Chemical Papers 69 (7) 938–949 (2015) DOI: 10.1515/chempap-2015-0101
ORIGINAL PAPER
Electrode and electrodeless impedance measurement for determination of orange juices parameters a
Romana Seidlová*, b Jaroslav Poživil, c Jaromír Seidl, d Stanislav Ďaďo, e Petra Průšová, b Lukáš Malec
a Department
of Computing and Control Engineering, Faculty of Chemical Engineering, University of Chemistry and Technology, Technická 6, 166 28 Prague 6, Czech Republic b Department
c Department
of Physics and Measurements, Faculty of Chemical Engineering, University of Chemistry and Technology, Technická 6, 166 28 Prague 6, Czech Republic
d Departments
e Department
of Information Technologies and Analytical Methods, University of Business in Prague, Spálená 76, 110 00 Prague 1, Czech Republic
of Measurement, Faculty of Electrical Engineering, Czech Technical University in Prague, Technická 2, 166 27 Prague 6, Czech Republic
of Food Preservation, Faculty of Food and Biochemical Technology, University of Chemistry and Technology, Technická 6, 166 28 Prague 6, Czech Republic Received 5 June 2014; Revised 23 November 2014; Accepted 12 December 2014
Electrical impedance spectroscopy (EIS) is a non-destructive, rapid and real-time measurement method which does not require special high-tech measurement devices and can be applied to food quality assessment. This method is rapid, effective and affords low-cost investigation of the product. The conventional EIS method requires a set of metal electrodes in direct contact with the medium to be measured. The complicated electrochemical processes on the electrodes–electrolyte interface could substantially affect the value of the impedance measured. The present study sought to explore the possibilities of using the impedance method for quality control in orange juices, to introduce the electrodeless method of electrolyte impedance measurement and to compare this with the conventional impedance methods. The electrical properties of the orange juices were described with the help of an equivalent circuit. An equivalent circuit was designed with constant phase element approximation. The values of the equivalent circuit components were fitted using a non-standard algorithm inspired by the behaviour of actual ant colonies. Implementing the electrodeless method obviated the electrodes phenomena effects and the behaviour of the electrolyte is similar to inductance. The proposed electrodeless method is generally applicable to measuring the electrochemical properties of electrolytes. c 2015 Institute of Chemistry, Slovak Academy of Sciences Keywords: electrical impedance spectrometry, transformer-based electrodeless impedance measurement, ant colony algorithm, equivalent circuits, assessment of juice parameters
Introduction Juice is an important citrus fruit product, containing more than half the fruit mass. Quality control of
fruit juices and the adulteration of fruit juices represent global problems. In the fifties, there was the first big case of fruit nectar adulteration in the USA (Nagy et al., 1988), then in Africa (Maireva et al.,
*Corresponding author, e-mail:
[email protected]
Brought to you by | New York University Bobst Library Technical Services Authenticated Download Date | 7/22/15 12:32 AM
R. Seidlová et al./Chemical Papers 69 (7) 938–949 (2015)
2013). The same problems have occurred in Europe but, due to the activities of the national fruit juice industries and their federations, the number of incidents has decreased. Over the years, fruit juices have frequently been subject to adulteration. As analytical techniques have become more objective and accurate, the methods used by the adulterators have also become more sophisticated. The most frequent method of deception is dilution of the concentrated juice product with sugar syrup and adjusting the flavour and colour levels as necessary. The impedance is a complex quantity composed of real (resistance) and imaginary (reactance) components; it is associated with the qualitative characteristics of the samples. Measurement of the impedance spectrum is achieved by determining impedance within the frequency range (EIS) and is primarily dependent on the structure and chemical composition of the samples. There are chemical components which can have a major effect on impedance: content of water, minerals, salts, organic acids and their degree of dissociation. Electrical Impedance Spectroscopy (EIS) is a nondestructive, rapid and real-time measurement method which does not require any special high-tech calibre measurement devices. EIS includes measurement of the linear electrical response of the material (including the electrode effects) and the subsequent analysis of the response to yield useful information on the physiological, biochemical and chemical properties (Macdonald, 1992). The dielectric properties of food have long been recognised (Zhang, 1995). When cells are damaged, for example during the freezing or heating process, the impedance loses its original character. EIS is a leading technique in various industries (Wu et al., 2005; Nandkumar et al., 2008), in the medical field (Kagan et al., 1977), in pathogen detection (Yang & Bashir, 2008) and in biosensors (Srinivasan et al., 2006; Daniels & Pourmand, 2007). In the food industry, it is used to detect various pathogens and also to check the quality of the food. It represents a powerful detection tool which can be used for real-time non-invasive process parameter control in food (orange juice) (Li, 2003), as a rapid and inexpensive method for the detection of changes during freezing and drying (Wu et al., 2008; Skierucha et al., 2012) and for detecting the fermentation process in wine (Zheng, 2010). Impedance measurement can be used for monitoring the moisture content changes caused by adulteration and the quality changes in fruit and vegetable juices (Katiyar, 2013). In some studies, the content of salt (Karásková et al., 2011; Badhe & Helambe, 2013; Rizo et al., 2013), the amount of supplemented sugar (Guo et al., 2011a; Badhe & Helambe, 2013), the amount of moisture (Li, 2003; Kobayashi et al., 2013), the fat content (Bertemes-Filho et al., 2010) and the quantity of phthalates in water and juices
939
