Synthesis, characterization and theoretical investigations of molybdenum carbonyl complexes with phosphorus/nitrogen/phosphorus ligand as bidentate and tridentate modes

Article abstract of DOI:10.1016/j.poly.2017.09.037

Carbonyl complexes of molybdenum with the phosphorus/nitrogen/phosphorus ligand, PNP-Mo(CO)_n, where n = 3, 4 and PNP is N,N-bis[2-(diphenylphosphino)ethyl]phenylamine have been synthesized and characterized by ¹H, ³¹P{¹H} NMR, IR (ATR and solution state), Raman, X-ray diffraction, MS and elemental analysis. DFT calculations of the complexes were also performed. The PNP-Mo(CO)₄ complex in which the binding mode is bidentate cis-phosphine and, facial and meridional isomers of PNP-Mo(CO)₃ in which the binding mode is tridentate were obtained. The facial isomer and the tetracarbonyl complexes were successfully synthesized and purified; however, it was not possible to isolate the meridional isomer. Only a few crystals of the meridional isomer were detected and used for X-ray analysis. When the facial isomer was dissolved in CH₂Cl₂ or in THF, it was found to undergo decomposition thereby generating the tetracarbonyl complex, meridional isomer, Mo(0) as a dark precipitate along with some facial isomer that was left intact in solution. The reason that the facial isomer was the preferred structure with respect to the meridional isomer was also investigated.

Full text of DOI:10.1016/j.poly.2017.09.037

Polyhedron 138 (2017) 206–217
Contents lists available at ScienceDirect
Polyhedron
journal homepage: www.elsevier.com/locate/poly
Synthesis, characterization and theoretical investigations of
molybdenum carbonyl complexes with phosphorus/nitrogen/
phosphorus ligand as bidentate and tridentate modes
⇑
Seyma Goren Keskin , Julie M. Stanley, Alan H. Cowley
Department of Chemistry, The University of Texas at Austin, 1 University Station, A5300, Austin, TX 78712-0165, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 1 April 2017
Accepted 21 September 2017
Available online 28 September 2017
Carbonyl complexes of molybdenum with the phosphorus/nitrogen/phosphorus ligand, PNP-Mo(CO)n,
where n = 3, 4 and PNP is N,N-bis[2-(diphenylphosphino)ethyl]phenylamine have been synthesized
and characterized by 1H, 31P{1H} NMR, IR (ATR and solution state), Raman, X-ray diffraction, MS and ele-
mental analysis. DFT calculations of the complexes were also performed. The PNP-Mo(CO)4 complex in
which the binding mode is bidentate cis-phosphine and, facial and meridional isomers of PNP-Mo(CO)3
in which the binding mode is tridentate were obtained. The facial isomer and the tetracarbonyl com-
plexes were successfully synthesized and purified; however, it was not possible to isolate the meridional
isomer. Only a few crystals of the meridional isomer were detected and used for X-ray analysis. When the
facial isomer was dissolved in CH2Cl2 or in THF, it was found to undergo decomposition thereby gener-
ating the tetracarbonyl complex, meridional isomer, Mo(0) as a dark precipitate along with some facial
isomer that was left intact in solution. The reason that the facial isomer was the preferred structure with
respect to the meridional isomer was also investigated.
Keywords:
Molybdenum carbonyl complexes
PNP ligand
Facial isomer
Meridional isomer
Ó 2017 Elsevier Ltd. All rights reserved.
1. Introduction
(PNP), bidentate monomeric (PP), and bidentate dimeric (PP)
modes (Fig. 1). Binding modes can be controlled by donor strength
Molybdenum complexes with low oxidation states have been
investigated extensively due to the roles that these complexes play
in the contexts of both chemical and biological nitrogen fixation.
Currently, more than fifty molybdenum containing enzymes are
known and the presence of molybdenum in these quantities is
required by eukaryotes [1]. The molybdenum complexes also play
important roles in organometallic chemistry. For example, molyb-
denum carbonyls have been used for Pauson – Khand reactions, i.e.
[2+2+1] cycloaddition in which an alkene, an alkyne and carbon
of N by having different R groups on it [4] (Fig. 2).
The tridentate PNP ligand N,N-bis[2-(diphenylphosphino)ethyl]
phenylamine that was used in the present study was synthesized
previously and coordinated to Mn, Co, Ni, Rh, Pd, Re and Pt
[5–10] metals as well as one example of a Mo complex (which
has an N2 monodentate ligand and no carbonyls) [11].
Previously, complexes of PNP ligand with molybdenum (cis-
PNP-Mo(CO)4, fac and mer isomers of PNP-Mo(CO)3) with different
R groups on the nitrogen atom than our ligand were reported
(Fig. 3) [3,12,13]. Similar crystal structures for the cis-PNP-Mo
(CO)4 and fac-PNP-Mo(CO)3 complexes (with different R groups
on the N compared with our complexes) were also shown [3,14].
The crystal structures for the meridional isomer of the PNP-Mo
(CO)3 type have not been reported previously. Furthermore, Siclo-
van et. al. reported that they could synthesize and purify the facial
isomer but they were not able to isolate the meridional isomer
[12]. The crystal structure for a meridional complex of Mo of the
type PNP-Mo(CO)n with the PNP ligand in which N is an amine,
the Mo is neutral and the monodentate ligands are all carbonyls
have not been reported previously. There are examples of
meridional complexes of Mo that have ionic Mo, or a Mo that has
six monodentate ligands, or a PNP ligand that has an imine N in
the literature [15–18].
monoxide combine to form a a,b-cyclopentenone. Pauson – Khand
reactions are mediated both by bimetallic and monometallic
transition metal complexes [2]. The coordination chemistry of
molybdenum complexes has also been studied under thermal or
photolytic conditions [3].
Many types of PNP (phosphorus/nitrogen/phosphorus) ligands
have been studied widely because the hemilabile property of the
nitrogen atom promotes some catalytic reactions and provides
different coordination geometries such as tridentate monomeric
⇑
Corresponding author at: The University of Texas at Rio Grande Valley,
Department of Chemistry ESCNE 3.324, 1201
78539-2999, USA.
W University Dr., Edinburg TX,
E-mail address: skeskin@utexas.edu (S. Goren Keskin).
https://doi.org/10.1016/j.poly.2017.09.037
0277-5387/Ó 2017 Elsevier Ltd. All rights reserved.

