Terminal acetylenes react to increase unsaturation in [(tBu 2PCH2SiMe2)2N]Re(H)4

Article abstract of DOI:10.1021/om049424i

(PNP^tBu)Re(H)₄, where PNP^tBu is ( ^tBu₂PCH₂SiMe₂)₂N, reacts at 23 °C with RC≡CH (R = ^tBu, SiMe₃, Ph) to give first H₂ and mirror-symmet

Full text of DOI:10.1021/om049424i

4934
Organometallics 2004, 23, 4934-4943
Ter m in a l Acetylen es Rea ct to In cr ea se Un sa tu r a tion in
[(tBu 2P CH2SiMe2)2N]Re(H)4
Oleg V. Ozerov, Lori A. Watson, Maren Pink, Mu-Hyun Baik, and
Kenneth G. Caulton*
Department of Chemistry, Indiana University, Bloomington, Indiana 47405
Received J uly 26, 2004
(PNPtBu)Re(H)4, where PNPtBu is (tBu2PCH2SiMe2)2N, reacts at 23 °C with RCtCH (R )
tBu, SiMe3, Ph) to give first H2 and mirror-symmetric (PNPtBu)ReH3(CCR), then H2 and C2ν
symmetric (PNPtBu)Re(CCR)2. The diacetylide compounds show temperature-independent
paramagnetism and 13C and 31P chemical shifts far beyond their normal values for other
(PNPtBu)ReXn compounds. Single-crystal X-ray diffraction shows very similar structures for
the cases R ) Ph and R ) SiMe3, each having an approximately C2v geometry with equivalent
acetylides with C-Re-C approximately 108°. No hydride or H2 ligands are detected in
final difference Fourier maps. DFT(B3PW91) calculations give minimum energy geometries
of these species, of their products upon adding H2, and of mechanistically significant
analogues [(H2PCH2SiH2)2N]ReHnR′mH2-m, with n ) 0, 2, m ) 1, 2, and R′ ) H or Ph. These
calculated geometries, when compared to those from X-ray diffraction, indicate that the
isolated compounds have no hydride or H2 ligands and are thus (PNP)ReIII(CCR)2, making
them more unsaturated than the reagent (PNP)ReV(H)4 by two electrons. Triplet state
geometries of (PNP)ReXY are calculated and analyzed, as are their frontier orbitals.
In tr od u ction
RCtCH, which might have been envisioned as candi-
dates for forming a π-acidic vinylidene ligand, RedCd
CHR, but instead form acetylide complexes by loss of
H2.
The molecule (PNPR)Re(H)4 (PNPR ) (R2PCH2-
SiMe2)2N) is a classical hydride of ReV, having a π-donor
ligand, the amide, which stabilizes the otherwise un-
saturated (16 valence electron) metal.1 We have termed
this “operational unsaturation” since the Re/N bond is
short enough to indicate multiple bonding, yet this
molecule forms a 1:1 adduct with Lewis bases, e.g.,
PMe3. Because an associative mechanism is thus avail-
able, (PNPR)Re(H)4 reacts rapidly at and below 23 °C
with ethylene.1 As H ligands are lost in this reaction,
the metal is reduced and its reducing power (π-basicity)
increases. As a result, ethylene is isomerized into a more
π-acidic ligand, ethylidyne (eq 1). With cyclic olefins,2
Resu lts
Syn th esis a n d Ch a r a cter iza tion . Terminal acety-
t
lenes RCCH (R ) Ph, SiMe3, and Bu) react rapidly with
(PNPtBu)Re(H)4 in benzene or pentane at 23 °C with
immediate gas evolution. 1H NMR monitoring of the
reactions identifies the gas as H2 and shows no signifi-
cant amount of the corresponding olefins among the
products. The synthetic reaction is thus an acid/base
reaction, with H2 elimination (eq 3). At an alkyne:Re
(PNPtBu)Re(H)4 + 2RCCH h
(PNPtBu)Re(CCR)2 + 3H2 (3)
mole ratio of 3.5:1, the product solution contains two
new complexes, blue (PNPtBu)ReH3(CCR) and (PNPtBu)-
Re(CCR)2. At lower mole ratios, the product yields
change consistent with (PNPtBu)ReH3(CCR) being an
intermediate. Following the time evolution of a reaction
with >3:1 alkyne:Re ratio in a closed system shows
gradual conversion of (PNPtBu)ReH3(CCR) to (PNPtBu)-
Re(CCR)2. The initial (kinetic) product of the reaction
of RCCH with (PNPtBu)Re(H)4 in benzene or pentane
at 23 °C is (PNPtBu)ReH3(CCR). In the presence of more
than 1 equiv of RCCH and with removal of H2, (PNPtBu)-
ReH3(CCR) is subsequently transformed into (PNPtBu)-
Re(CCR)2. However, even without the extra RCCH
present, (PNPtBu)ReH3(CCR) partially disproportionates
into (PNPtBu)Re(CCR)2 and (PNPtBu)Re(H)4. It was
since the conversion of the unsaturated hydrocarbon
cannot effect C-C bond scission (requisite for carbyne
formation), the reaction stops at the carbene stage. In
one instance, the isomerized olefin, cyclohexene, yields
a rare case of a â-agostic carbene (eq 2).
We report here on the distinctly different character
of the reaction of (PNPtBu)Re(H)4 with terminal alkynes,
* To whom correspondence should be addressed. E-mail: caulton@
indiana.edu.
(1) Ozerov, O. V.; Huffman, J . C.; Watson, L. A.; Caulton, K. G.
Organometallics 2003, 22, 2539.
(2) Ozerov, O. V.; Watson, L. A.; Pink, M.; Caulton, K. G. J . Am.
Chem. Soc. 2003, 125, 9604.
10.1021/om049424i CCC: $27.50 © 2004 American Chemical Society
Publication on Web 09/14/2004

Unsaturation in [(tBu2PCH2SiMe2)2N]Re(H)4
Organometallics, Vol. 23, No. 21, 2004 4935
independently established that Re-C bond formation
is reversible by exposing (PNPtBu)Re(CCPh)2 to D2 at
23 °C. This showed conversion to (PNPtBu)ReD3(CCPh),
then (PNPtBu)Re(D)4, together with PhCCD.
(PNPtBu)ReH3(CCR) showed a hydride 1H NMR signal
of intensity 3 H (a triplet from coupling to two P) and
equivalent phosphorus nuclei. Two PtBu signals, two
SiMe2 signals, and two SiCH2 signals indicate that the
hydride fluxionality does not create a time-averaged
mirror plane containing the ReNP2 plane; the molecule
has only the mirror symmetry defined by the NReCC
plane.
