Articles
Electrochemical Sensing of Hydrogen Sulfide Traces in Biological Samples
2024 Volume 92 Issue 2 Pages 022012
Details
2024 Volume 92 Issue 2 Pages 022012
Given that trace level H2S quantitation in biological samples currently requires expensive and time-consuming instrumental procedures, we herein developed an electrochemical Ag/C sensor for the detection of H2S in protein-containing neutral solutions. The detection principle was based on the increase in galvanic current that was elicited by the sulfidation of the Ag electrode and was linearly dependent on the concentration of H2S. This linear relationship enabled the quantitation of H2S in the concentration range of 0.037–6.7 µmol L−1 in mock biological samples without any pretreatment, highlighting the potential of the Ag/C sensor for the analysis of water quality and food, blood, and other biological samples.
Sensors are widely used in the analysis of environmental and biological samples, as exemplified by biosensing,1–4 immunoassays,5–7 and real-time PCR,8,9 offering several advantages over conventional techniques. In particular, the combination of highly sensitive and selective electrochemical sensors with compact power-saving devices enables rapid and simple target quantitation and monitoring.
Hydrogen sulfide (H2S) is known as a toxic gas with an odor like rotten eggs. It is normally present in blood at <0.05 µg mL−1 (no exposure), while levels of 0.08–0.50 µg mL−1 in poisoning deaths on exposure and 30.4–131 µg mL−1 in oral poisoning deaths with sulfide or polysulfide.10 H2S has recently attracted attention as a bioactive substance in the body, and various actions have been reported, including regulation of neurotransmission,11 smooth muscle relaxation,12 cytoprotection13 and regulation of insulin secretion.14 Owing to the pathological importance of H2S,15,16 it is anticipated that the H2S concentrations in aqueous biological samples will be used as diagnostic, therapeutic, and health indicators. Simple methods for measuring H2S concentration in aquas solutions include the lead acetate paper method and ion selective electrode method. However, the former method has a high detection limit (5 µg mL−1) and is therefore not applicable to trace level measurements, whereas the latter requires alkaline conditions and is therefore not applicable to protein-containing neutral solutions such as blood. Therefore, the analysis of trace H2S in blood and other biological samples is mainly accomplished using expensive instrumental techniques such as gas chromatography mass spectrometry17,18 and liquid chromatography mass spectrometry.
H2S is water and fat soluble. It is unstable in aqueous solutions due to volatilization into the gas phase and oxidation by dissolved oxygen. In samples containing proteins such as biological tissues, H2S quickly forms bound sulfur and acid-labile sulfur, and exits an unbound H2S.18,19 Hence, the quantitation of unbound H2S requires rapid pretreatment after sampling,15,20 highlighting the need for methods enabling the rapid and simple quantitation and monitoring of trace H2S concentrations in blood and other biological samples.
Previously, we reported on a biochemical corrosion monitoring (BCM) sensor that detects trace amounts of H2S released from living Escherichia coli (E. coli).21 It was proposed a Ag/C electrode to record the galvanic currents generated by bimetallic corrosion. A Fe/Ag sensor for atmospheric corrosion monitoring based on the same principle are currently used to evaluate metal corrosivity in atmospheric environments and are called ACM sensor.22–25 The new BCM sensor (hereafter referred to as Ag/C sensor) generates a galvanic current by the reaction of a Ag anode with a small amount of H2S, which is metabolized by E. coli in the culture medium during growth. The reaction results in the formation of an Ag2S film on the Ag electrode; the Ag/C sensor can monitor growth behavior without damaging the bacteria. A correlation between the number of E. coli and T20 (T20 is defined as the time required for the current to reach 20 nA) was found, and quantitative analysis of E. coli counts was reported. The Ag/C sensor showed a linear relationship between H2S and Igal in the H2S concentration range of 0.056–7.4 µmol mL−1 (0.0019–0.25 µg mL−1) in E. coli solution containing culture medium. It was suggested that this sensor can directly quantitate H2S in the presence of proteins and amino acids without any pretreatment such as separation, concentration, or fluorescent modification. However, Ag/C sensor direct response to H2S in trace H2S analysis has not been investigated. To apply Ag/C sensor to trace measurements, it is essential to analyze the behavior of the sensor response to H2S.
