19 th INTERNATIONAL CONGRESS ON ACOUSTICS MADRID, 2-7 SEPTEMBER 2007

Transcription

1 19 th INTERNATIONAL CONGRESS ON ACOUSTICS MADRID, 2-7 SEPTEMBER 2007 POWER ULTRASOUND AND SCALE-UP: FLOW-CELL REACTORS, CRYSTALLISATION, AND PARTICLE ENGINEERING PACS: Cs Ruecroft, Graham Prosonix Ltd, The Magdelen Centre, Robert Robinson Avenue, Oxford Science Park, Oxford, OX4 4GA, UK; ABSTRACT Power ultrasound is important in crystallisation control including nucleation, size distribution and morphology. Mesoscopic particles for drug inhalation are manufactured by primitive pharmaceutical technologies such as micronization. These particles can be prepared using power ultrasound assisted technologies such as Solution Atomisation and sonoxtallization (SAX TM ) technology being developed by Prosonix and the University of Bath, UK. This allows the production of spherical drug particles with superior geometrical, surface and performance properties, nanosuspensions, pharmaceutical co-crystals and combination-based products. Sononucleation is also very effective in conventional batch crystallisation and antisolvent precipitation enabling control of crystal size distribution, polymorph, morphological control, elimination of impurities, and improved solid-liquid separation behaviour. It is now possible to use power ultrasound at industrial scale for pharmaceutical manufacture. Industrial equipment to allow effective and focussed distribution of acoustic energy into a liquid by using a number of low-power transducers bonded to the outside of a cylindrical duct is now available. The scaleout feature of the technology ensures that success in the laboratory can be replicated across scale. The in-line continuous flow or batch mode process can be applied to intermediates, excipients, APIs, binders and sugars, and importantly can be validated across scale in cgmp environment. INTRODUCTION The application of power ultrasound ( khz) to crystallisation and chemical processing is an intensification technology that has undergone serious development over the past 15 years or so and has a significant future ahead. It has been known for 80 years that ultrasound can have interesting effects on chemical and biochemical systems as well as influencing events in crystallisation but the absence of scale up equipment led to that intransigence. However, recent advances in equipment have made its implementation at industrial scale feasible. There is a lot of interest in the application of ultrasound to crystallisation in the pharmaceutical and fine chemicals sectors of industry. Prosonix has used its experience in crystallisation and ultrasonic engineering to design and build ultrasonic processing equipment (referred to as Prosonitron for commercial purposes) to allow production of pharmaceutical ingredient with desirable morphology and particle size distribution [1]. The equipment allows distribution of acoustic energy into a liquid very effectively by using a number of low-power transducers (now 21 in a 5 L flow-cell) bonded to the outside of a cylindrical duct. This avoids the problems of using high-powered probe based equipment where metal particles can be shed into the crystallising liquor. Typical equipment for pharmaceutical manufacture fabricated from Hastelloy is shown in Figure 1. This equipment can be used as a recirculation or continuous flow-cell. Perhaps the ultimate in sonocrystallisation is the rapidly developing technology that to produce micro and nano-crystalline particles for drug delivery Solution Atomisation and Crystallisation by Sonication [2], SAX. SONONCRYSTALLISATION Sonocrystallisation [3] utilises power ultrasound and resultant acoustic cavitation to assist in nucleation of metastable solutions and subsequent crystal growth. The reasons why such local and transient energy concentrations assist with nucleation are difficult to explain but local 1

