PMCCPMCCPMCC

Search tips
Search criteria 

Advanced

 
Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
 
Nat Chem. Author manuscript; available in PMC 2011 January 1.
Published in final edited form as:
Published online 2010 May 16. doi:  10.1038/nchem.660
PMCID: PMC2892393
NIHMSID: NIHMS192087

Triggering N2 Uptake via Redox Induced Expulsion of Coordinated NH3 and N2 Silylation at Trigonal Bipyramidal Iron

Abstract

The biological reduction of nitrogen to ammonia may occur via one of two predominant pathways in which nitrogenous NxHy intermediates including hydrazine (N2H4), diazene (N2H2), nitride (N3-) and imide (NH2-) may be involved. To test the validity of hypotheses concerning iron’s direct role in the stepwise reduction of N2, iron model systems are needed. Such systems can test the chemical compatibility of iron with various proposed NxHy intermediates, and the reactivity patterns of such species. Here we describe a TBP (SiPR3)Fe-L scaffold (SiPR3 represents [Si(o-C6H4PR2)3]; R = Ph and iPr) where the apical site is occupied by nitrogenous ligands such as N2, N2H4, NH3 and N2R. The system accommodates terminally bound N2 in the three formal oxidation states (iron(0), +1, and +2). N2 uptake is demonstrated via displacement of its reduction partners NH3 and N2H4, and N2 functionalizaton is illustrated via electrophilic silylation.

Recent work from our group and several others has targeted the synthesis of a variety of Fe-NxHy small molecule model complexes motivated by two goals.i,ii,iii,iv First and foremost is the desire to develop synthetic catalysts whose mode/s of action might relate to, or at least stimulate hypotheses concerning, the manner by which biological nitrogenases reduce N2.v,vi,vii,viii Second, to be better positioned to interpret the spectroscopic data recently obtained for proposed intermediates of the N2-ase cofactor there is a timely need to build a library of Fe-NxHy model complexes as a point of reference. Of specific interest to us are Fe-NxHy complexes whose iron centers reside in geometries that are either 4- or 5-coordinate and feature local three-fold symmetry where the NxHy can be viewed as occupying an axial site.ix,x Such geometries may be relevant to some if not all of the intermediates of iron-bound N2 reduction cycles, as has been advanced elsewhere.v,xi

With these goals in mind we have recently begun working with monoanionic tetradentate trisphosphinosilyl SiPR3 ligands (SiPR3 represents [Si(o-C6H4PR2)3]; R = Ph and iPr) that accommodate mononuclear, open-shell 5-coordinate iron(II) and iron(I) species with a proclivity towards binding N2 in the axial site of a trigonal bipyramid (TBP) at a position that is trans to the silyl anchor.xii,xvi Preliminary reactivity data established that protonation of N2 can occur in modest yield to liberate N2H4.xii Hence, it became of interest to target hydrazine complexes and other open-shell iron complexes featuring nitrogenous ligands in the axial site. Established herein is that the (SiPR3)Fe template binds N2 axially trans to the silyl anchor in three distinct oxidation states that can be represented formally as iron(0), +1, and +2. To our knowledge, no previously established transition metal system has been characterized that can accommodate terminal N2 ligation across three oxidation states. In addition, the recycling of Fe(II)-NH3 and Fe(II)-N2H4 complexes to Fe(I)-N2 with expulsion of NH3 is illustrated; this transformation is of interest as a key step of a hypothetical catalyst cycle where the iron(I) oxidation state is used to trigger N2 uptake and NH3 release. Finally, we also establish that it is possible to directly silylate the coordinated N2 ligand to produce Fe-N2SiR3 products that appear to be far more stable than their Fe-N2H counter-parts. This reactivity pattern, while well established for certain molybdenum systems, is not well known for iron.xiii In sum, these chemical properties add motivation to the search for a molecular N2 reduction catalyst that uses iron as the redox active center to facilitate N2 binding and reduction.

An external file that holds a picture, illustration, etc.
Object name is nihms192087u1.jpg

The most convenient means of entry to the chemistry described herein proceeds via the iron(II) methyl complexes (SiPPh3)Fe(CH3) (1a) and (SiPiPr3)Fe(CH3) (1b). Addition of CH3MgCl to a mixture of ferrous chloride with the corresponding silane H[SiPR3] in tetrahydrofuran at −78 °C, followed by stirring overnight at RT, affords the red S = 1 methyl complexes 1a and 1b in good yield. While these species can be isolated in relatively pure form, trace amounts of the (SiPR3)Fe(N2) complex are typically present due to competitive reduction by CH3MgCl. The solid-state structures of 1a and 1b have been determined (see SI for details) and show nearly ideal TBP geometries at the iron centers (τ = 0.91 for 1a and 0.96 for 1b, where τ = 0.00 for a perfect square pyramid and τ = 1.00 for a TBP geometryxiv). The solid-state structures are noteworthy in that the methyl ligands occupy axial sites trans to the silyl anchors (see SI). Cyclic voltammetry of 1a reveals two reversible redox waves; E½ = −0.57 and −2.3 V (FeIII/II and FeII/I respectively; vs. Fc/Fc+, see SI). Corresponding redox events for the isopropyl derivative 1b are cathodically shifted by ~300 mV.

