Phase-selective solubility of poly(N-alkylacrylamide)s

Article abstract of DOI:10.1021/ja0349498

The extent of the phase-selective solubility of poly(N-alkylacrylamide)s was studied by UV-vis and fluorescence spectroscopy using poly(N-isopropylacrylamide) and poly(N-octadecylacrylamide) as representative polar and nonpolar poly(N-alkylacrylamide)s in a mixture of polar and nonpolar thermomorphic solvents. Phase-selective solubilities of greater than 10000:1 were seen with each labeled polymer in polar and nonpolar solvents such as heptane and DMF or heptane and 90% EtOH-H₂O. Using a poly(N-acryloxysuccinimide) as a common precursor, a pool-split synthesis was devised to prepare a library of poly(N-alkylacrylamide)s whose members varied only in the size of their N-alkyl substituent. The solubilities of these library members were measured in both the polar and nonpolar phases of a thermomorphic heptane/90% EtOH-H₂O mixture at 25°C. Such solvent mixtures are miscible hot (70°C) and biphasic cold (25°C). The results show that poly(N-pentylacrylamide) is selectively soluble (>99.5%) in the polar EtOH-rich phase at rest. Poly(N-alkylacrylamide)s with larger N-alkyl groups are predominantly (C₆, 85%; C₇, 95%) or exclusively (>C8, >99.5%) in the heptane-rich phase at rest.

Full text of DOI:10.1021/ja0349498

Published on Web 06/11/2003
Phase-Selective Solubility of Poly(N-alkylacrylamide)s
David E. Bergbreiter,* Reagan Hughes, Jacqueline Besinaiz, Chunmei Li, and
Philip L. Osburn
Contribution from the Department of Chemistry, Texas A&M UniVersity, P.O. Box 30012,
College Station, Texas 77842-3012
Received March 2, 2003; E-mail: bergbreiter@tamu.edu
Abstract: The extent of the phase-selective solubility of poly(N-alkylacrylamide)s was studied by UV-vis
and fluorescence spectroscopy using poly(N-isopropylacrylamide) and poly(N-octadecylacrylamide) as
representative polar and nonpolar poly(N-alkylacrylamide)s in a mixture of polar and nonpolar thermomorphic
solvents. Phase-selective solubilities of greater than 10000:1 were seen with each labeled polymer in polar
and nonpolar solvents such as heptane and DMF or heptane and 90% EtOH-H2O. Using a poly(N-
acryloxysuccinimide) as a common precursor, a pool-split synthesis was devised to prepare a library of
poly(N-alkylacrylamide)s whose members varied only in the size of their N-alkyl substituent. The solubilities
of these library members were measured in both the polar and nonpolar phases of a thermomorphic heptane/
90% EtOH-H2O mixture at 25 °C. Such solvent mixtures are miscible hot (70 °C) and biphasic cold (25
°C). The results show that poly(N-pentylacrylamide) is selectively soluble (>99.5%) in the polar EtOH-rich
phase at rest. Poly(N-alkylacrylamide)s with larger N-alkyl groups are predominantly (C6, 85%; C7, 95%)
or exclusively (>C8, >99.5%) in the heptane-rich phase at rest.
Polymer solubility is important in many applications of
macromolecular materials. Polymer solubility is particularly
important in work where soluble polymers serve as supports
for catalysts, reagents, sequestrants, or substrates.1 The most
common problem in designing a suitable soluble polymer
support is the need to achieve a high enough solubility so that
useful concentrations of catalysts, substrates, or reagents are
present. However, it is equally important to ensure that the
polymers used as supports have selective solubility because a
soluble polymer support is only useful if it is completely
separable from the solution under some other condition. It is
most common to achieve this result by a change of solvents or
conditions so that a soluble polymer support becomes completely
insoluble. Such a solubility change can be achieved by solvent
precipitation,2 pH change,3 cooling,1b,4 or heating.5 Alternatives
to separations that do not involve precipitation and filtration
are of interest too. For example, membrane filtration can separate
a solution of a large polymeric species and a smaller lower
molecular weight product.6 Membrane filtration is especially
useful with dendridic polymers whose shape facilitates this sort
of separation.7 More recently, work from our group and others
has shown that a polymer’s selective solubility in one phase of
a biphasic mixture can be as useful as either precipitation or
membrane filtration.8,9 Here we describe our studies of the
effects of polymer structure on the phase-selective solubility
of polymers in thermomorphic liquid/liquid biphasic mixturess
studies that show minor changes in polymer structure have a
profound effect on phase-selective solubility.
