sensors Article Article

Electrochemical Behavior Behavior and and Determination Determination of of Electrochemical Chlorogenic Acid Acid Based Based on on Multi-Walled Multi-Walled Carbon Carbon Chlorogenic Nanotubes Modified Modified Screen-Printed Screen-PrintedElectrode Electrode Nanotubes

Xiaoyan Ma 1,2, Hongqiao Yang 1,2, Huabin Xiong 1,2, Xiaofen Li 1,2, Jinting Gao 1,2 and Yuntao Gao 1,2,* Xiaoyan Ma 1,2 , Hongqiao Yang 1,2 , Huabin Xiong 1,2 , Xiaofen Li 1,2 , Jinting Gao 1,2 1 The and Yuntao Gao 1,2,Laboratory * Engineering of Polylactic Acid-Based Functional Materials of Yunnan, School of Chemistry and Yunnan Minzu University, Kunming 650500, China; [email protected] (X.M.); TheEnvironment, Engineering Laboratory of Polylactic Acid-Based Functional Materials of Yunnan, School of Chemistry [email protected] (H.Y.); [email protected] (H.X.); [email protected] (X.L.); and Environment, Yunnan Minzu University, Kunming 650500, China; [email protected] (X.M.); [email protected] (J.G) (H.Y.); [email protected] (H.X.); [email protected] (X.L.); [email protected] 2 Key Laboratory of Chemistry in Ethnic Medicinal Resources, State Ethnic Affairs Commission & Ministry of [email protected] (J.G.) Yunnan Minzu University, China 2 Education, Key Laboratory of Chemistry in EthnicKunming Medicinal650500, Resources, State Ethnic Affairs Commission & Ministry of * Correspondence: [email protected].; Tel.: +86-871-6591-0017 Education, Yunnan Minzu University, Kunming 650500, China

1

*Academic Correspondence: [email protected]; Tel.: +86-871-6591-0017 Editors: Jesus Iniesta Valcarcel and Craig E. Banks Received: 12 July 2016; Accepted: 12 October 2016; Published: Academic Editors: Jesus Iniesta Valcarcel and Craig E. Banks date Received: 12 July 2016; Accepted: 12 October 2016; Published: 27 October 2016

Abstract: In this paper, the multi-walled carbon nanotubes modified screen-printed electrode Abstract: In this was paper, the multi-walled carbon nanotubes electrode (MWCNTs/SPE) prepared and the MWCNTs/SPE was modified employedscreen-printed for the electrochemical (MWCNTs/SPE) wasantioxidant prepared and the MWCNTs/SPE was(CGAs). employed forofthe electrochemical determination of the substance chlorogenic acids A pair well-defined redox determination of the antioxidant substance chlorogenic acids (CGAs). A pair of well-defined peaks of CGA was observed at the MWCNTs/SPE in 0.10 mol/L acetic acid-sodium acetate buffer redox peaks of the CGAelectrode was observed at the in 0.10 mol/L acetic acid-sodium acetate (pH 6.2) and process wasMWCNTs/SPE adsorption-controlled. Cyclic voltammetry (CV) and buffer (pH 6.2) and the electrode process was adsorption-controlled. voltammetry (CV) and differential pulse voltammetry (DPV) methods for the determination Cyclic of CGA were proposed based differential pulse voltammetry (DPV) methods for the the determination of CGAexhibited were proposed based on the MWCNTs/SPE. Under the optimal conditions, proposed method linear ranges −5 on the0.17 MWCNTs/SPE. Under the linear optimal conditions, the proposed exhibited linearmol/L) ranges+ from to 15.8 µg/mL, and the regression equation was Ipamethod (µA) = 4.1993 C (×10 from 0.17 15.8 µg/mL, the linear regression was 0.12 Ipa (µA) = 4.1993 C (×10−5ofmol/L) + 1.1039 (r =to0.9976) and theand detection limit for CGAequation could reach µg/mL. The recovery matrine 1.1039 (r = 0.9976) and the detection limit for CGA could reach 0.12 µg/mL. The recovery of matrine was 94.74%–106.65% (RSD = 2.92%) in coffee beans. The proposed method is quick, sensitive, was 94.74%–106.65% (RSDfor = 2.92%) in coffee beans. The proposed method is quick, sensitive, reliable, reliable, and can be used the determination of CGA. and can be used for the determination of CGA. Keywords: chlorogenic acid; screen-printed electrode; multi-walled carbon nanotubes Keywords: chlorogenic acid; screen-printed electrode; multi-walled carbon nanotubes

1. Introduction 1. Introduction Chlorogenic acids (CGAs) (Figure 1) are a group of polyphenolic compounds common in Chlorogenic acids (CGAs) (Figure 1) are a groupfoods of polyphenolic compounds in in different different plant materials including many common and beverages [1], butcommon especially coffee, plant materials including many common foods and beverages [1], but especially in coffee, which has which has one of the highest concentrations of CGA of all plant constituents [2]. Many reports have one of the that highest concentrations CGA of allplays plantaconstituents [2]. Many reports have indicated a diet rich in CGA of compounds significant role in preventing manyindicated negative that a diet rich in CGA compounds plays a significant role in preventing many negative of effects of aging, as well various diseases associated with oxidative stress such aseffects cancer, aging, as well various diseases associated withdisease oxidative stress such as cancer, cardiovascular, aging cardiovascular, aging and neurodegenerative [3,4]. and neurodegenerative disease [3,4].

