Photoinduced conjugation of aldehyde functional polymers with olefins via [2 + 2]-cycloaddition

Article abstract of DOI:10.1021/ma2017748

Polymer end group modifications via [2 + 2] Paterno-Buechi reactions were carried out. Polyisobornyl acrylate (PiBoA, 3100 g?mol^-1) and polystyrene (PS, 2800 g?mol^-1) was synthesized by atom transfer radical polymerization (ATRP) emp

Full text of DOI:10.1021/ma2017748

ARTICLE
pubs.acs.org/Macromolecules
Photoinduced Conjugation of Aldehyde Functional Polymers with
Olefins via [2 + 2]-Cycloaddition
Matthias Conradi and Thomas Junkers*
Polymer Reaction Design Group, Institute for Materials Research IMO, Universiteit Hasselt, Agoralaan, Gebouw D,
BE-3590 Diepenbeek, Belgium
S Supporting Information
b
ABSTRACT: Polymer end group modifications via [2 + 2] PaternoÀ
B€uchi reactions were carried out. Polyisobornyl acrylate (PiBoA, 3100
g molÀ1) and polystyrene (PS, 2800 g molÀ1) was synthesized by atom
3
3
transfer radical polymerization (ATRP) employing an aldehyde functional
initiator. Good control over the polymerizations was achieved, and
material with high aldehyde functionality was obtained. The terminal
aldehyde was then reacted with alkenes under UV irradiation in
[2 + 2] cycloaddition reactions. Good yields of the modification reaction was
confirmed via NMR and by electrospray ionization mass spectrometry
(ESIÀMS) analysis and generally conversion of the aldehyde function into
the respective oxetanes of 90% or higher was observed. With the oxetane, a
variety of functional groups were introduced to the polymer chain ranging
from multifunctional allyl-compounds to disubstituted alkenes and amino- or hydroxyl-functional alkenes. In the photoreaction, the
integrity of the bromine end group of the ATRP polymers is retained. ATRP chain extensions can be performed after the
cycloaddition, demonstrating the versatility of this newly introduced polymer modification reaction.
’ INTRODUCTION
achieved by addition of oxidizing agents), the conjugation
reaction itself is in those cases, however, not driven by light.
It is evident from the criteria for efficient click reactions that
cycloadditions (CA) are generally good candidates to perform
conjugation reactions; they proceed in high yields on a single
reaction trajectory and À in most cases À form stable CÀC
bonds. While the thermally allowed [2 + 3] and [2 + 4]
cycloaddition reactions are already largely exploited, almost no
example in the polymer field is so far found for photochemically
allowed [2 + 2] CA reactions.15 Only the team of Barner-
Kowollik recently demonstrated that a [2 + 2]-like reaction
can in principle be performed using the dithioester thiocarbonyl
bond on a polymer derived by the reversible additionÀfragmen-
tation transfer polymerization (RAFT) technique.16 In their
study, these authors demonstrated that conjugation of an ene
can be achieved under UV-irradiation even though the ene was—
due to a rearrangement reaction—not conjugated in the form of
a four-membered ring, but rather inserted into the bond. So far
and to the best of our knowledge, this reaction is the only
example of a polymer modification/conjugation reaction where
the reaction itself is triggered by light and can thus be turned on
and off by will without delay and without use of additional
reagents to start or stop the reaction.
Not only since the introduction of the click concept,1,2
a
significant number of studies has been devoted to polymer con-
jugation and polymer end group modifications. On the basis of
(hetero)telechelic polymers,3 a broad variety of efficient chemis-
tries were developed4À7 that allow for conversion of materials for
an even broader range of applications. The most prominent
reactions in use are the copper catalyzed azid alkyne cycloaddi-
tion (CuAAC),8 thiolÀene and thiolÀyne9 or (hetero) DielsÀ
Alder10 reactions to name the most prominent examples. Most of
these reactions have, however, in common that they are triggered
via thermal activation. Thus, end group conjugations proceed
either upon heating or when the reactive compounds are mixed.
A control over the reaction in the time domain is hence not given.
To date, only few reactions can also be triggered by a light
stimulus. The radical chemistry variant of thiolÀene is based on
conventional radical initiators.11 Thus, these reactions can be also
started via UV-light or proceed under sunlight even when no
dedicated initiator is added via a self-initiation mechanism.12
Another example for a conjugation reaction started by light was
recently introduced by Bowman and co-workers, who used the
CuAAC reaction, but generated the active copper catalyst
in situ.13 Also, Barner-Kowollik showed that the light-induced for-
mation of o-quinodimethanes can be used to generate a poly-
meric enophiles that can undergo a classical DielsÀAlder con-
jugation.14 The above approaches are indeed UV-triggered
reactions and the first even allows for a “switch-off” after the
light was turned off (with the CuAAC method this can only be
Received: August 2, 2011
Revised:
September 17, 2011
Published: September 30, 2011
r
2011 American Chemical Society
7969
dx.doi.org/10.1021/ma2017748 Macromolecules 2011, 44, 7969–7976
|

Macromolecules
ARTICLE
Scheme 1. Mechanism of the PaternoÀBu€chi Reaction
round-bottom flask fitted with a condenser, a magnetic stirrer, a nitrogen
inletÀoutlet, and an addition funnel containing a mixture of 2-bromo-
propinoyl bromide (20 g, 0.096 mol) and 50 mL THF. The flask was
placed into an iceÀwater bath. The solution of 2-bromopropionyl
bromide was added dropwise over a period of 0.5 h under inert con-
ditions. Subsequently the mixture was allowed to reach room tempera-
ture and left stirring overnight. A white precipitate was filtered off. The
solvent was removed and the crude product was dissolved in dichlor-
omethane. The solution was washed three times with 1 M HCl and twice
with water. Finally, the solution was dried with MgSO4, and the solvent
was removed in vacuo. Further purification by column chromatography,
using hexane and ethyl acetate (hex:EtOAc = 5:1 f 3:1) as eluents,
yielded 15.16 g (77.3%) yellowish liquid product.
