Such confusions are indeed present in quite a few high profile review articles and it is very challenging to clear the muddy field. There are two key points worth mentioning, whilst I have given more critical analyses in my recent article which is published as an Open Access paper.

Chen GZ, Linear and non-linear pseudocapacitance with or without diffusion control, Prog. Nat. Sci.-Mater. Int., 31 (6) (2021) 792-800. https://doi.org/10.1016/j.pnsc.2021.10.011

The first point is that there are two types of Faradic charge storage mechanisms on electrode, namely the Nernstian mechanism (battery like behaviour) and capacitive mechanism (pseudocapacitance).

Secondly, both Nernstian and capacitive charge storage processes can be controlled by diffusion. It has been a misleading claim that the battery-like Faradaic storage process should be diffusion controlled (b=0.5). It is also wrong to claim the capacitive storage process is always surface controlled (b=1).

These confusions have started in the literature since Conway's first use of pseudocapacitance for what should have been more accurately termed as Faradaic capacitance.

It is now too late to stop the use of pseudocapacitance, but I believe we can at least differentiate between

(1) linear pseudocapacitance (truly capacitive with rectangular CVs whose charge storage is a linear function of the electrode potential or cell voltage) and

(2) non-linear pseudocapacitance (peak shaped CVs whose charge storage is a non-linear function of the electrode potential or cell voltage).

17 Feb 2022

Hi, Ayman,

Thank you for your further question.

If both Nernstian and capacitive charging processes co-exist in the electrode reaction, it is possible to judge their presences by electrochemical means, but I am NOT aware of any electrochemical method, such as cyclic voltammetry and chronopotentiometry (GCD) that is capable of differentiating them, particularly if the capacitive charging process is faradaic in nature.

If the capacitive charging is associated with the EDL charging which does not involve valence change of the active element, differentiation is possible between EDL and Faradaic (either Nernstian or pseudocapacitive) charging processes by simultaneous spectroscopic and electrochemical analyses, e.g. ESR (EPR) and cyclic voltammetry (see my paper mentioned above for more info).

About the phase/structural change caused by charging, there is no evidence that capacitive charging can cause these which are therefore more likely Nernstian in nature. However, it does not mean that the absence of phase change must be due to capacitive charging. The simple case is when charging occurs on dissolved redox active species in liquid or solid solutions, particularly when the solution is an insulator or poor conductor of electrons. (Here, solution means a homogeneous condensed phase.) In such cases, Nernstian charging is very common.

Nevertheless, in semiconducting materials (solid solutions), capacitive charging can proceed. A simple rule of thumb is if the redox species or charging site in the material can or cannot interacting with each other. If they cannot interact, it is surely Nernstian. If they can interact, the charging is more likely capacitive.

06 March 2022

Hi, Ayman Elkholy, Your understanding on interaction between redox sites is partially correct.

In semiconductors in which interactions occur between atoms, multivalent atoms share their valence electrons in an energy band. This band makes all the interacting atoms into a large cluster or "molecule" of atoms. This "molecule" can undergo many electron transfer, whilst each electron transfer to or from this "molecule" is shared by all atoms in the "molecule". In other words, the change of charge on each atom is fractional. As a result, we are now unable to use the concept of "valence change" which can only be an integer. This explanation follows the semiconductor band theory.

There is a long-time misunderstanding that battery-like behaviour corresponds to high energy and low power, and capacitive behaviour means high power and low energy. Charge capacity is an intrinsic property of materials, but the power capability is an issue of device design. In other words, it is possible to make a battery having much higher power capability (specific power) than the current commercial battery. For example you can achieve very high specific power in thin layer batteries in which the electrode is very thin. Similarly, if the electrode is sufficiently thick, a supercapacitor can also be of low specific power.

I do not recommend using the concept "intercalation pseudocapacitance". This concept creates at least two problems. Firstly, strictly speaking, the term intercalation refers to specifically to lithium ions intercalating into regular layer-structured or channel-structured materials, but many such materials are clearly Nernstian, such as LixCoO2 and LixMn2O4. Secondly, during the charging and discharging of most, if not all, pseudocapacitive materials, ingression and egression of charge balancing ions occur to maintain neutrality. Such processes are also broadly termed as intercalation and de-intercalation. However, these materials do not have regular layer- or channel-structures, but they offer a good level of porosity of irregular meso- or micro-pores. In both these cases, phase change does not happen or is not expected, but neither of these two processes agrees with the concept of intercalation pseudocapacitance, which is consequently very confusing or even misleading.

I hope the above explanation has addressed your question "...if the Faradic reaction (associated with valence change) is not accompanied by a phase change, can we adopt the "intercalation pseudocapacitance" concept to describe it?"

