Impedance Spectra of Different Capacitor Technologies
ANP109 BY DR. RENÉ KALBITZ
01. INTRODUCTION AND THEORETICAL BACKGROUND
Impedance and capacitance spectra (or scattering parameters) are common representations of frequency dependent electrical properties of capacitors. The interpretation of such spectra provides a wide range of electrochemical, physical and technical relevant information. Those need to be separated from the ever-present measurement artifacts as well as parasitic effects. Since it is sometimes not possible to provide all data in the data sheet, the engineer may have to utilize measured spectra to choose the suitable component for its circuit design. To provide the best possible database, Würth Elektronik eiSos has implemented the online tool REDEXPERT, were spectra but also other measurements are provided. In this note, we will recap the properties of such spectra and discuss how basic electrical characteristics can be inferred from it.
1.1 Equivalent Circuit of Capacitors
With the circuit, shown in Abbildung 1, it is possible to model frequency dependent impedance spectra of all capacitor types ranging from multilayer ceramic capacitor (MLCC) to Supercapacitors (SCs). 
Figure 1: Standard equivalent circuit as used for capacitors
The formula sign CS is the pure capacitance, which does not exist on its own as an electrical component. Any real capacitor has losses, which “slows down” the charge storing process. This phenomenon is modeled by the pure ohmic equivalent series resistance RESR (ESR). The resistance of the current collector and the leads also contribute to the ESR. The pure lossless capacitance is defined by a differential:
with dQ as the change of number of charges at the capacitor interface and dV as the change of voltage at the capacitor. Any alternating current in a metal conductor induces a magnetic field that opposes the current. In our model, this property is described by the equivalent series inductance LESL (ESL). Sometimes it is also referred to as parasitic inductance. CS, RESR as well as LESL are the most important parameters, necessary to describe the majority of all spectra. In the most basic approach, they are constants and do not change with frequency, which is sufficiently accurate for electrical engineering.
The loss of charges over time, i.e. the leakage current, is described in good approximation by the pure ohmic resistance RLeak. Usually RLeak is magnitudes larger than R_ESR and can often be neglected, i.e RLeak→ ∞. As we will see, its effect may only be visible in the spectra at very low frequencies below 1 Hz or so. A correct description of leakage current is, however, a physically complex issue, which may depend on further parameters such as pre-poling times and temperature. For technical measurement reasons, it is therefore common practice not to state RLeak but the value of leakage current, along with its measurement conditions in the datasheet.
This equivalent circuit in Abbildung 1 also yields the possibility to model practically any
- environmental (e.g. temperature) or
- nonlinear frequency dependency.
In this case, all model parameters are simply replaced by suitable mathematical functions or entire circuit sections are replaced by distributed networks. 
1.2 The Impedance and Capacitance Spectra
In the following paragraph, we define frequently used terms and measured quantities, such as capacitance and impedance.  The above circuit can be expressed as frequency dependent complex impedance , capacitance , scattering parameter (S-Paramter) , permittivity or any other measureable complex electrical quantity. Passive components are often characterized in terms of capacitance and impedance. We will therefore lay the focus on these two quantities. Impedance , is a complex quantity, with Re() and Im() as real and imaginary part, respectively. It is often expressed in terms of its magnitude and phase angle ϕ,
In a complex plane, as given in Figure 2, ϕ describes the angle between Re() (abscissa) and the complex vector . Physically, || represents the ratio of the voltage amplitude to the current amplitude, while ϕ gives the phase difference between voltage and current at a given frequency. The phase angle ϕ is related to the loss angle with
Figure 2: Vector representation of impedance in the complex plane. RLeak is nglected fot the sake of simplicity
In electrical engineering, it is also common to use magnitude || and its equivalent series resistance RESR = Re(). With the model in Figure 1 the equivalent series resistance is the real part of the impedance. In order to graphically show the relation between model and complex quantity , all model parameters (apart from RLeak) are also given in Figure 2 (Mathematical description is given in the appendix.) The impedance may also be transferred into complex capacitance with
All these quantities, such as Re(), Im(), || or δ, can be measured with impedance or network analyzers. Any electronic part (not only capacitors) can be characterized by a pair of frequency dependent quantities, such as Re() and Im() or Re() and Im(). However, it is only due to equivalent circuits such as given in Figure 1, by which measurement results can be interpreted. The model (also referred to as standard model) provides the mathematical means to extract the electric parameters CS, RESR, LESL and RLeak. It is not only possible to use the model for the extraction of parameters, it can also be used to calculate theoretical spectra.
