Functional, hyperbranched polyesters via Baylis-Hillman polymerization

Article abstract of DOI:10.1002/pola.24954

Hyperbranched polyesters are among the most common hyperbranched polymers. One of the interesting features of hyperbranched polyesters is that they contain unreacted hydroxyl and carboxylic acid groups at the linear and terminal structural units, which can be postmodified to adjust thermal, solubility, or mechanical properties, or to prepare core-shell type architectures. This article reports on the synthesis of a novel class of hyperbranched polyesters via an A₂ + B₃ type Baylis-Hillman polymerization of 2,6-pyridinedicarboxaldehyde and trimethylolpropane triacrylate. Baylis-Hillman polymerization generates highly functional polyesters that contain not only unreacted aldehyde and/or acrylate groups at the linear and terminal structural units but also chemically orthogonal vinyl and hydroxyl groups along the polymer backbone. Using 3-hydroxyquinuclidine as the catalyst, hyperbranched polymers with number-average molecular weights up to 7500 g/mol and degrees of branching up to 0.81 were obtained. To demonstrate the versatility of these hyperbranched polyesters to act as platforms for further derivatization, the orthogonal postpolymerization modification of the hydroxyl, vinyl, and pyridine functional moieties with phenyl isocyanate, methyl-3-mercaptopropionate, and methyl iodide is presented.

Full text of DOI:10.1002/pola.24954

WWW.POLYMERCHEMISTRY.ORG
ARTICLE
Functional, Hyperbranched Polyesters Via Baylis–Hillman Polymerization
Sanhao Ji,1 Bernd Bruchmann,2 Frederik Wurm,1 Harm-Anton Klok1
1
´
Ecole Polytechnique Fe´de´rale de Lausanne (EPFL), Institut des Mate´riaux and Institut des Sciences et Inge´nierie Chimiques,
Laboratoire des Polyme`res, Baˆtiment MXD, Station 12, CH-1015 Lausanne, Switzerland
2Polymer Research, BASF SE, D-67056 Ludwigshafen, Germany
Correspondence to: H.-A. Klok (E-mail: harm-anton.klok@epfl.ch)
Received 7 June 2011; accepted 11 August 2011; published online 9 September 2011
DOI: 10.1002/pola.24954
ABSTRACT: Hyperbranched polyesters are among the most com-
mon hyperbranched polymers. One of the interesting features of
hyperbranched polyesters is that they contain unreacted hydroxyl
and carboxylic acid groups at the linear and terminal structural
units, which can be postmodified to adjust thermal, solubility, or
mechanical properties, or to prepare core–shell type architec-
tures. This article reports on the synthesis of a novel class of
hyperbranched polyesters via an A2 þ B3 type Baylis–Hillman po-
lymerization of 2,6-pyridinedicarboxaldehyde and trimethylolpro-
pane triacrylate. Baylis–Hillman polymerization generates highly
functional polyesters that contain not only unreacted aldehyde
and/or acrylate groups at the linear and terminal structural units
but also chemically orthogonal vinyl and hydroxyl groups along
the polymer backbone. Using 3-hydroxyquinuclidine as the cata-
lyst, hyperbranched polymers with number-average molecular
weights up to 7500 g/mol and degrees of branching up to 0.81
were obtained. To demonstrate the versatility of these hyper-
branched polyesters to act as platforms for further derivatization,
the orthogonal postpolymerization modification of the hydroxyl,
vinyl, and pyridine functional moieties with phenyl isocyanate,
methyl-3-mercaptopropionate, and methyl iodide is presented.
C
V
2011 Wiley Periodicals, Inc. J Polym Sci Part A: Polym Chem
50: 25–34, 2012
KEYWORDS: A2 1 B3 approach; Baylis Hillman reaction; func-
tionalization of polymers; hyperbranched; polyesters
INTRODUCTION Hyperbranched polymers are an attractive
class of materials that possess some of the properties that
are characteristic for highly branched polymer architectures,
such as high functionality, enhanced solubility, low viscosity,
and low crystallinity, while obviating the tedious, stepwise
synthetic protocols that are required for the synthesis of
dendrimers.1–4 Over the past decades, a broad variety of
hyperbranched polymers has been synthesized. Among the
many types of hyperbranched polymers that have been
reported, hyperbranched polyesters are one of the most
dominating representatives.5
This contribution explores the feasibility of the Baylis–Hill-
man reaction for the synthesis of hyperbranched polyesters.
The Baylis–Hillman reaction involves the base-catalyzed reac-
tion of an a,b-unsaturated carbonyl compound with an alde-
hyde to form an a-methylene-b-hydroxycarbonyl compound
(Scheme 1).12–14 This reaction has attracted interest in or-
ganic synthesis as it is an atom-economical carbon–carbon
bond forming reaction that can be carried out with control
over stereochemistry and generates a polyfunctional scaffold
that can be converted to a variety of other products. Very
recently, also first examples of the use of the Baylis–Hillman
reaction for the synthesis of linear side-chain functional pol-
yesters were reported.15,16 Two characteristics that make
this reaction an attractive tool in polymer synthesis include:
(i) Baylis–Hillman polymerization results in polyesters with
two chemically orthogonal side-chain functional groups, viz.
a vinyl group and an hydroxyl group, which can subse-
quently be selectively modified via an appropriate postpoly-
merization modification reaction; (ii) as the vinyl and
hydroxyl groups are generated during the CAC bond forming
process, Baylis–Hillman polymerization allows the prepara-
tion of side-chain functional polyesters without the need for
laborious protective group chemistry. While the previous
Hyperbranched polyesters can be synthesized via various
approaches but are most often prepared via AB2 or A2 þ B3 type
step polymerizations.6–11 The resulting hyperbranched polymers
contain hydroxyl and/or carboxylic acid functional groups at the
linear and terminal structural units, which are amenable to post-
polymerization modification and can be used to tune the thermal,
solubility, and mechanical properties of these materials or to pre-
pare core–shell type architectures. Access to hyperbranched
polymers that contain multiple, chemically orthogonal functional
groups, which can be regioselectively addressed would enable
enhanced control of the properties of these macromolecules and
further expand their scope of applications.
