An Overview on the Development of Electrochemical Capacitors and Batteries – Part I

The Nobel Prize in Chemistry 2019 recognized the importance of Li-ion batteries and the revolution they allowed to happen during the past three decades. They are part of a broader class of electrochemical energy storage devices, which are employed where electrical energy is needed on demand and so, the electrochemical energy is converted into electrical energy as required by the application. This opens a variety of possibilities on the utilization of energy storage devices, beyond the well-known mobile applications, assisting on the decarbonization of energy production and distribution. In this series of reviews in two parts, two main types of energy storage devices will be explored: electrochemical capacitors (part I) and rechargeable batteries (part II). More specifi cally, we will discuss about the materials used in each type of device, their main role in the energy storage process, their advantages and drawbacks and, especially, strategies to improve their performance. In the present part, electrochemical capacitors will be addressed. Their fundamental difference to batteries is explained considering the process at the electrode/electrolyte surface and the impact in performance. Materials used in electrochemical capacitors, including double layer capacitors and pseudocapacitive materials will be reviewed, highlighting the importance of electrolytes. As an important part of these strategies, synthetic routes for the production of nanoparticles will also be approached (part I).


INTRODUCTION
In 2019, the Nobel Prize in Chemistry was awarded to Goodenough, Whittingham and Yoshino for the development of Li-ion batteries. Their fi ndings paved the way for the electric revolution in which we live today. From devices that enabled information access nearly everywhere to electric vehicles, they are all powered by a Li-ion battery. Moreover, the use of electrochemical energy storage devices in the power supply grid is on the rise, enabling power production to be more concentrated in intermittent renewable sources. Their works were developed during the 1980s and early 1990s, but obviously the technology kept improving over the years, and materials that can store more energy density have been described over the last decades. This is because there is a need for devices that can deliver energy for either a longer time (high energy density) or faster (high power density).
For instance, electrochemical capacitors are another class of electrochemical energy storage device, but, unlike batteries, they normally show low energy density yet high power. These differences in performance arise from the different electrochemical process that takes place during cell charge and discharge. In a typical double-layer capacitor, the only process occurring is the charge and discharge of the electric double-layer (interface between electrode and electrolyte), which is a fast process. The charge of the double-layer involves the charge compensation generated at the electrode surface by the ions in solution, this generated charge at the electrode surface comes from an external source, ie. when the device is plugged to the socket and it is being charged. The double-layer charging is a physical process (movements of ions and electrons) and it results in a high-power event (the process is fast, and it can deliver -or storethe energy quickly), but, since it is limited to the material surface area only, also results in low energy density (it cannot operate for much long). On the other hand, batteries show faradaic reactions, therefore, there is change on oxidation state of active materials during charge and discharge, which can happen in the bulk material as well as the surface, resulting in high energy. However, it depends on the slow kinetics of these faradaic reactions, resulting in low power. In between the materials for batteries and double-layer capacitors, there are the pseudocapacitive materials, which rely on fast faradaic reactions, mimicking a capacitor. These fast faradaic reactions occurs at the electrode surface. Figure 1 shows the relationship between energy and power density, the so-called Ragone plot, for various energy storage systems from electrostatic capacitors (physical) to combustion engines (chemical). The electrochemical systems need to be improved to gain more energy and power so they can compete with combustion engines, especially for transport applications. This review will show different strategies that have been used to build devices that can show such requirements.

ELECTROCHEMICAL CAPACITORS
Electrochemical capacitors show higher power and lower energy density when compared to batteries because the electrochemical processes involved in charge storage are different. These processes can vary even among electrochemical capacitors. Figure 2 compares an electrostatic capacitor (or physical capacitor), a doublelayer capacitor, and a pseudocapacitive material-based capacitor (polyaniline (PANI), as example). In this section, the former two kinds of capacitors will be reviewed, focusing on the electrochemical processes and strategies to improve these devices.

Electrochemical double-layer capacitor
As mentioned above, electrochemical doublelayer capacitors (EDLCs) rely on the physical interaction between the electrode surface and ions present in the electrolyte. This interaction is known as the electric double-layer. As one electrode is polarized positively, anions will migrate within the solution to locally compensate the positive charge, whereas the other electrode is polarized negatively, and therefore, cations will migrate to this negative surface to compensate the charge. Each electrode has a double-layer, one being defi cient and another having excess of electrons, which in turn will result in a doublelayer containing mostly anions and cations, respectively. The formation of both doublelayers is concomitant and requires a current fl ow between both electrodes, which is the capacitor charge process. This process occurs until the cell voltage reaches the maximum determined cell voltage. Once the EDLC is charged, the doublelayer is maintained and the stored energy can be used afterwards. While the EDLC is used, ie. the capacitor discharge process, a current will fl ow on the opposite direction, and both doublelayers will start to be undone. Once the device is completely discharged, cell voltage will be zero.

