Simple Portable Voltammetric Sensor Using Anodized Screen-Printed Graphene Electrode for the Quantitative Analysis of p-Hydroxybenzoic Acid in Cosmetics.

Screen-printed graphene electrodes (SPGEs) have become a potential option in electrochemical applications because of their outstanding properties and disposable approach to miniaturize the electrodes for onsite analysis. Herein, the detection of para-hydroxybenzoic acid (PHBA) in cosmetics using the anodized SPGE has been pioneered and reported. The simple anodization of the SPGE surface was operated by anodic pretreatment at a constant potential on SPGE. The surface morphologies and electrochemical behaviors of anodized SPGEs in different anodization electrolytes were examined. Using anodized SPGE in a phosphate-buffered solution, a nontoxic solution, the sensitivity of PHBA detection was significantly improved compared with pristine SPGE owing to the increase of the polar oxygen-containing functional group during the anodization. The anodized SPGE could detect a PHBA down to 0.073 μmol/L. Finally, the developed anodized SPGE presented high ability and feasibility for PHBA detection in cosmetics. Furthermore, a facile electrode preparation step with a nontoxic solution can present high reproducibility and compatibility with a portable potentiostat for onsite PHBA detection during manufacturing.

Introduction

Cosmetics and personal care products have become an important factor on a daily basis to enhance the appearance and confidence of women and men. The ingredients of cosmetics include water and organic and inorganic compounds that are necessary to prevent microbial contamination to ensure customers’ safety and increase their shelf-life.[1] Various compounds have been used as preservative substances, including aldehydes, glycol ether, isothiazolinones, organic acids, and parabens. 4-Hydroxybenzoic acid, known as para-hydroxybenzoic acid (PHBA), is a frequently used preservative substance in cosmetics. In the PHBA structure, the benzene ring of benzoic acid is substituted by a hydroxy group at C-4. Moreover, PHBA is the hydrolysis product of parabens, which is a popular preservative in cosmetics and foodstuff.[2,3] Although the toxicity of PHBA is not found to be acutely toxic, PHBA has an estrogenic activity that stimulates the growth of human breast cancer, and it is also considered the biomarker of gastric cancer.[2,4−6] At present, natural and organic cosmetic markets are growing. The customers want to know the source of ingredients or material of the products. In meeting the trend and demand of users, the production of cosmetics has been using natural, organic, and hypoallergenic substances; meanwhile, preservative no longer relies on the demands. Therefore, rapid and sensitive detection methods of PHBA are necessary to meet the recent global trend. Many analytical methods have been developed and applied for the investigation of PHBA in many cosmetics. The separation of PHBA using C18 column combined with spectrometry such as UV–vis,[7] photodiode-array,[8] and mass spectrometry[9] is the major technique for the detection of PHBA, which provides good sensitivity and selectivity. In addition, the quantification of PHBA using the fluorescence technique has been considered a potential analytical technique. Cyclodextrin inclusion allows trace analysis of PHBA.[10] However, these techniques require a bulk instrument and benchtop operation, which are not practical for onsite analysis.

To approach the onsite applications, electrochemical techniques have been developed to overcome the limitation of other techniques. A few researchers have proposed the electrochemical determination of PHBA using NiTiO3/CPE and carbon-disk electrodes, which are not portable and require complicated modification steps.[2,11,12] Screen-printed electrodes have been widely proposed in several developed electrochemical sensors because of their disposability, low cost, and easy fabrication, which enables mass production and can be integrated with a portable potentiostat.[13,14] Graphene screen-printed electrodes (SPGEs) have received considerable attention over carbon screen-printed electrodes (SPCEs) because of their potential electrocatalytic activity of graphene material resulting in high sensitivity.[15] Furthermore, using screen-printing ink for electrode fabrication can improve the stability and electrochemical performance of the sensor in practical use.[16] Therefore, proposing an electrochemical sensor using SPGEs can deliver the developed sensor to manipulate the assay onsite. Although the performance of SPGE is highly efficient, many researchers still further enhance the sensitivity of SPGE for trace analysis of various substances. Several methods for the modification of graphene have been proposed for a particular analyte such as drop-casting of the modifier on the electrode surface,[17] integration by mixing the modifier and ink before being screened on the substrate,[18−20] and electrochemical modification of electrocatalytic species on the surface.[21,22] Most of the modification methods require the addition of a reagent and extra step, which are nonenvironmentally friendly and complicated steps. Therefore, a simple modification method with an environmentally friendly reagent has received considerable interest.

