Nanoporous Carbon Synthesis: An Old Story with Exciting New Chapters Nanoporous Carbon Synthesis: An Old Story with Exciting New Chapters

Activated carbons are key materials in technological applications of multidisciplinary fields (e.g. adsorption, separation, and catalytic processes). The extensive use of these materials results from the combination of a well-developed pore network (micropores or micro + mesopores) along with the presence of heteroatoms (e.g. oxygen, nitrogen, and sulfur). The large scale production of nanoporous carbons is a well-established process since the first patents date from the beginning of the twentieth century. Conventional activation methodologies are divided between physical, using steam or CO 2 , and chemi - cal, being KOH, ZnCl 2 , and H 3 PO 4 the most commonly reported oxidizing agents. Due to the panoply of operational parameters that can be changed or added in the production of activated carbons, there is still room for R&D. In this chapter, both conventional and innovative synthetic processes are reviewed to offer an up-to-date picture regarding raw materials, carbonization step, activation process, and other approaches. Conventional activation of gels and chars obtained by novel approaches (i.e. sol-gel method, hydro thermal carbonization (HTC), and acid-mediated carbonization) and more innovative strategies (i.e. variations of HTC process, carbonization of organic salts and ionothermal approaches) are addressed. Textural, surface chemistry and morphological properties of the derived porous carbons were reviewed and critically rationalized.


Historic perspective
Carbon is the main element of the entire world's flora, and its atomic structure allows unique bonding possibilities leading to various structures with distinct properties. As consequence, carbon materials are used for so long that it is hard to determine the very first moment.
It was only during the eighteenth century that the mechanisms underlying the properties of charcoals started to be recognized. In 1773, Scheele measured the adsorptive properties of charcoals from distinct sources using various gases; a decade later, Lowitz studied the abilities of charcoals to adsorb odors from wounds [7,9]. This author published results regarding the adsorption of a range of organic vapors and also studied systematically the adsorption in various aqueous solutions, namely in the decolorization of contaminated tartaric acid solutions. The discussion of the contribution of charcoals to control odors from gangrenous ulcers was made by Kehl, in 1793, who also discovered that charcoal from animal origin (animal tissues) could be used to remove color from solution, including from sugar solutions [7,9]. The introduction of carbon materials in industrial processes took place in England, in 1794, when wood charcoals were applied as bleaching agents in the processing of sugar. By that time, the method for preparing the wood charcoals was kept a secret [9]. In 1805, the first large-scale sugar refining facility was introduced in France by Gruillon, working with ground and washed wood char [9]. Between 1805 and 1808, Delessert demonstrated the effectiveness of charcoal for decolorizing sugar-beet liquor contributing for the growth of this industry in France [9]. In 1808, all European sugar refineries were using charcoal as a decolorizer [9]. The discovery of bone-derived char as a more efficient raw sugar syrups decolorizing agent was made by Figuier in 1811, and quickly the sugar industry replaced wood charcoal by this improved material [5,9]. The regeneration of bone-derived char by heating was discovered in 1811, and sometime later granular bone-derived char started to be produced, thus enabling an easier regeneration [9]. In 1815, the majority of sugar refining facilities were using granular bone-derived char [9]. The dependence of the decolorizing properties on the char's source, the thermal processing and particle size was demonstrated by Bussy in 1822 [9]. He also produced a carbon material with higher decolorizing ability than bone-derived char by heating blood with potash, being this the first recorded example of an "activated carbon" material obtained by thermal and chemical processes [5,9].
The first large-scale application in gas phase took place in 1854 [8], with the installation of carbon filters in London sewage ventilation systems. In 1872, carbon filters were also used in the masks of the chemical industry workers for preventing the inhalation of mercury vapors.
Besides thousands of years of history, and the already large range of applications for charcoals, it was only in the beginning of the twentieth century that one of the most outstanding abilities of carbon materials was explored: the possibility of enclosing a huge porosity into the carbon material structure. This "revolution" was due to the discoveries of Raphael von Ostrejko that, between 1900 and 1903, patented two different methods for the industrial activation of charcoal and the production of activated carbon materials [10]. This set of patents mentions the bases of the chemical activation process (carbonization of lignocellulosic materials with metallic chlorides) and of the thermal or physical activation (slight gasification of chars with water vapor or carbon dioxide at elevated temperatures) and also the specific equipment for thereof. The factory for full-scale activated carbon production was built by that time in Ratibor, now named Raciborz (Poland), and still operates, being so the world's oldest industrial activated carbon manufacturing plant [10,11].
Although sugar industry was the first to apply activated carbons, in this case for whitening purposes, the starting point for the great development in the production and application of activated carbons was undoubtedly World War I, when these materials were used in gas masks [7,8]. The production and search for new activated carbons has been boosted decade after decade due to their fundamental role in various technological applications which are related to, namely, restricted environmental regulations, recovery of valuable chemical compounds, and catalysts support [8,[12][13][14].
Nowadays, the driving forces for the research in nanoporous carbons are related with the properties of the most recent carbon materials: fullerenes [15], carbon nanotubes [16] and graphene [17]. However, the excellent properties of these novel carbon forms also fostered the interest in the traditional porous carbons and, in the recent years, a considerable number of studies searching for new synthetic approaches have been published. The main objective is the preparation of highly porous materials with controlled porosity, and often also with tuned surface chemistry, to present enhanced behavior as, for example, electrode materials for supercapacitors.

Properties of nanoporous carbons
Describing nanoporous carbon materials as carbon structures with enclosed porosity is correct but it is somewhat incomplete since properties, as for example, surface chemistry, chemical composition, morphology or electrical conductivity are also essential to understand their potential and performance in a given application.
