Organometallics 1995,14, 4617-4624
4617
Toward One-Component Group 4 Homogeneous Ziegler-Natta Olefin Polymerization Catalysts: Hydroboration of Zirconium Bisalkyls with Pendant 2-Propenyl Groups Using '[(C&)gBH]2 Rupert E. v. H. Spence and Warren E. Piers*?' Guelph-Waterloo Centre for Graduate Work i n Chemistry, Guelph Campus, Department of Chemistry and Biochemistry, University of Guelph, Guelph, Ontario, Canada N l G 2W1 Received May 25, 1995@
Alkylation of bis[~5-(2-propenyl)cyclopentadienyllzirconiumdichloride, 2, with CH3Li, KCHzC&, or LiCHzSiMes led to the bisalkyl derivatives 3a-c (R = CH3, a; CH2C&t5, b; CH2SiMe3, c), which were isolated as >95% pure yellow oils. The new ligand [2-(2-propenyl)cyclopentadienyl](tert-buty1amino)dimethylsilane(CP"~~Y~S~NR) was synthesized in a one-pot procedure and attached to zirconium via reaction of its dilithio salt with ZrCl4*2THF, producing (Cpal1y1SiNR)ZrC1Z, 4, in 84% yield. Alkylation with the above reagents yielded alkyl derivatives 5a-c (R = CH3, a; C H Z C ~ Hb;~ CHzSiMe3, , c) again as >95% pure yellow oils. Reactions of dichlorides 2 and 4 with the electrophilic borane [(C6F5)2BHln,1,proceeded smoothly to the expected products in which the pendant double bond(s) was hydroborated, giving [~5-C5H4CH~CH~CH~B(C6F5)212ZrC12, 6, and [t-C4HsNSiMez-v5-C5H3CH~CH2B' (C6F&lZrC12, 7. Attempts to alkylate these compounds met with failure. Similarly, hydroborations of most of the alkyl derivatives 3a-c and 5a-c were not clean by virtue of the availability of alternate reaction pathways from hydroboration. Bisbenzyl complex 5b, however, reacted cleanly with 1 to yield a complex, 8, whose structure was elucidated using multinuclear and two-dimensional NMR techniques. In this complex, both benzyl groups on zirconium were observed to transfer to boron, giving a cationic zirconium compound stabilized by a coordinated benzyl group from the counterion [(C$&,CHZ)B(C~F~)Z].
Introduction Group 4 based "bent" metallocene dihalides are important catalyst precursors for a variety of oligomerization and polymerization processes, of which homogeneous Ziegler-Natta a-olefin polymerization2 is arguably the most important. In commercial catalyst systems, the active species, believed to be a cationic metallocene alkyl, [ c p ~ M R l +is, ~generated from a CpzMXZ precursor when it is treated with a Lewis acid cocatalyst such as methylaluminoxane (MAOI4 or B(c~F5)3.~ The latter coactivator is advantageous because it is required only in stoichiometric quantities to generate active catalysts, allowing for direct study of these elusive species.6 Ligand modifications have played a key role in the development of new catalyst systems in which a remarkable degree of control over such polymer properties Abstract published in Advance ACS Abstracts, September 1,1995. (1)Present address: Department of Chemistry, University of Calgary, 2500 University Drive N.W.,Calgary, Alberta, T2N 1N4, Canada. E-mail:
[email protected]. (2) Horton, A. D. Trends Polym. Sci. 1994, 2, 158. (3) Jordan, R. F. Adu. Organomet. Chem. 1991,32, 325. (4) Sinn, H.; Kaminsky, W.;Vollmer, H. J. Woldt, R. Angew. Chem., Znt. Ed. Engl. 1980, 19, 390. (5) (a) Yang, X.; Stern, C. L.; Marks, T. J. J. Am. Chem. SOC.1991, 113,3623. (b) Yang, X.; Stem, C. L.; Marks, T.J. Angew. Chem., Int. Ed. Engl. 1992, 31, 1375. (6)Yang, X.; Stern, C. L.; Marks, T. J. J . Am. Chem. SOC.1994,116, @
10015.
as ta~ticity,~ block structure,* and the level of incorporation of a comonomelg may be realized. Given the impact of ligand structure (both steric and electronic) on the microstructure of the polymer produced, catalyst design remains an extremely active and important facet of research in this area. One of the principle targets in catalyst design has been a single-component catalyst with the activities of the commonly employed twocomponent systems but without the need for cocatalysts. Neutral, group 3, or lanthanide-based catalysts, which are isoelectronic to the cationic group 4 species, have been developed in this regardlo but thus far do not exhibit the required activities for commercial use. Another approach t o this problem involves covalent attachment of the counteranion in cationic group 4 catalysts, forming a zwitterionic species.l' In theory, such a catalyst could be a self-activating single(7) (a) Kaminsky, W.; Kiilper, K.; Brintzinger, H.H.;Wild, F. R. P. Angew. Chem., Int. Ed. Engl. 1985, 24, 507. (b) Ewen, J. A. J . Am. Chem. SOC.1984, 106, 6355. (8) Coates, G. W.; Waymouth, R. M. Science 1995,267, 217. (9) (a) Stevens, J. C.; Timmers, F. J.; Wilson, D. R.; Schmidt, G. F.; Nickias, P. N.; Rosen, R. K.; Knight, G . W.; Lai, S.Y. (DOW)European Patent EP-416-815-A2,March 13, 1991. (10) (a) Burger, B. J.; Thompson, M. E.; Cotter, W. D.; Bercaw, J. E. J. A m . Chem. SOC.1990, 112, 1566. (b) Shapiro, P.J.; Cotter, W. D.; Schaefer, W. P.; Labinger, J. A.; Bercaw, J. E. J . Am. Chem. Soc. 1994, 116, 4623. ( c ) Watson, P. L. J . Am. Chem. SOC.1982, 104, 337. (d)Jeske, G.;Lauke, H.; Mauermann, H.; Swepston, P. N.; Schumann, H.; Marks, T. J. J. Am. Chem. Soc. 1985,107, 8091. (11)Hlatky, G. G.; Turner, H. W.; Eckman, R. R. J . Am. Chem. SOC. 1989,111, 2728.
