Treatment of [Ru(PPh3)(C6H4PPh2)2H][Li(THF)2] with AlMe2Cl and SnMe3Cl leads to elimination of LiCl and CH4 and formation of the heterobimetallic complexes [Ru(C6H4PPh2)2{PPh2C6H4AlMe(THF)}H] 5 and [Ru(PPh3)(C6H4PPh2)(PPh2C6H4SnMe2)] 6, respectively. The pathways to 5 and 6 have been probed by variable temperature NMR studies, together with input from DFT calculations. Complete reaction of H2 occurs with 5 at 60 °C and with 6 at room temperature to yield the spectroscopically characterized trihydride complexes [Ru(PPh2)2{PPh2C6H4AlMe}H3] 7 and [Ru(PPh2)2{PPh2C6H4SnMe2}H3] 8. In the presence of CO, 6 forms the acylated phosphine complex, [Ru(CO)2(C(O)C6H4PPh2)(PPh2C6H4SnMe2)] 9, through a series of intermediates that were identified by NMR spectroscopy in conjunction with 13CO labeling. Complex 6 undergoes addition and substitution reactions with the N-heterocyclic carbene 1,3,4,5-tetramethylimidazol-2-ylidene (IMe4) to give [Ru(IMe4)2(PPh2C6H4)(PPh2C6H4SnMe2)] 10, which converted via rare N-Me group C-H activation to [Ru(IMe4)(PPh3)(IMe4)'(PPh2C6H4SnMe2)] 11 upon heating at 60 °C and to a mixture of [Ru(IMe4)2(IMe4)'(PPh2C6H4SnMe2)] 12 and [Ru(PPh3)(PPh2C6H4)(IMe4-SnMe2)'] 13 at 120 °C.
Treatment of [Ru(PPh3)(C6H4PPh2)2H][Li(THF)2] with AlMe2Cl and SnMe3Cl leads to elimination of LiCl and CH4 and formation of the heterobimetallic complexes [Ru(C6H4PPh2)2{PPh2C6H4AlMe(THF)}H] 5 and [Ru(PPh3)(C6H4PPh2)(PPh2C6H4SnMe2)] 6, respectively. The pathways to 5 and 6 have been probed by variable temperature NMR studies, together with input from DFT calculations. Complete reaction of H2 occurs with 5 at 60 °C and with 6 at room temperature to yield the spectroscopically characterized trihydride complexes [Ru(PPh2)2{PPh2C6H4AlMe}H3] 7 and [Ru(PPh2)2{PPh2C6H4SnMe2}H3] 8. In the presence of CO, 6 forms the acylated phosphine complex, [Ru(CO)2(C(O)C6H4PPh2)(PPh2C6H4SnMe2)] 9, through a series of intermediates that were identified by NMR spectroscopy in conjunction with 13CO labeling. Complex 6 undergoes addition and substitution reactions with the N-heterocyclic carbene 1,3,4,5-tetramethylimidazol-2-ylidene (IMe4) to give [Ru(IMe4)2(PPh2C6H4)(PPh2C6H4SnMe2)] 10, which converted via rare N-Me group C-H activation to [Ru(IMe4)(PPh3)(IMe4)'(PPh2C6H4SnMe2)] 11 upon heating at 60 °C and to a mixture of [Ru(IMe4)2(IMe4)'(PPh2C6H4SnMe2)] 12 and [Ru(PPh3)(PPh2C6H4)(IMe4-SnMe2)'] 13 at 120 °C.
Heterobimetallic (HBM) complexes featuring
a transition metal (TM)
center in combination with a Lewis acidic s- or p-block metal (M′) continue to be the subject of
considerable interest, primarily due to the ability of such species
to bring about small molecule activation chemistry.[1,2] The
TM-Zn, -Ga and -Al complexes shown in Scheme represent three recent examples in which
[TM-M′] HBM complexes have been employed to bring about not
only small molecule activation but also a subsequent catalytic functionalization
step.
Scheme 1
Examples of Catalytic Transformations Mediated by [TM-M′]
Heterobimetallic Complexes
A commonly employed preparative route to [TM-M′]
HBM complexes
involves the reaction of a TM-hydride precursor with a Lewis acidic
metal alkyl reagent to give a [TM-M′] product following elimination
of an alkane.[3] In a recent study,[4] we employed such a reaction of [Ru(PPh3)3HCl] with LiMe, MgMe2, and ZnMe2 to give the bis-cyclometalated complexes[5] [Ru(PPh3)(C6H4PPh2)2H][M′] (M′ = Li(THF)21, MgMe(THF)22, and ZnMe 3)
shown in Scheme .
A combination of X-ray crystallography and DFT calculations showed
that the level of interaction between Ru–H and M′ increased
in the order of 1 < 2 < 3 such that 1 and 2 were best considered
as ruthenate anions with a group 1 or 2 countercation, whereas Zn
compound 3 exhibited far more covalent character. As
a result, the latter proved susceptible to reductive elimination of
the hydride ligand onto one of the metalated phosphines, to yield
the “dual unsaturated” isomer, [Ru(PPh3)2(C6H4PPh2)(ZnMe)] 4, which although only present in ca. 2%, allowed 3 to
react with H2 at −40 °C, ca. 100 °C lower
than the temperature required with either 1 and 2.
Scheme 2
Synthesis of [Ru-M′] Complexes 1 (M′ =
Li(THF)2), 2 (M′ = MgMe(THF)2) and 3 (M′ = ZnMe) and Equilibrium of the Latter
with 4
Prompted by the enhanced reactivity of the [Ru–Zn]
complex,
we have extended our studies to [Ru-M′] complexes in which
M′ = Al and Sn, on the basis that they would also exhibit strong
covalent interactions with the Ru center. We now describe the synthesis
and reactivity of the [Ru–Al] and [Ru–Sn] heterobimetallic
complexes [Ru(C6H4PPh2)2{PPh2C6H4AlMe(THF)}H] 5 and [Ru(PPh3)(C6H4PPh2)(PPh2C6H4SnMe2)] 6.
