Homologation of carbon monoxide is central to the heterogeneous Fischer-Tropsch process for the production of hydrocarbon fuels. C-C bond formation has been modeled by homogeneous systems, with [CnOn]2- fragments (n = 2-6) formed by two-electron reduction being commonly encountered. Here, we show that four- or six-electron reduction of CO can be accomplished by the use of anionic aluminum(I) ("aluminyl") compounds to give both topologically linear and branched C4/C6 chains. We show that the mechanism for homologation relies on the highly electron-rich nature of the aluminyl reagent and on an unusual mode of interaction of the CO molecule, which behaves primarily as a Z-type ligand in initial adduct formation. The formation of [C6O6]4- from [C4O4]4- shows for the first time a solution-phase CO homologation process that brings about chain branching via complete C-O bond cleavage, while a comparison of the linear [C4O4]4- system with the [C4O4]6- congener formed under more reducing conditions models the net conversion of C-O bonds to C-C bonds in the presence of additional reductants.
Homologation of carbon monoxide is central to the heterogeneous Fischer-Tropsch process for the production of hydrocarbon fuels. C-C bond formation has been modeled by homogeneous systems, with [CnOn]2- fragments (n = 2-6) formed by two-electron reduction being commonly encountered. Here, we show that four- or six-electron reduction of CO can be accomplished by the use of anionic aluminum(I) ("aluminyl") compounds to give both topologically linear and branched C4/C6 chains. We show that the mechanism for homologation relies on the highly electron-rich nature of the aluminyl reagent and on an unusual mode of interaction of the CO molecule, which behaves primarily as a Z-type ligand in initial adduct formation. The formation of [C6O6]4- from [C4O4]4- shows for the first time a solution-phase CO homologation process that brings about chain branching via complete C-O bond cleavage, while a comparison of the linear [C4O4]4- system with the [C4O4]6- congener formed under more reducing conditions models the net conversion of C-O bonds to C-C bonds in the presence of additional reductants.
The assembly of complex
molecules from carbon monoxide via C–C
bond formation represents a fundamental chemical challenge that has
key relevance to the production of hydrocarbon fuels via the heterogeneous
Fischer–Tropsch process.[1] Although
employed industrially since 1925, the mechanism for the conversion
of mixtures of CO and H2 to short-to-medium-chain alkane
products through the formation of C–C bonds remains the subject
of significant debate.[2] C=C bond formation to yield alkenes represents
a competing homologation process, and the product mixture typically
also includes oxygenated species along with methane—an undesirable
reduction product produced without the accompanying C–C bond
formation. The idealized ratio of H2/CO (ca. 2:1; eq , typically with 10 < n < 20) reflects the importance of the reductant (H2) in generating the desired alkane products. The much lower
proportion of dihydrogen in the coal-derived synthesis gas (“Syn
Gas”) feedstock is typically rectified through application
of the water gas shift reaction.[3]The amenability of homogeneous systems to small-molecule characterization
techniques has driven the investigation of organometallic compounds
capable of modeling the homologation of CO via C–C bond formation.[4] Prominent examples include d-block metal compounds (mirroring the use of heterogeneous transition-metal
catalysts such as iron and cobalt in the Fischer–Tropsch process),[5−17] with examples derived from low-valent s-, p-,[18−33] and f-block compounds[34−50] having also been reported (e.g., I–VII, Figure ). Prevalent
among the homologation products formed via reduction processes are
cyclic systems of the type [CO]2– (n = 3, 4,
and 6) and the related ethynediolate system [C2O2]2–,[6−8,11,18,20−23,29,40,41,44−48] formed by a formal two-electron reduction process, although a number
of systems featuring longer linear chains of carbon atoms have also
been reported (Figure ).[14,17,22,25,31,33,34,36,37,42,49] Within this sphere, a small number of studies have
emerged, which demonstrate the potential for control of homologation
processes at a constant oxidation level: the selective formation of
[CO]2– (n = 2, 3, or 4) by the cooperative
action of U(III) or Mg(I) centers has been shown to be influenced
by the steric bulk of ancillary ligands,[21,48] and the stepwise growth of C3 chains by the addition
of CO has been demonstrated at d-block metal systems
derived from metal carbonyl precursors.[17] Very recently, the formation of C4 or C5 chains
using aluminum(I) systems has been demonstrated, with the product
formed appearing to reflect the nuclearity of the metal reagent.[33]
Figure 1
Selected examples of CO homologation relevant to the current
study.
Selected examples of CO homologation relevant to the current
study.We have been developing the chemistry
of electron-rich aluminum
(“aluminyl”) compounds—stabilized by NON supporting
ligands (4,5-bis(2,6-diisopropylanilido)-2,7-di-tert-butyl-9,9-dimethylxanthene; Figure ).[51,52] Such systems have been shown
to be capable of the cleavage of C–H,[51] C–C,[53] and C–O bonds in
arene substrates[54] and to allow access
to reactive terminal imide species which can cleave the CO bond in
carbon monoxide.[55] Systems such as K2[(NON)Al]2 ([1-K]2) have also shown strongly reducing capabilities toward oxygen-containing
substrates,[56] and here, we report on the
reactivity of aluminyl compounds toward CO. We show that the homologation
to give the known [C4O4]4– fragment occurs with both mono- and dinuclear systems, a finding
rationalized computationally through a rate-determining C=C
bond-forming dimerization process. The initial interaction of the
CO molecule at aluminum is unusual, in that it behaves primarily as
a Z-type ligand, reflecting the highly electron-rich nature of the
aluminyl reagent. Subsequent formation of [C6O6]4– from [C4O4]4– shows for the first time a solution-phase CO homologation process
that brings about chain branching via complete C–O bond cleavage.
In addition, comparison of the linear [C4O4]4– system with the (unprecedented) [C4O4]6– congener formed under more reducing
conditions models the conversion of C–O bonds to C–C
bonds in the presence of additional reductants.
Figure 2
Aluminyl compounds central
to the current study: dimeric potassium
aluminyl compound K2[(NON)Al]2, [1-K]2, and monomeric lithium analogue (Et2O)2LiAl(NON), [1-Li(OEt2)2].
Aluminyl compounds central
to the current study: dimeric potassium
aluminyl compound K2[(NON)Al]2, [1-K]2, and monomeric lithium analogue (Et2O)2LiAl(NON), [1-Li(OEt2)2].
Results and Discussion
Synthesis of [C4O4]4– and [C6O6]4– via CO Homologation
The reaction of
potassium aluminyl compound [1-K]2[51] with CO (ca. 1.5 atm) in benzene
over 16 h at ca. 35 °C (followed by recrystallization from toluene)
leads to the isolation of the orange crystalline product [K(C7H8)]2[{(NON)Al}2(C4O4)], [2-K(C7H8)]2, which has been characterized by standard spectroscopic and
micro-analytical methods (Scheme and Figure ). [2-K(C7H8)]2 can be shown by X-ray crystallography to feature a centrosymmetric
structure in which two [(NON)Al] units are bridged by a [C4O4] fragment and two K+ counterions (which
each interact with the π system of a Dipp group of one NON ligand
and an additional toluene molecule; Figure ).
