Hydrogenolysis of [PhBP3]Fe≡N-p-tolyl: Probing the Reactivity of an Iron Imide with H2

Article abstract of DOI:10.1021/ja0399122

This paper describes the reductive hydrogenolysis of a low-spin (S = 1/2) iron(III) imide. Pseudotetrahedral [PhBP₃]Fe^III_≡N-p-tolyl is reduced by hydrogen at ambient temperature and pressure in benzene solution. The reduction appears to proceed in a stepwise fashion. An intermediate S = 2 iron(II) anilide, [PhBP₃]Fe(N(H)-p-tolyl), is observed and has been independently generated and structurally characterized. Prolonged hydrogenolysis in benzene results in the complete hydrogenolysis of the Fe?N linkage to release H₂N-p-tolyl. The major iron-containing product formed from this step is the diamagnetic cyclohexadienyl complex, [PhBP₃]Fe(η⁵-cyclohexadienyl), which has also been independently prepared and structurally characterized. Evidence is presented to suggest that the final [PhBP₃]Fe(η⁵-cyclohexadienyl) product is formed via benzene insertion into a reactive [PhBP₃]Fe^II-H intermediate. Copyright

Full text of DOI:10.1021/ja0399122

Published on Web 03/23/2004
Hydrogenolysis of [PhBP3]FetN-p-tolyl: Probing the Reactivity of an Iron
Imide with H2
Steven D. Brown and Jonas C. Peters*
DiVision of Chemistry and Chemical Engineering, Arnold and Mabel Beckman Laboratories of Chemical Synthesis,
California Institute of Technology, Pasadena, California 91125
Received December 3, 2003; E-mail: jpeters@caltech.edu
It is generally accepted that low-coordinate iron nitride species
are formed and hydrogenated on the surface of solid-state iron
catalysts in the Haber-Bosch process.1 Moreover, it has been
postulated that Fe-N (and/or Mo-N) multiple bonds may be
formed from N2 and further reduced by hydrogen equivalents (i.e.,
H+/e-) during turnover in nitrogenase enzymes.2 Studying the
reactivity patterns of hydrogen with molecular iron complexes
featuring Fe-N multiple bond linkages is therefore of considerable
interest but has been difficult to undertake due to the historical
Figure 1. 50% displacement ellipsoid representations of [PhBP3]Fe(N(H)-
absence of well-defined FedNR and/or FetN species.3
p-tolyl) (2), left, and [PhBP3]Fe(η5-cyclohexadienyl) (3), right. For clarity,
Our group has recently reported the synthesis of a number of
only the P3FeX core is shown. Selected bond lengths (Å) and angles (deg)
for 2: Fe1-N1, 1.913(2); Fe1-P1, 2.392(1); Fe1-P2, 2.452(1); Fe1-P3,
2.427(1); Fe1-N1-C46, 127.4(2). For 3: Fe-P1, 2.273(1); Fe-P2, 2.270-
(1); Fe-P3, 2.265(1); C46-C47, 1.477(4); C47-C48, 1.394(4); C48-C49,
1.415(4); C49-C50, 1.410(4); C50-C51, 1.385(4); C51-C46, 1.493(4).
mononuclear imides of iron and cobalt (e.g., [PhBPR ]M(NR) where
3
[PhBPR ] ) [PhB(CH2PPh2)3]- and [PhB(CH2PiPr2)3]-).4 These
3
trivalent imides are characterized by relatively robust MtNR triple
bonds (1σ + 2π) but are nonetheless able to release their imide
functionalities, an example being transfer to CO producing
isocyanate.4a,b This group transfer reactivity prompted us to survey
their respective reactivities toward hydrogen.
Anilide 2 features a weak N-H vibration at 3326 cm-1 (Nujol),
a temperature-independent magnetic moment of 4.92 µB (SQUID,
S ) 2), and relatively intense charge-transfer bands (450 nm, 3400
M-1 cm-1; 547 nm, 3000 M-1 cm-1). The solid-state structure of
2 (Figure 1) is related to that of 1 by the addition of a single H-atom
at nitrogen. Although both 1 and 2 are four-coordinate and
pseudotetrahedral, 2 features an Fe-N bond distance (1.913(2) Å)
and an Fe-N-C bond angle (127.4(2)°) very different from 1
(Fe-N ) 1.6578(2) Å; Fe-N-C ) 169.96(2)°).4a Of additional
note are the Fe-P bond distances, which are appreciably expanded
in the structure of 2 (Fe-Pavg for 1 ) 2.24 Å; Fe-Pavg for 2 )
2.42 Å). These structural differences reflect the low- versus high-
spin configurations of 1 and 2, respectively, in addition to the
diminished Fe-N π-bond character of 2 by comparison to 1.