Electrodes
The capacitance (C) of a capacitor can be calculated from equation (1): where ε 0 and ε r are the electric constant and the dielectric constant between charge separation, respectively; A is the electrode surface area; and d is the separation of two plates in a physical capacitor, or the thickness of the double-layer in the EDLC. There are two main reasons for EDLCs to present much higher capacitance than a physical capacitor, that can be observed from equation (1). Firstly, the thickness of doublelayer is several orders of magnitudes smaller than the separation between two plates in a physical capacitor, and secondly, the surface area achieved using carbonaceous materials is in order of a couple of thousands m 2 per gram. Therefore, activated carbon (AC) and other carbonaceous materials with high surface area are the most employed in EDLC development. AC is produced from many precursors, but most of commercially available ACs are prepared from coconut shell (Hall et al. 2010). AC preparation involves basically two steps. The fi rst is the carbonization of a precursor at high temperatures, which can be made under an oxygen-free atmosphere or by hydrothermal treatment . Then, the charcoallike product is activated, also at high temperature, either through a mixture of inert and oxidizing atmospheres or through chemical oxidation. Different processes lead to different textural properties, varying pore size distribution, surface area, and possibly chemical functionalization. These different properties will have an impact on capacitance. Not only changing the surface area will change the capacitance, but pore size can facilitate or hinder mass transport (Chmiola et al. 2008, Jäckel et al. 2016, Sillars et al. 2011, while chemical functionalization can increase interactions between ions and electrode surface, increasing the specific capacitance (Abbas et al. 2019). For instance, it is believed that nitrogen functionalization increases the AC specific capacitance through faradaic reactions, behaving as a pseudocapacitive material (Candelaria et al. 2012). However, even with the increment in capacitance, it has been difficult to determine the exactly role of functionalization and different pore size distribution (Simon & Gogotsi 2008).
As mentioned above, many precursors have been used for AC preparation. It is worth mentioning that carbonization of biomass has attracted great attention of many research groups (Titirici et al. 2015, Wu et al. 2013, 2016, due to the interest in adding value to biomass that would be treated as waste instead.
An obvious alternative to AC is the employment of graphene or graphene-like materials. Graphene is known to have a large surface area, comparable to AC, and, as a 2D material, this area could be doubled if accessible by the electrolyte in both sides. However, due to layer stacking during processability and electrode manufacturing, the achieved capacitance is below expectations and it is outperformed by AC (Béguin et al. 2014). A few strategies have overcome this problem, for instance, vertically aligned graphene can be growth directly onto Ni current collector, making use of both sides of the layer, resulting in a greater capacitance (Cai et al. 2014, Miller & Outlaw 2015. Changing the graphene textural properties is also an alternative to increase surface area (Kholmanov et al. 2013, Lopes et al. 2018, Salvatierra et al. 2013, and therefore the specific capacitance. This change can be done through activation of graphene with similar procedures described above .