To improve the sensitivity of PHBA detection, the surface of SPGE has been modified using the anodic pretreatment at a constant potential in a phosphate-buffered solution, a nontoxic solution. During the anodic pretreatment, the surface of SPGE was oxidized to be an oxidized film, which improved the durability and corrosion resistance of surfaces.[23,24] Typical oxidizing procedures are operated in aggressive organic or inorganic solutions such as sulfuric acid and sodium hydroxide. Many researchers have conducted experiments by replacing the use of aggressive and toxic solutions with a less harmful chemical in electrolytic oxidation to achieve “green chemistry.”[25] Moreover, applying anodization with SPGEs can remove any waste substances hindering the electrochemical kinetics of the electrode surface and redox species, which can improve the electrochemical performance.[26] Therefore, this work has applied anodic pretreatment as the modification method of SPGE to enhance the sensitivity of PHBA detection. The electrochemical cell has been miniaturized and designed to be a portable sensor, and electrochemical detection has been achieved using a portable potentiostat. To the best of our knowledge, no report has been found on the use of an anodized graphene material for the detection of PHBA. A nontoxic phosphate-buffered solution was used as an anodization solution; thus, the process can achieve green chemistry. The PHBA sensor demonstrated in this report presented sensitivity, portability, and environmental friendliness for PHBA detection in cosmetics.

Experimental Section

Chemical, Reagents, and Instruments

All chemicals applied in this work were of analytical grade and used as received. All of the solutions were prepared in DI water with a resistivity of 18.2 MΩ·cm. 4-Hydroxybenzoic acid (PHBA) was purchased from Sigma-Aldrich (St. Louis, MO). Sodium acetate and sodium hydroxide (NaOH) were traded from Tokyo Chemical Industry Co., Ltd. (Tokyo, Japan). Acetic acid and sulfuric acid (H2SO4) and potassium ferri/ferrocyanide [K3Fe(CN)6]/[K4Fe(CN)6] and potassium nitrate (KNO3) were purchased from Carlo Erba reagents (Lombardia, Italy) and Merck (Darmstadt, Germany), respectively.

The electrochemical measurements were operated using Sensit BT, a portable potentiostat (Palmsens, Houten, Netherlands), which was controlled via a tablet app, namely, “PStouch” (Palmsens, Houten, Netherlands), and the data were transmitted through Bluetooth communications. The three-electrode system was designed by Adobe Illustrator software (Adobe Systems, Inc.) consisting of working, counter, and reference electrodes. Silver/silver chloride paste (Ag/AgCl: 60/40) was purchased from Gwent Electronic Materials Ltd. (Bath, U.K.), which served as a reference electrode. Commercial carbon paste and graphene paste were purchased from Serve Science Co., Ltd. (Bangkok, Thailand), which served as the working and counter electrodes for screen-printed carbon electrode (SPCE) and screen-printed graphene electrode (SPGE), respectively. The homemade SPGE was fabricated by screen printing via a template block from Chaiyaboon Co., Ltd. (Bangkok, Thailand). In studying the surface morphologies of the electrodes, field emission scanning electron microscopy (FESEM) and energy-dispersive X-ray spectrometry (EDS) were operated using a MIRA3 model (TESCAN ORSAY HOLDING, Kohoutovice, Czech Republic).

Fabrication of SPGE

The single piece with a three-electrode system on the transparent sheet (polypropylene [PP]) substrate was designed using Adobe Illustrator software with a working electrode of 3.0 mm diameter. The PP sheet was cleaned with ethanol before use. For the fabrication of SPGE, the commercial graphene ink was first screened on the PP substrate for conducting pad and working and counter electrodes. Afterward, the Ag/AgCl ink was used as a reference electrode and painted on the prepared electrode sheet. Finally, the PVC ink was screened to limit the area of spreading the solution. In each step of screening and painting the layer, the electrode was allowed to remove the solvent and dried in an oven at 55.0 °C for 1 h. The finished SPGE was further modified in the next step. For the SPCE, commercial carbon ink was used instead of graphene ink.

