Azo-aromatic functionalized polyethylene by nitroxide radical coupling (NRC) reaction: Preparation and photo-physical properties

Article abstract of DOI:10.1016/j.polymer.2015.11.018

The nitroxide radical coupling (NRC) reaction between the 4-(phenylazo)-benzoyl-2,2,6,6-tetramethylpiperidine-1-oxyl radical (AzO-TEMPO) or the 4-(2-thienylazo)-benzoyl-2,2,6,6-tetramethylpiperidine-1-oxyl radical (ThiO-TEMPO) and polyethylene macroradicals allowed the preparation of "functional" polyolefins bearing covalently grafted azo-aromatic chromophores. A comparison of the photo-physical behavior of the free and grafted RO-TEMPO molecules was carried out by UV-Vis spectroscopy irradiating the free RO-TEMPO solutions and the functionalized polymer films at 366 and 254 nm. Results evidenced the transfer of the photo-physical properties of the chromophores to the polymer matrix. Interestingly, some different isomerization abilities and kinetics between the free and grafted RO-TEMPO moieties, depending on the nature of the aromatic group bonded to the diazo-moiety and on the structure of the polymer matrix, were observed. Indeed, when the chromophores were grafted to the polymer, the isomerization resulted to be less efficient in terms of photo-isomerization degree and isomerization rate than in solution, suggesting a strong effect of the semi-crystalline matrix especially in the case of the more rigid HDPE. Finally the determination of the water contact angle of the functionalized polymers, before and after photo-isomerization, confirmed the occurrence of the isomerization and evidenced an increase of the wettability of the polymer surface owing to the process.

Full text of DOI:10.1016/j.polymer.2015.11.018

Polymer 82 (2016) 366e377
Contents lists available at ScienceDirect
Polymer
journal homepage: www.elsevier.com/locate/polymer
Azo-aromatic functionalized polyethylene by nitroxide radical
coupling (NRC) reaction: Preparation and photo-physical properties
,
,
, c
Francesca Cicogna a *, Ilaria Domenichelli a b, Serena Coiai a, Fabio Bellina a
,
Marco Lessi c, Roberto Spiniello a, Elisa Passaglia a
a Istituto di Chimica dei Composti OrganoMetallici (ICCOM), Consiglio Nazionale delle Ricerche, SS Pisa, Via G. Moruzzi 1, 56124 Pisa, Italy
b Scuola Normale Superiore, Piazza dei Cavalieri 7, 56126 Pisa, Italy
c
ꢀ
Dipartimento di Chimica e Chimica Industriale, Universita degli Studi di Pisa, Via G. Moruzzi 13, 56124 Pisa, Italy
a r t i c l e i n f o
a b s t r a c t
Article history:
The nitroxide radical coupling (NRC) reaction between the 4-(phenylazo)-benzoyl-2,2,6,6-
tetramethylpiperidine-1-oxyl radical (AzO-TEMPO) or the 4-(2-thienylazo)-benzoyl-2,2,6,6-
tetramethylpiperidine-1-oxyl radical (ThiO-TEMPO) and polyethylene macroradicals allowed the prep-
aration of “functional” polyolefins bearing covalently grafted azo-aromatic chromophores. A comparison
of the photo-physical behavior of the free and grafted RO-TEMPO molecules was carried out by UV-Vis
spectroscopy irradiating the free RO-TEMPO solutions and the functionalized polymer films at 366 and
254 nm. Results evidenced the transfer of the photo-physical properties of the chromophores to the
polymer matrix. Interestingly, some different isomerization abilities and kinetics between the free and
grafted RO-TEMPO moieties, depending on the nature of the aromatic group bonded to the diazo-moiety
and on the structure of the polymer matrix, were observed. Indeed, when the chromophores were
grafted to the polymer, the isomerization resulted to be less efficient in terms of photo-isomerization
degree and isomerization rate than in solution, suggesting a strong effect of the semi-crystalline ma-
trix especially in the case of the more rigid HDPE. Finally the determination of the water contact angle of
the functionalized polymers, before and after photo-isomerization, confirmed the occurrence of the
isomerization and evidenced an increase of the wettability of the polymer surface owing to the process.
© 2015 Elsevier Ltd. All rights reserved.
Received 22 July 2015
Received in revised form
3 November 2015
Accepted 7 November 2015
Available online 2 December 2015
Keywords:
Nitroxide radical coupling reaction
Radical functionalization of polyolefins
Azobenzene derivatives
1. Introduction
“functional” materials can be prepared by the copolymerization of
a-olefins with functional unsaturated monomers. The polymeri-
Functional polymeric materials bearing covalently linked chro-
mophores, either as side chain or in the backbone, can find different
applications depending on the nature of the chromophore, the
polymer matrix and on the amount of functional groups per chain
[1e11]. The feasibility in providing these materials is subjected to
the synthetic procedure that depends on the nature and on the
specific reactivity of the starting chemicals. Generally, “functional”
polymers can be prepared by polymerization or copolymerization
of “functional” monomers and/or by post reactor modification of a
pre-formed matrices [12e14]. However, all processes involving
radical reactions are generally avoided unless they are carried out
in a controllable way to limit side reactions [14]. In the specific case
of polyolefins (POs), the polymer matrices chosen for this paper,
zation can proceed either by a free radical mechanism or by a
catalytic route, but both procedures suffer from the presence of
functional polar groups that, interfering with the radical mecha-
nism or with the catalyst, can negatively affect the process effi-
ciency. Two possible alternatives to prepare functional POs bearing
chromophores starting from the monomers are the ring-opening
metathesis polymerization, followed by hydrogenation, and the
acyclic diene metathesis poly-condensation; both routes generally
require complex synthesis of monomers and a careful choice of the
catalytic system [2,15,16]. Furthermore, the functionalization of POs
can be carried out by the classical radical post-reactor modification
of the matrix by using unsaturated monomers, like for example
maleic anhydride or its derivatives. However, this well-assessed
methodology does not respond to the requirement of selectivity
and does not guarantee the structural preservation of the pristine
macromolecular architecture. Moreover, only some functional
* Corresponding author.
E-mail address: francesca.cicogna@pi.iccom.cnr.it (F. Cicogna).
http://dx.doi.org/10.1016/j.polymer.2015.11.018
0032-3861/© 2015 Elsevier Ltd. All rights reserved.

