xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.loc.govstandardsmarcxmlschemaMARC21slim.xsd
leader nam 2200397Ka 4500
controlfield tag 001 002068402
007 cr mnu|||uuuuu
008 100413s2009 flu s 000 0 eng d
datafield ind1 8 ind2 024
subfield code a E14-SFE0003150
Measurement of dissolved gas concentrations in natural waters utilizing an in-situ, membrane inlet, linear quadrupole mass spectrometer
h [electronic resource] /
by Peter Wenner.
[Tampa, Fla] :
b University of South Florida,
Title from PDF of title page.
Document formatted into pages; contains 55 pages.
Thesis (M.S.)--University of South Florida, 2009.
Includes bibliographical references.
Text (Electronic thesis) in PDF format.
Mode of access: World Wide Web.
System requirements: World Wide Web browser and PDF reader.
Advisor: Robert H. Byrne, Ph.D.
ABSTRACT: Since its creation in the late 19th century, mass spectrometry has evolved into one of the most versatile analytical methods in science. To chart this evolution this thesis includes a historical overview of mass spectrometry and a description of the role of mass spectrometry in oceanography. The development and deployment of underwater mass spectrometers (UMS) at the University of South Florida's Center for Ocean Technology has made possible the collection of real-time data with greatly increased spatial and temporal density. The UMS was deployed via both remotely guided surface vehicles (GSV) and ship's cables to monitor a suite of dissolved gases and volatile organic compounds in saltwater and freshwater environments. Spectrometer data in Lake Maggiore, Florida were acquired at a rate of 0.7-3.6 seconds/sample for 2-3 hours.The resulting multi-analyte spectrometer data were recorded in real time with the Global Positioning System (GPS) observations of an associated surface vehicle and transmitted to a remote laptop computer via a wireless Ethernet link. These data were merged to create high-resolution maps of chemical distributions. Of particular interest were the co-varying oxygen and carbon dioxide mass spectrometer signals, diagnostic of photosynthesis-respiration processes, that were collected over a 10,800 square-meter area of the lake. The UMS was also deployed on a shipborne hydrowire in Saanich Inlet, a 200-meter deep fjord in the western Canadian province of British Columbia. The concentrations of a broad suite of dissolved gases were monitored on both downcast and upcast over a total depth range of 200 meters. Spectrometer data were acquired at a rate of 4.2 seconds/sample for the duration of the deployment.Mass spectrometer signals diagnostic of reduced species (CH, HS,) in the anoxic waters of the inlet below a depth of 100 meters were consistent with previous descriptions of the fjord's chemistry. The UMS was deployed on a remotely guided surface vehicle on the Hillsborough River in central Hillsborough County. Spectrometer data were acquired at a rate of 0.7 seconds/sample, and geographic location was recorded by an onboard GPS during a 2,640 meter transect of the river. Prior to the deployment, the mass spectrometer was calibrated using certified gas standards. The calibration experiments correlated mass spectrometer ion intensity data with dissolved gas concentrations, whereupon the mass spectrometer data collected during the deployment were reported in units of micromole/kilogram (mol/kg).The mass spectrometer recorded changes in gas concentrations associated with changing physical conditions and biological activity along the 2,640 meters of the river that was transited by the GSV.
Volatile organic compounds
Dissolved carbon dioxide
x Marine Science
t USF Electronic Theses and Dissertations.
M easurement of Dissolved Gas Concentrations in Natural Waters Utilizing an In Situ, Membrane Inlet Linear Quadrupole Mass Spectrometer by Peter Wenner A thesis submitted in partial fulfillment of the requirements for the degree of Master of Science College of Marine Science University of South Florida Major Professor: Robert Byrne, Ph.D. R. Timothy Short, Ph.D. Edward Van Vleet, Ph.D Date of Approval: November 16 2009 Keywords: calibration, volatile organic compounds, dissolved oxygen, dissolved nitrogen, dissolved carbon dioxide Copyright 2009, Peter Wenner
Dedication Many are the people who have been essential to the successful completion of this e ndeavor and to whom I dedicate this work Del and Emil Wenner, and my sister Ann Wenner, whose love and support never wavered though often were the times I tried their patience. To my beloved friend Anne Wenner who saw in me things I never imagined existed and gave me the strength and confidence to pursue my dreams. To Libby, my cherished wife, who has taught me the meaning of joy and entered my life near the end of this enterprise and never blinked through all the insanity. To Dr. Timothy Short, as fine a mentor and friend for which one could hope, may our adventures together take us far into the future.
Acknowledgments In addition to those people singled out in the dedication above I would like to t hank all those who provided so freely of their time and knowledge in helping me to complete this To my major professor, Dr. Bob Byrne, whose support and persistence kept me afloat during those times I found myself drowning in a sea of doubt. If it can be said that I am a mass spectrometrist it is to the credit of Gottfried Kibelka and his enthusiastic explanation of the finer points of membrane inlet mass spectrometry. Special thanks to Mike Hall and Chad Lembke for their friendship and mechanical engineering tutelage. Special thanks to Ray Hazen for patiently sharing his vast knowledge of electronics with a complete novice. Thanks t o Charlie Cullins and Darryl Ashley for their assistance in streamlining the Thanks also to Joe Kolesar for his assistance in assembling cables, attaching connectors, building batteries and countless other projects. Particula r thanks to the Underwater Mass Spectrometry Group: To Strawn Toler whose friendship and academic acumen got me through several periods of self doubt. To Friso van Amerom whose quiet brilliance and stolid support was, is and will continue to be a n inspiration. And, to Ryan Bell for his intelligence, insight, enthusiasm, programming expertise and steadfast commitment to get me to once again play hockey.
i Table of Contents List of Tables ii ii List of Figures iii iii Abstract v Chapter One Introduction 1 Chapter Two A Brief History of M ass Spectrometry 4 Chapter Three The Role of Mass Spectrometr y in Oceanography 9 The University of South Florida Underwater Mass Spectro meter 11 Chapter Four In Sit u Mass Spectrometer Deployments 14 Bayboro Harbor and Lake Maggiore 14 Saanich Inlet 25 Calibrat ion of Mass Spectrometer Data 33 Hillsborough River 41 Conclusions 4 8 References 51
ii List of Tables Table 1 Operational s pecifications of the 200 amu underwater MS system 13 Table 2 Mass to charge (m/z) values, diagnostic of volatile organic amu underwater MS during deployment in Bayboro Harbor 16 Table 3 Mass to charge ( m/z ) values, diagnostic of dissolved gas species scanned by the 200 amu underwater MS during the Lake Maggiore deployment. 17 Table 4 Mass to charge ( m/z ) values diagnostic of dissolved gas species scanned by the 200 amu mass spectrometer during the April 13, 2 004 Saanich Inlet deployment. 29 Table 5 Identification of gases contained in the three certified gas standards purchased from Airgas, Inc. and the ir respective concentrations. 35
iii List of Figures Figure 1 Schematic of the 200 amu linear qu adrupole underwater MS system 11 Figure 2 Schematic showing detail of the membrane housing (left) and m embrane probe with internal compression spr ing (right) 12 Figure 3 ENG Concepts guided surface vehicle deployed in Lake Maggiore 15 Figure 4 DOQQ map of B ayboro Harbor showing (a) m/z 91 (toluene) signal i ntensity mapped over the GPS plotted track of the guided surface vehicle 19 Figure 5 DOQQ map of deployment area in Lake Maggiore showing (a) m/z 32 (oxygen) signal intensity mapped over the GPS plotted track of the guided surface vehicle 20 22 Figure 6 D ata acquired by the 200 amu underwater MS system during the Bayboro Harbor deployment 21 Figure 7 D ata acquired by the the 200 amu underwater MS system deployed in Lake Maggiore (CO 2 and O 2 ). 22 Figure 8 Data acquired by the 200 amu underwater MS system deployed in Lake Magg iore ( CO2 O 2 and Ar). 23 Figure 9 Saanich Inlet location and bathymetry. 26 Figure 10 Un derwater MS system readied for deployment in Saanich Inlet from the deck of the Marine Sciences Vessel John Strickland 27 Figure 11 Temperature profile ( o C) from surface to 200 meters depth, d uring MS system deploymen t in Saanich Inlet on April 13 2004 30 Figure 12 Methane ( m/z 15) profile, from surface to 200 meters depth, during MS system deployment in Saani ch Inlet on April 13, 2004. 30 Figure 13 Carbon dioxide ( m/z 44) profile, from surface to 200 meters depth, during MS system deployment in Saani ch Inlet on April 13, 2004. 31 Figure 14 Oxygen ( m/z 32) profile, from surface to 200 meters depth, during MS system deployment in Saa ni ch Inlet on April 13, 2004. 31
iv Figure 15 Hydrogen sulfide and oxygen isotope ( m/z 34) profile, from surface to 200 meters depth, during MS system deployment in Saan ich Inlet on April 13, 2004. 32 Figure 16 Mass spectrometer calibration plot for dia tomic nitrogen (N 2 ) showing the ion intensity associated with three concentrations of nitrogen gas in Hillsborough River water. 36 Figure 17 Mass spectrometer calibration plot for diatomic oxygen (O 2 ) showing the ion intensity associated with three concentrations of oxygen gas in Hillsborough River water. 36 Figure 18 Mass spectrometer calibration plot for methane (CH 4 ) showing the ion intensity associated with three concentrations of methane gas in Hillsborough River water. 37 Figure 19 Mas s spectrometer calibration plot for carbon dioxide (CO 2 ) showing the ion intensity associated with three concentrations of carbon dioxide gas in Hillsborough River water. 37 Figure 20 Mass spectrometer calibration plot for argon (Ar) showing the ion intensity associated with three conce ntrations of argon gas in Hillsborough River water. 38 Figure 21 Calibration plots correlating mass spectrometer ion intensity data with dissolved gas concentrati on s. 40 Figure 22 Aerial photograph of the norther n Tampa Bay area with the portion of the Hillsborough River upon which the MS system was deployed on February 13, 2007. 42 Figure 23 Dissolved oxygen concentrations in a section of the lower Hillsborough River as recorded by the MS system during a 2,6 40 m et er transect in February 2007. 43 Figure 24 Dissolved carbon dioxide concentrations in a section of the lower Hillsborough River as recorded by the MS system during a 2,640 m eter transect in Febr ua ry 2007. 44 Figure 25 Dissolved methane concentrations in a section of the lower Hillsborough River as recorded by the MS system during a 2,640 m et er transect in February 2007. 46 Figure 26 Oxygen data as recorded by the MS system and the YSI DO sensor d uring the Hillsboro ugh River deployment of February 2007. 47
