Patricia Darcy

The use of isothermal microcalorimetry in the study of changes in crystallinity induced during the processing of powders

Abstract Isothermal microcalorimetry has been used to follow recrystallisation of amorphous regions of powder surfaces. Lactose monohydrate was taken as a model powder, and was processed by spray drying and micronisation. Spray drying produced an amorphous powder (as shown by X-ray diffraction), which was found to recrystallise when exposed to humidities over 50% RH. The recrystallisation process was extremely cooperative, with the entire sample recrystallising almost instantaneously, rather than a gradual process over the period of exposure to the water vapour. Similar results were noted when micronised material was investigated. The amount of amorphous material produced during micronisation was directly proportional to the intensity of the process. It proved possible to quantify the % amorphous content of powder sample with a resolution of at least 1%, which is considerably better than other techniques. The amorphous regions of the lactose crystallised as either α- or β-lactose. The difference between these samples could be detected by X-ray diffraction, and also could be seen by isothermal calorimetry, as the β-regions mutarotated to α-lactose. The application of isothermal microcalorimetry to studies of crystal properties of powders provides a quantitative characterisation of many aspects of crystallinity and crystal transition. The data obtained can subsequently be used to characterise the properties of the material, and to show how and when crystallisation will occur, and to aid predictions of the product of the crystallisation process. The demonstration of these applications provides a huge potetential for the use of isothermal microcalorimetry in this field of study.

The use of gravimetric studies to assess the degree of crystallinity of predominantly crystalline powders

An accurate humidity and temperature controlled microbalance system (dynamic vapour sorption) has been used to investigate the water sorption properties of mixtures of amorphous lactose (spray dried) and crystalline a-lactose monohydrate. From the shape of the sorption isotherms it could be seen that the first sorption process for each sample showed higher weight gain due to absorption into the amorphous regions, than was observed for the subsequent adsorption events on the sample. For the first sorption process, the weight fell after the humidity reached about 60% RH, as the amorphous material recrystallised (causing expulsion of absorbed water). On desorption there was always residual water after the first cycle, which correlated reasonably well with the residual that would be predicted if the amorphous material had been converted into the monohydrate form. The water sorption technique can readily be used to demonstrate the existence of amorphous material even for mixtures which contain as little as 0.05% w/w, and it is possible that the residual weight change may be used as an approximate quantification of the original amorphous content of the sample.

Assessment of disorder in crystalline powders--a review of analytical techniques and their application.

The need to be able to measure amorphous contents in crystalline powders is now recognised. In this review, calorimetric and gravimetric methods are reviewed in a way that should alert workers in the field to the theoretical, and practical considerations which are important to understanding how best to study crystalline samples which contain low levels of amorphous material. It is shown that vapour sorption techniques are very powerful as long as serious consideration is given to the choice of environmental conditions and the exact experimental methodology. As the amount of published work in this field grows, it becomes increasingly necessary to describe experimental and data manipulation methods in great detail.

Water mobility in amorphous lactose below and close to the glass transition temperature

Abstract The water sorption behaviour of amorphous lactose has been investigated gravimetrically. It was found that the kinetics of absorption at (especially) 40% and (also) 50% RH were bi-phasic. Although we have no explanation for this behaviour, it is noted that the inflection point between the two processes is at a 1:1 mole ratio of water:lactose. Equilibration at 40% RH results in an equilibrium uptake of 7% water, which is not sufficient to lower the Tg of lactose to the temperature of the experiment (T). Following from this, desorption is rapid and the rate proportional to the extent to which the RH has been lowered. If the sample is equilibrated to 50% RH the water content exceeds that which lowers the Tg below Tg this results in a collapse of the amorphous structure, but not in instantaneous recrystallisation. Exposure to higher humidities in an isothermal microcalorimeter revealed that the heat output for recrystallisation of the collapsed amorphous structure was indistinguishable from that produced on recrystallisation of the original expanded amorphous form. The rate of water desorption from the collapsed amorphous structure is slow and follows square root of time dependency. The rate of this diffusion controlled process is not altered by changing the external RH. The duration of exposure to 50% RH alters the extent of collapse, and hence alters the amount of water which is free to leave the sample rapidly and that which is released by the slow diffusion through the solid. After reducing the RH the water content of the collapsed structure remains high, but the recrystallisation is greatly delayed. These studies show that water can be held in different ways within amorphous lactose and this has implications for physical, chemical and potentially even microbiological stability of products.

