Synthesis of hydroxymethyl-functionalized polyimides and the facile attachment of an organic dye utilizing bis(isocyanates) and bis(acid chloride) linkers

Article abstract of DOI:10.1021/ma0503920

Two new monomers were prepared of the formula 3,X-bis(4-aminophenoxy) benzenemethanol (where X = 5, 3a; X = 4, 3b) and each copolymerized with 2,2-bis(4-aminophenyl)hexafluoropropane, 6-FDA, and 6 mol % of phthalic anhydride as an end-cap to afford polyimides 4 and 5, respectively. The CH ₂OH group in the polymers could be modified with excess 1,6-hexanediisocyanate to afford 100% modification of polymer 6 concomitant with formation of a pendent and reactive isocyanate group. This latter polymer was shown to react with disperse red 1 (DR-1) to produce a polyimide-supported dye (7, T_g 199°C) via formation of a urethane linkage. Treatment of the polymers 4 and 5 with excess adipoyl chloride afforded polymers 8 and 9, respectively, with a pendent acid chloride functional group. Further reaction with disperse red 1 (DR-1) in the presence of DMAP led to rapid esterification, yielding high-T_g dye-attached polyimides 11 (T_g 168°C) and 12 (T_g 175°C). Treatment of polymer 4 with sebacoyl chloride (a longer spacer) and then DR-1 in a similar manner afforded polymer 13 (T _g 156°C). All dye-attached polymers showed low optical loss as measured in solution at both 1300 (～0.5 dB/cm) and 1500 nm (～1.0 dB/cm) and showed excellent thermal stability at an elevated temperature (180°C) relative to guest/host analogues.

Full text of DOI:10.1021/ma0503920

10014
Macromolecules 2005, 38, 10014-10021
Synthesis of Hydroxymethyl-Functionalized Polyimides and the Facile
Attachment of an Organic Dye Utilizing Bis(isocyanates) and Bis(acid
chloride) Linkers
Michael E. Wright,* Stephen Fallis, Andy J. Guenthner, and Lawrence C. Baldwin
NAVAIR Engineering Science & Research Division, Chemistry Branch,
China Lake, California 93555-6100
Received February 23, 2005; Revised Manuscript Received September 30, 2005
ABSTRACT: Two new monomers were prepared of the formula 3,X-bis(4-aminophenoxy)benzenemethanol
(where X ) 5, 3a; X ) 4, 3b) and each copolymerized with 2,2-bis(4-aminophenyl)hexafluoropropane,
6-FDA, and 6 mol % of phthalic anhydride as an end-cap to afford polyimides 4 and 5, respectively. The
CH2OH group in the polymers could be modified with excess 1,6-hexanediisocyanate to afford 100%
modification of polymer 6 concomitant with formation of a pendent and reactive isocyanate group. This
latter polymer was shown to react with disperse red 1 (DR-1) to produce a polyimide-supported dye (7,
Tg 199 °C) via formation of a urethane linkage. Treatment of the polymers 4 and 5 with excess adipoyl
chloride afforded polymers 8 and 9, respectively, with a pendent acid chloride functional group. Further
reaction with disperse red 1 (DR-1) in the presence of DMAP led to rapid esterification, yielding high-Tg
dye-attached polyimides 11 (Tg 168 °C) and 12 (Tg 175 °C). Treatment of polymer 4 with sebacoyl chloride
(a longer spacer) and then DR-1 in a similar manner afforded polymer 13 (Tg 156 °C). All dye-attached
polymers showed low optical loss as measured in solution at both 1300 (∼0.5 dB/cm) and 1500 nm (∼1.0
dB/cm) and showed excellent thermal stability at an elevated temperature (180 °C) relative to guest/host
analogues.
Introduction
the potential dye sites (i.e., backbone phenolic -OH’s)
were utilized. In this paper we present a very chemically
efficient and versatile synthetic route for the attachment
of organic dyes to a polyimide backbone based on the
modification of a benzylic alcohol sites that affords
materials possessing low optical loss and high Tg’s.
The chemical modification and use of fluorinated
polymers plays a key role in a variety of today’s
advanced materials technologies.1 The practice of post-
polymer modification ranges dramatically in purpose
and style. We continue to see elegant applications of
post-polymer modification in diverse areas of polymer
chemistry that include approaches to combinatorial
synthesis2 and the development of novel electro- and
optoactive materials with conjugated polymer back-
bones.3
More specifically, we have had interest in developing
a strategy for the efficient and chemically mild incor-
poration of organic dyes into a robust polymeric mate-
rial. There are currently two principal means of attach-
ing an organic dye to a polymer. One method is to
prepare a monomer containing the optoactive group and
then co- or homopolymerize the molecule.4 This requires
the optoactive group to be stable to the polymerization
conditions and any other post-cure processes. The other
method is to link (i.e., chemically bond) a suitably
functionalized optoactive molecule to an existing poly-
mer backbone.5 The two methodologies have clear
advantages and disadvantages related to (1) chemical
efficiency (e.g., number of synthetic steps) and (2)
functional group tolerance.
With interest in utilizing polyimides (PIs) as the bulk
polymeric material, we found relatively few synthetic
procedures that provide a mild and efficient means of
attaching a dye to the polyimide backbone.5,6 In addition
to having mild reaction conditions and high attachment
efficiency, we sought synthetic methodology that would
permit systematic changes in Tg and isolation of the dye
from the backbone structure. Very recently, Luo and co-
workers7 reported on the modification of a phenolic-
polyimide backbone by use of a diacid linker and two
consecutive esterification steps. In that case, ∼50% of
Results and Discussion
Monomer Synthesis. From the readily available
methyl 3,5-dihydroxybenzoate we have developed a very
efficient synthetic scheme for the preparation of the new
monomer 3,5-bis(4-aminophenoxy)-1-hydroxymethyl-
benzene (3a) (Scheme 1).8 We find that 1a can be
quickly purified by trituration with hot ethanol. The
remaining two steps are performed in such a manner
as to require only simple filtrations to afford analytically
pure samples. It is interesting to note in the hydrogena-
tion of 1a that completion of the reaction is signaled by
a visually distinct agglomeration of the catalyst.
The synthesis of monomer intermediate 1b proves to
be more difficult and time-consuming. The primary
difficulty is reaction of the 4-fluoro-1-nitrobenzene at
the 3-position of the methyl benzoate compound after
the initial and faster reaction at the 4-position. Column
chromatography is utilized to remove the methyl 3-hy-
droxy-4-(4-nitrophenoxy)benzoate byproduct from the
desired product. The structure of the latter compound
is easily verified by proton NMR spectroscopy. Once 1b
was in hand, the remaining two steps proceed in
excellent yield and ultimately provide the monomer 3b
in 50% overall yield.
