Interlaboratory Assessment of Micro-XRF Silicon Drift Detector Systems for Forensic Elemental Analysis of Electrical Tape Evidence

A recent comprehensive interlaboratory study evaluated the performance of modern micro-X-ray fluorescence (μ-XRF) silicon drift detector (SDD) systems for forensic discrimination of electrical tape evidence, comparing results from eight participating laboratories. The research addressed the critical need for standardized analytical protocols in forensic laboratories analyzing pressure-sensitive adhesive tapes, which are frequently encountered in criminal investigations including drug trafficking, kidnapping, and improvised explosive device construction. The study employed two novel polyvinyl (PVC) matrix-matched standards and ten sets of black electrical tape samples to assess analytical performance through spectral contrast angle ratio (SCAR) analysis and spectral overlay methods. Results demonstrate the capability of modern μ-XRF SDD systems to achieve reliable discrimination with improved sensitivity over traditional scanning electron microscopy with energy-dispersive X-ray spectroscopy (SEM-EDS) methods, while offering non-destructive, cost-effective analysis. The findings support the implementation of consensus-based comparison criteria and standardized data interpretation protocols for forensic tape evidence examination using μ-XRF technology.

Spectroscopy spoke to Lacey Leatherland, lead author of the paper resulting from the study,¹ about the work and the findings.

What fundamental differences in excitation and detection mechanisms explain why modern μ-XRF systems with silicon drift detectors often outperform SEM-EDS for elemental discrimination of electrical tape backings?

Micro-XRF SDD systems outperform SEM-EDS primarily due to differences in background noise, sensitivity to trace elements, and interaction volume. A major difference is the smaller spot size associated with µ-XRF systems compared to SEM-EDS, which allows more tightly focused excitation. Moreover, the excitation source differs, as µ-XRF utilizes an X-ray beam (photons), while SEM-EDS employs an electron gun (electrons). Photons are more energetic than electrons and penetrate deeper into the electrical tape backings, thereby enhancing the signal and allowing more representative sampling.

Another critical difference lies in the background radiation produced during excitation. SEM-EDS uses a high-energy electron beam. When these electrons strike the sample, they undergo deceleration within the atomic fields, producing a continuous background known as Bremsstrahlung. This background noise often masks weak signals from trace elements.

On the other hand, µ-XRF uses Photon Excitation. Because X-rays are photons (uncharged), they do not produce Bremsstrahlung upon interaction with the sample. The resulting spectrum has a significantly higher peak-to-background ratio, allowing the detection of trace elements at the ppm (parts per million) level, which is critical for distinguishing between tape batches that may have identical major elemental compositions (e.g., lead stabilizers or calcium fillers) but different trace profiles (Mo, Sb, Ba, etc.).

Together, these benefits enable µ-XRF systems to provide greater information than SEM-EDS.

How do silicon drift detectors (SDDs) improve sensitivity, precision, and throughput in μ-XRF compared to traditional Si(Li) detectors, particularly for forensic trace evidence?

The shift from Si(Li) to SDD is not just an upgrade in hardware; it is a shift in the physics of how charge is collected, leading to massive gains in speed and sensitivity. Because improved sensitivity offers better signal-to-noise ratio of the elements of interest, the precision is also improved, generating better quality data.

The larger anodes—where the electrical signal is collected—used in Si(Li) detectors generate more noise, which must be reduced with longer processing times. SDDs, on the other hand, have larger sensor areas, better energy resolution at higher count rates, along with faster analytical time. For example, in traditional Si(Li) detectors, the anode covers the entire back surface of the silicon crystal. This makes the capacitance proportional to the detector's size. High capacitance leads to high electronic noise. An SDD uses a series of concentric "drift rings" to create a field gradient that funnels electrons toward a microscopic anode at the center. Because the anode is tiny, the capacitance is extremely low, regardless of the detector's actual active area.

Moreover, SDDs are cooled via the Peltier effect, where heat is pulled from one side of the device to the other, which is more efficient than Si(Li) detectors, which must be cooled with liquid nitrogen.