F. Cicogna et al. / Polymer 82 (2016) 366e377
367
groups can be grafted and often they need further modifications to
fulfill the final applications [16e18]. Some authors report the pos-
sibility to insert further modifiable functional groups on the surface
[12,19] or inside [20] a pre-formed polyolefin films by mild [21] or
severe photochemical techniques [22,23] or by chemical etching
[24]. By these methods functional POs bearing azo-aromatic groups
were prepared [23e27].
A feasible alternative to the previously described methods is the
nitroxide radical coupling (NRC) reaction (Scheme 1) that is the
method we used in this paper to prepare the POs functionalized
with RO-TEMPO derivatives.
authors reported studies of the kinetic behavior of azo-dyes cova-
lently grafted to semi-crystalline polyolefins (low density poly-
ethylene, LDPE, or polypropylene, PP) that were analyzed at room
temperature so well above their Tg and below their melt temper-
ature. In these examples, the kinetics of the cis-trans thermal back-
isomerization process was influenced by the crystallinity of the
matrix and by the steric hindrance of the azo-aromatic derivatives
[25,26].
In this paper, the synthesis of two nitroxide derivatives (RO-
TEMPO),
the
4-(phenylazo)-benzoyl-2,2,6,6-tetramethylpipe
ridine-1-oxyl radical (AzO-TEMPO) and the 4-(2-thienylazo)-ben-
zoyl-2,2,6,6-tetramethylpiperidine-1-oxyl radical (ThiO-TEMPO)
This reaction, which was already reported for the functionali-
zation of polyethylene [28,29] and polyesters [30,31], has many
advantages with respect to the classical radical post-reactor
modification of POs or polyesters with unsaturated monomers.
Indeed, it allows to insert specific/complex functionalities by a one-
step methodology because the process shows a great compatibility
with different functional groups. Moreover, by modulating the feed
ratio, it is possible to have a very good control of the macromo-
lecular architecture even though the functionalization was carried
out in the melt by using a peroxide as free radical initiator. Indeed,
the absence of the propagation and chain transfer steps, usually
present in the classical radical functionalization of POs with un-
saturated monomers, allows to limit the amount of free radicals
present during the functionalization process inhibiting or control-
ling all side reactions. Furthermore, it is possible to modulate the
grafting degree that depends on the quantity of reagents used in
the feed [28e31]. Generally, relatively low functionalization de-
grees (FDs) are planned and obtained with the aim of adding a
specific functionality without altering the thermal and mechanical
properties of the pristine material. The introduction of chromo-
phores (azo-aromatic derivatives, as in the present paper) by using
this methodology means to handle functional polymers with rela-
tively low FD values, however some interesting information can be
gathered also from less functionalized materials. Indeed, the photo-
physical properties of a chromophore are influenced by the sur-
rounding media rather than by the intramolecular interaction be-
tween different functional groups. The solidestate reactions are
generally characterized by a decrease of molecular mobility. As a
result, those processes that are unimolecular in solution are
frequently observed to progress by non-first-order mechanism in
the solid state due to the microscopically heterogeneous state of
aggregation or to the free-volume distribution of the solid medium
[32]. In this field, azobenzene and its derivatives were used as
microscopic photo-probes because their rate and extent of photo-
isomerization reflect the free-volume distribution and the local
mobility in the network structure. For example, from a careful ki-
netic evaluation of the trans-cis-trans photo- or thermal-
isomerization process of azo-compounds, it was possible to eval-
uate the free volume distribution of amorphous polymers below
the glass transition temperature (Tg) [33e35], to correlate the
physical aging of amorphous polymers with a reduction of the free
volume [36] or to study the extent of curing reactions [37]. Most
studies are dealing with amorphous polymer matrices that are
however analyzed in the glassy state below their Tg. Only some
(Fig. 1), as well as their grafting onto a copolymer ethylene/a-olefin
(co-EO) and high density polyethylene (HDPE) by the NRC reaction
are reported.
The aim of this work was first to assess the versatility of the NRC
reaction to obtain “functional” POs substituted with two different
chromophores that are able to isomerize under UV irradiation, then
to study the isomerization behavior of these chromophores in so-
lution and after grafting to the polymer matrix. The general pur-
pose was in fact to state this functionalization method as a good
tool to transfer the photo-physical properties of the free chromo-
phores to the polymer matrix and to verify how the polymer matrix
influences the photo-physical behavior of the chromophores.
The synthesis of ThiO-TEMPO is here reported for the first time
to prove the possibility of tuning the photo-physical properties of
the azo-aromatic moiety by modifying the nature/structure of the
benzene ring substituents [38e41] and to assess the compatibility
of the NRC reaction with the heteroaromatic group.
Both RO-TEMPO molecules were grafted to PE in the melt by
using a peroxide as a radical initiator. AzO-TEMPO was grafted onto
a copolymer ethylene/a-olefin (co-EO) that has a very low crystal-
linity (about 15%) and onto HDPE (crystallinity is about 70%),
whereas ThiO-TEMPO was grafted onto HDPE. These two matrices
were chosen to compare the photo-physical properties of AzO-
TEMPO grafted PE samples having very different crystallinity. The
functionalized polymers were characterized by FT-IR, TGA and
differential scanning calorimetry (DSC). An accurate photo-physical
analysis of both free RO-TEMPO derivatives and of functionalized
polymers was carried out by UVeVis spectroscopy under irradia-
tion at different wavelengths. Finally, water contact angle mea-
surements were used to confirm the photo-isomerization effects
onto wettability changes of the surface of these materials.
2. Experimental section
2.1. Materials
4-Hydroxy-2,2,6,6-tetramethylpiperidine-1-oxyl (HO-TEMPO)
(Fluka), 4-(phenylazo)-benzoyl chloride (Aldrich), triethylamine
(Aldrich), 4-aminobenzoic acid (Aldrich), tetrafluoroboric acid
(48 wt.
% in water, Aldrich), sodium nitrite (Aldrich), 2-
bromothiophene (Aldrich), di (tert-butylperoxy-isopropyl)-ben-
zene (mixture of isomers) (P, Perkadox 14S-FL, Akzo Nobel), tetra-
hydrofuran (THF, Aldrich), N,N0-dicyclohexylcarbodiimide (DCC,
Scheme 1. Nitroxide radical coupling (NRC) reaction.

368
F. Cicogna et al. / Polymer 82 (2016) 366e377
Fig. 1. Chemical structure of 4-(phenylazo)-benzoyl-2,2,6,6-tetramethylpiperidine-1-oxyl radical (AzO-TEMPO) and of 4-(2-thienylazo)-benzoyl-2,2,6,6-tetramethylpiperidine-1-
oxyl radical (ThiO-TEMPO).
Aldrich), 4-dimethylaminopyridine (DMAP, Aldrich), dimethylsulf-
oxide-d6 (DMSO-d6, Aldrich), dioxane (Aldrich), acetonitrile
(spectrophotometric grade, Aldrich) were used as received.
High density polyethylene (HDPE) Lacqtene 2070MN60 pro-
samples were obtained by potassium bromide pellet technique.
Spectra of polymers were obtained on films prepared by
compression molding at 190 ꢀC or 110 ꢀC under a pressure of about
6 tons by using the Carver press 3851-0.
duced by Arkema group having Mn
¼
21 000 g/mol,
Thermo gravimetric analyses (TGA) were carried out by using
the instrument Seiko EXSTAR 7200 TGA/DTA. In a typical experi-
ment, the sample (about 10 mg) was placed in an alumina sample
pan and the run was carried out at a standard rate of 10 ꢀC/min from
30 to 700 ꢀC under nitrogen flow.
Mw ¼ 86 000 g/mol, dispersity ¼ 4.10, density ¼ 0.96 g/cm3 and
MFR ¼ 7 g/10 min (2.16 kg at 190 ꢀC) and a random copolymer
ethylene/
a
-olefin (co-EO) having Mn
¼
50 000 g/mol,
Mw ¼ 102 000 g/mol, dispersity ¼ 2.15, density ¼ 0.87 g/cm3 and
MFR ¼ 4.7 g/10 (2.16 kg at 190 ꢀC) min were used as polymer
matrices. Fluka pre-coated aluminum F254 silica gel 60 sheets were
used for TLC analyses. Purifications by flash-chromatography were
performed using silica gel Merck 60 (particle size
0.040e0.063 mm). All the reactions were performed under argon
by standard syringe, cannula and septa techniques.
Differential Scanning Calorimetry (DSC) analysis was carried out
using a PerkineElmer DSC-4000 differential scanning calorimeter
thermal analyzer equipped with
a 3 stage cooler able to
reach ꢂ130 ꢀC. Thermal scans were carried out on 10e15 mg
samples under nitrogen atmosphere. Previously, the instrument
was calibrated by using indium (m.p. 156.6 ꢀC,
DH ¼ 28.5 J/g) and
zinc (m.p. 419.5 ꢀC). HDPE samples were heated from 30 to 180 ꢀC
then cooled to 30 ꢀC and heated again to 180 ꢀC at a cooling/
heating rate of 10 ꢀC/min. co-EO samples were cooled to ꢂ60 ꢀC,
heated to 130 ꢀC and cooled to 30 ꢀC at a cooling/heating rate of
10 ꢀC/min.
2.2. Characterization
Melting points were recorded on a hot-stage microscope
(Reichert Thermovar).
UVeVis absorption spectra were recorded at room temperature
with a PerkineElmer Lambda 25 UVeVis Spectrometer. Acetonitrile
solution of AzO-TEMPO and ThiO-TEMPO (about 5 ꢁ 10ꢂ5 M) and
films of functionalized polymers were analyzed. HDPE films were
prepared by compression molding; co-EO films were prepared by
solution casting onto a quartz plate.
Photo-activation of azobenzene, AzO-TEMPO, ThiO-TEMPO and
functionalized polymers was carried out with a Black Light equip-
ment from Helios Italquartz Triwood 25/36 system equipped with 2
quartz glass lamps screened at 366 nm (6 Watt each) and with 2
quartz glass lamps screened at 254 nm (6 Watt each).
GLC analyses were performed using two types of capillary col-
umns: an Alltech AT-35 bonded FSOT column (30 m ꢁ 0.25 mm i.d.)
and an Alltech AT-1 bonded FSOT column (30 m ꢁ 0.25 mm i.d.).
The mass spectra were acquired on an AB Sciex API 4000 triple
quadrupole mass spectrometer equipped with a turbo-V ionspray
source, coupled to a Perkin Elmer Series 200 Micro HPLC Pump. The
experimental conditions have been as follows: Curtain Gas (Ni-
trogen, CUR): 10; Gas Sources 1 & 2 (air; GS1 & GS2): 25; GS2
Temperature (TEM): 300 ꢀC; Ionspray Voltage (IS): 5.5 kV;
Declustering Potential (DP): 20 V; Entrance Potential (EP): 10 V;
Scan Range (Positive ions): 150e800 Th in 0.5 s; EI-MS spectra were
measured at 70 eV by GLC/MS.
Static contact angles were measured by the sessile drop method
with a FTA200 Camtel goniometer, using water (qw). The value of
NMR spectra (Varian VXR 300) were recorded at room tem-
perature, at 300 MHz (1H) or at 282.2 MHz (19F) and were referred
to TMS (1H) or C6F6 (19F).
static contact angle is obtained as the average of 10 measures. The
data were recorded after 25 s for every drop.
X-band EPR spectra were obtained with a Varian E112 spec-
trometer equipped with a Varian E257 temperature control unit.
The EPR spectrometer was interfaced with an IPC 610/P566C in-
dustrial grade Advantech computer by means of a data acquisition
system. This unit consists of an acquisition board capable of
acquiring up to 500 000 12-bit samples per second. The sample was
prepared by adding 0.1 mL of a solution of ThiO-TEMPO in
dichloromethane (3.4 ꢁ 10ꢂ4 M) to 45 mg of silica gel, after the
evaporation of the solvent the solid sample was transferred into the
quartz tube (3 mm internal diameter). Spectra were recorded at
room temperature before and after the heating of the sample at
190 ꢀC for 10 min.
2.3. Synthesis of 4-(phenylazo)-benzoyl-2,2,6,6-
tetramethylpiperidine-1-oxyl radical (AzO-TEMPO)
A
solution of 4-(phenylazo)-benzoyl chloride (0.71 g,
2.9 ꢁ 10ꢂ3 mol), HO-TEMPO (0.5 g, 2.9 mmol) and trietylamine
(0.47 mL, 2.9 mmol) in dichloromethane (20 mL) was mixed at
room temperature under nitrogen for 24 h. Crude mixture was
hydrolyzed with water (50 mL) and extracted with dichloro-
methane (3 ꢁ 50 mL). AzO-TEMPO was purified by column chro-
matography by using toluene/ethyl acetate (8/2) as eluting mixture.
Yield 80%. Melting point ¼ 130 ꢀC [42]; MS(CI): m/z: 381 [M þ 1],
FT-IR spectra were recorded by the Fourier Transform Spec-
trometer Perkin-Elmer Spectrum 100. Spectra of microcrystalline
380 [M]; FT-IR (KBr):
1279, 1240, 1179, 1112, 865, 776, 691 cm
n
¼ 3055, 2970, 2936, 1717, 1468, 1407, 1316,
ꢂ1
C H26N3O3 Calculated C,
22