Polymer Synthesis and Dye Attachment. Mono-
mers 3a and 3b are each copolymerized with 2,2-bis(4-
aminophenyl)hexafluoropropane and hexafluoroisopro-
pylidene-2,2-bis[4-phthlaic anhydride] (6-FDA) to afford
polyimides 4 and 5, respectively. The functionalized
monomers are used in a molar ratio that affords a final
10.1021/ma0503920 CCC: $30.25 © 2005 American Chemical Society
Published on Web 11/05/2005

Macromolecules, Vol. 38, No. 24, 2005
Hydroxymethyl-Functionalized Polyimides 10015
Scheme 1
Scheme 2
strategy, we sought to attach a functional group that
would react with an alcohol group under mild reaction
conditions at or near neutral pH. It is imperative to
avoid the use of nucleophilic agents or harsh reaction
conditions that might decompose some of the more
sensitive dyes that are in current use. In addition, to
retain good polymer properties and processability the
modification chemistry must avoid cross-linking and/
or backbone scission reactions.
The first approach involved treatment of 4 with an
excess (∼30-fold) of 1,6-diisocyanatohexane in the pres-
ence of triethylamine (Scheme 2). Complete reaction at
the benzyl alcohol group in polymer 4 is evident in the
proton NMR spectrum. The benzyl-methylene NMR
signal shows a downfield shift of nearly 0.5 ppm upon
conversion to the urethane. The pendent isocyanate in
polymer 6 is observed in the IR spectrum as a strong
absorbance peak at ∼2271 cm-1. The polymer retains
excellent solubility in THF, hence suggesting that little
if any cross-linking has occurred during polymer modi-
fication.
polyimide having ∼0.6 mequiv of -OH per gram of
polymer. In addition, all polymerizations are carried out
using 6 mol % of phthalic anhydride as end-cap to
provide control over molecular weight. Under the present
polymerization conditions, we obtain polymers with
number-average molecular weights in the range of
∼5500 to ∼6500 and polydispersities of ∼2. Polyimides
4 and 5 are found to have Tg’s of 268 and 283 °C,
respectively. These are typical of polyimides and fall into
a very useful range for many photonic-type applications.
The polymers are soluble in common organic solvents,
although we observe best solubility (20 wt % or more)
in oxygenated solvents such as THF and p-dioxane.
Polymer 6 does undergo reaction with the alcohol
functional group of DR-1 in the presence of DMAP to
afford polymer 7. Although NMR data show better than
90% attachment efficiency (i.e., comparison of the benzyl
methylene to the dye-CH2OCNH- signals), there is
evidence in the infrared spectrum of polymer 7 for
unreacted isocyanate groups. This is evidenced by an
infrared absorption at 2265 cm-1 that by relative
integration values may account for ∼5% of the starting
isocyanate sites being left unreacted. Elemental analysis
data for the polymer are in very good agreement with
the proposed structure. Analysis of polymer 7 by SEC
shows a small increase in polydispersity (2.4 w 2.8) with
a concomitant decrease in retention time. These data
coupled with excellent solubility are indicative that little
cross-linking has occurred through the dye attachment
chemistry. Although not explored in this work, the
unreacted isocyanate groups may prove useful for
subsequent cross-linking reactions.
Complete assignment of the proton NMR data for
polymers 4 and 5 has been completed and is placed in
the Experimental Section. In each case the benzyl
methylene is well separated from other NMR signals
and proves to be a valuable diagnostic probe for the
subsequent polymer modification chemistry presented
herein.
The second approach to dye attachment involves
treatment of polymers 4 and 5 with a 10-fold excess of
a diacid chloride in the presence of 2,6-di-tert-butyl-4-
methylpyridine and 4-(N,N-dimethylamino)pyridine
(DMAP). This affords the functionalized polymers 8 and
We have explored two new approaches for the attach-
ment of dye molecules to a polyimide backbone. In each

10016 Wright et al.
Macromolecules, Vol. 38, No. 24, 2005
Scheme 3
Figure 1. SEC traces for polymer 5 before chemical modifica-
tion and then polymer 12 with DR-1 attached.
We have followed the polymer modification in the
preparation of 12 using SEC analysis, and the results
are displayed in Figure 1. These data demonstrate
visually that the attachment chemistry does not modify
the polydispersity of the polymer at each stage of the
modification; however, it does afford the anticipated
increase in molecular weight upon dye attachment.
One of clear advantages to the attachment strategy
presented in this paper is the ability to tailor the length
of the spacer that connects the dye to the polymer
backbone. It is well-known that the length of the spacer
can directly affect the ability of the NLO-phore (i.e., dye)
to be oriented during an alignment process. To demon-
strate this ability, we choose a diacid chloride with 10
carbons, namely sebacoyl chloride. The increased linker
length leads to a DR-1 modified polymer with a signifi-
cantly lower Tg of 156 °C (∼20 °C in comparison to 11).
Polymer Physical Properties and Processing. To
successfully carry out the analysis of optical loss in
solution and thermal stability of polymer films, it is
necessary to select a common solvent for the polyimide
backbone (in this study that refers to polymer 4), free
DR-1, and the chromophore-attached polymer that
featured a sufficiently high boiling point to enable spin-
casting of high-quality films and that has a reasonably
simple absorption spectrum in the near-infrared region
of interest. Of the nearly one dozen potential solvents
identified, cyclohexanone was judged to best fulfill the
requirements stated above.
9, respectively, in high purity as determined by NMR
spectroscopy and good yield (Scheme 3). In the reaction
a precipitate is formed that is believed to be, in part,
the pyridinium hydrochloride salt. The precipitate is
removed from the reaction mixture by passing the
reaction mixture through a plug of Celite. It is impor-
tant to carry out the precipitation in ether and collection
of the polymers under the protection of dry nitrogen.
The polymers isolated after brief drying show complete
functionalization of the benzyl sites (i.e., loss of the NMR
signal at 4.63 ppm) with the expected proportion of
pendent acid chloride sites. The proton NMR spectrum
is again particularly useful for assigning the extent of
chemical modification and the integrity of the pendent
acid chloride functional group. Infrared spectroscopy
clearly showed the carbonyl stretch for the acid chloride
at 1785 cm-1, which upon reaction with dye moves into
the massive imide-carbonyl band.
