Sterically Induced Ligand Framework Distortion Effects on Catalytic

United States. Inorg. Chem. , 2018, 57 (6), pp 3451–3457. DOI: 10.1021/acs.inorgchem.8b00250. Publication Date (Web): February 27, 2018. Copyrig...
0 downloads 5 Views 1MB Size
Download PDF
Article pubs.acs.org/IC

Cite This: Inorg. Chem. XXXX, XXX, XXX−XXX

Sterically Induced Ligand Framework Distortion Effects on Catalytic Cyclic Ester Polymerizations Joahanna A. Macaranas, Anna M. Luke, Mukunda Mandal, Benjamin D. Neisen, Daniel J. Marell, Christopher J. Cramer,* and William B. Tolman*,‡ Department of Chemistry, Center for Sustainable Polymers, Chemical Theory Center, and Minnesota Supercomputing Institute (MSI), University of Minnesota, 207 Pleasant St. SE, Minneapolis, Minnesota 55455, United States ‡ Currently: Department of Chemistry, Washington University in St. Louis, One Brookings Drive, Campus Box 1134, St. Louis, Missouri 63130, United States S Supporting Information *

ABSTRACT: Aluminum alkoxide complexes supported by salen ligands [salen = N,N′bis(salicylaldimine)-2-methylpropane-1,2-diamine or N,N′-bis(salicylaldimine)-2,2-dimethylpropane-1,3-diamine] with o-adamantyl substituents have been synthesized and investigated for the polymerization of ε-caprolactone. Geometric analysis of the catalysts used for the reaction reveals the metal coordination geometries to be intermediate between square-pyramidal and trigonal-bipyramidal. A detailed kinetic study accompanied by density functional theory modeling of key mechanistic steps of the reaction suggest that, in addition to the length of the backbone linker, the o-aryl substituents have a significant impact on the catalyst’s reactivity. Bulky ortho substituents favorably distort the precatalyst geometry and thereby foster the achievement of the rate-limiting transition-state geometry at low energetic cost, thus accelerating the reaction.

■

INTRODUCTION Ring-opening transesterification polymerization (ROTEP) of cyclic esters catalyzed by metal alkoxide complexes is an attractive route toward the sustainable production of biodegradable polymers.1−7 With the ultimate aim of designing more efficient catalysts, a mechanistic understanding of these reactions has been sought through variation of the supporting ligand and metal ion.1,2,8−12 Particular focus has been placed on aluminum alkoxide complexes bound to salen [salen = N,N′bis(salicylaldimine)-2-methylpropane-1,2-diamine or N,N′-bis(salicylaldimine)-2,2-dimethylpropane-1,3-diamine] or related imine phenoxide ligands largely because of their ready synthesis, their modest and thus conveniently monitored ROTEP rates, the ease with which steric and electronic influences can be studied through ligand structural variation, and the high degree of molecular weight control that they exhibit. For example, previous studies1,2 of single-site (salen)AlOR catalysts 1 and 2 (Figure 1), which differed with respect to the length of the imine-backbone tether and remote substituents (R1 = OMe, Br, NO2), showed that the rates of ROTEP of ε-caprolactone (CL) were enhanced by electronwithdrawing groups and longer tether length. Detailed computational analysis revealed that catalysts having longer backbone tethers (e.g., 2 compared to 1) adopt “resting state” geometries that are closer to trigonal-bipyramidal (tbp) and thus more readily able to distort to the rate-limiting octahedral transition-state (TS) geometry at low energetic cost. A similar analysis focusing on the cost of distortion to the optimal octahedral TS structure rationalized the extremely low activity of a tetramethyl-5,7,12,14-dibenzo-1,4,8,11tetraaza[14]annulene (TMTAA)-based aluminum catalyst for © XXXX American Chemical Society

Figure 1. (salen)Al catalysts studied in polymerizations of cyclic esters. X = Me for 3−6; X = OiPr for 1, 2, 7, and 8; R = Me for all complexes except 5, for which R = H. This work focuses on 7 and 8.

CL polymerization, which requires almost 10 days for >90% conversion.13 Here, the rigid, near-square-pyramidal (sp) TMTAA-ligand framework strongly resists such distortion. These observations of such a structure−activity relationship led us to hypothesize that similar distortions of the precatalyst geometry and increased ROTEP rates might be imposed by enhancing the steric profile of the o-aryl substituents (R2). Previous studies of ROTEP of rac-lactide (LA) by catalysts 3 and 4 revealed slower rates for R1 = R2 = tBu than for smaller H or Cl units, but substituents larger than tBu were not examined.3 In other work, catalysts 5 and 6 having o-adamantyl (Ad) groups but different linker lengths exhibited essentially identical rates of ROTEP of LA, but comparisons to derivatives with tBu groups were not drawn, preventing comparative assessment of the steric effect of the Ad groups on ROTEP Received: January 28, 2018

A

DOI: 10.1021/acs.inorgchem.8b00250 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry rates.11,12 To address this issue and test the postulate that ortho substituents larger than tBu would enhance the rate of ROTEP of CL through favorable framework distortion, we targeted 7 and 8 for synthesis, ROTEP kinetics studies, and mechanistic evaluation by theory. As described in more detail below, comparisons to the tBu analogues 1 and 2 revealed a significant rate increase for the two-carbon linker system that was indeed correlated with significant geometric distortion of the starting catalyst structure by the Ad groups, but a much smaller effect was seen for the already tbp-like system with the three-carbon linker. Our results and accompanying theoretical evaluation further illustrate the range of the framework distortion analysis model as a means to evaluate and design metal alkoxide complexes as ROTEP catalysts.

