THE DEVELOPMENT AND MECHANISTIC STUDY OF DUAL H-BONDING ORGANIC CATALYSTS FOR THE ROP OF CYCLIC ESTERS

Organic catalysis for the Ring Opening Polymerization (ROP) of cyclic monomers is a rapidly emerging field of study that gained interest in 2005 with the advent of dual H-bonding catalysts. Synthesizing catalysts that produce fast reaction rates with superior reaction control over molecular weight (Mn) and molecular weight distributions (Mw/Mn) are of great interest for material applications. Current organic catalysts do not have the capabilities to satisfy these requirements, limiting the feasibility to pursue commercial scale applications. Analysis of polymerizations is done using a number of techniques. Nuclear Magnetic Resonance (NMR) is a power spectroscopy technique used to evaluate reaction progression for polymerization reactions. Through reaction conversions, the kinetics of each catalyst can be measured and compared with one another. Through NMR titration experiments, binding studies were used to compare and in some cases quantify the interactions between monomer and alcohol/chain end with the catalyst and cocatalysts respectively. Gel Permeation Chromatography (GPC) is another technique used for the analysis of polymers, which allows for the determination of the polymer molecular weight (Mn) and molecular weight distribution (Mw/Mn). The catalyst chosen to perform the ROP of monomer has a large impact on the control over the Mn and Mw/Mn. This method allows for the determination of polymer Mn and Mw/Mn, which translate to reaction control. Organic catalysis for the Ring Opening Polymerization (ROP) of cyclic monomers is a rapidly emerging field of study that gained interest in 2005 with the advent of dual H-bonding catalysts. Synthesizing catalysts that produce fast reaction rates with superior reaction control over Mn and Mw/Mn are of great interest for material applications. Current organic catalysts do not have the capabilities to satisfy both requirements limiting the feasibility to pursue commercial scale applications. First, a review of H-bonding organic catalysts and their relative reactivity will be discussed. The polymerization of cyclic esters by H-bonding (thio)urea has greatly increased since the first iterations of catalyst scaffolds. The incorporation of multi-armed H-bond donating species saw drastic increases in reaction rate. The incorporation of an oxygen (urea) in substitution of a sulfur (thiourea) saw an increase for all H-bond donors tested. These reactions also remained well controlled. These catalysts have been shown to be tolerant of solvent free polymerizations. The adoption of solvent free reactions is greatly valued by the commercial industry. Solvent free conditions allowed for the polymerization of several copolymers that were not possible through reactions within solvent. H-bonding (thio)urea catalysts used for the ROP of caprolactone were subjected to elevated temperatures (22-110°C). 1-O and 2-O produced linear Eyring plots out to 110°C (highest temperature evaluated). All other catalysts deviated from linearity at 80°C, due to decomposition of the H-bonding species. A switch to polar solvent alleviated decomposition for some H-bond donors while other remained curved. A mechanistic reasoning will be discussed. The introduction of a chiral architecture into the catalyst scaffold made kinetic resolution of racemic lactide possible. This chiral scaffold was responsible for an increase in isotacticity (Pm) of the resulting polymer. Multi armed chiral H-bond donors saw increase reaction rates but only small increase in Pm value versus mono-armed H-bond donors. A decrease in reaction temperature produced enhanced the Pm values. A new class of bifunctional, quinone derived catalyst was developed for the ringopening polymerization (ROP) of lactone monomers. Similar in architecture to other bifunctional catalysts, the quinone catalyst can activate monomer and alcohol/chain simultaneously. Attempts at ROP of both δ-valerolactone and L-lactide were unsuccessful. A mechanistic explanation is discussed. H-bonding urea or thiourea catalyst paired with a base cocatalyst have been employed for organocatalytic ring-opening polymerization (ROP) of aliphatic lactones (TOSUO, 4-MCL, 3,5-MCL and 6-MCL). Random copolymers with low dispersities were synthesized. A series of copolymers of CL and 3,5-MCL were produced and evaluated using TGA and DSC. Variation of the substituent along with its position on the monomer resulted in a different reaction rates. The relative rates of ROP for functionalized ε–caprolactone (4-MCL, 3,5-MCL, 6-MCL, and TOSUO) by H-bonding organic catalysts have been evaluated and a mechanistic reasoning discussed. H-bonding organic catalysts saw increased reaction rates and control for all monomers versus both metal and enzymatic catalysts.

x                     11 In practice, these conditions arise when a polymerization has a fast initiation rate relative to propagation rate and few to no side reactions. We shall refrain from pointing out when a catalyst (system) exhibits the characteristics of a 'living' polymerization, and rather point out when it is either especially well-controlled or exhibits low levels of control. Several, thorough reviews have been conducted in the wider field, 12-21 but not with quite the level of focus that the current platform provides.

