The Synthesis, Structure and Lewis Acidity of Bidentate Organotin Alkanes and Carboxylates
Claude H. Yoder1*, Jennifer E. Mihalick1, Wendy J. Kowalski1, Julie L. Ealy1, James N. Spencer1, Charles D. Schaeffer, Jr.2, Jennifer L. Green2, Kelly J. Sullivan2, Carolyn S. Yoder3 and Lisa C. Prokop3
1 Department of Chemistry, Franklin & Marshall College, Lancaster, PA 17604-3303, USA
2 Department of Chemistry, Elizabethtown College, Elizabethtown, PA 17022-2298, USA
3 Department of Chemistry, Millersville University, Millersville, PA 17551, USA
Received 29 June 1994; in revised form 19 October 1994
Abstract
A series of bis(diphenylbromostannyl)alkanes of the type (C6H5)2BrSn(CH2)nSnBr(C6H5)2, where n = 6, 10 and 12, and a series of bis(tributylstannyl)carboxylates of the type (n-C4H9)3SnO2C(CH2)nCO2Sn(n-C4H9)3, where n = 2, 6, 10, 12 and 14, were prepared. Tin-119 solid state NMR of the carboxylates indicated that the compounds contain five-coordinate, structurally-equivalent tins in the solid state. Adduct formation with triethylphosphine oxide (TEPO) for both the alkanes and carboxylates was monitored by phosphorus-31 NMR. Equilibrium constants for the alkanes were approximately independent of chain length from n = 6 to 12, while for the carboxylates, the constants for n = 2 and n = 14 were small. Equilibrium constants for the intermediate chains were approximately the same. Solid state NMR shows that the 1:1 TEPO adduct of the n = 12 carboxylate contains two different tin atoms, both of which are five-coordinate, and that the adduct is probably not symmetrically chelated.
Introduction
Multidentate Lewis acids have been subjected to increasing attention during the past several years1. Tin is an ideal atom to incorporate into organic molecules as a Lewis acidic site: most tin precursors are relatively cheap, synthetic methods are plentiful, and only one attached electronegative atom is necessary to confer considerable Lewis acidity. In an attempt to understand better the behavior of these potentially bidentate Lewis acids, we have prepared two series of tin compounds-- halides and carboxylates--and have determined the equilibria present in their reaction with triethylphosphine oxide, the equilibrium constants for 1:1 adduct formation, and the structure of the adducts formed.
Experimental
Grignard reagents were prepared under an argon atmosphere in oven-dried glassware. Other preparations were performed in dry glassware, but without an argon atmosphere.
The preparation of the bis(diphenylbromostannyl)alkanes followed the procedure of Newcomb et al2. Although the bis(triphenylstannyl)alkanes were separated by chromatography on alumina using toluene as eluent, reverse phase chromatography on Bakerbond Octadecyl (elution with acetonitrile), as suggested by Newcomb et al2., was necessary for the separation of the products of the HBr cleavage reaction from byproducts and starting materials. The spectral characteristics were identical to those reported by Newcomb et al2.
Bis(tributylstannyl) carboxylates were prepared by reaction of the appropriate dicarboxylic acid with tributyltin oxide in toluene. Water produced was removed by azeotropic distillation (Dean-Stark trap). Removal of toluene under vacuum led to precipitation of the compounds in good purity. Analytical data (Schwarzkopf Microanalytical Laboratories, Woodside, NY), melting points, yields, and carbon-13 NMR chemical shifts are reported in Tables 1 and 2.
Table 1. Melting points, yields, and analytical data
_________________________________________________________________________
Compound; n Mp, °C Yield, % Analytical Data, %a
_________________________________________________________________________
(n-C4H9)3SnO2C(CH2)nCO2Sn(n-C4H9)3
2 91-94 63 48.4/48.3, 8.6/8.4
6 118-120 72 51.1/51.1, 8.8/8.8
10 84-85 58 53.9/53.5, 8.9/9.2
12 85-86 55 54.6/54.6, 9.7/9.4
14 87 54 55.4/55.6, 9.6/9.6
_________________________________________________________________________
a Carbon, hydrogen, found/calc.