Synthesis and characterization of Fe-N2, Fe-N2+, and Fe-N2-

Exposure of the methyl complexes to acid sources selectively releases methane. In THF solvent with H(OEt2)2(B(ArF)4) as the added acid (B(ArF)4 = B(3,5-(CF3)2-C6H3)4), 1a is protonated to generate the cationic THF adduct {(SiPPh3)FeII(THF)}{B(ArF)4}, (2a). By contrast, exposure of the more electron releasing species 1b to H(OEt2)2(B(ArF)4) under nitrogen favors formation of the cationic nitrogen complex {(SiPiPr3)FeII(N2)}{B(ArF)4} (3) (Figure 1), which in THF solution under an atmosphere of nitrogen dominates the THF-adduct species by a ratio of ca. 6:1 as determined by UV-vis analysis. Alternatively, 3 can be obtained as a blue powder by adding H(OEt2)2(B(ArF)4) to a benzene solution of the red, previously reported N2 adduct (SiPiPr3)Fe(N2) (4b).xvi GC analysis confirms H2 as the byproduct of the latter reaction (see SI for details).

Figure 1
Synthetic scheme for the generation of Fe-N2+, Fe-N2, and Fe-N2- (3, 4b, 5 and 5’)

The presence of cationic (SiPiPr3)Fe(N2)+ species can also be gleaned by comparing the cyclic voltammetry of the neutral N2 adduct (SiPiPr3)Fe(N2) (4b) under a nitrogen or argon atmosphere in THF solution. Figure 2 shows four traces. Trace (a) provides the cyclic voltammogram of 4b under a nitrogen atmosphere. Two prominent and reversible waves are present at ca. -1.0 V and -2.2 V versus FeCp2/FeCp2+. These are assigned as the Fe-N2/Fe-N2+ and Fe-N2/Fe-N2- waves, respectively. The wave at -1.0 V shows a small shoulder on its negative side (-1.1 V) that we presume arises due to the generation of Fe-THF+ in addition to Fe-N2+ upon oxidation. As the sample is scanned cathodically a small feature appears at -1.9 V that we presume is due to the irreversible reduction of Fe-THF+. Indeed, when 4b is sparged for 30 sec with argon (b) the oxidation wave at -1.0 V corresponding to the oxidation of Fe-N2 is no longer reversible because oxidation leads to rapid loss of N2. Accordingly, the peak at -1.9 V increases in intensity because the generation of Fe-THF+ is favored under argon. Re-admission of N2 to the solution after removing most of the argon by rapid evacuation (c) gives rise to a partially recovered return wave at -1.0 V, which grows in intensity after thorough sparging with nitrogen (d) to provide a trace that is very similar to that observed initially (a), with the exception of a modest impurity appearing at ca. -1.7 V. One additional species to consider in the context of the assignments proposed above concerns trigonal pyramidal (SiPiPr3)Fe. The N2 ligand of 4b is modestly labile and it could therefore be that some of the minor features in the CV traces shown in Figure 2 arise from redox at such a 4-coordinate (SiPiPr3)Fe species, for example the wave at -1.9 V. Our preference for assigning this latter wave to the reduction of Fe-THF+ is in part due to the fact that when N2 is removed (trace (b)) the solution color (orange) is that of other 5-coordinate and divalent (SiPiPr3)Fe(L)+ species, for example the hydrazine adduct 6b (vide infra).

Figure 2
Cyclic voltammetry behavior of (SiPiPr3)Fe(N2) (4b)

We were gratified to find that the cationic complex 3 is sufficiently stable to be isolated and characterized. Its S = 1 spin state can be compared to that of its previously reported and nearly isostructural S = ½ relative 4b.xii Sodium naphthelide reduction of 4b affords the formally zerovalent congener {(SiPiPr3)Fe(N2)}{Na(THF)3} (5). Addition of two equiv of 12-crown-4 to 5 encapsulates the Na+ to provide terminally bonded {(SiPiPr3)Fe(N2)}{Na(12-C-4)2} (5’). High resolution crystal structures have now been obtained for 3, 5, and 5’ to accompany that which had been previously reported for 4b. These structural data collectively afford the only such data available for a terminally bonded N2 adduct of any transition metal in three distinct oxidation states (Figure 3, Table 1). Schrock has reported that the trivalent molybednum dinitrogen adduct [(HIPTNCH2CH2)3N]Mo(N2) (HIPT = 3,5-(2,4,6-i-Pr3C6H2)2C6H3) shows electrochemically reversible waves assigned as the Mo-N2+/0 and Mo-N20/-, where the neutral and anionic derivatives have been structurally characterized, the latter as a Mg adduct.xv Key to note for the present iron system is that the N2 ligand remains in the site trans to the Si anchor in each state of oxidation, and the iron center’s geometry is preserved in the cationic, neutral, and anionic species. Structural changes worth noting include an Fe-N bond distance contraction as the system is successively reduced, and a corresponding Fe-Si contraction upon successive reduction (Table 1).

