An In Situ Analyzer for Two-Dimensional Fe(II) Distribution in Sediment Pore Water Based on Ferrozine Coloration and Computer Imaging Densitometry.

A novel integrated analyzer was developed for the in situ determination of two-dimensional (2D) dissolved Fe(II) distributions in sediment pore water. The analyzer utilized gel enrichment and optical imaging techniques. An image probe mainly consisting of a gel holder and portable document scanner was designed to be inserted into sediment. The gel holder exposed to the sediment was made to hold a polyacrylamide gel strip (diffusive gel) and polyacrylamide gel strip impregnated with C18 and coated with ferrozine (concentrating gel). The concentrating gel strip could accumulate the dissolved Fe(II) in pore water and produce a magenta-colored Fe(II)-ferrozine compound on the gel strip in two dimensions. The portable document scanner sealed in a transparent box and stuck onto the back of the gel holder could record gel images from the back of the concentrating gel strip. Gel images with grayscale intensities were acquired and analyzed using ImageJ software, and Fe(II) concentration was determined based on a deployment time related calibration curve established in the laboratory. The measurement accuracy and precision were investigated. The quantitative range reached up to 200 μmol L-1. The method and analyzer exhibit distinct characteristics of in situ enrichment and measurement; they were successfully applied to determine the 2D Fe(II) distribution in lake and marine sediment pore waters.

Introduction

As one of the important biogenic elements, iron (Fe) is an essential nutrient for most marine and terrestrial organisms.[1,2] Iron exists in various chemical states in sediment pore waters and is involved in diverse biogeochemical reactions such as remineralization of organic matter, sulfur cycling, and phosphorus cycling.[2,3] Therefore, it is important to evaluate the speciation, content, and distribution of iron in pore water in sediment studies. Dissolved Fe(II) is the dominant species of iron in sediment pore water.[2−6] The spatial heterogeneity of both horizontal and vertical distributions of Fe(II) in sediment pore water is substantial, especially in the vertical direction.[4−6] It is highly desirable to obtain an accurate two-dimensional (2D) distribution of Fe(II) in sediment pore water. For example, Zhu et al. developed a single-use planar optical sensor for measuring high-resolution, 2D Fe2+ distributions in marine sediments.[5]

Sampling and sample processing are specialized in a conventional method to determine the vertical distribution of Fe(II) in sediment pore water. A sediment core is usually obtained with a suitable columnar sediment sampler[7] followed by layering in an inert atmosphere; then, the pore water in each layer is extracted for further treatment.[8,9] Although the conventional method is widely used, it is time-consuming and tedious with complex operation. Even more, Fe(II) is easily oxidized during sampling and processing.

Some microelectrodes are developed to measure the key dissolved redox species including iron in sediment using electrometric methods.[10−14] These microelectrodes have fast response speed and high sensitivity; however, they can be very fragile. Usually, core samples are taken, and measurements are carried out in laboratories.[12,13] Only a very few applications have been reported in real field situation.[14−16]

In recent years, some in situ sampling methods have been developed to measure Fe(II) concentrations in sediment pore water, including diffusive gradients in thin films (DGT), diffusive equilibration in thin films (DET), and high-resolution peeper (HR-Peeper). The DGT technique has great advantages in the determination of metals and organic pollutants in sediment, water, and soil.[17−22] After the DGT devices are deployed into a sediment for a specific time, they are collected and sent to the laboratory. The distribution of target metals can be evaluated using different methods.[21−31] The most widely used method involves dividing the binding gel layer into pieces, eluting the targets from each gel strip with acid, and measuring the extracted concentrations using common analytical methods such as atomic absorption spectrophotometry (AAS) and inductively coupled plasma-mass spectrometry (ICP-MS).[23,24] Using the equations established based on Fick’s first law, the amount of target analytes in the binding layer and sediment pore water can be calculated. Computer imaging densitometry (CID)[21,25] and laser ablation ICP-MS[26] are also used to analyze gel samples. All the methods can provide high-resolution 2D patterns of targets, and the spatial resolutions of CID and laser ablation ICP-MS methods are up to the micrometer level. The analytical method and performance of DET are similar to those of DGT.[27−31] For example, the polyacrylamide gel layer is used as a DET device for 2D Fe(II) concentration determination and as the diffusive layer of a DGT-like device for the sulfur species study.[28] Pagès et al. developed an in situ colorimetric DET method for reactive phosphate, which would provide 2D, high-resolution distributions in sediment pore water.[31]