(Zia et al., 2013) were measured, and the dielectric behaviour of beef (Damez et al., 2007; Mahapatra et al., 2010) examined. Other studies have focused on determination of the drying process (Zsivanovits et al., 2012), the ripening process (Guo et al., 2011b; Soltani et al., 2011), the dielectric properties of chestnut flour (Zhu et al., 2012), the detection of coconut water dilution in beverages (Franco et al., 2013), damage to food products caused by freezing (Wu et al., 2008; PérezEsteve et al., 2014) and on storage changes in quality parameters (Ragni et al., 2007). EIS can be applied as an alternative method for determination of the floral origin of honey (Scandurra et al., 2013), for detection of cell damage in biotechnology (Ando et al., 2014), for characterisation of milk, especially useful in determining the quality of dairy products (Halambre & Badhe, 2013) and also for the detection of milk adulteration (Guo et al., 2010; Das et al., 2011). The method is also appropriate for investigating bacterial contamination (Rahman et al., 2013), for determining the safety of frozen and unfrozen products (Vidaček et al., 2008; Fernández-Segovia et al., 2012) and for harvest process monitoring (Niu & Lee, 2000; Mizukami et al., 2007; Euring et al., 2011; Kuson & Terdwongworakul, 2013). EIS measuring can be used in a technological online process, where the results show the coagulation process of soymilk in real time in situ (Li et al., 2011), or during the process of cooking meat and biscuits (Li, 2003; Kobayashi et al., 2013). The present work investigates the possibilities of using the electrical impedance properties of fruits (oranges) and the relationship between impedance and quality criteria is explored. The quality of orange juices can differ according to the type of processing. The adulteration of orange juices has progressed from simple dilution with water and substitution of the juice with cheaper ingredients such as sugar, acid and other types of fruit to masking the adulteration process. It is possible to determine differences in the results of quality juice using such basic markers as acidity of juice, concentrations, sugars, Brix value. EIS was used to analyse orange juice and disclose the changes caused by adding water or a different concentrate of fruit component. The measuring method of impedance spectroscopy is especially applicable in orange juices processing technology. Part of the water is evaporated during the orange concentration production process and the concentrate resulting is then diluted with water. During the production process it is necessary to add the same amount of water to the concentrate as was removed in the previous operation. It is possible to make a mistake and to add the wrong amount of water due to human error or incorrect setting of the production line. The method of impedance measurement presented here can detect any possible aberration during the process to prevent distribution to the customers. This method
Brought to you by | New York University Bobst Library Technical Services Authenticated Download Date | 7/22/15 12:32 AM
940
R. Seidlová et al./Chemical Papers 69 (7) 938–949 (2015)
Table 1. Summary of samples; addition of water (15 vol. % of sample), sugar (80 g L−1 ) and ascorbic acid (0.5 g L−1 ) Sample 1 2 3 4 5 6 7 8
Fig. 1. Glass cell for electrode impedance measurement.
can also be used in other areas of food analysis prior to the application of more detailed analytical methods that are highly expensive (for example can be applied to official sampling during customs procedures, prior to them being dispatched to a Customs Technical Laboratory, etc.).
Experimental Electrode impedance measurement A glass cell of 16 cm × 7 cm × 9 cm was developed and constructed for the measurements of electrochemical properties in liquids. The planar electrodes (70 mm × 90 mm) were of stainless steel, positioned on the opposite sides of the cell at a distance of 158 mm. (See Fig. 1). For measuring, an Agilent 4294 Precision Impedance Analyser (Agilent Technologies, Japan) was used, connected via a 4-TP1M adapter to the cell with planar electrodes placed in a thermostatic bath. The following conditions pertained: voltage amplitude of 0.5 V, frequency range of 40 Hz–1 MHz; a laboratory temperature was measured and maintained constant. In the cell thus constructed, a homogeneous distribution of the electric field in the sample could be assumed due to the dimensions of the electrodes and to the distance of the electrodes. An AC voltage with variable frequency was applied to the electrodes and the subsequent current response was measured; the real and imaginary parts of the impedance were evaluated. The equivalent circuit modelling the behaviour of the system was determined by interpreting the measured spectrum. The measured data were fitted. The elements of the equivalent circuit should have a physical interpretation. The system was described by an electric diagram containing electrical components that modelled the real action that proceeded at the electrode–electrolyte interface. The equivalent circuit consisted of a series and paral-
Juice type Freshly squeezed Freshly squeezed Pfanner 100 % Freshly squeezed Pfanner 100 % Pfanner 100 % Freshly squeezed Pfanner 100 %
Water
Sugar
Ascorbic acid
No Yes Yes No No No No Yes
No No Yes Yes No Yes No No
No No No No No Yes Yes Yes
lel combinations of elements, some of which could be identified with real elements of electrical practice, but some were specific to electrochemistry. The physical interpretation of the resistance of the solution and the charge transfer resistance was the resistor. The capacitor was the capacity of the electrical double layer. Examples of electrochemistry-specific elements are Warburg impedance or a member of a constant phase shift CPE (Constant Phase Element). Warburg impedance describes the diffusion process control; CPE represents either the electrode surface roughness or describes the behaviour of systems where changes in electrochemical activity occur across the electrode. The elements assembled in an equivalent circuit then described the overall behaviour of an electrochemical system and depended, amongst others, on the structure and chemical composition of the samples – content of water, minerals, salts, organic acids and their degree of dissociation. In the case of the fruit juices, the real part was associated with resistance paths through the tissue, i.e. through the cytoplasm and symplast (the symplast of a plant is the inner side of the plasma membrane in which water and low-molecular-mass solutes can diffuse freely.) On the other hand, the imaginary part that was associated with the capacitive paths, for example through the membrane structure, was dependent on the frequency and was formed by inductive and capacitive reactance (Wu et al., 2005). Several samples of commercially produced 100 % orange juices and three variants of orange fruits with a declared origin were measured. All three types of fruits were mechanically squeezed using a kitchen juicer. Measurements were performed at a constant ambient temperature of 25 ◦C, which was continuously monitored. Measurements were repeated fifty times and the values were averaged for further statistical analysis. The present study used the 100 % orange juice Pfanner, ARANCIA, volume 1 L and a freshly squeezed juice. Eight samples were prepared with various combinations of added sugar (sucrose, glucose, fructose in the mass ratio of 1 : 1.1 : 1), ascorbic acid and a specific amount of water (dilution). All prepared samples are detailed in Table 1.
