H-H and Si-H bond addition to Fe≡NNR2 intermediates derived from N2

Article abstract of DOI:10.1021/ja400836u

The synthesis and characterization of Fe-diphosphineborane complexes are described in the context of N₂ functionalization chemistry. Iron aminoimides can be generated at room temperature under 1 atm N₂ and are shown to react with E-H

Full text of DOI:10.1021/ja400836u

Communication
pubs.acs.org/JACS
H−H and Si−H Bond Addition to FeNNR2 Intermediates Derived
from N2
Daniel L. M. Suess and Jonas C. Peters*
Division of Chemistry and Chemical Engineering, California Institute of Technology, Pasadena, California 91125, United States
S
* Supporting Information
Scheme 2
ABSTRACT: The synthesis and characterization of Fe−
diphosphineborane complexes are described in the context
of N2 functionalization chemistry. Iron aminoimides can
be generated at room temperature under 1 atm N2 and are
shown to react with E−H bonds from PhSiH3 and H2. The
resulting products derive from delivery of the E fragment
to Nα and the H atom to B. The flexibility and lability of
the Fe−BPh interactions in these complexes engender this
reactivity.
that might be more reactive than the Fe−triphosphineborane
[(TPB)Fe] system in 1. We now describe new (DPB)Fe
initrogen functionalization reactions using synthetic Fe
NNR2 complexes that react with nonpolar E−H bonds at RT,
D
complexes typically employ reductants in conjunction
thereby enabling the one-pot transformation of free N2 to an
Fe−hydrazido(−) species, the first such complex to be derived
from N2.
with electrophilic reagents;1 this strategy has allowed for the
reliable functionalization of Nβ in terminal Fe−N2 complexes
(Scheme 1). Whereas early transition metal N2 complexes
As an entry to useful (DPB)Fe synthons, we found that
reductive metalation of the isopropyl- and phenyl-substituted
DPB ligands8 PhB(o-iPr2PC6H4)2 (2a, iPrDPB) and PhB(o-
Ph2PC6H4)2 (2b, PhDPB) with FeBr2 and 1.0 equiv of Na/Hg
in C6H6 affords (DPB)FeBr complexes 3a and 3b in yields of
84 and 64%, respectively (Scheme 3). Brown 3a and 3b are
pseudotetrahedral S = 3/2 complexes (μeff = 3.8μB and 3.6μB in
C6D6 at RT, respectively) featuring η2-BC interactions that
have previously been observed in Cu and Ni complexes of this
ligand class.7,9 The structures determined by X-ray diffraction
(XRD) analysis (Figure 1 and Table 1 for 3a; for 3b, see the SI)
show close Fe−B [2.3243(11) and 2.330(4) Å, respectively]
and Fe−Cipso [2.2605(9) and 2.193(3) Å, respectively]
contacts, supporting the formulation of the η2-BC ligand as
both a donor via a filled arene π orbital and an acceptor via an
empty p orbital on B. In solution, 3a and 3b are Cs-symmetric
Scheme 1
display rich reactivity with nonpolar E−H (E = H, Si) bonds,2
such reactivity using Fe has little precedent. The addition of H2
to diiron-bridged nitrides has been demonstrated,3 and in a
recent report, N2-derived bridging iron nitrides were shown to
undergo hydrogenolysis to NH3.4 Hydrogenolysis of the
terminal Fe−imide functional group has likewise been
established,5 though these imides were not prepared from N2.
As such, we sought to generate Fe−aminoimides from N2 that
could undergo subsequent E−H bond addition across the Fe
NNR2 linkage as a method for Nα functionalization (Scheme
1).
1
as indicated by their H NMR spectra.
Further reduction of 3a with 1.0 equiv of Na/Hg under 1
atm N2 results in the formation of the dinuclear N2-bridged
1
complex (iPrDPB)Fe(μ-1,2-N2)Fe(iPrDPB) (4). The H NMR
spectrum of 4 in C6D6 indicates that the two Fe centers are
equivalent and have local Cs symmetry. The solution and solid-
state IR spectra of 4 lack an N−N stretch, suggesting that the
complex maintains its pseudocentrosymmetric dinuclear
structure in solution. The solution magnetic moment at RT
is 4.6μB, which is somewhat higher than the spin-only value of
4.0μB expected for two uncoupled S = 1 Fe centers.10 The two
pseudotetrahedral Fe centers in 4 have different local
geometries in the solid state. The geometry about FeA (Figure
In this context, we and others have studied Fe platforms that
can accommodate both N2 and terminal imide ligands,1a,c,5a,6
and we recently reported that Fe−aminoimide complex 1
(Scheme 2) can be derived from N2.1c Given previous studies
demonstrating H2 addition across the M−B bonds in related Ni
complexes,7 we expected that the Fe−B bond in 1 would
facilitate an E−H activation step. However, 1 does not react
with either H2 (1 atm) or PhSiH3 at room temperature (RT).
