Current Opinion in Solid State and Materials Science 8 (2004) 325331

Green synthesis of polymers using supercritical carbon dioxide

Colin D. Wood a, Andrew I. Cooper
a b

b,*

, Joseph M. DeSimone

a,c,*

Department of Chemistry, University of North Carolina at Chapel Hill, NC 27599, USA Donnan and Robert Robinson Laboratories, Department of Chemistry, University of Liverpool, Liverpool, L69 3BX, UK c Department of Chemical Engineering, North Carolina State University, Raleigh, NC 27695, USA

Abstract Supercritical carbon dioxide is the most extensively studied supercritical uid (SCF) medium for polymerization reactions and organic transformations. This can be attributed to a list of advantages ranging from solvent properties to practical environmental as well as economic considerations. Aside from these gains, CO2 nds particularly advantageous application in the synthesis and processing of uoropolymers and porous materials and as such, this will be the subject of this review. The overall aim is to highlight areas where the unique properties of SCF solvents can be exploited to generate polymers that would be dicult or inconvenient to obtain by other routes. 2005 Elsevier Ltd. All rights reserved.
Keywords: Supercritical; Fluoropolymer; Porous polymers; High-throughput

1. Introduction Supercritical carbon dioxide has emerged as the most extensively studied environmentally benign medium for organic transformations and polymerization reactions. This stems from a list of advantages ranging from solvent properties to practical environmental and economic considerations. Increasing environmental regulation has prompted a tremendous eort to develop and apply environmentally benign alternative media or solvent free processes to a variety of commercially signicant, solvent-intensive processes [1]. Carbon dioxide in the liquid and supercritical state is a potential alternative for many of the processes experiencing ever increasing regulation, owing to its non-toxic, non-ammable nature. Additionally, its ubiquity makes it inexpensive and readily available. Several excellent reviews are available that discuss the utilization of carbon dioxide for many solvent- and waste-intensive industries [25]. Aside from the gains
*

provided by CO2 as a reaction medium in general, it nds particularly advantageous application in the synthesis and processing of uoropolymers and porous materials [611] as such, this will be the subject of this review. In particular the aim is to highlight areas where the unique properties of supercritical uid (SCF) solvents can be exploited to generate polymers that would be dicult or inconvenient to obtain by other routes. The general properties of SCFs in relation to chemical synthesis [8] and extraction [2] have been reviewed previously and will not be a focus herein.

2. Fluoroolen polymerization in CO2 2.1. Introduction Fluoropolymers are typically synthesized in aqueous polymerization systems (both emulsion and suspension), non-aqueous systems (Freon-113), or in hybrid Freon113/aqueous hybrid systems [12]. Such processes require the use of large quantities of water, CFCs (for non-aqueous

Corresponding authors. E-mail address: aicooper@liverpool.ac.uk (A.I. Cooper).

1359-0286/$ - see front matter 2005 Elsevier Ltd. All rights reserved. doi:10.1016/j.cossms.2005.02.001

326

C.D. Wood et al. / Current Opinion in Solid State and Materials Science 8 (2004) 325331

polymerizations), and uorinated surfactants for emulsion polymerization. Many of the uorinated surfactants typically employed in aqueous emulsion and suspension polymerizations are currently under scrutiny due to bioaccumulation and environmental persistence [13,14]. These issues have collectively resulted in an increasingly urgent impetus for transition from the conventional uoropolymer synthesis platforms to alternatives that meet the requirements of emerging public and regulatory demands. Most commercially available uoropolymers are prepared from a relatively small group of olens including tetrauoroethylene (TFE), chlorotriuoroethylene (CTFE), vinylidene uoride (VF2), hexauoropropylene (HFP), ethylene, and peruoroalkyl vinyl ethers (PAVEs). Many of these monomers are ammable and some are explosive. For example, TFE is ammable when mixed with air and has a high propensity for explosion during expansion to a gas from its liquid phase under pressure. Further, TFE is highly explosive as a gas at elevated temperatures. In the presence of oxygen it will undergo autopolymerization, a process suciently exothermic to ignite an explosion. Aqueous suspension and dispersion polymerizations of copolymers of TFE with a variety of comonomers including hexauoropropylene (HFP) and various peruoroalkyl vinyl ethers (PAVEs) typically exhibit high levels of carboxylic acid end groups [7,15]. The presence of acid end groups often proves deleterious to the intended properties and function of the polymer not to mention the introduction of complications into certain post polymerization, melt-processing steps. This problem is particularly prevalent in copolymers including PAVE comonomers (peruoroalkyl vinyl ethers) which are an essential component in the control of crystallinity in TFE-based uoroplastics. In order to prevent decomposition, discoloration and emissions of