Brought to you by | New York University Bobst Library Technical Services Authenticated Download Date | 7/22/15 12:32 AM
941
R. Seidlová et al./Chemical Papers 69 (7) 938–949 (2015)
Table 2. Basic parameters of samples: refractive index of orange juice, pH value, titratable acidity, formol number and phosphorus content Refractive index, BRIX Sample
1 2 3 4 5 6 7 8
pH
◦
15.8 9.1 15 17 10.1 16.7 11 8.89
2.93 3.27 3.63 3.18 3.62 3.20 2.90 3.16
Acidity
Formol number
Ash phosphorus
g L−1
meq L−1
mg
11.72 8.19 5.18 8.78 6.53 9.92 13.73 9.73
624.9 296.5 780.5 317.1 977.6 920.9 334.6 816.4
179.19 150.82 205.50 209.28 320.16 265.13 223.49 218.61
Fig. 2. Configuration of transformer-based electrodeless method of impedance measurement (a). UG – Generator of measuring signal; Φ1 , Φ2 – magnetic flux in toroid cores of transformers; I1 , I3 – currents in windings of driving and sensing transformers, I2 – current in measured electrolyte; n1 , n2 , n3 – numbers of windings of driving and sensing transformers (n2 = 1). Equivalent circuit. Rx – Impedance of electrolyte in tube; Rz – input impedance of measuring instrument (b).
The impedance spectroscopy spectra, pH, Brix value, titratable acidity, formol number, ash content and phosphorus were measured in all samples. (See Table 2.) Besides the 100 % orange juice Pfanner, in some experiments other commercially produced 100 % orange juices marketed in the Czech Republic were also used. Specifically, these were the following products: HELLO 100 % Pomeranč (Linea Nivnice, Czech republic, batch – 14.2.2015 4:52, Brazil); TOMA 100 % Pomeranč (Pepsico, Czech republic, batch – 08.01.2015 L09: 34*2*008782); HAPPY DAY Pomeranč 100 % (Rauch, Rankweil, Austria, batch – 15.11.2014 16:04 1B0, country of origin: Brazil); RELAX 100 % Pomeranč (Tymbark, Poland, batch – 02.2015 0342228 B3 95872, country of origin: Brazil). Electrodeless impedance measurement The characterisation of disturbing phenomena on a double-layer interface metal electrode-electrolyte by e.g. Constant Phase Element (CPE) (McAdams, 2006) is only an approximation of the real situation. In a number of cases, the contact of the metal with the electrolyte could change the behaviour of the electrochemical or physiological process under observation. It might be difficult to distinguish between those processes taking place inside the electrolyte and those
occurring at the electrode-electrolyte interface. For example, implementation of the electrodeless method can avoid the effects of electrode phenomena and the electrolyte impedance is similar to the behaviour of inductance, as was observed (Bardos et al., 2005). Up to now, these effects were at least partially eliminated by using the same type of electrodes in all laboratories performing similar experiments, i.e. presuming that the same “distortion effects” occur in all experiments. Otherwise, the transferability of results would be jeopardised. In addition, electrodes from precious metals are expensive and generally can be used only once (difficulties with cleaning and sterilisation). For these reasons, methods of impedance measurement requiring no electrodes could be very useful. Unfortunately, up to now these methods have not been commercially available. One possible way of electrodeless measurement is based on the transformer principle (Salamon & Svitok, 1959). The conversion of a current through an isolated tube filled with a measured solution to a voltage by transformer (inductive method) has long been recognised (Salamon & Svitok, 1959) as an alternative for measuring the conductivity of electrolytes, but it is rarely used (Yang et al., 2013). The principle of the method is depicted in Figs. 2a and 2b. The isolated tube forms the secondary windings of the driving transformer and the primary winding of the current
Brought to you by | New York University Bobst Library Technical Services Authenticated Download Date | 7/22/15 12:32 AM
942
R. Seidlová et al./Chemical Papers 69 (7) 938–949 (2015)
Fig. 3. Cascade connection of integrating cells (a). Cascade connection of differentiating circuits (b).
(sensing) transformer. Normally, the voltage on the secondary winding n3 of the current transformer is measured. Due to the complicated equivalent circuit of the current transformer, the transfer function of the transformer is far from ideal and quite a complicated correction procedure is required in order to achieve the same transfer factor for all frequencies. In oscilloscope current probes, which might also be used for current I2 measurement, this difficult task is accomplished by carefully chosen circuit correction components. Each probe must be corrected individually and this costly process substantially increases the price of the probe and at the same time substantially decreases the sensitivity. Fortunately, in the present case dealing with harmonic signals, the inverse transfer function approach (“deconvolution”, calibration) could be implemented. The response of the output voltage to current I2 through the known impedance for different frequencies is measured (calibration curve). The output values measured for unknown impedances for different frequencies are then corrected using data obtained by calibration. The impedance of electrolyte has a complicated character and can be modelled by a combination of several basic cells containing terms (jω)ν with the noninteger exponent v. Terms with non-integer exponent correspond to the operations of fractional derivatives (Jacquelin, 1991, 1994; Jean, 1997). A simple example is the fractional derivative in the expression for voltage on a modified impedance element of inductance-type expressed as: uL = L
dv i dv t
(1)
A number of fractional derivatives of various functions can be found in the tables of Rieman– Liouville transforms (Debnath, 2012). An alternative to impedance modelling based on fractional derivatives is an approximation of the decreasing impedance with frequency by the cascade connection of simple cells containing resistors Rn and capacitors Cn (see Figs. 3a and 3b). In simple cases, the capacitor can be replaced with a member of the constant phase shift, defined by the parameters of CPE-T and CPE-P and its impedance
Fig. 4. Simplest equivalent circuit for interface metal electrode – electrolyte (a). CPE approximation (b).
