Microwave-Assisted Regioselective Suzuki Coupling of 2,4-Dichloropyrimidines with Aryl and Heteroaryl Boronic Acids

: Suzuki coupling reaction has been often used for the preparation of a diverse set of substituted pyrimidines. In this study, the Suzuki coupling of 2,4-dichloropyrimidines with aryl and heteroaryl boronic acids was investigated. A thorough screening of reaction conditions and the use of microwave irradiation led to a very efﬁcient and straightforward synthetic procedure providing C4-substituted pyrimidines in good to excellent yields. Short reaction time (15 min) and extremely low catalyst loading (0.5 mol%) are the main advantages of our tetrakis(triphenylphosphine)palladium(0) catalyzed microwave-assisted procedure, which could be used for quick and low-cost regioselective preparation of substituted pyrimidine rings. Contributions: Conceptualization, A.D. M.S.; methodology, A.D., K.M., M.S.; vali-dation, A.D. and K.M.; A.D., M.S.; writing—original preparation, writing—


Introduction
Substituted pyrimidine rings as scaffolds are of great interest for medicinal chemists being a part of many biologically active compounds [1,2]. The pyrimidine moiety is present in many natural compounds [3] (e.g., nucleic acids, alkaloids, folic acid, etc.) as well as synthetic analogs and also in approved drugs on the market, e.g., anticancer [4][5][6], antiviral [4,7], antibacterial [8], antilipidemic [9], anti-inflammatory [3,10], and antimalarial agents ( Figure 1). Large number of synthetic methods have been described for the preparation of pyrimidine-based compounds [11][12][13][14][15][16]. Cyclocondensation between guanidine, amidine or thiourea derivatives and 1,3-diketones or 1,3-diesters is the most classical method for the synthesis of the main pyrimidine core [15,17], whereas one of the approaches to prepare substituted pyrimidine rings is via halogenated pyrimidines [18][19][20], which are greatly commercially available. The most common reactions involving various halogenated pyrimidines are cross-coupling reactions since pyrimidine ring is an electron-deficient aromatic system being far more reactive in comparison with analogous benzene halides [17,20]. Halogenated pyrimidines are thus very convenient substrates for substantial assortment of nucleophilic aromatic substitutions. Suzuki coupling of halogenated pyrimidines with boronic acids has been a commonly used approach for the preparation of a diverse set of substituted pyrimidines [20].
The aim of our study was to develop a quick, efficient, and regioselective synthetic procedure for substituted pyrimidines at position C4 from readily available 2,4-dichloropyrimidines via Suzuki coupling with a diverse set of aryl and heteroaryl boronic acids. The selection of appropriate solvents, catalysts, and reactions conditions (temperature, time) was systemically performed. Last, but not least, the optimal procedure was also transferred to a microwave reactor and further optimization (i.e., temperature, reaction time catalyst loading) was carried out.

Results
One of the most straightforward approaches to prepare substituted pyrimidines is via Suzuki coupling of halogenated pyrimidines with boronic acids [20,21]. There have been a variety of studies performed in order to propose the reaction mechanism underlaying the Suzuki-Miyaura coupling, consisting of several key steps. The initial pre-catalyst activation allows the formation of Pd(0) species. In the second step, during oxidative addition of aryl halide to the palladium center, the insertion of metal atom into the Cipso and halide (X) bond occurs and the complex [aryl-Pd(II)(Ln)-X] forms, where L represents the potentially bound ligand to the palladium centre, with n ranging from 1 to 4 [28]. Then the oxygen-containing nucleophile (either hydroxide from base followed by addition of boronic acid RB(OH)2, or boronate RB(OH)3 -) replaces the halide group at palladium, which is later followed by transmetalation leading to the formation of aryl-Pd(II)(Ln)-R species. The final step includes reductive elimination of a product aryl-R with newly formed C-C bond, and regeneration of Pd(II) to Pd(0) [29]. In this study, Suzuki cross- Representative drugs with pyrimidine moiety on the market: anticancer (e.g., imatinib [6]), antiviral (e.g., zidovudine [7]), antibacterial (e.g., trimethoprim [8]), antihyperlipidemic (e.g., rosuvastatin [9,27]), antihypertensive (e.g., minoxidil), and antimalarial (e.g., pyrimethamine).