As a starting point in this research project, the alkene-enone
and the PaternoÀB€uchi reaction seem to be the best choice as
most data is available on these reactions in literature.15,17 Generally,
the reaction can proceed via two distinct pathways that lead to
cyclization. No cyclization can occur from the ground state
because of a phase mismatch of the frontier molecular orbitals.
If the ground state of the activated compound is promoted to the
S1 state via a photon, then cycloaddition is allowed and can occur
in a concerted mechanism. Via intersystem crossing, also a T1
level can be reached, which is then followed by cycloaddition via a
biradical intermediate (see Scheme 1A). Which pathway is
predominant is dependent on substituents, solvent and concen-
tration of the reaction partners. The exact influence of substit-
uents on the regio- and stereoselectivity of the cycloaddition is
hard to predict, but olefins with electron withdrawing substituent
generally proceed via a concerted mechanism, while a donor-
substituent preferentially leads to a radical pathway. The radical
intermediate pathway (B) is most commonly observed, which
leads to less predictable regio- and stereoselectivity; a stable
CÀC bond is however reached in any case. Thus, both reaction
pathways are viable routes toward stable conjugations (C + D).
Many high-yield PaternoÀB€uchi reactions are found in litera-
ture, where ethyl-phenylgloxylate,18 benzaldehyde, acetophe-
none, or benzophenone have been used as the carbonyl group
source to name some prominent examples. Generally, substrates
that are electron rich and stabilized via aromatic rings are good
choices for such reactions. The part of the alkene quencher can
be taken for example by allylic alcohol,19 oxazole,20 methyl vinyl
ketone,21 isobutene,22 1,2-dicyanoethane,23 furan (-derivatives),24
and octene,25 for which yields in the range >70À92% were reported.
In here, we want to report on the conjugation of a variety of
alkene compounds onto aldehyde-functional polymers that were
synthesized via atom transfer radical polymerization (ATRP) by
employing a ATRP initiator bearing a benzaldehyde-type func-
tional group.26 Examples for polystyrene (PS) and polyisobornyl
acrylate (PiBoA) are given and the conjugation products care-
fully analyzed using NMR spectroscopy as well as electrospray
ionization mass spectrometry (ESIÀMS).
1H NMR (300 MHz, CDCl3): δ = 9.96 (s, 1H, CHO), 7.92À7.88
(m, 2H, ortho to CHO), 7.30À7.23 (m, 2H, meta to CHO), 4.61À4.54
(q, 1H, CHBr, J = 6.86 Hz) and 1.92 (d, 3H, CH3, J = 6.86 Hz).
General ATRP Polymerization. The purified CuIBr (0.75 mmol,
108 mg, 1.5 equiv) was added under inert atmosphere into a sealed
Schlenk tube. A mixture of 0.076 mol (16 mL, 150 equiv) of the
monomer iBoA and 1.13 mmol (235 μL, 2.2 equiv) of PMDETA was
purged with nitrogen for 1 h to remove residual oxygen followed by
addition to the reaction flask via a degassed syringe. The EtOAc solution
(25 vol %) was degassed likewise and the reaction mixture was heated up
to 75 °C in an oil bath. Subsequently, the polymerization is started by
adding 0.5 mmol (128 mg, 1 equiv) of degassed initiator. After the
desired reaction time was reached, polymerization was stopped by
cooling in liquid nitrogen and a NMR sample was taken for conversion
determination. The polymer/monomer mixture was dissolved in THF
and the copper catalyst was removed by passing the diluted reaction
mixture over silica. After evaporating of the excess solvent, the polymer
was precipitated into a mixture of ice cold methanol:water (4:1; 10-fold
excess) yielding 5.32 g of polymer with Mn = 3200 g molÀ1 and PDI =
1.16 (by THFÀSEC).
Procedure for [2 + 2]-Cycloaddition. A typical procedure is
given: 55 mg (0.02 mmol, 1 equiv) of aldehyde functional polymer
(Mn = 3200 g molÀ1, PDI = 1.16) was mixed with 140 μL of 1-octene
(0.885 mmol, 50 equiv) and dissolved in 1.25 mL toluene. The flask was
sealed and flushed for 10 min with N2 and later stirred for 48 h at room
temperature while irradiated with UV-light using a Multilamp Reactor MLU
18 from Photochemical Reactor Ltd. The reactor consists of 12 Â 15 W
lamps with a peak emission of 254 nm and a distance of 12 cm to reaction
flask. After reaction the solvent and remaining olefin was removed in vacuo.