LIST OF TABLES
Hence, we will attempt to emphasize the virtues and deficits of the various catalysts, especially as they contrast to other organic catalysts for polymerization.

Thiourea-mediated Stereoselective ROP
The stereoselective ROP of rac-lactide is an attractive method for the generation of polylactides (PLAs) with highly regular or novel stereosequences, and the modular scaffold and rich diversity of chiral thiourea H-bond donors has proved an enticing target for several groups. The ROP of rac-or meso-lactide to generate highly tactic PLA has been well documented. [69][70][71] Briefly, stereoselective enchainment of the chiral monomer onto the chiral chain end can occur via control rendered by 1) the propagating chain end, 2) a chiral catalyst or 3) a mixed mechanism. 69,72,73 For the ROP of rac-LA, a high probability of propagating with retention of stereochemistry (P m = probability of meso enchainment) will result in a highly isotactic PLA. 3,69 Waymouth and Hedrick reported the (R,R)-1 mediated ROP of rac-lactide to proceed with modest selectivity (P m = 0.76); however, 2/(-)-sparteine catalyzed ROP of rac-LA rendered similar selectivity (P m = 0.77). 3 The polymers did not display a melting point, suggesting low stereoregularity. 3 Exceeding these P m values has become a benchmark of sorts for the stereoselective ROP of rac-lactide by H-bonding catalysts. Despite its successes, (-)-sparteine itself fell out of favor as an organocatalyst when it became scarce circa 2010, but a replacement base, benzyl bispidine, was disclosed which renders similar reaction rates and selectivity in the ROP of rac-lactide with 2, P m = 0. 74. 47,74 Recent research into photoresponsive azobenzene-based thiourea, 3, for the ROP of rac-lactide suggests a conceptual approach to switchable organocatalysts for ROP. 75,76 Catalysts that are switchable by external stimuli (i.e. redox pathways, lights, coordination chemistry etc.) 76 75 The ROP was proposed to proceed from the trans-isomer, presumably via a chain-end control mechanism. 3, 75 We make the safe prediction that switchable organic catalysts for ROP will play an important role in the next decade. 76 which was the only one of the examined structures to achieve full conversion in 24 h. 99 No epimerization was observed during polymerization. A classic H-bond mediated mechanism of enchainment was corroborated by NMR titration studies. 99 The H-bonding ability of squaramides is perturbed versus that of thioureas, 99 but they have approximately the same acidity (Schreiner's thiourea (8) pK a = 8.5; 6 pK a = 8.4; both in DMSO). 101,102 The altered structures possessing minimally altered pK a may have unseen implications for nascent imidate-mediated ROP, see below.

RATE-ACCELERATED DUAL CATALYSIS
From the very early days of the field, thiourea/base cocatalysts exhibited remarkably controlled ROP, so remarkable that the poor activity and productivity of the catalysts could be justified. However, with the application of N-heterocyclic carbene (NHC) and TBD organocatalysts to ROP, it became very clear that organocatalysts could possess activity to rival that of metal catalysts. 16,23,49 The dream of combining the rate of NHCs or TBD with the high selectivity of thiourea/base systems became an alluring research goal for several groups. One route that can be envisaged uses internal Lewis acids to stabilize the (thio)urea as it binds to monomer. The challenge became finding synthetically accessible (thio)ureas with Lewis acids that are compatible with ROP.

