On the non-classical course of Polonowski reactions of N-benzylmorpholine- N-oxide (NBnMO)

Article abstract of DOI:10.1016/j.tet.2005.02.001

The Polonowski reaction of NBnMO (4) afforded tropone (10) and the novel isoindole 11 besides the expected products benzaldehyde and acetmorpholide, in a temperature-dependent ratio. The reaction proceeded via two primary carbenium-iminium ion intermediates, an exo-centered species 5 which underwent a benzylium-tropylium type rearrangement, and a ring-centered species 6, which reacted further to isoindole 11 by intramolecular electrophilic substitution. The experimental findings were in good agreement with DFT computational data.

Full text of DOI:10.1016/j.tet.2005.02.001

Tetrahedron 61 (2005) 3483–3487
On the non-classical course of Polonowski reactions
of N-benzylmorpholine-N-oxide (NBnMO)
*
Thomas Rosenau, Peter Schmid and Paul Kosma
Christian-Doppler-Laboratory, Department of Chemistry, University of Natural Resources and Applied Life Sciences,
A-1190 Vienna, Austria
Received 19 October 2004; revised 31 January 2005; accepted 2 February 2005
Abstract—The Polonowski reaction of NBnMO (4) afforded tropone (10) and the novel isoindole 11 besides the expected products
benzaldehyde and acetmorpholide, in a temperature-dependent ratio. The reaction proceeded via two primary carbenium–iminium ion
intermediates, an exo-centered species 5 which underwent a benzylium–tropylium type rearrangement, and a ring-centered species 6, which
reacted further to isoindole 11 by intramolecular electrophilic substitution. The experimental findings were in good agreement with DFT
computational data.
q 2005 Elsevier Ltd. All rights reserved.
1. Introduction
available from previous NMR studies on conformational
equilibria of tertiary amine N-oxides,6 a different behavior
was anticipated: first, the corresponding exo-cation 5 was
supposed to be largely favored from the beginning due to
additional charge stabilization at the benzylic position.
Second, the rearrangement of ring-centered 6 into 5 would
be prevented as the required transition state was strongly
disfavored due to steric hindrance imposed by the bulky
benzyl group. Even though these expectations proved to be
partly true, Polonowski reactions of NBnMO revealed some
surprising results, which we wish to report herein, along
with related kinetic and computational studies.
Tertiary amine-N-oxide are frequently applied oxidants in
organic synthesis,1 mainly used in combination with
catalytic amounts of transition metal catalysts.2 N-Methyl-
morpholine-N-oxide (NMMO, 1) is moreover used in bulk
quantities as a cellulose solvent in the Lyocell process,
which is a new and environmentally benign industrial
approach to production of man-made cellulosic fibers.3
NMMO is able to dissolve cellulose directly without
chemical derivatization to give a dope which is spun simply
into air and water.
In previous studies, we have addressed the chemistry of the
Lyocell process4 and have shown that the NMMO-derived
carbenium–iminium ions (2, 3) play a key role in NMMO
and Lyocell chemistry. Later we reported on the first
example of a carbenium–iminium ion interconversion,
which was observed between these two Mannich inter-
mediates. In theory, Polonowski reactions of NMMO would
produce the ring-centered, thermodynamically favored 3 as
the main product besides small amounts of the exo-centered
2. However, intermediate 3 rearranges into the exo-centered
2 via a highly organized transition state involving one
molecule of water, the back reaction not being observed,5
so that the latter intermediate usually predominates. For
N-benzylmorpholine-N-oxide (NBnMO, 4), which was
2. Results and discussion
Polonowski reactions7 are intramolecular redox reactions in
tertiary aliphatic N-oxides. The nitrogen is reduced from the
formal oxidation stage K1 to K3, and an a-carbon is
oxidized from K2 to G0. The conversion represents a
heterolytic ‘self-oxidation’ which thus does not require an
external oxidant. Polonowski reactions are strictly hetero-
lytic processes.8 They are induced by O-acylation of the
Keywords: Amine N-oxides; Polonowski reactions; Carbenium–iminium
ions; Rearrangement.
* Corresponding author. Tel.: C43 1 36006 6071; fax: C43 1 36006 6059;
e-mail: thomas.rosenau@boku.ac.at
0040–4020/$ - see front matter q 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tet.2005.02.001

3484
T. Rosenau et al. / Tetrahedron 61 (2005) 3483–3487
amine N-oxide by organic or inorganic acid anhydrides or
halides, acetic anhydride being the traditional reagent used.
