|
|
|
|
A
New In Situ Chemical Analyzer for
Mapping Alfred K. Hanson Abstract-
A rapid-response, submersible chemical analyzer has been developed that is
particularly suited for mapping nutrient distributions in 3-D space and time and
for testing hydrodynamic models for chemical transport, within rivers,
estuaries, coastal and off shore waters. The
SubChemPak Analyzer™ is a novel
3-channel reagent delivery module that transforms underwater optical instruments
into sensitive chemical analyzers for rapid measurements of nutrients and
other environmentally important chemicals. A family of multi-nutrient analyzers
is being developed for selected nutrients: nitrate, nitrite, ammonia, urea,
phosphate, silicate, and iron. Continuous flow spectrophotometric and
flourometric methodologies have been optimized for rapid in
situ measurements of the nutrients. The operation, in
situ calibration and data acquisition for the instrument are remotely
controlled by a computer. The concentration readings for nutrients are
instantaneously displayed on the computer monitor. The SubChemPak
Analyzer may be co-deployed with
standard oceanographic electronic sensor packages (CTD) for vertical and/or
horizontal (to-yo) profiling. A
SubChemPak Analyzer was configured for
the simultaneous determination of dissolved nitrite and iron (II) and
co-deployed with a Sea Bird Electronics CTD system in the waters of Narragansett
Bay during December 1999. The CTD system included modular sensors for the
measurement of conductivity, temperature, pressure, dissolved oxygen, pH,
chlorophyll fluorescence, and light transmission and irradiance. The results
from laboratory and field tests in Narragansett Bay demonstrated that the SubChemPak
Analyzer is capable of producing high-resolution vertical nutrient profiles
in real time. The stable optical detection system and in
situ calibration feature enabled accurate nutrient determinations at trace
concentration levels (nanomolar to micromolar). The combination of this
sensitive real-time chemical profiling technique with concurrent acoustic
doppler current velocity and direction measurements will improve our ability to
detect and track chemical plumes in coastal waters. I.
INTRODUCTION
Characterization of the marine distributions of nutrients is critically
important because, depending upon their concentrations, these biologically
essential chemicals may greatly enhance or limit the growth of microscopic
plants in aquatic waters. Natural and man-made environmental events can lead to
dramatic changes in nutrient concentrations in aquatic waters, both in time and
space. Such episodic changes in nutrient concentrations can drastically
influence algal growth rates leading to troublesome eutrophication of lakes and
estuaries, harmful algal blooms, anoxia and fishery problems [1,2]. There has
been steady progress since the '70's in the development of analytical methods
[3] and multi-nutrient autoanalyzer systems for both off-site and on-site
determination of nutrients on collected water samples. However it is now
apparent that there are significant practical limitations to the temporal and
spatial resolution that can be obtained for nutrient measurements in aquatic
waters using traditional water sampling and bench-top autoanalyzer technologies.
One way to overcome these sampling limitations is in
situ chemical analysis [4, 5, 6]. During
1995 a prototype dual-nutrient, submersible chemical analyzer was designed,
developed and successfully tested in coastal waters [6]. The instrumentation was
developed for the NAVY to demonstrate the existence micro- to fine-scale
chemical gradients, associated with thin plankton layers, in stratified coastal
waters. The prototype analyzer simultaneously determined two nutrients,
dissolved nitrite and iron(II), at nanomolar concentration levels, in real-time
and with submeter-scale depth resolution. Subsequent
laboratory investigations demonstrated that additional nutrients of interest
(nitrate, ammonium and phosphate) could also be determined accurately with the
prototype analyzer, after kinetic optimization of the analytical methodologies
[7]. In this paper we describe the features of a
second-generation prototype, the SubChemPak Analyzer™.
We also
present some analytical results that were obtained while deploying the new
instrument package in the estuarine waters of Narragansett Bay, RI. II.
METHODOLOGY A.