By changing CS, RESR, LESL, RLeak, it is possible to describe or calculate the basic frequency behavior for all capacitors. This is exemplary demonstrated for impedance and capacitance spectra of a 4.7 µF and a 50 F capacitor, given in Figure 3 and Figure 4, respectively. In Figure 4, dashed-dotted lines designate the pure capacitive and inductive contributions. The parameters for the two examples are as follows:
- Supercapacitor (WCAP-STSC) with CS = 50 F, RESR= 5 mΩ, LESL = 5 nH and RLeak= 10 MΩ,
- Film capacitor (WCAP-FTBE) with CS = 4.7 µF, RESR= 15 mΩ, LESL = 5 nH and RLeak= 10 MΩ.
The parameters were chosen such, as to fit to actual Würth Elektronik eiSos products found under the match code for film capacitors WCAP-FTBE (4.7 µF) and SCs WCAP-STSC (50 F). In these plots, CS, RESR, LESL and RLeak were assumed to be constants and independent of frequency (Table 1). This assumption is in most cases in good agreement with actual measurements. However, especially for the RESR a frequency dependence in real measurements is noticeable, as will be discussed in the following sections.
Table 1: Electrical Parameters used for calculation of spectra.
As mentioned above, we may also think of it the other way round. With this single model, it is possible to derive the product parameters from a measured curve.
Before we look at measured graphs, it is worth the while to have a look at theoretical graphs. They have the advantage that they can be generated for any frequency range, which allows the representation of all features, such as characteristic frequencies, in one graph. 
Figure 3: Impedance spectra || for WCAP- FTBE (Top) and WCAPSTSC (Bottom) as calculated from the standard model
Figure 4: Capacitance spectra Re (), as calculated from the standard model. Graph for WCAP-STSC (orange) corresponds to left hand ordinate and the graph for the WCAP-FTBE (blue) corresponds to the right hand ordinate.
Generally, the position of the most prominent features of the spectra are described by four characteristic frequencies: Characteristic frequency of the RESR-C unit:
Characteristic frequency of the L-C unit:
Characteristic frequency of the RLeak-C unit:
Characteristic frequency of the RESR-L unit:
Two main situations can be distinguished in Figure 3 and Figure 4:
- Lorentz-Oscillation: fRC> fLC as in the case for CS = 4.7 µF (blue graph) and
- Debye-Relaxation: fRC < fLC as in the case for CS = 50 F (orange graph). 
Horizontal dash-dotted lines in both graphs signify the pure capacitive and inductive part. For the imaginary part of the capacitance, please refer to Figure 17 in the appendix.
fRC, the characteristic frequency of the R-C unit, is the frequency at which the capacitor can be charged and discharged. The inverse of the frequency is basically the charging time under ideal constant voltage charging. For the capacitor with CS = 50 F, the ideal charging time is about 2⋅π ⋅15 mΩ⋅50 F≈5 sec Below the frequency of (5 sec)^(-1) the capacitor can utilize its nearly full capacitance (>99.9 %). Above this frequency, the capacitor is not fully charged anymore (in reference to the maximum voltage of the AC signal).
At fRC the capacitance spectrum (Figure 4) of the Supercapacitor shows a shoulder. Below this frequency, the capacitance value can be inferred from the graph. Above fRC the impedance spectrum, given in Figure 3 (Bottom), shows a plateau at RESR.
fLC, the characteristic frequency of the L-C unit, is the frequency at which the coupling of parasitic inductance and capacitance leads to a resonant behavior (if fRC> fLC). Below this frequency the capacitor acts as capacitor, i.e. can be charged. Above this frequency the capacitor acts as inductor. The self-resonance results in a sharp minimum in the impedance spectrum (WCAP-FTBE), as given in Figure 3 (Top).
