Vol. 2

CRITICAL PHASE POLYMERIZATIONS

111

CRITICAL PHASE POLYMERIZATIONS

Introduction

A supercritical fluid (SCF) can be defined as a substance or mixture which has
both a temperature and pressure that exceed its critical temperature (T

c

) and

critical pressure (P

c

), and a density at or above its critical density (1) (Fig. 1).

Near this critical point, the density, transport properties (such as diffusivity and
viscosity), and other physical properties (such as dielectric constant and solvent
strength) can be varied in a continuum from gas-like to liquid-like, with relatively
small changes in temperature or pressure. The tunability of SCFs makes them
very attractive as solvents for polymerization reactions.

The first example of polymerization in an SCF and the most commercially

significant till this time is the high pressure synthesis of low density polyethylene
(PE) (2). In this process, ethylene from a recycle stream and a makeup source are
compressed to 250 MPa (2500 bar) in a tubular reactor at 250

◦

C. The free-radical

reaction is initiated with either peroxide or traces of oxygen. The polymer pro-
duced has an average molecular weight higher than 100,000 g/mol that is soluble
in the ethylene monomer at the process conditions. Polyethylene is precipitated
and collected in a separator downstream of the reactor section by reducing the
pressure of the solution to 35 MPa (350 bar), thus changing the solvent quality of
ethylene.

In an effort to reduce emission of volatile organic compounds, to completely

phase out the use of chlorofluorocarbons (CFCs), and to reduce the generation of
aqueous waste streams generated in the polymer industry, scientists have turned
to the use of supercritical carbon dioxide (scCO

2

). In the 1990s, CO

2

has proven

to be a suitable alternative solvent for critical phase polymerization reactions
(3). It has an easily accessible critical point with a T

c

of 31.1

◦

C and a P

c

of 7.38 MPa

Encyclopedia of Polymer Science and Technology. Copyright John Wiley & Sons, Inc. All rights reserved.

112

CRITICAL PHASE POLYMERIZATIONS

Vol. 2

solid

liquid

Supercritical

Fluid

Region

vapor

530

31.1

−57.0

7380

Pressure

, kP

a

Temperature,

°C

Fig. 1.

Supercritical fluid phase diagram. To convert kPa to bar, divide by 100.

−43°C

−23°C

−3°C

17

°C

97

°C

77

°C

57

°C

37

°C

1.25

1.00

0.75

0.50

0.25

0.00

Pressure, MPa

10

20

30

Density

, g

/mL

Fig. 2.

CO

2

density profile. To convert MPa to psi, multiply by 145.

(73.8 bar) (4). Within this region, scCO

2

exists as a low viscosity medium with a

tunable density (Fig. 2). It is nontoxic, nonflammable, relatively inexpensive, nat-
urally abundant, and generally regarded as safe. Because CO

2

is an ambient gas,

it can be separated from the polymer by depressurizing the reaction vessel, result-
ing in a dry polymer product. CO

2

is also inert toward free radicals and cations,

which lessens chain-transfer reactions. These aspects make CO

2

very attractive

Vol. 2

CRITICAL PHASE POLYMERIZATIONS

113

for use in synthesis of industrially important polymers. Numerous examples exist
of both chain-growth and step-growth polymerizations using CO

2

as a reaction

medium. This article focuses on the utilization of liquid and scCO

2

as a reaction

medium for polymer synthesis.

Knowledge of two important issues is important when CO

2

is used as a

polymerization solvent: solubility and plasticization. CO

2

is a nonpolar solvent

and should generally dissolve anything that dissolves in hexane. This concept is
valid for many low molar mass compounds that have appreciable vapor pressures;
however, it fails in the case of polymers that have negligible vapor pressures (5).
In fact, the only polymers that show good solubility in CO

2

under mild conditions

[

<100

◦

C,

<35 MPa (<350 bar)] are amorphous fluoropolymers, siloxanes (6,7),

and polyether carbonates (8). CO

2

has a very low polarizability per unit volume

(27.6

× 10

− 25

cm

3

) (9) and a low dielectric constant (1.01–1.67) (10), making

its solvating power similar to that of a fluorocarbon oil. Because of structural
symmetry, CO

2

lacks a dipole moment; however, it does have a strong quadrapole

moment that contributes to its solubility parameter. This quadrapole moment
along with the Lewis acidity of CO

2

allows it to participate in specific solute–

solvent interactions that allow certain polymers to have enhanced solubilities
in CO

2

. Experiments conducted using Fourier transform infrared spectroscopy

reveals that CO

2

is capable of having Lewis acid–base type interactions with

polymers that have electron-donating functional groups such as carbonyls (11).
Amorphous fluoropolymers have an especially high solubility that is attributed to
specific interactions such as the formation of weak CO

2

–fluoropolymer complexes

or favored clustering of the CO

2

near the more polar C F bonds of the polymer.

High resolution

1

H and

19

F nmr studies confirm that such interactions do exist,

thus causing magnetic shielding of the fluorine nuclei (12,13).

The unique solubility characteristics of CO

2

limit the types of polymerization

techniques that can be successfully employed. While amorphous fluoropolymers
may be synthesized homogeneously in CO

2

, most other polymers are insoluble

in CO

2

and must be made via heterogeneous processes such as precipitation,

dispersion, and emulsion techniques or via mass polymerizations swollen by CO

2

.

The plasticization of polymers by CO

2

, which causes a decrease in the glass-

transition temperature (T

g

) of the polymer, is another important feature that must

be taken into account. This plasticization facilitates the occurrence of important
effects that are essential to polymer synthesis. Plasticization of the polymer al-
lows enhanced diffusion of monomer and initiator into the polymer phase which
often results in increased polymerization rates in heterogeneous polymerizations,
removal of residual monomer, solvent, or catalyst from the polymer, and the forma-
tion of blends by polymerization within a CO

2

-swollen host polymer. An important

emerging issue is the role that CO

2

plays in manipulating the loci of reactants

such as monomer/comonomer/initiator partition coefficients. Each of these topics
is discussed in this chapter.

