Amperometric and Voltammetric Techniques for Fine Scale
Measurements of Sulfide in Sediments
Susie P. Escorcia, Texas A&M University, College Station, TX
Mentors: Carole Sakamoto, Kurt Buck
Keywords: amperometric, voltammetric, sulfide, microelectrodes
Microsensors used to determine sulfide concentrations in sediments can result in fine
scale data resolution not obtainable by most common methods currently used.
Amperometric and voltammetric sensors are currently being considered at MBARI as
methods to acquire data in high sulfide environments, such as the cold seeps in Monterey
Bay. These instruments require optimization to determine their capabilities and
usefulness to the environment intended. The project consists of investigating the use and
limitations of these different electrodes as well as learning their general operating
parameters. This includes lab experiments to determine the upper limits of sulfide
detection. Other lab experiments include a comparison of sulfide measurements from
cores collected from nearby Kirby Park and Elkhorn Slough using an amperometric
sensor, a voltammetric electrode, a methylene blue colorimetric technique, and an UV
absorption technique. In addition, these sensors were used to determine the fine scale
sulfide concentrations in cores collected from the cold seeps in Monterey Bay as well as
the Eel River.
Microelectrodes have been used to study specific chemical species, starting with the
oxygen electrodes (Revsbech et al. 1980, 1983). Traditional methods of porewater
collection, such as sediment squeezing or centrifugation, are incapable of assessing the
fine scale vertical resolution that microelectrodes provide, especially near the sediment-
water interface. These membrane electrodes, which depend on the flux of gas through a
membrane, have been used over the years to determine the concentration of various
single analytes, such as O2, H2S, pCO2, N2O and pH, while examining environmental
biogeochemical processes. On the other hand, solid-state microelectrodes allow for the
reduction of a species on the surface of the electrode, which allows for detection and
measurement of non-gaseous species. Solid state microelectrodes (Pt and C) were
originally developed in the field of physiology to make qualitative and quantitative
measurements in extremely small volumes (Chen et al. 1992, Lau et al. 1992).
Microelectrodes specifically used for this project were amperometric sulfide electrodes
and voltammetric solid-state mercury-gold amalgam microelectrodes.
MATERIALS AND METHODS
Amperometric electrodes made by Unisense were used to determine sulfide
concentrations. These electrodes varied in tip diameter from 50 to 500 um. Each
electrode was calibrated by standard addition of approximately 100 mM Na2S into 20 mls
of degassed buffer solutions of 6.8, 7.0 or 7.2 pH, up to a concentration of approximately
8 to 10 mM total sulfide. pH readings were taken of solutions after the sulfide standard
addition for each concentration to compensate for changes in pH. These electrodes sense
H2S by measuring the picoamp levels of current generated from H2S diffusing into the tip
through a silicone membrane and then undergo a chemical reaction within the
microelectrode (Jeroschewski et al. 1996).
Depending on pH, sulfide can be found in different forms such as HS-, H2S, S2-. Total
sulfide from unknowns can be calculated with pH and H2S concentrations determined
from standard calibrations and using the following formula:
[S2-tot] = (1+ (K1/[H3O+]))*[H2S] pK1 = 6.919
AMPEROMETRIC ELECTRODE EXPERIMENTS
Responsiveness was one of the questions with these new electrodes. We ran a timing
experiment by taking readings during a calibration at 30 second intervals for up to at least
two and a half minutes to determine equilibration time for the electrode.
The life span of an amperometric electrode is approximately six weeks (Jeroschewski et
al. 1996). An ongoing experiment was to run a standard curve every week to see if there
was a degradation of the amperometric signal. If signal degradation was noted, a
conditioning experiment followed to improve the signal and/or extend the life of the
Temperature was also a concern, since cores were to be run in the cold room, so
calibration took place both in the lab and the cold room to note any differences.
