Anal. Chem. 1991, 63,529-532
LITERATURE CITED
(1) Cady, P.; Dufour, s. W.; Shaw, J.; Kraeger, S.J. J . Clin. Microbiol. 1978, 7 , 265-272. (2) Fraatr, R. J.; Prakash, G.; Ailen, F. S . A m . Biotechnol. Lab. 1988, 6 , 24-28. (3) Shelly, D. C.; Quaries, J. M.; Warner, I. M. Anal. Left. 1981, 14(B13), 1111-1 124. (4) Steinkamp. J. A.; Fulwyler, M. J.; Couker. J. R.; Hiebert, R . D.; Horney, J. L.; Mullaney. P. F. Rev. Sci. Instrum. 1973, 44, 1301-1310. (5) Regina, F. J.; Lin, S. H.: Bolts, J. M. Abstracts of Papers, 199th National Meeting of the American Chemical Society, Boston, MA, April 22-27, 1990; American Chemical Society: Washington, DC, 1990; Abstract No. 51 (6) Lim. L. C. L.; Pennell, D. R.; Schell. R. F. J . Clin. Microbia/. 1990, 28, 670-675. Bej, A. K.; Steffan, R. J.; DiCesare, J.; Haff, L.; Atlas, R . M. Appi. Environ. Microbiol. 1990, 56, 307-314. Newman, J. S.:O'Brien, R . T. Appl. Microbiol. 1975, 30, 584-588. Bachrach. U.; Bachrach, Z. Appl. Microbiol. 1974, 2 8 , 169-171. Maoyu. Y . ; Zhang. Y. Appl. Environ. Microbiol. 1989. 55, 2082-2085. Libby, J. M.; Wada, H. G. J . Clin. Microbiol. 1989, 2 7 , 1456-1459. Graham, K.; Keller, K.; Ezzel, J.; Doyle, R . Eur. J . Clin. Microbiol. 1984, 3 , 210-212. Feng. P. C. S.: Hartman, P. A. Appl. Environ. Microbiol, 1982, 43, 1320- 1329. Warren, L. S.;Benoit, R . E.; Jessee, J. A. Appl. Environ. Microbiol. 1978, 3 5 , 136-141. Edberg, S. E.: Allen. M. J.; Smith, D. B. Toxic. Assess. 1988, 3 , 565-580. Godsey. J. H.; Matteo, M. R.; Shen, D.;Tolman, G.; Gohlke, J. R. J . Clin. Microbiol. 1981, 13, 483-490. Snyder, A. P.: Wang. T. T.: Greenberg, D. B. Appl. Environ. Microbiol. 1988. 5 1 . 969-977. Berg, J. D.; Fiksdal, L. Appl. Environ. Microbiol. 1968, 5 4 , 21 18-2122.
529
(19) Edberg, S.C.; Edberg, M. M. Yale J . Biol. Med. 1988, 67, 389-399. (20) Covert, T. C.: Shadix, L. C.; Rice, E. W.; Haines, J. R.; Freyberg, R. W. Appl. Environ. Microbiol. 1989, 55, 2443-2447. (21) St. Louis, R . H.; HI4 H. H., Jr. Crit. Rev. Anal. Chem. 1990, 21, 321-355. (22) Lawrence, A. H. J . Pulp Paper Sci. 1989. 15, J196-J199. (23) Eiceman, G. A.; Blyth, D. A.; Shoff, D. B.; Snyder, A. P. Anal. Chem. 1990, 62, 1374-1379. (24) Eiceman, G. A.; Snyder, A. P.; Biyth, D. B. Int. J . Environ. Anal. Chem. 1990, 38, 415-425. (25) Eiceman, G. A.; Shoff, D. 6.; Harden, C. S.;Snyder, A. P.;Martinez, P. M.; Fleischer, M. E.; Watkins, M. L. Anal. Chem. 1989, 61. 1093- 1099. (26) Karasek. F. W.; Kim, S. H.; Hill, H. H..Jr. Anal. Chem. 1976, 48, 1133-1 137. (27) Preston, J. M.; Rajadhyax, L. Anal. Chem. 1988, 6 0 , 31-34. (28) Eiceman, G. A.; Shoff, D. 6.; Harden, C. S.;Snyder, A. P. Int. J . Mass Spectrom. Ion Proc. 1968, 8 5 , 265-275. (29) Lawrence, A. H.; Neudorfl, P. Anal. Chem. 1988, 6 0 , 104-109. (30) Standard Methods for the Examination of Water and Wastewater, 16th ed.; Greenberg, A. E., Trussell, R. R., Clesceri, L. S.,Eds.; American Public Health Assoc.: Washington, DC, 1985; pp 876-886. (31) Rice, E. W.; Allen, M. J.; Edberg, S.C. Appl. Environ. Microbiol. 1990, 56. 1203-1205. (32) Novick, A.; Weiner. M. Proc. Natl. Acad. Sci. U . S . A . 1957, 4 3 , 553-566. (33) Bello, J. Science 1960, 165, 240-241. (34) Pardee, A. B.; Prestidge, L. S. Biochim. Biophys. Acta 1961, 49, 77-88.