Electrolytes
An EDLC electrolyte holds the ions that interact with the electrodes. Positive and negative electrodes usually interact with anions and cations, respectively, to form two electrical double-layers. Therefore, the ions need to have good mobility for a fast charge-discharge mechanism, which means that the electrolyte must have a high ionic conductivity.
Electrolytes can be classified by the solvent used in their formulation, and each of them has advantages and disadvantages when compared. The most common types are water, organic solvents, and ionic liquids (ILs). There are three main properties that have the most impact: viscosity, dielectric constant, and electrochemical stability. Each of these factors have a practical impact on the EDLC performance and metrics. Viscosity has great impact on ionic conductivity and capacitance itself. Firstly, as more viscous is the liquid, less mobility the ions possess, decreasing ionic conductivity, which in turn penalize the device power density. Secondly, the activated carbon micropores can be difficult to access when too viscous electrolytes are used, due to the low wettability of electrode/electrolyte. Moreover, the dielectric constant of the electrolyte solvent is directly proportional to the electrode capacitance, as shown in Equation 1, and normally the dielectric constant is higher for polar than for nonpolar solvents. However, polar solvents usually presents higher viscosities and possibly lower electrochemical stability, especially for polar protic solvents (Conway 1999). In addition, the electrochemical stability can depend on other aspects of the solvent, for instance, molecules with conjugated double bonds normally show a diminished electrochemical stability. The added salt can also play an important role in the electrochemical stability, since cations and anions can be reduced or oxidized at too high polarization. From equation (2) it is possible to see how important it is to have a high operating voltage, ie. high electrochemical stability of the whole system electrodes/electrolyte, to obtain a high energy density EDLC: where E, C and V are energy density (W h g -1 ), capacitance (F g -1 ), and operating voltage (V). For instance, increasing the operating voltage from 1 to 3 V results in an increment of 9 times in energy density.
Moreover, the ion size impacts on the electric double-layer and then on the obtained capacitance, since the relationship between ion size and pore size distribution plays a major role in the final capacitance, due to pore accessibility. Therefore, choosing the appropriate salt to the electrolyte requires one to consider the pore distribution of the activated carbon used in the electrode formulation.
Aqueous electrolyte can be acidic, alkaline or neutral water solutions. The issue with acidic and alkaline solutions is that the former will have a lower negative voltage limit due to the presence of H + and the H 2 evolution, and the latter will show lower positive voltage limit due to the presence of OHand the O 2 evolution. Overall, both of them will have a limited electrochemical stability which will decrease the EDLC energy density, as shown in Equation 2. On the other hand, the ionic conductivity of these electrolytes is high, guarantying a superior power density due to elevated ion mobility. Neutral aqueous solutions can present higher electrochemical stability but lower ionic conductivity than the others aqueous electrolyte counterpart. For instance, Fic et al. (2019) reported that an EDLC with aqueous solution of 1 mol L -1 Li 2 SO 4 can show an operating voltage up to 1.6 V, but they also showed that the cell internal pressure starts to increase when cell voltage exceeds 1.2 V, clearly indicating that cell optimization can be carried out to enhance operating voltage and mitigate electrolyte decomposition. For comparison, aqueous electrolyte containing KOH, H 2 SO 4 or Li 2 SO 4 show ionic conductivities of 540, 750 and 70 mS cm -1 , respectively, while the first two show operating voltages of ca. 0.8 V, while the last one could go up to 1.6, as mentioned above (Fic et al. 2019, Ramachandran & Wang 2018. To enhance the operating voltage, aprotic organic solvents can be used in the electrolyte formulation. The state-of-the-art regarding organic solvent electrolyte contains propylene carbonate (ε r = 66, at 20 °C) or acetonitrile (ε r = 36, at 20 °C), and considering that the latter has a flash point of 6 °C (T boiling = 82 °C) against 132 °C (T boiling = 242 °C) for the latter, the most common used between both is propylene carbonate (Gagliardi et al. 2007, Simeral & Amey 1970. The salt added in both is tetraethylammonium tetrafluoroborate (TEABF 4 ), normally at 1 mol L -1 (Béguin et al. 2014). Other salts are also investigated, as LiBF 4 and tetramethylammonium (TMABF 4 ), which present smaller cations, so the electric double-layer could host more cations and then increase capacitance. However, the solubility of these salts is lower than TEABF 4 (Nambu et al. 2013). Another strategy is changing the anion BF 4 for bis(oxalato)borate or difluoro(oxalato)borate, increasing the solubility of the TMA + salt in propylene carbonate (PC) (Nanbu et al. 2006a(Nanbu et al. , b, 2007. Schütter et al. (2015Schütter et al. ( , 2016 have proposed computational screening of solvents for electrolyte production. They scrutinized organic compounds considering several properties, as flash point and melting temperatures, viscosity, solubility of salts and electrochemical stability, and, on top of that, they considered only commercially available compounds. First, they correlated their method by using nitrile compounds as glutaronitrile, 2-methylglutaronitrile and adiponitrile. They concluded that the computational screening was successful on selecting appropriate compounds to be used as solvent for EDLC electrolytes (Schütter et al. 2015). Later, they widened the compounds class and selected the 3-cyanopropionic acid methyl ester. The electrolyte containing the selected solvent and TEABF 4 showed an operating voltage of 3.5 V and longterm stability (Schütter et al. 2016).
Another class of electrolyte that has received great attention is IL. ILs are salts that are liquid at room temperature. As they are composed only by ions, they show intrinsic ionic conductivity; in addition, they also show high thermal stability and wide electrochemical window (Eftekhari 2017, MacFarlane et al. 2016, Salanne 2017. There is a myriad of possible ILs since they can be produced by the appropriate matching of many cations and anions.  et al. 2008, 2010). The electrochemical stability of ILs is normally evaluated using inert electrodes like glassy carbon, however, when utilized as EDLC electrolyte, the stability of the current collector and the electrode (AC, binder and carbon black) must be taken into consideration, and an appropriate technique for maximum operation voltage determination must be done (Martins et al. 2018b, Weingarth et al. 2013. Two ILs have been mostly used and can be considered the benchmark when comparing new liquids: 1-ethyl-3-methyl imidazolium BF 4 ([Im 1,2 ][BF 4 ]) and N-butyl-N-methyl pyrrolidinium Tf 2 N ([Pyr 1,4 ][Tf 2 N]), achieving 3.8 and 3.7 V, respectively, depending on binder and AC (Martins et al. 2018a, b, Weingarth et al. 2013. A polymeric binder has shown to have a great impact on the operating voltage, increasing from 2.7 to 3.8 V using the same IL [Im 1,2 ][BF 4 ] (Martins et al. 2018b). Moreover, the use of a polymeric IL as binder in electrode preparation can increase the operating voltage and also the capacitance, being a promising alternative to inert binders (Martins et al. 2017b).
Different cations combined with Tf 2 Nwere also reported. The addition of an ether chain in a phosphonium cation remarkably increased the capacitance when compared with an analog cation with an alkyl chain, due to increased mobility of the ether chain (Rennie et al. 2013 All the anions cited above have fluorine in their composition, and one of the focus has been to find fluorine-free anions that still have competitive electrochemical properties. Dicyanamide (N(CN) 2 -) is one of the most promising fluorine-free anions, as ILs containing this anion show ionic conductivity similar to the state-ofthe-art electrolyte. For instance, [Pyr 1,4 ][N(CN) 2 ] shows 12.0 mS cm -1 against the 12.2 mS cm -1 of 1 mol L -1 TEABF 4 in PC (Hall et al. 2010, MacFarlane et al. 2002. However, these ILs suffer from low electrochemical stability, but even operating at 2.6 V, their EDLC shows similar energy density than an EDLC containing the Tf 2 Ncounterpart at 3.5 V (Wolff et al. 2015 ) also shows relatively high conductivity while keeping a wide electrochemical stability. EDLC containing the IL [Pyr 1,4 ][B(CN) 4 ] can operate up to 3.7 V and outperform the energy density of the EDLC using a Tf 2 Nanalogue at all tested power densities (Martins et al. 2018a). It is worth mentioning that strategies such as mixing different ILs to tailor the electric double-layer has also been considered and more symmetrical charge distribution between both electrodes has been observed (Van Aken et al. 2015).
Solid electrolytes, or gel polymer electrolytes, have also been considered for EDLCs. They have the advantage of being safer, since there is no possibility of leakage, and more importantly they are normally nonflammable. Mixture of poly(ethylene oxide) with different salts is the most common solid electrolyte, but other mixtures have been proposed, including the use of poly IL (Ayalneh Tiruye et al. 2015, Kim et al. 2013, Pont et al. 2009). The bottleneck of solid electrolytes is the low ion mobility, which results in low ionic conductivity compared to liquid electrolytes, hampering the device power density (Xue et al. 2015).

