Comparison of the Catalytic Activity of Mono- and Multinuclear Ga Complexes in the ROCOP of Epoxides and Cyclic Anhydrides

Graphical Abstract

The catalytically active site of tetranuclear Schiff-base complexes L^1–3₂Ga₄(t-Bu)₈ 1–3, which showed high activity and perfect selectivity (>99 %) in the alternating ROCOP of a series of epoxides and anhydrides to polyesters with high M_n and narrow dispersity Đ, was identified.

Abstract

Tetranuclear Schiff-base complexes L^1–3₂Ga₄(t-Bu)₈ 1–3 are highly active and selective (>99 %) catalyst in the alternating ring-opening copolymerization (ROCOP) of epoxides and anhydrides, yielding polyesters with high molecular weights (M_n) and narrow dispersity (Đ). The thermal properties (T_g) of the resulting polyester range from 18 °C to 124 °C and increase with increasing steric bulk or rigidity along the polymer backbone. Comparative studies using structurally related complexes L⁴Ga(t-Bu)₂ 4, [L⁵GaR₂]₂ (R=t-Bu 5, R=Me 6) and L⁶Ga(t-Bu)₂ 7 proved that the Ga₂O₂ core of catalyst 1 is the catalytically active species.

Introduction

Polyesters derived from renewable sources are potential sustainable alternatives to petroleum-based polymers.^1-3 They find applications including packaging, household goods and for the manufacture of biomedical devices due to their degradability and biocompatibility.^4-6 Although polylactide (PLA) and ϵ-polycaprolacton (PCL) are widely used in various technical applications, the properties of the polymers are often constrained by a small monomer pool, limited functional diversity of the substrate scope and unwanted post-polymerization alterations.⁷ These limitations can be overcome by controlled alternating ring-opening copolymerization (ROCOP) of epoxides and cyclic anhydrides, yielding a large variety of sustainable polyesters with adaptable thermal and mechanical properties and functionalities.^{7, 8, 9} Furthermore, many of the epoxides and anhydrides reported for ROCOP are derived from renewable and commercially accessible resources, including propylene oxide (PO), cyclohexene oxide (CHO), succinic anhydride (SA), maleic anhydride (MA), glutaric anhydride (GA), phthalic anhydride (PA), and diglycolic anhydride (DGA), respectively.^10-13

The last decades witnessed the development of not only highly active but also highly selective organometallic ROCOP catalysts,^{14, 15} which clearly outperformed early catalysts based on tertiary amines and SnBu₄.¹⁶ Highly active and selective β-diketiminate zinc catalyst allowed for the synthesis of aliphatic polyesters with high M_n and narrow molecular weight distribution (MWD) under mild reaction conditions.¹⁷ Williams et al. introduced heteroleptic bimetallic Mg and Zn complexes, which showed excellent catalytic activity and selectivity in the ROCOP of cyclic anhydrides and epoxide even in the absence of any nucleophilic co-catalysts.^18-21 Since then, active zinc,^{22, 23} magnesium,^24-26 iron,²⁷ chromium,²⁸ cobalt,^{29, 30} aluminum³¹ and manganese complexes³² were reported, with the cobalt and chromium complexes showing very high activities and regioselectivities in the epoxide-anhydride ROCOP.³³ Binuclear complexes showed far higher catalytic activities than mononuclear species,^{34, 35} indicating a bimetallic reaction mechanism in homopolymerization of epoxides³⁶ and copolymerization of epoxide and CO₂.^37-41 The origin of the higher catalytic activity of binuclear metal complexes in the copolymerization of epoxides and cyclic anhydrides results from a complex interplay of structural features of the catalysts in solution, including the metal size and coordination geometry, the metal-metal distance, M-OR bond strength and nucleophilicity of the initiator group, respectively.⁴²

We recently reported on tetranuclear gallium imino(phenolate) complexes L^1–3₂Ga₄(t-Bu)₈, which showed significantly higher catalytic activities compared to the corresponding Me-substituted complexes L^1–3₂Ga₄(Me)₈ both in the homo- and copolymerization of lactide (LA) and ϵ-caprolactone (ϵ-CL).^{43, 44} We herein extended our studies on copolymerization reactions of several epoxides and cyclic anhydrides (Scheme 1). Detailed mechanistical studies proved that the central Ga₂O₂ core of the tetranuclear complexes represents the catalytically active species.

Scheme 1

L^1–3₂Ga₄(t-Bu)₈ 1–3 and ROCOP of epoxides and anhydrides.

Results and Discussion

Epoxide/Anhydride Copolymerization

Complexes 1–3 were synthesized according to literature procedures⁴³ and their catalytic activity in the ROCOP of four epoxides including cyclohexene oxide (CHO), vinyl cyclohexene oxide (VCHO), cyclopropane oxide (CPO), and propylene oxide (PO) with a series of five cyclic anhydrides, succinic anhydride (SA), maleic anhydride (MA), glutaric anhydride (GA), phthalic anhydride (PA), and diglycolic anhydride (DGA), was investigated (Scheme 1). Copolymerization reactions were performed at 100 °C in toluene in the absence of any co-initiator and [epoxide] : [anhydride] : [cat.] molar ratios of 400 : 400 : 1, and the copolymers were characterized by NMR spectroscopy, GPC and MALDI-TOF-MS (Table 1).

Table 1. Copolymerization of epoxide and cyclic anhydride with tetranuclear complexes 1–3.^[a]

Entry	Monomer 1	Monomer 2	Catalyst	Time [h]	Conv. [%]^[b]	TOF [h^−1]^[c]	Mn^cal(GPC) [kg/mol]^[d]	Mn^b(theo) [kg/mol]^[e]	Đ	Tg [°C]^[f]
1	CHO	SA	1	8	99	100	35.6	78.5	1.3	60
2	CHO	SA	2	8	65	65	23.3	51.5	1.4	–
3	CHO	SA	3	8	82	82	27.6	65.1	1.3	–
4	CHO	MA	1	6	95	130	33.5	74.6	1.4	93
5	CHO	MA	2	8	85	85	30.2	66.8	1.5	–
6	CHO	MA	3	6	88	115	32.8	70.2	1.4	–
7	CHO	GA	1	24	30	10	12.7	25.5	1.7	25
8	CHO	PA	1	5	95	150	42.8	98.5	1.4	118
9	CHO	DGA	1	12	70	45	25.2	59.9	1.3	51
10	VCHO	SA	1	6	93	125	36.6	83.4	1.3	81
11	VCHO	SA	2	6	72	95	29.1	64.5	1.4	–
12	VCHO	SA	3	6	80	105	31.9	71.8	1.3	–
13	VCHO	MA	1	5	96	155	40.2	88.8	1.6	106
14	VCHO	GA	1	24	54	18	22.5	51.4	1.6	39
15	VCHO	PA	1	4	90	180	43.7	98.1	1.5	124
16	VCHO	DGA	1	12	83	55	35.2	79.1	1.4	60