Table 2. Carbon-13 NMR chemical shiftsa
________________________________________________________________________
Compound; n d(13C)
________________________________________________________________________
(C6H5)2BrSn(CH2)nSnBr(C6H5)2
6 17.4, 25.6, 32.7, 128.9, 130.0, 135.9, 138.4
10 17.9, 26.0, 29.0, 29.5, 33.6, 129.0, 130.1, 135.9, 138.6
12 17.7, 25.8, 28.9, 29.3, 29.4, 33.5, 128.8, 129.4, 135.7, 138.5
(n-C4H9)3SnO2C(CH2)nCO2Sn(n-C4H9)3
2 13.7, 16.4, 27.1, 27.9, 30.9, 180
6 13.7, 16.4, 25.8, 27.0, 27.9, 29.1, 34.9, 180
10 13.7, 16.4, 25.8, 27.1, 27.8, 29.4, 29.4, 29.5, 34.9, 180
12 13.7, 16.4, 26.0, 27.1, 27.9, 29.4, 29.5, 29.6, 29.7, 35.0, 180
14 13.7, 16.4, 26.0, 27.1, 27.9, 29.4, 29.5, 29.6, 29.7, 29.8, 35.0, 180
________________________________________________________________________
a Relative to internal TMS in ppm.
Equilibrium constants were determined from phosphorus-31 chemical shifts measured as a function of concentration using a schematic mapping program as described previously3. All compounds were vacuum dried and solvents were dried over molecular sieves before NMR analyses. All NMR spectra were obtained on a Varian UNITY 300 MHz instrument with an operating frequency of 75.43 MHz for carbon-13, 121.42 MHz for phosphorus-31, and 111.85 MHz for tin-119. Gated decoupling with no nuclear Overhauser effect was used for all nuclei. Phosphorus-31 spectra were obtained with a coaxial inner tube containing trimethylphosphate as a reference and acetone-d6 while the coaxial tube for tin-119 spectra contained tetraethyltin as a reference and acetone-d6. Phosphorus-31 spectra were examined at temperatures down to -40 °C and displayed no significant linewidth variation. Tin-phosphorus coupling was not detected. Solid-state CP/MAS tin-119 NMR spectra were obtained with a Doty 5 mm high speed probe. Three or four spin speeds from 1.5 to 4.5 kHz were used to establish the center band(s). Tetraphenyltin was used to establish the Hartman-Hahn conditions (2 ms contact time) and triphenyltin chloride was used as a replacement reference.
Results and Discussion
Several slight modifications of the Newcomb et al2. procedure for the synthesis of the bis(diphenylbromostannyl)alkanes were explored. For the cleavage of the phenyl group, hydrogen bromide gas was bubbled continuously into a solution of the bis(triphenylstannyl)alkane over a period of about 18 h. This procedure did not produce significant quantities of product. Likewise, concentrations of hydrogen bromide in methylene chloride of less than 0.15 M were less effective than concentrations of 0.2-0.3 M. In our hands, the most effective reaction conditions appeared to be slow addition of the hydrogen bromide solution over a 2 h period to a 0.2 M solution of the alkane in methylene chloride at -78 °C, followed by ten days at room temperature. The bromo derivatives are slightly sensitive to atmospheric moisture and require a minimum exposure to air.
Of the two procedures utilized for the synthesis of the bis(tributylstannyl)-carboxylates, the reaction of the tributyltin oxide with a dicarboxylic acid produced product in greater purity than the reaction of tributyltin chloride in methylene chloride with the sodium salt of the acid in water4.
In order to determine the structure of the carboxylates in solution the tin-119 NMR of both the solution state and the solid state were obtained. These shifts, shown in Table 3, reveal a significant difference between solution and solid, with the solid state shifts being to lower frequency. This trend is consistent with a change in coordination number from four to five. The solid state shifts are also in the area previously observed for five-coordinate carboxylates5. It is interesting that the magnitude of the shift depends very little on the number of carbons in the bridge and that there is only one type of tin in the lattice. Presumably, in the solid state the molecules are intermolecularly coordinated through a carbonyl-tin bond. In solution, the magnitude of the shift and difference from the solid state indicates that the compounds are monomeric and contain four-coordinate tins. This is substantiated by the concentration independence of the tin-119 chemical shift (although this would also be observed if there were intramolecular coordination).
