Using soluble polymers in latent biphasic systems

Article abstract of DOI:10.1021/ja0342852

A new strategy for carrying out reactions with a soluble polymer-bound reagent or catalyst is described. In this latent biphasic process, a solvent mixture at the cusp of immiscibility is prepared and used to carry out a reaction under homogeneous conditions. Then, after the reaction is complete, this mixture is perturbed by the addition of solvent or some other perturbing agent to produce a biphasic mixture. The product-containing phase is then separated under liquid/liquid conditions from the polymer-containing phase. The generality of this process is demonstrated using both dye-labeled polymers as surrogates for polymer-bound catalysts and with various polymer-bound organic and transition metal catalysts or reagents. In cases where a polymeric catalyst is used, the addition of fresh solvent and substrate reforms the original mixture allowing facile reuse of the catalyst.

Full text of DOI:10.1021/ja0342852

Published on Web 04/29/2003
Using Soluble Polymers in Latent Biphasic Systems
David E. Bergbreiter,* Philip L. Osburn, Thomas Smith, Chunmei Li, and
Jonathon D. Frels
Contribution from the Department of Chemistry, Texas A&M UniVersity, P.O. Box 30012,
College Station, Texas 77842-3012
Received January 22, 2003; E-mail: bergbreiter@tamu.edu
Abstract: A new strategy for carrying out reactions with a soluble polymer-bound reagent or catalyst is
described. In this latent biphasic process, a solvent mixture at the cusp of immiscibility is prepared and
used to carry out a reaction under homogeneous conditions. Then, after the reaction is complete, this
mixture is perturbed by the addition of solvent or some other perturbing agent to produce a biphasic mixture.
The product-containing phase is then separated under liquid/liquid conditions from the polymer-containing
phase. The generality of this process is demonstrated using both dye-labeled polymers as surrogates for
polymer-bound catalysts and with various polymer-bound organic and transition metal catalysts or reagents.
In cases where a polymeric catalyst is used, the addition of fresh solvent and substrate reforms the original
mixture allowing facile reuse of the catalyst.
Biphasic systems facilitate homogeneous catalyst recovery
and reuse and are widely used in catalysis.1 However, while
biphasic conditions facilitate catalyst/product separation, mul-
tiple phases introduce kinetic barriers and make the resulting
process more complex. Phase transfer catalysis and thermo-
morphic catalysis strategies address these issues but preserve
the underlying advantages of a biphasic system.2,3 Here, we
describe another alternative, a latent biphasic system. In such a
system, a solvent mixture at the cusp of immiscibility is prepared
and used to carry out a reaction under homogeneous conditions.
Then, after the reaction is complete, this mixture is perturbed
Figure 1. Latent biphasic process with a reaction occurring under
monophasic conditions and separation occurring under liquid/liquid biphasic
by the addition of some perturbing agent to produce a biphasic
mixture. A similar procedure was recently described as being
useful with dendrimer-bound catalysts.4 This paper describes
the generality of this simple alternative strategy using linear
polymers as supports. Dye-labeled polymers are used as
surrogates for catalysts on polymers to test separation efficiency
and catalysts on polymers are used to demonstrate feasibility
of this scheme in known homogeneous catalytic reactions.
Mixtures of solvents have a spectrum of phase behavior. For
example, ethanol/water or ethanol/heptane are miscible in all
proportions. However, mixtures containing all three of these
solvents vary in miscibility depending on the proportions of
solvents used, the temperature, and the presence or absence of
solutes. As an illustrative example, a 10.0:9.5:0.5 (vol:vol:vol)
mixture of heptane, EtOH and H2O is monophasic at 25 °C,
conditions after some perturbation.
but a perturbed 10.0:9.0:1.0 mixture of these solvents is biphasic
at 25 °C (Figure 1). We and others have used the temperature
dependence of the latter sort of mixture in thermomorphic
reactions.5-9 Here we show how addition of small amounts of
another solvent or salt to the former sort of monophasic mixture
that is at the cusp of miscibility (a latent biphasic system) can
be used as a new way to effect a separation after a homogeneous
reaction.
Although solvent mixtures can be designed to change from
a monophasic system to a biphasic system by addition of a small
volume of another solvent or by the addition of a small amount
of some other phase perturbing agent, that fact alone is not
sufficient to effect a practical separation after a homogeneous
reaction. Any practical separation scheme for a catalytic reaction
(1) A recent issue of Chem. ReV. reviewed this topic, see: Gladysz, J. A. Chem.
ReV. 2002, 102, 3215-3216 and also Baker, R. T.; Tumas, W. Science
1999, 284, 1477-1479.
(2) Goldberg, Y.; Alper, H. Phase-transfer catalysis and related systems. In
Applied Homogeneous Catalysis with Organometallic Compounds, 2nd ed.;
Cornils, B., Herrmann, W. A., Ed.; VCH: Weinheim, 2002; Vol. 2, 953-
974.
