Article pubs.acs.org/Macromolecules
Mechanistic Insights into the Organopolymerization of N‑Methyl N‑Carboxyanhydrides Mediated by N‑Heterocyclic Carbenes Laura Falivene, Miasser Al Ghamdi, and Luigi Cavallo* KAUST Calaysis Center (KCC), Physical Sciences & Engineering Division (PSE), King Abdullah University of Science and Technology (KAUST), Thuwal 23955-6900, Saudi Arabia S Supporting Information *
ABSTRACT: We report on a DFT investigation of initiation, propagation, and termination in the organopolymerization of N-methyl N-carboxyanhydrides toward cyclic poly(N-substituted glycine)s, promoted by N-heterocyclic carbenes (NHC). Calculations support the experimentally based hypothesis of two competing initiation pathways. The first leading to formation of a zwitterionic adduct by nucleophilic addition of the NHC to one of the carbonyl groups of monomer. The second via acid−base reactivity, starting with the NHC promoted abstraction of a proton from the methylene group of the monomer, leading to an ion-pair-type adduct, followed by nucleophilic attack of the adduct to a new monomer molecule. Chain elongation can proceed from both the initiation adducts via nucleophilic attack of the carbamate chain-end to a new monomer molecule via concerted elimination of CO2 from the carbamate chain-end. Energy barriers along all the considered termination pathways are remarkably higher that the energy barrier along the chain elongation pathways, consistent with the quasi-living experimental behavior. Analysis of the competing termination pathways suggests that the cyclic species determined via MALDITOF MS experiments consists of a zwitterionic species deriving from nucleophilic attack of the N atom of the carbamate chainend to the CO group bound to the NHC moiety.
■
imidazolium enolaluminate intermediates.23−26 This overview is completed by the NHC-mediated polymerization of Nsubstituted N-carboxyanhydride (NCA) to produce cyclic poly(N-substituted glycine)s, e.g., polypeptoids27,28 (see Chart 1).
INTRODUCTION Since Arduengo’s first isolation of a stable N-heterocyclic carbene (NHC),1 the impact of this class of molecules as ligand in organometallic catalysis and as initiator or catalyst itself in organic chemistry has become enormous. In the organometallic field several catalysts based on NHC are already used for a number of different chemical transformations approaching large scale industrial applications, such as Ru-mediated olefin metathesis,2−5 Pd-catalyzed cross-coupling reactions,6−8 Cucatalyzed reduction of ketones, C−H activation, hydrogenations, and Au-mediated transformations. This impressive number of applications is due to the flexible architecture of NHCs, which allows tuning the steric and the electronic properties of these ligands to achieve the desired catalytic activity.9−12 The use of NHCs as organocatalysts is growing to similar relevance, with applications in the polymerization of polar monomers as a striking example. Among the most relevant NHCs promoted polymerizations there is the ring-opening polymerization (ROP) of cyclic esters (e.g., lactides,13,14 βlactones15−17) to yield the respective homo- or copolyesters. Along the same line, the ROP of ethylene oxide to exclusively linear poly(ethylene oxide) was reported.18,19 Another remarkable example is the polymerization of α,β-unsaturated esters (or acrylic monomers) such as methyl methacrylate and methylene butyrolactones, realized using NHCs as nucleophilic initiators,20−22 as well as the rapid polymerization of the same monomers family by frustrated Lewis pairs consisting of bulky bases, and of sterically encumbered alane acids, via zwitterionic © XXXX American Chemical Society
Chart 1
As most of the NHC-promoted ROP, the reaction occurs via ring-opening of the heterocyclic monomer, in this case Ncarboxyanhydride. The main difference with respect to standard ROP is that a small molecule, CO2 in this case, is liberated at each polymerization step. Detailed mechanistic study on the ROP of N-butyl NCA suggested the appearance of two different initiating species in the reaction mixture after addition of the NHC to a solution of NCA.28 In particular, 1H NMR spectra evidenced two distinct sets of proton resonances due to the adducts from the nucleophilic attack of the NHC to a NCA molecule: one is consistent with a bis(isopropyl)imidazole-2-ylidenium species, as Received: August 9, 2016 Revised: September 14, 2016
A
DOI: 10.1021/acs.macromol.6b01722 Macromolecules XXXX, XXX, XXX−XXX
Article
Macromolecules Scheme 1. Monomer Activation Pathways and Energies (kcal/mol)
evidenced by the characteristic singlet at 10.8 ppm in the 1H NMR spectrum; the other was proposed to be an acyl-imidazole2-ylidenium species. As consequence of the two initiating species formed, two different propagating pathways can occur by monomer addition at the carbamate chain-end. The reaction proceeds in a quasi-living manner, leading to the synthesis of well-defined cyclic poly(α-peptoid)s.28 Intrigued from this quite complex mechanistic scenario, we decided to perform a DFT investigation aimed at clarifying basically all aspects of this polymerization reaction, from initiation, to the general chain growth step, to the possible chain termination mechanisms. This comprehensive understanding could help in the rational design of new initiators for the synthesis of poly(α-peptoid)s with diverse structures and controlled molecular mass.