Additional Supporting Information may be found in the online version of this article.
C
V
2011 Wiley Periodicals, Inc.
WWW.MATERIALSVIEWS.COM
JOURNAL OF POLYMER SCIENCE PART A: POLYMER CHEMISTRY 2012, 50, 25–34
25

ARTICLE
WWW.POLYMERCHEMISTRY.ORG
Finnigan SSQ 710C (Finnigan-MAT, Bremen, Germany)
equipped with an ESI interface. Data were acquired using
the ICIS software running on a Digital Unix workstation. Fou-
rier transform reflectance infrared spectra were acquired
using a nitrogen-purged Nicolet Magna-IR 560 spectrometer
equipped with a Micro Specular Reflectance accessory (Spe-
cac) and processed using the OMNIC ESP 5.1 software.
SCHEME 1 Baylis-Hillman reaction.
published reports exclusively focused on the synthesis of lin-
ear polymers, this contribution investigates the A2 þ B3 type
Baylis–Hillman polymerization of 2,6-pyridinedicarboxalde-
hyde and trimethylolpropane triacrylate (TMPTA) to gener-
ate highly functional, hyperbranched polyesters and explores
the postpolymerization modification of the vinyl, hydroxyl,
and pyridine moieties of these materials.
Procedures
2,6-Pyridinedicarboxaldehyde (2)
In a 500 mL round bottom flask equipped with a condenser,
2,6-pyridinedimethanol (5 g, 37 mmol) was dissolved in hot
CHCl3 (250 mL). The flask was cooled to room temperature,
and activated manganese oxide (60 g, 20 equiv) was added
as a solid. The mixture was vigorously stirred under reflux
for 4 days until most of the starting material was consumed.
After cooling to room temperature, the suspension was fil-
tered over Celite. The oxidized products were purified by
flash column chromatography on silica gel, eluting with ethyl
acetate and dichloromethane (1:1, v/v), which afforded 2 as
a white, pure solid (2.1 g, yield 42%).
EXPERIMENTAL
Materials
All reagents and solvents were of commercial grade and
used as received. TMPTA was purified by flash column chro-
matography on silica gel using hexane and ethyl acetate (4:1,
v/v) as the eluent. 2,6-Pyridinedimethanol was obtained
from Acros. All other chemicals and reagents were acquired
from Sigma–Aldrich (Buchs, Switzerland). Deuterated sol-
vents for NMR spectroscopy were acquired from Armar
Chemicals (Do¨ttigen, Switzerland).
1H NMR (400 MHz, CDCl3): 10.16 (s, 2H, HC¼¼O), 8.16 (d, J ¼
7.8 Hz, 2H, pyridine-H), 8.07 (t, J ¼ 7.8 Hz, 1H, pyridine-H).
13C NMR (101 MHz, CDCl3): 192.34, 152.98, 138.37, 125.32.
MS (ESI): 136.96 [MþH]þ. 1H NMR, 13C NMR, and ESI-MS
spectra are included in Supporting Information (Figs. S1–S3).
Analytical Methods
Mono(b-hydroxy-a-methylene-2-pyridinepropanoic
acid)-diacryloyl Trimethylolpropane (3)
Gel permeation chromatography (GPC) was performed on a
Waters Alliance GPCV 2000 system equipped with refractive
index, differential viscometer, and light scattering detectors.
Separation was carried out at 60 ꢀC with TSK-Gel Alpha
2500 þ 3000 þ 4000 columns, using vacuum distilled High-
performance liquid chromatography (HPLC) grade N,N-dime-
thylformamide (DMF) þ 0.5 g/L LiCl as eluent and a flow
rate of 0.6 mL/min. Molecular weights were determined
using conventional and universal calibration curves, which
were created with narrow polydispersity poly(methyl meth-
To a stirred mixture of 2-pyridinecarboxaldehyde (164 lL,
1.68 mmol) and TMPTA (500 mg, 1.68 mmol), 1,4-diazabicy-
clo[2.2.2]octane (DABCO) (180 mg, 1.6 mmol) and methanol
(120 lL, 3 mmol) were added. DMF (0.2 mL) was added to
help dissolve the reactants. The brown, homogeneous reaction
mixture was stirred at ambient temperature for 5 h, and the
progress of the reaction was monitored by thin layer chroma-
tography (TLC) (solvent: ethyl acetate). On completion, the
reaction mixture was purified by flash column chromatogra-
phy on silica gel, eluting with ethyl acetate to give the desired
product 3 as light yellow oil (545.6 mg, yield 80.5%).
acrylate) standards. Mark–Houwink
a parameters were
determined by universal calibration. Results were calculated
with the Empower Pro multidetection GPC software (Version
5.00). The interdetector volume was adjusted from the peak
position of uniform poly(ethylene glycol) (PEG) oligomers.
