ISSN 1560-0904, Polymer Science, Series B, 2017, Vol. 59, No. 2, pp. 147–156. © Pleiades Publishing, Ltd., 2017. Original Russian Text © I.E. Nifant’ev, P.V. Ivchenko, A.V. Shlyakhtin, A.V. Ivanyuk, 2017, published in Vysokomolekulyarnye Soedineniya, Seriya B, 2017, Vol. 59, No. 2, pp. 124–134.
CATALYSIS
Polymerization of Trimethylene Carbonate and Lactones in the Presence of Magnesium Monoionolate: A Comparative Theoretical and Experimental Study I. E. Nifant’eva,b, P. V. Ivchenkoa,b,*, A. V. Shlyakhtina, and A. V. Ivanyukb aFaculty
b
of Chemistry, Moscow State University, Moscow, 119991 Russia Topchiev Institute of Petrochemical Synthesis, Russian Academy of Sciences, Leninskii pr. 29, Moscow, 119991 Russia *е-mail:
[email protected] Received July 14, 2016; Revised Manuscript Received November 24, 2016
Abstract—Using quantum-chemical calculations (DFT, program Priroda), the formation of a catalyst species on the basis of magnesium complexes with 2,6-di-tert-butyl-4-methylphenol is discussed. A comparative theoretical and experimental study of the ring-opening polymerization of trimethylene carbonate, 1,4-dioxanone, δ-valerolactone, and ε-caprolactone in the presence of the monoionolate magnesium complex is performed. It is shown that the calculated values of activation barriers correlate with the observed order of activity of cyclic esters. The maximum rate of polymerization is exhibited by trimethylene carbonate. DOI: 10.1134/S1560090417020075
INTRODUCTION The design, synthesis, and study of properties of biodegradable and ecologically friendly polymeric materials constitute the domain of science that has been extensively developed during the two past decades at the boundary of macromolecular chemistry, coordination and organometallic chemistry, theory of catalysis, biochemistry, and medicine. Synthetic aliphatic polyesters obtained by ring-opening polymerization (Scheme (1)) attract the greatest attention of researchers [1–3]. This is associated with both the relative accessibility of original compounds, namely, cyclic esters, which may be isolated from renewable raw materials or synthesized with the use of biotechnologies, and the compositional variability of polymers, a wide scope of cyclic substrates, reaction mechanisms of ring-opening polymerization, and the used catalytic systems [4–8]: O
O ROH cat
O
RO
O H (1) n
The ring-opening polymerization of cyclic esters is catalyzed by various types of compounds. At present, the application of organocatalysis, enzymatic catalysis, and catalysis by coordination compounds of active metals is a topical problem [9–16]. A high catalytic activity is demonstrated, in particular, by the com-
plexes of magnesium with ionol (2,6-di-tert-butyl-4methylphenol, Butylated HydroxyToluene, BHT-H) of composition (BHT)2MgL2, where L is a donor ligand (tetrahydrofuran, dimethoxyethane, etc.) [17–20]. The ring-opening polymerization of ester substrates catalyzed by phenolates of active metals is currently assumed to proceed via two main mechanisms [20–22]: the monomer activation mechanism and the coordination–insertion mechanism. The key stage of both mechanisms is the nucleophilic attack at the carbonyl carbon atom of cyclic ester coordinated to the metal atom. The mechanisms differ in the nature of nucleophile: if the reaction occurs via the monomeractivation mechanism, a nucleophile is an alcohol molecule, while in the case of the coordination–insertion mechanism, an alkoxy group covalently bound to the metal atom serves as a nucleophile (Scheme (2)). Ring opening via the former mechanism necessarily occurs through formation of a complex, in which an alcoholic group is not coordinated to the metal atom, and for the reaction to continue, dissociation of the metal complex with the ester group followed by the back coordination of the alcoholic group is required. Such a reaction pathway is in disagreement with the experimentally observed living character of ringopening polymerization in the presence of active metal-complex catalysts which affords polymers with low polydispersity coefficients (Mw/Mn = 1.1–1.5). In the case of these systems, the coordination–insertion mechanism including the direct nucleophilic attack of
147
148
NIFANT’EV et al.
the alkoxy group bound to the metal atom at the carbonyl carbon atom of coordinated cyclic ester seems to be more probable. As a result of this attack, a catalytically active alkoxy complex is immediately formed. The
Ln M
O O
H O
H
O
O O
LnM
O
O
O
O
H
LnM
R
O
O
O
...