Results and Discussion
Synthesis and Characterization of [Ru(C6H4PPh2)2{PPh2C6H4AlMe(THF)}H] and [Ru(PPh3)(C6H4PPh2)(PPh2C6H4SnMe2)]
We showed previously that the [Ru–Li] salt 1 was a convenient precursor to both 2 and 3 upon treatment with MgMeCl and ZnMeCl, respectively (Scheme ), thanks to the
relative ease of removal of the LiCl byproduct.[4] Heating 1 with AlMe2Cl at 60 °C
led to full conversion through to yellow [Ru(C6H4PPh2)2{PPh2C6H4AlMe(THF)}H] 5, which was isolated in 69% yield, whereas
SnMe3Cl reacted with 1 at room temperature
to generate deep-blue [Ru(PPh3)(C6H4PPh2)(PPh2C6H4SnMe2)] 6 in a near quantitative amount (Scheme ). Comparison of Schemes and 3 shows that while both reactions were indeed accompanied by
loss of LiCl, the availability of additional M′-Me groups on
moving from ZnMeCl to AlMe2Cl and SnMe3Cl allowed
elimination of an extra molecule of CH4, resulting in metalation
of a further phosphine ligand (vide infra).[6]
Scheme 3
Synthesis of [Ru(C6H4PPh2)2{PPh2C6H4AlMe(THF)}H] 5 and [Ru(PPh3)(C6H4PPh2)(PPh2C6H4SnMe2)] 6
The X-ray crystal structure of 5 (Figure , Table ) showed a Ru atom
at the center of a highly
distorted octahedral arrangement of ligands (e.g., P(2)–Ru(1)–C(38)
= 150.40(6)°), including three metalated phosphines[7] in a mer-arrangement (cf. fac-RuP3 geometry of both 1 and 2). The ruthenium and aluminum centers formed part of a 6-membered
ring dimetalacycle in which the Al was attached to both a bridging
hydride ligand (located and refined with an Al(1)–H(1) distance
of 1.83(3) Å) and a phosphine phenyl group (Al(1)–C(2)
= 1.991(3) Å).[8,9] The structure of 6 (Figure , Table ) contained a 5-coordinate
Ru center (thus resembling 4) with a direct Ru–Sn
interaction (Ru(1)–Sn(1) = 2.5686(2) Å).[10] Phosphine metalation occurred onto Sn to generate a cyclostannylated
phosphine ligand, which bridges across the basal and axial sites of
the square pyramidal Ru complex. A related, albeit coordinatively
saturated, osmium derivative, [Os(PPh3)(CO)(C6H4PPh2)(PPh2C6H4SnMe2)], has been reported by Roper to form as a minor
product upon refluxing [Os(PPh3)2(CO)(SnMe3)Cl] with PPh3.[11−13]
Figure 1
Molecular structures
of (left) 5 and (right) 6. Ellipsoids at
30% level; all H atoms, except Ru–H–Al,
omitted for clarity. In 5, the solvent and the minor
disordered component have also been omitted for clarity.
Table 1
Selected Bond Lengths (Å) and
Angles (deg) in 5 and 6
5
6
Ru-PPh3
-
2.3504(6)
Ru-PPh2(C6H4Ru)
2.3854(6), 2.3555(6)
2.3616(7)
Ru-PPh2(C6H4Al)
2.3202(6)
-
Ru-PPh2(C6H4Sn)
-
2.3245(6)
Ru···E
(E = Al) 2.5911(7)
(E = Sn) 2.5686(2)
trans–P–Ru-P
166.16(2)
158.22(2)
Molecular structures
of (left) 5 and (right) 6. Ellipsoids at
30% level; all H atoms, except Ru–H–Al,
omitted for clarity. In 5, the solvent and the minor
disordered component have also been omitted for clarity.The solution NMR spectra of 5 and 6 (Figures S1–S7) were consistent
with their
solid-state structures.[14] Thus, the 1H NMR spectrum of 5 showed a broad triplet of
doublets Ru-H-Al signal at δ −6.29[15] with small 2JHP splittings (12 and 6 Hz) to the three cis-phosphorus nuclei.
In the 31P{1H} NMR spectrum, there were three
doublets, with those at δ 70 and δ −15 assigned
to the phosphines metalated onto Al and Ru respectively, based on
their mutually large (trans) coupling of 266 Hz, as well as the established
upfield shift associated with phosphines metalated onto a TM center
and downfield shift arising from 6-membered ring phosphine chelates.[16,17] The 1H NMR spectrum of 6 yielded very little
in the way of diagnostic information, but the presence of high (δ
75) and low (δ −29) frequency 31P{1H} NMR signals with a large mutual 2JPP splitting of 240 Hz was consistent with the presence
of cyclostannylated and cycloruthenated ligands respectively.
Pathways to Formation of 5 and 6
The very different structures of 5 and 6 led us to investigate their pathways to formation using variable
temperature NMR spectroscopy. Introduction of a frozen, yellow-orange
THF-d8 solution of 1 and
AlMe2Cl into a precooled (193 K) NMR probe revealed the
rapid formation of a 1:1 ratio of two intermediates, assigned as the
structures and shown in Scheme a. Characterization of these species (Figures S8–S12), as well as the higher temperature
intermediates and (Scheme a), was based on (i) the number of 31P NMR resonances
and their relative chemical shifts,[16,18] (ii) the relative
magnitudes of 2JPP/2JHP couplings, and (iii) 31P–1H HMQC connectivities.
Scheme 4
Proposed Structures
of Intermediates in the Formation of 5 and 6 (Based on Low-Temperature NMR Studies) Are Shown
in (a–c), with a Proposed Pathway from to 6 Illustrated in (d)
The * on in part (a) denotes uncertainty as to whether
THF is or is
not bound on Al.
Proposed Structures
of Intermediates in the Formation of 5 and 6 (Based on Low-Temperature NMR Studies) Are Shown
in (a–c), with a Proposed Pathway from to 6 Illustrated in (d)
The * on in part (a) denotes uncertainty as to whether
THF is or is
not bound on Al.Intermediate results from substitution
of the Li(THF)2 moiety in 1 by AlMe2 and was assigned based on the retention of a fac-RuP3 arrangement, comparable 31P chemical
shifts to those of 1 (especially the two low frequency
resonances for the metalated phosphines) and the presence of a low
frequency hydride resonance (δ −10.50 cf. δ −9.62
in 1),[4] attributed to the
bridging Ru-H-Al interaction. Intermediate showed a broad 1H singlet at
δ −2.53, in a 3:1 ratio with doublet of doublet of doublets
Ru-H-Al signal at δ −13.92, suggestive
of it being an isomer with a bridging Ru-Me-Al group
in place of the Ru-CH-Al bridge in . The replacement of one of the low-frequency 31P signals in by a new high
frequency signal for supported the
presence of a phosphine metalated onto just Al.Isomerization
of and to the mer- was
seen at 233 K, while further warming (to 273 K) generated a deeper-red
colored solution, consistent with formation of a coordinatively unsaturated
isomer . This showed just a single
AlMe proton resonance (cf. two resonances in ), consistent with a structure arising
out of the combination of the Ru-H-Al and one of
the two AlMe groups in and subsequent reductive elimination of methane. After 1
h at 273 K, - had been fully consumed, and represented ca. 80% of the reaction mixture. A final color change
from red to orange was observed at 298 K, concomitant with the formation
of the final product 5 through metalation of the third
phosphine ligand. After ca. 40 min at 298 K, 5 comprised
ca. 65% of all species in solution.[19]An analogous study of the formation of 6 (Scheme b; Figures S13–S16) failed to show any spectroscopic evidence
for the comparable initial substitution product (Scheme c).
This may imply that reductive elimination of Ru–H onto RuC6H4PPh2 in such a species is very fast,
supporting further the analogous behavior of [Ru–Sn] and [Ru–Zn]
species. Only a single (deep-blue) intermediate was observed between
168 and 273 K, which we propose is [Ru(PPh3)2(C6H4PPh2)SnMe3] () based on (i) the presence of only a single
SnMe resonance in the 1H NMR spectrum
(cf. two signals in 6 for the diastereotopic Me groups)
and (ii) the observation of a single low frequency (δ −29) 31P{1H} NMR signal for a cycloruthenated phosphine,
together with two “medium” frequency signals (δ
49, 41–cf. ) arising from two
PPh3 ligands. 6 began to appear above 273
K (Figures S13–S16). Following studies
by Wada[20] and Roper,[11b] a possible pathway for the transformation of to 6 involves Me group transfer from
Sn to Ru and generation of a transient Ru stannylene intermediate,
(Scheme d) that could
generate the final cyclostannylated phosphine containing complex 6 through attack on a phosphine phenyl C–H bond, followed
by elimination of methane.Determination of the free energies
of , , 5, and 6 by density functional theory
(DFT) calculations (BP86-D3BJ(C6H6)/BS2//BP86/BS1)
were in agreement with the experimental
findings. Thus, the free energy of was computed to be higher than that of 5 (Scheme ; see also Supporting Information for further details).