At the minimum of the impedance spectrum the RESR value can be read off.
The capacitance spectrum in Figure 4 shows a pole, which is a special type of singularity. It is actually a real physical behavior and not only a measurement artefact.
The measurement system, which consists of the capacitor and the parasitic inductance, behaves like a oscillator, i.e. resonator. 
At the increasing positive branch, the probing signal constructively contributes to the oscillations of the resonator. That is to say, the charge increase dQ at the interface is disproportionately high, although the magnitude of applied voltage signal dV remains the same. Since CS = dQ/dV, a strong increase of the capacitance is measured. At the maximum, the system is in phase (resonance) with the probing frequency.
A further increase of the probing frequency leads to an abrupt change in sign of the capacitance (singularity). at fLC.
At the negative branch, the probing signal destructively overlaps with the oscillations of the resonator. The current is actually flowing “in opposite direction” to the probing voltage.
Thus, the applied voltage signal dV leads effectively to a relative decrease of charge at the capacitor interface –dQ, which results in a negative capacitance. fLeak is the characteristic frequency of the RLeak-C unit. Below this frequency the capacitor acts like a resistor with resistance RLeak. That is to say, at very low frequencies, the leakage discharge is larger than the AC charging current. Usually this effect is rarely visible in the spectra. It requires either measurements to frequencies below 1 Hz or a rather low RLeak. fRL, the characteristic frequency of the RESR-L unit, is the frequency above which the capacitor acts like an inductor with inductance LESL. In cases where fRC < fLC, it signifies the onset of the increase in impedance at high frequencies. Above this frequency, it is exceedingly difficult to extract RESRvalues from measured impedance spectra.
02. MEASURED IMPEDANCE SPECTRA AND CAPACITANCE SPECTRA
The following sections will discuss spectra on different capacitor types, exemplarily chosen from the Würth Elektronik eiSos portfolio. The standard model, depicted in Figure 1, uses an frequency independent ohmic resistance RESR. However, physical processes as well as measurement artifacts may lead to deviations from the idealized ohmic behavior, as will be seen in the below section 
2.1 Supercapacitors WCAP-STSC
The below presented spectra were measured with the impedance analyzer Alpha-AK, POT/GAL from Novocontrol in a four-terminal Kelvin configuration. The four-terminal Kelvin configuration has the advantage of enabling a high phase angle resolution of about 0.001°, since the voltage measurement probes are independent from the current supply leads. The measured impedance spectrum of a SC with 50 F in Figure 5 shows the same features as the corresponding theoretical spectrum in Figure 3. In this case (fRC < fLC), fRC is below 1 Hz and thus several magnitudes smaller than fLC. As a result, the spectrum shows a flat bottom region from which RESR can be inferred. The increase of RESR towards lower frequencies is more clearly visible in the spectrum of Re(Z ̂ )=RESR in Figure 6. This frequency dependence is not a measurement artefact, however, attributed to real physical phenomena:
- distributed network of porous electrodes and
- the ionic charge transport in the electrolyte of the EDLC. 
A physical interpretation of the spectra is: The slower the SC is charged, the more pores can be infiltrated by charges, which leads to the increase of capacitance. Due to the viscous solvent, the ions require more time to insert the smaller pores, which in turn leads to an increase of RESR toward low frequencies.
Figure 5: Measured impedance spectrum || d of 50 F Supercapacitor, WCAP-STSC
The values of Re() = RESR below fRC and above fLC become unreliable, since the spectra become dominated by the capacitive and parasitic inductive behavior. The shoulder at fLeak is not visible, since it is practically difficult to measure toward such low frequencies. (Also for other capacitor technologies, it is not common practice to measure at such low frequencies. The shoulder at fLeak. is usually not depicted in impedance spectra.)
Figure 6: Measured spectrum of the real part of the impedance (Re()=RESR ) of 50 F Supercapacitor, WCAP-STSC.