Solvent effects were included with the PCM model using THF as the solvent.35,36 To these M06/TZVP electronic energies in solvent, zero point energy and thermal corrections were included from the gas-phase frequency calculations at the BP86/SVP level. The thermochemical analysis was performed at p = 1354 atm, rather than at 1 atm, to mimic better the condensed phase.37 We recall that the typical accuracy of energies predict by DFT calculations is within 3−5 kcal/mol of error.38
■
RESULTS AND DISCUSSION We start discussing possible initiation pathways using SIPr as reference NHC. For the sake of simplicity we used a methyl Nsubstituted NCA, rather than a n-butyl N-substituted NCA used experimentally.27,28 Chain Initiation. Starting from the anhydride activation event, we investigated the two competitive scenarios reported in Scheme 1. In the former pathway the NHC adds to a free monomer leading to the zwitterionic adduct 2. This adduct can further evolve through opening of the anhydride ring to the spirocyclic species 2′a or to the ring-open zwitterionic analogue 2′b. The second pathway is definitely more complex, and it starts with the NHC acting as a base extracting a proton from the CH2 moiety of the monomer, leading to the ion-pair 3/[NHC-H]+
■
COMPUTATIONAL DETAILS DFT calculations were performed at the generalized gradient approximation (GGA) level by using the Gaussian 09 package.29 Geometries were optimized with the BP86 functional30 and the SVP basis set.31 The reported free energies were built through single point energy calculations on the BP86/SVP geometries using the M06 functional32,33 and the triple-ζ TZVP basis set.34 B
DOI: 10.1021/acs.macromol.6b01722 Macromolecules XXXX, XXX, XXX−XXX
Article
Macromolecules
initiating species 2′a substantially isoenergetic with the starting species 1. Differently, initiation along the anionic pathway is substantially irreversible, since species 4 is predicted to be 7.7 kcal/mol below the starting species 1. In short, if species 2′a is not trapped by reaction with another monomer molecule, thus starting the growth of a polymer chain, initiation should converge toward the formation of the most stable species 4. Finally, according to our calculations the Munchnone initiating species 6, hypothesized by Zhang and co-workers,28 is thermodynamically less favored than species 4 by nearly 17 kcal/mol, and its formation is kinetically disfavored by roughly 30 kcal/mol. These differences are so large that despite any possible bias in the computational protocol favoring attack via the N atom, we can rule out this initiation pathway. Since all the above reactivity only involves the nonamidic carbonyl group labeled α in Chart 2, for completeness we also explored reactivity at the amidic carbonyl group labeled β in Chart 2. For the zwitterionc initiation pathway of Scheme 1, addition of the NHC molecule to the β carbonyl via transition state 1−2β requires 9.0 kcal/mol more than addition to the α carbonyl via transition state 1−2. For the anionic initiation pathway of Scheme 1, the energetics of the first step, which is the conversion of 1 to 3, remains unchanged since it consists in an H-transfer. We thus focused on the following monomer opening reaction by attack of 3 to the β carbonyl of a free NHC, and we compared the resulting energetics to the low energy attack of 3 to the α carbonyl of a free NHC. According to calculations, attack of 3 at the β carbonyl is kinetically unfavored, since transition state 3− 4β is 15.0 kcal/mol above transition state 3−4 of Scheme 1. On the basis of these results, we can exclude that the opening of the monomer occurs on the β carbon during initiation, and this result can be assumed to hold also during propagation. This result can be rationalized by considering the shape of the lowest unoccupied molecular orbital (LUMO) of the monomer, which shows a high participation of the α carbonyl group (see Figure 1).