Anodization of SPGE

One hundred microliters of 1.0 mol/L phosphate-buffered pH 6.5 solution was used as an anodization solution and introduced covering the SPGE surface. The constant potential of +2.0 V was applied on the surface of SPGE for 120 s. Afterward, the anodized SPGE was carefully rinsed with DI water and air-dried before being used. The type of electrolyte, pH, anodization potential, and anodization time were studied to obtain the optimal condition for anodization.

Electrochemical Detection of PHBA

The detection of PHBA was performed using a portable potentiostat (Sensit BT). The serial dilution of PHBA standard solution was prepared in 0.1 mol/L acetate buffer pH 4.5. For electrochemical measurement, 100 μL of PHBA standard solution was dropped down covering three electrodes. Differential pulse voltammetry (DPV) was performed with a potential range of 0.0 to +1.0 V, a step potential of 50 mV, a pulse amplitude of 175 mV, a pulse time of 50 ms, and a time interval of 700 ms. In addition, the electrochemical characterization of the developed sensor was operated using cyclic voltammetry (CV) and electrochemical impedance spectroscopy (EIS).

Interference Studies

The oxidation behavior of possible interferences found in cosmetics was studied, including K+, Na+, NH4+, Ca2+, Mg2+, Cl–, I–, NO3–, CO32–, benzoic acid, gallic acid, dopamine, uric acid, and ascorbic acid. The tolerance limit of studied interferences was investigated in a solution containing 20.0 μmol/L of PHBA. The interference that yielded a relative error equal to or less than 5% compared with the response of 20.0 μmol/L of standard PHBA was defined as not interfering.

Sample Analysis

Skin lotions were selected to examine the quantity of PHBA and estimate the ability of the proposed method to be applied to real samples. Five skin lotion samples were purchased from a local drug store in Thailand, and 5.0 mL of each skin lotion was diluted with 0.1 mol/L acetate buffer in a 25.0 mL volumetric flask, and then the standard PHBA solution was spiked. Afterward, 0.5 mL of this solution was diluted in a 25.0 mL volumetric flask, and the final concentration of PHBA in the spiked solution was calculated within the linear dynamic range. In determining the concentration of PHBA in the sample, 100.0 μL of the tested solution was introduced on the anodized SPGE, and the signal was measured using DPV. The linear equation obtained from the calibration curve was used to calculate the PHBA contained in the test solution.

Results and Discussion

Effect of Anodization on the Electrochemical Behavior of PHBA

The performance of the anodized electrode surface to the electrochemical oxidation of PHBA was studied on pristine SPCE, pristine SPGE, anodized SPCE, and anodized SPGE using CV. The anodization of SPCE and SPGE was first performed in 0.1 mol/L NaOH by applying a potential of 2.0 V vs Ag/AgCl for 120 s. Figure a shows a cyclic voltammogram of 20.0 μmol/L PHBA in 0.1 mol/L acetate buffer pH 4.5 on different electrodes. The result showed the occurrence of an irreversible oxidation peak of PHBA at a potential of 0.65 V. The absence of a cathodic peak indicated that the oxidation of PHBA was followed by the chemical reaction, which rapidly removed the generated product. However, the reversible peak at a low potential was observed on the anodized surface. The previous related research suggested that in the initial stage of the oxidation process, the cation-radical intermediate was generated.[27] When the pristine SPCE and SPGE were used, they have insufficient sensitivity to present the peak of this intermediate. However, when the anodization was applied, the polar oxygen-containing functional group and surface edge-plane sites were increased, which can act as a catalyst for the rate of electron transfer attributed to the presence of an intermediate peak at the lower potential and the enhanced sensitivity of the PHBA oxidation at the anodized surface.[28] Considering peak currents, no significant differences were observed when pristine SPCE (Pristine SPCE: PHBA) and pristine SPGE (Pristine SPGE: PHBA) were used. However, when the surfaces were anodized in 0.1 mol/L NaOH, the anodic peak currents of PHBA on anodized SPCE (Anodized SPCE: PHBA) and anodized SPGE (Anodized SPGE: PHBA) were significantly increased. Comparing the effect of anodization on the electrochemical behavior of PHBA in a particular electrode, DPV has been used to monitor the anodic peak current of PHBA. As shown in Figure b, after anodization, the anodic peak current was significantly increased on both anodized electrodes. However, in anodized SPGE, a 2-fold higher anodic peak current of PHBA was obtained. This finding could be attributed to the increase of the polar oxygen-containing functional group after anodization, and the high electroactive surface area of pristine SPGE increased the degree of anodization on the surface compared with the surface of SPCE.[27]