The work described here focuses on poly(N-alkylacrylamide)s
using a variety of solvent mixtures. These polymers were chosen
for study because we have already used these supports in
biphasic liquid/liquid separations. These polymers are accessible
by a synthetic route that we developed for the synthesis of a
(6) Vankelecom, I. F. J. Chem. ReV. 2002, 102, 3779-3810. Sirkar, K. K.;
Shanbhag, P. V.; Kowali, A. S. Ind. Eng. Chem. Res. 1999, 38, 3715-
3737. Dijkstra, H. P.; van Klink, G. P. M.; van Koten, G. Acc. Chem. Res.
2002, 35, 798-810.
(7) van Heerbeek, R.; Kamer, P. C. J.; van Leeuwen, P. W. N. M.; Reek, J. N.
H. Chem. ReV. 2002, 102, 3717-3756.
(8) (a) Bergbreiter, D. E.; Liu, Y.-S.; Osburn, P. L. J. Am. Chem. Soc. 1998,
120, 4251-4252. (b) Bergbreiter, D. E.; Osburn, P. L.; Wilson, A.; Sink,
E. M. J. Am. Chem. Soc. 2000, 122, 9058-9064. (c) Bergbreiter, D. E.;
Osburn, P. L.; Frels, J. D. J. Am. Chem. Soc. 2001, 123, 11105-11106.
(d) Mizugaki, T.; Murata, M.; Ooe, M.; Ebitani, K.; Kaneda, K. Chem.
Commun. 2002, 52-53. (e) Chiba, K.; Kono, Y.; Kim, S.; Nishimoto, K.;
Kitano, Y.; Tada, M. Chem. Commun. 2002, 1766-1767. (f) Ko¨llhofer,
A.; Plenio, H. Chem.-Eur. J. 2003, 9, 1416-1425. (g) Bergbreiter, D. E.;
Koshti, N.; Franchina, J. G.; Frels, J. D. Angew. Chem., Int. Ed. 2000, 39,
1040-1042.
(9) (a) Abatjoglou et al. (Abatjoglou, A. G.; Peterson, R. R.; Bryant, D. R.
Chem. Ind. 1996, 68, 133-139) describe such separations as ‘phase
decantantions’ and report how they can be used with low molecular weight
catalysts and products. (b) Bergbreiter, D. E.; Osburn, P. L.; Smith, T.; Li,
C.; Frels, J. D. J. Am. Chem. Soc. 2003, 125, 6254-6260. (c) Deng, G.-J.;
Fan, Q.-H.; Chen, X.-M.; Liu, D.-S.; Chan, A. S. C. Chem. Commun. 2002,
1570-1571. (d) Scurto, A. M.; Aki, S. N. V. K.; Brennecke, J. F. J. Am.
Chem. Soc. 2002, 124, 10276-10277.
(1) (a) Bergbreiter, D. E. Chem. ReV. 2002, 102, 3345-3383. (b) Bergbreiter,
D. E. J. Polym. Sci., Polym. Chem. Ed. 2001, 39, 2351-2363.
(2) Representative examples where solvent precipitation is used to recover
catalysts bound to polymers such as poly(alkene oxide)s, polystyrenes, and
chiral polymers can be found in recent reviews: (a) Dickerson, T. J.; Reed,
N. N.; Janda, K. D. Chem. ReV. 2002, 102, 3325-3343. (b) Toy, P. H.;
Janda, K. D. Acc. Chem. Res. 2000, 33, 546-554. (c) Bergbreiter, D. E.
In Chiral Catalyst Immobilization and Recycling; De Vos, D. E., Vankele-
com, I. F. J., Jacobs, P. A., Eds.; Wiley-VCH: Weinheim, Germany, 2000;
pp 43-80. (d) Fan, O.-H.; Li, Y.-M.; Chan, A. S. C. Chem. ReV. 2002,
102, 3385-3465.
(3) Malmstro¨m, T.; Andersson, C. J. Mol. Catal., A 2000, 157, 79-82.
