Synthesis and characterization of terminal [Re(XCO)(CO) 2(triphos)] (X=N, P): Isocyanate versus phosphaethynolate complexes

Article abstract of DOI:10.1002/chem.201202590

The terminal rhenium(I) phosphaethynolate complex [Re(PCO)(CO) ₂(triphos)] has been prepared in a salt metathesis reaction from Na(OCP) and [Re(OTf)(CO)₂(triphos)]. The analogous isocyanato complex [Re(NCO)(CO)₂(triphos)] has been likewise prepared for comparison. The structure of both complexes was elucidated by X-ray diffraction studies. While the isocyanato complex is linear, the phosphaethynolate complex is strongly bent around the pnictogen center. Computations including natural bond orbital (NBO) theory, natural resonance theory (NRT), and natural population analysis (NPA) indicate that the isocyanato complex can be viewed as a classic Werner-type complex, that is, with an electrostatic interaction between the Re^I and the NCO group. The phosphaethynolate complex [Re(P=C=O)(CO)₂(triphos)] is best described as a metallaphosphaketene with a Re^I-phosphorus bond of highly covalent character. The first phosphaethynolate complex [Re(PCO)(CO)₂(triphos)] and the analogous isocyanate complex have been synthesized and characterized. While the isocyanato complex is linear, the phosphaethynolate complex is strongly bent around the pnictogen center (see scheme). Copyright

Full text of DOI:10.1002/chem.201202590

FULL PAPER
DOI: 10.1002/chem.201202590
Synthesis and Characterization of Terminal [ReACHTNUGTRENNUG(XCO)(CO)2ACHTUNGTRENN(UGN triphos)]
(X=N, P): Isocyanate versus Phosphaethynolate Complexes
Simone Alidori,[a] Dominikus Heift,[a] Gustavo Santiso-Quinones,[a] Zoltꢀn Benko˝,[a]
Hansjçrg Grꢁtzmacher,*[a] Maria Caporali,[b] Luca Gonsalvi,[b] Andrea Rossin,[b] and
Maurizio Peruzzini*[b]
Abstract: The terminal rhenium(I)
phosphaethynolate complex [Re-
(PCO)(CO)2A(triphos)] has been pre-
pared in a salt metathesis reaction
from Na(OCP) and [Re(OTf)(CO)2-
(triphos)]. The analogous isocyanato
complex [Re(NCO)(CO)2A(triphos)] has
been likewise prepared for comparison.
The structure of both complexes was
elucidated by X-ray diffraction studies.
While the isocyanato complex is linear,
the phosphaethynolate complex is
strongly bent around the pnictogen
center. Computations including natural
bond orbital (NBO) theory, natural
resonance theory (NRT), and natural
population analysis (NPA) indicate
that the isocyanato complex can be
viewed as a classic Werner-type com-
plex, that is, with an electrostatic inter-
action between the ReI and the NCO
group. The phosphaethynolate complex
G
CHTUNGTRENNUNG
E
ACHTUNGTRENNUNG
ACHTUNGTRENNUNG
E
G
[Re
ACHUTNGTRENN(GU P=C=O)(CO)2ACHUTNGTREN(NUGN triphos)] is best de-
Keywords: density functional calcu-
scribed as
a
metallaphosphaketene
lations
· isocyanato complexes ·
with a ReI–phosphorus bond of highly
covalent character.
phosphaketenes · rhenium · X-ray
diffraction
Introduction
pared by reaction of carbonyl complexes with an azide salt,
ꢀ
ꢀ
ꢀ
ꢀ
according to [M] CO+N3 ![M] NCO +N2 (M=W, Cr,
Mo),[7] or inversely by reaction of azide complexes with
CO.[8] We observed that Fe(CO)5 reacts with NaPH2 to form
phosphaethynolate, however, in noncoordinated form as Na-
Phosphaethynolate (PCOꢀ) is the phosphorus analogue of
cyanate (NCOꢀ) and has been previously characterized in
form of its lithium[1] and calcium salt.[2] Recently, we have
reported a very simple synthesis of the sodium salts [Na-
AHCTUNGTRENNUNG
(OCP), while the iron-containing product is Na2[Fe(CO)4].[3]
ꢀ
ꢀ
(OCP)
E
E
G
Further attempts to prepare [M] PCO or [M] OCP com-
plexes by reaction of carbonyl complexes with NaPH2 failed
so far, and transition-metal complexes of terminal PCOꢀ
anions remain unknown. The reaction of cyclopentadienylir-
preparation of larger quantities of pure compound.[3] This
prompted us to further investigate the properties of phos-
phaethynolate, especially with respect to the well-known cy-
anate ion. The last belongs to the class of the so-called
pseudo-halogens[4] and as ambivalent anion may in principle
bind to transition-metal centers either through the nitrogen
or the oxygen center. So far, however, various structurally
characterized complexes indicate that the nitrogen center
preferentially binds to the transition-metal center.[5,6] Most
often, the M-N-C angles are weakly bent and vary in the
range from 143 to 1808. Cyanate complexes are often pre-
on(II) bromide complexes with LiACHTUNTRGNEUNG(OCP) has been investi-
ꢀ
gated and may proceed through [Fe] PCO intermediates,
but only a dimer with a central 1,3-diphosphetane-2,4-dione
ring was isolated in low yield.[9] The stability and electronic
properties[10] of the PCOꢀ ion and some ion pairs with alkali
metals have been computed. In the course of our study, we
revisited these results with respect to the specific differences
between NCOꢀ and PCOꢀ in free and metal-coordinated
form. Furthermore, we report the first example of a terminal
phosphaethynolate complex that may be regarded as metal-
laphosphaketene.[11–13]
[a] Dr. S. Alidori, D. Heift, Dr. G. Santiso-Quinones, Dr. Z. Benko,
Prof. Dr. H. Grꢀtzmacher
Department of Chemistry and Applied Biology
ETH Zꢀrich, 8093 Zꢀrich (Switzerland)
Fax : (+41)44-633-10-32
Results and Discussion
E-mail: hansjoerg.gruetzmacher@inorg.chem.ethz.ch
We turned our attention to classic salt metathesis reactions
[b] Dr. M. Caporali, Dr. L. Gonsalvi, Dr. A. Rossin, Dr. M. Peruzzini
Istituto di Chimica dei Composti Organometallici
Consiglio Nazionale delle Ricerche (ICCOM-CNR)
Via Madonna del Piano 10, 50019 Sesto Fiorentino, Firenze (Italy)
Fax : (+39)055-5225-203
of the type MX+Na
A
ACHTUNGTREN(NUNG OCP)+NaX. Simple on
paper, these metathesis reactions turned out to be unexpect-
edly complex, and, in most cases, dark precipitates formed,
which were insoluble in all common solvents. So far, we did
not further characterize these precipitates and suspect that
redox reactions may be the reason for this behavior. The
E-mail: mperuzzini@iccom.cnr.it
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.201202590.