Regarding the physical properties of nanoporous carbon materials, porosity is undoubtedly the most important one. IUPAC defines nanopores as pores with apertures up to around 100 nm and, according to their size, nanopores are classified as follows [18]: • Micropores-pores with width less than 2 nm • Mesopores-pores with width between 2 and 50 nm • Macropores-pores with width greater than 50 nm Due to their relevance for the adsorption process, micropores can be further distinguished between narrow micropores (i.e. ultramicropores) with width <0.7 nm and wide micropores (i.e. supermicropores) with width between 0.7 and 2 nm [18].
The limits for this pore classification were suggested by nitrogen adsorption-desorption data at −196°C [18] and rely on the fact that each pore size range corresponds to different pore filling mechanisms disclosed by the isotherm profile [19]. Micropore filling is regarded as a primary physisorption process and occurs at low relative pressures. Firstly, at very low relative pressure, adsorption occurs in the ultramicropores where due to the close vicinity of pore walls there is an overlapping of the adsorption forces favoring the enhancement of the adsorbent-adsorbate interaction, this process is termed "primary micropore filling." After this initial filling in the most energetic sites, the cooperative adsorbate-adsorbate interactions inside the supermicropores become more relevant (secondary process) occurring up to relative pressures ≈ 0.15 [18]. In mesopores, the physisorption comprises three stages: monolayer adsorption (i.e. all adsorbate molecules are in contact with the surface of the adsorbent) followed by consecutive adsorbate layers (multilayer adsorption) and finishing with capillary condensation [18]. The adsorption onto mesopores commonly presents a hysteresis loop [19]. The macropores are so wide that do not allow capillary condensation since behave as an open space (cannot be characterized by nitrogen adsorption whose maximum relative pressure occurs near the unity) [19].
In nanoporous carbons, the porosity results from the voids between the randomly cross-linked graphite-like crystallites that constitute the structure of the carbon skeleton, from the consumption of less organized carbonaceous matter and also from the removal of reactive carbon atoms in the crystallite during activation process (Figure 1). Surface area and pore size distribution (PSD) are the physical properties that will greatly influence the performance of nanoporous carbons in a given process. These parameters are commonly obtained from the analysis of nitrogen adsorption isotherm at −196°C. The surface area (apparent surface area for microporous materials) is commonly obtained applying the Brunauer-Emmett-Teller (BET) equation. Regarding the PSD, there are various mathematic models available, being those using the Non-Local Density Functional Theory (NLDFT) the most advanced ones. The use of nitrogen as adsorptive at −196°C has however diffusional limitations in ultamicroporous carbon materials; thus, for an accurate characterization of the micropore network, carbon dioxide adsorption at 0°C has been repeatedly proposed as a better alternative to nitrogen [20][21][22]. Besides the limitation of nitrogen for the characterization of ultramicropores, it is known that, due to its quadrupole moment, nitrogen molecule orientation during adsorption will be dependent on the surface chemistry of the adsorbent [18,22]. To overcome this issue, other adsorptive gases like argon or krypton can be used.
Carbon atoms are the major element of nanoporous carbon skeleton but the elemental composition of these materials also includes hydrogen and oxygen and, depending on the precursor, preparation route and post-synthesis functionalization may also have nitrogen, sulfur or phosphorus containing groups. These heteroatoms are mainly located at the edges of the basal planes due to the presence of unsaturated carbon atoms that are highly reactive. Figure 1 summarizes the most relevant oxygen, nitrogen, and sulfur-containing surface groups that may be present in the surface of nanoporous carbon materials, being important to mention that oxygen surface groups are, by far, the most common ones. These materials may also contain inorganic matter, sometimes attaining 20% (wt.), being this ash content mainly inherited from the carbon precursor [12].
The elemental composition and type of surface groups of a nanoporous carbon influence its performance in both gaseous and liquid phase processes, due to specific interactions with the adsorptive and also the solvent in the case of adsorption from solution. Properties like hydrophobicity/hydrophilicity or acid/base behavior are highly dependent on the surface chemistry of these materials. Carbons are in general hydrophobic, but the presence of increasing amounts of oxygen surface functionalities favors the adsorption of water molecules due to the formation of hydrogen bonds and, consequently, its wettability. This ability of porous carbon materials may be advantageous, or not, depending on the desired application. For example, in the impregnation of carbon supports with catalysts in aqueous phase, a high wettability will increase the impregnation degree; but in the adsorption of organic compounds from diluted aqueous solutions, higher wettability may led to the formation of water molecule clusters that hinder the diffusion of the target compounds towards the adsorption active sites (the same being valid for gas steams purification).
Regarding acid/base character, nanoporous carbons are considered amphoteric materials due to the presence of both acid and basic sites in their surface. Thus, depending on the amount and strength of all the surface groups, the materials may present net acid, basic or neutral surfaces. The surface chemistry of nanoporous carbons can be assessed through numerous techniques, and the best way to get a good characterization is to employ complementary techniques and combine the analysis of the results. For example, Boehm and potentiometric titrations provide qualitative and quantitative information on the surface of the nanoporous carbons, while temperature programmed desorption (TPD) detects more oxygen groups than Boehm titration, although with less quantitative information. On the other hand, X-ray photoelectron spectroscopy (XPS) and diffuse reflectance infrared spectroscopy (DRIFT) provide qualitative information about the surface of the nanoporous carbons [23].
Other properties of nanoporous carbon materials that may play important roles in specific applications are the morphology (powdered (PAC) and granular (GAC) forms are the most common ones), electrical conductivity, hardness and density.