According to the theory of diffusive equilibration, using multichambered HR-Peeper devices,[32−34] the target concentrations in the receiver solutions in each chamber are the same as those in the surrounding pore water. The distribution of the target analyte in pore water is obtained by retrieving and analyzing the solution in each chamber. The technique has the advantages of simple operation and availability of simultaneously measuring multiple chemicals. However, Fe(II) is highly sensitive to oxygen. Therefore, it is crucial to preserve the samples, or else the concentration measured can be much lower than the actual level.[33]

Although the abovementioned in situ sampling techniques enable tracing of target analyte distributions in sediments, the samples still need to be brought back to the laboratory for subsequent processing and analysis. Appropriate methods of sample preservation are required to prevent Fe(II) from being oxidized and contaminated. For fast analysis and less error, in situ determination is in high demand.

In our previous study,[35] a sensor system, which included a C18-ferrozine-based concentrating gel and an optical imaging device fitted with an LED light source and camera, is developed for in situ measurement of dissolved Fe(II) in sediment pore water. However, the device is rather preliminary, relatively cumbersome, and can only detect Fe(II) at one point (25 mm in diameter). The detection range, sensitivity, and accuracy are limited by the LED light source and camera. Moreover, since a camera needs a certain focal length, the sensor size would become relatively large. In this study, a compact and novel analyzer is proposed based on ferrozine coloration and CID for the in situ determination of 2D Fe(II) distribution in sediment pore water. The analyzer mainly consists of gel strips that could selectively color the accumulated Fe(II) from water and a portable document scanner that could record the gel color change and produce 2D images. On the contrary to the camera, a scanner can obtain the image as closer to the objective as possible and thus reduce the sensor size. Once the probe of the in situ analyzer is inserted into a sediment, the color intensity of gel is related to the Fe(II) concentration in the surrounding pore water and deployment time. The in situ analyzer acquires and saves the color gel images in real time. After the images are retrieved, the grayscale intensity of gel images stored therein is analyzed, and a 2D Fe(II) distribution is shown based on the calibration curve of grayscale intensity vs Fe(II) concentration previously prepared in the laboratory. With the developed analyzer, sampling and sample processing could be omitted, and sample contamination and preservation could also be avoided. The integrated analyzer was successfully applied in lake and marine sediments, and the in situ analysis of 2D distributions of dissolved Fe(II) was realized.

Experimental Section

Materials and Parts

The materials and parts for building the analyzer are as follows: a 6 mm polymethyl methacrylate (PMMA) plate (Shenzhen Hongwang Mould, China); a portable document scanner (I2, Guangzhou Netum Electronic Technology, China); a 35 mm stepper motor (FY35EC180A, Shenzhen Xingfengyuan Mechanical & Electrical, China); a 24 V, 5000 mAh lithium battery (Dongguan Mingbei Electronic Technology, China); a 11.1 V, 3400 mAh lithium battery pack (Dongguan Mingbei Electronic Technology, China); four synchronous pulleys and two synchronous belts (Suzhou Moerqi Hardware Mechanical & Electrical, China); an infrared remote control relay with a remote control (Dongguan Nuolings Electronic Technology, China); two DC-DC voltage regulator modules (Yujia Electronics, China); a system circuit board that contained an STM32F103C8T6 CPU (STMicroelectronics, USA).

Chemicals

Ultrapure water (pure water, resistivity ≥18.2 MΩ cm) was freshly obtained from a pure water system (Millipore, USA) and used for the preparation of all solutions. Ferrozine (3-(2-pyridyl)-5,6-diphenyl-1,2,4-triazine-p,p′-disulfonic acid monosodium salt hydrate) was obtained from Sigma-Aldrich (USA). Bondesil-C18 particles (40 μm) were obtained from Agilent Technologies (USA). All other chemicals used were of analytical or reagent grade and supplied by the Sinopharm Chemical Reagent (China).