followed by the addition of the formed anionic species 3 to another monomer molecule. For completeness, we investigated attack of 3 via the C1, O1, and N atoms, leading to the three possible initiating species 4, 5, and 6, respectively. The energetics of the two initiation pathways is also reported in Scheme 1. In analogy with the work of Hedrick and Waymouth on δ-valerolactone polymerization promoted by NHCs,35 we choose the weakly van der Waals complex 1, between the monomer and the NHC, as reference species at 0 kcal/mol. Along the zwitterionic initiation pathway, nucleophilic attack of the NHC to Me-NCA has to overcome an energy barrier of 7.0 kcal/mol to reach species 2, which presents an elongated ring C− O distance of 1.72 Å and lies 2.4 kcal/mol above the starting species 1. As can be noted from Scheme 1, species 2 preferentially evolves via transition state 2−2′a and a low energy barrier of 5.2 kcal/mol to the open isomer 2′a, 0.4 kcal/mol below the starting species 1. Conversion of 2 to the spirocycle isomer 2′b, 3.9 kcal/ mol above the starting species 1, via transition state 2−2′b at 5.7 kcal/mol, is disfavored both thermodynamically and kinetically. We explored also the possible decarboxylation of 2′a with formation of a 4-member sarcosine based spirocycle. However, this species was ruled out since it is almost 6 kcal/mol higher in energy than 2′b. The overall conclusion is that initiation along the zwitterionic pathway leads to a substantial equilibrium between the starting species 1 and the zwitterionic species 2′a. Moving to the anionic initiation pathway, the H transfer from the monomer to the NHC has an energy barrier of 16.6 kcal/mol and leads to the anionic species 3 interacting with [SIPr-H]+, the overall ion-pair laying 6.9 kcal/mol higher in energy relative to the starting species 1. Comparison of the three reaction pathways branching from 3 (see Figure S1) indicates that attack of 3 to a monomer molecule via the C1 atom with a transition state at 9.9 kcal/mol from 1 is clearly favored over attack via the O1 and N atoms, with transition states at 16.6 and 31.2 kcal/mol, respectively. Further, species 4, formed via attack of the C1 atom, lies 7.7 kcal/mol below 1, making the overall initiation step thermodynamically favored, whereas species 5 and 6, formed via attack of the O1 or N atoms, lie 6.7 and 9.3 kcal/mol above 1 and are thus disfavored relative to both 4 and 1. It is worth to note that transitions states 3−4, 3−5, and 3−6 correspond to a single step transformation with no intermediate identified. This holds also for 3−6, where attack of the N atom of 3 to the Cα atom of a new NCA molecule (see Chart 2) could lead to an addition Chart 2
Figure 1. Lowest unoccupied molecular orbital (LUMO) of Me-NCA.
Chain Propagation. In this section we describe the reaction profile for the propagation steps starting from the initiating species 2′a and 4. In all cases propagation occurs by nucleophilic attack of the carbamate chain-end to the α carbonyl of the monomer, with concerted liberation of CO2 from the chain-end. It is worth to note that no energetically accessible mechanism involving the earlier release of CO2 from the chain-end and the next addition of the new monomer has been found. The main difference between the zwitterionic and the anionic pathways is the presence of the SIPr chain-end starting from 2′a and of the [SIPr-H]+ cation next to the chain-end in the pathway starting from 4. The energy profiles for the complete zwitterionic and anionic pathways (initiation and first propagation step) are reported in Figure 2. We remind that the structure assumed as reference at 0 kcal/mol is complex 1, constituted by a monomer molecule weakly interacting with the NHC.
intermediate preceding CO2 elimination. Attempts to locate this structure always resulted in the starting geometry of the adduct breaking apart into separated 3 and NCA. At this point an overall analysis of the initiation events is possible. The first remark is that both initiation reactions hypothesized in the literature,28 and discussed in Scheme 1, are possible at moderate temperatures. In fact, the zwitterionic pathway requires the overcome of a barrier of 7.0 kcal/mol, while the anionic pathway requires the overcome of a barrier of 16.6 kcal/mol. The clear kinetic preference of 9.9 kcal/mol in favor of the zwitterionic initiation pathway is counterbalanced by the reversibility of the zwitterionic pathway, with the zwitterionic C
DOI: 10.1021/acs.macromol.6b01722 Macromolecules XXXX, XXX, XXX−XXX
Article
Macromolecules
Figure 2. Energy profiles for the initiation and first propagation step.
Figure 3. Structures and key distances of the first propagation product and the second addition transition state.
We comment here only on the propagation step since the initiation pathways have already been described in the previous section. Focusing on the zwitterionic pathway, the energy required to add a new monomer to the propagating species 2′a
and eliminate a CO2 molecule is 22.6 kcal/mol, and the corresponding addition product is 3.8 kcal/mol lower in energy than 2′a. Moving to the anionic pathway, the energy barrier for the first propagation step starting from species 4 has a barrier of D
DOI: 10.1021/acs.macromol.6b01722 Macromolecules XXXX, XXX, XXX−XXX
Article
Macromolecules
profiles for NHCs having different steric and electronic properties (see Chart 3 and Table 1).
