Equilibrator Inlet-Proton Transfer Reaction-Mass Spectrometry (EI

Sep 30, 2009 - In principle, dissolved DMS can be analyzed by PTR-MS by extraction of DMS ..... We also wish to thank Dr. Yutaka W. Watanabe of Hokkai...
0 downloads 0 Views 2MB Size
Download PDF
Anal. Chem. 2009, 81, 9021–9026

Equilibrator Inlet-Proton Transfer Reaction-Mass Spectrometry (EI-PTR-MS) for Sensitive, High-Resolution Measurement of Dimethyl Sulfide Dissolved in Seawater Sohiko Kameyama,† Hiroshi Tanimoto,*,† Satoshi Inomata,† Urumu Tsunogai,‡ Atsushi Ooki,† Yoko Yokouchi,† Shigenobu Takeda,§ Hajime Obata,| and Mitsuo Uematsu| National Institute for Environmental Studies, Tsukuba, 305-8506, Japan, Faculty of Science, Hokkaido University, Sapporo, 060-0810, Japan, Graduate School of Agricultural and Life Sciences, The University of Tokyo, Tokyo, 113-8657, Japan, and Ocean Research Institute, The University of Tokyo, Tokyo, 164-8639, Japan We developed an equilibrator inlet-proton transfer reaction-mass spectrometry (EI-PTR-MS) method for fast detection of dimethyl sulfide (DMS) dissolved in seawater. Dissolved DMS extracted by bubbling pure nitrogen through the sample was continuously directed to the PTRMS instrument. The equilibration of DMS between seawater and the carrier gas, and the response time of the system, were evaluated in the laboratory. DMS reached equilibrium with an overall response time of 1 min. The detection limit (50 pmol L-1 at 5 s integration) was sufficient for detection of DMS concentrations in the open ocean. The EI-PTR-MS instrument was deployed during a research cruise in the western North Pacific Ocean. Comparison of the EI-PTR-MS results with results obtained by means of membrane tube equilibrator-gas chromatography/mass spectrometry agreed reasonably well on average (R2 ) 0.99). EI-PTR-MS captured temporal variations of dissolved DMS concentrations, including elevated peaks associated with patches of high biogenic activity. These results demonstrate that the EI-PTR-MS technique was effective for highly time-resolved measurements of DMS in the open ocean. Further measurements will improve our understanding of the biogeochemical mechanisms of the production, consumption, and distribution of DMS on the ocean surface and, hence, the air-sea flux of DMS, which is a climatically important species. Dimethyl sulfide (DMS), a major natural source of atmospheric sulfur, plays an important role in climate regulation.1 Atmospheric DMS is photo-oxidized to form sulfate aerosols, which may affect the radiative budget of the atmosphere by serving as precursors of cloud condensation nuclei. The ocean surface is a major source * To whom correspondence should be addressed. E-mail: [email protected]. Fax: +81-29-850-2579. † National Institute for Environmental Studies. ‡ Hokkaido University. § Graduate School of Agricultural and Life Sciences, The University of Tokyo. | Ocean Research Institute, The University of Tokyo. (1) Charlson, R. J.; Lovelock, J. E.; Andreae, M. O.; Warren, S. G. Nature 1987, 326, 655–661. 10.1021/ac901630h CCC: $40.75  2009 American Chemical Society Published on Web 09/30/2009

