Improved Durability and Sensitivity of Bitterness-Sensing ... - MDPI

sensors Article

Improved Durability and Sensitivity of Bitterness-Sensing Membrane for Medicines Xiao Wu 1, *, Hideya Onitake 1 , Zhiqin Huang 1 , Takeshi Shiino 1 , Yusuke Tahara 2 , Rui Yatabe 2 , Hidekazu Ikezaki 3 and Kiyoshi Toko 1,2 1

2 3

*

Graduate School of Information Science and Electrical Engineering, Kyushu University, 744 Motooka, Nishi-ku, Fukuoka 819-0395, Japan; [email protected] (H.O.); [email protected] (Z.H.); [email protected] (T.S.); [email protected] (K.T.) Research and Development Center for Taste and Odor Sensing, Kyushu University, 744 Motooka, Nishi-ku, Fukuoka 819-0395, Japan; [email protected] (Y.T.); [email protected] (R.Y.) Intelligent Sensor Technology, Inc., 5-1-1 Onna, Atsugi-shi, Kanagawa 243-0032, Japan; [email protected] Correspondence: [email protected]; Tel.: +81-92-802-3762

Received: 19 September 2017; Accepted: 2 November 2017; Published: 4 November 2017

Abstract: This paper reports the improvement of a bitterness sensor based on a lipid polymer membrane consisting of phosphoric acid di-n-decyl ester (PADE) as a lipid and bis(1-butylpentyl) adipate (BBPA) and tributyl o-acetylcitrate (TBAC) as plasticizers. Although the commercialized bitterness sensor (BT0) has high sensitivity and selectivity to the bitterness of medicines, the sensor response gradually decreases to almost zero after two years at room temperature and humidity in a laboratory. To reveal the reason for the deterioration of the response, we investigated sensor membranes by measuring the membrane potential, contact angle, and adsorption amount, as well as by performing gas chromatography-mass spectrometry (GC-MS), liquid chromatography-tandem mass spectrometry (LC-MS/MS). We found that the change in the surface charge density caused by the hydrolysis of TBAC led to the deterioration of the response. The acidic environment generated by PADE promoted TBAC hydrolysis. Finally, we succeeded in fabricating a new membrane for sensing the bitterness of medicines with higher durability and sensitivity by adjusting the proportions of the lipid and plasticizers. Keywords: taste sensor; bitterness sensor; response deterioration; quinine hydrochloride

1. Introduction It is sometimes said that good medicine always tastes bitter. This concept has been reversed with the development of several bitterness-masking techniques. With the quality of life (QOL) of patients attracting increasing attention, especially for pediatric and geriatric patients [1], before marketing an oral bitter medicine, panelists have to take the medicine and evaluate its taste intensity. However, such tests have problems, such as low objectivity and reproducibility, as well as the potential impact of side effects from the medicine. Therefore, objective methods of bitterness evaluation are important for the pharmaceutical industry [2]. The research on taste sensing started in the mid-1990s, before the elucidation of the principle of vertebrate taste receptors [3–5]. Since then, many studies on electronic tongues [6–12] and bioelectronic tongues [13,14] have been carried out for taste assessment. The TS-5000Z taste sensor (Intelligent Sensor Technology Inc., Kanagawa, Japan) is a type of electronic tongue. It is equipped with several sensor electrodes to measure all five basic taste qualities, as well as astringency. Each sensor electrode has a lipid polymer membrane. In accordance with the physicochemical properties of taste qualities, each sensor membrane is designed using different types and amounts of lipids and plasticizers. Sensors 2017, 17, 2541; doi:10.3390/s17112541

www.mdpi.com/journal/sensors

Sensors 2017, 17, 2541

2 of 14

The lipids are mainly used to adjust the density of electric charges on the surface of the membrane while the plasticizers are added to form a membrane with flexibility. Both lipids and plasticizers affect the hydrophobicity of the membrane [4,15–17]. The taste sensor using the lipid polymer membrane has a unique characteristic called ‘global selectivity’, which means that each sensor does not distinguish each chemical substance, but distinguishes taste quality and taste intensity [18,19]. Therefore, this type of taste sensor has the advantage of better selectivity than other electronic tongues, as well as better sensitivity and durability than bioelectronic tongues in bitterness evaluation. In our previous studies, the taste sensor was applied to the objective evaluation of the taste of food and pharmaceutical products (e.g., beer, coffee, traditional Chinese medicines) [19–21]. A BT0 sensor is the sensor electrode of the taste sensor used for the bitterness quantification of medicines. The membrane of the BT0 sensor is composed of phosphoric acid di-n-decyl ester (PADE) as the lipid, bis(1-butylpentyl) adipate (BBPA) and tributyl o-acetylcitrate (TBAC) as the plasticizers [22]. In taste solutions, PADE produces negative charges on the surface of the membrane owing to the ionization of phosphate group, while BBPA and TBAC improve the membrane by increasing its flexibility and hydrophobicity. The composition ratios of PADE, BBPA, and TBAC affect the sensitivity and selectivity to bitter substances [22]. The BT0 sensor possesses good sensitivity and high bitterness sensory correlation to the bitterness of medicines, such as quinine hydrochloride, hydroxylammonium chloride, bromhexine hydrochloride, and so forth [19,23]. In addition, the sensor can evaluate the bitterness-masking effect caused by physical and biochemical masking materials [19]. In our previous study, we also provided an effective method of evaluating the bitterness suppression effect of high-potency sweeteners (functional masking materials) by using electrodes for sensing medicine bitterness and high-potency sweeteners [24–28]. Although the BT0 sensor has been used in the pharmaceutical sector, maintaining its long-term performance has become an unresolved issue for practical use. The sensor response gradually decreases with increasing storage time and decreases by half after one year, finally becoming almost zero after two years at room temperature and humidity in a laboratory. This problem not only affects the performance of the BT0 sensor, but also increases the cost of transportation and preservation. The objectives of this study are to clarify the decrease in the response of the BT0 sensor and to improve the sensor durability to extend its lifetime. In this study, we developed an accelerated deterioration process to reproduce a BT0 sensor after long-term deterioration in a shorter time than by natural deterioration. In addition, we found that the deterioration of the response was caused by the change in the surface charge density due to the hydrolysis of TBAC. We also demonstrated that increasing PADE greatly accelerates the deterioration of the response by creating an acidic environment that induces TBAC hydrolysis. Finally, a new membrane for sensing the bitterness of medicines that possesses higher durability and sensitivity was developed by adjusting the proportions of the lipid and plasticizers. 2. Experiment 2.1. Materials Bis(1-butylpentyl) adipate (BBPA) was purchased from Sigma-Aldrich, Inc. (St. Louis, MO, USA). Tributyl o-acetylcitrate (TBAC) and phosphoric acid di-n-decyl ester (PADE) were purchased from Tokyo Chemical Industry Co., Ltd. (Tokyo, Japan). Polyvinyl chloride (PVC), obtained from Wako Pure Chemical, Ltd. (Osaka, Japan), was used as the supporting material. The chemical structures of PADE, TBAC, and BBPA are shown in Figure 1. Tetrahydrofuran (THF) purchased from Sigma-Aldrich was used as the organic solvent.