O
O
O R
H R
O
R O
O LnM
O
H
H O
R
Monomer-activation mechanism
main side process during this living polymerization includes the formation of cyclic oligomeric products which can be effectively suppressed using complexes with sterically crowded ligands, for example, BHT:
R
LnM
O
Ln M
O
O
(2) O
R
Сoordination-insertion mechanism
LnM
O
O LnM
O
O
R
O
LnM
O
O O
O
R
O
LnM
O O O
O
O
R
O
R O
O
LnM
The mechanisms of ring-opening polymerization of cyclic ester substrates catalyzed by complex compounds of active metals have been extensively studied both theoretically and experimentally. Practically all studies devoted to the quantum-chemical simulation of ring-opening polymerization accept the coordination–insertion mechanism as the main one for various classes of complexes: carboxylates and alkoxy complexes of Sn(II) and alkoxy derivatives of Zn, Al, In, rare-earth metals, Cr, and U [23–47]. In a single paper devoted to quantum-chemical simulation with participation of magnesium complexes [48], the mechanism of the ring-opening polymerization of lactide is described, which differs appreciably from the mechanism of the ring-opening polymerization of lactones and cyclic carbonates. Thus, the ring-opening polymerization of lactones and cyclic carbonates catalyzed by complex magnesium compounds has hardly been studied to date from the viewpoint of mechanics. Recently [49], the mono-BHT complexes of Mg were synthesized and characterized. It was shown that, depending on the structural type (alkyl- or alkoxy-, ArO–Mg–R vs. ArO–Mg–OR), these complexes even in the presence of donor ligand THF have different degrees of association: in the presence of excess of tetrahydrofuran, alkyl complexes crystallize as monomers of composition (BHT)MgR(THF)2, while for alkoxy complexes the dimeric structure,
O
O LnM
O
O
R
...
O
[(BHT)Mg(μ-OR)(THF)]2, is more preferable. It was found that mono-BHT complexes compared with their analogs di-BHT are characterized by a higher catalytic activity with respect to various cyclic esters and reactions induced by the dimeric complex [(BHT)Mg(μ-OEt)(THF)]2 proceed with a marked induction period or need thermal activation. The authors of [49] assumed that catalyst species in all cases are monomeric mono-BHT alcoholates of magnesium and suggested pathways of their formation. The goals of this study are to perform the quantumchemical investigation of pathways for the formation of mono-BHT complexes of magnesium and for the ring-opening polymerization of trimethylene carbonate (TMC), 1,4-dioxanone (PDO), δ-valerolactone (δ-VL), and ε-caprolactone (ε-CL) catalyzed by these complexes and to experimentally verify the simulation results by polymerization experiments. In fact, this is the first theoretical study of the mechanism of polymerization of cyclic esters catalyzed by the alkoxy complexes of magnesium. EXPERIMENTAL General Remarks Commercially available monomers TMC, PDO, δ-VL, and ε-CL were distilled in vacuum and stored under argon. Toluene and tetrahydrofuran used as sol-
POLYMER SCIENCE, SERIES B
Vol. 59
No. 2
2017
POLYMERIZATION OF TRIMETHYLENE CARBONATE AND LACTONES
vents were boiled and distilled over sodium in the presence of benzophenone. Polymerization was conducted in thoroughly dried glass vessels equipped with a septa for loading reagents. The NMR spectra of sampled reaction mixtures and isolated polymers were registered on an Avance Bruker (400 MHz) instrument in CDCl3 using CH2Cl2 (δH = 5.3) as an internal standard. Polymer samples were analyzed on an Agilent PL-GPC 220 gel-permeation chromatograph equipped with a PLgel Olexis column (a nominal particle size of 13 μm, a maximum temperature of 220°C, and a molecular-mass range of 2 × 10 4–1 × 107); tetrahydrofuran (1 mL/min) was used as an eluent. Measurements were performed at a temperature of 40°C, and polystyrene was used as a standard. Quantum-Chemical Calculations All calculations were performed in terms of the density functional theory (DFT) in the Priroda program [50] using the Perdew−Burke−Ernzerhof (PBE) functional [51] and the 3ζ basis. Optimization of the geometry of complexes, frequency analysis, and calculation of entropy correction were performed for the gas phase at a temperature of 298.15 K. Intermediates (local minima) and transition states (first-order saddle points) were found by optimizing structures with respect to energy and calculating force-constant matrices (Hessians) followed