Figure 1

Electrochemical behavior of PHBA on nonanodized and anodized electrodes: (a) cyclic voltammogram and (b) differential pulse voltammogram of 20.0 μmol/L PHBA in 0.1 mol/L acetate buffer pH 4.5.

Before achieving the aim of this work and obtaining the highest performance for PHBA detection, anodization factors affecting the electrochemical responses of PHBA were investigated, including the type of anodization electrolyte, pH of anodization electrolyte, anodization potential, and time.[29,30] In this section, the reported current responses were obtained from the signal of the test subtracted by the background signal (signal-to-noise ratio [S–N]).

Optimization of Anodization Conditions

Type of Anodization Electrolytes

Considering that the anodic oxide film properties are affected by the type of electrolytes, 0.5 mol/L H2SO4, 1.0 mol/L PB and 1.0 mol/L NaOH were applied to oxidize the electrode surface. As shown in Figure a, among all applied electrolytes, a CV peak current response of PHBA was clearly observed in an electrode anodized by 1.0 mol/L PB solution. Similar to DPV results, well-defined peaks were obtained in an electrode anodized by 1.0 mol/L PB at a potential of 0.25 and 0.65 V vs Ag/AgCl. By contrast, in an electrode anodized by 1.0 mol/L NaOH, a small peak current response was observed at a potential of 0.65 V vs Ag/AgCl. In addition, no peak was observed under the 0.5 mol/L H2SO4 condition (Figure b). We hypothesize that anodization under a vigorous condition such as strong acid or strong basic solution not only anodized the surface but also eroded the surface during anodization. When the PB solution was used, the redox process is assisted by phosphate, resulting in a high number of oxygenated groups, which affects the electron mobility without eroding the surface.[29] Therefore, the electrode will be anodized by PB solution for the entire experiment. The morphologies of pristine and anodized surfaces were examined using different electrolytes to understand the effect of electrolytes.

Figure 2

Effect of the type of electrolytes represented by (a) cyclic voltammogram and (b) differential pulse voltammogram of each electrolyte.

SEM and EDS

A scanning electron microscope was used to characterize and compare the electrode surface morphology before and after anodization (Figure ). Figure a,b shows the rough surface of pristine SPCE and a flakelike surface of pristine SPGE, respectively. After pretreatment in NaOH, large pores were observed in anodized SPCE (Figure c) and SPGE (Figure d), which could support the hypothesis of the erosive phenomena in a strongly basic solution. However, among anodized SPGEs, the surface of anodized SPGE by PB solution presented more fractions and surface roughness because more edge regions have been created, resulting in a high electron transfer rate, which is inconsistent with the results shown in Figure .

Figure 3

SEM images of (a) pristine SPCE, (b) pristine SPGE, (c) anodized SPCE in NaOH, (d) anodized SPGE in NaOH, (e) anodized SPGE in H2SO4, and (f) anodized SPGE in PB.

EDS analysis was then further performed to investigate and compare the elemental component between pristine SPGE and anodized SPGE. After anodization with 1.0 mol/L NaOH, the percentage of oxygen was slightly increased from 4.78% for pristine carbon to 5.5% for anodized SPGE. The percentage of oxygen increased to 11.98 and 9.24% after oxidizing SPGE with 0.5 mol/L H2SO4 and 1.0 mol/L PB, respectively. This phenomenon indicated that anodization promoted the increase of polar oxygen-containing functional groups as previously mentioned. Hence, based on the results of previous studies, PB solution was selected as the electrolyte for the anodization of SPGE.

pH of Anodization Solution

The changes in pH influence the stability of the oxide film.[31] Therefore, the pH of 1.0 mol/L PB was also varied from 5.5 to 8.0 to achieve a high performance of the sensor. As shown in Figure a, the current response increased with the increase of pH until pH 6.5. However, at a pH higher than 6.5, the current response was negligibly decreased. Thus, 1.0 mol/L PB pH 6.5 will be used to perform anodization in further experiments.

