4894 Organometallics, Vol. 25, No. 20, 2006
Mitu and Baird
1
Propionic Acid. A series of H NMR experiments in CD2-
Note that the 11B chemical shift of the new species is similar to
that of [Me4N][MeCO2{B(C6F5)3}].6e
Cl2 and involving B(C6F5)3 and propionic acid in various
B(C6F5)3:CH3CH2CO2H molar ratios (0.25:1-2.75:1) were
carried out at -30 °C. Representative spectra in the range δ
1-14 are shown in Figure 2, and expanded spectra in the range
δ 1-3 in Figure 3. In view of the errors involved in measuring
the miniscule quantities of reagents involved, the ratios varied
somewhat from the nominal ratios stated in the Experimental
Section and in Figure 2; however, the trends in the ratios are
undoubtedly as stated.
n-Decanoic Acid. On treating a CD2Cl2 solution of n-
decanoic acid with B(C6F5)3 (nominal B(C6F5)3:n-decanoic acid
ratios ∼0.5:1-2:1) at -30 °C, the OH resonance of the free
acid at δ 12.56 first broadened then narrowed and shifted
ultimately to δ ∼9.5, as shown in Figure 7. Interestingly, at a
B(C6F5)3:n-decanoic acid ratio of ∼0.5:1, the CH3(CH2)7CH2-
CO2 resonance at δ 2.32 decreased in intensity and a new
resonance, of comparable intensity and attributable to the CH3-
(CH2)7CH2CO2 resonance of the 1:1 adduct, appeared at δ 2.61
(Figure 7b). Similarly the CH3(CH2)6CH2CH2CO2 resonance of
the free acid at δ 1.55 split into two resonances of comparable
intensity at δ 1.68 and 1.48. The resonance at δ 1.68 is pre-
sumably attributed to the CH3(CH2)6CH2CH2CO2 resonance of
the 1:1 adduct, and that at δ 1.48 possibly to free n-decanoic
acid hydrogen bonded to the adduct since the chemical shift is
not identical to that of the free acid. In Figures 7c,d, we show
the spectra of n-decanoic acid in the presence of 1.5 and 2 equiv
of B(C6F5)3, respectively. There is in both spectra clearly only
single sets of CH3(CH2)7CH2CO2 and CH3(CH2)6CH2CH2CO2
resonances, the chemical shifts of which correspond closely with
the chemical shifts of the pair of downfield resonances in Fig-
ure 7b and strengthen their assignments to the 1:1 adduct.
Changes in the aliphatic region were more subtle but, as can
be seen in Figure 3a, addition of ∼0.25 molar equivalent of
B(C6F5)3 to a solution of propionic acid resulted in the appear-
ance of very broad, weak methyl and methylene resonances at
δ ∼1.18 and ∼2.56, respectively. Both exhibited intensities
approximately one-third those of the corresponding methyl and
methylene resonances of the free acid although the breadth of
the lines made accurate measurements impossible. When the
B(C6F5)3:propionic acid ratio was increased to ∼0.5:1, the new
resonances shifted to δ 1.26 and 2.74 and gained in intensity
such that the ratio of their intensities to those of the free acid
resonances was now ∼0.5:1 (Figure 3b). (If, as is concluded
below, the new species is a 1:1 adduct, then these relative
integrations suggest that the nominal 2:1 ratio of reactants had
not been achieved exactly.)
These results are consistent with those discussed above for
propionic acid, although the chemical shift differences between
the CH3(CH2)nCH2CO2 and CH3(CH2)n-1CH2CH2CO2 reso-
nances of the free and coordinated acids are greater for
n-decanoic acid. 19F NMR spectra of the same B(C6F5)3/n-
decanoic acid solutions (not shown) also exhibited separate
resonances for the three pairs of ortho-F (δ -136.3), para-F
(δ -159.7), and meta-F (δ -166.0) resonances. These chemical
shifts are very similar to those reported above for the 1:1 pro-
pionic acid adduct.
Increasing the ratio to ∼1:1 resulted in disappearance of the
resonances of the free acid and observation of only the new
resonances (Figure 3d), while addition of excess B(C6F5)3
(up to a 2.75:1 ratio) resulted in no other changes in the spec-
trum. Interestingly, when the B(C6F5)3:propionic acid ratio was
∼0.75:1, the pairs of methylene and methyl resonances coa-
lesced (Figure 3c).