Chain-Growth Polymerizations in CO

2

Homogeneous Polymerizations.

Fluoropolymers are used in many

technologically demanding applications because of their balance of high

114

CRITICAL PHASE POLYMERIZATIONS

Vol. 2

performance properties. Several significant impediments to the synthesis of com-
mercially important fluoropolymers exist including their general insolubility in
most solvents, the undesirable effect of chain transfer of fluoroolefin radicals to
most hydrogen-containing solvents (hence the extensive use of CFC solvents),
and their high hydrophobicity (which necessitates the use of specially fluorinated
surfactants when synthesized in water). Preparing these amorphous low melt-
ing materials in scCO

2

can circumvent the environmental concerns for CFCs and

can avoid the ensemble of problems associated with their synthesis in water. The
feasibility of utilizing liquid and scCO

2

as a solvent for homogeneous solution

polymerization of high molar mass amorphous fluoropolymers by using thermal
free-radical initiation has been demonstrated (14–16). Several fluorinated acry-
lates have been polymerized using this methodology to give high yields of polymer.
For example, poly(1,1-dihydroperfluoroctyl acrylate) (PFOA) (50 w/v) was poly-
merized using 2,2



-azobis(isobutyronitrile) (AIBN) as the initiator at 59.4

◦

C and

20.7 MPa (207 bar) CO

2

for 48 h (eq. (1)).

(1)

The results from a uv spectroscopy study revealed that the decomposition

rate of AIBN in CO

2

is 2.5 times slower than those observed in more viscous

solvents (15). This was attributed to the difference in dielectric constants of the
solvents. In contrast, higher initiator efficiencies were observed in CO

2

, relative

to other solvents, as a result of a decreased solvent cage effect in the supercriti-
cal phase due to the inherently low solvent viscosity. Other fluorinated acrylates,
methacrylates, and perfluoroalkyl derivatized styrenes have been homogeneously
polymerized or copolymerized in CO

2

(17–19). The free-radical propagation ki-

netics of styrene, methyl methacrylate, and butyl methacrylate in CO

2

has been

evaluated using a pulsed laser polymerization (PLP) (20–22). The propagation rate
was found to be slightly lower than that in the bulk, which indicates that CO

2

does

not interfere with the chain-growth process. Homogeneous free-radical polymer-
izations have also been used to synthesize telomers of 1,1-difluoroethylene (23).
The perfluoroalkyl iodides prepared in this manner had both controlled molecular
weights and relatively narrow molecular weight distributions.

Precipitation Polymerizations.

A variety of polymers have been syn-

thesized in CO

2

by free-radical precipitation polymerization. In the earliest

studies, the free-radical polymerization of ethylene in CO

2

using either gamma

radiation or AIBN initiation was investigated (24–26). Other patents (27,28) dis-
closed the polymerization of several vinyl hydrocarbon monomers in liquid and
scCO

2

. Homopolymers of vinyl chloride, acrylonitrile, acrylic acid, vinyl acetate,

2-hydroxyethylacrylate, and some N-vinylcarboxamides as well as PS/PMMA
and PVC/PVAc random copolymers have been made. The polymerization re-
actions resulted in polymers with gravimetric yields ranging from 15 to 97%
and viscosity average molecular weights (M

v

) ranging from 1.2

× 10

4

to 1.6

×

10

6

g/mol depending on the monomer and the reaction conditions used. More re-

cently, the free-radical polymerization of high molecular weight poly(acrylic acid)

Vol. 2

CRITICAL PHASE POLYMERIZATIONS

115

(M

n

= 1.5×10

5

g/mol) has been achieved in scCO

2

despite the fact that the polymer

precipitates from solution at molecular weights greater than 1000 g/mol (29,30).
The molecular weight of the polymer could be controlled by either adding ethyl
mercaptan as a chain-transfer agent or by manipulating the temperature and
pressure parameters for the reaction. High performance fluoropolymers have also
been synthesized in CO

2

by free-radical precipitation polymerization. Homo- and

copolymers of tetrafluoroethylene (TFE) have generated particular interest, since
it was discovered that TFE can be handled more safely as a mixture with CO

2

(31).

In 1995, homopolymerizations (32) of TFE and copolymerizations (33,34) of TFE
with hexafluoropropylene (HFP) or perfluoropropyl vinyl ether (PPVE) yielding
high molecular weight polymers (

>10

6

g/mol) in good yield were reported. Copoly-

merization of TFE and perfluoroalkyl vinyl ether monomers in CO

2

occurs with vir-

tually no

β-scission, implying that propagation of terminal vinyl ether-containing

radicals is more favorable in CO

2

than in traditional solvents. A system for the con-

tinuous polymerization of various monomers in scCO

2

has been developed (35).

Continuous precipitation polymerization of vinylidene fluoride and acrylic acid
can occur in an intensely agitated continuous stirred tank reactor (CSTR). Rates
of polymerization of vinylidene fluoride in the CSTR are significantly higher than
the average rates of batch polymerization, under similar conditions. CO

2

has also

been employed as a solvent in precipitation polymerizations of cross-linked poly-
mers, using high levels of divinylbenzene cross-linker (36). The AIBN-initiated
reactions were performed at 65

◦

C for 12 h in 31 MPa (310 bar) CO

2

and yielded

spherical particles without the use of surfactants. Multifunctional monomers such
as trimethylolpropane trimethacrylate and ethylene glycol dimethacrylate were
also investigated (37). These reactions produced well-defined polymer monoliths
with pore diameters ranging from 20 to 8000 nm (see the “Porous Polymer Syn-
thesis” section). A clear advantage to the use of CO

2

in each of these precipitation

reactions is that the products are easily isolated from the depressurized reaction
vessels as dry, solvent-free materials.