Electro Spray Ionization Mass Spectrometry (ESIÀMS).
The Spectra were recorded on an LCQ mass spectrometer (Finigian
MAT) equipped with an atmospheric pressure ionization source oper-
ating in the nebulizer assisted electro spray mode. The instrument was
calibrated in the m/z range 220À2000 using a standard solution con-
taining caffeine, MRFA and Ultramark 1621. A constant spray voltage of
4.5 kV was used and nitrogen at a dimensionless auxiliary gas flow-rate of
10 and a dimensionless sheath gas flow-rate of 60 were applied. The
capillary voltage, the tube lens offset voltage and the capillary tempera-
ture were set to 34 V, 10 V, and 270 °C respectively. A 250 μL aliquot of a
polymer solution with concentration of 1 mg mLÀ1 was injected. A
mixture of dichloromethane and methanol (DCM:MeOH = 1:3), all
HPLC grade, were used as solvent.
THF Size Exclusion Chromatography (THFÀSEC). Analysis
of the MWDs of the polymer samples were performed on a Polymer
Laboratories Spectra Series (P100, AS100, Shodex RI-71), comprising
an autosampler, a PLgel 5.0 μm guard column (50 Â 7.5 mm), followed
by three PLgel 5 μm Mixed-C columns (300 Â 7.5 mm) and a differential
refractive index detector using THF as the eluent at 40 °C with a flow
rate of 1 mL minÀ1. The SEC system was calibrated using liÀne1ar narrow
polystyrene standards ranging from 474 to 7.5 Â 106 g mol (PS (K =
14.1 Â 10À5 dL gÀ1 and α = 0.70), PiBoA (K = 5.0 Â 10À5 dL gÀ1
’ EXPERIMENTAL SECTION
Materials. Isobornyl acrylate (iBoA, Aldrich, tech.), styrene (Sty,
Aldrich, 99%), and methyl methacrylate (MMA, Aldrich, 99%) monomers
were deinhibited over a column of activated basic alumina. Copper(I)
bromide (CuIBr, Sigma-Aldrich, 98%) was washed with acetic acid at
80 °C for 18 h to remove any soluble oxidized species before being
filtered, washed with absolute ethanol, to pH 7, then washed with ethyl
ether, and then dried under vacuum. 4-Hydroxybenzaldehyde (Aldrich,
98%), 2-bromopropanoyl bromide (Aldrich, 97%), triethylamine (TEA,
Sigma-Aldrich, 99%), 1-octene (Aldrich, 98%), 2-methylpentene (Aldrich,
98%), allylamine (Aldrich, 98%), 3-buten-1-ol (Aldrich, 96%), trimethy-
lolpropane allyl ether (Aldrich, 98%), and triallyl cyanurate (Aldrich,
97%) were used as received.
Synthesis of 4-Formylphenyl 2-bromopropanoate (FPBP).
The compound 4-hydroxybenzaldehyde (9.327 g, 0.076 mol), 200 mL of
THF and TEA (9.754 g, 0.096 mol) were placed into a 250 mL three-necked
7970
dx.doi.org/10.1021/ma2017748 |Macromolecules 2011, 44, 7969–7976

Macromolecules
ARTICLE
and α = 0.745) and PMMA (K = 9.44 Â 10À5 dL gÀ1 and α = 0.719))
and toluene as a flow marker.
region 4À5 ppm to the protons close to the oxygen and the
region of 5À6 ppm to the protons adjacent to the aromatic ring
(a zoom into this region of the spectrum can be found in the
Supporting Information). For the formation of the two different
regioisomers (see Scheme 1, parts C + D), indication is given that
the 1,2 disubstituted oxetane is preferentially formed compared
to the 1,3-substituted product. An abundant doublet at 5.91 ppm is
observed, which can only be assigned to the 1,2-substituted
product.27,28 Despite the large number of isomers, it can be
concluded that the oxetane formation out of the aldehyde is
almost quantitative, therefore demonstrating that the PaternoÀ
B€uchi reaction can serve as a conjugation tool for two individual
building blocks.
In order to transfer the results from small organic compounds
to a polymeric system, several design criteria need to be con-
sidered. Generally, high fidelity (hetero) telechelic polymers are
required to ensure a complete conjugation of substrates onto
polymer chains. Since the carbonyl function is the moiety that is
activated by UV light, higher reaction efficiencies are obtained
when the alkene is employed in excess. Hence, for purification
reasons, polymers carrying a terminal aldehyde moiety needed to
be synthesized to allow for an excess of alkene that can be re-
moved after reaction. A facile route to reach such structures is
given via the ATRP process employing an aldehyde-functional
initiator (see Scheme 2). A suitable initiator, 4-formylphenyl-2-
bromopropanoate (FPBP), carrying a benzaldehyde moiety has
been synthesized before by Yagci and his team and no inter-
ference of the aldehyde group with the ATRP process was
observed.29
Nuclear Magnetic Resonance Spectroscopy. 1H NMR spec-
tra were recorded in deuterated chloroform applying a pulse delay of 12 s
with two NMR spectrometer (300 and 400 MHz) from Oxford In-
struments Ltd. using a Varian probe (9 mm-4-nucleus AutoSWPFG).
