A High‐Performance Phosphate Potentiometric Sensor Based on Cobalt Nanoparticles Electroplated on Hierarchically Porous Laser‐Induced Graphene

Phosphate is an important fertilizer in agriculture and electrochemical sensors based on cobalt (Co) are promising for measurement of phosphate. The deposition of Co onto conductive substrates to form a nano/microstructured Co layer with rich active sites is a typical strategy to improve detection performances. However, the widely used substrates show smooth surfaces, which cannot provide extra active sites. Herein, a high‐performance phosphate sensor is fabricated by generating Co nanoparticles (NPs) on a hierarchically porous laser‐induced graphene (LIG) matrix with a large surface area. Specifically, a Co layer is first electroplated on the LIG matrix. Nevertheless, the pores of LIG will be blocked by the dense Co layer so the performances of the fabricated LIG‐Co sensor exhibit limited enhancements. Therefore, a “nearly all etching (NAE)” technique is developed to transform the dense Co layer to dispersive Co NPs and re‐expose the pores of LIG, both increasing the active sites. Compared to the LIG‐Co sensor, the LIG‐Co NP sensor displays overall superiority. The response time, sensitivity, and linearity are enhanced from ≈60 to ≈10 s, from 0.012 to 0.069 V dec−1, and from 0.973 to 0.998, respectively. The LIG‐Co NP sensor also shows a decent anti‐interference capability.


Introduction
Phosphorus is one of three major macronutrients essential for plant life, [1] which is acquired by plants mainly in the form of phosphate. [2,3]Appropriate fertilization of phosphate can significantly improve agricultural crop yield, but overuse will lead to low yield, soil erosion, and water eutrophication. [4,5]hus, it is of great importance to accurately determine the concentration of phosphate.[13][14] Electrochemical detection is an alternative method to overcome the above limitations.][17][18] Moreover, the equipment for electrochemical detection is portable and inexpensive, enabling in-field measurement at low cost.Electrochemical detection of phosphate including potentiometric, amperometric, and conductometric types. [8,19,20]Amongst, the potentiometric method, measuring phosphate concentration by correlating the open circuit potential (relative to the reference electrode) of ion-selective electrode (ISE) surface to the phosphate concentration, is widely applied with the features of low detection limit, high selectivity, and low requirement of equipment. [4,20,21]ccording to the above detection mechanism, ISE is the key for potentiometry.Although phosphate ISE can be fabricated by various materials, cobalt (Co) is commonly selected due to its high affinity to phosphate. [7,22,23]Co rod was initially used to detect phosphate [22] and then Co was coated on diverse substrates to form Co layer with micro or nanostructures. [24]The large surface areas of these structures offer more active sites, thus enhancing the detection performances. [23,25,26]The typical substrates include planar or cylindrical gold (Au) [27] and Co [26] with smooth surfaces, whose specific surface areas are small. [9,26]To this end, coating Co on large-surface-area substrates is hypothesized to provide more active sites and further improve the detection performances.
To conceptually verify this idea, we developed a phosphate potentiometric sensor based on Co nanoparticles (NPs) electroplated on hierarchically porous laser-induced graphene (LIG).In addition to large surface area derived from the hierarchically porous structure, [25,28] the advantages of simple fabrication and high conductivity have promoted the application of LIG as electrochemical sensors for the detection of nitrite (NO 2 − ), [29] glucose, [29] hydrogen peroxide (H 2 O 2 ), [30] and heavy metal ion (Cd 2+ and Pd 2+ ). [25,31,32]Here, LIG was first synthesized by a direct laser writing (DLW) technique from a polyimide (PI) substrate through a photothermal reaction [33,34] (Figure 1a,b).Then, Co was electroplated onto the LIG substrate and formed a dense Co layer (Figure 1c).However, this dense layer blocked the porous structure of LIG and the fabricated LIG-Co sensor exhibited limitedly enhanced performances compared to a pure Co rod sensor.To solve this problem, we invented a "nearly all etching (NAE)" technique to ablate 98.5% Co atoms through slow electrochemical corrosion [21] by immersing the LIG-Co sensor in a potassium hydrogen phthalate (KHP) solution for 4 days.As a result, the dense Co layer was transformed to dispersive Co NPs and the pores of LIG were re-exposed, offering abundant active sites for phosphate detection (Figure 1d).Compared to the LIG-Co sensor, the response time, sensitivity, and linearity (R 2 ) of the LIG-Co NP sensor were improved from ≈60 to ≈10 s, from 0.012 to 0.069 V dec −1 , and from 0.973 to 0.998, respectively.The LIG-Co NP sensor also demonstrated a wide linear detection range of 10 −5 -10 −1 mol L −1 and a good anti-interference capability both in a configured solution and a natural water solution.These performances are outstanding compared to the literature (Table S1, Supporting Information), endowing the LIG-Co NP sensor with great application potential in agriculture, industry, and household (Figure 1e).

