Electrochemical detection of choline at f-MWCNT/Fe3O4 nanocomposite modified glassy carbon electrode

Choline is employed as cholinergic activity marker in brain tissue in the field of clinical detection of diseases. Although, chromatographic methods and biosensors are the most commonly used techniques for choline detection, there is also an interest in exploring the efficacy of a cost effective non-enzyme based sensor for choline detection. Here, electrochemical sensors based on green synthesized metal oxides (iron (III) oxide nanoparticles) from Callistemon viminalis leaves and flowers extract (Fe3O4NPL and Fe3O4NPF) in combination of functionalized multi-walled carbon nanotube (f-MWCNT) supported on glassy carbon electrodes (GCE/f-MWCNT/Fe3O4NPL and GCE/f-MWCNT/Fe3O4NPF) were fabricated for choline detection. Morphological, structural and optical analysis of the nanocomposites were studied using scanning electron microscopy (SEM), fourier transform infrared spectroscopy (FTIR), X-ray diffractometer (XRD) and ultra violet-visible (UV–vis) spectroscopy accordingly. In contrast, electron transport properties on bare glassy carbon electrode (GCE) and nanocomposite modified electrodes (GCE/f-MWCNT/Fe3O4NPL and GCE/f-MWCNT/Fe3O4NPF) was examined through electrochemical characterization using cyclic voltammetry (CV) and electrochemical impedance spectroscopy (EIS). Electrochemical oxidation of choline was also studied through CV, EIS, square wave voltammetry (SWV) and chronoamperometry (CA). The result proved that f-MWCNT enhanced the reactivity of Fe3O4NP towards choline oxidation with voltammetric limit of detection (0.83 and 0.36 μM) for choline at GCE/f-MWCNT/Fe3O4NPL and GCE/f-MWCNT/Fe3O4NPF electrodes respectively. Designed sensors proved selective, reproducible, stable and applicable for real sample sensing in choline dietary supplements.


Introduction
Choline belongs to the family of water-soluble quaternary amine (trimethyl-β-hydroxyethyl ammonium), and officially choline is seen by the Institute of Medicine (IOM) as a vital nutrient for human with functions similar to vitamin B [1][2][3]. Choline is a precursor for neurotransmitter acetylcholine which participates in memory and muscle control, cell membrane signalling and transport of lipids [1,4,5]. In the area of clinical detection of diseases such as Parkinson's, Alzheimer's, multiple sclerosis and myasthenia gravis, choline is employed as a diagnostic of cholinergic activity in brain tissue [1,6,7]. Choline (Cho) aids the regulation of water transport in and out of the cells (Kidney glomerular) through oxidation process into osmolyte betaine [7,8]. Choline protects the foetus from environmental abuse such as alcohol that can lead to abnormalities in behaviour, organ structures, foetal loss and congenital disabilities [3,7]. Choline is obtained via intake of choline-rich food, supplements and de novo synthesis by the human body [2,3,7]. However, to sustain proper choline function in 2.3. Functionalization of multi-walled carbon nanotubes (MWCNT) MWCNT was functionalized by following a prescribed method with slight modification using a mixture of concentrated acid (H 2 SO 4 /HNO 3 ; 75:25) in a 3:1 ratio volume [43]. MWCNT (0.5 g) was mixed with 100 ml acid mixture stirred at constant temperature (50°C) for 30 min and centrifuged. The MWCNT suspension was washed several times with distilled water using a sintered glass until a pH of 7 was obtained and thereafter dried at 70°C for 12 h.

Preparation of nanocomposites (f-MWCNT-Fe 3 O 4 NP)
A 4.5 mg of Fe 3 O 4 NPL and 1.5 mg of functionalized MWCNT (f-MWCNT) were weighed into a glass vial with an addition of 1 ml dimethylformamide (DMF). The combination was ultrasonicated for 24 h at 25°C and oven dried at 55°C for 2 h to give f-MWCNT/Fe 3 O 4 NPL. Same step was followed in the preparation of f-MWCNT/Fe 3 O 4 NPF composite.

