Skip to main content
NCBI home page
ACS Omega logoLink to ACS Omega
. 2020 Sep 30;5(40):25704–25711. doi: 10.1021/acsomega.0c02985

Measurement Device for Ambient Carbonyl Sulfide by Means of Catalytic Reduction Followed by Wet Scrubbing/Fluorescence Detection

Kodai Abe †, Koki Shimohira †, Yusuke Miki ‡, Yasuo Hirose ‡, Shin-Ichi Ohira †, Kei Toda †,*
PMCID: PMC7557214  PMID: 33073096

Abstract

graphic file with name ao0c02985_0009.jpg

A portable chemical analysis system for monitoring ambient carbonyl sulfide (COS) was investigated for the first time. COS is paid attention to from the perspectives of photosynthesis tracer, breath diagnosis marker, and new process-use in the manufacture of semiconductors. Recently, the threshold level value of COS was settled at 5 ppm in volume ratio (ppmv) for workplace safety management. In this work, COS was converted to H2S by a small column packed with alumina catalyzer at 65 °C. Then, the H2S produced was collected in a small channel scrubber to react with fluorescein mercuric acetate (FMA), and the resulting fluorescence quenching was monitored using an LED/photodiode-based miniature detector. The miniature channel scrubber was re-examined to determine its robustness and easy fabrication, and conditions of the catalyzer were optimized. When the FMA concentration used was 1 μM, the limit of detection and dynamic range, which were both proportional to the FMA concentration, were 0.07 and 25 ppbv, respectively. Ambient COS in the background level and even contaminated COS in the nitrogen gas cylinder could be detected. If necessary, H2S was removed selectively by reproducible adsorbent columns. COS concentrations of engine exhaust were measured by the proposed method and by cryo-trap-gas chromatography-flame photometric detection, and the results obtained (0.5–5.9 ppbv) by the two methods agreed well (R2 = 0.945, n = 19). COS in ambient air and exhaust gases was successfully measured without any batchwise pretreatment.

Introduction

Carbonyl sulfide (COS) is the most abundant volatile sulfur compound in the atmosphere, and its background level is in the range 0.51–0.53 part per billion in volume ratio (ppbv).1 Its lifetime in the atmosphere, 7 years, is much longer than those of other volatile sulfur compounds (dimethyl sulfide (DMS): 0.5 days, H2S and SO2: 2 days), and it is retained in high concentrations in the atmosphere.2,3 Concentrations of COS in various environments are summarized in Table S1. COS is naturally produced from sediments4 and plants,5,6 especially in coastal areas7−12 and lakes.13,14 Volcanos are geological sources of COS.15−18 Indirect production also includes oxidative conversion from DMS and CS2.19 Anthropogenic emissions20,21 from the pulp, chemical fiber, pigment, aluminum and gas industries, agriculture, biomass burning,22 biofuel combustion, and food processing/waste can also produce COS.23 The elevated tropospheric COS levels impact the stratospheric aerosol layer.24 COS is recognized as a powerful tracer of ecosystem CO2 exchange because COS is taken up by vegetation and hydrolyzed by carbonic anhydrase.25,26 Also, COS in exhaled breath is a potential biomarker for different disease pathologies, such as liver disease,27 organ rejection,28 and cystic fibrosis.29,30 Recently, COS has been recognized as an important process gas for the manufacturing of advanced semiconductors to prepare conducting carbon films for the deep trench walls in oxygen plasma-assisted chemical vapor deposition.31 However, the toxicity of COS has become an issue in this decade.32−34 For example, inhalation exposure of COS produces altered brainstem auditory evoked responses. EPA has determined a threshold limit value of 5 ppmv for workplace COS.35 According to the recent industrial use and revealed toxicity of COS, a monitoring method for COS is urgently needed.

Nevertheless, a convenient and inexpensive monitoring method has not been developed for COS (see Table S2 for reported methods). Conventionally, gas samples collected in Tedlar bags are brought back to the laboratory to be analyzed offsite by cryo-trapping/thermal desorption followed by gas chromatography-flame photometric detection (CryT-GC-FPD). The CryT-GC-FPD method is troublesome and requires skilled analysts and long analysis time, thus it is not suited for onsite continuous monitoring. Recently, a quantum cascade laser was employed for long-path absorption spectroscopy (QCLS), cavity ring-down spectroscopy (CRDS), and off-axis integrated cavity output spectroscopy (OA-ICOS) to monitor COS, and the estimated limits of detection (LODs) were reported as 25036 or 0.185 ppbv.37 Kooijmans et al. evaluated a QCLS for COS measurement at an ambient level of around 0.5 ppbv.38 OA-ICOS has been recently developed to continuously measure dissolved COS in seawater with the use of an equilibrator.39,40 These ‘physical (or optical)’ methods provide continuous measurement with a special laser light source and a precious optical system for aligning kilometers of the optical axis. In this work, a chemical approach was employed to achieve a low-cost and sensitive system for COS monitoring.

COS itself is inactive and difficult to measure chemically. However, COS can be converted to H2S by catalytic hydrolysis. Zhao et al. investigated Zn–Ni–Fe hydrotalcite and Zn–Al hydrotalcite as catalysts.41−43 Hydrolysis with the former catalyst was performed at a high temperature of 250 °C. With the latter, the reaction proceeded at 50 °C, but oxygen in the air damaged the catalyst in a few hours. Apart from these, effective organic polymer catalysts,44 alumina/titania-based catalysts,45 and liquid-phase hydrolysis46 were also recently reported. We have found that the conventional alumina catalyst works well for this purpose at a low temperature (<100 °C) with a long lifetime. The produced H2S can be analyzed by wet scrubbing, followed by fluorescence detection, as previously demonstrated.47,48 The microsystem for H2S, called micro-gas analysis system (μGAS),49,50 is a highly sensitive portable device and could be applied for the measurement of CH3SH,51 HCHO,52,53 and breath NO54 for environmental and medical applications.55,56 However, fabrication of the microchannel scrubber, which is the key device of the μGAS, can be carried out only by skilled personnel. In this work, the channel scrubber was re-examined to make fabrication easier and the device more robust and maintaining the performance to establish a catalytic hydrolysis-miniature gas analysis system (CH-mGAS).