Figure Figure 1. 1. The The structure structure of of chlorogenic chlorogenic acid acid (CGA). (CGA). Sensors 2016, 16, 1797; doi:10.3390/s16111797 Sensors 2016, 16, 1797; doi:10.3390/s16111797

www.mdpi.com/journal/sensors www.mdpi.com/journal/sensors

Sensors 2016, 16, 1797

2 of 10

Several methods have been developed for the determination of chlorogenic acid and its derivatives in coffee beans and other plants. The most widely used methods are HPLC [5–7], capillary electrophoretic [8,9], and Miceller electrokinetic chromatography [10]. Although these developed methods have been efficient for the quantification of CGA and its derivatives, they have been criticized as being tedious and time consuming, and most of the instruments necessary for these methods are very expensive. In addition, UV-Vis spectrophotometer method is simple, fast and inexpensive for the determination of CGA in coffee beans; however, a direct spectral determination in coffee beans is relatively difficult, because of the spectral overlap with caffeine. In recent years, electrochemical methods have been widely investigated in the determination of phenolic compounds due to their simplicity, low cost, high sensitivity and rapid response [11,12]. Furthermore, caffeic acid as another important component of coffee can be broadly studied by cyclic voltammetry methods [13,14]. Nevertheless, to the best of our knowledge, the electrochemical determination of CGA has barely been reported. Screen-printed electrodes (SPEs) are especially recommended in the large-scale production of electrodes with easy-use and portability properties, which have been studied by Hart, Banks and Wang [15–18]. Also, these miniaturized screen-printed electrodes are suitable for working with sample microvolumes, and are disposable [19,20]. Screen-printed electrodes modified with multi-walled carbon nanotubes (MWCNTs/SPE) improve electron transfer properties, resulting in high sensitivity and low detection limits, decreased overpotentials, ease of mass production, and practicality [21]. Furthermore, they are described as useful electroanalytical tools for the development of analytical applications [22–25]. In this paper, we applied a simple and fast way to detect CGA with a highly sensitive voltammetric analysis method by using a modified screen-printed electrode with multi-walled carbon nanotubes. Here, the multi-walled carbon nanotubes material with a modified screen-printed electrode was prepared. The electrochemical behavior of CGA at MWCNTs/SPE was investigated, and a sensitive electrochemical analysis method of differential pulse voltammetry (DPV) was developed for the determination of CGA. Furthermore, the proposed method can be used in the quantitative determination of CGA in coffee beans. 2. Materials and Methods 2.1. Instruments, Materials and Reagents All electrochemical experiments were conducted with a ZAHNER Zennium IM6 Electrochemical Workstation (ZAHNER-elektrik GmbH and Co. KG, Kronach, Germany) with an integrated screen-printed three electrode device: a carbon working electrode, a carbon counter electrode, and an Ag/AgCl reference electrode. Scanning electron microscope (SEM, JSM-6360LV, JEOL, Co., Ltd., Tokyo, Japan). The carboxyl functionalized multi-walled carbon nanotubes (MWCNTs, purity > 95%, with a diameter of 10 nanometers, length of 5 nm) were purchased from Chengdu Institute of Organic Chemistry, Chinese Academy of Sciences; The screen-printed electrodes(work area of 3.1 square millimeter) were purchased from Methrom, Co., Ltd., Beijing, China; Coffee beans were sourced from Puèr University, Puèr, China; The specific concentration of chlorogenic acid (purity > 98%) was purchased from Sigma, St. Louis, MO, USA, and saved at 4 ◦ C. 0.10 mol/L sodium hydrogen phosphate-potassium dihydrogen phosphate buffer, 0.10 mol/L phosphate buffer solution (PBS), 0.10 mol/L citric acid buffer, 0.10 mol/L acetic acid-sodium acetate buffer, 0.10 mol/L sodium hydroxide solution, and 0.50 mmol/L potassium ferricyanide -potassium ferrocyanide solution (K3 Fe(CN)6 -K4 Fe(CN)6 ). Other reagents used were of analytical-reagent grade. Twice-distilled water was used throughout all experiments.

Sensors 2016, 16, 1797 Sensors 2016, 16, 1797

3 of 10 3 of 10

2.2. Experimental Methods 2.2. Experimental Methods 2.2.1. Purification and Functionalization of the Multi-Walled Carbon Nanotubes 2.2.1. Purification and Functionalization of the Multi-Walled Carbon Nanotubes In a 100 mL, three-necked, round-bottomed flask, the multi-walled carbon nanotubes of 500 mg a 100 mL, three-necked, round-bottomed flask,and the mixed multi-walled carbon nanotubes 500 mg and 50InmL concentrated nitric acid were firstly added homogenously. Next, theofmixture and 50 mL concentrated nitric acid were firstly added and mixed homogenously. Next, the mixture was constant-temperature reflowed for 12 h at 140 °C in an oil bath. Then, it was separated in the was constant-temperature for 12 h at 140 ◦nanotubes, C in an oilafter bath.centrifugal Then, it was separated centrifugal separator. Finally,reflowed the multi-walled carbon separation, werein the centrifugal separator. Finally, multi-walled nanotubes, after centrifugal separation, washed with distilled water and the then dried in a carbon vacuum oven [26]. Thus, the carboxylated were washed with distilled water and then dried in a vacuum oven [26]. Thus, the carboxylated multi-walled carbon nanotubes (MWCNT-COOH) were obtained. And the Raman spectra of multi-walled nanotubes (MWCNT-COOH) were2.obtained. Andspectra the Raman of MWCNT MWCNT and carbon MWCNT-COOH were shown in Figure Both of the have spectra the same pattern. and MWCNT-COOH were shown in Figure 2. Both of the spectra have the same pattern. Moreover, −1 Moreover, both Raman spectroscopy analyses showed a strong band at 1580 cm (G lines) which is −1 (G lines) which is the both Raman spectroscopy analyses showed strong band at and 1580a cm the Raman-allowed phonon high-frequency E2g afirst-order mode, disordered-induced peak at Raman-allowed phonon E2g first-order mode, andgraphene a disordered-induced peakasat 1358 cm−1 (D lines), which high-frequency may originate from defects in the curved sheets, tube ends, −1 (D lines), which may originate from defects in the curved graphene sheets, tube ends, 1358 cm well as the turbostratic structure of graphene in the materials [27,28]. Comparing the ratio of IG/ID of as well as the turbostratic graphene0.71 in the [27,28]. Comparing thewe ratio of Iable G /ID the two samples, which arestructure 0.95 for of MWCNT, for materials MWCNT-COOH, respectively, were the two samples, areof0.95 for MWCNT, 0.71 after for MWCNT-COOH, respectively, we were able toofdetermine that thewhich degree disorder is reduced carboxylation. Thus the carboxylation of to determine that the degree of disorder is reduced after carboxylation. Thus the carboxylation of MWCNTs might improve the electrochemical properties. MWCNTs might improve the electrochemical properties.

