In laboratory process development for the reduction of an amide carbonyl with borane-THF, headspace FTIR was used to measure the concentration of diborane in the vapor phase and determine the level of safety risk associated with the process. A constant nitrogen sweep was applied through the headspace of the reaction, and the IR spectra of the effluent gas were collected using an FTIR analyzer equipped with an on-line gas cell. The diborane peak at 2530-2480 cm-1 was used to quantify the headspace diborane level. Another major scale-up issue for this reduction was that the off-line LC assay required a time-consuming sample work-up, which would result in long down-times in the pilot plant. To address this issue, on-line liquid-phase FTIR was used to detect the amide and the amine-borane complex (product), and to determine the extent of reaction.

For pilot-plant scale-up, the laboratory results were used to design an optimized process. The optimized maximum pilot-plant diborane level for the reaction is ca. 1800 ppm, well below the lower explosive limit in air (5000 ppm). On the other hand, dummy reactions performed by charging borane-THF in the absence of amide substrate would generate ca. 18,000 ppm diborane at the optimized reaction conditions. Thus, it is important to ensure that the reaction is progressing during the borane-THF charge. The liquid-phase FTIR data show that the substrate is consumed rapidly, forming intermediates that convert to amine product. The data indicate that the reaction is complete after 4 hours, and the endpoint is confirmed by running the LC assay on an acid-treated end-of-reaction sample. Liquid-phase FTIR was used during pilot-plant scale-up of this reduction to ensure that the reaction occurred on the desired timescale, and to determine the reaction endpoint. Lab and pilot-plant results will be presented.