTRA system using a monochromatized AlKa X-ray source (hn = 1486.6 eV) at 75 W and detection pass energy of 20 eV. A low-energy electron flood gun (eFG) was ap-

4409

 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper plied for charge neutralization. To define binding energies (BE) of different elements, the C 1s line at 284.8 eV was taken as a reference.[40] Curve fitting analysis was based on linear or Shirley background subtraction and application of Gaussian-Lorentzian line shapes. Complexes 1–4 and Ph-PNN were prepared as previously reported.[26]

Scheme 3. Assignment of the carbon and hydrogen atoms of the PNN ligands.

Synthesis of complexes 5–7 [(tBu-PNN)Fe(CO)2] (5): Complex 2 (272 mg, 0.52 mmol) and zinc dust (ca. 2 g) were suspended in toluene (30 mL). The suspension was transferred to a 100 mL Schlenk tube and immersed in a bath of liquid nitrogen. The tube was evacuated on a high-vacuum line and 1 atm of carbon monoxide was added at 196 8C. The reaction mixture was allowed to warm up to ambient temperature and was heated to 105 8C for 108 h. The insoluble residue was filtered off, and all volatiles were removed in vacuo to afford 168.3 mg (74 %) of a dark purple to blue powder. Single crystals of the complex, suitable for X-ray diffraction, were obtained by cooling a saturated solution in DCM to 20 8C. 1H NMR (500 MHz, C6D6, 23 8C): d = 1.04 (d, 3JPH = 12.7 Hz, 18 H; H13), 3.19 (d, 2JPH = 8.9 Hz, 2 H; H11), 6.30 (t, 13JHH = 6.6 Hz, H; H2), 6.42 (d, 3JHH = 6.6 Hz, 1 H; H9), 6.71 (t, 3JHH = 7.5 Hz, 1 H; H3), 6.78 (t, 3JHH = 7.5 Hz, 1 H; H8), 7.31 (d, 3JHH = 8.2 Hz, 1 H; H7), 7.40 (d, 3JHH = 6.6 Hz, 1 H; H4), 9.82 ppm (br d, 3JHH = 6.6 Hz, 1 H; H1); 1H{31P} NMR (500 MHz, C6D6, 23 8C): d = 1.04 (s, 18 H; H13), 3.20 (s, 2 H; H11), 6.30 (vt, 3JHH = 6.4 Hz, 1 H; H2), 6.43 (vt, 3JHH = 6.6 Hz, 1 H; H9), 6.71 (m, H3), 6.78 (vt, 3JHH = 8.3 Hz, 1 H; H8), 7.31 (d, 3 JHH = 8.4 Hz, 1 H; H7), 7.40 (d, 3JHH = 8.3 Hz, 1 H; H4), 9.81 ppm (d, 3 JHH = 6.3 Hz, 1 H; H1); 13C{1H} NMR (125 MHz, C6D6, 23 8C): d = 29.3 (d, 2JPC = 3.7 Hz, C13), 35.4 (d, 1JPC = 16.4 Hz, C11), 37.5 (d, 1JPC = 12.5 Hz, C12), 112.3 (d, 3JPC = 12.5 Hz, C9), 116.4 (s, C2), 119.6 (s, C7), 122.0 (s, C4), 122.4 (s, C8), 124.4 (s, C3), 143.4 (d, 3JPC = 4.3 Hz, C6), 147.9 (s, C5), 154.4 (s, C1), 158.3 (d, 2JPC = 8.2 Hz, C10), 221.8 ppm (d, 2JPC = 13.6 Hz, CO); 31P{1H} NMR (162 MHz, C6D6, 23 8C): d = 137.9 ppm (s, tBu-PNN); 15N NMR (41 MHz, C6D6, 23 8C): d = 257.1 (N1), 261.2 ppm (N2); IR (DCM): n˜ = 1860 (nCO), 1920 cm1 (nCO); IR (thin film, NaCl): n˜ = 1838 (nCO), 1895 cm1 (nCO); ESI-MS: m/z: 353.23 [tBu-PNN + K] + = [C19H27N2KP] + , 337.25 [tBu-PNN + Na] + = [C19H27N2NaP] + ; magnetic susceptibility (Evans): meff = 0.0 mB (1,4-dioxane in C6D6, 23 8C). Chem. Eur. J. 2014, 20, 4403 – 4413