Characterization of nanocomposites
Structural properties of nanocomposites were studied with fourier transform infrared spectroscopy, a product of Opus Alpha-P, Brucker Corporation, Billerica, MA, USA. Ultraviolet visible spectra were obtained with Cary series UV-vis spectrophotometer, 300 UV-vis model, crystalline structures were studied on D8 advance x-ray diffractometer while morphological studies were conducted on a scanning electron microscopy.

Electrochemical studies
Electrochemical characterization of bare and modified glassy carbon electrodes was carried out using cyclic voltammetry (CV) at scan rate of 25 mVs −1 and electrochemical impedance spectroscopy (EIS) in 5 mM K 3 [Fe(CN) 6 ] solution prepared in 0.1 mM PBS, pH 7.4 at +0.2 V constant potential between 100 kHz and 0.1Hz frequency range. Electrochemical oxidation of choline was investigated in pH 7.3 lithium chloride solution using CV, EIS (at 0.5 V fixed potential), chronoamperometry (CA) and square wave voltammetry (SWV) at a frequency of 25 Hz. Selectivity studies were conducted using SWV and CA. Voltammetric and chronoamperometric experiment were conducted with an AUTO LAB potentiostat PGSTAT 302 (Eco Chemie, Utrecht, and The Netherlands) powered by the general purpose electrochemical system (GPES) software version 4.9 while EIS experiment was controlled by metrohm AUTOLAB frequency response analyser (FRA 32) controlled by NOVA software version 1.10.1.9.

Electrode modification
The modified electrodes were prepared by drop-drying method. Prior to the modification, GCE was cleaned by gentle polishing in aqueous slurry of alumina nano powder on a micro cloth pad followed by ultrasonically cleaned in methanol, thereafter in distilled water so as to remove particles of alumina nano powder and obtain a mirror like surface. The electrode was further sonicated in methanol, finally in distilled water and dried at 40°C for 3 min. Separate suspension of functionalized multi-walled carbon nanotube (f-MWCNT), metal oxide nanoparticles (iron (III) oxide nanoparticles representing Fe 3

Preparation of real samples
An opened capsule (powder) of CDP choline (citicholine) was dissolved in 2 ml LiCl solution and thereafter diluted 100 times using lithium chloride (LiCl) solution of pH 7.3. Same step was followed in the preparation of super B energy injection fizzy tablet containing choline bitartrate. A 2 ml of the diluted samples were further diluted in 100 ml standard flask and spiked with different concentrations of choline using addition methods. The amount of the analyte (choline) present in each aliquot solution was determined at GCE/ MWCNT/Fe 3 O 4 NPL and GCE/MWCNT/Fe 3 O 4 NPF electrodes using SWV.  [42], suggesting that Fe 3 O 4 NP entrapped in f-MWCNT retained its vital conformation, and has been successfully adsorbed on f-MWCNT [45].           between f-MWCNT and Fe 3 O 4 NP. However, electron transport properties were optimal at GCE/ f-MWCNT/Fe 3 O 4 NPL than GCE/f-MWCNT/Fe 3 O 4 NPF considering its current response which was approximately twice higher. Sequential order of oxidation peak current response of the electrodes is;

X-ray diffraction studies
The ratio of anodic to cathodic peak current response (I pa /I pc ) for GCE/f-MWCNT/Fe 3 O 4 NPL and GCE/f-MWCNT/Fe 3 O 4 NPF electrode was calculated to be 1.05 and 1.14 respectively which is greater 1. Peak potential separation (ΔEp) values of 163 and 102 mV were obtained at the respective nanocomposite modified electrodes which are greater than the theoretical value (59 mV) for a fast one-electron transport, indicating a quasi reversible reaction at the electrodes [48,49].