0276-733319512314-4617$09.00/0 0 1995 American Chemical Society
Spence and Piers
4618 Organometallics, Vol. 14, No. 10,1995
component system that lies dormant as a zwitterion in the absence of monomer. Another potential advantage of catalyst designs of this type would be control over catiodanion interactions; covalent attachment of the anion t o some location in the ligand structure may restrict ion.pairing and lead to more active catalysts.12 Covalent attachment of a Lewis acid coactivator to the ancillary ligand framework is one strategy for generating zwitterionic, self-activating catalysts, and, while attempts have been made,13successful incorporation of a Lewis acid activator of sufficient potency into an ancillary ligand framework remains a significant synthetic challenge. Given the reactive nature of a Lewis acid strong enough to abstract alkyls from a group 4 metal, it would be synthetically prudent to incorporate it into the catalyst in the last step of its synthesis. One possible means of carrying out this plan would be hydroboration of a pendant olefin functionality as the means of grafting a Lewis acidic boron center into a defined location in the supporting ligand. In light of the work of Marks et al., a logical borane reagent to employ would be [(C6F5)2BHIn,1, the synthesis of which we have recently reported.14 Borane 1 has proven to be an exceptional hydroboration reagent, as might be expected from its high electrophilicity. In this paper, we describe this reagent’s reactivity toward several biscyclopentadienyl and cyclopentadienyl(amido)zirconium dichlorides and bisalkyl derivatives with 2-propenyl groups dangling from the cyclopentadienyl donors.
Scheme 1
@
+
{L
E
(iw;ers)
Br
see ref. 15
1
I. n6uLi 2. 0.5 ZrCI4*2THF
2
3(1 R-CH3
3b R=CH&eHs 3c R=CH2SiMe3
Scheme 2
-‘ 1. MezSiCh 2.3 HzN:t-CdHg 3. 2 MeLi “Si
“‘i
I
ZrCI4.2THF
toluene
Results and Discussion The prevailing ancillary ligand arrays employed in homogeneous Ziegler-Natta catalysts are based on either a biscyclopentadienyl donor set or a chelating cyclopentadienyl-amido arrangement. We therefore chose to examine the hydroboration of catalyst precursors containing 2-propenyl groups attached to the cyclopentadienyl rings of these two families of compounds. The syntheses of biscyclopentadienyl type zirconium dichloride and bisalkyl precursors for reaction with borane 1 were straightforwardly accomplished using commonly employed metathetical routes as outlined in Scheme 1. The complex bi~[~~-(2-propenyl)cyclopentadienyllzirconium dichloride, 2, was synthesized using a procedure disclosed by Erker and Aul15 and was subsequently alkylated using alkyllithium reagents or benzyl potassium. The bisalkyl derivatives 3a-c were obtained as ’95% pure (by lH NMR spectroscopy) viscous yellow or orange oils. Proton and 13CNMR data for all new compounds reported herein are summarized in Table 1. On standing or cooling, bisalkyls 3a-b solidified into waxy solids which could be manipulated more conveniently, but attempts to purify these compounds via sublimation led to substantial decomposition. Crystals isolated from cooled solutions tended t o liquify upon isolation and warming to room temperature. In practice, therefore, these compounds were employed in further reactions as the crude products. (12)Deck, P. A.; Marks, T. J. J. Am. Chem. SOC.1995, 117, 6128. (13)Larkin, S. A.; Shapiro, P. J. Abstracts ofPapers, 209th Meeting of the American Chemical Society, Anaheim, CA, April 2-6, 1995; American Chemical Socieity, Washington, DC, 1995; INOR23. (14) Parks, D. J.; Spence, R. E. v H.; Piers, W. E. Angew. Chem., Int. Ed. Engl. 1995,34, 809;Angew. Chem. 1995, 107, 895. (15) Erker, G.; A d , R. Chem. Ber. 1991, 124, 1301.
5a R=CH3 5b R=CH&gHs 5c R=CHfiiMe3
4
A cyclopentadienyl-amido type ligand with a dangling 2-propenyl fhctionality, HzCpdYISiNR (R= t-CJI9) was obtained as a mixture of isomers in 90% yield after distillation from a one-pot procedure in which 2-propenylcyclopentadienyllithium was treated with MezSic12 followed by an excess of tert-butylamine (Scheme 2). Attachment of this ligand to zirconium was accomplished through reaction of its dilithio salt with ZrC42THF in toluene. Although the success of this methodology tends to be somewhat variable for these types of ligands, necessitating other approaches to ligand incorporation,16 we find LizCpal1y1SiNRt o be conveniently and reproducibly bound in this manner to form 4 in excellent yield. Alkylation of 4 with the same reagents employed for 2 again proceeded with no unusual observations to report. Like 3a-c, the bisalkyl derivatives 5a-c were either oils or gummy solids which defied purification and isolation by normal means. However, the com~
(16) Alternate routes are available: (a) Hughes, A. K.; Meetsma, A.; Teuben, J. H. Organometallics 1993,12,1936.(b) Piers, W. E.; Mu, Y.; MacGillivray, L. R; Zaworotko, M. J. Polyhedron 1995, 14, 1.
