Surface modified multiwalled carbon nanotube based molecularlyimprinted polymer for the sensing of dopamine in real samples usingpotentiometric method
Thayyath S. Anirudhan*, Sheeba Alexander, Aswathy LillyDepartment of Chemistry, School of Physical and Mathematical Sciences, University of Kerala, Kariavattom, Trivandrum 695 581, India
a r t i c l e i n f o
Article history:Received 21 May 2014Received in revised form26 July 2014Accepted 29 July 2014Available online 12 August 2014
The aim of this work was to test the application of multi-walled carbon nanotube based molecularlyimprinted polymer (MWCNTs-MIP) for the determination of dopamine (DA) in blood serum and urinesamples. The acrylamide grafted MWCNTs with vinyl group (MWCNTs-g-AAm-CH]CH2) was preparedby free radical graft copolymerization technique. MIP was synthesized by selective polymerization ofMWCNTs-g-AAm-CH]CH2 with itaconic acid (IA) as functional monomer in the presence of DA usingethylene glycol dimethacrylate (EGDMA) as a cross-linker and a,a0-azobisisobutyronitrile (AIBN) as theinitiator. The response time was ~2 min and the detection limit was 1.0 � 10�9 mol L�1, so the proposedsensor could be considered as a sensitive marker of DA depletion in Parkinson's disease. The MWCNTs-MIP sensor demonstrated a high selectivity, sensitivity and stability, in addition to that good repro-ducibility. The useful lifetime for the DA sensor was longer than 2 months and the sensor was suc-cessfully tested in real samples.
Dopamine [2-(3,4-dihydroxyphenyl)ethylamine, DA], a mono-amine neurotransmitter and hormone, is one of the biogenic cat-echolamines in the mammalian central nervous system. Theconcentration of DA in biological organisms is considerably low(0.01e1.0 mmol L�1). Abnormal levels of DA leads to several diseasessuch as epilepsy, senile dementia, HIV infection, Parkinson's, Alz-heimer's and Huntington's diseases [1]. In the patients of untreatedParkinson's disease most studies found that DA concentration de-creases considerably and it varies 1.22 ± 1.81 ng mL�1, to almostcomplete depletion [2]. Thus the quantification of DA in biologicalsystems is an important for diagnostic and pathological research.
Commonly used method for DA analysis involved relativelylaborious and expensive analytical techniques, such as mass spec-trometry [3], high performance liquid chromatography [4], flowinjection chemiluminescence [5] and optical absorption spectro-photometry in a microfluidic system [6] etc. Yet, these methodshave limitations of large sample volumes, expensive instrumenta-tion and environmental unfriendly solvents etc. Considering the
þ91 (0) 471 2308682.an).
above said factors along with the high electro activity of DA, elec-trochemical analytical technique for DA determination is anattractive method, owing to its low cost, ease of operation, fastresponse, high sensitivity and feasibility of miniaturization [7].
Molecular imprinting technique is a promising method for en-gineering three dimensional networks with a suitable recognitionsite complementary in size and shape to the template molecule forsubsequent rebinding process [8]. They were used in a variety ofapplications such as drug separations, template-assisted synthesisand catalysis and as receptor mimics biomimetic sensors andantibody mimics [9]. Yet, the bulk MIPs manifest high selectivity,there are certain drawbacks such as heterogeneous distribution ofbinding sites, the diffusion of the analyte across MIPs and the slowbinding kinetics along with poor site accessibility for templatemolecules [10]. Thus nanomaterials can be used as a supportingmaterial to overcome the above mentioned problems thatencountered with the use of MIPs. Multi-walled carbon nanotubes(MWCNTs) serve as an excellent support because it possesses largespecific surface area, highly porous nature and it consist of hollowstructures. This successful combination between MWCNTs andMIPs gives a superior route which is used for the large applicationof MIPs. Kan et al. [11] reported the immobilization of vinyl groupon the surface of MWCNTs for the selective determination of DA.Komathi et al. [12] reported nanomolar detection of DA through the
T.S. Anirudhan et al. / Polymer 55 (2014) 4820e4831 4821
use of a new nanocomposite made up of MWCNTs, a grafted silicanetwork and gold nanoparticles. Modified electrodes based onMWCNTs, such as MWCNTs modified gold electrode [13] and layer-by-layer assembled MWCNTs modified electrode [14] have beenemployed for the study of DA.
The presence of large surface area, high van der Waals force andhigh aspect ratio, these features make the MWCNTs to go throughself aggregation. A significant promising approach to recover theseproblems is the surface modification through graft polymerizationtechnique. The method is used to introduce alternate chemicalgroups on the surface of MWCNTs [15]. The hydrogen bondingamong the constituent acid groups present on the surface ofMWCNTs could lead to compact stacking of the CNTs and causebundling. Surface modifications of MWCNTs surface have been animportant research area due to their hydrophobicity properties inaqueous systems. It also effectively improves the dispersibility inaddition to reactivity of the MWCNTs for different applications.Chemical functionalization of MWCNTs during grafting canimprove its properties alongwith the polymermatrix produced canbe modified for specific applications. Acrylamide (AAm) is graftedon the surface of MWCNTs that increases its solubility. Zhang et al.(2010) [7] demonstrated the use of AAm for the grafting ofMWCNTs. In the literature of Kan et al. (2008) [11] firstly theMWCNTs was treated with nitric acid to give MWCNTs-COOH. Thiswill cause defects to the walls of the MWCNTs. But in this proposedmethod a direct grafting of AAm onto the surface of MWCNTs takesplace, which will minimize the defect formation and also increasesselectivity and sensitivity. The modification of the MWCNTs is acomplicated task and time-consuming, but once the material isprepared it is stable for a long time. In addition to that a minutequantity is required for sensing purposes and it can be reused forseveral times.
Previous works highlighted the MIP reports on electro-chemical sensors with capacitive [16], conductometric [17],amperometric and voltammetric [18] transduction. However, theabove said methods are concerned in the use of more sophisti-cated instrumentation, more difficult procedures; in addition tothat some of them suffer poor linearity of the calibration curve orextensive response times. The noteworthy feature of Potentio-metric sensors are simple design, good selectivity and givepromising results, which make it attractive [18]. Furthermore, forthe generation of membrane potentials, the potentiometric sen-sors do not need the template molecules to diffuse through theelectrode membranes [19]. Hence potentiometric technique is anexcellent option to combine through MIP technique. Moreover inthe proposed potentiometric method results a lower limit ofdetection of 1.0 � 10�9 M and good sensitivity compared to othermethods.
