1-acyl-4-benzylpyridinium tetrafluoroborates: Stability, structural properties, and utilization for the synthesis of acyl fluorides

Article abstract of DOI:10.1002/(SICI)1099-0690(199909)1999:9<2383::AID-EJOC2383>3.0.CO;2-T

1-Acyl-4-benzylpyridinium salts 4 containing nonnucleophilic anions X^- such as CF₃SO₃^-, FSO₃^-, and BF₄^- can be generated quantitatively and in situ from 1-acyl-4-alkylidene-1,4-dihydropyridines 1a-f and the corresponding acid, HX. The BF₄^- salts reveal an interesting and unexpected thermal instability which allows the convenient synthesis of carboxylic acid fluorides 5b-f. This procedure offers advantages over known methods: All operations can be performed in a standard glass apparatus and do not require high pressures. The formation of RCOF 5 is assisted by the pyridine moiety of 4, which splits off and functions as a Lewis base to intercept the BF₃ acid. The structural and electronic relationships as well as dominating differences between the very reactive cations of 4 and their almost 'inert' uncharged precursors, the dihydropyridines 1, are discussed both on the fundament of experimental evidence (X-ray structures of 1f and the extremely reactive and very labile 4f) and theoretical investigations (ab initio and DFT MO calculations).

Full text of DOI:10.1002/(SICI)1099-0690(199909)1999:9<2383::AID-EJOC2383>3.0.CO;2-T

FULL PAPER
1
-Acyl-4-benzylpyridinium Tetrafluoroborates: Stability, Structural Properties,
and Utilization for the Synthesis of Acyl Fluorides
[a]
[a]
[a]
[b]
Rüdiger Wagner, Bernd Wiedel, Wolfgang Günther, Helmar Görls,
and Ernst Anders*[
a]
Keywords: 4-Alkylidene-1,4-dihydropyridines / 1-Acyl-4-benzylpyridinium tetrafluoroborates / Carboxylic acid fluorides /
X-ray structures / Ab initio calculations / Density functional calculations
1
-Acyl-4-benzylpyridinium salts 4 containing nonnucleo-
by the pyridine moiety of 4, which splits off and functions as
a Lewis base to intercept the BF acid. The structural and
–
–
–, and BF
–
philic anions X such as CF
3
SO
3
, FSO
3
4
can be
3
generated quantitatively and in situ from 1-acyl-4-
electronic relationships as well as dominating differences
between the very reactive cations of 4 and their almost
“inert“ uncharged precursors, the dihydropyridines 1, are
discussed both on the fundament of experimental evidence
(X-ray structures of 1f and the extremely reactive and very
labile 4f) and theoretical investigations (ab initio and DFT
MO calculations).
alkylidene-1,4-dihydropyridines 1a–f and the corresponding
–
acid, HX. The BF
4
salts reveal an interesting and
unexpected thermal instability which allows the convenient
synthesis of carboxylic acid fluorides 5b–f. This procedure
offers advantages over known methods: All operations can
be performed in a standard glass apparatus and do not
require high pressures. The formation of RCOF 5 is assisted
Introduction
and reactive towards amines, thus providing an effective
peptide coupling method.[ Another application of acyl
fluorides is the stereoselective formation of α- and β-glyco-
syl esters in the presence of cesium fluoride.
Recently, we reported the synthesis and reactivity of the
14]
Acyl fluorides are conventionally prepared from the cor-
responding halides or anhydrides. Various reagents can be
utilized for this transformation. Without thought of com-
pleteness, we summarize some examples: Frequently used
[15]
1
-acylpyridinium salts 3 and 4, which are quantitatively ob-
[1]
inorganic salts like potassium fluoride, potassium hydro-
tained from the readily available and stable 1-acyldihydro-
[2]
[3]
gen fluoride, or potassium fluorosulfinate are just as
pyridine precursors 1 by protonation with strong protonic
suitable as sulfur tetrafluoride, benzoyl fluoride,[5] di-
[4]
[16]
acids 2a,b (Scheme 1).
[6]
alkylaminosulfur trifluorides, or 1,3-dimethyl-2-fluoropy-
[
7]
ridinium salts. In some cases it is possible to obtain aroyl
fluorides from aryl halides which had been treated with car-
bon monoxide and an alkali metal fluoride in the presence
[8]
of palladium phosphane complexes. Since use of anhy-
drous hydrogen fluoride as the fluorinating agent generally
requires superatmospheric pressure, modern fluorination
reactions alternatively employ the less volatile pyridinium
[9]
[10][11]
poly(hydrogen fluoride) or cyanuric fluoride.
Although carboxylic acid chlorides and bromides can be
utilized for standard acylation reactions, the corresponding
fluorides are a useful class of compounds. In special cases,
FriedelϪCrafts reactions proceed with better or even re-
versed regioselectivity if an acyl fluoride/BF system is used
3
[12]
instead of an acyl chloride/AlCl3.
Furthermore, there is
considerable current interest in the smooth formation of
Scheme 1. Synthesis of acylfluorides 5 from 1-acyl-4-benzylpyridi-
[13]
amino acid fluorides from FMOC-protected
amino acid nium tetrafluoroborates 4
chlorides. These fluorides are relatively stable to hydrolysis
Ϫ
The anion X significantly controls the properties of
[a]
Institut für Organische Chemie und Makromolekulare Chemie these compounds. The reactive trifluoromethanesulfonates
der Friedrich-Schiller-Universität
3
were especially suitable for the acylation of sensitive chiral
Humboldtstraße 10, D-07743 Jena, Germany
Fax: (internat.) ϩ49 (0)3641/ 948212;
E-mail: c5eran@rz.uni-jena.de
b]
(secondary) alcohols, while the use of the analogous (even
more reactive) tetrafluoroborates 4 generally resulted in low
yields of the desired esters. A more detailed analysis of the
crude reaction mixtures indicated that this must be due to
[
Institut für Anorganische und Analytische Chemie der
Friedrich-Schiller-Universität
Lessingstraße 8, D-07743 Jena, Germany
Eur. J. Org. Chem. 1999, 2383Ϫ2390

WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1999
1434Ϫ193X/99/0909Ϫ2383 $ 17.50ϩ.50/0
2383

R. Wagner, B. Wiedel, W. Günther, H. Görls, E. Anders
FULL PAPER
Table 1. Acyl fluorides 5bϪf prepared (cf. Scheme 1)
No.
R
method
melt.[a]
T [°C]
decomp.[a]
T [°C]
yield[b]
[%]
b.p. [°C(mbar)]
ref.
5
b
c
trans-C
6
H
5
CHϭCH
A
B
A
B
A
A
B
B
>116
120
85
45
77
55
64
74
48
84
65 Ϫ 67 (0.7)
84 Ϫ 85 (0.2)
[7], [10]
[3], [7]
5
CH
CH
3
(CH
2
)
11
> 76
Ϫ
> 110
> 120
Ϫ
85
Ϫ
124
127
Ϫ
5
5
d
e
3
20 (1013)
31 (0.5)
[9]
[9]
C
6
H
5
5
f
p-CH
3
OC
6
H
4
Ϫ
Ϫ
78 Ϫ 80 (0.3)
[53]
[
a]
Bath temperature. Ϫ [b] Yields of isolated and purified substances, structures confirmed by H- and 13C-NMR spectra (purity у 98%),
1
other analytical data are in accordance with those given in the literature.