can be expressed as: Z=
1 T (jω)
P
(2)
where T is the frequency of the independent member (“pseudocapacity”) with the DC unit, ω is the angular frequency. The exponent P ranges between 0 and 1. When the exponent P is equal to zero, the impedance of the element corresponds to the resistance. When P equals 1, the impedance of the element corresponds to the capacitor with capacity T. Appropriate equivalent circuits for the electrode method, taking into account the interface metal electrode – electrolyte, are shown in Figs 4a and 4b. The actual values of Rn , Cn or CPE can be found by using commercial impedance-fitting programmes or by using the algorithm based on an ant colony, presented below. This approach appears to be closer to physical reality as the electrolyte behaviour is similar to that of electrical circuits with distributed parameters. The experimental set-up depicted in Fig. 5 was constructed for the purpose of confirming the transformer-based principle electrodeless impedance measurement. The main aim was to verify the basic properties of the system. Impedance measurements were performed and the modulus and phase were evaluated. The resistivity of the electrolyte can be determined from these data. Assuming a constant current density along the tube with
Brought to you by | New York University Bobst Library Technical Services Authenticated Download Date | 7/22/15 12:32 AM
R. Seidlová et al./Chemical Papers 69 (7) 938–949 (2015)
943
Fig. 5. Experimental set-up comprising two toroid transformers in shielded cases and tube filled with measured electrolyte (juice) (a). The length of the tube is l = 611 mm, internal diameter d = 10 mm. driving transformer (1), sensing transformer (2), tube with electrolyte (3) and lock-in amplifier (4). Experimental set-up, detail (b).
the electrolyte, the relation between the electrical resistivity R and specific resistivity as: R=ρ
4l πd2
(3)
where is the resistivity of the electrolyte, l = 611 mm is the length of the tube and d = 10 mm is the internal diameter of the tube. Hence, the specific resistivity of the electrolyte is determined by the relation: ρ=
Rπd2 4l
(4)
In cases where the imaginary part of impedance Z is much less than the real, it can be expressed as: R = Re(Z)
(5)
Ant colony algorithm The ant colony algorithm (ACA) is inspired by the behaviour of ant colonies. The ants emit an aromatic substance, known as a pheromone, on their way to a food source. The pheromone quantity depends on the length of the path and the quality of the food source discovered. An ant chooses a path in relation to the intensity of the pheromone. The pheromone trail evaporates over time if no further pheromone is emitted. Other ants are attracted to follow the pheromone trail. The pheromone trail on paths leading to rich food sources close to the nest will be more frequented and, accordingly, will grow faster. In this way, the best solution has a more intensive pheromone and a higher probability of being chosen (Wang & Wu, 2001). The behaviour of ant colonies, as described above, can be used to solve combinatorial optimisation problems. Several algorithms have been developed based
on the principle of ant colonies, such as the shortest path-finding, solution of travelling salesman problem or classification rules discovery (Seidlová et al., 2012). The ACA principle can also be modified to fit the components values of the equivalent circuit. In this modified ACA algorithm, food sources are the equivalent circuit values and the amount of the pheromone is proportional to the difference between the analytical and simulated values. The probability of an ant choosing a path to the food source depends on this amount. In the present case, the amount of the pheromone is proportional to the difference between the measured and fitted impedances. The algorithm can be expressed as: there are m parameters for optimisation denoted as pi (1 ≤ i ≤ m), the set of N possible non-zero value of these parameters can be labelled as Qpi . Each ant starts from the set Qpi based on the pheromone τ j (Qpi ) of each element in set and the transition probability may be stated as: τj (QPi ) (6) P τjk (QPi ) = N g=1 τg (QPi An ant randomly chooses one element in set Qpi independently and a new level of pheromone is calculated: τj (QPi )(t + n) = ρτj (QPi )(t) + ∆τj (QPi )
(7)
where ρ is the evaporation coefficient, is the increased pheromone on j-th element which is determined using the difference between the analytical and measured output. The smaller the difference, the greater the volume of the pheromone (Jennings et al., 2008). The pheromone levels on all elements are normalised following their update so that the total amount of pheromone is constant and equal to Q.
Brought to you by | New York University Bobst Library Technical Services Authenticated Download Date | 7/22/15 12:32 AM
944
R. Seidlová et al./Chemical Papers 69 (7) 938–949 (2015)
This process is repeated until the optimised solution is found or the maximum admissible iteration times are reached (Chen et al., 2006). It was necessary to perform a careful optimisation of the parameters of this algorithm in order to achieve the required accuracy in an admissible calculation time. There is no theory support for the parameters’ setting; the optimised solution was achieved heuristically through repeated matching and adjustment. The specific values of the optimised parameters are given under Results and discussion. A diagram of this algorithm is shown in Fig. 6.
Results and discussion The impedances spectra for a number of samples of commercially produced and hand-squeezed orange juices were measured. There are several ways of displaying the frequency response data. The Nyquist plot displays the dependence of the imaginary part of impedance on the real part of impedance, using frequency as a parameter in the plot. The modified Nyquist plot for commercially produced orange juices is in Fig. 7. To find an equivalent circuit, a constant phase element (CPE) approach was used. CPE is an equivalent electrical circuit component that models the behaviour of a double-layer which is an imperfect capacitor. The electrical impedance of CPE can be calculated by using Eq. (2). The case P = 1 describes an ideal capacitor while the case P = 0 describes a pure resistor. The identified equivalent circuits corresponding to the present measurement are shown in Fig. 8, where R represents the bulk sample solution resistance, the CPE1 and CPE2 model the behaviour
Fig. 6. Diagram of modified ACA algorithm.
of a double-layer and the charge-transfer resistance associated with the double-layer (Fig. 9). The complex impedance for the equivalent circuit
Fig. 7. Relationship of real and imaginary parts of impedances (Nyquist plot) for several commercially produced orange juices marketed in the Czech Republic (frequency changes from 40 Hz to 1 MHz).