The flow diagram of CH-mGAS is shown in Figure 1a. The device is portable as shown in Figure S1. COS in an air sample was hydrolyzed in a catalysis reactor (CR). The H2S produced was collected at a honeycomb-patterned channel scrubber (HCS) in an alkaline absorbing solution and reacted with fluorescein mercuric acetate (FMA). Quenching of FMA fluorescence was monitored by a miniature fluorescence detector (μFD). If interference from ambient H2S must be considered, a H2S remover (Figure 1b) was placed at the sample inlet. See the Experimental Section for details of CH-mGAS.

Figure 1.

Figure 1

Open in a new tab

Flow diagram of the COS monitoring system. (a) Main system of CH-mGAS for COS measurement. (b) Selective H2S removal system as an option used when H2S interference might occur. RB: reagent bottle, μRP: micro-ring pump, DT: flow dumping tubes, HCS: honeycomb-patterned channel scrubber, μFD: microfluorescence detector, HB: humidifier bottle, CR: catalytic reactor, 3SV: three-way solenoid valve, SR: sulfide removal column, MT: mist trap, FS: airflow sensor, μNV: microneedle valve, μAP: miniature air pump, F: line filter, HSR: hydrogen sulfide remover, and SD: silica gel dryer column.

Results and Discussion

Miniature Channel Scrubber

We have previously developed a microchannel scrubber to collect water-soluble analyte gases.49−56 Fabrication of microchannels is advantageous for gas scrubbing. Similar approaches have been carried out by Guglielmino et al. for absorbance detection and Ueda et al. and Becker et al. for fluorescence detection of gaseous HCHO.57−59 An effective enrichment of the analyte gas was achieved by the microchannel scrubber, which could have a high ratio of the gas absorbing area to absorbing solution volume, and a small liquid volume of the scrubber makes the system response faster. However, the previous microchannel scrubber made of poly(dimethylsiloxane) (PDMS)50 was not robust for long operations. This was a serious disadvantage for field use. A gas-permeable membrane must be pasted strongly on the PDMS microchannel plate without adhesive; otherwise, the membrane could sometimes peel off from the PDMS plate due to the liquid flow pressure. If adhesive is used, it could spread out and clog the microchannel. Therefore, the device should be fabricated only by skilled personnel. To overcome this problem, new scrubbers were designed (Figures S2 and S3), so that the channels for absorbing solution could be fabricated by machining with a miniature end mill bit on a transparent poly(vinyl chloride) (PVC) plate. The gas absorbing channels were arranged in a honeycomb pattern, but the width and depth of the hexagon channels were 1.0 and 0.25 mm, while they were 0.22 and 0.06 mm in the previous device, respectively (Table 1, Figure S2). Though such an enlargement of the liquid channels might be disadvantageous due to the lowering of the absorbing-area-to-absorbing-solution volume ratio, higher collection efficiencies were obtained and comparable characteristics were achieved with the end-mill-machined device. This is because the pattern of gas channels was made the same as that of the liquid channels, while the gas channel of the previous PDMS device was a single large channel (Figure S3). Between the two newly developed devices, the bigger one had a better collection efficiency but the response speed was worse (Table 1); hence, the smaller PVC device was chosen as the best scrubber device.

Table 1. Comparison of the Channel Scrubbers.

  channel scrubber device
PDMSa PVC small PVC big
number of hexagonal unit 499 44 68
liquid channel      
width (mm) 0.22 1.0 1.0
depth (mm) 0.06 0.25 0.25
length (mm) 0.60 2.0 2.0
volume (μL) 13 86 116
gas absorbing area (mm2) 210 340 460
gas channel single honeycombb
width (mm) 14 1.0 1.0
depth (mm) 3.0 1.5 1.5
liquid flow rate (mL/min) 0.12 0.20 0.20
response (μM/H2S ppbv) 0.060 0.037 0.046
collection efficiency (%) 35 68 85
response T90% up (min)c 1.8 1.4 2.4
down (min) 1.4 1.3 1.3
Open in a new tab
a

The same device as in ref (53) was fabricated and examined in this study.

b

Gas channels of the PVC devices were honeycomb-patterned, same as the liquid channel. See Figure S2 for the design of the three scrubbers.

c

T90% up and down is 90% response time when the H2S concentration was changed from 0 to 50 ppbv and from 50 to 0 ppbv, respectively.

Clear PVC is a good material for the plates of the device because of its transparency and durability to reagents. However, the system that comprised the PVC scrubber did not provide response signals to H2S at low concentrations (<1 ppbv) probably because Pb additives in the material interacted with H2S. This problem was solved by the silanization of the gas channels. Not only did the dead zone in the calibration curve disappear but also the slope of the calibration curve became 1.8 times greater after the inactivation of the channel device.

Optimizations of Hydrolysis

The conditions for COS hydrolysis were examined. We succeeded in converting COS to H2S with the Al2O3 catalyst at a low temperature, as shown in the reaction in eq 1.

graphic file with name ao0c02985_m001.jpg 1

High amounts of catalyst and a high catalysis temperature resulted in poor recovery, as shown in Figure 2a,b. This was probably due to the fact that the produced H2S was decomposed on the catalyst.

Figure 2.

Figure 2

Open in a new tab

Optimizations of catalytic hydrolysis for (a) the amount of catalyst, (b) catalysis column temperature, and (c) sample airflow rate. Conversion efficiencies were obtained from response intensities for 100 ppbv COS with the catalyst and 50 ppbv H2S without the catalyst (200 ppbv COS was used when >10 g of the catalyst was packed in the column in (a) because of small conversion efficiency). The sample gas flow rate used was constant at 0.4 L/min in experiments (a, b). The amount of the catalyst packed in the column was 4 g in (b, c) and the catalyst temperature was 65 °C in (a, c). The humidifier temperature was set at 25 °C in all experiments.