■

RESULTS AND DISCUSSION Catalyst Synthesis and Characterization. The Adsubstituted salicylaldehydes used to make proligands H2Lx to support catalysts 7 (two-carbon linker) and 8 (three-carbon linker) were prepared analogously to previously reported procedures.1,2,14,15 A well-established1−12 condensation of the salicylaldehydes with the appropriate diamine afforded the desired ligands in good yield. Subsequent metalation of the proligands with aluminum isopropoxide at 85−90 °C for 3−4 days afforded complexes 7 and 8, the identities of which were confirmed by 1H and 13C NMR spectroscopy (Figures S3−S6). Most indicative of the conversion to the complexes was the disappearance of the salen hydroxyl hydrogen peaks (δ ≈ 14 ppm) in the 1H NMR spectra, as well as the appearance of peaks corresponding to the Al(OiPr) methine (δ ≈ 4 ppm, septet) and methyl (δ ≈ 1 ppm, doublet) hydrogen atoms. Also, upon metalation, the methylene protons in the backbone linker became diastereotopic and appeared as separate peaks. Single crystals of both 7 and 8 formed during the metalation reactions, enabling their X-ray structures to be determined (Figure 2). The complexes are mononuclear with geometries intermediate between the sp and tbp extremes, as reflected by τ5 values of 0.63 (7) and 0.81 (8) (where values of 0 and 1 are associated with the idealized sp and tbp geometries, respectively).16 These values are similar to those for the tBusubstituted variants 1 (0.52) and 2 (0.84),1,2 with the larger value for 7 compared to 1 suggesting enhanced twisting of the two-carbon backbone by the larger Ad groups in the former. These structural differences are further explored by density functional theory (DFT) calculations below. Polymerization Kinetics. The rates of polymerization of CL by catalysts 7 and 8 were determined by adding catalyst, internal standard [bis(p-trimethylsilyl)benzene], and monomer into an NMR tube in target concentrations of 0.007, 0.004, and 2.0 M, respectively, in toluene-d8 at temperatures between 273 and 300 K. The progress of the reactions (performed in triplicate at each temperature) were monitored via 1H NMR spectroscopy, by acquiring arrays of spectra until the nearcomplete disappearance of the monomer was observed (95− 99% monomer conversion). Integrations of monomer and polymer peaks from the spectral array data were used to determine the concentration as a function of time.17 Global fits using COPASI18 to first-order, second-order, and saturation (Michaelis−Menten) rate equations were attempted. Poor fits to first- and second-order rate laws contrasted with excellent fits to saturation kinetics (Figures S7, S8, and S10), which provided Keq and k2 values at each temperature (Tables S1 and S2). The saturation kinetics model was corroborated by an independent

Figure 2. Representation of the X-ray crystal structures of 7 (a) and 8 (b), showing all non-hydrogen atoms as 50% thermal ellipsoids. Selected interatomic distances (Å) and angles (deg) for 7: Al1−O1, 1.8215(15); Al1−O2, 1.8084(18); Al1−O3, 1.7240(19); Al1−N1, 2.039(2); Al1−N2, 2.014(2); O3−Al1−O2, 117.68(9); O3−Al1−O1, 97.10(9); O2−Al1−O1, 94.49(8); O3−Al1−N1, 96.24(9); O2−Al1− N1, 87.74(8); O1−Al1−N1, 163.57(9); O3−Al1−N2, 115.99(9); O2−Al1−N2, 125.51(9); O1−Al1−N2, 87.75(9); N1−Al1−N2, 77.79 (9). Selected interatomic distances (Å) and angles (deg) for 8: Al1− O1, 1.771(2); Al1−O2, 1.819(2); Al−O3, 1.730(2); Al1−N1, 2.041(3); Al1−N2, 1.982(3); O3−Al1−O2, 94.45(10); O3−Al1− O1, 125.00(11) ; O2−Al1−O1, 94.62(10); O3−Al1−N1, 88.97(10); O2−Al1−N1, 173.59(11); O1−Al1−N1, 87.82(10); O3−Al1−N2, 117.80(11); O2−Al1−N2, 89.53(10); O1−Al1−N2, 116.40(11); N1−Al1−N2, 84.08(11).

reaction progress kinetic analysis19 and an evaluation of Keq from the evaluation of NMR peaks associated with the catalyst (Figures S11−S13), which yielded similar kinetic and equilibrium parameters (Tables S3−S6). In addition, a Keq value of 0.7(4), in agreement with the values determined by the kinetic data, was measured by NMR spectroscopy for the binding of γ-butyrolactone (which is not polymerized) to 8 at 293 K (Figure S14), further supporting the kinetic model involving monomer coordination. The Keq values for 1, 2, 7, and 8 fall within a tight range (0.6−1.7 M−1), consistent with the low dependence of the polymerization rate on monomer binding, as noted previously.1,2 More significant temperature dependencies and disparities as a function of the catalyst were observed for the k2 values (Eyring plots are shown in Figure 3; activation parameters are listed in Table 1). It is evident that the polymerization rates for 7 and 8 are faster than those for 1 and 2. Notably, the differences in the k2 values for catalysts with the same linker lengths are significantly larger for 1 versus 7 (twocarbon linker; >5 orders of magnitude difference) than for 2 versus 8 (three-carbon linker; ∼2 orders of magnitude difference). In other words, the effect of replacing the tBu group by an Ad group is larger for the two-carbon linker systems (1 and 7) than for the three-carbon linker complexes (2 and 8). In order to understand these differences and their structural bases, we turned to DFT calculations. B