At higher temperatures [see the Supporting Information (SI)],
1 decomposes, and no tractable products were identified in the
presence of H2 or PhSiH3. We therefore explored the
development of an Fe−diphosphineborane [(DPB)Fe] system
Received: January 24, 2013
Published: March 11, 2013
© 2013 American Chemical Society
4938
dx.doi.org/10.1021/ja400836u | J. Am. Chem. Soc. 2013, 135, 4938−4941

Journal of the American Chemical Society
Communication
FeB. Since the two (DPB)Fe fragments are equivalent in
solution, the Fe−BCC interaction must be highly flexible, and
the solid-state bond metrics reflect the large range of local
geometries available to the Fe centers.
Scheme 3
Performing an identical reduction of the phenyl derivative 3b
using 1.0 equiv of Na/Hg does not trigger N2 binding but
instead generates the brown diamagnetic complex 5 that
contains an η7-BPh interaction; to our knowledge, this
coordination mode is unprecedented in the metal−borane
literature. The XRD structure of 5 (Figure 2) shows tight
Figure 2. Displacement ellipsoid (50%) representations of (left) 5 and
(right) 6a. For clarity, PR2 groups have been truncated and H atoms
and solvent molecules have been omitted. For 6a, only one of the two
molecules per asymmetric unit is shown. Selected distances (Å) for 5:
Fe−B, 2.2667(13); Fe−Cipso, 1.9669(11); Fe−Cortho, 2.090 (av); Fe−
Cmeta, 2.169 (av); Fe−Cpara, 2.1933(11).
Fe−η7-BPh distances. The bound Cipso atom is significantly
pyramidalized, as indicated by the sum of the C−C−C and two
B−C−C angles (342°). Further showing the significant
geometrical distortion of the bound arene is the contracted
B−Cipso−Cpara angle of 127.71(8)°. The η7-BPh binding mode
is maintained in solution, as evidenced by the upfield-shifted
1
aryl resonances in the H NMR spectrum (Hortho, 3.63 ppm;
H
meta, 3.24 ppm; Hpara, 6.25 ppm) and the 13C NMR spectrum
(Cipso, 106.77 ppm; Cortho, 99.41 ppm; Cmeta, 86.36 ppm; Cpara
78.73 ppm).
,
Each of the complexes 3a, 3b, 4, and 5 serves as a precursor
to an Fe−aminoimide complex derived from N2. In a procedure
nearly identical to that for generating 1,1c diamagnetic 6a and
6b can be accessed by stirring 3a and 3b, respectively, with 1.1
equiv of 1,2-bis(chlorodimethylsilyl)ethane and 3.1 equiv of
Na/Hg in THF under 1 atm N2. Alternatively, 4 or 5 may be
employed as starting materials in conjunction with 2.1 equiv of
Na/Hg. Both 6a and 6b are green in solution and dichroic
Figure 1. Displacement ellipsoid (50%) representations of (left) 3a
and (right) 4. For clarity, the PiPr2 groups have been truncated and the
H atoms have been omitted.
Table 1. Selected Bond Lengths (Å)
Fe−B
Fe−Cipso
Fe−Cortho
1
3a
2.3242(11)
2.3739(7)
2.3136(7)
2.3768(6)
2.4288(7)
2.859(5)
2.2605(9)
2.2516(6)
2.2133(6)
2.1492(5)
2.1440(6)
2.5483(11)
2.2714(7)
2.6642(7)
2.3403(6)
2.2266(6)
green/brown when crystalline. The H NMR spectrum of 6a
a
a
reveals its Cs symmetry in solution. In addition, the 1H
resonances of 6a attributed to the bound aryl ring are upfield-
shifted (Hortho, 5.10 ppm; Hmeta, 6.49 ppm; Hpara, 4.72 ppm).
