Evidence of entropy-driven bistability through 15N NMR analysis of a temperature- and solvent-induced, chiroptical switching polycarbodiimide

Article abstract of DOI:10.1021/ja4098803

The thermo- and solvo-driven chiroptical switching process observed in specific polycarbodiimides occurs in a concerted fashion with large deviations in specific optical rotation (OR) and CD Cotton effect as a consequence of varying populations of two dis

Full text of DOI:10.1021/ja4098803

Article
pubs.acs.org/JACS
Evidence of Entropy-Driven Bistability through 15N NMR Analysis of a
Temperature- and Solvent-Induced, Chiroptical Switching
Polycarbodiimide
James F. Reuther†,‡ and Bruce M. Novak*,‡
†Department of Chemistry, North Carolina State University, 2620 Yarbrough Drive, Raleigh, North Carolina 27695, United States
‡Department of Chemistry and Alan G. MacDiarmid NanoTech Institute, University of Texas Dallas, 800 West Campbell Road,
Richardson, Texas 75080, United States
S
* Supporting Information
ABSTRACT: The thermo- and solvo-driven chiroptical switching
process observed in specific polycarbodiimides occurs in a concerted
fashion with large deviations in specific optical rotation (OR) and CD
Cotton effect as a consequence of varying populations of two distinct
polymer conformations. These two conformations are clearly visible in
the 15N NMR and IR spectra of the 15N-labeled poly(15N-(1-naphthyl)-
N′-octadecylcarbodiimide) (Poly-3) and poly(15N-(1-naphthyl)-15N′-
octadecylcarbodiimide) (Poly-5). Using van’t Hoff analysis, the
enthalpies and entropies of switching (ΔHswitching; ΔSswitching) were
calculated for both Poly-3 and Poly-5 using the relative integrations of
both peaks in the 15N NMR spectra at different temperatures to measure
the populations of each state. The chiroptical switching (i.e.,
transitioning from state A to state B) was found to be an endothermic
process (positive ΔHswitching) for both Poly-3 and Poly-5 in all solvents
studied, meaning the conformation correlating with the downfield chemical shift (ca. 148 ppm, state B) is the higher enthalpy
state. The compensating factor behind this phenomenon has been determined to be the large increase in entropy in CHCl3 as a
result of the switching. Herein, we propose that the increased entropy in the system is a direct consequence of increased disorder
in the solvent as the switching occurs. Specifically, the chloroform solvent molecules are very ordered around the polymer chains
due to favorable solvent−polymer interactions, but as the switching occurs, these interactions become less favorable and disorder
results. The same level of solvent disorder is not achieved in toluene, causing the chiroptical switching process to occur at higher
temperatures.
of alterations to self-assembly of the polymer chains in gels,11
INTRODUCTION
■
liquid crystals,12 and Langmuir−Blodgett films.13 Specific
Synthesis and control of molecular switches presents important,
significant challenges in designing nanomachines capable of
minute changes to the overall primary or secondary structure at
the molecular level. Bistability, the presence of two specific
molecular structures that can interconvert between one another
upon influence from external stimuli, is a necessary attribute in
all molecular switches. Various external stimuli have been
molecular switch functional groups have also been incorporated
into several polymer systems including polymethacrylates,14
polypeptides,15 DNA,16 and polyisocyanates17 to name a few.
These systems offer a vast amount of potential applications in
the field of materials science.
In addition, formation of optically active, helical macro-
molecules with an excess single-handed screw sense has been an
shown to cause this phenomenon including alterations of
photochemical irradiation,1 solvent,2 heat,3 pressure,4 mag-
evergrowing field of interest in the past two decades due to
netic/electronic fields,5 or pH change,6 with irradiation and
their potential uses as asymmetric polymeric catalyst supports,
chiral sensors, polymeric liquid crystals, and/or stationary
heat being the predominant stimuli studied. Chiroptical
switches undergo molecular modifications that precisely alter
the chirality of the compound in question. The alteration to the
phases in chiral separations.18−26 Throughout this time, novel,
helical macromolecular scaffolds have emerged including
systems such as oligio-27 and poly(m-phenylene ethyny-
molecular structure of the switch can vary from helical
lenes),28,29 polyisocyanates,30,31 polyisocyanides,32−34 poly-
inversion7 to cis−trans isomerization8 to cyclization9 and
(phenyl acetylenes),35,36 poly(quinaxoline-2,3-dials),37,38 and
typically can be monitored by polarimetry, electronic circular
dichroism (ECD), vibrational circular dichroism (VCD), and
Raman spectroscopy.10 Applying these molecular switch
Received: September 27, 2013
functionalities to macromolecular systems offers possibilities
Published: December 6, 2013
© 2013 American Chemical Society
19292
dx.doi.org/10.1021/ja4098803 | J. Am. Chem. Soc. 2013, 135, 19292−19303

