CHIRALITY 26:243–248 (2014)
Comparison of a Molecular and an Immobilized Gadolinium(III)-tris [(1R,4S)-3-heptafluorobutanoyl-camphor] as Catalyst in the Asymmetric Danishefsky-Hetero-Diels-Alder-Reaction SKROLLAN STOCKINGER, AND OLIVER TRAPP* Ruprecht-Karls-Universität Heidelberg, Organisch-Chemisches Institut, Heidelberg, Germany
ABSTRACT On-column reaction gas chromatography combines the power of separation and rapid analysis of reactants and reaction products with screening of reactions in a single step. Not only conversions but the reaction rates at various temperatures can be obtained from single measurements, making this approach superior to the time-consuming measurements typically performed in reaction progress analysis. However, this approach has only been used in the investigation of interconversion processes, rearrangement reactions, and only a few examples of higherorder reactions are known. Here we present the screening of immobilized gadolinium(III)-tris [(1R,4S)-3-heptafluorobutanoyl-camphor] in the Danishefsky-hetero-Diels-Alder-reaction by enantioselective on-column reaction gas chromatography utilizing cryogenic focusing to achieve catalytic conversions in this higher-order reaction and subsequent separation of the enantiomeric product mixture to determine the enantiomeric ratio. The results obtained by this approach could be transferred to the conventional batch reaction at a larger scale, demonstrating that on-column reaction chromatography provides reliable results in the screening of enantioselective reactions. Chirality 26:243–248, 2014. © 2014 Wiley Periodicals, Inc. KEY WORDS: hetero-Diels-Alder-reaction; Danishefsky’s-diene; on-column reaction gas chromatography; enantioselective gas chromatography; asymmetric catalysis; combinatorial chemistry; screening INTRODUCTION
Since the discovery of the catalytic influence of aluminum chloride in Diels-Alder-Reactions by Yates and Eaten in 1960, the use of Lewis acids facilitates the controlled synthesis of six-membered ring systems.1,2 With up to four stereogenic centers built in a single step, this reaction is very attractive not only in the synthesis of natural products. Well-known catalysts are TiCl4 and SnCl4 as strong Lewis acids besides AlCl3, as well as softer Lewis acids like chiral complexes of lanthanides, transition metals, or main group elements.3–8 As chiral ligands 3-heptafluorobutanoylcamphorates (hfc), phosphanes like Chiraphos and binaphthol are widely used.4,5,9–11 Within this class of reactions the Danishefsky-hetero-Diels-Alder-reaction classifies the asymmetric transformation of (E)-1-methoxy3-trimethylsilyloxy-1,3-butadiene (Danishefsky’s diene) as activated diene with benzaldehyde to 2-phenyl-2,3-dihydro-4Hpyran-4-one.3,12,13 This reaction represents a lucent target for testing novel screening approaches of chiral catalysts, because the reactants are quite sensitive and there is a broad variation in the enantioselectivity depending on the type of ligand and metal.4,14 Lanthanide 3-heptafluorobutanoylcampherates like Eu(hfc)3 are known chiral selectors with applications as shift reagents in nuclear magnetic resonance (NMR) spectroscopy, as chiral selectors for the separation of enantiomers by enantioselective complexation gas chromatography (GC) and as chiral catalysts in Diels-Alder reactions.4,15–17 On the other hand, because of very long spin relaxation times gadolinium (III)-complexes have no application as shift reagents in NMR spectroscopy, but as contrast agents in magnetic resonance imaging (MRI) and as radiopharmaceuticals for imaging and show great potential in the treatment of tumors.18,19 Additionally, there are a few examples of gadolinium(III)-complexes as catalysts for Diels-Alder-reactions.8,20,21 © 2014 Wiley Periodicals, Inc.
In the present contribution we present gadolinium(III)-tris [(1R,4S)-3-heptafluorobutanoyl-camphor] (Gd(hfc)3) as a chiral catalyst in the Danishefsky-hetero-Diels-Alder-reaction. This catalyst has been covalently immobilized to polysiloxane and was used as a catalytically active stationary phase in enantioselective on-column reaction gas chromatography (ocRGC)22–25 as well as a molecular catalyst in conventional batch reactions for comparison. It has to be mentioned that already in 1961 Gil-Av and Herzberg-Minzly26 investigated the reaction kinetics of Diels-Alder reactions by ocRGC. They used nonvolatile dienophiles, so that the formed reaction product remained in the stationary phase. By contact time variations the degree of reaction can be determined and reaction rates estimated. A drawback of this method is that the reaction product and potentially formed by-products are not detected and may introduce an error in the calculation of reaction rate constants. Furthermore, such an approach is not suitable to investigate enantioselective processes, because the enantiomeric ratio cannot be determined. In order to ensure efficient catalyst improvement a reliable and versatile screening setup, which could even be speeded up by a high-throughput approach,27 is required to master the accumulating amount of data.28–33 Thus, our intention of using immobilized catalysts is not only to recycle the
Contract grant sponsor: European Research Council; Contract grant number: StG 258740. *Correspondence to: Oliver Trapp, Ruprecht-Karls-Universität Heidelberg, Organisch-Chemisches Institut, Im Neuenheimer Feld 270, 69120 Heidelberg, Germany. E-mail: [email protected] Received for publication 1 December 2013; Accepted 30 January 2014 DOI: 10.1002/chir.22312 Published online 19 March 2014 in Wiley Online Library (wileyonlinelibrary.com).
244
STOCKINGER AND TRAPP
material, but to create rapid screening platforms,34 where catalysts are prepared by complexing metals to the immobilized ligand at an analytical level, which lowers costs in the highthroughput screening of new catalysts and reactions.35–39 MATERIALS AND METHODS General All solvents and reagents were obtained from ABCR, Acros, SigmaAldrich, or VWR and were used without further purification unless 1 13 otherwise noted. Nuclear magnetic resonance ( H NMR and C NMR) spectra were recorded at 500 and 125 MHz, respectively, on a Bruker Avance 500 spectrometer (Rheinstetten, Germany) at room temperature. Chemical shifts (in ppm) were referenced to residual solvent protons.40 GC-MS measurements were performed on a Trace GC Ultra single quadrupole ISQ mass spectrometer (Thermo Scientific, San Jose, CA) equipped with a split injector (250°C), an on-column cold injector (2 min secondary N2-cooling time), and a flame ionization detector (250°C). Electron impact mass spectra were recorded at 70 eV. IR spectra were recorded on a Nicolet 6700 FT-IR with smart iTR ATR device (Thermo Scientific). [(1R,4S)-3-Heptafluorobutanoyl-10-propylenoxycamphor]polysiloxane in selector concentrations of 20% were prepared as reported previously.17 Fused silica capillaries (0.25 mm I.D.), obtained from Microquartz (Munich, Germany), were coated by the static method described by Grob.41 For cryo-focusing a cryogenic CO2 cold trap system by SGE Analytical Science (Victoria, Australia) was installed according to the experimental setup outlined in Figure 2.
Synthesis of Gadolinium(III)-tris[(1R,4S)-3heptafluorobutanoyl-camphor] (Gd(hfc)3) 1 Gadolinium(II)-acetate hydrate (337 mg, 1,00 mmol) was added to a solution of [(1R,4S)-3-Heptafluorobutanoyl-camphor] (200 mg, 574 μmol) and sodium carbonate (120 mg, 1.13 mmol) in dichloromethane (25.0 mL) and refluxed for 3 h. The reaction mixture was diluted with water and extracted with pentane. The organic extracts were washed with water, dried over MgSO4, and the solvent was removed under vacuum to yield gadolinium(III)-tris[(1R,4S)-3-heptafluorobutanoyl-camphor] (1, 405 mg, 338 μmol, 58%) as a yellow solid. ATR-FTIR: υ = 2965, 1647, -1 1522, 1342, 1200, 1212, 1115, 896, 748 cm . HR-MS (FAB+, m/z): calc. GdC42H42F21O6 [M]: 1199.1887, found: 1199.1866.