Why does elemental analysis provide stronger discrimination power than molecular techniques like FTIR for PVC-based electrical tape backings?

From the methods used for the examination of electrical tapes, elemental analysis offers the most discrimination, particularly on the backing layer, which is rich in inorganic components. The elements of interest are added during the formulation to provide the final desired characteristics to the product. For instance, chlorine is present as part of the PVC polymer, while common elements in the fillers include aluminum, silicon, calcium, titanium, iron, zinc, lead, and barium. Other elements are incorporated as part of the metallic driers and catalysts. The multi-functionality of inorganic components in the tape backing creates elemental profiles that are quite homogeneous within a tape roll but vary substantially between manufacturers, or even between the same manufacturers but different product quality or different geographical locations of the manufacturing plant. For example, electrical tape backings produced in the USA can differ from the same product manufactured in Poland or Taiwan, and small changes in the formulation are detected by sensitive methods like µ-XRF.

What sources of inter-instrumental and inter-examiner variability are most critical in μ-XRF spectroscopy, and how can standardized protocols mitigate these effects?

Sample preparation, instrument configuration and/or parameters, and spectral interpretation all have an influence on inter-instrumental and inter-examiner variability, which is why it’s important that standardized protocols are produced and adopted by laboratories, as they have been shown to limit these types of variation. When followed, standard operating procedures ensure that analyses are conducted consistently, regardless of the instrument operator.

While inter-instrument variation has impacted analysts’ ability to share and compare data collected on different instruments, this research suggests that the use of a spectral similarity metric, such as SCAR, can be used to “normalize” data across instruments to facilitate data-sharing, as the metric takes into consideration intra-instrument variation.

What matrix-related challenges arise when analyzing polymeric materials such as PVC with μ-XRF, and how do matrix-matched standards improve quantitative and comparative reliability?

Micro-XRF signal is dependent of the matrix. A heavy or dense matrix (like steel or glass) most often absorbs primary X-rays and converts them into fluorescence. In a light matrix, such as a PVC polymer, X-rays are more likely to scatter and thus the signal becomes dependent of the excitation source-to-matrix interaction. The density of the material also influences the interaction volume. Because polymers are less dense compared to metals or glass, the X-ray beam penetrates much deeper.

Matrix-matched standards are calibration samples that mimic the physical and chemical properties of the unknown. Unlike glass, which is another evidence type commonly examined with µ-XRF, there is no widely used matrix-matched standard for PVC materials, which hinders our ability to accurately quantify the elemental composition of the sample. Recent studies by researchers at the University of Amsterdam have developed matrix-match standards that can open opportunities to move into the quantitative direction.

How does the Spectral Contrast Angle Ratio (SCAR) differ from traditional spectral overlay methods, and why is SCAR advantageous for improving objectivity and reproducibility in forensic spectroscopy?

The comparison of spectral data can be done in various ways, but the spectral overlay is by far the most common method used by forensic examiners in comparing XRF data. With concerns regarding subjectivity, the community is interested in finding more objective, quantitative metrics for spectral interpretation and comparison. SCAR is based on the spectral contrast angle (SCA), or cosine similarity equation; however, SCAR takes an additional step. It employs a ratio of the average angle between spectra of two samples and the average angle within spectra of the same sample, to represent the relative similarity of the compared spectra. Calculating the average and standard deviation of the intra- and inter-sample ratios establishes ranges for indistinguishable and distinguishable classifications, respectively. Thus, SCAR can be used to minimize bias and increase confidence in results. We recommend using SCAR as a continuous metric as an input to score likelihood ratios to describe the weight of the evidence and to evaluate rates of misleading evidence, thereby allowing the use of a scale rather than a binary decision threshold.

How do instrumental parameters such as spot size, vacuum conditions, and detector configuration influence limits of detection in μ-XRF analysis of trace elements in tape backings?