• [a] Polymerization conditions: catalyst=12.73 μmol, in toluene; 100 °C; [CHO]=[Anhydride]=2.5 mmol equiv. to catalyst; [CHO] : [Anhydride] : [cat.]=400 : 400 : 1. [b] Monomer conversion as determined by ¹H NMR analysis. [c] Turnover frequency (TOF) h⁻¹. [d] M_n^GPC values were determined by GPC analysis in THF using PEO standards. [e] Calculated M_n^theo=[M_M1×([M₁]/[cat.])×x_p1+M_M2×([M₂]/[cat.])×x_p2]. [f] Determined by DSC.

Complexes 1–3 are highly active and highly selective catalysts, producing high molecular weight copolymers with narrow polydispersities and >99 % alternating microstructures. Complex 1 is the most active catalyst in all experiments, whereas complex 2 is the less active one as was observed in the homopolymerization of LA and ϵ-CL,⁴³ and the catalytic activity of complexes 1–3 is significantly higher compared to that observed for the corresponding Me-substituted complexes L^1–3₂Ga₄(Me)₈.⁴⁴

CHO/Anhydride Copolymerization

Complexes 1–3 are active catalysts (TOF 100 h⁻¹ 1, 65 h⁻¹ 2, 82 h⁻¹ 3) in the ROCOP of CHO with SA and MA at 100 °C in toluene (entries 1–6, Table 1) even without any co-initiator. High conversions were achieved within 8 h (1 99 %, 3 82 %, 2 65 %), yielding polymers with high molecular weights (31.3–45.6 kg/mol) and narrow dispersity (Đ=M_w/M_n) values (1.3–1.4). Complexes 1–3 showed higher catalytic activities in the polymerization of CHO and MA (TOF 130 h⁻¹ 1, 85 h⁻¹ 2, 115 h⁻¹ 3) under identical condition, most likely caused by the larger ring strain in the anhydride backbone of MA.^{45, 46} The strong influence of the ring strain of the anhydride monomer on the reactivity was also studied in a copolymerization experiment of CHO and PA using complex 1 (entry 8, Table 1), revealing a significantly higher activity with 95 % conversion within 5 h (TOF=150 h⁻¹) and formation of high molecular weight polymer. Both the catalytic activity and selectivity of complex 1 is higher compared to mononuclear and binuclear magnesium catalysts reported by Williams et al.^{47, 48} Copolymerization experiments of CHO with GA and DGA using complex 1 under identical reaction conditions gave lower conversions (GA: 30 % conversion after 24 h; DGA: 70 % conversion after 12 h). The overall catalytic activity of the CHO/anhydride copolymerization with catalyst 1 showed the following order: PA>MA>SA>DGA>GA.

¹H NMR spectra (electronic supplement) revealed the formation of perfectly alternating polyester copolymer chains (>99 %) in all experiments. The formation of polyether linkages can be excluded due to the absence of any resonance in the range of 3.5–3.6 ppm. DOSY NMR experiments of the resulting copolymers furthermore revealed that the proton signals of all monomeric units, which show the same diffusion coefficient, belong to the same polymer chain, thus confirming that this polymer is a copolymer (Figures S25–S27). GPC measurements revealed bimodal molecular weight distributions (MWD) of the corresponding polymers (Figures S28–S30), which is typical for the block copolymerization of epoxide.^{19, 49, 50} The molecular weights of the resulting copolymers are far below the theoretical values (Table 1); such observation is common in epoxide/anhydride copolymerization reactions^{45, 49, 50} due to the likely occurrence of intramolecular transesterification reactions or chain transfer reactions with protic impurities such as water during the propagation process of the polymerization. In addition, these results can also be explained in case that complexes 1–3 act as dual-site catalysts due to the presence of two active metal sites. The end group fidelity of the copolymers was determined by MALDI-TOF-MS analysis of low molecular weight copolymers, which were synthesized by reaction of CHO and the corresponding anhydride (SA, GA) with complex 1 at 100 °C in toluene using a [CHO] : [anhydride] : [cat.] molar ratio of 50 : 50 : 1. The MALDI-TOF-MS spectrum (Figures S31, S32) of the resulting copolymers showed two series of peaks, which are equally spaced by the collective mass of CHO and the corresponding anhydride (198 Da for CHO+SA; 212 Da for CHO+DGA), which is in concurrence with a perfectly alternating microstructure. Analysis of the peak distributions proved the presence of Na⁺ or proton adducts of linear polymer chains with α,ω-OH end groups to a full polymeric repeat unit most probably due to originate from water initiation and a second peak distributions of α,ω-OH end groups with one extra CHO molecule at the chain end in line with an alternating copolymerization mechanism. Furthermore, these results prove the initial coordination of the epoxide to the metal center, followed by a ring opening of the epoxide due to the nucleophilic attack from traces of protic impurities including water. A similar finding was observed for all other copolymerization epoxide and anhydride.

Copolymerization of VCHO with Cyclic Anhydrides

Copolymerization of VCHO with SA using complexes 1–3 with a [epoxide] : [anhydride] : [cat.] molar ratio of 400 : 400 : 1 at 100 °C in toluene (entries 10–12, Table 1) proceeds faster (TOF=125 h⁻¹ 1, 95 h⁻¹ 2, 105 h⁻¹ 3) than the copolymerization of CHO with SA (TOF 100 h⁻¹ 1, 65 h⁻¹ 2, 82 h⁻¹ 3), yielding copolymers with high M_n and narrow Đ values. Comparable results were observed in copolymerization reactions of VCHO with MA, GA, PA and DGA, respectively, using the most active complex 1 under identical reaction conditions (entries 13–16, Table 1). ¹H NMR analysis of the resulting copolymers (Figures S33–S42) confirmed the absence of ether linkages and established a virtually perfect microstructure selectivity (>99 %) toward formation of an alternating copolymer structure. DOSY NMR (Figure S43) and GPC analyses (Figure S44) of the resulting polymers as well as MALDI-TOF-MS spectra (Figure S45) of low molecular weight copolymers, which were synthesized at 100 °C in toluene by reaction of VCHO and MA with complex 1 with a [VCHO] : [anhydride] : [cat.] molar ratio of 50 : 50 : 1, are comparable to results obtained for the copolymerization of CHO with these anhydrides. In all experiments, complex 1 represents the most active and highly selective catalyst for the alternating copolymerization of epoxides and cyclic anhydrides.

