Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Linear Functionalized Polyethylene Prepared with Highly Active Neutral Ni(II) Complexes ERIC F. CONNOR,1 TODD R. YOUNKIN,1 JASON I. HENDERSON,1 SONJONG HWANG,1 ROBERT H. GRUBBS,1 WILLIAM P. ROBERTS,2 JOHNATHAN J. LITZAU2 1 Arnold and Mabel Beckman Laboratories of Chemical Synthesis, Division of Chemistry and Chemical Engineering, California Institute of Technology, Pasadena, California 91125 2 Cryovac, Incorporated, Sealed Air Corporation, 100 Rogers Bridge Road, Building A, Duncan, South Carolina 29334 Received 13 April 2002; accepted 22 May 2002 Published online 00 Month 2002 in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/pola.10370 Neutral Ni(II) salicylaldimine catalysts (pendant ligand ⫽ NCMe or PPh3) were used to copolymerize ethylene with monomers containing esters, alcohols, anhydrides, and amides and yielded linear functionalized polyethylene in a single step. ␣-Olefins and polycyclic olefin comonomers carrying functionality were directly incorporated into the polyethylene backbone by the catalysts without any cocatalyst, catalyst initiator, or other disturber compounds. The degree of comonomer incorporation was related to the monomer structure: tricyclononenes ⬎ norbornenes ⬎ ␣-olefins. A wide range of comonomer incorporation, up to 30 mol %, was achieved while a linear polyethylene structure was maintained under mild conditions (40 °C, 100 psi ethylene). Results from the characterization of the copolymers by solution and solid-state NMR techniques, thermal analysis, and molecular weight demonstrated that the materials contained a relatively pure microstructure for a functionalized polyethylene that was prepared in one step with no catalyst additive. © 2002 Wiley Periodicals, Inc. J Polym Sci ABSTRACT: Part A: Polym Chem 40: 2842–2854, 2002 Keywords: ␣-olefins; functionality; heteroatom-containing polymers; linear polyethylene; neutral nickel catalyst; norbornene; salicylaldamine; tricyclononene INTRODUCTION The incorporation of functional groups into the hydrocarbon backbones of polyolefins has been a highly desired modification for many years.1,2 It is expected to expand the scope of polyolefin applications by improving adhesion, polymer miscibility, gas diffusion characteristics, and toughness. In addition, the possibility of introducing reactive functional groups along a polyolefin backbone would allow for mild chemical modification and, Correspondence to: R. H. Grubbs (E-mail: rhg@its. caltech.edu) The supplementary material referred to in this article can be found at http://www.interscience.wiley.com/jpages/0887624X/Suppmat/2002/40/v40.2842.html. Journal of Polymer Science: Part A: Polymer Chemistry, Vol. 40, 2842–2854 (2002) © 2002 Wiley Periodicals, Inc. 2842 therefore, further alteration by the chemist.3 For polyethylene to function as a high-performance thermoplastic material, a high molecular weight and a crystalline melting point (Tm) are usually among the most desired properties.4 However, the incorporation of functionality directly into polyethylene often leads to a material that fails to meet one or both of these characteristics.5 Typical examples include commercial samples of poly(ethylene-co-alkylacrylate) that are prepared with extreme pressures (2000 atm) and high temperatures (⬎200°C) via the radical reaction of ethylene and acrylates.6 Materials formed in this fashion are characteristically branched copolymers that exhibit low Tm’s (⬃100 °C).7 The coordination polymerization of olefins offers an approach to more linear polymers (Scheme LINEAR FUNCTIONALIZED POLYETHYLENE Scheme 1 1). Polyolefin catalysts are typically electrophilic or oxophilic and, therefore, are poisoned by the presence of functionalized olefins.1 Heteroatoms have been incorporated into polyethylene using early metal and metallocene catalysts but often require the following reaction modifications: masking the functionality as an innocuous species (such as a borate), pre-complexation of functional groups by stoichiometric amounts of Lewis acidic species, and block copolymerization via two reactions.1,8 –15 More promise, however, has been observed with late metal catalysts because they are more tolerant of polar functionalized compounds.16 The neutral nickel catalysts based on a PO chelate used on an industrial scale in the Shell higher olefin process (SHOP) to generate ethylene oligomers have been reported to incorporate functional groups into polyethylene.16 Klabunde and Ittel17 showed that functionalized linear low-density polyethylene (LLDPE) could be prepared with POO chelate nickel complexes. For respectable catalytic turnover numbers (TONs), these catalysts often require high ethylene pressures and/or a cocatalyst to form higher molecular weight products. Longer chain ␣-olefins carrying functionality were reported to be incorporated into the polyethylene backbone.17,18 Recently Mecking demonstrated that neutral Ni complexes containing a SHOP NO type chelate operate in pure water to give high molecular weight polyethylene.19,20 Gibson recently reported the incorporation of acrylates into polyethylene with Ni catalysts with bulky PO ligands.21 Interestingly, acrylate incorporation was found at the chain ends of the polyethylene. Cationic nickel catalysts are believed to form intramolecular chelates with acrylate-type monomers, and so copolymerizations with these types of monomers have proven to be difficult.22 A significant breakthrough for these systems was re- 2843 ported recently by Johnson and McLain.23,24 They described the direct incorporation of alkyl acrylates into polyethylene