F igu r e 1. ORTEP drawing (50% probability ellipsoids) of
(PNPtBu)Re(CCSiMe3)2, showing selected atom labeling.
Crystallographic symmetry element is C2. Omitted for
(PNPtBu)Re(CCR)2 showed unusual NMR spectro-
scopic behavior (Table 1). While the 1H and 31P{1H}
NMR spectra were consistent with C2ν symmetry, the
1H chemical shifts (wholly in the 0-10 ppm region) were
significantly (as much as 5 ppm) shifted from their
normal1,2 regions. In addition, the 31P signals, while
sharp (half-width ≈ 0.2 ppm at 162 MHz and 23 °C),
were shifted to -250 to -400 ppm, while for (PNPtBu)-
ReH3(CCR), and indeed many (PNP)ReXn analogues,1,2
the 31P chemical shifts range from 0 to +85 ppm. These
31P chemical shifts are, moreover, unusually tempera-
ture dependent (∼1 ppm/°C). The 13C chemical shifts
are also found over a huge range, from -200 to +375
ppm. In contrast, the 1H NMR of (PNPtBu)Re(CCR)2
spectra are not abnormally temperature dependent
t
clarity: all H and the Bu methyl groups. Selected struc-
tural parameters (Å, deg): Re-C12, 1.980(2); Re-N1,
2.031(2); Re-P1, 2.4392(5); C12-C13, 1.224(3); C12-Re-
C12′, 108.75(12); N1-Re-P1′, 85.790(13); P1-Re-P1′,
171.58(3); Re-C12-C13, 176.6(2); N1-Re-C12, 125.62-
(6); C12-Re-P1′, 95.11(6); C12′-Re-P1′, 89.79(6).
1
(changing 0.1 ppm in 18 °C). Both H and 13C lines are
sharp and show coupling to 31P.
The earliest observations of strongly temperature-
dependent chemical shifts were in 59Co NMR, where
changes of 1-3.8 ppm/°C were observed.3 These were
attributed to TIP (better called “second-order paramag-
F igu r e 2. ORTEP drawing (50% probability ellipsoids) of
(PNPtBu)Re(CCPh)2 showing selected atom labeling. Omit-
t
ted for clarity: all H and the Bu methyl groups. Selected
1
netism”), where the authentically singlet A1g ground
structural parameters (Å, deg): Re-C23, 1.973(2); Re-
C31, 1.973(2); Re-N1, 2.0279(18); Re-P1, 2.4369(6); Re-
P2, 2.4431(7); C23-C24, 1.228(3); C31-C32, 1.232(4);
C23-Re-C31, 108.01(9); N1-Re-P1, 85.28(9); N1-Re-
P2, 85.81(9); P1-Re-P2, 170.86(2); Re-C23-C24, 174.6-
(2); Re-C31-C32, 176.7(2); N1-Re-C23, 129.15(2), N1-
Re-C31, 122.82(15); C23-Re-P1, 88.33(7); C31-Re-P1,
95.93(7), C23-Re-P2, 95.75(7); C31-Re-P2, 90.60(7).
1
state of Co(III)L6 was contaminated by mixing in a T1g
excited state. The temperature dependence of δ was due
to temperature-dependent vibrational averaging of this
excited state. This orbital angular momentum made a
paramagnetic contribution to the magnetic susceptibil-
ity ø, despite the purity of spin ) zero character.
A select set of ReIII and ReIV compounds have a long
1
history of unusual NMR behavior: large H chemical
shifts and no 31P NMR signal observable.4-9 However,
the compounds ReCl3(PR2Ph)3 and ReCl4(PRPh2)2 were
never subjected to a careful study of their solid state
magnetic susceptibility. We therefore collected SQUID
susceptibility data on both species (PNPtBu)Re(CCR)2 (R
) Ph and SiMe3). The measured susceptibilities of both
(PNPtBu)Re(CCR)2 are small down to at least 100 K and
relatively temperature independent. The plots (see
Experimental Section) show that the magnitude of the
susceptibility (after correcting for the diamagnetic
contribution, by using Pascal’s constants) is small
compared to the spin-only formula for even one unpaired
electron (øm‚T ) S(S + 1)/2 ) 0.375 cm3 mol-1 for S )
1/2). In addition, the observed temperature dependence
for two different acetylides can be fit over the full
temperature range (see øm versus T plots) by assuming
1
<0.8% paramagnetic (S ) /2) impurity, and then the
resulting temperature-independent susceptibility is small
(consistent with the conclusion above). Indeed, it is on
the order of the diamagnetic corrections. The molecules
(PNPtBu)Re(CCR)2 thus show very small and tempera-
ture-independent susceptibilities, and the unusual NMR
behavior of these diamagnetic molecules cannot be
attributed to unpaired electrons.
The X-ray structure determination (Figures 1 and 2)
shows substantial agreement between the structures of
the CCPh and CCSiMe3 analogues. Bond lengths Re-
N, Re-P, Re-C, CtC, and NRe-C show no chemi-
cally significant differences between the two. The silyl
example has crystallographic C2 symmetry (the Re-N
bond is the C2 axis), while the phenyl case has only
idealized 2-fold symmetry. In neither case were hydro-
gens detected attached to Re. Negative evidence for
hydrides by X-ray diffraction cannot be taken as certain,
despite our consistent ability to find and refine hydrides
using the SMART 6000 diffractometer. NMR cannot be
relied on for the decision here, due to the observation
of the unusual magnetism and NMR chemical shifts.
(3) Gielen, M., Willem, R., Wrackmeyer, B., Eds. Advanced Applica-
tions of NMR to Organometallic Chemistry; Wiley: New York, 1996.
(4) Chatt, J .; Leigh, G. J .; Mingos, D. M. P.; Paske, R. J . J . Chem.
Soc. A 1968, 2636.
(5) Leigh, G. J .; Gunz, H. P. J . Chem. Soc. A 1971, 2229.
(6) Shaw, D.; Randall, E. W. Chem. Commun. 1965, 82.
(7) Randall, E. W.; Shaw, D. Mol. Phys. 1965, 10, 41.
(8) Chatt, J .; Leigh, G. J .; Mingos, D. M. P.; Randall, E. W.; Shaw,
D. Chem. Commun. 1968, 419.
(9) Randall, E. W.; Shaw, D. J . Chem. Soc. A 1969, 2867.

4936 Organometallics, Vol. 23, No. 21, 2004
Ozerov et al.
