Polymer Synthesis by In Vitro Enzyme Catalysishomepages.rpi.edu/~grossr/_doc/_doc/publication/Chem....

Polymer Synthesis by In Vitro Enzyme CatalysisRichard A. Gross,* Ajay Kumar, and Bhanu Kalra

NSF IUCRC for Biocatalysis and BioProcessing of Macromolecules, Department of Chemistry and Chemical Engineering,Polytechnic University, Six Metrotech Center, Brooklyn, New York 11201

Received January 12, 2001

ContentsI. Introduction 2097II. Lipase-Catalyzed Enantioenriched Monomer,

Macromer, and Polymer Preparations2099

A. Lactones 2099B. Vinyl Monomers 2100C. Macromers 2102

III. Enzyme-Catalyzed Condensation Polymerizations 2102A. Self-Condensation Reactions 2102B. AA−BB Type Enzymatic

Polytransesterifications2103

C. Use of Activated Enol Esters in CondensationPolymerizations

2105

D. Combination of Condensation andRing-Opening Polymerizations

2107

IV. Enzyme-Catalyzed Ring-Opening Polymerization 2107A. Ring-Opening Polymerization of Lactones 2107B. Lipase-Catalyzed Poly(carbonate) Synthesis 2109C. Polyester Synthesis by Lipase-Catalyzed

Ring-Opening Copolymerizations2110

D. Poly(ester-co-carbonate) Synthesis byLipase-Catalyzed Ring-OpeningCopolymerizations

2110

V. Effect of Reaction Parameters on PolyesterSynthesis

2111

A. Solvent 2111B. Enzyme Concentration 2112C. Reaction Water Content 2112D. Reaction Temperature 2113

VI. Activation of Enzymes for Polymer Synthesis byImmobilization or Solubilization

2114

VII. Kinetic and Mechanistic Investigations ofLipase-Catalyzed Ring-Opening Polymerization ofLactones

2116

VIII. Transesterification of Polyester Substrates 2116IX. Modification of Carbohydrates 2117X. Peroxidase-Catalyzed Polymer Synthesis 2118XI. Challenges 2121XII. Additional Options: Enzyme Evolution 2122XIII. Acknowledgment 2122XIV. Abbreviations 2122XV. References 2122

I. IntroductionIn nature, living organisms are constantly produc-

ing different macromolecules for their metabolic

needs. These macromolecules, such as polysaccha-rides, polynucleotides, proteins, or polyesters, areessential to organism survival. Their synthesis gen-erally involves in vivo enzyme-catalyzed chain-growth polymerization reactions of activated mono-mers, which are generally formed within the cells bycomplex metabolic processes. Because of their diver-sity and renewability, microbial polymers such aspolysaccharides, bacterial polyhydroxyalkanoates (1)and polyanions such as poly(γ-glutamic acid) (2) havereceived increasing attention as candidates for in-dustrial applications. In many cases, microorganismscarry out polymer syntheses that are impractical orimpossible to accomplish with conventional chemis-try. Thus, microbial catalysts enable the productionof materials that might otherwise be unavailable.Microbial polymers are synthesized from renewablelow-cost feedstocks, and the polymerizations operateunder mild process conditions with minimal envi-ronmental impact. In addition, microbial polymersprovide products that, when disposed, can degradeto nontoxic products.

Certainly, there are many technical challenges tothe development of commercial processes that involvemicrobial polymerizations. The identification of amicroorganism with the desired synthetic capabilitiescan require an extensive search. Microbes are ofteneffective only in a restricted physiological set ofconditions (temperature, pH, in water) that are notcompatible with reactant stability or reaction ther-modynamics. Living catalysts are susceptible tochemical toxins and strain instability. Compared tochemical catalysis, microbial syntheses often occurwith relatively slow reaction kinetics. In addition, theisolation and purification of polymer products by low-cost processes often proves challenging. Furthermore,to better control in vivo processes where reactionsoccur within complex assemblies of natural sub-stances, fundamental studies are desperately neededto better define the structure of participating cata-

* Corresponding author e-mail: [email protected]; telephone: (718)-260-3024; fax: (718)875-9646.

2097Chem. Rev. 2001, 101, 2097−2124

10.1021/cr0002590 CCC: $36.00 © 2001 American Chemical SocietyPublished on Web 06/09/2001

lysts, their organization, pathway regulation, andmechanisms at a molecular level.

A separate activity that has taken on increasedimportance in recent years is the use of isolatedenzymes. These enzymes, harvested from livingorganisms, are finding increasing use as catalysts forpolymer synthesis in vitro. The hallmark of enzymecatalysis is its superior catalytic power and highselectivity under mild reaction conditions. In addi-tion, enzyme catalysis in polymer synthesis offersseveral advantages relative to chemical preparativer-outes. Enzymes, derived from renewable resources,(i) offer promising substrate conversion efficiency dueto their high selectivity for a given organic transfor-mation, (ii) have high enantio- and regioselectivity,(iii) offer catalyst recyclability, (iv) have the abilityto be used in bulk reaction media avoiding organicsolvents, and (v) can circumvent the use of potentiallytoxic catalysts. Polymers with well-defined structurescan be prepared by enzyme-catalyzed process. Incontrast, attempts to attain similar levels of polymerstructural control by conventional methods mayrequire undesirable protection-deprotection steps.Examples of well-defined polymeric structures thathave been prepared by using in vitro enzyme cataly-sis are as follows: (i) enantioenriched polyesters

synthesized by lipase-catalyzed stereoelective ring-opening polymerization,1 (ii) cellulase-catalyzed2 cel-lulose synthesis (3), (iii) horseradish peroxidase(HRP)-catalyzed polymerization of aromatic sub-strates (e.g., phenols, aniline, and their derivatives).3Recent advances in nonaqueous enzymology hassignificantly expanded the potential conditions thatthese reactions can be performed. Examples of alter-native reaction environments include various organic

Richard A Gross received his B.S. in Chemistry from SUNY at Albany in1979. In 1986, he received his Ph.D. in Organic-Polymer Chemistry underthe supervision of Professor Mark M. Green at the Polytechnic University(Brooklyn, NY). His doctoral work focused on macromolecular stereo-chemistry, and his dissertation title was “The Effect of Pendant GroupStructure on Polymer Conformation in Polyisocyanates and Polyisonitriles”.From 1986 to 1988, he worked as a postdoctoral fellow with Prof. RobertW. Lenz at the University of Massachusetts Amherst, specializing in thesynthesis of microbial polyhydroxyalkanoates (PHAs) by microbial andchemical routes as well as studies of structure−property relationships. In1988, he joined the University of Massachusetts Lowell as an assistantprofessor of chemistry. In 1990, he received a NSF Presidential YoungInvestigator Award. Dr. Gross was promoted to full professor in 1995. Dr.Gross has served as the President of the U.S. Society on BiodegradablePolymers (1999), Founder and Co-Editor of the Journal of EnvironmentalPolymer Degradation (1993−1998), and Co-Director of the NSF I/UCRCCenter Biodegradable Polymer Research Center (1993−1998). In 1998he joined the Polytechnic University (Brooklyn, NY) as the Herman F.Mark Chaired Professor. He is the Director of the NSF I/UCRC Centerfor Biocatalysis and Bioprocessing of Macromolecules (NSF−BBM). Hiscurrent interests include integrated chemical and biocatalytic strategiesfor the synthesis of monomers, polymers, and their selective modification;design/synthesis of bioresorbable polymers and study of their interactionswith biological systems; synthesis and selective modification of microbialglycolipids and studies of their biological properties; and polymertherapeutic systems for drug delivery.

Ajay Kumar received his B.Sc. and M.Sc. in Chemistry at the Universityof Delhi, India. He received his Ph.D. in 1997 under the supervision ofProfessor V. S. Parmar at the University of Delhi, India. His doctoral workwas on biotransformations of esters and amides of polyacetoxy aromaticcarboxylic acids and synthetic studies on nitrogen containing hetrocycliccompounds. He joined Professor Richard A. Gross in 1997 as apostdoctoral research associate and worked on enzyme-assisted ring-opening polymerizations. He is now a Senior Research Fellow in theNational Science Foundation I/UCRC Center for Biocatalysis and Bio-processing of Macromolecules at Polytechnic University, Brooklyn, NY.His current interest includes biotransformations on drug molecules, efficientand environmentally friendly synthetic routes to monomers, and polymerspreparations (biodegradable polymers of importance in medicine that aredifficult to synthesize using pure chemical means).

Bhanu Kalra received her B.Sc. and M.Sc. in Chemistry at the Universityof Delhi, India. She received her Ph.D. in 1997 under the supervision ofProfessor S. M. S. Chauhan at the University of Delhi, India. Her doctoralwork was on biomimetic oxidation reactions of selected phenolics. Shejoined Professor Richard A. Gross in 1998 as a postdoctoral researchassociate and worked on enzyme-mediated vinyl polymerizations, chemicalmodification of polysuccinimide with different phenolic amines to preparepolyaspartamides copolymers, and oxidative cross-linking of phenol-decorated water-soluble polymers with oxidases to form gels and to studythe gel properties. She is now a Senior Research Fellow in the NationalScience Foundation I/UCRC Center for Biocatalysis and Bioprocessingof Macromolecules at the Polytechnic University, Brooklyn, NY. Her currentinterest includes enzyme-catalyzed synthesis and polymerization of vinylmonomers from sugars, nucleoside bases, and other functional moieties.

2098 Chemical Reviews, 2001, Vol. 101, No. 7 Gross et al.

solvents, biphasic organic solvent-aqueous mixtures,reversed micelle systems, and supercritical fluids.Nontraditional reaction media have allowed newoptions for the control of polymer molecular weightas well as the morphology and architecture of poly-mer products. From a processing viewpoint, diversityin reaction environments facilitates the engineeringof systems to facilitate downstream product separa-tion and enzyme reuse.

In addition to providing important elements ofstructural control, in vitro enzyme-catalyzed polymersynthesis has many related environmental benefits.Historically, one of the major missions in polymerscience was to develop synthetic plastics to replacenatural materials such as plant fibers and animalproteins. Materials such as wool, cotton, silk, andwood have been targets for these developments sincethey were perceived as fluctuating in both qualityand availability. However, environmental pollutioncaused by the production and the disposal of petro-chemical-derived plastics has raised increasing con-cerns within the chemical industry. Alternativeapproaches using environmentally benign processesto synthesize plastics that are engineered to degrade-on-demand are actively being pursued. Enzymes, inaddition to the aforementioned remarkable features,are part of a sustainable environment. They comefrom natural systems, and when they are degraded,the amino acids of which they are made can bereadily recycled back into natural substances. Theirextended use in aqueous or solvent-less systems hasmade them particularly attractive to replace poten-tially toxic heavy metal catalysts for polymer syn-thesis. Furthermore, the products derived from en-zyme catalysis, whether polyesters, polyphenols,polysaccharides, proteins, or other polymers, are inmost cases biodegradable. Therefore, it is not sur-prising that the magnitude of research activity in thisfield has thrived in recent years.4

II. Lipase-Catalyzed Enantioenriched Monomer,Macromer, and Polymer Preparations

A. Lactones

Lipases in organic media have been used success-fully for lactonization of racemic γ-hydroxy esters (4),ω-hydroxy esters (5), and δ-hydroxy esters (6) intoenantioenriched γ-butyrolactones (7), ω-lactones (8),

and δ-lactones (9), respectively (Schemes 1 and 2).5-7Enzyme catalysis has generated stereochemical regu-larity and enantiopurity along chains. For example,the lipase-catalyzed polymerization of racemic R-meth-yl-â-propiolactone (MPL, 10) occurred with substan-tial enantioselectivity1 (Scheme 3). The polymeriza-

tion was catalyzed by lipase PS-30 (from Pseudomonascepacia) in different organic solvents. With respectto reaction rates and enantioselectivity, toluene wasthe preferred solvent for preparing [S]-enrichedPMPL (11) and [R]-enriched R-methyl-â-propiolac-tone (12). On the basis of the analysis of chainstereosequence distribution by 13C NMR, stereose-lectivity during propagation was said to result fromcatalyst enantiomorphic site control.