F. Cicogna et al. / Polymer 82 (2016) 366e377
369
69.45; H, 6.89; N, 11.04. Found: C, 69.40; H, 6.87; N, 11.00.
concentrated in vacuum and the residue was purified by flash
chromatography on silica gel using a mixture of n-hexane and ethyl
acetate (50:50) as eluent, to yield 2 (0.46 g, 15%) as an orange solid.
MS [m/z ]: 232 (86) [Mþ]; 121 (26); 111 (100); 83 (48); 65 (26);
39 (36).
2.4. Synthesis of 4-(2-thienylazo)-benzoyl-2,2,6,6-
tetramethylpiperidine-1-oxyl radical (ThiO-TEMPO)
1H NMR (DMSO-d6)
d
ppm 8.13 (m, J ¼ 8.5 Hz, 2H); 7.92 (dd,
The synthesis of ThiO-TEMPO was carried out as reported in
Scheme 2.
J ¼ 3.9 Hz, J ¼ 1.2 Hz, 1H); 7.88 (m, J ¼ 8.5 Hz, 2H); 7.66 (dd,
J ¼ 5.2 Hz, J ¼ 1.2 Hz, 1H); 7.25 (dd, J ¼ 5.2 Hz, J ¼ 3.9 Hz, 1H).
2.4.1. Synthesis of p-carboxybenzenediazonium tetrafluoroborate
(1) [43]
2.4.3. Synthesis of 4-(2-thienylazo)-benzoyl-2,2,6,6-
tetramethylpiperidine-1-oxyl radical (ThiO-TEMPO)
4-Aminobenzoic acid (6.86 g, 50 mmol) was dissolved in a so-
lution of tetrafluoroboric acid (48 wt. % in water, 26.2 mL, 36.58 g,
200 mmol) and water (50 mL). The mixture was stirred at room
temperature until the aminobenzoic acid was completely dissolved,
and then cooled at 0 ꢀC (some precipitation of solid occurred). The
solution was maintained at 0 ꢀC and a solution of sodium nitrite
(3.65 g, 53 mmol) in water (10 mL) was added dropwise within
30 min. During the addition it was observed the precipitation of a
solid. At the end of the addition the mixture was allowed to warm
to room temperature, which caused the complete dissolution of the
solid. The reaction mixture was then cooled to 5 ꢀC, and the
resulting solid was recovered by filtration, washed with cold
diethyl ether and dried in vacuo, giving 1 as a colorless solid (6.25 g,
To a suspension of 2 (0.46 g, 2 mmol) and 4-hydroxy-2,2,6,6-
tetramethylpiperidine-1-oxyl (0.35 g,
2 mmol) in dichloro-
methane (50 mL) was added N,N0-dicyclohexylcarbodiimide
(0.49 g, 2.4 mmol) and 4-dimethylaminopyridine (0.29 g,
2.4 mmol). The resulting clear solution was stirred at room tem-
perature for 30 h, during the reaction a precipitate was formed. At
the end of the reaction, dichloromethane was added and the sus-
pension was filtered on Celite. The crude product was purified by
flash chromatography on silica gel using a mixture of toluene and
ethyl acetate (85:15) as the eluent, to give ThiO-TEMPO as a deep
orange solid (0.54 g, 71%). Melting point ¼ 155e158 ꢀC; MS(CI) [m/
z]: 386 [Mþ]; FT-IR (KBr):
1316, 1279, 1240, 1179, 1112, 865, 776, 711 cmꢂ1. Anal. Calc. for
20H24N3SO3: C, 62.15; H, 6.26; N, 10.87; S, 8.30. Found: C, 62.10; H,
n
¼ 3077, 2970, 2936, 1717, 1468, 1411,
53% yield). 1H NMR (DMSO-d6)
d
ppm 8.80 (m, J ¼ 7.8 Hz, 2H); 8.45
(m, J ¼ 7.8 Hz, 2H). 19F NMR (DMSO-d6)
d
ppm ꢂ148.2; ꢂ148.3.
C
6.21; N, 10.82; S, 8.28%.
2.4.2. Synthesis of 4-(2-thienylazo)-benzoic acid (2)
According to the procedure reported in the literature for the
synthesis of 4-(2-thienylazo)-N,N-diethylaniline [44] to a suspen-
sion of magnesium turnings (0.734 g, 30.2 mmol) in tetrahydro-
furan (10 mL) was added dropwise a solution of 2-bromothiophene
(2.9 mL, 4.89 g, 30 mmol) in tetrahydrofuran (20 mL) at room
temperature (an exothermic reaction was observed). After all the
magnesium was consumed (3 h), the resulting solution was added
via cannula to a suspension of p-carboxybenzenediazonium tetra-
fluoroborate (1) (3.54 g, 15 mmol) in a mixture of tetrahydrofuran
and dioxane (5/2, 70 mL) cooled at ꢂ50 ꢀC. After the addition, the
resulting deep red mixture was allowed to warm to room tem-
perature and stirred overnight. The reaction mixture was then
quenched with a saturated aqueous sodium carbonate solution and
washed with ethyl acetate. The aqueous basic phase was acidified
to pH ¼ 5 with acetic acid and extracted with ethyl acetate
(4 ꢁ 30 mL) and with dichloromethane (2 ꢁ 30 mL). The collected
organic extracts were dried on anhydrous sodium sulfate,
2.5. Melt functionalization procedure
The melt reactions were carried out in an internal batch mixer
(Brabender Plastograph OHG47055) with a chamber of 30 mL.
Torque and temperature data were acquired by Brabender Mixing
software Win-Mix ver.1.0. Functionalization of HDPE was carried
out at 190 ꢀC, 50 rpm for 20 min. 20 g of HDPE were introduced in
the hot mixer and after the melting (about 5 min were necessary to
completely melt the polymer and to reach a constant torque signal),
0.145 mol% of ThiO-TEMPO or AzO-TEMPO were added and 1 min
later 0.043 mol% of P was introduced in the brabender chamber
(Table 1) [28]. Functionalized crude samples were cut in small
pieces and extracted with boiling acetone for 16 h and further
purified by dissolution in hot xylene and precipitation from
acetone. Samples were dried to constant weight and analyzed.
Melt functionalization of the random copolymer ethylene/
a-
olefin (co-EO) was carried out at 170 ꢀC, 50 rpm for 20 min by
Scheme 2. Synthesis of 4-(2-thienylazo)-benzoyl-2,2,6,6-tetramethylpiperidine-1-oxyl radical (ThiO-TEMPO).