Sensors 2017, 17, 2541 Sensors 2017, 2017, 17, 17, 2541 2541 Sensors

3 of 14 of 14 14 3 of

Figure1. 1.Chemical Chemicalstructures structuresof ofPADE, PADE, TBAC, TBAC, and and BBPA. BBPA. Figure Figure 1. Chemical structures of PADE, TBAC, and BBPA.

2.2. Fabrication of Sensor Electrode 2.2. 2.2. Fabrication Fabrication of of Sensor Sensor Electrode Electrode Lipid polymer membranes with different compositions were fabricated. The fabrication steps Lipid Lipid polymer polymer membranes membranes with with different compositions compositions were fabricated. The The fabrication fabrication steps steps were as as follows:Firstly, Firstly, PADE, BBPA, TBAC, the concentrations ofwill which will beindetailed in PADE, BBPA, TBAC, the concentrations of which be detailed Section 2.5, were asfollows: follows: Firstly, PADE, BBPA, TBAC, the concentrations of which will be detailed in Section 2.5, and 800 mg of dissolved PVC were in dissolved in 10and mLstirred THF and stirred for one hour. Secondly, the and 800 mg of PVC were 10 mL THF for one hour. Secondly, the resulting Section 2.5, and 800 mg of PVC were dissolved in 10 mL THF and stirred for one hour. Secondly, the resulting solution wasinto poured a 90 mm glass dish tothree dry for three a draft chamber. solution poured a 90 into mm Petri dishPetri to dry for days in adays draftin Thirdly, resultingwas solution was poured into glass a 90 mm glass Petri dish to dry for three days inchamber. a draft chamber. Thirdly, the membrane was pieces cut into pieces of 1and × 1 attached cm and attached to probe a sensor probe using an the membrane was cut into of 1 × 1 cm to a sensor using an adhesive Thirdly, the membrane was cut into pieces of 1 × 1 cm and attached to a sensor probe using an adhesive of 10 mL THF and 800 mg PVC. Fourthly, the sensor probe was filled with 0.2 mL of a of 10 mL THF PVC. sensor the probe was probe filled with mLwith of a0.2 solution adhesive of 10and mL 800 THFmg and 800 Fourthly, mg PVC.the Fourthly, sensor was 0.2 filled mL ofofa solution of and 3.3 M KCl andAgCl. saturated AgCl. sensor electrode was completed byAg/AgCl fixing an 3.3 M KCl saturated Finally, theFinally, sensor the electrode completed by fixing an solution of 3.3 M KCl and saturated AgCl. Finally, the sensorwas electrode was completed by fixing an Ag/AgCl electrode to the sensor probe. Figurethe 2 shows the sensor bitterness sensor The electrode. Theofthickness electrode to the sensor probe. Figure 2 shows bitterness electrode. thickness the lipid Ag/AgCl electrode to the sensor probe. Figure 2 shows the bitterness sensor electrode. The thickness of the lipid polymer membrane is about 0.73 mm. polymer membrane about 0.73ismm. of the lipid polymerismembrane about 0.73 mm.

Figure 2. Bitterness sensor electrode (BT0). The sensor electrode is composed composed of Ag/AgCl Ag/AgCl electrode Figure Figure 2. Bitterness Bitterness sensor electrode (BT0). The sensor electrode is composedof of Ag/AgCl electrode electrode and the sensor probe with a lipid polymer membrane. and the sensor probe and the sensor probe with a lipid polymer membrane.

2.3. Measurement System 2.3. 2.3. Measurement Measurement System System The sensor electrode and aa reference reference electrode (Ag/AgCl (Ag/AgCl electrode containing a solution of The The sensor sensor electrode electrode and and a reference electrode electrode (Ag/AgCl electrode electrode containing containing aa solution solution of of 3.3 M KCl and saturated AgCl) were used in the measurement. First, the bottom of the sensor 3.3 saturated AgCl) were usedused in theinmeasurement. First, the bottom the sensor electrode 3.3 M MKCl KCland and saturated AgCl) were the measurement. First, the of bottom of the sensor electrode with the membrane and theelectrode reference electrode were in immersed in asolution reference solution with the membrane and the reference immersed a reference comprising electrode with the membrane and the referencewere electrode were immersed in a reference solution comprising mM and 0.3 acid mM tartaric to obtain a reference potential Vr. solution The reference 30 mM KCl 30 and 0.3 KCl mM obtainacid a reference potential V r . The reference acted comprising 30 mM KCl tartaric and 0.3 mMtotartaric acid to obtain a reference potential Vr. The reference solution acted as an alternative to human saliva, and has almost no taste. Second, Vs was obtained by solution acted as an alternative to human saliva, and has almost no taste. Second, Vs was obtained by

Sensors 2017, 17, 2541

4 of 14

as an alternative to human saliva, and has almost no taste. Second, V s was obtained by placing the electrodes in a sample solution. Third, a different potential V r’ was obtained after the two electrodes were lightly rinsed with the reference solution. We defined the potential difference between V s and V r as the relative value and the potential difference between V r’ and V r as the CPA (the change in the membrane potential caused by adsorption) value [29,30]. In this paper, all the sensor responses are given as the CPA values. Finally, the CPA value was returned to V r by rinsing with a cleaning solution consisting of 30 vol % ethanol and 100 mM HCl. In this study, 0.1 mM quinine hydrochloride was used as a standard solution of a bitter medicine. 2.4. Accelerated Deterioration Process The reason for the deterioration in the response was investigated by comparing the characteristics of the BT0 sensor before and after the deterioration process. However, it is inefficient to wait for the BT0 sensor to degrade naturally. Therefore, we developed an accelerated deterioration process to reproduce a BT0 sensor after long-term deterioration in a shorter time than by natural deterioration. To find suitable conditions for accelerating the deterioration, we placed the sensor membrane in a chamber with constant temperature and humidity (YAMATO IG421, Tokyo, Japan) for 14 and 28 days. The sensor membrane was placed in a glass Petri dish without a cover during the accelerated deterioration process. The storage temperature was set to 45–80 ◦ C. The relative humidity (RH) was set to 20% (regarded as a dry condition), 40%, 60%, or 95% RH. After the accelerated deterioration process, sensor electrodes were fabricated using the deteriorated membranes and used to measure 0.1 mM quinine hydrochloride solution. 2.5. Effect of Membrane Components on the Deterioration Rate To determine the main component causing the response deterioration, the amounts of the lipid and the two plasticizers in the membrane were changed to observe the effect on the sensor response. The deterioration rate D was calculated using the following equation: D=

| Rafter − Rbefore | , Rbefore

(1)