This effect is not observed for 6b, perhaps because of
attenuated back-bonding in 6b due to the less electron-rich
metal center.
4: FeA
4: FeB
b
6a
7
−
−
b
a
FeA and FeB are the two unique Fe atoms per molecule. 6a contains
two molecules per asymmetric unit.
The solid-state structures of 6a (Figure 2) and 6b (see the
SI) are similar. For 6a, there are two molecules in the
asymmetric unit. The short Fe−N distances [1.6607(5) and
1.6657(5) Å for 6a; 1.6592(7) Å for 6b] are consistent with
those of other trigonal Fe(NR) linkages and imply an Fe
NNR triple bond.11 The bound arenes display alternating bond
lengths that vary between ca. 1.36 and 1.44 Å (see the SI).
Density functional theory (DFT) calculations (see SI) support
the formulation of 6a and 6b as typical pseudotetrahedral d6
Fe−imides11 that are similar to 1 except that one phosphine
donor in 1 has been replaced by the η3-BCC interaction in 6a
1) is distinguished by a short Fe−Cortho contact and relatively
long Fe−B and Fe−Cipso distances (Table 1), while FeB displays
somewhat shorter Fe−B and Fe−Cipso distances and a
negligible Fe−Cortho interaction. The phenyl ring bound to
FeA exhibits alternating C−C bond lengths between 1.4303(10)
and 1.3711(12) Å (see the SI), whereas this asymmetry is
negligible for the phenyl ring bound to FeB. These metrics
indicate that back-donation to the arene ring is more significant
for FeA and back-donation to the B atom is more significant for
4939
dx.doi.org/10.1021/ja400836u | J. Am. Chem. Soc. 2013, 135, 4938−4941

Journal of the American Chemical Society
Communication
and 6b. Although the presence of an Fe−B bond is not required
for the stability of pseudotetrahedral d6 Fe−imides,11,12 DFT
calculations on 6a, 1,1c and related Fe−imides6 show some
degree of Fe−B σ bonding. Quantifying the extent of Fe−B
bonding in these complexes thermodynamically is difficult
because the boranes are contained within the cage structures of
the ligands.
The reactions of E−H bonds with aminoimides 6a and 6b
were next examined. We were satisfied to observe that in
contrast to 1, the addition of 1.1 equiv of PhSiH3 to 6a at RT
readily generates a new orange species identified as the
trisilylhydrazido(−) product 7 resulting from hydrosilylation of
the Fe−N bond with delivery of SiH2Ph to Nα and H to B. To
our knowledge, this is the first Fe−hydrazido(−) complex to be
derived from N2, thereby adding to the body of previously
reported mononuclear Fe−hydrazido(−) model complexes.13
Having established this elementary step, we sought to combine
the formation of 6a and its subsequent hydrosilylation into a
single procedure. Accordingly, 7 can be generated in one pot
from 3a or 4 (eq 1).
H stretch at 3343 cm−1; the corresponding D2 addition product
shows νN−D at 2476 cm−1 (2441 cm−1 calcd). The broad,
intense νFe−H−B stretch at ca. 2100 cm−1 is also sensitive to
isotopic labeling (νFe−D−B ≈ 1550 cm−1). The connectivity of 8
was established by XRD analysis (Figure 3). In addition to the
formation of new N−H and Fe−H−B bonds, the structure of 8
reflects cleavage of the N−N bond, rearrangement of the
azadisilacyclopentane ring, and formal oxidation of one of the
phosphines. This overall transformation is sufficiently complex
that we felt it worthwhile to suggest a tentative mechanism
(Scheme 4). Hydrogenation of 6b to form A is analogous to the
Scheme 4
The IR spectrum of 7 exhibits an intense IR signal at 2090
cm−1 corresponding to the Si−H stretch and a broad, intense
IR stretch at ca. 2000 cm−1 corresponding to the Fe−H−B
functional group. The solution magnetic moment (μeff = 5.0μB,
C6D6, RT) indicates an S = 2 spin state. The N−N bond is
elongated from 1.326 Å (av) in 6a to 1.492(4) Å in 7 (Figure
3). Although both distances are consistent with N−N single
formation of 7 from 6a. The hydrazine rearrangement step to
form B has precedent for closely related disilylhydrazines.14
Intermediate B may be sufficiently unencumbered to allow for
η2-NN binding to give C, which could lead to direct N−N
cleavage and group transfer to a phosphine. Alternatively, the
N−N bond in C may be cleaved to form Fe(IV)−imide15 D,
which could then undergo group transfer to give 8. Cleavage of
the N−N bond of N2-derived ligands is rare for Fe,1c,4 and has
been observed for bimetallic early-transition-metal complex-
es.2h,16 In addition, the transformation of 6b to 8 is, to our
knowledge, only the second example of N2 functionalization
with H2 at a well-defined Fe complex.4
In conclusion, we have reported the generation of Fe−
aminoimides from N2 that undergo subsequent addition of
nonpolar E−H bonds. The significant flexibility of the Fe−BPh
interactions facilitates both the initial formation of the Fe−
aminoimide as well as the E−H activation step. Whereas
previous functionalization reactions of terminal Fe−N2 frag-
ments allow for derivatization of Nβ, this report demonstrates
that E−H addition to an unsaturated Fe−N bond is a viable
strategy for Nα functionalization.