’ RESULTS AND DISCUSSION
To find suitable reaction conditions, first model studies using
low-molecular weight compounds were carried out. As expected
from theory, best results were obtained with terminal olefins in
presence of benzaldehyde. When activated alkenes were used, e.g.,
an acrylate, homocyclization was observed, however the alkene
showed a high propensity to react with itself. Ketones, as de-
scribed in the literature as suitable reagents in PaternoÀB€uchi
reactions15 underwent the desired reactions, with however very
slow reaction rates. Figure 1 depicts the proton NMR spectrum
of the cycloaddition product from benzaldehyde and 1-octene.
Despite the simple starting materials, a complex spectrum is
obtained, which is due to the large number of stereo- and
regioisomers that are formed in the reaction. As in the depicted
case an excess of octene was used (removed in vacuo after re-
action) an almost complete disappearance of the characteristic
aldehyde peak close to 10 ppm is observed. Even though a clear
assignment is not easily done, the peak region between 2 and
2.5 ppm can be assigned to the protons at the carbon atom
opposite to the oxygen in the four-membered ring, the proton
Two polymers were synthesized employing the FPBP. To
cover different materials, a polystyrene and a polyisobornyl
acrylate was targeted as starting materials for later conjugation
reactions. The molecular weights and polydispersities of the
residual polymers are collated in Table 1. Low conversions were
chosen to prevent chain termination via chain transfer-to-poly-
33
mer, which would result in a lower end group fidelity. The
molecular weight distributions of the polymers are given in the
Supporting Information. In both cases, a successful ATRP could
be carried out. Also MMA was tested with FPBP, however,
significant amounts of dead polymer material was observed re-
peatedly, giving rise to broad molecular weight distributions with
low-molecular weight shoulders. It should be noted that NMR
analysis revealed a relatively high aldehyde functionality, indicat-
ing that the initiator was effective and that by changing the re-
action conditions, i.e., the ligand, well-controlled PMMA should
be obtainable with FPBP.
Figure 1. NMR spectrum from model studies with benzaldehyde and
1-octene (1:50) after 48 h UV irradiation.
Scheme 2. Synthesis of the Aldehyde Functional ATRP Initiator Followed by Polymerization of Styrene
Table 1. Average Molecular Weights and Functional Fidelity of the Aldehyde-Functional Polymers
polymer
Mn/ g molÀ1
Mw/ g molÀ1
PDI
conv, %
Mn,NMR/ g molÀ1
Mn,theo/ g molÀ1
end group funct,a %
3
3
3
3
I
PiBoA
PS
3100
2770
3600
3080
1.16
1.11
13.4
20
2700
3380
2705
2930
97
91
II
a Calculated by comparing peak integrals of the polymer end groups.
7971
dx.doi.org/10.1021/ma2017748 |Macromolecules 2011, 44, 7969–7976

Macromolecules
ARTICLE
Also for the other two polymers, PMMA and PiBoA, high end
group functionalities were obtained. Also, a good match between
expected average molecular weight, measured Mn from SEC
analysis and from NMR analysis was found. Molecular weights
from NMR were deduced via comparison of the characteristic
peak of the proton adjacent to the bromine at 4.1À4.25 ppm.
Functional fidelity was obtained by comparing the peaks of both
end groups. A representative NMR spectrum of the polystyrene
sample II is shown in Figure 2.
Scheme 3 depicts the PaternoÀB€uchi [2 + 2] reaction of the
polymeric substrate. In the scheme, R is a representative for the
different polymer types, PS or PiBoA. In a first step, the polymers
were all conjugated with 1-octene (a). To test for the conjugation
efficiency, trials were carried out with varying contents of a
whereby all other reaction conditions were kept constant, that is
reaction at room temperature for 2 days under constant UV
irradiation in a degassed toluene solution. The yield of the
photoreaction was determined by two different techniques after
isolation of the polymer by precipitation. NMR analysis was
carried out on all polymers to test for complete disappearance of
the aldehyde peak. It should thereby be noted that to stay on the
side of caution, a prolonged pulse delay time (12 s) and a larger
than normal number of scans was applied in the proton peak
acquisition to ensure that a reliable quantitative analysis of the
polymer end group could be performed. An example for a NMR
spectrum of a completely converted aldehyde-functional poly-
mer is also given in the lower part of Figure 2. Integration of the
aldehyde peak region reveals complete disappearance of the
aldehyde signal. At the same time, peaks according to the oxetane
appear, which are, however, not abundant enough to allow for
accurate integration.
