A combined magnetic circular dichroism and density functional theory approach for the elucidation of electronic structure and bonding in three- and four-coordinate iron(ii)–N-heterocyclic carbene complexes

Studies of electronic structure and bonding in iron(ii)–NHC complexes using a combined magnetic circular dichroism and DFT approach.


Introduction
During the last few decades, transition metal catalysts utilizing precious metals (i.e. Pd, Pt, Rh, Ir, Au, Ru) combined with NHC ligands have been successfully developed and employed for a variety of catalytic transformations including metathesis, C-H activation, hydrogenation and C-C bond formation reactions amongst many others. [1][2][3] While the coordination chemistry and catalysis of late transition metal-NHC complexes have been widely explored, the development and application of iron-NHCs has only recently begun to be widely investigated despite the fact that iron-NHC complexes have been known since the early 1970's. [4][5][6] Within the last decade, numerous reports of iron-NHC complexes with unique structures and novel catalytic applications have been reported. [7][8][9][10][11][12][13][14][15][16][17][18][19][20][21][22] A wide variety of applications of iron-NHC complexes to catalytic transformations have begun to emerge, including applications in hydrosilylation, carbometallation, cyclization, aziridination and allylic substitution reactions, demonstrating the signicant catalytic potential of iron-NHC complexes. [7][8][9] Of particular note, the combination of simple iron salts and NHCs results in the in situ generation of catalytically active iron-NHC species for C-C cross-coupling reactions including aryl-aryl, alkyl-aryl and alkyl-alkyl couplings (Scheme 1). NHC ligands have been particularly successful for iron cross-coupling as the aryl-aryl system is currently the best performing iron-based system for heteroaryl coupling and the alkyl-alkyl system is currently the only iron-based cross-coupling system capable of C(sp 3 )-C(sp 3 ) couplings in the presence of functional groups.
Within the catalytic applications of iron-NHCs, the nature of the specic NHC ligand that provides for the most effective catalysis varies signicantly from one system to another. For example, the optimal NHC for aryl-aryl cross-coupling catalysis is SIPr where the corresponding, unsaturated IPr ligand was found to be much less effective (Scheme 1). 23 By contrast, for the cross-coupling of non-activated chloroalkanes and aryl Grignards in Kumada-type couplings, the inverse of the NHC dependence on activity is observed with IPr outperforming SIPr. 24 However, these dramatic disparities in reactivity due to differences in NHC backbone saturation are not observed in alkyl-alkyl cross-coupling where similar product yields can be obtained using either IMes or SIMes. 25 In the aryl-alkyl crosscoupling system of Bedford and co-workers, NHCs lacking Naryl substituents (Cy, tBu) outperformed IMes. 26 Combined, these studies suggest that the NHC ring structure (i.e. saturated vs. unsaturated vs. substituted) and N-substitution may yield important differences in iron-NHC bonding, in situ iron-NHC speciation and, hence, iron-NHC reactivity.
Despite the observed dependence of catalytic performance on the NHC ligand structure, a detailed understanding of iron-NHC sand p-bonding and the effects of NHC ring perturbations on iron-NHC bonding is critically underdeveloped. In fact, detailed investigations of iron-NHC bonding have been limited to piano stool type iron(II) complexes with both cyclopentadienyl (Cp) and CO ligation where a combination of IR, electrochemical and theoretical methods suggested that the NHC ligand in these complexes can serve as both a s-donor and moderate p-acceptor. 27 However, fundamental insight into iron-NHC bonding in more electron decient, high-spin iron complexes is lacking. This deciency stands in stark contrast to precious metal NHC systems where IR studies of supporting CO ligands and DFT investigations have led to the general view of NHC ligands as strong s-donors (stronger than phosphine ligands) and weak p-acceptors, where the extent of p-bonding is dependent on the nature of the metal and supporting ligands. [28][29][30][31][32][33][34][35][36][37][38][39][40][41][42][43] While the general views of metal-NHC bonding from the precious metal systems currently drive much of the work in iron-NHC systems, rational catalyst development with iron-NHCs necessitates a fundamental understanding of the effects of NHC variations on electronic structure and bonding in paramagnetic iron systems that may be catalytically relevant. Such studies can also provide fundamental insight into the differences in electronic structures of iron-NHC and ironphosphine complexes lacking CO ligation that may be relevant to catalysis, including iron-catalyzed cross-coupling. Importantly, iron-NHC bonding in high-spin systems may differ signicantly compared to the low-spin, Cp and CO bound species more commonly investigated.