Ϫ
an unexpectedly enhanced tendency of the BF
anion to reactions have been reported by Winter et al. They obtained
4
Ϫ
decompose to F and BF . This decomposition is facili- special titanocene difluorides (containing silylated cyclo-
3
tated by the assistance of a Lewis base.
pentadienyl ligands) from the corresponding titanocene di-
chlorides after reaction with AgBF at room tempera-
4
[19]
ture.
F
A related example for the potential abstraction of
from BF 4 has been investigated by Caputo et al.
Ϫ
Ϫ
[20]
Results and Discussion
They reported that use of acetonitrile as the solvent resulted
The observation that the salts 4 are thermally unstable in the trimethylsilylnitrilium tetrafluoroborate 10. This re-
was the starting point for the development of a synthetically sult was refuted, however, by Bassinndale et al.[ who prov-
useful method for the preparation of acyl fluorides 5bϪf in ed the nonexistence of 10 as they managed to synthesize 9
21]
good yields, Scheme 1.
quantitatively from a mixture of trimethylchlorosilane and
In general, the extremely moisture-sensitive and reactive AgBF in both solvents. In this reaction, BF was obtained
4
3
salts 4a؊f were synthesized just prior to use, isolated if pos- as a by-product which was weakly coordinated to the sol-
sible, and heated under normal pressure (fluorides with low vent molecules (Scheme 2). These results as well as the ear-
boiling points) or in vacuo (higher boiling fluorides). At lier investigations by Lawton and Levy[ are not surpris-
their specific oil bath temperature (Table 1), the solid mate- ing, since the formation of the extremely strong silicon flu-
rials melted with the formation of an orange liquid. At this orine bond[ is the driving force for course of this reaction.
22]
21]
point, the fluorides 5bϪf were distilled off immediately
Method A). Alternatively, the thermal lability of these pyri- 1-trimethylsilyl-4-benzylpyridinium tetrafluoroborate (7)
dinium salts can be used for an in situ decomposition and (8) verified the lability of such compounds. In case of
Method B) of the compounds 4aϪf. This method is su- salt 7 (Scheme 2), the decomposition reaction is slow at
Previous investigation of 1-(4-methylphenyl)sulfonyl- and
(
(
1
perior for salts 4 which can neither be crystallized nor iso- room temperature and both sulfonic acid esters RSO R
3
lated due to their sensitivity. Furthermore, Method B is ad- and fluorides RSO F are formed in competition when 7 is
2
vantageous if the RCOF product stability demands moder- treated with alcohols.[ The existence of compound 8 has
ate temperatures. This is the case when additional func- been doubted, as there is, as yet, no evidence for its interme-
tional groups are present. Both methods are suitable for the diacy.[ Actually, the sulfonyl fluorides were identified by
synthesis of alkyl, aryl and α,β-unsaturated acyl fluorides mass spectroscopy and the suspected formation of trimeth-
23]
24]
1
and thus appear to be generally applicable. The experimen- ylfluorosilane (9) was derived from H-NMR spectroscopic
tal results for acyl fluorides 5b؊f, obtained either according data. Both decomposition reactions have therefore never
to Method A or B, are summarized in Table 1.
been utilized for preparative purposes.
In the course of our investigations, we detected a limi-
tation. If the 1-pivaloyl derivative 4a was employed, no flu-
oride 5a was detected in the distilled material obtained after
decomposition. This was independent of the decomposition
method employed. Instead nearly (NMR analytical) pure 4-
benzylpyridinium tetrafluoroborate (7) was obtained as the
only residue. The tert-butyl moiety was completely de-
[17]
stroyed.
Ϫ
Although the BF 4 anion normally behaves as a rela-
tively stable and weakly nucleophilic species, its stability is
sometimes overestimated. Some tris(3-tert-butylpyrazo-
lyl)hydroborato MCl (Mn, Fe, Co, Ni) complexes have been
employed by Gorrell and Parkin for the facile abstraction
Ϫ
[18]
of the fluoride from BF 4 ion of AgBF4.
Comparable Scheme 2. Lability of some tetrafluoroborates
2384
Eur. J. Org. Chem. 1999, 2383Ϫ2390

1-Acyl-4-benzylpyridinium Tetrafluoroborates: Stability, Structural Properties
FULL PAPER
At first sight, the properties of the title compounds 4 (an 1H-NMR spectroscopy. The components remained un-
X-ray structure is available for the 1-aroylpyridinium salt changed upon standing for 4 days at 50 °C. No trace of salt
f, vide infra for a detailed discussion), might be compared 4c was detected. We therefore draw the conclusion that the
4
to those of the TMSBF derivative 10. The relatively long decomposition is not an equilibrium reaction but that the
4
ϩ
N ϪCOϪbond [151.4(14) pm] could lead to the assump- rate constant is significantly temperature-dependent. The
tion that the labile salts 4 may dissociate into acylium tetra- reaction was completed after 18 h, when a sample of 4c in
fluoroborates 11b؊f, which then decompose giving the acyl CDCl was kept at 50 °C. Only traces of salt 4c could be
3
fluorides 5bϪf and the adduct 6 [Scheme 3, pathway (a)]. detected after this period. This provides an indication that
Such complexes 11 do indeed exist at low temperatures in yields obtained according to method B might be increased
a more or less nonnucleophilic environment and have been by prolonged heating.
[25]
[26]
intensively investigated by Olah et al.
and Seel.
Never-
We concluded from these results that the tendency of flu-
Ϫ
theless, pathway (a) in Scheme 3 seems to be not very re- oride abstraction from BF 4 is not only due to the acti-
alistic since the intermediate existence of extremely reactive vation of the carbonyl group bonded to the pyridinium ni-
acyl cations 11 in the presence of aromatic and nucleophilic trogen, but might also be an effect of the pyridinium moiety
compounds such as 12 would cause a manifold of side reac- itself. To check this assumption, we investigated the fluorin-
tions. At least the reverse reaction (11 ϩ 12 Ǟ 4) appears ating potential of 4-benzylpyridinium tetrafluoroborate and
to be more plausible.
HBF4 (2b) towards lauroyl chloride (15) under the same
Another related example is given in the literature. Akaba conditions as previously described for the Methods A and
[27]
et al.
irradiated aryl-alkyl ketones such as B (Scheme 4).
(
C H ) CHCOC H in the presence of 2,4,6-triphenylpyryl-
6
5 2
6
5
Ϫ
ium salts and molecular oxygen. Both the BF - and the
PF 6 -containing sensitizers formed significant yields of
4
Ϫ
benzoyl fluoride (5e). If pyrylium perchlorate was added,
benzoic acid was detected as an oxygenated product since
no fluorine was available. It was pointed out that the gener-
ation of benzoyl fluoride “could be very complex“ and the
intermediate formation of a benzoyl cation that degrades
Scheme 4. Fluorination of an acyl chloride 15 with 4-benzylpyridi-
nium tetrafluoroborate
Ϫ
the BF 4 moiety is only one of many conceivable alterna-
Surprisingly, an equimolar mixture of 4-benzylpyridin-
tives.
ium tetrafluoroborate and 15 in CH Cl , which was heated
2
2
The formation of 6 was proven in the course of the de- for 24 h, afforded fluoride 5c in 10% yield. The heating of
composition of 4c in CDCl . After heating this solution to the salt above its melting point (130°C) in the presence of
3
5
0°C (0.5 h), together with the signals of 5c, a singlet ap- acyl chloride 15 gave the same result. A mixture containing
1
peared in the H-NMR spectrum at δ ϭ 4.14 which was 12% of fluoride 5c and 88% of chloride 15 was distilled
assigned to the BF adduct 6. Compound 6 was indepen- off (in both cases the yields were determined by H-NMR
1
3
dently identified by comparing its spectra with those of a spectroscopy). For completeness, we investigated the be-
sample directly prepared from 4-benzylpyridine and havior of HBF (2b) against 15. Already at room tempera-
4
Et O BF ( H, 13C, 19F, B NMR).