Figure 3
Solid-state structures of 3, 5, and 5’
Table 1
Physical parameters for the N2 adduct species 3, 4b, 5 and 5’

Synthesis and characterization of Fe-NH3+, Fe-N2H4+, and Fe-N2H3B(C6F5)3

The cationic THF and N2 adducts are labile at the axial site trans to the silyl donor, and hence provide one pathway to the corresponding hydrazine adduct derivatives {(SiPPh3)FeII(N2H4)}{B(ArF)4}, {6a}{B(ArF)4} and {(SiPiPr3)FeII(N2H4)}{B(ArF)4}, {6b}{B(ArF)4)} via N2H4 addition. Alternatively, slow addition of the hydrazinium acid N2H5CF3SO3 to either 1a or 1b in THF generates dark red solutions of {6a}{OTf} and {6b}{OTf}, both of which can be isolated in > 90% yield. Their S = 1 spin states (μeff = 2.79 μB for 6a and 3.0 μB for 6b) are consistent with TBP structures, as confirmed by XRD analysis (τ = ~0.9 for 6a, 0.96 for 6b, Figure 4). Their solid-state structures reveal in each case a hydrazine ligand η1 coordinated to a 5-coordinate iron center in an axial site opposite the silyl anchor. For comparison, diamagnetic and 6-coordinate η2-hydrazine iron(II) complexes have been reported utilizing bidentate phosphine ligands.ii,xvii,xviii As indicated by Figure 4, the hydrazine moieties in {6a}{OTf} and {6b}{OTf} are hydrogen-bonded to the triflate anions in the solid-state, with average N-O distances of ~ 3 Å. In 6b, the hydrogen atoms could be located in the difference map at an average distance of ~ 2 Å for N -H(...)O. The N-H vibrations for these complexes show the presence of hydrogen bonds in the IR. These vibrations are broadened and shifted in solid-state spectra from that of their non-hydrogen bonded hydrazine derivatives {6a}{B(ArF)4} and {6b}{B(ArF)4}. Hydrazine adducts with an η1-binding mode to 5-coordinate metal complexes are uncommon.iv,xix,xx,xxi,xxii,xxiii,xxiv To our knowledge, the only other example of such a species showing approximate three-fold symmetry akin to 6a and 6b is a vanadium hydrazine complex supported by a tris(thiolate)amine ligand.xxv

Figure 4
Solid-state structures of {6b}{OTf}, 7a, N2H4B(C6F5)3, and 9b
An external file that holds a picture, illustration, etc.
Object name is nihms192087u2.jpg

The hydrazine ligand is quite labile for both 6a and 6b, and binding of the triflate anion with concomitant release of N2H4 can be observed by NMR spectroscopy in C6D6. Lability at the apical site, while potentially useful for a catalytic system, is problematic with regard to attempts to generate an Fe(HN=NH) complex via oxidation of 6a and 6b. For instance, our attempts to oxidize these hydrazine complexes with Pb(OAc)4 instead afforded mixtures of the neutral {(SiPR3)Fe(OTf)}xvi and {(SiPR3)Fe(OAc)} complexes (see SI). Perhaps more interesting is that 6b can be fully oxidized to {(SiPiPr3)FeII(N2)}+ by 3,5-di-tert-butyl-o-benzoquinone.xxvi

By analogy to the conversion of 6a and 6b, a THF solution of ammonia reacts with either 2a or 3 to afford the corresponding cationic ammonia adducts {(SiPPh3)FeII(NH3)}{B(ArF)4}, (7a), and {(SiPiPr3)FeII(NH3)}{B(ArF)4}, (7b), Figure 4. The NH3 ligand is substitutionally labile and hence obtaining rigorously pure samples by thorough drying is challenging: solvents in which the compounds dissolve (e.g., THF) partially substitute the NH3 ligand. Triplet 7beff = 3.27 μB) has been structurally characterized and as for the hydrazine derivatives features an NH3 ligand in the apical site opposite the silyl donor. While its structure (Figure 4) is unremarkable, it serves to underscore that the apical site of the {(SiPiPr3)Fe} system can accommodate N2 in the 0, +1, and +2 oxidation states, whereas NH3 ligation appears accessible only in the +2 oxidation state. Indeed, if one tries to reduce either 7a or 7b, NH3 is quantitatively released and the Fe(I)-N2 adducts 4a and 4b are generated, respectively. The significance of this transformation lies in the ability to recycle Fe(I)-N2 with release of NH3, key to the ultimate viability of a hypothetical Fe(I)-N2 catalyst system for generating ammonia. Also of note is that the reduction of the hydrazine adducts 6a and 6b leads to facile generation of 4a and 4b, respectively. In these cases, both N2H4 and NH3 are generated as determined by vacuum transfer of the volatiles.

