Precursors to dinitrogen reduction: Structures and reactivity of trans-[Fe(DMeOPrPE)2(η2-H2)H]+ and trans-[Fe(DMeOPrPE)2(N2)H]+

Article abstract of DOI:10.1039/b911066f

Trans-[Fe(DMeOPrPE)₂(H₂)H]⁺ and trans-[Fe(DMeOPrPE)₂(N₂)H]⁺ (DMeOPrPE = 1,2-bis(dimethoxypropylphosphino)ethane) were synthesized and their structures determined by X-ray crystallography. T

Full text of DOI:10.1039/b911066f

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/dalton | Dalton Transactions
Precursors to dinitrogen reduction: structures and reactivity of
trans-[Fe(DMeOPrPE)2(g2-H2)H]+ and trans-[Fe(DMeOPrPE)2(N2)H]+†
Justin L. Crossland, Douglas M. Young, Lev N. Zakharov and David R. Tyler*
Received 5th June 2009, Accepted 5th August 2009
First published as an Advance Article on the web 3rd September 2009
DOI: 10.1039/b911066f
trans-[Fe(DMeOPrPE)2(H2)H]+ and trans-[Fe(DMeOPrPE)2(N2)H]+ (DMeOPrPE =
1,2-bis(dimethoxypropylphosphino)ethane) were synthesized and their structures determined by X-ray
crystallography. These complexes are important species in a dinitrogen reduction scheme involving
protonation of an iron(0) dinitrogen complex to produce ammonia. The rates of substitution of the
coordinated H2 and N2 molecules with acetonitrile were monitored in a variety of organic solvents. The
coordinated N2 substituted ~6 times faster than H2, but surprisingly the solvent had little effect on the
observed rates. The results suggest that the H2 molecule in trans-[Fe(DMeOPrPE)2(H2)H]+ does not
participate in hydrogen bonding to the bulk solvent, as was previously observed in the analogous Ru
complex. The deprotonation of trans-[Fe(DMeOPrPE)2(N2)H]+ to yield Fe(DMeOPrPE)2N2 was
investigated in the presence of a variety of anions, and it was found that the anion facilitates the reaction
through an ion-pairing interaction in which the anion removes electron density from the hydride ligand.
Introduction
Since the discovery of the first dinitrogen complex by Allen and
Senoff in 1965,1 extensive research has focused on creating a
homogeneous catalyst capable of reducing dinitrogen to ammonia
at room temperature and pressure.2–4 Many remarkable advances
have been made toward this goal,5–7 including the achievement
of a catalytic cycle.8 The chemistry of iron dinitrogen complexes
represents a growing subset of this field,9,10 largely due to the fact
that increasing evidence implicates an iron reaction site in the
mechanism of nitrogenase enzymes.11–13 Several groups, including
our own, have found that addition of acid to five-coordinate iron
dinitrogen phosphine species yields varying amounts of ammonia
and/or hydrazine.14–19 Of particular interest to our group was
a cycle pioneered by Leigh (Scheme 1).20 In the Leigh cycle,
Scheme 1 Leigh cycle for the reduction of N2 to NH3 in Fe(P2)2N2
a trans-[Fe(P2)2(H2)H]+ complex (P2 = a bidentate phosphine)
systems. The direct reaction of the trans-[Fe(P2)2Cl2]+ with dihydrogen
is generated from trans-[Fe(P2)2Cl2]+, a hydride source, and an
represents our modification of the Leigh cycle.
acid. Following exchange of H2 for N2, the hydride ligand can
be “reductively” deprotonated to yield the five-coordinate iron(0)
dinitrogen complex Fe(P2)2N2. Protonation of this complex results
in moderate yields of ammonia and/or hydrazine. We were able to
modify this cycle by generating the trans-[Fe(P2)2(H2)H]+ complex
directly from the dichloride starting material and dihydrogen;21,22
thus, the electrons used for the reduction of dinitrogen to ammonia
come directly from H2 (Scheme 1).
details on the synthesis and structural characterization of trans-
[Fe(DMeOPrPE)2(H2)H]+ and trans-[Fe(DMeOPrPE)2(N2)H]+,
by exploring the substitution kinetics of these complexes
with acetonitrile in a variety of organic solvents, and by
describing the role of the anion in the deprotonation of
trans-[Fe(DMeOPrPE)2(N2)H]+.
We previously reported that the generation of
Fe(DMeOPrPE)2N2 (DMeOPrPE = 1,2-bis(dimethoxypropyl-
phosphino)ethane) and the protonation of this complex with
1 M triflic acid resulted in the formation of ammonia (15%) and
hydrazine (2%).16 Here we expand this report by describing further
Experimental
Materials and reagents
All manipulations were carried out in a Vacuum Atmospheres
Co. glove box (argon- or N2-filled) or on a Schlenk line using
argon or nitrogen. HPLC grade THF, hexane, and diethyl ether
(Burdick and Jackson) were dried and deoxygenated by passing
them, under an argon atmosphere, through commercial columns
of CuO followed by alumina. Toluene (Aldrich) was distilled under
N2 from CaH2 and degassed by three freeze–pump–thaw cycles.