(4) Osburn, P. L.; Bergbreiter, D. E. Prog. Polym. Sci. 2001, 26, 2015-2081.
(5) (a) Bergbreiter, D. E.; Caraway, J. W. J. Am. Chem. Soc. 1996, 118, 6092-
6093. (b) Huang, X.; Witte, K. L.; Bergbreiter, D. E.; Wong, C. H. AdV.
Synth. Catal. 2001, 343, 675-681. (c) Bergbreiter, D. E.; Case, B. L.; Liu,
Y.-S.; Caraway, J. W. Macromolecules 1998, 31, 6053-6062.
9
8244
J. AM. CHEM. SOC. 2003, 125, 8244-8249
10.1021/ja0349498 CCC: $25.00 © 2003 American Chemical Society

Polymer Phase-Selective Solubility
A R T I C L E S
Figure 1. Normal and inverse thermomorphic separations of a poly(N-
alkylacrylamide)-bound azo dye (a surrogate for a catalyst, reagent, or
substrate) in a thermomorphic heptane/90% EtOH-H2O mixture.
library of poly(N-alkylacrylamide)s. This synthesis provides a
route to poly(N-alkylacrylamide)s that vary only in the size of
their N-alkyl groups. It avoids the problem of variation in
molecular weight distribution that might complicate comparisons
of poly(N-alkylacrylamide)s prepared from N-alkylacrylamide
monomers.10 Polar/nonpolar solvent mixtures commonly used
in thermomorphic systems or latent biphasic systems already
shown to be useful in catalysis and synthesis were studied in
the most detail. The phase-selective solubility of poly(N-
isopropylacrylamide) in polar/polar mixtures too was briefly
examined. Finally, fluorescence labels were used as more
sensitive probes of the extent of polymer phase-selective
solubility. Using a dansyl probe, we were able to show that the
phase-selective solubility for polymers in nonpolar/polar solvent
mixtures can exceed 105:1.
Results and Discussion
Our past work showed that poly(N-isopropylacrylamide) 3
and poly(N-octadecylacrylamide) 4 labeled with p-Methyl Red
have very different solubilities.8a,c As shown in Figure 1, samples
of these two polymers labeled with the protonated form of
p-Methyl Red exhibited opposite phase-selective solubilities
when placed in a mixture of a nonpolar solvent (heptane) and
a polar solvent (90% ethanol-water). These polymers were
prepared by a conventional radical polymerization of the
respective N-isopropylacrylamide or N-octadecylacrylamide
monomer with a reactive ester comonomer (N-acryloxysuccin-
imide) (eq 1). The product copolymer’s (1 or 2) active ester
group was then allowed to react first with a primary amine
derivative of p-Methyl Red and then with isopropylamine or
octadecylamine (to consume any unreacted N-hydroxysuccin-
imide ester) to form the dye-labeled poly(N-isopropylacryla-
mide) (PNIPAM) 3 (eq 2) or poly(N-octadecylacrylamide)
(PNODAM) 4 (eq 3). The product polymers were labeled with
a 1-2 mol % loading of Methyl Red as the azo dye label and
were red in acidic solution or yellow in neutral solution. The
unprotonated form of the azo dye-labeled 3 had a λmax at 437
nm in EtOH. The same neutral dye supported on 4 dissolved in
heptane had a λmax at 419 nm. The protonated form of the azo
dye-labeled 3 had a λmax at 513 nm in a 90% EtOH-H2O
solution. The same protonated dye supported on 4 dissolved in
heptane had a λmax at 515 nm. A typical molecular weight (Mv)
of the PNIPAM polymer 3 (THF, 30 °C)11 was 6 × 105. Both
polymers 3 and 4 were readily soluble in a 70 °C miscible
mixture consisting of equal volumes of heptane and 90:10
(v/v) ethanol/water, but 3 and 4 had significantly different
solubilities at room temperature. Cooling a hot heptane/EtOH-
H2O homogeneous solution of either 3 or 4 produced a biphasic
mixture with a less dense nonpolar phase containing mostly
heptane (ca. 3% ethanol was in the heptane phase after cooling-
mediated phase separation) and a more dense polar phase
consisting mostly of ethanol and water (ca. 7% heptane is in
the polar phase after cooling-mediated phase separation). In the
case of the polar polymer 3, >99.9% of the polymer was in the
polar phase. This estimate is based on the absorbance difference
of the nonpolar and polar phases (the absorbance was <0.002
for the nonpolar phase and >2.00 for the polar phase). These
estimates assumed that 3 and 4 would have similar extinction
coefficients in the polar or nonpolar phase. The extent of phase
separation was also readily assayed visually for either the
protonated or nonprotonated dye-labeled polymers (photographs
of these phase-separated solutions of the protonated and
nonprotonated forms of 3 and 4 are provided in the Supporting
Information). In contrast, cooling a hot heptane/EtOH-H2O
solution containing 4 produced a biphasic mixture where
>99.9% of the nonpolar PNODAM-bound dye was in the
nonpolar phase. This differential phase-selective solubility of a
polar and nonpolar polymer-bound dye mirrors the phase-
selective solubility behavior of similar polymer-bound catalysts
(10) The variation in molecular weight distribution is a recognized problem that
affects the utility of N-alkylacrylamides in a variety of problems including,
for example, drug delivery applications and is one reason more well-defined
but more expensive polymers such as dendrimers are of increasing interest;
cf. Gillies, E. R.; Frechet, J. M. J. J. Am. Chem. Soc. 2002, 124, 14137-
14146.
