Effects of process parameters on the physicochemical properties of corn stalk hydrochar and the removal of alkali and alkaline earth metals

Funding information the General Programs of the National Natural Science Foundation of China, Grant/Award Number: 51676058; the Science and Technology fund of Liaoning Province Education Department, Grant/Award Number: LQN201705 Abstract The advantages of the direct combustion of agricultural biomass for power generation are restricted due to the drawbacks of this biomass and the deep peak shaving trend. Hydrothermal carbonization can improve the physical and thermochemical properties of raw materials. Corn stalk samples were selected as the research object in this study. Corn stalk hydrochars were prepared with reaction pressures of 0–3 MPa, reaction temperatures of 160–240 C, and residence times of 1–10 h. The effects of the process parameters on the physicochemical properties of hydrochar and on the removal of alkali and alkaline earth metals (AAEMs) were studied by measuring the hydrochar mass, energy yield, product composition, morphology, as well as the removal efficiency of AAEMs. The results indicated that the corn stalk hydrochar had higher mass and energy yields than the raw corn stalk. X-ray diffraction showed that the carbon content in the hydrochar was greater than that in raw corn stalk. Scanning electron microscopy showed that the reaction temperature played a more significant role in the hydrothermal carbonization process than the reaction pressure and residence time.


INTRODUCTION
Biomass has been widely recognized as a sustainable energy source and as one of the most potentially useful alternative energy sources [1,2]. Agricultural biomass is an abundant resource that accounts for 51% of biomass energy. Combusting agricultural biomass for power generation has become a feasible large-scale approach [3]. However, the bulk density and energy density of agricultural biomass are very low, but the demand for agricultural biomass for power generation is huge, resulting in a large collection radius. These factors significantly increase the transport and storage costs. Moreover, the high contents of alkali and alkaline earth metals (AAEMs) in the biomass cause the biomass ash to have a lower ash fusion point, which leads to severe ash deposition, slagging, and fouling issues on the heat surface of a boiler. As a result, the safe operation of the generating unit is at risk [4,5]. In addition, with the development of sustainable energy and grid-connected power generation, the generating unit needs to sustain deep peak shaving [6]. This may cause erratic combustion or even boiler flameout [7]. Currently, the combustion in biomass-fired boilers is usually stabilized through coal blending. However, instability in the collection and purchase markets has caused power plants to have high coal preparation requirements. In fact, this practice greatly reduces the advantages of agricultural biomass combustion power generation. Among many thermal pretreatment technologies, hydrothermal carbonization (HTC) has attracted great attention because it can directly process wet biomass and remove AAEMs [8]. HTC is a thermochemical conversion process using water as the reaction medium. The reaction temperature is usually in the range of 150-350 • C, and the reaction pressure is usually autogenous [9]. Through HTC technology, not only are the bulk density and energy density of the biomass greatly improved, but most water-soluble AAEMs in the raw biomass can be removed [10]. The solid residues of HTC are called hydrochars. Biomass hydrochar is a potential solid fuel that has a heating value almost equivalent to that of lignite. It is also a carbon-based material with rich oxygen-containing functional groups and high added value [11].
Corn stalk (CS) accounts for a large proportion of agricultural waste [23] and has broad application prospects, so it has gradually attracted the attention of researchers [24][25][26]. It is meaningful that the reasonable utilization of CS can solve the aforementioned problems of low energy content and high AAEMs to a certain extent. At present, there are few studies on the HTC of CS [27]. Especially the influence of HTC on the AAEM contents and on the hydrochar ash characteristics needs to be systematically conducted further.
In this study, typical CS from the Northeast District of China was investigated as the research subject. The changes in the composition and structure of CS hydrochar were analysed under a wider range of reaction pressure values, a longer residence time, and different temperatures compared to previous studies. The migration of AAEMs, which is vital to biomass combustion, was also observed under different conditions. The relevant properties of the CS hydrochar samples were compared with those of the raw CS. This study also supplied a series of basic data for the comprehensive utilization of biomass.

Material preparation
The CS for the current study was collected from a rural area in Liaoning Province, China. The raw CS was ground and sieved into different particle size ranges. The powder with a particle sizes ranges of 38-58 μm was selected, which was dried at 105 • C to a constant weight, and sealed in plastic bottles until further experiments were conducted.

Hydrothermal carbonization reaction process
The HTC experiment was carried out in a batch reactor made of 316L stainless steel. The maximum working pressure was 20 MPa and the working temperature was 300 • C. The reactor was electrically heated, and the heating rate was set to 5 • C/min. 10 g of the CS powder was dissolved into 200 mL of deionized water. The mixture was stirred for 12 h to fully swell and was then transferred into the reactor.
The reactor could be water-cooled to ambient temperature approximately 10 min after the reaction time, which was 1-10 h. The residues were filtered and washed repeatedly with deionized water. The solid products were dried to a constant weight at 105 • C. These hydrochar samples were marked as follows: the reaction temperature-reaction pressure-residence time Char. For example, 180-2-2 Char indicates corn stalk char obtained at 180 • C under a pressure of 2 MPa for 2 h.
The optimum operating conditions of HTC of CS as a coallike fuel were focused on this paper. It will acquire the hydrochar of CS which has good combustion and ash behaviours in favour of co-firing with lignite. The pressure values of 0-3 MPa were adopted to evaluate hydrochars derived from unsaturated water or saturated vapours. The initial pressure of 2 MPa was considered for use because the reactor was filled with unsaturated water since 200 • C. The residence time values of 1-10 h covered a broader range. Effect of the residence time on the characteristics of hydrochar of CS was investigated aiming to proper reaction time for above hydrochar. The HTC process parameters are as listed in Table 1.

