Synthesis, Structure, Spectroscopy, and Reactivity of Oxapentadienyl

Organometallics 2010, 29, 5057–5067 DOI: 10.1021/om100361a

5057

Synthesis, Structure, Spectroscopy, and Reactivity of Oxapentadienyl-Cobalt-Phosphine Complexes†,1 John R. Bleeke,* Bryn L. Lutes, Michael Lipschutz, Donastas Sakellariou-Thompson, and John Seonghyun Lee Department of Chemistry, Washington University, St. Louis, Missouri 63130

Nigam P. Rath Department of Chemistry and Biochemistry, University of Missouri;St. Louis, One University Boulevard, St. Louis, Missouri 63121 Received April 28, 2010

The first examples of oxapentadienyl-cobalt complexes have been synthesized and structurally characterized. Treatment of (Cl)Co(PMe3)3 with potassium oxapentadienide produces ((1,2,3-η)-5oxapentadienyl)Co(PMe3)3 (1), while the reaction of (Cl)Co(PMe3)3 with potassium 2,4-dimethyloxapentadienide generates ((1,2,3-η)-2,4-dimethyl-5-oxapentadienyl)Co(PMe3)3 (2). Both 1 and 2 undergo ligand substitution reactions when treated with excess carbon monoxide. Compound 1 reacts with 1 equiv of CO to produce ((1,2,3-η)-5-oxapentadienyl)Co(PMe3)2(CO) (3), while 2 undergoes a double CO substitution to generate ((1,2,3-η)-2,4-dimethyl-5-oxapentadienyl)Co(PMe3)(CO)2 (4). Compound 3 can also be synthesized by reacting (Cl)Co(PMe3)2(CO)2 with potassium oxapentadienide, but the analogous reaction involving potassium 2,4-dimethyloxapentadienide results in reduction of the cobalt starting material and production of the Co(0) dimer (CO)(PMe3)2Co(μCO)2Co(PMe3)2(CO) (5). Treatment of 1 with triflic acid (HO3SCF3) or methyl triflate (MeO3SCF3) results in electrophilic attack at oxygen and production of (η4-butadienol)Co(PMe3)3þO3SCF3- (6) or (η4-butadienyl methyl ether)Co(PMe3)3þO3SCF3- (7). Treatment of 2 with triflic acid or methyl triflate also results in electrophilic attack at oxygen; however, these products are unstable and rapidly lose their protonated or methylated ligands. The resulting Co(PMe3)4þO3SCF3-, generated in situ, reacts with excess carbon monoxide to produce Co(PMe3)3(CO)2þO3SCF3- (8). Compounds 1-8 have been characterized by single-crystal X-ray diffraction.

Introduction Transition-metal complexes containing heteropentadienyl ligands (i.e., pentadienyl analogues in which one terminal CH2 group has been replaced by a heteroatom)2 have attracted attention because of their potential to exhibit enhanced reactivity or even to serve as catalysts by shuttling between a variety of accessible η5, η3, and η1 bonding

modes.3 Our group has focused on electron-rich heteropentadienyl-M-phosphine complexes, including oxapentadienyl,4 thiapentadienyl,5 azapentadienyl,6 phosphapentadienyl,7 and most recently silapentadienyl systems.8 The choice of electron-rich complexes is based on the expectation that these systems will more readily undergo the desired η5 f η3 and η3fη1 ligand shifts. While most of our previous work has involved second- and third-row transition metals, primarily rhodium and iridium, we have begun to investigate the heteropentadienyl chemistry of cobalt, an attractive catalytic metal due to its low cost. We have recently published the results of our initial foray into thiapentadienyl-cobalt chemistry9 and, in this paper,

† Part of the Dietmar Seyferth Festschrift. Dedicated to Professor Dietmar Seyferth, the peerless Editor-in-Chief of Organometallics since its inception in 1982. Many thanks for your kindness and encouragement over the years! (1) Pentadienyl-Metal-Phosphine Chemistry. 37. For Part 36, see: Bleeke, J. R.; Lutes, B. L.; Rath, N. P. Organometallics 2009, 28, 4577– 4583. (2) In principle, the heteroatom could be located anywhere along the five-atom chain. However, for the vast majority of heteropentadienylmetal complexes, the heteroatom resides in a terminal position, and our work has focused exclusively on “terminal” heteropentadienyl ligands. (3) For recent reviews of heteropentadienyl-metal chemistry, see: (a) Bleeke, J. R. Organometallics 2005, 24, 5190–5207. (b) Paz-Sandoval, M. A.; Rangel-Salas, I. I. Coord. Chem. Rev. 2006, 250, 1071–1106. (4) (a) Bleeke, J. R.; Haile, T.; New, P. R.; Chiang, M. Y. Organometallics 1993, 12, 517–528. (b) Bleeke, J. R.; Donnay, E.; Rath, N. P. Organometallics 2002, 21, 4099–4112.

(5) (a) Bleeke, J. R.; Ortwerth, M. F.; Rohde, A. M. Organometallics 1995, 14, 2813–2826. (b) Bleeke, J. R.; Shokeen, M.; Wise, E. S.; Rath, N. P. Organometallics 2006, 25, 2486–2500. (6) Bleeke, J. R.; Luaders, S. T.; Robinson, K. D. Organometallics 1994, 13, 1592–1600. (7) Bleeke, J. R.; Rohde, A. M.; Robinson, K. D. Organometallics 1995, 14, 1674–1680. (8) Bleeke, J. R.; Thananatthanachon, T.; Rath, N. P. Organometallics 2008, 27, 2436–2446. (9) Bleeke, J. R.; Lutes, B. L.; Rath, N. P. Organometallics 2009, 28, 4577–4583.

r 2010 American Chemical Society

Published on Web 07/09/2010

pubs.acs.org/Organometallics

5058

Organometallics, Vol. 29, No. 21, 2010

Bleeke et al.

Scheme 1

describe an analogous investigation involving electron-rich oxapentadienyl-cobalt complexes.