The aim of our study was to develop a quick, efficient, and regioselective synthetic procedure for substituted pyrimidines at position C4 from readily available 2,4dichloropyrimidines via Suzuki coupling with a diverse set of aryl and heteroaryl boronic acids. The selection of appropriate solvents, catalysts, and reactions conditions (temperature, time) was systemically performed. Last, but not least, the optimal procedure was also transferred to a microwave reactor and further optimization (i.e., temperature, reaction time catalyst loading) was carried out.

Results
One of the most straightforward approaches to prepare substituted pyrimidines is via Suzuki coupling of halogenated pyrimidines with boronic acids [20,21]. There have been a variety of studies performed in order to propose the reaction mechanism underlaying the Suzuki-Miyaura coupling, consisting of several key steps. The initial pre-catalyst activation allows the formation of Pd(0) species. In the second step, during oxidative addition of aryl halide to the palladium center, the insertion of metal atom into the C ipso and halide (X) bond occurs and the complex [aryl-Pd(II)(L n )-X] forms, where L represents the potentially bound ligand to the palladium centre, with n ranging from 1 to 4 [28]. Then the oxygencontaining nucleophile (either hydroxide from base followed by addition of boronic acid RB(OH) 2 , or boronate RB(OH) 3 -) replaces the halide group at palladium, which is later followed by transmetalation leading to the formation of aryl-Pd(II)(L n )-R species. The final step includes reductive elimination of a product aryl-R with newly formed C-C bond, and regeneration of Pd(II) to Pd(0) [29]. In this study, Suzuki cross-coupling reaction between commercially available 2,4-dichloropyrimidine (1) and phenylboronic acid (2) was used as a model reaction to select appropriate catalyst, catalyst loading, solvent(s), and optimal coupling reaction between commercially available 2,4-dichloropyrimidine (1) and phenylboronic acid (2) was used as a model reaction to select appropriate catalyst, catalyst loading, solvent(s), and optimal reactions conditions (temperature, time). Our first attempt was to systemically screen the most common solvents used in Suzuki coupling reactions. Based on previous studies [21], tetrakis(triphenylphosphine)palladium(0) (Pd(PPh3)4) along with potassium carbonate K2CO3 as a base was selected as the starting catalytic system (Scheme 1). Scheme 1. Suzuki coupling of 2,4-dichloropyrimidine (1) with phenylboronic acid (2) as our model reaction.
A good selection of appropriate solvent for cross-coupling reactions is imperative. The screening of most common solvents for cross-coupling reactions (i.e., H2O, MeOH, THF, DMF, 1,4-dioxane, isopropanol, ethylene glycol) was performed at a room temperature (r.t.) and higher temperatures (60-100 °C), with yields determined by LC-MS being obtained after 1 h, 2 h, and 24 h (see Supporting Information, Table S1). The highest yields were obtained for less polar solvents, like isopropanol (64%) and 1,4-dioxane (72%), contrary to polar H2O, MeOH, DMF, and ethylene glycol, which corresponds to the observation made by Miyaura about Pd(PPh3)4 [30]. The choice of solvent (therefore its boiling point) defines the maximum operating temperature [31]. On the set of less polar solvents THF, isopropanol, and 1,4-dioxane with boiling points of 66 °C, 83 °C, and 101 °C, respectively, we confirmed that higher temperatures vastly improve the yield of reaction due to the high activation barriers reagents need to overcome for successful coupling. The C2 side product was formed in less than 8%, which is consistent with the literature reports [24].