As discussed in the article, limits of detection are comparable between the instrument configurations examined in this study. However, we noted several trends regarding sensitivity and detection limits, which can be attributed in part to differences in spot size. For example, smaller spot sizes yielded better sensitivity over larger spot sizes, as smaller spot sizes offer increased fluorescent yield. The influence of vacuum conditions was not assessed in this study, as it was not a configurable parameter for one of the instrument models that was evaluated, but empirical data not published in this study suggests the vacuum influences SNR.

The spot size is related to the flux density. In µ-XRF, X-rays are often focused using polycapillary optics into a microscopic spot (typically 10–50 µm). A smaller spot concentrates the X-ray tube's power into a tiny area. For a discrete trace particle (like a small grain of a lead stabilizer in the tape backing), a small spot provides a much higher signal relative to the surrounding matrix, lowering the LOD for that specific feature.

The vacuum atmosphere inside the sample chamber also influences light-element detection (elements with low atomic number, such as Na, Mg, Al, Si, P, S, and Cl). Air can act as a barrier that absorbs the signal. Applying a vacuum removes this barrier, leading to increased sensitivity and expanded elemental range.

Also, the physical arrangement of the SDDs relative to the sample dictates how much of the emitted signal is actually captured. Many modern µ-XRF systems use Dual-SDD arrays. By doubling the detector area or moving it closer to the sample, the system captures more photons per second. This directly increases the sensitivity, lowering the LOD without needing to increase the X-ray tube power (which would also increase background noise). The take-off angle also matters. Most systems are designed with the detector at a 90° angle relative to the primary beam. This geometry is chosen to minimize Rayleigh scattering (the background noise).

Given comparable discrimination power between μ-XRF and LA-ICP-MS for electrical tapes, what practical considerations (e.g., destructiveness, cost, speed, evidentiary preservation) drive technique selection in forensic laboratories?

Several things drive technique selection in the forensic laboratory. Obviously, cost is a primary factor that impacts this decision, which comes into play with the initial instrument cost and the cost of consumables and maintenance for the instrument over time. In both cases, µ-XRF provides a cheaper alternative to LA-ICP-MS. Micro-XRF is also non-destructive, whereas LA-ICP-MS is minimally destructive. This is important because we often deal with incredibly small samples where the laser ablation process can completely consume the sample.

Why is it important to evaluate both correct association and correct exclusion rates in spectroscopic interlaboratory studies, and how do false positive and false negative rates impact forensic conclusions?

Evaluating both correct association and discrimination power is important for these studies. First, we want to evaluate the correct association rate to ensure that the method can accurately associate samples from a common source. This metric allows us to assess the homogeneity within a single sample, which is important because if a sample were heterogenous, it would render any exclusion decision inconsequential. We also evaluate the discrimination power of a method to assess how well we can differentiate samples from distinct sources. Balancing these two rates is especially important in forensic science due to their impact on false exclusion and inclusion rates. In the case of the former, incorrectly excluding evidence could lead to a truly guilty suspect going free. On the other hand, techniques with large false inclusion rates, where dissimilar evidence is incorrectly associated, could lead to a genuinely innocent person being incriminated during an investigation.

What role does large interlaboratory studies play in the acceptance of spectroscopic techniques like μ-XRF in forensic science, and how can quantitative metrics such as SCAR support admissibility in court?

Conducting large interlaboratory studies is a vital step towards acceptance when new techniques are introduced to the forensic science community. These studies allow us to assess the reproducibility of methods and how results are influenced by different instrument configurations. This allows us to create truly consensus-based protocols and criteria that fit the need of the community.

Quantitative spectral similarity metrics, such as SCAR, can act as supporting evidence to increase analysts’ confidence when they are conducting spectral comparisons. The use of these metrics to support spectral overlay conclusions can make these decisions less subjective, improving their standing in the eyes of the court.

References

1. Leatherland, L.; Andrews, Z.; Neumann, C. et al. Assessing the Expanded Capacity of Modern μ-XRF SDD Systems for Forensic Analysis Through an Interlaboratory Study: Part I—Electrical Tapes. Forensic Chem. 2026, 47, 100719. DOI: 10.1016/j.forc.2025.100719