Synthesis of Gadolinium(III)-tris[(1R,4S)-3heptafluorobutanoyl-10-propylenoxycamphor]polysiloxane (Gd(hfpc)3@PS) 2 Gadolinium(III)-tris[(1R,4S)-3-heptafluorobutanoyl-10-propylenoxycamphor]-polysiloxane (2) was synthesized according to a procedure previously published.17 Therefore, [(1R,4S)-3-heptafluorobutanoyl-10propylenoxycamphor]-polysiloxane (50.0 mg) and gadolinium(III)-acetate hydrate (18.9 mg, 56.5 μmol) were reacted and purified to yield 49.1 mg gadolinium(III)-tris[(1R,4S)-3-heptafluorobutanoyl-10-propylenoxycamphor]polysiloxane as a red oil. ATR-FTIR: υ = 2962, 1653, 1648, 1559, 1540, 1521, -1 1507, 1457, 1258, 1231, 1008, 793, 700 cm .
Synthesis of 2-Phenyl-2,3-dihydro-4H-pyran-4-one 5 2-Phenyl-2,3-dihydro-4H-pyran-4-one was synthesized based on the method published by Schurig and colleagues.4 (E)-1-Methoxy-3trimethylsilyloxy-1,3-butadiene (Danishefsky’s diene, 107 mg, 0.62 mmol) and freshly distilled benzaldehyde (62.6 mg, 0.59 mmol) were added to a solution of europium(III)-tris[(1R,4S)-3-heptafluorobutanoyl-camphor] (37.0 mg, 5 mol%) in n-hexane (3.00 mL). The reaction mixture was stirred at room temperature till completion (monitored by thin-layer chromatography [TLC] and GC) and was quenched by adding of trifluoroacetic acid (0.5% in trifluoroacetic acid [TFA], 5.00 mL). The desired product was purified by column chromatography (silica gel, n-hexane/ diethyl ether 3:2) and isolated as a yellow oil (101 mg, 0.58 mmol, 98%). 1 H-NMR (500.13 MHz, CDCl3): δ = 7.49-7.40 (m, 6H, H-aryl, - = CHO-), 5.53 (d, 1H, J = 6.00 Hz, -COCH = -), 5.43 (dd, 1H, J = 14.53 Hz, J = 2.73 Hz, -CHAr), 2.92 (dd, 1H, J = 15.73 Hz, J = 15.73 Hz, -COCH2-), 2.66 (dd, 1H, Chirality DOI 10.1002/chir
13
J = 16.78 Hz, J = 2.82 Hz, -COCH2-) ppm. C-NMR (125.77 MHz, CDCl3): δ = 192.33, 163.35, 137.95, 129.09, 129.00, 126.24, 107.52, 81.24, 43.53 ppm. -1 ATR-FTIR: υ = 3031, 1670, 1592, 1267, 1226, 1037, 932, 756 cm . HR-MS (EI+, m/z): calc. for C11H10O2 [M]: 174.0681, found: 174.0672.
Off-Column Diels-Alder-Reaction A mixture of (E)-1-Methoxy-3-trimethylsilyloxy-1,3-butadiene (Danishefsky’s diene, 52.2 mg, 0.34 mmol), freshly distilled benzaldehyde (28.9 mg, 0.29 mmol), and tetradecene (7.68 μmol) in pentane (1.5 mL) was added to the catalyst (5 mol%, Eu(hfc)3: 18.5 mg, Gd(hfc)3: 18.6 mg). The flask was sealed and stirred under the given conditions. For the GC analysis, 0.10 mL was taken from the flask, filtered, and mixed with TFA (50% in DCM, 25.0 μL) and diluted with pentane. For quantitative analysis the free induction decay (FID) signal was used and to calculate conversions the integral was corrected by the number of carbon atoms.
On-Column Diels-Alder-Reaction (ocRGC) For the on-column measurements freshly distilled benzaldehyde (1 eq), (E)-1-Methoxy-3-trimethylsilyloxy-1,3-butadiene (Danishefsky’s diene, 0.9 eq) and tetradecene were mixed in 1 mL pentane. At regular intervals the sample was checked with a reference column (Chirasil-β-Dex, 17 m, 250 nm, 250 μm I.D.) to exclude changes or decomposition of the sample. The reaction mixture was injected with an on-column syringe from SGE following the general on-column cold injection techniques with a secondary N2-cooling time of 2 min. The cryogenic CO2 cooling was activated and deactivated according the measurement process outlined in Figure 2 or mentioned in the text. For quantitative analysis the FID signal was used and to calculate conversions the integral was corrected by the number of carbon atoms. For the unambiguous assignment of the formed hetero-Diels-Alder product, pyron 5 was dissolved in n-pentane and injected via split injection onto the metal(hfc) column (25 m, 250 nm, 250 μm I.D.) without cryogenic cooling or additional separation column.
RESULTS AND DISCUSSION
To extend the scope of ocRGC of higher-order reactions to asymmetric catalysis, we shifted our attention to the Danishefsky-hetero-Diels-Alder-reaction. In a prescreening we found some interesting activity of gadolinium(III)-tris [(1R,4S)-3-heptafluorobutanoyl-camphor] (Gd(hfc)3) complexes as chiral Lewis acid catalyst (Scheme 1). To perform more systematic studies we developed a strategy to immobilize Gd(hfc)3 1 to polysiloxane via a propyleneoxy linker attached to C10 in the camphor moiety, resulting in gadolinium (III)-tris[(1R,4S)-3-heptafluorobutanoyl-10-propyleneoxycamphor]polysiloxane (Gd(hfpc)3@PS) 2. This polymeric catalyst allows fabricating reactor capillaries for ocRGC that enables a fast and efficient screening of reactions of higher-order like the Diels-Alder-reaction where a diene and dienophile reacts to form six-membered rings. For this purpose Gd (hfc)3@PS 2 (Fig. 1), a polysiloxan bonded version of Gd (hfc)3 1, was synthesized with 20% selector content and coated onto the inner surface of fused-silica capillaries (0.25 mm I.D.) according to the static method described by Grob, resulting in a defined polymer film thickness of
Scheme 1. Danishefsky-hetero-Diels-Alder-reaction of (E)-1-methoxy-3trimethylsilyloxy-1,3-butadiene (Danishefsky’s diene, 3) with benzaldehyde 4 forming the chiral pyron 5 after simultaneous elimination of trimethylsilanol and methanol by traces of water at these low substrate concentrations.
ENANTIOSELECTIVE ON-COLUMN REACTION GC
Fig. 1. Gadolinium(III)-tris[(1R,4S)-3-heptafluorobutanoyl-camphor] (Gd(hfc)3, 1) and gadolinium(III)-tris[(1R,4S)-3-heptafluorobutanoyl-10-propylenoxycamphor]polysiloxane (Gd(hfc)3@PS, 2).