The simple comb-shaped two-electrode structure of the Ag/C sensor is fabricated by screen printing and facilitates ammeter miniaturization. Moreover, this sensor does not require labeling with biological materials such as antibodies, enzymes, or DNA or the use of mediators or metal nanoparticles. In this study, the response of the Ag/C sensor to H2S was investigated with the aim of applying it to trace measurement technology. We report the results of our investigations into changes in the concentration of H2S in solution, changes in the current of the sensor electrode due to H2S, and the relationship between the H2S concentration and the sensor current value.
We fabricated the BCM sensor consisting of a Ag/C galvanic couple, as shown in Fig. 1. The Ag/C sensor consists of two electrodes: Ag on the anode and C on the cathode. The outer dimensions of the sensor were 64 mm long, 64 mm wide, and 0.8 mm thick. The comb-shaped structure on the sensor had an anode (Ag) width of 0.7 mm and a cathode (C) width of 0.7 mm (Fig. 1b). The total electrode area was 320 mm2 for the anode and 1746 mm2 for the cathode, and the total length of the electrodes where the anode and cathode faced each other was 930 mm. However, when the sensor was attached to the test cell described in section 2.5, the cathode area in contact with the test solution was 1385 mm2. The sensor substrate was a 12 µm thick Ag-plated steel plate. A 20 µm thick layer of insulating epoxy resin (S-40, Taiyo Ink Mfg. Co., Ltd.) layer was screen-printed on the substrate and thermally cured at 140 °C for 30 min. Next, a 15 µm thick layer of carbon resin (Carboloid MRX-713J-A, Tamura Corp.) was printed in layer on the insulation layer and thermally cured at 130 °C for 30 min. The resistance of the insulating layer separating the electrodes exceeded 1 GΩ at room temperature (25 °C). The fabricated Ag/C sensor was stored in a desiccator and removed immediately before testing.
Schematic representation of the Ag/C sensor. (a) Plane view and (b) enlarged view of the cross section along line A–A′.
The LB liquid medium, which contained 1.0 % (w/v) tryptone, 0.5 % (w/v) yeast extract, and 0.5 % (w/v) sodium chloride (NaCl), was autoclaved at 121 °C for 15 min and cooled. The pH of the prepared LB liquid medium was measured to be 7.0. The E. coli solution was prepared by washing cultured E. coli DH5α with sterile saline followed by suspension in LB liquid medium.21 The concentration of E. coli in this solution was determined from its optical density at 600 nm (U-3900 spectrophotometer, Hitachi), with an optical density of unity corresponding to 108 colony-forming units cfu/mL.
2.3 Quantitation of H2S in test solutionSulfide ion were quantified using an ion chromatograph (Dionex ICS-3000, Thermo Fisher Scientific K.K., Japan) equipped with an ICE AS 1 fast column (φ 9 mm × 150 mm) and ACRS 500 suppressor.26,27 The eluent and reaction reagent were 2 mmol L−1 sulfuric acid (H2SO4) solution and 0.5 mol L−1 sodium hydroxide (NaOH) solution, respectively. The electrochemical detector was equipped with a Ag working electrode and Ag/AgCl reference electrode. The sulfide ion standard solution (31 mmol L−1) was prepared by dissolving 0.75 g sodium sulfide nonahydrate (Na2S·9H2O) in degassed 0.1 mol L−1 NaOH (100 mL). For each experiment, this solution was diluted with degassed 0.1 mol L−1 NaOH to prepare working solutions with sulfide-ion concentrations of 0.031–31 µmol L−1, which were used to obtain a linear calibration curve. Next, a 25 µL aliquot of the solution in the test cell was injected into the column of the ion chromatograph for sulfide-ion quantitation. The corresponding quantitation limit was experimentally determined as 0.008 µmol L−1 (= 10 × standard deviation of the blank). In aqueous solutions, H2S exists in the form of H2S, HS−, and S2−. The proportions of these ions and H2S can be determined from the acid dissociation constant of H2S (Eqs. 1–2) and solution pH. Herein, the ionic and neutral forms of H2S are collectively referred to as H2S, while the quantified S2− concentration is denoted as H2S concentration.