Results and Discussion A. Synthesis of ((1,2,3-η)-5-oxapentadienyl)Co(PMe3)3 (1). As shown in Scheme 1, the reaction of (Cl)Co(PMe3)310 with potassium oxapentadienide11 in tetrahydrofuran (THF) solvent leads to the formation of red-orange ((1,2,3-η)-5-oxapentadienyl)Co(PMe3)3 (1) in 94% yield. The 1H NMR spectrum of 1 shows a downfield peak at δ 6.53 due to H4 on uncoordinated C4. The protons on the coordinated allyl portion of the ligand all appear more upfield at δ 4.54 (H2), 3.41 (H3), 0.77 (H1), and 0.12 (H1). Similarly in the 13C{1H} NMR, C4 resonates at δ 170.2, while the other carbons all appear much further upfield at δ 69.0 (C3), 67.0 (C2), and ∼30.0 (C1). The 31P{1H} NMR spectrum at room temperature consists of a single peak at δ 2.8, indicating that the oxapentadienyl ligand is rotating rapidly on the allyl-Co axis, exchanging the positions of the three PMe3 ligands. When the temperature is lowered to -70 °C in acetone, the 31P{1H} NMR signal broadens but does not fully split into separate peaks. The infrared spectrum shows an intense, sharp peak at 1574.0 cm-1, assignable to the oxapentadienyl CdO stretch. The X-ray crystal structure of 1 is presented in Figure 1; selected bond distances are reported in the caption. The oxapentadienyl ligand is cis (anti) with respect to bond C2-C3 and is sickle-shaped. This ligand geometry is reflected in the observed C2-C3-C4-O1 torsion angle of 179.6(3)°, which closely approximates the value of 180° expected for an idealized sickle shape. The formyl group is rotated out of the allyl plane, resulting in a C1-C2-C3-C4 torsion angle of 20.2(4)°. This rotation causes C4 to lie 0.396 A˚ out of the plane. The cis geometry of the oxapentadienyl ligand may be stabilized by the contribution of a second resonance structure (II, Chart 1), featuring an η4-oxapentadienyl ligand. Consistent with this bonding picture is the relatively short bond length of C3-C4 (1.389(5) A˚). B. Synthesis of ((1,2,3-η)-2,4-dimethyl-5-oxapentadienyl)Co(PMe3)3 (2). Treatment of (Cl)Co(PMe3)310 with potassium 2,4-dimethyloxapentadienide4a in THF leads to the production of dark red ((1,2,3-η)-2,4-dimethyl-5-oxapentadienyl)Co(PMe3)3 (2) in 89% yield (Scheme1). In the 1H (10) Klein, H.-F.; Karsch, H. H. Inorg. Chem. 1975, 14, 473–477. (11) Heiszwolf, G. J.; Kloosterziel, H. Recl. Trav. Chim. Pays-Bas 1967, 86, 807.

Figure 1. Molecular structure of 1, using thermal ellipsoids at the 40% probability level. PMe3 methyl H’s are not shown. Selected bond distances (A˚): Co1-P1, 2.1873(8); Co1-P2, 2.1744(8); Co1-P3, 2.1644(6); Co1-C1, 2.044(3); Co1-C2, 1.946(3); Co1-C3, 2.112(3); C1-C2, 1.368(6); C2-C3, 1.443(5); C3-C4, 1.389(5); C4-O1, 1.223(4). Chart 1

NMR spectrum of 2, the signals for the allylic H’s appear, as expected, in the upfield region at δ 2.93 (H1), 2.62 (H3), and 1.48 (H1). In the 13C{1H} NMR, the uncoordinated carbon C4 resonates at δ 204.0, substantially downfield from the corresponding position of C4 in compound 1 (δ 170.2). Meanwhile, the allylic carbon signals appear at δ 84.0 (C2), 56.7 (C3), and 39.0 (C1). As in 1, the 31P{1H} NMR signal at room temperature is a singlet, indicating a rapid exchange of the PMe3 environments via dimethyloxapentadienyl ligand rotation. However, unlike the case for 1, the signal splits into three separate well-defined peaks upon cooling to -70 °C in toluene-d8, indicating that ligand rotation has been arrested. NMR line shape analysis has established a free energy of activation (ΔGq) of 10.9 ( 0.4 kcal/mol for this fluxional process. In the IR, the CdO stretching frequency appears at 1606.6 cm-1, as compared to 1574.0 cm-1 for 1. The X-ray crystal structure of 2 is presented in Figure 2. The dimethyloxapentadienyl ligand is cis with respect to C2-C3 and is U-shaped. The torsion angle C2-C3-C4-O1 is 7.4(3)° (0° expected for an idealized U shape), while the torsion angle C1-C2-C3-C4 is 38.1(3)°, indicating a substantial rotation of the acetyl group out of the allyl plane. As a result of this rotation, atom C4 is displaced 0.713 A˚ out of the C1-C2-C3 plane, as compared to 0.396 A˚ for 1. This enhanced displacement, together with the relatively long C3-C4 bond distance (1.459(3) A˚), suggests that the η4diene resonance structure (II, Chart 2) may be less important in this compound. Spectroscopic evidence is also consistent with this view, particularly the downfield shifting of C4 in the 13C NMR and the higher energy of the CdO stretch in the IR.

Article


5059

Scheme 2

Figure 2. Molecular structure of 2, using thermal ellipsoids at the 50% probability level. PMe3 methyl H’s are not shown. Selected bond distances (A˚): Co1-P1, 2.1655(5); Co1-P2, 2.2123(5); Co1-P3, 2.1790(5); Co1-C1, 2.0521(18); Co1-C2, 1.9733(17); Co1-C3, 2.1088(18); C1-C2, 1.422(3); C2-C3, 1.431(2); C2-C5, 1.509(2); C3-C4, 1.459(3); C4-C6, 1.523(3); C4-O1, 1.234(2). Chart 2

C. Reactions of Compounds 1 and 2 with Carbon Monoxide. When carbon monoxide is bubbled through a solution of compound 1 in THF, the color changes from red to orange, and ((1,2,3-η)-5-oxapentadienyl)Co(PMe3)2(CO) (3) is produced in 71% yield. As shown in Scheme 2, this substitution reaction most likely proceeds via an associative mechanism through an η1-oxapentadienyl-Co intermediate (A).12 However, a dissociative mechanism involving initial PMe3 loss cannot be ruled out. The 1H and 13C{1H} NMR spectra of 3 are very similar to those of 1, with all of the signals shifted downfield slightly. The main difference is seen in the 31P{1H} NMR, where two distinct peaks are observed, even at room temperature. This indicates that the asymmetry (12) This mechanism is “associative” in the sense that CO adds before PMe3 is lost. However, the initial oxapentadienyl shift from η3 to η1, which leads to the reactive 16e intermediate, is formally a ligand dissociation.

Figure 3. Molecular structure of 3, using thermal ellipsoids at the 50% probability level. PMe3 methyl H’s are not shown. Selected bond distances (A˚): Co1-P1, 2.1750(3); Co1-P2, 2.2167(3); Co1-C5, 1.7387(12); Co1-C1, 2.0569(12); Co1-C2, 1.9734(12); Co1-C3, 2.1134(12); C1-C2, 1.425(2); C2-C3, 1.4183(19); C3-C4, 1.4385(18); C4-O1, 1.2541(18).

of the phosphine/carbonyl ligand set has shut down the rapid oxapentadienyl rotation process observed in 1. In the IR, the oxapentadienyl CdO stretch is observed at 1612.3 cm-1, while the carbon monoxide ligand CtO stretch appears at 1935.0 cm-1. The X-ray crystal structure of 3 is presented in Figure 3. As in 1, the oxapentadienyl ligand is cis about C2-C3 and is sickle-shaped. The torsion angle C2-C3-C4-O1 is 178.94(12)°, while the torsion angle C1-C2-C3-C4 is 21.66(18)°. This rotation of the formyl group results in atom C4 being displaced 0.441 A˚ out of the allyl plane. One interesting aspect of the structure is the orientation of the phosphine and CO ligands. One of the large PMe3 ligands (P2) resides in the site under the open “mouth” of the allyl moiety, while the other PMe3 ligand (P1) lies under the allyl “edge” C2-C3 and the carbon monoxide ligand resides under the allyl “edge” C1-C2. In this way, the oxapentadienyl’s formyl moiety is “tucked in” between the two bulky PMe3’s. In contrast, when carbon monoxide is bubbled through a solution of compound 2 in THF, a double ligand exchange occurs, leading to red-orange ((1,2,3-η)-2,4-dimethyl-5-oxapentadienyl)Co(PMe3)(CO)2 (4) in 84% yield (Scheme 3). Again, we favor an associative mechanism for this reaction with monocarbonyl compound B (Scheme 3), the analogue of 3, serving as a reaction intermediate. The NMR spectra of 4, reported in the Experimental Section, are very similar to those of 2 and fully consistent with the proposed structure. In the IR spectrum, the dimethyloxapentadienyl CdO stretch is observed at 1668.2 cm-1, while the two