Journal of the American Chemical Society
Article
Figure 1. Chiroptical switching property of Poly-1 and Poly-2 observed in dilute solutions was hypothesized to be caused by the shutter-like rotation
of the aryl pendant group.
polycarbodiimides39 to name a few. Polycarbodiimides, a
relatively young class of helical macromolecules, have been an
intense area of interest in the Novak group since the discovery
of the controlled “living” Ti(IV)-mediated polymerization of
carbodiimides in 1994 due to their attractive properties such as
liquid crystallinity,40,41 high helical inversion barrier,42 and
chiroptical switching.43−46
In 2005, Tang et al. reported the first polycarbodiimide
possessing reversible solvent- and temperature-induced chirop-
tical switching in toluene.45 The achiral monomer was
polymerized with the chiral [(R)-(3,3′-dibromo-2,2′-
binaphthoxy]Ti(IV) di-tert-butoxide catalyst to form an
optically active polymer with excess single-handed helical
sense. The specific optical rotation (OR; [α]) and ECD cotton
effect of poly(N-(1-anthryl)-N′-octadecyl carbodiimide) (Poly-
1) were able to be modulated reversibly from positive to
negative values upon increasing the temperature of the dilute
polymer solution above 38.5 °C in toluene. In addition, doping
the polymer solution with 10% (v/v) THF or greater caused
the specific OR and ECD Cotton effect to exhibit the low
negative values observed in pure toluene at high temperatures.
Poly-1, however, showed no optical temperature dependence in
THF or CHCl3, and the ECD Cotton effect remained negative
at all temperatures.
Three possible molecular motions were hypothesized to be
the cause of this phenomenon including helical inversions,
imine inversions, and/or rotation about the N−Canthracene bond.
Full racemization of the polymer backbone for Poly-1 required
substantially more energy (ca. 100 h at +80 °C)42 to occur than
the low-energy switching process. The reversibility of the
process also contests the possibility that helical inversions play a
role since more than four heating/cooling cycles were
undergone without any significant racemization occurring.
Also, imine configuration interconversions in small molecule
analogs have been reported in the range of 20−26 kcal/mol
converting to analogous temperatures of 50−180 °C.47
Coupling the data collected with experimental and theoretical
VCD data, Tang et al. ruled out both helical inversions and
imine inversions as viable sources for the chiroptical changes
observed leaving rotation about the N−Caryl bond in a shutter-
like motion as a function of temperature and solvent polarity
(Figure 1).46 Upon influence of solvent or temperature, the aryl
groups synchronously reorient themselves to the slightly higher
energy state (B) aligning in the same general direction as the
helical director. The more polar solvents (CHCl3 or THF)
were believed to stabilize a more polar, high-energy state with
their signature low negative OR and ECD cotton effects.
Later in 2010, Kennemur et al. reported a new, more versatile
polycarbodiimide possessing amplified chiroptical switching
properties in numerous different solvents.43 Poly(N-(1-
naphthyl)-N′-octadecylcarbodiimide) (Poly-2) contains the
naphthyl pendant group capable of acting as the rotating flap
much like the more sterically bulky anthracene moiety. As a
result, the optical properties were enhanced with net specific
OR deviations of more than 1700° and the ability to switch, i.e.,
pass from positive OR to negative, in eight different solvents.
The different solvents shown to induce chiroptical switching
in Poly-2 are decisively diverse and include aromatic hydro-
carbons (toluene, xylenes, ethylbenzene, propylbenzene),
halogenated solvents (CHCl3, 1,2-dichlorobenzene, chloroben-
zene), and ethers (THF). Each solvent has varied effects on the
chiroptical switching of Poly-2, but chloroform, THF, and
toluene were found to possess the largest amplitude of
chiroptical switching. At first examination of Poly-1, the
solvent effects on the switching process seem intuitively
correlated to the polarity of the solvent, but no such trend
exists between the solvent used and the observed switching of
Poly-2.
Recently, we reported the use of 15N NMR spectroscopy to
directly probe the regioregularity of various nitrogen-15-
enriched polycarbodiimides.48,49 All polycarbodiimides wielding
two sterically inequivalent pendant groups, i.e., aromatic vs
primary aliphatic groups, were discovered to be completely
regioregular with the more bulky substituents solely located on
the CN backbone imine nitrogen. Poly-2 was initially chosen
for 15N enrichment to elucidate the regioregularity of
asymmetric polycarbodiimides and the presence of two C
N imine stretches in the IR spectrum.
The corresponding monomer was polymerized using the
chiral (R)-BINOL Ti(IV) diisopropoxide catalyst (Cat-1(R))
to form poly(15N-(1-naphthyl)-N-octadecylcarbodiimide)
(Poly-3) with an excess helical sense. This polymer was
found to be completely regioregular as evidence by the single
peak present in the 15N NMR spectra in toluene-d8. This result
along with the VT-IR data suggests that examining the imine
CN stretch in the FTIR spectra is not a sufficient handle for
regioregularity determination and that the stretch may offer
some insight into the switching process.
Herein, we propose a two-state model where the chiroptical
switching process is a consequence of two distinct polymer
conformations. We report the effect of solvent and temperature
on the 15N NMR spectra of Poly-3 and the newly synthesized,
dilabeled poly(15N-(1-naphthyl)-15N-octadecylcarbodiimide)
(Poly-5). In addition, the solvent and temperature effects
19293
dx.doi.org/10.1021/ja4098803 | J. Am. Chem. Soc. 2013, 135, 19292−19303

Journal of the American Chemical Society
Article
observed in the 15N NMR are compared with the specific OR,
ECD, and IR data of Poly-3. The synthesis and 15N NMR
analysis of poly(15N-(2-naphthyl)-N-octadecylcarbodiimide
(Poly-4) is also reported to compare with the chiroptical
switching Poly-3 and -5. The structures of Poly-3, -4, and -5
are shown in Figure 2.
nitrogen-15 enriched and used as received. Deuterated
chloroform-d (Sigma Aldrich, Milwaukee, MI), THF-d8
(Sigma Aldrich, Milwaukee, MI), toluene-d8 (Cambridge
Isotope Laboratories, Cambridge, MA), dichloromethane-d2
(Sigma Aldrich, Milwaukee, MI), and benzene-d6 (Cambridge
Isotope Laboratories, Cambridge, MA) were all purchased as
99.5% deuterated NMR solvents and used as received. Solvents
used in polymerizations of carbodiimide monomers were
distilled prior to use. All other solvents and reagents were
also used as received.
Instrumentation. 1H NMR spectra were collected on
Varian Mercury 400 and 300 MHz NMR spectrometers with
Oxford Narrow Bore superconducting magnets. The majority
of 15N NMR and 13C NMR spectra were conducted on the
Varian Mercury 400 MHz Spectrometer. Variable-temperature
15N NMR spectra for van’t Hoff analysis were collected on a
new Bruker AVANCE III 500 MHz NMR spectrometer for
better resolution. IR spectra were collected on a Jasco model
410 FTIR spectrometer either cast onto a KBr salt plate or
pressed into a KBr pellet. Specific OR measurements were
conducted on a Jasco P-1050 polarimeter in spectrophoto-
metric-grade solvents purchased from Sigma Aldrich (Milwau-
kee, MI). Mass spectrometry data was collected by the NCSU
Department of Chemistry Mass Spectrometry Facility on the
Agilent Technologies 6210 LC-TOF Mass Spectrometer.
Molecular weight determination was performed using a
Shimadzu Prominence Modular HPLC/GPC system connected
to a refractive index (RI) detector relative to a polystyrene
Figure 2. 15N-labeled polycarbodiimides synthesized to further analyze
the chiroptical switching process first observed in Poly-1 and Poly-2.
EXPERIMENTAL SECTION
■
Materials. 15N-labeled ammonium chloride (15NH4Cl,
Cambridge Isotope Laboratories, Cambridge, MA), sodium
nitrate (Na15NO3, Cambridge Isotope Laboratories, Cam-
bridge, MA), and potassium phthalimide (Cambridge Isotope
Laboratories, Cambridge, MA) were purchased as 99%
Scheme 1. Synthesis Scheme for Isotopically-Enriched Poly-3 and Poly-5 Starting 15N-Labeled Sodium Nitrate and Potassium
Phthalimide
19294
dx.doi.org/10.1021/ja4098803 | J. Am. Chem. Soc. 2013, 135, 19292−19303

