I. INTRODUCTION

Chemical warfare agents are highly toxic and dangerous substances, especially nerve agents such as Sarin and Soman ^[1]. When these extremely volatile chemical compounds enter the body through inhalation or skin contact, they can lead to the paralysis of neuromuscular transmissions within minutes, potentially resulting in death ^[2]. It is crucial to develop effective means to detect the use of warfare chemical agents to protect human life and ensure national security.

In recent years, the need for detection and monitoring systems for toxic gases has increased. For this purpose, various approaches have been developed and improved, including detection techniques based on surface acoustic wave devices ^[3], gas chromatography ^[4], and fluorometry ^[5].

Despite different approaches, the methods available so far have not been able to provide effective solutions for the rapid and real-time detection of nerve agents. This is largely due to the lack of portability, complexity of operation, high cost, and slow processing of existing systems ^[6].

Detection systems for these agents must meet the requirements of real-world applications with good sensitivity, ease of operation, and real-time response.

Dimethyl methyl phosphonate (DMMP) is often used as a simulation or surrogate for sarin gas in laboratory research, as it exhibits lower toxicity while having similar chemical properties. It is worth noting that the main form of contamination by these agents is through the air, making it more advantageous for detection systems to be more sensitive to gases than to liquids or fluids ^[7]. The DMMP palladium sensor research has focused on utilizing phthalocyanine-palladium thin film bilayer structures; these sensors exhibit promising electrical response behavior at room temperature ^[8]. Phthalocyanine-palladium thin film bilayer structures exhibit an electrical response when exposed to DMMP vapor. This makes them promise as sensors for detecting DMMP and similar organophosphorus compounds. The resistance of the phthalocyanine-palladium structure changes in response to different DMMP concentrations, allowing quantitative detection. The sensing mechanism is attributed to the strong interaction and binding between DMMP and the palladium layer in the bilayer structure ^[9]. Additionally, some researchers have explored the use of single-walled carbon nanotube networks for real-time detection of hazardous DMMP gas, showcasing the versatility of sensor materials ^[10].

To develop detection systems and sensors for chemical warfare agents, studies are being conducted to understand the complex reactions that occur during the adsorption of DMMP vapors on the surface of metal oxides ^[11].

Palladium-decorated carbon nanotubes (CNTs) show enhanced sensitivity and recovery compared to gold-CNT sensors, due to stronger interactions between DMMP and the palladium surface. Palladium-iron metal-organic frameworks (Pd-MOFs) are highly active and selective catalysts for reactions like Suzuki-Miyaura coupling and are more stable than other Pd-based catalysts. ^[12].

Palladium-based materials can effectively detect and quantify DMMP through strong interactions and binding between DMMP and the palladium component. This makes them promising candidates for chemical sensors targeting nerve agent simulants.

Microwave transmission line sensors have been developed for detecting different gases. These sensors utilize the near-field transmission-line technique to detect changes in the dielectric properties of the sensing material, allowing for the detection of various gases ^[13].

The total attenuation of the wave along the Coplanar Waveguide (CPW) is significantly affected by conductor losses, which are caused by electrical resistance. Variation in the geometry of the conductor, such as the width and thickness of the trace, influences conductor losses ^[14].

The S21 parameter directly quantifies the transmission loss or gain of the network, while the real part of Y11 indirectly reflects the transmission line loss through the power dissipation within the network. Both parameters provide complementary information about the performance of the transmission line ^[15].

This paper proposes the use of a palladium-based CPW microwave sensor as an alternative for detecting the concentration of DMMP vapor in air. For this application, a CPW transmission line was selected due to its simplicity. A layer of palladium was used as the active sensing area. The interaction between palladium and DMMP leads to changes in the conductive properties of the sensor, which can be detected using a vector network analyzer (VNA).

II. DESIGN AND FABRICATION

1. Design

CPWs are structures with a specific design consisting of a conductor line of width w and thickness t printed on a substrate with relative permittivity ε_r and height h. In addition, on both sides of the conductor line, other return conductors or ground planes are present, separated by a distance G. It is worth noting that the three conductors are in the same plane, which is a fundamental characteristic of these devices, coplanarity ^[16].

The fact that the conductors are in the same plane allows the electric and magnetic fields to have less confinement inside the substrate than if they were in opposite planes, e.g., a microstrip line ^[17], resulting in greater sensitivity to external elements to sense ^[18]. Such sensitivity is an important aspect in the application of CPW sensors, since variations in the dielectric characteristics of the material to be analyzed can significantly affect the distribution of electromagnetic fields within the structure, resulting in a change in the sensor response ^[19].

CPW structures have been widely used as sensing devices in various applications. For example, in ^[20], a CPW sensor with an interdigital capacitor resonator operating between 5 and 7 GHz was proposed for the characterization of chemical solvents. The sensor showed a shift in its resonance frequency of about 24 MHz for every 0.1 variation in the electric permittivity of the sample. In addition, the authors of ^[21] developed a planar CPW sensor for detecting adulteration in oil samples through perturbation in the S21 response of the system. In ^[22], a CPW sensor coated with a metal oxide composite film was successfully employed for sensing ambient humidity.

In this paper, we proposed a single palladium layer CPW microwave sensor for sensing DMMP gas over a fused quartz substrate. Fused quartz has been used in the literature as a substrate for gas sensors ^[23]. In ^[24], a hydrogen gas sensor based on zinc oxide was tested for two different substrates, quartz and PET, and it was found that quartz yielded good results due to its physical properties (low loss at microwave frequencies, operates at room temperature, good adhesion to metal deposition, and rigidity). During fabrication, seed layers of titanium (Ti) and gold (Au) were used to prevent peeling off of the palladium layer.