of atmospheric DMS.2,3 Investigators have proposed that phytoplankton releases dimethylsulfoniopropionate (DMSP), which is enzymatically cleaved by microorganisms to produce DMS.4,5 Much attention has focused on describing DMS concentrations in the sea surface layer, and several attempts have recently been made to establish empirical relationships between DMS distribution and biological, physical, and chemical variables for a full mechanistic understanding of the processes involved.6-10 However, these studies showed large uncertainties and spatial differences of sea surface DMS concentrations, and therefore, we have not yet arrived at comprehensive understanding of the variability of DMS production on the ocean surface. A purge-and-trap (P&T) method combined with gas chromatographic (GC) detection has been used to measure DMS dissolved in seawater.11 However, this technique requires onboard measurements of discrete seawater samples within 24 h after sampling. Recently, a solid-phase microextraction technique was used for extraction of dissolved DMS12-14 and, more importantly, for long(2) Keller, M. D.; Bellows, W. K.; Guillard, R. R. L. In Biogenic Sulfur in the Environment, Saltzmann, E., Cooper, W. J., Eds.; ACS Symposium Series 393; American Chemical Society: Washington, DC, 1989; pp. 167-182. (3) Liss, P. S.; Hatton, A. D.; Malin, G.; Nightingale, P. D.; Turner, S. M. Philos. Trans. R. Soc., B 1997, 352, 159–168. (4) Stefels, J.; Vanboekel, W. H. M. Mar. Ecol. Pro. Ser. 1993, 97, 11–18. (5) Ledyard, K. M.; Dacey, J. W. H. Mar. Ecol. Pro. Ser. 1994, 110, 95–103. (6) Kettle, A. J.; Andreae, M. O.; Amouroux, D.; Andreae, T. W.; Bates, T. S.; Berresheim, H.; Bingemer, H.; Boniforti, R.; Curran, M. A. J.; DiTullio, G. R.; Helas, G.; Jones, G. B.; Keller, M. D.; Kiene, R. P.; Leck, C.; Levasseur, M.; Malin, G.; Maspero, M.; Matrai, P.; McTaggart, A. R.; Mihalopoulos, N.; Nguyen, B. C.; Novo, A.; Putaud, J. P.; Rapsomanikis, S.; Roberts, G.; Schebeske, G.; Sharma, S.; Simo´, R.; Staubes, R.; Turner, S.; Uher, G. Global Biogeochem. Cycles 1999, 13, 399–444. (7) Anderson, T. R.; Spall, S. A.; Yool, A.; Cipollini, P.; Challenor, P. G.; Fasham, M. J. R. J. Mar. Syst. 2001, 30, 1–20. (8) Simo´, R.; Dachs, J. Global Biogeochem. Cycles. 2002, 16, 10, DOI: 1029/ 2001GB001829. (9) Belviso, S.; Moulin, C.; Bopp, L.; Stefels, J. Can. J. Fish. Aquat. Sci. 2004, 61, 804–816. (10) Watanabe, Y. W.; Yoshinari, H.; Sakamoto, A.; Nakano, Y.; Kasamatsu, N.; Midorikawa, T.; Ono, T. Mar. Chem. 2007, 103, 347–358. (11) Andreae, M. O.; Barnard, W. R. Anal. Chem. 1983, 55, 608–612. (12) Niki, T.; Fujinaga, T.; Watanabe, M. F.; Kinoshita, J. J. Oceanogr. 2004, 60, 913–917. (13) Yassaa, N.; Colomb, A.; Lochte, K.; Peeken, I.; Williams, J. Limnol. Oceanogr. Methods 2006, 4, 374–381. (14) Vogt, M.; Turner, S.; Yassaa, N.; Steinke, M.; Williams, J.; Liss, P. Mar. Chem. 2008, 108, 32–39.