Preparation of LIG
LIG was in situ fabricated on PI films by the DLW technology.Specifically, laser irradiation induced lattice vibrations in poly-imide (PI), resulting in high local temperatures.The high temperatures broke the chemical bonds of PI and rearranged carbon atoms to a graphite-like structure, while other atoms were recombined and released as gases. [35]A "two-square" shape was drawn in CAD software and written on the PI film.The upper part of the PI film was converted to LIG via a photothermal reaction, [33] whereas the lower part remained as PI, which can be used for a self-supported electrochemical sensor (Figure 1b).The fabricated sample is shown in Figure 2a where the larger square is the sensing area and the smaller square is the connecting area.Scanning electron microscopy (SEM) was employed to characterize the surface morphology of LIG (Figure 2b).A parallel pattern appears on LIG in line with the motion route of the laser beam.It also displays a highly porous structure due to the fast escape of gaseous by-products. [36,37]The porous structure is further unveiled to be hierarchical using high-magnification SEM, in which the pore size ranges from 0.052 to 21 μm (Figure 2b,c).The hierarchically porous structure owns a large specific surface area, offering abundant active sites for Co deposition and phosphate detection.Raman spectroscopy was performed to characterize the chemical structure of LIG.As shown in Figure 2d, three main characteristic peaks appear in the prepared LIG, i.e., D peak (1347.16cm −1 ), G peak (1585.15cm −1 ), and 2D peak (2694.1 cm −1 ) corresponding to structural defects, graphite carbons, and second-order zone boundary phonons, respectively [38] .The relatively high intensity of the D peak and the relatively low intensity of the 2D peak compared to the G peak uncover that the prepared LIG is composed of 3D porous graphene. [39,40]he conductivity is another factor determining the detection performance of electrochemical sensors, i.e., higher conductivity is associated with higher detection performance. [31,32]In order to improve the conductivity of LIG, we changed laser power and laser speed and took the sheet resistance as the indicator.It was found that as the laser power increases from 9.2 to 9.6 W, the sheet resistance of LIG decreases first and then increases (Figure 2e).When the power reaches 9.4 W, the sheet resistance reaches the lowest value of 8.5 Ω per □.The influence of laser speed on the sheet resistance shows the same trend as that of laser power, and the sheet resistance reaches the lowest value at the speed of 30 mm s −1 (Figure S1, Supporting Information).Here is the reason for these phenomena.During the laser writing process, low laser power or fast scanning speed will lead to deficient laser energy, which can only realize partial conversion of PI into LIG; high laser power or slow scanning speed will lead to superabundant laser energy, which can destroy the structure of the LIG and even burn through the PI film. [35,41,42]he laser with the optimum parameters (laser power = 9.4 W, laser scanning speed = 30 mm s −1 ) can provide moderate energy, thus generating LIG with the best structure and the lowest resistance.
In addition to electrical conductivity, electrochemical conductivity is a more important property of an electrochemical sensor.Electrochemical impedance spectroscopy (EIS) was performed on the prepared LIG.As shown in Figure 2f, the charge transfer resistance (R ct ) of LIG can be read as 3.5 Ω from the diameter of the minicircle, indicating the low electrochemical impedance and strong charge transfer ability of LIG. [32,43]