Similarly, Kobayashi and co-workers8 have per-formed ring-opening polymerization of (R)- and (S)-3-methyl-4-oxa-6-hexanolides (MOHELs, 13) usinglipase PC as catalyst in bulk at 60 °C (Scheme 4). A

comparison of the initial rate of polyMOHEL (14)formation from the (R) and (S) antipodes showed thatthe (S) enantiomer had an initial rate that was seventimes larger. Lipase PA and PF catalyzed the polym-erization of (S)-MOHEL but not (R)-MOHEL.

Chemical routes have shown variable levels ofsuccess for the preparation of enantioenriched lac-tones and their corresponding polymers. Examplesof chemically prepared optically active polyestersinclude poly(R-phenyl-â-propiolactone) (15),9 poly(R-ethyl-R-phenyl-â-propiolactone) (16),10,11 poly(R-meth-

Scheme 1. Lipase-Catalyzed StereoelectiveLactonization of γ-Hyroxy Esters

Scheme 2. Lipase-Catalyzed Lactonization ofω-Hyroxy Esters

Scheme 3. Lipase-Catalyzed StereoelectiveRing-Opening Polymerization ofr-Methyl-â-propiolactone

Scheme 4. Lipase-Catalyzed Ring-OpeningPolymerization of MOHEL

Polymer Synthesis by Enzyme Catalysis Chemical Reviews, 2001, Vol. 101, No. 7 2099

yl-R-ethyl-â-propiolactone) (17),12 and poly(lacticacid).13,14 Often, the synthetic methods used to pre-pare enantioenriched monomers either are tedious(requiring multistep reactions) or do not provide themonomer in sufficient enantiopurity. Attempts atcarrying out stereoelective polymerizations of racemiclactones using enantiomerically pure polymerizationcatalysts have often failed to produce useful results.For example, an organometallic product from Zn-(C2H5)2 and [R]-(-)-3,3-dimethyl-1,2-butanediol wasevaluated for its ability to catalyze the stereoelectivepolymerization of [R,S]-â-methyl-â-propiolactone (10).15The results of this work showed only a small enan-tiomeric enrichment in the final polymer. Recently,Coates and co-workers13 reported stereoelective co-polymerizations of lactides (LL/DD monomers) usinga zinc alkoxide complex. It acts as a single-site livinginitiator for the polymerization of rac-lactide thatgives heterotactic PLA.13 Heterotactic macromol-ecules are a rare type of polymer that have alternat-ing pairs of stereogenic centers in the main chain (i.e.,predominantly mr/rm diads are present).

B. Vinyl Monomers

Work has been carried out to prepare vinyl mono-mers from carbohydrates by selective enzyme-cata-lyzed transformations. Motivations for such workhave been to (i) selectively place vinyl functionalgroups at one site of multiple possible positions, thuscircumventing tedious protection-deprotection steps;(ii) synthesize polymers from renewable carbohydratefeedstocks; and (iii) develop a new family of functionalwater-soluble monomers and polymers. Since Kli-banov and co-workers16,17 first demonstrated selectivemonosaccharide acylation catalyzed by lipases, nu-merous reports have appeared in the literature onenzyme-catalyzed selective carbohydrate acylationsand deacylations.18,19 However, the intrinsic polarityof carbohydrate compounds causes them to be solublein only polar solvents (e.g., dimethyl sulfoxide (DMSO),dimethylformamide (DMF), and pyridine).20 In thesemedia, many enzymes lose their activity due to thestripping away of critical or essential water. Conse-quently, in polar solvents only a small number ofenzymes retain some activity that is manyfold de-creased relative to their use in nonpolar solvents.18,19Alternative strategies to improve the miscibilitybetween carbohydrates and hydrophobic organicsubstances have been pursued. Selected examples areas follows: (i) the use of solvent mixtures to achieveboth satisfactory lipase activities and good solubilityof the substrates,21 (ii) the use of organoboronic acidsto solubilize carbohydrates by complexation,22 (iii) thepreadsorption of carbohydrates on silica gel,23 (iv) theuse of tert-butyl alcohol that functions as a bulkypolar solvent,24 and (v) the prior modification ofsugars by alkylation25 and acetalization.26 For ex-ample, Adelhorst and co-workers26 performed theregioselective, solvent-free esterification of simple1-O-alkylglycosides (18-22) using a slight molarexcess of melted fatty acids (Scheme 5). A range of1-O-alkyl-6-O-acylglucopyranosides (18-22b) wereprepared in up to 90% yield, and the process has

recently undergone pilot-scale trials by Novo Nord-isk.27 As an extension of the work by Adelhorst andco-workers,26 our laboratory synthesized macromersby lipase-catalyzed ring-opening polymerizations oflactones from the hydroxyl moieties28,29 of carbohy-drates.30 In summary, ethylglucopyranoside (EGP)(19) was used as the multifunctional initiator, and�-caprolactone (23)/TMC (24) ring-opening polymer-ization was catalyzed by lipases (19 + 23 ) 25, 19 +24 ) 26, Scheme 6).30 Selective ring-opening from the

6-hydroxyl position was achieved. This work is dis-cussed in greater detail below in section IX.

Shibatani et al.31 demonstrated the synthesis ofvinyl sugar esters (e.g., 6-O-vinyladipoylglucose) (28)using glucose and activated fatty acids (27) in DMF(Scheme 7). Of the 34 different lipases and proteases

screened, an alkaline protease from Streptomyces sp.had the highest transesterification activity giving56% yield at 35 °C in 7 days. This protease was moreeffective than subtillisin from Bacillus subtilis.22,32Tokiwa and co-workers33 investigated the protease-catalyzed synthesis of vinyl pentose esters such asthat from arabinose and divinyl adipate. The trans-esterification catalyzed by the alkaline protease from

Scheme 5. Lipase-Catalyzed RegioselectiveAcylation of Glucose

Scheme 6. EGP-Initiated Polymerization of TMCand E-CL

Scheme 7. Protease-Catalyzed RegioselectiveSynthesis of 6-O-Vinyladipoylglucose


Streptomyces sp. of 0.5 M divinyl adipate (27) with0.25 M arabinose (29) in DMF (30 °C, 7 days) gave5-O-vinyladipoyl-D-arabinofuranose (30) in 50% yield(Scheme 8).33 The low productivity and the reduced

operational stability of enzymes in these polar aproticsolvents (e.g., DMF and DMSO) are some of thefactors that limit their use for these methods. Tokiwaand co-workers34 also studied the transesterificationreaction of 0.25 M thymidine (31) with 1 M divinyladipate (27) in DMF using the alkaline protease (5mg mL-1) from Streptomyces sp. (20 units mg-1min-1). After 7 days at 30 °C, this reaction gave 5′-O-vinyladipoylthymidine (32) in 77% yield (Scheme9).34

Our laboratory developed a chemoenzymatic routeto a vinyl-decorated glycolipid monomer (6-O-acryl-sophorolactone, Scheme 10).35 Sophorolipids (SLs)

were prepared in our laboratory by fermentation ofthe yeast Candida bombicola on a glucose/oleic acidmixture.36 The microbial synthesis of SLs results ina mixture of at least eight different compounds. SLsexist mainly in the lactonic (33) and acidic forms (34)and are acetylated to various extents at the 6′ and6′′ positions. Treatment of SL with sodium methoxideresulted in complete deacetylation and formation ofthe SL methyl ester (35). Novozyme-435 (immobilizedcatalyst containing the C. antarctica lipase B) wasused to lactonize the SL-methyl ester forming amacrolactone (36). This same catalyst was then usedto perform a highly selective vinyl esterification onthe macrolactone (37). The resulting vinyl-function-alized glycolipid (37) was successfully homopolymer-ized and subsequently copolymerized with acrylicacid (38) and acrylamide (39) using traditional freeradical initiation.35 We are not aware of an alterna-tive chemical route that could be used to prepare

this complex SL monomer.

Patil et al.37 used Proleather (alkaline proteasefrom Bacillus sp.) as a selective catalyst in pyridinefor the reaction of vinyl acrylate (41) with one of threesucrose (40) primary hydroxyl groups. The resultingmono acryl derivative (42) was polymerized by con-ventional free radical methods to give a novel poly-(sucrose acrylate) (42b, Scheme 11). Other examples

of chemoenzymatic methods that were used to pre-pare chiral monomers and functional polymers follow.Ritter and co-workers38,39 used lipases from C. cylin-dracea and C. antarctica to catalyze the esterificationof 11-methacryloylaminoundecanoic (43) acid with12-hydroxyundecanoic acid (44), isobutyl alcohol (45),menthol (46), cyclohexanol (47), cholesterol (48), andtestosterone (49). They also used C. antartica lipaseto catalyze the regioselective esterifications of methylR-D-glucopyranoside (18) and 3-O-methyl R-D-glu-copyranose (50) with 11-methacryloylaminounde-canoic acid (43).40 Subsequently, these monomerswere polymerized by conventional free radical tech-niques (Scheme 12).

Scheme 8. Protease-Catalyzed RegioselectiveSynthesis of 5-O-Vinyladipoyl-D-arabinofuranose

Scheme 9. Protease-Catalyzed RegioselectiveSynthesis of 5′-O-Vinyladipoylthymidine

Scheme 10. Chemoenzymatic Approach towardSynthesis of Poly(6-O-acryl sophorolactone) andCopolymers with Acrylic Acid or Acryamide

Scheme 11. Chemoenzymatic Synthesis of aSucrose Polyester


C. MacromersMacromers are functional oligomers or low molec-

ular weight polymers that are polymerizable. Koba-yashi and co-workers41 prepared polyesters withmethacryloyl end groups (53, 55). This was ac-complished by the polymerization of DDL (51) in thepresence of ethylene glycol methacrylate (52) andvinyl methacrylate (54) (Scheme 13). The acryl-

enzyme intermediate, formed by reaction of the lipaseand the vinyl ester, reacted to terminate propagatingchains.

