Rapid detection of multiple gas mixtures and evaluation of harmful gas removal efficiency in a deck decompression chamber using dynamic switching mass spectrometry

Abstract

The multiple gas mixtures in the deck decompression chamber (DDC) serve as a life-sustaining medium for divers to adapt to high-pressure environments and to safely decompress after saturation diving operations. However, due to the complexity of the gas composition and the wide concentration range, there is currently no technique that can rapidly detect all gases in the DDC. In this study, we developed a photoinduced chemical ionization (pCI) and electron impact (EI) dynamic switching mass spectrometry system. This technique was applied to investigate the spatial distribution uniformity of respiratory gases within the DDC at pressures of 0.13 MPa, 0.6 MPa, and 1.5 MPa, as well as the removal efficiency of toxic gases at 1.5 MPa. The switching stability time of pCI/EI dynamic switching mass spectrometry was 30 s per cycle, with optimal detection performance metrics including a detection limit down to 0.58 ppb (H₂S), a fastest response time of 2 s (for CO₂ and H₂S), sensitivity up to 62.97 counts per ppb (H₂S), and an overall concentration detection range spanning from 1.0 × 10⁻⁹ v/v (1 ppb) to 4.5 × 10⁻¹ v/v (45%), covering the requirements for all target gases. At the three pressure levels, the spatial distribution uniformity of O₂ in the DDC was 0.34%, 0.21%, and 0.08%, with the time to reach uniform distribution being 5 min, 3 min, and 1 min, respectively. For CO₂, the spatial distribution uniformity was 0.62%, 1.61%, and 1.30%, with uniform distribution achieved in 10 min, 7 min, and 5 min, respectively. The removal efficiencies for harmful gases such as CO₂, CH₄, NH₃, and H₂S were 39.57%, 10.75%, 14.29%, and 3.96%, respectively. Under high-pressure conditions, O₂ and CO₂ rapidly achieved uniform distribution within the DDC. The pCI/EI dynamic switching mass spectrometry technology provides a rapid detection method for multicomponent gases in enclosed chambers like the DDC. This technology holds significant implications for the life support and operational efficiency of personnel involved in shipwreck salvage, underwater construction, and deep-sea exploration of marine resources.

---

1 Introduction

The deck decompression chamber (DDC) is a temporary life support chamber in which the bodies of saturated divers gradually acclimate to varying water pressures at different depths or slowly (over several hours to several days) release high-pressure gases from their bodies.^1,2 The gas composition within the DDC is complex and spans a wide concentration range, constituting a multiple gas mixture. This mixture includes both life-sustaining gases such as oxygen (O₂),³ and gases related to human metabolism, such as carbon dioxide (CO₂), methane (CH₄), ammonia (NH₃), and hydrogen sulfide (H₂S).^4–7 Among these, NH₃ and H₂S are harmful odorous gases, detectable by human olfaction even at low concentrations.^8,9 According to Chinese national standards, the concentration threshold for H₂S in the DDC is at the ppb level, for NH₃ at the ppm level, and for CH₄ and CO₂ around 1%.¹⁰ Moreover, to prevent oxygen toxicity, the concentration of O₂ must be dynamically controlled during both pressurization and depressurization processes.¹¹ The gradual pressurization and depressurization inside the DDC subject divers to continuous physiological and psychological stress, and the multicomponent gas mixture directly affects their sensory experiences and overall health. Therefore, rapid detection of the multicomponent gases within the DDC is critical for ensuring the safety and operational efficiency of divers engaged in shipwreck salvage, underwater construction, and deep-sea exploration of marine resources.

Currently, the techniques used to detect multiple gas mixtures within the DDC include sensor technology, spectroscopy, and mass spectrometry. Sensor technology is suited for detecting specific types of gases. Lillo et al. integrated sensors for NH₃, carbon monoxide, chlorine, hydrogen chloride, hydrogen cyanide, H₂S, nitrogen dioxide, and sulfur dioxide gases to achieve real-time online detection of gases at ppm levels in high-pressure chambers.¹² Although combining multiple sensors can enable real-time online detection of multicomponent gases in the DDC, sensor technology has limited detection capabilities for trace gases such as NH₃ and H₂S. Spectroscopy, on the other hand, enables qualitative and quantitative analysis of gases by measuring the absorption of specific light frequencies by gas molecules. Lillo et al. developed an in situ online CO₂ monitor for high-pressure environments, enabling real-time monitoring of CO₂ in a dry deck shelter at six absolute atmospheres.¹³ However, since different gases require lasers of different frequencies, single-frequency spectroscopy cannot simultaneously detect multiple gas mixtures. While both sensor and spectroscopy technologies can operate directly under high-pressure conditions and enable rapid in situ detection of gases in the DDC, neither can simultaneously detect all multiple gas mixtures within the chamber.

