Modification of [8,8,8-(H)(PPh3)2-9-(Py)- Nido -8,7-RhSB9H9], Py = NC5H5, with monodentate phosphines: Reactivity and mechanistic insights

Article abstract of DOI:10.1021/om2009205

The [8,8,8-(H)(PPh₃)₂-9-(Py)-nido-8,7-RhSB ₉H₉] (2)/[1,1-(PPh₃)₂-3-(Py)-closo- 1,2-RhSB₉H₈] (3) pair catalyzes the hydrogenation of olefins through nido-to-closo transformations. Substitution of the phosphine ligands can lead to an improvement of the catalytic activity of this system. Therefore, the substitutional chemistry of 2 with PMePh₂, PMe ₂Ph, and PMe₃ has been studied, leading to the formation of [8,8,8-(H)(PPh₃)(PR₃)-9-(Py)-nido-8,7-RhSB ₉H₉], where R₃ = Me₂Ph (5) or Me₃ (6), and [8,8,8-(H)(PR₃)₂-9-(Py)-nido-8,7- RhSB₉H₉], where R₃ = MePh₂ (4) or Me₂Ph (7). Kinetic studies on the reaction of PMe₂Ph with 2 indicate that the substitutions follow a dissociative mechanism. The thermal dehydrogenation of 5-7 affords the corresponding closo-derivatives [1,1-(L)(PPh₃)-3-(Py)-closo-1,2-RhSB₉H₈], where L = PMe₂Ph (9) or PMe₃ (10), and [1,1-(L) ₂-3-(Py)-closo-1,2-RhSB₉H₈], where L = PMe ₂Ph (11) or PMe₃ (12). The substitution of PPh₃ by the more basic, less bulky phosphines facilitates hydrogen loss and consequent nido-to-closo transformations. The reaction of 5 and 6 with C ₂H₄ promotes a nido-to-closo cluster change and the consequent formation of 10 and 11 together with small amounts of C ₂H₄-ligated [1,1-(η²-C₂H ₄)(L)-3-(Py)-closo-1,2-RhSB₉H₈], where L = PPh₃ (13) or PMe₃ (15), characterized in situ by ¹H NMR spectroscopy. In the reactions with ethylene, ethane is detected in situ, indicating that the olefin is hydrogenated. The reactions of 5 and 6 with CO afford the CO-ligated [1,1-(CO)(L)-3-(Py)-closo-1,2-RhSB ₉H₈], where L = PMe₂Ph (16) or PMe₃ (17). The reactivity of the new PR₃-ligated nido-clusters versus H₂, C₂H₄, and CO is not improved with the phosphines studied in this work; however, the changes found in the chemical beahavior of this system are dramatic, confirming the tailorability of these 11-vertex rhodathiaboranes and the potential optimization of their catalytic activity by the judicious choice of the exo-polyhedral ligands.

Full text of DOI:10.1021/om2009205

Article
pubs.acs.org/Organometallics
Modification of [8,8,8-(H)(PPh3)2-9-(Py)-nido-8,7-RhSB9H9],
Py = NC5H5, with Monodentate Phosphines: Reactivity and
Mechanistic Insights
†
†
,†
†
†
́
́
Beatriz Calvo, Alvaro Alvarez, Ramon Macías,* Pilar García-Orduna, Fernando J. Lahoz,
́
̃
and Luis A. Oro†,‡
†Instituto de Síntesis Química y Catal
́ ́
isis Homogenea (ISQCH), Universidad de Zaragoza-Consejo Superior de Investigaciones
Científicas, Pedro Cerbuna 12, 50009-Zaragoza, Spain
‡King Fahd University of Petroleum and Minerals, KFUPM Visiting Professor, Dhahran 31261, Saudi Arabia
S
* Supporting Information
ABSTRACT: The [8,8,8-(H)(PPh3)2-9-(Py)-nido-8,7-
RhSB9H9] (2)/[1,1-(PPh3)2-3-(Py)-closo-1,2-RhSB9H8] (3)
pair catalyzes the hydrogenation of olefins through nido-to-
closo transformations. Substitution of the phosphine ligands
can lead to an improvement of the catalytic activity of this
system. Therefore, the substitutional chemistry of 2 with
PMePh2, PMe2Ph, and PMe3 has been studied, leading to the
formation of [8,8,8-(H)(PPh3)(PR3)-9-(Py)-nido-8,7-
RhSB9H9], where R3 = Me2Ph (5) or Me3 (6), and [8,8,8-
(H)(PR3)2-9-(Py)-nido-8,7-RhSB9H9], where R3 = MePh2 (4)
or Me2Ph (7). Kinetic studies on the reaction of PMe2Ph with
2 indicate that the substitutions follow a dissociative mechanism. The thermal dehydrogenation of 5−7 affords the corresponding
closo-derivatives [1,1-(L)(PPh3)-3-(Py)-closo-1,2-RhSB9H8], where L = PMe2Ph (9) or PMe3 (10), and [1,1-(L)2-3-(Py)-closo-
1,2-RhSB9H8], where L = PMe2Ph (11) or PMe3 (12). The substitution of PPh3 by the more basic, less bulky phosphines
facilitates hydrogen loss and consequent nido-to-closo transformations. The reaction of 5 and 6 with C2H4 promotes a nido-to-
closo cluster change and the consequent formation of 10 and 11 together with small amounts of C2H4-ligated [1,1-(η2-C2H4)(L)-
3-(Py)-closo-1,2-RhSB9H8], where L = PPh3 (13) or PMe3 (15), characterized in situ by 1H NMR spectroscopy. In the reactions
with ethylene, ethane is detected in situ, indicating that the olefin is hydrogenated. The reactions of 5 and 6 with CO afford the
CO-ligated [1,1-(CO)(L)-3-(Py)-closo-1,2-RhSB9H8], where L = PMe2Ph (16) or PMe3 (17). The reactivity of the new PR3-
ligated nido-clusters versus H2, C2H4, and CO is not improved with the phosphines studied in this work; however, the changes
found in the chemical beahavior of this system are dramatic, confirming the tailorability of these 11-vertex rhodathiaboranes and
the potential optimization of their catalytic activity by the judicious choice of the exo-polyhedral ligands.
The results of this work have provided new rhodathiaboranes
[8,8-(PPh3)(PR3)-nido-8,7-RhSB9H10], where R3 = Me3 and
Me2Ph, and [8,8-(PMePh2)2-nido-8,7-RhSB9H10], which
(a priori) were convenient starting materials for the synthesis
of new hydridorhodathiaboranes derived from 2; however, as
these compounds are obtained as mixtures with the
corresponding tris-PR3-ligated species, [8,8,8-(PR3)3-nido-8,7-
RhSB9H10], separatory work was necessary for the isolation of
the clusters in moderate to low yields. In view of these results,3
and as a continuation of our systematic tailoring of the
11-vertex rhodathiaborane system 2/3, we decided to study the
substitutional chemistry of 2.
INTRODUCTION
■
In recent years, we have developed the chemistry of pyridine-
ligated 11-vertex rhodathiaboranes that are synthesized from
[8,8-(PPh3)2-nido-8,7-RhSB9H10] (1).1 In particular, the system
formed by [8,8,8-(H)(PPh3)2-9-(Py)-nido-8,7-RhSB9H9] (2)
and [1,1-(PPh3)2-3-(Py)-closo-1,2-RhSB9H8] (3) has been
found to give rise to an interesting reaction chemistry that
embraces (i) nido-to-closo deshydrogenation, (ii) dihydrogen
promoted closo-to-nido transformations, (iii) oxidative addition
of sp C−H bonds, and (iv) catalysis of hydrogenation and
isomerization of olefins (Scheme 1).2
These 11-vertex clusters are easily prepared, versatile, and
stable, and, in view of the reactivity above-commented, they are
compounds with potential as homogeneous catalyst precursors.
Therefore, in an attempt to modify the reactivity of these
clusters, we have already carried out substitution of the exo-
polyhedral PPh3 ligands in 1 with the more basic (less bulky)
monodentate phosphines PMe3, PMe2Ph, and PPh2Me.3
Here we describe the reactivity of compound 2 with PMePh2,
PMe2Ph, and PMe3, the subsequent preparation of new
11-vertex hydrido-Rh nido-clusters, and the study of their
Received: October 1, 2011
Published: March 27, 2012
© 2012 American Chemical Society
2986
dx.doi.org/10.1021/om2009205 | Organometallics 2012, 31, 2986−2995

Organometallics
Article
Scheme 1. Formation of 2 and Ethylene Dihydrogen Mediated nido → closo → nido Conversion
Scheme 2
reactivity with small molecules such as CO, C2H4, and H2,
discussing trends and differences within the new series of PR3-
ligated clusters and with the previously reported PPh3-ligated
counterparts 2 and 3.
that contain the new disubstituted species [8,8,8-(H)(PMePh2)2-
9-(Py)-nido-8,7-RhSB9H9] (4) and the starting 2 (Scheme 2).
