A Flow-Based Synthesis of 2-Aminoadamantane-2-carboxylic Acid Claudio Battilocchio,* ,†,‡ Ian R. Baxendale, † Mariangela Biava, ‡ Matthew O. Kitching, † and Steven V. Ley † † Department of Chemistry, University of Cambridge, Lensfield Road, Cambridge CB2 1EW, United Kingdom ‡ Dipartimento di Chimica e Tecnologie del Farmaco, “Sapienza” Universita ̀ di Roma, P.le A. Moro 5, Roma 00185, Italy * S Supporting Information ABSTRACT: The development of a new, high-yielding, scalable and safe process for the preparation of 2-aminoadamantane-2- carboxylic acid (1) is described. This geminal, functionalized achiral amino acid has been reported to possess interesting biological activity as a transport mediator due to its unique physiochemical properties. We report herein on the use of various mesoreactor flow devices to expedite the lab-scale synthesis of this molecule by simplifying the processing requirements for use of several potentially hazardous reagent combinations and reaction conditions. ■ INTRODUCTION The adamantyl cage is an important motif 1 present in numerous biologically active compounds including a number of currently used therapeutic agents (Figure 1). 2 The unique symmetrical yet configurationally rigid geometry of the adamantane cage is often used to influence a structure’s physiological properties by providing a bulky lipophilic scaffold. 3 One particularly interesting adamantane derivative, namely, 2-aminoadamantane-2-carboxylic acid (1), is an unnatural, achiral amino acid which has been shown to possess novel transport inhibitory properties (Figure 2). 4 According to Tager and Christensen, 5 this molecule perfectly fulfills the theoretical requirements for transport inhibition of amino acids: (a) bulky side chain, (b) side chain apolarity, (c) catabolic resistance for the presence of a tertiary α-carbon, and (d) sufficient water solubility. We became interested in the synthesis of this amino acid as part of a research program directed towards the preparation of key probes for neurotensin receptors 1 and 2 (NTR1 and NTR2). 6a NTR1 and 2 are both seven-transmembrane G protein-coupled receptors (GPCR’s) with increasing relevance in several human cancers. 6 In the early 1990s, Sanofi-Aventis in a high throughput screening campaign identified SR 45398 (2) 7 as a binder of NTR in guinea pig brains (IC 50 40 μm). Through iterative lead optimization, this was developed into the NTR1 antagonist Merclinertant SR 48692 (3) 7,8 and then a second-generation analogue SR 142948A (4) 9 capable of preferentially activating NTR 2. In these molecules, the 2-aminoadamantane-2- carboxylic acid residue, 1, is reported to be responsible for key interactions allowing the selective recognition of the NTR receptor. 10 Wishing to prepare large quantities of both SR 48692 (3) and SR 142948A (4), we required a scalable and reliable route to this key fragment. ■ DISCUSSION Evaluating the preparative routes currently reported in the literature we identified two main synthetic strategies. Our initial investigations employed a Bucherer-Berg reaction as described by Nagasawa (Scheme 1, footnote a). 11 In the original report, treatment of ketone 5 with a buffered solution of sodium cyanide at elevated temperature and pressure yielded the spirohydantoin 6 which, without purification, was hydrolysed to the desired amino acid 1 in good overall reported yield. In order to more readily achieve the necessary elevated reaction temperature and pressure windows as outlined (Scheme 1, footnote a) we transposed the chemistry to a Biotage microwave reactor initially using sealed 20 mL reaction vials with only minor modifications of the previous batch route (Scheme 1, footnote b). This allowed us to perform the chemistry safely, yielding material identical to that obtained following the original batch conditions. (Note: this is an important point of consideration relating to later discussion concerning product purity.) Although batch processing of such relatively small volumes was straightforward when conducting optimization and test reactions, a microwave approach can be severely limiting when considering scale-up. It is not possible to directly scale a microwave cavity in order to increase output equivalent to a standard batch reactor. Physical restrictions, pertaining to the defined penetration depth of microwaves into a reactor, limit such an approach or certainly make reactor scaling a major engineering challenge. 12 We circumvented this issue in part by employing a CEM stop-flow microwave. 13 The adoption of stop-flow microwave handling allowed the rapid scale-up of the reaction sequence even though it required the processing of heterogeneous slurries (hydantoin input reaction solution and the resulting product mixture). Furthermore, the continuous monitoring of temperature and pressure combined with automatic safety cutoffs also minimized the risks associated with scale-up. Of particular value was the automation of the entire process which ensured reproducibility across the consecutively conducted batch operations whilst lowering the manual handling time necessary for a skilled chemist. In this fashion, material could be quickly processed through the two- step sequence on >400 mmol scale. Unfortunately, in our hands, isolation of the pure amino acid 1 was not possible. Following either the original reported Received: March 27, 2012 Published: April 17, 2012 Article pubs.acs.org/OPRD © 2012 American Chemical Society 798 dx.doi.org/10.1021/op300084z | Org. Process Res. Dev. 2012, 16, 798-810