Figure 2. Raman spectra of multi-walled carbon nanotube (MWCNT) and MWCNT-COOH. Figure 2. Raman spectra of multi-walled carbon nanotube (MWCNT) and MWCNT-COOH.

2.2.2. Preparation and Characteration of the Multi-Walled Carbon Nanotubes Modified Screen-Printed Electrode 2.2.2. Preparation and Characteration of the Multi-Walled Carbon Nanotubes Modified Screen-Printed Electrode Before modifying the working electrode at the integrated SPEs, the SPEs were first washed with distilled and dried by N2 stream. SPEs was pre-anodized in afirst 0.1washed M (pH with = 7.4) Beforewater modifying the working electrodeThen at thethe integrated SPEs, the SPEs were PBS containing 0.1dried M KCl by2 stream. applying an the anodic of +1.9 V in (vs. Ag/AgCl) Then SPEspotential was pre-anodized a 0.1 M (pH = for 7.4)120 PBSs. distilled water and by N The MWCNTs/SPEs were prepared by coating 5 µL 0.3 mg/mL of the MWCNTs homogeneous containing 0.1 M KCl by applying an anodic potential of +1.9 V (vs. Ag/AgCl) for 120 s. The suspension ontowere the SPEs and then at room All modified electrodes MWCNTs/SPEs prepared by dried coating 5 µL temperature 0.3 mg/mL overnight. of the MWCNTs homogeneous were cleaned by the cyclic voltammetric technique between –0.5 andovernight. +0.5 V at aAll scan rate of 50 mV/s in suspension onto SPEs and then dried at room temperature modified electrodes PBS (pH 7.4) until a stable cyclic voltammetric response was obtained, and then rinsed with water were cleaned by cyclic voltammetric technique between –0.5 and +0.5 V at a scan rate of 50 mV/s in and(pH dried stream [29]. The response SEM comparison of bare SPE andrinsed the multi-walled PBS 7.4)under until aa nitrogen stable cyclic voltammetric was obtained, and then with water carbon nanotubes modified stream screen-printed was shown Figure showed and dried under a nitrogen [29]. Theelectrode SEM comparison of in bare SPE 3. andThe theresults multi-walled that thenanotubes surface ofmodified the MWCNTs/SPEs waselectrode a kind of was reticulate cubic structure. It is results obviousshowed that the carbon screen-printed shown in Figure 3. The MWCNTs (withoflittle carbon impurities) distributed uniformlyIton surface of the SPE. that the surface the amorphous MWCNTs/SPEs was a kind of were reticulate cubic structure. is the obvious that The spaghetti-like MWCNTs formed a porous structure. The entangled cross-linked fibrils offered high MWCNTs (with little amorphous carbon impurities) were distributed uniformly on the surface of accessible surface area. MWCNTs formed a porous structure. The entangled cross-linked fibrils SPE. The spaghetti-like offered high accessible surface area.

Sensors 2016, 16, 1797 Sensors 2016, 16, 1797

4 of 10 4 of 10

(a)

(b)

Figure 3. The TheSEM SEMcomparasion comparasion bare screen-printed electrode (a)the and the multi-walled Figure 3. of of bare screen-printed electrode (SPE)(SPE) (a) and multi-walled carbon carbon nanotubes modified screen-printed electrode (b). nanotubes modified screen-printed electrode (b).