Mass spectrometry, including gas chromatography–mass spectrometry (GC-MS), is a commonly used technique for qualitative and quantitative analysis of gas mixtures.¹⁴ GC-MS requires offline sampling, followed by the selection of appropriate chromatographic columns and sample pretreatment methods (e.g., dilution or concentration) based on the polarity and concentration of the gases,^15,16 making the analysis process complex and time-consuming. Online mass spectrometry techniques, such as electron impact mass spectrometry (EI-MS), have been employed for gas detection in enclosed chambers. Arieli et al. directly inserted a high-resistance sampling capillary into a high-pressure chamber with pressures ranging from 0.10 MPa to 0.51 MPa and utilized EI-MS to conduct real-time online monitoring of high concentrations of O₂ (up to 21%) and CO₂ (up to 5%) within the chamber.¹⁷ To protect the health of astronauts aboard space stations, EI-MS has also been used for the direct monitoring of permanent gases such as O₂, N₂, CO₂, and CH₄ in the space station environment.^14,18 Although the 70 eV electrons in EI-MS can ionize most gases,¹⁹ interactions between electrons and gas mixtures often result in competition between high-concentration matrix gases like O₂ and CO₂ and trace gases such as H₂S and NH₃, limiting the ionization efficiency of trace gases.²⁰ In contrast, photoinduced chemical ionization mass spectrometry (pCI-MS) enables rapid and highly sensitive detection of H₂S and NH₃.^21,22 However, due to the high sensitivity of pCI-MS, it is generally used to detect ppbv or even lower concentration gases, which is not suitable for detecting high concentration gases such as O₂, CO₂, and CH₄ in DDC.

In this study, we aim to develop pCI and EI dynamic switching mass spectrometry capable of simultaneously detecting multiple gas mixtures within the DDC. First, the structure of the dual-source dynamic switching mass spectrometry system and dynamic switching mode will be introduced. Next, the instrument's performance will be evaluated. Finally, dual-source dynamic-switching mass spectrometry will be utilized to examine the uniformity of gas distribution under various operating conditions (0.13 MPa, 0.5 MPa, and 1.5 MPa) in an unmanned deck decompression chamber, as well as the removal efficiency of harmful gases at 1.5 MPa. This instrument offers a rapid and precise method for detecting multiple gas mixtures in the DDC and provides a refined tool for detecting the quality of breathing gas for divers engaged in saturation diving.

2 Materials and methods

2.1 Deck decompression chamber

The structure of the DDC and the process for sampling and analyzing the multiple gas mixtures inside the chamber are illustrated in Fig. 1. The DDC has a capsule-like shape with a volume of 2.3 m³. The internal pressure of the DDC is regulated by a pressurization station, which supplies synthetic air through a dedicated pipeline connected to the DDC. CO₂ and O₂ are introduced into the DDC via separate pipelines, and their distribution within the chamber is maintained using a gas equilibrium device (circulating fan). The experiment is an unmanned simulation. Seven spatial sampling points (S1, S2, S3, S4, S5, S6, and S7) are pre-arranged around the resting bed area for divers, and the gases inside the DDC are collected from these sampling points into sampling bags. The gases in the sampling bags are then analyzed by dual-source switching mass spectrometry.

Fig. 1 The structure of the DDC and the process for sampling and analyzing the multiple gas mixtures.

2.2 Photoinduced chemical ionization and electron impact dynamic switching mass spectrometry

The structure of the dual-source dynamic switching mass spectrometer is shown in Fig. 2. This mass spectrometer consists of a photoionization chemical ionization (pCI) source, an EI source, a quadrupole mass analyzer, and an electrical control system. The pCI source is used for the ionization of H₂S and NH₃, while the EI source is used for the ionization of high-concentration inorganic gases. Both the pCI and EI sources share a common sample introduction system and mass spectrometry detection system. The pCI and EI sources can operate independently or switch automatically between the two ion sources.

Fig. 2 The structure of the dual-source dynamic switching mass spectrometry system.

The pCI source comprises a photoionization Kr lamp and a radio frequency (RF) ion funnel drift tube. The Kr lamp emits photons with an energy of 10.60 eV.²³ The RF ion funnel drift tube is composed of stainless steel ring electrodes with a linearly varying inner diameter, sealed with polytetrafluoroethylene gaskets. The stainless steel electrodes are subjected to both RF and direct current (DC) voltages. During the operation of the pCI source, the frequency of the RF is set to 1 MHz, the amplitude of the RF voltage is 100 V, and the DC electric field strength within the reaction tube is 4.62 V cm⁻¹.

The ionization energies of H₂S and NH₃ are 10.46 eV and 10.07 eV, respectively, both of which are lower than the 10.60 eV energy of the Kr lamp. Therefore, both compounds can undergo single-photon ionization, resulting in the formation of molecular ion peaks (M⁺) and the release of electrons. The ionization process is as follows:

H₂S + hv → H₂S⁺ + e⁻ (1)

NH₃ + hv → NH₃⁺ + e⁻ (2)

In our experimental system, a large amount of H₃O⁺ ions was observed, which may originate from the reaction of electrons (e⁻) released during the photoionization process with residual water molecules, though the specific mechanism requires further investigation. The ionization process produces prominent H₃O⁺ ions (m/z 19), which subsequently undergo proton transfer reactions with H₂S/NH₃ to form H₃S⁺ (m/z 35) and NH₄⁺ (m/z 18), as shown by the characteristic peaks in Fig. S1.†