The interaction of these potentially bidentate Lewis acids with a Lewis base is of interest because of the possible structures that could be formed. The most likely are a 1:1 adduct with base attached to just one tin site and a 1:2 (acid to base) adduct with a base molecule attached to each of the tin sites. Less likely is a 1:1 adduct with the base attached to both tin sites. The equilibria involved in the reaction of triethylphosphine oxide (TEPO) and the alkane and carboxylate series were determined by monitoring the phosphorus-31 NMR chemical shift of the solution as a function of concentration. Only one resonance is observed in solution as a result of the lability of the adduct. If the initial concentrations of the acid and base are equal and only a 1:1 adduct is formed, then a plot of shift vs (shift/conc) will be linear and can be used to obtain the equilibrium constant3. For most of the compounds, five to eight solutions of different concentrations were used to establish these plots; in several cases where the lines were particularly curved, schematic mapping with 40-60 data points was used to determine the equilibria present3. Even in cases with considerable curvature, formation of the 1:1 adduct predominates. For 1,6-bis(bromo-diphenylstannyl)hexane an analysis with 56 points produced K (1:1) = 190 ± 15, K (1:2) = 670 ± 800. The F-test at the 95% confidence level showed that addition of the second equilibrium is statistically valid, but the magnitude of the equilibrium constant for the addition of a second mole of base to the 1:1 adduct is so small (3.5) that the majority of the reaction mixture exists in the form of the 1:1 adduct regardless of concentrations.
The data in Table 3 show that: (a) the alkanes and carboxylates of the same chain length have similar acidities toward TEPO (except perhaps for n = 6); (b) the alkanes have acidities similar to triphenyltin chloride (K = 2 x 102 in methylene chloride) while the bis carboxylates appear to be considerably more acidic than analogs such as tributyltin acetate (K = 13 with TEPO5), (c) for the carboxylates the equilibrium constants for 1:1 adduct formation increase with increasing chain length and then appear to level off or even decline; and (d) for the alkanes the equilibrium constants appear to be virtually independent of chain length from n = 6 to 12. If the effect of the solvent is taken into account
Table 3. Tin-119 NMR chemical shifts and equilibrium constants for 1:1 adduct formation with TEPO in CH2Cl2a
________________________________________________________________________
Compound; n d(119Sn), solution d(119Sn), solid K(1:1)
________________________________________________________________________
(C6H5)2BrSn(CH2)nSnBr(C6H5)2
6 2 x 102
10 3 x 102 b
12 2 x 102 b
(n-C4H9)3SnO2C(CH2)nCO2Sn(n-C4H9)3
2 104.6 -61.4 c
6 100.4 -58.6 9 x 10
10 100.9 3 x 102
12 101.5 -54.1 3 x 102
14 -56.1 c
________________________________________________________________________
a Solutions for tin-119 spectra ca. 0.1 M; estimated error in K ± 15%.
b K determined in toluene.
c Change in phosphorus-31 chemical shift as function of concentration is less than 0.4 ppm.
K could not be determined, but is likely to be small.
(for the TEPO adduct of (CH3)3SnCl, the constant in toluene is about 1.5 times greater than in methylene chloride6) the constants may even decrease slightly as the chain length increases.
The structure of the 1:1 adducts--whether mono- or bidentate-- is central to the analysis of the equilibrium constant data. The tin-119 solid-state NMR of the TEPO adduct of the bis carboxylate with n = 12 showed peaks of approximately the same intensity at -58.8 and -66.3 ppm. The presence of two peaks eliminates a symmetrical chelated structure (but does not eliminate an unsymmetrical chelated structure, although the similarity in shifts suggests that the coordination at the two tins is very similar). We favor a non-chelated structure that contains one tin coordinated to TEPO and the other to the carbonyl of a neighboring molecule. Thus, in solution, we suspect that the 1:1 adducts are all simple nonchelated monomeric species.