Journal of the American Chemical Society
Article
standard. A two-column system was employed consisting of an
Agilent Mixed-E and Mixed-C column. Polymer samples were
dissolved in HPLC-grade chloroform (ca. 1.8 mg/mL) with
0.5% (v/v) N,N-dimethylethylenediamine added to the sample.
Samples were passed through a 0.45 μm PTFE filter prior to
injection. The flow rate for all samples was 1.0 mL/min, and
the injection volume was 50 μL.
dissolved in 15 mL of DCM in an addition funnel. The amine
solution was then added dropwise into the cold, stirring
solution of thiophosgene. When the amine solution was
completely added, the ice bath was removed and the solution
was allowed to stir at room temperature for an additional 2.5 h.
The solution was then washed with brine (2 × 50 mL) and
dried over anhydrous sodium sulfate. The mixture was decanted
into a clean flask, and solvent was removed via rotatory
evaporation. Product was then further purified by column
chromatography (DCM as mobile phase) and collected as a
yellow solid. Product was then dried under high vacuum for 4
h.
Synthesis of 15N-Enriched Polycarbodiimides. Syn-
thesis schemes for the 15N-labeled Poly-3 and Poly-5 are
shown in Scheme 1. The procedure for synthesis of the labeled
1-naphthylamine and Poly-3 was previously reported. Synthesis
of the labeled isothiocyanate was carried out using a slightly
modified procedure reported by Kai et al.50 Synthesis the 15N-
octadecylphthalimide and removal of the phthalimide group
was performed following the procedure reported by Wood
albeit slightly modified.51 Synthesis and characterization of the
titanium catalysts used for polymerization has been previously
reported by Tian et al.52 Poly-4 was synthesized using the
original protocol for 15N-labeling polycarbodiimides reported
by DeSousa and Novak involving amidation of an acid chloride
with isotopically enriched ammonium chloride, Hofmann
rearrangement to the labeled methyl carbamate, and acidolysis
of the carbamate to the amine HCl salt.48 These reaction
procedures, yields, and characterization of compounds can be
found in the Supporting Information. Procedures for all
unreported reactions are outlined in the section below. For
the yields and characterization of all precursor compounds and
final 15N-enriched polycarbodiimides, see the Supporting
Information.
Formation of Asymmetric Ureas and Thioureas (5 and 7).
The specific amounts of reagents, yields, NMR data, MS data,
and FTIR data for each asymmetric urea are reported in the
Supporting Information.
The amine (1.0−1.2 equiv) was added to a 250 mL round-
bottom flask along with a stir bar and 70 mL of CHCl3. The
mixture was allowed to stir for 15 min prior to addition of the
corresponding isocyanate/isothiocyanate (1.0−1.2 equiv). The
reaction mixture was then allowed to stir at room temperature
when employing alkyl amines or heated to reflux when
employing aryl amines. In both cases, the solution was allowed
to stir for 10−15 h. Upon completion, solvent was removed
under reduced pressure and crude product was recrystallized
from ethanol. The product, typically a white powder, was
filtered and washed with ice-cold ethanol repeatedly. The solid
was then collected and dried under high vacuum.
Synthesis of 15N-Labeled Carbodiimide Monomers (6 and
8). Specific amounts of reagents, yields, NMR data, MS data,
Synthesis of 15N-Octadecylphthalimide (1). 15N-enriched
potassium phthalimide (1.03 g, 5.56 mmol, 1.0 equiv, 98%
isotope enriched) and 1.86 g of 1-bromooctadecane (5.59
mmol, ∼1.0 equiv) were added to a 250 mL round-bottom flask
along with ∼25 mL of DMF. The mixture was heated to 95 °C
and allowed to stir for 21 h. The flask was then allowed to cool
to room temperature, and the reaction mixture was diluted with
100 mL of DI water. Product was then extracted with diethyl
ether (4 × 50 mL), and the ether extracts were collected.
Extracts were dried over anhydrous sodium sulfate (Na2SO4)
and decanted into a clean flask. Solvent was then removed, and
the product was recrystallized from ethanol. Product was
filtered with suction as a white crystalline solid and dried under
high vacuum for 8 h.
and FTIR data are reported in the Supporting Information. 15
N
NMR data was not collected on any of the carbodiimide
monomers due to greater than 90 s delay times.
Dibromotriphenylphosphorane (PPh3Br2, 1.2 equiv) was
added to a round-bottom flask along with ∼5 mL of DCM.
The reaction flask was then placed under N2 atmosphere and
cooled to 0 °C in an ice bath. TEA (2.5 equiv) was then added
to the stirring mixture, and resulting vapors were allowed to
dissipate. 15N-labeled urea/thiourea (1.0 equiv) was then added
to the mixture, and the solution was allowed to stir until
completion (ca. 1−10 h) with the reaction progress monitored
by IR spectroscopy. Upon completion, the reaction was
quenched with ∼300 mL of pentane and the resulting
precipitate was filtered and washed with additional pentane.
Filtrate was collected, and solvent was removed via rotatory
evaporation. This quenching process was repeated twice more
to remove most of the byproducts from the system. Product
was further purified by column chromatography (DCM as
mobile phase) and collected. Solvent was removed, and the
monomer was dried under high vacuum to afford pure product.
Polymerization of Carbodiimides with Ti(IV) Catalysts.
All polymerizations were conducted in an inert N2 atmosphere
drybox. All glassware and solvents were dried prior to use for
any polymerizations. Polymerization times vary greatly depend-
ing on the nature of the monomer used. The preweighed
amount of catalyst was added directly to the monomer for
solvent-free polymerizations. For solution-based polymeriza-
tions, the specified amount of catalyst was weighed out in a
clean, dry vial and a catalyst solution was formed in dry CHCl3.
The necessary amount of catalyst solution was then injected
into the reaction vial containing the monomer and allowed to
stir for the allotted amount of time (see Supporting
Information). To form optically active polymers with a helical
Deprotection to n-Octadecylamine (2). 15N-Octadecyl
phthalimide (1.84 g, 4.61 mmol, 1.0 equiv) and 0.71 mL of
51% hydrazine hydrate (11.5 mmol, 2.5 equiv) were added to a
250 mL round-bottom flask along with 75 mL of ethanol. The
mixture was heated to reflux and allowed to stir for 5 h. Upon
cooling to room temperature, the resulting precipitate was
filtered off and washed repeatedly with DCM. Solvent was
removed, and product was redissolved in DCM. The solution
was then washed with 6 M NaOH (2 × 100 mL) and Brine (1
× 100 mL). The organic layer was then collected, dried over
anhydrous Na2SO4, and decanted into a clean flask. Solvent was
removed again via rotatory evaporation, and solid was collected
and dried under high vacuum to afford pure product as a white
solid.
Synthesis of 15N-Labeled 1-Napthylisothiocyanate (4).
Thiophosgene (0.64 mL, 6.35 mmol, 1.2 equiv) was added to
a 250 mL round-bottom flask along with 25 mL of DCM, and
the reaction mixture was cooled to 0 °C in an ice bath. 15N-
enriched 1-naphthylamine (1.01 g, 6.99 mmol, 1.0 equiv) and
triethylamine (TEA) (3.86 mL, 27.9 mmol, 4.0 equiv) were
19295
dx.doi.org/10.1021/ja4098803 | J. Am. Chem. Soc. 2013, 135, 19292−19303