| \begin{equation} \text{H$_{2}$S}\to \text{HS$^{-}$} + \text{2H$^{+}$}\quad \text{p$K_{\text{a1}}$} = 7.02 \end{equation} | (1) |
| \begin{equation} \text{HS$^{-}$}\to \text{S$^{2-}$} + \text{H$^{+}$}\quad \text{p$K_{\text{a2}}$} = 13.9 \end{equation} | (2) |
The polarization curves of the sensor electrodes were measured in the LB liquid medium, E. coli solution (108 cfu/mL), 0.5 % NaCl solution, and 0.5 % NaCl solution containing H2S. The E. coli solution was incubated at 37 °C for 90 min before the measurement and contained >0.74 µmol L−1 H2S according to the ion chromatography results. The potential dynamic polarization test was performed at a sweep rate of 20 mV/min from the rest potential to 1.0 V (Ag electrode) or −1.0 V (C electrode). Electrochemical measurements were performed at 25 °C in air using a potentiogalvanostat (273A, EG&G) and a 500 mL glass electrochemical cell containing the test solution (250 mL), a saturated calomel electrode reference electrode, and a Pt counter electrode. Ag/C sensors were covered with Teflon tape except for the electrode portion. The potentials values in this paper were regarded as the standard hydrogen electrode potential (SHE).
2.5 Galvanic current measurementGalvanic current response measurements (hereafter referred to as current response tests) were performed by dropping the H2S-containing LB liquid medium (test solution) onto the Ag/C sensor fixed to the test cell. The test cell consisted of a glass tube (42 mm ID, 32 mm height) and two resin plats was attached to the sensor to hold the test solution.21 The cell was assembled in an aseptic chamber after autoclaving each part at 121 °C for 15 min and placed in a constant temperature bath at 37 °C. After the test solution (10 mL) was dropped from the upper end of the test cell, the cell was sealed with a silicon culture plug. The Ag/C sensor was connected to a zero-resistance ammeter (model 2105, Keisokuki Center Co., Ltd.), and the galvanic current (Igal) between the electrodes was measured at 0.5 s intervals. The current generated when the Ag electrode of the Ag/C sensor functioned as an anode was expressed as positive. In the current response test, measurements were initiated with H2S-free LB liquid medium as the test solution. After the stabilization of Igal, the test solution was spiked with 3 mmol L−1 sulfide ion solution (prepared by dissolving Na2S·9H2O in degassed 0.1 mol L−1 NaOH) to achieve predetermined H2S concentrations between 3.1 and 15.6 µmol L−1. Then, a 0.3 mL aliquot of the test solution was collected and was immediately injected into the column of the ion chromatograph for sulfide ion quantitation.
Figure 2 shows polarization curves for Ag or C electrodes of Ag/C sensor in different test solutions. Figure 2a shows the polarization curves for Ag or C electrodes of Ag/C sensor in LB solution or E. coli solution, and Fig. 2b shows the polarization curves for them in 0.5 % NaCl solution, and 0.5 % NaCl solution containing 156 µmol L−1 H2S. The polarization curves of Figs. 2a and 2b are shown together in Fig. 2c. The horizontal axis of Fig. 2 is the current value is obtained as current density × electrode area (320 and 1746 mm2 for the Ag anode and C cathode, respectively). The intersections of the anodic and cathodic currents, showed by arrows (1), (2), (3), and (4) in Fig. 2, indicate the corrosion potential (Ecorr) and Igal of Ag/C sensor in each test solution. In Fig. 2a, the anodic current of the Ag electrode in LB liquid medium, shown by a pink line, rose above a rest potential of 0.25 V and sharply increased above 0.3 V. Given that the LB liquid medium contained NaCl, we expected AgCl to form (Eq. 3) on the Ag electrode. This hypothesis was supported by the X-ray fluorescence analysis results, which showed that the brown film formed on the Ag electrode after anodic polarization contained AgCl. The cathodic current of the C electrode in LB liquid medium, shown by a pink dotted line in Fig. 2a, increased slowly from a rest potential of 0.25 V to a constant current of approximately 300 µA below −0.5 V. The cathodic current was ascribed to the reduction of dissolved oxygen (Eq. 7), as the LB liquid medium was at pH 7. Given that the test was conducted in air, the partial pressure of oxygen was the same as that in the atmosphere. Therefore, 0.25 mmol L−1 (8 ppm) of dissolved oxygen is present in the aqueous solution at 25 °C, and the limiting diffusion current density of dissolved oxygen (iLO2) is determined as 20 µA/cm2.28 A constant current of 300 µA below −0.5 V corresponded to a current density of 17 µA/cm2 at the C electrode, which approximately agreed with iLO2. The Ecorr and Igal of the Ag/C sensor in the LB liquid medium, indicated by arrow (1) in Fig. 2a, were 0.28 V and 6 nA, respectively. Thus, Igal was the same as the stable current value of <10 nA observed before the Igal increment in the bacterial response test of the Ag/C sensor using E. coli.21 The anodic current of Ag electrode in the E. coli solution, shown by a red line in Fig. 2a, showed an increase around −0.30 V, which was a lower potential than that observed in LB liquid medium. Ecorr and Igal of Ag/C sensor in the E. coli solution, indicated by arrow (2) in Fig. 2a, were −0.23 V and about 1000 nA, respectively. This Igal value was close to the maximum current of 650 nA21 observed in the bacterial response test of the Ag/C sensor using 108 cfu ml−1 E. coli solution. As the pH of the E. coli solution was 7.0, i.e., equaled that of the LB liquid medium, the cathodic current, shown by a red dotted line in Fig. 2a, was ascribed to the reduction of dissolved oxygen (Eq. 7). However, the cathodic current in the E. coli solution was substantially lower than that in the LB liquid medium, which suggested that the growth of E. coli caused a decrease in the concentration of dissolved oxygen.
Polarization curves for Ag or C electrode of Ag/C sensor in different test solutions. (a) Test solution is LB liquid medium or a 108 cfu ml−1 E. coli solution incubated at 37 °C for 90 minutes. (b) Test solution is 0.5 % NaCl solution or 0.5 % NaCl solution containing 156 µmol L−1 H2S. The polarization curves of (a) and (b) are shown together in (c). Solid line: anodic current of the Ag electrode. Dotted line: cathodic current of the C electrode.
In H2S-free 0.5 % NaCl solution, the anodic current of the Ag electrode, shown by a blue line in Fig. 2b, started to increase around 0.25 V and rapidly increased above 0.3 V. After the polarization was completed, the formation of AgCl on the Ag electrode was confirmed. This rapid increase was ascribed to AgCl formation on the Ag electrode (Eq. 3). The cathodic current of the C electrode, shown by a blue dotted line in Fig. 2b, presumed a reduction current (Eq. 7) of dissolved oxygen, as a pH of H2S-free 0.5 % NaCl solution was 7. Ecorr and Igal of Ag/C sensor in H2S-free 0.5 % NaCl solution, indicated by arrow (3) in Fig. 2b, were 0.20 V and 8 nA respectively. The polarization curves of each electrode in H2S-free 0.5 % NaCl resembled those observed in the LB liquid medium, shown in Fig. 2c. By contrast, in H2S-containing 0.5 % NaCl, the anodic current of the Ag electrode, shown by a black line in Fig. 2b, increased around −0.15 V because of Ag sulfidation (Eqs. 4–6), as that in the E. coli solution. Ecorr and Igal, indicated by arrow (4) in Fig. 2b, were −0.10 V and 5 µA, respectively. The cathodic current of the C electrode, indicated by the black dotted line in Fig. 2b, could be attributed to the reduction of dissolved oxygen (Eq. 7) due to the pH of the 0.5 % NaCl solution containing 156 µmol L−1 H2S being 11. The change in this cathodic current was almost identical to that in H2S-free 0.5 % NaCl or LB liquid medium, shown in Fig. 2c. These cathodic currents were therefore considered to be due to the reduction reaction of dissolved oxygen. In summary, Igal increased in the presence of H2S, which was attributed to the resulting increase in the anodic current due to the sulfidation of the Ag electrode. Therefore, the increased Igal observed in the E. coli solution was attributed to the generation of H2S therein.