.
1
11
ture, as shown by an H-NMR experiment, the fluoride 5c
1
2
3
was initially formed Ϫ but disappeared completely after ca.
5
0 min. Thus, from a practical point of view, both fluorin-
ation reactions do not seem to be alternatives to the title
reaction presented in this paper.
At present, we conclude that the decomposition of 4 pro-
ceeds via pathway (b) (Scheme 3) which is initiated by the
formation of a loose complex 13 as found in the X-ray
Ϫ
structure in which a BF 4 is presumably in weak contact
with either C(10) or C(11). After an energetically inexpen-
sive internal rearrangement under inclusion of the carbonyl
C atom, the species 14 (an intermediate or transition struc-
Ϫ
ture) is formed, in which one of the BF 4 fluorines is prop-
erly positioned in order to allow the complex to collapse
into the products 5 and 6. This pathway avoids the dis-
sociation step under formation of 11 and 12. This step has
been modeled by DFT calculations.[ 4-Benzylpyridine
Scheme 3. Proposed reaction mechanism for the formation of acyl-
fluorides 5 from salts 4
28]
(12) serves as an internal proton acceptor in the case of
Based on the assumption that the process might be an acylation reactions with the trifluoromethane sulfonates 3
equilibrium reaction (cf. Scheme 3), equimolar amounts of and alcohols. In the course of the formation of the acyl
acyl fluoride 5c and 6 were mixed together in CDCl solu- fluorides 5b؊g, the pyridine derivative 12 remembers its
3
tion. The progress of the reaction was then monitored by Lewis base properties. This should be important for the suc-
Eur. J. Org. Chem. 1999, 2383Ϫ2390
2385

R. Wagner, B. Wiedel, W. Günther, H. Görls, E. Anders
FULL PAPER
cessful formation of aromatic acyl fluorides. Without the
synchronous formation of the equimolar amount of the
Lewis base, the free BF will catalyze the acylation of the
3
25]
[
aromatic ring moieties.
Crystal Structure and Ab initio Calculations
In spite of their great significance in preparative organic
synthesis, only a few crystal structures of 1-acylpyridinium
salts are available (Scheme 5). The first structure of such a
compound was that of 1-acetyl-4-dimethylaminopyridinium
[
29]
dimesylamide (16a).
This is not surprising since 1-acyl
DMAP derivatives are important acylating reagents.[ We
regard C4-heteroatom-substituted salts such as 16[ as
stabilized” representatives of cations in which the CϪN
30]
31]
ϩ
“
bond is stronger than in the salts 3 and 4 presented in this
work. In contrast to the 4-N(CH ) group, the 4-benzyl sub-
3
2
stituent of 3 and 4 has almost no influence on the stability
Figure 2. Crystal structure of 4f, one of the title compounds. Two
ϩ
of the remote CϪN bond. Another category of stabilized
Ϫ
BF
4
anions are interacting with the 1-acylpyridinium cation via
[32]
˚
pyridinium salts are the 1-acyloxy derivatives 17.
To the the C10/F1B atoms [distance: 3.01(2) A] and C11/F2A [distance:
˚
˚
3
.17(2) A]. The distance C13/F13 [3.29(2) A] is slightly larger than
the sum of the F and C van der Waals radii (3.17 A)
best of our knowledge, Figure 2 shows the first crystal
structure of an 1-acyl-4-alkylpyridinium salt reported in
the literature.
˚
˚
bonyl C atom C(13) [distance: 3.29(2) A]. Unfortunately,
the estimated standard deviation of this structure is rela-
tively large. Due to the experimental error, these interpret-
ations should be treated with caution.
[33]
The investigation of solid-state structures of type 4f
is
part of our ongoing studies on the selective bond lengthen-
ing caused by cation formation. Since the X-ray refinement
of structure 4f shows relatively large error bars, an X-ray-
based comparison of structural changes during protonation
of the dihydropyridine 1f[ producing the species 4f could
lead to misinterpretations. In order to obtain further insight
into the properties of these interesting compounds, we have
modeled the cation of 4f and its neutral precursor 1f with
34]
[
35][36]
[37]
[38][39]
MO methods (AM1,
PM3,
HF/6-31ϩG*,
and B3LYP/6-31ϩG*[40]).
Table 2 clearly indicates that the B3LYP/6-31ϩG* opti-
mized gas-phase structure of 1f is in very good agreement
with the geometry obtained from the excellent crystallo-
graphic analysis. In addition, all relevant structural proper-
ties of the B3LYP/6-31ϩG* calculated geometry of 4f
correspond well with those obtained from its poor X-ray
analysis (Table 3). Therefore, B3LYP is a suitable method
for describing both the uncharged precursor 1f and its cat-
ionic form 4f. The gas-phase data in combination with the
X-ray results provide a reliable foundation for structure dis-
cussions.
Scheme 5. Available crystal structures of 1-acylpyridinium salts
Protonation of dihydropyridine 1f giving structure 4f
leads to a selective lengthening of the exocyclic C(13)ϪN
single bond [139.9(2) pm to 151.4(14) pm by X-ray analysis,
Figure 1. Crystal structure of 1f, the precursor of 4f
Ϫ
In the solid-state structure two BF 4 anions are found to 141.4 pm to 152.7 pm by B3LYP/6-31ϩG* calculations].
be in contact with the positively charged (cf. Table 4) C(10) In addition, these structural changes parallel the significant
˚
and C(11) carbons [3.01(2) and 3.17(2) A]. Both distances changes in calculated atomic charges (resulting from a
are shorter than or equal to the sum of the van der Waals Natural Population Analysis [NPA],[ Table 4). The at-
41]
˚
[31c]
radii of C and F (3.17 A).