v M easurement of Dissolved Gas Concentrations in Natural Waters Utilizing an In Situ, Membrane Inlet Linear Quadrupole Mass Spectrometer Peter Wenner ABSTRACT Since its creation in the late 19 th century mass spectrometry has evolved into one of the most versatile analytical methods in science. To chart this evolution this thesis includes a historical overview o f mass spectrometry and a description of the role of mass spectrometry in oceanography The development and deployment of underwater mass spectrom eter s (UMS) at the University has made possible the collection of real time data with greatly increased spatial and temporal density. T he UMS was deployed via both remotely guided surface vehicle s (GSV) to monitor a suite of dissolved gases and volatile organic compounds in saltwater and freshwater environments. Spectrometer data in Lake Maggiore, Florida were acquired at a rate of 0.7 3.6 seconds/sample for 2 3 hours The resulting multi analyte spectrometer data were recorded in real time with the Global Positioning System (GPS) observations of an associated surface vehicle and transmitted to a remote laptop computer via a wireless Ethernet link. These data were merged to create high resolution maps of chemical distributions. Of particular interest were the co varying oxygen and carbon dioxide mass spectrometer signals diagnostic of photosynthesis respiration proces ses, that were collected over a 10,8 0 0 square meter area of the lake. The UMS was also deployed on a shipborne hydrowire in Saanich Inlet a 200 meter deep fjord in the western Canadian province of British Columbia. The
vi concentrations of a broad suite o f dissolved gases were monitore d on both downcast and upcast over a total depth range of 20 0 meters. Spectrometer da ta w ere acquired at a rate of 4.2 seconds/sample for the duration of th e deployment. M ass spectrometer signals diagnostic of reduced speci es (CH 4 H 2 S,) in the anoxic waters of the inlet below a depth of 100 meters were consistent with previous The UMS was deployed on a remotely guided surface vehicle on the Hillsborough River in central Hillsborough County S pec trometer data were acquired at a rate of 0.7 seconds/sample, and geographic location was recorded by an onboard GPS during a 2,640 meter transect of the river. Prior to the deployment the mass spectrometer was calibrated using certified gas stand ards. The calibration experiments correlated mass spectrometer ion intensity data with dissolved gas concentrations whereupon the mass spectrometer data collected during the deployment were reported in units of micromole/kilogram (mol/kg). The mas s spectrometer recorded changes in gas concentrations associated with changing physical conditions and biological activity along the 2,640 meter s of the river that was transited by the GSV.
1 Chapter One Introduction In the more than 100 years since its discovery, mass spectrometry has become one of the most versatile analytical methods in science. In the first half of the 20 th century investigations utilizing the mass spectrometer yielded important discoveries in physics. The discovery of isotopes, and their abundances, as well as the separation of uranium 235 from uranium 238 are examples of the early contributions made by th is analytical technique. The development of novel inlet systems, mass analyzers and ion detectors soon made the mass spectrometer an essential analytical tool throughout many scientific disciplines. M ass spectrometry has been widely used in oceanog raphic studies to better Inductively coupled plasma mass spectrometry (ICP MS) is used to analyze trace metal concentrations in seawater. Isotope ratio mass spectrometry ( IRMS) has been employed to determine isotopic 16 O/ 18 O fractionation in the tests of marine planktonic organisms to reconstruct historical temperature records. D evelopment of membrane introduction m ass spectrometry (MIMS) in 1963 (Hoch & Kok, 1963) is part icularly relevant to the work described in this thesis A membrane of polydimethylsiloxane (PDMS) isolates the internal vacuum of the mass spectrometer from the environment. The PDMS membrane is ideal for use in aquatic environments as it is permeable to dissolved gases and volatile nearly impermeable to water. The University of South
2 Florida underwater mass spectrometer (UMS) employs a PDMS membra ne inlet system allowing for in situ, real time measurement of dissolved gas This thesis describes four deployments of the UMS. The first deployment took place in Lake Maggiore, a hypereutrophic, urban lake in St. Petersburg, Florida. The UMS was deployed on a remotely operated guided surface vehicle (GSV) out fi tted with an onboard GPS system to monitor spatial distributions of dissolved gas concentrations. During post processing GPS and mass spectrometer data were meshed to produce geo referenced maps of ion intensity data along the track navigated by the GSV. M ass spectrometer ion intensity data were displayed as a color scale. The color scale was than superimposed upon the GPS plot to produce a plot of ion intensity vs. position on the place in the marine waters of Bayboro Harbor/Salt Creek which borders the University of South Florida campus in St. Petersburg, FL. This deployment was conducted in the same way as was the Lake Maggiore deployment, the exception being that the mass spectr ometer monitored for referenced maps plotting both the GPS track and mass spectrometer ion deployment. The third deployment took place in Saanich Inlet, located on the eastern shore of Vancouver Island, British Columbia, Canada. A series of deployments, to depths of up to 205 meters, were conducted over a four day period. Mass spectrometer ion intensity data were collected for a suite of dissolved gases and plotted to as a function of depth In t he fourth deployment the mass spectrometer was mount ed aboard the GSV along a section of the Hillsborough River in Tampa, Fl orida. The GSV navigated a 2,640 m e ter
3 portion of the river while the mass spectrometer collected data for argon, carbon dioxide, oxygen, methane and nitrogen. A series of calibration experiments were conducted prior to the deployment. These experiments correlated mass spectromet er ion intensity data directly to gas concentrations such that mass spectrometer data collected during the Hillsborough River deployment could be reported in units of mol/kg.
4 Chapter 2 A Brief History of Mass Spectrometry In 1898 W. Wien reported that the motion of ions could be deflected by electric and magnet ic fields (Wien, 1898) The true potential of this discovery began to be realized during a seri es of experiments conducted by J.J. Thom son from 1910 to 1920 with a parabola mass spectrograph (Thomson, 1912) Thom to separate ions based on their mass to charge ratio ( m/ z ) (Thomson, 1913) This was done by passing a collimated, positively charged ion beam through electrostatic and magnetic f ields of known magnitude s The ion beam was deflected both vertically and horizontally by the combined fields and the path trave rs ed by an ion in t his field was dependent upon ion m/z and velocity (Thomson, 1913) Ions impact ed a photog raphic plate where they were permanently recorded for later analysis. The instrument description is derived from the parabolic impact pattern s left on the photographic plate by ions of the same m/z ratio bu t differing velocities. This instrument was used in Thomson 1912 discovery of the mass 22 isotope of n eon. Though the parabola mass spectrograph represented a spectacular advance in instrumental analysis it lacked th e ability to focus an ion beam the only way to improve the resolving power of a mass spect rometer. Resolvin g power is the mea sure of a n (Smith, 1999) For ions with identical charge ( z ) but similar mass ( m ) separation (i.e., resolution) can be difficult. In 1919 F.W. Aston incorporated a velocity filter in his mass spectrograph allowing only ions of a specific velocity to enter the flight chamber (Aston, 1919) This velocity
5 focusing capability produced ion impacts along a straight line on photographic plate s thus eliminating the parabola produced During this same time period A.J. Dempster achieved direction focusing whereby deflection of ions through 180 o in a magnetic field provided a resolving power of 1 part in 100 (Dempster, 1918) Another of Dem use of a quadrant electrometer detector distinguished this instrument as the first true mass spectrometer. Dempster went on to d iscover many isotopes and determine their abundances. Bainbridge further conceptual improvement s comb ining both velocity and direction focusing created the double focusing mass spectrograph (Bainbridge, 1933) Applying the principles of their earlier theoretical work Mattauch and Herzog (Mattauch & Herzog, 1934) developed a mass spectrograph that was doub le focusing for all masses (Mattauch, 1936) By careful consideration of electric and magnetic field theory and manipulation of instrument geometry Mattauch and Herzog were able to construct an instrument that detected a ll ions as they impacted a linear plane photographic plate Of critical importance in this design was the location of the photograp hic plate at the exit pole boundary of the magnetic field ( Roboz, 1968) In 1940 A.O Nier introduced his sector an alyzer which served as the design template for many future instrument s (Nier, 1940) Nier replaced the large, heavy electromagnet neces sary to deflect an ion 180 o with a small w edge sector electromagnet that deflected ion path s by 60 o (Dulski, 1999) This design innovation reduced the weight and power consumption of the electromagnet and placed the ion source and detector (deLaeter, 2001) One of these instruments was used by Ni er in the early 1 separate nanogram quantities of
6 uranium 235 from uranium 238 (Nier, et al., 1940 ). The Calutron, a three story high 235 from uranium 238 for production of the first atomic bomb (Smith, 1947) and constructed mass spectrometers for commercial use These early commercial instruments were used primarily by chemists in the petroleum industry to measure the abun dances of small hydrocarbons in process streams (Griffiths, 2008) D uring this p eriod researchers in fields other than physics came to realize the quantitative analytical potential of mass spectrometry. Physicists at this time were still primarily intere sted in high precision isotope mass spectrometry but their scope of interest began to broaden into such areas as fragmentation patterns, ionization and theoretical molecular modeling (Diaz, 1999) In 1946 W.E. Stephens built the first time of fligh t (TOF) mass spectrometer (Stephens, 1946) Ions emanate d from a pulsed source and we re separated according to their m/z values as they traverse d a field free pat h of known length. Provided that ions leave the source at the same time and with the same energy thus the pulsed ion source, lighter ions travel the path from source to detector in less time than heavier ions. In 1948 S.A. Goudsmit developed a helical path mass spectrometer based on the orbital period of an ion circl ing in a uniform magnetic field (Goudsmit, 1948) This instrument which separated ions of different mass by measuring the time it took for different ions to travel a set number of revolutions along a helical flight pa th imposed by a uniform magnetic field. As in a linear time of flight instrument heavier ion s require a greater period of time to travel a given distance than do