Brought to you by | New York University Bobst Library Technical Services Authenticated Download Date | 7/22/15 12:32 AM
R. Seidlová et al./Chemical Papers 69 (7) 938–949 (2015)
945
Fig. 8. Equivalent circuit scheme for commercially produced orange juices in the arrangement of flat electrodes.
Fig. 10. Change in module of impedance with concentration measured by electrode method.
Fig. 9. Model of interface electrode–electrolyte.
in Fig. 8 can be described as follows: −1 P T1 (jω) 1 P2 + T2 (jω) Z= P R T1 (jω) 1 + 1
(8)
The values of the equivalent circuit elements were calculated with the help of the ant colony algorithm described above. This algorithm, implemented in MS Visual Studio 2010, includes several important parameters: Number of Ants, Pheromone Evaporation Factor, Total Amount of Pheromone, Initial Pheromone Value and Number of Loops. These parameters were analysed and optimised. The Pheromone Evaporation Factor and Number of Ants have the greatest influence on the accuracy and calculation time. Increasing the evaporation factor will result in a quicker convergence process, but reduced accuracy. Decreasing the evaporation factor requires the use of a greater number of ants. The calculation time increases with the number of ants. From previous experience (Seidlová et al., 2013) and experimental results, these parameters have been set as follows: number of loops Nc = 1000, pheromone evaporation factor ρ = 0.7, total amount of pheromone Q = 900, initial pheromone value τ 0 = 10, number of ants M = 500 or 1500, deviation E is limited to 2. (See Fig. 6). The values of the equivalent circuit element determined with help of the ant colony algorithm are presented in Tables 3 and 4. Fig. 10 shows the decrease in the impedance module with the concentration of orange juice in the arrangement of the flat electrode method. Fig. 11 presents and compares the results of the electrode and electrodeless methods. There is a remarkable difference in the curves. Saturation occurs, in case of the electrode method, during the measurement of orange juices with concentrations above 75 %.
Fig. 11. Measurement for samples of orange juice with electrode in contact with electrolyte and electrodeless methods. Measured at frequency of 1 MHz.
Fig. 12. Change in module of impedance with amount of added sugar, measured by electrode method.
The saturation is probably caused by the effects at the interface electrode – electrolyte (juice). Fig. 12 shows the increase in the impedance module with the amount of added sugar in the arrangement of the flat electrode method. Fig. 13 shows the increase in the impedance module with the amount of added ascorbic acid. Fig. 14 shows that the real part of the impedance increases with frequency in the arrangement of the electrodeless method. The measured impedance has the character of a serial connection of resistance and
Brought to you by | New York University Bobst Library Technical Services Authenticated Download Date | 7/22/15 12:32 AM
946
R. Seidlová et al./Chemical Papers 69 (7) 938–949 (2015)
Table 3. Values of equivalent circuit elements for freshly squeezed orange juices in the arrangement of flat electrodes: T and P are parameters of CPE1 and CPE2 elements, resp. T is a constant in F cm−2 sP −1 and P is related to the angle of rotation of a purely capacitive line on the complex plane plots (0 ≤ P ≤ 1), see Eq. (2); sugar added denotes the addition of 8 g per 100 mL of sugar: sucrose/glucose/fructose in the ratio 1 : 1.1 : 1; acid addition means of 0.5 g per 100 mL of ascorbic acid addition R
104 ·TCPE1
PCPE1
108 ·TCPE2
PCPE2
Water dilution?
Sugar added?
Ascorbic acid added?
Ω
F cm−2 sP −1
–
F cm−2 sP −1
–
Yes No Yes No
Yes Yes No No
Yes No No Yes
122.60 114.40 98.47 89.68
4.6786 4.8262 5.4326 4.5531
0.78384 0.79779 0.74968 0.83312
4.99 8.56 4.23 14.0
0.5127 0.50223 0.54095 0.49521
Table 4. Values of equivalent circuit elements for commercially produced orange juices in the arrangement of flat electrodes; sugar added denotes the addition of 8 g per 100 mL of sugar: sucrose/glucose/fructose in the ratio 1 : 1.1 : 1; acid addition means of 0.5 g per 100 mL of ascorbic acid addition R
104 ·TCPE1
PCPE1
1011 ·TCPE2
PCPE2
Water dilution?
Sugar added?
Ascorbic acid added?
Ω
F cm−2 sP −1
–
F cm−2 sP −1
–
Yes No Yes No
Yes Yes No No
No Yes Yes No
110.1 99.75 84.48 78.85
5.2392 5.3288 5.5724 5.6126
0.77581 0.79155 0.80381 0.7817
5.38 4.65 6.28 1.75
0.91738 0.93544 0.91995 0.85781
Fig. 15. Equivalent circuit scheme for commercially produced orange juices in arrangement of electrodeless method. Fig. 13. Change in module of impedance with amount of added ascorbic acid, measured by electrode method.
Fig. 14. Change in module of impedance with amount of added sugar, measured by electrodeless method.
inductance. The equivalent circuit scheme is shown in Fig. 15 and the values of equivalent circuit elements found for 100 % juice Pfanner are in Table 5. The
Fig. 16. Spectrum of real and imaginary parts of impedance for commercially produced orange juice Pfanner, in the arrangement of electrodeless method.
fitted and measured values of the real and imaginary parts of impedance for commercially produced orange juice Pfanner in the arrangement of the electrodeless method are shown in Fig. 16.
Brought to you by | New York University Bobst Library Technical Services Authenticated Download Date | 7/22/15 12:32 AM
947
R. Seidlová et al./Chemical Papers 69 (7) 938–949 (2015)
Table 5. Values of equivalent circuit elements for commercially produced orange juices in arrangement of electrodeless method Sample Juice Pfanner 100 % Added 2 g of sugar per Added 3 g of sugar per Added 4 g of sugar per Added 5 g of sugar per
100 100 100 100
mL mL mL mL
R1 /Ω
R2 /Ω
L1 /H
24612 24791 26137 27230 28829
17015 21345 24741 24965 28725
0.05274 0.06879 0.06816 0.07436 0.09204
No consideration is given to the self-heating of the juice due to the continuous flow of current. The voltage across the tube is only 0.5 V and for minimal electrolyte resistances in the order of hundreds of ohms the current is approximately in the order of milliamps. The power dissipated in the tube filled with the measured electrolyte is, hence, in the order of only a few milliwatts. Due to the relatively large area of the tube surface, its heat conductivity is high, hence any increase in temperature by current is negligible.