Electroplating Co Layer Onto LIG
After the preparation of LIG matrix, Co was electroplated onto LIG to provide phosphate responsivity.A Co plate and the LIG matrix were immersed in a standard plating solution as the anode and the cathode.The plating solution was composed of CoSO 4 , NaCl, and H 3 BO 3 , for the purposes of providing Co 2+ ions, increasing the solubility of the anode, and regulating the pH value, respectively. [44]A constant current (0.36 A) was applied to drive the electroplating.A magnetic stirring was performed to expel bubbles generated from hydrogen evolution reaction and speed up the ionic migration. [45,46]As the plating continued, Co was deposited onto the LIG matrix from the edge to the middle (Figure S2, Supporting Information) due to the higher cathode current concentrated on the margins. [45]The non-uniform current distribution would also cause a thicker Co layer on the margins.After 180 s, the sensing area of LIG was completely covered by a grey Co layer (Figure 3a).
SEM demonstrates that there is no more porous structure but dense metallic textiles on the surface of the LIG-Co electrode (Figure 3b).High-magnification SEM displays that the sizes of the Co grains are on a micrometer scale (Figure 3c).An energy dispersive spectrometer (EDS) was performed in this area to study the chemical components.Elemental mapping images show that C, Co, and O are uniformly distributed.Amongst, Co has the highest color intensity.Quantitatively, the atomic percentages of C, Co, and O are 15.17%, 77.91%, and 6.92%, respectively (Figure S3, Supporting Information).This result indicates that the Co layer almost covered the LIG matrix with slight oxidization.X-ray diffraction (XRD) spectrum (Figure S4, Supporting Information) displays three main diffraction peaks at 41.82, 47.64, and 76.08°, well matched with (220), (111), and (311) planes of face-centered cubic (fcc) Co. [47] After electroplating, the sensing performances of the LIG-Co electrode were characterized.A tree-electrode system was constructed using LIG-Co, platinum (Pt), and saturated calomel as working electrode (WE), counter electrode (CE), and reference electrode (RE), respectively.The LIG-Co electrode was immersed in solutions with varying phosphate concentrations and its potentials relative to the reference electrode were recorded.As shown in Figure 3d, in the solution with 0 mol L −1 phosphate, the LIG-Co sensor initially showed a potential of −0.508 V and then gradually decreased.After ≈61 s, the potential finally stabilized at −0.517 V.The negative potentials of the LIG-Co electrode are because Co would be spontaneously oxidized into CoO. [48]After phosphate was added to the solution, the potential noticeably decreased and moved in the negative direction.For example, in the solution with 10 −5 mol L −1 phosphate, the LIG-Co sensor initially showed a potential of −0.52 V and then stabilized at −0.525 V after ≈60 s.In the solutions with higher phosphate concentrations (10 −4 -10 −1 mol L −1 ), the potentials displayed a similar variation trend.
For quantitative analysis, the stabilization time and the stabilization potential were defined as response time and characteristic potential, [49] which were plotted in Figure 3e,f, respectively.Except for 10 −1 mol L −1 , in which the response time was ≈15 s, the response time for other concentrations was maintained at ≈60 s.The potential monotonically decreased from −0.525 to −0.574 V, as the concentration increased from 10 −5 to 10 −1 mol L −1 .Using the log of phosphate concentration (lg c) as the x-coordinate and the potential as the y-coordinate, their relationship can be fitted with a linear function: The absolute value of the slope, 0.012 voltage per decade (V dec −1 ), and coefficient of determination (R 2 ), 0.973, are defined as sensitivity and linearity, the key performances for electrochemical phosphate sensors. [50]The phosphate concentration of an unknown solution can be determined by using the electrochemical sensor to measure the characteristic potential and substituting it to Equation 1.Despite its practicability, the detection mechanism of Co-based phosphate sensors is subject to some debates, including mixed potential mechanism, [51] host-guest mechanism, [22] and Nernst potential mechanism. [4]o demonstrate the advantage of depositing Co onto LIG matrix, the sensing performances of a pure Co rod were also characterized (Figure S5, Supporting Information).Generally, the potential response of the Co rod to phosphate was analogous to that of the LIG-Co.However, the response time of the Co rod was much longer, i.e., >125 s for all concentrations in the range of 10 −5 -10 −1 mol L −1 .What is worse, the characteristic potential relatively irregularly changed with the concentration.The linear fitting between the potential and lg c is: Although the nominal sensitivity (0.025 V dec −1 ) of the Co rod is higher than that of the LIG-Co electrode, the low linearity (0.888) undermines its meaning.Therefore, electroplating Co onto LIG matrix has improved the sensing performances to a certain extent.However, compared to other electrochemical phosphate sensors, the sensing performances of the LIG-Co sensor are still relatively low (Table S1, Supporting Information).