Nanoporous carbon synthesis
There are a large range of carbon-rich precursors namely from vegetable (e.g. wood, fruit shells, stones or peels) or petrochemical origin that can be successfully transformed into activated carbons. However, the economically viable large-scale production of activated carbons requests the use of raw materials that gather high availability and constancy with high density and hardness, low inorganic content and, last but not the least, low cost. Thus, commercially available activated carbons are mainly produced from coal (i.e. lignite, sub-bituminous, bituminous and anthracite), coconut shell charcoal and wood charcoal. The raw material is a key issue for activated carbon producers since the development of the porous network occurs by consumption of the (char)coal precursor and the preparation yields can reach values as low as 10% (wt.). Consequently, assuring the supply of high quality precursors at a cost-effective price is of paramount importance to control the quality of the final material and the production cost. In 2015, the global activated carbon market was valued by Markets and Markets at USD 4.74 billion corresponding to 2743.7 kton being expected to continue to growth, mainly driven by more stringent government regulations to assure human and environmental protection. As already mentioned, (char)coal is the most important raw material for activated carbon producers but they are not the major (char)coal consumer. In fact, activated carbons are only a "by-product" of (char)coal production, which have the energy sector as major market (mainly for metallurgical, industrial and cooking fuel) that actually controls demand and prices. This reality, allied to the fact that most of raw materials came from the Asian continent, and that external factors (e.g. natural disasters) may compromise the quality, availability and cost of (char)coal, continues to drive the research for finding new, and more sustainable, raw materials and preparation routes for activated carbon production.
Conventional methods for activated carbon production request the use of high temperature kilns where, under a controlled atmosphere, the carbon-rich raw materials are transformed into nanoporous carbons. There are numerous procedures for the production of activated carbons, although the great majority of synthetic routes published and patents registered worldwide, including those allowing nowadays industrial production, fall in two major categories: the selective gasification of raw materials' carbon atoms (physical or thermal activation), and the co-carbonization of the precursor with oxidizing and/or dehydrating agents (chemical activation). The number of steps requested for the production of activated carbons depends on the characteristics of the raw material, morphological specifications of the final product, and on the type of activation. A general process flowsheet is presented in Figure 2.
The most commonly used pre-treatments are those aiming to obtain a given particle size or shape, washing steps for dirt removal and/or for the reduction of the inorganic content (i.e. acid washing) and pre-oxidation to prevent fluidization of coking coal during carbonization [8,12]. The raw material can be crushed and sieved to obtain powdered or granular particles with specific dimensions. Milling, binder addition and briquetting are other common options when working with precursors presenting low volatile content (e.g. medium-high rank coals, volatiles <20-30% (wt.)). To assure an efficient diffusion of the activating agent to all the individual coal particles of the briquette, it is common to mill and subsequently briquette the particles to obtain granular materials with a well-developed network of transport pores (i.e. meso and macropores) [8]. Direct activation of granular coal with low volatile content will not allow an uniform activation of the granules due to the incipient network of transport pores.
Physical or thermal activation is generally made in two consecutive heating steps: carbonization under inert atmosphere (usually nitrogen) to devolatilize the raw material, followed by activation that consists in the partial gasification of the obtained char with oxidizing agents (i.e. steam, carbon dioxide or a mixture of both) leading to the formation of the porous network. While the carbonization usually occurs at temperatures between 400 and 600°C, the gasification requests temperatures ranging from 800 to 1000°C. Depending on the raw material, it is also possible to avoid the carbonization step and proceed directly to the thermal activation. Activation with CO 2 must occur in conditions that assure chemical control (slow activation rate-days) instead of diffusional control which is faster but leads to external particle burning and, consequently, to poor porosity development [8]. The reactions of steam and carbon dioxide with carbon are endothermic, thus thermal activation needs external energy supply to maintain the requested high temperature. Oxygen (or air) is not commonly used as oxidizing agent since its reaction with carbon is highly exothermic and fast, being difficult to control and assure porosity development instead of particle consumption [8,12,24]. Due to this and to the safety issues related to the temperature control, oxygen activation is scarcely used. However, low amounts of oxygen (or air) can be added during thermal activation with steam or carbon dioxide to help maintaining the high temperatures by reacting with the gases produced during activation (i.e. CO and H 2 ) [24]. This approach has the advantage of lowering the pressure of CO and H 2 , both inhibiting gases for the activation, and increasing the partial pressure of the activating agent [24].
Chemical activation usually requests only one heating step: the raw material is mixed with an activating agent (e.g. ZnCl 2 , H 3 PO 4 , KOH) and is further treated under controlled atmosphere at temperatures from 400 to 900°C, depending on the selected activating agent. The solid product obtained need to be exhaustively washed with water to remove the chemicals and dried before storage. In this process, the mechanism of pore formation is dependent on the chemical agent: zinc chloride promotes the removal of water molecules from the lignocellulosic structures of the raw material while phosphoric acid combines chemically with them [24]. In none of these processes, the selective removal of carbon atoms occurs. With potassium hydroxide, the process is more complex since there is the disintegration of the structure followed by intercalation of the metallic potassium [24,28,29] into the "graphitic" laminar structure, breaking down particles and limiting granular activated carbons synthesis. Simultaneously, there is also some gasification of carbon atoms due to reaction with CO 2 and H 2 O, resulting from the redox reaction of carbon with potassium compounds [30]. During carbonization (i.e. heat treatment under inert atmosphere), the lignocellulosic precursor losses volume by contraction, but when chemical activation is applied there is incorporation of the activating reagent inside the particles inhibiting the expected contraction, that is, the activating agent may behave as a template for the formation of the microporosity [24]. In fact, in 2004, Molina-Sabio and Rodríguez-Reinoso [30] published an enlightening research that allowed to conclude that the dehydration of the carbon precursor (i.e. peach stones) by ZnCl 2 occurs along with templating by the activating agent, thus justifying the porosity evolution with the amount of ZnCl 2 impregnated. Chemical activation has advantages over physical process related with the use of a single heating step and lower temperatures, usually higher yields, shorter activation time (hours) and higher surface area and pore volumes. All these add-ons came with a price since this activation process is more corrosive (e.g. ZnCl 2 or KOH) than the physical one, and activating agents can be hazardous for the environment (i.e. ZnCl 2 and H 3 PO 4 ). Also, regardless the activating agent, a time-consuming washing step is needed to remove chemicals from the newly formed pore network and, if possible, to recover the activating agent. Actually, ZnCl 2 use is declining due to the environmental issues associated with zinc residues [12], low recovery efficiencies, corrosion problems and presence of residual zinc in final carbon [8].