Copolymerization of CPO and PO with Cyclic Anhydrides

We finally studied the ROCOP of CPO and PO with all five anhydrides using complex 1 under identical reaction conditions. Despite lower catalytic activities, perfect selectivities were observed, yielding copolymers with high M_n and narrow Đ values (Table 2). The catalytic activity of the epoxides with complex 1 follows the trend: VCHO>CHO>CPO>PO. The copolymerization of the anhydrides with CPO (TOF=64 h⁻¹ SA, 74 h⁻¹ MA, 80 h⁻¹ PA, 20 h⁻¹ GA and 44 h⁻¹ DGA) is faster than the copolymerization with PO (TOF=50 h⁻¹ SA, 57 h⁻¹ MA, 62 h⁻¹ PA, 10 h⁻¹ GA and 33 h⁻¹ DGA), while the same activity order was observed for both epoxides (PA>MA>SA>DGA>GA). ¹H NMR analyses of the resulting copolymers (Figures S46–S55, S61–S70) confirmed the absence of ether linkages and established a virtually of perfect alternating selectivity (>99 %) to the copolymer chain, while DOSY NMR (Figures S56–S57, S71–S72) proved the copolymeric nature of the obtained polyesters. MALDI-TOF-MS spectra (Figures S60, S75–S76) of low molecular weight copolymers, which were synthesized at 100 °C in toluene by reaction of CPO and PO with corresponding anhydrides using complex 1 with a [epoxide] : [anhydride] : [cat.] molar ratio of 50 : 50 : 1, are comparable to those obtained of CHO/anhydride copolymers.

Table 2. Copolymerization of epoxide and cyclic anhydride with complex 1.^[a]

Entry	Monomer 1	Monomer 2	Catalyst	Time [h]	Conv. [%]^[b]	TOF [h^−1]^[c]	Mn^cal(GPC) [kg/mol]^[d]	Mn^b(theo) [kg/mol]^[e]	Đ	Tg [°C]^[f]
1	CPO	SA	1	10	80	64	25.7	58.9	1.3	48
2	CPO	MA	1	10	92	74	30.1	67.1	1.4	71
3	CPO	PA	1	10	99	80	39.4	91.9	1.6	73
4	CPO	GA	1	10	25	20	7.8	19.8	1.3	19
5	CPO	DGA	1	10	55	44	18.9	44.1	1.3	41
6	PO	SA	1	12	75	50	22.1	47.4	1.4	30
7	PO	MA	1	12	86	57	26.8	53.7	1.7	34
8	PO	PA	1	12	93	62	33.2	76.7	1.6	41
9	PO	GA	1	12	16	10	5.1	11.0	1.2	NA
10	PO	DGA	1	12	50	33	15.7	34.8	1.4	18

• [a] Polymerization conditions: catalyst=12.73 μmol, in toluene; 100 °C; [Epoxide]=[Anhydride]=2.5 mmol equiv. to catalyst; [Epoxide] : [Anhydride] : [cat.]=400 : 400 : 1. [b] Monomer conversion as determined by ¹H NMR analysis. [c] Turnover frequency (TOF) mol⁻¹ h⁻¹. [d] M_n^GPC values were determined by GPC analysis in THF using PEO standards. [e] Calculated M_n^theo=[M_M1×([M₁]/[cat.])×x_p1+M_M2×([M₂]/[cat.])×x_p2]. [f] Determined by DSC.

Thermal Properties of the Copolymers

The glass transition temperatures (T_g) of the resulting polyesters were determined by DSC and are listed in Scheme 2. The results are summarized in Tables 1 and 2. All polyesters are amorphous with glass transition temperatures (T_g) ranging from 18 °C to 124 °C. The T_g values of polyesters obtained from copolymerization of CHO with SA (+60 °C), MA (+93 °C), PA (+118 °C), GA (+25 °C) and DGA (+51 °C) are comparable to those obtained by alternating coupling reactions of VCHO with SA (+82 °C), MA (+106 °C), PA (+124 °C), GA (+39 °C) and DGA (+60 °C), respectively (Figures S77–S79). In general, T_g depends on the M_n, crystallinity, and rigidity of the monomeric structure of polyesters as was previously reported.⁴⁵ Remarkably, thermal analyses of our copolymers showed that the steric demand of the pendant groups and rigidity of the monomeric structure substantially influence the thermal property (T_g) of the resulting polyester, since poly(CHO-alt-SA) (M_n=35.6 kg/mol) and poly(VCHO-alt-SA) (M_n=36.6 kg/mol) have close molecular weight whereas poly(VCHO-alt-SA) showed a higher T_g value than poly(CHO-alt-SA) (Table 1, entries 10 and 1). Similarly, poly(VCHO-alt-PA) (M_n=43.7 kg/mol) and poly(CHO-alt-PA) (M_n=42.8 kg/mol) have almost similar molecular weights, while poly(VCHO-alt-PA) showed higher T_g than the corresponding poly(CHO-alt-PA) (Table 1, entries 15 and 8). Poly(CHO-alt-MA) displayed a higher T_g value than poly(CHO-alt-SA) although poly(CHO-alt-MA) (M_n=33.5 kg/mol) has a slightly lower molecular weight compares to poly(CHO-alt-SA) (M_n=35.6 kg/mol) as summarized in Table 1 (entries 10 and 1). Poly(VCHO-alt-PA) (M_n=43.7 kg/mol) also showed a higher T_g value than the corresponding poly(VCHO-alt-MA) (M_n=40.2 kg/mol; Table 1, entries 15 and 13). These results are consistent with previous observations, according to which increasing steric bulk or rigidity along the polymer backbone results in higher T_g values.^{45, 50-52} However, the molecular weight of the polyester plays an important role to the T_g since poly(CHO-alt-GA) (M_n=12.7 kg/mol) and poly(VCHO-alt-GA) (M_n=22.5 kg/mol) have low molecular weight and showed lower T_g value of 25 °C and 39 °C, respectively. Nevertheless, the T_g values of alternating copolymers obtained from the copolymerization of CPO and anhydrides range from +19 °C to +73 °C (GA +19 °C, DGA +41 °C, SA+48 °C, MA +71 °C, PA +73 °C), while those of the copolymers formed by PO and SA (+30 °C), MA (+34 °C), PA (+41 °C), and DGA (+18 °C) are slightly lower (Figures S80, S81). Consequently, the T_g value of poly(CPO-alt-SA) (M_n=25.7 kg/mol) was higher than that of poly(PO-alt-SA) (M_n=22.1 kg/mol; Table 2, entries 1 and 6) and that of poly(CPO-alt-MA) (M_n=30.1 kg/mol) higher compared to poly(CPO-alt-SA) (M_n=25.7 kg/mol; Table 2, entries 2 and 1). Similarly, poly(CPO-alt-DGA) (M_n=18.9 kg/mol) exhibited comparatively higher T_g value than poly(PO-alt-DGA) (M_n=15.7 kg/mol; Table 2, entries 5 and 10). These results furthermore prove that the thermal properties of the copolymers can be tuned to some extent by the nature of the anhydride and the epoxide, for which the T_g values were found to increase with epoxide in the following order: VCHO>CHO>CPO>PO.

