a b s t r a c t

A novel potentiometric sensor based on a molecularly imprinted polymer (MIP) for determination ofpromethazine (PMZ) was prepared. Promethazine MIP particles were prepared and dispersed in 2-nitrophenyloctyl ether and then embedded in a polyvinyl chloride matrix. The effect of the monomerstype on the sensor performance was investigated, and an important role for this parameter was shown. Itwas shown that the membrane electrode with a MIP prepared by vinylbenzene and divinylbenzene had abetter performance in comparison to membrane electrodes containing MIPs prepared with methacrylicacid-ethylene glycol dimethacrylate or vinylbenzene-ethylene glycol dimethacrylate. After optimization,the membrane electrode constructed with a MIP of vinylbenzene-divinylbenzene exhibited a Nernstianresponse (31.2±1.0mVdecade−1) over a wide concentration range, from 5.0×10−7 to 1.0×10−1 M, witha low detection limit of 1.0×10−7Mand a response time of ∼50 s. The method has the requisite accuracy,sensitivity and precision to assay PMZ in syrup samples and biological fluids.

1. IntroductionMolecularly imprinted polymers (MIPs) are promising materialsfor continual use in sensor fields as the recognition element ormodifying agent.AnMIP is a synthetic polymer possessing selectivemolecular recognition properties for the shape and positioning offunctional groups because of its recognition sites within the polymermatrixthat are complementary to the analyte molecule. Thesematerials are similar to biological specific receptors in some waysbecause of their high selectivity to the target molecule and theirrecognition mechanism [1,2]. MIPs have been used for electrochemicalsensor development as the highly selective recognitionelement of the sensor [3–6].Promethazine is widely used for its antihistaminic, sedative,antipsychotic, analgesic and anticholinergic properties. However,promethazine hydrochloride can cause adverse effects in humans,such as endocrinal, cardiac and reproductive alterations. Therefore,its determination in commercial formulations and biologicalsamples is extremely important [7].Many analytical techniques such as titrimetric procedures[8–10], spectrophotometric methods [11], spectrofluorometry [12],high performance liquid chromatography [13] and voltammetry[14] have been employed for promethazine (PMZ) determination.Potentiometry is one of the simplest instrumental techniquesthat many chemists encounter. Potentiometric sensors provide an exciting and achievable opportunity to perform biomedical,environmental and industrial analyses away from a centralized laboratorybecause they make it possible to combine the ease of useand portability of potentiometry with simple, inexpensive fabricationtechniques. While potentiometry has been used for manyyears, the advances in the field of ion-selective electrodes make ita valuable technique in the modern laboratory [15,16].The unique feature of potentiometry with MIP-based sensorsis that the species do not have to diffuse through the membraneso there is no size restriction on the template compound. Despiteall these advantages, only a few MIP-based sensors have beenreported that utilize a potentiometric transducer [17–19]. Thesestudies describe potentiometric sensors created in several differentways: by dispersing MIP particles in plasticizer and embeddingthem in a polyvinyl chloride (PVC) matrix [20–23], by forming aglassy membrane [24], by assembling the template on the polarsurface of an indium tin oxide (ITO) glass plate [25,26], by depositinga MIP polymeric film on the gate surface of an ion-sensitivefield-effect transistor [27,28] and by embedding MIPs in the carbonpaste electrode [29].The potentiometric sensor that has already been reported forpromethazine determination [30,32] usually uses the ion-pairing

agent as the ionophore, which suffers from the main disadvantageof low selectivity toward the target molecule. The use of a MIP asthe ionophore in the membrane electrode for promethazine determinationwould be an interesting development in this field becauseit would provide improved selectivity in the developed sensor.In this work, a molecularly imprinted polymer with recognitionsites for promethazine was prepared and then used to fabricate the promethazine-selective potentiometric sensor, creating the firstMIP-based promethazine sensor. It was found that the MIP compositionas determined by the nature of the monomers used for theMIPpreparation had a considerable effect on the final sensor performance.After optimization of the parameters influencing the sensorperformance, the sensor was successively used for promethazinedetermination in pharmaceutical products and serum samples.

