Free radical polymerization of caffeine-containing methacrylate monomers

Article abstract of DOI:10.1002/pola.27756

The incorporation of acrylic functionality into caffeine enables the preparation of a vast array of novel thermoplastics and thermosets. A two-step derivatization provided a novel caffeine-containing methacrylate monomer capable of free radical polymerization. Copolymers of 2-ethylhexyl methacrylate and caffeine methacrylate (CMA) allowed for a systematic study of the effect of covalently bound caffeine on polymer properties. ¹H NMR and UV-vis spectroscopy confirmed caffeine incorporation at 5 and 13 mol %, and SEC revealed the formation of high molecular weight (co)polymers (>40,000 g/mol). CMA incorporation resulted in a multistep degradation profile with initial mass loss closely correlating to caffeine content. Differential scanning calorimetry, rheological, and thermomechanical analysis demonstrated that relatively low levels of CMA increased the glass transition temperature, resulting in higher moduli and elucidating the benefits of incorporating caffeine into polymers.

Full text of DOI:10.1002/pola.27756

JOURNAL OF
POLYMER SCIENCE
WWW.POLYMERCHEMISTRY.ORG
ARTICLE
Free Radical Polymerization of Caffeine-Containing Methacrylate
Monomers
Ashley M. Nelson, Sean T. Hemp, Jessica Chau, Timothy E. Long
Macromolecules and Interfaces Institute, Department of Chemistry, Virginia Tech, Blacksburg, Virginia 24061
Correspondence to: T. E. Long; (E-mail: telong@vt.edu)
Received 21 May 2015; accepted 17 June 2015; published online 28 July 2015
DOI: 10.1002/pola.27756
ABSTRACT: The incorporation of acrylic functionality into caf-
feine enables the preparation of a vast array of novel thermo-
plastics and thermosets. A two-step derivatization provided a
novel caffeine-containing methacrylate monomer capable of
free radical polymerization. Copolymers of 2-ethylhexyl meth-
acrylate and caffeine methacrylate (CMA) allowed for a system-
atic study of the effect of covalently bound caffeine on
polymer properties. 1H NMR and UV-vis spectroscopy con-
firmed caffeine incorporation at 5 and 13 mol %, and SEC
revealed the formation of high molecular weight (co)polymers
(>40,000 g/mol). CMA incorporation resulted in a multistep
degradation profile with initial mass loss closely correlating to
caffeine content. Differential scanning calorimetry, rheological,
and thermomechanical analysis demonstrated that relatively
low levels of CMA increased the glass transition temperature,
resulting in higher moduli and elucidating the benefits of incor-
C
V
porating caffeine into polymers.
2015 Wiley Periodicals, Inc.
J. Polym. Sci., Part A: Polym. Chem. 2015, 53, 2829–2837
KEYWORDS: photochemistry; renewable resources; structure-
property relations
depicted in Figure 1 below. Encapsulation, delivery, molecu-
larly imprinted polymers (MIPs), and the effect on oxygen
permeability of polyesters represent recent research topics
for caffeine.16–21 For example, Bowyer and coworkers20 syn-
thesized caffeine and theophylline MIPs using both micro-
wave and thermally induced polymerization. The
polymerization technique affected the rebinding affinity;
microwave MIPs exhibited higher dissociation constants than
the thermally induced MIPs. Caffeine nearly doubled the oxy-
gen barrier as an additive in 500-mL poly(ethylene tereph-
thalate) bottles.16 Oxygen transmission rates decreased from
0.046 cc/pkg/day for a bottle without caffeine to 0.0306 and
0.0261 cc/pkg/day with only 3 and 5 wt % caffeine incorpo-
ration respectively. The patent invokes an antiplasticization
effect of caffeine, however, the origin of this dramatic
increase in gas barrier has remained unexplored.
INTRODUCTION Nature plays an inspirational role in polymer
science, providing designs and templates for complex func-
tions and properties as well as offering an abundance of
renewable resources.1,2 Research dealing with the incorpora-
tion of renewable or bio-based compounds into polymeric
materials surged as environmental concerns and efforts
toward sustainability increased.3–5 Consideration of the effect
of synthetic materials on the environment is pervasive in
both academic and industrial laboratories.6,7 One means to
achieve the goals of green chemistry includes the synthesis
of functional materials solely comprised of bioderived com-
pounds, for example, poly(lactic acid).8 Alternatively, intro-
ducing small amounts of renewable compounds into a
polymeric material offers a route to achieve increased sus-
tainability while either attempting to maintain the original
material properties or further tuning properties.9 Monomers
derived from corn feedstock, particularly sugar-based alco-
hols and acids such as 1,4:3,6-dianhydrohexitols and succinic
acid found great success as they readily possess functional
groups for derivatization and polymerization.10–13
Natural polymers, chemicals, and chemical processes exhibit
a high level of structural complexity, bestowing intricate
functions and properties.22 Many strive toward mimicking
natural design when synthesizing polymers, such as the abil-
ity of a gecko to walk on walls, the color-changing capabil-
Caffeine, a well-known stimulant present in coffee and choc-
olate, belongs to a class of molecules referred to as xan-
ities of
a chameleon, and our own biological code—
deoxyribonucleic acid (DNA).23–25 Polymerizable functional
groups or reactive sites are essential for chemical derivatiza-
tion and polymerization. Unlike adenine, caffeine does not
thines.14,15 Other xanthine derivatives with
a
similar
chemical structure include theophylline and theobromine,
Additional Supporting Information may be found in the online version of this article.
C
V
2015 Wiley Periodicals, Inc.
WWW.MATERIALSVIEWS.COM
JOURNAL OF POLYMER SCIENCE, PART A: POLYMER CHEMISTRY 2015, 53, 2829–2837
2829

JOURNAL OF
POLYMER SCIENCE
ARTICLE
WWW.POLYMERCHEMISTRY.ORG
used as received. Ethanol (EtOH) was obtained from Decon
Laboratories, and used as received. Dry dichloromethane
(DCM) was obtained upon treatment with a Innovative Tech-
nology Puresolve solvent system.