where Rafter is the sensor response after the accelerated deterioration process and Rbefore is the sensor response before the accelerated deterioration process. We adopted the control variate method to arrange the sample concentrations. Nine membranes comprising 100% BBPA, 100% TBAC, and 33–278% PADE; five membranes comprising 100% PADE, 100% TBAC, and 33–200% BBPA; and six membranes comprising 100% PADE, 100% BBPA, and 3.3–100% TBAC were made. Here, 100% means that the absolute amount is the same as that in the BT0 sensor. After the accelerated deterioration process, we examined the responses of the sensors using these 20 membranes by measuring 0.1 mM quinine hydrochloride. 2.6. Quantitative Analysis by Liquid Chromatography-Tandem Mass Spectrometry Liquid chromatography–tandem mass spectrometry (LC-MS/MS) was adopted to compare the difference in the amounts of PADE, TBAC, and BBPA in the membrane before and after four-week accelerated deterioration. The sample solutions were prepared as follows. Firstly, 0.2 g of the sensor membrane was dissolved in a 100 mL screw tube. Secondly, the polymer was precipitated by slowly adding 100 mL of acetonitrile. Thirdly, some of the prepared solution was diluted with acetonitrile (diluted 20-fold for PADE quantification or 1000-fold for BBPA and TBAC quantification). Finally, the diluted solution was filtered through a PTFE filter and subjected to LC-MS/MS measurement. The conditions are summarized in Table 1.

Sensors 2017, 17, 2541

5 of 14

Table 1. LC-MS/MS conditions. LC instrument

Shimadzu LC-20A (Kyoto, Japan)

LC column

Cadenza CD-C18 (2.0 × 100 mm, 3 µm, Portland, OR, USA)

Column temperature

50 ◦ C

Mobile phase Flow rate

A: 10 mmol/L Ammonium acetate/H2 O B: Acetonitrile 0.3 mL/min 0.0 min → 10.0 min: B20% → B90% 10.0 min → 12.0 min: B90%

Gradient conditions

12.1 min → 20.0 min: B20% 0.0 min → 5.0 min: B60% → B95% 5.0 min → 15.0 min: B95% 15.1 min → 20.0 min: B60%

Injection volume

1 µL

MS instrument

API 4000 (AB SCIEX)

Ionization

ESI

Polarity

negative; positive SRM

Scan type

Q1: m/z 377.3 → Q3: m/z 237.1 Q1: m/z 399.3 → Q3: m/z 273.2 Q1: m/z 403.3 → Q3: m/z 329.3

2.7. Component Detection by Gas Chromatography-Mass Spectrometry To elucidate the physicochemical reaction occurring during the accelerated deterioration process, we compared the components contained in the BT0 membrane before and after the deterioration test. The membrane components were extracted by solid-phase micro-extraction (SPME) and analyzed by gas chromatography-mass spectrometry (GC-MS). By using the SPME/GS-MS method, the lipid polymer membrane can be measured directly without any solvent. The device used for GC-MS was a SHIMADZU QP2010 (Kyoto, Japan), in which electron ionization (EI) was performed at 70 eV. The column used for all analyses was a Stabilwax® -MS, in which the column was 30 m length, 0.25 mm I.D., 0.25 µm film thickness. The oven temperature was initially held at 40 ◦ C for 1 min, after which the temperature was increased to 270 ◦ C at a rate of 10 ◦ C/min. This temperature, 270 ◦ C, was maintained for 4 min. The split ratio was 1:25. The temperature of the inlet was set to 270 ◦ C and the interface temperature was set to 280 ◦ C. An 85 µm polyacrylate film fiber was used to extract the components in the membrane. The samples were cutoffs of BT0 membranes with a weight of 20 mg before and after the accelerated deterioration process. 2.8. Measurement of the Amount of Adsorbed Quinine Hydrochloride In order to determine whether the deterioration in the response was caused by the decrease in the adsorption of quinine hydrochloride, we measured the amount of quinine hydrochloride adsorbed on the membrane using an ultraviolet-visible spectrophotometer (UV-1800, Shimadzu Corporation, Kyoto, Japan). First, the relationship between the concentration and absorbance (calibration curve) was obtained by measuring the absorbance of a known concentration of quinine hydrochloride solution.

Sensors 2017, 17, 2541

6 of 14

Five milliliters of 0.1 mM quinine hydrochloride solution was added dropwise onto the surface of a membrane in a glass Petri dish. After allowing the quinine hydrochloride molecules to be adsorbed on the membrane for 30 s, and 3 mL of the quinine hydrochloride solution was removed from the glass Petri dish to measure its absorbance. By measuring the absorbance of the measured solution and the calibration curve, the concentration of the measured solution was calculated. We defined the difference between the amounts of the originally-added quinine hydrochloride solution and the solution after the 30 s adsorption process as the total amount of adsorbed quinine hydrochloride. By dividing by the area of the glass Petri dish, we obtained the amount of quinine hydrochloride adsorbed per square centimeter [30,31]. 2.9. Measurement of Surface Contact Angle To investigate the hydrophobicity of the membrane surface, 11 membranes were made for measurement comprising fixed amounts of BBPA and TBAC and 33–278% PADE. By measuring the contact angle of the membrane surface, we determined the change in the hydrophobicity of the membrane surface. The contact angle was measured by a DM500 contact angle meter (Kyowa Interface Science Co., Ltd., Saitama, Japan). In the measurement, a 2 µL water drop was placed on the membrane surface. 2.10. Fabrication of a Durable Bitterness Sensor We optimized the membrane composition to increase the durability (no decrease in response after accelerated deterioration process). To confirm the selectivity of the improved bitterness sensor to the basic taste qualities, sensor electrodes with the new membrane composition were fabricated to measure saltiness, sourness, umami, bitterness, and sweetness. Iso-alpha acid was used to examine the response to acidic bitter materials (bitterness (−)) [32,33], while quinine hydrochloride was used to examine hydrochloride salts (bitterness (+)) [34]. In addition, there are three requirements for the performance of the newly improved bitterness sensor for medicine: selectivity, durability, and concentration dependence. First, the CPA value for standard quinine hydrochloride solution should be at least 30 mV and not be affected by any other basic tastes. Second, the improved sensor should show dependence of the response of the improved bitterness sensor on quinine hydrochloride concentration. Third, the sensor response should remain almost unchanged after accelerated deterioration. 3. Results and Discussion 3.1. Conditions of the Accelerated Deterioration Process As shown in Figure 3a, the sensor response ratio gradually decreased during the natural deterioration process at room temperature (25 ◦ C) and humidity in a laboratory (uncontrolled, annual average RH of 67% at Kanagawa, Japan). The sensor response dropped by half after 400 days and was zero after 830 days. Although a higher temperature caused a faster deterioration, we used a temperature of 45 ◦ C to examine the effect of the humidity rather than a higher temperature to avoid the loss of membrane components due to evaporation. In Figure 3b, the CPA value remained almost unchanged at 45 ◦ C and 20% RH (dry condition). On the other hand, it decreased under wet conditions. The higher the humidity, the faster the decrease in the CPA value. This result shows that humidity is a factor causing the deterioration in response. In the case of natural deterioration, the response ratio decreased to about 45% after 443 days. In the case of the accelerated deterioration process, the response ratio reached 45% at 45 ◦ C and 95% RH after 28 days, i.e., the deterioration process was about 16 times faster than natural deterioration. In this case, the membrane after one month of accelerated deterioration can produce the same response characteristics as those after one year of natural deterioration. In the following experiment, a deteriorated BT0 membrane was obtained by placing a BT0 membrane on a Petri dish without a cover under the conditions of 45 ◦ C and 95% RH for 28 days.