2.2.3. Electrochemical Analysis 2.2.3. Electrochemical Analysis

were activated 0.10 sodium mol/L sodium hydroxide Before usingthe theMWCNTs/SPEs, MWCNTs/SPEs, they Before using they were activated in 0.10in mol/L hydroxide solution solution between the potential range of −0.5 V and 1.0 V at a scan rate of 10 mV/s, then the between the potential range of −0.5 V and 1.0 V at a scan rate of 10 mV/s, then the MWCNTs/SPEs activated were characterized in 0.50 ferricyanide-potassium mmol/L potassium ferricyanide-potassium MWCNTs/SPEs activated were characterized in 0.50 mmol/L potassium ferrocyanide solution 3 Fe(CN) 6 -K 4 Fe(CN) 6 ) at a scan rate of 0.15 V/s. ferrocyanide solution (K (K3 Fe(CN)6 -K4 Fe(CN)6 ) at a scan rate of 0.15 V/s. Cyclic voltammetry (CV) differential pulse pulse voltammetry voltammetry (DPV) performed in the Cyclic voltammetry (CV) and and differential (DPV) were were performed in the three-electrode in 0.10 0.10 mol/L mol/L acetic = 6.2) between the the three-electrode cell cell in acetic acid-sodium acid-sodium acetate acetate buffer buffer solution solution (pH (pH = 6.2) between potential range range of of − −0.5 V/s. The potential 0.5 V V and and +0.5 +0.5 V V at at aa scan scan rate rate of of 0.15 0.15 V/s. TheDPV DPV conditions conditions were: were: pulse pulse width width of ms, pulse pulse amplitude amplitude of of 180 180 mV mV and and pulse pulse interval interval of of 50 50 ms. ms. of 50 50 ms, 3. Results Resultsand andDiscussion Discussion 3.1. Cyclic Cyclic Voltammetry Voltammetry and and Differential Differential Pulse 3.1. Pulse Voltammetry Voltammetry Behaviors Behaviors of of CGA CGA at at MWCNTs/SPE MWCNTs/SPE Figure 44 displays acetate buffer buffer Figure displays the the CV CV curves curves of of CGA CGA in in the the 0.10 0.10 mol/L mol/L acetic acetic acid-sodium acid-sodium acetate solution (pH 6.2) at different electrodes included bare SPE (Figure 4a) and MWCNTs/SPE (Figure solution (pH 6.2) at different electrodes included bare SPE (Figure 4a) and MWCNTs/SPE (Figure 4b). 4b). The 0.5 VVtoto0.5 The result shows that there is no The scan scan rate rateisis0.15 0.15V/s V/s with withthe thepotential potentialrange rangefrom from−−0.5 0.5V.V. The result shows that there is electrochemical response; however, an obvious of redox were obtained at MWCNTs/SPE. no electrochemical response; however, an pair obvious pairpeaks of redox peaks were obtained at The oxidation peak (Epa ) andpotential reduction potential (Epcpeak ) of CGA were 0.08 −0.19 V MWCNTs/SPE. Thepotential oxidation peak (Epeak pa) and reduction potential (Epc)Vofand CGA were (vs. as ∆Erespectively, = Epa − Epc as = 0.27 TheAg/AgCl). ratio of the oxidation 0.08 Ag/AgCl), V and −0.19respectively, V (vs. Ag/AgCl), ΔE= V Epa(vs. − EAg/AgCl). pc = 0.27 V (vs. The ratio of peak current and peak (Ipa :Ipc ) peak was 0.42, that the electrode of CGA at the oxidation peakreduction current and reduction (Ipa:Ipcimplying ) was 0.42, implying that theprocess electrode process MWCNTs/SPE is quasi-reversible. of CGA at MWCNTs/SPE is quasi-reversible. 3.2. Influence Influence of of Supporting Supporting Electrolyte Electrolyte and 3.2. and pH pH Several supporting Several supporting electrolytes electrolytes such such as as 0.10 0.10 mol/L mol/L potassium potassium hydrogen hydrogen phosphate-potassium phosphate-potassium dihydrogen phosphate buffer (Figure 5a), 0.10 mol/L phosphate buffer 7%7% NaNa 2 HPO 4+ dihydrogen phosphate buffer (Figure 5a), 0.10 mol/L phosphate buffersolutions solutions(PBS, (PBS, 2HPO 4 1% KH PO + 90% NaCl + 2% KCl) (Figure 5b), 0.10 mol/L citric acid buffer (Figure 5c), and 0.10 mol/L 2 4 + 1% KH2PO4 + 90% NaCl + 2% KCl) (Figure 5b), 0.10 mol/L citric acid buffer (Figure 5c), and 0.10 mol/L acetic acid-sodium 5d)5d) were tested at MWCNTs/SPE. A pair CVofredox peaks acetic acid-sodiumacetate acetatebuffer buffer(Figure (Figure were tested at MWCNTs/SPE. A of pair CV redox are observed in the four supporting electrolytes. A better-defined CV response higherwith redox peak peaks are observed in the four supporting electrolytes. A better-defined CVwith response higher of CGA than in the other cases was obtained in 0.10 mol/L acetic acid-sodium acetate buffer. redox peak of CGA than in the other cases was obtained in 0.10 mol/L acetic acid-sodium acetate

buffer.

Sensors 2016, 16, 1797 Sensors 2016, 16, 1797

5 of 10 5 of 10

Sensors 16, 1797 Sensors 2016, 2016, 16, 1797

5 of 10 5 of 10

Figure 4. Cyclic voltammetry curves of CGA (a) at bare SPE and (b) at MWCNTs/SPE Figure 4. Cyclic voltammetry curves of CGA (a) at bare SPE and (b) at MWCNTs/SPE Figure 4. Cyclic voltammetry curves of CGA (a) at bare SPE and (b) at MWCNTs/SPE. Figure 4. Cyclic voltammetry curves of CGA (a) at bare SPE and (b) at MWCNTs/SPE

Figure 5. Influence of different supporting electrolytes on peak current: (a) 0.10 mol/L potassium Figure 5. Influence of different supporting electrolytes on peak current: (a) 0.10 mol/L potassium Figure 5. Influence of different supporting electrolytes peak current: 0.10 mol/L potassium hydrogen phosphate-potassium dihydrogen phosphateonbuffer; (b) 0.10(a) mol/L phosphate buffer hydrogen phosphate-potassium dihydrogen phosphate buffer; (b) 0.10 mol/L phosphate buffer hydrogen phosphate-potassium dihydrogen phosphate buffer; (b) 0.10 mol/L phosphate buffer solutions; (c) 0.10 mol/L citric acid buffer; (d) 0.10 mol/L acetic acid-sodium acetate buffer. solutions; 0.10 mol/L citric acid buffer; (d) 0.10electrolytes mol/L acetic acetate buffer. Figure(c)5. ofcitric different supporting onacid-sodium peak current: (a) 0.10 mol/L potassium solutions; (c)Influence 0.10 mol/L acid buffer; (d) 0.10 mol/L acetic acid-sodium acetate buffer. hydrogen phosphate-potassium dihydrogen buffer; (b) 0.10 mol/L phosphate buffer The influence of pH was investigated in 0.10phosphate mol/L acetic acid-sodium acetate buffer, as shown influence of 0.10 pHpH was investigated in in 0.10 mol/L acetic acid-sodium acetate buffer, asas shown solutions; mol/L citric acid buffer; (d) 0.10 mol/L acetic acid-sodium acetate buffer. The influence of was investigated 0.10 mol/L acetic acid-sodium acetate buffer, shown in The Figure 6, the (c) oxidation peak current of CGA increases with the increasing of pH from 4.0 to 6.2,

in in Figure 6, 6, thethe oxidation peak current CGA increases with the increasing ofof pH from 4.04.0 toto 6.2, Figure oxidation peak current of CGA increases with the increasing pH from 6.2, and then reaches its maximum at pH of 6.2, while the oxidation peak current decreases as pH increases The influence of pH was investigated in 0.10 mol/L acetic acid-sodium acetate buffer, as shown and then reaches its maximum at pH 6.2, while the oxidation peak current decreases as pH increases and then its maximum pH 6.2, while the oxidation peak(pH current asas pH increases above 6.2.reaches Therefore, 0.10 mol/Latacetic acid-sodium acetate buffer 6.2) decreases was chosen the optimal in 6.2. Figure 6, the oxidation peak current of CGAacetate increases with the6.2) increasing of pH from 4.0 to 6.2, above Therefore, 0.10 mol/L acetic acid-sodium buffer (pH was chosen asas the optimal above 6.2. Therefore, 0.10 mol/L acetic acid-sodium acetate buffer (pH 6.2) was chosen the optimal supporting electrolyte for subsequent experiments. and then reaches its at pH 6.2, while the oxidation peak current decreases as pH increases supporting electrolyte formaximum subsequent experiments. supporting electrolyte for subsequent experiments. above 6.2. Therefore, 0.10 mol/L acetic acid-sodium acetate buffer (pH 6.2) was chosen as the optimal supporting electrolyte for subsequent experiments.

