Methacryloyl chloride dimers: From structure elucidation to a manifold of chemical transformations

Article abstract of DOI:10.1016/j.tet.2014.07.019

We report the isolation and elucidation of structure, formation mechanisms, and reactivity of elusive methacryloyl chloride dimers. Upon standing of a sample of methacryloyl chloride (1) for prolonged periods of time dimer 3 forms via oxa-Diels-Alder reaction even at low temperatures. If traces of a Lewis acid are present, dimer 3 isomerizes into a mixture of 2-oxo-cyclopentanecarbonyl chlorides 4, 14, and traces of 5, with a relative ratio that depends on the nature of the Lewis acid and reaction time. Dimer 3 can also be separated and isolated by vacuum distillation and selectively converted into derivatives of 2,5-dimethyl-2-hydroxyadipic acid, as well as into the compound 14 by treating with a catalytic amount of titanium tetrachloride. Additionally, based on the study, suggestions on purification and handling of methacryloyl chloride are proposed.

Full text of DOI:10.1016/j.tet.2014.07.019

Tetrahedron 70 (2014) 6515e6521
Contents lists available at ScienceDirect
Tetrahedron
journal homepage: www.elsevier.com/locate/tet
Methacryloyl chloride dimers: from structure elucidation
to a manifold of chemical transformations
,
a
b
a
Jonas Warneke a , Ziyan Wang , Matthias Zeller , Dieter Leibfritz ,
*
Markus Plaumann a c, Vladimir A. Azov a
,
,
*
a Department of Chemistry, University of Bremen, Leobener Strasse NW 2C, D-28359 Bremen, Germany
b One University Plaza, Department of Chemistry, Youngstown State University, Youngstown, OH 44555-3663, USA
c Department of Biometry and Medical Informatics, Otto-von-Guericke University of Magdeburg, Leipziger Strasse 44, D-39120 Magdeburg, Germany
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 4 April 2014
Received in revised form 30 June 2014
Accepted 5 July 2014
Available online 9 July 2014
We report the isolation and elucidation of structure, formation mechanisms, and reactivity of elusive
methacryloyl chloride dimers. Upon standing of a sample of methacryloyl chloride (1) for prolonged
periods of time dimer 3 forms via oxa-DielseAlder reaction even at low temperatures. If traces of a Lewis
acid are present, dimer 3 isomerizes into a mixture of 2-oxo-cyclopentanecarbonyl chlorides 4, 14, and
traces of 5, with a relative ratio that depends on the nature of the Lewis acid and reaction time. Dimer 3
can also be separated and isolated by vacuum distillation and selectively converted into derivatives of
2,5-dimethyl-2-hydroxyadipic acid, as well as into the compound 14 by treating with a catalytic amount
of titanium tetrachloride. Additionally, based on the study, suggestions on purification and handling of
methacryloyl chloride are proposed.
Keywords:
Methacryloyl chloride
DielseAlder reaction
Lewis acid
Cationic rearrangement
Carboxonium ions
Ó 2014 Elsevier Ltd. All rights reserved.
1. Introduction
adipic acid in polymer chemistry4 as well as possible usefulness of
2 as a molecular building block. However, it was impossible to
In synthetic chemistry, a broad variety of simple and readily
available commercial reagents serve as building blocks for assem-
bly of diverse target structures of different complexity using com-
mon organic reactions. The chemistry of such reagents is usually
very well explored and it is quite unexpected to discover previously
unreported transformations of a simple organic compound that has
been known and intensively employed in synthesis for more than
a century. It is especially interesting, if a usual commercial reagent
affords upon standing a previously overlooked relatively complex
structure with multiple possibilities for further functionalization.
Recently we observed the formation of a white crystalline water
soluble substance after a sample of commercial methacryloyl
chloride1,2 (MAACl) 1 with a reported purity of more than 97% was
inadvertently exposed to air for several days. After detailed mo-
lecular structure analysis we were quite surprised to observe the
formation of previously undescribed 2,5-dimethyl-2-hydroxyadipic
acid3 2 (Fig. 1). This discovery attracted our attention not only due
to its astounding nature, but also because of the importance of
postulate a simple reaction pathway leading directly from MAACl to
the new substance 2. It was clear that in a first step, a dimerization
reaction with a bond formation between terminal alkene carbon
atoms should occur. Thorough literature searches provided very
scarce information about MAACl dimers, which have been de-
scribed sporadically in the literature as impurities in samples of
commercial methacryloyl chloride.5 To our knowledge, only one
publication describes a study on structure elucidation of MAACl
dimers.5a Therein, a mixture of three different dimers 3e5 (Fig. 2)
was proposed by Fisher et al. as a common MAACl impurity.
However, only the formation of dimers 4 and 5 was proved on the
basis of 1H NMR spectra, while the existence of 3 could not been
directly verified and no investigations on the mechanism of their
formation have been reported.6 In contrast to the observations of
* Corresponding authors. Tel.: þ49 421 218 63126; fax: þ49 421 218 63120;
e-mail addresses: j.warneke@uni-bremen.de (J. Warneke), vazov@uni-bremen.de,
vlazov@gmail.com (V.A. Azov).
Fig. 1. Molecular structures of methacryloyl chloride 1 and product 2, which was
isolated after exposure of a sample of commercial methacryloyl chloride to air.
http://dx.doi.org/10.1016/j.tet.2014.07.019
0040-4020/Ó 2014 Elsevier Ltd. All rights reserved.