2 dramatic temperature and pressure changes, shockwaves and rapid local cooling rates may all contribute to nucleation in the regions of the supersaturated solution or perhaps we can simply overcome the energy barriers associated with nucleation. Cavitation appears to be particularly effective as a means of inducing nucleation in a controlled and reproducible way and this provides a well-defined start point for the crystallisation process. Sononucleation can also eliminate the requirement to add seed crystals, which can be particularly advantageous in contained pharmaceutical crystallisation processes. Reduced acoustic power levels can lead to streaming effects, which can be useful to help crystal growth, rather than stable or violent transient cavitation. Figure 1.-Ultrasonic Processing Equipment for Pharmaceuticals Metastable solutions and ultrasound In many respects the ease or difficulty of carrying out a crystallisation process can be linked to an understanding of the metastable zone (MZ). Typically the application of high-intensity 20 khz ultrasound can lead to narrowing of the metastable zone width (MZW). This MZW narrowing can range from a few degrees to twenty or more when crystallizing sugars. By narrowing the MZW it is possible to tailor a crystal size distribution between the extreme cases of a short burst of ultrasound to nucleate at lower levels of supersaturation and allow growth to large crystals, and the production of small crystals via continuous (or perhaps a longer single burst) insonation throughout the duration of the process. The optimum needs to be determined by experimental investigation. Is possible that ultrasound may also induce secondary nucleation by mechanically disrupting crystals or loosely bound agglomerates that have already formed. The overall technique lends itself extremely well to almost any crystallisation process including valuable pharmaceuticals and polymorphic systems. An attractive feature of the sonocrystallisation technology is that it can be applied at any stage in a product pipeline. This scale-out feature of the technology ensures that success in the lab can be replicated across scale. The in-line continuous flow or batch mode process can be applied to intermediates, excipients, APIs, binders and sugars, and importantly can be validated across scale in cgmp environments. Polymorphism is common amongst organic materials resulting in the existence of two or more crystalline phases with different packing in the crystal lattice. Isolation of the wrong polymorph brings substantial problems in pharmaceutical applications. Application of ultrasound to a crystallizing system can help in producing the ground state polymorph (the most thermodynamically favoured and least soluble). For example in a system that exhibits enantiotropic polymorphism (Figure 2), sonocrystallisation, using a pilot scale recirculatory system (~500 L crystallizer and 5 L Prosonitron), allows preparation of the thermodynamically stable polymorph (cool along solid line), which has cubic type crystal habit, at low supersaturation. Conversely, at high supersaturation (cool along dotted line) fast nucleation kinetics, along with poorly controlled crystallisation, leads to the proliferation of the kinetic 2

3 (metastable) polymorph, which has a distinct needle habit, and in turn results in poorly stirred slurries and variable product bulk density. Figure 2.-Sonocrystallisation and Polymorph Control SCALE-UP AND EQUIPMENT DEVELOPMENT One of the most important barriers to the adoption of power ultrasound technology in manufacturing has been the lack of suitable equipment for use in industrial environments at the scale required particularly where flammable solvents are in use. Most discoveries in sonochemistry and sonoprocessing have been carried out in laboratories on the mg g scale using either high-intensity probe or bath systems. There is a fundamental requirement for equipment that may be operated simply and reliably at the kg tonne scale in the manufacturing of fine chemicals and pharmaceuticals and importantly in an explosion proof environment. For bulk commodity chemicals manufacturing, scales of at least an order of magnitude larger than this would be required. Although the cost-benefit basis of the technology makes it less attractive for this type of application, all processes should be examined on a caseby-case basis, as evidenced by our work in bulk alumina production [4]. The equipment designed and manufactured by Prosonix for sonoprocessing avoids resonance, standing waves and coherent wave relationships. There are a number of advantages in using non-coherent ultrasound: The more even distribution of the ultrasound through the working fluid is a key benefit. In addition one can design equipment with greater flexibility in terms of dimensions, frequency and configuration. We embarked on non-coherent designs after our early experience with multi-transducer systems gave rise to difficulties with transducers tuning in to each other and of mechanical resonance in system components. Such difficulties can undoubtedly be more easily overcome with modern transducers and system designs. The original multi-transducer designs, were based on a 4 5 L insonated volume as a cylindrical duct 120 mm in diameter, fitted with three radially mounted transducers.[5, 6] To reduce surface erosion and mechanical stresses at the point of contact, a liquid barrier was employed between the transducer and the duct wall. This unit was designed for non-coherent ultrasonic operation with a nominal frequency of 20 khz. Experience of operating the unit showed that the three transducers tuned to slightly different frequencies between 19 and 21 khz, and spot tests with aluminium foils and hydrophones indicated that the power input densities were reasonably uniform throughout the 4 5 L working volume. More recent developments have employed direct bonding of the transducer to the surface of the vessel [7]. Improvements in the bonding method, and a move to transducers with lower individual outputs, have enabled the move to systems with large numbers of transducers, to give an acoustic pattern that is uniform and non-coherent above the cavitational threshold throughout the working volume. The use of low-output transducers gives the additional advantage of avoiding the phenomenon of cavitational blocking (acoustic decoupling), which 3