Description of the Second Generation Prototype
The SubChemPak Analyzer (Fig. 1) is a 3-channel reagent delivery module that transforms underwater optical instruments into sensitive chemical analyzers for rapid measurements of nutrients and other environmentally important chemicals. Continuous flow spectrophotometric and flourometric methodologies have been optimized for rapid in situ determinations of several dissolved nutrients.
The
submersible components of the SubChemPak
Analyzer
system are the reagent delivery module and the electro-optical
detectors. The reagent delivery module is comprised of a cylindrical black
acetyl pressure housing (14 cm o.d. x 60 cm length) that contains the
electro-fluidic, data acquisition, instrument control and power regulation
systems. These submersible components are rated for depths up to 200 meters. A
single underwater cable connects the SubChemPak
Analyzer with a deck box that provides DC power to the submerged
instrumentation and a data communications interface (serial cable or
telemetry) to the computer. The SubChemPak can be readily configured for the determination of
different nutrients by changing the reagents, standards and the modular
electro-optical detectors. 1) Reagent
delivery module:
The primary components of the electro-fluidic flow system for the SubChemPak
Analyzer are shown schematically in Fig 2. The components of the include a
micro gear pump, flow rate and temperature sensors, a flexible tube heater,
miniature solenoid pumps for reagents and calibration standards, reagent bags
and a reaction manifold, reaction coil and debubbler. The three measurement
channels are internally configured to be either analytical (reagents added) or
reference (no reagents added). The reagents and calibration standards are contained in small
plastic bags located outside of the main pressure housing. They are usually
placed in a separate cylindrical housing that is exposed to ambient underwater
pressures. These bags are supplied pre-sterilized to help eliminate biofouling
of reagents and nutrient standards. Two to four reagent bags are typically
required for deployment; one to three 500 ml bags for the analytical reagents
and one 250 ml bag for the calibration standard. The
seawater and reagent flowrates are 20 ml/min/channel and 1 ml/min/channel,
respectively. At these reagent volumes and flowrates the instrument can be
operated continuously for ~8
hours without replenishing the reagents. A five point calibration by the
method of standard additions can be remotely initiated, while submerged, by
incrementally changing the flowrate of the standard solution from 0 to 5
ml/min. 2)
Electro-optical detectors:
Two modular optical detection systems, spectrophotometric and
fluorometric, have been developed by WET Labs, Inc. and tested with the SubChemPak Analyzer (Fig. 1b).
One optical detector, the A-Star, is a submersible
absorption meter with an optical flow cell (3 mm I.D. and 25 cm length). 3) Remote Data acquisition and instrument control: The data acquisition and control (DAC) technology for the SubChemPak Analyzer involves a remote windows PC, and the DAC software and hardware. National Instruments software, Lab VIEW™, and compatible data acquisition hardware have been adapted for this application. A stand-alone executable has been developed with a graphical user interface that provides user-friendly remote-control of all analyzer functions. The DAC hardware includes state-of-the-art chips for analog-digital signal conversion and two-way serial communication via a multi-drop RS 485 network. The multi-channel data acquisition rate is one reading per second.
4)
Remote power supply: The power
requirements for the SubChemPak Analyzer are remotely supplied by an AC to
DC switching power supply that is located in the deck box. The voltage and power
outputs of the main power supply and the length and gauge of the conducting
wires are designed to account for any voltage drop due to resistance losses over
the long underwater cables. Miniature DC-DC converters convert the unregulated
underwater DC supply into stabilized power (12 and 24 vdc) within the SubChemPak.
The maximum total power requirements of the SubChemPak,
are approximately 150 watts. The flexible tube heater is the major power
consuming component. This device is used to heat the flowing seawater (typically
to 30 C), in order to increase the rates of the color or fluorescence
development reactions. In warmer waters, less power is used for this task than
in cooler waters. The flexible tube heater is a requirement for profiled or
towed applications that require rapid response times (seconds).
A.
