2-(1-Thianthrenyl)-6-(N-morpholinyl)-4-pyranone

Before the preparation, anhydrous potassium carbonate (granulated, Aldrich) was powdered with a spatula and dried in a beaker in glassware-drying oven at 130C overnight. In a microwave-ready pressure vial equipped with a stirbar, 2-chloro-6-(N-morpholinyl)-4-pyranone 216mg (1.00mmol) mixture with thiantrene-1-boronic acid 390mg (1.5mmol, 90% pure, Aldrich), dry potassium carbonate 1.4g and activated 4A molecular sieves 1.1g (powder <5um, Aldrich) was suspended in anhydrous dioxane 16mL and the slurry was sparged with Ar gas for 10 min. Pd(PPh3)4 90mg (7.5mol%) was added and sparging was continued for additional 10 min. The vial was then capped under Ar and heated in a Biotage “Initiator” microwave at 130C for 2 hours. The cooled reaction mixture was poured onto a pad of silica 20g and the solids on the silica pad were washed with ethyl acetate 100mL – these first filtrates were discarded. The silica pad was then washed with a 2:1 chloroform + methanol mix 300mL; the methanolic filtrates were evaporated. Purification of the crude product on a silica column (40 g) in a methanol gradient in chloroform, 0 to 5% MeOH provided the pure product as a sticky glass. This purified product was dissolved in benzene (10mL) and the solution was freeze-dried on highvac. Y=345mg (87%) of a fluffy light-tan amorphous solid.

1H(d6-DMSO, 400MHz):  7.763(app d, 7.6Hz, 1H), 7.601(app t, 7.3Hz, 3H), 7.470(app t, 7.8Hz, 1H), 7.357(m, 2H), 6.254(d, 1.6Hz, 1H), 5.532(d, 1.6Hz, 1H), 3.690(br m, 4H), 3.383(br m, 4H); 13C(d6-DMSO, 100MHz): 177.77, 162.93, 157.95, 135.38, 135.04, 134.59, 133.58, 132.65, 131.26, 129.36, 129.10, 128.82, 128.78, 128.43, 128.24, 114.00, 89.41, 65.24(2C), 44.44(2C)

The procedure for making the 2-chloro-6-(N-morpholinyl)-4-pyranone was posted on Oct 3. The Suzuki procedure works equaly well with 5mol% of Pd-tetrakis (the 7.5mol% loading was due to weighing error).

The above procedure is an improvement over the patent procedure (published Y= 4% th). From reading the patent experimentals, I suspected that all Suzuki procedures reported there (with yields ranging from  2 to 13%) suffered from some kind of execution error – perhaps due to a lack of care when experiments were set and run in parallel. Also, experiment done under typical aqueous Suzuki reaction conditions (Na2CO3, Pd-tetrakis, DME+water)  returned no desired product at all. Next, I tried adding water to the anhydrous dioxane experiment – into a reaction mixture that contained the desired product – and this resulted in a rapid destruction of the product under the reaction conditions. So it became clear that the actual problem was the moisture sensitivity of the product in the presence of carbonate. Using a large excess of dry K2CO3 together with 4A molecular thieves supressed this decomposition problem.

For a few carbons more

I was trying to make diazaindoles from diaminobromopyrimidines by Pd-catalyzed arylation of a cyclic 1,3-diketone (dimedone) when I met these cyclized strangers:

Apparently the cyclization product resulted from a reductive Heck coupling with a condensation product of acetaldehyde plus dimedone. I used a fairly high Pd(II) pre-catalyst loading (10 mol%) so I guess the acetaldehyde equivalent was generated by the Pd(II)-promoted oxidation of triethylamine.

In this wild story the addition product does not beta-eliminate Pd(0) – as in typical Heck reaction – but undergoes a protonolysis instead. Pd(II) re-emerges from the sequence at the end, ready to hunt down and oxidize another molecule of triethylamine. (Credits: Dr. Ni for ROESY)