4 arises where power densities close to the delivery point are very high. In addition these multitransducer units very effectively concentrate ultrasonic intensity towards the central axis of the cylinder and away from the vessel walls, thus reducing problems of erosion and particle shedding. This vessel can be operated in batch mode or, for larger scale work, in continuous mode whereby units can be combined in a modular fashion for scale-out and increased residence time. In summary, a plurality of low electrical and acoustic power (~1-3 Wcm -2 ) transducers produces WL -1 (ideally WL -1 ). They are attached to a 150 mm diameter cylindrical duct fitted with ANSI flanges for easy installation. The power can be applied continuously or pulsed. We have measured acoustic cavitation, using a broadband acoustic sensor developed by the National Physics Laboratory [8], at a variety of positions using a 21 transducer cell then mathematically modelled the data so as to generate a 2-dimensional image as shown in Figure 3. The key point is that the most significant transient cavitation takes place in the centre of the device. Figure 3.- Representation of acoustic cavitation in a 5L Prosonitron This fundamental design and supporting patent has been crucial in developing more advanced flow-cells. One such device is the hastelloy explosion-proofed cell (for use with flammable solvents in a zoned process area) used for the manufacture of pharmaceuticals (Figure 1). Scaling out the fundamental flow-cell design in one dimension facilitated the design and manufacture of the 1.2 m flow-cell used in the bulk scale alumina production at Aughinish Alumina in Ireland and Alcoa- San Ciprian in Spain (Figure 4), [4]. The flow cells, with 40 bonded transducers, are manufactured form 316 stainless steel with a hard chrome internal surface for added corrosion resistance. To provide the desired residence time in this continuous system 6 of these 1.2 m cells were joined together in a modular fashion. Figure 4.- Typical equipment and installation of power ultrasound equipment for alumina production 4

5 CRYSTALS AND PARTICLES ENGINEERERING The Solution Atomisation and Crystallisation by Sonication (SAX) process is quite simple in philosophy but is more complex with respect to underlying physics and chemistry. The drug substance is dissolved in a suitable volatile solvent, which then undergoes careful atomisation, followed by controlled evaporation in a temperature regulated inert atmosphere to produce a highly concentrated and viscous non-crystalline droplet. These droplets (effectively concentrated whereby the solute is non crystalline and in the pre-glass state) are then passed into a non-solvent, contained within an ultrasonic flow-cell such as the Prosonitron (Figure 5). Only when power ultrasound is applied to the droplet via the antisolvent medium does it undergo nucleation and crystallisation. Without ultrasound the particles solidify with poor crystalline form. Evaporation of solvent from droplets Formation of small viscous droplets Product isolated from slurry Droplets enter the non-solvent and undergo sonication generating crystals Figure 5.- Principles of SAX and advanced laboratory rig (left) SAX is a sonocrystallisation technique for production of particles with optimum size and morphology suitable for formulation where microcrystallinity is essential. There are specific benefits in the production of particles for inhaled therapeutics, and also see potential in the production of nanosuspensions, pharmaceutical co-crystals and combination-based products where 2 drug substances are contained within the same particle. The technique allows production of spherical particles as shown for Budesonide (a potent inhaled corticosteroid) in Figure 5 within a well-defined size range and with control of the macroscopic morphology, including polymorphism and surface topology. These are important characteristics in defining the performance of mesoscopic particles, in turn defined by aerodynamic properties, shelf life, stability, bioavailability and efficacy. It is particle geometry that is the central design principle in controlling surface forces and hence interfacial interactions. No ultrasound With Ultrasound change of process parameters with ultrasound Figure 5.- Top: Influence of ultrasound on particles of Budesonide Bottom: further modification of conditions 5

6 CONCLUSIONS Power ultrasound can be applied to crystallisation at manufacturing scale both for traditional batch crystallisation and new technologies to produce mesoscopic micron and nano particles for drug inhalation. There is no reason why sonocrystallisation, including the core methods outlined above, should not become a core technology in the pharmaceutical and fine chemicals industries. Processing techniques such as SAX should become superior technology for the manufacture of microcrystalline drug substances, complex APIs, combination particles, and nanosuspensions. References: [1] J.P. Perkins: World Patent. WO 2000/35579 (2000) [2] R. Price, J.S. Kaerger: World Patent WO 2004/ (2004) Pharmaceutical Res 21 No.2 (2004) 372 [3] G. Ruecroft, D. Hipkiss, T. Ly, N. Maxted, P.W. Cains: Sonocrystallisation: The use of Ultrasound in Industrial Crystallisation. Organic Process Research and Development 9, (2005) (24) [4] G. Ruecroft, D. Hipkiss, M. Fennell: Improving The Bayer Process by Power Ultrasound Induced Crystallisation (sonocrystallisation) of Key Impurities. TMS 2005, TMS Light Metals (2005) 163 [5] P.D. Martin, L.D. Ward, Trans IchemE. 70A (1992) 296. [6] C.L. Desborough, R.B. Pike, L.D. Ward: US Patent , (1997) GB Patent A (1991) European Patent A (1991). [7] J.P. Perkins, World Patent. WO 2000/35579 B1 (2000). [8] M. Hodnett, B. Zaquiri, R. Chow: Ultrasonics Sonochemistry 11 (2004) 441 6