The acylation of the exogenous oxygen lowers the electron
density along the N–O bond, and facilitates proton
abstraction in a-position to the nitrogen. The classical
Polonowski reaction proceeds further with the loss of the
respective acid anion, for example, acetate, and simul-
taneous deprotonation from the a-position in a trans-
elimination process (Scheme 1). In this step the N–O bond is
cleaved, and an iminium ion is generated. Addition of
excess acetate in a-position with subsequent N–C bond
fission produces the secondary amines in the form of their
corresponding acetamides. In the overall process, the amine
oxide is dealkylated and acylated, and the a-position of the
cleaved alkyl substituent is oxidized to an aldehyde
function.9
Scheme 1. Classical pathway of the Polonowski reaction.
According to the pathway in Scheme 1, we expected
benzaldehyde (7) and acetmorpholide (N-acetylmorpholine,
8) to be the main products of the Polonowski reaction
between N-benzylmorpholine-N-oxide (4) and acetic anhy-
dride. Indeed, these two compounds were found in
equimolar amounts in an approx. 30% yield,10 but there
were three additional main products, however. One of them
was morpholine (9) in its non-acetylated form in about 50%
yield. Surprisingly, it was accompanied by tropone (10) in
the same amount. In contrast to tropone with its quite simple
NMR and mass spectra, identification of the third
unexpected product was rather intricate. Finally, the
compound was identified as oxazino-isoindole derivative
11, obtained in 20% yield. Starting from these experimental
results we set out to clarify the mechanism of this
conversion.
Scheme 2. Non-classical course of the Polonowski reaction of N-benzyl-
morpholine-N-oxide (4), percentages refer to approx. yields at 0 8C [50 8C].
As the first step in the reaction sequence N-benzylmorpho-
line-N-oxide is O-acylated by acetic anhydride. Fragmenta-
tion of the primary acylation product forms two expected
carbenium iminium intermediates, the exo-centered (ben-
zyl) cation 5 and the ring-centered 6 (Scheme 2), albeit the
latter in surprisingly large amounts. At 0 8C reaction
temperature, already about 20% of 6 was produced, and
its formation is more and more favored at increasing
temperatures. The dependence of the product distribution11
from the reaction temperature allowed estimating the
difference between the free activation energies for for-
mation of either 5 or 6, which was 12.8G2.3 kJ molK1
,
A first valuable clue as to the reaction mechanism was the
observed equimolarity between formed benzaldehyde and
acetmorpholide on the one hand as well as tropone and
morpholine on the other hand, indicating that the first
compound couple was formed according to one pathway,
and the latter according to a competitive one. Raising the
reaction temperature to rt, 35 and 50 8C had no effect on the
ratio between the two compound couples, whereas forma-
tion of isoindole 11, was increasingly favored. This
indicated that the two compound couples 7/8 and 9/10
originate from one intermediate and 11 from another one
which is favored at higher temperatures. From these
observations, the overall mechanism as shown in Scheme
2 was developed.
meaning that the activation energy for the formation of
5 was by about 13 kJ molK1 lower than for the formation
of 6.12
A possible explanation for these kinetic data was found by
means of computational chemistry. In the minimum
geometries of both 5 and 6 the positive carbons are sp2-
hybrids with trigonal planar geometry. The morpholinium
ring in 6 adopts a twisted chair geometry, as the positively
charged carbon has trigonal planar geometry, which anchors
the carbenium and the neighboring nitrogen in a rigid
structure while the rest of the ring remains flexible
(Scheme 3). This flexibility—mainly a flipping of the

T. Rosenau et al. / Tetrahedron 61 (2005) 3483–3487
3485
transition states leading to 5 and 6 predict an activation
energy difference of 14.5 kJ molK1 between the inter-
mediates 5 and 6, which is in satisfying agreement with
the kinetically determined value. The thermodynamic
stability of the two carbenium–iminium ions also was
assessed by means of DFT computations, which showed that
5 is by 19.2 kJ molK1 more stable than 6, which is roughly
the sixfold of the difference between the two NMMO-
derived carbenium–iminium ions 2 and 3 (2.9 kJ molK1).5
Once intermediate 6 was formed, it immediately underwent
intramolecular electrophilic substitution to afford isoindole
11. Apparently, the intramolecular path was much favored
over competitive intermolecular ones, as even in the
presence of excess methanol, morpholine or trimethyl-
hydroquinone as competing O-, N-, or C-nucleophiles,
respectively, only 11 but no trapping products were found.