Figure 2c shows response intensities for 50 ppbv H2S without the catalyzer and 100 ppbv COS with the catalyzer in blue and green plots, respectively, obtained by changing flow rates of the test gases. The H2S signal increased with an increase in the gas flow because the supply rate of the H2S analyte was proportional to the flow rate. But the curve bends downward because diffusion through the PTFE membrane limited the supply of H2S into the scrubbing solution. On the other hand, the signal for COS was constant above 0.4 L/min. This was because the conversion rate of the catalyst gradually became slow with the increase in the flow rate at this flow range (red curve). The hydrolysis conversion efficiency was maximum at the sampling rate of 0.4 L/min. From these results, we decided to use 4 g of the catalyst at 65 °C with 0.4 L/min sample flow rate as optimum parameters. In these conditions, the water vapor effect was examined by changing the temperature of the water vapor equilibrator. While excess water vapor (45 °C) decreased the response intensity again, the efficiency of conversion to H2S was almost the same between 15 and 35 °C. The conversion rate of COS to H2S increases with the increase in water vapor concentration,60 but too much water vapor clogs the micropore reaction sites by water deposition and decreases the hydrolysis rate.61 Therefore, the water bubbler was maintained at 25 °C or was used without temperature control for further experiments.

Analytical Performance of the Device

The fluorescence signal of FMA was quenched by the reaction of FMA with dissolved H2S, which was converted from COS. Thus, the fluorescence signal decreased with the introduction of COS. Figure 3a shows the response chart obtained in optimized conditions for 0–125 ppbv COS, triplicated for each concentration. The FMA concentration used was 5 μM. By switching a three-way solenoid valve (3SV in Figure 1a), baseline and response were alternatively obtained for 6 and 4 min, respectively, so that response peaks were obtained every 10 min. Good baseline stability and repeatable signals were obtained. A linear calibration curve was obtained from the results (R2 = 0.9984). The limit of detection (LOD), obtained from three times the noise and the calibration curve, was 0.32 ppbv for COS. The dynamic range could be enlarged by increasing the FMA concentration of the absorbing solution. On the contrary, a higher sensitivity could be achieved with a lower FMA concentration. With 1 μM FMA, for example, the LOD obtained was 0.07 ppbv, and the calibration curve was linear up to the dynamic range of 25 ppbv (R2 = 0.9998). The numbers in Figure 3 indicate concentrations of COS added in the test air. Even without COS mixing, obvious responses were observed as shown in the inset of Figure 3b. These responses were from COS existing in the experimental room. The air was used after passing through a Peltier dryer, silica gel/sodalime/active carbon columns, but COS was not removed by such conventional purifiers. Surprisingly, responses were obtained even for nitrogen supplied from a gas cylinder. COS removal is extremely difficult, and the proposed system can be used for quality control of gas manufacturing.

Figure 3.

Figure 3

Open in a new tab

Responses of CH-mGAS. (a) Responses obtained using 5 μM FMA absorbing/reaction solution for 0–125 ppbv COS, (b) high sensitivity mode using 1 μM FMA for 0–25 ppbv COS, and (c) wide range mode using 20 μM FMA and a short annular wet denuder for 0–7.5 ppmv. The test gases were introduced into the device to obtain peaks every 10 min, triplicated for each concentration. The numbers shown in the two panels are concentrations of COS added in the purified air. Enlarged response peaks in the inset of (b) were for COS contained in the purified air and nitrogen cylinder gas. B: Baseline obtained with keeping the solenoid valve (3SV in Figure 1a) on to remove the produced H2S.

Wide range monitoring was performed by replacing the channel scrubber and 5 μM FMA solution with a short annular wet denuder48 and 20 μM FMA solution. COS in ppmv levels was successfully monitored, as shown in Figure 3c, with a dynamic range of 7.5 ppmv (R2 = 0.9965). The dynamic range for COS can be enlarged by an increase in the FMA concentration and a decrease in the collection efficiency of the scrubber.

Reliability was examined for real applications. See Figure S4 in the Supporting Information. In 24 h of continuous monitoring, the relative standard deviation (RSD) was 3.3% for the response to 50 ppbv COS, which was obtained every 10 min (n = 144). The RSD for COS was a little lower than that for H2S (0.58%) but practically acceptable for continuous 24 h monitoring.

Response to 10 ppbv H2S was examined in a thermocontrolled chamber to examine the temperature effect on mGAS. The baseline raw signal (signal at a concentration of 0 ppbv) was affected by ambient temperature (≈−0.35 ppbv/°C between 5 and 45 °C). However, the baseline drift with temperature change was perfectly compensated by having a zero/measurement alternative mode (Figure S5). The span was a little affected by the ambient temperature, but the relative error was was corresponding to −0.9 ppbv at 15 °C and +0.0 ppbv at 35 °C for 10 ppbv H2S calibrated at 25 °C. The COS hydrolysis column was maintained at 65 °C for conversion to H2S. Thus, the proposed system works well under normal measurement conditions.

Elimination of Interference

We tested other volatile sulfur compounds that may interfere with the analysis. The tested volatile sulfur gases were 200 ppbv of H2S and 1000 ppbv of CH3SH, SO2, DMS, dimethyl disulfide (DMDS), and CS2. H2S and CH3SH had a 91 and 20% response to the same concentration of COS, respectively. H2S introduced in the analytical device was partially decomposed in the catalyzer, and the response intensity for H2S was slightly smaller than that for COS. The collection efficiency and FMA reaction rate of CH3SH were lower, and the response to CH3SH was smaller than that of H2S.48 The other sulfur gases did not show interference at all (Table S1).

We examined molecular sieve columns and alkaline absorbing solutions to remove H2S from the sample air. A bubbler was incorporated to supply water vapor for COS hydrolysis. An alkaline-added bubbler would work as a H2S absorber as well as a water vapor generator and effectively remove H2S. However, the removal efficiency dropped after 2 h of introducing 100 ppbv H2S (Figure 4). Although the lifetime doubled when the volume of the absorbing solution (0.1 M NaOH) was increased to 40 mL, the alkaline solution must be renewed every 4 h. The molecular sieve too removed H2S well. Finer powders (80/100 mesh) of the molecular sieve had longer lifetimes than the coarse adsorbent (1/8″), as shown in Figure 4. The molecular sieve column quantitatively removed 100 ppbv H2S for 3 h and 100 ppbv CH3SH for 2 h.

Figure 4.

Figure 4

Open in a new tab

Lifetime of the H2S removers. An impinger bottle containing 20 or 40 mL of 0.1 M NaOH, molecular sieve MS-3A 1/8″ pellets, and 80/100 mesh granular were examined to monitor the removal efficiency by introducing 100 ppbv H2S. The granular molecular sieve was examined with 100 ppbv CH3SH (solid plots) in addition to H2S.