250 nm (Gd(hfc)3@PS, 2).33 Additionally, the [(1R,4S)-3heptafluorobutanoyl-10-propyleneoxycamphor]-polysiloxane (hfc@PS) complexes of nickel(II), platinum(II), palladium (II), lanthanum(III), europium(III), and manganese(II) were synthesized and tested following the same procedure and conditions as for Gd(hfc)3@PS 2. Remarkably, Gd(hfc)3@PS 2 was the only catalytically active column of this library of metal complexes. In combination with the other mentioned
245
metal(hfc) columns, the injected reactant mixture could be detected without any pyron 5 formation, which shows the great potential of the immobilized catalyst Gd(hfc)3@PS 2. To make analysis and screening of higher-order reactions like the Diels-Alder-reaction feasible, the normal reaction GC setup had to be modified to increase the contact time Δt of the reactants and to make refocusing within a run possible, which not only helps to improve the conversion of the components, but also provides a means to improve sensitivity by focusing the peaks, which is necessary to precisely determine enantiomeric ratios. We found that the immobilized catalyst Gd(hfc)3@PS is highly catalytically active and the reactivity is sufficient in the ocRGC setup so that even by simple injection of the reactant mixture the desired pyron 5 is formed; however, interaction of the polymeric catalyst is so strong that only very broad peaks are observed, which do not allow quantifying the enantiomeric ratio. Hence, a cryogenic CO2 cold trap was installed in the GC oven to focus the formed pyron 5. Cryogenic focusing is a popular method to improve peak shapes or to concentrate substances above the limit of detection and was very recently proven to be a promising analytical device to improve detection of higher-order reactions in ocRGC systems by our research group.28 For this application a cryogenic CO2 cold trap was installed in the first part of the separation column (Fig. 2a). As separation column, a Chirasil-β-Dex42,43 fusedsilica column was used, which provided an excellent separation of the product enantiomers 5. An installation of the cryogenic focusing trap on the last part of the catalytically
Fig. 2. Experimental setup to perform enantioselective on-column reaction gas chromatography (a), the process protocol to perform reactions using cryogenic focusing (b), and schedule of the cryogenic cooling steps during an ocRGC experiment (c). Experimental conditions: On-column cold injection, 120°C, 10 kPa for 1 min, then 100 kPa/min to 100 kPa; He as inert carrier gas; catalytically active column: Gd(hfc)3@PS, 5 m, 250 nm film thickness, 250 μm ID; separation column: Chirasil-β-Dex, 10 m, 250 nm film thickness, 250 μm ID. Chirality DOI 10.1002/chir
246
STOCKINGER AND TRAPP
active fused-silica column was also tested but there were no differences detectable compared to the presented setup. To achieve an efficient separation of the formed product enantiomers the cryogenic focusing should only affect the formed pyron 5, and therefore the cooling was activated after 2.4 min of the injection of the reactants. As outlined in the process protocol (Fig. 2c), at this point the remaining reactants as well as the solvent and the internal standard already left the catalytically active column and some of these components already reached the detector. After 6 min the formed pyron 5 was fully focused by the cryogenic trap, the cooling was turned off, and the focused reaction product was released onto the separation column. For an unambiguous assignment of the formed heteroDiels-Alder product, pyron 5 was synthesized and measured onto a 25 m Gd(hfc)3@PS 2 column without using cryogenic cooling. The retention time of the reference sample was about 36 min compared to the retention time of the on-column synthesized product, which was already eluted after about 32 min. This difference in the retention times of about 4 min indicates the formation of the reaction product on the catalytically active column in the ocRGC experiment. Furthermore, without a catalytic column there was no reaction detectable under the same experimental conditions. The optimum length of the catalytically active Gd(hfc)3@PS 2 column was determined to be 5 m, which gives decreased measurement times and therefore higher throughput under maintained reactivity. In Figure 3 a resulting chromatogram of an ocRGC experiment of the described Danishefsky-hetero-Diels-Alder reaction using a 5 m catalytically active Gd(hfc)3@PS 2 column and a 10 m Chirasil-β-Dex separation column is depicted. Due to the high catalytic activity of the Gd(hfc)3@PS 2 column it was not possible to detect remaining (E)-1-methoxy-3trimethylsilyloxy-1,3-butadiene (Danishefsky’s diene 4), but only the transformed (E)-4-methoxybut-3-en-2-one 4* which was not separated from benzaldehyde 3 under the here applied experimental conditions. The enantiomers of pyron 5 could be baseline separated in the presented setup. The elution order of the enantiomers ((+)-5 before (–)-5) was assigned as described in the literature.4 With the here presented experimental on-column reaction chromatographic setup an enantiomeric ratio (er) of (+)-5 to (–)-5 of 44:56 corresponding to an enantiomeric excess (ee) of (–)-5 of 12% could be determined. An average standard deviation of 0.6% for the er values and 1.1% for the ee value could be
determined. Keeping in mind that the reaction time by the ocRGC experiment of the here described second-order reaction is well below 1 s and therefore also the conversion is low, it is quite remarkable that the er is detectable with high precision. In the next step we compared the results obtained by ocRGC with the results obtained by the conventional batch reaction. Of course, the conversions can be simply optimized by extending the reaction time; however, the er could differ because of different catalyst and substrate concentrations. For comparison, the molecular Gd(hfc)3 catalyst was synthesized and catalysis was performed in a conventional flask. For comparison, we also performed catalytic tests with the commercially available Eu(hfc)3 catalyst. Besides the use of Eu(hfc)3 in various applications, it is probably the best known catalyst for these types of Diels-Alder-reactions.2,4 In Table 1 the results of the catalytic experiments are summarized. Gd(hfc)3 1 shows approximately a similar reaction time compared to Eu(hfc)3 (at 20°C both require 1 day for completion of the reaction, at –30°C Gd(hfc)3 3 days and Eu(hfc)3 2 days, respectively). The er and ee catalyzed by Gd(hfc)3 1 are throughout these measurements slightly above the values obtained by using Eu(hfc)3. For example, Gd(hfc)3 1 gives an er of 33:67, whereas Eu(hfc)3 provides only a ratio of 35:65 at 20°C and at –30°C the difference exceeds 34:66 with Eu (hfc)3 and 30:70 with Gd(hfc)3 1. Further improvements of this reaction of Gd(hfc)3 1 to an er of 27:73 at –30°C could be achieved by adding an excess of the ligand (hfc) and TABLE 1. Enantiomeric ratio er of (+)-5 and (-)-5 and enantiomeric excess ee of (-)-5 obtained by conventional reaction conditions in a glass flask using molecular catalysts 1 and Eu(hfc)3 Catalyst
T [°C]
Reaction completed after
er [%]
ee [%]
Eu(hfc)3
20 -30 20 0 -30 -70
1d 2d 1d 1.5d 3d 3.5d
35:65 34:66 33:67 31:69 30:70 31:69
30% 32% 35% 38% 40% 38%
Gd(hfc)3
The reported values represent averages obtained from at least three GC measurements. The average standard deviation does not exceed ±1% for the er values and ±2% for the ee values.
Fig. 3. Chromatogram of a Gd(hfc)3@PS 2 catalyzed ocRGC Danishefsky-hetero-Diels-Alder-reaction using the cryogenic focusing setup outlined in Figure 2. Chirality DOI 10.1002/chir
ENANTIOSELECTIVE ON-COLUMN REACTION GC
increasing the amount of catalyst to 20 mol%, which also reduces the reaction time considerably. Surprisingly, the here measured er’s were lower than reported in the literature (ee’s of up to 42% at –33°C) by other groups. The only difference might be that we directly analyzed the enantiomeric composition using the crude reaction mixture and other groups perform purification by flash chromatography. It is well known that nonlinear effects might cause deviations from the original enantiomeric composition, leading to an enrichment of the major enantiomer. Nevertheless, the obtained results are comparable with respect to conversions under these reaction conditions. Compared to Eu(hfc)3, our findings indicate that Gd(hfc)3 1 gives slightly higher enantiomeric selectivities combined with an almost similar reactivity. CONCLUSION
In this report we demonstrated that enantioselective ocRGC of higher-order reactions as shown for the here-investigated Danishefsky-hetero-Diels-Alder reaction is feasible and it is a versatile tool to rapidly assess the enantioselectivity of catalyzed reactions. To improve enantioselective analysis cryogenic focusing by CO2 in a cold trap on the first part of the separation column was used. Cryogenic focusing overcomes peak broadening effects caused by strong interaction effects of the reaction product with the catalytically active stationary phase. As the catalytically active stationary phase used in the first section of the column setup of the ocRGC experiment, Gd(hfc)3 was immobilized by modification of the camphor moiety with a propyleneoxy linker at C10 and hydrosilylation to polysiloxane. It has to be pointed out that the enantioselectivity determined by the rapid screening approach correctly predicted the enantioselectivity of the reaction performed with the molecular catalyst in a conventional flask. Furthermore, the enantioselectivities, which are in general moderate for this type of reaction, are slightly better compared to the standard catalyst Eu(hfc)3. ACKNOWLEDGMENTS
We thank the Deutsche Forschungsgemeinschaft (DFG) for financial support of this research (SFB 623 ‘Molecular Catalysts: Structure and Functional Design’). O.T. was supported by the European Research Council under Grant Agreements No. StG 258740. We thank Chromasoft GmbH for the supply of the cryogenic CO2 cold trap system. LITERATURE CITED 1. Yates P, Eaton P. Acceleration of the Diels-Alder Reaction by Aluminum chloride. J Am Chem Soc 1960;82:4436–4437. 2. Fringuelli F, Taticchi A. The Diels-Alder reaction: selected practical methods. New York: John Wiley & Sons; 2002. 3. Bednarski M, Maring C, Danishefsky S. Chiral induction in the cyclocondensation of aldehydes with siloxydienes. Tetrahedron Lett 1983;24:3451–3454. 4. Keller F, Weinmann H, Schurig V. Chiral polysiloxane-fixed metal 1,3diketonates (chirasil-metals) as catalytic Lewis acids for a hetero Diels-Alder reaction — inversion of enantioselectivity upon catalyst – polymer binding. Chem Ber 1997;130:879–885. 5. Togni A. Asymmetric hetero Diels-Alder reactions catalyzed by novel chiral vanadium( IV) bis( 1,3-diketonato) complexes. Organometallics 1990;9:3106–3113. 6. Maruoka K, Oishi M, Yamamoto H. Methylaluminum bis(4-substituted-2,6di-ter-butylphenoxide) as an efficient non-chelating Lewis acid. Synlett 1993;9:683–685.