Cordova and co-workers42 prepared end-function-alized polycaprolactone (PCL) macromers using C.antarctica lipase B. �-Caprolactone (CL) (23) waspolymerized in the presence potential chain iniatorsthat included 9-decenol (56), 2-(3-hydroxyphenyl)-ethanol (57), 2-(4-hydroxyphenyl)ethanol (58), andcinnamyl alcohol (59) (Scheme 14). Alternatively, CLpolymerizations were conducted in the presence ofpotential chain-terminating carboxylic acids thatincluded n-decanoic acid (60), octadecanoic acid (61),oleic acid (62), linoleic acid (63), 2-(3-hydroxyphenyl)-acetic acid (64), 2-(4-hydroxyphenyl)acetic acid (65),and 3-(4-hydroxyphenyl)propanoic) acid (66) (Scheme15). In an effort to simultaneously control both thehydroxyl and the carboxyl end groups of macromers,esters [e.g., 9-decenyl oleate (67), 2-(4-hydroxyphe-nyl)ethyl acrylate (68), 9-decenyl-2-(4-hydroxyphe-nyl)acetate) (69), methyl linolenate (70)] or a se-quence of an alcohol and an acid were added atvarious times during the course of C. antarcticalipase B-catalyzed CL (23) polymerizations (Scheme

16). Similarly, a telechelic polymer (72) bearingcarboxylic acid groups at both chain ends was formedby carrying out the lipase-catalyzed polymerizationof DDL in the presence of divinyl sebacate (71)(Scheme 17).41 In this case, divinyl sebacate acts asa coupling agent so that the hydroxyl propagatingends of poly(DDL) chains react with the lipase-activated acyl intermediate of divinyl sebacate.

III. Enzyme-Catalyzed CondensationPolymerizations

A. Self-Condensation ReactionsPolymerizations may be conducted with A-B type

monomers, where the groups A and B can react with

Scheme 12. Enzymatic Synthesis of aPolymerizable Oligoester

Scheme 13. Enzymatic Synthesis of a FunctionalMacromer

Scheme 14. Enzymatic Synthesis ofEnd-Functionalized Polycaprolactone MonomersUsing Alcohol as Initiating Group

Scheme 15. Enzymatic Synthesis ofEnd-Functionalized Polycaprolactone MonomersUsing Carboxylic Acid as Chain-TerminatingGroup

Scheme 16. Enzymatic Synthesis ofEnd-Functionalized Polycaprolactone MonomersUsing Ester Functionality in the Reaction

Scheme 17. Enzymatic Synthesis of a FunctionalMacromer Using Dodecanolactone and DivinylSebacate


other B and A groups, respectively. When this occurswith the evolution of small molecule leaving groups,these reactions are referred to as A-B type conden-sation polymerizations. An advantage of A-B typemonomers relative to copolymerizations of A-A/B-Btype monomers is that the latter requires equimolarquantities of A-A and B-B monomers to obtain highmolecular weight polymers.

Gutman and co-workers6 studied PPL-catalyzedA-B type polymerization of â-, δ-, and �-hydroxyes-ters (4-6). Unsubstituted â-, δ-, and �-hydroxyestersundergo exclusively intermolecular transesterifica-tion to afford the corresponding oligomers whereassubstituted δ-methyl δ-hydroxyesters undergo lac-tonization (see section II). Knani and co-workers43were the first to consider the influence of enzymeorigin, solvent, concentration, reaction time, andother parameters on self-condensation reactions. Forthis, they chose methyl �-hydroxyhexanoate as amodel monomer (73). The degree of polymerization(DP) of the polyester formed followed a S-shapedbehavior with solvent log P (-0.5 < log P < 5) withan increase in DP around log P ∼2.5. Decreasingvalues of DP in good solvents for polyesters wereattributed to the rapid removal of product oligomersfrom the enzyme surface, resulting in reduced sub-strate concentration near the enzyme. In a separatestudy by Knani et al.,44 it was found that an increasein the size of hydroxyester lateral substituents frommethyl to ethyl to phenyl gave slower polymeriza-tions but higher enantioselectivity. Also Ritter andco-workers38 reported the formation of oligomers fromcholic acid (74, 75) by self-condensation reactionscatalyzed by the lipase from C. antartica (Scheme 18).

Hagan and Zaidi45,46 studied the C. cylindracealipase-catalyzed condensation polymerization of 10-hydroxydecanoic acid (76) and 11-hydroxydecanoicacid (77). The polymerization of 10-hydroxydecanoicacid, carried out in hexane for 48 h at 55 °C with3-Å molecular sieves, gave a polymer with Mn ) 9.3× 103.45,46 A time course study of 11-hydroxydecanoicacid polymerizations showed that oligomers areformed relatively rapidly and then later condense togenerate higher molecular weight polyesters. By thismethod, within 7 days, they reported the formationof product with molecular weights up to 35 × 103. Acompilation of lipase-catalyzed self-condensation po-lymerizations that have thus far appeared in theliterature is given in Table 1.

B. AA−BB Type EnzymaticPolytransesterifications

Okumara and co-workers47 were the first to at-tempt the lipase-catalyzed synthesis of oligoestersfrom reactions between dicarboxylic acids (BB) anddiols (AA). They showed that “trimer”, “pentamer”,and “heptamer” consisting of AA-BB-AA, AA-BB-AA-BB-AA, and AA-BB-AA-BB-AA-BB-AA,respectively, were formed. Klibanov and co-workers48used the stereoselectivity of lipases to prepare enan-tioenriched oligoesters. The reactions were con-ducted using racemic diester and an achiral diol or,conversely, a racemic diol and an achiral diester asmonomers.

Wallace and Morrow49,50 used halogenated alcoholssuch as 2,2,2-trichloroethyl to activate the acyl donorand thereby improve the polymerization kinetics.They also removed byproducts periodically during thereactions to further shift the equilibrium toward thegrowth of chains instead of chain degradation. Thus,these workers copolymerized bis(2,2,2-trichloroethyl)trans-3,4-epoxyadipate (78) and 1,4-butanediol (79)using porcine pancreatic lipase (PPL) as the catalyst.After 5 days, an enantioenriched polyester (80, 81)with Mw ) 7900 g/mol and an optical purity in excessof 95% was formed (Scheme 19).49

Linko et al.51 studied lipase-catalyzed polyesteri-fication reactions where the degree of activation ofthe diacid was systematically varied. Thus, thesynthesis of poly(1,4-butyl sebacate) (83a-c) fromreactions of 1,4-butanediol (79) with sebacic acid(82a), diethyl sebacate (82b), or bis(2,2,2-trifluoro-ethyl) sebacate (82c) were performed in veratrole ordiphenyl ether using the lipase from Mucor miehei(36.5 wt %). The effects of the removal of water oralcohol, the solvent character, and the substratestructure on the average molar mass of the productswere determined. When vacuum was used to removewater formed during the polymerization, sebacic acid(82a) was directly polymerized with 1,4-butandiol indiphenyl ether to give a product with Mw ) 42 × 103g/mol in 7 days at 37 °C (Scheme 20). The polymer-ization of bis(2,2,2-trifluoroethyl) sebacate (82c) with1,4-butanediol in diphenyl ether gave the corre-sponding polyester with Mw ) 46 400 g/mol in 72 hat 37 °C with periodic removal of trifluoroethanolunder vaccum (5 mmHg followed by 0.15 mmHg).52In addition, Linko et al.52 systematically varied thechain length of the monomeric dicarboxylic acid [C-4(84), C-6 (85), C-8 (61), C-10 (82a), and C-12 (86)]

Scheme 18. Candida antarctica CatalyzedSelf-Condensation Polymerization of Cholic Acid

Scheme 19. Porcine Pancreatic Lipase-CatalyzedSynthesis of Enantioenriched Polyester withEpoxy Groups in the Main Chain


Table 1. Lipase-Catalyzed Condensation Polymerizations

enzyme monomer refs

Alcaligenes sp. divinyl sebacate with glucose 67Aspergillus niger (lipase A) sebacic acid with 1,8-OL 8

1,13-tridecandedoic acid with 1,3-propanediol 52bis(2-chloroethyl)(+ -)2,5-bromoadipate with 1,6-hexanediol 47

Candida antartica lipase B 1,3-propanediol divinyl carbonate with 1,3-propanediol 61(Novozyme-435) 1,6-hexanediol divinyl carbonate with 1,2,4-butanetriol 61

1,4-butanediol divinyl carbonate with glycerol 61divinyl carbonate with 1,2,4-butanetriol 61divinyl carbonate with 1,3-propanediol 61divinyl carbonate with 1,10-decanediol 61divinyl carbonate with 1,12-dodecanediol 61divinyl carbonate with 1,9-nonanediol 61divinyl carbonate with 1,3- benzenedimethanol 61divinyl carbonate with 1,4-benzenedimethanol 61divinyl carbonate with 2,6-pyridinedimethanol 611,3-propanediol divinyl adipate with 1,3-benzenedimethanol 611,3-propanediol divinyl adipate with 1,4-benzenedimethanol 611,3-propanediol divinyl adipate with 2,6-pyridinedimethanol 611,2-benzenedimethanol 4,4-isopropylidenebis[2-(2,6-

dibromophenoxy)ethanol] bisphenol A59

divinyl isophthalate with 1,6-hexanediol divinyl terephthalatedivinyl p-phenylene diacetate

59

divinyl sebacate with p-xylene glycol 591,3-propanediol divinyl dicarbonate with 1,3-propanediol 681,3-propanediol divinyl dicarbonate with glycerol 681,3-propanediol divinyl dicarbonate with 1,2,4-butanetriol 681,3-propanediol divinyl dicarbonate with 1,2,6-trihydroxyhexane 681,4-butanediol divinyl dicarbonate with glycerol 681,4-butanediol divinyl dicarbonate with 1,2,4-butanetriol 681,4-butanediol divinyl dicarbonate with 1,2,6-trihydroxyhexane 681,6-hexanediol divinyl dicarbonate with glycerol 681,6-hexanediol divinyl dicarbonate with 1,2,4-butanetriol 681,6-hexanediol divinyl dicarbonate with 1,2,6-trihydroxyhexane 68divinyl adipate with glycerol 69divinyl adipate with 1,2,4-butanetriol 69divinyl adipate with 1,6-trihydroxyhexane 69adipic acid with 1,4-butanediol 55, 61divinyl adipate with 2,2,3,3-tetrafluoro-1,4-butanediol 62divinyl adipate with 2,2,3,3,4,4-hexafluoro-1,5-pentanediol 62divinyl adipate with 3,3,4,4,5,5,6,6-octafluorooctan-1,8-diol 62divinyl sebacate with glucose 67cholic acid 38

Candida cylinderacea 11-hydroxyundecanoic acid 46(lipase CCL) 10-hydroxyundecanoic acid 45

sebacic acid with 1,8-OL 8Candida rugosa (lipase AYS) bis(2,2,2-trifluoroethyl) sebacate with 1,4-butanediol 51Klebsiella oxytota (lipase K) sebacic acid with 1,8-OL 8Mucor meihei (lipozyme) sebacic acid with 1,4-butanediol 51

diethyl sebacate with 1,4-butanediol 51bis(2,2,2-trifluoroethyl) sebacate with 1,4-butanediol 51bis(2,2,2-trifluoroethyl) sebacate with 1,4- butanediol 51, 52bis(2,2,2-trifluoroethyl) sebacate with 1,2-ethanediol 51bis(2,2,2-trifluoroethyl) sebacate with 1,3-propanediol 51bis(2,2,2-trifluoroethyl) sebacate with 1,5-pentanediol 51bis(2,2,2-trifluoroethyl) sebacate with 1,6-hexanediol 51divinyl adipate with glycerol 69divinyl adipate with 1,2,4-butanetriol 9divinyl adipate with 1,6-trihydroxyhexane 69bis(2,3-butanedionemonooxime) glutarate with 1,6-hexanediol 65dichloroethyl fumarate with