The new 11-vertex hydridorhodathiaboranes 4−7 have been
characterized by multinuclear NMR spectroscopy and mass
spectrometry. In addition, a solid-state X-ray diffraction analysis
confirmed the molecular structure of 4 (Figure 1).
These rhodathiaboranes have an 11-vertex {RhSB9} core
geometry, formally derived from an icosahedron by the removal
of a vertex, which resembles the structures of the pyridine-
ligated counterpart 22a and the PMePh2-B(9)-substituted
[8,8,8-(H)(PMePh2)2-9-(PMePh2)-nido-8,7-RhSB9H9] (8).3
The framework structures correspond exactly to what one
expects for a 13-skeletal electron pair (sep) cluster with 11
vertices.4 This contrasts with the formal unsaturation found in
the parent compound, 1,1 which has only 12 sep, but a nido
structure. The discrepancy between the number of sep and the
structure is an interesting feature of 11-vertex heteroborane
RESULTS AND DISCUSSION
■
Reactions of [8,8,8-(H)(PPh3)2-9-(Py)-nido-8,7-
RhSB9H9] (2) with PR3. Treatment of the bis-PPh3-ligated
compound 2 with one equivalent of phosphine PR3 affords
[8,8,8-(H)(PPh3)(PR3)-9-(Py)-nido-8,7-RhSB9H9], where R3 =
Me2Ph (5) or Me3 (6). A 1 to 4 molar ratio of 2 versus PMe3
yields also the monosubstitued hydrido cluster 6, whereas
the reaction of 2 with four equivalents of PMe2Ph affords the
mixture, 5, which contains two monosubstituted isomers (vide
infra) 5a and 5b and the disubstitutued [8,8,8-(H)(PMe2Ph)2-
9-(Py)-nido-8,7-RhSB9H9] (7) (Scheme 2). In interesting contrast,
the reaction of 2 with PMePh2 in a 1:1 molar ratio gives mixtures
2987
dx.doi.org/10.1021/om2009205 | Organometallics 2012, 31, 2986−2995

Organometallics
Article
Table 1. trans-Effect of the Hydride Ligand in Some
11-Vertex nido-Rhodathiaboranes
compound
Rh(8)−S(7) (Å)
ref
a
[(PPh3)2RhSB9H10] (1)
2.3769(6)
2.3757(8)
2.3736(7)
2.431(2)
1
a
[(PMePh2)3RhSB9H10]
6
a
[(PMe3)3RhSB9H10]
3
b
[(H)(PPh3)2RhSB9H9(Py)] (2)
2b
b
[(H)(PMePh2)2RhSB9H9(Py)] (4)
2.4271(5)
2.4172(5)
this work
3
b
[(H)(PMePh2)2RhSB9H9(PMePh2)] (8)
a
b
Non-hydride-containing cluster. Hydride-containing cluster.
9-position of the cluster, and the difference between the two
Rh−P distances within each cluster is ca. 0.013(2) Å.
The structure of 4 is consistent with its spectroscopic data
and those of the analogues, 2, 5 (mixture of isomers), 6, and 7.
In this type of compound, the 11B NMR resonances are found in
the interval between δ(11B) +12.0 and −28.0 ppm. The pattern
does not change significantly through the series, with the
boron-11 resonance at the lowest field assigned to the pyridine-
ligated boron atom at the 9-postion.
Figure 1. Molecular structure of compound 4. Only the ipso-carbon
atoms on the phenyl groups are included to aid clarity. Ellipsoids are
shown at 50% probability level. Selected interatomic distances [Å] and
angles [deg]: Rh(8)−S(7) 2.4270(5), Rh(8)−P(1) 2.3091(6),
Rh(8)−P(2) 2.3224(6), Rh(8)-H 1.53(4), Rh(8)−B(3) 2.230(2),
Rh(8)−B(4) 2.209(2), Rh(8)−B(9) 2.214(2), B(9)-N 1.556(3);
P(1)−Rh(8)−P(2) 98.10(2), P(1)−Rh(8)−S(7) 105.08(2), P(2)−
Rh(8)−S(7) 97.82(2), P(1)−Rh(8)−B(9) 95.66(6), P(2)−Rh(8)−
B(9) 163.22(6), P(1)−Rh(8)−B(3) 157.35(7), P(1)−Rh(8)−B(4)
141.17(6), P(2)−Rh(8)−B(3) 87.61(6), P(2)−Rh(8)−B(4)
116.59(6), S(7)−Rh(8)−H 168.2(14).
The Rh−H hydride and B−H−B bridging hydrogen
resonances are quite informative, serving as diagnostic for the
formation of these clusters. The hydride signal appears close to
δ(1H) −12.5 ppm in the series: an apparent quartet for the bis-
PPh3- and bis-PMePh2-ligated clusters, 2 and 4, and a doublet
of doublets of doublets for the monosubstituted analogue 6.
In contrast, the PMe2Ph-monosubstituted counterpart 5
exhibits two hydride resonances in a 1:0.6 ratio, indicating
the presence of two isomers (5a and 5b, Scheme 2).
1
Scheme 3 depicts the shift of the hydride H resonance
found in 2 and 4−7. Monosubstitution with PMe2Ph and PMe3
at the position trans to the B(9) vertex results in a shielding of
the signal, the shift being larger for the more basic phosphine,
PMe3. It appears that substitution at the position trans to the
B(3)−B(4) edge results in a smaller high-field shift of the
hydride resonance (5b vs 5a).
An interesting characteristic that these Py-ligated hydrido-
rhodathiaboranes share is the temperature-dependent behavior
of their 31P{1H} NMR spectra (see Figures 2, S1, and S2).
Thus, one of the phosphorus resonances is significantly broader
than the other at room temperature, and it sharpens as the
temperature is decreased. First discovered for compound 2
(Figure S2), it was thought that this behavior is consistent with a
dissociation of the PPh3 ligand trans to the B(9) vertex.2b
However, this conclusion should be put into question since the
broadening could be due to 31P−11B couplings, which at lower
temperatures are reduced due to the thermal decoupling of the
quadrupolar 11B (I = 3/2) nucleus.7
frameworks that incorporate {Rh(L)2} or {Pt(L)2} (L =
phosphine ligands) fragments,5 making them intrinsically Lewis
acidic and, therefore, potentially labile. The Rh(8)−S(7)
distance of 2.4270(5) Å in 4 is close to the values found for
the previously reported analogue 2 (Schemes 1 and 2) and 8
(Chart 1, I),2a,3 where the hydride ligand is also trans to the
Chart 1
The 31P{1H} spectra of compounds 4 and 6 are assigned on
the basis of the commented relative broadening of the peaks
and the well-differentiated low-field chemical shift of the PPh3
ligand with respect to PPhMe2 and PMe3. The assignments for
the isomers 5a and 5b and the bis-PPh2Me-ligated cluster 7
were made by analysis of a series of 1H, 1H{31P}, 31P{1H}, and
1H−31P HSQC spectra (see Figures S8 to S14).
Substitution Reactions: Mechanistic Considerations.
In order to obtain some mechanistic insights into the
substitutional chemistry of 2 with monodentate phosphines,
small-scale reactions were studied at low temperatures by NMR
spectroscopy. Upon the addition of one equivalent of PMe3 to
a CD2Cl2 solution of 2 at 188 K, the 31P{1H} spectrum exhibits
the signals of compound 2, free PPh3, and PMe3 (Figure 2).
S(7) vertex (Table 1). In contrast, the Rh(8)−S(7) distance is
ca. 0.05 Å shorter in tris-PR3-ligated nido-rhodathiaboranes that
bear a phosphine ligand trans to the sulfur atom,3 confirming
the high structural trans influence of the hydride ligand
(Table 1). It is interesting to note that Rh−P distances in the
PMePh2-ligated cluster 4 are ca. 0.03 Å shorter than the
corresponding values in the PPh3 analogue 2, reflecting the
higher σ-donor capabilities of PMePh2 and its smaller steric
hindrance. In 2, 4, and 8, the longest Rh−P distance cor-
responds to the phosphine ligand trans to the substituted
2988
dx.doi.org/10.1021/om2009205 | Organometallics 2012, 31, 2986−2995

Organometallics
Scheme 3. Illustration of the H Chemical Shift of the Rh−H Hydride Ligand
Article
1
Figure 2. Variable-temperature 31P{1H} NMR spectra of 2 upon addition of one equivalent of PMe3 at low temperature (bottom spectrum).