Chem. Eur. J. 2012, 18, 14805 – 14811
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
14805

easy oxidation of four equivalents of LiACTHNUTRGNEG(NU OCP) with even I2
or SO2 to form a heterocyclic dianion (P4C4O4)2ꢀ, which con-
tains four OCP units forming a bicyclic structure with a cen-
[14]
ꢀ
tral P P bond, has been reported and indicates that OCP
salts may be rather strong reductants.
Therefore, we investigated solutions of NaACTHNUTRGNEUG(N OCP) in
DMSO or THF by cyclic voltammetry (CV). In both sol-
vents, irreversible oxidations at rather low anodic potentials
were observed (Eox =ꢀ60 mV in DMSO and Eox =ꢀ308 mV
in THF versus the ferrocenium/ferrocene (Fc+/Fc) couple;
Figure 1). The oxidation product is insoluble in DMSO and
Scheme 1. Synthesis of [ReACHTNUTRGENNUG 2CAHTUNGTRNE(NUGN triphos)] (L=N NCO, P PCO).
(h1-L)(CO)
ꢀ
ꢀ
Figure 1. Cyclic voltammogram of NaACTHNUGTRNEUNG(OCP) (5 mm) in DMSO containing
0.1m tBu4NBF4 as electrolyte (glassy carbon as working electrode, a plati-
num wire as counter electrode, and AgjAgCljKClsat as reference elec-
trode). The given potentials are referenced versus the Fc+/Fc couple.
Figure 2. ORTEP plot of the structure of [ReACHTNUTRGNENUG(NCO)(CO)2ACHTUNGTREN(NNGU triphos)] (3).
Selected bond lengths and angles are given in Table 1. Thermal ellipsoids
are drawn at 30% probability.
likely precipitates on a glassy carbon surface, which is in
line with the reported low solubility of Li2ACHTNUTRGNE(UGN P4C4O4). In the
inverse cathodic scan, no increase in current was observed
up to a potential of about E=ꢀ2 V, indicating the high sta-
bility of the oxidation product against reduction.
Consequently, we searched for a suitable metal-complex
precursor that is sufficiently stable against reduction by Na-
ACHTUNGTRENNUNG(OCP). We identified [ReAHCTUNTREGNN(GUN OTf)(CO)2ACHTNUGERTN(NUGN triphos)] (1, triphos=
ꢀ
ꢀ
ꢀ
MeCACHTUNGTRENNUNG(CH2PPh2)3, OTf =F3C SO3 ) with its especially
stable low-spin d6 valence-electron configuration at the
metal center as such a candidate.[15] Indeed, 1 reacts smooth-
ly with both K
in THF to give the new metal complexes [Re
(triphos)] (3) and [Re(PCO)(CO)2A(triphos)] (5) in excellent
A
ACHTUNGTRENN(UNG OCP) (4)
AHCTUNGTRENNUNG
A
R
CHTUNGTRENNUNG
yields (Scheme 1). Both complexes were recrystallized and
their structures determined with single crystals by using X-
ray diffraction methods. Plots of the molecular structures of
3 and 5 are shown in Figures 2 and 3, respectively.
Figure 3. ORTEP plot of the structure of [ReACHTNUTRGNENUG(PCO)(CO)2ACHTUNGTREN(NNGU triphos)] (5).
Selected bond lengths and angles are given in Table 1. Thermal ellipsoids
are drawn at 30% probability.
Structural Studies
anion binds through the X center to the ReI in both com-
plexes 3 (X=N) and 5 (X=P). The tripodal phosphane
ligand triphos adopts a fac conformation with two P atoms
trans to two carbonyls (denoted P2 and P3), and the last P
In both complexes, the coordination sphere at the rhenium
center is (distorted) octahedral as expected. The XCOꢀ
14806
www.chemeurj.org
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 14805 – 14811

Terminal [ReACHTUNGTRENNUNG(XCO)(CO)2AHCTUNGTRENN(UGN triphos)] (X=N, P) Complexes
FULL PAPER
atom is in a trans position to the XCO ligand (denoted P1,
X=N, P).
indicating a stronger trans effect of the CO ligand compared
ꢀ
ꢀ
with that of XCO . In the calculated structures, the Re CO
Unfortunately, a crystallographic disorder phenomenon in
both rhenium complexes 3 and 5, by which the XCO group
changes place with the CO groups, strongly diminishes the
accuracy of the bond parameters and does not allow any
meaningful discussion. Therefore, we computed the model
distances are very similar in both the NCO and PCO com-
ꢁ
plex, which indicates similar p-back donation Re!C O in
the two complexes. In agreement, the CO stretching fre-
quencies in the IR spectrum are observed at almost identical
wavenumbers (n˜ =1946 and 1885 cmꢀ1 in 3 and n˜ =1948 and
1888 cmꢀ1 in 5, see below).
complexes [ReACHTUNGTRENNUNG
(XCO)(CO)2(triphosMe)] (X=N: 3calcd and
X=P: 5calcd) in which the phenyl groups at the phosphorus
atoms were replaced by methyl groups. The computations
reproduce well the geometries of the two complexes. Only
The most striking difference between the two complexes
is the cyanate/phosphaethynolate coordination mode: while
the NCOꢀ ligand binds in an approximately linear fashion
to the ReI center (measured Re-N-C angle: 166.28, calculat-
ed angle: 178.68), the PCOꢀ shows a strongly bent coordina-
tion mode with a Re-P-C angle of 92.68 (calculated value:
97.78). Consequently, in complex 3 a description of the NCO
in [ReACHTUNGTRENNUNG
(PCO)(CO)2(triphosMe)] (5calcd), the PCO moiety has
a different orientation than in the crystal structure, probably
because of steric reasons in the crystal packing (the PCO
ꢀ
fragment is about 908 rotated around the Re P axis com-
ꢀ
ꢁ
pared to the X-ray structure; see the Supporting Informa-
tion for a structure plot). We believe that this modification
still allows a discussion of the electronic structure of the co-
ordinated XCO groups. Selected calculated and measured
bond distances and angles are collected in Table 1 for com-
parison.
fragment with a O C single and a C N triple bond seems to
ꢁ ꢀ
be appropriate (N C O), while in complex 5 the PCO frag-
ment is better described with an allene-like P=C=O reso-
nance structure.