with cationic nickel catalysts by the addition of an excess of borate salt that allowed the further insertion of ethylene. The addition of another catalyst component was required for the cationic nickel systems to operate in the presence of heteroatom functionality. However, the cationic palladium systems were singlecomponent for polyethylene-functionalized materials. The insertion of methacrylate has been shown to form four-, five, and six-membered chelates with the carbonyl coordinated to the palladium metal.25 The hydride elimination–reinsertion reaction cycle ring-expanded chelate structures and allowed the subsequent insertion of ethylene. Up to 20 mol % methacrylate could be incorporated into polyethylene. The polymers produced were highly branched (⬃100 branches per 1000 carbons), and the functionality was at the end of the branches. This provided novel materials with glass-transition temperatures (Tg’s) of ⫺67 to ⫺77 °C.26 –28 Functionalized norbornenes containing esters and free carboxylic acids have been homopolymerized via an insertion mechanism with cationic palladium catalysts to give high molecular weight polymers.29 –31 By reducing the steric bulk on the diimine ligands of the cationic late metal systems, Goodall et al. synthesized random copolymers of methacrylate and norbornene, obtaining wide control over the comonomer content.32 SHOP nickel catalysts were used by the same group to give ethylene-functionalized norbornene copolymers.33 Generally, the norbornene content was high in Goodall et al.’s materials (⬎40%), and this allowed the preparation of amorphous polymers with Tg’s ranging from 30 to 300 °C and higher, which depended on the bicyclic comonomer content. We have reported the neutral nickel catalysts 1 [pendant ligand (L) ⫽ NCMe] and 2 (L ⫽ PPh3), which provide high molecular weight, linear polyethylene in the presence of polar solvents such as ethers, esters, alcohols, amides, amines, and even water (Scheme 2).34,35 These catalysts contain specifically tailored bulky salicylaldamine ligands, which greatly increase both the catalyst activity and lifetime. The complexes operate at low pressures of ethylene (⬃1–7 atm), allowing reactions of less reactive olefins with respect to ethylene. Catalysts 1 and 2 operate by themselves, with no cocatalyst additive, and allow the incorporation of sensitive functionalities directly into polyethylene. 2844 CONNOR ET AL. Scheme 2 The resting state of the catalysts is believed to be the ligated species, in which NCCH3 or PPh3 occupies the open coordination site. Although dissociation of the pendant ligand leads to the formation of the same active species for complexes 1 and 2, 1H NMR experiments have shown that a higher percentage of initiation over the same timeframe and the same conditions occurs with the nitrile complex 1. Therefore, polymerizations conducted at lower ethylene pressures are more productive with catalyst 1. Accordingly, functionalities that do not strongly coordinate to the metal center were found to be successful with this system. We report the preparation of a number of functionalized polyolefins, ranging from semicrystalline materials (Tm ⬃ 80 –128 °C; 0.1– 6 mol % comonomer) to amorphous materials (Tg ⬃ 72 °C; 30 mol % comonomer). We describe the limitations and advantages of polymerization reactions that have been obtained with the single-component neutral nickel catalysts 1 and 2. EXPERIMENTAL available hex-1-ene epoxide and then hydrolyzed via stirring in dilute H2SO4 overnight. Extraction with diethyl ether was followed by stirring in the presence of Na2SO4. The diethyl ether was removed with reduced pressure, and this yielded the diol. Subsequent protection with 2-methoxypropene in the presence of amberlyst resin gave the dioxolane 5. The characterization matched published data.37,38 Preparation of Tricyclo[4.2.1.00,0]non-7-ene-3carboxylic Acid t-Butyl Ester (10) Quadricyclane (12 mL) and t-butyl acrylate (92 mL) were placed in a thick-walled Schlenk tube. The components were degassed and then sealed under an atmosphere of nitrogen. The reaction mixture was heated to 96 °C for 8 h. The product was obtained by distillation. The isolated yield with respect to quadricyclane was 80% (bp ⫽ 120 °C at 1.00 mmHg). NMR spectroscopy revealed that endo and exo isomers were present (3/1 ratio). The assignments of the endo isomer follow. 1 H NMR (CDCl3, 300 MHz, 25 °C): 5.90, 5.91 (2H, bs), 2.68 (1H, bs), 2.60 (1H, bs), 2.28 –2.27 (2H, m), 2.20 (1H, m), 1.98 (1H, d, JHH ⫽ 9.3 Hz), 1.61 (1H, d, JHH ⫽ 10.0 Hz), 1.41 (10H, s), 1.26 (1H, d, JHH ⫽ 10.0 Hz). 13C NMR (CDCl3, 75.47 MHz, 25 °C): 175.32, 135.78, 134.57, 79.76, 44.04, 43.83, 40.41, 40.27, 38.37, 34.11, 28.03, 23.66. IR (KBr): 1718.3 cm⫺1 {[C(O)]}. ELEM. ANAL. Calcd. for C14H20O2: C, 76.33%; H, 9.15%. Found: C, 76.41%; H, 9.12%. Materials All manipulations and polymerizations were carried out in an N2-filled drybox or with standard air-sensitive or vacuum-line techniques under argon. Argon was purified by passage through columns of BASF RS-11 (Chemalog) and Linde 4-Å molecular sieves. Ethylene was purified by passage through an O2 scrubber (Matheson model 6410 oxygen-absorbing purifier), and water was removed by molecular sieves. Solvents were rigorously degassed in 18-L reservoirs and passed through two sequential purification columns. Protic contaminants were removed with activated alumina, and a supported copper catalyst was used to remove trace oxygen from hydrocarbons.36 Compound 2 was synthesized according to a published procedure.34 The characterization and synthesis of 1 will be reported in a forthcoming article