This must be due to a pre-organizational effect, which holds
the attacking positive carbenium ion in optimal position for
electrophilic attack on the aromatic.
Scheme 3. Geometries of the carbenium–iminium ions 5 and 6. Grey
shaded areas indicate in-plane atoms, solid arrows denote conformational
flexibility (flipping or bond rotation), dotted arrows sterical hindrance.
Hypothetical geometries: 5-A: phenyl ring in carbenium–iminium plane,
maximum benzylic resonance stabilization, but H–H repulsion; 5-B: phenyl
ring perpendicular to carbenium–iminium plane, minimum benzylic
resonance stabilization.
Two competitive pathways start from intermediate 5: the
first one is the classical Polonowski pathway leading to 7
and 8, the second one constitutes a benzyl-tropylium
conversion, a rearrangement which is well known to occur
upon fragmentation of substituted aromatics in mass
spectrometry. The temperature independence of the ratio
between 7 (or 8) and 9 (or 10) indicated that the activation
energies for the addition of acetate and rearrangement into
the tropylium derivative were nearly equal. To support this
mechanistic view, we changed the acylating agent to
trifluoroacetic anhydride or sulfuryl chloride, since tri-
fluoroacetate or sulfate as the corresponding weakly
nucleophilic anions should disfavor the Polonowski path-
way thus promoting the competitive tropylium rearrange-
ment. The effect was even stronger than expected: formation
of 7 and 8 was completely prevented, and only 9 and 10 was
found besides isoindole 11. It should be noted that the
change of the acylating agent affected only the two
pathways extending from carbenium—iminium inter-
mediate 5, but not the pathways leading to it, that is, the
ratio between 5 and 6 is not influenced. This is in full
agreement with theory: less basic anions, such as trifluoro-
acetate or sulfate in comparison to acetate, will less
efficiently abstract a-protons from acylated 4. Since this
step is required for both 5 and 6 to form, both pathways are
equally affected although different protons are abstracted.
Thus, the overall reaction rate is lowered, but the ratio
between 5 and 6 remains constant. In contrast, only one
pathway extending from 5—the one leading to 7 and 8—
involves the respective anions, and only that one is
influenced by changes of the acylating agent.
conformationally free semi-chair—is increased at increas-
ing temperatures, whereas the rigidity of the carbenium–
iminium structure is not influenced. This structural element
is the basis of the carbenium–iminium resonance stabiliz-
ation, which is however only little affected by temperature
changes.
The case of intermediate 5, which actually is a benzyl
cation—carbenium–iminium ion hybrid, is different. Apart
from the stabilizing carbenium–iminium resonance effect,
there is strong additional stabilization by charge delocaliza-
tion into the aromatic ring. However, for this resonance
stabilization to become effective, the aromatic ring, the
benzylic CH and the C–N–C element of the morpholine ring
must lie in one plane. Optimum charge delocalization into
the phenyl moiety is only possible in periplanar benzyl
cations (cf. 5-A in Scheme 3), whereas the positive charge
remains localized at the benzyl position in perpendicular
benzyl cations (cf. 5-B in Scheme 3). At equilibrium
geometry of 5, the dihedral angle Ph-2–Ph-1–Cbenzyl–N is
278 and thus significantly different from 08, which would
guarantee full resonance stabilization. The steric repulsion
of the ortho-hydrogen in the phenyl ring and the a-hydrogen
in the morpholine ring prevents the latter geometry, the H–H
˚
distance still being relatively short with 2.08 A. Thus, the
carbenium–iminium ion in 5 does not experience the full
additional benzylic resonance stabilization. Moreover,
overcoming the rotational barrier of the benzyl cation (the
movement from periplanar into perpendicular conformation
and further to another periplanar one) becomes more and
more easy at higher temperatures. Hence, the net stabili-
zation of the carbenium–iminium ion 5 is decreased since the
phenyl moiety is increasingly adopting conformations other
than the fully periplanar one, which explains why 5 becomes
less stable at higher temperatures. Assuming also that the
thermodynamic stabilities of the products go parallel with
the activation energies according to the Hammond prin-
ciple, it becomes clear why the ratio between 5 and 6 is
shifted in favor of the latter with increasing temperature,
as was experimentally observed. Computations on the
The driving force for the rearrangement of 5 into the
(1-morpholinyl)tropylium cation (R-1 in Scheme 2) appears
to be the high resonance stabilization of the latter. NRT13
analysis suggests that there is a 54% participation of the
tropylium resonance form and a 46% participation of
the enamino-cyclohexatrienone canonical structure, which
is nearly the ‘ideal’ 50/50 partition. The mechanism can be
assumed to be that of the well-known benzyl–tropylium
rearrangement,14 from which the present case differs only in

3486
T. Rosenau et al. / Tetrahedron 61 (2005) 3483–3487
the presence of a morpholine ring as the benzylic substituent
which additionally stabilizes the primary intermediate R-1.