The molecular sieve column was reproducible. Two molecular sieve columns were connected in parallel at the sample inlet and used to remove H2S alternatively (Figure 1b). One of the columns not in use was heated at 200 °C with waste air flowing out from the sampling air pump. The H2S removal system could be used semipermanently by switching every 2.5 h with 2 h of heating and 0.5 h of cooling. The solid-state H2S remover did not interfere with COS analysis, and a good response was obtained for COS, as shown in Figure 5. Responses for the mixture of 50 ppbv COS, 500 ppbv H2S, and 500 ppbv SO2 were the same as those for 50 ppbv COS without H2S and SO2. Continuous and selective COS detection was succeeded by the reproducible H2S removal system.

Figure 5.

Figure 5

Open in a new tab

Responses of CH-mGAS with H2S remover obtained together using an ion molecule reaction-mass spectrometer (IMR-MS) and a UV-fluorescence (UV-FL)-based SO2 monitor. A mixture of COS, H2S, and SO2 and individual COS, SO2, and H2S were introduced into the measurement system. The concentration of COS was 50 ppbv and concentrations of H2S and SO2 were both 500 ppbv. The absorbing/reagent solution used was 5 μM FMA in 0.1 M NaOH.

Analysis of COS Emitted from Motorcycles

The developed system was evaluated for exhaust gases emitted from bicycle engines. Direct and continuous analysis of the exhaust by CH-mGAS together with IMR-MS was first evaluated. The exhaust gas from the bicycle placed outside the laboratory was introduced via a window into the CH-mGAS and ion molecule reaction-mass spectrometer (IMR-MS) placed in the laboratory. COS concentrations obtained by the two methods increased and decreased in a similar manner (Figure 6a). However, the values obtained by IMR-MS were much higher than those obtained by CH-mGAS. We assumed that COS concentrations obtained by IMR-MS were overestimated due to the coexistence of compounds, whose m/z was 60 the same as that of COS, such as CH3COOH produced by combustion. We tried to remove CH3COOH vapor by an alkaline bubbler, and then the values obtained by the two methods were in the same order in the engine acceleration periods. However, the overestimation by IMR-MS was not removed well, and the difference between signals of CH-mGAS and IMR-MS became larger with the time of operation due to the sticky property of the interference species (Figure 6b). After the engine was stopped, the output of CH-mGAS decreased quickly but the IMR-MS output did not reach zero within 40 min of the engine being stopped. This was probably due to the sticky organic acid being adsorbed in the sampling line and retained for a long time. Thus, COS in the exhaust gas cannot be measured by the mass spectrometer.

Figure 6.

Figure 6

Open in a new tab

Continuous measurement results for COS of engine exhaust gases by CH-mGAS and IMR-MS (a) without and (b) with NaOH scrubber to remove CH3COOH.

Next, batchwise comparison with gas chromatography was performed (Figure 7). The exhaust gas was collected in Tedlar bags and then measured by CH-mGAS (with the H2S removal system) and CryT-GC-FPD. The exhaust gas samples contained COS but not in high concentrations (0.5–6 ppbv). According to CryT-GC-FPD analysis, H2S existed in the exhaust, especially in the exhaust from old catalyst bikes, in several to several tens of ppbv, and SO2 was between 10 and 30 ppbv (Table S2). Even in such conditions, the values obtained by the two methods agreed well, with good correlation (R2 = 0.945) and a slope of 0.947, as shown in eq 2.

graphic file with name ao0c02985_m002.jpg 2

Furthermore, the CH-mGAS measured the bag sample in minutes while CryT-GC-FPD required a long analysis time of more than 30 min. Also, the analysis operator needs to prepare a cryo-bath of acetone/dry ice or liquid argon beforehand to perform CryT-GC-FPD.

Figure 7.

Figure 7

Open in a new tab

Comparison of measurement results obtained by CH-mGAS and CryT-GC-FPD. COS was measured for motorcycle exhaust gases and ambient air. The dashed lines show 95% confidence interval.

Conclusions

A new principle for measuring COS is proposed, which is based on online catalytic hydrolysis, followed by scrubbing/fluorescence detection, and an automated system, CH-mGAS, has been developed and characterized. The performance and features of the developed method is compared with those of conventional and reported methods in Table S2. CH-mGAS is the first analytical device to monitor COS chemically. The device shows good performance for COS sensing, and it is also portable and inexpensive. The LOD is dependent on the FMA concentration used, and the LODs are 0.07 and 0.32 ppbv when FMA concentrations are 1 and 5 μM, respectively. The sensitivity of the method with 1 μM FMA is comparable to those of optical methods, such as OA-ICOS, and the sensitivity and selectivity of the device are much higher than those of chemical ionization mass spectrometer simultaneously tested. Though it is difficult to monitor COS at an ambient level with the mass spectrometer due to low concentrations of COS and interference from coexisting compounds, the developed device can detect COS at the ambient level. Interference from H2S can be removed by a molecular sieve column system. Data analyzed by CH-mGAS agree well with data analyzed by CryT-GC-FPD. Also, the device has good repeatability and good selectivity. COS in ambient air and a nitrogen cylinder could be detected. The sensing system has the potential to be used for ambient COS monitoring, breath COS analysis, and quality control of gas manufacturing. Further investigation must be carried out to establish the crucial determination of COS at ambient levels to evaluate applicability in monitoring photosynthetic activity and atmospheric/oceanic COS content with incorporating calibration with 0.5 ppbv standard gas and further improvements of the system. In addition, it may also be applied in an alarm system for safety control in workplaces, such as semiconductor industries where COS is used for plasma-assisted vapor deposition. Wide range monitoring can be performed with FMA solution at high concentrations and with a short annular wet denuder.