The use of isothermal microcalorimetry in the study of changes in crystallinity of spray-dried salbutamol sulphate

Abstract Isothermal microcalorimetry has been used to monitor the recrystallisation of spray-dried salbutamol sulphate. The drug recrystallises in water vapour, by a cooperative process. The cooperative nature demonstrates that the water must first absorb to saturate the entire powder bed before recrystallisation occurs. Consequently, recrystallisation is slower for low humidities, due to a slower arrival of water vapour. The data have been compared with previous data for recrystallisation of spray-dried lactose. The heat change for the crystallisation was significantly lower for salbutamol sulphate than for lactose. In terms of apparent enthalpy of crystallisation, the large exothermic responses are indicative of the fact that the crystal form is the thermodynamically stable state. The salbutamol which had been recrystallised at the lower humidities showed that the process, whilst being rapid, was discontinuous. In each case, the exothermic recrystallisation was followed by an endothermic response for the expulsion of water as the amorphous region recrystallised. There was a repeating sequence of crystallisation, followed by water expulsion, followed by further recrystallisation. With each repeat of the cycle the responses decreased in size. This ability to follow crystallisation in real time provides a novel insight into the process.

The influence of additives on the recrystallisation of amorphous spray dried lactose

Abstract Amorphous material in crystals can constitute reactive ‘hot spots’, which can be centres for chemical degradation or physical transitions, leading to product instability. Problems have been encountered in studying small amounts of amorphous content for powdered systems, due to poor sensitivity of the majority of techniques. Isothermal microcalorimetry has been shown to have good resolution for cases where the amorphous content of the powder can be made to recrystallise in the instrument. In this study amorphous lactose has been investigated, being recrystallised by exposure to air at 75% RH. The lactose has been studied in two layers separated by varying amounts of glass beads (inert carrier), magnesium stearate (hydrophobic excipient), or microcrystalline cellulose (hygroscopic excipient). Significant differences were observed in the time needed to cause recrystallisation when amorphous material was separated by these different additives. Glass beads had only a small effect, but magnesium stearate caused an increased lag time prior to the recrystallisation event. In both these cases the lactose all recrystallised at one time, even though it was divided into two physically separated regions. A layer of microcrystalline cellulose between two layers of amorphous lactose resulted in a long lag time prior to recrystallisation, as it removed considerable amounts of water vapour from the atmosphere, thus preventing saturation of the lactose. By varying the weight of amorphous lactose in the upper and lower layers, and comparing data with the results obtained for homogeneous mixtures, it was possible to postulate a mechanism for the cooperative recrystallisation process. In essence, the water vapour is absorbed into the upper layers of the sample, and then transferred away yielding a concentration gradient through the entire sample in the cell. As the water content gradually increases in the lower layers, the rate of water absorption can become more rapid than the rate at which water is transferred away from the surface. After this time the surface saturates, starts to recrystallise thus liberating a great excess of water vapour, which is sufficient to cause the lower layers of powder to become saturated and also to recrystallise.