DOI: 10.1021/acs.inorgchem.8b00250 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry

measured by theoretical τ5 values of 0.77 and 0.79 for 2 and 8, respectively (Table 1). Because the turnover-limiting structure TS1 (Figures 4 and 5) involves an octahedral coordination

Figure 3. Eyring plots of ln(k2/T) versus 1/T for 1 (blue circles), 2 (black circles), 7 (blue squares), and 8 (black squares). Data for 1 and 2 are from refs 1 and 2, respectively, and those for 7 and 8 are from this work. Figure 4. Key stationary points along the reaction coordinate for ROTEP of CL. The turnover-limiting step is the alkoxide insertion step, TS1.

Computational Studies. To gain further insight into the experimentally observed trend in the polymerization rates, DFT modeling of the key stationary points on the relevant potential energy surfaces was undertaken. For computational efficiency, the isopropoxy group of the precatalyst was truncated to a methoxy group, [Al]-OMe. The Ad group is similar to the tBu group in terms of having a quaternary carbon attached to the aryl ring, but the remaining alkyl fragment is a polycyclic cage in the Ad case, as opposed to three methyl groups in the tBu case. To assess this change in the steric profile, we computed the fractional buried volume (% VBur) for each catalyst, a descriptor that quantitatively characterizes the steric environment associated with the catalytic pocket.20 To compute % VBur, a sphere having a particular van der Waals radius is centered on the active site of the catalystin this case, the aluminum atom. The percentage volume of the sphere that overlaps with the volume of other van der Waals spheres placed on ligand atoms, each having their own characteristic radius, is % VBur.21 For complexes 1, 2, 7, and 8, the % VBur values for a 5 Å metal radius are computed to be very similar: 54.1, 55.0, 55.0, and 56.7%, respectively, confirming that the ortho substituent does not explicitly block the approach to the active site (see the Supporting Information for details with other radii). Although there is not much change in the accessible volume across the various precatalysts, the close proximity of the orthosubstituting groups to one another, particularly in the system with a 2-carbon linker, is such that the ligand framework distorts and deviates more from planarity in the Ad case than t Bu. This can be quantified by the geometry index, τ5, which is calculated to be 0.35 for 1 but 0.67 for 7.1,22 The system with a longer 3-carbon linker, with its longer linker, is significantly distorted toward a tbp geometry with either substituent, as

geometry, the energetic cost required to distort the more tbp 7 to an octahedral geometry [referred to as the “framework distortion energy” (FDE)] is reduced compared to that of 1.23 FDE is a descriptor that can correlate well with a given catalyst’s reactivity when the geometries of the TS structures for a range of catalysts are all similar, such that the energy required to adopt those structures correlates closely with the overall activation free energy.2,24 A formal approach to computing the FDE involves computing the difference in the electronic energies of the frozen ligand framework of a resting precatalyst and that of its associated TS1, derived respectively by the removal of methoxy from the optimized former and methoxy and CL from the optimized latter.2 Consistent with the large geometric distortion associated with the Ad group, 7 compared to 1 has a FDE lower by 5.5 kcal/mol and a correspondingly lower free energy of activation associated with TS1 [ΔG⧧(TS1); lower by 2.9 kcal/mol; Table 1]. These results are entirely consistent with the experimental rate acceleration observed for 7 compared to 1. For the C3 system, by contrast, the FDE and associated ΔG⧧(TS1) values vary by substantially smaller values (Table 1), with the latter suggesting that no significant difference in the rates should be expected for 8 compared to 2, again in close agreement with the experimental results.

■

CONCLUSIONS Through the synthesis, structural characterization, and study of the kinetics of CL polymerization of complexes 7 and 8, comparisons to previously reported complexes 1 and 2 were

Table 1. Geometric Index τ5, Activation Parameters, FDEs, and Activation Free Energies through TS1 for Catalysts 1, 2, 7, and 8a τ5 catalyst

expt

theor

ΔH⧧ b

ΔS⧧b

ΔG⧧298b

ΔG⧧(TS1)c

FDEc

1d 7 2e 8

0.52 0.63 0.84 0.81

0.35 0.67 0.77 0.79

11.5(3) 12.4(5) 10.4(4) 10.3(2)

30(2) 16(1) 24(1) 22(1)

20.4(7) 17.2(5) 17.6(5) 16.9(3)

14.1 11.2 10.3 10.2

15.5 10.0 12.4 11.0

a SMD(toluene)//M06-2X/6-311+G(d,p)//M06-L/6-31+G(d,p). Energies in kcal/mol and entropies in cal/K·mol. bDetermined from the linear fits in Figure 3 using the Eyring equation. cDetermined from theory. dReference 1. eReference 2.