The volume of the injected loop was 0.214 mL, and the poly-
mer concentration was calculated to give a viscometric signal
less than 0.5% of the baseline level. NMR spectra were
recorded on Bruker ARX-400 and ARX-600 spectrometers.
CDCl3 was used as the solvent. For 1H NMR spectroscopy,
chemical shifts are reported in ppm relative to the solvent’s
residual 1H signal (CDCl3: 7.25 ppm). Coupling constants J
are given in Hz. 13C NMR spectra were recorded at 101 MHz.
The 13C signal of the solvent (CDCl3: 77 ppm) was used as
internal reference. Coupling constants J are given in Hz. 13C
NMR spectra of hyperbranched polymer 6 were recorded at
150.9 MHz. To allow integration, inverse gated proton
decoupled 13C NMR spectra were recorded with a pulse
interval of 10 s to allow complete recovery of all carbons.
Electrospray ionization-mass spectrometry (ESI-MS) analysis
was performed on a single quadrupole mass spectrometer
1H NMR (400 MHz, CDCl3): 8.53 (d, J ¼ 2 Hz, 1H, pyridine-H),
7.65 (t, J ¼ 2 Hz, 1H, pyridine-H), 7.38 (d, J ¼ 2 Hz, 1H, pyri-
dine-H), 7.18 (t, J ¼ 2 Hz, 1H, pyridine-H), 6.34 (t, J ¼ 3.6 Hz,
3H, ¼¼CH2), 6.07 (m, J ¼ 3.8 Hz, 2H, ACH¼¼CH2), 5.96 (s, 1H,
¼¼CH2), 5.83 (d, 2H, CH¼¼CH2), 5.56 (s, 1H, ACHAOH), 4.82 (s,
1H, AOH), 4.03 (s, AOACH2A, 6H), 1.43 (m, J ¼ 6.2 Hz, 2H,
ACH2ACH3), 0.84 (t, J ¼ 6.2 Hz, 3H, ACH2ACH3). 13C NMR
(101 MHz, CDCl3): 165.6, 159.24, 148.18, 141.33, 136.87,
131.21, 127.81, 122.65, 120.93, 72.25, 63.91, 40.79, 22.95, 7.22.
1
MS (ESI): 404.66 [MþH]þ. H NMR, 13C NMR, and ESI-MS spec-
tra are included in Supporting Information (Figs. S4–S6).
2-((Acryloyloxy)methyl)-2-ethylpropane-1,3-diyl
Bis(2-(hydroxy(pyridin-2-yl)methyl)acrylate) (4)
To a stirred mixture of 2-pyridinecarboxaldehyde (328 lL,
3.36 mmol) and TMPTA (500 mg, 1.68 mmol), DABCO (240
mg, 2.1 mmol) and methanol (120 lL, 3 mmol) were added.
DMF (0.2 mL) was added to help dissolve the reactants. The
26
JOURNAL OF POLYMER SCIENCE PART A: POLYMER CHEMISTRY 2012, 50, 25–34

WWW.POLYMERCHEMISTRY.ORG
ARTICLE
brown, homogeneous reaction mixture was stirred at ambi-
ent temperature for 5 h, and the progress of the reaction
was monitored by TLC (solvent: ethyl acetate). On comple-
tion, the reaction mixture was purified by flash column chro-
matography on silica gel, eluting with ethyl acetate to give
the desired product 4 as light yellow oil (643.3 mg, yield
75%).
tated hyperbranched polyester was isolated by centrifugation
and obtained as a white powder. The molecular weights of
the polymer samples analyzed at different reaction times are
summarized in Table 1.
1H NMR (400 MHz, CDCl3): 9.96 (HC¼¼O), 7.82 (pyridine-H),
7.61 (pyridine-H), 7.2 (pyridine-H), 6.31 (¼¼CH2), 6.13
(¼¼CH2), 5.47 (ACHAOH), 4.84 (AOH), 3.78 (AOACH2ACA),
1.23 (ACH2ACH3), 0.77 (ACH2ACH3). 13C NMR (150.9 MHz,
CDCl3): 192.43, 165.65, 158.68, 151.36, 140.94, 138.25,
127.72, 125.12, 120.34, 72.5, 64.07, 40.67, 22.85, 7.52. 1H
NMR is included in Supporting Information (Fig. S13).
1H NMR (400 MHz, CDCl3): 8.51 (d, J ¼ 2 Hz, 2H, pyridine-
H), 7.65 (t, J ¼ 2 Hz, 2H, pyridine-H), 7.36 (d, J ¼ 2 Hz, 2H,
pyridine-H), 7.18 (t, J ¼ 2 Hz, 2H, pyridine-H), 6.34 (d, J ¼
3.6 Hz, 3H, ¼¼CH2), 6.07 (m, J ¼ 3.8 Hz, 1H, ACH¼¼CH2), 5.93
(s, 2H, ACH¼¼CH2), 5.82 (d, 1H, ACH¼¼CH2), 5.50 (s, 2H,
ACHAOH), 4.84 (s, 2H, AOH), 3.98 (m, AOACH2A, 6H), 1.29
(m, J ¼ 6.2 Hz, 2H, ACH2ACH3), 0.77 (t, J ¼ 6.2 Hz, 3H,
ACH2ACH3). 13C NMR (101 MHz, CDCl3): 165.51, 159.28,
148.17, 141.3, 136.9, 131.21, 127.75, 122.65, 120.94, 72.24,
64.07, 40.77, 22.77, 7.14. MS (ESI): 511.63 [MþH]þ. 1H
NMR, 13C NMR, and ESI-MS spectra are included in Support-
ing Information (Figs. S7–S9).