Prolonged monitoring of the hydrogenation of 1 in C6D6 over a
period of days establishes a new diamagnetic product (31P NMR:
52 ppm (s)) and the release of H2N-p-tolyl (40% after 3 days). In
a preparative-scale reaction, imide 1 was exposed to an atmosphere
of H2 in benzene for 3 days in a sealed reaction flask. Extraction
of the crude solids into diethyl ether left behind the diamagnetic
species as a bright orange powder in 25% yield. XRD analysis of
a single crystal confirmed it to be the cyclohexadienyl complex
[PhBP3]Fe(η5-cyclohexadienyl) (3) (Figure 1). Complex 2 was then
isolated in 60% yield from the ethereal extract by crystallization.
The combined isolated yield of 2 and 3 accounted for 85% of the
total iron content of the reaction. Similar yields were provided by
The low-spin cobalt complex [PhBP3]CotN-p-tolyl4b is stable
to hydrogen pressure (1-3 atm) at modest temperatures (e70 °C).
By contrast, low-spin [PhBP3]FetN-p-tolyl (1)4a gives rise to a
fascinating reaction profile upon exposure to 1 atm of H2 at room
temperature. Both partial and complete hydrogenolysis of the
FetNR linkage is observed. Whereas H2-promoted reduction of
an imide to its corresponding amide was first described by
Wolczanski,5 reductive scission of a metal imide by hydrogen to
release amine has not to our knowledge been previously reported.6
Fingerprint resonances for imide 1 are conveniently monitored
by 1H NMR spectroscopy despite its paramagnetic S ) 1/2 ground
state.4a When a sealable NMR tube is charged with forest green 1
in C6D6 under an atmosphere of H2, its resonances fully decay
within a period of 3 h at room temperature. During this time, the
reaction solution remains homogeneous and the color changes to
an intense red/purple. The major reaction product at this stage is
the paramagnetic anilido complex [PhBP3]Fe(N(H)-p-tolyl) (2)
(Scheme 1).7 Assignment of 2 from its paramagnetically shifted
Scheme 1
1
in situ monitoring and integration of H NMR spectra of sealed
NMR tube experiments after 3 days. Ill-defined paramagnetic
resonances indicative of at least one side product account for the
remaining iron material (∼15%) in these sealed-tube experiments.
We suspect that the side-product(s) arise from the kinetically
competitive degradation of 2 during the course of the reaction. As
a control experiment, it is noted that the storage of 2 in benzene
under N2 leads to some degradation after 3 days but does not
produce 3.8
1H NMR resonances is aided by comparison to a spectrum of an
independently generated sample of 2 prepared by the reaction
between [Li][N(H)-p-tolyl] and [PhBP3]FeCl.4a
9
4538
J. AM. CHEM. SOC. 2004, 126, 4538-4539
10.1021/ja0399122 CCC: $27.50 © 2004 American Chemical Society

C O M M U N I C A T I O N S
To ensure that the H-atoms being delivered to both benzene and
the imide functionality arise from H2, the hydrogenation of 1 by
D2 in C6D6 was examined. The consumption of 1 proceeds in this
case much more slowly (krel ) k(H2)/k(D2) ) 5.6), and the product
yields after 3 days are consequently much lower. The overall
reaction is moreover less clean than for the case of H2 due to the
increased role of kinetically competitive side reactions given the
longer reaction time. Nonetheless, the expected products [PhBP3]-
Fe(N(D)-p-tolyl) (2b), [PhBP3]Fe(η5-cyclohexadienyl-d7) (3b), and
D2N-p-tolyl can be identified. Compound 2b is identified on the
basis of its 1H NMR resonances and the isotopically shifted N-D
stretch in its IR spectrum (Nujol: νND ) 2462 cm-1 (w)).
unsaturated substrates (e.g., azobenzene), appears to be stable in
aromatic solvents such as benzene.