We decided to repeat the reaction in four best solvents (i.e., THF, DMF, 1,4-dioxane and isopropanol) and isolate the pure product 3 by column chromatography. The isolated yields are presented in Table 1. The highest yield of 71% was obtained for non-polar 1,4dioxane, which has been reported to complement Pd(PPh3)4 [30], thus we decided to proceed with further optimization in this solvent. In a continuation of our study, we screened the most common commercially available palladium catalysts for this type of reaction. Tris(dibenzylideneacetone)palladium(0) (Pd2dba3) was not efficient for this type of reaction as has already been observed [32,33], Scheme 1. Suzuki coupling of 2,4-dichloropyrimidine (1) with phenylboronic acid (2) as our model reaction.
A good selection of appropriate solvent for cross-coupling reactions is imperative. The screening of most common solvents for cross-coupling reactions (i.e., H 2 O, MeOH, THF, DMF, 1,4-dioxane, isopropanol, ethylene glycol) was performed at a room temperature (r.t.) and higher temperatures (60-100 • C), with yields determined by LC-MS being obtained after 1 h, 2 h, and 24 h (see Supporting Information, Table S1). The highest yields were obtained for less polar solvents, like isopropanol (64%) and 1,4-dioxane (72%), contrary to polar H 2 O, MeOH, DMF, and ethylene glycol, which corresponds to the observation made by Miyaura about Pd(PPh 3 ) 4 [30]. The choice of solvent (therefore its boiling point) defines the maximum operating temperature [31]. On the set of less polar solvents THF, isopropanol, and 1,4-dioxane with boiling points of 66 • C, 83 • C, and 101 • C, respectively, we confirmed that higher temperatures vastly improve the yield of reaction due to the high activation barriers reagents need to overcome for successful coupling. The C2 side product was formed in less than 8%, which is consistent with the literature reports [24].
We decided to repeat the reaction in four best solvents (i.e., THF, DMF, 1,4-dioxane and isopropanol) and isolate the pure product 3 by column chromatography. The isolated yields are presented in Table 1. The highest yield of 71% was obtained for non-polar 1,4-dioxane, which has been reported to complement Pd(PPh 3 ) 4 [30], thus we decided to proceed with further optimization in this solvent. In a continuation of our study, we screened the most common commercially available palladium catalysts for this type of reaction. Tris(dibenzylideneacetone)palladium(0) (Pd 2 dba 3 ) was not efficient for this type of reaction as has already been observed [32,33], whereas as an adduct with chloroform (Pd 2 dba 3 · CHCl 3 ) and together with tri-tertbutylphosphonium tetrafluoroborate (TTBP · HBF 4 ) as a ligand it gave 35% and 23% yield, respectively ( Table 2, entries 1-3). With bis(triphenylphosphine)palladium(II) chloride (PdCl 2 (PPh 3 ) 2 ) similarly low yield of 36% was achieved ( namely tricyclohexylphosphine (PCy 3 ), triphenylphosphine (PPh 3 ), and triphenylphosphine bound of divinylbenzene (PPh 3 on DVB), with yields not significantly improved ( Table 2, entries 5-7). According to the mechanism proposed, Pd(II) pre-catalysts have to be reduced to Pd(0) prior to their involvement in oxidative addition. It has been reported that trimeric Pd(OAc) 2 is more susceptible to reduction in its monomeric form, whose formation has been shown to be proportional to the dipole moment of the solvent used [34]. In our screening, 1,4-dioxane with low dipole moment of only 0.45 D was used, which might be the reason for lower yields of Suzuki coupling with Pd(OAc) 2 . Solid-supported catalysts represent an attractive approach in the field of green organic chemistry, since they can often be recycled without loss of activity, in addition to the products and solution waste remaining free from metal contamination [35]. Despite several successful applications in Suzuki-Miyaura coupling [36][37][38], no product was formed when palladium on multiwall carbon nanotubes was used as a catalyst. Additionally, the screening was performed on a set of classic and solid-supported catalysts Pd(OAc) 2 with PPh 3 and Pd(OAc) 2 with PPh 3 on DVB, respectively, with the latter exhibiting inferior catalytic ability and yield in our model reaction. Only one catalyst ([1,1 -Bis(diphenylphosphino)ferrocene]dichloropalladium(II), in complex with dichloromethane (Pd(dppf)Cl 2 · CH 2 Cl 2 ) led to a higher yield of 70% ( Table 2, entry 8), which could be due to its ability (i.e., wide bite-angle P-Pd-P) of driving very effective reductive elimination [39]. Since there was no improvement compared to Pd(PPh 3 ) 4 ( Table 2, entry 9), all further reactions were performed with Pd(PPh 3 ) 4 . One parameter was screened additionally after selection of the appropriate catalyst, i.e., temperature (see Supporting Information, Table S2). Lowering the temperature significantly reduced the yield, thus all further reactions were performed at 100 • C.