(5) Bergbreiter, D. E.; Osburn, P. L.; Wilson, A.; Sink, E. M. J. Am. Chem.
Soc. 2000, 122, 9058-9064.
(6) Bergbreiter, D. E.; Osburn, P. L.; Frels, J. D. J. Am. Chem. Soc. 2001,
123, 11 105-11 106.
(7) Wende, M.; Meier, R.; Gladysz, J. A. J. Am. Chem. Soc. 2001, 123, 11490-
11491.
(3) Bergbreiter, D. E.; Liu, Y.-S.; Osburn, P. L. J. Am. Chem. Soc. 1998, 120,
4251-4252.
(4) Deng, G.-J.; Fan, Q.-H.; Chen, X.-M.; Liu, D. S.; Chan, A. S. C. Chem.
Commun. 2002, 1570-1571.
(8) Mizugaki, T.; Murata, M.; Ooe, M.; Ebitani, K.; Kaneda, K. Chem.
Commun. 2002, 52-53.
(9) Chiba, K.; Kono, Y.; Kim, S.; Nishimoto, K.; Kitano, Y.; Tada, M. Chem.
Commun. 2002, 1766-1767.
9
6254
J. AM. CHEM. SOC. 2003, 125, 6254-6260
10.1021/ja0342852 CCC: $25.00 © 2003 American Chemical Society

Soluble Polymers in Latent Biphasic Systems
A R T I C L E S
requires that a catalyst selectively partition itself into one of
the two phases formed after perturbation of the original
monophasic system. If a catalyst were used, then it should also
be possible to recycle the catalyst-containing phase through
multiple reaction cycles. A phase selective solubility of at least
200:1 for the catalyst is necessary if effective separable,
recyclable catalysts are to be developed. Similar considerations
have been discussed previously in connection with enzymatic
catalyst separations in aqueous biphasic systems and are
implicitly taken into account in any extraction.10 As shown
below, such selectivity and recyclability is possible with linear
polymer supports.
separation efficiency in this heptane/EtOH-H2O latent biphasic
system. When 0.5 mL of water was added to 8 mL of a 1:1
heptane-ethanol homogeneous solution of the dansyl-labeled 3,
two phases formed. More than 99.99% of the polymer stayed
in the aqueous ethanol phase based on fluorescence analysis of
both the heptane-rich and EtOH-rich phases for the dansyl
fluorophore. PNIPAM labeled with the p-methyl red azo dye
described above (i.e. polymer 4)11 was similarly phase selec-
tively soluble in the polar EtOH-rich phase of these heptane/
EtOH-H2O mixtures after phase separation, though the extent
of phase selective solubility could only be estimated at >99.9%.
These results and the PNODAM result above suggest that
separation and recovery of a catalyst bound to these polyacryl-
amides in either a nonpolar or a polar phase of a nonpolar/
polar mixture should be feasible.
Results and Discussion
To test the efficacy of soluble linear polymers for separations
in this latent biphasic strategy for catalyst recovery, we prepared
a variety of soluble polymers containing dye labels. First we
prepared poly(N-octadecylacrylamide) (PNODAM) containing
2 mol % of p-methyl red as a visually detectable surrogate for
a lipophilic polymer-bound catalyst (eq 1). We first showed that
this PNODAM-bound dye 2 is soluble in heptane or toluene
but is insoluble in an EtOH/H2O phase that contains >10% H2O.
We then showed that a visually homogeneous solution of this
lipophilic polymer-bound dye in a mixture of 10 mL each of
toluene and 95% EtOH/H2O becomes biphasic on addition of
0.5 mL of H2O. UV-visible analysis showed that >99.9% of
the polymer-bound dye stayed in the nonpolar toluene-rich
phase. Separation of this phase and addition of fresh aqueous
EtOH reformed the initial monophasic mixture. Repetition of
this cycle 5 times with no detectable loss of 2 in the EtOH-rich
phase showed that this approach is viable as a way to recover
and reuse polymeric catalysts. However, these immiscible phases
have similar density that makes phase separation slow. When a
1.00:0.96:0.04 mixture (vol:vol:vol) of heptane, EtOH, and H2O
was used to form a 20 mL monophasic solution, the addition
of just 10 drops of H2O was sufficient to induce phase
separation. This latter system’s phases had less similar densities,
making the separation cleaner, simpler, and faster. UV-visible
analysis showed that >99.9% of the polymer-bound dye stayed
in the nonpolar heptane-rich phase and the dye 2 was as
recyclable in this heptane/EtOH-H2O system as it was in the
toluene system above.
We believe that the most effective way to recover catalysts
in a latent biphasic reaction is to recover them in the nonpolar
phase of a nonpolar/polar mixture. This is illustrated by the
examples of catalysis described below. However, it might be
desirable to use polar/polar solvent mixtures in other situations.