We find that polymers 8-10 afford the highest and
most consistent dye attachment efficiencies if used
immediately after drying for a period of only 1 h at
reduced pressure (∼0.2 Torr). By NMR spectroscopy we
can observe residual ether; however, there appears to
be no interference with the dye attachment reaction. We
have also found the pendent acid chloride to be ex-
tremely sensitive to choice of the acid scavenger. For
example, triethylamine affords low dye attachment
efficiencies with concomitant destruction of the acid
chloride. At this point in time we do not have spectro-
scopic evidence to single out one particular pathway for
decomposition. Even with low attachment efficiencies,
we observe only one benzyl-CH2O2C- signal in the
proton NMR spectrum. This would indicate that the side
reaction is occurring at the acid chloride functional
group and does not seem to involve the benzylic position.
In a similar manner, we have taken polymer 5 and
treated it with excess adipoyl chloride and isolated the
acid chloride functionalized polymer and then attached
DR-1 using DMAP in dichloromethane. This process
affords polymer 12 in good yield. The latter polymer only
differs from 11 by the linkage substitution on the benzyl
alcohol core unit.
Since sample preparation involves use of filtered
solutions, it is necessary to ensure that the filtration
process is not inadvertently removing significant amounts
of either the polymer or the chromophore; a simple
comparison of spectra using filtered and unfiltered
solutions was performed over a wavelength of 500-1600
nm in order to observe the peak absorption band edge
of DR-1. Both the magnitude and location of the band
edge remained unchanged before and after filtration,
strongly supporting the hypothesis that no DR-1 is
removed during filtration.
Figure 2 shows the optical loss spectra of the chro-
mophore-attached polyimide along with the backbone
polymer and the backbone/DR-1 mixtures from 1000 to

Macromolecules, Vol. 38, No. 24, 2005
Hydroxymethyl-Functionalized Polyimides 10017
the chromophore-attached polymer are thus as low or
lower than those of the starting materials. Moreover,
the losses of the chromophore-attached polyimide are
quite similar to those of other polyimides measured in
our laboratory9 and to other well-known host polymers
for optical applications, such as amorphous polycarbon-
ate.10
The most likely explanation for the unusually high
loss values in the backbone polyimide and guest/host
polymer systems above 1400 nm is the presence of
hydroxyl groups on both the DR-1 and on the backbone
itself (i.e., polymer 4). Whereas in the backbone, there
is one hydroxyl group per ∼1600 amu equivalent weight;
in DR-1 there is one hydroxyl group per 314 amu
equivalent weight. Thus, if hydroxyl groups were the
major constituent of the excess optical loss, then loss
should increase with increasing DR-1 content in guest/
host systems; however, it should be greatly reduced as
the attachment reaction consumes both DR-1 and
backbone hydroxyl groups.
Perhaps the most important conclusion that can be
drawn from the absorption data is that the completion
of the attachment reaction is seen to introduce no major
sources of absorption into the near-infrared spectrum.
The only clearly visible “new” peaks that coincide with
the completion of attachment appear in the 1150-1200
nm range and are small. Moreover, the measured losses
for the chromophore-attached polymer are generally at
or below about 0.5 dB/cm at 1300 nm and 1 dB/cm at
1500 nm. Although these results apply only to “intrinsic”
absorption in solution and a technique such as photo-
thermal deflection spectroscopy11 would be needed to
make definitive statements about solid-state behavior,
the presence of broad regions free of major absorption
peaks in the solution spectrum of the chromophore-
attached polymer is a very positive indicator for low loss
behavior in solid-state films.
Thermal Stability of the Dye-Modified Polyim-
ides. While avoiding increased optical losses is an
important achievement in developing chromophore at-
tachment reactions, by itself it does not demonstrate
that a chromophore-attached polymer system would be
any more desirable than the equivalent guest/host
system for electrooptical applications. It is therefore
important to validate that the chromophore-attached
materials do indeed deliver at least some of their
promised benefits. These include but are not limited to
increased physical thermal stability of the chromophore,
restricted motion (especially below Tg), and a decrease
in tendencies for phase separation (i.e., crystallization
or aggregation of the dye).
Although it is possible to perform dynamic nonlinear
optical thermal stability measurements on poled elec-
trooptical devices to test the effects of chromophore
attachment on inhibiting the free motion of the chro-
mophore molecules, a much simpler test involves the
linear optical stability of very thin films. Many guest/
host systems, including those containing DR-1, are
known to suffer from unwanted sublimation of the
chromophore molecules at elevated temperatures.12 This
effect is easily measured by examining the absorption
spectrum of DR-1 in submicron-thick films as a function
of time at elevated temperatures, where it is found that
when sublimation takes place, there is a rapid loss in
absorption intensity over a periods as short as a few
minutes. Figure 3 shows the relative decrease in ab-
sorption intensity (at the peak absorption wavelength)
Figure 2. Optical loss, as measured by near-infrared spec-
troscopy, for the SC-DR-1 polymer and guest/host mixtures of
the backbone and chromophore starting materials.
1600 nm. The calculated volume fractions of solute were
3.7% for the polyimide backbone, 3.8% for the 17 wt %
DR-1 in polyimide and for the chromophore-attached
polyimide, and 4.1% for the 33 wt % DR-1 in polyimide.
The resulting density values of 1.3 g/cm3 for the
backbone polyimide, 1.25 g/cm3 for the chromophore-
attached polyimide, and 1.25 and 1.15 g/cm3 for the 17
and 33 wt % guest/host systems, respectively, are quite
reasonable. Examination of the spectra as a function of
the proposed volume fraction of solute shows that the
error in the volume fraction estimate is about (0.1%
or, in terms of density, about 0.05 g/cm3. Further
examination of the spectra show that the observed loss
values at 1300 and 1550 nm are very insensitive to
variations of the proposed volume fractions, changing
by only about 0.05 dB/cm per 0.1% change in volume
fraction.
For all but the chromophore-attached polymer, the
spectra reveal two distinct trends. At wavelengths below
1400 nm, the losses are highest for the backbone
polyimide and decrease with increasing DR-1 content.
These samples show broad regions where no significant
absorption peaks are present at 1000-1100 and 1240-
1320 nm, in which the losses range from 0.5 to 0.7 dB/
cm for the polyimide backbone to 0.2-0.4 dB/cm for the
backbone with 33 wt % DR-1. Thus, neither the poly-
imide nor the DR-1 is likely to cause significant loss
issues in these regions.