C

DOI: 10.1021/acs.inorgchem.8b00250 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry

Figure 5. Standard-state free energies (kcal/mol) at the SMD(toluene)//M06-2X/6-311+G(d,p)//M06-L/6-31+G(d,p) level of theory25−29 for CL ring opening relative to the infinitely separated precatalyst and monomer with line drawings of relevant stationary points. Solid lines refer to tBusubstituted catalysts (1 and 2), while dotted lines refer to the corresponding Ad-substituted catalysts (7 and 8). protonated solvents were degassed and passed through a solvent purification system (Glass Contour, Laguna, CA) prior to use. Deuterated solvents were dried over calcium hydride, degassed through freeze−pump−thaw techniques, and distilled before storage under dinitrogen. NMR spectroscopies were performed with a Bruker Avance III (500 MHz) spectrometer equipped with a BBFO SmartProbe. Chemical shifts for 1H and 13C NMR spectra were referenced to residual protium in the deuterated solvent (for 1H NMR) and the deuterated solvent itself (for 13C NMR). Elemental analyses were performed by either Robertson Microlit Laboratories (Ledgewood, NJ) or Atlantic Microlab, Inc. (Norcross, GA). The starting material, 2-(1-adamantyl)-4-bromophenol, was analogously synthesized according to published procedures.30,31 Synthesis of 3-(1-Adamantyl)-5-bromo-2-hydroxybenzaldehyde. In a nitrogen-filled glovebox, paraformaldehyde (0.219 g, 7.3 mmol), magnesium chloride (anhydrous beads, −10 mesh, 0.621 g, 6.5 mmol), and tetrahydrofuran (THF; 3 mL) were added to an ovendried screw-cap bomb flask equipped with a stirbar. To this slurry, triethylamine (0.91 mL, 6.5 mmol) was added dropwise with stirring. A solution of 2-(1-adamantyl)-4-bromophenol (0.100 g, 3.3 mmol) in THF (7 mL) was then prepared and subsequently added dropwise to the mixture. The bomb flask was sealed, removed from the glovebox, and left to stir at 80 °C in an oil bath overnight. The reaction mixture was then cooled to room temperature before dilution with chloroform (60 mL). Hydrochloric acid (3 mL, 1.0 M) and deionized water (50 mL) were added, and the mixture was additionally stirred overnight at room temperature. The organic phase was extracted and washed with water (3 × 150 mL) and brine (1 × 150 mL), dried over sodium sulfate, and filtered before removal of the solvent in vacuo. Recrystallization of the crude product from hot chloroform and isooctane in a −30 °C freezer overnight gave a yellow-orange, crystalline solid. Yield: 0.809 g (73%). 1H NMR (500 MHz, CDCl3): δ 11.76 (s, 1H, OH), 9.80 (s, 1H, CHO), 7.51 (d, J = 2.5 Hz, 1H, ArH), 7.50 (d, J = 2.5 Hz, 1H, ArH), 2.10 (app s, 9H, AdH), 1.78 (app s, 6H, AdH). 13C NMR (126 MHz, CDCl3): δ 196.26, 160.61, 141.38, 137.32, 133.57, 121.81, 111.56, 40.07, 37.53, 36.98, 28.95.

enabled that focused on the effect of changing the size and shape of the ortho substituents on the supporting ligand. The new Ad-substituted systems exhibited saturation kinetics, with differences in the overall rate being primarily due to disparities in the values of the rate constants k2, the temperature dependencies of which were analyzed by the Eyring equation. Importantly, the Ad-substituted systems 7 and 8 exhibited faster CL polymerization rates than the corresponding ones with o-tBu groups (1 and 2), with larger effects for the complexes having the less flexible and shorter ligand backbone linker. A detailed computational study revealed that increased steric repulsion between ortho groups in the aryloxide ligand for models of 7 and 8 results in a significant difference in the reactant geometries, such that the FDEs to reach the TS geometry are reduced. This prediction of an enhancement in the polymerization rate with a structural change from tBu to Ad ortho substitution is consistent with the detailed kinetic analyses. While the sensitivity to substitution is large for the related catalysts 1 and 7, which have significantly different computed FDE and ΔG⧧(TS1) values, it is less pronounced for catalysts 2 and 8, which both have relaxed catalyst geometries already well characterized as tbp-like, leading to more similar FDE and ΔG⧧(TS1) values. Using theory to assess the hypothetical catalyst geometries and FDEs should prove useful in ongoing design efforts with respect to maximizing activity.

■

EXPERIMENTAL SECTION

Materials, Methods, and General Considerations. All reactions containing air- and/or water-sensitive compounds were performed within the inert atmosphere of a nitrogen-filled glovebox or using Schlenk-line techniques, unless otherwise noted. All reagents were purchased from commercial sources and used as received, unless otherwise noted. ε-Caprolactone (CL) and β-butyrolactone (BL) were purified by drying over calcium hydride and subsequent distillation. All D