Pseudocapacitive materials
There is an effort to enable electrochemical capacitors with higher energy and power densities to go beyond the employment of electrodes based on carbon-derived materials (Lin et al. 2018b). Such goal can be achieved by the use of the so-called pseudocapacitive materials. These materials, differently from the carbonaceous ones, do not retain charge exclusively from ion adsorption at the electrode/ electrolyte interface, but from reversible and fast redox reactions near or at the electrode material surface (Augustyn et al. 2014, Conway 1999. These sorts of electrochemical processes, despite their faradaic nature, exhibit capacitive features, such as a rectangular-shaped potentiodynamic profile and almost linear galvanostatic charge-discharge curves. For that reason, those phenomena are called pseudocapacitive (from greek: pseudos, pretended and not real) (Conway 1999, Brousse et al. 2017. Traditionally, according to Conway (Conway 1999, Conway et al. 1997, several faradaic mechanisms can result in capacitive electrochemical features: (1) intercalation pseudocapacitance, (2) underpotential deposition (3) redox pseudocapacitance or specific adsorption with charge transfer, and (4) charge compensation in conducting polymers.
These four processes have different physical natures and can occur in plenty of materials, revealing, however, the same electrochemical behavior. Intercalation pseudocapacitance, for instance, takes place when ions intercalate (due to charge transfer ) into the active sites of redox materials through tunnel or channels, without crystallographic phase change (Augustyn et al. 2014, Jiang & Liu 2019. Underpotential deposition, on the other hand, occurs when depositing a metal ion on a different metal surface, above its redox potential, forming a monolayer or submonolayer of deposited metal. One straightforward example is the deposition of Pb 2+ onto Au surface (Chen et al. 1993). Besides, redox pseudocapacitance arises during the adsorption of ions at the electrode surface concomitantly with partial charge transfer, due to specific interactions between the ion's valence electrons and the electrode's surface orbitals, like in a Lewis acid-base donoracceptor interaction (Conway 1991). It is worth to mention, as an example, the adsorption of H + into Ru 2 O.nH 2 O surface (Ardizzone et al. 1990). The latter pseudocapacitive process listed here appears during the electrochemical oxidation/ reduction of conducting polymers. As long as polymeric chains are being oxidized/reduced, ions from the electrolyte are incorporated in the polymer backbone, in order to compensate the generated charges. All those mechanisms are schematized in Figure 3.
As mentioned before, those processes arise from different physical natures, but the materials display the same electrochemical features. Such singularity is due to the relationship between the electrical potential and the extent of charge developed at the electrode surface as a result of interfacial phenomena or along the so-called inner surface of materials (Ardizzone et al. 1990). The following expression was proposed by Conway (1999) and Conway et al. (1997): wherein E is the electrode potential, R is the real gas constant, n is the number of electrons, F is the Faraday constant, T is the temperature and X is the fractional coverage, which varies for each type of pseudocapacitive process. Besides such a thermodynamic approach, pseudocapacitive materials hold the promise for better electrochemical capacitors materials, due to their kinetic behavior. The energy storage in such materials is limited to reactions that occur at electrode surface or are limited to surface availability, differently from solid-state diffusion (as battery materials), displaying high rate capability (Augustyn et al. 2014). There is a significant difference between transition metal oxides that display a pseudocapacitive behavior and those that do not. The latter are employed as battery materials. However, some of such battery-type transition metal oxides, when nanostructured from bulk to a few nanoscale dimensions, display a capacitive electrochemical signature, as shown in Figure 4 ( Okubo et al. 2007). This is because, as pointed out, pseudocapacitive processes are surface dependent. The increase in surface area via nanostructuration increases the availability of redox sites at the surface, promoting the so-called "surface redox processes" (in contrast with "bulk redox processes"). Such nanostructuration, also suppresses phase transformation of some battery-type materials to a large extent and stimulate high-rate capability due to the shortening of diffusion pathway length. All those factors give raise to capacitive electrochemical signatures, despite their redox nature. RuO 2 was one of the earliest pseudocapacitve materials reported. It presents high specific capacitance (over 700 F g -1 ) (Augustyn et al.