Sensors 2017, 17, 2541 Sensors2017, 2017,17, 17,2541 2541 Sensors

7 of 14 14 77ofof14

(a) (a)

(b) (b)

Figure3.3.The Therelationship relationshipbetween betweenthe theresponse responseratio ratioand andpreservation preservationtime time(a) (a)atatroom roomtemperature temperature Figure Figure 3. The relationship between the response ratio and preservation time (a) at room temperature andhumidity humidityininaalaboratory laboratoryand and(b) (b)atat45 45°C °Cwith withdifferent differentlevels levelsofofhumidity. humidity. and and humidity in a laboratory and (b) at 45 ◦ C with different levels of humidity.

3.2.Relationship Relationshipbetween betweenDeterioration DeteriorationRate Rateand andLipid LipidMass MassRatio Ratio 3.2. 3.2. Relationship between Deterioration Rate and Lipid Mass Ratio Figure44illustrates illustratesthe thedeterioration deteriorationrate ratedefined definedby byEquation Equation(1) (1)when whenthe thePADE PADEmass massratio ratio Figure Figure 4 illustrates the deterioration rate defined by TBAC, Equation (1) when the PADE mass were ratio increased from 0.66% to 5.36%. The mass ratios of PADE, and BBPA in the BT0 sensor increased from 0.66% to 5.36%. The mass ratios of PADE, TBAC, and BBPA in the BT0 sensor were increased from 0.66% to 5.36%. The mass ratios of PADE, TBAC, and BBPA in the BT0 sensor were 1.96%,65.36%, 65.36%,and and32.68%, 32.68%,respectively. respectively.The Thecircled circledgreen greenpoints pointsrepresent representthe thesensor sensormembranes membranes 1.96%, 1.96%, 65.36%, and 32.68%, respectively. The circled green points represent the sensor membranes withthe theTBAC TBACmass massratio ratioless lessthan than49%. 49%.Excluding Excludingthe thecircled circledgreen greenpoints, points,the thedeterioration deterioration rate with rate with the TBAC mass ratio less than 49%. Excluding the circled green points, the deterioration 2 = 0.85). This finding may indicate that (1) rate has a strong correlation with the PADE mass ratio (R the 2 has a strong correlation with the PADE mass ratio (R 2= 0.85). This finding may indicate that (1) the has a strong correlation with the PADE mass ratio (R = 0.85). This finding may indicate that (1) the deteriorationrate rateofofBT0 BT0was wasproportional proportionaltotothe thePADE PADEmass massratio ratiowhen whenthe themass massratio ratioofofTBAC TBAC deterioration deterioration rate of BT0 was proportional to the PADE mass ratio when the mass ratio of TBAC exceeded49%; 49%;(2) (2)when whenthe theTBAC TBACmass massratio ratioininthe themembrane membranewas wasless lessthan than49%, 49%,the theresponse response exceeded exceeded 49%; (2)suppressed. when the TBAC mass ratio in the membrane was less than 49%, the response deterioration was deterioration was suppressed. deterioration was suppressed.

Figure 4. 4. The The relationship relationship between between deterioration deterioration rate rate and and lipid lipid mass mass ratio. ratio. The The circled circled green green points points Figure Figure 4. The relationship between deterioration rate and lipid mass ratio. The circled green points represent the sensor membranes with the TBAC mass ratio less than 49%. The regression equation and representthe thesensor sensormembranes membraneswith withthe theTBAC TBACmass massratio ratioless lessthan than49%. 49%.The Theregression regressionequation equation represent R2 were calculated without the green points. 2 andRR2 were werecalculated calculatedwithout withoutthe thegreen greenpoints. points. and

3.3. LC-MS/MS LC-MS/MS Analysis Analysis Results Results 3.3. 3.3. LC-MS/MS Analysis Results Figure 55 shows shows that that the the amount amount of of TBAC TBAC was was significantly significantly reduced reduced as as aa result result of deterioration. deterioration. Figure Figure 5 shows that the amount of TBAC was significantly reduced as a result ofofdeterioration. The amounts of PADE and BBPA did not change during the deterioration process, even though PADE Theamounts amountsofofPADE PADEand andBBPA BBPAdid didnot notchange changeduring duringthe thedeterioration deteriorationprocess, process,even eventhough though The PADEhad hadaastrong strongeffect effecton onthe thedeterioration deteriorationrate. rate.This Thisindicates indicatesthat thatPADE PADEpromoted promotedthe theprocess process PADE causing the deterioration in the response while remaining unchanged. causing the deterioration in the response while remaining unchanged.

Sensors 2017, 17, 2541

8 of 14

had a strong effect on the deterioration rate. This indicates that PADE promoted the process causing Sensors 88of Sensors2017, 2017,17, 17,2541 2541 of14 14 the deterioration in the response while remaining unchanged.

the main components obtained by LC-MS/MS. Figure Figure5.5.Quantitative Quantitativecomparison comparisonof ofthe themain maincomponents componentsobtained obtainedby byLC-MS/MS. LC-MS/MS.The Theresults resultsare are ±±SD expressed SD (n 3). expressedas asthe themean mean± SD(n (n== =3). 3).

3.4. ofofthe the Membrane Components 3.4. GC-MS Analysis 3.4.GC-MS GC-MSAnalysis Analysisof theMembrane MembraneComponents Components Butyl citrate was detected as product of TBAC. The of Butyl degradation the plasticizer amount butyl Butylcitrate citratewas wasdetected detectedas asaaadegradation degradationproduct productof ofthe theplasticizer plasticizerTBAC. TBAC.The Theamount amountof ofbutyl butyl citrate increased as a result of deterioration (Figure 6), indicating that some TBAC molecules were citrate citrate increased increased as asaaresult resultof ofdeterioration deterioration(Figure (Figure6), 6),indicating indicatingthat thatsome someTBAC TBACmolecules moleculeswere were hydrolyzed during the deterioration process. hydrolyzed hydrolyzedduring duringthe thedeterioration deteriorationprocess. process.

Figure of the accelerated deterioration Figure6.6. Change Change in in the the amount amount of of butyl butylcitrate citrateas asaaresult resultof ofthe theaccelerated accelerateddeterioration deteriorationprocess process measured by GC–MS. The results are expressed as the mean ± SD (n = 3). results are expressed as the mean ± SD (n = 3). measured by GC–MS. The results are expressed as the mean ± SD (n = 3).

