Synthesis of Some" Cobaloxime" Derivatives: A Demonstration of

Kimberly Freed, Sean Basquill, Angela Mendel, and William J. Shoemaker. Department of Chemistry, Gettysburg College, Gettysburg, PA 17325. Cobaloximes...
2 downloads 0 Views 134KB Size
In the Laboratory

Synthesis of Some Cobaloxime Derivatives: A Demonstration of “Umpolung” in the Reactivity of an Organometallic Complex Donald L. Jameson,* Joseph J. Grzybowski,* Deb E. Hammels, Ronald K. Castellano, Molly E. Hoke, Kimberly Freed, Sean Basquill, Angela Mendel, and William J. Shoemaker Department of Chemistry, Gettysburg College, Gettysburg, PA 17325

Cobaloximes have been prepared and studied extensively as model compounds for the coenzyme vitamin B12 (1, 2 ). These compounds, which exist at the interface of classical coordination chemistry, organometallic chemistry, and bioinorganic chemistry, offer the opportunity to expose students to these diverse areas. As part of an advanced laboratory course in organic and inorganic synthesis, we have developed a sequence of experiments (Scheme I) that demonstrate the scope of reactivity of this class of compounds.

O

H2 N

O NH 2

O

O

CH3

H2 N

NH2

N

N

Co

CH3 N

N

H2 N

O N

O

O

N

NH 2

O NH O

OP O OH O

CoBr2

+

2 H

O

HO

dimethylglyoxime (dmgH2)

N

Br O HO N N Co N N O O H H Br

1

Co(dmgH)(dmgH2)Br2

4-t-BuPy Br

1. NaBH4 2. CH2CH2I

O HO N N Co N N O H O N

N O N

methylcobaloxime vitamin B12 model complex

O

methylcobalamin vitamin B12 derivative

Structural comparison of a vitamin B12 derivative and a cobaloxime model complex.

[O]

CH2CH3 O HO N N Co N N O H O N

H

O

N

N N

H O N

O

Co

H

3. PhMgBr

O HO N N Co N N O H O N

3

2

4

Co(dmgH)2(4-t-BuPy)Et

Co(dmgH)2(4-t-BuPy)Br

Co(dmgH)2(4-t-BuPy)Ph

Scheme I: synthesis of cobaloxime derivatives

The metal site of vitamin B12 consists of a cobalt atom coordinated by a tetrapyrrole macrocyclic ligand called a corrin (see structures below). One of the axial sites is occupied by a 5,6-dimethylbenzimidazole nucleotide. The key feature of the vitamin’s biochemical activity is its ability to form metal–alkyl (Co–C) bonds in the second axial site. The N4 macrocycle in the cobaloxime model complexes is made up of two monodeprotonated dioxime molecules linked at two points by hydrogen bonding. The result is an essentially planar macrocyclic ligand. The most common cobaloximes utilize the dimethylglyoxime ligand. The structure shows a typical cobaloxime model complex. A very large number of model complexes have been prepared in which both the neutral ligand and the alkyl ligand are varied (1, 2).

Alkyl or aryl derivatives of cobaloximes are routinely prepared by two general routes. A carbanionic reagent reacting with a Co(III) halide results in a nucleophilic displacement of the halide ion (eq 1). In the second method, Co(III) is converted to Co(I) using a variety of reducing agents (eq 2). In acidic or neutral solution, a cobalt hydride species exists, while under sufficiently basic conditions, the Co(I) species has significant anionic character and behaves as an exceedingly powerful nucleophile. Reaction of the Co(I) species with a wide variety of electrophiles generates an organocobaloxime (eq 3). In the organocobaloxime compounds, the cobalt formally exists in a +3 oxidation state, with the carbon ligand being assigned a charge of ᎑1. Br H O O N N Co N N O O H N

Br H O O N N Co N N O O H N

H O O N N Co N N O O H N

+ PhMgBr

NaBH4

+ CH3I

H O O N N Co N N O O H N

+ MgBr2

(1)

H O O N N Co N N O O H N

CH3 H O O N N Co N N O O H N

(2)

+

I-

(3)

*Corresponding authors.