Internal Lewis Acid Enhanced H-Bond Donors
A urea H-bond donating catalyst with an internal boronate ester, 9, displayed enhanced activity versus its parent urea, 10 ( Figure 1.4). HBD 9 was applied with sparteine cocatalyst for the ROP of LA at room temperature (k 2 /k 9 ~ 1). 103  An activated-(thio)urea mechanism is proposed for multi-H-bond donor mediated ROP in non-polar solvent, but urea H-bond donors remain highly-active in polar solvent.
Kinetic studies on the several systems in benzene-d 6 reveal the (thio)urea ROPs to be first order in monomer, initiator, and cocatalysts, suggesting one mono-/bis-/tris-H-bond donor acting at one monomer in the transition state. 48,53,54,105 H-bonds are electrostatic in nature and have low directionality, 106 which allows for the possibility of multi-(thio)ureas directly activating monomer in a multi-activation mechanism. Computational models suggest that tristhiourea 14 is C3 symmetric (all H-bonded), 105 and an analogue of 15 with n-propyl (versus ethyl) linking arms is highly inactive for ROP, 107 suggesting that the (thio)urea moieties prefer to bind to themselves. These experiments, along with computational studies, suggest an activated-(thio)urea mechanism is operative in non-polar solvent. 105 Traditional H-bonding catalysts (e.g. 2/base) become very inactive in polar solvent, which limits their utility. 3 The urea HBDs, however, remain highly active in polar solvents (e.g. acetone and THF). 105,108 Recent, and still-evolving, studies suggest that a different mechanism involving urea anions is operative in polar solvent. [58][59][60] Urea and Thiourea Anions The deprotonation of urea or thiourea with strong bases (alkoxides or metal hydrides) has been shown to produce the corresponding urea anion or thiourea anion (also: imidate or thioimidate) which are incredibly active for the ROP of lactones. 59 times more active than thiourea anion motif. 59,60 Polymerizations with VL and CL were also completed within seconds. 59 An ROP with similar activity can be achieved by a urea (e.g. 16) plus strong organic base (e.g. MTBD, DBU, BEMP) cocatalyzed ROP. 108 The latter method may be operationally simpler, and urea plus organic base cocatalyzed ROP may be more controlled, especially post polymerization. 108 The rates of the two methods appear to be very similar and mark a departure from early H-bond mediated ROP: seconds instead of hours or days! Remarkably, the ROPs remain highly controlled. An antibacterial compound, triclocarban (TCC, Scheme 1.6), was shown to be a very effective H-bond donating catalyst for the ROP of lactones when used with organic base cocatalysts. 108 It was proposed that this compound effects ROP through the same mechanism as other urea/strong base mediated polymerizations, and TCC/BEMP displays the same approximate rate and control behavior as trisurea (15)/BEMP, although the trisurea is more active (k 15 /k TCC ~4, VL). 105,108 We anticipate that the movement towards readily available reagents will prompt wider adoption of organocatalysts and facilitate new applications; the success of TBD may be due, at least in part, to its commercial availability. To demonstrate this point, TCC/base cocatalyzed ROP was applied to the solvent-free polymerization of several lactones, which was previously limited due to 1) the presumed inactivity of urea HBDs in polar (monomer) solvent, and 2) the large amounts of catalyst required for neat conditions. 58 Solvent-free ROP catalyzed by TCC/base allowed for the one-pot synthesis of di-and tri-block copolymers, and TCC/alkylamines were effective for the solvent-free ROP of LA, 58 a longstanding challenge. 110 The reactions remained highly controlled and 'living' in nature despite solidifying prior to full conversion.

Sulfonamides, Phosphoric and Phosphoramide H-bond Donor/Acceptors
A selection of mono-and bis-sulfonamide HBDs which have been applied with base cocatalysts for the ROP of LA are shown in Figure 1

Phenol and Benzyl Alcohol H-bond Donors
Considering their efficacy for the ROP of several monomers, electron deficient

Electrostatic Monomer Activation by Cations
H-bonds -a very poor name for the phenomenon -require no orbital overlap and are a type of electrostatic interaction. 106  Benzoic acid, which is widely used to quench organic catalysts by protonating amine bases, 2 forms an active ROP cocatalyst when mixed 1:1 with DBU. 128