Figure 6. Influence of buffer solution pHpH to peak current. Figure 6. Influence of buffer solution to peak current. Figure 6. Influence of buffer solution pH to peak current.

Figure 6. Influence of buffer solution pH to peak current.

Sensors 2016, 16, 1797

6 of 10

Sensors 2016, 16, 1797 Sensors 2016, 16, 1797

6 of610 of 10

3.3. Influence of the Scan Rate 3.3. Influence of the Scan Rate Figures 7 and 8 shows the effect of scan rate on the CV response of CGA at MWCNTs/SPE in the 3.3. Influence of the Scan Rate 0.10Figures mol/L 7acetic acetate buffer (pH 6.2). It of is CGA foundatthat both the oxidation and 8acid-sodium shows the effect of scan ratesolution on the CV response MWCNTs/SPE in the Figures shows the effectbuffer of scan on(pH the 6.2). CV response of CGA at MWCNTs/SPE peak current (I7paand )acid-sodium and8 reduction peak current (Irate pc) are linear to the scan rate (υ) inboth the range of 0.03 toin 0.10 mol/L acetic acetate solution It is found that the oxidation thecurrent 0.10 the mol/L acetic acid-sodium (pH Itrate is found the 0.15 V/s, regression equations of Ibuffer pa and Isolution pclinear are Ipato (µA) =scan 15.887 ν(υ) + 2.8809 = 0.9969) andtoIpc peak (Ilinear pa) and reduction peak acetate current (Ipc ) are the6.2). inthat the(rboth range of oxidation 0.03 peak ) and reduction peak current )are areindicates the scan (υ) in range of 0.03 (µA) =current 16.798 ν(I+pa4.1028 (r = 0.9955), respectively. This theνIprate proportional to the vIbut 0.15 V/s, the linear regression equations of Ipa and(IIpcpc Ilinear pa (µA)to =that 15.887 +is 2.8809 (r the = 0.9969) and pc to 1/2 0.15 V/s, the regression equations of Ipa This and Iindicates (µA)the = 15.887 ν +adsorption-controlled. 2.8809 (rto= the 0.9969) not . Therefore, the process of CGA atIpa MWCNTs/SPE is pc are (µA) =v16.798 ν +linear 4.1028 (relectrochemical = 0.9955), respectively. that Ip is proportional v butand Ipc = 16.798 ν +signal-to-noise 4.1028 (r = 0.9955), respectively. indicates that the Ip isof proportional toscan the v The maximum peak ratio for CGA wasThis achieved at the scan 0.15 V/s. The not v1/2(µA) . Therefore, the electrochemical process of CGA at MWCNTs/SPE is rate adsorption-controlled. 1/2 but not v V/s . peak Therefore, the electrochemical process CGA at MWCNTs/SPE rate of 0.15 wassignal-to-noise therefore selected work. The maximum ratiofor forthis CGA wasofachieved at the scan rateisofadsorption-controlled. 0.15 V/s. The scan The maximum peak signal-to-noise ratio for CGA was achieved at the scan rate of 0.15 V/s. The scan rate of 0.15 V/s was therefore selected for this work. rate of 0.15 V/s was therefore selected for this work.

Figure 7. Cyclic voltammetry curves of CGA at different scan rates. Figure 7. Cyclic voltammetry curves of CGA at different scan rates. Figure 7. Cyclic voltammetry curves of CGA at different scan rates.

Figure 8. The peak current of CGA at different scan rate. Figure 8. The peak current of CGA at different scan rate.

Figure 8. The peak current of CGA at different scan rate.

Therelationship relationshipbetween betweenthe thepeak peakcurrent current(I) (I)and andelectron electrontransfer transfernumber number(n) (n)comply complywith with The Equation (1) in the electrode reaction according to the Laviron theory [30]. Equation (1) in the electrode according to the [30].number (n) comply with The relationship betweenreaction the peak current (I) andLaviron electrontheory transfer Equation (1) in the electrode reaction according to the Laviron theory [30]. n 2 Fn22 FA2 AΓ  T vv nFQv nFQ v (1)(1) I  I2 = 2 4RT T = 4RT n F 4ARTT v n F4QR vT (1) I   T ·cm 4 R vT(mV/s) in the formula are the Faraday 2 4 R(mol −2 ) and 2), ),ΓΓ −2) and While,FF(96485 (96485 C·mol C·mol−1−),1 ),AA(cm (cm T T(mol·cm v (mV/s) in the formula are the Faraday While, −1), A area, 2), −2)quantity constant, the electrode surface the andand theinscan rate, respectively. Q = nFAΓ constant, electrode surface area, the adsorption quantity the rate, respectively. Q =T , (cm ΓTadsorption (mol·cm and v (mV/s) thescan formula are the Faraday While, the F (96485 C·mol Q is the peak area of area asurface single of cyclic (with of rate, electricity). The oxidation nFAΓ T, Q is the peak of aprocess single process ofvoltammetry cyclicquantity voltammetry (with quantity of electricity). constant, the electrode area, the adsorption andquantity the scan respectively. QThe = peak electron transfer number (n) was calculated to be 2.01 (v = 0.15 V/s) in this electrode reaction, oxidation electron number（n） was calculated to be(with 2.01 (v = 0.15 V/s) in this electrode nFAΓ T, Q is peak the peak area transfer of a single process of cyclic voltammetry quantity of electricity). The the oxidation peak potential (Enumber（n） reduction peak potential 0.08 Vwere and − 0.19VV pa ) and (E pc ) of reaction, the oxidation peak potential pa) and peak (Epcwere )V/s) of CGA 0.08 oxidation peak electron transfer was reduction calculated to be(Epotential 2.01 (v CGA = 0.15 in this electrode (vs. −0.19 Ag/AgCl), ∆E = (E Epa Epcreduction (vs. Ag/AgCl), the ofwere the oxidation and (vs. respectively. Ag/AgCl), respectively. ΔE == Epa0.27 − EVpcpeak = 0.27 V (vs. (E Ag/AgCl), the ratio of V the reaction, the Voxidation peak potential pa) − and potential pc ) ofratio CGA 0.08

and −0.19 V (vs. Ag/AgCl), respectively. ΔE = Epa − Epc = 0.27 V (vs. Ag/AgCl), the ratio of the