Analytical Chemistry, Vol. 81, No. 21, November 1, 2009

9021

term preservation of DMS, enabling samples to be analyzed in the laboratory rather than at sea.15 Thus this technique could enhance oceanic DMS data sets; however, like the P&T method, it is not a continuous technique. Tortell16 recently developed a membrane inlet mass spectrometry (MIMS) technique for the highly time-resolved measurement of dissolved DMS and made continuous measurements of DMS dissolved in marine continental shelf waters and coastal waters.17,18 These investigators pointed out that previous ship-board surveys may often have underestimated the variability in surface water DMS concentrations, and proposed real-time measurement as a powerful tool to identify hot spots of elevated DMS concentrations. However, the detection limit of their system is 2 nmol L-1, prohibiting MIMS from measurements of DMS in the open ocean.16 Therefore, a highfrequency, high-sensitivity measurement system is needed for detection of DMS in the open ocean to improve our understanding of oceanic production and global and regional distributions of DMS. Proton transfer reaction-mass spectrometry (PTR-MS) allows online measurement of atmospheric volatile organic compounds (VOCs) with high sensitivity (pptv levels) and rapid response time (0.1-10 s).19,20 PTR-MS is potentially advantageous compared to GC because the former does not require sample pretreatment such as dehydration or preconcentration. In principle, dissolved DMS can be analyzed by PTR-MS by extraction of DMS from the liquid phase to the gas phase. Williams et al.21 used a P&T PTR-MS method to measure several VOCs, including DMS, dissolved in surface seawater. To measure dissolved DMS with both high sensitivity and timeresolution, we combined PTR-MS with equilibration of DMS between seawater and air. There are three general types of equilibrators: shower type,22-24 membrane type,25-27 and bubbling type.28,29 A potential weakness of the shower type is that marine organisms might block the holes of the shower-head during longterm use. The membrane type might be subject to adsorptive loss of DMS to the membrane surface depending on the material of membrane. Although the bubbling type generally requires a lot of space on research vessels, it is less subject to interruption of the water stream and to adsorptive loss. Therefore, we use a bubbling-type equilibrator as an inlet for continuous detection of dissolved DMS in seawater. We evaluated the equilibrator inlet (15) Sakamoto, A.; Niki, T.; Watanabe, Y. W. Anal. Chem. 2006, 78, 4593– 4597. (16) Tortell, P. D. Limnol. Oceanogr. Methods 2005, 3, 24–37. (17) Tortell, P. D. Geochem. Geophys. Geosyst. 2005, 6, 10, DOI: 1029/ 2005GC000953. (18) Nemcek, N.; Ianson, D.; Tortell, P. D. Global Biogeochem. Cycles. 2008, 22, 10, DOI: 1029/2006GB002879. (19) Lindinger, W.; Hansel, A.; Jordan, A. Int. J. Mass Spectrom. 1998, 173, 191–241. (20) de Gouw, J.; Warneke, C. Mass Spectrom. Rev. 2007, 26, 223–257. (21) Williams, J.; Holzinger, R.; Gros, V.; Xu, X.; Atlas, E.; Wallace, D. W. R. Geophys. Res. Lett. 2004, 31, 10, DOI: 1029/2004GL020012. (22) Keeling, C. D.; Rakestraw, N. W.; Waterman, L. S. J. Geophys. Res. 1965, 70, 6087-6097. (23) Weiss, R. F. J. Chromatogr. Sci. 1981, 19, 611–616. (24) Butler, J. H.; Elkins, J. W.; Thompson, T. M.; Hall, B. D.; Swanson, T. H.; Koropalov, V. J. Geophys. Res. 1991, 96, 22347–22355. (25) Groszko, W.; Moore, R. M. Chemosphere 1998, 36, 3083–3092. (26) Ooki, A.; Yokouchi, Y. Environ. Sci. Technol. 2008, 42, 5706–5711. (27) Loose, B.; Stute, M.; Alexander, P.; Smethie, W. M. Water Resour. Res. 2009, 45, 10, DOI: 1029/2008WR006969. (28) Takahashi, T. J. Geophys. Res. 1961, 66, 477–494. (29) Frankignoulle, M.; Borges, A.; Biondo, R. Water Res. 2001, 35, 1344–1347.

9022

Analytical Chemistry, Vol. 81, No. 21, November 1, 2009

(EI)-PTR-MS technique in the laboratory, including the extraction efficiency, instrumental response time, and detection limit for DMS. We then deployed the EI-PTR-MS system on a research cruise in the western North Pacific Ocean and compared the resulting DMS data obtained by a GC method. EXPERIMENTAL SECTION PTR-MS Instrument. We used a commercially available PTRMS instrument (PTRMS-FDT-hs, IONICON Analytik GmbH, Innsbruck, Austria).30 Briefly, the PTR-MS instrument consists of a discharge ion source to produce the H3O+ ions; a drift tube, in which the proton-transfer reactions between H3O+ and VOCs including DMS take place and a quadruple mass spectrometer (QMS) for detection of reagent and product ions. In a hollow cathode discharge ion source, H3O+ ions were produced from pure water vapor (flow rate 7.8 sccm). Sample air was introduced into the drift tube at approximately 22 sccm; the drift tube pressure was held at 2.1 mbar. The sampling inlet and the drift tube were held at 105 °C. The drift tube (9.2 cm long) consisted of stainless steel ring electrodes connected to a resistor network, which divided the overall drift voltage (Udrift) into a homogeneously increasing voltage and established a homogeneous electric field inside the drift tube. The electric field was applied along the drift tube to avoid substantial formation of cluster ions, H3O+ · (H2O)n, n ) 1, 2, ... H3O+ + H2O T H3O+·H2O