Towards this goal, an approach combining magnetic circular dichroism (MCD) studies and density functional theory (DFT) investigations of well-dened iron(II)-NHC complexes has been utilized to directly investigate electronic structure and bonding in high-spin iron(II)-NHC complexes. The results provide direct insight into the ligand-eld strength of NHC ligands compared to amine and phosphine ligands, the effects of NHC ring variations on bonding and the extent of donation and back donation contributions to bonding in iron(II)-NHC complexes as a function of coordination number and geometry.

Mössbauer spectroscopy
All solid samples for 57 Fe Mössbauer spectroscopy were run on non-enriched samples of the as-isolated complexes. All samples were prepared in an inert atmosphere glove box equipped with a liquid nitrogen ll port to enable sample freezing to 77 K within the glove box. Each sample was loaded into a Delrin Mössbauer sample cup for measurements and loaded under liquid nitrogen. Low temperature 57 Fe Mössbauer measurements were performed using a See Co. MS4 Mössbauer spectrometer integrated with a Janis SVT-400 He/N 2 cryostat for measurements at 80 K with a 0.07 T applied magnetic eld. Isomer shis were determined relative to a-Fe at 298 K. All Mössbauer spectra were t using the program WMoss (SeeCo).

Magnetic circular dichroism spectroscopy
All samples for MCD spectroscopy were prepared in an inert atmosphere glove box equipped with a liquid nitrogen ll port to enable sample freezing to 77 K within the glove box. MCD samples were prepared in 6 : 1 (v/v) toluene-d 8 : benzene-d 6 (to form low temperature optical glasses) in copper cells tted with quartz disks and a 3 mm gasket. Low temperature MCD experiments were conducted using two Jasco spectropolarimeters. Both instruments utilize a modied sample compartment incorporating focusing optics and an Oxford Instruments SM4000-7T superconducting magnet/cryostat. This set-up permits measurements from 1.6 K to 290 K with magnetic elds up to 7 T. A calibrated Cernox sensor directly inserted in the copper sample holder is used to measure the temperature at the sample to AE0.001 K. UV-visible (UV-vis) MCD spectra were collected using a Jasco J-715 spectropolarimeter and a shielded S-20 photomultiplier tube. Near-infrared (NIR) MCD spectra were collected using a Jasco J-730 spectropolarimeter with a liquid nitrogen cooled InSb detector. The spectral range accessible with this NIR MCD setup is 2000-600 nm. All MCD spectra were baseline-corrected against zero-eld scans. VTVH-MCD spectra were analyzed using previously reported tting procedures. 48 For VTVH-MCD tting, both negative and positive zero-eld splitting models were evaluated. The reported error bars were determined via evaluation of the effects of systematic variations of the t parameters on the quality of the overall t. D and |E/D| values are obtained directly from the t parameters using the relationships E ¼ (d/6) + 1/3 [(d 2 /2) + dE s ] 1/2 and ÀD ¼ E + (E s /3) À (d/6) for S ¼ 2 as previously described. 48 Electronic structure calculations Spin unrestricted DFT calculations were performed with the Gaussian 09 package. 49 All geometry optimization calculations were performed with the B3LYP exchange-correlation functional 50,51 with the TZVP 52 basis set on all atoms and inclusion of solvation effects using the Polarized Continuum Model (PCM) with toluene as the solvent. 53 The dispersion correction of Grimme (GD3) combined with the damping function of Becke and Johnson (BJ) was used in geometry optimizations of all four-coordinate complexes. 54 The geometries of all complexes were fully optimized starting from X-ray crystal structures (when available) with initial optimization performed with cep-4g before optimizing at the TZVP level. All optimized geometries had frequencies found to be positive.