23.4 kcal/mol, remarkably similar to that along the zwitterionic pathway, despite the different interaction between the anionic chain-end and either the cationic NHC ring or the [NHC-H]+ ion. For the sake of completeness, we calculated also a second chain growth step. Addition of the second monomer along the zwitterionic pathway has an energy barrier of 25.4 kcal/mol, similar to that for addition of the first monomer molecule, while the stability of the product increases, with an energy gain of 10.6 kcal/mol with respect to the first addition product. The reduced exothermicity of the first addition can be explained considering that the electrostatic interaction between the chain-ends in 2′a requires the growing chain assuming a quite stable conformation corresponding to a relaxed 6-membered ring, while the same interaction in the first product requires the closure of a less favored 9-membered ring. Moving to the second propagation step along the anionic pathway, it has a barrier of 22.5 kcal/mol, and it is again more exothermic than the first addition step, by roughly 7 kcal/mol thanks to the more relaxed conformation the longer chain can assume (see below for the structure of the growing chains). Overall, these results are also in quite good agreement with the experimentally estimated activation energy of 26 kcal/mol.28 To better rationalize the role of the NHC in both propagation pathways, a structural comparison of the first propagation product and of the second addition transition state is reported in Figure 3. The propagating species along the zwitterionic pathway maintains a pseudocyclic zwitterionic geometry, with the two charged chain-ends in close contact due to Coulombic attraction and with a distance of only 2.35 Å between the negatively charged oxygen of the carbamate chain-end and the positively charged carbon atom of the NHC moiety. In the propagation transition state, the carboxylate chain-end moiety points away from the NHC, basically resulting in a CO2 molecule being removed from the chain-end, allowing an interaction of the N atom of the chainend with the incoming monomer. Conversely, the propagating species along the anionic pathway shows a more open structure with the carboxylate chain-end interacting with the NHC-H+ moiety through a hydrogen bond of 1.71 Å. It is worth to note that the growing chain assumes a folded conformation, rather than a linear all-trans conformation, in line with the experimental evidence that polysarcosine assumes a random coil conformation.39 In the corresponding propagation transition state, the oxygen atoms of both the CO2 moiety at the carbamate chainend and of the CO2 group of the new monomer are found at short distances from the hydrogen atom of the NHC−H+, 2.28− 2.35 Å. Also in this transition state the CO2 moiety at the carbamate chain-end basically corresponds to a CO2 molecule being removed from the chain-end. Moreover, as consequence of the chain-end, the two propagating transition states are structurally quite different: in the zwitterionic pathway the N− C forming bond of 2.08 Å and the C−O ring-opening bond of 1.43 Å resemble the reactants; conversely, in the anionic pathway, the N−C forming bond is already 1.51 Å and the ringopening bond is 2.23 Å, reflecting the products features. The role of the NHC moiety is clearly crucial in both pathways: in addition to initiating the chain growth, the NHC also facilitates the chain propagation by mediating monomer addition through counterion effects. Catalyst Structure Effect. As pointed out in the previous section, the NHC structure is expected to influence the monomer activation and the chain propagation rate whatever mechanism occurs. In the following, we investigate the energy
Chart 3
Table 1. Free Energies (kcal/mol in THF) of Species Involved in the Initiation and First Propagation Step for the NHCs in Chart 3 1 1−2 2′a 2′a−P1 P1 1−3 3 3−4 4 4−P2 P2
SIPr
SIMes
StBu
SMe
IPr
0.0 7.0 −0.4 22.2 −4.2 16.6 6.9 9.9 −7.7 15.7 −13.8
0.0 6.6 −5.2 23.9 −3.7 14.4 7.2 6.8 −8.2 14.9 −14.4
0.0 26.6 8.9 30.9 5.0 15.2 1.3 5.0 −11.8 15.2 −17.1
0.0 8.4 −9.3 19.5 −4.7 13.7 8.8 6.4 −8.6 15.9 −13.3
0.0 6.5 −2.1 19.7 −10.1 14.4 10.9 9.4 −6.9 15.6 −15.3
a P1 and P2 = first propagation product along the zwitterionic and the anionic pathways, respectively. The energy of 2−2′a has been omitted because this step is practically barrierless for all the NHCs considered.
