Low-Cost Activated Grape Seed-Derived Hydrochar through Hydrothermal Carbonization and Chemical Activation for Sulfamethoxazole Adsorption

Activated carbons were prepared by chemical activation with KOH, FeCl3 and H3PO4 of the chars obtained via hydrothermal carbonization of grape seeds. The hydrochars prepared at temperatures higher than 200 °C yielded quite similar proximate and ultimate analyses. However, heating value (24.5–31.4 MJ·kg−1) and energy density (1.04–1.33) significantly increased with carbonization temperatures between 180 and 300 °C. All the hydrochars showed negligible BET surface areas, while values between 100 and 845 m2·g−1 were measured by CO2 adsorption at 273 K. Activation of the hydrochars with KOH (activating agent to hydrochar ratio of 3:1 and 750 °C) led to highly porous carbons with around 2200 m2·g−1 BET surface area. Significantly lower values were obtained with FeCl3 (321–417 m2·g−1) and H3PO4 (590–654 m2·g−1), showing these last activated carbons important contributors to mesopores. The resulting materials were tested in the adsorption of sulfamethoxazole from aqueous solution. The adsorption capacity was determined by the porous texture rather than by the surface composition, and analyzed by FTIR and TPD. The adsorption equilibrium data (20 °C) fitted the Langmuir equation well. The KOH-activated carbons yielded fairly high saturation capacity reaching up to 650 mg·g−1.


Introduction
Biomass is a widely available source of energy, particularly important in developing countries, but cannot be considered a technically ideal fuel due to physical and chemical properties, each often having fibrous nature, high moisture content, volatile components, alkali and alkaline earth metallic content and a relatively low bulk density and heating value [1][2][3]. Thus, pre-treatment of biomass is in many cases convenient for efficient use as energy source. A broad range of biological (mainly anaerobic digestion and fermentation) and thermochemical (torrefaction and pyrolysis) treatments are typically used to improve the fuel properties of raw biomass [2,4,5]. From those processes, liquid or gaseous biofuels and even some valuable products are derived.
As an alternative to the classical thermochemical methods, hydrothermal carbonization (HTC), also referred to as wet torrefaction, is becoming an increasingly attractive way of biomass conversion. It operates in presence of water, at comparatively mild temperatures (180-300 • C) and a corresponding saturation pressure [6][7][8][9][10][11]. The resulting solid product, usually called hydrochar (HC), is more stable and has higher carbon content than the starting substrate and improved (higher) heating value with respect to the char resulting from slow-pyrolysis or conventional carbonization at the same temperature [12][13][14]. In addition to the hydrochar, HTC gives rise to a high organic load aqueous stream [15,16], and a gas consisting mainly of CO 2 .
In general, hydrochars show relatively low BET surface areas and pore volumes, which hinders their application as adsorbents [24]. Nevertheless, Titirici [23] reported the presence of ultramicropores (narrow micropores) in several sugar-derived hydrochars, as determined by CO 2 adsorption. Low-cost adsorbents can be obtained from HC upon further activation with different agents (CO 2 , steam, acids, bases, or salts), allowing substantial pore volume and surface area increases, and even changes of the morphological structures of HTC chars [23][24][25]. A number of works have reported on the application of activated hydrochars in aqueous-phase adsorption for the removal of heavy metals [24,[26][27][28], dyes [24,29], phenolic compounds [30] and emerging pollutants [18,31]. Gas-phase applications have been also studied, including, VOCs [32] and CO 2 [31,33] adsorption. Hydrochars treated by H 2 O 2 oxidation have been used for lead removal from water [34].