Scheme 2

Alternating copolymers obtained from epoxide/anhydride copolymerization reactions with complex 1.

Study of the Active Site of the Catalysts

The tetranuclear complexes 1–3 each contain two general structural motifs, which potentially serve as catalytically active sites: the mononuclear core A containing the N,O chelating Schiff-base ligand and the binuclear core B consisting of a four-membered Ga₂O₂ ring (Figure 1). To elucidate which unit is responsible for the catalytic activity, we synthesized structurally related model complexes with ligands L⁴H and L⁵H. To further compare the activity of the binuclear core B with a related mononuclear species, we expanded the study to monomeric complex t-Bu₂GaOCPh₃ 7 containing ligand L⁶H⁵³ as well as dinuclear phenoxoimine complex 8 containing five-coordinated gallium atoms.

Figure 1

Structural motifs of catalyst 1 (Core A and B) and ligands L^4–6H.

Equimolar reactions of L^4,5H with t-Bu₃Ga in toluene at 25 °C gave complexes 4 and 5 in high yields. Complex 6 was formed by analogous reaction of equimolar amounts of Me₃Ga with L⁵H (Scheme 3) and complex 8 by equimolar reaction of 2,4-di-tert-butyl-6-{[(3-hydroxypropyl) imino]methyl}phenol with Me₃Ga in toluene at 100 °C. Complexes 4–6 and 8 were characterized by heteronuclear NMR (¹H, ¹³C) and IR spectroscopy and their purity confirmed by elemental analysis. The ¹H and ¹³C NMR spectra of complexes 4–6 show the expected resonances of the ligands. Interestingly, the ¹H NMR spectrum of complex 6 shows two sets of resonances for the ligand (L⁵), whose intensity changes with temperature as was proven by variable-temperature (VT) ¹H NMR study from −40 °C to 80 °C (Figure S13). We assign these findings to the formation of a dimer/trimer equilibrium in solution, which shifts with increasing temperature from the trimer to the dimer (entropy effect). The ¹H and ¹³C NMR spectra of complex 8 show the expected resonances of the ligand. The methylene proton resonances of the 1-propanolate unit (OCH₂CH₂CH₂N) each split into two multiplets, revealing their diastereotopic nature upon coordination of the ligand to the gallium atom as was reported for tetranuclear zinc and magnesium complexes (Figure S10).⁵⁴

Scheme 3

Synthesis of complexes 4–8.

Single Crystal X-ray Diffraction Studies

Single crystals of complexes 4–6 and 8 were grown from saturated toluene solutions at 0 °C (4, 8) and −30 °C (5, 6). Complex 4 (Figure 2) crystallizes in the monoclinic space group P2₁ with two molecules in the unit cell and complexes 5 (Figure 3), 6 (Figure 4) and 8 (Figure 5) in the triclinic space group with one (5) and two molecules (6, 8) in the unit cell. Crystallographic data are summarized in Table S1.

Figure 2

ORTEP representation of solid-state structure of the complex 4. H atoms have been omitted for clarity and thermal ellipsoids are shown with 50 % probability level. Selected bond lengths [Å] and angles [°]: Ga1-O1 1.8985(12), Ga1-C19 2.0011(17), Ga1-C23 2.0056(17), Ga1-N1 2.0196(14), O1-Ga1-C19 104.70(6), O1-Ga1-C23 107.50(6), C19-Ga1-C23 126.53(7), O1-Ga1-N1 92.77(5), C19-Ga1-N1 115.15(6), C23-Ga1-N1 104.72(6).

Figure 3

ORTEP representation of solid-state structure of the complex 5. H atoms and the second orientation of the disordered CH₂CH₂CH₂Ph group have been omitted for clarity and thermal ellipsoids are shown with 50 % probability level. The part in pale colours was generated by symmetry. Selected bond lengths [Å] and angles [°]: Ga1-O1 1.952(2), Ga1-C10 2.014(3), Ga1-C14 2.017(4), O1-Ga1-O1′ 76.67(10), O1-Ga1-C10 111.97(12), O1′-Ga1-C10 110.86(13), O1-Ga1-C14 112.15(13), O1′-Ga1-C14 112.67(13), C10-Ga1-C14 123.18(15).

Figure 4

ORTEP representation of solid-state structure of the complex 6. H atoms have been omitted for clarity and thermal ellipsoids are shown with 50 % probability level. Selected bond lengths [Å] and angles [°]: Ga1_-1-O1_-1 1.9379(9), Ga1_-1-O1_-3 1.9466(9), Ga1_-2-O1_-1 1.9435(9), Ga1_-2-O1_-2 1.9540(9), Ga1_-3-O1_-2 1.9301(9), Ga1_-3-O1_-3 1.9360(9), O1_-1-Ga1_-1-O1_-3 96.81(4), O11-Ga11-C111 106.67(5), O1_-1-Ga1_-2-O1_-2 95.76(4), O1_-2-Ga1_-3-O1_-3 93.01(4), C10_-2-Ga1_-2-C11_-2 125.19(6).

