Trace evidence examination using laboratory and synchrotron X-ray diffraction techniques

On 7 August 2007, Corryn Rayney, a Supreme Court registrar, failed to return home after a dance class at a nearby Community Centre in Bentley, a suburb of Perth, Western Australia (Fig. 1). Her body was subsequently recovered on 16 August 2007 after transmission oil from her broken-down, abandoned car lead police to a clandestine grave site along a dirt track in Kings Park, also located in Perth (Fig. 1). Soil evidence from the victim's body, clothing, home, vehicle, Community Centre and grave site were analysed extensively in order to identify the likely movement of the victim after she left the Community Centre. The Perth coastal plain consists of predominantly coarse sandy soils that show only minor mineralogical variation throughout the region, and consequently bulk soil mineralogy was of little forensic use. However, very small (often <0.5 mm) red particles were observed on the victim's body, hair, clothing (mainly her bra), in her vehicle and in the body bag used by the Coroner to transport the body. These red particles had the appearance of brick fragments. The red particles were not found at the surrounds of the grave site and therefore, if the source location of the red particles could be determined, they could provide information on the movement of the victim prior to burial. The question that had to be answered by forensic investigators was: could these very small red particles be characterized to such an extent that they could be associated with bricks from a known location?

Recycled red house bricks were used extensively around the victim's home for the paving of driveways (Fig. 2), footpaths and pergola areas. These bricks were recycled from houses constructed prior to the 1940s when weaker lime mortar was replaced with stronger Portland cement mortars: lime mortars being more easily remove from bricks than Portland cement mortars. Brick pavers were also present on pathways at the Bentley Community Centre. Raw materials for house bricks and pavers are sourced from many localities in the foothills to the east of the Perth central business district (CBD), and have extensive variability in mineralogical composition, manufacturing processes and firing conditions. The bricks used at the victim's home for paving areas (Fig. 2) and at the Bentley Community Centre range in colour from red to yellow/pale-brown, hardness from soft to very hard, and durability from weak to very strong. To clarify the movement of the victim prior to burial, the very small red particles and the bricks at the victim's home and at the Bentley Community Centre needed to be subjected to rigorous forensic analysis.

The aim of soil forensic analysis is to compare earth materials taken from questioned items, such as clothing, shovels or vehicles, with a putative source or sources (e.g. specific control locations or crime scenes). Elimination of sources is usually unequivocal but failure to exclude is possible association. Earth materials are powerful, perhaps ideal, pieces of contact trace evidence that help in criminal investigations, as outlined by Fitzpatrick (2013a, b). A wide variety of earth materials such as soils, rocks, minerals and human-made mineral particles (e.g. bricks) can be used to indicate or compare provenance, and therefore be used as intelligence and subsequently evidence to narrow areas of search during an investigation. Evaluative comparison of soil on one article of evidence compared to another, or compared to a known location, can and has been used as evidence in courts of law (e.g. Kugler 2003; Smith 2003; Pye 2007; Ritz et al. 2008; Ruffell & McKinley 2008; Fitzpatrick et al. 2009; Murray 2011; Fitzpatrick & Raven 2012a, b; Fitzpatrick 2013a, b).

The characteristics of bricks are diverse but their geochemistry and mineralogy can be identified with high precision (e.g. Mishirky & Siniansky 1971; Livingston et al. 1998; Cultrone et al. 2005), and as such can potentially provide important clues to provenance, transport pathways and depositional conditions. Despite the long-standing and widespread use of bricks in construction in urban environments globally (Pye 2007, p. 50), to our knowledge there are only two brief forensic case studies published as abstracts in the grey literature by Smith (2003) and Schneck (2007) where bricks have been used as forensic evidence in a homicide and rape investigation, respectively.