2014)
, large potential window, high proton conductivity, high rate capability, and metallic type electronic conductivity. Also, the hydrated state of the oxide and its crystallography have influence over RuO 2 pseudocapacitance behavior and specific capacitance values. Moreover, for electrogenerated material, factors like surface treatment of obtained fi lms, such as annealing and anodization, and deposition potential range seem to play a considerable role at RuO 2 pseudocapacitive behavior (Ahn et al. 2006, Patake et al. 2009). Nevertheless, the high cost of such material, its toxicity and scarcity still limit its practical application as electrode material. However, MnO 2 arises as a cheaper and more environmentally friendly electrode material. Its pseudocapacitive behavior was Capacitance values near 200 F g -1 were recorded, despite its theoretical specifi c capacitance near 1233 F g -1 (assuming a potential window ~0,9 V, for one-electron transfer). Such difference stems from MnO 2 poor electronic conductivity (10 -5 to 10 -6 S cm -1 (Zhao et al. 2017)), limiting the charge storage to a thick layer of surface. However, in the form of thin or ultrathin fi lms, specific capacitances over 1000 F g -1 were achieved (Toupin et al. 2004). Several attempts were made in order to enlarge such conductivity and specific capacitance values, such as changing the crystallographic structure, using different synthetic pathways, doping with metal ions ( Differently from others spinel materials, Co 3 O 4 and NiCo 2 O 4 can act as battery-type materials, due to the formation of oxyhydroxides along the charge storage process in alkaline media, which gives higher capacitance values in comparison with others spinel-type metal oxides. However, during galvanostatic chargedischarge processes, in some cases, small plateaus corresponding to purely faradaic reactions are observed, instead of capacitor-like triangular profi le . Aside, some cobalt oxides obtained by successive ionic layer adsorption and reaction revealed reasonable pseudocapacitive properties (Kandalkar et al. 2008).
The majority of electrodes based on such materials are studied in aqueous electrolytes (Brousse et al. 2017). However, the efforts to push towards pseudocapacitive-materialbased devices lead to the use of non-aqueous electrolyte. Most of them are based on nonaqueous lithium-ion or sodium-ion electrolytes. In such media some transition metal oxides exhibit intercalation pseudocapacitance due to the intercalation of Na + or Li + ions. Also, the electrolyte enables capacitance over a wider potential range in comparison with aqueous media, and therefore higher energy densities

MXenes
MXenes are 2D transition metal carbides, nitrides and carbonitrides obtained by the selective chemical etching of A element layers in an MAX structure, where "M" is an early transition metal oxide (such as Sc, Ti, Zr,Hf, V,Nb,Ta,Cr,Mo, etc.), "A" is an A group element, mainly group A-13 and A-14 (such as Al, Si, S, Ga, Ge, As, Cd, In, Sn etc.) and "X" is carbon/nitrogen (Naguib et al. 2012). In general, the MAX phase is stable, however, the relative weakness of the M-A bound in comparison with the M-X bond allows etching by treatment with HF (Naguib et al. 2011). Also, the method enables functionalization of MXene with -F or -OH, influencing the physicochemical properties of obtained materials (Alhabeb et al. 2017a). In terms of pseudocapacitive materials, Ti 3 C 2 T x (T x = OH / F) is undoubtedly the most studied MXene, being obtained by selective etching of an Al layer in Ti 3 AlC 2 phase.
Most MXenes are studied in aqueous electrolytes. In the case of Ti 3 C 2 T x (Ghidiu et al. 2014, Lukatskaya et al. 2013, Naguib et al. 2012, it is known that its pseudocapacitance is related to Ti redox changes accompanied by cation intercalation (Lin et al. 2018a). Interestingly, some cations readily intercalate into Ti 3 C 2 T x by only immersing it in a suitable cation solution (Lukatskaya et al. 2013). Also, some promising results are observed for free standing clay-type Ti 3 C 2 T x paper electrode e.g a volumetric capacitance exceeding 900 F cm -3 , in H 2 SO 4 electrolyte (Ghidiu et al. 2014). Also, still employing paper electrode, a range of 300 -400 F cm -3 is observed in neutral media (Lukatskaya et al. 2013). However, despite those values, the narrow potential window allowed by aqueous media hinders the widespread use of MXene based electrochemical capacitors.

