Safe Near-Infrared Spectroscopic Identification of Chemical Warfare Agents Using a 3D-Printed Glass Liquid Cell

Rapid and reliable on-site identification of chemical warfare agents (CWAs) is critical due to their extreme toxicity and continued misuse. A research team from the Netherlands explored near-infrared spectroscopy (NIR) spectroscopy as a fast, noninvasive, and portable alternative for CWA detection, addressing limitations of existing on-site analytical methods. A sealable liquid cell fabricated from 3D-printed glass and incorporating a polytetrafluoroethylene (PTFE) reflector was developed to safely analyze highly toxic liquid agents without sample alteration or exposure. Optimized for short optical path length, the device enables reproducible NIR measurements of nerve and blister agents using a portable spectrometer, with results confirmed by benchtop NIR and supported by density functional theory calculations. Spectroscopy spoke to Jelle C. de Koning, corresponding author of the paper resulting from this work (1).

Can you explain the primary challenges associated with the rapid detection of chemical warfare agents (CWAs) in the field?

There are multiple challenges in the rapid detection of CWAs in the field. The first is the chemical properties of the compounds. Some are volatile and can easily be detected above the surface using techniques like ion mobility spectrometry (IMS). Others display a low volatility and can not easily be detected above the surface and therefore require another approach. A second challenge is the extreme toxicity of these compounds, making both rapid identification in case of suspected use and avoiding contact with the sample very important. Contactless detection is not possible for every technique and comes with practical challenges, especially combined with earlier mentioned limitations. Many detectors are commercially available but in addition to the practical limitations, there is often a trade off in sensitivity and selectivity. This makes that there is no technique available that can do everything needed and investigation into new techniques is needed.

What are the advantages and limitations of near-infrared (NIR) spectroscopy compared to other portable spectroscopic techniques, such as Raman or Fourier transform-infrared (FT-IR), for CWA detection?

NIR spectroscopy has several practical advantages. While some methods can take up to minutes to complete the analysis, NIR measurements are very rapid and typically only take seconds to complete. Because of technological developments, NIR has been miniaturized to a MEMS sensor that is smaller than 3 centimeters in all directions. This means that NIR can be used in very small portable devices, allowing for on-site analysis. In addition, the ability to miniaturize this technique makes that a lot of commercial detectors are available and that this technique is relatively low cost compared to other portable detection devices. NIR itself allows for very reproducible measurements and has the advantage of being able to measure through certain containers of glass or plastic, allowing for analysis of materials without having to open them, which is important when dealing with highly toxic chemicals. Sample integrity is also maintained as NIR only emits small amounts of energy to the samples and is therefore considered to be non-destructive. A disadvantage of NIR, like many other portable detectors, is the limited sensitivity, making it mostly suitable for bulk analysis.

How does sample form (solid vs. liquid) affect NIR spectroscopy measurements, and what strategies are employed to analyze highly toxic liquid CWAs safely?

The differences in both sample forms is in the way light is reflected. Solid samples generally allow for direct diffuse reflectance measurements, where light is reflected from the solid material to the detector. In liquids, not much light is reflected from the detector, and the detector thinks a lot of light has been absorbed by the sample. This is not a problem in a set-up suitable for transmittance measurements, but most NIR detectors are set up for diffuse reflectance measurements. To circumvent this, we needed a way to reflect light back after passing through the liquid. This was done by adding reflective material behind the liquid, where the light passes through the liquid, reflects on the material, passes through the liquid again and reaches the detector. Such reflection measurements are usually done for liquids in an open environment. To ensure safety while doing the measurements of extremely dangerous chemicals, we packed the space for the sample and the reflective material into a sealable glass container, so after sampling the operator does not have to meet the harmful agent.

There are also differences in solid and liquid materials in terms of the results. Liquids tend to be more homogeneously distributed and do not show scattering effects like solids do. Therefor measurements show even more overlap in liquids.

Describe the design and functional advantages of the 3D-printed glass liquid cell used for NIR analysis of CWAs.

The design was created according to a few requirements needed for safe use in the field. First, the obtained NIR signal needed to be of good quality, we found that this was the best when the reflecting material directly traps the liquid against the glass. Second, the design needed to be safe and sealable, keeping the liquids inside and avoiding any spills of materials. The conical design allows for easy insertion of the insert while avoiding too much air displacement within the cell that can lead to spills of liquid. The design also needed to be repeatable in production, to allow comparison of measurements from different cells. To achieve this, 3D printing of glass was used, which has a high reproducibility in the final product. This was done by an expert at the University of Amsterdam workshop.

The liquid cells also needed to be affordable. With the chemicals we are investigating, opening and cleaning of the cells after exposure to CWAs comes with extra risks, and is preferably not done. The final design was therefore developed as a product that is cheap enough to be used as a disposable device. Finally, the cell needed to be from materials that are relatively inert, as it is important to maintain sample integrity. This was achieved by looking at different materials and assessed by investigating the stability of different samples over time. The final design provides a high quality NIR signal and is also usable as a storage container, in which samples can be safely transported.

Why was polytetrafluoroethylene (PTFE) chosen as the reflective material in the liquid cell design, and what properties make it suitable for this application?