hydrogen uoride, polymers containing high levels of acid end groups may require high-temperature hydrolysis and uorination nishing steps. A propagating polymeric radical derived from a terminal PAVE monomer unit has one of two possible reaction pathways (Scheme 1). First, and most obvious, it can cross propagate to monomer continuing the polymerization reaction or it can undergo b-scission, resulting in an acid uoride terminated polymer and a peruoroalkyl radical capable of initiating further polymerization which is essentially a chain-transfer-to-monomer step [7,15]. The use of CFCs circumvents these problems; however, CFCs have fallen under exceedingly strict regulation due to environmental concerns and as a result, are no longer viable to be used as large scale reaction media. 2.2. TFE-based materials Tetrauoroethylene (TFE)-based copolymers have become premium high performance materials for a broad range of applications that demand chemical and thermal resistance, and melt processing capability [16]. As mentioned earlier, caution must be taken when TFE is used as a monomer due to potential explosions. It has been demonstrated that CO2/TFE mixtures are far less susceptible to ignition as TFE forms a pseudo azeotrope with CO2 [17]. This makes handling and delivering monomer much safer. A range of TFE-based uoropolymers have been successfully synthesized in CO2. These include FEP, PFA, ETFE, TFE/vinyl acetate polymers, Naon-type materials, and Teon AF-type materials (Fig. 1) the details have recently been reviewed elsewhere [11]. Further, it is worth noting that uoropolymers synthesized in sc-CO2 exhibit signicantly diminished levels of acid end groups leading to very high molecular weight materials [15]. Indeed, in such instances, it is even necessary to add chain transfer agents to control molecular weight and maintain meltprocessability of the product [18]. IR analysis has conrmed that uoropolymers synthesized under these conditions exhibit acid end group levels an order of magnitude below (03 end groups per 106 carbon atoms) those which are synthesized in conventional organic and aqueous reaction systems [14,18]. The most likely explanation for this is that the tremendous plasticizing capability that CO2 provides for ecient transport of monomer into the polymer phase, maintaining high eective concentrations of monomer in the vicinity of the active chain ends, thus favoring the bimolecular propagation step over the unimolecular b-scission process (Scheme 1). The other explanation may be the lower temperatures that one can use with CO2 which may favor cross propagation over b-scission. Additionally, it has been demonstrated that dense CO2 is exceptionally inert to free radical chemistry in general [19]. This demonstrates the broad applicability of liquid and supercrit-

F CF2 C O CF2 CF2 CF3

TFE Cross Propagation

F O CF2 CF2 CF3

F F

CF2 C CF2 C

-scission
F CF2 C O CF2 CF2 CF3

O CF2 C F

+
F3C CF2 CF2

Scheme 1. Reaction sequence for b-scission in uoroolen polymerization [7,15].