To test the generality of linear polymers in latent biphasic
separations, we have also briefly explored the use of linear
polymers in this circumstance too.
The polar solvent mixture we examined was a mixture of
tert-butyl methyl ether (TBME), ethanol and water. First, we
prepared a miscible mixture of these solvents using 10 mL of
TBME, 6 mL of EtOH, and 4 mL of H2O. PNIPAM labeled
with methyl red (4) dissolved in this mixture. On addition of
2.5 mL of water to this homogeneous solution, a biphasic
mixture formed. The dye labeled polymer 4 was exclusively in
the TBME-rich layer at that point.
PNIPAM was phase selectively soluble in one phase of the
perturbed latently biphasic TBME/EtOH/H2O mixture. We
reasoned it should be possible to design a polymer that is phase
selectively soluble in either phase as was the case for the
PNIPAM/PNODAM polymers used in heptane/aqueous EtOH
mixtures above. This required a more polar polymer than
PNIPAM. A candidate more polar polymer, poly(N-acryletha-
nolamide) 6, was prepared by reaction of an active ester
polymer, poly(N-acryloxysuccinimide) (PNASI),12 with a mix-
Poly(N-isopropylacrylamide) (PNIPAM), a polymer previ-
(11) Bergbreiter, D. E.; Case, B. L.; Liu, Y.,-S.; Caraway, J. W. Macromolecules
1998, 31, 6053-6062.
(12) (a) Mammen, M.; Dahmann, G.; Whitesides, G. M. J. Med. Chem. 1995,
38, 4179-4190. (b) Batz, H. G.; Franzmann, G.; Ringsdorf, H. Angew.
Chem., Int. Ed. Engl. 1972, 11, 1103-1104.
ously used in thermomorphic systems,3 also showed good
(10) Pollak, A.; Whitesides, G. M. J. Am. Chem. Soc. 1976, 98, 289-291.
9
J. AM. CHEM. SOC. VOL. 125, NO. 20, 2003 6255

A R T I C L E S
Bergbreiter et al.
ture of ethanolamine and an amine-derivative of p-methyl red
(eq 2). As was true in the examples above, the methyl red
indicator dye was incorporated both to serve as a surrogate
for a catalyst and as a convenient UV-visible label for the
polymer.
The total volume of the recovered H2O-rich phase in each of
the 4 subsequent cycles was not substantially different from
the volume of the H2O-rich phase in the first cycle because 2.5
mL of H2O was being added every cycle to induce phase
separation.
The absorbance in the TBME (top) layers through five cycles
was as follows; cycle 1-0.108 at 421 nm; cycle 2-0.034 at
426 nm; cycle 3-0.022 at 422 nm; cycle 4-0.037 at 423 nm;
cycle 5-0.011 at 426 nm. The absorbance of the polymer-bound
dye containing layer was generally too high to measure without
dilution (A > 3.5). Qualitatively, it was the same shade of orange
through each cycle. The loss of some material in cycles 1
through 5 may be due to low molecular weight PNAEAM
present in the sample of 6 used. The PNASI precursor used to
prepare 6 was polydisperse since it was derived from a standard
radical polymerization. Although the polymer 6 was not directly
analyzed, the PNASI used to prepare the PNAEAM was
characterized indirectly. This was accomplished by allowing the
common PNASI precursor polymer to react with isopropylamine
to form PNIPAM. The PNIPAM so prepared (and by inference
the PNASI) had a PDI of 1.89 and an Mn of 1.8 × 105 Da
based on GPC analysis. In prior experiments using a PEG-bound
dye as a surrogate for a PEG-bound catalyst in a thermomorphic
system,5 we noted that it can be necessary to extract low
concentrations of less phase selectively soluble polymer from
a polydisperse sample in a liquid/liquid extractor if a >99.5%
recovery of a polymer is to be achieved. This was not done in
this case prior to the use of 6 in this latent biphasic system,
whereas recovery of PNAEAM was >99.5% by cycle 5.
The dye-labeled polymer poly(N-acrylethanolamide)
(PNAEAM, 6) (20 mg) was first dissolved in 6 mL of ethanol.
Heating was used to speed dissolution, but the polymer was
soluble in ethanol at room temperature. Then an approximately
4 mL portion of water was added. The polymer stayed in
solution. Finally, 10 mL of TBME was added to form an
orangish transparent solution. This latently biphasic system was
then perturbed into a biphasic state by the addition of 2.5 mL
of water. The denser orangish water-rich phase and less dense
TBME-rich phase thus formed separated readily in less than
five minutes. Once separation occurred, the polymer-bound azo
dye was recovered as an aqueous solution by simple liquid/
liquid separation.