A stock solution of 10.0 mmol L–1 Fe(II) was prepared by dissolving ferrous ammonium sulfate in 1 mol L–1 hydrochloric acid and stored at 4 °C while not in use. Working solutions of Fe(II) were freshly prepared by diluting the stock solution with pure water and adding solid ascorbic acid to a concentration of about 0.1 mol L–1 to maintain the redox species of Fe(II).

Preparation of Polyacrylamide Gel and C18-Ferrozine Concentrating Gel

In this study, polyacrylamide gel and polyacrylamide-based gel impregnated with C18 and coated with ferrozine (C18-ferrozine concentrating gel) were used as the diffusive phase and concentrating phase, respectively. Polyacrylamide gel is a common gel used as a diffusion phase. Ferrozine exhibits good selectivity toward Fe(II), forming a stable magenta-colored complex. C18 was selected because it could tightly absorb ferrozine and could be well mixed in polyacrylamide gel.[35]

The formulas of the two gels are based on our previous study.[35] Briefly, a polyacrylamide gel solution was made of acrylamide, N,N′-methylene diacrylamide, and pure water. Freshly prepared ammonium persulfate solution and N,N,N′,N′,-tetramethylethylenediamine were added to the polyacrylamide gel solution, mixed well, and pipetted into a prepared casting assembly to produce the polyacrylamide gel. The C18-ferrozine concentrating gel was prepared from a polyacrylamide hydrogel impregnated with C18 and coated with ferrozine. About 36 h was needed for the production of a stable and usable C18-ferrozine gel sheet, including gel preparation, C18-ferrozine gel impregnation, and cleaning.

Based on previous study results,[35] the gel preparation processes were further optimized to improve the reproducibility of gels. The thickness of diffusion gel was maintained at 0.8 mm, and that of the C18-ferrozine concentrating gel increased from 0.5 to 0.8 mm for holding more ferrozine, thus increasing the capacity. As a result, the upper limit of the measurement range increased from 100 to 200 μmol L–1.

In this study, gel strips (212 × 21 mm) were placed in a rectangular gel holder and used for measurement in the field. Round gels (a diameter of 25 mm) were placed in a round device (similar to a commercial DGT device) and used for the characterization tests in the laboratory. For installing, a C18-ferrozine concentrating gel, diffusion gel, and polyether sulfone membrane (0.45 μm) were placed onto the gel holder in sequence.

Design and Construction of the In Situ Analyzer

The in situ analyzer consists of an image probe and a control box, which communicate with each other via a waterproof cable and connection plugs, as shown in Figure A,B. The key components of the image probe consist of a portable scanner together with its driving parts that were sealed in a waterproof housing and gels together with their holder that were stuck onto the outer housing.

Figure 1

In situ analyzer: (A) schematic view of the analyzer (strips: white, polyether sulfone membrane; blue, diffusion gel; red, concentrating gel), (B) photograph of the analyzer, and (C) schematic diagram of the synchronous belt driving mechanism.

The waterproof housing of the image probe is made of transparent acrylic plates, which allowed a light transmittance of over 90%. The probe contained a modified portable document scanner and a synchronous belt driving mechanism (Figure C) that drove the scanner to reciprocate in a straight line. The scanner was adapted from a commercial handheld scanner and fitted with a memory card, which could be programmed to record and save the color images of gels.

The synchronous belt drive mechanism consists of a standard 35 mm stepper motor, a drive shaft, a driven shaft, two sets of synchronous wheels, two synchronous belts, and two limit switches. In operation, the stepper motor drove the drive shaft to rotate, and then the rotary motion was converted into a linear motion of synchronous belts through synchronous wheels. After the scanner moved to the specified position, the limit switch was triggered, and a scan was completed. Subsequently, the stepper motor moved in reverse, driving the scanner back to the initial position and waiting for the next scan. The effective travel distance of the synchronous belt linear drive mechanism was 52 mm. When the scanner was working, the image area obtained from one scan was 217 × 52 mm2, while 212 × 21 mm2 with gel coloring remained, and the rest parts were removed. The scanning time could be adjusted, and it was set as 75 s in this study.