by analysis of the obtained frequency set (free energy minimum and absence of imaginary frequencies for intermediates, the first-order saddle point, and the single imaginary frequency for transition states). The true character of transition states was confirmed by scanning over the reaction coordinate with a step of 1 × 10–2 Å upon variation in geometric parameters. Experiments on the Polymerization of Cyclic Substrates Experimental technique. To a solution of cyclic substrate (1 mol/L, 7 mL, 7 mmol) in toluene, a solution of (BHT)MgBu (0.35 mol/L, 0.2 mL, 0.07 mmol) in tetrahydrofuran and benzyl alcohol (7.2 μL, 0.07 mmol) used as an initiator were added. The reaction mixtures (50 μL) were sampled through the septa, dissolved in CDCl3 (600 μL), and analyzed by 1H NMR spectroscopy. The ratio between the polymeric products and the unreacted substrate was determined by integrating triplet signals of –OCH2– fragments having chemical shifts (ppm) δ = 4.41 (monomer) and 4.35 (polymer) for TMC, δ = 4.37 (monomer) and 4.17 (polymer) for δ-VL, and δ = 4.19 (monomer) and 4.14 ppm (polymer) for ε-CL. In the case of PDO, the precipitate formed after 1 min was filtered off, washed with pentane, and dried in vacuum. The yield of the product was determined by weighing. After reaction for 5 min, the polymers POLYMER SCIENCE, SERIES B
Vol. 59
No. 2
2017
149
obtained from TMC, δ-VL, and ε-CL were precipitated via addition of diethyl ether (50 mL), filtered off, and dried in vacuum. Poly(trimethylene carbonate). Yield of the product, 676 mg (95%). 1H NMR (CDCl3, 25°C): δ 2.07 (quint, 3J = 6.0 Hz, 2H), 4.35 (t, 3J = 6.0 Hz, 4H) ppm; Mn = 19200 and Mw/Mn = 1.31 (GPC). Poly(δ-valerolactone). Yield of the product, 652 mg (93%). 1H NMR (CDCl3, 25°C): δ 1.76 (br, 4H), 2.44 (br, 2H), 4.17 (t, 3J = 5.8 Hz, 2H) ppm; Mn = 16900 and Mw/Mn = 1.44 (GPC). Poly(ε-caprolactone). Yield of the product, 744 mg (93%). 1H NMR (CDCl3, 25°C): δ 1.44 (quint, 3J = 8.3 Hz, 2H), 1.67–1.76 (m, 4H), 2.37 (t, 3J = 7.3 Hz, 2H), 4.14 (t, 3J = 6.9 Hz, 2H) ppm; Mn = 18000 and Mw/Mn = 1.39 (GPC). RESULTS AND DISCUSSION Formation of a Catalyst Species Geometry, relative stability and dissociation of the dimeric forms of mono-BHT complexes. According to X-ray diffraction [49, 52, 53], (BHT)MgR and (BHT)Mg(OR') are of dimeric nature in the absence of solvating solvents or substrate molecules capable of solvation. In the former case, BHT plays the role of the bridging fragment and the complex corresponds to formula [(μ-BHT)MgR]2, while in the latter case the alkoxy group serves as a bridging group and the dimer [(BHT)Mg(μ-OR')]2 is formed. In the presence of the solvating solvent tetrahydrofuran, the alkyl complex forms monomeric solvate (BHT)Mg(THF)2R (R = Bu) [49]. This solvate, when activated by alcohol, efficiently catalyzes the ring-opening polymerization of various substrates. The dimeric structure of the alkoxy complex is preserved during solvation, and [(BHT)Mg(THF)(μ-OR')]2 (R' = Et) is formed [49]. Ring-opening polymerization in the presence of this complex needs thermal activation. These facts with consideration for the previously published data [17– 20, 30, 36, 38, 39, 48] made it possible to state that the catalyst species are monomeric complexes of composition (BHT)Mg(S)2OR', where S is the molecule of a solvating solvent or substrate, cyclic ester. In the present study, pathways of formation of similar species were examined using quantum-chemical calculations (DFT). To this end, we calculated the free energy of model dimeric complexes [(μBHT)MgMe]2 (1) and [(BHT)Mg(μ-OMe)]2 (2); the products of interaction of 1 and 2 with tetrahydrofuran and substrate, trimethylene carbonate, of composition [(μ-BHT)(S)MgMe]2 (3) and [(BHT)(S)Mg(μ-
150
NIFANT’EV et al.
G, kJ/mol
G, kJ/mol TMC THF
Dissociation Solvation
BHT
8
S
Mg
Mg
O
6
BHT
100
BHT
Mg
8 5
10
3 9
7 33.1
10
22.6
15.9
7
3
BHT
S
BHT
BHT
Mg
0
4 −20.5 −21.8
MeOH
O
S
S
BHT
BHT Mg O
BHT Mg
Mg
1
2
O
S S
Mg
Mg Mg
9
S
BHT
BHT
S
BHT Mg
S Mg
50
57.3
43.1
23.4
21.3
1
0
59.8
52.3
49.0 49.8
−50
100
100.0
5
50
O
6
1 → 3 → 7 → 9 → 10
4
Mg BHT
2
Mg BHT O
O S
Fig. 1. Formation of catalyst species, monomeric complexes (BHT)Mg(S)2OMe (10), on the basis of dimeric mono-BHT complexes with allowance for change in free energy during dissociation and solvation of dimeric complexes [(μ-BHT)MgMe]2 (1) and [(BHT)Mg(μ-OMe)]2 (2). The values of ΔG are given per 1 mole for 5–10.