ꢀ
In the two complexes, the Re P bond trans to the XCO
ligand has almost the same length (2.350 and 2.347 ꢂ in
Vibrational Spectroscopy
3calcd and 5calcd, respectively, and 2.403 and 2.404 ꢂ in 3 and
5, respectively) and is significantly shorter than the Re P
distances trans to the carbonyl groups (2.437–2.450 ꢂ in the
calculated and 2.442–2.469 ꢂ in the measured structures),
To gain a deeper understanding of the bonding and elec-
tronic structure in these complexes, we analyzed the vibra-
tional frequencies of the XCO groups. Table 2 lists the ex-
perimental and computed infrared (IR) absorptions of the
coordinated and free XCOꢀ anions and the coordinated CO
ꢀ
Table 1. Calculated and experimentally (given in italics) obtained select-
ed bond distances [ꢂ] and bond angles [8] for the XCOꢀ anions and the
Table 2. Measured (n˜exp) and calculated (n˜calcd) IR stretching frequencies
[cmꢀ1] and calculated intensities I [km molꢀ1].
complexes [Re
(XCO)(CO)2A(triphos)] 3 and 5 (X=N, P; see Figures 2 and 3).
ACHTUNGTRENNUNG
(XCO)(CO)2(triphosMe)] 3calcd and 5calcd and [Re-
A
CHTUNGTRENNUNG
NCOꢀ
2214[a] 2235
3
PCOꢀ
1780, 1755[a] 1846
5
NCOꢀ
3calcd/3
PCOꢀ
5calcd/5
exp
asym
XCO
n˜
ꢀ
Re X
2.111
2.110(7)
1.188
1.066(7)
1.204
1.261(8)
2.350
2.403(2)
2.445
2.469(2)
2.444
2.442(2)
1.932
1.954(7)
1.932
1.977(5)
1.164
1.083(7)
1.164
1.106(7)
178.6
166.2(6)
178.4
2.596
2.561(7)
1.648
1.62(2)
1.182
1.23(2)
2.347
2.403(3)
2.437
2.435(3)
2.450
2.430(4)
1.936
2.12(1)
1.931
2.01(1)
1.162
0.86(2)
1.164
1.06(1)
97.7
93.7(7)
177.5
calcd [b]
asym
asymmetric vibration n˜
2139[c] 2280
1798[d]
1879
1.195[a]
1.229[a]
1.625[b]
1.203[b]
ꢀ
X C
I[b]
807
1515
–
1225
637
–
(X)C O
XCO
n˜
1213[a]
–
exp
[e]
[e]
[e]
ꢀ
asym
[c]
calcd [b]
asym
symmetric vibration
n˜
1216[c] 1326
787[d]
0.1
754
2.4
ꢀ
Re P1
I[b]
61
57
1946
1960
[d]
ꢀ
Re P2
[d]
exp
asym
ꢀ
Re P3
2 CO
n˜
n˜
1950
1966
calcd [b]
asym
ꢀ
Re C1(O1)
symmetric vibration
Re C2(O2)
I[b]
1146
1884
1027
1890
ꢀ
exp
2 CO,
n˜
asym
ꢀ
(Re)C1 O1
calcd [b]
asym
asymmetric vibration n˜
1915
1919
ꢀ
(Re)C2 O2
I[b]
998
1144
Re-X-C
X-C-O
[a] The experimental values for the sodium salts Na
are given. For Na(OCP), the data for two different crystalline compounds
[Na(OCP)(DME)2]2 and [Na(OCP)(dioxane)2.5
is given.[3] [b] Calculat-
ed for the [Re
(XCO)(CO)2(triphosMe)] compounds. [c] Previous compu-
ACHTUNGTREN(NGNU XCO) with X=N, P
AHCTUNGTRENNUNG
180.0
180.0
A
U
T
A
]
1
175.5(8)
174.1(2)
AHCTUNGTRENNUNG
tations give 2166 and 1228 cmꢀ1 for n˜asym and n˜sym, respectively.[16] [d] Pre-
vious computations at the CCSD(T) level give 1803.7 and 786.2 cmꢀ1 for
n˜asym and n˜sym, respectively. [e] The intensity was too low for an assign-
ment; see also ref. [3].
ꢀ
ꢀ
[a] Calculated values from ref. [16]: 1.193 ꢂ for N C and 1.229 ꢂ for C
ꢀ
O
bond. [b] Calculated values from ref. [17]: 1.639 ꢂ for P C and
1.204 ꢂ for C O bond. [c] In trans position to the XCO ligand. [d] In
trans position to the CO ligands.
ꢀ
Chem. Eur. J. 2012, 18, 14805 – 14811
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
14807

H. Grꢀtzmacher, M. Peruzzini et al.
groups of 3 and 5 compared to 3calcd and 5calcd by using DFT
and the PBE1 functional. Generally, the agreement between
the computed and measured data of the OCX and CO
groups in the Re complexes are satisfactory (a scaling factor
was used as proposed in the literature, see the computation-
al details). Since the stretching vibrations in the OCX
groups (n˜) are often rather strongly coupled, they are denot-
ed as n˜asym and n˜sym here. This is especially true in the free
1.182 ꢂꢀ1.203 ꢂ=ꢀ0.021 ꢂ); however, the C P bond is
ꢀ
significantly elongated in the complex when compared to
the free anion (Dd=1.648 ꢂꢀ1.625 ꢂ= +0.023 ꢂ). This ob-
servation further supports the idea that a change from a
ꢀ
(O C P) species to a phosphaketene-type unit (O=C=P)ꢀ
ꢀ ꢁ
occurs upon coordination to a metal center.