on the catalysts. Compound 5 was prepared from commercially Synthesis of Tricyclo[4.2.1.00,0]non-7-ene-3,4dicarboxylic Acid t-Butyl Ester (11) A previously reported adduct of quadricyclane and maleic anhydride39 was hydrolyzed by 16.2 g of the white solid being placed in a flask with 100 mL of H2O. NaOH (0.4 g) was added to the solution, which was then set to reflux for 8 h. The water was removed to give a white solid that was dried under reduced pressure. t-BuOH (50 mL), the white solid, and 1.5 mL of concentrated H2SO4 were placed into a glass bomb with a pressure gauge. Isobutylene gas (70 mL) was condensed in the glass bomb (⫺70 °C). The flask was allowed to warm to room temperature. After the reaction was stirred for 8 h, the flask was heated to 84 °C (70 psig in the flask) for 3 h. The flask was cooled to ⫺70 °C, and the isobutylene gas was allowed to evaporate. The oil that remained was dissolved in pentane and washed with a saturated LINEAR FUNCTIONALIZED POLYETHYLENE water solution of K2CO3. After the pentane layer was dried with MgSO4, 22.0 g of the crude product was obtained, and it crystallized upon standing. The product was purified by column chromatography with a 1/9 mixture of ethyl acetate and hexanes. NMR spectroscopy revealed that exo and endo isomers were present (3/1 ratio). Spectral Data for the Endo Isomer 1 H NMR (CDCl3, 300 MHz, 25 °C): 5.97 (2H, bs), 2.74 (2H, bs), 2.58 (2H, m, JHH ⫽ 3.5 Hz), 2.23 (2H, m, JHH ⫽ 2.1 Hz), 1.5–1.3 (20H, maxima at 1.48, 1.43, and 1.36). 13C NMR (CDCl3, 75.47 MHz, 25 °C): 172.32, 135.33, 80.42, 43.45, 40.89, 40.81, 37.90, 28.07. Spectral data for the Exo Isomer 1 H NMR (CDCl3, 300 MHz, 25 °C): 5.98 (2H, bs), 3.34 (2H, m, JHH ⫽ 9.6 Hz), 3.14 (2H, bs), 2.12 (2H, m, JHH ⫽ 9.3 Hz), 2.02 (1H), 1.43 (18H, bs), 1.26 (1H). 13C NMR (CDCl3, 75.47 MHz, 25 °C): 169.98, 136.49, 80.00, 42.98, 41.83, 37.88 (two overlapping peaks), 28.07. IR (KBr): 1721.5 cm⫺1 {[C(O)]}. ELEM. ANAL. Calcd. for C14H20O2: C, 71.22%; H, 8.81%. Found: C, 71.22%; H, 8.76%. Synthesis of N-Butyl-tricyclo[4.2.1.00,0]non-7-ene3,4-dicarboxyimide (13) A previously reported adduct of quadricyclane and maleic anhydride39 was placed in a flask with toluene and 1.1 equiv of n-butyl amine. The mixture was refluxed for 12 h with a Dean–Stark apparatus. The product was isolated by distillation to give a light yellow solid. The isolated yield with respect to the anhydride starting material was 85% (bp ⫽ 142 °C at 0.25 mmHg). 1 H NMR (CDCl3, 300 MHz, 25 °C): 6.01 (2H, bs), 3.51 (2H, t, J ⫽ 7.2 Hz), 2.94 (2H, bs), 2.65 (2H, bs), 2.10 (2H, bs), 1.6 –1.4 (6H, m), 1.3–1.2 (6H, m), 0.89 (3H, t, 7.5 Hz). 13C NMR (CDCl3, 75.47 MHz, 25 °C): 179.10, 135.81, 44.54, 41.71, 41.05, 40.95, 38.92, 30.10, 20.37, 14.05. IR (KBr): 1769.3, 1697.2 cm⫺1 {[C(O)]}. ELEM. ANAL. Calcd. for C15H19NO2: C, 73.44%; H, 7.81%; N, 5.71%. Found: C, 73.34%; H, 7.94%; N, 5.76%. General Procedure for the Polymerization of Ethylene with Comonomers via Ni Complexes A 6- or 12-oz Fisher–Porter glass pressure bottle was charged with the appropriate amount of the Ni complex under an atmosphere of nitrogen (for 2845 the polymerization at 14 atm, a 200-mL autoclave was used). Toluene (80 mL) was then cannulatransferred into the reactor and was followed by a solution of comonomer in toluene (10 mL). The ethylene pressure was raised to a particular value and maintained for the specified times. MeOH/ HCl workup afforded polyethylene, which was filtered and dried in vacuo. General Procedure for the High-Resolution Liquid C NMR Characterization of Polyethylene Copolymers 13 It was shown by Randall40 that the nuclear Overhauser effect (NOE) factor for many polyethylene systems is nearly full and constant for all carbons of the copolymer. Therefore, WALTZ (wideband alternating phase low power technique for zero residual splitting) decoupling sequences were used for an improved signal-to-noise ratio, which allowed the number of scans to be reduced. This is important when the polymer of interest is sparingly soluble in the solvent blend. When the integrations did not agree with each other via WALTZ, an inverse-gated experiment was run to check if some of the signals were being selectively enhanced. Little difference in the integration was observed in most polyethylenes; with calculations based on the averaging of several signals, any error was minimized. Other things that were considered were the potentially long relaxation times (T1’s) of end groups (e.g., methyls and methylenes at or near the end of the polymer chain); these signals were not used for quantitation because they would not have fully relaxed when the next pulse fired and, as a result, signal was lost in these carbons. In a typical copolymer, the branch, ␣, and  carbons were used to integrate. The sum of all of these signals was made and then divided by the number of carbons that they were attributed to get an averaged integration unit per carbon for that constituent. A 90° pulse (based on the main methylene resonance at 30.0 ppm) and at least 32,000 data points were collected on a Bruker DMX-400 MHz NMR instrument with high-temperature capabilities. Three thousand scans, with a 10-s recycle delay, which amounted to an 8-h run, constituted a typical experiment. This was done to satisfy the requirements for the signal-to-noise ratio (ASTM Method D 5017-96). The NMR solvent mixture [⬃150-mg sample, added in small aliquots, with 2.0 mL of 1,2,4trichlorobenzene (protonated), 0.75 mL of benzene-d6 (lock), and 0.25 mL of hexamethyldisiloxane (internal reference); 10-mm tubes] worked 2846 CONNOR ET AL. Scheme 3 well for most polyolefins. The