7 of approximately 0.01 sec w hich yielded a mass accuracy of 10 3 amu. It was an instrument of this type that was used to measure heavy elements such as 208 Pb and 209 Bi (Hays, et. al., 1951). The continuing evolution of ion cyclotron mass spectrometry culminated in 1974 with the deve lopment of Fourier transform ion cyclotron resonance (FT ICR) which represented a major breakthrough in the field of mass spectrometry (Comisarow & Marshall, 1974). Investigation of ion interactions in two and three dimensional quadrupole fields b y Wolfgang Paul (Paul & Steinwedel, 1954) and coworkers in 1953 led to the development of the quadrupole mass analyzer. Alt hough early quadrupole instruments had poor resolving power and low mass range through use of sweeping electric potentials (Gross, 2004) they offered advantages of high transmission, light weight, small size, comparatively low price and high scan speed In 1957 t he need to separate and analyze complex mixtures led to one of the most powerful tools in instrumental analysis: coupled gas chromatograph y mass spectrometry (Gohlke & McLafferty, 1993) The success of this approach led to the development of other hyphenated methods : liquid chromatography mass spectrometry (Ar drey, 2003) capillary zone electrophoresis mass spectrometry (Tanaka, et. al., 1998) and supercritical fluid chromatography mass spectrometry (Smith, et. al., 1982) The coupling s of instrumental techniques further broadened the range of disciplines to w hich mass spectrometry could be applied. C ontinued improvements in in strument design and performance moved m ass spectrometry beyond the realm of physics i nto biology, chemistry, geology and materials science as well as many other disciplines. Of p articular importance were developments
8 taking place in ionization techniques and ion detection. Most early mass spectrometers utilized electron impact ionization (EI) to create ions. This technique limited the suite of compou nds that could be analyzed by mass spectrometry In 1984 (Fenn & Yamashita, 1984) introduction of electrospray ionization made direct transfer of ions from solution to the gas phase possible, allowing for analysis of large, non volatile molecules such as proteins and nucleic acid polymers (Amad, et. al., 2000) Matrix assisted laser desorption ionization (MALDI) enabled the mass spectrometric analysis of large (100,000 amu) biological molecules (Karas, et. al. 1987) In this development sample s crystalliz ed within an organic mat rix were evapora ted and ionized by a laser pulse. It is with MALDI TOF instruments that peptide (Spengler, et. al., 1991) and oligonucleotide (Wang, et. al., 1997) sequencing was first accomplished. Beyond what has been presented here is a vast arr ay of mass spectrometric techniques and applications that make mass spectrometry one of the most versatile methods of instrumental analysis in all of science.
9 Chapter 3 The Role of Mass Spectrometry in Oceanography With successive refinements, as discussed in the previous chapter m ass spectrometry became an impor tant analytical technique in interdisciplinary realm s and began to make important contributions to biological, chemical and geological oceanography (Short, et. al., 2009). T raditional oceanographic sampling techniques involve sample collection in the field, transport of samples to a land based laboratory, chemical processing and subsequent analysis Modern instrumentation, including portable ship based mass spectrometers now allow prompt analysis of samples at sea. I n the area of biological oceanography mass spectrometry has been utilized for measurement of specific organic molecules in assessments of marine food web dynamics (Volkman & Tanoue, 2002) for determinations of organic matter provenance (Peters, et. al., 2005) and for assessment of past sea surface temperatures in paleoclimate reconstruction (Wakeham, 1993) Inductively coupled plasma mass spectrometry (IC P MS) has become an essential tool for multi elemental analysis of trace element concentrations in seawater. Trace elements play an important role as micronutrients and toxicants for a variety of marine organisms (Morel & Price, 2003) ICP MS is uniquely suited as well to provide input for models of estuarine, benthic, coastal and pelagic mass transport processes. Of particular importance to the disciplines of geological, biological and chemical oceanography was the development of isotope ratio m r (Urey, 1947)
10 development of the isotope ratio mass spectrometer in 1940 (Nier, 1940) led to the use of isotopic ratios to understand chemical pathways in t he environment Application of these technique s has enabled climatologists to co nstruct a Membrane introduction mass spectrometry (MIMS) can be used to make high frequency, real time measurements of both dissolved gases (Hemond & Camilli, 2002) (Kana, et. al., 1994) solved gases and volatile organic com pounds (VOCs) in aquatic environments pr ovide s important information about dyna mic biogeochemical processes (Short, et. al, 2001) Submersible MIMS systems have been developed to enable the collection of data with high spatial and temporal resolution (Wenner, et. al., 2004) The MIMS system de veloped at the University of South Florida will be described in the next section of this thesis
11 The University of South Florida Underwater Mass Spectrometer The underwater MS system described here is similar in construction to the 100 amu underwater linear quadrupole system repo rted earlier (Short, et. al., 1999) The 2 00 amu instrument uses a linear quadrupole mass analyzer (Transpector CPM 200 Residual Gas Analyzer, Inficon, Inc., Syracuse, NY, USA) with a closed ion source. A new high pressure membrane introduction system, designed and m achined at the Center for Ocean Technology, extends the depth capability of the instrument well beyond the 30 m limit imposed by the previous design (Short, et al., 2001) The high pressure membrane introduction probe has been pressure tested to a depth of 250 m eters. A schematic of the overall underwater MS system is shown in Figure 1 and details of the high pressure membrane probe are shown in Figure 2. Other minor diff erences between our current and previo us designs are noted below (Short, et. al., 1999) Figure 1. Schematic of the 200 amu linear quadrupole underwater MS system.
12 The underwater MS syste m shown in Figure 1 is modular. The main pressure v essel houses the primary vacuum system, mass analyzer, high pressure sampling system and control electronics. A separate, smaller pressure vessel contains two diaphragm pumps. These pumps provide the backin g vacuum for the turbomolecular/molecular drag pump. Table 1 summarizes the critical specifications of the system. The housing of the high pressure membrane probe (Figure 2) was machined from 3.18 cm diameter 316 stainless st eel stock. Tubing ( 1.59 mm (OD ) by 1.02 mm (ID) 316 stainless steel) was coiled around the exterior of the membrane probe housing (5 revolutions) and silver soldered in place to provide good thermal contact between coils and ho using. Water flowed through this tubi ng and then entered the central channel of the probe h ousing where it flowed over Figure 2. Schematic showing detail of the membrane housing (left) and membrane probe with the internal compression spring (right). (Graphic contributed by Richard Hildebrand of Johns Hopkins University, Applied Physics Laboratory)
13 the PDMS membrane. A 1 mm diameter spring (Gutekunst and Co., Metzingen, Germany) was inserted into the interior of the PDMS membr ane to provide internal support at increased pressures. T wo cartrid ge heaters (Hottwatt, Danvers, MA, USA) inserted into the back of the membrane probe housing heated the probe housing and consequently the water within the stainless steel tubing. The probe temperature w as monitored via a thermocouple (Minco, Minneapolis, M N) within the probe housing. An onboard micro controller, deve loped in house, regulated power to the cartridge h eaters and stabilized the probe temperature. The temperature of the sample water was adjusted to optimize an alyte pervaporation relative to wat er vapor load across t he membrane A magnetic piston pump (Inducti ve Pump Corporation, Barneveld, NY, USA), constructed of stainless steel and pressure tested to 60 atm ( ~ 600 m), pu lled ambient water into the sample tubing, through the membrane probe and back to the environment. The piston pump could be adju sted to provide flow rates of 1 10 mL/min. Table 1. Operational specifications of the 200 amu underwater MS system Type Linear quadrupole MS Mass Range 200 amu Inlet System Membrane introduction Power Consumption 105 W Operating Voltage 24 V Deployment Time 10 14 days Diameter 19 cm (7.5 in.) Length 114 cm (45 in.) Weight 33 kg (72.7 lb) Depth 200 meters
14 Chapter 4 In situ Mass Spectrometer Deployments Bayboro Harbor and Lake Maggiore Data presented in this section were taken from two deployments of the underwater MS. In both cases, the instrument was deployed aboard a remotely guided surface vehicle GSV) (ENG Concepts St. Petersburg, FL, USA) (Figure 3). The 200 amu instrument was first deployed in Bayboro Harbor adjacent to the USF campus in St. Petersburg, FL, USA. The second deployment was in Lake Maggiore, a 1.41 km 2 hypereutrophic, urban lake within the city limits of St. Petersburg, FL, USA ( CHM Hill, 1991) In both deployments, the instrument was suspended beneath the vehicle at a depth of 0.30 m and water at this depth was sampled at 5 mL/min. The forward speed of the v ehicle was maintained at approx imately 0.5 m/s and vehicle movements we re controlled through an rf link using a hand held controller. A GPS (Trimble BD 112) mounted aboard the surface vehicle recorded the position (latitude/longitude) of the vehicle every second. A 803.11 wireless connection enabled communication between th e underwater MS and a shore side laptop computer. Vehicle location plus MS data and operating parameters were monitored in real time on the same laptop computer. During the Bayboro Harbor deployment, MS and GPS power was provided by two 24 V DC lead acid battery packs. Battery capacity limited the deployment to 2 3 hours. A gasoline powered 1000 Watt Honda generator secured to the surface vehicle provided power to the underwater MS and GPS during the Lake Maggiore deployment. The use of this generator, with a 2.27 liter fuel capacity, extended the deployment time to 5 hours.
