Milling Made Easy: Nanoindentation as a Predictor of Bulk Properties

December 1, 2004
M.J. Snowden, J.C. Mitchell, Lisa J. Taylor, P.J. Dunn, D.G. Papadopoulos
Pharmaceutical Technology Europe
Volume 16, Issue 12

Size reduction of materials through comminution is employed in many industries, including agrochemicals, minerals, ceramics and pharmaceuticals. The reasons for particle size reduction depend on the industry in question. Within the pharmaceutical industry, a large percentage of products are formed from powders and undergo processing to improve dosage form properties. Particle size reduction prior to compacting to tablets can aid with dissolution and homogeneity. Such processing of powders is, in part, dependent on their mechanical properties and balancing these properties is crucial in achieving desired manufacturing performance. Generally, pilot-scale milling trials are run to determine the most effective and efficient mill and operating conditions for each material. These trials, however, require relatively large quantities of material as well as time, and are normally run in early development when sufficient material becomes available. Hence, it would be highly beneficial to identify a physical property..

Size reduction of materials through comminution is employed in many industries, including agrochemicals, minerals, ceramics and pharmaceuticals. The reasons for particle size reduction depend on the industry in question. Within the pharmaceutical industry, a large percentage of products are formed from powders and undergo processing to improve dosage form properties. Particle size reduction prior to compacting to tablets can aid with dissolution and homogeneity. Such processing of powders is, in part, dependent on their mechanical properties and balancing these properties is crucial in achieving desired manufacturing performance. Generally, pilot-scale milling trials are run to determine the most effective and efficient mill and operating conditions for each material. These trials, however, require relatively large quantities of material as well as time, and are normally run in early development when sufficient material becomes available. Hence, it would be highly beneficial to identify a physical property of a material that could be linked to the fracture behaviour of a drug substance during milling and which could be determined using a small-scale technique. This could enable the prediction of bulk material comminution from a small amount of substance, leading to significant reductions in both time and costs during production.

Understanding particle fracture

When a stress is applied to a material, two common mechanisms occur. First, the substance could undergo deformation (either plastically and/or elastically) or it could fracture. In practice, a combination of these mechanisms takes place. It is believed that both deformation and fracture mechanisms play an important role in particle breakage and thus milling performance. Whilst these mechanisms have been described and analysed individually, there have only been a few attempts made to describe their combined effect and their relationship to bulk behaviour. Therefore, a breakage parameter should ideally provide a representation of the competing mechanisms that a material undergoes when under stress. One parameter that has been reported in the literature is a brittleness index,

1

which is the ratio of material hardness (resistance to deformation) and toughness (resistance to fracture). It is hypothesized that materials that fragment extensively during milling have a high brittleness index.

The technique

Indentation is one small-scale technique that enables a measure of both hardness and toughness, and was the method employed by Lawn

et al

.

1

The advantage of using indentation methods compared with traditional compaction-based techniques is that the mechanical properties are measured on single crystals in their constituent form, thus the formation of compacts before testing is not required. As such, the amount of material required for investigations is minimal, the brittleness index of pharmaceutical materials can be determined from a sample containing only ten suitable crystals (crystals should free of debris and be larger than 100×50 μm). An additional benefit is that the properties measured are those of the crystal not the bulk material that has undergone compaction and, therefore, deformation and fracture processes.

Indentation allows the measurement of hardness (resistance to plastic deformation) and Young's modulus (resistance to elastic deformation) at the near surface, by loading and unloading a probe into a sample. The analysis method used to determine plastic and elastic behaviour depends on the type of indentation being employed. Continuous recording or depth-sensing indentation, the technique employed in this study, works by measuring the displacement of a sharp indenter into the sample as a function of the applied load. Because of the low loads applied, this method is also known as nanoindentation. The resulting plot is known as a hysteresis curve, which can act as mechanical fingerprint. Analysis of this curve enables the calculation of both hardness and Young's modulus. There are a number of methods available for the analysis of load-displacement curves; the technique employed here was developed by Oliver and Pharr.2 The data required for such calculations is peak load, maximum depth and contact stiffness at the initial stage of unloading.

Indentation of brittle materials can often lead to cracking around the indent impression. The formation of these cracks enables the calculation of fracture toughness using fracture mechanics. Numerous semi-empirical models have been proposed to describe indentation fracture toughness.3 A common one is shown in Equation 1.

Table I: Mechanical properties of test materials. These models remain semi-empirical, as the stress fields produced during indentation are complex and not fully resolved. In general, the calculation of toughness is dependent on the elastic and plastic stress field around the indent, applied load and the cracking around the indentation impression. One important feature in this analysis is the relationship between applied load and indentation cracking, which is found to be constant for a given material.

Table II: Comparison of brittleness indices and size reduction ratios.

As a feasibility study, nanoindentation has been used to measure the mechanical properties and thus calculate the brittleness index of single crystals for a number of active pharmaceutical materials prepared by commercial crystallization processes.