Polymers
Pseudocapacitance in conducting polymers arises from their redox reactions, which are followed by entrapment of electrolyte ions at the polymer's backbone in order to compensate the generated charges, which is often called doping process. Conducting polymers usually display higher specific capacitance values in comparison with some transition metal oxides (Muzaffar et al. 2019), relative low cost and can be readily synthetized chemically or electrochemically (González et al. 2016, Huguenin & Torresi 2008. However, conducting polymer-based electrodes present lower cyclability (Beguin & Frackowiak 2013, González et al. 2016) and conductivity loss, due to polymer swelling and shrinking along innumerous cycles (Beguin & Frackowiak 2013). The most common employed conducting polymers are PANI (Banerjee et al. 2019, Karami et al. 2003, Neves & Polo Fonseca, 2002, Ryu et al. 2002, polypyrrole (PPy) (Fan & Maier 2006, Snook et al. 2004, Wang et al. 2007, polythiofene (PTh) (and its derivates) (Snook et al. 2011, Villers et al. 2003, and poly(3,4-ethylenedioxythiophene) (PEDOT) (Her et al. 2006, Snook et al. 2011, Villers et al. 2003). Each of them shows electroactivity in various potential ranges, depending on the electrolyte (organic and aqueous) (Maia et al. 1996, Varela et al. 2001. It is important to point out that the electrolyte displays a great importance over charge compensation in conducting polymers. Such influence was deeply explored for polymers like PPy (Maia et al. 1996, Peres et al. 1992, PANI (De Albuquerque Maranhão & Torresi 1999a, b, Torresi et al. 1993 and PEDOT (Lé et al. 2019) by means of in situ electrochemical techniques such as electrochemical crystal quartz microbalance (De Albuquerque Maranhão & Torresi 1999a, 1999b, Lé et al. 2019, Maia et al. 1996, Peres et al. 1992, Torresi et al. 1993 and optical beam deflection . In general, the specific capacitance of chemically synthetized conducting polymer-based electrode ranges from 100 to 600 F g -1 , depending on the media. For organic electrolytes , values below 300 F g -1 are usually observed (Laforgue et al. 2002, Soudan et al. 2001, Villers et al. 2003. In order to circumvent problems regarding volumetric changes, composite materials such as polymerceramic, polymer-metal, polymer-carbon (Lin et al. 2018a) are being employed. One strategy to improve the conducting polymer electrochemical performance is using 3D nanostructured morphologies, since those materials display small charge transfer resistance and favorable ion transport within the 3D continuous framework (PANI e.g) (Shi et al. 2015). Expressive enhancements were observed for devices based on 3D nanostructured polypyrrole (Antonio et al. 2016(Antonio et al. , 2019.

Electrolytes
The electrolyte will play an important role in supercapacitors when pseudocapacitive materials are used, since the redox reactions occurring during charge/discharge require certain specificity from the electrolyte. Ions present in the electrolyte and the media pH are important for redox reactions, as already mentioned above for some examples. It is worth mentioning that most of the research on pseudocapacitive materials is performed using aqueous electrolyte, but the narrow electrochemical voltage window results in low energy density. In order to overcome this problem, pseudocapacitive materials can also be used with organic solvent-based electrolytes, mainly employing electrolytes also used in metal-ion batteries, since many of the pseudocapacitive material used in such configuration shows metal ion intercalation (Augustyn et al. 2014, Choi et al. 2020. Conducting polymers have also been tested with ILs, showing different potentiodynamic profile than in aqueous electrolytes, due to the different activity in presence of such IL structure ). An asymmetrical capacitor using two different conducting polymer and aqueous gel electrolyte operating up to 1.6 V was also proposed, showing expanded operating voltage and capacitive behavior, in addition to the improved safety due to solid electrolyte (Kurra et al. 2015). Besides the enlarged operating voltage that non-aqueous electrolytes show, problems with lower capacitance must be overcome, especially when MXenes are employed. Such type of material shows outstanding capacitance in aqueous electrolyte but still fails when ILs are used. For instance, Ghidiu et al. (2014) reported a MXene of Ti 3 C 2 T x (T = OH, O or F) showing a gravimetric capacitance of 222 F g -1 in aqueous H 2 SO 4 electrolyte (at 20 mV s -1 ), or 821 F cm -3 . On the other hand, Lin et al. (2016) using a similar electrode preparation achivied only 70 F g -1 at the same 20 mV s -1 but using an ILs as electrolyte. That is more than three times higher capacitance when aqueous electrolyte is used.. In a different strategy, IL was dissolved in acetonitrile and used as electrolyte. As prepared MXene showed around 30 F g -1 at 20 mV s -1 , and the addition of carbon nanotubes to the electrode improved the gravimetric capacitance to around 75 F g -1 (Dall et al. 2016), which is still far lower than the aqueous counterpart.

MATERIALS DESIGN FOR ELECTROCHEMICAL ENERGY STORAGE
Electrode materials are the fundamental components in electrochemical energy storage devices, in both batteries and supercapacitors. In many cases, material selection plays an essential role in limiting the level of electric energy storage due to their theoretical specific capacity and their voltage. For either devices, the overall system performance relies on electroactivity and stability of electrode materials (Chen & Xue 2016). The main issues to address in the rational design of electrode materials are related to (i) energy density; (ii) kinetics of charge and mass transport; and (iii) structure stability in operando conditions.
For an electrode material with a given chemical composition, potential, structure stability, electronic conductivity, and charge transport are dependent on its crystal structure and the material's dimensionality. Different crystal structures present different atomic bonding and internal energy that can release different amounts of chemical energy. Moreover, when in nanometer scale, materials present a higher surface to volume ratio, which decreases the ion transport pathway and contributes to enhancing the capacity of electrodes. Nanostructured materials, such as hollow core/ shell nanoparticles and nanowires, can also reduce electrode fatigue due to expansion/ contraction during cyclability (Neto et al. 2020, Zhou et al. 2019a). In the following section, we will briefly present the most common procedures to synthesize electrode materials, and discuss the advantages and limitations related to each method. Table I summarizes some materials and the strategies used produce each of them.