Dispersion and Emulsion Polymerizations.

The relative insolubility

of many industrially important polymers in scCO

2

necessitates the use of het-

erogeneous polymerization techniques conducted with added stabilizers. Conven-
tional heterogeneous dispersion polymerizations of unsaturated monomers are
performed in either aqueous or organic dispersing media, with the addition of
interfacially active agents to stabilize the colloidal dispersion that forms (38,39).
Successful stabilization of the polymer colloid during polymerization results in
the formation of high molar mass polymers with high rates of polymerization. A
similar analogy applies to dispersion polymerization in CO

2

where stabilization

can be accomplished with the aid of nonionic homopolymers, block copolymers,
and reactive macromonomer surfactants (40) (Fig. 3). Typically, a free-radical
dispersion polymerization in scCO

2

starts as a one-phase, homogeneous system

such that both the monomer and the polymerization initiator are soluble in the
polymerization medium. As the reaction progresses, the resulting growing poly-
mer phase separates into primary particles that are stabilized by amphiphilic
materials so as to prevent particle flocculation and aggregation. Polymer colloids
produced by dispersion polymerizations in CO

2

are usually stabilized by a steric

mechanism as compared with an electrostatic mechanism that is common to col-
loidal stabilization in aqueous environments because of the low dielectric con-
stant of CO

2

. Steric stabilization of a colloidal dispersion is usually imparted by

116

CRITICAL PHASE POLYMERIZATIONS

Vol. 2

Fig. 3.

Surfactants for CO

2

.

Vol. 2

CRITICAL PHASE POLYMERIZATIONS

117

amphiphilic macromolecules that become adsorbed onto the surface of the dis-
persed phase. These macromolecules contain an anchoring segment that attaches
to the particle usually by physical adsorption and stabilizing moieties that are sol-
uble in the continuous phase. The stabilizer prevents aggregation of particles by
coating the surface of each particle and imparting long-range repulsions between
them.

A substantial number of studies on the dispersion polymerization of lipophilic

monomers in CO

2

have focused on methyl methacrylate (MMA). The first success

in this area exploited the amphiphilic nature of the homopolymer PFOA as a
stabilizer (40) (eq. (2)).

(2)

MMA was polymerized heterogeneously to very high conversions (

>90%)

and high degrees of polymerization (

>3000 g/mol) in scCO

2

. These polymeriza-

tions were conducted at 65

◦

C and 20.7 MPa (207 bar) CO

2

with AIBN or fluo-

rinated derivatives of AIBN as initiators and PFOA (M

n

= 1.1 × 10

4

or 2.0

×

10

5

g/mol) as the surfactant. The reactions resulted in kinetically stable disper-

sions of micrometer-sized particles with a narrow size distribution. In contrast, re-
actions that were performed without PFOA as a stabilizer resulted in precipitated
polymer with an unstructured, nonspherical morphology. Several other monomers
were also polymerized by dispersion polymerization methods in CO

2

(Table 1).

Table 1. Summary of Dispersion and Emulsion Polymerizations in CO

2

Monomer(s)

Stabilizers

Reference

Acrylamide

Amide end-capped PFPE

41

2,6-Dimethylphenylene

PFOA, PFOA-based random

42

oxide

copolymers, PS-b-PFOA

Divinylbenzene and

Fluorinated methacrylate–PMMA

37

ethylvinylbenzene

block copolymer

Methyl methacrylate

PFOA

40,43,44

PDMS macromonomer

(45,46)–(47)

Fluorinated graft or block copolymers

48

Styrene

PFOA, PS-b-PFOA

49

PDMS, PS-b-PDMS

50

FVE-b-MVE

51

Carboxylic acid terminated PFPE

52

Vinyl acetate

PFOA, PDMS, PDMS macromonomer,

53

PVAc-b-PDMS and PVAc-b-PFOA

Vinyl acetate and ethylene

PDMS, PVAc-b-PFOA

53

1-Vinyl-2-pryrrolidinone

PFOA

54

Acrylonitrile

PFOA, PS-b-PFOA

55

118

CRITICAL PHASE POLYMERIZATIONS

Vol. 2

Typical emulsion polymerizations utilize oil-soluble monomers dispersed in

an aqueous media with a water-soluble initiator, whereas “inverse” emulsions
employ water-soluble monomers dispersed in an organic medium containing an
oil-soluble initiator. The insoluble polymer particles that result from these re-
actions are stabilized as colloids in solution by repulsive forces imparted by a
small molecule ionic surfactant and/or an amphiphilic macromolecular stabilizer
(56,57). These reactions produce high molecular weight, spherical polymer par-
ticles with sizes typically smaller than 1

µm. Since most common monomers

are highly soluble in CO

2

, there are very few examples of CO

2

-based emulsion

polymerizations. However, acrylamide is a rare example of a vinyl monomer that
has a low solubility in CO

2

at moderate temperatures and pressures. The AIBN-

initiated water-in-oil or “inverse” emulsion polymerization of acrylamide has been
attempted at 65

◦

C in 35.2 MPa (352 bar) CO

2

(41,58) (eq. (3)).

(3)

Reactions performed with an amide end-capped perfluoropolyether (PFPE)

appeared to form a milky-white latex, but produced polymers with conversions
and molecular weights similar to those polymers precipitated in reactions with-
out a stabilizer. More inverse emulsion studies are expected with the utiliza-
tion of surfactants that have the ability to form water in CO

2

microemulsions.