After the first optimization process, we decided to further upgrade the reaction procedure with the aim to improve the yields and shorten the reaction time by performing it in a microwave reactor. We started with the solvent scan, where H 2 O due to favorable dielectric constant for microwave heating was added in different proportions to THF or 1,4-dioxane. With this experiment, where total solvent volume was 6 mL, we tried to find the optimal solvent ratio (v/v) to obtain the highest yield of 3 and as little as possible of any side products (Table 3). Generally, the yield of 3 was improved, when H 2 O was present in the reaction mixture, with the peak at solvent ratio 1:1 or 2:1 in the favor of non-polar solvent for THF and 1,4-dioxane, respectively (Table 3, entries 1 and 8). In addition to possessing dipole moment, the presence of water is also beneficial due to increased amount of hydrophilic boronates RB(OH) 3 - [31]. Overall, solvent mixture of 1,4-dioxane and H 2 O in ratio 2:1 appeared to be the most optimal, with 80% yield.  Table S3). Initial total volume of 6 mL in addition to 4.5 mL were found as the most suitable, while the yield significantly decreased when smaller solvent volumes were used. All further experiments were performed in total volume of 6 mL. Microwave-assisted reaction conditions, i.e., temperature and time, were further evaluated. First, the reaction mixture was subjected to temperature scan ranging from 60 • C to 140 • C with the interval of 20 • C ( Table 4). The temperature of 100 • C appeared to be the most optimal leading to 80% yield. Higher temperatures (i.e., 120 • C and 140 • C, entries 7 and 10, respectively) did not improve the yield, since the side product 2,4-diphenylpyrimidine was formed. Additionally, time of the reaction was monitored for temperatures equal or above 100 • C. Reaction time of 15 min at 100 • C (yield 81%, entry 4) was found as equally efficient as 20 min. Therefore, these conditions were used for all further reactions. To investigate the optimal amount of catalyst needed for the microwave-assisted Suzuki coupling, we performed the scan of catalyst loading, ranging from 5 mol% to 0.05 mol%. Contrary to the results from the experiments, performed in a flask, the minimal amount of catalyst needed for microwave-assisted reaction was only 0.5 mol% (Table 5), with lower amounts shown to be insufficient for all starting material to react. A scale-up of the model reaction on proposed reaction conditions has further been performed. The reaction was performed on 4.0 mmol scale of starting reagents, with obtained yield of 3 being 53%, and 30% of 1 still remaining unreacted in the reaction mixture. Therefore, we prolonged the microwave reaction time for 5 min to 20 min and improved the yield of 3 to 74%, with only 19% of 1 still being present in the reaction mixture. Taken together, we can conclude that our microwave-assisted procedure is also suitable for production of aryl pyrimidines on larger scales.
With the optimal conditions being determined, the scope of several aryl and heteroaryl boronic acids was investigated (Table 6). On the series of methoxyphenyl boronic acids, different substituent positions on the aromatic ring were evaluated (products 26, 27, and 28); however, no significant effect on the yield and selectivity was found between ortho, meta or para position of a methoxy group. Therefore, we preferentially focused on meta-substituents in the following reactions. Overall, the main product was consistently the 4-substituted 2-chloropyrimidine, with little or usually none of 2-substituted 4-chloropyrimidine being formed during the reaction. Several different substituents were screened for and it was noticed that yields of phenylboronic acids with electron-withdrawing groups on meta position (30 to 33) were rather high and comparable to the model reaction (3). Other boronic acids with carboxylic and electron-donating groups (i.e., amino group) led to lower yields. As expected, the naphthyl boronic acid (product 37) displayed similar yield to our model reaction with phenylboronic acid 2 (3).