In the present work, a vinyl functionalized MWCNTs were pre-pared as a substrate, above which DA imprinted polymer wassynthesized by means of free radical polymerization. In the firstpart grafting of AAm onto the surface of MWCNTs was carried outusing (NH4)2S2O8 as a radical initiator and trimethylolpropane tri-methacrylate (TRIM) as a cross-linking agent. The double bonds onthe surface of MWCNTs were opened by initiator molecules. In thepresence of vinyl monomer AAm, the radical is added to the doublebond of the monomer resulting in a covalent bond between themonomer and the MWCNTs to form MWCNTs-g-AAm [20]. TheMWCNTs-g-AAm upon further reaction with allyl alcohol in thepresence of FeCl3 and CH3NO2 to formMWCNTs-CH]CH2 [21]. TheMWCNTs-CH]CH2 undergoes polymerization reaction with IA inpresence of EGDMA as the cross-linker, AIBN as the initiator and DAas the template molecule. The surface physical and chemicalproperties of the resulted materials were characterized andconfirmed by SEM, FTIR, XRD, TEM and Raman spectroscopy.
The mechanism of the formation of MWCNTs-g-AAm is shownbelow.
2. Initiation
Creation of primary radicals.
S2O2�8 /2SO�
4�
2SO�4
� þ H2O/HSO�4 þ �OH
SO�4
� and �OH are primary radicals.Creation of secondary radicalic sites on the surface of MWCNT.
SO�4
�.
�OHþMWCNTeH/MWCNT� þHSO�
4 or H2O
MWCNT�is secondary radical.
3. Propagation and termination
MWCNT�þCH2 ¼ CHeCOeNH2/MWCNTeCH2eCH2eCOeNH2
4. Material and methods
4.1. Materials
MWCNTs were obtained from Shenzhen Carbon Nanotechnol-ogies Co. Ltd. Dopamine hydrochloride, TRIM, EGDMA, IA, AAm,AIBN, allyl alcohol were purchased from SigmaeAldrich, MO, USA.Tetrahydrofuran (THF), acetonitrile, toluene, (NH4)2S2O8 anddimethyl formamide (DMF) were obtained from E.Merck IndiaLimited. Phosphate buffer solution (PBS, pH 7.0, ionic strength0.1 mol L�1) was used as a supporting electrolyte. Standard stocksolution (0.1 mol L�1) of DA was set via deionized water. For theanalytical applications, human blood serum and urine sampleswere collected from a clinical laboratory located at Thir-uvananthapuram City. These collected samples were stored in arefrigerator at �4 �C before use.
4.2. Instruments for characterization
The saturated calomel electrode (SCE) was used as referenceelectrode for potential measurements. Potentiometric measure-ments were made with 4 ½ Digit True RMS Multimeter (MODEL1085). pH measurements studies were determined using Systronic(model m pH system 362)-pH meter (Systronic India Ltd). Spectro-photometric determinations of DA were identified out using aJASCO UV-Visible (model V-530, Japan) spectrophotometer. Ramanspectra were collected with a micro-Raman spectrometer Lab RamUV HR, Jobin-Yvon and 10 mW intensity is used, in addition to that784.9 nm excitation wavelength of a diode laser was focused ontothe sample. FTIR spectra were recorded on a Perkin Elmer FTIRspectrophotometer. It was collected under room temperature/hu-midity control after background correction. Scans number was 32for both samples as well as background and X-Axis was wavenumber, ranging from 0 to 4000 cm�1 and Y-axis was % trans-mittance. The surface morphology study of MWCNTs-MIP andMWCNTs-NIP were performed using a JEOL JSM 6390 LA scanningelectron microscopy (SEM). SEM study was conducted at roomtemperature via using low vacuum mode. The transmission elec-tron microscopy (TEM) images were taken by Philips CM12 in-strument. X-ray diffraction (XRD) patterns were recorded using aRigaku Dmax IC model (Japan) X-ray diffractometer. HPLC analysiswas performed using Diomex Ultimate 3000 UHPLC and the
T.S. Anirudhan et al. / Polymer 55 (2014) 4820e48314822
detector used was FLD3000. The specific surface area of the ma-terial was obtained from the N2/77 K adsorption isotherm byapplying the Brunauer, Emmett and Teller (BET) method using across-sectional area of 0.162 nm2. A microprocessor controlledadsorption meter type Autosorb-1-MP instrument (QuantachromCorp., New York USA) was used.
4.3. Methods
4.3.1. Synthesis of MWCNTs-g-AAmAbout 30.0 mg of crude MWCNTs was dispersed in 20.0 mL of
DMF; the mixture was stirred and sonicated for 2 h to maintainhomogeneity. 0.01 g of AAm is added to the mixture solution ofMWCNTs and DMF. TRIM is used as the cross linker and it isdropped into the doubled necked flask in the molar ratio 28:1,followed by initiator (NH4)2S2O8. The resulting mixture was purged
Scheme 1. A schematic representatio
with N2 gas and it is stirred using a magnetic stirrer at 60 �C for20 h. Subsequent to cooling at room temperature, the AAm-g-MWCNTs was washed with anhydrous THF and filtered through a0.22 mm polycarbonate membrane filter paper and washed withdistilled water continuously. After a series of complete filtrationand washing, the solid material was dried under vacuum and formsthe product, MWCNTs-g-AAm.
4.3.2. Synthesis of vinyl group functionalized MWCNTsMWCNTs-g-AAm was suspended in the mixture of allyl alcohol
and nitromethane in the presence of catalytic amount of FeCl3under reflux for 24 h. After that the mixture was filtered through a0.22-mm polycarbonate membrane and washed thoroughly withanhydrous THF and finally with distilled water. The filtered solidwas vacuum dried to obtain vinyl group functionalized MWCNTs(MWCNTsCH]CH2).
n of formation of MWCNTs-MIP.
Fig. 1. The experimental set up for the detection of DA using MWCNTs-MIP sensor.
T.S. Anirudhan et al. / Polymer 55 (2014) 4820e4831 4823
4.3.3. Synthesis of MWCNTs-MIPThe formation of MWCNTs-MIP onto the surface of
MWCNTs-CH]CH2 is explained in Scheme 1. MWCNTs-CH]CH2was added to the mixture of 40 mL of acetonitrile and 8 mL oftoluene in a 500-mL round-bottom flask and purged with N2 undermagnetic stirring. DA and IA dissolved in 5 mL of DMF were addedto the reactor and mixed for 30 min to form a complex of templatemolecule and functional monomer. To this mixture, EGDMA andAIBN were also added, followed by increasing the temperature to80 �C. In addition to that the reaction mixture was allowed toproceed for 4 h. The finally obtained solid material was collectedand washed with ethanol to discard the unreacted reagents. Theresultant polymer was ground to powder and then was soxhletextracted by the mixture of methanol and acetic acid (9:1, v/v) forseveral times to extract the DA. The complete removal of DA fromMIP particles was confirmed by reading the absorbance of thewashout solution. The particles were repeatedly washed until DAwas no longer detected. Non-imprinted polymer (NIP) was pre-pared following the above said procedure in the absence of DAmolecule. The three different ratios of IA/EGDMA/AIBN were takento find out the optimized composition, i.e., to find out the moreselectiveMIP ratio and it is found to be 3/18/0.15, 3/12/0.15 and 3/9/0.15, respectively.