4,4′-isopropylidenebis[2-(2,6-dibromophenoxy)ethanol]70

adipic acid with 1,4-butanediol 71porcine pancreatic lipase (PPL) divinyl sebacate with glucose 67

sebacic acid + 1,8-OL 8bis(2,2,2-trifluoroethyl) sebacate with 1,4-butanediol 58, 71bis(2,2,2-trifluoroethyl) glutarate with 1,4-butanediol 72bis(2,2,2-trichloroethyl) adipate with 1,4-butanediol 73bis(2,2,2-trichloroethyl) trans-3-hexanedioate (racemic mixture) with

1,4-butanediol49

methyl-5-hydroxypentanoate 43methyl-6-hydroxyhexanoate

Pseudomonas aeruginosa(lipase PA)

sebacic acid with 1,8-OL 8

Pseudomonas cepacia(lipase PS, PS-30)

sebacic acid + 1,8-OL 8

Pseudomonas fluoresenes(lipase PF)

sebacic acid with 1,8-OL 8

divinyl adipate with 1,4-butanediol 57bis(2,2,2-trifluoroethyl) sebacate with 1,4-butanediol 51

Rhizopus delemer (lipase RD) divinyl sebacate with glucose 67Bacillus sp. Proleather sucrose with bis(2,2,2-trifluoroethyladipate) 37Pseudomonas sp. lipase (PSL) divinyl sebacate with glucose 67Steptomyces sp. alkaline protease divinyl adipate with arabinose 33

divinyl sebacate with glucose 67


and diol [C-2 (87), C-3 (88), C-4 (79), C-5 (89), andC-6 (90)] used for the polycondensation polymeriza-tion. Of the lipases and solvents screened, the M.miehei lipase and diphenyl ether, respectively, werefound to be preferred. It was also observed that themolecular weight of the polymer increased with anincrease in the substrate concentration of up to about0.83 M. The reaction of adipic acid (85) with differentdiols showed the following trend with respect topolymer DP: 1,6-hexanediol > 1,4-butanediol > 1,5-decanediol > 1,3-decanediol >1,2-butanediol.52 Simi-larly, the reaction of 1,6-hexanediol with differentacids showed the following trend toward polymerDP: adipic acid > sebacic acid > octanedioic acid >dodecanoic acid > succinic acid.52 M. miehei catalyzedthe condensation polymerization of adipic acid (85)and hexanediol (90) in diphenyl ether at 37 °C for 7days under reduced pressure (0.15 mmHg) to give thecorresponding product with Mw ) 77.4 × 103 and PDI) 4.4.52 Subsequently, using the preferred M. mieheilipase from above, Linko et al.53studied copolymeriza-tions using an aromatic diacid (terephthalic or isoph-thalic) (91, 92) and an aliphatic diol (1,4-butane- or1,6-hexanediol) (79, 90). Even at temperatures up to70 °C, the polymerization of these diacids built fromaromatic moieties was unsuccessful. However, usingNovozyme-435 as the catalyst, the polymerization ofaromatic diacids was accomplished. For example,while the Novozyme-435-catalyzed reaction of isoph-thalic acid with butanediol yielded oligomers, asimilar reaction between the C-6 diol and isopthalicacid at 70 °C yielded a polymer with Mw ) 55 × 103.

Kobayashi and co-workers54 studied the potentialof carrying out condensation reactions in solventlessor bulk reactions. They reported the preparation ofaliphatic polyesters with Mw > 10 × 103 by reactingsebacic acid (82a) with 1,4-butanediol (79) in solvent-free system. The reactions were conducted underreduced pressure, using the C. antarctica lipase.Binns et al.55 studied copolymerizations of adipic acid(85) and 1,4-butanediol (79) using Novozyme-435.They reported that under solvent-free conditions, themixture was refluxed at 40 °C for 4 h, followed byheating at 60 °C for 10 h under pressure. The

polymerization proceeds by a step growth mechanismto give a homogeneous mixture. The GPC of reactionmixture after 4 and 14 h showed very differentproduct distribution, the former showing a discretearray of predominantly hydroxy-terminated oligo-mers and latter showing the polyesters with weightaverage molecular weight of 2227 and polydispersityof 1.5. A compilation of lipase-catalyzed AA-BB typecondensation polymerizations that have thus farappeared in the literature is given in Table 1.

C. Use of Activated Enol Esters in CondensationPolymerizations

Much of the work carried out on lipase-catalyzedcondensation polymerizations has focused on the useof activated diacids such as enol esters. Enol or vinylesters both accelerate the rate of acyl transfer andshift the reaction equilibrium toward polymer sincethe byproduct rapidly tautomarizes to a ketone oraldehyde. Acyl transfer using enol esters has beenshown to be about 10 times slower than hydrolysisand about 10-100 times faster than acyl transferusing other activated esters.56 For example, enzy-matic hydrolysis of nonactivated esters such as ethylesters reacts at rates about 10-3-10-4 times slowerthan the corresponding enol esters. Uyama et al.57used the Pseudomonas fluorescens lipase to investi-gate polymerizations of divinyl adipate (27) withdifferent diols [ethylene glycol (87), 1,4-butanediol(79), 1,6-hexanediol (90), 1,10-decanediol (93)]. Thesepolymerizations were performed in isopropyl etherfor 48 h at 45 °C. The polymerization of divinyladipate and 1,4-butanediol catalyzed by P. fluorescensin isopropyl ether at 45 °C for 48 h produced acorresponding polyester in 50% yield with weightaverage molecular weight of 6700 and polydispersityof 1.9. The P. fluorescens catalyzed polymerizationof divinyl adipate with ethylene glycol, 1,6-hexane-diol, or 1,10-decanediol as glycols resulted in poly-esters of lower molecular weights (2000, 5900, and2700, respectively) than the polyester obtained fromthe polymerization of divinyladipate and butanediol.In an effort to more efficiently conduct lipase-catalyzed polymerizations of divinyl isophthalate(94), terephthalate (95), p-phenylene diacetate (96),and sebacate (97) diesters, Uyama et al.58 expandedon the work above. The importance of lipase origin(C. antartica, C. cylinderacea, M. meihei, P. cepacia,P. fluorecens, and porcine pancreas), temperature(45-75 °C), solvent (e.g., heptane, acetonitrile, cy-clohexane, isooctane, THF, and toluene), and effectsof chain length in R,ω-alkylene glycols (79, 87, 90,93, 98) (Scheme 21) were evaluated. Of the lipasesscreened, lipase CA gave polyesters having the high-est molecular weights. Also, nonpolar solvents suchas heptane and cyclohexane were preferred. Further-more, the maximum yields and product molecularweights were obtained at 60 °C. For example, lipaseCA catalyzed polymerization of divinyl isophthalate(94) and 1,6-hexanediol (90) in heptane at 60 °Cresulted in polyester formation in 74% yield with Mnand PDI of 5500 and 1.6, respectively, in 48 h.

Russell and co-workers59 studied the Novozyme-435-catalyzed bulk polymerization of divinyl adipate

Scheme 20. Lipase-Catalyzed CondensationPolymerization of Sebacic Acid Ester withButanediol


(27) and 1,4-butanediol (79). The preferred examplewas a 72-h polymerization at 50 °C that gave thecorresponding polyester with Mw ) 23.2 ×103 g/mol.These workers found that the product molecularweight was decreased when the reaction was con-ducted without taking proper precautions to excludereactions of water with reactive divinyl ester groups.They also found excellent agreement between theirexperimental data and that predicted by a math-ematical model.60 Such agreement allowed Russelland co-workers61 to conclude that the polymerizationsoccur by a step-condensation mechanism where therate of polymerization is a function of substratemolecular weight.

Russell and co-workers61 also studied Novozyme-435-catalyzed A-A/B-B type condensation polymer-izations to prepare aromatic polyesters and polycar-bonates. Polymerizations between divinyl esters ordicarbonates with aromatic diols, conducted for 24 hin bulk catalyzed by Novozyme-435 (1/10 wt % ofmonmers) at preferably 70 °C, gave low molecularweight polycarbonates and polyesters, respectively.The aromatic diols included in their work were asfollows: 1,2-benzenedimethanol, 1,3-benzenedimeth-anol, 1,4-benzenedimethanol, 2,6-pyridinedimethanolor 4,4-isopropylidenebis(2-(2,6-dibromophenoxy)etha-nol), and bisphenol A. The Mw of these polycarbonatesand polyesters did not exceed 5.2 × 103 and 3.5 ×103, respectively. When various isomers of benzene-dimethanol (98-100) were used, Novozyme-435 ex-hibited regioselectivity. Thus, p-benzenedimethanol(98) reacted to a greater extent than the correspond-ing meta or ortho isomers (99 and 100, respectively).The regioselectivity of lipases can be exploited in suchwork so that selected isomers from complex mixturesare preferentially polymerized.

Russell and co-workers62 have also studied lipase-catalyzed polymerizations of activated diesters andfluorinated diols. The effects of reaction time, con-tinuous enzyme addition, enzyme concentration, anddiol chain length were studied to determine thefactors that would limit chain growth. Potentiallimiting factors considered were enzyme inactivation,enzyme specificity, reaction thermodynamics, hy-drolysis of activated esters, and polymer precipita-tion. As was seen many times above, Novozyme-435was found to be the effective catalyst. Polymermolecular weight steadily increased and then leveled

off at 50 °C after 30 h at Mw ∼ 1773. Enzymespecificity toward shorter chain fluorinated diolsappeared to be a prominent factor that limited chaingrowth. An increase in the product molecular weightresulted when the fluorinated diol contained an ad-ditional CH2 spacer between CF2 and hydroxyl groups.For example, no polymer was produced for the reac-tions of divinyl adipate (27) with 1H,1H,9H,9H-per-fluoro-1,9-nonanediol [HOCH2(CF2)7CH2OH], 1H,1H,-12H,12H-perfluoro-1,12-dodecanediol [HOCH2(CF2)10-CH2OH], fluorinated oligomers of Mw ∼ 727 fromcondensation of 2,2,3,3,4,4-hexafluoro-1,5-pentanediol[HOCH2(CF2)3CH2OH] with divinyl adipate (27), andfluorinated polyesters with Mw ∼ 1095 were producedfrom reaction of divinyl adipate (27) with 2,2,3,3-tetrafluoro-1,4-butanediol [HOCH2(CF2)2CH2OH]. Re-actions between divinyl adipate and 3,3,4,4,5,5,6,6-octafluorooctan-1,8-diol [HOCH2CH2(CF2)4CH2CH2OH]showed that an additional methylene spacer betweenthe fluorine atoms and the hydroxyl groups has asignificant effect on the polymerization in THF andin solvent-free conditions resulting in polyesters ofMw ) 3245 and 8094, respectively.62

Other researchers have exploited enzyme regiose-lectivity to prepare sugar-based building blocks forcondensation63 and vinyl chain polymerizations.64 Forexample, lipases including C. cylindracea, PPL, andP. fluorescens and proteases including Proleather andprotease N were screened as catalysts for the bu-tyrylation of sucrose (40).64 On the basis of this modelstudy, Proleather in pyridine at 45 °C was selectedand used to synthesize polyesters (101) from sucroseand bis(2,2,2-trifluoroethyl) sebacate (82c). By em-ploying this strategy, a sucrose-based polyester wasobtained in 20% yield with DP 11 after 5 days(Scheme 22).

