Chiral scale-up using S MB chromatography
Hanns-Ingolf Paul and Tobias Reichelt from Bayer AG, Germany, discuss the industrial production of pure chiral compounds using simulated moving bed chromatography
Hanns-Ingolf Paul and Tobias Reichelt from Bayer AG, Germany, discuss the industrial production of pure chiral compounds using simulated moving bed chromatography
Due to their highly specific activity, optically active substances play an important role in the pharmaceutical industry. Chiral compounds can be produced by various methods and the classical workhorse for enantiomer production consists of chemical methods. However, successful custom chiral manufacturing demands a fairly complete tool box in order to provide cost efficient and fast services, and physical methods, such as chromatography, offer interesting alternatives.
Chromatography is a well-known and widely established analytical separation technique due to its high selectivity, and in pharmaceutical r&d it is well established for the isolation of target substances on a preparative scale.
However, obstacles to industrial applications, including dilution of the separation products and batch mode operation, impose practical limits for scale-up.
Simulated moving bed (SMB) chromatography, is a continuous countercurrent chromatography process (see figure 1) that offers distinct improvements regarding efficiency and avoids these scale-up limitations.
specialised plant
SMB is of particular interest in those cases where separation using classical methods, e.g. crystallisation or diastereomeric separation, is not possible. Chromatography requires a specialised, sophisticated plant and is therefore expensive. As a consequence, it is usually preferred for high-value products.
Although SMB has been available for a number of years, typical applications are still on the small to medium scale, but the manufacturers of the SMB units as well as the producers of chiral stationary phases (CSP) point out that the practical limits of scale-up, even for chiral separations, have not yet been reached.
The most striking advantage of SMB separation is the time factor. Typically, separations are well explored for analytical purposes during early product development. Later mg to g amounts of substance is required for in vitro toxicology and efficacy studies. These are often generated using preparative batch chromatography. Hence, at this point of clinical development, there is already some knowledge about chromatographic separation, i.e. CSP and solvents, has already been obtained or elucidated.
The next step is optimisation of the separation system and the application to SMB technology. Once the process change from preparative scale to SMB is made, further scale-up to pilot and industrial scale is straightforward and risk-free.
Another strong indicator for reasonable economic potential for an SMB separation is the possibility of racemisation of the unwanted enantiomer. SMB typically allows the isolation of both optical enantiomers in fairly high purity, and recirculation of the 'wrong' enantiomer significantly improves the overall recovery of the chiral separation step and boosts the economy of the process. Therefore, SMB is beneficial for separations where racemisation is possible and requires the 'wrong' enantiomer chemically unchanged with high purity and yield.
The ongoing innovations in the field of CSP, as well as the continuously growing database of investigated separation systems, has dramatically improved the likelihood of successful screening.
Initiated by a customer inquiry, an SMB project begins with screening experiments using a sample of approximately 2-5g of the racemate. Once a separation system has been identified, loading experiments are performed. On the basis of these, process simulations will be conducted to provide an initial estimation of the SMB process parameters and its productivity.
experimental confirmation
A typical time frame for this work is two to four weeks, depending on the complexity of the system under consideration. At the end, the customer is provided with a feasibility statement.
In the case of a positive evaluation, the process development phase using about 20-100g of racemate is initiated. These investigations use a lab-scale SMB or micro-SMB and serve to offer experimental confirmation to the previously estimated process parameters. Further experimental runs and simulations are aimed at developing the SMB process according to the customer's specification e.g. purity and yield of the extract and/or raffinate.
The resulting database is used to determine a set of process parameters, which serves as the basis for the offer (cost and time) to the customer. Additionally, samples of extract and raffinate are supplied for analytical evaluation. This second set of results can usually be submitted in three to five weeks.
The next project phase involves the production of pure enantiomer at a scale of up to 100kg using either Bayer's lab SMB unit or at a scale of up to five metric tonnes per year using the company's cGMP pilot-scale SMB unit within its ZeTO pilot plant. This service can be combined in a beneficial way with almost any kind of chemical synthesis upstream or downstream of the separation step, allowing multi-step syntheses within one plant.
The synthesis of complex chiral molecules consists of multiple chemical steps, and the chiral separation of a racemic intermediate may be one of these steps.
A case study analyses the separation costs for the production of a chiral compound using a chemical resolution step compared with an SMB process.