To allow for a better assignment of the products, the polymers
obtained from the cycloadditions were subjected to ESIÀMS
mass analysis to further test the conjugation efficiency. The
advantage of a mass spectrometric analysis over the NMR char-
acterization is that both the disappearance of the starting material
and emergence of the products can be traced in higher detail and
with a largely increased accuracy. Also, all of the different stereo-
isomers of the formed oxetane are isobaric and hence appear as
only one apparent species in the mass spectrum. In view of the
present study this is an advantage since the exact configuration of
the conjugated moiety is not of relevance. In the upper part of
Figure 3, a representative section of the full mass spectrum is
depicted. The peak assignment alongside experimental and
theoretical m/z can be found in Table 2. Several peaks are seen
which may practically all be assigned according to the expecta-
tions of an ATRP polymerization. The most abundant peak in
each monomer repeat unit (i.e., 1735.7 m/z) can be assigned to
Figure 2. NMR spectra of polystyrene II before (upper part) and after
cycloaddition with 1-octene at room temperature (II to octene = 1:50).
Figure 3. ESIÀMS spectrum of polyisobornyl acrylate sample I before
(top) and after (bottom) conjugation with 1-octene.
Scheme 3. End Group Modification of the Aldehyde-Terminal Polymers via [2 + 2] Cycloaddition
7972
dx.doi.org/10.1021/ma2017748 |Macromolecules 2011, 44, 7969–7976

Macromolecules
ARTICLE
Table 2. Species Identified in the Mass Spectra As Shown in Figure 3
number
name
formula
experiment/ (m/z)
theory/ (m/z)
δ
Na+ Adducts
K+ Adducts
II
I
Br-M7-oxe
C109H165BrO17Na+
C101H149BrO17Na+
C101H150O17Na+
1847.80
1735.67
1658.73
1848.11
1735.99
1658.08
0.3
0.3
0.6
Br-M7-CHO
H-M7-CHO
V
IV
III
VI
VI
Br-M7-oxe
C109H165BrO17K+
C101H149BrO17K+
C101H150O17K+
C101H149O17K+
1863.80
1751.73
1673.73a
1673.73a
1864.09
1751.96
1674.05
1672.04
0.3
0.2
Br-M7-CHO
H-M7-CHO
d-M7-CHO
Na22+ Adducts
2+
Idc
IIdc
VI
Br-M15-CHO
Br-M15-oxe
C205H309BrO33Na2
C213H325BrO33Na2
2+
1711.93
1767.93
1673.73a
1673.73a
1712.07
1768.14
1673.12
1672.11
0.1
0.2
2+
H-M15-CHO
d-M15-CHO
C205H310O33Na2
2+
VI
C205H309O33Na2
a The monoisotopic peak of the species is overlapped, m/z 1673.73 is the maximum of the full peak.
species I (Br-M7-CHO) (see Table 2), which is a polymer species
consisting of seven monomer units, the aldehyde-functional
initiator on one side and a bromine end group on the other side.
A relatively good match (0.3 Da) between theoretical and ex-
perimental mass is found considering the mass spectrometer type
used and mass range under observation. The peak pattern of the
main peak (and all others) can be adequately simulated when
taking the natural abundances of isotopes into account. Other
peaks in the spectrum represent double-charged products which
can be also assigned to the main ATRP product. Low abundant
single-charged species (III and V) also appear in the spectrum,
which can be assigned to potassium adducts of the main product
(III) and disproportionation products stemming from conven-
tional termination during polymerization (V). Occurrence of
such termination product is not uncommon for ATRP polymer-
izations in the observed amounts.30 Regardless, even though
species V cannot be reactivated in further ATRP polymerization,
it nevertheless carries the desired aldehyde end group and can
hence be functionalized. Satisfyingly, no side products resulting
from transfer to polymer reactions, which are often observed in
acrylate polymerizations, are observed in significant concentra-
tions. Thus, all species present in the mass spectrum can in
principle be conjugated with alkenes and a complete shift of all
peaks should be observable after a successful PaternoÀB€uchi
reaction. Indeed, when inspecting the ESIÀMS spectrum of the
polymer after conjugation, a more or less complete shift is
observed, as indicated in Figure 3. The main product, both the
single charged as well as the double charged species are shifted
according to the mass of octene (112.12 and 56.06 Da for the
double charged chains, respectively). Only few, very low abun-
dant species are seen in the depicted case. Success of the reaction
can thus also be confirmed from the side of the product, where
the NMR only confirmed the conversion of the starting material.
For the analysis of the product mass spectra, it is important to
note that Barner-Kowollik and co-workers, when trying the
cycloaddition of RAFT CdS bonds with alkenes obtained
products where an insertion rather than a cycloaddition had taken
place.16 Such product is isobaric and can, at least in principle, also
contribute to the spectrum in the present case. Via MS/MS
Table 3. Conjugation Efficiencies for Aldehyde Functional
PiBoA I with 1-Octene
[a]/[polymer]
yieldMS/%
yieldNMR/%
1000
100
50
90
90
85
44
32
29
33
25
100
100
100
75
40
30
75
25
63
20
75
15
55
experiments it can, however, be shown that no analogue species
as in the RAFT reaction is present.