Group 4 Ziegler-Natta Catalysts
Organometallics, Vol. 14, No. 10,1995 4619
pounds were seen to be '95% clean by lH and 13C{lH} NMR spectroscopy (Table 1)and were used as the crude oils. As can be seen from inspection of this data, the planar chirality imparted by the 1,3-substituted cyclopentadienyl ligand renders all protons with the same connectivity diastereotopic. To our knowledge the effect of an asymmetric ligand environment of this type on polymer microstructure has not been investigated. Hydroboration of the dichloride derivatives 2 and 4 proceeded smoothly upon treatment with 2 and 1equiv, respectively, of the borane reagent 1, yielding the compounds 6 and 7 as shown in eq 1 and 2. Like
6
hydroboration of other olefinic substrates with 1, the reactions were extremely rapid and proceeded to completion essentially upon mixing of the reagents as indicated by the immediate disappearance of characteristic olefinic resonances of the 2-propenyl group in the 'H NMR and complete dissolution of the normally insoluble, dimeric borane l.14 In contrast, hydroboration of 2 with 9-BBN took several hours to reach c o n ~ l u s i o n .Within ~~ the detection limits of lH NMR spectroscopy, these hydroborations were highly regioselective, producing exclusively the 1,2 addition products shown. Both 6 and 7 are prone to pick up 1or 2 equiv of any Lewis base present in the system, and so donor solvents are to be avoided when handling these compounds. Complex 6, for example, may be viewed as a chelating Lewis acid, and consistent with this notion is its propensity toward binding 1 equiv of diethyl ether or THF when exposed t o these solvents. In the absence of donors, there does not appear to be any association between the Lewis acidic boron centers and the chloride ligands on zirconium. This is most convincingly demonstrated by the llB NMR chemical shifts found for 6
and 7,which are a t 73.9 and 73.0 ppm, respectively, These resonances are in the chemical shift range typical of neutral, three-coordinate boron centers.17 Attempts to cleanly alkylate these compounds with a variety of alkylating agents (MeLi, KCH2CsH5, AlMe3, ZnMe2) met with failure, producing complex mixtures of products. Similar results were found regardless of the reaction conditions employed. These observations were not completely unanticipated in light of the high electrophilicity of the boron centers in these molecules and confirm the notion that incorporation of these Lewis acid activators into the molecular structure of the catalyst is best left as the ultimate step in the synthesis. We therefore prepared the bisalkyl derivatives as described above and treated them with the borane 1in the appropriate equivalency. The compounds 3a-c and 5a-c were all found to react immediately with the borane, but only the reaction between 1 and the (Cpa"Y1SiNR)ZrBz2 derivative 5b led to a stable, isolable product (vide infra). The outcomes of the other reactions were highly dependent on the size of the alkyl groups. For example, the reactions involving the'two dimethyl compounds 3a and 5a were more complex than the reactions involving the bisalkyls incorporating larger (trimethylsily1)methyl or benzyl groups. In addition t o the desired hydroboration of the pendant olefin functionality, the borane also reacted with the zirconium carbon bonds as evidenced by the presence of both CH4 and CH3B(CsF5)2 in the reaction product mixtures.18In a separate study, 1 was found to be highly reactive toward the parent zirconocene bisalkyls; we have reported details of the reactions of CpzZr(CH3)z with 1 e1~ewhere.l~ Reactions of the bis(trimethylsily1)methyl derivatives 3c and 5c with 1 proceeded more selectively to give hydroborated products, as evidenced by the complete disappearance of olefinic signals in the lH N M R spectrum. Although spectra taken soon after mixing the reagents showed the reactions to be quite clean, the products were unstable in solution; decomposition to several products was observed within 30 min. In both cases, one of the compounds in the final mixture was identified as TMSCHZB(C~F&.~~ Unlike the above described chemistry, reaction between l and the dibenzyl complex 5b, when carried out (17) Kidd, R.G.In NMR of Newly Accessible Nuclei; Laszlo, P., Ed. Academic Press: New York, 1983; Vol. 2. (18)In addition to sharp signals for molecular species in the product mixtures of the reactions of dimethyl derivatives 3a and 5a with 1, broad, featureless resonances associated with a viscous oil were observed in the lH NMR spectrum. We speculate that this component of the product mixture results from the oligomerization of the pendant a-olefin functionality by active catalysts formed early in the reaction upon successful hydroboration and methyl abstraction. (This demonstrates another potential synthetic pitfall to be circumvented in the generation of these types of catalysts). Aside from the broad 'H NMR spectra, we note in support of this notion that only sharp signals were observed in the 'H NMR when a complex analogous to 5a but with a 2-methyl-2-propenyl group dangling from the Cp ring was reacted with 1 (Spence, R. E.v H.; Piers, W. E. Unpublished results), presumably because the gem-disubstituted olefin cannot be oligomerized with catalysts of this type. However, this reaction was also extremely complex due to the variety of reaction pathways possible and the production of several diastereomers in each pathway. (19) Spence, R. E. v H.; Parks, D. J.; Piers, W. E.; MacDonald, M. A,; Zaworotko, M. J., Rettig, S.J. Angew. Chem., Int. Ed. Engl. 1995, 34, 1230; Angew. Chem. 1995,107,1337. (20) 'H NMR, C6D6: 2.01, CH2 (m, 2H, 2 J =~1.5~Hz); -0.075, CH3 (s,9H). This compound was independently generated from CpzZr(CH2TMSh and ClB(CsFd2.