Procedure 2. TMPTA 1 (408 lL, 1.5 mmol) was dissolved in
0.5 mL DMF and added dropwise over 30 min into a mixture
of 2,6-pyridinedicarboxaldehyde 2 (204.7 mg, 1.5 mmol) and
3-HQD (100 mg, 0.8 mmol), which was dissolved in 0.5 mL
DMF. The homogeneous reaction mixture was stirred at ambi-
ent temperature. The course of the reaction was followed by
1H NMR spectroscopy and GPC. To this end, aliquots were
withdrawn from the reaction mixture, and polymer samples
were isolated as described in Procedure 1.
2-Ethyl-2-(((2-(hydroxy(pyridin-2-yl)methyl)acryloyl)oxy)-
methyl)propane-1,3-diyl Bis(2-(hydroxy(pyridin-2-yl)
methyl)acrylate) (5)
Procedure 3. To a stirred mixture of 2,6-pyridinedicarboxal-
dehyde 2 (256.1 mg, 1.9 mmol) and TMPTA 1 (334 lL, 1.2
mmol), 3-HQD (99 mg, 0.8 mmol) was added. DMF (1.0 mL)
was added to help dissolve to reactants, and the homogene-
ous reaction mixture was stirred at ambient temperature.
The course of the reaction was followed by 1H NMR spec-
troscopy and GPC. To this end, aliquots were withdrawn
from the reaction mixture, and polymer samples were iso-
lated as described in Procedure 1.
To a stirred mixture of 2-pyridinecarboxaldehyde (530 lL,
5.57 mmol) and TMPTA (487 mg, 1.64 mmol), DABCO (360
mg, 3.2 mmol) and methanol (180 lL, 4.5 mmol) were added.
DMF (0.2 mL) was added to help dissolve the reactants. The
brown, homogeneous reaction mixture was stirred at ambient
temperature for 5 h, and the progress of the reaction was
monitored by TLC (solvent: ethyl acetate). On completion, the
reaction mixture was purified by flash column chromatogra-
phy on silica gel, eluting with ethyl acetate to give the desired
product 5 as a light yellow oil (780 mg, yield 77%).
Procedure 4. TMPTA 1 (342 lL, 1.3 mmol) was dissolved
in 0.5 mL DMF and added dropwise over 30 min into a mix-
ture of 2,6-pyridinedicarboxaldehyde
2 (257.9 mg, 1.9
1H NMR (400 MHz, CDCl3): 8.42 (d, J ¼ 2 Hz, 3H, pyridine-
H), 7.59 (t, J ¼ 2 Hz, 3H, pyridine-H), 7.31 (d, J ¼ 2 Hz, 3H,
pyridine-H), 7.11 (t, J ¼ 2 Hz, 3H, pyridine-H), 6.25 (s, 3H,
¼¼CH2), 5.88 (s, 3H, ¼¼CH2), 5.55 (s, 3H, ACHAOH), 5.04 (br,
3H, AOH), 3.8 (m, AOACH2A, 6H), 1.08 (m, J ¼ 6.2 Hz, 2H,
ACH2ACH3), 0.64 (t, J ¼ 6.2 Hz, 3H, ACH2ACH3). 13C NMR
(101 MHz, CDCl3):165.38, 159.32, 148.12, 141.28, 136.86,
127.58, 122.58, 120.9, 72.19, 63.87, 40.72, 22.56, 7.02. MS
(ESI): 618.56 [MþH]þ. 1H NMR, 13C NMR, and ESI-MS spec-
tra are included in Supporting Information (Figs. S10–S12).
mmol) and 3-HQD (105.4 mg, 0.8 mmol), which was dis-
solved in 0.5 mL DMF. The homogeneous reaction mixture
was stirred at ambient temperature. The course of the reac-
1
tion was followed by H NMR spectroscopy and GPC. To this
end, aliquots were withdrawn from the reaction mixture, and
polymer samples were isolated as described in Procedure 1.
Postpolymerization Modification of 6 with Methyl 3-Mer-
captopropionate (7). Polymer 6 (120 mg, Table 1, Proce-
dure 3 after 2 h) was dissolved in dry THF (0.2 mL), and pyri-
dine (0.2 mL) was added to the mixture. Then, a large excess
of methyl 3-mercaptopropionate (56 lL mL, 10 mmol) was
added. The solution was stirred overnight at ambient tempera-
ture. The postmodified polymer was precipitated by addition
of cold diethyl ether (100 mL) and isolated by centrifugation
to afford 7 as a white powder (93.5 mg). Mn ¼ 15,400; Mw/Mn
¼ 2.58. The GPC elugram is included in Supporting Informa-
tion (Fig. S14). 1H NMR spectroscopy indicated quantitative
conversion of the double bonds.
