Catalyst Design

As a reference point, a set of dibenzhydryl[ ²³ , ²⁴ ] (Ar₂CH‐) substituted salicylaldimines covering a range of steric and electronic properties were generated, and converted to corresponding Ni^II complexes (Ni2 to Ni10, Scheme 2; for full synthesis and characterization of metal complexes and ligands cf. the Supporting Information, SI). Additionally, the prototypical reference Ni1 was prepared by a reported procedure (Scheme 2). ^[17]

Scheme 2

Reference Ni^II catalysts studied.

Without addition of any activating cocatalysts, these neutral single‐component Ni^II catalysts were evaluated for ethylene polymerization under 8 bar at various temperatures. Despite their considerably different steric and electronic substitution patterns, Ni1–Ni10 did not differ dramatically in their catalytic properties (Table 1 and Figure 1; cf. SI for full polymerization data).

Figure 1

Polymer molecular weights produced by Ni1–Ni12 at 30 °C and 90 °C, respectively (8 bar ethylene).

Table 1

Ethylene polymerization with reference Ni^II catalysts.^[a]

Entry	cat.	T [°C]	Yield [g]	act. [×106][b]	M n [×104][c]	M w [×104][c]	M w/M n [c]	brs[d]	T m [e] [°C]
1	Ni1	30	0.72	0.58	11.8	16.5	1.40	3	129
2	Ni1	90	0.56	0.45	0.6	0.8	1.44	63	/
3	Ni2	30	1.94	1.55	27.3	31.9	1.17	3	134
4	Ni2	90	1.33	1.06	0.6	1.1	1.77	61	81
5	Ni3	30	0.61	0.49	8.5	12.6	1.47	3	135
6	Ni3	90	3.31	2.65	0.4	0.4	1.15	43	101
7	Ni4	30	0.88	0.70	14.6	19.0	1.30	3	134
8	Ni4	90	2.18	1.74	0.7	1.2	1.55	65	80
9	Ni5	30	0.63	0.50	15.0	18.2	1.21	2	135
10	Ni5	90	2.53	2.02	0.9	1.2	1.35	63	79
11	Ni6	30	0.84	0.67	17.8	22.0	1.24	3	136
12	Ni6	90	2.32	1.86	0.6	0.9	1.46	60	85
13	Ni7	30	1.60	1.28	25.6	30.4	1.19	2	133
14	Ni7	90	2.29	1.83	0.7	1.3	1.95	62	78
15	Ni8	30	0.83	0.66	14.3	17.7	1.24	10	133
16	Ni8	90	2.53	2.02	0.7	1.4	1.92	53	85
17	Ni9	30	0.94	0.75	14.0	18.7	1.34	6	129
18	Ni9	90	1.22	0.98	0.3	0.5	1.48	84	/
19	Ni10	30	2.31	1.85	22.5	31.4	1.40	11	118
20	Ni10	90	0.80	0.64	1.1	1.8	1.70	55	/
21	Ni11	30	2.93	2.34	57.4	67.2	1.17	3	132
22	Ni11	50	3.94	3.15	39.9	70.4	1.77	14	123
23	Ni11	70	1.87	1.50	32.4	53.1	1.64	23	106
24	Ni11	90	1.65	1.32	9.5	16.9	1.78	46	84
25	Ni12	30	2.00	1.60	63.7	73.9	1.16	4	128
26	Ni12	50	1.75	1.40	44.2	71.9	1.63	17	110
27	Ni12	70	0.95	0.76	18.5	32.8	1.77	42	83
28	Ni12	90	0.78	0.62	10.0	17.6	1.76	74	/
29[f]	Ni11	50	4.61	3.69	36.1	56.0	1.55	14	118

[a] Reaction conditions: Ni catalyst (5 μmol), toluene (100 mL), ethylene (8 bar), polymerization time (15 min), all entries are based on at least two runs, unless noted otherwise. [b] Activity is in unit of 10⁶ g mol⁻¹ h⁻¹. [c] Determined by GPC in 1,2,4‐trichlorobenzene at 150 °C using a light scattering detector. [d] brs=Number of branches per 1000C, as determined by ¹H NMR spectroscopy. [e] Determined by DSC (second heating). [f] THF (100 mL) as solvent.

To rationalize these findings, steric maps based on single crystal structures were optimized by a DFT method (BP86‐D3 method, cf. SI for computational details) and the buried volume (V_bur) as a measure of the steric environment around the active Ni^II site was calculated ^[25] (Figure 2). Notably, despite their sterically demanding substituents Ni8 (p‐Ph‐Ph) and Ni9 (Np) feature a similar—even slightly lower—V_bur% value than Ni2 (Ph) (45 % and 46 % vs. 48 %).