Figure 3. Displacement ellipsoid (50%) representations of (left) 7 and
(right) 8. For clarity, PR2 groups have been truncated and H atoms
not located in the difference map and solvent molecules have been
omitted.
ASSOCIATED CONTENT
* Supporting Information
Experimental details, spectra, DFT calculations, and XRD
tables. This material is available free of charge via the Internet
at http://pubs.acs.org.
■
bonds, the comparatively short bond in 6a is due to the sp
hybridization of Nα and some degree of N−N multiple-bond
character. The very long N−N bond in 7 (longer than that in
free N2H4) is likely due to a high degree of steric pressure
exerted by its bulky Si and Fe substituents. The sum of the C−
B−C angles is 334°, reflecting the tetrahedral geometry of the
borohydride ligand.
Addition of 1 atm H2 at 50 °C to a C6H6 solution of 6b
affords a pale-brown solution from which colorless solids
identified as the S = 2 product 8 (μeff = 4.8μB, C6D6, RT) can
be isolated. Its IR spectrum contains a peak assigned to an N−
S
AUTHOR INFORMATION
Corresponding Author
jpeters@caltech.edu
Notes
■
The authors declare no competing financial interest.
4940
dx.doi.org/10.1021/ja400836u | J. Am. Chem. Soc. 2013, 135, 4938−4941

Journal of the American Chemical Society
Communication
ACKNOWLEDGMENTS
■
We acknowledge the NIH (GM070757) and the Beckman
Institute for funding and thank Lawrence Henling and Dr. Jens
Kaiser for assistance with XRD studies. We also acknowledge
the Gordon and Betty Moore Foundation, the Beckman
Institute, and the Sanofi-Aventis BRP at Caltech for their
support of the Molecular Observatory at Caltech. The SSRL is
operated for the DOE and supported by its Office of Biological
and Environmental Research, the NIH NIGMS (including
P41GM103393), and the NCRR (P41RR001209).
REFERENCES
■
(1) (a) Betley, T. A.; Peters, J. C. J. Am. Chem. Soc. 2003, 125, 10782.
(b) Lee, Y.; Mankad, N. P.; Peters, J. C. Nat. Chem. 2010, 2, 558.
(c) Moret, M.-E.; Peters, J. C. J. Am. Chem. Soc. 2011, 133, 18118.
(d) Yuki, M.; Tanaka, H.; Sasaki, K.; Miyake, Y.; Yoshizawa, K.;
Nishibayashi, Y. Nat. Commun. 2012, 3, 1254.
(2) (a) Fryzuk, M. D.; Love, J. B.; Rettig, S. J.; Young, V. G. Science
1997, 275, 1445. (b) Fryzuk, M. D.; MacKay, B. A.; Patrick, B. O. J.
Am. Chem. Soc. 2003, 125, 3234. (c) Pool, J. A.; Lobkovsky, E.; Chirik,
P. J. Nature 2004, 427, 527. (d) Pool, J. A.; Bernskoetter, W. H.;
Chirik, P. J. J. Am. Chem. Soc. 2004, 126, 14326. (e) Hirotsu, M.;
Fontaine, P. P.; Epshteyn, A.; Sita, L. R. J. Am. Chem. Soc. 2007, 129,
9284. (f) Hirotsu, M.; Fontaine, P. P.; Zavalij, P. Y.; Sita, L. R. J. Am.
