An analysis of the structural, thermal and optical characteristics as well as the electrical resistivity of tert-butyldiphenylsilyl substituted poly(norbornene-dicarboximide)s

Article abstract of DOI:10.1016/j.polymer.2014.11.043

A sequence of well controlled tert -butyldiphenylsilyl substituted poly(norbornene-dicarboximide)s with ascending molecular weights have been prepared using the Grubbs 1st generation catalyst in anhydrous chloroform. The kinetics of the polymerization was examined using nuclear magnetic resonance spectroscopy and gel permeation chromatography. By decreasing the Grubbs catalyst concentration, the polymer molecular weights increased linearly. The glass transition temperatures and thermal decomposition temperatures initially increased with polymer molecular weight, then reached a plateau. Regardless of molecular weight the polydispersities remained narrow. The polymers exhibited a predominantly trans microstructure and matrix-assisted laser desorption/ionization mass spectrometry was utilized to determine polymer end groups and to calculate the absolute molecular weights. The residual ruthenium content of the polymers was quantified using inductively coupled plasma mass spectrometry and the influence on the resistivity, refractive indices and thin film optical transmittance was evaluated.

Full text of DOI:10.1016/j.polymer.2014.11.043

Polymer 56 (2015) 189e198
Contents lists available at ScienceDirect
Polymer
journal homepage: www.elsevier.com/locate/polymer
An analysis of the structural, thermal and optical characteristics as
well as the electrical resistivity of tert-butyldiphenylsilyl substituted
poly(norbornene-dicarboximide)s
Andrew M. Spring a, Daisuke Maeda b, Masaaki Ozawa b, Keisuke Odoi b, Feng Qiu a,
,
*
Kazuhiro Yamamoto a, Shiyoshi Yokoyama a
a Institute for Materials Chemistry and Engineering, Kyushu University, 6-1 Kasuga-koen, Kasuga City, Fukuoka 815-8580, Japan
b Nissan Chemical Industries, Ltd, 2-10-1 Tsuboi Nishi, Funabashi, Chiba 274-8507, Japan
a r t i c l e i n f o
a b s t r a c t
Article history:
A sequence of well controlled tert-butyldiphenylsilyl substituted poly(norbornene-dicarboximide)s with
ascending molecular weights have been prepared using the Grubbs 1st generation catalyst in anhydrous
chloroform. The kinetics of the polymerization was examined using nuclear magnetic resonance spec-
troscopy and gel permeation chromatography. By decreasing the Grubbs catalyst concentration, the
polymer molecular weights increased linearly. The glass transition temperatures and thermal decom-
position temperatures initially increased with polymer molecular weight, then reached a plateau.
Regardless of molecular weight the polydispersities remained narrow. The polymers exhibited a pre-
dominantly trans microstructure and matrix-assisted laser desorption/ionization mass spectrometry was
utilized to determine polymer end groups and to calculate the absolute molecular weights. The residual
ruthenium content of the polymers was quantified using inductively coupled plasma mass spectrometry
and the influence on the resistivity, refractive indices and thin film optical transmittance was evaluated.
© 2014 Elsevier Ltd. All rights reserved.
Received 8 October 2014
Received in revised form
13 November 2014
Accepted 18 November 2014
Available online 22 November 2014
Keywords:
ROMP
Poly(NDI)s
Well-controlled
1. Introduction
optical transparency. When incorporated into devices, polymers
are often exposed to high operating temperatures and voltages for
Poly(norbornene-dicarboximides) are a well-researched type of
living polymer which were originally prepared by ring opening
metathesis polymerization (ROMP) in 1992 [1]. Since then, the
polymers have been functionalized with numerous active groups,
tailoring them to a varied range of applications [2e5]. As a conse-
quence of their outstanding thermal features, superior optical
transparency, effective film forming characteristics, low moisture
absorption and robust mechanical properties they have been uti-
lized as both the passive and active components in organic elec-
tronic and optical applications [6e8]. A central requirement for
their utilization is a high purity. Polymers must be free from re-
sidual catalyst or other fragments associated with their synthesis as
if impure a degradation in their thin film optical transparency,
stability and a lower electrical resistivity may result. Important
physical characteristics include the thermal decomposition tem-
prolonged time periods, in such cases the presence of residual
metal catalyst may reduce the device lifetime. Extensive efforts
have been made to reduce the concentration of residual ruthenium
in ROMP produced polymers, including by aqueous extraction [9],
silica gel/activated carbon [10], the use of small molecule ruthe-
nium scavengers and polymer supported catalysts [11].
In this manuscript we have synthesized a tert-butyldiphenylsilyl
substituted norbornene-dicarboximide monomer. The tert-butyl-
diphenylsilyl group is a well-known bulky substituent in organic
synthesis as well as being an excellent protecting group for alco-
hols, which can be easily removed by treatment with lithium ace-
tate acting as a bifunctional Lewis acideLewis base catalyst [12] or
by treatment with 1,8-diazabicyclo[5.4.0]undec-7-ene (DBU) [13].
The reactivity of this monomer was gauged by kinetic experiments
which monitored the increase in polymer proton nuclear magnetic
resonance (1H NMR) signals and the reduction of monomer 1H NMR
signals. Gel permeation chromatography (GPC) was utilized to
follow the growth in polymer molecular weight. After optimization
of the polymerization conditions, a sequence of polymers was
perature (Td), glass transition temperature (Tg), resistivity (r) and
* Corresponding author.
E-mail address: s_yokoyama@cm.kyushu-u.ac.jp (S. Yokoyama).
http://dx.doi.org/10.1016/j.polymer.2014.11.043
0032-3861/© 2014 Elsevier Ltd. All rights reserved.