www.chemeurj.org

[(iPr-PNN)Fe(CO)2] (6): Complex 4 (170.0 mg, 0.340 mmol) and zinc dust (ca. 2 g) were suspended in toluene (30 mL). The suspension was transferred to a 100 mL Schlenk tube and immersed in a bath of liquid nitrogen. The tube was evacuated on a high-vacuum line and 1 atm of carbon monoxide was added at 196 8C. The reaction mixture was allowed to warm up to ambient temperature and was heated to 105 8C for 16 h. The insoluble residue was filtered off, and all volatiles were removed in vacuo to afford 68.0 mg (50 %) of a dark purple to blue powder. Single crystals of the complex, suitable for X-ray diffraction, were obtained by cooling a saturated solution in toluene/pentane (2:1) to 20 8C. 1H NMR (400 MHz, [D8]toluene, 23 8C): d = 0.89 (dd, 3JPH = 14.2 Hz, 3JHH = 7.0 Hz, 6 H; H13), 0.96 (dd, 3JPH = 15.0 Hz, 3JHH = 7.1 Hz, 6 H; H13’), 1.84 (dsept, 2 JPH = 14.0 Hz, 3JHH = 6.9 Hz, 2 H; H12), 2.99 (br d, 2JPH = 9.4 Hz, 2 H; H11), 6.30 (vt, 3JHH = 6.6 Hz, 5JPH = 1.4 Hz, 1 H; H2), 6.45 (br d; 3JHH = 6.7 Hz, 1 H; H9), 6.70 (br, 1 H; H3), 6.74 (br, 1 H; H8), 7.26 (d, 3JHH = 8.3 Hz, 1 H; H7), 7.35 (d, 3JHH = 8.5 Hz, 1 H; H4), 9.75 ppm (br d, 4 JPH = 5.6 Hz, 1 H; H1); 1H{31P} NMR (400 MHz, [D8]toluene, 23 8C): d = 0.89 (d, 3JHH = 7.0 Hz, 6 H; H13), 0.96 (d, 3JHH = 7.0 Hz, 6 H; H13’), 1.84 (sept, 3JHH = 7.0 Hz, 2 H; H12), 2.99 (br, 2 H; H11), 6.30 (vt, 3JHH = 6.3 Hz, 1 H; H2), 6.45 (d, 3JHH = 6.6 Hz, 1 H; H9), 6.71 (br, 2 H; H3 + H8), 7.26 (d, 3JHH = 8.3 Hz, 1 H; H7), 7.35 (d, 3JHH = 8.4 Hz, 1 H; H4), 9.74 ppm (br, 1 H; H1); 13C{1H} NMR (100 MHz, [D8]toluene, 23 8C): d = 17.9 (br m, C13 + C13’), 28.5 (br d, 1JPC = 21.7 Hz, iPr-CH), 37.1 (br d, 1JPH = 20.6 Hz, CH2P), 112.5 (br d, 3JPC = 10.1 Hz, C9), 116.2 (br s, C2), 119.7 (br, C7), 122.0 (br s, C4), 122.2 (br s, C8), 125.0 (br s, C3), 143.9 (br s, C6), 149.1 (br s, C5), 155.0 (br s, C1), 158.1 ppm (br, C10); 13 1 C{ H} NMR (125 MHz, [D8]toluene, 80 8C): d = 17.5 (br m, C13 + C13’), 26.4 (br, C12), 36.7 (br d, 1JPH = 18.1 Hz, C11), 112.7 (br, C9), 116.5 (br s, C2), 119.6 (br, C7), 122.0 (br s, C4), 122.2 (br s, C8), 125.6 (br s, C3), 143.6 (br, C6), 148.8 (br s, C5), 154.9 (br s, C1), 158.1 (br, C10), 219.3 ppm (br, CO); 13C{1H} NMR (125 MHz, [D8]toluene, 80 8C): d = 18.2 (br m, C13 + C13’), 28.7 (d, 1JPH = 21.4 Hz, C13), 37.6 (d, 1 JPH = 19.8 Hz, C11), 112.7 (d, 3JPC = 10.2 Hz, C9), 116.2 (s, C2), 119.9 (s, C7), 122.2 (d, 3JPC = 0.8 Hz, C8), 122.3 (s, C4), 125.1 (s, C3), 144.2 (d, 3JPC = 4.5 Hz, C6), 149.3 (s, C5), 155.2 (s, C1), 158.2 (d, 2JPC = 8.4 Hz, C10), 219.3 ppm (d, 2JPC = 16.9 Hz, CO); 31P{1H} NMR (162 MHz, [D8]toluene, 23 8C): d = 121.5 ppm (br, iPr-PNN); 31P{1H} NMR (202 MHz, [D8]toluene, 80 8C): d = 120.7 ppm (s, iPr-PNN); 15 N NMR (41 MHz, [D8]toluene, 40 8C): d = 262.5 (N1), 265.6 ppm (N2); IR (DCM): n˜ = 1856 (nCO), 1913 cm1 (nCO); IR (thin film, NaCl): n˜ = 1847 (nCO), 1907 cm1 (nCO); ESI-MS: m/z: 325.21 [iPr-PNN + K] + = [C17H23N2KP] + , 309.23 [iPr-PNN + Na] + = [C17H23N2NaP] + ; magnetic susceptibility (Evans): meff = 0.0 mB (1,4-dioxane in C6D6, 23 8C). [(Ph-PNN)Fe(CO)2] (7): Ph-PNN (200 mg, 0.56 mmol) was placed in a 100 mL Schlenk tube and a solution of [Fe(CO)5] (200 mg, 1.02 mmol) in dioxane (15 mL) was added. The reaction mixture was heated to 95 8C for 26 h in a closed reaction vessel. The Schlenk tube was quickly evacuated and refilled with N2 three times (after 1.5 h, after 3.5 h, and after 15.5 h). The reaction mixture was allowed to reach room temperature, the insoluble residue was filtered off and washed with dioxane (2  5 mL). The solutions were combined, all volatiles were removed in vacuo, and the residue was dried for 6 h in high vacuum. The resulting residue was extracted with toluene (4  6 mL), all volatiles were removed from the combined solutions under reduced pressure, and the residue was dried for 24 h in high vacuum to give 248 mg (95 %) dark purple to blue powder. Single crystals of the complex, suitable for X-ray diffraction, were obtained by cooling a saturated solution in benzene/pentane (1:10) to 20 8C. 1H NMR (400 MHz, C6D6, 23 8C): d = 3.72 (d, 2JPH = 10.6 Hz, 2 H; H11), 6.30 (br vt, 3JHH = 5.6 Hz, 1 H; H2), 6.45 (br vd, 3JHH = 6.1 Hz, 1 H; H9), 6.74 (m, 2 H; H3 + H8), 6.95 (m, 6 H; H14 + H15), 7.30 (vd, 3JHH = 7.9 Hz, 1 H; H7), 7.39 (vd, 3JHH =

