Rapid catalyst evaluation for Sonogashira coupling in continuous flow

Selection of the most appropriate catalyst for Pd-catalyzed Sonogashira coupling often requires time and resources. We applied a continuous flow device for rapid catalyst screening of 4 heterogeneous catalysts over 6 Sonogashira coupling reactions together with longevity estimation. The screening procedure requires only 8 hours. For coupling the alkynes with 4-iodo-anisole Fibrecat types of catalysts showed the highest performance, while when 4-bromo-anisole was applied as an aromatic halide PdCl 2 (PPh 3 ) 2 DVB gave superior results. For phenyl-acetylene and ethynyl-cyclohexene the coupling of the bromo derivative resulted in relatively low conversion as well as selectivity for most of the polymer-bound catalysts. Applying the cheap catalyst 10 % Pd/C the coupling reaction could only be achieved with the rather reactive 4-iodo-anisole although with moderate conversion.


Introduction
Pd-catalyzed Sonogashira coupling is currently the most practical method for synthesizing aryl-(vinyl) acetylenes from the corresponding terminal alkynes and aryl halides.The reaction was first reported by Kenkichi Sonogashira and Nobue Hagihara in 1975. 1 Since then numerous variations of the reaction were reported 2 under traditional batch conditions including copper-free applications that reduced the dimerization (self-coupling) of the acetylenic species (Glaser coupling).Over 100 different conditions were applied for the synthesis of 1-methoxy-4-(2phenylethynyl)benzene from 4-iodo-anisole and phenyl-acethylene.The different users varied the catalyst (heterogeneous or homogeneous Pd species with or without copper and/or phosphine ligands) as well as other (post)transition metal catalysts such as Fe, Sm, In), inorganic or organic bases, various protic solvents including water were applied.The reaction can be carried out from r.t. to 175 °C, although heating reduces the reaction time significantly.MW irradiation is particularly efficient, Awuah reported a homogeneous phase Pd catalyzed copper-free procedure under less than an hour (t = 0.5h, T= 90 °C). 3 Pure microwave-assisted Sonogashira reaction was also attempted without any catalysts but the yield was significantly lower than the catalytic process (tetrabutylammomium bromide, sodium carbonate in water, Time= 0.25h, T= 175 °C). 4,5nterestingly Pd-free copper catalyzed reaction was also reported with excellent yield using tetrabutylammomium bromide, potassium carbonate, triphenylphosphine and copper(l) iodide in water: reaction time= 0.7 h, T= 120 °C under microwave irradiation. 6ontinuous flow (micro)reactors offer several advantages.They extend the parameter space of chemical reactions significantly (up to 350 °C and 200 bar); increase reaction rates via enhanced heat/mass transfer (flash heating); enhance reproducibility and speed up optimization of reaction parameters due to fast and accurate changing of temperature and pressure parameters; prevent by-product formation by moving the reagents away from the reaction zone right after they form the desired product; allow the control of the selectivity through fine tuning the residence time.Flow reactors offer a viable alternative solution to microwave-assisted batch chemistry. 7Various microfluidic/flow or microreactor approaches were reported in the recent years for Sonogashira coupling.For example the reaction was performed in an ionic liquid ([BMIm][PF6]) in the absence of a copper salt using PdCl2(PPh3)2. 8Kawanami applied high T/p water containing sodium hydroxide and palladium dichloride in a homogeneous microfluidic system (with fluid channels typically in the submillimeter range) applying residence time= 0.1-4.0s, T= 250 °C , p= 160 bar. 9 In another homogenous flow approach Sugimoto applied a loop reactor (1000 μm i.d., 10 m length) mixing the alkyl halide and the alkyne in DMF and using PdCl 2 (PPh 3 ) 2 (1 mol %), CuBr (2 mol %), and diisopropylethylamine as a base at 120 °C. 10 The application of heterogeneous catalysis in conjunction with microreactor technology can facilitate a cleaner/green and scalable flow methodology for coupling reactions.As the most simple solution microtubular ("loop") reactors were coated with thin catalytic metal inner surface (Pd or Pd− Cu alloy) and applied in high pressure and high-temperature water (HPHT− H2O) in flow.Under the optimized reaction conditions of 250 °C and 160 bar of pressure the coupling reaction was carried out in good yield and 100% selectivity within a very short residence time of ~1.6 s. 11 More frequently immobilized/heterogenized catalysts 12 have been utilized to promote a range of cross-coupling reactions including Heck, Sonogashira, Suzuki, etc.. 13,14 Transition metal catalysts such as palladium, copper, ruthenium, and nickel are described on silica, monolithic, magnetic nanoparticles and polymer supports. 15Flow-through microreactors using poly(chloromethylstyrene-co-divinylbenzene) monolith in capillaries of internal diameter 250 um were used successfully for Suzuki-Miyaura and Sonogashira reactions, with very low palladium leaching. 16e applied immobilized catalysts in a fixed-bed flow reactor developed for heterogeneous flow hydrogenation, H-Cube. 17The H-Cube system is also suitable for any heterogeneous catalytic reactions including cross coupling using immobilized transition metal catalysts in prepacked-columns " catalyst cartridges" (CatCart).The CatCart system was designed to make the use of pyrophoric catalysts safer and easier to handle, and reusable with different substrates. 18he CatCart system has number of advantages: it has a filter at each end which means the catalyst stays in the CatCart, therefore, no filtering is needed and the catalyst can be reused.Furthermore, the catalyst handling is easy because no weighing out or disposal are required.For mg to g scale operation 70 mm standard catalyst cartridges are applied and we have chosen this size in the current study.The void volume for 70 mm cartridges is between 0.5-0.8mL (for 10% Pd/C: 0.669 mL), thus, the estimated residence time at 0.1 mL/min flow rates are between 5-8 min.Such short residence time means that reactions are less likely to react again with themselves and form side products.In the CatCart system the catalyst should have a greater longevity since any product which may poison the catalyst is removed continuously from the catalyst.In a batch reactor, the product remains with the catalyst and can lead to deactivation.The objective of our study was to determine which commercial immobilized catalysts perform best in the CatCart system for Sonogashira reaction.We tested 3 polymer-bound catalysts and in addition 10 % Pd/C (Table 1