The amount of MWCNTs and the final polymer (MWCNTs-MIP)analysis were done by calculating the percentage of grafting usingthe equation that.
% of grafting ¼ W2 �W1ð ÞW1
� 100 (1)
Where W2 is the weight of the polymer after grafting andW1 is theoriginal weight of MWCNTs. The % of grafting of polymer(MWCNTs-MIP) was found to be 86%.
4.4. Electrochemical determinations
To study the electrochemical measurements, MWCNTs-MIPs(50 mg) were dispersed in 0.5 mL of methanol with ultrasonic for40 min. About 2 mL solution from the dispersed medium wasdropped on a clean Cu electrode surface, in addition to that; it isdried at room temperature for 1 h. 20 mL of 1% (w/v) agaroseaqueous solution was placed on the surface of the above electrodetill the gelling of agarose takes place. Subsequent to the completegelling process, the modified electrode was incubated in1.0 � 10�3 mol L�1 analyte solution and it was rinsed with distilledwater carefully. The electrode was then transferred into fresh PBSsolution for further electrochemical measurements.
Electrochemical measurements were carried out using a twoelectrode system consisting of Cu electrode as the working elec-trode and calomel electrode as the reference electrode. These twoelectrodes were dipped in a solution of phosphate buffer (pH¼ 7.0).The potential measurements were carried out using a multimeter.The experimental set up for the detection of DA using MWCNTs-MIP sensor is shown in the Fig. 1.
4.5. Selectivity studies
Matched Potential Method (MPM) was used to find out poten-tiometric selectivity coefficients. The initial concentration of pri-mary ion was set to 1.0 � 10�3 mol L�1 (aA) and the activity of theprimary ion solution added into this solution was1.0 � 10�1 mol L�1 (a0A). The corresponding potential change wasabout 10 mV. The interference of NH4
þ, Naþ, Kþ, Ca2þ, Mg2þ,cysteine, citric acid and ascorbic acid was assessed by the additionof small aliquots of the corresponding solutions into the primary
ion solution of aA until the same potential changewas observed, i.e.,until the increment of 10 mV was reached. It ensures that the finalconcentration of primary ion was not altered by more than 5.0%.
4.6. Real sample analysis
The serum samples were collected from the blood samples usinga simple standard method described by Kannan and John [22] andthe standard addition technique was used for the determination ofDA. Briefly, the blood samples obtained from the clinical laboratorywere centrifuged for 10 min at 5000 rpm and kept at 5� C. For thestudy, sample volume of 50 mL of the serum sample was diluted to5.0 mL with 0.1 mol L�1 PBS (pH 7.0). The human urine sample wascollected in a bottle. Before the sample analysis, it was centrifugedfor 10 min at 5000 rpm in order to remove precipitated proteinsand other particulate matters [23]. About 50 mL of the urine samplewas diluted to 5.0 mL with PBS (pH 7.0). About 1 and 1.5 ng mL�1 ofDA were spiked to urine and serum samples. The potential dis-played by the test solution was measured before and after theaddition of DA solution. To show the accuracy of the method, theresults obtained by current method were compared with thoseobtained from HPLC.
5. Results and discussion
The proposed reaction scheme is shown in Scheme 1. Grafting ofAAm onto the surface of MWCNTswas carried out using (NH4)2S2O8as a radical initiator and TRIM as a cross-linking agent. The per-sulphate initiator decomposed on heating to generate sulfateanion-radicals and these radicals start to initiate the grafting re-action by a free radical mechanism. The double bonds on the sur-face of MWCNTs were opened by initiator molecules. In thepresence of vinyl monomers, the radical is added to the doublebond of the monomer resulting in a covalent bond between themonomer and the MWCNTs to form MWCNT-g-AAm. This reactswith allyl alcohol in the presence of FeCl3 and CH3NO2 to give vinylfunctionalized MWCNTs-CH]CH2. MWCNTs-CH]CH2 reacts withIA in presence of EGDMA as the cross-linker and AIBN as theinitiator in presence of DA as the template molecule gives DAloaded MWCNTs-MIP. The template was removed by soxhletextraction to give MWCNTs-MIP. Three different ratios of IA/EGDMA/AIBN were taken such as 3/18/0.15, 3/12/0.15 and 3/9/0.15.Among these ratios, 3/18/0.15 gives more selectivity compared toother two ratios, so we optimize the ratio of IA/EGDMA/AIBN as 3/18/0.15. Three different types of porogens were used for the studysuch as acetonitrile; chloroform and DMSO, among these solvents
T.S. Anirudhan et al. / Polymer 55 (2014) 4820e48314824
DMSO give high sensitivity compared to other two solvents. Thevalues of the specific surface area were calculated according to theBET equation, and were found to be 165.89 and 195.28 m2/g forMWCNTs and MWCNTs-MIP respectively.
5.1. Characterization
The powder XRD spectra of MWCNTs, MWCNTs-g-AAm,MWCNTs-CH]CH2, DA, DA loaded MWCNTs-MIP, MWCNTs-MIPand MWCNTs-NIP are presented in Fig. 2. The diffraction spectraof MWCNTs-g-AAm matches well with the characteristic peaks ofAAm at 12.2�, 19.6�, 24.3� and 28.8� (2q) nevertheless with thedecreased intensity, which confirms the grafting of AAm onto thesurface of MWCNTs in presence of initiator. The attachment ofthe AAm on the surfaces of the MWCNTs causes a shift in the 2-theta value of the spectrum which reveals the crystalline state ofthe MWCNT is changed. The shift in the peak indicates that afterfunctionalization the crystallinity of MWCNTs is decreasedbecause of the weakened interactions present between the car-bon atoms. In the spectrum of MWCNTs-CH]CH2, the peakspresent at 12.5�, 28.2� and 31.5� indicate that the reaction withallyl alcohol is occurred and the vinyl group is attached on thesurface of MWCNTs. In the spectrum of DA a number of peakswere present at 18�, 22� 25�, 32�, 42� and 44� and in the spec-trum of DA loaded MWCNTs-MIP the peaks at 18�, 22� and 32�
indicate the presence of DA molecule. Nevertheless in the spec-trum of MWCNTs-MIP the peaks at 18�, 22� and 32� were absent,illustrating that the molecule is completely leached out from theMIP matrix. In the spectrum of MWCNTs-NIP the specific peaks
0 10 20 30 40 50 60 70 80 90 100
Cou
nts
2θ (degrees)
MWCNTs
MWCNTs-g-AAm
MWCNTs--CH=CH2
DA
DA loaded MWCNTs-MIP
MWCNTs-MIP
MWCNTs-NIP
Fig. 2. XRD spectra of MWCNTs, MWCNTs-g-AAm, MWCNTs-CH]CH2, DA, DA loadedMWCNTs-MIP, MWCNTs-MIP and MWCNTs-NIP.