Journal of the American Chemical Society
Article
Figure 3. 15N NMR spectra of Poly-3 in CDCl3, THF-d8, and toluene-d8 with corresponding specific values OR (c = 2.00 mg/mL, λ = 435 nm, T =
19 °C) in each solvent.
bias, Cat-1(R) and Cat-1(S) was utilized for chiral induction
upon polymerization of all polymers. All polymers were worked
up by redissolving the crude product in a minimal amount of
chloroform and precipitating the pure product upon addition to
methanol, a typical antisolvent. Polymer product was then
filtered, collected, and dried under high vacuum. The
Kennemur et al. albeit with slightly diminished amplitude
(see Supporting Information).
Effect of Solvent on 15N NMR Spectra of Poly-3. Taking
another look at the CN imine stretch in the IR spectra of
Poly-3 reveals an interesting result. As previously stated, the IR
spectrum of the polymer dissolved in CHCl3 displays two peaks
in the imine CN region. In this solvent, the peak at 1619
cm−1 appears larger in intensity than the lower wavenumber
peak in the region at 1602 cm−1. These two peaks were first
attributed to regioirregular placement of the pendant groups
along the polymer backbone but was later disproven by VT-
FTIR.53 It is noteworthy to point out that two peaks
corresponding to the imine nitrogen are lower in wavenumber
than the values reported by Kennemur et al. (ca. 21 and 16
cm−1, respectively).43,53 This is attributed to the significant 15N-
isotope shift experienced by the imine stretch in the IR spectra
upon specific isotope labeling. When Poly-3 is dissolved in
THF, however, the two peaks switch in intensities and the peak
at 1602 cm−1 becomes more pronounced, further disproving
regioirregularity. Switching the solvent to toluene results in
complete disappearance of the peak at 1619 cm−1, leaving a
single stretch at 1601 cm−1 corresponding to the imine bond.
These are the same peaks that change in intensity as a function
of temperature.
1
monomer-to-catalyst ratio, yield, H NMR, IR, and 15N NMR
data of each polymer are reported in the Supporting
Information as well as scans of the IR spectra.
15N NMR Collection of Precursor Compounds and
Polymers. 15N NMR collection of the polymers and precursor
compounds was achieved using the same procedure previously
reported.48,49 Each spectrum was referenced externally to 15N-
enriched benzamide set to δ = 0.00 ppm. For the specific
procedure of spectra acquisition see the Supporting Informa-
tion.
RESULTS AND DISCUSSION
■
Preparation of the mono- and dilabeled polymers, Poly-3 and
Poly-5, is displayed in Scheme 1. Molecular weight (MW)
determination for Poly-3, -4, and -5 via gel permeation
chromatography (GPC) was performed with an additive, N,N-
dimethylethylenediamine, in order to reduce the affinity of the
nitrogen-rich backbone to the stationary phase of the column.
The number-average MW (Mn), weight-average MW (Mw), and
polydispersity (PDI) for the newly reported polymers are as
follows: (a) Poly-3, Mn = 79 000 Da, Mw = 200 000 Da, PDI =
2.53; (b) Poly-4, Mn = 34 000 Da, Mw = 81 000 Da, PDI =
2.38; (c) Poly-5, Mn = 60 000 Da, Mw = 140 000 Da, PDI =
2.33. Variable-temperature polarimetry of Poly-3 displays a
similar profile to the chiroptical switching reported by
Quantification of the populations of the two states is not
possible using IR spectroscopy because of the substantial
overlap present. Hence, we turned to 15N NMR to further
confirm the proposed two-state model and quantify their
relative populations. 15N NMR spectra of Poly-3 dissolved in
CDCl3 (top), THF-d8 (middle), and toluene-d8 (bottom) are
shown in Figure 3. A similar trend to that observed in the IR
spectra of the polymer dissolved in each solvent is seen in the
19296
dx.doi.org/10.1021/ja4098803 | J. Am. Chem. Soc. 2013, 135, 19292−19303

Journal of the American Chemical Society
Article
Figure 4. 15N NMR spectra of Poly-3 in DCM-d2 and benzene-d6 displays predominantly one peak in each, albeit the opposite, signifying that the
majority of the polymer chains are in a single conformation. Spectrum of Poly-4 in CDCl3 displays one peak unlike the very similar Poly-3, verifying
that when chiroptical switching is removed there is the presence of two specific conformations.
15N NMR spectra. In CDCl3, two peaks are present with the
polymer system, so they were the natural choice for these
inaugural NMR studies. In an attempt to establish the
population extremes of the system, multiple solvents were
studied to determine which solvents provide the maximum
positive values and minimum negative values of OR. At room
temperature, dichloromethane was found to provide the highest
specific OR, ca. [α]22435 nm = 625°, while benzene provided the
lowest value, ca. [α]22435 nm = −152°, of the solvents tested. The
specific OR of Poly-3 in DCM and benzene, however, does not
vary greatly as a function of temperature and never passes
through 0° at all feasible temperatures. 15N NMR of the
polymer in DCM-d2 displays one peak corresponding to the
imine nitrogen at δ = 139 ppm, which agrees with the trend
between the vast majority of the polymer chains adopting state
A at higher positive specific OR values. Also, in concurrence
with the observed trends, the spectrum in benzene-d6 displays
predominantly one peak at δ = 148 ppm. This suggests that the
vast majority of the naphthyl pendant groups attached to the
imine nitrogen are in a single conformation in both DCM and
benzene, albeit opposite from one another. In addition, the
chemical shifts of the two distinct states remained constant in
both benzene-d6 and DCM-d2, further supporting the
hypothesis that the two conformations are the same in all
solvents. It should be noted that the higher field peak at ca. 138
ppm never fully disappears, and whether this is a “non-
responsive” portion of the polymer chains or an inability to
uncover conditions allowing for full negative OR values remains
to be seen.
peak at δ = 148 ppm being the predominant chemical shift in
the spectra and the relative integration of each peak equal to
about 2:1. In THF-d8, both peaks are shifted slightly farther
downfield and switch in intensities with the relative integration
of the peaks at δ = 153 and 143 ppm equal to about 1:2,
respectively. In toluene-d8, the downfield peak is nearly
completely eliminated with one peak at δ = 139 ppm in the
imine region. The polymer is still regioregular since both 15N
NMR peaks correspond to the imine nitrogen. Both the IR and
the 15N NMR peaks corresponding to the imine nitrogen
appear to be directly associated with the polymer conformation
leading to the different specific OR measured. High positive OR
values correlate to the upfield peak in the 15N NMR and the IR
stretch at 1602 cm−1 (state A), whereas the downfield peak and
the peak at 1619 cm−1 correlate to the lower negative values
(state B). This data supports our hypothesis that the switching
process occurs in a concerted fashion between two states with
the observed OR changes resulting from the equilibrium
populations of these two distinct conformations. Chemical
shifts of both states remain relatively constant in all three
solvents, supporting the polymer chains adopting largely similar
conformations in each solvent. It is also noted that the rate of
switching between these two states is very rapid when
compared to the NMR time scale, so measuring the kinetics
of the switching process is not possible using this technique.
Chloroform, THF, and toluene were found to be the most
versatile solvents for the chiroptical switching process in this
19297
dx.doi.org/10.1021/ja4098803 | J. Am. Chem. Soc. 2013, 135, 19292−19303