Amino-Oxazoline
The structures of amino-oxazolines and thiazolines are analogous to that of TBD.
An initial screening of the thiazoline catalyzed ROP of LA determined that thiazolines with electron withdrawing groups resulted in reduced ROP activity and produced atactic Mechanistic studies suggest that ROP is initiated from the CD and that the lactone/CD inclusion complex is vital to catalysis. When ROP is attempted using an acylated CD (no free hydroxyls), no conversion to polylactone is observed, which suggests that CDs are covalently attached to the polylactone chain end in a normal CDcatalyzed ROP. 140 Further, suppression of the ROP of VL was noted with a β-CD/adamantane inclusion complex catalyst system. The adamantane guest is strongly inserted in the β-CD cavity, which excludes VL, suggesting that lactone/CD inclusion complexes are essential for ROP. 140 The mechanistic picture that emerges suggests that, initially, a complex is formed between lactone and CD at a ratio of 1:1, and a hydroxyl group at the C 2 -position attacks the monomer to begin enchainment. Further development of these or similar extremely mild catalysts for ROP could provide new and exciting methods of ultra-controlled ROP.

CONCLUSION
The narrative of this chapter can be summarized by following the circular evolution of dual catalysts away from and back towards the popular organocatalyst, TBD.
When the TBD catalyzed ROP of lactones was disclosed in 2006, 23 it was the perfect storm of a successful catalyst. It is easy to use, readily available, highly active and exhibits decent selectivity for monomer and control (M w /M n ~ 1.2). While TBD was originally proposed to operate via a nucleophilic mechanism of enchainment, an H-bond mediated, bifunctional, mechanism was also envisaged. 23 This mechanism has been much debated, and it is not entirely certain which mechanism is operative and when. 32 By no means is this story complete, and as of January 2018 our mechanistic understanding of nascent urea/strong base mediated ROP is still evolving. Indeed, the broader field of organocatalytic polymerization is a bridge between the disparate worlds of materials chemist (ease of use) and synthetic polymer chemist (mechanistic interest).
We assert that the cooperative and collegial nature of our community has facilitated the synergistic evolution of new mechanism to new abilities -in monomer scope, polymer architecture and level of reaction control. We hope that this will continue.

INTRODUCTION
The H-bonding catalysts for ring-opening polymerization (ROP) stand out among the highly controlled polymerization methods for their ability to tolerate functional groups while precisely controlling molecular weight and polydispersity. 1-7 H-bond donating cocatalysts are believed to effect a "living" ROP via dual activation of monomer by a H-bond donor, usually a thiourea (TU), and activation of alcohol chain end by base cocatalyst. 8,9 The exquisite and remarkable combination of rate and selectivity present in other fields (e.g., olefin polymerization catalysis) 10,11 has yet to be paralleled in organocatalytic ROP, especially H-bond mediated transformations. The development of organocatalysts for polymerization has largely proceeded along divergent pathways toward highly selective 1,9,[12][13][14][15] or highly active [16][17][18][19] catalysts. Indeed, the low activity of organocatalysts for ROP has been specifically identified as a shortcoming of the field, whereas highly active metal-containing catalysts for ROP are well-known. 20   This is consistent with previous suggestions that H-bond-mediated ROP operates via dual activation of monomer by 1 and of alcohol chain end by base. 1 Because H-bonds require no orbital overlap and are electrostatic in nature, 26 we cannot rule out a dual-thiourea activated mechanism, Eq. 2.1. However, computational studies for the activation of lactones by 2-S suggest an activated-TU mechanism is preferred over a dual-thiourea activation mechanism, Eq. 2.1; this assertion is also supported by the 2-S/alkylamine cocatalyzed ROP of lactide. 22,27 The series of thiourea H-bond donating catalysts was extended to a trisTU H-bond donor, 3-S, but this catalyst exhibits significantly reduced activity versus 1-S or 2-S in the TU/base cocatalyzed ROP of lactones, Initiation of a CL ROP from 1-pyrenebutanol produces PCL with overlapping refractive index and UV traces in the GPC, suggesting end-group fidelity; the "living" alcohol chain end is susceptible to chain extension by repeated additions of monomer, (see  Table 2.4). suggests that the activated-urea mechanism is the more robust proposal. In small molecule transformations, urea H-bond donating catalysts have been observed to possess similar activity to their heavy chalcogen counterparts. 33 The development of urea and thiourea H-bond donating catalysts continued apace until the turn of the millennium when several reports emerged that extolled the operational (e.g., increased solubility) 34,35 and synthetic (e.g., higher yields and enantioselectivities) [35][36][37] benefits of thioureas over ureas. In our estimation, the ubiquity of the thiourea motif in H-bond mediated transformations may be more due to the coincidental timing of these reports than any general superiority of thioureas over urea H-bonding catalysts. Indeed, ureas are more polar than thioureas and should be expected to be better H-bond activators, 33 and in some catalysis applications, urea catalysts are clearly superior. 38,39 The with slightly reduced reaction rates or M w /M n (see Table 2.5).