Ta ble 1. Tem p er a tu r e Dep en d en ce of 31P NMR
Ch em ica l Sh ifts
temp, °C
(PNPtBu)Re(CtCSiMe3)2
(PNPtBu)Re(CtCPh)2
-40
0
10
20
21.7
30
40
50
60
-277.7
-253.6
-262.2
-271.6
-272.6
-281.1
-290.8
-300.9
-311.3
-340.4
-354.4
-355.6
-369.5
-384.7
-400.2
-416.1
Multiple efforts at growing the larger crystals required
for a neutron diffraction structure determination were
unsuccessful. Thus, comparison of the experimental
non-hydrogen atom structure to structures of various
stoichiometries determined by DFT(B3PW91) geometry
optimization might be useful. This use of DFT geom-
etries for “quantitative and qualitative analysis” has
been recently demonstrated for the FeMoCo enzyme
center,10-13 and our approach is in the same spirit.
DFT Geom etr y Optim ization s for Sin glet [N(SiH2-
CH2P H2)2]ReHn RmH2-m. Minimum DFT(B3PW91) en-
ergy structures were determined for singlet spin states
of the title species with R ) H, CtCH, and Ph, and with
m ) 0, 1, 2 and n ) 0, 2. The last variation is intended
to probe whether the presence or absence of two H
ligands has a detectable influence on the non-hydrogen
atom skeleton; this is our intended method of learning
whether the X-ray structure can indirectly tell whether
these molecules have H ligands. The study of three R
groups is intended to probe how robust our conclusions
are to variation among three different one-electron
donor ligands.
F igu r e 3. SOMO orbital contour diagrams and energies
for phenyl, methyl (middle), and ethynyl (lower) radicals
(DFT (B3LYP/6-31G**)/unrestricted) at the fixed geometry
of their H-R analogues. Contours are drawn at 0.025 au
intervals, and the outmost contour is 0.025, as indicated.
All three plots are drawn to scale.
1.46-1.53 Å. That is, H and phenyl ligands make the
metal more reducing than acetylide. We suggest that
this distinction arises because acetylide is less of a
σ-donor to the metal than either H or Ph; superficially,
acetylide is more of a pseudohalide than it is a pure
σ-donor like H or phenyl. The orbital explanation is that
the larger s component of the sp hybrid orbital in
acetylide compared to the sp2 and sp3 orbitals of the
phenyl and methyl groups gives rise to a notably lower
orbital energy for the acetylide than the other two
ligands. Therefore, the acetylide is a weaker σ-donor
than the phenyl group. Thus, the main distinction
between acetylide and methyl or phenyl as a ligand can
be traced to the energy of the σ-orbital that will form
the M-C bond, and these are shown in Figure 3.
Illustrated are the singly occupied orbitals of the radical
species C6H5, CH3, and CCH, but calculated at the
frozen geometry of the corresponding molecular species
H-R. The species CCH has a σ-orbital with an energy
of -10.72 eV, which is notably lower than the corre-
sponding orbitals of CH3 or C6H5, at -7.02 and -6.90
eV, respectively. The relatively low orbital energy of the
acetylide increases the orbital energy difference from
the σ-accepting d-orbital of the metal center, with the
consequence that there will be much less mixing of d
with the CCH σ-orbital. All three orbitals are drawn to
scale in Figure 3, and the larger s component of the
acetylide can be readily recognized by comparing the
shapes of the orbitals. An additional contributing factor
favoring an intact H2 co-ligand may be that acetylide is
more π-acidic than the other R groups studied.
The starting geometries for DFT energy minimization
for n ) 2 in every case (X, Y ) R, H) involved
nonbonding H/H distances (i.e., >1.3 Å) and all ligands
in the same plane, I, that are perpendicular to the PNP
plane. As will be seen, this involves considerable
“migration” of the 2 H (i.e., to the PNP plane), to get to
the most stable geometry, and it is unbiased with regard
to hydride versus H2 final geometry.
The results in Table 2 show the following.
(a) Re-P distances are invariant ((0.01 Å) as are the
acetylide CtC distances ((0.01 Å).
(b) The ground state geometries for n ) 2 have a
dihydrogen (intact H2) ligand for one or two acetylide
ligands (m ) 1, 2), but are dihydride complexes (ReV)
for one or two phenyl ligands, and also for R ) H. The
difference is an H/H distance of 0.90-0.99 Å versus
(10) Huniar, U.; Ahlrichs, R.; Coucouvanis, D. J . Am. Chem. Soc.
2004, 126, 2588.
(11) Hinnemann, B.; Norskov, J . K. J . Am. Chem. Soc. 2003, 125,
1466.
(12) Dance, I. Chem. Commun. 2003, 324.
(13) Lovell, T.; Liu, T.; Case, D. A.; Noodleman, L. J . Am. Chem.
Soc. 2003, 125, 8377.
(c) Regardless of dihydride or dihydrogen structures,
the case n ) 2 has an angle between the two R ligands

Unsaturation in [(tBu2PCH2SiMe2)2N]Re(H)4
Organometallics, Vol. 23, No. 21, 2004 4937
Ta ble 2. Str u ctu r a l P a r a m eter s for (P NP H)ReHn Rm H2-m fr om DF T Op tim iza tion
Ta ble 3. Com p a r ison of Ca lcu la ted (DF T) a n d
X-r a y-Deter m in ed Str u ctu r a l P a r a m eter s for
(P NP )ReHn (CCR)2
(b) Re-P lengthens by 0.04 Å and Re-N lengthens
by 0.07 Å in the triplet, suggesting that the LUMO of
the singlet is antibonding in both Re-P and Re-N
bonds. Figure 5 confirms this.
(c) The C-Re-C angle is 120° in the singlet but only
a very small 94.5° in the triplet. This is consistent with
avoidance by carbon of the singly occupied Nπ-dπ orbital
in the triplet. The triplet state is calculated to lie 13.3
kcal/mol higher than the singlet, which is consistent
with the measured magnetic susceptibilities.
F in a l Con clu sion on H Liga n d s. The data collected
in Table 3 compare the DFT-calculated structural
parameters for n ) 0 (singlet and triplet) and for n ) 2
to the X-ray values. We choose to put the greatest
reliance on the angular location of the acetylide CR (i.e.,
C-Re-C), and by this criterion the composition with
no H on Re (i.e., consistent with the absence of such
electron density in the final X-ray Fourier maps) is
clearly preferred.
Oth er Tr ip let Sta tes. To understand the influence
of the anionic ligands X and Y on the singlet/triplet
energy gap among (PNPH)ReXY species, we have cal-
culated minimum triplet energy structures for (PNPH)-
Re(H)2 and (PNPH)ReH(CCH), to compare to (PNPH)-
Re(CCH)2. These three are compared in Figure 4,
together with their energies relative to their singlet
states in their minimum energy geometries. In every
case, the singlet is more stable, and the singlet-triplet
energy gap shows no consistent trend. In each case, the
triplet state has the longer Re-N distance and the
smaller X-Re-Y, the latter generally near 90°. For
(PNPH)Re(H)(CCH), the structure begins to resemble a
larger by 15-30° than for n ) 0. Predictably, the smaller
of these two numbers is for dihydride and the larger is
for dihydrogen, because the “bulk” of H2 lies closer to
the C-Re-C plane.