DOI: 10.1021/acs.inorgchem.8b00250 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry Synthesis of Ligands H2L7 and H2L8. To an oven-dried round bottom flask equipped with a stirbar and reflux condenser, 3-(1adamantyl)-5-bromo-2-hydroxybenzaldehyde (2.57 g, 7.66 mmol) was added and dissolved in absolute ethanol (0.63 M) with stirring. To this mixture was added dropwise a diamine (either 2,2-dimethylpropane1,3-diamine or 2-methylpropane-1,2-diamine; 0.5 equiv) before heating to reflux for 2 h. The reaction mixture was cooled to room temperature before being placed in a −30 °C freezer overnight. The crude solid was isolated through vacuum filtration and washed with hexanes before recrystallization from hot methanol and hexanes. The purified product was isolated through vacuum filtration as brightyellow crystals. After drying overnight in a vacuum oven, the ligand was stored under dinitrogen in a glovebox. Yield: H2L7, 338 mg (52%); H2L8, 423 mg (77%). H2L7. 1H NMR (500 MHz, CDCl3): δ 14.35 (s, 1H, OH), 13.85 (s, 1H, OH), 8.28 (s, 1H, CHN), 8.24 (s, 1H, CHN), 7.30 (d, J = 2.5 Hz, 1H, ArH), 7.28 (d, J = 2.5 Hz, 1H, ArH), 7.22 (d, J = 2.5 Hz, 1H, ArH), 7.18 (d, J = 2.5 Hz, 1H, ArH), 3.71 (s, 2H, NCH2C(CH3)2N), 2.11 (m, 12H, AdH), 2.08 (m, 6H, AdH), 1.77 (m, 12H, AdH), 1.43 (s, 6H, NCH2C(CH3)2N). 13C NMR (126 MHz, CDCl3): δ 167.60, 166.53, 161.63, 160.18, 159.96, 140.49, 132.76, 132.50, 131.81, 131.65, 120.03, 119.95, 110.23, 110.04, 70.33, 60.46, 40.16, 40.14, 37.46, 37.45, 37.18, 37.16, 29.10, 25.60, and 25.60. Anal. Calcd for C38H46Br2N2O2: C, 63.16; H, 6.42; N, 3.88. Found: C, 62.28; H, 6.38; N, 3.97. H2L8. 1H NMR (500 MHz, CDCl3): δ 14.04 (s, 2H, OH), 8.28 (s, 2H, CHN), 7.32 (d, J = 2.5 Hz, 2H, ArH), 7.23 (d, J = 2.5 Hz, 2H, ArH), 3.49 (s, 4H, NCH2C(CH3)2CH2N), 2.15 (m, 12H, AdH), 2.10 (m, 6H, AdH), 1.80 (m, 12H, AdH), 1.10 (s, 6H, NCH2C(CH3)2CH2N). 13C NMR (126 MHz, CDCl3): δ 165.64, 160.02, 140.50, 132.60, 131.57, 120.02, 110.21, 68.41, 40.18, 37.49, 37.17, 36.51, 29.12, 24.68. Anal. Calcd for C39H48Br2N2O2: C, 63.59; H, 6.57; N, 3.80. Found: C, 63.63; H, 6.69; N, 3.84. Synthesis of Complexes 7 and 8. To an oven-dried screw-cap bomb flask equipped with a stirbar were added a ligand (H2L7, 0.276 g, 0.38 mmol, H2L8, 0.430 g, 0.584 mmol) and aluminum isopropoxide, which had been dissolved in toluene (0.167 M) while in a nitrogenfilled glovebox. The sealed flask was pumped out of the box, heated to 80 °C, and stirred for 3−4 days at this temperature. After cooling to room temperature, the bomb flask was pumped back into the glovebox, and the solvent was removed in vacuo. After washing with pentane, the product was isolated through vacuum filtration and then recrystallized from minimal toluene layered with an equal volume of pentane in a −40 °C freezer overnight. The purified product was isolated as a light-yellow solid and stored under dinitrogen. Yield: 7, 150 mg, (48%); 8, 42.2 mg (63%). Compound 7. 1H NMR (500 MHz, CDCl3): δ 8.27 (s, 1H, CHN), 8.18 (s, 1H, CHN), 7.41 (d, J = 2.6 Hz, 1H, ArH), 7.39 (d, J = 2.6 Hz, 1 H, ArH), 7.24 (d, J = 2.6 Hz, 1H, ArH), 7.19 (d, J = 2.6 Hz, 1H, ArH), 4.23 (d, J = 11.7 Hz, 1H, NCH2C(CH3)2N), 3.68 (sep, J = 5.7 Hz, 1H, OCH(CH3)2), 3.25 (d, J = 11.7 Hz, 1H, NCH2C(CH3)2N), 2.26 (m, 3H, AdH), 2.19 (m, 6H, AdH), 2.04 (m, 6H, AdH), 1.96 (m, 3H, AdH), 1.76 (m, 3H, AdH), 1.068 (m, 6H, AdH), 1.66 (m, 3H, AdH), 1.54 (s, 3H, NCH2C(CH3)2N), 1.23 (s, 3H, NCH2C(CH3)2N), 0.93 (d, J = 5.66 Hz, 3H, OCH(CH3)2), 0.92 (d, J = 5.66 Hz, 3H, OCH(CH3)2). 13C NMR (126 MHz, CDCl3): δ 167.96, 165.97, 164.75, 164.25, 144.69, 144.16, 136.33, 135.39, 133.12, 132.50, 129.18, 128.37, 125.44, 121.07, 120.71, 108.43, 107.35, 65.81, 62.92, 60.63, 40.59, 40.29, 37.93, 37.86, 37.16, 37.08, 34.28, 29.29, 29.05, 27.99, 27.63, 25.92, 22.50, 21.61, 14.22. Anal. Calcd for C41H51AlBr2N2O3: C, 61.05; H, 6.37; N, 3.47. Found: C, 62.11; H, 6.31; N, 3.19. Compound 8. 1H NMR (400 MHz, CDCl3): δ 8.00 (s, 2H, CHN), 7.36 (d, J = 2.7 Hz, 2H, ArH), 7.19 (d, J = 2.7 Hz, 2H, ArH), 3.90 (sep, J = 5.5 Hz, 1H, OCH(CH3)2), 3.85 (d, J = 11.8 Hz, 2H, NCH2C(CH3)2CH2N), 3.22 (d, J = 12.1 Hz, 2H, NCH2C(CH3)2CH2N), 2.13 (m, 6H, AdH), 2.07 (m, 6H, AdH), 1.95 (m, 6H, AdH), 1.64 (m, 12H, AdH), 1.11 (s, 3H, NCH2C(CH3)2CH2N), 0.97 (d, J = 5.54 Hz, 6H, OCH(CH3)2), 0.93 (s, 3H, NCH2C(CH3)2CH2N). 13C NMR (126 MHz, CDCl3): δ 167.86, 164.90, 147.84, 144.27, 135.74, 132.85, 129.19, 128.38, 125.45, 120.76, 107.90, 69.67, 40.28, 37.80, 37.05, 36.04, 34.29, 29.14, 25.72, 22.50, 14.22. Anal. Calcd for C42H53AlBr2N2O3: C, 61.47; H, 6.51; N, 3.41. Found: C, 62.20; H, 6.31; N, 3.19.

