The current methods for identifying ignitable liquid residues in fire debris are heavily based on the holistic, qualitative interpretation of chromatographic patterns with the mass spectral identification of selected peaks. The identification of neat, unweathered ignitable liquids according to ASTM 1618 using these methods is relatively straightforward for the trained analyst. The challenges in fire debris analysis arise with phenomena such as evaporation, substrate interference, and biodegradation. These phenomena result in alterations of chromatographic patterns which can lead to misclassifications or false negatives. The biodegradation of ignitable liquids is generally known to be more complex than evaporation [20], and proceeds in a manner that is dependent on numerous factors such as: composition of the petroleum product/ignitable liquid, structure of the hydrocarbon compound, soil type, bacterial community, the type of microbial metabolism that is occurring, and the environmental conditions surrounding in the sample. While nothing can be done to prevent the biodegradation, continued research on biodegraded ignitable liquids and the characterization of the trends observed may be able to provide insight into how an analyst can identify a biodegraded ignitable liquid residue. This research utilized normalized abundance values of select ions from pre-existing gas chromatography-mass spectrometry (GC-MS) data on samples from three different gasoline and diesel biodegradation studies. A total of 18 ions were selected to indicate the presence of six hydrocarbon classes (three each for alkanes, aromatics, cycloalkanes, naphthalenes, indanes, and adamantanes) based on them being either base peaks or high abundance peaks within the electron impact mass spectra of compounds within that hydrocarbon class. The loss of ion abundance over the degradation periods was assessed by creating scatter plots and performing simple linear regression analyses. Coefficient of determination values, the standard error of the estimate, the slope, and the slope error of the best fit line were assessed to draw conclusions regarding which classes exhibited desirable characteristics, relative to the other classes, such as a linear degradation, low variation in abundance within the sampling days, and a slow rate of abundance loss over the degradation period. Additional analyses included two-way analysis of the variance (ANOVA), to assess the effects of time as well as different soil type on the degradation of the hydrocarbons, stepwise multinomial logistic regressions to identify which classes were the best predictors of the type of ignitable liquid, and one-way ANOVAs to determine where the differences in the ratios of hydrocarbon classes existed within each of the ignitable liquids, as well as between the two liquids. Hydrocarbon classes identified as exhibiting characteristics such as slow and/or reliable rates of abundance loss during biodegradation are thought of as desirable for future validation studies, where specific ranges of hydrocarbon class abundance(s) may be used to identify the presence of a biodegraded ignitable liquid. Classes of hydrocarbons that have experienced biodegradation that maintain an abundance close to that of a neat, non degraded counterpart, or that reliably degrade and have predictable abundance levels given a particular period of degradation, would be instrumental in determining whether or not an unknown sample contains an ignitable liquid residue. It is the hope that these assessments will not only provide helpful information to future researchers in the field of fire debris analysis, but that they will create interest in the quantitative, statistical assessment of ignitable liquid data for detection and identification purposes.