Agricultural wastes are recognized as interesting feedstocks for inexpensive carbon materials [35]. Among them, grape seeds represent up to 15% of the solid wastes from the wine industry, where they are mostly burnt as fuel. Some papers have reported on the preparation of activated carbons by both physical and chemical activation from this precursor [36][37][38][39][40]. Activated carbon is widely available at low cost in many regions and its structural characteristics (granular morphology, size and preferential distribution of the lignocellulosic material in the periphery of the seed) make it particularly attractive for that purpose [36]. Essentially microporous activated carbons have been obtained by physical (CO 2 ) and chemical (KOH and K 2 CO 3 ) activation [36,40]. Using phosphoric acid as the activating agent allowed obtaining also predominantly microporous carbons but with a broader porous texture with some significant contribution of mesoporosity [36]. Recently, Purnomo et al. [41] reported a two-stage process consisting of HTC followed by KOH-activation of the resulting hydrochar for the preparation of microporous carbons with high BET surface area under milder operating conditions than those used for direct single-step activation with the same agent.
The aim of this work was to study the valorization of grape seeds into carbon materials with the double purpose of obtaining high energy-density solid fuels upon hydrothermal carbonization and low-cost activated carbons by chemical activation of the resulting hydrochar. Temperature and biomass:water ratio have been analysed as operating conditions for HTC and KOH. FeCl 3 and H 3 PO 4 have been tested as activating agents under different conditions. Hydrochars and activated carbons have been characterized by several techniques covering proximate and ultimate analyses, 77 K N 2 adsorption-desorption and CO 2 adsorption isotherms for porous texture, SEM for morphological examination and FTIR and TPD for surface functional groups assessment. Sulfamethoxazole, a sulfonamide antibiotic used to treat urinary infections was selected as the model compound to test the potential of the activated carbons as adsorbents for the removal of water pollutants. That compound is one of the most frequently emerging contaminants found in municipal wastewater, and in many cases is still present in the effluents from sewage treatment plants.

Preparation of Hydrochars
Grape seeds (GSs) provided by a local winery ("Tinta de Toro", Zamora) were used as the hydrochar precursor. Table 1 shows the proximate and ultimate analyses. The HTC process was carried out in a Teflon-lined stainless steel vessel (100 mL), using 20 g of dried grape seeds and different water to GS ratios (10-60 wt.% of GSs). The reactor was sealed, inserted in a muffle furnace (Hobersal serie 8B Mod 12 PR/400, Hobersal, Barcelona, Spain) and heated up to 180-300 • C for 16 h. The hydrochar was recovered by filtration, washed with distilled water and oven-dried at 105 • C for 24 h (Nabertherm R 60/750/12-C6, Nabertherm, Bremen, Germany). The hydrochars were denoted as GS followed by the carbonization temperature and the dry GS percentage in the mixture. For instance, GS-260-40 represents the hydrochar obtained from grape seeds at 260 • C with 40% of dry GS. Table 1 includes also the hydrochar yield values calculated as the mass of the hydrochar per unit mass of GS, both on a dry basis. The energy yield or energy recovery efficiency is calculated by [42] (also known as the energetic biomass utilization efficiency (BUE E ) [43]): Energy recovery efficiency = Hydrochar yield·Energy density, where the energy density is the ratio of the higher heating values of the char relative to that of the grape seeds.

Preparation of Activated Carbons
The activated carbons (ACs) were prepared using either potassium hydroxide (KOH), iron chloride (FeCl 3 ) or phosphoric acid (H 3 PO 4 ), which were physically mixed with the ground hydrochar at different activating agent to HC ratios of 4:1, 3:1 and 2:1 (w:w). In the case of H 3 PO 4 , the mixture was left overnight at 60 • C and then heated at 500 • C for 2 h under continuous N 2 flow (100 NmL·min −1 ) in a stainless steel tube placed in an electrical furnace. The activation temperature was reached at 10 • C·min −1 heating rate. Activation with KOH and FeCl 3 was performed similarly but at 750 • C for 1 h [44][45][46][47]. After cooling under nitrogen flow, the ACs were washed with HCl or NaOH 0.1 M until neutral pH; then were washed with distilled water; and finally, were dried at 105 • C overnight. The ACs were denoted as GSHC, followed by the activating agent and the activating agent:hydrochar (w:w) ratio. For instance, GSHC-KOH-3 represents the activated carbon obtained with KOH using 3:1 activating agent to hydrochar ratio.