6516
J. Warneke et al. / Tetrahedron 70 (2014) 6515e6521
vinyl ketone,9 and methacrolein10 known for more than 60 years.11
The reaction is thermally induced and is not accelerated by addition
of a Lewis acid (LA).12 Samples of freshly distilled pure methacryloyl
chloride stored at 40 ꢁC for two months reach an equilibrium state
containing ca. 40 mol % of 3, while samples kept at 4 ꢁC for several
months show only traces of the dimerization product. Heating of
a sample of MAACl or of 3 at 110 ꢁC in a sealed tube for 4 h results
in formation of an equilibrated mixture with the 94:6 molar ratio of
1/3, indicating that the fraction of dimer 3 in the equilibrium
mixture drops with the temperature, as would be expected from
considerations of entropy. Thus, dimer 3 can be generated in an
equilibrium reaction and can be then separated from methacryloyl
chloride by vacuum distillation. Dissociation of 3 into 1 was studied
Fig. 2. Previously reported dimers of methacryloyl chloride.5a The structure of 3 is
shown in brackets as the formation of this compound has not been directly verified.
Fisher et al., we observed the selective formation of one particular
dimerization product 2 with a distinctly different chemical struc-
ture. Therefore, we decided to look into the chemical processes that
lead to the formation of 2 from a commercial methacryloyl chloride
sample upon its contact with air.
ꢁ
in dilute toluene-d8 solution at 110 C and afforded a first order
kinetic dissociation constant13 of 6ꢂ10ꢀ3 sꢀ1 (see Supplementary
data for the details). Thermal decomposition of 1 is not observed
at this temperature.14
2. Results and discussion
The methacryloyl dimer 3 contains a relatively uncommon
structural element, i.e., a CeC double bond with a geminal chlorine
and oxygen substituents at one of the carbon atoms. We decided to
investigate the reactivity of 3 (Scheme 1) in more details. As ex-
pected, hydrolysis of 3 using D2O led to the formation of 2a, con-
We started with the analysis of several samples of commercially
available methacryloyl chloride. Independent of supplier, label,
stabilizer, and storage time (all samples were stored at ꢀ20 ꢁC as
advised), we found that all these samples always contain about
10e15 mol % of another substance (determined by integration of 1H
NMR spectra), although the labels stated the purity at 97%. Samples
could also contain minor quantities of other impurities (Fig. 3), such
as compound 6. Besides, one commercial sample, packed in a bottle
with a crown cap and a septum, contained relatively small amount
of one additional impurity not found in all other samples we have
tested.
taining three DO-groups, as well as one deuterium in the a-position
to one of the carboxyl groups. Upon action of MeOH, the formation
of dimethyl ester 7 was observed.15 The acid chloride group can be
selectively transformed into ester 8 or amide 9 through the reaction
with the corresponding amine or alcohol under basic conditions. In
successive reactions the second ‘masked’ carboxyl group can be
esterified under acidic condition with the formation of 7 or 10,
respectively.16 Thus, selective preparation of the derivatives of 2
with different substituents of two carboxyl groups are readily
achievable, making it possible to use 3, for example, as an in-
termediate in macromolecular chemistry for the production of
polyester/polyamide polymers comprising two different diamide/
diester linkers.
To clarify the reaction process leading to the ring opening of 3
with the formation of 2, we followed up the hydrolysis of 3 by
exposing a sample of pure 3 to air and measuring NMR spectra at
different stages of the process.17 Analysis of 1D and 2D NMR spectra
(Figs. S45, S46)18 gave evidence for the formation of one key in-
termediate, to which structure 11 (Scheme 2) can be tentatively
ascribed.19 During the hydrolysis process, the amount of 11 first
rises up to 50% of the sample composition accompanied by disap-
pearance of 3. After the maximum amount of 11 is reached, it starts
to vanish slowly, whereas the signals of the final product 2 arise in
the NMR spectra. Subsequently the latter starts to crystallize at the
bottom of the flask. All attempts to isolate 11 in pure form were
unsuccessful, and no other intermediates could be detected in any
significant (>5%) amounts before completion of the process. We
conclude that after the first stage of hydrolysis of 3 either carboxylic
acid 12, or more likely, anhydride 13, is formed (Scheme 2), which is
then attacked by HCl and cyclizes into 11. After most of highly re-
active 3 is consumed, hydrolysis of 11 becomes dominant, affording
final product 2 at the end of the process.
Removal of MAACl under vacuum resulted in a slightly tanned
liquid, which was in turn purified by distillation under high vacuum
at ca. 1 mbar affording a colorless liquid with a pungent odor.7
A
detailed molecular structure analysis enabled identification of this
compound as the cyclic dimer 3 (Fig. 2). Compound 3 has been
suggested before as a common impurity of methacryloyl chloride
due to the presence of its esters or amides in esterification or
amidation reactions of MAACl,5aec respectively, but it was never
isolated and characterized in a pure form. Our investigations have
now clearly shown that 3, but not the dimers 4 and 5, constitutes
the main impurity in MAACl. Moreover, dimer 3 is readily available
upon vacuum distillation and may serve as a useful synthetic
intermediate.
The formation of 3 can be formally explained by an oxo-Diels
eAlder reaction between two methacryloyl chloride molecules,
which is similar to the dimerization reactions of acrolein,8 methyl
Thus, we have proved that the previously undescribed adipic
acid derivative 2 was generated via the dominant contamination 3
present in all MAACl samples. However, it was still unclear why the
samples investigated by Fischer et al. contained significant
amounts of dimers 4 and 5 and how they could form from 1.
Therefore, we turned our attention to the sample packed in a bottle
with a crown cap that contained small amounts of an additional
impurity. 1H NMR analysis of the sample after removal of 1 under
vacuum showed the presence of 3 together with another minor
component (Fig. S51). Comparison of the 1H NMR spectrum of the
unknown compound with the spectra of other dimers reported
previously,5a as well as the analysis of 2D NMR spectra identified
Fig. 3. 1H NMR spectrum (360 MHz, 8.4 T, CDCl3) of a commercial sample of meth-
acryloyl chloride. Signals of methacryloyl chloride 1 are shown in red, signals of the
dimer 3 are marked in blue. Compound 6, the product of HCl addition to MAACl, was
also detected in small amounts in some samples, signals are marked green.