(1)

H3O+·(H2O)n + H2O T H3O+·(H2O)n+1

(2)

In the drift tube, DMS in the sample gas was ionized by protontransfer reactions, because of the higher proton affinity of DMS (198.6 kcal mol-1) than that of H2O (165.2 kcal mol-1): H3O+ + CH3SCH3 f CH3SCH3·H+ + H2O

(3)

A fraction of the reagent ion, H3O+, and the product ions was extracted through a small orifice into the QMS where they were detected by a secondary electron multiplier operated in the ion pulse counting mode. The mass dependence of the transmission efficiency of the QMS was calibrated by the manufacturer. The field strength, E/N, of the drift tube, where E is the electric field strength (V cm-1) and N is the buffer gas number density (molecules cm-3), was set to 108 Td (Td ) 10-17 cm2 V molecule-1) to reduce fragmentation of the detected VOCs. The source current and U4, U5, Udrift, U1, and UNC of the PTRMS instrument were 6.0 mA and 50, 120, 400, 48, and 5.8 V, respectively. The water vapor concentration in the sample air, [H2O]sample, was determined from the best-fit curve for the plot of [H2O]sample versus the relative intensity of H3O+ · H2O to H3O+ (M37/M19), as shown in Inomata et al.30 The presence of DMS was indicated by ion signals at m/z 63 (CH3SCH3 · H+) with this identification being in accordance with previous studies.19,31 (30) Inomata, S.; Tanimoto, H.; Kameyama, S.; Tsunogai, U.; Irie, H.; Kanaya, Y.; Wang, Z. Atmos. Chem. Phys. 2008, 8, 273–284. (31) Williams, J.; Poschl, U.; Crutzen, P. J.; Hansel, A.; Holzinger, R.; Warneke, C.; Lindinger, W.; Lelieveld, J. J. Atmos. Chem. 2001, 38, 133–166.

Figure 1. Schematic diagram of the EI-PTR-MS system used to measure DMS dissolved in seawater.

The detection sensitivity under dry conditions was determined with a dynamic dilution system consisting of two mass flow controllers (FC-795C, Advanced Energy Japan, Tokyo, Japan). A gas mixture containing DMS (5-100 ppbv) was produced by dynamic dilution of gravimetrically prepared standard gases balanced with nitrogen (5.002 ppmv for DMS, Takachiho Chemical Industrial Co., Tokyo, Japan) with ultrapure N2 gas (>99.99995%, Japan Fine Products Co., Kawasaki, Japan). Equilibrator. The equilibrator, modeled on the equilibrator for fCO2 measurements developed by the National Institute for Environmental Studies,32 consisted of a brown (to prevent photolysis) vertical glass tube (i.d. 15.3 cm; height 80 cm; volume and column height of water are 10 L and 60 cm, respectively.) equipped at the bottom with a bubbler head (mesh size 20-30 µm) for the carrier gas and a seawater outlet and at the top with a carrier gas outlet and a seawater inlet (Figure 1). Dissolved DMS was extracted into DMS-free carrier gas (ultrapure N2). The carrier gas and the sample seawater stream in the equilibrator flowed in opposite directions, and part of the extracted gas was continuously directed to the PTRMS at ambient pressure without pretreatment. Teflon tubings (3/8” and 1/8” o.d.) and fittings were used to deliver samples to the PTR-MS to minimize loss of DMS. Tygon tubings (SaintGobain, Courbevoie, France) were used for seawater samples. (32) Murphy, P. P.; Nojiri, Y.; Fujinuma, Y.; Wong, C. S.; Zeng, J.; Kimoto, T.; Kimoto, H. J. Atmos. Oceanic Technol. 2001, 18, 1719–1734.