NCOꢀ anion and in complex 3calcd in which the N C and C
O vibrations are so strongly coupled that it does not allow
the separation of the two vibration modes. In the free PCOꢀ
anion and its ReI complex 5calcd, the asymmetric vibration
(n˜asym) at higher wavenumbers can be attributed to the CO
stretch, while the absorption at lower wavenumbers corre-
Natural Bond Orbital Calculations
ꢀ
ꢀ
We inspected the electronic structure of both the NCOꢀ and
PCOꢀ anions in the free and coordinated form in more
detail by using a natural resonance theory (NRT) analysis.
The two most important resonance structures of NCOꢀ and
PCOꢀ anions are shown in Figure 4.
ꢁ
sponds more to the C P stretching frequency (n˜sym), al-
though also here both modes show substantial coupling. Nu-
merically, our calculated data for the free PCOꢀ anion agree
well with the ones obtained by Hꢀbler and Schwerdtfeger at
the coupled cluster level (CCSD(T); n˜asym =1803.7 and n˜sym
=
786.2 cmꢀ1).[17] However, it is worth noting that these au-
thors assign the absorption at higher wavenumbers n˜asym, to
n˜Cꢁ and the lower wavenumber absorption to n˜CꢀO, which
P
we believe is erroneous and contradicts our assignment. The
experimental data were obtained for the sodium salts [Na-
Figure 4. The two most important resonance structures and their weights
A
E
(OCP)
U
(B3LYP/6-31+G*) for NCOꢀ and PCOꢀ.
calculated stretching frequencies for the free OCNꢀ anion
(n˜ =2139 cmꢀ1) is at substantially lower wavenumbers when
compared to the experimentally obtained data for the
The relative weights of the resonance structures are col-
lected in the table below the structures (PBE1/A level of
theory, some other resonance structures have also been
found with very low weights). Since the relative weights of
both resonance structures are remarkably large for both
anions, these anions can be regarded as bifunctional re-
agents: the attack of an electrophile can be expected either
on the element X (N or P) or on the oxygen atom. Howev-
er, it is important to highlight that the ratio of the relative
weights of the ketene-like resonance structure A and the
alkyne-like resonance structure B is larger in the case of the
P-analogue than for the N-analogue, suggesting the impor-
tance of the ketene-like resonance structure in the descrip-
tion of the electronic structure of PCOꢀ. The calculated nat-
ural population analysis (NPA) charges q also support the
ambivalent character of these anions (for NCOꢀ: q(N)=
ꢀ0.81 and q(O)=ꢀ0.75 e; for PCOꢀ: q(P)=ꢀ0.44 and
q(O)=ꢀ0.65 e, see Table 3). Remarkably, in the case of
NCOꢀ, the nitrogen is slightly more negatively charged than
the oxygen. On the contrary, for PCOꢀ the oxygen is, as ex-
sodium salt NaACHTUNGTRENNUNG
(OCN) (n˜ =2214 cmꢀ1); however, the calcu-
lated asymmetric vibration of the PCOꢀ anion is somewhat
larger than the measured data. These inconsistencies may be
attributed to crystal-packing effects.
A shift (Dn˜ =n˜complexꢀn˜anion) to higher wavenumbers is ob-
calcd
asym
calcd
frequencies when
sym
served for both calculated n˜
and n˜
the anion is bound in the molecular complex [Re-
ACHTUNGTRENNUNG
(NCO)(CO)2(triphosMe)]. For 3calcd, these shifts amount to
calcd
sym
Dn˜acsaylmcd =141 and Dn˜
=110 cmꢀ1. This is likewise reflected
in the experimental data: Dn˜aesxypm =21 cmꢀ1. The situation is
very different for the [Re
complex: while the v˜asym, which can be assigned more to the
(PCO)(CO)2(triphosMe)] (5calcd
ACHTUNGTERNNUNG )
calcd
asym
CO stretching, is shifted to higher wavenumbers (Dn˜
=
81 cmꢀ1 and Dn˜aesxypm =66 cmꢀ1), the symmetrical stretching
n˜sym (mainly the CP stretching) is shifted to lower wavenum-
bers (Dn˜acsaylmcd =ꢀ33 cmꢀ1; the data cannot be determined ex-
perimentally, because of the low intensity of this mode).
ꢁ
This indicates weakening of the C P bond and strengthen-
ing of the C=O bond upon formation of the ReI complex.
ꢀ
ꢀ
Furthermore, we analyzed the change in the O C and C
X bond distances when the small free anion OCXꢀ is trans-
ferred from the gas phase into a molecular compound like
Table 3. NPA partial charges q [atomic units] of XCOꢀ anions and [Re-
AHCTUNGTRENNUNG
(XCO)(CO)2(triphosMe)] complexes (3calcd: X=N, 5calcd: X=P).
[Re
ACHTUNGTRENNUNG
(XCO)(CO)2(triphosMe)] in the gas phase. In the case of
the formation of complex 3calcd, the computed O C bond
NCOꢀ
3calcd
PCOꢀ
5calcd
ꢀ
length decreases from 1.224 to 1.204 ꢂ (Dd=dcomplexꢀdanion
=
q(Re)
q(X)
q(C)
q(O)
–
ꢀ1.77
ꢀ0.64
+0.76
ꢀ0.65
ꢀ0.53
–
ꢀ2.11
+0.09
+0.24
ꢀ0.55
ꢀ0.22
ꢀ
ꢀ0.020 ꢂ), and the C N bond also contracts slightly by
ꢀ0.81
+0.56
ꢀ0.75
ꢀ1.00
ꢀ0.44
+0.09
ꢀ0.65
ꢀ1.00
0.007 ꢂ (in the complex 1.188 and in the anion 1.195 ꢂ). In
ꢀ
the case of the PCO-analogues, the O C bond is also short-
q(X)+q(C)+q(O)
er in the molecular complex than in the free anion (Dd=
14808
www.chemeurj.org
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 14805 – 14811

Terminal [ReACHTUNGTRENNUNG(XCO)(CO)2AHCTUNGTRENN(UGN triphos)] (X=N, P) Complexes
FULL PAPER
pected, more negatively charged (see above). Since both the
[Re(XCO)(CO)2A(triphos)] (kX) and [Re(OCX)(CO)2-
atom is now slightly positive (+0.09 e), while in the free
G
E
G
anion it is quite negative. Furthermore, the XCO fragment
(triphos)] (kO) compounds (X=N or P) can be expected
itself becomes almost neutral in the PCO-complex 5calcd
,
from the reaction of the rhenium-complex with XCOꢀ
anions, the relative energy of the two isomers was comput-
however, it remains partially negative (ꢀ0.53 e) in the
NCO-analogue 3calcd. This is in good agreement with the rel-
ꢀ
ed. For both anions, the [Re
N
ative contribution of the atomic hybrids in the Re X local-
pounds are remarkably more stable than the isomers [Re-
ized bonds: the contribution of the Re atomic hybrid in the
ꢀ
ꢀ
A
Re P bond is 39.4%, while it is 21.7% in the Re N bond.