J. Warneke et al. / Tetrahedron 70 (2014) 6515e6521
6517
Scheme 1. Reactivity of the MAACl dimer 3.
this substance as dimer 4 (Fig. 2). Because of the thermal lability of
3, which undergoes retro-DielseAlder reaction forming volatile 1,
the dimer 4 could be isolated by boiling off 1 at 120 ꢁC for several
minutes. This way, thermally stable 4 could be isolated before by
Fischer et al., while the dominant dimer 3 decomposed during the
distillation of the mixture.5a
Inspection of the crown cap after its removal has shown that it
was significantly corroded from inside and traces of oxidized metal
in form of chlorides could get in contact with the contents of the
bottle. Since transition metal chlorides in high oxidation states are
known to be strong Lewis acids (LA), we studied the influence of
various LAs on the dimer 3 adding catalytic amounts of AlCl3, TiCl4,
for generation of 4 and 14 was achieved with TiCl4 and Et2O$BF3 as
catalysts, although in the case of the latter the isomerization process
was slower. Moreover, it was observed that upon standing, com-
pound 4 slowly isomerizes into its diastereomer 14, which pre-
sumably is thermodynamically more stable than 4. The fastest
reaction was detected for the sample containing TiCl4, for which
isomerization into 14 was complete within 1e2 weeks (Fig. 4).
Relative configuration of the two stereocenters in 4 had been
determined before by X-ray crystallography of its aromatic amide
15a22 (Scheme 3). To confirm the expected relative configuration of
the two stereocentersin14, analogous amides16a and16b have been
prepared by us from 14. Amide 16a readily afforded crystals suitable
for X-ray structural analysis18 that established the cis-disposition of
two methyl substituents of the cyclopentanone ring (Fig. 5), in con-
trast to the trans-disposition in the previously reported amide 15a,
thus proving the diastereomeric identity of compound 14. The con-
formation of 16a is stabilized by two intramolecular hydrogen bonds
with a S11(6) graph set motif:23 one classical N1eH1/O2 H-bond a,
andonenon-classicalC14eH14/O1H-bond b. Quite surprisingly, no
classical intermolecular hydrogen bonds are present in crystals of
16a. The crystal packing is dominated by the stacking of aromatic
SnCl4, FeCl3, and Et2O$BF3 to ca. 300 mL samples of 3. For all sam-
ples, fast darkening of the reaction mixture was observed, and in
the case of TiCl4 the reaction showed significant exothermicity.
NMR analysis20 has evidenced that all reaction mixtures contained
predominantly 4, as well as its previously undescribed di-
astereomer 14 (ca. 30e50 % initially). The isomerization of 3 was
completed within several days (Scheme 3).
Notably, the reaction rate depended on the activity of the Lewis
acid,21 and the fastest reactionwas observed with TiCl4. In the case of
AlCl3, traces of the third dimer 5 described by Fischer et al. have also
been detected. Unquestionable identificationofcompounds 4, 5, and
14 in the reaction mixture was possible by correlation of the methyl
group hydrogens in HMBC NMR spectra (Fig. S48). All samples
showed formation of minor (<3e5%) byproducts, with the SnCl4-
containing sample being the most impure one. The best selectivity
ꢀ
rings with an interplanar distance of 3.4382(5) A (distance between
ꢀ
ring centroids of 3.6983(9) A), as well as a number of borderline
(distance is slightly below or equal to the sum of VdW radii) contacts
involving Cl-atoms and O-atom of the amido-group with different
aliphatic andaromatic H-atoms, whichcan be classified asweaknon-
classical H-bonds.24
Scheme 2. Proposed pathway for the hydrolysis of 3.

6518
J. Warneke et al. / Tetrahedron 70 (2014) 6515e6521
Scheme 3. Lewis acid (LA) catalyzed transformation of 3 and preparation of aromatic amides 15a, b and 16a, b. Amides 15a, b have been prepared before.5a
A tentative mechanism for the Lewis acid catalyzed trans-
compound 6. Still, in open vials, pure 3 does not isomerize to 4 due
to the fast escape of HCl, but hydrolyzes selectively forming 2.
Finally, the following suggestions on handling of methacryloyl
chloride can be proposed. To obtain pure MAACl, commercial
samples should be distilled and then kept at temperatures prefer-
entially not higher than ꢀ20 ꢁC without any contact with metal
surfaces. Under these conditions, MAACl can be stored without any
detectable formation of dimer 3 for at least several months. To
recycle MAACl from the remaining dimer 3, which can constitute
up to 15% of a batch, the distillation residue should be heated
slightly above the boiling temperature of MAACl until dimer 3 starts
to decompose into 1, which can be collected upon cooling using
a common distillation setup.
formation of 3 into 4, 5, and 14 is suggested in Scheme 4. At first, the
chloroformyl group is activated by a Lewis acid affording a highly
reactive electrophile 17, which intramolecularly attacks the double
bond forming a cyclic carboxonium cation 18.25,26 Breakage of the
CeO bond leads to tertiary carbocation 19, which can either split off
a proton giving 5, or attach Clꢀ with the formation of 4 or 14. As was
noted above, product 4 is likely to be the kinetically preferred one,
whereas upon standing compound 14 is formed predominantly
(most likely via the same carbocation 19), indicating its higher
thermodynamic stability.