dissolution heating block and NMR experiment were maintained at 130 °C because good temperature control was important (benzene-d6 begins to boil at ⬃130 °C, and going above this could prove disastrous). Small slits in the top of the NMR tube caps were cut to avoid pressure buildup in the tube. to be approximately 80 and 0.1 s at 4.7 T for the crystalline and amorphous phases, respectively. There was no noticeable NMR signal growth observed when the repetition time was increased to 1000 s. RESULTS AND DISCUSSION General Procedure for Gel Permeation Chromatography (GPC) Analysis GPC analysis was determined with a Waters 150 GPCV liquid chromatograph. It was equipped with three detectors: a Waters differential refractometer, a Waters single-capillary viscometer, and a Wyatt DAWN DSP light scattering detector. Several samples were prepared and yielded approximately 0.1 wt % solutions in 1,2,4-trichlorobenzene. Four Waters high-temperature Styragel columns (106, 105, 104, and 103 Å) at 140 °C were used to determine the molecular weight distributions (MWDs) of the polymers. The instrument was calibrated with TSK narrowpolydispersity polystyrene standards. A universal calibration curve was created to determine the MWDs. The data were collected and calculated with Water’s Millennium software. General Procedure for Solid-State NMR The solid-state 13C NMR experimental procedure was as follows. All magic-angle spinning (MAS) 13 C NMR experiments were carried out with a Bruker DSX 200 (50 MHz for 13C) and a 7-mm cross-polarity/magic-angle spinning (CP-MAS) probe, except for a sample of 10 mol % comonomer incorporation for which a Bruker DSX 500 (125 MHz for 13C) and a 4-mm BL CP-MAS probe were used because of the limited amount of the sample. Polymer powder (160 mg) was packed in a 7-mm MAS rotor and was spun at 4.0 kHz and 4.7 T. The 13C block decay signal was acquired with a 4-s, 90° pulse with high-power H-decoupling, and the typical repetition time used was 500 s. The spin–lattice relaxation times were estimated Copolymerization of Ethylene and ␣-Olefins Initially, we tested the commercially available ethyl undecylenoate 3 to determine whether catalysts 1 and 2 could incorporate functionalized ␣-olefins and produce high molecular weight ethylene copolymers (Scheme 3). Trials 1a– d (Table 1) outline our discoveries with catalysts 1 and 2. In analogy to our observations for ethylene homopolymerization, the acetonitrile catalyst 1 demonstrated a slightly higher activity for ethylene/ethyl undecylenoate copolymerizations than the phosphine-containing catalyst 2 (cf. trials 1a and b to trials 1c and d). The observed turnover numbers (TONs) for both catalysts are ⬃2 ⫻ 105 grams of polymer ⫻ mol of catlayst⫺1 ⫻ h⫺1, which is approximately 1 order of magnitude slower than for ethylene homopolymerizations performed in a noncoordinating solvent (e.g., toluene or benzene, ⬃6 ⫻ 106) but compares well with the rates observed for ethylene homopolymerization in a more coordinating medium (e.g., ether, ⬃3.1 ⫻ 106, or ethyl acetate, ⬃6.8 ⫻ 105). This rate suppression could be due to the coordination of the functional moiety or merely a reflection of a decreased insertion rate for ␣-olefins. Klabunde and Ittel reported a similar effect with their POO chelate nickel complexes.41 However, a threefold increase in the concentration of 3 led to only a minor decrease in the activity of both catalysts 1 and 2 (cf. trials 1a and b and 1c– d). Higher incorporation of 3 was obtained from both catalysts in more concentrated solutions. When a large excess of ethylene was employed, the conversion of comonomer 3 into the polymer was low (Table 2, trial 2c). However, high molec- 2847 LINEAR FUNCTIONALIZED POLYETHYLENE Table 1. Comparison of Catalysts 1 and 2 for the Copolymerization of Ethylene and Ethyl Undecyleneoate 3a Trial Catalyst Comonomer 3 (M) TON (⫻105)b Branch Contentc Tm (°C) Comonomer Incorporatedd 1a 1b 1c 1d 1 1 2 2 0.62 2.07 0.62 2.07 1.88 1.80 1.76 1.32 5 6 9 10 118 108 125 110 1.0 3.8 0.2 1.8 a Conditions: 25 mol of catalyst (2.5 mM); volume of toluene ⫹ comonomer 3 ⫽ 10 mL; 120 psig ethylene; reaction time ⫽ 15 min 40 °C bath. b TON ⫽ (g of polymer) ⫻ (mol of catalyst)⫺1 ⫻ (h)⫺1. c Total Me ⫹ Et ⫹ Pr ⫹ Bu branches per 1000 carbons as determined by 13C NMR. d Incorporation ⫽ molar percentage. cause of protonation of the polymer chain on the ligand from the catalyst. Protecting the diol functionality as the dioxolane monomer 5 led to higher ethylene consumption by catalyst 1, and this was reflected in both the molecular weight and TON (cf. trials 2d and e). The copolymerization of 1-octene 6 under similar conditions demonstrated a few noticeable differences (Table 3). First, it indicated that the presence of functionality on the monomer affected comonomer incorporation into polymer. Higher contents of ␣-olefins (12.5 mol %, trial 3f) could be obtained with 1-octene at lower concentrations than for the functionalized ␣-olefins previously discussed (Tables 1 and 2). As the molarity of the ␣-olefin in the polymerization mixture increased, a corresponding decrease in the polymer molecular weight resulted (trials 3a, b, d, and e). This contrasts with the functionalized ␣-olefins, for which the molecular weight was less affected by the comonomer concentration (trials 2a– c). ular weight, linear copolymers were accessible, and these materials fell into the category of highdensity polyethylene.4 When the polymerization was run in almost neat functionalized ␣-olefins, comonomer incorporation into polyethylene could reach levels of nearly 10 mol %, and so the material moved toward