190
A.M. Spring et al. / Polymer 56 (2015) 189e198
prepared with ascending molecular weights. This enabled an
assessment of the living nature of the polymerization, the influence
of the molecular weight on the thermal properties and an investi-
gation of the impact of residual ruthenium on the optical properties
and the resistivity. The polymer sequence was characterized using
1H NMR, GPC and matrix-assisted laser desorption/ionization mass
spectrometry (MALDI-MS) to determine their absolute molecular
weights and end groups. The molecular weights obtained by
MALDI-MS were very similar to those measured by GPC and the
ruthenium content of the polymers was quantified using induc-
tively coupled plasma mass spectrometry (ICP-MS). The impact of
this residual ruthenium on the refractive indices, thin film optical
transmittance and electrical resistivity was evaluated. The syn-
thesis of the precursors 1e4, monomer (5) and the polymer series
(6aej) is presented in Scheme 1.
with the samples being encapsulated in a standard aluminium DSC/
TGA pan. To ensure a consistent measurement each sample was run
twice, with the second run being quoted in the manuscript. For DSC
analysis the samples were heated from 30 ꢀC to 250 ꢀC and for TGA
measurement the samples were heated from 30 ꢀC to 500 ꢀC both
under a nitrogen atmosphere. TGA measurements quoted corre-
spond to the temperature at which 5% mass loss occurred. The
polymer molecular weights were determined by gel permeation
chromatography (GPC) calibrated against narrow polydispersity
index polystyrene standards using a Shodex GPC K-804L column on
a JASCO LC2000 instrument at 40 ꢀC in chloroform with a flow rate
of 1 mL/min. Positive-ion MALDI-MS analysis was undertaken on a
JEOL JMS-S3000 Spiral TOF instrument in Spiral and Linear modes.
The Spiral mode was used for accurate end group analysis. The
Linear mode was used to measure the complete distribution. In
both cases the samples were prepared by mixing the matrix trans-
2-[3-(4-tert-butylphenyl)-2-methyl-2-propenylidene]malononi-
trile (DCTB) in THF (30 mg/mL), polymer in THF (30 mg/mL) and
salt, NaI in methanol (30 mg/mL) in a ratio of 1:1:1. The residual
ruthenium content of the polymers was determined by Inductively
Coupled Plasma Mass Spectrometry (ICP-MS) on an Agilent 7500cs
instrument. The polymer sample (50 mg) was decomposed using a
microwave decomposing device and was added to a 5 g mixture of
HCl and HNO3. UVevis transmission (400e1000 nm) measure-
2. Experimental
2.1. Techniques
1H and 13C nuclear magnetic resonance (NMR) spectra were
recorded on a JNM-LA400 spectrometer at 400 MHz and 125 MHz
frequencies respectively in deuterated chloroform (CDCl3) or
dimethylsulfoxide (DMSO) with tetramethylsilane (TMS) as the
internal standard. Thermogravimetric analysis (TGA) and differen-
tial scanning calorimetry (DSC) were performed at a heating rate of
10 ꢀC/min on an SII-TG/DTA 6200 and SII-DSC 6220 respectively,
ments were obtained on
a Jasco V-670 spectrophotometer.
Refractive index measurements were obtained using a Sairon
Technology Inc. Prism Coupler SPA-4000 on slide glass at 1310 and
Scheme 1. Synthesis of compounds (1), (2), (3), (4), (5) and the polymer sequence (6aej).

A.M. Spring et al. / Polymer 56 (2015) 189e198
191
1550 nm. CHN elemental analysis was performed using a Yanako
CHN corder MT-6. Film thickness was determined by using a KLA
Tencor MEMS surface profiler and gold electrodes were sputtered
onto ITO glass using a magnetron sputter at 40 mA for 6 min.
was obtained by precipitation into water. The resulting white solid
was filtered off, washed with excess water and dried in a vacuum
oven at 60 ꢀC for 12 h. The crude material was then recrystallized
from a mixture of boiling ethanol and THF to give a white crystal-
line solid in a yield of 39%. 1H NMR (400 MHz, DMSO)
d: 8.94 (bs,
2.2. Materials
1H), 6.25 (d, J ¼ 8 Hz, 2H), 6.06 (d, J ¼ 8 Hz, 2H), 5.56 (s, 2H), 2.00 (s,
2H), 1.71 (s, 2H), 0.65 (d, J ¼ 9 Hz, 1H), 0.60 (d, J ¼ 9 Hz, 1H) p.p.m.
Dicyclopentadiene, 4-aminophenol and tert-butyl(chloro)
diphenylsilane were purchased from Tokyo Chemical Industry Co
Ltd (TCI). Maleic anhydride, tetrahydrofuran, methanol, anhydrous
toluene, triethylamine, o-dichlorobenzene, ethyl vinyl ether, and
chlorobenzene were obtained from Kanto Chemical Co Inc. and
used without prior purification. Chloroform was obtained from
Kanto Chemical Co Inc. and was dried under calcium hydride and
distilled onto microwave oven dried (5 min) molecular sieves. The
Grubbs 1st generation initiator was purchased from Sigma-Aldrich
Co. and was used as received.
13C NMR (125 MHz, DMSO)
d: 178.17, 158.35, 138.76, 129.13, 124.35,
116.37, 48.25, 45.84, 43.52 p.p.m. High-resolution mass spectrom-
etry (Electron Ionization, EI) calculated for 15H13NO3: m/z
C
255.2730; found: m/z 255.0898. Elemental analysis, calculated; C:
70.58%, H: 5.13%, N: 5.49%, found; C: 70.20%, H: 5.07%, N: 5.50%.