(11) Winnik, F. M. Macromolecules 1990, 23, 233-242.
9
J. AM. CHEM. SOC. VOL. 125, NO. 27, 2003 8245

A R T I C L E S
Bergbreiter et al.
Table 1. Spectroscopic Estimates of Solubility for p-Methyl Red-
and Dansyl-Labeled PNIPAMs and PNODAMs and p-Methyl
Red-Labeled Poly(ethylene glycol) in Various Thermomorphic or
Biphasic Solvent Mixtures
another temperature). Immiscible mixtures of polar and nonpolar
solvents (e.g., water and heptane) or other solvents with varying
temperature-dependent miscibility (e.g., toluene and 85% EtOH-
H2O) are not listed here. Generally, such solvents have
comparable (>99.9%) phase-selective solubilities for the polar
and nonpolar polymer-bound dyes 3 and 4 in the polar and
nonpolar phases of these biphasic mixtures. The phase-selective
solubilities of polymers in solvents such as those listed in Table
1 were studied in more detail in this work because similar
polymer-supported species and similar solvent mixtures are
useful in homogeneous catalytic reactions and catalyst recovery.
High selective solubility of a dye-labeled polymer can be
presumed to be predictive of high levels of separation of a
similar ligand- or catalyst-containing polymer in synthetic
chemistry since the dye or fluorophore in these polymers (3-
6) is a surrogate for a catalyst, reagent, or substrate.
The high phase-selective solubility of these polymers has
important consequences for catalysis. Whitesides previously
noted12 that biphasic systems using enzymes as macromolecular
catalysts in aqueous biphasic reactors often cannot be readily
recovered without resort to countercurrent operation with
multiple partitions of the product mixture because enzyme phase
selectivities (distribution coefficients) are typically between 0.1
and 10 in ternary aqueous biphasic systems. However, the phase-
selective solubility of up to >100000:1 seen for these linear
poly(N-alkylacrylamide) supports and the use of roughly equal
volumes of the polar and nonpolar phases means that a
separation can easily effect >99.99% polymer (and presumably
catalyst) recovery.
nonpolar phase
polymer probe
solvent mixture
solubilitya (%)
Methyl Red-PNIPAM
Methyl Red-PNIPAM
dansyl-PNIPAM
Methyl Red-PNODAM
Methyl Red-PEG5000
Methyl Red-PNODAM
dansyl-PNODAM
dansyl-PNIPAM
heptane-EtOH(aq)b
heptane-DMFc
heptane-EtOH(aq)b
heptane-EtOH(aq)b
heptane-EtOH(aq)b
heptane-DMFc
heptane-EtOH(aq)b
heptane-DMFc
heptane-DMFc
Et3N-waterd
<0.1
<0.1
<0.01
>99.9
<0.1
>99.9
>99.99
<0.001
dansyl-PNODAM
Methyl Red-PNIPAM
dansyl-PNIPAM
>99.99
>99.9 (Et3N phase)
>99.9 (Et3N phase)
Et3N-waterd
a The relative amounts of polymer-bound dye or fluorophore were
measured in each phase after a miscible hot homogeneous mixture of
solvents and polymer was cooled to room temperature to form a biphasic
mixture. Relative solubilities (e.g., >99.9% or <0.001%) were estimates
based on the minimum detectable amount of probe after a series of serial
dilutions. b This solvent mixture consisted of a 1:1 (vol/vol) mixture of
heptane and EtOH-H2O (90% EtOH by volume). c This solvent mixture
consisted of a 1:1 (vol/vol) mixture of heptane and dimethylformamide.