Figure 4

Parameters affecting electrochemical anodization: (a) pH of anodization electrolyte, (b) anodization potential, and (c) anodization time on current responses.

Anodization Potential

The anodic potential also affects the current response. In this experiment, the range of potentials were measured from +1.8 to +2.2 V using 1.0 mol/L PB pH 6.5 as an electrolyte solution. The results shown in Figure b revealed that at a +2.0 V potential, the highest current response was obtained. However, at more positive potentials, the current responses evidently decreased because of the high background signal (data not shown). In the next experiment, 1.0 mol/L PB pH 6.5 solution will be applied to oxidize the electrode at +2.0 V potential vs Ag/AgCl.

Anodization Time

After obtaining the optimized potential, the anodization time was also investigated to complete the oxidation of the graphene surface. Figure c shows that the current responses gradually increased with longer applied anodization time until 120 s. Afterward, the current responses decreased. Therefore, 120 s anodization time is suitable to oxidize the electrode.

In addition, 1.0 mol/L PB pH 6.5 will be used as an electrochemical anodization solution at a potential of +2.0 V vs Ag/AgCl for 120 s in the following experiments.

Characterization of Anodized SPGE

CV and EIS were performed using 5.0 mmol/L [Fe(CN)63–]/[Fe(CN)64–] in 0.1 mol/L KNO3 to characterize the electron transfer properties in particular electrodes. The results shown in Figure S1 describe that the redox current of all electrodes was a reversible electrochemical reaction. This behavior can be described using the Randles–Sevcik equation (eq ), which accounted for the electroactive surface area calculation.where n is the number of electrons; A is the electroactive surface area (cm2); D0 is the diffusion coefficient (cm2/s); ν is the scan rate (V/s); and C0* is the concentration (mol/cm3). The electroactive surface areas of studied electrodes were calculated on the basis of the known parameters, including n = 1, D0 = 7.6 × 10–6 cm2/s, C0* = 5 × 10–6 mol/cm3, and Ip ν–1/2 is the slope of the plot between the square root of scan rate ((V/s)1/2) and peak current (A), and found to be 7.032 mm2 for SPCE, 7.680 mm2 for SPGE, 8.256 mm2 for anodized SPCE, and 10.05 mm2 for anodized SPGE. Among these electrodes, the anodized SPGE exhibited the highest electroactive surface area. As shown in Figure a, cyclic voltammograms of the studied electrodes were compared, and the results showed that the anodized SPGE presented the fastest charge transfer because of the smallest peak separation and the highest current response, which is consistent with the calculated electroactive surface area. These results were in agreement with the characterization by EIS. The Nyquist plots indicated a semicircle and a straight line with a 45.0° angle, which represented the charge transfer resistance (Rct) and diffusion-controlled redox process, respectively. A high Rct indicated poor electron transfer in the system. As shown in Figure b, Rct obtained in each electrode followed the order pristine SPCE > pristine SPGE > anodized SPCE > anodized SPGE. These results indicated that anodized SPGE provided more excellent electron transfer than others because of the generation of edge planes in anodized SPGE. Consequently, lateral charge transfer occurs and electrochemical activity is enhanced.[32,33]

Figure 5

(a) Cyclic voltammogram at a scan rate of 100 mV/s and (b) EIS spectra of pristine SPCE, pristine SPGE, anodized SPCE, and anodized SPGE in 5.0 mmol/L [Fe(CN)63–]/[Fe(CN)64–] in 0.1 mol/L KNO3.

Electrochemical Mechanism of PHBA on Anodized SPGE

After confirming that the anodized SPGE electrode provided excellent electrochemical performances, the possible electrochemical mechanism of anodized SPGE toward the oxidation of PHBA was investigated. CV was performed using 20.0 μmol/L of PHBA in 0.1 mol/L acetate buffer pH 4.6 in different scan rates. As shown in Figure , the irreversible oxidation peak current at a potential of approximately 0.70 V was observed (Figure a), and a linear relationship with the square root of scan rate (Figure b) was observed in a scan rate ranging from 20 to 40 mV/s, which is the characteristic of the diffusion-controlled current. Therefore, the oxidation of PHBA on anodized SPGE is controlled by diffusion.

