Synthesis of 6-azido-1-(phenylsulfonyl)hexanol from 6-bromohexanol Abstract 6-bromohexanol was reacted with a series of reagents through three reactions in order to yield 6-azido-1-(phenylsulfonyl), an azide-functionalized linker used to attach a TAK-242 prodrug to tissue surfaces for sustained release.1 The first reaction utilizes an Sn2 mechanism to remove the halogen and add an azido group and had an 85.6% yield. The trichloroisocyanuric acid (TCCA) deprotonates the alcohol in the second reaction and allows TEMPO to deprotonate the alpha carbon in order to create an aldehyde with a 42.6% yield. In the final reaction n-BuLi deprotonates the methyl group of the phenyl methyl sulfone (PMS) allowing it to attack the carbonyl carbon of the aldehyde. This creates the final product in a 12.2% yield (4.45% overall). Scheme I: 6-bromohexanol to 6-azidohexanol under microwave Scheme II: 6-azidohexanol to 6-azidohexanal using trichloroisocyanuric acid Scheme III: 6-azidohexanal to 6-azido-1-(phenylsulfonyl)hexanol using n-BuLi The FTIR and NMR data used to track the progress of the reaction show that the products at each stage are fairly pure, except for the crude NMR taken before the column separation. Introduction Diabetes is an important area of research in modern medicinal chemistry. In diabetics, the pancreas either no longer makes insulin due to the destruction of beta cells due to an autoimmune condition (Type I) or progressive insulin resistance (Type II). Now, pancreatic islet cell transplantation gives those with diabetes a chance to naturally produce insulin through new cells. A recurring issue with transplantation is the rejection or damage of the transplanted organ or cells, especially through inflammation. A new method of reducing inflammation is modifying the surface of the islet cells to slowly “bleed” anti-inflammatory drugs, like TAK-242. In order to attach TAK-242 to the islet cells, a linker molecule must be used that can bind to both the prodrug and the cell surface. Figure I: Scheme of the attachment of linker to TAK-242, the cell, and subsequent release1 This experiment was done in order to create a linker that can attach a prodrug to the islet cell surfaces in order to be hydrolyzed to release the prodrug to reduce inflammation.1 When producing molecules to be used for pharmaceutical purposes, it is important to have high purity, yield, and, ideally, low cost. Results and Discussion The FTIR data (included at end) for this experiment showed the reactions were progressing as expected. There were a few important peaks that were expected at every step in the reaction. The most obvious peak was the wide alcohol peak at ~3300cm-1. This peak was seen in the starting material and the 6-azidohexanol, but disappeared when the hydroxide group was converted to a aldehyde. At this point in time three new peaks appeared. The carbon oxygen double bond stretch showed sharply at ~1700cm-1 while the carbon hydrogen bond had two peaks at ~2730/2830cm-1. These peaks disappeared, and the alcohol peak reappeared after the addition of PMS. Another bond that was tracked was the azido group’s. There was initially no peak because the starting material contained a halide, but when the azido group was added, a new peak showed at ~2070cm-1 and remained for the rest of the experiment. It was more difficult to tell which peaks were both new and important after the addition of PMS. From available FTIR data it would be expected that two new peaks would appear in the 1225-980cm-1 stretch for the sulfur/oxygen double bonds and in 710-570cm-1 for the S-C stretch.2 NMR data (included at end) also showed a good progression of the reactions. The initial reactant NMR was very clean, showing the expected triplets at 3.6 and 3.4ppm for the carbons adjacent to the alcohol and halide, respectively. The alcohol hydrogen showed lower than expected at 2.6ppm. The rest of the alkane hydrogens showed at 1.4, 1.55, and 1.85 as expected. The next NMR for 6-azidohexanol showed similar