Azatriquinanes: Synthesis, Structure, and Reactivity

Article abstract of DOI:10.1021/jo980788s

Azatriquinane, a tricyclic amine with rigid, hemispherical topology, is synthesized in six steps from pyrrole. This first example of a [2.2.2]cyclazine can be oxidatively dimerized to a novel, highly strained heptacycle or oxidized with chlorine to give an azatriquinacene, a theoretical precursor to diazadodecahedrane via Woodward dimerization.

Full text of DOI:10.1021/jo980788s

6016
J . Org. Chem. 1998, 63, 6016-6020
Aza tr iqu in a n es: Syn th esis, Str u ctu r e, a n d Rea ctivity
Nicholas M. Hext,† J ens Hansen,† Alexander J . Blake,† David E. Hibbs,‡
Michael B. Hursthouse,‡ Oleg V. Shishkin,† and Mark Mascal*,†
Department of Chemistry, University of Nottingham, Nottingham NG7 2RD, U.K., and Department of
Chemistry, University of Wales Cardiff, P.O. Box 912, Cardiff CF1 3TB, U.K.
Received April 28, 1998
Azatriquinane, a tricyclic amine with rigid, hemispherical topology, is synthesized in six steps from
pyrrole. This first example of a [2.2.2]cyclazine can be oxidatively dimerized to a novel, highly
strained heptacycle or oxidized with chlorine to give an azatriquinacene, a theoretical precursor to
diazadodecahedrane via Woodward dimerization.
In tr od u ction
Triquinacene, the tricyclic hydrocarbon C10H10 (1), was
first prepared by Woodward in an early attempt to
synthesize dodecahedrane.1 It was demonstrated that
the photochemical [2 + 2 + 2 + 2 + 2 + 2]π-electrocyclic
process which constitutes dimerization was theoretically
allowed, although the reaction has never actually been
experimentally observed.2 The reason for this may be
that the endo-endo superimposition which leads to the
reactive configuration for dimerization of 1 would rarely
be attained by random collisions of two hemispherical
bodies.3
In the more than 30 years since the synthesis of
triquinacene, it is interesting that no attempt has been
made to prepare aza-analogues of this topologically
interesting molecule, despite a long-standing interest in
the higher [2.2.3]-, [2.3.3]-, and [3.3.3]cyclazines. For
instance, a great deal of work has been published
concerning the electronic properties of the unsaturated
[2.2.3] and [3.3.3] heterocycles,4 while the [2.3.3] and
[3.3.3] ring systems turn up in no less than eight natural
products,5 of which crepidine (2) (ex. orchids),6 and
coccinelline (3) (ex. ladybird defense secretions)7 are
representative examples.
We recently developed an interest in the [2.2.2]cycla-
zine (i.e. 10-azatriquinane) molecule (4) from a number
of perspectives. This rigid, convex tricycle whose amine
function cannot invert could, for example, be desymme-
trized by bis-R-substitution to give a novel type of chiral
acid/base system.8 It could also be incorporated into
cryptand macrobicycles (e.g. 5), where the lone pairs of
the apical nitrogens would be compelled to point into the
cavity, leading to even more effective metal chelation.
Finally, if azatriquinacene (6) could be prepared, the
prospect of Woodward dimerization could be reexam-
ined: The critical difference between 6 and 1 is that the
former would be expected to possess some aqueous
solubility (particularly as the hydrochloride). This would
lead to states of aggregation in solution which could
dramatically increase the probability of electrocyclic
reaction. This phenomenon has already been described
for [4 + 2] reactions, where very substantial acceleration
of cycloaddition rates vs those in organic media are
observed.9 The achievement of 1,16-diazadodecahedrane
by such means would be an important result not only
from a fundamental standpoint but also in the provision
of functional handles on a molecular Platonic solid.
Although cage hydrocarbons and related polycycles have
classically been of interest as major synthetic mile-
stones,3,10 they have gained more current relevance in
the wider context of providing structurally well-defined,
functionally interconnectable particles for nanotechno-
logical purposes.11
* Corresponding author. Tel: +115-951-3541. Fax: +115-951-3564.