Several researchers have demonstrated that various amphiphilic surfactants
like the graft copolymer PFOA-g-PEO (59), an ammonium carboxylate PFPE
[(OCF

2

-CF(CF

3

))

n

(OCF

2

)

m

]OCF

2

COO

−

NH

4

+

(60), and the hybrid hydrocarbon–

fluorocarbon C

7

F

15

CH(OSO

3

−

Na

+

)C

7

H

15

(61) are able to stabilize water-in-CO

2

microemulsions. In principle, these same surfactants could be used to form mi-
croemulsions of water and water-soluble monomers in CO

2

for inverse emulsion

polymerizations.

In Situ Synthesis of Polymer Blends.

The ability of scCO

2

to plasticize

most polymers enables the synthesis of polymer blends that are not easily accessi-
ble by other methods. Supercritical CO

2

can be used to swell a CO

2

-insoluble poly-

mer matrix and subsequently infuse CO

2

-soluble monomers and initiators into the

matrix for polymerization. In one system, poly(chlorotrifluorethylene) (PCTFE),
poly(4-methyl-1-pentene) (PMP), high density polyethylene (HDPE), nylon-6,6-
poly(oxymethylene), and bisphenol A polycarbonate (PC) were each swollen with
CO

2

and infused with styrene monomer and either AIBN or tert-butyl perbenzoate

initiators (62,63). Polymerization was thermally initiated within the polymer ma-
trix either in the presence of CO

2

or N

2

to generate a polymer blend. Mass uptakes

of up to 118% based on the original mass of the polymer were observed. In the
case of PCTFE, extraction of polystyrene from the matrix and subsequent gpc anal-
ysis showed high molecular weight polymer (M

n

> 1 × 10

5

g/mol) in the blend.

Transmission electron microscopy and energy dispersive x-ray analysis were used
to demonstrate that the polystyrene exists as discrete phase-segregated regions

Vol. 2

CRITICAL PHASE POLYMERIZATIONS

119

throughout the matrix polymer. Thermal analysis results revealed that radical
grafting reactions are not significant. Other materials including PS/PE compos-
ites (64) and FEP/PS blends (65) can be synthesized by related routes. In a simi-
lar study this strategy has been used as a route to polymer surface modification.
Styrene (with and without various cross-linkers) was infused into three polymer
substrates (PTFE, PCTFE, and FEP), polymerized, and subsequently sulfonated
in order to provide surface modified semi-interpenetrating networks (66). An in-
crease in wettability of the modified fluoropolymer matrices was shown using
water contact angle measurements.

Cationic Polymerizations.

Homogeneous and heterogeneous cationic

polymerizations of various monomers in liquid and scCO

2

have been performed

(Table 2). Several different catalyst systems were employed in these polymeriza-
tions, all of which proved to be effective. The early experiments conducted using
monomers such as isobutylene (67) and formaldehyde (70–72) normally resulted
in a low yield of low molecular weight products. Nonetheless, proof of concept
for the use of CO

2

as an effective solvent for cationic chain-growth reactions was

achieved. Other polymers based on different substituted alkenes (73), vinyl ethers
(74,75) (eq. (4)), and oxetanes (75) (eq. (5)) have been synthesized.

Table 2. Summary of Cationic Polymerizations in CO

2

a

Monomer (s)

Catalyst System

Reference

Isobutylene

AlBr

3

/ethyl bromide (cosolvent)

67

TiCl

4

/isopropyl chloride

67

2-(2,4,4-trimethylpentyl) chloride

68

(TMPSCl)/TiCl

4

/methyl chloride

Isobutylene/styrene

TMPSCl/TiCl

4

/methyl chloride

69

Ethyl vinyl ether

SnCl

4

or BF

3

-O(C

2

H

5

)

2

28

Formaldehyde

Acetic acid or trifluoroacetic acid

70,71–72

3-Methyl-1-butene

Water/AlCl

3

/methyl chloride

73

4-Methyl-1-pentene

Water/AlCl

3

/methyl chloride

73

Isobutyl vinyl ether (IBVE)

Acetic acid adduct of IBVE

74,75

C

2

H

5

AlCl

2

C

2

H

5

OOCCH

3

(Lewis base deactivator)

2-(N-propyl-N-perfluorooctyl-

Acetic acid adduct of FVE

74,75

sulfonamide ethyl vinyl ether
(FVE)

C

2

H

5

AlCl

2

/C

2

H

5

OOCCH

3

Bis(ethoxymethyl)oxetane

BH

3

-THF/NaOH

b

76

3-Methyl-3



[1,1-dihydrohepta-

BH

3

-THF/NaOH

76

fluorobutoxy)methyl]-oxetane

FVE/methyl vinyl ether

Acetic acid adduct of FVE C

2

H

5

AlCl

2

/

76

C

2

H

5

OOCCH

3

Styrene

TiCl

4

c

77

a

reactions are homogeneous unless noted.

b

heterogeneous precipitation reaction.

c

heterogeneous dispersion reaction.

120

CRITICAL PHASE POLYMERIZATIONS

Vol. 2

(4)

(5)

Fast reaction kinetics generated by the high reactivity of carbocations is a

major benefit in cationic polymerizations. However, this same reactivity leads to
unwanted side reactions such as chain transfer to monomer and early chain ter-
mination by abstraction of acidic hydrogen atoms beta to the carbocation. CO

2

has

several features that make it a suitable solvent for cationic polymerization reac-
tions. It is relatively inert toward the carbocationic propagating species and is not
incorporated into the polymer backbone (75). CO

2

as a solvent mimics the typical

nonpolar solvents that are conventionally used by affecting the equilibrium be-
tween contact pairs and solvent-separated pairs. This in turn affects the activation
energy for chain transfer and termination reactions. To minimize the limitations of
cationic polymerizations, lower reaction temperatures (typically

−70

◦

C to

−30

◦

C)

as well as controlled polymerization techniques are used. Living cationic polymer-
ization methods that involve the stabilization of the cationic propagating species
with a nucleophilic counterion or an added Lewis base have been developed and
allow well-defined polymer and block copolymers to be prepared (75).