Figure 5

ORTEP representation of solid-state structure of the complex 8. H atoms have been omitted for clarity and thermal ellipsoids are shown with 50 % probability level. Only one of two independent molecules is displayed. The part in pale colors is generated by symmetry. Selected bond lengths [Å] and angles [°]: Ga1_-1-O1_-1 1.9178(8), Ga1_-1-O2_-1 1.9501(8), Ga1_-1-O2_-1 1.9992(8 Ga1_-1-C19_-1 1.9716(11), Ga1_-1-N1_-1 2.0820(9), O1_-1-Ga1_-1-O2_-1 129.48(4), O1_-1-Ga1_-1-N1_-1 88.27(3), O2_-1-Ga1_-1-N1_-1 87.54(3), Ga1_-2-O1_-2 1.9546(8), Ga1_-2-O2_-2 2.0516(8), Ga1_-2-O2_-2 1.9005(8), Ga1_-2-C19_-2 1.9674(12), Ga1_-2-N1_-2 2.0284(9), O1_-2-Ga1_-2-O2_-2 153.79(4), O1_-2-Ga1_-2-N1_-2 88.54(3), N1_-2-Ga1_-2-O2_-2 86.19(3).

The Ga atoms in complexes 4–6 each adopt distorted tetrahedral coordination spheres, and the central structural parameters within the ligands as well as the Ga−O, Ga−C and Ga−N bond lengths are within the expected ranges. Complex 5 forms a centro-symmetric dimer with a planar four membered Ga₂O₂ ring, whereas complex 6 shows a non-planar six-membered Ga₃O₃ ring in the solid state with O1_-3 and Ga1_-2 being the “bow” and “stern” of the boat-type conformation. The Ga−O bond lengths within the Ga₂O₂ ring in complex 5 (1.952(2) Å, 1.961(2) Å) and the six-membered Ga₃O₃ ring in complex 6, which range from 1.9301(9) Å to 1.9540(9) Å, are approximately equal and elongated compared to the Ga−O bond lengths in complex 4 (1.8985(12) Å). The Ga atoms in complex 8 adopt distorted square-pyramidal coordination spheres, sharing one of the base edges with its symmetry equivalent, forming a planar four-membered Ga₂O₂ ring. The Ga−O bond lengths are comparable to those of 4–6.

A Comparative Catalytic Study

The catalytic activities of complexes 4 and 5, which serve as structural model systems for core A and core B of catalyst 1, were tested in copolymerization reactions of CHO with both SA and PA under the identical reaction conditions as applied for the copolymerization of CHO with SA and PA using complex 1. Complex 4 showed poor catalytic activity to the copolymerization and only 10 % conversion was achieved for CHO with SA (TOF=3 h⁻¹, Table 3, entry 3) and 14 % for CHO with PA (TOF=5 h⁻¹ Table 3, entry 4) after 24 hours, while the copolymer chain contained 75 % and 84 % of poly(ester) linkage. In contrast, complex 5 showed excellent catalytic activity in the copolymerization reactions and 96 % conversion was achieved for CHO with SA (TOF=96 h⁻¹, Table 3, entry 5) after 8 hours and 90 % for CHO with PA (TOF=144 h⁻¹, Table 3, entry 6) after 5 hours, respectively. In both cases high molecular weight copolymers with a narrow polydispersity were obtained and the copolymers chain contained more than 99 % poly(ester) linkage. According to these findings we propose that core B in complex 1 is the catalytically active site in the ROCOP of epoxides and anhydrides. Complex 5 also shows higher catalytic activities in the copolymerization reactions when compared to the Me-substituted complex 6, for which 80 % conversion was achieved after 8 hours in the CHO/SA copolymerization (TOF=80 h⁻¹, Table 3, entry 7) and 75 % conversion after 5 hours in the CHO/PA copolymerization (TOF=120 h⁻¹, Table 3, entry 8), respectively. A comparable activity trend was found for L¹₂Ga₄(t-Bu)₈ 1 and the Me-substituted counterpart L¹₂Ga₄(Me)₈.⁴⁴ In contrast, complex 7 is far less active in the ROCOP of CHO with both SA and PA, achieving only 27 % conversion after 8 hours in the CHO/SA copolymerization (TOF=27 h⁻¹, Table 3, entry 9) and 30 % after 5 hours in the CHO/PA copolymerization (TOF=48 h⁻¹, Table 3, entry 10). Since both binuclear complexes 5 and 6 showed much higher catalytic activities compared to the mononuclear complex 7, we propose a bimetallic reaction mechanism as was reported for epoxide and anhydride homo and copolymerization reactions with bimetallic complexes.^55-57

Table 3. Comparative studies for the copolymerization of CHO with SA and PA using complexes 1 and 4–8.^[a]

Entry	Monomer-1	Monomer-2	Cat.	Time [h]	Conv. [%]^[b]	TOF [h^−1]^[c]	Poly(ester) linkage [%]^[d]	Mn^cal(GPC) [kg/mol]^[e]	Đ
1	CHO	SA	1	8	99	100	99	35.6	1.3
2	CHO	PA	1	5	95	150	99	42.8	1.4
3	CHO	SA	4	24	10	3	75	4.6	1.5
4	CHO	PA	4	24	14	5	84	6.8	1.7
5	CHO	SA	5	8	96	96	99	33.9	1.3
6	CHO	PA	5	5	90	144	99	37.4	1.3
7	CHO	SA	6	8	80	80	99	30.3	1.4
8	CHO	PA	6	5	75	120	99	35.2	1.5
9	CHO	SA	7	8	27	27	80	6.7	1.5
10	CHO	PA	7	5	30	48	75	9.5	1.6
11	CHO	SA	t-Bu3Ga	8	42	42	85	13.8	1.7
12	CHO	SA	8	8	60	60	99	23.5	1.5

• [a] Polymerization conditions: catalyst=12.73 μmol, in toluene; 100 °C; [Epoxide]=[Anhydride]=2.5 mmol equiv. to catalyst; [Epoxide] : [Anhydride] : [cat.]=400 : 400 : 1. [b] Monomer conversion as determined by ¹H NMR analysis. [c] Turnover frequency (TOF) h⁻¹. [d] Poly(ester) linkages as determined by ¹H NMR analysis. [e] M_n^GPC values were determined by GPC analysis in THF using PEO standards.