247
7. Smith MB, March J. March’s advanced organic chemistry. Hoboken, NJ: John Wiley & Sons; 2007. 8. Evans DA, Wu J. Enantioselective rare-earth catalyzed quinone Diels Alder reactions. J Am Chem Soc 2003;125:10162–10163. 9. Maruoka K, Itoh T, Shirasaka T, Yamamoto H. Asymmetric heteroDiels-Alder reaction catalyzed by a chiral organoaluminum reagent. J Am Chem Soc 1988;110:310–312. 10. Faller JW, Smart CJ. Use of organometallic complexes of ruthenium in the Lewis acid catalyzed hetero Diels-Alder reaction. Tetrahedron Lett 1989;30:1189–1192. 11. Ishitani H, Kobayashi S. Catalytic asymmetric aza Diels-Alder reactions using a chiral lanthanide Lewis acid. Enantioselective synthesis of tetrahydroquinoline derivatives using a catalytic amount of a chiral source. Tetrahedron Lett 1996;37:7357–7360. 12. Danishefsky S, Bednarski M. Lanthanide catalysis of cycloadditions of heterodienes with enol ethers. Tetrahedron Lett 1984;25:721–724. 13. Danishefsky S, Maring C. A new approach to the synthesis of hexoses: an entry to (+)-fucose and (+)-daunosamine. J Am Chem Soc 1985;107:1269–1274. 14. Gerstberger G, Palm C, Anwander R. The Danishefsky hetero Diels-Alder reaction mediated by organolanthanide-modified mesoporous silicate MCM-41. Chem Eur J 1999;5:997–1005. 15. Hesse M, Meier H, Zeeh B. Spektroskopische Methoden in der organischen Chemie. Berlin: Georg Thieme; 2005. 16. Schurig V, Bürkle W. Extending the scope of enantiomer resolution by complexation gas chromatography. J Am Chem Soc 1982;104:7573–7580. 17. Spallek MJ, Storch G, Trapp O. Straightforward synthesis of poly (dimethylsiloxane) phases with immobilized (1R)-3-(perfluoroalkanoyl) camphorate metal complexes and their application in enantioselective complexation gas chromatography. Eur J Organ Chem 2012;2012:3929–3945. 18. Morill TC. Lanthanide shift reagents in stereochemical analysis. Rochester, NY: VCH Publisher; 1986. 19. Frullano L, Meade TJ. Multimodal MRI contrast agents. J Biol Inorg Chem 2007;12:939–949. 20. Ma Y, Qian C, Xie M, Sun J. Lanthanide chloride catalyzed imino Diels Alder reaction. one-pot synthesis of pyrano[3,2-c]- and furo[3,2-c]quinolines. J Organ Chem 1999;64:6462–6467. 21. Yu Y, Zhou J, Yao Z, Xu F, Shen Q. Stereoselective synthesis of pyrano [3,2-c]- and furano[3,2-c]quinolines: gadolinium chloride catalyzed onepot aza–Diels–Alder reactions. Heteroatom Chem 2010;21:351–354. 22. Trapp O. Unified equation for access to rate constants of first-order reactions in dynamic and on-column reaction chromatography. Anal Chem 2006;78:189–198. 23. Trapp O, Weber SK, Bauch S, Hofstadt W. High-throughput screening of catalysts by combining reaction and analysis. Angew Chem Int Ed 2007;46:7307–7310; Angew Chem 2007;119:7447–7451. 24. Trapp O, Weber SK, Bauch S, Bäcker T, Hofstadt W, Spliethoff B. High throughput kinetic study of hydrogenations over palladium nanoparticles — combination of reaction and analysis. Chem Eur J 2008;14:4657–4666. 25. Trapp O. Investigation of the stereodynamics of molecules and catalyzed reactions by CE. Electrophoresis 2010;31:786–813. 26. Gil-Av E, Herzberg-Minzly Y. Gas-chromatographic study of the rate of Diels-Alder addition. Proc Chem Soc 1961;316. 27. Boelens HFM, Iron D, Westerhuis JA, Rothenberg G. Tracking chemical kinetics in high-throughput systems. Chem Eur J 2003;9:3876–3881. 28. Becker S, Höbenreich H, Vogel A, Knorr J, Wilhelm S, Rosenau F, Jaeger K-E, Reetz MT, Kolmar H. Single-cell high-throughput screening to identify enantioselective hydrolytic enzymes. Angew Chem Int Ed 2008;47:5085–5088; Angew Chem 2008;120:5163–5166. 29. de Bellefon C, Tanchoux N, Caravieilhes S, Grenouillet P, Hessel V. Microreactors for dynamic, high throughput screening of fluid/liquid molecular catalysis. Angew Chem Int Ed 2000;39:3442–3445; Angew Chem 2000;112(19):3584–3587. 30. Kluwer AM, Detz RJ, Abiri Z, van der Burg AM, Reek JNH. Evolutionary catalyst screening: iridium-catalyzed imine hydrogenation. Adv Synth Catal 2012;354:89–95. 31. Krabbe SW, Hatcher MA, Bowman RK, Mitchell MB, McClure MS, Johnson JS. Copper-catalyzed asymmetric hydrogenation of aryl and heteroaryl ketones. Organ Lett 2013;15:4560–4563. 32. Cai C, Chung JYL, McWilliams JC, Sun Y, Shultz CS, Palucki M. From high-throughput catalyst screening to reaction optimization: detailed Chirality DOI 10.1002/chir
248
33.
34. 35.
36.
37.
38.
STOCKINGER AND TRAPP
investigation of regioselective Suzuki coupling of 1,6-naphthyridone dichloride. Organ Proc Res Dev 2007;11:328–335. Lefort L, Boogers JF, deVries AM, deVries J. High throughput screening of Monophos instant ligand library leads to a ton-scale asymmetric hydrogenation process. Top Catal 2006;40:185–191. Trapp O. High-throughput monitoring of interconverting stereoisomers and catalytic reactions. Chem Today 2008;26:26–28. Troendlin J, Rehbein J, Hiersemann M, Trapp O. Integration of catalysis and analysis is the key: rapid and precise investigation of the catalytic asymmetric Gosteli–Claisen rearrangement. J Am Chem Soc 2011;133:16444–16450. Stockinger S, Trapp O. Integrating reaction and analysis: investigation of higher-order reactions by cryogenic trapping. Beilstein J Organ Chem 2013;9:1837–1842. Lang C, Gartner U, Trapp O. Catalysts by the meter: rapid screening approach of N-heterocyclic carbene ligand based catalysts. Chem Commun 2011;47:391–393. Weber SK, Bremer S, Trapp O. Integration of reaction and separation in a micro-capillary column reactor—palladium nanoparticle catalyzed C–C bond forming reactions. Chem Eng Sci 2010;65:2410–2416.
Chirality DOI 10.1002/chir
39. Sandel S, Weber SK, Trapp O. Oxidations with bonded salen-catalysts in micro capillaries. Chem Eng Sci 2012;83:171–179. 40. Fulmer GR, Miller AJM, Sherden NH, Gottlieb HE, Nudelman A, Stoltz BM, Bercaw JE, Goldberg KI. NMR chemical shifts of trace impurities: common laboratory solvents, organics, and gases in deuterated solvents relevant to the organometallic chemist. Organometallics 2010;29:2176–2179. 41. Grob K. Making and manipulating capillary columns for gas chromatography. Heidelberg, Basel, New York: Dr. Alfred Hüthig Verlag; 1986. 42. Schurig V, Schmalzing D, Schleimer M. Enantiomer separation on immobilized Chirasil-Metal and Chirasil-Dex by gas chromatography and supercritical fluid chromatography. Angew Chem Int Ed 1991;30:987–989; Angew Chem 1991;103:994–996. 43. Cousin H, Trapp O, Peulon-Agasse V, Pannecoucke X, Banspach L, Trapp G, Jiang Z, Combret JC, Schurig V. Synthesis, NMR characterisation and polysiloxane-based immobilization of the three regioisomeric monooctenylpermethyl-β-cyclodextrins - application in enantioselective GC. Eur J Org Chem 2003;3273–3287.