4410

 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper 8.1 Hz, 1 H; H4), 7.54 (br, 4 H; H13), 9.82 ppm (br vd, 3JHH = 4.9 Hz, 1 H; H1); 1H{31P} NMR (400 MHz, C6D6, 23 8C): d = 3.72 (s, 2 H; H11), 6.30 (br vt, 3JHH = 5.5 Hz, 1 H; H2), 6.44 (br vd, 3JHH = 6.1 Hz, 1 H; H9), 6.74 (m, 2 H; H3 + H8), 6.95 (m, 6 H; H14 + H15), 7.30 (vd, 3JHH = 8.2 Hz, 1 H; H7), 7.39 (vd, 3JHH = 8.0 Hz, 1 H; H4), 7.54 (br vd, 3JHH = 3.4 Hz, 4 H; H13), 9.82 ppm (br vd, 3JHH = 5.4 Hz, 1 H; H1); 13C{1H} NMR (125 MHz, C6D6, 23 8C): d = 43.9 (d, 1JPC = 26.8 Hz, C11), 114.1 (d, 3JPC = 11.6 Hz, C9), 116.9 (s, C2), 120.0 (s, C7), 122.1 (s, C4), 122.9 (s, C8), 125.5 (s, C3), 128.6 (d, 3JPC = 8.9 Hz, C14), 130.0 (d, 4JPC = 2.3 Hz, C15), 132.0 (d, 2JPC = 10.9 Hz, C13), 136.2 (d, 1JPC = 40.1 Hz, C12), 144.5 (d, 3JPC = 4.5 Hz, C6), 149.2 (s, C5), 155.2 (s, C1), 155.9 (d, 3 JPC = 9.4 Hz, C10), 218.2 ppm (d, 2JPC = 16.9 Hz, CO); 31P{1H} NMR (162 MHz, C6D6, 23 8C): d = 101.4 ppm (s, Ph-PNN); 15N NMR (41 MHz, [D8]toluene, 23 8C): d = 259.9 (N1), 264.3 ppm (N2); IR (DCM): n˜ = 1863 (nCO), 1925 cm1 (nCO); IR (thin film, NaCl): n˜ = 1857 (nCO), 1917 cm1 (nCO); ESI-MS: 393.18 [iPr-PNN + K] + = [C23H19N2KP] + , 377.14 [iPr-PNN + Na] + = [C23H19N2NaP] + ; magnetic susceptibility (Evans): meff = 0.0 mB (1,4-dioxane in C6D6, 23 8C).

