Synthesis, characterization and the thermal isomerization mechanism of a series of trans-tetracarbonyl-bis(phosphinite)molybdenum(0) complexes and an evaluation of their potential to form heterobimetallic compounds

Article abstract of DOI:10.1016/j.jorganchem.2013.10.015

The reactions of coordinated phosphorus-donor ligands have previously been used to prepare bimetallic trans-[cis-Mo(CO)₄(PPh ₂NHCH₂CH₂NCH(o-C₆H₄)O) ₂]M (M = Ni(II), Cu(II)) complexes. Similar bimetallic complexes containing trans-tetracarbonyl-bis(phosphinite)molybdenum(0) centers could be even more interesting because they could serve as molecular gyroscopes. In an attempt to prepare such complexes, the three intermediate compounds trans-Mo(CO)₄(PPh₂Cl)₂ (1), trans-Mo(CO) ₄(PPh₂NHCH₂CH₂NH₂) ₂ (2), and trans-Mo(CO)₄(PPh₂NHCH ₂CH₂NCH(o-C₆H₄)OH)₂ (3) were successfully synthesized and fully characterized by multinuclear NMR spectroscopy, elemental analysis and X-ray crystallography. However, the reaction of 3 with Ni(OAc)₂(H₂O)₄ yielded a mixture of crude products with the previously reported trans-[cis-Mo(CO) ₄(PPh₂NHCH₂CH₂NCH(o-C6H ₄)O)₂]Ni being the only product successfully isolated. The cis product was formed due to facile transecis isomerization during the coordination of the Ni(II) to 3. To gain more insight into such isomerizations, the transecis isomerizations of 2 and 3 have been followed by ³¹P{¹H} NMR spectroscopy. The reactions follow reversible first order kinetics and appear to display intramolecular mechanisms due to the lack of carbonyl substitutions of the phosphinamidite ligands when the isomerizations are performed under CO atmospheres. Eyring analyses performed on the two complexes show that they have nearly identical activation energies (2: ΔH = 10 ± 2 kJ mol^-1, ΔS = 39±5 J mol ^-1 K^-1; 3: ΔH = 11±2 kJ mol^-1, ΔS = 41 ± 7 J mol^-1 K^-1).

Full text of DOI:10.1016/j.jorganchem.2013.10.015

Journal of Organometallic Chemistry 751 (2014) 695e702
Contents lists available at ScienceDirect
Journal of Organometallic Chemistry
journal homepage: www.elsevier.com/locate/jorganchem
Synthesis, characterization and the thermal isomerization mechanism
of a series of trans-tetracarbonyl-bis(phosphinite)molybdenum(0)
complexes and an evaluation of their potential to form
heterobimetallic compounds
*
Justin R. Martin, Samantha D. Hastings, Jason L. Freeman, Gary M. Gray
University of Alabama at Birmingham, 901 14th Street South, Birmingham, AL 35294, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
The reactions of coordinated phosphorus-donor ligands have previously been used to prepare bimetallic
trans-[cis-Mo(CO)4(PPh2NHCH2CH2NCH(o-C6H4)O)2]M (M ¼ Ni(II), Cu(II)) complexes. Similar bimetallic
complexes containing trans-tetracarbonyl-bis(phosphinite)molybdenum(0) centers could be even more
interesting because they could serve as molecular gyroscopes. In an attempt to prepare such complexes,
the three intermediate compounds trans-Mo(CO)4(PPh2Cl)2 (1), trans-Mo(CO)4(PPh2NHCH2CH2NH2)2 (2),
and trans-Mo(CO)4(PPh2NHCH2CH2NCH(o-C6H4)OH)2 (3) were successfully synthesized and fully char-
acterized by multinuclear NMR spectroscopy, elemental analysis and X-ray crystallography. However, the
reaction of 3 with Ni(OAc)2(H2O)4 yielded a mixture of crude products with the previously reported
trans-[cis-Mo(CO)4(PPh2NHCH2CH2NCH(o-C6H4)O)2]Ni being the only product successfully isolated. The
cis product was formed due to facile transecis isomerization during the coordination of the Ni(II) to 3. To
gain more insight into such isomerizations, the transecis isomerizations of 2 and 3 have been followed
by 31P{1H} NMR spectroscopy. The reactions follow reversible first order kinetics and appear to display
intramolecular mechanisms due to the lack of carbonyl substitutions of the phosphinamidite ligands
when the isomerizations are performed under CO atmospheres. Eyring analyses performed on the two
Received 13 August 2013
Received in revised form
6 September 2013
Accepted 9 October 2013
Keywords:
Molybdenum
Phosphinamidite ligands
Isomerization
Eyring analysis
Synthesis
complexes show that they have nearly identical activation energies (2:
D
K
ꢁ1).
H ¼ 10 ꢀ 2 kJ molꢁ1
,
D
S ¼ 39 ꢀ 5 J molꢁ1
K
ꢁ1; 3:
D
H ¼ 11 ꢀ 2 kJ molꢁ1
,
D
S ¼ 41 ꢀ 7 J molꢁ1
Ó 2013 Elsevier B.V. All rights reserved.
1. Introduction
sequences have also been used to prepare trinuclear complexes
with bridging
gands [8].
u-salicylaldiminato-substituted phosphinamidite li-
Transition metal complexes with ligands that give rise to steri-
cally well-defined coordination environments are useful in homo-
geneous catalysis and in the design of molecular machines, such as
rotors [1e5]. One potential approach to actualize these well-
defined environments is to incorporate a second metal center
into the complex that would provide a fixed coordination envi-
ronment in the ligand backbone. We have previously demonstrated
that it is possible to accomplish this through functionalization of
coordinated phosphinamidite ligands in octahedral cis-tetra-
carbonyl-bis(phosphinamidite)molybdenum(0) complexes with
terminal salicylaldiminato moieties that can form trans square
planar complexes with Ni(II) [6] and Cu(II) [7]. Similar reaction
In light of the successful syntheses of the bi- and tri-metallic
complexes described in the previous paragraph, we were inter-
ested in using similar methods to generate isomers of the bimetallic
complexes containing trans-tetracarbonyl-bis(phosphinamidite)
molybdenum(0) centers. Such complexes can be considered ‘mo-
lecular gyroscopes’ because they restrict molecular rotation to the
equatorial carbonyl ligands about the octahedral metal center.