Michael Sams kidnapped Julie Dart in July 1991 and later murdered her after she tried to escape. Sams subsequently kidnapped Stephanie Slater in January 1992, eventually releasing her after the payment of a ransom. According to Smith (2003), police were sent on a long trail across the north of England to deliver the ransom money and secure the release of Slater. Significantly, Sams had left a series of instructions for the police to follow. These instructions were mostly attached to bricks left on the hard shoulder of motorways from Yorkshire though to Manchester. The bricks were painted silver and had a little battery-powered light on the top. It was the provenance of these bricks that would eventually lead police back to Sams’ workshop in Newark. The geochemistry (X-ray fluorescence (XRF)) and mineralogy (X-ray diffraction (XRD)) of the bricks were checked against a database, which enabled the identification of the raw materials from which the bricks were made. These data were used to trace the quarry, brickworks and manufacturer, and was followed by an investigation of sales ledgers, invoices and delivery notes, intermingled with bankruptcy and sale of the original company. This evidence led to the identification of a consignment of these bricks delivered to a depot in Newark, 200 m from Sams’ rented workshop. The similarities between the bricks retrieved from Sams’ workshop and the bricks used in the extortion of the ransom were, according to Smith (2003), regarded to be ‘perfect’, and were later to be referred to as ‘the brick's fingerprint’. Schneck (2007) briefly described an investigation where a young boy was sodomized adjacent to a baseball park infield in El Monte, California. A suspect was later arrested, and his shoes examined for soil and compared to soil from the baseball park. Traces of red brick were observed in the soil from the suspect's shoes. This particular city used crushed recycled red brick for the baseball infield. Microscopical examination of the brick particles from the shoes and the infield showed sufficient similarities in colour, texture and composition to be used as evidence.

Forensic earth scientists, such as soil scientists and geologists, are now using advanced automated techniques that have the ability to acquire information from very small samples. Consequently, earth forensics is being used increasingly in criminal investigations. This paper expands on a preliminary report by Raven et al. (2016), and provides more details on the advantages and successful use of advanced synchrotron X-ray diffraction (XRD) methods. This paper demonstrates how pedological and laboratory approaches, especially those involving traditional and advanced synchrotron XRD methods, have been critical in developing predictive soil–regolith models, from microscopic to landscape scales, to help determine the possible source of the very small red brick fragments associated with the murder of Corryn Rayney.

A total of 157 individual pieces of forensic evidence were submitted to the Centre for Australian Forensic Soil Science (CAFSS) by the Western Australian Police during the period of the investigation between December 2009 and April 2011 (Fitzpatrick & Raven 2010, 2012a; Fitzpatrick et al. 2011). Evidence materials included: minute quantities of soil, brick and paint fragments extracted from bulk material by previous investigators; bulk soil samples collected from the victim's home and surrounding garden beds (Figs 2, 3 & 4), the Bentley Community Centre, and the Kings Park grave site; particulate material from the victim's clothing, especially from the victim's bra (Fig. 5), hair combings and loose material (i.e. soil and Liquidamber styraciflua seed pods) from the victim's body; brick paving chips from the victim's home (40 samples) and Bentley Community Centre (eight samples); and vacuumings from the victim's vehicle.

Initial visual examination and subsequent microscopical examination of the materials were performed with a Wild Leitz M420 stereo microscope illuminated with Schott LED light source (Fitzpatrick & Raven 2016). A 10 megapixel Lumenera Infinity 4 digital camera was used to photograph the smaller samples under the stereo microscope. Larger items were photographed with a 10 megapixel Canon 40D digital camera.

Morphological descriptions of all soil materials were conducted according to the USDA Field Book for Describing and Sampling Soils, Version 2.0 (Schoeneberger et al. 2012) and the Australian Soil and Land Survey Field Handbook (McDonald & Isbell 2009). Soil morphological descriptors such as colour, texture, quartz grain shape, ped structure, segregations (organic or brick fragments), effervescence class and water repellence class are some of the most useful properties for visual soil characterization (e.g. Fitzpatrick et al. 2003) and assessing soil conditions (e.g. Fitzpatrick et al. 1999).

Soil colour was determined on dry samples using Munsell Soil colour notation (Munsell Soil Color Book 2009). Soil colour is usually the first property recorded in a morphological description of soils (and may be the only feature of significance to a layperson). Soil colour provides an indicator of drainage or redox status because soil colour relates to soil aeration and organic matter content (Fitzpatrick et al. 1999).

Bulk soil samples were prepared by grinding small representative subsamples in an agate mortar and pestle. The resulting fine powders were lightly front-pressed onto silicon low background holders for XRD analysis. The 48 brick paver subsamples from the victim's home and the Bentley Community Centre were prepared for XRD analysis by grinding subsamples of the materials for 20 s in a mechanical tungsten carbide ring and puck mill. The resulting fine powders were lightly back-pressed into stainless steel sample holders for XRD analysis.