For wavelengths in excess of 1400 nm, however, the
trend is reversed. Of the backbone and guest/host
systems, the polyimide with no DR-1 shows the lowest
losses, while increasing DR-1 leads to increased loss,
especially in the regions of strongest absorption around
1450 nm. At 1550 nm, the losses range from just under
2 dB/cm for the backbone to near 2.3 dB/cm for the 33
wt % DR-1 system. Thus, at these wavelengths, optical
losses from the starting materials for the chromophore-
attached polymer are potentially troublesome. When the
losses from the chromophore-attached polymer itself are
considered, however, the differences between wave-
lengths above and below 1400 nm become even more
apparent. Below 1400 nm, the chromophore-attached
polymer behaves similarly to the guest/host systems,
with a very low loss of 0.3 dB/cm at a wavelength of
1300 nm. Above 1400 nm, however, the losses in the
chromophore-attached polymer are significantly lower
than in the guest/host systems, with a broad region free
of absorption peaks between about 1480 and 1560 nm,
in which the loss is just 0.8-1.0 dB/cm. The losses for

10018 Wright et al.
Macromolecules, Vol. 38, No. 24, 2005
significant reduction in chromophore mobility has been
achieved by the chemical attachment, and this new
attachment chemistry is quite stable at high tempera-
tures for extended periods of time.
Concluding Remarks
In this work we have demonstrated that a benzyl
alcohol-functionalized diamine can be polymerized with
bis(anhydrides) to a highly functionalized polyimide that
exhibits exceptional solubility and processability. Fur-
thermore, we have shown that the benzyl alcohol group
in the new polymers can be chemically modified with a
bis-functional linking agent, for example bis(isocyan-
ates) or bis(acid chlorides), to afford reactive side-chain
polyimides that contain a reactive isocyanate or acid
chloride group. The “pendent functional” group has been
shown to provide an excellent site for attachment of an
alcohol-containing dye (like DR-1). Furthermore, by
changing the length of the linker (i.e., diacid chloride),
a means of tuning the polymers final Tg is possible with
this new attachment chemistry. The polyimides pre-
sented here with DR-1 attached were shown to have
excellent thermal properties as well as low optical loss.
Work continues to more fully utilize the full potential
of this very efficient attachment/polymer modification
chemistry.
Figure 3. Relative amounts of chromophore lost to sublima-
tion or degradation after exposure to elevated temperature for
the SC-DR-1 polymer and guest/host mixtures of the backbone
and chromophore starting materials.
for film ranging in thickness from 0.3 to 0.6 µm of the
chromophore-attached DR-1 polymer and two corre-
sponding guest/host systems having 17 and 33 wt % DR-
1, respectively, after being exposed to various temper-
atures for a period of 30 min. The general trend at all
temperatures is for the 33 wt % guest/host system to
suffer the largest changes, followed by the 17 wt %
guest/host system and, last, by the chromophore-
attached polymer. The differences are most pronounced
at 180 °C, where the chromophore-attached polymer’s
peak absorption changes by just over 4%, compared to
over 10% for the 17 wt % guest/host system and over
30% for the 33 wt % guest/host system.
The loss of absorption intensity in these systems at
elevated temperature could be due to any one or a
combination of three likely causes: (1) changes in the
absorption spectrum of the host (or backbone) polyimide,
(2) sublimation of DR-1, or (3) chemical degradation of
DR-1. To check the first possibility listed above, the
spectrum of the polyimide host (which is also the
backbone for the dye-attached polymer) was compared
before and after exposure to 180 °C for 30 min. The
observed spectral changes were negligible, indicating
that the backbone polyimide is very stable, as expected,
at these temperatures. Distinguishing between the
second and third possibilities is more difficult, but DR-1
is not known to suffer significant chemical degradation
in a nitrogen atmosphere at 180 °C, and, if it were, we
would expect such degradation to affect the chro-
mophore-attached and guest/host systems to a similar
extent. Thus, although a small portion of the loss in
absorption, particularly in the chromophore-attached
polymer may be attributed to degradation, the major
component of the differences between the guest/host and
chromophore-attached systems must be ascribed to
differences in the sublimation rate of DR-1.
Experimental Section
General Synthetic Methods. All manipulations of com-
pounds and solvents were carried out using standard Schlenk
techniques. Tetrahydrofuran (THF), ether, N-methylpyrroli-
dinone (99.5%), and dichloromethane solvents were purchased
as the anhydrous grade and inhibitor-free from Aldrich and
used as received. 1H and 13C NMR measurements were
performed using a Bruker AC 200 or Bruker Avance 400 MHz
1
instrument. H and 13C NMR chemical shifts are reported vs
1
the deuterated solvent peak (solvent, H, 13C: CDCl3, δ 7.27
ppm, δ 77.0 ppm; DMSO-d6, δ 2.50 ppm, δ 39.5 ppm). Methyl
3,5-dihydroxybenzoate, 4-fluoronitrobenzene, LAH (95%, pow-
der), cyclohexanone (99+%), and phthalic anhydride were
purchased from Aldrich Chemical Co. and used as received.
The 6-FDA (lot #104\65) and 6F-4,4′-diamine were purchased
from CHRISKEV Co. and used as received. Elemental analyses
were performed at Atlantic Microlab, Inc., Norcross, GA.
Optical Loss Measurements. The guest/host systems
were mixed by adding enough solvent to 5:1 and 2:1 by weight
mixtures of the backbone polyimide and the DR-1 to make 5
wt % solutions (for spectrophotometry) or 10 wt % solutions
(for film casting) and stirring vigorously at room temperature
for 2 h. Corresponding solutions of only the polyimide backbone
and only the chromophore-attached polymer were similarly
prepared. Unless otherwise noted, all solutions were passed
through Teflon filters (pore size 0.2 µm) prior to use.
Optical measurements were carried out in a Cary 5 spec-
trophotometer using quartz cuvettes following a previously
published procedure.9 With these careful procedures in place,
a precision of 0.0005-0.001 in absorbance could be obtained,
which translated into around (0.2 dB/cm in solution optical
loss.
Thermal Stability Studies. Solutions in cyclohexanone at
10 wt % were prepared of the chromophore-attached polymer,
the backbone polyimide, and guest/host mixtures of the
backbone polyimide and DR-1 in a 5:1 and 2:1 weight ratio,
which are referred to as the 17 and 33 wt % mixtures,
respectively. The solutions were filtered and then spun onto
clean glass microscope slides at 2400 rpm for 30 s using a
Headway Research, Inc., model EC101 spin-casting apparatus.