present in the MWCNTs-MIP 18�, 22� and 32� were absent, itdemonstrates that template selective binding cavity is absent inthe MWCNTs-NIP. The decrease in the number of peaks demon-strates that the lesser number of hydrogen bonding in theMWCNTs-NIP, which also confirms smaller number of selectivebinding sites.
FTIR spectra were applied to characterize the structural changesof MWCNTs. The main functional groups of the predicted structurecan be observed with corresponding infrared absorption peaks. Theband around 1549 cm�1 corresponds to the graphitic structure ofMWCNTs (Fig. 3). In the spectrum of MWCNTs-g-AAm, absorptionbands at 3350 and 3190 cm�1 corresponds to the stretching vi-bration of the NH2 group. The absorption bands at 1670 and1620 cm�1 are attributed to stretching vibration of C]O andbending vibration of eNH2. The absorption band at 1455 cm�1 in-dicates the presence of bending vibration of eCH2e group. Theabove said data further confirms the grafting reaction of MWCNTsand AAm [24]. In the spectrum of MWCNTs-CH]CH2 the peak at1629 cm�1 was assigned to C]C stretch vibration. The band at1250 cm�1 corresponds to N–C stretching. The bands at 3010 and3053 cm�1 were attributed to the stretching vibration of the vinyliceCH2- bonds. This confirms the presence of vinyl functionalizedMWCNTs [25]. In the spectrum of DA, the peaks at 1590 and1491 cm�1 represent the absorption bands of the NeH bending andCeC stretching respectively. The peak at 3356 cm�1 is the typicalstretching vibration of eOH. The absorption bands at 1615 and1484 cm�1 are assigned to the bending vibration of NeH bond andCeC aromatic ring stretching respectively. The peaks present at3331and 2930 cm�1 were attributed to the stretching vibration ofeOH and eCH2e groups respectively.
Fig. 3. FTIR spectra of MWCNTs, MWCNTs-g-AAm, MWCNTs-CH]CH2, DA, DA loadedMWCNTs-MIP, MWCNTs-MIP and MWCNTs-NIP.
T.S. Anirudhan et al. / Polymer 55 (2014) 4820e4831 4825
The band at 1273 cm�1 represents the stretching vibrations ofCeO and aliphatic primary amine vibration. In the spectrum of DAloaded MWCNTs-MIP, the peak was considerably strong in theunwashed MWCNTs-MIP spectrum which reveals the presence ofthe formation of hydrogen bonding between DA andMWCNTs-MIP.In the spectrum of MWCNTs-MIP the peaks at 1590 and 1491 cm�1
were absent shows that DA molecule is completely removed fromthe MIP matrix. The absorption bands of 3500 cm�1 and 1728 cm�1
are assigned to eOH and C]O stretching vibrations of carboxylicgroups, respectively. Compared to the FTIR spectrum of MWCNTs-CH]CH2, the peaks corresponding to a C]C stretching vibrationare absent in MWCNTs-MIP indicating a polymerization reactionbetween the vinyl group of MWCNTs-CH]CH2 and that of themonomer IA. In the spectrum of DA loaded MWCNTs-MIP, thepeaks present is the combination of MWCNTs-MIP and DA indi-cating that DA is encapsulated in MWCNTs-MIP. In the spectrum ofDA loaded MWCNTs-MIP the peak 1590 and 1491 cm�1 are due tothe presence of DA molecule. In the spectrum of MWCNTs-NIP thepeaks present in the MIP were absent, demonstrating that astructural modification is occurred and it reveals the absence oftemplate selective binding cavity. The peaks present at 3500 and1728 cm�1 are due to the presence of OeH stretching and C]Ostretching vibrations of carboxylic groups.
The morphological features of the prepared samples werestudied using SEM analysis (Fig. 4). In the SEM image of pureMWCNTs before the surface modification it appears to be smoothand tidy. After the surface modification with AAm, the smoothsurface of MWCNTs has become rough and debundled. This
Fig. 4. SEM images of a) MWCNTs, b) MWCNTs-g-AAm, c) MWCNTs-CH]C
supports the grafting reaction of AAm on to the surface ofMWCNTs. MWCNTs-g-AAm further reaction with allyl alcoholforming MWCNTs-CH]CH2. In the SEM image of MWCNTs-CH]CH2, a homogeneous coating was appeared on the surface of theMWCNTs. This confirms that the reaction between AAm and allylalcohol was occurred. The coating is fluffy and demonstrates aporous characteristic. The MWCNTs-CH]CH2 upon polymeriza-tion reaction with DA molecule in the presence of EGDMA, AIBNand IA. The SEM image of DA loaded MWCNTs-MIP appears ahighly ordered and perfect alignment over the surface ofMWCNTs. This indicates that MIPs-imprinted form was success-fully polymerized on its surface. After being washed with themixture of methanol and acetic acid (9:1, v/v) the DA moleculeswere completely removed from the polymer network. In the SEMimage of MWCNTs-MIP some visible, irregular and large numbersof additional imprinted pores were present on the surface of MIP.This confirms that DA molecule is completely removed from thepolymer matrix. The average diameter of MWCNTs-MIP is greaterthan MWCNTs due the grafting reaction with AAm and allylalcohol. The surface of MWCNTs-NIP is homogeneous as thecomplementary site for the recognition of DA is not present.
SEM image of coating on the electrode is shown in the Fig. 5. Theimprinted materials (MWCNTs-MIP) show an agglomeratedappearance by means of irregular shape in the size range of50e100 mm. The surface of the membrane was seemed to be moreporous and the imprinted shell was successfully attached to thesurface of the material and the porous nature helps the effectivesensing of analyte molecule.
H2, d) DA loaded MWCNTs-MIP, e) MWCNTs-MIP and f) MWCNTs-NIP.
Fig. 5. SEM image of coating on the electrode.