The fluorine F(2A) seems to tractive charge difference between C(13) and N decreases
prefer contact with C(11) over coordination with the car- from 1.15 (1f) to 1.08 (4f). Furthermore, the Wiberg-Bond-
2386
Eur. J. Org. Chem. 1999, 2383Ϫ2390

1-Acyl-4-benzylpyridinium Tetrafluoroborates: Stability, Structural Properties
FULL PAPER
Table 2. Selected bond distances and bond angles of structure 1f, determined by X-ray structure analysis, and derived from semiempirical,
ab initio, and DFT calculations
Atoms[a]
X-ray[b]
AM1[b]
PM3[b]
HF/6-31ϩG*
B3LYP/6-31ϩG*
NϪC(13)
139.9(2)
122.1(2)
140.9(2)
140.0(2)
133.2(2)
133.2(2)
145.6(2)
145.7(2)
136.0(2)
131.1(2)
127.3(2)
112.3(14)
119.5(14)
144.1
124.3
140.2
140.6
135.6
135.6
145.7
145.3
135.2
126.5
124.8
114.4
119.4
144.0
121.9
142.5
142.6
134.6
134.7
145.8
145.2
135.2
128.5
125.6
115.1
118.5
139.1
119.8
140.0
139.9
132.8
133.0
146.8
146.8
134.0
128.1
126.3
112.6
120.4
141.4
122.6
140.4
140.2
134.9
135.0
145.9
145.9
137.0
129.7
126.5
112.9
119.9
C(13)ϪO(1)
NϪC(11)
NϪC(10)
C(11)ϪC(12)
C(9)ϪC(10)
C(8)ϪC(9)
C(8)ϪC(12)
C(7)ϪC(8)
C(1)ϪC(7)ϪC(8)
C(7)ϪC(8)ϪC(9)
C(9)ϪC(8)ϪC(12)
NϪC(13)ϪO(1)
[a]
Numbering cf. X-ray, Figure 1. Ϫ [b] Bond lengths in [pm], angles in [°].
Table 3. Selected bond distances and angles of structure 4f, determined by X-ray structure analysis and derived from semiempirical, ab
initio, and DFT calculations
Atoms[a]
X-ray[b]
AM1[b]
PM3[b]
HF/6-31ϩG*
B3LYP/6-31ϩG*
NϪC(13)
151.4(14)
120.7(13)
134.3(13)
137.6(14)
138.(2)
148.4
122.7
137.1
137.3
139.4
139.5
140.6
140.9
148.6
114.1
122.2
117.9
115.3
153.4
120.1
137.8
137.8
138.6
138.8
140.1
140.3
149.4
109.9
120.7
118.6
113.5
148.7
117.9
134.7
134.3
136.4
137.0
139.5
140.4
151.6
116.2
123.5
117.5
115.5
152.7
120.6
135.8
135.6
138.0
138.2
140.5
140.9
151.7
116.2
122.9
117.2
114.9
C(13)ϪO(1)
NϪC(11)
NϪC(10)
C(11)ϪC(12)
C(9)ϪC(10)
C(8)ϪC(9)
134.(2)
143.(2)
C(8)ϪC(12)
C(7)ϪC(8)
C(1)ϪC(7)ϪC(8)
C(7)ϪC(8)ϪC(9)
C(9)ϪC(8)ϪC(12)
NϪC(13)ϪO(1)
135.7(14)
145.0(2)
114.0(10)
120.8(12)
118.0(11)
116.0(10)
[a]
Numbering cf. X-ray, Figure 2. Ϫ [b] Bond lengths in [pm], angles in [°].
Indices[42] (Table 5) from the NBO analysis clearly support Conclusion
the existence of this effect (bond-order NϪC(13): 1f 1.03;
4
f 0.80). No other bond (except the exocyclic double/single
The method described here represents a new and versatile
bond change: from 136.0(2) pm to 145.0(2) pm by X-ray preparative route to carboxylic acid fluorides 5. There is no
analysis and from 137.0 pm to 151.0 pm by B3LYP/6- need for high pressure and/or polyethylene equipment and
3
1ϩG* calculations; bond orders: from 1.61 to 1.02) is no hazardous gases are involved. All transformations can
[28]
characterized by such a significant change.
therefore be performed in conventional glass flasks. The
The smaller changes in the bond relations in the central source of fluorine is the inexpensive tetrafluoroboric acid
pyridinium ring [N, C(8), C(9), C(10), C(11), and C(12)] are (2b), similar to the historical Schiemann reaction.[ These
in accord with what one would expect. Furthermore, it is reactions can be performed on normal laboratory scales.
noteworthy that all quantum-mechanical methods predict a The conditions employed are mild (especially those of
bond-shortening of the C(13)ϪO(1) bond by protonation method B), and the yields are satisfactory relative to those
of 1f [1f 122.1(2) pm and 4f 120.7(13) pm by X-ray analysis; of the other fluorination reactions mentioned above and ad-
43]
[
44]
1
f 122.6 pm and 4f 120.6 pm by B3LYP/6-31ϩG* calcu- ditional methods with perfluoro compounds,
cesium
lations; bond orders: 1f 1.68, 4f 1.80], which is in excellent fluoroxysulfate[ and the most used potassium fluoride,
45]
agreement with the X-ray analysis. This shortening can be whose major drawback is its very low solubility in all but a
explained by a slight delocalization of one lone pair of the few protic solvents.
Ϫ
O(1) under the influence of the adjacent acyl C(13) carbon
With the evidence provided in this work that the BF 4
atom. The decrease of negative charge at the equivalent ion formally decomposes to fluoride and BF , a more gen-
3
O(1) reflects this effect (1f Ϫ0.60, 4f Ϫ0,.50)
eral synthetic problem appears to be illuminated. Some au-
Eur. J. Org. Chem. 1999, 2383Ϫ2390
2387

R. Wagner, B. Wiedel, W. Günther, H. Görls, E. Anders
FULL PAPER
Table 4. Selected NPA-charges (B3LYP/6-31ϩG*) of structures 1f and 4f
C(7)
C(8)
C(9)
C(10)
C(11)
C(12)
N
C(13)
O(1)
[
a]
1
f
f
Ϫ0.25
Ϫ0.50
Ϫ0.08
ϩ0.10
Ϫ0.27
Ϫ0.25
Ϫ0.01
ϩ0.09
Ϫ0.02
ϩ0.07
Ϫ0.25
Ϫ0.25
Ϫ0.46
Ϫ0.38
ϩ0.69
ϩ0.70
Ϫ0.60
Ϫ0.50
[a]
4
[a]
Numbering cf. X-ray, Figures 1 and 2.
Table 5. Specific Wiberg-Bond-Indices B3LYP/6-31ϩG*, structures 1f and 4f
C(7)ϪC(8)
C(8)ϪC(9)
C(9)ϪC(10)
C(10)ϪN
NϪC(13)
C(13)ϪO(1)
[
a]
1
f
f
1.61
1.02
1.11
1.37
1.75
1.48
1.04
1.24
1.03
0.80
1.68
1.80
[a]
4
[a]
Numbering cf. X-ray, Figures 1 and 2
ϩ
thors who tried to utilize N-acyl-N tetrafluoroborates de- ∆φ ϭ 1°) at Ϫ90°C. Data were corrected for Lorentz and polari-
47]
zation effects, but not for absorption.[ Structure Solution and Re-
finement: The structures were solved by direct methods
rived both from pyridines and aliphatic tertiary amines
complained about the limited applicability of such com-
48]
(
SHELXS[ ) and refined by full-matrix least squares techniques
pounds but failed to give a plausible explanation for the
2
[49]
against Fo (SHELXL-97 ). The hydrogen atoms for 1f were lo-
cated by difference Fourier synthesis and refined isotropically. The
hydrogen atoms for 4f were included at calculated positions with
fixed thermal parameters. All non-hydrogen atoms were refined an-
isotropically. XP (SIEMENS Analytical X-ray Instruments, Inc.)
[46]
instability of such cations.