PTFE has several advantages that made it a suitable candidate for use in the liquid cell. The material is known to have a high chemical inertness; this is important as we are dealing with highly reactive chemicals and we want to maintain the sample inside the liquid cell as it is. PTFE material can be easily machined into different shapes, which provided us the possibility to design it into the cell in any shape needed. Both the insert and the spacer were made from PTFE material, which resulted in proper sealing of the cell and provides us flexibility in the use of the cell. The material also reflects NIR light, based on the properties of the material. While materials were used that showed a better reflectance, like aluminum, these has undesirable properties like potential reactions with the CWAs, resulting in sample degradation. PTFE ended up being the most practical option while still providing good quality NIR spectra for the chemicals of interest.

How does the path length in transmission NIR measurements influence sensitivity and spectral detail for CWA analysis?

A higher path length leads to more sensitivity in the spectrum, which can be used to unveil more details in some regions of the spectrum. The first region of the spectrum (before 1900) could benefit from more sensitivity. The later region of the spectrum shows more signal using the current set-up. There is a certain trade-off in using larger path lengths, as using this detector we observed that increasing the path length can also lead to saturation effects. Saturation effects are caused by a too high absorption which results in a loss of spectral detail. This was already observed in the later region of the spectrum for some samples analyzed in this work. Different options in path length were tested in the experimental stage of this work. The optimal path length for a compound can also be calculated but is more difficult to determine for a group of compounds with varying properties. For this reason, 0.5 mm was used as a compromise between the highest possible sensitivity and the spectral detail obtained.

Discuss how NIR spectroscopy data can be validated using both portable and benchtop instruments, and the implications of observed differences in spectral resolution.

Portable instrumentation has different requirements in both practical use and spectral detail in the results. The portable device, the Puck, has a smaller wavelength range (1350-2550), measures 257 data points over this range and has a resolution of 16 nm (FWHM). The benchtop instrument has a higher resolution, 10 nm (FWHM), a broader range (350-2500) and collects 2151 data points over this range. In previous research using both of this instruments, good agreement was observed between these two instruments. As nothing has previously been published on CWAs, the comparison between the two instruments was made to see if any detail would be visible in the results of the high-resolution instrument compared to the portable device. It can be possible that more details are visible, or that features become visible at a higher resolution that are not visible on the lower resolution portable device. With less data points and a lower resolution, it can well be that features merge in the spectrum and become (partially) invisible that can be resolved in the higher resolution systems. The added detail can be used to give more confidence in distinguishing closely related chemicals. In this study, some extra detail was observed in in the spectra of benchtop instrument; overall, the results were in good agreement.

What role does density functional theory (DFT) calculations play in supporting NIR-based identification of CWAs, and what are the current limitations of this approach?

As this is the first time spectra of CWAs are published, there was no material or database to compare the spectra obtained in this study to. Other ways to confirmation our results obtained with the portable NIR device were sought. The benchtop NIR was used for this, and theoretical (DFT) calculations were done as previously published work showed consensus between spectra and calculated spectra to obtain an extra level of certainty. The investigation into the DFT calculations were done in collaboration with experts of the FELIX group of the Radboud University.

Using DFT, the spectrum can be relatively well reproduced. It was found from the calculations of the different molecules that there are two limitations at this stage with the used method. While the position of the features could be accurately calculated, calculating the intensity of the feature is computationally intensive. Either a lot of time and computer power should be reserved, or it should be accepted that a difference in intensity of the spectrum is observed. The other limitation we faced is the calculation of compounds that can form intermolecular hydrogen bonds. Which gives rise to a broad, elevated signal in the NIR spectrum which is not observed in the DFT calculations.

In practical scenarios, what considerations must be made for using the 3D-printed liquid cell under varying environmental conditions, such as temperature or humidity?

All the experiments were done under controlled conditions in our laboratory. Both the detector and the liquid cell were not exposed to more extreme temperature or humidity conditions. These could potentially influence both the functioning of the detector and the functioning of the liquid cell. The leak tightness of the device can also be affected when dealing with different materials (glass, PTFE) under the influence of temperature, as heat expansion and cold contraction effects will be different for the two materials. It should be investigated if any effects occur at changing conditions that can lead to both chemicals leaking out of the cell and potentially also chemicals, like water at a high humidity, enteringthe cell and leading to sample degradation. Large amounts of water in or outside of the cell can lead to a disturbance in NIR signal, as water shows a strong NIR signal.

How could the methodology develop for intact liquid CWAs be adapted for more complex matrices, such as soil or mixtures, to enhance real-world applicability of NIR detection?

To ensure correct results when applying this method in the real world, different circumstances that may be faced should be investigated. First, mixtures of chemicals should be investigated. As can be seen from the stability measurements conducted in this study, some of these chemicals are prone to degradation and result in a mixed spectrum of two compounds, which must be unraveled to identify the chemical using multivariate data analysis. Databases of either mixture of pure compounds must be established that can be found in the field to feed these data analysis approaches.

To be able to identify these chemicals on backgrounds, different backgrounds must be investigated to test for any compounds or signals interfering with the NIR analysis. In addition, limits of the methodology must be investigated by adding small amounts of these chemicals into the background to see what levels can still be accurately determined. This includes the investigation into false positive and false negative results. The process of identifying these chemicals in complex backgrounds heavily relies on data analysis and is currently under investigation.

References

1. de Koning, J. C.; van der Schans, M. J.; Chau, L. F. et al. Rapid On-Site NIR Spectroscopic Characterization of CWA Liquids Using a Novel 3D-Printed Glass Cell. Anal. Chem. 2025, 97 (38), 20973-20981.  DOI: 10.1021/acs.analchem.5c03719