Figure 2

Steric maps of Ni2, Ni3, Ni8, Ni9, and Ni11 optimized by a DFT computational method (BP86‐D3).

Viewing the substituents in single crystal maps, we find that due to their free rotation, these bulky substituents such as Ph, p‐Ph‐Ph, and Np consistently orient away from the Ni^II active center (cf. SI. Notably, solution NMR data of Ni2 confirms free rotation of the phenyl groups around the CH‐Ph bond, cf. SI). As a consequence, shielding of the axial sites to retard β‐H elimination is limited. Most notably, the axial sites above and below the square planar coordinated Ni^II center are not both shielded (Figure 2 a–d).

To appropriately restrict conformational freedom and force an aryl ring into each apical position, we sought to annulate a given dibenzhydryl moieties’ aryl groups to a suitable mutual ring structure. In view of its considerable tension, a seven‐membered ring in the form of a dibenzosuberyl motif ^[26] was targeted.

Remarkably, the key 2,6‐disubstituted aniline is accessible in a single step from commercially available compounds in good yield (70 %, Scheme 3). This contrasts the multi‐step synthesis of the aniline component for state‐of‐the‐art N‐terphenyl‐ and N‐8‐arylnaphthyl‐ salicylaldiminato Ni^II catalysts[ ²¹ , ²² , ²⁷ , ²⁸ ] (cf. Scheme 1).

Scheme 3

Synthesis of C _s‐shielded neutral salicylaldiminato Ni^II catalysts.

Condensation with salicylaldehydes, and reaction with [(tmeda)NiMe₂] and pyridine yielded dibenzosuberyl‐based 9‐anthracenyl‐ and meta‐terphenyl‐substituted salicylaldiminato Ni^II complexes Ni11 and Ni12, respectively (Scheme 3. For full characterization data cf. the SI). Ni11 and Ni12 were undoubtedly corroborated by NMR spectroscopy and elemental analysis. Based on single crystal structure data (cf. SI for details), the buried volume of Ni11 was determined by the aforementioned DFT methods. This yields a significantly higher value compared to the reference cases (V _bur% value 52 %). Most significantly, two dibenzosuberyl substituents very effectively shield the axial sites above and below the square planar Ni^II center (Figure 2 e). This effective shielding by the suberyl aryl groups is also reflected by an upfield ¹H NMR resonance of the Ni‐Me moiety (Ni11: −0.93 ppm) relative to the reference catalyst precursors (Ni2: −0.29 ppm). Most notably, in pressure reactor ethylene polymerization experiments with Ni11 and Ni12 yield much higher molecular weights vs. all reference systems (Table 1 and Figure 1).

Polymerization properties

Ethylene polymerization with Ni11 and Ni12 as single‐component catalyst precursors was studied in more detail (Table 1, entries 21–28 and Table 2). Notably, Ni11 and Ni12 are thermally robust to give a high molecular weight (M _w=16.9×10⁴ g mol⁻¹) even at a high polymerization temperature of 90 °C, more than an order of magnitude higher than those produced by Ni1–Ni10 (Figure 1), indicating the effectiveness of our C _s‐type shielding strategy.

Table 2

Effect of time and pressure on ethylene polymerization with Ni11.^[a]

Entry	T [°C]	P [bar]	T [min]	Yield [g]	act. [×106][b]	M n [×104][c]	M w [×104][c]	M w/M n [c]	brs[d]	T m [e] [°C]
1[f]	30	8	5	0.31	2.19	22.3	25.5	1.15	3.5	131/135
2[f]	30	8	10	0.60	2.12	39.5	45.4	1.15	2.9	133/136
3[f]	30	8	15	0.91	2.14	57.8	69.5	1.20	2.7	133/138
4[f]	30	8	20	1.09	1.92	70.9	86.5	1.22	2.9	132/ 138
5[g]	30	30	15	2.61	6.96	208.7	242.3	1.16	0.2[k]	136/ 144
6[h]	30	40	15	2.26	12.05	248.9	372.2	1.50	0.1[k]	134/142
7[h]	30	40	60	2.95	3.93	304.7	505.1	1.66	<0.1[k]	135/144
8[i]	90	30	15	2.94	3.92	21.1	31.8	1.51	29	99/106
9[i]	90	40	15	3.70	4.93	22.4	34.1	1.52	28	101/106
10[h,j]	30	40	60	2.31	3.08	141.3	202.0	1.43	0.3[k]	135/143

[a] Reaction conditions: toluene (100 mL), all entries are based on at least two runs, unless noted otherwise. [b] Activity is in unit of 10⁶ g mol⁻¹ h⁻¹. [c] Determined by GPC in 1,2,4‐trichlorobenzene at 150 °C using a light scattering detector. [d] brs=Number of branches per 1000C, as determined by ¹H NMR spectroscopy. [e] Determined by DSC at 10 K min⁻¹ (second heating/ first heating). [f] Ni11 (1.7 μmol). [g] Ni11 (1.5 μmol). [h] Ni11 (0.75 μmol). [i] Ni11 (3 μmol). [j] THF (100 mL). [k] brs=Number of branches per 1000C, as determined by ¹³C NMR spectroscopy.