d This solvent mixture consisted of a 1:1 (vol/vol) mixture of triethylamine
and H2O and differed from the other solvent mixtures in that it was miscible
cold (0 °C) and biphasic hot (25 °C).
and led us to explore in greater detail both the extent of phase-
selective solubility of such polymers and the effect of polymer
structure on such phase-selective solubility.
To further test the extent of phase-selective solubility, we
prepared analogues of 3 and 4 that contain dansyl groups using
the reactions shown in eqs 4 and 5. The resultant polymers 5
Most of the examples in Table 1 above are for the two
polymers we have used in catalysis and synthesissPNIPAM
and PNODAM.8 PEG derivatives too have been widely used
as supports in polymer-supported chemistry and in thermomor-
phic systems.2a,8a We briefly examined the phase-selective
solubility of one examplescommercially available PEG5000. This
polar polymer has phase-selective solubility like that of the polar
PNIPAM derivative. However, unlike the poly(N-alkylacryl-
amide)s, it would not be as simple to tinker with the polymer
microstructure to design a heptane-soluble support. Moreover,
our work showed that high phase-selective solubility was only
obtained for this polymer after continuously extracting an
aqueous solution of this polymer with heptane either before
Methyl Red labeling or as a Methyl Red-labeled derivative. We
believe this extraction removes lower molecular weight PEG
derivatives that are not as phase selectively soluble.
A caveat for Table 1 is that the poly(N-alkylacrylamide)s
listed here only contain N-isopropyl and N-octadecyl substitu-
ents. These substitutents of these poly(N-alkylacrylamide)s are
significantly different. We also wished to ascertain how sensitive
phase-selective solubility is to the size (hydrophobicity) of the
pendant N-alkyl groups of these poly(N-alkylacrylamide)s. We
already knew from others’ work that the N-alkyl substituent’s
hydrophobicity affects the poly(N-alkylacrylamide)s’ temper-
ature-dependent solubility.13 Such effects are useful in synthesis,
and varying the nature of an N-alkyl substituent can produce
quite subtle changes.5,14
and 6 behaved exactly the same as the Methyl Red-labeled
polymers 3 and 4. However, assays of the concentration of these
fluorophore-labeled polymers were more sensitive. This allowed
us to both use a lower loading of probe on the polymer and
estimate that phase-selective solubilities of 5 and 6 are as high
as >105:1 in polar solvents or >104:1 in nonpolar solvents.
Table 1 lists the phase-selective solubility of the both dye-
and fluorophore-labeled polymers in several solvent mixtures.
The solvent mixtures in this list are all thermomorphic (they
are biphasic at one temperature and completely monophasic at
(12) Pollak, A.; Whitesides, G. M. J. Am. Chem. Soc. 1976, 98, 289-291.
(13) Taylor, L. D.; Cerankowski, L. D. J. Polym. Sci., Polym. Chem. Ed. 1975,
13, 2551-2570. Schild, H. G. Prog. Polym. Sci. 1992, 17, 163-249.
(14) Mao, H.; Li, C.; Zhang, Y.; Bergbreiter, D. E.; Cremer, P. S. J. Am. Chem.
Soc. 2003, 125, 2850-2851.