The resonances of the substitution product 6 are not well
detected in the 31P{1H} spectra until the temperature reaches
273 K. At room temperature, the reaction is completed
[Figure 2: the broadness of the resonance close to δ(31P) −
to the B(9) vertex. Following the same procedure, the treat-
ment of 2 with one equivalent of PMe2Ph at low temperatures
does not afford significant amounts of substitution products
until the temperature reaches 273 K (see Figures S4 and S5).
At room temperature, there is no free PMe2Ph left and there
are two sets of major signals in the 31P{1H} NMR spectrum. As
commented above, the spectrum is consistent with the presence
of two {Rh(H)(PPh3)(PMe2Ph) }-to-{SB9H9(Py)} isomers:
the signals at the lowest field are assigned to the PPh3 ligands,
whereas the peaks at around δ(31P) −5 ppm correspond to the
PMe2Ph groups. In the same conditions, the reaction of 2 with
one equivalent of PMePh2 also starts to take place at significant
rates above 273 K (see Figures S6 and S7), but in contrast,
there is formation of the PMePh2-disubstituted cluster, 4.
2
15.0 ppm arises most probably from J(31P−11B) coupling as
commented above]. The proton NMR spectrum reflects these
results, showing the disappearance of the starting material 2
and the consequent formation of its substitution product 6
(Figure S3). There is also a weak Rh−H hydride signal at
−12.70 ppm, which may correspond to a minor {Rh(H)-
(PMe3)(PPh3)}-to{SB9H9(Py)} isomer. In any event, these
data reveal that the reaction of 2 with PMe3 affords the mono-
substituted product 6 with good regioselectivity: the sub-
stitution takes place at the exo-polyhedral position that is trans
2989
dx.doi.org/10.1021/om2009205 | Organometallics 2012, 31, 2986−2995

Organometallics
Article
These results indicate clearly that the reactions of the
monodentate phosphines, PMePh2, PMe2Ph, and PMe3, with 2
occur at lower rates than with the previously studied
rhodathiaborane, 1.3 Variable-temperature experiments sug-
gested that the most plausible mechanism for the substitution
of the phosphine ligands in the parent compound 1 (Scheme 1)
is associative, implying the binding of a third phosphine at the
rhodium center.3 For the herein described reactions of 2 with
monodentate phosphines, however, it is not clear if, as discussed
above, the broadening of one of the two 31P resonances in this
new series of hydridorhodathiaboranes, 4−7, is due to 31P−11B
coupling or to dissociative processes (or both).
In compound 6, DFT calculations have confirmed the
assignment of the broad 31P signal at the highest field as the
PMe3 ligand trans to the pyridine-substituted B(9) vertex
(Table S10); in addition, previous DFT studies revealed that
elongation of the Rh−PH3 bond trans to the B(9)−Py vertex in
the model [8,8,8-(H)(PH3)2-9-(Py)-nido-8,7-RhSB9H9] results
in a smaller energy increase of the cluster compared with the
elongation of the Rh−PH3 bond trans to the B(3)−B(4)
edge.2b Thus, these theoretical results suggest that dissociation
of the PR3 ligand trans to the {B(9)−Py} vertex in 2 and 4−7
may play an important role in the reactivity of these clusters.
Kinetic Studies. The following kinetic studies try to shed
more light on the reaction mechanism. The reaction between 2
and PMe2Ph was studied by 1H NMR spectroscopy, monitoring
the decrease of the hydride integral resonance versus time. The
concentration of PMe2Ph was kept in excess to hold pseudo-
first-order conditions. The value of kobs, summarized in Table 2,
although the commented-above temperature-dependent broad-
ening in the 31P{1H} spectra of 2−7 may arise mainly from
the effects of “thermal decoupling” on the boron nuclei,
dissociation of the phosphine ligand trans to B(9), as previously
suggested by us,2b appears to direct the reactivity of these
clusters.
Reactivity Studies. The modification of the exo-polyhedral
ligands in the 11-vertex rhodathiaboranes 2 and 3 is a
reasonable approach for the fine-tuning of the reactivity of
this system. The adequate combination of exo-polyhedral
ligands bound either to the rhodium atom or to the B(9) vertex
can (a priori) result in an improvement of the catalytic activity
of this kind of cluster. Therefore, we carried out a systematic
study of the reactivity of the new PR3-ligated clusters, 4−7,
having the PPh3-ligated clusters 2 and 3 as reference.
Dehydrogenation. The hydridorhodathiaboranes 2 and 5−7
are stable in solution at room temperature; however, heating
at reflux temperature in dichloromethane results in the loss of
dihydrogen and the concomitant nido-to-closo transformation of
the clusters (Scheme 4). In contrast, the PMePh2-ligated coun-
terpart, 4, does not undergo dehydrogenation. Quantitative studies
1
on the dehydrogenation reactions were carried out by H NMR
spectroscopy at 50 °C in CD2Cl2 (Tables S4−S6). The results
were adjusted to first-order kinetics, yielding the reaction rates and
half-life constants listed in Table 3.
Table 3. Rate Constants and Half-Lives for the
Dehydrogenation Reactions of 2, 5a, 5b, 6, and 7 in CD2Cl2
at 50 °C
compound
k × 104 min−1
t1/2 (h)
Table 2. Rate Constants for PPh3 Substitution in
Compound 2
a
2
26 1.4
41 0.5
25 0.4
45 1.6
83 1.5
4.4
2.8
4.6
2.6
1.4
5a
5b
6
b
T (K)
[PMe2Ph] (equivalents)
102kobs (min−1
)
273
273
263
268
273
279
283
5
20
10
10
10
10
1.4
1.4
7
0.13
0.27
1.4
These data show that the fastest rate of dehydrogenation
corresponds to the bis-PMe2Ph-ligated 7: roughly three times
as fast as the parent cluster 2 and the isomers 5b. Next in the
series, we find the isomer 5a and the PMe3-monosubstituted
derivative 6, which exhibit a rate of dehydrogenation
approximately two times faster than 2 (and 5b). This trend
indicates that the substitution of PPh3 with more basic and
less bulky phophine lingads facilitates the dehydrogenation of
the clusters and the consequent nido-to-closo transformation.
In addition, the different rate between the isomers 5a and 5b
suggests that a trans arrangement of PR3, where R3 = Me2Ph
(5a and 5b) and Me3 (6), to the B(9) vertex results in an
increase of the dehydrogenation rate. It is somewhat surprising,
2.5
10
14.3
a
b
[2] = 0.02 M in CD2Cl2. Values obtained from three different
measurements.
is independent of the concentration of PMe2Ph, suggesting
that the substitution of one PPh3 ligand in 2 takes place in a
dissociative process. The activation parameters obtained from
the Eyring plot (Figure S15), ΔH‡ = 33 3 kcal mol−1 and
ΔS‡ = 45 9 eu, are consistent with dissociation of PPh3 as
the most plausible mechanism for this reaction. Therefore,
Scheme 4
2990
dx.doi.org/10.1021/om2009205 | Organometallics 2012, 31, 2986−2995

Organometallics
Article
however, that in the same conditions hydrogen loss is
completely hindered for the PMePh2-ligated compound 4.
The thermal treatment of 5, 6, and 7 is a convenient route to
the synthesis of the mixed-ligated closo-derivatives [1,1-(L)
(PPh3)-3-(Py)-closo-1,2-RhSB9H8], where L = PMe2Ph (9) or
PMe3 (10), and the bis-PR3-ligated analogues [1,1-(PR3)2-3-
(Py)-closo-1,2-RhSB9H8], where L = PMe2Ph (11) or PMe3
(12). These compounds have been fully characterized by
multinuclear NMR and mass spectrometry. X-ray diffraction
analysis has confirmed the molecular structures of 10 and 11
(Figures 3 and Figure S16, respectively).
activity of the system (see Scheme 1 above).2b Therefore, one
of the objectives of this work was to change the phosphine
ligands to optimize the activation of dihydrogen by 11-vertex
closo-rhodathiaboranes and subsequently the transfer of the two
hydrogen atoms to unsaturated organic molecules. However, in
the same conditions as 3, the 11-vertex closo-rhodathiaboranes
9−12 do not react with dihydrogen. Thus, although the loss of
H2 is facilitated by the presence of the more basic and less
bulky PMe2Ph and PMe3 ligands in 5−7, the reverse reaction
with H2 (to regenerate these nido-hydrido clusters) is hindered.