E-mail: pczmm@vax.ccc.nottingham.ac.uk.
† University of Nottingham.
‡ University of Wales Cardiff.
(1) Woodward, R. B.; Fukunaga, T.; Kelly, R. C. J . Am. Chem. Soc.
1964, 86, 3162.
(2) Bertz, S. H.; Kourouklis, G. A.; J ayaraman, A.; Lannoye, G.;
Cook, J . M. Can. J . Chem. 1993, 71, 352.
(3) For a general discussion of this strategy, see: Paquette, L. A.
Chem. Rev. 1989, 89, 1051.
(4) Flitsch, W.; Kra¨mer, U. Adv. Heterocycl. Chem. 1978, 22, 321.
(5) Chilocorine, coccinelline, crepidine, exochomine, hippodamine,
myrrhine, precoccinelline, and poranthericine: Dictionary of Natural
Products, Chapman and Hall, version 5.1, 1996.
(6) Elander, M.; Leander, K.; Rosenblom, J .; Ruusa, E. Acta Chem.
Scand. 1973, 27, 1907.
(7) Tursch, B.; Daloze, D.; Dupont, M.; Pasteels, J . M.; Tricot, M.-
C. Experientia 1971, 27, 1380.
(8) For a current review of chiral electrophile transfer to enolates,
see Yanagisawa, A.; Ishihara, K.; Yamamoto, H. Synlett Special Issue
1997, 411 and references therein.
We have reported in brief on the synthesis of aza-
triquinane 4, its dimer 13, and its oxidation to perchlo-
roazatriquinacene 14.12 An updated and full account is
(9) (a) Rideout, D. C.; Breslow, R. J . Am. Chem. Soc. 1980, 102, 7816.
(b) Grieco, P. A.; Yoshida, K.; Garner, P. J . Org. Chem. 1983, 48, 3137.
(c) Breslow, R.; Maitra, U.; Rideout, D. Tetrahedron Lett. 1983, 24,
1901.
(10) Vo¨gtle, F. Fascinating Molecules in Organic Chemistry, Wiley:
Chichester, 1992.
(11) Drexler, K. E. Nanosystems, Molecular Machinery, Manufactur-
ing, and Computation; Wiley: Chichester, 1992.
S0022-3263(98)00788-9 CCC: $15.00 © 1998 American Chemical Society
Published on Web 07/30/1998

Azatriquinanes
J . Org. Chem., Vol. 63, No. 17, 1998 6017
Sch em e 1
F igu r e 1. X-ray crystal structure of 4‚HBF4. The counterion
is omitted for clarity.
-
the involvement of the BF4 counterion in a short
hydrogen bond to the protonated nitrogen, with an N‚‚‚
F distance of 2.817 Å and N-H-F angle of 176°.
The ionization potential of the nitrogen lone pair of 4
is 7.8 eV.13 This low value16 is reflected in the molecule’s
high basicity, estimated to be about 0.5 pKa units beyond
that of quinuclidine17 by acid-base equilibrium measure-
ments.18 We therefore proposed making use of these
electrons in their rigid framework by incorporation into
a macrobicycle (i.e. 5) and for chiral electrophile delivery
(H+, X+), as mentioned earlier. In either case this would
involve substitution at the positions R to the nitrogen
atom. A classical method for effecting this transforma-
tion is N-oxide formation followed by deprotonation and
quenching with an electrophile. Previous work had
shown that the reaction was successful for quinuclidine,19
and accordingly the oxide 12 was prepared from 4 using
hydrogen peroxide in methanol. Treatment of 12 with
tert-butyllithium at -78 °C followed by addition of
iodopentane did result in the formation of a new com-
pound, but without incorporation of the electrophile.