When the EI source is in operation, the RF ion funnel drift tube does not apply RF or DC voltages. CH₄, O₂, and CO₂ pass directly through the drift tube into the mass spectrometry chamber. In the mass spectrometry chamber, the EI source generates electrons with an energy of 70 eV. These electrons collide with CH₄, O₂, and CO₂, producing molecular ion peaks and fragment peaks. During the operation of either the pCI or the EI source, the temperature of the drift tube is maintained at 80 °C, and the pressure within the reaction tube is maintained at 0.8 hPa. The sample introduction line consists of passivized stainless steel capillaries, with a sample flow rate of 48 mL min⁻¹. The quadrupole mass analyzer has a mass analysis range of 0.1 amu to 510 amu and operates at a pressure of 3.2 × 10⁻⁶ hPa (the pressure remains constant during ion source switching). The electrical control system manages the dynamic switching of the pCI and EI sources and the corresponding mass spectrometry parameters, as well as data acquisition. The switching process is automatically completed by the control system based on the set parameters. Additionally, to demonstrate the characteristics of the dual ion source system, as shown in the ESI Fig. S2,† we compared the performance of pCI-MS and EI-MS when detecting typical target compounds. The pCI mode exhibits higher selectivity and sensitivity for H₂S and NH₃ and can effectively suppress background gas interference. In contrast, the EI mode is suitable for detecting inorganic gases like N₂.

2.3 Data quality assurance and control

To ensure the accuracy of the qualitative and quantitative data from the instrument, the mass axis was checked every 24 hours and recalibrated if any deviations were detected. The mass spectrometer was calibrated every 24 hours using multi-point calibration with standard gases. The standard gases, including H₂S, NH₃, O₂, CO₂, and CH₄, were purchased from Shanghai Wetry Standard Gas Analysis Technology Co., Ltd. Aluminum foil sampling bags with a capacity of 3 L were purchased from Ningbo Hongpu Laboratory Technology Co., Ltd.

3 Results and discussion

3.1 Dynamic switching mode of dual-source switching mass spectrometry

The signal response changes of monitored compounds during ionization source switching under actual working conditions are illustrated in Fig. 3. During the dynamic switching between the EI and pCI sources, the drift tube temperature (80 °C), pressure (0.8 hPa), and mass spectrometry chamber pressure (3.2 × 10⁻⁶ hPa) remained constant. In the EI-MS mode, the Kr lamp of the pCI source was turned off, and no photons were emitted. The electron energy in the EI-MS mode was set to 70 eV. When gases are subjected to 70 eV electrons, molecular ions and fragment ions are generated. Although these fragment ions provide structural information, the molecular ion peak of CH₄ (CH₄⁺) is at m/z 16.0425, while fragment ions of O₂ and CO₂ produced under 70 eV electrons appear at O⁺ (m/z 15.9994). Given the mass resolution of 1 for the quadrupole mass spectrometer, it is impossible to directly obtain the signal intensity of CH₄ at m/z 16. However, based on the standard EI-MS mass spectrum of CH₄, the primary signal peak for CH₄ is at m/z 15. By selecting m/z 15, the signal intensity of CH₄⁺ can be identified, avoiding interference from O⁺. In EI-MS mode, the mass spectrometric data for CH₄, O₂, and CO₂ are saved in EI-MS files. During the operation of the pCI source, the filament of the EI source was turned off. The mass spectrometric data for NH₃ and H₂S were saved in pCI-MS files. The EI-MS and pCI-MS modes were switched cyclically: EI-MS detected CH₄, O₂, and CO₂ with a dwell time of 0.1 s per ion, and pCI-MS detected NH₃ and H₂S with a dwell time of 0.1 s per ion. Therefore, the theoretical fastest switching frequency between EI-MS and pCI-MS was 5 Hz. However, in practice, it took time for the ion sources to switch and the signal to stabilize.

Fig. 3 The signal response changes of (a) CH₄, (b) CO, (c) O₂, (d) H₂S and (e) NH₃ during ionization source switching under actual working conditions.

As illustrated in Fig. 3(a)–(c), when switching from pCI-MS to EI-MS, the ionic signals of CH₄, CO₂, and O₂ require approximately 30 seconds (t₁) to achieve stability. During the initial activation of the EI source, significant signal oscillations are observed (potentially associated with electron emission stabilization and plasma formation processes), with signals beginning to approach a steady state at approximately t1/2.

In contrast, as demonstrated in Fig. 3(d) and (e), the reverse transition (EI → pCI) exhibits a markedly reduced stabilization time of less than 1 second (t₂) for NH₃ and H₂S ionic signals. This temporal disparity primarily stems from the distinct physical mechanisms governing the two switching processes: t₁ (pCI → EI) encompasses both the physical transition time of the ion source and the period required for the EI source to establish electron emission equilibrium and stabilize plasma conditions, whereas t₂ (EI → pCI) predominantly involves the deactivation of the EI source and the instantaneous initiation of the Kr lamp, which rapidly establishes a stable photoionization state.

These findings validate the practical utility of the dual ion source system in real-world applications: although the transition from pCI mode to EI mode necessitates a certain preheating period, the system can promptly resume analytical capabilities upon returning from EI mode to pCI mode, rendering it suitable for analytical requirements demanding flexible alternation between the two modes. Additionally, these observations provide valuable reference for experimental protocol design—the extended preheating time of EI-MS may require consideration in time-sensitive applications.