S. Goren Keskin et al. / Polyhedron 138 (2017) 206–217
207
Fig. 1. Possible coordination geometries of PNP ligand [4].
complex to determine how much each ligand and metal orbital
contributes to the structure and if that causes the stability differ-
ences among the various complexes. The second DFT calculations
were carried out only for the PNP ligand for the meridional geome-
tries and facial geometries in order to determine if any steric effects
coming from the rings that formed upon chelation play a role in the
stability of the complex.
Fig. 2. Coordination change by change in donor strength of hemilabile N [5].
2. Experimental
2.1. Materials and physical measurements
The air- and moisture-sensitive reactions were carried out in
oven-dried glassware using standard Schlenk techniques either
under an inert nitrogen atmosphere or in a glove box with an argon
atmosphere. The N-phenyldiethanolamine and the trimethylamine
N-oxide dihydrate were purchased from TCI America. The
trimethylamine N-oxide dihydrate was purified by sublimation
prior to use in order to remove the water of hydration [21]. The
p-toluenesulphonyl chloride and hexacarbonylmolybdenum were
purchased from Alfa Aesar. The phosphorus oxychloride was pur-
chased from Acros, and the potassium diphenyl phosphide and
(bicyclo[2.2.1]hepta-2,5-diene) tetracarbonylmolybdenum were
purchased from Aldrich. The cycloheptatriene molybdenum tricar-
bonyl was purchased from Strem. The dry solvents were obtained
from an Innovative Technologies Pure-Solv 400 solvent purification
system. All other chemicals were used as received from commer-
cial suppliers. The 1H NMR spectra were recorded with a Varian
Unity + 300 spectrometer and were referenced to residual solvent
peaks. The 31P{1H} NMR spectra were referenced to a H3PO4 exter-
nal standard. All peak positions were listed in ppm and all coupling
constants were listed in Hertz (Hz). The mass spectrometry was
carried out using a Thermo Finngan TSQ 700 spectrometer. The tar-
get ligand, N,N-bis[2-(diphenylphosphino)ethyl]phenylamine 3,
was prepared according to Kostas’ method [8]. The syntheses of
fac-PNP-Mo(CO)3 4 and cis-PNP-Mo(CO)4 5 were carried out by
techniques similar to those of Blower [22] and Ainscough [15].
Fig. 3. PNP-Mo(CO)n, n = 3–4 complexes previously reported [3,12,13].
In the present study, molybdenum carbonyl complexes with an
N,N-bis[2-(diphenylphosphino)ethyl]phenylamine ligand have
been synthesized and fully characterized. Two isomers for the
PNP-Mo(CO)3, i.e. facial and meridional isomers, and one tetracar-
bonyl complex, cis-PNP-Mo(CO)4 were obtained along with their
crystal structures. Ligand binding modes, conversion between the
three complexes and the preferred geometry of the facial over
the meridional isomer were also investigated. The explanation for
the preferred geometry of the facial isomer is based on the trans
2.2. Spectroscopy
effect and
p-back donation, Jahn–Teller effect, bond angles and
bond distances, and DFT calculations of the molecular orbitals.
However, the choice of the facial or the meridional isomer was
explained on the basis of electronics rather than sterics as in other
journal articles [1,19,20].
The IR spectra were recorded with Nicolet Avatar 330 FT-IR
spectrophotometer. The solid-state IR measurements were carried
using an ATR accessory. The solution IR data were recorded using
DCM solutions (ꢁ10–2 M).
The DFT calculations of the three complexes namely, cis-PNP-
Mo(CO)4, fac-PNP-Mo(CO)3 and mer-PNP-Mo(CO)3 were also per-
formed to determine if the calculations show any stability differ-
ences among the various complexes. The energy levels for
HOMOꢀ6 to LUMO+6 were displayed as both graphs and molecu-
lar orbital pictures. The molecular orbital compositions for these
energy levels were also calculated and presented as tables for each
The Raman spectra were recorded with a Princeton Instruments
Acton SpectraPro 2500i spectrophotometer using epi-illumination
through a X20 air objective in an Olympus IX-71 inverted micro-
scope with a 0.5 mW 532 nm excitation. Each spectrum was car-
ried out using an average of 10 acquisitions, each of which had
an exposure time of 5 s. All samples were in the solid-state and
placed on glass substrates.

208
S. Goren Keskin et al. / Polyhedron 138 (2017) 206–217
Scheme 1. Synthesis of the PNP pincer ligand.
Scheme 2. Synthesis of PNP-Mo(CO)n complexes and isomers.
2.3. Preparation of ligand (3) and molybdenum complexes (4–6)
filtration. The precipitate was then recrystallized using a hot
mixture of toluene/ethanol (1:1) (crystallized at 0 °C). The white
crystalline product was collected by filtration (75%, 5.6438 g,
11.53 mmol). 1H NMR (300 MHz, CDCl3): 7.69 (d, 4H, J = 8.4),
7.25 (d, 4H, J = 8.1), 7.11 (dd, 2H, J = 8.7), 6.68 (t, 1H, J = 8.1),
6.40 (d, 2H, J = 8.1), 4.06 (t, 4H, J = 6.0), 3.53 (t, 4H, J = 6.0),
2.40 (s, 6H). 1H NMR, 13C{1H} NMR and ESI MS were previously
reported [8].
Scheme 1 shows the routes that were used to synthesize Ligand
3 and Scheme 2 shows the routes that were used to synthesize the
molybdenum complexes.
2.3.1. N,N-bis[2-(p-tolylsulfonoxy)ethyl]phenylamine (1)
N-phenyldiethanolamine (2.7855 g, 15.37 mmol) was dissolved
in dry pyridine and cooled to 0 °C. p-Toluenesulfonylchloride
(7.4134 g, 38.89 mmol) was also dissolved in dry pyridine sepa-
rately and then added to the N-phenyldiethanolamine solution
dropwise while stirring at 0 °C. After complete addition, the mix-
ture was allowed to warm to room temperature and stirred over-
night. The reaction contents were poured into a rapidly stirred
water/ice bath. The resulting precipitate was collected via vacuum
2.3.2. N,N-bis(2-chloroethyl)phenylamine (2)
N-phenyldiethanolamine (0.5 g, 2.759 mmol) was dissolved in
dry benzene and then cooled to 0 °C. Phosphorus oxychloride
(0.9 mL, 96.56 mmol) was added dropwise. When the addition
was complete, the resulting mixture was refluxed at 90 °C for
1 hour, then cooled to room temperature and stirred overnight.