5060


Bleeke et al. Scheme 3

Figure 4. Molecular structure of 4, using thermal ellipsoids at the 50% probability level. PMe3 methyl H’s are not shown. Selected bond distances (A˚): Co1-P1, 2.1673(10); Co1-C7, 1.784(4); Co1-C8, 1.766(3); Co1-C1, 2.120(3); Co1-C2, 2.039(3); Co1-C3, 2.087(3); C1-C2, 1.396(4); C2-C3, 1.430(5); C2-C5, 1.510(5); C3-C4, 1.468(4); C4-C6, 1.512(5); C4-O1, 1.228(4).

carbon monoxide ligands give rise to CtO stretches at 1992.5 and 1939.3 cm-1. The X-ray crystal structure of 4 is presented in Figure 4. As in compound 2, the dimethyloxapentadienyl ligand is cis about C2-C3 and is U-shaped. The key torsion angles C2-C3-C4-O1 and C1-C2-C3-C4 are 0.8(5) and 37.6(5)°, respectively. C4 is displaced 0.719 A˚ out of the allyl plane. The orientation of the ligands in 4 is exactly opposite that in 3; the carbon monoxide ligands reside under the “mouth” and the C2-C3 “edge” of the allyl moiety, while the PMe3 ligand sits under the C1-C2 “edge”. Figure 5 presents an overlay view of structures 3 and 4. In this view, the allyl moieties of 3 (C1-C2-C3) and 4 (C10 -C20 -C30 ) are overlaid, allowing a comparison of the positions of the carbon monoxide and phosphine ligands. In 3 the PMe3 ligands (P1 and P2) are situated so that the formyl moiety of the oxapentadienyl (C4-O1) can tuck in between them. Carbon monoxide (C5-O2) lies opposite the formyl moiety. In 4, on the other hand, the bulky acetyl moiety of the dimethyloxapentadienyl (C60 -C40 -O10 ) is flanked by small carbon monoxide ligands (C70 -O20 and C80 -O30 ), while the bulky PMe3 (P10 ) lies opposite. From the overlay view, it is clear that if a PMe3 ligand were situated under the allyl “mouth” in 4, it would encounter unfavorable steric contacts with the acetyl group. The formation of the double carbonyl substitution product 4 is probably driven by electronic as well as steric factors. Due to the inductive effect of the methyl groups, the dimethyloxapentadienyl ligand is a better electrondonating ligand than oxapentadienyl. Hence, the metal center

Figure 5. Overlay view of structures 3 and 4. In this view, the allyl moieties of the oxapentadienyl ligand in 3 (C1, C2, C3, C4, O1) and the dimethyloxapentadienyl ligand in 4 (C10 , C20 , C30 , C40 , C50 , C60 , O10 ) are overlaid, revealing the relative orientation of the remaining ligands. The phosphine and carbonyl ligands of 3 are labeled P1, P2, and C5O2, while those of 4 are labeled P10 , C70 O20 , and C80 O30 . Oxapentadienyl carbons C2 and C20 lie at the center of the picture and are unlabeled.

is more electron-rich and can better accommodate electronwithdrawing ligands such as CO’s. D. Alternative Synthetic Approach to Compounds 3 and 4. We were interested in whether carbonyl-containing compounds 3 and 4 could be synthesized from the alternative starting material, (Cl)Co(PMe3)2(CO)2.10 As shown in Scheme 4, treatment of this species with potassium oxapentadienide in THF does, in fact, produce 3 in 59% yield, along with 1 equiv of CO. However, the reaction of (Cl)Co(PMe3)2(CO)2 with potassium 2,4-dimethyloxapentadienide does not proceed as expected. As shown in Scheme 4, this reaction does not produce a dimethyloxapentadienyl-Co complex but instead leads to the production of the orange dimer (CO)(PMe3)2Co(μ2-CO)2Co(PMe3)2(CO) (5) in 43% yield. While the detailed mechanism of this reaction is not known, it seems likely that the dimethyloxapentadienide serves as a reducing agent, generating the 17e Co(0) radical A (Scheme 4), which then dimerizes to 5. The greater electron-donating ability of dimethyloxapentadienyl as compared to oxapentadienyl (vide supra) explains the divergent reactions in Scheme 4. Compound 5 has been previously observed,13 and the spectroscopic data reported in the literature (1H NMR and IR) match our spectra. We have confirmed the structure of 5 by X-ray diffraction, as shown in Figure 6. The molecule possesses an approximate 2-fold rotation axis which renders the ligands on Co1 chemically equivalent to those on Co2: i.e., P1 equivalent to P4, P2 equivalent to P3, and C1O1 (13) Klein, H.-F.; Karsch, H. H. Chem. Ber. 1975, 108, 944–955.

Article


5061

Scheme 4

Figure 6. Molecular structure of 5, using thermal ellipsoids at the 50% probability level. PMe3 methyl H’s are not shown. Selected bond distances (A˚): Co1-Co2, 2.5209(8); Co1-P1, 2.1800(13); Co1-P2, 2.2136(13); Co2-P3, 2.2199(13); Co2P4, 2.1742(13); Co1-C1, 1.761(5); Co1-C2, 1.920(4); Co1-C3, 1.897(4); Co2-C2, 1.914(4); Co2-C3, 1.916(4); Co2-C4, 1.753(5).

equivalent to C4O4. The two bridging CO’s lie on the same side of the M-M bond and are also equivalent to each other by virtue of the 2-fold rotation axis. The room-temperature 31 P{1H} NMR spectrum for 5 consists of a single peak, indicating a fluxional process that renders the two sets of phosphine ligands equivalent. Although the detailed mechanism of the fluxional process is not known, it may involve exchange of the bridging and terminal carbonyl ligands, as shown in Scheme 5. In the IR spectrum, the bridging and terminal CO stretches are observed at 1718.1 and 1918.1 cm-1, respectively. E. Reactions of Compounds 1 and 2 with Triflic Acid and Methyl Triflate. Oxapentadienyl-cobalt compounds 1 and 2 possess two potential sites for electrophilic addition, the oxapentadienyl oxygen atom and the cobalt(I) center. In order to probe the relative reactivities of these sites, we treated each of these compounds with triflic acid (HO3SCF3) and with methyl triflate (MeO3SCF3). As shown in Scheme 6, treatment of 1 with triflic acid in diethyl ether results in the immediate precipitation of (η4-butadienol)Co(PMe3)3þO3SCF3- (6) as a yellow-orange powder in 65% yield. The 1 H NMR spectrum of 6 shows a downfield peak at δ 7.18,