C.D. Wood et al. / Current Opinion in Solid State and Materials Science 8 (2004) 325331

327

CF2

CF2

CF2

CF
n

CF2

CF2

CF2

CF
n

CF3 Fluorinated Ethylene Propylene Resin (FEP)

ORf Perfluoroalkoxy Resin (PFA)

CF2

CF2

CH2

CH2
n

CF2

CF2

CH2

CH
n

O Ethylene Tetrafluoroethylene Resin (ETFE) CH3 Tetrafluoroethylene-co-vinyl acetate O

CF2

CF2

CF2

CF
n

CF2

CF2 O

CF

CF
n

OCF2CF(CF3)OCF2CF2SO2F Nafion

F3C TeflonAF

CF3

Fig. 1. Tetrauoroethylene-based uoropolymers synthesized in carbon dioxide.

ical CO2 in the synthesis and processing of uorinated polymers which has led to the commercialization of a CO2-based approach in the manufacturing of certain grades of polymers based on tetrauoroethylene (TeflonTM). DuPont has recently invested an initial $40 million in the construction of a TeonTM FEP production facility in North Carolina. The process employs CO2 as the continuous phase instead of 1,1,2-trichloro1,2,2-triuoroethane or water and surfactant [1,2023]. The use of carbon dioxide results in a surfactant free system with no issues of chain-transfer-to-solvent. This process not only avoids many of the environmental and safety concerns but also provides a product with superior properties. This is the rst commercial example of uoropolymer resins made using supercritical carbon dioxide as the solvent. Teon AF is an amorphous copolymer of tetrauoroethylene (TFE) and 2,2-bis(triuoromethyl)-4,5diuoro-1,3-dioxole (PDD). It combines the properties of amorphous plastics, such as optical transparency and solubility in organic solvents, with those of peruorinated polymers, including high thermal stability, excellent chemical stability and low surface energy. Moreover Teon AF exhibits the lowest dielectric constant (1.90 for Teon AF 2400) and the lowest refractive index (1.29 for Teon AF 2400) known for any solid organic polymer [12]. As such, Teon AF is well-suited for use as an optical material. Teon AF-based materials have also been synthesized in supercritical carbon dioxide (see Fig. 1) [24]. A range of copolymers with various compositions and dierent molecular weights were pre-

pared in yields as high as 74% and their properties were compared to commercially available Teon AF. The glass transition temperatures for the materials ranged from 67 to 334 C. The synthesis of these copolymers in CO2 has several key advantages compared to conventional polymerization techniques. The low reaction temperature and the use of a peruorinated initiator results in copolymers with properties which are comparable to those of the analogous commercial product. However, an additional uorination step is not necessary and the product is isolated directly from the reactor without contamination from solvents or surfactants. Moreover, the synthesis uses TFE/CO2 mixtures instead of pure TFE which has inherent safety advantages [17]. 2.3. VF2-based materials VF2 has been successfully synthesized in CO2 in a continuous precipitation process using a continuous stirred tank reactor [25]. In such a process monomer, solvent (CO2), and initiator are continuously added to the reactor. The advantage of a continuous process over a batch process is that the monomer and the CO2 can be removed and recycled, leading to a high rate of polymerization. The authors later reported that, although this was a precipitation polymerization, a model based on the kinetics for a homogeneous polymerization t the experimental data, conrming the reaction order of 0.5 and 1 for the initiator and monomer respectively [26]. Further, the eect of VF2 ow rate, residence time and temperature were investigated in this system [27].

328

C.D. Wood et al. / Current Opinion in Solid State and Materials Science 8 (2004) 325331