| \begin{equation} \text{Ag} + \text{Cl$^{-}$}\to \text{AgCl} + e^{-} \end{equation} | (3) |
| \begin{equation} \text{2Ag} + \text{H$_{2}$S}\to \text{Ag$_{2}$S} + \text{2H$^{+}$} + \text{2$e^{-}$} \end{equation} | (4) |
| \begin{equation} \text{2Ag} + \text{HS$^{-}$}\to \text{Ag$_{2}$S} + \text{H$^{+}$} + \text{2$e^{-}$} \end{equation} | (5) |
| \begin{equation} \text{2Ag} + \text{S$^{2-}$}\to \text{Ag$_{2}$S} + \text{2$e^{-}$} \end{equation} | (6) |
| \begin{equation} \text{O$_{2}$} + \text{2H$_{2}$O} + \text{4$e^{-}$}\to \text{4OH$^{-}$} \end{equation} | (7) |
H2S is water and fat soluble. Due to its high solubility in water of 2.257 cm3/cm3 (25 °C),29 1 g of H2S dissolves in 292 ml water at 25 °C. The pKa1 of H2S (7.02) suggests that at pH 7, the H2S and HS− are present in equal amounts (Eq. 1). However, small amounts of H2S in aqueous solutions are unstable due to volatilization into the gas phase, adsorption on glass, and oxidation by dissolved oxygen. Moreover, H2S is converted to bonded sulfur, which forms disulfide bonds with the -SH group ends of proteins and peptides contained in the LB liquid medium, and acid-unstable sulfur, which binds to the metal atoms in enzymes. Unbound H2S is present and is called free H2S.18,19 Next, it was examined the time variation on the concentration of H2S in the LB liquid medium. The test cell used for current response test was loaded with the LB liquid medium (10 mL) containing 15.6 µmol L−1 H2S as the test solution and placed in a thermostatic chamber held at 37 °C. The H2S concentration in the test solution was analyzed by ion chromatography at 0, 0.5, and 20 min. The same test was performed for 0.5 % NaCl and 30 mmol L−1 HEPES buffer (pH 7.0). Figure 3 shows the changes in H2S concentration in each test solution at 0, 5, and 10 minutes. At 0 min, the H2S concentration in the NaCl solution was 15.3 µmol L−1. At 0.5 and 20 min, the H2S concentration in this solution decreased to 15.0 and 13.3 µmol L−1, accounting for 96 % and 84 % of the spiked H2S concentration (15.6 µmol L−1), respectively. And its concentration was maintained until 60 min (shown in Fig. S1 of the Supporting Information). The H2S concentration in the HEPES buffer was 11.0 µmol L−1 (70 % of the spiked H2S concentration) at 0 min, decreasing to 8.4 µmol L−1 (53 %) at 20 min and 3 µmol L−1 (20 %) at 60 min (Fig. S1 of the Supporting Information). The H2S concentration in the LB liquid medium was 7.2 µmol L−1 (46 % of the spiked H2S concentration) at 0 min, decreasing to 6.3 µmol L−1 (40 %) at 0.5 min, and decreasing to 0.15 µmol L−1 (<1 %) at 20 min. After 20 min, H2S could not be detected in the LB liquid medium (shown in Fig. S1 of the Supporting Information). The pH values of each test solution spiked with 15.6 µmol L−1 H2S were 10.3 for 0.5 % NaCl and 7.0 for 30 mmol L−1 HEPES buffer and LB liquid medium. The volatilization of H2S in the NaCl solution was assumed to be suppressed because the fraction of non-dissociative form, H2S, in this solution (pH 10.3) was less than 0.01 %. The slight decrease in H2S concentration in the NaCl solution was ascribed to adsorption on glass and oxidation by dissolved oxygen. The HEPES buffer with pH 7 was considered to lose H2S to the gas phase, as the H2S fraction was about 50 %, and the shift of the equilibrium to H2S further contributed to volatilization. The decrease in H2S concentration in the LB liquid medium was considerably faster than that in the HEPES buffer with the same pH, which was ascribed to the binding of H2S to proteins and peptides in addition to volatilization. We also inferred that binding was prioritized over volatilization. The Ag/C sensor is thought to detect the H2S remaining in the test solution. However, the binding of H2S to proteins and peptides proceeded rapidly immediately after the addition of H2S to the LB liquid medium.