due to the OH group. Meanwhile, the peak due to H4 is shifted substantially upfield to δ 3.59 (from δ 6.53 in 1), indicating that C4 is now coordinated to cobalt. The remaining η4-diene protons resonate at δ 5.21 (H2), 5.05 (H3), 1.29 (H1), and -0.70 (H1). In the 13C{1H) NMR, coordinated C4 shifts upfield to δ 96.9 (from δ 170.2 in 1), while the remaining diene carbons appear at δ 80.4 (C2), 79.3 (C3), and 35.7 (C1). The 31P{1H} NMR spectrum of 6 at room temperature is a broad singlet, due to facile rotation of the dienyl ligand about the η4-diene-Co axis. This rotation exchanges the chemical environments of the three phosphine ligands. When the temperature is lowered to -70 °C, the rotational process is slowed down, and the singlet splits into two broad peaks of intensity 2:1. However, the expected limiting spectrum of three separate peaks is not observed down to -70 °C. In the IR spectrum, the O-H stretch appears at 3306.9 cm-1, while no significant peaks are observed in the carbonyl region (1500-2000 cm-1). The X-ray crystal structure of 6 has been obtained and is presented in Figure 7. The η4-butadienol ligand in 6 retains the sickle shape of the η3-oxapentadienyl ligand in precursor 1, as reflected in the value of 175.7(2)° observed for the torsion angle C2-C3-C4-O1. The torsion angle C1-C2C3-C4 is 1.9(3)°, and the four dienol carbons (C1-C4) are almost coplanar with a mean deviation of 0.0066 A˚. Atom O1 lies 0.11 A˚ out of this plane. The dienol carboncarbon bond lengths are essentially equivalent (see Figure 6 caption), while C4-O1 lengthens to 1.380(3) A˚ from its value of 1.223(4) A˚ in 1. The cobalt-carbon bonds exhibit substantial variation in length. Carbons C1-C3 are bound tightly, with Co-C distances in the range of 2.026(2)-2.066(2) A˚, while Co-C4 is significantly longer at 2.173(2) A˚. The hydrogen atom of the OH group was located and refined; it resides cis to the hydrogen atom on C4 (H4). Treatment of 1 with methyl triflate in diethyl ether also leads to electrophilic addition at the oxapentadienyl oxygen and immediate formation of an orange precipitate of (η4-butadienyl methyl ether)Co(PMe3)3þO3SCF3- (7) (Scheme 6) in 83% yield. The 1H NMR spectrum of 7 shows a sharp singlet at δ 3.55, assignable to the ether methyl group. In the 13C{1H} NMR, the ether methyl resonates at δ 60.6. The peaks due to the η4-butadiene moiety are very similar to those observed in 6. As with 6, the 31P{1H} NMR signal of 7 at room temperature is a broad singlet due to facile rotation of the butadienyl methyl ether ligand. When the temperature is

5062



Scheme 6

Figure 8. Molecular structure of the cation in 7, using thermal ellipsoids at the 50% level. Only one of the two crystallographically distinct molecules is described here. PMe3 methyl H’s are not shown. Selected bond distances (A˚): Co1-P1, 2.2070(3); Co1-P2, 2.2026(3); Co1-P3, 2.2307(3); Co1-C1, 2.0702(11); Co1-C2, 2.0188(11); Co1-C3, 2.0629(10); Co1-C4, 2.1693(11); C1-C2, 1.4231(16); C2-C3, 1.4008(16); 1.411(3); C3-C4, 1.4177(15), C4-O1, 1.3849(14), O1-C5, 1.4283(15).

Figure 7. Molecular structure of the cation in 6 using thermal ellipsoids at the 50% probability level. PMe3 methyl H’s are not shown. Selected bond distances (A˚): Co1-P1, 2.2108(7); Co1P2, 2.1898(6); Co1-P3, 2.2371(6); Co1-C1, 2.065(2); Co1-C2, 2.026(2); Co1-C3, 2.066(2); Co1-C4, 2.173(2); C1-C2, 1.423(3); C2-C3, 1.404(3); C3-C4, 1.408(3); C4-O1, 1.380(3).

lowered to -70 °C, this signal splits into two peaks with a 2:1 ratio but does not fully decoalesce to three separate peaks. The X-ray crystal structure of 7, presented in Figure 8, closely resembles that of 6. The butadienyl methyl ether ligand retains the sickle shape of precursor 1, as reflected in the torsion angle C2-C3-C4-O1 (177.30(9)°). The torsion angle C1-C2-C3-C4 is 0.87(15)°, and atoms C1C2-C3-C4 are essentially coplanar with a mean deviation of just 0.0030 A˚ from the best plane. Atom O1 lies 0.065 A˚ from this plane. Methyl carbon C5 resides cis to the hydrogen on C4 (H4) but is rotated out of the ligand plane toward cobalt. This results in a C3-C4-O1-C5 torsion angle of 165.60(10)° and a displacement of 0.27 A˚ for C5 out of the

C1-C2-C3-C4 plane. The carbon-carbon bond lengths within the diene moiety are all essentially equivalent, indicating full delocalization, but the cobalt-carbon bond distances show substantial variation. As in 6, carbons C1-C3 are tightly bound (2.019(1)-2.070(1) A˚), while C4 interacts more weakly (2.169(1) A˚). Treatment of ((1,2,3-η)-2,4-dimethyl-5-oxapentadienyl)Co(PMe3)3 (2) with triflic acid leads to a dramatically different result. Instead of the yellow-orange product observed in the analogous reaction involving 1 (vide supra), a deep green product is obtained. The green compound is paramagnetic and not amenable to NMR characterization. Its infrared spectrum shows no peaks in the CO or OH stretching regions, indicating that the dimethyloxapentadienyl ligand or its protonated derivative are no longer coordinated to the metal center. At first we speculated that the green compound might be the 16e tetrahedral species Co(PMe3)3(O3SCF3). However, this neutral compound would be expected to have some solubility in nonpolar solvents such as pentane and ether,14 while our compound is soluble only in polar solvents such as THF and methylene chloride. This suggested an ionic formulation, and the deep green color implicated Co(PMe3)4þO3SCF3-, a compound whose tetraphenylborate salt had been previously reported by Klein to be green.10 We reproduced Klein’s synthesis of Co(PMe3)4þBPh4- and found that, indeed, the colors of the two compounds matched, as did their UV-visible spectra. As a further test, we treated both salts with excess carbon monoxide and observed, in each case, a rapid color change from green to orange and production of Co(PMe3)3(CO)2þO3SCF3- (8) or its tetraphenylborate analogue. (14) For example, the neutral acetate analogue Co(PMe3)3(O2CCH3) has substantial pentane solubility.10

Article


5063

(Scheme 6). The scavenging of PMe3 by 14e “Co(PMe3)3þ” is also not surprising. In fact, Dartiguenave and Beauchamp16 have reported that treatment of (Br)Co(PMe3)3 with NaþBPh4- in the absence of added PMe3 produces Co(PMe3)4þBPh4-. When compound 2 is treated with methyl triflate, the dark green compound is again obtained. As shown in Scheme 7, the same mechanism is thought to be operating: i.e., methylation at oxygen, followed by loss of the methylated ligand and production of Co(PMe3)4þO3SCF3-. As above, treatment of the green solution with carbon monoxide produces 8 in high yield.