Department of Chemistry, University of Oregon, Eugene, Oregon 97403,
USA. E-mail: dtyler@uoregon.edu
† Electronic supplementary information (ESI) available: NMR spectra
and crystallographic data including the.cif files for I, II, and III. CCDC
reference numbers 735060–735062. For ESI and crystallographic data in
CIF or other electronic format see DOI: 10.1039/b911066f
This journal is
The Royal Society of Chemistry 2009
Dalton Trans., 2009, 9253–9259 | 9253
©

View Article Online
23
-3
-1
Commercially available reagents were used as received. TlBArF
˚
1.221 g cm , m = 0.396 mm , Z = 2, l = 0.71073 A, T = 100 K,
24 706 reflections collected, 13 266 independent (Rint =0.0130), R1
[I > 2s(I)] = 0.0311, wR2 [I > 2s(I)] = 0.0811, R1 (all data) =
0.0346, wR2 (all data) = 0.0842.
24
and TlBF4 were synthesized by literature procedures. trans-
Fe(DMeOPrPE)2Cl225 and trans-[Fe(DMeOPrPE)2(CH3CN)H]+22
were synthesized as previously reported. Deuterated solvents were
purchased from Cambridge Isotope Laboratories and used as
received.
Crystal data for trans-[Fe(DMeOPrPE)2Cl][BPh4] (II).
¯
C60H100BClFeO8P4, M = 1175.39, triclinic, space group P1, a =
˚
13.5949(11), b = 15.4389(13), c = 16.0182(13) A, a = 97.9400(10),
Instrumentation
◦
3
˚
b = 94.0320(10), g = 103.9340(10) , V = 3213.0(5) A , Dc =
31P{ H} and H NMR spectra were recorded on either a Varian
Unity/Inova 300 spectrometer at an operating frequency of
299.94 (1H) and 121.42 (31P) MHz or a Varian Unity/Inova 500
spectrometer at an operating frequency of 500.62 (1H) and 202.45
(31P) MHz. The 1H and 31P chemical shifts were referenced to the
solvent peak and to an external standard of 1% H3PO4 in D2O,
respectively. NMR samples were sealed under argon or nitrogen in
7 mm J. Young tubes. Note that the 1H NMR data for the methyl
and methylene regions in complexes containing the DMeOPrPE
ligand were generally broad and uninformative and therefore
are not reported in the synthetic descriptions below. Elemental
analyses were performed by Robertson Microlit Laboratories.
Mass spectra were obtained using an Agilent 1100 LC/MS Mass
Spectrometer. The samples were dissolved in Et2O and introduced
into the ionization head (ESI) using the infusion method.
˚
1
1
-3
-1
1.215 g cm , m = 0.425 mm , Z = 2, l = 0.71073 A, T = 173 K,
30 038 reflections collected, 13 652 independent (Rint =0.0450), R1
[I > 2s(I)] = 0.0625, wR2 [I > 2s(I)] = 0.1304, R1 (all data) =
0.0933, wR2 (all data) = 0.1504.
Crystal data for trans-[Fe(DMeOPrPE)2(N2)H][BPh4] (III).
¯
C60H101BFeN2O8P4, M = 1168.97, triclinic, space group P1, a =
˚
13.466(2), b = 15.443(3), c = 16.219(3) A, a = 97.2619(3), b =
◦
3
-3
˚
94.169(3), g = 104.408(3) , V = 3221.5(10) A , Dc = 1.205 g cm ,
-1
˚
m = 0.384 mm , Z = 2, l = 0.71073 A, T = 173 K, 36 541
reflections collected, 14 419 independent (Rint =0.0194), R1 [I >
2s(I)] = 0.0451, wR2 [I > 2s(I)] = 0.1264, R1 (all data) = 0.0561,
wR2 (all data) = 0.1367.
Methods
Synthesis of trans-[Fe(DMeOPrPE)2(H2)H][BPh4] (I). trans-
Fe(DMeOPrPE)2Cl2 (1.86 g, 2.086 mmol), NaBPh4 (1.43 g,
4.172 mmol), and Proton Sponge (0.45 g, 2.086 mmol) were
combined as solids in a 120 mL Fischer-Porter tube. THF (15 mL)
and Et2O (15 mL) were then added, and the resulting solution was
immediately charged with 1 atm of H2. The solution turned from
green to orange to a faint yellow with the production of a white
precipitate (NaCl) over the course of several hours. The reaction
was allowed to stir for 48 h to ensure completion. The solution
was then filtered through Celite under an argon atmosphere. The
complex was precipitated as an oil by addition of hexane. The
oil was redissolved in toluene, filtered through Celite, and again
precipitated with hexane. The remaining oil was triturated with
hexane to yield 2.32 g (97% yield) of a tan colored powder. X-
Ray quality crystals were grown by slow evaporation of a THF
solution. Anal. calcd. for C60H103BFeO8P4: C, 63.02; H, 9.09%.