S. Goren Keskin et al. / Polyhedron 138 (2017) 206–217
209
The resulting reaction mixture was then poured into ice-water,
stirred for 1 hour, following which the mixture was transferred
into a separatory funnel. The aqueous phase was then washed
twice with benzene. The organic phases were then combined,
washed twice with brine, dried over MgSO4, vacuum filtered and
the filtrate was evaporated. The crude product was then purified
by column chromatography (silica gel). The eluent was a mixture
of hexanes/CH2Cl2 (2:1). The first fraction was the product. The sol-
vent was then evaporated to obtain the pure product as a yellowish
oil (66%, 0.399g, 1.829 mmol). 1H NMR (300 MHz, CDCl3): 7.34 (t,
2H, J = 7.5), 6.86 (t, 1H, J = 7.5), 6.76 (d, 2H, J = 8.7), 3.8 – 3.66 (m,
8H).
molybdenum(0) (NBD-Mo(CO)4 (0.087 g, 0.29 mmol), was added
to the solution and stirred overnight at room temperature. The sol-
vent was evaporated and the resulting residue was dissolved in
DCM and subsequently precipitated by the addition of hexanes.
The resulting suspension was filtered through a frit. The filtrate
was evaporated to afford a yellow solid (54%, 0.113 g, 0.156 mmol).
1H NMR (300 MHz, CD2Cl2): 7.53–7.25 (m, 20H), 7.08 (t, 2H,
J = 8.4–6.9), 6.73 (t, 1H, J = 7.2), 6.38 (d, 2H, J = 8.4), 3.32 (m, 4H),
2.65 (m, 4H). 13C{1H} NMR (75 MHz, CD2Cl2): 148.9 (s), 138.9
(m), 132.6 (m), 129.8 (s), 129 (s), 128.7 (m), 118.6 (s), 114.9 (s),
48.8 (s), 30.26 (t, J = 15). 31P{1H} NMR (121 MHz, CH2Cl2): 25.8
(s). Low resolution MS (ESI): calculated for cis-PNP-Mo(CO)4 (M+)
724.11, found 727.5. Both the solution state and solid-state IR data
were obtained. (Solid state, in ATR) IR: 2016 cmꢀ1, 1898 cmꢀ1
(combination of two peaks), 1863 cmꢀ1. (Solution in CH2Cl2 in
between KBr discs) IR: 2020 cmꢀ1 and 1903 cmꢀ1 (combination
of 3 bands). Raman spectrum: 2014 cmꢀ1, 1899 cmꢀ1 (combination
of two bands), 1865 cmꢀ1. Elemental analysis: calculated for cis-
PNP-Mo(CO)4 C: 62.90, H: 4.58, N: 1.93; found% C: 62.92%, H:
4.45%, N: 1.80. Mp: decomposed approximately at 240 °C. Hexanes
were layered on a concentrated CH2Cl2 solution of 5 thereby gen-
erating an X-ray diffraction quality single crystal (Fig. 7).
As a second approach, Mo(CO)6 (0.0963 g, 0.3648 mmol) was
dissolved in acetonitrile (ACN) and refluxed under nitrogen for
1–2 hours until a clear yellow solution was obtained. The ligand,
3, (0.1879 g, 0.363 mmol) was dissolved in CH2Cl2 and cannula
transferred into the metal solution. The resulting mixture was stir-
red for 2 days in the dark. The removal of solvent resulted in the
yellow-gray solid, 5 as a crude product mixture in low yield. Hex-
anes were layered onto a concentrated CH2Cl2 solution of 5 thereby
generating an X-ray diffraction quality single crystal that was iden-
tical to the one described above (Fig. 7).
A third way to synthesize compound 5 was developed by simply
loading CO gas into a solution of 4. For this purpose, 35 mg of com-
pound 4 was dissolved in THF in a screw top NMR tube in a glove
box, then the tube of which was connected to the Schlenk line. The
line was vacuumed/N2 three times. Meanwhile the NMR tube was
cooled down to 0 °C. The screw top was slightly opened and
quickly vacuumed, then CO gas was loaded for 15 minutes while
the CO tank was open. The reaction solution was kept under a pos-
itive pressure of CO (the tank was closed) for 50 minutes. The 31P
{1H} NMR was taken and the major product was found to be com-
pound 5. The NMR tube was then connected to the Schlenk line.
The same steps were followed and CO was loaded again this time
15 minutes with open tank and 45 minutes with closed tank. The
conversion of 4 to 5 was quantitative by 31P{1H} NMR spectra.
2.3.3. N,N-bis[2-(diphenylphosphino)ethyl]-phenylamine (PNP) (3)
The ditosylate starting material 1 (2.058 g, 4.2 mmol), was dis-
solved in dry THF under nitrogen in a Schlenk flask. For this, KPPh2
(17.23 ml, 8.6 mmol, 0.5 M in THF) was added via a glass air tight
syringe. The resulting mixture was stirred for 2 hours at room tem-
perature then quenched with ꢁ3.4 mL of dry methanol and stirred
for 30 minutes. The solvents were evaporated and the resulting
Schlenk flask was then transferred into the glovebox. The residue
was washed/suspended in toluene and the resulting washings
were passed through a frit in which there were celite/silica gel/
celite layers. The filtrate was evaporated to dryness and the result-
ing residue was washed with ether and dried again. The analyti-
cally pure product was recovered as a white precipitate (1.9 g,
3.67 mmol, 88%). 1H NMR (300 MHz, CD2Cl2): 7.45–7.32 (m,
20H), 7.07 (t, 2H, J = 7.8), 6.60 (t, 1H, J = 6.6), 6.37 (d, 2H, J = 8.1)
3.35 (m, 4H), 2.29 (t, 4H, J = 7.8). 31P{1H} NMR (121 MHz, CD2Cl2):
ꢀ20 (s). 1H NMR, 13C{1H} NMR and MS data were reported previ-
ously and matched to our data [8].
Another approach to synthesize the ligand using N,N-bis(2-
chloroethyl)phenylamine as the starting material instead of N,N-
bis[2-(p-tolylsulfonyl)ethyl]phenylamine was also used and result-
ing in a ꢁ60% yield.
2.3.4. fac-PNP-Mo(CO)3 (4)
fac-N,N-bis[2-(diphenylphosphino)ethyl]phenylamine(tricar-
bonyl)molybdenum(0)
Both the ligand 3 (0.1g, 0.0001932 mol) and cycloheptatrienet-
ricarbonyl
molybdenum(0)
(CHT-Mo(CO)3)
(0.053 g,
0.0001937 mol) were dissolved in toluene separately in a glovebox
and then the metal solution was added to the ligand solution via a
pipette. The resulting mixture was stirred overnight in a glovebox
which afforded a precipitate. The resulting precipitate was filtered
through a frit containing Celite. The resulting solid was washed
twice with toluene then dissolved in dry DCM and then dried in
vacuo to give 4 as a yellow solid (59%, 0.08 g, 0.0001147 mol).1H
NMR (300 MHz, CD2Cl2): 7–8 ppm (m, 25H, J = 3.0), 3.67 (m, 4H),
3.16 (m, 4H). 31P{1H} NMR (121 MHz, CD2Cl2): 40 (s). HRMS (CI)
calculated for fac-PNP-Mo(CO)3 (M+) 671.1041, found 671.1052.
Hexanes were layered on a CH2Cl2 solution of the product thereby
generating a crystal that was suitable for X-ray analysis (Fig. 6).
Both solution state and solid-state IR data were obtained. Solid
state IR: 1923 cmꢀ1, 1836 cmꢀ1, 1784 cmꢀ1. Solution state IR (solu-
tion in CH2Cl2): 1932 cmꢀ1, 1838 cmꢀ1, 1802 cmꢀ1. Raman Spec-
2.3.6. mer-PNP-Mo(CO)3 (6)
mer-N,N-bis[2-(diphenylphosphino)ethyl]phenylamine(tricar-
bonyl)molybdenum(0)
50 mg (0.0689 mmol) of 5 was transferred into a 10 mL Schlenk
flask and then dissolved in CH2Cl2 to get a clear, burnt orange solu-
tion. A stock solution of (CH3)3NO was prepared by dissolving
49 mg (0.0006524 mol) material in 10 mL of CH2Cl2. 1 equivalence
(1.06 mL) of the (CH3)3NO solution was then transferred into the
Schlenk flask via a syringe following which it was stirred overnight
in the dark at room temperature. 31P{1H} NMR and 1H NMR were
obtained. However, there was no proof of the product formation.
Accordingly, hexanes were layered on top of the NMR solution in
order to grow crystals. Compound 6 was obtained as a minor pro-
duct of the reaction and observed by X-ray crystallography (Fig. 8).
The remainder of the product consisted of the unreacted starting
material 5.
trum: 1917 cmꢀ1
,
1828 cmꢀ1
,
1796 cmꢀ1 and shoulder at
1782 cmꢀ1. Elemental Analysis: calculated for fac-PNP-Mo(CO)3-
ꢂ2H2O% C: 60.58%, H:5.08%, N: 1.91%; found% C: 61.08%, H: 4.81%,
N: 1.88%. Mp: decomposed around 240 °C.
2.3.5. cis-PNP-Mo(CO)4 (5)
cis-N,N-bis[2-(diphenylphosphino)ethyl]phenylamine(tetracar-
bonyl)molybdenum(0)
Ligand 3, (0.15 g, 0.29 mmol), was dissolved in toluene in the
glovebox following which (2,5-norbornadiene)tetracarbonyl-
As a second approach to the synthesis of the meridional isomer,
the facial isomer 4 was dissolved in THF and stirred for three weeks
in the dark in a glovebox. Some of the facial isomer remained