Figure 6

(a) Cyclic voltammograms in various scan rates from 20 to 60 mV/s and (b) linear relationship with the square root of scan rate ((mV/s)1/2) performed by an anodized SPGE using 20.0 μmol/L of PHBA in 0.1 mol/L acetate buffer pH 4.5.

Optimization of Conditions for Quantitative Analysis of PHBA

For the quantitative analysis of PHBA, DPV was performed. DPV parameters affecting electrochemical signals were investigated to obtain the best performance of the PHBA-developed sensor. The results of the following experiments will be reported as S–N.

pH of Supporting Electrolyte

The pH of acetate buffer solution, a supporting electrolyte of PHBA detection, was studied because an anodized SPGE electrode presented rich ionizable carboxyl and hydroxyl groups, which can be protonated/deprotonated by changing the pH.[34] An acetate buffer solution was adjusted to different pH ranges from 3.5 to 5.5 (Figure a). The anodic current of PHBA increased with the increase of pH until reaching 4.5. Beyond pH 4.5, the signal drastically decreased. Therefore, pH 4.5 was appropriate for PHBA detection.

Figure 7

Investigation of parameters affecting quantitative analysis of PHBA, including (a) pH of 0.1 mol/L acetate buffer, (b) step potential (mV), (c) pulse amplitude (mV), (d) pulse time (ms), and (e) interval time (ms) tested by DPV using 20.0 μmol/L of PHBA.

Step Potential

The step potential of DPV was studied at 10, 25, 50, 75, and 100 mV. The signal increased from 10 to 50 mV and then decreased (Figure b). Therefore, 50 mV was fixed for further experiments.

Pulse Amplitude and Time

Pulse amplitude and time were subsequently investigated. The pulse amplitude was studied from 25 to 275 mV. The results in Figure c showed that the signal increased with the increase of pulse amplitude. After reaching 175 mV, the signal was steady. Therefore, 175 mV of pulse amplitude was selected for use in the next experiments. Then, after obtaining the optimal pulse amplitude, pulse time was also studied. As shown in Figure d, at 50 ms, the highest signal response was achieved. The longer the pulse time applied, the less signal response was obtained. Thus, 50 ms of pulse time was the optimal time to apply in the following experiments.

Interval Time

The interval time was also optimized by fixing at 400, 500, 600, 700, and 800 ms (Figure e). The obtained signal response increased from 400 to 700 ms. After 700 ms, the signal response decreased. Therefore, 700 ms was the suitable interval time in this experiment.

Analytical Performance

After an investigation of all influencing parameters, the analytical performance of the developed PHBA sensor was then examined. All optimized parameters were fixed, and the concentration of PHBA in 1.0 mol/L acetate buffer pH 4.5 varied in the range of 0.0–50.0 μmol/L. The DPV results are shown in Figure a. The high peak current was detected when the concentration of PHBA increased from 0.0 to 50.0 μmol/L. Figure b displays a calibration plot between concentrations of PHBA (μmol/L) and current (μA). The plot showed that a couple of linear ranges were obtained in the range of 0.3–10.0 μmol/L (Figure b, inset i) and 10.0–50.0 μmol/L (Figure b, inset ii) with a good linear correlation (R2) of 0.9983 and 0.9973, respectively. The limit of detection (LOD, 3SD/slope) for PHBA detection was found to be 0.0726 and 0.240 μmol/L for the linear response of insets i and ii, respectively. The result showed that our proposed sensor offers a lower LOD than the maximum level of application of PHBA identified by the U.S. Food and Drug Administration (FDA) (0.1% in food and 0.4% in cosmetics).[35]

Figure 8

(a) Differential pulse voltammogram of anodized SPGE in 0.1 mol/L acetate buffer pH 4.5 containing different concentrations of PHBA ranging from 0.0 to 50.0 μmol/L. (b) Calibration plots between the concentration of PHBA (μmol/L) vs anodic current (μA); inset (i): linear response of PHBA concentrations from 0.0 to 10.0 μmol/L; inset (ii): linear response of PHBA concentrations from 10 to 50 μmol/L.