OMe)]2 (4); and monomeric complexes 5–10 having various degrees of solvation. An analysis of the calculation data (Fig. 1) makes it possible to the make the following generalizations. First, the dissociation of nonsolvated dimers 1 and 2 in the absence of the solvating agent requires thermal activation. In this case, the consumption of energy for the dissociation of alkyl complex 1 to afford 5 is much lower than the energy needed for the dissociation of alkoxy complex 2 to afford complex 6. Second, in the presence of the substrate or solvating solvent alkyl, complex 1 forms unstable solvated dimer 3, which readily dissociates to give rise to monomeric alkyl complex 7. Further solvation of complex 7 to produce 9 is accompanied by an insignificant increase in free energy. Third, in the presence of the substrate, alkoxy complex 2 forms stable dimer 4, whose thermal dissociation to afford complex 8 is hampered. At the same time, the solvation of complex 8 followed by transition to 10 is energetically favorable. The total ΔG for the transition of 4 to 10 is 74.5 kJ/mol (S = THF) and 63.6 kJ/mol (S = TMC), in agreement with the experimentally determined need for the thermal activation of [(BHT)Mg(THF)(μ-OEt)]2 [49]. A catalyst species of ring-opening polymerization. In our opinion, complex 10 is a catalyst species of ringopening polymerization. Possible pathways of its formation with allowance for the calculated values of the relative energy of complexes 1–9 are also presented in Fig. 1. When a single-component catalyst, namely, stable dimer 4 is used, complex 10 is formed during the thermal dissociation complex 4 followed by coordina-
tion of the second molecule of S. This process is energetically unfavorable. In contrast, the formation of monomeric complexes 7 and 9 from dimer 1 needs minimal energy expenditure, because the process is additionally favored by the instability of solvated dimer 3. Monomeric alkyl complexes 7 and 9, which are Grignard reagents, readily interact with alcohol (ROH) used as an initiator to give rise to solvated complexes 8 and 10, respectively. This reasoning confirms the experimental fact that exactly alkyl complexes activated by ROH are the most effective catalysts of ring-opening polymerization for the studied series of mono-BHT complexes of Mg [49]. Mechanism of Ring-Opening Polymerization in the Presence of Alcoholates of Active Metals On the basis of assumption about common pathways for the ring-opening polymerizations of lactones and carbonates catalyzed by metal alkoxy complexes [28–47], it may be proposed that the reaction catalyzed by the BHT complex of magnesium occurs via the coordination–insertion mechanism through formation of four key intermediates and three transition states at each stage (Scheme (3)). At the stage of first insertion, the process includes the nucleophilic attack of the alkoxy group of the alcohol initiator which is connected to the metal atom accompanied by opening of the cyclic substrate and its coordination to the magnesium atom by the end alkoxy group. At the subsequent stage of chain growth, the role of initiator is played by the end alkoxy group. Simulation of ringopening polymerization with the aim to compare reactivities of a number of cyclic substrates may be reduced
POLYMER SCIENCE, SERIES B
Vol. 59
No. 2
2017
POLYMERIZATION OF TRIMETHYLENE CARBONATE AND LACTONES
to analysis of the stage of first insertion with the use of a methoxy group as the simplest structure initiator. Hence, the catalyst species is the tetrahedral complex of composition (BHT)–Mg(S)2–OMe, I-1. The nucleophilic attack of OMe at the carbonyl group of cyclic ester I in complex I-1 proceeds via the transition state TS-12 and leads to the formation of intermediate I-2, in which the endocyclic oxygen atom is coordinated to the magnesium atom. I-2 through TS-23 via ring opening transforms into complex I-3, which then through TS-34 forms “open,” coordination unsaturated alcoholate I-4, able to coordinate the next substrate molecule. This coordination leads to I-1p (R = MeOC(O)–Z–), from which the next catalytic cycle begins with the growth of polymer chains. For the comparative estimation of energy profiles of reaction, it is sufficient to consider intermediates and transition states for the stage of first insertion and intermediate I-1p for the stage of chain growth. The change in the free energy of the system during the
reaction may be estimated from the difference in the energies of intermediates I-1p and I-1. The generally accepted scheme may be substantially complemented by considering the alternative pathway of transition I-3 to I-1p which includes the coordination of a substrate molecule to afford intermediate I-3' transforming into I-1p, bypassing the stage of coordination of unsaturated intermediate I-4. An even more substantial addition to the generally accepted mechanism is based on the fact that the transition of I-1 to I-2 should include the stage of rotation around bond Mg–O through transition state TS-12' (Scheme (3)), whose geometry and energy, in our opinion, require additional estimation. The main intermediates and transition states for the ring-opening polymerization of lactones and carbonates catalyzed by mono-BHT complexes of magnesium (R = Me, insertion; R = MeOC(O)–Z–, chain growth) may be depicted as follows:
≠ O RO
O Mg BHT
O I-1
Mg O
TS-12
O
BHT
O
BHT
O
BHT
O
I-2
O
O
Mg
Mg
TS-12′
O
O
O
O
≠
RO
RO
O
RO
O
Mg BHT
≠
O
RO
151
O
TS-23
O
O
O O
RO O
RO O
O
O
O
O
BHT
O
BHT
O I-4
I-1p
TS-34
O
O
RO O
O
Mg
Mg
Mg BHT
≠
RO
O
RO
O
BHT
O
O O
O
O
Mg
BHT
O
O
Mg
BHT
TS-34′
O
I-3′ O
POLYMER SCIENCE, SERIES B
O O
Vol. 59
O No. 2
2017
O
I-3 O
O
O
≠
(3)
Mg
O
RO
O
152
NIFANT’EV et al.