210
S. Goren Keskin et al. / Polyhedron 138 (2017) 206–217
Table 2
intact. However, some of the product decomposed to give the
meridional isomer, and tetracarbonyl complex 5 in solution, and
Mo(0) as a dark precipitate. 31P{1H} NMR (121 MHz, CD2Cl2):
57.6 (s) meridional product, 40 (s) facial product, 25 (s) tetracar-
bonyl complex.
The bond angles around octahedral Mo atom and Mo–C–O for compound 4.
fac-PNP-Mo(CO)3
Angle
Ideal angle
Difference
C1–Mo1–C2
C1–Mo1–C3
C1–Mo1–N1
C1–Mo1–P1
C1–Mo1–P2
C2–Mo1–C3
C2–Mo1–N1
C2–Mo1–P1
C2–Mo1–P2
C3–Mo1–N1
C3–Mo1–P1
C3–Mo1–P2
N1–Mo1–P1
N1–Mo1–P2
P1–Mo1–P2
C38–Mo2–C39
C38–Mo2–C40
C38–Mo2–N2
C38–Mo2–P3
C38–Mo2–P4
C39–Mo2–C40
C39–Mo2–N2
C39–Mo2–P3
C39–Mo2–P24
C40–Mo2–N2
C40–Mo2–P3
C40–Mo2–P4
N2–Mo2–P3
N2–Mo2–P4
P3–Mo2–P4
93.0(1)
82.3(1)
90.5(1)
86.1(1)
169.0(1)
87.5(1)
171.7(1)
94.1(1)
98.0(1)
100.4(1)
168.4(1)
97.6(1)
78.67(7)
78.68(7)
93.62(3)
92.1(1)
81.2(1)
94.3(1)
82.8(1)
170.1(1)
88.9(1)
169.0(1)
94.3(1)
96.9(1)
100.9(1)
163.7(1)
94.6(1)
77.65(7)
77.62(7)
100.84(3)
90
90
90
90
180
90
180
90
90
90
180
90
90
90
90
90
90
90
90
180
90
180
90
90
90
180
90
90
90
90
ꢀ3
3.03
ꢀ0.5
ꢀ3.1
11
2.4. X-ray crystal structure analysis
ꢀ2.5
ꢀ8.3
4.1
Single crystals of 4, 5 and 6 suitable for X-ray diffraction analy-
sis were grown by using the methods that were described in
Section 2.3.
The single crystal diffraction data were collected on a Nonius
Kappa CCD diffractometer using a graphite monochromator with
8
10.4
ꢀ11.6
7.6
ꢀ11.32
ꢀ11.31
3.62
2.1
Mo Ka radiation (k = 0.71073 Å) and an Oxford Cryostream low
temperature device. Absorption corrections were applied by using
Multi-scan [23]. The data reductions were performed using
DENZO-SMN [24]. The structure was solved by direct methods
using the program SIR97 [25] and refined by full-matrix least-
squares on F2 with anisotropic displacement parameters for the
non-H atoms using SHELXL-97 [26]. The hydrogen atoms were calcu-
lated in ideal positions with isotropic displacement parameters set
to 1.2xUeq of the attached atom (1.5xUeq for methyl hydrogen
atoms). Neutral atom scattering factors and values used to calcu-
late the linear absorption coefficient are taken from the Interna-
tional Tables for X-ray Crystallography (1992) [27]. The
crystallographic data and structure refinement for complexes 4, 5
and 6 are listed in Table 1. The selected bond lengths and angles
for 4, 5 and 6 are presented in Tables 2–7.
ꢀ8.8
4.3
ꢀ7.2
ꢀ9.9
ꢀ1.1
ꢀ11
4.3
6.9
10.9
ꢀ16.3
4.6
ꢀ12.34
ꢀ12.37
10.84
Carbonyl groups
O1–C1–Mo1
O2–C2–Mo1
O3–C3–Mo1
O4–C38–Mo2
O5–C39–Mo2
O6–C40–Mo2
176.0(3)
178.7(3)
171.5(3)
176.2(3)
177.0(3)
171.4(3)
180
180
180
180
180
180
ꢀ4
3. Results and discussion
ꢀ1.3
ꢀ8.5
ꢀ3.8
ꢀ3
3.1. Syntheses and characterization of the complexes
ꢀ8.6
The PNP ligand, 3, was prepared according to Kostas’ method
[8]. It was obtained as a white solid. There were two synthetic
Table 3
The bond angles around octahedral Mo atom and Mo–C–O for compound 5.
Table 1
Crystal data and structure refinement for compound 4, 5 and 6 respectively.
cis-PNP-Mo(CO)4
Angle
Ideal angle
Difference
CCDC
762871
762870
762872
C1–Mo–C2
C1–Mo–C3
C1–Mo–C4
C1–Mo–P1
C1–Mo–P2
C2–Mo–C3
C2–Mo–C4
C2–Mo–P1
C2–Mo–P2
C3–Mo–C4
C3–Mo–P1
C3–Mo–P2
C4–Mo–P1
C4–Mo–P2
P1–Mo–P2
87.4(1)
85.4(1)
86.1(1)
174.5(1)
90.5(1)
172.7(1)
88.3(2)
90.7(1)
94.0(1)
90.0(1)
96.37(9)
87.25(9)
88.7(1)
175.8(1)
94.78(3)
90
90
90
180
90
180
90
90
90
90
90
180
90
180
90
ꢀ2.6
ꢀ4.6
ꢀ3.9
ꢀ5.5
ꢀ0.5
ꢀ7.3
ꢀ1.7
0.7
Formula
FW
T (K)
Crystal system
Space group
a (Å)
C
37H33NO3P2Mo C38H33NO4P2Mo C37H33NO3P2Mo
697.52
153(2)
monoclinic
C2
19.456(4)
12.232(2)
13.608(3)
90
95.38(3)
90
3224.2(11)
4
725.53
153(2)
monoclinic
P21/n
697.52
153(2)
monoclinic
P21/c
10.027(2)
19.158(4)
18.646(4)
90
17.620(4)
17.198(3)
20.971(4)
90
b (Å)
c (Å)
4
0
a
(°)
b (°)
102.853(1)
90
100.40(3)
90
6.37
ꢀ2.74
ꢀ1.3
ꢀ4.2
4.78
c
(°)
V (Å3)
3492.3(12)
4
3224.2(11)
8
Z
q
l
(g/cm3)
(mmꢀ1
1.437
0.54
1.38
1.482
)
0.51
0.56
F(000)
1432
1488
2864
Carbonyl groups
O1–C1–Mo
O2–C2–Mo
O3–C3–Mo
O4–C4–Mo
Crystal size (mm) 0.09 ꢃ 0.05 ꢃ 0.03 0.30 ꢃ 0.02 ꢃ 0.02 0.46 ꢃ 0.06 ꢃ 0.07
175.4(3)
175.2(3)
173.1(3)
177.4(3)
180
180
180
180
ꢀ4.6
ꢀ4.8
ꢀ6.9
ꢀ2.6
h (°)
Index ranges
1.0 to 27.5
1.0 to 27.5
1.0 to 27.5
ꢀ25 ꢄ h ꢄ 24
ꢀ15 ꢄ k ꢄ 15
ꢀ2 ꢄ l ꢄ 17
Multi-scan
ꢀ13 ꢄ h ꢄ 13
ꢀ24 ꢄ k ꢄ 23
ꢀ24 ꢄ l ꢄ 24
Multi-scan
ꢀ20 ꢄ h ꢄ 22
ꢀ22 ꢄ k ꢄ 18
ꢀ27 ꢄ l ꢄ 17
Multi-scan
Absorption
correction
Maximum and
minimum
0.953 and 0.984 0.990 and 0.863 0.994 and 0.896
routes to obtain 3, i.e., by using 2 or by using 1. Both were success-
ful to prepare and used to synthesize 3. The yield of 1 was higher
than that of 2 by approximately 9%. The yield of 3 by using 1 as
the starting material is 88%. The yield of 3 by using 2 as the starting
material is ꢁ60%.
transmission
GOF on F2
1.16
1.02
1.15
R1, R2 I > 2
r
(I)
0.0955, 0.1822
0.1571, 0.2087
1.70 and ꢀ1.45
0.0640, 0.1452
0.0758, 0.1487
0.66 and ꢀ0.63
0.0585, 0.1225
0.0783, 0.1391
0.94 and ꢀ0.93
R1, R2 (all data)
Largest diff. peak
and hole (e Åꢀ3
The molybdenum reagents, (Mo(ACN)x(CO)y, pip2Mo(CO)4,
NBD-Mo(CO)4 and CHT-Mo(CO)3) were allowed to reacted with
)