2.3.4. Synthesis of tert-butyldiphenylsilyl substituted monomer (5)
N-Phenol-norbornene-5,6-dicarboximide
(27.00
g,
105.77 mmol) and imidazole (19.57 g, 287.45 mmol) were dissolved
in DMF (150 mL). To this stirred solution tert-butyl(chloro)diphe-
nylsilane (34.96 g, 134.47 mmol) was added drop wise. After the
complete addition, the mixture was stirred for 3 h at room tem-
perature. The crude product was obtained by precipitation into
water and filtration. The white solid was washed with excess water
and dried in a vacuum oven overnight at 60 ꢀC and then recrys-
tallized from a mixture of boiling ethanol and THF to give the pure
2.3. Monomer synthesis
2.3.1. Synthesis of exo-norbornene-5,6-dicarboxylic anhydride (2)
Initially norbornene-5,6-dicarboxylic anhydride (1) was syn-
thesized in an exo/endo mixture. The pure exo form (2) was ob-
tained by three recrystallizations from boiling chlorobenzene. The
exact procedure employed was as follows; into a solution of maleic
anhydride (188.24 g, 1.92 mol) in o-dichlorobenzene (200 mL) at
200 ꢀC, dicyclopentadiene (128.66 g, 0.96 mol) was firstly melted
by heating to 50 ꢀC and then added drop-wise. The mixture was
heated to reflux for 1.5 h (200 ꢀC), and then allowed to cool slowly
to room temperature. After 12 h, the exo/endo mixture crystallized
out of the o-dichlorobenzene. This white solid was filtered off and
the remaining o-dichlorobenzene was discarded. The obtained
white crystalline solid was recrystallized multiple times from
boiling chlorobenzene to obtain the exo isomer (2) as a white
crystalline material in a yield of approximately 45%. 1H NMR
material in a yield of 44%. 1H NMR (400 MHz, CDCl3)
d: 7.71 (d,
J ¼ 8 Hz, 4H), 7.39 (m, 6H), 6.98 (d, J ¼ 8 Hz, 2H), 6.82 (d, J ¼ 8 Hz,
2H), 6.31 (s, 2H), 3.35 (s, 2H), 2.79 (s, 2H), 1.56 (s, 2H), 1.09 (s, 9H)
p.p.m. 13C NMR (125 MHz, CDCl3)
d: 177.69, 156.02, 138.36, 135.87,
132.86, 130.44, 128.28, 127.64, 120.47, 48.05, 46.10, 43.21, 26.76,
19.73 p.p.m. High-resolution mass spectrometry (Electron Ioniza-
tion, EI) calculated for C31H31NO3Si: m/z 493.6780; found: m/z
494.2146. Elemental analysis, calculated; C: 75.42%, H: 6.33%, N:
2.84%, found; C: 75.28%, H: 6.24%, N: 2.79%.
2.3.5. Polymerization kinetics
Kinetic studies were performed on the polymerization of
monomer (5) in order to determine the optimum reaction time for
monomer conversion. The investigation was carried out as follows;
the monomer (5) (0.5 g, 1.01 mmol) was charged into 7 Radley's
Carousel tubes. To each of these tubes was added anhydrous
chloroform (10 mL), the solutions were stirred at room temperature
for 10 min. The Grubbs 1st generation catalyst (n ¼ 100, 8.33 mg,
1.01 ꢁ 10ꢂ5 mol) was charged into 7 flasks (10 mL), dissolved in
anhydrous chloroform (1 mL) and quickly transferred to each of the
reaction tubes. The polymerizations were then quenched using
ethyl vinyl ether (1 mL) at intervals of 12 min, 24 min, 36 min,
48 min, 60 min, 120 min and 180 min. The quenching reaction was
allowed to proceed for 30 min in each case and then the solvent and
ethyl vinyl ether were evaporated under vacuum. The 7 kinetic
samples were then analyzed by 1H NMR and GPC, to determine at
what reaction time monomer conversion occurred.
(400 MHz, CDCl3) d: 6.32 (s, 2H), 3.43 (s, 2H), 2.99 (s, 2H), 1.66 (d,
J ¼ 11 Hz, 1H), 1.41 (d, J ¼ 11 Hz, 1H) p.p.m. 13C NMR (125 MHz,
CDCl3)
d: 172.05, 138.22, 49.05, 47.13, 44.35 p.p.m. Elemental
analysis, calculated; C: 65.85%, H: 4.91%, N: 0.00%, found; C: 65.83%,
H: 4.83%, N: 0.00%.
2.3.2. Synthesis of N-phenol-amic acid (3)
exo-Norbornene-5,6-dicarboxylic anhydride (2) (33.20 g,
0.20 mol) was dissolved in tetrahydrofuran (200 mL). To this stirred
solution, 4-aminophenol (20.96 g, 0.19 mol) was added slowly.
After addition the mixture was refluxed for 4 h and was then
allowed to cool overnight after which a white precipitate formed.
The precipitate was filtered off and washed with an excess of cold
tetrahydrofuran. Drying in a vacuum oven overnight at 60 ꢀC gave
the pure material in a yield of 78%. The solubility of this compound
in chloroform was poor, therefore its NMR spectra were acquired in
DMSO.
2.3.6. Polymer synthesis
1H NMR (400 MHz, DMSO)
d
: 11.10 (bs, 1H), 8.89 (s, 1H), 8.31 (bs,
The polymers were synthesized in one batch using a Radley's
Carousel 12 Plus parallel synthesizer. The procedure used was as
follows; the monomer (5) (2.0 g, 4.05 mmol) was transferred to 10
Radley's Carousel tubes and the solid stirred under vacuum for
10 min. After this period the tubes were backfilled with nitrogen
and then nitrogen degassed (30 min) anhydrous chloroform
(10 mL) was added by syringe. The monomer solutions were stirred
for an additional 10 min at room temperature. The Grubbs 1st
generation catalyst was charged into 10 flasks (10 mL) in varying
quantities to prepare the polymer series, (n ¼ 10, 333 mg,
4.05 ꢁ 10ꢂ4 mol), (n ¼ 20, 166 mg, 2.03 ꢁ 10ꢂ4 mol), (n ¼ 30, 111 mg
1.35 ꢁ 10ꢂ4 mol), (n ¼ 40, 83 mg, 1.01 ꢁ 10ꢂ4 mol), (n ¼ 50, 66 mg,
8.10 ꢁ 10ꢂ5 mol), (n ¼ 60, 55 mg, 6.75 ꢁ 10ꢂ5 mol), (n ¼ 70, 47 mg,
1H), 6.53 (d, J ¼ 9 Hz, 2H), 5.86 (d, J ¼ 9 Hz, 2H), 5.43 (s, 2H), 2.15 (s,
1H), 2.03 (s, 1H), 1.81 (d, J ¼ 9 Hz, 1H), 1.69 (m, 1H), 1.57 (d, J ¼ 9 Hz,
1H), 1.45 (d, J ¼ 9 Hz, 1H), 0.47 (d, J ¼ 9 Hz, 1H) p.p.m. 13C NMR
(125 MHz, DMSO) d: 175.38, 171.70, 154.13, 139.21, 139.10, 132.14,
122.08, 115.84, 48.53, 47.88, 47.23, 46.00, 44.65 p.p.m. High-
resolution mass spectrometry (Electron Ionization, EI) calculated
for C15H15NO4: m/z 273.1001; found: m/z 273.1003.