Results of Quantum-Chemical Calculations for TMC, PDO, δ-VL, and ε-CL Using TMC as a substrate, the optimized geometry and free energy of intermediates and transition states were calculated in accordance with the mechanism of reaction shown in Scheme (3). For each intermediate, the conformations of cyclic fragments and the mutual orientation of BHT and substrate were varied in order to search for the most stable states. Transitions between the optimized structures of intermediates were studied by scanning over key geometric parameters (distances MeO–C(O) for TS-12, MeO–Mg parallel to O(cycle)–Mg for TS-12', MeO(O)C–O–CH2 for TS-23, and C=O–Mg for TS-34). An examination of free energies of intermediates and transition states made it possible to infer that the reaction pathway including intermediate I-3' and passing through highenergy state TS-34' is treated as energetically less favorable. In the case of TMC, the most stable intermediate of reaction is a complex with the product of ring opening I-3, in which the magnesium atom occurs in the tetrahedral configuration, is covalently bound to the oxygen atoms of BHT and MeOC(O)O(CH2)3O, and is coordinated to two carbonyl groups (TMC and MeOC(O)O(CH2)3O). In accordance with the commonly accepted viewpoint, we expected that the activation barrier of reaction for TMC will correspond to the energy of the first transition state TS-12 (according to calculations of the TMC polymerization catalyzed by Zn complexes [30], exactly this transition state has the maximum energy; structures similar to TS-12' were not examined at all in [30]). However, our calculations supplemented the generally accepted scheme by transition state TS-12', whose energy for TMC is much higher than that for TS-12. During the comparative analysis of the energy profiles of reaction for four main cyclic substrates, we confined ourselves to the five types of intermediates (I-1-I-4, I-1p) and four transition states (TS-12, TS12', TS-23, TS-34). For PDO, δ-VL, and ε-CL, two conformers of intermediates I-1, which are distinguished by the presence of a weak coordination of ‒OMe to the carbonyl atom of carbon, were found. The first conformer, in which such a coordination is present, is structurally similar to I-1 (TMC), and distances MeO–C(=O) are 2.315 Å for PDO, 2.535 Å for δ-VL, and 2.648 Å for ε-CL (for TMC, this value is 2.571 Å). The second conformer, I-1', has an open structure. For TMC, this type is not fixed, and for PDO, δ-VL, and ε-CL, the energies of I-1' are lower than the energies of I-1. As a consequence, in construction of energy profiles, I-1 was selected as a ground state for TMC and I-1' was selected as a ground state for PDO, δ-VL, and ε-CL. Similar data
were obtained for final intermediates I-1p: open intermediate I-1p' was not fixed for TMC, and for PDO and δ-VL, its energy is lower than the energy of I-1p. In contrast, intermediate I-1p' is not fixed for ε-CL. However, the most important result was obtained by comparing the values of free energy of TS-12 and TS-12' for a number of substrates. As was mentioned above, a new, supplementing the generally accepted scheme, transition state TS-12' for TMC has a higher energy than that of TS-12. Calculations of free energies of TS-12 and TS-12' for processes involving PDO and lactones showed that, in the case of PDO and δ-VL, TS-12 has the maximum energy. In the case of ε-CL, the energy of TS-12' is ~4 kJ/mol higher than the energy of TS-12. Estimation of the reactivity of the substrate from the value of the barrier of nucleophilic attack (TS-12) is inapplicable in the case of ε-CL. It is possible that just an increase in the activation energy corresponding to the transition state of rotation of TS-12' is responsible for a relatively low reactivity in the ring-opening polymerization of lactones containing expanded rings (ω-pentadecalactone) and ringsubstituted cyclic esters. The calculated values of relative free energy are summarized in the table, and the energy profiles of ring-opening polymerization are shown in Fig. 2. In the ring-opening polymerization of four substrates, activation barriers calculated for the gas-phase reaction correspond to the following sequence of reactivities: TMC > PDO ~ δ-VL ~ ε-CL. Hence, as follows from calculations, TMC is the most active substrate. Monomers δ-VL and ε-CL should demonstrate close reactivities, and the reactivity of δ-VL is slightly higher than that of ε-CL. The general change in free energy on the transition of I-1 to I-1p ΔGtot = G(I-1p) – G(I-1) – G (Sub) for PDO is a positive value equal to 14.2 kJ/mol; the ring-opening polymerization of this monomer in solution should be reversible, as opposed to the polymerization of ε-CL (ΔGtot = –4.2 kJ/mol) and TMC (ΔGtot = ‒2.1 kJ/mol). The values of activation barriers obtained for catalysis with mono-BHT complexes of magnesium are qualitatively different from the data reported in [11] for ring-opening polymerization in the presence of 1,5,7triazabicyclo[4,4,0]decene-5 (TBD). In the presence of this organic catalyst, the sequence of reactivities of the monomers may be written as PDO > δ-VL > TMC ~ ε-CL. Note that the reaction has much higher (by ~20 kJ/mol) activation energies compared with the processes catalyzed by (BHT)Mg(OMe).