Journal of the American Chemical Society
Article
Figure 5. 15N NMR spectra of Poly-5 in CDCl3, THF-d8, and toluene-d8 at 20 °C showing the effect of the chiroptical switching phenomenon on
the amine nitrogen.
The chiroptical switching properties of the polymer are
completely removed upon changing the connectivity of the
naphthyl pendant group as evident by switching the imine
connection from the 1 position to the 2 position of the ring.43
For this reason, poly(15N-(2-naphthyl)-N-octadecylcarbodii-
mide) (Poly-4) was synthesized using the original protocol
designed by DeSousa et al.48 15N NMR spectra of Poly-4 in
CDCl3 displayed only one peak in the imine region unlike the
very similar, but differently connected, Poly-3. Also, the peak is
shifted farther upfield close to the 15N NMR chemical shift of
poly(15N-phenyl-N-hexylcarbodiimde), suggesting that Poly-4
is closer conformationally to this polymer rather than Poly-3.
This suggests that the presence of two conformations is a
function of the attachment of the naphthyl ring just like the
optical rotation dependence on solvent and temperature. 15N
NMR spectra of Poly-4 in CDCl3 and Poly-3 in DCM-d2 and
benzene-d6 are shown in Figure 4.
Effect of Switching Process on the Amine Nitrogen.
The chiroptical switching process is known to have a great deal
of effect on the imine nitrogen and its environment. It is still
unclear, however, whether the amine nitrogen is affected by the
switching process. To probe the effect of the process on the
amine nitrogen, a new synthesis scheme was created to produce
the dilabeled poly(15N-(1-naphthyl)-15N-octadecylcarbodii-
mide) (Poly-5). The enantiomer of the catalyst used for
polymerization of Poly-3 was employed to ensure that the
opposite handed helix of Poly-5 formed upon polymerization
would not affect the presence of the peaks in the imine region
of the 15N NMR. The specific OR profiles of Poly-5 reversed
compared to Poly-3 (see Supporting Information), confirming
the opposite handed helix was formed upon polymerization.
The synthesis scheme for Poly-5 is shown in Scheme 1. The
initial 15N NMR spectra were collected in the same solvents as
in studies of Poly-3 (CDCl3, THF-d8, and toluene-d8 in Figure
5).
The resulting peaks corresponding to the imine nitrogen of
Poly-5 mimicked the peaks seen in the spectra of Poly-3 with
the same pattern seen in all three solvents. The amine nitrogen
also displayed two peaks in a mirror image of the imine
nitrogen chemical shifts. In CDCl3, the peak at δ = 148 ppm for
imine nitrogen and the peak at δ = 11.5 ppm corresponding to
the amine nitrogen predominate in the spectra where the peaks
at δ = 138 and 19.9 ppm correlate with the minority population
of polymer chains. The spectrum in THF-d8 continues the
trend with the intensities of both sets of peaks switching.
Another interesting observation is the shift downfield of all four
peaks. This was also observed in the spectrum of Poly-3, where
the two imine nitrogen chemical shifts migrate downfield a total
of 5 ppm. This can most likely be attributed to some specific
solvent−backbone interaction that adds to the magnetic
shielding of the backbone nitrogens. In toluene-d8, the peaks
are also shifted slightly downfield. This observation differs from
the peak identified in the imine region of Poly-3 in toluene-d8
in which the peak corresponding to the imine nitrogen is
located at δ = 139 ppm. The source of this observation is still
unknown and under further investigation. The populations of
polymer conformations, however, correlate perfectly with the
expected trend seen with Poly-3. The peaks δ = 151 and 14.7
19298
dx.doi.org/10.1021/ja4098803 | J. Am. Chem. Soc. 2013, 135, 19292−19303

Journal of the American Chemical Society
Article
ppm are nearly completely diminished with only one peak in
both the imine and the amine regions.
toluene at 45 °C is 98.9°, slightly less than the polymer in THF
(ca. [α]19 = 175°), so the slightly increased prominence of
435
Using Variable-Temperature 15N NMR To Measure the
Thermodynamics of Switching. In addition to changing the
solvent composition of dilute polymer solutions, altering the
temperature in solutions of Poly-3 and Poly-5 causes the
specific OR and CD to vary drastically in a reversible fashion.
As expected, variable-temperature 15N NMR provided further
insight into how the populations of each specific polymer
conformation are modulated as a function of temperature. 15N
NMR spectra at 20, 30, 45, and 60 °C in toluene-d8 are
displayed in Figure 6. When the temperature is increased and
the downfield peak correlates with the trend. At 60 °C, the
trend continues the conformation population analogous to the
downfield peak growing in intensity integrating relative to the
upfield chemical shift to 1:1.5, respectively. In all solvents, state
B is populated with increasing abundance as temperatures
increase as evident by the 15N chemical shift at ca. 150 ppm and
the negative specific OR values.
VT-15N NMR spectra of Poly-3 and Poly-5 in THF-d8 are
significantly different with the populations of each conforma-
tion altering much more drastically than in toluene-d8 (see
Supporting Information). To observe the most significant
alterations of the polymer conformation in THF, temperatures
of 20, 30, and 40 °C were chosen for variable-temperature
studies. At 20 °C, the upfield chemical shift in the 15N NMR
spectrum of Poly-3 predominates significantly with the relative
integration of the downfield peak with respect to the upfield
peak equal to about 1:2.2. Increasing the temperature of the
NMR solution to 30 °C results in an increase in the relative
population of the downfield chemical shift at 151 ppm and a
decrease in the specific OR from [α]20 = 175° to [α]30
=
435
435
−2.0°. Further heating the solution to 40 °C causes the specific
OR to reach the minimum value in THF of [α]40435 = −55° and
the relative ratio of the downfield peak to the upfield peak equal
to about 1:1.1. Interestingly, the populations of each Poly-3
conformation never favor the state corresponding to the
downfield shift in THF similar to when dissolved in toluene-d8;
however, the population change is slightly more significant.
In CDCl3, the VT-15N NMR spectra display a considerably
more pronounced population change over a smaller range of
temperatures when compared to toluene-d8 and THF-d8
(Figure 7). At 0 °C, the specific OR of Poly-3 is at the
maximum value (ca. [α]19 = 418°) in this particular solvent.
435
15N NMR analysis at this temperature reveals a single peak in
the spectra matching nicely with the trend seen in previous
spectra. The conformation corresponding to the downfield shift
grows significantly in population with respect to the upfield
peak upon heating the polymer solution to 10 °C with relative
integrations of the two peaks equal to about 1:2. The downfield
conformation becomes even more prominent when the
temperature is increased further to 15 °C. At 20 °C, the
equilibrium shifts in favor of the conformation corresponding
with the downfield shift with the relative integrations of the two
peaks equal to ∼1.2:1. VT-15N NMR spectra of Poly-5 in
toluene-d8, THF-d8, and CDCl3 mimic the spectra collected for
Poly-3 in the same solvents (see Supporting Information). This
data continues to match nicely with the observed trend
between the chiroptical changes and the population change of
two distinct conformations.
Since we can now quantify the specific proportions of each
polymer conformation in solution at different temperatures, it is
now possible to calculate the change in enthalpy and entropy
during the equilibrium process. This can be accomplished using
the van’t Hoff relationship shown below in eq 1 where Keq is
the equilibrium constant, R is the gas constant (1.989 cal/(mol·
K)), and T is the temperature in Kelvin. This method is used
extensively to study biological macromolecular systems in order
to examine specific functions and dynamic behavior in DNA
and peptides.54,55
Figure 6. Variable-temperature 15N NMR spectra of Poly-3 in
toluene-d8 at 20, 30, 45, and 60 °C showing the conformational
dependence on the temperature of the polymer solution.
the specific OR of the polymer sample decreases the intensity
of the peak at δ =152 ppm slowly increases, suggesting that the
population of the secondary conformation corresponding to the
downfield chemical shift is also increasing. At 45 °C, the relative
integration of the downfield peak and the peak at δ = 142 ppm
reveals a ∼1:2 ratio, which mimics the polymers behaviors in
THF at room temperature. The specific OR of the polymer in
ΔHswitching
ΔSswitching
ln(Keq) = −
+
(1)
RT
R
19299
dx.doi.org/10.1021/ja4098803 | J. Am. Chem. Soc. 2013, 135, 19292−19303