Experimental Section

Reagents

COS test gases were prepared by dilution of standard cylinder gas (100 ppm COS balanced with nitrogen, Japan Fine Products, Kawasaki, Japan) with purified air using mass flow controllers. Gas dilutions (uncertainty of the first step dilution accuracy was within 2.7%) were conducted twice to obtain ppbv levels of COS. The gas absorbing solution was 0.1 M NaOH with the addition of 1 or 5 μM FMA. FMA was purchased from Sigma Aldrich (St. Louis, MO). A molecular sieve 3A (pore size 30 μm, 1/8″, Nacalai Tesque, Kyoto, Japan, or 80/100 mesh, Shimadzu, Kyoto, Japan) was used for selective H2S removal. A minicolumn packed with soda lime No.2 (Wako Pure Chemical, Osaka, Japan) and active carbon (Nacalai) was used to remove H2S produced by hydrolysis of COS to obtain a baseline signal. Alumina catalyst NT-9710 (50 m2/g) was obtained from Nissan Girdler Catalyst (Tokyo, Japan). A silanizing agent (L-25, Fuji Systems, Tokyo) was used to make the gas channels inactive.

Analytical System

The flow diagram of the developed system is shown in Figure 1a. All components were installed in a portable box (Figure S1). The air sample was aspirated using a miniature airpump (CM-15-12, Enomoto Micro Pump, Tokyo) and controlled by a flow sensor and a needle valve (PFMV510 and AS1002F, SMC, Tokyo) at a flow rate of 0.4 L/min. The COS air sample was passed through a humidifying bubbler and a catalytic reactor to convert COS to H2S. The produced H2S was collected in an alkaline FMA reagent solution in the honeycomb-patterned channel scrubber. In the case of wide range measurement, a short annular wet denuder,48 comprising a porous polypropylene tube (o.d. 2.7 mm × i.d. 1.8 mm × effective length 10 mm, Accurel PP S6/2, Membrana, Wuppertal, Germany) jacketed in a 1/4″ Teflon tube, was used instead of the channel scrubber. Fluorescence quenching by the absorbed sulfide was monitored by a microfluorescence detector that comprised a blue LED (NSPB500s λmax 460 nm, Nichia, Anan, Japan) and a photodiode (OPT301M, Texas Instruments, Dallas, TX) covered with a plastic long-pass filter (SC-52, Fujifilm, Tokyo, Japan) to cut off light <520 nm.48 The FMA solution flowed through the other channel side by a micro-ring pump (RP-Q1.2N-P20A-DC3V, Aquatech, Osaka, Japan) with a flow dumper made of silicone tubing (o.d. 6 mm × i.d. 4 mm × 30 mm) and a Teflon tubing coil (i.d. 0.3 mm × 1 m) connected in series.

Analysis of COS Emitted from Engines

To validate the developed system, exhaust from bicycle engines were measured by an ion molecule reaction-mass spectrometer (IMR-MS) (Airsense Net, V&F Analyse-und Mestechnik GmbH, Absam, Austria) and CryT-GC-FPD together with the CH-mGAS device. The continuous analysis was performed using a bicycle placed outside the laboratory and the exhaust was aspirated via a window to CH-mGAS and IMR-MS placed in the laboratory. An impinger containing 0.1 M NaOH solution was placed occasionally at the sample inlet to eliminate acidic gases. Also, batchwise analysis was conducted by collecting 3 L of exhaust gas in a Tedlar bag. The bag samples were analyzed by CryT-GC-FPD, IMR-MS, and CH-mGAS. See the Supporting Information for details of CryT-GC-FPD analyses.

Acknowledgments

This work was supported by KAKENHI, Grants-in-Aid for Basic Research (B) (16H04168) and (A) (19H00955), Bilateral Programs Joint Research Projects (open partnership) from the Japan Society for the Promotion of Science, and the A-STEP Regional Value Program (VP29217943622) from Japan Science and Technology Agency.

Supporting Information Available

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acsomega.0c02985.

  • Concentrations of COS in various environments; comparison of COS analytical methods; pictures of CH-mGAS; designs of previous and new channel scrubbers; repeatability test of CH-mGAS; effect of ambient temperature on the measurement results; interference from the other sulfur gases; and comparative analysis with IMR-MS and CryT-GC-FPD (PDF)

The authors declare no competing financial interest.

Supplementary Material

ao0c02985_si_001.pdf (833.4KB, pdf)