370
F. Cicogna et al. / Polymer 82 (2016) 366e377
Table 1
Radical functionalization of polyolefins: feed composition, functionalization degree (FD) and thermal properties.
Sample name
Feed composition
RO-TEMPOb (mol%)
Functionalization degreea
FDFT-IRc (mol%)
Thermal properties
d
e
f
Peroxide (mol%)
FDTGA (mol%)
T5% (ꢀC)
Tmax (ꢀC)
co-EOg
e
e
e
e
440
456
430
447
455
470
488
475
486
487
HDPEg
e
e
e
e
(co-EO)-g-(AzO-TEMPO)
HDPE-g-(AzO-TEMPO)
HDPE-g-(ThiO-TEMPO)
0.28
0.14
0.14
0.09
0.04
0.04
0.18
0.12
0.11
0.20
0.09
0.08
a
Functionalization Degree (FD): moles of the grafted functional groups with respect to 100 mol of monomer repeating units.
RO-TEMPO: functional TEMPO derivative.
FD evaluated by FT-IR analysis.
FD evaluated by TGA analysis under nitrogen.
Temperature corresponding to 5% weight loss.
b
c
d
e
f
Maximum degradation temperature.
co-EO and HDPE starting polymers.
g
adopting the same procedure described for the preparation of
asymmetric stretching of the diazo group (See Fig. 1 in Ref. [47])
[48].
functionalized HDPE samples. In this case 0.28 mol% of AzO-TEMPO
and 0.09 mol% of P were used (Table 1). Crude sample was extracted
with boiling acetone for 16 h in order to remove all low molecular
weight sub-products and it was further purified by dissolution in
hot toluene (80 ꢀC) and collected by precipitation of the toluene
solution in acetone. The purified sample was dried to constant
weight and analyzed.
The functionalization degree (FD) of the functionalized poly-
mers was evaluated as previously reported by using the FT-IR
spectroscopy [28]. Particularly, in the case of functionalized co-
EO, the FD was evaluated by using a previously described calibra-
tion curve [28] where the ratio between the area of the band at
1112 cmꢂ1 (associated to the grafted species) and that of the
methylene rocking of polyethylene at 720 cmꢂ1 was considered. In
the case of HDPE, in the spectral region near 700 cmꢂ1 two bands
are present at about 720 cmꢂ1 and 730 cmꢂ1 attributed to the
crystalline phase of HDPE. These two bands are superimposed to a
broad band centered at about 723 cmꢂ1 associated with the
amorphous phase of the polymer matrix [45]. To evaluate the FD of
HDPE functionalized samples (Table 1), the total area of all bands in
the region between 700 and 740 cmꢂ1 was considered as repre-
sentative of the total amount of methylene groups [46], and the
same calibration curve used in the case of co-EO was applied.
Thermograms of RO-TEMPO derivatives showed
a single
degradation step in the case of AzO-TEMPO (Tonset ¼ 270 ꢀC) and a
gradual weight loss between 100 and 240 ꢀC followed by the main
degradation step (Tonset ¼ 248 ꢀC) in the case of ThiO-TEMPO (See
Fig. 2 in Ref. [47]). The initial weight loss was attributed to the
volatilization of some side products or reagent residues derived
from the synthesis of ThiO-TEMPO (see Experimental Section) and
in a very limited extent to the volatilization/degradation of the
nitroxide as supported by EPR measurements collected after
treating the nitroxide at 190 ꢀC for 10 min (See Fig. 3 in Ref. [47]).
Functionalization of HDPE and co-EO was carried out as previously
reported [28,29] at 190 ꢀC and 170 ꢀC, respectively (See Experi-
mental section and Table 1).
FT-IR spectra of functionalized HDPE and co-EO, collected after a
careful purification of the crude samples to remove all low mo-
lecular weight sub-products, confirmed the grafting of the RO-
TEMPO derivatives (Figs. 2 and 3) as evidenced by the presence of
a band at 1722 cmꢂ1 due to the carbonyl stretching of the grafted
functionalities.
From the FT-IR spectra of functionalized HDPE samples, and by
using the data of a previously reported calibration curve [28], it was
possible to roughly evaluate the functionalization degree (FD) of
the HDPE-g-(RO-TEMPO) samples which was about 0.1 mol% for
both RO-TEMPO derivatives. In the case of co-EO and by using the
same calibration curve, an FD value of about 0.18 mol% was calcu-
lated reflecting the different feed conditions used to functionalize
3. Results and discussion
3.1. Preparation and characterization of RO-TEMPO functionalized
HDPE and co-EO
The
synthesis
of
4-(phenylazo)-benzoyl-2,2,6,6-
tetramethylpiperidine-1-oxyl radical (AzO-TEMPO) was carried
out by modifying a procedure already reported in the literature
[42],
whereas
the
4-(2-thienylazo)-benzoyl-2,2,6,6-
tetramethylpiperidine-1-oxyl (ThiO-TEMPO) (Scheme 2) was ob-
tained by esterification of the (E)-4-(thiophen-2-yldiazenyl)-ben-
zoic acid (2) prepared from 4-carboxybenzenediazonium
tetrafluoroborate (1) and 2-thienylmagnesium bromide, according
to the procedure reported by Moilan, McNelis and coworkers [44].
After the purification of the two RO-TEMPO derivatives by col-
umn chromatography, they were characterized by HPLC-MS anal-
ysis and by FT-IR spectroscopy both confirming their molecular
structure. The infrared spectra of RO-TEMPO derivatives showed a
band at 1715 cmꢂ1 due to the carbonyl stretching and a band at
3055 cmꢂ1, in the case of AzO-TEMPO, and at 3077 cmꢂ1, in the case
of ThiO-TEMPO, attributable to the CeH stretching of the benzene
and of the thiophene, respectively. Both spectra showed also a very
weak band at about 1410 cmꢂ1 that can be attributed to the
Fig. 2. FT-IR spectra of HDPE, HDPE-g-(AzO-TEMPO) and HDPE-g-(ThiO-TEMPO).

F. Cicogna et al. / Polymer 82 (2016) 366e377
371
Fig. 5. TGA thermograms and their first derivative of HPDE, HDPE-g-(AzO-TEMPO) and
HDPE-g-(ThiO-TEMPO). The analyses were carried out under nitrogen flow at a heating
rate of 10 ꢀC/min.
Fig. 3. FT-IR spectra of co-EO and (co-EO)-g-(AzO-TEMPO).
the HDPE and the co-EO. On the basis of the FD values, and by
considering the number average molecular weight of the pristine
polymer (See Experimental section), it was possible to state that
about 1 functional group per each HDPE polymer chain and about
2e3 functional groups per each co-EO chain were grafted. As dis-
cussed in the introduction, due to these low FDs and to the random
arrangement of the functional groups onto the polymer chain, the
intermolecular interactions between chromophores can be
excluded. For this reason, the trans-cis-trans photo- and thermal-
isomerization of the chromophores grafted to the polyolefins
were studied and compared with those characteristics of the free
RO-TEMPO derivatives.
analysis, can be lost by decomposition or volatilization. The per-
centage of this weight loss roughly corresponds to the FD evaluated
by FT-IR analysis (Table 1) confirming the attribution of this TGA
step to the detachment of TEMPO moieties.
3.2. Photo-physical properties of free RO-TEMPO derivatives
The comparison of the UVeVis spectra of both RO-TEMPO de-
rivatives with that of the azobenzene, all obtained from diluted
acetonitrile solutions, is shown in Fig. 6 and all spectral parameters
are reported in Table 2. The main absorption band of AzO-TEMPO at
To evaluate the thermal stability of the functionalized polymers
and to state if at high temperature a homolytic bond cleavage of the
>NeOeC bond can cause the loss of the grafted TEMPO from the
polymer matrix, the samples were analyzed by TGA under nitrogen.
Thermograms of functionalized polymers evidenced a first weight
loss (about 1 wt%) starting at about 270 ꢀC (Fig. 4 and Fig. 5), fol-
lowed by the main degradation step of the polymer matrix at about
485 ꢀC, in the case of HDPE, and at about 470 ꢀC, in the case of co-EO
(See Figs. 4 and 5 in Ref. [47]). Both temperatures are very close to
the degradation temperature of the two pristine polymers sug-
gesting that the functionalization process did not alter the thermal
degradation mechanism of the matrix. As regards the first weight
loss, it can be due to the detachment of the grafted TEMPO moiety
that, at high temperature and under the gas flow used during the
323 nm (p-p* transition) is red shifted of about 10 nm and has a
higher molar extinction coefficient (ε) with respect to the same
band of azobenzene [42]. The other two bands of AzO-TEMPO at
228 nm (s-s* transition) [49] and 446 nm (n-p* transition) have
the same position as those of azobenzene. The main absorption
band of ThiO-TEMPO is even more red shifted (about 53 nm) with
respect to that of azobenzene. This causes the partial overlap of the
ThiO-TEMPO main band with the band at 446 nm that in this
spectrum looks like a long tail of the main absorption. This large red
shift is due to the presence of the electron donating thiophene ring
(substituent of the N]N bond) that makes ThiO-TEMPO classifiable
as “amino-benzene” type chromophore characterized by an energy
of the main
[38].
p-p* transition similar to that of the n-p* transition
Fig. 4. TGA thermograms and their first derivative of (co-EO) and (co-EO)-g-(AzO-
TEMPO). The analyses were carried out under nitrogen flow at a heating rate of 10 ꢀC/
Fig. 6. UVeVis absorption spectra of azobenzene, AzO-TEMPO and ThiO-TEMPO in
min.
acetonitrile.