Analytical Methods
1H NMR spectroscopy was performed on a Varian Inova or
Agilent U4-DD2 400 MHz spectrometer with samples pre-
pared in deuterated chloroform (CDCl3) or deuterated
dimethyl sulfoxide (DMSOd6). Dynamic light scattering (DLS)
using a Malvern Zetasizer Nano ZS at 25 8C provided the
hydrodynamic diameter of the copolymers in THF. The
(co)polymer solutions were filtered through a 0.45 lm PTFE
syringe filter prior to analysis and the results are an average
of three runs. Size exclusion chromatography (SEC) was per-
formed in THF on a Waters 515 HPLC pump equipped with
a Waters 717plus autosampler. Samples were analyzed at a
concentration of 1 mg/mL in THF through a series of three
Polymer Laboratories PLgel 5 lm MIXED-C columns at a
flow rate of 1 mL/min and results obtained using a Waters
2414 refractive index (RI) detector and Wyatt Technology
Corporation Mini-dawn MALLS detector. Absolute molecular
weight data was obtained using the dn/dc value for poly(-
EHMA) in THF determined offline with a Wyatt Technology
FIGURE 1 Chemical structures of caffeine (with atoms num-
bered for reference), theophylline, and theobromine.
contain a readily reactive functional group such as an amine.
This aromatic heterocycle, which contains hydrogen bond
accepting carbonyls, noncovalently interacts with receptors
and remains unaltered chemically.26 In the early 1970’s, Sal-
omon and coworkers27,28 reported the ability to incorporate
hydroxyl functionality at the 8-position of caffeine, adenine,
adenosine, guanosine, and 2’-deoxyguanosine. Figure
2
depicts the chemical similarity between adenosine, caffeine,
and the nucleobase adenine. More recently, Alami et al.29
alkenylated at the 8-position of caffeine with a Pd/Cu cata-
lyzed coupling reaction. On careful analysis of the literature,
caffeine is present in many publications; however, derivatiz-
ing and covalently incorporating caffeine into a copolymer
remains unexplored.
R
OptilabV T-rEX refractive index detector. Thermal transitions
of the (co)polymers were obtained using a TA Instruments
Q50 TGA and TA Instruments Q1000 DSC affording degrada-
tion temperatures (Td) and glass transition temperatures
(Tg), respectively. A heating rate of 10 8C/min was used for
the TGA and a heat/cool/heat method of 10 8C/min from
room temperature to 100 8C, 10 8C/min from 100 to 280 8C,
and 10 8C/min from 280 to 100 8C was used for the DSC
and thermograms reported from the second heating, ensur-
ing all samples had identical thermal history.
In this manuscript, we describe for the first time the synthe-
sis, polymerization, and physical characterization of caffeine-
containing copolymers. A novel caffeine methacrylate (CMA)
monomer, which is available from a photoreaction to intro-
duce an alcohol and subsequent acid chloride reaction, read-
ily underwent free radical copolymerization with 2-
ethylhexyl methacrylate (EHMA). Poly(EHMA) served as a
control, and incorporation of various amounts of CMA
allowed for an investigation of the effect of caffeine on
acrylic copolymer properties. 1H NMR spectroscopy con-
firmed CMA incorporation, and CMA altered the thermal,
thermomechanical, rheological, and UV-vis absorbance prop-
erties. These copolymers demonstrate the capability and fea-
sibility of incorporating the renewable resource caffeine into
acrylic polymers and elucidate the influence of the heterocy-
clic structure on polymer properties.
Photoreactions
Photoreactions were conducted in an Ace Glass photochemi-
cal safety cabinet using a 24.4475 cm 450 W ultraviolet (UV)
immersion lamp connected to a 120 V, 60 Hz, 450 W UV
power supply. The UV immersion lamp was placed inside a
R
water-cooled condenser which fit inside a PyrexV apparatus
equipped with a N2 inlet and outlet.
Thermomechanical Analysis
EXPERIMENTAL
Dynamic mechanical analysis (DMA) was performed on a TA
Instruments Q800 DMA in tension mode conducted at a fre-
quency of 1 Hz. Samples were equilibrated at 250 8C, held
Materials
R
Caffeine (99%), methacryloyl chloride (MAC; 97%), LuperoxV
DI (di-tert-butyl peroxide, DBP; 98%), and dioxane (spectro-
photometric grade, 991%) were obtained from Sigma
Aldrich and used as received. Triethylamine (TEA; Sigma
Aldrich; ꢀ 99%) was distilled from calcium hydride prior to
use. 2-ethylhexyl methacrylate (EHMA; Sigma Aldrich; 98%)
was eluted through an aluminum oxide column to remove
inhibitor prior to use. 4,4’-Azobis(4-cyanopentanoic acid) (V-
501; Sigma Aldrich; ꢀ98%) was recrystallized from ethanol
prior to use. Chloroform (CHCl3), methanol (MeOH), isopro-
panol (IPA), and tetrahydrofuran (THF; stabilized with
0.025% BHT) were purchased from Spectrum Chemicals and
FIGURE 2 Chemical structures of adenosine, caffeine, and ade-
nine (with atoms numbered for reference).
2830
JOURNAL OF POLYMER SCIENCE, PART A: POLYMER CHEMISTRY 2015, 53, 2829–2837

JOURNAL OF
POLYMER SCIENCE
WWW.POLYMERCHEMISTRY.ORG
ARTICLE
for 5 min, and then heated at a rate of 3 8C/min until flow.
Polymer films were solution-cast from CHCl3 onto a Mylar
sheet, maintained at 20 8C overnight, placed in a vacuum
oven at ca. 50 8C for at least 7.5 h, and finally vacuum was
closed and the film remained at 50 8C for at least 22 h.
224.09. 1H NMR (400 MHz, DMSO-d6, d): 3.90 (s, 3H,
ANACH3), 3.39 (s, 3H, ANACH3), 3.20 (s, 3H, ANACH3),
4.57 (d, 2H, ACH2A), 5.64 (t, 1H, AOH).