Molecular machines that can perform rotor functions such as these
are becoming increasingly popular in the fields of nanofluidics and
photonics [1,9]. Additionally, the incorporation of an electric dipole
moment in these complexes could be a method to drive unidirec-
tional rotation, which is a necessary feature of functional rotor
systems [10].
In this manuscript, we describe our attempts to use the stepwise
* Corresponding author. Tel.: þ1 205 934 8094.
synthesis shown in Scheme
1 to prepare the trans-[trans-
E-mail addresses: martin89@uab.edu (J.R. Martin), sdelmont@uab.edu
(S.D. Hastings), gmgray@uab.edu (G.M. Gray).
Mo(CO)4(PPh2NHCH2CH2NCH(o-C6H4)O)2]Ni, 4, heterobimetallic
0022-328X/$ e see front matter Ó 2013 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.jorganchem.2013.10.015

696
J.R. Martin et al. / Journal of Organometallic Chemistry 751 (2014) 695e702
Scheme 1. Proposed reaction scheme for the preparation of the trans-[trans-Mo(CO)4(PPh2NHCH2CH2NCH(o-C6H4)O)2]Ni, 4, heterobimetallic molecular gyroscope.
molecular gyroscope. The intermediates have been characterized
with elemental analysis, multinuclear NMR spectroscopy and X-ray
crystallography. During the course of this work, we discovered that
the trans-Mo(CO)4(Ph2PNHR)2 intermediates undergo relatively
coordination geometries. The trends in the 13C{1H} NMR chemical
shifts of the carbonyl ligands are quite similar to those previously
reported for the carbonyls cis to similar phosphinite ligands in
Mo(CO)5(Ph2PX) complexes (X ¼ Cl,
d
13C{1H} ¼ 208.23 ppm;
rapid
transecis
isomerizations
whereas
the
trans-
X ¼ NHnPr,
d
13C{1H} ¼ 211.07 ppm). The more downfield 13C{1H}
Mo(CO)4(Ph2PCl)2 intermediate does not. To gain more insight into
the factors that affect the transecis isomerizations in these com-
plexes, the manner in which the changes in the P-donor ligands
affect the mechanism of the isomerization reaction has been
studied, and the results from these studies are presented.
chemical shifts of carbonyl ligands in the complexes with phos-
phinamidite ligands as compared to those of complexes with
chlorodiphenyphosphine ligands are consistent with their lower
CO stretching frequencies [13] and demonstrate that they are better
electron donors. The other 13C{1H} and 1H NMR assignments were
consistent with those previously reported for the cis isomers [7].
The synthesis of 4 was expected to be more difficult than that of
the previously reported cis isomer [6,7] due to an increased po-
tential for the formation of oligomeric and/or polymeric products.
Such products are more likely with a trans coordination geometry
because the two salicylaldiminato groups in 3 are on opposite sides
of the molecule and thus could preferentially coordinate to
different nickel(II) centers. The reaction is further complicated by
the lack of a common solvent for the Ni(OAc)2(H2O)4 and 3. Because
of these issues, a stoichiometric amount of Ni(OAc)2(H2O)4 was
dissolved into a large quantity of a mixture of methanol and THF,
and this solution was then added dropwise over a 26 h period to a
solution of 3. However, even under these conditions, multiple
resonances were observed in the 31P{1H} NMR spectrum of the
crude product as shown in Fig. 1. The chemical shifts of the 31P{1H}
NMR resonances suggested that both the cis and trans isomers of
the Mo(CO)4(PPh2NHR)2 group were present. Only the cis isomer of
the desired product 4, trans-[cis-Mo(CO)4(PPh2NHCH2CH2NCH(o-
C6H4)O)2]Ni, subsequently denoted as cis-4, was isolated by
precipitating a dichloromethane solution of the reaction products
into a rapidly stirred 1:10 v/v mixture of dichloromethane and
hexanes. All complexes except for cis-4 precipitated, and pure cis-4
2. Results and discussion
2.1. Synthesis and characterization
The syntheses of the intermediate complexes shown in Scheme
1, 1, 2 and 3, were relatively straightforward. Complex 1 was pre-
pared via the thermal substitution reaction of Mo(CO)6 with PPh2Cl
as has been previously reported [11,12]. The consumption of the
PPh2Cl in the substitution reaction was monitored using 31P{1H}
NMR spectroscopy. After 28 h, the 31P{1H} NMR spectrum of the
reaction mixture indicated that both 1 and its cis isomer had
formed along with the monosubstituted Mo(CO)5(PPh2Cl) complex.
The three species were successfully separated by column chroma-
tography followed by recrystallization. The syntheses of 2 and 3
were essentially identical to those of their cis isomers [6,7] and
gave the complexes in high yields.