Unsurprisingly, coordination of THF stabilized both structures. 6 was calculated to be more stable than , as well as 7.8 kcal/mol more stable than 5Sn, the Sn analogue of [Ru–Al] complex 5 (Scheme ).
Scheme 5
Free Energies (BP86-D3BJ(C6H6)/BS2//BP86/BS1)
Relative to 5 (kcal/mol)
Scheme 6
Free Energies (BP86-D3BJ(C6H6)/BS2//BP86/BS1)
Relative to (kcal/mol)
Reactivity of 5 and 6 with H2
Complex 5 showed a similar reluctance to 1 and 2 in reacting with H2 only at
elevated temperature (60 °C) to yield a single product, which
was characterized as the trihydride species [Ru(PPh2)2{PPh2C6H4AlMe}H3] 7 (Scheme ) based on NMR spectroscopy (Figures S17–S22).[21] We were unable to crystallize the
product which decomposed in the absence of a H2 atmosphere
to a mixture of species, two of which were identified as [Ru(PPh3)3(η2-H2)H2] and [Ru(PPh3)4H2].[22] The fate of the aluminum was not determined.
Scheme 7
Proposed Structure of 7 from Reaction of 5 with H2 (1 atm, 60°C)
The 31P{1H} NMR spectrum
of 7 exhibited three signals with JPP values
indicative of a mer-RuP3 geometry. Based
on the findings for 5, the high frequency (δ 74,
doublet of doublets) signal was attributed to the Al-metalated phosphine,
with a doublet of doublets at δ 62 and a triplet at δ
59 arising from the two PPh3 ligands. The low frequency
region of the room temperature 1H NMR spectrum of 7 showed an AlMe resonance at δ −0.39,
which integrated to 3 relative to three hydride signals (each of relative
integral 1) at δ −8.46 (broad doublet), −8.72
(triplet of doublets) and −11.07 (doublet of triplets).[23,24] The lowest frequency hydride signal was assigned to the bridging
hydride Hc (Scheme ) based on the presence of (i) a NOESY peak to the AlMe resonance and (ii) a 54 Hz 2JHP doublet splitting, indicative of a pseudo-trans PPh3 ligand. The well-resolved appearance of this signal suggests
it is more closely associated with Ru than quadrupolar Al. The highest
frequency hydride signal was assigned to Hb based on the
presence of a NOESY peak to the AlMe signal. The
signal stayed broad between 223 and 337 K (Figure S19), consistent with it being associated more with Al (i.e.,
Ru···H-Al).[25] The magnitude of the 2JHP splittings (28 and 14 Hz) on the resonance at δ −8.72
(Ha) support it being cis to three phosphine ligands.[26] No 2JHH coupling was observed on any of the hydride resonances in the 1H{31P} NMR spectrum.[27]NMR spectra of the analogous reaction with D2 showed
that the three hydride signals were present in the proton NMR spectrum,
but all in an integral ratio of <1 relative to the AlMe resonance. This, together with broad 31P resonances,
indicates that both H2/D2 addition as well as
phosphine cyclometalation must be reversible, allowing H/D exchange
to take place into the ortho-positions of PPh3 ligands.In contrast to 5, [Ru–Sn] complex 6 showed behavior that aligned with [Ru–Zn] complex 3 in reacting with H2 at room temperature, to yield what
we assign as the trihydride complex [Ru(PPh2)2{PPh2C6H4SnMe2}H3] 8 (Scheme ). A gradual color change from a blue to a colorless/pale-yellow
was observed when H2 was allowed to diffuse slowly through
a sample of 6, although if H2 was added and
the sample shaken vigorously, a colorless precipitate (which failed
to redissolve in most common solvents) formed almost instantaneously.
The IR spectra of the precipitate and material from solution were
identical,[28] implying that 8 is the product in both cases. We assume that 8 sits
right on the edge of solubility and that shaking results in precipitation.
Scheme 8
Proposed Structure of 8 from the Room Temperature Reaction
of 6 with H2 (1 atm)
The 1H NMR spectrum of a homogeneous
solution of 8 formed upon slow diffusion of H2 exhibited a
single SnMe2 resonance (cf. two different
SnMe resonances for 6) of integral 6
at room temperature, along with a single, broad (fwhm = 36 Hz) hydride
signal at δ −7.6 of relative integral 3 (Figures S23–S27) with a 1JHSn coupling of 184 Hz. The magnitude is suggestive
of some degree of interaction between Ru-H and Sn
centers,[29,30] although the hydride T1 value of 390 ms (400 MHz, 298 K) would exclude any appreciable
nonclassical behavior. Variable temperature NMR measurements were
consistent with 8 being fluxional in solution. Thus,
cooling to 223 K (THF) only broadened the hydride signal, whereas
warming to 332 K resolved it into a single doublet of triplets, with JHP values of 16 and 7 Hz, indicative of the
hydride ligands being cis to all three phosphorus nuclei. The fluxionality
was mirrored in the 31P{1H} NMR spectrum, which
comprised at low temperature of a triplet (δ 85, cyclostannylated
phosphine), together with a broad singlet (δ 56, two PPh3 ligands) that resolved into a doublet upon warming to (or
above) room temperature. The mutual 2JPP splitting of 98 Hz is in-between the values typically
associated with trans- and cis–P–Ru−P arrangements.[31]The ease with which 8 precipitated
thwarted all attempts
to generate single crystals suitable for X-ray crystallography, even
via a solid-state transformation.[32]Scheme shows a structure
for 8 (Table S2) that is based
on other group 8 metal derivatives [Ru(PR3)3(ER3′)H3] (ER3′ =
SiR3′, SnR3′),[29a,30,33] which all feature a common tetrahedral arrangement
of Si/Sn and 3PR3 units with hydride ligands capping the
Si/Sn(PR3)2 faces.
Reactivity of 6 with Lewis Bases
CO
Additional studies of small molecule reactivity
focused on [Ru–Sn] precursor 6. As shown in Scheme , both addition and
insertion of CO took place when 6 was heated under 1
atm CO at 80 °C, to ultimately form the acylated phosphine complex
[Ru(CO)2(C(O)C6H4PPh2)(PPh2C6H4SnMe2)] 9, which could be isolated in 60% yield. Typically, acylated phosphine
ligands are generated by oxidative addition of phosphino substituted
aldehydes,[34] rather than by CO insertion
into a M–aryl bond,[35] although the
latter route does have precedence with ruthenium.[35b]
Scheme 9
Spectroscopically Detected Intermediates – in the
Formation
of 9
The X-ray structure (Figure ) of 9 revealed an octahedral
coordination sphere
with a trans-arrangement of the P atoms of the stannylated and acylated
phosphines, leaving the -SnMe2 and -C(O)(aryl) groups trans
to the two carbonyl ligands. As a result of this geometry, the Ru–Sn
(2.6879(2) Å) and Ru–P (2.3869(6) Å) distances of
the stannylated phosphine were significantly longer than in 6. The Ru–C(O) distance (2.129(2) Å) was comparable
to that in [Ru(PPh3)(CO)2(C(O)C6H4PPh2)H] (2.110(1) Å).[34c] In the 31P NMR spectrum, there was only a minor
change in the chemical shift of the stannylated phosphine relative
to 6, whereas the phosphine metalated onto Ru moved ca.
