Chemistry and catalysis advances in organometallic chemistry and catalysis
Download 11.05 Mb. Pdf ko'rish
|
- Bu sahifa navigatsiya:
- Figure 12.3
|
Scheme 12.8 M = Fe or Ru; R = H or Me; R 1 = H or Me; X = H, Me, or Ph. P R
P R X Fe Mo-I
– CH 2 =CH 2 Fe P R X P R X
R = Me,
t Bu, Ph, allyl; X = –(CH 2
4 −.
162 SYNTHESIS OF METALLOCENES VIA METATHESIS IN METAL COORDINATION SPHERES + cat.
racemate (
) (
) R 1 R 1 R 2 Fe R 1 R 1 R 2 Fe R 1 R 1 R 2 Fe cat. =
N i Pr
Pr O
Mo Me Me Ph t Bu
Bu
R 1 = t Bu, Cy, or SiMe 3 , R
2 = H or Me. expected ansa products in reasonable yields, albeit rather high catalyst loadings (10–20%) and prolonged reaction time under reflux conditions were required. Enantioselective RCM of planar-chiral ferrocenes was accomplished with a Mo-chiral catalyst (Scheme 12.10) [17]. The racemic substrates were kinetically resolved to yield (R)-ansa products and the recovered (S)-substrates with ee up to >99.5% and 95%, respectively. However, the presence of a methallyl substituent in the monosubstituted Cp ligand (R 2 = Me) was essential for the metathesis to proceed with high enantioselectivity. The authors suggested that the initial metathesis reaction of chiral (R)-[Mo] catalyst should occur preferably with the olefinic group in the planar-chiral trisubstituted Cp in order to provide significant stereoselectivity. Accordingly, the olefinic group that was remote from the planar-chiral moiety in the substrate should be less reactive than the other group (i.e., methallyl vs allyl) to render the reaction stereoselective. Enantioselective synthesis of planar-chiral phosphaferrocenes was recently achieved by asymmetric ring-closing synthesis using an appropriate chiral Mo catalyst (Scheme 12.11) [18]. The reactions proceeded with high yields (up to 95%) and ee (up to 99%) in the presence of the chiral (R)-[Mo] catalyst that was efficient in the kinetic resolution of the racemic planar-chiral substrates [17]. The stereochemical outcome of the reaction strongly depended on the structure of the allylic group in the phospholyl ligand: for R = H, R
1 =
Bu (Scheme 12.11), the bridged product was obtained in 65% yield but with only marginal ee (1%). Fortunately, for R = Me, the
Interligand ring-closing ene–yne metathesis of 1-alkenyl-1 -propargyl ferrocenes provided the expected [4]- or [5]ferrocenophanes in yields up to 84% (Scheme 12.12) [19]. These reactions could be performed with Ru-I or Ru-II catalysts in benzene or CH 2 Cl
at moderate temperature (40–60
◦ C); however, Ru-II was much more effective than the first-generation catalyst. In the presence of Schrock-Mo catalyst, the substrate (R 1 = Me, R 2 = R
3 = H, n = 1) was completely consumed into an undefined mixture of oligomeric products. A reaction of the substrate with a terminal alkyne moiety (R 1 = H) was moderately successful only under an ethylene atmosphere (15% yield). In this case, products of intermolecular metathesis with ethylene were also formed (17% yield). The complex with R 1 = SiMe
3 did not react under these conditions. In the case of the complex with a butenyl substituent (n = 2), because of the competition between intra- and intermolecular metathesis, high dilution conditions (0.01 mol/l) and higher temperature (60 ◦ C) were needed to facilitate 81% yield of the cyclic product. The ring-closing reaction was also Fe P R R X Fe P R R X – CH
2 =CH
2 cat.
