Polymers on the crime scene: How can analytical chemistry help to exploit the information from these mute witnesses?

Abstract

Polymers are ubiquitous in everyday life, so it is very likely that they may be encountered on a crime scene as well. In order to exploit to the fullest extent the amount of information that these items contain, it is necessary to properly characterise them. The state of the art and the most recent advances in the forensic characterisation of polymeric items are presented. The qualitative and quantitative determination of the formulation is discussed, along with more innovative approaches, that focus on the features directly related to the macromolecular nature of such traces (molecular weight, degree of crystallinity, presence of comonomers, etc.).

---

Valerio Causin

Valerio Causin is a researcher of Industrial Chemistry at the University of Padova. His research interests are focused on the application in a forensic science context of polymer characterization techniques, especially those focused on the material's structure and morphology.

---

1. Introduction

Due to its fascinating nature of interface between the intricacy of law and the rigour of science, forensic science has attracted a wide interest among the academic community, as well as within the general public. Forensic science can be generally defined as science applied to the administration of justice. This definition does not fully reflect the peculiarity of this discipline. An emerging view within the scientific community focuses on traces as the characterizing elements on which forensic science is based, i.e. forensic science is the study of traces.¹ Traces are the remnants of an activity,² and they can be considered as the most basic “material or physical” information on crime. This “trace-centered” definition more clearly portrays the distinguishing features and implications of this discipline.

First of all, forensic science rarely deals with samples, but with specimens. The specimens have the particular properties of being traces whose representativity of the whole source is unknown and can only be supposed, just as the homogeneity of the source has to be supposed.

Analyses in the laboratory are just a part of a much wider process that relies on proper detection, selection, collection and observation of specimens at the crime scene. Once detected and collected, such traces follow a complex set of interrelated processes that ultimately compose the whole information system.¹ They are collated and interpreted in order to provide knowledge that is used to make decisions at various levels of security systems and criminal justice.¹ This relates to some key issues in the process of interpretation of the information. It is quite widely accepted that the forensic scientist should consider two different propositions that, commonly, represent the positions of the offence and of the defence. The possible propositions that may be employed to guide the interpretation of evidence can be framed into the following hierarchy: the source level (e.g. the glass fragment came from window X/The glass fragment came from some other broken glass object), the activity level (e.g. Mr A is the man who smashed window X/Mr A was not present when window X was smashed) and the offence level (e.g. Mr A committed the burglary/Another person committed the burglary).³ It is not the purpose of this paper to cover the complex inference processes involved in the interpretation of scientific evidence. Suffice here to say that the propositions that may be formulated, and therefore which level within the hierarchy delineated before is tackled, depend on the circumstances of the case, the observations that have been made and the available background data.^3,4 In other words, the possibility of identifying the source of a trace or to reconstruct the dynamics of a crime are not only dependent on how the analytical process was carried out, but on a variety of information and circumstances that need to be assessed on a case-by-case basis. It is nevertheless beyond debate that the analytical data obtained in the laboratory play a crucial part in contributing to the whole process of attribution of significance to forensic evidence. Despite the significant contribution of criminalistics to critical legal matters and the perceptions of popular culture, there are branches of forensic science that are still in need of a considerable research effort, in order to allow a full exploitation of their potential. The characterisation of polymeric traces is among these latter areas of knowledge. Academic research and published protocols are rather scarce, primarily due to the vastness of possible samples and thus to the difficulty of obtaining general methods useful for all the items possibly encountered in actual investigations.

A very large number of the objects that surround us in our daily lives are polymeric in nature, so it is highly likely that polymers may be connected to the commission of crimes also. Among the many possible examples, we can cite textile fibres that are shed during a struggle, paint chips transferred or reflectors broken in a car accident, or adhesive tape that is often used in kidnappings. These can be considered the most common items, but also latex gloves, bits of polyurethane foam, resins contained in inks, and plastic bags are more unusual materials that a forensic scientist could be called to deal with.

One of the major reasons that delayed the exploitation of polymeric items in forensic science is their alleged “commonness”. Even though most plastic objects are industrial products very similar for structure and properties, a certain degree of differentiation between them occurs. This possibility of discrimination lies in the influence that slightly different production processes have on the structure of the polymer obtained. Most of the plastic bags on the market, for instance, are made of polyethylene. A large number of manufacturers exist, though, each one of them using different raw materials, additives and processing procedures. A scrupulous characterisation approach, targeted to the individuation of the manufacturer's profile, is fundamental for a proper use of polymeric items in investigations. Polymers, as other traces, are class evidence, and as such they acquire evidential value only if they can be placed in as small a class as possible. Sound analytical procedures capable of yielding the largest amount of information are therefore necessary to achieve this aim.

Practically no plastic object is made of 100% polymer. Most polymers in fact do not possess adequate aesthetical and/or functional properties to fit common end uses, so additives are extensively used in industrial practice. Plastic objects are therefore usually more aptly described as composites. As such, their characterisation can be made, on a first level, individuating the formulation of the matrix polymer and of the inorganic and organic additives (Fig. 1). A second level, though, is much more underrepresented in the forensic science literature, and makes use of methods and characterisation techniques which are typically applied by polymer scientists. In this approach, focus is posed on the polymer matrix. Features such as the average molecular weight, molecular weight distribution, isotacticity, or presence of comonomers, differentiate macromolecules from ordinary small molecules, and are strictly dependent on the synthetic pathway chosen to produce the polymer (Fig. 1). Investigating the structure and morphology of the semicrystalline framework can shed light on the processing imposed during manufacturing (Fig. 1). This information dependent on the synthesis and manufacturing process, may allow to narrow significantly the class in which the item can be placed, dramatically increasing the significance of the evidence.

Fig. 1 Characterisation scheme for a polymeric item of forensic interest.

The aim of this critical review is that of surveying the state of the art of the analytical procedures proposed to characterise polymeric traces in a forensic science context, offering suggestions for further developments and research. An effort will be made to merge polymer science with forensic science, in order to promote a point of view, in the design of a characterisation procedure for polymeric items of forensic interest, which may contribute to maximise the significance of this kind of trace evidence.

References were chosen as much as possible within analytical journals, although in some cases papers from materials or polymer science were cited to give examples of the described procedures. Although DNA is by any means a polymer, it will not be taken into account in this review, as the analysis of this molecule, fundamental in biology, life, and forensic science, is very well covered by the literature. Focus will be posed instead on polymers of interest as non-biological contact traces.

The organisation of the paper will be according to the type of information that can be gathered on the item, i.e. formulation, synthesis-dependent features, structure and morphology.

2. Formulation

As said in the introduction, objects based on polymeric materials usually contain additives which can be organic or inorganic. Being able to determine the nature and quantity of these additives in the formulation would enable the forensic scientist to make very significant comparisons.

An established example of such a strategy is the analysis of dyes and pigments in textile fibres, which is very well developed and has been well covered by a number of books and recent papers,^5–7 where it is shown that many techniques ranging from chromatography, to spectroscopy to mass spectrometry have been explored and proposed.

A more original approach to the determination of the formulation of textile fibres was proposed by Vann et al.⁸ who applied Raman microspectroscopy to distinguish between rutile and anatase forms of the inorganic pigment titanium dioxide (TiO₂) and to make quantitative measurements of titania loading. TiO₂ is a very common additive used in the fibre industry as a delustering agent, to confer an optimal aesthetical appearance to the final product. Classical reference textbooks⁶ recommend an examination of the fibre under the microscope, noting the presence of delusterant and roughly quantifying it, in comparison to the reference sample. This work on Raman spectroscopy is especially interesting because it introduces an objective identification and quantification procedure. Fibres are iconic examples that allow to understand the difficulties involved in the development of a forensic characterisation procedure. They are extremely small (the average item can be 3 mm long and about 20 μm in diameter) and very anisotropic in shape. Microspectroscopy, either in the UV/visible or IR range, or Raman, is then a requested choice.⁶ Polarization properties of the fibre on the incoming and transmitted light must be taken into account.⁸ The effect of the cross section shape, which influences the transmission with edge or interference effects^8,9 should also be considered. Not always the filler distribution is homogeneous, so a suitable number of replicates must be performed in the measurements. The variability within the fibre and between fibres should be assessed with suitable experiments made on a collection of samples.