(d) Going from n ) 0 to n ) 2, the Re-N distance is
longer by 0.07 Å for dihydride ground states, but it
lengthens by less than 0.03 Å for a dihydrogen ground
state.
Exp lor in g th e P a r a m a gn etic Alter n a tive. We also
calculated [(H2PSiH2CH2)2N]Re(CCH)2 in its lowest
energy triplet state. Geometries of singlet and triplet
are compared in Table 3 and Figure 4. These show the
following.
(a) Re-C and CtC distances are nearly invariant to
spin state.
F igu r e 4. DFT minimum energy geometries for triplet states of species shown, together with their energy relative to the
corresponding singlet.

4938 Organometallics, Vol. 23, No. 21, 2004
Ozerov et al.
This general answer can be supplemented by one that
is specific to an ReX3(PR3)2 case. We begin by noting
that Re(Ph)3(PEt2Ph)2 is an authentic 14 valence elec-
tron molecule and is diamagnetic.22,23 To see how the
change from one phenyl to one amide can influence the
ground spin state, analysis of the relevant d-orbital
splitting diagram is helpful. Contrasted below are the
d-orbital splitting for D3h RePh3(PR3)2 and C2ν (PNP)-
Re(CCR)2, both in the same axis system. The key
difference is that symmetry reduction to C2v and the
change to more halide-like ligands split the two d-orbital
degeneracies of D3h, and with more independent energy
levels, a small energy gap ∆E becomes possible, when
the electronic character of the X ligand is right. The
triplet state can then become energetically competitive.
The amide nitrogen lone pair (illustrated) does not have
the symmetry to mix with any of the three lowest
frontier orbitals depicted for (PNP)Re(CCR)2, and so xy,
which is π*ReN, remains empty; N f Re π donation is
thus unimpaired, to first order, by going from singlet
to triplet states.
F igu r e 5. Orbital isosurface plots of the two singly
occupied MOs of triplet (PNPH)Re(CCH)2.
Rea ction En er gies (DF T) for P oten tia l Rea ction
In ter m ed ia tes. The species described above, together
with additional ones, were located at DFT-optimized
geometries to establish energetically viable intermedi-
ates in the synthesis of (PNPtBu)Re(CCR)2. Since these
are all energy minima, this view of the potential energy
surface alone cannot fully define the mechanism; transi-
tion states were not calculated, in part because of the
large number of minima considered. Nevertheless, spe-
cies whose energy is very high make a mechanism
passing through that intermediate too slow; the DFT
energies can thus help to exclude certain mechanisms.
In addition, a high reaction energy is informative since
it indicates some destabilizing features of that product.
As such, this permits better understanding of the
bonding potential of the [(R2PCH2SiMe2)2N]Re substruc-
ture: its reducing power, the influence of changing
metal oxidation state, trans effect, and the character
(bonding or antibonding) of certain frontier orbitals.
(a ) Acetylid es. Scheme 1 shows reaction energies
(∆E + ZPE differs only insignificantly from enthalpies)
square pyramid, where N-Re-C is 154.2°. In sum,
all these ReIII species are predicted to be more stable
as singlets and with a near 90° X-Re-Y angle.
Discu ssion
The synthetic reaction employed gives a product
whose valence electron count has decreased by two. The
reaction transforms ReV to ReIII, a d4 species. Why is
the triplet state only 13 kcal/mol above the singlet? One
possible answer is that, even if the coordinated amide
nitrogen involves its lone pair in π donation to unsatur-
ated Re, the result is a 16-electron species. Numerous
examples have shown14-18 that, for the early and middle
transition metals, it can occasionally be preferable to
“save” the spin pairing energy and half fill two frontier
orbitals rather than doubly occupy one and leave a small
HOMO/LUMO gap (the 18e rule generally promises
a small gap for the singlet state of a 16-electron
species).19-21
(17) Poli, R. Comments Inorg. Chem. 1992, 12, 285.
(18) Abugideiri, F.; Keogh, D. W.; Poli, R. J . Chem. Soc., Chem.
Commun. 1994, 2317.
(19) Hessen, B.; Teuben, J . H.; Lemmen, T. H.; Huffman, J . C.;
Caulton, K. G. Organometallics 1985, 4, 946.
(20) Hessen, B.; Lemmen, T. H.; Luttikhedde, H. J . G.; Teuben, J .
H.; Petersen, J . L.; Huffman, J . C.; J agner, S.; Caulton, K. G.
Organometallics 1987, 6, 2354.
(21) Hessen, B.; Meetsma, A.; Teuben, J . H. J . Am. Chem. Soc. 1988,
110, 4860.
(22) Carroll, W. E.; Bau, R. J . Chem. Soc., Chem. Commun. 1978,
825.
(14) Harvey, J . N.; Poli, R.; Smith, K. M. Coord. Chem. Rev. 2003,
238-239, 347.
(15) Smith, K. M.; Poli, R.; Harvey, J . N. New J . Chem. 2000, 24,
77.
(23) Chatt, J .; Garforth, J . D.; Rowe, G. A. J . Chem. Soc. A 1966,
1834.
(16) Poli, R. Acc. Chem. Res. 1997, 30, 494.

Unsaturation in [(tBu2PCH2SiMe2)2N]Re(H)4
Organometallics, Vol. 23, No. 21, 2004 4939
Sch em e 1a
a
∆(E + ZPE), kcal/mol.