Polymer 6
Procedure 1. To a stirred mixture of 2,6-pyridinedicarboxal-
dehyde 2 (195.4 mg, 1.4 mmol) and TMPTA 1 (390 lL, 1.4
mmol), 3-hydroxyquinuclidine (3-HQD; 100 mg, 0.8 mmol)
was added. DMF (1 mL) was added to help dissolve the reac-
tants, and the homogeneous reaction mixture was stirred at
ambient temperature. The course of the reaction was fol-
lowed by 1H NMR spectroscopy and GPC. To this end, 450
lL aliquots were taken from the reaction mixture at defined
time intervals. These samples were diluted with chloroform
(50 mL) and then washed with a saturated solution of aque-
ous NaHCO3, filtered and concentrated under reduced pres-
sure. The resulting material was dissolved in chloroform (2
mL) and precipitated in diethyl ether (100 mL). The precipi-
1H NMR (400 MHz, CDCl3): 9.99 (HC¼¼O), 7.72, 7.32 (pyri-
dine-H), 5.04 (ACHAOH), 4.45 (AOH), 3.88 (AOACH2ACA),
3.61 (AOCH3, 6H), 3.15 (ACHACH2ASA), 2.84 (ACH2ASA
CH2ACH2), 2.6 (ASACH2ACH2A), 2.42 (ASACH2ACH2A),
1.32 (ACHACH3), 0.76 (ACHACH3).
WWW.MATERIALSVIEWS.COM
JOURNAL OF POLYMER SCIENCE PART A: POLYMER CHEMISTRY 2012, 50, 25–34
27

ARTICLE
WWW.POLYMERCHEMISTRY.ORG
TABLE 1 Summary of Reaction Conditions Evaluated for the Baylis–Hillman Polymerization of 1 and 2 and Molecular
Characterization Data of the Resulting Polymers
a
b
c
b
Monomer Molar
Ratio (A2:B3)
Polymerization
Time (h)
Mw
Mn
Mn
Mw/Mn
(ꢁ)
Procedure
Addition Mode
(g/mol)
(g/mol)
(g/mol)
ac
DBd
1
1
1
2
1:1
1:1
1:1
1:1
1
2
3
1
One-pot
One-pot
One-pot
6,800
8,500
3,700
4,200
2,600
5,000
3,600
4,200
2,300
5,400
1.9
1.9
2.6
3.3
0.26
0.25
0.24
0.31
0.57
0.66
0.36
0.65
7,200
Slow monomer
addition
13,400
2
2
1:1
1:1
2
3
Slow monomer
addition
27,100
28,100
7,300
7,500
9,300
9,700
4.1
4.0
0.35
0.36
0.72
0.68
Slow monomer
addition
3
3
3
4
3:2
3:2
3:2
3:2
1
2
3
1
One-pot
One-pot
One-pot
11,600
19,700
17,300
8,800
2,800
5,100
3,300
4,300
4,200
8,700
5,700
4,500
2.3
2.2
3.1
2.7
0.30
0.29
0.27
0.30
0.79
0.61
0.63
0.81
Slow monomer
addition
4
4
a
3:2
3:2
2
3
Slow monomer
addition
20,300
5,200
–
6,100
–
4.6
–
0.33
–
0.72
–
Slow monomer
addition
Gelation
c
Determined by GPC with a laser light scattering detector.
From GPC determined using universal calibration.
Calculated from eq 1.
b
d
Measured by GPC using a conventional calibration curve, which was
created with PMMA standards.
RESULTS AND DISCUSSION
Postpolymerization Modification of 6 with Phenyl Isocya-
nate (8). To polymer 6 (96.2 mg, Table 1, Procedure 3 after
2 h) dissolved in chloroform (1.2 mL), phenyl isocyanate
(120 lL, 1.1 mmol) was added, and the solution was stirred
for 6 h at ambient temperature. The postmodified product
was precipitated by the addition of cold diethyl ether (100
mL) and isolated by centrifugation. Polymer 8 was obtained
as a transparent oil (103.3 mg). Mn ¼ 3900; Mw/Mn ¼ 11.6.
The GPC elugram is included in Supporting Information (Fig.
Polymerization
To evaluate the feasibility of the Baylis–Hillman reaction for
the synthesis of hyperbranched polyesters, this report investi-
gates the A2 þ B3 polymerization of 2,6-pyridinedicarboxalde-
hyde (2) and TMPTA (1) (Scheme 2). These monomers were
chosen as the Baylis–Hillman reactions between 2,6-pyridine-
dicarboxaldehyde and methyl acrylate17 and 2-pyridinecar-
boxaldehyde and methyl acrylate18 have been reported to pro-
ceed with 93–100% yield, which is an important requirement
to obtain high molecular weight polymers in a step-type poly-
merization. All polymerizations of 1 and 2 were carried out in
DMF, which was found to be a good solvent for both the mono-
mers and the polymer and ensured homogeneous reaction
conditions, and used 3-HQD as the catalyst. A number of poly-
merization procedures were evaluated, which differed with
respect to the molar ratio of monomers used, the mode of
monomer addition as well as reaction time. The resulting poly-
mers were analyzed with GPC and NMR. Table 1 summarizes
the reaction conditions that were evaluated and the character-
istics of the final polymers.
1
S14). H NMR spectroscopy indicated 83% conversion of the
hydroxyl groups.
1H NMR (400 MHz, CDCl3): 6.5–8.0 (aromatic ring), 6.26
(¼¼CH2), 5.82 (¼¼CH2), 5.31 (ACHAOH), 3.9 (AOACH2ACA),
1.25 (ACH2ACH3), 0.77 (ACH2ACH3).