Comparison of a Molecular and an Immobilized Gadolinium(III)-tris [(1R,4S)-3-heptafluorobutanoyl-camphor] as Catalyst in the Asymmetric Danishefsky-Hetero-Diels-Alder-Reaction SKROLLAN STOCKINGER, AND OLIVER TRAPP* Ruprecht-Karls-Universität Heidelberg, Organisch-Chemisches Institut, Heidelberg, Germany
ABSTRACT On-column reaction gas chromatography combines the power of separation and rapid analysis of reactants and reaction products with screening of reactions in a single step. Not only conversions but the reaction rates at various temperatures can be obtained from single measurements, making this approach superior to the time-consuming measurements typically performed in reaction progress analysis. However, this approach has only been used in the investigation of interconversion processes, rearrangement reactions, and only a few examples of higherorder reactions are known. Here we present the screening of immobilized gadolinium(III)-tris [(1R,4S)-3-heptafluorobutanoyl-camphor] in the Danishefsky-hetero-Diels-Alder-reaction by enantioselective on-column reaction gas chromatography utilizing cryogenic focusing to achieve catalytic conversions in this higher-order reaction and subsequent separation of the enantiomeric product mixture to determine the enantiomeric ratio. The results obtained by this approach could be transferred to the conventional batch reaction at a larger scale, demonstrating that on-column reaction chromatography provides reliable results in the screening of enantioselective reactions. Chirality 26:243–248, 2014. © 2014 Wiley Periodicals, Inc. KEY WORDS: hetero-Diels-Alder-reaction; Danishefsky’s-diene; on-column reaction gas chromatography; enantioselective gas chromatography; asymmetric catalysis; combinatorial chemistry; screening INTRODUCTION
Since the discovery of the catalytic influence of aluminum chloride in Diels-Alder-Reactions by Yates and Eaten in 1960, the use of Lewis acids facilitates the controlled synthesis of six-membered ring systems.1,2 With up to four stereogenic centers built in a single step, this reaction is very attractive not only in the synthesis of natural products. Well-known catalysts are TiCl4 and SnCl4 as strong Lewis acids besides AlCl3, as well as softer Lewis acids like chiral complexes of lanthanides, transition metals, or main group elements.3–8 As chiral ligands 3-heptafluorobutanoylcamphorates (hfc), phosphanes like Chiraphos and binaphthol are widely used.4,5,9–11 Within this class of reactions the Danishefsky-hetero-Diels-Alder-reaction classifies the asymmetric transformation of (E)-1-methoxy3-trimethylsilyloxy-1,3-butadiene (Danishefsky’s diene) as activated diene with benzaldehyde to 2-phenyl-2,3-dihydro-4Hpyran-4-one.3,12,13 This reaction represents a lucent target for testing novel screening approaches of chiral catalysts, because the reactants are quite sensitive and there is a broad variation in the enantioselectivity depending on the type of ligand and metal.4,14 Lanthanide 3-heptafluorobutanoylcampherates like Eu(hfc)3 are known chiral selectors with applications as shift reagents in nuclear magnetic resonance (NMR) spectroscopy, as chiral selectors for the separation of enantiomers by enantioselective complexation gas chromatography (GC) and as chiral catalysts in Diels-Alder reactions.4,15–17 On the other hand, because of very long spin relaxation times gadolinium (III)-complexes have no application as shift reagents in NMR spectroscopy, but as contrast agents in magnetic resonance imaging (MRI) and as radiopharmaceuticals for imaging and show great potential in the treatment of tumors.18,19 Additionally, there are a few examples of gadolinium(III)-complexes as catalysts for Diels-Alder-reactions.8,20,21 © 2014 Wiley Periodicals, Inc.
In the present contribution we present gadolinium(III)-tris [(1R,4S)-3-heptafluorobutanoyl-camphor] (Gd(hfc)3) as a chiral catalyst in the Danishefsky-hetero-Diels-Alder-reaction. This catalyst has been covalently immobilized to polysiloxane and was used as a catalytically active stationary phase in enantioselective on-column reaction gas chromatography (ocRGC)22–25 as well as a molecular catalyst in conventional batch reactions for comparison. It has to be mentioned that already in 1961 Gil-Av and Herzberg-Minzly26 investigated the reaction kinetics of Diels-Alder reactions by ocRGC. They used nonvolatile dienophiles, so that the formed reaction product remained in the stationary phase. By contact time variations the degree of reaction can be determined and reaction rates estimated. A drawback of this method is that the reaction product and potentially formed by-products are not detected and may introduce an error in the calculation of reaction rate constants. Furthermore, such an approach is not suitable to investigate enantioselective processes, because the enantiomeric ratio cannot be determined. In order to ensure efficient catalyst improvement a reliable and versatile screening setup, which could even be speeded up by a high-throughput approach,27 is required to master the accumulating amount of data.28–33 Thus, our intention of using immobilized catalysts is not only to recycle the
Contract grant sponsor: European Research Council; Contract grant number: StG 258740. *Correspondence to: Oliver Trapp, Ruprecht-Karls-Universität Heidelberg, Organisch-Chemisches Institut, Im Neuenheimer Feld 270, 69120 Heidelberg, Germany. E-mail: [email protected] Received for publication 1 December 2013; Accepted 30 January 2014 DOI: 10.1002/chir.22312 Published online 19 March 2014 in Wiley Online Library (wileyonlinelibrary.com).
244
STOCKINGER AND TRAPP
material, but to create rapid screening platforms,34 where catalysts are prepared by complexing metals to the immobilized ligand at an analytical level, which lowers costs in the highthroughput screening of new catalysts and reactions.35–39 MATERIALS AND METHODS General All solvents and reagents were obtained from ABCR, Acros, SigmaAldrich, or VWR and were used without further purification unless 1 13 otherwise noted. Nuclear magnetic resonance ( H NMR and C NMR) spectra were recorded at 500 and 125 MHz, respectively, on a Bruker Avance 500 spectrometer (Rheinstetten, Germany) at room temperature. Chemical shifts (in ppm) were referenced to residual solvent protons.40 GC-MS measurements were performed on a Trace GC Ultra single quadrupole ISQ mass spectrometer (Thermo Scientific, San Jose, CA) equipped with a split injector (250°C), an on-column cold injector (2 min secondary N2-cooling time), and a flame ionization detector (250°C). Electron impact mass spectra were recorded at 70 eV. IR spectra were recorded on a Nicolet 6700 FT-IR with smart iTR ATR device (Thermo Scientific). [(1R,4S)-3-Heptafluorobutanoyl-10-propylenoxycamphor]polysiloxane in selector concentrations of 20% were prepared as reported previously.17 Fused silica capillaries (0.25 mm I.D.), obtained from Microquartz (Munich, Germany), were coated by the static method described by Grob.41 For cryo-focusing a cryogenic CO2 cold trap system by SGE Analytical Science (Victoria, Australia) was installed according to the experimental setup outlined in Figure 2.
Synthesis of Gadolinium(III)-tris[(1R,4S)-3heptafluorobutanoyl-camphor] (Gd(hfc)3) 1 Gadolinium(II)-acetate hydrate (337 mg, 1,00 mmol) was added to a solution of [(1R,4S)-3-Heptafluorobutanoyl-camphor] (200 mg, 574 μmol) and sodium carbonate (120 mg, 1.13 mmol) in dichloromethane (25.0 mL) and refluxed for 3 h. The reaction mixture was diluted with water and extracted with pentane. The organic extracts were washed with water, dried over MgSO4, and the solvent was removed under vacuum to yield gadolinium(III)-tris[(1R,4S)-3-heptafluorobutanoyl-camphor] (1, 405 mg, 338 μmol, 58%) as a yellow solid. ATR-FTIR: υ = 2965, 1647, -1 1522, 1342, 1200, 1212, 1115, 896, 748 cm . HR-MS (FAB+, m/z): calc. GdC42H42F21O6 [M]: 1199.1887, found: 1199.1866.