for various (PNPH)ReHn(CCH)1 or 2 species, representing
the addition of one or two acetylenes and release of
various amounts of H2. An initial loss of H2, when
compensated by T‚∆S° of about 8 kcal/mol,24,25 charac-
terizes one possible first step, generating the exception-
ally reactive (PNP)Re(H)2 (note reaction energy for eq
f). However, combining these two into a displacement
of H2 by HCCH gives better energetics (eq g versus a +
f). A second possible product of attack of HCCH on intact
(PNP)Re(H)4 is h, with a very favorable reaction energy
to give the η1-vinyl complex. H2 is only weakly bound
in this species (BDE 7.0 kcal/mol, which will be com-
pensated by 8 kcal/mol T‚∆S), so one rapidly forms the
R-agostic monohydride (eq i), which is only slightly more
stable than (PNP)Re(H)2(HCCH). The energies of two
isomers of this η2-HCCH species, a structural isomer
(eq j, 28.5 kcal/mol higher) and a redox isomer (eq k,
13.1 kcal/mol higher), show that (PNP)Re(H)2(HCCH)
lies in a remarkably deep energy well, such that it might
reach an experimentally detectable concentration if it
were kinetically accessible. Its stability is also reflected
in the unreasonably large (versus experiment) positive
reaction energy, +35.1 kcal/mol, to lose H2 and form an
acetylide, eq l. Thus, an attractive alternative mecha-
nism would be oxidative addition of the C-H bond of
HCCH to transient (PNP)Re(H)2 (reaction b, energy
-26.6 kcal/mol), followed by loss of H2 (eq c, +10.4 kcal/
mol) to give (PNP)ReH(CCH). Oxidative addition of the
C-H of a second molecule of acetylene to this inter-
mediate is favorable (eq d, -15.0 kcal/mol), and the
resulting dihydrogen (H2) complex easily (eq e, -0.6
kcal/mol) loses H2 to give the observed product. The
energy of this last reaction clearly confirms our ag-
gregate conclusion from above that the isolated reaction
product from HCCPh or HCCSiMe3 contains no H
ligands on Re.
It also bears mention that any reaction could be made
47.1 kcal/mol more favorable (E + ZPE) by the addition
of hydrogenation of 1 mol of HCCH. Either this could
benefit a reaction shown as “-H2”, or it could consume
2H from an ReH2 substructure currently shown in
Scheme 1 as persisting throughout the reaction. It
should also be noted that the use of H in the DFT
calculations fails to model steric effects. Therefore,
energies of associative and of dissociative steps involve
errors, but of predictable character.
(b) P h en yls. For comparison of H, C6H5, and CCH
as substituents on such a conversion of (PNP)Re(H)4 to
(PNP)ReHnRmH2-m, Scheme 2 shows transformations
related to Scheme 1. Analogous reactions are shown
with the same letter label plus a prime symbol. All of
the species not shown in bold in Scheme 1 are of course
absent in Scheme 2. The reaction energies in Scheme 2
show that Re-H and Ph-H are thermodynamically
better than Re-Ph and H-H to a similar extent for Re-
(H)4 (eq x′, +9.7 kcal/mol) and Re(H)3Ph (eq y′, +10.5
kcal/mol) as test cases; there is thus no major conse-
quence for σ-bond metathesis thermodynamics of phenyl
versus H as “spectator” substituent on Re. On the other
hand (eq a′ versus eq c′ or e′) H2 binds much stronger
to Re(H)2 than to Re(H)Ph or Re(Ph)2. The higher
reactivity of H-CCH compared to H-Ph is also evident
in the more favorable ∆(E + ZPE) for reactions with
Re(H)2 (b versus b′) and ReH(Ph) (d versus d′). A
compensating reflection of the reactivity of Re(H)2
versus Re(H)(Ph) is that oxidative addition of benzene
to the former (eq b′, -12.2 kcal/mol) is much more
favorable than to the latter (eq d′, -0.4 kcal/mol). As a
result, while an associative path for eq x′ is preferred
for H-Ph oxidative addition to Re(H)4, it is not (eq y′
versus c′) for benzene oxidative addition to Re(H)3Ph.
(24) Watson, L. A.; Eisenstein, O. J . Chem. Educ. 2001, 79, 1269.
(25) Minas de Piedade, M. E.; Martinho Simoes, J . A. J . Organomet.
Chem. 1996, 518, 167.

4940 Organometallics, Vol. 23, No. 21, 2004
Ozerov et al.
Sch em e 2a
a
∆(E + ZPE), kcal/mol.
Finally (eq e′), binding of H2 to (PNP)Re(Ph)2 (1.5 kcal/
mol) is insignificantly strong, just as it was to (PNP)-
Re(H)2(CCH)2. Regarding structure, neither the (PNP)-
ReH(Ph) nor the (PNP)Re(Ph)2 species is ortho-agostic
to the unsaturated metal.
Mech a n istic Sp ecu la tion . The low activation en-
ergy suggested by the rapid reaction of (PNPtBu)Re(H)4
with RCCH at 23 °C rules out preliminary H2 loss from
the Re reagent (a in Scheme 1). The large steric bulk of
the (PNPtBu) ligand probably favors products (and
intermediates) of smaller steric profile. This may con-
tribute to the preference for the observed σ-alkynyl
products and intermediates as opposed to the η2-alkyne
complexes because the reagent complex is very crowded.
We therefore favor an acid/base reaction (eq 4), which
immediately evolves H2. The intermediates (PNPtBu)-
mentally (via reaction with H2 or D2) that hydrogen
converts (PNPtBu)Re(CCR)2 back to (PNPtBu)ReH3(CCR),
then to (PNPtBu)Re(H)4, making vacuum removal of H2
an effective way to drive the reaction fully to (PNPtBu)-
Re(CCR)2. During this (slow) hydrogenolysis reaction,
there is no evidence (>1%) of H2 coordinating to, or even
broadening the NMR signals of, (PNPtBu)Re(CCR)2.
Con clu sion s. The reason for the title conclusion is
the weak binding of H2 to either of the (PNP)ReR2
species discussed here (R ) CCR′ or Ph). This contrasts
with the case of R ) H (Scheme 1) and illustrates the
fundamental reluctance of unsaturated polyhydrides to
form, and thus the high reactivity, once formed, of
unsaturated polyhydrides (e.g., (PNP)Re(H)2). Finally,
Scheme 1 shows clearly that the most thermodynami-
cally stable product of a 1:1 reaction of (PNP)Re(H)4 and
HCCH is (PNP)ReH(tC-CH3) (cf. eq 5), and thus the
observed product is the result of kinetic control. In fact,
the two unimolecular H migration reactions shown in
the box in Scheme 1, have both been shown28,29 to have
high barriers, and these are likely the steps that effect
kinetic control in the experiments reported here.
(PNPtBu)Re(H)4 + RCCH* f
(PNPtBu)Re(H)3(CCR) + H-H* (4)
t
Re(H)3(CCR) were detected for R ) Ph, SiMe3, and -
Bu, and these diamagnetic species have had their
hydride content quantitated by1H NMR integration and
quartet splitting in the selectively hydride-coupled 31P
NMR spectrum. If indeed the acidity of RCCH is the
origin of the first mechanistic step, it is clear why the
reaction does not go in the direction of vinyl, vinylidene,
or carbyne products shown in Scheme 1. In this regard,
olefins convert (PNPtBu)Re(H)4 to a hydrido carbyne
species.1 The mild conditions employed also enable
kinetic control; the migration of hydrogen from M or
from CR in η1-vinyl complexes has been generally
found26,27 to involve a high (>30 kcal/mol) barrier.