4620 Organometallics, Vol. 14, No. 10,1995
Spence and Piers
Table 1. lH and W{'H} N M R Data for Isolated New Compoundsa
Compound
lH N M R Data 6@Pm) Assignment
#
136.5 132.5 16.0, 1.6 116.5 11.0 116.7, 112.3 6.8 34.6
5.84 (m, 2H) 5.69, 5.50 (m. 8H) 5.02 (m, 2H) 4.98(m,H) 3.11 (d, 4H) -0.21 (6, 6H)
137.6 125.6 17.0, 1.5 115.7 10.0 111.2, 108.1 6.4 34.5 30.7 7.2
3b
7.22 (d, 4H) 6.90 (m,6H) 5.67 (m,2H) 5.57, 5.48 (m,8H) 4.91 (m. W ) 4.92 (m, 2H) 2.77 (d, 4H) 1.83 (s, 4H)
152.5 137.2 128.5, 126.1, 121.3 125.7 17.0, 1.4 115.8 10.5 113.5, 111.6 6.4 61.7 34.0
3c
6.03, 5.50 (m,8H) 5.91 (m. 2H) 5.03 (m. 2H) 5.00(m, 2H) 3.18 (d, 4H) 0.13 (s, 18H) 0.04 (s. 4H)
137.5 124.7 17.1. 1.8 115.8 10.1 112.1, 106.8 6.8 43.5 34.9 3.8
6.41, 6.10, 5.98 (m, 3H) 5.86 (m, 1H) 4.99 (m, 1H) 4.96 (m, 1H) 3.30 (d, 2H) 1.29 (s, 9H) 0.25 (s, 6H)
138.6 Cipo 135.8 =cH 17.0, 1.8 122.5, 121.8, 120.3 CpCH 10.0 117.1 =CH* 6.8 109.0 CpC-Si 57.7, 32.7 NC(CH3)3 34.1 CH2 0.78, 0.63 SiCH3
3a
5 a
6.38, 6.00, 5.82 (m, 3H) 5.85 (m, 1H) 5.06 (m, 1H) 4.98 (m, 1H) ~3.24(m, 2H) 1.34(s, 9H) 0.30(s, 6H) 0.11, 0.04(s,6H) 7.11, 7.05 (m,4H) 6.95, 6.90 (m,2H) 6.80, 6.60 (44H) 6.04, 5.84, 5.76 (m, 3H) 5.76 (m.1H) 4.97 (m,W ) 3.16(dd, 1H) , 2.95(dd, 1H) 2.17, 1.63 (4 2H) 1.56, 1.21 (d 2H) 1.19 (I, 9H) 0.33, 0.28(s. 6H)
/e-@
'eM
CHpPh
'f
'CHzPh
=cH
5.88, 5.72 (m, 8H) 5.84 (m, 2H) 4.97 (m, 2H) 4.96 (m,2H) 3.36 (4 4H)
2b
si
J(Hz)
13CllHl . -NMR Data )"6 Assignment
5b
CpH
=cHc H 2 =cHH =cHH =ccH2 WW3)3 SiCH3 m
17.2 10.2 6.7
3
7.3, 7.4 7.7, 7.1
15.8. 6.2 15.8, 6.6 9.5 10.6
cipso
=CH2
cfl (332
=ai Cipso
ICH2
cfl cH2
ZrCHJ
=cH Cipso
"2
cpclr CH2 -2
SiCH3
136.9 =cH 132.7 tip, 120.1, 119.0, 116.8 CpCH 116.2 " 2 102.9 cpc-si 55.5, 34.3 Wm3h 36.4 CH2 34.0d -3 1.59, 1.54 SiCH3 147.3, 142.5 137.4 132.1 130.4, 129.2, 127.5, 127.1, 123.2, 122.1 119.8, 119.3, 118.3 115.7 105.7 57.3, 33.9 55.5, 52.9 33.9 1.7. 1.2
Group 4 Ziegler-Natta Catalysts
Organometallics, Vol. 14, No. 10,1995 4621
Table 1 (Continued) 1~
Compound
#
NMR~ a t a
G@P@
Assignment
6.59, 6.14, 6.07 (m,3H) 5.93 (m,1H) 5.08 (m,1H) 5.02 (m,1H) 3.41 (dd, 1H) 5 C 3.32 (dd, 1H) 1.36 (8, 9H) 0.87, 0.20 (d, W ) 0.63, 0.05 (d, W) 0.37, 0.35(s, 6H) 0.17, 0.16(s, 18H)
CpH
5.89, 5.57 (m, 4H) 2.75 (t. 4H) 1.90(brt,4H) 1.76 (m, 4H)
CPH CCH2
6
7
6.41, 6.12, 5.99 (m,3H) 2.78 (m,1H) 2.60 (m,1H) 1.86(brfW) 1.77 (m,W) 1.29 (9, 9H) 0.27 (s, 6H) 7.18 (m,3H) 7.05 (m,W ) 6.94, 6.66 (d, 2H) 6.07, 6.81 (m,W) 5.82, 5.41, 4.96 (m,3H) 5.76 (m,1H) 2.97 (br I, W) 2.93, 2.83 (d, 2H) 2.24, 1.12 (m,W) 2.11 (m, W) 0.89, 0.31 (m, W) 0.77 (s, 9H) 0.24, 0.05 (s, 6H)
13C(1H)N M R Data J(Hz) G(ppm) Assignment 137.1 =cH 132.7 16.8 119.9, 119.0. 116.7 10.1, 1.7 116.1 15.3, 6.4 103.4 15.3, 6.6 56.0, 34.3 52.6, 52.6 11.2 34.4 11.6 3.8, 3.3 1.9, 1.7
B W
W
6.5 7.5
133.4 117.7. 110.8 32.7 32.0 (br) 26.2
CipdJCP) CpCH
139.9 122.7, 121.5 120.1 109.5 58.0, 32.6 32.4 32.0 26.5 0.6, 0.5
tips* (CP) CpCH CpC-Si NCW3)3 CCHz CHzB
cm2
CH2B a
2
m 2
SiCH3
159.7, 151.1. 147.7 129.1, 127.8, 123.8 128.7, 127.4 128.6, 128.2 127.3, 115.1, 112.7 121.9 113.6 59.0, 34.3 50.3 40.5 37.9, 33.3 27.2 5.0, 0.46
a lH NMR spectra were recorded in C& at room temperature at 400 MHz except where noted. 13C{lH) NMR spectra were obtained in C j D g at room temperature at 100 MHz except where noted. Spectra were
referenced to solvent peaks. b NMR data for compound 2 in CDC13 were reported in reference 15. C 1H and 13C{lH} spectra recorded at 200 and 50 MHz, respectively. The resonance for the other diastereotopic ZrCH3 group was obscured by the signal for the t-butyl group. e 1H spectrum recoded in C7Dg at -2bT.
in hexanes, led to a product, 8 , in 61% yield as a yellow powder that precipitated from solution (eq 3). The
precise structure of this compound is unknown due to our inability to obtain single crystals for X-ray analysis; the compound tends t o oil out of the aliphatic solvent systems employed and decomposes in chlorinated or donor solvents. Microanalytical data are consistent with an empirical formula comprising the components of the reaction, 5b and 1. Several lines of spectroscopic evidence led us to propose the structure depicted in eq 3 in which an exchange of both benzyl groups from zirconium to boron has taken place and in which the propyl side chain on the ring is tethered to the zirconium center. Immediately evident from lH, 19F,and llB{lH} NMR data was the fact that hydroboration and alkyl abstraction by the boron center had taken place. Olefinic resonances from the 2-propenyl group were no longer