XRD patterns of the bulk soil and brick paver samples were collected with a PANalytical X'Pert Pro Multi-purpose Diffractometer using iron-filtered Co Kα radiation, an automatic divergence slit and a X'Celerator Si strip detector. XRD patterns were recorded from 3° to 80° 2-theta (2θ) in steps of 0.0167° 2θ for a total counting time of approximately 30 min.

Large numbers of small (<3 mm size) red fragments were observed in many of the bulk soil samples from the victim's home (Figs 2, 3 & 4), and various exhibits from the victim's body, hair, clothing, bra (Fig. 5) and vehicle. Brick fragments for XRD examination were extracted from the bulk soils and exhibits using fine-pointed tweezers under a binocular microscope.

The size of the fragments varied considerably from approximately 3 mm down to less than 0.1 mm in dimension (Figs 3, 4 & 5). Fragments between 0.3 and 1 mm were selected for XRD analysis because the XRD beam would not penetrate particles that were larger than 1 mm and particles smaller than 0.3 mm would not produce acceptable XRD patterns on the laboratory-based XRD instrument. Small fragments were prepared by carefully mounting the unground fragments onto the end of 0.5 mm glass capillaries using cyanoacrylate glue (Fig. 6). The capillaries were secured into a brass holder using melted wax with the aid of a soldering iron and positioned on a Huber tri-axial goniometer head, which was mounted onto the sample spinner attachment on the XRD instrument (Fig. 7).

XRD patterns of the small fragments were collected with a PANalytical X'Pert Pro Multi-purpose Diffractometer using a modified instrument configuration. Iron-filtered Co Kα radiation was used with the X-ray tube positioned in point-focus mode and the beam aligned along the axis of the diffractometer. A 0.5 mm monocapillary focusing attachment was used to focus the X-ray beam onto the fragments mounted on a sample spinner stage (Fig. 7), with the diffracted X-rays collected by an X'Celerator Si strip detector. The diffraction patterns were recorded in steps of 0.0167° 2θ with a total counting time of between 1 and 5 h. The longer data collection times were performed only on samples deemed to be crucial to the investigation. This afforded better discrimination of the mineral phases from very small particles and samples where confirmation of the minor mineral components was necessary.

Initially, a proof of concept experiment was conducted at the Australian Synchrotron powder diffraction beamline (AS103/PD/3019) to ascertain if the superior intensity and resolution could be a useful tool to provenance brick fragments extracted from bulk bricks. The purpose of the proof of concept experiment was to determine if very small brick fragments could be sufficiently characterized to be compared to the bricks from which the fragments originated. A total of 40 of the 48 brick pavers from the victim's home and the Bentley Community Centre (Fig. 1) were selected for analysis at the Australian Synchrotron. For XRD analysis of the bulk brick material, ground samples of the bulk bricks were poured into 0.5 mm diameter glass capillaries and gently tapped on the bench to compact the powder into the end of the glass capillaries. The capillaries were secured into a brass holder using melted wax with the aid of a soldering iron. The brass holder was secured onto a Huber goniometer head, placed onto the powder diffraction beamline sample spinner stage and aligned for XRD measurement.

Numerous small brick fragments were also produced from each of the intact brick pavers by mechanically chipping the bulk pavers using a small pestle. Several fragments were generated from each brick and the most appropriate was chosen by selecting only those that appeared to be fine-grained composites, were less than 0.5 mm in diameter and representative of the bulk of the brick sample. Fragments containing dominantly coarse, opaque or semi-translucent grains (i.e. quartz or feldspar) were avoided where possible.

The fragments were prepared by carefully mounting them onto the ends of 0.5 mm glass capillaries using cyanoacrylate glue (Fig. 6). The capillaries were mounted into brass holders and positioned on a Huber goniometer head, which was mounted onto the sample spinner attachment and aligned on the synchrotron powder diffraction beamline.

The proof of concept experiment was followed up with a commercial in confidence experiment (PD/C3560) to determine if any of the small red fragments collected during the police investigation could be successfully compared with the bulk bricks collected from the victim's home or the Bentley Community Centre (Fig. 1). A total of 25 red fragments, collected as evidence from the victim's body, clothing, body bag, vehicle and home were prepared for synchrotron XRD analysis using the method detailed above.