An external file that holds a picture, illustration, etc.
Object name is nihms192087u3.jpg

A rare NXHy ligand for iron that we sought within this system is the hydrazido (N2H3-) ligand.iii We reasoned that the hydrazine adducts 6a and 6b might afford access to such complexes via deprotonation. While this did not turn out to be the case, the reaction that results is interesting. When a THF solution of 6a{OTf} is exposed to a stoichiometric equiv of N1,N1,N8,N8-tetramethylnaphthalene-1,8-diamine (proton sponge) a reaction ensues affording the paramagnetic ammonia adduct complex {[(Si(o-C6H4PPh2)2(o-C6H4P(=NH)Ph2)]Fe(NH3)}{OTf}, (8), which has been identified by XRD analysis. Its structure reveals that one arm of the SiPPh3 ligand is oxidized to P(V) via formal insertion of NH into the Fe-P bond. The N=P bond distance in 8 of 1.5945(1) Å is very close to other reported N=P double bonds.xxvii,xxviii,xxix,xxx,xxxi,xxxii IR spectroscopy shows N-H vibrations at 3339, 3256 and 3168 cm−1. Complexes 6a and 8 are hence structural isomers of one another and the role of the base thereby appears to be catalytic. The details of this reaction are, however, unclear and complicated by the presence of unidentified byproducts.

While the terminally bonded N2H3 ligand is elusive for these {(SiP3)Fe} systems, such a ligand can be generated in the presence of the Lewis Acid acceptor B(C6F5)3. Thus, the addition of (C6F5)BNH2NH2 to 1a or 1b leads to the formation of the neutral and zwitterionic iron(II) hydrazido-borane complexes (SiPPh3)FeII(N2H3B(C6F5)3) (9a) and (SiPiPr3)FeII(N2H3B(C6F5)3) (9b), Figure 4. The hydrazine-borane adduct N2H4B(C6F5)3 was synthesized from the 1:1 mixture of hydrazine and tris(pentafluorophenyl)borane in THF. Both 9a and 9b give easily resolved 19F NMR signals at −123, −157, −162 ppm despite their triplet ground states (μeff = 2.90 μB and 2.83 μB). Solid-state crystal structures reveal η1-bound hydrazido-borane ligands with the borane terminating the β–NH (see Figure 4). To our knowledge, this ligand type is unique.

The B-N bonds (1.544(5) Å for 9a and 1.553(3) Å for 9b) are much shorter than that of the precursor N2H4B(C6F5)3 (1.6316(19) Å). The N-N bond distances are 1.449(4) and 1.442(10) Å for 9a and 9b, respectively, which are slightly shortened from that in free N2H4B(C6F5)3. Interestingly, hydrogen bonds between the hydrogen atoms of hydrazine (and hydrazido) and ortho-fluorine atoms of B(C6F5)3 are exhibited in 9a and 9b, and also in the precursor N2H4B(C6F5)3, Figure 4. These intra-molecular N–H(...)F-C hydrogen bonds are relatively unusual examples of hydrogen bonding in the literature.xxxiii,xxxiv,xxxv,xxxvi,xxxvii All H-bonded H-atoms can be located from the difference maps of the corresponding X-ray crystallographic data, and display distances in the H…F hydrogen bond range; 2.158 Å ~ 2.356 Å.

Synthesis and characterization of Fe-N2Ph and Fe-N2SiMe3

To attempt the synthesis of a mono-substituted hydrazido derivative the methyl complex 1a was exposed to phenylhydrazinium triflate, Figure 5. The reaction instead afforded a mixture of species presumed to contain paramagnetic {(SiPPh3)Fe(NH2-NHPh)}{OTf} and (SiPPh3)Fe(OTf). To try to isolate a well-defined FeII(NH-NHPh) species the addition of base was pursued. However, while addition of base appears to remove H+ it also triggers formal loss of H2 to afford the phenyldiazenido complex {(SiPPh3)Fe(N2C6H5)}, (10). This is true for bases such as proton sponge and also phenylhydrazine, PhNH-NH2, which is necessarily present in solution. Dark brown crystals of 10 can be isolated from the reaction mixture in good yield (~ 70%, Figure 5), and IR spectroscopy reveals an N-N vibration at 1623 cm−1. The solid-state crystal structure (Figure 5) confirms an η1-phenyldiazenido ligand in an axial position trans to the silyl donor, with short N-N and Fe-N distances (1.233(7) and 1.690(5) Å) reflecting multiple bond character in each linkage. The N-N-C angle (122.5(5)°) establishes sp2 hybridization at Nβ. Diazenido 10 is structurally distinct by virtue of having a diazenido ligand occupying an axial position of a TBP geometry. For the few 5-coordinate iron diazenido complexes that have been structurally characterized, the diazenido ligand occupies an equatorial site.xxxviii,xxxix,xl,xli

Figure 5
Synthesis and characterization of (SiPPh3)FeII(N2C6H5) (10) and {(SiPPh3)FeII(N2C6H5)}{B(C6H3(CF3)2)4} (11)