Further calculations of molecular orbitals (MOs) and TD-DFT used the B3LYP functional with the TZVP basis set on all atoms. MO compositions and analyses were calculated using the AOMix program. 55,56 Atomic charges and spin densities were calculated using Mulliken population analysis (MPA). Orbitals from the Gaussian calculations were plotted with the Chem-Cra program. TD-DFT was used to calculate the electronic transition energies and intensities of the 30-40 lowest-energy states. The analysis of the MO compositions in terms of fragment orbitals, Mayer bond orders, total overlap populations and the charge decomposition analysis (CDA) 57,58 were performed using AOMix-FO. 55 CDA and its applications have been previously described in detail by Gorelsky and co-workers. 59,60 Results and discussion Spectroscopic and electronic structure studies of (IMes) 2 FeCl 2 Initial studies focused on (IMes) 2 FeCl 2 as a representative example of distorted tetrahedral (NHC) 2 FeX 2 complexes which have been widely explored synthetically and as potential iron(II) pre-catalysts for a variety of reactions (Scheme 2). The 80 K 57 Fe Mössbauer spectrum of polycrystalline (IMes) 2 FeCl 2 ( Fig. 1A) is well-t as a single iron species with d ¼ 0.80 mm s À1 and DE Q ¼ 2.12 mm s À1 , where the observed isomer shi falls within the expected range for high-spin iron(II), S ¼ 2 distorted tetrahedral species. 61 The 5 K, 7 T NIR MCD spectrum of (IMes) 2 FeCl 2 contains two low-energy ligand-eld (LF) transitions at 5440 cm À1 and 6520 cm À1 (10Dq(T d ) ¼ 5980 cm À1 ) (Fig. 1B). For a tetrahedral S ¼ 2 iron(II) complex, only the 5 E / 5 T 2 transition is spin allowed. For the distorted tetrahedral environment as is present in (IMes) 2 FeCl 2 , the degeneracy of both the ground and excited states is removed, resulting in two LF transitions as observed by MCD spectroscopy. The saturation magnetization behavior for (IMes) 2 FeCl 2 collected at 5917 cm À1 is welldescribed by a S ¼ 2 negative zero-eld split (ÀZFS) non-Kramers doublet model with ground-state spin-Hamiltonian (Fig. 1B, inset). The 5 K, 7 T UV-vis MCD spectrum of (IMes) 2 FeCl 2 contains multiple charge transfer (CT) transitions in the 30 000-35 000 cm À1 Scheme 2 Distorted tetrahedral (NHC) 2 FeCl 2 complexes with IMes and Cl IMes ligands. region ( Fig. 1C) which are assigned and discussed using TD-DFT calculations in the ESI. † Spin unrestricted DFT calculations were used to further analyze the electronic structure of (IMes) 2 FeCl 2 . Geometry optimization with B3LYP/TZVP and the GD3BJ dispersion correction yielded overall structural features, bond lengths and angles in good agreement with those observed by crystallography ( Table 1). The optimized (IMes) 2 FeCl 2 complex is best described as a distorted tetrahedral complex with Fe-IMes bond lengths of 2.161Å and 2.163Å, Fe-Cl bond lengths of 2.334Å, a C(IMes)-Fe-C(IMes) bond angle of 126.36 and a Cl-Fe-Cl bond angle of 107.45 . This optimized geometry correlates well with the literature structure, which has Fe-IMes bond lengths of 2.139Å and 2.157Å, Fe-Cl bond lengths of 2.310Å and 2.292Å, a C(IMes)-Fe-C(IMes) bond angle of 125.20 and a Cl-Fe-Cl bond angle of 106.66 . Both the experimental and computational studies of (IMes) 2 FeCl 2 are indicative of a high-spin iron(II) complex (S ¼ 2). The molecular orbitals and their corresponding energies as well as electronic transition energies were calculated from the optimized structure.