A few published works have reported the detection of PHBA using an electrochemical technique (Table ). The NiTiO3-modified CPE has been proposed in two reports. Although one of these PHBA sensors exhibited a slightly lower detection limit than our proposed device,[11] tedious modification steps were necessary, resulting in inconsistency in each electrode surface. This result could affect the reproducibility of the assay. Moreover, capillary electrophoresis coupled with electrochemical detection has been proposed for PHBA detection.[12] This proposed device provided a higher LOD than our developed sensor, and it cannot be used outside the laboratory, making it inconvenient for onsite application.

Table 1

Comparison of the Analytical Performance of PHBA-Based Electrochemical Detection

techniques	electrode modification	applied method	linear range (μmol/L)	LOD (μmol/L)	references
electrochemistry	NiTiO3/CPEa	DPV	10.0–90.0 and 90.0–1000.0	0.10	(2)
electrochemistry	NiTiO3/CPE	DPV	0.7–80.0 and 80.0–1000.0	0.062	(11)
CE-EDb	carbon-disk electrode	hydrodynamic voltammetry	2–500	1.40	(12)
electrochemistry	anodized SPGE	DPV	0.3–10.0 and 10.0–50.0	0.073	this work

Carbon paste electrode modified with nickel titanate nanoceramics.

Capillary electrophoresis-electrochemical detection.

Compared with our developed work, only electrode anodization is required in the electrode preparation step for enhancing the sensitivity of PHBA detection. For end-users, a single-step sample addition should be performed. Moreover, the anodized SPGE can be applied using a portable potentiostat, which is accessible and eligible to the end-users to operate onsite.

Interferences

Several interferences affecting the selectivity of PHBA detection were also investigated as the tolerance limit (Table ). The investigated substances should not affect the oxidation current of PHBA more than ±5% after adding to the system.[11] Before the test, the concentration of PHBA in 0.1 mol/L acetate buffer pH 4.5 was fixed at 20.0 μmol/L, whereas other substances were particularly added until the oxidation peak current of PHBA was higher than ±5% compared with the original response signal. The normalized current of 20.0 μmol/L of PHBA before and after spiked the studied interferences with the concentration at the tolerance limit is shown in Figure S3. The results revealed that the ionic substances have no effects on PHBA detection, whereas uric and ascorbic acid can interfere after adding a concentration 50 times higher than PHBA. These phenomena are attributed to the different chemical structures and electrochemical behaviors between PHBA and the studied interferences. In the case of uric acid and ascorbic acid, the oxidation potentials of both species were found to be 0.30 and 0.40 V, respectively, which is closed to the oxidation of PHBA caused a lower tolerance limit than those obtained from other studied species. The results could conclude that almost studied interferences found in cosmetics did not interfere with the detection of PHBA. A proposed PHBA sensor was demonstrated as an effective device to further apply for practical use in real samples.

Table 2

Interference Studies of PHBA Detection as the Tolerance Limit

interferences	tolerance limit (Winterference/WPHBA)
K+, Na+, NH4+	500
Ca2+, Mg2+	500
Cl–, I–, NO3–, CO32–	500
benzoic acid	200
gallic acid	200
dopamine	100
uric acid	50
ascorbic acid	50

Repeatability of Anodized SPGE

The repeatability of anodized SPGE was also investigated in this work. Ten different SPGEs were anodized in PB pH 6.5 and tested with 20.0 μmol/L PHBA in 0.1 mol/L acetate buffer pH 4.5. The oxidation current of tested electrodes is presented in Figure S2 as S–N. The relative standard deviation (RSD) of these anodized SPGEs was found to be 2.5%. Therefore, the anodized SPGE provides good repeatability among electrodes. Moreover, this preparation method is easy, rapid, and convenient, and it can be conducted for mass production. Furthermore, the stability of anodized SPGE was examined after storing at room temperature, and we found that the proposed anodized SPGE indicated great stability, which could retain 95.0% of its original response after storing longer than a month after anodization, as shown in Figure S4.