Journal of the American Chemical Society
Article
Figure 8. van’t Hoff plot of Poly-3 and Poly-5 in toluene-d8, THF-d8,
and CDCl3 used to determine the thermodynamics of the chiroptical
switching process.
This is due to the larger quantity of polymer chains occupying
the higher energy state B compared to the switching process in
toluene. The driving force behind this phenomenon is the very
large entropy change associated with the switching in CHCl3.
The increase in the enthalpies and entropies of switching for
both polymers when the solvent is changed from toluene to
THF to chloroform implies that the change in disorder of the
system (polymer and solvent) is substantially larger when the
polymer is dissolved in CHCl3. If we assume that the entropy
change associated with the specific molecular motion of the
polymer will be the same in all solvents then the large
difference in ΔSswitching could be attributed to increased disorder
in the chloroform solvent molecules as the polymer undergoes
the switching process. This assumption is reasonable since both
conformations present in solution are the same or very similar
in all three solvents as evident by the comparable chemical shift
values. This is most likely due to ordered solvent−polymer
interactions between the chloroform solvent molecules and the
polymer chains adopting state A. As switching from state A to
state B occurs, the solvent−polymer ordering becomes less,
thus increasing the entropy of the system. An even stronger
solvent−polymer ordered interaction in toluene causes the
switching process to become less favorable and, in turn, inhibits
the polymer chains from predominantly adopting the higher
energy state B. We propose that the toluene solvent molecules
may associate with the polymer chains more favorably with
respect to CHCl3 through ordered π−π stacking between the
solvent and the naphthyl pendant groups. Due to this
increasingly ordered interaction, the switching phenomenon
from state A to state B is much less favorable in toluene due to
the diminished propensity of the toluene molecules to
dissociate upon switching. This hypothesis supports the higher
observed switching temperatures of Poly-3 and Poly-5 in
toluene as well as the limited number of polymer chains
adopting state B.
Figure 7. Variable-temperature15N NMR spectra of Poly-3 in CDCl3
at 0, 10, 15, and 20 °C displaying the distinct population shifts as the
temperature of the solution is altered.
For our system where the equilibrium consists of
interconversions of two specific states, the relative integrations
of the peaks in the 15N NMR spectra at various temperatures
allows for a qualitative measure of the specific proportions of
each conformation in solution. With these proportions, one can
now calculate Keq simply by dividing the relative concentration
of state B by the relative concentration of state A (i.e., Keq
=
[B]/[A]). Plotting the values of ln(Keq) vs 1/T should result in
a linear plot with the slope equal to −ΔHeq/R and the y
intercept equal to ΔSeq/R. Results of this analysis are shown in
Figure 8.
This analysis provides consistent results between different
batches of identical polymers. Results of the van’t Hoff analysis
and errors associated to the error of integration for both Poly-3
and Poly-5 are outlined in Table 1. van’t Hoff plots for Poly-5
match up very well with the plots of Poly-3 providing similar
energy values, further substantiating the validity of this analysis.
These polymers display positive enthalpy changes in all
solvents indicating that we are populating a higher energy state
as the conformations are shifted from state A to state B.
Interestingly, both Poly-3 and Poly-5 have the highest
observed ΔHswitching when dissolved in CHCl3 despite the
entire switching process occurring below room temperature.
Although absolute identification of the conformational
reorientation cannot be unambiguously determined via 15N
NMR, new insights can be drawn from the calculated energy
parameters, furthering the understanding of this dynamic
equilibrium. Herein, we propose two hypothesized pendant
group conformations (Figure 9) believed to be the two
observed states visible in the 15N NMR spectra of Poly-3 and
Poly-5. The naphthyl pendant group in the proposed; lower
energy state A is positioned out of the plane of the backbone
imine allowing for no p-orbital overlap between the conjugated
naphthyl ring and the imine π bond. Upon switching, the
naphthyl ring is hypothesized to reorient in plane with the
19300
dx.doi.org/10.1021/ja4098803 | J. Am. Chem. Soc. 2013, 135, 19292−19303