Experimentally, when this acid/base mechanism is
eliminated, so too is the rapid reaction: benzene does
not compete with RCCH in reaction with (PNPtBu)Re-
(H)4 because the acid/base mechanism is not relevant
for arenes. Exchange of (PNPtBu)Re(H)4 hydrides with
C6D6 (neat) is not detectable over 3 h at 23 °C, and no
phenyl complexes accumulate, consistent with the reac-
tion energies in Scheme 2.
Integrating the several observations and the DFT
computations on (PNPtBu)Re(CCR)2 species, these ap-
pear to have essentially diamagnetic ground states but
with very minor mixing of an excited state into the
ground state wave function, together with vibrational
averaging, which leads to the anomalous NMR behav-
ior: unusually large 13C and 31P chemical shifts (these
nuclei being closer to the metal center) that show large
temperature dependence (31P), concurrent with rela-
tively ordinary and temperature-insensitive 1H NMR
chemical shifts. It remains to be seen whether Re(III)
and Os(IV) are the only metals in which these effects
manifest themselves, or whether selected d4 species of
other metals will also show these phenomena. It is
worth noting that the cases reported here have a
different coordination number than the previous ReCl3L3
and OsCl4L2 examples, so octahedral geometry is not
demanded.
It is also relevant that we observe at most 10% of
hydrogenated alkyne as product, and we do observe
either free H2 or products that put H on Re. The steps
in Scheme 1 involving H2 evolution are thus the most
relevant. At the same time, we demonstrated experi-
(26) Ipaktschi, J .; Mohsseni-ala, J .; Uhlig, S. Eur. J . Chem. 2003,
4313.
(27) Perez-Carreno, E.; Paoli, P.; Ienco, A.; Mealli, C. Eur. J . Chem.
1999, 1315.
(28) Dolker, N.; Frenking, G. J . Organomet. Chem. 2001, 617-618,
225.
(29) Stegmann, R.; Frenking, G. Organometallics 1998, 17, 2089.

Unsaturation in [(tBu2PCH2SiMe2)2N]Re(H)4
Organometallics, Vol. 23, No. 21, 2004 4941
F igu r e 6. Plots of magnetic susceptibility of (PNPtBu)Re(CCR)2 for R ) Ph (above) and R ) SiMe3 (below).
2000 (300 MHz 1H; 121 MHz 31P; 75 MHz 13C), a Varian Unity
The products with unusual magnetic and especially
1
Inova instrument (400 MHz H; 162 MHz 31P; 101 MHz 13C),
NMR behavior here are those with low coordination
1
number, 5, in (PNPtBu)Re(CCR)2. In general, low coor-
or a Varian Unity Inova instrument (500 MHz H; 126 MHz
13C). Elemental analyses were performed by CALI, Inc. (Par-
dination number causes less splitting of the five d-
sippany, NJ ). ORTEP plots were generated using Ortep-3 for
orbitals, and thus other spin states can become com-
Windows.30
petitive with the singlet. It can sometimes become
preferable to half fill orbitals when the sub-18 electron
Ma gn etic Mea su r em en ts. Magnetic measurements were
performed on a Quantum Design MPMS-XL SQUID magne-
count would leave an orbital completely empty, thus
apparently contributing nothing to species stability. It
is also relevant that the unusual magnetism is observed
with the acetylide ligand, which our analysis has shown
is the weakest sigma donor, the most halide-like (cf.
“weak-field” ligands). The unusal NMR phenomena
represent additional evidence that the product is Re-
(III) and thus argue against the presence of undetected
hydride ligands.
tometer equipped with a 70 kG (7 T) magnet. Direct current
susceptibility data were collected on powdered microcrystalline
samples contained in evacuated, sealed NMR tubes, using a
field of 1 kG (0.1 T) in the 300-306 K temperature range.
Pascal’s constants were used to estimate the diamagnetic
correction for the complex, which was then applied to the
experimental susceptibility to give the molar paramagnetic
susceptibility (øM), Figure 6. The different susceptibility be-
havior for different R below 10 K suggests that this is not an
intrinsic property of (PNPtBu)Re(CCR)2, but instead an impurity-
derived phenomenon.
Attempts to fit the temperature dependence of the 31P
chemical shifts show (Supporting Information) that a δ versus
T fit is modestly better than a δ versus 1/T plot. We also
considered a singlet ground state (chemical shift δS) in thermal
equilibrium with a triplet state (δT) lying higher by ∆G°. The
average chemical shift of such a system is given by
Exp er im en ta l Section
Gen er a l Con sid er a tion s. All manipulations were per-
formed using standard Schlenk techniques or in an argon-filled
glovebox unless otherwise noted. Pentane, n-hexane, C6D6, and
C7D8 were vacuum transferred from NaK/benzophenone/18-
crown-6. Alkynes were stirred over and then vacuum trans-
ferred from CaH2. (PNPtBu)Re(H)4 was prepared as reported
elsewhere.1 1H NMR chemical shifts are reported in ppm
relative to protio impurities in the deuterated solvents. 31P
spectra are referenced to external standards of 85% H3PO4 (at
0 ppm). NMR spectra were recorded with a Varian Gemini
δS
1 + e-∆G°/RT 1 + e-∆G°/RT
We modeled the temperature dependence of δ (31P) for different
δT
δav
)
+

4942 Organometallics, Vol. 23, No. 21, 2004
Ozerov et al.
assumed values of δT, the (unknown) chemical shift of the
(unobservably broad) triplet species: -102, -103, and -104 ppm
(see Supporting Information). For a ∆G° value equal to the
singlet/triplet difference calculated by DFT, the variation of δ
is extremely small, even for δT ) -104. For a smaller ∆G°, δ
changes more, but the population of the triplet becomes large
enough that varying line broadening should be seen in the
averaged signal, contradicting experiment.
NMR Mon itor in g of Rea ction s of (P NP tBu )Re(H)4 w ith
Ter m in a l Alk yn es. Gen er a l. The reactions of (PNPtBu)Re-
(H)4 with 1-10 equiv of RCtCH produced mixtures of
(PNPtBu)Re(H)4, (PNPtBu)ReH3(CtCR), (PNPtBu)Re(CtCR)2,
RCtCH, and H2, whose time evolution is described below.
These mixtures generally had a navy blue color of (PNPtBu)-
ReH3(CtCR). Selectively hydride-coupled 31P NMR spectra
yielded a quartet, indicating the presence of three H ligands
on Re. Initially, (PNPtBu)ReH3(CtCR) was the major compo-
nent, but over time (at 22 °C in a closed J . Young NMR tube)
the fraction of (PNPtBu)ReH3(CtCR) decreased. In all cases a
small (<10%) amount of RCHdCH2 was also observed.