In summary, the low-spin iron(III) imide 1 undergoes partial and
then complete hydrogenation under ambient conditions to release
aniline in what appears to be a well-defined, stepwise process. We
can directly observe the intermediate Fe(II) anilido species, 2, and
have provided evidence for the subsequent intermediacy of a
reactive FeII-H species that is trapped by benzene solvent to provide
3. Interesting mechanistic issues remain to be resolved and are
currently under investigation.
Acknowledgment. We thank the NSF (CHE-01232216) and
the DOE (PECASE) for financial support. The authors acknowledge
David M. Jenkins and Larry M. Henling for technical assistance.
1
Diamagnetic 3b is assigned from its 31P and H NMR spectra. In
particular, no cyclohexadienyl ring resonances are present in its
1H NMR spectrum due to complete deuteration. Compound 3b also
exhibits the expected ES-MS molecular ion peak at 828 m/z. The
organic byproduct D2N-p-tolyl is identified by GC-MS as the only
amine-containing product of the reaction.
Supporting Information Available: Detailed experimental pro-
cedures, characterization data, and crystallographic information. This
material is available free of charge via the Internet at http://pubs.acs.org.
The reduction of 1 to anilide 2 and then to 3 appears to proceed
in a stepwise fashion (Scheme 1). The detailed mechanism by which
these steps occur is clearly of interest but somewhat difficult to
unravel due to the paramagnetic nature of 1 and 2, in addition to
the presence of at least one paramagnetic side-product(s) that is
formed during the course of the reaction. A reasonable mechanistic
outline to suggest is as follows: The first step (1 + 1/2 H2 f 2)
involves the addition of H2 to 1 to generate an unobservable species
“[PhBP3]FeIII(H)(NHAr)”, which, if formed, must bimolecularly
release H2 to provide observable 2. A second addition of H2 would
occur at 2 to generate an unobservable “FeII-H” source (with loss
of H2N-p-tolyl) that adds to benzene via insertion.9 Evidence
consistent with a reactive “FeII-H” intermediate comes from the
following set of observations: First, the incubation of 2 under an
atmosphere of H2 at 25 °C in C6D6 slowly generates 3 (30% after
3 days) with concomitant evolution of H2N-p-tolyl (50% after 3
days). Second, the addition of KHBEt3 to a benzene solution of
[PhBP3]FeIICl generates 3 in high yield (77% isolated). Moreover,
when this latter reaction is carried out in THF rather than benzene
a new and diamagnetic species is generated that can be assigned
as the complex [PhBP3]FeII(HBEt3) (4) on the basis of its solution
References
(1) Smil, V. Enriching the Earth; MIT Press: Cambridge, MA, 2001.
(2) (a) Howard, J. B.; Rees, D. C. Chem. ReV. 1996, 96, 2965 and references
cited therein. (b) Burgess, B. K.; Lowe, D. J. Chem. ReV. 1996, 96, 2983.
(c) Einsle, O.; Tezcan, A.; Andrade, S. L. A.; Schmid, B.; Yoshida, M.;
Howard, J. B.; Rees, D. C. Science 2002, 297, 1696. (d) Thorneley, R.
N. F.; Lowe, D. In Molybdenum Enzymes; Spiro, T. G., Ed.; Wiley-
Interscience: New York, 1985. (e) Sellmann, D.; Sutter, J. Acc. Chem.
Res. 1997, 30, 460 (f) Yandulov, D. V.; Schrock, R. R. Science 2003,
301, 76. (g) Chatt, J.; Dilworth, J. R.; Richards, R. L. Chem. ReV. 1978,
78, 589. (h) Richards, R. L. Coord. Chem. ReV. 1996, 154, 83.
(3) (a) Verma, A. K.; Nazif, T. N.; Achim, C.; Lee, S. C. J. Am. Chem. Soc.
2000, 122, 11013. (b) Wagner, W. D.; Nakamoto, K. J. Am. Chem. Soc.
1989, 111, 1590. (c) Meyer, K.; Bill, E.; Mienert, B.; Weyhermuller, T.;
Wieghardt, K. J. Am. Chem. Soc. 1999, 121, 4859. (d) Jensen, M. P.;
Mehn, M. P.; Que, L., Jr. Angew. Chem., Int. Ed. 2003, 42, 4357.