19.6 kcalmolꢀ1 and for X=P the energy difference is
20.6 kcalmolꢀ1 at the PBE1/A level of theory), which is in
good agreement with the experimental findings.
This suggests that the Re P bond in the PCO-complex (X=
ꢀ
P, 5calcd) is rather covalent, while the Re N bond in the
ꢀ
NCO-complex (X=N, 3calcd) is much more ionic.
The similarity between the experimental and computed
structures shows that the bent coordination mode of the
PCO unit in the complexes is not the result of crystal-pack-
ing effects. Indeed, the calculated barrier to linearity is re-
Conclusion
markably high for [Re
A
We synthesized the first transition-metal complex bearing a
terminal phosphaethynolate (PCOꢀ) ligand. The analogous
isocyanato (NCO) complex has been prepared for compari-
son as well. Although both ligands coordinate through the
pnictogen center, the coordination behavior of the two li-
gands is very different: in the isocyanato complex, the coor-
dination sphere around the pnictogen atom is close to
linear, while the phosphaethynolate complex is strongly
bent. The analysis of the bonding parameters, vibration fre-
quencies, and the electronic structure of the complexes and
the free anions let us conclude that the phosphaethynolate
to 22.4 kcalmolꢀ1 at the PBE1/A level. On the contrary, this
barrier is very low (0.1 kcalmolꢀ1) for the cyanate complex
[ReACHTUNGTRENNUNG
(NCO)(CO)2(triphosMe)] (3calcd).
Natural bond orbital (NBO) analyses have been per-
formed to gain further insights in the differences of the elec-
tronic structure of NCO versus PCO complexes. This NBO
analysis suggests the two different Lewis structures for the
complexes 3calcd and 5calcd shown in Figure 5 with the best fit-
ting (least of largest deviations).
ꢀ
compound is best described as a metallaphosphaketene (M
P=C=O) with a highly covalent metal–phosphorus bond,
whereas the isocyanate analogue is a classic Werner-type
!
ꢀ
complex (M N+ C O ). Remarkably, the differences in
ꢁ ꢀ
ꢀ
the electronic structure are not reflected in the Re P bond
lengths to the triphos ligand nor in the stretching frequen-
cies of the carbonyl groups. Both the classic ionically bound
cyanate ligand NCOꢀ and the covalently bound phosphae-
thynolate PCOꢀ exert indistinguishable effects on the
Figure 5. The resonance structures suggested by the NBO analysis for
3calcd and 5calcd
.
[Re(CO)2ACTHNUGRTENUNG(triphos)] fragment.
For [Re
found between the N and C atoms. However, in the case of
[Re
(PCO)(CO)2(triphosMe)] (5calcd), a P=C double bond and
ACHTUNGTRENNUNG
(NCO)(CO)2(triphosMe)] (3calcd), a triple bond was
ACHTUNGTRENNUNG
Experimental Section
a phosphorus lone pair with high s character (68.0% at
PBE1/A) were obtained. This is in good agreement with the
well-known inert s-orbital effect.[18] Additionally, the find-
ings from the NBO analysis are in accordance with the NRT
analysis of the free anions XCOꢀ: the contribution of the
resonance structure B with one lone pair on the atom X
(see Figure 4) is much larger for NCOꢀ than for PCOꢀ, and
this strongly influences the coordination fashion of the
ligand.
The charge distribution of the complexes was also investi-
gated by using natural population analysis (NPA) charges,
and the results are summarized in Table 3. In both com-
plexes, the Re atom carries a negative partial charge
Materials and methods: Unless otherwise specified, all reactions were
performed by using the standard Schlenk procedures under a dry nitro-
gen or argon atmosphere. All solvents were purified by standard meth-
ods. The rhenium complex [Re(h1-OSO2CF3)(CO)
CHTUNGTREN(NUNG triphos)] [triphos=
2A
MeCAHCTNUGTERNNNUG
(CH2PPh2)3] was prepared as previously described.[15] All the other
reagents and chemicals were reagent grade and, unless otherwise stated,
used as received from commercial suppliers. Deuterated solvents (Al-
drich) were degassed by three freeze–pump–thaw cycles before use.
NMR spectra were recorded on a BRUKER AVANCE II 400 MHz spec-
trometer, operating at 400.13 MHz (1H), 100.61 MHz (13C), and
161.97 MHz (31P) and being equipped with a low-temperature measure-
ment device. 1H and 13C NMR chemical shifts are reported in parts per
million (ppm) downfield of tetramethylsilane (TMS) and were calibrated
against the residual protiated resonance (1H) or the deuterated solvent
multiplet (13C), while 31P{1H} NMR spectra were referenced to 85%
H3PO4 with downfield shift taken as positive. The computer simulation of
the second-order 13C NMR CO multiplets was carried out with Daisy. IR
spectra were collected on a Perkin–Elmer Spectrum BX II spectropho-
tometer as solution spectra in a Pike liquid cell that is equipped with
KBr windows. The IR spectra in solid state were measured using a
(ꢀ1.77 e for [Re
more negative, ꢀ2.11 e for [Re
(5calcd)). In the [Re(NCO)(CO)2(triphosMe)] (3calcd) complex,
ACHTUNGTRENNUNG
AHCTUNGTRENNUNG
ACHTUNGTRENNUNG
the partial charges within the NCO fragment are similar to
the free anion; however, for the phosphorus analogue the P
Chem. Eur. J. 2012, 18, 14805 – 14811
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
14809

H. Grꢀtzmacher, M. Peruzzini et al.
Golden gate unit under inert atmosphere or in KBr. Elemental analyses
(C, H, N) were performed by using a Carlo Erba model 1106 elemental
analyzer of the Microanalytical Service of the Department of Chemistry
at the University of Florence.