2.3.3. Synthesis of N-phenol-norbornene-5,6-dicarboximide (4)
N-Phenol-amic acid (3) (75.00 g, 274.44 mmol) and triethyl-
amine (41.65 g, 411.66 mmol) were dissolved in dry DMF (300 mL)
and heated to 120 ꢀC for 3 h. After the reaction the crude product

192
A.M. Spring et al. / Polymer 56 (2015) 189e198
5.78 ꢁ 10ꢂ5 mol), (n ¼ 80, 41 mg 5.06 ꢁ 10ꢂ5 mol), (n ¼ 90, 37 mg,
4.50 ꢁ 10ꢂ5 mol), (n ¼ 100, 33 mg, 4.05 ꢁ 10ꢂ5 mol). The catalyst
was then dissolved in chloroform (1 mL) and quickly added to each
of the 10 reaction tubes by syringe. Each of the 10 polymerizations
were allowed to proceed for 180 min at room temperature, after
which the polymerizations were quenched by addition of an excess
of ethyl vinyl ether (1 mL). The quenching reaction was allowed to
proceed for 30 min at room temperature after which the solvent
and ethyl vinyl ether were evaporated under vacuum. The polymers
were then re-dissolved in the minimum amount of chloroform and
precipitated into stirred methanol. To ensure maximum catalyst
residue removal, this procedure was repeated twice, initially the
filtrate was a light brown colour, indicating the presence of the
catalyst residue, after the 2nd dissolution and precipitation the
filtrate became clear. Finally the pure polymers (6aej) were ob-
tained by filtration and dried in a vacuum oven at 60 ꢀC for 12 h.
Polymer yields were all high (>80%) reflecting monomer conver-
sion, at lower molecular weights however it was rather difficult to
collect all of the very fine precipitate which easily passed through
the filter, thus reducing the yield in these cases. As the structure of
the polymers was unchanged with molecular weight, the following
may serve as an example for all polymers.
(4) with tert-butyl(chloro)diphenylsilane in the presence of imid-
azole. A high monomer purity was achieved by multiple re-
crystallizations from a mixture of boiling ethanol and THF. Fig. 1G
displays the 1H NMR spectrum and Fig. 1H the 13C NMR spectrum of
monomer (5). Key environments that can be observed in Fig. 1G are
the multiplet at 7.71 p.p.m which corresponds to the four tert-
butyldiphenylsilyl aromatic protons closest to the silicon (a) and
the multiplet at 7.39 p.p.m which corresponds to the remaining six
tert-butyldiphenylsilyl aromatic protons (b). The doublet at
6.98 p.p.m is generated by the two protons of the linking 1,4-
disubstituted benzene closest to the oxygen, which are mirrored
by a similar set of doublets at 6.82 p.p.m generated by the two
protons adjacent the nitrogen. Both doublets have a coupling
constant of 8 Hz confirming correct assignment. The singlet at
6.31 p.p.m corresponds to the two protons of the alkene. Finally the
singlet at 3.35 p.p.m integrating to two protons corresponds to the
bridging environments (f), the singlet at 2.79 p.p.m integrating to
two protons corresponds to the bridging environments (g). The
singlet at low chemical shift 1.56 p.p.m, integrating to two protons
corresponds to (h). Finally the large singlet at 1.09 p.p.m can be
ascribed to the protons of the tert-butyl group. The 13C NMR
spectrum in Fig. 1D indicates the presence of thirteen carbon sig-
nals. These are the five signals of the norbornene-dicarboximide
unit at 177.69 (1), 135.87 (3), 48.05 (10), 46.10 (11), 43.21 (12) and
the four signals of the 1,4-disubstituted benzene linker at 156.02
(2), 132.86 (5), 127.64 (6) and 120.47 (9) ppm. Finally there are four
signals which can be attributed to the terminal tert-butyldiphe-
nylsilyl group at 138.36 (4), 130.44 (6), 128.28 (7) and 19.73 (14)
ppm.
Polymer (6j): 1H NMR (400 MHz, CDCl3)
d: 7.69 (3.89H, m)cis/
trans, 7.37 (5.90H, m)cis/trans, 6.97 (1.34H, d, J ¼ 8 Hz)trans, 6.88
(0.33H, d, J ¼ 8 Hz)cis, 6.80 (1.37H, d, J ¼ 8 Hz)cis, 6.69 (0.31H, d,
J ¼ 8 Hz)cis, 5.75 (1.36H, m)trans, 5.52 (0.03H, m)cis, 3.45 (0.16H, m)
cis, 3.18 (0.06H, m)cis, 3.05 (1.37H, m)trans, 2.76 (1.20H, m)trans,
2.15 (0.48H, m)cis/trans, 1.63 (2.22H, m)cis/trans, 1.08 (9.00H, s) cis/
trans p.p.m. Yield ¼ 97%.
3. Results and discussion
3.2. Polymer synthesis and characterization
3.1. Monomer synthesis and characterization
To ascertain the optimum polymerization condition for the
synthesis of the polymer sequence (6aej), a kinetic investigation
of monomer (5) was executed. An identical quantity of monomer
(5) was charged into seven Radley's Carousel tubes. The solid was
stirred under vacuum and then the vessel filled with nitrogen.
Anhydrous chloroform was added to each vessel and the solid
allowed to dissolve. A solution of the Grubbs 1st generation
catalyst was added to each of the seven vessels. The mass of the
catalyst was kept constant (8.33 mg) to yield a polymer with 100
repeat units. The sequential polymerizations were quenched at
12 min, 24 min, 36 min, 48 min, 60 min, 120 min and 180 min by
the addition of ethyl vinyl ether. 1H NMR and GPC analysis was
used to monitor monomer conversion after each time period.