POLYMER SCIENCE, SERIES B
Vol. 59
No. 2
2017
POLYMERIZATION OF TRIMETHYLENE CARBONATE AND LACTONES
153
Relative free energies for intermediates and transition states of ring-opening polymerization in the presence of (BHT)Mg(OMe) for trimethylene carbonate (TMC), 1,4-dioxanone (PDO), δ-valerolactone (δ-VL), and ε-caprolactone (ε-CL) Intermediate (transition state)
Relative free energy (kcal/mol) for substrates TMC
PDO
δ-VL
ε-CL
I-1
0
10.5
5.0
9.2
I-1'
No data
0
0
0
TS-12
33.9
50.6
49.8
47.3
TS-12'
38.1
35.6
43.9
51.5
I-2
12.6
17.6
23.0
14.6
TS-23
23.4
43.9
49.4
19.2
I-3
–7.5
2.9
5.4
8.8
TS-34
33.5
38.9
41.0
No data
I-4
13.8
38.9
35.6
20.9
I-1p
–2.1
21.8
10.5
No data
I-1p'
No data
14.2
3.8
−4.2
* ΔG tot
−2.1
14.2
3.8
−4.2
E α**
38.1
50.6
49.8
51.5
* ΔG = G(I-1p) – G(I-1) – G(Sub). tot **Ε is the relative free activation energy. α
A Comparative Experimental Estimation of Productivity of BHT–Mg Complexes TMC, PDO, δ-VL, and ε-CL were polymerized in the form of 1 mol/L solutions in toluene at 20°C in the presence of 1 mol % (BHT)Mg(THF)2Bu activated by benzyl alcohol. Reaction mixtures were sampled at equal time intervals, and the product-to-substrate ratio was calculated from analysis of high-resolution 1H NMR spectra. Figure 3 shows structural fragments, the signals of which were taken into account in analysis of the spectra, together with a number of 1H NMR spectra for the ring-opening polymerization of ε-CL as an example. The values of monomer conversion were approximated for a 1-min interval of reaction, and on the basis of this approximation, the productivity of the catalysts was estimated. The value of TOF (Turnover Factor, which is the number of monomer insertions per catalyst molecule per unit time) was 71 min–1 for δ-VL and 64 min–1 for ε-CL. The polymerization of TMC was complicated by formation of a poorly soluble polymeric product, and estimation of the reaction rate at the initial time (10 s) yielded TOF ~170 min–1. The polymerization of PDO performed under the same conditions for 1 min afforded a poorly soluble polyPOLYMER SCIENCE, SERIES B
Vol. 59
No. 2
2017
mer, whose yield was determined by weighing of the product (conversion of 82%, TOF ~80 min–1). In this case, the solution was practically free of polymeric products. It appears that a low solubility of the polymeric product in the case of PDO was an additional factor that shifted the equilibrium of PDO polymerization, which is thermodynamically forbidden at room temperature, to the right. CONCLUSIONS Using the model reaction including the interaction of BHT–Mg–OMe with cyclic esters, energy profiles have been calculated by the DFT method for the ringopening polymerization of trimethylene carbonate, 1,4-dioxanone, δ-valerolactone, and ε-caprolactone. It has been shown that the values of activation barriers correlate with the experimentally observed rates of ring-opening polymerization. For six-membered substrates PDO and δ-VL, transition states at the stage of nucleophilic attack of the alkoxy group at the carbonyl group of substrate (TS-1) have the maximum energy. In contrast, for TMC and ε-CL, a key feature is a qualitatively different transition state TS-1', which for the first time has been fixed in this study for ring-opening polymerization induced
154
NIFANT’EV et al.
G, kJ/mol TMC O
RO
RO
O Mg BHT O
RO
O
Mg BHT
O
BHT
O
Mg O
PDO
RO
ε-CL
O
O
O
TS-12
δ-VL
O
O O
O
Mg
TS-23
TS-12'
BHT O O
50
TS-34
25
I-2 RO O
0 I-1
O
O
RO
I-1p
Mg BHT
O
O Mg BHT O
O
I-3
O
RO
−25
I-4
O
O
O
Mg O
BHT
RO
O O
O
O
Mg BHT
O RO
BHT
O Mg O
O O O
Fig. 2. Energy profiles of the ring-opening polymerization of trimethylene carbonate (TMC), 1,4-dioxanone (PDO), δ-valerolactone (δ-VL), and ε-caprolactone (ε-CL) in the presence of (BHT)Mg(OMe). H
(a) H H
H
O
O
O
(b)
H H H H O
O
H H
O
1.00
n
4.25
4.20
0.96
4.15
H H
4.10
4.05
4.00