Hydrothermal and solvothermal syntheses
Hydrothermal (in water) and solvothermal (in other solvents such as ethylene glycol or acetone) syntheses can be described as methods for formation and growth of crystals through a chemical reaction in a reactor sealed with a solution above ambient temperature and pressure (Feng & Xu 2001, Kharissova et al. 2019). Among all methods, hydrothermal and solvothermal syntheses are ideal for controlling the morphology, particle size distribution, shape, uniformity, area, crystallinity, nucleation, and a technique to facilitate an excellent reproducibility of nanomaterials especially one to three-dimensional nanocrystals (Kharissova et al. 2019, Shi et al. 2013). The precise control over hydrothermal and solvothermal conditions is indispensable for the preparation of structures. These parameters can be controlled and involve changing pressure, pH, temperature, solvent, organic additives, reaction time, and the precursor source (Byrappa & Adschiri 2007, Kharissova et al. 2019, Shi et al. 2013. Under this scenario, the hydrothermal and solvothermal syntheses would greatly influence the performance of electrochemical energy storage/ conversion applications.

Co-precipitation method
Together with hydrothermal, the co-precipitation method is one of the most traditional processes to prepare transition metal oxides, and it involves the formation of insoluble products of inorganic oxides. Two strategies are mainly used, where one is the proper metal oxide directly precipitated and the other occurs with an intermediated by-product precipitation followed by oxide crystallization upon temperature treatment. This method is versatile to produce 1,2,3-d transition metal oxide structures and allows the combination with other methodologies to reach desired crystalline phases, sizes, and morphologies. Koenig et al. (2011) have tailored lithiated manganese and nickel oxides in a reactor system with NH 4 OH/Na 2 CO 3 for LIB cathode. Their report on the internal gradient of composition presented excellent retention of discharge capacity over 100 cycles. He et al. (2018) reported octahedral Li 1.2 Mn 0.54 Ni 0.13 Co 0.13 O 2 core-shell Lirich with a spinel-layer via co-precipitation/ gel method. According to the authors, the LIB cathodic material contained enough oxygen vacancies in the lattice allowing the reinsertion of Li + (He et al. 2018). Compared to the traditional co-precipitation material, the proposed coprecipitation/gel approach yielded higher discharge capacity and coulombic efficiency.

Sol-gel method
The sol-gel method is an alternative to hydrothermal and solvothermal synthesis, it offers some particular advantages to produce complex inorganic materials at lower temperature and pressure, and shorter synthesis times (Danks et al. 2016). Moreover, sol-gel chemistry is capable of controlling the morphology and the size of particles (Haetge et al. 2012, Trewyn et al. 2007). The sol-gel method implies hydrolysis and condensation of a metal alkoxide precursor to obtain a sol composed of colloidal particles and, finally, an interconnect rigid network structure called a gel (Danks et al. 2016, Hench & West 1990, Lakshmi et al. 1997). In the sol-gel chemistry, there are many different ways to control the structures with small changes in conditions relatively easy to achieve as pH, the metal alkoxide/H 2 O ratio, and temperature (Hench & West 1990).

Pechini method
The Pechini method (Pechini 1967) (and its modified approaches) is versatile, capable of fine engineering the material surface, and it overcomes most of the difficulties and disadvantages of traditional alkoxide based sol-gel process (Dimesso 2016). The synthesis follows two major steps to achieve oxide structures: i) A desired cation polyesterification resin is generated by the combination of the chosen metal precursor and hydroxycarboxylic acid to polyhydroxy alcohol. The cationicspecies homogeneity is guaranteed by choosing a soluble/time-stable precursors and heating the solution at mild temperatures. ii) Calcination of the resin. The evaporation of the organics is conducted upon a thermal treating step, resulting in the crystallization of the desired oxide. Different Pechini-modified approaches are found to achieve desirable structural characteristics in the nanomaterials (Dimesso et al. 2011).

Chemical vapor deposition method
The chemical vapor deposition method can be considered as a gas-phase synthetic process. In vapor phase, the mechanism of formation can be divided into three main events: precursor vaporization, nucleation, and growth of crystals. This approach is very simple, versatile, and also operates continuously with high product yield for desired nanoparticle preparation (Maduraiveeran et al. 2019). However, to produce materials with controlled morphology, it is required to use a template or to modify the substrate to allow nanostructures to grow (Zhou et al. 2019a). The multistep deposition of metals or metal oxides offers an alternative to produce multilayered electrochemical devices.

Microwave irradiation method
Microwave irradiation is another interesting tool to produce metal oxides for not depending on high pressure or temperature. Particularly, this method stands out by its simplicity and rapidity with which reactions occur, besides being a methodology that allows scale-up without heat diffusion problems (Herring et al. 2013). The microwave heating depends on the reactants ability to adsorb such irradiation. Generally, irradiated metal precursors with a large microwave irradiation absorption feature tend to reach the effective temperature instantaneously, allowing them to achieve small nanoparticles in a fast way.