Submerged Tests in Tanks and Narragansett Bay
After bench top testing, evaluation and refinement, a SubChemPak
Analyzer was installed on a
winch-deployable, electronic profiling package. The profiling package also
included a Sea Bird Electronics Model 25 Sea Logger CTD with modular sensors for
the measurement of conductivity, temperature, pressure, dissolved oxygen, pH,
chlorophyll fluorescence, and light transmission and irradiance. A 30-meter long
umbilical was assembled that included two sea-cables and a plastic tube for
water sample collection by a pump-to-surface technique. Two notebook computers
were used to control the operation of the submerged instruments and acquire and
display the real-time data stream from the CTD sensors and the SubChemPak
Analyzer. The
integrated SubChemPak Analyzer and CTD
system were then
tested in a series of submerged experiments in out door test tanks filled
with Narragansett Bay seawater (5 meters deep). On December 20, 1999 the CTD and
chemical profiling system was deployed in Narragansett Bay (Fig. 3) aboard a URI
Ocean Engineering research vessel, the CT-1. For these tank and field tests the SubChemPak Analyzer was configured for the simultaneous
determination of the nutrients, dissolved nitrite and iron(II).
Three
A-Star detectors (520 nm) were used for the
chemical measurements. One A-Star detector was used to determine Nitrite
by a spectrophotometric method that is based upon the formation of a colored azo
dye. Nitrite reacts with sulfanilamide to form a diazonium ion that is
subsequently coupled with N-(1-napthyl)-ethylenediamine dihydrochloride (molar
absorbtivity = ~46,000 at 532 nm) [3]. A second A-Star detector (520 nm) was
used to determine iron(II) using the classical Ferrozine methodology (molar
absorbtivity = ~26,500 at 560 nm, somewhat less at 520 nm) [6,9]. The third
A-Star detector was used as a reference detector (no reagents added or color
formation) to correct for any background absorption due to the presence of
colored dissolved organic matter (CDOM) in the seawater.
During
the December cruise in Narragansett Bay, four vertical casts were completed
using the SubChemPak Analyzer and CTD
system near three station locations (Stations 6, 7, and 8 as shown on Fig. 3)
while drifting with surface currents. The package descent rates were 2-3 m/min.
Vertical profiles of current velocity and direction (RDI ADCP) were also
recorded during the casts. III.
RESULTS A.
Instrument Internal Calibration – In situ
One of the early goals of the submerged tank and field tests was to evaluate the accuracy and precision of the internal calibration while the instrument is submerged. The results of several calibration experiments are compiled and shown in Fig. 4. The calibration data (Fig. 4) includes results from a six point in situ calibration run, conducted at 9.5 m depth at station 7. The calibration coefficients obtained for nitrite and iron(II) at 9.5 m depth, in the cold (~6 C) bottom waters of the Bay, were statistically indistinguishable from those obtained in the lab and other submerged environments. The detection limits for nitrite and iron(II) were estimated to be 1.0 nM and 2.0 nM, respectively.
B.
Narragansett Bay Profiles
The vertical profiles obtained at
Stations 6, 7 and 8 are shown in Figs. 5-8 for some of the parameters that were
recorded during the SubChemPak Analyzer
deployments. All parameters were measured simultaneously, and visualized in
real-time during the deployment. All data was binned at 1 sample per second.