Summary

Figure 9. Molecular structure of the cation in 8 using thermal ellipsoids at the 50% level. PMe3 methyl H’s are not shown. Selected bond distances (A˚) and bond angles (deg): Co1-P1, 2.2489(3); Co1-P2, 2.1996(3); Co1-P3, 2.2029(3); Co1-C1, 1.7470(10); Co1-C2, 1.7519(9); P1-Co1-P2, 97.060(10); P1Co1-P3, 96.987(10); P2-Co1-P3, 165.945(10); P1-Co1-C1, 112.98(4); P1-Co1-C2, 117.32(3); P2-Co1-C1, 87.39(3); P2-Co1-C2, 86.27(3); P3-Co1-C1, 86.70(3); P3-Co1-C2, 87.71(3); C1-Co1-C2, 129.69 (5).

Although Co(PMe3)3(CO)2þBPh4- has been previously prepared by a different synthetic route,15 the structure of the cation had not been reported. We were able to grow X-rayquality crystals of Co(PMe3)3(CO)2þO3SCF3- (8) and obtain the structure, which is presented in Figure 9. The coordination geometry is trigonal bipyramidal, with two of the bulky PMe3 ligands (P2 and P3) occupying the axial sites. However, these ligands “lean” toward the smaller CO ligands (and away from the equatorial phosphine P1) so that the P2-Co1-P3 angle is 165.94(1)°, rather than the idealized value of 180°. The 1H NMR spectrum of 8 at room temperature exhibits a single doublet at δ 1.77 (JH-P = 8.7 Hz), indicating that the phosphines are exchanging. Similarly, the 31 P{1H} NMR consists of a single peak (a singlet) at δ 8.7. These signals do not split upon cooling to -70 °C in acetoned6. In the infrared, the carbonyl ligands give rise to peaks at 1988.5 and 1929.1 cm-1. As shown in Scheme 7, the formation of deep green C probably involves the initial protonation of the dimethyloxapentadienyl oxygen in 2 to produce a transient η4-butadienol-Co species (A, Scheme 7). Unlike its analogue 6, this species is unstable with respect to dissociation of the protonated ligand, leading to the formation of “Co(PMe3)3þO3SCF3-”, which then scavenges 1 equiv of PMe3 to produce Co(PMe3)4þO3SCF3- (C). The relative instability of A (as compared to 6) is not surprising when one considers the structures of precursors 1 and 2. Recall that, in 2, carbon atom C4 is displaced much further out of the allyl (C1-C2C3) plane than in 1. This displacement probably results from unfavorable steric interactions involving the acetyl moiety in 2 (C6-C4-O1) and the phosphine ligands. Hence, one would expect a much weaker interaction between Co and C4 in protonation product A (Scheme 7) than in the analogous 6 (15) Attali, S.; Poilblanc, R. Inorg. Chim. Acta 1971, 6, 475–479.

In this paper, we have reported the synthesis, structure, and spectroscopy of the first oxapentadienyl-cobalt complexes. These compounds show a strong preference for the (1,2,3-η)-5-oxapentadienyl bonding mode, both in the tris(PMe3) parent compounds and in their CO-substituted derivatives. The electrophilic reagents triflic acid and methyl triflate react exclusively at the oxapentadienyl oxygen, producing η4-butadienol-cobalt or η4-butadienyl methyl ethercobalt products. These products are stable when the starting oxapentadienyl is unmethylated but unstable when the starting oxapentadienyl ligand is 2,4-dimethylated. In the latter case, Co(PMe3)4þO3SCF3- is generated in situ and can be converted to Co(PMe3)3(CO)2þO3SCF3- upon treatment with CO. Current work in our laboratory is focused on using bulkier phosphine ligands to stabilize alternative (especially η5) oxapentadienyl bonding modes.

Experimental Section General Comments on Experimental Techniques. All manipulations were carried out under a nitrogen atmosphere, using either glovebox or double-manifold Schlenk techniques. Solvents were stored under nitrogen after being distilled from the appropriate drying agents. Deuterated NMR solvents were obtained in sealed vials and used as received. (Cl)Co(PMe3)3 and (Cl)Co(PMe3)2(CO)2 were prepared by the procedure of Klein and Karsch.10 Potassium oxapentadienide11 and potassium 2,4-dimethyloxapentadienide4a were prepared by literature procedures. Highpurity carbon monoxide gas was obtained from Praxair and used as received. Triflic acid (trifluoromethanesulfonic acid) and methyl triflate (methyl trifluoromethanesulfonate) were obtained from Aldrich and used as received. NMR experiments were performed on a Varian Unity Plus300 spectrometer (1H, 300 MHz; 13C 75 MHz; 31P, 121 MHz), a Varian Mercury-300 spectrometer (1H, 300 MHz; 13C, 75 MHz; 31 P, 121 MHz), a Varian Unity Plus-500 spectrometer (1H, 500 MHz; 13C, 125 MHz; 31P, 202 MHz), or a Varian Unity-600 spectrometer (1H, 600 MHz; 13C, 150 MHz; 31P, 242 MHz). 1H and 13C spectra were referenced to tetramethylsilane, while 31P spectra were referenced to external H3PO4. HMQC (1H-detected multiple quantum coherence), HMBC (heteronuclear multiple-bond correlation), and COSY (correlation spectroscopy) experiments aided in assigning some of the 1H and 13C peaks. Note: carbonyl ligands always gave rise to very weak 13C NMR signals and could not be reliably assigned. Synthesis of ((1,2,3-η)-5-oxapentadienyl)Co(PMe3)3 (1). Potassium oxapentadienide (0.65 g, 6.0 mmol) and (Cl)Co(PMe3)3 (1.00 g, 3.10 mmol) were weighed into a 125 mL Erlenmeyer (16) DeCarvalho, L. C. A.; Peres, Y.; Dartiguenave, M.; Dartiguenave, Y.; Beauchamp, A. L. Acta Crystallogr. 1989, C45, 159–161.