The presence of catalyst residues or of inorganic additives can be detected from the inorganic analysis of the sample. Atomic spectroscopic methods and X-ray fluorescence are covered by excellent periodical reviews, appearing annually in the Journal of Applied Atomic Spectroscopy.^10,11 From these surveys it emerges that applications in forensics almost invariably are focused on glass and paints, whereas the determination of trace elements in plastics is done as part of industrial or basic polymer science research. Cross fertilisation between these fields is desirable, because usage of these techniques for forensic items such as plastic automotive parts, plastic bags, or latex gloves, would be very advantageous in increasing their evidential value. A similar example was proposed by Aziz et al., who applied established petroleum characterization methods for analyzing adhesive tapes.¹²

From the point of view of the forensic scientist interested in polymeric items, the most desirable advances in the field of elemental analysis regard the techniques that allow inflicting the minimum damage to the sample, such as laser ablation-induced coupled plasma-mass spectrometry (LA-ICP-MS) and laser induced breakdown spectroscopy (LIBS). LIBS is rapidly gathering pace, especially in soil and glass analysis, but it is not yet exploited to its fullest extent for polymer items. Among the very few applications of this technique to polymeric materials there are works by Wise and Almirall on the detection of rare earth chemical taggants placed on objects¹³ and by Sarkar et al. on the forensic differentiation of paper.¹⁴ A very attractive feature of LIBS is the availability of portable instruments, that allows to “move the lab to the crime scene”. This emerging trend in the management of casework is a natural consequence of the importance, also noted in the Introduction, of a proper detection, selection, collection and observation of specimens at the crime scene. A preassesment of the specimens directly on the crime scene allows to focus on the more interesting and information-rich traces, thus avoiding overloading the laboratory with useless specimens.

LA-ICP-MS is surely the technique that was most applied for the forensic characterisation of the inorganic content of polymeric items, especially paint chips^15,16 or microdebris.^17,18 By LA-ICP-MS, trace elements in the sub-ppb range can be detected, allowing an excellent discriminative power. Since no digestion of the sample is necessary, two advantages are obtained. The first one is that of avoiding manipulation of the sample. Sample preparation is often exploited by lawyers to induce doubts about the validity of the analytical results. The second advantage achieved by LA-ICP-MS is that the absence of a chemical treatment for digesting the sample decreases the risk of introducing agents that interfere with the analysis. For example, in their work on the ICP-MS analysis of the inorganic profile of adhesive tapes, Dobney and coworkers noted that the use of hydrochloric acid in the digestion step interfered with an accurate determination of Cr.¹⁹ When performing comparisons based on the inorganic profile, caution must be exercised, because all sources of contamination must be taken into account. Many crime scenes are exposed to the elements, and weathering may severely alter the elemental profile of the items therein. LA-ICP-MS has a very good spatial discrimination, and it allows to exclude macroscopically altered regions of the sample. More importantly, lengthening the time of exposure to the laser increases the depth of the crater and thus the sampling depth. This allows on one hand to exclude from the analysis the outer layer of the sample, which is more prone to contamination and weathering, and on the other hand to obtain the inorganic profile through all the thickness of the sample.^15,16 This is extremely significant when dealing with specimens with a layered structure, such as paints or multilayered plastic films.

Matrix effects must always be considered, especially when high concentrations of a prevalent inorganic additive are present in the organic matrix, as noted by Deconkinck et al. in their work on car paints.¹⁵

Although the performance of LA-ICP-MS is probably unrivalled for forensic purposes, other methods of elemental analysis have been and are routinely applied.

Traditional ICP-MS is a viable and economic alternative when large sample sizes are available and when a partial destruction of the item is acceptable. An example may be in the analysis of paper in questioned document examination. Spence et al. confirmed the efficacy of such an approach, being able to differentiate seventeen paper samples on the basis of just two elements (Mn and Sr). Intralot and intrasample experiments were carried out to check the robustness of the analysis, taking into account nine of the most abundant and homogeneously distributed elements.²⁰

Scanning electron microscopy with energy dispersive X-ray analysis (SEM-EDX), has a good spatial resolution, but lacks enough sensitivity to enable trace analysis. Among the most recent reports that made use of this technique, we can cite the work of Higashikawa et al., in which SEM-EDX was used to discriminate single fibres on the basis of the inorganic antimicrobial agents that they contained.²¹ X-ray fluorescence is an established technique for the analysis of paints or for some kinds of fibres,^22,23 especially because of its increased sensitivity with respect to SEM-EDX. Micro-beam X-ray fluorescence,^24,25 although the sampled depth depends on the energy of the X-ray photons and, hence, on the analyte element, offers probably the best currently available compromise for the analysis of the inorganic profile of small specimens, due to its very good spatial resolution and to the portability of devices which allows for direct intervention of analytical techniques on the crime scene. Very promising work has been done in the last few years on the hyphenation of elemental analysis, such as micro-beam X-ray fluorescence or SEM-EDX, and Raman microspectroscopy.^25,26 Cross reference of the elemental profile with the molecular information from Raman allows for unambiguous and complete identification of the formulation of the item.

A fascinating approach for the investigation of the inorganic additives in analysed items is the use of cathodoluminescence (CL). CL consists in the emission of visible radiation from a sample that has been impinged by an electron beam. This technique is not new, its discovery dates back to 1879, and is routinely applied in geology for provenance studies based on the content of trace elements or compounds, and in physics and material science to detect the presence of inorganic contaminants. Palenik and Buscaglia offered an excellent review of the potential of such a technique in which every instrumental and technical detail is well described.²⁷ The electron beam that induces luminescence can be produced by a cold cathode, which avoids the necessity of coating the sample with a conductive layer, and allows operation in low vacuum with a beam of large dimension, up to 2.5 cm. The drawback is that the obtained cathodoluminescence is low in intensity and quantitative measurements are not very reliable. Using a hot cathode source, such limitations are overcome and higher resolution is possible by a finer beam, but coating of the sample and high vacuum are necessary. A limit linked to sample preparation is that CL is a surface technique, therefore, in order to obtain complete data also on the bulk and to rule out the effect of contamination, a cross section should be prepared. Moreover, standardisation problems affect image collection and the acquisition of spectra. Despite these drawbacks, that can be minimised operating with due diligence, CL has been shown to be extremely effective in the comparison of paint and duct tape (Fig. 2).²⁷ The advantage of CL is that it complements elemental analysis with information on the phase of the minerals present in the sample. For example, the two phases of CaCO₃, calcite and aragonite, luminesce orange-red and yellow-green, respectively. Trace elements such as Mn or Fe, which can be related to the geologic origin of the mineral, or can derive from its anthropogenic synthesis, further modify the CL spectrum.²⁷ In case of complex samples it can be very hard to identify the various phases of the mineral additives by traditional methods such as X-ray diffraction. Associating CL with elemental analysis can be a much more immediate approach, also improved by possibility to visualise the dispersion and location of the fillers. A further detail that can enrich the comparison is the observation of the evolution of luminescence with time during exposure to the beam. Phenomena like defect migration, annealing or heating cause in some instances a fading or a modification of luminescence (Fig. 2d).

Fig. 2 Luminescence of paint and duct tape. (a) By reflected light (left) a paint chip appears homogeneous, whereas CL shows the presence of four layers differing by inorganic content. (b) Schematic of duct tape construction and (c) a CL image of duct tape, which shows that each layer contains a luminescent component. (d) Six duct tapes can be differentiated by CL, the top row shows CL immediately after exposure to the electron beam, the lower row shows the same samples after 10 min of electron beam exposure. Reprinted from ref. 27 with permission from Wiley.

CL can be considered one example of an emerging class of methods in forensic science, i.e. chemical imaging. Methods exploiting visible, near infrared and mid-infrared absorption and photoluminescence have been proposed for chemical imaging, showing the potential of visually conveying spectral and chemical information.^28–32 This aspect is very important in forensic science, because the audience of the analytical results are judges, lawyers, and jurors who do not have specific scientific training and often have difficulties in completely understanding the significance of the work of the scientist. It is obvious that critical decisions like those taken in a courtroom should not be based on poorly understood analyses results.

The advantages of synchrotron radiation in X-ray absorption, fluorescence, diffraction and scattering, and in IR spectroscopy as well, were presented by Kempson et al.³³ Synchrotron radiation allows for greater spatial resolution, more rapid analysis and higher sensitivity. The high brilliance of the incoming light beam dramatically decreases the minimum sample size necessary for the cited analyses. However, access to synchrotron sources is limited, and usually beamtime is allotted on the basis of research projects presented many months in advance. Use of such techniques, albeit highly effective, can be considered just for research purposes or for very high profile cases, and can be rarely if ever applied in normal casework.

Stable isotope ratio analyses are now routinely applied in provenance studies in the environmental, biological and geological fields. Their application on trace evidence items is quite underrepresented:³⁴ only a few reports exist in the literature on post-blast debris,^35–39 architectural paints,⁴⁰ plastic bags,⁴¹ and adhesive tapes.^12,42 Isotope ratio mass spectrometry (IRMS) can detect minute variations in stable isotope composition, therefore yielding information on the source and origin of a given material, which no other analytical method can provide. The principle exploited by IRMS is that, although the isotope ratios of the whole Earth are fixed because they were determined at the time of the planet's formation, a number of natural and artificial fractionation processes occur, that locally modify the relative abundances of stable isotopes of an element.³⁴ Use of raw materials of different origin, processed in different ways by different manufacturers will thus result in different stable isotope ratios, and therefore in an unparalleled opportunity for forensic discrimination of otherwise similar items. Since all polymers contain carbon and hydrogen, and most of them also oxygen atoms, these elements are those that deserve more focus, although in particular cases of highly filled samples, it can be worth extending IRMS to other elements too. Care must be taken, while setting the experimental parameters, that the analysis itself does not produce fractionation.