Indeed, the same outcome was observed after deproto-
nation and simple quenching with methanol. Spectro-
scopic data indicated that the product was formally an
oxidative dimer of 4, and the evidence pointed to the
rather remarkable structure 13. Fortunately, single
crystals of this material were available and the assign-
ment could be secured by X-ray diffraction (Figure 2). In
stark contrast to 4, compound 13 is essentially nonbasic,
and models confirm that protonation at nitrogen seriously
violates the van der Waals radii of the adjacent hydro-
gens on the ethylene bridge. This hitherto unknown ring
system shows the evidence of extraordinary ring strain
in the length of the C-C bonds which connect the two
tricycles together (C6-C9A, 1.584 Å) and in the 25° range
in bond angles (from 100.0° for N1-C6-C9A to 125.3°
for C5-C6-C9A). The observation of 13 is presumably
the result of the elimination of LiO- from the molecule
and deprotonation of the resulting iminium salt to give
the corresponding azomethine ylide, followed by 3 + 3
now given of this work, in addition to structural studies
of each of these three molecules. A study of the electronic
properties of 4 has appeared elsewhere.13
Resu lts a n d Discu ssion
The synthetic route taken to azatriquinane is pre-
sented in Scheme 1. The starting material, 2,5-pyrrole-
dipropanoic acid dimethyl ester 7, could be prepared on
a large scale by bis-addition of pyrrole to methyl acry-
late.14 Hydrogenation over rhodium on alumina gave a
mixture of pyrrolidines 8a and 8b in 88% yield. Although
the desired cis isomer predominated, no separation could
be effected at this point. Cyclization to the pyrrolizines
was accomplished by prolonged heating at reflux in
toluene, which gave a separable mixture of 9a (84%) and
9b (11%). Compound 9a was carried on to the next stage,
consisting of dry distillation from an intimate mixture
with sodalime.15 Careful workup allowed the isolation
of the enamine 10 in 55% yield; however, better conver-
sion to tricyclic product was achieved if the condensate
was treated directly with aqueous acid to give the stable
hemiaminal 11. The parent triquinane 4 could then be
derived from 11 in good yield by reduction with lithium
aluminum hydride.
Compound 4 is a volatile white solid with a melting
point just above room temperature. A crystalline tet-
rafluoroborate salt could be prepared, and single crystals
suitable for X-ray diffraction were grown from CH2Cl2/
ether. The structure of 4‚HBF4 is shown in Figure 1.
Although the molecule has averaged C3 symmetry in
solution (by NMR), no 3-fold crystallographic symmetry
is imposed. An interesting feature of this structure is
(16) Takahashi, M.; Watanabe, I.; Ikeda, S. J . Electron. Spectrosc.
Relat. Phenom. 1986, 37, 275.
(12) Mascal, M.; Hext, N. M.; Shishkin, O. V. Tetrahedron Lett. 1996,
37, 131.
(13) Galasso, V.; Hansen, J .; J ones, D.; Mascal, M. J . Mol. Struct.
(THEOCHEM) 1997, 392, 21.
(14) Treibs, A.; Michl, K.-H. Liebigs Ann. Chem. 1954, 589, 163.
(15) Method: Miyano, S.; Somehara, T.; Nakao, M.; Sumoto, K.
Synthesis 1978, 701.
(17) Formally 1-azabicyclo[2.2.2]octane; a consensus of literature
values places the pKa of this compound at about 11.0.
(18) Method: Alder, R. W.; Eastment, P.; Hext, N. M.; Moss, R. E.;
Orpen, A. G.; White, J . M. J . Chem. Soc., Chem. Commun. 1988, 1528.
(19) Barton, D. H. R.; Beugelmans, R.; Young, R. N. Nouv. J . Chim.
1978, 2, 363.

6018 J . Org. Chem., Vol. 63, No. 17, 1998
Hext et al.
F igu r e 3. X-ray crystal structure of 14.