3.5. 3.5.Amount AmountofofAdsorbed AdsorbedQuinine QuinineHydrochloride Hydrochloride

Figure shows that the same amount quinine hydrochloride was adsorbed before Figure amount ofof quinine hydrochloride was adsorbed before andand afterafter the Figure777shows showsthat thatthe thesame same amount of quinine hydrochloride was adsorbed before and after the deterioration. As we reported in previous studies, the sensor response was affected by both the deterioration. As we reported in previous studies, the sensor response was affected by both the surface the deterioration. As we reported in previous studies, the sensor response was affected by both the surface charge density and amount of [15,35]. indicates that charge the amount adsorption [15,35]. Therefore, the result the indicates the change in surfacedensity chargeand density andthe theof amount ofadsorption adsorption [15,35].Therefore, Therefore, theresult resultthat indicates thatthe the change in charge of membrane during the deterioration the surface charge density of thedensity BT0 membrane during the accelerated process probably change in the the surface surface charge density of the the BT0 BT0 membrane duringdeterioration the accelerated accelerated deterioration process probably caused the decrease in the sensor response. caused the decrease in the sensor response. process probably caused the decrease in the sensor response.

Sensors 2017, 17, 2541 Sensors 2017, 17, 2541 Sensors 2017, 17, 2541

9 of 14 9 of 14 9 of 14

Figure 7. Amounts Amounts of adsorbed quinine hydrochloride before and after after the the accelerated accelerated deterioration deterioration Figure 7. 7. Amountsof ofadsorbed adsorbedquinine quininehydrochloride hydrochloridebefore before and and after the accelerated deterioration Figure process. The results are expressed as the mean ± SD (n = 3). process. The Theresults results are are expressed expressed as as the the mean mean ±±SD SD(n(n==3). 3). process.

3.6. Reference Potential and Contact Angle of Lipid Lipid Polymer Polymer Membrane Membrane 3.6. 3.6. Reference ReferencePotential Potential and and Contact Contact Angle Angle of of Lipid Polymer Membrane The reference potential potential (V (Vrr)))was was significantly lower after the accelerated deterioration process, The The reference reference potential (V significantlylower lowerafter afterthe theaccelerated accelerateddeterioration deteriorationprocess, process, r wassignificantly which indicates that negatively-changed substances were generated, which increased the surface which which indicates indicates that that negatively-changed negatively-changed substances substances were were generated, generated, which which increased increased the the surface surface charge density (Figure 8). charge charge density density (Figure (Figure 8). 8).

Figure 8. The relationship between the reference potential and lipid concentration. The results are Figure 8. 8. The relationship relationship between between the the reference reference potential potential and and lipid lipid concentration. concentration. The The results results are are Figure expressedThe as the mean ± SD (n = 4). expressed as the mean ± SD (n = 4). expressed as the mean ± SD (n = 4).

The contact contact angle angle decreased decreased as as aa result of deterioration (Figure 9). This result result indicates indicates that that the the The result of of deterioration deterioration (Figure (Figure 9). 9). This This The contact angle decreased as a result result indicates that the surface became more hydrophilic during the deterioration process. During the deterioration process, surface became became more more hydrophilic hydrophilic during during the the deterioration deterioration process. process. During During the the deterioration deterioration process, process, surface TBAC was hydrolyzed into butyl citrate and acetic acid, which caused the membrane surface to to TBAC was hydrolyzed into butyl citrate and acetic acid, which caused the membrane surface TBAC was hydrolyzed into butyl citrate and acetic acid, which caused the membrane surface to become hydrophilic hydrophilic and and increased increased the the negativity negativity of of the the reference reference potential. potential. The The role role of of PADE PADE during during become become hydrophilic and increased the negativity of the reference potential. The role of PADE during the deterioration was to create an acidic condition in the membrane allowing TBAC to hydrolyze, the deterioration deterioration was was to to create create an an acidic acidic condition condition in in the the membrane membrane allowing allowing TBAC TBAC to to hydrolyze, hydrolyze, the because PADE is a phosphate ester. In addition, the PADE content affects the deterioration rate more more because PADE PADE is is aa phosphate phosphate ester. ester. In In addition, addition, the the PADE PADE content contentaffects affects the the deterioration deterioration rate rate because more than the TBAC content. Therefore, it is considered that the reduction in the PADE content improves than the the TBAC TBAC content. content. Therefore, Therefore, itit is is considered considered that that the the reduction reduction in in the the PADE PADE content content improves improves than the durability durability of the sensor. the of the sensor. the durability of the sensor.

Sensors 2017, 17, 2541 Sensors Sensors 2017, 2017,17, 17,2541 2541

10 of 14 10 10 of of 14 14

Figure 9. 9. The The relationship relationship between between the the surface surface contact contact angle angle and and lipid lipid concentration. concentration. The The results results are are Figure Figure 9. The relationship between the surface contact angle and lipid concentration. The results are expressed as the mean ± SD (n = 4). expressed as the mean ± SD (n = 4). expressed as the mean ± SD (n = 4).

3.7. BT0 Sensor with Improved Durability and Sensitivity 3.7. 3.7. BT0 BT0 Sensor Sensor with with Improved Improved Durability Durability and and Sensitivity Sensitivity As shown in Figure 10, we found that the deterioration rate of the CPA value was reasonably As in 10, found that thethe deterioration raterate of the CPACPA value was was reasonably low As shown shown inFigure Figure 10,we we that deterioration the value reasonably low when 33% or less PADE wasfound added. This membrane might beofconsidered as relatively durable when 33% or less PADE was added. This membrane might be considered as relatively durable because low when or less PADE added. This membrane might considered asabout relatively durable because of 33% the relatively weakwas acidic environment. However, thebe CPA value was 25 mV, only of the relatively weak acidic environment. However, the CPA value was about 25 mV, only 70% of that because of the relatively weak acidic environment. However, CPA valuehydrochloride) was about 25 mV, 70% of that of the conventional BT0 sensor (about 36 mV to 0.1the mM quinine andonly did of theofconventional BT0 sensor (about 36 mV(about to 0.1 mM quinine hydrochloride) and did not satisfy the 70% that of the conventional BT0 sensor 36 mV to 0.1 mM quinine hydrochloride) and not satisfy the required response. On the other hand, the CPA value increased when the amountdid of required response. On the other hand, the CPA value increased when the amount of the plasticizer not plasticizer satisfy the TBAC required response. OnBT0 the other hand,was the CPA value However, increased when amount of the included in the membrane decreased. TBACthe is considered TBAC included in the BT0 membrane decreased. However, TBAC is considered to beisnecessary to thebe plasticizer included in the was BT0 membrane was decreased. However, TBAC to necessaryTBAC to match the results of bitterness sensory tests for various medicines in considered the sensor match the resultsto of bitterness sensory tests for various medicines in sensormedicines development stage [22]. to be necessary theshown resultsinof bitterness sensory tests of forthe various in the development stagematch [22]. As Figure 11, the responses sensor electrodes with 17%,sensor 34%, As shown in Figure 11, the responses of sensor electrodes with 17%, 34%, and 50% TBAC exhibited development stage [22]. Asno shown in Figure 11, the before responses sensor 17%, 34%, and 50% TBAC exhibited significant difference andofafter the electrodes acceleratedwith deterioration, no difference and after difference the accelerated indicating high durability, andsignificant 50% TBAC exhibitedbefore no significant beforedeterioration, and after the accelerated deterioration, indicating high durability, as well as a sufficient response to quinine hydrochloride comparable to as well as a sufficient response to quinine hydrochloride comparable to that of the conventional BT0 indicating durability, as sensor well as[19]. a sufficient response to quinine to that of the high conventional BT0 Therefore, the new sensor ishydrochloride not expected comparable to deteriorate sensor [19]. Therefore, the BT0 new sensor sensor is notTherefore, expected tothe deteriorate during oneexpected year of storage at room that of the conventional [19]. new sensor is not to deteriorate during one year of storage at room temperature and humidity in a laboratory. temperature and of humidity laboratory. during one year storage in at aroom temperature and humidity in a laboratory.