The use of the pCI source enabled efficient soft ionization of trace gases such as NH₃ and H₂S, while most matrix gases (e.g., N₂, O₂, and CO₂) could not be detected due to their inability to undergo proton transfer reactions. In contrast, when ionizing trace gases, the EI source exhibited limited ionization efficiency, making it difficult for mass spectrometry to detect these trace gases. However, the EI source produced clear signals for high-concentration matrix gases such as N₂, O₂, and CO₂. By flexibly combining and switching between the EI and pCI sources, rapid detection of a wide range of gas concentrations was achieved without the need to replace hardware. This approach saved significant time and improved experimental efficiency. However, the rapid switching between the dual sources imposes higher demands on the reliability of key components.

3.2 The performance of dual-source dynamic switching mass spectrometry

The calibration curves for O₂, CH₄, CO₂, NH₃, and H₂S are shown in Fig. 4. The concentration range of the O₂ calibration curve is 5% to 45%, which covers the concentrations typically found in atmospheric and pressurized chambers. The CH₄ calibration curve spans a concentration range of 100 ppm to 1500 ppm, while the CO₄ calibration curve ranges from 150 ppm to 3000 ppm. The NH₃ calibration curve covers a range of 1 ppm to 30 ppm, and the H₂S calibration curve ranges from 1 ppb to 30 ppb. All of these calibration curves have correlation coefficients (R²) exceeding 0.9975, with CH₄ and CO₂ reaching an R² of 0.9999, indicating that the pCI/EI-MS system provides excellent linear response for these gases. The instrument's response time was determined using the first-order system transient response method,²⁴ as shown in Table 1. The response times for CO₂ and H₂S are approximately 2 seconds, the shortest among the five gases. CH₄ and O₂ have response times of approximately 4 seconds, while NH₃ has a response time of about 11 seconds. Due to NH₃'s strong adsorption to the pipelines, the gas transfer lines are made of passivated stainless steel to reduce response time. Using the slope (k) of the calibration curve, the standard deviation (SD) in the instrument's background signal, and a signal-to-noise ratio of 3 (S/N = 3), the limits of detection (LODs) for these five gases were calculated (LOD = 3 × SD/k). The instrument's background signal was obtained by introducing high-purity nitrogen gas. As shown in Table 1, the LODs for CH₄, CO₂, NH₃, and H₂S are well below the limits specified in the Chinese standard for diver breathing gas and detection methods (GB-18435-2007), demonstrating that the performance of this system meets the requirements for practical applications.

Fig. 4 (a) Standard curve of O₂ in EI-MS mode; (b) standard curves of CH₄ and CO₂ in EI-MS mode; (c) the standard curve of NH₃ in pCI-MS mode; (d) standard curve of H₂S in pCI-MS mode.

Table 1 Response times and LODs for pCI/EI-MS of CH₄, O₂, CO₂, NH₃ and H₂S

			m/z	Response times (s)	LODs	GB-18435-2007 (China)
1	EI-MS	CH4	16	4	210.60 ppm	872.46 ppm
2		O2	32	4	841.90 ppm	—
3		CO2	44	2	20.13 ppm	33.67 ppm
4	pCI-MS	NH3	18	11	0.13 ppm	0.68 ppm
5		H2S	35	2	0.58 ppb	4.83 ppb

---

3.3 Gas equilibrium experiments in the DDC under 0.13 MPa, 0.6 MPa, and 1.5 MPa

The gas equilibrium test was conducted to assess whether the distribution of respiratory gases, O₂ and CO₂ is uniform at different locations within the DDC under varying pressures. The objective was to ensure that personnel inside the chamber have stable respiratory conditions while moving within the DDC. O₂ serves as a life-support gas, while CO₂ simulates the exhalation produced by divers.

During the gas equilibrium test, the ambient temperature remained stable at approximately 23 °C, and the relative humidity inside the DDC ranged from 50% to 60%. The pressure control device was operated to pressurize the DDC with synthetic air (before reaching the target pressure). High-purity O₂ and CO₂ were then introduced into the DDC simultaneously from high-pressure O₂ and CO₂ cylinders (until the DDC pressure reached the target value). Three pressure conditions were tested in the DDC: 0.13 MPa, 0.6 MPa, and 1.5 MPa. Under each condition, 500 L of O₂ and 80 L of CO₂ (at 23 °C and 0.1 MPa) were introduced. The gas equilibrium system was activated, and from time zero, air samples were collected at regular intervals from seven spatial sampling points in the DDC. For each point, five gas samples were collected using sampling bags, and the samples were analyzed using the pCI/EI-MS. The test primarily evaluated the spatial distribution uniformity of O₂ and CO₂. During the mass spectrometric analysis of the samples, only the EI source was used, with a dwell time of 1 s.

