Development of Thermal Flash Heating/Rapid Cooling System Using a Heated Needle Coupled with an Alternating Current (AC) Corona Discharge Ion Source

Abstract

In previous work, a flash desorption/rapid-cooling system for the analysis of nonvolatile samples was developed using a heated needle. The needle was moved down to the sample surface using a linear actuator. In this system, it was necessary to adjust the lowest position of the needle manually. In the present work, a flash-heating system that requires no manual adjustment was implemented and combined with an alternating current (AC) corona discharge ion source. By using a movable needle inserted into a heated ceramic tube, the needle stops and remains on the sample surface at the moment the needle tip touches it, thereby preventing deep penetration. After contact for 200 ms, the needle was lifted up, enabling rapid cooling of the heated sample. Thus, the movable needle in the ceramic tube acts as a soft-touch sensor. This system, coupled with an AC corona discharge ion source, was successfully applied to the analysis of industrial polymers.

INTRODUCTION

Structural information on the main chains and additives of polymer products is essential for understanding their properties, quality, and performance. In polymer analysis, mass spectrometry is particularly useful because it provides mass information on repeating units derived from polymer backbones as well as on additives, offering high identification capability.¹⁾ For the analysis of nonvolatile polymers by mass spectrometry, a thermal-desorption process is required to generate characteristic degradation products while minimizing secondary reactions. Thermal desorption and pyrolytic approaches are useful because they provide information on repeating units, terminal groups, and additives. However, when the sample is heated too slowly, secondary thermal decomposition occurs, and the resulting mass spectra become overly complex.¹⁾ In addition, if the heated sample is not cooled rapidly, further decomposition of both the sample and the desorbed molecules continues, reducing spectral clarity and analytical throughput. Therefore, a technique enabling both rapid heating and rapid cooling is required for practical and reproducible polymer analysis.

To address this issue, we previously developed a thermal flash-heating/rapid-cooling desorption method.^2,3) In this method, a heated needle was brought into momentary contact with the sample surface and then immediately withdrawn to achieve flash heating and rapid cooling. As a result, characteristic thermal degradation patterns reflecting the repeating units of polymers could be obtained within a short time, enabling simple analysis without pretreatment. However, the method still had limitations for rapid analysis, such as the need to manually adjust the needle position for each measurement. In addition, a stable baseline, in which the background ion signals remain constant during the measurement without upward or downward drift or sudden fluctuations, was important for reliable measurements. In long-term measurements using a direct current (DC) power supply, contamination of the discharge needle tip was a potential concern. Thus, some issues still remain in the use of DC corona discharge for reliable and stable measurements.

In the present work, we introduced a movable needle inserted into a heated ceramic tube so that the needle stops and remains on the sample surface upon contact, thereby avoiding deep penetration and eliminating the need for manual adjustment. By using this configuration, a simplified process involving flash heating for 200 ms followed by rapid cooling was achieved. In addition, alternating current (AC) corona discharge was adopted instead of DC corona discharge for the ionization of desorbed gaseous products. DC corona discharge has long been used as an ion source for atmospheric-pressure chemical ionization (APCI),^4–17) whereas AC corona discharge has recently been shown to offer several advantages, including stronger ion signals, suppression of transition from corona to arc, and stable proximity ionization.^18–26) By combining the flash-heating/rapid-cooling system with AC corona discharge ionization, we aimed to establish a practical method for analyzing gaseous products generated by the thermal decomposition of synthetic polymers.

EXPERIMENTAL

Chemicals and samples

Polystyrene (PS, APS-M5) was purchased from Acrysunday Co., Ltd. (Tokyo, Japan). Polyoxymethylene (POM, H1-101), polymethyl methacrylate (PMMA, E2-101A), and polyamide 6 (PA6, K1-104) were purchased from Spectra Forum (Yokkaichi, Japan). Polypropylene (PP, 2-9221-02) was purchased from AS ONE Corp. (Osaka, Japan). Nichrome filament (FCHW2, φ 0.5 mm) was purchased from BioChromato, Inc. (Fujisawa, Japan). The average molecular weights of the purchased polymers were not available in the catalogs.

AC corona ion source

A conceptual diagram of the experimental system is shown in Fig. 1. The AC corona discharge ion source was built in a closed stainless-steel chamber. The desorbed molecules were introduced into the ion source and ionized by AC corona discharge using a custom-made radiofrequency (RF) power source manufactured by ARIOS Co., Ltd. (Akishima, Tokyo, Japan). According to our previous study,²³⁾ an RF of 15–20 kHz produced the strongest ion intensities. In this work, an 18-kHz RF voltage of ±3.0 kV_pp, which was approximately ±0.5 kV_pp above the threshold voltage for the occurrence of AC corona discharge, was applied to the corona needle (outer diameter: 0.65 mm; curvature diameter of the needle tip: ~1 μm; Seirin, Shizuoka, Japan). The distance between the tip of the discharge needle and the counter electrode positioned in front of the inlet of the mass spectrometer (MS) was 1.5 mm. At this distance, a highly stable AC corona discharge was maintained without the occurrence of arc discharge. The ionized molecules were analyzed using a Quadrupole Time-of-Flight mass spectrometer (compact QTOF, Bruker Daltonics, Bremen, Germany; mass resolution: 30,000). The MS inlet voltage was set to −500 V for the MS used in this experiment. However, the ion source developed in this work is readily applicable to any type of ion source kept at ground potential. The ion source temperature was set to room temperature, while the curtain gas temperature was set to 200°C and supplied at a flow rate of 2 L/min. Mass spectra were acquired over an m/z range of 50–1000.