8-(1-Hydroxyethyl)21,3,7-Trimethyl-3,7-Dihydro-1H-Purine-2,6-
Dione (EtOH Derivative). HRMS: calcd for C10H14N4O3, 238.11;
found, 238.05. 1H NMR (400 MHz, CDCl3, d): 3.95 (s, 3H,
ANACH3), 3.49 (s, 3H, ANACH3), 3.33 (s, 3H, ANACH3), 4.94
(q, 1H, ACHA), 1.58 (d, 3H, ACH3A), 2.85 (br, AOH).
Ultraviolet-Visible Spectroscopy
A Ocean Optics Inc. USB 2000 spectrometer coupled to a CUV
sample holder providing
a
1
cm pathlength enabled
ultraviolet-visible (UV-Vis) spectroscopy. An Analytical Inst. Sys-
tems Model DT 1000CE UV light source provided a wavelength
range of 190–850 nm. All solutions were prepared in dioxane
and measurements occurred in a quartz cuvette. 1024 M caf-
feine control solutions and 2.6(1025) M caffeine-containing
copolymer solutions (normalized to caffeine content) were
suitable for quantitative absorbance measurements.
8-(2-Hydroxypropan-2-yl)21,3,7-Trimethyl-3,7-Dihydro-1H-Purine-
2,6-Dione (IPA Derivative, Crude). H NMR (400 MHz, CDCl3, d):
4.10 (s, 3H, ANACH3), 3.43 (s, 3H, ANACH3), 3.27 (s, 3H,
ANACH3), 1.62 (s, 6H, A (CH3)2), 2.72 (br, AOH).
1
Synthesis of Caffeine Methacrylate
Methacrylates were synthesized from both the MeOH and
EtOH caffeine derivatives. Starting from the MeOH caffeine
derivative, the following procedure describes the synthesis
of (1,3,7-trimethyl-2,6-dioxo-2,3,6,7-tetrahydro-1H-purin-8-yl)
methyl methacrylate (CMA). Dry DCM was added to the crude
hydroxyl functionalized caffeine reaction mixture. The dis-
solved mixture in DCM was cannulated into a dry round-
bottomed flask equipped with a magnetic stir bar. Using dry
syringe techniques, TEA (0.1 mol equivalent excess compared
to MAC) was added to the reaction and the contents were
stirred in an ice bath. MAC was added to the reaction and the
reaction was allowed to proceed, slowly warming up to room
temperature with stirring overnight. A rotary evaporator
removed excess reagents. Conventional free-radical polymer-
ization was performed on the crude reaction mixture; how-
ever, dissolving the crude mixture in 20:1 CHCl3:MeOH and
eluting with 2:1 CHCl3:acetone on a silica gel column allowed
for CMA isolation. After collecting the appropriate fractions
from the column, rotary evaporation removed the solvent and
the product was dried in a vacuum oven at 50 8C overnight.
Melt Rheology
A
TA Instruments Discovery HR-2 hybrid rheometer
equipped with an environmental test chamber (ETC) pro-
vided time-temperature superposition (TTSP) melt rheology
data. The measurements were performed using an 8-mm
parallel plate steel geometry. An amplitude sweep was con-
ducted first to ensure frequency measurements were per-
formed at a strain within the viscoelastic regime of the
material, and master curves were shifted to a reference tem-
perature (Tref) of 30 8C. Complex viscosity data was also
reported for all samples at 30 8C.
Synthesis of Hydroxyl Functionalized Caffeine
The procedure involved an adaptation of
a previously
reported procedure27 and the MeOH, EtOH, and IPA caffeine
derivatives were all prepared in a similar manner. The pho-
toreactions were conducted in the setup described above.
For example, the synthesis of the MeOH derivative, 8-
(hydroxymethyl)21,3,7-trimethyl-3,7-dihydro-1H-purine-2,6-
dione, follows. 10.0 g of caffeine, 87 mL of reverse osmosis
(RO) water, and 350 mL of MeOH were added to the photo-
reactor glassware with a magnetic stir bar. 25 mL of DBP
was then added to the reaction and a 15 min N2 purge with
stirring was allowed prior to UV irradiation. Upon irradia-
tion, aliquots were extracted to monitor the progress of the
reaction. Excess reagents were removed using a rotary evap-
orator and the crude mixture was dried in a vacuum oven
(ꢁ50 mmHg) at about 90 8C overnight. The crude reaction
mixture was used in the subsequent acid chloride reaction;
however, isolation of the compound was possible using silica
gel chromatography. The crude reaction mixture was dis-
solved in 1:1 CHCl3:MeOH, loaded onto the column and
eluted with a 9:1 CHCl3:MeOH solvent mixture. Prior to run-
ning the column, thin layer chromatography (TLC) was per-
formed to ensure good separation. Using the aforementioned
solvent system, caffeine exhibited an Rf value of 0.46 and the
product exhibited an Rf value of 0.23. After collecting the
product from the column, a rotary evaporator removed the
solvent and the product was dried in a vacuum oven at ca.
40 8C. The isolated product (93% pure via NMR) was a
white solid. HRMS: calcd for C9H12N4O3, 224.09; found,
1
HRMS: calcd for C13H16N4O4, 292.12; found, 292.12. H NMR
(400 MHz, CDCl3, d): 4.03 (s, 3H, ANACH3), 3.57 (s, 3H,
ANACH3), 3.40 (s, 3H, ANACH3), 5.27 (s, 2H, ACH2A), 6.16
(s, 1H, AH), 5.65 (t, 1H, AH), 1.95 (s, 3H, ACACH3).
1-(1,3,7-trimethyl-2,6-dioxo-2,3,6,7-tetrahydro-1H-purin-8-yl)ethyl
methacrylate (from EtOH caffeine derivative). HRMS: calcd for
C14H18N4O4, 306.13; found, 306.13. 1H NMR (400 MHz, CDCl3, d):
4.01 (s, 3H, ANACH3), 3.56 (s, 3H, ANACH3), 3.38 (s, 3H,
ANACH3), 6.02 (q, 1H, ACHA), 1.73 (d, 3H, ACACH3), 6.15 (t,
1H, AH), 5.63 (t, 1H, AH), 1.94 (t, 3H, ACACH3).