S. Goren Keskin et al. / Polyhedron 138 (2017) 206–217
211
Table 4
5. The Mo(ACN)x(CO)y reagent was obtained in situ by refluxing
Mo(CO)6 in ACN for 1–2 hours. However, the commercial NBD-
Mo(CO)4 reagent afforded the purest product for the synthesis of
cis-PNP-Mo(CO)4 which has been fully characterized. In order to
obtain the meridional product mer-PNP-Mo(CO)3, 6, cis-PNP-Mo
(CO)4 was allowed to react with (CH3)3NO. (CH3)3NO was prepared
by sublimation of (CH3)3NOꢂ2H2O. The reaction mixtures of com-
plex 5 and (CH3)3NO resulted in a mixture of both the meridional
product and the unreacted cis-PNP-Mo(CO)4 as crystals.
The bond angles around octahedral Mo atom and Mo-C-O for compound 6.
mer-PNP-Mo(CO)3
Angle
Ideal angle
Difference
C35–Mo–C36
C35–Mo–C37
C35–Mo–P1
C35–Mo–P2
C35–Mo–N
C36–Mo–C37
C36–Mo–P1
C36–Mo–P2
C36–Mo–N
C37–Mo–P1
C37–MoP2
C37–Mo–N
N–Mo–P1
78.7(5)
163.0(5)
99.8(3)
89.0(3)
103.1(4)
84.5(5)
101.4(4)
99.6(4)
178.1(4)
86.4(3)
99.6(4)
90.8(3)
79.1(2)
79.8(2)
158.5(1)
90
180
90
90
90
90
90
90
180
90
ꢀ11.2
ꢀ16.9
9.8
ꢀ1
13.1
ꢀ5.4
11.4
9.6
The second approach to the synthesis of the meridional product
involved the use of fac-PNP-Mo(CO)3. For this purpose, the facial
isomer was dissolved in THF. The resulting 31P{1H} NMR spectrum
has a singlet peak initially at 40 ppm. The solution was stirred in
the dark for three weeks, following which the 31P{1H} NMR spec-
trum of the solution was measured. The three singlets that were
observed: the peak at 57.6 ppm belonging to meridional product
6, the peak at 40 ppm belonging to facial product 4 and the peak
at 25 ppm belonging to the tetracarbonyl complex 5. Refluxing of
4 in THF overnight resulted in the same outcome. An attempt
was made to load PR3 (R = Et3, Ph3, Et2Ph, EtPh2) into a solution
of the facial isomer. However, the same results were evident: Some
of the facial isomer decomposed thereby generating the tetracar-
bonyl product 5, the meridional isomer 6, some unreacted facial iso-
mer 4, some free PNP ligand 3 and a black precipitate Mo(0).
The experimental observations suggest that the conversion of
the facial to the meridional isomer occurs via a dissociative mecha-
nism. There is an equilibrium between the isomers and the facial
isomer must be the thermodynamic product. Because, stirring a
THF solution of pure facial isomer 4, gives a black precipitate Mo
(0), cis-PNP-Mo(CO)4 5, mer-PNP-Mo(CO)3 6, as well as intact fac-
PNP-Mo(CO)3 4. Some of the fac-PNP-Mo(CO)3 must have decom-
posed to give Mo(0), and some of the excess carbonyls must have
been picked up by other fac-PNP-Mo(CO)3 to give cis-PNP-Mo
(CO)4, the rest of the free carbonyls must be picked up by the five
coordinate intermediate resulted from the dissociation of fac-PNP-
Mo(CO)3 thereby generating the mer-PNP-Mo(CO)3 isomer. This
opinion also explains why the meridional product is the minor
product. Furthermore, this view is also supported by the literature.
Crabtree explained the reason that the ML3(CO)3 complexes
undergo a dissociative mechanism. He also specifically provided
the molybdenum complex of MoL3(CO)3 as an example [28].
All the complexes were found to be soluble in CH2Cl2. Solution
and solid-state IR data and solid-state Raman data were obtained
for the fac-PNP-Mo(CO)3 and cis-PNP-Mo(CO)4 (Figs. 4 and 5) and
found to be consistent with the proposed molecular structures.
Each of the three complexes adopted the point group Cs which is
as evident from each of the crystal structures. Group theory calcu-
lations of the CO stretches for the three complexes resulted in the
ꢀ1.9
ꢀ3.6
9.6
90
90
90
90
0.8
ꢀ10.9
ꢀ10.2
ꢀ21.5
N–Mo–P2
P1–Mo–P2
180
Carbonyl groups
O1–C35–Mo
O2–C36–Mo
O3–C37–Mo
173(1)
177(1)
175(1)
180
180
180
ꢀ7
ꢀ3
ꢀ5
Table 5
The bond lengths Å around Mo atom for compound 4.
fac-PNP-Mo(CO)3
Bond lengths
Mo1–C1
Mo1–C2
Mo1–C3
Mo1–N1
Mo1–P1
Mo1–P2
Mo2–C38
Mo2–C39
Mo2–C40
Mo2–N2
Mo2–P3
Mo2–P4
1.979(4)
1.933(3)
1.977(3)
2.466(3)
2.496(1)
2.512(1)
1.977(4)
1.937(3)
1.977(3)
2.489(3)
2.495(1)
2.543(1)
Table 6
The bond lengths Å around Mo atom for compound 5.
cis-PNP-Mo(CO)4
Bond lengths
Mo–C1
Mo–C2
Mo–C3
Mo–C4
Mo–P1
Mo–P2
1.982(3)
2.027(4)
2.049(3)
1.978(3)
2.551(1)
2.5404(9)
following:
= 3A0 for the complex 6. All bands were found to be both Raman
and IR active.
C C
= 2A0 + A00 for complex 4, = 3A0 + A00 for complex 5,
C
Table 7
The bond lengths Å around Mo atom for compound 6.
The IR spectrum of 4 revealed three bands as expected from the
above calculation. However, the Raman spectrum evidenced three
bands and a shoulder (1782 cmꢀ1). The Raman and IR spectra for
complex 5 resulted in three stretches in the measurements while
calculation afforded four stretches. But, one of the stretches (in
both spectra) is broad and appears to look like a combination of
two bands (1898 cmꢀ1). However, the solution state IR spectrum
of 5 consisted of two stretches. It appears that the three bands
between 1800 and 1900 cmꢀ1 combined to result in a broad peak
mer-PNP-Mo(CO)3
Bond lengths
Mo–C35
Mo–C36
Mo–C37
Mo–P1
Mo–P2
N–Mo
2.00(1)
1.93(1)
1.99(1)
2.458(3)
2.442(3)
2.424(9)
at 1903 cmꢀ1. The remaining peak was observed at 2020 cmꢀ1
.
one equivalent of the ligand in order to obtain the corresponding
PNP-Mo(CO)n complexes. The reaction of CHT-Mo(CO)3 with the
PNP ligand 3, resulted in the generation of fac-PNP-Mo(CO)3, 4.
The reactions of Mo(ACN)x(CO)y, pip2Mo(CO)4 and NBD-Mo(CO)4
with the ligand, 3, resulted in the generation of cis-PNP-Mo(CO)4,
3.2. Structure of complexes 4, 5 and 6
The solid-state structures of complexes 4, 5 and 6 were deter-
mined by X-ray diffraction analysis and the resulting ORTEP repre-