Table 8. Crystal data and summary of data collection and refinement for 5–7.a

formula crystal description crystal size [mm3] Mr [g mol1] space group crystal system a [] b [] c [] a [8] b [8] g [8] V [3] Z 1cacld [g cm3] m [mm1] reflns unique reflns 2qmax [8] Rint parameters final R[b] final R[c] gooF

General procedure for the synthesis of 5 and 6 by reactions of 1–4 with NaBHEt3 under CO atmosphere: The corresponding complex 1, 2, 3, or 4 (0.02 mmol) was suspended in C6D6 (0.6 mL) and a solution of NaBHEt3 in toluene (0.1 mol L1) was added (0.04 mmol). The reaction mixture was transferred in a J. Young NMR tube, the solvent was frozen, and the atmosphere was replaced by CO. The reaction mixture was allowed to warm up to room temperature and the freeze pump thaw procedure was repeated. 1H and 31P{1H} NMR measurements showed the formation of 5 and 6, respectively. Alternative synthesis of 5 by a reaction of 2 with NaBH4 under CO atmosphere: Complex 2 (264 mg, 0.5 mmol) was dissolved in DCM (15 mL) and CH3CN (5 mL), NaBH4 (190 mg, 5.0 mmol) was added in one portion. The solution was stirred for 1 h, transferred to a 100 mL Schlenk tube and immersed in a bath of liquid nitrogen. The tube was evacuated on a high-vacuum line and 1 atm. of carbon monoxide was added at 196 8C. The reaction mixture was allowed to warm up to ambient temperature. After several hours under carbon monoxide atmosphere the color changed from red to purple. This procedure was repeated after 16 h. After one day the suspension was filtered, all volatiles of the solution were removed in vacuo. The residue was extracted pentane (5  60 mL) to give a dark blue solution. Evaporation of all volatiles in vacuo yielded 85.0 mg (40 % yield) of a dark blue powder. NMR characterization was identical with the product derived from the reaction with Zn (see above).

X-ray structure determinations X-ray data were collected on Bruker APEX-II KappaCCD diffractometer equipped with Miracol optics and graphite monochromator at 100 K. Crystal data and summary of data collection and refinement for complexes 5–7 are given in Table 8. CCDC-973748 (complex 5), 973747 (complex 6), and 973746 (complex 7), contain the supplementary crystallographic data for this paper. These data can be obtained free of charge from The Cambridge Crystallographic Data Centre via www.ccdc.cam.ac.uk/data_request/cif.

Computational Methods DFT calculations were performed using Gaussian 09, Revision C.01.[41] The M06 (Minnesota06)[42] hybrid meta-GGA functional including the second version of Grimme’s empirical dispersion correction[43] has been used for geometry optimizations. The functional was successfully applied for similar systems.[44] The optimizations were carried out with the B1 basis set, which corresponds to a comChem. Eur. J. 2014, 20, 4403 – 4413

www.chemeurj.org

5

6

7

C21H27FeN2O2P black prism 0.40  0.26  0.08 426.27 P1¯ triclinic 8.243(8) 10.591(9) 12.194(3) 103.59(4) 101.93(3) 97.89(3) 992.8(13) 2 1.426 0.859 55715 7563 66.28 0.0403 250 0.0358 0.0513 1.051

C19H23FeN2O2P black prism 0.50  0.42  0.15 398.21 Pbca orthorhombic 15.0347(11) 15.5252(10) 15.9757(12) 90 90 90 3729.0(5) 8 1.419 0.909 97075 7161 66.60 0.0571 230 0.0304 0.0473 1.020

C25H19FeN2O2P black plate 0.40  0.21  0.07 466.24 P21/c monoclinic 15.9656(12) 9.2198(7) 14.7499(10) 90 106.776(3) 90 2078.8(3) 4 1.490 0.828 34752 5186 56.82 0.0341 280 0.0282 0.0362 1.040

[a] Collected using MoKa radiation (l = 0.71073 ). [b] For data with I > 2s(I). [c] For all data.

bination of the Huzinaga–Dunning double-z basis set (D95(d,p))[45] on the lighter atoms and Stuttgart–Dresden (SDD)[46] basis set in conjunction with a relativistic-effective-core potential on iron. The optimized structures were checked for the presence of imaginary frequencies at the same level of theory as the geometry optimization. All minima contain only real frequencies. The singlet states of 5–7 were optimized in restricted Kohn–Sham formalism because this electronic configuration is dominant in the CASSCF wavefunction (see discussion and ref. [39]). The triplet and quintet states of 5–7 were optimized in unrestricted Kohn–Sham formalism. Singlepoint energy calculations were carried out using Orca version 2.9.0.[47] The natural orbitals were generated at the RI-MP2 level of theory[48] using the Def2-SVP basis set.[49] The RIJCOSX approximation[50] was applied to speed up the single-point energy calculations. In the RIJCOSX approximation, the two electron integrals are approximated by the RI-J approximation[51] while the exchange integrals are approximated by ‘chain-of-spheres’ approximation (COSX). The RI-MP2 method requires using an auxiliary basis sets. The Ahlrichs Def2-SVP/J and Def2-SVP/C auxiliary basis sets were selected as implemented in Orca. The natural orbitals were used as the guess for the CASSCF method. Twelve orbitals containing twelve electrons (i.e., CASSCF(12,12)) were selected for the active space (see Supporting Information). The CASSCF calculations were run with the Def2-TVZPP basis set.[48] The calculations were accelerated using the RI and RIJCOSX approximations. These approximations require using auxiliary basis sets. The Def2-TZVPP/C and Def2-TZVPP/J auxiliary basis sets were applied as implemented in Orca. The N-electron valence state perturbation theory (NEVPT2) method[52] was used in addition to CASSCF to describe dynamic correlation with the same basis sets. Natural bond order (NBO) analysis of the complexes was performed by the NBO program version 6.0.[53]

