Putting chromium on the map for N2 reduction: Production of hydrazine and ammonia. A study of: Cis -M(N2)2 (M = Cr, Mo, W) bis(diphosphine) complexes

Article abstract of DOI:10.1039/c6cc03449g

The first complete structurally and spectroscopically characterized series of isostructural Group 6 N₂ complexes is reported. Protonolysis experiments on cis-[M(N₂)₂(P^EtN^RP^Et)₂] (M = Cr, Mo, W; R = 2,6-difluorobenzyl) reveal that only Cr affords N₂H₅⁺ and NH₄⁺ from the reduction of the N₂ ligands.

Full text of DOI:10.1039/c6cc03449g

ChemComm
View Article Online
View Journal
COMMUNICATION
Putting chromium on the map for N reduction:
2
production of hydrazine and ammonia. A study of
Cite this:DOI: 10.1039/c6cc03449g
cis-M(N ) (M = Cr, Mo, W) bis(diphosphine)
complexes†
2
2
Received 25th April 2016,
Accepted 15th June 2016
a
a
a
a
Jonathan D. Egbert, Molly O’Hagan, Eric S. Wiedner, R. Morris Bullock,
DOI: 10.1039/c6cc03449g
b
b
a
Nicholas A. Piro, W. Scott Kassel and Michael T. Mock*
www.rsc.org/chemcomm
The first complete structurally and spectroscopically characterized of complexes for all the group 6 metals with identical ligands has
series of isostructural Group 6 N
2
complexes is reported. Protonolysis prevented an unambiguous comparison based only on metal
Et
R
Et
7a,11
experiments on cis-[M(N
2
)
2
(P N P )
2
] (M = Cr, Mo, W; R = 2,6- identity.
Herein we report the first spectroscopic, electro-
+
+
difluorobenzyl) reveal that only Cr affords N
H
2 5
and NH
4
from chemical, and protonolysis study of a structurally identical series
Et
R Et
the reduction of the N ligands.
2
of group 6 bis(dinitrogen) complexes, cis-[M(N
M = Cr (1), Mo (2), W (3); R = 2,6-difluorobenzyl). The results
Transition metal dinitrogen complexes have been studied for of this metal-based comparison show the ability of zero-valent Cr
several decades, and have revealed a myriad of information to serve as an active metal for the reduction of N . Notably, of the
2 2 2
) (P N P ) ]
(
2
relevant to the mechanism of ammonia formation in biological group 6 metal complexes examined in this study, the Cr analogue
1
and heterogeneous N reduction processes. Large-scale ammonia exhibits the most activated N ligands and is the only complex to
2
2
synthesis from N and H by the Haber–Bosch process is critical produce N -derived hydrazine and ammonia upon the addition
2
2
2
2
for making fertilizers to maintain worldwide food production.
However, concerns over CO emissions from this century-old
process are motivating the development of alternative approaches.
For example, an electrocatalytic system for N reduction via ligand in THF, Scheme 1. Importantly, the reaction time and
of excess acid.
III/IV
2
Complexes 1–3 were prepared by Mg reduction of M
precursors in the presence of two equiv. of the PNP diphosphine
3
2
addition of protons and electrons, (akin to the mild reaction temperature were critical parameters in the synthesis of 1 and 2.
4
conditions employed by nitrogenase ) would provide a carbon- Complex 1 was prepared at ꢀ5 1C, as the reduction carried out
neutral approach to NH production.
at room temperature resulted in nearly no formation of 1.
3
Seminal studies on zero-valent group 6 complexes have Stirring for 60 h afforded a (B9 : 1) mixture of 1 and trans-
5
Et 2,6-F2-Bn Et
primarily focused on N
reports deciphering N reactivity on the basis of the identity of isomers were separated by precipitating 1 from a THF solution
the metal and phosphine ligands examined only Mo and W, as by adding cold pentane. In solution, only the Cr isomers show
2
reactivity at Mo and W. Moreover, [Cr(N
2
)
2
(P N
2
P ) ], trans-1 (Fig. S1 and S2, ESI†). The
2
6
12
very few related Cr(N
2
) complexes were known due to limited reversible interconversion, suggesting they are similar in energy.