100 ppm to higher frequency as a result of CO insertion (Figures S32–S35).[35b]
Figure 2
Molecular structure of 9. Ellipsoids are shown at
30% level with all H atoms omitted for clarity. Selected bond lengths
(Å) and angles (deg): Ru1–P1 2.3214(6), Ru1–P2
2.3869(6), Ru1–C1 1.920(2), Ru1–C2 1.942(2), Ru1–C3
2.129(2), Ru1–Sn1 2.6879(2), P1–Ru1–P2 170.79(2).
Molecular structure of 9. Ellipsoids are shown at
30% level with all H atoms omitted for clarity. Selected bond lengths
(Å) and angles (deg): Ru1–P1 2.3214(6), Ru1–P2
2.3869(6), Ru1–C1 1.920(2), Ru1–C2 1.942(2), Ru1–C3
2.129(2), Ru1–Sn1 2.6879(2), P1–Ru1–P2 170.79(2).In situ NMR measurements, in conjunction with 13CO labeling,
revealed a series of intermediates on the pathway to 9 (Scheme ; Figures S37–S42). Thus, shaking 6 with 13CO (1 atm) brought about an instantaneous
change in color at room temperature from blue to yellow, concomitant
with formation of the 18-electron CO addition species . This was identified by the appearance of three
doublet of doublet of doublet 31P resonances, each with
a cis-sized 2JPC coupling (8–11
Hz) to a single 13CO ligand, which resonated in the 13C{1H} NMR spectrum at δ 207 as a doublet
of triplets. Isomerization of occurred
overnight
at room temperature to yield , which
exhibited one metalated phosphorus signal with a much greater 2JPSn splitting (970 Hz vs 180
Hz), consistent with a change in orientation to trans P–Ru–SnMe2. There was also a small amount of the dicarbonyl species (identified on the basis of two multiply
coupled high frequency signals in the 13C{1H}
NMR spectrum), which increased in intensity upon heating at 80 °C,
leaving it as the main product in solution after 1 h. Further heating
converted to the final product 9. No signals attributable to (Scheme ), a “Ru-SnPhos”
analogue of [Ru(ZnPhos)(CO)3], which we have shown to be
the product formed when a mixture of [Ru(PPh3)3HCl] and LiCH2TMS/ZnMe2 was heated under CO,[36] were observed at any point in the overall reaction.
1,3,4,5-Tetramethylimidazol-2-ylidene (IMe4)
Treatment of 6 with ca. 3 equiv of the N-heterocyclic
carbene 1,3,4,5-tetramethylimidazol-2-ylidene (IMe4) led
to full consumption of the starting material over the course of ca.
1 h to form the coordinatively saturated product [Ru(IMe4)2(PPh2C6H4)(PPh2C6H4SnMe2)] 10 (Scheme ), which
was isolated as an orange microcrystalline solid in 55% yield. The 31P{1H} NMR spectrum of 10 displayed
doublets at both high (δ 69) and low (δ −36) frequency,
consistent with retention of both the cyclostannylated and cycloruthenated
phosphines, although the magnitude of 2JPP (18 Hz) now implied they were in a cis-configuration
(Figures S43–S48). Two inequivalent
IMe4 ligands were evident from the appearance of four NMe and four NCMe resonances in the 1H NMR spectrum and the presence of two 13C carbenic
resonances (δ 191, 2JCP = 86 and 16 Hz; δ 200, 2JCP = 8 and 2 Hz).
Scheme 10
Synthesis of bis-IMe4 Complex 10 and Formation
of 11–13 upon Heating
As shown in Figure and Table , X-ray
crystallography revealed that the Ru center in 10 was
significantly distorted from regular octahedral. Accommodation of
the two IMe4 ligands caused a reduction (relative to 6) in the bite angles of both the cyclostannylated (84.371(17)°
to 80.092(15)°) and cycloruthenated (67.93(7)° to 66.42(6)°
respectively) phosphines, as well as an acute trans–P–Ru−Sn angle (155.047(15)°). Incorporation
of a (phosphine) ligand trans to tin increased the Ru–Sn distance
to 2.6345(2) Å from the value of 2.5686(2) Å in 6.
Figure 3
Molecular structures of one of the molecules in the asymmetric
unit of (left) 10 and (right) 11. Ellipsoids
at 30% level. All H atoms and solvent have been omitted for clarity
in 10, while all hydrogens, with the exception of those
attached to C1, have been omitted in 11.
Table 2
Selected Bond Lengths (Å) and
Angles (deg) in the IMe4 Complexes 10–13
10
11
12
13
Ru-CIMe4
2.121(2), 2.132(2)
2.119(3)
2.109(2), 2.087(3)
-
Ru-CIMe4′
-
2.091(3)
2.089(3)
2.032(3)
Ru-CH2
-
2.246(3)
2.224(3)
-
Ru-PC6H4
2.4073(6)
-
-
2.3445(9)
Ru-C6H4P
2.132(2)
-
-
2.084(3)
Ru-PC6H4Sn
2.3451(5)
2.3162(8)
2.3267(6)
-
Ru-PPh3
-
2.3296(7)
-
2.3118(8)
Ru-Sn
2.6345(2)
2.6435(3)
2.6604(3)
2.5223(4)
CIMe4-Ru-CIMe4
88.50(8)
-
-
-
CIMe4′-Ru-Sn
-
99.05(9)
96.38(8)
81.25(11)
Molecular structures of one of the molecules in the asymmetric
unit of (left) 10 and (right) 11. Ellipsoids
at 30% level. All H atoms and solvent have been omitted for clarity
in 10, while all hydrogens, with the exception of those
attached to C1, have been omitted in 11.Heating 10 at 60 °C in THF or benzene
brought
about metalation of one of the IMe4 ligands to give [Ru(IMe4)(PPh3)(IMe4)′(PPh2C6H4SnMe2)] (11, Scheme ). The 31P{1H} NMR spectrum showed replacement of the low frequency
signal for the cycloruthenated phosphine in 10 by a resonance
at δ 54, arising from a Ru-PPh3 resulting from reductive
elimination of Ru–H (resulting from IMe4 activation)
onto the Ph2PC6H4Ru ligand. The 1H NMR spectrum exhibited a total of seven carbene methyl resonances,
and also showed two doublets of doublets at δ 2.42 and 2.22
(each of relative integral 1) for the diastereotopic protons of the
Ru-CH2 arm (Figures S49–S51).NMR monitoring of the reaction indicated that optimum conversion
of the starting material (ca. 80–85%) occurred over ca. 2 h
at 60 °C to yield 11 as the main reaction product,
although always alongside a number of other, smaller, unidentifiable
species, which became more prominent with longer heating. While we
were therefore unable to isolate 11 as an analytically
pure material, a combination of multinuclear NMR spectra and a crystal
structure determination (achieved by picking of a single crystal)
identified 11 unequivocally.In contrast to the
well-known metalation of NHCs bearing N-aryl
or bulky N-alkyl substituents,[37] C–H
activation of N-methylated carbenes is restricted to a very small
number of examples,[38] most likely because
of the severe structural constraints imposed by forming a four-membered
ring metalacycle. These structural impositions are apparent in the
X-ray structure of 11 (Figure ), which shows a dramatically tilted carbene
ring with very different N(1)–C(2)–Ru(1) and N(2)–C(2)–Ru(1)
angles (99.3(2)° and 156.1(3)° respectively, Δ = 56.8°).