Journal of the American Chemical Society
Article
Table 1. Empirically Calculated van’t Hoff Enthalpies and Entropies of the Reversible, Chiroptical Switching Process Using the
Relative Integration the Nitrogen-15 NMR Peaks Corresponding to the Imine Nitrogen To Measure the Populations of Each
Polymer Conformation
Poly-3
Poly-5
solvent
toluene
THF
chloroform
toluene
THF
chloroform
a
specific OR changes
15−375°
5.8 ( 2.9)
17 ( 9.0)
+0.77
−50−406°
6.4 ( 3.3)
20 ( 11)
+0.36
−132−488°
15 ( 3.4)
53 ( 12)
−0.37
−99−3°
5.5 ( 1.4)
16 ( 4.5)
+0.71
−99−30°
6.1 ( 2.2)
19 ( 7.5)
+0.40
−76−46°
14 ( 2.4)
49 ( 8.3)
−0.39
ΔHswitching (kcal/mol)
ΔSswitching (cal·mol−1·K−1
)
b
ΔGswitching at 25 °C (kcal/mol)
a
b
All specific OR measurements recorded for 2.00 mg/mL dilute; polymer solutions at 435 nm. Free energy of switching estimated using the van’t
Hoff enthalpies and entropies.
Figure 9. Proposed conformational structures of states A and B represented in chemical structures, and optimized 24mer model showing the
hypothesized naphthyl orientations in and out of plane of the backbone imine nitrogen.
backbone imine, allowing for conjugation to occur. This state
also correlates with the higher energy state B due to the added
steric interactions between adjacent aryl pendant groups.
Currently, new studies are underway to pinpoint the exact
pendant group orientations using VCD spectroscopy coupled
with density functional theory (DFT) calculations.
(Table 1). The positive free energy values for both polymers in
toluene and THF match nicely with the higher switching
temperatures of the polymers in both solvents and the fact that
the equilibrium never shifts in favor of state B at any
temperature observed. In addition, the estimated free energy
of switching for Poly-3 and Poly-5 in chloroform reveals
negative ΔGswitching, which is also in concurrence with the
observed trend due to the entire switching process occurring
The free energy of the switching process was then estimated
using the van’t Hoff energies measured for Poly-3 and Poly-5
19301
dx.doi.org/10.1021/ja4098803 | J. Am. Chem. Soc. 2013, 135, 19292−19303

Journal of the American Chemical Society
Article
below room temperature and the equilibrium shifting in favor
of the downfield conformation at temperatures greater than 18
°C. This new data offers new insight in what has proven to be
an intensely interesting polymer behavior.
for their aid in NMR acquisition. We thank Dr. Dennis Smith
(UTD) and Ben Batchelor (UTD) for their help in the MW
determination of the reported polymers via GPC. We gratefully
acknowledge the NSF-MRI grant (CHE-1126177) used to
purchase the Bruker Advance III 500 NMR instrument. We
thank the NCSU Department of Chemistry Mass Spectrometry
Facility funded by the North Carolina Biotechnology Center
and NCSU Department of Chemistry for all mass spectral data
collected.
CONCLUSIONS
■
Synthesis of 15N-labeled poly(15N-1-naphthyl-N-octadecylcar-
bodiimide) (Poly-3) and poly(15N-1-naphthyl-15N′-octadecyl-
carbodiimide) (Poly-5) has provided ample new understanding
on the driving forces behind the solvent effects of chiroptical
switching process. We discovered evidence suggesting that the
process occurs in a two-state model and the populations can be
determined using 15N NMR spectroscopy. The changes in
specific OR results from a rapid equilibrium between these two
states and the position of equilibrium is a function of both
solvent and temperature. We further propose that the two
conformations are similar in all solvents, but the large
alterations in entropy are related to the disorder of the solvents
in the solvation sphere during the switching process. These
populations were found to possess potent temperature and
solvent dependence in all cases studied. Using variable-
temperature 15N NMR to measure the proportions of the
polymer chains in each specific conformation, we were able to
calculate the van’t Hoff enthalpies and entropies of switching.
This analysis provided different values for chloroform and
toluene associated with significantly larger changes in entropy
when the switching process occurs in chloroform. The
additional order of the system in toluene has contributions
from ordering of the local solvent molecules in some specific
manner around the polymer chains (possibly π−π interactions
between the toluene and the naphthyl pendant group). These
added interactions are believed to be the root of the higher
switching temperatures as well as the limited number of
polymer chains adopting state B in toluene. In addition, the free
energy of switching was calculated for both polymer systems,
showing that ΔGswitching is positive in toluene and negative in
chloroform at 25 °C. This also correlates with the trend due to
the switching process occurring below room temperature in
chloroform and above room temperature in toluene. To further
elucidate the mechanism of the chiroptical switching process,
VCD measurements and calculations are currently ongoing to
help identify the conformations present in the system.
REFERENCES
■
(1) Agati, G.; McDonagh, A. F. J. Am. Chem. Soc. 1995, 117, 4425−
4426.
(2) Suk, J.-m.; Naidu, V. R.; Liu, X.; Lah, M. S.; Jeong, K.-S. J. Am.
Chem. Soc. 2011, 133, 13938−13941.
(3) Zou, G.; Jiang, H.; Zhang, Q.; Kohn, H.; Manaka, T.; Iwamoto,
M. J. Mater. Chem. 2010, 20, 285−291.
(4) Canary, J. W.; Mortezaei, S.; Liang, J. Coord. Chem. Rev. 2010,
254, 2249−2266.
(5) Li, D.; Wang, Z. Y.; Ma, D. Chem. Commun. 2009, 1529−1531.
(6) Zhang, G.; Liu, M. J. Mater. Chem. 2009, 19, 1471−1476.
(7) Feringa, B.; Wynberg, H. J. Am. Chem. Soc. 1977, 99, 602−603.
(8) Murata, K.; Aoki, M.; Nishi, T.; Ikeda, A.; Shinkai, S. Chem.
Commun. 1991, 1715−1718.
(9) Yamaguchi, T.; Uchida, K.; Irie, M. J. Am. Chem. Soc. 1997, 119,
6066−6071.
(10) Feringa, B. L.; van, D. R. A.; Koumura, N.; Geertsema, E. M.
Chem. Rev. 2000, 100, 1789−1816.
(11) Effing, J. J.; Kwak, J. C. T. Angew. Chem., Int. Ed. Engl. 1995, 34,
88−90.
(12) Ikeda, T.; Tsutsumi, O. Science 1995, 268, 1873−1875.
(13) Yamazaki, I.; Ohta, N. Pure Appl. Chem. 1995, 67, 209−216.
(14) Oosterling, M. L. C. M.; Schoevaars, A. M.; Haitjema, K. J.;
Feringa, B. L. Isr. J. Chem. 1997, 36, 341−348.
(15) Ueno, A.; Takahashi, K.; Anzai, J.; Osa, T. J. Am. Chem. Soc.
1981, 103, 6410−6415.
(16) Mammana, A.; Carroll, G. T.; Areephong, J.; Feringa, B. L. J.
Phys. Chem. B 2011, 115, 11581−11587.
(17) Pijper, D.; Jongejan, M. G. M.; Meetsma, A.; Feringa, B. L. J.
Am. Chem. Soc. 2008, 130, 4541−4552.
(18) Green, M. M.; Peterson, N. C.; Sato, T.; Teramoto, A.; Cook,
R.; Lifson, S. Science 1995, 268, 1860−1866.
(19) Gellman, S. H. Acc. Chem. Res. 1998, 31, 173−180.
(20) Cornelissen, J. J. L. M.; Rowan, A. E.; Nolte, R. J. M.;
Sommerdijk, N. A. J. M. Chem. Rev. 2001, 101, 4039−4070.
(21) Hill, D. J.; Mio, M. J.; Prince, R. B.; Hughes, T. S.; Moore, J. S.
Chem. Rev. 2001, 101, 3893−4012.
ASSOCIATED CONTENT
* Supporting Information
■
S
(22) Nakano, T.; Okamoto, Y. Chem. Rev. 2001, 101, 4013−4038.
(23) Lam, J. W. Y.; Tang, B. Z. Acc. Chem. Res. 2005, 38, 745−754.
(24) Yashima, E.; Maeda, K. Macromolecules 2008, 41, 3−12.
(25) Yashima, E.; Maeda, K.; Iida, H.; Furusho, Y.; Nagai, K. Chem.
Rev. 2009, 109, 6102−6211.
(26) Kennemur, J. G.; Novak, B. M. Polymer 2011, 52, 1693−1710.
(27) Prest, P.-J.; Prince, R. B.; Moore, J. S. J. Am. Chem. Soc. 1999,
121, 5933−5939.
Further experimental characterization on all precursor com-
pounds and polymers including NMR, IR, specific OR, and
mass spectral data. This material is available free of charge via
the Internet at http://pubs.acs.org.
AUTHOR INFORMATION
Corresponding Author
E-mail: Bruce.novak@utdallas.edu
■
(28) Zhao, D.; Moore, J. S. J. Am. Chem. Soc. 2002, 124, 9996−9997.
(29) Banno, M.; Yamaguchi, T.; Nagai, K.; Kaiser, C.; Hecht, S.;
Yashima, E. J. Am. Chem. Soc. 2012, 134, 8718−8728.
(30) Green, M. M.; Gross, R. A.; Crosby, C., III; Schilling, F. C.
Macromolecules 1987, 20, 992−999.
Notes
The authors declare no competing financial interest.
(31) Lifson, S.; Green, M. M.; Andreola, C.; Peterson, N. C. J. Am.
Chem. Soc. 1989, 111, 8850−8858.
ACKNOWLEDGMENTS
■
Funding for this work was provided by the Faculty start-up
fund from the University of Texas at Dallas (UTD), the
Endowed Chair for Excellence at UTD, and the Howard
Schaeffer Distinguished Chair at NCSU. Our special thanks to
Dr. Sabapathy Sankar (NCSU) and Dr. Hien Nguyen (UTD)
(32) Nolte, R. J. M.; Van, B. A. J. M.; Drenth, W. J. Am. Chem. Soc.
1974, 96, 5932−5933.
(33) Van, B. A. J. M.; Nolte, R. J. M.; Naaktgeboren, A. J.; Zwikker, J.
W.; Drenth, W.; Hezemans, A. M. F. Macromolecules 1983, 16, 1679−
1689.
19302
dx.doi.org/10.1021/ja4098803 | J. Am. Chem. Soc. 2013, 135, 19292−19303