These experiments were complemented by 19F, 13C, and 11
B
NMR spectra of solutions (all in CD2Cl2 at -30 °C) containing
B(C6F5)3 and propionic acid in various ratios. On the addition
of approximately 1 equiv of propionic acid to a solution of
B(C6F5)3, the resonances of free B(C6F5)3 at δ -128.6 (ortho-
F), -144.3 (para-F), and -158.7 (meta-F) (Figure 4a) disap-
peared and a new set of somewhat broadened resonances at δ
-136.9 (ortho-F), -158.7 (para-F), and -165.4 (meta-F)
appeared (Figure 4b); these 19F chemical shifts are very similar
to those of the acetate adduct [Me4N][MeCO2{B(C6F5)3}].6e The
spectrum of a 2:1 reaction mixture (Figure 4c) exhibited both
this set of resonances and also those of free B(C6F5)3 at δ
-128.6 (ortho-F), -144.3 (para-F), and -158.7 (meta-F), the
three pairs of ortho-, meta-, and para-resonances being of
comparable intensities and somewhat broadened because of
exchange. Several sharp, weak resonances were also observed,
presumably a result of slow cleavage reactions of the type
discussed above.
The 13C NMR experiments were carried out using propionic
acid enriched in 13C at the carboxylic carbon. On the addition
of ∼0.75 equiv of B(C6F5)3 to a solution of propionic acid, the
resonance of free acid at δ 182 weakened, broadened, and shifted
somewhat while a new, broad resonance at δ ∼189 appeared
(Figure 5b). When the ratio of B(C6F5)3 to propionic acid was
1:1 and higher, only the (now sharp) resonance at δ 189
remained (Figures 5c-e). Similar 11B NMR experiments (Fig-
ure 6) showed that the broad resonance of free B(C6F5)3 at δ
62.0 was replaced completely by a new, somewhat narrower
resonance at δ 0.2 in the presence of g1 equiv of propionic
acid, but that the two resonances were present in comparable
intensities when the B(C6F5)3 to propionic acid ratio was 2:1.
Nature of the Adduct(s) in Solution. On the basis of the
IR spectroscopic data for propionic acid, it seems that a 1:1
adduct, C2H5CO2H‚B(C6F5)3, is formed as in Scheme 1 (A, R
) CH3CH2), although slow, proton-induced B-C6F5 cleavage
even at low temperatures impaired all attempts to grow crys-
tallographically useful crystals. The IR spectral data show clearly
that ν(OH) and ν(CdO) of free propionic acid disappear
completely on the addition of 1 molar equiv of B(C6F5)3 and
that addition of excess B(C6F5)3 results in little further change
in the IR spectrum. It was disturbing initially when we found
that the adduct did not exhibit ν(CdO) in the range 1666-
1700 cm-1 as anticipated on the basis of IR data for 1:1 adducts
of B(C6F5)3 with other carbonyl Lewis bases,11 but we success-
fully prepared and characterized spectroscopically and crystallo-
graphically the corresponding 1:1 adduct of methyl propionate,
C2H5CO2Me‚B(C6F5)3 (see below). The ester in this compound
coordinates to the borane via the carbonyl oxygen atom, as in
the corresponding ethyl benzoate adduct,11 but ν(CdO) shifts
from 1746 cm-1 in the free ester to 1601 cm-1 on coordination,
a ∆ν(CdO) of 145 cm-1, which contrasts markedly with the
∆ν(CdO) of 49 cm-1 reported previously for the 1:1 adduct of
ethyl benzoate.11 The reasons for this apparent discrepancy are
discussed below, but observation of an apparent ν(CdO) of
C2H5CO2H‚B(C6F5)3 at 1561 cm-1 now seems quite reasonable.
Turning now to the 1H, 19F, 11B, and 13C NMR spectroscopic
evidence for the carboxylic acid systems, the data support the
conclusion that 1:1 adducts are formed essentially quantitatively.
For both propionic and n-decanoic acids, addition of ∼0.5 equiv
of B(C6F5)3 to a CD2Cl2 solution of the acid at -30 °C resulted
in spectra that exhibited resonances of free and coordinated acid