(4) (a) Brown, S. D.; Betley, T. A.; Peters, J. C. J. Am. Chem. Soc. 2003,
125, 322. (b) Jenkins, D. M.; Betley, T. A.; Peters, J. C. J. Am. Chem.
Soc. 2002, 124, 11238. (c) Betley, T. A.; Peters, J. C. J. Am. Chem. Soc.
2003, 125, 10782.
(5) Cummins, C. C.; Baxter, S. M.; Wolczanski, P. T. J. Am. Chem. Soc.
1988, 110, 8731.
(6) For recent examples of terminal imide H2 chemistry, see: (a) Schmidt, J.
A. R.; Arnold, J. Organometallics 2002, 21, 3426. (b) Burckhardt, U.;
Casty, G. L.; Gavenonis, J.; Tilley, T. D. Organometallics 2002, 21, 3108.
(c) Cameron, T. M.; Ortiz, C. G.; Ghiviriga, I.; Abboud, K. A.; Boncella,
J. M. J. Am. Chem. Soc. 2002, 124, 922.
(7) Anilido complexes of divalent iron have been reported recently from direct
amination of an auxiliary ligand (see ref 3d) and by thermal degradation
of an intermediate hydrazido complex of iron(II). See: Smith, J. M.;
Lachicotte, R. J.; Holland, P. L. J. Am. Chem. Soc. 2003, 125, 15752.
(8) While 2 can be isolated and thoroughly characterized, including a
satisfactory combustion analysis, its 1H NMR spectra invariably show
minor impurities because of its tendency to slowly degrade upon
dissolution (see Supporting Information).
(9) Tilley has observed the intramolecular hydrogenation of an aromatic ring
upon exposure of a d0 tantalum imide to hydrogen. This process is believed
to occur via an intermediate hydride. See: Gavenonis, J.; Tilley, T. D. J.
Am. Chem. Soc. 2002, 124, 8536.
(10) Kaesz, H. D.; Saillant, R. B. Chem. ReV. 1972, 72, 231.
(11) See for example: (a) Rigaut, S.; Delville, M. H.; Astruc, D. J. Am. Chem.
Soc. 1997, 119, 11132. (b) Lee, S. S.; Lee, I. S.; Chung, Y. K.
Organometallics 1996, 15, 5428. (c) Meng, W. D.; Stephenson, G. R. J.
Organomet. Chem. 1989, 371, 355. (d) Whitesides, T. H.; Arhart, R. W.
J. Am. Chem. Soc. 1971, 93, 5296.
IR and NMR data in THF (νBH ) 2448 cm-1 31P NMR 55 ppm
;
(br s); 11B NMR δ 25.5 ppm (br s, HBEt3), -12.8 ppm (s,
PhB(CH2PPh2)3-)). Most importantly, complex 4 serves as a
“[PhBP3]FeII-H” equivalent and is instantly converted to 3 with
loss of BEt3 upon addition of benzene to a THF solution. We also
note that the hydrogenation of 1 in CD2Cl2 generates H2N-p-tolyl
and the chloride complex [PhBP3]FeIICl. Reactive metal hydrides
are known to exchange with halocarbons,10 and it is reasonable to
expect that a “[PhBP3]FeII-H” intermediate might behave similarly.
Whereas iron cyclohexadienyl complexes structurally related to
3 are known,11-13 their formation from the insertion of benzene
into a reactive Fe-H bond is, to our knowledge, unprecedented. A
curious reactivity comparison to note in this context concerns
Holland’s four-coordinate iron hydride dimer {LFeIIH}2, in which
L represents a bulky â-diketiminate ligand.7 This low-coordinate
hydride system is isolable and, while reactive toward certain
(12) Turculet, L.; Feldman, J. D.; Tilley, T. D. Organometallics 2003, 22, 4627.
(13) Metal hydride systems are known to catalyze the conversion of benzene
to cyclohexane in the presence of hydrogen. For a recent lead reference,
see: Su¨ss-Fink, G.; Faure, M.; Ward, T. R. Angew. Chem., Int. Ed. 2002,
41, 99.
JA0399122
9
J. AM. CHEM. SOC. VOL. 126, NO. 14, 2004 4539