Free Radical Copolymerization of (1,3,7-Trimethyl-2,6-
Dioxo-2,3,6,7-Tetrahydro-1H-Purin-8-yl)Methyl
Methacrylate (CMA)
Conventional free radical polymerization afforded poly(ethyl-
hexyl methacrylate) (poly(EHMA)) and caffeine-containing
copolymers. All polymerizations were conducted in dioxane
at 10% solids using 4,40-azobis(4-cyanopentanoic acid) (V-
501) as the free radical initiator (0.67 mol % for EHMA and
1.0 mol % for the copolymerizations). The reagents were
added to a single-neck, round-bottomed, flask equipped with
a magnetic stir bar, sparged with an inert gas, and then
WWW.MATERIALSVIEWS.COM
JOURNAL OF POLYMER SCIENCE, PART A: POLYMER CHEMISTRY 2015, 53, 2829–2837
2831

JOURNAL OF
POLYMER SCIENCE
ARTICLE
WWW.POLYMERCHEMISTRY.ORG
SCHEME 2 Synthesis of caffeine methacrylate (CMA) monomer.
SCHEME 1 Synthesis of hydroxyl-functionalized caffeine using
a peroxide-initiated photoreaction.
ence between the three carbon radical intermediates
(18 < 28 < 38). Murgida et al.30 studied the photoreaction of
caffeine, theophylline, and theobromine (Fig. 1) initiated with
benzophenone. The proposed mechanism included a caffeine
radical intermediate formed upon reaction of caffeine with
the alcohol-based radical. The final product, hydroxyl-
functionalized caffeine, ultimately formed from coupling of
the radical intermediate of caffeine with either the alcohol-
based radical or the benzophenone radical. Two routes pro-
hibit the necessary caffeine radical intermediate: (1) combi-
nation of two alcohol-based radicals, and (2) combination of
the photoinitiator radical with an alcohol-based radical. The
molar excess of alcohol to both caffeine and peroxide in our
reaction (173:1:6, respectively, for the MeOH derivative)
favors alcohol-based radical combination/termination, pre-
sumably eliminating the effect of radical stability on caffeine
intermediate formation, resulting in similar conversion for
each reaction.
allowed to react for approximately 24 h at 70 8C. Polymers
were isolated upon precipitation into MeOH for poly(EHMA)
and poly(EHMA95-co-CMA5), 80:20 MeOH:H2O for poly(-
EHMA87-co-CMA13); all products were filtered and dried in
vacuo at 50 8C overnight.
RESULTS AND DISCUSSION
Synthesis and Copolymerization of Caffeine-Containing
Methacrylate
Caffeine lacks functional groups capable of undergoing poly-
merization, thus chemical modification was necessary to
achieve
a caffeine-based monomer. Previous literature
reported the methodology to introduce hydroxyl functional-
ity at the 8-position of a variety of purine derivatives using a
peroxide-initiated photoreaction between the purine and
alcohol.27 This reaction served as a first step, placing the
reactive hydroxyl group on caffeine allowing for further deri-
vatization to a methacrylate. Scheme 1 depicts the photore-
action and conditions used to synthesize the methanol
caffeine derivative. The bond created in the photoreaction
forms at the carbon directly adjacent to the hydroxyl group,
therefore, different alcohols (methanol, ethanol, and isopro-
panol) afforded caffeine with a primary, secondary, and terti-
ary hydroxyl, respectively.
Placing the hydroxyl group on caffeine provided a reactive
chemical site for further derivatization to a polymerizable
methacrylate. A commonly employed reaction with methacry-
loyl chloride (MAC), as depicted in Scheme 2, provided a caf-
feine methacrylate (CMA) monomer.31,32 This second and final
reaction in the monomer synthesis succeeded on the crude
photoreaction mixtures, eliminating challenging isolation and
purification steps. Supporting Information Figure S1 displays
the 1H NMR spectrum for CMA. 2-ethylhexyl methacrylate
(EHMA), which is a well-studied monomer providing a low
glass transition temperature (Tg), served as a comonomer to
observe the influence of caffeine on thermomechanical prop-
erties.33–35 Conventional free radical copolymerization in diox-
ane of commercially available EHMA, CMA, and V-501 as a
radical initiator, as represented in Scheme 3, generated
copolymers containing covalently bound caffeine. Varying feed
ratios of CMA to EHMA afforded a series of caffeine-
containing copolymers.
Figure 3 depicts the chemical structures and caffeine conver-
sion as a function of exposure time for the three hydroxyl-
functionalized caffeine derivatives. Caffeine conversion pla-
teaued at 30% for each reaction despite the stability differ-
Structural Confirmation and Molecular Weight Analysis
of Poly(EHMAx-co-CMAy)
Copolymerization proceeded on crude CMA to expedite syn-
thesis and eliminate unnecessary solvent and waste from
FIGURE 3 Photoreaction kinetics of caffeine with various alco-
hols as a function of time, determined using 1H NMR spectros-
copy, with primary (methanol), secondary (ethanol), and
tertiary (isopropanol) radicals showing the independence of
radical intermediate stability.
SCHEME 3 Synthesis of caffeine-containing copolymers using
conventional free radical polymerization.