4.10
4.05
4.00
4.05
4.00
HH H H
H H
O
O
O
H H O
1.00
n
4.25
4.20
1.74
4.15
HH H
H H
O
O
H
O O H H
n
1.00
4.25
3.46
4.20
4.15
4.10
Fig. 3. (а) Structural fragments of monomers and corresponding polymers used in the analysis of 1H NMR spectra and (b) the region of 1H NMR spectra of reaction mixtures of the ring-opening polymerization of ε-CL used to estimate efficiency of the catalyst. POLYMER SCIENCE, SERIES B
Vol. 59
No. 2
2017
POLYMERIZATION OF TRIMETHYLENE CARBONATE AND LACTONES
by complexes of active metals. Most likely, this transition state may be treated as a key one for the polymerization of expanded and substituted lactones catalyzed by metal alkoxides. This assumption makes it possible to explain the experimentally observed low reactivity of such substrates as ω-pentadecalactone. As was theoretically predicted and experimentally confirmed, the mono-BHT magnesium complex, as opposed to organocatalyst (TBD), more efficiently catalyzes the polymerization of trimethylene carbonate compared with lactones. The finding that the activation barriers of ringopening polymerization catalyzed by the mono-BHT complexes of magnesium are close and small makes it possible to assume that these catalysts show promise for the synthesis of random copolymers of trimethylene carbonate, 1,4-dioxanone, δ-valerolactone, and ε-caprolactone, which are promising biomedical materials. In our opinion, the technique of quantum-chemical simulation of the mechanism of ring-opening polymerization used in this study is applicable to the comparative analysis of more complex substrates and the catalytic behavior of sterically crowded phenolates of metals other than magnesium. Hence, the goal of similar investigations is the development of new catalysts of ring-opening polymerization that allow the synthesis of a wide scope of polymers with substantially new properties. ACKNOWLEDGMENTS This work was supported by the Russian Science Foundation, grant no. 16-13-10344. REFERENCES 1. S. Slomkowski, Macromol. Symp. 253, 47 (2007). 2. O. Dechy-Cabaret, B. Martin-Vaca, and D. Bourissou, Chem. Rev. 104, 6147 (2004). 3. J. N. Hoskins and S. M. Grayson, Polym. Chem. 2, 289 (2011). 4. K. Yao and C. Tang, Macromolecules 46, 1689 (2013). 5. N. G. Ricapito, C. Ghobril, H. Zhang, M. W. Grinstaff, and D. Putnam, Chem. Rev. 116, 2664 (2016). 6. D. J. Liuab and E. Y.-X. Chen, Green Chem. 16, 964 (2014). 7. C. M. Thomas, Chem. Soc. Rev. 39, 165 (2010). 8. G.-Q. Chen and M. K. Patel, Chem. Rev. 112, 2082 (2012). 9. N. E. Kamber, B. G. G. Lohmeijer, and J. L. Hedrick, Chem. Rev. 107, 5813 (2007). 10. A. P. Dove, ACS Macro Lett. 1, 1409 (2012). 11. I. Nifant’ev, A. Shlyakhtin, V. Bagrov, B. Lozhkin, G. Zakirova, P. Ivchenko, and O. Legon’kova, React. Kinet., Mech. Catal. 117, 447 (2016). 12. S. Shoda, H. Uyama, J. Kadokawa, S. Kimura, and S. Kobayashi, Chem. Rev. 116, 2307 (2016). POLYMER SCIENCE, SERIES B
Vol. 59
No. 2
2017
155
13. Y. Sarazin and J.-F. Carpentier, Chem. Rev. 115, 3564 (2015). 14. S. M. Guillaume, E. Kirillov, Y. Sarazin, and J.-F. Carpentier, Chem. Eur. J. 21, 7988 (2015). 15. A. Plichta, Z. Florjańczyk, A. Kundys, A. Frydrych, M. Dębowski, and N. Langwald, Pure Appl. Chem. 86, 733 (2014). 16. F. M. Garcia-Valle, R. Estivill, C. Gallegos, T. Cuenca, M. E. G. Mosquera, V. Tabernero, and J. Cano, Organometallics 34, 477 (2015). 17. H.-J. Fang, P.-S. Lai, J.-Y. Chen, S. C. N. Hsu, W.-D. Peng, S.-W. Ou, Y.-C. Lai, Y.-J. Chen, H. Chung, Y. Chen, T.-C. Huang, B.-S. Wu, and H.-Y. Chen, J. Polym. Sci., Part A: Polym. Chem. 50, 2697 (2012). 18. J. A. Wilson, S. A. Hopkins, P. M. Wright, and A. P. Dove, Polym. Chem. 5, 2691 (2014). 19. J. A. Wilson, S. A. Hopkins, P. M. Wright, and A. P. Dove, Macromolecules 48, 950 (2015). 20. H.-Y. Chen, L. Mialon, K. A. Abboud, and S. A. Miller, Organometallics 31, 5252 (2012). 