Fig. 1. Conceptual diagram of the thermal flash-heating/rapid-cooling system coupled with an AC corona discharge ion source. AC, alternating current.

Thermal flash desorption system

In this study, industrial plastic pellets were used as test samples. As shown in Fig. 1, a polymer pellet was fixed onto an SUS substrate using glue (adhesive putty). The x and y positions of the sample to be analyzed were adjusted to match the default x and y positions of the heated needle tip along the perpendicular z-axis. The sample was placed 5 mm from the inlet of the gas-introducing ceramic tube (i.d.: 4 mm; o.d.: 6 mm; 65 cm in length).

The most important part of the present flash desorption system is the configuration of the heating system. A nickel alloy needle, with an outer diameter of 1.0 mm and a length of 65 mm, was inserted into the heated ceramic tube with an inner diameter of 1.2 mm. The curvature diameter of the needle tip was ~54 μm, as shown in the inset of Fig. 1. With this configuration, the needle moved almost freely through the ceramic tube. As shown in Fig. 1, the tip of the needle was adjusted to protrude from the terminal end of the ceramic tube by 4 mm, which is adjusted using a stopper similar to a nail head.

The needle was indirectly heated by the ceramic tube, which was heated with a nichrome wire. The temperature of the needle was controlled by the nichrome wire filament current and measured using a pyrometer (Optris Xi 80, Optris GmbH & Co. KG, Berlin, Germany).

The working distance of the needle, controlled by a linear actuator, was 50 mm. The needle started to move downward at a speed of 600 mm/s over a distance of 30 mm and then slowed down to zero at a distance of 50 mm. The Z position of the sample was adjusted to be 49 mm below the tip of the needle. At this distance, the speed of the needle was slowed down to approximately 33 mm/s. Thus, the needle tip touched the polymer sample at this speed. When the needle touched the sample surface, the needle did not penetrate into the sample but floated approximately 1 mm within the ceramic tube after touching, as shown in Fig. 1. The penetration depth of the heated needle into the polymer sample depends on the heater temperature, the weight of the needle (4.1 g), and the contact time of the needle with the sample. In this experiment, the contact time was adjusted to 200 ms. Under these conditions, the penetration depth was estimated using the depth-of-focus method with a customized microscope (SKM-Z200C-PC; SAITOH KOUGAKU Co., Ltd., Yokohama, Japan). The penetration depths for PS and PMMA at a needle temperature of 350°C were 144 and 150 µm, respectively. Similarly, the penetration depths for POM and PMMA at a needle temperature of 450°C were 270 and 226 µm, respectively. Furthermore, the penetration depths for PP at needle temperatures of 300, 400, and 500°C were 87, 133, and 216 µm, respectively. The penetration depth increased with increasing temperature. After touching, the needle was lifted up at a speed of 600 mm/s, enabling rapid cooling of the polymer sample and of desorbed molecules after flash heating. The desorbed molecules were transferred to the ion source through a suction tube by the MS vacuum and the suction pump (Fig. 1).

RESULTS AND DISCUSSION

Under the present APCI conditions using AC corona discharge ionization, H₃O⁺ generated from the moisture in the air may act as the major protonating reagent ions. All the major ions were detected as protonated forms or as cluster ions with NH⁺₄.

$${\mathrm{H}}_2{\mathrm{O}}^{+}+{\mathrm{H}}_2\mathrm{O}\to {\mathrm{H}}_3{\mathrm{O}}^{+}+\mathrm{O}\mathrm{H}\cdot$$ (1)

$${\mathrm{H}}_3{\mathrm{O}}^{+}+{\mathrm{N}\mathrm{H}}_3\to {{\mathrm{N}\mathrm{H}}_4}^{+}+{\mathrm{H}}_2\mathrm{O}$$ (2)

$${{\mathrm{N}\mathrm{H}}_4}^{+}+\mathrm{M}+\text{third body}\to {[\mathrm{M}+{\mathrm{N}\mathrm{H}}_4]}^{+}+\text{third body}$$ (3)

A third body (N₂ or O₂) is necessary to stabilize the intermediate complex [M + NH₄]^+* to form the cluster ion [M + NH₄]⁺. NH₃ originated from impurities in the air.

Because the polymers used in this study have different chemical structures and therefore different thermal decomposition temperatures, each material was measured at an appropriate needle-contact temperature. When polymers are heated below their thermal decomposition temperatures, they do not exhibit characteristic decomposition patterns specific to the material, whereas excessive decomposition at higher temperatures can lead to overly complex spectra that hinder interpretation. For these reasons, the needle temperature was adjusted according to the polymer material. For unknown polymers, the analysis can be performed by increasing the temperature stepwise from 350°C to 650°C in 100°C increments until informative spectra are obtained, enabling flexible application of the method.