The dimensions employed in the sensor design were determined to have a 50-ohm impedance transmission line, with a size suitable for testing under controlled conditions. Fig. 1 shows the CPW sensor dimensions. In this study, the CPW gap was set to 0.08 mm, determined primarily by the mechanical dimensions of the test fixture—specifically, the width of the central pin and the necessary clearance to the adjacent ground planes to prevent shorting and ensure reliable RF contact. The length of the CPW line was selected to provide mechanical stability during repeated insertion into the test fixture, allowing consistent alignment and secure pin engagement.

Fig. 1. Schematic view of the CPW: (a) top view and (b) cross-section view.

2. Fabrication

The process consisted of depositing a thin palladium film by e-beam evaporation over a quartz substrate using a spin-coated mold made of photoresist and its removal using the lift-off technique. The photoresist spin coating process ^[25] and subsequent lift-off after metal deposition is an extensively employed procedure for defining the shape of thin metal films onto uniform and flat surfaces. A seed layers of Ti and Au were used to prevent peeling off of the palladium layer. Fig. 2 shows the fabricated sensor.

Fig. 2. Fabricated CPW microwave sensor.

III. MEASUREMENT RESULTS

Fig. 3 shows the measurement environment setup. The measurements were performed at the chemical defense area of the National Institute for Aerospace Technology (INTA) in Madrid, Spain. The DMMP vapor concentration provided was set at 400 parts per million (ppm) for this experiment.

Measurements involving the highly toxic chemical compound DMMP require a highly controlled experimental environment. The experimental arrangement used for measurements consisted of the CPW sensor connected to an Agilent E8361A vector network analyzer (measuring S-parameters from 0 to 15 GHz) through an Anritsu 3680V universal test device. In addition, the setup was equipped with a small tube designed to provide DMMP gas coming from a gas generator directly to the sensor. The strategic position of this tube near the top of the sensor allowed precise analysis of the interaction between the DMMP vapor and sensor.

Fig. 3. Experimental setup for DMMP vapor detection performed at INTA.

The S21 response values are obtained for the sensor exposed to 400 ppm of DMMP vapor in air. In Fig. 4 shows the S21 responses for different frequency ranges and the variation of the insertion loss at the presence of DMMP. This variation was expressed in percentage of changes.

Fig. 4. DMMP Measurements results of S21.

Fig. 5 presents the frequency-dependent attenuation constants (α) of the CPW sensor under air and 400 ppm DMMP exposure conditions, calculated from the ABCD parameters of the transmission line. Distinct variations in α are observed across the frequency range, with several pronounced peaks in the GHz band showing noticeable shifts between the two environments. These spectral differences suggest that DMMP adsorption on the Pd surface alters the local surface impedance, thereby modulating the RF signal attenuation. The appearance of frequency-specific peaks further indicates that the sensor is capable of spectrally resolved detection, which may enhance its chemical selectivity. These results demonstrate that the proposed CPW sensor not only quantifies overall transmission line loss but also responds sensitively to chemical surface interactions, validating its potential for high-frequency, real-time gas sensing applications.

Fig. 5. Calculated attenuation constant of the CPW sensor. (a) Attenuation constant of the CPW sensor in air and after DMMP exposure. (b) Difference in attenuation constant (Δα).

The S21 parameter directly quantifies the transmission loss or gain of the network, while the real part of Y11 indirectly reflects the transmission line loss through the power dissipation within the network. Both parameters provide complementary information about the performance of the transmission line. The real part of the Y11 parameter (input admittance) of a transmission line exhibits resonant peaks at specific frequencies. This occurs due to the standing wave pattern formed by the reflected waves at the input of the transmission line. At resonant frequencies, the input impedance of the transmission line is purely real (resistive), resulting in a peak in the real part of the Y11 parameter. The resonant frequencies depend on the length of the transmission line and the termination impedance at the other end ^[15]. The real part of the Y11 parameter is a measure of the power dissipation within the transmission line, which directly contributes to the overall transmission line loss. A higher real part of Y11 indicates greater conductance, which may correspond to increased power dissipation in the transmission line. In our experiment, we believe that conductance losses are more dominant than dielectric losses when DMMP is exposed to the CPW. Consequently, DMMP exposure leads to higher transmission losses. Fig. 6 shows the results of Y11 calculated from the measurements of the S parameters.

Fig. 6. Results of Y11 obtained from the DMMP measurements of the S parameters. (a) Magnitude, (b) real part, (c) imaginary part.

To further clarify the sensor’s response to DMMP exposure, Fig. 7 presents zoomed-in views of the measured real part of Y11 at selected resonant frequencies. A noticeable difference is observed between the 0 ppm and 400 ppm DMMP conditions, particularly at 2.4, 4.9, 7.3, 9.7, and 12.1 GHz. These changes in the real part of Y11 are attributed to changes in the surface conductivity of the Pd layer upon adsorption of DMMP molecules. As the real part of the Y11 is directly influenced by the electrode’s surface conductivity, the results strongly support the hypothesis that Pd–DMMP interactions modulate the electrical characteristics of the sensor. This observation validates the proposed high-frequency sensing mechanism.

Fig. 7. Results of Y11 obtained from the DMMP measurements of the S parameters. (a) Magnitude, (b) real part, (c) imaginary part.

IV. CONCLUSIONS

This study introduces a novel palladium-based Coplanar Waveguide (CPW) sensor for detecting DMMP vapor at microwave frequencies. The sensor successfully detected DMMP at a concentration of 400 ppm by measuring changes in electrical conductivity. The straightforward design demonstrates potential for chemical defense and warfare agent detection. Future research should focus on improving sensitivity, exploring lower detection thresholds, and evaluating performance across varied environmental conditions.