The vertical cast at station 6 in the Providence River (Fig. 5) was taken
at approximately 12:17 PM local time, during slack low tide. Water velocities at
that station were appropriately low, below 10 cm/s, and spanned a range of
directions. The CTD data indicated
a stratified water column with cooler, lower salinity water overlying denser
waters that were warmer and saltier. An intense plume of phytoplankton (see
chlorophyll data) was embedded in the water column from 1 to 4 meters depth. These
eutrophic waters also exhibited higher oxygen levels and a distinct pH gradient
to higher levels with depth. Dissolved nitrite generally decreased with depth
with a sharper gradient in deeper waters. Surprisingly high and variable levels
of dissolved iron(II) (100-500 nM) were present in the water column at this
station. We hypothesize that this anomalous plume of iron(II), is a
thermodynamically unstable or transient signal, emanated from a sewage treatment
plant outflow located near the sampling site. This chemical plume containing
iron(II) was not apparent at Stations 7 or 8 (see below). Similar plumes of
iron(II) were also observed during our earlier work in East Sound, WA [6]. The vertical cast
at station 7, near the mouth of the Providence River, (Fig. 6) was taken at
approximately 1:13 PM local time, at the beginning of the incoming tide. The CTD
data indicated a shallow layer of cool, lower salinity water (0-2 m) overlying
denser waters (2-8 m) that were warmer and saltier. The shallow surface layer
had elevated chlorophyll pigment levels, higher oxygen and lower pH levels, than
the bottom layers. Dissolved nitrite decreased with depth and iron (II)
increased with depth in the water column. Water velocities (not shown) were
quite low in the top four meters, but increased to approximately 18 cm/s in the
lower water column, at an average direction of 315 degrees.
The in situ trace chemical profiles
for nitrite and iron(II) (Fig. 7b) exhibited a sharp discontinuity near 5.5
meters depth that is attributed to current shear associated with incoming tidal
currents in deeper waters (northerly velocity component). IV.
DISCUSSION A. Nutrient
Determinations
in Marine Waters There has been steady progress in the development of water sample
collection and preservation techniques, and bench-top analytical methods and
instrumentation, for the determination of nutrients in aquatic waters. Water
samples are generally collected by tripping water samplers at discrete depths
during hydro or CTD-rosette casts or by sampling seawater pumped from depth.
Collected water samples are often filtered and frozen to preserve them for
future nutrient determinations. Well characterized spectrophotometric and
fluorometric analytical methodologies have been established for the
determination of nutrients in water samples [3]. Optical instrumental techniques
have also been developed for important micronutrients like iron [10,12]. Several
multi-nutrient autoanalyzer systems, that apply segmented (air-bubble)
continuous flow analysis, are commercially available and routinely used for off-site
determinations of nutrients on collected water samples. Autoanalyzer systems
have also been used on-site, either in portable field laboratories or
on-board coastal and off-shore research vessels, in an effort to obtain faster
and more reliable nutrient data. On-site nutrient measurements are often
conducted continuously, or semi-continuously, by connecting the autoanalyzer to
a flowing sample stream. Such "on-line" systems allow for continuous
vertical profiling [13, 14] or horizontal surface mapping [15, 16] of chemical
distributions. It is well known that on-site field measurements can offer cost
and time advantages over sample analyses by off-site laboratories. In spite of this notable progress, there are significant practical
limitations to the temporal and spatial resolution that can be obtained for
nutrient measurements in aquatic waters using these established sampling and
autoanalyzer technologies. Although off-site measurements by university, state,
federal and private laboratories persist as the primary method for many
environmental chemical analyses, now there is a need for enhanced on-site
measurement capabilities and, particularly for chemical sensors and analyzers
that operate in situ [4, 5, 6,17]. B. In Situ Nutrient Analyzers for Environmental Monitoring
In
situ sensors for physical (salinity,
temperature), bio-optical (radiometers, chlorophyll fluorometers, light
transmissometers) and some chemical parameters (oxygen, pH) have been available
for many years to the marine research and environmental monitoring community.
These sensors have become so reliable that it is now a rare occurrence to have
water samples collected for salinity or oxygen measurements. Fortunately, our
evolving environmental monitoring and oceanographic research needs for in
situ chemical measurements has coincided with considerable technological
advancement in our ability to construct submersible chemical analyzers and
sensors that may operate in remote environments. Advances in flow-injection and
continuous flow analysis techniques, osmotic and electro-osmotic pumps,
fiber-optics technology, electrochemical sensors and biosensors offer exciting
opportunities for the development of submersible instrumentation to monitor most
chemical constituents of interest. There has been considerable progress during
the past ten years in the development of submersible chemical analyzers for
either stationary monitoring or profiling for selected chemical
constituents [4, 5, 6, 16, 18,19, 20, 21]. Two profiling chemical analyzers for nutrients have been
developed and described in the scientific literature. Johnson et al. [18]
were the first to successfully develop a submersible chemical analyzer (the dual
channel "scanner") by applying continuous flow analysis, without
air-bubble segmentation. The "scanner" system was successfully applied
to obtain continuous profiles for nitrate, sulfide, iron, manganese and hydrogen
peroxide in marine waters [16,18,19]. A similar approach has also used [21] to
construct a 3 channel instrument, the "in
situ-CFA", with a demonstrated applicability for nitrate profiling.