XRD data were collected on the powder diffraction bending magnet beamline (10BM-1) at the Australian Synchrotron using an energy of 13.005 keV (X-ray wavelength 0.9534 Å) with a MYTHEN strip detector. The detector consists of 16 modules each covering approximately 4.8° 2θ with approximately a 0.2° 2θ gap between modules. This set-up on the beamline results in each pixel covering approximately 0.00375° 2θ with a total collection angle of approximately 80° 2θ. To overcome the gap between modules, the diffraction pattern is collected at two detector positions offset by approximately 0.5° 2θ. The two diffraction patterns are then merged into one using the JAVA program CONVAS2 (Rowles 2010). The high-intensity X-ray beam was collimated in the vertical direction to approximately 0.5 mm and in the horizontal direction to approximately 20 mm for the powdered brick samples and 0.5 mm for the brick fragments. XRD data from the bulk powdered brick samples were collected for 3 min at each detector position and for approximately 27 min each for the fragments. This resulted in total data collection times of 6 and 54 min for the bulk samples and fragments, respectively. The mounted brick fragments were recovered from the ends of the glass capillaries by dissolving the cyanoacrylate glue with acetone.

All the control soils sampled at the victim's home had sandy textures with quartz being the dominant mineral (Fitzpatrick & Raven 2010). The shapes (roundness and sphericity) of the quartz particles ranged from rounded tabular (dominant) to subrounded (subdominant). Of significance, is the obvious presence of small red brick fragments (usually few to very few and fine). The Munsell soil colours of all the small red brick fragments in these samples are mostly Red (10 R Hue) (Fitzpatrick & Raven 2010). Most of the control samples overlie brick paving (i.e. ‘technic hard rock’ according to the IUSS Working Group WRB (2014)), which starts within 5 cm of the soil surface and covers about 95% or more of the horizontal extent of these control soils (i.e. Ekranic). As a consequence, the majority of these soils classify as Ekranic Technosols (Arenic) in accordance to the IUSS Working Group WRB (2014) and Urbic Anthroposols using the Australian Soil Classification (Isbell & The National Committee on Soils and Terrain 2016).

All control soils from Kings Park (grave site) have sandy textures with quartz being the dominant mineral. The shapes (roundness and sphericity) of the quartz particles are mostly subrounded. All samples were non-effervescent in 6 N HCl, indicating they do not contain significant amounts of carbonate minerals. No red brick fragments or other anthropogenic features were observed in any of the Kings Park samples. Consequently, the majority of these ‘natural soils’ classify as Rubic and Albic Hydrophobic, Arenosols (Dystric) in accordance to the IUSS Working Group WRB (2014) and Arenic Orthic Tenosols using the Australian Soil Classification (Isbell & The National Committee on Soils and Terrain 2016).

XRD analysis of the bulk soil samples from all of the sites of interest showed the mineralogy as typical of the vast coastal plain of the Perth metropolitan area, which is situated on extensive sand dune deposits. Quartz was by far the most dominant mineral, comprising an estimated 90% or more of the bulk soils. Other common components included traces of feldspar (albite and/or anorthite and microcline and/or orthoclase). Several samples also contained traces of calcite. Organic matter was evident in most samples by optical microscopy showing variations in colour from light brown to dark brown and black; however, this material is not observable with XRD analysis at the levels found in the soil samples. Soil samples vacuumed from the front passenger floor of the vehicle footwell also contained amphibole, chlorite and mica. This has been identified as likely to have originated from blue metal aggregate commonly used in bitumen road construction. Some of these minerals were also identified in a small, blue-grey 2 × 1 mm grain recovered from the victim's body bag.

The mineralogical composition of the paving brick subsamples from the victim's home (Fig. 2) showed similar mineralogical compositions. Common mineral species found were quartz, hematite, mullite, corundum and pseudobrookite. Less common mineral species found were cristobalite, feldspar (albite and/or anorthite and microcline and/or orthoclase), spinel and mica (muscovite). Three of the brick paving samples from the victim's home also contained calcite, which was later identified as a component of the lime mortar used in the original use of the recycled brick pavers. In addition to the crystalline mineral components identified, significant amounts of amorphous or ‘glassy’ material were also present in the brick paver samples from the victim's home. This is material where, because of relatively rapid cooling, the constituent elements have been frozen in non-regular arrangements. Consequently, for this glassy material, the normally sharp, intense XRD peaks characteristic of well-crystalline minerals are replaced with very broad, relatively low-intensity, bands.