15 During the Bayboro Harbor deployment, the surface vehicle navigated a track across the Harbor and transited 150 meters up Salt Creek, which empties into the Harbor. The underwater MS monitored a total of seven ions as described in Table 2. Three of these ions (78, 91, 92) are diagnostic of com ponents in gasoline and can enter the environment via fuel spills or internal combustion engine exhaust. Another three ions (47, 83, 85) are diagnostic of chlorinated compounds found in tap water, and the seventh (62) is associated with dimethyl sulfide, a compound produced by certain planktonic organisms (Andreae, et. al., 1994) Each of the seven ions was scanned for 512 milliseconds (ms) resulting in a total scan time of 3.6 seconds (s) The MS used an electron multiplier ion detector operated at 110 0 volts (V) to record ion intensities. Sample transit time (the Figure 3. ENG Concepts guided surface vehicle deployed in Lake Maggiore
16 elapsed time between sample acquisition and actual measurement) was 242 seconds. All MS data were adjusted to properly correlate MS data with geographic position. Table 2. Mass to charge ( m/ z ) values, diagnostic of volatile organic compounds scanned by the 200 amu underwater MS during deployment in Bayboro Harbor. m/z Value Compound Associated with 47 Chloroform Tap water 47 Dimethyl sulfide (DMS) Plankton 62 Dimethyl sulfide Plankton 78 Benzene Internal combustion engine exhaust 83 Chloroform Ta p water 85 Chloroform Tap water 91 Toluene Internal combustion engine exhaust 92 Toluene Internal combustion engine exhaust During the Lake Maggiore deployment, the underwater MS monitored ions diagnostic of nitrogen ( m/z 28, 29), oxygen ( m/z 32, 34), carbon dioxide ( m/z 44, 45, 46) and argon ( m/z 40) as described in Table 3. Each of these masses was scanned for 64 ms. The time required to scan and record the 8 ions was 650 ms. The electron multiplier detector was operated at 900 V to record ion intensities. The sample transit time for this experiment was 270 s. The longer sample transit time for this deployment resulted from a longer sam ple inlet tube. The guided surface vehicle navigated transects in a small embayment in the southeast corner of Lake Maggiore that encompassed approximately 53,000 m 2 The overall track of the measurements was approximately 1.5 km. MS observations were s ynchronized with GPS measurements prior to deployment. To coordinate the two streams of data, a digital filter (PeakFit 4.0, Systat Software Inc., Richmond, CA, USA) was used to match the frequencies of MS and GPS observations.
17 The two datasets for each experiment were then combined into one array, with 242 and 270 seconds subtracted from the time stamps of the Bayboro and Lake Maggiore MS data, respectively, to account for the sample transit times mentioned earlier. The resulting Table 3. Mass to charg e ( m/z ) values, diagnostic of dissolved gas species, scanned by the 200 amu underwater MS during the Lake Maggiore deployment. m/z Value Detected Ion s Isotopic form 28 N 2 + 14 N 14 N 29 N 2 + 15 N 14 N 32 O 2 + 16 O 16 O 34 O 2 + 16 O 18 O 40 Ar + 40 Ar 44 CO 2 + 12 C 16 O 16 O 45 CO 2 + 13 C 16 O 16 O 46 CO 2 + 12 C 18 O 16 O spliced datasets were plotted on digital orthographic quarter quad (DOQQ) maps representing the regions of the Bayboro Harbor and Lake Maggiore deployments (Figures 4 and 5, respectively). The positional error s of the mapped chemical tracks including all sources of error (GPS, MS sampling offset, and map resolution) are estimated to be 3 m. Further reduction of this error i s difficult as t he two primary contributors, GPS accuracy and DOQQ resolution, are factors over which there are no experimental controls. The MS data for m/z 62, 91 and 92 from the Bayboro Harbor deployment were plotted (Figure 6 ) as ion intensity vs. time of day. Ion intensity traces for m/z 91 and 92 exhibited very similar behavior. S ignificant increases in intensity at approximately 11:35 a.m. and 11:50 a.m. indicated enhanced toluene concentrations. The m/z 91:92 ratio is c lose to the value for toluene reported in the National Institute of Standards and Technology (NIST) database ( NIST, 1998) The m/z 62 sign al had two significant peaks, th e first of which occurred at approximately 11:55 a.m. and t he second, slightly smalle r (a) (b)
18 at approximately 11:58 a.m. The m/z 62 trace cannot be attributed to a specific chemical species. The MS fingerprint of dimethyl sulfide has significant peaks at m/z 47 and 62, with the m/z 62:47 ratio equal to approximately one. The Bayboro Harbor d ata showed no increase in the m/z 47 ion intensity in the region of the m/z 62 ion intensity peaks. Additional analyses are required to identify the species responsible for m/z 62 peaks in the absence of peaks at m/z 47 It should be noted that underwate r MS, as applied here, The instrumentation is not currently capable of positively identifying individual VOC species in the environment. Lake Maggiore is hypereutrophic because of extreme nutr ient loading from urban storm water run off, and at the time of this investigation, was scheduled for extensive dredging to remove nutrient laden sediment. The underwater MS was deployed to acquire data relating to dissolved gas concentrations in a stres sed freshwater environment. MS data for m/z 32 (oxygen) and m/ z 44 (carbon dioxide) from the Lake Maggiore deployment are shown in Figure 7 as ion intensity vs. time of day. Ion intensities for a second isotopic form of oxygen ( m/z 34) and carbon dioxide ( m/z 45) were also recorded to verify identifications of the m/z 32 and m/z 44 peaks (Figure 8). Between approximately 10:15 a.m. and 11:03 a.m. the m/z 32 and m/z 34 signal intensities show distinct inverse relationships to those of m/z 44 and m/z 45. There are 13 significant minima in m/z 32 ion intensity in this region of the data. Eight of these minima correspond to ion intensity maxima in the m/z 44 signal. This co variation in the m/z 32 and m/z 44 signal intensities could b e evidence of fine scale spatial variations in photosynthesis and respiration by lake organisms (Jumars, 1993) In those regions of the lake where respiration dominates photosynthesis, one would expect to see a decrease in
19 dissolved oxygen coupled with an increase in dissolved carbon dioxide. Additional work is necessary to confirm that this co variation in the m/z 32 and m/z 44 signal intensities Figure 4. DOQQ map of Bayboro Harbor showing (a) m/z 91(toluene) signal intensity mapped over the GPS plotted track of the guided surface vehicle; (b) m/z 62 (dimethyl sulfide) signal intensity mapped over the GPS plotted track of the guided surface vehicle.
20 Figure 5. DOQQ map of deployment area in Lake Maggiore showing (a) m/z 32 (oxygen) signal intensity mapped over the GPS plotted track of the guided surface vehicle; (b) m/z 44 (carbon dioxide) signal intensity mapped over the GPS plotted track of the guided surface vehicle.
21 Figure 6. Data acquired by the 200 amu underwater MS system during the Bayboro Harbor deployment. The m/z 91 and m/z 92 signals are diagnostic of toluene and m/z 62 is diagnostic of dimethyl sulfide. is attributable to organic/inorganic transformations of particulate/dissolved carbon (organic carbon/dissolved CO 2 ) with attendant covariations in dissolved oxygen. The ion intensity for argon ( m/z 40) was also recorded during the Lake Maggiore deployment (Figure 8). Argon is biologic ally inactive and its ion intensity should remain stable during deployments in waters that are approximately isothermal. The absence of significant changes in the m/z 40 signal indicated that all MS system components were functioning normally. The underwater MS data for m/z 91 (Figure 4a) and m/z 62 (Figure 4b) were displayed on a DOQQ map of Bayboro Harbor. Mapping of the Bayboro Harbor data correlated the m/z 62 and m/z 91 time series data in Figure 6 to geographic locations within the harbor All m/z values that were monitored by the underwater MS could be mapped as well. As discussed above, the identity of the compound responsible for the increase in m/z 62 was impossible to determine without additional analyses. Both
22 areas of increased m/z 91 and m/z 92 intensity (Figure 6) are likely toluene signals associated with the exhaust of boats that had traversed the path of the surface vehicle/MS. The Lake Maggiore underwater MS data for carbon dioxide (Figure 5a) and oxygen (Figure 5b) were displayed on a DOQQ map of Lake Maggiore. The pattern of oxygen and carbon dioxide concentrations in Lake Maggiore were associated with specific locations at the time of the measurements. General conclusions about the relative significance of benthic and water column biogeochemical contributions to the observed signals require additiona l supporting data (e.g., chlorophyll, particulate organic carbon Figure 7. Da ta acquired by the 200 amu underwater MS system in Lake Maggiore. The m/z 32 signal is diagnostic of oxygen and m/z 44 is diagnostic of carbon dioxide.