Kinetic Measurements and Analysis. Polymerization kinetics were performed by adding 450 μL of a stock solution of catalyst in toluene-d8 (9.7 mM) and 5 μL of a stock solution of internal standard (bis[p-(trimethylsilyl)]benzene) in toluene-d8 (0.5 M) to a new NMR tube inside a nitrogen-filled glovebox. The NMR tube was then capped with a rubber septum and wrapped with electrical tape. A gastight syringe charged with 170 μL of a stock solution of CL in toluene-d8 (7.35 M) was also capped with a septum to prevent air contamination. The target concentrations for the polymerization reaction were 0.007 M catalyst, 0.004 M internal standard, and 2.0 M CL. Both the NMR tube and syringe were pumped outside the box and brought to the NMR instrumentation center. A methanol standard was used to calibrate the temperature of the NMR spectrometer (500 MHz Bruker Avance III) to ensure an accurate temperature reading. A sample 1H NMR spectra of the catalyst and internal standard was collected to ensure appropriate shimming, and the NMR tube was ejected from the instrument to load the monomer into the tube. Once the monomer was injected into the NMR tube, the tube was shaken vigorously and immediately injected into the spectrometer. The time between monomer injection and the spectral array was measured in minutes, and the spectral array was started as soon as the NMR tube had locked and shimmed. An array of spectra was taken every 48 s (four scans) with a relaxation delay of 10 s, a gain of 10, and an acquisition time of 2 s. The sample was spun, and autoshim was used for the kinetic array to ensure proper shimming through the entirety of the reaction. The reaction was monitored until there was a complete disappearance of the monomer peaks in the spectrum. For each catalyst at each temperature, the polymerization reactions were done in triplicate. The obtained NMR data were analyzed through MestReNova software by integration. The integrated peak and determined concentration of the internal standard allowed for the absolute concentrations of all species as a function of time to be calculated. Then, the reaction time was calculated in seconds by using the known time duration of each spectrum and the amount of time between injecting the NMR tube into the spectrometer and beginning the 1H NMR spectrum acquisition. The concentration versus time data were entered into COPASI software and fit to eq S1 to obtain the respective KM and Vmax values. Using these values, the rate constants Keq and k2 were calculated, and the reaction rates were plotted as a function of [CL] (full fit details are included in the Supporting Information). All curve fits were performed with Origin software (OriginLab, Northampton, MA). All computational details involved with this work are provided in the Supporting Information.

■

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.inorgchem.8b00250. Experimental data, concentration versus time plots, and computational details (PDF) Accession Codes

CCDC 1819919−1819920 contain the supplementary crystallographic data for this paper. These data can be obtained free of charge via www.ccdc.cam.ac.uk/data_request/cif, or by emailing [email protected], or by contacting The Cambridge Crystallographic Data Centre, 12 Union Road, Cambridge CB2 1EZ, UK; fax: +44 1223 336033.