The outer shell of the probe housing contained a gel holder (Figure A), including a bottom plate and cover plate with a rectangular sampling window. For more details about the probe size, see the Supporting Information Figure S1. The gels including C18-ferrozine concentrating gel and diffusion gel and a membrane were paced onto the gel holder in sequence. Then, the cover plate was buckled on the holder and tightened with polypropylene (PP) screws.

The control box was a waterproof box containing a control module for the probe, mainly composed of a lithium battery, a single-chip microcomputer, and an infrared remote control relay. The core component, STM32F103C8 single-chip microcomputer, controlled the coordinated operation of the analyzer. The infrared remote control relay was placed between the lithium battery and single-chip microcomputer to switch on/off the probe within a distance of ∼10 m.

The pixel resolution of the image provided by the developed in situ analyzer was 330 pixels in width and 2560 pixels in height. Both horizontal and vertical resolutions were 300 dpi.

Deployment of the Analyzer

When the analyzer was applied in a field, the image probe with gels on it was slowly and carefully inserted into the sediment, and the control box could be put on a nearby mud flat or floated on water. Care should be taken to avoid too much disturbance of the sediment. The dissolved Fe(II) in the pore water would diffuse through the polyacrylamide gel from the rectangular sampling window, be concentrated onto the C18-ferrozine gel, and react with the ferrozine in the gel to form a color complex. The image of the produced magenta color of the Fe(II)-ferrozine compound on the concentrating gel strip in 2D was recorded using a scanner.

CID Analysis

ImageJ 1.46r was used to convert the scanned images of gels to grayscale intensity. The average grayscale intensities of magenta-colored zones in red-green-blue (RGB) channels were analyzed. Only the grayscale intensities of the G channel were used in subsequent analysis because of its highest sensitivity toward Fe(II) concentration.

Application Sites

A lake with a water depth of about 1 m located in the Xiang’an Campus of Xiamen University, China, was selected as one of the sites for field application. Another site was located at a marine mangrove area in Jimei District of Xiamen, China. The water depth was 0 m at low tide and ∼0.5 m at high tide. The salinity of pore water was ∼15‰. The applications were carried out in December 2018.

Results and Discussion

Adsorption Ability of Ferrozine in the Concentrating Gel

For this study, ferrozine should be firmly attached to the C18-ferrozine concentrating gel when in use. Because ferrozine is soluble in water, its adsorption ability in the gel should be investigated. Two pieces of prepared polyacrylamide-based gel impregnated with C18 were immersed in a ferrozine solution of 0.01 mol L–1 for 1 h. The gels coated with ferrozine were taken out, rinsed with pure water three times, and immersed in 400 mL of pure water for extraction. 0.5 mL of extraction solution was taken at a designed time, and the concentration of ferrozine in the solution was determined using a spectrophotometric method.[36]Figure shows the concentration change in ferrozine in the solution as the extraction time increased. Additionally, after 24 h, the extraction solution was replaced with pure water, and the gels were further extracted for 12 h. Then, the solution was analyzed for ferrozine, and no ferrozine was detected. Then, the solution was analyzed for ferrozine. The experiment was repeated three times, and no ferrozine was detected.

Figure 2

Ferrozine concentrations at different extraction times.

Figure shows that once the ferrozine-coated gels were immersed in pure water, the unattached ferrozine was dissolved into the extraction solution. After 2 h, no more ferrozine was released from the gels. Then, the prepared C18-ferrozine concentrating gels were soaked in pure water for 8 h before use.

Measurement Repeatability of the Scanner

To evaluate the measurement repeatability of the scanner, a round device containing gels was exposed to a 75 μmol L–1 Fe(II) solution for 60 min and then scanned 27 times using a scanner. A blank was taken for comparison. The pH of solutions was adjusted to 6, and ascorbic acid was added to each solution to maintain the solution in reductive conditions. The images were analyzed, and the results are shown in Figure . The grayscale intensities of the blank are 85.5 ± 0.7, and the RSD is 0.8%. The grayscale intensities of the concentrating colored gel are 49.9 ± 0.8, with an RSD of 1.5%. The results indicate good measurement repeatability of the scanner.