Figure 10. The relationship between the sensor response and lipid concentration. The results are Figure 10. The relationship between the sensor response and lipid concentration. The results are expressed the relationship mean ± SD (n = 4). the sensor response and lipid concentration. The results are Figure 10.as The between expressed as the mean ± SD (n = 4). expressed as the mean ± SD (n = 4).

Sensors 2017, 17, 2541 Sensors 2017, 17, 2541

11 of 14 11 of 14

Figure 11. The relationship between the sensor response and TBAC concentration. The The results are Figure 11. The relationship between the sensor response and TBAC concentration. The results are expressed as the mean ± ± SD SD(n (n==4). 4). expressed as the mean ± SD (n = 4).

As in Figure 12, 12, all all three three sensors sensors showed showed good good sensitivity sensitivity to to bitterness bitterness (+) (+) and met the AsAsshown shown shownininFigure Figure 12, all three sensors showed good sensitivity to bitterness (+)and andmet metthe the response requirement of over 30 mV to 0.1 mM quinine hydrochloride. In addition, the sensor did response requirement of over 30 mV to 0.1 mM quinine hydrochloride. In addition, the sensor did not response requirement of over 30 mV to 0.1 mM quinine hydrochloride. In addition, the sensor did not respond to any other basic tastes, showing good selectivity to bitterness From viewpoint respond to any basic tastes, showing good selectivity to bitterness (+).(+). From thethe viewpoint of not respond toother any other basic tastes, showing good selectivity to bitterness (+). From the viewpoint of sensitivity, the optimized membrane composition is 33% PADE, 100% BBPA, and 17% TBAC. sensitivity, the optimized membrane composition is 33% PADE, 100% BBPA, and 17% TBAC. Therefore, of sensitivity, the optimized membrane composition is 33% PADE, 100% BBPA, and 17% TBAC. Therefore, we the measured the concentration of this sensor for quinine hydrochloride. we measured concentration dependencedependence of this sensor Figure 13 Therefore, we measured the concentration dependence offor thisquinine sensor hydrochloride. for quinine hydrochloride. Figure 13 shows that the sensor response of the improved sensor was proportional to the logarithm shows that the sensor response of the improved sensor was proportional to the logarithm of the Figure 13 shows that the sensor response of the improved sensor was proportional to the logarithm of the concentration of quinine hydrochloride, which expresses the bitterness intensity, in accordance concentration of quinine hydrochloride, which expresses the bitterness intensity, in accordance with of the concentration of quinine hydrochloride, which expresses the bitterness intensity, in accordance with the Weber-Fechner law [36]. The improved sensor shows thesame samelinear linearresponse responserange range as as the the Weber-Fechner law [36]. The improved sensor shows the the with the Weber-Fechner law [36]. The improved sensor shows the same linear response range as the conventional BT0 sensor of between between 0.01 0.01 and and 11 mM mM quinine quinine hydrochloride. hydrochloride. Since the bitterness bitterness conventional BT0 sensor of Since the conventional BT0 sensor of between 0.01 and 1 mM quinine hydrochloride. Since the bitterness intensity corresponding to between 0.01 and 0.1 mM quinine hydrochloride is the common range for intensity corresponding to between 0.01 and 0.10.1 mM quinine is is the intensity corresponding to between 0.01 and mM quininehydrochloride hydrochloride thecommon commonrange rangefor for sensory scores felt by humans, we chose this range to compare the sensor sensitivity. The improved sensory improved sensoryscores scoresfelt feltbybyhumans, humans,we wechose chosethis thisrange rangetotocompare comparethe thesensor sensorsensitivity. sensitivity.The The improved sensor has aa higher response-concentration slope than the conventional sensor, which indicates that sensor which indicates sensorhas has ahigher higherresponse-concentration response-concentrationslope slopethan thanthe theconventional conventionalsensor, sensor, which indicatesthat that the improved sensor has has higher higher sensitivity sensitivity in in this this bitterness bitterness range. range. the improved sensor the improved sensor has higher sensitivity in this bitterness range.

Figure 12. The sensor response to five basic tastes. The reference solution (RS) comprised 30 mM Figure 12. The sensor response to five basic tastes. The reference solution (RS) comprised 30 mM KCl, KCl, and 12. 0.3 The mMsensor tartaricresponse acid; thetosaltiness sample 300 solution mM KCl,(RS) andcomprised 0.3 mM tartaric acid; Figure five basic tastes.comprised The reference 30 mM KCl, and 0.3 mM tartaric acid; the saltiness sample comprised 300 mM KCl, and 0.3 mM tartaric acid; the the sourness sample comprised 30 mM KCl, and 3 mM tartaric acid; the umami sample comprised and 0.3 mM tartaric acid; the saltiness sample comprised 300 mM KCl, and 0.3 mM tartaric acid; the sourness sample comprised 30 mM KCl, and 3 mM tartaric acid; the umami sample comprised 10 mM 10sourness mM sodium glutamate and30 RS;mM theKCl, bitterness 0.1 mMsample quinine hydrochloride sample comprised and 3 (+) mMsample tartariccomprised acid; the umami comprised 10 mM sodium glutamate and RS; the bitterness (+) sample comprised 0.1 mM quinine hydrochloride and and RS; the bitterness ( − ) sample comprised 0.01 vol % iso-alpha acid and RS; and the sweetness sodium glutamate and RS; the bitterness (+) sample comprised 0.1 mM quinine hydrochloride and RS; the bitterness (−) sample comprised 0.01 vol% iso-alpha acid and RS; and the sweetness sample sample comprised M sucrose RS. RS; the bitterness1 (−) sample and comprised 0.01 vol% iso-alpha acid and RS; and the sweetness sample comprised 1 M sucrose and RS. comprised 1 M sucrose and RS.

Sensors 2017, 17, 2541 Sensors 2017, 17, 2541

12 of 14 12 of 14

Figure response of Figure 13. 13. Dependence Dependence of of the the response of the the improved improved bitterness bitterness sensor sensor and and conventional conventional bitterness bitterness sensor on quinine hydrochloride concentration. sensor on quinine hydrochloride concentration.