Part 1: The emergence of environmentally friendly commercial products and their impact on fire debris analysis. Environmentally friendly products (i.e. green products) are environmentally preferable choices relative to comparable commercial products. They are readily available to the public, often highly flammable, and can be used by criminals as accelerants to facilitate the start and/or spread of fire. It is critical for analysts to have an understanding of their composition and chromatographic characteristics. Green products include paint thinners, solvents, removers, and cleaning and surface preparation products. As the composition of commercial products continually change over time, the fire debris community needs to be aware of the variety of environmentally friendly ignitable liquids that could be encountered during casework. Traditionally, when fire debris analysts have been trained, they are taught that most of the ignitable liquid residues they will encounter in casework are petroleum-based products. With the increasing emergence of non-petroleum based green products in the consumer marketplace, such products may be encountered more often than ever before in fire debris evidence submitted to forensic laboratories. Analysts should become familiar with the chromatographic features of these products as neat liquids as well as when present in fire debris samples. The purpose of this study is to introduce fire debris analysts to the prevalence of green products and increase knowledge regarding a variety of green product compositions and the characteristics they exhibit when analyzed as neat liquids and in "mock" fire debris samples. Several green products were analyzed as neat liquid samples and subsequently extracted from fire debris samples using typical fire debris extraction and analysis techniques in order to familiarize fire debris analysts with the chromatographic and mass spectral features of these products. General information about different types of green commercial products, their chromatographic and mass spectral characteristics, and their interpretation will be summarized. Analytical methods were developed for the analysis of environmentally friendly products and included considerations of gas chromatography oven temperature and ramp rate, hold times, and flow rate, as well as the scan rate and range of the mass spectrometer. Analyses involving common substrates were performed, including spiking green products onto various substrates with subsequent analysis and comparison of burned and unburned samples. Part 2: Application of fast GC/MS analysis for the identification of ignitable liquids in fire debris samples. Fire debris samples that contain ignitable liquid residues undergo a two-step process of extraction, most commonly via passive adsorption elution (PAE) onto an activated carbon strip, and instrumental analysis by gas chromatography/mass spectrometry. Upon completion of PAE, adsorbed compounds are eluted from the adsorbent with a suitable solvent and analyzed using (GC/MS) for the potential identification of ignitable liquid residues. A thorough evaluation of the literature revealed the average run time for gas chromatography of fire debris samples that contain hydrocarbon or petroleum based ignitable liquids to be 30 minutes. Additionally, a blank sample is run before an evidentiary sample to ensure solvent purity and to ensure any chromatographic carry over has not occurred between subsequent injections. The average run time, along with case volume, extraction times and case reviews contributes significantly to the backlog of samples to be analyzed in most crime laboratories around the country. Fast-GC/MS would significantly reduce analysis time, lower operating costs and would use less consumables. Based on a process known as pattern recognition, an initial goal of a fire debris analyst is to identify a pattern that is consistent with an ignitable liquid class. The standard method followed by most fire debris analysts use or base standard operating procedures (SOPs) on the American Society of Testing and Materials (ASTM) E1618, which defines the classes of commercial ignitable liquids based on chemical composition and boiling point range (or volatility). This study was conducted to optimize current methods of ignitable liquid detection and to optimize fast-GC/MS conditions for the identification of ignitable liquids in fire debris samples. Additionally, this study was conducted to determine if fast-GC/MS can reduce chromatographic separation times without sacrificing peak resolution and subsequently allow for ignitable liquid discrimination. Method development included considerations of flow rate, initial GC oven temperature, ramp rate, and mid and end temperature hold times. Fast-GC/MS conditions were tested on neat ignitable liquids from all nine ASTM E1618 classes. Optimizing fast-GC/MS method parameters led to an increase in sample throughput in comparison to traditional GC/MS methods. As a result, the GC/MS identification of ignitable liquids and their residues was performed in a quarter of the amount of time when compared to traditional methods.

Identification of ignitable liquids in fire debris analysis using pattern recognition is an important step in determining the nature of a suspicious fire. Complex mixtures that make up ignitable liquids are susceptible to microbial degradation when fire debris evidence is presented in the form of soil. Microbial degradation results in a selective metabolism of certain classes of compounds required for identification of an ignitable liquid. Various ignitable liquids that may be used to initiate or propagate a fire contain different classes of organic compounds. These include normal alkanes, branched alkanes, cycloalkanes, aromatics, terpenes, and others. In this work, microbial degradation of nine ignitable liquids in soil was evaluated over a period of twenty-six days. The degradation of aromatic compounds in gasoline was faster with toluene and C2-alkylbenzenes than in C3-alkylbenzenes. However, the overall loss of aromatics made gasoline chromatographically unidentifiable. The complete loss of n-alkanes in medium and petroleum distillates resulted in patterns that resembled naphthenic-paraffinic products. Normal alkanes were more susceptible to microbial degradation than isoalkanes, which was specifically demonstrated in medium and heavy petroleum distillates. In diesel, pristane and phytane remained prominent in comparison to the normally prevalent n-alkanes, which could no longer be detected post-degradation. The degradation of isoalkanes and cycloalkanes was evaluated in a naphthenic-paraffinic product. Isoalkanes were degraded significantly faster than cycloalkanes. The remaining peaks in the naphthenic-paraffinic pattern consisted solely of cycloalkane compounds, and could no longer be classified as a naphthenic-paraffinic product. The terpene compounds in turpentine were also observed to be susceptible to degradation by microorganisms. The loss of !-pinene, limonene, and camphene was significantly noticeable in comparison to other terpene compounds, such as 1,4-cineole. Microbial biodegradation in different soil types was investigated. The difference in soil texture can affect the rate of metabolism of ignitable liquids due to the variance of available oxygen, nutrients and mobility of the microbial population. The degradation of isoalkanes, cycloalkanes, aromatics and heavier normal alkanes was faster in clay, whereas normal alkanes of lower molecular weight were degraded more readily in sand. There has been no explanation of this occurrence within the scientific literature, however it could be hypothesized that the difference in microbial flora and water saturation levels could affect the selective degradation between the two soil types. Fire debris evidence is often stored for long periods of time before analysis due to case backlogs. The storage condition of arson-related soil samples is a sensitive subject. If evidence, containing soil, is stored at room temperature, petroleum compounds in any ignitable liquid residues that are present will be degraded within a week. Therefore, it is important to freeze or refrigerate soil samples. The storage of both refrigerated and frozen soil samples containing gasoline were evaluated over six months. Less than 6% of the aromatic compounds distinctive of gasoline remained when stored at 5 °C, while minimal change was observed in the same compounds when stored at -15 °C. Microbial degradation of petroleum-based ignitable liquids is advantageous from the environmental perspective. However, within the forensic community the effect of microbial action could lead to misclassification or inability to identify the presence of an ignitable liquid in fire debris evidence.