The mineralogical composition of seven out of the eight paving brick subsamples from the Bentley Community Centre showed almost identical mineralogical compositions. The common mineral species found were quartz, mullite, feldspar (albite and/or anorthite), spinel, cristobalite, hematite, corundum and pseudobrookite. Gypsum was found as a trace component in only one brick subsample from the carport area near the path to the car park of the Bentley Community Centre. The origin of gypsum is uncertain as gypsum would have transformed to anhydrite during the brick firing process. The other single brick paver subsample taken from the path leading to the building adjacent to the Community Centre and in front of a playground contained quartz, mullite, feldspar (albite and/or anorthite and microcline and/or orthoclase), spinel, cristobalite, hematite and pseudobrookite.

Cluster analysis was performed on the laboratory XRD patterns from the bulk paving subsamples using HighScore Plus (Version 4.5) from PANalytical B.V., Almelo, The Netherlands (Fig. 8). The XRD patterns are clustered into five discrete groupings, which are directly related to the variable mineralogical compositions of the bricks. While most of the bricks contain several common minerals (quartz, hematite, mullite, cristobalite, corundum and pseudobrookite), cluster analysis showed sufficient variability to separate the XRD patterns into the following categories showing: (1) significant spinel, along with potassium feldspar and mica; (2) no spinel; (3) high mullite content; (4) high albite/anorthite but low mullite content; and (5) broad mullite and cristobalite peaks with high spinel content. Three of the XRD patterns were unclustered, indicating substantial differences in the mineralogical composition from the other bricks. Importantly, all the brick pavers from the Bentley Community Centre, with the exception of the brick from near the adjacent playground, were grouped into cluster 5 (Fig. 8). None of the bricks from the victim's home fell into this cluster. This shows that the seven Community Centre bricks were likely to have been manufactured by the same company using raw materials from the same sources. The bricks from the playground nearby the Community Centre were likely to have been sourced from a different manufacturer or a different blend of raw materials used in their manufacturing process.

The higher resolution and intensity of the synchrotron XRD data from the bulk brick samples from the victim's home and the Bentley Community Centre provided additional information such as the identification of several trace phases (e.g. zircon, rutile and two mica phases) that were not observed in the laboratory XRD data. One disadvantage of the synchrotron data collection was that the use of glass capillaries limits the detection of ‘glassy’ materials in the sample because the X-ray scattering from the capillary can swamp the scattering from the ‘glassy’ material in the sample. This additional scattering from the glass capillary meant that when cluster analysis was performed on the synchrotron XRD data from the bulk bricks, the clusters obtained were not as well defined as the clusters obtained from the laboratory XRD data. In broad terms, the synchrotron XRD data showed that the mineralogy of the bulk paving bricks from the victim's home and the Bentley Community Centre could be grouped into six categories with five individual bricks that could not be placed in any of the five groups. These five bricks had one or more significant mineralogical differences that excluded them from being placed in any one of the six categories. The additional category in the synchrotron XRD data (cf. five categories from the laboratory XRD data) is because of a combination of the additional information obtained using synchrotron radiation and, to some extent, the information added by X-ray scattering from the glass capillaries. Most importantly for the forensic investigation, just as for the laboratory XRD data, seven of the eight bricks from the Bentley Community Centre fell into a single category with no bricks from the victim's home in this category. The synchrotron XRD patterns of the bricks from the Bentley Community Centre show the close similarities of the XRD patterns from these bricks (Fig. 9). More generally, comparison of the synchrotron XRD patterns of the diverse range of bricks from the victim's home demonstrates the capability of synchrotron XRD data to match bricks of similar compositions. For example, Figure 10 shows the synchrotron XRD patterns of five bricks chosen from the same cluster (corresponding to cluster 1 of the laboratory XRD data). The close relationhip of the XRD patterns from these five bricks is clearly evident. Comparison of Figure 9 and Figure 10 clearly shows the differences in the XRD patterns and, therefore, the composition of bricks from the victim's home and from the Bentley Community Centre.

The mineralogy of the red coloured fragments showed fewer mineral components than the bulk brick pavers. This is mainly because the resolution and detection limits when using the laboratory XRD configuration are substantially below that achievable using the ‘standard’ Bragg–Brentano configuration. Quartz and hematite were found in the vast majority of samples, several samples contained mullite, albite and/or anorthite and cristobalite, and a few samples contained goethite and kaolin. The identification of mullite indicated that the red fragments had been heated to a high temperature, and supported the supposition that the red particles were brick fragments.