7
stability toward binding N
2
. Recently, in efforts to develop
electrocatalysts for N reduction, our group has made advances
2
in understanding N bonding and reactivity of Cr by studying
2
8
9
mono- and bis(dinitrogen) complexes with cyclic 8-, 12-, and
10
1
6-membered phosphine ligands containing pendant amines.
While these unique Cr–N complexes provided a qualitative
assessment of spectroscopic and acid reactivity patterns between
2
2
N complexes of Cr, Mo, and W, the absence of an isolable series
a
Center for Molecular Electrocatalysis, Pacific Northwest National Laboratory,
P.O. Box 999, K2-57, Richland, WA 99352, USA. E-mail: michael.mock@pnnl.gov
Department of Chemistry, Villanova University, Villanova, PA, USA
b
†
Electronic supplementary information (ESI) available: Experimental procedures,
Et 2,6-F2-Bn Et
crystallographic details, and additional spectroscopic and electrochemical data. CCDC
445490 (1), 1445489 (trans-1), 1445488 (2), 1445491 (3). For ESI and crystallographic
Scheme 1 Synthesis of cis-[M(N
2 2
) (P N
2
P ) ]; (M = Cr, Mo, W),
1
X = Cl, Br. Conditions (temperature, reaction time): 1, ꢀ5 1C, 72 h; 2, ꢀ5 1C,
8 h; 3, 25 1C, 24 h.
data in CIF or other electronic format see DOI: 10.1039/c6cc03449g
This journal is ©The Royal Society of Chemistry 2016
Chem. Commun.

View Article Online
Communication
ChemComm
Fig. 1 Molecular structures of complexes 1 (left), 2 (middle), and 3 (right). Thermal ellipsoids are drawn at 50% probability. Hydrogen atoms are omitted
for clarity. Crystals of 2 and 3 contain two molecules per asymmetric unit with similar metric parameters; only one molecule is shown. Selected bond
distances (Å): (1) Cr–N1 = 1.864(2); Cr–N3 = 1.871(2); N1–N2 = 1.132(2); N3–N4 = 1.130(2); (2) Mo–N1 = 2.027(7); Mo–N3 = 2.020(6); N1–N2 = 1.112(8);
N3–N4 = 1.117(8); (3) W–N1 = 2.004(3); W–N3 = 1.991(3); N1–N2 = 1.125(4); N3–N4 = 1.134(4).
In THF at room temperature, a 3 : 1 mixture of trans-1 to 1
3
1
isomerized to 490% 1, by IR and P NMR spectroscopy (Fig. S3
and S4 ESI†). Heating the same sample to 40 1C for 3 days
results in partial conversion (ca. 20%) back to trans-1 with a
small amount of free PNP ligand present (Fig. S5, ESI†). In the
case of Mo, the reduction performed at ꢀ5 1C was vital to isolate
2
, the kinetic product, after 8 h of stirring. In a previous study,
Et 2,6-F2-Bn Et
we reported the formation of trans-[Mo(N
2 2
) (P N
2
P ) ] by
an analogous procedure after stirring for 20 h under a N
2
1
3
atmosphere at 25 1C. In the present case, reaction times
exceeding 8 h, even at ꢀ5 1C, result in partial isomerization to
the thermodynamically favored trans isomer. In contrast to 1
and 2, the synthesis of 3 was performed at 25 1C, and after 24 h
of stirring, only the cis isomer was observed. Isolated yields of
Et 2,6-F2-Bn Et
Fig. 2 Infrared spectra showing the nNN bands of cis-[M(N
2
)
2
(P N
2
P ) ]
recorded in THF at 25 1C: M = Cr (1), black, 1990, 1911; Mo (2), blue, 2012,
ꢀ1
1
950; W (3), red, 1987, 1925 cm .
2 2
the cis-M(N ) products increase going down the group, ca. expected to decrease going down the group on the basis of
6
,15
4
% for 1, 41% for 2, and 62% for 3. The higher yields for Mo more electron-releasing dp donation of the metal,
complex 2
ꢀ
1
and W may reflect a greater affinity to bind N
oxidation states during the reduction process.