RECEIVED for review August 2, 1990. Accepted November 19, 1990.
Extraction of Methane from Seawater Using Ultrasonic Vacuum Degassing Manfred Schmitt* Geochemische Analysen, Wilhelmstrasse 36, 0-3160 Lehrte, FRG
Eckhard Faber Bundesanstalt fur Geowissenschaften und Rohstoffe, Stilleweg 2, 0-3000 Hannover 51, FRG
Reiner Botz and Peter Stoffers Geologisch-PalaontologischesInstitut, Universitat Kiel, Ohlshausenstrasse 40-60, 0-2300 Kiel, FRG
INTRODUCTION Methane (and higher hydrocarbons) is a well-known constituent of marine and freshwater sediments and the overlying water column (1-5). Degassing of sediments has been shown to deliver reliable information on both the gas quantities and the stable isotopic composition of methane in sediments (6-8). Hydrocarbon gases in sediments generally consist of more than 80% methane, with 6 13CH4between -100% and -20% (8), depending on the thermal or bacterial origin and the postgenetic history of gases. Although the saturation concentration of methane in water is near 23 mg/L, such high values are virtually never found. Methane concentrations of ocean waters usually range from several nanograms/liter to several micrograms/liter (4,9,10). This may be explained by low methane input into the water, gas release into the atmosphere before equilibration, and/or oxidation of methane in the water column. It has been shown that high methane concentrations in seawater are characteristic of active hydrothermal areas. Mapping of methane concentration anomalies in seawater has proved to be a reliable indicator for active hydrothermal areas (9,10). In general, the methane concentration in seawater is determined directly on board research vessels, as sample storage is not necessary then. To extract the dissolved gases, a helium
stripping method is commonly used (11).It requires relatively large amounts of purified helium, which is lost after stripping to the atmosphere. To recover methane from the carrier gas (helium), an adsorption-desorption technique is required. This method delivers reliable results on the methane concentration in water samples. However, isotopic fractionation of the extracted methane is possible due to incomplete adsorption-desorption processes. To overcome this problem, a new water-degassing system has been developed: methane (and other dissolved gases) is extracted in a vacuum system while the water sample is exposed to ultrasonic energy. A capillary gas chromatograph is used to determine the methane concentration. The main advantage of the new system is that within a short period of time (approximately 20 min/sample) reproducible data are available for the methane concentration in water. This may contribute to further cruise planning in the evaluation of hydrothermal areas and furthermore provides isotopically unfractionated methane for later stable carbon isotope investigations.
EXPERIMENTAL SECTION Water Sampling. Water samples are taken by Niskin bottles in water depths ranging from the surface down to 6000 m. After
that, the water is transferred from the Niskin bottles through short silicone tubes into evacuated glass bottles. As the analysis time
0003-2700/91/0363-0529$02.50100 1991 American Chemical Society
530
0
ANALYTICAL CHEMISTRY, VOL. 63, NO. 5, MARCH 1, 1991
Gas Sample
Q
Water Reserve
Water TrOD B o t l l e
U l t r a sonic R o l h
Figure 1. Ultrasonic vacuum degassing line.