In addition to the major phases present, some samples also contained one of the following minerals: corundum, pseudobrookite, maghemite, calcite and amphibole. One sample had a single peak that was unidentified. The identification of kaolin, goethite and maghemite in several red fragments was likely to be due to the presence of pisoliths (small rounded red accretions) that probably originated from the nearby Darling Ranges to the east of Perth. Pisoliths are quite common among the lateritic bauxites deposits of the Darling Ranges (e.g. Sadleir & Gilkes 1974; Anand & Paine 2002). XRD analysis of a small white fragment showed quartz and calcite as the only two minerals identified. Calcite is likely to be remnant lime mortar that was used when the bricks were originally used in housing construction.

While the laboratory XRD data from the small fragments were not able to show conclusively that a particular fragment originated from a single brick paver or a group of similar bricks, the laboratory XRD was sufficient to show that mineral phases in the small fragments could be identified and differences in the composition of the fragments could be detected. This outcome led to an application for synchrotron access and subsequent beamtime on the high-resolution high-intensity powder diffraction beamline at the Australian Synchrotron.

For the initial proof of concept experiment, synchrotron XRD patterns were collected from 27 brick fragments that were mechanically produced from the bulk bricks from the victim's home and the Bentley Community Centre. The synchrotron XRD patterns from these fragments were compared with the synchrotron XRD data from the powdered bulk brick samples. The quality of the XRD data produced from the synchrotron source was far superior to the laboratory source, both in terms of resolution and intensity (Fig. 11). The mineralogy of 26 of the 27 brick fragments were successfully matched to one of the six groups of bulk brick from which they were mechanically produced. One of the fragments chosen for the initial proof of concept experiment was from a brick with XRD patterns that fell outside the five groups. The XRD pattern of the fragment from this brick compared only with the XRD pattern of the single bulk brick from which the fragment originated (Fig. 12). The most likely reason that the XRD pattern of the one fragment that compared poorly with the XRD pattern of the brick from which the fragment originated (and the XRD patterns of all bulk bricks) is that the fragment selected from the numerous fragments mechanically generated was not representative of the bulk brick.

When comparing the XRD data for small fragments with the bulk brick XRD data, several assumptions and considerations were made. Bricks are composed of coarse-grained non-clay minerals such as quartz, feldspar (albite, anorthite and microcline) and mica (muscovite, phlogopite) and very fine clay minerals. The coarse minerals are likely to be present in much lower amounts and orientate strongly in the small (<0.5 mm) fragments. Therefore, when comparing the XRD patterns, these components are generally of secondary importance. Furthermore, many of the brick components undergo significant elemental substitutions during manufacture, which results in shifts of peak positions and peak widths. Allowance was also made for the X-ray scattering from the glass capillary used to mount the bulk brick powder. The X-ray scattering from the glass capillary gives rise to an increase in the background X-ray intensity in the angle range from 10° to 20° 2θ (Fig. 12).

Having established that, by using XRD patterns collected on a synchrotron, very small brick fragments could be matched to different brick types, XRD patterns were collected from 20 red fragments recovered from the victim's body, clothing and vehicle, and from the body bag used by the Coroner to transport the body. Based on the XRD data, each of these 20 red fragments could be compared to the paving bricks located at the victim's home (e.g. Fig. 13). Importantly, the Bentley Community Centre could be excluded as a possible source of these fragments. The XRD evidence in conjunction with other forensic evidence was accepted by the trial judge as proof that the victim had returned home from the Bentley Community Centre and that the victim was likely to have been dragged across the red brick pavers located at her home (Martin 2012a, b). However, as it was the victim's place of residence, the judge would not exclude involuntary transfer not connected with the crime. Therefore, while the origin of the brick particles was accepted, the mode of transfer remained an issue and contributed to a not guilty verdict.

In August 2007 the body of Corryn Rayney was found buried in Kings Park, Perth, Australia. Tracing the victim's movements from the Bentley Community Centre where she was last seen to the discovery of her body was vital to the police investigation. Consequently, police called on advice from specialists from a number of forensic science disciplines. Soil evidence from the victim's body, clothing, home, vehicle, the Bentley Community Centre and the grave site were analysed extensively in order to identify the likely movement of the victim after she left the Bentley Community Centre. All the control soils collected at the victim's home classify as Anthroposols or ‘man-made soils’ (Ekranic Technosols (Arenic)). In contrast, the control soils at Kings Park (grave site) and the alibi soils at the Bentley Community Centre classify as ‘natural soils’ (i.e. Hydrophobic, Arenosols).