2
at higher metal has the highest nNN bands at 2012, 1950 cm . The nNN bands
for 3 are 1987 and 1925 cm , and notably, 1 exhibits nNN bands
Complexes 1–3 were characterized by X-ray crystallography, at 1990 and 1911 cm and the lowest N–N force constant of the
ꢀ1
ꢀ
1
1
6
allowing for the first structural comparison of a full series of series (Table S1, ESI†). 1 also displays slight asymmetry in
group 6 bis(dinitrogen) complexes with identical ligands. In the band intensities in THF, indicative of dissimilarity in the
each case, single crystals were obtained by evaporation of a N–M–N bond angle compared to 2 and 3. Thus, enhanced N2
concentrated Et
in Fig. 1, are essentially isostructural, exhibiting a 6-coordinate bonding orbitals, or based on our earlier bonding assessment
octahedral geometry at the metal, and dinitrogen ligands in the of Cr–N complexes, a stronger polarization of the N ligand in
cis conformation. The P–M–P angle is B861 consistently across 1 decreases the N–N vibrational frequency.
2 2
O solution. The molecular structures, shown activation for 1 could be due to better dp overlap with N anti-
1
7
2
2
8
3
1
15
the metal series. Metal–ligand bond lengths are similar for
Characterization of complexes 1–3 by P and N NMR
complexes 2 and 3; the M–P and M–N bonds are ca. 2.45 Å spectroscopy afforded additional trends based on metal identity.
3
1
15
and 2.01 Å, respectively. However, those for 1 are shorter at ca. The P and N NMR spectra for complexes 1–3 are shown in
3
1
1
2
.35 Å and 1.87 Å.
The N–N bond lengths for the entire series fall within the complex displays two multiplets of an AA BB pattern, as
range 1.11–1.13 Å, consistent with an NRN triple bond of the expected for two distinct phosphorus environments. Descending
Fig. 3a and b, respectively. In the P{ H} NMR spectra, each
0
0
31
end-on bound N
compared to free N
phosphine complexes with cis N
2
ligands. This observed elongation of N
2
,
the group, the P NMR resonances appear at higher field.
at 1.0975 Å, is analogous to other group Resonances for 1 appear at d 37.9, 32.2, while trans-1 shows
2
8
,14
183
6
2
ligands.
The variance in a singlet at d 37.1. The resonances for 3 have
W satellites,
N–N bond lengths is quite small (ca. 0.02 Å). Distances for the
N–N bonds in 1 and 3 are ca. 1.13 Å, and, the N–N distances in 2
are ca. 1.11 Å.
JPW = 297 and 303 Hz.
1
5
15N
15N
15N
The N -labelled isotopologues 1 , 2 , and 3
were
2
1
5
prepared for N NMR spectral studies by exposing a degassed
1
5
The N
2
stretching frequencies in the infrared spectra provide THF-d
8
solution of the complex to a headspace of
N
2
gas.
1
5
a more quantitative measure of the activation of the N
2
ligands. ligand exchange was rapid in all cases, resulting in
N
2
N
2
1
5
1
The IR spectra collected in THF contain two nNN bands for 1–3, incorporation into the product within minutes. The N{ H}
as shown in Fig. 2. Although the energy of the nNN bands are NMR spectrum for each complex displays two signals for the
Chem. Commun.
This journal is ©The Royal Society of Chemistry 2016

View Article Online
ChemComm
Communication
3
1
1
15
1
15N 15N
8
Fig. 3 (a) P{ H} NMR spectra (202.4 MHz) of complexes 1–3 recorded at 25 1C in THF-d . (b) N{ H} NMR spectra (50.7 MHz) of complexes 1 , 2
15N
31
15
15N
and 3
recorded at 25 1C in THF-d
8
. The selected P and N NMR spectra of 1 contain trans-1 and trans-1 , respectively, for a comparison of data for
the cis and trans stereoisomers.
1
5
31
end-on bound
coupling) for the proximal nitrogen atoms (N
J15N–15N ca. 6–7 Hz) for the distal nitrogen atoms (N ). As in the kinetic potential shift resulting from rapid N2 loss upon Cr
N
2
ligands, a broad resonance (due to
P
Epa, for 1 is ca. 300 mV more negative than the E1/2 of 2 and 3.