4. Conclusions We a We investigated investigatedthe thereason reasonfor forthe thedeterioration deteriorationininthe theresponse responseofofthe theBT0 BT0sensor sensorand andfabricated fabricated new lipid polymer membrane for bitterness with improved durability and sensitivity. We concluded a new lipid polymer membrane for bitterness with improved durability and sensitivity. We that the proximate for the deterioration in the response the BT0 sensor is thesensor increase in concluded that the cause proximate cause for the deterioration in theofresponse of the BT0 is the the surface density. Experimental results showed that TBAC wasthat involved hydrolysis increase in charge the surface charge density. Experimental results showed TBACinwas involvedand in produced negatively-charged substances, which increased the surface charge density. It was also hydrolysis and produced negatively-charged substances, which increased the surface charge density. revealed that the ratethat of deterioration is stronglyispromoted by increasing the mass ratio of PADE It was also revealed the rate of deterioration strongly promoted by increasing the mass ratio because acidic environment generated bygenerated PADE promotes TBAC hydrolysis. In addition, a low of PADEthebecause the acidic environment by PADE promotes TBAC hydrolysis. In ratio of TBAC suppress the deterioration rate and improve theimprove sensitivity the sensor. addition, a low can ratioboth of TBAC can both suppress the deterioration rate and theof sensitivity of Therefore, results motivated to improve theimprove sensor by the amountsthe of amounts PADE and the sensor.the Therefore, the resultsus motivated us to thereducing sensor by reducing of TBAC theTBAC membrane. conclusion, producedwe a new sensor amembrane consisting of 33% PADE, PADEin and in the In membrane. Inwe conclusion, produced new sensor membrane consisting 100% BBPA, and100% 17% TBAC. showed high durability andhigh sensitivity compared with the of 33% PADE, BBPA, This and sensor 17% TBAC. This sensor showed durability and sensitivity conventional BT0 sensor. compared with the conventional BT0 sensor. The next step is to measure commercially available medicines to confirm the correlation between the sensor response response and and the the bitterness bitterness sensory sensory score. score. Acknowledgments: The The paper paperisisbased basedononresults results obtained from a project commissioned by New the New Energy Acknowledgments: obtained from a project commissioned by the Energy and and Industrial Technology Development Organization (NEDO). Industrial Technology Development Organization (NEDO). Author Author Contributions: Contributions: The work presented here was carried out in collaboration between all authors. Xiao Wu, Yusuke Tahara, Rui RuiYatabe, Yatabe,Hidekazu Hidekazu Ikezaki, Kiyoshi defined the research Xiao Wu, Yusuke Tahara, Ikezaki, andand Kiyoshi TokoToko defined the research theme.theme. Xiao Wu, Hideya Hideya Onitake, and Zhiqin Huang designed method and experiments. Xiao Wu, Hideya Onitake, Zhiqin Huang, Onitake, and Zhiqin Huang designed method and experiments. Xiao Wu, Hideya Onitake, Zhiqin Huang, and and Takeshi Shiino carried out the experiments and analyzed the data, and Xiao Wu interpreted the results Takeshi Shiino carriedYusuke out the Tahara, experiments and analyzed the data, andand XiaoKiyoshi Wu interpreted the results and wrote and wrote the paper. Rui Yatabe, Hidekazu Ikezaki, Toko provided directions for the paper. Yusuke Tahara, Rui Yatabe, Hidekazu Ikezaki, and Kiyoshi Toko provided directions for experimental methods, analysis of data, interpretation of the results, and writing of the paper. All authors have contributed to,methods, seen, andanalysis approved the manuscript. experimental of data, interpretation of the results, and writing of the paper. All authors have contributed to, seen, and approved the manuscript. Hidekazu Ikezaki is the president of Intelligent Sensor Technology, Inc., the maker of the Conflicts of Interest: taste sensing system TS-5000Z. Kiyoshi Toko holds stock in Intelligent Sensor Technology, Inc. Conflicts of Interest: Hidekazu Ikezaki is the president of Intelligent Sensor Technology, Inc., the maker of the taste sensing system TS-5000Z. Kiyoshi Toko holds stock in Intelligent Sensor Technology, Inc.

References

References 1. Kawano, Y.; Ito, A.; Sasatsu, M.; Machida, Y. Preparation of orally disintegrating tablets with taste-masking function: Masking effect in M.; granules prepared with correctives using the dry tablets granulation method and 1. Kawano, Y.; Ito, A.; Sasatsu, Machida, Y. Preparation of orally disintegrating with taste-masking evaluation of tablets prepared using the taste-masked granules. Yakugaku Zasshi 2010, 130, 81–86. function: Masking effect in granules prepared with correctives using the dry granulation method and (In Japanese) [PubMed] evaluation of[CrossRef] tablets prepared using the taste-masked granules. Yakugaku Zasshi 2010, 130, 81–86. (in 2. Anand, V.; Kataria, M.; Kukkar, V.; Saharan, V. The latest trends in the taste assessment of pharmaceuticals. Japanese) Drug Discov. Today 2007, 12, 257–265. [CrossRef] [PubMed] 2. Anand, V.; Kataria, M.; Kukkar, V.; Saharan, V. The latest trends in the taste assessment of pharmaceuticals. Drug Discov. Today 2007, 12, 257–265.

Sensors 2017, 17, 2541

3. 4. 5. 6. 7.

8. 9.

10. 11.

12.

13.

14.

15.

16. 17. 18. 19.

20. 21.

22.

23. 24.