It was found that monomer feed concentration had a pronounced eect upon the modality of the molecular weight distribution. At a high monomer concentration, the resulting polymer exhibited a bimodal molecular weight distribution. The authors suggested that at high monomer concentrations the kinetics may no longer adhere to the homogeneous kinetic model proposed earlier [26]. Several other parameters were investigated such as chain branching and agitation but they were unable to explain the bimodality of the polymer. 2.4. Fluoroalkyl acrylate polymerizations in CO2 One of the more dicult issues associated with the use of CO2 as a polymerization medium has been the limited solubility of most high molecular weight polymers in CO2. As such, much of the work reported for polymerizations in CO2 are typically heterogeneous precipitation, dispersion or emulsion processes as discussed above. However, it has been demonstrated that oligomeric peruoropolyethers and oligomeric poly(chlorotriuoroethylene) oils are soluble in liquid CO2 [2]. Additionally, it has been reported that highly uorinated polyacrylates of high molecular weight (>250,000) exhibit exceptional solubility in supercritical CO2. Not surprisingly, since this discovery, it has been demonstrated that high molecular weight, amorphous uoropolymers can be synthesized homogeneously in CO2 utilizing free radical initiators [21,2833]. Most amorphous uoropolymers exhibit solvent resistance to common organic solvents and therefore conventional approaches to synthesis and processing have depended on chlorouorocarbon (CFC) platforms. Owing to the previously mentioned environmental concerns with the use of CFC refrigerants and solvents, CFCs are no longer acceptable for large-scale commercial use. As an eective reaction medium for homogeneous polymerizations under mild conditions, CO2, again, is an increasingly attractive, inexpensive and harmless platform option for the synthesis of highly uorinated monomers (as domestic and international regulation of CFCs becomes ever more restrictive in the coming years). Various uorinated acrylate monomers, such as 1,1-dihydroperuorooctyl acrylate (FOA) have been successfully polymerized in CO2. It is also noted that reaction conditions for these polymerizations are mild in terms of temperature and pressure and not entirely unlike those encountered in conventional solvent based processes [21]. Further, studies have shown CO2 to be inert to radical chemistry, eliminating chain transfer to solvent as a side reaction [8]. Additionally the low viscosity of sc-CO2 removes issues of Tromsdorf eect and autoacceleration [8]. Therefore, as a reaction medium, CO2 is ideally suited to free radical polymerizations. In all cases, the polymerizations remained homogeneous throughout the reaction, illustrating the

high solubility of the polymers in supercritical CO2. The solution properties of poly(FOA) in CO2 have also been investigated using small-angle neutron scattering (SANS) and it was shown that CO2 is an excellent solvent for poly(FOA) [34]. 2.5. Photooxidation of uoroolens in liquid CO2 Another class of uoropolymer known to be readily soluble in CO2 are peruoropolyethers (PFPEs) [2]. As a unique class of uoropolymers, PFPE polymers and copolymers have been established as high performance materials, exhibiting low surface energies and low moduli, as well as excellent thermal and chemical stabilities. PFPEs are primarily found in high performance lubricant applications such as for magnetic data storage media and as heat exchanger uids. Recently, DeSimone et al. developed a novel solvent resistant photocurable PFPE that can be used for the fabrication of microuidic devices [35] and as materials for high resolution soft lithography to replicate sub-100 nm sized features with no indications of limits to going to even smaller in size [36]. Moreover, the material has several specic advantages over traditional glass and PDMS-based materials and broadens the utility of both processes. In addition, the process exemplies a solvent-free approach. One of the main industrial processes for the production of PFPEs is photooxidation of uoroolens [37]. Currently, only TFE and HFP are used commercially in this process. Typically, HFP is photooxidized in bulk owing to its very low reactivity while TFE requires an inert diluent in order to prevent homopolymerization of the olen. DeSimone and coworkers reported recently the photooxidation of various concentrations of HFP in CO2 in concert with parallel reactions carried out in bulk HFP and in peruorocyclobutane for purposes of comparison [38]. Based on the data collected in this study, it was demonstrated that there is a strong dependence of molecular weight and composition on HFP concentration (i.e., lower HFP concentrations gave lower peroxide content). Additionally, viscosity eects in liquid CO2 appeared to signicantly reduce the amount of peroxidic linkages in the nal product. 2.6. Porous materials and supercritical uids Porous materials are used in a wide range of applications, including catalysis, chemical separations, and tissue engineering [9]. However, the synthesis of these materials is often solvent intensive. Supercritical carbon dioxide as an alternative solvent for the production of functional porous materials can circumvent this issue as well as aording a number of specic physical, chemical, and toxicological advantages listed below. For example, in order to dry porous materials often requires energy intensive steps whereas; the transient dry nat-