Transition-Metal-Catalyzed Polymerizations.

Metal-catalyzed poly-

merizations (76) have been performed in scCO

2

(77,78). The ring-opening metathe-

sis polymerization (ROMP) of norbornene has been performed with CO

2

and

CO

2

/methanol mixtures using a Ru(H

2

O)

6

(O

3

SC

6

H

4

CH

3

-p)

2

catalyst as the ini-

tiator (78) (eq. (6)).

(6)

These reactions were carried out at 65

◦

C and 6–34.5 MPa (60–345 bar)

CO

2

and yielded polymers with conversions, molecular weights, PDIs, and poly-

mer microstructures similar to those obtained in other solvent systems. Addi-
tion of methanol as a cosolvent to the polymerization medium accomplished sev-
eral things: it enabled the ruthenium catalyst that was insoluble in neat CO

2

to be solubilized in the CO

2

/monomer mixture; it allowed for increased rates of

Vol. 2

CRITICAL PHASE POLYMERIZATIONS

121

polymerization yielding similar conversions in a shorter period of time compared
to polymerizations done in pure CO

2

; and it strongly affected the cis–trans ratio

of the polymer microstructure by favoring the trans carbene propagating species
(76,79). Both a ruthenium-based Grubbs catalyst (80) and a molybdenum-based
Shrock catalyst (81) have shown even greater activity in the ROMP of norbornene,
producing high molecular weight polymer (10

5

–10

6

g/mol) in good yields (up to

97%) at milder reaction conditions [25–45

◦

C and 11 MPa (110 bar) CO

2

]. The

ruthenium catalyst was also used to polymerize cis-cyclooctene in 50% yield and
a molecular weight of 10

5

g/mol.

An insoluble rhodium catalyst [(nbd)Rh(acac)] has been used to synthesize

polyphenylacetylene in the presence of liquid and scCO

2

(82). The catalyst is sol-

ubilized in the reaction mixture by addition of a perfluoroalkyl-substituted triph-
enylphosphine ligand. Polyphenylactylene formed in good yields (60–75%) had a
molecular weight of about 4

× 10

4

g/mol for THF soluble fractions, and molecular

weight distributions in the range of 8–10.

A number of metal-catalyzed polymerizations have utilized CO

2

as both a

solvent and as a reagent in the reactions. Precipitation copolymerization of either
propylene oxide (83) or cyclohexene oxide (84) with CO

2

in scCO

2

has been cat-

alyzed using heterogeneous zinc catalysts. Copolymerizations of CO

2

and propy-

lene oxide formed PCs with a molecular weight of about 10

4

g/mol and incor-

poration of CO

2

at greater than 90% (eq. (7)). A small percentage of propylene

carbonate by-product was also observed.

(7)

The cyclohexene oxide reactions yielded polymers with M

w

= 3.8 × 10

5

g/mol

and 91% CO

2

incorporation. Catalyst efficiencies for these two systems were low

to average (3–400 g of polymer per gram of Zn), possibly because of the insolubility
of the zinc complexes in the CO

2

/monomer mixture (85). Molecular weight distri-

butions were also very broad (PDI

= 2.4–27). A partially fluorinated porphyrin

complex was found to be soluble in CO

2

, and CO

2

/monomer mixtures catalyzed

the polymerization of cyclohexene oxide and CO

2

(in the presence of a base) form-

ing polymer in good yields with high degrees of carbonate linkages (90–97%) (86).
This system demonstrated substantially higher catalyst efficiencies (up to 3900 g
of polymer per gram of Cr) than other systems and very narrow molecular weight
distributions (PDI

< 1.4).

Thermal Ring-Opening Polymerization.

The thermal ring-opening

polymerization of a silicone-bridged [1]-ferroceneophane in scCO

2

was performed

at 20.7 MPa (207 bar) and 75–130

◦

C (87) (eq. (8)). The polymer molecular weights

were of the same order of magnitude as the polymers prepared by the solvent-free
methods, but were uniformly lower (M

w

= 2.87 × 10

5

vs 5.2

× 10

5

g/mol).

122

CRITICAL PHASE POLYMERIZATIONS

Vol. 2

(8)

In addition to the lower molecular weights, the PDIs of the polymers made

in CO

2

were broader (3.0–5.1 vs 1.5). Increasing the monomer concentration by

75% for the same reaction increased the M

w

of the polymer to 5.9

× 105 g/mol

and decreased the PDI to 2. This synthesis represents a comprehensible way to
make organometallic polymers from highly strained precursors without dealing
with the limitations imposed by melt-phase (88,89) or polycondensation routes
(90).

Step-Growth Polymerizations in CO

2

Melt-Phase Condensation Polymerizations.

Many condensation poly-

merizations are performed in the melt phase to produce high molecular weight
material without the need for organic solvents. These polycondensation reactions
are driven by the removal of small molecule condensates such as water, methanol,
and phenol. Current industrial processes require high temperatures and high
vacuum to allow the removal of the condensate molecule from the polymer melt
phase. The drawback to this method is that the high molecular weight polymers
produced have high viscosities that make polymer processing difficult and costly,
as well as limit the molecular weight of the polymer. CO

2

is capable of acting as a

plasticizing agent for the polymer melt phase, thus increasing the free volume of
the melt and decreasing the viscosity of the melt for greater processability. Using
CO

2

also facilitates the driving force for polycondensation reactions. Plasticiza-

tion of the polymer melt phase provides a greater surface area for condensate
removal, and solubilization of the small molecule condensate in CO

2

assists in

carrying it out of the reactor. This enhancement in condensate removal results in
greater mobility of chain ends, leading to higher reaction rates and higher molec-
ular weight products. This strategy has been pioneered by DeSimone to synthe-
size polymers by step-growth methods, including polycarbonates, polyesters, and
polyamides.