In other words, the energy barrier for the dehydrogenation of
the nido-clusters is lower than for the PPh3-ligated compound
2, whereas the activation energy of the reaction of the 11-vertex
closo-rhodathiaboranes 9−12 with H2 appears to be higher than
for 3, precluding the H2 addition to the closo-clusters. In this
context, once more, it is noteworthy that 4 does not undergo
hydrogen loss in the same conditions as 2 and 5−7, indicating
that for the PMePh2-ligated counterpart the activation energy
barrier toward the formation of the hypothetical closo-cluster,
[1,1-(PMePh2)2-3-(Py)-closo-1,2-RhSB9H8], by loss of H2 is
significantly higher. In any event, it is evident that the nature of
the phosphine ligands plays a crucial role in tuning the reactivity
of the closo-cluters toward the activation of dihydrogen;
therefore, the right choice of ligands should improve the
reactivity of the system toward small molecules.
Reactions with CO and C2H4. Both reagents CO and C2H4
react with 2 to give CO- and C2H4-ligated closo-rhodathiabor-
anes [1,1-(L)(PPh3)-3-(Py)-closo-1,2-RhSB9H8], where L =
C2H4 (13, Scheme 1 above) or CO (14, Chart 2, II);2a,b in
Chart 2
Figure 3. Molecular structure of 10. Only the ipso-carbon atoms on
the phenyl rings of the PPh3 ligand are included to aid clarity.
Ellipsoids are shown at 50% probability level. Selected interatomic
distances (Å) and angles (deg): Rh(1)−S(2) 2.3841(9), Rh(1)−P(1)
2.2889(10), Rh(1)−P(2) 2.2710(9), Rh(1)−B(3) 2.087(4), Rh(1)−
B(4) 2.401(4), Rh(1)−B(5) 2.471(4), Rh(1)−B(6), 2.380(4),
Rh(1)−B(7) 2.354(4); P(1)−Rh(1)−P(2) 96.68(4).
The molecular structures of 10 and 11 are based on the
canonical structure of an octadecahedron that may be
rationalized by simple application of the Wade/Mingos
approach.4 The rhodium atom occupies the apical cluster
position connected to six atoms in the {SB9} fragment with two
exo-polyhedral metal coordination sites occupied by the PPh3
and PR3 ligands. Compounds 9−12 are isoelectronic with 3 and
the organometallic derivatives [1,1-(η2-L)(PPh3)-3-(Py)-closo-
1,2-RhSB9H8], where L = η2-C2H4 (13, Scheme 1 above) or η2-
C2(CO2Me)2,2c as well as other rhodathiaboranes that have
different exo-polyhedral ligands on the rhodium center and the
B(3) vertex.2c,8 Not surprisingly, the longest Rh−P distance
[2.3278(13) Å] corresponds to 3, which bears two bulky PPh3
ligands; however, it is interesting to notice that in 10 the P−Rh
distance of the PMe3 ligand [2.2889(10) Å] is longer than that
of PPh3 [2.2710(9) Å], suggesting that the steric effects of PPh3
dominate the better σ-donor capabilities of PMe3.
the reaction with ethylene, there is also formation of ethane
(detected in situ by 1H NMR spectroscopy), demonstrating that
the alkene is hydrogenated. In the current study, it has been
found that the reaction of 5 (mixture of isomers) and 6 with
ethylene promotes the nido-to-closo transformation, yielding 10
and 11, respectively, at room temperature, also formed
thermally in longer reaction times as discussed above. Under a
higher pressure of ethylene (see Experimental Section), the
reaction of 6 has been found to afford, in addition to the major
component 10, small amounts of the ethylene-ligated
rhodathiaborane, 13, and the PMe3-ligated counterpart,
[1,1-(η2-C2H4)(PMe3)-3-(Py)-closo-1,2-RhSB9H8] (15). This
new compound has been characterized in situ by NMR
spectroscopy, and the chemical nonrigidity of the Rh−C2H4
Reactions with H2(g). The closo-rhodathiaborane 3 reacts
with dihydrogen to give the nido-cluster 2.2a This is a significant
reaction that implies the heterolytic splitting of H2 on the closo-
cluster and completes a stoichiometric cycle of dehydroge-
nation/hydrogenation that represents the essence of the catalytic
1
linkage studied by variable-temperature H NMR experiments
(Figure S17). The activation energy of this rotational fluxional
process is 4 kJ/mol higher for 15 than for 13, suggesting a
stronger Rh−(η2-C2H4) bond as the result of an increase in the
2991
dx.doi.org/10.1021/om2009205 | Organometallics 2012, 31, 2986−2995

Organometallics
Article
Scheme 5
π-back-donation from the metal center to the olefin ligand in the
former compound, which is expected from the presence of the
better σ-donor ligand, PMe3 in 15. In the same conditions, in
contrast, the mixture of isomers 5 leads to only 13 in small
amounts. Both reactions afford ethane (10-to- and 11-to-C2H6
ratio 1:0.2), indicating that although the major reaction appears
to be H2 loss, the transfer of the two hydrogen atoms to the
CC bond of the olefin also occurs for the new PR3-ligated
hydrido-ligated clusters (Scheme 5).
With carbon monoxide, the mixture of isomers 5 reacts to
give the CO-ligated closo-cluster 14 (Chart 2, II) and the new
derivative [1,1-(CO)(PMe2Ph)-3-(Py)-closo-1,2-RhSB9H8] (16)
in a 1:4 ratio. Similarly, compound 6 reacts to give a mixture of
14 and [1,1-(CO)(PMe3)-3-(Py)-closo-1,2-RhSB9H8] (17), this
time in a 1:5 ratio. Therefore, carbon monoxide substitutes
one phosphine ligand in the hydridorhodathiaboranes, causing
the nido-to-closo transformation of the cluster by hydrogen loss.
The PPh3 ligand is preferably substituted by the entering CO
molecule, reflecting the fact that PMe2Ph in 5 and PMe3 in 6 are
worse leaving groups than PPh3.
The new 11-vertex nido-hydridorhodathiaboranes, 5−7,
undergo faster thermal dehydrogenation than the parent
compound 2 to give the corresponding closo-clusters, 9−11,
which do not react with H2 in the same reaction conditions used
for the bis-PPh3-ligated derivative 3. Therefore, the dehydroge-
nation/hydrogenation cycle found for the bis-PPh3-ligated 2/3
pair is not completed (in the conditions of pressure and
temperature studied) with the new rhodathiaboranes reported in
this work. Ethylene promotes hydrogen loss and consequent
nido-to-closo transformation, but compared with 2, the new
species 5 and 6 do not exhibit an enhanced reactivity with the
olefin; perhaps reflecting the fact that the dissociation of the
Rh−PMe2Ph and Rh−PMe3 bonds in 5 and 6 is more difficult
than the dissociation of Rh−PPh3 in 2. This fact together with
the lack of reactivity of the new closo-clusters with dihydrogen,
essential for subsequent hydrogenation of unsaturated organic
molecules, indicates that the change of PPh3 in 2 with the
phosphines studied in this work would not lead to an improve-
ment in the catalytic activity of the system. Likewise, the new
CO-ligated closo-clusters 16 and 17 are inert in the presence
of molecular hydrogen, matching this time the reactivity of the
PPh3-ligated counterpart 14.
Although the objective of improving the reactivity of the 2/3
pair versus dihydrogen and olefins is not accomplished by
substitution of PPh3 with the studied phosphines, the treat-
ment of 2 with PR3, where R3 = Me3, Me2Ph, and MePh2, is a
convenient synthetic route for the alteration of the exo-
polyhedral ligand sphere, confirming the tailorability of this
11-vertex rhodathiaborane system. The lack of reactivity of the
bis-PMePh2-ligated cluster 4 is an interesting contrast, which
suggests that other exo-polyhedral metal ligands may tune the
clusters toward the opposite behavior: an enhancement of the
catalytic activity with respect to 2.