The electronic ground state of (IMes) 2 FeCl 2 is described by the frontier molecular orbitals (FMOs) with focus on the unoccupied MOs, in conjunction with their occupied counterparts, to showcase the major contributions to bonding. The MOs and the corresponding energy diagram are shown in Fig. 2 Charge decomposition (CDA), fragment molecular orbital (FO) and Mayer bond order (MBO) analyses were completed for the (IMes) 2 FeCl 2 complex. The MOs of a complex can be described as linear combinations of the occupied and unoccupied MOs of dened molecular fragments, termed fragment molecular orbitals or FOs. In CDA, donation and back donation between different molecular fragments can be evaluated based on the overlap and coefficients of the MO-LCFO matrix (where LCFO ¼ linear combination of fragment molecular orbitals). In the case of (IMes) 2 FeCl 2 , CDA was completed to quantify the total charge donation and back donation between the Fe-Cl 2 fragment and two IMes ligand fragments (i.e. 3 total fragments). When the Fe-Cl 2 fragment and IMes ligand fragments are combined to form the complex, there is a mixing of the occupied fragment orbital of the donor with the unoccupied fragment orbital of the acceptor which allows for the transfer of electron density (charge donation) from the donor to the acceptor. From the CDA, it can be seen that there is a total charge donation (IMes 2 / Fe-Cl 2 ) of 0.954 electrons and total back donation (Fe-Cl 2 / IMes 2 ) of 0.391 electrons. This results in a net charge donation to Fe-Cl 2 of 0.563 electrons. The FO analysis provides information about changes in occupancies of the fragment molecular orbitals of the donor and acceptor upon (B, inset) VTVH-MCD data (dots) and fit (lines) of (IMes) 2 FeCl 2 collected at 5917 cm À1 . MCD spectra were collected on a 3 mM solution of (IMes) 2 FeCl 2 in 6 : 1 (toluene-d 8 : benzene-d 6 ). Peak fits are shown for the MCD spectra (dashed lines). manifold, there is also a signicant electronic polarization, particularly in the Fe-Cl 2 fragment, which allows for a redistribution of charge from an occupied Fe-Cl 2 fragment orbital with mostly d character to unoccupied fragment orbitals of the Fe-Cl 2 fragment (see Fig. 3). Furthermore, the FO analysis indicates the presence of back donation in both the a and b manifolds from occupied Fe-Cl 2 fragment orbitals to unoccupied IMes fragment orbitals (i.e. $9.6% and $10.1% occupancy changes for each IMes fragment in both the a and b manifolds, respectively). However, in contrast to the donation in this system, the occupancy changes corresponding to back donation are spread over many occupied Fe-Cl 2 and unoccupied IMes FOs. For this distorted tetrahedral complex, the Mayer bond order (MBO) between Fe and each IMes is 0.760 and 0.767. In both cases, the b contribution to the bond order is greater than that of the a, 0.390 and 0.393 for b and 0.370 and 0.374 for a. When compared to the occupied a-spin MOs, the occupied b-spin MOs contribute more to the overall bond order due to a small increase in s donation from the NHCs to the unoccupied b-spin orbitals on Fe.