The capacitance spectrum in Figure 7 shows the typical shoulder situated at the characteristic frequency of the RESR-C unit fRC. The height of the shoulder is at about 51 F (0.01 Hz) with a slight increase toward lower frequencies. This increase is especially pronounced in electrolytic capacitors such as Supercaps. It is, as already mentioned above, mainly caused by the charge storing at porous interfaces as well as pseudo capacitive processes. Mathematically, it can be well described by a set of distributed R-C networks.
The characteristic frequency is, as shown above, governed by the term RESR CS. Due to the relatively high capacities, the characteristic frequencies has to shift to frequencies around or even below 1 Hz. The indicated frequency fRC Figure 7 is about 0.21 Hz. The inverse of fRC can be interpreted as lower limit for the constant voltage charging time, which in this case is:
Figure 7: Measured capacitance spectrum of 50 F Supercapacitor, WCAP-STSC
The following characteristic values can be extracted from the measured spectra above:
- CS (0.01 Hz)=51 F
- RESR (fRC=0.2 Hz) = 0.012 Ω
- RESR (fLC=160 Hz) = 0.007 Ω
2.2 Aluminum Electrolytic Capacitors, WCAP-AIG8
The following spectra were measured with the impedance analyzer Alpha-AK, POT/GAL from Novocontrol in a fourterminal Kelvin configuration.
In principle, the charge storing mechanism of the aluminum electrolytic capacitor (E-Cap) is comparable to the one of the SC. The energy is stored by electrolytic charges at a porous interface. However, the E-Cap utilizes a thin porous aluminum oxide layer as a dielectric. The porosity of this layer again leads to large effective surfaces and relatively large capacitance. It is due to the large capacitance that for this type of capacitor fRC is often (not always!) still lower than fLC, as indicated in the measured impedance spectrum of a 270 µF capacitor in Figure 8. Compared to the SC, fLC has shifted towards higher frequencies. The capacitive contribution is more pronounced and thus appears also further shifted to higher frequencies. The contribution of the parasitic inductance is about the same. Hence, the flat bottom region is much smaller than for the SC.
Figure 8: Measured impedance spectrum || of 270 µF aluminum electrolytic capacitor.
In general, all features of the spectra in Figure 9 and Figure 10 appear shifted to higher frequencies. The interpretation of the capacitance spectrum in Figure 9, with its characteristic shoulder, is similar to the one for the SC.
Figure 9: Measured capacitance spectrum of 270 µF aluminum electrolytic capacitor.
The equivalent series resistance Re()=RESR, given in Figure 10, also shows the increase toward low frequencies. However, below 1 kHz or so the spectrum shows a strong increase toward lower frequencies. This strong increase is however, probably not due to any real physical effect, it is a measurement artefact. It is a common effect that always happens, if the loss angle is very close to zero. In this region, the equipment is not able to clearly separate the real and imaginary part from each other, which results in an overestimated output of the real part.
Hence, the correct interpretation of the spectra at low loss angles, which corresponds to low or high frequencies with respect to fLC, is usually difficult, since the loss angle resolution of the most LCR meters is not smaller than 0.1 degree or so. However, that also means that generally the values around or at fLC are most trustworthy.
At the end, a correct interpretation of any impedance spectra is only possible with the experimental details of the measurement.
Figure 10: Measured spectrum of the real part of the impedance (Re() = RESR ) of 270 µF aluminum electrolytic capacitor.
The following characteristic values can be extracted from the measured spectra above. The dissipation factor DF = RESR/Xc = 2 π f CS RESR is stated to improve comparability with data sheets and other documentations. The results for frequencies ≪fLC are given for the sake of completeness. They may contain a large error as is explained in the Appendix and elsewhere. 