The C1–Ru–C2 angle subtended at Ru (63.36(12)°)
is more similar to that in [Os(PPr3)2(CO)(IMe2)′Cl] (63.01(16)/63.11(16)°)
reported by Esteruelas[38b] than [TpYb(IMe4)(IMe4)′]
(55.4(2)°) described by Ferrence et al.,[38a] most likely due to the presence of both the bigger lanthanide
and the very different ligand coordination spheres.When 10 was heated to 120 °C in toluene, very
different activation chemistry of the carbene took place with the
IMe4 ligands from two molecules of 10 undergoing
redistribution to give a mixture of the six-coordinate, tris-carbene
product [Ru(IMe4)2(IMe4)′(PPh2C6H4SnMe2)] 12 and five-coordinate, monocarbene species [Ru(PPh3)(PPh2C6H4)(IMe4-SnMe2)′] 13 (Scheme ). As for 11, we were able to manually
separate yellow crystals of 12 and purple crystals of 13 to allow their structural characterization, but were unable
to separate enough clean material for elemental analyses or measurement
of pristine NMR spectra. Purple 13 was more obvious to
identify and manually separate, resulting in NMR spectra that were
typically cleaner than those of 12 (Figures S52–S59).The structure of 12, which is shown in Figure (metrics in Table ), displayed a cis-arrangement
of two intact IMe4 ligands, one of which was trans to the
carbenic carbon of the third, metalated IMe4. The difference
in the two Ru–C–N angles (Δ = 57.2°) showed
that this was even more distorted in terms of coordination than that
in 11, although the bite angle did not change (C(1)–Ru(2)–C(7)
= 63.47(11)°). A cyclostannylated phosphine occupied the last
two coordination sites of the highly distorted (e.g., C(7)–Ru(2)–Sn(1)
= 159.35(8)°) octahedral Ru coordination sphere. In accord with
the structure, the 1H and 13C NMR spectra showed
separate resonances for each of the 11 inequivalent NMe and NCMe groups.
Figure 4
Molecular structures of one of the molecules
in the asymmetric
unit in (left) 12 and (right) 13. Ellipsoids
at 30% level. In 12, the minor disordered component and
hydrogens, with the exception of those attached to C7, have been omitted
for clarity. In 13, the minor disordered component and
hydrogens, with the exception of those attached to C6, have been omitted
for clarity.
Molecular structures of one of the molecules
in the asymmetric
unit in (left) 12 and (right) 13. Ellipsoids
at 30% level. In 12, the minor disordered component and
hydrogens, with the exception of those attached to C7, have been omitted
for clarity. In 13, the minor disordered component and
hydrogens, with the exception of those attached to C6, have been omitted
for clarity.The most striking feature of 13 (Figure , Table ) was the formation of a novel
chelating
stannylcarbene ligand, attached to Ru at the apical and equatorial
positions of a distorted square pyramid through very short Ru–C
and Ru–Sn bond lengths (Ru(2)–C(1) = 2.032(3) Å,
Ru(2)–Sn(1) = 2.5223(4) Å). While metal-bound NHCs with
p-block functionalized N-substituents are quite common, they are typically
preformed prior to either addition onto or substitution at a metal
center,[39] as opposed to through a bond
activation reaction as seen here.[40] We
are unaware of any examples of bidentate NHC-Sn ligands prepared by
any route,[41] although Tilley has recently
described a bidentate P–Sn ligand arising from C–H activation
of an Fe-PMeiPr2 ligand onto Sn.[42]
Summary and Conclusions
The synthesis and reactivity
of the heterobimetallic PPh3-derived [Ru–Al] and
[Ru–Sn] complexes [Ru(C6H4PPh2)2{PPh2C6H4AlMe(THF)}H] 5 and [Ru(PPh3)(C6H4PPh2)(PPh2C6H4SnMe2)] 6 has been described
in a study that represents a continuation of our ongoing research
line, in which we attempt to describe and rationalize the effects
of M′ on heterobimetallic Ru-main group metal M′ complexes.
In conjunction with previous studies on [Ru–Li], [Ru–Mg]
and [Ru–Zn] systems 1–3 (Scheme ), we can conclude thatThe nature of M′ strongly affects
both the structure and reactivity of such heterobimetallic complexes,
with ate-type chemistry predominant in the case of
more “ionic” M′ metals such as Li, Mg and Al,
whereas more “covalent” behavior is observed for M′
= Sn and Zn, with direct Ru–Sn (and Ru–Zn) bonds prevalent
(Scheme ).
Scheme 11
Representation of the Two Extremes of [Ru-M′]
Complexes Arsing
in Our Work
Reactivity toward H2 provides
a means to discriminate ate-[Ru-M′] from bonded
[Ru-M′] type complexes; the former react sluggishly, while
the latter react instantaneously, as a result of the presence of a
coordinatively unsaturated Ru center.Combining (reversibly) cyclometalated,
and substitutionally labile, Ru-PPh3 ligands and an increasing
number of M′-Me groups across M′ = Li, Mg, Zn, Al, and
Sn provides a valuable route to heterobimetallic [Ru-M′] containing
new ligand frameworks (e.g., chelating stannylcarbene ligand in 13) with potential for further interesting small molecule
reactivity.
Experimental Section
General Comments
All manipulations were carried out
at room temperature under argon using standard Schlenk, high vacuum,
and glovebox techniques using dry and degassed solvents. C6D6, C6D5CD3, and THF-d8 were vacuum transferred from potassium. NMR
spectra were recorded at 298 K (unless otherwise stated) on Bruker
Avance 400 and 500 MHz NMR spectrometers and referenced as follows:
C6D6 (1H, δ 7.16; 13C, δ 128.0), C6D5CD3 (1H, δ 2.09), THF-d8 (1H, δ 3.58; 13C, δ 25.3). 31P{1H} spectra were referenced externally to 85% H3PO4 and 119Sn to SnMe4. IR
spectra were recorded on a Nicolet Nexus spectrometer and UV–vis
spectra on a PerkinElmer Lambda 35 spectrometer. Elemental analyses
were performed by Elemental Microanalysis Ltd., Okehampton, Devon,
U.K. [Ru(PPh3)3HCl]·toluene,[43] [Ru(PPh3)(C6H4PPh2)2H][Li(THF)2] (1)4 and IMe4,[44] were
prepared according to literature methods. Prior to use, [Ru(PPh3)3HCl]·toluene was dried under high vacuum
and ground to a fine powder affording a material with ca. 1 molecule
of toluene per Ru (based on 1H NMR analysis). IMe4 was purified by sublimation. LiCH2TMS was used as a colorless
solid obtained upon cooling a commercial 1.0 M solution in pentane
at −32 °C, separating the resulting colorless crystals
by decantation in a glovebox and drying under vacuum. AlMe2Cl (1.0 M solution in hexane, Merck) and SnMe3Cl (Merck)
were used as received.