The reaction of {(SiPPh3)Fe(THF)}{B(ArF)4} 2a and phenyl hydrazine gives a red solution of {(SiPPh3)Fe(NH2-NHPh)}+, with hydrazine like 1H-NMR signatures based on comparison with the spectra of 6a and 6b. N-H vibrations are observed at 3346, 3271, 3230 cm−1. The red product is unstable and is slowly converted to the ammonia adduct 7a at RT presumably as a result of disproportionation of the iron-bound phenylhydrazine. The major organic product is aniline, as identified by 1H-NMR and GC. Use of an excess of phenylhydrazine instead gave rise to the inky black product [(SiPPh3)Fe(N2C6H5)]+ (11). The same product is also obtained by the addition of {Cp2Fe}{B(ArF)4)} to 10 in C6D6 solution. The N-N vibrational frequency of 11 is 1690 cm−1, revealing comparatively less back donation from iron to the N-N π* orbital than for 10. This is also supported by the solid-state structure of 11, which shows a shorter N-N bond distance and a longer Fe-N bond distance than for 10 (Figure 5). Harder to explain is the curious lengthening of the Fe-P bond distances upon oxidation of 10 to 11.

Our isolation of the phenyldiazenido complexes 10 and 11 motivated us to explore whether we might be able to prepare related diazenido complexes by direct functionalization of the iron-bound N2 ligand in the neutral adduct complexes 4a, 4b or the anion 5. Whereas we have previously shown that the N2 ligand in 4a can be protonated in modest yield to release hydrazine (46% in the presence of CrCl2), trapping a derivatized N2 ligand still bound to the iron center has proven elusive for the phenyl decorated (SiPPh3)Fe system. When H(OEt2)2(B(ArF)4) or CH3OTf are added to 5 in THF at low temperature, thermally unstable and as yet uncharacterized species appear that eventually decay to the iron(I) N2 adduct 4b. At this stage we can only speculate as to the presence of Fe-N2H and Fe-N2Me intermediates. The use of silyl electrophiles has proven more fruitful with regard to isolation of products. Thus, treatment of 5 with TMSCl or TMSOTf in frozen THF followed by gradual warming of the solution affords the desired dark red diazenido complex (SiPiPr3)Fe(N2SiMe3) (12) with concomitant salt elimination, Figure 6. Complex 12 can also be generated directly from 4b if Na/Hg amalgam is used as a reductant in the presence of TMSCl. The analogous complexes (SiPiPr3)Fe(N2SiiPr3) and (SiPiPr3)Fe(N2SiPh3) are obtained using triisopropylsilyl trifluoromethanesulfonate (TIPS-OTf) and triphenylsilyl chloride, respectively. In contrast to its S = 1 relative 10, diazenido 12 is diamagnetic. Two 29Si-NMR resonances are present in the 29Si-NMR spectrum at 84.3 ppm (q, 2JSiP = 38 Hz) and −15.6 ppm (s). A 15N-NMR spectrum of the labeled complex 12-15N shows two resonances at 418.5 and 270.9 ppmxlii shifted from corresponding peaks for the 15N-enriched precursor 5-15N (340.3 and 309.7 ppm). Large separation between these two 15N signals is fully consistent with functionalization at the dinitrogen ligand by the TMS group, as for the related molybdenum complex [HIPTN3N]Mo-NNH species.xliii The N-N vibrational frequency of 12 is 1748 cm-1 (1694 cm−1 for 12-15N).

Figure 6
Synthesis and characterization of (SiPiPr3)FeII(N2SiMe3) (12)

Dark red crystals of 12 were obtained and an XRD analysis reveals a TMS group bound to Nβ of the TBP iron scaffold (τ = 0.93; Figure 6). The relatively short Fe-N1 distance (1.695(2) Å) implies multiple bond character between the iron center and Nα. The N-N bond distance of 1.195(3) Å establishes further reduction of the N2 unit relative to its precursor 5 (1.147(4) Å), where a Na+ cation interacts with Nβ. A single point DFT calculation (see SI for details) of 12 illuminates the multiple bond character between iron and Nα nicely, revealing that both HOMO and HOMO-1 possess significant π bonding character between the Fe and N atoms (Figure 6). The difference in magnetic behavior between diazenidos 10 and 12 is curious and is the subject of ongoing studies in our lab. We tentatively suggest that complex 12 is best formulated as a d8 iron anion, akin to 5 and 5’, that strongly backbonds into the N2SiMe3+ π* orbitals. Such a configuration for a TBP structure is expected to produce a diamagnet. By contrast, perhaps 10 is better formulated as a d6 iron center, which for a TBP structure provides a spin triplet in accord with the numerous other S = 1 TBP iron(II) complexes described herein. The angle [for all]Nα-Nβ-Si (165.6(3)°) for 12 is far less bent than the [for all]Nα-Nβ-C in complex 10 (122.5(5)°), which is consistent with this comparative description.