The sensitivity of IRMS is at the same time a strong and weak point of the technique for forensic applications. On one hand it allows to enrich the data by sampling different zones of the items that may be isotopically distinct from each other. For example grip seal plastic bags are manufactured assembling the seal to the plastic bags, each component having a different origin,⁴¹ or adhesive tapes are composite materials that can be schematically described as glue on a plastic backing, each coming from a different source as well.⁴² Analysing each component of the object greatly increases the significance of the comparison. On the other hand, manufacturing variation requires to sample several locations of the item, before concluding an association. As an example, Quirk et al. published a work in which they used IRMS to associate post-blast residues of two-way radios with the undamaged radio used to initiate explosive devices. They found that when less than three fragments were recovered, positive association was not possible because of the variations in the manufacturing of the items: apparati with sequential serial numbers were not necessarily made from all of the same components.³⁵

Not all the additives used in the polymer industry are inorganic in nature. Especially when organic fillers are to be analysed, mass spectrometry is the technique of choice for their identification, although to date quantification is not possible yet by this approach.^43–47 The most recent advances in mass spectrometry, which are enriching the forensic science toolbox, are ambient ionization techniques. DESI (desorption electrospray ionization) and DART (direct analysis in real time) are the most established, but a number of other ambient ionization techniques have been proposed in the last 5 years.⁴⁸ In DART, the only one for which applications to high molecular weight polymers were reported, the excited-state species present in a plasma are used to desorb the analyte molecules. The basics of the technique are detailed elsewhere,⁴⁹ suffice here to say that it allows a remarkable flexibility, being able to analyse solid, liquid and gaseous samples, under ambient and solventless conditions, and without exposing the operator and the sample to high voltage or ionizing radiations. Laramée et al. presented a number of applications that confirmed the potential of DART for forensic science.⁵⁰ When approaching polymeric items, DART can yield information both on the low molecular mass additives and on the nature of the polymeric matrix. Performing the analysis at low gas temperature, additives, plasticizers and other low molecular mass compounds are ionized and detected by the mass spectrometer. Increasing the temperature, polymer fragments are observed.⁵⁰ In DART, fragmentation starts at much lower temperatures than those necessary for pyrolysis, so DART-specific mechanisms must be invoked. DART spectra are simpler than those obtained by pyrolysis (Fig. 3). The applicability of DART-MS to glues, plastics, fibres and inks was shown.⁵⁰

Fig. 3 (a) DART and (b) Pyr-GC-MS spectrum of nylon 6. Reprinted from (a) ref. 50 with permission from Wiley and (b) from ref. 51 with permission from Elsevier.

To the knowledge of the author, the cited applications of DART are the only works in mass spectrometry dealing with the forensic characterisation of high molecular weight matrices. However, these methods are still of interest for the polymer scientist working in forensic science, because they are very well suited for the analysis of low molecular weight synthetic polymers, e.g. resins in inks⁵² or in pressure sensitive adhesives,⁵³ or lubricants for condoms.⁵⁴

The same is true for capillary electrophoresis (CE), another technique widely used in other branches of forensic science, i.e. DNA profiling, but comparatively less applied in the contact trace field. It is a very suitable technique for the detection of dyes and pigments in a variety of matrices,^7,55–57 but its usefulness for the analysis of polymers is limited to biomacromolecules or oligomers such as lubricants for condoms.⁵⁷ A further drawback of CE is that it requires extraction of the analyte. This sample preparation step alters the nature of the specimen, and therefore it is not desirable in a procedure applied in forensic science.

3. Polymer microstructure

Among the analytical methods useful for the characterisation of the polymeric matrix in a plastic item, there is no doubt that IR and Raman spectroscopy are the most widely used. Hyphenation with optical microscopes and the invention of attenuated total reflection (ATR) IR spectroscopy widened the range of analysable forensic items to very tiny fibres,^6,58,59 paint smears^60,61 or plastic particles transferred in an assault.⁶² Transflection sampling mode is also an option for very small samples. In transflection, the sample is deposited on a reflective surface (called the substrate) and the incident light transmits through the sample onto the surface of the substrate from which it is diffusely reflected back through the sample to the detector.⁶³ It has been successfully applied on a variety of samples in near infrared spectroscopy. Diffuse and specular reflected light occurring during measurements may negatively influence spectral quality, but the potential of this approach, also in the mid-IR region, has been shown by Kocak et al., although only applications in the illegal drug sector were privileged so far.^63,64

The applications of IR and Raman in the forensic analysis of materials are many, among these the identification of contaminants in pharmaceutical products,^65,66 the analysis of lubricants in sexual assault cases,⁶⁷ the analysis of paper in questioned document examination,^68,69 the characterization of adhesive tapes,⁷⁰ the verification of authenticity in works of art,⁷¹ and the discrimination of architectural^72–74 or automotive paints.⁷⁵ A beautiful and very explicative example of how rich is the amount of information that can be contained in a single IR spectrum is shown by Grieve and coworkers in their work on acrylic fibres.^76,77 A single acrylic fibre can yield a spectrum that tells the analyst what comonomer was used in the synthesis of the fibre, what are the dyes used to confer its colour and which solvent and which technologies were used in the spinning process.

Application of IR spectroscopy as a tool for quantitation is quite a controversial matter. Some reports on the analysis of fibres showed, though, that fairly precise measurements can be performed.^9,78 IR peaks relative to the comonomer and signals due to the polyacrylonitrile backbone were integrated, their ratios were calculated and used to assess quantitatively the comonomer content in acrylic fibres, obtaining data with a relative standard deviation below 6%, and a remarkable differentiation of samples which had been deemed indistinguishable by previous microscopic examination.^9,78

Raman selection rules are different from those of IR. Moreover the Raman effect is a scattering process, and unlike IR it does not suffer from problems associated to light transmission. These aspects make Raman and IR complementary techniques, which can yield data on all the components of a complex sample. It is especially worth noting that Raman allows an in situ analysis of dyes, for example in fibers, without pretreatment or other sample preparation, therefore avoiding more labour intensive and partially destructive methods such as chromatographic procedures.^79,80 The complementarity of Raman and IR often allows to obtain separate information on the polymer substrate and on the additives. Jochem and Lehnert,⁵⁹ for example, worked on automotive paints and noted that while Raman proved very useful for the analysis of pigments and did not give information on the polymeric binders, the opposite happened for IR spectroscopy, which was well suited for studying the polymers, but failed to provide useful data on dyes or pigments. Applying both techniques therefore allows for a complete characterisation of the material.

Another widely used technique, useful for the identification of the polymeric matrix composing trace evidence items, is pyrolysis hyphenated with gas chromatography and mass spectrometry (Pyr-GC-MS). Pyrolysis overcomes the difficulties encountered in mass spectrometry for the ionization and detection of large macromolecules, because they are fragmented by application of a heat impulse. Reproducibility has always been an issue in Pyr-GC-MS, however many literature works showed that, by a careful experimental design, very reliable and informative results can be obtained. Examples of items successfully analysed by Pyr-GC-MS include paints,⁸¹ wood fragments and vegetable fibres,⁸² glitter lip gloss,⁸³ acrylic fibres.⁸⁴ Pyr-GC-MS data are not always easily interpretable, since the detected fragments come from the polymer matrix as well as from other substances present in the formulation of the analysed item. When an identification of the nature of the fragments is possible, it allows to gain further insight and to offer a chemical interpretation of the comparison between items. A quite detailed picture of the chemical structure of the polymer can be obtained, sometimes, if necessary, with simple derivatisation techniques, as shown by the pioneering work of Challinor.^85–87 Kristensen et al. were for instance able to distinguish if wood fragments came from angiosperms or gymnosperms on the basis of the ratio of syringyl to guayacyl lignin fragments in thermally assisted hydrolysis and methylation (THM) Pyr-GC-MS.⁸² The same authors tried to extend a similar approach to the discrimination of cellulosic textile fibres without the same degree of success, yet confirming the intrinsic value of Pyr-GC-MS in acquiring useful characterisation data for a thorough and complete description of items of forensic interest. Pyr-GC-MS can be particularly useful when dealing with copolymers. The presence or absence of particular fragments can be indicative of the presence or absence of a particular comonomer, complementing data from IR spectroscopy. For example in pyrograms obtained from acrylic fibres, the presence of acetic acid was noted only when vinyl acetate was used as a comonomer of acrylonitrile, and it was absent in the case of other fibres in which acrylonitrile was copolymerized with methyl methacrylate, styrene sulfonate, methyl acrylate or vinyl pyrrolidone.⁸⁴

A technique which can be considered related to pyrolysis is thermogravimetry (TG).⁸⁸ TG allows to follow changes in sample mass originated by thermal events, especially degradation or combustion, of plastic items during a temperature program. Such a behaviour is strictly dependent on the microstructure of the material and on its formulation.^89,90 Ihms and Brinkman⁹⁰ and unpublished results by the author confirm that TGA shows excellent discriminating results even when samples are contaminated with silicates, organics, moisture, and char.