Complex 8 was also tested in the copolymerization reactions of CHO with SA in a [CHO] : [SA] : [cat] molar ratio of 400 : 400 : 1 under identical reaction condition as applied for the copolymerization using complex 5 and 6. Complex 8 showed good catalytic activity (TOF=60 h⁻¹, Table 3, entry 12). 60 % conversion was achieved after 8 hours, and the copolymer chain contained 99 % of poly(ester) linkage. The copolymerization experiments show that complex 8 is catalytic more active and selective than the mononuclear complexes 4 and 7 but shows a lower catalytic activity in the copolymerization reactions compared to the dinuclear complexes 5 and 6. However, the Ga atoms in complexes 5 and 6 are four-coordinated and each adopt distorted tetrahedral coordination spheres, whereas the Ga atom in complex 8 is five-coordinated and adopts a distorted square-pyramidal coordination sphere. We also extended the copolymerization study to the reaction of cyclohexene oxide (CHO) with succinic anhydride (SA) using t-Bu₃Ga with a [CHO] : [SA] : [t-Bu₃Ga] molar ratio of 400 : 400 : 1 at 100 °C in toluene. t-Bu₃Ga showed moderate catalytic activity (TOF=42 h⁻¹, Table 3, entry 11). 42 % conversion was achieved after 8 hours, and the copolymer chain contained 85 % of poly(ester) linkage. The copolymerization result prove that t-Bu₃Ga is more reactive than the mononuclear complex 7 but exhibits lower reactivity and selectivity compared to the binuclear complexes 1, 5 and 6.

To further prove that the binuclear Ga₂O₂ core of these complexes do not decompose to mononuclear Ga complexes after monomer coordination in polymerization process, we investigated complex 2 in a CHO/SA copolymerization reaction with a [CHO] : [SA] : [2] molar ratio of 100 : 100 : 1 at 100 °C in toluene-d₈ in a J-Young NMR tube. The reaction progress was monitoring by ¹H NMR spectroscopy at 100 °C for 1 hour. The analysis of ¹H NMR spectra (Figures S82, S83) revealed that complex 2 do not decompose after addition of the monomers. The copolymer formation started immediately as proven by the increasing copolymer concentration (poly(CHO-alt-SA)) and decreasing monomer concentrations. We further compared the ¹H NMR spectrum of complex 2 in toluene-d₈ at 100 °C and the ¹H NMR spectra of the copolymerization reaction solution of complex 2 in toluene-d₈ at 100 °C after addition of the monomers. The proton resonances of complex 2 in the polymerization study are slightly shifted to lower field compared to pure complex 2, while the proton resonance of the t-Bu groups are shifted to higher field, most likely due to the coordination of the monomer to the Ga atom. The copolymerization reaction was monitored by ¹H NMR spectroscopy for one hour and ¹H NMR spectra were recorded every 10 minutes. All proton resonances of complex 2 remain identical within this period of time, proving that complex 2 do not decompose during the copolymerization process.

The experimental results from the comparative studies using complexes 1–3 as well as the model complexes 4–8 are fully consistent with the assumption, that the binuclear Ga₂O₂ core (core B) in complexes 1–3 represent the catalytically active site. However, the different activities of complexes 1–3 indicate that the Schiff-base ligands also have a distinct influence on the catalytic activity of complexes 1–3, since their catalytic activity and selectivity was found to be influenced by the electronic and steric nature of the Schiff-base ligand. The Schiff-base ligands hence do not only serve a simple spectator ligand, even though their role in the polymerization reactions is yet not clear.

Conclusion

Tetranuclear gallium complexes 1–3 showed high catalytic activities and perfect selectivities (>99 %) in the ROCOP of a series of epoxides and cyclic anhydrides without using any co-catalyst. The resulting polyesters contain alternating monomer units and show high M_n values and narrow dispersity. The catalytic activity of the epoxides (VCHO>CHO>CPO>PO) and the anhydrides (PA>MA>SA>DGA>GA) with the most active catalyst 1 strongly depends on the ring strain and steric demand of the monomers. The resulting amorphous polyesters showed T_g ranging from 18 °C to 124 °C. Comparative studies using structurally related model complexes revealed that the binuclear Ga₂O₂ core (core B) in complexes 1–3 represent the catalytically active site. The enhancement of catalytic activity and selectivity for the binuclear complexes is likely attributed to the occurrence of double metal effects.

Experimental Section

General experimental details. All reactions were performed under a dry argon atmosphere using standard Schlenk and glovebox techniques. Argon gas was purified by passing the gas through preheated Cu₂O pellets and molecular sieves columns. n-Hexane was dried using a mBraun Solvent Purification System, degassed, and stored in Schlenk flasks under argon atmosphere, whereas toluene was dried by heating under reflux for 12 h over sodium and benzophenone and freshly distilled prior to use. Deuterated solvents were dried over activated molecular sieves (4 Å) and degassed prior to use. Karl Fischer titration of the solvents showed water contents below 2 ppm. t-Bu₃Ga, Schiff-base complexes 1–3⁴³ as well as complex 7⁵³ were synthesized according to literature procedures. ¹H and ¹³C NMR spectra were recorded at 297 K in CDCl₃, C₆D₆ or toluene-d₈ solution using a Bruker Avance 300 spectrometer with a QNP probe head (¹H: 300 MHz, ¹³C: 75 MHz) or Bruker Avance 400 (¹H: 400 MHz, ¹³C: 125 MHz). ¹H and ¹³C NMR spectra were referenced to internal CDCl₃ (¹H=7.26 ppm, ¹³C=77.16 ppm), C₆D₆ (¹H=7.16 ppm, ¹³C=128.16 ppm), and toluene-d₈ (¹H=7.00 ppm, ¹³C=128.33 ppm), respectively. IR spectra were recorded in a glovebox under Ar atmosphere using an ALPHA–T FT-IR spectrometer equipped with a single reflection ATR sampling module. Elemental analyses were performed with a Perkin Elmer Series 11 analyzer at the Elemental Analysis Laboratory of the University Duisburg-Essen. Melting points were determined in sealed glass capillaries and are not corrected. The number-average molecular weight (M_n), molecular weight distribution (MWD) and dispersity (Đ) of all polylactide polymers were determined by Gel Permeation Chromatography (GPC) on a 1260 Infinity instrument (Polymer Standard Service, Mainz) equipped with 3 SDV columns (pore sizes 10⁶, 10⁵, 10³ Å) using HPLC grade THF as eluent at a flow rate of 1.0 mL min⁻¹ at 40 °C (column oven TCC6000). M_n and Đ of the resulting polymers were determined relative to 12 narrow molecular weight PEO standards (Polymer Standard Service, Mainz) and the WinGPC UniChrom software. Matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF-MS) measurements were carried out on a Bruker UltrafleXtreme MALDI-TOF mass spectrometer (BrukerDaltonik, Bremen, Germany). Two different procedures for the measurements of MALDI-TOF-MS were applied: a) Poly(CHO-alt-GA), poly(VCHO-alt-MA) and poly(CPO-alt-MA) copolymers were dissolved in THF (10 mg/mL) and dihydroxybenzoic acid was used as a matrix in THF (20 mg/mL). b) Poly(CHO-alt-SA), poly(PO-alt-GA) and poly(PO-alt-DGA) copolymer were dissolved in THF (10 mg/mL), and 5 μL of the solution was mixed with 5 μL of trans-2-[3-(4-tert-butylphenyl)-2-methyl-2-propenylidene]malononitrile (DCTB) (20 mg/mL THF) and 0.1 μL of sodium trifluoroacetate (10 mg/mL THF). A droplet of as-prepared samples was thereof applied on the sample target. The reflective positive ion mode was used to acquire the mass spectra of the samples. The calibration was done externally with poly(methyl methacrylate) standards using the nearest-neighbor positions. Differential scanning calorimetry (DSC) data were collected using a DSC 3+/700/866/Argon from Mettler Toledo with a heating and cooling rate of 10 K min⁻¹. Salicylaldehyde, amines, calcium hydride, epoxides and anhydrides were purchased from Sigma-Aldrich and used as received.

