DEVELOPMENT AND VALIDATION OF A TECHNIQUE FOR ANALYSIS OF TONER PRINTED DOCUMENTS USING MAGNETIC FLUX MEASUREMENTS
Toner printed documents are frequently submitted for forensic examination, yet there is no standard methodology for examination of the documents, and laboratories are recommended to implement ‘complementary analyses’ using multiple different types of instrumentation and analytical techniques which provide categorical information about the production of a document until they have enough information to make a determination about its authenticity, for example. Laboratories often face difficulties in implementing the ‘complementary analyses’ approach, because it requires access to multiple types of costly instrumentation, lengthy analyses and data interpretation, and the results are often inconclusive due to the categorical nature of the data provided by most techniques. It has been acknowledged that there is a need for the development of a method for toner analysis that is non-destructive, cost-efficient, and provides quantitative data that can easily be interpreted using statistical methods. The purpose of this dissertation was to develop a method for toner analysis that fulfilled this need using magnetic flux measurements. To achieve this aim, the project objectives were to 1) assess the printing and device factors which affect the repeatability of measurements; 2) determine the critical method and instrument parameters which affect measurement uncertainty as well as to develop a method protocol which considers those parameters and could be implemented in a forensic laboratory setting; 3) conduct a method validation and apply the refined method protocol to a representative population sample of toners.
First, an initial assessment of the magneto-optical imaging technology was conducted which assessed the repeatability and reproducibility of magnetic flux measurements of toners, as well as sample stability. Stability was assessed by measuring the magnetic properties of documents produced by five electrophotographic devices which had been analyzed one year prior. A representative population of 150 toner samples was then assessed using the magnetic flux as a function of toner area as a normalizing unit to determine if significant variation could be observed. The results indicated that magnetic flux does remain stable over time and that there is a relationship between toner area and magnetic flux. The relationship between toner area and magnetic flux allowed for the normalization of data and assisted in comparison of samples with variable text and print features, which is necessary for casework application. The variation present in the population distribution was large enough to allow for discrimination of most samples based on magnetic flux measurement. After the initial assessment, method development was completed, which included assessing the instrument and method parameters which affect measurement uncertainty and developing a method protocol which considers them. At this stage, multiple analyses were conducted to analyze the impact of the variables which had been identified as potential contributors to the measurement uncertainty. One study assessed the impact of changing toner area within the field of the sensor and controlling the grey value variation during the area determination process using seven samples from different electrophotographic devices. This study found that the toner area could vary without significantly impacting measurement uncertainty as long as contacts with the sensor periphery were minimized, and that limiting the variation of the mean grey value to ± 0.5
pixels was effective in reducing measurement uncertainty. The next study assessed the impacts of hysteresis effects and biasing induction current spatial effects on measurement uncertainty, using five samples from different devices. This study found that both induction current spatial effects and hysteresis effects significantly increased intra sample variation, but that hysteresis effects were reciprocal in different planes of rotation and their impacts can be mitigated using a sampling methodology which conducts measurements from four different orientations. Induction current spatial effects could not be completely mitigated via sampling methodology, but there was a relationship between the percent of toner surface area incident to the induction current and measurement enhancement, so this relationship could be exploited to reduce the impacts of biasing induction current spatial effects, and potentially a method could be developed with further research to correct for the signal enhancement. The last stage of method development involved assessing factors which impacted the reproducibility of measurements via an inter-laboratory study. A method protocol was designed which controlled the variables known to impact measurement uncertainty, and this method protocol was issued together with three toner samples to laboratories which were recruited to participate in the study. Laboratory conditions including the instrument version, software update version, operator, and laboratory setup were found to have a significant impact on the reproducibility of results, though the implementation of a standard method did reduce the levels of measurement uncertainty from those observed in the initial assessment, where the operator was the only reproducibility variable assessed. Though measurements cannot be compared between laboratories at this time, it is possible that the development and deployment of a known reference material for use with
the instrument would further reduce measurement uncertainty and allow for meaningful inter-laboratory comparisons. Finally, the developed method was validated using a published forensic validation guideline in the field of forensic toxicology. The positive quality control sample provided by the instrument manufacturer was used to assess the bias and precision of measurements, and the potential for matrix effects and interferences were also assessed using five matrix samples from different sources and a collection of 25 different media types that may be encountered on a mixed media document. The bias and precision were found to be within acceptable levels, and no matrix effects or interferents were identified. After validation, the method was applied to a representative population sample. The distribution of measurements in the population sample was similar to the distribution in the initial assessment study, and it was determined that toner populations consistently exhibited variation which provides the potential to discriminate between samples of different origin. Magnetic flux measurements of toners were found to provide the potential for a robust, non-destructive, cost-efficient method for toner examination which provides continuous quantitative data which can be easily interpreted using commonly employed statistical testing methodologies. Although there is still a need for the development of a known reference material for use with the instrument to reduce measurement uncertainty to a level which would allow for inter-laboratory data comparisons, magnetic flux measurement methods can be validated to forensic laboratory standards, allowing practitioners to deploy another analytical tool for toner printed documents analysis.