Synthesis of 1-[3,5-bis(trifluoromethyl)phenyl thiourea]-3-aminopropane.
Preliminary studies suggest that these catalysts exhibit the same reactivity trends in small molecule transesterification and, hence, may have general applicability beyond ROP. The transesterification of ethyl acetate (1.6 M) with benzyl alcohol (1.6 M) was conducted in C 6 D 6 . Observed rate constants (k obs ) at early reaction time were measured for each H-bond donor/MTBD cocatalyzed transesterification. These rate constants show the same trends in catalyst activity that were observed for the ROP reactions: 3-O is the most rapid catalyst and it is 1−2 orders of magnitude more rapid than 1-S, (see Table   2.3). This suggests a general role for the increased activation of esters by urea H-bond donors (versus thioureas), yet the slower rates for the transesterification of s-trans (versus s-cis) esters accounts for the low rate of transesterification post polymerization, (see Table 2.6).

Computational Data
Dual-thiourea activation in DCM       (Thio)ureas have made great advances in the ROP community, rivaling some of the fastest known metal catalysts available. [18][19][20][21] Although they have been shown to be fast and selective, little work has been done to illustrate their ability towards stereoselective polymerizations. 22 Due to their non-covalent, H-bonding interactions, stereospecific reactions could be perceived as too difficult. However, many small molecule transformations have been performed using chiral thioureas and with great success. [23][24][25] The functional group tolerance of (thio)urea H-bonding catalysts signifies a great opportunity for catalyst optimization for stereocontrolled ROP of rac-LA. 18   The ability to perform stereoselective polymerizations would mean little if the ability to analyze the polymers was not adequate. Determination of stereo sequences has been described previously and is used widely for the analysis of these polymers. 28,29 Through the selective 1 H decoupled NMR, stereo sequences can be analyzed (Figure 3.1).