JChemEd.chem.wisc.edu • Vol. 75 No. 4 April 1998 • Journal of Chemical Education

447

In the Laboratory

The reversal of reactivity displayed by this organometallic system has parallels in classical organic chemistry. Alkyl halides (carbon in a higher oxidation state) behave as electrophiles and readily undergo nucleophilic displacement of the halide atom (eq 4). This sense of reactivity can be reversed by reducing the alkyl halide (for instance by conversion to a Grignard reagent using elemental magnesium; eq 5). These reagents (where carbon is now in a lower oxidation state) are very nucleophilic and can react with a variety of electrophiles (eq 6). Seebach has coined the term “umpolung” to describe this reversal of reactivity (3). RX + R′ ᎑ → R–R′

(4)

RX + M0 → RMX

(5)

RMX + R′–X′ → R–R′ + MXX′

(6)

Syntheses of cobaloxime derivatives have been reported in two volumes of Inorganic Synthesis (4, 5) and there are at least five published laboratory experiments based on the preparation of cobaloximes (6–10). The original Inorganic Synthesis preparation of cobaloxime derivatives (4) suffers drawbacks that preclude its conversion to a useful laboratory exercise. Several of the derivatives are of such low solubility that characterization by NMR is difficult. Subsequent work by Marzilli has shown (11) that the procedure for preparation of Co(dmgH)2(Py)Cl leads to large amounts of a salt, [Co(dmgH)2(Py)2][Co(dmgH)2Cl2], in addition to the desired complex. In that same paper, Marzilli describes a more reliable, two-step procedure for preparing Co(dmgH)2(RPy)Cl. Both Marzilli (11) and Bulkowski et al. ( 5), in the second Inorganic Synthesis procedure, have used 4-tert-butylpyridine as the axial base, thus increasing solubility and facilitating characterization by NMR. The final refinement of the original Inorganic Synthesis procedure is to prepare organocobaloximes via alkylation of a Co(I) intermediate prepared by reduction of the Co(III) precursor. Although Bulkowski et al. (5 ) state that this procedure gives more reproducible yields and higher quality product, it is still common practice to prepare organocobaloximes in situ from mixtures of cobalt(II) salt, dioxime ligand, sodium borohydride, and electrophile (12). Three of the published experiments describe the in situ preparation of organocobaloximes without adding NaBH4 (7–9). In this case, the Co(II) intermediate disproportionates to the reactive Co(I) and Co(III). This method, while very rapid and technically straightforward, is seldom encountered in the research literature because it suffers the obvious disadvantage of a maximum 50% yield based on cobalt and ligand. The procedure described here incorporates all of the aforementioned improvements. The preparation of organocobaloximes by the reaction of a Co(III) complex with a Grignard reagent is essentially the procedure of Schrauzer (4). This experiment accomplishes an additional objective for our advanced lab course. There are few examples of the preparation of compounds that involve a very air-sensitive intermediate, but yield a final product that is air-stable enough to characterize by routine methods. The Co(I) intermediate generated in the course of preparing the ver y stable Co(dmgH)2(4-t-BuPy)Et is extremely air-sensitive and failure to maintain strictly anaerobic conditions results in greatly diminished yields and very impure product. Careful planning allows for the preparation of all compounds in two four-hour lab periods. Compounds 1 and 2 448

(Scheme I) may be made in a single day, and compounds 3 and 4 (Scheme I) may be prepared during the second lab period. We have utilized commercially available anhydrous solvents and Grignard reagents in order to facilitate timely completion of the syntheses. These solvents and reagents may be transferred by cannula or syringe. In addition to avoiding the sometimes capricious preparation of phenylmagnesium bromide, the use of the commercial Grignard reagent, whose approximate concentration is known, allows for the more consistent synthesis of Co(dmgH)2(4-t-BuPy)Ph. In this reaction, three equivalents of Grignard reagent are required for conversion to the organocobaloxime; the first two remove the acidic bridging hydrogens of the macrocycle. Unlike classical Grignard reactions, where a large excess of reagent can often be tolerated, the reaction of excess PhMgBr with cobaloximes leads to diminished yields of the organocobalt complex. The addition of more than three equivalents of PhMgBr to compound 2 is accompanied by a green color change, suggesting possible reduction of the cobalt to the +2 or +1 oxidation state. This color change allows for the reaction to be performed as a crude titration with essentially identical results. The low-spin d6 compounds are readily characterized by NMR spectroscopy. Peak assignments are made by analogy to compounds reported in the literature. In the case of the imine carbon and the pyridine α -carbon, which both appear between 148 and 152 ppm in the 13C spectra, assignments are confirmed by the use of DEPT (Distortionless Enhanced Polarization Transfer) 135. Noteworthy aspects of the NMR are the broad peak above 15 ppm for the bridging O–H of the macrocyclic ligand (observed in two of the three derivatives) and the absence of a Co–C signal in the two organocobalt complexes, due to coupling of the carbon nuclear spin with the cobalt (I = 7/2) nuclear spin. Finally, these experiments can serve as the starting point for a wide variety of more independent projects. In addition to the dioxime macrocycle, other planar ligands that support organocobalt species are Schiff bases (13) and mixed Schiff base oximes (14 ). Other commonly used neutral axial bases are phosphines (15 ) and imidazoles (16 ). Electrophiles capable of reacting with the Co(I) intermediate are acyl halides (17 ), vinyl halides (12a), activated aryl halides (18), and epoxides (19). Experimental Procedure All reagents and solvents were used as supplied. Phenylmagnesium bromide (a 3.0 M solution in diethyl ether) and anhydrous tetrahydrofuran were purchased from Aldrich in Sure-Seal bottles and were transferred either by cannula or syringe. 1H and 13C NMR were obtained at 80 and 20 MHz, respectively, using an IBM NR-80 spectrometer. The spectra were obtained as CDCl3 solutions using TMS as the internal standard. NMR data and peak assignments are collected in Table 1.