- CS (120 Hz) = 265 µF
- RESR (120 Hz) = 0.14 Ω Ω
- RESR (fLC= 68.5 kHz) = 0.04 Ω
- DF (120 Hz) = RESR/Xc =2 π f CS RESR = 2.8 %
- DF (fLC = 68.5 kHz) = 2 π f CS RESR = 0.8 %. (Capacitive Reactance: Xc = 1/(2 π f Cs )
2.3 Film Capacitors, WCAP-FTBE
The below spectra were measured with Agilent E5061B Network Analyzer in a shunt through configuration.  The measured impedance spectrum of a 470 nF Film capacitor, in Figure 11, shows in principle the same features as the calculated spectra, given in Figure 3 (Top). Due to the low capacitance fRC > fLC, as indicated by the dashed lines, which results in a graph with a sharp minimum at fLC= 1.94 MHz. The resistance value at the minimum is roughly the RESR at fLC, which in this case is about 0.04 Ω.
Figure 11: Measured impedance spectrum of 470 nF film capacitor.
The measured spectrum of the equivalent series resistance Re()=RESR, given in Figure 12, shows a bathtub-like shape with a minimum around fLC. The increase at high and low frequencies is very likely not the actual physically correct ESR behavior of the capacitor, but a measurement artifact from the low capacitance (high impedance) and the parasitic inductance.
The separation of a small real part from a large imaginary part is technically difficult. At the end the accuracy and the resolution of measurable loss angle, i.e phase angle, determines the accuracy of the measured impedance. Accuracy plots for the used equipment (Figure 22, Figure 24) and an error calculation for this measurement is given in the Appendix. At these demanding conditions, analyzers often overestimate Re(). 
The network analyzer has no separate current and voltage connections, which usually leads to a lower phase accuracy than for four-terminal Kelvin configurations. Hence, it is a common phenomenon that ESR values, if measured in such a configuration, are overestimated. For the sake of prudency, it is best to consider those values as conservative estimates.
Hence, it is practically difficult to say to what extend the measured Re() in Figure 12 is correct. Certainly around fLC the results of approximately 0.04 Ω are most trustworthy and can be considered as an conservative estimate.
Figure 12: Measured spectrum of the real part of the impedance (Re()=RESR) of a 470 nF film capacitor
The measured capacitance spectrum in Figure 13 shows the overall features of the calculated spectrum, given Figure 4. It has a plateau region at low frequencies and a singularity, positioned at fLC. The measured capacitance, as read of from the plateau region, is about 495 nF. The measured capacitance with its hyperbolic behavior is physically correct and not some sort of measurement artefact (see section 1.2).
Figure 13: Measured capacitance spectrum of a 470 nF film capacitor.
The parasitic inductance may change as function of the length of circuit path or temperature. As a consequence, also the position of fLC may change accordingly. It is therefore in practice important to operate the application not in the vicinity of fLC. The following characteristic values can be extracted from the measured spectra above. The dissipation factor DF = RESR/Xc = 2 π f CS RESR is calculated to improve comparability with data sheets and other documentations. The results for frequencies ≪fLC are given for the sake of completeness. They may contain a large error as explained in the Appendix and elsewhere. 
- CS (1 kHz) = 495 nF
- RESR(1 kHz) = 2.2 Ω
- RESR(fLC=1.94 MHz) = 0.04 Ω
- DF(1 kHz) = 2 π f CS RESR = 0.68 %
- DF(fLC=1.94 MHz) = 2 π f CS RESR= 24 %.
2.4 Multilayer Ceramic Chip Capacitor, WCAP-CSGP
The following spectra were measured with Agilent E5061B Network Analyzer in a shunt through configuration. The impedance spectrum, given in Figure 14, is qualitatively the same as for the film capacitor in section 2.3. Due to the lower rated capacitance of 22 nF the impedance spectrum is shifted to higher frequencies with fLC= 45.8 MHz. As in the last chapter fRC>fLC, which leads to a sharp minimum at fLC
Figure 14: Measured impedance spectrum of 22 nF MLCC.
The ESR, as inferred from the minimum at fLC, is about 0.06 Ω, which for the same reasons as discussed in section 2.3, could be considered as conservative estimate. The actual ESR might be even lower.
The capacitance spectrum in Figure 15 shows, as discussed above, the typical resonant behavior. From its constant plateau region we may read a capacitance of about 23 nF.