■

AUTHOR INFORMATION

Corresponding Authors

*E-mail: [email protected] (C.J.C.). Twitter: @ChemProfCramer. *E-mail: [email protected] (W.B.T.). Twitter: @WBTolman. E

DOI: 10.1021/acs.inorgchem.8b00250 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry ORCID

(8) Rosen, T.; Popowski, Y.; Goldberg, I.; Kol, M. Zinc Complexes of Sequential Tetradentate Monoanionic Ligands in the Isoselective Polymerization of Rac-Lactide. Chem. - Eur. J. 2016, 22, 11533−11536. (9) Gupta, K. C.; Sutar, A. K. Catalytic Activities of Schiff Base Transition Metal Complexes. Coord. Chem. Rev. 2008, 252, 1420− 1450. (10) Cozzi, P. G. Metal-Salen Schiff Base Complexes in Catalysis: Practical Aspects. Chem. Soc. Rev. 2004, 33, 410−421. (11) Agatemor, C.; Arnold, A. E.; Cross, E. D.; Decken, A.; Shaver, M. P. Aluminium Salophen and Salen Initiators in the Ring-Opening Polymerisation of Rac-Lactide and Rac-β-Butyrolactone: Electronic Effects on Stereoselectivity and Polymerisation Rates. J. Organomet. Chem. 2013, 745−746, 335−340. (12) Cross, E. D.; Allan, L. E. N.; Decken, A.; Shaver, M. P. Aluminum Salen and Salan Complexes in the Ring-Opening Polymerization of Cyclic Esters: Controlled Immortal and Copolymerization of Rac-β-Butyrolactone and Rac-Lactide. J. Polym. Sci., Part A: Polym. Chem. 2013, 51, 1137−1146. (13) Stasiw, D. E.; Mandal, M.; Neisen, B. D.; Mitchell, L. A.; Cramer, C. J.; Tolman, W. B. Why So Slow? Mechanistic Insights from Studies of a Poor Catalyst for Polymerization of ε-Caprolactone. Inorg. Chem. 2017, 56, 725−728. (14) Gademann, K.; Chavez, D. E.; Jacobsen, E. N. Highly Enantioselective Inverse-Electron-Demand Hetero-Diels-Alder Reactions of α,β-unsaturated Aldehydes. Angew. Chem., Int. Ed. 2002, 41, 3059−3061. (15) Bryliakov, K. P.; Talsi, E. P. Titanium-Salan-Catalyzed Asymmetric Oxidation of Sulfides and Kinetic Resolution of Sulfoxides with H2O2 as the Oxidant. Eur. J. Org. Chem. 2008, 19, 3369−3376. (16) Addison, A. W.; Rao, T. N.; Reedijk, J.; van Rijn, J.; Verschoor, G. C. Synthesis, Structure, and Spectroscopic Properties of Copper(II) Compounds Containing Nitrogen−sulphur Donor Ligands; the Crystal and Molecular Structure of aqua[1,7-bis(N-Methylbenzimidazol-2′-Yl)-2,6-Dithiaheptane]copper (II) Perchlorate. J. Chem. Soc., Dalton Trans. 1984, 1349−1356. (17) Mestrelab Research SL, Feliciano Barrera 9B, Bajo, 15706 Santiago de Compostela, Spain; www.mestrelab.com. (18) Hoops, S.; Sahle, S.; Gauges, R.; Lee, C.; Pahle, J.; Simus, N.; Singhal, M.; Xu, L.; Mendes, P.; Kummer, U. COPASI: a COmplex PAthway SImulator. Bioinformatics 2006, 22, 3067−74. (19) Blackmond, D. Reaction progress kinetic analysis: A powerful methodology for mechanistic studies of complex catalytic reactions. Angew. Chem., Int. Ed. 2005, 44, 4302−4320. (20) Hillier, A. C.; Sommer, W. J.; Yong, B. S.; Petersen, J. L.; Cavallo, L.; Nolan, S. P. A Combined Experimental and Theoretical Study Examining the Binding of N-Heterocyclic Carbenes (NHC) to the Cp*RuCl (Cp* = η5-C5Me5) Moiety: Insight into Stereoelectronic Differences between Unsaturated and Saturated NHC Ligands. Organometallics 2003, 22, 4322−4326. (21) Falivene, L.; Credendino, R.; Poater, A.; Petta, A.; Serra, L.; Oliva, R.; Scarano, V.; Cavallo, L. SambVca 2. A Web Tool for Analyzing Catalytic Pockets with Topographic Steric Maps. Organometallics 2016, 35, 2286−2293. (22) The agreement between theory and experiment for τ is less good for 1 than for 7, but when a structure for 1 is computed, constraining τ to its experimental value of 0.52, the increase in energy is only 0.5 kcal/mol, suggesting that, at least in this range of the τ coordinate, the system is fairly flexible and solid-state packing effects may contribute to differences in the observed crystal and fully optimized gas-phase structures. (23) Although the energetics of a single cycle of propagation must be independent of the catalyst, Figure 5 shows small variations in net energetics as a function of the catalyst. This is because the step being studied is initiation, not propagation, and the terminal methoxy remains close enough to have different interactions with the different catalyst structures (which would be eliminated with continued chain growth).

Joahanna A. Macaranas: 0000-0001-7610-4453 Anna M. Luke: 0000-0001-9486-5635 Mukunda Mandal: 0000-0002-5984-465X Daniel J. Marell: 0000-0003-2453-9062 Christopher J. Cramer: 0000-0001-5048-1859 William B. Tolman: 0000-0002-2243-6409 Author Contributions

The manuscript was written through contributions of all authors. All authors have given approval to the final version of the manuscript. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS Funding for this project was provided by the Center for Sustainable Polymers at the University of Minnesota, a National Science Foundation (NSF)-supported Center for Chemical Innovation (Grant CHE-1413862). The X-ray diffraction experiments were performed using a crystal diffractometer acquired through NSF-MRI Award CHE1229400. The authors acknowledge the MSI at the University of Minnesota for providing resources that contributed to the research results reported within this paper. We also thank Dr. Manuel Ortuño for helpful discussions on computational modeling. NMR instrumentation was supported by the Office of the Vice President of Research, College of Science and Engineering, and the Department of Chemistry at the University of Minnesota; the Bruker HD NMR spectrometer was supported by the Office of the Director, National Institutes of Health under Award S10OD011952. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.