Compared to these polymerizations in toluene, catalytic activity is even higher in tetrahydrofuran (THF) as a polar reaction medium, although molecular weight decreases slightly (Table 1, entries 22 vs. 29). Remarkably, polymerization even occurs in this challenging reaction medium—which is not compatible with early transition metal and also many late metal polymerization catalysts[ ³ , ⁴ , ⁵ , ⁶ , ²⁰ ]—to high molecular weight polymer (Table 2).

The polymerization to high molecular weight polyethylenes proceeds in a living fashion at 30 °C and 8 bar (Table 2, entries 1–4). Both yield and molecular weight increase linearly with polymerization time (Figure 3). Within 20 min M _w=86.5×10⁴ g mol⁻¹ is reached. Notably, polymer precipitation during the reaction possibly has a small effect on molecular weight distributions.

Figure 3

Living nature of ethylene polymerization with Ni11 at 30 °C (Table 2, entries 1–4).

Polymerization at higher 40 bar ethylene pressure results in remarkably higher catalyst activities (TOF: 4.3×10⁵ h⁻¹) and molecular weights well in the ultrahigh molecular weight regime (M _w: 3.7×10⁶ g mol⁻¹), while branching is further reduced to only 0.1/1000C (Table 2, entry 6). This compares favorably with the state‐of‐the‐art neutral Ni^II catalysts[ ¹⁶ , ²¹ , ²⁸ ] and other catalysts.[ ²⁹ , ³⁰ , ³¹ , ³² ]

Due to the triple characteristics of high activity, low degree of branching, and ultrahigh molecular weight, rapid precipitation of the product polymer occurred during polymerization, which may impact the reaction. Thus, an even lower catalyst loading of 7.5×10⁻⁷ mol was explored. Notably, despite the absence of any alkyl aluminum cocatalysts or scavenger, polymerization performance is retained (Table 2, entries 6 and 7). At a prolonged reaction time of 1 h UHWMPE with M _w of 5.1×10⁶ g mol⁻¹ (M _n: 3.1×10⁶ g mol⁻¹) is formed. Most importantly, the degree of branching is below the detection limit (<0.1/1000C) (Figure 4, Table 2, entry 7). This is also reflected in a very high melting point (T _m) of up to 144 °C. As anticipated, higher ethylene pressure of 30 bar and 40 bar at challenging 90 °C resulted in elevated activities and enhanced molecular weights (M _w: 3.4×10⁵ g mol⁻¹) (Table 2, entries 8 and 9). Most notably, in THF as a reaction medium Ni11 even enabled the formation of linear UHMWPE (M _w: 2.0×10⁶ g mol⁻¹; 0.3 branches/1000C) at 40 bar (Table 2, entry 10).

Figure 4

¹³C NMR spectra (100 MHz, C₂D₂Cl₄, 110 °C) of strictly linear UHMWPEs (conc.: 100 g L⁻¹): a) Table 2, entry 7; b) Table 2, entry 5.

Ni11 was also capable of promoting copolymerization of ethylene and ester‐functionalized undecylenic acid methyl ester (cf. SI, Table S2). Studies at various temperatures from 30 °C to 90 °C gave copolymers with high molecular weights (3.9 to 26.4×10⁴ g mol⁻¹), tunable degrees of branching (4 to 28/1000C), and reasonable incorporations of comonomer (0.2 to 1.5 mol %).

In summary, the approach reported enables an effective shielding of both apical positions of a square‐planar coordinated active metal site, through a C _s‐type arrangement of both shielding groups by attachment via the same N‐donor atoms’ framework. This overcomes the general propensity of late transition metal catalysts for β‐H elimination, and hinders chain transfer and branching pathways very effectively, as evidenced by a living nature of the polymerization and the formation of strictly linear ultrahigh molecular weight polyethylene. Such a microstructure is desirable for this important material, but difficult to access, particularly with late transition metal catalysts. Notably, these novel catalysts are also thermally robust and tolerant to polar reaction media. At the same time, this novel type of shielding is easily accessible. The key aniline building block is formed from commercially available compounds in a single step, with good yield. Thus, this chemistry provides broader perspectives for polymerization catalysis and beyond.

Conflict of interest

The authors declare no conflict of interest.