Mechanochemical synthesis
Mechanochemistry has emerged as one of the alternatives to conventional solvent and thermal-based routes for chemistry and materials preparation ( . Ball milling devices are the most common apparatuses used for such a purpose, enabling to generate materials for several applications with eco-friendly based protocols, as most of the reactions are carried out directly in the solid state. In this procedure, the raw reactant materials, mostly powders, are introduced in the milling jars along with balls as milling bodies. As the milling starts, mixing and comminution occurs, i.e. the breakdown of the solids into smaller objects, increasing the surface area of the overall system. Concomitantly, the accumulation of stress in the solid particles in the form of plastic deformation and defects and the continuous generation of fresh reactive surfaces lead to the chemical transformations in reactive mixtures. The ongoing milling enhances these described processes and may push, many times, the reactions to completion in shorter times when compared to classical thermal routes. The ball to powder mass ratio and the type of mill (planetary ball mill, vibratory ball mill, attritor) (Baláž et al. 2013) influence the effectiveness of the milling in terms towards the desired product. The type of mechanical energy input (shock, impact, shear stress) can also drive to different products (Michalchuk et al. 2013).
I n t h e e l e c t r o c h e m i s t r y f i e l d , mechanochemistry has been used for the elaboration of electrocatalysts such as nanoparticles, and materials for batteries and supercapacitors (Kosova & Devyatkina 2004, Kosova 2010, Muñoz-Batista et al. 2018, Ning et al. 2004, Tarascon et al. 2005. Fine particulate compounds used as electrodes with spinel, layered, framework structure and other amorphous materials were synthesized and tested as negative and positive electrodes and solid electrolytes (Kosova, 2010, Lenain et al. 1998, Soiron et al. 2001, Tarascon et al. 2005, Yang et al. 2010.

MXenes
The production of 2D materials can be classified into two different approaches, the bottomup, which relies on the reaction of molecular building blocks to form covalently linked sheets, and the top-down, that consists of the chemical or mechanical exfoliation of bulk crystal to obtain individual sheets . The latter approach is the most used for the production of graphene and graphene-analog materials, such as molybdenum disulfide (Mo 2 S) and hexagonal boron nitride, because the interlayer interactions are much weaker than the bonds within the layers (Nicolosi et al. 2013).
In the case of MXenes, etching of MAX phases followed by exfoliation is the most common strategy of synthesis, and several etching methods have been reported in the literature (Alhabeb et al. 2017a, Naguib et al. 2012, Verger et al. 2019a. The first reported MXene was Ti 3 C 2 T x and was produced by etching Al atoms in Ti 3 AlC 2 with concentrated hydrofluoric acid (HF) for few hours at room temperature (Naguib et al. 2011 Etching conditions depend on HF concentration, frequently from 5 to 30 wt%, or on the formation of HF in situ by using LiF in acidic media (Ghidiu et al. 2014, Peng et al. 2016 or using NH 4 HF 2 (Halim et al. 2014, Lipatov et al. 2016. To obtain individual sheets of Ti 3 C 2 T x , depending on the etching protocol, it is required to use intercalating compounds to expand interlayer space in MXene flakes. When lithium fluoride is employed as an etchant, solvated Li + ion intercalation can be adjusted, and delamination of individual sheets can be achieved by sonication in 5 mol L -1 LiF/ 6 mol L -1 HCl, or by manual shaking in 12 mol L -1 LiF/ 9 mol L -1 HCl (Sang et al. 2016), called minimally intensive layer delamination (MILD). The MILD method is more suitable for applications where larger and less defective MXene sheets are required (Sang et al. 2016). When harsh etching conditions are employed, such as 30 wt % of HF, MAX phases expand during the process and accordion-like structures were formed ( Figure  5c), while 5 wt % HF, NH 4 F, and MILD-etched delamination produced negligible opening in MXenes lamellas (Alhabeb et al. 2017b).
Despite the extraordinary properties of MXene materials, there are some concerns regarding the potential hazards during their synthesis and processing (Lakhe et al. 2019) ( Figure 5), which remains as one of the main drawbacks that hinder scaling up their production. Hazards related to MXene synthesis includes handling toxic and corrosive reactants, and also the production of potentially explosive by products, such as H 2 (Lakhe et al. 2019). Properly wash MXene is another important problem to address, because fluoride is not completely removed after the synthesis, the exposure to nanostructured materials with high surface area with hazardous functional groups is of great concern to human health (Fadeel et al. 2018). The large-scale production of MXenes and its application in functional devices will only be achieve when safety considerations were better understood, from raw materials until its final disposal.

CONCLUSIONS
In the present part of this review series, it was intended to show the importance that electrochemical energy storage technologies have gained after decades of development that culminated on the Nobel Prize 2019 in Chemistry. Particularly, electrochemical capacitors were explored, highlighting their fundamental difference when compared to batteries. Different types of electrochemical capacitors were explored: double-layer capacitor and pseudocapacitive capacitors, including metal oxides, MXenes and conducting polymers. In addition, this part also explored the most common strategies and synthetic routes to obtain nanostructured materials to be applied in both electrochemical capacitors and batteries. The next part of the review will address the advances in rechargeable batteries.  TMO: transition metal oxide. MXenes: 2D transition metal carbides, nitrides or carbonitrides. MAX: M is for a transition metal, A for an element from group A and X is carbon or nitrogen. PPy: polypyrrolye. PTh: polythiofene. PEDOT: poly(3,4-ethylenedioxythiophene). MILD: minimally intensive layer delamination.