Model experiments implicate a benzoquinoneketene intermediate in LCP synthesis

Article abstract of DOI:10.1021/ma201105q

LCP synthesis from 4-hydroxybenzoic acid (HBA) and comonomers involves the formation of phenolic anion end groups at the later stages of polymerization. Model reactions show that these end groups can undergo cleavage to form 4-oxo-2,5-cyclohexadienylideneketene OCK (a benzoquinoneketene intermediate) with loss of a resonance-stabilized phenolic anion. With base, ethyl 4-(4′-hydroxybenzoyloxy)benzoate formed ethyl 4-hydroxybenzoate and poly-HBA, while ethyl 4-(4′-methoxybenzoyloxy)benzoate did not react under the same conditions, indicating an E1cB mechanism involving OCK as an intermediate. Attempted trapping of the OCK intermediate using cycloadditions failed, but it was successfully trapped with secondary amines. These results make it likely that LCP synthesis occurs at least partly via OCK. It is a ubiquitous intermediate in LCP polymerization and characterization.

Full text of DOI:10.1021/ma201105q

ARTICLE
pubs.acs.org/Macromolecules
Model Experiments Implicate a Benzoquinoneketene Intermediate in
LCP Synthesis
Jeffrey M. Robertson, Cristina G. Contreras, Robert B. Bates, and H. K. Hall, Jr.*
C. S. Marvel Laboratories, Department of Chemistry and Biochemistry, University of Arizona, Tucson, Arizona 85721, United States
ABSTRACT: LCP synthesis from 4-hydroxybenzoic acid
(HBA) and comonomers involves the formation of phenolic
anion end groups at the later stages of polymerization. Model
reactions show that these end groups can undergo cleavage to
form 4-oxo-2,5-cyclohexadienylideneketene OCK (a benzoqui-
noneketene intermediate) with loss of a resonance-stabilized
phenolic anion. With base, ethyl 4-(40-hydroxybenzoyloxy)-
benzoate formed ethyl 4-hydroxybenzoate and poly-HBA,
while ethyl 4-(40-methoxybenzoyloxy)benzoate did not react
under the same conditions, indicating an E1cB mechanism
involving OCK as an intermediate. Attempted trapping of the OCK intermediate using cycloadditions failed, but it was successfully
trapped with secondary amines. These results make it likely that LCP synthesis occurs at least partly via OCK. It is a ubiquitous
intermediate in LCP polymerization and characterization.
’ INTRODUCTION
In a separate physical organic chemistry investigation, Cevasco
et al.11 studied the reaction of 2,4-dinitrophenyl 4-hydroxybenzo-
ate with aniline in alkaline conditions; on the basis of sophisti-
cated kinetic measurements, they concluded that more anilide
was formed than expected. The reaction proceeded by two
different pathways, as shown in Scheme 2: the usual attack of
aniline at the carbonyl group via a tetrahedral intermediate led to
the expected anilide (BAc2 mechanism), but aniline also acted as
a base, forming the phenolic anion which eliminates OCK; the
latter reacts with aniline to form the anilide (E1cB elimination).
In these reactions of 2,4-dinitrophenyl 4-hydroxybenzoate, the
2,4-dinitrophenoxide anion provided an excellent leaving group.
It occurred to us that in alkaline conditions the phenolic end
groups in polymers from HBA are similarly constructed for OCK
formation, with the leaving group now being the polymeric
phenoxide anion stabilized by a 4-aryloxycarbonyl group. We
report on our study with model compounds to evaluate the
possibility that OCK elimination occurs under LCP polymeriza-
tion conditions from phenoxide ends of 4-hydroxybenzoic
acid units:
4-Hydroxybenzoic acid (HBA) is an essential component of
high-tech liquid crystalline polymers (LCP’s). Its structure leads
to all para-linked aromatic polyesters whose rodlike structures
give rise to thermotropic liquid crystal character. The homo-
polymer is intractable, but certain copolymers have excellent
physical properties and still permit processing and molding.1ꢀ3
The copolymerizations with terephthalic acid and hydroquinone
are performed by acetylating the p-hydroxyl group of the
monomers, followed by thermal polymerization at ∼300 °C.
