Skip to main content
All Chemistry

2.4 Experimental section

Sijbren Otto - Catalysis of Diels-Alder Reactions in Water

Materials

Cyclopentadiene (2.5) was prepared from its dimer (Merck-Schuchardt) immediately before use. Dimineralised water was distilled twice in a quartz distillation unit. Ethanol (Merck) was of the highest purity available. Acetonitrile (Janssen) was run over basic aluminium oxide prior to use. 2,2,2-Trifluoroethanol (Acros) was purified by distillation (bp 79°C). Co(NO3)2·6H2O, Ni(NO3)2·6H2O, Cu(NO3)2·3H2O, Zn(NO3)2·4H2O and KNO3 were of the highest purity available. Substituted 3-phenyl-1-(2-pyridyl)-2-propene-ones (2.4a-e) were prepared by an aldol condensation of the corresponding substituted benzaldehyde with 2-acetylpyridine, following either of two modified literature procedures64.

2.4a and 2.4b

To a stirred solution of 0.5 ml of 10% aqueous sodium hydroxide and 8.25 mmol of the appropriate aldehyde in 10 ml of ethanol, 8.25 mmol of 2-acetylpyridine was added dropwise during 2-3 hours. The temperature was kept at 0°C. After stirring for another 2 hours the reaction mixture was filtered, yielding almost pure solid 2.4a (7.26 mmol, 88%) or 2.4b (7.76 mmol, 94 %). After crystallisation from ethanol the melting points were recorded and the compounds were characterised by 1H-NMR and mass spectroscopy. 2.4a: mp 158.2-158.5°C (lit. 156°C82; 154-5°C83), 1H NMR (200 MHz, CDCl3) δ 7.55 (m,1H), 7.86 (d,2H), 7.91 (m,2H), 8.22 (m,1H), 8.27 (d,2H), 8.45 (d,1H), 8.77 (d,2H). Exact mass: calcd 254.069; found: 254.069. 2.4b: mp 102.2-102.5°C (lit. 104°C82; 91-92°C83), 1H NMR (200 MHz, CDCl3) δ 7.39 (d,2H), 7.50 (m,1H), 7.67 (d,2H), 7.88 (m,2H), 8.19 (m,1H), 8.29 (d,1H), 8.75 (m,1H).

2.4c, 2.4d and 2.4e

17 mmol of 2-acetylpyridine and 16.5 mmol of the appropriate benzaldehyde were introduced into 100 ml of water at 0-5 °C. The mixture was shaken thoroughly in order to obtain a finely dispersed emulsion. 10 ml of a 10% sodium hydroxide solution was added. The mixture was again shaken and left overnight undisturbed at 4°C. The solution should not be stirred since this results in a phase separation and lower yields due to transport limitations. The product separated as an oil that solidified upon shaking. Filtration and washing with water gave the almost pure product in satisfactory yields: 2.4c: 95%, 2.4d: 84%, 2.4e: 96%. After crystallisation from ethanol the melting points were recorded and the compounds were characterised by 1H-NMR. 2.4c: mp 74.5-75.3 (lit. 74 °C82; 71°C83), 1H NMR (200 MHz, CDCl3) δ 7.46 (m,4H), 7.74 (m,2H), 7.86 (m,1H), 7.95 (d,1H), 8.20 (m,1H), 8.32 (d,1H), 8.75 (m,1H). 2.4d: mp 84.8-85.3 (lit. 83°C82), 1H NMR (200 MHz, CDCl3) δ 2.40 (s,3H), 7.23 (d,2H), 7.49 (m,1H), 7.64 (d,2H), 7.87 (m,1H), 7.93 (d,1H), 8.19 (m,1H), 8.27 (d,1H), 8.74 (m,1H). 2.4e: mp 84.6-85.2 (lit. 84-85°C82; 84°C83), 1H NMR (200 MHz, CDCl3) δ 3.85 (s,3H), 6.93 (d,2H), 7.47 (m,1H), 7.69 (d,2H), 7.86 (m,1H), 7.92 (d,1H), 8.19 (d,1H), 8.19 (m,1H), 8.73 (m,1H).

2.4f and 2.4g

2.4f and 2.4g have been prepared as outlined in Scheme 2.5. Yields were not optimised. p-(Bromomethyl)benzaldehyde (2.8) has been prepared by reaction of p-(bromomethyl)benzonitrile (2.7, Acros) with diisobutylaluminium hydride following a literature procedure65.