Real Samples

Five samples of skin lotions were determined to verify the practical performance of the proposed anodized SPGE. The samples were diluted with 0.1 mol/L acetate buffer pH 4.5 and spiked with the standard solution of PHBA to the total final concentration of PHBA within the linear range. As shown in Table , the %RSD and recoveries were found to be 0.4–3.6 and 97.2–104.4%, respectively. These estimated values are well accepted with the defined RSD lower than 15% and recoveries from 70.0 to 110.0% established by the European Commission. Moreover, high-performance liquid chromatography (HPLC), a standard method for PHBA analysis,[36] was also performed parallelly using the same samples. The obtained results are revealed in Table , indicating that the calculated t-value was significantly lower than the critical t-value between two assays. It could be concluded that there is no significant difference between the proposed PHBA sensor and the conventional HPLC method. Thus, the results reveal that the proposed anodized SPGE has a highly effective feasibility for the determination of PHBA in cosmetics.

Table 3

Determination of PHBA in Different Samples (n = 3) by the Proposed Anodized SPGE and the Conventional HPLC Methoda

		found (μmol/L)		%recovery		%RSD
samples	added (μmol/L)	this work	HPLC	this work	HPLC	this work	HPLC
1	0.00	1.03 ± 0.04	1.00 ± 0.01	ND	ND	3.6	1.3
	2.00	3.00 ± 0.08	2.99 ± 0.05	99.5	100.3	2.8	1.6
	5.00	6.06 ± 0.11	6.01 ± 0.09	100.2	101.2	1.8	0.8
	8.00	9.20 ± 0.16	9.15 ± 0.10	101.9	102.5	1.7	1.0
2	0.00	ND	ND	ND	ND	ND	ND
	2.00	1.97 ± 0.06	2.00 ± 0.04	100.0	98.4	3.0	2.0
	5.00	4.98 ± 0.03	5.09 ± 0.01	101.8	99.6	0.7	0.4
	8.00	7.96 ± 0.06	8.10 ± 0.03	101.25	99.5	0.8	0.5
3	0.00	ND	ND	ND	ND	ND	ND
	2.00	1.84 ± 0.04	2.04 ± 0.02	102.0	91.95	2.3	1.3
	5.00	4.86 ± 0.11	5.16 ± 0.08	103.2	97.2	2.3	2.1
	8.00	7.97 ± 0.05	8.07 ± 0.03	100.9	99.8	2.9	2.0
4	0.00	ND	ND	ND	ND	ND	ND
	2.00	2.06 ± 0.04	2.02 ± 0.01	101.0	103.2	2.2	2.0
	5.00	5.12 ± 0.06	5.02 ± 0.05	100.4	102.4	1.2	0.9
	8.00	8.06 ± 0.03	8.00 ± 0.02	100.0	100.8	0.4	0.4
5	0.00	ND	ND	ND	ND	ND	ND
	2.00	2.09 ± 0.02	1.99 ± 0.02	99.5	104.4	1.3	1.1
	5.00	5.14 ± 0.06	5.04 ± 0.05	100.8	102.9	1.1	1.1
	8.00	8.09 ± 0.06	7.99 ± 0.04	99.9	101.2	0.7	0.5
paired two-tail test	t values	–0.516
	t critical	2.145

ND = Not detected.

Conclusions

The anodized SPGE was proposed in this work for PHBA detection with a simple and cost-effective modification. A PB solution was used as an anodization solution, which showed the highest current response after the PHBA solution was applied. The facile electrode preparation for PHBA detection and the utilization of PB solution as an anodization solution is environmentally friendly that could be counted as a green chemistry process. Under optimal conditions, the detection limit for PHBA was achieved at 0.073 and 0.240 μmol/L. Moreover, real samples were used in this work and the results showed the high effective feasibility for the detection of PHBA in cosmetics. The developed device exhibited great potential, reliability, and feasibility for PHBA detection with a facile electrode preparation step, which enables mass production. Furthermore, this PHBA sensor can be used together with a portable potentiostat, which is deliverable and practical for onsite analysis. However, there is some limitation that the anodization process could cause a higher background signal than the pristine surfaces and facilitate the oxidation of water causing the narrow potential window of the anodized electrode, which limited the application of substances with greater positive oxidation potential.