The resulting films were baked at 70 °C for 30 min under dry
nitrogen and then allowed to sit at room temperature for
another 24 h. This procedure is known to be effective for
In comparing the guest/host and chromophore-
attached systems, it is also worth noting that the rate
of sublimation (relative to the initial amount of chro-
mophore present) should increase with decreasing film
thickness if diffusion of the chromophore within the film
is a limiting factor in mass transfer. In our case, the
thinnest film, at 0.3 µm, was the chromophore-attached
polymer, followed by the 17 wt % guest/host system at
0.45 µm and the 33 wt % system at 0.6 µm. Thus, if the
effects of film thickness were significant, they would
only serve to mask the intrinsic differences among the
systems studied by reducing the rate of guest/host
chromophore sublimation relative to that in the at-
tached chromophore polymer. Hence, it is quite clear a

Macromolecules, Vol. 38, No. 24, 2005
Hydroxymethyl-Functionalized Polyimides 10019
removing practically all cyclohexanone from the resultant
films, which ranged in thickness from 0.3 to 0.6 µm.
Anal. Calcd for C19H18N2O3: C, 70.83%; H, 5.63%; N, 8.69%.
Found: C, 70.81%; H, 5.89%; N, 8.62%.
Preparation of Methyl 3,4-Bis(4-nitrophenoxy)ben-
zoate (1b). A flask was charged with methyl 3,4-dihydroxy-
benzoate (5.00 g, 29.7 mmol), DMF (50 mL), K2CO3 (13 g,
mmol), and 4-fluoronitrobenzene (9.22 g, 6.93 mL, 65.3 mmol),
in that order. The mixture was heated at ∼70 °C for a period
of 48 h with stirring. The mixture was diluted with dichlo-
romethane (200 mL) and then water (400 mL). This mixture
was stirred vigorously for 15 min, and the layers were
separated. The organic layer was washed in this manner with
water (4 × 100 mL) and brine (100 mL) and then dried over
magnesium sulfate. The drying agent was removed, and the
solvents removed under reduced pressure. Final purification
was achieved by column chromatography (2 × 20 cm, silica
gel, hexanes/dichloromethane, 1/4, v/v, respectively). The
product was the second major band collected off the column
and after removal of the solvents afforded 1b as a light-yellow
After the initial drying was complete, a spectrum of the film
was recorded from 800 to 350 nm at 0.6 nm intervals and an
averaging time of 0.1 s, with the spectrum of an uncoated but
similarly cleaned microscope slide as a baseline. The film was
placed in the spectrophotometer beam in a sample holder that
allowed for accurate control of the beam location on the surface
of the film. After collecting the spectrum, the sample was again
heated for 30 min under dry nitrogen. Separate samples were
used with heating temperatures of 120, 140, and 180 °C. After
heating, the samples were quickly cooled to room temperature,
aligned in the sample holder so that the same location would
be illuminated by the sample beam, and measured again using
the same procedure. The spectra of the samples (except for
the backbone polyimide with no DR-1) consisted of a single,
large peak of absorbance 0.8-2.4 at 475-500 nm. Although
the spectrum of the backbone polyimide showed a small
absorbance at this wavelength, its effect on the relative change
in the value of the peak absorbance was considered negligible.
Thus, the amount of DR-1 lost during the exposure to high
temperature was computed on the basis of the relative peak
absorption between 475 and 500 nm as compared before and
after exposure. The error in the measured peak values has
been demonstrated to be around 1%, which is easily small
enough to allow for the desired comparisons.
Preparation of Methyl 3,5-Bis(4-nitrophenoxy)ben-
zoate (1a). A flask was charged with methyl 3,5-dihydroxy-
benzoate (23.0 g, 137 mmol), DMF (200 mL), K2CO3 (65 g,
0.471 mol), and 4-fluoronitrobenzene (32 mL, 260 mmol), in
that order. The mixture was heated at ∼60 °C for a period of
48 h with stirring. The mixture was diluted with dichlo-
romethane (500 mL) and washed several times with water (5
× 300 mL) in an Erlenmeyer flask with vigorous strirring. The
organic layer was washed with brine (100 mL) and then dried
over MgSO4. The solvents were removed under reduced
pressure, and the crude product was triturated using hot
ethanol (∼100 mL) to afford light yellow crystals of 1a (43.6
g, 78%, mp 176-179 °C).
1
glassy-like solid (8.52 g, 70%). H NMR (CDCl3, 200 MHz) δ:
8.05 (d, J ) 7.0 Hz, 2H), 8.04 (d, J ) 7.0 Hz, 2H), 7.90 (dd, J
) 2.0, 8.5 Hz, 1H), 7.82 (d. J ) 2.0 Hz, 1H), 7.19 (d, J ) 8.5
Hz, 1H), 6.85 (d, J ) 7.0 Hz, 2H), 6.81 (d, J ) 7.0 Hz, 2H), 3.
81 (s, 3H, OCH3). Anal. Calcd for C20H14N2O8: C, 58.54%; H,
3.44%. Found: C, 58.70%; H, 3.50%.
Preparation of Methyl 3,4-Bis(4-aminophenoxy)ben-
zoate (2b). A Schlenk flask was charged with methyl 3,4-bis-
(4-nitrophenoxy)benzoate (3.50 g, 10.0 mmol), THF (75 mL),
and platinum oxide (350 mg). The mixture was placed under
a hydrogen atmosphere (∼2 psig) and allowed to react with
stirring for 12 h. At the end of the reaction the platinum oxide
catalyst is seen to aggregate at the bottom of the flask. The
mixture was filtered through a plug of Celite and the THF
removed under reduced pressure to afford pure 2b as a white
solid (2.82 g, 94%). 1H NMR (CDCl3, 200 MHz) δ: 7.55 (dd, J
) 2.0, 8.5 Hz, 1H), 7.48 (d, J ) 2.0 Hz, 1H), 6.80 (d, J ) 8.7
Hz, 4H), 6.75 (d, J ) 8.5 Hz, 1H), 6.59 (d, J ) 8.7 Hz, 4H), 3.
75 (s, 3H, OCH3), 3.53 (br s, 4H, NH2). Anal. Calcd for
C20H18N2O4: C, 68.56%; H, 5.18%. Found: C, 68.44%; H,
5.02%.