Obviously, the reactivity of
fluoride ions as intermediates always has to be taken into
Ϫ
account in the presence of BF 4 and quaternized nitrogens
as they could cause a rapid formation of a variety of fluori-
nated or deprotonated products. Ketenes are typical ex- was used for structure representations.
amples. For efficient acylation reactions we recommend, in
4
-Benzylidene-1-lauroyl-1,4-dihydropyridine (1c): Yellow crystals,
[
16a]
accord with our recent results,
trifluormethanesulfonates 3.
the exclusive use of the
Ϫ1
m.p. 83Ϫ84 °C (acetone).Ϫ IR (KBr pellet): ν˜ ϭ 1678, 1655 cm
1
(
CϭO, CϭC). Ϫ H NMR (CDCl
3
, 400 MHz): δ ϭ 0.86 (t, 3 H,
), 1.70 (m, 2 H, COCH CH ),
), 5.87, 6.44, 6.76, 7.31 (br, 4 H, dihydropyri-
dine-moiety: both rotamers), 5.85 [s, 1 H, CϭCH(α)], 7.12Ϫ7.29
CH
2
3
), 1.19Ϫ1.31 (m, 16 H, (CH
2
)
8
2
2
.49 (t, 2 H, COCH
2
Experimental Section
1
3
(
(
3
1
m, 5 H, PhH). Ϫ C NMR (CDCl
CH ), 22.65, 24.43, 29.17, 29.30, 29.32, 29.43, 29.56, 29.57, 31.86,
3.12 (CH ), 109.87, 110.51, 116.74, 117.50, 122.45, 123.35, 124.42,
25.38 (br, C-2, C-3, C-5, C-6, dihydropyridine-moiety: both rota-
3
, 100.6 MHz): δ ϭ 14.03
General: All reactions were carried out under a positive pressure
of purified N . Standard syringe techniques were used to transfer
solvents and to add liquid reagents. Glassware was flame-dried and
flushed with N before use. Ϫ The dihydropyridines 1a,d؊f and
the 1-acylpyridinium tetrafluoroborates 4a,d؊f are either known in
the literature or derivatives 1c and 4c were synthesized according to
literature procedures. All acyl fluorides 5bϪf prepared are known
compounds, and their physical properties agree with those reported
3
2
2
2
mers), 115.83 [C(α)], 125.76, 127.68, 128.30, 129.33, 137.86, 168.28
CϭO). Ϫ C24 33NO (351.5): calcd. C 82.00, H 9.46, N 3.98; found
(
H
C 82.25, H 9.66, N 3.91
4-Benzyl-1-lauroylpyridinium Tetrafluoroborate (4c): 1H NMR
in the literature (cf. Table 1). HBF
containers (Fluka) and had a content of 54% in Et
and toluene were distilled from sodium benzophenone ketyl;
4
(2b) was taken from original (CDCl , 400 MHz): δ ϭ 0.84 (t, 3 H, CH ), 1.22Ϫ1.27 (m, 14 H,
3
3
2
O. Diethyl ether
(CH ) ), 1.36 (m,
2
H, COCH CH CH ), 1.77 (m, 2 H,
2
7
2
2
2
COCH CH ), 3.30 (t, 2 H, COCH ), 4.30 [s, 2 H, C(α)H ],
2
2
2
2
1
1
CH
Cl
2 2
was purified by column chromatography (alumina). Ϫ B-,
1
7.15Ϫ7.32 (m, 5 H, PhH), 7.87 (d, 2 H, Py H-3, H-5), 9.11 (d, 2
13
13
C-, and H-NMR spectra were determined with Bruker DRX400
H, Py H-2, H-6). Ϫ C NMR (CDCl , 100.6 MHz): δ ϭ 14.02,
3
operating at 128.4, 100.6 and 400 MHz. Chemical shifts (δ) are 22.58, 23.79, 28.36, 29.13, 29.27, 29.33, 29.43, 29.53, 31.81, 33.39;
reported relative to CDCl
ϭ 2.49, δ ϭ 39.7) as an internal standard and are given in
ppm; the standard for B-NMR spectroscopy was BF
NMR spectra were determined with a Bruker AC200 operating at
88.3 MHz and chemical shifts (δ) are reported relative to CCl
3
(δ
H
ϭ 7.24, δ
C
ϭ 77.0) or [D
6
]DMSO
41.98 [C(α)], 127.32 (Ph C-p), 127.76 (Py C-3, C-5), 129.32, 129.38
(Ph C-o,oЈ, C-m,mЈ), 135.16 (Ph C-i), 139.07 (Py C-2, C-6), 168.51
(Py C-4), 169.82 (CϭO).
(
δ
H
C
11
19
3
2
· Et O. F-
Acyl Fluorides 5b؊g; General Procedures
1
3
F
as an external standard. Coupling constants J are given in Hz. Ϫ
IR spectra were recorded with a Nicolet Impact 400 spectrophoto-
meter. Ϫ Microanalysis were obtained with a LECO CHNS-932
element analyzer. Ϫ Melting points were taken with a copper block
apparatus (Linström) and are uncorrected.
Method A: Freshly prepared 1-acylpyridinium tetrafluoroborates
4a؊g were weighed into a 100 mL flask which was connected to a
microdistillation apparatus. The salt was heated by means of an oil
bath to slightly above the melting point and a clear orange solution
was formed. Then the temperature slowly was raised until distil-
lation was complete.
Crystal Structure Analysis of 1f and 4f: Data Collection: The inten-
sity data for the compounds were collected on a Nonius Kap-
Method B: The dihydropyridines 1a؊g were dissolved in 100 mL
paCCD diffractometer, using graphite-monochromated Mo-K
radiation and the φ-scan technique (180 frames, 30 s per frame,
α
2 2 4
of CH Cl and the equimolar amount of HBF (2b) was added.
The mixture was brought to reflux and heated for 3.0 to 4.0 h.
2388
Eur. J. Org. Chem. 1999, 2383Ϫ2390

1-Acyl-4-benzylpyridinium Tetrafluoroborates: Stability, Structural Properties
FULL PAPER
Then CH
2
Cl
2
was evaporated off and the residue was dissolved in
O. The insoluble components were filtered off, the
tallized upon cooling to Ϫ40°C. The solid was filtered off, washed
with 20 mL of dry Et O, and dried in vacuo giving 11.0 g of 6
30 mL of Et
2
2
1
solvent was removed with a rotatory evaporator and the crude (92%), colorless solid. Ϫ H NMR (CDCl
product was distilled.
3
, 400 MHz): δ ϭ 4.15
), 7.09Ϫ7.40 (m, 5 H, benzylic H), 7.51 (d, J ϭ 6.3
Hz, 2 H, Py H-3, H-5), 8.53 (d, J ϭ 6.3 Hz, 2 H, Py H-2,H-6). Ϫ
(s, 2 H, CH
2
From theses residues (both methods), benzylpyridine (12) was reco-
vered in 85Ϫ90% yield after distillation of the acyl fluorides: As
exemplified for the synthesis of 5c,e,f, the residue was dissolved in
1
3
3 2
C NMR (CDCl , 100.6 MHz): δ ϭ 41.11 (s, CH ), 126.19 (s, Py
C-3, C-5), 127.22, 129.14, 129.40, 136.45 (s, benzylic C), 142.85 (s,
1
9
Py C-2, C-6), 163.54 (s, Py C-4). Ϫ F NMR (CDCl
δ ϭ Ϫ152.0 (q, J ϭ 11.3 Hz, N-BF
). Ϫ 11B NMR (CDCl
MHz): δ ϭ 3.57 (q, J ϭ 10.9 Hz, N-BF ).