4411

 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper Acknowledgements This research was supported by the European Research Council under the FP7 framework (ERC No. 246837) and by the MINERVA Foundation. T.Z. received a postdoctoral fellowship from the MINERVA Foundation, P.M. received a postdoctoral fellowship from the Feinberg Graduate School of the Weizmann Institute of Science, and D.M. holds the Israel Matz Professorial Chair of Organic Chemistry.

[13]

[14] [15]

[16]

Keywords: bipyridine · iron · pincer ligands · redox chemistry · pi interactions

[1] a) K. Hindson, B. de Bruin, Eur. J. Inorg. Chem. 2012, 340 – 342; b) H. Grtzmacher, Angew. Chem. 2008, 120, 1838 – 1842; Angew. Chem. Int. Ed. 2008, 47, 1814 – 1818. [2] a) W. I. Dzik, J. I. van der Vlugt, J. N. H. Reek, B. de Bruin, Angew. Chem. 2011, 123, 3416 – 3418; Angew. Chem. Int. Ed. 2011, 50, 3356 – 3358; b) W. Kaim, Inorg. Chem. 2011, 50, 9752 – 9765; c) J. I. van der Vlugt, Eur. J. Inorg. Chem. 2012, 363 – 375; d) P. J. Chirik, K. Wieghardt, Science 2010, 327, 794 – 795; e) O. R. Luca, R. H. Crabtree, Chem. Soc. Rev. 2013, 42, 1440 – 1459; f) V. K. K. Praneeth, M. R. Ringenberg, T. R. Ward, Angew. Chem. Int. Ed. 2012, 51, 10228 – 10234; g) K. G. Caulton, Eur. J. Inorg. Chem. 2012, 435 – 443; h) W. Kaim, Eur. J. Inorg. Chem. 2012, 343 – 348; i) S. Blanchard, E. Derat, M. Desage-El Murr, L. Fensterbank, M. Malacria, V. Mouris-Mansuy, Eur. J. Inorg. Chem. 2012, 376 – 389; j) K. Ray, T. Petrenko, K. Wieghardt, F. Neese, Dalton Trans. 2007, 1552 – 1566; k) D. Sieh, M. Schlimm, L. Andernach, F. Angersbach, S. Nckel, J. Schçffel, N. Sˇusˇnjar, P. Burger, Eur. J. Inorg. Chem. 2012, 444 – 462; l) V. Lyaskovskyy, B. de Bruin, ACS Catal. 2012, 2, 270 – 279; m) A. I. O. Suarez, V. Lyaskovskyy, J. N. H. Reek, J. I. van der Vlugt, B. de Bruin, Angew. Chem. 2013, 125, 12740 – 12760; Angew. Chem. Int. Ed. 2013, 52, 12510 – 12529. [3] a) B. L. Small, M. Brookhart, A. M. A. Bennett, J. Am. Chem. Soc. 1998, 120, 4049 – 4050; b) B. L. Small, M. Brookhart, J. Am. Chem. Soc. 1998, 120, 7143 – 7144. [4] a) G. J. P. Britovsek, V. C. Gibson, S. J. McTavish, G. A. Solan, A. J. P. White, D. J. Williams, G. J. P. Britovsek, B. S. Kimberley, P. J. Maddox, Chem. Commun. 1998, 849 – 850; b) G. J. P. Britovsek, M. Bruce, V. C. Gibson, B. S. Kimberley, P. J. Maddox, S. Mastroianni, S. J. McTavish, C. Redshaw, G. A. Solan, S. Strçmberg, A. J. P. White, D. J. Williams, J. Am. Chem. Soc. 1999, 121, 8728 – 8740. [5] a) S. C. Bart, E. Lobkovsky, P. J. Chirik, J. Am. Chem. Soc. 2004, 126, 13794 – 13807; b) R. J. Trovitch, E. Lobkovsky, E. Bill, P. J. Chirik, Organometallics 2008, 27, 1470 – 1478; c) S. K. Russell, J. M. Darmon, E. Lobkovsky, P. J. Chirik, Inorg. Chem. 2010, 49, 2782 – 2792; d) S. C. Bart, E. Lobkovsky, E. Bill, P. J. Chirik, J. Am. Chem. Soc. 2006, 128, 5302 – 5303. [6] a) A. M. Tondreau, E. Lobkovsky, P. J. Chirik, Org. Lett. 2008, 10, 2789 – 2792; b) A. M. Tondreau, J. M. Darmon, B. M. Wile, S. K. Floyd, E. Lobkovsky, P. J. Chirik, Organometallics 2009, 28, 3928 – 3940; c) A. M. Tondreau, C. C. H. Atienza, J. M. Darmon, C. Milsmann, H. M. Hoyt, K. J. Weller, S. A. Nye, K. M. Lewis, J. Boyer, J. G. P. Delis, E. Lobkovsky, P. J. Chirik, Organometallics 2012, 31, 4886 – 4893; d) S. K. Russell, C. Milsmann, E. Lobkovsky, T. Weyhermller, P. J. Chirik, Inorg. Chem. 2011, 50, 3159 – 3169; e) A. M. Tondreau, C. C. H. Atienza, K. J. Weller, S. A. Nye, K. M. Lewis, J. G. P. Delis, P. J. Chirik, Science 2012, 335, 567 – 570. [7] J. V. Obligacion, P. J. Chirik, Org. Lett. 2013, 15, 2680 – 2683. [8] a) K. T. Sylvester, P. J. Chirik, J. Am. Chem. Soc. 2009, 131, 8772 – 8774; b) J. M. Hoyt, K. T. Sylvester, S. P. Semproni, P. J. Chirik, J. Am. Chem. Soc. 2013, 135, 4862 – 4877. [9] J. Y. Wu, B. N. Stanzl, T. Ritter, J. Am. Chem. Soc. 2010, 132, 13214 – 13216. [10] J. Y. Wu, B. t. Moreau, T. Ritter, J. Am. Chem. Soc. 2009, 131, 12915 – 12917. [11] B. t. Moreau, J. Y. Wu, T. Ritter, Org. Lett. 2009, 11, 337 – 339. [12] J. Raynaud, J. Y. Wu, T. Ritter, Angew. Chem. 2012, 124, 11975 – 11978; Angew. Chem. Int. Ed. 2012, 51, 11805 – 11808; corrigendum: J. Raynaud, Chem. Eur. J. 2014, 20, 4403 – 4413