Focusing on initiation steps 1−2a′ along the zwitterionic pathway 1 and 1−4 along the anionic pathway, analysis of the values reported in Table 1 indicates that that SIMes and SMe behave similarly to SIPr, with transition state 1−2 lower in energy by 6−9 kcal/mol with respect to transition state 1−3. Moreover, since the zwitterionic initiating species 2′a is clearly more stable than the initial species 1 for SIMes and SMe, while it is isoenergetic with 1 for SIPr, initiation along the zwitterionic pathway is thermodynamically preferred for SIMes and SMe. On the contrary, initiation along the zwitterionic pathway can be ruled out for StBu being transition state 1−2 almost 10 kcal/mol higher in energy with respect to transition state 1−3 one. This result is in agreement with the experimental finding that only the NMR signal assigned to the [NHC−H]+ species is found when the StBu NHC is involved. Analysis of the electronic and steric of the four NHCs considered in this work allows a better understanding of the impact of the NHC structure in discriminating between the two competing polymerization mechanisms. The calculated electrophilicity index, ω (see Supporting Information),40−44 shows that the nucleophilic character is decreasing in the order StBu > SMe > SIMes > SIPr. On the contrary, the steric hindrance increases from SMe to StBu with SIMes smaller than SIPr, as indicated by the values of the buried volume, %VBur (see Supporting E
DOI: 10.1021/acs.macromol.6b01722 Macromolecules XXXX, XXX, XXX−XXX
Article
Macromolecules Scheme 2. Termination Pathways and Energies (kcal/mol)
Information).45−47 Focusing on the energy of species along the zwitterionic pathway in Table 1, both the initiation and the propagation mechanism seem to correlate better with the %VBur of the NHC, with SIPr, SIMes, and SMe showing much lower energy barriers and more stable intermediates with respect to StBu. These results point out the steric hindrance of the NHCs as the key factor in determining the energetics along the zwitterionic pathway. As a consequence, for StBu initiation along the zwitterionic pathway is not favored, despite being the most nucleophilic NHC. Conversely, the data relative to the anionic pathway in Table 1 suggest that the steric hindrance of the NHC has a less relevant role. This result is strictly connected to the mechanism occurring since the NHC is not directly bonded to the growing chain along the anionic pathway: the catalyst approaches to the monomer to abstract the hydrogen in the 1−3 step; after the proton transfer, the [NHC−H]+ moiety is not directly connect to the main chain, and its steric hindrance has minor impact on the chain growth. For the sake of completeness, we also investigated the unsaturated NHC IPr of Chart 3. The whole scenario is not meaningfully affected by the difference in the backbone; compare SIPr with IPr in Table 1. Species along the zwitterionic pathway result to be thermodynamically more stable with IPr, since the greater delocalization of the π-system in the unsaturated heterocycle stabilizes more the positive charge on the heterocycle ring in the zwitterionic species. Conversely, species 3 along the
anionic pathway is less stable with IPr respect to SIPr as consequence of the stronger σ-donor behavior of saturated NHCs. Chain Termination. To shed light on the possible termination mechanism that could compete with chain growth, we have investigated the pathways reported in Scheme 2. We have considered as starting point the polymeric chain with two monomer units, and for straightforward comparison, we have reported also the energetics of chain growth. Focusing on the propagating species P3 along the zwitterionic polymerization pathway, it can propagate via transition state P3−P5 and an energy barrier of 27.6 kcal/mol to the next propagation intermediate P5. In competition, it can eliminate CO2 along three possible pathways. The first, via transition state P3−T1 leading to the cyclic polypeptoid species T1, also liberates the NHC molecule. The second, via transition state P3−D1 and concerted attack of the N atom of the chain-end to the CO moiety bound to the NHC, leading to the cyclic zwitterion species D1. The third, via transition state P3−D2 and simultaneous attack of the N atom of the chain-end to the carbenic C atom of the NHC, leading to the spirocyclic zwitterion species D2. Intermediates D1 and D2 can be considered more as dormient intermediates, since the NHC moiety is still incorporated in the polypeptoid chain, and polymerization can be in principle restarted by reactivity of intermediates D1 and D2 with another monomer molecule. F
DOI: 10.1021/acs.macromol.6b01722 Macromolecules XXXX, XXX, XXX−XXX
Article
Macromolecules
studied are much higher in energy relative to the chain propagation. Comparison between four NHCs with different steric and electronic properties pointed out that the steric hindrance of the NHC is the main factor impacting the energetics of the zwitterionic pathway and also indicated that electronic properties play an important role in balancing the steric hindrance of the NHC along the anionic pathway.
These three termination pathways have energy barriers in the range of 39−44 kcal/mol, which makes them clearly unfavored relative to chain propagation. Moving to the propagating species P4 along the anionic polymerization pathway, it can propagate via transition state P4−P6 and an energy barrier of 28.9 kcal/mol to the next propagation intermediate P6. In competition, it can terminate via CO2 elimination and transition state P4−T2, with release of a N-carboxyanhydride anion. Also, this termination pathway, with an energy barrier of 40.1 kcal/mol, is clearly disfavored relative to chain propagation. The high energy barrier predicted for all the termination reactions considered are consistent with the quasi-living behavior experimentally observed. Nevertheless, based on MALDI-TOF MS experiments, it was hypothesized that among the possible termination reactions, the one leading to formation of spirocycle structures via CO2 elimination, such as species D2 in Scheme 2, should be favored. However, our calculations indicate that formation of species D2 is unlikely for steric reasons. In fact, despite the carbene carbon atom is calculated to be slightly more nucleophilic than the adjacent acyl carbon, the short distances between the Nsubstituent of the heterocycle and the Me substituent on the N atom on the cyclic chain in D2 destabilize this species. Instead, our calculations suggest the formation of the zwitterionic species D1, also reported in Scheme 2, since its formation is favored kinetically, transition state P3−D1 is 4.9 kcal/mol lower in energy than transition state P3−D2, and thermodynamically, species D1 is 6.0 kcal/mol more stable than the spirocyclic species D2.