The 13C{1H} and 1H NMR spectra of the intermediate complexes,
2 and 3, were completely assigned using a combination of 1H{31P}
NMR, 1He13C HSQC and 1He1H COSY experiments. Each complex
exhibited a single downfield 1:2:1 triplet 13C NMR resonance for
the four carbonyl ligands indicating that both complexes have trans

J.R. Martin et al. / Journal of Organometallic Chemistry 751 (2014) 695e702
697
Fig. 1. 31P{1H} NMR showing the separation of cis-4 (76.65 ppm) from the crude mixture of products obtained from the reaction of 3 with Ni(OAc)2(H2O)4.
was obtained in moderate yields after two precipitations. The for-
mation of cis-4 could occur either by transecis isomerization of 3
followed by the coordination of the Ni(II) or by coordination of the
Ni(II) to 3 to form the desired trans monomer followed by trans to
cis isomerization of this material. The latter is less likely due to the
fact that the isolated trans species were not observed to convert to
the cis product.
intramolecular metal (sp) carbonyl (
reported by our group [14].
p)/p interactions recently
The edge-to-face (T-p) stacking interactions between the aro-
matic groups in 1e3 are also of interest. In 1, intermolecular T-
p
interactions between the phenyl rings of two molecules with a
ꢀ
centroid-carbon distance of 3.778 (4) A are observed. In contrast, 2
exhibits intramolecular T-p stacking interactions between two
phenyl rings on the same phosphinamidite ligand, with a centroid-
ꢀ
carbon distance of 3.972 (6) A, as a result of different rotations
2.2. X-ray crystallographic studies
about PeCPh-i bonds. Similar intramolecular interactions are also
seen for one of the phosphinamidite ligands in 3, with a centroid-
carbon distance of 3.946 (4) A. However, the other phosphinami-
The molecular structures of 1, 2 and 3 are shown in Figs. 2e4,
respectively, and selected bond distances and angles are given in
Table 1. All three compounds crystallize in the triclinic space group
P-1. The coordination environment of the molybdenum in each
complex is a slightly distorted octahedron with the largest de-
viations from the ideal 90ꢂ PeMoeC angles being 3.65(9)ꢂ for 1,
1.31(15)ꢂ for 2 and 5.17(10)ꢂ for 3.
ꢀ
dite ligand in 3 exhibits an intermolecular T-p stacking interaction
between the phenylene of the salicylaldiminato group and a phenyl
group of an adjacent molecule with a centroid-carbon distance of
ꢀ
3.895 (4) A.
The orientations of the substituents of the phosphorus-donor
ligands relative to the tetracarbonylmolybdenum(0) center can be
described as either staggered or eclipsed based on the smallest Xe
PeMoeC (X ¼ Cl (1) or NHR (2 or 3)) torsion angle. These torsion
angles are given in Table 2 and the Newman projections are shown
in Fig. 5. All of the complexes exhibit phosphorus ligands in stag-
gered confirmations relative to the carbonyls except the PeCl1
ligand of 1, which features an eclipsed confirmation. Interestingly,
2.3. Thermal transecis isomerization mechanism of 2 and 3
To better understand the factors that may be contributing to
isomerization of the trans-tetracarbonyl-bis(phosphinamidite)
molybdenum center during the reaction of 3 with Ni(OAc)2(H2O)4,
the thermodynamics and kinetics of the transecis isomerizations of
the intermediates 1, 2, 3 and cis-4 were investigated. No isomeri-
zation of 1 or cis-4 was observed at temperatures up to 338 K in
benzene-d6 for periods as long as 14 days. In contrast, isomeriza-
tions of both 2 and 3 did occur under similar reaction conditions.
all of these conformations allow face-to-face
p interactions be-
tween a orbital of one of the carbonyl ligands and the aromatic
p
p
system of an eclipsed phenyl group of the phosphorus-donor ligand
on each side of the complex as shown in Fig. 6 (selected distances
are listed in Table 3). These interactions are similar to other
Fig. 2. X-ray crystal structure of 1. Thermal ellipsoids are drawn at 25% and hydrogen
Fig. 3. X-ray crystal structure of 2. Thermal ellipsoids are drawn at 25% and hydrogen
atoms have been omitted for clarity.
atoms have been omitted for clarity.

698
J.R. Martin et al. / Journal of Organometallic Chemistry 751 (2014) 695e702
Table 2
Selected torsion angles (ꢂ) for 1e3.
Bonda
1
2
3
COeMoeP (1)eX (1)
COeMoeP (10)eX (10)
COeMoeP (1)eC (3)
COeMoeP (10)eC (30)
COeMoeP (1)eC (9)
COeMoeP (10)eC (90)
27.3 (1)
ꢁ6.6 (1)
ꢁ50.2 (3)
50.2 (3)
70.1 (3)
ꢁ41.9 (1)
ꢁ35.7 (1)
74.1 (1)
146.7 (1)
109.3 (1)
ꢁ87.6 (1)
ꢁ70.1 (3)
80.1 (1)
ꢁ167.7 (3)
ꢁ162.1 (2)
ꢁ126.7 (1)
167.7 (3)
ꢁ158.2 (2)
a
CO is defined as the carbonyl ligand closest to the X substituent.
constants for the reverse (cisetrans) reactions were calculated from
the reversible first-order rate constants for the forward reactions
and the transecis equilibrium constants.
An Eyring analysis [15] was used to extract the enthalpies of
activation (DH) and the entropies of activation (DS) for the transe
cis equilibria of 2 and 3. The data in Table 5 show that these acti-
vation energies are in excellent agreement with the kinetic pa-
rameters for the activation processes of the forward and reverse
isomerization reactions ð
D
H ¼
D
Hfz ꢁ
D
Hrz;
D
S ¼
D
Szf ꢁ
D
SzrÞ. The
positive
D
H values indicate that enthalpy favors the trans isomer of
Fig. 4. X-ray crystal structure of 3. Thermal ellipsoids are drawn at 25% and hydrogen
atoms have been omitted for clarity.
each complex, however the equilibrium lies in favor of the cis
isomer due to a favorable free energy of activation term at 298 K (2:
D
G ¼ ꢁ1.6 kJ molꢁ1; 3:
D
G ¼ ꢁ1.2 kJ molꢁ1). It is particularly
These results suggest that the X substituent on the phosphorus-
donor atom (1: X ¼ Cl; 2 and 3: X ¼ NHR) has a significant effect
on the rate at which the isomerization occurs. The fact that cis-4
does not undergo cisetrans isomerization suggests that the pres-
ence of the square-planar nickel(II) center in this molecule favors
the cis isomer.