Figure 3

Repeatability of image grayscale intensities obtained using the scanner. Fe(II) concentration, 75 μmol L–1; pH, 6.0; deployment time, 60 min.

Uniformity of the C18-Ferrozine Concentrating Gel

The uniformity of gels, especially the C18-ferrozine concentrating gel, significantly affected the accuracy and reproducibility of determination results and was evaluated in this experiment. Fifteen round concentrating devices with gels prepared at different times were deployed into 75 μmol L–1 Fe(II) solution for 60 min. The gel images were acquired using a scanner, and the average grayscale intensity from the spots of each image is shown in Figure . The average grayscale intensity of the 15 concentrating gels is 52.9, and the RSD is 8.4%, showing good uniformity and reproducibility of the concentrating gels.

Figure 4

Grayscale intensities of different C18-ferrozine concentrating gels. Fe(II) concentration, 75 μmol L–1; pH, 6.0; deployment time, 60 min.

Establishment of the Calibration Curve

Based on our previous study,[35] 60 min was selected as the optimal deployment time. A series of Fe(II) solutions were prepared, and ascorbic acid was added to each solution to maintain the solution in reductive conditions. The pH of solutions was adjusted to 6. Two C18-ferrozine concentrating round devices were placed in each solution for 60 min. The gel images and grayscale intensities were obtained.

More than five grayscale intensities of blank gels measured in this study were averaged and defined as Gblank, and the mean grayscale intensity of the two sample concentrating gels was set as Gsample. The grayscale intensity difference between Gsample and Gblank is represented as “blank-corrected grayscale intensity”. A calibration curve was established using the blank-corrected grayscale intensity and Fe(II) concentration. The data were fitted with a quadratic equation with an R2 value of 0.9877, as shown in Figure . When the Fe(II) concentration was higher than 200 μmol L–1, the blank-corrected grayscale intensity did not increase, demonstrating that the concentrating gel became saturated with Fe(II) and reached its binding capacity. Therefore, the upper limit of detection of this method is considered to be 200 μmol L–1. In other words, quantification analysis of the analyzer was limited at up to 200 μmol L–1. Some microenvironment information with Fe(II) concentration higher than 200 μmol L–1 could be missed or underestimated, and attention should be paid in these cases. However, this range is already suitable for most of the marine sediments and normal lake sediments. The binding capacity of the concentrating gel is briefly discussed in our previous study.[35]

Figure 5

Calibration curve of blank-corrected grayscale vs Fe(II) concentration. pH, 6.0; deployment time, 60 min.

Effect of Salinity

The effect of salinity on analyzer performance was evaluated. Six C18-ferrozine concentrating round devices were soaked in a series of 75 μmol L–1 Fe(II) solutions with different salinities for 60 min, and their grayscale intensities were determined. The results shown in Figure indicate that even though there might be a very slight grayscale intensity decrease with an increase in salinity, no significant correlation was concluded from a statistical calculation. Moreover, compared with those in our previous study,[35] the gel preparation processes were further optimized and may improve the properties of gels, including the ability against salinity interference.

Figure 6

Effect of salinity on the grayscale intensity of the C18-ferrozine concentrating gel (n = 3). Fe(II) concentration, 75 μmol L–1; pH, 6; deployment time 60 min.

Recovery and Comparison with a Conventional Method

Experiments were carried out to evaluate the recovery and accuracy of this method. A lake (see Section ) surface sediment was collected and extracted with pure water (about 1:10 in weight) to prepare a sediment solution. A portion of the solution was taken, and its pH was adjusted to pH 6. Three C18-ferrozine concentrating round devices were deployed in solution for 60 min. Four Fe(II) solutions with different concentrations were used for the test, and the dissolved Fe(II) concentration was determined. The results are shown in Table ; the recoveries of Fe(II) were in the range of 91–105%. This is acceptable for sediment pore water analysis.