The voltammetric solid-state mercury-gold amalgam microelectrodes were developed
principally to determine concentration of oxygen, dissolved (reduced) iron (Fe2+),
manganese (Mn2+) and sulfide (S2-) in sediments (Brendel & Luther III 1995) using
square and cyclic wave voltammetry. These electrodes can also measure I-, Cd+2, and
Pb+2. Voltammetry is the application of a potential ramp with the subsequent
measurement of current, when a chemical species reacts at the electrode (Brendel 1995).
The system used in these experiments was from Bioanalytical Systems, Inc. A glass 10
um electrode was polished, plated, and polarized in order to optimize frequency, step,
amplitude, and filter.
The bacterial mats we are mainly looking at are of Thioploca, a filamentous sulfur
bacteria that can produce dense bacterial mats. These mats contain sheaths that can reach
down five to ten centimeters in depth (Otte et al. 1999, Schulz et al. 1996). These bacteria
oxidize sulfide found in the sediments. For our sampling, the majority of cores were
collected from the center of the bacterial mat, the bacterial mat edge, and a control site, a
meter or more away.
The electrode is positioned and inserted into the sediment core at two-millimeter
increments using a micromanipulator (the increment can even be submillimeter if
desired). After taking readings in a two or three centimeter section of the core, the
electrode is removed, the core is sliced the appropriate amount, and the electrode is
repositioned to continue profiling. A ten to twelve centimeter core takes approximately
an hour to process. pH measurements are also taken.
Two cores from Kirby Park in Elkhorn Slough were collected in order to compare four
methods to determine sulfide. They were an amperometric electrode, a voltammetric
electrode, an UV absorption method, and the methylene blue colorimetric method. Two
of the methods used were fine scale measurements of sulfide (amperometric and
voltammetric) and the other two methods were utilized after the core was sectioned in
two centimeter increments, centrifuged to collect the porewater and then measured for
sulfide after appropriate preparation or addition of reagent. Salinity and pH were also
measured from the centrifuged porewaters.
EXTROVERT CLIFF (JULY 21, 1999)
Three cores from Extrovert Cliff were collected: one from the center of a bacterial mat,
one from along the edge of the bacterial mat, and a control a meter or two away from the
mat. The cores were sampled for sulfide using the amperometric electrode under a
CLAM FIELD SOUTH SITE (JULY 27, 1999)
Three cores from Clam Field South site were collected for sulfide sampling with the
amperometric electrode. A similar core collection pattern was employed: one from the
center of a bacterial mat, one from along the edge of the bacterial mat, and a control a
meter or two away from the mat. The cores were sampled for sulfide using the
amperometric electrode under a nitrogen atmosphere.
EEL RIVER (AUGUST 21 & 23, 1999)
Seven cores were collected from various sites, such as a methane seeps, clam beds,
bacterial mats and a control. These cores were sampled in the field using the
amperometric sulfide electrodes and pH electrodes simultaneously.
Laboratory experiments performed resulted in finding that H2S was linear to up to the
equivalent of at least 8 to 10 mM total sulfide. Early calibrations of total sulfide showed
a plateau after about 6mM total sulfide. But after taking into account pH of the solutions
and calibrating H2S, instead of total sulfide directly, we can measure total sulfide up to 10
Other lab experiments for responsiveness show that each electrode is different and can
change slightly after use, but they are relatively quick to equilibrate. The electrode’s life
span is limited. Standard curves over time do not seem to change much, but once they
start to show signs of degradation, they do not recuperate very much despite changes in
conditioning as recommended (Jeroschewski et al. 1996). Temperature differences when
calibrating do not appear to affect results.
Results of our intercalibration are mixed, with like methods resulting in like results. In
one of the cores (Figure 1) the amperometric and voltammetric concentrations were
similar, but in a second core, the voltammetric electrodes were saturated and were unable
to take readings. The two methods using extracted porewaters were similar in
concentrations. The fine scale data recorded higher concentrations of total sulfide with
depth. The two coarser scale measurements are similar, but did not record the same level
of concentrations in the deeper sediments. These samples were above the upper limits of
their respective techniques. This resulted in dilutions in order to bring the concentrations
into appropriate detection ranges.