Trivially, one could assert that polymers differ from small molecules because they have a very large molecular weight. Albeit true, the situation is quite a lot more complicated, because the synthetic pathways and the processing necessary for the industrial production of plastic items introduce a number of features that radically differentiate polymers from traditional small molecules. A very distinctive parameter characterising polymers is their average molecular weight. It is beyond the scope of this review to discuss the details of this fundamental topic of polymer science, it will be sufficient to note that, since it is not possible to perfectly control the reactions involved in the formation of macromolecular chains, the product of a polymerisation will be invariably a mixture of molecules composed of a variable number of repeating units derived from the monomer. If we compare a sample of water and one of polypropylene, we will see that all water molecules are composed by two atoms of hydrogen and one of oxygen, i.e. all water molecules have the same molecular weight. On the other hand, in the polypropylene sample there will be molecules composed of the same type of repeating units, but differing in length. Some chains will be shorter and some will be longer. To describe such a system it is necessary to define the molecular weight of the sample by its mean (the most common definitions of such quantity are the number average molecular weight and the weight average molecular weight) and its distribution. These features are obviously critically dependent on the synthesis procedure adopted to prepare the polymer, and so they may be exceptionally useful for a forensic scientist. Very surprisingly, just one report exists in the literature where this approach was followed. Kumooka successfully applied size exclusion chromatography to differentiate rubber based pressure sensitive adhesive tapes.⁹¹ The necessary sample size is not very small, Kumooka used pieces of adhesive tape 1 cm long and 2 cm wide, but cases involving adhesive tape usually yield quite large quantities of this material. Moreover, this approach is applicable in cases in which relatively large plastic items are present, i.e. car accident scenes, failure analysis or improvised explosive devices.

In addition to size exclusion chromatography, the number average molecular weight of a polymer sample can be investigated by an end group analysis. This technique can only be applied to polymers with suitable end groups, e.g. carboxyls that can be titrated.⁹² IR or NMR spectroscopy can also be employed by ratioing signals due to the end groups with respect to those related to atoms or bonds located in the backbone of the chain.^93–96 The larger the number of end groups per unit sample mass (or per unit backbone segment), the lower the average molecular weight. Sensitivity issues limit the maximum molar mass measurable by end group analysis. Since only one or two end groups exist in each chain, in order to assess molecular weights of 10⁴ g mol⁻¹, one would need an analytical method capable of detecting 10⁻⁵ moles. A practical upper limit for the number average molecular weight that can be measured by end group analysis is 15000 g mol⁻¹, so this method is not useful for commercial structural plastic materials, but could be useful for the analysis of oligomers used in lubricants, additives, paints or adhesive tapes.

Due to the extensive sample mass required, the other methods of determination of the polymer average molecular weight, such as those based on measures of colligative properties, light scattering or viscosity, are not likely to find any application in forensic science, and thus deserve much less attention. When comparing two polymeric samples on the basis of molecular weight, possible degradation pathways likely to decrease the size of polymer chains should be taken into account, especially for items coming from a crime scene which has been exposed for long times to the elements (UV light especially).

NMR is, in principle, a very powerful tool for the forensic characterisation of polymers. It allows to probe synthesis-dependent features such as the degree of isotacticity, i.e. the regularity of the sequence of configurations of the stereogenic centres in polymers derived from monosubstituted vinyl monomers, and the mode of addition of these monomers (regioregularity),⁹⁷i.e. with the substituents on adjacent carbon atoms (head-to-head) or on alternate carbon atoms (head-to-tail).

Modern catalysts allow the synthesis of almost completely isotactic polymers, so differentiation of commercial materials on this basis seems unlikely to succeed. Regioregularity may yield more interesting results under a forensic point of view. Commercial poly(vinylidene fluoride) (PVDF), a polymer used in the electronics industry, and thus possibly found in investigations related to explosive devices, can have significant quantities of head-to-head sequences, which are detectable by NMR or IR spectroscopy.^98–101

NMR is also a very efficient method to investigate and quantify the type and number of ramifications along the main chain. NMR is for example one of the most sensitive and selective methods to differentiate between high density polyethylene (HDPE), low density polyethylene (LDPE) and linear low density polyethylene (LLDPE).⁹⁴ Such polymers differ by the presence of ramifications in the macromolecular chains, which are absent in HDPE, random in number, location and length in LDPE and controlled in length and number in LLDPE (which is actually a copolymer of ethylene and α-olefins). Many polyethylene objects can be made with any of these 3 kinds of materials, so different producers could manufacture the same item with different types of PE, therefore offering to the forensic scientist an opportunity for discrimination.

Quite large sample sizes are needed for NMR analyses, so the method, although attractive, can be applied just to a limited set of forensic items that are more likely to be found in large quantities at a crime scene, e.g. automotive plastic parts at a car crash location. Other problems to be overcome in NMR measurements are the difficulty in dissolving polymers and the high cost of the instrumentation.

4. Polymer structure and morphology

When a low molecular weight substance crystallises, all its atoms or molecules become ordered in a crystalline lattice. On the contrary, polymers which are able to crystallise are called semicrystalline, because, due to entropic constrains, large macromolecular chains are unable to form a crystalline lattice extended to the whole sample. The structure and morphology of polymers is therefore better described as an alternation of crystalline and amorphous domains. The degree of crystallinity represents the fraction (in mass or volume) of sample which is arranged in the crystalline domains, and it is a very important and peculiar parameter from which many of the physical mechanical properties of the material depend. From a forensic point of view, the attractive feature of the degree of crystallinity is that it is strictly dependent on the processing imposed to the polymer, and so in many cases it reflects a distinctive mark of a particular manufacturer of a mass produced item. To make an example, plastic bags are almost invariably made from polyethylene, a semicrystalline polymer. However, the process chosen to transform the raw material into the finished plastic bags involves melting of the plastic and its extrusion. The temperature, the devices, the time frame for the production of similar products by different manufacturers will be different, and they will induce different crystallisation conditions, which will in turn yield a different structure and thus degree of crystallinity. Examination of the structure is therefore potentially a very powerful approach to the forensic characterisation of polymeric items, but very few literature exist on the subject.

The main technique for the measurement of the degree of crystallinity is powder X-ray diffraction (XRD). XRD has found application in forensic science mainly for the analysis of inorganic species and minerals in soils or metals,^102,103 whereas just a very few examples exist of its use for polymeric items.^69,104–106 Both regions of the semicrystalline framework contribute to the XRD pattern of a polymeric sample: crystalline domains originate sharp reflections, whereas the amorphous zones produce a wide and diffused halo. A fitting procedure of the experimental patterns can be performed, by which the contributions of the crystalline and amorphous domains are deconvoluted¹⁰⁷ (Fig. 4). A certain degree of subjectivity exist in such data treatment, especially in the choice of the fitting functions (usually Gaussians or Lorentzians) and of the width and position of the amorphous halo. In order to make the results independent of such choices, it is paramount to use, in the comparison of samples, always the same width and position of the amorphous halo, changing just its intensity, and the same kind of fitting functions throughout the sample series. Crystallinity can thus be evaluated as the ratio between the area of crystalline peaks over the total area of the diffractograms.

Fig. 4 Example of the deconvolution procedure applied to calculate the degree of crystallinity from the XRD trace of polypropylene.

The crystallinity degree affects many properties of the material, which can in turn be exploited as another way to assess the crystallinity degree by an indirect way. Measurement of refractive index is an example of such a strategy.¹⁰⁸ The approach of characterising materials of forensic interest on the basis of their physical mechanical properties, e.g. tensile properties¹⁰⁹ or ability to transmit light,¹¹⁰ is quite original and would deserve much more attention from the forensic scientific community. Its main drawback is that these tests often require a relatively large number of specimens of standardised shape, which are not always available or obtainable.

Manufacturing processes are not always isotropic in nature, and therefore they tend to impart to the polymer different structures along different directions. A typical example of this is found in polypropylene backings for adhesive tapes. In the production of monoaxially oriented polypropylene (MOPP), the forming polymer film is stretched along a preferential direction, causing the chains to line up anisotropically. This causes the material to exhibit different properties if stress is applied along the direction of stretch or perpendicularly to it. MOPP is used in hand-tearable adhesive tapes because it is resistant along the axis of the tape but it can be easily broken if torn perpendicularly to it.¹¹¹ If XRD are taken of MOPP, mounting the sample along two different directions perpendicular to one another, different patterns will be obtained, reflecting the anisotropic structure of the polymer. Such anisotropy was detected in the development of structural-based characterisation procedures of plastic bags¹⁰⁵ (Fig. 5) and paper.⁶⁹ Quantification of such anisotropy can be carried out evaluating the degree of crystallinity on the XRD patterns from each direction of sample mounting.