Scheme 1
Homologation of CO by Aluminyl Compounds Yielding Products Containing
[C4O4]4– or [C6O6]4– Fragments
Four-electron
reduction of CO
by [1-K]2 or 2 equiv of [1-Li(OEt2)2] to give [2-K(C7H8)]2/[2-Li(OEt2)]2, featuring the topologically linear [C4O4]4– fragment; onward reaction of [2-K(C7H8)]2 with CO to yield the [C6O6]4– system, 3.
Figure 3
Molecular structures of (upper) [K(C7H8)]2[{(NON) Al}2(C4O4)], [2-K(C7H8)]2, and (lower)
the anionic component of the asymmetric unit of K[K(THF)2][{(NON)Al}2(C6O6)], 3, in the solid state as determined by X-ray crystallography. Hydrogen
atoms omitted and selected groups represented in the wireframe format
for clarity; thermal ellipsoids drawn at the 50% probability level.
Key metrical parameters are listed in Table .
Molecular structures of (upper) [K(C7H8)]2[{(NON) Al}2(C4O4)], [2-K(C7H8)]2, and (lower)
the anionic component of the asymmetric unit of K[K(THF)2][{(NON)Al}2(C6O6)], 3, in the solid state as determined by X-ray crystallography. Hydrogen
atoms omitted and selected groups represented in the wireframe format
for clarity; thermal ellipsoids drawn at the 50% probability level.
Key metrical parameters are listed in Table .
Table 1
Key Metrical Parameters for Aluminum
Complexes Containing [C4O4] Ions (n = 2, 4, and 6)
[C4O4]4–
[C4O4]6–
[C4O4]2–
d/Å
[2-K(C7H8)]2
[2-K(THF)]2
[2-Li(OEt2)]2
[4-K2(BEt3)]2
5
C–C
1.507(1)
1.525(4)
1.482(5)
1.481(3)
1.438(2)
1.373(1)
1.489(3)
1.357(4)
1.377(4)
1.461(2)
1.376(3)
C–O
1.233(1)
1.213(4)
1.246(5)
1.382(3)
1.257(2)
1.383(1)
1.247(3)
1.391(4)
1.440(3)
1.265(2)
1.373(3)
1.381(3)
Al–C
2.038(1)
2.053(2)
2.063(4)
2.008(2)
n/a
2.041(3)
Al–O
1.871(1)
1.872(2)
1.877(2)
1.874(2)
1.933(1)
1.860(2)
1.994(1)
Homologation of CO by Aluminyl Compounds Yielding Products Containing
[C4O4]4– or [C6O6]4– Fragments
Four-electron
reduction of CO
by [1-K]2 or 2 equiv of [1-Li(OEt2)2] to give [2-K(C7H8)]2/[2-Li(OEt2)]2, featuring the topologically linear [C4O4]4– fragment; onward reaction of [2-K(C7H8)]2 with CO to yield the [C6O6]4– system, 3.Overall charge balance implies that the bridging
unit is [C4O4]4–, a finding
consistent with
the presence in the (dimeric) aluminyl starting material of two Al(I)
reductants, that is, with four-electron reduction of four molecules
of CO. To a first approximation, C–C distances within the [C4O4]4– unit reflect the presence
of two C–C single bonds (1.507(1) Å) and a central C=C
double bond (1.373(1) Å), while the C–O distances (1.383(1)/1.233(1)
Å) are consistent with C–O single bonds linked to the central carbon atoms [i.e., C(2)] and
C=O double bonds involving the Al-bound carbons C(1).Four-carbon homologation has some literature precedent. In terms
of a metal-bound ligand system, the C/O skeleton bears some resemblance
connectivity-wise to the enedione diolate [R2C4O4]2– system reported by Marks and co-workers,
derived from the insertion of two molecules of CO into a thorium(IV)
alkyl bond (IV)[34,36] and to the mixed d-block/Al
systems reported by Crimmin et al. (VI).[17] Structurally, however, it is most closely related to Stephan’s
B/N-functionalized system Cy2N{C4O4(BCy2)2}NCy2 derived from the trapping
of CO by a combination of LiNCy2 and ClBCy2,
and which also features a C(=O)–C(−O)=C(−O)–C(=O)
skeleton (V).[31] In addition,
a similar C4 system was reported very recently by Coles
and co-workers (VII).[33]The isolated yield of [2-K(C7H8)]2 obtained from [1-K]2 and CO
is low (ca. 10%), and in situ 1H NMR monitoring is consistent
with the formation of other (NON)Al-containing species under these
reaction conditions. Hypothesizing that this is due to the formation
of other homologation products of the type [CO]4–, we sought
(i) to probe the reactivity of isolated samples of [2-K(C7H8)]2 with CO under more forcing
conditions, to see if homologation could be driven toward longer C–C-bonded
chains and (ii) to investigate if greater control of the delivery
of the carbon monoxide (and with it the product distribution) could
be achieved by the use of alternative CO-containing reagents. Targeting
the second of these objectives, we employed Fe(CO)5 as
a labile (and readily soluble) source of the CO molecule in reactions
with [1-K]2. Reaction with 4 equiv of Fe(CO)5 in benzene-d6 in the presence
of the crown ether 12-crown-4 (to aid crystallization) leads to the
formation of [2-K(12-c-4)]2, the solid-state
structure of which is very closely related to that of [2-K(C7H8)]2 (see the Supporting Information). While the potential
involvement of the iron center in Fe(CO)5 in this chemistry
cannot be ruled out, it is noteworthy that conversion to the same
[C4O4]4– fragment via this
approach is markedly higher (20–30%) than that obtained using
gaseous CO.The possibility for assembling [CO]4– chains
with n > 4 was probed via the reaction of [2-K(C7H8)]2 with CO at higher
temperatures/pressures.
Use of a 2 atm pressure in THF-d8 at ca.
65 °C over 4 d leads to quantitative conversion by 1H NMR (through at least three intermediate species) to a single product
featuring two inequivalent NON ligands, which can be shown by X-ray
crystallography to be K[K(THF)2] [{(NON)Al}2(C6O6)], (3; Figure ). 3 features
a [C6O6]4– fragment formed
by the assimilation of two further equivalents of CO, which presents
distinct (O,O and C,O) coordination modes at the two aluminum centers.