I/0
p
), and a doublet The position of the Cr wave is likely due in part to a negative
(
d
3
1
15
P NMR spectra, the N resonances appear upfield upon oxidation. Kinetic analysis suggests that the formal reduction
descending the group. For 1 , the N signals appear close potential (E1 ) of 1 is also more negative than the E1/2 values
1
5N
15
0
0
together at d ꢀ7.3, ꢀ11.5, for N
d
and N
p
, respectively. For observed for 2 and 3. For example, E1 for 1 would be ꢀ1.12 V
1
5
9
ꢀ1
comparison between isomers, the N signals for N
d
and N
p
if the first-order rate constant for N
2
loss were 1 ꢁ 10
s
1
5N
of trans-1
28.0, ꢀ22.6, respectively. For 2
are nearly identical, however the position of the signals differ
appear at higher field in opposite positions, at d (Table S2, ESI†). Accordingly, these data are consistent with 1
1
5N
15N
ꢀ
and 3
the N
d
resonances being easier to oxidize than 2 and 3.
Protonolysis experiments revealed the outcome of metal-
1
5N
15N
+
for N at d ꢀ39.1 for 2
and d ꢀ60.4 for 3 . Thus, it is clear dependence on the reduction of N to ammonium (NH ) and
p
2
4
1
5
+
that the magnetic environment of the N ligands display a hydrazinium (N H ) products. In a typical reaction, a THF-d
8
2
2
5
periodic trend of increased magnetic shielding upon descending solution of complex 1, 2, or 3 was treated with 100 equiv. of
1
8
15
the group. However, the trend in N chemical shifts does not triflic acid (CF
3
SO
3
H) at ꢀ40 1C, according to Scheme 2. Results
15b
+
correlate to metal center basicity to release electron density to from five runs revealed complex 1 formed N
2
H
5
(avg. 0.22
+
+
+
the N
2
ligands based on the values of the nNN bands.
equiv. N
2
H
5
per Cr atom), and NH
4
(avg. 0.08 equiv. NH
ligands
oxidation potentials for this series of complexes to assess the (electrons originate from Cr), (Table S3, ESI†). The yields of
4
per
Cyclic voltammetry (CV) experiments probed the metal Cr atom) from the protonation and reduction of the N
2
1
9
+
+
5
1
trend of the electron density at the metal center with an NH4 and N H were quantified by H NMR spectroscopy using
2
identical ligand set. This trend could be correlated to the nNN 1,3,5-trimethoxy-benzene as an internal integration standard
1
5N
stretching frequencies and acid reactivity at N
2
. The CV of the (for additional details see ESI†). Acid addition to 1
confirmed
I/0
20
M
couple for complexes 1–3 were recorded in THF at a scan-rate the products are derived from the dinitrogen ligands. To
ꢀ1
of 0.1 V s (Fig. S6, ESI†). Complexes 2 and 3 exhibit quasi- obtain reduced N
reversible waves corresponding to the Mo and W redox
2
products, low temperatures are required, as
were not formed in HOTf addition to 1 at 25 1C.
couple with a half-wave potential, E1/2 = ꢀ1.04 V and ꢀ1.03 V, In contrast, treatment of 2 and 3 with excess triflic acid at
I/0
I/0
+
+
2 5 4
N H and NH
+
/0
I/0
2
respectively (vs. Cp Fe ). The M waves of 2 and 3 are similar
ꢀ1
despite 3 having nNN bands that appear 25 cm lower in energy.
The quasi-reversible nature of the M wave at low scan-rates,
ca. 0.1 V s , is likely due to N
I/0
ꢀ
1
2
ligand loss upon metal
oxidation. The waves do not become reversible at faster scan
ꢀ
1
rates, ca. n = 1 V s
.
In contrast to 2 and 3, complex 1 exhibits an irreversible,
I/0
anodic wave corresponding to the Cr oxidation, Epa = ꢀ1.36 V.
ꢀ
1
This wave is irreversible at all attempted scan rates (up to 20 V s ),
even when the CV was performed at ꢀ30 1C. The irreversibility of
I/0
Ph Bn
8
the Cr wave is analogous to the CV of cis-[Cr(N ) (P
N
2
) ] ,
2 2
2
2
I/0
but differs from the quasi-reversible Cr
wave of trans-
)]. Irreversibility of this wave suggests that
Ph Bn
10
[
Cr(N
2
)
2
(P
2 2
(PNP) ] is unstable, and that N ligand loss is rapid
4
N
4
I
+
[Cr (N
2
)
2
Et 2,6-F2-Bn Et
upon metal oxidation. Remarkably, the anodic peak potential, Scheme 2 Reaction of cis-[M(N
)
2 2
(P N
P )
2
] with excess HOTf.