To gain more insight into the effects of the X ¼ NHR substituents
on transecis isomerizations of trans-tetracarbonyl-bis(phosphina-
midite)molybdenum complexes, the transecis isomerizations of 2
and 3 were followed by 31P{1H} NMR spectroscopy at a range of
temperatures. The concentrations of the trans and cis isomers were
determined from the integrations of the resonances located at
interesting that DS is positive because typically the trans isomer of
Mo(CO)4(P-donor)2 complexes are favored entropically [16e19].
It is of importance to note that 2 and 3 have identical activation
energies at 328 K, yet the kinetic parameters for the forward and
reverse reactions are different. The fact that the kinetic parameters
for the reverse isomerizations are less positive/more negative in
both complexes is consistent with results reported by Darensbourg
et al. for isomerization reactions of M(CO)4(P-donor)2 (M ¼ Group
VIA; P-donor ligand ¼ phosphine or phosphite) [16e19].
To gain additional qualitative information about the mechanism
of the transecis isomerization in 2 and 3, both isomerizations were
carried out under a CO atmosphere using an established experi-
mental design [16e20]. These experiments did not result in the
formation of any monosubstituted complex Mo(CO)5(PPh2NHR),
which would be expected if dissociation of a phosphinamidite
ligand were occurring. The absence of this dissociation process
suggests that the transecis isomerizations are intramolecular.
Previous isomerization studies of Mo(CO)4(phosphine)2 complexes
d
¼ 89.52 ppm and
d
¼ 78.97 ppm for 2 and at ¼ 90.42 ppm and at
d
d
¼ 80.38 ppm for 3, respectively. These concentrations were used
to calculate the equilibrium constants in Table 4. The reversible
first-order rate constants for the forward (trans to cis) reactions at
each temperature for each complex were extracted from graphs of
[cis] vs. time (2: Fig. 7; 3: Fig. 8). The reversible first-order rate
have shown that positive DS values should afford a ligand dissoci-
ation process [16]. However, the non-dissociative pathway
observed in this experiment is further supported by the relatively
Table 1
ꢂ
ꢀ
Selected bond distances (A) and angles ( ) for 1e3.
Bond
1
2
3
small magnitude of DH, which is less than one third of the reported
energy needed for ligand dissociation [18]. The pathway for a
nondissociative, intramolecular rearrangement process in
Mo(CO)4(phosphine)2 complexes has previously been shown to
proceed through a trigonal-prismatic transition state [16e18,21].
MoeP (1)
2.434 (8)
2.425 (8)
2.028 (3)
2.044 (3)
2.026 (3)
2.028 (3)
2.083 (10)
2.089 (13)
1.135 (4)
1.121 (4)
1.137 (4)
1.135 (3)
114.55 (4)
114.90 (4)
88.10 (8)
94.41 (8)
89.68 (9)
89.78 (9)
91.10 (9)
86.35 (9)
90.58 (8)
89.98 (8)
177.40 (2)
2.455 (15)
2.455 (15)
2.017 (5)
2.030 (5)
2.017 (5)
2.030 (5)
1.667 (4)
1.667 (4)
1.147 (6)
1.133 (6)
1.147 (6)
1.133 (6)
111.85 (17)
111.85 (17)
88.70 (15)
88.70 (15)
91.31 (14)
91.31 (14)
88.70 (15)
88.70 (15)
91.31 (14)
91.31 (14)
180.00 (0)
2.455 (11)
2.453 (11)
2.047 (4)
2.043 (4)
2.008 (4)
2.019 (4)
1.681 (3)
1.687 (3)
1.129 (4)
1.132 (4)
1.145 (4)
1.143 (4)
110.75 (11)
110.99 (11)
88.76 (10)
89.61 (10)
95.17 (10)
85.20 (10)
88.09 (11)
93.54 (11)
87.71 (10)
91.93 (10)
178.33 (3)
MoeP (10)
MoeC (1)
MoeC (2)
MoeC (10)
MoeC (20)
P (1)eX (1)
P (10)eX (10)
C (1)eO (1)
3. Conclusions
C (2)eO (2)
Three octahedral trans-tetracarbonyl-bis(phosphorus-donor)
molybdenum(0) complexes (1, 2 and 3) that are potential in-
termediates in the formation of a bimetallic molecular rotor (4)
have been synthesized and fully characterized by multinuclear
NMR spectroscopy and X-ray crystallography. All three complexes
crystallize in the triclinic space group P-1 with the molybdenum
centered in a slightly distorted octahedron. A variety of face-to-face
C (10)eO (10)
C (20)eO (20)
MoeP (1)eX (1)
MoeP (10)eX (10)
C (1)eMoeP (1)
C (1)eMoeP (10)
C (2)eMoeP (1)
C (2)eMoeP (10)
C (10)eMoeP (1)
C (10)eMoeP (10)
C (20)eMoeP (1)
C (20)eMoeP (10)
P (1)eMoeP (10)
and edge-to-face
complexes.
p interactions have been observed in all three
A mixture of the thermally stable cis-4 as well as either 4 or its
oligomers was formed during the reaction of 3 with Ni(OA-
c)2(H2O)4, but only cis-4 could be isolated from this mixture. The

J.R. Martin et al. / Journal of Organometallic Chemistry 751 (2014) 695e702
699
O
C
O
C
X
X
Ph
C
O
O
C
C
O
C
O
Ph
Ph
C
O
C
O
Ph
staggered
eclipsed
Fig. 5. Newman projections showing the orientation of the substituents on phosphorus-donor ligands relative to the carbonyl ligands.
formation of this product indicates that transecis isomerization of
the tetracarbonyl-bis(phosphinamidite)molybdenum(0) center
must have occurred during the reaction of 3 with Ni(OAc)2(H2O)4.
Similar isomerizations were observed for 2 and 3 upon heating.
However, cisetrans equilibrium was not observed when solutions
of 1 or cis-4 were heated. These results suggest that the rigidity of
the square-planar nickel(II) center in cis-4 favors the cis isomer, and
also that the substituent on the phosphorus-donor atom (1 ¼ Cl, 2
and 3 ¼ NHR) has a significant effect upon the rate of the
isomerization.