C.D. Wood et al. / Current Opinion in Solid State and Materials Science 8 (2004) 325331

329

ure of CO2 circumvents these issues. Pore collapse can occur in certain materials when removing conventional liquid solvents; this can be avoided using SCFs because they do not possess a liquidvapour interface. Porous structures are important in biomedical applications (e.g., tissue engineering) where the low toxicity of CO2 oers specic advantages in terms of minimizing the use of toxic organic solvents. In addition, the wetting properties and low viscosity of CO2 oers specic benets in terms of surface modication. 2.7. Free-radical polymerization using sc-CO2 as a pressure-adjustable porogen Recently, Cooper et al. demonstrated the formation of permanently porous crosslinked poly(acrylate) and poly(methacrylates) monoliths using sc-CO2 as the porogenic solvent [39,40]. Materials of this type [41,42] are useful in applications such as high-performance liquid chromatography, high-performance membrane chromatography, capillary electrochromatography, microuidics, [43] and high-throughput bioreactors [44]. In this process, no organic solvents are used in either the synthesis or purication. It is possible to synthesize the monoliths in a variety of containment vessels (Fig. 2), including chromatography columns and narrow-bore capillary tubing. Moreover, the variable density associated with SCF solvents was exploited in order to ne-tune the polymer morphology. Fig. 3 shows the variation in BrunnauerEmmetTeller (BET) surface area for a series of crosslinked monoliths synthesized using sc-CO2 as the porogen over a range of reaction pressures [45]. As can be seen the BET surface area and the average pore size in these materials could be continuously varied over a broad range simply by varying the SCF solvent density. It is interesting to note the minimum in area (and a maximum in the average pore diameter, not shown) was observed at a reaction pres-

350 300

Surface Area (m2/g)

250 200 150 100 50 100

150

200

250 2

300

350

400 4

45 450

Pressure (bar)

Fig. 3. Surface area (as measured by N2 adsorptiondesorption using the BET method) as a function of the CO2 pressure employed during the polymerization. Reproduced with permission; copyright 2003, American Chemical Society [45].

sure of around 180 bar. This can be rationalized by considering the variation in solvent quality as a function of CO2 density and the resulting inuence on the mechanism of nucleation, phase separation, aggregation, monomer partitioning, and pore formation [46]. This approach was also extended to synthesize well-dened porous, cross-linked beads by suspension polymerization, again without using any organic solvents [47]. The surface area of the of the beads could be tuned over a broad range (5500 m2/g) simply by varying the CO2 density. 2.8. Templating of supercritical uid emulsions Emulsion templating is useful for the synthesis of highly porous inorganic, [4851] and organic materials [5254]. In principle, it is possible to access a wide range of porous hydrophilic materials by reaction-induced phase separation of concentrated oil-in-water (O/W) emulsions. A signicant drawback to this process is that large quantities of a water-immiscible oil or organic solvent are required as the internal phase (usually >80 vol.%). In addition, it is often dicult to remove this oil phase after the reaction. Inspired by studies on SCF emulsion stability and formation [55], we have developed methods for templating high internal phase CO2-in-water (C/W) emulsions (HIPE) to generate highly porous materials in the absence of any organic solventsonly water and CO2 are used [56]. If the emulsions are stable one can generate low-density materials (0.1 g/cm3) with relatively large pore volumes (up to 6 cm3/g) from water-soluble vinyl monomers such as acrylamide and hydroxyethyl acrylate. Fig. 4 shows a crosslinked polyacrylamide material synthesized from a high internal phase C/W emulsion, as characterized by SEM and confocal microscopy (scale = 230 lm 230 lm). Comparison of the two images illustrates quite clearly how the porous structure shown in the SEM image is templated from the C/W emulsion (as

Fig. 2. Polymer monolith, synthesized using sc-CO2 as the porogen as isolated from the reaction vessel. The monolith adopts the shape of the reactor. Reproduced with permission; copyright 2003, American Chemical Society [45].