References

  1. Kettle A. J.; Kuhn U.; von Hobe M.; Kesselmeier J.; Andreae M. O. Global budget of atmospheric carbonyl sulfide: Temporal and spatial variations of the dominant sources and sinks. J. Geophys. Res.: Atmos. 2002, 107, ACH-25. [Google Scholar]
  2. Lelieveld J.; Roelofs G. J.; Ganzeveld L.; Feichter J.; Rodhe H. Terrestrial sources and distribution of atmospheric sulphur. Philos. Trans. R. Soc. B 1997, 352, 149–158. 10.1098/rstb.1997.0010. [DOI] [Google Scholar]
  3. Seinfeld J. H.; Pandis S. N.. Atmospheric Trace Constituents. In Atmospheric Chemistry and Physics: From Air Pollution to Climate Change, 3rd ed.; Seinfeld J. H.; Pandis S. N., Eds.; Wiley, 2016; pp 18–68. [Google Scholar]
  4. Yang Z.; Kanda K.; Tsuruta H.; Minami K. Measurement of biogenic sulfur gases emission from some Chinese and Japanese soils. Atmos. Environ. 1996, 30, 2399–2405. 10.1016/1352-2310(95)00247-2. [DOI] [Google Scholar]
  5. Stimler K.; Montzka S. A.; Berry J. A.; Rudich Y.; Yakir D. Relationships between carbonyl sulfide (COS) and CO2 during leaf gas exchange. New Phytol. 2010, 186, 869–878. 10.1111/j.1469-8137.2010.03218.x. [DOI] [PubMed] [Google Scholar]
  6. Yi Z.; Wang X. Carbonyl sulfide and dimethyl sulfide fluxes in an urban lawn and adjacent bare soil in Guangzhou, China. J. Environ. Sci. 2011, 23, 784–789. 10.1016/S1001-0742(10)60478-0. [DOI] [PubMed] [Google Scholar]
  7. Watts S. F. The mass budgets of carbonyl sulfide, dimethyl sulfide, carbon disulfide and hydrogen sulfide. Atmos. Environ. 2000, 34, 761–779. 10.1016/S1352-2310(99)00342-8. [DOI] [Google Scholar]
  8. Delaune R. D.; Devai I.; Lindau C. W. Flux of reduced sulfur gases along a salinity gradient in Lousiana coastal marshes. Estuarine, Coastal Shelf Sci. 2002, 54, 1003–1011. 10.1006/ecss.2001.0871. [DOI] [Google Scholar]
  9. Whelan M. E.; Min D.-H.; Rhew R. C. Salt marsh vegetation as a carbonyl sulfide (COS) source to the atmosphere. Atmos. Environ. 2013, 73, 131–137. 10.1016/j.atmosenv.2013.02.048. [DOI] [Google Scholar]
  10. Bodenbender J.; Wassmann R.; Papen H.; Rennenberg H. Temporal and spatial variation of sulfur-gas-transfer between coastal marine sediments and the atmosphere. Atmos. Environ. 1999, 33, 3487–3502. 10.1016/S1352-2310(98)00351-3. [DOI] [Google Scholar]
  11. Devai I.; DeLaune R. D. Formation of volatile sulfur compounds in salt marsh sediment as influenced by soil redox condition. Org. Geochem. 1995, 23, 283–287. 10.1016/0146-6380(95)00024-9. [DOI] [Google Scholar]
  12. Angert A.; Said-Ahmad W.; Davidson C.; Amrani A. Sulfur isotopes ratio of atmospheric carbonyl sulfide constrains its sources. Sci. Rep. 2019, 9, 741 10.1038/s41598-018-37131-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Du Q.; Mu Y.; Zhang C.; Liu J.; Zhang Y. H.; Liu C. Photochemical production of carbonyl sulfide, carbon disulfide and dimethyl sulfide in a lake water. J. Environ. Sci. 2017, 51, 146–156. 10.1016/j.jes.2016.08.006. [DOI] [PubMed] [Google Scholar]
  14. Richards S. R.; Rudd J. W. M.; Kelly C. A. Organic volatile sulfur in lakes ranging in sulfate and dissolved salt concentration over five orders of magnitude. Limnol. Oceanogr. 1994, 39, 562–572. 10.4319/lo.1994.39.3.0562. [DOI] [Google Scholar]
  15. Belviso S.; Nguyen B. C.; Allard P. Estimate of carbonyl sulfide (OCS) volcanic source strength deduced from OCS/CO2 ratios in volcanic gases. Geophys. Res. Lett. 1986, 13, 133–136. 10.1029/GL013i002p00133. [DOI] [Google Scholar]
  16. Muller E.; Philippot P.; Rollion-Bard C.; Cartigny P. Multiple sulfur-isotope signatures in Archean sulfates and their implications for the chemistry and dynamics of the early atmosphere. Proc. Natl. Acad. Sci. U.S.A. 2016, 113, 7432–7437. 10.1073/pnas.1520522113. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Oppenheimer C.; Kyle P.; Eisele F.; Crawford J.; Huey G.; Tanner D.; Kim S. C.; Mauldin L.; Blake D.; Beyersdorf A.; Buhr M.; Davis D. D. Atmospheric chemistry of an Antarctic volcanic plume. J. Geophys. Res.: Atmos. 2010, 115, D04303 10.1029/2009JD011910. [DOI] [Google Scholar]
  18. Vasilakos C.; Maggos T.; Bartzis J. G.; Papagiannakopoulos P. Determination of atmospheric sulfur compounds near a volcanic area in Greece. J. Atmos. Chem. 2005, 52, 101–116. 10.1007/s10874-005-8853-5. [DOI] [Google Scholar]
  19. Modiri Gharehveran M.; Shah A. D. Indirect photochemical formation of carbonyl sulfide and carbon disulfide in natural waters: role of organic sulfur precursors, water quality constituents, and temperature. Environ. Sci. Technol. 2018, 52, 9108–9117. 10.1021/acs.est.8b01618. [DOI] [PubMed] [Google Scholar]
  20. Lee C.-L.; Brimblecombe P. Anthropogenic contributions to global carbonyl sulfide, carbon disulfide and organosulfides fluxes. Earth-Sci. Rev. 2016, 160, 1–18. 10.1016/j.earscirev.2016.06.005. [DOI] [Google Scholar]
  21. Cheng Y.; Zhang C.; Zhang Y. J.; Zhang H.; Sun X.; Mu Y. Characteristics and anthropogenic sources of carbonyl sulfide in Beijing. J. Environ. Sci. 2015, 28, 163–170. 10.1016/j.jes.2014.05.052. [DOI] [PubMed] [Google Scholar]
  22. Notholt J.; Kuang Z.; Rinsland C. P.; Toon G. C.; Rex M.; Jones N.; Albrecht T.; Deckelmann H.; Krieg J.; Weinzierl C.; Bingemer H.; Weller R.; Schrems O. Enhanced upper tropical tropospheric COS: impact on the stratospheric aerosol layer. Science 2003, 300, 307–310. 10.1126/science.1080320. [DOI] [PubMed] [Google Scholar]