Fig. 6. Raman spectra of MWCNTs, MWCNTs-g-AAm, MWCNTs-CH]CH2, DA loadedMWCNTs-MIP, MWCNTs-MIP and MWCNTs-NIP.
Fig. 7. TEM images of MWC
T.S. Anirudhan et al. / Polymer 55 (2014) 4820e48314826
The Raman spectra of the MWCNTs and functionalizedMWCNTs (Fig. 6) shows two characteristic bands of MWCNTs, Dband at 1350 cm�1 and G-band at 1580 cm�1. The said bandswere attributed to the disorder-induced peaks and tangentialmode peaks [26]. The presence of G band is related to the vi-bration of sp2 hybridized carbon atoms and the D is related to thedefects in the hexagonal graphitic layers [27]. The ratio of in-tensity of the G bands to D bands (R ¼ IG/ID) depends on thestructure of MWCNTs [28]. It typically serves as a measurementof the disordered sites on carbon nanotubes walls, helping as theindicator of the level of the covalent functionalization of theCNTs. If the ratio of IG/ID is higher, it shows the presence of moredisordered carbon. The presence of D' mode at ~1615 cm�1 alsoserves to demonstrate the disorder and defects of graphitic walls.The D line intensity in MWCNTs-g-AAm has increased, whichindicates an increase in defects or sp3 hybridized carbons. Thisinformation reveals that AAm was covalently grafted to the sur-face of the MWCNTs. It is noticeably interesting that the maximaof the D and G bands are changed to lower frequencies forfunctionalized MWCNTs. These above said results show that thereaction in Scheme 1 was completed successfully. The grafting tothe surface of the MWCNTs through the formation of covalentbonds had done successfully, which further supports the resultsobtained by FTIR analysis.
In the Raman spectrum of MWCNTs-g-AAm a weak shoulder D0
band around 1610 cm�1 is present which indicates the function-alization of MWCNTs. In the Raman spectrum of DA loadedMWCNTs-MIP, the peaks present at 1277, 1448, and 1618 cm �1 aredue to the CeO, eCH2e, and phenyl ring stretching modes,respectively [29]. The absence of two characteristic peaks of DA at1265 cm�1 and 1489 cm�1 in the spectrum of MWCNTs-MIP con-firms the complete removal of DA from MIP.
TEM images of MWCNTs and MWCNTs-MIP are shown inFig. 7. In the structure of MWCNTs, the individual tubes wereclearly separated from one another and it possesses an averagethickness of about 140.7 nm. Another peculiarity is that the wallof the MWCNTs is quite smooth and clean as well as it does notshow a covered structure with any additional phase. In contrast,the MWCNTs-MIP contains an extra phase and appears to bespherical as well as the size of the particle was also decreased.The extra phase is due to the structural change of MWCNTs-MIPfrom tubular to spherical. This result confirms that the surfacemodification of MWCNTs is occurred not only at the tips but alsoon the whole surface. The decrease of particle size is owing to thepresence of imprinted shell.
NTs and MWCNTs-MIP.
Fig. 9. Potential response of DA sensor in different concentrations.
T.S. Anirudhan et al. / Polymer 55 (2014) 4820e4831 4827
5.2. Effect of pH
The effect of pH on the MWCNTs-MIP was studied bymeasuring the electrode response at DA concentrations of1.0 � 10�3 mol L�1 (Fig. 8). Potentiometric measurements wereperformed in the pH range 3.0e8.0, adjusted by adding smallvolumes of concentrated HCl or concentrated NaOH to the testsolution. The DA exists mainly in protonated form at pH valueslower than the pKa1 value 8.57. The deprotonated form of the DAmolecule exists at pH values higher than the pKa2 value 10.08 andit exist mainly as a neutral molecule between pKa1 and pKa2. Thusthe protonated DA interacts weakly with the functional groups ofthe recognition site [30]. EMF observed initially in the pH range3.0e5.0 could be due to the presence of positive charge bearing DAand the acidic media. The sensor response is high in MWCNTs-MIPbecause of the presence of selective binding sites compared to theMWCNTs-NIP (Fig. 8).
5.3. Sensitivity and detection limit
The characteristic performance of the sensor membrane wascalculated under the optimal circumstances (Fig. 9). A graph isplotted by E vs log [DA], the sensor gave linear calibration plot with2 different slopes in the range of 10�9 to 10�6 mol L�1 and 10�6 to10�5 mol L�1 for MWCNTs-MIP and MWCNTs-NIP, respectively. Inthe plot of MWCNTs-MIP, the slopes obtained for 10�9 to 10�6 is25.3 and for 10�6 to 10�5 is 54 mv/decades respectively where as inthe plot of MWCNTs-NIP, slope obtained for 10�9 to 10�6 is 7.85 andfor 10�6 to 10�5 is 27 mv/decades.
Due to the presence of specific binding sites in the MWCNTs-MIP membrane it shows the stable potential value. The limit ofdetection of the prepared sensor obtained was1.0 � 10�9 mol L�1 (LOD is equal to 3 times the value of the
Fig. 8. Effect of pH for the sensing of DA using MWCNTs-MIP.
standard deviation of the output signal, s) for imprinted polymermembrane, which is based on IUPAC suggestion and it alsoequals to 0.15 ng mL�1 [31]. The sensor performance of theMWCNTs-NIP is also shown in the Fig. 9. In order to express therepeatability/reproducibility of the method, the replica mea-surements of DA with the electrode was calculated under thesame conditions (n ¼ 5). The R.S.D. value was calculated from theexperiment in a number of times and it is found to be 1.5%, it isthe precision obtained for the sensor. Additionally, the relevanterror percentage and accuracy were calculated in each case. Theerror percentages obtained from the experiment were 0.6%, 0.9%,1.1%, 1.3% and 1.5% respectively. Another important thing is thatthe response of the sensor maintained over 70e80% of theoriginal value for 20 days. The non-specific electrostatic inter-action occurs between the eCOOH group of the IA and eNH2group of dopamine molecule. This non specific interaction favorsthe adsorption of dopamine molecules to the electrode/sensorsurface [32,33]. The MWCNTs possesses high surface area andnumber of the sites present in the MWCNTs-MIP is also high andthis property of the material enhances detection of dopamine atlow concentration levels.
5.4. Stability
The long-term stability of the MWCNTs-MIP was tested over atwo-month period. The stability of the sensor was calculated and itwas found to be 2 months, by storing the sensor in 0.1 mol L�1 PBSat room temperature. There is no obvious changes were observed inthe first week and the response obtained is maintained by 95.0% ofthe initial response. Subsequent to storing the MIP material for 2months, the potentiometric response also maintained 93.0%compared with the initial steady state value. The results demon-strated that the stability of the sensor was satisfied for an extensivetime.