The flow of seawater was controlled with a peristaltic tubing pump (MasterFlex 7554-95, Cole-Parmer Instrument Company, IL), and continuously monitored with a flow meter (LD10TATAAA-RC and DU-5TGS1, HORIBA STEC, Kyoto, Japan). Water temperature was monitored with a temperature probe and a logger (TidbiT V2 Temp, Onset Computer Corporation, MA) at the bottom of the equilibrator. Laboratory Experiments on Artificial Seawater. The EIPTR-MS system was evaluated in the laboratory using artificial seawater containing DMS. The extraction efficiency, detection limit, and response time were determined. The water flow rate to the equilibrator was controlled at 1 L min-1. We added NaCl to pure water and controlled its salinity at 35 psu. We used the apparatus described by Ooki and Yokouchi26 to dissolve DMS in the salt water (Figure 1). DMS-containing artificial seawater was continuously produced with a silicone hollow fiber membrane module (NAGASEP, Nagayanagi Co., Tokyo, Japan; 6000 silicone tubes; length 20 cm; o.d. 0.25 mm; i.d. 0.17 mm). Sample gas containing 3 ppbv DMS prepared by dynamic dilution of the gravimetric gas standard was supplied to the module at 5 L min-1. The sample gas passing through the module was bubbled through water in a bucket. Thus DMS in the sample gas permeated the silicone of the module and dissolved in the circulation water. Using this procedure, we obtained DMScontaining artificial seawater in equilibrium with the sample gas at ppbv levels. Analytical Chemistry, Vol. 81, No. 21, November 1, 2009

9023

Quantification of Dissolved DMS. To calculate the DMS concentrations in sample water, we applied correction factors to the concentrations in the sample gas extracted from the equilibrator. Given that DMS is in equilibrium between the gas and liquid phases, its concentration in the liquid phase (ca) should equal the product of the Henry’s law constant (kH) and the partial pressure of DMS in the gas phase (pg): ca ) kH × pg

(4)

Henry’s law constants depend on temperature (T): kH ) kH0 × exp[-∆solnH/R × (1/T - 1/T0)]

(5)

where kH0 is the Henry’s law constant under standard conditions (T0 ) 298.15 K), ∆solnH is enthalpy of solution, and R is the gas constant. We used data for the Henry’s law constant reported in previous studies under seawater conditions (salinity ≈ 35 psu).33-37 We estimated the Henry’s law constant and its temperature dependence by averaging all the data (shown in Supporting Information (SI) Figure S-1). We calculated kH0 and -∆solnH/R to be 0.42 M atm-1 and 3593 K, respectively, and we used the calculated Henry’s law constant to calculate the concentrations of DMS dissolved in seawater. Field Deployments and Comparison with Membrane Equilibrator-Gas Chromatography/Mass Spectrometry Data. Field observations were made during the SPEEDS/SOLAS (Subarctic Pacific Experiment for Ecosystem Dynamics Study/ Surface Ocean Lower Atmosphere Study) 2008 cruise of the R/V Hakuho-Maru in the western subarctic North Pacific Ocean in July-August 2008. We collected natural seawater samples from the sea surface with continuous pumping systems, and part of the collected seawater was continuously introduced to the equilibrator. For underway seawater sampling, we used a pumping system with a towed torpedo-shaped fish to keep the sampling inlet under the sea surface. Because the seawater samples were shared by several groups, the seawater flow to the EI-PTR-MS was limited to ∼1.0 L min-1. Therefore, we set the flow rate of seawater to the equilibrator at 1 L min-1. When the towed torpedo-shaped fish was unavailable (i.e., when the ship hove to), we collected samples with a built-in pumping system of R/V Hakuho-Maru. Although a much higher flow rate was obtainable with this built-in system, we kept the flow rate at 1 L min-1 for consistency. The DMS concentrations obtained with the two sampling systems were not significantly different (shown in SI Figure S-2). We set the carrier gas flow rate for the equilibrator inlet to 75 sccm during the observations. PTR-MS measurements were conducted in multiple-ion detection mode (5 s data collection per cycle); we obtained mass signals for DMS at 1 min intervals. (33) Cline, J. D.; Bates, T. S. Geophys. Res. Lett. 1983, 10, 949–952. (34) Przyjazny, A.; Janicki, W.; Chrzanowski, W.; Staszewski, R. J. Chromatogr. 1983, 280, 249–260. (35) Dacey, J. W. H.; Wakeham, S. G.; Howes, B. L. Geophys. Res. Lett. 1984, 11, 991–994. (36) De Bruyn, W. J.; Swartz, E.; Hu, J. H.; Shorter, J. A.; Davidovits, P.; Worsnop, D. R.; Zahniser, M. S.; Kolb, C. E. J. Geophys. Res. 1995, 100, 7245–7251. (37) Wong, P. K.; Wang, Y. H. Chemosphere 1997, 35, 535–544.