At the beginning the polymerization proceeds via the expected
acidolysis mechanism (Scheme 1). As the polymerization pro-
ceeds, decarboxylation of the carboxylic acid end groups leads to
phenyl ester ends, while ketene loss from the acetate end groups
forms phenolic ends. These two new end groups can combine,
catalyzed by sodium or potassium acetate, to evolve phenol and
the molecular weight continues to increase during the later
stages.4,5
There are still many open questions regarding this high-
temperature polymerization. The mass spectra of these copo-
lymers showed evidence of stretches of homo-HBA,6 and it
has been an open question if these homopolymer stretches
were formed during the polymerization or due to scrambling
of 120 Da mass units during cationic mass spectral structure
determinations.7 In addition, the loss of 120 amu in the MS-
MS of anions from HBA oligomers was noted.8 4-Oxo-2,
5-cyclohexadienylideneketene (OCK) was proposed to be
responsible for both these observations. OCK may also be an
intermediate in the radical chain decomposition of acetate
groups to form the phenolic end groups.9 Though unstable
under ordinary conditions, OCK has been studied spectrally in
an argon matrix.10
’ RESULTS AND DISCUSSION
Synthesis of Model Compounds. Ethyl 4-(40-hydroxyben-
zoyloxy)benzoate 1 (hydroxyl dimer) was selected as a model for
Received: May 13, 2011
Revised:
June 13, 2011
Published: June 27, 2011
r
2011 American Chemical Society
5586
dx.doi.org/10.1021/ma201105q Macromolecules 2011, 44, 5586–5589
|

Macromolecules
ARTICLE
Scheme 1. Change of Mechanism in LCP Polymerization
Scheme 2. Reaction of 2,4-Dinitrophenyl 4-Hydroxybenzoate with Aniline in Basic Conditions11
Scheme 3. Syntheses of Hydroxyl Dimer 1 and Methoxyl
Dimer 2
Table 1. Molecular Formulas of MS Anions Observed by
ESIꢀa
mass
anion
mol form
calcd
obsd
strength
BOHꢀ
C7H5O3
137.024
165.055
257.045
285.076
377.066
405.097
497.087
525.119
137.026
165.055
257.046
285.077
377.064
405.096
497.091
525.119
m
m
s
BOEtꢀ
B2OHꢀ
B2OEtꢀ
B3OHꢀ2
B3OEtꢀ
B4OHꢀ
B4OEtꢀ
C9H9O3
C14H9O5
C16H13O5
C21H13O7
C23H17O7
C28H17O9
C30H21O9
s
m
m
w
w
a B = OCK unit [ꢀp-OCH6H4(CdO)ꢀ], e.g., BOHꢀ = p-HOC6H4-
(CdO)Oꢀ; s = strong, m = medium, w = weak.
The methoxyl dimer 2 was synthesized similarly from p-anisic
acid 6.
an end group in the later stages of the LCP polymerization.12
With the hydroxyl dimer, base could produce OCK by elimina-
tion; the methoxyl dimer, ethyl 4-(40-methoxybenzoyloxy)-
benzoate 2, served as a control which would not be able to
undergo this elimination. Dimers 1 and 2 were synthesized as
shown in Scheme 3. 4-Acetoxybenzoic acid 3 was converted to
the acid chloride and coupled with ethyl 4-hydroxybenzoate 4
to give ester 5. The acetyl group of dimer 5 was cleanly removed
by aminolysis with n-butylamine to give hydroxyl dimer 1.
Model Reactions with Base. Reactions were performed with
dimers 1 and 2 in attempts to implicate OCK as an inter-
mediate in the reactions of hydroxyl dimer 1. Hydroxyl dimer 1
was heated to 130 °C in the presence of potassium acetate and
18-crown-6 and stirred for 3 h, which are considerably milder
conditions than the authentic LCP polymerization. Poly-HBA
and ethyl 4-hydroxybenzoate 4 were the only products, as
shown by mass spectra which defined the sizes of the oligomers
5587
dx.doi.org/10.1021/ma201105q |Macromolecules 2011, 44, 5586–5589

Macromolecules
ARTICLE
as well as the end groups. Table 1 shows the anions observed in
negative ESI and Table 2 the cations in positive ESI. Cyclics
showed only in the cation spectra, since they have no protons
acidic enough to give anions for the anion spectra. These linear
and cyclic poly-HBA molecules are all presumed to be formed
by oligomerization of OCK.