At a pressure of 0.13 MPa, the concentration changes of O₂ and CO₂ inside the DDC are shown in Fig. 5(a and d). At 0.13 MPa, the time required for O₂ to reach spatial equilibrium was 5 minutes. During this process, the average O₂ concentration across the seven spatial sampling points exhibited a linear increase, and the relative standard deviation (RSD) of the concentration distribution among the seven sampling points was 0.49%. After 5 minutes, the concentrations at all seven sampling points reached equilibrium, with an RSD of 0.23%, indicating a 56.36% reduction in spatial fluctuation. This result suggests that the seven sampling points were not spatially equivalent, leading to differences in O₂ concentrations during equilibrium. These differences were likely due to the varying distances of the sampling points from the gas equilibrium device. For CO₂, the time to reach spatial equilibrium at 0.13 MPa was 10 minutes. The CO₂ concentrations in the seven spatial sampling bags also exhibited a linear increase, and as shown in Fig. 5(d), the concentrations at sampling points S7, S5, and S1 were noticeably higher than those at S2, S3, S4, and S6 during equilibrium, further indicating spatial variation among the sampling points. The equilibrium time for CO₂ was twice that of O₂ at 0.13 MPa. Despite having the same average molecular kinetic energy at the same temperature, O₂ has a lower molecular weight than CO₂, resulting in faster diffusion of O₂. Furthermore, when comparing the relative concentration increases before and after equilibrium, O₂ exhibited a 14.81% increase, while CO₂ showed a 178.66% increase. Considering the diffusion speed and concentration differences before and after equilibrium, O₂ reached equilibrium faster with a smaller increase in concentration, resulting in a shorter spatial equilibrium time. The RSD of CO₂ spatial fluctuation before equilibrium was 6.33%, while after equilibrium, it was 2.52%, representing a 60.19% reduction. This outcome was also due to the varying distances of the sampling points from the gas equilibrium device.

Fig. 5 The concentration change of O₂ (a–c) and CO₂ (d–f) at seven spatial sampling points with time under 0.13 MPa, 0.6 MPa and 1.5 MPa pressure in DDC.

At a pressure of 0.6 MPa, the changes in O₂ and CO₂ concentrations within the DDC are shown in Fig. 5(b and e). Compared to the spatial equilibrium time of 5 minutes at 0.13 MPa, the time for O₂ to reach equilibrium in the DDC decreased to 3 minutes at 0.6 MPa. At 1.5 MPa, the equilibrium time further reduced to 1 minute, as shown in Fig. 5(c). This reduction in equilibrium time is related to the increased diffusion rate of O₂ and the smaller concentration increase before and after equilibrium. At constant temperature, the average kinetic energy of O₂ remains unchanged, but with increasing pressure, the frequency of O₂ collisions with surrounding molecules increases, leading to a faster diffusion rate in the DDC. Additionally, the concentration increments of O₂ at 0.6 MPa and 1.5 MPa were 1.47% and 1.19%, respectively, further contributing to the reduced equilibrium time. The spatial equilibrium time for CO₂ was approximately 7 minutes at 0.6 MPa and decreased to 5 minutes at 1.5 MPa. As the pressure in the DDC increased, the spatial equilibrium times for both O₂ and CO₂ shortened.

As shown in Fig. 5(a–c), the spatial equilibrium times for O₂ in the DDC were 5 minutes, 3 minutes, and 1 minute, respectively. After equilibrium, the O₂ concentrations were 23.16% (±0.34%), 20.47% (±0.21%), and 20.04% (±0.08%). The results indicate that the distribution deviations of O₂ in the DDC under different conditions were consistent, with the RSD values not exceeding 0.34%, demonstrating good equilibrium across different pressures. Moreover, as the pressure in the DDC increased, the time required for O₂ to reach spatial equilibrium decreased, allowing for quicker adjustments of O₂ concentration in the pressurized chamber, thereby ensuring the quality of the breathing air for divers. As shown in Fig. 5(d–f), the spatial equilibrium times for CO₂ were 10 minutes, 7 minutes, and 5 minutes, respectively, with post-equilibrium concentrations of 1871.43 ppm (±0.62%), 707.14 ppm (±1.61%), and 561.23 ppm (±1.30%). The equilibrium times for both O₂ and CO₂ decreased as pressure increased, and the spatial distribution uniformity for CO₂ in the DDC showed an RSD of no greater than 1.61%. The fluctuations in O₂ and CO₂ concentrations were partly due to system noise and partly due to the inherent uneven distribution of the gases. Thus, the actual RSD values for the concentration distribution at the seven spatial sampling points are likely lower than the calculated values. When the partial pressure of O₂ in human breathing air exceeds 0.14 MPa, central nervous system oxygen toxicity can occur.^25–27 In high-pressure environments, this high-precision, rapid mass spectrometry detection technique plays a critical role in acquiring basic data on individual tolerance and preventing oxygen toxicity.

Based on the gas equilibrium experiments conducted at 0.13 MPa, 0.6 MPa, and 1.5 MPa, we conclude that the distribution of respiratory gases (O₂ and CO₂) within the DDC achieves satisfactory uniformity after reaching equilibrium. The relative standard deviations in the spatial distribution of O₂ did not exceed 0.34% across all pressure conditions, while CO₂ showed RSDs no greater than 1.61%. These low RSD values confirm that after equilibrium is reached, personnel can move freely within the DDC while experiencing consistent respiratory conditions. The equilibrium time decreases with increasing pressure—from 5 to 1 minute for O₂ and from 10 to 5 minutes for CO₂—when pressure increases from 0.13 MPa to 1.5 MPa. This knowledge of equilibrium timing and distribution uniformity is crucial for establishing proper ventilation protocols and ensuring safe breathing conditions during hyperbaric operations.