The transecis isomerizations of 2 and 3 have been shown to
follow reversible first order kinetics, and the thermodynamic and
kinetic parameters for these reactions have been determined. In
addition, the isomerization mechanism has been determined to
follow a non-dissociative process that does not involve any
breaking of the molybdenumephosphorus bonds. These results,
coupled with the thermodynamic and kinetic parameters obtained
from an Eyring analysis, suggest that the isomerizations of 2 and 3
proceed via an intramolecular mechanism involving a trigonal-
prismatic transition state similar to those reported for related
complexes.
CaH2 and finally by distillation from Na/benzophenone. It was
stored over molecular sieves (3 A; 8e12 mesh) and used
ꢀ
within a few hours. Toluene was dried by distillation from Na
ꢀ
and stored over molecular sieves (3 A; 8e12 mesh). Other
solvents were reagent grade and were degassed using high-
purity (99.998%) nitrogen before use. Ethylenediamine was
dried by allowing it to stand over KOH followed by fractional
distillation. The fraction that distilled at 118 ꢂC was collected
ꢀ
and stored over molecular sieves (3 A; 8e12 mesh) until use.
o-Salicylaldehyde was allowed to stand over molecular sieves
ꢀ
(3 A; 8e12 mesh) for several days before being distilled under
reduced pressure. The distilled o-salicylaldehyde was stored in
a sealed 50 mL round bottom flask under dry nitrogen until
use. Reagent grade PPh2Cl was found to be 94.000% pure by
31P{1H} NMR spectra, while Mo(CO)6 was of sufficient purity
as received from the supplier.
4.2. Characterization
Multinuclear 31P{1H}, 13C{1H}, 1H{31P} and 1H NMR spectra were
recorded on either a Bruker DRX-400 or a Bruker DPX-300 NMR
spectrometer. Chloroform-d solutions of the complexes were pre-
pared under a stream of N2 gas. The 31P{1H} NMR spectra were
referenced to external 85% H3PO4 in a coaxial tube that also con-
tained CDCl3, and the 13C{1H} and 1H NMR spectra were referenced
to internal SiMe4 (TMS). Some assignments in 3 were based on 2D
NMR spectra (1He13C HSQC, 1He1H COSY). Elemental analyses (C
and H) were performed by Atlantic Microlabs, Inc.
4. Materials and methods
4.1. Materials
Tetrahydrofuran (THF) was dried by first allowing it to
stand over MgSO4 for at least 12 h, then by distillation from
Fig. 6. X-ray crystal structures of 1 (left), 2 (middle) and 3 (right) with intramolecular short contacts. Thermal ellipsoids are drawn at the 25% probability level, all hydrogens are
omitted and the non-aromatic substituent on each ligand is truncated for clarity. Centroids are drawn for both the aryl and carbonyl groups.

700
J.R. Martin et al. / Journal of Organometallic Chemistry 751 (2014) 695e702
Table 3
0.025
Selected bond distances for 1e3.
Compound
1
2
3
0.02
0.015
0.01
0.005
0
(C1^O1) centroid to ringa
(C2^O2) centroid to ringb
(C10^O10) centroid to ringc
(C10^O10) centroid to ringd
3.597 (1)
3.634 (1)
3.829 (1)
3.508 (1)
3.823 (1)
337 K
334 K
329 K
328 K
a
Ring is defined as C30, C40, C50, C60, C70, C80.
Ring is defined as C3, C4, C5, C6, C7, C8.
Ring is defined as C9, C10, C11, C12, C13, C14.
Ring is defined as C90, C100 C110, C120, C130, C140.
b
c
d
0
50000
100000
150000
200000
4.3. X-ray data collection and solution
Time (s)
Suitable single crystals of 1, 2, and 3 were glued on glass fibers
with epoxy and mounted upon a Bruker-Nonius CAD4 single-
crystal diffractometer under aerobic conditions. Standard peak
search and automatic indexing routines followed by least square
fits of 25 accurately centered reflections resulted in accurate unit
cell parameters for each. The space groups of the crystals were
assigned on the basis of systematic absences and intensity statis-
tics. All data collections were carried out using the CAD4-PC soft-
ware [22], and details of the data collections are given in Table 6.
Analytical scattering factors of each compound were corrected for
Fig. 7. Reversible first-order kinetics plots of the trans to cis isomerizations of 2 at four
temperatures.
physically separated. 31P{1H} NMR (162 MHz; CDCl3):
d 137.05 (s).
1H NMR (400 MHz; CDCl3):
d 7.69e7.74 (m, 4H, HPh-o), 7.45 (m, 6H,
HPh-m,p). 13C{1H} NMR (101 MHz; CDCl3):
d
208.23 (t, j2JCPj ¼ 10. Hz,
CO), 141.46 (aq, j1JCP
þ
3JCP
j
¼
28 Hz, CPheCi), 131.29 (aq,
j2JCP
þ
4JCPj ¼ 15 Hz, CPheCo), 131.06 (s, CPheCp), 128.80 (aq,
j3JCP
þ
5JCP
j
¼
10. Hz, CPheCm). Anal. Calcd. for
both
D D
f0 and i f00 components of anomalous dispersion. All data
[C28H20Cl2MoO4P2$0.1CH2Cl2]: C 51.31, H 3.10; Found: C 51.16, H
3.02.
were corrected for Lorentz and polarization effects and empirical
absorption corrections were applied when necessary.
4.5. trans-Mo(CO)4(PPh2NHCH2CH2NH2)2 (2)
All crystallographic calculations were performed with the
SHELXTL-PC program package [23]. All heavy atom positions were
located using Direct Methods, and all remaining non-hydrogen
atoms were located using difference Fourier maps. Full-matrix re-
finements of the positional and anisotropic thermal parameters for
all non-hydrogen atoms versus F2 were carried out. All hydrogen
A solution of 4.0 g (6.1 mmol) of 1 in THF was added dropwise at
room temperature to an excess (11 mL) of H2NCH2CH2NH2. The
reaction mixture was stirred under N2 for 30 min after the addition
was completed, and then the solvent was removed under vacuum.