NAE of Co to Form LIG-Co NP Sensor
As mentioned above, although electroplating Co onto LIG demonstrates certain enhancements, the sensing performances are still relatively low.This result is against the initial hypothesis that the hierarchically porous structure of LIG would provide much more active sites for electrochemical detection and thus improve the sensing performances.It could be ascribed to the structure of LIG-Co.As shown in Figure 3b, first, Co was electroplated as a dense bulky layer whose specific surface area should be low; secondly, the pores of LIG were blocked by the dense Co layer.To solve these problems, we developed the "nearly all etching (NAE)" technique to transform the dense Co layer to dispersive Co NPs and expose the pores of LIG, both increasing the active sites.
An accidental discovery prompted the invention of the method.For Co-based phosphate sensors, they are typically soaked into KHP solutions for tens of minutes to establish a steady-state pH and initial potential. [51]When testing LIG-Co, it was also soaked in KHP solution for 20 min. [22]However, on one occasion, it was soaked for a few days.We found that the Co layer on LIG partially disappeared, but the performances were surprisingly enhanced.To systematically investigate this positive phenomenon, LIG-Co electrodes were soaked in KHP solution for varying durations.As shown in Figures 4a and S6 (Supporting Information), little vis-ible morphological change could be observed on Day 1 and Day 2. However, on Day 3, many spotted vacancies appeared on the Co layer.On Day 4 (Figure 4a), the Co layer on the central area nearly all disappeared.The Co layer on the margins still existed because of the thicker Co.On day 5, the entire Co layer disappeared.The disappearance of Co could be ascribed to the slow electrochemical corrosion. [22]he samples soaked for 4 days were found to the optimal sensing performances (narrated later), so we focused on the material characterizations of these samples.As shown in Figure 4b, the morphology of the central area of Day-4 samples significantly changed compared to the original LIG-Co.The dense Co layer disappeared and the porous structure of LIG was exposed again.Elemental maps display that C, Co, and O were still uniformly distributed but the color intensity of C was remarkably improved compared to the original LIG-Co (Figure 3c), indicating the increase in content.Specifically, the atomic percentage of C increased from 15.17% to 92.80% (Figure S7, Supporting Information).In contrast, the atomic percentage of Co decreased from 77.91% to 1.22%, called "nearly all etched (NAE)".The content of O was maintained at a close level, 5.98% compared to 6.92%.High-magnification SEM image illustrates that the hierarchically porous structure of LIG was re-exposed after NAE (Figure 4c).Moreover, the pore walls of LIG are no longer smooth (Figure 2c) but distributed with numerous particles, which should be Co.The sizes of Co particles are in nanoscale, ranging from ≈87 to 173 nm as shown in Figure 4d.These results confirm that the developed NAE technique has transformed the dense Co layer to dispersive Co NPs and re-exposed the hierarchically porous structure of LIG.