This journal is ©The Royal Society of Chemistry 2016
Chem. Commun.

View Article Online
Communication
ChemComm
+
+
ꢀ
40 1C did not produce any detectible amount of NH
4
or N
by H NMR spectroscopy. Furthermore, treatment of a THF
solution of complexes 1, 2, or 3 with 100 equiv. H SO at 25 1C
H
2 5
6 W. Hussain, G. J. Leigh, H. M. Ali, C. J. Pickett and D. A. Rankin,
1
21
J. Chem. Soc., Dalton Trans., 1984, 1703–1708.
7
(a) H. H. Karsch, Angew. Chem., Int. Ed., 1977, 16, 56–57;
(b) G. S. Girolami, J. E. Salt, G. Wilkinson, M. Thornton-Pett and
M. B. Hursthouse, J. Am. Chem. Soc., 1983, 105, 5954–5956.
2
4
+
+ 22
did not form detectible amounts of NH4 or N H .
2
5
8
M. T. Mock, S. Chen, R. Rousseau, M. J. O’Hagan, W. G. Dougherty,
W. S. Kassel, D. L. DuBois and R. M. Bullock, Chem. Commun., 2011,
In conclusion, we prepared N complexes of Cr, Mo, and W
with identical ligands and examined their spectroscopic and
electrochemical properties to more clearly understand the
2
47, 12212–12214.
9 M. T. Mock, A. W. Pierpont, J. D. Egbert, M. O’Hagan, S. Chen,
R. M. Bullock, W. G. Dougherty, W. S. Kassel and R. Rousseau, Inorg.
Chem., 2015, 54, 4827–4839.
metal-dependent N
complex contains the most activated N
negative oxidation potential of the series. Greater N
for 1 may reflect better dp overlap with N2 anti-bonding
2
activation and reactivity. The Cr–N
ligands and the most
activation
2
2
10 M. T. Mock, S. Chen, M. O’Hagan, R. Rousseau, W. G. Dougherty,
W. S. Kassel and R. M. Bullock, J. Am. Chem. Soc., 2013, 135,
2
1
1493–11496.
1
1 The group 6 complexes cis-[M(N ) (PMe ) ] (M = Cr, Mo, W) have
2 2 3 4
+
+
orbitals. Cr was the only complex to form NH4 and N H
from reduction of the N ligands upon addition of acid. Future
2
5
been reported. The Cr analogue is unstable under ambient condi-
tions (ref. 7a); (a) E. Carmona, A. Galindo, M. L. Poveda and
R. D. Rodgers, Inorg. Chem., 1985, 24, 4033–4039; (b) E. Carmona,
J. M. Marin, M. L. Poveda, J. L. Atwood and R. D. Rogers, J. Am.
Chem. Soc., 1983, 105, 3014–3022.
2
studies aim to identify Cr–N
pathway, and to utilize Cr to develop an electrocatalytic system
for N reduction to N and NH
x y 2
H intermediates in the N reduction
H
2 4
3
.
12 H. Gailus, C. Woitha and D. Rehder, J. Chem. Soc., Dalton Trans.,
994, 3471–3477.
2
1
This work was supported as part of the Center for Molecular
Electrocatalysis, an Energy Frontier Research Center funded by
the U.S. Department of Energy (U. S. DOE), Office of Science,
Office of Basic Energy Sciences. Pacific Northwest National
Laboratory is operated by Battelle for the U. S. DOE.
1
3 L. A. Labios, C. J. Weiss, J. D. Egbert, S. Lense, R. M. Bullock,
W. G. Dougherty, W. S. Kassel and M. T. Mock, Z. Anorg. Allg. Chem.,
2015, 641, 105–117.