Some heteroaryl boronic acids were also screened, with the reaction being selective and successful for both furanylboronic acids (38 and 39) with comparable yield. However, the coupling of 2,4-dichloropyrimidine with thiophen-2-boronic acid failed due to the sulphur poisoning of the palladium catalyst [40]. Furthermore, in case of pyridinylboronic acids, the reaction was also unsuccessful.

Chemistry and Chemical Characterization of Compounds
The reagents and solvents were purchased from commercial sources (Sigma-Aldrich, St. Louis, MO, USA; Acros Organics, Antwerp, Belgium; Alfa Aesar, Haverhill, MA, USA; TCI, Tokio, Japan) and used without further purification. The microwave-assisted reactions were performed using an Anton Paar Monowave 200 microwave reactor (Anton Paar GmbH, Graz, Austria). Flash column chromatography was performed on Merck silica gel 60 (mesh size, 70-230), using the indicated solvents. Monitoring of the purification was done by thin-layer chromatography on silica-gel plates (Merck DC Fertig-platten Kieselgel 60 GF254, Merck, Darmstadt, Germany) and visualized under UV light. Yields refer to the purified products or were determined by LC-MS, as stated, and were not optimized. rimidine (23) led to the production of both isomers 2-chloro-4-methoxy-6-phenylpyrimidine (44A) and 4-chloro-6-methoxy-2-phenylpyrimidine (44B) with no selectivity between them.  Taken together, the reaction is less susceptible to different substitutions on aryl boronic acids, with most of them giving high yield and selectivity at coupling with 2,4-dichloropyrimidine. Since one of the major obstacles in coupling with heteroaryl chlorides is their potentially insufficient electron-deficiency, substitutions on the scaffold have immense effect on the outcome of Suzuki-Miyaura cross-coupling reactions.

Chemistry and Chemical Characterization of Compounds
The reagents and solvents were purchased from commercial sources (Sigma-Aldrich, St. Louis, MO, USA; Acros Organics, Antwerp, Belgium; Alfa Aesar, Haverhill, MA, USA; TCI, Tokio, Japan) and used without further purification. The microwave-assisted reactions were performed using an Anton Paar Monowave 200 microwave reactor (Anton Paar GmbH, Graz, Austria). Flash column chromatography was performed on Merck silica gel 60 (mesh size, 70-230), using the indicated solvents. Monitoring of the purification was done by thin-layer chromatography on silica-gel plates (Merck DC Fertig-platten Kieselgel 60 GF254, Merck, Darmstadt, Germany) and visualized under UV light. Yields refer to the purified products or were determined by LC-MS, as stated, and were not optimized. rimidine (23) led to the production of both isomers 2-chloro-4-methoxy-6-phenylpyrimidine (44A) and 4-chloro-6-methoxy-2-phenylpyrimidine (44B) with no selectivity between them.  Taken together, the reaction is less susceptible to different substitutions on aryl boronic acids, with most of them giving high yield and selectivity at coupling with 2,4-dichloropyrimidine. Since one of the major obstacles in coupling with heteroaryl chlorides is their potentially insufficient electron-deficiency, substitutions on the scaffold have immense effect on the outcome of Suzuki-Miyaura cross-coupling reactions.