Thanks, Prof. George Zheng Chen. I will read your review article which looks interesting. I have a question: as long as the capacitive charge storage process can be controlled by diffusion as the Nernstian (battery-type) process, how to distinguish both of them electrochemically? Is it through the shape of CV and GCD curves or this should be supported by physical characterization techniques?

For example, if there are significant structural phase transformations during charging/discharging, can we say it is a battery-type behavior? And if there is slight or no phase change, e.g. shift in XRD peaks, can we say it is a capacitive behavior?

Hi, Ayman,

If both Nernstian and capacitive charging processes co-exist in the electrode reaction, it is possible to judge their presences by electrochemical means, but I am NOT aware of any electrochemical method, such as cyclic voltammetry and chronopotentiometry (GCD) that is capable of differentiating them, particularly if the capacitive charging process is faradaic in nature.

If the capacitive charging is associated with the EDL charging which does not involve valence change of the active element, differentiation is possible between EDL and Faradaic (either Nernstian or pseudocapacitive) charging processes by simultaneous spectroscopic and electrochemical analyses, e.g. ESR (EPR) and cyclic voltammetry (see my paper mentioned above for more info).

About the phase/structural change caused by charging, there is no evidence that capacitive charging can cause these which are therefore more likely Nernstian in nature. However, it does not mean that the absence of phase change must be due to capacitive charging. The simple case is when charging occurs on dissolved redox active species in liquid or solid solutions, particularly when the solution is an insulator or poor conductor of electrons. (Here, solution means a homogeneous condensed phase.) In such cases, Nernstian charging is very common. Nevertheless, in semiconducting materials (solid solutions), capacitive charging can proceed. A simple rule of thumb is if the redox species or charging sites in the material can or cannot interact with each other. If they cannot interact, it is surely Nernstian. If they can interact, the charging is more likely capacitive.

Ayman Elkholy Please find the attached file for this review, which explains Intercalation pseudocapacitance.

Intercalation pseudocapacitance in electrochemical energy storage:

recent advances in fundamental understanding and materials

developmen.

Thanks again, Prof. George Zheng Chen. I have a question regarding the last rule you mentioned in the latest reply. I think you mean by the redox species the charge carriers that undergo redox reactions with the material's active site. Is this right? If so, the interaction between both of them may be associated with valence change and/or phase change. As far as I understand, Faradaic reaction should be accompanied by a valence change. Then, if the phase change happens during a Faradaic process, the process will be slow and I think the charge storage mechanism here can be called a battery-like mechanism. In other words, the process output will be high energy and low power. On the other hand, if the Faradic reaction (associated with valence change) is not accompanied by a phase change, can we adopt the "intercalation pseudocapacitance" concept to describe it? This is the core of my original question. It is understandable that surface reactions always occur regardless of the type of the active material. What I mean in this discussion is how to describe the mechanism with the major contribution to the charge storage process.

Thanks, Dalia M Elgendy . I already referred in the original post to the review article that you shared

Ayman Elkholy Your understanding on interaction between redox sites is partially correct.

In semiconductors in which interactions occur between atoms, multivalent atoms share their valence electrons in an energy band. This band makes all the interacting atoms into a large cluster or "molecule" of atoms. This "molecule" can undergo many electron transfer, whilst each electron transfer to or from this "molecule" is shared by all atoms in the "molecule". In other words, the change of charge on each atom is fractional. As a result, we are now unable to use the concept of "valence change" which can only be an integer. This explanation follows the semiconductor band theory.

There is a long-time misunderstanding that battery-like behaviour corresponds to high energy and low power, and capacitive behaviour means high power and low energy. Charge capacity is an intrinsic property of materials, but the power capability is an issue of device design. In other words, it is possible to make a battery having much higher power capability (specific power) than the current commercial battery. For example you can achieve very high specific power in thin layer batteries in which the electrode is very thin. Similarly, if the electrode is sufficiently thick, a supercapacitor can also be of low specific power.

I do not recommend using the concept "intercalation pseudocapacitance". This concept creates at least two problems. Firstly, strictly speaking, the term intercalation refers to specifically to lithium ions intercalating into regular layer-structured or channel-structured materials, but many such materials are clearly Nernstian, such as LixCoO2 and LixMn2O4. Secondly, during the charging and discharging of most, if not all, pseudocapacitive materials, ingression and egression of charge balancing ions occur to maintain neutrality. Such processes are also broadly termed as intercalation and de-intercalation. However, these materials do not have regular layer- or channel-structures, but they offer a good level of porosity of irregular meso- or micro-pores. In both these cases, phase change does not happen or is not expected, but neither of these two processes agrees with the concept of intercalation pseudocapacitance, which is consequently very confusing or even misleading.

I hope the above explanation has addressed your question "...if the Faradic reaction (associated with valence change) is not accompanied by a phase change, can we adopt the "intercalation pseudocapacitance" concept to describe it?"

Thanks, Prof. George Zheng Chen.