(a ) (P NP tBu )Re(H)4 + Me3SiCtCH. (PNPtBu)Re(H)4 (24.0
mg, 37 µmol) was mixed with Me3SiCtCH (21.1 µL, 148 µmol)
in 0.6 mL of C6D6. Integration of the 1H NMR spectrum (20
min after mixing at 22 °C) revealed that the ratio between
(PNPtBu)Re(CtCSiMe3)2, (PNPtBu)ReH3(CtCSiMe3), and
(PNPtBu)Re(H)4 was 10:90:0 (plus excess free Me3SiCtCH).
After 24 h at 22 °C in a closed J . Young tube the ratio changed
to 32:68:0. After 48 h, the ratio was 40:60:0. At this time, the
solution was degassed by two freeze-pump-thaw cycles. Ten
minutes after degassing, the ratio was 77:23:0. Another 24 h
later the ratio changed to 87:13:0. The NMR tube was then
back-filled with 1 atm of H2 to test for reversibility of the
reaction; 10 min later the ratio was 69:31:0.
(P NP tBu )Re(CtCP h )2. (PNPtBu)Re(H)4 (125 mg, 193 µmol)
was dissolved in 5 mL of pentane, and PhCtCH (85 µL, 772
µmol) was added to it. This caused rapid gas evolution and
change of color to deep blue. The mixture was stirred while
being periodically exposed to vacuum for 10-15 s periods, until
most of the volatiles were removed. Then another 5 mL of
pentane was added and the slow removal of volatiles was
repeated. The purplish brown residue was dissolved in a
minimal amount of pentane and placed into a -30 °C freezer
overnight. The next day the brownish crystalline product was
separated by decantation, washed with cold pentane, and dried
in vacuo. Yield: 115 mg (70%). The supernatant contained only
the product (besides solvent and traces of phenylacetylene).
1H NMR (C7D8, 22 °C): δ 8.64 (t, 8 Hz, 4H, m-Ph), 2.91 (d,
8 Hz, 4H, o-Ph), 2.31 (t, 8 Hz, 2H, p-Ph), 2.16 (t, J HP ) 5 Hz,
36H, PC(CH3)3), 2.01 (br s, 4H, PCH2Si), 0.23 (s, 12H, SiCH3).
31P{1H} NMR (C7D8, 21.7 °C): δ -272.6 (s). 13C{1H} NMR
(C7D8, 22 °C): δ 374.6, 200.1, 147.5, 116.2, 57.4 (t, 7 Hz), 44.2,
41.5, 32.8, 15.1, -140.8 (m). Anal. (C24H51N2P2ReSi2) Calcd
(Found): C, 54.51 (54.37); H, 7.46 (7.65); N, 1.67 (1.66).
1H NMR (C7D8, +40 °C): δ 8.75 (t, 8 Hz, 4H, m-Ph), 2.60
(d, 8 Hz, 4H, o-Ph), 2.03 (t, 8 Hz, 2H, p-Ph), 2.20 (t, J HP ) 6
Hz, 36H, PC(CH3)3), 2.08 (br s, 4H, PCH2Si), 0.21 (s, 12H,
SiCH3).
Da ta for (P NP tBu )ReH3(CtCSiMe3). 1H NMR (C6D6, 22
°C): δ 1.49 (t, J HP ) 6 Hz, 18H, PC(CH3)3), 1.20 (t, J HP ) 6
Hz, 18H, PC(CH3)3), 1.05 (br dt, 2H, PCH2Si), 0.80 (br dt, 2H,
PCH2Si), 0.35 (s, 6H, SiCH3 of PNP), 0.32 (s, 9H, CtCSi-
(CH3)3), 0.25 (s, 6H, SiCH3 of PNP), -6.33 (br, 3H, ReH3). 31P-
{1H} NMR (C6D6, 22 °C): δ 72.8 (s).
(b) (P NP tBu )Re(H)4 + P h CtCH. (PNPtBu)Re(H)4 (24.0 mg,
37 µmol) was mixed with PhCtCH (4.1 µL, 37 µmol) in 0.6
mL of C6D6. Integration of the 1H NMR spectrum (10 min after
mixing at 22 °C) revealed that the ratio between (PNPtBu)Re-
(CtCPh)2, (PNPtBu)ReH3(CtCPh), and (PNPtBu)Re(H)4 was
4:89:7 (plus ca. 10% free PhCtCH). After 18 h at 22 °C in a
closed J . Young tube the ratio changed to 25:37:38, while free
PhCtCH was completely consumed.
(P NP tBu )Re(CtCTol)2. (PNPtBu)Re(CtCTol)2 (Tol ) para-
tolyl ) C6H4Me-p) was prepared analogously from (PNPtBu)-
Re(H)4 (130 mg, 204 µmol) and TolCtCH (70 mg, 600 µmol).
Yield: 130 mg (74%).
1H NMR (C6D6, 22 °C): δ 8.56 (d, 8 Hz, 4H, m-Tol), 5.85 (s,
6H, Ar-CH3), 2.92 (d, 8 Hz, 4H, o-Tol), 2.23 (t, J HP ) 5 Hz,
36H, PC(CH3)3), 2.06 (br s, 4H, PCH2Si), 0.24 (s, 12H, SiCH3).
31P{1H} NMR (C7D8, 25 °C): δ -282.4 (s).
1
Da ta for (P NP tBu )ReH3(CtCP h ). H NMR (C6D6, 22 °C):
(P NP tBu )Re(CtCSiMe3)2. (PNPtBu)Re(H)4 (0.50 g, 0.77
mmol) was dissolved in 5 mL of pentane, and Me3SiCtCH
(0.71 mL, 5.0 mmol) was added to it. This caused rapid gas
evolution and change of color to deep blue. The mixture was
stirred for 2 h while being periodically exposed to vacuum for
ca. 1-2 s periods, followed by complete removal of volatiles in
vacuo. The residue was treated with Me3SiCtCH (0.2 mL, 1.4
mmol) and 3 mL of pentane to complete the conversion of
hydrides to (PNPtBu)Re(CCSiMe3)2. A homogeneous purple
solution formed and was stirred for 30 min. Then the volatiles
were removed in vacuo, and the residue was dissolved in a
minimal amount of n-hexane and filtered through a pad of
Celite to remove a minute amount of insolubles. Crystallization
from n-hexane at -30 °C produced highly crystalline material
(0.46 g), which was washed with cold n-hexane and dried in
vacuo. Similar recrystallization from the supernatant gave
another 0.11 g of the product. Total yield: 0.57 g (89%). The
product is extremely lipophilic. In order for the crystallization
to succeed it was found imperative that Me3SiCtCH be freshly
vacuum transferred from CaH2 and (PNPtBu)Re(H)4 be freshly
recrystallized.