  23. Wu T.; Wang X.; Li D.; Yi Z. Emission of volatile organic sulfur compounds (VOSCs) during aerobic decomposition of food wastes. Atmos. Environ. 2010, 44, 5065–5071. 10.1016/j.atmosenv.2010.09.019. [DOI] [Google Scholar]
  24. Brühl C.; Lelieveld J.; Crutzen P. J.; Tost H. The role of carbonyl sulphide as a source of stratospheric sulphate aerosol and its impact on climate. Atmos. Chem. Phys. 2012, 12, 1239–1253. 10.5194/acp-12-1239-2012. [DOI] [Google Scholar]
  25. Yang F.; Qubaja R.; Tatarinov F.; Stern R.; Yakir D. Soil–atmosphere exchange of carbonyl sulfide in a Mediterranean citrus orchard. Atmos. Chem. Phys. 2019, 19, 3873–3883. 10.5194/acp-19-3873-2019. [DOI] [Google Scholar]
  26. Kitz F.; Gomez-Brandon M.; Eder B.; Etemadi M.; Spielmann F. M.; Hammerle A.; Insam H.; Wohlfahrt G. Soil carbonyl sulfide exchange in relation to microbial community composition: insights from a managed grassland soil amendment experiment. Soil Biol. Biochem. 2019, 135, 28–37. 10.1016/j.soilbio.2019.04.005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Sehnert S. S.; Jiang L.; Burdick J. F.; Risby T. H. Breath biomarkers for detection of human liver diseases: preliminary study. Biomarkers 2002, 7, 174–187. 10.1080/13547500110118184. [DOI] [PubMed] [Google Scholar]
  28. Studer S. M.; Orens J. B.; Rosas I.; Krishnan J. A.; Cope K. A.; Yang S.; Conte J. V.; Becker P. B.; Risby T. H. Patterns and significance of exhaled-breath biomarkers in lung transplant recipients with acute allograft rejection. J. Heart Lung Transplantation 2001, 20, 1158–1166. 10.1016/S1053-2498(01)00343-6. [DOI] [PubMed] [Google Scholar]
  29. Kamboures M. A.; Blake D. R.; Cooper D. M.; Newcomb R. L.; Barker M.; Larson J. K.; Meinardi S.; Nussbaum E.; Rowland F. S. Breath sulfides and pulmonary function in cystic fibrosis. Proc. Natl. Acad. Sci. U.S.A. 2005, 102, 15762–15767. 10.1073/pnas.0507263102. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Steiger A. K.; Zhao Y.; Pluth M. D. Emerging roles of carbonyl sulfide in chemical biology: sulfide transporter or gasotransmitter?. Antioxid. Redox Signaling 2018, 28, 1516–1532. 10.1089/ars.2017.7119. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Kim J. K.; Cho S. I.; Kim N. G.; Jhon M. S.; Min K. S.; Kim C. K.; Yeom G. Y. Study on the etching characteristics of amorphous carbon layer in oxygen plasma with carbonyl sulfide. J. Vac. Sci. Technol., A 2013, 31, 021301 10.1116/1.4780122. [DOI] [Google Scholar]
  32. Bartholomaeus A. R.; Haritos V. S. Review of the toxicology of carbonyl sulfide, a new grain fumigant. Food Chem. Toxicol. 2005, 43, 1687–1701. 10.1016/j.fct.2005.06.016. [DOI] [PubMed] [Google Scholar]
  33. Haritos V. S.; Dojchinov G. Carbonic anhydrase metabolism is a key factor in the toxicity of CO2 and COS but not CS2 toward the flour beetle Tribolium castaneum [Coleoptera: Tenebrionidae]. Comp. Biochem. Physiol., Part C: Toxicol. Pharmacol. 2005, 140, 139–147. 10.1016/j.cca.2005.01.012. [DOI] [PubMed] [Google Scholar]
  34. Herr D. W.; Graff J. E.; Moser V. C.; Crofton K. M.; Little P. B.; Morgan D. L.; Sills R. C. Inhalational exposure to carbonyl sulfide produces altered brainstem auditory and somatosensory-evoked potentials in Fischer 344N rats. Toxicol. Sci. 2007, 95, 118–135. 10.1093/toxsci/kfl146. [DOI] [PubMed] [Google Scholar]
  35. EPA, U. S., Provisional peer-reviewed toxicity values for carbonyl sulfide (CASRN 463-58-1). Superfund Health Risk Technical Support Center, N. C. f. E. A., Office of Research and Development, Ed. Cincinnati, OH, 2015.
  36. Bielecki Z.; Stacewicz T.; Wojtas J.; Mikołajczyk J. Application of quantum cascade lasers to trace gas detection. Bull. Pol. Acad. Sci.: Tech. Sci. 2015, 63, 515–525. 10.1515/bpasts-2015-0059. [DOI] [Google Scholar]
  37. Banik G. D.; Maity A.; Som S.; Pal M.; Pradhan M. An external-cavity quantum cascade laser operating near 5.2 μm combined with cavity ring-down spectroscopy for multi-component chemical sensing. Laser Phys. 2018, 28, 045701 10.1088/1555-6611/aaaa55. [DOI] [Google Scholar]
  38. Kooijmans L. M. J.; Uitslag N. A. M.; Zahniser M. S.; Nelson D. D.; Montzka S. A.; Chen H. Continuous and high-precision atmospheric concentration measurements of COS, CO2, CO and H2O using a quantum cascade laser spectrometer (QCLS). Atmos. Meas. Tech. 2016, 9, 5293–5314. 10.5194/amt-9-5293-2016. [DOI] [Google Scholar]
  39. Lennartz S. T.; Marandino C. A.; von Hobe M.; Andreae M. O.; Aranami K.; Atlas E.; Berkelhammer M.; Bingemer H.; Booge D.; Cutter G.; Cortes P.; Kremser S.; Law C. S.; Marriner A.; Simó R.; Quack B.; Uher G.; Xie H.; Xu X. Marine carbonyl sulfide (OCS) and carbon disulfide (CS2): a compilation of measurements in seawater and the marine boundary layer. Earth Syst. Sci. Data 2020, 12, 591–609. 10.5194/essd-12-591-2020. [DOI] [Google Scholar]
  40. Lennartz S. T.; Marandino C. A.; von Hobe M.; Cortes P.; Quack B.; Simo R.; Booge D.; Pozzer A.; Steinhoff T.; Arevalo-Martinez D. L.; Kloss C.; Bracher A.; Röttgers R.; Atlas E.; Krüger K. Direct oceanic emissions unlikely to account for the missing source of atmospheric carbonyl sulfide. Atmos. Chem. Phys. 2017, 17, 385–402. 10.5194/acp-17-385-2017. [DOI] [Google Scholar]