peaks, but the azido group is not nearly as polar so the hydrogens on the adjacent carbon were more shielded than those next to the halide. This resulted in a triplet showing at 3.25ppm. A multiplet appeared around 3.4ppm and could be the alcohol peak that is not seen anywhere else, but it should, theoretically, just be a rough singlet. After the conversion to 6-azidohexanal, another NMR was taken. The most important, and distinct, change in this graph was the addition of the triplet aldehyde peak at 9.8ppm. This also further deshielded the carbons adjacent to the aldehyde and moved their peak up to 2.5ppm. This NMR showed two unexpected peaks at 3.3 and 3.4 along with heavy interference peaks from acetone and chloroform. The crude NMR taken before the column separation was, as expected, messy, but surprisingly similar to the purified product NMR. The crude NMR had an additional significant peak at 1.9ppm. The final NMR had the expected benzene peaks in the 7.58.0ppm range, with the hydrogens closer to the S=O bonds being more deshielded. The alcohol peak showed at 3.4ppm as expected and the adjacent carbon also showed the expected quartet at 4.1ppm. This is more deshielded because it sits next to the alcohol but is also adjacent to the sulfur which is double bonded to two oxygens. The alkane peaks remained the same as prior. There was still interference from acetone at 2.17ppm and chloroform at 7.26ppm along with an unexpected, strong singlet at 2ppm. This could be the beta carbon from the alcohol, as the strong deshielding from the alcohol may have further deshield the next carbon, but it would still be expected to be a quartet. Unknown multiplets also appeared at 1.5ppm. These appeared on the crude NMR and are most likely leftover contaminants that made it through the column. 1H NMR (300 MHz, Chloroform-d) δ 3.61 (t, J = 6.6 Hz, 2H), 3.39 (t, J = 6.8 Hz, 2H), 2.63 – 2.55 (m, 1H), 1.85 (dq, J = 8.1, 6.7 Hz, 2H), 1.64 – 1.27 (m, 4H). Reported coupling constants for 6-bromohexanol 1H NMR (300 MHz, Chloroform-d) δ 3.66 (t, J = 6.5 Hz, 1H), 3.28 (t, J = 6.9 Hz, 1H), 2.30 – 2.11 (m, 1H), 1.61 (dt, J = 9.9, 6.9 Hz, 1H), 1.45 – 1.33 (m, 0H). Reported coupling constants for 6-azidohexanol 1H NMR (300 MHz, Chloroform-d) δ 9.79 (t, J = 1.6 Hz, 1H), 3.48 – 3.35 (m, 1H), 3.30 (t, J = 6.8 Hz, 2H), 2.55 – 2.30 (m, 2H), 2.00 – 1.70 (m, 1H), 1.74 – 1.55 (m, 2H), 1.54 – 1.34 (m, 1H). Reported coupling constants for 6-azidohexanal 1H NMR (300 MHz, Chloroform-d) δ 8.02 – 7.83 (m, 3H), 7.77 – 7.53 (m, 4H), 4.13 (q, J = 7.2 Hz, 3H), 3.42 (q, J = 3.6, 2.2 Hz, 0H), 3.39 (s, 2H), 3.41 – 3.12 (m, 4H), 3.08 (d, J = 0.6 Hz, 1H), 2.06 (d, J = 0.6 Hz, 3H), 1.93 – 1.81 (m, 1H), 1.27 (td, J = 7.1, 0.6 Hz, 4H), 0.87 (s, 0H). Reported coupling constants for 6-azido-1-(phenylsulfonyl)hexanol The only UV active compounds used in the experiment were PMS and the final product, so they were TLC’d after the final purification and to track the progress of the column. All TLC’s were run in 25/75 ethyl acetate/hexanes. Molecule PMS Final Product Rf value 0.20 0.55 Table I: TLC data The reactions had good yields throughout the experiment until the column chromatography was done. The 85.6% yield for the first reaction was fair, but considering the simplicity of the reaction and its mechanism would indicate that a quantitative yield could possibly be achieved. The next reaction progressed with a 42.6% yield, but was a considerably more complicated reaction. The final reaction was similarly complicated, using n-BuLi to deprotonate PMS which would act as a nucleophile to attack the aldehyde. A similar yield would be expected, but product was most likely lost in the column due to user error leading to the 12.2% yield. Reaction Reactant mass (g) Halide -> Azide 0.601 Alcohol -> Aldehyde 0.407 Aldehyde -> Sulfonyl 0.171 Reactant moles 0.00332 0.00284 0.00121 Mol Ratio 1 to 1 1 to 1 1 to 1 Moles Product 0.00284 0.00121 0.000148 Theoretical Yield (g) 0.475 0.401 0.343 Actual Yield (g) 0.407 0.171 0.042 Percent