The studies above with dye-labeled polymers show that a
latent biphasic strategy can effect separations. Studies with
catalysts bound to phase selectively heptane-soluble polymer
supports show the utility of this latent biphasic chemistry in
known catalytic reactions. Three reactions were studied. The
first was Heck catalysis using an SCS-Pd(II) catalyst we
described previously (eq 3).5 Catalyst 7 is soluble at 25 °C in
a miscible 1:1 heptane-dimethylacetamide (DMA) mixture along
with acrylic acid, iodobenzene and Et3N. After a homogeneous
reaction (120 °C for 12 h), cooling to 25 °C formed a biphasic
mixture of 7 and the Heck product due to formation of
triethylammonium iodide. Isolation of the heptane phase
containing 7 and addition of a fresh amount of DMA, substrates,
and base to this heptane solution allowed us to recycle this
catalyst.
Catalysis was not studied in this TBME/EtOH/H2O system.
However, recycling the polymer bound azo dye was feasible.
This recycling was followed both visually and by UV-visible
spectroscopy. First, the less dense TBME-rich layer from the
first cycle described above was separated from the H2O-rich
phase and analyzed by UV-visible spectroscopy. A small
absorbance due to the dye (<3% of the original dye) was
detected in this first cycle in the TBME rich phase. Then, 10
mL of fresh pure TBME was added to the denser bottom layer
(the H2O-rich, polymer-bound dye-containing phase) along with
4 mL of ethanol. These additions of solvent reformed a single
miscible orangish homogeneous solution that was visually
indistinguishable from the starting solution. Water addition to
form a biphasic system, separation, analysis of the TBME-rich
phase for 6 and reforming of the miscible solution by further
addition of TBME and EtOH was then carried out for several
cycles. Less than 1% of the polymer 6 was lost in the TBME
phase in cycles 2-5 as determined by UV-visible spectroscopy.
9
6256 J. AM. CHEM. SOC. VOL. 125, NO. 20, 2003

Soluble Polymers in Latent Biphasic Systems
A R T I C L E S
The generality of this latent biphasic approach in homoge-
neous catalyst recovery was shown using a polymer-bound
trifunctional base catalyst 8 (eq 4) as a DMAP analog13 in
acylation of 2,6-dialkylphenols by (Boc)2O in either a 1:1
heptane-aqueous EtOH or a 1:1 heptane-DMF system (eq 5).
Yields in this chemistry gradually increased from cycle to cycle
as the heptane-rich phase became saturated in the product
carbonate. Yields through 6 cycles for R ) -H were 59%, 80%,
95%, 99%, 99%, and 99% and for R ) -CH3 were 35%, 66%,
89%, 99%, 99%, and 99% when a heptane/ethanol system was
used. A similar experiment was also carried out using 2,6-
diisopropylphenol in a heated thermorphic heptane/DMF mix-
ture. In this latter case, the product yields (from the DMF-rich
phase) were 44%, 57%, 76%, 90%, 98%, 99%, and 99% through
the first seven cycles. In the first two of these cases, the initially
monophasic reaction mixture was perturbed to be biphasic by
the addition of 10 vol % H2O. The heptane phase so formed
was then recycled by addition of fresh EtOH (or DMF), substrate
and (Boc)2O. Heptane was added as needed to replace heptane
lost in the EtOH- or DMF-rich phases. In all three of these series,
the polymeric catalyst used was labeled with a methyl red label.
The absence of detectable dye-labeled polymer in any of the
cycles in either the polar EtOH-rich or DMF-rich phases
indicated quantitative separation of the catalyst into the nonpolar
heptane-rich phase.
Any phase selectively soluble polymer support should work
in this latent biphasic strategy. To show this, a third example
of catalyst recycling under latent biphasic conditions used
another nonpolar phase-soluble polymer as a support for a Lewis
acid catalyst. This polymer has not been used before as a soluble
polymer support, but it is similar to a material we used earlier
to support hydrogenation catalysts.14 This new polymer (9) was
prepared by an alternating copolymerization of the commercially
available monomers, octadecyl vinyl ether and maleic anhy-
dride.15 Post polymerization amidation of the anhydride groups
of this copolymer with a 1:1 mixture of octadecylamine and
morpholine produced an amic acid-containing polymer that had
a 10:1 (mole:mole) mixture of secondary and tertiary amides.
Heating induced imidization of the secondary amide-acids of
this amic acid-containing polymer to produce a terpolymer
containing N-octadecylmaleimide and octadecylvinyl ether
repeating units with a ca. 10% loading of unimidized amic acids
derived from the secondary amine morpholine. Ion exchange
of some of the protons of these remaining carboxylic acid groups
of these remaining amic acids with silver acetate produced an
Ag(I)-polymer 5 that was then used as a heptane-soluble Lewis
acid (eq 6).
