Chris Blatchford 1, Mark Saunders 2, Kalyan Potluri 2, Simon Gaisford, 3, Graham Buckton, 3 1 3M HealthCare, Drug Delivery Systems Division, Loughborough, United Kingdom; 2 Pharmaterials Ltd, London, United Kingdom; 3 The London School of Pharmacy, London, United Kingdom During the manufacture and processing of active drug substances, amorphous regions or defects may be introduced into the sample due to the shear physical forces. These forces can result from man processes including milling, impaction, compression, attrition, etc. Although the percentage of amorphous content introduced in this way is usually low (of the order of ~1% w/w), its location mainly on the surface of what are usually small particles gives it a disproportionate influence on the properties of the material, such as altering the flow properties, specific surface area, compressibility etc. It is therefore important to control and monitor the concentration of amorphous material in a product. Amorphous material is in a metastable physical state relative to the crystalline counterpart and this can have a detrimental affect on product stability, in particular on storage at elevated temperatures/humidities. It is necessary to have appropriate analytical methods and standards to be able to quantify amorphous content. In this study, the recrystallisation kinetics and stabilities of amorphous  - lactose monohydrate was investigated using the techniques of Gravimetric Vapour Sorption Analysis and Hyper TM -Differential Scanning Calorimetry (DSC).. Introduction Methods From these studies, a varied and reduced physical stability of the amorphous component of processed  -lactose monohydrate (via micronisation) was observed. This has not been reported previously. It is postulated there is a reduction in the activation energy barrier for crystallisation due to crystal seeding (secondary nucleation) and suggests that care must be taken when producing standards for determining amorphous content and determining the physical stability of processed materials. Conclusions 3M Drug Delivery Systems Figures 1 and 2 show the resulting fast-scan DSC profiles of crystalline and micronised  -lactose monohydrate heated at a scan rate of C Header Here From the DSC thermogram shown in figure 1, two distinct endothermic transitions can be seen onset ca. 150 and C, corresponding to the thermally induced dehydration and subsequent melt of the anhydrous  -lactose crystalline lattice. However, for the micronised sample (figure 2) the dehydration endotherm was split, with an additional endotherm at lower temperature and the onset temperature reduced to ca C. This two-phase water loss was not observed using conventional scanning rates (10 0 C/min), however a low temperature broadening and onset of the dehydration endotherm was observed (fig 3). Note the thermograms in Figure 2 and 3 were terminated at a lower temperature, prior to the melt temperature. Figure 4: Kinetic water sorption,desorption isotherm obtained for micronised  -lactose monohydrate. Inset shows the relatively rapid kinetics of the first process Assessment of the re-crystallisation behaviour of micronised alpha-lactose monohydrate Enabling your success Figure 1: DSC thermogram of 100 % crystalline  -lactose monohydrate heated at a scan rate of C Results Materials Pure  -lactose monohydrate (supplied by 3M) was micronised in an air jet microniser (Glen Creston, type: 81 – 010). Compressed nitrogen at 50 psi was used for the process of micronisation. Five cycles of micronisation was performed to achieve a thoroughly micronised sample Methods Hyper TM -DSC Approximately 8-10 mg of sample (accurately weighed) was placed into an aluminium DSC pan and crimp sealed to ensure a tight fit. The sample was then loaded into a Pyris 1 DSC instrument (Perkin-Elmer) fitted with an intra- cooler 2P cooling unit and subsequently cooled to C. Once a stable heat- flow response was attained, the sample was then heated to C at a heating rate of either 10 or C/min and the resulting heat-flow response monitored. Prior to analysis the instrument was temperature and heat-flow calibrated using indium and zinc reference standards. A nitrogen purge gas at a flow rate of 20 ml/min was used to provide an inert atmosphere. Gravimetric Vapour Sorption Analysis (GVS) Micronised  -lactose monohydrate (ca. 20 mg accurately weighed) was loaded into a 42µm mesh cone shaped stainless steel sample container and suspended upright within the sample chamber of the IgaSorp GVS instrument (Hiden Analytical, UK) held constant at 25 0 C (+/ C). The sample was then dried by maintaining a 0 % humidity environment until no further weight change was recorded. Subsequently, the sample was subjected to a ramping profile, maintaining the sample at each step until equilibration had been attained (99.5 % step completion). Upon reaching equilibration, the % RH within the apparatus was ramped to the next step and the equilibration procedure repeated. After completion of the sorption cycle, the sample was again dried to constant weight at 0 % RH. The weight change during the sorption/desorption cycles was calculated so that the hygroscopic nature of the sample could be determined. Figure 2: DSC thermogram of micronised  -lactose monohydrate heated at a scan rate of C Figure 3: DSC thermogram of micronised  -lactose monohydrate heated at a scan rate of 10 0 C It is proposed that the water lost during dehydration occurs via a two-stage process. The initial fast phase is the direct result of desorption of water from the surface of the smaller micronised particles (higher surface area available for water evaporation) followed by the somewhat slower desorption of water from the bulk or core of the particles. Prior to the dehydration step a small exothermic response was observed which has been attributed to the collapse and recrystallisation of amorphous material formed during micronisation. This transition was not observed under slow scanning conditions (10 0 C/min). It is considered this is due to the decreased sensitivity of the instrument at slower scanning conditions. GVS analysis was performed on the micronised  -lactose monohydrate sample to determine the stability of the sample (See Fig 4). Upon exposure to mild conditions of relative humidity (40 % RH), an initial fast weight loss profile was observed. This weight loss event is common for samples containing significant quantities of amorphous material, where the glass transition temperature is reduced to below that of experimental study temperature (25 0 C), a result of plasticisation by the water vapour present. This is then followed by the subsequent crystallisation of the amorphous fraction produced during micronisation. After completion of the initial rapid transition a much slower crystallisation process was observed at higher %RH. This process was seen at each incremental increase of % RH, indicating that complete crystallisation does not occur until continued exposure to high levels of relative humidity (> 80 % RH). This effect has previously been observed during Atomic Force Microscopy studies (Price and Young, 2004), which showed that only a portion of the bulk collapsed particles undergo primary nucleation and crystal growth, with complete crystallisation was achieved after high elevated % RH levels (94 %). In these studies, the initial crystallisation event of the amorphous phase was seen to occur at lower relative humidity levels than previously reported for amorphous  -lactose monohydrate (Briggner et al.) suggesting a destabilisation of the amorphous state and crystallisation under milder conditions. It is proposed that this initial fast phase crystallisation event occurs as a direct result of crystal seeding (secondary nucleation) of the amorphous fractions located on the surface of the crystalline particles, which dramatically reduces the activation energy barrier required for nucleation and subsequent crystal growth. References International Journal Pharmaceutics (2): , L. Briggner, G. Buckton, K. Bystrom and P. Darcy Journal Pharmaceutical Sciences (1):155-64, R. Price, P. Young