In the course of our study, we found that 3 is highly sensitive to
residual moisture in closed vessels. Addition of traces of water leads
to fast decomposition of 3 with the formation of 4, as well as several
other decomposition products of unknown nature. Presumably, the
catalytic activity of HCl, a strong Brønsted acid, generated upon
reaction of water with the acid chloride, is similar to the one of
other LAs tested in the study, although it is affected by the possi-
bility of an attack of HCl on the double bond of 3 (Scheme 2). On the
other hand, samples of 3 containing methacryloyl chloride are
much more stable due to deactivation of HCl by binding to the
double bond of MAACl in Michael fashion with the formation of
3. Conclusion
To summarize, we have found that commercial methacryloyl
chloride, even if marked as a pure reagent, usually contains sig-
nificant amounts of the DielseAlder dimer 3, as well as, occasion-
ally, other high-boiling impurities, and, thus, should be always
distilled before use to avoid the formation of non-desired reaction
byproducts. We were able to isolate and characterize dimer 3 and
have studied its reactivity. Upon action of different reagents it can
be easily transformed into various derivatives of 2,5-dimethyl-2-
hydroxyadipic acid. Isomerization pathways of 3 leading to the
build-up of acid chlorides 4, 5, and 14 have been studied and the
mechanism of their formation upon action of Lewis acids has been
proposed. Future work will focus on the development of conditions
for selective transformation of 3 and 14 into other potentially useful
compounds.
Fig. 4. Isomerization of 3 followed using 1H NMR spectra (CDCl3, 360 MHz) recorded:
(a) before the addition of TiCl4 and (b) 1 day after, (c) 3 days after, (d) 10 days after, and
(e) 17 days after addition of TiCl4. Signals are shown in red for 3, green for 4, and blue
for 14.
Fig. 5. ORTEP plot of 16a confirming the cis-disposition of two Me-groups and
showing two weak intramolecular H-bonds. Thermal ellipsoids are drawn at the 50%
probability level.

J. Warneke et al. / Tetrahedron 70 (2014) 6515e6521
6519
Scheme 4. Mechanism for the Lewis acid (LA) catalyzed transformation from 3 to 5 or 14. Although there are two possibilities for activation of the chloroformyl group by an LA,
either by formation of an oxocarbenium cation or by formation of a donoreacceptor complexes by binding to the oxygen atom;27 only the first option is shown.
4. Experimental section
4.1. General information
d 2.40e2.31 (m, 1H), 1.76e1.69 (m, 2H), 1.63e1.51 (m, 1H), 1.31e1.21
(m, 1H), 1.33 (s, 3H), 1.11 (d, 3J¼7.0 Hz, 3H). 13C NMR (90 MHz,
CD3CN):
(360 MHz, D2O):
d
177.9, 177.7, 74.8, 39.4, 38.0, 28.2, 26.1, 17.1. 1H NMR
d
2.37e2.28 (m, 1H), 1.69e1.56 (m, 1H), 1.62e1.52
Methacryloyl chloride was purchased from Aldrich, Fluka, and
Alfa Aesar. Reagent grade chemicals and solvents were used without
further purification unless stated otherwise. NMR spectra were
recorded using Bruker Avance WB-360 or Bruker Avance DRX600
(m, 1H), 1.56e1.42 (m, 1H), 1.27e1.17 (m, 1H), 1.26 (s, 3H), 0.98
(d, 3J¼7.0 Hz, 3H). MS(ESIꢀ): m/z (%) 189 (100) [MꢀH]ꢀ, 171 (10)
[MꢀH2OꢀH]ꢀ. HRMS (ESIꢀ): m/z [MꢀH]ꢀ calcd for C8H13O5
189.07684, Found 189.07637.
spectrometers. NMR chemical shifts (d) are reported in parts per
million (ppm) relative to TMS, the residual solvent signals were
used as reference CDCl3 (7.26 ppm for 1H, 77.0 ppm for 13C). 1H NMR
coupling constants (J) are reported in Hertz (Hz) and multiplicity is
indicated as follows: b (broad signal), s (singlet), d (doublet), t
(triplet), q (quartet), or m (multiplet). Low resolution mass spectra
were measured on Finnigan MAT 95 spectrometers for electron
ionization (EI, 70 eV) and on a Bruker Esquire-LC spectrometer for
electrospray ionization (ESI). High resolution MS-spectra (HRMS)
were measured on a Finnigan MAT 95 spectrometer (EI, 70 eV), and
on a Thermo Fisher Scientific LTQ Orbitrap spectrometer (ESI).
Melting points were determined using a capillary melting point
apparatus and are uncorrected. Flash chromatography (FC) was
4.2.2. 6-Chloro-3,4-dihydro-2,5-dimethyl-2H-pyran-2-carbonyl
chloride (3). Compound 3 can be easily isolated from commercial
methacryloyl chloride samples. First, methacryloyl chloride 1 is
removed at 5 mbar at normal temperature, then dimer 3 is distilled
at 0.5 mbar and 30e40 ꢁC. Typically it is possible to obtain ca.
3e4.5 g of 3 from 25 g of commercial 1. After keeping a sample of
1 at 40 ꢁC for two months, about 5 g (47 mmol, 50%) of 3 can be
obtained from 10 mL (10.7 g, 102 mmol) of 1 after vacuum distil-
lation. d¼1.25 g/mL. 1H NMR (360 MHz, CDCl3):
d 2.46e2.39
(m, 1H), 2.16e1.97 (m, 2H), 1.93e1.83 (m, 1H), 1.69 (s, 3H), 1.60
(s, 3H). 13C NMR (90 MHz, CDCl3):
d
176.8, 136.1, 105.0, 86.5, 30.þ3,
24.9, 24.3, 18.1. MS (EI): m/z (%) 208 (35) [M]þꢃ, 145 (75) [MeCOCl] ,
69 (100) [C4H5O]þ. HRMS (EI): m/z [M]þ calcd for C8H10Cl2O2
ꢃ
carried out using 230e440 mesh (particle size 36e70
mm) silica gel.