2. Experimental2.1. ReagentsPromethazine hydrochloride and clozapine were obtained fromFluka (Switzerland). Methacrylic acid (MAA), 4-vinylpyridine (VB),divinylbenzene (DVB), ethylene glycol dimethacrylate (EGDMA), 2,2-azobisisobutyronitrile (AIBN), 2-nitrophenyl octyl ether(NPOE), di-n-octylphthalate (DOP), dibutylphthalate (DBP), bis(2-ethylhexyl) sebacate (BEHS) and high-molecular-mass poly(vinylchloride) (PVC) were purchased from Aldrich (USA). All otherchemicals were of analytical reagent grade and were obtainedfrom Merck (Germany). Deionized water was used throughout.Drug-free human serum was obtained from the Iranian bloodtransfusion service (Ardabil, Iran) and stored at −20 ◦C until useafter gentle thawing.

2.2. Preparation of PMZ-imprinted and non-imprinted polymerparticlesPMZ-imprinted polymer particles were prepared by taking1mmolof PMZand 7mmolof functional monomer (MAA or VB) in a50-ml round-bottom flask. The mixture was then left in contact for5min for prearrangement. Subsequently, 32mmol of cross-linker(EGDMA or DVB) and AIBN (0.1 g) were added. The mixture waspurged withN2 for 10 min and the flask was sealed under this atmosphere.It was then kept stirring in an oil bath maintained at 60 ◦C tostart the polymerization process. After 12 h, the obtained polymermaterials were ground and sieved. The PMZ and unpolymerizedmonomers were removed by Soxhlet extraction with 100 ml ofmethanol by refluxing for 12 h. Then, the particles were suspendedin acetone and allowed to settle for 4 h. The sedimented particleswere discarded and those not sedimented were collected by centrifugation.The particles collected were suspended in acetone againand allowed to settle for 4 h, followed by centrifugation. This procedurewas repeated four times. The resulting MIP particles weredried to a constant weight under vacuum at 60 ◦C and were used inthe following experiments. Non-imprinted polymer (NIP) particleswere prepared analogously without the addition of PMZ during thepolymer material preparation.

2.3. Fabrication of the PMZ sensorThe PVC membrane sensors were fabricated by followingthe general procedure described below. PMZ-imprinted or nonimprintedpolymer particles were dispersed in NPOE (DOP orBEHS or DBP) and were added to 2.5 ml of tetrahydroforan (THF)-containing PVC. The resulting solution was homogenized in asonicator and then poured into a Teflon mold. The THF was allowed

to evaporate at room temperature. The polymer membranes thusobtained had a thickness of ∼0.5mm. The membranes were gluedto one end of a glass tube. The tube was then filled with an internalfilling solution of 10−3M of PMZ. The sensor was kept in air whennot in use.

2.4. Analytical procedureThe sensor was conditioned in 25 ml of 0.1Mbuffer with a pH of2.5 for 3 h. The response of the sensor was examined by measuringthe electromotive force (EMF) of the following electrochemical cell:Ag, AgCl, 1.0×10−3M PMZ|PVC membrane|sample solution||KCl(saturated)|Hg2Cl2, Hg. The potential response of the sample solutioncontaining varying amounts ofPMZin 50ml of 0.1Mbuffer (pH2.5) was measured. The measurements were conducted from lowto high concentration of PMZ. The EMF was plotted as a function ofPMZ concentration.

2.5. Syrup sample preparation and determinationSyrup containing 5mg/ml of promethazine–HCl was dilutedwith distilled water. An aliquot containing 1×10−6 to 1×10−2Mwas taken, the above procedure was followed and the membranepotentials were measured. The quantity of promethazine–HCl per ml of syrup was calculated from the standard calibration graph andrequired calculations.