This Ag(I)-polymer 10 as well as 716 both catalyze atom
efficient oxazoline formation from ethyl isocyanoacetate and
benzaldehyde (eq 7) under latent biphasic conditions. Recycling
studies however only used the polymer-bound Ag(I) catalyst
10 since it was a more readily available catalyst.
Recycling 10 proved to be simple and effective. After 20
mL of a miscible 1:1 (vol:vol) EtOH-heptane solution containing
10 (215 ppm Ag(I)) was used for oxazoline formation (overnight
in a flask shielded from light), the catalyst 10 was quantitatively
recovered as a heptane phase by the addition of 15 drops of
water. Recycling 10 proved to be possible and involved addition
of 10 mL of fresh EtOH, 5 mL of heptane and fresh reagents.
Analysis of the polar phase showed traces of Ag(I) (0.44, 0.18,
and 0.14 ppm Ag(I) in cycles 1-3, respectively) in the recovered
EtOH-rich phase. The origin of this <0.11% loss of Ag(I) per
cycle is uncertain. It may reflect loss of polymer to the other
phase since this value is at the UV-visible detection limit for
(13) Bergbreiter, D. E.; Osburn, P. L.; Li, C. Org. Lett. 2002, 4, 737-740.
(14) Bergbreiter, D. E.; Liu, Y.-S. Tetrahedron Lett. 1997, 38, 3703-3706.
(15) Culberston, B. M. In Encyclopedia of Polymer Science and Engineering;
Kroschwitz, J., Ed.; Wiley & Sons: New York, 1987; Vol. 9, p 225.
(16) Gimenez, R.; Swager, T. M. J. Mol. Catal. A: Chem. 2001, 166, 265-
273.
9
J. AM. CHEM. SOC. VOL. 125, NO. 20, 2003 6257

A R T I C L E S
Bergbreiter et al.
a dye-labeled analogue of 10. Alternatively, this could reflect
low but variable Ag(I) leaching due to glassware contamination
by ppb levels of anionic surfactants. Indeed, a repetition of this
experiment with more cycles but with ordinary glassware
(glassware that had been used and washed with phosphate-
containing detergents) showed a larger and more variable
average loss of <0.15% of the starting Ag(I)/cycle (0.59, 0.05,
0.58, 0.20, 0.20 ppm Ag(I) in cycles 1-5).
A final example of latent biphasic chemistry is the use of
the polymer 9 as a selective primary amine sequestrant. Earlier
work using a similar polymer as a support for phosphine ligands
had noted that similar maleic anhydride copolymers do not
readily react with 2° amines.14 Thus, we expected 4 would be
effective as a selective sequestrant for 1° amines.
This heptane-soluble sequestrant was used to remove excess
amine used in a propylene oxide alkylation of amines to form
ethanolamines (eq 8). In this reaction, excess 1° amine was used
to minimize over alkylation. Addition of 9 then consumed the
excess 1° amine and perturbation of the resulting heptane-EtOH
solution separated the product 2° amine from the scavenger resin
and the sequestered 1° amine. A series of reactions used 2 equiv
of butyl-, amyl-, hexyl-, octyl-, or decylamine as the substrates
using ethanol as the solvent. After heating to 70 °C for 2 h, the
mixture was cooled to room temperature, and a heptane solution
of 3 equiv of 9 was added to form a homogeneous heptane-
EtOH mixture. The excess amine was consumed within 1 min.
after addition by this solution of 9. Addition of H2O (the added
water was 10% of the total volume) then produced two phases.
Any 1° amine was at that point bound to the polymer and
sequestered in the heptane phase. The polar phase containing
the hydroxylamine products was then separated. A limitation
of this chemistry was that the products derived from the more
hydrophobic amine substrates have some heptane solubility. In
those cases, high yields of product can only be obtained if the
heptane solution is extracted with aqueous EtOH (the polymer
exclusively remains in the heptane phase even after the
extractions).
support for a reagent or catalyst can be designed to be selectively
soluble in a particular phase, products can partition to either
phase. In a repetitive process like the phenyl carbonate synthesis
described above, this problem will be of minimal consequence
as the recycled phase will eventually be saturated in product.
In other cases, extraction of the recycled phase might be
necessary to obtain a high isolated yield of a product in a single
cycle. This problem is a general problem with any biphasic
liquid/liquid separation strategy. Finally, it should be recognized
that reactions carried out in this way involve a mixed solvents
mixed solvent conditions will not always mirror conventional
chemistry carried out in a single pure solvent.
The results here show that this latent biphasic approach is
an alternative to more standard biphasic chemistry using soluble
and insoluble polymers since polymers can be designed to have
large distribution coefficients and phase selective solubility.