9024

Analytical Chemistry, Vol. 81, No. 21, November 1, 2009

We compared the EI-PTR-MS data to data obtained by means of membrane equilibrator-gas chromatography/mass spectrometry (ME-GC/MS).26 Surface seawater samples for ME-GC/MS were obtained with the built-in sampling system. Seawater for ME-GC/ MS was continuously passed at 15 L min-1 through the silicone membrane tube which consisted of six silicone tubes (length 10 m; o.d. 2.0 mm; i.d. 1.5 mm; Q7-4780, 80SH, Fuji System Co., Tokyo, Japan) housed in a polyvinyl chloride pipe (i.d. 16 mm). Dissolved DMS permeated the silicone membrane into the gas phase within the silicone tubing. Ultrapure air was continuously supplied to the equilibrator inlet at 25 sccm. The collected air samples were dehydrated with a Nafion dryer (Perma Pure LLC, NJ) and transferred to a preconcentration/ capillary GC/MS system. A gravimetrically prepared standard gas (DMS 560 pptv; Taiyo Nippon Sanso Co., Tokyo, Japan) was analyzed for quantification. Standard gases independently prepared for EI-PTR-MS and ME-GC/MS agreed within 2%. The ME-GC/MS was originally developed for the automatic measurements of trace marine halocarbons which typical concentrations are in the range of 0.1 ∼ hundreds pmol L-1. For DMS, the system suffers from a little blank problem due to carryover (2-10%, expressed as an inversion function of DMS amount in the previous sample) from the previous sample on the Nafion dryer, and therefore, the concentrations of DMS were used after correction for that. Analytical precision (2σ) based on the repeated analyses of the standard gas was 6%. Measurements of DMS as well as several halocarbons with this system were done every 70 min during the cruise. RESULTS AND DISCUSSION Humidity Dependence of the Detection Sensitivity. In the EI-PTR-MS system, the sample gases extracted through the equilibrator contained a lot of water vapor. Therefore, we considered the humidity dependence of the detection sensitivity for DMS by means of a procedure described in detail by Inomata et al.30 The humidity was varied from 0 to 36 mmol mol-1 with a humidity generator/ controller (SRG-1R-10, SHINYEI, Kobe, Japan). The detection sensitivity of DMS under dry conditions was 6.0 ncps ppbv-1. The product ion count rates are given as normalized counts per second (ncps; normalized to an H3O+ intensity of 106 cps). The detection sensitivity depended markedly on the humidity, indicating that DMS reacted with H3O+ ions and with cluster ions in the drift tube. Because the number of cluster ions increased with increasing humidity, a simple function was used to correlate humidity with detection sensitivity. The humidity dependence of the detection sensitivity (RS) was better fitted by the following quadratic function than a linear one: RS ) (0.000253 ( 0.000067) × {[H2O]sample}2 + 1, R2 ) 0.98 ([H2O]sample ) 0-36 mmol mol-1)

(6)

Equilibration Efficiency. The extraction efficiency was calculated by taking the ratio of DMS signals in sample gases extracted from and dissolved to the artificial seawater. Extraction efficiencies reached 100% (within signal variability) at a flow rate ranging from 75 to 500 sccm, which indicates that DMS completely equilibrated between the carrier gas and seawater in the equilibrator (Figure 2).

Figure 2. Extraction efficiency of DMS as a function of carrier gas flow rate. Vertical bars represent standard deviations (2σ).

Figure 4. An example of temporal variation of DMS dissolved in surface seawater together with surface seawater temperature during the SPEEDS/SOLAS 2008 cruise. Data collected at 5 s integration are plotted at 1 min intervals for EI-PTR-MS (b), overlaid with MEGC/MS data (O). Table 1. Statistics for Concentrations of DMS Dissolved in Surface Seawater Measured during the SPEEDS/SOLAS 2008 Cruise DMS concentration (nmol L-1)

Figure 3. Response time of the EI-PTR-MS system to step changes in the DMS concentration of the artificial seawater stream. The artificial seawater samples were switched from DMS-containing to DMS-free batches at the time of 0 min. Data collected at 1 s integration are plotted every 7 s.

average

SDa

minimum

median

maximum

4.6

3.9

1.1

3.4

30.4

a

SD, standard deviation.