The X-ray structure of 16a was solved by direct methods and refined
208.00579, Found 208.00537.
by full-matrix least-squares analysis using SHELXTL28 and ShelXle.29
4.2.3. trans-3-Chloro-1,3-dimethyl-2-oxocyclopentanecarbonyl chlo-
ride (4). Compound 4 was isolated from ca. 20 mL of commercial
methacryloyl chloride kept in a glass bottle with a metal crown cap.
First, 1 was removed by vacuum distillation, then the residue in
amount of 3 mL was heated to 120 ꢁC for 20 min under the flow of
nitrogen to remove 1 formed by decomposition of 3. Compound 4
remained in the residue in amount of ca. 0.3 mL as a tarry brown
4.2. Synthetic procedures
4.2.1. 2-Hydroxy-2,5-dimethylhexanedioic acid (2). Compound
2
was first obtained by leaving 1 mL (1.07 g, 10 mmol) of commercial
methacryloyl chloride in an open flask for 3e4 weeks resulting in
formation of 0.29 g (1.5 mmol) of 2. Alternatively, 2 can be prepared
by leaving 3 in an open vial for 2e3 weeks. On a larger scale,
compound 2 can be prepared as follows: 0.50 g (2.4 mmol) of 3
were dissolved in MeCN (4 mL), and H2O (1 mL) was added to the
solution. Reaction mixture was stirred over the week-end and then
evaporated to dryness. After drying over P2O5 0.456 g (2.4 mmol,
100%) of the product were obtained as a colorless crystalline
liquid. 1H NMR (600 MHz, CDCl3):
(m, 1H), 2.35e2.30 (m, 1H), 2.25e2.18 (m, 1H), 1.70 (s, 3H), 1.64
d 2.60e2.54 (m, 1H), 2.44e2.39
(s, 3H). 13C NMR (150 MHz, CDCl3):
d
205.7, 172.9, 68.6, 64.2, 38.þ0,
32.1, 24.5, 23.9. MS (EI): m/z (%) 208 (10) [M]þꢃ, 173 (10) [MꢀCl] ,
145 (10) [MꢀCOCl]þ.
powder. NMR analysis showed the presence of
2
as
a
di-
4.2.4. 1,3-Dimethyl-2-oxocyclopent-3-ene-1-carbonyl chloride (5).
Compound 5 could not be separated in a pure form, but could be
characterized by NMR in the reaction mixture formed after addition
astereomeric mixture3 with >95% purity together with several
trace impurities of unknown origin. Recrystallization form MeCN
afforded 0.298 g (1.57 mmol, 65%) of 2 in the diastereomerically
pure form. Mp: 128e129 ꢁC. 1H NMR (360 MHz, CD3CN):
of AlCl3 to dimer 3. 1H NMR (360 MHz, CDCl3):
d
7.42e7.39 (m, 1H),
3.34e3.25 (m, 1H), 2.65e2.55 (m, 1H), 1.86e1.84 (m, 3H), 1.44 (s,

6520
J. Warneke et al. / Tetrahedron 70 (2014) 6515e6521
3H). 13C NMR (90 MHz, CDCl3):
40.0, 20.8, 10.2.
d
203.5, 172.2, 156.2, 132.7, 62.2,
was purified by column chromatography on SiO2 (concentration
gradient: CH2Cl2/CH2Cl2/MeOH, 5%). After solvent evaporation
0.280 g (1.08 mmol, 67%) of 10 were obtained as a colorless liquid.1H
4.2.5. 2-Hydroxy-2,5-dimethylhexanedioic acid dimethyl ester
(7). Compound 7 was prepared by dissolving of 0.50 mL of 3
(0.625 g, 3.0 mmol) in MeOH (10 mL). After that several drops (3e5)
of water were added to the reaction mixture, which was then
stirred for two days. The solvent was removed under vacuum, the
product was purified by filtration through a short SiO2 plug
(CH2Cl2). Evaporation of the solvent afforded 0.593 g (2.7 mmol,
90%) of the product as a colorless liquid. Alternative method:
compound 8 (0.410 g, 2.0 mmol) was dissolved in MeOH (5 mL),
then 3e5 drops of concd HCl were added to the reaction mixture,
which was then stirred for two days. Work up as above yielded
0.406 g (1.86 mmol, 93%) of the product as a colorless liquid. 1H
NMR (360 MHz, CDCl3): d 6.88 (br s, 1H), 3.64 (s, 3H), 3.37 (br s, 1H),
3.26e3.16 (m, 2H), 2.46e2.33 (m, 1H), 1.94e1.83 (m, 1H), 1.79e1.64
(m,1H),1.52e1.45 (m,1H),1.49e1.40 (m, 2H),1.37e1.29 (m, 2H),1.36
(s, 3H), 1.34e1.25 (m, 1H), 1.12 (d, 3J¼7.3 Hz, 3H), 0.90 (t, 3J¼7.3 Hz,
3H). 13C NMR (90 MHz, CDCl3):
d 177.4, 175.3, 75.5, 51.7, 39.3, 38.9,
37.9, 31.6, 27.4, 27.0, 20.0, 17.5, 13.7. MS (ESIþ): m/z (%) 298 (12)
[MþK]þ, 282 (100) [MþNa]þ, 260 (5) [MþH]þ. HRMS (ESIþ): m/z
[MþNa]þ calcd for C13H25NO4Na 282.16758, Found 282.16760.