X-Ray crystallography
X-Ray diffraction data were collected at 100
K (trans-
[Fe(DMeOPrPE)2(H2)H][BPh4]) or 173 K ([Fe(DMeOPrPE)2Cl]-
[BPh4] and trans-[Fe(DMeOPrPE)2(N2)H][BPh4]) on a Bruker
Apex CCD diffractometer with Mo-Ka radiation (l = 0.71073
26
˚
A). Absorption corrections were applied by SADABS. The
structures were found by direct methods and calculations of
difference Fourier maps. All non-hydrogen atoms were refined
with anisotropic thermal parameters. The C and O atoms in
two of the –(CH2)3OMe groups of [Fe(DMeOPrPE)2Cl][BPh4]
are disordered over two positions (in 71/29 and 74/26 ratios)
and were refined with restrictions; the standard C–C and C–O
distances were used in the refinements as targets for corresponding
bond distances. H atoms in such groups were treated in calculated
positions and refined in a rigid group model. All H atoms in
trans-[Fe(DMeOPrPE)2(H2)H][BPh4] were found on the difference
F-map and refined with isotropic thermal parameters. It was
found that the Fe atom in trans-[Fe(DMeOPrPE)2(H2)H][BPh4]
was disordered over two positions (in a 88 : 12 ratio) related to
two opposite orientations of the HFe(H2) fragment. The Fe1 and
Fe1a positions are out from the average plane of the four P atoms
1
Found: C, 62.98; H, 9.30%. 31P{ H} NMR (toluene-d8): d 85.0
2
1
(s). 31P NMR (toluene-d8): d 85.0 (d, JP–H = 47 Hz). H NMR
(toluene-d8) of the hydride region: d -11.1 (s, br) and d -15.1
(quintet, 2JH–P = 49 Hz).
Synthesis of [Fe(DMeOPrPE)2Cl][BPh4] (II). NaBPh4
(0.384 g, 1.12 mmol) was added to a Et2O solution of trans-
Fe(DMeOPrPE)2Cl2 (0.1 g, 0.112 mmol) under argon. After
stirring for 1 h the solvent was evaporated. The yellow residue
was extracted into toluene and the solution was filtered through
Celite. The complex was precipitated as a yellow solid by addition
of hexane. The yellow powder was washed with hexane followed
by diethyl ether (0.097 g, 74% yield). Crystals suitable for X-ray
diffraction were grown from a saturated diethyl ether solution.
˚
bonded to the Fe atom on 0.14 and 0.22 A, respectively. The H2
and H atoms corresponding to one of two possible orientations of
the Fe atom were found on the residual density and refined without
restrictions with occupation factor m = 0.88 similar to that for the
Fe1 atom. The second orientation of this fragment, corresponding
to an occupation factor of m = 0.12, was not found and these H
atoms were not taken into consideration. All calculations were
performed by the Bruker SHELXTL 6.10 package.
1
31P{ H} NMR (toluene-d8) at 193 K: d 55.9 (s). 31P NMR
Crystal data for trans-[Fe(DMeOPrPE)2(H2)H][BPh4] (I).
(toluene-d8) at 193 K: d 55.9 (s). No 31P resonances were observed
at room temperature. Anal. calcd. for C60H100BClFeO8P4: C,
61.31; H, 8.58%. Found: C, 60.98; H, 8.39%. ESI (Et2O, +ve):
855.4 [Fe(DMeOPrPE)2Cl]+.
¯
C60H103BFeO8P4, M = 1142.96, triclinic, space group P1, a =
˚
10.8357(10), b = 16.9005(16), c = 18.2311(17) A, a = 90.013(2),
◦
3
˚
b = 105.896(1), g = 103.852(2) , V = 3109.8(5) A , Dc
9254 | Dalton Trans., 2009, 9253–9259
=
This journal is
The Royal Society of Chemistry 2009
©

View Article Online
Synthesis of trans-[Fe(DMeOPrPE)2(N2)H][BPh4] (III).
A
following time intervals were used: a spectrum was taken every
30 s for the first 3 min; a spectrum was collected at 1 min intervals
in the following 6 min; a spectrum was collected at 2 min intervals
in the following 12 min; and the interval was increased to 5 min
for the remaining 30 min.
THF–Et2O solution of I prepared by the method above was
charged with 2 atm of N2 in a Fischer-Porter tube and stirred
for 48 h. The Fischer-Porter tube was vented and refilled with N2
several times over the course of the reaction to remove free H2.
The solution color changed from light yellow to light brown over
the course of the reaction. The solvent was evaporated under an
N2 atmosphere and the resulting brown oil was triturated with
hexanes to yield 2.28 g of a tan solid (94% yield). X-Ray quality
crystals were grown by hexane diffusion into a THF solution.
Anal. calcd. for C60H101BFeN2O8P4: C, 61.62; H, 8.71; N, 2.40%.
Found: C, 61.94; H, 8.97; N, 2.25%. IR(KBr): (nNN) 2088 cm-1.
Test for hydrogen bonding in trans-[Fe(DMeOPrPE)2(H2)H]-
[BPh4]. A solution of I in toluene-d8 (0.00746 M, 0.3 mL) was
placed in an NMR tube fitted with a septum. The tube was then
placed into the NMR magnet and allowed to equilibrate at -40 ◦C
1
for 15 min. After this equilibration period, the H spectrum was
acquired and referenced to the toluene resonances. The sample
was then ejected, 10 mL (60 equivalents) of acetone-d6 was added,
and the sample placed back into the magnet. The procedure was
repeated until 1000 equivalents of acetone had been added. The
H2 resonance shifted ~0.04 ppm downfield upon addition of the
acetone; however, other resonances of I also underwent shifts of
similar magnitudes, both downfield and upfield in direction.