F igu r e 2. X-ray crystal structure of 13. The molecule lies on
a crystallographic inversion center. Hydrogen atoms have been
omitted and the ellipsoids are represented at the 20% prob-
ability level for clarity.
consequent enhancement of its basicity. An unexpected
reaction of the N-oxide 12 links two molecules of 4
through a common six-membered ring to produce a novel
heptacycle 13. Azatriquinane 4 can be oxidized with
chlorine to give perchloroazatriquinacene 14, from which,
however, the parent azatriquinacene 6 could not be
derived. Work continues along varied lines on the goal
of substituting azatriquinane R to the nitrogen in order
to exploit its unique structural properties, and on the
eventual achievement of the triene 6.
Sch em e 2
Exp er im en ta l Section
Dim eth yl cis-2,5-P yr r olid in ed ip r op a n oa te (8a ) a n d
Dim eth yl tr a n s-2,5-P yr r olid in ed ip r op a n oa te (8b). Rho-
dium on alumina (5%, 663 mg) and a solution of pyrrole 714
(29.9 g, 125 mmol) in acetic acid (125 mL) were placed in an
1-L Parr apparatus. After multiple cycles of evacuation and
flushing with hydrogen, the mixture was shaken under 40 psi
of hydrogen for 24 h. The solvent was evaporated and the
residue partitioned between dichloromethane (200 mL) and
saturated aqueous K2CO3 (200 mL). The aqueous phase was
extracted with CH2Cl2 (2 × 200 mL). The combined organic
extracts were washed with saturated aqueous K2CO3 (200 mL)
and dried over MgSO4. Chromotography (4:1 ether/MeOH)
gave a mixture of the pyrrolidines 8a and 8b (26.9 g, 88%) as
a pale yellow oil: 1H NMR (400 MHz, CDCl3) (major isomer
8a only) δ 3.67 (6 H, s), 3.05 (2 H, m), 2.38 (4 H, dt, J 1.5 and
7.6), 1.99 (1 H, br s), 1.86 (2 H, m), 1.76 (4 H, q, J 7.1), 1.35 (2
H, m); 13C NMR (100.6 MHz; CDCl3) (major isomer 8a only) δ
174.1 (C), 58.1 (CH), 51.5 (CH3), 31.7 (CH2), 30.8 (CH2); HRMS
(FAB) found, m/z 244.15309 (M + H), C12H22NO4 requires
244.15488.
dimerization.20 The molecule is not particularly stable
in solution and decomposes in the course of a few days
into an as yet unidentified rearrangement product.
A major objective of this research program was the
attainment of azatriquinacene 6, and efforts turned
toward the dehydrogenation of 4 (Scheme 2). The Cl2
oxidation-partial reduction protocol by which J acobson
was able to convert triquinane into triquinacene21 ap-
peared promising, and we therefore irradiated a solution
of 4 in sulfuryl chloride and in this way obtained the first
example of an azatriquinacene (14) in excellent yield.
This nonachlorotricycle could be crystallized from pen-
tane and the X-ray crystal structure is shown in Figure
3. Unfortunately, dechlorination with lithium and tert-
butyl alcohol as described21 led only to decomposition, and
further attempts across a wide range of reductive condi-
tions have failed to produce any characterizable product.
In conclusion, the first synthesis of the [2.2.2]cyclazine
ring system has been accomplished by sequential anne-
lations onto pyrrole.22 The rigid tricyclic framework
enforces pyramidalization at the apical nitrogen with
cis-Hexa h yd r o-5-[2-(m eth oxyca r bon yl)eth yl]-3H-p yr -
r olizin -3-on e (9a ) a n d tr a n s-Hexa h yd r o-5-[2-(m eth oxy-
ca r bon yl)eth yl]-3H-p yr r olizin -3-on e (9b). A mixture of
pyrrolidines 8a and 8b (33.2 g, 136 mmol) was heated at reflux
in toluene (250 mL) for 10 d under nitrogen. The solvent was
evaporated and the residue chromatographed (30:1 f 10:1
ether/MeOH) to give 9a (24.3 g, 84%) and 9b (3.27 g, 11%) as
colorless oils.