The detection and identification of the oxygenated class of ignitable liquids is a complex issue for the fire debris analyst. The oxygenated compounds are difficult to recover using traditional analytical techniques since their chemical characteristics are vastly different from those of the petroleum products that compose the majority of the ignitable liquid classes. Previous research has demonstrated that the use of zeolite 13X as an adsorbent in heated passive headspace concentration provides increased recovery of oxygenated compounds in comparison to the conventional activated charcoal adsorbent. This hypothesis was further tested in this work using more realistic casework scenarios. Various carpet, carpet padding and wood types were utilized in a number of burn conditions in order to determine if any substrate interferences were present; as well as to monitor the recovery of oxygenated compounds from these substrates using the proposed zeolite extraction method. The substrates explored did not contribute significant background interference to complicate the identification of the oxygenated compounds. In addition, small volumes of the oxygenated ignitable liquids were easily recovered and identified from all burn states using the zeolite method. A dual-mode extraction with both zeolites and activated charcoal strips as adsorbents was utilized with mixtures of oxygenated compounds and petroleum products to determine if a variety of ignitable liquid classes could be detected and identified in the presence of a variety of substrate matrices within a single extraction protocol. The dual-mode extraction showed that both the oxygenated compounds and petroleum products could be detected and identified using a single extraction protocol in the presence of various substrate matrices. Lastly, an experiment was devised to compare the recovery of the oxygenated compounds using various total available surface areas of both zeolites and activated charcoal strips in order to determine which adsorbent exhibits a greater recovery when all other experimental conditions remain constant. When the surface areas were equalized between the zeolites and activated charcoal strips, the activated charcoal exhibited a greater recovery of the oxygenated compounds. However, the cost effectiveness of the zeolites allows for a greater amount of zeolite beads to be used in order to achieve the same recovery as the activated charcoal strips for a much lower price. Therefore, the findings from this work, in combination with previous research, continue to support the use of zeolite 13X as an alternative adsorbent for the recovery of oxygenated ignitable liquids from fire debris evidence.

In this dissertation research, two fundamental studies involving gas chromatography mass spectrometry of n-butylbenzene and pyrolysis products are presented. In the first study, fragmentation pathways of n-butylbenzene in quadrupole ion trap have been investigated. At low energy, product ion corresponding to m/z 92 and m/z 91 are formed via competitive parallel dissociation. Studies have also shown that at higher energy m/z 92 has sufficient internal energy to undergo further fragmentation yielding m/z 91 via consecutive dissociation. Thus in order to discern the fragmentation pathways of n-butylbenzene, the technique of two-dimensional correlation spectroscopy (2DCOS) was applied to the mass spectral data. Application of 2DCOS resulted in two 2D correlation spectra namely synchronous and asynchronous. A third spectra known as coherence spectra was obtained from the ration of asynchronous to synchronous correlation intensities. For the elucidation of n-butylbenzene fragmentation pathways, all the three spectra were utilized in this study. The second study in this dissertation involves investigation of pyrolysis products to aid in fire debris analysis. One of the major concerns in fire debris analysis is that pyrolysis products can mask the patterns of compounds of interest and make the chromatographic results interpretation extremely difficult. One of the approaches for investigating the formation of pyrolysis products is to subject the commonly found building materials to controlled heating in laboratory. In this study, new heating methodologies for controlled heating of substrates involving furnace, paint-cans and flat steel pans have been developed. The substrates used for investigating pyrolysis products were polystyrene, polyvinylchloride, polybutadiene, yellow-pine, nylon carpet and padding. Experiments were also performed to investigate the influence of hydrocarbons on the formation of pyrolysis.

Current fire debris analysis procedure involves using the chromatographic patterns of total ion chromatograms, extracted ion chromatograms, and target compound analysis to identify an ignitable liquid according to the American Society for Testing and Materials (ASTM) E 1618 standard method. Classifying the ignitable liquid is accomplished by a visual comparison of chromatographic data obtained from any extracted ignitable liquid residue in the debris to the chromatograms of ignitable liquids in a database, i.e. by visual pattern recognition. Pattern recognition proves time consuming and introduces potential for human error. One particularly difficult aspect of fire debris analysis is recognizing an ignitable liquid residue when the intensity of its chromatographic pattern is extremely low or masked by pyrolysis products. In this research, a unique approach to fire debris analysis was applied by utilizing the samples' total ion spectrum (TIS) to identify an ignitable liquid, if present. The TIS, created by summing the intensity of each ion across all elution times in a gas chromatography-mass spectrometry (GC-MS) dataset retains sufficient information content for the identification of complex mixtures . Computer assisted spectral comparison was then performed on the samples' TIS by target factor analysis (TFA). This approach allowed rapid automated searching against a library of ignitable liquid summed ion spectra. Receiver operating characteristic (ROC) curves measured how well TFA identified ignitable liquids in the database that were of the same ASTM classification as the ignitable liquid in fire debris samples, as depicted in their corresponding area under the ROC curve. This study incorporated statistical analysis to aid in classification of an ignitable liquid, therefore alleviating interpretive error inherent in visual pattern recognition. This method could allow an analyst to declare an ignitable liquid present when utilization of visual pattern recognition alone is not sufficient.