372
F. Cicogna et al. / Polymer 82 (2016) 366e377
Table 2
UVeVis absorption and photo-physical properties of azobenzene, AzO-TEMPO and ThiO-TEMPO in solution (CH3CN).
a
Sample
l max (nm)
ε max (Mꢂ1 cmꢂ1
)
PSS366 (min)
35
Composition at PSS366
100% Cis
Dark cis-transb h (% trans)
16 h (8.5%)d
PSS254c (min)
35
Composition at PSS254
n.d.
Azobenzene
228
317
446
228
323
446
237
369
14500
23000
550
14100
25500
266
AzO-TEMPO
35
6
100% Cis
91% Cis
24 h (10%)
6 h (80%)
35
20
0% Cis
ThiO-TEMPO
11260
23200
38% Cis
a
Irradiation time necessary to reach the Photo Stationary State under 366 nm emitting lamp (PSS366).
Time after which the tabulated composition is regained during thermal cis-trans back-isomerization.
b
c
Irradiation time necessary to reach the Photo Stationary State under 254 nm emitting lamp (PSS254). All samples were irradiated from the cis rich form.
d
5 ꢁ 10ꢂ5 M in THF [49].
The irradiation of the acetonitrile solutions of azobenzene and
of RO-TEMPO derivatives at 366 nm caused the photo-
isomerization of all compounds from the trans isomer to a new
Photo Stationary State (PSS366) rich in the cis isomer (Figs. 7 and 8,
and Fig. 6 in Ref. [47] for azobenzene). In all cases isosbestic points
were detected (two in the case of azobenzene and AzO-TEMPO and
three in the case of ThiO-TEMPO) confirming that the isomerization
involves two species that are in equilibrium between each other.
The irradiation of azobenzene and AzO-TEMPO solutions at
366 nm caused the decrease of the absorption band at 320 nm and
the increase of the band at about 460 nm [42] (See Fig. 7 in
Ref. [47]). The new photo stationary states (PSS366) were reached in
about 35 min of irradiation and were mainly composed of the pure
cis isomer [50]. The irradiation of ThiO-TEMPO solution with the
same lamp caused the decrease of the band at 369 nm and the
increase of a band at 284 nm (See Fig. 8 in Ref. [47]). The decrease of
the main band at 369 nm can screen and partially compensate the
expected increase of the less intense band at 440 nm that appar-
ently decreases during irradiation. In this case, the PSS366 was
reached in about 6 min of irradiation. The faster isomerization of
ThiO-TEMPO with respect to that of AzO-TEMPO can be ascribed to
the higher molar extinction coefficient of ThiO-TEMPO at the irra-
diation wavelength rather than to a different photo-isomerization
efficiency of the two chromophores.
Fig. 8. Absorption spectra of ThiO-TEMPO in acetonitrile solution after irradiation at
366 nm and 254 nm.
in the UV region, the difference between the molar extinction co-
efficient of the cis and trans isomers is lower. Only at about 240 nm
the cis isomer of azobenzene absorbs more than the trans isomer.
Anyway, the presence of substituents onto the benzene ring of
azobenzene derivatives can shift the absorption maxima both of
the cis and trans isomer thus, in some cases, the cis isomer can
absorb more than the trans isomer both in the visible and in the UV
region. In these cases, the cis-trans photo-isomerization can be
promoted by irradiation with visible or UV light [51e53]. In the case
of AzO-TEMPO the cis isomer absorbs more than the trans species
both in the visible region and in the UV between 240 and 280 nm,
whereas, the cis isomer of ThiO-TEMPO has a higher extinction
coefficient only in the UV region between 250 and 300 nm. As a
consequence, to photo-isomerize the cis isomer of AzO-TEMPO
both the visible or the UV light can be used, whereas, in the case
of ThiO-TEMPO only UV irradiation can be effective. To verify this
hypothesis, irradiation at 254 nm was carried out. As hypothesized,
the irradiation of the cis isomer of AzO-TEMPO with a 254 nm
emitting lamp completely restored the trans isomer in 35 min. The
same conversion was observed also in the case of azobenzene
showing that its cis isomer can be photo-isomerized by irradiation
in the UV region. Irradiation of both cis and trans ThiO-TEMPO with
the same lamp gave a new PSS254 (Fig. 8) having a spectral pattern
that appears as a combination of the spectrum of the trans and cis
isomers. This result is confirmed by the TLC analysis of the irradi-
ated solution that evidenced the presence of two products. To
evaluate the composition of this PSS254, the absorption spectrum of
the pure cis isomer of ThiO-TEMPO would be necessary; however
by considering that the separation of the two isomers was not
possible neither by column chromatography nor by other
The back cis-trans isomerization of azobenzene derivatives can
be obtained thermally or by irradiation with opportune wavelength
light. Generally, the photo-isomerization of azobenzene is carried
out by visible light because in this spectral region the cis isomer has
a higher extinction coefficient than the trans isomer [38], whereas
Fig. 7. Absorption spectra of AzO-TEMPO in acetonitrile solution after irradiation at
366 nm and 254 nm.

F. Cicogna et al. / Polymer 82 (2016) 366e377
373
techniques, the spectrum of the cis isomer of ThiO-TEMPO was
esteemed on the basis of the Fischer method [54]. To apply this
method two requirements are necessary: first, upon irradiation of a
sample by two different wavelengths, two PSSs having different
composition have to be reached; second, the trans-cis system has to
be sufficiently thermally and photo-chemically stable to make
possible the establishment of true photo-stationary states. It is also
necessary to make the assumption that the ratio between the
quantum yields for the trans-cis and cis-trans isomerization are
constant, independently of the irradiation wavelength. In the case
of ThiO-TEMPO the first two requirements are fulfilled, whereas for
the last assumption there is no possibility to evaluate the quantum
yield at different wavelengths and it was assumed that this hy-
pothesis is also valid.
Application of the Fischer method to the two PSSs, obtained
upon irradiation of ThiO-TEMPO solution with 254 nm and 366 nm
lamps, allowed to extrapolate the spectrum of the cis isomer (Fig. 8)
and to esteem the composition of the mixture at PSSs (Table 2). To
account for the different compositions at PSSs, it is necessary to
consider that the trans-cis isomerization is both a photo-chemically
and thermally reversible process. Both processes are inter-
connected and depend on the nature of azobenzene derivatives.
Particularly, the rate of the photochemical process depends on the
quantum yield of the cis-trans isomerization and on the extinction
coefficients of each isomer at the irradiation wavelength; whereas,
the thermal reversibility is mainly related to the nature of sub-
stituent on the azobenzene and to the isomerization medium. In
the case of ThiO-TEMPO, irradiation at 254 nm gives a mixture of
the two isomers because this wavelength is quite near to one iso-
sbestic point of this molecule (Fig. 8). At this wavelength the
extinction coefficient of the cis and trans isomer is similar, so both
the isomers are absorbing during irradiation and an equilibrium
composition is obtained.
A rough evaluation of the quantum yield for the trans-cis
isomerization of AzO-TEMPO and ThiO-TEMPO was carried out
upon irradiation at 366 nm. For the calculation, the equation and
the method reported in the literature by Gauglitz [55] and Rau [56]
were applied. To evaluate the incident irradiation intensity, the
photo-isomerization of azobenzene was used as actinometer [57].
Quantum yields obtained by this method for both AzO-TEMPO and
ThiO-TEMPO were quite low being about 0.05, but they are in
agreement with the value reported in the literature for AzO-TEMPO
[42].
Thermal back-isomerization from cis to trans isomer was
observed for both RO-TEMPO derivatives, but the process is quite
slow especially in the case of AzO-TEMPO that regained only about
35% of trans form in 24 hs. Faster back isomerization is observed in
the case of cis ThiO-TEMPO that needs about 6 hs to recover 90% of
trans isomer. The rate of the back thermal isomerization depends
on the mechanism of the process. It can occur either by inversion of
a nitrogen center or via rotation where the latter seems to be
favored for derivatives with strong dipole moment [49], even if a
combination of the two mechanisms cannot be excluded. The
presence of a polar substituent on the phenyl ring, generally, ac-
celerates the isomerization rate by increasing the dipole moment of
the molecule and by lowering the activation barrier of the thermal
relaxation. This effect can be responsible for the faster thermal
isomerization detected in the case of ThiO-TEMPO that has the
thiophene ring, an electron donor group, substituted to N]N bond
[49].
region, whereas the cis and trans ThiO-TEMPO isomers have similar
extinction coefficients in the visible and then a partial conversion
can be obtained by irradiation at this wavelength.
Finally, more irradiation cycles were repeated both in the case of
AzO-TEMPO and ThiO-TEMPO giving in all cases the same compo-
sition and the same conversion as a demonstration of the fact that
trans-cis-trans isomerization is completely reversible and the
process is repeatable.
3.3. Photo-physical properties of HDPE-g-(RO-TEMPO) and (co-
EO)-g-(AzO-TEMPO)
The photo-physical properties of RO-TEMPO functionalized POs
were evaluated by collecting spectra of polymer films before and
after irradiation with an opportune wavelength lamp. The ab-
sorption spectra of the RO-TEMPO grafted to the POs are quite
similar to the spectra recorded for the free chromophores in solu-
tion, suggesting that they are molecularly dispersed in the matrices
(Fig. 9). Noteworthy, the main absorption band of AzO-TEMPO
grafted to HDPE or to co-EO shows a partially resolved vibrational
structure that is more evident in the spectra of functional POs than
in the spectrum of Azo-TEMPO recorded in solution (See Fig. 7 in
Ref. [47]). The presence of a low resolved vibrational structure is
described in the literature also for azobenzene solutions [38].
Interestingly, a higher vibrational resolution was observed in the
UV spectrum of azobenzene embedded in polyethylene films or in
solution cooled to 77 K [58]. Therefore, the fact that in the spectra of
AzO-TEMPO recorded in solution, the vibrational structure is less
evident than in the spectra recorded from polymers, suggests that
AzO-TEMPO grafted to the polymer matrices is probably organized
in a rigid environment.
The comparison between the free and grafted RO-TEMPO de-
rivatives evidenced also that the grafted chromophores exhibited
the trans-cis-trans isomerization processes (Figs. 10 and 11 and
Fig. 9 in Ref. [47]) already observed for the free RO-TEMPO de-
rivatives, suggesting that the photo- and thermal-isomerization in
the polymer matrix proceeded without side reactions.
Differences, especially in terms of isomerization efficiency and
isomerization kinetics, are anyway detected between chromo-
phores free in solution or embedded, after grafting, in the polymer
matrix.
Particularly, the degree of photo-isomerization (R) at PSS366 or
at PSS254 (Table 3), evaluated from the relation: R ¼ [(A0 ꢂ A∞)/
A0] ꢁ 100 (where A0 is the absorbance before the irradiation and A∞
Exposition of cis isomers of both RO-TEMPO derivatives to
natural light (laboratory light) gives two new PSSs containing about
92% trans isomer in the case of AzO-TEMPO and 40% of trans isomer
in the case of ThiO-TEMPO. This behavior confirms that the cis AzO-
TEMPO absorbs more than the trans isomer in the visible and UV
Fig. 9. Superimposition of the UVeVis absorption spectra of pristine HDPE, HDPE-g-
(AzO-TEMPO) and HDPE-g-(ThiO-TEMPO).