In contrast, the methoxyl dimer 2 was recovered unchanged
under these conditions. This is strong evidence for OCK as an
intermediate and shows that OCK molecules can combine to
form oligomers.
Trapping Experiments. To definitely show that OCK was
involved, trapping experiments were designed. OCK resem-
bles p-benzoquinone and appeared amenable to trapping by
cycloaddition. The trapping agents tried were N-vinylcarba-
zole, N,N-dimethylisobutenylamine, diphenylisobenzofuran,
1-methoxycyclohexadiene, 1-methoxy-3-(trimethylsiloxy)-
1,3-butadiene, anthracene, and anthrone. In each case, the
hydroxyl dimer was heated in the presence of the trapping
reagent, potassium acetate, and 18-crown-6 to generate the
required phenoxy anion. In no case was a cycloadduct from
OCK characterized.
Accordingly, we returned to amines as traps. Since we
wished to work in bulk to mimic polymerization conditions,
dilute solutions as used by Cevasco were inappropriate.11 To
find conditions which exclude the direct displacement path-
way, the amine reactions were run using the methoxyl dimer
2: if no direct displacement (BAc2 mechanism) by an amine
took place with methoxyl dimer 2, it would not take place with
hydroxyl dimer 1. In fact, the methoxyl dimer 2 would react
faster than the hydroxyl dimer 1 because p-methoxyl is less
displacement, so p-anisidine was excluded as a trap. Day et al.13
showed that secondary amines are much less reactive than
primary amines in ester aminolysis. We tried N-methylpiper-
azine and cis-2,6-dimethylpiperidine, but again direct displace-
ment occurred. However, the more hindered diethylamine and
the less nucleophilic morpholine gave no direct displacement
on methoxyl dimer 2. Thus, these secondary amines should be
unreactive to the ester group but are able to deprotonate the
phenolic hydroxyl group. If OCK formed as an intermediate, it
could then combine with the amine to form the corres-
ponding amide.
Accordingly, these amines, acting as both base and trapping
agents, were heated with hydroxyl dimer 1 (Scheme 5). The
diethylamine reaction product was extracted three times with
hot water, leaving a small amount of insoluble gum. The
cooled water extracts were filtered to remove precipitated
ethyl p-hydroxybenzoate, and the filtrate was rotary evapo-
rated. The residue was a mixture of diethylamine, ethyl p-
hydroxybenzoate, and the desired p-hydroxyamide 7a. Mor-
pholine gave analogous amide 7b. No oligomers were found in
these experiments. In separate control experiments, ethyl
p-hydroxybenzoate was heated with these amines and recovered
unchanged.
Exclusion of Self Condensation To Form Oligomers. We
regard the HBA-oligomers as forming from polyadditions
of OCK. However, one other possibility was considered:
nucleophilic attack by the phenoxide anion on the acyl group
of the hydroxyl dimer might occur, again with displace-
ment of the stabilized p-ethoxycarbonylphenoxide anion.
This would lead successively to trimer, tetramer, etc. To
examine this possibility, model reactions were again em-
ployed. 4-Hydroxyacetophenone, simulating the anion of
the hydroxyl dimer 1, was allowed to react under our
standard conditions with methoxyl dimer 2. Only a very
small extent of reaction occurred (less than 5%). Accordingly,
this reaction can offer if anything only a minor route to the
oligomeric poly-HBA.
electron-donating than p-hydroxyl (Hammett σOCH3
=
ꢀ0.12, σOH = ꢀ0.38) and permits a more electophilic p-
carbonyl group.
Heating p-anisidine with methoxyl dimer 2 at 120 °C for 3 h
gave N-(p-methoxyphenyl)-p-methoxybenzamide by direct
Table 2. Molecular Formulas of MS Cations Observed by
ESI+a
mass
Scheme 5. E1cB Mechanism in the Aminolysis of Dimer 1
with Morpholine and Diethylamine
cation
mol form
calcd
obsd
strength
+
BOEtH2
C9H11O3
C14H9O4
C14H11O5
C16H15O5
C21H13O6
C21H15O7
C23H19O7
C28H17O8
C35H21O10
167.071
241.050
259.060
287.094
361.071
379.081
407.113
481.092
601.113
167.070
241.050
259.061
287.093
361.072
379.080
407.113
481.080
601.101
m
m
m
s
B2H+
+
B2OH3
+
B2OEtH2
B3H+
m
w
m
w
w
+
B3OH3
+
B3OEtH2
B4H+
B5H+
a B = OCK unit [ꢀp-OCH6H4(CdO)ꢀ], e.g., BOEtH2+ = EtOC6H4C-
+
(OH)2 ; italics = cyclic; s = strong, m = medium, w = weak.