for one sample is only 20 min, no long sample storage is required prior to analysis. Gas Extraction. The dissolved gases are extracted a t room temperature in a vacuum system at nearly water vapor pressure while ultrasonic energy is applied to the water samples (Sonorex RK102,35 kHz, 2 X 240 W).The ultrasonic treatment leads to gas release into the head-space. To prevent chemical reactions of the dissolved gases, like cracking of the C2+components possibly existing (ethene, ethane, propane, and other homologues up to hexane) or oxidation (12),the ultrasound is applied frequently for seconds only during a time interval of 5 min. The degassing line shown in Figure 1consists of a sample bottle (A) (volume > 1 L) that is connected via a valve to a gas buret (B). An evacuated gas container (C) is attached on top of the gas buret. If the gas quantities are high, this container may be filled with gas for later stable isotope analysis. Gas for on-board gas chromatography (GC) analysis is sampled by a syringe via a silicone septum (D).A t the lower end of the sample bottle (A), the water sample is connected by a glass tube and a valve via a second glass bottle (E) (water reserve) to the vacuum system (protected by a water trap bottle (F)). The degassing system is made of glass, which may accidently implode due to the applied vacuum. This is not very likely to occur. However, to avoid injuries by broken glass pieces, all glass bottles are wrapped with plastic tape. The whole apparatus is mounted in a metal rack of size 80 X 80 X 80 cm for easy transportation and rapid installation. For sampling, bottle A (volume approximately 1.2 L)is disconnected from the line, evacuated, and, via V1, completely filled with water from the Niskin samplers. For analysis, bottle A is attached to the buret. Some water is pumped off into bottle E until bottle A contains approximately 1L of water. The vacuum of the head-space in bottle A at this time is near water vapor pressure. The ultrasonic energy is applied frequently over a time period of 5 min. A t the end of the gas extraction, the vacuumdegassed water of bottle B is added to the water sample in bottle A. Via V1 the water level raises into the buret until atmospheric pressure is reached. The gas quantity can be read on the scale of the buret (commonly between 5 and 20 mL). One milliliter of gas is taken through a septum for immediate GC analysis. The remaining gas may be transferred into a gas container for further analysis (e.g., isotopic analysis in the case of sufficient gas quantities).
lr
n
Figure 2. Typical gas chromatogram of a hydrocarbon gas mixture (methane to hexane, each 1000 ppm in helium). Conditions: column, 30-m GS-Q quartz capillary; carrier gas, helium 4 mL/min; detector, FID, range 3; column temperature, initial 50 OC, 2 min, raise 3 OC/min, final 180 OC, 5 min; injection, 1-mL gas mixture, each 1000 vppm.
Gas Chromatographic Analysis. A Shimadzu GC-14APSF gas chromatograph is used for methane determination. One milliliter of the gas sample is injected into a 30-m-long, megabore (0.53-pm) GS-Q-coated quartz capillary column. The carrier gas is nitrogen of high purity grade with a flow rate of approximately 4 mL/min. The column temperature normally is 30 "C. The hydrocarbons are recorded by a flame ionization detector (FID). Peak processing, integration, and reporting are done by a programmable integrator (Shimadzu C-RGA). The gas consumption is low (4 mL N2/min, 15 mL Hn/min, 200 mL/min air provided by a membrane pump combined with a cleaning system). The GC system is permanently mounted in a metal rack (size: 80 X 80 X 80 cm) for easy transportation and rapid installation. With a GC system as described above, a C1-C6gas mixture providing a very good CI-C2peak separation can be determined in less than 11min (Figure 2). The linearity of the GC system was determined by injection of different quantities of methane (1-loo0 nL). The correlation factor of 0.9998shows a good linear correlation of the signal according to the injected amount of gas. The signal-to-noise ratio of the system was checked by direct injection of 5 nL of methane into the gas chromatograph. The measured signal-to-noise ratio of 101 is sufficient to determine methane background quantities down to 1 nL of CHI.
-
ANALYTICAL CHEMISTRY, VOL. 63, NO. 5, MARCH 1, 1991
531
Table I. Recovery and Reproducibility of Methane Concentration in Tap Water (CH,= 58 nL of Methane/L of Water) Equilibrated with Air (CH,= 1.7 vppm)
no.
degassed water, mL
gas volume, mL
methane concn, nL CHI/L
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
750 700 750 750 780 750 750 760 750 780 770 7 80 750 790 750
14.0 12.3 12.4 12.1 11.5 14.0 14.6 12.9 14.4 14.5 14.9 14.1 13.5 14.1 14.3
62.5 57.6 51.4 50.8 48.7 41.2 53.2 41.9 55.2 53.5 49.1 53.7 47.1 45.2 56.3
13.6 1.1
51.2 5.8
mean
l a (std dev)
0
(
13C/12Csample - 13C/12Cstandard 13C/12Cstandard
1
1000
40000
60000
80000
C H 4 [ n l / L I US Figure 3. Methane concentrations gained by helium stripping and ultrasonic vacuum degassing. The line given represents values of equal concentrations for both methods.