Spence and Piers
4622 Organometallics, Vol. 14, No. 10, 1995
been removed from organozirconium compounds using the Lewis acid abstractor B(C6F5)3.6*21In the absence of stronger bases, the cationic zirconium center obtains electrons'from the Jc-system of the nonfluorinated arene ring of the benzyl group. Because of the asymmetric coordination environment about zirconium in 8 all five phenyl protons of the coordinated ring are inequivalent; a COSY experiment allowed the ortho, metu, and pura protons to be assigned as given in Table 1. The data discussed to this point do not distinguish between the structure proposed for 8 in eq 3 and one akin to that shown in A which would have resulted if
:
7.0
6.0
5.0
3.0
4.0
2.0
1.0
ppm
Figure 1. 400 MHz lH NMR spectrum of 8 (C7D8, -20 "C).
/t ,,,
I
3.0
2.5
2.0
1.5
1.0
0.5
0.0
FI (ppm)
Figure 2. COSY45 plot of the 0-3.3 ppm region of the 1H NMR spectrum of 8. The cross-peak indicated by the arrow is due to the hexane present in the sample. present. Furthermore, the chemical shift of the llB resonance (-9.1 ppm) indicated a four-coordinate, anionic boron center. The pattern of resonances in the 19F NMR spectrum (-132.9, ortho; -164.1, para; -167.3, meta) in which the resonance for the para fluorine atom was shifted upfield by about 15 ppm in comparison to that in neutral, three-coordinate compounds, was also typical of such a boron environmente6 Finally, in the 13C{lH} NMR spectrum, two broad resonances attributable to carbons bonded to boron were detected a t 33.3 and 37.9 ppm. Only one such resonance would be expected if the reaction had stopped a t simply hydroborating the pendant double bond. The proton NMR spectrum of 8, accumulated at -20 "C, is shown in Figure 1. In the aromatic region of the lH NMR spectrum, in addition to a normal pattern of resonances for a typical benzyl group, five separate resonances were observed for the phenyl protons of the other benzyl moiety. The upfield shift of these resonances from their usual positions is indicative of coordination of the arene ring to the zirconium center. Similar features were evident in the l3C(lH} NMR spectrum (Table 1). This sort of behavior has been observed in other instances where benzyl groups have
A
the reaction between 5b and 1 had stopped at hydroboration and benzyl abstraction. Although a more complex reaction sequence must take place to arrive at 8, structure A was rejected as a result of a heteronuclear multiple quantum coherence (HMQC) experiment22 which clearly showed that the two broad resonances in the 13C{lH} NMR spectrum assigned to carbons bonded to boron correlated to the broad resonance at 2.97 ppm and to the doublets at 2.93 and 2.83 in the 'H NMR spectrum which we ascribed to the benzyl methylene groups. If the assignment of the benzyl protons is correct, this result implies that both benzyl groups have been transferred to boron. In support of this assignment, the COSY experiment showed that the resonances around 2.9 ppm do not correlate to any other peaks in the spectrum (Figure 2 shows the COSY plot for the 0-3.5 ppm region of the spectrum). Also, a separate NOESY experiment on the sample showed that these benzyl resonances correlate only with ortho protons on the phenyl rings, as would be expected on the basis of their physical proximity. The HMQC and COSY experiments also aided in assignment of the six inequivalent protons and carbon resonances for the three methylenes of the propyl chain connecting the Cp ring to zirconium. Of particular interest were the two proton resonances at 0.31 and 0.89 ppm which were correlated to the carbon resonance a t 50.3 ppm. The COSY plot shown in Figure 2 clearly establishes the connectivity of the protons in the CCH2CH2CHS (X = B or Zr) and shows that these two protons and the carbon are due to the methylene unit bonded to X. The low-field position of the carbon peak is characteristic of carbons bonded to cationic zirconium centers; protons on such carbons typically also have rather high-field proton resonance^.^^ Although carbon atoms bonded to the B(C6F& fragments typically resonate near 50 ppm, RCH~B(CGF&protons typically (21),(a)Pellachia, C.; Immirzi, A,; Grassi, A.; Zambelli, A. Organometallics 1993,12, 4473.(b) Pellachia, C.;Immirzi, A.; Grassi, A. J. Am. Chem. SOC.1993,115, 1160.(c) Bochmann, M.; Lancaster, S. J.; Hursthouse, M. B.;Abdul Malik,K. M. Organometallics 1994,13,2235. (d) van der Linden, A,; Schaverien, C. J.; Meijboom, N.; Ganter, C.; Orpen, A. G . J . Am. Chem. SOC.1996,117,3008. (22) Bax, A.; Subramanian, S. J. Magn. Reson. 1986,67,565. (23) Jordan, R. F.; LaPointe, R. E.;Bradley, P. K.; Baenziger, N. Organometallics 1989,12, 2892.