Diazaindole cyclization

The acetylene compound 92.5mg (0.3 mmol) and CuI 114.5mg (0.6mmol) was suspended in anh. dimethylacetamide 0.8mL in a 25mL14/20 round flask equipped with a straight stopcock. The mixture was degassed by highvac/Ar purge. Barton base 0.15mL (0.75mmol, Fluka) was added through the stopcock and the mixture was stirred at 135-140C oil bath under Ar in a closed flask for 3 hours. The mixture was cooled to RT, methanol 1mL was added followed by TMS-CN 0.25mL. The mixture  was diluted with dichloromethane 10mL, sonicated for 10 min and then stirred at RT for 1 hour. The precipitated CuCN was removed by filtration through a pad of silica. The silica was washed with 50mL of a 10:1 mixture of dichloromethane with methanol. The filtrates were evaporated and the residue was dried on highvac. The crude product was purified on a column of silica (35g) in EtOAc gradient in dichloromethane, 10 to 30% of EtOAc (in 45 min, 20mL/min). Y= 77mg(83%) of a light tan solid

1H(CDCl3, 400MHz):8.330(br s, 1H), 7.393(m, 2H), 7.298(m, 2H), 7.223(m, 1H), 6.025(s, 1H), 5.707(br s, 1H), 4.677(s, 2H), 4.071(q, 7.0Hz, 2H), 2.635(t, 7.8Hz, 2H), 1.686(m, 2H), 1.453(m, 2H), 1.280(t, 7.0Hz, 3H), 0.973(t, 7.4Hz, 3H)

The above procedure was succesfully used to cyclize a large number of diaminopyrimidine acetylenes with excellent yields (82 to 96%Y). Various substituent groups on amines and Phe, Bu, H on acetylene were tolerated. Only TMS-protected acetylene compounds failed to provide diazaindoles but the deprotected acetylenes (R=H) cyclized smoothly.

Barton base is a low-boiling strong base. It serves here both as a copper ligand and a base – it keeps things soluble and it deprotonates the aminopyrimidine NH. (I have not tried other bases like DBU but the reaction works even without base – the reaction was slower and formation of voluminous precipitate was a complication). Since the product binds strongly to Cu, cyanide excess was used in the workup to break the complex. Trimethylsilyl cyanide is very poisonous, it has to be handled with gloves in the hood.

End of daily updates

Sonogashira on 2,4-diamino-5-bromopyrimidine

A solution of 5-bromo-2,4-dichloropyrimidine 4.989g (21.89mmol, Aldrich) in anh THF 20 mL was cooled to 0C. With vigorous stirring, a solution of 2M ethyl amine in THF 24 mL (48 mmol, Aldrich) was added dropwise over 15min (a voluminous precipitate formed). The cooling bath was removed and the mixture was stirred at RT for 45 min. The reaction mixture was filtered through a 40g plug of silica and silica was washed with 5:1 mixture dichloromethane-ethyl acetate 240mL. The filtrates were evaporated and the residue was dried on highvac overnight. The obtained oily residue was dissolved in neat benzylamine 14.4mL (132 mmol) and the mixture was stirred under argon at 60C for 11 hours (overnight).  The reaction mixture was concentrated on highvac to dryness. The residue was dissolved in a mixture 4:1 dichloromethane-ethyl acetate, filtered through a pad of silica (40g) and silica was washed with 250mL of the same mixture. The filtrates were evaporated and dried on highvac. The solidified residue was re-crystallized from hexane. Y=5.631g (84%) of a white crystalline solid. (The minor regioisomer remained in the supernatants). 1H(CDCl3, 400MHz): 7.710(br s, 1H), 7.333-7.221(m, 5H), 5.852(br s, 1H), 5.098(s, 1H), 4.542(d, 6Hz, 2H), 3.416(m, 2H), 1.174(t, 7.4Hz, 3H); 13C(CDCl3, 100MHz): 161.05, 158.16, 155.66, 139.65, 128.44, 127.533, 127.01, 45.72, 35.78, 14.74

5-bromo-2-benzylamino-4-ethylaminopyrimidine 276.5mg (0.90 mmol) from the previous step was combined with 15 mg of PdCl2(dppf).CH2Cl2 (2 mol%) and 2.5mL of anh. THF in a 4mL vial. Triethylamine 0.19mL(1.35mmol) and 1-hexyne 0.155mL (1.35 mmol) was added and the mixture was stirred in a closed vial for 10 min. Solid CuI 2.9mg (1.7mol%) was added, the vial was briefly flushed with Ar, capped and placed on oil bath. The mixture was stirred at 60C for 20h. (The progres of reaction was monitored by observing formation of Et3N.HBr precipitate). The reaction mixture was filtered through a plug of silica (2g, washed with 5:1 DCM-EtOAc mix 45mL). The filtrates were evaporated and the residue was purified on a column of silica 10g  using a ethyl acetate gradient in dichloromethane, 0 to 15% EtOAc. Y=249mg (90%) of a pale-yellow solid.