While dendrimers can be used similarly, these sorts of soluble
polymers are not needed to ensure efficient recovery or
separation of a polymer-bound catalyst or reagent. The addition
of a low volume percent of a third solvent or the addition of a
phase perturbing agent should also be useful in more normal
biphasic reactions such as fluorous biphasic chemistry where
solvent or ligand/catalyst loss is a problem.
Experimental Section
All reagents and solvents were obtained from commercial sources
1
and used without further purification unless otherwise specified. H
NMR spectra were recorded on Varian VXR-300 or Unity p300
spectrometers at 300 MHz. Chemical shifts are reported in ppm with
chloroform (7.27 ppm) as an internal standard.
Tests of Polymer Separability/Recovery in the Toluene-Etha-
nol-Water System. A sample of poly(N-octadecylacrylamide) con-
taining 1 or 2 mol % of p-methyl red as a dye label (2) was prepared
according to a reported procedure6 and dissolved in a mixture of 10
mL of toluene and 10 mL of 86% EtOH-H2O. Addition of 0.5 mL of
water (2.5% of the total volume) to the solution resulted in layer
separation with the dye-labeled polymer visually residing in the upper
toluene-rich layer. Analysis by UV-visible spectroscopy showed the
polymer had >1500:1 preference for the toluene layer.
Tests of Polymer Separability/Recovery in the Heptane-Etha-
nol-Water System. A sample of the dye-labeled poly(N-octadecyl-
acrylamide) 2 was dissolved in 20 mL of a solution containing equal
portions of heptane and 96% EtOH-H2O. Addition of 9-10 drops of
water produced a biphasic mixture. As in the toluene-aqueous ethanol
system above, this dye-labeled PNODAM polymer was exclusively
soluble in the heptane layer. UV-visible spectroscopy showed that
the phase selective solubility was >1500:1. Similar experiments with
a dansyl-labeled PNIPAM (3) or a methyl red-labeled PNIPAM (4)
had 3 or 4 staying exclusively in the polar EtOH-rich phase. The
necessary PNIPAM derivatives were prepared by reaction of an amine
derivative of dansyl or methyl red17 with a PNIPAM-NASI copolymer
using a previously described procedure.11
N-Octadecylacrylamide. Octadecylamine (26.7 g) was suspended
in 500 mL of CH2Cl2. Then 16.9 mL of triethylamine was added, and
the mixture was cooled to 0 °C using an ice-water bath. Dropwise
addition of 50 mL of a CH2Cl2 solution containing 8.9 mL of acryloyl
chloride to the amine suspension and stirring at room temperature for
1 h yielded a solution of the product monomer. To isolate the monomer,
the CH2Cl2 was removed under reduced pressure and the residue was
taken up in 250 mL of chloroform, washed with water and brine, dried
over sodium sulfate and filtered. The filtrate was concentrated and the
Conclusions
This latent biphasic strategy should be a general way to carry
out homogeneous reactions and separate and recover catalysts,
reagents or products by biphasic separation. However, it is
important to recognize that this approach like any approach has
some limitations. First, it is unlikely that any of the polymers
described here would be suitable for all reactions. A specific
reaction may require a different polymersa polymer without
N-H bonds for example. Second, because hundreds of different
solvent mixtures might be used, it would presumably be
necessary to test the phase selective solubility of a polymer with
the specific solvents that are to be used. Third, the solvent
mixtures used should be different in density in their biphasic
state. Fourth, loss of solvent in the recovered/reused phase can
occur. This problem can be minimized by using a second phase
that is presaturated with this solvent. Fifth, although the polymer
(17) Bergbreiter, D. E.; Franchina, J. G.; Kabza, K. Macromolecules 1999, 32,
4993-4998.
9
6258 J. AM. CHEM. SOC. VOL. 125, NO. 20, 2003

Soluble Polymers in Latent Biphasic Systems
A R T I C L E S
residue was taken up in 100 mL of acetone with warming. After the
addition of 80 mL of acetone, the mixture was allowed to sit at room
temperature in the dark to crystallize. The product was isolated by
filtration and had mp 70-71 °C (lit18 74.5-75.1 °C); 1H NMR (CDCl3)
δ 6.3 (d, 15.4 Hz, 1H), δ 6.1 (dd, J ) 10.2, 15.4 Hz, 1H), 5.6 (d, 10.2
Hz, 1H), 3.3 (q, 6.9 Hz, 2H), 1.5 (t, 6.9 Hz, 2H), 1.3 (bs, 30H), 0.9 (t,
6.6 Hz, 3H).
was stirred overnight. An acidic ion-exchange resin, Amberlyst-15 (4
g), was added to the reaction and the mixture was shaken overnight to
remove any unreacted amines. The resin was filtered and washed with
CHCl3. The filtrate was freed of solvent under reduced pressure to yield
the product amic acid-containing polymer as a solid (IR stretches at
1716, 1641, 1575, and 1467 cm-1). This polymer was then dissolved
in 250 mL of toluene and heated at reflux overnight using a Dean-
Stark trap to remove water. The product solution was then concentrated
under reduced pressure to yield a solid that was dissolved in 100 mL
of CH2Cl2. The solvent was removed under reduced pressure using a
rotary evaporator to give 1.1 g (65%) of a light yellow solid. 1H NMR
(CDCl3) δ 3.3-3.5 (br s, 2.2H), 2.8 (br s, 1H), 1.4-1.6 (br s, 2H),
1.1-1.4 (br s, 27H), 0.83 (t, 6.6 Hz, 3H). IR (KBr, cm-1) 1772, 1700,
1465, 1382.