Group 4 Ziegler-Natta Catalysts
Organometallics, Vol. 14, No. 10, 1995 4623
Scheme 3
boration to incorporate Lewis acidic boron centers into the ancillary ligand structure appears to be viable, but several observations herein provide guidance as t o the conditions under which such chemistry may successfully produce a catalyst with the desired features. For example, since borane reagent 1 is also reactive toward metal-carbon bonds, alkyl groups or pendant functionalities must be chosen to bias the reagent toward hydroboration, which must be irreversible. It also appears from this study that the three-carbon tether is too long to avoid the exchange shown in Scheme 3 which leads to a complex such as 8. Close inspection of the lH NMR spectrum of 8 shown in Figure 1 indicates that another species is present in minor amounts as evidenced by small ligand signals in the baseline. This is perhaps indicative of a species related t o the desired structure A in which such an exchange has not taken place. This exchange is likely entropically favored in this particular system but may be enthalpically disfavored in related complexes with two- or one-carbon chains linking the boron to the cyclopentadienyl donor. We are currently exploring these and other strategies to circumvent the problems exposed in this work.
resonate in between 2 and 3.5 ppm.14 This data also favors 8 over the alternate structure A. Other observations compatible with our assignment were the observed temperature dependence of the lH and 19FNMR spectra. Consistent with the presence of a loosely coordinated arene ring, exchange of substituents on boron was observed t o be facile. In the 19F spectrum of 8 at room temperature, only one set of sharp signals for the diastereotopic CsF5 groups was observed. Upon being cooled, each resonance broadened, the signal due to the ortho fluorines most severely. At around -50 "C coalescence was observed, the signals emerging as four peaks at -129.9, -131.7, -134.4, and -135.3 ppm (average = -132.8 ppm) due to four diastereotopic ortho fluorines in the slow exchange regime. Chemical shift dispersion in the meta and para fluorines was not as extensive, but signficant broadening was observed in these resonances also. In the lH NMR spectrum, the slow exchange spectrum was observed at -20 "C (Figure 1). Upon warming, the resonances assigned to the benzyl methylene protons broadened and coalesced into one broad resonance a t 2.88 ppm; similarly all of the resonance for the phenyl protons coalesced into a broad peak centered around 6.89 ppm. Other resonances in the spectrum, most notably the signals for the diastereotopic methyl groups on silicon, remained sharp. These observations may be rationalized on the basis of an exchange process involving dissociation of the coordinated arene22aand recoordination of the phenyl ring of the other benzyl group. The details of how 8 forms are unknown, but the steps which must take place are depicted in Scheme 3. Likely hydroboration takes place first; subsequently an exchange between one of the benzyl groups (presumably the one closest to the dangling boron center) and the boron propyl moiety occurs. Whether this is a concerted process as shown in Scheme 3 or a stepwise sequence is unknown. The intermediacy of the neutral products is also in question; no evidence for production of CsH5CH2B(CsF& was observed, but, as mentioned above, in the hydroboration of 5c some TMSCH2B(CsF& was identified in the mixture of products. If produced, abstraction of the remaining benzyl group by C6H5CHzB(CsFd2 would lead to 8 as shown; alternatively, the exchanges necessary could be concerted in nature.