■
ASSOCIATED CONTENT
S Supporting Information *
The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.macromol.6b01722. Cartesian coordinates of all the species discussed in this work (PDF)
■
AUTHOR INFORMATION
Corresponding Author
*E-mail:
[email protected] (L.C.). Notes
The authors declare no competing financial interest.
■ ■
ACKNOWLEDGMENTS This work was supported by funding from King Abdullah University of Science and Technology (KAUST).
■
REFERENCES
(1) Arduengo, A. J.; Harlow, R. L.; Kline, M. A Stable Crystalline Carbene. J. Am. Chem. Soc. 1991, 113, 361. (2) Huang, J.; Stevens, E. D.; Nolan, S. P.; Petersen, J. L. Olefin Metathesis-Active Ruthenium Complexes Bearing a Nucleophilic Carbene Ligand. J. Am. Chem. Soc. 1999, 121, 2674. (3) Vyboishchikov, S. F.; Buhl, M.; Thiel, W. Mechanism of Olefin Metathesis with Catalysis by Ruthenium Carbene Complexes: Density Functional Studies on Model Systems. Chem. - Eur. J. 2002, 8, 3962. (4) Scholl, M.; Ding, S.; Lee, C. W.; Grubbs, R. H. Synthesis and activity of a new generation of ruthenium-based olefin metathesis catalysts coordinated with 1,3-dimesityl-4,5-dihydroimidazol-2-ylidene ligands. Org. Lett. 1999, 1, 953. (5) Grubbs, R. H. Handbook of Olefin Metathesis; Wiley-VCH: Weinheim, Germany, 2003. (6) Assen, E.; Kantchev, B.; O’Brien, C. J.; Organ, M. G. Palladium Complexes of N-Heterocyclic Carbenes as Catalysts for Cross-Coupling Reactions - A Synthetic Chemist’s Perspective. Angew. Chem., Int. Ed. 2007, 46, 2768. (7) Miyaura, N.; Suzuki, A. Palladium-Catalyzed Cross-Coupling Reactions of Organoboron Compounds. Chem. Rev. 1995, 95, 2457. (8) Diez-Gonzalez, S.; Marion, N.; Nolan, S. P. N-Heterocyclic Carbenes in Late Transition Metal Catalysis. Chem. Rev. 2009, 109, 3612. (9) Diez-Gonzalez, S.; Nolan, S. P. Stereoelectronic parameters associated with N-heterocyclic carbene (NHC) ligands: A quest for understanding. Coord. Chem. Rev. 2007, 251, 874. (10) Dorta, R.; Stevens, E. D.; Scott, N. M.; Costabile, C.; Cavallo, L.; Hoff, C. D.; Nolan, S. P. Steric and Electronic Properties of NHeterocyclic Carbenes (NHC): A Detailed Study on Their Interaction with Ni(CO)4. J. Am. Chem. Soc. 2005, 127, 2485. (11) Kelly, R. A., III; Clavier, H.; Giudice, S.; Scott, N. M.; Stevens, E. D.; Bordner, J.; Samardjiev, I.; Hoff, C. D.; Cavallo, L.; Nolan, S. P. Determination of N-Heterocyclic Carbene (NHC) Steric and Electronic Parameters using the [(NHC)Ir(CO)2Cl] System. Organometallics 2008, 27, 202. (12) Jacobsen, H.; Correa, A.; Poater, A.; Costabile, C.; Cavallo, L. Understanding the M-(NHC) (NHC = N-heterocyclic carbene) bond. Coord. Chem. Rev. 2008, 253, 2784.