Fig. 2A shows a plot of the molecular weight against the reaction
time. The curve exhibits a steep gradient at short reaction times (6,
12, 24, 36, 48 min) and then begins to level at 60 min, with
monomer conversion being complete at 180 min. The inset table of
Fig. 2A shows that the molecular weight increase from 60 min to
180 min was relatively small (3 K) as compared to the increase
from 6 to 48 min (67 K). Despite the bulky tert-butyldiphenylsilyl
group, the reaction kinetics appear similar to those of a cyclohexyl
substituted norbornene-dicarboximide monomer, which under-
went monomer conversion after 120 min under the same condi-
tions [15,7]. Fig. 2C shows the key areas of the polymer 1H NMR
spectrum which allows the progress of the polymerization to be
followed. The singlet marked with a star (*) corresponds to the
alkene of monomer (5). The broad peak marked with a cross (þ)
corresponds to the trans vinylene of the polymer. As the poly-
merization progresses, the area of the peak attributed to the
monomer (*) is reduced while the area of the peak attributed to
the polymer (þ) increases. After 180 min, the area of the monomer
peak (*) is immeasurably small as compared to that of the polymer
Previous investigations by Yoon et al. have verified that endo-
norbornene-dicarboximides are considerably less reactive than
their exo counterparts due to steric hindrance around the alkene
[14]. The removal of the endo isomer from the mixture at the first
stages can facilitate a clean polymerization profile of the monomer.
Consequently exo-norbornene-5,6-dicarboxylic anhydride (2) was
purified carefully to ensure it was endo isomer free. Fig. 1A presents
an 1H NMR spectrum of the exo/endo mixture after the reaction
between cyclopentadiene and maleic anhydride. The proton envi-
ronments a to e have been assigned on the spectrum. Indicated
with a star (*) are the signals corresponding to the endo isomer and
marked with a cross (þ) are the signals corresponding to the exo
isomer. The exo/endo ratio is 1:0.65 in the as formed mixture. Fig.1B
displays an 1H NMR spectrum of the exo isomer after multiple re-
crystallizations from boiling monochlorobenzene. The signals
attributed to the endo isomer (*) are almost absent, with those of
the exo isomer predominating. The following step involved the
synthesis of N-phenol-amic acid (3). This molecule was acquired by
the slow addition of 4-aminophenol to a stirred solution of exo-
norbornene-5,6-dicarboxylic anhydride (2) in tetrahydrofuran. The
material was insoluble in chloroform, therefore the NMR spectra
were acquired in deuterated DMSO. Fig. 1C presents the 1H NMR
spectrum and Fig.1D the 13C NMR spectrum of this compound. Both
the NMR and high resolution mass spectrometry confirmed suc-
cessful synthesis and purification. Treatment of the amic acid (3)
with an excess of triethylamine afforded the monomer precursor N-
phenol-norbornene-5,6-dicarboximide (4). NMR analysis was
again accomplished in DMSO solution. Both NMR analysis shown in
Fig. 1E, F and elemental analysis confirmed a high purity. Finally the
desired tert-butyldiphenylsilyl substituted monomer (5) was syn-
thesized by the functionalization of the phenol group of precursor
(þ). The small peak labelled with a delta (
D) can be attributed to

A.M. Spring et al. / Polymer 56 (2015) 189e198
193
Fig. 1. 1H NMR spectra of exo/endo-norbornene-5,6-dicarboxylic anhydride (1) (A) and exo-norbornene-5,6-dicarboxylic anhydride (2) (B) measured in deuterated chloroform. 1
H
NMR (C) and 13C NMR (D) spectra of precursor (3) measured in deuterated DMSO. 1H NMR (E) and 13C NMR (F) spectra of N-phenol-norbornene-5,6-dicarboximide (4) measured in
deuterated DMSO and the 1H NMR (G) and 13C NMR (H) spectra of monomer (5) measured in deuterated chloroform.
the residual endo isomer. This peak is more prominent in the
180 min 1H NMR spectrum as the proportion of the exo isomer is
diminished, confirming that the endo isomer is inactive at room
temperature. Fig. 2B shows the molecular weight distribution
curves associated with the various quenching times. The distri-
butions are broad at short quenching times (6e12 min), after
which they become narrower as the polymer increases in length
(24e180 min).

194
A.M. Spring et al. / Polymer 56 (2015) 189e198
Fig. 2. The molecular weights from the kinetic investigation. Polymerizations were quenched at 6, 12, 24, 36, 48, 60, 120 and 180 min (A). The molecular weight distributions (B). 1
NMR spectra showing the reduction in the alkene peak of the monomer (*) and the increase in the trans vinylene peak of the polymer (þ) as reaction time is increased. Residual
unreacted endo isomer can also be observed ( ) (C).
H
D
In recent work it was determined that the polymerization of
cyclohexyl substituted norbornene-dicarboximides proceed by a
living mechanism. Strictly speaking this meant that the polymeri-
zation could be characterized as a chain polymerization from which
chain-transfer and chain-termination reactions were largely absent
[15,7]. Given that tert-buyldiphenysilyl is a much bulkier substitu-
ent than cyclohexyl, we also investigated the extent to which
polymers (6aej) could be described as living. IUPAC states that
essential criteria for living polymerizations are; complete and rapid
initiation, irreversible propagation steps and the absence of chain
termination and chain transfer [16]. If all these criteria are satisfied
by a polymer system, then there should be a linear dependence of
the number average molecular weight (Mn) on monomer conver-
sion, control of molecular weight and narrow molecular weight
distributions. To evaluate the extent to which these polymers were
living, a range of polymers of ascending molecular weights were
produced. The number of repeat units was varied from 10 to 100 by
adjusting the [monomer]/[Initiator] ratio. After allowing the poly-
merizations to proceed for 180 min, the living polymers were
quenched with an excess (1 mL) of ethyl vinyl ether and precipi-
tated multiple times from chloroform into methanol. Previous work
has established that the stereochemistry of the polymerization
Fig. 3. The annotated 1H NMR spectrum of the polymer (6j).