1
31P{ H} NMR (toluene-d8): d 75.8 (s). 31P NMR (toluene-d8): d
2
1
75.8 (d, JP–H = 49 Hz). H NMR (toluene-d8) of the hydride
region: d -18.6 (quintet, 2JH–P = 49 Hz).
Generation of trans-[Fe(DMeOPrPE)2(H2)H]X (X = BF4, OTf,
PF6, BArF). These compounds were prepared analogously to I
using the appropriate counter-ion source: TlBF4, TlOTf, TlPF6, or
TlBArF. The NMR characterization of these complexes matched
that reported for the BPh4 complex. These complexes were not
isolated.
Results and discussion
Synthesis of dihydrogen and dinitrogen complexes
Generation of trans-[Fe(DMeOPrPE)2(N2)H]X (X = BF4, OTf,
PF6, BArF). These compounds were prepared analogously to
III using the appropriate trans-[Fe(DMeOPrPE)2(H2)H]X starting
2
The iron dihydrogen complex trans-[Fe(DMeOPrPE)2(h -H2)H]+
-
was previously synthesized as the PF6 salt; however, attempts to
isolate this complex as a solid were unsuccessful.16,22 Following the
same synthetic procedure, but using NaBPh4 instead of TlPF6, an
isolable solid (I) was obtained (eqn (1)).
1
material. These complexes were not isolated. 31P{ H} NMR
(toluene-d8): d 75.5–76.9 (s). 31P NMR (toluene-d8): d 75.5–76.9
2
1
(d, JP–H = 49 Hz). H NMR (toluene-d8) of the hydride region:
2
BF4, d -18.35 (quintet, JP–H = 49 Hz); OTf, d -18.4 (quintet,
2
2JP–H = 49 Hz); BPh4, d -18.6 (quintet, JP–H = 49 Hz); BArF,
d -18.8 (quintet, 2JP–H = 49 Hz).
(1)
Generation of Fe(DMeOPrPE)2(N2). trans-[Fe(DMeOPrPE)2-
(N2)H][X] (X = BPh4, BF4, OTf, PF6, BArF) (30 mg) was dissolved
t
in 0.6 mL of toluene-d8. BuOK (2 equivalents) was added as a
solid to the solution under an N2 atmosphere. The solution color
changed from pale brown to bright orange over the course of the
reaction. The reaction was agitated for 16 h. These complexes
were not isolated. All of these complexes were shown to undergo
The solution characterization of
I
by NMR spec-
troscopy matched the previously reported data for the trans-
2
[Fe(DMeOPrPE)2(h -H2)H][PF6] complex,22 with a single res-
1
onance observed in the 31P{ H} spectrum (85.0 ppm) and a
1
deprotonation except the BArF containing complex. 31P{ H}
2
broad singlet (-11.1 ppm) and quintet (-15.1 ppm, JHP
=
NMR (toluene-d8): d 79.5 (s). 31P NMR (toluene-d8): d 79.5 (s).
◦
1
49 Hz) observed in the low temperature (-40 C) H spectrum.
X-Ray quality crystals of I were grown by slow evaporation
of a THF solution (Fig. 1). All hydrogen atoms were located,
Substitution kinetic experiments. In an NMR tube fitted with
a septum, trans-[Fe(DMeOPrPE)2(H2)H]BPh4 (I) (0.3 mL of a
0.00746 M solution in toluene-d8, 0.00224 mmol), a solvent
(0.3 mL of either toluene, acetone, tetrahydrofuran, dimethyl-
formamide, dimethylacetamide, hexamethylphosphoramide), and
acetonitrile (12 mL, 0.230 mmol) were sequentially added under
Ar. Immediately after addition of the acetonitrile, the reaction
2
allowing visualization of the intact h dihydrogen ligand. The
structure closely matches the previously published X-ray struc-
tures of trans-[Fe(DPPE)2(H2)H]+ and trans-[Fe(DMPE)2(H2)H]+
(DPPE = 1,2-bis(diphenylphosphino)ethane; DMPE = 1,2-
bis(dimethylphosphino)ethane).27,28 Although the dihydrogen lig-
and was clearly found, X-ray methods are unreliable in accurately
measuring the H–H bond distance, as evidenced by the fact that
1
was monitored by 31P{ H} NMR spectroscopy. For the first
25 min, a spectrum was taken every 5 min. In the following
30 min, a spectrum was collected at 10 min intervals. For the
remaining 6 h, the interval was increased to 30 min. The con-
centration of trans-[Fe(DMeOPrPE)2H(H2)]BPh4 was determined
˚
the measured H–H bond length (0.69 A) is shorter than that of
˚
free dihydrogen (0.74 A) and significantly shorter than the bond
length determined by NMR methods.29
1
by 31P{ H} peak integrations. The rate constants were then
Previous work showed that the reaction of trans-
Fe(DMeOPrPE)2Cl2 with H2 proceeds through a stepwise mech-
anism (Scheme 2); displacement of one chloride ligand with H2
occurs first, followed by heterolysis of the coordinated H2 to form
the hydride ligand, and finally displacement of the second chloride
with another equivalent of H2.22
obtained by fitting the data (concentration versus time) with a
single parameter exponential decay function using SigmaPlot
software. The kinetic data for the reaction of acetonitrile with
trans-[Fe(DMeOPrPE)2(N2)H]BPh4 (III) were acquired in the
same manner (identical concentration and solvents), except the
This journal is
The Royal Society of Chemistry 2009
Dalton Trans., 2009, 9253–9259 | 9255
©

View Article Online
Scheme 2 Mechanism of the reaction of trans-Fe(DMeOPrPE)2Cl2 with H2.