3
, 188.3 MHz):
2
00 mL of half conc. HCl and then washed with 2 ϫ 100 mL of
Et O . The aqueous solution was made alkaline with conc. NaOH
solution and extracted with 3 ϫ 100 mL of CH Cl . The organic
layer was dried (Na SO ), then the solvent was evaporated and fi-
nally the crude product was purified by distillation.
3
3
, 128.4
2
3
2
2
2
4
Cinnamoyl Fluoride (5b): Method A: from pyridinium salt 4b (11.2 Acknowledgments
g, 24.3 mmol), yield: 3.1 g (85%), m.p. ϭ 31.5°C. Ϫ Method B:
Financial support by the Deutsche Forschungsgemeinschaft, the
Fonds der Chemischen Industrie and by the Thüringer Ministerium
für Wissenschaft, Forschung und Kultur (Erfurt, Germany) is
gratefully acknowledged. We thank the HewlettϪPackard Com-
pany for providing us with CPU time. We thank Dr. M. Kunert
from dihydropyridine 1b (10.2 g, 34.1 mmol) and 4.6 mL of HBF
4
1
(
2b) in CH
2 2
Cl (150 mL), heating: 4.0 h, yield: 2.3 g (45%). Ϫ H
[50]
3
3
NMR (CDCl
3
, 400 MHz) : δ ϭ 6.36 (dd, JHF ϭ 7.4 Hz, JHH
ϭ
1
6.0 Hz, 1 H, CHϭCHCOF,); 7.39Ϫ7.50 (m, 3 H, Ph H-p-,H-
3
m,mЈ); 7.54Ϫ7.62 (m, 2 H, Ph H-o, oЈ); 7.82 (d, JHH ϭ 16.0 Hz, 2
H, CHϭCHCOF,). Ϫ C NMR (CDCl
and A. Jansen of the Max Planck Gesellschaft, Forschungsgruppe
13
3
, 100.6 MHz): δ ϭ 111.71
19
2
CO -Chemie, Jena for the performance of F-NMR measure-
2
(d, JCF ϭ 67.1 Hz, CHϭCHCOF), 128.26 (Ph C-o,oЈ), 128.90 (Ph
ments. We thank Dr. Jennie Weston (Jena) for useful discussions.
4
C-m,mЈ); 131.12 (Ph C-p), 132.84 (d, JCF ϭ 0.8 Hz, C-i), 151.22
3
1
(d, JCF ϭ 6.14 Hz, CHϭCHCOF), 156.94 (d, JCF ϭ 338.2 Hz,
[
1] [1a]
A. G. Pittman, D. L. Sharp, J. Org. Chem. 1966, 31,
COF).
1b]
316Ϫ2318. Ϫ [ 18-crown-6 complex: C. L. Liotta, H. P.
2
Lauroyl Fluoride (5c): Method A: from pyridinium salt 4c (8.5 g,
Harris, J. Am. Chem. Soc. 1974, 96, 2250Ϫ2252.
[2]
G. A. Olah, S. Kuhn, S. Beke, Chem. Ber. 1956, 89, 862Ϫ864.
F. Seel, J. Langer, Chem. Ber. 1958, 91, 2553Ϫ2557.
W. R. Hasek, W. C. Smith,V. A. Engelhardt, J. Am. Chem. Soc.
19.35 mmol), yield: 3.0 g (77%). Ϫ Method B: from dihydropyri-
[3]
dine 1c (10.8 g, 30.7 mmol) and 4.2 mL of HBF (2b) in 100 mL
4
[4]
1
of CH
2
Cl
2
, heating: 4.0 h, yield: 3.4 g (54.8%). Ϫ H NMR (CDCl
3
,
1
960, 82, 543Ϫ551.
[5]
4
1
00 MHz): δ ϭ 0.86 (t, 3 H, CH
3
), 1.15Ϫ1.40 (m, 16 H, (CH
2
)
8
),
G. A. Olah, S. J. Kuhn,. Org. Synth. 1965, 45, 3Ϫ7.
L. N. Markovski, V. E. Pashinnik, Synthesis 1975, 801Ϫ802.
T. Mukaiyama, T. Tanaka, Chem. Lett. 1976, 303Ϫ306, and
references cited therein.
[6]
3
3
2
.65 (m, 2 H, β-CH ), 2.47 (dt, JHH ϭ 7.4 Hz, JHF ϭ 1.1 Hz, 2
[7]
). Ϫ 1 C NMR (CDCl
3
H, α-CH
2
3
, 100.6 MHz): δ ϭ 14.00 (s, CH
3
),
3
22.59, 23.84 (d, JCF ϭ 1.7 Hz, β-CH
2
,), 28.61 (s, δ-CH ), 28.98 (s,
2
[8] [8a]
T. Sakakura, M. Chaisupakitsin, T Hayashi, M Tanaka, J.
2
[8b]
γ-CH
α-CH
2
), 29.22, 29.25, 29.44, 29.48, 31.80, 32.04 (d, JCF ϭ 32.0 Hz,
Organomet. Chem. 1987, 334, 205Ϫ211. Ϫ
T. Okano, N.
,), 163.55 (d, JCF ϭ 360.7 Hz, COF). Ϫ 19F NMR (CDCl
1
Harada, J. Kiji, Bull. Chem. Soc. Jpn. 1992, 65, 1741Ϫ1743.
G. A. Olah, J. T. Welch, Y. D. Vankar, M. Nojima, I. Kerekes,
J. A. Olah, J. Org. Chem. 1979, 44, 3872Ϫ3881.
2
3
,
[
9]
88.3 MHz)[51]: δ ϭ 44.9.
1
[10]
G. A. Olah, M. Nojima, I. Kerekes, Synthesis 1973, 487Ϫ488.
Acetyl Fluoride (5d): Method A: from pyridinium salt 4d (9.0 g,
[11]
1
For further fluorinating reagents for halogen exchange and re-
3
0.0 mmol), yield: 1.2 g (64%). Ϫ H NMR (CDCl
3
, 400 MHz):
, 100.6 MHz): δ ϭ
3
), 160.8 (d, JCF ϭ 354.4 Hz, COF).
[11a]
lated reactions cf.:
ZnF : K. D. Goggin, J. F. Lambert, S.
2
3
13
δ ϭ 2.23 ( JHF ϭ 7.1 Hz). Ϫ C NMR (CDCl
3
[11b]
W. Walinsky, Synlett 1994, 162Ϫ164. Ϫ
AgClO
4 4
/TiF : J.
2
1
1
8.7 (d, JCF ϭ 58.3 Hz, CH
Jünnemann, I. Lundt, J. Thiem, Liebigs Ann. Chem. 1991,
59Ϫ764. Ϫ [
Chim. Acta 1985, 68, 283Ϫ287. Ϫ
Penglis, Adv. Carbohydr. Chem. Biochem. 1981, 38, 195Ϫ285. Ϫ
11c]
BF -activation: H. Kunz, W. Sager, Helv.
7
3
Benzoyl Fluoride (5e): Method A: from pyridinium salt 4e (5.9 g,
[11d]
AgF: Review: A. A. E.