3-O in vacuum
From analysis of these NMRs, a P m value can be obtained, which denotes the probability to propagate a meso stereocenter. A value equal to one indicates isotactic polymer.
Initial studies into the kinetic resolution of rac-lactide by thioureas have been done previously within our research group with promising results. 30 This study looked at various H-bond donors for the kinetic resolution of rac-LA and the resulting P m values.
The catalysts shown in Figure 3.2, contain the same thiourea backbone, with changes only of a single substituent resulting in altered P m values. 24 The polymerization of rac-LA The decrease in temperature produced an increase in P m value to 0.82. The reaction temperature was dropped even lower to -78°C. However, due to solubility issues, no conversion of monomer was seen after 20 hours of monitoring (Table 3.2).          zero conversion to polymer after 20 days, see Figure 4.5. We presume that the observed conversions are due to initiation from base. 30 Despite being inert separately, the combined solutions can yield an ROP so rapid, that the combined solvent-free solutions  The ROP appeared to be living in nature but sluggish, reaching full conversion to poly(valerolactone) (PVL) in days and displaying a broadened M w /M n . 31 We believe the ability to conduct rapid and highly controlled ROP of lactones like VL and CL under solvent-free and non-melt conditions constitutes an advantage of the TCC/base cocatalysts over other (organo)catalyst systems.  38 Further, organic catalysts are susceptible to charring/decomposition at high temperature. 39 However, neither deactivation nor decomposition appear to be a concern for TCC/base cocatalyzed ROP at 80°C (Table   4.3). TCC/BEMP were also applied for the solvent-free ROP of PDL from benzyl alcohol (  17 and ureas with fewer electron withdrawing substituents have been observed to be more active. 16 A screen of base cocatalysts with 2-S showed PMDTA cocatalyst to exhibit a good combination of high rate and control (Table 4. For the ROP of lactide, the effects of reaction conditions on polymer tacticity must also be considered. For each polymerization in Table 4.4, the percent isotacticity was determined from the isolated polymer by 1 H decoupled 13 C NMR using previously established tacticity-dependent chemical shifts (Experimental Section). 13  impurities can be a concern. 39 The 2-S/PMDTA cocatalyzed ROP (Table 4 17,36 This suggests that the relative activity of urea vs thiourea is not dictated by solvent, and the various monomers seem to exhibit a preference for urea vs thiourea. Our group previously described the activity of thiourea/amine base cocatalysts in the ROP of LA as being related to the nature of cocatalyst binding (i.e. enthalpic vs solvophobic binding), 34 and understanding the preference exhibited by LA for thioureas vs ureas may require a full study of the solution interactions at play during an ROP catalyzed by the various catalysts. We are unable to measure a urea/LA binding constant due to poor solubility in non-hydrogen bonding solvents.                         , strong base cocatalysts and polar solvent. [5][6][7][8][9] Urea H-bond donors have been shown to be more active than the corresponding thioureas. 5,7,10,11 These trends also hold for the multi-(thio)urea H-bond donors developed by our group for ROP of esters. 10,12 The internal H-bond stabilization rendered by the extra (thio)urea moieties is thought to be the source of the augmented activity (versus mono-(thio)urea donors). 10,12 The active catalytic forms -H-bonding and imidate -are in rapid equilibrium unless a strong inorganic base (e.g. alkoxide or hydride) is applied, in which case the imidate is the catalytic species, Figure 5.1. 7,9,10 One advantage of the H-bonding class of catalysts is their efficacy for room temperature ROP, 1 but some applications mandate the application of elevated temperatures. For example, the solvent-free ROP of lactide (LA) had been identified as a challenge for organocatalysts. 13 The high temperature required to melt the polymer (180°C) typically results in charring when organic catalysts are applied. 13 We recently  The activation parameters of enchainment are superimposed with those of catalyst dynamics/reagent binding, 9,17 and these observed activation parameters (ΔH ≠ obs and ΔS ≠ obs ) are also given in Table 5.1. 18 The ROP of CL was chosen because the slower reaction kinetics (versus VL or lactide) facilitate monitoring by aliquot or 1 H NMR, and the ROP of CL features a high ceiling temperature (T ceil = 261 °C) 19 , which suggests that any temperature dependent observations are not due to substantially diminished enchainment equilibrium constants. weaker H-bonding at higher temperatures. 17 Binding to monomer is also exothermic. 15,17 These observations suggest that Arrhenius reaction acceleration outpaces the weakening of H-bonding resulting in faster ROP until catalyst decomposition >80°C. The higher ΔH ≠ obs in non-polar solvent (vs polar) corroborates the suggestion that Arrhenius behavior is resisted by catalyst dynamics -that is, thermal rate effects are partially offset by weakened H-bonding. 18 The to occur. 11,16 Our group recently disclosed that 2-S/PMDETA was optimal -in terms of rate and isotacticity -for the solvent-free ROP of LA, 11 and this was the only catalyst system whose temperature dependent kinetics were examined herein. We have                                 Recently, the deprotonation of (thio)ureas to produce (thio)imidate species ( Figure 6.1.b) have been shown to be very active for the ROP of cyclic esters. 2,[9][10][11] Reminiscent of TBD, the (thio)imidate can activate monomer and alcohol/chain end through a bifunctional process. However, with the negative charge on the (thio)imidate, it is possible for the species produce undesired side reactions, potentially decreasing the overall control of the reaction. As an alternative, we proposed that a quinoidal catalyst ( Figure 6.1.c), which is structurally similar to the (thio)imidate character but remains chemically neutral, might be the best of both worlds with the rate of the imidate catalyst and the control of a neutral (thio)urea species in an easily accessible bifunctional scaffold.  2b from the initial bright red crystals to brown and black crystals also suggests decomposition. This is not surprising given previous reports of acyl substituted quinoidal species described as very reactive and susceptible to nucleophilic attack. 13,14 Decomposition of the oxidized products, 1b and 2b, could be due to similar reactions arising from the highly electron withdrawing character of the trifluoromethyl substituted aryl ring. In response to this hypothesis, two new urea-based catalysts with less electron withdrawing character were synthesized (Figure 6.3.3b and 4b). 3b containing a 3,5-   (Figure 6.6).
The conversion of monomer to polymer in organocatalytic polymerization is due to the activation of both the monomer and initiator/chain end using a catalyst (dual or bifunctional), without these species, conversion does not happen. 4 The lack of turnover of either δ-VL or L-LA found for 5b, suggests such weaknesses could be present.