Synthesis of Gadolinium(III)-tris[(1R,4S)-3heptafluorobutanoyl-10-propylenoxycamphor]polysiloxane (Gd(hfpc)3@PS) 2 Gadolinium(III)-tris[(1R,4S)-3-heptafluorobutanoyl-10-propylenoxycamphor]-polysiloxane (2) was synthesized according to a procedure previously published.17 Therefore, [(1R,4S)-3-heptafluorobutanoyl-10propylenoxycamphor]-polysiloxane (50.0 mg) and gadolinium(III)-acetate hydrate (18.9 mg, 56.5 μmol) were reacted and purified to yield 49.1 mg gadolinium(III)-tris[(1R,4S)-3-heptafluorobutanoyl-10-propylenoxycamphor]polysiloxane as a red oil. ATR-FTIR: υ = 2962, 1653, 1648, 1559, 1540, 1521, -1 1507, 1457, 1258, 1231, 1008, 793, 700 cm .
Synthesis of 2-Phenyl-2,3-dihydro-4H-pyran-4-one 5 2-Phenyl-2,3-dihydro-4H-pyran-4-one was synthesized based on the method published by Schurig and colleagues.4 (E)-1-Methoxy-3trimethylsilyloxy-1,3-butadiene (Danishefsky’s diene, 107 mg, 0.62 mmol) and freshly distilled benzaldehyde (62.6 mg, 0.59 mmol) were added to a solution of europium(III)-tris[(1R,4S)-3-heptafluorobutanoyl-camphor] (37.0 mg, 5 mol%) in n-hexane (3.00 mL). The reaction mixture was stirred at room temperature till completion (monitored by thin-layer chromatography [TLC] and GC) and was quenched by adding of trifluoroacetic acid (0.5% in trifluoroacetic acid [TFA], 5.00 mL). The desired product was purified by column chromatography (silica gel, n-hexane/ diethyl ether 3:2) and isolated as a yellow oil (101 mg, 0.58 mmol, 98%). 1 H-NMR (500.13 MHz, CDCl3): δ = 7.49-7.40 (m, 6H, H-aryl, - = CHO-), 5.53 (d, 1H, J = 6.00 Hz, -COCH = -), 5.43 (dd, 1H, J = 14.53 Hz, J = 2.73 Hz, -CHAr), 2.92 (dd, 1H, J = 15.73 Hz, J = 15.73 Hz, -COCH2-), 2.66 (dd, 1H, Chirality DOI 10.1002/chir
13
J = 16.78 Hz, J = 2.82 Hz, -COCH2-) ppm. C-NMR (125.77 MHz, CDCl3): δ = 192.33, 163.35, 137.95, 129.09, 129.00, 126.24, 107.52, 81.24, 43.53 ppm. -1 ATR-FTIR: υ = 3031, 1670, 1592, 1267, 1226, 1037, 932, 756 cm . HR-MS (EI+, m/z): calc. for C11H10O2 [M]: 174.0681, found: 174.0672.
Off-Column Diels-Alder-Reaction A mixture of (E)-1-Methoxy-3-trimethylsilyloxy-1,3-butadiene (Danishefsky’s diene, 52.2 mg, 0.34 mmol), freshly distilled benzaldehyde (28.9 mg, 0.29 mmol), and tetradecene (7.68 μmol) in pentane (1.5 mL) was added to the catalyst (5 mol%, Eu(hfc)3: 18.5 mg, Gd(hfc)3: 18.6 mg). The flask was sealed and stirred under the given conditions. For the GC analysis, 0.10 mL was taken from the flask, filtered, and mixed with TFA (50% in DCM, 25.0 μL) and diluted with pentane. For quantitative analysis the free induction decay (FID) signal was used and to calculate conversions the integral was corrected by the number of carbon atoms.
On-Column Diels-Alder-Reaction (ocRGC) For the on-column measurements freshly distilled benzaldehyde (1 eq), (E)-1-Methoxy-3-trimethylsilyloxy-1,3-butadiene (Danishefsky’s diene, 0.9 eq) and tetradecene were mixed in 1 mL pentane. At regular intervals the sample was checked with a reference column (Chirasil-β-Dex, 17 m, 250 nm, 250 μm I.D.) to exclude changes or decomposition of the sample. The reaction mixture was injected with an on-column syringe from SGE following the general on-column cold injection techniques with a secondary N2-cooling time of 2 min. The cryogenic CO2 cooling was activated and deactivated according the measurement process outlined in Figure 2 or mentioned in the text. For quantitative analysis the FID signal was used and to calculate conversions the integral was corrected by the number of carbon atoms. For the unambiguous assignment of the formed hetero-Diels-Alder product, pyron 5 was dissolved in n-pentane and injected via split injection onto the metal(hfc) column (25 m, 250 nm, 250 μm I.D.) without cryogenic cooling or additional separation column.
RESULTS AND DISCUSSION
To extend the scope of ocRGC of higher-order reactions to asymmetric catalysis, we shifted our attention to the Danishefsky-hetero-Diels-Alder-reaction. In a prescreening we found some interesting activity of gadolinium(III)-tris [(1R,4S)-3-heptafluorobutanoyl-camphor] (Gd(hfc)3) complexes as chiral Lewis acid catalyst (Scheme 1). To perform more systematic studies we developed a strategy to immobilize Gd(hfc)3 1 to polysiloxane via a propyleneoxy linker attached to C10 in the camphor moiety, resulting in gadolinium (III)-tris[(1R,4S)-3-heptafluorobutanoyl-10-propyleneoxycamphor]polysiloxane (Gd(hfpc)3@PS) 2. This polymeric catalyst allows fabricating reactor capillaries for ocRGC that enables a fast and efficient screening of reactions of higher-order like the Diels-Alder-reaction where a diene and dienophile reacts to form six-membered rings. For this purpose Gd (hfc)3@PS 2 (Fig. 1), a polysiloxan bonded version of Gd (hfc)3 1, was synthesized with 20% selector content and coated onto the inner surface of fused-silica capillaries (0.25 mm I.D.) according to the static method described by Grob, resulting in a defined polymer film thickness of
Scheme 1. Danishefsky-hetero-Diels-Alder-reaction of (E)-1-methoxy-3trimethylsilyloxy-1,3-butadiene (Danishefsky’s diene, 3) with benzaldehyde 4 forming the chiral pyron 5 after simultaneous elimination of trimethylsilanol and methanol by traces of water at these low substrate concentrations.
ENANTIOSELECTIVE ON-COLUMN REACTION GC
Fig. 1. Gadolinium(III)-tris[(1R,4S)-3-heptafluorobutanoyl-camphor] (Gd(hfc)3, 1) and gadolinium(III)-tris[(1R,4S)-3-heptafluorobutanoyl-10-propylenoxycamphor]polysiloxane (Gd(hfc)3@PS, 2).