[Ru(C6H4PPh2)2{PPh2C6H4AlMe(THF)}H] 5
AlMe2Cl (135 μL of a 1.0 M solution in hexane, 0.135
mmol) was added to an agitated suspension of [Ru(PPh3)(C6H4PPh2)2H][Li(THF)2] (1, 94 mg, 0.09 mmol) and PPh3 (47 mg,
0.18 mmol)[45] in benzene (1.5 mL), and the
reaction mixture was heated at 60 °C for 1 h. After the mixture
was cooled to room temperature, the precipitate of LiCl was separated
by cannula filtration and the filtrate reduced to dryness. The residue
was dissolved in THF (2 mL), layered with hexane (2 mL), and left
to crystallize at −20 °C (3 days). The yellow crystalline
product was separated by decantation, washed with hexane (2 ×
1 mL), and dried under vacuum. Yield: 71 mg (69%; contains ca. 3 molecules
of THF per Ru based on 1H NMR analysis; Figure S1). 1H NMR (500 MHz, C6D6): δ 8.32 (t, J = 9.0 Hz, 2H, Ar),
7.90 (d, J = 6.9 Hz, 1H, Ar), 7.84–7.74 (m,
4H, Ar), 7.50 (t, J = 6.9 Hz, 1H, Ar), 7.40 (br s,
1H, Ar), 7.30–7.14 (m, 7H, Ar; partially overlapped with residual
C6D5H), 7.05−6.99 (m, 4H, Ar), 6.93 (br
m, 2H, Ar), 6.88–6.78 (m, 5H, Ar), 6.75–6.61 (m, 10H,
Ar), 6.45 (t, J = 7.9 Hz, J = 2.3
Hz, 2H, Ar), 6.06 (t, J = 9.0 Hz, 2H, Ar), 5.83 (br
m, 1H, Ar), 3.56 (m, 11H, THF), 1.40 (m, 11H, THF), −1.00 (s,
3H, AlMe), −6.30 (m, 1H, RuH). 31P{1H} NMR (202 MHz, C6D6): δ 70.0 (dd, 2JPP = 266 Hz, 2JPP = 23 Hz),
−15.4 (dd, 2JPP = 266
Hz, 2JPP = 30 Hz), −25.9
(dd, 2JPP = 30 Hz, 2JPP = 23 Hz). Anal. Calcd. for C55H46AlP3Ru·2.75THF (1126.1): C
70.38, H 6.09. Found: C 70.49, H 6.28.
[Ru(PPh3)(C6H4PPh2)(PPh2C6H4SnMe2)] 6
A THF suspension (10 mL) of [Ru(PPh3)3HCl]·toluene (509 mg, 0.50 mmol) was treated with
LiCH2TMS (97 mg, 1.03 mmol) and stirred for 30 min a J.
Young resealable ampule to afford an orange solution. A solution of
SnMe3Cl (100 mg in 3 mL C6H6, 0.50
mmol) was added dropwise over ca. 3 min with stirring (the vial containing
the SnMe3Cl solution was washed with C6H6 (2 × 1 mL), and the washings added to the reaction).
The resulting dark blue solution was stirred (2 h), and the volatiles
were then removed under vacuum. The residual blue oil was treated
with 20 mL hexane and 10 mL of benzene to precipitate LiCl. The suspension
was cannula filtered, the residue washed with hexane (5 mL), and the
combined filtrate and washings were concentrated under vacuum to yield
a blue oil. Recrystallization from benzene/hexane (1:2 ratio) at room
temperature for 24 h, and then at −20 °C for 72 h, afforded 6 as dark blue crystals, which were separated, washed with
hexane (2 × 1 mL), and dried under vacuum to give 520 mg of product
(94% yield). 6 is present in solution together with ca.
5% of a minor isomer, which we propose to have the structure 6′ shown in Figure S4. Data
for 6: 1H NMR (500 MHz, THF-d8): δ 7.74 (t, J = 6.5 Hz, 1H,
Ar), 7.66 (t, J = 8.4 Hz, 2H, Ar), 7.58 (d, J = 6.9 Hz (JHSn = 25.0 Hz),
1H, Ar), 7.47 (t, J = 8.8 Hz, 2H, Ar), 7.34–7.16
(m, 9H (partially overlaps with C6H6), Ar),
7.14–6.99 (m, 5H, Ar), 6.95–6.84 (m, 14H, Ar), 6.77–6.66
(m, 2H, Ar), 6.64 (t, J = 7.3 Hz, 1H, Ar), 6.55 (t, J = 7.5 Hz, 2H, Ar), 6.40–6.28 (m, 4H, Ar), 0.40
(s (2JHSn = 41 Hz), 3H, SnMe), −1.10 (s (2JHSn = 46 Hz), 3H, SnMe). 31P{1H} NMR (202 MHz, THF-d8): δ
74.8 (dd, 2JPP = 241 Hz, 2JPP = 16 Hz (2JPSn = 150 Hz)), 39.9 (dd, 2JPP = 25 Hz, 2JPP = 16 Hz (2JPSn =
76 Hz)), −28.5 (dd, 2JPP = 241 Hz, 2JPP = 25 Hz (2JPSn = 148 Hz)). 119Sn{1H} NMR (187 MHz, THF-d8): δ 21.4 (td, 2JSnP = 150 Hz, 2JSnP = 78 Hz).
UV/vis (toluene, nm): λmax = 600 (ε = 2080
dm3 mol–1 cm–1), 486
(ε = 1890 dm3 mol–1 cm–1). Anal. Calcd. for C56H49P3RuSn·C6H6 (1112.8): C 66.92, H 4.98. Found: C 68.36, H
5.17. Repeated attempts at analysis gave consistently a high%C value,
which might be attributable to an incorrect formulation for 6′. Selected NMR data for 6′. 1H NMR (500 MHz, THF-d8): δ
0.11 (s (2JHSn = 47 Hz), 3H,
SnMe), −0.20 (s (2JHSn = 44 Hz), 3H, SnMe). 31P{1H} NMR (202 MHz, THF-d8): δ 50.1 (dd, 2JPP =
243 Hz, 2JPP = 17 Hz), 37.4
(dd, 2JPP = 27 Hz, 2JPP = 17 Hz), 3.5 (dd, 2JPP = 243 Hz, 2JPP = 27 Hz).