Zinc chlorine and phosphoric acid lead to higher activation yields (≈40%) compared to physical activation since the use of these compounds inhibit the formation of tar and other by-products formed in physical activation [12]. In the case of potassium hydroxide, the more complex activation mechanism gathering K intercalation in carbon lattice and gasification allows to prepare activated carbon with a large range of yields depending on the precursor: from 80 and 70% [31][32][33] to values, as low as, 20-10% [34][35][36]. KOH allows obtaining superactivated carbons (over 3000 m 2 g −1 ) when high KOH/precursor weight ratios (i.e. between 2 and 4) are used. Table 1 gathers information regarding the most suitable precursors for the commonly used activating agents, as well as, the type of porosity obtained, and the effect of experimental conditions on the pore size distribution. It is important to emphasize that the effect of experimental conditions on pore size distribution presented in Table 1 is a general trend; thus, for an in-depth knowledge of the effect of activating agent and experimental conditions on the porous characteristics of a given raw material, a comprehensive study is always needed.
The above-mentioned precursors and activating agents are those more representative for industrial scale production of activated carbons. However, regarding the production of specialty carbons (low volume processes) and for research works, the number of options increases exponentially. Regarding raw materials, hard biomass residues like shells and stones allow to obtain high quality PACs or GACs; phenol formaldehyde polymers yield high surface area porous carbons, while polyacrylonitrile (PAN), acrylonitrile textile or rayon are adequate to synthetize activated carbon fibers or cloths [24]. When aiming nitrogen doped materials, for example for the adsorption of sulfur species, the use of polymers containing nitrogen (e.g. PAN) is a common synthetic route. In what concerns activating agents, besides ZnCl 2 , H 3 PO 4 and KOH, alkali hydroxides and salts (e.g. NaOH, K 2 CO 3 , Na 2 CO 3 ) and metal chlorides (e.g. AlCl 3 , FeCl 3 , NH 4 Cl, CuCl 2 ) can also be used [24].
At industrial scale, steam activation continues to be the most widely used method to produce activated carbons. The advantages of steam activation are related with the fact that does not request post-activation work-up, namely a final washing step, is less expensive and has less environmental constrains than the chemical activation. Besides, for precursors with less than 10% of ash, carbon materials with surface areas of 1000 m 2 g −1 can be easily obtained with activation yields of 50% [8], thus allowing a good compromise between production cost and porosity development.
In general, the above-described activating methodologies produce activated carbons with wide pore size distributions, so the need for carbon materials with more regular porosity prompted the research on the synthesis of ordered mesoporous carbons (OMCs) by applying hard or soft templating approaches. The advantages of OMCs over conventional activated carbons are related with the ordered and hierarchical pore network, but their multi-step preparation procedures are time consuming, use hazardous chemicals to remove the inorganic templates and, consequently, have very low atomic economy and high production costs. This class of nanostructured materials is out of the scope of the present chapter since the first works were published in the late 1990s, and comprehensive reviews are available in the literature, providing a broad overview and up-to-date information on synthesis and properties of this class of porous carbon materials [37][38][39]. In the following, the recent developments regarding new synthesis approaches-i.e. precursors, activating agent and routes-for preparing nanoporous carbon materials are presented.

Conventional activation of gels and chars obtained by novel approaches
As already mentioned, classical preparation of activated carbons usually requests an initial thermochemical process (carbonization) to remove other elements than carbon and transform Micro -KOH/precursor ratio has more influence on adsorption capacity and PSD than activation temperature -Increase of KOH/precursor ratio widens pores from narrow to large micropores and, in less extend, to small mesopores; also hinders the granular morphology (particle disintegration leads to powders) Table 1. Appropriate precursors, kinetic of activation, and type of porosity usually obtained for the most common activating agents (information gathered from reference literature [8,12,24,29]). the raw material (i.e. coal or biomass) in a carbon-rich material (charcoal) that will be subsequently activated to improve the pore network. This thermochemical conversion is particularly important when using biomass whose carbon content usually is between 30 and 50% (wt.) [25]. However, the need for more specialized porous carbon materials, as well as the growing interest in developing more sustainable and greener processes, has prompted the research community to explore alternatives to the conventional carbonization step which, in some cases, also allow to make feasible the use of novel precursors.
In the category of alternative methods to obtain carbon-rich materials for further activation via conventional methodologies, processes as distinct as sol-gel polymerization reaction, hydrothermal carbonization (HTC) and acid-mediated carbonization (AMC) can be grouped. While sol-gel process involves the use of synthetic compounds-resorcinol and formaldehyde-as reagents for the polycondensation reaction in the presence of a catalyst, both HTC and AMC are mainly applied to transform biomass into a coal-like material.
The first resorcinol-formaldehyde organic gel was synthetized by Pekala in 1989 being the aqueous polycondensation performed under alkaline conditions [40]. One of the major advantages of this process is its flexibility since the main properties of the gel can be tuned during the synthesis and drying process. The synthesis starts with the formation of the wet gel and, during this step, the most important parameters controlling the properties of the final gel are the precursors' concentration, the catalyst type and concentration and the time and temperature of curing [41]. After drying, aerogels (supercritical drying), xerogels (subcritical drying), and cryogels (freeze drying) can be obtained. The drying procedure is one of the most important steps regarding the porous properties of the organic gel and is also crucial when the preparation of a thermally stable carbon gel is foreseen. The organic gels are mainly mesoporous materials thus if the preparation of a micro + mesoporous material is envisage a thermochemical step involving carbonization and/or activation must be considered. The first works focused on the carbonization and activation of organic gels were published during the 1990s, and nowadays, these methodologies continue to be studied to better tune the pore network of these materials and explore novel applications.