Both NMR and ESIÀMS are suitable to determine the yield of
the reaction. A series of experiments was carried out at constant
reaction time with varying amounts of alkene. The results from
these experiments are collated in Table 3. In this table, yields
based on NMR, that is disappearance of aldehyde in the product
spectrum, and yields based on ESIÀMS are given. ESIÀMS
yields were obtained by integration of all single charged species
within a representative monomer repeat unit in the mass spectra
that are unambiguously assignable to a polymeric species (that
are peaks that show a typical isotopic pattern). Such an approach
can be taken for acrylates to obtain relative concentrations of all
polymer species with different end groups.31 As one can see from
the table, NMR predicts higher yields for the reaction that is
observable with ESIÀMS. This may have two reasons:
(i) occurrenceof side products, which are toolow inconcentration
to be detectable in NMR, and (ii) peaks that represent the
desired product but have been degraded (e.g., ring-opened)
during the ESI process are falsely accounted for. Regardless,
the ESI-based yields represent a lower limit and the true yields
may lay between both numbers obtained.
Overall, with decreasing equivalents of alkene present in the
reaction mixture a decreasing yield is observed. Close to quanti-
tative conversion of the aldehyde is only seen when at least
50 equiv are employed at the chosen reaction time of 2 days.
7973
dx.doi.org/10.1021/ma2017748 |Macromolecules 2011, 44, 7969–7976

Macromolecules
ARTICLE
Table 4. Yields of Reaction of Polyisobornyl Acrylate I with
Various Alkenes
alkene
CCdO/CCdC
yieldMS, %
1-octene (a)
1:40
1:50
1:50
1:50
1:50
1:50
95
60
34
88
86
90
allylamine (b)
3-buten-1-ol (c)
trimethylolpropane allyl ether (d)
triallyl cyanurate (e)
2-methylpentene (f)
It should be noted that in order to avoid side-reactions, relatively
low intensity UV light was used and therefore long reactions
times were chosen to compensate. Higher amounts of alkene in
the reaction mixture or higher light intensities will lead to an
increase in reaction rate and further refinement of the exact
conditions will be required.
Figure 4. Molecular weight distributions of the chain-end modified
polyisobornyl acrylate I (full line) and chain extended polymer via
subsequent ATRP (dashed line).
After a relatively good correlation between NMR analysis and
ESIÀMS is established, estimates on the overall quality of the
reaction can be made based on NMR analysis alone. This is
important to allow for an easy screening of reaction conditions.
Also, an even more important, NMR also allows for reaction
control in case where no mass spectra can be obtained. ESIÀMS
is in most cases limited to low molecular weight compounds and
polymers with sufficient polarity in order to ionize. Thus, unpolar
polymers such as polystyrene can only be analyzed via mass
spectrometry with comparatively high effort.32 In the case of the
polystyrene (II) that was prepared via ATRP with the aldehyde
functional polymer, no mass analysis could be carried out. With
identical reaction conditions and 50 equiv of 1-octene, however,
were quantitative yields indicated by NMR, thus allowing for the
conclusion that also for these type of polymers successfully end
group modifications via the [2 + 2] cycloaddition could be
carried out.
In the subsequent step, different alkenes carrying various
functional groups were subjected to PaternoÀB€uchi end group
modifications. Largely successful modifications were achieved in
case of the functional alkenes depicted in Scheme 3 as well as
2-methylpentene to also test a 1,1-disubstituted alkene. In all
cases, complete conversion of the terminal aldehyde was ob-
served by NMR. Yields based on ESIÀMS are given in Table 4.
Satisfying results were obtained in most cases. Somewhat sur-
prisingly, even the disubstituted methylpentene exhibited very
high yields, demonstrating that steric hindrance might not be too
high. It may be assumed that with a disubstituted compound, a
higher regioselectivity was potentially achieved; however, no
further characterization was carried out to test for this hypothesis.
It should be noted again that the MS technique most likely
leads to an underestimation of the true yield of the reaction. For
example, in case of the alcohol c, the expected oxetane product
peak is clearly the most abundant. However, due to unfavorable
signal-to-noise and presence of many small background peaks, a
low yield was numerically deduced. Regardless, yields close to or
higher than 90% are observed (with the exception of b and c),
indicating good success of the end group modifications and
demonstrating that the [2 + 2] photoreaction is relatively
tolerant toward other functional groups. This is a very important
observation, since only with a high functional tolerance, applic-
ability of the method to a large number of polymers consisting of
functional monomers is given. In particular, the cyanurate e is
therefore of high interest, since it introduces two alkenes to the
polymer chain, hence allowing for branch point for further
modification reactions. An interesting effect is seen with the
analysis of the conjugation product with trimethylolpropane allyl
ether d. In the mass spectra, peaks according to an internal
cyclization of the conjugated molecule appear as the most abundant
species in the mass spectrum. Since such cyclization was not
observed when irradiating d alone by UV light (as confirmed by
NMR), it can be concluded that this specific peak is a result of a
condensation reaction occurring during the ESI process. Thus,
both the product peak as well as the cyclized product together
were counted to calculate the yield. All mass spectra and assign-
ment of peaks can be found in the Supporting Information.