(10 mmol) was added dropwise to a solution of morpholine
(20 mmol) in chloroform (100 ml). The mixture was stirred
for 30 min at room temperature and refluxed for 2 h. After
cooling to 0 8C in ice water and addition of n-hexane
(50 ml), the white, crystalline precipitate formed (morpho-
linium chloride) was removed by filtration, and the solvents
were evaporated under reduced pressure. The residue, crude
N-benzylmorpholine, was dissolved in ethanol (50 ml) and
hydrogen peroxide (30% in H2O, 5 ml) was added. The
mixture was stirred overnight, excess H2O2 was destroyed
by slowly adding the reaction mixture to a suspension of
MnO2 (10 mg) in ethanol (5 ml). Removal of MnO2 by
filtration and evaporation of the solvent under reduced
pressure provided 7, which was recrystallized from acetone
(24% overall yield, 0.47 g, mp 136 8C). The synthesis was
repeated and up-scaled to obtain sufficient material.
1H NMR (CDCl3, 0.1 M): d 2.94 (dd, 2H, N–CHeq), 3.53
(dt, 2H, N–CHax), 3.78 (dd, 2H, O–CHeq), 4.16 (dt, 2H,
O–CHax), 4.43 (s, 2H, N–CH2–Ph), 7.44 (m, 3H, Ph), 7.58
(m, 2H, Ph). 13C NMR (CDCl3, 0.1 M): d 62.40 (O–CH2),
64.40 (N–CH2), 75.96 (N–CH2–Ph), 129.46 (Ph, d.i., C-3,
C-5), 130.26 (Ph, C-1), 130.80 (Ph, C-4), 134.20 (Ph, d.i.,
C-2, C-6). Anal. calcd for C11H15NO2$H2O (211.27): C
62.54, H 8.11, N 6.63; found C 62.16, H 8.32, N 6.48.
NMMO (1) readily undergoes O-alkylation by the NMMO-
derived carbenium–iminium ions 2 and 3 leading to an
autocatalytic degradation cycle.15 Interestingly, the carbe-
nium–iminium ions derived from NBnMO do not enter
a similar path. Evidently, the reaction to the observed
products 7–11 is energetically favored over O-alkylation of
NBnMO by either of the carbenium–iminium ions 5 or 6.
The implications of this observation for the stabilization of
NMMO solutions will be discussed elsewhere.
3. Conclusions
The Polonowski reaction of N-benzylmorpholine-N-oxide
(4) proved to be rather complex. Instead of the expected
high yields, only approx. 30% of the two ‘traditional’
products 7 and 8 were obtained at 0 8C, the majority of the
starting material being converted into the novel isoindole 11
and morpholine (9)/tropone (10). The reaction path and the
observed product distribution can be explained by the
intermediacy of two competing carbenium–iminium ions 5
and 6. The differing tolerance of these intermediates
towards temperature changes can be utilized to tune the
product distribution. The experimental findings agree very
well with DFT computational data.
4.2. General experimental procedure for Polonowski
reactions of NBnMO
A solution of NBnMO (4, 10 mmol) in CH2Cl2 (50 ml) was
added dropwise to a solution of acetyl chloride or acetic
anhydride (10 mmol) in the same solvent (150 ml) under
stirring and efficient cooling with an ice/NaCl bath. Also
inorganic acid chlorides, such as POCl3 or SOCl2, were used
with the same result. (CAUTION! Degradation reactions
of amine N-oxides are known to easily become uncontrol-
lable!19 Work in an efficient hood and wear appropriate eye
protection!). The organic phase was washed thoroughly
with a concentrated aqueous sodium hydrogencarbonate
solution until evolution of CO2 ceased, and was dried over
NaSO4. Evaporation of the solvent in vacuo yielded a
yellow syrup, which was chromatographed on silica gel
using n-hexane/chloroform (v/vZ5:1) to elute the products
in the order tropone (10), benzaldehyde (7), the novel
isoindole 11, acetmorpholide (8), and morpholine (9). If
inorganic acid chlorides were used instead of acetyl chloride
or acetic anhydride, no acetmorpholide was found.