Sodium (p-oxomethylphenyl)methylsulfonate (2.9)

A suspension of 3.90 g (19.6 mmol) of p-(bromomethyl)benzaldehyde (2.8) and 4.00 g (31.7 mmol) of sodium sulfite in 40 ml of water was refluxed for two hours, after which a clear solution was obtained. The reaction mixture was cooled on an ice bath resulting in precipitation of some sodium sulfite. After filtration, the solvent was evaporated. Ethanol was added to the remaining solid and the suspension was refluxed for 10 minutes. After filtering the hot solution, the filtrate was allowed to cool down slowly to −18 °C whereupon sodium (p-oxomethylphenyl)methylsulfonate (2.9) separated as colourless crystals. The extraction procedure was repeated two more times, affording 2.29 g (10.3 mmol, 53%) of the desired product. 1H-NMR (200 MHz, D2O) δ(ppm) = 4.10 (s,2H); 7.44 (d,2H); 7,76 or 7.76 (in the editors’ view) (d,2H); 9.75 (s,1H).

Sodium (4-(3-oxo-3-(2-pyridyl)-1-propenyl)phenyl)methylsulfonate (2.4f)

A solution of 75 ml of ethanol and 3.75 ml of a 10% solution of sodium hydroxide in water was cooled to 0 °C and 2.13 g (9.57 mmol) of sodium (p-oxomethylphenyl)methylsulfonate (2.9) and 1.28 g (10.6 mmol) of 2-acetylpyridine were added. The solution was stirred for 16 hours at 0°C and filtered affording 2.47 g (7.66 mmol, 80%) of crude 2.4f. Crystallisation from methanol yielded 1.08 g (35%) of the desired product, mp 220 °C (decomposition). 1H-NMR (200 MHz, D2O) δ(ppm) = 4.00 (s,2H); 7.36 (d,2H); 7.55 (m,1H); 7.58 (d,2H); 7.61 (m,2H); 7.81 (m,2H); 8.43 (d,1H). Anal. Calcd for C15H12NO4SNa: C, 55.4; H, 3.72; N, 4.31; S, 9.84; Na, 7.07. Found: C, 54.7; H, 3.74; N, 4.19; S, 9.41; Na, 6.97.

3-(p-(Bromomethyl)-phenyl)-1-(2-pyridyl)-2-propen-1-one (2.10)

A solution of 10 g of sodium hydroxide in 1.0 L of water was cooled to 0-5 °C. 5.95 g (49.2 mmol) of 2-acetylpyridine and a solution of 7.25 g (36.4 mmol) p-(bromomethyl)benzaldehyde (2.8) dissolved in a minimal amount of ethanol were added. The resulting suspension was stirred for 48 hours at 0-5 °C. The product was filtered and washed extensively with water until the smell of the 2-acetylpyridine had disappeared. After drying at 50 °C under vacuum, 10.5 g (34.8 mmol, 96%) of 3-(p-(bromomethyl)phenyl)-1-(2-pyridyl)-2-propen-1-one (2.10) was obtained. 1H NMR (200 MHz, CDCl3) δ(ppm) = 4.50 (s,2H); 7.42 (d,2H); 7.44 (m,1H); 7.70 (d,2H); 7.85 (m,1H); 7.91 (d,1H); 8.20 (d,1H); 8.31 (d,1H); 8.75 (d,1H).

(4-(3-Oxo-3-(2-pyridyl)-1-propenyl)phenyl)methyltrimethylammonium bromide (2.4g)

5.00 g (16.6 mmol) of 3-(p-(bromomethyl)phenyl)-1-(2-pyridyl)-2-propen-1-one (2.10) was suspended in 350 ml of dry ether under a nitrogen atmosphere. 20 ml of a 4.2 M solution of trimethylamine in ethanol (Fluka) was added. The reaction mixture was stirred for 24 hours at room temperature under a nitrogen atmosphere. Evaporation of the solvents and excess of trimethylamine afforded crude 2.4g in quantitative yield. The very hygroscopic product can be crystallized from anhydrous 1-propanol. Removal of this solvent from the crystals is not possible by conventional methods. However, dissolving the product in water, filtration and subsequent freeze drying afforded 3.10 g (8.59 mmol, 52%) of 2.4g, mp 212.5 °C (decomposition). 1H-NMR (200 MHz, CDCl3) δ(ppm) = 3.44 (s,9H); 5.16 (s,2H); 7.50 (m,1H); 7.75 (s,4H); 7.84 (d,1H); 7.85 (m,1H); 8.01 (d,1H); 8.31 (d,1H); 8,72 or 8.72 (in the editors’ view) (d,1H). Anal. Calcd for C18H21N2BrO: C, 60.0; H, 5.88; N, 7.78; Br, 21.92. Found: C, 59.8; H, 5.97; N, 7.58; Br, 21.77.