3
ZnBr · MeCN: W. Tyrra, D. Naumann, S. V. Pasenok,
1
1
6.3 mmol), yield: 1.5 g (74%). Ϫ Method B: from dihydropyridine
e (9.2 g, 33.7 mmol) and 4.6 mL of HBF (2b) in CH Cl (150
]DMSO,
00 MHz): δ ϭ 7.61 (m, 2 H, H-m,mЈ), 7.81 (m, 1 H, H-p), 8.00
[11e]
CF
Y. L. Yaguploskii, J. Fluorine Chem. 1995, 70, 181Ϫ186. Ϫ
4
2
2
[11f]
1
mL), heating: 3.5 h; yield: 2.0 g (48%). Ϫ H NMR ([D
4
6
W. Tyrra, D. Naumann, J. Prakt. Chem. 1996, 338, 283Ϫ286;
R. Miethchen, C. Hager, M. Hein, Synthesis 1997, 159Ϫ164.
J. A. Hyatt, P. W. Raynolds, J. Org. Chem. 1984, 49, 384Ϫ385.
FMOC : [(9-Fluorenylmethyl)oxy]carbonyl.
13
[12]
13]
6
(m, 2 H, H-o,oЈ). Ϫ C NMR ([D ]DMSO, 100.6 MHz): δ ϭ 124.9
[
2
4
(d,
J
CF ϭ 60.5 Hz, C-1), 129.0 (d,
J
CF ϭ 1.13 Hz, C-3, C-5),
[14] [14a]
3
1
L. A. Carpino, D. Sadat-Aalaee, H. G. Chao, R. H. De-
1
3
31.4 (d, JCF ϭ 4.1 Hz, C-2, C-6), 135.3 (s, C-4), 157.34 (d, JCF
44.3 Hz, COF). Ϫ 19F NMR (CDCl
ϭ
[14b]
Selms, J. Am. Chem. Soc. 1990, 112, 9651Ϫ9652. Ϫ
J.-N.
H. Wenschuh, M Beyer-
3
, 188.3 MHz): δ ϭ 17.52.
Bertho, A. Loffet, C. Pinel, F. Reuther, G. Sennyey, Tetrahedron
Lett. 1991, 32( 10) , 1303Ϫ1306. Ϫ [
14c]
Anisoyl Fluoride (5f): Method B: from dihydropyridine 1f (9.85 g,
mann, E. Krause, M.Brudel, R. Winter, M. Schümann, L. A.
Carpino, M. Bienert, J. Org. Chem. 1994, 59, 3275Ϫ3280.
S. Shoda, T. Mukaiyama, Chem. Lett. 1982, 861Ϫ862. Ϫ cf:
3
2.5 mmol) and 4.4 mL of HBF
.0 h, yield 4.2 g (84%). Ϫ H NMR (CDCl
4
(2b) in CH
2
Cl
2
(90 mL), heating
1
[15]
4
3
, 400 MHz): δ ϭ 3.86
[11]
3 3
s, 3 H, OCH ,
), 6.96 (m, 2 H), 7.94 (m, 2 H). Ϫ 1 C NMR (CDCl
3
ref.
16] [16a]
(
1
3
[
00.6 MHz)[52]: δ ϭ 56.59 (s, OCH
4
R. Wagner, W. Günther, E. Anders, Synthesis 1998,
[16b]
3
), 115.37 (d, JCF ϭ 1.3 Hz, C-
8
83Ϫ888. Ϫ
E. Anders, W. Will, A. Stankowiak, Chem.
[16c]
2
3
, C-5), 117.12 (d, JCF ϭ 61.8 Hz, C-1), 134.69 (d, JCF ϭ 4.2 Hz,
Ber. 1983, 116, 3192Ϫ3204. Ϫ
E. Anders, Synthesis 1978,
E. Anders, W. Will, Synthesis 1982, 390Ϫ392.
1
[16d]
C-2, C-6), 158.23 (d, JCF ϭ 339.7 Hz, COF); 166.18 (s, C-4). Ϫ
586Ϫ588. Ϫ
[17]
19
The instability of similar compounds and their analogous de-
composition has been described in the literature: cf. C. Lohse,
S. Hollenstein, T. Laube, Angew. Chem. 1991, 103, 1678Ϫ1679;
Angew. Chem. Int. Ed. Engl. 1991, 30, 1656Ϫ1658 and refer-
ences cited therein.
I. B. Gorrell, G. Parkin, Inorg. Chem. 1990, 29, 2452Ϫ2456.
C. H. Winter, X.-X. Zhou, M. J. Heeg, Inorg. Chem. 1992, 31,
1808Ϫ1825.
3
F NMR (CDCl , 188.3 MHz): δ ϭ 15.40.
Preparation of Borontrifluoride Pyridine Complex 6 as a Reference
Compound: A solution of 8.5 g (50.2 mmol) of 4-benzylpyridine
(
(
12) in 100 mL of Et
2
O was cooled with an ice bath. Then 13.6 mL
(48% in Et O) was added and the mix-
[18]
[19]
50.2 mmol) of BF ϫ OEt
3
2
2
ture was stirred for 1 h. A yellowish oil precipitated which crys-
Eur. J. Org. Chem. 1999, 2383Ϫ2390
2389

R. Wagner, B. Wiedel, W. Günther, H. Görls, E. Anders
FULL PAPER
R. Caputo, C. Ferreri, G. Palumbo, E. Wenkert, Tetrahedron
Lett. 1984, 25( 5) , 577Ϫ578.
21]
[20]
21 EZ, U.K., on quoting the depository number CSD-410404
(1f) and CSD-410405 (4f), the names of the authors, and the
journal citation.
35]
[
A. R. Bassinndale, T. Stout, Tetrahedron Lett. 1984, 25( 15) ,
[
1631Ϫ1632.
J. J. P. Stewart, QCPE 455, Vers. 93, 1993, PC-Version: B. Wie-
del, 1996.
36]
[22]
E. A. Lawton, A. Levy, J. Am. Chem. Soc. 1955, 77, 6083.
E. Anders, A. Stankowiak, Synthesis 1984, 1039Ϫ1041.
A. Stankowiak, Dissertation, Universität Erlangen-Nürnberg,
[
[23]
M. J. S. Dewar, E. G. Zoebisch, E. F. Healy, J. J. P. Stewart, J.
Am. Chem. Soc. 1985, 107, 3902Ϫ3909.
37]
[24]
[
J. P. Stewart, J. Comput. Chem. 1989, 10, 209Ϫ220.
M. J. Frisch, G. W. Trucks, H. B. Schlegel, P. M. W. Gill, B. G.
Johnson, M. A. Robb, J. R. Cheeseman, T. A. Keith, G. A.
Petersson, J. A. Montgomery, K. Raghavachari, M. A. Al-La-
ham, V. G. Zakrzewski, J. V. Ortiz, J. B. Foresman, C. Y. Peng,
P. Y. Ayala, M. W. Wong, J. L. Andres, E. S. Replogle, R. Gom-
perts, R. L. Martin, D. J. Fox, J. S. Binkley, D. J. Defrees, J.
Baker, J. P. Stewart, M. Head-Gordon, C. Gonzalez, and J. A.
Pople, Gaussian 94 (Revision E.1), (Ed. Gaussian, Inc.), Pitts-
burgh PA, 1995.
1986.
[38]
[
25]
G. A. Olah, S. J. Kuhn, W. S. Tolgyesi, E. B. Baker, J. Am.
Chem. Soc. 1962, 84, 2733Ϫ2740.