L⁴Ga(t-Bu)₂ (4). A solution of L⁴H (400 mg, 1.45 mmol) in 5 mL of toluene was added dropwise at −30 °C to a stirred solution of t-Bu₃Ga (350 mg, 1.45 mmol) in 5 mL of toluene. The solution was stirred for 5 minutes at −30 °C, warmed to ambient temperature and stirred for additional 3 h. The solvent was evaporated under reduced pressure to yield a yellow solid, which was recrystallized from a saturated solution in toluene at 0 °C, isolated by filtration and dried under vacuum. Yield: 610 mg (58 %). Mp 110 °C. Anal. calc. For C₂₆H₄₆GaNO: C, 68.13; H, 10.12; N, 3.06; found: C 68.34, H 9.89, N 3.19. ¹H NMR (400 MHz, toluene-d₈, 25 °C): δ=0.68 (t, ³J_H-H=10 Hz, 3 H, CH₃), 1.26 (s, 18 H, C(CH₃)₃), 1.29 (s, 9 H, C(CH₃)₃), 1.36–1.49 (m, 2 H, CH₂), 1.61 (s, 9 H, C(CH₃)₃), 3.01 (t, ³J_H-H=10.4 Hz, 2 H, NCH₂), 6.79 (d, ⁴J_H-H=3.6 Hz, 1 H, Ar-H), 7.55 (s, 1 H, CH=N), 7.61 (d, ⁴J_H-H=3.6 Hz, 1 H, Ar-H). ¹³C NMR (100 MHz, toluene-d₈, 25 °C): δ=11.26 (CH₃), 23.28 (Me₃C), 23.80 (CH₂), 29.94 (Me₃C), 30.78 (Me₃C), 31.49(Me₃C), 34.02 (Me₃C), 35.69 (Me₃C), 59.41 (CH₂N), 117.71 (Ar-C), 131.47 (Ar-C), 136.81 (Ar-C), 137.42 (Ar-C), 141.10 (Ar-C), 165.91 (Ar-C), 171.45 (Ar-CH=N). IR: ν=2946, 2913, 2867, 2832, 1616, 1538, 1458, 1434, 1412, 1386, 1358, 1320,1255, 1170, 1136, 1106, 1051, 1020, 890, 838, 810, 785, 746, 716, 632, 538, 517, 483, 442, 389 cm⁻¹.

L⁵₂Ga₂(t-Bu)₄ (5). A solution of one equivalent L⁵H (400 mg, 2.94 mmol) in 5 mL of toluene was added dropwise at −30 °C to a stirred solution of one equivalent of t-Bu₃Ga (708 mg, 2.94 mmol) in 5 mL of toluene. The solution was stirred for 5 minutes at −30 °C and for 3 h at ambient temperature (25 °C). The solvent was evaporated under reduced pressure to yield a white solid, which was recrystallized from a saturated solution in toluene at 0 °C, isolated by filtration and dried under vacuum. Yield: 885 mg (94 %). Mp 112 °C. Anal. calc. For C₃₄H₅₈Ga₂O₂: C, 63.98; H, 9.16; found: C 63.57, H 9.36. ¹H NMR (400 MHz, toluene-d₈, 25 °C): δ=1.42 (s, 36 H, C(CH₃)₃), 2.05–2.13 (m, 4 H, CH₂), 2.55 (t, ³J_H-H=7.6 Hz, 4 H, CH₂), 3.96 ((t, ³J_H-H=7.6 Hz, 4 H, OCH₂), 7.16–7.19 (m, 6 H, Ar-H), 7.25–7.29 (m, 4 H, Ar-H). ¹³C-NMR (100 MHz, toluene-d₈, 25 °C): δ=26.30 (Me₃C), 31.85 (Me₃C), 32.19 (CH₂), 36.62 (CH₂), 66.97 (CH₂O), 126.34 (Ar-C), 128.69 (Ar-C), 137.44 (Ar-C), 141.29 (Ar-C). IR: ν=3078, 3051, 3016, 2921, 2857, 2751 2687, 1601, 14,89, 1460, 1441, 1383, 1356, 1255, 1178, 1074, 1042, 1019,976, 936, 852, 814, 748, 698, 580, 535, 487, 453 cm⁻¹.

L⁵₂Ga₂Me₄ (6). Compound 6 was synthesized by an identical procedure as applied for the synthesis of complex 5, using L⁵H (400 mg, 2.94 mmol) and Me₃Ga (337 mg, 2.94 mmol). Yield: 630 mg (91 %). Mp 104 °C. Anal. Calc. for C₃₃H₅₁Ga₃O₃: C, 56.23; H, 7.29. Found: C, 56.71; H, 7.64. ¹H NMR (400 MHz, toluene-d₈, 25 °C K) [Trimer]: δ=−0.140 (t, 18 H, CH₃), 1.69–1.75 (m, 6 H, CH₂), 2.40 (t, ³J_H-H=8 Hz, 6 H, CH₂), 3.67 (t, ³J_H-H=6 Hz, 6 H, OCH₂, 7.00–7.15 (m, 15 H, Ar-H). [Dimer]: −0.08 (s, 12 H, CH₃), 1.59–1.64 (m, 4 H, CH₂), 2.49 (t, ³J_H-H=7.6 Hz, 4 H, CH₂), 3.47 (t, ³J_H-H=6 Hz, 4 H, OCH₂), 7.00–7.15 (m, 10 H, Ar-H). ¹³C-NMR (100 MHz, toluene-d₈, 25 °C ): [Trimer]: δ=−5.70 (Me), 32.26 (CH₂), 35.51 (CH₂), 64.83 (OCH₂), 126.17 (Ar-C), 128.61 (Ar-C), 137.43 (Ar-C), 141.58 (Ar-C). [Dimer]: δ=−6.64 (Me), 32.32 (CH₂), 35.48 (CH₂), 63.37 (OCH₂), 126.24 (Ar-C), 128.69 (Ar-C), 137.43 (Ar-C), 141.79 (Ar-C). IR: ν=3078, 3052, 3017, 2953, 2935, 2924, 2911, 2858, 1601, 1489, 1460, 1441, 1383, 1364, 1356, 1256, 1202, 1075, 1027, 995, 976, 939, 910, 819, 814, 748, 698, 595, 532, 501 467, 423 cm⁻¹.

