A Crystallizable f‐Element Tuck‐In Complex: The Tuck‐in Tuck‐over Uranium Metallocene [(C 5 Me 5 )U{μ‐η 5 :η 1 :η 1 ‐C 5 Me 3 (CH 2 ) 2 }(μ‐H) 2 U(C 5 Me 5 ) 2 ]
2008; Wiley; Volume: 47; Issue: 27 Linguagem: Inglês
10.1002/anie.200801062
ISSN1521-3773
AutoresWilliam J. Evans, Kevin A. Miller, Antonio G. DiPasquale, Arnold L. Rheingold, Timothy Stewart, Robert Bau,
Tópico(s)Radioactive element chemistry and processing
ResumoLong sought structural data on an f-element tuck-in complex have been obtained for the title compound 1 that contains the first example of both tuck-in and tuck-over bonding in a ligand derived from C5Me5− by metalation (see scheme). One of the features of the C5Me5− group that makes it such a desirable ligand in organometallic chemistry is the fact that it is relatively inert to the CH activation that often complicates the chemistry of C5H5− metallocene complexes.1–4 Although C5Me5− is more resistant to attack on the CH bond, its methyl groups can be metalated with some highly reactive metal species.5–10 Since these CH activated ligands arise from extremes in reactivity, they have been involved in unprecedented reactions. For example, the homogeneous CH activation of methane was first discovered with [{(C5Me5)2LuMe}n]5 based on a mechanism involving the "tucked-in"6 complex "[(C5Me5){C5Me4(CH2)}Lu]". Tuck-in6 intermediates have also played prominent roles in explaining the CH activation in complexes such as [(C5Me5)2ScMe]6 and [(C5Me5)2Th(CH2CMe3)2].11 Despite the repeated use of tuck-in complexes in mechanistic schemes involving f elements, no spectroscopic or crystallographic evidence has ever been presented to support the existence of such an intermediate in a lanthanide or actinide complex. Tuck-in complexes have been crystallographically characterized with transition metals.7, 9, 10, 12 For example, in the {(C5Me5)2TiH} system, [(C5Me5)Ti{η5:η1-C5Me4(CH2)}] (1) could be isolated.7 However, in these transition-metal tuck-in compounds, low-oxidation-state tetramethylfulvalene resonance structures can contribute to the stability of the complexes.7, 9, 10, 12, 13 With the limited oxidation states available for f elements, this is less likely. 1 Crystallographic evidence for an alternative type of C5Me5− metalation has been obtained with lanthanides in the form of "tuck-over" complexes [(C5Me5)2Ln(μ-H)(μ-η1:η5-CH2C5Me4)Ln(C5Me5)] (2; Ln=Y,14 La,15 Sm,16 and Lu17) in which the methylene group formed by CH activation is attached to a second metal atom. In some cases, double CH activation at two methyl groups resulted in C5Me3(CH2)23− ligands, for example, in [{(C5Me5)3Ln2[C5Me3(CH2)2]}2] (3; Ln=Ce18 and Sm19).13, 20) The only crystallographic evidence for this type of CH activation of a C5Me5− ligand in an actinide complex involves a methylene group attached to nitrogen not the metal. Thermolysis of the U6+–imido complex [(C5Me5)2U(NAd)2] (Ad=1-adamantyl) formed [(C5Me5)U{η1:η5-NAd(CH2C5Me4)}(NHAd)] (4,21)—a reaction that could involve a tuck-in (or tuck-over) intermediate. The only other suggestion of C5Me5− metalation in actinide chemistry is kinetic data on a multiple-pathway transformation involving formation of [(C5Me5)2Th(μ-CH2)2EMe2] from [(C5Me5)2ThR2] (R=CH2EMe3; E=C, Si).11 We report here the first crystallographically characterized tuck-in complex of an f element and the first crystallographic data on uranium tuck-in and tuck-over structures. Both are found in the same structure (Figure 1). This bonding mode has not been observed previously in any metal complex of ligands derived from C5Me5− to our knowledge. Molecular structure of 7 (thermal ellipsoid drawn at the 50 % probability level). X-ray crystallography revealed the structure of complex 7 (Figure 1). A doubly-metalated [μ-η5:η1:η1-C5Me3(CH2)2]3− ligand is engaged in both tuck-in and tuck-over binding to uranium. The large thermal ellipsoids in the C1–C5 ring are likely the result of multiple ring orientations. However, resolution limits in the X-ray data were not adequate to propose a disordered model. Consequently, the disorder was treated as high thermal activity. Although the bridging hydride ligands were not located in the X-ray crystal structure, their presence was established by other means. Attempts to prepare a diamagnetic thorium