Figure 1. Intercalibration profiles for total sulfide from core in Kirby Park, CA on July 2, 1999.
EXTROVERT CLIFF (JULY 21, 1999)
Two cores were processed, a core from the edge and the control. The electrode broke
when it hit a clam in the core from the edge. We decided to not start the center core
when we were advised that there were clams present in the core. The edge core profile
was depleted at the surface and increased up to approximately 6 mM (Figure 2). The
control core had no sulfide present.
Figure 2. Total sulfide profile of a bacterial mat edge at Extrovert Cliff on July 21, 1999.
CLAM FIELD SOUTH SITE (JULY 27, 1999)
Similar cores were collected from the Clam Field South Site, one from the center, one
from the edge of the bacterial mat, and a control. Note that the sulfide profile from the
center of the mat is depleted in sulfide deeper than the other two cores (Figure 3).
Figure 3. Total sulfide concentrations in cores taken from the center of a bacterial mat, the edge of the
bacterial mat and a control.
EEL RIVER (AUGUST 21 & 23, 1999)
We processed seven cores during the trip and have preliminary data, but will only present
two profiles at this time (Figure 4). The control was depleted in sulfide in the upper
sediments and increased with depth. The clam core was extremely high in total sulfide as
it increased with depth up to a concentration of over 17 mM. This core had
approximately six clams in the upper couple of centimeters. The larger diameter
electrode was able to penetrate the core eleven centimeters before breaking.
Figure 4. Total sulfide concentrations in cores taken from a control and a clam site in the Eel River on
August 23 and 21, 1999, respectively.
The voltammetric electrodes were optimized under different parameters for sulfide.
Comparisons in changes in frequency, step, amplitude and filter were done and plotted
against each other to compare results. In all cases, except for filtering, the faster or
higher the parameter improved the peak. Filtering was slightly different in that in most
cases, a higher filter did not improve the signal.
The bulk of the experimentation occurred with the amperometric electrode. Because they
had a limited life span, we concentrated on our understanding of these electrodes.
Through trial and error, we learned the intricacies of these electrodes and were able to
apply them to take measurements in cores collected in various locations. Fine scale data
of sulfide in the Monterey Canyon does not exist and the use of these electrodes could
elucidate interactions within different communities present. Accurate sulfide fluxes into
or out of the sediments can be calculated with this fine detail of concentration gradients at
the sediment water interface.
These amperometric electrodes appear to have a high measurement range, as we were
able to standardize up to 10 mM without apparent saturation. They are portable as they
were taken on the road up to the Eel River cruise at the end of August. They have a
quick response, so lots of measurements can be taken in a relatively short amount of time.
The calculations needed to generate total sulfide concentrations are relatively easy to do.
pH is a major constituent to obtain accurate data.
Comparison of amperometric and voltammetric electrodes is difficult, since there are
fundamental differences in techniques. The amperometric electrodes only measure H2S,
but have a higher detection range. Voltammetric electrodes are capable of measuring
more than one species in one scan, but the interpretation of the raw data may be more
involved. From the limited experiments performed on the voltammetric electrodes for
this project, optimization can be done and in general, faster and bigger seem to make a
positive difference in scan performance. An increase in the filter was the only parameter
that did not improve peak resolution. But this should not be considered universal, since
these parameters were limited in scope.
Our intercalibrations of different sulfide measurement techniques proved to be most
interesting, as there was no definitive answer as to which is the best technique. It appears
as if there is some loss due to oxidation perhaps. Without further comparison with a
more deliberate attempt to control more variables, such as limiting exposure to the
atmosphere and taking pH readings in the sediment and after centrifuging, absolute
answers are not possible. Each technique has its advantages and disadvantages and their
use must be made on a case by case basis.