Computational details: Gas-phase geometry optimizations, molecular or-
bitals, and frequency calculations were performed on a model tripodal
”theoretical” triphos ligand, that is, MeCACHTUNRTGNE(UNG CH2PMe2)3 in which the phenyl
rings on the triphos P atoms were replaced by methyl groups (abbreviat-
ed as triphosMe), to have a reasonable compromise between accurate
system description and affordable computational time. The DFT func-
tional of choice was PBE1,[19] implemented within the Gaussian09 soft-
ware package;[20] inner core electrons of the rhenium atom were ex-
pressed through an SDD pseudopotential.[21] This specific functional and
pseudopotential combination was proven to perform very well in predict-
ing the geometries of third-row transition-metal-complexes.[22] An SDD
basis set with an additional f-polarization function[23] was also employed
Synthesis of [Re
(79.8 mg, 0.984 mmol, 10 equiv) in degassed water (0.5 mL) was added to
a solution of [Re(h1-O-OSO2CF3)(CO)
2A(triphos)] (100 mg, 0.0984 mmol,
1 equiv) in THF(20 mL). After stirring the reaction mixture for 24 h at
ACHTUNGTRENNUNG 2ACHTUNGTREN(NGUN triphos)] (3): A solution of KOCN
(h1-N-NCO)(CO)
CTHUNGTRENNUNG
ACHTUNGTRENNUNG
room temperature, a pale-yellow precipitate was formed. The yellow
powder was collected by filtration on a sintered glass-frit and washed
with diethyl ether (5 mL), before being dried by pumping for three hours
under vacuum. Concentration of the mother liquor gave another crop of
product (62.6 mg, 0.0688 mmol, 70%). Crystals of 3 suitable for an X-ray
diffraction analysis were obtained by slow evaporation of an acetone sol-
ution and diffusion of an ethanol/dichloromethane (1:1) solution.
1H NMR (CD2Cl2, 258C): d=7.1–7.9 (m, 30H; Ar-H), 2.6–2.5 (br s, 6H;
3CH2), 1.5 ppm (s, 3H; CH3); 13C{1H} NMR (CD2Cl2, 258C): d=197.1
for the metal center, while all the other atoms were given a 6-31+GACHTUNGTRENNUNG(d,p)
basis. This basis set combination is abbreviated as A. To check the relia-
bility of the calculations, they were also performed by using different
functionals (B3LYP, BP86) and a different basis set (cc-pVDZ); these
data are collected in the Supporting Information. For the vibrational fre-
quencies, a scaling factor was used.[24] The NBO analysis was performed
with the NBO 5.0 program.[25]
(m, AMM’X spin system, 2J
(A,X)=7.0 Hz, 2J
(M,M’)=28.5 Hz, CO), 194.5 (br s, NCO), 145.5–127.9
(Caromatic), 39.5 (q, 3J(C,P)=10.2 Hz, CH3) 39.1 (q, 2J
(C,P)=4.0 Hz,
CCH3) 36.2 (dt, 1J(CH2,PA)=26.5 Hz, 3J
(CH2,PM)=3.4 Hz, CH2Pax),
33.2 ppm (dd, N=J(CH2,PM)+J(CH2,PM’)=18.5 Hz, J(CH2,PA)=4.8 Hz,
A
N
A
ACHTUNGTRENNUNG
X-ray crystal structure of [Re
were collected on an Oxford Diffraction XCALIBUR 3 diffractometer,
equipped with CCD area detector by using MoKa radiation (l=
(h1-N-NCO)(CO)
ACHUTGTNRENUNG 2ACHTUNGTNER(NUGN triphos)]: X-ray data
A
ACHTUNGTRENNUNG
A
ACHTUNGTRENNUNG
a
A
R
ACHTUNGTRENNUNG
0.7107 ꢂ) at 150(2) K. The program used for the data collections was
CrysAlis CCD 1.171.[26] Data reductions were carried out with the pro-
gram CrysAlis RED 1.171,[27] and the absorption corrections were ap-
plied with the program ABSPACK 1.17.[26] Direct methods implemented
in SIR97[28] were used to solve the structures, and the refinements were
performed by full-matrix least squares against F2 implemented in
SHELX97.[29] The NCOꢀ ligand is partially disordered over the three
”stool” positions (there is scrambling with CO), but the disorder was not
explicitly treated during the refinement, because no significant improve-
ment of the final R1 value was observed.
CH2Peq); a simulated carbon spectrum can be found in the Supporting In-
formation; 31P{1H} NMR (CD2Cl2, 258C): AM2 spin system, d=1.1 (t, 2J-
A
ACHTNUGTNER(UNNG PM,PA)=15.8 Hz, 2PM); IR
(KBr): n˜ =2235 (vs, NCO), 1946 (vs, CO), 1885 cmꢀ1 (s, CO); IR (THF):
n˜ =2240 (s, NCO), 1956 (vs, CO), 1895 cmꢀ1 (s, CO); elemental analysis
calcd. (%) for C44H39NO3P3Re: C 58.14, H 4.32, N 1.54; found: C 58.01,
H 4.52, N 1.49.