Journal of the American Chemical Society
Article
(34) Takei, F.; Yanai, K.; Onitsuka, K.; Takahashi, S. Angew. Chem.,
Int. Ed. 1996, 35, 1554−1556.
(35) Kwak, G.; Masuda, T. J. Polym. Sci., Part A: Polym.Chem. 2001,
39, 71−77.
(36) Aoki, T.; Kaneko, T.; Maruyama, N.; Sumi, A.; Takahashi, M.;
Sato, T.; Teraguchi, M. J. Am. Chem. Soc. 2003, 125, 6346−6347.
(37) Ito, Y.; Ihara, E.; Murakami, M. Angew. Chem., Int. Ed. 1992, 31,
1509−1510.
(38) Ito, Y.; Ihara, E.; Murakami, M.; Sisido, M. Macromolecules 1992,
25, 6810−6813.
(39) Goodwin, A.; Novak, B. M. Macromolecules 1994, 27, 5520−
5522.
(40) Kim, J.; Novak, B. M.; Waddon, A. J. Macromolecules 2004, 37,
1660−1662.
(41) Kim, J.; Novak, B. M.; Waddon, A. J. Macromolecules 2004, 37,
8286−8292.
(42) Tang, H.-Z.; Lu, Y.; Tian, G.; Capracotta, M. D.; Novak, B. M. J.
Am. Chem. Soc. 2004, 126, 3722−3723.
(43) Kennemur, J. G.; Clark, J. B.; Tian, G.; Novak, B. M.
Macromolecules 2010, 43, 1867−1873.
(44) Kennemur, J. G.; Kilgore, C. A.; Novak, B. M. J. Polym. Sci., Part
A: Polym. Chem. 2011, 49, 719−728.
(45) Tang, H.-Z.; Boyle, P. D.; Novak, B. M. J. Am. Chem. Soc. 2005,
127, 2136−2142.
(46) Tang, H.-Z.; Novak, B. M.; He, J.; Polavarapu, P. L. Angew.
Chem., Int. Ed. 2005, 44, 7298−7301.
(47) Jennings, W. B.; Boyd, D. R. J. Am. Chem. Soc. 1972, 94, 7187−
7188.
(48) DeSousa, J. D.; Novak, B. M. ACS Macro Lett. 2012, 1, 672−
675.
(49) Reuther, J. F.; D., D. J.; Novak, B. M. Macromolecules 2012, 45,
7719−7728.
(50) Kai, H.; Morioka, Y.; Koriyama, Y.; Okamoto, K.; Hasegawa, Y.;
Hattori, M.; Koike, K.; Chiba, H.; Shinohara, S.; Iwamoto, Y.;
Takahashi, K.; Tanimoto, N. Bioorg. Med. Chem. Lett. 2008, 18, 6444−
6447.
(51) Wood, G. W. J. Chem. Soc. 1953, 3327.
(52) Tian, G.; Lu, Y.; Novak, B. M. J. Am. Chem. Soc. 2004, 126,
4082−4083.
(53) Kennemur, J. G.; DeSousa, J. D.; Martin, J. D.; Novak, B. M.
Macromolecules 2011, 44, 5064−5067.
(54) Ambrosone, L.; Ragone, R. Int. J. Biol. Macromol. 2000, 27,
241−244.
(55) Lavelle, L.; Fresco, J. R. Nucleic Acids Res. 1995, 23, 2692−2705.
19303
dx.doi.org/10.1021/ja4098803 | J. Am. Chem. Soc. 2013, 135, 19292−19303