Sensors 2016, 2016, 16, 16, 1797 Sensors 1797

of 10 10 77 of

oxidation peak current and reduction peak (Ipa/Ipc) was 0.42, implying that the electrode process of peak current and reduction peak (Ipa /Ipc ) was 0.42, implying that the electrode process of CGA at CGA at MWCNTs/SPE is quasi-reversible. MWCNTs/SPE is quasi-reversible. 3.4. 3.4. The The Linear Linear Range Range and and Detection Detection Limit Limit A well-defined oxidation oxidation peak peakDPV DPVresponses responseswith withaahigh highpeak peakcurrent currentofof CGA was observed A well-defined CGA was observed at at MWCNTs/SPE. MWCNTs/SPE was the working electrode. The different concentrations of CGA MWCNTs/SPE. MWCNTs/SPE was the working electrode. The different concentrations of CGA were were into acid-sodium acetic acid-sodium acetate buffer(pH solution (pHthe 6.2), then the pulse differential pulse addedadded into acetic acetate buffer solution 6.2), then differential voltammetry voltammetry analysis was used in a pulse width of 50 ms, a pulse amplitude of 180 mV, and pulse analysis was used in a pulse width of 50 ms, a pulse amplitude of 180 mV, and pulse interval of 50 ms interval of 50 ms in the potential range of −0.5–0.5 V. We found that the oxidation peak current value in the potential range of −0.5–0.5 V. We found that the oxidation peak current value was linearly was linearly related to the concentration of CGA in the range of 0.17 to 15.8 µg/mL and the detection related to the concentration of CGA in the range of 0.17 to 15.8 µg/mL and the detection limit was limit was 0.12 µg/mL (shown in Figure 9). The regression equation was: Ipa (µA) = 4.1993 c (10−5 0.12 µg/mL (shown in Figure 9). The regression equation was: Ipa (µA) = 4.1993 c (10−5 mol/L) + 1.1039 mol/L) + 1.1039 (r = 0.9976). (r = 0.9976).

Figure 9. The The peak peak current current of CGA at different concentration. Figure

The detection limit limitand andlinear linearrange rangeofofthe theproposed proposed method have been compared with of The detection method have been compared with thatthat of the the other previously reported methods for determination the determination of CGA shown in Table It is evident other previously reported methods for the of CGA shown in Table 1. It 1. is evident that that the proposed electrochemical method shows high sensitivity with the lower detection limit, the proposed electrochemical method shows high sensitivity with the lower detection limit, indicating indicating that MWCNTs/SPE used for as athe sensor for the sensitive electrochemical detection of that MWCNTs/SPE can be usedcan as abe sensor sensitive electrochemical detection of CGA in many CGA in many samples [31–34]. samples [31–34]. Table with other other methods methods for for CGA CGA detection. detection. Table 1. 1. Comparison Comparison of of our our research research with Measurement Methods Measurement Methods

Linear Range

Detection Limit

Linear Range (µg/mL) (µg/mL) (μg/mL) Detection Limit (μg/mL)

UV-Vis spectroscopy UV-Vis spectroscopy Capillary electrophoresis with chemiluminescence Capillary electrophoresis with chemiluminescence HPLC Square-waveHPLC voltammetry Square-wave voltammetry Differential pulse voltammetry (DPV) using a MWCNTs/SPE

Differential pulse voltammetry (DPV) using a MWCNTs/SPE

10.7−39.010.7−39.0 1.10−110 1.10−110 0.8−20.0 1.77−17.70.8−20.0 0.17−15.81.77−17.7

16 16 0.5 0.5 0.32 0.27 0.32 0.12 0.27

0.17−15.8

0.12

References References [29] [29] [30] [30] [31] [31] [32] [32] this method

this method

3.5. Determination of CGA in Green Coffee Bean 3.5. Determination of CGA in Green Coffee Bean The coffee beans (From Puèr University, Puèr, China) were ground to powder with a mortar, The coffee beans (From Puèr University, Puèr, China) were ground to powder with a mortar, and the ground coffee sample was defatted with hexane (1:6; w/v) for 8 h in a Soxhlet extraction system. and the ground coffee sample was defatted with hexane (1:6; w/v) for 8 h in a Soxhlet extraction Then, CGA was extracted from the defatted coffee powder with water using the microwave-assisted system. Then, CGA was extracted from the defatted coffee powder with water using the extraction (MAE) lab station (Shanghai new apparatus of Microwave Chemical Technology Co., LTD, microwave-assisted extraction (MAE) lab station (Shanghai new◦ apparatus of Microwave Chemical Shanghai, China) for 5 min under the conditions of 800 w, 50 C and liquid-solid ratio of 5:1 [35]. Technology Co., LTD, Shanghai, China) for 5 min under the conditions of 800 w, 50 °C and According to the proposed method, the MWCNTs/SPE was applied to the determination of CGA in liquid-solid ratio of 5:1 [35]. According to the proposed method, the MWCNTs/SPE was applied to coffee beans, and the result is shown in Table 1. The standard deviation (RSD %) was found to be the determination of CGA in coffee beans, and the result is shown in Table 1. The standard deviation

Sensors 2016, 16, 1797

8 of 10

1.33%–4.77% and the recovery was 94.74%–106.65%. The result of CGA determination was in good agreement with that specified by HPLC (shown in Table 2.). Table 2. Measurement results of CGA in coffee beans (n = 5). By This Method

CGA Sample

(mg/g)

RSD (%)

1 2 3 4

13.23 21.12 25.21 18.74

4.77 3.14 2.47 1.33

Added (mg/g)

Found (mg/g)

Recovery (%)

By HPLC * (mg/g)

10.00 10.00 10.00 10.00

24.11 30.01 36.03 28.92

106.65 94.74 103.25 100.96

14.71 19.78 23.92 19.03

* The tested conditions of HPLC: The temperature of 25 ◦ C, the flow rate of 1.0 mL/min, the mobile phase was a mixture of acetonitrile (solvent A) and water–glacial acetic acid (99:1, v/v, pH 2.8) (solvent B).