The residue was washed first with a 30 mL aliquot of de-ionized
water, next with two 10 mL aliquots of ethanol and finally with
two 10 mL aliquots of diethyl ether to yield a yellow powder. The
yellow powder was dissolved in 300 mL of benzene and the excess
H2NCH2CH2NH2 was extracted with two washes of deionized wa-
ter. The benzene layer was evaporated to dryness to yield 3.1 g
(85%) of the crude product as a yellow powder. Recrystallization
from a 1:1 solution of dichloromethane/hexanes yielded analyti-
atoms were placed in calculated positions with the appropriate
ꢀ
molecular geometry and
d
(CeH) ¼ 0.96 A, except for H1N1, H2N2
and H3N2 in 2. The isotropic thermal parameter of each hydrogen
atom was fixed equal to 1.2 times the Ueq value of the atom to which
it was bound. Selected bond lengths, angles and torsion angles for
all of the compounds are given in Tables 1 and 2.
4.4. trans-Mo(CO)4(PPh2Cl)2 (1)
cally pure 2. 31P{1H} NMR (162 MHz; CDCl3):
d
89.52 (s). 1H NMR
(400 MHz; CDCl3): d 7.60e7.64 (m, 4H, HPh-o), 7.40 (m, 6H, HPh-m,p),
A solution of 12.5 g (47.3 mmol) of Mo(CO)6 and 18.1 mL
(66.7 mmol) of PPh2Cl in 150 mL of toluene was refluxed under N2,
and the progress of the reaction was monitored by 31P{1H} NMR
spectroscopy. After 28 h of reflux, the black solution was evaporate
to dryness, and the residue was purified by chromatography on
silica gel using a 3:1 v/v mixture of dichloromethane and hexanes
as the eluent. The yellow solid from the chromatography was
recrystallized from a dichloromethane/hexanes solution to yield
14.4 g (48.8%) of a mixture of green [cis-Mo(CO)4(PPh2Cl)2; 10.1 g]
and yellow [trans-Mo(CO)4(PPh2Cl)2 (1); 4.3 g] crystals, which were
2.67 (q, 2H, j3JHHj ¼ 5.9 Hz, HEn-1), 2.53 (t, 2H, j3JHHj ¼ 5.8 Hz, CEn-2),
2.50 (t, 1H, j3JHHj ¼ 6.3 Hz, HN), 0.95 (s, 2H, H2N). 13C{1H} NMR
(101 MHz; CDCl3):
d
211.07 (t, j2JCPj ¼ 9 Hz, CO),
d 139.62 (aq,
j1JCP þ 3JCPj ¼ 39 Hz, CPheCi), 131.67 (aq, j2JCP þ 4JCPj ¼ 13 Hz, CPheCo),
0.025
0.02
0.015
Table 4
328 K
Equilibrium constants and reversible first-order rate constants for the forward
(transecis) and reverse (cisetrans) isomerizations of 2 and 3.
0.01
335 K
Compound
T (K)
kf (10ꢁ5 sꢁ1
)
kr (10ꢁ6 sꢁ1
)
K
337 K
0.005
2
337
334
329
328
337
335
328
3.29 (4)
2.79 (6)
1.64 (4)
1.41 (5)
2.18 (8)
2.06 (6)
1.41 (3)
10.4 (3)
9.23 (3)
5.63 (2)
5.03 (2)
6.67 (5)
6.60 (5)
4.79 (4)
3.15 (8)
3.02 (7)
2.91 (6)
2.83 (13)
3.27 (13)
3.17 (28)
2.94 (10)
0
0
50000
100000
150000
200000
250000
300000
Time (s)
3
Fig. 8. Reversible first-order kinetics plots of the trans to cis isomerizations of 3 at
three temperatures.

J.R. Martin et al. / Journal of Organometallic Chemistry 751 (2014) 695e702
701
Table 5
Equilibrium and kinetic parameters for the activation processes of the transecis isomerizations of 2 and 3.
Compound
D
Ha
D
Sa
D
Hfz
D
Sfz
D
Hrz
D
Srz
b
b
b
b
(kJ molꢁ1
)
(J molꢁ1 Kꢁ1
)
(kJ molꢁ1
)
(J molꢁ1 Kꢁ1
)
(kJ molꢁ1
)
(J molꢁ1 Kꢁ1
)
2
3
10 (2)
11 (2)
39 (5)
41 (7)
85 (12)
43 (4)
118 (36)
ꢁ10 (12)
74 (14)
33 (7)
76 (42)
ꢁ49 (20)
a
Equilibrium activation parameters relative to the cisetrans isomerizations obtained from linear plots of eRln K vs. 1/T (Supplementary data).
Kinetic activation parameters relative to the forward (f) or reverse (r) pathways obtained from linear plots of ln(k/T) vs. 1/T (Supplementary data).
b
129.54 (s, CPheCp), 128.39 (aq, j3JCP þ 5JCPj ¼ 9 Hz, CPheCm), 47.24 (aq,
(t, 1H, j3JHHj ¼ 6.9, NH). 13C{1H} NMR (101 MHz; CDCl3):
d 210.70 (t,
j2JCP þ 4JCPj ¼ 7 Hz, CEn-1), 43.48 (aq, j3JCP þ 5JCPj ¼ 6 Hz, CEn-2). Anal.
