Multikilogram Synthesis of 4-D-Erythronolactone via Batch and Continuous Processing Loretta L. Wong,* Run Ling Wong, Gabriel Loh, Phyllis E.W. Tan, Soo Khean Teoh, Salim M. Shaik, Paul N. Sharratt, Wee Chew, Suat Teng Tan, and David Wang Process Science and Modelling, Institute of Chemical and Engineering Sciences, 1 Pesek Road, Jurong Island, Singapore 627833 ABSTRACT: The conversion of a batch process to continuous (flow) operation has been investigated. The manufacture of 4,D-erythronolactone at kilogram scale was used as an example. Fully continuous processing was found to be impracticable with the available plant because of the difficulty in carrying out a multiphase isolation step continuously, so hybrid batch-continuous options were explored. It was found that very little additional laboratory or process safety work other than that required for the batch process was required to develop the hybrid process. A hybrid process was chosen because of the difficulty caused by the precipitation of solid byproduct during the isolation stage. While the project was a technical success, the performance benefits of the hybrid process over the batch were not seen as commercially significant for this system. ■ INTRODUCTION The replacement of batch processing with continuous and/or batch-continuous hybrid processes has been a theme attracting significant interest for some years in the fine/specialty chemicals and pharmaceuticals industries. 1 The potential to access more efficient, safer, and less polluting processes is attractive to sectors that face a range of regulatory and commercial pressures. How- ever, adoption has been slower than many had hoped. There are a variety of reasons for this, for example existing batch capital assets, a perception of difficulty, fear of the novel, lack of suitable resources to explore the possibilities, and concerns about the potential time to develop continuous processes. In many cases where a serious exploration has taken place (or even when a continuous process has been implemented) the business benefits are often not found to be overwhelmingly attractive versus the traditional batch options. It has recently been esti- mated that the pharmaceutical industry has invested $600 million in continuous processing research but has rather little to show in terms of benefits. 2 The work reported here is part of a broader programme in the investigation of the opportunities and challenges for innovative (and often continuous) processing for high-value chemicals. The aim is to overcome at least some of the barriers to uptake through developing and making available experience and comparative studies at a credible scale. This work set out to explore the conversion of a simple batch process to continuous operation, and to demonstrate batch and continuous processes at a small pilot scale The production rate was chosen to deliver the equivalent quantity of product to a 2-3 kg batch process. This allowed for a direct comparison of efficiency between the batch and continuous process. There were multiple objectives in doing this: • To provide a real example for the comparison of batch and continuous processes making the same product • To understand where and how the deployment of continuous processing brings benefits and problems • To act as a motivating example for the design and opera- tion of a multipurpose continuous facility [though this is not described in any detail here] • To explore how various process development and scale- up tools [PAT, calorimetry, laboratory experimentation, knowledge capture tools] performed in the batch-to- continuous conversion and to look for gaps where new methods and tools could be useful The synthesis of 4,D-erythronolactone was used as the work- ing example. A batch process was adapted from the literature using normal practices in scale-up. The approach to continuous process design was broadly to convert the batch process step- by-step to continuous stages, looking to make only small changes (for example in concentration to maintain mobility), and avoiding the presence of solids where practicable. Kinetic and thermal measurements were used to support sizing and/or rating of equipmentand where practicable the ones used were those that had already been carried out for development of the batch process. Broadly, the example is seen as equivalent to the conversion of an existing batch process based on reasonable (though not excessive) process understanding already available. The aim was not to develop an “optimal” process from first principles. This is a realistic position for many in the fine chemicals and pharmaceuticals industry, where the opportunity to obtain step change improvements in an existing process performance might be considered to address cost or regulatory concerns. Cohen et al. 3 published a 1 L laboratory-scale batch procedure for the synthesis of the isopropylidene acetal of 4-DEL which also formed the basis for a patented procedure towards the total synthesis of Swainsonine. 4 A 20 L laboratory- scale procedure (704 g) was published by Dunigan 5 in 1991 following Cohen’s procedure (Scheme 1). Special Issue: Continuous Processes 2012 Received: December 4, 2011 Published: April 10, 2012 Article pubs.acs.org/OPRD © 2012 American Chemical Society 1003 dx.doi.org/10.1021/op200352k | Org. Process Res. Dev. 2012, 16, 1003-1012