The influence of heating/drying on the crystallisation of amorphous lactose after structural collapse

This study was designed to investigate the influence of collapse of amorphous lactose on its subsequent behaviour during drying, or other processes which cause increases in the temperature of the material. Amorphous lactose was prepared by spray drying from aqueous solution. The solid was dried and then exposed to 50% RH for various times in order to induce different amounts of collapse in the amorphous structure. All the samples remained amorphous for the range of exposure times used. During heating in a differential scanning calorimeter, the non-collapsed material crystallised at ca. 180°C to give mostly α-lactose, with some β-lactose present. The collapsed lactose crystallised at ca. 70°C and yielded mostly β-lactose, with some α-lactose monohydrate present. It can be concluded that the collapsed structure will crystallise on drying at lower temperatures than the non-collapsed lactose. The non-collapsed material rapidly loses its sorbed water (this would occur during the early stages of drying), whilst the collapsed lactose loses its water suddenly during crystallisation. Thermogravimetric analysis revealed (generally) three distinct water loss peaks for the collapsed structure, two of which were believed to be due to crystallisation occurring and the final one being the loss of water of crystallisation. The sudden loss of water from the collapsed material will make a substantial contribution to the free water content of a formulation and as such could cause confusion during drying processes. Material which was partially collapsed behaved in an intermediate manner between non-collapsed and totally collapsed samples.

The extent of errors associated with contact angles 3. The influence of surface roughness effects on angles measured using a Wilhelmy plate technique for powders

Abstract Assessment of the wettability of powders is known to be problematic. It is known that the use of sessile drops on liquid saturated powder compacts can tend to underestimate the value of contact angle. The use of compacted plates of powder in a Wilhelmy plate experiment allows the calculation of a contact angle for a liquid on a powder without the need to pre-saturate the powder bed. As part of the Wilhelmy plate approach, it is necessary to measure the perimeter of the plate. In this study the effect of surface roughness on the error in measured perimeter is estimated. Propyl p-hydroxybenzoate was compacted into rectangular plates which were each divided into half. One half was used for contact angle measurements and the other viewed by scanning electron microscopy. The effect of surface roughness on measured angle was estimated. By coating the plates with gold, and then measuring a contact angle, the fact that the surface energy of the plate was known was used to allow an estimate of the effective (rather than measured) plate perimeter. It was found that the perimeter was 1.78 times greater than the measured value, giving a significant underestimate of the contact angle.

Quantitative assessments of powder crystallinity: Estimates of heat and mass transfer to interpret isothermal microcalorimetry data

Abstract Isothermal microcalorimetry is used to study small quantities of amorphous materials in crystalline powders. The aim of this work is to better understand the isothermal microcalorimetry measurement with regard to the quantification of amorphous contents of materials. Amorphous lactose was crystallized in a sealed ampoule in an isothermal microcalorimeter at a range of temperatures (25–60°C) and humidities. Identical heat changes for crystallization were observed at all humidities at 25°C; however, the measured heat varied with humidity at higher temperatures. The heat measured by isothermal microcalorimetry was approximately the difference between the heat of crystallization and the heat of vaporization of the desorbed water. The isothermal microcalorimetry output for this process is now better understood and it can be seen that, in order to obtain quantitative data for crystallinity, it is necessary to have a slow supply of vapor. As the measured heat change is related to the extent of water desorption, care must be taken when using microcalorimetry to quantify the amorphous content of powders, especially when comparing data generated at different environmental conditions.

Crystallization of bulk samples of partially amorphous spray-dried lactose.

The crystallization of partially amorphous spray-dried lactose was studied as a function of sample size. Crystallization occurred gradually over a period of 80 hr for a 95-g sample. The water content during crystallization was lower than that needed to cause crystallization if it had been distributed evenly throughout the bed, thus the absorbed water must have been unevenly distributed. The weight of the sample continued to change for days after crystallization was completed, because of the slow desorption of condensed water and the very slow formation of the hydrate form. Surprisingly, all samples with a weight between 42 and 95 g were found to take up the same mass (not percent) of water at the same time. This provides further evidence that the water was not evenly distributed throughout the sample. Water loss after this peak differed in the different weight samples, with the largest weights resulting in the lowest residual weight after 2 weeks. Only the sample of 22 g load had a different peak weight and a much lower weight loss after crystallization. This study provides detail of how partially amorphous bulk samples crystallize.