Similar to enol esters, oximes can also be used asirreversible acyl transfer agents for lipase catalysis.Thus, instead of a dienol ester, Athavale et al.65polymerized diols with bis(2,3-butanedione mon-oxime) alkanedioate using lipozyme IM-20. The re-sults obtained by activation with enol esters, andtheir corresponding oximes were comparable. Noattempts were made to analyze the end group of thepolyester.

A compilation of lipase-catalyzed condensationpolymerizations that have thus far appeared in theliterature where activated diesters were used is givenin Table 1.

Scheme 21. Lipase-Catalyzed CondensationPolymerization of Divinyl Sebacate with Diols ofVarying Length

Scheme 22. Synthesis of a Sucrose PolyesterCatalyzed by Proleather in Anhydrous Pyridine


D. Combination of Condensation andRing-Opening Polymerizations

An example of the versatility of lipase reactions istheir ability to concurrently catalyze condensationand ring-opening chain-type polymerizations. Jed-linski et al.66 reported that PPL catalyzes 12-hy-droxydodecanoic acid (103)/â-butyrolactone (102) co-polymerizations at 45 °C in organic media (104,Scheme 23). Using the preferred solvent (toluene),

after 72 h, the copolymer was formed in 70% yieldwith Mn of 1 800 and PD 1.1. The electrosprayionization mass spectrum (ESI-MS) of the copolymershowed that the chain segments formed containedvarious compositions of 3-hydroxybutanoate and 12-hydroxydodecanoate units. Kobayashi and co-work-ers8 found that the lipase from P. cepacia (lipase PC)catalyzed the concurrent copolymerization of mac-rolides, divinyl esters, and glycols (Scheme 24). The

copolymerization of PDL (105), divinyl sebacate (97),and 1,4-butanediol (79) in isopropyl ether at 60 °Cfor 72 h gave a copolymer (106) in 80% yield withMw ) 6.5 × 103 g/mol. The authors did not provideinformation on the repeat unit sequence distributionof the products.

A compilation of examples and literature referencesthat describe lipase-catalyzed polymerizations withconcurrent condensation and ring-opening step po-lymerization reactions is given in Table 1.

IV. Enzyme-Catalyzed Ring-OpeningPolymerization

A. Ring-Opening Polymerization of LactonesIn contrast to condensation polymerizations, ring-

opening polymerizations of lactones and carbonatesdo not generate a leaving group during the course ofthe reactions.28,29,74,75 This characteristic alleviatesconcerns that the leaving group, if not efficientlyremoved, might limit monomer conversion or polymermolecular weight.76 Research has been conducted onenzyme-catalyzed ring-opening polymerizations of(()-R-methyl-â-propiolactone (10),1 �-caprolactone (�-CL) (23),28,29,41,42,74,75 â-methyl-â-propiolactone (102),77

dodecanolactone (DDL) (51),8,41,78 pentadecanolactone(PDL) (105),40,78-81 â-propiolactone (107),77,82 δ-vale-rolactone (δ-VL) (108),74 8-octanolide (8-OL) (109),83undecanolactone (UDL) (110),8,41,78,79,84 hexadecano-lactone (HDL) (111),41 R-decenyl-â-propiolactone(112),85 R-dodecenyl-â-propiolactone (113),85 benzyl-â-D,L-malonolactonate (114),86 R-methyl-�-caprolac-tone (116),87 R-methyl-δ-valerolactone (115),87 γ-bu-tyrolactone (7e),77,85 1,4-dioxane-2-one (117),88 andothers (Schemes 25 and 26). A compilation of thelactones polymers, the enzymes used, and the corre-sponding citation(s) is given in Table 2.

Common difficulties reported in the publicationslisted in Table 2 are low product molecular weightsand slow polymerization kinetics. In an effort toovercome these difficulties, our laboratory has inves-tigated reaction parameters such as solvent, temper-ature, monomer concentration in solvents, enzymeconcentration, water content, propagation kinetics,and mechanism of C. antarctica lipase (Novozyme-435)-catalyzed �-CL polymerizations.75,89 It was ob-served that different combinations of these param-eters could be used to control the polymerization rate,molecular weight, and polydispersity.

In contrast to 4-, 6-, and 7-membered lactonepolymerizations that can be conducted with goodefficiency using traditional chemical initiators andcatalysts,90 the polymerization of macrolactones bytraditional chemical methods proceeds slowly to givelow molecular weight polymers.91 Enzyme-catalyzedpolymerizations of macrolactones have thus far provedadvantageous relative to chemical preparative routes.Kobayashi and co-workers41,78,79,84 were first to in-vestigate the enzyme-catalyzed polymerization ofω-undecanolide (UDL) (110), ω-dodecanolide (DDL)(51), and ω-pentadecanolide (PDL) (105) (12-, 13-,and 16-membered lactones) (Scheme 25). Screeningof enzymes for the polymerization of UDL (110), DDL(51), and PDL (105) using lipases including thosefrom Aspergillus niger, C. cylindracea (lipase B),Candida rugosa, Rhizopus delmar, Rhizopus javani-cus, P. fluorescens (lipase P, Cosmo Bio.) Pseudomo-nas sp. (lipase PS, Amano), phospholipase, and

Scheme 23. Ring-Opening Copolymerization ofâ-Propiolactone in the Presence of Hydroxy Acid

Scheme 24. Copolymerization Using Combinationof Ring-Opening and Condensation Reactions

Scheme 25. Lactones That Have Been Studied asMonomers for Lipase-Catalyzed Ring-OpeningPolymerizations

Scheme 26. Ring-Opening Polymerization ofSubstituted Lactones


Table 2. Lipase Catalyzed Ring-Opening Polymerizations

enzyme monomer/comonomers refs

Aspergillus niger (lipase A) �-CL 8, 114DDL 8PDL 78, 805-methyl-5-benzyloxycarbonyl-1,3-dioxan-2-one 93

Candida antartica lipase B �-CL 42, 115, 116(Novozyme-435) 8-OL 8, 83

1,4-dioxane-2-one 88â-BL(R,RS) 117TMC 89, 92PDL 805-methyl-5-benzyloxycarbonyl-1,3-dioxan-2-one 93R-Me-γ-VL 87R-Me-�-CL 878-OL, 8-OL-co-�-CL, 8-OL-co-DDL 83

Candida cylinderacea (lipase CCL) 8-OL 83TMC 92, 95R-Me-â-PL 1�-CL 8, 41, 114δ-VL 8, 113PDL 8, 41, 78â-BL (R,RS) 117â-PL 82DDL 8, 41UDL 8, 41

Candida rugosa (lipase AYS) 5-methyl-5-benzyloxycarbonyl-1,3-dioxan-2-one 935-methyl-5-benzyloxycarbonyl-1,3-dioxan-2-one-co-TMC 98PDL 78

Mucor javanicus â-BL(R,RS) 117(lipase M, map-10) PDL 80

TMC 92Mucor meihei (lipozyme) â-BL(R,RS) 117

PDL 78TMC 92, 955-methyl-5-benzyloxycarbonyl-1,3-dioxan-2-one 93

Pencillium rorueforti (lipase PR) PDL 78porcine pancreatic lipase (PPL) DDL 8

â-BL(R,RS) 117�-CL 8, 28, 29, 41, 114γ-VL 8, 114TMC 92, 95, 305-methyl-5-benzyloxycarbonyl-1,3-dioxan-2-one 93â-BL 77R-Me-â-PL 1PDL 78â-BL-co-12-hydroxydodecanoic acid 66lactide-co-TMC 67

Pseudomonas aeruginosa �-CL 8(lipase PA) DDL 8

S-MOHEL 88-OL 83

Pseudomonas cepacia â-BL(R,RS) 117(lipase PS, PS-30) δ-DL 116

δ-DDL 8, 103â-BL 77, 116�-CL 8, 41, 113, 116γ-VL 116γ-CL 116TMC 92, 95PDL 78, 805-methyl-5-benzyloxycarbonyl-1,3-dioxan-2-one 938-OL 8, 83DDL 8MOHELs 8HDL 41PDL-co-�CL 788-OL-co-�CL 838-OL-co-DDL 83R-Me-â-PL 76

Pseudomonas fluoresenes �-CL 8, 41, 66, 78, 114(lipase PF) δ-VL 8, 114

S-MOHEL 8UDL 8, 78DDL 8, 41, 78PDL 8, 41, 78HDL 418-OL 8, 83PDL-co-γ-VL 78PDL-co-�-CL 78PDL-co-DDL 78PDL-co-UDL 78TMC 92


porcine pancreatic lipase was carried out.78,79,84 Quan-titative conversions of UDL to poly(UDL) (110b) wereachieved within 120 h using lipase P and PS.79 Thehighest number average molecular weight (Mn )25 000 g/mol) reported by these workers was for poly-(DDL) (51b) synthesis (75 °C, 120 h) using theimmobilized lipase PS from a Pseudomonas sp. (li-pase PS, Toyobe Co.).84 Our laboratory evaluatedlipase PS-30 immobilized on Celite as a catalyst forbulk PDL polymerization and reported the synthesisof poly(PDL) with Mn ) 62 000 and PDI 1.9.80Recently, instead of in bulk, Novozyme-435-catalyzedpolymerization of PDL was conducted in toluene (1:1wt/vol). These conditions resulted in 31% monomerconversion within only 1 min and the formation ofpoly(PDL) (105b) with the highest Mn (86 000 g/mol)thus far reported for lipase-catalyzed polymeriza-tion.81 Also, lipase catalysis has been explored for 1,4-dioxan-2-one117 ring-opening polymerization (Scheme27).88 An immobilized form of lipase CA, at 60 °C,

exhibited high catalytic activity for this monomer.The weight average molecular weight of poly(1,4-dioxanone,118) was 40 000 after 15 h using 5 wt % ofthe immobilized lipase CA.

B. Lipase-Catalyzed Poly(carbonate) SynthesisLipases have also been used to catalyze the ring-

opening polymerization of cyclic carbonate mono-mers.92-95 Novozyme-435, PPL, PS-30, AK, CCL,MAP, and lipozyme-IM were evaluated by us ascatalysts for the bulk polymerization of trimethylenecarbonate (TMC, 1,3-dioxan-2-one) (24).92 Of thesecatalysts, Novozyme-435 has been most effective. Inone example, Novozyme-435-catalyzed polymeriza-tion of TMC at 70 °C for 120 h gave 97% monomerconversion, poly(TMC) (119) with Mn ) 15 000,without decarboxylation during propagation (Scheme28).92 Similarly Matsumura et al.95 claimed an ex-

traordinarily high molecular weight poly(TMC) (Mw) 156 000) was obtained by using low quantities ofPPL (0.1 wt %) as the catalyst at very high reactiontemperature (100 °C). The thermal polymerizabilityof TMC in the absence of enzyme cast questions asto whether the polymerization proceeded by anenzyme-catalyzed mechanism. In contrast, Kobayashiet al.94 reported the formation of low molecularweight poly(TMC) (Mn ) 800) by PPL (50 wt %)catalyzed polymerization at 75 °C. The lipase-catalyzed polymerization of the disubstituted TMCanalogue 5-methyl-5-benzyloxycarbonyl-1,3-dioxan-2-one (MBC, 120 + 121 ) 122) was also studied(Scheme 29).93 The bulk polymerization, catalyzed by

lipase AK (from Pseudomonas fluorescens) for 72 hat 80 °C, gave 97% monomer conversion and product(123) with Mn ) 6100. The benzyl ester protectinggroups of poly(MBC) were removed with Pd/C inethyl acetate to give the corresponding functionalpolycarbonate with pendant carboxylic acid groups(124, Scheme 29).