The structural variation in equilibrium constants can be rationalized by the effect of the second Lewis site on the acidity of the first. It would appear that the electronic effect of a diphenylbromostannyl group or a tributylcarboxyl group on the acidity of a given tin site should be greater with shorter bridging chain lengths. However, an examination of space-filling models indicates that the end groups can actually approach more closely as the chain length increases due to greater conformational mobility. When the chain is very small (n = 2) the group is large enough and close enough to produce considerable steric hindrance at the acidic site. Thus, with very small chains, it appears that steric hindrance may inhibit adduct formation, but with larger chains the electrostatic field effect (probably primarily a charge induced-dipole effect in the adduct) stabilizes the longer chain adducts more than the medium chain adducts. As the chain length increases beyond a certain point, the additional conformational mobility does nothing to decrease the distance between the end sites and moreover the decrease in entropy resulting from a significant amount of end-to-end interactions results in a decrease in equilibrium constant.
The greater difference in acidity of the bis(tributylstannyl)carboxylates (relative to simple tributylstannyl carboxylates, such as the acetate) compared to the bis(bromodiphenylstannyl)alkanes (relative to triphenyltin chloride) is probably a result of the greater dipole moment of the carboxylate site.
Acknowledgements
The authors are indebted to the Hackman Scholar program, the Merck & Company Inc. Foundation's Undergraduate Science Research Program, the Elizabethtown College Faculty Grants Committee, the Millersville University Grants Program, the Camille and Henry Dreyfus Foundation Scholar/Fellow Program for Undergraduate Insitutions, the National Science Foundation, and the Petroleum Research Fund, administered by the American Chemical Society, for support of this work. We are grateful to the Council on Undergraduate Research and Pfizer Inc. for an AIURP summer fellowship (J.L.G.). We thank the National Science Foundation, Armstrong World Industries, and Franklin & Marshall College for assistance in acquiring a Varian UNITY 300-MHz NMR spectrometer and facilities. The authors are also indebted to Beth Buckwalter for NMR support.
References
1. M.T. Blanda, J.H. Horner, and M. Newcomb, J. Org. Chem., 54 (1989) 4626;
H.E. Katz, J. Org. Chem., 54 (1989) 2179; M.E. Jung and H. Xia, Tetrahedron Lett., 29 (1988) 297; T.J. Karol, J.P. Hutchinson, J.R. Hyde, H.G. Kuivila, and J.A. Zubieta, Organometallics, 2 (1983) 106 ; M. Newcomb, J.H. Horner, and M.T. Blanda, J. Am. Chem. Soc., 109 (1987) 7878 ; J.D. Wuest and B. Zacharie, J. Am. Chem. Soc., 109 (1987) 4714; M. Gielen and K. Jurkschat, J. Organomet. Chem., 273 (1984) 303; M. Gielen, K. Jurkschat, J. Meunier-Piret, and M. Van Meerssche, Bull. Soc. Chim. Belg., 93 (1984) 379. For recent work on multidentate Lewis acids, see: F. Nadeau, M. Simard, and J.D. Wuest, Organometallics, 9 (1990) 1311.
2. Y. Azuma and M. Newcomb, Organometallics, 3 (1984) 9.
3. C.H. Yoder, D. Mokrynka, S.M. Coley, J.C. Otter, R.E. Haines, A. Grushow, L.J. Ansel, J.W. Hovick, J. Mikus, M.A. Shermak, and J.N. Spencer, Organometallics, 6 (1987) 1679; J.N. Spencer, B.G. Enders, A. Grushow, S.P. Kneizys, W.L. Nachlis, D. Mokrynka, S.M. Coley, J.C. Otter, and C.H. Yoder, J. Organomet. Chem., 362 (1989) 53.
4. M. Frankel, D. Gertner, D. Wagner, and A. Zikha, J. Organomet. Chem., 9 (1967) 83.
5. C.H. Yoder, R.A. Morreall, C.I. Butoi, W.J. Kowalski, and J.N. Spencer, J. Organomet. Chem., 448 (1993) 59.
6. J.N. Spencer, T.F. Ganunis, A.I. Zafar, C.M. Salata, S. Gupta, S. Puppala, H.J. Eppley, J.L. Ealy, and C.H. Yoder, J. Phys. Chem., 95 (1991) 4910.