Table 1The comparison of the proposed MWCNTs-MIPs with other electrochemicalmethods for the detection of DA.
Method of detection Limit ofdetection(mol L�1)
Linear range (mol L�1) References
MIP-QCM 1.0 � 10�8 1.0 � 10�4 to 1.0 � 10�3 [33]MIP-CFME 3.8 � 10�7 1.0 � 10�6 to 1.0 � 10�4 [34]MIP-hybrid sensor 2.2 � 10�7 2.2 � 10�7 to 1.44 � 10�4 [35]MWCNTs-MIP/CCE 1.3 � 10�9 4.8 � 10�5 to 2.2 � 10�7 [36]MIP hybrid MWCNTs-CE 1.0 � 10-9 9.7 � 10�9 to 6.6 � 10�7 [37]MWCNTs-MIP 1.0 � 10�9 1.0 � 10�9 to 1.0 � 10�5 This work
T.S. Anirudhan et al. / Polymer 55 (2014) 4820e48314828
5.5. Comparison of MWCNTs-MIP with other methods
The result of MWCNTs-MIPs is compared with other publishedworks based on MIPs and it is shown in Table 1. MIP-QCM, MIP-CFME, MIP-hybrid sensor, MWCNTs-MIP/CCE, MIP hybridMWCNTs-CE having the detection limits of 1.0 � 10�8, 3.8 � 10�7,2.2 � 10�7, 1.3 � 10�9 and 1.0 � 10�9 mol L�1 respectively. The re-sults show that the value of detection limits in the present methodis lower and it gives a wider linear range of the prepared MWCNTs-MIPs sensor. It also gives an attractive viewpoint for its broadapplication. The LOD of the prepared sensor was found to be1.0 � 10�9 mol L�1 for DA. So the material is suitable for thedetection of DA in nano level concentrations. From this data it isclear that the current sensitivity and quantification range areenough for any potential clinical applications [11]. The MWCNTs-MIP used in the present work possesses the advantages of lowerdetection limit, high sensitivity and wider range of quantification(Table 1) [34e38].
5.6. Reusability
As sensor materials, the true value of MIP would be realized ifMIP can be reused many times and the reusability of the sameMIP-based potentiometric sensor was measured by using the responsesignal nine times in a continuous manner, and results are repre-sented in Fig. 10. After each use, the electrode was renewed by themethod of template extraction and stored at room temperature.Subsequent to each addition of DA and the attainment of equilib-rium, the electrode was recovered by sequential washes withmethanol/acetic acid (9:1, v/v) and distilled water until the initial
Fig. 10. Effect of reproducibility of the MWCNTs -MIP sensor in different cycles.
potential was attained. After the usage of the electrode it can bestored in methanol at 4 �C. The result reveals that the sensor showsexcellent reproducibility (Fig. 10). In the ninth cycle response signalof the sensor has the loss of 1.8%. Another peculiarity of the pre-pared sensor is that it posses long-term strength and the responseof the sensor does not change even after two months.
5.7. Sensor selectivity
Bakker and Pretsch [39] reported that the selectivity of an ISEdepends on the selectivity of the template/ionophore binding.Furthermore it is based on the standard free energies of therespective ions in the aqueous and organic phases. The selectivity ofDA over other interfering ions (X) present in the solution could beestablished in terms of the potentiometric selectivity coefficients(KMPM
DA;X ). Matched potential method (MPM) is widely used for thecalculation of potentiometric selectivity coefficients, which is rec-ommended by IUPAC [40]. At first, a known activity (a0DA) of DAsolution is added into a reference solution that contains a fixedactivity (aDA; 1 � 10�3 mol L�1) of DA, and the corresponding po-tential change (DE) is recorded. Then the interfering ions (X) aresuccessively added to an identical reference solution to provide thesame potential change (DE). The change in potential producedmustbe the same in both cases.
The selectivity coefficient (KMPMDA;X ), for interferences was calcu-
and ascorbic acid were tested as possible interfering species pre-sent in biological samples. The value of log KMPM
DA;X shows the vari-ation from �1.6 to-0.35 (Table 2). The effect of variation of theinterference of glucose and ascorbic acid was studied; it shows thatglucose has less interfering effect compared to uric acid. The effectof interfering ions varies in the order of Naþ<Mg2þ<NH4
þ<Kþ<Ca2þ.The MPM selectivity coefficients for various interfering ions arelisted in Table 2. The results in the table show that almost inter-fering ions interfere with DA ions that were present in the solution.In the cases of amino acids and other ions, the interfering ions wereunable to reach the potential changes that obtained for the DAmolecules. This is indicated by the missing values in Table 2. Thismay indicate very low interfering effect of the ions. In the case ofascorbic acid there is no influence with DA because it has lowsensitivity to the modified electrode. The concentration of DA isbelow the detecting limit (BDL), so it cannot be found directly. Inaddition to that theMWCNTs-MIP is highly selective towards DA, sothe interference with ascorbic acid is less. Another point is that,blood samples containing several other ions other than ascorbic
Table 2Potentiometric selectivity coefficients were assessed by thematched potential method (MPM).
T.S. Anirudhan et al. / Polymer 55 (2014) 4820e4831 4829
acid, in the prepared sensor, ascorbic acid does not interferencewith dopamine. But while conducting blood analysis there is apossibility for the interference of other ions.
5.8. Real sample analysis
Potentiometric methods were helpful for the determination ofDA in real samples by using the developed electrodes. Table 3shows the potentiometric results obtained from the analysis ofDA in real samples. The concentration of DA in blood of a patientsuffering from Parkinson's diseases is 1.22 ± 1.81 ng mL�1 and theexperiment is conducted using blood samples with 1 and1.5 ng mL�1. The results show that the obtained concentration is0.98 and 1.45 ng mL�1, so it is successfully used for the determi-nation of DA in real samples. The recovery values and the relativestandard deviation which were calculated when four samples ofthe same product were analyzed with three different electrodes. Anaverage recovery of about 93.5%, in addition to that a relative de-viation less than 1% was obtained. The method is used for thedetermination of human urine and blood samples. The obtainedvalues for the revival and R.S.D were satisfactory. So the proposedmethod could be efficiently used for the determination of DA incommercial samples. To check the accuracy of the method, theresults obtained in the present method were compared withanother validated technique like HPLC (Table 4). The results weresatisfactory by the HPLC method and recovery and RSD obtainedwere 97.08% and less than 1%, respectively.