Melt polymerization of bisphenols (bisphenol A, bisphenol P, bisphenol AF,

and bisphenol Z) with diphenyl carbonate in CO

2

with several catalysts have been

achieved (91,92). Polymers with number-average molecular weights ranging from
2.2

× 10

3

to 1.1

× 10

4

g/mol (M

w

= 4.5 × 10

3

to 2.7

× 10

4

) were obtained over a

range of reaction temperatures (180–250

◦

C) and CO

2

pressures [20.7–24.1 MPa

(207–241 bar)]. Reaction conditions were chosen to ensure efficient removal of solu-
bilized condensate (phenol) without extracting the reactants (diphenyl carbonate).
Polycarbonate synthesis from bisphenol A and diphenyl carbonate catalyzed by
tetraphenylphosphonium tetraphenyl borate were also performed in scCO

2

(93)

(eq. (9)).

Vol. 2

CRITICAL PHASE POLYMERIZATIONS

123

(9)

Polymers with high molecular weights (M

n

up to 1.3

× 10

4

g/mol) were ob-

tained at 270

◦

C and 29.6 MPa (296 bar) CO

2

. In this technique a slight excess of

diphenyl carbonate is required because of monomer extraction. Supercritical CO

2

has also been used to induce crystallization of PC prepolymer beads for solid-state
polymerization (94). The amorphous PC beads (M

w

= 2500 g/mol) were rendered

19% crystalline by treatment with scCO

2

and were subsequently polymerized at

temperatures from 180 to 240

◦

C yielding polymers with molecular weights as high

as 1.4

× 10

4

g/mol.

Poly(ethylene terephthalate) (PET) is commonly prepared by melt-phase

polycondensation reactions. Bis(hydroxyethyl) terephthalate (BHET) was con-
verted to PET under a variable flow (2–10 mL/min) of CO

2

at 20.7 MPa (207 bar)

(95). Molecular weights were obtained over the range 3

× 10

3

to 6.3

× 10

3

g/mol and

increased significantly with flow rate and/or reaction time. These values are low
compared to normal molecular weights obtained in commercial processes (approx.
2

× 10

4

g/mol) and may be attributed to the decreased solubility of the ethylene

glycol condensate in CO

2

compared with phenol in the polycarbonate system.

Polyamides have also been synthesized in the melt phase in the presence of

CO

2

by the nylon salt route (96). A 1:1 salt of hexamethylenediamine and adipic

acid was heated at 220

◦

C for 2 h and then at 280

◦

C for 3 h in 20.7 MPa (207 bar)

CO

2

producing nylon-6,6 with molecular weights up to 2.45

× 10

4

g/mol.

Oxidative Coupling.

The synthesis of polypyrrole, an organic conducting

polymer known for its good thermal stability, has been extended to SFCs. Polypyr-
role was synthesized in supercritical fluoroform and in scCO

2

from pyrrole that is

generated in situ by decarboxylation of 2-carboxypyrrole and a chemical oxidant
[either FeCl

3

or Fe(CF

3

SO

3

−

)

3

] (97). Fe(CF

3

SO

3

−

)

3

was more soluble in CO

2

and

gave higher yields compared to FeCl

3

(50–60% vs. 87% yield). These yields are still

lower than those obtained from control reactions run in toluene (92%). Unlike the
globular structure normally obtained when polypyrrole is prepared in traditional
solvents, these polymers possess a different morphology with fibers ranging from
100 to 200 nm in diameter and several micrometers long.

Poly(2,6-dimethylphenylene oxide) (PPO) is synthesized via oxidative cou-

pling in a CO

2

continuous phase (43). The reactions proceed at 34.5 MPa (345 bar)

and at either room temperature or 40

◦

C for 20 h using a CuBr/amine/O

2

catalyst

system (eq. (10)). PPO with yields as high as 83% and Mn up to 1.7

× 10

4

g/mol were

precipitated. Small molecule amines (pyridine and dimethylethylamine) and CO

2

-

soluble polymeric amines (block copolymers of FOA and either 4-vinylpyridine or
DMAEA) were used in the reaction.

124

CRITICAL PHASE POLYMERIZATIONS

Vol. 2

(10)

The pyridine proved to be the most effective amine for the system. Although

the polymeric amines produced milky white latexes and served to stabilize the
PPO as a colloid, no increase in yields or molecular weight associated with disper-
sion polymerizations were observed.

Porous Polymer Synthesis.

Supercritical CO

2

is an attractive solvent

for the preparation of porous polymers since conventional processes often require
large volumes of “porogenic” organic solvents that tend to become trapped in
the polymer matrix. Microcellular polyurethane foams have been synthesized in
scCO

2

by treating a range of diisocyanates with various propylene oxide and ethy-

lene oxide polyols (98). Phase separation was found to lead to the generation of
a polymer-rich phase and a CO

2

-rich phase that became porous upon removal

of CO

2

. Another example of the synthesis of porous polymers is the direct for-

mation of aerogels by sol–gel polymerization of alkoxysilanes in scCO

2

(99). A

“water-free” sol–gel polymerization technique is utilized in order to account for
the low miscibility of water and CO

2

(100). In this technique, alkoxysilanes such

as tetramethoxysilane (TMOS) and 1,4-bis-(triethoxysilyl) benzene (BESP) can
react with anhydrous formic acid at 35–40

◦

C in the presence of 41 MPa (410 bar)

scCO

2

for up to 12 h. Opaque white silica aerogels were obtained in nearly quanti-

tative yields (after aging for 12–18 h). The silica gels contained pores larger than
those found in gels made in traditional solvents and had interconnected particles
100–200 nm in diameter. Surface areas of silica gels (260–308 m

2

/g) were similar

to gels formed in ethanol or formic acid. BESP aerogels had surface areas lower
than those prepared by aqueous sol–gel polymerizations in organic solvents (471–
586 m

2

/g vs 1000–1600 m

2

/g) (101). A technique has been developed that exploits

the selation of CO

2

to form well-defined macroporous materials based on acrylate

monomers (IOZ) and by using suspension polymerization methods (103).