www.chemeurj.org

[17]

[18] [19] [20]

[21]

[22] [23] [24] [25] [26]

[27] [28] [29] [30]

[31]

4412

J. Y. Wu, T. Ritter, Angew. Chem. 2012, 124, 13095; Angew. Chem. Int. Ed. 2012, 51, 12920. a) P. L. Holland, Dalton Trans. 2010, 39, 5415 – 5425; b) S. N. Brown, Inorg. Chem. 2012, 51, 1251 – 1260; c) A. Das, T. M. Scherer, S. M. Mobin, W. Kaim, G. K. Lahiri, Chem. Eur. J. 2012, 18, 11007 – 11018; d) P. Chaudhuri, C. N. Verani, E. Bill, E. Bothe, T. Weyhermller, K. Wieghardt, J. Am. Chem. Soc. 2001, 123, 2213 – 2223. C. C. Scarborough, K. Wieghardt, Inorg. Chem. 2011, 50, 9773 – 9793. a) C. Gunanathan, D. Milstein, Top. Organomet. Chem. 2011, 37, 55 – 84; b) D. Milstein, Top. Catal. 2010, 53, 915 – 923; c) C. Gunanathan, D. Milstein, Science 2013, 341, 1229712. a) E. Balaraman, C. Gunanathan, J. Zhang, L. J. W. Shimon, D. Milstein, Nat. Chem. 2011, 3, 609 – 614; b) E. Balaraman, Y. Ben-David, D. Milstein, Angew. Chem. 2011, 123, 11906 – 11909; Angew. Chem. Int. Ed. 2011, 50, 11702 – 11705; c) E. Balaraman, B. Gnanaprakasam, L. J. W. Shimon, D. Milstein, J. Am. Chem. Soc. 2010, 132, 16756 – 16758; d) D. Srimani, M. Feller, Y. Ben-David, D. Milstein, Chem. Commun. 2012, 48, 11853 – 11855; e) E. Balaraman, E. Fogler, D. Milstein, Chem. Commun. 2012, 48, 1111 – 1113. a) D. Srimani, E. Balaraman, B. Gnanaprakasam, Y. Ben-David, D. Milstein, Adv. Synth. Catal. 2012, 354, 2403 – 2406; b) D. Srimani, Y. Ben-David, D. Milstein, Angew. Chem. 2013, 125, 4104 – 4107; Angew. Chem. Int. Ed. 2013, 52, 4012 – 4015; c) D. Srimani, Y. Ben-David, D. Milstein, Chem. Commun. 2013, 49, 6632 – 6634. E. Balaraman, E. Khaskin, G. Leitus, D. Milstein, Nat. Chem. 2013, 5, 122 – 125. E. Khaskin, D. Milstein, ACS Catal. 2013, 3, 448 – 452. a) S. Enthaler, K. Junge, M. Beller, Angew. Chem. 2008, 120, 3363 – 3367; Angew. Chem. Int. Ed. 2008, 47, 3317 – 3321; b) K. Gopalaiah, Chem. Rev. 2013, 113, 3248 – 3296; c) M. Darwish, M. Wills, Catal. Sci. Technol. 2012, 2, 243 – 255; d) K. Junge, K. Schrçder, M. Beller, Chem. Commun. 2011, 47, 4849 – 4859; e) S. Gaillard, J.-L. Renaud, ChemSusChem 2008, 1, 505 – 509; f) R. H. Morris, Chem. Soc. Rev. 2009, 38, 2282 – 2291; g) S. Chakraborty, H. Guan, Dalton Trans. 2010, 39, 7427 – 7436; h) H. Nakazawa, M. Itazaki, in Top. Organomet. Chem. 2011, 33, 27 – 81. R. Langer, Y. Diskin-Posner, G. Leitus, L. J. W. Shimon, Y. Ben-David, D. Milstein, Angew. Chem. 2011, 123, 10122 – 10126; Angew. Chem. Int. Ed. 2011, 50, 9948 – 9952. T. Zell, B. Butschke, Y. Ben-David, D. Milstein, Chem. Eur. J. 2013, 19, 8068 – 8072. R. Langer, G. Leitus, Y. Ben-David, D. Milstein, Angew. Chem. 2011, 123, 2168 – 2172; Angew. Chem. Int. Ed. 2011, 50, 2120 – 2124. R. Langer, M. A. Iron, L. Konstantinovski, Y. Diskin-Posner, G. Leitus, Y. Ben-David, D. Milstein, Chem. Eur. J. 2012, 18, 7196 – 7209. D. Srimani, Y. Diskin-Posner, Y. Ben-David, D. Milstein, Angew. Chem. 2013, 125, 14381 – 14384; Angew. Chem. Int. Ed. 2013, 52, 14131 – 14134. T. Zell, R. Langer, M. A. Iron, L. Konstantinovski, L. J. W. Shimon, Y. Diskin-Posner, G. Leitus, E. Balaraman, Y. Ben-David, D. Milstein, Inorg. Chem. 2013, 52, 9636 – 9649. L. Zhang, D. Peng, X. Leng, Z. Huang, Angew. Chem. 2013, 125, 3764 – 3768; Angew. Chem. Int. Ed. 2013, 52, 3676 – 