To better evaluate the relative state of oxidation of the diamagnetic diazenido species 12 by comparison to the other (SiPiPr3)Fe species described herein we collected Mössbauer spectra for solid samples of (SiPiPr3)Fe(Cl), (SiPiPr3)Fe(N2)+ (3), (SiPiPr3)Fe(N2) (4b), {(SiPiPr3)Fe(N2)}{Na(THF)3} (5), {(SiPiPr3)Fe(N2)}{Na(12-C-4)2} (5’), and (SiPiPr3)Fe(N2SiMe3) (12) in zero external magnetic field at 77 K. Each of the spectra shows single quadrupole doublets as shown in Figure 7. Their isomer shifts and quadrupole splittings are listed in Figure 7. The isomer shift of cationic 3 is very close to that of (SiPiPr3)Fe(Cl) (S = 1) and consistent with other ferrous complexes.ix The isomer shift decreases by ca. 0.1 to 0.15 mm/s per formal state of oxidation from 3 to 4b, and from 4b to 5 and 5’. The isomer shift of the silyldiazenido species 12 is closer to that of 5 than 5’ in accord with our supposition that the TMS group capping the N2 ligand is electronically comparable to the Na(THF)3+ cation. Therefore, an Fe(0) d8 assignment is best accorded to complex 12, at least to the extent that such an assignment is appropriate for diamagnetic 5 and 5’. Because these complexes are highly covalent our primary intent here is to compare their relative states of oxidation with respect to one another.

Figure 7
Zero field Mössbauer spectra

In summary, the {(SiPR3)Fe} scaffold continues to show its effectiveness in stabilizing nitrogenous donor ligands in the apical site of a trigonal bipyramid, trans to the silyl anchor of the ligand auxiliary. In particular, terminal N2 binding is structurally established for the formal oxidation states Fe(0), Fe(I), and Fe(II). In addition, all of the 5-coordinate iron(II) structures described herein are open shell triplets. The synthesis of open-shell Fe-NxHy systems is of timely interest for comparison of their spectroscopic parameters with related data being obtained for the cofactor of nitrogenase under catalytic turnover conditions. The demonstration that the {(SiPR3)Fe} scaffold can accommodate N2, NH3, and N2H4 in the apical site, and that (SiPR3)FeII-NH3+ and (SiPR3)FeII-N2H4+ species can be recycled to (SiPR3)FeI-N2 via chemical reduction with concomitant liberation of NH3, suggests to us that an iron-mediated nitrogen fixation catalyst system based upon three-fold symmetry may yet be accessible. A promising lead is that the iron-bound N2 ligand reacts with electrophiles at the Fe(0) state, which for the silyl derivatives afford stable Fe-N2SiR3 diazenido products.

Acknowledgments

We acknowledge the NIH (GM-070757). Funding for the MIT Department of Chemistry Instrumentation Facility has been provided in part by the NSF (CHE-0234877). Dr. Peter Mueller provided assistance with XRD analyses. NPM is grateful for an NSF graduate fellowship. We thank Professors R. H. Holm and T. A. Betley at Harvard University for providing us with generous access to a Mössbauer spectrometer.

Footnotes

Author contributions

Y.L., N.P.M. and J.C.P. conceived and designed the experiments, Y.L. and N.P.M. performed the experiments and Y.L. and J.C.P. co-wrote the paper.

Additional information

Supplementary information and chemical sompound information accompany this paper at www.nature.com/naturechemistry. Reprints and permission information is available online at http://npg.nature.com/reprintsandpermissions/. Correspondence and requests for materials should be addressed to J.C.P.