23 Figure 8. D ata acquired by the 200 amu underw ater MS system deployed in Lake Maggiore. The signals of two isotopic forms of carbon dioxide ( m/z 44 and 45) oxygen ( m/z 32 and 34) and argon ( m/z 40) are shown. Degradation of all signals between time 9:40 a.m. and 10:10 a.m. is caused by out gassing associated with the MS start up. measurements, etc.) and additional underwater MS calibrations that allow transfor ma tion of ion intensities to in situ gas concentrations. In situ measurements by an underwater MS can eliminate many problems inherent in traditional sampling methods and provide data with spatial and temporal resolutions that are difficult to obtain by other means. High density data mapping can be a powerful t ool in both aquatic biogeochemical process studies and studies aimed at the identification of point sources for anthropogenic chemicals. The PDMS membrane is highly permeable to
24 small, volatile, non polar molecules, making it an ideal interface for the ty pes of compounds examined in this section. Development of different types of MS sample introduction interfaces will be required for analysis of other types of aqueous chemical species. Future plans include the development of a solid phase micro extractio n (SPME) interface for analysis of larger, somewhat more polar and less volatile molecules. This will greatly expand the types of compounds that are detectable using underwater MS. Addi tional work was undertaken in February 2007 to establish quantitative relationships between MS ion intensities and analyte concentrations. The results of some portions of that work will b e presented later in this chapter
25 Saanich Inlet Saanich Inlet is located on the eastern shore of Vancouver Island in the Canadian provinc e of British Columbia (Figure 9a) The inlet has a surface area of 65 km 2 and a maximum depth of 225 meters (Binqiu, 1989). A sill located at a depth of 70 meters at the northern mouth of the inlet isolates basin waters (Figure 9b) such that for about 8 months of the year the bottom waters are anoxic, contain no nitrate or nitrite, and significant concentrations of free hydrogen sulfide (H 2 S) are present (German & Elderfi eld, 1989). In April 2004 members of the USF Mass Spectrometry Group col laborated with colleagues from the University of Victoria (Victoria, British Columbia) and the Wood s Hole Oceanographic Institute ( Woods Hole, Massachusetts ) in evaluations of in situ instrumentation for measure ment of dissolved gas concentrations, specifically methane. During the dates of April 13 16, 2004 the underwater MS system was deployed in Saanich Inlet four times. D eployments took place aboard the University of Victo Marine Sciences Vessel John Strickland T he MS system along with a CTD probe (Applied Microsystems Ltd, Sydney, British C olumbia, Canada) that provide d temperature salinity and depth measurements, was coupled to a steel cable (Figure 10) and lowered to depth using an onboard deck crane A 220 meter tether connected the MS system to both the real time data transmission and overall system control, and delivered power to the MS during a deployment of the MS system to a depth of 200 meters on April 13, 2004.
26 Figure 9. Saanic h Inlet location and bathymetry: (a) chart showing location of Saanich (a) (b) LATITUDE (decimal degrees) Location of April 13, 2004 MS deployment Lat: 48 o Lon: 123 o LONGITUDE (decimal degrees) Approximate path of MS downcast and upcast on April 13, 2004 deployment
27 Figure 10 Underwater MS system readied for deployment in Saanich Inlet from the deck of the Marine Sciences Vessel John Strickland The approximate location of the deployment , was latitude: 48 o o The m/z values used to monitor dissolved gas species for the April 13, 2004 deployment are listed in Table 4. During this deployment the electron multiplier was set to 900 volts and the mass spectrometer dwell time for each m/z was set to 64 ms resulting in a scan time of approximately 0.90 seconds AML Micro CTD 220 meter deployment tether provides MS power and real time data monitoring Pressure vessel containing the MS and peripheral components Instrument harness, mates MS to deck crane cable
28 Table 4. Mass to charge ( m/z ) values diagnostic of dissolved gas species scanned by the 200 amu mass spectrometer during the April 13, 2004 Saanich Inlet deployment. m/z V alue Detected Ions I sotopic F orm 12 C + 12 C 14 CH 2 + N + 12 C 1 H 1 H, 14 N 15 CH 3 + N + 12 C 1 H 1 H 1 H, 15 N 16 CH 4 + O + 12 C 1 H 1 H 1 H 1 H, 16 O 28 N 2 + 14 N 14 N 29 N 2 + 14 N 15 N 32 O 2 + 16 O 16 O 34 O 2 + H 2 S + 16 O 18 O, 1 H 1 H 32 S 40 Ar + 40 Ar 44 CO 2 + 12 C 16 O 16 O 45 CO 2 + 13 C 16 O 16 O Figures 11 through 15 display the vertical trends associated with temperature, methane ( m/z 15), carbon dioxide ( m/z 44), oxygen ( m/z 32) and hydrogen sulfide plus 18 O 16 O ( m/z 34). The most abundant peak in the mass spectrum of methane occurs at m/z 16. It is not practical, however, to monitor this value for trends in methane concentration as the m/z 16 signal is overwhelmed by atomic oxygen derived from fragmentation of CO 2 H 2 O and O 2 at the MS ion source The next most abundant peak in the methane mass spectrum occurs at m/z 15 corresponding to the CH 3 ioniza tion fragment of methane. The absence of substantial interferences makes m/z 15 the best ion for monitoring trends in methane concentration. The vertical temperature profile in Figure 11 shows a ~ 5.5 o C drop in temperature (from 13.5 o C to 8.5 o C) in the first 25 meters. From 8 5 meters to 12 5 meters the temperature increased from 8.5 o to 9.5 o C. Below 125 meters the temperature was essentially constant In Figure 12 the m/z 15 ion intensity (diagnostic of CH 4 ) remains fairly constant through the first 100 meters A sharp incr ease in m/z 15 intensity is seen below 100 meters as increasingly anoxic conditions stabilize reduced species such as
29 methane and H 2 S A chemocline depth at 100 meter s is also seen in the ion intensity plot of ca rbon dioxide ( Figure 13 ) A gradual in crease in m/z 44 intensity between 7 0 and 100 meters gives way to a sharp in crease at 100 meters. Between 100 and 125 mete rs the carbon dioxide plot increase s more gradually and then decreases between 125 and 200 meters, possibly indicating that methanoge nic bacteria are converting CO 2 to methane in the bottom waters and sediments of the fjord Figure 14 displays the ion intensity plot associated with oxygen at m/z 32 T he ion intensity exhibits a steady decline with increasing depth between 25 an d 100 meters, with only small changes below this depth The ion intensity plot for H 2 S plus 18 O 16 O (both m/z 34 ) is displayed in Figure 15. The m/z 34 trend closely tracks that of m/z 32 from the surface to a depth of 140 meters. From 140 to 200 meters t he m/z 34 intensity plot steadily increases whereas the m/z 32 plot remains relatively constant. This increase in the m/z 34 ion intensity plot is attributable to the presence of H 2 S in the deep waters of Saanich Inlet. Although the PDMS membrane of th e MS system inlet is relatively impermeable to H 2 S in appreciable quantities, much like H 2 O, it will migrate across the PDMS membrane and enter the mass spectrometer. The data presented here from the April 2004 deployments of the USF MS system demo nstrated MS system capabilities to provide data of unprecedented spatial an d temporal density at depths up to 200 meters. The ion intensity measurements obtained during this work allowed qualitative comparison of trends in gas concentrations but not absol ute concentration measurements T he next section of this thesis describe s MS calibration s that relate ion intensities directly to gas concentrations.