Although much has been learned during these experiments on the amperometric
electrode, further refinements could be made to limit artifacts from the processing the
core. It is recommended that a less disruptive way of removing sediment from the core
be explored to limit exposure of the underlying sediment. In addition, more robust pH
electrodes are recommended to survive the potential harsh environments of clam beds
and remnant shells in the Monterey Canyon sediments. Increasing the diameter of the
sulfide electrode in the second set of electrodes acquired seemed to increase their
durability. The down side appears to have been a slight increase in responsiveness.
Perhaps fine-tuning is necessary to find the optimum diameter for these types of
sediments. Further experiments are necessary to determine an upper limit of these
electrodes and in situ capabilities also provide an interesting aspect in measuring sulfide.
I’d like to acknowledge Carole Sakamoto and Kurt Buck for all their help in expanding
my knowledge in sulfide and microelectrodes. There is much left to be done, but alas my
time is up. Many thanks for the patience and tolerance before, during, and after my
sulfide experiments, of all my fellow lab mates, Peter Walz, Thomas Chapin, and Josh
Plant. Special thanks to George Matsumoto for all the terrific things he did for us, from
the administrative to the entertainment. And last, but not least, my fellow interns who
made the whole experience complete. I wish them well in their future endeavors.
Barry JP, Greene HG, Orange DL, Baxter CH, Robison BH, Kochevar RE, Nybakken
JW, Reed DL, and McHugh CM (1996) Biologic and geologic characteristics of cold
seeps in Monterey Bay, California. Deep-Sea Research I 41(11-12), 1739-1762.
Brendel PJ (1995) Development of a mercury thin film voltammetric microelectrode for
the determination of biogeochemically important redox species in porewaters of marine
and freshwater sediments. Doctor of Philosophy in Chemistry, Dissertation, University
of Delaware. 141 p.
Brendel PJ, Luther III GW (1995) Development of a gold amalgam voltammetric
microelectrode for the determination of dissolved Fe, Mn, O2, and S(-II) in porewaters of
marine and freshwater sediments. Environmental Science and Technology 29(3), 751-
Buck K, Barry JP (1998) Monterey Bay cold seep infauna: quantitative comparison of
bacterial mat meiofauna with non-seep control sites. Cah. Biol. Mar. 39, 333-335.
Chen TK, Lau YY, Wong DKY, Ewing AG (1992) Pulse voltammetry in single cells
using platinum microelectrodes. Anal. Chem. 64, 1264-1268.
Jeroschewski P, Steuckart C, and Kuhl M (1996) An amperometric microsensor for the
determination of H2S in aquatic environments. Anal. Chem. 68, 4351-4357.
Lau YY, Abe T, Ewing AG (1992) Voltammetric measurement of oxygen in single
neurons using platinized carbon ring electrodes. Anal. Chem. 64, 1702-1705.
Otte S, Kuenen JG, Nielsen LP, Paerl HW, Zopfi J, Schulz HN, Teske A, Strotmann B,
Gallardo VA, and Jorgensen BB (1999) Nitrogen, carbon, and sulfur metabolism in
natural Thioploca samples. Applied and Environmental Microbiology 65(7), 3148-3157.
Revsbech NP, Jorgensen BB, Blackburn TH (1983) Microelectrode studies of the
photosynthesis and O2, H2S, and pH profiles of a microbial mat. Limnol. Oceanogr.
Revsbech NP, Jorgensen BB, Blackburn TH (1980) Distribution of oxygen in marine
sediments measured with microelectrodes. Limnol. Oceanogr. 25(3), 1062-1074.
Schulz HN, Jorgensen BB, Fossing HA, Ramsing NB (1996) Community structure of
filamentous, sheath-building sulfur bacteria, Thioploca spp., off the coast of Chile.
Applied and Environmental Microbiology 62(6), 1855-1862.