CONCLUSIONS
■
The parent hydrido-Rh cluster 2 reacts readily with mono-
dentate phosphines, affording new 11-vertex derivatives with
compositions and configurations that depend on the entering
phosphine. Thus the treatment of 2 with PMePh2 forms
exclusively the bis-substituted cluster 4. In contrast, the less
bulky and more basic phosphines PMe3 and PMe2Ph afford
monosubstituted species 5 (mixture of isomers) and 6; with
PMe2Ph, it is also possible to prepare the bis-substituted
derivative 7. Kinetic studies demonstrate that the substitution
of the PPh3 ligand in 2 follows a dissociative mechanism that
most likely involves the phosphine ligand trans to the B(9)
position. These results support previous conclusions dealing
with the reactivity of 2 with olefins in which dissociative
mechanisms were invoked.2a,b
2992
dx.doi.org/10.1021/om2009205 | Organometallics 2012, 31, 2986−2995

Organometallics
Article
Crystal data for compound 11: C21H35B9NP2RhS·CH2Cl2, M =
680.67; red prism, 0.203 × 0.149 × 0.117 mm3; monoclinic, P21/c;
a = 9.1278(13) Å, b = 24.263(4) Å, c = 14.915(2) Å; β = 103.397(2)°;
Z = 4; V = 3213.2(8) Å3; Dc = 1.407 g/cm3; μ = 0.878 mm−1, min. and
max. transmission factors 0.798 and 0.902; 2θmax = 58.82°; 34 469
reflections collected, 8363 unique [Rint = 0.0270]; number of data/
restrains/parameters 8363/4/470; final GoF 1.170, R1 = 0.0591 [7700
reflections, I > 2σ(I)], wR2 = 0.149 for all data.
Calculations. All calculations were performed using the Gaussian
03 package.15 Structures were initially optimized using standard
methods with the STO-3G* basis sets for C, B, P, S, and H and with
the LANL2DZ basis set for the rhodium atom. The final optimizations,
including frequency analyses to confirm the true minima, together with
GIAO nuclear-shielding calculations, were performed using B3LYP
methodology, with the 6-31G* and LANL2DZ basis sets. The GIAO
nuclear shielding calculations were performed on the final optimized
geometries, and computed 11B shielding values were related to
chemical shifts by comparison with the computed value for B2H6,
which was taken to be δ(11B) +16.6 ppm relative to the BF3(OEt2) =
0.0 ppm standard.
EXPERIMENTAL SECTION
■
General Procedures. Reactions were carried out under an argon
atmosphere using standard Schlenk-line techniques. Solvents were
obtained dried from a solvent purification system from Innovative
Technology Inc. The commercially available phosphines PMe3,
PMe2Ph, and PMePh2 were used as received without further
purification. The 11-vertex rhodathiaborane 2 was prepared according
to the literature methods.2a,b Preparative thin-layer chromatography
(TLC) was carried out using 1 mm layers of silca gel G (Fluka, type
GF254) made from water slurries on glass plates of dimensions
20 × 20 cm and dried in air at 25 °C. Infrared spectra were recorded
on a Perkin-Elmer Spectrum 100 spectrometer, using a Universal ATR
sampling accessory. NMR spectra were recorded on Bruker Avance
300-MHz and AV 400-MHz spectrometers, using 31P{1H}, 1H-31P
HSQC, 11B, 11B{1H}, 1H, 1H{11B}, and 1H{11B(selective)} techniques.
1H NMR chemical shifts were measured relative to partially deuterated
solvent peaks but are reported in ppm relative to tetramethylsilane. 11
B
chemical shifts are quoted relative to [BF3(OEt)2)], and 31P chemical
shifts are quoted relative to 85% aqueous H3PO4. Mass spectrometric
data were recorded on a MICROFLEX instrument operating in either
positive or negative modes, using matrix-assisted laser desorption/
ionization (MALDI). A nitrogen laser of 337 nm (photon energy of
3.68 eV) was used for the ionization processes, and the molecules
under study were protected with a matrix of trans-2-[3-(4-tert-
butylphenyl)-2-methyl-2-propenylidene]malononitrile.
Kinetic Analysis. The kinetics of the substitution reaction between
2 and PMe2Ph were measured in 0.02 M solutions of the
hydridorhodathiaborane in CD2Cl2. The decay of the hydride complex
resonance was monitored by 1H NMR spectroscopy (Tables S4 and S5).
The integrals were normalized relative to the solvent peak that was used
as an internal standard.
The activation parameters, ΔH‡ and ΔS‡, were obtained from a
linear least-squares fit of ln(k/T) vs 1/T (Eyring equation). Errors
were computed by published methods.16 The error in temperature was
assumed to be 1 K; the error in kobs was estimated as 10%.
X-ray Crystallography. Crystals of compounds 4, 10, and 11
suitable for X-ray diffraction analysis were grown by slow diffusion of
hexane into a concentrated solution of each rhodathiaborane in
dichloromethane. X-ray diffraction data were collected at low temper-
ature (100(2) K) on the BM16 CRG beamline at the ESRF in the case
of 4, and for 10 and 11 using an automatic Bruker Kappa APEX DUO
CCD area detector diffractometer equipped with graphite-monochro-
matic Mo Kα radiation (λ = 0.71073 Å) using narrow frames (0.3° in ω).
In all cases, single crystals were mounted on micromount supports and
were covered with a protective perfluoropolyether. In the case of 4, data
were measured in a single-axis HUBER diffractometer, equipped with an
Oxford 600 Cryosystem open-flow nitrogen cryostat (100(1) K), using
monochromatic silicon(111) synchrotron radiation (λ = 0.7382 Å).
Intensities were integrated including Lorentz and polarization effects
with the HKL2000 suite (4) or SAINT-Plus program9 (10 and 11)
and corrected for absorption using multiscan methods applied with the
SORTAV (4) or SADABS (10 and 11) program.10 The structures
were solved using the SHELXS-86 program.11 Refinements were carried
out by full-matrix least-squares on F2 with SHELXL-97,12 including
isotropic and subsequent anisotropic displacement parameters for all
non-hydrogen atoms. All hydrogen atoms for compounds 4 and 10
were included from observed positions and refined isotropically. For 11,
most of the hydrogen atoms were observed, but some others were
included in calculated positions and refined with a riding model. The
programs ORTEP-313 and PLATON14 were used to prepare the figures
and the crystallographic data.
Synthesis of [8,8,8-(H)(PMePh2)2-9-(Py)-nido-8,7-RhSB9H9] (4). A
82.7 mg (0.083 mmol) amount of 2, placed in a Schlenk tube, was
dissolved in 15 mL of CH2Cl2 under an atmosphere of argon. Three
equivalents of PMePh2 (58.06 mL, 0.29 mmol) was added, and the
resulting solution was stirred for 4 h. Solvent was reduced in volume,
and hexane added to form a yellow precipitate, which was separated
by decantation, washed with hexane, and dried under vacuum. The
resulting yellow product was characterized as 4 (39.2 mg, 0.054 mmol,
56%). IR (ATR): νmax/cm−1 2512 m (BH), 2011 m (RhH). 11B{1H}
NMR (160 MHz; CDCl3; 298 K): δ 12.2 (1B, br, BH), 6.0 (1B, d,
1JB−H = 112 Hz, BH), 1.9 (1B, br, BH), −0.5 (1B, br, BH), −4.5 (1B, br,
1
BH), −10.1 (1B, d, JB−H = 142 Hz, BH), −19.7 (1B, br, BH), −27.1
(2B, d, 1JB−H = 142 Hz, BH). 1H{11B} NMR (300 MHz; CD2Cl2; 298 K):
δ 8.17 (2H, m, Py), 7.85 (1H, m, Py), 7.54 (1H, m, Py), 7.45 (1H,
m, Py), 7.32−6.85 (20H, aromatic, 2PMePh2), 3.86 (1H, br, BH), 3.46
(1H, br, BH), 2.86 (1H, br, BH), 2.39 (1H, br, BH), 1.84 (1H, br, BH),
2
2
1.83 (3H, d, JP−H = 7.1 Hz, PMePh2), 1.64 (3H, d, JP−H = 7.0 Hz,
PMePh2), 1.39 (1H, s, BH), 1.32 (1H, s, BH), 0.92 (1H, s, BH), −1.34
(1H, s, BHB), −12.54 (1H, q, 1JRh−H ≈ 2JP−H = 13 Hz, Rh−H). 31P{1H}
NMR (161 MHz; CDCl3; 223 K): δ 17.4 (1P, dd, 1JRh−P = 104 Hz, 2JPP
too broad to be resolved), 12.6 (1P, dd, 1JRh−P = 128 Hz, 2JP−P = 18 Hz).
MS m/z (MALDI): 520 [M+ − (PPh2Me + 2H), isotope envelope;
PC18H26RhSB9N requires 520; C11H33B9NP2RhS requires 722].