Nanoindentation was carried out using a Nanotester 600 (Micro Materials Ltd, UK). A schematic of the instrument is shown in Figure 1. The indenter is loaded against the sample by passing a current through the coil that is then drawn to a permanent magnet. Displacement of the indenter into the sample is measured by the variation in voltage between the capacitance plates. The sample holder is aligned with the indenter by means of three micrometer stages. This arrangement is mounted on a separate stage, and allows movement between the indenter and a high-resolution zoom microscope, which allows high precision selection of areas for indentation.

After indentation, a scanning electron microscope (Amray 1820T; Amray, USA) was used to acquire micrographs of the indented crystals. The indent diagonals and the cracks were measured and from these the fracture toughness (Kc) and brittleness index were calculated.

The bulk milling behaviour of the test materials was investigated using pilot-scale Apex hammer mill trials to generate size reduction data. The data was used as a comparison to the nanoindentation brittleness indices. A number of particle sizing methods were employed to measure both input and output particle size, depending on the difficulty of handling the crystals. The size reduction was then calculated as a percentage using Equation 3.

Results

Deformation behaviour

Comparison of hardness and Young's modulus indicated that the technique was able to clearly distinguish between different materials (Table I). Good reproducibility was observed using a minimum sample size of five crystals. Some deviation in results is expected for two reasons. The contact between crystal surface and indenter varies between runs and there is a high degree of surface roughness associated with bulk-supplied crystals as their growth is not controlled to allow the formation of flat smooth crystals. A wide range of deformation behaviour was observed for the materials examined. As the hardness and Young's modulus decreased the amount of deformation increased. Therefore, compound B is very plastic, whereas compound A is relatively hard and undergoes little deformation.

Fracture behaviour

On examination of the SEM micrographs of the indented samples, a wide range of breakage behaviour was observed for the drug substances examined. Figure 2 shows the indentation of compound A. Plastic deformation during the process has left a clear impression of the indenter. From two of the three corners of the indent a crack can be seen, a third crack began midway along one side of the indent. Once these cracks were initiated, the material fractured very easily. At the other extreme, compound B was found not to fracture under the load conditions employed in this study. As Figure 3 shows, a clear indentation impression can be seen with a small quantity of microtears in the centre of the indent. Also a small quantity of material was seen to have built-up around the edges of the indent, known as pile-up. Both of these features are typical of the occurrence of plastic deformation.

Figure 1: Schematic of Nanotester 600.

Predicting milling behaviour

The brittleness indices indicated that compound A was very brittle and easy to mill, whereas compound B, which showed no sign of fracture, was the least brittle. The importance of using the brittleness index is highlighted by a comparison between compounds A and C. On the basis of the fracture toughness values alone (0.019 MPa m

1/2

and 0.02 MPa m

1/2

respectively) it would be assumed that the two compounds would show similar breakage propensity. When the plasticity of the materials was taken into account, as part of the brittleness index, this was not the case. The much lower hardness value for compound C, compared with compound A, suggests a greater amount of plastic deformation took place as a result of the applied stress.

Table II provides a comparison between the brittleness indices and the size reduction ratios determined from the milling trials. A correlation between the two sets of data was observed, which may have the potential to classify materials with respect to their milling behaviour. The results can be clearly divided into three categories of difficulty, which can then be related to preferred mill type. Category one would be materials easy to break and are likely to be suitable for hammer milling. Category two would be moderate materials, which require more energetic milling to reach a suitable particle size reduction, in such cases air-classifying mills would be favoured. The final category would be those compounds that are difficult to mill and these compounds would generally require jet milling.

Figure 2: SEM micrograph of compound A at 32000 magnification.

Summary

The brittleness index has been identified as a breakage parameter, which correlates to bulk milling behaviour and measured using a nanoindentation technique. The results indicate that particle size reduction from pilot-scale milling trials could be related to the brittleness index. Hence, the brittleness index determined from crystal experiments can be employed to provide a guide to the most suitable mill type for a material. Whilst this technique is promising, other factors such as particle size and shape should be considered, as these may have a strong influence on the choice of a suitable mill.

References

1. B.R. Lawn and D.B. Marshall, "Hardness, Toughness, and Brittleness: An Indentation Analysis,"

J. Am. Ceram. Soc.

62

(7-8), 347-350 (1979).

2. W.C. Oliver and G.M. Pharr, "An Improved Technique for Determining Hardness and Elastic Modulus Using Load and Displacement Sensing Indentation Experiments," J. Mater. Res. 7 (6), 1564-1583 (1992).

3. G.R. Anstis et al., "A Critical Evaluation of Indentation Techniques for Measuring Fracture Toughness: I. Direct Crack Measurements," J. Am. Ceram. Soc.64 (9), 533-538 (1981).

4. R.D. Dukino and M.V. Swain, "Comparative Measurement of Indentation Fracture Toughness with Berkovich and Vickers Indenters," J. Am.Ceram.Soc.75 (12), 3299-3304 (1992).