  41. Zhao S.; Yi H.; Tang X.; Jiang S.; Gao F.; Zhang B.; Zuo Y.; Wang Z. Q. The hydrolysis of carbonyl sulfide at Low temperature: a review. Sci. World J. 2013, 2013, 1–8. 10.1155/2013/739501. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Zhao S.; Yi H.; Tang X.; Kang D.; Wang H.; Li K.; Duan K. Characterization of Zn–Ni–Fe hydrotalcite-derived oxides and their application in the hydrolysis of carbonyl sulfide. Appl. Clay Sci. 2012, 56, 84–89. 10.1016/j.clay.2011.11.026. [DOI] [Google Scholar]
  43. Zhao S.; Yi H.; Tang X.; Song C. Low temperature hydrolysis of carbonyl sulfide using Zn–Al hydrotalcite-derived catalysts. Chem. Eng. J. 2013, 226, 161–165. 10.1016/j.cej.2013.04.039. [DOI] [Google Scholar]
  44. Mi J.; Liu F.; Chen W.; Chen X.; Shen L.; Cao Y.; Au C.; Huang K.; Zheng A.; Jiang L. Design of efficient, hierarchical porous polymers endowed with tunable structural base sites for direct catalytic elimination of COS and H2S. ACS Appl. Mater. Interfaces 2019, 11, 29950–29959. 10.1021/acsami.9b09149. [DOI] [PubMed] [Google Scholar]
  45. He E.; Huang G.; Fan H.; Yang C.; Wang H.; Tian Z.; Wang L.; Zhao Y. Macroporous alumina- and titania-based catalyst for carbonyl sulfide hydrolysis at ambient temperature. Fuel 2019, 246, 277–284. 10.1016/j.fuel.2019.02.097. [DOI] [Google Scholar]
  46. Palma V.; Vaiano V.; Barba D.; Colozzi M.; Palo E.; Barbato L.; Cortese S.; Miccio M. Study of the carbonyl sulphide hydrolysis reaction in liquid phase. Chem. Eng. Trans. 2019, 73, 247–252. [Google Scholar]
  47. Toda K.; Dasgupta P. K.; Li J.; Tarver G. A.; Zarus G. M. Fluorometric field instrument for continuous measurement of atmospheric hydrogen sulfide. Anal. Chem. 2001, 73, 5716–5724. 10.1021/ac010602g. [DOI] [PubMed] [Google Scholar]
  48. Toda K.; Ohira S.-I.; Tanaka T.; Nishimura T.; Dasgupta P. K. Field instrument for simultaneous large dynamic range measurement of atmospheric hydrogen sulfide, methanethiol, and sulfur dioxide. Environ. Sci. Technol. 2004, 38, 1529–1536. 10.1021/es034450d. [DOI] [PubMed] [Google Scholar]
  49. Toda K.; Ohira S.-I.; Ikeda M. Micro-gas analysis system μGAS comprising a microchannel scrubber and a micro-fluorescence detector for measurement of hydrogen sulfide. Anal. Chim. Acta 2004, 511, 3–10. 10.1016/j.aca.2004.01.040. [DOI] [Google Scholar]
  50. Ohira S.-I.; Toda K. Micro gas analysis system for measurement of atmospheric hydrogen sulfide and sulfur dioxide. Lab Chip 2005, 5, 1374–1379. 10.1039/b511281h. [DOI] [PubMed] [Google Scholar]
  51. Toda K.; Kuwahara H.; Kajiwara H.; Hirota K.; Ohira S.-I. Gas analyzer for continuous monitoring of trace level methanethiol by microchannel collection and fluorescence detection. Anal. Chim. Acta 2014, 841, 1–9. 10.1016/j.aca.2014.06.019. [DOI] [PubMed] [Google Scholar]
  52. Toda K.; Tokunaga W.; Gushiken Y.; Hirota K.; Nose T.; Suda D.; Nagai J.; Ohira S.-I. Mobile monitoring along a street canyon and stationary forest air monitoring of formaldehyde by means of a micro-gas analysis system. J. Environ. Monit. 2012, 14, 1462–1472. 10.1039/c2em10935b. [DOI] [PubMed] [Google Scholar]
  53. Yanaga A.; Hozumi N.; Ohira S.-I.; Hasegawa A.; Toda K. Formaldehyde vapor produced from hexamethylenetetramine and pesticide: Simultaneous monitoring of formaldehyde and ozone in chamber experiments by flow-based hybrid micro-gas analyzer. Talanta 2016, 148, 649–654. 10.1016/j.talanta.2015.05.060. [DOI] [PubMed] [Google Scholar]
  54. Toda K.; Koga T.; Kosuge J.; Kashiwagi M.; Oguchi H.; Arimoto T. Micro gas analyzer measurement of nitric oxide in breath by direct wet scrubbing and fluorescence detection. Anal. Chem. 2009, 81, 7031–7037. 10.1021/ac901131d. [DOI] [PubMed] [Google Scholar]
  55. Ohira S.-I.; Someya K.; Toda K. In situ gas generation for micro gas analysis system. Anal. Chim. Acta 2007, 588, 147–152. 10.1016/j.aca.2007.01.069. [DOI] [PubMed] [Google Scholar]
  56. Ohira S.-I.; Toda K. Micro gas analyzers for environmental and medical applications. Anal. Chim. Acta 2008, 619, 143–156. 10.1016/j.aca.2008.05.010. [DOI] [PubMed] [Google Scholar]
  57. Guglielmino M.; Allouch A.; Serra C. A.; Calvé S. L. Development of microfluidic analytical method for on-line gaseous formaldehyde detection. Sens. Actuators, B 2017, 243, 963–970. 10.1016/j.snb.2016.11.093. [DOI] [Google Scholar]
  58. Ueda M.; Teshima N.; Sakai T. Development of formaldehyde standard gas generator based on gravitational dispensing-vaporization and its application to breath formaldehyde determination. Bunseki Kagaku 2008, 57, 605–611. 10.2116/bunsekikagaku.57.605. [DOI] [Google Scholar]
  59. Becker A.; Andrikopoulou C.; Bernhardt P.; Ocampo-Torres R.; Trocquet C.; Le Calve S. Development and optimization of an airborne formaldehyde microfluidic analytical device based on passive uptake through a microporous tube. Micromachines 2019, 10, 807. 10.3390/mi10120807. [DOI] [PMC free article] [PubMed] [Google Scholar]
  60. West J.; Williams B. P.; Young N.; Rhodes C.; Hutchings G. J. Low temperature hydrolysis of carbonyl sulfide using γ-alumina catalysts. Catal. Lett. 2001, 74, 111–114. 10.1023/A:1016681030065. [DOI] [Google Scholar]
  61. Akimoto M. Role of reduction sites in vapor-phase hydrolysis of carbonyl sulfide over alumina catalysts. J. Catal. 1980, 62, 84–93. 10.1016/0021-9517(80)90423-6. [DOI] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

ao0c02985_si_001.pdf (833.4KB, pdf)

Articles from ACS Omega are provided here courtesy of American Chemical Society

RESOURCES