5064



flask, and 70 mL of tetrahydrofuran (THF) was added. The resulting solution was stirred at room temperature for 45 min. After removal of the solvent under vacuum, the product was extracted with 100 mL of diethyl ether and the extract filtered through Celite. The ether was then removed under vacuum. The resulting powder was dissolved in a minimal quantity of diethyl ether and the solution cooled to -30 °C, producing red-orange crystals of 1 overnight. Yield: 1.04 g (94%). Anal. Calcd for C13H32CoOP3: 43.82; H, 9.07. Found: C, 44.09; H, 8.81. 1H NMR (acetone-d6, 22 °C): δ 6.53 (d, JH-H = 8.4 Hz, 1, H4), 4.54 (br s, 1, H2), 3.41 (br s, 1, H3), 1.28 (s, 27, PMe3’s), 0.77 (br s, 1, H1), 0.12 (br s, 1, H1). 13C{1H} NMR (acetone-d6, 22 °C): δ 170.2 (s, C4), 69.0 (d, JC-P = 12.8 Hz, C3), 67.0 (d, JC-P = 9.6 Hz, C2), ∼30.0 (obscured by solvent, C1), 21.3 (complex m, PMe3’s). 31 P{1H} NMR (acetone-d6, 22 °C): δ 2.8 (br s, PMe3’s). Note: when the temperature was lowered to -70 °C, the 31P NMR signal broadened but did not split. IR (Nujol mull): 1574.0 cm-1 (CdO). Synthesis of ((1,2,3-η)-2,4-dimethyl-5-oxapentadienyl)Co(PMe3)3 (2). Potassium 2,4-dimethyloxapentadienide (0.40 g, 2.9 mmol) and (Cl)Co(PMe3)3 (0.47 g, 1.5 mmol) were weighed into a 125 mL Erlenmeyer flask, and 70 mL of THF was added. The resulting solution was stirred at room temperature for 45 min. After removal of the solvent under vacuum, the product was extracted with 100 mL of diethyl ether and the extract filtered through Celite. The ether was then removed under vacuum. The resulting powder was dissolved in a minimal quantity of diethyl ether and the solution cooled to -30 °C, producing dark red crystals of 2 overnight. Yield: 0.50 g (89%). Anal. Calcd for C15H36CoOP3: C, 46.87; H, 9.46. Found: C, 46.38; H, 9.34. 1H NMR (benzened6, 22 °C): δ 2.93 (br s, 1, H1), 2.62 (br s, 1, H3), 2.13 (br s, 3, H5’s), 1.98 (br s, 3, H6’s), 1.48 (br s, 1, H1), 1.01 (br s, 27, PMe3’s). Note: when the temperature was lowered to -70 °C in toluene-d8, the signal at δ 1.01 split into three peaks at δ 1.13 (br s, 9, PMe3), 0.76 (br s, 9, PMe3), 0.62 (br s, 9, PMe3). 13C{1H} (benzene-d6, 22 °C): δ 204.0 (s, C4), 84.0 (s, C2), 56.7 (s, C3), 39.0 (s, C1), 32.9 (s, C6), 27.5 (s, C5), 22.4 (br s, PMe3’s). 31P{1H} (benzene-d6, 22 °C): δ0.5 (br s, PMe3’s). Note: when the temperature was lowered to -70 °C in toluene-d8, the 31P NMR signal split into three peaks at δ 16.3 (br s, 1, PMe3), 2.3 (br s, 1, PMe3), -9.5 (br s, 1, PMe3). IR (Nujol mull): 1606.6 cm-1 (CdO). Synthesis of ((1,2,3-η)-5-oxapentadienyl)Co(PMe3)2(CO) (3). Compound 1 (0.39 g, 1.1 mmol) was dissolved in 30 mL of THF in a 50 mL Erlenmeyer flask, capped with a rubber septum. Carbon monoxide gas was then bubbled through the solution for 1 h, using a needle that penetrated the septum. During this exposure, the solution changed from red to red-orange. After

filtration through Celite, the solvent was removed under vacuum. The resulting powder was dissolved in a minimal quantity of diethyl ether and the solution cooled to -30 °C, causing orange crystals of 3 to form overnight. Yield: 0.24 g (71%). Anal. Calcd for C11H23CoOP2: C, 42.86; H, 7.54. Found: C, 42.39; H, 7.49. 1H NMR (acetone-d6, 22 °C): δ 7.10 (d, JH-H = 8.4 Hz, 1, H4), 5.09 (m, 1, H2), 3.78 (m, 1, H3), 1.88 (m, 1, H1), 1.44-1.40 (m, 18, PMe3’s), 1.17 (dd, J = 16.2 Hz, 9.6 Hz, 1, H1). 13 C{1H} NMR (acetone-d6, 22 °C): δ 177.0 (s, C4), 74.2 (s, C2), 71.2 (s, C3), 36.2 (s, C1), 19.8 (dd, JC-P = 22.6 Hz, 2.4 Hz, PMe3), 19.5 (dd, JC-P = 25.0 Hz, 4.0 Hz, PMe3). 31P{1H} NMR (acetone-d6, 22 °C): δ 9.7 (s, 1, PMe3), -3.7 (s, 1, PMe3). IR (Nujol mull): 1935.0 (CtO), 1612.3 cm-1 (CdO). Alternative Synthesis of ((1,2,3-η)-5-oxapentadienyl)Co(PMe3)2(CO) (3). Potassium oxapentadienide (0.072 g, 0.67 mmol) and (Cl)Co(PMe3)2(CO)2 (0.16 g, 0.53 mmol) were weighed into a 50 mL Erlenmeyer flask, and 20 mL of THF was added. The resulting solution was stirred at room temperature for 45 min. After removal of the solvent under vacuum, the product was extracted with pentane and the extract filtered through Celite. The pentane was then removed under vacuum. The resulting powder was dissolved in a minimal quantity of pentane and the solution cooled to -30 °C, producing orange crystals of 3 after several days. Yield: 0.097 g (59%). Synthesis of ((1,2,3-η)-2,4-dimethyl-5-oxapentadienyl)Co(PMe3)(CO)2 (4). Compound 2 (0.54 g, 1.4 mmol) was dissolved in 30 mL of THF in a 50 mL Erlenmeyer flask, capped with a rubber septum. Carbon monoxide gas was then bubbled through the solution for 1 h, using a needle that penetrated the septum. During this exposure, the solution changed from dark red to red-orange. After filtration through Celite, the solvent was removed under vacuum. The resulting powder was dissolved in a minimal quantity of diethyl ether and the solution cooled to -30 °C, causing red-orange crystals of 4 to form overnight. Yield: 0.34 g (84%). Anal. Calcd for C11H18CoO3P: C, 45.84; H, 6.31. Found: C, 45.81; H, 6.10. 1H NMR (benzene-d6, 25 °C): δ 3.77, 3.75 (overlapping s’s, 2, H1 and H3), 2.71 (d, J = 7.5 Hz, 1, H1), 2.06 (s, 3, H6’s), 1.70 (s, 3, H5’s), 0.76 (d, JH-P = 8.5 Hz, 9, PMe3). 13C{1H} NMR (benzene-d6, 25 °C): δ 200.5 (s, C4), 100.2 (s, C2), 63.1 (s, C3), 56.7 (s, C1), 29.5 (s, C6), 25.9 (s, C5), 19.5 (d, JC-P = 29.9 Hz, PMe3). 31P{1H} NMR (benzene-d6, 25 °C): δ 22.1 (s, PMe3). IR (Nujol mull): 1992.5 (CtO), 1939.3 (CtO), 1668.2 cm-1 (CdO). Reaction of Potassium 2,4-Dimethyloxapentadienide with (Cl)Co(PMe3)2(CO)2. Synthesis of (CO)(PMe3)2Co(μ2-CO)2Co(PMe3)2(CO) (5). Potassium 2,4-dimethyloxapentadienide (0.75 g, 5.5 mmol) and (Cl)Co(PMe3)2(CO)2 (1.66 g, 5.48 mmol) were weighed into a 50 mL Erlenmeyer flask, and 25 mL of THF