4. Conclusions The multi-walled carbon nanotubes material modified screen-printed electrode (MWCNTs/SPE) was prepared and it was applied to the electrochemical behavior research and determination of CGA. The differential pulse voltammetry (DPV) method for the determination of CGA was proposed based on the MWCNTs/SPE. This secured several advantages, as the method is quick, sensitive and reliable. This proposed method can be used for the determination of CGA in the coffee beans. The advantages of this proposed method are: high sensitivity, simplicity of preparation at short time and good reproducibility. Acknowledgments: This work was supported by the National Natural Science Fundation of China (21367025), Program for Innovative Research Team (in Science and Technology) in University of Yunnan Province (2010UY08, 2011UYN09), Program for Yunnan Provincial Innovation Team (2011HC008), Program for State Ethnic Affairs Commission of the China (2014YNZ012) and Key Laboratory of Ethnic Medicine Resource Chemistry of State Ethnic Affairs Commission & Ministry of Education (MZY1302). Author Contributions: Xiaoyan Ma and Yuntao Gao designed the overall system for estimating the accuracy of the determination of CGA based on electrochemical method. In addition, they wrote and revised the paper. Hongqiao Yang, Huabin Xiong, Xiaofen Li and Jinting Gao helped the experiments. Conflicts of Interest: The authors declare no conflict of interest.

References 1. 2. 3.

4. 5.

6. 7. 8.

Leopoldini, M.; Russo, N.; Toscano, M. The molecular basis of working mechanism of natural polyphenolic antioxidants. Food Chem. 2011, 125, 288–306. [CrossRef] Adriana, F.; Tomas, D.O. Effect of roasting on the formation of chlorogenic acid lactones in coffee. J. Agric. Food Chem. 2005, 53, 1505–1513. Alvarez, J.L.; Wijngaard, H.; Arendt, E.K. Polyphenol composition and in vitro, antioxidant activity of amaranth, quinoa buckwheat and wheat as affected by sprouting and baking. Food Chem. 2010, 119, 770–778. [CrossRef] Mussatto, S.I.; Machado, E.M.S.; Martins, S.; Teixeira, J.A. Production, Composition, and Application of Coffee and Its Industrial Residues. Food Bioprocess Technol. 2011, 4, 661–672. [CrossRef] Weisz, G.M.; Schneider, L.; Schweiggert, U.; Kammerer, D.R.; Carleal, R. Sustainable sunflower processing—I. Development of a process for the adsorptive decolorization of sunflower [Helianthus annuus, L.] protein extracts. Innov. Food Sci. Emerg. Technol. 2010, 11, 733–741. [CrossRef] Upadhyay, R.; Mohan, R.L.J. An outlook on chlorogenic acids-occurrence, chemistry, technology, and biological activities. Crit. Rev. Food Sci. Nutr. 2013, 53, 968–984. [CrossRef] [PubMed] Butt, M.S.; Sultan, M.T. Coffee and its Consumption: Benefits and Risks. Crit. Rev. Food Sci. Nutr. 2011, 51, 363–373. [CrossRef] [PubMed] Hurtado-Fernández, E.; Gómez-Romero, M.; Carrasco-Pancorbo, A.; Fernández-Gutiérrez, A. Application and potential of capillary electroseparation methods to determine antioxidant phenolic compounds from plant food material. J. Pharm. Biomed. Anal. 2010, 53, 1130–1160. [CrossRef] [PubMed]

Sensors 2016, 16, 1797

9. 10.

11.

12.

13.

14.

15.

16.

17. 18. 19. 20. 21. 22.

23.

24.

25.

26. 27. 28.