A great advantage of conventional activation of organic gels over the previously mentioned carbon precursors is the possibility of coupling the mesopore network of the organic gels with micropores created during carbonization or activation step. Tamon et al. studied the effect of carbonization on the porous properties of aerogels and cryogels proving that both precursors preserved the mesoporous structure after the thermal treatment and that cryogels are more prone to develop micropores [42]. Regarding activation, literature data reveal that the gasification of aerogels with CO 2 allows retaining the mesopore network of the organic gels and increasing the volume of micropores [43,44]. Moreno-Castilla and co-workers reported the activation of a mesoporous aerogel with CO 2 showing that for a 22% burn-off the increase and widening of the precursor micropore network practically did not change the mesopore structure [44]. In a recent work, Ania and co-workers [45] evaluated the effect of carbonization and activation (physical and chemical) of xerogels which, due to their drying in subcritical conditions, may suffer stronger changes upon thermal treatment. In fact, both carbonization and CO 2 activation lead to the increase of the micropore volume at the expense of a severe decrease in the mesopore volume and shrinkage of the average pore size. However, under controlled experimental conditions (i.e. impregnation methodology and temperature), the chemical activation of the xerogel with KOH or K 2 CO 3 allowed to suppress the shrinkage and structural collapse, forming a micropore structure associated with the enlargement and/ or preservation of the mesopore network of the pristine xerogel [45]. For further discussion and information regarding the control of the properties of resorcinol-formaldehyde organic and carbon gels, the reader can refer to the reviews by Al-Muhtaseb and co-workers [41,46].
Hydrothermal carbonization (HTC) is probably one of the most promising alternatives to conventional carbonization of biomass, as it is clearly shown by the ever-increasing number of publications focused in this process since the beginning of the twentieth century. HTC is inspired in the natural processes of coal formation that take millions years and request temperature and pressure in a nonoxidizing atmosphere to transform biomass into a carbon-rich material. Interestingly, the first research work on HTC was published in 1913 by Friedrich Bergius, a Nobel Laureate in recognition of his studies regarding carbon-water reactions at high pressure and mild temperature to successfully produce H 2 by avoiding CO generation [47]. When performing the experiments, Bergius noticed that when peat was used as carbon source the solid residue obtained had an elemental composition similar to that of coal, leading him to investigate the HTC decomposition of plant-based compounds into coal-like materials [48]. As reported in [49], with the exception of the works by Berl and Schmidt (1932) and van Krevelen et al. (1960), the interest in the solid carbon materials obtained from HTC process was forgotten until the work published by Wang et al. [50] in 2001. These authors reported the synthesis of carbon spheres from sugar under hydrothermal conditions (190°C and self-generated pressure in a high-pressure vessel). HTC is a cost effective and eco-friendly process; since to convert biomass into carbon-rich materials, it uses water as solvent, mild temperatures, self-generated pressure and occurs in few hours with no CO 2 emission. Since 2001, the interest in this carbonization process has been exponentially increasing, and in 2016, around 400 papers mentioning hydrothermal carbonization were published, and the works focusing in this process received more than 11,000 citations (source Web of Science, Sept 2017). Some advantages of HTC-derived materials, commonly known as hydrochars, over charcoals obtained by conventional carbonization is their high content in oxygenated groups and the possibility of morphological control (e.g. spherical morphology).
Hydrochars have an incipient pore network; thus, conventional activation has been commonly employed to obtain specialized nanoporous carbon materials, mainly regarding surface functionalization, morphology features and ultrahigh surface areas and pore volumes. The use of hydrochars as activated carbon precursors was first reported by Zhao et al. [51] aiming to prepare a nitrogencontaining porous carbon material to be used as supercapacitor. The authors performed HTC of D-glucosamine (carbon source) and the hydrochar containing 6.7% of nitrogen was further activated with KOH allowing to obtain a nitrogen-doped porous carbon with a BET area of 600 m 2 g −1 .
As it can be clearly seen in the overview of hydrochar-derived activated carbons presented in Table 2, the great majority of studies concerning hydrochar activation used KOH as activating agent, mainly because, in general, this oxidizing agent allows obtaining the most developed microporosities reaching BET surface areas higher than 3000 m 2 g −1 . It is also interesting to notice that there are studies reporting the use of hydrochars for the synthesis of micro + mesoporous Rattan stalks (Lacosperma secundiflorum) 1135 V Meso /V Total = 72% [55] materials via chemical activation with zinc chloride [52,53] or hydroxides [54][55][56], with the mesopore volume corresponding to at least 50% of the total pore volume, i.e. textural characteristics that are not easy to obtain by the most conventional carbonization + activation process. A recent work reports the preparation of an N-doped nanoporous carbon with similar volumes of micro and mesopores and a BET area near 1000 m 2 g −1 by using HTC of bamboo shoots in the presence of water and 0.5 cm 3 of H 2 SO 4 followed by the carbonization of the obtained hydrochar, so in the absence of any activating agent [57]. The work by Fuertes and Sevilla enables the synthesis of superactivated porous carbons with 50% of mesopores by adding melamine during the HTC process, thus the final materials also present 1.3-1.7% of nitrogen [56]. Besides opening the possibility of preparing mesoporous carbon materials, the use of hydrochars as activated carbon precursors also allows better control of the micropore size distribution [34,35,58].
When starting from simple carbohydrates, the use of HTC followed by activation may enable preparing carbons with spherical morphology (diameters in the micrometer range) and smooth surfaces [34,35,59,60]. It is also possible to obtain a nanoporous carbon material with spherical shape, but with blackberry morphology, by adding acrylic acid to starch during the HTC step [61].