The Perth region is a coastal plain where the soil consists predominantly of coarse, sandy soils (Northcote et al. 1960–68; Johnston et al. 2003) that show only minor mineralogical variation throughout the region. Consequently, pedology and bulk soil mineralogy was of little use in identifying the origin of crucial soil evidence. However, very small (often <0.5 mm) red particles were observed on the victim's body, hair, clothing (mainly her bra), in her vehicle and in the body bag used by the Coroner to transport the body. These red particles had the appearance of brick fragments and were not found in the soils that surrounded the grave site at Kings Park.

Recycled red house bricks were extensively used at the victim's home for paving of the driveway, footpaths and pergola area. The paving bricks at the victim's home were recycled from houses constructed prior to 1940. Bricks manufactured prior to 1940 were sourced from several localities and were often subjected to less than ideal firing conditions. The red bricks used at the victim's home therefore ranged in hardness and durability from soft to very hard. In contrast, the paving bricks at the Bentley Community Centre were fired relatively recently, and showed little variation in colour and texture from brick to brick and were very hard.

Laboratory X-ray diffraction (XRD) analysis of 42 bricks from the victim's home and the Bentley Community Centre showed that, based on the mineralogy of the bricks, the bricks could be classified into five or six distinct groups (Fitzpatrick et al. 2011; Raven et al. 2016). Bricks from the Bentley Community Centre all classified into a single group and this group did not contain any bricks from the victim's home.

Laboratory-based XRD techniques using low-background Si wafer holders are useful for measuring XRD patterns from samples with weights as low as several milligrams. However, these techniques are generally not sensitive enough to measure statistically meaningful XRD patterns from samples weighing less than 1 mg or comprising a few submillimetre particles. Synchrotron XRD with high X-ray intensity provides far greater sensitivity and resolution than laboratory-source XRD systems. This enables identification of minute amounts of mineral components.

Powdered bulk and small fragments extracted from the 42 bricks were measured on the powder diffraction beamline at the Australian Synchrotron. The synchrotron results showed that each of the small brick fragments examined could be effectively compared to the whole brick group from which the fragment originated. Fragments of red particles from the forensic evidence were subsequently analysed at the Australian Synchrotron and were shown to be consistent with having originated from the victim's home.

The soil forensic work on this investigation was often complex and painstaking, with the CAFSS team spending nearly 2 years successfully linking minute brick particles in the murder victim's bra and hair to a collection of old brick pavers in the front yard of the victim's home. The CAFSS report and presentations during cross-examination in the Perth Supreme Court provided a ‘predictive, soil-regolith model, from microscopic to landscape scale’, which established that soil and brick particles/fragments found on the victim's bra and hair originated from the front yard of the victim's home. The trial was before a judge only and he (Judge Martin) concluded that the mineralogy data from the brick particles on the victim's clothing and the bricks from her front driveway suggested she could have been attacked in her front yard and then transported to Kings Park where her body was buried (Martin 2012a, b).

Our analysis of bricks and brick fragments demonstrates that XRD is a highly effective technique for determining the source of trace materials for forensic investigations. XRD data (particularly data collected at the Australian Synchrotron) show that there is a strong correlation between the XRD patterns from small, brick fragments and the XRD pattern from a bulk sample of the brick (or batch of bricks) from which the fragments are known to have originated. More importantly, because of the diverse mineral composition of clay bricks, XRD analysis enables brick fragments to be exclusively classified and categorized within specific groups of clay bricks.

The authors would like to thank Greg Rinder for the art and graphics of Figure 1. We thank Professor Ken Pye, Director Kenneth Pye Associates Ltd, for kindly providing the authors with a copy of the abstract by Smith A.S. (2003) entitled ‘A Brick's a brick, isn't it? Michael Sams – Case Study’. The authors acknowledge the assistance of the beamline scientist Kia Wallwork. We thank Professor Hilton Kobus from Flinders University for constructive comments on the draft manuscript.

The authors wish to acknowledge the Australian Synchrotron funding (grant No. AS103/PD/3019) for beamtime.

MDR: Data Curation (Equal), Formal Analysis (Lead), Investigation (Equal), Methodology (Lead), Validation (Equal), Writing – Original Draft (Lead); RWF: Investigation (Supporting), Project Administration (Equal), Writing – Review & Editing (Equal); PGS: Formal Analysis (Equal), Investigation (Equal), Methodology (Supporting), Validation (Equal), Writing – Review & Editing (Equal).