9 of 10

Dmitrienko, S.G.; Kudrinskaya, V.A.; Apyari, V.V. Methods of extraction, preconcentration, and determination of quercetin. J. Anal. Chem. 2012, 67, 299–311. [CrossRef] Zarrin, E.; Golsefidi, M.A.; Saify, A.; Tanha, A.A.; Rezaeifaral, Z. Carbon nanotube reinforced hollow fiber solid/liquid phase microextraction: A novel extraction technique for the measurement of caffeic acid in Echinacea purpurea, herbal extracts combined with high-performance liquid chromatography. J. Chromatogr. A 2010, 1217, 2768–2775. Barroso, M.F.; De-Los-Santos-Álvarez, N.; Delerue-Matos, C.; Oliveira, M.B.P.P. Towards a reliable technology for antioxidant capacity and oxidative damage evaluation: Electrochemical (bio)sensors. Biosens. Bioelectron. 2011, 30, 1–12. [CrossRef] [PubMed] Yola, M.L.; Atar, N.; Üstündag, ˘ Z.; Solak, A.O. A novel voltammetric sensor based on p-aminothiophenol functionalized graphene oxide/gold nanoparticles for determining quercetin in the presence of ascorbic acid. J. Electroanal. Chem. 2013, 698, 9–16. [CrossRef] Apetrei, C.; Alessio, P.; Constantino, C.J.L.; Saja, J.A.D.; Rodriguez-Mendez, M.L. Biomimetic biosensor based on lipidic layers containing tyrosinase and lutetium bisphthalocyanine for the detection of antioxidants. Biosens. Bioelectron. 2011, 26, 2513–2519. [CrossRef] [PubMed] Abou Samra, M.; Sanda Chedea, V.; Economou, A.; Calokerinos, A.; Kefalas, P. Antioxidant/prooxidant properties of model phenolic compounds: Part I. Studies on equimolar mixtures by chemiluminescence and cyclic voltammetry. Food Chem. 2011, 125, 622–629. [CrossRef] Rawson, F.J.; Jackson, S.K.; Hart, J.P. Voltammetric Behavior of DNA and Its Derivatives Using Screen Printed Carbon Electrodes and Its Possible Application in Genotoxicity Screening. Anal. Lett. 2010, 43, 1790–1800. [CrossRef] Honeychurch, K.C.; Hart, J.P.; Cowell, D.C.; Arriganal, D.W.M. Voltammetric studies of lead at calixarene modified screen-printed carbon electrodes and its trace determination in water by stripping voltammetry. Sens. Actuators B Chem. 2001, 77, 642–652. [CrossRef] Tan, F.; Metters, J.P.; Banks, C.E. Electroanalytical applications of screen printed microelectrode arrays. Sens. Actuators B Chem. 2013, 181, 454–462. [CrossRef] Wang, J.; Pedrero, M.; Pamidi, P.V.A.; Caial, X. Metal-dispersed screen-printed carbon electrodes. Electroanalysis 1995, 7, 1032–1034. [CrossRef] Bandodkar, A.J.; Hung, V.W.; Jia, W.; Valdés-Ramírez, G.; Windmiller, J.R. Tattoo-based potentiometric ion-selective sensors for epidermal pH monitoring. Analyst 2013, 138, 123–128. [CrossRef] [PubMed] Washe, A.P.; Lozano, P. Facile and versatile approaches to enhancing electrochemical performance of screen printed electrodes. Electrochim. Acta 2013, 91, 166–172. [CrossRef] Gao, J. The Electrochemical Analysis of Cadmium in Choline Chloride/Urea Deep Eutectic Solvent Electrolyte at Carbon Nanotubes Modified Electrode. Int. J. Electrochem. Sci. 2016, 11, 6306–6314. [CrossRef] Zhu, J.; Xi, Q.; Yangal, W. A high performance electrochemical sensor for acetaminophen based on single-walled carbon nanotube–graphene nanosheet hybrid films. Sens. Actuators B Chem. 2012, 161, 648–654. Jain, R.; Rather, J.A. Voltammetric determination of antibacterial drug gemifloxacin in solubilized systems at multi-walled carbon nanotubes modified glassy carbon electrode. Colloids Surf. B 2011, 83, 340–346. [CrossRef] [PubMed] Dogan-Topal, B.; Bozal-Palabıyık, B.; Uslu, B.; Ozkanal, S.A. Multi-walled carbon nanotube modified glassy carbon electrode as a voltammetric nanosensor for the sensitive determination of anti-viral drug valganciclovir in pharmaceuticals. Sens. Actuators B Chem. 2013, 177, 841–847. [CrossRef] Liu, M.; Xiang, J.; Zhou, J.; Ding, H. A disposable amperometric sensor for rapid detection of serotonin in the blood and brain of the depressed mice based on Nafion membrane-coated colloidal gold screen-printed electrode. J. Electroanal. Chem. 2010, 640, 1–7. [CrossRef] Lee, G.W.; Kim, J.; Yoon, J.; Bae, J.S.; Shin, B.C. Structural characterization of carboxylated multi-walled carbon nanotubes. Thin Solid Films 2008, 516, 5781–5784. [CrossRef] Tan, P.; Zhang, S.L.; Yue, K.T.; Huang, F.; Shi, Z. Comparative Raman Study of Carbon Nanotubes Prepared by D.C. Arc Discharge and Catalytic Methods. J. Raman Spectrosc. 1997, 28, 369–372. [CrossRef] Li, W.; Zhang, H.; Wang, C.; Zhang, Y.; Xu, L. Raman characterization of aligned carbon nanotubes produced by thermal decomposition of hydrocarbon vapor. Appl. Phys. Lett. 1997, 70, 2684–2686. [CrossRef]

Sensors 2016, 16, 1797

29.

30.

31. 32. 33.

34. 35.

10 of 10

Zhang, L.; Li, Y.; Zhang, L.; Li, D.W.; Karpuzoval, D. Electrocatalytic Oxidation of NADH on Graphene Oxide and Reduced Graphene Oxide Modified Screen-Printed Electrode. Int. J. Electrochem. Sci. 2011, 6, 819–829. Ye, L.; Xiang, M.; Zhang, Y.; Luo, L.; Gao, Y. A Novel Electrochemical Method for Sensitive Detection of Anticancer Drug Picoplatin with Graphene Multi-walled Carbon Nanotubes Modified Glassy Carbon Electrode. Int. J. Electrochem. Sci. 2013, 8, 12726–12734. Belay, A.; Gholap, A.V. Characterization and determination of chlorogenic acid (CGA) in coffee beans by UV-Vis spectroscopy. Afr. J. Pure Appl. Chem. 2009, 3, 234–240. Han, S. Capillary electrophoresis with chemiluminescence detection of rutin and chlorogenic acid based on its enhancing effect for the luminol-ferricyanide system. Anal. Sci. 2005, 21, 1371–1374. [CrossRef] [PubMed] Zhang, Q.; Li, J.; Wang, C.; Sun, W.; Zhang, Z. A gradient HPLC method for the quality control of chlorogenic acid, linarin and luteolin in Flos Chrysanthemi Indici, suppository. J. Pharm. Biomed. Anal. 2007, 43, 753–757. [CrossRef] [PubMed] Ludwig, I.A.; Peña, M.P.; Cid, C.; Crozier, A. Catabolism of coffee chlorogenic acids by human colonic microbiota. Biofactors 2013, 39, 623–632. [CrossRef] [PubMed] Upadhyay, R.; Ramalakshmi, K.; Rao, L.J.M. Microwave-assisted extraction of chlorogenic acids from green coffee beans. Food Chem. 2012, 130, 184–188. [CrossRef] © 2016 by the authors; licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC-BY) license (http://creativecommons.org/licenses/by/4.0/).