Postpolymerization Modification of 6 with Methyl Iodide
(9). To the polymer 6 (40 mg, Table 1, Procedure 3 after 2 h)
dissolved in DMF (0.1 mL), methyl iodide (50 lL, 0.8 mmol)
was added, and the solution was stirred for 1 h at ambient
temperature. The removal of methyl iodide and DMF in vacuo
gave the postmodified product 8 as yellow transparent oil (52
1
mg). Mn ¼ 10,200; Mw/Mn ¼ 6.5. H NMR spectroscopy indi-
In a first series of polymerization experiments, monomers 1
and 2 were mixed in a 1:1 molar ratio and polymerized at
room temperature in the presence of 3-HQD as the catalyst
(Table 1, Procedure 1). Using this protocol, polymers were
obtained with a number-average molecular weight (Mn) of
4200 after 2 h of reaction time. Longer reaction times (3 h)
did not result in an increase, but rather in a decrease of
cated 47% conversion of the pyridine residues. The GPC elu-
gram is included in Supporting Information (Fig. S14).
1H NMR (400 MHz, CDCl3): 9.91 (HC¼¼O), 7.75 (pyridine-H),
7.4 (pyridine-H), 7.32 (pyridine-H), 6.25 (¼¼CH2), 6.03
(¼¼CH2), 5.9 (ACHAOH), 5.56 (AOH), 3.95 (AOACH2ACA),
1.39 (ACH2ACH3), 0.79 (ACH2ACH3).
28
JOURNAL OF POLYMER SCIENCE PART A: POLYMER CHEMISTRY 2012, 50, 25–34

WWW.POLYMERCHEMISTRY.ORG
ARTICLE
SCHEME 2 Synthesis of hyperbranched polyesters via Baylis-Hillman polymerization.
molecular weight, pointing toward a possible depolymeriza-
tion, which would be consistent with the potential reversible
character of the Baylis–Hillman polymerization (as long as
catalyst is present). Instead of feeding polymerization reac-
tion with a mixture of both monomers, a second series of
experiments explored the slow addition of 1 to a solution
containing 2. This approach has been frequently explored for
the synthesis of hyperbranched polymers via the A2 þ B3
strategy.19–23 Slow addition of 1 to a DMF solution contain-
ing 2 and 3-HQD indeed afforded polymers with significantly
increased molecular weights, for example 7300 g/mol
instead of 4200 g/mol after 2 h (Table 1, Procedure 2).
Instead of polymerizing 1 and 2 at an equimolar monomer
ratio, the polymerization can also be carried out at
SCHEME 3 Postpolymerization modification of hyperbranched polyesters prepared via Baylis-Hillman polymerization.
WWW.MATERIALSVIEWS.COM
JOURNAL OF POLYMER SCIENCE PART A: POLYMER CHEMISTRY 2012, 50, 25–34
29

ARTICLE
WWW.POLYMERCHEMISTRY.ORG
monomer ratio afforded polymers with slightly lower molec-
ular weights after 1 h and slightly increased molecular
weights after 2 and 3 h as compared to the analogous 1:1 A2
þ B3 polymerization (Table 1, Procedure 3 and Procedure 1).
As was observed above (cf. Procedure 1 and Procedure 2,
Table 1), slow addition of 2 equiv of 1 to a solution of 3
equiv 2 resulted in polymers with increased molecular
weights as compared to the one-pot polymerization at equiv-
alent reaction times (Procedure 4, Table 1). However,
whereas the one-pot procedure afforded a soluble polymer
even after 3 h, gelation was observed after 3 h when 1 is
slowly added to 2.
GPC was used to investigate the kinetics of the hyper-
branched polymerization of 1 and 2. As a typical example,
Figure 1 shows a series of GPC traces that were taken over a
period of 3 h, and which allow to monitor monomer con-
sumption and the evolution of polymer molecular weight
(see also Supporting Information Figs. S15–S20). The results
in Figure 1 were obtained by a one-pot polymerization of 1
and 2 using stoichiometric amounts of acrylate and aldehyde
groups (Procedure 3). With increasing polymerization time,
the GPC traces in Figure 1 broaden and shift to smaller
FIGURE 1 GPC elugrams of samples taken at regular time inter-
vals during the Baylis–Hillman polymerization of 1 and 2 (Pro-
cedure 3).
stoichiometric concentration of acrylate and aldehyde
groups, that is at a 3:2 molar ratio of the A2 and B3 mono-
mers. One-pot polymerization of 2 and 1 at a 3:2 molar
FIGURE 2 13C NMR spectrum (150.9 MHz, CDCl3) of hyperbranched polyester 6 (Procedure 3, Table 1, polymerization time 1 h).