3.4 Harmful gas removal experiment in the DDC under 1.5 MPa

Using the pressure control device, synthetic air was injected into the DDC, followed by the introduction of specific amounts of CO₂, CH₄, NH₃, and H₂S into the chamber. The initial concentrations of these gases are shown in Table 2. The gas equilibration device was then activated. To verify the effectiveness of the toxic and harmful gas removal module in the unmanned simulation chamber, the pCI/EI-MS was employed to simultaneously measure the concentrations of CO₂, CH₄, NH₃, and H₂S before and after the module's operation, with a dwell time of 0.1 s per ion and a switching stability time of approximately 30 seconds per cycle. It is worth noting that before starting the dual-source switching detection, the system needs to wait for the gas signal to completely stabilize (≥11 seconds), and the detection results are presented in Table 2.

Table 2 The harmful gas removal test under 1.5 MPa air conditions in the DDC

		Initial concentration	Adsorption time (min)	Concentration after adsorption	Concentration clearance	Removal efficiency (%)
1	CO2	904.52 ppm	20	528.59 ppm	375.93 ppm	39.57%
2	CH4	867.00 ppm	60	773.83 ppm	93.17 ppm	10.75%
3	NH3	0.21 ppm	20	0.18 ppm	0.03 ppm	14.29%
4	H2S	6.57 ppb	20	6.32 ppb	0.26 ppb	3.96%

---

In this experiment, CO₂ in the DDC was removed using soda lime. Prolonged exposure to low concentrations of CO₂ can cause headaches, dizziness, impaired concentration, and memory loss,^28,29 while short-term exposure to high concentrations of CO₂ can lead to respiratory center paralysis and even death.^30,31 According to the International Marine Contractors Association (IMCA) standards, each diver's oxygen consumption rate is calculated to be 0.5 L min⁻¹, regardless of diving depth.³² A diver's respiratory quotient typically ranges from 0.7 to 1.1,³³ meaning that the rate of CO₂ production from a diver's respiration is between 0.35 L min⁻¹ and 0.55 L min⁻¹. To ensure that the partial pressure of CO₂ in the decompression chamber remains within a reasonable range, the CO₂ removal rate in the simulation chamber must be no less than 0.55 L min⁻¹. As shown in Table 2, the CO₂ removal efficiency was 39.57%, with a removal rate of 0.65 L min⁻¹, which exceeds the required rate of 0.55 L min⁻¹.

In this experiment, CH₄ in the simulation chamber was removed using specially designed activated carbon. The CH₄ in the DDC is produced by human metabolism. Table 2 shows that the CH₄ removal efficiency was 10.75%, with a removal rate of 0.05 L min⁻¹. Although CH₄ is non-toxic, it can act as an asphyxiant at high concentrations. To increase the CH₄ removal rate, the amount of CH₄ adsorbent could be increased. Activated carbon was also used in this experiment to remove NH₃ and H₂S, both of which are toxic and odorous gases. Inhaling these gases can affect the physiological health and emotional stability of divers. According to China's national standards for diving breathing gases (GB-18435-2007), the NH₃ concentration in the chamber must be less than or equal to 0.68 ppm, and the H₂S concentration must be less than or equal to 4.83 ppb. As shown in Table 2, the current absorption conditions effectively remove NH₃, but the efficiency of H₂S removal needs to be improved. The results indicate that further enhancements are necessary to improve H₂S removal efficiency.

4 Conclusions

In this study, we developed a pCI and EI dynamic switching mass spectrometry system to detect multiple gas mixtures in a simulated multi-depth DDC. The pCI/EI-MS can dynamically switch ionization sources, achieving efficient ionization of high-concentration gases such as CH₄, O₂, and CO₂, as well as trace gases like NH₃ and H₂S. The LODs of the EI-MS ranged from 20.13 ppm to 841.90 ppm and the concentration range of the EI-MS was 100 ppm–45%, with response times between 2 and 4 seconds, while the LODs of the pCI-MS ranged from 0.58 ppb to 0.13 ppm, and the concentration range of the pCI-MS was 1 ppb–30 ppm, with response times between 2 and 11 seconds. Under pressure conditions of 0.13 MPa, 0.5 MPa, and 1.5 MPa, the RSDs of O₂ spatial distribution in the DDC were 0.34%, 0.21%, and 0.08%, respectively, and the RSDs of CO₂ spatial distribution were 0.62%, 1.61%, and 1.30%, respectively. At a pressure of 1.5 MPa, we assessed the removal efficiency of harmful gases in the DDC, with removal rates ranging from 3.96% to 39.57%. This system is capable of rapidly and accurately detecting the concentrations of multiple gas mixtures in enclosed chambers, which is crucial for safeguarding the psychological well-being and life safety of saturation divers. However, the technology also has some limitations, such as the quadrupole mass analyzer's resolution of 1, which prevents the system from distinguishing between CO and N₂. Additionally, the sampling method relies on sample bags, making real-time, online detection of gases in the DDC impossible. In future work, we plan to integrate high-resolution time-of-flight mass spectrometry to enable the detection and differentiation of CO and N₂ and design pressure conversion valves to allow continuous online monitoring of high-pressure gases in the DDC.