Chem. Soc. 2007, 129, 12690. (g) Pun, D.; Bradley, C. A.; Lobkovsky,
E.; Keresztes, I.; Chirik, P. J. J. Am. Chem. Soc. 2008, 130, 14046.
(h) Semproni, S. P.; Lobkovsky, E.; Chirik, P. J. J. Am. Chem. Soc.
2011, 133, 10406.
(3) Brown, S. D.; Mehn, M. P.; Peters, J. C. J. Am. Chem. Soc. 2005,
127, 13146.
(4) Rodriguez, M. M.; Bill, E.; Brennessel, W. W.; Holland, P. L.
Science 2011, 334, 780.
(5) (a) Brown, S. D.; Peters, J. C. J. Am. Chem. Soc. 2004, 126, 4538.
(b) Bart, S. C.; Lobkovsky, E.; Bill, E.; Chirik, P. J. J. Am. Chem. Soc.
2006, 128, 5302.
(6) Moret, M.-E.; Peters, J. C. Angew. Chem., Int. Ed. 2011, 50, 2063.
(7) Harman, W. H.; Peters, J. C. J. Am. Chem. Soc. 2012, 134, 5080.
(8) (a) Bontemps, S.; Gornitzka, H.; Bouhadir, G.; Miqueu, K.;
Bourissou, D. Angew. Chem., Int. Ed. 2006, 45, 1611. (b) Sircoglou, M.;
Bontemps, S.; Mercy, M.; Saffon, N.; Takahashi, M.; Bouhadir, G.;
Maron, L.; Bourissou, D. Angew. Chem., Int. Ed. 2007, 46, 8583.
(9) (a) Emslie, D. J. H.; Cowie, B. E.; Kolpin, K. B. Dalton Trans.
2012, 41, 1101. (b) Sircoglou, M.; Bontemps, S.; Mercy, M.; Miqueu,
K.; Ladeira, S.; Saffon, N.; Maron, L.; Bouhadir, G.; Bourissou, D.
Inorg. Chem. 2010, 49, 3983.
(10) (a) Smith, J. M.; Lachicotte, R. J.; Pittard, K. A.; Cundari, T. R.;
Lukat-Rodgers, G.; Rodgers, K. R.; Holland, P. L. J. Am. Chem. Soc.
2001, 123, 9222. (b) Hendrich, M. P.; Gunderson, W.; Behan, R. K.;
Green, M. T.; Mehn, M. P.; Betley, T. A.; Lu, C. C.; Peters, J. C. Proc.
Natl. Acad. Sci. U.S.A. 2006, 103, 17107. (c) Field, L. D.; Guest, R. W.;
Turner, P. Inorg. Chem. 2010, 49, 9086.
(11) Saouma, C. T.; Peters, J. C. Coord. Chem. Rev. 2011, 255, 920.
(12) Brown, S. D.; Peters, J. C. J. Am. Chem. Soc. 2005, 127, 1913.
(13) Selected examples: (a) Smith, J. M.; Lachicotte, R. J.; Holland,
P. L. J. Am. Chem. Soc. 2003, 125, 15752. (b) Crossland, J. L.;
Balesdent, C. G.; Tyler, D. R. Dalton Trans. 2009, 4420. (c) Saouma,
C. T.; Kinney, R. A.; Hoffman, B. M.; Peters, J. C. Angew. Chem., Int.
Ed. 2011, 50, 3446.
(14) Pitt, C. G.; Skillern, K. R. Inorg. Nucl. Chem. Lett. 1966, 2, 237.
(15) Thomas, C. M.; Mankad, N. P.; Peters, J. C. J. Am. Chem. Soc.
2006, 128, 4956.
(16) Selected examples: (a) Peters, J. C.; Cherry, J.-P. F.; Thomas, J.
C.; Baraldo, L.; Mindiola, D. J.; Davis, W. M.; Cummins, C. C. J. Am.
Chem. Soc. 1999, 121, 10053. (b) Fryzuk, M. D. Acc. Chem. Res. 2009,
42, 127. (c) Knobloch, D. J.; Lobkovsky, E.; Chirik, P. J. Nat. Chem.
2009, 2, 30. (d) Knobloch, D. J.; Lobkovsky, E.; Chirik, P. J. J. Am.
Chem. Soc. 2010, 132, 10553.
4941
dx.doi.org/10.1021/ja400836u | J. Am. Chem. Soc. 2013, 135, 4938−4941