Table 2. (Continued)

enzyme monomer/comonomers refs

Rhizopus delemer (lipase RD) �-CL 114PDL 78

Rhizopus japonicus (lipase RJ) �-CL 41, 114γ-VL 114PDL 78

HE PDL 78PD PDL 78PR PDL 78CR PDL 78clonezyme ESL-001 library HEC-modified CL 118Pseudomonas sp. lipase (PSL) �-CL, â-BL, γ-BL, δ-DCL, δ-DDL, PDL 85

ethyl 4-hydroxybutyrate, ethyl-6-hydroxyhexanoate,ethyl-3-hydroxybutyrate, ethyl 5-hydroxy-hexanoate, ethyl 5-hydroxylaurate, ethyl15-hydroxypentadecanoate

�-CL-co-ethyl lactate, �-CL-co-lactide, �-CL-co-γ-BL,�-CL-co-ethyl 4-hydroxybutyrate, �-CL-co-PDL,�-CL-co-ethyl 15-hydroxypentadecanoate,�-CL-co-lactide-co-cyclopentadecanolide

Scheme 28. Ring-Opening Polymerization of1,3-Dioxan-2-one

Scheme 27. Ring-Opening Polymerization of1,4-Dioxan-2-one Scheme 29. Synthesis of Substituted TMC and Its

Ring-Opening Polymerization Using Lipase asCatalyst


C. Polyester Synthesis by Lipase-CatalyzedRing-Opening Copolymerizations

The lipase-catalyzed copolymerization of two ormore monomers presents additional challenges andquestions. For example, knowledge of the relativereactivity of comonomers as well as the kinetic andthermodynamic distribution of repeat units alongchains is particularly important. Since the use oflipase catalysis for lactone polymerization is a rela-tively new area of study, lactone copolymerizationshas thus far received little attention. Namekawa andco-workers96 first studied the enzyme-catalyzed co-polymerization of â-propiolactone (107) with �-CL(23). Uyama and co-workers78 performed copolymer-izations of PDL (105) with DDL (51), UDL (110), VL(108), and CL (23) using the lipase from P. fluore-scens, in bulk for 240 h at 60-75 °C (Scheme 30).

The rates of these reactions were slow and yieldedlow molecular weight (Mn < 6000 g/mol) copolymers.Kobayashi and co-workers83 reported copolymeriza-tion of δ-valerolactone (108) and �-CL (23) using thelipase from P. fluorescens. They also published theuse of a lipase from C. antarctica for the copolymer-ization of 8-OL (109) with CL (23) and DDL (51) for48 h at 60 °C in isooctane.83 In this later report,copolymers were formed with percent conversions of∼80% in 48 h at 60 °C with Mn values up to 9000(0.10 g of lipase/1.0 mmol of monomer). Dong and co-workers85 reported copolymerizations (bulk, 45 °C, 20days) of �-caprolactone with some cyclic and linearmonomers catalyzed by the lipase from a Pseudomo-nas sp. (40 mg of lipase/0.1 mmol of monomer).Among the copolymerizations performed, that of�-caprolactone (23) with pentadecalactone (105) gavethe highest product Mn (8.4 kDa).85 The molecularweights of copolymers of �-caprolactone (23) withlactones (102, 107, 112, 113) were higher than thoseof copolymers prepared from the corresponding linearhydroxyesters (Scheme 31).

Our laboratory investigated copolymerizations of�-caprolactone (23) and pentadecalactone (105) (1:1molar feed ratio, 45 min, 70 °C) catalyzed by No-vozyme-435 in toluene. The product, obtained in 88%isolated yield, had an Mn of 20 000 (Figure 1).81 The

reactivity ratios of the two monomers were calcu-lated, and PDL polymerization was found to be 13×faster than that of �-caprolactone. Despite the dif-ference in reactivities of the two monomers, thecopolymers formed were random. These results wereexplained by rapid Novozyme-435-catalyzed polymer-polymer transacylation or transesterification reac-tions.81

D. Poly(ester-co-carbonate) Synthesis byLipase-Catalyzed Ring-OpeningCopolymerizations

There has been a significant effort to copolymerizeTMC with lactones and other carbonate monomers.Matsumura and co-workers97 performed copolymer-izations of lactide (125) with trimethylene carbonate(24) using porcine pancreatic lipase at 100 °C for 168h. They claimed that random copolymers were formed(126, Scheme 32) with Mw values up to 21 000.

However, since trimethylene carbonate is known tothermally polymerize at 100 °C, the extent of polym-erization that occurs due to activation of monomersat the lipase catalytic triad versus by thermal orother chemical processes is questionable.92 LipaseAK-catalyzed copolymerizations of 1,3-dioxan-2-one(TMC, 24) with 5-methyl-5-benzyloxycarbonyl-1,3-dioxan-2-one (MBC,122) were carried out in bulk at80 °C for 72 h. Although TMC reacted more rapidlythan MBC, the product isolated at 72 h appeared tohave a random repeat unit distribution.98 Recently,using Novozyme-435 in toluene at 70 °C, trimethyl-ene carbonate/PDL copolymerizations (24, 105) wereperformed and gave random copolymers (127, Scheme

Scheme 30. Ring-Opening Copolymerization ofLactone Monomers

Scheme 31. Ring-Opening Copolymerization ofCaprolactone and â-Propiolactone and ItsDerivatives

Figure 1. Conversion profile of monomer during thepolyesterification of PDL at 60 °C.

Scheme 32. Ring-Opening Copolymerization ofLactide and TMC


33).99 Varying the feed ratio of the comonomersallowed regulation of the copolymer composition. Theisolated yield and Mn of poly(PDL-co-43 mol % TMC)formed after 24 h (feed 2:1 PDL:TMC) was 90% and30.9 × 103 g/mol, respectively. Thus far, an alterna-tive chemical route to random poly(PDL-co-TMC) isnot known. For example, PDL/TMC copolymeriza-tions with chemical catalyst such as stannous oc-tanoate, methylaluminoxane, and aluminum triiso-proxide resulted either in homo-polyTMC or blockcopolymers of poly(TMC-co-PDL).99 Chemical cata-lysts have thus far favored TMC over PDL polymer-ization. In contrast, by lipase catalysis, PDL wasmore rapidly polymerized than TMC. Thus, hereinlie important differences in the inherent catalyticproperties of lipases as opposed to chemical catalyststhat can be exploited to give unique copolymers.

Recently, we studied lipase-catalyzed copolymeri-zation of PDL (105) with a sugar carbonate (IPXTC,128) in toluene at 70 °C. 100 Novozyme-435 was foundto be the most effective lipase catalyst based on itsability to form PDL/IPXTC copolymers. For example,by this method, poly(PDL-co-19 mol % IPXTC) wasprepared in 38% isolated yield in 5 days with Mn )4070. The copolymer formed consisted of PDL blockswith random interruptions by IPXTC units or shortsegments.

V. Effect of Reaction Parameters on PolyesterSynthesis

To improve propagation kinetics and increaseproduct molecular weight, reaction parameters suchas solvent, temperature, monomer/solvent ratio, en-zyme concentration, enzyme source, and reactionwater content were studied.43,51,52,59-61,75,78,80,81,84,101-111

A. SolventFor any enzyme-catalyzed process in organic me-

dia, the solvent plays a crucial role in determiningenzyme stability and regulating the partitioning ofsubstrates and products between the solvent and theenzyme. Enzyme specificity can be altered or specif-ically fine-tuned by proper selection of the organicsolvent. The nature of organic solvents is well-knownto be crucial for the maintenance of a critical watercontent necessary for catalytic activity. More hydro-philic solvents tend to strip the essential hydration

water from the enzyme, thus distorting its catalyticconformation. Alternatively, more hydrophobic sol-vents retain enzyme catalytic activity, thus leavinga water layer that adheres to the enzyme surface andacts as a protective sheath.101 When selecting thesolvent for polymer synthesis, it is important toconsider the viscosity and solubility of the monomerand corresponding polymer in that solvent. Thischoice will effect the confirmation of the polymer insolution as well as the diffusivity of the polymer andsubstrates. The polymer conformation and radius ofgyration in a solvent may alter its ability to reactwith, for example, an activated monomer or chainend. However, deciphering these relationships isdifficult since a change in the solvent will simulta-neously effect multiple parameters such as enzymeactivity and polymer chain extension in that solvent.Recently, Russel and co-workers studied the role ofdiffusion in biocatalytic polytransesterification. 112

Knani and co-workers43 showed that the solventsolubility parameter effects the self-condensation ofmethyl �-hydroxyhexanoate (73). The DP of thepolyester decreased with increasing solvent solubilityparameter [6-12 (cal/mL)1/2]. The DP of the synthe-sized polyester followed an S-shaped curve relation-ship with solvent log P (-0.5 < log P < 5). At a log P∼ 2.5, the polymer DP increased. Decreasing valuesof DP in good solvents for polyesters were attributedto better solubility and the removal of product oligo-mers from the enzyme surface, resulting in reducedsubstrate concentration near the enzyme. Examplesof solvent effect on polyester synthesis by condensa-tion or step-growth polymerization follows. Shuai etal.113 reported PPL-catalyzed polycondensation of3-hydroxybutyric acid (129) and 12-hydroxydode-canoic acid (104) in diethyl ether at room tempera-ture and toluene at temperatures of 55 and 75 °C.The yields of the products varied with solvent andreaction time. The best-isolated yield of 93% resultedfrom polymerizations in diethyl ether at room tem-perature (112 h, Mn ) 230). The highest Mn of 2900resulted from polymerizations in diethyl ether atroom temperature in toluene at 75 °C (56 h, 36%yields), respectively. Binns et al.55 studied copoly-merizations of adipic acid (85) and 1,4-butanediol (79)using Novozyme-435. They reported that in solvent-free conditions, the polymerization proceeds by astep-growth mechanism. However, in toluene, theyfound that transesterification reactions between chainstake on greater importance.