Crystal data for compound 4: C31H41B9NP2RhS·2(CH2Cl2),
M = 891.74; yellow block, 0.078 × 0.065 × 0.052 mm3; monoclinic,
P21/n; a = 10.1578(3) Å, b = 25.6316(2) Å, c = 16.3428(3) Å;
β = 100.3430(10)°; Z = 4; V = 4185.88(15) Å3; Dc = 1.415 g/cm3;
μ = 0.812 mm−1, min. and max. correction factors 0.80 and 1.24;
[8,8,8-(H)(L)(PPh3)-9-(Py)-nido-8,7-RhSB9H9] where L = PMe2Ph
(5), PMe3 (6). Compound 5. The procedure was the same as for
4, using 38.5 mg (0.045 mmol) of 2 and two equivalents of
PMe2Ph (19.4 μL, 0.14 mmol) and stirring the resulting solu-
tion for 15 min. The resulting orange product was characterized
as 5 (30.5 mg, 0.042 mmol, 93%), which is a mixture of two
isomers, 5a and 5b. Both isomers appear to exhibit the same
11B{1H} NMR (160 MHz; CDCl3; 298 K): δ 12.1 (1B, s, BN),
2θmax = 56.24°; 42 060 reflections collected, 8895 unique [Rint
=
0.0731]; number of data/restrains/parameters 8895/0/624; final GoF
1.058, R1 = 0.0420 [8707 reflections, I > 2σ(I)], wR2 = 0.112 for all
data.
1
4.3 (1B, br BH), −0.9 (2B, d, JB−H = 102 Hz, BH), −4.9 (1B,
Crystal data for compound 10: C26H37B9NP2RhS . CH2Cl2,
M = 742.69; red plate, 0.195 × 0.094 × 0.037 mm3; triclinic, P1;
br, BH), −9.6 (1B, d BH), −21.4 (1B, br, BH), −25.9 (1B, d,
̅
1
1JB−H = 159 Hz, BH), −27.6 (1B, d, JB−H = 142 Hz, BH);
similarly, the directly boron-bound proton resonances were
a = 10.8930(11) Å, b = 12.0718(12) Å, c = 14.3942(14) Å; α =
87.299(2)°, β = 69.5940(10)°, γ = 81.4180(10)°; Z = 2; V = 1754.1(3)
Å3; Dc = 1.406 g/cm3; μ = 0.811 mm−1, min. and max. transmission
factors 0.798 and 0.943; 2θmax = 59.08°; 19 013 reflections collected,
8827 unique [Rint = 0.0474]; number of data/restrains/parameters
8827/0/544; final GoF 1.010, R1 = 0.0481 [6398 reflections, I > 2σ(I)],
wR2 = 0.115 for all data.
1
indistinguishable for the isomers: H{11B} NMR (300 MHz;
CD2Cl2; 298 K): δ 3.86 (1H, s, BH), 3.48 (1H, s, BH), 3.27
(1H, s, BH), 2.68 (1H, s, BH), 1.73 (1H, s, BH), 1.43 (1H, s,
BH), 1.29 (1H, s, BH), 0.83 (1H, s, BH), −1.27 (1H, s, BHB).
MS m/z (MALDI): 721 [(M+ − H), isotope envelope;
2993
dx.doi.org/10.1021/om2009205 | Organometallics 2012, 31, 2986−2995

Organometallics
Article
P2C31H41RhSB9N requires 722], 582 [(M+ − PMe2Ph − 2H),
isotope envelope].
Synthesis of [1,1-(PMe3)2-3-(Py)-closo-1,2-RhSB9H8] (12). A
Schlenk tube was charged with 2 (112.8 mg, 0.1333 mmol). The
rhodathiaborane was dissolved in 10 mL of CH2Cl2, and then three
equivalents of PMe3 (399.9 μL, 0.399 mmol) was added under an argon
atmosphere. The resulting solution was heated at 70 °C in an argon
atmosphere for 12 h. Then, solvent was evaporated, and the resulting
solid crystallized in CH2Cl2/hexane, affording 12. Yield: 56.1 mg,
0.1189 mmol, 89.2%. 31P{1H} NMR (121 MHz; CDCl3; 300 K): δ −9.3
(2P, d, 1JRh−P = 147 Hz). Anal. Calcd for C11H31B9NP2RhS: C, 28.02; H,
6.62; N, 2.97; S, 6.80. Found: C, 27.86; H, 6.59; N, 2.86; S, 6.48.
Synthesis of [1,1-(CO)(PMe2Ph)-3-(Py)-closo-1,2-RhSB9H8]
(16). A 51.02 mg portion of the isomer mixture of 5 (0.0707 mmol)
was loaded in a Schlenk tube and dissolved in 10 mL of CH2Cl2.
A balloon filled with CO(g) was attached to the Schlenk tube; the
system was then cooled in liquid nitrogen and evacuated under vacuum.
The system was exposed to the carbon monoxide atmosphere created
upon opening of the balloon, and the reaction was stirred overnight at
room temperature. The solvent was evaporated under vacuum, and the
red solid studied by NMR. The data demonstrated formation of the
PPh3-ligated closo-rhodathiaborane 14 and the CO-ligated analogue 15
in a 1:4 ratio, which results in a yield of 53% for 16 (nonisolated,
calculated from the NMR data). NMR data of 15: IR(ATR): νmax/cm−1
2507 vs (BH), 1969 vs (CO), 1620 m, 1482 m, 1458 m, 1434 m, 1259 m,
1093 m, 1006 s, 905 s, 799 m, 682 s, 524 m, 488 m. 31P{1H} NMR
Additional NMR data for 5a: 1H{11B} NMR (400 MHz, CD2Cl2):
2
δ +1.53 (3H, doublet, JH−P = 7.5 Hz, PMe2Ph), +1.36 (3H, doublet,
2JH−P = 7.5 Hz, PMe2Ph), −13.02 (1H, q, 1JH−Rh ≈ 2JH−P = 20 Hz, Rh−
H). 31P{1H} NMR (162 MHz; CD2Cl2; 223 K): δ +35.2 (1P, dd,
2
1
1JP−Rh = 128 Hz, JPP = 28 Hz, PPh3), −2.9 (1P, dd, JP−Rh = 105 Hz,
2JP−P = 27 Hz, PMe2Ph).
Additional NMR data for 5b: H{11B} NMR (300 MHz, CDCl3):
1
δ +1.44 (3H, d, 2JH−P = 7.4 Hz, PMe2Ph), +1.01 (3H, d, 2JH−P = 7.7 Hz,
PMe2Ph), −12.78 (1H, q, JH−Rh ≈ JH−P = 22 Hz, Rh−H). 31P{1H}
1
2
1
NMR (162 MHz; CD2Cl2; 223 K): δ +39.2 (1P, dd, JP−Rh = 108 Hz,
2JP−P = 19 Hz, PPh3), −2.1 (1P, dd, JP−Rh = 128 Hz, JP−P = 19 Hz,
1
2
PMe2Ph).
Compound 6. Following the same synthetic steps described for
compounds 4 and 5, 76.1 mg (0.069 mmol) of 2 was treated with two
equivalents of PMe3 (18.5 mL, 0.18 mmol). Yield: yellow product,
0.023 g (0.048 mmol, 54%). Anal. Calcd for C26H39B9NP2RhS: C,
47.33; H, 5.96; N, 2.12; S, 4.86. Found: C, 47.07; H, 5.88; N, 1.45; S,
3.95. IR (ATR): νmax/cm−1 2910 m (BH), 2112 m (RhH). 31P{1H}
1
NMR (161 MHz; CD2Cl2; 223 K): δ 36.9 (dd, JRh−P = 129 Hz,
1
2JP−P = 30 Hz, PPh3), −12.9 (br d, JP−Rh = 103 Hz, PMe3). MS m/z
(MALDI): 409 [M+ − (PPh3 + 1H), isotope envelope;
PC8H33RhSB9N].
1
(202 MHz; CDCl3; 300 K): δ 1.9 (d, JRh−P = 129 Hz, PMe2Ph).