30
JOURNAL OF POLYMER SCIENCE PART A: POLYMER CHEMISTRY 2012, 50, 25–34

WWW.POLYMERCHEMISTRY.ORG
ARTICLE
FIGURE 3 13C NMR spectra of (A) hyperbranched polymer 6 (Procedure 3, after a polymerization time of 1 h), (B) model compound
3, (C) model compound 4, and (D) model compound 5.
elution volumes, which is consistent with the formation of
hyperbranched polymer 6. Furthermore, monitoring the in-
tensity of the peak corresponding to 1 reveals that monomer
consumption is quantitative after 15 min.
tively broad polydispersities (Mw/Mn) at the end of the poly-
merization reaction, which are characteristic of many
hyperbranched polymers.1 Furthermore, the use of triple detec-
tion GPC allowed to determine Mark–Houwink a parameters,
which give information about the solution structure of the poly-
mers. Typical values for a linear, statistical coil in a good sol-
vent are 0.5 < a < 1, whereas for hyperbranched polymers
usually smaller values of 0.2 < a < 0.5 are reported.3,5,19,21
For polymers obtained from 1 and 2, a parameters were found
that ranged from 0.24 to 0.36 (Table 1), indicating a compact
Structural Characterization
In addition to providing insight into polymerization kinetics,
GPC experiments were also used to obtain information about
the molecular architecture of polymers obtained by polymer-
ization of 1 and 2. A first indication comes from the rela-
WWW.MATERIALSVIEWS.COM
JOURNAL OF POLYMER SCIENCE PART A: POLYMER CHEMISTRY 2012, 50, 25–34
31

ARTICLE
WWW.POLYMERCHEMISTRY.ORG
FIGURE 4 1H NMR spectra (400 MHz, CDCl3) of (A) polymer 6 (Procedure 3, after a polymerization time of 1 h), (B) polymer 7, (C)
polymer 8, and (D) polymer 9 (in d6-DMF).
32
JOURNAL OF POLYMER SCIENCE PART A: POLYMER CHEMISTRY 2012, 50, 25–34

WWW.POLYMERCHEMISTRY.ORG
ARTICLE
solution structure as would be expected for a hyperbranched
polymer. No significant dependence of the a parameter either
on the molecular weight of the polymer or on the synthetic
strategy was observed.
Further information about the architecture of hyperbranched
polymer 6 was obtained from NMR experiments. Polymer-
ization of 1 and 2 results in polymers with an irregular,
hyperbranched structure, which are composed of three dif-
ferent structural units, that is linear (L), dendritic (D), and
terminal (T) units (Scheme 2). Estimation of the relative
amounts of these three different structural units allows to
calculate the degree of branching (DB), which provides a
quantitative approach to assess and compare the architec-
tures of hyperbranched polymers produced in different poly-
merization experiments. The relative amounts of D, L, and T
units could be determined using 13C NMR spectroscopy. As
a typical example, Figure 2 shows the full 13C NMR spec-
trum of a hyperbranched polyester prepared from 1 and 2,
whereas Figure 3 represents a magnification of the 6–8 ppm
region, which shows the methyl (ACH3) resonances of the
TMPTA units. The methyl region of the 13C NMR spectrum
shows three major and one smaller resonances. The three
major resonances could be assigned using model compounds
3, 4, and 5, which, respectively, represent the T, L, and D
units. The minor, low field shifted resonance at 7.12 ppm is
tentatively ascribed to result from intramolecular cyclization
or other intermolecular side reactions. Experimental evidence
that would allow an unambiguous assignment of this signal,
however, is lacking at the moment. As this signal is relatively
small compared with the other three resonances, it was not
further taken into account for the determination of the DB.
The use of an inversed gated, proton decoupled 13C NMR pro-
tocol allowed to integrate the different methyl resonances
and subsequently estimate the fractions of T, L, and D units.
From these numbers, the DB was calculated using:24
FIGURE 5 FTIR spectra of (A) polymer 6 (Procedure 3, after a
polymerization time of 1 h), (B) polymer 7, (C) polymer 8, and
(D) polymer 9.
well as the disappearance of the C¼¼C vibration at 1631
cmꢁ1 in the FTIR spectrum (Fig. 5). Reaction of 6 with phe-
nyl isocyanate results in the appearance of aromatic resonan-
1
ces in the H NMR spectrum as well as two new CAH vibra-
tions at 906 and 725 cmꢁ1 in the FTIR spectrum, which
reflect the introduction of the carbamate groups. Comparison
1
of the H NMR integrals of peak ‘‘m’’ (6.9–7.1 ppm, aromatic
protons of the carbarate side chain) with that of peak ‘‘c’’
(tertiary CAH proton) allowed to estimate a 83% hydroxyl
group conversion. Finally, quaternization of the pyridine
groups with methyl iodide afforded a water-soluble hyper-
branched polyester. Quaternization is evident from the signal
at 3.5 ppm in the 1H NMR spectrum of 9 and is also sup-
ported by a new FTIR band at 1629 cmꢁ1, which is due to
the new CAN bond.25 The conversion of pyridine groups
was estimated to 47% based on comparison of the 1H NMR
integrals of signals ‘‘n’’ and ‘‘c.’’
DB ¼ 2D=ð2D þ LÞ
(1)
For the hyperbranched polyesters 6 prepared in this contri-
bution, DBs ranging from 0.36 to 0.81 were determined
(Table 1). Samples that were prepared via slow monomer
addition seemed to show a tendency toward slightly higher
DB values.
CONCLUSIONS
This manuscript has investigated the feasibility of the Baylis–
Hillman reaction for the synthesis of hyperbranched polymers.
Using 2,6-pyridinedicarboxaldehyde and TMPTA as monomers
and 3-HQD as catalyst, hyperbranched polyesters with num-
ber-average molecular weights of up to 7500 g/mol were
obtained via an A2 þ B3 type approach. Carrying out these
polymerizations following a slow monomer addition protocol
resulted in increased molecular weights as compared to
materials that were obtained in a one-pot polymerization.