C₃₈H₆₀Ga₂N₂O₄ (8). A solution of one equivalent of 2,4-di-tert-butyl-6-{[(3-hydroxypropyl)imino]methyl}phenol (400 mg, 1.37 mmol) in 5 mL of toluene was added dropwise at ambient ttemperature (25 °C) to a stirred solution of one equivalent of Me₃Ga (158 mg, 1.37 mmol) in 5 mL of toluene. The solution was stirred for 5 minutes at 25 °C and for 12 h at 100 °C. The solvent was evaporated under reduced pressure to yield a yellow solid, which was recrystallized from a saturated solution in toluene at 0 °C, isolated by filtration and dried under vacuum. Yield: 445 mg (87 %). Mp 175 °C. Anal. calc. For C₃₈H₆₀Ga₂N₂O₄: C, 60.99; H, 8.08; N, 3.74; found: C 61.20; H 8.24; N, 3.70. ¹H NMR (400 MHz, toluene-d₈, 25 °C): δ=0.04 (s, 6 H, CH₃), 1.36 (s, 18 H, C(CH₃)₃), 1.43–1.47 (m, 2 H, CH₂), 1.55 (s, 18 H, C(CH₃)₃), 1.78–1.86 (m, 2 H, CH₂), 2.85–2.95 (m, 2 H, CH₂), 3.57 (t, ³J_H-H=11.6 Hz, 2 H, CH₂), 4.04 (t, ³J_H-H=9.2 Hz, 2 H, OCH₂), 4.56–4.60 (m, 2 H, OCH₂), 6.72 (d, ⁴J_H-H=2.8 Hz, 2 H, Ar-H), 7.27 (s, 2 H, N=CH), 7.55 (d, ⁴J_H-H=2.4 Hz, 2 H, Ar-H). ¹³C-NMR (100 MHz, toluene-d₈, 25 °C): δ=−9.99 (CH₃), 29.77 (Me₃C), 31.68 (Me₃C), 32.27 (CH₂), 34.04 (Me₃C), 35.57 (Me₃C), 61.93 (CH₂), 65.22 (CH₂O), 117.30 (Ar-C), 128.46 (Ar-C), 129.46 (Ar-C), 136.56 (Ar-C),140.22 (Ar-C). 165.28 (Ar-C), 129.46 (Ar-C), 167.60 (C=N). IR: ν=2943, 2895, 2855, 1633, 1613, 1544, 1530, 1457, 1408, 1381, 1353, 1318, 1298, 1267, 1254, 1239, 1194, 1166, 1081 1071, 1022, 1003, 977, 929, 873, 832, 804, 780, 743, 721,702, 628, 564, 530, 517, 462, 444 cm⁻¹.

General procedure for the copolymerization of anhydrides and epoxide using complex 1–7. The anhydride (5 mmol), epoxide (5 mmol) and catalyst (12.7 μmol) were dissolved in 5 ml of toluene and stirred at 100 °C for the desired time (Table 1, Table 2 and Table 3). The polymerization reaction was terminated by adding acidic methanol (10 % HCl) followed by removal of all volatiles under dynamic vacuum, yielding a viscous residue. A small aliquot was taken from the viscous residue to determine the conversion rate of the monomer by ¹H NMR analysis. The resulting residue was then dissolved in a minimum amount of CH₂Cl₂ and poured into 10 % acidic (HCl) methanol, immediately resulting in the precipitation of the polymer, which was isolated by filtration, dried under dynamic vacuum to constant weight and used for GPC and NMR analysis.

Single-crystal X-ray analyses. The crystals were mounted on nylon loops in inert oil. Data were collected on a Bruker AXS D8 Venture diffractometer (4, 5) with Photon II detector (mono-chromated Cu_Ka radiation, λ=1.54178 Å, microfocus source) and on a Bruker AXS D8 Kappa diffractometer (6, 8) with APEX2 detector (mono-chromated Mo_Ka radiation, λ=0.71073 Å) at 100(2) K. The structures were solved by Direct Methods (SHELXS-97)⁵⁸ and refined anisotropically by full-matrix least-squares on F² (SHELXL-2017).^{59, 60} Absorption corrections were performed semi-empirically from equivalent reflections on basis of multi-scans (Bruker AXS APEX2). Hydrogen atoms were refined using a riding model or rigid methyl groups. The absolute structure parameter of 4 suggests twinning by inversion with a small second component, however a refinement as inversion twin yielded a BASF of zero within standard uncertainty for the minor component, thus this model was discarded.⁶¹ The crystal of 5 was a non-merohedral twin of two components and the model refined against de-twinned HKLF4 data. The phenyl propanoate ligand in 5 is disordered over two positions. The bond lengths and angles of the phenyl ring were restrained to be equal (SADI) and RIGU restraints were applied to the displacement parameters of its atoms. In 8 a tert-butyl group shows rotational disorder. Two alternate positions were identified and refined; however, the displacement ellipsoids suggest that this is still a rather crude model. The corresponding displacement parameters were restrained with RIGU and SIMU.

Deposition Numbers 2110980 (4), 2110981 (5), 2110982 (6), and 2118668 (8) contain the supplementary crystallographic data for this paper. These data are provided free of charge by the joint Cambridge Crystallographic Data Centre and Fachinformationszentrum Karlsruhe Access Structures service www.ccdc.cam.ac.uk/structures.

Acknowledgements

Financial support by the DFG through the Emmy-Noether Program (A.H.G.; GR 5075/2-1) and the University of Duisburg-Essen (S.S.) is acknowledged. We thank Y. Schulte (University of Duisburg-Essen) for the collection of the X-ray data of 4 and 5. Open Access funding enabled and organized by Projekt DEAL.

Conflict of interest

The authors declare no conflict of interest.

Open Research

Data Availability Statement

The data that support the findings of this study are available in the supplementary material of this article.