Figure 15: Measured capacitance spectrum of 22 nF MLCC.
Figure 16 shows the ESR spectrum with a shallow bathtub like shape. As discussed in the last chapter, the increase toward low and high frequencies is probably the result of incorrect separation of real and imaginary part of the impedance by the measurement equipment, i.e. small loss angle δ
Figure 16: Measured spectrum of the real part of the impedance (Re()=RESR) of 22 nF MLCC
The increase toward lower frequencies could be partially explained with the increase in dielectric loss due to ionic polarization of larger domain structures or charge hopping induced conductivity. However, the actual value of about Re() = 3000 Ω at 1 kHz clearly suggests, that the increase is largely attributed to an insufficient resolution of the loss angle. As mentioned before, it is technically difficult to separate the real part from a relatively large imaginary part, i.e. small loss angle (see Appendix, Figure 23).
The increase toward higher frequencies could be due to the skin effect. The geometrical factor of the electrodes of an MLCC makes a scientifically sound analysis of this effect difficult. Usually studies of skin effect are conducted on cylindrical geometries. Those results cannot be quantitatively used for measurements of MLCCs. 
Due to the resolution limits of the loss angle, the Re() = RESR spectra have often a minimum around fLC and strongly increasing slopes toward low and high frequencies. This measurement effect will superimpose any potential skineffect (see Figure 25)  To cut a long story short, ESR spectra of LCR resonators are most trustworthy in the region around fLC. Beyond that, they require, not always but often, further technical knowledge for interpretation. The following characteristic values can be extracted from the measured spectra above. The dissipation factor DF = RESR/Xc = 2 π f CS RESR is given to improve comparability with data sheets and other documentations. The results for frequencies ≪fLC are given for the sake of completeness. They may contain a large error as is explained in the Appendix and elsewhere. 
- CS (1 kHz) = 23 nF
- RESR(fLC= 46 MHz) = 0.06 Ω
- DF(fLC= 46 MHz) = 2 π f CS RESR= 38 %
The standard model, given in Figure 1, is a suitable means to interpret the technically important features of all commercially relevant capacitors. In most cases, it is even sufficient to use an even simpler model, which neglects RLeak. It was furthermore demonstrated how the model parameters, such as CS and RESR , can be extracted from measured spectra. The deviations from that model as well as aspects of the measurement accuracy have been exemplarily discussed on measured spectra. Especially the separation of real and imaginary part at small loss angles is defective. In the selfresonant case, the RESR spectra can only be correctly measured around fLC. Due to this reason the physical phenomena, such as skin effect, are very likely superimposed by a large error and cannot be studied in the measured RESR spectra.
A.1 Graphs and Figures
Figure 17: Capacitance spectrum Im() as calculated from the standard model. Corresponds to spectrum Re(̂) in Figure 4. Gray lines indicate Re().
Interpretation of Im() (Figure 17) for R-C unit: Im() describes the dissipation of energy in a capacitive system, associated to the movement of charges. This dielectric relaxation, i.e. absorption band, is caused by the reorientation of permanent molecular dipoles, induced by the applied alternating electric field. Im() is mathematically related to the Re() by the Kramers Kronig relation. In case of a Debye relaxation, Im() can be conveniently used to read off fRC. The height of peak of Im() at fRC is for the Debye relaxation CS/2. An example of calculation for a Lorentz oscillation (fRC> fLC) is given in Figure 18. The parameters for the plot are such as to show all details of the curve progression. The graph does not necessarily correspond to a specific capacitor product.
Figure 18: Example of calculation for a Lorentz relaxation (fRC > fLC), Parameters are such as to see all details.
Figure 19: Phase shift angle ϕ(f) for WCAP-FTBE as calculated from the standard model. Corresponds to impedance spectra in Figure 3. Polarization contributions at higher frequencies, such as electronic polarizations, are neglected, i.e. C(f→∞)=0
Figure 20: Loss angle δ(f) for WCAP-FTBE as calculated from the standard model. Corresponds to impedance spectra in Figure 3. Polarization contributions at higher frequencies, such as electronic polarizations, are neglected, i.e. C(f→∞)=0
Figure 21: P Phase shift angle ϕ(f) for WCAP-STSC as calculated from the standard model. Corresponds to impedance spectra in Figure 3. Polarization contributions at higher frequencies, such as electronic polarizations, are neglected, i.e. C(f→∞)=0.