■

REFERENCES

(1) Miranda, M. O.; Deporre, Y.; Vazquez-Lima, H.; Johnson, M. A.; Marell, D. J.; Cramer, C. J.; Tolman, W. B. Understanding the Mechanism of Polymerization of ε-Caprolactone Catalyzed by Aluminum Salen Complexes. Inorg. Chem. 2013, 52, 13692−13701. (2) Marlier, E. E.; Macaranas, J. A.; Marell, D. J.; Dunbar, C. R.; Johnson, M. A.; DePorre, Y.; Miranda, M. O.; Neisen, B. D.; Cramer, C. J.; Hillmyer, M. A.; et al. Mechanistic Studies of ε-Caprolactone Polymerization by (salen)AlOR Complexes and a Predictive Model for Cyclic Ester Polymerizations. ACS Catal. 2016, 6, 1215−1224. (3) Hormnirun, P.; Marshall, E. L.; Gibson, V. C.; Pugh, R. I.; White, A. J. P. Study of Ligand Substituent Effects on the Rate and Stereoselectivity of Lactide Polymerization Using Aluminum SalenType Initiators. Proc. Natl. Acad. Sci. U. S. A. 2006, 103, 15343−15348. (4) Nomura, N.; Ishii, R.; Yamamoto, Y.; Kondo, T. Stereoselective Ring-Opening Polymerization of a Racemic Lactide by Using Achiral Salen- and Homosalen-Aluminum Complexes. Chem. - Eur. J. 2007, 13, 4433−4451. (5) Zhong, Z.; Dijkstra, P. J.; Feijen, J. Controlled and Stereoselective Polymerization of Lactide: Kinetics, Selectivity, and Microstructures. J. Am. Chem. Soc. 2003, 125, 11291−11298. (6) Chisholm, M. H.; Gallucci, J. C.; Phomphrai, K. Comparative Study of the Coordination Chemistry and Lactide Polymerization of Alkoxide and Amide Complexes of Zinc and Magnesium with a βDiiminato Ligand Bearing Ether Substituents. Inorg. Chem. 2005, 44, 8004−8010. (7) Chen, H. Y.; Peng, Y. L.; Huang, T. H.; Sutar, A. K.; Miller, S. A.; Lin, C. C. Comparative Study of Lactide Polymerization by Zinc Alkoxide Complexes with a β-Diketiminato Ligand Bearing Different Substituents. J. Mol. Catal. A: Chem. 2011, 339, 61−71. F

DOI: 10.1021/acs.inorgchem.8b00250 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry (24) Bickelhaupt, F. M.; Houk, K. N. Analyzing Reaction Rates with the Distortion/Interaction-Activation Strain Model. Angew. Chem., Int. Ed. 2017, 56, 10070−10086. (25) Zhao, Y.; Truhlar, D. G. A New Local Density Functional for Main-Group Thermochemistry, Transition Metal Bonding, Thermochemical Kinetics, and Noncovalent Interactions. J. Chem. Phys. 2006, 125, 194101. (26) Zhao, Y.; Truhlar, D. G. The M06 Suite of Density Functionals for Main Group Thermochemistry, Thermochemical Kinetics, Noncovalent Interactions, Excited States, and Transition Elements: Two New Functionals and Systematic Testing of Four M06-Class Functionals and 12 Other Function. Theor. Chem. Acc. 2008, 120, 215−241. (27) Hehre, W. J.; Ditchfield, R.; Pople, J. A. SelfConsistent Molecular Orbital Methods. XII. Further Extensions of Gaussian Type Basis Sets for Use in Molecular Orbital Studies of Organic Molecules. J. Chem. Phys. 1972, 56, 2257−2261. (28) Krishnan, R.; Binkley, J. S.; Seeger, R.; Pople, J. A. SelfConsistent Molecular Orbital Methods. XX. A Basis Set for Correlated Wave Functions. J. Chem. Phys. 1980, 72, 650−654. (29) Marenich, A. V.; Cramer, C. J.; Truhlar, D. G. Universal Solvation Model Based on Solute Electron Density and on a Continuum Model of the Solvent Defined by the Bulk Dielectric Constant and Atomic Surface Tensions. J. Phys. Chem. B 2009, 113, 6378−6396. (30) Charpentier, B.; Bernardon, J.-M.; Eustache, J.; Millois, C.; Martin, B.; Michel, S.; Shroot, B. Synthesis, Structure-Affinity Relationships, and Biological Activities of Ligands Binding to Retinoic Acid Receptor Subtypes. J. Med. Chem. 1995, 38, 4993−5006. (31) Liu, Z.; Xiang, J. A High Yield and Pilot-Scale Process for the Preparation of Adapalene. Org. Process Res. Dev. 2006, 10, 285−288.

G

DOI: 10.1021/acs.inorgchem.8b00250 Inorg. Chem. XXXX, XXX, XXX−XXX