Article was added. The resulting solution was stirred at room temperature for 45 min. After removal of the solvent under vacuum, the product was extracted with pentane and the extract filtered through Celite. The pentane was then removed under vacuum. The resulting powder was dissolved in a minimal quantity of pentane and the solution cooled to -30 °C, producing orange crystals of 5 overnight. Yield: 0.63 g (43%). Anal. Calcd for C16H36Co2O4P4: C, 35.97; H, 6.81. Found: C, 35.99; H, 6.46. 1H NMR (acetone-d6, 22 °C): δ 1.30 (virtual t, JH-P = 6.9 Hz, PMe3’s). 13C{1H} NMR (acetone-d6, 22 °C): δ 18.7 (virtual t, JC-P = 25.6 Hz, PMe3’s). 31P{1H} NMR (acetone-d6, 22 °C): δ 0.7 (br s, PMe3’s). IR (Nujol mull): 1918.1 (CtO), 1718.1 cm-1 (μ2-CO). Synthesis of (η4-butadienol)Co(PMe3)3þO3SCF3- (6). Compound 1 (0.12 g, 0.35 mmol) was dissolved in 8 mL of diethyl ether and the solution cooled to -30 °C. The resulting red solution was treated with a 1.66 M solution of triflic acid in diethyl ether (0.18 mL, 0.30 mmol), causing a yellow-orange precipitate to form immediately. The precipitate was allowed to settle, and the supernatant was decanted. The remaining solid was washed twice with diethyl ether and twice with pentane. It was then dissolved in a minimal quantity of acetone and the solution cooled to -30 °C, causing yellow crystals of 6 to form overnight. Yield: 0.098 g (65%). Anal. Calcd for C14H33CoF3O4P3S: C, 33.20; H, 6.58. Found: C, 33.34; H, 6.46. 1H NMR (acetone-d6, 25 °C): δ7.18 (s, 1, OH), 5.21 (s, 1, H2), 5.05 (br s, 1, H3), 3.59 (s, 1, H4), 1.52 (br s, 27, PMe3’s), 1.29 (s, 1, H1), -0.70 (s, 1, H1). Note: at -70 °C, the PMe3 signal split into two peaks at δ 1.60 (br s, 9, PMe3) and 1.40 (br s, 18, PMe3’s). 13C{1H} NMR (acetone-d6, 25 °C): δ 96.9 (d, C4), 80.4 (d, JC-P = 26.6 Hz, C2), 79.3 (d, JC-P = 28.9 Hz, C3), 35.7 (s, C1), 19.9 (br s, PMe3’s). 31 P{1H} NMR (acetone-d6, 25 °C): δ 4.7 (br s, PMe3’s). Note: at -70 °C, the 31P NMR signal split into two peaks at δ 7.2 (br s, 2, PMe3’s) and 6.2 (s, 1, PMe3). IR (Nujol): 3306.9 cm-1 (O-H). There were no significant peaks in the carbonyl region (1500-2000 cm-1). Synthesis of (η4-butadienyl methyl ether)Co(PMe3)3þO3SCF3(7). Compound 1 (0.46 g, 1.3 mmol) was dissolved in 15 mL of diethyl ether and the solution cooled to -30 °C. The resulting red solution was treated with a 1.45 M solution of methyl triflate in diethyl ether (0.89 mL, 1.3 mmol), causing a yellow-orange precipitate to form immediately. The reaction mixture was stirred at room temperature for 5 min and then cooled to -30 °C for 10 min before collecting the precipitate by filtration. After it was washed with pentane, the precipitate was dissolved in acetone and the solution filtered. The acetone was removed under vacuum, and the resulting powder was dissolved in a minimal quantity of acetone. Addition of a small quantity of pentane, followed by cooling to -30 °C, caused orange crystals of 7 to form overnight. Yield: 0.56 g (83%). Anal. Calcd for C15H35CoF3O4P3S: C, 34.62; H, 6.79. Found: C, 34.70; H, 6.56. 1H NMR (acetone-d6, 22 °C): δ 5.20 (s, 1, H3 or H2), 5.10 (s, 1, H2 or H3), 3.55 (s, 3, Me), 2.90 (s, 1, H4), 1.79 (br s, 9, PMe3), 1.44 (br s, 18, PMe3’s), 1.32 (s, 1, H1), -0.56 (s, 1, H1). 13C{1H} NMR (acetone-d6, 22 °C): δ 101.0 (s, C4), 81.2 (d, JC-P = 30.2 Hz, C2 or C3), 79.6 (d, JC-P = 25.2 Hz, C3 or C2), 60.6 (d, JC-P = 18.3 Hz, Me), 36.5 (s, C1), 21.7 (br s, PMe3), 20.0 (br s, PMe3), 19.1 (br s, PMe3). 31P{1H} NMR (acetone-d6, 22 °C): δ 4.4 (br s, PMe3’s). Note: at -70 °C, the 31P NMR signal split into two peaks at δ 7.1 (br s, 1, PMe3) and 6.4 (br s, 2, PMe3’s). IR (Nujol): there were no significant peaks in the carbonyl region (1500-2000 cm-1). Reaction of Compound 2 with Triflic Acid. Synthesis of Co(PMe3)3(CO)2þO3SCF3- (8). Compound 2 (0.54 g, 1.4 mmol) was dissolved in 70 mL of THF and the solution cooled to -30 °C. Upon addition of a 1.66 M solution of triflic acid in diethyl ether (0.84 mL, 1.4 mmol), the solution changed immediately from red-orange to dark green. Carbon monoxide was then bubbled through the solution for 1 h, causing the color to change from green to orange. After filtration through Celite, the solvent was removed under vacuum. The resulting powder was dissolved in a


5065

minimal quantity of THF/ether and the solution cooled to -30 °C, causing yellow crystals of 8 to form overnight. Yield: 0.56 g (81%). Anal. Calcd for C12H27CoF3O5P3S: C, 29.28; H, 5.54. Found: C, 29.01; H, 5.32. 1H NMR (acetone-d6, 22 °C): δ 1.77 (d, JH-P = 8.7 Hz, PMe3’s). 13C{1H} NMR (acetone-d6, 22 °C): δ 19.9-20.6 (complex m, PMe3’s). 31P{1H} NMR (acetone-d6, 22 °C): δ 8.7 (s, PMe3’s). IR (Nujol): 1988.5 (CtO), 1929.1 cm-1 (CtO). Note: (1) when the temperature was lowered to -70 °C, the 1H and 31P NMR signals did not split; (2) reaction of 2 with methyl triflate, followed by treatment with CO, also produced 8 in high yield. Calculation of ΔGq for Compound 2 from Variable-Temperature (VT) NMR Data. Theoretical line shapes were calculated for a series of rates using the method of Johnson.17,18 The experimental VT 31P{1H} NMR spectra were then matched against the theoretical spectra, and in this way, exchange rate constants were determined for each temperature. These exchange rate constants, k, were then used to calculate the free energy of activation, ΔGq, at each temperature, T, using the Eyring equation.19 The reported ΔGq value is the average value over all the temperatures in the simulation, and the uncertainty is the estimated standard deviation. X-ray Diffraction Studies. Crystals of X-ray diffraction quality were obtained for compounds 1-8. In all cases crystals of appropriate dimensions were mounted on MiTeGen microloops20 in random orientations. Preliminary examination and data collection were performed using a Bruker Kappa Apex II charge coupled device (CCD) detector system single-crystal X-ray diffractometer equipped with an Oxford Cryostream LT device. All data were collected using graphite-monochromated Mo K R radiation (λ = 0.710 73 A˚) from a fine-focus sealed-tube X-ray source. Preliminary unit cell constants were determined with a set of 36 narrow frame scans. Typical data sets consisted of combinations of φ and φ scan frames with typical scan width of 0.5° and counting time of 15-30 s/frame at a crystal to detector distance of 4.0 cm. The collected frames were integrated using an orientation matrix determined from the narrow frame scans. Apex II and SAINT software packages21 were used for data collection and data integration. Analysis of the integrated data did not show any decay. Final cell constants were determined by global refinement of reflections from the complete data set. Collected data were corrected for systematic errors using SADABS or TWINABS.21 Crystal data and intensity data collection parameters are given in Table 1. Structure solution and refinement were carried out using the SHELXTL-PLUS software package.22 The structures were solved by direct methods andPrefined with full-matrix leastsquares refinement by minimizing w(Fo2 - Fc2).2 All non-hydrogen atoms were refined anisotropically to convergence. Specific experimental details for individual structure are given below. For compounds 1, 4, 5, 7, and 8, all H atoms were added in the calculated positions and were refined using appropriate riding models (AFIX m3). For 2, the H atoms on C1 (H1A, H1B) and C3 (H3) were located and refined using geometric constraints (SADI). All other H atoms were added in the calculated positions and were refined using appropriate riding models. For 3, the H’s on the oxapentadienyl ligand (H1A, H1B, H2, H3, and H4) were located and refined freely. All other H atoms were added in the calculated positions and were refined using appropriate riding models. For 6, the H atoms of the dienol (17) Johnson, C. S., Jr. Am. J. Phys. 1967, 35, 929–933. (18) Martin, M. L.; Martin, G. J.; Delpeuch, J.-J. Practical NMR Spectroscopy; Heydon: London, 1980; pp 303-309. (19) Lowry, T. H.; Richardson, K. S. Mechanism and Theory in Organic Chemistry; Harper and Row: New York, 1976; pp 99-101. (20) MiTeGen, LLC, PO Box 3867, Ithaca, NY 14852. (21) Apex II and SAINT; Bruker Analytical X-Ray, Madison, WI, 2008. (22) Sheldrick, G. M. SHELXTL-Bruker. Acta Crystallogr. 2008, A64, 112–122.