Table 1

Recoveries of Dissolved Fe(II) from Spiked Water Samples of a Lake Sediment (n = 3)

spiked Fe(II) (μmol L–1)	Fe(II) (μmol L–1) found (mean ± SD)	RSD (%)	recovery (%)
0	not detectable
10	9.9 ± 2.0	20.3	99.3
50	45.7 ± 0.2	0.5	91.4
75	78.2 ± 7.8	10.0	104

The proposed in situ analyzer and a conventional sampling method with spectrophotometric analysis[36] were used at the same time to determine the vertical distributions of Fe(II) in the pore water of a lake (see Section ) sediment. The distance between the two sites for the two methods was about 1 m. The results of the two methods were compared to evaluate the reliability of the proposed analyzer.

Using the conventional sampling method, the pore water in 0–200 mm depth sediment was extracted in layers at intervals of 20 or 30 mm, and the concentrations of Fe(II) were determined using a spectrophotometric method.[36] A very small amount of pore water could be extracted from the sediment at 160 to 200 mm depth. The volume of pore water extracted at 180 mm depth was less than 1 mL, not enough for Fe(II) analysis; thus, the datum at this point is missing.

The vertical distributions of Fe(II) measured using the two methods are shown in Figure ; similar distribution patterns are observed. The Fe(II) concentrations at the same depth determined using the two methods were considered to be close, indicating that the proposed analyzer has reliable measurement accuracy. The reasons for deviation in the two methods include systemic error, measurement error of grayscale intensity, and especially, sampling and analytical errors with conventional methods because Fe(II) in samples might be oxidized during handling. The concentration of Fe(II) on the water–sediment interface obtained from in situ measurement was about 40 μmol L–1. The abnormal result could be due to the fact that the insertion position of the image probe was too deep. Since the lake water was turbid and depth was about 1 m, when the probe was inserted into the lake sediment, it was difficult to control the depth position.

Figure 7

Dissolved Fe(II) distribution pattern in sediment pore water obtained using the proposed in situ analyzer and conventional methods.

Field Application

Field experiments were conducted in both lake and marine sediments. The gel strips were assembled onto the holder of the image probe, and the probe was inserted into sediment until the gels were completely submerged in the sediment.

The probe was taken out after being developed for 60 min. The grayscale intensities of the gel images at deployment times of 0 and 60 min were analyzed, and the blank-corrected grayscale intensities were used to calculate the Fe(II) concentrations based on the calibration curve described in Section .

Figure shows the 2D distribution of Fe(II) obtained from the field experiment. The corresponding 2D Fe(II) distribution patterns (200 × 18 mm) in Figure B,D were created from data of the center part (dotted frame area) of the original gel strips (212 × 21 mm) in Figure A,C. For the Gblank images of Figure A,C, see the Supporting Information Figure S2. Clearly, in the vertical direction of both lake and marine sediments, the deeper the sediment, the higher the concentration of dissolved Fe(II) in pore water. In the horizontal direction, no obvious regularity was observed for Fe(II) distribution. The results are consistent with those reported in other studies.[4,5] In natural sediments, heterogeneity is significant; therefore, the in situ 2D measurement technique is very important.

Figure 8

Distributions of Fe(II) in sediment pore waters: (A) original gel picture of lake sediment, (B) the corresponding 2D Fe(II) distribution pattern converted from the dotted frame area in panel (A), (C) original gel picture of marine sediment, and (D) the corresponding 2D Fe(II) distribution pattern converted from the dotted frame area in panel (C).

Conclusions and Perspective

A gel enrichment technique and CID technique were used in this study to develop a novel in situ method and analyzer for the determination of the 2D Fe(II) distribution pattern in sediments. The method exhibits outstanding characteristics of real in situ analysis with a fast response and simple operation. The analyzer was successfully applied in lake and marine sediments, and the results were comparable with those of conventional sampling and analytical methods, verifying the feasibility and reliability of the analyzer.

In the future, the method and analyzer can be extended to measure other target analytes in sediments, such as sulfide, using different concentrating gels with selective reagents. The design of this analyzer could be improved, for example, to reduce the thickness of the probe so that the sediment is less disturbed. The performance of this analyzer could be improved by optimizing the preparation process of gel and adjusting the internal light source of the probe. Compared with our previous study, the analyzer weight was reduced to about 2/3 and the upper limit of the measurement range increased from 100 to 200 μmol L–1. Most importantly, 2D measurement was realized in the present study.