Conclusions The present study investigated the possibilities of using measuring with the impedance method for determining the quality of orange juices. Electrical impedance measurement is a powerful tool used for real-time non-invasive process parameter control. The method is very fast, effective and offers a low-cost investigation of the product. The results showed that the addition of water, sugar or ascorbic acid had an influence on the impedance spectra. The electrical properties of orange juices can be described with the help of an equivalent circuit. The parameters of these circuits are dependent, among others, on the production method of the juices and may reveal some methods of falsification. In order to find the equivalent circuits, the constant phase element approximation was adopted to model the electrical response of the orange juice. The values of the equivalent circuit elements were fitted using an algorithm based on the ant colony algorithm. The original algorithm was designed and used for another type of tasks. The ACA was modified to be used as a new and original application as demonstrated in this paper. The simulation results revealed that the impedance of the equivalent circuit calculated by this algorithm are in good agreement with the experimental results throughout the frequency range. The algorithm thus developed is easy to apply, it is robust, gives good results across the range of the spectrum and facilitates the distributed parallel computation. In this study, two different types of measurement – electrode and electrodeless were compared. It was confirmed that the metal electrode in contact with the electrolyte affects the saturation of the electrode, hence the measurement of concentrations of 75 % and higher is inaccurate. The inductive impedance nature
of the electrolyte has a striking feature; it shows that measurement by contact methods, i.e. with electrodes, conceals (“masks”) electrochemical phenomena and the electrolytes measured thereby act as capacitors. The proposed electrodeless method is generally applicable to measurement of the electrochemical properties of electrolytes. It substantially facilitates the finding of the electrical circuit model of electrolyte and analysis of the corresponding electrochemical processes. A further advantage of this method using impedance measuring is that it enables the analysis of more samples within a shorter time. This method can be recommended for fast monitoring in the processing line for producing fresh juice.
Symbols Cn CPE d i I j l L P R Rn t T uL Z ν ρ ω
capacitors constant phase element diameter current density current imaginary unit length induction P member of CPE resistivity resistors time T member of CPE voltage impedance dimensionless parameter specific resistivity angular frequency
F m A m−2 A m H Ω Ω s
F cm−2 sP −1 V Ω Ωm s−1
References Ando, Y., Mizutani, K., & Wakatsuki, N. (2014). Electrical impedance analysis of potato tissues during drying. Journal of Food Engineering, 121, 24–31. DOI: 10.1016/j.jfoodeng. 2013.08.008. Badhe, S. G., & Helambe, S. N. (2013). Electrical impedance analysis of commonly used preservatives NaCl and C12 H22 O11 . Science Research Reporter, 3, 239–242. Bardos, A., Zare, R. N., & Markides, K. (2005). Inductive behavior of electrolytes in AC conductance mea-
Brought to you by | New York University Bobst Library Technical Services Authenticated Download Date | 7/22/15 12:32 AM
948
R. Seidlová et al./Chemical Papers 69 (7) 938–949 (2015)
surements. Chemical Physics Letters, 402, 274–278. DOI: 10.1016/j.cplett.2004.12.047. Bertemes-Filho, P., Valicheski, R., Pereira, R. M., & Paterno, A. S. (2010). Bioelectrical impedance analysis for bovine milk: Preliminary results. Journal of Physics: Conference Series, 224, 012133. DOI: 10.1088/1742-6596/224/1/012133. Chen, Y. G., Gu, X., Shen, Y. H., & Xing, S. Z. (2006). Optimization of active power filter system pi parameters based on improved ant colony algorithm, mechatronics and automation. In Proceedings of the 2006 International Conference on Mechatronics and Automation, June 25–28, 2006 (pp. 2189–2193). Henan, Luoyang, China: IEEE. DOI: 10.1109/icma.2006.257633. Damez, J. L., Clerjon, S., Abouelkaram, S., & Lepetit, J. (2007). Dielectric behavior of beef meat in the 1–1500 kHz range: Simulation with the Fricke/Cole–Cole model. Meat Science, 77, 512–519. DOI: 10.1016/j.meatsci.2007.04.028. Daniels, J. S., & Pourmand, N. (2007). Label-free impedance biosensors: Opportunities and challenges. Electroanalysis, 19, 1239–1257. DOI: 10.1002/elan.200603855. Das, S., Sivaramakrishna, M., Biswas, K., & Goswami, B. (2011). Performance study of a “constant phase angle based” impedance sensor to detect milk adulteration. Sensors and Actuators A: Physical, 167, 273–278. DOI: 10.1016/j.sna. 2011.02.041. Debnath, L. (2012). Nonlinear partial differential equations for scientists and engineers (pp. 675–687). New York, NY, USA: Springer. DOI: 10.1007/978-0-8176-8265-1. Euring, F., Russ, W., Wilke, W., & Grupa, U. (2011). Development of an impedance measurement system for the detection of decay of apples. Procedia Food Science, 1, 1188–1194. DOI: 10.1016/j.profoo.2011.09.177. Fernández-Segovia, I., Fuentes, A., Ali˜ no, M., Masot, R., Alca˜ niz, M., & Barat, J. M. (2012). Detection of frozenthawed salmon (Salmo salar) by a rapid low-cost method. Journal of Food Engineering, 113, 210–216. DOI: 10.1016/j. jfoodeng.2012.06.003. Franco, A. P., Tadini, C. C., & Gut, J. A. W. (2013). Dielectric properties of simulated green coconut water from 500 to 3,000 MHz at temperatures from 10 to 80 C◦ . In Proceedengs of the 2013 AIChE Annual Meeting Global Challenges for Engineering a Sustainable Future, November 3–8, 2013. San Francisco, CA, USA: AIChE. Guo, W. C., Zhu, X. H., Liu, H., Yue, R., & Wang, S. J. (2010). Effects of milk concentration and freshness on microwave dielectric properties. Journal of Food Engineering, 99, 344– 350. DOI: 10.1016/j.jfoodeng.2010.03.015. Guo, W. C., Liu, Y., Zhu, X. H., & Wang, S. J. (2011a). Temperature-dependent dielectric properties of honey associated with dielectric heating. Journal of Food Engineering, 102, 209–216. DOI: 10.1016/j.jfoodeng.2010.08.016. Guo, W. C., Zhu, X. H., Nelson, S. O., Yue, R., Liu, H., & Liu, Y. (2011b). Maturity effects on dielectric properties of apples from 10 to 4500 MHz. LWT – Food Science and Technology, 44, 224–230. DOI: 10.1016/j.lwt.2010.05.032. Halambre, S. N., & Badhe, S. G. (2013). Characterization of milk using impedance analysis technique. Deccean Current Science International Research Journal, 8, 132–136. Jacquelin, J. (1991). A number of models for CPA impedances of conductors and for relaxation in non-Debye dielectrics. Journal of Non-Crystalline Solids, 131–133, 1080–1083. DOI: 10.1016/0022-3093(91)90728-o. Jacquelin, J. (1994). Theoretical impedance of rough electrodes with smooth shapes of roughness. Electrochimica Acta, 39, 2673–2684. DOI: 10.1016/0013-4686(94)00296-7. Jean, J. (1997). The phasance concept: A review. Current Topics in Electrochemistry, 4, 127–136.