The commercial development
of submersible chemical sensors and analyzers has been economically hindered
within the United States [17] limiting their availability for use by our
environmental and oceanographic research communities. Two in situ nutrient analyzers for nitrate and phosphate are
commercially available in Europe for ‘longer-term’ stationary monitoring of
nutrient levels in marine waters: 1) W.S Ocean Systems “Nutrient Monitor” and 2) Chelsea Instruments “AquaSensor”. Since
several minutes (or more) are required per analysis, these instruments are
presently not designed or suited for determining
nutrient concentrations at the high data rates required for vertical
profiling and 3-D mapping of nutrient distributions. A.
The SubChemPak Analyzer in Narragansett Bay
The field deployments in the Providence River and Narragansett Bay
demonstrated the unique capability of the SubChemPak
Analyzer to profile trace chemical concentrations, in real time, and
continuously (1 reading per second) while submerged. The vertical depth
resolution was estimated to be 20 cm at a 3 m/min descent rate. The instrument
operated in a consistent manner at room temperatures (20 C), submerged in the
cold waters (6-9 C) of Narragansett Bay, and when subjected to natural gradients
of salinity, temperature, biology and chemicals. The experimental results
demonstrate that the in situ
calibration feature works very well, and that the SubChemPak Analyzer is capable of producing accurate and
reproducible analytical results while submerged. The high-resolution vertical profiles obtained for nitrite and
iron(II) in the stratified waters of Narragansett Bay exhibited micro- to
fine-scale chemical gradients. These observations provide further supporting
evidence for the fine-scale chemical gradients and thin layers that were
detected in East Sound, WA during
an earlier investigation [6]. Continuous vertical profiling with an in
situ chemical analyzer is the only way to document such dynamic chemical
variability. These chemical gradients can not be detected by the placement of in
situ chemical analyzers (stationary monitors) at selected fixed depths
within the estuary. V. CONCLUSIONS The results from laboratory and field tests in Narragansett Bay
demonstrate that the SubChemPak Analyzer
is capable of producing high-resolution vertical nutrient profiles in real time.
The stable optical detection system, fast data acquisition rates and in
situ calibration feature enabled accurate nutrient determinations at trace
concentration levels (nanomolar to micromolar). The instrument can be readily
integrated with a winch-deployable oceanographic profiling package that includes
a CTD system components and auxiliary sensors. The combination of this
sensitive, rapid, real-time chemical profiling technique with concurrent
acoustic doppler current velocity and direction measurements will improve our
ability to detect and track chemical plumes in coastal waters. ACKNOWLEDGMENTS Casey Moore, Eugene Morin and Percy Donaghay provided invaluable
assistance with engineering and instrumentation. The assistance of John King,,
Chris Kincaid, Beth Lacey-Laliberte, William Deleo and Captain Fred Pease with
the field deployments is also greatly appreciated. This
work was supported by funds provided by the RI-Economic Policy Commission and
the URI-Ocean Technology Center, the NOAA/UNH
Cooperative Institute for Coastal and Estuarine Environmental Technology,
the Office of Naval Research and the National Ocean Partnership Program. REFERENCES [1]
ECOHAB, The
Ecology and Oceanography of Harmful Algal Blooms: A National Research Agenda,
Donald M. Anderson, Ed., Woods Hole Oceanographic Institution Publication, 1995,
66 p. [2]
D.
