QCT and RUT were purchased from Extrasynthese (Lyon, France). PAP and APAP were purchased from JEULIN and used as received. 0.2 g/L QCT and RUT, then 0.5 g/L PAP and APAP were prepared as stock solutions. Phosphate buffer (PBS, 0.1 M, pH 6.0) was used as working solution. All other aqueous solutions were prepared using distilled water.

For the determination of APAP in pharmaceutical formulations, a commercial tablet of Doliprane® was carefully weighed. It was then finely powdered and dissolved in 200 mL PBS. A 15 µL aliquot of this solution was then diluted to the mark with PBS in a 10 mL volumetric flask. The sample was finally spiked with known standard amounts of both PAP and APAP, and the unknown concentration of APAP in the tablet was determined using the standard addition method on OAG-CPE. Human urine was also used as a real matrix for quantitative analyses of QCT and RUT. Thus, a sample of urine (5 mL, pH 5.8) was diluted with 20 mL of PBS and enriched successively with 3, 5 and 7 mg/L of both QCT and RUT. The mixture was stirred for two minutes using a homogenizer and the solutions were then left to rest for 5 minutes. QCT and RUT extracted from spiked urine samples were determined using SWV on OAG-CPE and recovery rates were then calculated.

The Arabic Gum (AG) was collected from Acacia senegal trees in Ouagadougou (Burkina Faso) and was pretreated as previously described [19] [20] . The procedure for the AG acylation was adapted from the methodology reported by Olatunde et al. [21] . Briefly, 50 g of the raw gum powder was dissolved into 250 mL of distilled water and stirred for 20 min. The pH of the mixture was adjusted to 8.0 using 1 M NaOH. 5.1 g of acetic anhydride was added over a period of 1 h, while maintaining a pH range of 8.0 - 8.5. The reaction was allowed to proceed for 5 min after the addition of acetic anhydride. The pH was finally adjusted to 4.5 with 0.5 M HCl. The acylated AG was then precipitated from the solution using ethanol, filtered and washed with ethanol and air-dried.

In the same way, the oxidation of AG consisted of the introduction of 50 g of raw gum powder into 500 mL flask containing 250 mL of distilled water and stirred for 20 min to ensure complete dissolution. The pH of the mixture was adjusted to 9.5 with 2 M NaOH. 10 g of sodium hypochlorite (15% available chlorine) were slowly added into the solution over a period of 30 min while maintaining the pH range at 9.0 - 9.5. The reaction proceeded for 10 min after addition of NaOCl (15% available chlorine). After a few minutes, the pH of the mixture was adjusted to 7.0 with 1 M H₂SO₄ and the oxidized AG was then precipitated from the solution using ethanol, filtered and washed with ethanol and air-dried. Acylated and oxidized gums were powdered to obtain fine fractions less than 40 µm and then used without any further modifications for the preparation of modified carbon paste electrodes.

(SVW) Experiments were performed with a DY2300 electrochemical analyzer (Digi-IVY Instruments, USA) running with the DY2300 EN software and conducted using a three-electrode system, with the modified CPEs as working electrode, a platinum wire as the counter electrode, and a saturated calomel electrode (SCE) as the reference electrode. Fourier transform infrared (FTIR) study was performed over the wavenumber range of 4000 - 400 cm⁻¹ by the ART technique with a Bruker α-P FT-IR spectrophotometer with a resolution of 4 cm⁻¹. A sequence of 200 scans was recorded for each spectrum.

The unmodified CPE was prepared according to previous work [13] . Briefly, the unmodified CPE was prepared by thoroughly hand mixing 30 mg of silicon oil with 70 mg of graphite powder (analytical grade, ultra F, <325 mesh, from Alfa) in a mortar. A portion of the composite mixture was packed into the cylindrical hole of a Teflon tube equipped with a copper wire, serving as an electrical contact with the rest of the circuit [19] . The surface exposed to the solution was polished on a weighing paper to give a smooth aspect before use. When not in use, CPEs were removed from the supporting electrolyte and stored at room temperature. Acylated and oxidized AGs modified carbon paste electrodes were prepared as described in [22] for the bare CPE using 65 mg of graphite powder, 30 mg of silicon oil and 5 mg of either acylated or oxidized gum powder.

The FT-IR spectra of raw, acetylated and oxidized AG are presented in Figure 1 .