Synthesis of [Re
OSO2CF3)(CO)2A(triphos)] (47 mg, 0.0463 mmol) in THF (5.0 mL) was
added dropwise to solution of Na(OCP)·2.5dioxane (14 mg,
ACHTUNGTRENNUNG 2ACHTUNGTRENNUGN
(h1-P-PCO)(CO) (triphos)] (5): A solution of [Re(h1-O-
CHTUNGTRENNUNG
a
ACHTUNGTRENNUNG
X-ray crystal structure of [ReACHTUNTRGENNUG 2ACHTUNGTERN(NUGN triphos)]: Pale-yellow
(h1-P-PCO)(CO)
0.0463 mmol) in THF (3.0 mL) by using a microsyringe. The reaction
mixture was stirred for 30 min at room temperature, then the solvent was
evaporated under reduced pressure. The orange-brown residue was dis-
solved in dichloromethane (15 mL) and filtered over Celite to eliminate
sodium triflate. Then, the solvent was removed under reduced pressure
and the orange powder dried for three hours under vacuum (30.0 mg,
0.0324 mmol, 70%). Crystals of 5 were obtained by slow evaporation of a
saturated THF solution under inert atmosphere. 1H NMR ([D8]THF,
258C): d=6.9–7.9 (m, 30H; Ar-H), 2.9–2.4 (br m, 6H; 3CH2), 1.4 ppm (s,
3H, CH3); 13C {1H} NMR ([D8]THF, 208C): d=198.0 (m, AMM’QX spin
rectangular-platelet crystals of 5 were obtained at room temperature by
slow evaporation of a saturated THF solution. X-ray data were collected
on a Bruker SMART Apex II diffractometer with a CCD area detector
by using MoKa radiation (l=0.7107 ꢂ) at 100(2) K. Empirical absorption
correction was performed with SADABS-2008/1 (Bruker) and refinement
against full matrix (versus F2) with SHELXTL (version 6.12) and
SHELXL-97. The structure was solved (SHELX 6.14 8/6/00) by direct
methods and successive interpretation of the difference of Fourier maps,
followed by least-square refinement. All non-hydrogen atoms were re-
fined anisotropically. The contribution of all the hydrogen atoms, in their
calculated positions, was included in the refinement by using a riding
model.
system, CO), 175.4 (dt, 1J
127.4 (Caromatic), 39.1 (q, 3J
10.2 Hz, CH3C), 34.5 (br d, 2J
AHCTUNGTRENNUNG
C
ACHTUNGTRENNUNG
A
ACHTUNGTRENNUNG
CCDC-875351 and 858778 contain the supplementary crystallographic
data for this paper. These data can be obtained free of charge from The
Cambridge Crystallographic Data Centre via www.ccdc.cam.ac.uk/data_
request/cif
system, d=ꢀ4.6 (dt, 2J
G
(A,M)=17.6 Hz, PA), ꢀ19.4
(dd, 2J(A,M)=17.6 Hz, 2J
ACHTUNGTRENNUNG ACHTUNGTRENNUNG
ACHTUNGTRENNUNG ACHTUNGTRENNUNG(M,Q)=14.3 Hz, PQ); IR (Golden gate): n˜ =1950 (vs,
(A,Q)=37.5 Hz; 2J
CO), 1890 (vs, CO), 1846 cmꢀ1 (s, PCO); IR (THF): n˜ =1960 (vs, CO),
1901 (vs), 1860 cmꢀ1 (s, P=C=O); elemental analysis calcd. (%) for
C44H39O3P4Re: C 57.08, H, 4.25; found: C 56.89, H 4.36.
Electrochemical study of Na
and general procedure: The cyclic voltammetry experiments were per-
formed on the ligand Na(OCP), dissolved in DMSO by using a solution
ACHTUNGTRENN(UNG OCP) by cyclic voltammetry—equipment
Acknowledgements
AHCTUNGTRENNUNG
5 mm with respect to the ligand and 0.1m with respect to tBu4NBF4,
which was employed as conductivity buffer. All experiments were per-
formed by means of a PARSTAT 2773 galvanostat/potentiostat (Prince-
ton Applied Research) that was equipped with Dr. Bobꢃs Cell (Gamry)
by using the classic three-electrode topology based on a AgjAgCljKClsat
(Gamry) reference electrode, 3 mm diameter glassy carbon as working
electrode, and platinum wire as counter electrode. All solutions were pre-
pared in a glove box under inert atmosphere, while the electrochemical
cell was purged with nitrogen prior to use and maintained under a posi-
tive nitrogen pressure during the measure. After measuring the ligand, a
2 mm solution of ferrocene in DMSO and 0.1m in tBu4NBF4 was meas-
ured to calibrate the potential (all potentials are referenced to the Fc/Fc+
redox couple).
This work was supported by the ETH Zꢀrich. Marie Curie Action FP7-
PEOPLE-2012-ITN-317404 ”SusPhos” and MIUR project PRIN 2009
are thanked for supporting this research activity. Thanks are also ex-
pressed to COST Action CM0802 (PhoSciNet) for granting a STSM for
S.A. to ICCOM-CNR. Dr. J. Filippi (ICCOM-CNR) is thanked for help
in the electrochemical measurements. The Sciex-NMSch fellowship (grant
number 10.256) is acknowledged for financial support (Z.B.).
[1] G. Becker, W. Schwarz, N. Seidler, M. Westerhausen, Z. Anorg.
Allg. Chem. 1992, 612, 72–82.
[2] M. Westerhausen, S. Schneiderbauer, H. Piotrowski, M. Suter, H.
Nçth, J. Organomet. Chem. 2002, 643–644, 189–193.
14810
www.chemeurj.org
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 14805 – 14811

Terminal [ReACHTUNGTRENNUNG(XCO)(CO)2AHCTUNGTRENN(UGN triphos)] (X=N, P) Complexes
FULL PAPER
[3] F. F. Puschmann, D. Stein, D. Heift, C. Hendriksen, Z. A. Gal, H.-F.
Grꢀtzmacher, H. Grꢀtzmacher, Angew. Chem. 2011, 123, 8570–
8574; Angew. Chem. Int. Ed. 2011, 50, 8420–8423. Note that in this
paper erroneously vibrations at n˜ =1262 (1247) cmꢀ1 were assigned
[19] C. Adamo, V. Barone, J. Chem. Phys. 1999, 110, 6158–6170.
[20] Gaussian 09, Revision A.02, M. J. Frisch, G. W. Trucks, H. B. Schle-
gel, G. E. Scuseria, M. A. Robb, J. R. Cheeseman, G. Scalmani, V.