2832
JOURNAL OF POLYMER SCIENCE, PART A: POLYMER CHEMISTRY 2015, 53, 2829–2837

JOURNAL OF
POLYMER SCIENCE
WWW.POLYMERCHEMISTRY.ORG
ARTICLE
FIGURE 4 Structural confirmation of poly(EHMA95-co-CMA5) using 1H NMR spectroscopy. [Color figure can be viewed in the online
issue, which is available at wileyonlinelibrary.com.]
workup procedures. This strategy prevented a determination
of an accurate CMA feed ratio, therefore, NMR after polymer
purification provided actual mol % CMA incorporation. Figure
UV-Vis spectroscopy is a common analytical tool for evaluating
conjugated molecules and polymers.36,37 Researchers utilized
UV-Vis measurements to quantify and detect caffeine for vari-
ous applications including determining the concentration in
coffee and investigating mutagenic activity.38–40 Figure 5 dis-
plays the UV-Vis absorption spectra for caffeine, poly(EHMA95-
co-CMA5), and poly(EHMA87-co-CMA13), confirming successful
caffeine incorporation. Neat caffeine exhibited a maximum
wavelength (kmax) of 276 nm while poly(EHMA95-co-CMA5)
and poly(EHMA87-co-CMA13) revealed similar elevated kmax
values of 282 nm and 283 nm, respectively. Increased conju-
gation enhances the stability of the p* state, which simultane-
ously decreases the energy required for the transition
ultimately increasing the wavelength of absorption.36,37 There-
fore, the additional p bond and resonance present in the
CMA-containing copolymers from the ester could account for
the observed bathochromic shift.
1
4 depicts the H NMR spectra of poly(EHMA95-co-CMA5). The
region below 2 ppm corresponds to the backbone peaks and
pendant alkyls on EHMA; however, integration of the distinct
resonances for the three methyl groups on caffeine and the
methylene adjacent to the ester for each monomer quantified
mol % CMA. Purified control poly(EHMA) and CMA copoly-
mers containing 5 and 13 mol % CMA (using integrations
from resonance a and d in Fig. 4) allowed for an investigation
of the impact of CMA on polymeric properties. Precipitation
revealed an initial solubility difference; poly(EHMA87-co-
CMA13) required an 80:20 MeOH:H2O mixture as opposed to
100% MeOH for poly(EHMA) and poly(EHMA95-co-CMA5).
FIGURE 5 Normalized absorbance spectra for CMA-containing
copolymers, including a caffeine control in dioxane showing a
distinct red shift when caffeine is bound to the polymer
backbone.
FIGURE 6 Light scattering SEC chromatograms of poly(EHMA)
control and caffeine-containing copolymers.
WWW.MATERIALSVIEWS.COM
JOURNAL OF POLYMER SCIENCE, PART A: POLYMER CHEMISTRY 2015, 53, 2829–2837
2833

JOURNAL OF
POLYMER SCIENCE
ARTICLE
WWW.POLYMERCHEMISTRY.ORG
TABLE 1 Molecular Weight and Thermal Degradation of
aggregation peak near 100 nm increased with increasing
CMA content from 7% for poly(EHMA) to 30 and 49% for
poly(EHMA95-co-CMA5) and poly(EHMA87-co-CMA13), respec-
tively. This aggregation could account for the observed
shoulder in the poly(EHMA87-co-CMA13) SEC trace. Support-
ing Information Figure S2 displays the hydrodynamic diame-
ters for the (co)polymers as a function of mean volume %
(a) and mean intensity % (b).
Caffeine-Containing Copolymers
Mol % Wt %
Mw
Td1,onset Wt. Lossd1 Tdfinal,onset
CMAa CMA (Kg/mol)b PDIb
(8C)c
(%)c
(8C)c
0
0
7
99.2
78.5
40.1
2.13
1.89
2.12
269
195
194
N.D.
10
N.D.
313
315
5
13
18
13
Thermal, Thermomechanical, and Rheological
a
Calculated from 1H NMR.
b
Characterization of Caffeine-Containing Copolymers
Thermogravimetric analysis (TGA) provided weight loss as a
function of temperature in a N2 atmosphere for the (co)poly-
mer series. Poly(EHMA) displayed a one-step degradation
occurring at 228 8C while both poly(EHMA95-co-CMA5) and
Absolute molecular weight data obtained from THF SEC at 40 8C using
dn/dc 5 0.088.
c
TA Instruments Q50 thermogravimetric analyzer (TGA) heating at 10
8C/min
N.D. not detected
poly(EHMA87-co-CMA13) displayed multistep degradation
Size exclusion chromatography (SEC) in THF provided abso-
profiles. Figure 7 depicts the decomposition curves and
Table 1 specifies the degradation temperatures (Td) and cor-
responding % weight loss. CMA-containing copolymers dis-
played an initial degradation lower than poly(EHMA),
however, the largest degradation step of the copolymers
increased ꢁ50 8C compared to poly(EHMA). The final degra-
dation temperature increased from 269 8C for poly(EHMA)
to 313 and 315 8C for poly(EHMA95-co-CMA5) and poly(-
EHMA87-co-CMA13), respectively. Additionally, the absence of
early weight loss suggested the absence of volatile by-
products.
lute molecular weight analysis of poly(EHMA), poly(EHMA95
-
co-CMA5), and poly(EHMA87-co-CMA13). Figure 6 depicts the
light scattering (LS) traces and Table 1 displays the corre-
sponding weight-average molecular weight (Mw) and polydis-
persity index (PDI) for each copolymer. Offline determination
of the refractive index increment, dn/dc, for poly(EHMA)
afforded absolute molecular weight. Using the dn/dc value
addressed the chemical dissimilarity between the CMA-
containing series and the polystyrene chromatography col-
umns; however, it is important to emphasize that the poly(-
EHMA) dn/dc generated the molecular weight data for all
three (co)polymers in the series. A separate determination of
the dn/dc value for each copolymer composition would mini-
mize the effect of compositional differences on the results.