9
8246 J. AM. CHEM. SOC. VOL. 125, NO. 27, 2003

Polymer Phase-Selective Solubility
A R T I C L E S
We expected that the phase-selective solubility of polymers
too would be affected by the hydrophobicity of the N-alkyl
substituent. However, it was not clear if small changes in
polymer substituent structure would lead to dramatic changes
in polymer solubility. The difference between PNIPAM and
PNODAM shows that there is a spectrum of phase-selective
solubility that is dependent on the size of the N-alkyl substit-
uents. However, these N-alkyl substituents are so different that
they are likely to be at the extremes of the polar-phase-soluble
or nonpolar-phase-soluble regions of this phase-selective solu-
bility behavior spectrum. Predicting the effects intermediate-
sized N-alkyl groups have on solubility is less certain. A
comparison of solubility parameters calculated using substituent
effects for two poly(N-alkylacrylamide)s that differ by only one
or two methylene groups in their N-alkyl substituents suggests
that changing the alkyl group size by a few carbons would lead
to a less than 1% change in the solubility parameter δ.15
However, the solubility of poly(N-alkylacrylamide)s such as
PNIPAM and PNODAM is affected by polar effects and
hydrogen bonding, so the relative solubility of poly(N-alky-
lacrylamide)s with similar alkyl substituents is best tested
experimentally.
There are some experimental problems associated with
directly comparing the phase-selective solubility of a series of
poly(N-alkylacrylamide)s prepared by polymerization reactions
such as eq 1. Specifically, a comparison between different poly-
(N-alkylacrylamide)s is complicated because conventional radi-
cal polymerization leads to polymers that do not necessarily
have the same polydispersity and degree of polymerization.
Indeed, it is difficult to prepare samples of the same polymer
in batch to batch reactions such that the degree of polymerization
and polydispersity are identical. While it is possible that the
degree of polymerization and the polydispersity are not impor-
tant factors in phase-selective solubility, we preferred to avoid
this ambiguity. We therefore used the synthesis of polymeric
active esters developed originally by Ringsdorf16 as a route to
a library of poly(N-alkylacrylamide) polymers that can be
compared without the ambiguities of variable polydispersity and
degree of polymerization. Whitesides had previously used a
similar approach with sequential amidation of the polymeric
active ester groups by various amines as a synthetic route to
polymeric inhibitors useful in biology.17
First, to verify that the procedure in Scheme 1 produces polymer
products that behave the same as the products prepared using
eqs 1-3, we separately prepared dye-labeled PNIPAM and
PNODAM polymers 8 and 15 via the reactive polyacrylate 7
and compared these polymers to 3 and 4. In this synthesis, a
homopolymer of NASI was prepared using AIBN initiation. The
product active ester homopolymer was then allowed to react
with a 50% molar excess of a 100:1 (mol/mol) mixture
containing either isopropylamine or octadecylamine and an
amine-containing derivative of Methyl Red. In this first
synthesis, the product PNIPAM and PNODAM polymers 8 and
15 were separated from excess amines and isolated by solvent
precipitation. The phase-selective solubility of these Methyl Red-
labeled PNIPAM and PNODAM polymers was identical to that
seen for 3 and 4 prepared via the more conventional route (eqs
2 and 3).
Next we developed a purification procedure for removal of
the excess amines in the syntheses of Scheme 1. Excess amine
is required in Scheme 1 to ensure no reactive ester remains in
the product polymer. We had previously shown that unreacted
NASI esters could hydrolyze readily to form acrylic acid
groups5csan event that would produce a terpolymer whose
phase-selective solubility might be quite different from that of
the desired poly(N-alkylacrylamide). However, while excess
amine avoids this problem, the resulting presence of unreacted
low molecular weight dye complicates the analysis of the phase-
selective solubility of the products. Thus, we required a
purification step that would remove low molecular weight
amines without any fractionation of the product poly(N-
alkylacrylamide)s. Such a separation can be accomplished by
solvent precipitation (vide supra). However, solvent precipitation
is a potential fractionation step, and we wanted to maintain in
our library synthesis the same polydispersity and degree of
polymerization for all samples. Therefore, we instead removed
excess low molecular weight amines from our product poly-
(N-alkylacrylamide) library members using a sulfonated ion-
exchange resin that we had previously shown to be a useful
amine sequestration agent in parallel synthesis of amines from
carbamates.18 In a control experiment, we showed that polymers
such as 8 and 15 do not react with this sequestration agentsa
sulfonated ion-exchange resin (Amberlyst 15). This was ex-
pected since our earlier work had shown that this resin readily
reacts chemioselectively with simple amines and/or BOC-
protected amines but that it does not react with amides. The
lack of reactivity of poly(N-alkylacrylamides) with the sul-
fonated cross-linked polystyrene was confirmed by our observa-
tion (UV-vis analysis) of no change in the solution concen-
tration of the chromophore due to 8 on treatment of a THF
solution of 8 with this polymeric sulfonic acid.