Preparation of 3,4-Bis(4-aminophenoxy)benzene-
Methanol (3b). A THF (50 mL) solution containing methyl
3,4-bis(4-aminophenoxy)benzoate (2.50 g, 7.14 mmol) was
added dropwise to a THF (75 mL) solution/suspension contain-
ing LAH (1.50 g, 39.5 mmol). A gentle reflux was maintained
during the addition as well as controlling evolution of hydrogen
gas. After complete addition of the ester solution, the mixture
was heated at reflux for an additional 16 h with stirring. The
mixture was carefully quenched with water (1.5 mL), 15%
aqueous NaOH (1.5 mL), and water (4.5 mL) and then dried
over magnesium sulfate. The solution was filtered and the
THF removed under reduced pressure to afford pure 3b as a
colorless oil (2.33 g, 80%). The yield is adjusted for the sample
containing ∼3 wt % of THF as determined by proton NMR
spectroscopy. 1H NMR (DMSO-d6, 400 MHz) δ: 6.90 (dd, J )
8.3 Hz, J ) 1.8 Hz, 1H, H2), 6.82 (d, J ) 8.3 Hz, 1H, H3), 6.81
(d, J ) 1.8 Hz, 1H, H6), 6.71 (d, J ) 8.7 Hz, 2H, Ar′H2), 6.69
(d, J ) 8.7 Hz, 2H, Ar′H2), 6.56 (d, J ) 8.7 Hz, 2H, Ar′H3),
6.55 (d, J ) 8.7 Hz, 2H, ArH3), 5.14 (t, J ) 5.6 Hz, 1H, -OH),
4.91 (s, 2H, -NH2), 4.87 (s, 2H, -NH2), 4.36 (d, J ) 5.6 Hz,
2H, -CH2). 13C NMR (DMSO-d6) δ: 148.6, 147.3, 146.6, 146.5,
145.0, 144.6, 138.0 (aromatic C’s), 120.9, 119.7, 119.0, 116.6,
114.9, 114.8 (aromatic CH’s), 62.3 (CH2). Anal. Calcd for
C19H18N2O3: C, 70.83%; H, 5.63%; N, 8.69%. Found: C,
70.51%; H, 5.44%; N, 8.54%.
Preparation of CH2OH-Functionalized Polyimide (4).
A Schlenk flask was charged with monomer 3a (5.430 g, 16.14
mmol), 6F-4,4′-diamine (8.094 g, 24.21 mmol), and NMP (70
mL). Once the solution was homogeneous the amine solution
was added to a NMP (70 mL) solution 6-FDA (16.853 g, 37.93
mmol) and phthalic anhydride (717 mg, 4.84 mmol, 6 mol %
end-cap). The mixture was stirred at ambient temperature for
∼16 h and then fitted with a reflux condenser and immersed
in an oil bath heated to 180 °C for 6 h. The reaction vessel
was removed from the oil bath, allowed to spontaneously cool
for 30 min, and then poured into a vigorously stirred solution
Preparation of Methyl 3,5-Bis(4-aminophenoxy)ben-
zoate (2a). A Schlenk flask was charged with methyl 3,5-bis-
(4-nitrophenoxy)benzoate (10.0 g, 24.4 mmol), THF (150 mL),
and platinum oxide (800 mg). The mixture was placed under
a hydrogen atmosphere (∼2 psig) and allowed to react with
stirring for 6-10 h. At the end of the reaction the platinum
oxide catalyst is seen to aggregate at the bottom of the flask,
signaling completion of the reaction. The mixture is filtered
through a plug of Celite and the THF removed under reduced
pressure to afford pure methyl 3,5-bis(4-aminophenoxy)ben-
zoate as a white powder (8.40 g, 98%, mp 159-161 °C, lit.8
1
165-167). H NMR (DMSO-d6) δ: 6.94 (d, J ) 2.4 Hz, 2 H),
6.80 (d, J ) 8.8 Hz, 4 H), 6.70 (t, J ) 2.4 Hz, 1 H), 6.61 (d, J
) 8.8 Hz, 4 H), 5.06 (br s, 4 H, NH2), 3.75 (s, 3 H, OCH3).
Preparation of 3,5-Bis(4-aminophenoxy)benzene
Methanol (3a). A THF (200 mL) solution containing methyl
3,5-bis(4-aminophenoxy)benzoate (8.40 g, 24.0 mmol) was
added dropwise to a THF (200 mL) solution/suspension
containing LAH (4.10 g, 108 mmol). A gentle reflux was
maintained during the addition as well as controlling evolution
of hydrogen gas. After complete addition of the ester solution,
the mixture was heated at reflux under a nitrogen atmosphere
for an additional 18 h with stirring. The mixture was carefully
quenched with water (4.1 mL), 15% aqueous NaOH (4.1 mL),
and water (12.3 mL) and then dried over magnesium sulfate
(∼2 g). The solution was filtered, and the THF was removed
under reduced pressure to afford pure monomer 3a (as an oil,
which after standing, may crystallize, 6.58 g, 85%, mp 97-99
°C). The yield is adjusted for the sample containing 0.1-0.3
mol equiv of THF as determined by proton NMR spectroscopy.
1H NMR (DMSO-d6) δ: 6.75 (d, J ) 8.7 Hz, 4 H), 6.59 (d, J )
8.7 Hz, 4 H), 6.42 (d, J ) 2.2 Hz, 2 H), 6.23 (t, J ) 2.2 Hz, 1
H), 5.15 (br s, 1H, OH), 4.96 (s, 4H, NH2), 4.35 (s, 2H, CH2).
13C NMR (DMSO-d6) δ: 160.1, 145.7, 145.6, 145.2 (aromatic
C’s), 121.0, 114.8, 107.3, 103.2 (aromatic CH’s), 62.5 (CH2).

10020 Wright et al.
Macromolecules, Vol. 38, No. 24, 2005
of methanol (2 L). Vigorous stirring of the methanol was
maintained during the entire precipitation process. The pre-
cipitated polymer was then collected on a medium-porosity
glass frit, washed with methanol (∼2 L), and then dried under
reduced pressure for 12 h at 80 °C. Polyimide 4 was isolated
as an off-white solid (23.65 g, 80%, Mn 6500, polydispersity
2.3, Tg 268 °C). 1H NMR (CDCl3) δ: 7.85-8.10 (m, 17 H), 7.50-
7.65 (m, 10 H), 7.38 (br s, 4 H), 7.13 (br s, 4 H), 6.83 (s, 2 H),
6.72 (s, 1 H), 4.63 (s, 2 H). Selected 13C NMR (CDCl3) signals:
δ 166.3, 166.2, 165.9, 165.8 (carbonyls), 158.2, 156.9 (aromatic
C’s), 131.3, 128.2, 126.6, 119.5, 112.7 (aromatic CH’s), 64.6
(CH2OH). Anal. Calcd for 4: C, 58.37%; H, 2.28%; N, 3.84%.