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

References

i. Saouma CT, Müller P, Peters JC. Characterization of structurally unusual diiron NxHy complexes. J Am Chem Soc. 2009;131:10358–10359. [PubMed]
ii. Field LD, Li HL, Magill AM. Base-Mediated Conversion of Hydrazine to Diazene and Dinitrogen at an Iron Center. Inorg Chem. 2009;48:5–7. [PubMed]
iii. Crossland JL, Balesdent CG, Tyler DR. Intermediates in the reduction of N2 to NH3: synthesis of iron η2 hydrazido(1-) and diazene complexes. Dalton Trans. 2009:4420–4422. [PubMed]
iv. Sellmann D, Shaban SY, Heinemann FW. Syntheses, Structures and Reactivity of Electron-Rich Fe and Ru Complexes with the New Pentadentate Ligand Et2NpyS4-H2 {4-(Diethylamino)-2,6-bis[(2-mercaptophenyl)thiomethyl]pyridine} Eur J Inorg Chem. 2004:4591–4601.
v. Peters JC, Mehn MP. Activation of Small Molecules Ch 3. Wiley-VCH; Weinheim: 2006.
vi. Barney BM, et al. Diazene (HN=NH) Is a Substrate for Nitrogenase: Insights into the Pathway of N2 Reduction. Biochemistry. 2007;46:6784–6794. [PMC free article] [PubMed]
vii. Barney BM, et al. A methyldiazene (HN=N-CH3)-derived species bound to the nitrogenase active-site FeMo cofactor: Implications for mechanism. Proc Natl Acad Sci USA. 2006;103:17113–17118. [PubMed]
viii. Barney BM, et al. Intermediates Trapped during Nitrogenase Reduction of N[equivalent]N, CH3–N=NH, and H2N–NH2. J Am Chem Soc. 2005;127:14960–14961. [PubMed]
ix. Hendrich MP, Gunderson W, Behan RK, Green MT, Mehn MP, Betley TA, Lu CC, Peters JC. On the feasibility of N2 fixation via a single site FeI/FeIV cycle - Spectroscopic studies of FeI(N2)FeI, FeIV[equivalent]N, and related species. Proc Natl Acad Sci U S A. 2006;103:17107–17112. [PubMed]
x. MacBeth CE, Harkins SB, Peters JC. Synthesis and Characterization of Cationic Iron Complexes Supported by the Neutral Ligands NPi-Pr3, NArPi-Pr3, andNSt-Bu3. Can J Chem. 2005;83:332–340.
xi. Hinnemann B, NØrskov JK. Modeling a Central Ligand in the Nitrogenase FeMo Cofactor. J Am Chem Soc. 2003;125:1466–1467. [PubMed]
xii. Mankad NP, Whited MT, Peters JC. Terminal FeI-N2 and FeII…H-C Interactions Supported by Tris(phosphino)silyl Ligands. Angew Chem Int Ed. 2007;46:5768–5771. [PubMed]
xiii. Betley TA, Peters JC. Dinitrogen Chemistry from Trigonally Coordinated Iron and Cobalt Platforms. J Am Chem Soc. 2003;125:10782–10783. [PubMed]
xiv. Addison AW, Rao TN, Reedijk J, van Rijn J, Verschoor GC. Synthesis, Structure and Spectroscopic Properties of Copper(II) Compounds Containing Nitrogen-Sulfur Donor Ligands; the Crystal and Molecular Structure of Aqua[1,7-bis(N-methylbenzimidazol-2′- yl)-2,6-dithiaheptane]copper(II) Perchlorate. J Chem Soc Dalton Trans. 1984:1349–1356.
xv. Yandulov DV, Schrock RR. Studies Relevant to Catalytic Reduction of Dinitrogen to Ammonia by Molybdenum Triamidoamine Complexes. Inorg Chem. 2005;44:1103–1117. [PubMed]
xvi. Whited MT, Mankad NP, Lee Y, Oblad PF, Peters JC. Dinitrogen Complexes Supported by Tris(phosphino)silyl Ligands. Inorg Chem. 2009;48:2507–2517. [PubMed]
xvii. Field LD, Li HL, Dalgarno SJ, Turner P. The first side-on bound metal complex of diazene, HN=NH. Chem Comm. 2008:1680–1682. [PubMed]
xviii. Crossland JL, Zakharov LN, Tyler DR. Synthesis and Characterization of an Iron(II) η2-Hydrazine Complex. Inorg Chem. 2007;46:10476–10478. [PubMed]
xix. Yu Y, Brennessel WW, Holland PL. Borane B–C Bond Cleavage by a Low-Coordinate Iron Hydride Complex and N–N Bond Cleavage by the Hydridoborate Product. Organometallics. 2007;26:3217–3226. [PMC free article] [PubMed]
xx. Shakir M, Parveen S, Begum N, Azim Y. Interaction of manganese(II), iron(II), cobalt(II), nickel(II), copper(II) and zinc(II) with acetylhydrazine, formed in situ; first crystal structure of tris(acetylhydrazine) nickel(II) perchlorate. Transition Metal Chem. 2004;29:916–920.
xxi. Albertin G, Antoniutti S, Bordignon E, Chimisso F. Preparation of bis(hydrazine) complexes of iron(II) Inorg Chem Commu. 2001;4:402–404.
xxii. Sellmann D, Soglowek W, Knoch F, Ritter G, Dengler J. Transition-Metal Complexes with Sulfur Ligands. 88. Dependence of Spin State, Structure, and Reactivity of [FeII(L)(‘NHS4’)] Complexes on the Coligand L (L = CO, N2H2, N2H4, NH3, Pyridine, NHCH3NH2, CH3OH, THF, P(OCH3)3, P(OPh)3): Model Complexes for Iron Nitrogenases (‘NHS42– = Dianion of 2,2‘-Bis[(2- mercaptophenyl)thio]diethylamine) Inorg Chem. 1992;31:3711–3717.