Yield 85.6 42.6 12.2 Table II: Stoichiometric ratio of reaction and yields Conclusions From the experimental data, it would appear that the prodrug linker 6-azido-1(phenylsulfonyl)hexanol can be produced with fair yields and high purity. With more practice high yields could probably be achieved. The main area of loss with this experiment was in the column. The step-wise eluent setup may not have been ideal for running the molecule in question and might need to be adjusted for the future. In addition, the time spent in the column needs to be reduced so the diffusion of the molecule has less of an effect, although this is a user error and will improve with time. Another probable error occurred upon the addition of NH4Cl used to quench the reaction. The protocol said to wait until the solution was clear and then add the quenching agent. Due to time constraints, the NH4Cl was added while there was still color in the solution. Upon addition, the solution immediately turned clear, indicating the reaction may have been cut short. This drug is meant to act as a mediator of slow drug release in-vivo for islet cell transplantation. A 2018 paper by the Kane lab at Baylor University concluded that this linker effectively releases TAK-242 over several hours after transplantation, effectively providing protection against inflammatory damage. Furthermore, it was found that the conjugation chemistry this linker is used for can be applied to other tissues and organs, indicating a possible, broader application.1 Experimental Section 0.6g of 6-bromohexanol and 0.43g of sodium azide were heated in 4ml of water in the microwave for 1 hour at 100 degrees C. The mixture was diluted with 25ml of water and extracted with three 25ml aliquots of DCM. The organic layer was separated with a separatory funnel, dried with magnesium sulfate, and rotovapped to give a colorless oil. NMR and FTIR characterizations were taken at this point. Two days later the 6-azidohexanol (0.407g) was vigorously stirred with TEMPO (0.01g) and sodium bicarbonate (0.284g) in 5.7ml of dichloromethane and 0.59ml of water. After this mixture is thoroughly stirred, 0.264g of solid trichloroisocyanuric acid was added in small portions and allowed to stir for an additional 30 minutes. The organic phase was separated and washed successively with saturated aqueous NaHCO3 and brine. This mixture was again dried with magnesium sulfate and rotovapped to provide the 6-azidohexanal. Additional NMR and FTIR characterizations were done at this time. Five days later, 0.172g phenyl methyl sulfone was stirred with 13ml anhydrous THF (should have been 3.3ml) and cooled to -78 degrees C. At this point, 0.688ml of 1.6M n-BuLi was added dropwise and the solution was allowed to warm to 0 degrees C over 30 minutes. The mixture was then re-cooled to -78 degrees C and 0.172g 6-azidohexanal was added. The mixture was stirred for 15 minutes and removed from the cooling bath to warm to room temperature. When the mixture became clear, 5ml of saturated aqueous NH4Cl was added to quench the reaction. This mixture was left to sit for two days. The mixture was finally diluted with ethyl acetate and washed with water and brine, dried with magnesium sulfate, and rotovapped to yield 0.299g of crude product. A crude NMR was taken at this point. A column was set up with silica gel and run with a step gradient of 0, 25, and 50% ethyl acetate in hexanes. The progress of the column was tracked with TLC run in 25/75 ethyl acetate/hexanes. After purification, 0.042g of 6-azido-1(phenylsulfonyl)hexanol, a pale, yellow oil was yielded. Final TLC, NMR, and FTIR characterization was done at this time. References 1. Chang, Charles A. Et Al. Ex-vivo generation of drug-eluting islets improves transplant outcomes by inhibiting TLR4-Mediated NFkB upregulation. Biomaterials. 2018. 159, 13-24. 2. Hampton, Demoin. Et Al. (2010). Vibrational spectroscopy tutorial: sulfur and phosphorus. Available from http://faculty.missouri.edu/~glaserr/8160f10/A03_Silver.pdf.