Interestingly, the initial wt. loss in poly(EHMA95-co-CMA5)
and poly(EHMA87-co-CMA13) closely correlated to the loss of
pendant caffeine. Beta hydrogen elimination is a commonly
reported degradation route for methacrylates with pendant
side chains or groups.42–44 CMA, however, does not contain
any beta hydrogens eliminating the common route as a pos-
sible mechanism for caffeine loss. Fodor et al.45 extensively
researched the thermal decomposition of poly(N-vinylimida-
zole) using pyrolysis-gas chromatography/mass spectrometry
(Py-GC/MS). This highly sensitive analytical technique identi-
fied 10 different small molecule byproducts, one of them
The three LS traces in Figure 6 show poly(EHMA) eluting
first corresponding to the highest Mw of 99.2 kg/mol (Table
1). Although lower, both poly(EHMA95-co-CMA5) and poly(-
EHMA87-co-CMA13) exhibited suitable Mws for conventional
free radical polymerization of 78.5 kg/mol and 40.1 kg/mol,
respectively. A higher concentration of V-501 initiated the 5
and 13 mol % CMA copolymers compared to the poly(-
EHMA) control. Since polymerization rate scales according to
the square root of initiator concentration, the increased V-
501 concentration lowers the final attainable polymer molec-
ular weight and presumably contributed to the observed
decreased Mw.41 Poly(EHMA) and poly(EHMA95-co-CMA5)
displayed monomodal LS traces, whereas poly(EHMA87-co-
CMA13) contained a high molecular weight shoulder and
slight low molecular weight shoulder. This could originate
from the isolation procedure and/or poor THF dissolution
during chromatography for this particular copolymer compo-
sition. Dynamic light scattering (DLS) in THF of the (co)poly-
mers
showed
monomodal
hydrodynamic
diameter
distributions (99.9 % of mean volume) of 12 6 5 nm,
9 6 3 nm, and 8 6 2 nm for poly(EHMA), poly(EHMA95-co-
CMA5), and poly(EHMA87-co-CMA13), respectively. These
results provided evidence for good solubility in THF to
ensure reliable SEC data. In lieu of the high molecular weight
shoulder present in poly(EHMA87-co-CMA13), DLS also speci-
fied hydrodynamic diameter as a function of intensity. An
FIGURE 7 Two-step thermal degradation profile of caffeine-
containing copolymers.
2834
JOURNAL OF POLYMER SCIENCE, PART A: POLYMER CHEMISTRY 2015, 53, 2829–2837

JOURNAL OF
POLYMER SCIENCE
WWW.POLYMERCHEMISTRY.ORG
ARTICLE
FIGURE 10 Melt rheology exhibits higher moduli with increas-
ing CMA content. [Color figure can be viewed in the online
issue, which is available at wileyonlinelibrary.com.]
FIGURE 8 Proposed thermal degradation route of caffeine-
containing copolymers resulting in the loss of free caffeine.
(Reproduced from Ref. 51 with permission from American
Chemical Society).
from the plateau of the step-wise Tg shifting to the right for
poly(EHMA95-co-CMA5) and higher for poly(EHMA87-co-
CMA13). The lack of any endothermic transitions indicated
amorphous (co)polymers, and a single Tg suggested random
copolymerization of the two methacrylates without
microphase-separation. Placing hydrogen bonding sites pend-
ant to the polymer backbone, Zhang et al.46 observed highly
ordered lamellar morphologies. Cheng et al.47 also observed
H-bonding driven self-assembly upon synthesizing poly(acry-
lates) containing 7 mol % of an adenine-containing acrylate.
Surprisingly, polymerizing a thymine-containing acrylate in
concentrations up to 30 mol % incorporation did not exhibit
any self-assembly. Caffeine does not hydrogen bond with
itself, however, the aromatic structure is prone to p-p stack-
ing.48–50 It is proposed that placing the caffeine pendant
group further from the backbone will increase mobility, ena-
bling the aromatic groups to p-p stack and potentially result-
ing in a distinct caffeine-rich phase. Additionally, increasing
being imidazole. Figure 8 depicts the proposed mechanism
(left) constructed to support imidazole as a decomposition
product. As discussed above, the photoreaction in the first
step of CMA synthesis occurs through a radical intermediate
at the same position on caffeine in which the proposed deg-
radation mechanism occurs. With the propensity of caffeine
to lose a hydrogen at the same attachment point, the right
side of Figure 8 proposes a mechanism of side group elimi-
nation and subsequent hydrogen abstraction to explain the
observed caffeine loss.
Differential scanning calorimetry (DSC) revealed a broad
glass transition temperature for poly(EHMA), poly(EHMA95
-
co-CMA5) and poly(EHMA87-co-CMA13), visible in Figure 9.
Although difficult to assign an exact value due to the breadth
of the transition, the observed Tg of poly(EHMA) corre-
sponded well with the literature value of ꢁ210 8C.33 Incor-
porating caffeine increased the Tg of the polymers, visible
FIGURE 9 DSC shows broad Tg’s for the copolymers, trending
FIGURE 11 Complex viscosity of poly(EHMA) and caffeine-
containing copolymers at 30 8C.
upward with increasing CMA content.
WWW.MATERIALSVIEWS.COM
JOURNAL OF POLYMER SCIENCE, PART A: POLYMER CHEMISTRY 2015, 53, 2829–2837
2835

JOURNAL OF
POLYMER SCIENCE
ARTICLE
WWW.POLYMERCHEMISTRY.ORG
exhibited a slightly higher Tg than poly(EHMA), with a tan
delta maximum occurring at 55 and 49 8C, respectively. Addi-
tionally, a higher storage modulus with CMA was consistent.
CONCLUSIONS
The successful synthesis and polymerization of a novel
caffeine-containing methacrylate afforded copolymers contain-
ing covalently bound caffeine pendant groups. A free radical
induced photoreaction and acid chloride reaction generated
CMA requiring minimal isolation prior to polymerization.
Using 2-ethylhexyl methacrylate as a comonomer, film-forming
polymers containing 0, 5, and 13 mol % CMA provided insight
into the effect of caffeine on physical properties. Caffeine
influenced both thermal stability and Tg, showing a unique
multistep thermal degradation profile similar to sidegroup
elimination as observed for poly(N-vinylimidazole).45 1H NMR
confirmed caffeine content, and UV-vis spectroscopy exhibited
a characteristic absorbance for caffeine, shifted to higher
wavelengths attributed to the extra p bond and resonance in
CMA-containing copolymers. Finally, caffeine improved poly-
mer moduli suggesting the aromatic structure stiffened poly-
mer chains. These novel materials demonstrated the effect of
the naturally occurring caffeine structure on polymer proper-
ties, including both thermal and thermomechanical influences.