xxiii. Casey MT, Guinan P, Canavan A, McCann M, Cardin C, Kelly NB. Reaction of 1,1′-diacetylferrocene with hydrazine hydrate: Synthesis and X-ray crystal structures of bis(hydrazine)bis(hydrazinecarboxylato-N′,O)iron(II), [Fe(N2H4)2(O2CNHNH2)2], and the cyclic biferrocene diazine, [---N(Me)CC5H4FeC5H4C(Me)N---]2. Polyhedron. 1991;10:483–489.
xxiv. Goedken VL, Peng S-M, Molin-Norris J, Park Y-A. Carbon monoxide complexes of iron(II): synthesis and structural studies of five- and six-coordinate complexes of the macrocyclic ligand, C22H22N42– J Am Chem Soc. 1976;98:8391–8400. [PubMed]
xxv. Davies SC, et al. Vanadium Complexes of the N(CH2CH2S)33– and O(CH2CH2S)22– Ligands with Coligands Relevant to Nitrogen Fixation Processes. Inorg Chem. 2000;39:3485–3498. [PubMed]
xxvi. It is likely that the N2 in the product derives from both the original N2H4 in the precursor and atmospheric N2 due to rapid ligand exchange.
xxvii. Alhomaidan O, Hollink E, Stephan DW. Main Group Heterocycles from Lithiated Phosphinimines. Organometallics. 2007;26:3041–3048.
xxviii. Aguilar D, Contel M, Navarro R, Urriolabeitia EP. Organogold(III) Iminophosphorane Complexes as Efficient Catalysts in the Addition of 2-Methylfuran and Electron-Rich Arenes to Methyl Vinyl Ketone. Organometallics. 2007;26:4604–4611.
xxix. Bielsa R, Larrea A, Navarro R, Soler T, Urriolabeitia EP. Synthesis, Structure, Reactivity, and Catalytic Activity of C,N- and C,N,N-Orthopalladated Iminophosphoranes. Eur J Inorg Chem. 2005:1724–1736.
xxx. Chan KTK, et al. Anionic Phosphinimine-Chelate Complexes of Rhodium and Iridium: Steric and Electronic Influences on Oxidative Addition of CH2Cl2. Organometallics. 2004;23:381–390.
xxxi. Spencer LP, et al. Pyridine– and Imidazole–Phosphinimine Bidentate Ligand Complexes: Considerations for Ethylene Oligomerization Catalysts. Organometallics. 2003;22:3841–3854.
xxxii. LePichon L, Stephan DW, Gao Z, Wang Q. Iron Phosphinimide and Phosphinimine Complexes: Catalyst Precursors for Ethylene Polymerization. Organometallics. 2002;21:1362–1366.
xxxiii. Mountford AJ, et al. Intra- and Intermolecular N–H(...)F–C Hydrogen-Bonding Interactions in Amine Adducts of Tris(pentafluorophenyl)borane and –alane. Inorg Chem. 2005;44:5921–5933. [PubMed]
xxxiv. Hyla-Kryspin I, Haufe G, Grimme S. Weak Hydrogen Bridges: A Systematic Theoretical Study on the Nature and Strength of C-H(...)F-C Interactions. Chem Eur J. 2004;10:3411. [PubMed]
xxxv. Brammer L, Bruton EA, Sherwood P. Understanding the Behavior of Halogens as Hydrogen Bond Acceptors. Cryst Crowth Des. 2001;1:277.
xxxvi. Takemura H, et al. The C-F(...)Cation Interaction: An Ammonium Complex of a Hexafluoro Macrocyclic Cage Compound. Chem Eur J. 2000;6:2334. [PubMed]
xxxvii. Howard JAK, Hoy VJ, O’Hagan D, Smith GT. How good is fluorine as a hydrogen bond acceptor? Tetrahedron. 1996;52:12613.
xxxviii. Davies SC, Hughes DL, Richards RL, Sanders JR. Molybdenum and tungsten complexes of the N(CH2CH2S)33– (NS3) ligand with oxide, sulfide, diazenide, hydrazide and nitrosyl co-ligands. J Chem Soc Dalton Trans. 2000:719–725.
xxxix. Albertin G, et al. Reactivity of Hydrides FeH2(CO)2P2 (P = Phosphites) with Aryldiazonium Cations: Preparation, Characterization, Xray Crystal Structure, and Electrochemical Studies of Mono- and Binuclear Aryldiazenido Complexes. Inorg Chem. 1998;37:5602–5610. [PubMed]
xl. Albertin G, Antoniutti S, Pelizzi G, Vitali F, Bordignon E. Bis(aryldiazene) derivatives of iron(II): preparation, characterization, and properties of the first complexes containing two diazene ligands bonded to the same central metal. The x-ray crystal structures of hexacoordinate [FeH(4-CH3C6H4NNH)[P(OEt)3]4]+, and pentacoordinate [Fe(4-CH3C6H4N2)[P(OEt)3]4]+ cations. J Am Chem Soc. 1986;108:6627–6634.
xli. Haymore BL, Ibers JA. Aryldiazo complexes. Structure of an iron-aryldiazo complex, dicarbonyl bis(triphenylphosphine) benzenediazonium iron(1+) tetrafluoroborate (1-) Inorg Chem. 1975;14:1369–1376.
xlii. By 29Si-NMR measurements, 15N-29Si coupling was detected at −15.9 ppm (dd, 1,2JSiN = 10, 2.2 Hz).However, no coupling was detected in the 31P-NMR spectrum of the 15N-enriched sample, even at −90 °C
xliii. Yandulov DV, Schrock RR. Reduction of Dinitrogen to Ammonia at a Well-Protected Reaction Site in a Molybdenum Triamidoamine Complex. J Am Chem Soc. 2002;124:6252–6253. [PubMed]