Fig. 5 XRD traces of the same plastic bag mounted in two directions, perpendicular to one another. The XRD patterns radically change, reflecting preferential orientation of the polymer crystallites.

Another effective approach for the quantitation of the anisotropy induced by the fabrication process is by polarised FTIR spectroscopy.^112–115 Increasing draw ratios in the spinning process of fibres or in the manufacturing of films induce the orientation of the macromolecular chains (and thus of their functional groups) along the axis of the fibre. This can be detected monitoring the intensity of the signal due to the functional groups present in the polymer chains as a function of the polarization of the incoming IR beam. Dichroic ratios quantify this effect and can be related, by appropriate calibration, to the draw ratio exerted in the manufacturing process. Cho et al. were able to classify polyester fibres into individual groups by this approach¹¹⁴ and applied it to acrylic fibres also.¹¹²Fig. 6 schematises the potential of polarized FTIR in acquiring information on the structure and orientation of macromolecular chains.

Fig. 6 Schematic showing the different orientation of carbonyl groups in an unoriented and an oriented polyester, with the correspondent expected polarized infrared absorption spectra (red and blue traces correspond to an impinging IR beam polarized along the transverse and drawing direction, respectively).

FTIR and Raman are also suitable methods for the quantification of the degree of crystallinity of some kinds of polymers. For example vibrational bands ascribable to crystalline and amorphous domains have been identified for poly(ethylene terephthalate),^116,117 polypropylene^118,119 and polyethylene.¹²⁰ The relative intensities of such bands allow to assess the relative quantities of material which is ordered in the crystalline domains and of material which is disordered in amorphous zones. Although not a polymer, also the crystallinity of hydroxyapatite was estimated by IR spectroscopy, which can be of interest in the characterisation of bone remains in forensic and archeoforensic applications.^121,122

Other interesting structural information which can be investigated by vibrational spectroscopy are the crystalline phases attained by the polymer. Polymorphism is a very common feature of polymers, which, depending on crystallisation conditions, can order themselves according to different unit cells. Moreover, different phases can coexist in the same material.^123,124 Obviously XRD is the technique of choice for studying the structure of materials, but it requires quite large sample sizes, not always available in the items commonly found in casework. FTIR or Raman are therefore more convenient methods, since many polymers exhibit different vibrational bands for different crystal phases. Two examples are polyvinylidene fluoride^98,125 and nylon.^126,127

Another family of techniques, widely used by the polymer scientist, but neglected by the forensic scientist is that of thermal analytical methods.⁸⁸ Although in the last few years some examples of applications of such approaches appeared in the forensic literature,^{89,90,106,128–133} the potential of thermal analysis appears to be still underexploited. Sample sizes of about 5 mg are necessary for performing this type of analysis, quite too large for fibres or paints, but affordable in cases of post-blast residues,^90,132 failure analysis,¹³¹ drug packaging or latex gloves examination.^89,106,130 Automotive plastic items on car crash scenes or materials involved in civil liability lawsuits are other possible examples in which characterisation by thermal analysis can contribute. In particular, two techniques are particularly helpful for analysing polymers: TG and differential scanning calorimetry (DSC).⁸⁸ TG has been already covered in section 2. The most useful data for a forensic scientist obtainable by DSC are those related to the melting behaviour. Melting temperature and enthalpy are in fact related to the structure and morphology of the polymer. The melting enthalpy is proportional to the degree of crystallinity, thus DSC is still another method useful for the measurement of this feature.^130,131 The shape of the melting peak is significant, also. Often multiple melting peaks are observed, which are indicative of multiple populations of crystallites, and thus can be directly connected to the thermal history experienced by the sample during processing.^130,132 Although DSC data are very informative, overinterpretation should be avoided. The melting and crystallisation behaviour depends on a number of factors, the most important being the formulation, the microstructure of the macromolecular chain and the crystallisation conditions. DSC, and more widely all the techniques that focus on the structure of the polymer semicrystalline framework, should not be used as means of identification of a certain type of polymer. FTIR or NMR are more suitable for that purpose. There are instances in the literature where XRD and DSC have been proposed as methods to discriminate between LDPE, LLDPE and HDPE,^86,114 but this is not rigorous, because similar structures could be attained by different polymers, tuning formulation and crystallisation conditions. DSC and XRD must always be intended, in a forensic context, as a complement to the techniques capable of yielding information on the polymer microstructure (section 2.), and can not substitute them. In other words, information on the polymer nature and microstructure should be acquired first, subsequently structure and morphology must be investigated, and the data from both these approaches will lead to significant conclusions.

The glass transition temperature (T_g) is a polymer-specific parameter that is easily, reliably and accurately measured by DSC. Below T_g the material behaves as a glassy solid, beyond T_g the polymer acquires elastomeric properties. Since its value is fundamental for determining the suitability of the material to the intended end use, in the plastics industry, T_g is often tailored by addition of additives, called plastifiers, and by control of the structure and morphology. Polyvinylchloride (PVC) is a classical example of a polymer whose T_g is widely varied, especially by the use of plastifiers, as a function of the intended application. T_g is an indicator of the manufacturing process and thus may be of help to the forensic scientist.

DSC is often the method of choice in failure analysis investigations. For thermosets, it allows to check if the reticulation was optimal.¹⁰⁶ In the case of thermoplastics, the dependence of the crystallisation behaviour both on the formulation and on the crystallisation conditions can be exploited to acquire information on what went wrong in a defective production batch (and thus who to blame for damage or failure). The deconvolution of formulation and processing can be made by applying a triple ramp temperature program. The first heating is aimed at determining the melting behaviour of the sample as a function of both the raw materials, i.e. formulation, and of its thermal history, i.e. processing. After the first heating, the previous thermal history experienced by the sample is cancelled. The signals obtained by the subsequent cooling and heating ramps, since they follow the erasure of the thermal history, are dependent only on the microstructure of the polymer.¹³⁰ Comparison with a non-defective batch, allows then to individuate if the non-conformity happens in the first or second heating run, therefore the reason for the inefficient behaviour can be ascribed to the correct source, provided that it has structural reasons.

In the development of an analytical method focused on the structural characterisation, the effect of weathering must always be taken into account because the structure of polymers is sensitive to annealing or thermal treatments, although very extreme environmental conditions (very high temperatures for very long times) are normally required to produce measurable structural modifications.¹¹⁷ The possible effect of weathering should be considered also when interpreting the chemical information included in spectral data. For example, suppose that the IR spectrum of a questioned item differs from that of its possible source object by the presence of a carbonyl band due to oxidation at around 1700 cm⁻¹. The analyst should determine if that oxidation can be due to the crime scene environment, rather than to the history experienced by the object during its life until the commitment of the crime. In other words, for a significant and probative result, the influence of the crime scene environment on shaping the structure and morphology of the trace item must be ruled out, either by tests simulating the effect of the crime scene conditions¹¹⁷ on a similar plastic item or by analysing the circumstances of the case (e.g. the crime scene is protected, such as the inside of a building not exposed to extreme temperature changes or meteorological events).

An attractive technique that could help shed light on the structure and morphology of a variety of polymers items, in a nondestructive way, is NMR relaxometry. This technique basically studies the decay of the excitation of the protons that return to their equilibrium state after the pulse of a NMR experiment. The free induction decay (FID) signal so obtained is followed as a function of time. Solid matter damps oscillations and produces a fast signal decay. In a liquid, the more mobile surroundings cause less damping and thus gives a slower decay of the signal. Semicrystalline polymers often have within their bulk both stiff and hard regions, e.g. crystalline phase, and softer regions, e.g. amorphous zones. Some materials, such as polyurethanes, contain hard and soft segments in their macromolecular chains. Polymer chains in rubbers, elastomers and crosslinked materials have a different mobility if they are close or far from a crosslinking point.

A structural and morphological characterisation of such materials can be attained by analysis of the FID signal shape, which depends on the relative quantities of the solid, i.e. stiff or hard regions, and liquid, i.e. soft, components of the sample (Fig. 7).^134,135

Fig. 7 ¹H NMR T₂ relaxation decay for crosslinked and non-crosslinked ethylene-propylene-diene-monomer rubber samples. Reprinted with permission from ref. 134

No application in the forensic analysis of plastic items was found in the literature for NMR relaxometry, but its low cost with respect to standard NMR equipment, its portability, its non-destructivness, its reduced need for sample preparation call for a greater consideration by the scientific community.

5. Conclusions

A number of analytical and characterisation techniques have been presented, which are able to yield information on the formulation, on the microstructure, and on the structure and morphology of polymeric items of forensic interest. Some of the methods are more established and are routinely applied, others are less exploited in contact trace analysis and deserve more study and more interest by the forensic community.