cat. = (R)-[Mo] = N
Pr
Pr O O Mo Me Me Ph R 1 R 1 Scheme 12.11 R = H, Me; X = –(CH 2 ) 4 –, –(CH 2 ) 5 –, –CH
2 OCH
2 –; R
1 =
Bu, –C 6
3 -3,5-(CF
3 ) 2 , –C 6 H 3 -2,5-(CF
3 ) 2 . SYNTHESIS OF METALLOCENES BY RING-CLOSING METATHESIS 163 Ru-II
R 3 R 2 R 2 R 3 R 1 Fe ( ) n R 3 R 3 Fe R 2 R 1 R 2 ( ) n Scheme 12.12 R 1 = Me, R 2 , R 3 = H, Me, or t Bu, n = 1 or 2. accomplished with two complexes bearing trisubstituted Cp ligands (R 2 or R
3 =
Bu). The molecular structure of the cyclic product (R 1 = Me, R
2 = R
3 = H, n = 1) was determined by X-ray diffraction. The two Cp ligands are nearly eclipsed and slightly tilted (the dihedral angle = 5.04
◦ ), which is comparable to other [4]ferrocenophanes [15]. Nickelocene is a unique, moderately stable metallocene with 20 valence electrons (VE). Cyclopentadienyl ligands of nickelocene are labile and at least one of them is easily substituted in majority of its reactions [20]. Therefore, its stability in the presence of the Ru-metathesis catalysts was questionable; however, RCM of 1,1 -bis(alkenyl)nickelocenes proved to be successful (Schemes 12.13 and 12.14) [21]. Compound 3-1 is the first example of ansa-nickelocene. It is a dark-green, paramagnetic solid (mp = 95–97
◦ C), stable under an inert atmosphere, and soluble in common organic solvents. The molecular structure of 3-1 reveals that both cyclopentadienyl rings are flat and not parallel (Fig. 12.3). They are slightly inclined toward the bridge with the angle between the two Cp planes of 8.1(2) ◦ . The average Ni–C (2.179 ˚ A) and cyclopentadienyl C–C (1.424 ˚ A) bond distances in 3-1 are close to those determined for nickelocene at 101 K (2.185 ˚ A and 1.423 ˚ A, respectively [22]). Cyclopentadienyl rings in the compound 3-1 are in an eclipsed conformation (unlike in nickelocene), which is caused by the rigid bridge. The stereochemistry of the double bond in crystalline 3-1 is Z. When butenyl-substituted nickelocene was used, a mixture of the expected ansa-nickelocene (3-2) and a dinickelocene (3-3) resulting from homocoupling of two molecules of the substrate were produced (Scheme 12.14) [21]. X-ray studies reveal that the molecule of 3-3 is centrosymmetric consisting of two nickelocenes coupled by two 3- hexenylene bridges with E stereochemistry of the double bonds. The Cp ligands and the bridges are in an eclipsed conformation (Fig. 12.4). The rings are flat and slightly inclined (in an opposite direction than in 3-1) with a dihedral angle of 5.4(2) ◦ . The authors attributed this bending to the steric repulsion of the bridging hydrocarbon chains. Average Ni–C (2.183 ˚ A) and C–C cyclopentadienyl (1.423 ˚ A) bond distances are the same as in nickelocene. Consequently, metallocene derivatives featuring four alkenyl substituents with 1,3,1 ,3 -pattern (3-4, 3-5) [23] seemed as challenging substrates for any metathesis catalyst. Ring-closing reactions of 3-4 and 3-5 might provide a considerable number of regio- and stereoisomers with an unprecedented connectivity between the cyclopentadienyl ligands. Indeed, while the Ru-I was effective for the closure of only one bridge, employment of the more reactive Ru-II provided the first diansa-nickelocene
A single-crystal X-ray structure analysis revealed that the cyclopentadienyl rings in complex 3-6 were linked with 1,2 ,3,4 -connectivity by two E-hex-3-enylene chains (Fig. 12.5). Consequently, the molecule exhibits screw-shape geometry (i.e., axial chirality). The cyclopentadienyl rings are in an eclipsed conformation and are not parallel (dihedral angle 9.3 ◦ ).