PMOAVE-Silver Salt (10). The polymer prepared as described
above (500 mg, 0.0795 mmol) was dissolved in 70 mL of heptane and
added to 30 mL of EtOH. Silver acetate (132 mg, 0.795 mmol) was
added to the reaction as a solid. The reaction was stirred at room
temperature in the dark for 48 h. Water (10 mL) was added to the
reaction to separate the layers. The mixture was transferred to a
separatory funnel and the aqueous layer was removed. The organic
layer was filtered and dried to give 400 mg (79%) of a dark yellow
powder. IR (KBr, cm-1) 1698, 1646, 1585.
Poly(N-octadecylacrylamide-co-NASI) (PNODAM-NASI) (10:1).
Benzene (50 mL) was degassed with N2 for 30 min. Then octadecyl-
acrylamide (4.5 g, 13.9 mmol) and N-acryloxysuccinimide (0.24 g, 1.39
mmol) were added. The reaction mixture was warmed to 80 °C and a
solution of AIBN (12.6 mg, 0.077 mmol) in 2 mL of benzene was
added. The reaction mixture was stirred at 80 °C under nitrogen for 36
h, at which point the solvent was removed and the residue was taken
up in 30 mL of chloroform. The product was isolated by solvent
precipitation by slow addition of this solution into 150 mL of cold
methanol to give 4.1 g (87%) of a white powder. IR (KBr, cm-1) 1812,
1
1783, 1737, 1654, 1542. H NMR (CDCl3) δ 3.1 (br s, 20H), 2.8 (br
s, 4H), 1.5 (br s, 27H), 1.2 (br s, 395H), 0.8 (br t, 30H). Polymers with
different monomer ratios were obtained similarly using different feed
ratios of monomers.
Dye-labeled PNODAM-supported DAAP Catalyst. A sample of
a 4:1 PNODAM-PNASI copolymer (4 g) prepared as described above
was dissolved in chloroform and first allowed to react with 50 mg of
amine-terminated methyl red14 at room-temperature overnight. Then
500 mg of the amine-terminated DAAP catalyst15 in chloroform was
added. After overnight stirring, a final portion of 1 g of N-octadecyl-
amine was added to quench any remaining active esters. After overnight
stirring, 3.7 g of the red polymer product was isolated by methanol
precipitation. The polymer was characterized by UV-visible spectros-
copy (λmax ) 419 nm in heptane). Based on GPC analysis (DMF, 0.01
M LiBr, TOSOH BIOSEP column) of an analogously prepared
PNIPAM polymer, the molecular weight (Mn) of this product and the
PDI were 800,000 Da and 1.9, respectively.
Latent Biphasic Heck Catalysis. The catalyst used and the
procedure used was identical to the normal thermomorphic procedure
previously described with the difference being that the starting solution
was a single phase.6
Latent Biphasic Oxazoline Synthesis. The polymeric Ag(I) catalyst
10 prepared above (125 mg, 0.02 mmol of Ag) was dissolved in 20
mL of heptane. Benzaldehyde (212 mg, 2 mmol), ethyl isocyanoacetate
(226 mg, 2 mmol), and Hunig’s base (35 µL, 0.2 mmol) were dissolved
in 10 mL EtOH containing 1 mmol dodecane as an internal standard.
The alcoholic solution was then added to the solution of the polymer
to form a homogeneous system. The reaction was protected from light
and stirred at room-temperature overnight. At this point, 15 drops (∼600
mg) of water were added to induce separation of the layers. The lower
layer was removed and stored for metal analysis. A fresh solution of
reagents in ethanol was added to the reaction flask containing the
heptane solution of 10. An additional 5 mL of heptane was added to
regenerate a new solution of substrate, catalyst and solvents that could
be used in a subsequent reaction cycle.
Sequestration of Primary Amines by a Heptane-Soluble Polymer
Under Latent Biphasic Conditions. In a typical reaction, 2 mmol of
a primary amine and 70 µL of propylene oxide (1 mmol) were dissolved
in 10 mL of ethanol and heated at 70 °C for 2 h. Then 10 mL of a
heptane solution of 1.25 g of the polymer sequestrant PMAOVE (3
mmol) was added to this solution. After stirring 10 min, this miscible
mixture of product, sequestered excess amine and polymer was induced
to phase separate into two phases by the addition of 1.5 mL of water.