A.M. Spring et al. / Polymer 56 (2015) 189e198
195
depends on the ligands attached to the Grubbs initiator, with the
P(Cy)3 group of Grubbs 1 resulting in a predominantly trans
microstructure and the bulky N-heterocyclic carbene of Grubbs 2
giving an approximately 1:1 cis/trans mixture [7,15,17]. The 1H NMR
spectrum of polymer (6j) is displayed in Fig. 3. Clearly most signals
are a mirror of the monomer (5) spectrum in Fig.1G, however as the
polymers are composed of a cis/trans mixture, there are additional
signals. The multiplet at 7.96 p.p.m corresponds to the tert-butyl-
diphenylsilyl (a), the multiplet at 7.36 p.p.m corresponds to the
remaining tert-butyldiphenylsilyl protons (b). These signals are a
combination of both cis and trans. The doublets at 6.96 and
6.88 p.p.m correspond to the trans and cis aromatic protons (c). As
the polymers are predominantly trans, this peak is much larger
than the cis. The doublets at 6.80 and 6.69 p.p.m correspond to the
remaining trans and cis protons of the 1,4-disubstituted aromatic
(d). The doublets at 5.72 and 5.51 p.p.m can be assigned to the trans
and cis alkene protons (e). The two multiplets at 3.45 and 3.18 p.p.m
can be assigned to the cis bridging protons (f) and (g) respectively.
The trans bridging protons (f) and (g) appear much larger at 3.05
and 2.75 p.p.m. Finally the cis/trans mixture of (h) and (i) generate
multiplets at 2.13 and 1.56 p.p.m. Fig. 4A, displays the level of
control achieved using a plot of molecular weight (Mn) vs [Mono-
mer]/[Initiator] compared to the molecular weight expected from
calculation. In most cases the experimental values (circles) were
close to the calculated values (squares). The relationship between
molecular weight (Mn) and [Monomer]/[Initiator] clearly demon-
strates the living nature of the polymerization. By reducing the
initiator, the polymer molecular weight increases in a linear
fashion.
Fig. 4B displays the molecular weight distribution curves of the
resulting polymers. All polymers (6aej) appear well controlled as
their distributions are narrow. The peaks shift along the Log M axis
as would be expected. It can be concluded that the bulky tert-
butyldipheylsilyl substituent is no hindrance to a well-controlled
polymerization. The narrow PDIs (1.1e1.2) as shown in Table 1,
indicate that there is no evidence of chain termination steps.
To confirm the accuracy of the GPC measurements, polymers 6a
and 6b were analyzed by MALDI-MS. The values obtained by
MALDI-MS were found to be very similar to those measured by GPC
as can be observed in Fig. 5. Providing that the polymers have a
narrow molecular weight distribution (1.1e1.2), MALDI-MS can
allow an accurate and absolute determination of the polymer mo-
lecular weights, end groups and the repeat unit [18]. The presence
of an additional sequence of peaks in the MALDI-MS spectrum can
suggest the occurrence of chain termination reactions and cyclic
polymers. GPC analysis revealed that the PDI of polymers 6a and 6b
was both 1.2, making them viable for MALDI-MS analysis. The
spectra were acquired by using trans-2-[3-(4-tert-butylphenyl)-2-
methyl-2-propenylidene]malononitrile (DCTB) as the matrix, so-
dium iodide (NaI) as the salt and mixing with the polymer in a ratio
of 1:1:1 in THF on a JMS-S3000 Spiral-TOF instrument calibrated
against PEG standards. Operation in Spiral mode allowed an accu-
rate determination of polymer end groups, although this mode was
unable to acquire the full distribution at higher molecular weights.
In contrast, the Linear mode allowed the measurement of the
complete molecular weight distribution for polymers 6a and 6b. By
using this mode the Mn, Mw and PDIs were calculated and con-
trasted to values obtained from GPC. The MALDI calculated Mn were
6,100 and 9,700 for polymers 6a and 6b. This contrasted to GPC
measured values of 5,800 and 9,700 respectively. Fig. 5A shows the
Spiral MALDI-MS spectrum of polymer 6a, the calculated Mn was
4,200 which is significantly lower than that obtained by the Linear
mode (6,100). Fig. 5B shows the end group analysis of polymer 6a
using the Spiral mode spectrum, the values obtained match exactly
those expected from calculation, and furthermore the repeat unit of
493 is observed. Fig. 5C and D show the Linear mode MALDI-MS
spectra of polymers 6a and 6b respectively. In all cases there are
one set of peaks confirming the well-controlled nature of the
polymerization. There are no peaks attributed to cyclic polymers.
Table 2 presents an analysis of the expanded spectrum from
Fig. 5B. The polymer repeat units from 5 to 11 were analyzed.
Calculated m/z refers to the expected weight of the polymer chain
including end groups and sodium.
Although it has been verified previously that the Tg of
substituted
poly(norbornenedicarboximides)
depends
Table 1
The calculated molecular weights (calc Mw) compared with the experimental (Mn)
and PDI measured by GPC as well as the Tg and Td for the polymer series (6aej).
Polymer
Calc Mw (g molꢂ1
)
Mn (g molꢂ1
)
PDI
Tg (ꢀC)
Td (ꢀC)
6a
6b
6c
6d
6e
6f
6g
6h
6i
4,931
9,861
5,800
9,700
1.2
1.2
1.2
1.2
1.2
1.2
1.2
1.2
1.1
1.1
159
177
187
188
192
196
196
200
198
196
339
408
419
423
420
421
421
426
426
421
14,792
19,722
24,653
29,583
34,514
39,444
44,375
49,305
13,600
17,300
20,100
25,100
28,200
31,100
40,600
47,100
Fig. 4. Plots of the polymer molecular weight vs [monomer]/[initiator] for polymers
(6aej) (A). The molecular weight distributions for polymers (6aej) measured by GPC
(B).