This review was focused on the analytical techniques and on the information which can be obtained therefrom. Table 1 summarises the different groups of techniques.

Table 1 Summary of the techniques which may be of interest for the analysis of polymeric items in a forensic context

Group of techniques	Technique	Target analytes	Information	Remarks
Vibrational spectroscopy	Infrared spectroscopy	Polymer matrix, additives	Formulation, polymer microstructure and structure	Very small sample size, no sample preparation, complementary information on polymer matrix and additives
	Raman spectroscopy	Polymer matrix, additives	Formulation, polymer microstructure
Elemental analysis	LA-ICP-MS	Inorganic additives	Formulation	No sample preparation, good spatial resolution
	LIBS	Inorganic additives	Formulation	No sample preparation, portability of instruments
	ICP-MS	Inorganic additives	Formulation	Sample preparation necessary
	SEM-EDX	Inorganic additives	Formulation	Good spatial resolution, low sensitivity
	XRF	Inorganic additives	Formulation	Good spatial resolution, portability of instruments, hyphenation with Raman spectroscopy
Chemical imaging	Cathodoluminescence	Inorganic additives	Formulation	Complements elemental analysis, direct visualisation of the location and dispersion of additives, information on the structure of additives
	IR/Raman/VIS absorption spectral imaging	Polymer matrix, additives	Formulation	Direct visualisation of the location and dispersion of additives
Mass spectrometry	Stable isotope ratio analysis	Polymer matrix, additives	Formulation	Provenance studies
	DESI	Organic additives	Formulation	Ambient ionization technique, no sample preparation
	DART	Polymer matrix, organic additives	Formulation	Ambient ionization technique, no sample preparation
Separation processes	Capillary electrophoresis	Additives	Formulation	Sample preparation necessary
	Pyr-GC-MS	Polymer matrix, organic additives	Polymer microstructure	Small sample size, reproducibility is an issue
	SEC	Polymer matrix	Polymer microstructure	Very rarely applied in forensic science, relatively large sample size
Thermal analysis	TGA	Polymer matrix, additives	Formulation	Relatively large sample size
	DSC	Polymer matrix	Polymer structure	Relatively large sample size
Magnetic spectroscopy	NMR	Polymer matrix	Polymer microstructure	Relatively large sample size, sample preparation necessary
	NMR relaxometry	Polymer matrix	Polymer microstructure	No sample preparation, non-destructive, portability of instruments
X-ray diffraction	X-ray diffraction	Polymer matrix, inorganic additives	Polymer structure, formulation	Relatively large sample size

---

It was shown that a large array of data can be obtained on the various facets of polymer materials. If on one hand this is indeed an advantage, it can also be, on the other hand, a source of confusion, because among so much information it can be difficult to identify the significant data. The data and the features that carry the most discriminative information can be individuated by methods such as principal component analysis,^82,84,136 clustering techniques,^82,137 likelihood ratio models,¹³⁸ or other multivariate data analysis techniques.¹⁴ By these statistical techniques it is possible to identify which features are the most significant for the description of the sample, and therefore which analytical methods are best suited for an effective characterization protocol.

Although it is not correct to equate the possibility to discriminate between different samples with an increased significance of evidence, it is nevertheless true that oftentimes forensic scientists deal with comparisons. The discriminating power of each proposed technique should therefore be determined before applying the method to casework, to aid in the interpretation of the significance of the results when reporting in Court.^{4,29,69,89,105,130}

Suitable interpretation approaches must be developed, which are statistically sound and which may offer a tangible contribution to the verification of the hypotheses put forth by prosecution and defence.^4,139,140

Many aspects of forensic science can take benefit from a cross-fertilisation with polymer science.

Despite the widespread use of plastics, most of the published literature in the field of forensic engineering, failure analysis, toolmark analysis and product liability still deal with metallic materials.^{106,141–145} Availability of peer reviewed literature is vital, if one wants to support an opinion in Court.

Association of data on the packaging of substances of abuse, along with the composition of the illegal drug, can be of significant value in forensic intelligence, i.e. in the process used to infer and characterise links between samples that originate from the same and different seizures with the aim of reconstructing the broad dynamics of organized crime.^{105,130,146–150}

A better knowledge of pyrolysis processes of polymers is necessary for a thorough interpretation of data in fire-related investigations.¹⁵¹ Polymer science could help in shedding light on topics like these, allowing to further optimise already established analytical protocols.

On the other hand, an inspection of the polymer science literature can readily show that the approach in using most of the methods cited in the review is quite different from that of the analytical chemistry field. Little attention is paid to statistical analysis and results are usually employed to interpret trends of a specific property in a given series of samples, rather than to obtaining significant absolute values for a measurement. Amending these flaws with a more analytical mindset will no doubt benefit the polymer science community as well.