Evaluation of the relative ligand eld strengths of NHC ligands in L 2 FeCl 2 distorted tetrahedral iron(II) complexes A series of distorted tetrahedral iron(II) complexes containing two chloride ligands and two additional ligands (e.g. diamine, two monodentate phosphines or NHCs) were investigated by NIR MCD spectroscopy in order to evaluate the relative LF strength of NHC ligands in distorted tetrahedral iron(II) complexes. Such an evaluation is important as it has been previously suggested that NHCs may be stronger eld ligands than phosphines and, hence, may be able to serve as   Table 2. The diamineiron(II)-dichloride complex (tmpn)FeCl 2 (tmpn ¼ N,N,N 0 ,N 0 -tetramethylpropane-1,3-diamine) contains very low energy LF transitions at 5260 and 6140 cm À1 (10Dq(T d ) ¼ 5700 cm À1 ). Similarly, (teeda)FeCl 2 (teeda ¼ N,N,N 0 ,N 0 -tetraethylethylenediamine) gives LF transitions at 5290 and 6310 cm À1 (10Dq(T d ) ¼ 5800 cm À1 ). Both NHC ligated tetrahedra gave slightly higher energy LF bands, where ( Cl IMes) 2 FeCl 2 exhibited LF transitions at 5520 cm À1 and 6540 cm À1 (10Dq(T d ) ¼ 6030 cm À1 ), similar to those observed for (IMes) 2 FeCl 2 (10Dq(T d ) ¼ 5980 cm À1 ). The phosphine-iron(II)-dichloride complexes yielded the highest energy ligand-eld transitions, with (PPh 3 ) 2 FeCl 2 exhibiting LF bands at 5590 cm À1 and 7590 cm À1 (10Dq(T d ) ¼ 6590 cm À1 ) and (PMe 3 ) 2 FeCl 2 at 6340 cm À1 and 7600 cm À1 (10Dq(T d ) ¼ 6970 cm À1 ). In contrast to the NHC and diamine complexes, both phosphine species exhibit a pseudo-A term in their LF MCD transitions, where such pseudo-A terms in MCD (i.e. a pair of temperature-dependent C-terms with opposite sign) arise from spin orbit coupling (SOC) between two excited states that are close in energy to which two orthogonal transitions occur from a single ground state. [64][65][66] While the physical origin of the pseudo-A term is currently elusive in the phosphine complexes as it requires a detailed understanding of the SOC mechanism, it is noteworthy that previous MCD studies of distorted tetrahedral bisphosphine complexes have also exhibited pseudo-A term LF transitions 67 and the generality and origin of this behavior for tetrahedral phosphine complexes will be a focus of future study.
To obtain further insight into iron(II)-NHC bonding compared to phosphine and amine ligands, CDA and MBO analyses were completed for the L 2 FeCl 2 complexes. The MBO between Fe and the NHCs in (IMes) 2 FeCl 2 is the highest of the series (see Table 3). Notably, the MBOs between Fe and PPh 3 or PMe 3 are slightly smaller. Lower still are the MBOs for the Fe-diamine bonds in (tmpn)FeCl 2 and (teeda)FeCl 2 , 0.362 and 0.362 for (tmpn)FeCl 2 and 0.352 and 0.336 for (teeda)FeCl 2 . However, the MBO between Fe and Cl is much higher in both the phosphine and diamine complexes when compared to the NHC complex. This is indicative of a strong trans-type inuence of the NHC ligand that causes a pronounced weakening of the Fe-Cl bond. This inuence can be seen experimentally in the (IMes) 2 FeCl 2 crystal structure where the Fe-Cl bonds are elongated (avg Fe-Cl ¼ 2.30Å) in comparison to the analogous bisphosphine (avg Fe-Cl ¼ 2.23Å) and diamine (avg Fe-Cl ¼ 2.26Å) complexes (see Table 1). The calculated total overlap population (TOP) between the