an LLDPE.4 With high levels of comonomer incorporation, the resulting polymer exhibited a decrease in Tm as predicted (in trial 1b, 3.8 mol % incorporation reduced Tm to 108 °C).42 13C NMR spectra of poly(ethylene-coethyl undecylenoate)s of different compositions are displayed in Figure 1. One can observe more functional groups than branches (cf. the intensity of the C1 peak with that of the methyl peak in Fig. 1). When one considers the branch content in functionalized polyethylenes that are produced as a result of side reactions in a free radical polymerization process, we find that both spectra illustrate that these catalysts produced a material with a relatively pure microstructure.5,42 ␣-Olefins with free alcohols (5) could be incorporated (trial 2e). Again, the catalyst lifetime was affected, as for the homopolymerization of ethylene in the presence of methanol,35 possibly be- Branch Analysis When the 13C NMR spectra of copolymers from 1-octene and ethyl undecylenoate were examined Table 2. Copolymerization of Ethylene and Functionalized-␣-Olefin Comonomers with Catalyst 1a Trial 2a 2b 2c 2d 2e 3 3 3 4 5 Comonomer (M) Mw (⫻103) PDI (Mw/Mn) Branch Contentb Tm (°C) Ethylene TONc Comonomer TONc Comonomer Incorporatedd 3.61 2.16 0.72 0.72 0.72 176 204 165 172 54 2.4 2.1 2.5 2.3 2.4 8 7 8 5 8 125 128 124 124 125 5,396 6,011 12,400 6,080 1,362 17 10 25 19 5 0.4 0.2 0.2 0.2 0.3 Conditions: 65 mol of catalyst (0.72 mM); 90 mL of toluene; 100 psig ethylene; 40 °C bath; reaction time ⫽ 8 h. Total Me ⫹ Et ⫹ Pr ⫹ Bu branches per 1000 carbons as determined by 13C NMR. TON ⫽ mol of substrate converted/mol of catalyst. d Incorporation ⫽ molar percentage. a b c 2848 CONNOR ET AL. Figure 1. 13C NMR spectra (75% C6H3Cl3, 25% C6D6, 400 MHz, 130 °C) of a copolymer of ethylene and monomer 3 produced by catalyst 1. The polymer contains (a) 10 mol % 3 or (b) 0.4 mol % 3. at similar incorporation levels, differences in the comonomer insertion mechanism became apparent. Therefore, when we compared the branch content of polymers produced in trials 1b and 3a, it was apparent that the presence of functionality in the reaction solution reduced the branch content from 58 branches per 1000 carbons (trial 3a) to 6 branches per 1000 carbons. An examination of Table 2 demonstrates that the presence of heteroatoms, whether in the solvent or in the monomer, resulted in polymers with fewer branches in comparison with the comonomers produced with octene (Table 3). With neutral nickel salicylaldamine catalysts, copolymerizations with vinyl-functionalized or al- lyl-functionalized monomers such as vinyl acetate or methyl acrylate were unsuccessful under the conditions used in this study. With methyl acrylate, no polymer was obtained. A solution of methyl acrylate with catalyst 2, examined by NMR, showed no observable interaction of the monomer and the catalyst. Polymerizations with catalysts 1 and 2 in the presence of methyl acrylate gave no isolable material; this suggested that intramolecular chelation occurred during polymerization and prevented polymer chain growth. With vinyl monomers such as vinyl ether, polyethylene was produced, but the incorporation of the comonomer into the polyethylene backbone was not observed. LINEAR FUNCTIONALIZED POLYETHYLENE 2849 Table 3. Copolymerization of Ethylene and 1-Octene 6 with Catalysts 1 and 2a 1-Octene Trial Catalyst mL mM Activityb Mw (⫻103) PDI (Mw/Mn) Branch Contentc Comonomer Incorporatedd Tm (°C) 3a 3b 3c 3d 3e 3f 1 1 2 1 1 1 0.1 0.5 1.0 1.0 2.5 10.0 2.2 10.5 22.5 22.5 55.0 235.0 5.27 2.75 3.27 2.48 1.22 0.95 12.8 8.8 —e 9.3 7.4 3.3 3.30 2.50 —e 2.10 3.40 2.10 58 67 76 66 82 106 1.7 2.4 1.7 4.0 6.5 12.5 111.5 104.8 120.0 89.4 Broad Broad a Standard conditions (unless stated otherwise): 28 mol of catalyst (2.8 mM); volume of toluene ⫹ comonomer 6 ⫽ 10 mL; 120 psig ethylene; 40 °C bath; reaction time ⫽ 1 h. b Activity ⫽ 105 g of polymer ⫻ (mol of catalyst)⫺1 ⫻ (h)⫺1. c Total Me ⫹ Et ⫹ Pr ⫹ Bu branches per 1000 carbons as determined by 13C NMR. d Incorporation ⫽ molar percentage. e Not determined. Copolymerization of Functionalized Norbornenes Although ␣-olefins were successfully copolymerized with ethylene, the results suggest that we could employ a slightly more reactive olefin for the reaction process (Scheme 4). Norbornene monomers contain 19.2 kcal/mol of ring strain,43 and the relief of strain energy via insertion into the saturated bicyclic structure (14.4 kcal/mol) has been considered to be an important factor for their increased activity in comparison with that of an ␣-olefin. In addition, it has been observed that the -hydrogen elimination reaction occurs rarely with norbornene monomers, presumably because of the unfavorable geometry of the  hydrogen on the bicyclic monomer.44 There are reports that under standard conditions, diimine Pd and Ni polyethylene catalysts give no isolable ethylene– norbornene copolymers, and so a study of these types of monomers was conducted and reported.45 Reaction data for these monomer types with catalysts 1 and 2 (Table 4) show that they were readily incorporated into the polyethylene backbone with an improved molar percentage with respect to that of ␣-olefins. Although the norbornene monomers were incorporated in a high molar percentage, the overall TON was noted to be lower than with the functionalized ␣-olefins. It is likely that this reflects the reduced rate of Scheme 4 insertion of ethylene into the metal norbornene bond due to steric factors. The polycyclic ethylene copolymers gave lower molecular weight materials, but the comonomer yield into the polymer was higher. Functionalized