1H(CDCl3, 400MHz): 7.880(br s, 1H), 7.331-7.258(m, 5H), 5.387(br s, 1H), 5.274(br s, 1H), 4.598(d, 5.5Hz, 2H), 3.437(m, 2H), 2.442(t, 7.3Hz, 2H), 1.575(m, 2H), 1.465(m, 2H), 1.188(t, 7.2Hz, 3H), 0.950(t, 7.4Hz, 3H); 13C(CDCl3, 100MHz): 162.14, 160.64, 157.73, 139.72, 128.45, 127.52, 127.01, 96.50, 73.77, 45.45, 35.39, 31.03, 22.06, 19.44, 14.85, 13.63

Coordinating properties of diaminopyrimidine make these bromides a relatively troublesome substrate class in the Sonogashira coupling. By comparison with various frequently-used sonogashira systems, the above conditions – THF, NEt3, PdCl2(dppf), CuI – were far superior to many other systems for these diaminopyrimidine bromides. Phenyacetylene and TMS-acetylene also coupled with high yields (>80%). With less electron rich alkynes like phenylacetylene, 5mol% of the catalyst was usualy applied to make the reaction complete within one day at 60C. Degassing is not essential. It is important that CuI is added very last. Small quantities of CuI can be difficult to weight (a cut tip of Pasteur pippet was used for weighing) but it is important not to overload the reaction with Cu salt – Cu is not rate-controlling and its excess promotes side-reactions. CuBr can be used in place of CuI.  The products have tendency to tail on silica therefore the gradient elution was used. Since the stating material is usualy close to the product, it is important to ensure a complete conversion of the starting material.

The analogous diaminopyrimidine iodides had faster initial rate but curiously failed to provide a complete conversion of the starting material under these reaction conditions. It is possible that the reaction progress was in this case inhibited by accumulation of NEt3.HI together with coordinating properties of the pyrimidines.

4-(2′,4′-difluorobenzyl)-piperidine hydrochloride

 

TFA 100mL was added to 1-acetyl-4-(2′,4′-difluorobenzoyl)-piperidine 5.00g (Astatech; 18.7 mmol) slurry in triethylsilane 20mL (125mmol). The mixture was stirred at RT for 30 min, then cooled to -20C on salt/ice bath under argon. Boron trifluoride etherate 15mL (119 mmol) was added dropwise over 5 min and the mixture was stirred at -20 to -10C for 4 hours, then at -10 to -5C for 1 hour. The reaction mixture was concentrated to a small volume from ambient bath, the residue was diluted with water 50mL and ether 150mL. Saturated NaHCO3 solution 300mL was slowly added (CO2 evolution) and the mixture was stirred at RT overnight (16 hours). The mixture was extracted twice with ether (2x200mL), the extracts were washed with brine 200mL, combined, dried (MgSO4) and evaporated. The residue was purified on a column of silica (80g) in a EtOAc gradient in hexane, 0 to 100%EtOAc, followed by straight EtOAc.

The obtained pure 1-acetyl-4-(2′,4′-difluorobenzyl)-piperidine was dissolved in conc HCl 80mL. Water 80mL was added and the mixture was refluxed on oil bath (140C) for 5 hours. The obtained solution was cooled and evaporated,  the residue was re-evaporated from a 2:1 chloroform-methanol mixture (50mL) and dried on highvac. The residue was suspended in ethyl acetate 10mL, TBME 200mL was added, the slurry was briefly sonicated (5 min), filtered quickly (hygroscopic!), the precipitate washed with TBME and dried on highvac. Y=4.212g (91%Y) of a white solid, pure by HPLC, LC/MS(+cESI): 212(M+1)

BF3 in TFA is much stronger acid than TFA alone. (Without BF3 the reduction stops at benzylic alcohol in this case). Addition of BF3 etherate is also advantageous for easier workup on large scale: TES-F is much easier to evaporate than TES-OH.

The End Is Nigh

 

The regular daily updates will end this Friday, Nov 17. I will keep Org Prep (Monthly) up as my personal chemistry page. I will continue adding experimental procedures here but this will happen infrequently. So thank you all for your visits during the past few weeks!

Milkshake