Summary and Concluding Remarks Although we have not yet arrived at an internally activated catalyst system, our results show that such a goal is a realistic one. The strategy of using hydro-
Experimental Section General Procedures. All manipulations were performed either in a Braun MB-150 inert atmosphere glovebox or on greaseless vacuum lines equipped with Teflon needle valves (Kontes) using swivel frit-type glassware. Solvents were purified by distillation from a n appropriate drying agent and stored in evacuated pots over either "titanocene" 24 (aliphatic solvents), sodium benzophenone ketyl (diethyl ether, THF), or CaHz (chlorinated solvents). &-Benzene was dried sequentially over activated 3 A sieves and "titanocene" and was stored in the glovebox; other NMR solvents were dried analogously to the perprotio solvents. The following reagents were synthesized using literature [(C&'5)2BH12,l4 and procedures: [q5-C&T&H~CH=CH21~ZrC12,15 K C H Z C ~ HAll ~ . other ~ ~ materials were obtained from Aldrich and either used as received or dried and distilled prior to use. Samples were analyzed by NMR spectroscopy on Varian Gemini 200 MHz and Unity 400 MHz instruments. 'H spectra were referenced to residual protons in the deuterated solvent medium. l3C(lH} NMR were referenced to solvent signals; signals broadened severely due to bonding to quadrupolar boron nuclei were detected using the HMQC pulse sequence. 19FNMR spectra were recorded at 376.2MHz and referenced to CFCl3. llB{lH) NMR spectra were recorded at 128.3MHz and referenced to BFyEt20 at 0.0 ppm. Microanalyses were performed by Oneida Research Services, Inc., One Halsey Rd., Whitesboro, NY 13492. Preparation of [t15-CeH4CH2CH=CH212Zr(C~)2, 3a. To dichloa sluny of bis[q5-(2-propenyl)clopentadienyllzirconium ride (300mg, 0.81 mmol) in diethyl ether at -78 "C was added a 1.4M diethyl ether solution of methyllithium (1.15mL, 1.61 mmol). The cooling bath was removed, and the reaction was stirred at room temperature for 15 min. The diethyl ether was removed in vacuo,hexanes were added, and the reaction was filtered. Removal of the hexanes gave a '95% pure (by NMR) yellow oil that could not be further purified. Preparation of [t15-C5~CH2CH=CH212Zr(CH2CsH5)2, 3b. THF (10 mL) was condensed into an evacuated flask containing bis[q5-(2-propenyl)cyclopentadienyllzirconium dichloride (624mg, 1.68mmol) and benzylpotassium (436mg, 3.35 (24)Manrich, R. H.; Brintzinger,H.H.J. Am. Chem. SOC.1971, 93, 2046. (25) Schlosser; M.; Hartmann, J.Angew. Chem., Int. Ed.Engl. 1973, 12, 508.
4624 Organometallics, Vol. 14,No. 10, 1995
Spence a n d Piers
Preparation of CpaUylSiNRZr(CHzCeH&, 5b. THF (30 mmol) at -78 "C. The reaction was warmed to room temperature, at which point it became very dark in color. The THF mL) was condensed into an evacuated flask containing Cpal1y1was removed in uacuo, and hexanes (20 mL) were added. SiNRZrClz (1.05 g, 2.65 mmol) and benzylpotassium (0.716g, Filtration of the reaction mixture gave an orange solution that 5.5 mmol) a t -78 "C. The reaction was warmed to room yielded a '95% pure (by NMR) oil on solvent removal. temperature, at which point the red color of the benzylpotasPreparation of [q6-C&CH~CH=CH~1~Zr(CH~SiMes)z,sium was observed to disappear. The volatiles were removed 3c. To a slurry of bis[q5-(2-propenyl)cyclopentadienyl]zirin uacuo, hexanes (25mL) were added, and the reaction was conium dichloride (588 mg, 1.58 mmol) in diethyl ether (15 taken t o reflux and filtered. Removal of the hexanes gave a mL) at room temperature was added a 1 M pentane solution '95% pure yellow oil, which solidified to a gummy solid on of [(trimethylsilyl)methyl]lithium(3.16mL, 3.16mmol). The standing: yield, 1.209 g, 2.39 mmol, 90%. reaction was stirred for 15 min before the volatiles were Preparation of CpdylSiNRZr(CHzSiMe&, 5c. To a removed in uacuo. Hexanes (10mL) were added, the reaction solution of CpdylSiNR.ZrClz(198mg, 0.5 mmol) in diethyl ether was filtered, and the hexanes were removed to give a '95% (15mL) at 0 "C was added a 1.0M solution of [(trimethylsilyl)pure (by NMR) oil that solidified on standing: yield, 688 mg, methylllithium in pentane (1.0mL, 1.0 mmol). The reaction 92%. was warmed to room temperature, and the volatiles were Preparation of [2-(2-propenyl)cyclopentadienyl](tertremoved in uacuo. Hexanes (10 mL) were added, and the butylamino)dimethylsilane,H2Cpdy1SiNR. To a solution reaction was filtered, leaving