Isotope Analysis. The preparation of gas samples for isotopic analysis is done in a preparative gas chromatographic system. Methane is separated from air and other components by lowtemperature gas chromatography with a packed column (Porapak Q, 1/4 in., 3 m) and then quantitatively converted into COz in an oxidation oven filled with purified CuO at a temperature of 900 OC. The reaction products (C02and HzO) are frozen in a trap at liquid nitrogen temperature (-196 "C). COz is separated from HzO by raising the temperature of the cold trap to -78 "C (dry ice). The released C02 is collected, and the carbon isotope ratio is measured on a Finnigan MAT 251 mass spectrometer. Isotope ratios are given in the usual 6 notation: 6 13CH4=
20000
[%0]
All 13C/12Cisotope ratios are given relative to the PDB (Pee Dee Belemnite) standard. Reproducibility of the isotope analysis is about *1L for 10 WLof methane (13). RESULTS Recovery of Methane by the Ultrasonic Vacuum Method. T o establish methane equilibration at the room temperature, 20 L of tap water was exposed to the atmosphere for 2 weeks. The calculated (after Henry's law) methane equilibration concentration of this water is about 58 nL of CH4/L of water (using a Bunsen factor of a = 0.034 16 L of CH4/L of H 2 0 at 20 "C and atmospheric methane concentration of 1.7 vppm (14)). After that, the water was degassed 15 times with the ultrasonic vacuum method. The results are The mean (atmosphere-equilibrated) listed in Table I:
methane concentration is 51 nL of CH4/L ( l a = f 5.8 nL of CH4/L), which represents a recovery rate of 88% by ultrasonic vacuum degassing. Higher recoveries are possible by multiple degassing of the water (IO). Comparison of Both Methods: Ultrasonic Degassing and Helium Stripping. For comparison purposes, the following water samples were extracted by both methods: sample 1, water from Lake Hohnhorst near Hannover; sample 2, water from a swamp in Lehrte close to Hannover; sample 3, artificial gas-water mixture (several liters of tap water were equilibrated with a methane-air mixture. The carbon isotope ratio of the methane used was 6 13CH4= -33.2%0 PDB). The gases released during ultrasonic vacuum treatment were transferred into gas containers (Figure 1) for isotope measurements and immediate methane concentration analysis. The gases released during helium stripping were collected in a cool trap (1/8-in.stainless steel, filled with activated charcoal a t liquid nitrogen temperature (-196 "C)). After stripping, the adsorbed hydrocarbons were removed from the charcoal by heating the trap to 200 "C. The gas was analyzed by GC analysis, and the carbon isotope ratio of methane was determined by a mass spectrometer (see Table 11). It can be seen that both techniques deliver similar methane yields when the methane concentration in the water is low, and relatively high He quantities (1250 mL,samples 7 and 8) are used for stripping. However, the comparison presented in Table I shows that ultrasonic vacuum degassing delivers higher membrane concentrations than He stripping when the methane concentration in the water is high. T o extract large methane quantities, several liters of stripping helium is obviously required for complete methane extraction (Table 11, samples 1-6; Figure 3). Isotopic analysis shows the m0:hane degassed by helium stripping (-36.5%0 I 6 13CH4I - 3 6 k ) to be enriched in 12C by some 3% relative to the original methane (6 13CH4= -33.2700, Table 11, samples 4-6). Whether this is due to incomplete degassing of water by helium stripping or due to adsorption-desorption processes, as suggested, cannot be decided. However the value 6 13CH4of -33.6%0of the methane
Table 11. Methane Concentrations and Carbon Isotope Ratios of Methane Degassed from Water Samples by Helium Stripping and Ultrasonic Vacuum Degassin@ ultrasonic sample
no.
CHrH20 lake
1 2 3 4 5 6 7 8
swamp
CHI-HZO CHd-HZO CHI-HZO lake lake
CHI, nL/L
c
15 780 5 280 650 34 200 74 000 47 100 250 340
6 WH4, %O nd
nd nd -33.7 -34.0 -33.7 -27.0 -27.7
helium stripping CH4, nLfL 6 13CH4,L
c
4 820 1920 180 26 000 37 400 31 600 340 330
Samples 1 and 4-6 are equilibrated with methane of 6 WH, = -33.2%. nd = not determined.
nd nd nd -36.5 -36.0 -36.6 -30.1 -31.1
quantity of strip gas, mL 130 130 130 1250 1250 1250 1250 1250
Anal. Chem. 1991, 63,532-533
532
::I I
I
I.
.1
/
/ /
,. ............................. mithine M/LI
'\
\
/
'.