CONCLUSION
The quinoidal bifunctional catalyst species, 5b, was inactive for the ROP of both L-LA and δ-valerolactone. This is likely a result of minimal binding interaction between the catalyst and both monomer and alcohol/chain end, which was shown through multiple binding experiments between the species. The lack of activity when compared to the highly active imidate or guanidine base TBD represents the importance of binding interactions for ROP. The absence of binding interactions for 5b are likely due to the substituents on either side of the urea moiety. Without the strong electron withdrawing group found in the 1b or 2b substituted structure, the ability of H-bonding for 5b to a monomer is hindered. Attempts at synthesizing a more active quinoidal catalyst remains difficult given the decomposition observed with those bearing aromatic groups. However, an electron deficient alkyl chain could be created potentially circumventing this problem.
The quinoidal substituted side also lacks the ability to accept an H-bond, which could be due to the electron density of the quinone group moving away from the nitrogen, rendering the basic nitrogen weaker. TBD on the other hand has a much more basic nitrogen with perhaps just enough H-bonding characteristic to be capable of ROP which is likely the reason for is high activity and it's low controllability. The imidate catalyst, b ( Figure 6.1), formed from a proton transfer is active to H-bonding due to the electron withdrawing aryl ring and strong basic character to accept an H-bond due to the negative charge of the imidate itself. While it is possible the quinone derived catalyst species could be active for ROP, the current scaffolds lack the capabilities. Further work within this avenue could produce active catalysts that are highly active and controlled.     III. IV.

INTRODUCTION
The production of aliphatic polyesters are of interest due to their biodegradability and biocompatibility for potential applications within the medical 1-4 and material fields. [5][6][7] These materials are typically synthesized through polycondensation reactions, but the need for more controlled reactions has led to the ring-opening polymerization (ROP) of cyclic esters using metal catalysts. These catalysts have shown good control over molecular weight (M n ) and molecular weight distribution (M w /M n ). For example, Breteler, using a chiral salen AlOiPr complex was able to produce P-6-MCL with M w /M n of 1.04. 8 The polymerization of other methyl substituted ε-caprolactones (ε-CL) have been identified as well, with moderate control over M w /M n (1.12 -2.8). [8][9][10][11][12][13][14][15] In an effort to avoid metal catalysts, some have resorted to enzymatic catalysts for ROP of cyclic esters. [16][17][18][19][20][21][22] Enzymatic catalysts however, are plagued by poor solubility, long reaction times and low conversions. Several lipases have been employed for the polymerization of the aliphatic monomer 4-MCL in the monomer bulk at 60°C, 23 but were slow and did not yield high conversions (< 35% conversion after 10 days). 1,4,8-trioaspiro[4.6]-9undecanone (TOSUO) has also been evaluated using metal catalysts. 24-28 TOSUO is of particular interest due to the ease of post polymerization modification due to the ketal component within the backbone.
As an alternative to metal and enzymatic catalysts, the ROP of aliphatic lactones using H-bond donating (thio)urea catalysts has the potential to be fast, controlled and free of metal contamination. Here in, we report on the (co)polymerizations of TOSUO and methyl functionalized ε-caprolactone monomers (4-MCL, 3,5-MCL and 6-MCL) using H-bond donating (thio)ureas and base cocatalysts. To understand the variation in rate constants for functionalized CL monomers, a series of binding experiments were conducted. Equilibrium constants (K eq , Scheme 7.1) between TOSUO, 3,5-MCL, 6-MCL and ε-CL to 1-S were calculated to be 14.1, 14.8, 18.9 and 42 M -1 respectively, using 1 H NMR titration experiments. 35 The high K eq value of CL indicates stronger binding for the un-substituted ring. However, the rate constant of ε-CL is similar to TOSUO. Also TOSUO, 3,5-MCL and 6-MCL have similar K eq , yet        However, integration over the ranges for both monomer (red) and polymer (blue) resonances separately allows for the calculation of conversions.  M w /M n = 1.07). Temp program: start 25°C, cool to -40°C at 5°C/min, heat to 100°C at 5°C/min, cool to -40 at 5°C/min, heat to 100°C at 1°C/min, cool to -40°C at 1°C/min.
Green line shows temperature program and the red line is the thermal response.