4.2.9. cis-3-Chloro-1,3-dimethyl-2-oxocyclopentanecarbonyl
chlo-
ride (14). Compound 14 was prepared by adding a drop of TiCl4 to
0.30 mL (0.37 g, 1.78 mmol) of 3. The reaction mixture immediately
turned brown. After stirring for three weeks at rt, the sample
turned deep blue and contained >97% of 14 according to NMR
analysis. The reaction mixture can be used for acylation without
additional purification, or it can be distilled under vacuum (1 mbar)
affording 0.23 g (1.10 mmol, 62%) of 14 as a colorless liquid. 1H NMR
NMR (360 MHz, CDCl3):
d 3.21 (br s, 1H), 2.43e2.33 (m, 1H),
1.84e1.69 (m,1H),1.77e1.65 (m,1H),1.66e1.52 (m,1H), 3.75 (s, 3H),
3.64 (s, 3H),1.36 (s, 3H),1.30e1.18 (m,1H),1.11 (d, 3J¼7.1 Hz, 3H). 13
C
NMR (90 MHz, CDCl3): d 177.4, 176.7, 74.3, 52.7, 51.5, 39.1, 37.3, 27.5,
25.9, 16.8. MS (ESIþ): m/z (%) 241 (100) [MþNa]þ. HRMS (ESIþ): m/z
[MþNa]þ calcd for C10H18O5Na 241.10464, Found 241.10489.
(600 MHz, CDCl3):
d 2.97e2.90 (m, 1H), 2.52e2.44 (m, 1H),
2.09e2.04 (m, 1H), 2.02e1.97 (m, 1H), 1.69 (s, 3H), 1.48 (s, 3H). 13
C
4.2.6. 6-Chloro-3,4-dihydro-2,5-dimethyl-2H-pyran-2-carboxylic
acid methyl ester (8). To a solution of 0.40 mL (0.50 g, 2.4 mmol) of
3 in MeCN (4 mL) were added 0.51 mL (0.367 g, 3.63 mmol) of Et3N,
reaction mixture was cooled to 0 ꢁC, and then MeOH (0.15 mL,
0.120 g, 3.7 mmol) was added dropwise. The reaction mixture was
allowed to warm to rt and stirred overnight, then triethylamine
hydrochloride was filtered off, and the filtrate was evaporated to
dryness. The residue was dissolved in hexane and filtered once
again to remove the remains of triethylamine hydrochloride. After
removal of the solvent, the residue was purified by column chro-
matography on SiO2 (cyclohaxene/CH2Cl2, 1:1) affording 0.391 g
(1.91 mmol, 80%) of 8 as a colorless liquid after solvent evaporation.
NMR (150 MHz, CDCl3):
d
203.7, 173.1, 68.2, 63.2, 36.5,þ31.3, 24.7,
21.9. MS (EI): m/z (%) 208 (25) [M]þꢃ, 173 (15) [MꢀCl] , 145 (20)
[MꢀCOCl]þ, 69 (100) [C4H5O]þ. HRMS (EI): m/z [M]þ calcd for
ꢃ
C8H10Cl2O2 208.00579, Found 208.00670.
4.2.10. cis-3-Chloro-N-(3,5-dichlorophenyl)-1,3-dimethyl-2-
oxocyclopentanecarboxamide (16a). To
a
solution containing
0.145 g (0.90 mmol) of 3,5-dichloroaniline and 0.15 mL (0.109 g,
1.08 mmol) of Et3N in THF (3 mL), 0.18 g (0.86 mmol) of crude 14
(containing catalytic amounts of TiCl4) were added and stirred
overnight at 50 ꢁC. Then the reaction mixture was filtered through
a plug of Celite, and the solvent was removed under vacuum. The
residue was dissolved in CH2Cl2 and purified by column chroma-
tography on SiO2 (CH2Cl2) affording 0.182 g (0.544 mmol, 63%) of
16a as off-white crystals. X-ray quality crystals were grown by slow
evaporation of CH2Cl2/hexane solution. Mp: 126e128 ꢁC. 1H NMR
1H NMR (360 MHz, CDCl3):
d
3.75 (s, 3H), 2.33e2.26 (m, 1H),
d
2.10e1.93 (m, 2H), 1.85e1.74 (m, 1H), 1.67 (s, 3H), 1.53 (s, 3H). 13
C
NMR (90 MHz, CDCl3):
d
173.0, 136.1, 103.3, 81.0, 52.6, 30.5, 25.þ4,
24.7, 18.2. MS (EI): m/z (%) 204 (25) [M]þꢃ, 145 (20) [MꢀCOOMe] ,
69 (100) [C4H5O]þ. HRMS (EI): m/z [M]þ calcd for C9H13ClO3
(360 MHz, CDCl3):
2.88e2.78 (m,1H), 2.46e2.38 (m,1H), 2.08e2.02 (m,1H), 2.02e1.96
(m, 1H), 1.69 (s, 3H), 1.48 (s, 3H). 13C NMR (90 MHz, CDCl3):
212.5,
d 8.56 (br s, 1H), 7.53 (s, 2H), 7.09 (s, 1H),
ꢃ
204.05532, Found 204.05516.
d
4.2.7. 6-Chloro-3,4-dihydro-2,5-dimethyl-N-butyl-2H-pyran-2-
carboxamide (9). Solution of 0.40 mL (0.50 g, 2.4 mmol) of 3 in
MeCN (4 mL) was cooled to 0 ꢁC, and then BuNH2 (0.59 mL, 0.437 g,
6.0 mmol) was added dropwise. The reaction mixture was allowed
to warm to rt and stirred overnight, then butylamine hydrochloride
was filtered off, and the filtrate was evaporated to dryness. The
residue was dissolved in hexane and filtered once again to remove
the remains of butylamine hydrochloride. After removal of the
solvent the residue was purified by column chromatography on
SiO2 (CH2Cl2) affording 0.418 g (1.70 mmol, 71%) of 8 as a colorless
168.7, 139.3, 135.2, 124.4, 118.1, 69.7, 54.8, 35.7, 2þ9.4, 24.8, 24.1. MS
(EI): m/z (%) 333 (35) [M]þꢃ, 173 (40) [MeNHAr] . HRMS (EI): m/z
[M]þꢃ calcd for C14H14Cl3NO2 333.00901, Found 333.00842.