Characterization of Hydrochar and Activated Carbons
The proximate analyses were performed by the ASTM methods D3173-11 (moisture), D3174-11 (ash) and D3175-11 (volatile matter (VM)) using a Mettler Toledo apparatus (TGA/SDTA851e). The elemental analyses were carried out in a LECO Model CHNS-932 elemental analyzer. The higher heating value (HHV) of hydrochar samples was determined using an IKA Calorimeter System C2000 according to the ASTM D5865 procedure.
The porous texture of the AC samples was characterized by 77 K N 2 adsorption-desorption in a Micromeritics apparatus (Tristar 3020, Micromeritics, Narcross, GA, USA). The samples were previously degassed (150 • C for 6 h) in a Micromeritics VacPrep 061 device (Micromeritics, Narcross, GA, USA). The surface area was calculated by the BET equation within the 0.05-0.30 relative pressure range and the micropore volume (V micro ) was obtained by the t-method. The difference between the volume of N 2 adsorbed at 0.95 relative pressure (as liquid) and the micropore volume was taken as the mesopore volume (V meso ). The external or non-micropore surface area (A ext ) was also obtained from the t method. Surface area and micropore volume of the samples were also determined by CO 2 adsorption at 273 K in the same equipment. The corresponding surface area (S DA ) and the micropore volume (V micro-DA ) values were calculated by the Dubinin-Astakhov equation [48].
The surface composition of the hydrochars and activated carbons was analyzed by FTIR spectroscopy using a FTIR Bruker IFS66v spectrophotometer (Bruker Corporation, Billerica, MA, USA), with a KBr disc, having a resolution of 4 cm −1 from 4000 to 550 cm −1 and 250 scans. Temperature programmed desorption (TPD) was also used to assess the amount of oxygen surface groups. The samples (0.1 g) were heated up to 900 • C at 10 • C·min −1 in a vertical quartz tube under continuous N 2 flow of 1 NL·min −1 . The amounts of CO 2 and CO evolved were analyzed by non-dispersive infrared absorption in a Siemens model Ultramat 22 equipment (Siemens Aktiengesellschoff, Munich, Germany).
The morphologies of hydrochars and activated carbon samples were studied by scanning electron microscopy (SEM) with a Hitachi S-3000N apparatus (Hitachi Ltd, Tokyo, Japan). The samples were metalized with gold using a Sputter Coater SC502 (Quorum, East Sussex, UK). Images were obtained in the high vacuum mode under an accelerating voltage of 20 kV, using secondary and backscattered electrons.
The pH slurry of the carbons was determined measuring the pH (pH-meter, Crison) of an aqueous suspension of the sample (1 g) in distilled water (10 mL) [49].

Adsorption Tests
The potential application of the activated carbons as adsorbents in aqueous phase was evaluated using sulfamethoxazole (SMX) as model impurity, which has a solubility in water of 610 mg·L −1 at 298 K. Samples of AC (12.5 mg, ≈100 µm) were contacted in stoppered glass bottles with 50 mL of SMX aqueous solutions (25 to 150 mg·L −1 ). A commercial activated carbon (C: 89.5 wt.%, A BET : 800 m 2 ·g −1 ; Vmicro: 0.67 cm 3 ·g −1 ; Vmeso: 0.53 cm 3 ·g −1 ), supplied by Merck, was also used as reference for comparison purposes. Experiments were carried out at 20 • C and the natural pH of the SMX solution (4.6) in a thermostatized shaker (Optic Ivymen System, Biotech, Madrid, Spain) at equivalent 200 rpm. A kinetic experiment showed that 5 days was time enough to reach the adsorption equilibrium in the operating conditions. SMX concentration was determined by UV-vis spectrophotometry (Cary 60 UV-Vis, Agilent Technologies, Santa Clara, CA, USA) at 265 nm wavelength. The Langmuir equation was used to fit the equilibrium data. The reported results are the average values from triplicate runs, being the standard errors always below 5%. Table 1 shows the hydrochar yields and the proximate and ultimate analyses of the raw GSs and hydrochars. Hydrochar yields were almost constant ( 60-65 wt.%) in the HTC runs performed with a GS relative amount higher than 20% in the GS + water mixture within the temperature range of 220-280 • C (aprox.). At a lower GS percentage the HC yield decreased since a higher relative amount of water favors the extraction and transfer of solid components to the liquid phase [12,50]. The effect of the temperature in the HTC process can be seen at fixed GS percentage in the starting mixture (40 wt.%) and temperatures below and above the aforementioned range. At temperatures ≤200 • C, high HC yields were obtained associated to lower carbonization as can be seen from the C content of the samples. At temperature above 280 • C gasification reactions gain significance, giving rise to lower yield but a more carbonized solid [51]. It is well known that dehydration and decarboxylation reactions are the main responsible of the mass loss upon HTC [52,53].