We used a pool-split synthesis route to prepare libraries of
poly(N-alkylacrylamide)s. This chemistry employs a common
polymer precursor and no fractionation steps. Thus, this
chemistry leads to a library of polymer samples, all of which
have the same degree of polymerization and polydispersity but
which vary in the structure of their pendant N-alkyl groups. By
combining this synthesis with a labeling procedure using the
azo dye described above, we have been able to study the effect
of N-alkyl group structure on the spectrum of phase-selective
solubility for poly(N-alkylacrylamide)s. The results allow us to
directly assay and compare the effects of polymer structure on
polymer phase-selective solubility.
Once procedures were in hand to remove the excess amine
and dye used in the synthesis of poly(N-alkylacrylamide)s
prepared via the polymeric active ester 7, we were able to
directly prepare a small library of poly(N-alkylacrylamide)s
8-15 containing N-isopropyl, N-n-butyl, N-n-pentyl, N-n-hexyl,
N-n-heptyl, N-n-octyl, N-n-dodecyl, and N-n-octadecyl groups.
A common precursor polymer (7) was prepared, and portions
of this polymer were added to DMF solutions containing a 1.5-
fold excess of a 100:1 mixture of a primary amine and a primary
Before using this chemistry to make a library to test phase-
selective solubility, we carried out several control experiments.
(15) Stevens, M. P. Polymer Chemistry, An Introduction; Oxford University
Press: New York, 1999; pp 38-40.
(16) Batz, H. G.; Franzmann, G.; Ringsdorf, H. Angew. Chem., Int. Ed. Engl.
1972, 11, 1103-1104.
(17) Mammen, M.; Dahmann, G.; Whitesides, G. M. J. Med. Chem. 1995, 38,
4179-4190.
(18) Liu, Y.-S.; Zhao, C.; Bergbreiter, D. E.; Romo, D. J. Org. Chem. 1998,
63, 3471-3473.
9
J. AM. CHEM. SOC. VOL. 125, NO. 27, 2003 8247

A R T I C L E S
Bergbreiter et al.
Scheme 1. Pool-Split Synthesis of a Library of Poly(N-alkylacrylamide)s That Vary Only in the Size of Their N-Alkyl Substituent
amine derivative of Methyl Red. The reaction mixtures were
heated to 50 °C to ensure solubility of all the reaction
components. After 24 h, the reaction mixture was cooled, and
a sulfonic acid containing ion-exchange resin (Amberlyst 15)
was added. Shaking the resulting mixture for 24 h removed any
excess low molecular weight amine or unbound dye. At this
point, a qualitative measure of the similarity of the samples of
polymers 8-15 was the visual similarity of color for the eight
different solutions, all of which had the same concentration of
chromophore. The remaining solvent was then removed at
reduced pressure to yield samples of 8-15 that were used
directly in phase-selective solubility studies.
The results of these studies were that Methyl Red-labeled
poly(N-alkylacrylamide) polymers with eight, twelve, and
eighteen carbon chains had phase-selective solubility in heptane.
No (<0.5%) Methyl Red-labeled polymer could be detected in
the polar phase in these three cases. Poly(N-alkylacrylamide)s
with three, four, or five carbon chains had phase-selective
solubility in the polar ethanol phase. No (<0.5%) Methyl Red-
labeled polymer could be detected in the nonpolar heptane phase
in these two cases. Intermediate phase-selective solubility was
seen for poly(N-alkylacrylamide)s containing six or seven
carbons in the pendant N-alkyl amido groups. Poly(N-hexyl-
acrylamide) was 85% soluble in the heptane phase (average of
two experiments). Poly(N-heptylacrylamide) had 91% phase-
selective solubility in the heptane phase (average of two
experiments). These results are summarized graphically in
Figure 2. These experiments all involved analysis of the Methyl
Red-containing polymers under conditions where the Methyl
Red is not protonated.