No description available.

Fire debris analysis currently relies on visual pattern recognition of the total ion chromatograms, extracted ion profiles, and target compound chromatograms to identify the presence of an ignitable liquid. This procedure is described in the ASTM International E1618-10 standard method. For large data sets, this methodology can be time consuming and is a subjective method, the accuracy of which is dependent upon the skill and experience of the analyst. This research aimed to develop an automated classification method for large data sets and investigated the use of the total ion spectrum (TIS). The TIS is calculated by taking an average mass spectrum across the entire chromatographic range and has been shown to contain sufficient information content for the identification of ignitable liquids. The TIS of ignitable liquids and substrates were compiled into model data sets. Substrates are defined as common building materials and household furnishings that are typically found at the scene of a fire and are, therefore, present in fire debris samples. Fire debris samples were also used which were obtained from laboratory-scale and large-scale burns. An automated classification method was developed using computational software that was written in-house. Within this method, a multi-step classification scheme was used to detect ignitable liquid residues in fire debris samples and assign these to the classes defined in ASTM E1618-10. Classifications were made using linear discriminant analysis, quadratic discriminant analysis (QDA), and soft independent modeling of class analogy (SIMCA). The model data sets iv were tested by cross-validation and used to classify fire debris samples. Correct classification rates were calculated for each data set. Classifier performance metrics were also calculated for the first step of the classification scheme which included false positive rates, true positive rates, and the precision of the method. The first step, which determines a sample to be positive or negative for ignitable liquid residue, is arguably the most important in the forensic application. Overall, the highest correct classification rates were achieved using QDA for the first step of the scheme and SIMCA for the remaining steps. In the first step of the classification scheme, correct classification rates of 95.3% and 89.2% were obtained using QDA to classify the crossvalidation test set and fire debris samples, respectively. For this step, the cross-validation test set resulted in a true positive rate of 96.2%, a false positive rate of 9.3%, and a precision of 98.2%. The fire debris data set had a true positive rate of 82.9%, a false positive rate of 1.3%, and a precision of 99.0%. Correct classifications rates of 100% were achieved for both data sets in the majority of the remaining steps which used SIMCA for classification. The lowest correct classification rate, 69.2%, was obtained for the fire debris samples in one of the final steps in the classification scheme. In this research, the first statistically valid error rates for fire debris analysis have been developed through cross-validation of large data sets. The fire debris analyst can use the automated method as a tool for detecting and classifying ignitable liquid residues in fire debris samples. The error rates reduce the subjectivity associated with the current methods and provide a level of confidence in sample classification that does not currently exist in forensic fire debris analysis.

This work presents a background on the chemical complexity of fire debris analysis, including an ever-present matrix of pyrolysis products as the catalyst that led to the creation of the National Center for Forensic Science's Fire Debris Database. A selection of these 1,000+ casework-relevant ground truth samples was used to create two newly proposed analyst workflows to connect the current method of categorical reporting with evaluative reporting practices reflective of the strength of the evidence. Both workflows use linear sequential unmasking to help mitigate bias, a discrete scoring system for quantification of the analysis, and receiver operating characteristic (ROC) curves to bridge together categorical and probabilistic reporting by indicating the optimum decision threshold the analysts are operating from when they make a decision. Both workflows also allow a machine-learning component to be included in evaluating the evidence and are practical methods for obtaining validated performances for human and machine decisions. The second workflow includes subjective logic, which provides a means of determining the uncertainty inherent to the opinion made by the analyst and the machine learning computational model. ‘Fuzzy categories' and an opinion triangle connect the opinion offered by the analyst given their perceived uncertainty to the ROC curve so a categorical decision can be made. For each workflow, three analysts independently assessed 20 randomly chosen samples from the Fire Debris Database and followed the ASTM E1618-19 standard fire debris analysis method. The resultant area under the ROC curve for each analyst for each workflow was 0.90 or higher, indicating that all were in the very good to excellent range for diagnostic classifiers, as was the machine learning model tested in the second workflow. Recommendations for implementing a performance validation workflow, how repetitive engagement can help the individual analyst and insights on using these for performance validation and training purposes are also included.