Examples of solvent effects on lactone ring-openingpolymerizations are described below. Dong et al.102studied �-caprolactone (23) polymerization at 37 °Cfor 240 h in a series of solvents (log P from 0.46 to5.0) using the lipase from a Pseudomonas sp. Enzy-matic polymerization in solvents with log P < 2 (THF,chloroform, 1,2-dichloroethane) gave low monomerconversion and molecular weight. In contrast, forsolvents with log P between 2 and 4 (benzene,toluene, diisopropyl ether, and cyclohexane), themonomer conversion and molecular weight was rela-tively higher. They were not able to directly correlatetheir results with log P values. Instead, they believethat the solvent dipole moment, substrate or polymer

Scheme 33. Ring-Opening Copolymerization ofPDL and TMC


solubility, and enzyme specificity all are factors thatultimately determine the outcome of solvent effectson polymerization reactions. Similar results havebeen reported by Kobayashi and co-workers114 forlipase CC-catalyzed polymerization of �-CL (45 °C for5-10 days) and γ-valerolactone (45 °C for 1-15 days).In our laboratory, Novozyme-435-catalyzed polymer-izations of �-CL at 70 °C were carried out. Solventshaving log P values from -1.1 to 0.49 showed lowpropagation rates (e30% �-CL conversion in 4 h) andgave products of short chain length (Mn e 5200g/mol). In contrast, solvents with log P values from1.9 to 4.5 showed enhanced propagation rates andafforded polymers of higher molecular weight (Mn )11 500-17 000 g/mol) (Figure 2).75

The monomer concentration in solvent is also animportant reaction parameter.81,75 In one study car-ried out in our laboratory using Novozyme-435 at 70°C for 4 h, ratios of toluene to �-CL of 0:1 (solventless),1:1, 1.5:1, 2:1, 2.5:1, 3:1, 4:1, 5:1, and 10:1 (wt/vol)were investigated (Figures 3 and 4).75 The region of

toluene/�-CL 1.5:2.5 gave substantially increasedreaction rates and product molecular weights. Forexample, solventless and polymerizations in toluene(toluene/�-CL 2:1 vol/vol) gave percent of �-CL con-version, Mn, and polydispersity values of 41 and 85%,10 800 and 17 200, and 2.1 and 1.8, respectively. Also,variation in the toluene to monomer ratio from 0:1to 5:1 to 10:1 gave products with polydispersityvalues of 2.09, 1.52, and 1.38, respectively. This trendin dispersity is explained by a decrease in lipase-

catalyzed transesterification reactions between chainsas the polymer solution concentration is diluted.

B. Enzyme ConcentrationDeng et al.89 reported on how variation in No-

vozyme-435 concentration affected �-caprolactonepolymerizations (bulk, 70 °C). Plots of monomerconversion as a function of reaction time for threedifferent catalyst concentrations are displayed inFigure 5. As anticipated, Figure 5 shows that by

increasing the catalyst concentration, the percent ofmonomer conversion rate increased. The total num-ber of polymer chains per millimole of monomer ([Np])increased as the Novozyme-435 concentration wasincreased. Thus, Mn and Novozyme-435 concentrationwere inversely related. Russell and co-workers 119studied the effect of enzyme/substrate concentrationon the condensation polymerization of divinyl adipate(27) and 1,4 butanediol (79). From a series of kineticexperiments, it was determined that, by increasingenzyme/substrate (E/S) concentration, higher molec-ular weights were attained in a relatively shorterreaction time. This is despite that at higher E/S ratiosthe extent of hydrolysis of vinyl ester groups washigher. Since the enzyme component of the reactionmixture normally has the relatively highest weightpercent of water, it follows that higher E/S ratios willlead to greater reaction water contents and, there-fore, greater extents of vinyl ester hydrolysis.

C. Reaction Water ContentThe importance of water concentration on various

small molecule lipase-catalyzed transformations in

Figure 2. Effect of solvent on the polymerization of�-caprolactone.

Figure 3. Effect of �-caprolactone concentration in tolueneon the polymerization rate.

Figure 4. Effect of �-caprolactone concentration in tolueneon the polymer Mn.

Figure 5. Effect of enzyme concentration on the rate of�-CL polymerization.


organic media has been well-documented.5,120 Asdiscussed above, water bound to the enzyme surfaceplays an important role in maintaining an enzyme’sconformational flexibility. Lipase-catalyzed polymer-izations create a new dimension to the importanceof water concentration in reactions. This is due totheir role during chain initiation. The reports belowfurther elaborate on the multiple functions of waterin lipase-catalyzed polymerizations.

Studies in our laboratory on PDL polymerizationat 70 °C catalyzed by lipase PS-30 (Amano, im-mobilized on Celite) showed that by lowering thereaction water content, large increases in poly(PDL)molecular weight resulted.80 This increase in productmolecular weight was accompanied by a decrease inpropagation kinetics. Furthermore, our analysis of aplot of the total number of chains ([Np]) versus thereaction water content for Novozyme-435-catalyzed�-CL polymerizations showed a positive slope and alinear relationship. Thus, by increasing the watercontent in reactions, there was a predictable increasein the number of chains.103 Matsumoto and co-workers104 used lipase-catalyzed ring-opening polym-erization of lactones such as 11-undecanolide (110)and 15-pentadecanolide (105) to estimate the watercontent in a reaction mixture. Using the P. fluore-scense lipase in cyclohexane, at 65 °C, they found thatthe number average molecular weight values cor-related to the water content in the organic solvent.Dong et al.102 also reported on the function of waterduring �-CL polymerizations (in bulk, 45 °C, 240 h).They observed that, during the initial stages ofreactions, the rate of monomer conversion increasedwith increases in water content from 0.1 to 1.0, 4.0and 16.0% (vol/vol) (Figure 6). The plot of Mn as a

function of time (Figure 7) shows that at 16% water,high amounts of low molecular weight product wasformed. Also, the profiles for the low molecularweight product at 0.1 and 16% water were nearlyidentical. The above studies agree that, within acertain range of water contents, the propagation rateincreases and the product molecular weight decreasesas the reaction water content increases.102-104 Thishas been explained as resulting from an increase inthe number of propagating chains with increasedreaction water.

D. Reaction TemperatureAttempts to increase the propagation rates of

lipase-catalyzed polymerizations by an increase inthe reaction temperature must take into account thethermal stability of the lipase. Efforts to increase thethermal stability of enzymes using immobilizationtechniques, changing the solvent, accessing lipasesfrom extremophiles, and intervening genetically areall relevant to opportunities for operating in vitropolymerizations at higher temperatures. Anothergeneral consideration is that lipase-catalyzed poly-merizations in solventless or concentrated solutionscan encounter diffusion constraints. Work in viscousreaction media can benefit by the ability to operateat higher temperatures.

When chain-type lactone ring-opening polymeriza-tion catalyzed by lipases does not have a terminationreaction (see section E) and in the absence of diffusionlimitations, the number of chain-initiating and -prop-agating events will control the chain length. In theabsence of adding nucleophiles to the polymerizationsor in the presence of impurities, water is the promi-nent initiating molecule. Thus, anything that tendsto perturb the concentration of water in the systemwill also alter the product molecular weight. Forexample, as the reaction temperature is increased,hydrogen bonding between protein and water will bedisrupted. Water that was once enzyme- or matrix-bound may be increasingly accessible. Thus, at highertemperatures, water may become increasingly avail-able in the polymerization to initiate new chains. Thestudies described below provide experimental datathat ultimately must be reconciled with the above orother hypotheses that explain temperature/molecularweight/enzyme activity relationships.

Kobayashi and co-workers78,84 found large increasesin percent of conversions of DDL (51) and PDL (105)by increasing the reaction temperature from 60 to75 °C. In their work, the reactions were conductedin bulk where diffusion constraints may be encoun-tered. The lipases used were those from P. fluorescensand lipase CC (C. cylindracea). For example, usinglipase CC, after a 65% PDL conversion at 75 °C and85% PDL conversion at 60 °C, the product Mn valueswere 16.2 × 103 and 6.7 × 103, respectively. Thus,by increasing the reaction temperature in viscousreaction mixtures, the mobility of monomer andpolymer components is increased and, therefore, sois the product molecular weight.

Figure 6. Effect of water content on the monomer conver-sion of PDL.

Figure 7. Effect of water content on the polymer Mn ofPDL.


Our laboratory studied the effects of temperaturesranging from 60 to 105 °C on �-caprolactone polym-erization catalyzed by Novozyme-435.75 To study thereactions as a function of time, toluene-d was usedas the solvent (�-caprolactone/toluene-d 1:5 v/v), andthe polymerizations were monitored in situ by NMR.The percent of monomer conversion as a function oftime increased with an increase in temperature from60 to 90 °C (Figure 8). A further increase in the

reaction temperature from 90 to 105 °C gave de-creased reaction rates. Study of the dependence ofmolecular weight with temperature showed thatincreasing the temperature from 60 to 105 °C re-sulted in a decrease in Mn (from ∼16 × 103 to 7.0×103) at similar (25%) conversion (Figure 9). It is

believed that, for solution polymerizations such as�-caprolactone in toluene-d, the thermal require-ments to overcome diffusional constraints are muchlower than was encountered for bulk systems. Thus,at lower temperatures where relatively less water is“free” for chain initiation, higher molecular weightproducts result. Similar results were also observedfor polymerization of PDL (105) in toluene.81

VI. Activation of Enzymes for Polymer Synthesisby Immobilization or Solubilization

Enzyme immobilization can result in improvedenzyme stability, recyclability, and activity for vari-

ous organic transformations.80,121 An alternative toenzyme immobilization is to solubilize the protein inthe organic phase. Techniques to achieve this objec-tive have been reviewed elsewhere;122 however, abrief description of these methods is provided belowfor the convenience of the reader.

At least two approaches to dissolve enzymes inorganic media have been successfully developed. Inthe first method, Takahashi and co-workers105 co-valently linked poly(ethylene glycol) (130) to exterioramino acids of the catalytic protein. Alternatively,the hydrophilic surface of an enzyme was nonco-valently associated with a surfactant.106 These meth-ods result in catalytic proteins that have improvedsolubility in the reaction medium and, therefore, givehigher reaction rates as compared to the enzymepowder.

Association of catalytic proteins with surfactantsresults in a hydrophobic outer environment that cansolubilize protein-surfactant complexes in nonpolarmedia (Scheme 34).106-109 Goto and co-workers110

showed that “surfactant coating” can result in lipaseesterification rates that are much greater than thecorresponding insoluble lipase powder. They alsodemonstrated that the pH of the buffer solution thatis used to form protein-surfactant complex must becarefully regulated to ensure that the complex formedhas the desired effect on the lipase activity. Goto andco-workers110 used the nonionic surfactant glutamatedioleyl ester ribitol amide (131). By adjusting phos-phate buffer solutions to the mid-pH range, thefollowing lipases were surfactant-coated: lipasesfrom C. cylindracea, Aspergillus niger, Rhizopus sp.,Muccor javanicus, and Pseudomonas sp. (see Scheme34). By surfactant coating the lipase from Pseudomo-nas sp., the reaction rate for esterification of lauricacid (132) with benzyl alcohol (133) increased byabout 2 orders of magnitude relative to the nonor-ganic soluble lipase powder. Other studies by Oka-hata and co-workers111 showed that dialkyl am-phiphilic lipid coated lipases from Rhizopus delemar,Pseudomonas fragi, P. fluorescens, Rhizopus niveus,and wheat germ had high activity in organic mediafor both glyceride ester synthesis106,111 and resolutionsof racemic alcohols.123

The above methods to activate catalytic proteinshave been applied to enzyme-catalyzed polymer

Figure 8. Effect of reaction temperature on the polym-erization rate of CL.