To measure the partial pressure of CO2 in surface seawater, Frankignoulle et al.29 used a bubbling-type equilibrator similar in size (height 80 cm; diameter 10 cm) to ours. Equilibration of CO2 was complete within 2 min at a carrier gas flow rate of 3 L min-1 and a seawater flow rate of 3 L min-1. The solubility of DMS (Henry’s law constant, 0.42 M atm-1 at 25 °C) is 1 order of magnitude higher than that of CO2 (0.034 M atm-1 at 25 °C; Wilhelm et al.38). The achievement of equilibrium depends not only on the residence time of the carrier gas and sample water but also on the solubility of the target gas.39 More-soluble species reach equilibrium more quickly, which agrees with our experimental results for equilibration of DMS. Response Time. To estimate the response time for the EIPTR-MS system, we evaluated the decay in DMS concentration when two artificial seawater batches of greatly different concentrations were switched at the inlet of the equilibrator. We prepared a DMS-containing batch (section 2.3) and a DMS-free batch consisting of salted Milli-Q water with no added DMS. By transferring the intake tube from the DMS-containing batch to the DMS-free batch, we induced an instantaneous decrease in DMS concentration. The response was quite rapid, with a response time (defined as e-folding time) of ∼1 min (Figure 3). The residence times of the carrier gas in the headspace and of seawater in the equilibrator were calculated to be 8 and 10 min, respectively; that is, the instrumental response time for DMS was much shorter than the residence times in the equilibrator. Not only residence time of carrier gas and seawater but also solubility of target gas contributes to the response time. The shorter-than-expected response

time likely resulted from the achievement of equilibrium for DMS between seawater directed from the inlet and gases residing in the headspace, owing to the relatively high solubility of DMS. Detection Limit. To determine the background signals for DMS for the system, we analyzed N2-purged pure water samples. The pure water stored in the equilibrator was purged with pureN2 gas at 1.0 L min-1 for 6 h, and then the DMS signals were analyzed at a carrier gas flow rate of 75 sccm. More than 99% of the dissolved DMS was extracted into the gas phase by means of this degassing treatment. All the oceanic data were corrected by subtraction of the background signal. We calculated the detection limit to be 50 pmol L-1 at a signalto-noise (S/N) ratio of 2 with an integration time of 5 s by assuming the water temperature to be 25 °C and the water vapor concentration to be the saturated vapor concentration at 25 °C. The DMS detection limit of the EI-PTR-MS system was 2 orders of magnitude lower than that of the MIMS system.16 This is low enough to detect DMS in the open ocean, in comparison to database of DMS concentration (0.04-315.69 nmol L-1).6 Field Observation and Comparison of EI-PTR-MS to MEGC/MS. The large temporal variations of DMS dissolved in surface seawater captured by EI-PTR-MS (Figure 4) likely reflected great spatial variations in DMS distributions associated with variable biological activity in the western North Pacific Ocean. High DMS concentrations were associated with regions of relatively high water temperature. The mean DMS concentration for all samples was 4.6 ± 3.9 nmol L-1 (Table 1), which agrees well with summertime levels of 4.3 ± 3.2 nmol L-1 in the same region obtained by means of a general GC method.40

(38) Wilhelm, E.; Battino, R.; Wilcock, R. J. Chem. Rev. 1977, 77, 219–262. (39) Johnson, J. E. Anal. Chim. Acta 1999, 395, 119–132.

(40) Aranami, K.; Tsunogai, S. J. Geophys. Res. 2004, 109, 10, DOI: 1029/ 2003JD004288.

Analytical Chemistry, Vol. 81, No. 21, November 1, 2009

9025

EI-PTR-MS and ME-GC/MS data. Other possible explanations include uncertainties in correction factors for background signals (±1% at 10 nmol L-1) for EI-PTR-MS. As for ME-GC/ MS, the background signals due to artifacts may have been underestimated at higher concentrations. Uncertainty in water temperature measurements was also a contributing factor (±4%). We conclude that EI-PTR-MS and ME-GC/MS were in good agreement for DMS measurements within their uncertainties and reducing these uncertainties will be considered in the future work.