Jennings, A. L., Ordonez, R., & Ceccarelli, N. (2008). An Ant Colony Optimization using training data applied to UAV way point path planning in wind. In Proceedings of the IEEE Swarm Intelligence Symposium, September 21–23, 2008. St. Louis, Missouri, USA: IEEE Computational Intelligence Society. DOI: 10.1109/sis.2008.4668302. Kagan, R. L., Schuette, W. H., Zierdt, C. H., & MacLowry, J. D. (1977). Rapid automated diagnosis of bacteremia by impedance detection. Journal of Clinical Microbiology, 5, 51–57. Karásková, P., Fuentes, A., Fernández-Segovia, I., Alca˜ niz, M., Masot, R., & Barat, J. M. (2011). Development of a lowcost non-destructive system for measuring moisture and salt content in smoked fish products. Procedia Food Science, 1, 1195–1201. DOI: 10.1016/j.profoo.2011.09.178. Katiyar, V. (2013). Food adulteration: The demonic onslaught on health. Saarbr¨ ucken, Germany: Lambert Academic Publishing. Kobayashi, A., Mizutani, K., Wakasuki, N., & Maeda, Y. (2013). Changes of electrical impedance characteristic of pork in heating process. International Proceedings of Chemical, Biological & Environmental Engineering, 50, 74–78. DOI: 10.7763/ipcbee.2013.v50.16. Kuson, P., & Terdwongworakul, A. (2013). Minimally-destructive evaluation of durian maturity based on electrical impedance measurement. Journal of Food Engineering, 116, 50–56. DOI: 10.1016/j.jfoodeng.2012.11.021. Li, X. B. (2003). Impedance spectroscopy for manufacturing control material physical properties. Ph.D. thesis, Seattle, WA, USA: University of Washington. Li, X., Toyoda, K., & Ihara, I. (2011). Coagulation process of soymilk characterized by electrical impedance spectroscopy. Journal of Food Engineering, 105, 563–568. DOI: 10.1016/j.jfoodeng.2011.03.023. Macdonald, J. R. (1992). Impedance spectroscopy. Annuals of Biomedical Engineering, 20, 289–305. DOI: 10.1007/bf02368 532. Mahapatra, A. K., Jones, B. L., Nguyen, C. N., & Kannan, G. (2010). Experimental determination of the electrical resistivity of beef. Agricultural Engineering International: CIGR Journal, 12, 124–128. Maireva, S., Usai, T., & Manhokwe, S. (2013). The determination of adulteration in orange based fruit juices. International Journal of Science and Technology, 2, 365–372. McAdams, E. (2006). Bioelectrodes. In J. G. Webster (Ed.), Encyclopedia of medical devices and instrumentation. New York, NY, USA: Wiley. DOI: 10.1002/0471732877.emd013. Mizukami, Y., Yamada, K., Sawai, Y., & Yamaguchi, Y. (2007). Measurement of fresh tea growth using electrical impedance spectroscopy. Agricultural Journal, 2, 134–139. Nagy, S., Attaway, J. A., Rhodes, M. E. (1988). Adulteration of fruit juice beverages. New York, NY, USA: CRC Press. Nandakumar, V., La Belle, J. T., Reed, J., Shah, M., Cochran, D., Joshi, L., & Alford, T. L. (2008). A methodology for rapid detection of Salmonella typhimurium using label-free electrochemical impedance spectroscopy. Biosensors and Bioelectronics, 24, 1039–1042. DOI: 10.1016/j.bios.2008.06.036. Niu, J., & Lee, J. Y. (2000). A new approach for the determination of fish freshness by electrochemical impedance spectroscopy. Journal of Food Science, 65, 780–785. DOI: 10.1111/j.1365-2621.2000.tb13586.x. Pérez-Esteve, E., Fuentes, A., Grau, R., Fernández-Segovia, I., Masot, R., Alca˜ niz, M., & Barat, J. M. (2014). Use of impedance spectroscopy for predicting freshness of sea bream (Sparus aurata). Food Control, 35, 360–365. DOI: 10.1016/j.foodcont.2013.07.025. Ragni, L., Al-Shami, A., Berardinelli, A., Mikhaylenko, G., & Tang, J. (2007). Quality evaluation of shell eggs during stor-
Brought to you by | New York University Bobst Library Technical Services Authenticated Download Date | 7/22/15 12:32 AM
R. Seidlová et al./Chemical Papers 69 (7) 938–949 (2015)