2.6. Preparation of serum samples and extraction procedureFor the preparation of serum standard solutions, 1ml of PMZaqueous solution was transferred to a 5 ml volumetric flask, andthen the solution was diluted to the mark with urine and vortexedfor 1min. Then, the solution was adjusted to a pH of 10by the addition of a concentrated sodium hydroxide solution. Forthe determination of PMZ, 2ml dichloromethane was added to1ml of serum samples and vortexed for 2min. The mixture wascentrifuged at 1000rpm for 3min to separate the aqueous andorganic layers. After removal of the organic layer, the extractionwas repeated on the residual aqueous layer. The dichloromethanelayers were pooled and dried at 40 ◦C under a gentle stream of nitrogen.After drying, the samples were reconstituted with 15 ml ofbuffer. Then, the analysis was conducted, as indicated in the generalanalytical procedure. The calibration curve for serum sampleswas also prepared using buffer solution.

3. Results and discussion3.1. Primary evaluation of MIP and NIP particles prepared withdifferent compositionThree different monomer compositions were used for preparationof MIP and NIP particles under identical synthesisconditions and at the same mole ratios of cross-linker/functionalmonomer/template.Fig. 1 shows the scanning electron microscopy (SEM) images ofthe MIPs and NIPs having structures of MAA-EGDMA, VB-EGDMAand VB-DVB. It can be seen that in all three polymers there is nodifferences between the MIP and NIP with the same structure withregards to the size and surface morphology of the polymeric particles.In addition, it is clear that the MIPs and NIPs with the VB-DVBstructure have a larger size and a different morphology in comparisonto the other synthesized polymers. The PMZ adsorption capabilities of MIPs and NIPs created fromthe three polymeric structures were also investigated. In this study,a simple batch extraction procedure coupled with a colorimetricdetection method [33] was applied. For this procedure, 5×10−5M

of PMZ solution was prepared and 0.05 g ofMIP or NIP powder wascontacted with PMZ solution for 30 min with stirring. Then, theMIP or NIP was separated from the solution and the non-adsorbedPMZ was measured in the solution by the colorimetric method. Theresults obtained are illustrated in Fig. 2. It is evident that in all polymericstructures, the adsorption capability of the MIP is better thanthat of the NIP. Regarding the similarity between the MIP and theNIP according to the SEM images, it can be concluded that the presenceof recognition cavities in the MIP is the main reason for thehigher adsorption capability of the MIP over the NIP in all polymericstructures. It can be seen that both the MIP and NIP preparedwith MAA and EGDMA adsorb the PMZ more than the two otherpolymers. However, washing of the MIP and the NIP with a properwashing solution (water/ethanol, 70:30, v/v), removed 11% of theadsorbed PMZ from the MIP prepared with VB and DVB. For thesame conditions, the washing could remove 46% and 55% of thePMZ adsorbed on the MIP made of VB-EGDMA and MAA-EGDMA,respectively (shown in the inset of Fig. 2). In addition, it was foundthat most of the PMZ adsorbed on the NIP was removed by washingfor all the polymeric structures. Because the washing mainlyremoved the PMZ molecules adsorbed on the polymer surface (socalledsurface adsorption), it can be concluded that a considerableportion of the PMZ uptaken by the MAA-EGDMA polymer occursbecause of a weak surface adsorption mechanism. The high surfaceadsorption capability of MAA-EGDMA-based MIPs and NIPs can beattributed to the smaller size of these polymers and the presenceof carboxylic acid functional groups in the polymer structure thatseem to be able to interact strongly with promethazine molecules.It must be mentioned that the amount of PMZ adsorbed by recognitioncavities on the MIP in the case of the MAA-EGDMA-basedpolymer was also high because, even after the washing step, thePMZ remaining on the MAA-EGDMA MIP was still larger than forother types of MIPs at the same conditions. For the VB-DVB MIP,although the total adsorbed amount of PMZ was low, most of thePMZ uptaken was fixed at recognition cavities. This conclusion wasmade by comparison of the amount of PMZ present before and afterthe washing step.The binding energy (_E) of PMZ with MAA and VB was calculatedin order to theoretically evaluate the interaction between the PMZ and the polymers prepared by monomers of MAA andVB. The higher the template–monomer binding energy, the moreintensive the template interaction with the polymer backboneand also with polymer-selective sites containing that monomer.For the computational approach, PMZ was chosen as the templatemolecule and MAA and VB were chosen as the functionalmonomers. The _E of the interactions of the template withthese monomers was calculated by the density functional theorymethod at the B3LYP/6-31G level (Gaussian 98 software) afterpreliminary energy minimization by molecular mechanics. Thecalculation was carried out in the gas-phase state. The bindingenergy of PMZ with the monomer was obtained from the followingequation:_E = E(PMZ-monomer) − E(monomer) − E(PMZ).The calculated energies for the PMZ-MAA and PMZ-VB interactionswere equal to 176.518 and 11.549 kJ/mol, respectively. The optimizedconformations of PMZ and the mentioned monomers areshown in Fig. 3. From these results, it can be concluded that PMZcan interact more strongly with the MAA-based polymer as comparedto the VB-based polymer. Because the mentioned functionalgroups exist on both the surfaces and the selective cavities of the