The use of water as solvent constitutes a major advantage of HTC over convention carbonization that requests dry biomass. In fact, HTC makes feasible the use of wet biomass as is the case of algae [62,63] or sewage sludge [53] that after being carbonized can be activated to obtain a carbon material with well-developed pore structure and also heteroatom-rich surface (i.e. nitrogen and sulfur). HTC is, in fact, a valuable process when envisaging heteroatom-doped carbon materials, since if precursors with nitrogen and/or sulfur-e.g. glucosamine, algae, sewage V Meso and V Total , microporous and total pore volume, respectively; PSD, pore size distribution. sludge, tobacco rods or bamboo shoots-were selected this methodology allows retaining higher percentages of heteroatoms compared to conventional carbonization, and thus assures relevant amounts of these atoms after the activation step [53,54,57,62,63].
Following a somewhat different approach, White et al. reported the preparation of nitrogendoped carbogel materials by reacting glucose and ovalbumin (secondary biomass precursor) in HTC conditions [64]. The novelty was the saturation of the monolithic hydrochar with ethanol followed by extraction under supercritical conditions, resembling the last step of the synthesis of organic aerogels. The final material was obtained after carbonization (350-900°C). When using 550°C, a material with an interconnected 3D pore system, BET area of 476 m 2 g −1 and 5-7% (wt.) nitrogen content was obtained.
Very recently, Sevilla et al. proposed an alternative to KOH activation of biomass-derived hydrochars by using potassium oxalate (K 2 C 2 O 4 ) and melamine to obtain N-doped superactivated carbons with high yields (40-46%) [65]. The materials attained BET areas near 3000 m 2 g −1 , presenting mainly micropores (70%) and nitrogen content between 0.5 and 0.9% irrespectively the amount of melamine used. Although the mixture of potassium oxalate/melamine is presented by the authors as an alternative to the corrosive KOH, it is important to highlight that this synthesis route is restricted for lab-scale (under restricted security conditions) since the KCN present in the solid will generate dangerous toxic vapors of HCN during the washing with HCl.
For a deeper overview regarding the HTC process and derived carbon materials, a recent book chapter by Titirici et al. [48] and a minireview [66] are recommended, and for deeper analysis on hydrochar-derived activated carbons, the following review [67] can be consulted.
Acid-mediated carbonization (AMC) is so far the less studied alternative method to obtain a carbon-rich material. However, considering that acid catalysis is a common practice to extract sugars from lignocellulosic biomass [86] and that, as just discussed, sugars can be successfully used as activated carbon precursors (i.e. via HTC followed by activation), this process must also be explored. This methodology allows processing largely available and hopefully low cost biomass residues. Moreover, AMC occurs at atmospheric pressure and, in some cases, at low temperature (≈ 100°C) allowing processing carbon precursors that do not gather the adequate properties to be carbonized through the conventional process or by HTC. In fact, AMC is able to maximize the availability of cellulose and hemicellulose in carbon precursors with high inorganic content (e.g. rice husk) [87,88] and is also efficient for producing carbonrich materials from liquid carbon precursors, as is the case of glycerol [89]. Depending on the acid catalyst used (e.g. H 2 SO 4 and/or H 3 PO 4 ), the obtained oxygen-rich chars also contain relevant amounts of sulfur or phosphorus groups, thus enhancing the reactivity for subsequent activation and allowing to join a well-developed pore network with heteroatom doping in the final activated carbon.
Wang et al. evaluated sulfuric acid hydrolysis of rice husk followed by the dehydration, polymerization and carbonization of the sugars to yield chars, which can be named as acidchars. These materials were activated with H 3 PO 4 [87,88] and KOH [88] to produce nanoporous carbons with BET areas higher than 2400 m 2 g −1 , and in the case of phosphoric acid activation, materials with micro + mesopore networks. In 2010, the authors reported the effect of H 2 SO 4 concentration during carbonization step, as well as the reaction temperature and time onto the yield of the obtained acidchar that attained 32% (wt.) for the reaction at 95°C during 10 h when 72% H 2 SO 4 solution was used in both hydrolysis and carbonization [88]. Regarding the effect of the H 2 SO 4 concentration (42-72%) during acid carbonization onto the textural properties of the activated carbon obtained by H 3 PO 4 activation, the authors concluded that although acidchars prepared under distinct acid concentrations present similar elemental analysis, the BET area and pore volumes increase as the concentration of the H 2 SO 4 decreases [87]. These findings were rationalized by the authors considering that high H 2 SO 4 concentrations promote the aggregation of carbon nanoparticles during carbonization, limiting the uniform H 3 PO 4 impregnation of the particles and consequently the development of the pore network in the inner particle [87].
In a recent publication, Cui and Atkinson [89] systematically investigated liquid glycerol AMC using various acid catalysts (i.e. H 2 SO 4 , H 3 PO 4 , HCl and CH 3 COOH) aiming to study the effect of the acid carbonization conditions onto the textural properties of glycerol-derived nanoporous carbon materials obtained by subsequent physical activation with steam and CO 2 . The AMC of glycerol was made under nitrogen between 400 and 800°C, and the highest carbonization yields were obtained at 400°C for 10:3 volumetric mixtures glycerol:acid (30% yield for H 2 SO 4 and 50% yield for H 3 PO 4 ). Upon activation, materials with BET area values ranging between 990 and 2470 m 2 g −1 and tailored porosity were obtained. The H 3 PO 4 -char originated micro + mesoporous carbon materials with the volume of mesopores being more than 50% of the total porosity regardless the physical activating agent and the amount of acid during the AMC, while the steam activation of the H 2 SO 4 -char lead to materials with 40-44% mesopore volume, and the CO 2 activation only 22-25% mesopore volume. The elemental analysis of the nanoporous carbon materials obtained by H 2 SO 4 carbonization revealed the presence of 0.35-0.71% of sulfur and the materials carbonized with H 3 PO 4 and activated with CO 2 attained even higher heteroatom doping with phosphorus content between 2.04 and 4.34%.