26]
[
F. Seel, Z. Anorg. Allgem. Chem. 1943, 250, 331Ϫ351.
R. Akaba, Y. Niimura, T. Fukushima, Y. Kawai, T. Tajima, T.
Kuragami, A. Negishi, M. Kamata, H. Sakuragi, K. Tokumaru,
J. Am. Chem. Soc. 1992, 114, 4460Ϫ4464.
28]
[27]
[
As expected, the comparison of the heterolytic dissociation ener-
gies reveals an identical trend, 1f: 147.78 kcal/mol, 4f: 39.96
kcal/mol (B3LYP/6-31ϩG*//B3LYP/6-31ϩG* calculations).
P. G. Jones, K. Linoh, A. Blaschette, Z. Naturforsch. 1990,
[39]
T. Clark, J. Chandrasekhar, G. W. Spitznagel and P. v. R. Schle-
yer, J. Comp. Chem. 1983, 4, 294.
[29]
45b, 267Ϫ270.
[40] [40a]
[40b]
A. D. Becke, J. Chem. Phys. 1993, 98, 5648. Ϫ
W. Yang, R. G. Parr, Phys. Rev. B, 1988, 37, 785.
[41] [41a]
C. Lee,
[30]
G. Höfle, W. Steglich, H. Vorbrüggen, Angew. Chem. 1978, 90,
6
02Ϫ615; Angew. Chem. Int. Ed. Engl. 1978, 17, 569.
A. E Reed, R. B. Weinstock, F. Weinhold, J. Chem. Phys.
[31] [31a]
G. L. Bryant, J. A. King, Acta Cryst. 1992, C48,
[41b]
1
985, 83, 735. Ϫ
A. E. Reed, L. A. Curtiss, F. Weinhold,
2
036Ϫ2039. Ϫ [
31b]
S. Hollenstein, C. Lohse, T. Laube, Croat.
Chem. Rev. 1988, 88, 899.
Chem. Acta 1992, 65( 3) , 727Ϫ736.
[42]
[43]
K. B. Wiberg, Tetrahedron 1968, 24, 1083.
G. Balz, G. Schiemann, Ber. Dtsch. Chem. Ges. 1927, 60,
[32]
[17,31a]
[32a]
Cf. refs.
1
and ref.
G. L. Bryant, J. A. King, Acta Cryst.
[32b]
993, C49, 551Ϫ554. Ϫ
P. Singh, D. L. Comins, S. P.
[32c]
1
186Ϫ1190.
Joseph, Acta Cryst. 1994, C50, 25Ϫ27. Ϫ
T. Laube, Chem.
[44] [44a]
[44b]
N. Ishikawa, S. Sasaki, Chem. Lett. 1976, 1407Ϫ1408. Ϫ
Rev. 1998, 98, 1277.
S. Yanagida, Y. Noji, M. Okahara, Tetrahedron Lett. 1977,
[33]
ϩ
Ϫ
[44c]
Crystal Data for 4f: [C20
H
18NO
2
] [BF
4
] , Mr ϭ 391.16
27, 2337Ϫ2340. Ϫ
S. Yanagida, Y. Noji, M. Okahara, Bull.
Ϫ1
gmol , colourless prism, size 0.20 ϫ 0.20 ϫ 0.04 mm, mono-
clinic, space group P2 /n, a ϭ 8.445(2), b ϭ 21.086(5), c ϭ
Chem. Soc. Jpn. 1981, 54, 1151Ϫ1158.
[45]
1
S. Stavber, Z. Planinsek, M. Zupan, J. Org. Chem. 1992, 57,
5334Ϫ5337 and literature cited therein.
[46] [46a]
˚ ˚
1.363(2) A, β ϭ 109.06(1)°, V ϭ 1912.5(7) A , Tϭ Ϫ90°C,
Ϫ3 Ϫ1
3
1
Z ϭ 4, rcalcd. ϭ 1.359 gcm , µ(Mo-K
08, 5060 reflections in h(Ϫ9/0), k(Ϫ23/23), l(Ϫ11/12), meas-
ured in the range 4.26° Յ Θ Յ 23.27°, 2595 independent reflec-
tions, Rint ϭ 0.156, 842 reflections with F > 4σ(F ), 254 param-
eters, Robs ϭ 0.098 wR 2obs ϭ 0.247, GOOF ϭ 1.488 largest
α
) ϭ 1.13 cm , F(000) ϭ
J. A. King, G. L. Bryant, J. Org. Chem. 1992, 57,
[46b]
8
5136Ϫ5139. Ϫ
J. V. Paukstelis, M. Kim, J. Org. Chem. 1974,
11, 1499Ϫ1503. Ϫ [
46c]
J. V. Paukstelis, M. Kim, J. Org. Chem.
o
o
1974, 11, 1503Ϫ1507.
[47]
Z. Otwinowski, W. Minor, Processing of X-ray Diffraction Data
Collected in Oscillation Mode, in: Methods in Enzymology, Vol
276, Macromolecular Crystallography, Part A (Eds.: C. W.
Carter, R. M. Sweet), Academic Press, 1996, pp. 347Ϫ368.
G. M. Sheldrick, Acta Crystallogr. Sect. A 1990, 46, 467Ϫ473.
G. M. Sheldrick, SHELXL-97, University of Göttingen, Ger-
many 1997.
˚
Ϫ3
difference peak and hole: 0.236/Ϫ0.216 e A
.
[34]
H17NO
2
, Mr ϭ 303.35 gmolϪ1, colour-
Crystal Data for 1f: C20
less prism, size 0.35 ϫ 0.20 ϫ 0.10 mm, monoclinic, space
˚
1
group P2 /c, a ϭ 17.0256(9), b ϭ 14.6550(8), c ϭ 6.1546(3) A,
3
[48]
[49]
˚
β ϭ 92.397(3)°, V ϭ 1534.29(14) A , Tϭ Ϫ90°C, Z ϭ 4,
Ϫ3
Ϫ1
r
calcd. ϭ 1.313 gcm , µ(Mo-K
α
) ϭ 0.85 cm , F(000) ϭ 640,
[50]
4
056 reflections in h(Ϫ18/18), k(Ϫ16/16), l(0/6), measured in the
range 3.59° Յ Θ Յ 23.28°, 2198 independent reflections, Rint
.033, 1811 reflections with F
R1obs ϭ 0.0380, wR 2obs ϭ 0.0968, R1all ϭ 0.049, wR all
.105, GOOF ϭ 1.046, largest difference peak and hole: 0.165/
L.-H. Gan, Aust. J. Chem. 1988, 41, 677Ϫ681.
S. Rozen, I. Ben-David, J. Fluorine Chem. 1996, 76, 145Ϫ147.
B. D. Song, W. P. Jencks, J. Am. Chem. Soc. 1989, 111,
8470Ϫ8479.
[51]
ϭ
[52]
0
o o
> 4σ(F ), 277 parameters,
2
ϭ
[53]
0
A. Sekiya, N. Ishikawa, Bull. Chem. Soc. Jpn. 1978, 51( 4) ,
1267Ϫ1268.
˚
Ϫ3
Ϫ0.226 e A
.
Further details of the crystal structure investigations are avail-
able on requests from CCDC, 12 Union Road, Cambridge CB
Received November 6, 1998
[O98500]
2390
Eur. J. Org. Chem. 1999, 2383Ϫ2390