Effect of varying scan rate at GCE/f-MWCNT/Fe 3 O 4 NP electrodes
The influence of scan rate on the redox process of the nanocomposite modified electrodes surface was studied by CV (cyclic voltammetry) in 5 mM K 3 [Fe(CN) 6 ] solution in the range of 25-500 mVs −1 potential sweeps. An increase in peak currents with increasing scan rate was noticed as shown in figures 8(a) and (b). Also, anodic and cathodic peak potentials shifted more to the positive and negative values accordingly as scan rate (v) increases. The graph of peak currents (I p ) against square root of scan rate (v 1/2 ) gave 0. 99213   = values were quite higher than the theoretical value (0.118 Vdec −1 ) for one-electron process in the rate determining process which could be due to surface adsorption of reactants at the electrode surface [51]. Electron transfer rate constant (ks) for GCE/f-MWCNT/Fe 3 O 4 NPL and GCE/f-MWCNT/Fe 3 O 4 NPF was found to be 0.09 and 0.29 s −1 respectively using equation (7).  6 ] at the electrode surface [45]. Zѡ is warburg impedance which appeared as a semicircle-infinite linear diffusion [45], C dl is the double layer capacitance and Q is the constant phase element. Summary of the fitted impedance data is shown in table 2 where x 2 is the chi-square. The negative chi-square values and the percentage errors in parentheses confirmed successful fitting of EIS data. A large R ct value of 1402 Ω was obtained for bare GCE at the   Figure 12 represents possible electrochemical redox reaction mechanism of choline at nanocomposite modified electrodes.

Effects of varying scan rate at constant choline concentration (2 mM) in LiCl solution
Scan rate has great impact on the redox process of electrode surface. Hence, cyclic voltammograms of nanocomposite modified electrodes (GCE/f-MWCNT/Fe 3 O 4 NPL and GCE/f-MWCNT/Fe 3 O 4 NPF were recorded in the scan rate range from 25 400 mVs −1 ((figures 13(a) and (b)) in 10 mM LiCl solution containing 2 mM choline (Cho) concentration. Oxidation (anodic) peak currents increased linear with increase in scan rate, confirming a diffusion controlled process (figures 13(a) and (b)) while the peak potentials shifted to the more positive values with increasing scan rate. Disappearance of peak potentials (redox peaks) noticed from 250 mVs −1 scan rate in figure 13(b) could be ascribed to great polarization since the redox reaction potential      (7)). The disparity in Ks value as well as the number of n could be ascribed to different levels of catalysis at the modified nanocomposite electrodes surface and the electrode materials [55]. Tafel slope (b) values of 163 and 182 mVdec -1 were obtain for GCE/f-MWCNT/Fe 3 O 4 NPL and GCE/f-MWCNT/Fe 3 O 4 NPF) employing equation (6). The values were quite higher than the theoretical value (118 mVdec -1 ) which could possibly be attributed to adsorption of choline on the surface of the electrodes.   In addition, estimated electron transfer rate constant of (k 0 ) was found to be 1.99×10 -3 and 3.69×10

Selectivity of designed sensors
The common challenge in the electrochemical detection of choline is the interference from coexisting substances such as ascorbic acid (AA) and dopamine (DA). Selectivity of the designed electrochemical sensors in the presence of these possible interfering species was investigated using square wave voltammetry and chronoamperometry. Voltammetric response of choline in company of AA and DA was carried out using square wave voltammetry, within −0.2 to 0.8 V potential windows, 0.005 V, 0.01 V, 25 Hz, 5 ss, and +0.58 V versus Ag/AgCl sat'd 3 M KCl potential step, amplitude, frequency, equilibration time and applied potential accordingly in LiCl

Analytical application of designed sensors
The feasibility of designed sensors for determination of choline in real samples was demonstrated with pharmaceutical samples (choline dietary supplements); CDP choline (citicholine) and choline bitartrate in super B energy injection fizzy tablet using SWV. Standard addition experiment was conducted by adding known amount of choline standard solution into prepared samples. Choline concentration was evaluated using SWV set at +0.58 V in +0.2-0.8 V potential windows. Table 4 presents the obtained result with % recovery for n=3 at 95%, indicating that the designed sensors are reliable for determining choline from its drug.
The result of t-test, F-test carried out at 95% confidence level signifies there was no significant difference in the result. The Q-test analysis indicated that the outliers (80, 86 recoveries) be retained since Qcal was<Qtab.  Significant increase in current response of choline upon several additions of AA and DA was noticed from the interference study using chronoamperometry connoting non-interference of AA and DA signal with that of choline. The designed sensors were successfully applied for direct choline detection in real samples (choline dietary supplements) with good percentage recoveries. The study shows that, the use of electrochemical sensor for choline detection can be explored.