Fig. 2 Molecular structure of [Fe(DMeOPrPE)2Cl]+ (II). Ellipsoids are
drawn at 50% probability. The tetraphenylborate anion and hydrogen
atoms have been omitted for clarity.
2
Fig. 1 Molecular structure of trans-[Fe(DMeOPrPE)2(h -H2)H][BPh4]
(I). Ellipsoids are drawn at 50% probability. The tetraphenylborate anion
and hydrogen atoms of the phosphine ligands have been omitted for clarity.
I by a ligand substitution reaction with N2 (eqn (2)) at 1 atm.
Because I is slightly more stable than III, the reaction vessel must
be purged with N2 several times to remove any residual H2 to
achieve complete conversion.
Substitution of the first chloride with H2 likely proceeds through
a dissociative mechanism because this reaction does not proceed
in organic solvents without the addition of a chloride abstracting
reagent. In support of this hypothesis, [Fe(DMeOPrPE)2Cl]+
(II) was synthesized by reacting trans-Fe(DMeOPrPE)2Cl2 with
NaBPh4 in diethyl ether in the absence of H2.30 The product was
isolated as a yellow powder and X-ray quality crystals were grown
from a saturated diethyl ether solution (Fig. 2). The five-coordinate
structure is best described as having a slightly distorted square-
pyramidal geometry: using the angular structural parameter t
defined by Addison et al., the complex has only 13% trigonal
bipyrimidal distortion.32 The chloride ligand of II occupies the
apical position and the P–Fe–Cl angles range from 92.69◦ to
(2)
As with the trans-[Fe(DMeOPrPE)2(H2)H]+ complex, previous
attempts to isolate the trans-[Fe(DMeOPrPE)2(N2)H]+ complex
-
-
as the PF6 salt were unsuccessful.22 Using the BPh4 counterion,
complex III was isolated as a tan solid. Again, the NMR
characterization of III matched with the previously reported
data,22 with a singlet at 75.8 ppm in the 31P{ H} spectrum and a
1
◦
98.93 . The Fe–Cl bond length (2.35 A) of II is unchanged
hydride resonance at -18.6 ppm (2JHP = 49 Hz) in the 1H spectrum.
To determine the solid-state structure, light brown crystals of III
were grown by layering a THF solution with hexanes and allowing
the solution to stand under an N2 atmosphere for ~1 week. The
molecular structure of III shows the end-on bonded dinitrogen
ligand trans to the hydride ligand (Fig. 3). The iron–phosphorus
˚
˚
from that of the dichloride starting material (2.35 A), while the
Fe–P bonds (2.28–2.32 A) are slightly lengthened. The five-
25
˚
coordinate complex II readily binds H2 in solution to form trans-
[Fe(DMeOPrPE)2(H2)Cl]+, as evidenced by 31P and 1H NMR
spectroscopy, consistent with the mechanism in Scheme 2.33
The dihydrogen complex I is a convenient starting material
for the generation of trans-[Fe(DMeOPrPE)2(N2)H]+. Complex
III, trans-[Fe(DMeOPrPE)2(N2)H][BPh4], was synthesized from
˚
bond lengths in III range from 2.23–2.24 A; these bond lengths
closely match the DMPE34 and DEPE35 analogs. The N–N bond
˚
length of 1.11 A shows slight elongation compared with free
9256 | Dalton Trans., 2009, 9253–9259
This journal is The Royal Society of Chemistry 2009
©

View Article Online
Table 1 Summary of rate constants for I and III in various solvents, as
well as solvent hydrogen-bond accepting strengths (b) and solvent polarity
(ET(30)) parameters42
Solvent
I kobs (¥ 10-4)/s-1
III kobs (¥ 10-4)/s-1
b
ET(30)
Toluene
Acetone
THF
DMF
DMA
1.70 0.03
2.16 0.06
1.71 0.04
1.71 0.02
1.58 0.04
1.57 0.02
11.22 0.16
13.10 0.26
11.36 0.21
8.40 0.13
8.40 0.14
8.12 0.18
0.11
0.43
0.55
0.69
0.76
1.05
33.9
42.2
37.4
43.8
43.7
40.9
HMPA
(0.00746 M) of I or III in toluene-d8 (0.3 mL) was mixed with an
equal volume of one the solvents listed above and to that solution
a 100-fold excess of acetonitrile was added to ensure pseudo first-
order kinetics. The rate of substitution was then determined by
1
monitoring the disappearance of I or III in the 31P{ H} NMR
spectrum (Fig. 4). The hydrogen bond accepting strengths of the
solvents were quantified using the b parameter. Because these
complexes are charged and the substitution is expected to proceed
by a dissociative mechanism, solvent polarity could also affect the
observed rates. Consequently, the solvent polarity was quantified
using the ET(30) parameter.42 The data and results are summarized
in Table 1.
Fig. 3 ORTEP representation of trans-[Fe(DMeOPrPE)2(N2)H]+ (III).