Conclusion

The many examples cited in this article suggest that CO

2

is a viable alternative

solvent for numerous polymerizations. The use of CO

2

provides not only environ-

mental advantages over traditional solvents but also performance advantages.
And its tunable properties allow it to be used as a continuous phase in a vari-
ety of both step-growth and chain-growth polymerization techniques with con-
trol heretofore unattainable with conventional solvents. The rapid growth of CO

2

technology in the past ten years has provided fundamental knowledge that will
undoubtedly enable the preparation of more advanced materials in the future.

BIBLIOGRAPHY

1. J. A. Darr and M. Poliakoff, Chem. Rev. 99, 565 (1999).
2. K. M. Scholsky, J. Supercrit. Fluids 6, 103 (1993).
3. A. I. Cooper, J. Mater. Chem. 10, 207 (2000);

J. L. Kendall and co-workers, Chem.

Rev. 99, 543 (1999).

Vol. 2

CRITICAL PHASE POLYMERIZATIONS

125

4. E. L. Quinn and C. L. Jones, Carbon Dioxide, Reinhold Publishing Corp., New York,

1936.

5. J. A. Hyatt, J. Org. Chem. 49, 5097 (1984).
6. M. A. McHugh and V. J. Krukonis, Supercritical Fluid Extraction: Principles and

Practice, 2nd ed., Butterworth-Heineman, Stoneham, 1993.

7. P. E. Jessop and W. Leitner, eds., Chemical Synthesis Using Supercritical Fluids,

Wiley-VCH, Weinheim, 1999.

8. T. Sarbu, T. Styranec, and E. J. Beckman, Nature 405, 165 (2000).
9. R. C. Reid, J. M. Praunsnitz, and B. E. Poling, The Properties of Gases and Liquids,

4th ed., McGraw-Hill Book Co., Inc., New York, 1987.

10. F. G. Keyes and J. G. Kirkwood, Phys. Rev. 36, 754 (1930).
11. S. G. Kazarian and co-workers, J. Am. Chem. Soc. 118, 1729 (1996).
12. A. Dardin and co-workers, Macromolecules 30, 3593 (1997).
13. A. Dardin, J. M. DeSimone, and E. T. Samulski, J. Phys. Chem. B 102, 1775 (1998).
14. J. M. DeSimone, Z. Guan, and C. S. Elsbernd, Science 257, 945 (1992).
15. Z. Guan and co-workers, Macromolecules 26, 2663 (1993).
16. Int. Pat. Appl. PCT/US 93/01626 (1993), J. M. DeSimone.
17. Z. Guan and co-workers, Polym. Prepr. (Am. Chem. Soc., Div. Polym. Chem.) 34, 329

(1993).

18. K. K. Kapellen, C. D. Mistele, and J. M. DeSimone, Polym. Prepr. (Am. Chem. Soc. Div.

Polym. Mater. Sci. Eng.) 74, 256 (1996).

19. M. Shaw, J. Ryan, and C. Erkey, Abstr. Pap. Am. Chem. Soc. 214, 243 (1997).
20. A. M. van Herk and co-workers, Macromolecules 30, 4780 (1997).
21. M. A. Quadir and co-workers, Macromolecules 31, 6481 (1998).
22. S. Beuermann and co-workers, Macromol. Chem. Phys. 199, 1209 (1998).
23. J. R. Combes, Z. Guan, and J. M. DeSimone, Macromolecules 27, 865 (1994).
24. M. Hagiwara and co-workers, J. Polym. Sci., Part A-1 6, 603 (1968).
25. M. Hagiwara and co-workers, J. Polym. Sci., Part A-1 6, 609 (1968).
26. U.S. Pat. 3471463 (1969), T. Kagiya and co-workers.
27. Fr. Pat. 1524533 (1968) (Sumitomo Chemical Co.).
28. U.S. Pat. 3522228 (1970), K. Fukui and co-workers.
29. T. J. Romack, E. E. Maury, and J. M. DeSimone, Macromolecules 28, 912 (1995).
30. H. Q. Hu and co-workers, Acta Polym. Sinica 740 (1998).
31. U.S. Pat. 5345013 (1994), D. J. Van Bramer, M. B. Shiflett, and A. Yokozeki.
32. T. J. Romack, B. E. Kipp, and J. M. DeSimone, Macromolecules 28, 8432 (1995).
33. T. J. Romack and J. M. DeSimone, Macromolecules 28, 8429 (1995).
34. J. P. DeYoung, T. J. Romack, and J. M. DeSimone, Polym. Prepr. (Am. Chem. Soc., Div.

Polym. Chem.) 38, 424 (1997).

35. P. A. Carpentier and co-workers, Macromolecules 32, 5973 (1999).
36. A. I. Cooper, W. P. Hems, and A. B. Holmes, Macromol. Rapid Commun. 19, 353 (1998).
37. A. I. Cooper, W. P. Hems, and A. B. Holmes, Macromolecules 32, 2156 (1999).
38. K. E. J. Barret, Dispersion Polymerization in Organic Media, John Wiley & Sons Inc.,

New York, 1975

39. E. D. Sudol, in J. M. Asua, ed., Polymeric Dispersions: Principles and Applications,

Vol. 335, Kluwer Academic Press: Boston, Mass. 1997, pp. 141–153.