Chemistry and Chemical Characterization of Compounds
The reagents and solvents were purchased from commercial sources (Sigma-Aldrich, St. Louis, MO, USA; Acros Organics, Antwerp, Belgium; Alfa Aesar, Haverhill, MA, USA; TCI, Tokio, Japan) and used without further purification. The microwave-assisted reactions were performed using an Anton Paar Monowave 200 microwave reactor (Anton Paar GmbH, Graz, Austria). Flash column chromatography was performed on Merck silica gel 60 (mesh size, 70-230), using the indicated solvents. Monitoring of the purification was done by thin-layer chromatography on silica-gel plates (Merck DC Fertig-platten Kieselgel 60 GF254, Merck, Darmstadt, Germany) and visualized under UV light. Yields refer to the purified products or were determined by LC-MS, as stated, and were not optimized.

General Procedure for Catalyst Screening
50 µL of reaction mixture was diluted with 950 µL of internal standard solution (prepared by 102 mg of acetanilide in 100 mL of CH 3 CN). The sample was further diluted (1/10) with CH 3 CN and filtered to obtain the final sample for LC-MS analysis.
In case of Pd(dppf)Cl 2 ·CHCl 2 and Pd(PPh 3 ) 4 as catalyst, after 24 h, the reaction mixture was extracted with EtOAc, brine and dried over anhydrous Na 2 SO 4 . Solvents were removed under reduced pressure, the remaining solid was purified using flash column chromatography with EtOAc/n-Hex as an eluent.
50 µL of reaction mixture was diluted with 950 µL of internal standard solution (prepared by 102 mg of acetanilide in 100 mL of CH 3 CN). The sample was further diluted (1/10) with CH 3 CN and filtered to obtain the final sample for LC-MS analysis. (Table 3) 1 (75 mg, 0.5 mmol) was dissolved in a mixture of THF or 1,4-dioxane and H 2 O (total 6 mL), where the air (oxygen) was displaced with argon before use. Then K 2 CO 3 (207 mg, 1.5 mmol), 2 (61 mg, 0.5 mmol), and 3 mol% of Pd(PPh 3 ) 4 (17.3 mg, 0.015 mmol) were added and stirred under argon atmosphere. The reaction mixture was then stirred for 20 min at 100 • C in a microwave reactor. 50 µL of reaction mixture was diluted with 950 µL of internal standard solution (prepared by 102 mg of acetanilide in 100 mL of CH 3 CN). The sample was further diluted (1/10) with CH 3 CN and filtered to obtain the final sample for LC-MS analysis.
3.8. General Procedure for Scale-Up 1 (600 mg, 4.0 mmol) was dissolved in a mixture of 1,4-dioxane (10 mL) and H 2 O (5 mL), where the air (oxygen) was displaced with argon before use. Then K 2 CO 3 (1.656 g, 12.0 mmol), 2 (488 mg, 4.0 mmol), and 0.5 mol% of Pd(PPh 3 ) 4 (23.12 mg, 0.02 mmol) were added and stirred under argon atmosphere. The reaction mixture was then stirred for a specified amount of time (i.e., 15, 20 min) at 100 • C in a microwave reactor. 50 µL of reaction mixture was diluted with 950 µL of internal standard solution (prepared by 102 mg of acetanilide in 100 mL of CH 3 CN). The sample was further diluted (1/10) with CH 3 CN and filtered to obtain the final sample for LC-MS analysis.
3.9. General Procedure for Screening of Boronic Acids (Table 6) 1 (75 mg, 0.5 mmol) was dissolved in a mixture of 1,4-dioxane (4 mL) and H 2 O (2 mL), where the air (oxygen) was displaced with argon before use. Then K 2 CO 3 (207 mg, 1.5 mmol), appropriate boronic acid (0.5 mmol), and 0.5 mol% of Pd(PPh 3 ) 4 (2.9 mg, 0.0025 mmol) were added and stirred under argon atmosphere. The reaction mixture was then stirred for 15 min at 100 • C in a microwave reactor. Afterwards, reaction mixture was extracted with EtOAc, brine, and dried over anhydrous Na 2 SO 4 . Solvents were removed under reduced pressure, the remaining solid was purified using flash column chromatography with EtOAc/n-Hex as an eluent.

Conflicts of Interest:
The authors declare no conflict of interest.