Sensing Performances of LIG-Co NP Sensor
The aforementioned material characterizations have approved that NAE produces a new structure sensor with Co NPs distributed on the hierarchically porous LIG.This structure is expected to offer abundant active sites for the electrochemical detection of phosphate, thus improving the sensing performances.To verify this hypothesis, we immersed the LIG-Co NP sensor into solutions with varying phosphate concentrations and recorded its potential relative to the reference electrode.As shown in Figure 5a, the potential in all solutions stabilized very fast compared to LIG-Co without NAE (Figure 3d).Specifically, the response time for 10 −5 -10 −1 mol L −1 was all <≈10 s (Figure 5b).The potential in the solution without phosphate was −0.258 V, much higher than the LIG-Co without NAE (−0.517V).This feature would significantly enlarge the variation range of the potential when varying concentrations of phosphate were added, which has great potential of improving sensitivity.After 10 −5 mol L −1 phosphate was added, the potential of the LIG-Co NP sensor slightly decreased to −0.267 V.In solutions with higher concentrations, the potential gradually decreased and finally reached −0.521 V at 10 −1 mol L −1 .The relationship between the potential and log of phosphate concentration was plotted in Figure 5c, which can be fitted with a linear equation: Therefore, besides the response time being shortened from ≈60 to ≈10 s, the sensitivity and the linearity were enhanced from 0.012 to 0.069 V dec −1 and from 0.973 to 0.998 after NAE treatment.Owing to the short response time and the high sensitivity, the LIG-Co NP sensor can distinguish the phosphate concentration variation in real-time (Figure 5d).These performances are outstanding compared to literature (Table S1, Supporting Information).The above results successfully verify the original hypothesis that NAE can convert the dense Co layer to dispersive Co NPs and re-expose the hierarchical porous structure of LIG were exposed, thus offering abundant active sites for phosphate detection.One possible problem of this verification is that there was a dense Co layer remained in the margins after 4-day NAE (Figure 4a), which could affect the sensing performances.To solve this problem, we trimmed the margins and only characterized the central area.As shown in Figure S8 (Supporting Information) the sensing performances of the central area are almost identical to the entire LIG-Co NP sensor.The response time is even shorter (≈8 s), the sensitivity is slightly lower (0.052 V dec −1 ), and the linearity is the same (0.998).
We also investigate the influence of the NAE duration on the sensing performances.The response curves and characteristic curves of LIG-Co with NAE durations of 1, 2, and 3, 5 days are shown in Figures S9 and S10 (Supporting Information).After linear fitting, their sensitivities and linearities were plotted in Figure 5e,f combined with the Day 4 data.As the NAE duration increases, the sensitivity first increases from 0.010 V dec −1 on Day 1 to 0.069 V dec −1 on Day 4 and then decreases to 0.001 V dec −1 (regarded as no responsivity) on Day 5. A similar variation trend can be observed for the linearity.It first increases from 0.643 on Day 1 to 0.998 on Day 4 and then decreases to 0.038 on Day 5. Thus, the optimal NAE duration is 4 days.This phenomenon can be explained by the structure evolution.In the first few days of NAE, the dense Co layer was gradually etched.More Co NPs were generated and more pores of LIG were exposed so the sensing performances were gradually enhanced.After 5-day NAE, Co was totally etched and the sensor lost the responsivity to phosphate.On Day 4, the Co NPs and the pores of LIG reached the optimal levels and provided the best sensing performances.Reducing electroplating time may also help to maintain the porous structure but the linearity (R 2 ) of the sensing characteristic curve degraded to 0.721 (for 60 s) and 0.865 (for 120 s), respectively (Figure S11, Supporting Information), mainly attributed to the non-uniform distribution of Co layer (Figure S2, Supporting Information).
In addition to the sensing performances, the anti-interference capability of the sensor is also crucial for practical application considering various ions would coexist with phosphate in a test environment.According to literature, [24] the anti-interference capability was characterized by separately adding common anions (Cl − , NO 3 − , SO 4 2− , and CH 3 COO − ) to phosphate solutions and recording corresponding potentials.The potential difference between the solutions with and without interfering ions (ΔP/Pp) is considered as the detection interference.The concentration of phosphate was set as 10 −4 mol L −1 while the concentrations of interfering ions were set as 10 −2 mol L −1 to amplify the possible interference.As shown in Figure 6a, the interferences induced by inorganic anions (Cl − , NO 3 − , SO 4 2− ) are negligible (<1.01%), and the interference induced by organic anion (CH 3 COO − ) is slight (4.51%).Overall, the LIG-Co NP sensor demonstrates an excellent anti-interference capability against these ions.
The above anti-interference experiment was performed in standard experimental conditions.Nevertheless, in practical applications, the types and concentrations of interfering components are unknown, such as ions, dirt, and microorganisms.In order to further characterize the anti-interference capability of the sensor in an unknown condition, the deionized water was replaced with water from a natural pond (on the Pukou campus of Nanjing Agricultural University) to prepare test solutions.
According to the volume of the pond water, KH 2 PO 4 with certain masses was weighted and dissolved in the pond water to obtain phosphate concentrations of 10 −5 , 10 −4 , 10 −3 , 10 −2 , and 10 −1 mol L −1 .The LIG-Co NP sensor was immersed into the solutions and recorded the potentials.Then the recorded potentials were substituted into Equation 3 to calculate the concentrations (found concentrations).As shown in Figure 6b, the found concentration and the added concentration are well-fitted with the y = x equation, indicating the high anti-interference capability of the LIG-Co NP sensor in unknown conditions.However, the cyclic sensing performance of the LIG-Co NP sensor is poor (Figure S12, Supporting Information), which is consistent with the Co-based phosphate sensors. [12,23,25]Thus, it should be used as a disposal sensor.