13 of 14

Banerjee, R.; Tudu, B.; Bandyopadhyay, R.; Bhattacharyya, N. A review on combined odor and taste sensor systems. J. Food Eng. 2016, 190, 10–21. [CrossRef] Tahara, Y.; Toko, K. Electronic tongues—A review. IEEE Sens. J. 2013, 13, 3001–3011. [CrossRef] Bachmanov, A.A.; Beauchamp, G.K. Taste receptor genes. Annu. Rev. Nutr. 2007, 27, 389–414. [CrossRef] [PubMed] Winquist, F. Voltammetric electronic tongues—Basic principles and applications. Microchim. Acta 2008, 163, 3–10. [CrossRef] Facure, M.H.M.; Mercante, L.A.; Mattoso, L.H.C.; Correa, D.S. Detection of trace levels of organophosphate pesticides using an electronic tongue based on graphene hybrid nanocomposites. Talanta 2017, 167, 59–66. [CrossRef] [PubMed] Ciosek, P.; Wroblewski, W. Sensor array for liquid sensing-electronic tongue systems. Analyst 2007, 132, 963–978. [CrossRef] [PubMed] Marx, I.; Rodrigues, N.; Dias, L.G.; Veloso, A.C.A.; Pereira, J.A.; Drunkler, D.A.; Peres, A.M. Sensory classification of table olives using an electronic tongue: Analysis of aqueous pastes and brines. Talanta 2017, 162, 98–106. [CrossRef] [PubMed] Citterio, D.; Suzuki, K. Smart taste sensors. Anal. Chem. 2008, 80, 3965–3972. [PubMed] Nuñez, L.; Cetó, X.; Pividori, M.I.; Zanoni, M.V.B.; del Valle, M. Development and application of an electronic tongue for detection and monitoring of nitrate, nitrite and ammonium levels in waters. Microchem. J. 2013, 110, 273–279. [CrossRef] Daikuzonoa, C.M.; Dantas, C.A.R.; Volpati, D.; Constantino, C.J.L.; Piazzetta, M.H.O.; Gobbi, A.L.; Taylor, D.M.; Oliveira, O.N., Jr.; Riul, A., Jr. Microfluidic electronic tongue. Sens. Actuators B Chem. 2015, 207, 1129–1135. [CrossRef] Song, H.S.; Kwon, O.S.; Lee, S.H.; Park, S.J.; Kim, U.K.; Jang, J.; Park, T.H. Human Taste Receptor-Functionalized Field Effect Transistor as a Human-Like Nanobioelectronic Tongue. Nano Lett. 2013, 13, 172–178. [CrossRef] [PubMed] Ha, D.; Sun, Q.; Su, K.; Wan, H.; Li, H.; Xu, N.; Sun, F.; Zhuang, L.; Hu, N.; Wang, P. Recent achievements in electronic tongue and bioelectronic tongue as taste sensors. Sens. Actuators B Chem. 2015, 207, 1136–1146. [CrossRef] Toko, K.; Hara, D.; Tahara, Y.; Yasuura, M.; Ikezaki, H. Relationship between the amount of bitter substances adsorbed onto lipid/polymer membrane and the electrical response of taste sensors. Sensors 2014, 14, 16274–16286. [CrossRef] [PubMed] Ikezaki, H.; Kobayashi, Y.; Toukubo, R.; Naito, Y.; Taniguchi, A.; Toko, K. Techniques to control sensitivity and selectivity of multichannel taste sensor using lipid membranes. Proc. Transducers 1999, 99, 1634–1637. Cui, H.; Habara, M.; Ikezaki, H.; Toko, K. Selectivity control in a sweetness sensor using lipid/polymer membranes. Sens. Mater. 2005, 17, 385–390. Del Valle, M. Electronic tongues employing electrochemical sensors. Electroanalysis 2010, 22, 1539–1555. [CrossRef] Kobayashi, Y.; Habara, M.; Ikezaki, H.; Chen, R.G.; Naito, Y.; Toko, K. Advanced taste sensors based on artificial lipids with global selectivity to basic taste qualities and high correlation to sensory scores. Sensors 2010, 10, 3411–3443. [CrossRef] [PubMed] Liu, J.; Yang, J.; Wang, W.; Fu, S.; Shi, Y.; Men, H. Automatic evaluation of sensory information for beer at a fuzzy level using electronic tongue and electronic nose. Sens. Mater. 2016, 28, 785–795. [CrossRef] Lin, R.; Gao, X.; Wang, J.; Dai, L.; Kang, B.; Zhang, L.; Shi, J.; Gui, X.; Liu, P.; Li, X. Traditional human taste panel and taste sensors methods for bitter taste masking research on combined bitterness suppressants of berberine hydrochloride. Sens. Mater. 2017, 29, 105–116. [CrossRef] Kobayashi, Y.; Hamada, H.; Yamaguchi, Y.; Ikezaki, H.; Toko, K. Development of an artificial lipid-based membrane sensor with high selectivity and sensitivity to the bitterness of drugs and with high correlation with sensory score. IEEJ Trans. 2009, 4, 710–719. [CrossRef] Akitomi, H.; Tahara, Y.; Yasuura, M.; Kobayashi, Y.; Ikezaki, H.; Toko, K. Quantification of tastes of amino acids using taste sensors. Sens. Actuators B Chem. 2013, 179, 276–281. [CrossRef] Ito, M.; Wada, K.; Yoshida, M.; Hazakawa, K.; Abe, K.; Chen, R.G.; Habara, M.; Ikezaki, H.; Uchida, T. Quantitative evaluation of bitterness of H1-receptor antagonists and masking effect of acesulfame potassium, an artificial sweetener, using a taste sensor. Sens. Mater. 2013, 25, 17–30.

Sensors 2017, 17, 2541

25. 26.

27. 28. 29.

30.

31. 32. 33. 34.

35. 36.

14 of 14

Takagi, S.; Toko, K.; Wada, K.; Ohki, T. Quantification of suppression of bitterness using an electronic tongue. J. Pharm. Sci. 2001, 12, 2042–2048. [CrossRef] Wu, X.; Onitake, H.; Haraguchi, T.; Tahara, Y.; Yatabe, R.; Yoshida, M.; Uchida, T.; Ikezaki, H.; Toko, K. Quantitative prediction of bitterness masking effect of high-potency sweeteners using taste sensor Voltammetric electronic tongues—Basic principles and applications. Sens. Actuators B Chem. 2016, 235, 11–17. [CrossRef] Yasuura, M.; Okazaki, H.; Tahara, Y.; Ikezaki, H.; Toko, K. Development of sweetness sensor with selectivity to negatively charged high-potency sweeteners. Sens. Actuators B Chem. 2014, 201, 329–335. [CrossRef] Yasuura, M.; Tahara, Y.; Ikezaki, H.; Toko, K. Development of a sweetness sensor for positively charged high-potency sweeteners. Sensors 2014, 14, 7359–7373. [CrossRef] [PubMed] Harada, Y.; Tahara, Y.; Toko, K. Study of the relationship between taste sensor response and the amount of epigallocatechin gallate adsorbed onto a lipid-polymer membrane. Sensors 2015, 15, 6241–6249. [CrossRef] [PubMed] Fukagawa, T.; Tahara, Y.; Yasuura, M.; Habara, M.; Ikezaki, H.; Toko, K. Relationship between taste sensor response and amount of quinine adsorbed on lipid/polymer membrane. J. Innov. Electron. Commun. 2012, 2, 1–6. Hara, D.; Fukagawa, T.; Tahara, Y.; Yasuura, M.; Toko, K. Examination of amount of astringent substances adsorbed onto lipid/polymer membrane used in taste sensor. Sens. Lett. 2014, 12, 1172–1176. [CrossRef] Gastl, M.; Hanke, S.; Back, W. Analytical investigations to evaluate bitter sensation using a taste sensing system. Brew. Sci. 2007, 60, 48–54. Gastl, M.; Hanke, S.; Back, W. Drinkability—Balance and harmony of components as well as an incentive for continuing to drink. BRAUWELT Int. 2008, 3, 148–153. Lyall, V.; Heck, G.L.; Vinnikova, A.K.; Ghosh, S.; Phan, T.H.; Alam, R.I.; Russell, O.F.; Malik, S.A.; Bigbee, J.W.; DeSimone, J.A. The mammalian amiloride-insensitive non-specific salt taste receptor is a vanilloid receptor-1 variant. J. Physiol. 2004, 588, 147–159. [CrossRef] [PubMed] Harada, Y.; Noda, J.; Yatabe, R.; Ikezaki, H.; Toko, K. Research on the changes to the lipid/polymer membrane used in the acidic bitterness sensor caused by preconditioning. Sensors 2016, 16, 230. [CrossRef] [PubMed] Pfaffmann, C. The sense of taste. In Handbook of Physiology; Field, I.J., Ed.; American Physiological Society: Washington, DC, USA, 1959; Volume 1, pp. 507–533. © 2017 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/).