From Figure 1 , the obtained spectra show a broad peak characteristic of polysaccharide units in the region 3500 - 3200 cm⁻¹, which indicates the presence of alcohols, phenols or carboxylic acids [23] . As seen, the peak at 3352 cm⁻¹ increases significantly when the raw AG is oxidized ( Figure 1(c) ) and decreases when raw AG is acylated ( Figure 1(b) ). This can be explained by an increase and a decrease of OH functional groups at the surface of AG after oxidation and acylation respectively. In fact, during the acylation reaction of AG, carboxyl groups are gradually converted to COOCH₃ by eliminating OH functions. While AG oxidation introduces in addition to carboxyl, alcohol and phenol functions, which make the surface of the oxidized AG with OH functions [24] . The peak at 2940 cm⁻¹, indicates the presence of asymmetric and symmetric C-H stretching of aliphatic methyl and methylene. The peaks observed at 1718 cm⁻¹ and 1604 cm⁻¹ correspond to both carbonyl (C=O) of carboxylic acids and carboxylate, respectively. Minor peaks localized between 1500 cm⁻¹ and 1244 cm⁻¹ may indicate the symmetrical bending vibration of alkane bonds (−CH₂). The presence of C-O, C-H or C-C stretching vibrations of carboxyl groups (−COOH) is indicated by absorption peak observed at 1244 cm⁻¹. The peak at 1041 cm⁻¹ is probably due to the stretching vibrations of C-O of the alcohol on the glycosyl group [18] . The obtained results from this section clearly show that new chemical functions such as alcohol, carboxylic acids or carbonyl were successfully formed at the surface of raw AG during acylation and oxidation reactions. The newly formed functions will positively interact by forming electrostatic bonds with alcohol functions of PAP, APAP and flavonoids, leading to the production of better voltammetric signals.

Cyclic voltammetry was applied in 0.1 M pH 6.0 PBS containing 0.5 mg/L for each QCT and RUT and 10 mg/L for each PAP and APAP on different elaborated electrodes and Figure 2 shows the obtained results.

Globally, from Figure 2 , the oxidation signals of QCT, RUT, PAP and APAP obtained on both modified electrodes (AAG-CPE and OAG-CPE) are more pronounced than those on the unmodified CPE. Table 1 gives the summary of oxidation peak intensities of each analyte obtained on different tested electrodes.

From Table 1 , the oxidation signal of QCT is about 4.57 µA on the unmodified CPE and increases to 5.72 µA on both acylated and oxidized modified CPE ( Figure 2(a) ). Similar results are observed with APAP where the oxidation signal increases from 10.30 µA to 12.12 µA and then to 12.72 µA on CPE, AAG-CPE and OAG-CPE respectively ( Figure 2(b) ). With RUT, a clear difference is observed between the bare CPE and the two modified electrodes with a twofold increase in the oxidation peak current (5.71 µA and 10.14 µA /10.29 µA respectively) ( Figure 2(a) ). The oxidation signal of PAP increases slightly from 12.10 µA to 12.12 µA on CPE and AAG-CPE respectively and becomes more pronounced (16.36 µA) on OAG-CPE ( Figure 2(b) ). These results can be explained by the fact that the acylation of AG introduces new functions at its surface, which interact positively with either alcohol or phenol groups of detected analytes by increasing their voltammetric signals. The OH functions increase and become more important at the surface of oxidized AG and this justifies the improvement of oxidation signals at OAG-CPE on the detections of QCT and RUT ( Figures 2(a)-(c) ), and of PAP and APAP ( Figure 2(b) and Figure 2(c) ). These results also demonstrate that both acylated and oxidized AG were successfully inserted into the carbon paste.

Figure 3 shows cyclic voltammetric signals of 0.5 mg/L of QCT and RUT ( Figure 3(a) ), and of 10 mg/L of the mixture PAP and APAP ( Figure 3(b) ) on OAG-CPE at different scan rates.

Insets of Figure 3 show both oxidation and reduction peak currents of QCT, RUT, PAP and APAP on OAG-CPE, plotted against the square root of scan rates (Ip = f(ν^1/2)). As seen, excellent linear relationships were obtained between anodic and cathodic peak currents for both QCT and RUT ( Figure 3(a) ) and for both PAP and APAP ( Figure 3(b) ), and the square root of scan rates (ν) from 0.01 to 0.5 V/s (insets of Figure 3(a) and Figure3(b) ). They are expressed respectively by the following equations: Ip_a(QCT) (µA) = 9.78υ^1/2 (V/s)^1/2 − 0.88 (R² = 0.997) and Ip_c(QCT) (µA) = −11.37υ^1/2 (V/s)^1/2 + 1.23 (R² = 0.993); Ip_a(RUT) (µA) = 10.20υ^1/2 (V/s)^1/2 − 0.88 (R² = 0.995) and Ip_c(RUT) (µA) = −11.23υ^1/2 (V/s)^1/2 + 1.23 (R² = 0.992); Ip_a(PAP) (µA) = 45.94υ^1/2 (V/s)^1/2 − 0.12 (R²= 0.997) and Ip_c(PAP) (µA) = −66.13υ^1/2 (V/s)^1/2 + 2.07 (R² = 0.998); Ip_a(APAP) (µA) = 31.92υ^1/2 (V/s)^1/2 − 2.12 (R²= 0.990). These results clearly indicate that the electrode reactions correspond to a diffusion-controlled process [12] [13] [19] .