The geometry of the [C6O6]4– group is unique, featuring a six-membered pyran core, with the mono-branched
nature of the six-carbon chain implying that C–C bond formation
is also accompanied by complete cleavage of one of the C–O
bonds. While the homologation of CO by highly reducing molecular species
to give linear or cyclic products has a literature precedent, to our
knowledge, the formation of branched carbon chains
in this fashion, with the accompanying cleavage of the exceptionally
strong C–O bond, is unprecedented. Examples of branched-chain
carbon skeletons have been reported previously—but only via
subsequent derivatization reactions of linear CO-derived chains with
electrophiles such as CO2 or MeI.[50]
Mechanistic Studies: CO Coordination and C–C/C=C
Bond Formation
While the two-electron reduction of CO by
a number of main group compounds to give systems of the type [CO]2– has been reported,[18,20−23,29] four-electron processes yielding the topologically linear [C4O4]4– fragment have a more limited
literature precedent.[31,33] We were therefore keen to probe
the mechanism of this transformation through both experimental and
quantum chemical methods. As a key mechanistic principle, we first
set out to determine whether the dimeric nature of the aluminyl reagent,
[1-K]2, is critical in driving the formation
of the (four-electron reduction) product, [C4O4]4–.[1-K]2 has been
shown by DOSY NMR measurements to retain a dinuclear structure in
solution in arene solvents. With this in mind, we targeted the reactivity
of CO with related aluminyl compounds known to possess a mononuclear
structure. The reaction of the (monomeric) lithium aluminyl complex
(Et2O)2LiAl(NON) ([1-Li(OEt2)2])[57] with CO was therefore
probed under similar conditions. Intriguingly, this chemistry leads
to the formation of a very similar dinuclear product, [2-Li(OEt2)]2, featuring an analogous [C4O4]4– fragment, and two encapsulated
Li+ counterions each ligated by two of the oxygen atoms
of the [C4O4]4– unit and an
additional molecule of diethyl ether. The structural metrics relating
to the Al2(C4O4) unit (Table and the Supporting Information) are very similar to those measured
for [2-K(C7H8)]2, being
consistent with a centrosymmetric four-carbon chain incorporating
two C–C and one C=C bonds. Moreover, the formation of
[2-Li(OEt2)]2 from [1-Li(OEt2)2] in this manner (in greater conversion
to [2-K(C7H8)]2, 40
vs 10%) suggests that the state of aggregation of the aluminyl reagent
is not critical in determining the stoichiometry of this carbon-containing
product. This observation is consistent with the mechanism proposed
by Marks and co-workers for the formation of [Cp*2Th{OC(CH2Bu)C(O)}]2, in which
the central C=C bond is postulated as being constructed in
the final mechanistic step through the dimerization of two mono-metallic
carbene units (themselves accessed by bending of a coordinated ketene
ligand).[36] Coles has also proposed that
the dimerization of a dioxocarbene moiety is the final step in the
formation of an ethenetraolate ligand, [C2O4]4–.[58]In
the case of the (NON)Al-derived systems presented here, this
hypothesis is consistent with the results of mechanistic calculations
carried out using density functional theory—for a model system
in which the backbone Bu and Dipp Pr groups are replaced by Me and the cation
is omitted for computational efficiency (Figure ). The rate-determining step for the lowest
energy pathway is found to involve dimerization of two carbene fragments,
with these each being derived from the coupling of two molecules of
CO. Kinetically, the transition state energy for the C=C forming
step is located ca. 34 kcal mol–1 above the starting
materials (and ca. 37 kcal mol–1 above the preceding
singlet carbene intermediate). This barrier height reflects, at least
in part, the sterically encumbered (and anionic) nature of the two
NON-supported fragments being brought together and presumably underpins
the relatively low yields associated with systems such as [2-K(C7H8)]2 (and the possibility
for forming longer C-based chains via carbene/CO coupling).[33]
Figure 4
Proposed mechanism for CO homologation. The lowest energy
pathway
(blue) involves iso-carbonyl binding of one molecule of CO (I2); the alternative (higher barrier) pathway shown in red
involves binding of the second CO through carbon (I2′) and then isomerization to I2. (Calculations carried
out at the B3LYP/def2-TZVP//B3LYP/def2-SVP level with solvation modeled
with CPCM (benzene) and Bu and Pr groups replaced by Me for computational efficiency.)
Proposed mechanism for CO homologation. The lowest energy
pathway
(blue) involves iso-carbonyl binding of one molecule of CO (I2); the alternative (higher barrier) pathway shown in red
involves binding of the second CO through carbon (I2′) and then isomerization to I2. (Calculations carried
out at the B3LYP/def2-TZVP//B3LYP/def2-SVP level with solvation modeled
with CPCM (benzene) and Bu and Pr groups replaced by Me for computational efficiency.)In terms of the initial interaction of the CO molecule(s)
with
the (model) [(NON′)Al]− unit, we find that
the end-on approach of CO perpendicular to the AlN2 plane
in a manner analogous to that determined for CO-adducts of isoelectronic
silylene systems, X2Si·CO,[59] does not correspond to a local minimum on the potential energy surface.
Rather, reflecting the nucleophilic nature of the aluminyl reagent,
the approach of the carbon monoxide unit gives rise to a bent Al–C–O
unit, and the predominant orbital interaction involves electron donation from Al to the π* orbital of CO. For the model C-bound
bis (carbonyl) adduct [(NON′)Al(CO)2]−, the HOMO (Figure ) defines a three-center interaction involving the aluminum-centered
lone pair and the π* orbitals of the two carbonyl ligands. Moreover,
extended transition state with
natural orbitals for chemical valence (ETS-NOCV) analysis determines
that this three-center “back-bonding” interaction accounts
for the majority (261 kcal mol–1) of the orbital
interaction energy (ΔEorb).[60] By contrast, the complementary three-center
interaction involving donation from the CO lone pairs to the Al-centered
pπ-orbital, accounts for only 25 kcal mol–1. By means of comparison, a related ETS-NOCV analysis on the mono-carbonyl
adduct of a cationic borole (in which the CO ligand behaves primarily
as a σ-donor) yields 26 kcal mol–1 (back-bonding
to CO) and 106 kcal mol–1 (sigma donation from CO).[61] As such, the primary role of the CO donors in
[(NON′)Al(CO)2]− is strongly suggested
to be as Z-type ligands.
Figure 5
(Upper) HOMO of the model C-bound bis(carbonyl)
adduct [(NON′)Al(CO)2]− (shading
denotes phasing of the wavefunction); (lower) predominant orbital
interactions determined by ETS-NOCV methods for the binding of the
two carbonyl ligands (shading denotes deformation densities: blue/green—depletion/enhancement
of electron density).
(Upper) HOMO of the model C-bound bis(carbonyl)
adduct [(NON′)Al(CO)2]− (shading
denotes phasing of the wavefunction); (lower) predominant orbital
interactions determined by ETS-NOCV methods for the binding of the
two carbonyl ligands (shading denotes deformation densities: blue/green—depletion/enhancement
of electron density).While the bis-C-ligated adduct [(NON′)Al(CO)2]− (I2′) represents
the lowest energy form of the dicarbonyl species, onward transformation
into carbene intermediate I3 necessarily involves rotation
of one of the CO ligands to generate the isomeric species [(NON′)Al(CO)(OC)]− (I2) featuring one isocarbonyl donor.[62] The barrier to this isomerization within the
coordination sphere of aluminum (ca. 40.2 kcal mol–1) is found to be significantly higher than that for simple dissociation/re-association.
Moreover, while I2 is ca. 17.5 kcal mol–1 higher in energy than I2′, it is capable of
undergoing a very facile C–C coupling to generate carbene intermediate I3, which is lower in energy than either of the carbonyl adducts.Interestingly, although C=C bond formation represents the
key thermodynamic driving force for the overall homologation reaction
(with the final dimerization process being energetically downhill
by ca. 70 kcal mol–1), experimental evidence for
the lability of the central C=C bond in [C4O4]4– systems can be obtained
from the reactions of isolated samples of [2-K(C7H8)]2 with alkali metal hydroborate
reagents. In the case of the reaction with Li[Bu3BH] in benzene/THF, simple cation metathesis
is observed, generating [Li(THF)]2 [{(NON)Al}2(C4O4)], [2-Li(THF)]2, the structure of which closely resembles both [2-K(C7H8)]2 and [2-Li(OEt2)]2 (see the Supporting Information). However, thermolysis with K[Et3BH] at ca. 65–70
°C in the same solvent system leads to 70% conversion over a
period of 14 d to a new species, the apparently analogous potassium
salt [2-K(THF)]2. The structure of this species,
however, can be shown by X-ray crystallography to feature the alternative syn disposition of substituents about the central C=C
double bond (Scheme and Figure ).[33] The C=C/C–C and C=O/C–O
bond lengths within the carbon chain are not significantly different
from the anti-form, but the [2-K(THF)]2 units are linked in the solid state via bridging K+ cations into tetra-aluminum “dimer of dimers” (see
the Supporting Information). The lower
solubility of this aggregate presumably drives the anti-to-syn isomerization process.
Scheme 2
Isomerization of [C4O4]4–: anti to syn Isomerization of
the [C4O4]4– Fragment in the
Presence of K[Et3BH] in Benzene/THF at ca. 60 °C
Figure 6
Molecular structure of
[K(THF)]2[{(NON)Al}2(C4O4)], [2-K(THF)]2, in the solid state as determined
by X-ray crystallography, showing
dimeric unit analogous to [2-K(C7H8)]2. Further aggregation into “dimer of dimers”
shown in Figure S16 (Supporting Information). Hydrogen atoms omitted and selected
groups represented in a wireframe format for clarity; thermal ellipsoids
drawn at the 50% probability level. Key metrical parameters are listed
in Table .
Molecular structure of
[K(THF)]2[{(NON)Al}2(C4O4)], [2-K(THF)]2, in the solid state as determined
by X-ray crystallography, showing
dimeric unit analogous to [2-K(C7H8)]2. Further aggregation into “dimer of dimers”
shown in Figure S16 (Supporting Information). Hydrogen atoms omitted and selected
groups represented in a wireframe format for clarity; thermal ellipsoids
drawn at the 50% probability level. Key metrical parameters are listed
in Table .Kinetically, both heat and the presence of additional
K+ cations can be shown to be necessary for the transformation
to occur.
C=O/C=C conjugation within the [C4O4]4– unit (manifested through resonance structures
featuring a central C(2)–C(2′) single bond) presumably
lowers the barrier to rotation about the central C=C double
bond and allows mechanistically for the anti-to-syn isomerization process.
[C4O4]6– and [C4O4]2– Systems: Synthesis and
Structural Comparisons
While isolated samples
of [2-K(C7H8)]2 are
resistant to further reduction by the hydride component of Li[Bu3BH] or K[Et3BH],
the addition of K[Et3BH] to a mixture of [1-K]2/CO in benzene/THF in situ allows a more reduced product
to be obtained. In this case, the reaction leads to evolution of dihydrogen
(as judged by 1H NMR spectroscopy) and the formation of
the further reduced species K4[{(NON)Al}2(C4O4)(BEt3)2] ([4-K2(BEt3)]2), containing an unprecedented
formally hexa-anionic [C4O4]6– chain (Scheme and Figure ). Mechanistically,
the fact that isolated samples of [2-K(C7H8)]2 are not reduced (either
chemically or electrochemically) suggests that the formation of [4-K2(BEt3)]2 involves a reduction
event which occurs prior to the dimerization of the
carbene intermediate I3.
Scheme 3
Synthesis of a Dialuminum System Containing the [C4O4]6– Fragment: Synthesis of K4[{(NON)Al}2(C4O4) (BEt3)2], [4-K2(BEt3)]2, via the Reaction of [1-K]2 with
CO in the Presence of the Hydride Source K[Et3BH]
Figure 7
Molecular structure of [4-K2(BEt3)]2 in the solid state as
determined by X-ray crystallography.
Hydrogen atoms omitted and selected groups represented in a wireframe
format for clarity; thermal ellipsoids drawn at the 50% probability
level. Key metrical parameters are listed in Table .
Molecular structure of [4-K2(BEt3)]2 in the solid state as
determined by X-ray crystallography.
Hydrogen atoms omitted and selected groups represented in a wireframe
format for clarity; thermal ellipsoids drawn at the 50% probability
level. Key metrical parameters are listed in Table .[4-K2(BEt3)]2 is
extremely reactive, particularly in solution, and has limited solubility
in compatible solvents; as such, it can be characterized by 1H NMR spectroscopy and X-ray crystallography only. The centro-symmetric
structure revealed crystallographically reveals four K+ cations encapsulated by a combination of O-atom and aryl π
system ligation. Two molecules of triethylborane are also incorporated
into the molecular framework, bound to the terminal O-atoms of the
[C4O4]6– chain. The central
C–C distance is consistent with a single bond (1.481(3) Å),
and the two terminal CC linkages feature bond lengths (1.377(4) Å)
reflective of a C=C bond within an enolate functionality. The
C–O distances[(1.382(3) and 1.440(3) Å] are consistent
with single bonds. As such, the structural metrics for [4-K2(BEt3)]2 are consistent with
a system based around a butadiene-tetraolate skeleton, C(−O)=C(−O)–C(−O)=C(−O).To allow for a complete structural comparison of the series of
systems of the type [C4O4] (n = 2, 4, 6) within the same aluminum
coordination sphere, we also targeted the corresponding [C4O4]2– (squarate) system. Such systems
have been synthesized in the past by homologation of CO using metal
centers (or combinations of metal centers) capable of effecting a
two-electron reduction.[20,44] In our hands, however,
dimeric Al(II) systems such as (NON)Al–Al(NON)[51] do not react with CO under either thermal or photolytic
conditions, presumably reflecting the stronger nature of the metal–metal
bond compared, for example, to dimeric Mg(I) reductants.[20−22] For structural comparison, we therefore examined alternative synthetic
approaches involving the reactions of pre-formed squarate salts M2[C4O4] with the readily available Al(III)
iodide precursor (NON)AlI. In the event, (NON)Al(C4O4)Al(NON) (5) is most easily accessed by combining
the disilver salt of squaric acid, Ag2[C4O4], with (NON)AlI in Et2O. 5 has been
characterized by standard spectroscopic and analytical methods and
by X-ray crystallography (Figure ). Its molecular structure features a planar carbocyclic
[C4O4]2– ligand bridging between
the two aluminum centers. The carbon-containing fragment closely resembles
those reported previously,[20,44] featuring C–C
and C–O distances of 1.438(2)/1.461(2) and 1.257(2)/1.265(2)
Å, respectively, consistent with the presence of a delocalized
π system involving all four CO units, albeit with the fourfold
symmetry of the free squarate ion being disrupted by chelation to
the aluminum centers. The two different Al–O distances [1.933(1)
and 1.994(1) Å] reflect a geometry at aluminum, which is between
square pyramidal and trigonal bipyramidal (τ = 0.55), with the
longer bond length being associated with the axial O-donor (assuming
the TBP limit).
Figure 8
Molecular structure of the dialuminum-supported [C4O4]2– compound {(NON)Al}2(C4O4), 5, in the solid
state as determined
by X-ray crystallography. Hydrogen atoms omitted and selected groups
represented in a wireframe format for clarity; thermal ellipsoids
drawn at the 50% probability level. Key metrical parameters are listed
in Table .
Molecular structure of the dialuminum-supported [C4O4]2– compound {(NON)Al}2(C4O4), 5, in the solid
state as determined
by X-ray crystallography. Hydrogen atoms omitted and selected groups
represented in a wireframe format for clarity; thermal ellipsoids
drawn at the 50% probability level. Key metrical parameters are listed
in Table .The availability of structural data for 5, [2-K(C7H8)]2, and [4-K2(BEt3)]2 provides a unique opportunity
to probe the effects on the carbon/oxygen aggregate of stepwise (formal)
reduction of the carbocyclic fragment (Figure )—notwithstanding the fact that (experimentally)
both chemical oxidation and electrochemical oxidation or reduction
of [2-K(C7H8)]2 are
not facile. The two π-electron squarate dianions have been described
as possessing a “moderate” aromaticity,[63] a factor which underpins the prevalence of this planar
carbocycle in four-carbon systems derived from two-electron reduction
processes of CO.[20,44] In the case of 5, DFT calculations imply that an alternative (linear) isomer featuring
a chain of four −C(=O)– units (akin to a doubly
oxidized form of [2-K(C7H8)]2) is not an energetic minimum, rearranging to give a system
featuring a bridging ethynediolate fragment ([OCCO]2–) and two aluminum-bound CO ligands (which even then lies ca. 10.8
kcal mol–1 above the squarate form; see the Supporting Information). The topologically linear
butenedione diolate structure associated with the more heavily reduced
[C4O4]4– tetra-anion avoids
the formation of a Hückel 4π anti-aromatic carbocycle
derived from the addition of an additional pair of electrons to the
LUMO of the squarate system. An alternative carbocyclic isomer featuring
a more pronounced rectangular “squarate” core (akin
to the Jahn–Teller distorted structure of the related 4π
electron system cyclobutadiene)[64] is found
to lie significantly higher in energy (by >48 kcal mol–1).
Figure 9
Comparison of the structures of the [C4O4] fragments within 5,
[2-K(C7H8)]2, and
[4-K2(BEt3)]2: (upper)
important resonance structures of [C4O4]2–, [C4O4]4–, and [C4O4]6– fragments;
(lower) Key frontier orbitals: HOMO-1 and LUMO of [2-K(C7H8)]2 and HOMO of [4-K2(BEt3)]2.
Comparison of the structures of the [C4O4] fragments within 5,
[2-K(C7H8)]2, and
[4-K2(BEt3)]2: (upper)
important resonance structures of [C4O4]2–, [C4O4]4–, and [C4O4]6– fragments;
(lower) Key frontier orbitals: HOMO-1 and LUMO of [2-K(C7H8)]2 and HOMO of [4-K2(BEt3)]2.The HOMO-1 of [2-K(C7H8)]2 [Figure (lower)]
corresponds to a π-bonding orbital across the central CC linkage
of the [C4O4]4– ligand. The
formal addition of a further pair of electrons would then generate
the [C4O4]6– hexa-anion, which
adopts a butadienetetraolate structure via reduction of the enedione
component to the corresponding bis(enolate). The structural changes
associated with the CC and CO linkages on transitioning from [C4O4]4– to [C4O4]6– are consistent with the form of the
LUMO of [2-K(C7H8)]2 (and the HOMO of [4-K2(BEt3)]2), which features out-of-phase π interactions in the
terminal CO and central CC units (Figure ).The overall transition from [C4O4]2– to [C4O4]4– to [C4O4]6– is characterized by net bond cleavage
(C–C/C–O bonds: 11 to 10 to 9), consistent (at a simplistic
level) with successive filling of anti-bonding molecular orbitals.
A comparison between the closely related [C4O4]4– and [C4O4]6– fragments in [2-K(C7H8)]2 and [4-K2(BEt3)]2, respectively, reflects the net conversion of C–O bonds to
C–C bonds in the presence of an additional reductant. The roles
of the [Et3BH]− reagent in bringing about
this transformation (acting as a source of both electrons and of the
BEt3 fragment bound at C–O–) suggest
that it acts as a surrogate for H2 in the Fischer–Tropsch
process.[2]
Conclusions
We
have shown here that four- or six-electron reduction of CO can
be accomplished by the use of anionic aluminum(I) (“aluminyl”)
compounds of the type M[(NON)Al] to give topologically linear or branched
C4/C6 chains depending on the reaction conditions.
The mechanism for homologation to the [C4O4]4– system proceeds via the rate-limiting formation of
the central C=C double bond from two carbene fragments and
rationalizes the synthesis of this system from both monomeric (M =
Li) and dimeric (M = K) precursors. Initial adduct formation with
CO relies on the highly electron-rich nature of the aluminyl reagent,
which drives an unusual (primarily Z-type) mode of interaction of
the CO molecule with the metal center. The onward formation of [C6O6]4– from [C4O4]4– demonstrates for the first time a homogeneous
process which brings about chain branching via complete C–O
bond cleavage. A comparison of the linear [C4O4]4– system with the [C4O4]6– congener formed under more reducing conditions
reflects the net conversion of C–O bonds to C–C bonds
in the presence of an additional reductant.
Experimental
Section
Complete details of the synthetic methods and characterizing
data,
crystallographic information, and details of quantum chemical studies
are provided in the Supporting Information. Starting materials NONAlI,[51a] K2[NONAl]2,[51a] [Li(OEt2)2] [NONAl],[57] and Ag2(squarate)[65] were prepared according to literature methods. All other
reagents were used as received.
[K(C7H8)]2[{(NON)Al}2(C4O4)], [2-K(C7H8)]2
To a 25 mL reaction
bomb were added
[1-K]2 (250 mg, 0.34 mmol) and benzene (7.5
mL). The resulting mixture was degassed twice using the freeze–pump–thaw
method and sealed under vacuum. The bomb was then warmed to 307 K
in an oil bath, and after a few minutes, the headspace was charged
with CO (1.5 atm, ca. 1.2 mmol). After sealing the vessel, the mixture
was vigorously shaken for 30 s and subsequently allowed to stand undisturbed
at 307 K for 16 h. Recrystallization from toluene led to the formation
of small orange crystals of [2-K(C7H8)]2·4(C7H8); isolated yield
of single crystals: 37 mg, 11%. Anal. Calcd for C98H124Al2K2N4O6 +
1.5 toluene: C, 75.57; H, 7.95; N, 3.25. Found: C, 75.52; H, 8.04;
N, 3.22. 1H NMR (400 MHz, THF-d8): δ 7.07–6.95 (m, 6H, Dipp-Ar-CH),
6.42 (d, J = 1.9 Hz, 2H, XA-CH1), 5.62 (d, J = 1.9 Hz, 2H, XA-CH3), 3.30–3.14 (m, 4H, CHMe2), 1.73 (s, 3H, C(CH3)2), 1.56 (s, 3H, C(CH3)2), 1.11 (d, J = 6.8 Hz, 6H, CH(CH3)2), 1.04 (s, 18H, C(CH3)3), 0.86 (d, J = 6.8 Hz, 6H, CH(CH3)2), 0.68 (d, J = 6.8 Hz,
6H, CH(CH3)2), 0.43 (d, J = 6.8 Hz, 6H, CH(CH3)2). 13C NMR (126 MHz, THF-d8): δ 148.2 (Dipp-o-C), 147.8 (CtBu), 147.6 (Dipp-o-C), 146.8 (XA-CN), 144.8
(Dipp-i-C), 142.5 (XA-CO), 133.8 (CCMe2), 131.3 (C(=O)CO), 129.2 (C6H6), 125.4 (Dipp-Ar-CH), 125.2 (Dipp-Ar-CH), 123.6 (Dipp-m-CH), 111.0 (XA-C3H), 105.3 (XA-C1H), 37.9 (CCMe2), 35.5 (CMe3), 32.6 (C(CH3)2), 32.2 (C(CH3)3), 29.1 (CH(CH3)2), 28.5 (CH(CH3)2), 26.3 (CH(CH3)2),
25.5* (CH(CH3)2), 24.7 (CH(CH3)2), 24.5 (CH(CH3)2), 23.8 (C(CH3)2). *Overlapped with solvent signal—located in
HSQC and HMBC. Due to quadrupolar broadening, the resonance associated
with the Al–C(=O) group was not observed
in the range −30 to 300 ppm. The corresponding signal is measured
for the related derivative [2-K(12-crown-4)]2 at 296.2 ppm.
[K(12-c-4)]2[{(NON)Al}2(C4O4)], [2-K(12-c-4)]2
(Method
1): To a suspension of [2-K(C7H8)]2 (15 mg, 0.009 mmol) in benzene-d6 (0.5 ml) was added a solution of 12-crown-4 in toluene (0.18
mL, 0.1 M, 0.018 mmol, 2.05 equiv), and the reaction mixture was heated
at 358 K for 2 h. Volatiles were removed under vacuum, and the product
was obtained as an orange powder. Yield 13 mg, 80% (quantitative by
NMR). (Method 2): To a solution of [1-K]2 (0.120
g, 0.163 mmol) in toluene (5 mL) was added a solution of 12-crown-4
in toluene (3.2 mL of a 0.1 M solution, 0.320 mmol). The resulting
red solution was stirred at room temperature for 5 min before addition
of Fe(CO)5 (22 μL, 0.161 mmol). The resulting dark
red solution was stirred overnight. Removal of volatiles yields a
yellow oil, which can be dissolved in minimal benzene and stored at
room temperature overnight to yield single crystals of [2-K(12-c-4)]2·4(C6H6) suitable
for X-ray crystallography. Isolated yield of single crystals ca. 5
mg, 20% (30% by NMR). 1H NMR (600 MHz, THF-d8): δ 7.01 (t, J = 7.0 Hz, 2H,
Dipp-p-CH), 6.96 (d, J = 7.0 Hz, 4H, Dipp-m-CH), 6.40
(d, J = 2.0 Hz, 2H, XA-CH1), 5.59 (d, J = 1.9 Hz, 2H, XA-CH3), 3.60 (s, 16H, 12-crown-4), 3.42–3.34 (m, 2H,
CHMe2), 3.14 (sept., J = 6.9 Hz, 2H, CHMe2), 1.73 (s, 3H, C(CH3)2), 1.58 (s, 3H, C(CH3)2), 1.24 (d, J = 6.7 Hz,
6H, CH(CH3)2), 1.03 (s, 18H,
C(CH3)3), 0.79 (d, J = 6.8 Hz, 6H, CH(CH3)2), 0.77 (d, J = 6.8 Hz, 6H, CH(CH3)2), 0.33 (d, J = 6.8 Hz,
6H, CH(CH3)2). 13C NMR (151 MHz, THF-d8): δ 296.2
(AlCO), 147.7 (Dipp-o-C), 147.5 (Dipp-o-C), 147.4 (CtBu), 146.9 (XA-CN), 145.2
(Dipp-i-C), 142.7 (XA-CO), 134.1 (CCMe2), 130.8 (C(=O)CO), 125.9 (Dipp-m-CH),
125.1 (Dipp-p-CH), 123.1 (Dipp-m-CH), 111.4 (XA-C3H), 105.3 (XA-C1H), 70.5 (OCH2), 37.8 (CMe2),
35.4 (C(CH3)3), 32.2 (C(CH3)3), 32.0 (C(CH3)2), 29.4 (CHMe2), 28.7 (CHMe2), 26.5 (CH(CH3)2), 25.0 (CH(CH3)2), 24.6 (CH(CH3)2), 24.3 (CH(CH3)2), 23.6 (C(CH3)2).
[Li(OEt2)][{(NON)Al}2(C4O4)], [2-Li(OEt2)]2 and [Li(THF)]2[{(NON)Al}2(C4O4)], [2-Li(THF)]2
Ether Adduct
(Method 1) A solution of [1-Li]
(15 mg, 0.017 mmol) in C6D6 (0.5 mL) was
degassed twice using the freeze–pump–thaw method. The
headspace was then charged with CO (2 bar) and briefly shaken. 1H NMR reveals the formation of a mixture of products, including
[2-Li(OEt2)]2 in ca. 40% yield.
Upon concentration (to ca. 1/3 volume) and standing for 2 days, X-ray
quality crystals of [2-Li(OEt2)]2 formed.(Method 2) [2-K(C7H8)]2 (30 mg, 0.018 mmol) and LiI (5 mg, 2.1 equiv) were
dissolved in Et2O (15 mL) and briefly heated to 318 K in
a sealed ampoule, leading to a color change from orange to yellow.
The mixture was allowed to cool to room temperature, and the resulting
cloudy mixture was filtered onto benzene (8 mL). After mixing, the
solution was concentrated until it turned turbid (ca. 5 mL). At this
point, a sample (0.2 mL) was taken and diluted with C6D6 (0.3 mL), and a 1H NMR spectrum was measured.
The resulting spectrum shows the presence of a single species, which
has the same spectroscopic signals as those obtained using method
1 in a J. Young’s NMR tube. 1H NMR (400 MHz, C6D6): δ 6.67 (d, J = 2.0
Hz, 2H, XA-CH1), 6.00 (d, J = 2.0 Hz, 2H, XA-CH3), 3.72 (sept, J = 6.9 Hz, 2H), 1.70 (s, 3H, C(CH3)2), 1.61 (s, 3H, C(CH3)2), 1.52 (d, J = 6.9 Hz, 6H, CH(CH3)2), 1.19 (s, 18H, C(CH3)3), 0.36 (d, J = 6.9 Hz,
6H, CH(CH3)2). Samples of [2-Li(OEt2)]2 lose Et2O under
continuous vacuum. Dissolution in THF-d8, however, yields spectroscopic signals indistinguishable from samples
of [2-Li(THF)]2 prepared as set out below.
THF Adduct
A solution containing [2-K(C7H8)]2 (ca. 20 mg, 0.012 mmol) was prepared
as described above, and Li[HBsBu3] (1 mL of
a 1 M solution in THF) was added. The mixture was heated at 307 K
for 16 h, affording small yellow crystals. Analytically pure samples
can be obtained by recrystallization from minimal benzene or THF and
washing with THF. Isolated yield of single crystals: 7 mg, 30% over
last step. Anal. Calcd for C106H140Al2Li2N4O8 + 1.75 THF: C, 75.73; H,
8.66; N, 3.13. Found: C, 75.13; H, 8.66; N, 3.21. 1H NMR
(400 MHz, C6D6): δ 7.25 (m, 6H, Dipp-Ar-CH), 6.70 (d, J = 2.0 Hz, 2H, XA-CH1), 6.05 (d, J = 2.0 Hz, 2H,
XA-CH3), 3.77 (sept, J = 6.9 Hz, 2H, CHMe2), 3.70 (s, m, 4H,
THF), 3.41 (sept, J = 6.9 Hz, 2H, CHMe2), 1.69 (s, 3H, C(CH3)2), 1.60 (s, 3H, C(CH3)2), 1.58 (d, J = 6.9 Hz, 6H, CH(CH3)2), 1.32 (m, 4H, THF), 1.21 (s, 18H, C(CH3)3), 1.15 (overlapping d, J = 6.9 Hz, 12H, CH(CH3)2), 0.36 (d, J = 6.9 Hz, 6H, CH(CH3)2). 13C NMR (151 MHz, C6D6): δ 147.8 (CtBu),
147.2 (Dipp-o-C), 146.3 (Dipp-o-C), 145.7 (XA-C4), 143.6 (Dipp-i-C), 141.4
(XA-CO), 133.1 (CCMe2), 125.8 (Dipp-p-C and Dipp-m-C), 123.0 (Dipp-m-C), 111.3 (XA-C3H), 106.4 (XA-C1H), 68.6 (THF-CO), 37.0 (CMe2), 35.0 (CMe3), 32.1 (C(CH3)2), 31.9 (C(CH3)3), 29.5 (CHMe2), 28.2 (CHMe2), 26.3 (CH(CH3)2), 25.4 (THF), 24.1 (CH(CH3)2), 23.9 (CH(CH3)2), 23.2 (C(CH3)2), 23.2 (CH(CH3)2). Due to quadrupolar broadening (and overlap with the solvent),
the resonances associated with the [C4O4]4– fragment were not observed in the range −30
to 300 ppm. The corresponding signals are measured for the related
derivative [2-K(12-crown-4)]2 at 130.8 and
296.2 ppm. 7Li NMR (156 MHz, C6D6) δ 0.61.
K[K(THF)2][{(NON)Al}2(C6O6)], 3
[2-K(C7H8)]2 (25 mg, 0.015 mmol) was
suspended in
THF-d8 (0.5 mL) in a J. Young’s
NMR tube and degassed twice using the freeze–pump–thaw
method. The headspace was then charged with CO (2 atm) and heated
at 339 K for 4 d. During this time, [2-K(C7H8)]2 dissolved, and a color change from yellow
to red to orange was observed. 1H NMR monitoring shows
quantitative conversion to a single product. After concentrating the
solution to a quarter of its original, small pale orange plates formed
at the surface of the solution. Anal. Calcd for C108H140Al2K2N4O10 +
2 THF: C, 72.16; H, 8.14; N, 2.90. Found: C, 72.13; H, 8.06; N, 2.55. 1H NMR (600 MHz, THF-d8): δ
7.43 (t, J = 7.6 Hz, 2H, Dipp-p-CH), 7.32 (t, J = 7.6 Hz, 2H, Dipp-p-CH), 7.25 (dd, J = 7.8,
1.6 Hz, 2H, Dipp-m-CH), 7.16 (dd, J = 7.6, 1.6 Hz, 2H, Dipp-m-CH), 7.09 (dd, J = 7.8, 1.6 Hz, 2H, Dipp-m-CH), 7.00 (dd, J = 7.7, 1.6 Hz,
2H, Dipp-m-CH), 6.49 (d, J = 1.9 Hz, 2H, XA-C1H), 6.42
(d, J = 1.9 Hz, 2H, XA-C1H), 5.36 (d, J = 1.9 Hz, 2H, XA-C3H), 5.34 (d, J = 1.9 Hz, 2H, XA-C3H), 3.46–3.36 (overlapping sept., 4H, CHMe2), 3.32 (sept., J = 7.2
Hz, 2H, CHMe2), 3.16 (sept., J = 6.9 Hz, 2H, CHMe2), 1.75 (s, 3H, C(CH3)2), 1.73 (s, 3H, C(CH3)2), 1.57 (s, 3H, C(CH3)2), 1.54 (s, 3H, C(CH3)2), 1.20 (d, J = 7.0 Hz, 6H, CH(CH3)2), 1.03 (s, 18H, C(CH3)3), 1.00 (s, 18H, C(CH3)3), 0.95 (d, J = 6.9 Hz, 6H,
CH(CH3)2), 0.87 (d, J = 6.7 Hz, 6H, CH(CH3)2), 0.83 (d, J = 6.8 Hz, 6H, CH(CH3)2), 0.80 (d, J = 6.7 Hz,
6H, CH(CH3)2), 0.73 (d, J = 6.7 Hz, 6H, CH(CH3)2), 0.68 (d, J = 6.8 Hz, 6H, CH(CH3)2), 0.51 (d, J = 6.8 Hz,
6H, CH(CH3)2). 13C NMR (151 MHz, THF-d8): δ 258.6
(C6O6), 173.5 (C6O6),
169.0 (C6O6), 156.2 (C6O6), 149.4 (Dipp-o-C), 149.1 (Dipp-o-C), 148.0 (Dipp-o-C), 147.6 (CtBu), 147.5 (Dipp-o-C), 147.3 (CtBu), 146.1 (XA-C4), 145.8 (Dipp-i-C), 145.7 (Dipp-i-C), 145.6 (XA-C4), 142.7 (XA-CO), 141.5 (XA-CO), 133.7 (CCMe2), 133.3 (CCMe2), 129.2
(free C6H6), 127.4 (Dipp-p-CH), 127.3 (Dipp-p-CH),
127.2 (Dipp-m-CH), 126.2 (Dipp-m-CH), 124.4 (Dipp-m-CH), 123.8 (Dipp-m-CH),
118.2 (C6O6), 111.3 (XA-C3H), 110.5 (C6O6), 106.5 (XA-C1H), 105.9 (XA-C1H), 37.9 (CMe2), 37.8 (CMe2), 35.5 (CMe3), 35.4 (CMe3), 32.2 (C(CH3)3), 32.1 (C(CH3)3), 32.1 (C(CH3)2), 32.0 (C(CH3)2), 29.2 (CHMe2), 28.8 (CHMe2), 28.1 (CHMe2), 28.0 (CHMe2), 26.2 (CH(CH3)2), 26.1 (CH(CH3)2), 26.0 (CH(CH3)2), 25.3 (CH(CH3)2), 25.2 (CH(CH3)2), 24.9 (CH(CH3)2), 22.9 (C(CH3)2), 22.9 (C(CH3)2).
K4[{(NON)Al}2(C4O4) (BEt3)2], [4-K2(BEt3)]2
To a stirred solution containing [1-K]2 (220 mg, 0.30 mmol), benzene (6.6 mL), and
CO (ca. 1.5 bar) in a 25 mL reaction bomb was added K[HBEt3] (1.5 mL of a 1 M solution in THF, 1.5 mmol). After 24 h, the solution
was concentrated to a fifth of its original volume under reduced pressure.
Prolonged standing led to the formation of colorless extremely sensitive
crystals of [4-K2(BEt3)]2. Yield 12 mg, 4%. 1H NMR (400 MHz, THF-d8): δ 7.27 (br s, 6H, Dipp-Ar-CH), 6.36 (br s, 2H, XA-CH1), 5.28 (s,
2H, XA-CH3), 3.79 (br s, 2H, CHMe2), 3.47 (sept, J = 6.6 Hz,
2H, CHMe2), 1.68 (s, 6H, C(CH3)2), 1.28 (br s, 6H, CH(CH3)2), 1.00 (s, 18H, C(CH3)3), 0.80 (d, J = 6.6 Hz, 13H),
0.57 (t, J = 7.7 Hz, 9H, BCH2CH3), 0.06 (q, J = 7.7 Hz, 6H,
BCH2CH3). Due to the low solubility
and stability of this compound in solution, no satisfactory 11B and 13C spectra could be obtained.
[(NON)Al]2(C4O4), 5
(NON)AlI
(500 mg, 0.60 mmol, 2.0 equiv) and Ag2(C4O4) (110 mg, 0.33 mmol) were suspended in
Et2O (10 mL), and the reaction mixture was refluxed at
308 K for 48 h in the dark. Volatiles were removed in vacuo, and the
residue was extracted into toluene (15 mL), filtered, and layered
with hexane (15 mL). Upon standing for several days, crystals of 5 suitable for single-crystal X-ray diffraction were obtained.
Yield 180 mg, 39%. Anal. Calcd for C98H124Al2N4O6 + 0.5 hexane: C, 78.21; H, 8.51;
N, 3.61. Found 78.08; H, 8.76; N, 3.34. 1H NMR (400 MHz,
C6D6): δ 7.20 (m, 2H, Dipp-p-CH), 7.14 (m, 4H, Dipp-p-CH), 6.72 (d, J = 1.9 Hz, 2H, XA-CH1), 5.98 (d, J = 1.9 Hz, 2H,
XA-CH3), 3.54 (sept, J = 6.8 Hz, 4H, CHMe2), 1.58 (s, 6H, C(CH3)2), 1.15 (s, 18H, C(CH3)3), 1.06 (d, J = 6.8 Hz,
12H, CH(CH3)2), 0.91–0.83
(br s, 12H, CH(CH3)2). 13C NMR (126 MHz, C6D6): δ 194.0
(C4O4), 148.9 (CtBu), 147.0 (Dipp-o-C), 144.4 (XA-CN), 141.0, 140.8 (Dipp-i-C, XA-CO), 133.3 (CCMe2), 127.0 (Dipp-p-C), 125.1 (Dipp-m-C), 112.1 (XA-C3H), 108.2 (XA-C1H), 37.2 (CMe2), 35.1(CMe3), 31.7 (C(CH3)3), 28.4 (CHMe2), 26.4 (br, C(CH3)2), 25.4 (CH(CH3)2), 24.7(CH(CH3)2).
Authors: Andrey V Protchenko; Petra Vasko; Dinh Cao Huan Do; Jamie Hicks; M Ángeles Fuentes; Cameron Jones; Simon Aldridge Journal: Angew Chem Int Ed Engl Date: 2019-01-03 Impact factor: 15.336
Authors: Holger Braunschweig; Theresa Dellermann; Rian D Dewhurst; William C Ewing; Kai Hammond; J Oscar C Jimenez-Halla; Thomas Kramer; Ivo Krummenacher; Jan Mies; Ashwini K Phukan; Alfredo Vargas Journal: Nat Chem Date: 2013-10-13 Impact factor: 24.427
Authors: Owen T Summerscales; F Geoffrey N Cloke; Peter B Hitchcock; Jennifer C Green; Nilay Hazari Journal: J Am Chem Soc Date: 2006-08-02 Impact factor: 15.419
Authors: Owen T Summerscales; F Geoffrey N Cloke; Peter B Hitchcock; Jennifer C Green; Nilay Hazari Journal: Science Date: 2006-02-10 Impact factor: 47.728
Authors: Helen R Sharpe; Ana M Geer; Laurence J Taylor; Benjamin M Gridley; Toby J Blundell; Alexander J Blake; E Stephen Davies; William Lewis; Jonathan McMaster; David Robinson; Deborah L Kays Journal: Nat Commun Date: 2018-09-14 Impact factor: 14.919
Authors: Jamie Hicks; Petra Vasko; Andreas Heilmann; Jose M Goicoechea; Simon Aldridge Journal: Angew Chem Int Ed Engl Date: 2020-09-02 Impact factor: 15.336
Authors: Samuel Grams; Jonathan Eyselein; Jens Langer; Christian Färber; Sjoerd Harder Journal: Angew Chem Int Ed Engl Date: 2020-05-25 Impact factor: 15.336