30 Figure 11. Temperature profile (Deg C) from surface to 200 meters depth, during MS system deployment in Saanich Inlet on April 13, 2004. Figure 12. Methane ( m/z 15) profile, from surface to 200 meters depth, during MS system deployment in Saanich Inlet on April 13, 2004. 1 2 Depth (m) Temperature Profile April 13, 2004 Temperature (Deg C) 0 50 100 1 5 0 2 00 8 9 10 1 1 1 2 1 3 1 4 1 4 Depth (m) Methane Profile April 13, 2004 m/z 15 (amps x 10 14 ) 0 50 100 1 5 0 2 00 2 3 4 5 6 7 8
31 Figure 13. Carbon dioxide ( m/z 44) profile, from surface to 200 meters depth, during MS system deployment in Saanich Inlet on April 13, 2004. Figure 14. Oxygen ( m/z 32) profile, from surface to 200 meters depth, during MS system deployment in Saanich Inlet on April 13, 2004 Depth (m) Carbon Dioxide Profile April 13, 2004 m/z 44 (amps x 10 9 ) 0 50 100 1 5 0 2 00 0 0.5 1 1.5 Depth (m) Oxygen Profile April 13, 2004 m/z 32 (amps x 10 9 ) 0 50 100 1 5 0 2 00 0.5 .5 1 1.5 2 2.5 3
32 Figure 15. Hydrogen sulfide and oxygen isotope ( m/z 34) profile, from surface to 200 meters depth, during MS system deployment in Saanich Inlet on April 13, 2004. Depth (m) Hydrogen Sulfide/Oxygen Profile April 13, 2004 m/z 34 (amps x 10 11 ) 0 50 100 1 5 0 2 00 0 0.5 1 1.5
33 Calibration of Mass Spectrometer Data F ield deployments of the USF underwater MS system described in previous sections of this chapter, have shown that mass spectro meter s can be us ed in situ to produce data that a re diagnostic of biogeochemical p rocesses. However, dissolved gas data are of limited use unless MS observations are reported directly in terms of concentrations. In order to address this problem c ertified gas standards were used for calibration of the UMS. Subsequently, m ass spectromet er data obtained in field studies along a 2,640 m e ter section of the Hillsborough River were expressed in units of mol/kg. A portable calibration system was constructed for use with certified gas standards that were obtained from Airgas, Inc. (Radnor, PA). The composition of the calibration standards are given in Table 5 Equilibrations with certified gases, and subsequent calibrations, were conducted using two 2,000 mL volumetric flasks. Standardized dissolved gas solutions were produced in Flask One. This flask, containing approximately 1,500 mL of Hillsborough River water and a magnetic stir bar, was placed atop a waterproof magnetic stirrer. Both the flask and the magnetic stirrer were immersed in a constant temperature water bath. A t hermometer immersed in the water bath was used to monitor solution temperature. The opening of the flask was covered terminated just above the bottom of the flask and was fitted with a cylindrical stone gas stainless steel tube was approximate ly 7 cm in length and served as a pressure relief to
34 prevent super connected to a two way sampling valve which was connected to the MS system sample inlet. The second flask (Flask 2) cont ained approximately 1,500 mL of Hillsborough Table 5. Identification of gases contained in the three certified gas stand ards purchased from Airgas Inc. and their respective concentrations. Tank # Gas Gas Concentration (%) 1 Argon (Ar) Carbon dioxide (CO 2 ) Methane (CH 4) Nitrogen (N 2 ) Oxygen (O 2 ) 2.04 2.03 1.01 74.96 14.96 2 Argon (Ar) Carbon dioxide (CO 2 ) Methane (CH 4) Nitrogen (N 2 ) Oxygen (O 2 ) 5.07 0.62 3.01 85.34 5.95 3 Argon (Ar) Carbon dioxide (CO 2 ) Methane (CH 4) Nitrogen (N 2 ) Oxygen (O 2 ) 0. 50 0.30 4.98 92.20 2.00 River water that was open to the atmosphere. One end of a diameter stainless steel tubing was inserted into the water in the flask, t he other end was fitted to the same two way sampling valve to which the first flask was connected This configuration made it possible to instantaneously switch from one solution to the other while the mass spectrometer was actively scanning The mass spectrometer was calibrated one day prior to the deployment on the Hill sborough River A solution of standardized dissolved gas es was produced for each of the three certified gas standards T he first solution was produced from Tank 3, the second solution from Tank 1 and the final solution from Tank 2. At the beginning of th e
35 calibration run wa ter from Flask 2 was pumped through the mass spectrometer sampling system while gas from Tank 3 was bubbled into Flask 1. The solution in Flask 1 was allowed to equilibrate for 30 minutes at which point the flow of gas into the flask was stopped Immediately after the gas was turned off the two way sampling valve was switched so that the equilibrated dissolved gas standard in Flask 1 was pumped through the mass spectrometer sampling system. Each standard solution was sampled until the mass spectrometer signal stabilized and remained stable for at least 10 minutes. This process was repeated for the other two dissolved gas standards. Figur es 16 20 plot mass spectrometer ion intensity vs. scan number for all 5 gases during the three calibration runs. The ion intensity values in Figure s 16 20 span several hundred scans at each gas single ion intensity value associated with a specific gas conce ntration. The five values were chosen near the end of the equilibration run to ensure that the solution had reached equilibrium. For example: In Figure 16 intensity values for the plateau observed between scans 2449 and 2976 resulted from a solution equilibrated with a certified gas standard that was 74.96% nitrogen Of the 527 ion intensity data points in this range t he five values between scans 2906 and 2911 were used to produce an average intensity value of 3.56(10) 8 amps. Ion intensity value s for all remaining gases were related to gas concentrations in the same manner for each of the five gases being measured.
36 Figure 16. Mass spectrometer calibration plot for diatomic nitrogen (N 2 ) showing the ion intensity associated with three conce ntrations of nitrogen gas in Hillsborough River water. (Reported water temperatures are the temperatures at which each solution was equilibrated; atmospheric pressure assumed to be 101.325 kPa) Figure 17. Mass spectrometer calibration plot for diatomic oxygen (O 2 ) showing the ion intensity associated with three concentrations of nitrogen gas in Hillsborough River water. (Reported water temperatures are the temperatures at which each solution was equilibrated; atmospheric pressure assumed to be 101.325 kPa) 0.00E+00 5.00E 09 1.00E 08 1.50E 08 2.00E 08 2.50E 08 3.00E 08 3.50E 08 4.00E 08 4.50E 08 1 185 369 553 737 921 1105 1289 1473 1657 1841 2025 2209 2393 2577 2761 2945 3129 3313 3497 3681 3865 4049 4233 4417 4601 4785 4969 5153 5337 Intensity (Amps) Scan # N 2 Calibration for the Hillsborough River (m/z 28) 92.2000% N 2 Water Temp: 74.96% N 2 Water Temp: 85.345% N 2 Water Temp: 0.00E+00 2.00E 09 4.00E 09 6.00E 09 8.00E 09 1.00E 08 1.20E 08 1.40E 08 1.60E 08 1.80E 08 2.00E 08 1 185 369 553 737 921 1105 1289 1473 1657 1841 2025 2209 2393 2577 2761 2945 3129 3313 3497 3681 3865 4049 4233 4417 4601 4785 4969 5153 5337 Intensity (Amps) Scan # O 2 Calibration for Hillsborough River ( m/z 32) 2.000% O 2 Water Temp: 24.8 0 C 14.96% O 2 Water Temp: 24.5 0 C 5.949% O 2 Water Temp: 24.5 0 C
37 Figure 18. Mass spectrometer calibration plot for methane (CH 4 ) showing the ion intensity associated with three concentrations of methane gas in Hillsborough River water. (Reported water temperatures are the temperatures at which each solution was equilibrated; atmospheric pressure assumed to be 101.325 kPa) Figure 19 Mass spectrometer calibration plot for metha ne (CO 2 ) showing the ion intensity associated with three concentrations of methane gas in Hillsborough River water. (Reported water temperatures are the temperatures at which each solution was equilibrated; atmospheric pressure assumed to be 101.325 kPa) 0.00E+00 5.00E 10 1.00E 09 1.50E 09 2.00E 09 2.50E 09 3.00E 09 3.50E 09 1 257 513 769 1025 1281 1537 1793 2049 2305 2561 2817 3073 3329 3585 3841 4097 4353 4609 4865 5121 5377 Intensity (Amps) Scan # CH 4 Calibration for the Hillsborough River ( m/z 15) 5.000% CH 4 Water Temp:24.8 0 C 1.010% CH 4 Water 3.010% CH 4 Water 3.00E 09 8.00E 09 1.30E 08 1.80E 08 2.30E 08 2.80E 08 3.30E 08 1 184 367 550 733 916 1099 1282 1465 1648 1831 2014 2197 2380 2563 2746 2929 3112 3295 3478 3661 3844 4027 4210 4393 4576 4759 4942 5125 5308 Intensity (Amps) Scan # CO 2 Calibration for the Hillsborough River ( m/z 44) 0.3000% CO 2 Water Temp: 24.8 0 C 2.030% CO 2 Water Temp: 24.5 0 C 0.6260% CO 2 Water Temp: 24.5 0 C
38 Figure 20. Mass spectrometer calibration plot for argon (Ar) showing the ion intensity associated with three c oncentrations of methane gas in Hillsborough River water. (Reported water temperatures are the temperatures at which each solution was equilibrated; atmospheric pressure assumed to be 101.325 kPa) Next, equations fitted to the solubility data from several sources (Garcia & Gordon, 1992, Hamme, 2004, Weiss, 1974, Wiesenburg & Guinasso, 1979) were used to calculat e the Bunsen solubility coefficient for each gas at its respective equilibration temperat ure (for the purpose of these calculations salinity was assumed to be zero and pressure was assumed to be 101 .325 kPa). Dissolved gas concentrations could then be calculated with C i = s i p i (1) Where s i is the Bunsen solubility coefficient, p i is the partial pressure of the gas being measured and C i is the dissolved gas concentration in mol/kg. This was done for each 0.00E+00 1.00E 09 2.00E 09 3.00E 09 4.00E 09 5.00E 09 6.00E 09 7.00E 09 1 185 369 553 737 921 1105 1289 1473 1657 1841 2025 2209 2393 2577 2761 2945 3129 3313 3497 3681 3865 4049 4233 4417 4601 4785 4969 5153 5337 Intensity (Amps) Scan # Ar Calibration for the Hillsborough River 0.5000% Ar Water Temp: 24.8 0 C 2.040% Ar Water Temp: 24.5 0 C 5.070% Ar Water Temp: 24.5 0 C
39 of the gases in the three certified gas standards. In this way dissolved gas concentr ations, in units of moles/kilogram, were correlated to mass spectrometer ion intensity values. These data were the n plotted (Figure 23) to produce 3 point calibration curves for each of the five gases to be measured during the Hillsborough River deploym ent. These calibration curves were then used to convert the mass spectrometer ion intensity data collected during the deployment to dissolved gas concentrations. Interferences arising from molecular fragmentation occurring in the mass spectromete r ion source result s in three of the calibra tion plots in Figure 23 (oxygen at m/z 32 and 34 and nitrogen at m/z 28 ) having a positive y axis intercept. During electron impact ionization a percentage of H 2 O and CO 2 molecules entering the ion source through the membrane inlet are fragmented. Two ions formed in this process, O 2 + and CO + add to the ion intensity signal measured for m/z 32, 34 and m/z 28 respectively. This increase in the ion intensity signal for all oxygen and nitrogen calibration standards resulted in a positive offset of the calibration curve. It was assumed that the rate of fragmentation and subsequent increase in the m/z 28, 32 and 34 signals remained constant and was not accounted for in the final calibra tion.