Figure 22: Loss angle δ(f) for WCAP-STSC as calculated from the standard model. Corresponds to impedance spectra in Figure 3 Polarization contributions at higher frequencies, such as electronic polarizations, are neglected, i.e. C(f→∞)=0
Figure 23: Accuracy plot for E5061B ENA Vector Network Analyzer from Keysight. Conditions for 10 % measurement accuracy range (Source: externally  )
Figure 24: Phase accuracy for E5061B ENA Vector Network Analyzer from Keysight. (Source: externally  )
Figure 25: Surface resistivity and depth of penetration for different materials at room temperature. (Source: externally  )
A.2 Formulas and Notes
Mathematical description of the equivalent circuit in section 1.1 (temporal frequency: f, angular frequency: ω=2⋅π⋅f)
Table 2: Values for exemplary error calculation
 Editors: E. Barsoukov, J. R. Macdonald, Impedance Spectroscopy: Theory, Experiment, and Applications, Third Edition, John Wiley & Sons (2018)
 Luis Moura and Izzat Darwazeh, Introduction to Linear Circuit Analysis and Modelling: From DC to RF, Newnes an imprint of Elsevier (2005)
 B. E. Conway, Electrochemical Supercapacitors - Scientific Fundamentals and Technological Applications, Kluwer Academic Publishers, Springer, Boston (1999)
 A. F. J. Levi, Essential Classical Mechanics for Device Physics, Morgan & Claypool Publishers (2016)
 R. M. Hill and L. A. Dissado, Debye and non-Debye relaxation, Journal of Physics C: Solid State Physics, 18(19):3829-3836 (1985)
 D Edwards, J.-H Hwang, S.J Ford, T.O Mason, Experimental limitations in impedance spectroscopy: Part V. Apparatus contributions and corrections, Solid State Ionics, 99(1,2):85-93 (1997)
 Jong-Sook Lee et al., Impedance Spectroscopy Models for X5R Multilayer Ceramic Capacitors, Journal of the Korean Ceramic Society, 49(5):475-483 (2012)
 R. Kötz et al., Temperature behavior and impedance fundamentals of supercapacitors, Journal of Power Sources, 154:550–555 (2006)
 M. Schönleber, E. Ivers-Tiffée, Approximability of impedance spectra by RC elements and implications for impedance analysis, Electrochemistry Communications, 58:15-19 (2015)
 Bing-Ang Mei et al., Physical Interpretations of Nyquist Plots for EDLC Electrodes and Devices, The Journal of Physical Chemistry C, 122(1): 194-206 (2018)
 M.R. Hasyim, D. Ma and R. Rajagopalan and C. Randall, Prediction of Charge-Discharge and Impedance Characteristics of Electric Double-Layer Capacitors Using Porous Electrode Theory, The Electrochemical Society, 164(13): A2899-A2913 (2017)
 V. Srinivasan and J. W. Weidner, Mathematical Modeling of Electrochemical Capacitors, Journal of The Electrochemical Society, 146: 1650-1658 (1999)
 M. Itagaki, Electrochemical Impedance and Complex Capacitance to Interpret Electrochemical Capacitor, Electrochemistry, 75(8):649-655 (2007)
 T. Stolzke et al., Comprehensive accuracy examination of electrical power loss measurements of inductive components for frequencies up to 1 MHz, Journal of Magnetism and Magnetic Materials, 497: 166022 (2020)
 Online Manual, E5061B Network Analyzer, Keysight, URL: http://ena.support.keysight.com/e5061b/manuals/webhelp/eng/?nid=- 32496.1150148.00&cc=CA&lc=eng&id=1790874
 Data Sheet, E5061B Network Analyzer, Keysight URL: http://ena.support.keysight.com/e5061b/manuals/webhelp/eng/product_information/specifications.htm
 H.A. Wheeler, Formulas for the Skin Effect, Proceedings of the IRE, 30(9):412-424 (1942)
 A. E. Kennelly, F. A. Laws and P. H. Pierce, Experimental Researches on Skin Effect in Conductors, in Transactions of the American Institute of Electrical Engineers, 34(2):1953-2021 (1915)
 G. Schröder J. Kaumanns, R. Plath, Advanced measurement of AC resistance on skin-effect reduced large conductor power cables ,8th International Conference on Insulated Power Cables, Versailles – France (2011)