Synthesis of Dibromo(dimethylglyoxime)dimethylglyoximato)cobalt(III) (1) Dimethylglyoxime (2.20 g, 19.0 mmol) was added to a stirred solution of 3.00 g of cobalt(II) bromide hydrate (9.17 mmol) in 50 mL of acetone. As a very gentle stream of air was played over the solution a green solid was deposited. After 30 min, the solution was chilled on ice, filtered, and washed with 2 × 15 mL of cold acetone. The yield of green

Journal of Chemical Education • Vol. 75 No. 4 April 1998 • JChemEd.chem.wisc.edu

In the Laboratory Table 1a. 1H NMR Spectroscopic Data for Cobaloxime Derivatives, Co(dmgH)2(4-t-BuPy)X Oxime Me

X Br Ph Et

Pyridine CMe3

Pyridine α

Pyridine β

O–H– –O

Ph

Ethyl CH2

Ethyl CH3







6.8–7.0 (m, 3H) 7.2–7.5 (m, 2H)





2.40

1.24

8.07

7.18

ca. 17.5

(s, 12H)

(s, 9H)

(d, 2H, J = 7 Hz)

(d, 2H, J = 7 Hz)

(v br s, 2H)

not obsd

2.03

1.29

8.58

7.30

(s, 12H)

(s, 9H)

(d, 2H, J = 7 Hz)

(d, 2H, J = 7 Hz)

2.13

1.29

8.44

7.27

18.1

(s, 12H)

(s, 9H)

(d, 2H, J = 7 Hz)

(d, 2H, J = 7 Hz)

(br s, 2H)

Table 1b.

13C



1.70

0.34

(q, 2H, J = 8 Hz)

(t, 3H, J = 8 Hz)

NMR Spectroscopic Data for Cobaloxime Derivatives, Co(dmgH)2(4-t-BuPy)X

X

Oxime Me

Oxime C=N

Pyridine CMe3

Pyridine CMe3

Pyridine α

Pyridine β

Pyridine γ

Ph

CH2CH3

Br

13.1

152.8

30.2

35.0

150.0

122.9

163.8





Ph

12.1

150.2

30.3

34.9

149.7

122.5

162.2

123.7, 126.6, 135.0 (Co–C not obsd)

Et

11.9

148.8

30.3

34.9

149.6

122.2

161.7



microcrystalline product was 3.5–3.9 g (85–95 % yield based on CoBr2). The compound has very low solubility and was therefore not characterized spectroscopically.

Synthesis of Bromo(4-tert-Butylpyridine)cobaloxime (2) A quantity of 3.5 g (7.78 mmol) of the green product (1) from the first step was suspended in 80 mL of methanol and 2.20 g (2.42 mL, 16.3 mmol) of 4-tert-butylpyridine was added. The mixture was stirred until the green solid had disappeared and was replaced by a brown crystalline solid (20–30 min). Water (125 mL) was added with stirring and the suspension was cooled in ice for 10 min. The product was collected by suction filtration and washed with 3 × 15 mL of 2:1 water/methanol and 2 × 10 mL of diethyl ether. The yield of the brown microcrystalline solid was about 3.5 g (89%). Synthesis of Ethyl(4-tert-butylpyridine)cobaloxime (3) Three or four KOH pellets (0.3 –0.4 g) were dissolved in 40 mL of methanol in a 200-mL two-neck flask and 1.50 g (3.00 mmol) of bromo(4-tert-butylpyridine)cobaloxime (2) was added. The homogeneous mixture was chilled in an ice/ salt bath (to about ᎑10 °C) and was thoroughly deaerated with nitrogen bubbling (≈10 min). Sodium borohydride (0.15 g, 3.9 mmol) was added (carefully) against a positive nitrogen flow and the mixture rapidly turned dark blue-green. The solution was stirred for 5 min and 0.50 mL (0.98 g, 6.3 mmol) of ethyl iodide was added via graduated pipet (against a positive nitrogen flow). The mixture immediately turned red-brown and was stirred in the ice bath for 20 to 30 min. The reaction was quenched by the addition of acetone (5 mL) and water (60 mL). The volume was reduced to 50 mL on a rotary evaporator and an orange crystalline solid was deposited. The flask was chilled in an ice bath and the orange solid was filtered and washed with water. The yields for this step were 1.0–1.2 g (73–88%). The compound was sufficiently pure as judged by NMR, but it may be recrystallized by dissolving in methanol, adding an equal volume of water, and reducing to half volume on a rotary evaporator. Synthesis of Phenyl(4-tert-butylpyridine)cobaloxime (4) A dry 200-mL round-bottom flask was charged with 1.50 g of bromo(4-tert-butylpyridine)cobaloxime (3.00