The higher Tg and storage moduli of low CMA contents (<
15 mol %) suggested that CMA could perform as a comono-
mer in acrylic polymers to impart specific physical properties
and increase sustainability. Furthermore, future placement of
caffeine further from the polymer backbone may encourage
noncovalent p-p stacking and expand the library of caffeine-
containing polymers, promoting microphase-separation for the
synthesis of elastomers and broadening the understanding
and potential of caffeine in material applications.
FIGURE 12 Thermomechanical analysis of 5 mol % CMA copol-
ymer, showing an increased and extended glassy modulus
compared to poly(EHMA) control.
CMA content will promote p-p stacking through a proximity
effect.
Time-temperature superposition (TTSP) melt rheology and
dynamic mechanical analysis (DMA) further investigated the
effect of caffeine on thermal properties. Figure 10 displays
the master curves, obtained using a 30 8C reference tempera-
ture (Tref), for the (co)polymer series. All polymers adhered
to TTSP, which provided a broad range of rheological behav-
ior over angular frequencies of 1026 to ꢁ104 rad/s. The
crossover of the storage (G’) and loss (G”) moduli (G’ 5 G”)
occurred in each material and shifted to lower frequencies
with increasing CMA.51 Modulus also increased with increas-
ing caffeine content. Contrary to the SEC results discussed
previously where poly(EHMA) exhibited the highest Mn, a
shift of the crossover point to lower frequencies often indi-
cates higher molecular weight.52 The shifted crossover point
combined with the enhanced modulus suggested caffeine
pendant groups increased chain stiffness. Similarly complex
viscosity, depicted in Figure 11, followed an increasing trend
with CMA incorporation. Complex viscosity often scales with
molecular weight, thus these results further supported that
pendant caffeine functional groups largely influenced chain
stiffness.
ACKNOWLEDGMENTS
The authors would like to thank the U.S. Army Research Labora-
tory and the U.S. Army Research Office under the Army Materi-
als Center of Excellence Program, contract W911NF-06-2-0014
for financial support. The authors would like to thank Michael
H. Allen, Jr. and Richard N. Carmean for insightful discussions
and assistance. The authors would also like to thank Dr. David
Inglefield and Mingtao Chen for the offline determination of
poly(EHMA) dn/dc in THF.
All three (co)polymers formed films with solution casting
from chloroform. Figure 12 shows the DMA curves of poly(-
EHMA) and poly(EHMA95-co-CMA5) obtained from their
respective films. Poly(EHMA87-co-CMA13) formed a film,
REFERENCES AND NOTES
however, films were glassy at room temperature, unlike the
other two samples, and shattered while attempting to obtain
a rectangular piece for thermomechanical testing. The stiffer,
glassy, nature of poly(EHMA87-co-CMA13) correlated well to
TTSP and viscosity data, demonstrating the influence of the
caffeine pendant group on properties. Poly(EHMA) and pol-
y(EHMA95-co-CMA5) exhibited DMA traces expected of amor-
phous materials; beginning with a glassy modulus at low
temperatures and proceeding through a single transition (Tg)
leading to flow. As observed in DSC, poly(EHMA95-co-CMA5)
1 A. E. Barron, R. N. Zuckerman, Curr. Opin. Chem. Biol. 1999,
3, 681–687.
2 M. N. Belgacem, A. Gandini, Monomers, Polymers and Com-
posites from Renewable Resources; Elsevier: Oxford, UK, 2008.
3 M. A. R. Meier, J. O. Metzger, U. S. Schubert, Chem. Soc.
Rev. 2007, 36, 1788–1802.
4 A. K. Mohanty, M. Misra, L. T. Drzal, J. Polym. Environ. 2002,
10, 19–26.
5 A. J. Ragauskas, C. K. Williams, B. H. Davison, G. Britovsek,
J. Cairney, C. A. Eckert, W. J. Frederick, J. P. Hallett, D. J. Leak,
2836
JOURNAL OF POLYMER SCIENCE, PART A: POLYMER CHEMISTRY 2015, 53, 2829–2837

JOURNAL OF
POLYMER SCIENCE
WWW.POLYMERCHEMISTRY.ORG
ARTICLE
C. L. Liotta, J. R. Mielenz, R. Murphy, R. Templer, T.
Tschaplinski, Science 2006, 311, 484–489.
32 M. Rikkou-Kalourkoti, P. A. Panteli, C. S. Patrickios, Polym.
Chem. 2014, 5, 4339–4347.
6 A. Gandini, T. M. Lacerda, Prog. Polym. Sci., in press.
33 C. L. Elkins, T. Park, M. G. McKee, T. E. Long, J. Polym. Sci.,
Part A: Polym. Chem. 2005, 43, 4618–4631.
7 C. A. Koch, Ed.; AVery Dennison Corporation: US. U. S. Pat-
ent 8,796,351 B2, 2014.
34 M. Sierant, S. Kazmierski, A. Rozanski, P. Paluch, U. Bienias,
B. J. Miksa, N. J. Chem. 2015, 39, 1506–1516.
8 K. Madhavan Nampoothiri, N. R. Nair, R. P. John, Bioresour.
Technol. 2010, 101, 8493–8501.
35 F. Fleischhaker, A. P. Haehnel, A. M. Misske, M. Blanchot, S.
Haremza, C. Barner-Kowollik, Macromol. Chem. Phys. 2014,
215, 1192–1200.
9 L. Yu, K. Dean, L. Li, Prog. Polym. Sci. 2006, 31, 576–602.
10 F. Fenouillot, A. Rousseau, G. Colomines, R. Saint-Loup, J.
P. Pascault, Prog. Polym. Sci. 2010, 35, 578–622.
36 P. Y. Bruice, Organic Chemistry, 7th ed.; Pearson Education:
Upper Saddle River, NJ, 2014.
11 H. Song, S. Y. Lee, Enz. Microbial Technol. 2006, 39, 352–
361.