a '95% pure oil upon removal of [(2-propenyl)cyclopentadienylllithium(8.0 g, 71.4mmol) in of the solvent. THF (200mL) at -78 "C was rapidly added MezSiClz (9.21g, 71.4mmol) via syringe. The reaction was warmed t o room Preparation of [q5-C~H4CH~CH~CH~B(CsFa)~lzZrClz, 6. temperature and was stirred for 15 h. To this reaction was Bis(pentafluorophenyl)borane,1 (279mg, 0.81mmol), and bisadded tert-butylamine (20mL, 190 mmol), and a precipitate [qs-(2-propenyl)cyclopentadienyl]zirconiumdichloride (150mg, immediately formed. After the solution had been stirred for 0.40 mmol) were combined in a flask, and =lo mL of CHzClz 2 days, the THF was removed in uucuo and hexanes (200mL) was vacuum transferred into the flash at -78 "C. The reaction were added. After filtration, the hexanes were removed from was stirred while being warmed to room temperature. The the orange solution and the resulting oil was distilled. The solvent was removed in uacuo, and the resulting gummy solid product was obtained as a mixture of isomers: yield, 15.1 g, was triturated with hexanes. The white solid was isolated by 64.3mmol, 90%; bp 49 "C, Torr. lH NMR (CDC13, ppm): filtration and washed with hexanes: yield, 329 mg (77%). 19F 6.72,6.45,6.21,6.11,CpH, 5.92,5.18,4.98,CH-CH2; 3.39, NMR (C6D6): -132.0 (ortho, 2F), -149.0 (para, IF), -162.7 NH, 3.17,CCH2; 3.03,2.96,CpH; 1.20, 1.11,NCH3;0.20,0.19, (meta, 2F). llB{lH} NMR (C6D6): 73.9. -0.04,SiCH3 Preparation of [ ~ - C ~ H ~ N S ~ ~ M ~ ~ - ~ ~ - C ~ H ~ C Preparation of Li2Cpdy1SiNR. To a solution of [2-(2(CsF&lZrC12, 7. Hexane (20 mL) was condensed into a n propenyl)cyclopentadienyll(tert-butyla~no)dimethylsilane (6.7 evacuated flask containing [(tert-butylamino)dimethyl(2-allylg, 28.5 mmol) in diethyl ether (150mL) was added a 1.4 M cyclopentadienyl)silane]zirconium dichloride (229 mg, 0.578 diethyl ether solution of methyllithium (45 mL, 63 mmol). mmol) and bis(pentafluoropheny1)borane (200mg, 0.578mmol) Addition was completed over 10 min and led to an orange at -78 "C. The reaction was warmed to 35 "C until all the solution. After a few hours a voluminous white precipitate reagents went into solution. The solvent volume was reduced started to form, and the reaction was stirred for a total of 15 by half, and the reaction was cooled t o -78 "C. The white h. The product was isolated by filtration and washed with a little diethyl ether: yield, 5.22g, 21.1 mmol, 74%. solid that precipitated was isolated by filtration: yield, 288 Preparation of CpdylSiNRZrClz,4. Toluene (50 mL) was mg, 67%. Anal. Calcd: C, 42.12;H, 3.26;N, 1.89. Found: condensed into an evacuated flask containing LizCpa1IY1SiNR c , 42.56;H, 3.00;N, 1.76. "F NMR (C6D6): -131.8 (ortho, (2.118g, 8.58 mmol) and ZrCld(THF)2(3.236g, 8.58 mmol) at 2F),-148.7 (para, lF), -162.5 (meta, 2F). "I3 NMR -78 "C. The reaction was warmed to room temperature and 73.0. was stirred for 40 h. The toluene was removed in uacuo, and Synthesis of 8. To a stirred suspension of bis(pentaflu0hexanes (50 mL) were added to the residues. After heating ropheny1)borane (823mg, 2.39 mmol) in hexanes (5 mL) was the hexanes to reflux, the reaction was filtered to leave an added a solution of Cpa11Y1SiNRZr(CH2C6Hs)2 (1.209g, 2.39 orange solution. The solvent volume was reduced t o apmmol) in hexanes (10mL). As the reaction proceeded a yellow proximately 30 mL, the solution was cooled to -78 "C, and precipitate formed. After 2 h, the reaction was cooled to -78 the resulting white crystals were isolated by filtration: yield, "C and filtered: yield, 1.247 g, 1.46mmol, 61%. C, 56.33;H, 2.86 g, 7.23 mmol, 84%. Anal. Calcd: C, 42.51;H, 5.86;N, 4.49; N, 1.64. Found: C, 56.65;H, 4.96;N, 1.68. 19F NMR 3.54. Found: C, 43.29;H, 5.81; N, 3.50. (C6D6) -132.9 (ortho, 2F), -164.1 (para, IF), -167.3 (meta, Preparation of Cpa"YISiNRZr(CH&,5a. To a solution 2F). llB{lH} NMR (C6D6) -9.1. of Cpa"Y1SiNRZrC12(554mg, 1.4mmol) in diethyl ether (15mL) at room temperature was added a 1.4M diethyl ether solution of methyllithium (2 mL, 2.8 mmol). After the solution had Acknowledgment. Funding for this work was probeen stirred for 10 min, the diethyl ether was removed in uacuo vided by the Novacor Research and Technology Center and hexanes (10mL) were added. The reaction was filtered, of Calgary, Alberta, and is gratefully acknowledged. and the hexanes were removed to give the product as a pale yellow oil that was pure by lH NMR spectroscopy. OM9503858