Fe and Cl moieties in (IMes) 2 FeCl 2 (avg Fe-Cl ¼ 0.39) is signicantly lower than those calculated for the bisphosphine and diamine complexes (avg Fe-Cl ¼ 0.49 and 0.50, respectively). The amount of electron density between the Fe and Cl atoms in each complex displays an inverse relationship with the amount of electron density donated by each L-type ligand (Table 3). This correlation suggests that the strongest donor to the distorted tetrahedral Fe(II) center will cause the most pronounced weakening of the Fe-Cl bond. This rationalization using TOP values is consistent with the perturbation approach of Burdett and Albright, 68 which shows that systems of lowered symmetry can, through orbital mixing, exhibit metal-ligand bond weakening effects similar to the trans inuence typically seen in square-planar and octahedral complexes. The results of the CDA show that the complex with the highest net charge donation to the Fe-Cl 2 fragment from the L-type ligand fragments is (IMes) 2 FeCl 2 with a net   70 The NIR MCD spectrum of (IPr)Fe(CH 2 TMS) 2 contains two LF transitions at 6660 cm À1 and 9260 cm À1 (Fig. 5A). The saturation magnetization behavior for (IPr)Fe(CH 2 TMS) 2 collected at 9852 cm À1 is well-described by a S ¼ 2 ÀZFS non-Kramers doublet model  with ground-state spin-Hamiltonian parameters of d ¼ 2.1 AE 0.2 cm À1 and g ll ¼ 9.9 AE 0.2 with D ¼ À20 AE 2 cm À1 and |E/D| ¼ 0.20 AE 0.02 (Fig. 5A, inset). The observed D value is similar to those previously determined for (IPr)Fe{N(SiMe 3 ) 2 } 2 (D ¼ À18.2 cm À1 ) and (IMes)Fe{N(SiMe 3 ) 2 } 2 (D ¼ À23.3 cm À1 ) by Layeld and coworkers. 71 The 5 K, 7 T NIR MCD spectra of (SIPr)Fe(CH 2 TMS) 2 ( Fig. 5B) and ( Cl IPr)Fe(CH 2 TMS) 2 (Fig. 5C) yielded similar LF transitions and ZFS parameters to (IPr)Fe(CH 2 TMS) 2 (Table 4). Thus, NIR MCD spectroscopy supports the presence of similar electronic structures in the 3C (NHC)Fe(CH 2 TMS) 2 complexes. Lastly, UV-vis MCD spectra of the three complexes indicate small differences in the CT transitions (see ESI †). Spin unrestricted DFT calculations were used to further analyze the electronic structures of (IPr)Fe(CH 2 TMS) 2 , (SIPr) Fe(CH 2 TMS) 2 , and ( Cl IPr)Fe(CH 2 TMS) 2 . Geometry optimizations with B3LYP/TZVP yielded overall structural features, bond lengths and angles in good agreement with those observed by crystallography (Table 5; note that no crystal structure is available for ( Cl IPr)Fe(CH 2 TMS) 2 ). Both experimental and computational studies of the series of 3C NHC complexes are indicative of high-spin iron(II) complexes (S ¼ 2). Details of the MO analyses are given in the ESI. † Importantly, CDA and MBO analyses were completed for the series of 3C Fe(II) NHC complexes ( Table  6). From the CDA, it can be seen that the total donation from the NHC fragment to the Fe-(CH 2 TMS) 2 fragment is the highest for (IPr)Fe(CH 2 TMS) 2 , suggesting that this complex is the strongest s-donor of the series, followed by (SIPr)Fe(CH 2 TMS) 2 , and ( Cl IPr)Fe(CH 2 TMS) 2 (see Table 6). While it can also be seen that (SIPr)Fe(CH 2 TMS) 2 has the highest back donation of the series, the overall back donation to the NHC fragment (average ¼ 0.076 electrons) for all three complexes is small when compared to the total donation to the Fe-(CH 2 TMS) 2 fragment (average ¼ 0.