Ellipsoids are drawn at 50% probability. The tetraphenylborate counter-ion
and hydrogen atoms of the phosphine ligands have been omitted for clarity.
˚
˚
dinitrogen (1.10 A), with the N–N bond length falling in between
34
35
˚
the DMPE (1.13 A) and DEPE (1.07 A) complexes.
Effect of solvent on the rate of H2 and N2 substitution
The substitution of H2 and N2 by various small molecules is a
commonly observed reaction because these ligands are typically
weakly bonded.36,37 Both I and III are important species in
the Leigh-type dinitrogen reduction cycle and their substitution
reactivity was studied to gain insights into how to improve the
yields of ammonia. It was previously shown with the trans-
[Ru(DMeOPrPE)2(H2)H]+ complex that a coordinated dihydrogen
molecule can act as a hydrogen bond donor to a neutral acceptor
molecule in solution, an interaction termed dihydrogen hydrogen
bonding (DHHB).38 If the coordinated H2 in I were capable of
donating a hydrogen bond to the bulk solvent, we wanted to
explore how this would affect the reactivity. The substitution
reactions of H2 and N2 in I and III, respectively, by acetonitrile
were chosen for study because these reactions have been studied
in great detail in analogous systems39–41 and the product of
the reaction, trans-[Fe(DMeOPrPE)2(MeCN)H]+, has previously
been characterized.22 Complexes I and III are amenable to studies
where the solvent needs to be varied because the DMeOPrPE
ligand provides solubility in a spectrum of solvents. The BPh4
anion also provides a weakly interacting counterion, minimizing
any potential ion-pairing effects.
To probe the effect of solvent on the rate of substitution, the
following solvents were used: toluene, acetone, tetrahydrofuran
(THF), dimethylformamide (DMF), dimethylacetamide (DMA),
and hexamethylphosphoramide (HMPA). These solvents were
chosen because both I and III were stable in them and they
represent a spectrum of hydrogen-bond accepting strengths with-
out having any hydrogen-bond donating ability. A stock solution
Fig. 4 Sample kinetic trace of the concentration of I as a function of time
in a toluene–DMF (50 : 50) solvent mixture.
As can be seen in Table 1, the rate of N2 substitution with
acetonitrile occurs ~6 times faster than H2 substitution. The
solvent has little effect on the rates of substitution. Upon changing
from a non-polar solvent like toluene to a polar solvent like
HMPA, the rate constants change by less than a factor of 2
for both I and III. Furthermore, the rate constants for both
I and III do not trend with hydrogen bond accepting strength
(b) or with solvent polarity (ET(30)). This result suggests that
either hydrogen bonding to the coordinated dihydrogen in I is
very weak and does not affect the lability of the H2 molecule
or that hydrogen bonding to coordinated H2 is not occurring
in this system. Unfortunately, two of the three tests previously
used to determine the hydrogen bonding ability of H2 in the
trans-[Ru(DMeOPrPE)2(H2)H]+ complex could not be used for I
because the pyridine-N-oxide probe molecule used in the prior
study of the trans-[Ru(DMeOPrPE)2(H2)H]+ complex readily
displaced the H2 molecule in I.43 The third test for hydrogen
-
This journal is
The Royal Society of Chemistry 2009
Dalton Trans., 2009, 9253–9259 | 9257
©

View Article Online
bonding, titrating I with a hydrogen bond accepting solvent and
monitoring the chemical shift of the H2 resonance, yielded no
shift in the H2 resonance attributable to hydrogen bonding. This
result, combined with the kinetic data in Table 1, suggests that the
H2 molecule in I is either unable to donate a hydrogen bond, in
contrast to the Ru analog, or that if DHHB is occurring then it is
too weak to have any effect on the substitution reactivity.
anion is too large to approach the hydride ligand, and the hydride
ligand is not activated toward deprotonation. This hypothesis is
merely speculative and further studies would be needed to confirm
this hypothesis. However, this data does show that ion-pairing
interactions can greatly affect the reactivity of metal hydrides, and
thus the choice of anion needs to be considered when designing
cationic metal hydride systems.
Anion effects in the deprotonation of
Conclusion
trans-[Fe(DMeOPrPE)2(N2)H]+
The trans-[Fe(DMeOPrPE)2(H2)H]+ and trans-[Fe(DMeOPrPE)2-
(N2)H]+ complexes were synthesized and characterized both in
solution and in the solid-state. Both molecules are important
species in a Leigh-type dinitrogen reduction cycle and their
substitution reactivity was studied to gain insights into how to
improve the yields of ammonia. The coordinated H2 and N2
molecules in these complexes are weakly bonded and can be
displaced by a wide variety of small molecules. The substitution
rates with acetonitrile were monitored to determine the relative
substitution rates of H2 versus N2. It was determined that the rate
of N2 substitution occurred ~6 times quicker than H2 substitution,
but neither complex showed any rate dependence on the solvent.