Figure 9. Effect of reaction temperature on Mn of poly-caprolactone.

Scheme 34. Immobilization of Lipases UsingCarbohydrate-Based Surface-Active Compounds


synthesis. Noda and co-workers124 used glutamatedioleyl ester ribitol amide (131) to coat and activatelipase PS. By surfactant-coating lipase PS, the per-cent conversion of PDL as a function of time at 60°C in cyclohexane increased by a factor >100-foldrelative to the nonsurfactant-coated enzyme powder(Figure 10). Furthermore, poly(PDL) formed by using

the native lipase PS powder, and the surfactant-coated lipase PS, had Mw values of 9.8 × 103 and21.3 × 103, respectively. Moreover, by using thesurfactant-coated lipase PS in cyclohexane, the po-lymerization of PDL proceeded to 99% conversion in72 h at 30 or 45 °C. In comparison with the nativelipase PS powder at temperatures up to 60 °C,percent of PDL conversion reached only 41% after72 h.

Recently, Dordick and co-workers125,126 observedthat the protease Subtilisin Carlsberg from Bacilluslicheniformis can be ion paired with a surfactant andthen extracted from aqueous solutions (8.5 mM,HEPES buffer, pH 7.8, containing 6 mM KCl) intoorganic solvents (e.g., isooctane). This catalyst systemretains its activity in isooctane for the acylation ofnearly all-available surface primary hydroxy groupsof amylose (136, thin film or fine powder). By solu-bilization of the enzyme and an acyl-substrate suchas vinyl caproate (135), oligo- or polysaccharidesubstrates (134) were the only insoluble reactioncomponent (Scheme 35). An alternative strategy is

to solubilize the polysaccharide in a polar aproticsolvent and to use the enzyme in its insoluble state.However, the later approach is flawed since it is well-

known that catalytic lipases and proteases have poorcatalytic activity in polar aprotic solvents. Further-more, the use of such solvents is generally undesir-able.

Our laboratory investigated the affect of enzymeimmobilization on ω-pentadecalactone (105, PDL)ring-opening polymerization.80 Solventless polymer-izations of PDL at 75 °C using nonimmobilizedpowdered lipase PS-30 as well as lipase PS-30 thatwas physically immobilized on Celite-521 were per-formed. In comparison to nonimmobilized lipase PS-30, the Celite-immobilized enzyme gave a muchgreater rate of monomer conversion and polyestersof higher molecular weight (Figures 11 and 12,respectively).

To further illustrate the importance of enzymeactivation, Novo Nordisk currently immobilizes lipaseB from C. antarctica on a polyacrylate resin. Thisproduct, marketed as Novozyme-435, has shownextraordinary versatility for a broad range of lactonestructural variants.65,92,127 This has been attributedto the absence of a “lid” that regulates the access ofsubstrate to the enzyme active site.128 Examples ofthe exceptional activity of this catalyst for manypolymer-related reactions are found throughout thisreview. Comparative experiments for polyester syn-thesis using the nonimmobilized form of lipase Bfrom C. antarctica have thus far shown reaction rates

Figure 10. Effect of immobilization (surfactant coated) onlipase-catalyzed polymerization of PDL.

Scheme 35. Enzymatic Modification of Amylose

Figure 11. Monomer conversion as a function of time forthe bulk polymerization of PDL catalyzed by NI- and I-PS30 at 70 °C (reaction water content of 0.20% w/w).

Figure 12. Number average molecular weight as a func-tion of percent conversion for the bulk polymerization ofPDL catalyzed by NI- and I-PS 30 at 70 °C (reaction watercontent of 0.20% w/w).


that are orders of magnitude lower than for No-vozyme-435.

VII. Kinetic and Mechanistic Investigations ofLipase-Catalyzed Ring-Opening Polymerization ofLactones

On the basis of publications from our labora-tory29,129 and others,84,42 it is believed that lipase-catalyzed lactone ring-opening polymerization pro-ceeds by an enzyme-activated monomer (EAM)mechanism (Scheme 36). For the case of PPL and

other lipases that contain an active serine moiety,the serine residue participates in the nucleophilicattack on a lactone such as �-CL to form an enzyme-activated monomer (acyl-enzyme intermediate) com-plex (EAM). For initiation, a nucleophile such aswater can react with the EAM complex to form themonoadduct (ω-hydroxyhexanoic acid). If other nu-cleophiles such as alcohols or amines are added tothe polymerization, they can replace water as theinitiating species.29 Changes in the initiator used canresult in faster or slower initiation rates and efficien-cies. Furthermore, the nucleophile that initiates achain will ultimately become the group at the chainend. Polymer chain growth or propagation takesplace when the ω-hydroxyl terminal group of a chainacts as the nucleophile that reacts with the EAMcomplex to give a product that is elongated by onerepeat unit.

For the �-CL/PPL monomer-enzyme system, threenucleophiles (butanol, water, and butylamine) wereused by us to study their affects on chain initiationand propagation.29 An experimental rate equationwas derived, and the living/immortal nature of thepolymerization was tested by a classical polymerchemistry approach. The results indicated that mono-mer consumption followed a first-order rate law andwas independent of the type and concentration of thenucleophile. Lipase-catalyzed polymerization of �-CLwas found to share many features with that of animmortal polymerization. Later, Deng and Gross89extended these studies to the �-CL/Novozyme-435monomer-enzyme system. They reported that therate of chain initiation (ki) was relatively slow ascompared to chain propagation (kp). To further assessthe “living” or “immortal” nature of the polymeriza-tion, first-order plots for three catalyst concentrationswere constructed after correcting for the amount ofmonomer consumed in initiation (where Np is con-stant).89 The plot of ln{([M]0 - [M]i)/([Mt] - [M]i)}versus time was linear (r2 ranged from 0.992 to0.996), indicating that there was no chain termina-

tion and that monomer consumption followed a first-order rate law. The absence of chain transfer and/orchain termination in these reaction systems impliedthat the product molecular weight may be controlledby adjusting the stoichiometry of the reactants. Theextent that the molecular weight distribution ofproducts is broadened by lipase-catalyzed transes-terification is discussed below.

Most recently, Henderson and Gross130 showed thatthe bulk polymerization of ω-pentadecalactone (PDL)at 50 °C catalyzed by an immobilized lipase from aPseudomonas sp. (I-PS-30) deviated from an idealliving system. Analysis of the kinetic data indicatedthat the polymerizing system consisted of propagat-ing chains that were slowly increasing in number andthat Mn was not a simple function of the [monomer]0to [initiator]0 ratio. It was also shown that the Mw/Mn (MWD) becomes narrower with increasing con-version and that only a fraction of the water availablein the system was used to initiate the formation ofchains. On the basis of these results, comparativestudies between slow initiation and slow exchangesinvolving dormant and active chains were made. Thecurrent model used to describe these observations isthat which governs the slow dynamics of exchangeduring propagation. Interestingly, the rate at whichmonomer was consumed was found to vary from firstto second order when examining the kinetics atdifferent stages of the reaction (i.e., 3-13% and 13-40% conversion, respectively). The fact that theconcentration of polymer chains ([R ∼ OH]) increasedwith monomer conversion may explain the increasedsensitivity of the reaction rate to monomer concen-tration as the reaction progresses.

VIII. Transesterification of Polyester SubstratesLipase-catalyzed transesterification reactions in

organic media have been used extensively for theresolution of alcohols and carboxylic acids fromracemic low molar mass esters.131 Later, step-growthcondensation polymerizations that involve transes-terification reactions between chain ends were usedto prepare various polymers including those that areenantioenriched (see section III). Recently, our labo-ratory reported the ability of Novozyme-435 to cata-lyze transesterification reactions between preformedpolyester substrates.132,133 In one example, transacy-lation reactions in bulk at 70-75 °C between poly-pentadecalactone (105b, 4 300 g/mol) and polycapro-lactone (23b, 9 200 g/mol) were studied. After 1 h,analysis of the product by 13C NMR showed that themixture of two homopolymers had been transformedinto random copolyester.133 An increase in the mo-lecular weight of the initial homopolymers (polyca-prolactone, Mn ) 44 000, PDI 1.65; polypentadeca-lactone, Mn ) 40 000, PDI 1.71), under identicalreaction conditions, resulted in the formation of amultiblock copolymer (Mn ) 64 700, PDI 1.97). Anincrease in the reaction time from 1 to 24 h resultedin a random copolymer. Thus, the kinetics of No-vozyme-435 transacylation reactions was chain-length dependent. From the above observations andthe currently excepted mechanisms of lipase-cata-lyzed polymerizations,133 the following mechanism fortransacylation between preformed polyesters was

Scheme 36. Proposed Mechanism forLipase-Catalyzed Ring-Opening Polymerization ofE-CL


postulated (Scheme 37). In the first step, lipasecatalysis results in the cleavage of an intrachain esterlinkage. The resulting enzyme-activated polymercomplex (EAPC) then reacts with the terminal hy-droxyl group at a chain end to form an ester linkage.These reactions continue and thereby reshuffle therepeat unit sequence distribution. In the exampleabove, the redistribution of repeat units occurred tothe extent that random copolymer resulted.132-134

IX. Modification of CarbohydratesMultifunctional groups on carbohydrate substrates

present unique challenges for selective modifications.There is great opportunity to direct selective modi-fications to create building blocks that are relevantto polymeric systems. Often, traditional chemicalmethods that lead to selective carbohydrate modifica-tion involve tedious introduction and removal ofprotecting groups.109 In contrast, enzymes can pro-vide high levels of regioselectivity so that simple one-pot reactions may replace multistep processes.

Our laboratory reported a method where the initia-tion of lactone or cyclic carbonate ring-opening po-lymerization occurs selectively from the 6-hydroxylposition of multifunctional alkyl glucosides (Scheme6). Since R,â-ethyl glucose (19) and �-CL (23) are bothliquids, this reaction was studied without solvent. Forexample, at 70 °C for 96 h, PPL catalyzed the highlyregioselective initiation of �-CL to give 1-ethyl-6-oligo-(�-CL)-glucopyranoside [6-oligo(�-CL)-EGP] (25) (e.g.,Mn ) 1120 g/mol, Mw/Mn ) 2.16).30 In a separatereport published shortly after us, Cordova et al.described a similar result by using Novozyme-435 foracylation of methyl galacto- and glucopyranoside(137, 18).135 The ω-hydroxyl group of oligo(�-CL) wasregioselectively esterified with vinyl acetate (41)using lipase PS-30 (from P. cepacia) (138).115 Then,the three free hydroxyl groups located exclusively atthe EGP headgroup of the resulting product (138)were used as sites for initiation of stannous oc-tanoate-catalyzed L-lactide (125, L-LA) ring-openingpolymerization. This gave a multi-arm heteroblockcopolymer (139) with spatially well-defined armsaround a carbohydrate core (Scheme 38).

Relative to monosaccharides, polysaccharides areeven a more difficult class of materi

Polymer Synthesis by In Vitro Enzyme Catalysishomepages.rpi.edu/~grossr/_doc/_doc/publication/Chem....

Documents

Transcript of Polymer Synthesis by In Vitro Enzyme Catalysishomepages.rpi.edu/~grossr/_doc/_doc/publication/Chem....