In this section, CV technique was used to investigate the simultaneous and electrochemical behavior of both QCT and RUT, then PAP and APAP on different tested electrodes, and to assess the impact of acylated and oxidized Arabic gums on the carbon paste. The coming section will consist of studying the sensibility of the OAG-modified CPE towards the simultaneous quantification of both QCT and RUT, then PAP and APAP using the SWV technique.

The analytical performance of the OAG-modified CPE for simultaneous detection of both QCT and RUT ( Figure 4(a) ), and of both PAP and APAP ( Figure 4(b) ) was studied with SWV. Figure 4 shows that the oxidation peak currents of the tested compounds increase linearly with increasing concentrations of QCT, RUT, PAP and APAP in the range from 0.02 to 0.1 mg/L for both flavonoids and from 1 to 9 mg/L for both PAP and APAP. The obtained regression equations were Ip_a(QCT) (µA) = 1.11C (µM) – 1.21 (R² = 0.998); Ip_a(RUT) (µA) = 0.22C (µM) +3.69 (R² = 0.982); Ip_a(PAP) (µA) = 5.00C (µM) + 2.78 (R² = 0.997) and Ip_a(APAP) (µA) = 2.69C (µM) − 2.33 (R² = 0.990) for QCT, RUT, PAP and APAP respectively (insets of Figure 4(a) & Figure 4(b) ).

The limits of detection (LOD) were found to be 0.005 mg/L, 0.023 mg/L, 0.039 mg/L and 0.105 mg/L for QCT, RUT, PAP and APAP respectively, calculated using 3 S/m, where S is the standard deviation of the interception and m is the slope of the regression line [25] [26] . Table 2 gives the obtained values of standard deviation and the slope of the regression for QCT, RUT, PAP and APAP.

The comparison of linear ranges and LOD of the proposed method to others previously reported in the literature is shown in Table 3 .

^anitrogen doped porous carbon, ^bpoly(3,4-ethylenedioxythiophene), ^cGold nanoparticles/tetraaminophenylporphyrin functionalized multiwalled carbon nanotubes nanocomposites.

Globally, from Table 3 , LOD of OAG-modified CPE are closer and even better in comparison to others previously reported in the literature. The principal advantage of the proposed approach is the possibility to easily and efficiently functionalize AG, in contrast to other materials used as electrode modifiers like poly (3, 4-ethylenedioxythiophene) or tetraaminophenylporphyrin [14] [15] .

Knowing that PAP is the main by-product of APAP, the quantification of APAP in any real matrices requires to be carried out in presence of PAP. Thus, OAG-CPE was applied for the direct detection of APAP in the commercial tablets of Doliprane® 500 using the common analytical method of internal standards. Figure 5 shows SWV curves recorded on a commercial tablet of Doliprane® 500 and the insert gives the relation between peak currents and APAP concentrations.

From Figure 5 , a linear relationship was obtained between the peak intensity (observed at 0.46 V versus SCE) and concentrations of added APAP. It is expressed by the following equation, Ip_a(APAP) (µA) = 48.67C (mg/L) + 5.95 (R² = 0.997). The obtained results ( Table 4 ) were in the range of the commercial limit of quality admitted by the European drugs regulation (3%) and the recovery was found to be 94.3% [12] .

Secondly, OAG-CPE was used to detect QCT and RUT in human urine simultaneously. In fact, human urine samples were enriched successively with 3, 5 and 7 mg/L of both QCT and RUT and their exact amounts were determined using SWV on OAG-CPE. The results obtained are summarized in Table 5 and Figure 6 shows typical curves recorded. All the samples were determined three times under the same conditions.

From Table 4 , recoveries of RUT ranged from 97% to 99%, and those of QCT ranged from 106% to 122% in human urine samples. Comparing both flavonoids, better recovery values were obtained with RUT, showing that this method could be effective and reliable for RUT detection in urine samples as previously reported [13] . The poor determination of QCT in urine samples may be due to the presence of some interfering compounds such as urea, creatinine or uric acid, which are always present in the urine, and may interfere by increasing the voltammetric signal of QCT. This clearly indicates that such samples should be pre-filtrated before any analyses. The results also show that the present sensor, which uses oxidized AG, could be useful for the direct detection of both QCT and RUT in human urine.