3680. D. F. Evans, J. Chem. Soc. 1959, 2003 – 2005. E. M. Pelczar, T. J. Emge, K. Krogh-Jespersen, A. S. Goldman, Organometallics 2008, 27, 5759 – 5767. For an excellent overview on iron(0) complexes and their geometries, ranging from trigonal bipyramidal to square pyramidal, see reference [5a]. Note that the complex described there was later described as an iron(II) complex. a) M. Wang, J. England, T. Weyhermuller, S. L. Kokatam, C. J. Pollock, S. DeBeer, J. Shen, G. P. Yap, K. H. Theopold, K. Wieghardt, Inorg. Chem. 2013, 52, 4472 – 4487; b) C. C. Scarborough, S. Sproules, T. Weyhermller, S. DeBeer, K. Wieghardt, Inorg. Chem. 2011, 50, 12446 – 12462; c) M. Irwin, L. R. Doyle, T. Krmer, R. Herchel, J. E. McGrady, J. M. Goicoechea, Inorg. Chem. 2012, 51, 12301 – 12312; d) C. C. Scarborough, S. Sproules, C. J. Doonan, K. S. Hagen, T. Weyhermller, K. Wieghardt, Inorg. Chem. 2012, 51, 6969 – 6982; e) A. C. Bowman, J. England, S. Sproules, T. Weyhermller, K. Wieghardt, Inorg. Chem. 2013, 52, 2242 – 2256; f) J. England, K. Wieghardt, Inorg. Chem. 2013, 52, 10067 – 10079; g) M. Wang, T. Weyhermuller, J. England, K. Wieghardt, Inorg. Chem. 2013, 52, 12763 – 12776.  2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper [32] M. H. Chisholm, J. C. Huffman, I. P. Rothwell, P. G. Bradley, N. Kress, W. H. Woodruff, J. Am. Chem. Soc. 1981, 103, 4945 – 4947. [33] E. Gore-Randall, M. Irwin, M. S. Denning, J. M. Goicoechea, Inorg. Chem. 2009, 48, 8304 – 8316. [34] H. Bock, J.-M. Lehn, J. Pauls, S. Holl, V. Krenzel, Angew. Chem. 1999, 111, 1004 – 1008; Angew. Chem. Int. Ed. 1999, 38, 952 – 955. [35] D. Benito-Garagorri, L. G. a. Alves, M. Puchberger, K. Mereiter, L. F. Veiros, M. J. Calhorda, M. D. Carvalho, L. P. Ferreira, M. Godinho, K. Kirchner, Organometallics 2009, 28, 6902 – 6914. [36] a) P. Brant, R. D. Feltham, J. Electron Spectrosc. Relat. Phenom. 1983, 32, 205 – 221; b) A. P. Grosvenor, B. A. Kobe, M. C. Biesinger, N. S. McIntyre, Surf. Interface Anal. 2004, 36, 1564 – 1574. [37] D. Briggs, P. Seah, Practical Surface Analysis: Auger and X-ray Photoelectron Spectroscopy, Wiley, Chichester (UK), 1990. [38] a) P. Brant, R. D. Feltham, Inorg. Chem. 1980, 19, 2673 – 2676; b) R. D. Feltham, P. Brant, J. Am. Chem. Soc. 1982, 104, 641 – 645; c) J. Blomquist, P. Kjll, B. Folkesson, R. Larsson, Struct. Chem. 1991, 2, 223 – 229; d) J. Blomquist, U. Helgeson, B. Folkesson, R. Larsson, Chem. Phys. 1983, 76, 71 – 78. [39] For a comprehensive study evaluating the use of different methods for the description of bipyridine ligands in potentially redox non-innocent metal complexes see: P. Milko, M. A. Iron, J. Chem. Theory Comput. 2014, 10, 220 – 235. [40] G. Beamson, D. Briggs, High-Resolution XPS of Organic Polymers: The Scienta ESCA 300 Database, Wiley, Chichester (UK), 1992. [41] Gaussian 09, Revision C.01, M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria, M. A. Robb, J. R. Cheeseman, G. Scalmani, V. Barone, B. Mennucci, G. A. Petersson, H. Nakatsuji, M. Caricato, X. Li, H. P. Hratchian, A. F. Izmaylov, J. Bloino, G. Zheng, J. L. Sonnenberg, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, T. Vreven, J. A. Montgomery, J. E. Peralta, F. Ogliaro, M. Bearpark, J. J. Heyd, E. Brothers, K. N. Kudin, V. N. Staroverov, R. Kobaya-