Identification of the key parameters for catalyst testing
For setting the parameters for catalyst evaluation we have chosen the following model reaction (Scheme 1) applying 1.2 eq.alkyne (1), 1 eq.aromatic halide, 3 eq.sodium hydroxide as a base and methanol as a solvent (c=0.05M to the aromatic halide).The optimal concentration was found as 0.05M, more concentrated solution led to incomplete reaction (decreased conversion).

Scheme 1.
Model reaction for the identification of the parameter set for catalyst testing.In order to save the large cartridges the parameter "scanning" was started with standard 30 mm cartridges.As a result 0.1 mL/min flow rate, 100 bar and 100 ºC was selected.During the reactions no acetylene dimerization was observed, which was expected in copper-free conditions.The definition of selectivity is widely used in the catalytic science as the following: Si = ni/nk * 100, where i is the target product in a mixture of k products, excluding the reagent residues.

Investigation of catalyst longevity
Catalyst longevity is one of the key concerns in any catalytic processes.We investigated one of the best performing catalysts (PdCl2(PPh3)2, DVB) in 4 reactions (Entry 2,3,4,5) for repeated performance.We collected 4 mL samples (Fractions 1-5) in continuous operation and the samples were analyzed by LC-MS for conversion.As it is shown in Table 5. the longevity is strongly dependent on the structures.Interestingly for Entry 2 the conversion dropped very rapidly.Since as it was previously shown the selectivity was rather poor for PdCl2(PPh3)2, DVB in Entry 2, which might be due to significant leaching of the catalyst and the polymers in the presence of phenyl-acetylene that led to unidentified impurities (see also under Conclusion).
Similarly in Entry 5 where the coupling agent was again phenyl-acetylene the conversion also dropped but the starting value was also rather poor.In the other two cases the stability and the performance was acceptable for this polymer-bound catalyst.

Conclusions
The present study aimed at demonstrating the potential of flow technology in real-time catalyst screening.Table 6 shows the best performing catalysts for each reaction (Entry 1-6).For coupling the alkynes with 4-iodo-anisole the Fibrecat types of catalysts showed the highest performance if both the conversion and the selectivity were taken into consideration.For 4-iodo-anisole PdCl2(PPh3)2, DVB showed higher conversion but LC-MS showed significant amount of unknown impurities, thus, the selectivity was lower particularly for Entry 1 (51%).On the other hand when 4-bromo-anisole was applied as the aromatic halide coupling reagent PdCl2(PPh3)2, DVB gave superior results including both conversion and selectivity.For phenylacetylene and ethynyl-cyclohexene the coupling of the bromo species resulted in low conversion as well as selectivity for most of the polymer-bound catalysts but Fibrecat 1001 showed relatively good selectivity for ethynyl-cyclohexene.This study confirms that catalyst evaluation using flow devices allows a rapid evaluation and selection of the most appropriate catalysts and since the performance is strongly dependent on nature of reactants.For the cheap and simple catalyst, 10% Pd/C coupling reaction was only achieved with the more reactive 4-iodo-anisole (highest conversion: 53%, Entry 3), while with 4-bromo-anisole essentially no coupling reaction was observed.
In conclusion, if considering that each catalyst evaluating reaction lasts approx.10 min plus another approx.10 min for washing the system, for completing such reaction matrix of 24 reactions (6 reactions x 4 catalyts) it requires 8 hours.Performing the HPLC evaluation can be done in parallel.Thus, the striking advantage of continuous flow devices is the rapid parameter optimization as well as rapid screening to find the best catalyst for a particular reaction.Performing catalyst screening in flow can be automated using pre-programmed catalyst cartridge changing (e.g.CatCart Changer 22 ) and a fraction collector, thus, the evaluation experiments can be carried out overnight and unattended.

Experimental Section
General.Sonogashira coupling was performed using H-Cube fixed-bed flow reactor in "no gas" mode. 23The conversion and selectivity was determined by HPLC-MS comparing the area percentage with calibration results obtained at various concentrations.For LC-MS experiments Agilent 1100 liquid chromatograph was linked with Waters Micromass ZQ mass spectrometer using atmospheric-pressure chemical ionization (APCI) technique. 1H NMR spectra were recorded on Bruker Avance II 400 MHz.

Table 1 .
.) .List and properties of commercial immobilized catalysts involved in the study

Table 2 .
Parameters investigated for finding the optimal parameter set (flow rate, T, p) for the model reaction

Table 3 .
The reaction matrix applied in the catalyst screening study O 8

Table 4 .
The performance of the 4 catalysts in the 6 coupling reactions between terminal alkynes and aryl halides leading to aryl-(vinyl) acetylenes(6-8)

Table 5 .
Conversion rates of the coupling reaction of various reactants under continuous operation in the subsequently collected fractions

Table 6 .
Summary table of the best performing catalysts based on the results obtained in the test reactions