tricyclononene monomers (10– 13) contain an exocyclobutane (Scheme 5 and 6). It was predicted that these monomers would be incorporated more readily than their norbornene analogues as a result of the functionality being held further away from the inserting olefin (cf. trials 5c and g). These monomers were also easy to prepare in one step according to the method described by Tabushi.39 The copolymerization data obtained for ethylene and 10 with the neutral nickel catalyst are displayed in Table 4. The molecular weight decreased moderately when the ethylene pressure was decreased from 7 to 1 bar (cf. trials 5d– g). As observed, very high levels of comonomer incorporation (30 mol %) could be obtained with these monomers by a lower pressure of ethylene being applied (⬃5 psig) to initiate the complex in the presence of comonomer (trial 5d). Because of the reactivity difference between the smaller ethylene monomer and the less reactive cyclic olefin, it was thought that with the concentration of ethylene decreasing in the polymerization solution, the reaction of the less reactive monomer would occur more frequently. A decrease in the TON was observed for both ethylene and the polycyclic comonomer at lower ethylene pressures. It is believed that this resulted because of a combination of incomplete catalyst initiation and catalyst decomposition. 2850 CONNOR ET AL. Table 4. Ethylene Polymerization with Catalysts 1 and 2 in the Presence of Monomers 7–13a Trial R Comonomer/ Ni Ethylene (psig) Ethylene TONb Comonomer TONb Branch Contentc Mw (⫻103)d PDI (Mw/Mn) Tm (°C) Comonomer Incorporatede 5a 5b 5c 5d 5e 5f 5g 5h 5ih 5j 7 8 9 10 10 10 10 11 12 13 0.16 0.14 0.36 0.14 0.15 0.15 0.36 0.18 0.02 0.13 100 100 100 ⬍5 10 30 100 100 20 100 458 1540 2474 102 2014 2600 8585 1360 3240 7600 34 39 60 51 129 123 200 5 20 8 9 9 9 —f 11 14 14 5 43 7 17 70 74 30 55 41 76 59 —i —i 1.6 2.1 1.6 1.2 1.6 1.6 2.3 2.9 —i —i 105 110 109 —g 80 90 112 124 112 124 5 4 2 31 6 4 2 0.3 0.5 0.1 a Conditions: 65 mol (0.72 mM) of catalyst; 90 mL of toluene; 40 °C bath; reaction time 8 h. Trials 5a– b were conducted with catalyst 2, and trials 5c–j were obtained with catalyst 1. b TON ⫽ mol of substrate converted/mol of catalyst. c Total Me ⫹ Et ⫹ Pr ⫹ Bu branches per 1000 carbons. d In g ⫻ mol⫺1. e Incorporation ⫽ molar percentage, as determined by 13C NMR. f No branches observed. g No Tm observed. h 2,6-Di-tert-butylpyridine was added to the monomer solution before the reaction. i Data not recorded. It was observed that the polymers with increasing comonomer incorporation reflected a corresponding decrease in Tm. With 6 mol % of the comonomer 10 in polyethylene, Tm, as determined by differential scanning calorimetry (DSC), appeared broad, and as the comonomer content in the polymer increased, the melting endotherm disappeared (Fig. 2). With 30 mol % comonomer incorporation (trial 5d), DSC thermal analysis showed an endothermic thermal transition between 68 and 75 °C (maximum rate, 71 °C) that was assigned as Tg; no melting endotherm was observed. These results agree with data from ethylene–norbornene copolymers made with early metal metallocenes.46 The 13C NMR spectroscopic analysis of these ethylene copolymers indicated that the branch content within these copolymers was generally less than 15 branches per 1000 carbons, with methyl branches being the most prevalent (trials 5e– g, which contained 11, 14, and 14 methyl branches per 1000 carbons, respectively). The 13C NMR spectra for the polycyclic Scheme 5 copolymers described in this article indicated that there were no comonomer dyads present, even when the comonomer incorporation was high (see Fig. 3). Resonances with a shift downfield from ␦ ⫽ 46.0 ppm, typical of norbornene dyads in polyethylene, were not observed in our spectra.47 No homopolymer was obtained from any norbornene monomer with the single-component Ni catalysts. Therefore, the theoretical limit of polycyclic olefin incorporation for catalysts 1 and 2 is likely to be 50 mol %. It is probable that norbornene insertion occurs preferentially at alternating ethylene units because of steric encounters that occur with the norbornene monomer and the ligand on catalyst 1. This characteristic is frequently exhibited with metallocene systems.48 For this reason, Goodall et al. modified the cationic nickel and palladium catalysts containing diimine ligands so that ethylene–norbornene copolymers could be prepared.32 Although we have not yet found the appropriate conditions or monomer, it should be possible to form a strictly alternating ethylene– tricyclononene or ethylene–norbornene copolymer with these catalysts. Moreover, monomer 10 contained the acid-sensitive tert-butyl ester group, and it can be seen from the 13C NMR spectrum (Fig. 2) that this functionality remained fully intact. We think that this is illustrative of how mild these nickel catalysts are and that this demonstrates the advantage of using a neutral- LINEAR FUNCTIONALIZED POLYETHYLENE 2851 Scheme 6 based catalyst rather than a cationic complex more prone to generating highly acidic byproducts during the polymerization. Copolymerizations of monomer 12 with ethylene demonstrated the incorporation of anhydride functionality into an ethylene copolymer (trial 5i). The catalyst TON was increased by treatment of the anhydride monomer solution with a proton sponge for the removal of trace amounts of acid because carboxylic acids have been shown to protonate the ligand from the metal catalyst. For further improvements of the TONs, the di-tertbutyl ester analogue 11 was prepared and incorporated into polyethylene (trial 5h). The