resulting polymers, it can be concluded that the polymer based onthe MAA monomer uptakes PMZ more intensively than the polymerwith the VB structure. This theoretical finding is in agreementwith the previously explained adsorption results and the followingpotentiometric results.3.2. Effect of polymer structure on the potential responses ofmembrane sensorsThe plots of the potential responses versus the decade of PMZconcentrations for the three described MIP- and corresponding NIPbasedmembranes are comparatively illustrated in Fig. 4. As can beseen, in the case of MAA-EGDMA, the responses of the MIP- andNIP-based electrode to PMZ are identical for all the tested PMZconcentrations. It is also clear that in the case of VB-EGDMA, thereis a similarity in the responses of the MIP- and NIP-based electrodesto PMZ. However, in this case, partial distinction of the MIPin comparison to the NIP is possible. However, for VB-DVB, thereis a noticeable difference between the MIP- and NIP-based membranesin response to PMZ. Here, the MIP-based sensor responds tothe PMZ in a near-Nernstian manner, whereas the NIP-based sensor’s response to the PMZ is not noticeably regular for most testedvalues of PMZ concentration.As was described above, in the polymer with the MAA-EGDMAcomposition, both the MIP and the NIP adsorbed PMZ intensively.It was also proved that most of this adsorption occurred becauseof surface adsorption, which generally has a non-selective nature.Although the washing removed most of the PMZ from the MIP withthe MAA-EGDMA structure, the PMZ remaining on the MIP afterwashing was still high, indicating the strong recognition capabilityof the MIP. However, in the potentiometric method, unlike otherelectrochemical methods such as voltammetry [3,4], it was notpossible to differentiate the surface-adsorbed PMZ from PMZ recognizedby the selective recognition sites present on the surface orin the proximity of the MIP surface. In the mentioned voltammetricsensor, the use of a simple washing step before the electrochemicaldetermination and after the extraction step could remove theweakly adsorbed analyte (surface adsorption) from the MIP. In thepotentiometric sensor, PMZ is not required to diffuse through amembrane and the PMZ ions adsorbed on the surface (by simplesurface adsorption and by the recognition sites on the MIP surfaceor those in the proximity of the surface) are responsible for themembrane potential creation. For this reason, it can be seen thatthe MAA-EGDMA NIP responded to the PMZ concentration changelike a MIP. However, in the case of the VB-DVB MIP, the surface adsorption was not large, and most of the PMZ adsorbed to the MIPwas present because of the selective recognition of PMZ by recognitionsites on the MIP. This led to noticeable differences betweenthe MIP- and NIP-based membranes in response to PMZ.As a general conclusion from this study, it can be said that theresponse of the MIP electrode originates from two sources: selectivesite interactions (presented as cavities on the surface of theMIP particles) and the surface adsorption effect. Thus, if the surfaceadsorption capability of the MIP for a target molecule is high,the membrane potential response to the analyte will be intensivelyaffected by the surface adsorption, and the selective sites of theMIP will not properly show their recognition capabilities. Becauseof the nonselective nature of the surface adsorption, the response ofthe electrode under such conditions is not selective enough towardPMZ. Because the NIP contains no predesigned selective cavities forPMZ and the only way for the NIP to uptake PMZ is through surfaceadsorption, the large similarity in the responses of MIP- and

