Green Stabilizers in Nitrocellulose Propellants: VOC Fingerprinting of Accelerated Aging by HS-SPME–GC×GC–TOF-MS

Propellant powders store large amounts of chemical energy and require stabilizers to slow degradation of energetic nitrate-ester groups. Traditional stabilizers, such as diphenylamine, akardite II, and ethyl centralite, raise safety and toxicity concerns, so green alternatives such as α-ionone, α-tocopherol, and hydroxyl-terminated polybutadiene (HTPB), are being adopted.

A research team comprised of members of the University of Liège (Belgium) and the College of William & Mary (Williamsburg, Virginia) developed a robust analytical workflow to profile volatile organic compounds (VOCs) emitted during accelerated (STANAG 4582) thermal aging of smokeless powders formulated with those green stabilizers. Headspace solid-phase microextraction coupled with comprehensive two-dimensional gas chromatography–time-of-flight mass spectrometry (HS-SPME–GC×GC–TOFMS) provided high peak capacity and sensitive VOC detection. Multivariate analysis (principal component analysis) was used to track compositional changes in the headspace over aging and to identify candidate degradation markers.

The results demonstrated that distinct VOC fingerprints that evolve with thermal aging and differ by stabilizer type, enabling discrimination of degradation stages and identification of marker species linked to stabilizer degradation pathways. The workflow offers a comprehensive, high-resolution approach for routine quality control of nitrate-ester propellants containing green stabilizers and supports wider adoption of safer stabilizer chemistries by providing a means to monitor long-term stability and decomposition under controlled conditions.

LCGC International spoke to Pierre-Hugues Stefanuto, lead scientist and senior lecturer at the University of Liège, and lead author of the paper resulting from this research (1), about this workflow and its potential implications for future research.

What motivated the shift from classical stabilizers (such as diphenylamine) to green alternatives like α‑ionone, α‑tocopherol, and HTPB in propellant formulations?

The main driver is the change in the regulation for chemical in the European Union (EU), but also in other countries. The European Chemicals Agency regulations have recently begun encouraging greener chemistry across all sectors. Armament manufacturers are very interested in moving forward by replacing propellant stabilizers with natural products, which aligns with the green chemistry principle of designing safer chemicals.

There is a real exposition hazard for people expose during the production and the utilization of different type of propellants. Most of the current stabilizing option are potential carcinogenic compounds.

How do the chemical properties and volatilities of α‑ionone, α‑tocopherol, and HTPB affect their behavior and degradation in nitrate‑ester propellants?

Nitro-based propellants are highly energetic and unstable during aging. Exposition to light and heat can quickly deteriorate the propellants making dangerous to use or inefficient. There is a strict regulation, the STANAG 4582 (NATO Standardization Agreement), about aging simulation and evaluation.

α‑ionone, α‑tocopherol, and HTPB are all used in other field of chemistry as green stabilizers. For example, their performances in plastic have been surpassing the ones from other synthetic stabilizers. These three compounds have the capacity to undergo oxidation protecting the main material, their by-products are armless chemicals.

α‑ionone smells like violet, and it is present in numerous flower-based essential oils. α‑tocopherol is usually called vitamin E, it is present in a large variety of food in the European diet. The two compounds and their degradation products are volatile, making them highly suitable for headspace monitoring, decreasing the need of complex sample preparation for dangerous material monitoring.

Hydroxyl-terminated polybutadiene (HTPB), a rubber-like material, is non-volatile as for its by-product. The headspace analysis is less efficient, even if small changes can be identified.

How did you design the accelerated aging protocol (including reference to STANAG 4582), and how well do you think thermal aging models actual field aging of propellant powders?

This is a good question. Aging evaluation is always a model of what is happening in the field the STANAG 4582 has been developed and tested on classical propellants with a high success. Here, we are monitoring the lifetime of the stabilizer, but other physico-chemical evaluation needs to be performed to guarantee the stability of the product properties.

What key patterns did principal component analysis (PCA) reveal about early versus late‑stage aging and how could these patterns be applied in routine quality control?

Samples of less than five years had profiles that indicated the release of VOCs from the stabilizer additive itself. Samples of more than five years clustered in the center of the PCA, demonstrating a stable emission of VOCs from common components of the smokeless powders. The differences likely originated from the volatility differences of the three green stabilizers, and PCA could be a useful tool to visualize the stabilizer aging process.

Here, we are using PCA for its true essence, which is the display of the variance comprise in a high dimensional data set. PCA is a powerful visualization tools to follow up trends. I hope it will be more accepted and implemented for routine workflow in the future.

How does detecting these VOCs complement solid‑phase analyses, and what does that mean for understanding stabilization mechanisms?

Identifying the same degradation products over two extraction methods is a confirmation that the by-products are formed during aging and not during the extraction steps. As mentioned before, these stabilizers are found in other materials. We hope that demonstrating the capacity of headspace monitoring here will promote the use of solvent-free methods on other materials.

What data‑processing workflow (filtering, alignment, significance thresholds) do you recommend for robust marker discovery in GC×GC datasets?

Don’t shoot a fly with a bazooka. The simplest workflow is always the best. In the present study, all the data were acquired in a short period, so the retention time shifts were minimal. In addition, the chromatograms were not too complex making the peak picking easier. The main goal of the alignment, here, was to generate a usable global peak table and guarantee reliable peak annotation across the different runs. From there, we used basic pre-processing by normalizing on the solid-phase microextraction (SPME) peak in combination with auto-scaling. Next, we used t-test and PCA to identify and visualize the different chemical profiles and their evolutions. T-test and PCA are our starting tools in all our non-targeted studies. If the analytical steps are well controlled, they usually perform well and provide highly interpretable results. It is sometime tempting to use more advanced tools, such as machine learning. Here, they were not required to answer our research question.

Which degradation markers or analyte classes would you prioritize for translation to routine 1D GC methods and why?

In the present case, we are convinced that 1D GC-FID could provide the required performances for routine monitoring. Here, we applied a non-targeted approach since we did not know what to expect from the analysis. GC×GC is the best tool for discovery, and it can perform well in routine. Nevertheless, as I say before, don’t shoot a fly with a bazooka. If you can, always make it simpler.

References

1. Damseaux, C.; Scholl, G.; Perrault Uptmor, K. A. et al. Analysis of Volatile Degradation Products from Green Stabilizers in Smokeless Propellants with Comprehensive Two-Dimensional Gas Chromatography. Sep. Sci. Plus 2025, 8, e70130. DOI: 10.1002/sscp.70130