41 Hillsborough River Deployment The Hillsborough River, flowing 54 miles from its source in the Green Swamp to Hillsborough Bay, is an important recreational and drinking water resource (Pillsbury, 2004) The lower 12 miles of the river passe s through a heavily urbanized and in dustrialized area including the City of Tampa (Figure 22) This portion of the river is also subject to restricted flow due to construction of a dam that create d the 6.1 billio n liter Hillsborou gh Reservoir. Additionally, the lower Hillsborough is impacted by a high volume of groundwater flow in the form of freshwater springs and seeps. T he MS system was deployed on the lower Hillsborough River aboard a guided surface vehicle (see Bayboro Harbor and Lake Maggiore section of this chapter for the GSV description) in February of 2007. The GSV navigated a 2,640 meter stretch of the river (Figure 22 ) while the MS system sampled the surface water of the river and collected mass spectrometer data for the dissolved gases listed in Table 5. The GSV was also outfitted with a GPS, a conductivity, temperature and d epth (CTD) probe and a dissolved oxygen (DO) probe. T he deployment produced four data sets ( gas concentrations, temperature, DO, latitude and longitude ) that were subsequently combined by use of a Matlab script, to produce geo referenced maps of the d issolved gas concentrations along the GSV track. Intermittent loss of GPS signal during the deployment resulted in gaps in the concentration maps shown in Figures 23 25 The mass spectrometer sampled continuously during the deployment, at a rate of 0.7 seconds/sample, resulting in the collection of mass spectrometer data over the entire 2,640 meter section of the river.
42 Dissolved oxygen (DO) concentr ations are reported in Figure 23 A steady increase in DO concentration was observed over the first half of the deployment track. A sharper inc rease in DO concentration occurred in the vicinity of t he U.S. 275 overpass and remained elevated in the area of the river adjacent to an activ e spring vent located on the property of Sulfur Springs. Elevated DO concentrations in this portion of the river were not anticipated as spring water, typically, has low DO concentrations. A large aeration basin on th e Sulfur Springs site (Figure 23 In set), used to increase spring water DO Figure 22: Aerial photograph of the northern Tampa Bay area with the portion of the Hillsborough River upon which the MS system was deployed on February 13, 2007
43 Figure 23. Dissolved oxygen concentrations in a section of the lower Hillsborough River as recorded by the MS system during a 2,640 meter transect in February 2007. Inset: Sulfur Springs aeration pond, like ly source of elevated oxygen concentrations observed in this segment of the river. concentrations before it enters the Hillsborough River, was the source of the elevated DO concentrations. Just east of the Sulfur Springs influence the DO signal begins to decrease until the last ~300 meters of the track where the DO signal begins a stepwise increase.
44 Comparison of this portion of the DO data with the corresponding portion of CO 2 data (Figure 26) shows that the two signals co vary over this section of the river. As with the Lake Maggiore data, this O 2 CO 2 signal covariation is indicative of biochemical transformations associated with photosynthesis/respiration. Figure 24. Dissolved carbon dioxide concentrations in a section of the lower Hillsborough River as recorded by the MS system during a 2,640 meter transect in February 2007.
45 Dissolved carbon dioxide concentrations (Figure 24) remained relatively constant over the first ~1,760 meters of the track. East of Sulfur Springs the CO 2 signal begins to increase until the section of the river where the O 2 CO 2 signals were observed to co vary. The United States Geological Survey conducted an investigation of groundwat er inputs to the lower Hillsborough by measuring radon (Rn) concentrations (59) a natural groundwater tracer. The investigation showed steadily increasing Rn concentrations, indicative of increased groundwater input, along the section of the river where e levated CO 2 concentrations were observed. Groundwater can be supersaturated relative to CO 2 and often leads to supersaturation of CO 2 in streams and rivers (Jones & Mulholland, 1998). It is possible that increased groundwater flow could be contributing t o elevated CO 2 concentrations in certain sections of the river. Dissolved methane concentrations (Figure 25) remained relatively constant over the first ~1,200 meters of the track. In the area of Sulfur Springs, the CH 4 concentration fluctuates br iefly before beginning a steady increase over the final 1,000 meters of the track. Increasing methane concentrations along this section of the river could be groun dwater inputs. In order to validate MS system performance during this deployment mass spectrometer DO data were compared to the YSI DO sensor data. The data plotted in Figure 26 show excellent agreement but the mass spectrometer DO concentrations were consistentl y lower than those recorded by the YSI DO sensor. The discrepancy in the two data sets was attributed to fine scale changes (alteration of membrane geometry, change in sample flow rate, change in sample temperature, etc
46 Figure 25. Dissolved methane concentrations in a section of the lower Hillsborough River as recorded by the MSS system during a 2,640 meter transect in February 2007. in the MS system that could have affected mass spectrometer analyte measurements.
47 In order to compensate for these effects an argon correction factor was derived. [Ar] s,c [O 2 ] m = [O 2 ] corr (2) [Ar] m Where [Ar] s,c is the calculated argon saturation concentration along the deployment track, [Ar] m is the argon concentration measured by the mass spectrometer, [O 2 ] m is the oxygen concentration measured by the mass spectrometer and [O 2 ] corr is the corrected oxygen concentration. A Matlab script was written for calculation of the argon corrected oxygen data. Mass spectrometer argon corrected oxygen data plotted against the YSI DO sensor oxygen data were in excellent agreement All data plotted in Figures 23, 24 and 25 were argon corrected in this way Oxygen Time Series YSI Oxygen Sensor Mass Spectrometer Argon Corrected Time (UTC) 300 25 0 2 00 15 0 1 00 5 0 0 2:24 PM 3:21 PM 4:19 PM 5:16 PM 6:14 PM Figure 26. Oxygen data as recorded by the MS system and the YSI DO sensor during the Hillsborough River deployment of Febr uary 2007. The green trace represents mass spectrometer data that have been argon corrected.
48 Conclusions The work presented in this thesis provides evidence that a mass spectrometer can be deployed in aquatic environments and collect unique and meaningful data. Much work remains to be done to realize the full potential of this new technology. Improvement in the accuracy and precision of in situ mass spectrometer data is crucial if the intrument is to provide data that can be employed in interpreting the fine scale, biogeochemically important fluctuations in the dissolved gas concentrations of natural systems. Measurement of dissolved gases has long been an integral part of understanding the biogeochemical cycling of mar ine and freshwater systems. These measurements are of ss. Traditional methods for the collection of dissolved gas samples are labor and time intensive as well as being prone to sample loss. Typically single grab samples are collected and analyzed leading to poor spatial and temporal data densities, leaving conditions in most of the water column, which is not sampled, to be interpolated. The USF UMS system, utilizing a m embrane inlet is well su ited to address many of the shortcomings of traditional di ssolved gas sampling methods. Continuous sampling, with high through put of sample, and rapid scan rates provide real time, temporally dense data sets. Deployment of the USF UMS system on a variet y of platforms such as a GSV has resulted in the collection of dissolved gas data over large areas and at great depths. The UMS system has the unique capability of simultaneously monitoring
49 multiple analytes, eliminating the need to dep loy several analyte specific instruments. The power of this method was evident in the measuring of covarying oxygen and carbon dioxide signals diagnostic of biochemical processes, in both Lake Maggiore and the Hillsborough River with concomitant collect ion of GPS data. These data were than plotted, with high precision, on georeferenced maps providing the exact geographic location where the biochemical processes occurred. Absent either the oxygen or carbon dioxide signal the significance of these data m ight not have been recognized. It is anticipated that UMS systems will become an integral component of ocean observing networks. Network infrastructure in most cases, will provide power and data transmission such that UMS will collect data over ex tended periods of time. UMS, outfitted with membrane inlets, have been de ployed to depth at MC118 located in the Gulf of Mexico at depths of ~900 meters, off the coast of Mississippi, to detect methane seeps associated with methane hydrates. Advances i n micro fabrication techniques have led to the miniaturization of mass spectrometer sensors which require less power. Manufacture of smaller vacuum components (diaphragm backing pumps, turbo/molecular pumps) will further reduce the footprint and power re quirements of entire mass spectrometer systems This miniaturization of mass spectrometer systems makes possible the manufacture of other mass spectrometer geometries that can be packaged as in situ instruments. Efforts are underway at the University of South Florida to develop an in situ magnetic sector mass spectrometer with a Mattauch Herzog geometry that will have the capability of measuring carbon, nitrogen and oxygen isotopes in aquatic environments.
50 On site, real time analysis of large organic molecules by in situ MALDI TOF mass spectrometers could lead to better understanding of the biochemistry of coastal and estuarine waters. Similarly, in situ ICP MS could provide further insight into the cycling and complexation of trace metals in aquatic environments. In situ hyphenated methods such as GC MS and HPLC MS also hold great potential to produce data that will further our understanding of aquatic environments. In time, in situ mass spectrometry will be as robust and versatile an analytical method as its lab borne predecessor.