Author contributions

Qu Liang: writing – original draft, experiments, validation, and data curation. Pingxiao Liu: investigation, methodology, and visualization. Lei Zhao: project administration. Xuejun Wang: visualization. Jun Zou: methodology. Xun Bao: monitor software and data acquisition software. Qiangling Zhang: PTR-MS instrument. Wei Xu: visualization. Xue Zou: methodology. Shifeng Wang: conceptualization. Chaoqun Huang: formal analysis. Chengyin Shen: funding acquisition. Yannan Chu: supervision.

Data availability

The processed data supporting the findings of this study are available within the article and its ESI.† The raw datasets generated during the current study cannot be publicly shared due to confidentiality restrictions. Further inquiries regarding specific aspects of the restricted datasets may be directed to the corresponding author and may be considered on a case-by-case basis subject to confidentiality restrictions.

Conflicts of interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgements

This paper was financially supported by the National Natural Science Foundation of China [22376198, 22306187, 22076190, 21705152 and 62171433], the Youth Innovation Promotion Association, CAS, China [Y2023122], the Anhui Provincial Natural Science Foundation [2408085J020], the HFIPS Director's Fund [BJPY2021B08, YZJJ202305-CX], the Anhui Provincial Key R&D Program [2022a05020034], the Research Fund of Anhui Institute of Translational Medicine [2023zhyx-C78], the Health Research Program of Anhui [AHWJ2024Aa10005] and the Anhui Postdoctoral Scientific Research Program Foundation [2024C876].