212
S. Goren Keskin et al. / Polyhedron 138 (2017) 206–217
octahedral environment. The coordination environment around
the metal center is defined by two phosphorus atoms from the
PNP ligand, one nitrogen atom from the PNP ligand and three car-
bon atoms from the carbonyl ligands. The phosphorus atoms are
arranged trans to each other and the carbonyl ligands have the
meridional arrangement.
The bond lengths and angles for 4, 5 and 6 were compared with
those of similar complexes that can be found in the literature.
Complex 4 is consistent with the literature values in both angles
and bond lengths [13]. The literature complexes similar to those
5 have slightly different angles. For example, the P-Mo-P angle
for [3] is 86.75° while that for [13] is 88.1 (Fig. 3, PNP-Mo(CO)4
complex upper left is from [3] and the PNP-Mo(CO)3 complex
upper right is from [13]). Compound 5 has P1-Mo-P2 angle of
94.78° but, bond lengths for the literature complexes and 5 are
consistent. However, for the complex 6, carbonyl carbons that
are trans to each other (C35-Mo-C37) have a bond angle of 163°.
On the other hand, the complex in the literature (ligand has an
imine N) has an angle of 171.07°. The trans phosphines of 6 and
the literature values are consistent. However, the trans carbonyl
carbons do not match those of the bond angles [15] (the complex
[15] is abbreviated as mer-[Mo(CO)3(PNCHP)] in the literature.
The N-donor in the complex from [15] is an imine nitrogen). All
bond lengths are also consistent except for that of the Mo–N bond.
6 that has a Mo–N distance of 2.424 Å, while that of the literature
complex has a value of 2.237 Å.
Fig. 4. Solid state IR and Raman spectra of fac-PNP-Mo(CO)3.
3.3. Stability of complexes 4, 5 and 6
A similar mer-PNP-Mo(CO)3 product has also been reported by
Siclovan et. al. However, these authors also could not isolate the
meridional isomer, despite the fact that they could successfully
synthesize and purify the facial isomer [12] (Fig. 3, complex upper
left). The choice of a facial or meridional isomer was explained on
the basis of electronics rather than sterics in the literature: When
a monodentate ligand is negatively charged such as a halogen,
the preferred geometry becomes meridional due to the electrostatic
repulsions between the halogens. In the case of the neutral mon-
odentate ligands such as a carbonyl ligand, the preferred geometry
is facial in the complexes [1,19,20]. The explanation of the stability
of the complexes of meridional and facial isomers –electronics of
the monodentate ion plays a role rather than sterics- was not gen-
uinely explaining experimental observations in the present study.
Because in those journal articles, the only reason provided was
repulsion of negative charges on halogens resulted in the merid-
ional geometry. The reason for the facial geometry being the pre-
ferred geometry in PNP-Mo complexes with neutral monodentate
ligand i.e. carbonyls was not provided. Moreover, the Jahn–Teller
Fig. 5. Solid state IR and Raman spectra of cis-PNP-Mo(CO)4.
sentation can be seen in Figs. 6, 7 and 8, and the crystallographic
and structural refinement data are presented in Table 1. In the
crystal structure of complex 4 (Fig. 6), the unit cell contains two
fac-PNP-Mo(CO)3 complexes, both of which a have six-coordinate
Mo (0) that lies at the center of an octahedral environment. The
coordination environment around the metal center is defined by
two phosphorus atoms from the PNP ligand that are arranged cis
to each other. Namely, one N atom from the PNP ligand and three
carbon atoms from the carbonyl ligands that have a facial arrange-
ment. The difference between these two facial complexes depends
on the orientation of the five membered rings that were formed
upon chelation, which means that C16 and C17 in one structure
have a different orientation than C53 and C54 in the other struc-
ture. In the case of complex 5, the six-coordinate Mo (0) lies at
the center of an octahedral environment. The coordination environ-
ment around the metal center is defined by two phosphorus atoms
from the PNP ligand and four carbon atoms from the carbonyl
ligands. The P atoms are arranged cis to each other. In the case of
complex 6, the six-coordinate Mo (0) also lies at the center of an
effect, trans effect, p-back donation and sterics were not taken into
account in those journal articles. In the present study, a variety of
the aforementioned effects have been considered and discussions
of each are provided.
The first argument concerns the trans effect,
p-back donation
and bond distances. The main reason for compound 4 being more
stable than compound 6 is that all of the t2g-type d orbitals in
the facial geometry are effectively involved in Mo–CO
p-back
bonding. In the case of compound 6 which has a meridional geom-
etry, the two CO ligands are located trans to each other compete
each another as
p acids and therefore one of the t2g-type d orbitals
is less effectively involved compared with the facial isomer. As a
result, longer metal–carbon bond distances are observed for the
meridional complex and therefore, the facial isomer is more stable.
Tables 5, 6 and 7 show the distances of each atom to Mo center in
the octahedral environment. The Mo–C distances are slightly
shorter for complex 4 than those for complex 6. The carbonyls that
are located trans to the phosphines benefit from the p-back dona-

S. Goren Keskin et al. / Polyhedron 138 (2017) 206–217
213
Fig. 6. Structure of fac-PNP-Mo(CO)3 (4).
Fig. 8. Structure of mer-PNP-Mo(CO)3 (6).
Fig. 7. Structure of cis-PNP-Mo(CO)4 (5).
tion and are more tightly held by the metal, as a result, the M–C
distances are shorter than those for the carbonyls [28]. This is
exactly what was observed for the facial complex 4 for which all
the M–C distances are less than 2 Å (1.933–1.979 Å). In the case
of complex 6, phosphines that have less trans effect than the car-
bonyls, are trans to each other and the M–P bonds are shorter. As
expected, the carbonyls that are located trans to each other have
longer M–C distances (Mo–C35: 2.00 Å and Mo–C37: 1.99 Å) which
Table 8
The bond angles around octahedral Mo atom for the compound mer-PNP-MoCl3 [1].
mer-PNP-MoCl3
Angle
Ideal angle
Difference
N1–Mo–P2
N1–Mo–P3
N1–Mo–Cl5
N1–Mo–Cl6
N1–Mo–Cl7
P2–Mo–Cl5
P2–Mo–Cl6
P2–Mo–Cl7
P2–Mo–P3
P3–Mo–Cl5
P3–Mo–Cl6
P3–Mo–Cl7
Cl5–Mo–Cl6
Cl5–Mo–Cl7
Cl6–Mo–Cl7
80.5
81.1
87
90
90
90
90
180
90
90
90
180
90
90
90
180
90
90
ꢀ9.5
ꢀ8.9
ꢀ3
91.1
177.2
86.8
88.9
101.2
161.2
88.5
95.2
97
1.1
ꢀ2.8
ꢀ3.2
ꢀ1.1
11.2
ꢀ18.8
ꢀ1.5
5.2
results in a poor p-back donation. Furthermore, the M–C distances
in which carbonyls are trans to the N atom for both complex 4 and
6 are shorter (for 4 Mo1–C2: 1.933 Å, Mo2–C39:1.937 Å; for 6 Mo–
C36: 1.93 Å) than the other M–C distances (1.98 Å for 4, 1.99–
2.00 Å for 6) (trans to either phosphine or carbonyl). The reason
for this is that there is a slight
p-back donation from the metal
7
to the phosphine; therefore, the N atom becomes a better electron
donor than the phosphine and carbonyls trans to the N benefit
from that electron donation [29].
175.5
91
91.1
ꢀ4.5
1
1.1