complete conversion of the di-tert-butyl ester to the anhydride was accomplished by the heating of the functionalized polyethylene to 200 °C under re- duced pressure (Scheme 7). This conversion could be effected in a polymer melt during processing. The structures of the anhydride– ethylene copolymers prepared from both 11 and 12 were found, with 13C NMR spectroscopic analysis, to be similar. Tricyclononene n-butyl imide 13, prepared from the anhydride 11, was also incorporated into ethylene polymers. Polymerizations carried out with 13 proceeded in a noticeably controlled fashion. This observation is reflected in the high ethylene TONs observed (trial 5j) and demonstrates that certain functionalities actually stabilized the nickel complexes toward side reactions over the course of the polymerization. The concentration of the methanol fraction used to precipitate the polymer and analysis by 1 H NMR spectroscopy revealed the following Figure 2. DSC heating curves for a copolymer of ethylene and monomer 10 (in air at 5 °C/min after cooling from 140 °C at 10 °C/min) showing both the melt transition and the transition for thermal cleavage of the t-butyl ester: (a) polyethylene with 6 mol % 10 incorporated, (b) polyethylene with 4 mol % 10 incorporated, and (c) polyethylene with 2 mol % 10 incorporated. 2852 CONNOR ET AL. Figure 3. 13C NMR spectrum (75% C6D3Cl3, 25% C6D6, 400 MHz, 130 °C) of a copolymer of ethylene and monomer 10 produced by catalyst 1. The polymer contains 31 mol % 10. when strained cyclic monomers were used. No low molecular weight polyethylene fractions were ever observed; only unreacted monomer and ligand were identified. An analysis of the unreacted methyl ester norbornene monomer 9 revealed that there did not appear to be a significant change in the initial–final isomer distribution. This indicates that norbornene monomers did not appear to exhibit an obvious reactivity difference between the endo and exo isomers for the neutral nickel catalysts 1 and 2.49 This was confirmed from an NMR analysis of the polymers that contained an endo/exo ratio of 60/40, which was similar to the ratio found in the monomer feed. Similarly, 10 and 11 were found to contain anti/syn isomer ratios of 40/60 and 45/55, respectively. The tricyclononenes differed from the norbornene monomers in that a reactivity difference between stereoisomers was observed. The monomer isomer ratio found in the methanol from polymer precipitation indicated that the anti isomer was consumed faster than the syn isomer. The preference for one isomer is also reflected by the isomer Scheme 7 ratio (determined by 13C NMR spectroscopic analysis) present in the isolated polymers. A study was carried out with solid-state 13C NMR techniques to characterize the crystallinity within linear polyethylenes containing comonomer 3 (further details are given in the supplementary material). The polyethylene samples examined ranged from 3.8 to 0.1 mol % in comonomer content. Spectra from these samples exhibited solid-state 13C NMR signals between 26 and 35 ppm, and these were fitted with three Lorentzian lines, as shown in the Figure 4. The peak at 31.3 ppm with a full width at half-maximum of 126 Hz represents the amorphous-phase methylene carbons, whereas the two relatively narrow peaks at 32.8 and 34.1 ppm with a full width at half-maximum of 30 – 45 Hz were assigned as methylene carbons in the crystalline phase.50 –52 The two crystalline peaks indicate the presence of two crystallographically distinct phases, which are called the orthorhombic phase (32.8 ppm) and the monoclinic phase (34.1 ppm) in previous reports.50,51 The quantification of the three different phases was carried out from the spectral peak area of fitted lines and is summarized in Table 5. The content of the crystalline phase for the samples examined was over 60%, and over 20% of this phase was found to be the monoclinic crystalline phase. Within the comonomer incorporation range examined (which is expressed as a molar percentage), the crystallinity increased as the incorporation decreased, as expected. No strong cor- LINEAR FUNCTIONALIZED POLYETHYLENE 2853 Figure 4. Solid-state 13C NMR spectra of functionalized polyethylene containing monomer 3. (a) Experimentally obtained spectra are shown on the top, and the bottom curve shows the Lorentzian fit with the resulting simulated curve in the middle. (b) Solid-state spectra are shown that were obtained experimentally for polyethylene with different monomer loadings. relation between the incorporation and the fraction of the monoclinic phase was observed. CONCLUSIONS Neutral, late transition-metal catalysts similar to the SHOP catalyst were shown to be useful in preparing a number of functional-group-containing, high molecular weight, lightly branched, semicrystalline ethylene copolymers under mild laboratory conditions. Significantly, these were obtained directly from monomers containing heteroatoms with a single catalyst system. The potential that catalysts of the neutral nickel type may have for making a variety of useful functionalized polyethylene materials was demonstrated. High ethylene TONs were observed for catalysts 1 and 2, demonstrating a tolerance of these catalysts for both olefin functionality and a variety of different oxygen and nitrogen functionalities. The tricyclononene and norbornene monomers exhibited higher incorporation into the polyethylene Table 5. Relative Ratio of the Crystalline Phases Trial Incorporation (mol %) Crystalline Part (%) Fraction of Monoclinic Phase (%) 2b 2c 2a 1b 0.2 0.2 0.4 3.8 63.3 65.5 60.5 66.7 26.7 30.9 28.4 10.6 backbone in comparison with ␣-olefins such as ethyl undecylenoate. Functional groups that were successfully incorporated directly into ethylene copolymers in this manner included ester, imide, and anhydride groups. The indirect incorporation of anhydride groups was also demonstrated with the thermolysis of vicinal di-tert-butyl ester groups. Support was provided by the National Institute for Standards and Technology (through the Advanced Technology Program) and Cryovac, a division of Sealed Air Corp. GPC and polymer NMR analyses were carried out by Cryovac. REFERENCES AND NOTES 1. Boffa, L. S.; Novak, B. M. Chem Rev 2000, 100, 1479 –1493. 