δ 7.35 (d, 8 Hz, 2H, o-Ph), 7.18 (t, 8 Hz, 2H, m-Ph), 6.86 (t, 8
Hz, 1H, p-Ph), 1.46 (t, J HP ) 6 Hz, 18H, PC(CH3)3), 1.22 (t,
J HP ) 6 Hz, 18H, PC(CH3)3), 1.08 (br dt, 2H, PCH2Si), 0.83
(br dt, 2H, PCH2Si), 0.39 (s, 6H, SiCH3) 0.28 (s, 6H, SiCH3),
-6.10 (t, J HP ) 24 Hz, 3H, ReH3). 31P{1H} NMR (C6D6, 22 °C):
δ 72.1 (s).
(c) (P NP tBu )Re(H)4 + TolCtCH. (PNPtBu)Re(H)4 (13.0 mg,
20 µmol) was mixed with TolCtCH (5.1 µL, 40 µmol) in 0.6
mL of C6D6. Integration of the 1H NMR spectrum (30 min after
mixing at 22 °C) revealed that the ratio between (PNPtBu)Re-
(CtCTol)2, (PNPtBu)ReH3(CtCTol), and (PNPtBu)Re(H)4 was
22:71:7 (plus some free TolCtCH).
Da ta for (P NP tBu )ReH3(CtCTol). 1H NMR (C6D6, 22
°C): δ 7.32 (d, 8 Hz, 2H, m-Tol), 7.02 (d, 8 Hz, 2H, o-Tol), 1.92
(s, 3H, Ar-CH3), 1.49 (t, J HP ) 6 Hz, 18H, PC(CH3)3), 1.24 (t,
J HP ) 6 Hz, 18H, PC(CH3)3), 1.10 (br dt, 2H, PCH2Si), 0.86
(br dt, 2H, PCH2Si), 0.40 (s, 6H, SiCH3) 0.28 (s, 6H, SiCH3),
-6.16 (t, J HP ) 24 Hz, 3H, ReH3). 31P{1H} NMR (C6D6, 22 °C):
δ 71.3 (s).
(d ) (P NP tBu )Re(H)4 + tBu CtCH. (PNPtBu)Re(H)4 (24.0 mg,
t
1H NMR (C7D8, 22 °C): δ 2.36 (t, J HP ) 6 Hz, 36H,
PC(CH3)3), 2.16 (br t, J HP ) 4 Hz, 4H, PCH2Si), 0.56 (s, 18H,
Si(CH3)3), 0.14 (s, 12H, Si(CH3)2 of PNP). 31P{1H} NMR (C7D8,
21.7 °C): δ -355.6 (s). 13C{1H} NMR (C7D8, 22 °C): δ 377.1
(s), 62.6 (t, J ) 7 Hz), 49.4 (s), 39.7 (s), 34.4 (s), 18.3 (s), -201.6
(m). 1H NMR (C7D8, -40 °C): δ 2.19 (br s, 36H, PC(CH3)3),
1.91 (br s, 4H, PCH2Si), 0.50 (s, 18H, Si(CH3)3), 0.22 (s, 12H,
Si(CH3)2 of PNP).
37 µmol) was mixed with BuCtCH (18.3 µL, 148 µmol) in 0.6
mL of C6D6. The mixture was shaken well for 5 min, and then
the solution was degassed by two freeze-pump-thaw cycles.
1
Integration of the H NMR spectrum of the resulting mixture
revealed that the ratio between (PNPtBu)Re(CtCBut)2, (PNPtBu)-
ReH3(CtCBut), and (PNPtBu)Re(H)4 was 26:52:22 (plus excess
t
free BuCtCH).
Da ta for (P NP tBu )ReH3(CtCBu t). 1H NMR (C6D6, 22
°C): δ 1.49 (t, J HP ) 6 Hz, 18H, PC(CH3)3), 1.23 (t, J HP ) 6
Hz, 18H, PC(CH3)3), 0.96 (br dt, 2H, PCH2Si), 0.86 (br dt, 2H,
(30) Faruguia, L. J . J . Appl. Crystallogr. 1997, 30, 565.

Unsaturation in [(tBu2PCH2SiMe2)2N]Re(H)4
Organometallics, Vol. 23, No. 21, 2004 4943
PCH2Si), 0.35 (s, 6H, SiCH3) 0.23 (s, 6H, SiCH3), -6.52 (br,
3H, ReH3). 31P{1H} NMR (C6D6, 22 °C): δ 71.6 (s).
peaks was <2% of the theoretical value expected for the fully
protio components. PhCCD was concomitantly observed (no
detectable acetylenic H signal).
Da ta for (P NP tBu )Re(CtCBu t)2. 1H NMR (C6D6, 22 °C):
δ 1.96 (t, J HP ) 5 Hz, 36H, PC(CH3)3), 1.81 (t, 4 Hz, 4H, PCH2-
Si), 1.58 (s, 18H, CtC-C(CH3)3), 0.27 (s, 12H, SiCH3).
Rea ction of (P NP tBu )Re(CtCP h )2 w ith D2. (PNPtBu)Re-
(CtCPh)2 (28.0 mg, 33.0 µmol) was dissolved in 0.6 mL of C6D6
in a J . Young NMR tube. The solution was degassed and the
NMR tube was filled with ca. 300 Torr of D2 (ca. 2 mL, ca. 35
µmol). The reaction was monitored by 1H NMR. Only traces
(<2%) of (PNPtBu)ReH3(CtCPh) were observed after 10 min.
After 24 h ca. 20% of (PNPtBu)Re(CtCPh)2 underwent hydro-
genolysis, but protio Re-H peaks of (PNPtBu)ReH3(CtCPh)
and (PNPtBu)Re(H)4 were not observed. After 4 days at 22 °C
the ratio of (PNPtBu)Re(CtCPh)2, (PNPtBu)ReD3(CtCPh), and
1
Ack n ow led gm en t. This work was supported by the
National Science Foundation. We thank Dr. Cristina
Can˜ada Vilalta for facilitating the SQUID measure-
ments.
Su p p or t in g In for m a t ion Ava ila b le: Full crystallo-
graphic information, including cif files, on two determinations,
together with DFT optimized geometries on molecules de-
scribed in the text are available free of charge via the Internet
at http://pubs.acs.org.
1
(PNPtBu)Re(D)4 was 27:18:55. The intensity of the H hydride
OM049424I