analog according to Equation (1) were unsuccessful. The infrared spectrum of 7 displayed a broad band centered at 1164 cm−1 in the region typical for U-H-U stretching modes. This absorption band is similar to those observed for both 5 and 6,22, 24 which give rise to broad bands centered at 1188 and 1176 cm−1, respectively. Attempts to synthesize a deuterium analog of 7 were thwarted by the fact that neither [{(C5Me5)2UD2}2] nor [{(C5Me5)2UD}2] were accessible due to hydride exchange with the C5Me5− rings.24 The infrared spectrum of the product made according to Equation (1), but from precursors synthesized from [(C5Me5)2UMe2] and D2, was identical to that of 7. The presence of hydrides in 7 was probed by measuring the gas evolution during the synthesis of 7 from 622 by means of a Toepler pump. Only one equivalent of dihydrogen per two uranium atoms was obtained as shown in Equation (1). Two equivalents of dihydrogen would be expected, if a U3+ complex without hydride ligands was formed instead of 7. The reactions of 7 with phenol and C6H5OD gave further support for the presence of hydrides. Reaction of 7 with phenol yields H2 and the bis(phenoxide) complex [(C5Me5)2U(OPh)2] (8) in 92 % yield [Eq. (3)]. The identity of 8 was confirmed by X-ray crystallography (Figure 2). Reaction of 7 with C6H5OD gave HD28 and a product that had 2D NMR resonances at 3.1 ppm consistent with the presence of deuterium in place of hydrogen in C5Me5− rings of 8. Molecular structure of 8 (thermal ellipsoid drawn at the 50 % probability level). The structure is similar to those of [(C5Me5)2U(EPh)2] (E=S, Se)26 except the U-O1-C21 and U-O2-C27 angles (174(1) and 172(1)°) are larger.27 The structure of 7 contains two uranium atoms separated by 3.7917(5) Å, a distance intermediate between the U⋅⋅⋅U distances in tetravalent 5 (3.606(6) Å) and trivalent 6 (3.8530(7) and 3.8651(7) Å).22 The larger distance in 7 vs 5 is consistent with the more extensive bridging structure that includes the tuck-over unit. The presence of the two "tuck" units in the C11–C15 ring led to significant variation of the UC bond lengths to that ring, although the ring planarity is not affected and the CC distances are equivalent within the error limits. The ring carbon atoms attached to the methylene groups, C11 and C15, have the shortest U1Cring distances (2.422(6) and 2.436(7) Å), for C12 and C14 these distances are 2.709(6) and 2.783(8) Å, and the ring carbon most distant from the methylene groups has a U1C13 bond length of 2.935(7) Å. The U2C16 tuck-over linkage (2.640(1) Å) is longer than typical UCalkyl single bonds (for example, 2.414(7) and 2.424(7) Å for UCMe in [(C5Me5)2UMe2]29). This is consistent with the long LnC bonds of the CH2 tuck-over groups observed in lanthanide complexes.14–19 The U1C16 distance (2.722(8) Å) is in the broad range of UC distances, and hence C16 may also transfer electron density to U1. A similar situation is seen in the lanthanide tuck-over complexes.14–19 The U1C20 distance, the first structurally characterized bond length between an f element and the carbon atom of a CH2 tuck-in group, is 2.564(1) Å. Hence, this bond is longer than a terminal alkyl bond, as expected for bridging ligands, but it is not as long as for a methylene bridge attached to a second metallocene. In comparison, the TiCH2 distance in 1 is very similar to a Ti3+Calkyl bond.7 The isolation of 7 raises several basic questions about the reactivity of UH groups. Although it is reasonable that a UH group in 6 can metalate a methyl group in C5Me5−, as has previously been observed with LnH bonds,14, 16 a sigma-bond metathesis with elimination of H2 would lead to a trivalent [(C5Me5)U(C5Me4CH2)] moiety from one of the {(C5Me5)2UH} units in dimeric 6. It is not clear how or why a second metalation would occur at the C5Me4CH22− ligand to form the observed tuck-in and tuck-over structure. It is possible that the hydride ligands in 7 are formed by reduction of H2 with U3+ ions in the same way that hydride ligands in 5 are formed from H2 and 6 in the reverse of the equilibrium in Equation (1). Hence, as the C5Me5− rings are being metalated, the H2 produced in this process may react with the U3+ centers before it leaves the metal coordination sphere. Clearly, the chemistry of uranium metallocene hydrides has some extra dimensions that have not yet been fully explored. New uranium hydrides are accessible that combine hydride and alkyl functionality. In addition, double CH activation is possible in this class of compounds to form tuck-in and tuck-over structures in a single complex. In any case, the long sought crystallographic evidence for the postulated f element tuck-in intermediates has been obtained and the existence of both tuck-in and tuck-over structures for actinides has been established. 