Notes and references

1. O. Ribaux, A. Baylon, C. Roux, O. Delémont, E. Lock, C. Zingg and P. Margot, Forensic Sci. Int., 2010, 195, 10–16.
2. E. Locard, L'enquète criminelle et les méthodes scientifiques, Flammarion, Paris, 1920.
3. R. Cook, I. W. Evett, G. Jackson, P. J. Jones and J. A. Lambert, Sci. Justice, 1998, 38, 231–239.
4. C. Aitken and F. Taroni, Statistics and the Evaluation of Evidence for Forensic Scientists, Wiley, Chichester, 2004.
5. J. V. Goodpaster and E. A. Liszewski, Anal. Bioanal. Chem., 2009, 394, 2009–2018.
6. J. Robertson and M. Grieve, Forensic Examination of Fibres, Taylor & Francis, London, 1999.
7. C. R. Dockery, A. R. Stefan, A. A. Nieuwland, S. N. Roberson, B. M. Baguley, J. E. Hendrix and S. L. Morgan, Anal. Bioanal. Chem., 2009, 394, 2095–2103.
8. B. C. Vann, S. M. Angel, J. E. Hendrix, E. G. Bartick and S. L. Morgan, Appl. Spectrosc., 2009, 63, 407–411.
9. M. W. Tungol, E. G. Bartick and A. Montaser, Appl. Spectrosc., 1993, 47, 1655–1658.
10. S. Carter, A. S. Fisher, P. S. Goodall, M. W. Hinds, S. Lancaster and S. Shore, J. Anal. At. Spectrom., 2009, 24, 1599–1656.
11. M. West, A. T. Ellis, P. J. Potts, C. Streli, C. Vanhoof, D. Wegrzynek and P. Wobrauschek, J. Anal. At. Spectrom., 2009, 24, 1289–1326.
12. N. Aziz, P. F. Greenwood, K. Grice, R. J. Watling and W. van Bronswijk, J. Forensic Sci., 2008, 53, 1130–1137.
13. S. H. Wise and J. R. Almirall, Appl. Opt., 2008, 47, G15–G20.
14. A. Sarkar, S. K. Aggarwal and D. Alamelu, Anal. Methods, 2010, 2, 32–36.
15. I. Deconinck, C. Latkoczy, D. Gunther, F. Govaert and F. Vanhaecke, J. Anal. At. Spectrom., 2006, 21, 279–287.
16. A. L. Hobbs and J. R. Almirall, Anal. Bioanal. Chem., 2003, 376, 1265–1271.
17. C. J. Scadding, R. J. Watling and A. G. Thomas, Talanta, 2005, 67, 414–424.
18. D. B. Aeschliman, S. J. Bajic, D. P. Baldwin and R. S. Houk, Anal. Chem., 2004, 76, 3119–3125.
19. A. M. Dobney, W. Wiarda, P. de Joode and G. J. Q. van der Peijl, J. Anal. At. Spectrom., 2002, 17, 478–484.
20. L. D. Spence, A. T. Baker and J. P. Byrne, J. Anal. At. Spectrom., 2000, 15, 813–819.
21. Y. Higashikawa, M. Kasamatsu, Y. Suzuki and S. Suzuki, Bunseki Kagaku, 2007, 56, 1183–1186.
22. G. Gautier, A. Bezur, K. Muir, F. Casadio and I. Fiedler, Appl. Spectrosc., 2009, 63, 597–603.
23. R. D. Koons, J. Forensic Sci., 1996, 41, 199–205.
24. C. Roux, J. Cartier and M. Grieve, J. Forensic Sci., 1997, 42, 1019–1026.
25. J. Zieba-Palus, R. Borusiewicz and M. Kunicki, Forensic Sci. Int., 2008, 175, 1–10.
26. V. Otieno-Alego, J. Raman Spectrosc., 2009, 40, 948–953.
27. C. S. Palenik and J. Buscaglia, in Forensic analysis on the cutting edge. New metods for trace evidence analysis., ed. R. D. Blackledge, Wiley, Hoboken, NJ, 2007, p. 141–174.
28. C. Ricci, K. L. A. Chan and S. G. Kazarian, Appl. Spectrosc., 2006, 60, 1013–1021.
29. G. Payne, C. Wallace, B. Reedy, C. Lennard, R. Schuler, D. Exline and C. Roux, Talanta, 2005, 67, 334–344.
30. L. Flynn, R. O'Leary, C. Roux and B. J. Reedy, J. Forensic Sci., 2006, 51, 586–596.
31. K. Flynn, R. O'Leary, C. Lennard, C. Roux and B. J. Reedy, J. Forensic Sci., 2005, 50, 832–841.
32. P. R. Ng, S. Walker, M. Tahtouh and B. J. Reedy, Anal. Bioanal. Chem., 2009, 394, 2039–2048.
33. I. M. Kempson, K. P. Kirkbride, W. M. Skinner and J. Coumbaros, Talanta, 2005, 67, 286–303.
34. S. Benson, C. Lennard, P. Maynard and C. Roux, Forensic Sci. Int., 2006, 157, 1–22.
35. A. T. Quirk, J. M. Bellerby, J. F. Carter, F. A. Thomas and J. C. Hill, Sci. Justice, 2009, 49, 87–93.
36. S. J. Benson, C. J. Lennard, D. M. Hill, P. Maynard and C. Roux, J. Forensic Sci., 2010, 55, 193–204.
37. S. J. Benson, C. J. Lennard, P. Maynard, D. M. Hill, A. S. Andrew, K. Neal, H. Stuart-Williams, J. Hope, G. Stewart Walker and C. Roux, J. Forensic Sci., 2010, 55, 205–212.
38. S. J. Benson, C. J. Lennard, P. Maynard, D. M. Hill, A. S. Andrew and C. Roux, Sci. Justice, 2009, 49, 73–80.
39. S. J. Benson, C. J. Lennard, P. Maynard, D. M. Hill, A. S. Andrew and C. Roux, Sci. Justice, 2009, 49, 81–86.
40. L. J. Reidy, W. Meier-Augenstein and R. M. Kalin, Rapid Commun. Mass Spectrom., 2005, 19, 1899–1905.
41. E. Taylor, J. F. Carter, J. C. Hill, C. Morton, N. NicDaeid and R. Sleeman, Forensic Sci. Int., 2008, 177, 214–220.
42. J. F. Carter, P. L. Grundy, J. C. Hill, N. C. Ronan, E. L. Titterton and R. Sleeman, Analyst, 2004, 129, 1206–1210.
43. E. O'Neill, D. Harrington and J. Allison, Anal. Bioanal. Chem., 2009, 394, 2029–2038.
44. S. Stachura, V. J. Desiderio and J. Allison, J. Forensic Sci., 2007, 52, 595–603.
45. T. Mukai, H. Nakazumi, S. Kawabata, M. Kusatani, S. Nakai and S. Honda, J. Forensic Sci., 2008, 53, 107–115.
46. C. Berger-Karin, U. Hendriks and J. Geyer-Lippmann, J. Forensic Sci., 2008, 53, 989–992.
47. L. J. Soltzberg, A. Hagar, S. Kridaratikorn, A. Mattson and R. Newman, J. Am. Soc. Mass Spectrom., 2007, 18, 2001–2006.
48. D. R. Ifa, A. U. Jackson, G. Paglia and R. G. Cooks, Anal. Bioanal. Chem., 2009, 394, 1995–2008.
49. R. B. Cody, J. A. Laramée and H. D. Durst, Anal. Chem., 2005, 77, 2297–2302.
50. J. A. Laramée, R. B. Cody, J. M. Nilles and H. D. Durst, in Forensic analysis on the cutting edge. New metods for trace evidence analysis., ed. R. D. Blackledge, Wiley, Hoboken, NJ, 2007, p. 196.
51. S. Czernik, C. C. Elam, R. J. Evans, R. R. Meglen, L. Moens and K. Tatsumoto, J. Anal. Appl. Pyrolysis, 1998, 46, 51–64.
52. J. H. Bugler, H. Buchner and A. Dallmayer, J. Forensic Sci., 2005, 50, 1209–1214.
53. Y. Kumooka, Forensic Sci. Int., 2008, 176, 111–120.
54. J. J. Thomas, Z. X. Shen, R. Blackledge and G. Siuzdak, Anal. Chim. Acta, 2001, 442, 183–190.
55. F. Tagliaro and F. Bortolotti, Electrophoresis, 2006, 27, 231–243.
56. F. Tagliaro and F. Bortolotti, Electrophoresis, 2008, 29, 260–268.
57. F. Burger, M. Dawson, C. Roux, P. Maynard, P. Doble and P. Kirkbride, Talanta, 2005, 67, 368–376.
58. I. P. Keen, G. W. White and P. M. Fredericks, J. Forensic Sci., 1998, 43, 82–89.
59. G. Jochem and R. J. Lehnert, Sci. Justice, 2002, 42, 215–221.
60. P. Buzzini, G. Massonnet and F. M. Sermier, J. Raman Spectrosc., 2006, 37, 922–931.
61. Y. S. Giang, S. M. Wang, L. L. Cho, C. K. Yang and C. C. Lu, J. Forensic Sci., 2002, 47, 625–629.
62. J. F. Amick and C. W. Beheim, J. Forensic Sci., 2002, 47, 619–624.
63. A. Koçak, L. M. De Cotiis and D. B. Hoffman, Forensic Sci. Int., 2010, 195, 36.
64. A. Kocak, J. P. Lucania and S. L. Berets, Appl. Spectrosc., 2009, 63, 507–511.
65. X. Cao, Z. Wen, A. Vance and G. Torraca, Appl. Spectrosc., 2009, 63, 830–834.
66. G. Li, G. Torraca, W. Jing and Z. Wen, Vib. Spectrosc., 2009, 50, 152–159.
67. P. Maynard, K. Allwell, C. Roux, M. Dawson and D. Royds, Forensic Sci. Int., 2001, 124, 140–156.
68. A. Kher, M. Mulholland, B. Reedy and P. Maynard, Appl. Spectrosc., 2001, 55, 1192–1198.
69. V. Causin, C. Marega, A. Marigo, R. Casamassima, G. Peluso and L. Ripani, Forensic Sci. Int., 2010, 197, 70–74.
70. P. Maynard, K. Gates, C. Roux and C. Lennard, J. Forensic Sci., 2001, 46, 280–287.
71. R. J. H. Clark, Appl. Phys. A: Mater. Sci. Process., 2007, 89, 833–840.
72. S. E. J. Bell, L. A. Fido, S. J. Speers and W. J. Armstrong, Appl. Spectrosc., 2005, 59, 100–108.
73. S. E. J. Bell, L. A. Fido, S. J. Speers, W. J. Armstrong and S. Spratt, Appl. Spectrosc., 2005, 59, 1333–1339.
74. S. E. J. Bell, L. A. Fido, S. J. Speers, W. J. Armstrong and S. Spratt, Appl. Spectrosc., 2005, 59, 1340–1346.