Calcd. for [C32H34MoN4O4P2$0.4CH2Cl2]: C 53.27, H 4.80; Found: C
53.07, H 4.82.
j2JCPj ¼ 9 Hz, CO), 166.97 (s, CH]N), 161.44 (s, CCH]N), 139.40 (aq,
j1JCP
þ
3JCPj ¼ 39 Hz, CPheCi), 132.71 (s, CHCHCOH), 131.92 (s,
CHCCH]N), 131.60 (aq, j2JCP þ 4JCPj ¼ 13 Hz, CPheCo), 129.71 (s, CPhe
Cp),128.50 (aq, j3JCP þ 5JCPj ¼ 9 Hz, CPheCm),119.06 (s, COH),118.99 (s,
CHCHCCH]N), 117.34 (s, CHCOH), 61.24 (aq, j3JCP þ 5JCPj ¼ 4 Hz, CEn-
4.6. trans-Mo(CO)4(PPh2NHCH2CH2NCH(o-C6H4)OH)2 (3)
2), 44.87 (aq, j2JCP
þ
4JCP
j ¼ 7 Hz, CEn-1). Anal. Calcd. for
C46H42MoN4O6P2: C 61.07, H 4.68; Found. C 60.53, H 4.84.
Freshly distilled o-salicylaldehyde (580 mL) was added dropwise
at room temperature to a stirred solution of 1.9 g (0.27 mmol) of 2
in THF, and then the reaction mixture was stirred under N2 for an
additional 1.5 h. Evaporation of the solvent yielded brownish-
yellow oil that became a hard brown solid under vacuum. The
solid was recrystallized from a 1:1 solution of dichloromethane/
4.7. Attempted synthesis of trans-[trans-
Mo(CO)4(PPh2NHCH2CH2NCH(o-C6H4)O)2]Ni (4)
A mixture of 0.15 g (0.60 mmol) of NiOAc$4H2O in w250 mL of a
1:1 THF/methanol solution was sonicated until the salt dissolved,
and then the solution was added dropwise at room temperature to
a stirred solution of 0.54 g (0.59 mmol) of 3 in 100 mL of THF over a
26 h period. Evaporation of the hazel colored solution to dryness
yielded a mixture of products indicated by multiple resonances in
the 31P{1H} NMR spectrum. This mixture was dissolved into
approximately 20 mL of dichloromethane, and this solution was
added to approximately 800 mL of a hexanes. The liquid was
hexanes to yield 2.1 g (86%) of brown, analytically pure crystals. 31
P
{1H} NMR (162 MHz; CDCl3): 90.42 (s). 1H NMR (400 MHz; CDCl3):
d
d
13.12 (s, 1H, OH), 8.12 (s, 1H, CH]N), 7.57e7.61 (m, 4H, HPh-o), 7.38
(m, 6H, HPh-m,p), 7.32 (ddd, 1H, j3JHHj ¼ 8.4 Hz, j3JHHj ¼ 7.2 Hz,
j4JHH
j
¼
1.6 Hz, CHCHCOH), 7.18 (dd, 1H, j3JHH
j
¼
7.7 Hz,
j4JHHj ¼ 1.7 Hz, CHCCH]N), 6.95 (d, 1H, j3JHHj ¼ 8.3 Hz, CHCOH),
6.86 (td, 1H, j3JHH
j
¼
7.5 Hz, CHCHCCH]N), 3.32 (t, 2H,
j3JHHj ¼ 5.7 Hz, HEn-2), 2.98 (q, 2H, j3JHHj ¼ 6.3 Hz, HEn-1), 2.25e2.29
Table 6
Crystal data and structure refinement for 1, 2 and 3.
1
2
3
Empirical formula
Formula weight
Temperature
C28H20Cl2MoO4P2
649.22
C32H34MoN4O4P2
696.51
C46H42MoN4O6P2
904.72
293(2) K
293(2) K
293(2) K
ꢀ
ꢀ
ꢀ
Wavelength
0.71073 A
0.71073 A
0.71073 A
Crystal system
Space group
Unit cell dimensions
Triclinic
P-1
Triclinic
P-1
Triclinic
P-1
ꢀ
ꢀ
ꢀ
a ¼ 10.372(2) A
a ¼ 8.964(18) A
a ¼ 10.157(2) A
ꢀ
ꢀ
ꢀ
b ¼ 11.551(2) A
b ¼ 10.452(2) A
b ¼ 14.942(3) A
ꢀ
ꢀ
ꢀ
c ¼ 13.486(3) A
c ¼ 10.842(2) A
c ¼ 14.941(3) A
a
b
g
¼ 111.05(3)ꢂ
¼ 103.71(3)ꢂ
¼ 96.47(3)ꢂ
a
b
g
¼ 118.74(3)ꢂ
¼ 105.94(3)ꢂ
¼ 96.43(3)ꢂ
a
b
g
¼ 78.56(3)ꢂ
¼ 87.27(3)ꢂ
¼ 82.27(3)ꢂ
3
ꢀ
Volume (A )
1429.6(5)
2
819.7(3)
2201.7(8)
2
Z
1
dcalc (mg/m3)
1.508
0.789
652
1.411
0.539
358
1.365
0.422
932
Absorption coefficient (mmꢁ1
F (000)
)
Crystal size
0.2 ꢃ 0.4 ꢃ 0.5 mm3
2.00e24.97ꢂ
ꢁ12 ꢄ h ꢄ 1
ꢁ13 ꢄ k ꢄ 13
ꢁ15 ꢄ l ꢄ 16
5930
0.3 ꢃ 0.1 ꢃ 0.1 mm3
2.23e24.97ꢂ
ꢁ1 ꢄ h ꢄ 10
ꢁ12 ꢄ k ꢄ 12
ꢁ12 ꢄ l ꢄ 12
5949
0.4 ꢃ 0.4 ꢃ 0.15 mm3
2.02e24.97ꢂ
ꢁ12 ꢄ h ꢄ 1
ꢁ17 ꢄ k ꢄ 17
ꢁ17 ꢄ l ꢄ 17
9152
q
range for data selection
Index ranges
Reflections collected
Independent reflections
5019 [R(int) ¼ 0.0187]
2886 [R(int) ¼ 0.0423]
7715 [R(int) ¼ 0.0305]
Completeness to
q
¼ 24.97ꢂ
99.9%
100.0%
100.0%
Absorption correction
Max. and min. transmission
Refinement method
j
-scan
j
-scan
j-scan
0.8994 and 0.7606
Full-matrix least squares on F2
5019/0/335
0.8590 and 0.6812
Full-matrix least squares on F2
2886/0/208
0.4566 and 0.3680
Full-matrix least squares on F2
7715/0/540
Data/restraints/parameters
Goodness-of-fit on F2
1.096
1.026
1.038
Final R indices [I > 2
R indices (all data)
q
(I)]
R1 ¼ 0.0295, wR2 ¼ 0.0753
R1 ¼ 0.0518, wR2 ¼ 0.1259
R1 ¼ 0.0363, wR2 ¼ 0.0946
R1 ¼ 0.0382,
R1 ¼ 0.0795,
R1 ¼ 0.0714,
wR2 ¼ 0.0795
wR2 ¼ 0.1382
wR2 ¼ 0.1079
ꢀꢁ3
Largest diff. peak and hole (e A
)
0.349 and ꢁ0.359
0.973 and ꢁ0.881
0.345 and ꢁ0.441

702
J.R. Martin et al. / Journal of Organometallic Chemistry 751 (2014) 695e702
decanted from the precipitate and then filtered through a layer of
silica gel. The liquid was evaporated to dryness, and the precipi-
tation was repeated a second time. Evaporation of the filtrate to
dryness yielded 0.26 g (45.3%) of the spectroscopically pure cis
Appendix B. Supplementary data
Supplementary data related to this article can be found at http://
dx.doi.org/10.1016/j.jorganchem.2013.10.015.