330

C.D. Wood et al. / Current Opinion in Solid State and Materials Science 8 (2004) 325331

[57]. In this study it was shown that all of the problems associated with the initial approach could be overcome and it was possible to synthesize C/W emulsion-templated polymers at relatively modest pressures (60 70 bar) and low temperatures (20 C) using inexpensive and readily available hydrocarbon surfactants. Moreover, we demonstrated that this technique can in principle be extended to the synthesis of emulsion-templated HEA and HEMA hydrogels that may be useful, for example, in biomedical applications [5860]. 2.9. High throughput solubility measurements in CO2 As discussed CO2 is an excellent environmentally benign alternative solvent for the synthesis of uoropolymers and porous materials. Moreover, a range of environmental as well as practical advantages can be realized from using CO2, which has resulted in the commercial synthesis of certain grades of Teon by DuPont. However, the discovery of inexpensive CO2-soluble surfactants, ligands and phase transfer catalysts would greatly increase the utility of CO2. As mentioned the only polymers with appreciable solubility in CO2 are amorphous uoropolymers and polysiloxanes [61]. Inexpensive poly(ether carbonate) copolymers have been reported to be soluble in CO2 under moderate conditions [62,63] and could function as building blocks for inexpensive surfactants, but numerous practical diculties remain. Recently, Cooper et al. reported a new method which allows for the rapid parallel solubility measurements for libraries of materials in supercritical uids [64]. The technique was used to evaluate the solubility of a mixed library of 100 synthetic polymers including polyesters, polycarbonates, and vinyl polymers. This method is at least 50 times faster than other techniques in terms of the rate of useful information that is obtained and has broad utility in the discovery of novel SCF-soluble ligands, catalysts, biomolecules, dyes, or pharmaceuticals for a wide range of materials applications.

Fig. 4. Emulsion-templated crosslinked polyacrylamide materials synthesized by polymerization of a high-internal phase CO2-in-water emulsion (C/W HIPE): (a) SEM image of sectioned material and (b) confocal image of same material, obtained by lling the pore structure with a solution of uorescent dye. As such (a) shows the walls of the material while (b) show the holes formed by templating the sc-CO2 emulsion droplets. Both images = 230 lm 230 lm. Ratio of CO2/ aqueous phase = 80:20 v/v.d Pore volume = 3.9 cm3/g. Average pore diameter = 3.9 lm. Adapted from Butler et al. [56].

3. Conclusions represented by the confocal microscopy image of the pores). In general, the confocal image gives a better measure of the CO2 emulsion droplet size and size distribution immediately before gelation of the aqueous phase. Initially we used low molecular weight (Mw  550 g/ mol) peruoropolyether ammonium carboxylate surfactants to stabilize the C/W emulsions [56] but there are some practical disadvantages of using these surfactants in this particular process such as cost and the surfactant is non-degradable. It was subsequently shown that it is possible to use inexpensive hydrocarbon surfactants to stabilize C/W emulsions and that these emulsions can also be templated to yield low-density porous materials In general CO2 is an attractive solvent alternative for the synthesis of polymers because it is environmentally friendly, non-toxic, non-ammable, inexpensive, and readily available in high-purity from a number of sources. Product isolation is straightforward because CO2 is a gas under ambient conditions, removing the need for energy intensive drying steps. Moreover, additional advantages are realized in the manufacture of uoropolymers such as increased safety in handling explosive monomers and enhanced polymer properties in many cases. As demonstrated CO2 is not only an adequate alternative, but in many cases a superior one. This

C.D. Wood et al. / Current Opinion in Solid State and Materials Science 8 (2004) 325331

331

is demonstrated no more poignantly that DuPonts recent $275 million investment in its CO2 based TeonTM production process. In addition, CO2 aords a number of specic advantages that can be derived from the use of SCFs for the synthesis and modication of porous materials. There are a number of specic chemical advantages that are realized by using CO2 for the synthesis of a range of materials but moreover, they oer the potential of reducing organic solvent usage in the production of materials.