374
F. Cicogna et al. / Polymer 82 (2016) 366e377
sites.
The time necessary to reach the PSSs under irradiation is longer
for the RO-TEMPO derivatives grafted to the polymer matrix than
for the free chromophores in acetonitrile solution. This effect can be
well evidenced by reporting the absorbance variation as a function
of the irradiation time (Figs. 12 and 13). The comparison of the
curves, even considering the different starting absorbance, clearly
shows that the isomerization of the free molecules in solution is
faster and gives a larger degree of photo-isomerization.
The absorbance data collected during the irradiation can be used
to evaluate the kinetics of the trans-cis photo-isomerization pro-
cess and, as a first approximation, kinetic data can be fitted by a first
order kinetic law (Eq (1)):
ln½ðA0 ꢂ A∞Þ=ðA0 ꢂ AtÞꢃ ¼ kt
(1)
Fig. 10. UVeVis absorption spectra of HDPE-g-(AzO-TEMPO) collected during irradi-
ation at 366 nm.
where A0, At and A∞ are the absorbance values before irradiation, at
time of irradiation and at PSS, respectively, and I0
t
k
¼
(εtFtc þ εcFct)ln10 þ kct, where I0 is the incident irradiation in-
tensity, εt and εc are the molar extinction coefficient of the trans and
cis form at the irradiation wavelength, respectively; Ftc and Fct are
the isomerization quant yields of the trans-cis and cis-trans photo-
isomerization and kct is the rate constant for the thermal cis-trans
isomerization. It is necessary to point out that the photo-
isomerization of an azo-aromatic compound is a very complex
process since it involves many different aspects that it is difficult to
consider. The process is not only a pure photochemically reversible
process, but it involves a thermal conversion that can play a sig-
nificant role [60]. A more complex mathematical and experimental
evaluation is far behind the aim of this paper, accordingly, equation
(1) was used as a tool to easily compare the experimental results
reported in Figs. 12 and 13.
By applying Eq (1), higher rates constant were obtained (Figs. 14
and 15, and Table 4) for the trans-cis isomerization of both RO-
TEMPO molecules in solution than for the same chromophores
grafted to the polymers. Moreover, a lower isomerization rate of
AzO-TEMPO was observed when it is grafted to HDPE than to co-EO.
To account for these differences, it is necessary to consider that the
trans-cis isomerization of azo-aromatic derivatives is accompanied
by the rearrangement of the hindered azobenzene moiety. For
example, in the case of azobenzene, the distance between the para
carbon atoms of the aromatic rings goes from 0.90 nm (trans) to
0.55 nm (cis) [11]. If the photo-isomerization of the chromophore is
occurring when it is embedded in a polymer matrix, a change of the
Fig. 11. UVeVis absorption spectra of HDPE-g-(ThiO-TEMPO) collected during irradi-
ation at 366 nm.
Table 3
Degree of photo-isomerization R ¼ [(A0 ꢂ A∞)/A0] ꢁ 100 at PSS reached upon
irradiation at 366 nm (PSS366) and at 254 nm (PSS254).
a
Sample
R at PSS366
R at PSS254
AzO-TEMPO (CH3CN solution)
ThiO-TEMPO (CH3CN solution)
(co-EO)-g-(AzO-TEMPO)
HDPE-g-(AzO-TEMPO)
82
83
57
60
64
e
34
e
e
HDPE-g-(ThiO-TEMPO)
20
a
The PSS was reached upon irradiation of the film having the absorption spec-
trum typical of the trans isomer of grafted ThiO-TEMPO.
is the absorbance at PSS at the same wavelength) [59] evidenced
that the process is less effective when RO-TEMPO derivatives are
grafted to the polymer than when they are free in solution. For
example R is about 60% for RO-TEMPO grafted to polymer chains,
and analyzed as polymer films, whereas it is about 80% when the
molecules are irradiated in solution. Evidently, the presence of the
polymer network affects the isomerization of the azo-moiety and
only a fraction of the grafted molecules can effectively be iso-
merized. As discussed later, these deviations of the final conver-
sions from that obtained in solution can be due to the restriction of
the mobility of chromophores in polymer solids and to the het-
erogeneous distribution of local free-volume at the isomerization
Fig. 12. Absorbance vs. irradiation time of AzO-TEMPO (acetonitrile solution) and (co-
EO)-g-(AzO-TEMPO) and HDPE-g-(AzO-TEMPO) (polymer films). Irradiation wave-
length 366 nm.