250 nm (Gd(hfc)3@PS, 2).33 Additionally, the [(1R,4S)-3heptafluorobutanoyl-10-propyleneoxycamphor]-polysiloxane (hfc@PS) complexes of nickel(II), platinum(II), palladium (II), lanthanum(III), europium(III), and manganese(II) were synthesized and tested following the same procedure and conditions as for Gd(hfc)3@PS 2. Remarkably, Gd(hfc)3@PS 2 was the only catalytically active column of this library of metal complexes. In combination with the other mentioned
245
metal(hfc) columns, the injected reactant mixture could be detected without any pyron 5 formation, which shows the great potential of the immobilized catalyst Gd(hfc)3@PS 2. To make analysis and screening of higher-order reactions like the Diels-Alder-reaction feasible, the normal reaction GC setup had to be modified to increase the contact time Δt of the reactants and to make refocusing within a run possible, which not only helps to improve the conversion of the components, but also provides a means to improve sensitivity by focusing the peaks, which is necessary to precisely determine enantiomeric ratios. We found that the immobilized catalyst Gd(hfc)3@PS is highly catalytically active and the reactivity is sufficient in the ocRGC setup so that even by simple injection of the reactant mixture the desired pyron 5 is formed; however, interaction of the polymeric catalyst is so strong that only very broad peaks are observed, which do not allow quantifying the enantiomeric ratio. Hence, a cryogenic CO2 cold trap was installed in the GC oven to focus the formed pyron 5. Cryogenic focusing is a popular method to improve peak shapes or to concentrate substances above the limit of detection and was very recently proven to be a promising analytical device to improve detection of higher-order reactions in ocRGC systems by our research group.28 For this application a cryogenic CO2 cold trap was installed in the first part of the separation column (Fig. 2a). As separation column, a Chirasil-β-Dex42,43 fusedsilica column was used, which provided an excellent separation of the product enantiomers 5. An installation of the cryogenic focusing trap on the last part of the catalytically
Fig. 2. Experimental setup to perform enantioselective on-column reaction gas chromatography (a), the process protocol to perform reactions using cryogenic focusing (b), and schedule of the cryogenic cooling steps during an ocRGC experiment (c). Experimental conditions: On-column cold injection, 120°C, 10 kPa for 1 min, then 100 kPa/min to 100 kPa; He as inert carrier gas; catalytically active column: Gd(hfc)3@PS, 5 m, 250 nm film thickness, 250 μm ID; separation column: Chirasil-β-Dex, 10 m, 250 nm film thickness, 250 μm ID. Chirality DOI 10.1002/chir
246
STOCKINGER AND TRAPP
active fused-silica column was also tested but there were no differences detectable compared to the presented setup. To achieve an efficient separation of the formed product enantiomers the cryogenic focusing should only affect the formed pyron 5, and therefore the cooling was activated after 2.4 min of the injection of the reactants. As outlined in the process protocol (Fig. 2c), at this point the remaining reactants as well as the solvent and the internal standard already left the catalytically active column and some of these components already reached the detector. After 6 min the formed pyron 5 was fully focused by the cryogenic trap, the cooling was turned off, and the focused reaction product was released onto the separation column. For an unambiguous assignment of the formed heteroDiels-Alder product, pyron 5 was synthesized and measured onto a 25 m Gd(hfc)3@PS 2 column without using cryogenic cooling. The retention time of the reference sample was about 36 min compared to the retention time of the on-column synthesized product, which was already eluted after about 32 min. This difference in the retention times of about 4 min indicates the formation of the reaction product on the catalytically active column in the ocRGC experiment. Furthermore, without a catalytic column there was no reaction detectable under the same experimental conditions. The optimum length of the catalytically active Gd(hfc)3@PS 2 column was determined to be 5 m, which gives decreased measurement times and therefore higher throughput under maintained reactivity. In Figure 3 a resulting chromatogram of an ocRGC experiment of the described Danishefsky-hetero-Diels-Alder reaction using a 5 m catalytically active Gd(hfc)3@PS 2 column and a 10 m Chirasil-β-Dex separation column is depicted. Due to the high catalytic activity of the Gd(hfc)3@PS 2 column it was not possible to detect remaining (E)-1-methoxy-3trimethylsilyloxy-1,3-butadiene (Danishefsky’s diene 4), but only the transformed (E)-4-methoxybut-3-en-2-one 4* which was not separated from benzaldehyde 3 under the here applied experimental conditions. The enantiomers of pyron 5 could be baseline separated in the presented setup. The elution order of the enantiomers ((+)-5 before (–)-5) was assigned as described in the literature.4 With the here presented experimental on-column reaction chromatographic setup an enantiomeric ratio (er) of (+)-5 to (–)-5 of 44:56 corresponding to an enantiomeric excess (ee) of (–)-5 of 12% could be determined. An average standard deviation of 0.6% for the er values and 1.1% for the ee value could be
determined. Keeping in mind that the reaction time by the ocRGC experiment of the here described second-order reaction is well below 1 s and therefore also the conversion is low, it is quite remarkable that the er is detectable with high precision. In the next step we compared the results obtained by ocRGC with the results obtained by the conventional batch reaction. Of course, the conversions can be simply optimized by extending the reaction time; however, the er could differ because of different catalyst and substrate concentrations. For comparison, the molecular Gd(hfc)3 catalyst was synthesized and catalysis was performed in a conventional flask. For comparison, we also performed catalytic tests with the commercially available Eu(hfc)3 catalyst. Besides the use of Eu(hfc)3 in various applications, it is probably the best known catalyst for these types of Diels-Alder-reactions.2,4 In Table 1 the results of the catalytic experiments are summarized. Gd(hfc)3 1 shows approximately a similar reaction time compared to Eu(hfc)3 (at 20°C both require 1 day for completion of the reaction, at –30°C Gd(hfc)3 3 days and Eu(hfc)3 2 days, respectively). The er and ee catalyzed by Gd(hfc)3 1 are throughout these measurements slightly above the values obtained by using Eu(hfc)3. For example, Gd(hfc)3 1 gives an er of 33:67, whereas Eu(hfc)3 provides only a ratio of 35:65 at 20°C and at –30°C the difference exceeds 34:66 with Eu (hfc)3 and 30:70 with Gd(hfc)3 1. Further improvements of this reaction of Gd(hfc)3 1 to an er of 27:73 at –30°C could be achieved by adding an excess of the ligand (hfc) and TABLE 1. Enantiomeric ratio er of (+)-5 and (-)-5 and enantiomeric excess ee of (-)-5 obtained by conventional reaction conditions in a glass flask using molecular catalysts 1 and Eu(hfc)3 Catalyst
T [°C]
Reaction completed after
er [%]
ee [%]
Eu(hfc)3
20 -30 20 0 -30 -70
1d 2d 1d 1.5d 3d 3.5d
35:65 34:66 33:67 31:69 30:70 31:69
30% 32% 35% 38% 40% 38%
Gd(hfc)3
The reported values represent averages obtained from at least three GC measurements. The average standard deviation does not exceed ±1% for the er values and ±2% for the ee values.
Fig. 3. Chromatogram of a Gd(hfc)3@PS 2 catalyzed ocRGC Danishefsky-hetero-Diels-Alder-reaction using the cryogenic focusing setup outlined in Figure 2. Chirality DOI 10.1002/chir
ENANTIOSELECTIVE ON-COLUMN REACTION GC
increasing the amount of catalyst to 20 mol%, which also reduces the reaction time considerably. Surprisingly, the here measured er’s were lower than reported in the literature (ee’s of up to 42% at –33°C) by other groups. The only difference might be that we directly analyzed the enantiomeric composition using the crude reaction mixture and other groups perform purification by flash chromatography. It is well known that nonlinear effects might cause deviations from the original enantiomeric composition, leading to an enrichment of the major enantiomer. Nevertheless, the obtained results are comparable with respect to conversions under these reaction conditions. Compared to Eu(hfc)3, our findings indicate that Gd(hfc)3 1 gives slightly higher enantiomeric selectivities combined with an almost similar reactivity. CONCLUSION
In this report we demonstrated that enantioselective ocRGC of higher-order reactions as shown for the here-investigated Danishefsky-hetero-Diels-Alder reaction is feasible and it is a versatile tool to rapidly assess the enantioselectivity of catalyzed reactions. To improve enantioselective analysis cryogenic focusing by CO2 in a cold trap on the first part of the separation column was used. Cryogenic focusing overcomes peak broadening effects caused by strong interaction effects of the reaction product with the catalytically active stationary phase. As the catalytically active stationary phase used in the first section of the column setup of the ocRGC experiment, Gd(hfc)3 was immobilized by modification of the camphor moiety with a propyleneoxy linker at C10 and hydrosilylation to polysiloxane. It has to be pointed out that the enantioselectivity determined by the rapid screening approach correctly predicted the enantioselectivity of the reaction performed with the molecular catalyst in a conventional flask. Furthermore, the enantioselectivities, which are in general moderate for this type of reaction, are slightly better compared to the standard catalyst Eu(hfc)3. ACKNOWLEDGMENTS
We thank the Deutsche Forschungsgemeinschaft (DFG) for financial support of this research (SFB 623 ‘Molecular Catalysts: Structure and Functional Design’). O.T. was supported by the European Research Council under Grant Agreements No. StG 258740. We thank Chromasoft GmbH for the supply of the cryogenic CO2 cold trap system. LITERATURE CITED 1. Yates P, Eaton P. Acceleration of the Diels-Alder Reaction by Aluminum chloride. J Am Chem Soc 1960;82:4436–4437. 2. Fringuelli F, Taticchi A. The Diels-Alder reaction: selected practical methods. New York: John Wiley & Sons; 2002. 3. Bednarski M, Maring C, Danishefsky S. Chiral induction in the cyclocondensation of aldehydes with siloxydienes. Tetrahedron Lett 1983;24:3451–3454. 4. Keller F, Weinmann H, Schurig V. Chiral polysiloxane-fixed metal 1,3diketonates (chirasil-metals) as catalytic Lewis acids for a hetero Diels-Alder reaction — inversion of enantioselectivity upon catalyst – polymer binding. Chem Ber 1997;130:879–885. 5. Togni A. Asymmetric hetero Diels-Alder reactions catalyzed by novel chiral vanadium( IV) bis( 1,3-diketonato) complexes. Organometallics 1990;9:3106–3113. 6. Maruoka K, Oishi M, Yamamoto H. Methylaluminum bis(4-substituted-2,6di-ter-butylphenoxide) as an efficient non-chelating Lewis acid. Synlett 1993;9:683–685.