The inability of solvent to impact the substitution of the H2
complex was of particular interest because our recent work showed
that coordinated H2 can participate in hydrogen bonding in the
analogous Ru complex.38 However, 1H NMR experiments suggest
that the H2 ligand is not involved in hydrogen bonding in the
trans-[Fe(DMeOPrPE)2(H2)H]+ complex. Consequently, this is
one reason why the solvent has no influence on the substitution
rate in the Fe complex. In a final set of experiments, the effect of the
anion on the deprotonation of III to yield Fe(DMeOPrPE)2N2 was
explored. The results suggest that an ion-pairing interaction of the
anion with the hydride ligand removes electron density from the
hydride ligand and thereby facilitates the deprotonation reaction.
The deprotonation of the trans-[Fe(DMeOPrPE)2(N2)H]+ com-
plex is a key step in the Leigh cycle for producing ammonia
(Scheme 1). In order to determine if there was an anion effect on
the deprotonation reaction, the trans-[Fe(DMeOPrPE)2(N2)H]+
complex was synthesized with various anions. The complexes
with various counterions were synthesized analogously to I and
III using the appropriate chloride abstractor and anion source
(Scheme 3). Spectroscopic characterization by NMR (31P and 1H)
showed the resulting metal complexes to be identical to the metal
complexes in molecules I and III.
Scheme
3
Synthetic scheme for trans-[Fe(DMeOPrPE)2(N2)H][X]
complexes.
Deprotonation of the trans-[Fe(DMeOPrPE)2(N2)H][X] com-
t
plexes was then performed using 2 equivalents of BuOK, and
the completion of the reaction was determined by a shift in
1
the 31P{ H} resonance (~3 ppm downfield), the loss of hydride
coupling in the 31P spectrum, and the disappearance of the hydride
1
peak in the H NMR spectrum (Table 2). All of the complexes,
-
-
with the exception of the BArF complex (BArF = tetrakis(3,5-
bis(trifluoromethyl)phenyl)borate), were shown to undergo depro-
tonation to yield Fe(DMeOPrPE)2N2 within 16 h (Table 2).
Closer inspection of the 1H NMR spectra (Table 2) reveals
that the hydride resonance shifts downfield as the size of the
anion decreases, with the largest anion (BArF-) being the only
complex that was not deprotonated. This could be explained
by an ion-pairing phenomenon in which the anion is closely
associated with the hydride ligand and assists in the deprotonation
reaction.45–47 It is proposed that the interaction of the anion with
the hydride ligand decreases the electron density of the hydride
Acknowledgements
We thank the NSF (CHE-0809393) for funding.
References
1 A. D. Allen and C. V. Senoff, J. Chem. Soc., 1965, 621–622.
2 J. Chatt, J. R. Dilworth and R. L. Richards, Chem. Rev., 1978, 78,
589–625.
3 M. Hidai and Y. Mizobe, Chem. Rev., 1995, 95, 1115–1133.
4 B. A. MacKay and M. D. Fryzuk, Chem. Rev., 2004, 104, 385–401.
5 M. D. Fryzuk, J. B. Love, S. J. Rettig and V. G. Young, Science, 1997,
275, 1445–1447.
t
and helps facilitate the deprotonation by BuOK. As the anions
6 Y. Nishibayashi, S. Iwai and M. Hidai, Science, 1998, 279, 540–544.
7 J. A. Pool, E. Lobkovsky and P. J. Chirik, Nature, 2004, 427, 527–529.
8 D. V. Yandulov and R. R. Schrock, Science, 2003, 301, 76–78.
9 J. M. Smith, A. R. Sadique, T. R. Cundari, K. R. Rodgers, G. Lukat-
Rodgers, R. J. Lachicotte, C. J. Flaschenriem, J. Vela and P. L. Holland,
J. Am. Chem. Soc., 2006, 128, 756–769.
become larger, the approach to the hydride ligand becomes more
hindered due to steric interactions with the phosphine ligands,
and the ion-pairing interaction is diminished. Thus, the BArF
-
Table 2 NMR data of trans-[Fe(DMeOPrPE)2(N2)H][X] complexes
10 T. A. Betley and J. C. Peters, J. Am. Chem. Soc., 2003, 125, 10782–10783.
11 P. C. Dos Santos, R. Y. Igarashi, H. Lee, B. M. Hoffman, D. R. Dean
and L. C. Seefeldt, Acc. Chem. Res., 2005, 38, 208–214.
12 B. M. Barney, R. Y. Igarashi, P. C. Dos Santos, D. R. Dean and L. C.
Seefeldt, J. Biol. Chem., 2004, 279, 53621–53624.
13 P. C. Dos Santos, R. Y. Igarashi, H. Lee, B. M. Hoffman, D. R. Dean
and L. C. Seefeldt, Biochemistry, 2003, 42, 9102–9109.
14 D. Hall and G. J. Leigh, J. Chem. Soc., Dalton Trans., 1996, 3539–
3541.
Anion
Initial 31P d (ppm)
31P d w/base (ppm)
Hydride d (ppm)
-
BF4
PF6
76.9
75.8
76.0
75.7
75.5
79.5
79.8
79.3
79.6
75.5
-18.35
-18.4
-18.4
-18.6
-18.8
-
OTf-
-
BPh4
-
BArF
9258 | Dalton Trans., 2009, 9253–9259
This journal is
The Royal Society of Chemistry 2009
©

View Article Online
15 T. A. George, D. J. Rose, Y. Chang, Q. Chen and J. Zubieta, Inorg.
Chem., 1995, 34, 1295–1298.