Characterization of the Hydrochars
A van Krevelen diagram was plotted from the ultimate analyses ( Figure 1) where the points corresponding to the HCs have been placed. As the HTC temperature increases, the resulting HCs evolve from a typical peat composition towards the lignite region, approaching even sub-bituminous coal at the highest HTC temperature tested (300 • C). A van Krevelen diagram was plotted from the ultimate analyses ( Figure 1) where the points corresponding to the HCs have been placed. As the HTC temperature increases, the resulting HCs evolve from a typical peat composition towards the lignite region, approaching even sub-bituminous coal at the highest HTC temperature tested (300 °C).  Table 2 summarizes the energy properties of hydrochars, where, clearly, there was a significant increase of the higher heating value as the HTC temperature increased. The energy density of the most carbonized solid was improved by one-third with respect to the starting precursor. Taking into account the GS to HC yield, the highest energy recovery efficiency does not correspond to the highest HTC temperature tested but to a significantly lower value somewhere around 200 °C. However, since the energy density of the resulting fuel is also an important issue, the information provided by Table  2 points to an optimum HTC temperature in the range of 260-280 °C. Moreover, this determination, also known as a BUEE, provides an extra information because it allows a comparison between different fuels and options for biomass utilization [43].   Table 2 summarizes the energy properties of hydrochars, where, clearly, there was a significant increase of the higher heating value as the HTC temperature increased. The energy density of the most carbonized solid was improved by one-third with respect to the starting precursor. Taking into account the GS to HC yield, the highest energy recovery efficiency does not correspond to the highest HTC temperature tested but to a significantly lower value somewhere around 200 • C. However, since the energy density of the resulting fuel is also an important issue, the information provided by Table 2 points to an optimum HTC temperature in the range of 260-280 • C. Moreover, this determination, also known as a BUE E , provides an extra information because it allows a comparison between different fuels and options for biomass utilization [43]. The 77 K N 2 adsorption isotherms of the hydrochars (supporting material) corresponded to Type II of the IUPAC classification, associated to non-porous solids. The highest BET surface area was only 8 m 2 ·g −1 , corresponding to the GS-220-40 sample. To learn more on the porous texture of these hydrochars the CO 2 adsorption isotherms at 273 K were obtained (Supplementary Material). The surface area values from those isotherms (S DA ) varied from 100 to 845 m 2 ·g −1 , the highest corresponding to the above mentioned sample, with a micropore volume of 0.513 cm 3 ·g −1 . These huge differences between the BET and S DA values are indicative of microporous texture consisting essentially of very narrow micropores, the so-called ultramicropores (≤0.5 nm width), corresponding most probably to slit-shaped pores. This type of microporosity has been also observed in chars from HTC of different sugars at 180 • C calcined at high temperatures in an inert atmosphere [23].

Activated Carbons by Chemical Activation of Hydrochars
Previous works have proven that hydrochars can be more effective adsorbents than thermal (pyrolytic) biochars due to their diverse structures and surface functional groups [54]. Looking at the characteristics of the hydrochars, the GS-220-40 was selected for further activation to achieve a well-developed porosity typical of a carbon-based adsorbent. In previous works [23] HTCs obtained at 180-240 • C were more prone to yield a better developed porosity upon KOH-activation than those obtained above 260 • C, although from fairly different precursors than in the current work. Table 3 summarizes the porous textures of the activated carbons. Representative N 2 adsorption-desorption isotherms are depicted in Figure 2. All of them correspond to essentially microporous solids; there was a higher contribution of mesoporosity in the case of H 3 PO 4 -activation. The rest of isotherms are given as Supplementary Material (Figures S1-S3). KOH was by far the activating agent allowing the highest porosity development, with BET surface area values up to ca. 2200 m 2 ·g −1 . These carbons were essentially microporous with increased mesoporosity at increasing KOH to hydrochar mass ratio. The activated carbon prepared at 4:1 mass ratio had almost 1800 m 2 ·g −1 BET surface area in spite of its high ash content (38.2%), mostly due to the remaining inorganic matter derived from the activating agent. In fact, its BET surface area represented as much as ca. 2900 m 2 ·g −1 on an ash-free basis.