Scheme 4. Reaction of Hydroxyl Dimer 1 with Base
5588
dx.doi.org/10.1021/ma201105q |Macromolecules 2011, 44, 5586–5589

Macromolecules
ARTICLE
’ CONCLUSIONS
(d, J = 8.5 Hz, OCCCH). 13C NMR: δ 121.62, 127.74, 134.79, 134.79,
151.19, 169.13, 170.46.
A model reaction of HBA dimer simulating LCP synthesis
gave poly-HBA by an E1cB mechanism via the ketene inter-
mediate OCK under very mild conditions. Cycloaddition trap-
ping of OCK failed, but it was successfully trapped with
secondary amines. Since LCP synthesis involves end groups like
dimer 1 at much higher temperatures than were used in these
model reactions, these results indicate that LPC synthesis occurs
partly via OCK. Some OCK must distill out and some must add
to phenoxide polymer ends, thus scrambling 4-hydroxybenzoic
acid units in the polymer ends. The unstable substance OCK
appears to be a ubiquitous intermediate in anionic, cationic, and
free radical aspects of LCP synthesis and characterization.
(4-Hydroxyphenyl)(morpholino)methanone 7b. Dimer 1
(0.5 g, 1.7 mmol) was mixed with morpholine (0.23 g, 2.6 mmol) at
120 °C for 3 h. HBA was extracted 3ꢁ with hot water and rotary
evaporated, leaving 4 and 7b (184 mg, 52.3%, mp 115ꢀ120 °C). 1H and
13C NMR spectra as reported.14
’ ACKNOWLEDGMENT
We are indebted to Solvay Advanced Polymers, Alpharetta,
GA, to University of Arizona/NASA Space Grant Program, and
to Dr. Anne Padias and Samiul Ahad for help with the
manuscript.
’ EXPERIMENTAL SECTION
NMR and Mass Spectra. 1H NMR spectra at 500 MHz and 13
C
’ REFERENCES
spectra at 125 MHz were recorded in CDCl3 except where noted on a
Bruker DRX 500 spectrometer. ESI mass spectra were obtained on a
Bruker 9.4 T ApexQh FT-ICR instrument with electrospray ionization,
both in the positive and negative ion modes.
Poly-HBA from Dimer 1. A mixture of dimer 1 (100 mg, 0.35
mmol), potassium acetate (2 mg, 0.0175 mmol), and 18-crown-6 (5 mg,
0.0175 mmol) was heated at 130 °C for 3 h. The resulting solid was
triturated 3ꢁ with ether; the ether solutions contained ethyl 4-hydro-
xybenzoate 4. The residue was shown to be HBA oligomers by NMR on
a Bruker DRX 600 in hexafluoroisopropanol (HFIP) and by mass
spectrometry. Both cyclic and linear oligomers formed as indicated in
Tables 1 and 2.
(1) Jin, J. I.; Kang, C. S. Prog. Polym. Sci. 1997, 22, 937.
(2) Han, H.; Bhowmik, P. K. Prog. Polym. Sci. 1997, 22, 1431.
(3) Negi, Y. S.; Goyal, R. K. Int. J. Plast. Technol. 2003, 7, 99.
(4) Han, X.; Williams, P. A.; Padias, A. B.; Hall, H. K., Jr.; Sung,
H. N.; Linstid, H. C.; Lee, C. Macromolecules 1996, 29, 8313.
(5) Leblanc, J. P.; Huang, J.; Padias, A. B.; Hall, H. K., Jr. J. Polym. Sci.,
Polym. Chem. Ed. 1992, 30, 2321.
(6) Somogyi, A.; Bojkova, N.; Padias, A. B.; Hall, H. K., Jr. Macro-
molecules 2005, 38, 4067.
(7) Elandaloussi, E. H.; Somogyi, A.; Padias, A. B.; Bates, R. B.; Hall,
H. K., Jr. Macromolecules 2006, 39, 6913.
(8) Hall, H.; Ahad, S.; Bates, R.; Bertucci, M.; Contreras, C.;
Dospinoiu, A.; Lin, G.; Singletary, N.; Somogyi, A. Polymers 2011, 3, 367.