Unsatisfying yields are determined for b and c. As noted above,
for c this may be attributed to the appearance of many smaller
product peaks that cannot be assigned unambiguously, thus
reducing the calculated yield of the reaction. For the amine
derivative b, less products are observed. In this case, the desired
product is only low abundant in the spectrum. Therefore, a peak
according to an oxidized version of the expected functional
oxetane is seen. Since the reaction is carried out under inert
conditions, it is therefore assumed that the oxidation to the
substituted hydroxylamine occurs during mass analysis and the
conversion of the reaction was calculated based on this peak. The
overall yield is, however, still low (60%), and the reaction with
the amine must be considered as unsuccessful at this point in
time. It should, however, be stressed again that the presence of
unassigned peaks in the spectrum does not necessarily mean that
the reaction was unsuccessful. In this specific case, side reactions—
before or after cycloaddition—may occur such as oxetane ring-
opening, imine formation, or bromine substitution reactions or
further oxidations during the ESI measurement. Further studies
as well as analysis with higher resolution mass spectrometry may
help to elucidate the exact situation.
The mass spectra described above indicated that the bromine
end group is retained in the end groups modification and is not
disturbed by the UV light. Therefore, it should be possible to
carry out ATRP chain extensions after the photoreaction had
taken place. Therefore, the chain-end modified I was subjected to
further polymerization under ATRP conditions. Figure 4 depicts
the molecular weight distributions before and after the ATRP
7974
dx.doi.org/10.1021/ma2017748 |Macromolecules 2011, 44, 7969–7976

Macromolecules
ARTICLE
reaction. The full line represents I after successful cycloaddition
with 1-octene where all polymer characteristics such as average
molecular weight and polydispersity of the original ATRP
polymer was retained. The dashed line shows the polymer after
chain extension. In the given example, the number-average molec-
’ AUTHOR INFORMATION
Corresponding Author
*E-mail: thomas.junkers@uhasselt.be. Telephone: +32 (11)
268318. Fax: +32 (11) 268399.
ular weight increased from 3100 to 4100 g molÀ1 and the PDI
3
increased from 1.16 to 1.32. While the extended polymer is hence
slightly broader in its distribution than usually achievable in
ATRP block polymerizations, it still shows a complete shift of the
whole distribution toward higher molecular weights, which is an
independent proof for the retention of the active bromine
functionality at the polymer chain end.
’ ACKNOWLEDGMENT
M.C. and T.J. are grateful for financial support via the BOF
funds of the UHasselt and wish to thank J.-P. Noben from the
biomedical research institute for providing access to the ESIÀMS.
’ REFERENCES
(1) Kolb, H. C.; Finn, M. G.; Sharpless, K. B. Angew. Chem., Int. Ed.
2001, 40, 2004–2021.
(2) Barner-Kowollik, C.; Du Prez, F. E.; Espeel, P.; Hawker, C. J.;
Junkers, T.; Schlaad, H.; Van Camp, W. Angew. Chem., Int. Ed. 2011,
50, 60–62.
(3) Tasdelen, M. A.; Kahveci, M. U.; Yagci, Y. Prog. Polym. Sci. 2011,
36, 455–567.
(4) Mansfeld, U.; Pietsch, C.; Hoogenboom, R.; Becer, R.; Schubert,
U. S. Polym. Chem. 2010, 1, 1560–1598.
(5) Barner-Kowollik, C.; Inglis, A. J. Macromol. Chem. Phys. 2009,
210, 987–992.
(6) Lutz, J. F. Angew. Chem., Int. Ed. 2007, 46, 1018–1025.
(7) Iha, R. K.; Wooley, K. L.; Nystrom, A. M.; Burke, D. J.; Kade,
M. J.; Hawker, C. J. Chem. Rev. 2009, 109, 5620–5686.
(8) Lutz, J. F.; Schlaad, H. Polymer 2008, 49, 817–824.
(9) Lowe, A. B. Polym. Chem. 2010, 1, 17–36.
(10) (a) Sinnwell, S.; Inglis, A. J.; Davis, T. P.; Stenzel, M. H.; Barner-
Kowollik, C. Chem. Commun. 2008, 2052–2054. (b) Inglis, A. J.;
Sinnwell, S.; Stenzel, M. H.; Barner-Kowollik, C. Angew. Chem. 2009,
48, 2411–2414. (c) Sinnwell, S.; Synatschke, C. V.; Junkers, T.; Stenzel,
M. H.; Barner-Kowollik, C. Macromolecules 2008, 41, 7904–7912.
(11) Koo, S. P. S.; Stamenoviꢀc, M. M.; Prasath, R. A.; Inglis, A. J.; Du
Prez, F. E.; Barner-Kowollik, C.; Van Camp, W.; Junkers, T. J. Polym. Sci.,
Part A: Polym. Chem. 2010, 48, 1699–1713.
It must be noted that the experiments described herein
resemble a fundamental study to resolve whether the PaternoÀ
B€uchi reaction is suitable as an efficient polymer modification
reaction. Compared to typical click-reactions, the photocycload-
dition may not fulfill the stringent criteria of the click concept2 as
reaction times are generally somewhat too high and also rela-
tively large excesses of the alkene compound need to be em-
ployed in order to achieve complete conversions of the aldehyde
compounds. Nevertheless, the required starting materials are
easily obtained at low cost via the route described and reaction
conditions may still be further refined to allow for faster reactions
and lower excess amounts of alkene. Such investigations are
currently underway in our laboratories. Overall, the PaternoÀ
B€uchi CA has a distinct synthetic potential: A relatively high
tolerance toward other functional groups is achieved and thus
secondary functionalities can be introduced at the chain end
under relatively mild reaction conditions. Combinations of the
CA with other efficient modification reactions or ATRP block
polymerizations can be made, giving rise to interesting synthesis
pathways toward more complex macromolecular architectures. It
should hereby be also noted that the oxetane ring that is formed
in the CA can potentially be ring-opened and hence serve as a
secondary conjugation point in the polymer. Investigations into
this particular feature are also currently underway in our laboratories.