NIP-based MAA-EGDMAmembranes can be attributed to the dominanceof the surface adsorption over the selective site interactions.The following selectivity study was conducted to prove thisstatement. Fig. 5 shows the potential response behaviors of threemembrane electrodes prepared by using different MIP structuresfor chloropromethazine, chosen as a similar compound to promethazine.It can be seen that the MAA-EGDMA-based sensor could not differentiate these two molecules, whereas the electrode preparedwith VB-DVB demonstrated considerable selectivity to PMZ overchloropromethazine for most of the tested concentrations. As mentionedbefore, this can be attributed to the low surface adsorptionin the case of VB-DVB-based MIPs.Taking into account the demonstrated results and consideringthe differences between the potentiometric sensors prepared withMIPs and NIPs of VB-DVB polymer, we selected this structure forfurther investigation.3.3. Response of MIP particles (VB-DVB) to PVC ratioIt has been proved that the ratio of ionophore to PVC influencesthe working concentration range, slope and response timein the cases of conventional ionophore-based sensors [34,35] andimprinted, polymer-based, ion-selective electrodes [23,24,36].Weobserved that the ratio ofPVCto imprinted polymer particlesplayed a key role in the efficiency of the sensors because the amountof imprinted polymer particles determines the number of bindingsites available for recognition. Hence, a study was conducted withMIP- and NIP-based sensors and the results obtained are shown inTable 1. From the presented results, it is clear that the membranewith aMIP particle toPVCratio of 1:2 gave the best response. It mustbe mentioned that in order to select the best ratio, the considerableobserved difference between the MIP- and NIP-based sensorsis necessary, in addition to the proper response of the MIP-basedsensor. Both criteria can be seen in the MIP to PVC ratio of 1:2. 3.4. Effect of membrane compositionThere are a great number of reports on conventional potentiometricsensors that show that the response behavior of the sensordepends on various features of membranes such as the propertiesof the plasticizer and the nature and amount of ion recognizingmaterial used [23,37]. Thus, different aspects of the membranepreparation using PMZ-imprinted polymer particles were optimizedalong similar lines.The addition of an appropriate plasticizer leads to the properphysical properties and ensures high mobility of thePMZions in themembrane.These solvent mediators strongly influence the workingconcentration range of potentiometer sensors [36].The effect of different plasticizers—NPOE, BEHS, DBP andDOP—on the performance of the PMZ sensor was investigated. Itwas found that using BEHS, DBP and DOP as plasticizers resulted incalibration curve slopes of 47.5 (±2.2), 50.8 (±1.7) and 53.3 (±3.3),respectively. These slopes were larger than that of the membraneprepared withNPOE(31.2 (±1.0)). However, the NPOE-basedmembraneshowed a lower detection limit and a wider linear range. Inaddition, the difference between the MIP-based membrane and theNIP-containing membrane was very noticeable when NPOE wasapplied as a plasticizer. For the other mentioned plasticizers, theresponse difference between the MIP- and NIP-based membraneswas not as large.Overall, the PMZ sensor gave a narrower linear response rangeand poor detection limits with plasticizers like BEHS, DBP andDOP, in comparison to NPOE. The appropriate characteristics of the membrane containing NPOE can be directly related to the higher