9a : 1H NMR (400 MHz, CDCl3) δ 3.91 (1 H, septet, J 5.2),
3.67 (3 H, s), 3.60 (1 H, br m), 2.74-2.64 (1 H, m), 2.63-2.54
(1 H, m), 2.48 (1 H, dd, J 8.6 and 16.4), 2.44-2.29 (2 H, m),
2.28-2.15 (2 H, m), 1.96-1.88 (2 H, m), 1.86-1.76 (1 H, m)
1.73-1.62 (1 H, m), 1.47-1.38 (1 H, m); 13C NMR (100.6 MHz;
CDCl3) δ 173.4 (C), 171.9, (C), 64.2 (CH), 52.7 (CH), 51.5 (CH3),
37.6 (CH2), 33.5 (CH2), 31.0 (CH2), 29.3 (CH2), 28.2 (CH2), 26.4
(CH2); HRMS (EI) found, m/z 211.1204 (M+), C11H17NO3
requires 211.1208.
(20) Such dimerizations have been observed for acyclic azomethine
ylides: Ouali, M. S.; Vaultier, M.; Carrie´, R. Bull. Soc. Chim. Fr. II
1979, 633.
(21) J acobson, I. T. Acta Chem. Scand. 1967, 21, 2235.
(22) The fortuitous synthesis of a polycycle containing the [2.2.2]-
cyclazine ring system has been reported: Lamm, B.; Andersen, L.;
Aurell, C.-J . Tetrahedron Lett. 1983, 24, 2413.
9b: 1H NMR (400 MHz, CDCl3) δ 3.89-3.82 (1 H, m), 3.79
(1 H, quintet, J 7.1), 3.59 (3 H, s), 2.60 (1 H, dt, J 10.0 and
16.7), 2.37 (2 H, t, J 7.7), 2.32 (1 H, ddd, J 2.7, 9.8, and 16.8),

Azatriquinanes
J . Org. Chem., Vol. 63, No. 17, 1998 6019
2.26-2.18 (2 H, m), 2.01-1.94 (1 H, m), 1.74 (2 H, q, J 7.5),
1.70-1.55 (2 H, m), 1.24 (1 H, quintet, J 10.2); 13C NMR (100.6
MHz; CDCl3) δ 175.6, 173.6, 61.0, 53.5, 51.3, 34.3, 33.4, 32.6,
(400 MHz, C6D6) δ 3.50 (3 H, m), 1.67-1.57 (6 H, m), 1.25-
1.15 (6 H, m); 13C NMR (100.6 MHz; C6D6) δ 65.8, 30.9; HRMS
(EI) found, m/z 137.1255 (M+), C9H15N requires 137.1205.
cis,cis,cis-10-Azatr icyclo[5,2,1,01,10]decan e N-Oxide (12).
Hydrogen peroxide (30%, 0.50 mL) was added to a solution of
amine 4 (0.254 g, 1.85 mmol) in methanol (10 mL) at room
temperature. After standing for 45 min, another portion of
hydrogen peroxide (30%, 0.50 mL) was added. After 38 h a
final aliquot of hydrogen peroxide (30%, 0.50 mL) was added.
Stirring was then initiated and a small amount of platinum
black introduced. When gas evolution had ceased, additional
platinum black was added until no further gas evolution was
observed. The suspension was filtered through Celite, and the
solid residues were washed with methanol. The solvent was
evaporated to give a colorless oil which solidified on drying
(40 °C, 0.1 Torr, 72 h). Oxide 12 is extremely hygroscopic and
therefore no reliable yield determination was possible. It
decomposes gradually on standing and was used directly in
the next step. 1H NMR (400 MHz, CDCl3) δ 4.07 (3 H, m),
2.48-2.39 (6 H, m), 1.85-1.76 (6 H, m); 13C NMR (100.6 MHz;
CDCl3) δ 84.9, 29.3; HRMS (FAB) found, m/z 154.1290 (M +
H), C9H16NO requires 154.1232.