Found: C, 58.22%; H, 2.17%; N, 3.95%.
Preparation of CH2OH-Functionalized Polyimide (5).
In a similar manner as above, a Schlenk flask was charged
3b (1.55 g, 4.56 mmol), 6F-4,4′-diamine (2.285 g, 6.83 mmol),
and NMP (25 mL) and stirred until homogeneous. In one
portion was added 6-FDA (4.760 g, 10.72 mmol) and phthalic
anhydride (202 mg, 6 mol % end-cap). This mixture was
allowed to react with stirring at ambient temperature for 12
h and place in an oil bath maintained at 180 °C for 6 h. The
mixture was removed from the oil bath and after ∼30 min
poured into methanol (∼1 L) and isolated as above. This
afforded polymer 5 as an off-white powder (6.77 g, 79%, Mn ∼
5300, polydispersity 1.7, Tg 283 °C). 1H NMR (CDCl3) δ: 8.12-
7.8 (m, ∼16 H), 7.6-7.55 (m, ∼11 H), 7.34 (d, J ) 8.8 Hz, 4
H), 7.18 (br s, 2 H), 7.01 (d, J ) 8.8 Hz, 4 H), 4.62 (s, 2 H).
Anal. Calcd for 5: C, 58.37%; H, 2.28%; N, 3.84%. Found: C,
57.41%; H, 2.27%; N, 3.64%.
Attachment of an Organic Dye to Acid Chloride-
Functionalized PI. A typical procedure is as follows: The
functionalized polyimide (8, 9, or 10, 1.00 g, 0.49, 0.49, and
0.43 mequiv, respectively) was dissolved in dichloromethane
(25 mL) and then treated with DR-1 (500 mg, 1.59 mmol). This
mixture was stirred for ∼10 min, and then in one portion
DMAP (100 mg, 0.82 mmol) was added. The mixture was
allowed to react with stirring for 3 h at ambient temperature.
The mixture was added to a rapidly stirred solution of
methanol (200 mL), and the polymer was collected on a glass
frit (60 mL, medium). The resulting polymer was washed with
methanol, ether (50 mL), methanol (100 mL), and ether (50
mL) and then dried under reduced pressure (∼0.2 Torr) for
16 h. This typically yielded 1.0 g (85% yield) of polymer with
near-quantitative attachment of organic dye as measured by
NMR, Tg, and UV-vis spectroscopy. Selected data for the
polymers:
1
Polymer 11 (Tg 168 °C). H NMR (CDCl3) δ: 8.30 (d, J )
8.5 Hz, 2 H), 8.10-7.82 (m, ∼16 H), 7.65-7.55 (m, ∼10 H),
7.41 (d, J ) 8.8 Hz, 4 H), 7.16 (d, J ) 8.8 Hz, 4 H), 6.9-6.8
(m, 3 H), 6.75 (s, 1 H), 5.06 (s, 2 H), 4.29 (apparent t, 2 H),
3.68 (apparent t, 2 H), 3.53 (m, 2 H), 2.45-2.30 (m, 4 H), 1.7-
1.5 (m, 4 H), 1.25 (t, J ) 7 Hz, 3H). Anal. Calcd: C, 59.10%;
H, 2.93%; N, 5.61%. Found: C, 59.39%; H, 3.04%; N, 6.14%.
1
Polymer 12 (Tg 175 °C). H NMR (CDCl3) δ: 8.31 (d, J )
8.5 Hz, 2 H), 8.12-7.8 (m, ∼16 H), 7.6-7.55 (m, ∼10 H), 7.34
(d, J ) 8.8 Hz, 4 H), 7.18 (br s, 3 H), 7.01 (d, J ) 8.8 Hz, 4 H),
6.79 (d, J ) 8.5 Hz, 2 H), 5.08 (s, 2 H), 4.29 (t, J ) 6 Hz, 2 H),
3.68 (t, J ) 6 Hz, 2 H), 3.53 (q, J ) 7 Hz, 2 H), 2.40-2.25 (m,
4 H), 1.70-1.55 (m, 4 H), 1.27 (q, J ) 7 Hz, 3 H). Anal. Calcd
for 12: C, 59.10%; H, 2.93%; N, 5.61%. Found: C, 58.91%; H,
2.98%; N, 5.92%.
Reaction of Polyimide with Hexamethyldiisocyanate
(6). A homogeneous THF (10 mL) solution containing 4 (1.00
g, ∼0.63 mequiv of -CH2OH) and triethylamine (5 mL) was
heated to 55 °C and then treated with 1,6-diisocyanatohexane
(∼3.0 mL, 18.4 mmol), and the reaction continued at 55 °C
for 2 h with stirring. The warm solution was poured into 200
mL of ether to precipitate the polymer. The off-white polymer
was collected on a glass frit and washed with additional ether
(∼200 mL) and then hexanes (∼300 mL). The polymer was
dried under reduced pressure to afford the functionalized
isocyanate polymer 6 (1.00 g, 90%). Selected 1H NMR (DMSO-
d6) data: δ 4.98 (s, 2H, CH2O), 3.1-2.8 (m, 4 H, CH2NCO and
CH2NHCO), 1.4-1.0 (m, 8 H, CH2’s). IR (film) νNdCdO 2271
1
Polymer 13 (Tg 156 °C). H NMR (CDCl3) δ: 8.30 (d, J )
8.5 Hz, 2 H), 8.10-7.80 (m, ∼16 H), 7.65-7.55 (m, ∼10 H),
7.41 (d, J ) 8.8 Hz, 4 H), 7.16 (d, J ) 8.8 Hz, 4 H), 6.78 (br s,
2 H), 6.75-6.68 (m, 2 H), 5.06 (s, 2 H), 4.29 (t, J ) 6 Hz, 2 H),
3.68 (t, J ) 6 Hz, 2 H), 3.53 (q, J ) 7 Hz, 2 H), 2.40-2.25 (m,
4 H), 1.70-1.40 (m, 4 H), 1.35-1.10 (m, 11 H). Anal. Calcd
for 13: C, 59.75%; H, 3.21%; N, 5.42%. Found: C, 59.89%; H,
3.46%; N, 5.72%.