214
S. Goren Keskin et al. / Polyhedron 138 (2017) 206–217
Table 9
isomers. The angles around the Mo atom were compared for each
complex to determine if there is any angle distortion that could
cause the difference in stability. The angles of the meridional iso-
mer around the octahedral Mo atom are the most distorted of the
three complexes. Furthermore, the angles between the Mo and
the carbonyl groups that are supposed to be 180° are again the
most distorted for the meridional isomer. The most distorted angles
for the meridional isomer 6 are as follows; C35–Mo–C36: 78.8°,
C35–Mo–C37: 163.1°, C35–Mo–N: 103.1°, N–Mo–P1: 79.1°, C36–
Mo–P1: 101.4°, N–Mo–P2: 79.8°, P1–Mo–P2: 158.5°. The difference
between the ideal angles changes between 10.2° and 21.5° (7
angles). The most distorted angles for the facial isomer 4 are as fol-
lows (angles greater than 10° distortions); C3–Mo1–N1: 100.4°,
C3–Mo1–P1: 168.4°, N1–Mo1–P1: 78.68°, C1–Mo1–P2: 169.0°,
N1–Mo1–P2: 78.69°, C40–Mo2–N2: 100.9°, C40–Mo2–P3: 163.7°,
The bond lengths around Mo atom for compound mer-
PNP-MoCl3 [1].
mer-PNP-MoCl3
Bond lengths (Å)
Mo–N
2.332
2.524
2.541
2.417
2.395
2.376
Mo–P2
Mo–P3
Mo–Cl5
Mo–Cl6
Mo–Cl7
The second argument is whether having a less angle distortion
is the reason for the facial complex being more stable than the
meridional one or not; in other words, whether the Jahn–Teller
effect plays a role in the stability between the meridional and facial
Table 10
Composition of selected HOMOs and LUMOs of the complex 4, expressed in terms of composing fragments.
MO
eV
P2 (PPh2CH2)
P1 (PPh2CH2)
Mo
C2 (CO)
C3 (CO)
N (N(CH2)2Ph)
C1 (CO)
L+6
L+5
L+4
L+3
L+2
L+1
LUMO
HOMO
H-1
H-2
H-3
H-4
H-5
H-6
ꢀ0.48
ꢀ0.56
ꢀ0.64
ꢀ0.75
ꢀ0.85
ꢀ1.05
ꢀ1.11
ꢀ4.86
ꢀ4.88
ꢀ4.98
ꢀ6.52
ꢀ6.68
ꢀ6.93
ꢀ6.96
46
8
44
29
75
8
81
0
3
4
2
5
44
49
10
81
17
59
13
90
14
1
1
5
20
71
54
48
4
2
9
4
4
0
1
62
63
61
4
3
1
0
1
2
2
1
0
0
16
16
2
1
0
2
0
3
3
1
0
1
2
13
15
0
0
1
33
7
20
1
3
1
1
3
1
1
3
1
4
1
2
0
1
16
3
12
1
1
0
0
72
19
1
0
0
1
0
2
Table 11
Composition of selected HOMOs and LUMOs of the complex 5, expressed in terms of composing fragments.
MO
eV
P2 (PPh2CH2)
P1 (PPh2CH2)
C3 (CO)
Mo
C1 (CO)
C4 (CO)
N (N(CH2)2Ph)
C2 (CO)
L+6
L+5
L+4
L+3
L+2
L+1
LUMO
HOMO
H-1
H-2
H-3
H-4
H-5
H-6
ꢀ0.72
ꢀ0.82
ꢀ0.91
ꢀ1.01
ꢀ1.15
ꢀ1.2
77
92
3
87
3
10
5
3
3
3
1
6
15
1
95
4
53
19
37
3
2
1
2
22
46
25
1
1
0
3
16
29
21
1
1
2
1
1
1
3
0
1
0
1
1
9
8
8
2
0
1
1
2
0
0
1
1
0
2
9
6
5
8
0
1
0
2
1
0
0
0
0
0
35
40
6
15
71
28
3
2
1
0
3
17
29
22
1
2
11
9
9
12
11
40
37
56
52
1
ꢀ1.33
ꢀ5.32
ꢀ5.51
ꢀ5.66
ꢀ5.73
ꢀ6.79
ꢀ6.83
ꢀ6.92
3
10
10
0
0
0
0
0
0
21
67
3
2
Table 12
Composition of selected HOMOs and LUMOs of the complex 6, expressed in terms of composing fragments.
MO
eV
P2(PPh2CH2)
P1(PPh2CH2)
Cdown(CO)
Mo
C2(CO)
N(N(CH2)2Ph)
C1(CO)
L+6
L+5
L+4
L+3
L+2
L+1
LUMO
HOMO
H-1
H-2
H-3
H-4
H-5
H-6
ꢀ0.58
ꢀ0.62
ꢀ0.67
ꢀ0.79
ꢀ0.85
ꢀ0.93
ꢀ0.98
ꢀ4.65
ꢀ5.01
ꢀ5.11
ꢀ6.8
23
4
13
86
13
9
55
17
3
3
4
2
0
1
0
0
0
0
0
23
1
12
0
0
0
0
6
1
5
2
2
8
9
67
61
56
2
1
0
8
0
11
3
6
17
16
2
14
14
0
40
7
9
1
12
3
4
17
17
1
14
14
0
42
16
11
10
51
3
6
0
46
63
100
4
18
68
21
32
4
0
1
1
1
51
34
0
ꢀ6.82
ꢀ6.92
ꢀ7.04
0
0
0
1
0
1
0
0
0
94
0

S. Goren Keskin et al. / Polyhedron 138 (2017) 206–217
215
Fig. 9. Some of the molecular orbitals of compound 4, 5 and 6.
Fig. 10. The energy levels of the complexes 4, 5 and 6.
N2–Mo2–P3: 77.66°, N2–Mo2–P4: 77.63°, P3–Mo2–P4: 100.84°.
The difference between the ideal angles for Mo1 ranges between
10.4° and 11.32° (5 angles); for Mo2, changes between 10.84°
and 16.3°. The most distorted angle for complex 5 was found to
be only 7.3°. The less distorted the angles are the more stable the
molecule should be. Moreover, the P–Mo–P angle for the merid-
ional isomer 6 is distorted by 21.5°. While the C–Mo–C (trans)
angle is distorted by 16.9°. The largest angle distortions between
the C–Mo–C atoms in the facial isomer 4 are 8.8° for C38–Mo2–
C40 and 3.03° for C1–Mo1–C3. P–Mo–P angle in complex 4 is dis-
torted by 3.62° for P1–Mo1–P2 and 10.84° for P3–Mo2–P4. The dis-
tortions in bond angles for compound 4 are significantly smaller
than those for compound 6. Tables 2, 3 and 4 show the octahedral
angles around the Mo atom and the Mo–C–O angles for each com-