(12) Gress, A.; Volkel, A.; Schlaad, H. Macromolecules 2007,
40, 7928–7933.
(13) Adzima, B. J.; Tao, Y.; Kloxin, C. J.; DeForest, C. A.; Anseth,
K. S.; Bowman, C. N. Nature Chem. 2011, 3, 256–259.
(14) Gruendling, T.; Oehlenschlaeger, K. K.; Frick, E.; Glassner, M.;
Schmid, C.; Barner-Kowollik, C. Macromol. Rapid Commun. 2011, 32,
807–812.
(15) Bach, T. Synthesis 1998, 683–703.
(16) Gruendling, T.; Kaupp, M.; Blinco, J. P.; Barner-Kowollik, C.
Macromolecules 2011, 44, 166–174.
’ CONCLUSION
The successful implementation of UV-induced PaternoÀ
B€uchi cycloaddition of alkenes onto aldehyde-functional poly-
mers was demonstrated. A polyisobornyl acrylate as well as a
polystyrene derivative, both carrying a terminal aldehyde end
group were synthesized utilizing an aldehyde-functional ATRP
initiator. The efficiency of the CA reaction was tested via NMR
and ESIÀMS analysis and for the conjugation of 1-octene,
conversions of 90À100% were obtained after mild UV irradiation
for 2 days at room temperature, when 50 equiv of the alkene were
used. Via this modification reaction, different functional groups were
introduced ranging from multifunctional allyl-compounds to disub-
stituted alkenes and amino or hydroxyl-functional alkenes. In the
photoreaction, the integrity of the bromine end group of the ATRP
polymers is unchanged and subsequent ATRP chain extensions can
be performed after the cycloaddition, demonstrating the versatility
of this newly introduced polymer modification reaction.
(17) B€uchi, G.; Inman, C. G.; Lipinsky, E. S. J. Am. Chem. Soc. 1954,
76, 4327–4331.
(18) Nehrings, A.; Scharf, H.; Runsink J. Angew. Chem. Int. Ed. 1985,
24, 877–878.
(19) Adam, W.; Stegmann, V. R. Synthesis 2001, 2001, 1203–1214.
(20) Griesbeck, A. G.; Bondock, S. Aust. J. Chem. 2008, 61, 573–580.
(21) Du, J.; Yoon, T. P. J. Angew. Chem. Int. Ed. 2009, 131,
14604–14605.
(22) Caldwell, R. A.; Sovocool, G. W.; Gajewski, R. P. J. Angew.
Chem. Int. Ed. 1973, 95, 2549–2557.
(23) Dalton, J. C.; Wriede, P. A.; Turro, N. J. J. Angew. Chem. Int. Ed.
1970, 92, 1318–1326.
(24) Griesbeck, A. G.; Stadtm€uller, S. Chem. Ber. 1990, 123, 357–362.
(25) D’Auria, M.; Racioppi, R. Curr. Org. Chem. 2009, 13, 939–954.
(26) Matyjaszewski, K.; Xia, J. Chem. Rev. 2001, 101, 2921–2990.
(27) Bach, T.; Bergmann, H.; Brummerhop, H.; Lewis, W.; Harms,
K. Chem.—Eur. J. 2001, 20, 4512–4521.
(28) Hammaecher, C.; Portella, C. Chem. Commun. 2008, 5833–5835.
(29) Yurteri, S.; Cianga, I.; Yagci, Y. Macromol. Chem. Phys. 2003,
204, 1771–1783.
’ ASSOCIATED CONTENT
S
Supporting Information. Detailed evaluation of NMR
b
and ESI spectra of the discussed polymer samples and molecular
weight distributions of the starting materials. This material is
available free of charge via the Internet at http://pubs.acs.org/.
7975
dx.doi.org/10.1021/ma2017748 |Macromolecules 2011, 44, 7969–7976

Macromolecules
ARTICLE
(30) Tsarevsky, N. V.; Bencherif, S. A.; Matyjaszewski, K. Macro-
molecules 2007, 40, 4439–4445.
(31) Koo, S. P. S.; Junkers, T.; Barner-Kowollik, C. Macromolecules
2009, 42, 62–69.
(32) Gr€undling, T.; Hart-Smith, G.; Davis, T. P.; Stenzel, M. H.;
Barner-Kowollik, C. Macromolecules 2008, 41, 1966–1971.
(33) Junkers, T.; Barner-Kowollik, C. J. Polym. Sci., Part A: Polym.
Chem. 2008, 46, 7585–7605.
7976
dx.doi.org/10.1021/ma2017748 |Macromolecules 2011, 44, 7969–7976