Variations of hydrothermal carbonization (HTC) process
In parallel with the studies centered in the activation of hydrochars, in the last few years, the scientific community also started to explore new synthesis routes to obtain a porous structure during the HTC step, being prepared porous carbons with BET surface areas up to 700m 2 g −1 .
The methodologies proposed avoid the need of further thermal or chemical activation, may enable the synthesis of heteroatom-doped solids [90] and can also allow the synthesis of hierarchical materials [91].
Fechler et al. reported the synthesis of porous carbon materials with BET areas between 425 and 672 m 2 g −1 through HTC (180°C overnight) of glucose mixed with several eutectic salt mixtures (i.e. LiCl/ZnCl 2 , NaCl/ZnCl 2 , and KCl/ZnCl 2 ) in the presence of a small amount of water [90]. The authors proved that both the amount of water added and the salt composition are determinant for the successful synthesis of materials, which are formed by very small particles aggregation, identical to aerogels. The use of the eutectic mixture LiCl/ZnCl 2 originated the carbon materials with the highest surface area. When 2-pyrrol-carboxyaldehyde was added as co-reagent of the mixture glucose and ZnCl 2 , a material with 3% of nitrogen and BET area of 576 m 2 g −1 was obtained. Fellinger et al. used glucose as carbon source and borax (Na 2 B 4 O 7 ) as both catalyst and structure-directing agent to prepare hierarchical structured carbogels under HTC conditions [91]. Further carbonization under nitrogen at 550 or 1000°C allowed to obtain a carbon material with a BET area of 614 m 2 g −1 and 70% of mesopore volume.

Carbonization of organic salts
In 2010, Zhou and co-workers developed mesoporous carbons by the carbonization of organic salts (magnesium and barium citrates) evaluating the influence of the temperature of the thermal treatment (600-800°C) [92]. In the case of magnesium salt, BET areas up to 2322 m 2 g −1 were attained, and the increase of temperature resulted in a progressive increase of the mesopore volume percentage (from 50 to 100%). The barium citrate-derived materials are mainly mesoporous (>90%) and, independently of the temperature, pores between 10 and 20 nm are obtained. Calcium citrate was tested by other authors who reported the paramount importance of temperature in the textural properties of the mesoporous carbons [93].
Atkinson and Rood proposed the use of dichloroacetates of alkaline metals as carbon precursors for the synthesis of nanoporous carbons, being the pore networks, once again, dependent on the cation [94]. The pore structure is produced by fast pyrolysis (15 s to 7 min) under nitrogen flow at temperatures between 300 and 1100°C, and materials attained BET areas of 740 m 2 g −1 . This methodology allowed the synthesis of microporous materials when lithium salt was used and micro + mesoporous solids for the other two metals. In the same research line, Xu and co-workers reported the pyrolysis of EDTA salts to obtain nitrogendoped porous carbons with BET areas reaching 1800 m 2 g −1 and porosity characteristics dependent on the thermal treatment temperature (higher the temperature, higher the mesopore volume) [95,96].
The protocol of organic salts carbonization was extended to gluconates and alginates along with citrates to understand the mechanism of the porosity development [97][98][99]. The results shown that the textural properties of the porous carbon materials obtained by this methodology are heavily dependent on the type of the cation in the organic salt: while potassium salts originate essentially microporous solids, for sodium and calcium the amount of mesopores is also relevant [98]. In the case of calcium citrate-derived material, the development of the mesoporosity was attributed to the formation of CaO nanoparticles, which act as endotemplates during the carbonization. It was also shown that the nature of the organic salt has a great impact on the morphology, with sodium gluconate leading to the formation of large carbon nanosheets, while sodium citrate originates sponge-like particles. The synthesis of nitrogendoped porous carbons was also explored by mixing potassium gluconate with melamine, which allowed to obtain a microporous material gathering 22.9% of nitrogen with 660 m 2 g −1 of BET area [97]. The endotemplate approach on the carbonization of organic salts was further explored with iron, calcium and zinc citrates [99]. The carbonization of these nonalkali organic salts produces mesoporous materials with BET areas between 950 and 1610 m 2 g −1 and distinct pore size distributions: monomodal distribution centered at 11 nm for calcium citrate, bimodal distribution centered at 9 and 20 nm for iron citrate, and bimodal distribution centered at 3 and 10 nm for iron citrate. These carbons can be post-functionalized by heat treatment in the presence of melamine to obtain materials gathering high BET area and mesopore volume with high nitrogen content (8-9%).

Ionothermal approaches
The synthesis of porous carbon materials via ionothermal concept derives from the methodologies proposed in 2004 by Morris and co-workers to prepare zeolite analogues (i.e. inorganic materials) using ionic liquids or eutectic mixtures as both solvents and inorganic structure directing agents (templates) [100]. The ionothermal approach is analogous to hydrothermal and solvothermal processes, where the solvents are predominantly molecular (water or nonaqueous solvents, respectively) but, as the name means, it refers to processes occurring in ionic solvents (e.g. ionic liquids or eutectic mixtures) [100]. The low vapor pressure of the ionic solvents is a great advantage of ionothermal reaction over the hydrothermal or solvothermal concepts, since it allows avoiding the safety concerns connected with the high pressures required to prevent molecular solvent evaporation [101].
The preparation of carbon materials following the ionothermal principles is reported in the literature as both ionothermal and molten salt synthesis processes. Nowadays, there is still not a generally accepted terminology for these processes what turns difficult to understand the classifications and underlying procedures followed by the distinct authors. In fact, the use of ionothermal/molten salt process can be linked either to the preparation of carbon materials with incipient porosity obtained at temperatures ≈ 200°C (ionothermal carbonization-ITC) [102] or to the preparation of porous carbons by a two-step process including the previously mentioned ITC followed by a thermal treatment of the ionothermal derived carbon at high temperatures (attaining 1000°C or more) [103,104]. Ionothermal/molten salt process is also considered in the case where the mixture of the carbon precursor and the ionic solvent is directly thermally treated at high temperature [105,106].