dielectric constant of the NPOE plasticizer [23]. Thus, NPOE wasselected as the plasticizer. Afterwards, a study was conducted tofind the proper amount of NPOE. It was found that using 0.2 mlof NPOE in the membrane led to better physical properties andresulted in optimum analytical characteristics such as a wider linearrange, a lower detection limit and a Nernstian slope.It was also found that PMZ membranes were brittle in theabsence of plasticizer and could not be used for recording sensorperformance.3.5. Effect of test solution pHThe effect of the pH of the solution on the performance of thePMZ sensor was studied by varying the pH in the range of 1.0–6.0.The results are illustrated in Fig. 6. As can be seen, the potentialsremained constant in the range of 2.0–5.0. The observed potentialdrift at lower pH values may be attributed to the membraneresponse to H+ and at higher pH values (pH > 5) could be due tothe change of the promethazine ionic charge. Therefore, a pH of2.5, fixed with monochloroacetic acid–based buffer, was chosen forfurther promethazine determination by the proposed sensor.3.6. Response time and memory effect evaluationIn Fig. 7, the resulting potential–time responses of themembrane electrode prepared with VB-DVB are presented. Thepotential–time response behaviors were obtained upon changingthe promethazine concentration from 1.0×10−6 to 1.0×10−4M(by fast injection of _l amounts of a concentrated solution; the first raising portion of the curve). This was followed by changing thePMZ concentration from 1.0×10−4 to 1.0×10−3Mat the next step(the second raising portion of the curve). Afterwards, in order tocheck thememoryor hysteric effect of the sensor, the concentrationwas reduced from 1.0×10−3 to 1.0×10−4M (by appropriate dilutionof the solution; the first descending portion of the curve). Theconcentration was further reduced from 1.0×10−4 to 1.0×10−6Mat the next step (the second descending portion of the curve). Itis evident that the potentiometric response of the electrode preparedwith MIP particles was rapid (50 s) and reversible, althoughthe time needed to reach the equilibrium value for the case of high-to-low sample concentration was longer than that of the lowto-high sample concentration because of the filling of the cavityin the imprinted polymer with the target molecule. However, themeasurements performed in the sequence of high-to-low concentrationindicate that the response of the MIP-based electrodes wasreversible. In addition, it is evident that in spite of the requiredlonger time (120–145 s) needed to return the potential to its initialvalue in the dilution step, no unwanted memory or hysteric effectwas found for the sensor.3.7. Sensitivity and detection limitThe potential responses of PMZ-imprinted and control polymermembrane sensors fabricated under optimal conditions, asobtained from the above studies, were checked. The calibrationplot obtained from MIPs and NIPs of VB-DVB polymer (based ontriplicate measurements at each concentration) is shown in Fig. 8.The linear working concentration plot (shown in the inset of Fig. 7)demonstrated a linear response in the range from 5×10−7 to1×10−1 M. The limit of detection was calculated to be 1×10−7Mbased on the IUPAC definition. The slope of the obtained calibrationcurve, 31.2 mV, indicates that the sensor obeys the Nernstianlaw and also proves that the [PMZ]2+ species of promethazineare mainly responsible for creating the potential responses. Onthe other hand, the non-imprinted membrane did not respondto PMZ below 1×10−4M and gave a linear response for PMZ