formed mixed crystals (Scheme 12.16). Ni Ni – CH 2 =CH 2 Ru-I
3-1 Scheme 12.13 Ni Ni – CH 2 =CH 2 + Ru-I Ni Ni
3-3 Scheme 12.14 164 SYNTHESIS OF METALLOCENES VIA METATHESIS IN METAL COORDINATION SPHERES Figure 12.3 The molecular structure of 3-1 showing its eclipsed conformation. Figure 12.4 The molecular structure of 3-3. Ni – CH
2 =CH
2 Ru-I
or Ru-II Ni Ni – CH 2 =CH 2 Ru-II
3-4 3-6 Scheme 12.15 SYNTHESIS OF METALLOCENES BY CROSS-METATHESIS 165 Figure 12.5 The molecular structure of 3-6. Hydrogen atoms were omitted for clarity. Fe − 2CH
2 =CH
2 Ru-II
Fe 3-5 3-7 Scheme 12.16 M Cl Cl M Cl Cl – CH
2 =CH
2 Ru-I
( ) n ( )
n ( )
n ( )
n Scheme 12.17 M = Zr or Hf, n = 1 or 4. GC/MS analyses of the reaction mixture confirmed the presence of three isomers of the diansa-ferrocene, presumably E,E, E,Z, and Z,Z isomers. The structural data of the E,E-3-7 and E,Z-3-7 isomers were extracted from the X-ray data of the mixed crystals. Unexpectedly, the structure of the metallocene core of 3-7 compounds was not affected by the geometry of the double bonds in the bridges [24]. RCM was also efficiently performed for 16-electron, early transition metal metallocene dichlorides. Two research groups concurrently [15, 25] reported the intramolecular Ru-catalyzed reactions of bis(cyclopentadienylalkenyl) group IV bent metallocene derivatives shown on Scheme 12.17. Products with Z-double bonds were obtained for allyl-substituted complexes (n = 1), while the longer chain (n = 4) produced the cyclic product with the E-configured double bond. Similarly to the substituted ferrocene derivatives, significant diastereoselectivity was observed for the substrates bearing additional substituents in the Cp ligands (Scheme 12.18). When a mixture of meso and rac complexes was treated with Ru-I catalyst, the meso-diastereoisomer reacted faster that the rac-diastereoisomer. The rac-diastereoisomer underwent RCM reaction with the Ru-II catalyst. In another report, rac- and meso-metallocene dichlorides were separated by recrystallization. The metathesis of meso- isomer with Ru-I was inconclusive and no product was identified. However, the rac-isomer produced cleanly the expected rac–ansa complex in 87% yield (Scheme 12.19) [26]. 12.4 SYNTHESIS OF METALLOCENES BY CROSS-METATHESIS The first example of self-metathesis of a metallocene derivative, namely, vinylferrocene 4-1, to yield the E-homodimer, appeared in 1993 (Scheme 12.20) [10]. This reaction was considered as a model for ADMET polymerization of
166 SYNTHESIS OF METALLOCENES VIA METATHESIS IN METAL COORDINATION SPHERES R R
R M Cl Cl R R R R M Cl Cl R R R R M Cl Cl R R R R M Cl Cl rac rac meso meso – CH
2 =CH
2 Ru-I
– CH 2 =CH 2 Ru-II
Scheme 12.18 M = Zr, R = t Bu.
– CH 2 =CH 2 Ru-I
rac meso Inconclusive results rac R R Zr Cl Cl R R Zr Cl Cl R R Zr Cl Cl Ru-I
Scheme 12.19 R = t Bu.
Fe 2 – CH 2 =CH
2 Mo-III
Fe Fe
Scheme 12.20 divinylferrocenes and, owing to the isolation of the expected product in 54% yield, prompted further studies in this field (Schemes 12.5 and 12.6). Cross-metathesis was recognized as a convenient route to introduce various functionalities into organometallic frameworks, initially including ferrocenes. For example, catalyst with a methylideneferrocenyl ligand was prepared by a stoichiometric reaction of vinylferrocene (4-1) with the Schrock catalyst (Scheme 12.21) [3]. Polymers with one ferrocenyl redox-active end group were obtained with this unique initiator. Suitably substituted ferrocenyl alcohols and ketones were elaborated into more complex molecules using Ru-I catalyst under reflux conditions in CH 2 Cl 2 (Scheme 12.22) [27]. Predominantly E products were isolated in good yields together with small amounts of the self-metathesis products. Fe NAr Mo t -BuO
t -BuO
t -Bu
NAr Mo
-BuO
-BuO
Fe t-Bu − + Mo-II 4-1 Scheme 12.21 SYNTHESIS OF METALLOCENES BY CROSS-METATHESIS 167 Z + – CH 2 =CH 2 Ru-I
Fe R 1 R 2 Fe R 1 R 2 Z Scheme 12.22 R 1 = H, R 2 = OH, or R 1 = R
2 = O; Z = Ph, CH 2 SiMe
3 , CH
2 OAc, CO
2 CH 3 . Fe Ar Fe Ar + – CH 2 =CH 2 Mo-I
4-1 Scheme 12.23 Ar = Ph, 4-Me-C 6 H 4 , 2-thiophene, 2-furane, 2-naphthalene, 4-biphenyl. Ni + 2 Z − 2CH
2 =CH
2 Ru-II
Z Z Ni 4-2 Scheme 12.24 Z = CO 2 Me, COMe, CO 2 C(Me)