Conclusion
In summary, we have demonstrated a high-performance phosphate potentiometric sensor based on Co NPs electroplated on LIG matrix.The hierarchically porous structure of LIG has provided extra active sites for electrochemical detection.The developed NAE technique could transform the dense electroplated Co layer to dispersive Co NPs and re-expose the pores of LIG.Compared to the pure Co rod and the LIG-Co, the detection performances of the LIG-Co NP sensors have been significantly enhanced, including the short response time of ≈10 s, high sensitivity of 0.069 V dec −1 , large linearity of 0.998, and strong antiinterference capability.The fabricated sensor would promote the in-field measurement of phosphate and the proposed strategy would shed light on the development of high-performance electrochemical sensors.

Experimental Section
Synthesis of LIG: PI films (thickness 100 μm, Shenzhen Lexin Plastic Co., LTD.) were irradiated by a CO 2 laser engraving machine (wavelength 10.6 μm, beam size 200 μm, maximum power 50 W, maximum scanning speed 500 mm s −1 , KB-4060) to synthesize LIG.AutoCAD software was used to design the shape of LIG.Laser parameters were set in CorelLASER software.

Electroplating of Co-Layer:
The plating solution composed of CoSO 4 . 7H 2 O (0.445 mol L −1 , 99%, Aladdin), NaCl (0.291 mol L −1 , 99.5%, Xilong Scientific), and H 3 BO 3 (0.485 mol L −1 , 99.9%, MREDA) was first prepared.The LIG matrix and Co plate were connected to a power supply (Maisheng ms-3010D) as the anode and the cathode.A constant current (0.36 A) was applied to drive the electroplating and a magnetic mixing (300 r min −1 ) was performed to remove the bubbles.The electroplating continued for 180 s.
Materials Characterizations: SEM and EDS were performed using a Hitachi Regulus 8100 scanning electron microscope.Raman spectroscopy was performed by HORIBA Scientific LabRAM HR Evolution with a 532 nm laser.The sheet resistance measurement was achieved by the HPS2524 precision square resistance measuring instrument.XRD spectra were obtained with a Japanese Rigaku Smartlab 9KW X-ray diffractometer.EIS was measured in a solution containing 0.5 mol L −1 KCl (99%, Meilunbio) and 1 mmol/L [Fe(CN) 6 ] 3− (Tianjin Xiensi Biochemical Technology Co., Ltd) using CHI660e electrochemical workstation (Shanghai Chenhua Co., LTD.) following the traditional three-electrode configuration.The scanning frequencies varied from 0.1 kHz to 1000 kHz and amplitude was set as 5 mV.

Figure 1 .
Figure 1.Schematic diagram of preparation and application of the LIG-Co NP sensor.a) PI film.b) LIG fabricated by DLW on the PI film.c) Dense Co layer formed by electroplating on the LIG substrate.d) Co NPs generated by NAE, in which the dense Co layer was immersed in the KHP solution for a 4-day electrochemical corrosion.

Figure 2 .
Figure 2. Material characterizations of LIG.a) Optical photo.b) SEM image.c) high-magnification SEM image, and d) Raman spectrum of LIG.e) Influence of laser power on LIG sheet resistance at a constant laser scanning speed of 30 mm s −1 .f) EIS and equivalent circuit diagram of LIG.

Figure 3 .
Figure 3. Material and performance characterizations of LIG-Co sensor.a) Optical photo.b) SEM image, and c) high-magnification SEM image and elemental maps of LIG-Co (electroplating time = 180 s).d) Representative potential response curves of LIG-Co in solutions with phosphate concentrations varying from 0 to 10 −1 mol L −1 .e) Response time in relation to phosphate concentration.f) Relationship between the potential and the phosphate concentration.

Figure 4 .
Figure 4. Material characterizations of LIG-Co NP sensor.a) Optical photo.b) SEM image and elemental maps, and c), d) high-magnification SEM images of LIG-Co NP sensor.

Figure 5 .
Figure 5. Sensing performance characterizations of LIG-Co NP sensor.a) Representative potential response curves of the LIG-Co NP sensor in solutions with phosphate concentrations varying from 0 to 10 −1 mol L −1 .b) Response time in relation to phosphate concentration.c) Relationship between the potential and the phosphate concentration.d) Real-time response of the potential to the phosphate concentration variation.e) The relationship between the sensitivity and the NAE duration.f) The relationship between the linearity and the NAE duration.

Figure 6 .
Figure 6.Anti-interference capability of LIG-Co NP sensor.a) Potential difference between the solutions with and without interfering ions.b) The relationship between the found concentration and the added concentration of phosphate in solutions prepared by pond water with unknown interfering components.