Chem. Eur. J. 2014, 20, 4403 – 4413

www.chemeurj.org

[42] [43] [44] [45] [46] [47] [48] [49] [50]

[51] [52] [53]

shi, J. Normand, K. Raghavachari, A. Rendell, J. C. Burant, S. S. Iyengar, J. Tomasi, M. Cossi, N. Rega, J. M. Millam, M. Klene, J. E. Knox, J. B. Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R. E. Stratmann, O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. W. Ochterski, R. L. Martin, K. Morokuma, V. G. Zakrzewski, G. A. Voth, P. Salvador, J. J. Dannenberg, S. Dapprich, A. D. Daniels, Farkas, J. B. Foresman, J. V. Ortiz, J. Cioslowski, D. J. Fox, Wallingford CT, 2009. Y. Zhao, D. Truhlar, Theor. Chem. Acc. 2008, 120, 215 – 241. S. Grimme, J. Comput. Chem. 2006, 27, 1787 – 1799. C. Gunanathan, B. Gnanaprakasam, M. A. Iron, L. J. W. Shimon, D. Milstein, J. Am. Chem. Soc. 2010, 132, 14763 – 14765. T. H. Dunning Jr., P. J. Hay, in Modern Theoretical Chemistry, Vol. 3 (Ed.: H. F. Schaefer III), Plenum, New York, 1976, pp. 1 – 28. M. Dolg, H. Stoll, H. Preuss, Theor. Chim. Acta 1993, 85, 441 – 450. F. Neese, WIREs Comput. Mol. Sci. 2012, 2, 73 – 78. C. Møller, M. S. Plesset, Phys. Rev. 1934, 46, 618 – 622. F. Weigend, R. Ahlrichs, Phys. Chem. Chem. Phys. 2005, 7, 3297 – 3305. a) F. Neese, F. Wennmohs, A. Hansen, U. Becker, Chem. Phys. 2009, 356, 98 – 109; b) S. Kossmann, F. Neese, Chem. Phys. Lett. 2009, 481, 240 – 243; c) R. Izsak, F. Neese, J. Chem. Phys. 2011, 135, 144105. F. Neese, J. Comput. Chem. 2003, 24, 1740 – 1747, and references therein. C. Angeli, R. Cimiraglia, J.-P. Malrieu, J. Chem. Phys. 2002, 117, 9138 – 9153, and references therein. E. D. Glendening, J. K. Badenhoop, A. E. Reed, J. E. Carpenter, J. A. Bohmann, C. M. Morales, C. R. Landis, F. Weinhold, Theoretical Chemistry Institute, University of Wisconsin, Madison, 2013.

Received: November 26, 2013 Published online on March 3, 2014

4413

 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim