242
A. R. Jupp
borane (HB(C 6 F 5 ) 2 ) to 1 yielded the Lewis adducts 2 and 3, respectively (Fig. 7.2a).
A range of solid-state NMR techniques confirmed the presence of the P–B bond in
both compounds, and diffuse reflectance infrared Fourier transform (DRIFT) spectroscopy showed the diagnostic bands from the C 6 F 5 rings in both products, as well as
the B–H stretching frequency at 2405 cm
−1 in 3. Although the original FLP systems
precluded formation of the Lewis adduct through sterics, it has since been shown
that the same FLP reactivity can be observed if an equilibrium can be established
between the “free” and “bound” acid and base components [31]. 2 and 3 were therefore explored as heterogeneous catalysts for the selective hydrogenation of 3-hexyne
to Z-3-hexene (Fig. 7.2b). Although 2 gave rather poor conversion under the conditions tried, 3 was an efficient catalyst for this reduction, giving quantitative conversion
and high selectivity towards Z-3-hexene (approximately 84%), with small amounts
of E-3-hexene, Z-2-hexene, E-2-hexene and hexane also produced. The significantly
enhanced catalytic ability of 3 over 2 led the authors to propose that the mechanism proceeds by initial hydroboration of the alkyne, followed by hydrogen splitting
and protonation, as previously suggested by Pápai and Repo for a different FLP
system [32]. Recycling studies of 3 as a catalyst were also carried out, leading to
a drastic drop in conversion, but when the Piers’ borane was refreshed in between
runs then the high conversion was maintained. Finally, the catalytic activity of 3
was also studied in the hydrogenation of other alkynes, namely diphenylacetylene,
(hex-2-yn-1-yloxy)trimethylsilane and methyl octadec-9-ynoate, but in all cases the
conversion was low (5–25%).
Thomas et al. employed a different approach to obtain a solid Lewis base [33].
Instead of grafting the phosphine to a solid support, they synthesised porous polymer
networks based on two different triarylphosphine moieties with varying degrees of
steric hindrance close to the phosphorus centre. Accordingly, a Yamamoto polymerisation process was used to afford 4 and 5 (Fig. 7.3a) [34]. Suspension of the polymer
networks with B(C 6 F 5 ) 3 in DCM led to rapid swelling of the polymer and impregnation of the Lewis acidic borane into each porous network, to afford adducts 6 and
7. The formation of an adduct is observed in the solid state in each case by MASNMR (MAS: magic angle spinning) spectroscopy. This may be initially surprising,
as the “molecular” counterpart to 7, modelled as a combination of P(2,6-Me 2 -C 6 H 3 ) 3
and B(C 6 F 5 ) 3 , is fully frustrated in solution, that is the NMR chemical shifts of the
individual components are unchanged on mixing. But on removal of the solvent, the
“dry” solid mixture of the acid and the base has a non-negligible interaction, and this
is shown in the solid-state NMR spectrum of the dry mixture, which is distinct from
each separate component. Although the P–B interaction is present in 6 and 7, the
coordinating solvents (such as Et 2 O, THF, methanol or acetone) preferentially bind
to the borane, and thus simple washing of 6 and 7 with such solvents restores the
free networks 4 and 5, respectively. The ability of these heterogeneous systems to
activate dihydrogen was subsequently explored. A combination of either 4 or 5 with
equimolar B(C 6 F 5 ) 3 in cyclohexane-d 12 under a pressure (6 bar) of 1:1 H 2 /D 2 led
to the reversible activation of dihydrogen, as observed by the emergence of a 1:1:1
triplet signal attributable to HD in the
1 H NMR spectrum at 4.54 ppm (Fig. 7.3b).
The isotopic scrambling indicates that these species could be promising candidates
A. R. Jupp
borane (HB(C 6 F 5 ) 2 ) to 1 yielded the Lewis adducts 2 and 3, respectively (Fig. 7.2a).
A range of solid-state NMR techniques confirmed the presence of the P–B bond in
both compounds, and diffuse reflectance infrared Fourier transform (DRIFT) spectroscopy showed the diagnostic bands from the C 6 F 5 rings in both products, as well as
the B–H stretching frequency at 2405 cm
−1 in 3. Although the original FLP systems
precluded formation of the Lewis adduct through sterics, it has since been shown
that the same FLP reactivity can be observed if an equilibrium can be established
between the “free” and “bound” acid and base components [31]. 2 and 3 were therefore explored as heterogeneous catalysts for the selective hydrogenation of 3-hexyne
to Z-3-hexene (Fig. 7.2b). Although 2 gave rather poor conversion under the conditions tried, 3 was an efficient catalyst for this reduction, giving quantitative conversion
and high selectivity towards Z-3-hexene (approximately 84%), with small amounts
of E-3-hexene, Z-2-hexene, E-2-hexene and hexane also produced. The significantly
enhanced catalytic ability of 3 over 2 led the authors to propose that the mechanism proceeds by initial hydroboration of the alkyne, followed by hydrogen splitting
and protonation, as previously suggested by Pápai and Repo for a different FLP
system [32]. Recycling studies of 3 as a catalyst were also carried out, leading to
a drastic drop in conversion, but when the Piers’ borane was refreshed in between
runs then the high conversion was maintained. Finally, the catalytic activity of 3
was also studied in the hydrogenation of other alkynes, namely diphenylacetylene,
(hex-2-yn-1-yloxy)trimethylsilane and methyl octadec-9-ynoate, but in all cases the
conversion was low (5–25%).
Thomas et al. employed a different approach to obtain a solid Lewis base [33].
Instead of grafting the phosphine to a solid support, they synthesised porous polymer
networks based on two different triarylphosphine moieties with varying degrees of
steric hindrance close to the phosphorus centre. Accordingly, a Yamamoto polymerisation process was used to afford 4 and 5 (Fig. 7.3a) [34]. Suspension of the polymer
networks with B(C 6 F 5 ) 3 in DCM led to rapid swelling of the polymer and impregnation of the Lewis acidic borane into each porous network, to afford adducts 6 and
7. The formation of an adduct is observed in the solid state in each case by MASNMR (MAS: magic angle spinning) spectroscopy. This may be initially surprising,
as the “molecular” counterpart to 7, modelled as a combination of P(2,6-Me 2 -C 6 H 3 ) 3
and B(C 6 F 5 ) 3 , is fully frustrated in solution, that is the NMR chemical shifts of the
individual components are unchanged on mixing. But on removal of the solvent, the
“dry” solid mixture of the acid and the base has a non-negligible interaction, and this
is shown in the solid-state NMR spectrum of the dry mixture, which is distinct from
each separate component. Although the P–B interaction is present in 6 and 7, the
coordinating solvents (such as Et 2 O, THF, methanol or acetone) preferentially bind
to the borane, and thus simple washing of 6 and 7 with such solvents restores the
free networks 4 and 5, respectively. The ability of these heterogeneous systems to
activate dihydrogen was subsequently explored. A combination of either 4 or 5 with
equimolar B(C 6 F 5 ) 3 in cyclohexane-d 12 under a pressure (6 bar) of 1:1 H 2 /D 2 led
to the reversible activation of dihydrogen, as observed by the emergence of a 1:1:1
triplet signal attributable to HD in the
1 H NMR spectrum at 4.54 ppm (Fig. 7.3b).
The isotopic scrambling indicates that these species could be promising candidates