Variable Temperature NMR Study of the Formation of 5
Variable Temperature NMR Study of the Formation of 6
SnMe3Cl (2.5 mg, 0.012 mmol, in 0.1 mL THF-d8) was injected into a J. Young resealable NMR
tube containing a frozen THF-d8 (0.4 mL)
solution of 1 (11.5 mg, 0.011 mmol). The reaction mixture
was warmed until the THF melted (165 K), at which point the color
changed from yellow-orange to dark blue. The blue solution was inserted
into a precooled (193 K) NMR spectrometer and 1H, 31P{1H} and 1H{31P} NMR spectra
were acquired over the range 193–233 K. A single species assigned
as (Scheme ) was observed up to 273 K, at which point,
the final product 6 was also observed (Figures S13–S16). Selected NMR data for . 1H NMR (400 MHz, THF-d8, 193 K): δ −0.21 (s (2JHSn = 38 Hz), 9H, SnMe3). 31P{1H} NMR (162 MHz, THF-d8, 193 K): δ 48.6 (dd, 2JPP = 241 Hz, 2JPP = 14 Hz (2JPSn =
177 Hz)), 40.6 (dd, 2JPP =
24 Hz, 2JPP = 14 Hz (2JPSn = 155 Hz)), −36.1 (dd, 2JPP = 241 Hz, 2JPP = 24 Hz (2JPSn = 198 Hz)).
[Ru(PPh3)2(PPh2C6H4AlMe)H3] 7
A C6D6 (0.5 mL) solution of 5 (10 mg,
0.009 mmol) in a J. Young resealable NMR tube was freeze–pump–thaw
degassed (×3) and placed under 1 atm H2. Heating at
60 °C for 2 h resulted in complete conversion (based on 31P{1H} NMR spectroscopy) to [Ru(PPh3)2(PPh2C6H4AlMe)H3] 7, which was characterized by 1H
and 31P{1H} NMR spectroscopy. Selected NMR data
for 7: 1H NMR (500 MHz, C6D6): δ −0.39 (s, 3H, AlMe), −8.46
(br d, 1H, Ru···H-Al), −8.72
(td, 2JHP = 28.2 Hz, 2JHP = 13.9 Hz, 1H, Ru-H), −11.07 (dt, 2JHP = 53.5 Hz, 2JHP = 23.3 Hz,
1H, Ru-H···Al). 31P{1H} NMR (202 MHz, C6D6): δ 73.5
(dd, 2JPP = 240 Hz, 2JPP = 28 Hz), 62.1 (dd, 2JPP = 240 Hz, 2JPP = 23 Hz), 58.8 (br t, 2JPP = 25 Hz). Exposure of the sample to vacuum for 2 h followed
by redissolution of the residue in C6D6 revealed
complete degradation of 7 and appearance of a number
of resonances in both the 1H and 31P NMR spectra
(Figure S22), including signals for both
[Ru(PPh3)3(η2-H2)H2] (1H: δ −7.08 (s); 31P{1H}: δ 57.7 (s)) and [Ru(PPh3)4H2] (1H: δ −10.13 (m); 31P{1H}: δ 49.3 (t), 41.1 (t)).[25] The fate of the Al metal was not established.
[Ru(PPh3)2(PPh2C6H4SnMe2)H3] 8
A C6D6 (0.5 mL) or THF-d8 (0.5 mL) solution of 6 (12 mg, 0.011 mmol) in
a J. Young resealable NMR tube was placed under 1 atm of H2, which was then allowed to slowly diffuse through the sample to
yield a pale yellow-colorless homogeneous solution of [Ru(PPh3)2(PPh2C6H4SnMe2)H3] 8. This was characterized by 1H and 31P{1H} variable temperature NMR
spectroscopy. 1H NMR (500 MHz, C6D6): δ 8.0 (d, JHH = 7.0 Hz (JHSn = 32 Hz), 1H, Ar), 7.51 (dd, JHH = 7.8, JHH = 4.5 Hz, 1H,
Ar), 7.39 (t, JHH = 8.9 Hz, 4H, Ar), 7.28
(t, JHH = 8.8 Hz, 12H, Ar), 7.02 (t, J = 7.3 Hz, 1H, Ar), 6.94 (t, JHH = 7.6 Hz, 6H, Ar), 6.88–6.79 (m, 18H, Ar), −0.50 (s
(2JHSn = 47 Hz), 6H, SnMe), −7.51 (br m (2JHSn = 183 Hz), 3H, RuH). 31P{1H} NMR (162 MHz, C6D6): δ 84.7
(t, 2JPP = 99 Hz (2JPSn = 124 Hz)), 56.6 (d, 2JPP = 99 Hz (2JPSn = 95 Hz)). 1H NMR (500 MHz, THF-d8): δ 7.89 (d, JHH = 7.3 Hz (JHSn = 33 Hz), 1H,
Ar), 7.39 (m, 1H, Ar), 7.13 (t, JHH =
7.2 Hz, 7H, Ar), 7.08 (t, JHH = 9.0 Hz,
3H, Ar), 7.04–6.89 (m, 23H, Ar), 0.05 (s (2JHSn = 48 Hz), 6H, SnMe), −7.91
(br m (2JHSn = 179 Hz), 3H,
RuH). 31P{1H} NMR (162 MHz,
THF-d8): δ 85.1 (t, 2JPP = 97 Hz (2JPSn = 127 Hz)), 56.2 (d, 2JPP = 97 Hz (JPSn = 89 Hz)).
Selected 1H NMR (500 MHz, THF-d8, 332 K): δ 0.04 (s (2JHSn = 48 Hz), 6H, SnMe), −7.95 (dt, 2JHP = 16.4 Hz, 2JHP = 7.4 Hz (2JHSn = 180 Hz), 3H, RuH). 31P{1H} NMR (162 MHz, THF-d8, 332 K): δ
84.0 (t, 2JPP = 96 Hz (JPSn = not determined)), 54.6 (d, 2JPP = 96 Hz (JPSn ∼ 96 Hz)). IR (KBr, cm–1): 1967 (νRuHSn), 1746 (νRuHSn).The formation
of 8 could also be performed in the solid-state. Stirring
a microcrystalline sample of 6 (15 mg, 0.013 mmol) under
1 atm of H2 in a J. Young resealable ampule for 24 h brought
about a color change from purple to off-white. Conversion to 8 was proven by IR spectroscopy (Figure S30). In an attempt to prepare an isolable derivative of 8, an excess of pyridine (50 μL, 0.49 mmol) was added
to a C6D6 (0.5 mL) solution of 6 (10 mg, 0.022 mmol) to give [Ru(NC5H5)(C6H4PPh2)(PPh2C6H4SnMe2)] (Figure S31), assigned from the appearance of two doublets (δ 80.1 (d, 2JPP = 285 Hz), −26.1 (d, 2JPP = 285 Hz)) in the 31P{1H} NMR spectrum. Addition of 1 atm H2 to
the crude sample rapidly yielded 31P{1H} NMR
signals of 8 at ca. δ 84 and 55.
[Ru(CO)2(C(O)C6H4PPh2)(PPh2C6H4SnMe2)] 9
A benzene (2 mL) solution of 6 (111
mg, 0.10 mmol) was placed under CO (1 atm), and the solution was stirred
at 80 °C for 4 h. The resulting yellow solution was filtered
through a pad of Celite. The pad was washed with 1 mL C6H6 and the combined filtrate and washings layered with
hexane (6 mL). An initial batch of yellow crystals of 9 were formed. Treatment with additional hexane (6 mL) and cooling
to −32 °C for 24 h afforded yellow crystalline needles.