7: In a glovebox, a green-brown solution of a 1:1 mixture of 5 and 6 (248 mg, 0.244 mmol) in toluene (10 mL) was heated to 110 °C for 3 minutes with frequent venting. The mixture was allowed to cool slowly to room temperature, and solvent was removed under reduced pressure to yield a green oil. A concentrated solution in toluene (3 mL) produced dark green crystals of 7 (188 mg, 0.185 mmol, 75 %) after 2 days at −35 °C. 1H NMR (C6D6, 298 K): δ=−23.9 (br s, 30 H, C5Me5, Δν1/2=600 Hz), −2.6 ppm (s, 15 H, C5Me5, Δν1/2=8 Hz), [−0.3 (s, 2 H), 1.5 (s, 3 H), 1.1 (s, 2 H), C5Me5(CH2)2]. 1H NMR (C6D6, 343 K): δ=−19.3 (s, 30 H, C5Me5, Δν1/2=70 Hz), −7.7 (s), −1.5 (s, 15 H, C5Me5, Δν1/2=11 Hz), 3.4 (s), 1.2(s), 0.4(s). 13C NMR (C6D6, 343 K): δ=−47.5 (C5Me5), −58.6 (C5Me5), 126.0 (C5Me5), 129.7 ppm (C5Me5). IR (KBr): =2966 (m), 2903 (vs), 2854 (vs), 2722 (w), 1434 (m), 1377 (m), 1164 (s), 1020 (s), 903 (m), 799 (m), 586 (m) cm−1. Elemental analysis calcd for C40H60U2: C 47.24, H 5.95, U 46.81; found: C 47.58, H 5.91, U 46.43. 7: A sealable Schlenk flask fitted with a Teflon stopper was charged with 6 (499 mg, 0.490 mmol) in toluene (20 mL). After four freeze-pump-thaw cycles, the solution was heated to 110 °C for 20 minutes. The reaction mixture was frozen in liquid nitrogen and evacuated using a Toepler pump equipped with a U-trap cooled in liquid nitrogen. The non-condensable gas was collected (9.8 μmol, 0.93 equiv) and subsequently analyzed by 1H NMR spectroscopy in C6D6 to be H2 (single resonance at 4.46 ppm). The solvent from the reaction mixture was evaporated to dryness yielding 7 as a dark brown crystalline material (0.469 g, 94 %). 8: In a glovebox, a solution of PhOH (56 mg, 0.59 mmol) in toluene (3 mL) was added to a stirred solution of dark green 7 (152 mg, 0.150 mmol) in toluene (5 mL). The solution immediately turned dark orange. After the mixture was stirred overnight, the solution was evaporated to dryness to yield 8 as a dark orange crystalline powder (192 mg, 92 %). Crystals of 8 suitable for X-ray diffraction studies were grown at −35 °C from a concentrated toluene solution. 1H NMR (500 MHz, C6D6): δ=3.19 (s, 30 H, C5Me5, Δν1/2=5 Hz), 3.9 (t, 2 H, 3JH,H=8 Hz, p-H), 1.8 (t, 4 H, 3JH,H=8 Hz, m-H), −13.6 ppm (d, 4 H, 3JH,H=8 Hz, p-H). 13C NMR (125 MHz, C6D6): δ=−32.8 (C5Me5), 141.1 (C5Me5), 125.1 (m-Ph), 104.1 (o-Ph), 108.1 ppm (p-Ph). IR (KBr): =2972 (m), 2907 (m), 2857 (m), 1588 (vs), 1489 (vs), 1475 (vs), 1377 (w), 1252 (vs), 1276 (vs), 1160 (s), 1065 (m), 998 (m), 873 (vs), 863 (vs), 754 (vs), 691, 604 (s) cm−1. C,H analysis calcd for C32H40O2U: C 55.33, H 5.80; found: C 55.62, H 5.50. In a similar experiment, 7 (12 mg, 0.012 mmol) in C6D6 was added to a J-Young tube containing a frozen slurry of (Et3NH)BPh4 (10 mg, 0.024 mmol) in C6D6. The J-Young tube was capped immediately and a color change from brown-green to brown was observed. 1H NMR spectroscopy showed quantitative conversion of starting material to the previously characterized [(C5Me5)2U]BPh425 and H2, which exhibited a 1H NMR resonance at 4.46 ppm. Compound 7 crystallizes in the space group P with a=10.5198(17), b=11.0156(17), c=16.281(3) Å, α=89.529(3), β=81.943(3), γ=80.842(3)°, V=1844.0(5) Å3, Z=2, ρcalcd=1.828 Mg m−3, R1=0.0436 [I > 2σ(I)], wR2=0.1114, GOF=1.047. Compound 8 crystallizes in the space group P with a=9.409(2), b=9.667(2), c=17.020(4) Å, α=99.753(4), β=96.714(4), γ=108.070(4)°, V=1426.6(6) Å3, Z=1, ρcalcd=1.617 Mg m−3, R1=0.078 [I > 2σ(I)], wR2=0.184. Detailed facts of importance to specialist readers are published as "Supporting Information". Such documents are peer-reviewed, but not copy-edited or typeset. They are made available as submitted by the authors. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.
Referência(s)