6j

196
A.M. Spring et al. / Polymer 56 (2015) 189e198
Fig. 5. The Spiral MALDI-MS spectrum of polymer (6a). (A). An expansion of the spectrum from 2,500 to 6,000 m/z showing the repeat unit and the peak positions. (B) The Linear
MALDI-MS spectrum of polymer 6a (C). The Linear MALDI-MS spectrum of polymer 6b (D).
predominantly on the size of the substituent [19], it is also influ-
enced by the polymer molecular weight. However this dependence
is only significant when the molecular weights are low [15]. As can
be observed in Fig. 6A, which is a plot of Tg vs polymer molecular
weight, when the molecular weight was increased beyond
25,000 g molꢂ1 the Tg increase ceased. From 5,000 g molꢂ1 to
25,000 g molꢂ1 the Tg increased from 159 ꢀC to 200 ꢀC. This trend
was also observed for cyclohexyl substituted polymers [15]. Inter-
estingly the bulky tert-butyldiphenyl silyl group did not result in a
significant elevation in the polymer Tg as compared to cyclohexyl
substituted polymers at similar molecular weights [15]. At 10
repeat units the cyclohexyl polymer exhibited a Tg of 178 ꢀC as
compared to 200 ꢀC for the tert-butyldiphenylsilyl. At 100 repeat
units the cyclohexyl substituted exhibited a Tg of 208 ꢀC which was
the same as the tert-butyldiphenylsilyl [15]. In contrast the Tg of an
n-octyl substituted polymer exhibited a Tg of only 52 ꢀC at 100
repeat units. The polymer Tds ranged from 339 ꢀC to 426 ꢀC as
shown in Fig. 6B.
As could be expected the Td also increased as the polymer mo-
lecular weight increased. This increase in Td halted at approxi-
mately 20,000
g . By comparison to the cyclohexyl
molꢂ1
substituted case, at 10 repeat units a value of 407 ꢀC was measured,
Table 2
The analysis of the Spiral MALDI-MS expansion of Fig. 5B, showing the calculated m/
z, experimental m/z, error and repeat unit.
Repeat units (n)
Calculated m/z
Measured m/z
Error
Repeat
5
6
7
8
9
10
11
2,595
3,089
3,583
4,076
4,570
5,064
5,557
2,595
3,089
3,583
4,076
4,570
5,064
5,557
0
0
0
493
494
493
494
493
494
e
ꢂ1
0
ꢂ1
0
Fig. 6. The variation in polymer Tg (A) and polymer Td (B) as the molecular weight was
increased for polymers (6aej).

A.M. Spring et al. / Polymer 56 (2015) 189e198
197
while at 100 repeat units it was 408 ꢀC. Therefore by the incorpo-
ration of a significantly bulkier tert-butyldiphenylsilyl group it was
not possible to significantly enhance the polymers thermal prop-
erties compared to cyclohexyl substituted polymers.
Table 4
The refractive indices of the polymers spin coated onto slide glass at 1310 and
1550 nm.
Polymer
Film thickness (
m
m)
RI (1310 nm)
RI (1550 nm)
A high polymer purity is essential for organic electronic and
optical applications. Even low concentrations of residual ruthenium
complex or other impurities associated with the synthesis of the
polymers may lead to a dramatic reduction in their utility,
furthermore as the polymers are designed to be used under the
application of a high voltage, ruthenium based impurities may also
influence the resistivity. Consequently all polymers in the series
(6aej) were analyzed by CHN elemental analysis, ICP-MS, thin film
UVevis transmission spectroscopy and their electrical resistivity
measured. Polymers with a low molecular weight such as (6a)
required a much larger mass of Grubbs catalyst to prepare (333 mg)
than polymers with a higher molecular weight such as (6j) (33 mg).
Therefore it is a reasonable assumption that even after purification,
polymer 6a should contain much higher levels of ruthenium than
polymer 6j. All polymers were purified in an identical manner by
multiple precipitations from chloroform into stirred methanol.
Various sources in the literature cite the difficulty in the removal of
these Grubbs catalyst residues after the polymerization of norbor-
nenedicarboximides [20,21]. Table 3 shows the CHN elemental
analysis and ICP-MS results for the polymer series (6aej).
6a
6b
6c
6d
6e
6f
6g
6h
6i
0.7
0.7
0.8
0.9
1.0
1.0
1.0
1.2
1.0
1.1
1.56
1.57
1.56
1.57
1.57
1.57
1.57
1.57
1.57
1.57
1.57
1.57
1.56
1.55
1.56
1.56
1.56
1.57
1.57
1.56
6j
thickness of the polymer films ranged from 0.8 mm to 1.2 mm as
measured by a MEMS surface profiler. Higher molecular weight
polymers having a higher viscosity resulted in slightly thicker films.
A circular gold electrode was sputtered directly onto the polymer
films with a radius of 2 mm. Although a visual inspection of the
polymer solutions in chloroform revealed that 6a and 6b had a light
brown colouration, UVevis (400e1000 nm) transmittance mea-
surements of the thin films showed little variation, with all poly-
mers having between 99 and 95% transmittance. This indicates that
the residual ruthenium as measured by ICP-MS has little or no in-
fluence on the thin film UVevis transmission at 400e1000 nm. To
In all cases the C% expected from calculation was lower than the
measured value, verifying that the polymers did contain residual
impurities after workup. It could be expected that polymer (6a)
may exhibit the greatest difference between the expected and
measure the resistivity, a voltage of 100 V/mm was applied across
each of the polymers. As can be observed from Table 3, the poly-
mers all exhibited a similar resistivity of between 3 ꢁ 1015
U m to
measured values (DC%). The DC% varied from 0.07 to 0.39 for the
6 ꢁ 1015
U m regardless of the residual ruthenium content. Inter-
polymers (6aej), with the smallest change in fact being attributed
to polymer 6a. Therefore in terms of the total impurities, polymer
6a had the highest purity as determined by CHN elemental analysis.