The lowest surface area development occurred with FeCl 3 as activating agent, due in part to the very high ash content of the resulting activated carbons. The highest BET surface area of these carbons on an ash-free basis was 770 m 2 ·g −1 , only slightly more than one-quarter that of the abovementioned KOH-activated carbon. FeCl 3 -activation can be addressed to the preparation of carbon-supported Fe catalysts. Bedia et al. [44] and Mena et al. [55] prepared this type of catalyst, and tested theirs in catalytic wet peroxide oxidation (CWPO), from FeCl 3 -activation of sewage sludge.
Activation with H 3 PO 4 gave rise to microporous carbons but with much higher relative contribution of mesoporosity, showing Type IV isotherms of the IUPAC classification. The BET surface area values are relatively low compared to the obtained by the same method with other precursors. The ash content was significantly lower than by the two other activation procedures. Increasing the H 3 PO 4 to HC ratio decreased the BET surface area, affecting mostly, the mesopore contribution.   Figure 3 shows SEM images of the grape seeds, hydrochars and activated carbons. The raw GSs exhibited a well-defined morphological structure in three different layers, while in the hydrochar image shows some partial degradation of the structure occurred, although the morphology was still preserved in great part, which indicates good thermomechanical stability for the GSs. This allows obtaining granular activating carbons from this starting material. KOH activation produced a spongy-like structure with many holes. In the case of FeCl3-activation, the formation of microspheres was visible; after the activation with H3PO4, prismatic particles can be seen.  Figure 3 shows SEM images of the grape seeds, hydrochars and activated carbons. The raw GSs exhibited a well-defined morphological structure in three different layers, while in the hydrochar image shows some partial degradation of the structure occurred, although the morphology was still preserved in great part, which indicates good thermomechanical stability for the GSs. This allows obtaining granular activating carbons from this starting material. KOH activation produced a spongy-like structure with many holes. In the case of FeCl 3 -activation, the formation of microspheres was visible; after the activation with H 3 PO 4 , prismatic particles can be seen.  Figure 4 shows the FTIR spectra of the raw GS, GS-220-40 hydrochar and the activated carbons obtained with an activating agent to hydrochar ratio of 3. A decreasing intensity of the O-H stretching of the hydroxyl groups at 3400 cm −1 was observed for the hydrochar, because the dehydration occurred upon HTC [56,57]. The bands appearing in the 2950 to 2750 cm −1 range, associated to aliphatic carbon −CHx stretching vibration, tended to be less intense for the GS-220-40 hydrochar, probably due to the evolution of nonpolar alkyl carbon structure [58]. In addition, transmission intensities at 1750-750 cm −1 arising from C=O stretching of ketones and other carbonyl structures were reduced. Furthermore, the decline of the peak at 1030 cm −1 (the C-O and C-O-C stretching) indicates that both decarboxylation and dehydration occur, consistently with the van Krevelen representation of Figure 1.
The FTIR spectra of activated carbons showed a dramatic decrease of functionalities in agreement with the reported in general in the literature relative to activation of carbon precursors. A broad band (3600-3100 cm −1 ) appears in the three activated carbons, related to the existence of hydroxyl functional groups. Those related to C=C and aromatic rings vibration (1600-1500 cm −1 ) and the associated to C-O bonds at 1157-1090 cm −1 , related to the presence of alcohol groups were more intense in the GSHC-FeCl3-3, and, mainly, in the GS-H3PO4-3 activated carbons [59,60].