age using a dielectric technique. Transactions of the ASABE, 50, 1331–1340. DOI: 10.13031/2013.23610. Rahman, M. S. A., Mukhopadhyay, S. C., Yu, P. L., Goicoechea, J., Matias, I. R., Gooneratne, C. P., & Kosel, J. (2013). Detection of bacterial endotoxin in food: New planar interdigital sensors based approach. Journal of Food Engineering, 114, 346–360. DOI: 10.1016/j.jfoodeng.2012.08.026. Rizo, A., Fuentes, A., Fernández-Segovia, I., Masot, R., Alca˜ niz, M., & Barat, J. M. (2013). Development of a new salmon salting–smoking method and process monitoring by impedance spectroscopy. LWT – Food Science and Technology, 51, 218–224. DOI: 10.1016/j.lwt.2012.09.025. Salamon, M., & Svitok, P. (1959). A low frequency electrodeless conductometer for measuring the electrical conductivity of solutions. Abingdon, UK: Atomic Energy Authority. Scandurra, G., Tripodi, G., & Verzera, A. (2013). Impedance spectroscopy for rapid determination of honey floral origin. Journal of Food Engineering, 119, 738–743. DOI: 10.1016/j.jfoodeng.2013.06.042. Seidlová, R., Poživil, J., & Hanta, V. (2012). Classification rule extracting with ant colony algorithms. In Proceedings of the 39th International Conference of Slovak Society of Chemical Engineering, May 25–29, 2012 (pp. 664–671). Bratislava, Slovakia: Slovak Society of Chemical Engineering. Seidlová, R., Seidl, J., Poživil, J., & Hanta, V. (2013). Application of ant colony algorithm to model electrical impedance spectra of orange juices. In Proceedings of the 40th International Conference of Slovak Society of Chemical Engineering, May 27–31, 2013 (pp. 903–910). Bratislava, Slovakia: Slovak Society of Chemical Engineering. Skierucha, W., Wilczek, A., & Szyplowska, A. (2012). Dielectric spectroscopy in agrophysics. International Agrophysics, 26, 187–197. DOI: 10.2478/v10247-012-0027-5. Soltani, M., Alimardani, R., & Omid, M. (2011). Evaluating banana ripening status from measuring dielectric properties. Journal of Food Engineering, 105, 625–631. DOI: 10.1016/j.jfoodeng.2011.03.032. Srinivasan, B., Tung, S., Li, Y. B., & Varshney, M. (2006). Simulation of electrical impedance based microfluidic biosensor for detection of E. coli cells. In Proceedings of the COMSOL Users Conference 2006, October 22–24, 2006. Boston, MA, USA: COMSOL. Vidaček, S., Medi´c, H., Botka-Petrak, K., Nežak, J., & Petrak, T. (2008). Bioelectrical impedance analysis of frozen sea bass (Dicentrarchus labrax). Journal of Food Engineering, 88, 263–271. DOI: 10.1016/j.jfoodeng.2008.02.010.
949
Wang, L., & Wu, Q. D. (2001). Ant system algorithm for optimization in continuous space. In Proceedings of the 2001 IEEE International Conference on Control Applications, September 7–7, 2011. Mexico City, Mexico: IEEE. DOI: 10.1109/cca.2001.973897. Wu, J., Ben, Y. X., & Chang, H. C. (2005). Particle detection by electrical impedance spectroscopy with asymmetricpolarization AC electroosmotic trapping. Microfluidics and Nanofluidics, 1, 161–167. DOI: 10.1007/s10404-004-0024-5. Wu, L., Ogawa, Y., & Tagawa, A. (2008). Electrical impedance spectroscopy analysis of eggplant pulp and effects of drying and freezing–thawing treatments on its impedance characteristics. Journal of Food Engineering, 87, 274–280. DOI: 10.1016/j.jfoodeng.2007.12.003. Yang, L., & Bashir, R. (2008). Electrical/electrochemical impedance for rapid detection of foodborne pathogenic bacteria. Biotechnology Advances, 26, 135–150. DOI: 10.1016/j. biotechadv.2007.10.003. Yang, B., Meng, F., & Dong, Y. G. (2013). A coil-coupled sensor for electrolyte solution conductivity measurement. In Proceedings of the 2013 2nd International Conference on Measurement, Information and Control, August 16–18, 2013. Harbin, Heilongjiang, China: IEEE. DOI: 10.1109/mic.2013. 6757915. Zhang, H. (1995). Electrical properties of foods. Food Engineering, 1, 152. Zheng, S. (2010). An investigation on electrical properties of major constituents of grape must under fermentation using of electrical impedance spectroscopy. Ph.D. thesis. Melbourne, Australia: RMIT University. Zhu, X. H., Guo, W. C., Wu, X. L., & Wang, S. J. (2012). Dielectric properties of chestnut flour relevant to drying with radiofrequency and microwave energy. Journal of Food Engineering, 113, 143–150. DOI: 10.1016/j.jfoodeng.2012.04.014. Zia, A. I., Syaifudin, A. R. M., Mukhopadhyay, S. C., Yu, P. L., Al-Bahadly, I. H., Gooneratne, C. P., Kosel, J., & Liao, T. S. (2013). Electrochemical impedance spectroscopy based MEMS sensors for phthalates detection in water and juices. Journal of Physics: Conference Series, 439, 012026. DOI: 10.1088/1742-6596/439/1/012026. Zsivanovits, G., Brashlyanova, B., & Karabadzhov, O. (2012). Dielectric impedance monitoring of heat pump drying of apple slices. Journal of Food Physics, 24–25, 50–58.
Brought to you by | New York University Bobst Library Technical Services Authenticated Download Date | 7/22/15 12:32 AM