4. Experimental
4.1. General
All chemicals were commercially available. Thin layer
chromatography (TLC) was performed on silica gel 60
plates (5!10 cm, 0.25 mm) with fluorescence detection
under UV light at 254 nm. Column chromatography was
performed on silica gel G60 (40–63 mm). Melting points,
determined on a Kofler-type micro hot stage with Reichert-
Biovar microscope, are uncorrected. 1H NMR spectra were
recorded at 300.13 MHz, 13C NMR spectra at 75.47 MHz in
CDCl3 as the solvent if not stated otherwise and TMS as the
internal standard. Data are given in ppm. 13C peaks were
assigned by means of APT, HMQC and HMBC spectra;
“d.i.” denotes peaks with double intensity.
1
4.2.1. Acetmorpholide (8). H NMR (DMSO-d6, 110 8C):
d 2.04 (s, 3H, CH3), 3.47 (m, 4H, N–CH2), 3.68 (t, 4H,
O–CH2). 13C NMR: d 19.7; 47.4; 67.3; 169.4.
Computations, as implemented through Spartan Pro 02 by
Wavefunction, Inc., Irvine, CA, USA, were carried out on
geometries pre-optimized by the semi-empirical PM3
method. For full geometry optimization the widely
employed B3LYP hybrid method, which includes a mixture
of HF and DFT exchange terms and the gradient-corrected
correlation functional of Lee, Yang and Parr16 parametrized
by Becke,17 was used, along with the double-zeta split
valence basis sets 6-31CG*,18 which includes diffuse
functions. Transition states and minima were confirmed
by analysis of the calculated vibrational spectrum, and by
intrinsic reaction coordinate analysis. For all transition
states the number of imaginary frequencies was 1, for all
minimum geometries it was 0.
4.2.2. Morpholine (9). 1H NMR: d 1.73 (s, b, 1H, NH), 2.87
(m, 4H, N–CH2, JZ4.7 Hz), 3.68 (m, 4H, O–CH2, JZ
4.7 Hz). 13C NMR (CDCl3, 0.1 M): d 46.4, 64.1.
1
4.2.3. Tropone (10). H NMR: d 6.97–7.17 (m, 6H, CH).
13C NMR: d 136.7, 136.1, 142.1, 188.1.
4.2.4. 3,4,6,10b-Tetrahydro-1H-[1,4]-oxazino[3,4-a]iso-
1
3
2
indole (11). H NMR: d 2.52 (dd, 1H, JZ4.4 Hz, JZ
11.2 Hz, N–CH2), 2.66 (ddd, 1H, 3JZ4.4, 6.1 Hz, 2JZ
11.2 Hz, N–CH2), 3.55–3.58 (m, 1H, Ar–CH2–N, O–CH2–
CH2), 3.63 (m, 1H, O–CH2–CH), 3.65–3.72 (m, 2H,
4.1.1. N-Benzylmorpholine-N-oxide (4). Benzyl chloride

T. Rosenau et al. / Tetrahedron 61 (2005) 3483–3487
3487
nucleophiles to yield a-substituted tertiary amines. In the
case of water or hydroxyl ions as the nucleophile, these
products represent semi-aminals which yield secondary amine
and carbonyl compound. Also high-yield variants, so-called
Polonowski–Potier reactions are known, which use trifluoro-
acetic anhydride.
Ar–CH2–N, O–CH2–CH2), 3.76 (m, 1H, O–CH2–CH), 4.31
3
3
(t, 1H, JZ6.6 Hz), 6.86 (d, 1H, JZ7.8 Hz), 7.13 (t, 1H,
3
3
3JZ7.8 Hz), 7.29 (t, 1H, JZ7.8 Hz), 7.62 (d, 1H, JZ
7.8 Hz). 13C NMR: d 46.7 (N–CH2–CH2), 56.2 (Ar–CH2–
N), 66.0 (N–CH), 66.2 (O–CH2–CH2), 72.9 (O–CH2–CH),
121.8, 124.1, 125.8, 128.0, 132.4, 140.1. Anal. calcd for
C11H13NO (175.23): C 75.40, H 7.48, N 7.99; found C
75.21, H 7.71, N 8.13. Anal. calcd for C11H14ClNO (211.69,
hydrochloride of 11): C 62.41, H 6.67, Cl 16.75, N 6.62;
found C 62.36, H 6.83, Cl 16.34, N 6.32.