216
S. Goren Keskin et al. / Polyhedron 138 (2017) 206–217
plex, as well as the deviation from the ideal angles. One can find
the corresponding pictures for crystal structures to see the labeled
atoms in the Figs. 6, 7 and 8.
ring strain formed upon the chelation. The angles around each
atom in the chelation rings were also examined to determine if
there is any difference between the isomers. Distortions were very
similar for the two isomers (facial and meridional); this outcome
also supports the results of the DFT calculation for the ring strain
upon chelation was not the reason for the stability of the facial
isomer.
The bond angles and bond distances for mer-PNP-MoCl3 that
were reported by Fernandez-Trujillo et al. was also studied in order
to compare the data with 6 [1]. It was not surprising that the mer-
PNP-MoCl3 complex is more stable than mer-PNP-Mo(CO)3 and can
synthetically be made because Clꢀ has a less trans effect than CO
groups and also than phosphines; in addition, there are ionic bonds
between Mo3+ and each Clꢀ which should be stronger than the
Mo–CO bonds. There are still angle distortions for mer-PNP-MoCl3
compared to complexes 4 and 6. The angle distortions which are
greater than 10° around the Mo atom in the complex mer-PNP-
MoCl3 taken from the literature are P2–Mo–P3: 161.2° and P2–
Mo–Cl7: 101.2° and their differences from ideality are 18.8° and
11.2°, respectively. The remaining angles and distortions can be
found in Table 8. The distances of the atoms to Mo in octahedral
environment for mer-PNP-MoCl3 are listed in Table 9.
The third argument is whether composition of the molecular
orbitals play a role in compound 4 being the more stable isomer
than compound 6 or not. Some DFT calculations have been per-
formed to determine the composition of the molecular orbitals.
GaussSum version 2.1 has been used to find the orbital contribu-
tions. As it is indicated in Tables 10, 11 and 12, the biggest contri-
bution in the HOMO level of each complex is coming from the
molybdenum atom. In the case of compound 5, the percent contri-
bution of the Mo atom (40%) to the HOMO is less than compound 4
and 6 (62% and 67% respectively). However, nitrogen group con-
tributes more in 5 (35%), and interestingly there is no bond
between N and Mo in this compound. The percent contribution
from N for complexes 4 and 6 are 3% and 0% respectively. The low-
est HOMO level (which should make the complex more stable)
among all of the complexes also belongs to 5. LUMO levels are
mainly ligand based in all three complexes. It can be concluded
that HOMO to HOMOꢀ2 levels of complexes 4 and 6 are metal
base. HOMO level of complex 4 is slightly more stable in which
Mo is contributing 62% and two of the carbonyl groups are con-
tributing 16% each. On the other hand, HOMO level of complex 6
is composed of 67% metal orbital and 23% is coming from one of
the carbonyls. Contribution of the two carbonyl groups to the
molecular orbital in complex 4 compared with complex 6 that
has only one carbonyl overlap with the metal, should be the reason
for the lower level HOMO for 4. HOMOꢀ1 of complexes 4 and 6 are
not much different because the main contribution is coming from
the metal and two of the carbonyls in both complexes. However,
HOMOꢀ2 is more stable in complex 6, because metal and all the
carbonyls are contributing to the molecular orbital. Overlap of
the metal d orbital and three carbonyls are effectively engaged
(Fig. 9). The metal carbonyl bonds can be clearly seen in HOMO
to HOMOꢀ2 levels for each complex and ligand base orbitals of
LUMOs. Fig. 10 shows the energy levels (HOMOꢀ6 to LUMO+6)
for each complex in (kcal/mol).
4. Conclusions
Three different PNP-Mo(CO)n complexes where n = 3, 4 were
synthesized, characterized, and the conversion conditions between
them were investigated. The synthesis and purification of fac-PNP-
Mo(CO)3 and cis-PNP-Mo(CO)4 were successfully achieved, but the
synthesis and isolation of mer-PNP-Mo(CO)3 could never be con-
trolled. mer-PNP-Mo(CO)3 was observed as a minor product in
the syntheses reactions by 31P{1H} NMR and X-ray diffraction mea-
surements. The solution of pure fac-PNP-Mo(CO)3 undergoes
decomposition thereby generating the tetracarbonyl complex, the
meridional isomer, Mo(0) as a dark precipitate along with some
facial isomer that was left intact in solution.
The facial isomer is favored compared with the meridional iso-
mer mainly because of trans effect and p-back donation (effective
involvement of all t2g orbitals in the facial geometry). Possible other
effects such as Jahn–Teller effect (angle distortions) has been dis-
cussed. The reason for the preferred geometry being the facial
arrangement is not electronics as it was reported in the literature
[1,19,20]. DFT calculations for orbital compositions and DFT calcu-
lations for ring strain in the ligand orientation support the
conclusions.
Acknowledgements
We gratefully acknowledge the Robert A. Welch Foundation [F-
1631], the Petroleum Research Fund administered by the American
Chemical Society [47022-G3], the National Science Foundation
[CHE-0639239, CHE-0741973 and CHE-0847763], the American
Heart Association [0765078Y], the UT-CNM and UT-Austin.
The Raman measurements were performed in Dr. Willet’s labo-
ratory in UT Austin Department of Chemistry and we gratefully
acknowledge IGERT National Science Foundation grant [DGE-
0549417].
We would like to thank Dr. Bradley J. Holliday and Drew Lewis.
Appendix A. Supplementary data
CCDC 762871, 762870, and 762872 contain the supplementary
crystallographic data for 4, 5 and 6. These data can be obtained free
of charge via http://www.ccdc.cam.ac.uk/conts/retrieving.html, or
from the Cambridge Crystallographic Data Centre, 12 Union Road,
Cambridge CB2 1EZ, UK; fax: (+44) 1223-336-033; or e-mail:
deposit@ccdc.cam.ac.uk.
The last possible effect studied is whether the ring strain which
can form upon chelation, affects the stability of the complexes or
not. The facial and meridional isomers were taken, molybdenum
and the monodentate ligands (carbonyls) were removed to elimi-
nate any effect coming from the stability formed upon chelation,
and then the DFT calculations were performed for the both orien-
tations (facial and meridional) of the PNP ligand. Only one of the
two facial complexes shown in Fig. 6 was used for the calculations.
Energy difference between the facial and meridional orientation of
the PNP ligand was found as 0.13347799 Ha (=83.7587026 kcal/-
mol) and surprisingly the meridional orientation of the PNP ligand
is the more stable one in terms of the ring strain than the facial ori-
entation in the calculation. Therefore, the reason for the facial com-
plex is more stable than the meridional one is not coming from the
References
[1] M.J. Fernandeztrujillo, M.G. Basallote, P. Valerga, M.C. Puerta, D.L. Hughes, J.
Chem. Soc. Dalton Trans. 11 (1991) 3149.
[2] P. Maji, W. Wang, A.E. Greene, Y. Gimbert, J. Organomet. Chem. 693 (10) (2008)
1841.
[3] N.M.S. Ballester, M.R.J. Elsegood, M.B. Smith, G.M. Brown, Acta Crystallogr.
Sect. E Struct. Rep. Online 63 (2007) m719.
[4] C.Q. Zhao, M.C. Jennings, R. Puddephatt, J. Dalton Trans. 9 (2008) 1243.
[5] K.K.M. Hii, M. Thornton-Pett, A. Jutand, R.P. Tooze, Organometallics 18 (10)
(1999) 1887.
[6] K.D. Welch, W.G. Dougherty, W.S. Kassel, D.L. DuBois, R.M. Bullock,
Organometallics 29 (20) (2010) 4532.
[7] Q. Dong, M.J. Rose, W. Wong, H.B. Gray, Inorg. Chem. 50 (20) (2011) 10213.
[8] I.D. Kostas, J. Organomet. Chem. 626 (1–2) (2001) 221.
[9] D. Michos, X.L. Luo, R.H. Crabtree, J. Chem. Soc. Dalton Trans. 10 (1992) 1735.

S. Goren Keskin et al. / Polyhedron 138 (2017) 206–217
217
[10] I.D. Kostas, G. Antonopoulou, C. Potamitis, C.P. Raptopoulou, V. Psycharis, J.
Organomet. Chem. 828 (2017) 133.
[21] Y.L. Shi, Y.C. Gao, Q.Z. Shi, D.L. Kershner, F. Basolo, Organometallics 6 (7) (1987)
1528.
[11] S. Hinrichsen, A.-C. Schnoor, K. Grund, B. Flöser, A. Schlimm, C. Näther, J.
Krahmer, F. Tuczek, Dalton Trans. 45 (37) (2016) 14801.
[22] P.J. Blower, J.C. Jeffery, J.R. Miller, S.N. Salek, D. Schmaljohann, R.J. Smith, M.J.
Went, Inorg. Chem. 36 (8) (1997) 1578.
[12] O.P. Siclovan, R.J. Angelici, Inorg. Chem. 37 (3) (1998) 432.
[13] A. Hessler, S. Kucken, O. Stelzer, W.S. Sheldrick, J. Organomet. Chem. 553 (1–2)
(1998) 39.
[14] R. Sherwood, F.G. de Rivera, J.H. Wan, Q. Zhang, A.J.P. White, O. Rossell, G.
Hogarth, J. Wilton-Ely, Inorg. Chem. 54 (9) (2015) 4222.
[15] E.W. Ainscough, A.M. Brodie, P.D. Buckley, A.K. Burrell, S.M.F. Kennedy, J.M.
Waters, J. Chem. Soc. Dalton Trans. 16 (2000) 2663.
[16] G.J. Kubas, G.D. Jarvinen, R.R. Ryan, J. Am. Chem. Soc. 105 (7) (1983) 1883.
[17] R.G. Finke, G. Gaughan, C. Pierpont, J.H. Noordik, Organometallics 2 (10) (1983)
1481.
[23] Higashi, T., ABSCOR. Rigaku Corporation: Tokyo, Japan., 2001.
[24] Z. Otwinowski, W. Minor, DENZO-SMN. Processing of X-ray diffraction data
collected in oscillation mode, in: Carter, C.W.J., Sweets, R.M.(eds.), Methods in
Enzymology, vol. 276, Academic Press, , 1997, pp. 307–326.
[25] A. Altomare, M.C. Burla, M. Camalli, G.L. Cascarano, C. Giacovazzo, A.
Guagliardi, A.G.G. Moliterni, G. Polidori, R. Spagna, SIR97. A program for
crystal structure solution, J. Appl. Crystallogr. 32 (1999) 115.
[26] G.M. Sheldrick, SHELX 97, A software package for the solution and refinement
of X-ray data, University of Göttingen, Göttingen, Germany, 1997.
[27] D.T. Cromer, J.T. Waber, International Tables for X-Ray Crystallography,
Kynoch Press, Birmingham, 1992.
[18] E.O. Fischer, D. Wittmann, D. Himmelreich, R. Cai, K. Ackermann, D.
Neugebauer, Chem. Ber. Recl. 115 (9) (1982) 3152.
[19] H.A. Mayer, W.C. Kaska, Chem. Rev. 94 (5) (1994) 1239.
[20] B.E. Owens, R. Poli, A.L. Rheingold, Inorg. Chem. 28 (8) (1989) 1456.
[28] R.H. Crabtree, The Organometallic Chemistry of the Transition Metals, third
ed., John Wiley and Sons, USA, 2001, pp. 96–98.
[29] F.A. Cotton, W.G., Advanced Inorganic Chemistry, 5 ed., John Wiley and Sons,
USA, 1988, pp. 60.