4.2.11. cis-3-Chloro-N-(4-nitrophenyl)-1,3-dimethyl-2-
oxocyclopentanecarboxamide (16b). To a solution containing 0.130 g
(0.94 mmol) of 4-nitroaniline and 0.15 mL (0.109 g, 1.08 mmol) of
Et3N in THF (3 mL), 0.18 g (0.86 mmol) of crude 14 (containing
catalytic amounts of TiCl4) were added and stirred overnight at rt.
Then the reaction mixture was filtered through a plug of Celite, and
the solvent was removed under vacuum. The residue was dissolved
in CHCl3 and purified by column chromatography on SiO2 (CHCl3)
affording 0.146 g (0.47 mmol, 55%) of 16b as pale yellow crystals.
liquid after solvent evaporation. 1H NMR (360 MHz, CDCl3):
d 6.41
(br s, 1H), 3.38e3.28 (m, 1H), 3.20e3.09 (m, 1H), 2.32e2.22 (m, 1H),
2.11e1.90 (m, 2H),1.78e1.68 (m, 1H), 1.66 (s, 3H), 1.50e1.38 (m, 2H),
1.43 (s, 3H), 1.35e1.21 (m, 2H), 0.88 (t, 3J¼7.3 Hz, 3H). 13C NMR
Mp: 129e130 ꢁC. 1H NMR (360 MHz, CDCl3):
d 8.89 (br s, 1H),
(90 MHz, CDCl3):
d
172.6, 134.7, 105.0, 82.8, 38.8, 31.5, 29.4, 25.2,
8.26e8.16 (m, 2H), 7.78e7.70 (m, 2H), 2.91e2.78 (m, 1H), 2.49e2.40
(m, 1H), 2.12e2.05 (m, 1H), 2.05e1.98 (m, 1H), 1.75 (s, 3H), 1.51 (s,
24.3, 19.7, 18.2, 13.6. MS (EI): m/zþ(%) 245 (20) [M]þꢃ, 209 (2þ0)
[MꢀHCl]þꢃ, 145 (10) [MꢀCONHBu] , 110 (100) [MꢀCONHBuCl]
,
3H). 13C NMR (90 MHz, CDCl3):
d 212.4, 168.9, 143.7, 143.3, 125.0,
ꢃ
69 (35) [C4H5O]þ. HRMS (EI): m/z [M]þ calcd for C12H20ClNO2
119.3, 69.7, 55.0, 35.6, 29.4, þ25.0, 24.1. MS (EI): m/z þ(%) 310 (85)
ꢃ
[M]þꢃ, 173 (85) [MꢀNHAr] . HRMS (EI): m/z [M]
calcd for
ꢃ
245.11826, Found 245.11787.
C
14H15ClN2O4 310.07203, Found 310.07170.
4.2.8. 5-Hydrohy-2,5-dimethyl-6-butylamino-6-oxohexanoic
acid
methyl ester (10). Compound 10 was prepared by dissolving 0.395 g
of 9 (1.61 mmol) in MeOH (5 mL). After that several drops (3e5) of
concd HCl were added to the solution, which was then stirred for
two days. The solvent was removed under vacuum and the product
Acknowledgements
€
We are grateful to Dr. T. Dulcks and Ms. D. Kemken (MS) and Dr.
Uli Papke (HR-ESI-MS, TU Braunschweig) for their help with the

J. Warneke et al. / Tetrahedron 70 (2014) 6515e6521
6521
10. (a) Whetstone, R. R. Dihydropyran Derivatives, U.S. Patent 2,479,283, Aug 16,
1949. (b) Stoner, G. G.; McNulty, J. S. J. Am. Chem. Soc. 1950, 72, 1531e1533; (c)
Hall, R. H. J. Chem. Soc. 1953, 1398e1402.
11. Methacrylic acids was also reported to form dimers, although at higher tem-
peratures and through a different mechanism: (a) Crawford, J. W. C. J. Soc. Chem.
Ind. 1947, 66, 155e156; (b) Albisetti, C. J.; England, D. C.; Hogsed, M. J.; Joyce, R.
M. J. Am. Chem. Soc. 1956, 78, 472e475.
characterization of new compounds. The X-ray diffractometer
was funded by NSF Grant 0087210, Ohio Board of Regents Grant
CAP-491, and by Youngstown State University. We are also indebted
€
to Prof. J. Beckmann, Dr. T. Dulcks, and Prof. Dietmar Kuck (Uni-
€
versitat Bielefeld) for helpful discussions.
12. Addition of AlCl3, TiCl4, SnCl4, or Et2O$BF3 does not lead to accelerated con-
version of 1 into 3.
13. Taking into account the composition of the 1/3 equilibrium mixture at the same
temperature and making an assumption that the kinetics of the dissociation of
3 is not significantly influenced by the media, we can estimate the second order
Supplementary data
Crystallographic data for the structure 16a has been deposited to
the Cambridge Crystallographic Data Centre under deposition
number CCDC 987551. Supplementary data associated with this
article and containing NMR spectra with signal assignments, MS
spectra, X-ray data of 16a. Supplementary data related to this article
can be found at http://dx.doi.org/10.1016/j.tet.2014.07.019.
dimerization constant of MAACl to be 4ꢂ10ꢀ4 mol sꢀ2
.
14. Pathways of thermal decomposition of 1 at 1500 K employing infrared laser-
powered homogeneous pyrolysis (IR LPHP) technique have been investigated
before, see: Allen, G. R.; Russell, D. K. New J. Chem. 2004, 28, 1107e1115.