10. As minor byproduct, cinnamic acid was found, evidently
produced by a Perkin reaction of 7 with Ac2O.
11. Estimated by the 1H NMR integrals of the crude reaction
mixture.
12. With R being the ratio 5 and 6 follows: RZexp (K(DE#6K
DE#5)/RT), so that a plot ‘ln R versus 1/T’ gives the slope
KD(DE#)/R, and thus the difference of activation energies.
13. NRT (Natural Resonance Theory) is included in NBO 5.0, see:
Glendening, E. D.; Badenhoop J. K.; Reed, A. J.; Carpenter, J.
E.; Bohmann, J. A.; Morales, C. M.; Weinhold, F. 2001, http://
www.chem.wisc.edu/wnbo5.
14. (a) Rylander, P. N.; Meyerson, S. J. Am. Chem. Soc. 1956, 78,
5799. (b) Rylander, P. N.; Meyerson, S.; Grubb, H. M. J. Am.
Chem. Soc. 1957, 79, 842. (c) Meyerson, S.; Rylander, P. N.
J. Chem. Phys. 1957, 27, 901. (d) Meyerson, S.; Rylander,
P. N.; Eliel, E. L.; McCollum, J. D. J. Am. Chem. Soc. 1959,
81, 2606. (e) Pottie, R. F.; Lossing, F. P. J. Am. Chem. Soc.
1961, 83, 2634. (f) Sevin, F.; Sokmen, I.; Duz, P.; Shevlin,
P. B. Tetrahedron Lett. 2003, 44(16), 3405–3408.
15. Rosenau, T.; Potthast, A.; Kosma, P.; Chen, C. L.; Gratzl, J. S.
J. Org. Chem. 1999, 64, 2166–2167.
Acknowledgements
The authors would like to thank Dr. Andreas Hofinger,
Department of Chemistry at the University of Natural
Resources and Applied Life Sciences Vienna, for recording
the NMR spectra.
References and notes
1. Albini, A. Synthesis 1993, 263–277.
¨
2. For a review see: Schroder, M. Chem. Rev. 1980, 80, 187–213.
3. (a) Chanzy, H. J. Polym. Sci., Polym. Phys. Ed. 1980, 18,
1137–1144. (b) Chanzy, H.; Nawrot, S.; Peguy, A.; Smith, P.
J. Polym. Sci. 1982, 20, 1909–1924.
16. (a) Lee, C.; Yang, W.; Parr, R. G. Phys. Rev. 1988, B37, 785.
(b) Miehlich, B.; Savin, A.; Stoll, H.; Preuss, H. Chem. Phys.
Lett. 1989, 157, 200.
4. For a review see: Rosenau, T.; Potthast, A.; Sixta, H.; Kosma,
P. Progr. Polym. Sci. 2001, 26(9), 1763–1837.
5. Rosenau, T.; Potthast, A.; Kosma, P. Tetrahedron 2003, 60,
301–306.
17. (a) Becke, A. Phys. Rev. 1988, A38, 3098. (b) Becke, A.
J. Chem. Phys. 1993, 98, 5648.
18. (a) Hariharan, P. C.; Pople, J. A. Theor. Chim. Acta 1973, 28,
213. (b) Francl, M. M.; Pietro, W. J.; Hehre, W. J.; Binkley,
J. S.; Gordon, M. S.; DeFrees, D. J.; Pople, J. A. J. Chem. Phys.
1982, 77, 3654.
6. Rosenau, T.; Hofinger, A.; Potthast, A.; Kosma, P. Polymer
2003, 44(20), 6153–6158.
7. Polonowski, M.; Polonowski, M. Bull. Soc. Chim. Fr. 1927,
41, 1190.
19. (a) Brandner, A.; Zengel, H. G. German Patent DE-OS
3,034,685, 1980. (b) Brandner, A.; Zengel, G. H. Chem. Abstr.
1982, 97 7727d CA. (c) Ringel, C. Z. Chem. 1969, 9, 188. (d)
see Ref. 15.
8. (a) Huisgen, R.; Bayerlein, F.; Heydkamp, B. Chem. Ber.
1959, 92, 3223–3241. (b) Bayerlein, F.; Huisgen, R. Chem.
Ber. 1959, 92, 3242.
9. Modifications of the Polonowski reaction use competing