F. Cicogna et al. / Polymer 82 (2016) 366e377
375
Table 4
Kinetic parameters for the trans-cis isomerization of AzO-TEMPO and ThiO-TEMPO
in solution or grafted to HDPE or co-EO and crystallinity of co-EO and HDPE samples.
Sample
k (secꢂ1
)
Crystallinity (%)a
AzO-TEMPO solution
(co-EO)-g-(AzO-TEMPO)
HDPE-g-(AzO-TEMPO)
ThiO-TEMPO solution
HDPE-g-(ThiO-TEMPO)
2.0 ꢁ 10ꢂ3 9 ꢁ 10ꢂ6
1.6 ꢁ 10ꢂ3 2 ꢁ 10ꢂ5
9.0 ꢁ 10ꢂ4 9 ꢁ 10ꢂ6
1.4 ꢁ 10ꢂ2 3 ꢁ 10ꢂ4
6.7 ꢁ 10ꢂ3 5 ꢁ 10ꢂ5
e
13.3b
65.0c
e
63.2c
a
Evaluated by DSC analysis considering that the melting enthalpy of poly-
ethylene 100% crystalline is 290 J/g.
b
Evaluated on the first heating scan [61]. The crystallinity of pristine co-EO was
about 15%.
c
Evaluated on the second heating scan. The crystallinity of pristine HDPE was
about 70%.
the chromophores, the process occurs faster. On the contrary, those
polymers that are more rigid and have small free volumes hinder
the isomerization and slow down the trans-cis isomerization pro-
cess. The reverse is true for the back thermal isomerization process;
indeed, those polymers that are more flexible are also able to adapt
more rapidly to the size and shape of the cis isomer, as a conse-
quence the thermal back isomerization to the trans form is slower
because of the greater stabilization of the cis isomer. In a rigid
media, the polymer chains change their conformation slowly than
the isomerization process; as a consequence the cis isomer is, for
longer time, surrounded by a polymer conformation that resembles
the template shape of the trans isomer rather than that of the cis
form. In these conditions, the cis isomer quickly re-isomerizes to
the trans form [25]. On the basis of these considerations, two as-
pects have to be considered to compare the isomerization rate
between different media: first the mobility of the polymer matrix
or generally of a solution, and second the free volume distribution.
With regard to the isomerization of Azo-TEMPO grafted to co-EO or
HDPE, the data reported in Table 4 suggest that the process is faster
in co-EO than in HDPE. Evidently, the co-EO chains can more easily
change their conformation with respect to HDPE probably because
the co-EO chains are more flexible and/or the free volume of HDPE
is smaller. Both contributes can be related to the crystallinity of the
two matrices; indeed the crystalline phase of a polymer is rigid and
also the interfacial region between the amorphous and crystalline
phase is considered to be stiffer than the flexible amorphous phase.
As a consequence, a polymer that is characterized by a higher
crystallinity is also more rigid [25,26,62] or less flexible, than a less
crystalline matrix. So, with the aim to relate the isomerization ki-
netics observed for the functionalized polymers, with the crystal-
linity of the matrix, all samples prepared in this paper were
analyzed by DSC (Table 4, See Figs. 10e13 in Ref. [47]). The analysis
evidenced that co-EO has a lower crystallinity than HDPE and that
after grafting of RO-TEMPO, it further decreases [63]. Moreover,
both RO-TEMPO moieties were expected to be localized in the
amorphous region, or at least at the interfacial region between the
amorphous and the crystalline fraction of co-EO or HDPE because
the functional groups are too sterically hindered to enter in the
crystalline region of the polymer [25,64].
Fig. 13. Absorbance vs. irradiation time of ThiO-TEMPO (acetonitrile solution) and
HDPE-g-(ThiO-TEMPO) polymer film. Irradiation wavelength 366 nm.
Fig. 14. First order kinetic plot for the trans-cis isomerization of AzO-TEMPO in
acetonitrile solution, (co-EO)-g-(AzO-TEMPO) and HDPE-g-(AzO-TEMPO).
The more rigid and crystalline HDPE environment has the role to
disfavor and to slow down the formation of the less kinetically
stable and more hindered cis isomer, whereas when the chromo-
phores are grafted to the less rigid matrix (co-EO) the effect is less
evident. On the basis of these considerations, the slower formation
of the cis isomer embedded in the more rigid polymer matrix can
be attributed to the fact that the molecules experienced a more
strained conformation accounting for a general slowing down of
the process [25,26]. The thermal back-isomerization from cis to
trans isomer was observed for all our samples; in the case of HDPE-
Fig. 15. First order kinetic plots for the trans-cis isomerization of ThiO-TEMPO in
acetonitrile solution and HDPE-g-(ThiO-TEMPO).
conformation of the polymer chains is necessary to accommodate
the cis isomer. The polymers that are more flexible are able to adapt
more rapidly to the size and shape of the cis isomer upon its for-
mation from the trans form. If the polymer chain rearrangement
occurs on the time scale associated with the photo-isomerization of

376
F. Cicogna et al. / Polymer 82 (2016) 366e377
Table 5
Water contact angle of pristine polymers and functionalized polymers recorded before and after irradiation at 366 nm.
Sample
Contact angle (ꢀ)
Contact angle (ꢀ) after irradiation at 366 nm
HDPE
co-EO
99.1 2.2
98.2 2.1
96.7 1.1
98.3 2.6
97.8 0.8
e
e
HDPE-g-(AzO-TEMPO)
HDPE-g-(ThiO-TEMPO)
(co-EO)-g-(AzO-TEMPO)
93.3 1.5
94.8 2.1
95.1 1.7
g-(ThiO-TEMPO) 5e6 hs are necessary to recover about 80% of the
starting absorbance, whereas in the case of HDPE-g-(AzO-TEMPO)
the same conversion requires about 30 h. This comparison con-
firms that, as occurred in solution, ThiO-TEMPO isomerizes quickly
than AzO-TEMPO, also when it is grafted and embedded in a
polymer matrix. Comparison between the thermal back-
isomerization of AzO-TEMPO grafted to HDPE or co-EO evidenced
that the process is slower in co-EO (60% of the trans form of AzO-
TEMPO grafted to co-EO is recovered in 30 hs from the cis form)
confirming that the less rigid matrix allows for the formation, upon
irradiation, of a less strained cis isomer that slowly rearranges to
the more thermodynamically stable trans isomer.
were likely localized in the amorphous region, or at least at the
interfacial region between amorphous and crystalline phase,
polymers having higher crystallinity are more rigid. On the other
hand, the isomerization of the azo-aromatic moiety is strongly
dependent on the conformational mobility of the chromophore and
of the polymer chains, so the process can be hindered by a less
flexible matrix as occurred in the case of HDPE. Interestingly, the
surface of the functionalized polymers shows a change in the
wettability properties after isomerization of the chromophores.
The functional POs prepared by NRC reaction and described in this
paper can be classified as stimuli responsive materials able to
respond to light irradiation. Finally the NRC reaction can be
considered a feasible functionalization procedure by which it is
possible to plan the preparation of POs bearing specific designed
functionalities.
Finally, the water contact angle on the surface of the pristine
polymers and of the functionalized polymer was determined before
and after photo-isomerization from the trans to the cis rich PSS366
.
The aim is to highlight the change of the surface wettability caused
by the isomerization of the azo-aromatic group [1]. Results
(Table 5) evidenced that the radical functionalization of the matrix
does not change the wettability of the polymer matrix whereas,
after isomerization, a small decrease of the contact angle values is
observed. This behavior is in agreement with the fact that the trans
isomer has no dipole moment, whereas the dipole moment of the
non planar cis isomer is 3D [11]. The differences between the values
reported in the Table 5 are small, but significant if related to the
functionalization degree. Finally, the more polar ThiO-TEMPO
seems to cause a larger effect than AzO-TEMPO confirming the
potential usefulness of this group and of the NRC reaction for the
preparation of smart polyolefins.
Acknowledgments
We thank Dr. A. Raffaelli (CNR-IBF Pisa) for the acquisition of
mass-spectra, Dr. C. Pinzino (CNR-ICCOM SS Pisa) for the acquisition
of the EPR spectra, and Prof. G. Cicogna (Pisa University) for his
contribute to calculation. This research was funded by the National
Project ‘New Polymer Systems with Electric and Optical Function-
alities via Nano and Micro Adhesive Dispersion to Produce Mate-
rials and Devices for Smart Applications’ POLOPTEL 2011e2014, La
Fondazione CARIPISA conv. 167/09, and by MIUR under the pro-
gram FIRB 2010-Futuro in Ricerca, project title ‘GREENERꢂTowards
multifunctional, efficient, safe and stable “green” bio-plastics based
nanocomposites of technological interest via the immobilization of
functionalized nanoparticles and stabilizing molecules’ (Project
code: RBFR10DCS7).
4. Conclusion
The preparation of thermoplastic polymers, like POs, showing
optical responses is still an open question. The dye-POs physical
mixtures based on PE or PP generally show phase separation during
casting and drying due to the incompatibility between the phases
that limits the miscibility. To overcome these problems, in this
paper, chromophoric moieties able to isomerize under UV irradia-
tion were covalently grafted on PE samples. The NRC reaction of
AzO-TEMPO or ThiO-TEMPO with PO macroradicals was here
exploited to prepare functional HDPE and co-EO samples. The
functionalized polymers obtained by this procedure showed photo-
physical behavior similar to those observed for the low molecular
weight RO-TEMPO derivatives, thus suggesting the complete
transfer of the optical properties of the chromophores to the
polymer matrices. The two azo-aromatic derivatives, both as a free
radical and grafted moieties, exhibited different photo-physical
properties associable to the nature of the aromatic ring
substituted to the N]N double bond. In particular, it is observed
that both cis isomers have higher extinction coefficient than the
trans isomers in the UV region allowing to control the isomeriza-
tion process by modulation of the irradiation wavelength.
References
[1] N. Wagner, P. Theato, Polymer 55 (2014) 3436e3453.
[2] F. Ciardelli, G. Ruggeri, A. Pucci, Chem. Soc. Rev. 42 (2013) 857e870.
[3] H. Mengab, G. Li, J. Mater Chem. A 1 (2013) 7838e7865.
[4] A. Emoto, E. Uchida, T. Fukuda, Polymer 4 (2012) 150e186.
[5] A.A. Beharry, G.G. Woolley, Chem. Soc. Rev. 40 (2011) 4422e4437.
[6] M. Saphiannikova, V. Toshevikov, J. Ilnytskyi, Nonlin Opt. Quant. Opt. 41
(2010) 27e57.
[7] J.-M. Schumers, C.-A. Fustin, J.-F. Gohy, Macromol. Rapid Commun. 31 (2010)
1588e1607.
[8] K.M. Lee, H. Koerner, R.A. Vaia, T.J. Bunning, T.J. White, Macromolecules 43
(2010) 8185e8190.
[9] F. Liu, M.W. Urban, Prog. Polym. Sci. 35 (2010) 3e23.
[10] A. Natansohn, P. Rochon, Chem. Rev. 102 (2002) 4139e4175.
[11] G.S. Kumar, D.C. Neckers, Chem. Rev. 89 (1989) 1915e1925.
[12] T. Yoshida, S. Kanaoka, S. Aoshima, J. Polym. Sci. Part A Polym. Chem. 43
(2005) 4292e4297.
[13] F.J. Pavinatto, J.Y. Barletta, R.C. Sanfelice, M.R. Cardoso, D.T. Balogh,
C.R. Mendonça, O.N. Oliveira Jr., Polymer 50 (2009) 491e498.
[14] D. Kuckling, W. Urban, in: M.W. Urban (Ed.), Handbook of Stimuli Responsive
Materials, Wiley-VCH Verlag, GmbH & Co. KGaA, Weinheim, Germany, 2011,
pp. 1e26.
[15] L. Ding, L. Zhang, H. Han, W. Huang, C. Song, M. Xie, Y. Zhang, Macromolecules
42 (2009) 5036e5042.
[16] N.M. Franssen, J.N.H. Reek, B. de Bruin, Chem. Soc. Rev. 42 (2013) 5809e5832.
[17] N.K. Boaen, M.A. Hillmyer, Chem. Soc. Rev. 34 (2005) 267e275.
[18] E. Passaglia, S. Coiai, F. Cicogna, F. Ciardelli, Polym. Int. 63 (2014) 12e21.
[19] K.T. Lee, J.M. Goddard, J.H. Hotchkiss, Packag Technol. Sci. 22 (2009) 139e150.
Moreover, the differences in terms of isomerization kinetics and
isomerization efficiency observed between the free and grafted
chromophores were attributed to the matrix effect and particularly
to the polymer crystallinity. Indeed, although the functional groups