CONCLUSIONS We have reported a DFT mechanistic investigation of initiation, growth, and termination reactions in the organopolymerization of N-substituted N-carboxyanhydrides promoted by NHCs. Our calculations support the experimental evidence that initiation can occur via two mechanisms in competition: the zwitterionic mechanism, consisting in the addition of the NHC to a free monomer generating a zwitterionic adduct, and the anionic mechanism, consisting in the abstraction of a proton from the CH2 moiety of the monomer by the NHC. Initiation along the anionic pathway generates an ion pair ready to activate a new monomer via addition to the carbon atom and CO2 elimination. The alternative mechanism consisting in the formation of a Munchnone initiating species via addition of the new monomer to the N atom can be ruled out, since it is energetically inaccessible. Both initiation mechanisms are possible at moderate temperatures. However, the zwitterionic mechanism is kinetically preferred but thermodynamically reversible, while the anionic mechanism is favored thermodynamically. Moving to the chain growth, propagation occurs by nucleophilic attack of the carbamate chain-end to the α carbonyl of the monomer, with concerted liberation of CO2 from the chain-end. No energetically accessible mechanism involving the earlier release of CO2 from the chain-end and the next addition of the new monomer has been found. The main difference between the zwitterionic and the anionic propagation pathways is the presence of the NHC as initiating group at the beginning of chain-end in the case of zwitterionic pathway and of the [NHC− H]+ cation next to the anionic chain-end in the case of the anionic pathway. The two chain elongation pathways have similar barrier, consistent with the experimental monomodal distribution of the molecular mass of the achieved polymers. In agreement with the quasi-living experimental behavior, all the termination pathways G
DOI: 10.1021/acs.macromol.6b01722 Macromolecules XXXX, XXX, XXX−XXX
Article
Macromolecules
(33) Zhao, H.; Dang, L.; Marder, T. B.; Lin, Z. DFT Studies on the Mechanism of the Diboration of Aldehydes Catalyzed by Copper(I) Boryl Complexes. J. Am. Chem. Soc. 2008, 130, 5586. (34) Schaefer, A.; Huber, C.; Ahlrichs, R. ully optimized contracted Gaussian-basis sets of triple zeta valence quality for atoms Li to Kr. J. Chem. Phys. 1994, 100, 5829. (35) Tomasi, J.; Mennucci, B.; Cammi, R. Quantum mechanical continuum solvation models. Chem. Rev. 2005, 105, 2999. (36) Barone, V.; Cossi, M. Quantum calculation of molecular energies and energy gradients in solution by a conductor solvent model. J. Phys. Chem. A 1998, 102, 1995. (37) Martin, R. L.; Pratt, P. J. Hydrolysis of Ferric Ion in Water and Conformational Equilibrium. J. Phys. Chem. A 1998, 102, 3565. (38) Peverati, R.; Truhlar, D. G. Quest for a universal density functional: the accuracy of density functionals across a broad spectrum of databases in chemistry and physics. Philos. Trans. R. Soc., A 2014, 372, 20120476. (39) Gangloff, N.; Luxenhofer, R. Peptoids for Biomimetic Hierarchical Structures. Adv. Polym. Sci. 2013, 262, 389. (40) Parr, R. G.; Szentpály, L. v.; Liu, S. Electrophilicity index. J. Am. Chem. Soc. 1999, 121, 1922. (41) Geerlings, P.; De Proft, F.; Langenaeker, W. Conceptual density functional theory. Chem. Rev. 2003, 103, 1793. (42) Parr, R. G.; Pearson, R. G. Absolute hardness: companion parameter to absolute electronegativity. J. Am. Chem. Soc. 1983, 105, 7512. (43) Koopmans, T. Ordering of wave functions and eigenenergies to the individual electrons of an atom. Physica 1934, 1, 104. (44) Parr, R. G.; Donnelly, R. A.; Levy, M.; Palke, W. E. Electronegativity: the density functional viewpoint. J. Chem. Phys. 1977, 68, 3801. (45) Poater, A.; Cosenza, B.; Correa, A.; Giudice, S.; Ragone, F.; Scarano, V.; Cavallo, L. SambVca: A Web Application for the Calculation of the Buried Volume of N-Heterocyclic Carbene Ligands. Eur. J. Inorg. Chem. 2009, 2009, 1759. (46) Poater, A.; Cavallo, L. comparing families of olefin polymerization precatalysts using the percentage of buried volume. Dalton Trans. 2009, 8885. (47) Clavier; Caijo, F.; Borré, E.; Rix, D.; Boeda, F.; Nolan, S. P.; Mauduit, M. Towards Long-Living Metathesis Catalysts by Tuning the N-Heterocyclic Carbene (NHC) Ligand on TrifluoroacetamideActivated Boomerang Ru Complexes. Eur. J. Org. Chem. 2009, 25, 4254.