 E. Ivers-Tiffée and Aweber, Evaluation of electrochemical impedance spectra by the distribution of relaxation times, Journal of the Ceramic Society of Japan, 125(4):193-201 (2017)
The Application Note is based on our knowledge and experience of typical requirements concerning these areas. It serves as general guidance and should not be construed as a commitment for the suitability for customer applications by Würth Elektronik eiSos GmbH & Co. KG. The information in the Application Note is subject to change without notice. This document and parts thereof must not be reproduced or copied without written permission, and contents thereof must not be imparted to a third party nor be used for any unauthorized purpose. Würth Elektronik eiSos GmbH & Co. KG and its subsidiaries and affiliates (WE) are not liable for application assistance of any kind. Customers may use WE’s assistance and product recommendations for their applications and design. The responsibility for the applicability and use of WE Products in a particular customer design is always solely within the authority of the customer. Due to this fact it is up to the customer to evaluate and investigate, where appropriate, and decide whether the device with the specific product characteristics described in the product specification is valid and suitable for the respective customer application or not. The technical specifications are stated in the current data sheet of the products. Therefore the customers shall use the data sheets and are cautioned to verify that data sheets are current. The current data sheets can be downloaded at www.we-online.com. Customers shall strictly observe any product-specific notes, cautions and warnings. WE reserves the right to make corrections, modifications, enhancements, improvements, and other changes to its products and services. WE DOES NOT WARRANT OR REPRESENT THAT ANY LICENSE, EITHER EXPRESS OR IMPLIED, IS GRANTED UNDER ANY PATENT RIGHT, COPYRIGHT, MASK WORK RIGHT, OR OTHER INTELLECTUAL PROPERTY RIGHT RELATING TO ANY COMBINATION, MACHINE, OR PROCESS IN WHICH WE PRODUCTS OR SERVICES ARE USED. INFORMATION PUBLISHED BY WE REGARDING THIRD-PARTY PRODUCTS OR SERVICES DOES NOT CONSTITUTE A LICENSE FROM WE TO USE SUCH PRODUCTS OR SERVICES OR A WARRANTY OR ENDORSEMENT THEREOF. WE products are not authorized for use in safety-critical applications, or where a failure of the product is reasonably expected to cause severe personal injury or death. Moreover, WE products are neither designed nor intended for use in areas such as military, aerospace, aviation, nuclear control, submarine, transportation (automotive control, train control, ship control), transportation signal, disaster prevention, medical, public information network etc. Customers shall inform WE about the intent of such usage before design-in stage. In certain customer applications requiring a very high level of safety and in which the malfunction or failure of an electronic component could endanger human life or health, customers must ensure that they have all necessary expertise in the safety and regulatory ramifications of their applications. Customers acknowledge and agree that they are solely responsible for all legal, regulatory and safety-related requirements concerning their products and any use of WE products in such safety-critical applications, notwithstanding any applications-related information or support that may be provided by WE. CUSTOMERS SHALL INDEMNIFY WE AGAINST ANY DAMAGES ARISING OUT OF THE USE OF WE PRODUCTS IN SUCH SAFETY CRITICAL APPLICATIONS
Application Notes : https://we-online.com/en/support/knowledge/application-notes
Contact : https://we-online.com/en/support/contact
Würth Elektronik eiSos GmbH & Co. KG
Max-Eyth-Str. 1, 74638 Waldenburg, Germany
Tel.: +49 (0) 7942 / 945 – 0