The first studies reporting the preparation of porous carbon materials through ionothermal process used ionic liquids as carbon precursors. In 2009, Lee et al. synthesized N-doped materials (2-3%) with BET area values between 640 and 780 m 2 g −1 , and the authors demonstrated the influence of the ionic liquid nature on the development of mainly microporous or micro + mesoporous materials [107]. In a further work, the authors report a similar process for obtaining materials with up to 17% of nitrogen content, although with lower porosity development [108]. In 2010, other approach of the same research group gathered an ionic liquid with simple carbohydrates allowing to obtain carbon materials with a highly developed mesopore network after treatment at only 200°C during 20 h in a nonpressurized chamber [102]. Xie et al. reported the synthesis of magnetic hierarchical porous carbons by using several carbohydrates and an iron containing ionic liquid, and the authors proposed that the ionic liquid has a triple role: salt template, solvent and catalyst [103].
In the subsequent research in ionothermal approaches (Table 3), the introduction of salts as co-reagents became common and generally accepted as a hard template route. Zinc chloride is by far the most frequently reported salt, used alone or in mixtures (eutectic or not) with other salts. This strategy allowed to synthesize porous carbons with ultrahigh surface area (easily around 2000 m 2 g −1 ), hierarchical structure and high heteroatom doping (i.e. nitrogen (>5%) and sulfur). All these features are of fundamental importance for boosting the application of these materials in energy storage processes. Actually, in the great majority of publications, the authors report high performance of the ionothermal-derived porous carbons as supercapacitors. While initially nitrogen-containing ionic liquids were used as carbon precursors [109][110][111], along the years, the number of studies exploring carbohydrates, or even biomass, as carbon source has increased [105,106,[112][113][114][115][116][117][118][119][120]. More elaborated synthesis schemes have been reported, as is the case of the work developed by Chang et al. [121] where organic solvents were added dropwise to molten ZnCl 2 at 550°C. By changing the solvent, the authors were able to obtain nitrogen (14%) or sulfur (13%) doped porous carbons with different morphologies (i.e. aerogel, nanosheets or hyperbranch). There are also reports on ITC of biomass with nitrogen-containing ionic liquid followed by conventional KOH activation to yield nitrogen doping up to 1.59% and apparent surface areas of 2838 m 2 g −1 [124]. The ionothermal approach is actually a powerful route to synthesize porous carbon materials with valuable graphene-like (2D) structures [125][126][127][128].
It is also possible to found reports on highly mesoporous carbon materials obtained by ZnCl 2 -mediated ionothermal/molten salt synthesis [129][130][131]. Although these routes are presented as novel synthesis procedures, the high mesopore volumes reported are most probably the result of the complex ZnCl 2 activation mechanism when very high amount of ZnCl 2 is added. As it was previously mentioned, the behavior of chemical activating agents as templates during carbonization cannot be disregarded, and for the particular case on ZnCl 2, the results obtained by Molina-Sabio and Rodríguez-Reinoso [30] allowed the authors to conclude that this chemical acts as template for the creation of porosity. So, although new materials are being produced under "novel" synthesis procedures with appealing names, in some cases, the experimental route and underlying mechanism for pore creation seems to be the one accepted for conventional chemical activation.  For more information regarding the preparation of porous carbon materials via ionothermal/ molten salt approaches and also their main applications, several review works are available in the literature [132][133][134][135][136].

Conclusion
The technological relevance of porous carbon materials continues to prompt the scientific community and companies to explore alternative routes to the conventional methods, in order to develop specialized materials or improve the production process (e.g. optimizing energy costs and minimizing wastes).
The great majority of conventional and innovative processes allowing the synthesis of porous carbons involve heat treatments at moderate-to-high temperatures for carbonization to occur and subsequent formation of the porous carbon skeleton. In the case of chemical activation or ionothermal/molten salt processes, a washing step with water or 10% HCl is required to remove chemical compounds clog the porosity.
Conventional methods are based on a solid and structured carbon material and therefore activation occurs by gasification, selective oxidation of the most reactive carbon atoms and heteroatoms or intercalation processes. Thus, the conventional methodologies are considered top-down processes. Regarding innovative approaches both top-down and bottom-up routes are reported in the literature. HTC or inonothermal reactions promote a top-down process when the carbon precursor is a biomass, but is a bottom-up route when discrete entities (e.g. carbohydrates or ionic liquids) are the starting materials. The acid-mediated degradation of biomass is a top-down route that converts biomass in a carbohydrate-rich acid liquor that is carbonized in a bottom-up approach. Some of the advantages of these novel approaches over the conventional ones are related to the possibility of producing highly porous carbon materials with easier pore size distribution tuning, high amounts of heteroatoms in the surface, and fine control of morphology (e.g. sponge-like, aerogel-like, spherical, sheets (2D)). In light of a more sustainable and circular economy, it is also relevant that some of the novel synthetic approaches (e.g. HTC) allow enabling future largescale production based on high moisture containing biomass residues and even liquids (e.g. glycerol).
The driving force for the development of new porous carbons throughout these nonconventional methods is the search for high performing materials for electrochemical applications and energy storage, which request hierarchically porous structures ideally doped with electron-rich nonmetallic elements (e.g. nitrogen, sulfur) to increase the conductivity.
The carbon atom is a versatile element that since the Stone Age has reinvented itself. So, the overlook of this chapter allows to predict that in the near future the number of novel synthesis routes to feed the demand for even more specialized porous carbons will continue to increase. This may occur by revisiting synthesis routes established for other classes of materials or by the discovery of completely novel processes.