16 J. D. Gilbertson, N. K. Szymczak and D. R. Tyler, J. Am. Chem. Soc.,
2005, 127, 10184–10185.
17 N. P. Mankad, M. T. Whited and J. C. Peters, Angew. Chem., Int. Ed.,
2007, 46, 5768–5771.
31 O. Franke, B. E. Wiesler, N. Lehnert, C. Nather, V. Ksenofontov, J.
Neuhausen and F. Tuczek, Inorg. Chem., 2002, 41, 3491–3499.
32 A. W. Addison, T. N. Rao, J. Reedijk, J. van Rijn J. and G. G. Verschoor,
J. Chem. Soc., Dalton Trans., 1984, 1349–1356.
33 trans-[Fe(DMeOPrPE)2(H2)Cl]+ has been previously synthesized and
characterized.22.
18 It should be noted that Komiya’s group reported that addition of acid
to Fe(DEPE)2N2 yielded no ammonia or hydrazine.19.
19 M. Hirano, M. Akita, T. Morikita, H. Kubo, A. Fukuoka and S.
Komiya, J. Chem. Soc., Dalton Trans., 1997, 3453–3458.
20 G. J. Leigh, J. Organomet. Chem., 2004, 689, 3999–4005.
21 J. D. Gilbertson, N. K. Szymczak and D. R. Tyler, Inorg. Chem., 2004,
43, 3341–3343.
22 J. D. Gilbertson, N. K. Szymczak, J. L. Crossland, W. K. Miller, D. K.
Lyon, B. M. Foxman, J. Davis and D. R. Tyler, Inorg. Chem., 2007, 46,
1205–1214.
34 A. Hills, D. L. Hughes, M. Jimenez-Tenorio and G. J. Leigh,
J. Organomet. Chem., 1990, 391, C41–C44.
35 I. E. Buys, L. D. Field, T. W. Hambley and A. E. D. McQueen, Acta
Crystallogr., Sect. C: Cryst. Struct. Commun., 1993, 49, 1056–1059.
36 P. G. Jessop and R. H. Morris, Coord. Chem. Rev., 1992, 121, 155–284.
37 P. Giannoccaro, M. Rossi and A. Sacco, Coord. Chem. Rev., 1972, 8,
77–79.
38 N. K. Szymczak, L. N. Zakharov and D. R. Tyler, J. Am. Chem. Soc.,
2006, 128, 15830–15835.
39 C. A. Helleren, R. A. Henderson and G. J. Leigh, J. Chem. Soc., Dalton
Trans., 1999, 1213–1220.
23 R. P. Hughes, D. C. Lindner, A. L. Rheingold and G. P. A. Yap, Inorg.
Chem., 1997, 36, 1726–1727.
40 M. G. Basallote, J. Dura´n, M. J. Ferna´ndez-Trujillo and M. A. Ma´n˜ez,
J. Organomet. Chem., 2000, 609, 29–35.
24 F. J. Arna´iz, J. Chem. Ed., 1997, 74, 1332–1333.
25 W. K. Miller, J. D. Gilbertson, C. Leiva-Paredes, P. R. Bernatis, T. J. R.
Weakley, D. K. Lyon and D. R. Tyler, Inorg. Chem., 2002, 41, 5453–
5465.
41 M. G. Basallote, J. Dura´n, M. J. Ferna´ndez-Trujillo, G. Gonza´lez, M. A.
Ma´n˜ez and M. Mart´ınez, Inorg. Chem., 1998, 37, 1623–1628.
42 Y. Marcus, Chem. Soc. Rev., 1993, 22, 409–416.
26 G. M. Sheldrick, SADABS (2.01), Bruker/Siemens Area Detector
Absorption Correction Program, Bruker AXS, Madison, WI, 1998.
27 J. S. Ricci, T. F. Koetzle, M. T. Bautista, T. M. Hofstede, R. H. Morris
and J. F. Sawyer, J. Am. Chem. Soc., 1989, 111, 8823–8827.
28 A. Hills, D. L. Hughes, M. Jimenez-Tenorio, G. J. Leigh and A. T.
Rowley, J. Chem. Soc., Dalton Trans., 1993, 3041–3049.
43 If an excess of pyridine-N-oxide is added to I it replaces all ligands
forming the hexakis pyridine-N-oxide complex.44.
44 J. L. Crossland, L. N. Zakharov and D. R. Tyler, Acta Crystallogr.,
Sect. E: Struct. Rep. Online, 2007, 63, m1196–m1197.
45 M. Y. Darensbourg and M. M. Ludvig, Inorg. Chem., 1986, 25, 2894–
2898.
29 The H–H bond length was previously determined in solution using the
46 J. M. Hanckel and M. Y. Darensbourg, J. Am. Chem. Soc., 1983, 105,
21
˚
˚
T1(min) (0.85–1.07 A) and J(H,D) (0.92 A) methods. .
30 An analogous five-coordinate complex, [Fe(DEPE)2Cl][BPh4], was
previously synthesized.31.
6979–6980.
47 L. M. Epstein, E. S. Shubina, A. N. Krylov, A. Z. Kreindlin and M. I.
Rybinskaya, J. Organomet. Chem., 1993, 460, 87–89.
This journal is
The Royal Society of Chemistry 2009
Dalton Trans., 2009, 9253–9259 | 9259
©