While sequestration of primary amines in the heptane-soluble polymer
was quantitative, the amino alcohol products from longer chain amines
(e.g., octylamine) had some solubility in the heptane phase. In these
cases, the nonpolar phase was extracted with aqueous ethanol to
completely recover the product. Yields of amino alcohol products were
typically in the range of 80-95% (GC yields). The absence of any
contaminating primary amine in the amino alcohol product was verified
by GC chromatography using dodecane as an internal standard.
Synthesis of PNAEAM. PNASI (1 g, 5.9 mmol of active ester)
was first prepared following a literature procedure.12a This reactive ester
polymer was then dissolved in 30 mL of dry DMF with heating. To
this solution was added 22 mg (0.059 mmol) of amine-terminated
methyl red.17 This solution was stirred for 4 h. Then amino ethanol (1
g, 16.4 mmol) was added and the reaction stirred at room temperature.
During this time a white precipitate formed. The reaction was allowed
to stir overnight. The mixture was transferred to a centrifuge tube and
centrifuged for about an hour. The red supernatant was removed with
Acylation Using PNODAM-Supported DAAP Catalyst. An
equimolar mixture of 1 mmol of the substituted phenol and 1.02 mol
of Boc anhydride was dissolved in 10 mL of ethanol. Then 10 mL of
a heptane solution containing 97 mg of the PNODAM-supported DAAP
catalyst (1 mol %) was added and the mixture was stirred for 30 min.
Phase separation was achieved by the addition of 1.5 mL of water,
and the aqueous ethanol phase was concentrated and extracted with
dichloromethane to obtain the product. The heptane phase was recovered
and mixed with fresh substrate solution in ethanol for the next catalytic
cycle.
Poly(maleic anhydride-c-octadecyl vinyl ether) (PMAOVE).
Maleic anhydride was recrystallized from benzene before use. In a
typical copolymerization, maleic anhydride (3.3 g, 33.7 mmol) and
octadecyl vinyl ether (10 g, 33.7 mmol) were dissolved in 30 mL of
benzene. The reaction flask was then evacuated and backfilled with
N2 5 times. After heating the resulting solution to 60 °C, 2 mL of a
toluene solution of benzoyl peroxide (10 mg, 0.041 mmol) was added
to initiate the polymerization. After 20 h at 60 °C, the reaction mixture
was diluted with 40 mL of ether and poured into 900 mL of acetone to
provide a white powder. This precipitate was isolated by filtration and
1
dried to yield 10.5 g (79%) of the desired product polymer. H NMR
(CDCl3) δ 3.7-3.2 (br s, 2H), 1.7-1.2 (br m, 37H), 0.9 (t, 6.6 Hz,
3H). IR (KBr, cm-1) 1859, 1781, 1724, 1465. GPC analysis of a
N-benzyl imide derivative of this product showed it had Mn ) 730 000
and Mw ) 1 950 000.
PMAOVE Substituted with Octadecylamine and Morpholine.
PMAOVE (1 g, 2.5 mmol) was dissolved in 50 mL of CH2Cl2 and
allowed to react with a mixture of octadecylamine (1.36 g, 5.0 mmol)
and morpholine (442 mg, 5.0 mmol) in 25 mL of CH2Cl2. The reaction
(18) Roy, R.; Tropper, F. D.; Romanowska, A. J. Chem. Soc., Chem. Commun.
1992, 1611-1613.
9
J. AM. CHEM. SOC. VOL. 125, NO. 20, 2003 6259

A R T I C L E S
Bergbreiter et al.
a pipet and the solvent was removed under vacuum. The resulting red
solid was dissolved in 20 mL ethanol with heating. This dissolution
was also slow. The red solution was added dropwise via pipet to 150
mL of hexane to form an orange powder. The powder was isolated by
filtration. While on the filter, the solid started to become tacky, possibly
from water in the air, so the solid was transferred to a vial and dried
under vacuum to give 460 mg of an orange solid. This product was
characterized by 1H NMR spectroscopy but the spectrum was of poor
quality with broad peaks. 1H NMR (DMSO-d6) δ 7.4-7.7 (br m, 1 H),
4.8-5.1 (br m, 1 H), 3.0-3.4 (br m, 5 H), 1.8-2.1 (br m, 1 H), 1.2-
1.4 (br m, 2 H).
Acknowledgment. Support of this work by the National
Science Foundation (CHE-0010103), Polium Technologies and
the Robert A. Welch Foundation is gratefully acknowledged.
P.L.O. acknowledges the National Science Foundation for
support from a Graduate Research Fellowship Award.
JA0342852
9
6260 J. AM. CHEM. SOC. VOL. 125, NO. 20, 2003