51 References Amad, M. H., N.B. Cech, G.S. Jackson, C.G. Enke, Importance of Gas Phase Proton Affinities in Determining the ESI Response for Analytes and Solvents. Journal of Mass Spectrometry 2000, 35, 784 789. Andreae, T. W., M.O. Andreae, G. Schebeske,, Biogenic sulfur emissions and aerosols over the tropical South Atlantic 1. Dimethylsulfide in sea water and in the atmospheric boundary layer. Jour nal of Geophysical Research 1994, 99, (D11), 22819 22829. Ardrey, R. E., Liquid Chromatography Mass Spectrometry An Introduction John Wiley & Sons: Chichester, 2003. Aston, F. W., A Positive Ray Spectrograph. Philosophical Magazine V. 1919, 38, 709. Ba inbridge, K. T., Comparison of the Mass of He and H 1 on a Mass Spectrograph. Physical Review 1933, 43, 103. Binqiu, Z., Sea Flux of Trace Metals in Saanich Inlet. Chinese Journal of Oceanology and Limnology 1989, 7, (4), 345 354. CHM Hill, I nc Lake Magg iore Environmental Assessment Phase II ; St. Petersburg, 1991. Comisarow, M. B., A.G. Marshall, Fourier Transform Ion Cyclotron Resonance Spectroscopy. Chemical & Physical Letters 1974, 25, 282 283. deLaeter, J. R., Applications of Inorganic Mass Spectrometry John Wiley & Sons, inc.: New York, 2001; p 474. Dempster, A. J., A New Method of Positive Ray Analysis. Physical Review 1918, 11, 316. Diaz, J. A. D. Sub Miniature Double Focusing Mass Spectrometer for In situ Environmental Monitoring and Application to Volcanic Gaseous Emissions. University of Minnesota, 1999. Dole, M., L.L Mack, P. Kralik, A. Rheude, Molecular Beams of Macroions, II. Journal of Chemical Physics 1970, 52, (10), 4977. Dulski, T. R., Trace elemental analysis of metals: Met hods and Techniques Marcel Decker, Inc.: 1999; p 581.
52 Garcia, H. E., L.I. Gordon, Oxygen in seawater: Better fitting equations. Limnology and Oceanography 1992, 37, (6), 1307 1312. German, C. R., Henry Elderfield, Rare earth elements in Saanich Inlet, B ritish Columbie, a seasonally anoxic basin. Geochimica et Cosmochimica Acta 1989, 53, 2561 2571. Gohlke, R. S., F.W. McLafferty, Early Gas Chromatography/Mass Spectrometry. Journal of the American Chemical Society Mass Spectrometry 1993, 4, 367 371. Goud smit, S. A., A Time of Flight Mass Spectrometer. Physical Review 1948, 74, 622. Griffiths, J., A Brief History of Mass Spectrometry. Analytical Chemistry 2008, 80, 5678 5683. Gross, J. H., Mass Spectrometry, A Textbook 1st ed.; Springer Verlag: Heidelbe rg, 2004; p 518. Hamme, R. C., The solubility of neon, nitrogen and argon in distilled water and seawater. Deep Sea Research 2004, 51, (11), 1517 1528. Hays, E. E., P.I. Richards, and S.A. Goudsmit, Mass Measurements with a Magnetic Time of Flight Mass S pectrometer. Physical Review 1951, 84, 824. Hemond, H., R. Camilli, NEREUS: engineering concept for an underwater mass spectrometer. Trends in Analytical Chemistry 2002, 21, 526. Hoch, G., B. Kok, A mass spectrometer inlet system for sampling gases disso lved in liquid phases. Archives of Biochemistry and Biophysis 1963, 101, (1), 160 170. Jumars, P. A., Concepts in Biological Oceanography, An Interdisciplinary Primer Oxford University Press: New York, 1993. Kana, T. M. C. D., M. Duane Hunt, James B. Oldham, George E. Bennett, and Jeffrey C. Cornwell, Membrane Inlet Mass Spectrometer for Rapid High Precision Determination of N 2 O 2 and Ar in Environmental Water Samples. Analytical Chemistry 1994, 66, 4166 4170. Karas, M., U. Bahr, D. Bachmann, F. Hillenkamp, Matrix Assisted Ultraviolet Laser Desorption of Non Volatile Compounds. International Journal of Mass Spectrometry. Ion Processes 1987, 78, 53 68. Mattauch, J., R. Herzog, Uber einen neuen Massenspektrographen. Z. Physik 1934, 89, (11 12), 786 Mattauch, J., A Double Focusing Mass Spectrograph and the Masses of N 15 and O 18 Journal of Physical Review 1936, 50, 617, 1089.
53 Nier, A. O., A mass spectrometer for routine isotope abundance measurements. Review of Scientific Instruments 1940, 11, 212 Nier, A. O., E.T. Booth, J.R. Dunning & A.V. Grosse, Nuclear fission of separated uranium isotopes. Physics Review 1940, 57, (6), 546. Nier, A. O., A Mass Spectrometer for Routine Isotope Abundance Measurements. Review of Scientific Instruments 1940, 11 212 216. Paul, W., H. Steinwedel, A New Mass Spectrometer without Magnetic Field. Z. Naturforsch 1953, 8A, 448 450. Peters, K. E., Clifford C. Walters and J. Michael Moldowan, The Biomarker Guide Biomarkers and Isotopes in the Environment and Human His tory Cambridge University Press: Cambridge, 2005; Vol. 1. Pillsbury, L. A. Spatial and Temporal Chemical Variations in the Hillsborough River System. University of South Florida, St. Petersburg, FL, 2004. Price, F. M. M. M. a. N. M., The Biogeochemical Cycles of Trace Metals in the Oceans. Science 2003, 300, 944. Roboz, J., Introduction to Mass Spectrometry: Instrumentation and Techniques Interscience Publishers: 1968; Vol. 3, p 539. Short, R. T., D.P. Fries, S.K. Toler, C.E. Lembke, R.H. Byrne, Deve lopment of an underwater mass spectrometry system for in situ chemical analysis. Measurement Science Technology 1999, 10, (12), 1195 1201. Short, R. T., D.P. Fries, M.L. Kerr, C.E. Lembke, S.K. Toler, P.G. Wenner, and R.H. Byrne, Underwater Mass Spectrome ters for in situ Chemical Analysis of the Hydrosphere. Journal of the American Society of Mass Spectrometry 2001, 12, 676 682. Short, R. T., D.P. Fries, G.P.G. Kibelka, M.L. Kerr, S.K. Toler, P.G. Wenner, R.H. Byrne In Field Chemical Analysis using real time in water mass spectrometry Oceans 2001, MTS/Ieee Proceedings, Honolulu, HI, 2001; Honolulu, HI, 2001. Short, R. T., R.H. Byrne, David Hollander, Johan Schijf, Strawn K. Toler, and Edward S. vanVleet, Mass Spectrometry Instrumentation, Interpret ation and Applications J. Wiley & Sons: Hoboken, 2009; p 371. Smith, L. P., W.E. Parkins, A.T. Forrester, On the Separation of Isotopes in Quantity by Electromagnetic Means. Physics Review 1947, 72, 989 1002. Smith, R. D., W.D. Felix, J.C. Fjeldsted, M. L. Lee, Capillary Column Supercritical Fluid Chromatography Mass Spectrometry. Analytical Chemistry 1982, 54, (1883 1885).
54 Smith, R. M., Understanding Mass Spectra: A Basic Approach John Wiley and Sons, Inc.: 1999; p 290. Spengler, B., D. Kirsch, R. Kauf mann, R.J. Cotter, Metastable decay of peptides and proteins in matrix assisted laser desorption mass spectrometry. Rapid Communications in Mass Spectrometry 1991, 5, 198. Stephens, W. E., A Pulsed Mass Spectrometer with Time Dispersion. Physical Review 1946, 69, 691. Tanaka, Y., Y. Kishimotos, K. Otsuga, S. Terabe, Strategy fo r Selecting Separation Solutions in CE MS. Journal of Chromatography A 1998, 817, 49 57. Tanoue, J. K. V. a. E., Chemical and Biological Studies of Particulate Organic Matter in t he Ocean. Journal of Oceanography 2002, 58, (2), 265 279. Technology, N. I. o. S. a., NIST/EPA/NIH mass spectral database, standard reference database 1A and the NIST mass spectral search program, version 1.6d.cd 1998. Thomson, J. J., Further Experiments on Positive Rays. Philosophical Magazine V. 1912, 24, 209. Thomson, J. J. Rays of positive electricity and their application to chemical analysis ; Longmans Green and Co.: 1913. Thomson, J. J., Rays of positive electricity Proceedings of the Royal Socie ty, 1913; 1913; pp 1 20. Urey, H. C., The Thermodynamic Properties of Isotopic Substances. Journal of the Chemical Society 1947 562. Wakeham, S. G., Reconstructing past oceanic temperatures from marine organic biogeochemistry, chemical fossils and molec ular stratigraphy. Environmental Science and Technology 1993, 27, (1), 28 33. Wang, B. H., C.E. Hopkins, A.B. Belenky, A.S. Cohen, Sequencing of modified oligonucleotides using in source fragmentation and delayed pulsed ion extraction marix assisted laser desorption ionization time of flight mass spectrometry. International Journal of Mass Spectrometry and Ion Processes 1997, 169 170, 331 350. Weiss, R. F., Carbon Dioxide in Water and Seawater: The solubility of a non ideal gas. Marine Chem istry 1974, 2, 203 215. Wenner, P. G., R.J. Bell, F.H.W. van Amerom, S.K. Toler, J.E. Edkins, M.L. Hall, K.Koehn, R.T. Short, R.H. Byrne, Environmental Chemical Mapping Using an Underwater Mass Spectrometer. Trends in Analytical Chemistry 2004, 23, (4), 2 88 295.
55 Wien, W., Ueber die Fragen, welche die translatorische Bewegung des Lichtathers betreffen. Annalen Der Physik 1898, 301, (3). Wiesenburg, D. A., N.L. Guinasso, Jr., Equilibrium Solubilities of Methane, Carbon Monoxide, and Hydrogen in Water and Sea water. Journal of Chemical and Engineering Data 1979, 24, (4), 356 360.