F. Cicogna et al. / Polymer 82 (2016) 366e377
377
[20] C. Wang, R.G. Weiss, Macromolecules 32 (1999) 7032e7039.
[21] N. Sarkar, S. Bhattacharjee, S. Sivaramor, Langmuir 13 (1997) 4142e4149.
[22] M.R. Sanchis, V. Blanes, M. Blanes, D. Garcia, R. Balart, Eur. Polym. J. 42 (2006)
1558e1568.
[23] N. Delorme, J.-F. Bardeau, D. Nicolas-Debarnot, A. Bulou, F. Poncin-Epaillard,
Langmuir 19 (2003) 5318e5353.
[43] H. McNab, L.C. Monahan, J. Chem. Soc. Perkin Trans I (1989) 419e424.
[44] C.R. Moylan, B.J. McNelis, L.C. Nathan, M.A. Marques, E.L. Hermstad,
B.A. Brichler, J. Org. Chem. 69 (2004) 8239e8243.
[45] H. Hagemam, R.G. Snyder, Macromolecules 22 (1989) 3600e3606.
[46] D. Daoust, S. Bebelman, N. Chaupart, R. Legras, J. Devaux, J. Costa, J. Appl.
Polym. Sci. 75 (2000) 96e106.
[24] D.E. Bergbreiter, M.H. Hein, K.J. Huang, Macromolecules 22 (1989)
4650e4652.
[47] Structural, thermal and photo-physical characterizations of azo-aromatic
TEMPO derivatives before and after their grafting to polyolefins. F. Cicogna,
I. Domenichelli, S. Coiai, F. Bellina, M. Lessi, R. Spiniello, E. Passaglia. Data in
Brief submitted.
[48] A.V. Zemskov, G.N. Rodinova, Yu G. Tuchin, V.V. Karpov, J. Appl. Spectr. 49
(1988) 1020e1024.
[25] C. Wang, R.G. Weiss, Macromolecules 36 (2003) 3833e3840.
[26] N. Sarkar, A. Sarkar, S. Sivaram, J. Appl. Polym. Sci. 81 (2001) 2923e2928.
ꢁ
[27] E. Rivera, del P. Carreon-Castro M, R. Salazar, G. Huerta, C. Becerril, L. Rivera,
Polymer 48 (2007) 3420e3428.
[28] F. Cicogna, S. Coiai, E. Passaglia, I. Tucci, L. Ricci, F. Ciardelli, A. Batistini,
J. Polym. Sci. Part A Polym. Chem. 49 (2011) 781e795.
[29] F. Cicogna, S. Coiai, C. Pinzino, F. Ciardelli, E. Passaglia, React. Funct. Polym. 72
(2012) 695e702.
[30] F. Cicogna, S. Coiai, P. Rizzarelli, S. Carroccio, C. Gambarotti, I. Domenichelli,
C. Yang, N.Tz. Dintcheva, G. Filippone, C. Pinzino, E. Passaglia, Polym. Chem. 5
(2014) 5656e5667.
[49] U. Georgi, P. Reichenbach, U. Oertel, L.M. Eng, B. Voit, React. Funct. Polym. 72
(2012) 242e251.
[50] G. Zimmerman, L.Y. Chow, U.J. Paik, J. Am. Chem. Soc. 80 (1958) 3528e3531.
[51] A. Yabe, Y. Kawabata, H. Niino, M. Tanaka, A. Ouchi, H. Takahashi, S. Tamura,
W. Tagaki, H. Nakahara, K. Fukuda, Chem. Lett. 17 (1988) 1e4.
[52] T. Seki, K. Tanaka, K. Ichimura, Poly. J. 30 (1998) 646e652.
[53] R.M. Parker, J.C. Gates, H.L. Rogers, P.G.R. Smith, M.C. Grossel, J. Mater. Chem.
20 (2010) 9118e9125.
[31] I. Domenichelli, S. Coiai, F. Cicogna, C. Pinzino, E. Passaglia, Polym. Int. 64
(2015) 631e640.
[54] E.J. Fischer, Phys. Chem. 71 (1967) 3704e3706.
[32] R.G. Weiss, V. Ramamurthy, G.S. Hammond, Acc. Chem. Res. 26 (1993)
530e536.
[55] G. Gauglitz, S.J. Hubig, Photochem. 15 (1981) 255e257.
[56] H.J. Rau, Photochem. 26 (1984) 221e225.
[33] K. Yoshii, S. Machida, K. Horie, J. Polym. Sci. Part B Polym. Phys. 38 (2000)
3098e3105.
[57] G.J. Gauglitz, Photochem. 5 (1976) 41e47.
[58] H. Rau, Photoisomerization of azobenzenes, in: S. Zouheir, W. Knoll (Eds.),
“Photoreactive Organic Thin Films”, Academic Press, 2002, pp. 3e104.
[59] A. Airinei, N. Fifere, M. Homocianu, C. Gaina, V. Gaina, B.C. Simionescu, Int. J.
Mol. Sci. 12 (2011) 6176e6193.
[60] M. Maafi, R.G. Brown, Photochem. Photobiol. Sci. 7 (2008) 1360e1372.
[61] L. Calucci, F. Cicogna, C. Forte, Phys. Chem. Chem. Phys. 15 (2013)
15584e15594.
[62] W. Gu, A.J. Hill, X. Wang, C. Cui, R.G. Weiss, Macromolecules 33 (2000)
7801e7811.
[63] C. Rosales, L. Marquez, J. Gonzalez, R. Perera, B. Rojas, M. Vivas, Polym. Eng.
Sci. 36 (1996) 2247e2252.
[34] C. Barrett, A. Natansohn, P. Rochod, Chem. Mater. 7 (1995) 899e903.
[35] J.C. Victor, J.M. Torkelson, Macromolecules 20 (1987) 2241e2250.
[36] L. Lamarre, C.S.P. Sung, Macromolecules 16 (1983) 1729e1736.
[37] W.-C. Yu, C.S.P. Sung, Macromolecules 21 (1988) 365e371.
[38] H.M.D. Bandara, S.C. Burdette, Chem. Soc. Rev. 41 (2012) 1809e1825.
[39] P.J. Coelho, L.M. Carvalho, J.C.V.P. Moura, M.M.M. Raposo, Dyes Pigm. 82
(2009) 130e133 and reference therein.
[40] P.J. Coelho, C.M. Sousa, M.C.R. Castro, A.M.C. Fonseca, M.M.M. Raposo, Opt.
Mater. 35 (2013) 1167e1172.
[41] A.D. Towns, Dyes Pigm. 42 (1999) 3e28.
[42] S. Nakatsuji, M. Fujino, S. Hasegawa, H. Akutsu, J. Yamada, V.S. Gurman, A
Kh Vorobiev, J. Org. Chem. 72 (2007) 2021e2029.
[64] P. Phillips, J. Chem. Rev. 90 (1990) 425e436.