5.9. Response time
Mehran et al. [41] reported that response time is the averagetime necessary for the sensor to reach ±1 mV of the magnitude ofthe equilibrated potential signal, all of them having a 10 fold con-centration difference. The concentrations of the solutions taken forthe study are 1�10�9 to 1�10�5 mol L�1. The response time of theMWCNTs-MIP was quite fast, within 2 min sensor gives betterresponse and attained equilibrium within 5 min (Fig. 11). It isevident that, the better performance of the MIP based potentio-metric sensor is due to the specific interaction of DAmolecules withthe recognition sites of MIP in the polymeric membrane and it isalso effective for the specific recognition of the target ions.
Table 4Analysis of urine and serum samples using the HPLC method.
Sample Concentration of DA (ng mL�1) RSD (%) Recovery (%)
To find out the dispersion ability of pure MWCNTs andMWCNTs-MIP, a small amount of each MWCNTs samples is placedin acetone and stirred using a water bath ultrasonicator for 10 min.Photograph demonstrating the dispersion of (a) MWCNTs, (b)MWCNTs-MIP and (c) MWCNTs-MIP after six months are shown inFig. 12. The figure shows that the MWCNTs-MIP remained as stablesuspension in the acetone for more than six months than the pureMWCNTs. The pure MWCNTs give unstable suspension after 5 h. Astable colloidal form of MWCNTs-MIP dispersion is obtained due tothe presence of functional groups present on the surface of thecarbon nanotubes. The weak tubeetube interactions present be-tween the MWCNTs floss after the debundling process. In additionto that their enhanced polarity by the surface modification onMWCNTs surface has improved the dispersibility and also it is astabilized form. The reason is that the carbonyl (C]O) group pre-sent in acetonewhich is polar in nature, it forms the hydrogen bondwith the carboxylated functional groups of MWCNTs-MIP. Anotherfact is that the sonication technique is applied during the func-tionalization and in the dispersion process. It moreover gives en-ergetic effect in getting the MWCNTs bundles and starts to lose.Upon refluxing, the ultrasonic waves were produced during thewater vibration; it has efficiently slowed down the disturbance onMWCNTs structure. So far it is stabilized and the MWCNTs wereidentified to be colloidal in aqueous media.
Fig. 11. Dynamic response for the sensing of DA of MWCNTs-MIP and MWCNTs-NIPsensor, A (10�6 M), B (10�5 M), C (10�4 M), D (10�3 M), E (10�2 M).
Fig. 12. Photograph demonstrating the dispersion of MWCNTs (a), MWCNTs-MIP (b)and (c) MWCNTs-MIP after six months.
T.S. Anirudhan et al. / Polymer 55 (2014) 4820e48314830
6. Conclusions
In the present study, a novel molecular imprinted polymerbased electrode was developed for potentiometric detection of DA.Compared with the non-imprinted polymers, the MIP matrixdemonstrated much higher binding capacity. The pH study ofMWCNTs-MIP, demonstrated that the adsorption of DA was higherat pH 7.0. The potentiometric method employing the molecularimprinting technology provided an attractive alternative for the DAassessment. The potentiometric method is promising for the spe-cific, rapid and simple detection of DA in blood serum and urinesamples. The prepared MWCNTs-MIP possesses long-term stabilityfor more than 2 months. Even though the modification of theMWCNTs is very complicated and time-consuming, once the ma-terial is prepared it is stable for a long time (2 months). Besides thisfor sensing purposes, a minute quantity is required and it can bereused for several times. The MWCNTs-MIP demonstrated goodpracticability, selectivity, stability and regeneration. In thecompetitive adsorption environment, it can be seen that theMWCNTs-MIP prepared can display highly specific recognition ofDA in presence of interfering ions. Furthermore, the modifiedelectrode can also be used to determinate the concentration of DAwith a linear range from 10�9 to 10�5 mol L�1. All these propertiessuggest its application in the clinical laboratories in the determi-nation of DA.
Acknowledgments
The authors are thankful to the Head, Department of Chemistry,University of Kerala, Trivandrum, for providing laboratory facilities.
References
[1] Ali SR, Ma Y, Parajuli RR, Balogun Y, Lai WYC, He H. A nonoxidative sensorbased on a self-doped polyanline/carbon nanotube compo-site for sensitiveand selective detection of the neurotransmitter dopamine. Anal Chem2007;79:2583e7.
[2] Engelborghs S, Marescau B, De Deyn PP. Amino acids and biogenic amines incerebrospinal fluid of patients with Parkinson’s disease. Neurochem Res2003;28:1145e50.
[3] Holdiness MR, Rosen MT, Justice JB, Neill DB. Gas chromatographic-massspectrometric determination of dopamine in subregions of rat brain.J Chromatogr 1980;198:329e36.
[4] Takeuchi T, Murase N, Maki H, Mukawa T, Shinmori H. Dopamine selectivemolecularly imprinted polymers via post-imprinting modification. Org BiomolChem 2006;4:565e8.
[5] Deftereos NT, Calokerinos AC, Efstathiou CE. Flow injection chemilumino-metric determination of epinephrine, norepinephrine, dopamine and L-dopa.Analyst 1993;118:627e32.
[6] Maminski M, Olejniczak M, Chudy M, Dybko A, Brz�ozka Z. Spectrophotometricdetermination of dopamine in microliter scale using microfluidic systembased on polymeric technology. Anal Chim Acta 2005;540:153e7.
[7] Zhang MS, Huang JR, Yu P, Chen X. Preparation and characteristics of proteinmolecularly imprinted membrance on the surface of multiwalled carbonnanotubes. Talanta 2010;81:162e6.
[8] Haupt K, Mosbach K. Molecularly imprinted polymers and their use in bio-mimetic sensors. Chem Rev 2000;100:2495e504.
[9] Vlatakis G, Andersson LI, Müller R, Mosbach K. Drug assay using antibodymimics made by molecular imprinting. Nature 1993;361:645e7.
[10] Gao BJ, Niu QY, Du RK. Preparation and recognition performance of cytisinealkaloid-imprinted material prepared using novel surface molecularimprinting technique. J Sep Sci 2010;33:1338e48.
[11] Kan X, Zhao Y, Geng Z, Wang Z, Zhu JJ. Composites of multiwalled carbonnanotubes and molecularly imprinted polymers for dopamine recognition.J Phys Chem C 2008;112:4849e54.
[12] Komathi S, Gopalan AI, Lee KP. Nanomolar detection of dopamine at multi-walled carbon nanotube grafted silica network/gold nanoparticle functional-ised nanocomposite electrodes. Analyst 2010;135:397e404.
[13] Zhang P, Wu FH, Zhao GC, Wei XW. Selective response of dopamine in thepresence of ascorbic acid at multi-walled carbon nanotube modified goldelectrode. Bioelectrochemistry 2005;67:109.