Mass spectra obtained for some representative polymers will be described. Figure 2 shows the positive-mode mass spectrum obtained for PS measured at 350°C. The monomer units of [R1-(C₈H₈)_n-R2 + H]⁺ with (R1 + R2) = C₁₆H₁₄O were observed For PS, the observed repeating units corresponded to the styrene unit (C₈H₈), and the sum formula for both terminal groups was suggested to be C₁₆H₁₄O. Because the terminal-group structure could not be determined from the present data, only the combined terminal-group composition was reported.

Fig. 2. Mass spectrum of PS measured at 350°C. PS, polystyrene.

Figure 3 shows the mass spectrum of POM measured at 450°C. The monomer units of [R1-(CH₂O)_n-R2 + NH₄]⁺ with (R1 + R2) = CH₂O and [R1′(CH₂O)_n-R2′ + NH₄]⁺ with (R1′ + R2′) = C₂H₄O were observed. For POM, repeating units of POM (CH₂O) were observed, and the combined terminal-group composition was CH₂O. Another series with terminal groups corresponding to C₂H₄O was also detected, likely originating from differences in polymer end-group structures.

Fig. 3. Mass spectrum of POM measured at 450°C. POM, polyoxymethylene.

Figure 4 shows the mass spectrum of PMMA measured at 350°C. Monomer units of [R1-(C₅H₈O₂)_n-R2 + H]⁺ with (R1 + R2) = C₇H₈O₃ were detected. For PMMA, the repeating unit of methyl methacrylate (C₅H₈O₂) was observed, and the combined terminal-group composition was suggested to be C₇H₈O₃.

Fig. 4. Mass spectrum of PMMA measured at 350°C. PMMA, polymethyl methacrylate.

Figure 5 shows the mass spectrum of PA6 measured at 450°C. The monomer units of [R1-(C₆H₁₁NO)_n-R2 + H]⁺ with (R1 + R2) = C₆H₁₁NO were observed. For PA6, repeating units of C₆H₁₁NO were detected, and the combined terminal-group composition was assigned as C₆H₁₁NO.

Fig. 5. Mass spectrum of PA6 measured at 450°C. PA6, polyamide 6.

Figure 6 shows the mass spectra of PP measured at 300, 400, and 500°C. At 300°C, thermal degradation of the main chain was minimal, and only additives were observed (m/z 571.4794 for DMTDP and m/z 647.4604 for Irgafos168). At 400 and 500°C, repeating units of propylene (C₃H₆, Δm/z 42) originating from main-chain cleavage became dominant. The combined terminal-group composition was assigned as C₅H₈O₂ (unsaturation 2, two oxygen atoms) at 400°C, whereas it shifted to C₆H₁₀O (unsaturation 2, one oxygen atom) at 500°C. The stronger oxidation at 400°C was attributed to slower heat transfer, which allowed oxidization to proceed during decomposition. At 500°C, monomer units of [R1′-(C₃H₆)_n-R2′ + H]⁺ with (R1′+ R2′) = C₆H₁₀O became the dominant species. High-temperature flash heating was suggested to suppress secondary oxidative reactions by rapidly vaporizing the sample, thereby preventing the formation of overly complex mass spectra. The present method is characterized by the ability to precisely control both temperature and reaction time on an extremely short timescale, which is considered to contribute to the effective suppression of secondary reactions. Further detailed analyses of the mass spectra are currently in progress in our laboratory.

Fig. 6. Mass spectra of PP measured at 300, 400, and 500°C. PP, polypropylene.

Signals observed below m/z 210 were attributed to fragments generated by the thermal decomposition of the polymers. The fragmentation patterns differed among the samples due to differences in polymer structures, resulting in distinct low-mass spectral features.

In addition, this method requires no sample pretreatment and enables localized analysis of bulk materials. Under the present operating conditions, no carryover attributable to residue adhesion on the needle tip was observed. When measurements are performed below the thermal decomposition temperature, residual material can be effectively removed by brief heating of the needle above the thermal decomposition temperature after measurement.

CONCLUSION

In this study, a thermal flash desorption/rapid point analysis system coupled with AC corona discharge ionization was developed for the analysis of industrial polymers. By adopting AC corona discharge instead of a DC corona discharge, a highly stable corona plasma could be maintained without the occurrence of arc discharge.

For the thermal decomposition of nonvolatile samples such as industrial polymers and for simultaneous thermal desorption, a needle inserted in a heated ceramic tube was used. Because the needle is freely movable within the heated ceramic tube, it stops at the moment of contact with the sample surface; that is, the overrun of the needle is avoided. The great advantage of the present system is that no touch sensor is needed because the stopping point of the needle just corresponds to the contact position with the sample surface.

Thus, the present technique would be highly useful for rapid point analysis of nonvolatile materials, especially the industrial polymer materials investigated in this study. In addition, further applications of this technology to other materials, such as rubber, are anticipated.

Mass Spectrom (Tokyo) 2026; 15(1): A0194