Acknowledgments We thank the NSF STC (grant number 537494), EPA (grant number 535325), ONR (grant number 535747), EPSRC. A.I.C. acknowledges the support of the Royal Society for the provision of a Research Fellowship.

References
[1] DeSimone JM. Science 2002;297:799803. [2] McHugh MA, Krukonis V. Supercritical uid extraction. 2nd ed. Butterworth Heinemann; 1994. [3] Bartle KD, Cliord AA, Jafar SA, Shilstone GF. J Phys Chem, 1991. Ref. Data 20713778. [4] Johnson KP, Penninger, editors. Supercritical uid science and technology. Washington, DC: American Chemical Society; 1989. [5] Phelphs CL, Smart NG, Wai CM. J Chem Ed 1996;73:11638. [6] Kendall JL, Canelas DA, Young JL, DeSimone JM. Chem Rev 1999;99:543. [7] DeYoung JP, Romack TJ, DeSimone JM. In: Hougham G, Cassidy PE, Johns K, Davidson T, editors. Fluoropolymers 1: synthesis. New York: Kluwer Academic/Plenum Publishers; 1999. p. 191205. [8] Davidson TA, DeSimone JM. In: Jessop PG, Leitner W, editors. Chemical synthesis using supercritical uids. New York, Chichester, Brisbane, Singapore, Toronto: Wiley-VCH Weinheim; 1999. p. 297. [9] Cooper AI. Adv Mater 2003;15:1049. [10] Cooper AI. J Mater Chem 2000;10:207. [11] Wood CD, Yarbrough JC, Roberts GW, DeSimone JM. In: Kemmere M, Meyer T, editors. Polymer processes in carbon dioxide. Wiley: New York, NY; in press. [12] Resnick RP, Buck WH. In: Scheirs J, editor. Modern uoropolymers. Chichester: John Wiley & Sons; 1997. p. 397. [13] Moody CA, Field JA. Environ Sci Technol 2000;34:386470. [14] Hogue CC & EN. August 30, 2004, p. 17. [15] DeYoung JP, Romack TJ, DeSimone JM. Polym Prepr (Am Chem Soc Div Polym Chem) 1997;38:424. [16] Fiering AE. In: Banks RE, Smart BE, Tatlow JC, editors. Organouorine chemistry: principles and commercial applications. New York: Plenum Press; 1994. [17] Van Bramer DJ, Schiett MB, Yokozeki A. United States Patent 5,345,013, 1994. [18] Romack TJ, DeSimone JM. Macromolecules 1995;28:8429. [19] Bunyard WC, Kadla JF, DeSimone JM. J Am Chem Soc 2001;123:7199. [20] Wood CD, Michel U, Rolland JP, DeSimone JM. J Fluorine Chem 2004;125:1671. [21] DeSimone JM, Guan Z, Elsbernd C. Science 1992;257:945. [22] Romack TJ, DeSimone JM. Macromolecules 1995;28:842931.