Figure 5. Scatter plots of DMS measured by EI-PTR-MS against data measured by ME-GC/MS (black circles, left axis). The solid line is a regression line; the dashed indicates line 1:1 correspondence. Differences between the two methods are also shown (gray crosses, right axis).

In general, the EI-PTR-MS and ME-GC/MS data agreed well (Figure 4). On 15 August, episodes of high DMS concentrations, exceeding 25 nmol L-1, were detected by EI-PTR-MS; the high concentrations likely resulted from rapid production and/or consumption due to biophysicochemical parameters such as phytoplankton biomass, mixed layer depth, seawater temperature, and nutrient concentrations. ME-GC/MS missed the top of the high peaks in these episodes (although the method tracked the basic pattern of variations) owing to its limited measurement frequency (i.e., 70 min intervals at maximum). Figure 5 shows a detailed comparison of DMS concentrations determined by EI-PTR-MS and ME-GC/MS. We used reducedmajor-axis (RMA) regression,41 rather than standard linear leastsquares regression, to obtain the slope and intercept because both data sets were measured variables and were, thus, both subject to error. RMA regression is a bilinear method that allows for errors in both variables. The EI-PTR-MS values compared here were averaged over 30 min to match the time resolution of ME-GC/MS. Both data sets were tightly correlated. The slope of the overall regression line for DMS was 0.90 ± 0.02, and the intercept was negligible (-0.03 ± 0.30 nmol L-1, R2 ) 0.99). EI-PTR-MS tended to underestimate DMS concentrations compared to ME-GC/ MS at high concentrations (>10 nmol L-1). High DMS concentrations (>10 nmol L-1) were detected in a relatively warm water region (>16 °C) (Figure 4), where the humidity of extracted gases reached 35 mmol mol-1. At high humidity range, the uncertainty associated with the fitting curve for humidity correction was large (±4% at 35 mmol mol-1 humidity). This uncertainty likely contributed to the difference between the (41) Ayers, G. P. Atmos. Environ. 2001, 35, 2423–2425.

9026

Analytical Chemistry, Vol. 81, No. 21, November 1, 2009

CONCLUSIONS We developed an EI-PTR-MS system for continuous measurement of concentrations of DMS dissolved in seawater. We investigated several important correction factors for accurate quantification of dissolved DMS using the EI-PTR-MS system with high sensitivity and time-resolution. The detection limit of EI-PTRMS for DMS was low enough to detect DMS in the open ocean, suggesting that EI-PTR-MS would greatly increase the amount of data for DMS dissolved in surface seawater over wide regions of the ocean. The EI-PTR-MS system required a relatively large space (about 2 × 2 × 2 m3); therefore, the system will need to be made smaller if it is to be widely used onboard ships. EI-PTR-MS has the potential to enhance the spatial and temporal coverage of dissolved DMS concentrations in the open ocean surface and, thereby, will be a useful tool for improving our understanding of the global budget and production processes of DMS. ACKNOWLEDGMENT We express our sincere thanks to Dr. Yukihiro Nojiri of National Institute for Environmental Studies for fruitful comments on the development of the equilibrator. We thank Dr. Atsushi Tsuda of the University of Tokyo; Dr. Jun Nishioka, Dr. Daisuke D. Komatsu, Mr. Uta Konno, and Ms. Satoko Daita of Hokkaido University; Dr. Hirofumi Tazoe of Nihon University; and the captain and crew of the R/V Hakuho-Maru for support during the SPEEDS/SOLAS 2008 cruise. We also wish to thank Dr. Yutaka W. Watanabe of Hokkaido University for valuable comments on DMS measurements. This research was supported by a Grantin-Aid (No. 1867001) for Scientific Research in Priority Areas “Western Pacific Air-Sea Interaction Study (W-PASS)”. The underway-sampling system used in this work was developed as a part of the Japan EOS Promotion Program sponsored by MEXT. This research is a contribution to the Surface Ocean Lower Atmosphere Study (SOLAS) Core Project of the International Geosphere-Biosphere Programme (IGBP). SUPPORTING INFORMATION AVAILABLE Figure S-1: Temperature dependence of Henry’s law constants for DMS. Figure S-2: Sampling methods on the R/V Hakuho-maru. This material is available free of charge via the Internet at http://pubs.acs.org. Received for review July 22, 2009. Accepted September 21, 2009. AC901630H