(9) Stuparu, M. C.; Xu, J.; Hall, H. K., Jr. Tetrahedron Lett. 2009,
50, 6743.
Ethyl 4-(40-Hydroxybenzoyloxy)benzoate 1. A solution of
4-acetoxybenzoic acid 3 (3.92 g, 21.7 mmol), thionyl chloride (8 mL,
109 mmol), and DMF (4 drops) was stirred under a stream of argon
until excess thionyl chloride was removed (30 min). A solution of ethyl
4-hydroxybenzoate 4 (3.61 g, 21.7 mmol) in pyridine (5 mL) was added
at 0 °C. After 2 h, the mixture was poured onto ice and warmed to room
temperature. The mixture was filtered and washed with a solution of
K2CO3 (8.92 g, mp 102ꢀ105 °C). The resulting powder was dissolved
in n-butylamine (1.58 g, 21.7 mmol), stirred for 1 h, poured onto ice, and
warmed to room temperature. The mixture was filtered, yielding 112 (5.4
g, 87%, mp 164 °C). 1H NMR: δ 1.406 (t, J = 7.3 Hz, CH3), 4.394 (q, J =
7.3 Hz, CH2), 6.929 (d, J = 9.0 Hz, H10), 7.283 (d, J = 9.0 Hz, H4), 8.102
(d, J = 8.5 Hz, H9), 8.119 (d, J = 8 Hz, H3). 13C NMR: δ 14.32, 61.18,
115.53, 121.24, 121.79, 127.88, 131.14, 132.68, 154.69, 160.97, 161.71,
164.43.
(10) Sander, W.; Mueller, W.; Sustmann, R. Angew. Chem. 1988,
100, 577.
(11) Cevasco, G.; Guanti, G.; Hopkins, A. R.; Thea, S.; Williams, A.
J. Org. Chem. 1985, 50, 479.
(12) Kimura, T.; Duan, X.; Kato, M.; Okada, S.; Yamada, S.;
Matsuda, H.; Nakanishi, H. Polymer 1998, 39, 491.
(13) Arnett, E.; Miller, J. G.; Day, A. R. J. Am. Chem. Soc. 1951,
73, 5393.
(14) O’Mahony, G.; Pitts, A. Org. Lett. 2010, 12, 2024.
Ethyl 4-(4-Methoxybenzoyloxy)benzoate 2. A solution of p-
anisic acid 6 (1 g, 6.6 mmol), thionyl chloride (3 mL, 41 mmol), and
DMF (2 drops) was stirred under a stream of argon until excess thionyl
chloride was removed (30 min). A solution of ethyl 4-hydroxybenzoate 4
(1.09 g, 6.6 mmol) in pyridine (2 mL) was added at 0 °C. After 1 h, the
mixture was poured onto ice and warmed to room temperature. The
1
mixture was filtered, yielding 212 (1.71 g, 86%, mp 92ꢀ94 °C). H
NMR: δ 1.405 (t, J = 7.3 Hz, CH3), 3.905 (s, OCH3), 4.381 (q, J = 7.3
Hz, CH2), 6.994 (d, J = 9.0 Hz, H10), 7.288 (d, J = 8.5 Hz, H4), 8.120 (d,
J = 9.0 Hz, H9), 8.154 (d, J = 9.0 Hz, H3). 13C NMR: δ 14.34, 55.54,
61.05, 113.92, 121.37, 121.75, 127.92, 131.10, 132.38, 154.67, 164.09,
164.33, 165.89.
N,N-Diethyl-4-hydroxybenzamide 7a. Dimer 1 (1 g, 3.4
mmol) was mixed with diethylamine (0.38 g, 5.2 mmol) at 120 °C for
3 h. The mixture was extracted 3ꢁ with hot water, leaving a viscous
liquid. After cooling to room temperature, 5 was filtered off and the
solvent was removed by rotary evaporation to give 7a (26 mg, 8%, mp
1
132ꢀ135 °C). H NMR: δ 1.145 (br s, CH3), 1.235 (br s, CH3),
3.320 (br s, CH2), 3.557 (br s, CH2), 6.768 (d, J = 8.0 Hz, OCCH), 7.417
5589
dx.doi.org/10.1021/ma201105q |Macromolecules 2011, 44, 5586–5589