The amount of surface oxygen groups in the activated carbons was assessed in terms of CO2 and CO evolved upon TPD. The results are summarized in the inside Table of Figure 4. In all the cases, the ratio CO2/CO was within 27%-30%, suggesting that the oxygen groups must be ascribed majorly to CO-evolving functionalities (phenol, carbonyl, anhydride ether, quinone), mainly associated to weak acid and neutral/basic surfaces. However, there are also some significant amounts of evolved CO2, such that carboxylic, lactone and anhydride groups may be also present in some extent [61,62]. H3PO4, and in particular, FeCl3, appear to be more prone than KOH to create oxygen groups on the carbon surface upon activation. Rey et al. [62] arrived to similar conclusion working with Fe(NO3)3 as a precursor.   Figure 4 shows the FTIR spectra of the raw GS, GS-220-40 hydrochar and the activated carbons obtained with an activating agent to hydrochar ratio of 3. A decreasing intensity of the O-H stretching of the hydroxyl groups at 3400 cm −1 was observed for the hydrochar, because the dehydration occurred upon HTC [56,57]. The bands appearing in the 2950 to 2750 cm −1 range, associated to aliphatic carbon −CH x stretching vibration, tended to be less intense for the GS-220-40 hydrochar, probably due to the evolution of nonpolar alkyl carbon structure [58]. In addition, transmission intensities at 1750-750 cm −1 arising from C=O stretching of ketones and other carbonyl structures were reduced. Furthermore, the decline of the peak at 1030 cm −1 (the C-O and C-O-C stretching) indicates that both decarboxylation and dehydration occur, consistently with the van Krevelen representation of Figure 1.
The FTIR spectra of activated carbons showed a dramatic decrease of functionalities in agreement with the reported in general in the literature relative to activation of carbon precursors. A broad band (3600-3100 cm −1 ) appears in the three activated carbons, related to the existence of hydroxyl functional groups. Those related to C=C and aromatic rings vibration (1600-1500 cm −1 ) and the associated to C-O bonds at 1157-1090 cm −1 , related to the presence of alcohol groups were more intense in the GSHC-FeCl 3 -3, and, mainly, in the GS-H 3 PO 4 -3 activated carbons [59,60].
The amount of surface oxygen groups in the activated carbons was assessed in terms of CO 2 and CO evolved upon TPD. The results are summarized in the inside Table of Figure 4. In all the cases, the ratio CO 2 /CO was within 27-30%, suggesting that the oxygen groups must be ascribed majorly to CO-evolving functionalities (phenol, carbonyl, anhydride ether, quinone), mainly associated to weak acid and neutral/basic surfaces. However, there are also some significant amounts of evolved CO 2 , such that carboxylic, lactone and anhydride groups may be also present in some extent [61,62]. H 3 PO 4 , and in particular, FeCl 3 , appear to be more prone than KOH to create oxygen groups on the carbon surface upon activation. Rey et al. [62] arrived to similar conclusion working with Fe(NO 3 ) 3 as a precursor.  Figure 5 shows the adsorption isotherms (20 °C) of SMX with the GSAC-KOH-3, GSAC-FeCl3-3 and GSAC-H3PO4-3 activated carbons and a commercial active carbon. The adsorption capacity seems to be mainly determined by the BET surface area of the carbons rather than the amount and character of the surface groups ( Figure 4). Consistently with its much higher surface area, GSHC-KOH-3 yielded the highest adsorption capacity. The only deviation of that behavior occurred with the GSHC-H3PO4-3 activated carbon, which showed lower adsorption than the expected from its BET surface area. This can be explained in terms of SMX ionization and the pH slurry of this carbon. At solution pH 4.6, SMX (pKa1: 1.7, pKa2: 5.6 [63]) must be majorly made of neutral and anion species; since the pH slurry of GSHC-H3PO4-3 is 2.2, its surface will be negatively charged, which induces repulsion forces between the molecules of SMX and the carbon surface [64,65].