15. Traces of water are necessary to complete the hydrolysis; reaction with dry
MeOH led to formation of a non-pure product.
16. Although compounds 7 and 10 should be isolated as diastereomeric mixtures,
no presence of additional signals is observed in NMR spectra of compound 7. In
13C NMR of 10, separate signals are observed for several resonances, giving
evidence for the predominant formation of one diastereomer with the di-
astereomeric ratio comparable to the one of compound 2.
References and notes
17. ca. 0.3 mL of 3 were left to hydrolyze in 5 mL round bottom flask under a hood.
For the analysis, small aliquots of a sample were taken and NMR spectra in dry
CDCl3 were measured.
1. Methacryloyl chloride is a simple derivative of methacrylic acid, an industrial
chemical produced in amount of ca. 2ꢂ106 t/year and known since 1865, see:
Bauer, W., Jr. Methacrylic Acid and Derivatives. Ullmann’s Encyclopedia of
Industrial Chemistry; Wiley-VCH GmbH, 2011.
2. SciFinder search for methacryloyl chloride using its CAS number (920-46-7)
gives more than 7000 hits with nearly 1:1 ratio between patents and research
papers (as of Jan. 2014).
3. Although compound 2 was isolated as a mixture of two diastereomeric pairs,
1H NMR spectrum of 2 could not be used for the determination of the di-
astereomeric ratio due to the complete signal overlap. The diastereomeric ratio
of ca. 5.5:1 was roughly estimated from 13C NMR spectrum, in which separate
resonances are observed for several corresponding carbon atoms of two di-
astereomers. Only the resonances of the major diastereomer are reported in the
Experimental section.
18. See Supplementary data for 1D and 2D NMR spectra, NMR signal assignment,
MS spectra, complete X-ray data, and description of kinetics measurement.
19. Formation of a similar tricyclic lactone lacking Cl-atom has been reported be-
fore for the oxidation product of methacrolein dimer, see Refs. 10b,c.
20. Reaction mixtures were analyzed using NMR in CDCl3. For FeCl3 sample direct
measurements could not be performed due to the paramagnetic character of Fe
(III). In this case, the mixture of acid chlorides was first esterified with MeOH,
purified, and then NMR analysis was performed.
21. Although several methods for relative comparison of strength of Lewis acids are
known, none of those afford a universal scale, because coordinative efficiency
of Lewis acids vary against different Lewis bases, see: Olah, G. A.; Prakash, G. K.
ꢂ
ꢂ
S.; Molnar, A.; Sommer, J. Superacid Chemistry; John Wiley & Sons: Hoboken, NJ,
2009; pp 1e34.
4. Musser, M. T. In Adipic Acid, Ullmann’s Encyclopedia of Industrial Chemistry;
Wiley-VCH GmbH: 2000.
22. X-ray structure of 15a has been previously reported by Fischer et al.5a (CCDC
ref. code entry: DAPJOT).
ꢁ
5. (a) Fischer, W.; Bellus, D.; Adler, A.; Francotte, E.; Roloff, A. Chimia 1985, 39,
23. For descriptions of graph-set notations, see: Bernstein, J.; Davis, R. E.; Shimoni,
L.; Chang, N.-L. Angew. Chem., Int. Ed. Engl. 1995, 34, 1555e1573.
24. (a) Desiraju, G. R.; Steiner, T. The Weak Hydrogen Bond in Structural Chemistry
and Biology; Oxford University: Oxford, UK, 1999; (b) Steiner, T. Angew. Chem.,
Int. Ed. 2002, 41, 48e76.
19e20; (b) Boldt, P.; Zippel, S.; Haucke, G. J. Chem. Res. 1997, 386e387; (c)
ꢂ
ꢂꢁ
Husar, B.; Moszner, N.; Lukac, I. Beilstein J. Org. Chem. 2012, 8, 337e343; (d)
Sinya, N. High-purity Methacryloyl Halides and Methacrylic Acid Derivatives
with Reduced Dimer Content, and Their Manufacture. Jpn. Kokai Tokkyo Koho
JP 2002187868 A, 20020705, 2002.
25. For properties and reactivity of carboxonium ions, see: Olah, G. A.; Prakash, G.
6. Although compound 3 was found to be present in PubChem (CID 13396407 and
CID 641639), no details about its preparation and purification have ever been
reported, no spectral characterization is available, and it has been never de-
scribed as a major impurity of commercial methacryloyl chloride.
7. On average, 3e5 g of 3 can be isolated from 25 mL samples of commercial
MAACl.
ꢂ
ꢂ
K. S.; Molnar, A.; Sommer, J. Superacid Chemistry; John Wiley & Sons: Hoboken,
NJ, 2009; pp 172e192.
26. For bi- and tricyclic carboxonium ions, prepared by acylation of cycloalkenes
with acylium salts, see: Lubinskaya, O. V.; Shashkov, A. S.; Chertkov, V. A.; Smit,
W. A. Synthesis 1976, 742e745.
€
27. Chevrier, B.; Weis, R. Angew. Chem., Int. Ed. Engl. 1974, 1, 1e10.
8. (a) Alder, K.; Ruden, E. Chem. Ber. 1941, 74B, 920e926; (b) Schulz, H.; Wagner, H.
28. Sheldrick, G. M. Acta Cryst. 2008, A64, 112e122.
Angew. Chem. 1950, 62, 105e118.
€
€
29. Hubschle, C. B.; Sheldrick, G. M.; Dittrich, B. J. Appl. Cryst. 2011, 44, 1281e1284.
9. Alder, K.; Offermanns, H.; Ruden, E. Chem. Ber. 1941, 74B, 905e920.