75. J. De Gelder, P. Vandenabeele, F. Govaert and L. Moens, J. Raman Spectrosc., 2005, 36, 1059–1067.
76. M. C. Grieve, R. M. E. Griffin and R. Malone, Sci. Justice, 1998, 38, 27–37.
77. M. C. Grieve, Sci. Justice, 1995, 35, 179–190.
78. V. Causin, C. Marega, S. Schiavone and A. Marigo, Forensic Sci. Int., 2005, 151, 125–131.
79. G. Massonnet, P. Buzzini, G. Jochem, M. Stauber, T. Coyle, C. Roux, J. Thomas, H. Leijenhorst, Z. Van Zanten, K. Wiggins, C. Russell, S. Chabli and A. Rosengarten, J. Forensic Sci., 2005, 50, 1028–1038.
80. J. Thomas, P. Buzzini, G. Massonnet, B. Reedy and C. Roux, Forensic Sci. Int., 2005, 152, 189–197.
81. J. M. Milczarek and J. Zieba-Palus, J. Anal. Appl. Pyrolysis, 2009, 86, 252–259.
82. R. Kristensen, S. Coulson and A. Gordon, J. Anal. Appl. Pyrolysis, 2009, 86, 90–98.
83. M. Zellner and L. Quarino, J. Forensic Sci., 2009, 54, 1022–1028.
84. V. Causin, C. Marega, S. Schiavone, V. Della Guardia and A. Marigo, J. Anal. Appl. Pyrolysis, 2006, 75, 43–48.
85. J. M. Challinor, Forensic Sci. Int., 1983, 21, 269–285.
86. J. M. Challinor, J. Anal. Appl. Pyrolysis, 1989, 16, 323–333.
87. J. M. Challinor, J. Anal. Appl. Pyrolysis, 1991, 18, 233–244.
88. J. D. Menczel and R. B. Prime, Thermal Analysis of Polymers. Fundamentals and Applications, Wiley, Hoboken, NJ, 2009.
89. V. Causin, C. Marega, A. Marigo, P. Carresi, V. Della Guardia and S. Schiavone, Forensic Sci. Int., 2009, 188, 57–63.
90. E. C. Ihms and D. W. Brinkman, J. Forensic Sci., 2004, 49, 505–510.
91. Y. Kumooka, Forensic Sci. Int., 2007, 171, 5–8.
92. D. Bikiaris, V. Karavelidis and E. Karavas, Molecules, 2009, 14, 2410–2430.
93. D. D. Lu, J. C. Yuan and Z. Q. Lei, Polym. Adv. Technol., 2009, 20, 536–540.
94. J. L. Koenig, Spectroscopy of polymers, American Chemical Society, Washington, 1992.
95. D. Campbell, R. A. Pethrick and J. R. White, Polymer Characterization. Physical Techniques, Stanley Thornes, Cheltenham, UK, 2000.
96. S. Al-AbdulRazzak, E. A. Lofgren and S. A. Jabarin, Polym. Int., 2002, 51, 174–182.
97. Z. Shi, X. Su, Z. Zheng and X. Wang, e-polymers, 2006 , art. no. 077.
98. M. Kobayashi, K. Tashiro and H. Tadakoro, Macromolecules, 1975, 8, 158.
99. N. A. Hakeem, H. I. Abdelkader, N. A. El-sheshtawi and I. S. Eleshmawi, J. Appl. Polym. Sci., 2006, 102, 2125–2131.
100. H. M. G. Correia and M. M. D. Ramos, Ferroelectrics, 2006, 338, 185–191.
101. A. H. Oraby, Polym. Test., 2000, 19, 865–878.
102. D. F. Rendle, Adv. X-Ray Anal., 2003, 46, 17–24.
103. W. Kugler, Adv. X-Ray Anal., 2003, 46, 1–16.
104. T. Hashimoto, D. G. Howitt, D. P. Land, F. A. Tulleners, F. A. Springer and S. Wang, J. Forensic Sci., 2007, 52, 1082–1088.
105. V. Causin, C. Marega, P. Carresi, S. Schiavone and A. Marigo, Forensic Sci. Int., 2007, 168, 37–41.
106. V. Causin, C. Marega and A. Marigo, Eng. Failure Anal., 2007, 14, 1394–1400.
107. A. M. Hindeleh and D. J. Johnson, J. Phys. D: Appl. Phys., 1971, 4, 259–263.
108. V. Causin, C. Marega, S. Schiavone and A. Marigo, Forensic Sci. Int., 2005, 149, 193–200.
109. R. Kumar, Forensic Sci. Int., 2010, 195, 128–131.
110. C. E. H. Berger, Forensic Sci. Int., 2009, 192, 1–6.
111. J. M. Smith, in Forensic analysis on the cutting edge. New metods for trace evidence analysis., ed. R. Blackledge, Wiley, Hoboken, NJ, 2007.
112. L. L. Cho, J. A. Reffner, B. M. Gatewood and D. L. Wetzel, J. Forensic Sci., 1999, 44, 275–282.
113. M. Boulet-Audet, T. Lefevre, T. Buffeteau and M. Pezolet, Appl. Spectrosc., 2008, 62, 956–962.
114. L. Cho, J. A. Reffner and D. L. Wetzel, J. Forensic Sci., 1999, 44, 283–291.
115. C. Pellerin, M. E. Rousseau, R. E. Prud'homme and M. Pézolet, Appl. Spectrosc., 2002, 56, 17–23.
116. V. Causin, C. Marega, G. Guzzini and A. Marigo, Appl. Spectrosc., 2004, 58, 1272–1276.
117. V. Causin, C. Marega, G. Guzzini and A. Marigo, J. Forensic Sci., 2005, 50, 519–525.
118. K. Rajakumar, V. Sarasvathy, A. Thamarai Chelvan, R. Chitra and C. T. Vijayakumar, J. Polym. Environ., 2009, 17, 191–202.
119. A. R. Horrocks, J. Mwila, M. Miraftab, M. Liu and S. S. Chohan, Polym. Degrad. Stab., 1999, 65, 25–36.
120. S. S. Cherukupalli and A. A. Ogale, Polym. Eng. Sci., 2004, 44, 1484.
121. G. Nagy, T. Lorand, Z. Patonai, G. Montsko, I. Bajnoczky, A. Marcsik and L. Mark, Forensic Sci. Int., 2008, 175, 55–60.
122. T. J. U. Thompson, M. Gauthier and M. Islam, J. Archaeol. Sci., 2009, 36, 910–914.
123. V. Causin, C. Marega, R. Saini, A. Marigo and G. Ferrara, J. Therm. Anal. Calorim., 2007, 90, 849–857.
124. A. Marigo, V. Causin, C. Marega and P. Ferrari, Polym. Int., 2004, 53, 2001–2008.
125. V. Causin, M. L. Carraro, C. Marega, R. Saini, S. Campestrini and A. Marigo, J. Appl. Polym. Sci., 2008, 109, 2354–2361.
126. M. Kyotani and S. Mitsuhashi, J. Polym. Sci. Part A: Polym. Chem., 1972, 10, 1497.
127. O. Monticelli, Z. Musina, F. Ghigliotti, S. Russo and V. Causin, e-polymers, 2007 , art. no. 124.
128. S. Raja, P. S. Thomas, B. H. Stuart, J. P. Guerbois and C. O'Brien, J. Therm. Anal. Calorim., 2009, 98, 173–176.
129. R. I. Medeiros, N. R. Antoniosi Filho and M. I. Goncalves Leles, J. Therm. Anal. Calorim., 2009, 97, 337–342.
130. V. Causin, C. Marega, P. Carresi, S. Schiavone and A. Marigo, Forensic Sci. Int., 2006, 164, 148–154.
131. M. Sajwan, S. Aggarwal and R. B. Singh, Forensic Sci. Int., 2008, 175, 130–133.
132. T. Tsukame, M. Kutsuzawa, H. Sekine, H. Saitoh and Y. Shibasaki, J. Therm. Anal. Calorim., 1999, 57, 847–851.
133. A. Riga, Thermochim. Acta, 1998, 324, 151–163.
134. R. A. Orza, P. C. M. M. Magusin, V. M. Litvinov, M. van Duin and M. A. J. Michels, Macromolecules, 2007, 40, 8999–9008.
135. S. C. Chinn, A. Cook-Tendulkar, R. Maxwell, H. Wheeler, M. Wilson and Z. H. Xie, Polym. Test., 2007, 26, 1015–1024.
136. R. H. Brody, H. G. M. Edwards and A. M. Pollard, Anal. Chim. Acta, 2001, 427, 223–232.
137. F. Ratle, C. Gagne, A. Terrettaz-Zufferey, M. Kanevski, P. Esseiva and O. Ribaux, Chemom. Intell. Lab. Syst., 2008, 90, 123–131.
138. G. Zadora and T. Neocleous, Anal. Chim. Acta, 2009, 642, 266–278.
139. D. B. Hibbert and N. Armstrong, Chemom. Intell. Lab. Syst., 2009, 97, 211–220.
140. V. Causin, S. Schiavone, A. Marigo and P. Carresi, Forensic Sci. Int., 2004, 141, 159–170.
141. I. Juarez-Ferreras, C. Gonzalez-Nicieza, A. Menendez-Diaz, A. E. Alvarez-Vigil and M. L. Alvarez-Fernandez, Eng. Failure Anal., 2009, 16, 2357–2370.
142. C. R. Gagg and P. R. Lewis, Eng. Failure Anal., 2009, 16, 1775–1793.
143. F. Albermani, S. Kitipornchai and R. W. K. Chan, Eng. Failure Anal., 2009, 16, 1922–1928.
144. D. Muller and O. Israelsohn-Azulay, J. Forensic Sci., 2009, 54, 128–131.
145. V. Causin, C. Marega and S. Schiavone, Forensic Sci. Int., 2005, 148, 157–162.
146. P. Esseiva, S. Ioset, F. Anglada, L. Gaste, O. Ribaux, P. Margot, A. Gallusser, A. Biedermann, Y. Specht and E. Ottinger, Forensic Sci. Int., 2007, 167, 247–254.
147. R. Sugita, K. Sasagawa and S. Suzuki, J. Forensic Sci., 2009, 54, 1341–1348.
148. J. Huttunen, M. Dawson, C. Roux and J. Robertson, Australian J. Forensic Sci., 2009, 41, 63–72.
149. J. Huttunen, P. A. Doble, M. Dawson, C. Roux and J. Robertson, Australian J. Forensic Sci., 2008, 40, 73–83.
150. J. Huttunen, C. E. Austin, M. Dawson, C. Roux and J. Robertson, Australian J. Forensic Sci., 2007, 39, 93–106.
151. J. R. Almirall and K. G. Furton, J. Anal. Appl. Pyrolysis, 2004, 71, 51–67.

---