Flgure 4. Histograms of methane concentrationsof 656 water samples
from different working areas.
degassed by ultrasonic treatment (samples 4-6) is nearly identical with the value of methane (6 13CH4= -33.2%) used for the equilibration experiment. This suggest that ultrasonic vacuum degassing gives reliable information not only on the amount of methane in water samples but also on its carbon isotopic composition. The differences in 6 13CH4values of gases released by both methods (He stripping and ultrasonic degassing) (samples 7 and 8; Table 11) are explained by an
isotopic fractionation occurring when the helium stripping technique is used. Application. The ultrasonic vacuum degassing system was used on board research vessels RV Sonne, RV Polarstern, and RV Meteor operating in various sea areas. During these cruises, 656 water samples were analyzed. The aim of this study was always to identify hydrothermally active areas by detection of methane plumes in the water column. All data obtained by the new system are presented in histograms (Figure 4). A detailed discussion of the gas data will be presented elsewhere. The methane concentration of 15 nL of methane/L of seawater reflects the normal or slightly elevated background value similar to published data gained by helium stripping. High methane concentrations (in one case up to 31615 nL of CH4/L) are probably due to hydrothermal input. Isotopic analysis will help to clarify whether the methane is of abiotic or biogenic origin.
LITERATURE CITED Craig, H. Geochim. Cosmochim. Acta 1053, 3 , 53-92. Claypool, G. E.; Kaplan. I.R. Natural Gases in Marine Sediments; Plenumi New York, 1974; pp 99-139. Stahl, W.; Faber, E., Carey, B. D., Jr.; Kirksey, D. L. AAFGBull. 1081, 9. 1543-1550. ..... Welhan, J. A.; Lupton, J. E. AAPG Bull. 1987, 2 , 215-223. Whiticar, M. J.; Faber, E.; Schoell, M. Geochim. Cosmochim. Acta 1986, 50, 693-709. Horvitz, L. AAPG Stud. Geol. 1978, 10, 241-269. Faber, E.; Schmltt, M.; Stahl, W. Initial Reports of the Deep Sea Drilling Project, Vol. XLII, p 2, Washington, 1978. Faber, E.; Rehder, S.; Stahi, W. &&I, Kohle. Erdgas, Petrochem. 1083, 8, 357-361. Horibe, Y.; Kim, K. R.; Craig, H. Nature 1988, 324, 131-133. Kim, K. R. Ph.D. Thesls, University of Califwnia, San Diego. 1983. Swinnerton, J. W.; Linnenbom, J. Gas Chromatogr. 1967, 5 , 570-573. Susiick, S. K. Ultrasound: Its Chemical, Physical and Biologlcal f f fects; VCH Verhgsgesellschaft mbH: New York and Weinheim, 1988; pp 138-163. Dumke, I.; Faber, E.; Poggenburg, J. Anal. Chem. 1089, 61, 2 149-2154. Cicerone, R. J.; Oremland, R . S. Global Biogeochem. Cycles 1988? 2(4), 299-327.
RECEIVED for review July 23, 1990. Accepted November 21, 1990.
Voltammetric Method for the Determination of Borohydride Concentration in Alkaline Aqueous Solutions Michael V. Mirkin and Allen J. Bard* Department of Chemistry, The University of Texas at Austin, Austin, Texas 78712 INTRODUCTION Sodium and other borohydrides have been used as the reducing agents in many inorganic ( I , 2) and organic (2,3) reactions. For carrying out these reactions, as well as in studies of methods for the synthesis of borohydride, one needs a simple, rapid, and accurate analytical method to control its concentration in a solution. None of several reported analytical procedures for borohydrides meets all of these requirements. The most accurate hydrogen evolution method ( 4 ) is quite complicated. A number of titrimetric methods (2),including the iodate method, are less accurate and are also not selective. These methods, as well as polarographic (5) and spectrophotometric (6)ones, cannot be used directly in a reaction system (for example, in an electrochemical cell) to measure continuously the borohydride concentration. To our knowledge, no one has studied the complicated electrode reaction of borohydride oxidation by cyclic voltam-
metry. We also found no previous references to the electrooxidation of BH4- ion at a gold electrode. In studying this process, we found that the linear sweep voltammograms a t a gold electrode (unlike those a t Ni or Pt electrodes) possess a well-defined shape suitable for analytical determinations. The high stability of the gold electrode and the very low magnitude of the background current in the potential region corresponding to borohydride oxidation lead to good precision of the analytical procedure described below.
EXPERIMENTAL SECTION Cyclic voltammetry of aqueous sodium borohydride solutions was performed in a three-electrode cell comprising a gold working electrode in a disk form (area 0.12 cm2),a RuOz (dimensionally stable) counter electrode and a saturated mercurous sulfate reference electrode (SMSE). The working electrode was polished before a series of measurements with 0.05-pm a-alumina paste (Buehler, Lake Bluff, IL). No additional polishing was done
0003-2700/91/0363-0532$02.50/00 1991 American Chemical Society