Time variations of H2S concentration in the LB liquid medium, 30 mmol L−1 HEPES buffer (pH 7.0), and 0.5 % NaCl spiked with 15.6 µmol L−1 H2S and kept at 37 °C.
Figure 4 shows the time variation of Igal and H2S concentration in the LB liquid medium spiked H2S (spiked H2S concentration = 15.6 µmol L−1). After spiking, Igal rapidly increased, which could be considered a transient response, sharply decreased to 290 nA after 20 s, and then decreased more slowly. The concentration of H2S dropped from 15.6 to 7.23 µmol L−1 upon spiking, subsequently experiencing a slow decrease. The decreases in Igal and H2S concentration were ascribed to the volatilization of H2S and its binding to proteins and other substances. Therefore, to investigate the relationship between Igal and H2S concentration in the test solution 30 s after the spike, when the transient current change stabilizes, it was performed the current response test by adding the H2S solution to the LB liquid medium. Figure 5 summarizes the relationship between Igal and H2S concentration in the analyzed test solution. It was revealing that the Ag/C sensor could detect H2S at levels above 0.037 µmol L−1 and that Igal increased with increasing H2S concentration in the LB liquid medium. Equation 8 presents the linear relationship between Igal and H2S concentration in the range of 0.037–6.7 µmol L−1. Thus, the Ag/C sensor could quantify traces of H2S in neutral protein-containing solutions at concentrations of 0.037–6.7 µmol L−1.
| \begin{equation} I_{\text{gal}}\ (\text{nA}) = 52.2\ [\text{H$_{2}$S}]\ (\text{$\unicode{x00B5}$mol$\,$L$^{-1}$}) + 6.18\quad (R^{2} = 0.955) \end{equation} | (8) |
Changes in the galvanic current and H2S concentration in the LB liquid medium spiked with 15.6 µmol L−1 H2S at 37 °C.
Relationship between H2S concentration and Igal in the LB liquid medium 30 s after H2S addition at 37 °C.
We examined the response of the Ag/C sensor to low H2S concentrations, the effects of time on the concentration of H2S in LB liquid solution, and the relationship between the H2S concentration and the sensor current. The sensor was responsive to H2S in 0.5 % NaCl and the LB liquid medium, as evidenced by an increase in the galvanic current (Igal), which, in turn, was attributed to the increase in the anodic current due to the sulfidation of the Ag anode by H2S. The cathodic (C electrode) current was attributed to the reduction of dissolved oxygen. Therefore, the increase in Igal in the E. coli solution was ascribed to the generation of H2S therein. The concentration of H2S in the LB liquid medium decreased to <1 % of the spiked concentration within 20 min after spiking because of the binding of H2S by proteins and other substances. In the concentration range of 0.037–6.7 µmol L−1, Igal linearly increased with increasing H2S concentration in the LB liquid medium, which showed that this analyte could be reliably quantified in neutral samples containing proteins and other substances without any pretreatment. Therefore, our sensor can be applied to the analysis of water quality and food, blood, and other biological samples.
The data that support the findings of this study are openly available under the terms of the designated Creative Commons License in J-STAGE Data at https://doi.org/10.50892/data.electrochemistry.24785187.
Chiyako Touge: Conceptualization (Equal), Data curation (Equal), Formal analysis (Equal), Investigation (Lead), Methodology (Lead), Writing – original draft (Lead)
Michiyo Nakatsu: Conceptualization (Lead), Data curation (Lead), Formal analysis (Lead), Investigation (Lead), Methodology (Lead), Visualization (Lead), Writing – original draft (Equal), Writing – review & editing (Lead)
Mai Sugimoto: Data curation (Equal), Investigation (Equal), Methodology (Equal)
Hiroaki Sakamoto: Conceptualization (Equal), Project administration (Equal), Writing – review & editing (Equal)
The authors declare no conflict of interest in the manuscript.
C. Touge, M. Nakatsu, and H. Sakamoto: ECSJ Active Members
M. Sugimoto: ECSJ Student Member