40. J. M. DeSimone and co-workers, Science 265, 356 (1994).
41. F. A. Adamsky and E. J. Beckman, Macromolecules 27, 312 (1994).
42. K. K. Kapellen, C. D. Mistelle, and J. M. DeSimone, Macromolecules 29, 495 (1996).
43. Y. L. Hsiao and co-workers, Macromolecules 28, 8159 (1995).
44. Y. L. Hsiao and J. M. DeSimone, J. Polym. Sci., Part A: Polym. Chem. 35, 2009 (1997).
45. K. A. Shaffer and co-workers, Macromolecules 29, 2704 (1996).
46. M. L. O’Neill and co-workers, Macromolecules 31, 2848 (1998).
47. M. L. O’Neill and co-workers, Macromolecules 31, 2838 (1998).

126

CRITICAL PHASE POLYMERIZATIONS

Vol. 2

48. C. Lepilleur and E. J. Beckman, Macromolecules 30, 745 (1997).
49. D. A. Canelas, D. E. Betts, and J. M. DeSimone, Macromolecules 29, 2818 (1996).
50. D. A. Canelas and J. M. DeSimone, Macromolecules 30, 5673 (1997).
51. M. R. Clarke, J. L. Kendall, and J. M. DeSimone, Macromolecules 30, 6011 (1997).
52. P. Christian and S. M. Howdle, Macromolecules 33, 237 (2000).
53. D. A. Canelas and co-workers, Macromolecules 31, 6794 (1998).
54. T. Carson, J. Lizotte, and J. M. DeSimone, Macromolecules, 33, 1917–1920 2000.
55. H. Shiho and J. M. DeSimone, Macromolecules, 33, 1565–1569 2000.
56. D. C. Blackley, Emulsion Polymerization, John Wiley & Sons, Inc., New York, 1975.
57. R. G. Gilbert, Emulsion Polymerization: A Mechanistic Approach, Academic Press,

London, 1995.

58. F. A. Adamsky and E. J. Beckman, Macromolecules 27, 523 (1994).
59. E. E. Maury and co-workers, Polym. Prepr. (Am. Chem. Soc., Div. Polym. Chem.) 6,

205 (1993).

60. K. P. Johnston and co-workers, Science 271, 624 (1996).
61. K. Harrison and co-workers, Langmuir 10, 3536 (1994).
62. J. J. Watkins and T. J. McCarthy, Macromolecules 27, 4845 (1994).
63. J. J. Watkins and T. J. McCarthy, Macromolecules 28, 4067 (1995).
64. E. Kung, A. J. Lesser, and T. J. McCarthy, Macromolecules 31, 4160 (1997).
65. K. A. Arora, A. J. Lesser, and T. J. McCarthy, Macromolecules 32, 2562 (1999).
66. P. Rajagopalan and T. J. McCarthy, Macromolecules 31, 4791 (1998).
67. R. Biddulph and P. Plesch, J. Chem. Soc. 3913 (1960).
68. T. Perneker and J. P. Kennedy, Polym. Bull. 32, 537 (1994).
69. U.S. Pat. 5376744 (1994), J. P. Kennedy and T. Pernecker.
70. H. Yokota and co-workers, J. Polym. Sci., Part A-1 42, 1412 (1969).
71. T. Pernecker, G. Deak, and J. P. Kennedy, Macromol. Rep. A 32, 969 (1995).
72. M. R. Clarke and J. M. DeSimone, Polym. Prepr. (Am. Chem. Soc., Div. Polym. Chem.)

35, 482 (1994).

73. M. R. Clarke and J. M. DeSimone, Macromolecules 28, 3002 (1995).
74. T. Higashimura, M. Sawamoto, and M. Miyamoto, Macromolecules 17, 265 (1984).
75. M. R. Clarke, J. L. Kendall, and J. M. DeSimone, Macromolecules 30, 6011 (1997).
76. S. T. Nguyen, R. H. Grubbs, and J. W. Ziller, J. Am. Chem. Soc. 115, 9858 (1993).
77. C. D. Mistle, H. H. Thorp, and J. M. DeSimone, J. Macromol. Sci., A: Pure Appl. Chem.

33, 953 (1996).

78. J. G. Hamilton and co-workers, Macromolecules 31, 4387 (1998).
79. P. Schwab and co-workers, Angew. Chem. Int. Ed. Engl. 34, 2039 (1995).
80. P. Schwab, R. H. Grubbs, and J. W. Ziller, J. Am. Chem. Soc. 118, 100 (1996).
81. R. R. Schrock and co-workers, J. Am Chem. Soc. 112, 3875 (1990).
82. H. Hori, C. Six, and W. Lietner, Macromolecules 32, 3178 (1999).
83. D. J. Darensbourg, N. W. Stafford, and T. Katsurao, J. Mol. Cat. A: Chem. 104, L1–L4

(1995).

84. D. J. Darensbourg and M. W. Holtcamp, Macromolecules 28, 7577 (1995).
85. D. J. Darnsbourg and M. W. Holtcamp, Coord. Chem. Rev. 153, 155 (1996).
86. S. A. Mang and co-workers, Macromolecules, in press.
87. P. Christrian and co-workers, Polym. Prepr. 8, 434 (1997).
88. C. U. Pittman and S. Yukihiko, Chem. Lett. 383 (1975).
89. D. A. Foucher, B. Z. Tang, and I. Manners, J. Am. Chem. Soc. 114, 6246 (1992).
90. E. W. Neuse and L. Bednarik, Macromolecules 12, 187 (1979).
91. P. G. Odell and G. K. Hamer, Polym. Prepr. 38, 470 (1997).
92. U.S. Pat. 5698665 (1997), P. G. Odell.
93. S. M. Gross and co-workers, Macromolecules 31, 25, 9090 (1998).
94. S. M. Gross and co-workers, Macromolecules 32, 3167 (1999).
95. A. L. Burke, G. Maier, and J. M. DeSimone, Polym. Mater. Sci. Eng. 74, 248 (1996).