Once a collection of Methyl Red-labeled poly(N-alkylacryl-
amide)s 8-15 with differently sized N-alkyl side chains was in
hand, the phase-selective solubility of these polymers was tested
using visible spectroscopy. In each test, an individual poly(N-
alkylacrylamide) was added to a mixture of a 1:1 (vol/vol)
mixture of heptane and 90% aqueous ethanol (EtOH:H2O )
90:10 (v/v)). Heating to 70 °C produced a single miscible
solution from this solvent mixture and also dissolved the
polymer. At this point, the mixtures were cooled to room
temperature, at which time two phases, a nonpolar heptane phase
and a polar aqueous ethanol phase, formed. In general, phase
separation occurred readily. If phase separation were slow,
centrifugation would be used to facilitate complete phase
separation. Each phase was then sampled and analyzed by UV-
vis spectroscopy.
While the solubility differences seen for polymers 8-15
mirrored those seen for polymers made via eqs 2 and 3, the
degree of polymerization of the reactive polyester 7 used in
Scheme 1 is not as large as is the case for polymer prepared by
eq 2 or 3. The lower degree of polymerization of 7 was
measured indirectly by analyzing the Mn and Mw of the PNIPAM
derivative of 7 by GPC using 0.01 M LiBr in DMF as the eluent
on a TOSOH BIOSEP column. This analysis showed that this
PNIPAM derivative has a Mn and Mw of 1.8 × 105 and 3.4 ×
9
8248 J. AM. CHEM. SOC. VOL. 125, NO. 27, 2003

Polymer Phase-Selective Solubility
A R T I C L E S
Conclusions
Phase-selective solubility can readily be engineered into a
polymer. Small changes in the size of the N-alkyl group of a
poly(N-alkylacrylamide) lead to significant changes in phase-
selective solubility in situations where a polymer can dissolve
in either a polar or a nonpolar phase of a thermomorphic mixture
of heptane/90% EtOH-H2O. The homopolymer of N-acryloxy-
succinimide is a useful precursor to these poly(N-alkylacryl-
amide)s, and the product polymers can be separated from excess
amines either by conventional precipitations or through the use
of a polymeric reagent. Future work will examine poly(N-
alkylacrylamide)s containing mixtures of N-alkyl groups, less
polydisperse samples, and other polymers.
Figure 2. Heptane solubility versus the number of carbons in the N-alkyl
group of p-Methyl Red-labeled poly(N-alkylacrylamides) prepared according
to the method shown in Scheme 1 as measured by UV-vis spectroscopy
of the polar and nonpolar phases of a mixture of heptane/90% ethanol-
water (1:1 vol/vol) at 25 °C. In all cases, the phase-selective solubility was
measured after the mixture had been heated to 70 °C (miscibility) and then
allowed to cool to a resting biphasic state.
Acknowledgment. Support of this work by the National
Science Foundation (Grant NSF-CHE-0010103) and the Robert
A. Welch Foundation is gratefully acknowledged. J.B. also
acknowledges support of her efforts during the summer of 2002
through Texas A&M’s NSF REU program. Mr. Albert Resines
is also acknowledged for his contributions during the summer
of 2001 in a Texas A&M Chemistry-funded REU program for
his repetition of the chemistry in Scheme 1. Finally, we are
grateful to our colleague Professor Paul Cremer, who brought
the thermomorphic liquid/liquid systems that have an LCST to
our attention.
105, respectively. This suggests that phase-selective solubility
can be seen for PNIPAM and PNODAM polymers with varying
degrees of polymerization.
Phase-selective solubility of polymers can also be achieved
in thermomorphic systems that employ two polar solvents. An
example of such a system would be triethylamine and water.
This system differs from other thermomorphic mixtures we have
studied in that this thermomorphic solvent mixture has an LCST
for its miscibility; an equal volume of these two solvents is
miscible at 0 °C and immiscible at 25 °C. In this case, a
PNIPAM-bound Methyl Red dye or dansyl probe was found to
reside exclusively in the Et3N phase after a homogeneous 0 °C
1:1 (vol/vol) mixture of Et3N and H2O was warmed to room
temperature and phase separated. In this experiment, the
PNIPAM was present at 2 wt %. This separation was completely
reversible.
Supporting Information Available: Full experimental details
and photographs of PNIPAM and PNODAM resting biphasic
mixtures illustrating the clean and complete separation of phases
containing either protonated or nonprotonated p-Methyl Red dye
labels. This material is available free of charge via the Internet
at http://pubs.acs.org.
JA0349498
9
J. AM. CHEM. SOC. VOL. 125, NO. 27, 2003 8249