The yellow solids were combined, washed with hexane (2 × 1 mL)
and dried under vacuum. Yield: 54 mg (63%). 1H NMR (400
MHz, C6D6): δ 8.30–8.23 (m, 2H,
Ar), 7.89–7.79 (m, 3H, Ar), 7.74 (d, J = 7.3
Hz (JHSn = 26 Hz), 1H, Ar), 7.58–7.46
(m, 3H, Ar), 7.42 (t, J = 7.2 Hz, 1H, Ar), 7.27–6.87
(m, 18H, Ar), 0.24 (s (2JHSn = 41 Hz), 3H, SnMe), −0.24 (s (2JHSn = 43 Hz), 3H, SnMe). 31P{1H} NMR (162 MHz, C6D6): δ 66.9 (d, 2JPP = 210 Hz (2JP-117Sn = 140 Hz, 2JP-119Sn = 146 Hz)), 64.2 (d, 2JPP = 210 Hz (2JP-117Sn = 134, 2JP-119Sn =
140 Hz)). 13C{1H} NMR (126 MHz, C6D6): δ 202.9 (t, 2JCP = 8 Hz, Ru-CO), 200.0 (t, 2JCP = 10 Hz, Ru-CO), 158.1 (d, JCP = 40 Hz, Ar), 155.5 (dd, JCP = 61 Hz, JCP = 4 Hz, Ar),
140.7 (d, JCP = 42 Hz, Ar), 140.1 (d, JCP = 4 Hz, Ar), 139.6 (br m, Ar), 139.3 (br
m, Ar), 138.0 (dd, JCP = 42 Hz, JCP = 4 Hz, Ar), 136.2 (d, JCP = 24 Hz, Ar), 135.5 (d, JCP = 9 Hz, Ar), 134.8 (dd, JCP = 39 Hz, JCP = 3 Hz, Ar), 132.9 (d, JCP = 13 Hz, Ar), 132.6 (d, JCP = 9 Hz, Ar), 132.5 (d, JCP = 9 Hz, Ar),
131.7 (d, JCP = 13 Hz, Ar), 131.6 (d, JCP = 11 Hz, Ar), 130.6 (d, JCP = 9 Hz, Ar), 130.4 (d, JCP = 5 Hz, Ar), 130.1 (s, Ar), 129.9 (s, Ar), 129.4 (s, Ar), 128.7
(d, JCP = 11 Hz, Ar), 122.0 (d, JCP = 17 Hz, Ar), −5.8 (s, SnMe), −8.2 (s, SnMe). 119Sn{1H} NMR (187 MHz, C6D6): δ 115.3
(t, 2JSnP = 143 Hz). IR (KBr,
cm–1): 2010 (νCO), 1966 (νCO), 1963 (νCO), 1954 (νCO), 1596 (νC(O)C6H4), 1568 (νC(O)C6H4). IR (C6D6, cm–1): 2008
(νCO), 1961 (νCO). Anal. Calcd.
for C41H34O3P2RuSn (856.4):
C 57.50, H 4.00. Found: C 57.88, H 4.09.
Variable Temperature/13CO NMR Study of the Formation
of 9
Data for 5, 9, 10, and 13 were collected on
an Agilent Xcalibur diffractometer (using a Mo Kα radiation)
while the 6, 11, and 12 data
sets were obtained using an Agilent SuperNova instrument and a Cu
Kα source (Table S1). All experiments
were conducted at 150 K, solved by employing either the solution program
native to Olex2[47] or SHELXT.[48] Refinements were conducted using SHELXL[49] via the Olex2 interface. Convergence of the
models was largely unremarkable and only exceptional points of note
will be outlined herein. In particular, the asymmetric unit in 5 was seen to contain one molecule of the organometallic complex
and two molecules of THF. The hydride ligand in the main feature was
located and refined without restraints, while C58 from the ligated
THF was modeled to take account of 55:45 disorder. One of the guest
THF molecules also resolved satisfactorily into two disordered components
(60:40 ratio) with the inclusion of some distance and ADP restraints
in the final least-squares. The second solvent moiety was readily
identifiable as a THF, but disorder was messy, and it prevailed beyond
two fractions. As such, this was ultimately treated using the solvent
mask algorithm in Olex2, and an allowance for the same was made in
the formula as presented.In 10, the asymmetric
unit was noted to comprise one molecule of the Sn–Ru
complex, one full-occupancy molecule of benzene and another benzene
moiety which was modeled to take account of 72:28 disorder. Each component
of the latter was treated as a rigid hexagon in the refinement. The
highest residual electron density peaks in the difference Fourier
map are at chemically insignificant distances from atoms in the main
feature. Indeed, they may point toward some very minor disorder, at
a level which negates modeling.The hydrogen atoms attached
to C1 in 11 were located
and refined subject to being a common distance from the parent atom.
The highest residual electron density peak is located at a chemically
insignificant distance from Sn1. One molecule of the organometallic
complex and a region of solvent correspond to the asymmetric unit
in the structure of 12. The hydrogen atoms attached to
C7 were located and refined freely. Analysis of the electron density
indicated 5% disorder of the tin center (at location Sn1a) and this
was accounted for in the model. However, no effort was made to model
the necessary 5% disorder of the phosphine ligand attached the main
group metal, as it would be imprudent to invest in location of 5%
disorder for first row elements with the expectation of credibility.
The solvent moiety was very disordered and was ultimately treated
using the solvent mask algorithm available in Olex2, with an allowance
made for the presence of one molecule of toluene, per asymmetric unit,
in the formula as presented.In 13, the hydrogen
atoms attached to C6 were located
and refined freely. The electron density indicated 9% disorder of
the tin center (at location Sn1a) and this was accounted for in the
model. However, (using similar rationale to that employed for 6) no effort was made to model the necessary 9% disorder of
the phosphine ligand attached the minor tin component. Distance and
ADP restraints were included for the minor tin component.
Computational Methodology
DFT calculations were run
with Gaussian 16 (C.01).[50] The Al, P, Ru
and Sn centers were described with the Stuttgart RECPs and associated
basis sets,[51] and the 6-31G** basis set
was used for all other atoms (BS1).[52] A
polarization function was also added to Al (ζd =
0.190), P (ζd = 0.387) and Sn (ζd = 0.180).[53] Initial BP86 optimizations[54] were performed using the ‘grid = ultrafine’
option, with all stationary points being fully characterized via analytical
frequency calculations as minima. All energies were recomputed with
a larger basis set (BS2) featuring 6-311++G** basis sets on all atoms,
with the exception of Ru and Sn which employed the basis set aug-cc-pVTZ-PP.
Corrections for the effect of solvent (benzene: ε = 2.2706;
THF: ε = 7.4257) solvent were employed using the polarizable
continuum model and BS1.[55] Single-point
dispersion corrections to the BP86 results employed Grimme’s
D3 parameter set with Becke-Johnson damping as implemented in Gaussian.[56] Natural Bonding Orbital (NBO 3.1)[57] analyses were performed on the BP86/BS1 optimized
geometries at the BP86/BS2 level.
Authors: Christian P Sindlinger; Sebastian Weiss; Hartmut Schubert; Lars Wesemann Journal: Angew Chem Int Ed Engl Date: 2015-02-05 Impact factor: 15.336
Authors: Melvyn B Ansell; Debbie E Roberts; F Geoffrey N Cloke; Oscar Navarro; John Spencer Journal: Angew Chem Int Ed Engl Date: 2015-04-09 Impact factor: 15.336
Authors: Mateusz K Cybulski; Nicholas A Beattie; Stuart A Macgregor; Mary F Mahon; Michael K Whittlesey Journal: Chemistry Date: 2020-07-28 Impact factor: 5.236