References

1. J. P. Imbert, L. Matity, J. Y. Massimelli and P. Bryson, Review of Saturation Decompression Procedures used in Commercial Diving, Diving Hyperb. Med., 2024, 54(1), 23–38.
2. J. P. Imbert, J. Y. Massimelli, A. Kulkarni, L. Matity and P. Bryson, A Review of Accelerated Decompression from Heliox Saturation in Commercial Diving Emergencies, Diving Hyperb. Med., 2022, 52(4), 245–259.
3. J. Risberg, P. J. van Ooij and O. S. Eftedal, Decompression Procedures for Transfer under Pressure (‘TUP’) Diving, Diving Hyperb. Med., 2023, 53(3), 189–202.
4. I. D. Weiner, W. E. Mitch and J. M. Sands, Urea and Ammonia Metabolism and the Control of Renal Nitrogen Excretion, Clin. J. Am. Soc. Nephrol., 2015, 10(8), 1444–1458.
5. A. M. M. Bennett and T. Cibis, The Underwater World and Diving Physiology. Engineering and Medicine in Extreme Environments, Springer, 2022, pp. 57–87.
6. L. Ernst, B. Steinfeld, U. Barayeu, T. Klintzsch, M. Kurth and D. Grimm, et al., Methane Formation Driven by Reactive Oxygen Species Across all Living Organisms, Nature, 2022, 603(7901), 482–487.
7. G. Cirino, C. Szabo and A. Papapetropoulos, Physiological Roles of Hydrogen Sulfide in Mammalian Cells, Tissues, and Organs, Physiol. Rev., 2023, 103(1), 31–276.
8. J. Jiang, F. Wang, J. Wang and J. Li, Ammonia and Hydrogen Sulphide Odour Emissions from Different Areas of a Landfill in Hangzhou, China, Waste Manage. Res., 2021, 39(2), 360–367.
9. Z. Han, F. Qi, H. Wang, R. Li and D. Sun, Odor Assessment of NH3 and Volatile Sulfide Compounds in a Full-Scale Municipal Sludge Aerobic Composting Plant, Bioresour. Technol., 2019, 282, 447–455.
10. China. Breathing Gases for Divers and Their Determination Methods. Beijing: General Administration of Quality Supervision, Inspection and Quarantine of the People's Republic of China, Standardization Administration of the People's Republic of China; 2007. p. 11.
11. China. Requirements of Preparing Gases for Heliox Saturation Diving to 200 m. Beijing: Ministry of Transport of the People's Republic of China; 2009. p. 10.
12. R. S. Lillo and J. M. Caldwell, Development and Evaluation of a Hyperbaric Toxic Gas Monitor (SUBTOX) for Disabled Submarines, 2013.
13. R. S. Lillo, W. R. Porter, A. Ruby, W. H. Mints, J. Cladwell and J. F. Himm, Development and Evaluation of Hyperbaric Carbon Dioxide Analyzer for Dry Deck Shelter Operations, 1998.
14. M. B. Darrach, B. Bae, D. Fu, V. Garkanian, M. Homer, R. Kidd, C. Jung-Kubiak, H. Kraus, F. Maiwald, S. Madzunkov, C. Malone, D. Nikolic, M. Rais-Zadeh, J. Simcic, T. Tillmans and F. Zhong, Calibration and Performance of the Spacecraft Atmosphere Monitor, an Air Constituent Monitor for Human Spaceflight. 52nd International Conference on Environmental System; Calgary, Canada 2023.
15. G. Nyerges, D. Szieberth, J. Matyasi and J. Balla, Cave Air Analysis with Gas Chromatography Mass Spectrometry, Period. Polytech.: Chem. Eng., 2021, 65(3), 416–423.
16. B. Gruber, F. David and P. Sandra, Capillary Gas Chromatography-Mass Spectrometry: Current Trends and Perspectives, TrAC, Trends Anal. Chem., 2020, 124, 9.
17. R. N. Arieli, Y. Daskalovic, M. Eynan, Y. Arieli, O. Ertracht and A. Shupak, Use of a Mass Spectrometer for Direct Respiratory Gas Sampling from the Hyperbaric Chamber, Aviat., Space Environ. Med., 2001, 72(9), 799–804.
18. W. T. Wallace, T. F. L. Limero, J. Leslie, P. D. Mudgett and D. B. Gazda, Monitoring of the Atmosphere on the International Space Station with the Air Quality Monitor, International Conference on Environmental Systems (ICES 2017) ; Charleston, South Carolina, 2017.
19. A. Glineur, M. Beccaria and G. Purcaro, Exploring 20 eV Electron Impact Ionization in Gas Chromatography-Tandem Mass Spectrometry for the Determination of Estrogenic Compounds, J. Chromatogr. A, 2021, 1652, 5.
20. E. V. S. Maciel, N. G. Pereira dos Santos, D. A. Vargas Medina and F. M. Lanças, Electron ionization mass spectrometry: Quo vadis?, Electrophoresis, 2022, 43(15), 1587–1600.
21. J. Jiang, Y. Wang, K. Hou, L. Hua, P. Chen and W. Liu, et al., Photoionization-Generated Dibromomethane Cation Chemical Ionization Source for Time-of-Flight Mass Spectrometry and Its Application on Sensitive Detection of Volatile Sulfur Compounds, Anal. Chem., 2016, 88(10), 5028–5032.
22. M. Dang, M. Li, J. Li, R. Liu, Y. Guo and K. Hou, Emissions of Formamide and Ammonia from Foam Mats: Online Measurement Based on Dopant-Assisted Photoionization TOFMS and Assessment of Their Exposure for Children, Environ. Sci. Technol., 2023, 57(14), 5646–5654.
23. A. Neumann, O. Tiemann, H. J. Hansen, C. P. Rueger and R. Zimmermann, Detailed Comparison of Xenon APPI (9.6/8.4 eV), Krypton APPI (10.6/10.0 eV), APCI, and APLI (266 nm) for Gas Chromatography High Resolution Mass Spectrometry of Standards and Complex Mixtures, J. Am. Soc. Mass Spectrom., 2023, 34(8), 1632–1646.
24. Q. Liang, Z. Ma, X. Bao, W. Xu, Q. Zhang and X. Zou, et al., Imaging VOC Distribution and Tracing Emission Sources in Surface Water by a Mobile Shipborne Spray Inlet Proton Transfer Reaction Mass Spectrometry, J. Cleaner Prod., 2023, 414, 137655.
25. P. G. Harch and S. Rhodes, Acute and Chronic Central Nervous System Oxidative Stress/Toxicity during Hyperbaric Oxygen Treatment of Subacute and Chronic Neurological Conditions, Front. Neurol., 2024, 15, 11.
26. E. P. Manning, Central Nervous System Oxygen Toxicity and Hyperbaric Oxygen Seizures, Aerosp. Med. Hum. Perform., 2016, 87(5), 477–486.
27. K. Jomova, R. Raptova, S. Y. Alomar, S. H. Alwasel, E. Nepovimova and K. Kuca, et al., Reactive Oxygen Species, Toxicity, Oxidative Stress, and Antioxidants: Chronic Diseases and Aging, Arch. Toxicol., 2023, 97(10), 2499–2574.
28. K. Azuma, N. Kagi, U. Yanagi and H. Osawa, Effects of Low-Level Inhalation Exposure to Carbon Dioxide in Indoor Environments: A Short Review on Human Health and Psychomotor Performance, Environ. Int., 2018, 121, 51–56.
29. J. Maniscalco, F. Hoffmeyer, C. Monse, B. Jettkant, E. Marek and T. Bruening, et al., Physiological Responses, Self-reported Health Effects, and Cognitive Performance during Exposure to Carbon Dioxide at 20 000 ppm, Indoor Air, 2022, 32(1), e12939.
30. R. van der Schrier, M. van Velzen, M. Roozekrans, E. Sarton, E. Olofsen and M. Niesters, et al., Carbon Dioxide Tolerability and Toxicity in Rat and Man: A Translational Study, Front. Toxicol., 2022, 4, 1001709.
31. P. Halpern, Y. Raskin, P. Sorkine and A. Oganezov, Exposure to Extremely High Concentrations of Carbon Dioxide: A Clinical Description of a Mass Casualty Incident, Ann. Emerg. Med., 2004, 43(2), 196–199.
32. IMCA, Minimum Quantities of Gas Required Offshore, International Marine Contractors Association, 2020, 10.
33. S. J. Troutman and P. Webb, Instrument for Continuous Measurement of O2 Consumption and CO2 Production of Men in Hyperbaric Chambers, J. Biomech. Eng., 1978, 100, 1–6.

---

Footnotes

† Electronic supplementary information (ESI) available. See DOI: https://doi.org/10.1039/d5an00132c

‡ These authors are contributed equally to this work.

---