Justic, N.N. Rabalais, R.E. Turner, and W.J. Wiseman, Jr., “Seasonal coupling
between riverborne nutrients, net productivity and hypoxia”, Marine
Pollut Bull 26:184-189, 1993. [3]
K.
Grasshoff and F. Koroleff, “Determination of nutrients”, in Methods
of Seawater Analysis, K. Grasshoff, M. Ehrhardt and K. Kremling eds., Verlag
Chemie, 1983, 125-187. [4]
K.S.
Johnson, K. Coale and H.W. Jannasch, “Analytical chemistry in oceanography”,
Analytical Chemistry, 64:1065A-1075A,
1992. [5]
K. S.
Johnson. and H.W. Jannasch, “Analytical chemistry under the sea surface:
monitoring ocean chemistry”, In Situ..
Naval Research Reviews. 3: 4-12, 1994. [6]
A.K. Hanson,
Jr. and P.L. Donaghay. “Micro to fine-scale chemical gradients and layers in
stratified coastal waters”, Oceanography,
Vol. 11, No. 1, 10-17, 1998. [7]
A.K.
Hanson, Jr., “A dual-channel profiling chemical analyzer for nutrients and
iron(II) in marine waters”, AGU Ocean Sciences, San Diego, CA. EOS,
Transactions, American Geophysical Union, 79(1): 2, 1998. [8]
N.M.
Price and P.J. Harrison, 1987. “Comparison of methods for the analysis of
dissolved urea in seawater”, Marine
Biology 94:307-317, 1987. [9]
L.L.
Stookey,
‘Ferrozine - A new spectrophotometric reagent for iron”, Analytical
Chemistry 42(7) 779-781, 1970.
[10]
D.W.
O'Sullivan, A.K.
Hanson, Jr. and D.R. Kester, “The distribution and redox chemistry of iron in
the Pettaquamscutt Estuary”. Estuarine ,
Coastal and Shelf Science, 45:769-788, 1998.
[11]
R.
Kerouel, R. and A. Aminot, “Fluorometric determination of ammonia in sea and
estuarine waters by direct-segmented flow analysis”, Marine Chemistry, 57:265-275, 1997.
[12]
D.W.
O’Sullivan, A.K. Hanson, Jr. and D.R. Kester, “Stopped flow luminol
chemiluminescence determination of Fe(II) and reducible iron in seawater at
subnanomolar levels” Marine Chemistry, 49:65-77, 1995.
[13]
J.J.
Anderson and A. Okubo, “Resolution of chemical properties with a vertical
profiling pump”, Deep-Sea Research,
29(8A): 1013-1019, 1982.
[14]
L.A.
Codispotti, G.E. Friederich, J.W. Murray and C.M. Sakamoto, “Chemical
variability in the Black Sea: implications of continuous vertical profiles that
penetrated the oxic/anoxic interface”, Deep-Sea
Research, 38:2, S691-S710, 1991.
[15]
F.P. Chavez,
H.W. Jannasch, K.S. Johnson, C.M. Sakamoto, G.E. Friederich, G.D. Thurmond, R.A.
Herlien and L.A. Codispotti, “The MBARI Program for obtaining real-time
measurements in Monterey Bay”. IEEE,
327-333, 1992.
[16]
K.H. Coale,
C.S. Chin, G.J. Massouth, K.S. Johnson and E.T. Baker. “In situ chemical mapping of dissolved iron and manganese in
hydrothermal plumes”, Nature
352:325-328, 1991.
[17]
MarChem,
MarChem 93: “The Proceedings of a Workshop on Marine Chemistry
Instrumentation”, S. J. Martin, Ed., Martin Laboratories, 1993.
[18]
K.S.
Johnson, C.L. Beehler, and C.M. Sakamoto-Arnold. 1986a. “A submersible flow
analysis system”, Analytica Chimica Acta.
179:245-257, 1986. |
|
[Home] [Products] [Services] [Company Info] [News] Information: Hanson@subchem.com Copyright © 1998
SubChem Systems, Inc.
|