[23] Parkinson G. Chem Eng 1999;106:17. [24] Michel U, Resnick P, Kipp BE, DeSimone JM. Macromolecules 2003;36:710713. [25] Charpentier PA, Kennedy KA, DeSimone JM, Roberts GW. Macromolecules 1999;32:59735. [26] Charpentier PA, DeSimone JM, Roberts GW. Ind Eng Chem Res 2000;39:458896. [27] Saraf MK, Gerrard S, Wojcinski LM, Charpentier PA, DeSimone JM, Roberts GW. Macromolecules 2002;35:797685. [28] Guan Z, Elsbernd CS, DeSimone JM. Polym Prepr (Am Chem Soc, Div Polym Chem) 1992;34:329. [29] Guan Z, Combes JR, Elsbernd CS, DeSimone JM. Polym Prepr (Am Chem Soc Div Polym Chem) 1993;34:446. [30] Ehrlich P. Chemtracts: Org Chem 1993;6:92. [31] DeSimone JM. US Pat. 5,496,901, 1996. [32] DeSimone JM. US Pat. 5,688,879, 1997. [33] DeSimone JM. US Pat. 5,739,223, 1998. [34] McClain JB, Londono JD, Combes JR, Romack TJ, Canelas DA, Betts DE, et al. J Am Chem Soc 1996;118:917. [35] Rolland JP, Van Dam RM, Schorzman DA, Quake SR, DeSimone JM. J Am Chem Soc 2004;126:2322. [36] Rolland JP, Hagberg EC, Denison GM, Carter KR, DeSimone JM. Angew Chem Int Ed 2004;43:5796. [37] Sianesi D, Marchionni G, DePasquale RJ. In: Banks RE, Smart BE, Tatlow JC, editors. Organouorine chemistry: principles and commercial applications. New York: Plenum Press; 1994. p. 431. [38] Bunyard WC, Romack TJ, DeSimone JM. Macromolecules 1999;32:8224. [39] Cooper AI, Holmes AB. Adv Mater 1999;11:1270. [40] Cooper AI, Wood CD, Holmes AB. Ind Eng Chem Res 2000;39:4741. chet JMJ. Science 1996;273:205. [41] Svec F, Fre chet JMJ. Ind Eng Chem Res 1998;38:34. [42] Svec F, Fre chet JMJ. J Polym Sci A, Polym Chem [43] Yu C, Xu MC, Svec F, Fre 2002;40:755. chet JMJ. Biotechnol Bioeng 1996;49:355. [44] Petro M, Svec F, Fre [45] Hebb AK, Senoo K, Bhat R, Cooper AI. Chem Mater 2003;15:2061. [46] Sherrington DC. Chem Commun 1998;21:2275. [47] Wood CD, Cooper AI. Macromolecules 2001;34:5. [48] Imhof A, Pine DJ. Adv Mater 1998;10:697. [49] Schmidt-Winkel P, Lukens WW, Yang PD, Margolese DI, Letlow JS, Ying JY, et al. Chem Mater 2000;12:686. [50] Manoharan VN, Imhof A, Thorne JD, Pine DJ. Adv Mater 2001;13:447. [51] Zhang H, Hardy GC, Rosseinsky MJ, Cooper AI. Adv Mater 2003;15:78. [52] Cameron NR, Sherrington DC. Adv Polym Sci 1996;126:163. [53] Busby W, Cameron NR, Jahoda CAB. Biomacromolecules 2001;2:154. [54] Zhang H, Cooper AI. Chem Mater 2002;14:4017. [55] Lee CT, Psathas PA, Johnston KP, deGrazia J, Randolph TW. Langmuir 1999;15:6781. [56] Butler R, Davies CM, Cooper AI. Adv Mater 2001;13:1459. [57] Butler R, Hopkinson I, Cooper AI. J Am Chem Soc 2003;125:14473. [58] Song J, Saiz E, Bertozzi CR. J Am Chem Soc 2003;125:1236. [59] Luo Y, Dalton PD, Shoichet MS. Chem Mater 2001;13:4087. [60] Chiellini F, Bizzari R, Ober CK, Schmaljohann D, Yu TY, Solaro R, et al. Macromol Rapid Commun 2001;22:1284. [61] Shaer KA, Jones TA, Canelas DA, DeSimone JM. Macromolecules 1996;29:2704. [62] Sarbu T, Styranec T, Beckman EJ. Nature 2000;405:165. [63] Sarbu T, Styranec T, Beckman EJ. Ind Eng Chem Res 2000;39:4678. [64] Bray CL, Tan B, Wood CD, Cooper AI. J Mater Chem 2005;15:456.