3 . The chemoselective cross-metathesis was further investigated for vinylferrocene 4-1 and a series of vinylarenes with the Mo-Schrock catalyst to yield π-conjugated molecules (Scheme 12.23) [28]. The desired heterodimers were obtained in good yields together with small amounts of homodimers. Only E products were reported in this system. Following these successful reports, olefin cross-metathesis of suitably substituted ferrocene derivatives was used in the synthesis of complex molecules, including dinuclear Zr/Fe polymerization catalysts (see Scheme 12.26) [29], rotaxanes [30], dendrimers [31]. In this context, taking into account that the Cp ligands in nickelocene are labile [20], we decided to develop novel nickelocene derivatives with polar functional groups by means of selective cross-metathesis. Thus, reactions of 1,1 - diallylnickelocene with α,β-unsaturated carbonyl compounds catalyzed with Ru-II were performed, using an excess of the organic olefin (Scheme 12.24) [32a]. Compound 4-2 (Z = COMe) crystallized in the monoclinic space group P2 1 /c with the Ni atom at the inversion center. The molecule adopts a staggered anti conformation. The average Ni–C (2.170 ˚ A) and C–C(cyclopentadienyl ring) (1.417 ˚ A) bond lengths are slightly shorter than those in nickelocene. The molecule of complex 4-2 (Z = CO 2 Me) in the solid state also adopts a staggered conformation; however, unlike in 4-2 (Z = COMe), the substituents are slightly inclined toward the Ni center [32b]. In both complexes, the substituents are approximately flat with E double bonds (Fig. 12.6). Cross-metathesis of the allyl-Cp substituted titanocene dichloride leads to the formation of dinuclear titanium complexes (Scheme 12.25, M = Ti, n = 1) [33]. The composition of the reaction mixture depends on the catalyst used. Treatment of the substrate with Ru-I (3 mol%) in benzene, toluene, or dichloromethane gave a mixture of Z and E isomers in a 1 : 1 ratio, while application of the Ru-II catalyst resulted with the formation of the pure E isomer. Similar reactions of the allyl-, 3-butenyl-, and 4-pentenyl-Cp substituted zirconocene dichloride have been described by Kuwabara et al. [29] (Scheme 12.25, M = Zr, n = 1 ÷ 3). Metathesis of the allyl-substituted zirconocene dichloride with Ru-II (5 mol%, room temperature) gave a mixture of E and Z isomers at 99 : 1 ratio. When Ru-I was used at 40 ◦ C, the ratio of isomers has changed to 6 : 1, still in favor of the E isomer. Heterodinuclear Zr/Fe complex has been prepared by cross-metathesis of 3-butenyl-Cp-substituted zirconocene dichloride with ferrocenylmethyl acrylate (Scheme 12.26) [29]. This Zr/Fe complex exhibited catalytic activity in ethylene and propylene polymerization similar to those of the starting Zr complex. Selective cross-metathesis of allyl-substituted ansa-zirconocene dichloride with Pd, Co, and Ni complexes having the acrylate pendant resulted in the synthesis of heterobimetallic complexes Zr/Pd, Zr/Co, and Zr/Ni (Scheme 12.27) [34]. 168 SYNTHESIS OF METALLOCENES VIA METATHESIS IN METAL COORDINATION SPHERES (a) (b)
Figure 12.6 The molecular structures of 4-2; (a) Z = CO 2
= COMe. (See insert for color representation of the figure.) – CH
2 =CH
2 Ru-I
or Ru-II 2 M Cl Cl ( ) n ( )
n M Cl Cl M Cl Cl ( )
n Scheme 12.25 M = Ti [33], Zr [29]; n = 1 ÷ 3. + – CH
2 =CH
2 Ru-II
Zr Cl Cl Fe O O Fe O O Zr Cl Cl Scheme 12.26 + N M O O N X X – CH 2 =CH 2 Ru-II
Zr Cl Cl Si Zr Cl Cl N M O O N X X Si Scheme 12.27 MX 2 = PdCl 2 ; CoCl 2 ; NiBr
2 (H 2 O). These complexes have been tested as initiators for ethylene polymerization in order to afford branched polyethylene. Catalytic activity of complexes, properties, and branched structures of polymers depended on the type of the late transition metal.
Allyl-Cp-substituted ansa-zirconocene dichloride also undergoes self-metathesis (Scheme 12.28) [33]. Owing to the planar chirality of the substrate and the possibility of the formation of Z and E isomers, coupling of two such molecules should lead to four possible stereoisomeric products. Reactions with Ru-I or Ru-II catalysts in various solvents (dichloromethane, benzene, or toluene) gave mixtures of products, in which a single isomer predominates (from circa 80% to 95%). The authors assumed that it was one of the two E isomers, but it was not clear whether rac- or mesoisomer was formed. A few examples of Mo-catalyzed alkyne cross-metathesis have been reported for ferrocenes and ruthenocenes. These reactions are summarized on Schemes 12.29 [35] and 12.30 [36].
|