Synthesis of [8,8,8-(H)(PMe2Ph)2-9-(Py)-nido-8,7-RhSB9H9]
(7). 2 (12 mg, 0.014 mmol) was treated with PMe2Ph (7.8 mg, 8
mL, 0.057 mmol) in 10 mL of CH2Cl2. The resulting solution was
stirred under an atmosphere of argon for 2.5 h. Solvent was
evaporated, and the solid subjected to NMR studies in CD2Cl2. The
NMR spectra showed that the product of the reaction contained the
isomers 5a and 5b together with the bis-PMe2Ph-substituted cluster 7
in a 1:1:2 ratio. Isolation of compound 7 was not possible due to its
instability toward dehydrogenation; therefore, the characterization
of 7 was carried out in situ by NMR spectroscopy. 31P{1H} NMR (161
Synthesis of [1,1-(CO)(PMe3)-3-(Py)-closo-1,2-RhSB9H8] (17).
An orange solution of 6 (10 mg, 0.015 mmol) in 2 mL of CD2Cl2
was stirred under an atmosphere of CO(g) at room temperature for
15 min, during which time the initial bright orange solution became
yellow. The final mixture was washed repeatedly with hexane to give
5.9 mg of 17 (0.014 mmol, 92%). IR (ATR): νmax/cm−1 2521 s (BH),
2475 s (BH), 1979 vs (CO), 1620 m, 1480 m, 1457 m, 1434 m, 1156
m, 1091 m, 1006 m, 933 m, 799 m, 681 m. 11B{1H} NMR (160 MHz;
1
CD2Cl2; 298 K): δ 55.0 (1B, s, B-py), 27.4 (1B, d, JB−H = 129 Hz,
BH), 0.7 (1B, br, BH), −1.8 (1B, d, 1JB−H = 120 Hz, BH), −14.1 (1B,
1
MHz; CD21Cl2; 298 K): δ 3.2 (d, JRh−P = 77 Hz, PB trans to B(9)),
br, BH), −25.8 (1B, br, BH), −32.3 (1B, d, 1JB−H = 149 Hz, BH), −33.8
2
−1.8 (dd, JRh−P = 124 Hz, JP−P = 24 Hz, PA trans to B(3)−B(4)
1
1
(1B, d, JB−H = 146 Hz, BH). H NMR (300 MHz; CD2Cl2; 300 K):
δ 9.41 (2H, d, J = 5.6 Hz, Py), 8.26 (1H, m, Py), 7.82 (2H, m, Py), 4.40
(1H, s, BH), 2.51 (1H, s, BH), 2.17 (1H, s, BH), 1.83 (1H, s, BH), +1.35
(3H, d, 2JP−H = 10.3 Hz, PMe3), 0.38 (1H, s, BH), 0.32 (1H, s, BH), 0.05
(1H, s, BH), 0.02 (1H, s, BH). 31P{1H} NMR (202 MHz; CD2Cl2;
300 K): δ −6.1 (d, 1JRh−P = 137 Hz, PMe3); m/z (MALDI−) 421 (M − 4H;
isotope envelop; PC9H22ORhSB9N requires 425).
edge).
Synthesis of [1,1-(PPh3)(PMe2Ph)-3-(Py)-closo-1,2-RhSB9H8]
(9). A 5.9 mg (0.0082 mmol) amount of a mixture of isomers 5a and
5b was dissolved in CD2Cl2 in an NMR tube and heated at 70 °C for
3 h. Solvent was reduced in volume, and hexane added to form a
yellow precipitate, which was separated by decantation, washed with
hexane, and dried under vacuum. The resulting red product was
characterized as 9 (4.2 mg, 0.059 mol, 72%). 31P{1H} NMR (161
MHz; CDCl3; 223 K): δ 50.7 (dd, 1JRhP = 155 Hz, 2JPP = 30 Hz, PPh3),
Reactions of 9−12, 16, and 17 with H2(g). In a typical reaction,
around 10 mg (0.016 mmol) of the closo-rhodathiaboranes was placed
in a screw-capped NMR tube and dissolved in 0.6 mL of CD2Cl2. The
NMR tube was cooled in liquid nitrogen, evacuated, and then filled
with H2. The samples were monitored at different intervals overnight
by 1H NMR spectroscopy, without evidence of changes in the spectra.
Reactions of 4−6 with C2H4. In a Schlenk tube, around 10 mg of
the hydridorhodathiaboranes 4−6 was dissolved in CH2Cl2 (∼8 mL).
A balloon filled with C2H4 was attached to the Schlenk tube; the
system was then cooled in liquid nitrogen and evacuated under
vacuum. The hydrorhodathiaborane solutions were exposed to the
ethylene atmosphere created upon the opening of the balloon, and the
reaction was stirred overnight at room temperature. In the case of
compound 4, the NMR data demonstrated that there was no reaction
with the olefin. In the case of 6, however, there was formation of the
closo-derivative 10. In the same conditions, a mixture of the isomers 5a
and 5b leads to the formation of the bis-PMe2Ph derivative 11.
Compound 6 with Ethylene. Alternatively, the reaction with
compound 6 (5.5 mg, 0.0083 mmol, in 0.4 mL of CD2Cl2) was carried
out in a screw-capped NMR tube, which was exposed to 1.5 bar of
1
2
−6.2 (dd, JRhP = 138 Hz, JPP = 32 Hz, PMe2Ph).
Synthesis of [1,1-(PPh3)(PMe3)-3-(Py)-closo-1,2-RhSB9H8]
(10). A 10 mg (0.021 mmol) sample of 6, placed in an Schlenk
tube, was dissolved in CH2Cl2 and heated to reflux temperature under
an atmosphere of argon for 3 h. Solvent was reduced in volume, and
hexane added to form a yellow precipitate, which was separated by
decantation, washed with hexane, and dried under vacuum. The
resulting yellow product was characterized as 10 (6.9 mg, 0.015 mmol,
69%). 31P{1H} NMR (161 MHz; CDCl3; 300 K) ordered as δP [DFT-
1
2
calcd δP]: 49.2 [67.7] (dd, JRh−P = 157 Hz, JP−P = 29 Hz, PPh3),
1
−15.4 [−3.2] (1P, dd, JRh−P = 146 Hz, PMe3). Anal. Calcd for
C26H37B9NP2RhS: C, 47.47; H, 5.67; N, 2.13; S, 4.87. Found: C,
47.07; H, 5.88; N, 2.39; S, 3.77.
Synthesis of [1,1-(PMe2Ph)2-3-(Py)-closo-1,2-RhSB9H8] (11). A
Schlenk tube was charged with the mixture of isomers 5 (10.1 mg,
0.012 mmoL). The mixture of hydridorhodathiaboranes was dissolved
in 10 mL of CH2Cl2, and then PMe2Ph (2.5 μL, 0.024 mmol) was
added under an argon atmosphere. The resulting solution was heated
at reflux in an argon atmosphere for 12 h. Solvent was evaporated,
and the red residue crystallized in CH2Cl2/hexane to give 6.0 mg
of compound 11 (0.010 mmol, 84%). 31P{1H} NMR (161 MHz;
1
ethylene. After 3 days, the H NMR spectrum at room temperature
exhibits two broad multiplets at +2.27 and +2.05 ppm, which
correspond to the previously reported ethylene-ligated rhodathiabo-
rane, 13. In addition, there are two broad multiplets of higher intensity
at +2.38 and +2.16 ppm, which can be assigned to the rhodium-bound
ethylene ligand of the PMe3-ligated analogue [1,1-(PMe3)(η2-C2H4)-
3-(Py)-closo-1,2-RhSB9H8] (15). At lower temperatures, the two
1
CD2Cl3; 300 K): δ 2.9 (d, JRh−P = 140 Hz). Anal. Calcd for
C21H35B9NP2RhS: C, 42.34; H, 5.92; N, 2.35; S, 5.38. Found: C,
41.88; H, 5.87; N, 2.18; S, 4.70. MS m/z (MALDI): 596 (M+, isotope
envelope; C21H35B9NP2RhS requires 596).
2994
dx.doi.org/10.1021/om2009205 | Organometallics 2012, 31, 2986−2995

Organometallics
Article
signals of the ethylene ligand in 15 split into four peaks, which coalesce
in pairs over the range 196−300 K to give a free energy for the
REFERENCES
■
(1) Ferguson, G.; Jennings, M. C.; Lough, A. J.; Coughlan, S.;
Spalding, T. R.; Kennedy, J. D.; Fontaine, X. L. R.; Stibr, B. J. Chem.
Soc., Chem. Commun. 1990, 891.
rotation of the ethylene ligand of ΔG273 = 53 kJ/mol (vs ΔG252
=
49 kJ/mol for 13;2b see Supporting Information, Figure S17).