Acknowledgment. The authors are grateful to the
Office of Naval Research for funding of this work. We
thank Mr. Dan Bliss for his technical assistance in
obtaining SEC data on the polymers and Dr. Geoff
Lindsay for partaking in many helpful discussions.
cm-1
.
Reaction of the Pendent Isocyanate with DR-1 (7). A
methylene chloride (20 mL) solution of polymer 6 (1.00 g) was
treated with DMAP (203 mg) and stirred for 10 min. DR-1
(0.314 g, 1.00 mmol) was added in one portion, and the
resulting dark red solution was allowed to stir overnight with
stirring under nitrogen. The solution was poured into MeOH
(300 mL) to precipitate the polymer. The polymer was collected
on a glass-frit and washed with MeOH (200 mL), ether (150
mL), MeOH (100 mL), and ether (100 mL) and then dried
under reduced pressure for 16 h. This afforded polymer 7 (0.96
g, 74%, Tg 199 °C) as a bright red powder. 1H NMR (CDCl3) δ:
8.30 (d, J ) 8.5 Hz), 8.10-7.82 (m), 7.70-7.55 (m), 7.41 (d, J
) 8.8 Hz), 7.16 (d, J ) 8.8 Hz), 6.9-6.70 (m), 5.06 (s, 2 H),
4.77 (br s, NH), 4.25 (m, OCH2CH2N), 3.68 (m), 3.50 (m), 3.30-
2.90 (m, -CH2NH), 1.7-1.2 (m, CH2’s and CH3). Anal. Calcd:
C, 58.66%; H, 3.12%; N, 6.69%. Found: C, 58.18%; H, 3.27%;
N, 6.55%.
References and Notes
(1) Resinger, J. J.; Hillmyer, M. A. Prog. Polym. Sci. 2002, 27,
971-1005 and references therein.
(2) Guillier, F.; Orain, D.; Bradley, M. Chem. Rev. 2000, 100,
2091-2157 and references therein. Krchnak, V.; Holladay,
M. W. Chem. Rev. 2002, 102, 61-91 and references therein.
(3) Wang, Y.; Erdogan, B.; Wilson, J. N.; Bunz, U. H. F. Chem.
Commun. 2003, 14, 1624-1625. Wagner, M.; Nuyken, O.
Macromolecules 2003, 36, 6716-6721 and references therein.
(4) Burland, D. M.; Miller, R. D.; Walsh, C. A. Chem. Rev. 1994,
94, 31 and references therein Davey, M. H.; Lee, V. Y.; Wu,
L. M.; Moylan, C. R.; Volksen, W.; Knoesen, A.; Miller, R.
D.; Marks, T. J. Chem. Mater. 2000, 12, 1679-1693 and
references therein. Gubbelmans, E.; Verbiest, T.; Van Beylen,
M.; Persoons, A.; Samyn, C. Polymer 2002, 43, 1581-1585.
Lee, J. Y.; Bang, H. B.; Kang, T. S.; Park, E. J. Eur. Polym.
J. 2004, 40, 1815-1822 and references therein.
General Procedure for the Modification of PI-CH2OH
with a Diacid Chloride. A THF solution (10 mL) containing
4 or 5 (1.00 g, 0.53 mequiv), 2,6-di-tert-butylpyridine (200 mg,
0.97 mmol), and DMAP (20 mg, 0.16 mmol) was treated with
the appropriate diacid chloride (∼7 mmol). Some precipitate
forms over the reaction a period of 16 h, and the mixture was
allowed to react with stirring for time ranging from 16 to 20
h. The mixture is filtered through a plug of Celite (1 × 2.5
cm) into a rapidly stirred solution of ether (150 mL). The
polymer was collected by filtration taking care to protect the
solution and polymer from air and then dried under reduced
pressure (0.2 Torr) in Schlenk flask for 1 h to afford ∼800-
900 mg of modified polymer. The material does contain
residual ether (∼5 wt %) and is used immediately for the next
reaction without purification.
(5) For examples of Mitsunobu coupling of dyes to phenol-
containing polyimides see: Chen, T. A.; Jen, A. K.-Y.; Cai,
Y. M. J. Am. Chem. Soc. 1995, 117, 7295-7296. Marder, S.
R.; Kippelen, B.; Jen, A. K.-Y.; Peyghambarian, N. Nature
(London) 1997, 338, 845. Saadeh, H.; Yu, D.; Wang, L. M.;
Yu, L. P. J. Mater. Chem. 1999, 9, 1865.
(6) For additional examples of chemical modification of polyimide
backbones containing phenolic groups see: Ueda, M.; Na-
kayama, T. Macromolecules 1996, 29, 6427-6431. Yu, H.-
S.; Yamashita, T.; Horie, K. Macromolecules 1996, 29, 1144-
1150. Chen, H.; Yin, J. J. Polym. Sci., Part A: Polym. Chem.
2004, 42, 1735-1744. For pendent carboxylic acid groups

Macromolecules, Vol. 38, No. 24, 2005
Hydroxymethyl-Functionalized Polyimides 10021
see: Wind, J. D.; Staudt-Bickel, C.; Paul, D. R.; Koros, W.
A., Eich, M., Kuzyk, M. G., Eds.; Proc. SPIE; SPIE: Belling-
ham, WA, 2004; Vol. 5517, pp 175-186.
Macromolecules 2003, 36, 1882-1888.
(10) Zhang, C.; Dalton, L. R.; Oh, M. C.; Zhang, H.; Steier, W. H.
Chem. Mater. 2001, 13, 3043-3050.
(7) Luo, J.; Haller, M.; Li, H.; Tang, H. Z.; Jen, A. K.-Y.; Jakka,
K.; Chou, C.-H.; Shu, C.-F. Macromolecules 2004, 37, 248-
250 and references therein.
(8) An alternative and less efficient synthetic route to compound
2a has been reported previously: Yang, G.; Jikei, M.;
Kakimoto, M. Macromolecules 1999, 32, 2215-2220.
(11) Wu, L. M.; Knudsen, A. J. Polym. Sci., Part B: Polym. Phys.
2001, 39, 217.
(12) Banach, M. J.; Clarson, S. J.; Beaucage, G.; Benkoski, J.;
Mates, T.; Kramer, E. J.; Vaia, R. A. J. Appl. Polym. Sci.
2002, 86, 2021.
(9) Guenthner, A. J.; Pentony, J. M.; Lindsay, G. A. In Linear
and Nonlinear Optics of Organic Materials IV; Norwood, R.
MA0503920