Also the DB of the polymers prepared via slow monomer
addition seemed to be slightly increased as compared to
polymers obtained via one-pot A2 þ B3 hyperbranched poly-
merization. The Baylis–Hillman reaction is an attractive tool
for the synthesis of functional hyperbranched polyesters as
it generates functional groups that are amenable to further
postpolymerization modification. The hyperbranched polyest-
ers prepared in this contribution contain three orthogonally
Postpolymerization Modification
Hyperbranched polymer 6 contains three chemically orthog-
onal functional groups, viz. hydroxyl, vinyl, and pyridine moi-
eties that are amenable to further functionalization via post-
polymerization modification. As a first proof-of-concept, the
modification of 6 with methyl-3-mercaptopropionate, phenyl
isocyanate, and methyl iodide was investigated (Scheme 3).
These postpolymerization modification reactions were moni-
1
tored with H NMR and FTIR spectroscopy.
Postpolymerization modification of 6 with methyl-3-mercap-
topropionate proceeded with quantitative conversion of the
vinyl groups, as evidenced by the disappearance of the dou-
ble bond resonances in the 1H NMR spectrum (Fig. 4) as
WWW.MATERIALSVIEWS.COM
JOURNAL OF POLYMER SCIENCE PART A: POLYMER CHEMISTRY 2012, 50, 25–34
33

ARTICLE
WWW.POLYMERCHEMISTRY.ORG
10 Parzuchowski, P. G.; Grabowska, M.; Jaroch, M.; Kusznerc-
zuk, M. J Polym Sci Part A: Polym Chem 2009, 47, 3860–3868.
reactive functional groups, viz. hydroxyl, vinyl, and pyridine
moieties, and in first proof-of-concept experiments, the suc-
cessful postpolymerization modification of these groups was
demonstrated. The presence of multiple, orthogonal function
groups make these polyesters of interest, for example as pre-
cursors for dual cure coatings as well as a variety of other
applications.
11 Erber, M.; Boye, S.; Hartmann, T.; Voit, B. I.; Lederer, A.
J Polym Sci Part A: Polym Chem 2009, 47, 5158–5168.
12 Basavaiah, D.; Reddy, B. S.; Badsara, S. S. Chem Rev 2010,
110, 5447–5674.
13 Basavaiah, D.; Rao, A. J.; Satyanarayana, T. Chem Rev 2003,
103, 811–891.
This research was supported by BASF SE. SJ thanks Dr. Quoc
Tuan Nguyen for his help with the GPC analysis.
14 Langer, P. Angew Chem Int Ed 2000, 39, 3049–3052.
15 Venkitasubramanian, P.; Hagberg, E. C.; Bloom, P. D. Polym
Prepr (Am Chem Soc Div Polym Chem) 2008, 49, 914–915.
16 Ji, S. H.; Bruchmann, B.; Klok, H.-A. Macromolecules 2011,
44, 5218–5226.
REFERENCES AND NOTES
17 Yu, C. Z.; Liu, B.; Hu, L. Q. J Org Chem 2001, 66, 5413–5418.
1 Voit, B. I.; Lederer, A. Chem Rev 2009, 109, 5924–5973.
2 Gao, C.; Yan, D. Prog Polym Sci 2004, 29, 183–275.
18 Aggarwal, V. K.; Emme, I.; Fulford, S. Y. J Org Chem 2003,
68, 692–700.
3 Voit, B.
J Polym Sci Part A: Polym Chem 2005, 43,
19 Lin, Q.; Long, T. E. Macromolecules 2003, 36, 9809–9816.
2679–2699.
20 Unal, S.; Oguz, C.; Yilgor, E.; Gallivan, M.; Long, T. E.; Yil-
4 Hult, A.; Johansson, M.; Malmstrom, E. Adv Polym Sci 1999,
gor, I. Polymer 2005, 46, 4533–4543.
143, 1–34.
21 Unal, S.; Lin, Q.; Mourey, T. H.; Long, T. E. Macromolecules
5 McKee, M. G.; Unal, S.; Wilkes, G. L.; Long, T. E. Prog Polym
2005, 38, 3246–3254.
Sci 2005, 30, 507–539.
22 Schmaljohann, D.; Voit, B. Macromol Theory Simul 2003,
6 Fan, Z. R.; Lederer, A.; Voit, B. Polymer 2009, 50, 3431–3439.
12, 679–689.
7 Sato, M.; Nakashima, A.; Sato, Y.; Yamaguchi, I. J Polym Sci
23 Reisch, A.; Komber, H.; Voit, B. Macromolecules 2007, 40,
Part A: Polym Chem 2008, 46, 6688–6702.
6846–6858.
8 Stumbe, J. F.; Bruchmann, B. Macromol Rapid Commun
2004, 25, 921–924.
24 Holter, D.; Burgath, A.; Frey, H. Acta Polym 1997, 48, 30–35.
9 Khalyavina, A.; Schallausky, F.; Komber, H.; Al Samman, M.;
25 Kalbasi, R. J.; Kolahdoozan, M.; Massah, A.; Shahabian, K.
Radke, W.; Lederer, A. Macromolecules 2010, 43, 3268–3276.
Bull Korean Chem Soc 2010, 31, 2618–2626.
34
JOURNAL OF POLYMER SCIENCE PART A: POLYMER CHEMISTRY 2012, 50, 25–34