The study considers the production of 500kg of a chiral compound through a six step synthesis. The following assumptions are made:
The case study examines the costs for the six-step sequence with chromatographic and chemical enantiomer separation at the beginning and the end of the sequence (figure 2). The results show that there are significant savings in production costs if the racemate separation is performed early on. This is due to the savings in material consumption and smaller batch sizes. The same holds true for the chromatographic separation.
Additionally, the higher yields of the SMB separation result in reduced material costs. For the example chosen, this advantage of better recovery of SMB over-compensates for the higher conversion costs compared with the classical chemical separation. Note: the incorporation of the chromatographic resolution step further downstream of the reaction sequence may be advisable due to potential racemisation.
Studies show that SMB chromatography is possibly more expensive than classical chemical resolution for a single chemical step, but it is clearly more cost- effective for a multi-step process.
Prior to production using the pilot plant, a lab-scale SMB unit is employed to determine SMB separation parameters and to produce small batches.
Bayer's commercial lab unit with 12 columns allows process optimisation and, subsequently, the production of up to 100kg of the desired enantiomer. The product streams are concentrated with two rotary-film evaporators. This lab unit is operated in a cleanroom environment.
purity versus productivity
The three dimensional chart (figure 3) shows an example of the correlation of extract and raffinate purity versus productivity for an executed separation. Firstly, the data indicates that both enantiomers can be obtained at close to 100% optical purity. The steep decline of productivity is observed as the purity of the product streams approaches 100%. The chart clearly demonstrates that a compromise regarding purity of extract and raffinate enhances the productivity of the separation significantly. This supports the frequently recommended combination of SMB separation with a subsequent crystallisation step.
Bayer's SMB unit is designed for an annual capacity of 100 to 5000kg of enantiomer. The setup allows the use of a two-component eluent with automatically adjustable composition. The racemate is fed to the SMB unit using two agitated vessels held at a constant temperature. This setup allows the identification of defined batches which are separately tested according to GMP rules. A typical snap-shot concentration profile of a successful SMB-separation for a racemate separation with the pilot plant is displayed in figure 4. The internal concentration of both enantiomers is detected using the on-line UV detector. The curves demonstrate the concentration profile during the start-up phase of the separation. The blue and brown curves represent steady-state conditions. Both product streams (extract and raffinate) meet the required specifications quickly.
The process parameters have been derived from earlier experiments using the lab-scale SMB unit and adjusted to the larger scale using linear extrapolation based on the square of the ratio of column diameters. This confirms the assumption that the scale-up of the chiral separation using the simulated moving bed process is a straight-forward process, at least up to the order of eight.
conclusions
The impact of the improved efficiency of CSP on chromatographic separations has been described. Modern CSPs are stable, versatile, and readily available. Their development provided the basis for the successful realisation of SMB separations as an alternative to classical chiral resolution methods.
Typically, larger-scale separations using SMB are economically attractive if productivity values are in excess of 1000 grac/(kgCSPd); very favourable cases show productivities of more than
5000 grac/(kgCSPd). But specific conditions and constraints must be considered.
The key economic advantage of SMB is the high recovery of the target enantiomer. SMB separations are usually introduced at the end of the sequence of synthesis steps. However, the chromatographic resolution step is also advantageous at an earlier stage.
Another positive feature of the separation is the fact that the 'wrong' enantiomer is usually isolated as well. In many cases, this allows racemisation and recirculation of the racemate to reduce material costs even further.
Finally, the typical sequence of activities for an SMB project shows that, for a suitable separation, the chiral compound can be delivered in a comparably short period of time with limited experimental efforts. This helps in the need for the faster development of new drugs.
New class of chiral chromatography
ZirChrom Separations, a US manufacturer of zirconia-based high performance chromatographic materials, has launched a project to develop two new classes of chiral stationary phases (CSPs). The first of the new CSPs is based on the permanent covalent attachment of a variety of chiral selectors to carbon-clad, porous zirconia microspheres. The second new class of CSPs is based on the modification of zirconia through the use of catecholic anchors via coordination chemistry. The goal of the project is to test the feasibility of developing at least two different type phases for analytical scale liquid chromatography. If successful, the new phases will allow for increased speed in analytical separations and higher throughput in preparative scale separations. The project will use the carbon chemical vapour deposition technology developed by ZirChrom. One of the major innovations in our approach is the use of a nonpolar surface," said Dr Clayton McNeff, vice president and technical director. When used with normal phase eluents, this new concept offers several important advantages over current state-of-the-art polar substrates, such as silica. We anticipate significant improvements in enantiomeric selectivity by minimising the achiral contributions to retention that arise from the polar interactions of analytes associated with silica surfaces," he explained.