However the CHN analysis of polymers is known to be a relatively
inaccurate method for the quantification of low levels of metals or
inorganic residues and the residual ruthenium within the polymers
may lead to the greatest degradation in key properties. Upon
consideration of the ICP-MS analysis results shown in Table 3,
polymer 6a and 6b contain the highest level of residual ruthenium
(7,800 p.p.m and 3,000 p.p.m) as would be expected compared to
polymers 6c to 6j which contained 1,600e540 p.p.m. By consider-
ation of both the CHN elemental analysis and the ICP-MS it could be
concluded that the polymers were all of a high purity. Given that 6a
contained 7,800 p.p.m of residual ruthenium the polymer was still
99.93% pure as determined by CHN elemental analysis which
measures the total impurities including ruthenium. The most
impure polymer (6i) was still 99.49% pure. For transmittance and
resistivity evaluations the polymers were dissolved in 1,2-dichlo-
roethane (100 mg/1 mL) and spin coated onto ITO glass slides. The
estingly the polymers (6aej) had a slightly higher resistivity than a
film of commercially available poly(methylmethacrylate), who's
resistivity was measured for comparison. This suggests that the
presence of residual ruthenium at these concentrations has little or
no influence on the polymer resistivity.
Finally the refractive indices of the polymers were measured on
slide glass at 1310 nm and 1550 nm. The polymers were spin coated
onto slide glass and the film thickness measured using a MEMS
surface profiler. The refractive indices were measured using a prism
coupler. As can be observed from Table 4 the refractive indices
ranged from 1.57 to 1.55 at 1550 nm and from 1.57 to 1.56 at
1310 nm. There was no correlation between the ICP-MS measured
ruthenium content and the refractive index of the polymers.
4. Conclusion
We have prepared a series of well controlled tert-butyldiphe-
nylsilyl substituted poly(norbornene-dicarboximide)s using the
Grubbs 1st generation catalyst. Kinetic investigations showed that
the polymerization was complete after 180 min in chloroform at
room temperature. A sequence of polymers were synthesized with
ascending molecular weights. The polymer structure was charac-
terised using NMR, GPC and MALDI-MS. Both the polymer Tgs and
Tds initially increased due to the increasing molecular weights then
reached a plateau. The residual ruthenium content of the polymers
was quantified using ICP-MS. The influence of the residual ruthe-
nium on the refractive indices, optical transmittance and resistivity
was evaluated. It was found that although the ruthenium content
varied widely across the polymer series, this had little impact on
the electronic and optical properties of the polymers.
Table 3
CHN elemental analysis, ICP-MS, UVevis thin film transmittance (%T) and resistivity
of films of polymers 6aej spin-coated onto ITO glass.
between calculated C% and measured C%.
DC% refers to the difference
Polymer
[Ru]
(ppm)
DC%
%T
Filmthickness
(mm)
Resistivity (U m)
PMMA
6a
6b
6c
6d
6e
6f
6g
6h
6i
0
7,800
3,000
1,400
1,100
1,100
1,400
930
e
100
99
99
99
99
99
99
98
99
98
95
2.0
0.8
0.9
1.0
1.0
1.1
1.2
1.1
1.1
1.1
1.2
9 ꢁ 1014
6 ꢁ 1015
5 ꢁ 1015
4 ꢁ 1015
4 ꢁ 1015
3 ꢁ 1015
3 ꢁ 1015
4 ꢁ 1015
3 ꢁ 1015
3 ꢁ 1015
3 ꢁ 1015
0.07
0.14
0.19
0.28
0.31
0.28
0.39
0.36
0.51
0.20
Acknowledgements
570
540
1,600
The authors acknowledge financial support from the following
funding sources. This work was supported by the Cooperative
6j

198
A.M. Spring et al. / Polymer 56 (2015) 189e198
[8] Jia Y, Spring AM, Qiu F, Yu F, Yamamoto K, Aoki I, et al. Jpn J Appl Phys
2014;53.
[9] Hong SH, Grubbs RH. Org Lett 2007;9:1955.
[10] Cho JH, Kim BM. Org Lett 2003;5:531.
[11] Al-Hashimi M, Hongfa C, George B, Bazzi HS, Bergbreiter DE. J Polym Sci Part A
Polym Chem 2012;50:3954.
Research Program of “Network Joint Research Center for Materials
and Devices” of the Ministry of Education, Culture, Sports, and
Science and Technology, Japan and JSPS KAKENHI Grant Number
26289108.
[12] Wang B, Sun H-X, Sun Z-H. J Org Chem 2009;74:1781.
[13] Yeom C-E, Kim HW, Lee SY, Kim BM. Synlett 2007;2007:0146.
[14] Yoon KH, Kim KO, Schaefer M, Yoon DY. Polymer 2012;53:2290.
[15] Spring AM, Yu F, Qiu F, Yamamoto K, Yokoyama S. Polym J 2014:1e8. http://
dx.doi.org/10.1038/pj.2014.26.
[16] Grubbs RH. Handbook of metathesis. Wiley-VCH; 2003.
[17] Spring AM, Yu CY, Horie M, Turner ML. Chem Commun 2009:2676.
[18] Tatro SR, Baker GR, Fleming R, Harmon JP. Polymer 2002;43:2329.
[19] Contreras AP, Cerda AM, Tlenkopatchev MA. Macromol Chem Phys 2002;203:
1811.
References
[1] Asrar J. Macromolecules 1992;25:5150.
[2] Gao AX, Liao L, Johnson JA. ACS Macro Lett 2014:854.
[3] Rao N,V, Ganivada MN, Sarkar S, Dinda H, Chatterjee K, Dalui T, et al. Bioconjug
Chem 2013;25:276.
[4] Eo M, Han D, Park MH, Hong M, Do Y, Yoo S, et al. Eur Polym J 2014;51:37.
[5] Miao W-K, Yan Y-K, Wang X-L, Xiao Y, Ren L-J, Zheng P, et al. ACS Macro Lett
2014;3:211.
[6] Jia Y, Spring AM, Yu F, Yamamoto K, Aoki I, Otomo A, et al. Thin Solid Films
2014;554:175.
[7] Yu F, Spring AM, Li L, Qiu F, Yamamoto K, Maeda D, et al. J Polym Sci Part A
Polym Chem 2013;51:1278.
[20] Al-Hashimi M, Hongfa C, George B, Bazzi HS, Bergbreiter DE. J Polym Sci Part A
Polym Chem 2012;50:5211.
[21] Patel PR, Kiser RC, Lu YY, Fong E, Ho WC, Tirrell DA, et al. Biomacromolecules
2012;13:2546.