– 15.8 (Co–C not obsd)

mmol) and an egg-shaped stirring bar and the flask was capped with a septum punctured with a syringe needle as a pressure vent. Anhydrous THF (30 mL) was added via syringe and the flask was chilled in ice. A 3.2 mL portion of phenylmagnesium bromide (3.0 M solution; 9.6 mmol) was added dropwise via syringe over a 10-min period. The reaction became homogeneous and underwent a series of color changes as the Grignard reagent was added. When the addition was complete, the mixture was green. The mixture was stirred until the green color was replaced by a yellow-tan suspension (about 30–40 min). The septum was removed, the mixture was again cooled in ice, and any excess Grignard quenched by the careful addition of 1 mL of water. A 100-mL portion of 1 M HCl was added, the organic solvents were removed on a rotary evaporator, and the yellow solid collected by suction filtration. The solid was washed with water, methanol, and diethyl ether. The yields for this step were about 1.2 g (76%). If the concentration of the PhMgBr is not known or if less than strictly anhydrous THF is used, the reaction may be conducted as a crude colorimetric titration. Excess Grignard reagent acts as a scavenger for small amounts of water. The first onset of green as the PhMgBr is added indicates that the reaction is complete. Yields using this method are in the 50–70% range. Literature Cited 1. Toscano, P. J.; Marzilli, L. G. Prog. Inorg. Chem. 1984, 31, 105– 204 and references therein. 2. Bresciani-Pahor, N.; Forcolin, M.; Marzilli, L. G.; Randaccio, L.; Summers, M. F.; Toscano, P. J. Coord. Chem. Rev. 1985, 63, 1– 125 and references therein. 3. Seebach, D. Angew. Chem. Int. Ed. Eng. 1979, 18, 239–258. 4. Schrauzer, G. N. Inorg. Synth. 1968, 11, 61–70. 5. Bulkowski, J.; Cutler, A.; Dolphin, D.; Silverman, R. B. Inorg. Synth. 1980, 20, 127–134. 6. Clifford, B.; Cullen, W. R. J. Chem. Educ. 1983, 60, 554–555. 7. Brown, T. M.; Cooksey, C. J. Educ. Chem. 1987, 24, 77–80. 8. Brown, T. M.; Dronsfeld, A. T.; Cooksey, C. J.; Crich, D. J. Chem. Educ. 1990, 67, 434–435. 9. Brown, T. M.; Dronsfeld, A. T.; Cooksey, C. J.; Crich, D. J. Chem. Educ. 1990, 67, 973–974. 10. Jones, C. J. In Inorganic Experiments; Woolins, J. D., Ed.; VCH: New York, 1994; Chapter 3.7, pp 83–93.

JChemEd.chem.wisc.edu • Vol. 75 No. 4 April 1998 • Journal of Chemical Education

449

In the Laboratory 11. Trogler, W. C.; Stewart, R. C.; Epps, L. A.; Marzilli, L. G. Inorg. Chem. 1974, 13, 1564–1570. 12. For example: (a) Stang, P. J.; Datta, A. K.; Dixit, V.; Wistrand, L. G. Organometallics 1989, 8, 1020–1023. (b) Wright, M. W.; Smalley, T. L., Jr.; Welker, M. E.; Rheingold, A. L. J. Am. Chem. Soc. 1994, 116, 6777–6791. 13. Summers, M. F.; Marzilli, L. G.; Bresciani-Pahor, N.; Randaccio, L. J. Am. Chem. Soc. 1984, 106, 4478–4485. 14. Parker, W. O., Jr.; Bresciani-Pahor, N.; Zangrando, E.; Randaccio, L.; Marzilli, L. G. Inorg. Chem. 1985, 24, 3908–3913.

450

15. Trogler, W. C.; Marzilli, L. G. Inorg. Chem. 1975, 14, 2942– 2948. 16. Charland, J.-P.; Attia, W. M.; Randaccio, L.; Marzilli, L. G. Organometallics 1990, 9, 1367–1375. 17. Schrauzer, G. N.; Windgassen, R. J. J. Am. Chem. Soc. 1967, 89, 1999–2007. 18. Brown, K. L.; Awtrey, A. W.; LeGates, R. J. Am. Chem. Soc. 1978, 100, 823–828. 19. Brown, K. L.; Ingraham, L. L. J. Am. Chem. Soc. 1974, 96, 7681–7686.

Journal of Chemical Education • Vol. 75 No. 4 April 1998 • JChemEd.chem.wisc.edu