31.1, 30.7, 26.1; HRMS (EI) found, m/z 211.1206 (M+), C11H17
-
NO3 requires 211.1208.
cis-10-Aza tr icyclo[5.2.1.01,10]d ec-1-en e (10). An intimate
mixture of pyrrolizine ester 9a (2.11 g, 10.0 mmol) and
powdered sodalime (7.0 g) was flame pyrolyzed under an argon
atmosphere in a microdistillation apparatus. Heating was
continued until no more liquid could be seen condensing out
of the reaction and the residue had become gray. The distillate
was transferred into a mixture of water (5 mL) and CH2Cl2 (5
mL). The aqueous layer was extracted with CH2Cl2 (2 × 5
mL), and the combined organic phases were dried over MgSO4.
Evaporation of the solvent under vacuum at room temperature
gave the crude product as a brown oil (1.16 g). Short path
distillation (Kugelrohr) at 1 Torr gave the enamine 10 (0.74
g, 55%) as a pale yellow oil: 1H NMR (400 MHz, CDCl3) δ 4.32
(1 H, s), 3.52 (1 H, q, J 7.3), 3.47-3.41 (1 H, m), 2.94-2.87 (1
H, m), 2.45 (1 H, dd, J ) 3.1 and 14.9), 2.39-2.32 (1 H, m),
2.24-2.19 (2 H, m), 1.80-1.71 (1 H, m), 1.67-1.53 (3 H, m),
1.40-1.33 (1 H, m); 13C NMR (100.6 MHz; CDCl3) δ 156.3, 93.2,
64.0, 59.8, 39.4, 34.4, 32.2, 29.5, 23.8; HRMS (EI) found, m/z
135.1054 (M+), C9H13N requires 135.1048.
cis,cis,cis-10-Aza tr icyclo[5,2,1,01,10]d eca n -1-ol Hyd r o-
ch lor id e (11). Compound 9a (1.00 g, 4.73 mmol) was treated
as described above in the synthesis of 10, but the distillate
was transferred with the aid of a little CH2Cl2 into 1.0 M HCl
(30 mL). All volatiles were removed under reduced pressure
to give a brown solid which was chromatographed (CH2Cl2 f
4:1 CH2Cl2/MeOH) to give 11 (0.628 g, 70%) as a beige solid:
mp 160 °C (dec); 1H NMR (400 MHz, CDCl3) δ 4.44 (2 H,
quintet, J ) 5.8), 2.54-2.40 (4 H, m), 2.22-2.13 (2 H, m),
2.12-2.03 (2 H, m), 1.83-1.73 (4 H, m); 13C NMR (100.6 MHz;
CDCl3) δ 109.4, 65.2, 35.8, 29.7, 29.6; HRMS (FAB) found, m/z
154.1257 (M + H), C9H16NO requires 154.1232. Anal. Calcd
for C9H16ClNO: C, 56.99; H, 8.50; N, 7.38; Cl, 18.69. Found:
C, 56.77; H, 8.63; N, 7.37; Cl, 18.76.
19,20-Dia za h ep t a cyclo[7.5.2.22,8.11,9.12,8.05,20.012,19]eico-
sa n e (13). Tert-butyllithium (1.6 M in hexanes, 1.0 mL, 1.6
mmol) was added dropwise to a stirred suspension of 12 (0.125
g, 0.82 mmol) in THF (20 mL) at -78 °C. The resulting pale
yellow, homogeneous solution was stirred for 20 min before
being allowed to slowly warm to 0 °C. Methanol (1 mL) was
then added, and the solvents were evaporated. Chromatog-
raphy (CH2Cl2 f 4:1 CH2Cl2/ether) gave 13 (0.053 g, 48%) as
a white solid: 1H NMR (400 MHz, CDCl3) δ 3.43-3.36 (2 H,
m), 1.95-1.81 (16 H, m), 1.61-1.50 (4 H, m), 1.11-1.08 (4 H,
m); 13C NMR (100.6 MHz; CDCl3) δ 79.6 (C), 58.9 (CH), 39.7
(CH2), 35.2 (CH2), 33.3 (CH2); HRMS (FAB) found, m/z
271.2188 (M + H), C18H27N2 requires 271.2174.
cis,cis,cis-1,2,3,4,5,6,7,8,9-Non a ch lor o-10-a za t r icyclo-
[5.2.1.01,10]d eca -2,5,8-tr ien e (14). A solution of 4‚HBF4
(0.225 g, 1.00 mmol) in freshly distilled sulfuryl chloride (40
mL) was irradiated with a 240 W tungsten lamp for 16 h. The
heat from the lamp maintained the system at gentle reflux
during the irradiation period. The solution was cooled to room
temperature and the excess sulfuryl chloride was evaporated.