Barone, B. Mennucci, G. A. Petersson, H. Nakatsuji, M. Caricato, X.
Li, H. P. Hratchian, A. F. Izmaylov, J. Bloino, G. Zheng, J. L. Son-
nenberg, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa,
M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, T. Vreven,
J. A. Montgomery, Jr., J. E. Peralta, F. Ogliaro, M. Bearpark, J. J.
Heyd, E. Brothers, K. N. Kudin, V. N. Staroverov, R. Kobayashi, J.
Normand, K. Raghavachari, A. Rendell, J. C. Burant, S. S. Iyengar,
J. Tomasi, M. Cossi, N. Rega, J. M. Millam, M. Klene, J. E. Knox,
J. B. Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R. E.
Stratmann, O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. W.
Ochterski, R. L. Martin, K. Morokuma, V. G. Zakrzewski, G. A.
Voth, P. Salvador, J. J. Dannenberg, S. Dapprich, A. D. Daniels, ꢅ.
Farkas, J. B. Foresman, J. V. Ortiz, J. Cioslowski, and D. J. Fox,
Gaussian, Inc., Wallingford CT, 2009.
ꢀ
to the C O stretching frequencies, which are in fact too low in in-
tensity to be observed experimentally.
[4] PATAI’S Chemistry of Functional Groups: The Chemistry of Halides,
Pseudo-Halides and Azides (Eds.: S. Patai, Z. Rappoport), Parts 1
and 2, Wiley, 1983.
[5] U. Ray, D. Banerjee, B. G. Chand, J. Cheng, T.-H. Lu, C. Sinha, J.
Coord. Chem. 2005, 58, 1105–1113.
[6] J. L. Zuffa, W. L. Gladfelter, J. Am. Chem. Soc. 1986, 108, 4669–
4671.
[7] W. Beck, J. Organomet. Chem. 1990, 383, 143–160.
[8] W. Beck, W. P. Fehlhammer, P. Pollmann, H. Schachl, Chem. Ber.
1969, 102, 1976–1987.
[9] L. Weber, B. Torwiehe, G. Bassmann, H.-G. Stammler, B. Neumann,
Organometallics 1996, 15, 128–132.
[10] a) P. Pyykkç, Y. Zhao, Mol. Phys. 1990, 70, 701–714; b) T. Veszprꢄ-
[21] M. Dolg, U. Wedig, H. Stoll, H. Preuss, J. Chem. Phys. 1987, 86,
866–872.
mi, T. Pasinszki, M. Fehꢄr, Inorg. Chem. 1996, 35, 2132–2135.
[22] M. Bꢀhl, C. Reimann, D. A. Pantazis, T. Bredow, F. Neese, J. Chem.
Theory Comput. 2008, 4, 1449–1459.
[23] A. W. Ehlers, M. Bçhme, S. Dapprich, A. Gobbi, A. Hçllwarth, V.
Jonas, K. F. Kçhler, R. Stegmann, A. Veldkamp, G. Frenking, Chem.
Phys. Lett. 1993, 208, 111–114.
ꢀ
[11] For tBu P=C=O, see: R. Appel, W. Paulen, Tetrahedron Lett. 1983,
24, 2639–2642.
ꢀ
[12] For (2,4,6-tBu3C6H2) P=C=O, see: R. Appel, W. Paulen, Angew.
Chem. 1983, 95, 807–808; Angew. Chem. Int. Ed. Engl. 1983, 22,
785.
[24] Y. Tantirungrotechai, K. Phanasant, S. Roddecha, P. Surawatana-
wong, V. Sutthikhum, J. Limtrakul, J. Mol. Struct. THEOCHEM
2006, 760, 189–192.
[25] NBO115.0, E. D. Glendening, J. K. Badenhoop, A. E. Reed, J. E.
Carpenter, J. A. Bohmann, C. M. Morales, F. Weinhold, Theoretical
Chemistry Institute, University of Wisconsin, Madison, 2001.
[26] CrysAlisCCD 1.171.31.2 (release 07-07-2006), CrysAlis171.NET,
Oxford Diffraction Ltd.
[27] CrysAlis RED 1.171.31.2 (release 07-07-2006), CrysAlis171.NET,
Oxford Diffraction Ltd.
[28] A. Altomare, M. C. Burla, M. Camalli, G. L. Cascarano, C. Giaco-
vazzo, A. Guagliardi, A. G. G. Moliterni, G. Polidori, R. Spagna, J.
Appl. Crystallogr. 1999, 32, 115–119.
ꢀ
[13] For a recent report on metal coordinated Me P=C=O, see: Y. Mao,
Z. Wang, R. Ganguly, F. Mathey, Organometallics 2012, 31, 4786–
4790.
[14] G. Becker, G. Heckmann, K. Hubler, W. Schwarz, Z. anorg. allg.
Chem. 1995, 621, 34–46.
[15] a) C. Bianchini, A. Marchi, L. Marvelli, M. Peruzzini, A. Romerosa,
R. Rossi, A. Vacca, Organometallics 1995, 14, 3203–3215; b) P. Ber-
gamini, F. F. DeBiani, L. Marvelli, N. Mascellani, M. Peruzzini, R.
Rossi, P. Zanello, New J. Chem. 1999, 23, 207–217.
[16] C. Lꢄonard, H. Gritli, G. Chambaud, J. Chem. Phys. 2010, 133,
124318.
[17] K. Hꢀbler, P. Schwerdtfeger, Inorg. Chem. 1999, 38, 157–164. The
+
ꢀ
species Li
(OCP) with the Li -ion binding side on to the C O unit
[29] G. M. Sheldrick, SHELXL-97, Universitꢆt Gçttingen, Gçttingen,
1997.
was likewise computed with n˜asym =1720.7 and n˜sym =824.6 cmꢀ1
.
[18] For the original paper on the inert s-pair effect, see: N. V. Sidgwick,
Ann. Reports 1933, 20, 120; for an application to the bonding in
main-group-element compounds, see: W. Kutzelnigg, Angew. Chem.
1984, 96, 262–286; Angew. Chem. Int. Ed. Engl. 1984, 23, 272–295.
Received: July 23, 2012
Published online: September 28, 2012
Chem. Eur. J. 2012, 18, 14805 – 14811
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
14811