1
4 (a) M. Yuki, Y. Miyake, Y. Nishibayashi, I. Wakiji and M. Hidai,
Organometallics, 2008, 27, 3947–3953; (b) C. J. Weiss, A. N. Groves,
M. T. Mock, W. G. Dougherty, W. S. Kassel, M. L. Helm, D. L. DuBois
and R. M. Bullock, Dalton Trans., 2012, 41, 4517–4529.
1
5 (a) F. Tuczek, K. H. Horn and N. Lehnert, Coord. Chem. Rev., 2003,
Notes and references
245, 107–120; (b) D. F. Shriver, Acc. Chem. Res., 1970, 3, 231–238;
1
(a) D. V. Yandulov and R. R. Schrock, Science, 2003, 301, 76–78;
(c) G. J. Leigh, Acc. Chem. Res., 1992, 25, 177–181.
(
(
(
b) B. A. MacKay and M. D. Fryzuk, Chem. Rev., 2004, 104, 385–401; 16 F. A. Cotton and C. S. Kraihanzel, J. Am. Chem. Soc., 1962, 84, 4432–4438.
c) R. R. Schrock, Angew. Chem., Int. Ed., 2008, 47, 5512–5522; 17 (a) D. Sellmann, Angew. Chem., Int. Ed., 1974, 13, 639–649;
d) K. Arashiba, Y. Miyake and Y. Nishibayashi, Nat. Chem., 2011,
(b) R. Hoffmann and D. L. DuBois, Nouv. J. Chem., 1977, 1, 479–492.
, 120–125; (e) J. S. Anderson, J. Rittle and J. C. Peters, Nature, 2013, 18 S. Donavon-Mtunzi, R. L. Richards and J. Mason, J. Chem. Soc.,
01, 84–87; ( f ) K. Grubel, W. W. Brennessel, B. Q. Mercado and
Dalton Trans., 1984, 469–474.
3
5
+
+
2 5 4
P. L. Holland, J. Am. Chem. Soc., 2014, 136, 16807–16816; (g) I. Coric, 19 The peak area of N H (broad singlet, 10.8 ppm) and NH (1 : 1 : 1
B. Q. Mercado, E. Bill, D. J. Vinyard and P. L. Holland, Nature, 2015,
triplet, 7.0 ppm, JHN = 51 Hz) were determined by manual integra-
tion and by line-fitting analysis using the CRAFT NMR software
program.
5
4
26, 96–99; (h) J. Rittle and J. C. Peters, J. Am. Chem. Soc., 2016, 138,
243–4248; (i) T. J. Del Castillo, N. B. Thompson and J. C. Peters,
15N
15
+
15
+
15
J. Am. Chem. Soc., 2016, 138, 5341–5350.
20 Protonolysis of 1
afforded
N
2
H
5
and NH
4
by N NMR
2
3
V. Smil, Enriching the Earth: Fritz Haber, Carl Bosch, and the Trans-
formation of World Food Production, MIT Press, Cambridge, MA,
spectroscopy at ꢀ326 ppm and ꢀ363 ppm, respectively. In addition,
Et 2,6-F2-Bn Et
1
31
[HP N
spectroscopy.
P ][OTf ] was observed by
H
and
P NMR
2
001.
(a) S. Licht, B. Cui, B. Wang, F. F. Li, J. Lau and S. Liu, Science, 2014, 21 The formation of M–H products for 2 and 3 was apparent based
31
3
45, 637–640; (b) C. J. van der Ham, M. T. Koper and
upon multiplet resonances (due to P coupling) in the high-field
1
D. G. Hetterscheid, Chem. Soc. Rev., 2014, 43, 5183–5191.
region of the H NMR spectra.
4
5
B. M. Hoffman, D. Lukoyanov, Z. Y. Yang, D. R. Dean and 22 (a) Y. Tanabe, S. Kuriyama, K. Arashiba, Y. Miyake, K. Nakajima and
L. C. Seefeldt, Chem. Rev., 2014, 114, 4041–4062.
J. Chatt, J. R. Dilworth and R. L. Richards, Chem. Rev., 1978, 78,
Y. Nishibayashi, Chem. Commun., 2013, 49, 9290–9292; (b) K. Arashiba,
S. Kuriyama, K. Nakajima and Y. Nishibayashi, Chem. Commun., 2013,
49, 11215–11217.
5
89–625.
Chem. Commun.
This journal is ©The Royal Society of Chemistry 2016