21. A. Duda, “ROP of Cyclic Esters. Mechanisms of Ionic and Coordination Processes,” in Polymer Science: A Comprehensive Reference, Ed. by K. Matyjaszewski and M. Möller (Amsterdam, Elsevier, 2012), Vol. 4, Chap. 4.11, pp. 213–246. 22. C. A. Wheaton, P. G. Hayes, and B. J. Ireland, Dalton Trans., No. 25, 4832 (2009). 23. A. Kowalski, J. Libiszowski, T. Biela, M. Cypryk, A. Duda, and S. Penczek, Macromolecules 38, 8170 (2005). 24. N. Lawan, S. Muangpil, N. Kungwan, P. Meepowpan, V. S. Lee, and W. Punyodom, Comput. Theor. Chem. 1020, 121 (2013). 25. C. Sattayanon, N. Kungwan, W. Punyodom, P. Meepowpan, and S. Jungsuttiwong, J. Mol. Model. 19, 5377 (2013). 26. C. Sattayanon, W. Sontising, W. Limwanich, P. Meepowpan, W. Punyodom, and N. Kungwan, Struct. Chem. 26, 695 (2015). 27. C. Sattayanon, W. Sontising, J. Jitonnom, P. Meepowpan, W. Punyodom, and N. Kungwan, Comput. Theor. Chem. 1044, 29 (2014). 28. D. Cheshmedzhieva, I. Angelova, S. Ilieva, G. S. Georgiev, and B. Galabov, Comput. Theor. Chem. 995, 8 (2012). 29. I. del Rosal, P. Brignou, S. M. Guillaume, J.-F. Carpentier, and L. Maron, Polym. Chem. 6, 3336 (2015). 30. I. del Rosal, P. Brignou, S. M. Guillaume, J.-F. Carpentier, and L. Maron, Polym. Chem. 2, 2564 (2011). 31. Z. Dai, J. Zhang, Y. Gao, N. Tang, Y. Huang, and J. Wu, Catal. Sci. Technol. 3, 3268 (2013). 32. J. S. Klitzke, T. Roisnel, E. Kirillov, O. de L. Casagrande, Jr., and J.-F. Carpentier, Organometallics 33, 5693 (2014). 33. E. E. Marlier, J. A. Macaranas, D. J. Marell, C. R. Dunbar, M. A. Johnson, Y. DePorre, M. O. Miranda, B. D. Neisen, C. J. Cramer, M. A. Hillmyer, and W. B. Tolman, ACS Catal. 6, 1215 (2016).
156
NIFANT’EV et al.
34. K. Ding, M. O. Miranda, B. Moscato-Goodpaster, N. Ajellal, L. E. Breyfogle, E. D. Hermes, C. P. Schaller, S. E. Roe, C. J. Cramer, M. A. Hillmyer, and W. B. Tolman, Macromolecules 45, 5387 (2012). 35. M.-C. Chang, W.-Y. Lu, H.-Y. Chang, Y.-C. Lai, M. Y. Chiang, H.-Y. Chen, and H.-Y. Chen, Inorg. Chem. 54, 11292 (2015). 36. S. Tabthong, T. Nanok, P. Sumrit, P. Kongsaeree, S. Prabpai, P. Chuawong, and P. Hormnirun, Macromolecules 48, 6846 (2015). 37. S. K. Roymuhury, D. Chakraborty, and V. Ramkumar, Eur. Polym. J. 70, 203 (2015). 38. M. O. Miranda, Y. DePorre, H. Vazquez-Lima, M. A. Johnson, D. J. Marell, C. J. Cramer, and W. B. Tolman, Inorg. Chem. 52, 13692 (2013). 39. J. Fang, I. Yu, P. Mehrkhodavandi, and L. Maron, Organometallics 32, 6950 (2013). 40. J.-F. Carpentier, Organometallics 34, 4175 (2015). 41. N. Susperregui, M. U. Kramer, J. Okuda, and L. Maron, Organometallics 30, 1326 (2011). 42. J. Liu, J. Ling, X. Li, and Z. Shen, J. Mol. Catal. A: Chem. 300, 59 (2009). 43. J. Linga, J. Shen, and T. E. Hogen-Esch, Polymer 50, 3575 (2009).
44. X. Ni, Z. Liang, J. Ling, X. Li, and Z. Shen, Polym. Int. 60, 1745 (2011). 45. M. Bouyahyi, N. Ajellal, E. Kirillov, C. M. Thomas, and J.-F. Carpentier, Chem. – Eur. J. 17, 1872 (2011). 46. R. Reichardt, S. Vagin, R. Reithmeier, A. K. Ott, and B. Rieger, Macromolecules 43, 9311 (2010). 47. A. Walshe, J. Fang, L. Maron, and R. J. Baker, Inorg. Chem. 52, 9077 (2013). 48. E. L. Marshall, V. C. Gibson, and H. S. Rzepa, J. Am. Chem. Soc. 127, 6048 (2005). 49. I. E. Nifant’ev, A. V. Shlyakhtin, A. N. Tavtorkin, P. V. Ivchenko, R. S. Borisov, and A. V. Churakov, Catal. Commun. 87, 106 (2016). 50. D. N. Laikov and Y. A. Ustynyuk, Russ. Chem. Bull. 54, 820 (2005). 51. J. P. Perdew, K. Burke, and M. Ernzerhof, Phys. Rev. Lett. 77, 3865 (1996). 52. J. Gromada, A. Mortreux, T. Chenal, J. W. Ziller, F. Leising, and J.-F. Carpentier, Chem. – Eur. J. 8, 3773 (2002). 53. K. W. Henderson, G. W. Honeyman, A. R. Kennedy, R. E. Mulvey, J. A. Parkinson, and D. C. Sherrington, Dalton Trans., No. 7, 1365 (2003).
POLYMER SCIENCE, SERIES B
Translated by T. Soboleva
Vol. 59
No. 2
2017