Adsorption of Sulfamethoxazole
The experimental data were fitted to the Langmuir equation:  Table. 3.3. Adsorption of Sulfamethoxazole Figure 5 shows the adsorption isotherms (20 • C) of SMX with the GSAC-KOH-3, GSAC-FeCl 3 -3 and GSAC-H 3 PO 4 -3 activated carbons and a commercial active carbon. The adsorption capacity seems to be mainly determined by the BET surface area of the carbons rather than the amount and character of the surface groups ( Figure 4). Consistently with its much higher surface area, GSHC-KOH-3 yielded the highest adsorption capacity. The only deviation of that behavior occurred with the GSHC-H 3 PO 4 -3 activated carbon, which showed lower adsorption than the expected from its BET surface area. This can be explained in terms of SMX ionization and the pH slurry of this carbon. At solution pH 4.6, SMX (pKa1: 1.7, pKa2: 5.6 [63]) must be majorly made of neutral and anion species; since the pH slurry of GSHC-H 3 PO 4 -3 is 2.2, its surface will be negatively charged, which induces repulsion forces between the molecules of SMX and the carbon surface [64,65].
The experimental data were fitted to the Langmuir equation: where q e is the equilibrium adsorbate loading onto the adsorbent (mg·g −1 ), C e the equilibrium liquid-phase concentration of the adsorbate (mg·L −1 ), q L the monolayer saturation capacity (mg·g −1 ) and K L the Langmuir constant (L·mg −1 ). This equation described the equilibrium data well, as can be seen in Figure 5. The values of the fitting parameters and the correlation coefficients are given in Table 4. The calculated monolayer Langmuir capacity of GSHC-KOH-3 reached 650 mg·g −1 (notice that in Figure 5 the represented experimental SMX values did not reach the saturation level). The SMX adsorption capacity of this material is higher than the exhibited by other adsorbents, such as activated carbon from pine tree or coconut shell, coal, carbon nanotubes, graphene and organo-montmorillonites at similar operating conditions, which are characterized by maximum adsorption capacities within the range of 122-242 mg·g −1 [65][66][67].
Appl. Sci. 2019, 9, x 10 of 14 where qe is the equilibrium adsorbate loading onto the adsorbent (mg g −1 ), Ce the equilibrium liquidphase concentration of the adsorbate (mg L −1 ), qL the monolayer saturation capacity (mg g −1 ) and KL the Langmuir constant (L mg −1 ). This equation described the equilibrium data well, as can be seen in Figure 5. The values of the fitting parameters and the correlation coefficients are given in Table 4. The calculated monolayer Langmuir capacity of GSHC-KOH-3 reached 650 mg g −1 (notice that in Figure  5 the represented experimental SMX values did not reach the saturation level). The SMX adsorption capacity of this material is higher than the exhibited by other adsorbents, such as activated carbon from pine tree or coconut shell, coal, carbon nanotubes, graphene and organo-montmorillonites at similar operating conditions, which are characterized by maximum adsorption capacities within the range of 122-242 mg g −1 [65][66][67].

Conclusions
Hydrothermal carbonization of grape seeds is a promising way of valorization of that biomass waste from the wine industry. It allows obtaining chars of fairly good properties to be used as improved solid fuel. Further activation of those chars leads to higher value-added products, activated carbons. Activation with FeCl3, H3PO4 and KOH led to high BET surface area carbons of predominantly microporous texture, which can be useful adsorbents for liquid-phase applications, as in water pollutant removal. In fact, a fairly high adsorption capacity was experimentally found with sulfamethoxazole, an emerging contaminant, as the target compound.

Conclusions
Hydrothermal carbonization of grape seeds is a promising way of valorization of that biomass waste from the wine industry. It allows obtaining chars of fairly good properties to be used as improved solid fuel. Further activation of those chars leads to higher value-added products, activated carbons. Activation with FeCl 3 , H 3 PO 4 and KOH led to high BET surface area carbons of predominantly microporous texture, which can be useful adsorbents for liquid-phase applications, as in water pollutant removal. In fact, a fairly high adsorption capacity was experimentally found with sulfamethoxazole, an emerging contaminant, as the target compound.

Conflicts of Interest:
The authors declare no conflict of interest.