Kinetic measurements

All kinetic measurements were performed using UV-Vis spectroscopy (Perkin Elmer λ2, λ5 or λ12) monitoring the disappearance of the absorption of the dienophile at 25 ± 0.1 °C in the presence of a known excess of diene. Measurements were performed in cuvettes of 1 cm path length, initially containing only a solution of the dienophile. After equilibration, a 10-25 µl of a stock solution of cyclopentadiene in 1-propanol was added and the absorption was monitored. Two different methods were used to determine the second-order rate constants. The rates of the faster reactions (half-lives not more than a few hours) were determined by following the reaction during at least four half-lives. From the absorbance data, pseudo-first-order rate constants were obtained using a fitting-program. The rate constants of the slower reactions in organic solvents and the reactions with cyclopentadiene in water with half-lives of more than 15 minutes were determined using initial rate kinetics84. Using a known excess of cyclopentadiene, the following expression was used to calculate the second-order rate constants:

$$k_2 = \frac{d(A_{\mathbf{2.4}})}{dt} \cdot \frac{1}{d \cdot (\varepsilon_{\mathbf{2.4}} - \varepsilon_{\mathbf{2.6}}) \cdot [\mathbf{2.4}]_0 \cdot [\mathbf{2.5}]_0}$$

Herein d(A2.4)/dt is the slope of the plot of the absorption of the dienophile versus time during the first five percent of the reaction and d is the pathlength of the cuvette. The extinction coefficients of the dienophile and the product were determined separately under the same conditions as used in the kinetic runs. This method has been successfully tested by comparing the results with rate constants obtained by traditional pseudo-first-order kinetics. Typical concentrations were: [dienophile] = 1·10−5 M, [cyclopentadiene] = 1·10−3 M and [catalyst] = 1·10−2 M. All rate constants were measured at least three times. Those obtained by the traditional method were reproducible to within 3%, whereas the initial rate method gave a reproducibility of 5%.

Equilibrium constants

Measurements were performed employing a Perkin Elmer λ2, 5 or 12 UV-Vis spectrophotometer at 25 ± 0.1°C. Equilibrium constants were determined by measuring the extinction coefficient at a suitable wavelength of the partially complexed dienophile (εobs) as a function of the concentration of metal ion. The following expression can be derived85:

$$\frac{[\mathrm{M}^{n+}]}{\varepsilon_{\mathbf{2.4}} - \varepsilon_{obs}} = \frac{1}{(\varepsilon_{\mathbf{2.4}} - \varepsilon_{compl}) \cdot K_a} + \frac{[\mathrm{M}^{n+}]}{\varepsilon_{\mathbf{2.4}} - \varepsilon_{compl}}$$

After determining the extinction coefficient of the uncomplexed dienophile (ε2.4), [Mn+]/(ε2.4−εobs) was plotted versus [Mn+] yielding a straight line. The equilibrium constant now equals the ratio of the slope and the intercept of this line. Very accurate measurements of the extinction coefficients are a prerequisite for obtaining reliable equilibrium constants. Crucial in this respect were the choice of the wavelength and the choice of the appropriate metal-ion concentrations. The most accurate results were obtained at the wavelength of maximal difference between the extinction coefficients of uncomplexed and complexed dienophile. The metal-ion concentrations were chosen so as to cover the largest possible change in εobs with the smallest possible change in [M2+]. Solutions of different [M2+] with total ionic strength of 2.00 M were prepared. KNO3 was used as the background electrolyte. Extinction coefficients were determined by filling the cuvet with an accurately known volume of this solution and measuring the absorption after injection of 3-10 µl of a stock solution of the dienophile in 1-propanol. Typical concentration ranges were: [dienophile] = 6·10−6 - 2·10−5 M and [M2+] = 5·10−3 - 2·10−5 M. The reproducibility of the equilibrium constants varied between 5-10%.