isomer of 4 (cis-4). 31P{1H} NMR (162 MHz; CDCl3):
d 76.65 (s).
References
4.8. Reversible first-order kinetics experiments
[1] G.D. Hess, F. Hampel, J.A. Gladysz, Organometallics 26 (2007) 5129e5131.
[2] M.A. Garcia-Garibay, Nat. Mater. 7 (2008) 431e432.
[3] R. Franke, D. Selent, A. Borner, Chem. Rev. 112 (2012) 5675e5732.
[4] G.T. Whiteker, C.J. Cobley, Top. Organomet. Chem. 42 (2012) 35e46.
[5] K. Skopek, M.C. Hershberger, J.A. Gladysz, Coord. Chem. Rev. 251 (2007)
1723e1733.
[6] C.S. Kraihanzel, E. Sinn, G.M. Gray, J. Am. Chem. Soc. 103 (1981) 960e962.
[7] C. Duffey, S. Stepleton, M.E. Anderson, D. Cox, M. Ready, H. Byrd,
C.A. Bloomfield, J.L. Freeman, G.M. Gray, J. Chem. Crystallogr. 41 (2011) 1560e
1567.
[8] G.M. Gray, A.L. Zell, H. Einspahr, Inorg. Chem. 25 (1986) 2923e2925.
[9] T. Akutagawa, D. Endo, F. Kudo, S. Noro, S. Takeda, L. Cronin, T. Nakamura,
Cryst. Growth Des. 8 (2008) 812e816.
[10] T. Shima, F. Hampel, J.A. Gladysz, Angew. Chem. Int. Ed. Engl. 43 (2004) 5537e
5540.
[11] E.H. Wong, F.C. Bradley, Inorg. Chem. 20 (1981) 2333e2335.
[12] P.W. Dyer, J. Fawcett, M.J. Hanton, R.D. Kemmitt, R. Padda, N. Singh, Dalton
Trans. (2002) 104e113.
[13] G.M. Gray, Inorg. Chim. Acta 81 (1983) 157e162.
[14] S.D. Hastings, H. Byrd, L.N. Gray, M.J. Jablonsky, J.L. Freeman, G.M. Gray, Eur. J.
Inorg. Chem. (2013) 2900e2911.
[15] E.V. Anslyn, D. Dougherty, Modern Physical Organic Chemistry, University
Science Books, Sausalito, CA, 2006.
[16] D.J. Darensbourg, A.H. Graves, Inorg. Chem. 18 (1979) 1257e1261.
[17] D.J. Darensbourg, Inorg. Chem. 18 (1979) 14e17.
[18] D.J. Darensbourg, R. Kudaroski, W. Schenk, Inorg. Chem. 21 (1982) 2488e
2491.
A 0.028 M solution of each complex in benzene-d6 was prepared
in a 5 mm sealed J-Young NMR tube. The solution was then frozen
in liquid nitrogen, evacuated and finally back-filled with gaseous N2
three times to remove oxygen. The trans to cis equilibration was
followed by 31P{1H} NMR spectroscopy on a Bruker DRX-400 NMR
spectrometer. The temperature of the spectrometer remained
constant throughout the experiment. The relative amounts of the
trans and cis isomers were determined from the integrations of the
31P{1H} NMR resonances.
Acknowledgments
The authors would like to acknowledge the National Science
Foundation (NSF) for support under grant EPS-0447675. Justin
Martin thanks the Alabama EPSCoR Graduate Research Scholars
Program (GRSP) for a Graduate Fellowship. Special thanks to Dr.
Mike Jablonsky of the Chemistry Department at UAB for his assis-
tance in interpreting the 2-D NMR spectra of 3.
Appendix A. Supplementary material
[19] D.J. Darensbourg, R.L. Gray, Inorg. Chem. 23 (1984) 2993e2996.
[20] D.J. Darensbourg, J.R. Andreatta, S.M. Stranahan, J.H. Reibenspies, Organo-
metallics 26 (2007) 6832e6838.
[21] S. Datta, B. Dezube, J.K. Kouba, S.S. Wreford, J. Am. Chem. Soc. 100 (1978)
4404e4412.
CCDC 1: CCDC 928162, 2: CCDC 928163, 3: CCDC 928164 contain
the supplementary crystallographic data for this paper. These data
can be obtained free of charge from The Cambridge Crystallo-
graphic Data Centre via www.ccdc.cam.ac.uk/data_request/cif.
[22] CAD4-PC Version 1.2, 1988.
[23] G.M. Sheldrick, SHELXTL NT Version 6.12, 2001.