Chromatography (petroleum ether) gave 14 (0.415 g, 94%) as
a white solid: 13C NMR (100.6 MHz; CD2Cl2, -60 °C) δ 131.0,
94.0; MS (EI) isotopic cluster for (C9Cl8N)+, m/z 412 (6.4), 410
(25.1), 409 (6.3), 408 (63.3), 407 (10.0), 406 (100), 405 (9.0),
404 (91.4), 403 (3.5), 402 (35.5).
cis,cis,cis-10-Aza tr icyclo[5,2,1,01,10]d eca n e (4). Lithium
aluminum hydride (0.67 g, 18 mmol) was added to a stirred
suspension of 11 (0.843 g, 4.44 mmol) in THF (100 mL). When
gas evolution had ceased, a further portion of lithium alumi-
num hydride (0.90 g, 24 mmol) was added and the suspension
was heated at reflux for 62 h. The mixture was cooled to 0 °C
and water (0.5 mL) was cautiously added, followed by 2 M
aqueous NaOH (10 mL). The slurry was stirred at room
temperature for 30 min and K2CO3 (20 g) added. The mixture
was stirred for a further 1 h at room temperature, before being
filtered through layers of Celite and K2CO3. The solid residues
were washed with CH2Cl2-MeOH (20:1, 20 × 10 mL). To the
combined filtrates was added 0.5 M aqueous HCl (30 mL), and
all volatiles were removed under reduced pressure to give a
pale yellow oil. Chromatography (CH2Cl2 f 4:1 CH2Cl2/
MeOH) gave the hydrochloride salt (4‚HCl) as a hygroscopic
beige solid (0.746 g, 97%): 1H NMR (400 MHz, CDCl3) δ 12.75
(1 H, br s), 4.25 (3 H, br s), 2.19 (6 H, br m), 1.62 (6 H, br m);
13C NMR (100.6 MHz; CDCl3) δ 66.5, 29.7. The compound was
analyzed as its crystalline tetrafluoroborate salt (4‚HBF4),
which is produced in essentially quantitative yield by dissolv-
ing 4‚HCl in a 20-fold molar excess of saturated aqueous
NaBF4 followed by extraction with CH2Cl2: mp 200 °C (dec);
1H NMR (400 MHz, CDCl3) δ 8.89 (1 H, br s), 4.33 (3 H, br s),
2.29-2.21 (6 H, m), 1.88-1.80 (6 H, m); 13C NMR (100.6 MHz;
CDCl3) δ 68.3, 29.5; HRMS (FAB) found, m/z 138.1284 (M +
H), C9H16N requires 138.1283. Anal. Calcd for C9H16BF4N:
C, 48.04; H, 7.17; N, 6.22. Found: C, 47.87; H, 7.18; N, 6.15.