Product analysis

Endo-exo product mixtures were isolated using the following procedure. A solution of cyclopentadiene (concentration 2·10−3 M in water and 0.4 M in organic solvents) and the dienophile (concentration 1-5 mM) in the appropriate solvent, eventually containing a 0.01 M concentration of catalyst, was stirred at 25 °C until the UV-absorption of the dienophile had disappeared. The reaction mixture (diluted with water in the case of the organic solvents) was extracted with ether. The ether layer was washed with water and dried over sodium sulfate. After the evaporation of the ether the adducts were obtained in quantitative yields and almost invariably as oils. Only the reaction of 2.4c and 2.5 in water with 10 mM HCl gave a white precipitate. The product mixtures were analysed with respect to their endo-exo ratio by 1H-NMR spectroscopy. By repeating the extraction-drying procedure it was confirmed that the work-up procedure did not influence the endo-exo ratio of the isolated product mixture.

We have purified only the products of 2.4a and 2.4c by crystallisation from 1-propanol and ethanol, respectively. The purified products were still a mixture of endo and exo isomers. Elemental analyses of these compounds are given below. The Diels-Alder adducts of 2.4b, 2.4d-g were characterised by comparison of their NMR spectra with those of 2.4a and 2.4c. We will report here only the NMR data for the endo isomer, since the signals of the minor (7-12%) exo isomer partly coincide with the larger signals of the endo isomer and no attempts were made to separate the two.

2.6a: Anal. (C19H16N2O3) calcd. C: 71.22, H: 5.04, N: 8.75; found C: 70.82, H: 4.93, N: 8.66. 1H NMR (300 MHz, CDCl3) δ 1.65 (dd,1H), 1.99 (d,1H), 3.11 (d,1H), 3.52 (d,1H), 3.59 (s,1H), 4.46 (dd,1H), 5.85 (dd,1H), 6.47 (dd,1H), 7.21 (m,3H), 8.0 (m,5H), 8.6 (d,1H).

2.6b: 1H NMR (300 MHz, CDCl3) δ 1.61 (dd,1H), 2.00 (d,1H), 3.04 (d,1H), 3.40 (dd,1H), 3.54 (s,1H), 4.45 (dd,1H), 5.82 (dd,1H), 6.47 (dd,1H), 7.21 (m,5H), 7.45 (m,1H), 7.82 (m,1H), 7.99 (d,1H), 8.66 (d,1H).

2.6c: Anal. (C19H17NO) calcd. C: 82.87, H: 6.23, N: 5.09; found C: 82.28, H: 6.24, N: 5.21. 1H NMR (300 MHz, CDCl3) δ 1.61 (dd,1H), 2.05 (d,1H), 3.07 (d,1h), 3.43 (dd,1H), 3.53 (s,1H), 4.51 (dd,1H), 5.81 (dd,1H), 6.47 (dd,1H), 7.21 (m,5H), 7.41 (m,1H), 7.80 (m,1H), 7.99 (m,1H), 8.65 (m,1H).

2.6d: 1H NMR (300 MHz, CDCl3) δ 1.60 (dd,1H), 2.07 (d,1H), 3.06 (d,1H), 3.42 (d,1H), 3.54 (s,1H), 4.53 (dd,1H), 5.83 (dd,1H), 6.49 (dd,1H), 7.09 (d,2H), 7.22 (d,2H), 7.43 (m,1H), 7.80 (m,1H), 8.00 (d,1H), 8.67 (d,1H).

2.6e: 1H NMR (300 MHz, CDCl3) δ 1.59 (dd,1H), 2.05 (d,1H), 3.02 (d,1H), 3.39 (d,1H), 3.52 (s,1H), 4.49 (dd,1H), 5.81 (dd,1H), 6.48 (dd,1H), 6.82 (d,2H), 7.23 (d,2H), 7.43 (m,1H), 7.79 (m,1H), 7.99 (d,1H), 8.67 (d,1H).

2.6f: 1H NMR (300 MHz, D2O) δ 1.49 (dd,1H), 1.90 (d,1H), 2.98 (d,1H), 3.15 (d,1H), 3.27 (s,1H), 4.07 (s,2H), 4.25 (m,1H), 5.75 (m,1H), 6.44 (m,1H), 7.26 (s,4H), 7.55 (m1H), 7.87 (m,2H), 8.52 (d,1H).

2.6g: 1H NMR (300 MHz, CDCl3) δ 1.62 (dd,1H), 1.98 (d,1H), 3.07 (d,1H), 3.36 (s,9H), 3.43 (d,1H), 3.55 (2,1H) or (s,1H) (in the editors’ view), 4.45 (m,1H), 4,92 or 4.92 (in the editors’ view) (s,2H), 5.82 (m,1H), 6.47 (m,1H), 7.36 (d,2H), 7.46 (m,1H), 7.52 (d,2H), 7.83 (dt,1H), 7.99 (d,1H), 8.66 (d,1H).