2. Britovsek, G. J. P.; Gibson, V. C.; Wass, D. F. Angew Chem Int Ed 1999, 38, 429 – 447. 3. Novak, B. M.; Tanaka, H. Abstr Pap Am Chem Soc 1999, 217, U414 –U414. 4. Fatou, J. G.; Macia, I. G.; Marco, C.; Gomez, M. A.; Arribas, J. M.; Fontecha, A.; Aroca, M.; Martinez, M. C. J Mater Sci 1996, 31, 3095–3107. 5. Watson, M. D.; Wagener, K. B. Macromolecules 2000, 33, 5411–5417. 6. Deckers, A.; Schulund, R.; Klimesch, R. U.S. Patent 5,473,025, 1995. 7. Marr, D. W. M. Macromolecules 1995, 28, 8470 – 8476. 8. Stehling, U. M.; Malmstrom, E. E.; Waymouth, R. M.; Hawker, C. J. Macromolecules 1998, 31, 4396. 2854 CONNOR ET AL. 9. Lu, B.; Chung, T. C. Macromolecules 1998, 31, 5943–5946. 10. Lu, B.; Chung, T. C. Macromolecules 1999, 32, 2525–2533. 11. Lu, B.; Chung, T. C. J Polym Sci Part A: Polym Chem 2000, 38, 1337–1343. 12. Zheng, L.; Farris, R. J.; Coughlin, E. B. J Polym Sci Part A: Polym Chem 2001, 39, 2920 –2928. 13. Wang, Y. B.; Hillmyer, M. A. J Polym Sci Part A: Polym Chem 2001, 39, 2755–2766. 14. Desurmont, G.; Tanaka, M.; Li, Y.; Yasuda, H.; Tokimitsu, T.; Tone, S.; Yanagase, A. J Polym Sci Part A: Polym Chem 2000, 38, 4095– 4109. 15. Matyjaszewski, K.; Teodorescu, M.; Miller, P. J.; Peterson, M. L. J Polym Sci Part A: Polym Chem 2000, 38, 2440 –2448. 16. Ittel, S. D.; Johnson, L. K.; Brookhart, M. Chem Rev 2000, 100, 1169 –1203. 17. Klabunde, U.; Ittel, S. D. J Mol Catal 1987, 41, 123–134. 18. Klabunde, U.; Roe, D. C.; Ittel, S. D. Abstr Pap Am Chem Soc 1987, 193, 285-INOR. 19. Mecking, S.; Held, A.; Bauers, F. M. Angewandte Chemie-International Edition 2002, 41 , 545–561. 20. Bauers, F. M.; Mecking, S. Angewandte ChemieInternational Edition 2001 , 40, 3020 –3022. 21. Gibson, V. C.; Tomov, A. Chemical Commun 2001, 1964 –1965. 22. Correia, S. G.; Marques, M. M.; Ascenso, J. R.; Ribeiro, A. F. G.; Gomes, P. T.; Dias, A. R.; Blais, M.; Rausch, M. D.; Chien, J. C. W. J Polym Sci Part A: Polym Chem 1999, 37, 2471–2480. 23. Johnson, L. K.; A., B.; Dobbs, K.; Hauptman, E.; Lonkin, A.; Ittel, S.; McCord, E.; McLain, S. J.; Yin, Z.; Wang, Y.; Brookhart, M. Presented at the American Chemical Society Meeting, Orlando, FL, April 2002. 24. McLain, S. J.; Sweetman, K. J.; Johnson, L. K.; McCord, E. F. Presented at the American Chemical Society Meeting, Orlando, FL, April 2002. 25. Mecking, S.; Johnson, L. K.; Wang, L.; Brookhart, M. J Am Chem Soc 1998, 120, 888 – 899. 26. Johnson, L. K.; Killian, C. M.; Brookhart, M. J Am Chem Soc 1995, 117, 6414 – 6415. 27. Johnson, L. K.; Mecking, S.; Brookhart, M. J Am Chem Soc 1996, 118, 267–268. 28. Guan, Z. B.; Cotts, P. M.; McCord, E. F.; McLain, S. J. Science 1999, 283, 2059 –2062. 29. Mehler, C.; Risse, W. Makromol Chem Rapid Commun 1992, 13, 455– 459. 30. Mathew, J. P.; Reinmuth, A.; Melia, J.; Swords, N.; Risse, W. Macromolecules 1996, 29, 2755–2763. 31. Goodall, B. L.; McIntosh, L. H.; Rhodes, L. F. Macromol Symp 1995, 89, 421– 432. 32. Goodall, B. L.; McIntosh, L. H. (BFGoodrich). WO 9856839, 1998. 33. Goodall, B. L.; Makovetsky, K. L.; Finkelshtein, E. S.; Bykov, V. I.; Bagdasaryan, A. K.; Rhodes, L. F. (BFGoodrich). U.S. Patent 05,929,181, 1998. 34. Wang, C. M.; Friedrich, S.; Younkin, T. R.; Li, R. T.; Grubbs, R. H.; Bansleben, D. A.; Day, M. W. Organometallics 1998, 17, 3149 –3151. 35. Younkin, T. R.; Connor, E. F.; Henderson, J. I.; Friedrich, S. K.; Grubbs, R. H.; Bansleben, D. A. Science 2000, 287, 460 – 462. 36. Pangborn, A. B.; Giardello, M. A.; Grubbs, R. H.; Rosen, R. K.; Timmers, F. J. Organometallics 1996, 15, 1518 –1520. 37. Boaz, N. W. Tetrahedron: Asymmetry 1995, 6, 15– 16. 38. Ramaswamy, S.; Oehlschlager, A. C. Tetrahedron 1991, 47, 1145–1156. 39. Tabushi, I.; Yamamura, K.; Yoshida, Z. J Am Chem Soc 1972, 94, 787. 40. Randall, J. C. J Macromol Sci Rev Macromol Chem Phys 1989, 29, 201–317. 41. Klabunde, U.; Mulhaupt, R.; Herskovitz, T.; Janowicz, A. H.; Calabrese, J.; Ittel, S. D. J Polym Sci Part A: Polym Chem 1987, 25, 1989 –2003. 42. Watson, M. D.; Wagener, K. B. Macromolecules 2000, 33, 8963– 8970. 43. Wiberg, K. B. Angew Chem Int Ed Engl 1986, 25, 312–322. 44. Mehler, C.; Risse, W. Macromolecules 1992, 25, 4226 – 4228. 45. Johnson, L. K.; Brookhart, M. S.; Killian, C. M.; Arthur, S. D.; Feldman, J.; McCord, E. F.; McLain, S. J.; Kreutzer, K. A.; Bennett, M. A.; Coughlin, E. B.; Ittel, S. D.; Parthasarathy, A.; Tempel, D. J. WO-A 96/23010 CE. I. DuPont De Ne mours & University of North Carolina at Chapel Hill [Chem Abstr 1996, 125, 222773t] 1996. 46. Rische, T.; Waddon, A. J.; Dickinson, L. C.; MacKnight, W. J. Macromolecules 1998, 31, 1871–1874. 47. Arndt-Rosenau, M.; Beulich, I. Macromolecules 1999, 32, 7335–7343. 48. McKnight, A. L.; Waymouth, R. M. Macromolecules 1999, 32, 2816 –2825. 49. Breunig, S.; Risse, W. Makromol Chem Macromol Chem Phys 1992, 193, 2915–2927. 50. VanderHart, D. L.; Khoury, F. Polymer 1984, 25, 1587. 51. Earl, W. L. ; William, D. L. S. Macromolecules 1979, 12, 762–767. 52. Han, O. C.; Chae, S. A.; Han, S. G. Woo, S. K. Bull Korean Chem Soc 1996, 17, 1074.