(13) Culkin, D. A.; Jeong, W.; Csihony, S.; Gomez, E. D.; Balsara, N. P.; Hedrick, J. L.; Waymouth, R. M. Zwitterionic Polymerization of Lactide to Cyclic Poly (Lactide) by Using N-Heterocyclic Carbene Organocatalysts. Angew. Chem., Int. Ed. 2007, 46, 2627. (14) Jeong, W.; Shin, E. J.; Culkin, D. A.; Hedrick, J. L.; Waymouth, R. M. Zwitterionic polymerization: a kinetic strategy for the controlled synthesis of cyclic polylactide. J. Am. Chem. Soc. 2009, 131, 4884. (15) Shin, E. J.; Brown, H. A.; Gonzalez, S.; Jeong, W.; Hedrick, J. L.; Waymouth, R. M. Zwitterionic copolymerization: synthesis of cyclic gradient copolymers. Angew. Chem., Int. Ed. 2011, 50, 6388. (16) Jeong, W.; Hedrick, J. L.; Waymouth, R. M. Organic spirocyclic initiators for the ring-expansion polymerization of β-lactones. J. Am. Chem. Soc. 2007, 129, 8414. (17) Acharya, A. K.; Chang, Y. A.; Jones, G. O.; Rice, J. E.; Hedrick, J. L.; Horn, H. W.; Waymouth, R. M. Experimental and Computational Studies on the Mechanism of Zwitterionic Ring-Opening Polymerization of Δ-Valerolactone with N-Heterocyclic Carbenes. J. Phys. Chem. B 2014, 118, 6553. (18) Raynaud, J.; Absalon, C.; Gnanou, Y.; Taton, D. N-Heterocyclic Carbene-Induced Zwitterionic Ring-Opening Polymerization of Ethylene Oxide and Direct Synthesis of α,ω-Difunctionalized Poly(ethylene oxide)s and Poly(ethylene oxide)-b-poly(ε-caprolactone) Block Copolymers. J. Am. Chem. Soc. 2009, 131, 3201. (19) Raynaud, J.; Absalon, C.; Gnanou, Y.; Taton, D. N-Heterocyclic Carbene-Organocatalyzed Ring-Opening Polymerization of Ethylene Oxide in the Presence of Alcohols or Trimethylsilyl Nucleophiles as Chain Moderators for the Synthesis of α,ω-Heterodifunctionalized Poly(ethylene oxide)s. Macromolecules 2010, 43, 2814. (20) Zhang, Y.; Chen, E. Y. X. Conjugate-Addition Organopolymerization: Rapid Production of Acrylic Bioplastics by NHeterocyclic Carbenes. Angew. Chem., Int. Ed. 2012, 51, 2465. (21) Hong, M.; Chen, E. Y. X. Proton-Transfer Polymerization (HTP): Converting Methacrylates to Polyesters by an N-Heterocyclic Carbene. Angew. Chem., Int. Ed. 2014, 53, 11900. (22) Zhang, Y.; Schmitt, M.; Falivene, L.; Caporaso, L.; Cavallo, L.; Chen, E. Y. X. Organocatalytic Conjugate-Addition Polymerization of Linear and Cyclic Acrylic Monomers by N-Heterocyclic Carbenes: Mechanisms of Chain Initiation, Propagation, and Termination. J. Am. Chem. Soc. 2013, 135, 17925. (23) Zhang, Y.; Miyake, G. M.; John, M. G.; Falivene, L.; Caporaso, L.; Cavallo, L.; Chen, E. Y. X. Lewis pair polymerization by classical and frustrated Lewis pairs: acid, base and monomer scope and polymerization mechanism. Dalton Trans. 2012, 41, 9119. (24) Zhang, Y.; Miyake, G. M.; Chen, E. Y. X. Alane-Based Classical and Frustrated Lewis Pairs in PolymerSynthesis: Rapid Polymerization of MMA and Naturally RenewableMethylene Butyrolactones into HighMolecular-Weight Polymers. Angew. Chem., Int. Ed. 2010, 49, 10158. (25) Stephan, D. W.; Erker, G. Frustrated Lewis Pairs: Metal-free Hydrogen Activation and More. Angew. Chem., Int. Ed. 2010, 49, 46. (26) Stephan, D. W.; Erker, G. Frustrated Lewis pair chemistry: development and perspectives. Angew. Chem., Int. Ed. 2015, 54, 6400. (27) Guo, L.; Zhang, D. Cyclic poly (α-peptoid) s and their block copolymers from N-heterocyclic carbene-mediated ring-opening polymerizations of N-substituted N-carboxylanhydrides. J. Am. Chem. Soc. 2009, 131, 18072. (28) Guo, L.; Lahasky, S. H.; Ghale, K.; Zhang, D. N-Heterocyclic Carbene-Mediated Zwitterionic Polymerization of N-Substituted NCarboxyanhydrides toward Poly(α-peptoid)s: Kinetic, Mechanism, and Architectural Control. J. Am. Chem. Soc. 2012, 134, 9163. (29) Gaussian09; Gaussian, Inc.: Pittsburgh, PA. (30) Becke, A. D. Density-functional exchange-energy approximation with correct asymptotic behavior. Phys. Rev. A: At., Mol., Opt. Phys. 1988, 38, 3098. (31) Schaefer, A.; Horn, H.; Ahlrichs, R. Fully optimized contracted Gaussian basis sets for atoms Li to Kr. J. Chem. Phys. 1992, 97, 2571. (32) Zhao, H.; Lin, Z.; Marder, T. B. Density Functional Theory Studies on the Mechanism of the Reduction of CO2 to CO Catalyzed by Copper(I) Boryl Complexes. J. Am. Chem. Soc. 2006, 128, 15637. H
DOI: 10.1021/acs.macromol.6b01722 Macromolecules XXXX, XXX, XXX−XXX