37 D. A. Skoog, F. J. Holler, S. R. Crouch, Instrumental Analy-
sis; Brooks/Cole, 2007.
12 S. Chatti, G. Schwarz, H. R. Kricheldorf, Macromolecules
2006, 39, 9064–9070.
38 A. Belay, K. Ture, M. Redi, A. Asfaw, Food Chem. 2008, 108,
310–315.
13 Q. Lin, J. Pasatta, T. E. Long, J. Polym. Sci., Part A: Polym.
Chem. 2003, 41, 2512–2520.
ꢁ
ꢁ
39 L. Lopez-Martıꢁnez, P. L. Lopez-de-Alba, R. Garcıꢁa-Campos, L.
ꢁ
M. De Leon-Rodrıꢁguez, Anal. Chim. Acta 2003, 493, 83–94.
14 E. Looser, T. W. Baumann, H. Wanner, Phytochemistry
1974, 13, 2515–2518.
ꢁ
40 A. Woziwodzka, G. Gołunski, D. Wyrzykowski, R.
15 R. Matissek, Z Lebensm Unters Forsch 1997, 205, 175–184.
ꢁ
Kazmierkiewicz, J. Piosik, Chem. Res. Toxicol. 2013, 26, 1660–
1673.
16 Y. Shi, F. M. Schloss, D. A. Weemes, PCT, Eds.; The Coca-
Cola Company: USA. Eastman Chemical Company: USA,
2007.
41 G. Odian, Principles of Polymerization, 4th Ed.; Wiley: Hobo-
ken, NJ, 2004.
17 L. Paseta, G. Potier, S. Abbott, J. Coronas, Org. Biomol.
Chem. 2015, 13, 1724–1731.
42 S. D. Smith, T. E. Long, J. E. McGrath, J. Polym. Sci., Part
A: Polym. Chem. 1994, 32, 1747–1753.
ꢀ
18 A. Belscˇak-Cvitanovic´, D. Komes, S. Karlovic´, S. Djakovic´, I.
ꢁ
43 J. M. Cervantes-Uc, J. V. Cauich-Rodrıguez, W. A. Herrera-
ꢀ
ꢀ
ꢀ
Spoljaric´, G. Mrsic´, D. Jezek, Food Chem. 2015, 167, 378–386.
ꢁ
ꢁ
Kao, H. Vazquez-Torres, A. Marcos-Fernandez, Polym. Degrad.
19 E. G. Kuznetsova, O. M. Kuryleva, L. A. Salomatina, V. I.
Sevast’yanov, Biomed. Eng. 2008, 42, 141–144.
Stab. 2008, 93, 1891–1900.
44 S. Zulfiqar, M. Zafar-uz-Zaman, A. Munir, I. C. McNeill,
Polym. Degrad. Stab. 1997, 55, 89–94.
20 N. W. Turner, C. I. Holdsworth, S. W. Donne, A. McCluskey,
M. C. Bowyer, N. J. Chem. 2010, 34, 686–692.
ꢁ
ꢁ
45 C. Fodor, J. Bozi, M. Blazso, B. Ivan, Macromolecules 2012,
21 H. Zayas, C. I. Holdsworth, M. C. Bowyer, A. McCluskey,
Org. Biomol. Chem. 2014, 12, 6994–7003.
45, 8953–8960.
46 K. Zhang, G. B. Fahs, M. Aiba, R. B. Moore, T. E. Long,
Chem. Commun. 2014, 50, 9145–9148.
22 R. S. Trask, H. R. Williams, I. P. Bond, Bioinspirat. Biomimet.
2007, 2, P1–P9.
47 S. Cheng, M. Zhang, N. Dixit, R. B. Moore, T. E. Long, Mac-
romolecules 2012, 45, 805–812.
23 N. Huebsch, D. J. Mooney, Nature 2009, 462, 426–432.
24 S. S. Hardaker, R. V. Gregory, MRS Bull. 2003, 28, 564–567.
25 S. T. Hemp, T. E. Long, Macromol. Biosci. 2012, 12, 29.
48 G. M. Hill, D. M. Moriarity, W. N. Setzer, Scientia Pharma-
ceut. 2011, 79, 729–747.
26 S. Goswami, A. K. Mahapatra, R. Mukherjee, J. Chem. Soc.
Perkin Trans. 1 2001, 2717–2726.
49 V. V. Kostjukov, N. M. Khomytova, A. A. Hernandez
Santiago, A.-M. C. Tavera, J. S. Alvarado, M. P. Evstigneev, J.
Chem. Thermodyn. 2011, 43, 1424–1434.
27 J. Salomon, D. Elad, J. Org. Chem. 1973, 38, 3420–3421.
28 D. Elad, H. Steinmaus, I. Rosenthal, J. Org. Chem. 1971, 36,
3594–3598.
50 H.-M. Zhang, T.-T. Chen, Q.-H. Zhou, Y.-Q. Wang, J. Mol.
Struct. 2009, 938, 221–228.
29 S. Sahnoun, S. Messaoudi, J.-D. Brion, M. Alami, Eur. J.
Org. Chem. 2010, 2010, 6097–6102.
51 A. Y. Malkin, A. I. Isayev, Rheology—Concepts, Methods,
and Applications, 2nd Ed.; ChemTec Publishing: Toronto, Can-
ada, 2012.
ꢁ
30 D. H. Murgida, P. F. Aramendıa, R. Erra-Balsells, Photochem.
Photobiol. 1998, 67, 487–494.
52 T. G. Mezger, The Rheology Handbook: For Users of Rota-
tional and Oscillatory Rheometers; Vincentz Network: Hann-
over, Germany, 2006.
31 T. Saito, B. D. Mather, P. J. Costanzo, F. L. Beyer, T. E.
Long, Macromolecules 2008, 41, 3503–3512.
WWW.MATERIALSVIEWS.COM
JOURNAL OF POLYMER SCIENCE, PART A: POLYMER CHEMISTRY 2015, 53, 2829–2837
2837