[1,1-(PMe3)(η2-C2H4)-3-(Py)-closo-1,2-RhSB9H8] (15). NMR data
were assigned from a mixture that contains 10 (major), 13 (minor),
and 15 (medium). 11B{1H} NMR (160 MHz; CD2Cl2; 298 K): δ 54.5
(1B, s, B-py), 24.9 (1B, b, BH), 0.9 (1B, b, BH), −1.1 (1B, b, BH),
−20.5 (1B, b, BH), −27.6 (1B, b, BH), 30.5 (1B, b, BH), −31.7 (1B, b,
́
(2) (a) Alvarez, A.; Macías, R.; Fabra, M. J.; Lahoz, F. J.; Oro, L. A.
J. Am. Chem. Soc. 2008, 130, 2148. (b) Alvarez, A.; Macias, R.; Bould, J.;
Fabra, M. J.; Lahoz, F. J.; Oro, L. A. J. Am. Chem. Soc. 2008, 130,
11455. (c) Alvarez, A.; Macias, R.; Bould, J.; Cunchillos, C.; Lahoz,
F. J.; Oro, L. A. Chem.Eur. J. 2009, 15, 5428.
1
BH). H{11B} NMR (300 MHz; CD2Cl2; 298K): δ 4.37 (1H, s, BH),
(3) Calvo, B.; Kess, M.; Macias, R.; Cunchillos, C.; Lahoz, F. J.;
Kennedy, J. D.; Oro, L. A. Dalton Trans. 2011, 40, 6555.
+1.18 (3H, d, 2JH−P = 10.2 Hz, P-CH3), 0.46 (1H, s, BH), 0.01 (1B, s,
BH), −0.57 (1B, s, BH). 31P{1H} NMR (161 MHz; CD2Cl2; 300 K):
(4) (a) Wade, K. Inorg. Nucl. Chem. Lett. 1972, 8, 559. (b) Wade, K.
Adv. Inorg. Chem. Radiochem. 1976, 18, 1. (c) Mingos, D. M. P. Nature
(London), Phys. Sci. 1972, 99, 236. (d) Mingos, D. M. P.; Wales, D. J.
Introduction to Cluster Chemistry; Prentice Hall: London, 1990.
(5) (a) Kennedy, J. D. In The Borane-Carborane-Carbocation
Continuum; Casanova, J., Ed.; Wiley: New York, 1998. (b) Macías,
1
δ −5.1 (1P, d, JP−Rh = 134 Hz, PMe3).
1
Regarding the 11B and H NMR data, the assignments are tentative
since there is overlapping of signals with the other two rhodathiabor-
anes, 10 and 13. The 31P{1H} data, however, can be unambiguously
assigned.
In a higher scale, 50.2 mg (0.059 mmol) of 6 was placed in a screw-
capped Schlenk tube and dissolved in 25 mL of CH2Cl2. The resulting
solution was frozen at the liquid nitrogen temperature, and the system
evacuated under vacuum and then exposed to 1.5 bar of ethylene. After
7 days of stirring at room temperature, the solvent was evaporated to
dryness and applied to preparative TLC plates. The chromatogram was
developed using CH2Cl2/hexane (3:1) as mobile phase to give broad
yellow (Rf = 0.33) and red (Rf = 0.13) bands. The former was a mixture
of compounds 13 (1.71 mg, 0.0028 mmol, 5%) and 15 (4.28 mg,
0.010 mmol, 17%) in a 1:4 ratio, whereas the second was compound 10
(12 mg, 0.018 mmol, 31%). It was found that compounds 13 and 15
decompose largely on the plates; thus, a second preparative TLC plate
resulted in the loss of both species.
̌
R.; Bould, J.; Holub, J.; Kennedy, J. D.; Stibr, B.; Thornton-Pett, M.
Dalton Trans. 2007, 2885. (c) Murphy, M. P.; Spalding, T. R.; Cowey,
C.; Kennedy, J. D.; Thornton-Pett, M.; Holub, J. J. Organomet. Chem.
1998, 550, 151.
(6) Ferguson, G.; Lough, A. L.; Coughlan, S.; Spalding, T. R. Acta
Crystallogr., Sect. C: Cryst. Struct. Commun. 1992, C48, 440.
(7) Beall, H.; Bushweller, C. H.; Dewkett, W. J.; Grace, M. J. Am.
Chem. Soc. 1970, 92, 3484.
(8) Macías, R.; Rath, N. P.; Barton, L. Organometallics 1999, 18,
3637.
(9) Otwinowski, Z.; Minor, W. In Methods in Enzymology; Carter,
J. R. M. S. C. W., Ed.; Academic Press: New York, 1997; Vol. 276.
(10) Blessing, R. H. Acta Crystallogr. 1995, A51, 33.
In the same conditions, using screw-capped NMR and Schlenk
tubes and 1.5 bar of ethylene, the mixture of isomers 5 afforded
compound 11 as the major component and small amounts of the
ethylene-ligated cluster 13. However, we did not detect the ethylene-
ligated PMe2Ph counterpart.
(11) Sheldrick, G. M. SHELXS-86; University of Gottingen:
Gottingen, Germany, 1997.
(12) Sheldrick, G. M. Acta Crystallogr. 2008, A64, 112.
(13) Farrugia, L. J. J. Appl. Crystallogr. 1999, 32, 837.
̈
̈
(14) Spek, A. L. J. Appl. Crystallogr. 2003, 36, 7.
(15) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.;
Robb, M. A.; Cheeseman, J. R.; Montgomery, J. A., Jr.; Vreven, T.;
Kudin, K. N.; Burant, J. C.; J. M.Millam; Iyengar, S. S.; Tomasi, J.;
Barone, V.; Mennucci, B.; Cossi, M.; Scalmani, G.; Rega, N.;
Petersson, G. A.; Nakatsuji, H.; Hada, M.; Ehara, M.; Toyota, K.;
Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao,
O.; Nakai, H.; Klene, M.; Li, X.; Knox, J. E.; Hratchian, H. P.; Cross, J.
B.; Bakken, V.; Adamo, C.; Jaramillo, J.; ; Gomperts, R.; Stratmann, R.
E.; Yazyev, O.; Austin, A. J.; Cammi, R.; Pomelli, C.; Ochterski, J.;
Ayala, P. Y.; Morokuma, K.; Voth, G. A.; Salvador, P.; Dannenberg,
J. J.; Zakrzewski, V. G.; Dapprich, S.; Daniels, A. D.; Strain, M. C.;
Farkas, O.; Malick, D. K.; Rabuck, A. D.; Raghavachari, K.; Foresman,
J. B.; Ortiz, J. V.; Cui, Q.; Baboul, A. G.; Clifford, S.; Cioslowski, J.;
Stefanov, B. B.; Liu, G.; Liashenko, A.; Piskorz, P.; Komaromi, I.;
L.Martin, R.; Fox, D. J.; Keith, T.; Al-Laham, M. A.; Peng, C. Y.;
Nanayakkara, A.; Challacombe, M.; Gill, P. M. W.; Johnson, B. G.;
Chen, W.; Wong, M. W.; Gonzalez, C.; Pople, J. A.Gaussian 03;
Gaussian, Inc.: Wallingford, CT, 2004.
ASSOCIATED CONTENT
■
S
* Supporting Information
31P{1H} NMR spectra at different temperatures of compounds
2 and 6; NMR spectra of the reactions of 2 with PMe3,
PMe2Ph, and PMePh2 at different temperatures; series of
1H−31P HSQC, 31P{1H}, and 1H NMR spectra of samples that
contain mixtures of 5a, 5b, and 7; tables of the kinetics studies;
the Eyring plot; ORTEP-type picture for compound 11;
experimental and DFT-calculated coordinates and energy for 6
and 10; CIF files for 4, 10, and 11. This material is available
free of charge via the Internet at http://pubs.acs.org.
AUTHOR INFORMATION
■
Corresponding Author
́
*E-mail: rmacias@unizar.es.
(16) Morse, P. M.; Spencer, M. D.; Wilson, S. R.; Girolami, G. S.
Organometallics 1994, 13, 1646.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
We acknowledge the Spanish Ministry of Science and
Innovation (CTQ2009-10132, CONSOLIDER INGENIO,
CSD2009-00050, MULTICAT, and CSD2006-0015, Crystal-
lization Factory) for support of this work. B.C. thanks the
́ ́
“Diputacion General de Aragon” for a predoctoral scholarship.
P.G.O. acknowledges financial support from the CSIC JAE-
DOC program, contract co-funded by the ESF. We also thank
ESRF BM16 beamline staff for their support of data acquisition
of 4.
2995
dx.doi.org/10.1021/om2009205 | Organometallics 2012, 31, 2986−2995