[14] Zhang MN, Gong KP, Zhang HW, Mao LQ. Layer-by-layer assembled carbonnanotubes for selective determination of dopamine in the presence ofascorbic acid. Biosens Bioelectron 2005;20:1270.
[15] Meng H, Sui GX, Fang PF, Yang R. Effects of acid- and diamine-modifiedMWNTs on the mechanical properties and crystallization behavior of poly-amide 6. Polymer 2008;49:610e20.
[16] Panasyuk-Delaney T, Mirsky VM, Ulbricht M, Wolfbeis OS. Impedometricherbicide chemosensors based on molecularly imprinted polymers. Anal ChimActa 2001;435:157e62.
[17] Suedee R, Intakong W, Dickert FL. Molecularly imprinted polymer-modifiedelectrode for on-line conductometric monitoring of haloacetic acids in chlo-rinated water. Anal Chim Acta 2006;569:66e75.
[18] Gomez-Caballero A, Unceta N, Goicolea MA, Barrio RJ. Voltammetric deter-mination of metamitron with an electrogenerated molecularly imprintedpolymer microsensor. Electroanalysis 2007;19:356e63.
[19] Blanco-L�opez MC, Lobo-Casta MJ,en�on AJ, Miranda-Ordieres P. Tuen�on-Blanco,Electrochemical sensors based on molecularly imprinted polymers. TrendsAnal Chem 2004;23:36e48.
[20] Tasis D, Tagmatarchis N, Bianco A, Prato M. Chemistry of carbon nanotubes.Chem Rev 2006;106:1105e36.
[21] Jana U, Maiti S, Biswas S. An efficient FeCl3-catalyzed amidation reaction ofsecondary benzylic and allylic alcohols with carboxamides or p-toluene-sulfonamide. Tetrahedron Lett 2008;49:858e62.
[22] Kannan P, John SA. Ultrasensitive detection of L-cysteine using gold-5-amino-2-mercapto-1, 3, 4-thiadiazole core-shell nanoparticles film modified elec-trode. Biosens Bioelectron 2011;30:276e81.
[23] Weng Q, Xu G, Yuan K, Tang P. Determination of monoamines in urine bycapillary electrophoresis with field-amplified sample stacking and ampero-metric detection. J Chromatogr B 2006;835:55e61.
[24] Kuo SW, Kao HC, Chang FC. Thermal behavior and specific interaction in highglass transition temperature PMMA copolymer. Polymer 2003;44:6873e82.
[25] Li YS, Vecchio NE. Studies of thiol additions of silane coupling agents byvibrational spectroscopy. Spectrochim Acta, Part a 2007;67:1321.
[26] Jorio A, Pimenta MA, Filho AGS, Saito R, Dresselhaus G, Dresselhaus MS.Characterizing carbon nanotube samples with resonance Raman scattering.New J Phys 2003;5:1391.
[27] Hong CY, You YZ, Pan CY. Functionalized multi-walled carbon nanotubes withpoly(N(-2-hydroxypropyl)methacrylamide) by RAFT polymerization. J PolymSci Part a: Polym Chem 2006;44:2419e27.
[28] Kastner J, Pichler T, Kuzmany H, Curran S, Blau W, Weldon DN, et al. Reso-nance raman and infrared spectroscopy of carbon nanotubes. Chem Phys Lett1994;221:53e8.
[29] Lee NS, Hsieh YZ, Paisley RF, Morris MD. Surface-enhanced Raman spectros-copy of the catecholamine neurotransmitters and related compounds. AnalChem 1988;60:442e6.
[30] Herlinger E, Jameson RF, Linert W. Spontaneous autoxidation of dopamine.J Chem Soc Perkin Trans 1995;2:259e63.
[31] Buck RP, Lindener E. Recomendations for nomenclature of ion-selectiveelectrodes. Pure Appl Chem 1994;66:2527e36.
[32] Mary W, Abhinav R, Christopher O. Polymer brushes for electrochemicalbiosensors. Soft Matter 2011;7:297e302.
[33] Roongnapa S, Vatcharee S, Bhutorn C, Chitchamai O, Gary PM. Temperaturesensitive dopamine-imprinted (N,N-methylene-bis-acrylamide cross-linked)polymer and its application to the selective extraction of adrenergic drugsfrom urine. J Chromatogr A 2006;114:239e49.
[34] Agnieszka P, Subramanian S, Wlodzimierz K, Eranda M, Melvin EZ, Francis DS.Molecularly imprinted poly[bis(2,20-bithienyl)methane] film with built-inmolecular recognition sites for a piezoelectric microgravimetry chemosensorfor selective determination of dopamine. Bioelectrochemistry 2010;80:62e72.
[35] Gomez-Caballero Alberto, Ana U, Sanchez-Ortega Alicia, Nora UM,Aranzazu G, Ramon JB. Molecularly imprinted poly[tetra(o-aminophenyl)porphyrin] as a stable and selective coating for the development of voltam-metric sensors. J Electroanal Chem 2010;638:246e53.
T.S. Anirudhan et al. / Polymer 55 (2014) 4820e4831 4831
[36] Dhana L, Alessandra B, Michael JW, Iva C, Steven AF, Sreenath S, et al. Elec-trochemical sensor for catechol and dopamine based on a catalytic molecu-larly imprinted polymer-conducting polymer hybrid recognition element.Anal Chem 2009;81:3576e84.
[37] Bhim BP, Darshika J, Mahavir PT. A dual-template imprinted polymer-modified carbon ceramic electrode for ultra trace simultaneous analysis ofascorbic acid and dopamine. Biosens Bioelectron 2013;50:19e27.
[38] Bhim BP, Deepak K, Rashmi M, Mahavir PT. Solegel derived multiwalledcarbon nanotubes ceramic electrode modified with molecularly imprintedpolymer for ultra trace sensing of dopamine in real samples. ElectrochimicaActa 2011;56:7202e11.
[39] Bakker E, Buhlmann P, Pretsch E. Carrier-based ion-selective electrodes andbulk Optodes. 1. General Charact Chem Rev 1997;97:3083e132.
[40] Umezawa Y, Buhlmann P, Umezawa K, Tohda K, Amemiya S. Potentiometricselectivity coefficients of ion-selective electrodes. Pure Appl Chem 2000;72:1851e2082.
[41] Mehran J, Solmaz EF, Ali M, Majid A, Mohammad RG, Parviz N, et al. Molec-ularly imprinted polymer based potentiometric sensor for the determinationof hydroxyzine in tablets and biological fluids. Anal Chim Acta 2008;612:65e74.