in the concentration range from 1×10−3 to 1×10−1 M. The betterresponse characteristics of the PMZ-imprinted sensor over thenon-imprinted polymer-based sensor over the entire concentrationrange are attributed to a significant imprinting effect.3.8. Interference studyThe potentiometric selectivity coefficients were measured bythe matched potential method (MPM) [38]. The coefficients describe the preference of the developed membrane electrode foran interfering ion, X, with reference to the promethazine ion.According to the MPM method, the specified activity (concentration)of the primary ions is added to a reference solution(1.0×10−6M promethazine), and the potential is measured. Inanother experiment, the interfering ions (X) are successively addedto an identical reference solution until the measured potentialmatches that obtained before the addition of the primary ions. TheMPM selectivity coefficient, KMPMPMZ,X, is then given by the resultingprimary ion activity (concentration) to the interfering ion activityratio:KMPMPMZ,X= apmzaX.The MPM selectivity coefficients for the promethazine ionselectiveelectrode prepared with MIP and its correspondingNIP at the constant pH value of 2.5 are listed in Table 2. Whenthe MIP sensor is used to measure promethazine, all the othersubstances (except for chloropromethazine) hardly interferewith the determination. In most cases, the selectivity coefficientswere small enough to cause interference in the promethazinedetermination. The MPM selectivity sequence of the employedMIP for different drugs and organic materials approximatelyobeys the following order: chloropromethazine > methyleneblue > clozapine > hydroxyzine >metchachlorpramide > pyrrole >aniline. For membranes constructed with NIP, the obtained selectivitiesfor the tested compounds were noticeably lower thanthose of the MIP-based membrane electrode. This again proves thepresence and proper functioning of selective cavities in the MIP.3.9. Stability and reusabilityTwo important criteria required for any sensing device, in additionto sensitivity and selectivity, are stability and reusability. Theabove-developed PMZ sensor was found to be stable (deviation lessthan 1mV for 1×10−4Mof PMZ) for 2months and could be reusedmore than 10 times without any loss in sensing ability.3.10. Accuracy and reproducibilityThe accuracy of the described potentiometric measurementswas checked by calculating the recovery of a known promethazineconcentration (1×10−5 M). The mean percentage recovery,obtained by applying the calibration curve method, was 103.3%(n = 5).The reproducibility of the sensor was evaluated with fiverepeated potentiometric measurements of the 1.0×10−5Mpromethazine solution. The precision of the described procedurein terms of relative standard deviation was 5.3%.

3.11. Analytical application3.11.1. Analysis of promethazine in syrup sampleThe proposed potentiometric procedure was successfullyapplied for promethazine determination in syrup samples. The

resulting data obtained from the calibration curve procedure werestatistically compared with the labeled amounts on the syrupand those obtained by the spectrophotometric method [33]. Thismethod is based on the oxidation of promethazine by potassiumpersulfate and measurement of the absorbance of the resultingcolored product at a wavelength of 510nm (_ max). The resultsare presented in Table 3. As can be seen, satisfactory results wereobtained from the proposed sensor.3.11.2. Promethazine assay in spiked human serumThe proposed potentiometric procedure was also successfullyapplied to an assay of promethazine in spiked human serum. Theresults of the recovery studies are listed in Table 4. The recoveriesof the methods were in the range of 96–110% for the spiked serum.Consequently, it was concluded that the suggested method wassensitive and precise.4. ConclusionThe potentiometric sensor employing a MIP as the ionophorein a PVC membrane electrode provided an attractive alternativefor the measurement of promethazine. In this work, we investigatedthe effect of MIP composition on the potentiometric sensorperformance. It was shown that the presence of selective cavitiesin the MIP structure was not enough of a criterion for properfunctioning of the resulting sensor. It was shown that properattention to the polymer structure and the surface adsorption onMIPs can be effective in appropriately designing and preparingpotentiometric sensors that utilize MIPs. The potentiometric MIPsensor prepared with VB-DVB showed a considerable differencein response to promethazine in comparison to the corresponding NIP-based sensor. The described sensor introduces a new strategyfor constructing potentiometric chemosensors for specific, rapidand simple promethazine detection in pharmaceutical formulationsand human blood. Its accuracy, reproducibility, simplicity andselectivity suggest its application would be appropriate in qualitycontrol analysis and clinical laboratories