5066


Bleeke et al.

Table 1. X-ray Diffraction Structure Summary

formula fw cryst syst space group a, A˚ b, A˚ c, A˚ R, deg β, deg γ, deg V, A˚3 Z cryst dimens, mm calcd density, g/cm3 radiation: λ, A˚ temp, K θ range, deg data collected h k l total decay no. of data collected no. of unique data Mo KR linear abs coeff, mm-1 abs cor applied data to param ratio final R indices (obsd data)a R1 wR2 R indices (all data) R1 wR2 goodness of fit

formula fw cryst syst space group a, A˚ b, A˚ c, A˚ R, deg β, deg γ, deg V, A˚3 Z cryst dimens, mm calcd density, g/cm3 radiation: λ, A˚ temp, K θ range, deg data collected h k l total decay no. of data collected no. of unique data Mo KR linear abs coeff, mm-1 abs cor applied data to param ratio final R indices (obsd data)a R1 wR2 R indices (all data) R1 wR2 goodness of fit a I > 2σ(I).

1

2

3

4

C13H32CoOP3 356.23 orthorhombic Pna21 18.8012(14) 7.9575(6) 12.0680(10) 90 90 90 1805.5(2) 4 0.41 0.37 0.19 1.311 0.710 73 100(2) 2.75-30.78

C15H36CoOP3 384.28 monoclinic P21/n 8.7078(5) 16.0967(8) 14.2582(8) 90 93.930(2) 90 1993.83(19) 4 0.38 0.29 0.09 1.280 0.710 73 100(2) 2.66-27.54

C11H23CoO2P2 308.16 monoclinic P21/n 9.7576(7) 11.1403(8) 14.0125(10) 90 100.413(3) 90 1498.11(19) 4 0.47 0.33 0.26 1.366 0.710 73 100(2) 2.80-35.73

C11H18CoO3P 288.15 monoclinic P21/n 6.2947(18) 13.870(4) 16.203(5) 90 95.183(14) 90 1408.9(7) 4 0.20 0.12 0.11 1.359 0.710 73 100(2) 2.52-25.42

-26 to þ27 -11 to þ11 -17 to þ17 none obsd 53 265 5610 1.206 numerical 32.43




0.0350 0.0871

0.0301 0.0605

0.0304 0.0726

0.0395 0.0996

0.0413 0.0901 1.060

0.0461 0.0660 1.023

0.0409 0.0772 1.051

0.0568 0.1099 1.058

5

6

7

8

C16H36Co2O4P4 534.19 monoclinic P21/c 10.9720(9) 13.2483(11) 18.0190(15) 90 107.156(4) 90 2502.7(4) 4 0.34 0.18 0.17 1.418 0.710 73 140(2) 3.00-25.48

C14H33CoF3O4P3S 506.30 triclinic P1 8.4154(6) 10.1041(6) 13.1407(10) 90.567(4) 98.475(4) 92.082(4) 1104.29(13) 2 0.27 0.13 0.09 1.523 0.710 73 100(2) 1.57-28.97

C15H35CoF3O4P3S 520.33 monoclinic P21/c 16.5274(11) 15.3020(11) 18.7319(12) 90 90.855(3) 90 4736.8(6) 8 0.24 0.19 0.17 1.459 0.710 73 100(2) 1.23-36.48

C12H27CoF3O5P3S 492.24 monoclinic P21/c 14.8592(9) 8.3390(5) 18.0971(11) 90 107.585(3) 90 2137.6(2) 4 0.52 0.14 0.12 1.529 0.710 73 100(2) 2.36-37.30

-13 to þ13 -15 to þ16 -21 to þ21 none obsd 39 554 4598 1.598 semiempirical 18.62

-11 to þ11 -13 to þ13 0 to þ17 none obsd 56 729 5786 1.130 semiempirical 22.00

-27 to þ27 -25 to þ25 -31 to þ31 none obsd 214 380 22 719 1.056 semiempirical 44.81

-25 to þ25 -14 to þ14 -30 to þ30 none obsd 137 054 11 095 1.168 semiempirical 47.21

0.0476 0.1134

0.0307 0.0712

0.0321 0.0642

0.0257 0.0569

0.0772 0.1345 1.026

0.0389 0.0751 1.087

0.0576 0.0729 1.013

0.0381 0.0619 1.039

Article ligand (H1, H1A, H1B, H2, H3, and H4) were located and their positions were refined with thermal parameters riding on the parent non-H atoms. All other H’s were added in the calculated positions and were refined using appropriate riding models. Compound 6 was a twinned crystal, and data reduction was carried out using two domains. Refinements were performed using merged HKLF 5 data. Twin refinement resulted in a ratio of 53:47 for the two domains (BASF/HKLF 5 refinement in SHELXTL). Significant improvement in structure quality (R1, wR2, goodness of fit, and residual densities) was observed when using merged HKLF 5 data, as compared to HKLF 4 or unresolved (nontwin) data. For compound 7, there were two independent molecules in the unit cell. The structures of these molecules were nearly identical; therefore, only one is presented in Figure 8.


5067

Acknowledgment. The regional X-ray Facility at the University of Missouri;St. Louis was funded in part by the National Science Foundation’s MRI Program (No. CHE-0420497). D.S.-T. and J.S.L. were high school student participants in the STARS (Students and Teachers As Research Scientists) program. We thank the STARS program sponsors. Supporting Information Available: Tables, figures, and CIF files giving structure determination summaries, final atomic coordinates, thermal parameters, bond lengths, bond angles, and torsion angles for compounds 1-8. This material is available free of charge via the Internet at http:// pubs.acs.org.

Synthesis, Structure, Spectroscopy, and Reactivity of Oxapentadienyl

Recommend Documents