Conclusions
While signicant progress has been made in the understanding of metal-NHC bonding and electronic structure in precious metal systems, especially electron rich systems with CO ligation, the elucidation of electronic structure and bonding in high-spin iron-NHC systems remains limited. In the present study, the rst application of an approach combining magnetic circular dichroism studies to evaluate LF transitions, 10Dq(T d ) and metal-centered charge-transfer transitions with detailed DFT studies including Mayer bond order and charge decomposition analyses is used to evaluate electronic structure and bonding in electron poor iron-NHCs. In contrast to IR-based methods reliant upon CO ligation, this approach is a direct probe of electronic structure in iron-NHC complexes and broadly applicable to any S > 0 iron-NHC species and, hence, is not limited to the low-spin iron species generally present with CO ligation. In terms of catalysis, the ability to probe electronic structure and bonding in more electron decient iron species is essential as it has been proposed that such iron(I) and iron(II) species may be active in cross-coupling. Near-infrared MCD studies of distorted tetrahedral (IMes) 2 FeCl 2 compared to a series of L 2 FeCl 2 distorted tetrahedral (L ¼ phosphine or amine) permit the rst direct elucidation of 10Dq(T d ) and hence, ligand eld strength, of NHCs relative to phosphine and diamine ligands. From these studies, 10Dq(T d ) for (IMes) 2 FeCl 2 is found to be intermediate in magnitude relative to the corresponding phosphine (largest 10Dq(T d )) and diamine (smallest 10Dq(T d )) complexes. While the observed 10Dq(T d ) values initially appear to contradict existing views of NHCs as stronger eld ligands than phosphines, 35,72 Mayer bond order and charge decomposition analyses indicate that the NHC is a stronger donor ligand than phosphines. The origin of the reduced 10Dq(T d ) value with NHC ligation is found to reect the signicant weakening of the Fe-Cl bonds (leading to reduced Fe-Cl bond orders and reduced charge donation from Cl to Fe) in the NHC complexes relative to the comparable amine and phosphine complexes. While strong trans-type inuences in metal-NHCs have been previously proposed, 72 this study provides the rst direct quantitative evaluation of this in terms of the resulting energies of both the occupied and unoccupied d orbitals. These effects are signicant in L 2 FeCl 2 complexes as indicated by the large differences in 10Dq(T d ), clearly demonstrating that NHCs are not simple analogues of phosphine ligands for iron. Importantly, the differences in the electronic structures of iron-NHC vs. iron-phosphine complexes are likely a signicant contributing factor to the differing catalytic performances observed with these ligands. For example, the variations in the d orbital energies (both occupied and unoccupied) could directly affect the reaction barriers for homolytic R-X cleavage or oxidative addition proposed in iron-catalyzed crosscoupling. [73][74][75] Furthermore, the signicant effect of iron-NHC coordination on the ligation strength of other ligands to iron, such as nucleophile-derived ligands in cross-coupling, could provide a pathway to modulate the reactivity of coordinated ligands that ultimately form new C-C bonds upon reaction with electrophiles. For example, it is anticipated that the trans-type inuence due to the NHC ligation will result in a trans effect where the rate of Fe-Cl substitution/transmetalation will be increased. Similarly, for transmetalated iron species of the form (IMes) 2 FeRX or (IMes) 2 FeR 2 , the rate of Fe-R bond dissociation required for cross-coupled product generation should also be increased due to this effect.
Analogous studies of electronic structure and bonding as a function of NHC backbone structure (e.g. saturated, unsaturated, chlorinated) in both three-and four-coordinate iron(II)-NHC complexes were also performed due to the significant differences reported in catalytic systems as a function of NHC backbone structure. In these high-spin iron(II) systems (S ¼ 2), minimal effects were observed as a function of NHC backbone structure. However, the small differences in back donation observed are likely amplied at lower oxidation states of iron due to the resulting decrease in the energy separation between the occupied iron d orbitals and the unoccupied NHC p* orbitals as the iron oxidation state is reduced below iron(II). While the iron(I)-NHC complex, (IMes) 2 FeCl, has been recently reported in the literature, 76 iron(I)-NHC complexes with varied NHC backbone structures are not yet known. However, preliminary DFT and spectroscopic studies of (IMes) 2 FeCl are consistent with a signicant increase in Fe-Cl / IMes 2 backdonation upon reduction to iron(I). Once iron(I) complexes with varied NHC ligands are synthetically accessible, electronic structure studies can provide further insight into the effects of NHC structure on iron-NHC bonding as a function of iron oxidation state.
The continued application of the combined MCD and DFT approach employed herein to additional iron-NHC systems, including N-substituent variations in iron(II)-NHCs and iron(I)-NHCs as a function of coordination number, geometry and supporting ligands, should continue to expand our fundamental understanding of iron-NHC electronic structure and bonding. Ultimately, such studies will continue to provide critical insight into the molecular-level origins of variations in catalytic performance as a function of NHC structure with catalytically relevant supporting ligands. also thank Dr William W. Brennessel for assistance in the collection and analysis of X-ray crystallographic data.