To obtain the free base 4, the hydrochloride salt 4‚HCl (121
mg, 0.700 mmol) was introduced into a mixture of ether (5
mL) and 2 M aqueous KOH (0.7 mL). After vigorous agitation
for 2 min the layers were separated, and the aqueous layer
was extracted with ether. The combined organic phase was
dried over MgSO4 and the solvent cautiously evaporated to
Cr ysta llogr a p h ic Stu d ies. Crystal data for 4‚HBF4:
C9H16N+BF4-, M ) 225.04, orthorhombic, space group P212121,
a ) 6.7150(6), b ) 9.3320(9), c ) 16.814(5) Å, U ) 1053.6(3)
Å3 (by least squares refinement of the setting angles for 250
reflections with θ ) 2.4-25.0°), T ) 120(2) K, Z ) 4, Dx
)
1.419 g cm-3, µ(Mo KR) ) 0.132 mm-1. Data were collected
using a FAST TV area detector diffractometer following
previously described methods.23 From the ranges scanned,
3666 data were recorded (2θmax ) 50°) and merged to give 1647
unique reflections (Rint ) 0.044) with intensity > 0. The
structure was solved using automatic direct methods.24 Re-
finement was by full-matrix least squares25 on F2 with all non-
hydrogen atoms anisotropic. The hydrogen atoms were located
from the difference map and refined isotropically. The weight-
ing scheme, w ) 1/[σ2(Fo)2 + (0.0444P)2], where P ) 1/3[(Fo)2 +
2
2Fc ], gave satisfactory agreement analyses. Final R1 [F g 4σ-
(I)] ) 0.0362 and wR2[all F2 data] ) 0.0838, S[F2] ) 0.99 for
all 1647 reflections and 200 parameters. The final ∆F
synthesis showed no peaks above 0.24 e Å-3
.
Crystal data for 14: C9Cl9N, M ) 441.14, orthorhombic,
space group Pbca, a ) 14.814(6), b ) 12.722(2), c ) 15.847(3)
Å, U ) 2986(14) Å3 [from 2θ values of 32 reflections measured
at (ω (14 e θ e 16°, λ ) 0.71073 Å, T ) 298 K)], Z ) 8, Dx )
(23) Darr, J . A.; Drake, S. R.; Hursthouse, M. B.; Malik, K. M. A.
Inorg. Chem. 1993, 32, 5704.
(24) Sheldrick, G. M. Acta Crystallogr., Sect. A, 1990, 46, 467.
(25) Sheldrick, G. M. SHELXL93, University of Go¨ttingen, 1993.
1
give 4 (89 mg, 93%) as a white solid: mp ca. 30 °C; H NMR

6020 J . Org. Chem., Vol. 63, No. 17, 1998
Hext et al.
1.962 g cm-3, µ(Mo KR) ) 1.667 mm-1. Data were collected
using a Stoe Stadi-4 four-circle diffractometer, graphite-
monochromated Mo KR X-radiation, ω/θ scans, 2687 reflections
measured (2θmax ) 50°), 2630 being unique [Rint) 0.005 after
application of absorption corrections (T range 0.0464-0.0525)
based on ψ scans], giving 2068 with F g 4σ(F) and 2623 which
were retained in all calculations. No crystal decay was
observed. The structure was solved using automatic direct
methods.24 Refinement was by full-matrix least squares25 with
all atoms anisotropic. The weighting scheme w-1 ) [σ2(Fo2) +
(0.027P)2 + 3.11P], where P ) 1/3[MAX(Fo2,0) + 2Fc2], gave
satisfactory agreement analyses. Final R1 [F g 4σ(F)] )
0.0333, wR2 [all F2 data] ) 0.0768, S[F2] ) 1.04 for 173 refined
parameters. An extinction correction25 refined to 0.00026(10)
Ack n ow led gm en t. Dr. Alexander Khvat is thanked
for having repeated the preparation of compound 10.
This work was financially supported by the Engineering
and Physical Sciences Research Council and the Euro-
pean Commission.
Su p p or t in g In for m a t ion Ava ila b le: 1H NMR and 13C
NMR spectra for compounds 8a /b, 9a , 9b, 10, 4‚HCl, 4, 12,
and 13, the 13C NMR spectrum for compound 14, and tables
of crystallographic data for compounds 4‚HBF4 and 14 (29
pages). This material is contained in libraries on microfiche,
immediately follows this article in the microfilm version of the
journal, and can be ordered from the ACS; see any current
masthead page for ordering information.
and the final ∆F synthesis showed no peaks above 0.35 e Å-3
For the crystal data for compound 13 consult ref 12.
.
J O980788S