247
7. Smith MB, March J. March’s advanced organic chemistry. Hoboken, NJ: John Wiley & Sons; 2007. 8. Evans DA, Wu J. Enantioselective rare-earth catalyzed quinone Diels Alder reactions. J Am Chem Soc 2003;125:10162–10163. 9. Maruoka K, Itoh T, Shirasaka T, Yamamoto H. Asymmetric heteroDiels-Alder reaction catalyzed by a chiral organoaluminum reagent. J Am Chem Soc 1988;110:310–312. 10. Faller JW, Smart CJ. Use of organometallic complexes of ruthenium in the Lewis acid catalyzed hetero Diels-Alder reaction. Tetrahedron Lett 1989;30:1189–1192. 11. Ishitani H, Kobayashi S. Catalytic asymmetric aza Diels-Alder reactions using a chiral lanthanide Lewis acid. Enantioselective synthesis of tetrahydroquinoline derivatives using a catalytic amount of a chiral source. Tetrahedron Lett 1996;37:7357–7360. 12. Danishefsky S, Bednarski M. Lanthanide catalysis of cycloadditions of heterodienes with enol ethers. Tetrahedron Lett 1984;25:721–724. 13. Danishefsky S, Maring C. A new approach to the synthesis of hexoses: an entry to (+)-fucose and (+)-daunosamine. J Am Chem Soc 1985;107:1269–1274. 14. Gerstberger G, Palm C, Anwander R. The Danishefsky hetero Diels-Alder reaction mediated by organolanthanide-modified mesoporous silicate MCM-41. Chem Eur J 1999;5:997–1005. 15. Hesse M, Meier H, Zeeh B. Spektroskopische Methoden in der organischen Chemie. Berlin: Georg Thieme; 2005. 16. Schurig V, Bürkle W. Extending the scope of enantiomer resolution by complexation gas chromatography. J Am Chem Soc 1982;104:7573–7580. 17. Spallek MJ, Storch G, Trapp O. Straightforward synthesis of poly (dimethylsiloxane) phases with immobilized (1R)-3-(perfluoroalkanoyl) camphorate metal complexes and their application in enantioselective complexation gas chromatography. Eur J Organ Chem 2012;2012:3929–3945. 18. Morill TC. Lanthanide shift reagents in stereochemical analysis. Rochester, NY: VCH Publisher; 1986. 19. Frullano L, Meade TJ. Multimodal MRI contrast agents. J Biol Inorg Chem 2007;12:939–949. 20. Ma Y, Qian C, Xie M, Sun J. Lanthanide chloride catalyzed imino Diels Alder reaction. one-pot synthesis of pyrano[3,2-c]- and furo[3,2-c]quinolines. J Organ Chem 1999;64:6462–6467. 21. Yu Y, Zhou J, Yao Z, Xu F, Shen Q. Stereoselective synthesis of pyrano [3,2-c]- and furano[3,2-c]quinolines: gadolinium chloride catalyzed onepot aza–Diels–Alder reactions. Heteroatom Chem 2010;21:351–354. 22. Trapp O. Unified equation for access to rate constants of first-order reactions in dynamic and on-column reaction chromatography. Anal Chem 2006;78:189–198. 23. Trapp O, Weber SK, Bauch S, Hofstadt W. High-throughput screening of catalysts by combining reaction and analysis. Angew Chem Int Ed 2007;46:7307–7310; Angew Chem 2007;119:7447–7451. 24. Trapp O, Weber SK, Bauch S, Bäcker T, Hofstadt W, Spliethoff B. High throughput kinetic study of hydrogenations over palladium nanoparticles — combination of reaction and analysis. Chem Eur J 2008;14:4657–4666. 25. Trapp O. Investigation of the stereodynamics of molecules and catalyzed reactions by CE. Electrophoresis 2010;31:786–813. 26. Gil-Av E, Herzberg-Minzly Y. Gas-chromatographic study of the rate of Diels-Alder addition. Proc Chem Soc 1961;316. 27. Boelens HFM, Iron D, Westerhuis JA, Rothenberg G. Tracking chemical kinetics in high-throughput systems. Chem Eur J 2003;9:3876–3881. 28. Becker S, Höbenreich H, Vogel A, Knorr J, Wilhelm S, Rosenau F, Jaeger K-E, Reetz MT, Kolmar H. Single-cell high-throughput screening to identify enantioselective hydrolytic enzymes. Angew Chem Int Ed 2008;47:5085–5088; Angew Chem 2008;120:5163–5166. 29. de Bellefon C, Tanchoux N, Caravieilhes S, Grenouillet P, Hessel V. Microreactors for dynamic, high throughput screening of fluid/liquid molecular catalysis. Angew Chem Int Ed 2000;39:3442–3445; Angew Chem 2000;112(19):3584–3587. 30. Kluwer AM, Detz RJ, Abiri Z, van der Burg AM, Reek JNH. Evolutionary catalyst screening: iridium-catalyzed imine hydrogenation. Adv Synth Catal 2012;354:89–95. 31. Krabbe SW, Hatcher MA, Bowman RK, Mitchell MB, McClure MS, Johnson JS. Copper-catalyzed asymmetric hydrogenation of aryl and heteroaryl ketones. Organ Lett 2013;15:4560–4563. 32. Cai C, Chung JYL, McWilliams JC, Sun Y, Shultz CS, Palucki M. From high-throughput catalyst screening to reaction optimization: detailed Chirality DOI 10.1002/chir
248
33.
34. 35.
36.
37.
38.
STOCKINGER AND TRAPP
investigation of regioselective Suzuki coupling of 1,6-naphthyridone dichloride. Organ Proc Res Dev 2007;11:328–335. Lefort L, Boogers JF, deVries AM, deVries J. High throughput screening of Monophos instant ligand library leads to a ton-scale asymmetric hydrogenation process. Top Catal 2006;40:185–191. Trapp O. High-throughput monitoring of interconverting stereoisomers and catalytic reactions. Chem Today 2008;26:26–28. Troendlin J, Rehbein J, Hiersemann M, Trapp O. Integration of catalysis and analysis is the key: rapid and precise investigation of the catalytic asymmetric Gosteli–Claisen rearrangement. J Am Chem Soc 2011;133:16444–16450. Stockinger S, Trapp O. Integrating reaction and analysis: investigation of higher-order reactions by cryogenic trapping. Beilstein J Organ Chem 2013;9:1837–1842. Lang C, Gartner U, Trapp O. Catalysts by the meter: rapid screening approach of N-heterocyclic carbene ligand based catalysts. Chem Commun 2011;47:391–393. Weber SK, Bremer S, Trapp O. Integration of reaction and separation in a micro-capillary column reactor—palladium nanoparticle catalyzed C–C bond forming reactions. Chem Eng Sci 2010;65:2410–2416.
Chirality DOI 10.1002/chir
39. Sandel S, Weber SK, Trapp O. Oxidations with bonded salen-catalysts in micro capillaries. Chem Eng Sci 2012;83:171–179. 40. Fulmer GR, Miller AJM, Sherden NH, Gottlieb HE, Nudelman A, Stoltz BM, Bercaw JE, Goldberg KI. NMR chemical shifts of trace impurities: common laboratory solvents, organics, and gases in deuterated solvents relevant to the organometallic chemist. Organometallics 2010;29:2176–2179. 41. Grob K. Making and manipulating capillary columns for gas chromatography. Heidelberg, Basel, New York: Dr. Alfred Hüthig Verlag; 1986. 42. Schurig V, Schmalzing D, Schleimer M. Enantiomer separation on immobilized Chirasil-Metal and Chirasil-Dex by gas chromatography and supercritical fluid chromatography. Angew Chem Int Ed 1991;30:987–989; Angew Chem 1991;103:994–996. 43. Cousin H, Trapp O, Peulon-Agasse V, Pannecoucke X, Banspach L, Trapp G, Jiang Z, Combret JC, Schurig V. Synthesis, NMR characterisation and polysiloxane-based immobilization of the three regioisomeric monooctenylpermethyl-β-cyclodextrins - application in enantioselective GC. Eur J Org Chem 2003;3273–3287.