Analysis methods
The proximate analysis results of the CS are shown in Table 2.
The contents of cellulose, hemicellulose, and lignin of CS were determined by the Van Soest fibre abstersion method, which mainly involved determining of the neutral detergent fibre, acid detergent fibre, acid detergent lignin, and acid insoluble ash contents [28]. The measured contents of the components (dry basis) were 42.16% cellulose, 29.29% hemicellulose, and 5.24% lignin. Compared with rice husk and coconut husk, the CS had a lower lignin content and was easier to carbonize. The C, H, N, and S element compositions of the raw CS and hydrochars samples were analysed with a Vario EL Ш element analyser. The heating value was measured with an IKA C200 oxygen-bomb calorimeter. The crystal structure was characterized by a D/MAX-2500 PC X-ray diffractometer using Cu radiation in the 2θ range of 10-80 • . After the samples were digested, the K, Ca, Na, and Mg contents in the samples were determined by using an Optima 7000DV, which is an inductively coupled plasma optical emission spectrometer (ICP-OES). The microscopic morphologies of the raw CS and hydrochar samples were characterized using a Zeiss Supra 55 scanning electron microscope (SEM).
The mass yield of the hydrochar was calculated using Equation (1): Hydrochar mass yield = Hydrochar mass ( dry basis ) (1) The energy yield of the hydrochar was calculated using Equations (2) and (3): Higher heating value (HHV) of hydrochar ( dry basis ) and Energy yield = Hydrochar mass yield × Energy density (3) The AAEM contents in the hydrochar changed upon HTC. The AAEM removal efficiency was calculated using Equation (4):

Mass content of AAEMs in hydrochar × Mass of hydrochar
Mass content of AAEMs in raw CS × 100% (4)

Hydrochar mass yield and energy content
The reaction pressure, reaction temperature, and residence time of the HTC are important factors influencing the thermal stability of the hydrochar. The changes in the mass yield, as well as the energy yield of the CS hydrochar under different process parameters were obtained.

Mass and energy yields of CS hydrochar at different reaction pressures
The effect of pressure on the mass and energy yields of the hydrochar during the HTC of the CS are described in Figure 1. The mass yields of hydrochar were 46.20%, 44.88%, 44.46%, and 26.85% at 0-3 MPa, respectively. The mass yield of the hydrochar decreased by 19.35% as the pressure increased from 0 to 3 MPa. The mass yield of hydrochar did not change significantly until the pressure increased more than 2 MPa. Overall, the mass yield of the hydrochar declined as the pressure in the reactor increased.
It is worth noting that the initial pressure of 2 MPa was the change point for the state of deionized water in the reactor when the reaction temperature was 200 • C. The actual pressure reached 1.3-1.5 MPa from the initial 0 MPa in the reactor at the temperature of 200 • C. At this pressure and temperature, the water in the reactor was in the form of wet saturated steam. When the initial pressure was 2 MPa, the actual pressure in the reactor was 3.9-4 MPa at the temperature of 200 • C, and the liquid in the reactor was unsaturated water. The mass yield of the hydrochar produced by unsaturated water was lower than that produced by wet saturated steam. It was lower because the reaction rate increased as a result of the higher heat transfer of unsaturated water [29,30].
The energy yield is commonly used to evaluate the energy retention of biomass hydrochar. The maximum energy yield for biomass could be used to indicate the optimal HTC conditions for the purpose of alternative fuel use. As the reaction pressure increased from 0 to 3 MPa, the energy yields first increased and then decreased in Figure 1. The maximum value (94.05%) appears at 1 MPa. In fact, the water shows the transitional behaviour from saturated vapour to unsaturated water under that pressure. The energy yield was reduced to a certain extent when the pressure exceeded 2 MPa.
The results in terms of water behaviour under different pressures could be further supported. As the reaction pressure increased, the ionic reaction was enhanced, the dielectric constant of water decreased, and the chemical reaction rate increased [31,32]. The higher ionization could accelerate the hydrolysis of the CS, while a lower dielectric constant might dissolve the organic compound [1].

Mass and energy yields of CS hydrochar at different reaction temperatures
As shown in Figure 2, the mass and energy yields of CS hydrochar were investigated at different reaction temperatures (180, 200, 220 and 240 • C) with 2 MPa reaction pressure and a 2 h residence time.
The mass yields of hydrochar at 180, 200, 220 and 240 • C were 45.26%, 44.46%, 28.47% and 24.22%, respectively. During the HTC process, the mass yield of hydrochar decreased gradually as the reaction temperature increased. Between 180 and 200 • C, the decrease in the mass yield of the hydrochar was not obvious. The mass yield of the hydrochar decreased steadily after the temperature was increased to 200 • C. As the temperature increased from 200 to 240 • C, the mass yield decreased from 44.46% to 24.22%, which was a total decline of 20.23%. The reason was that, at 180 • C, only hemicellulose began to decompose, but when the temperature exceeded 200 • C, both cellulose and lignin began to hydrolyse and decompose. As a result, the mass yield showed a trend of rapid decline above 200 • C [9].
Because of the reduction of the mass yield and the high heating value (HHV), the energy yield declined from 81.35% at 180 • C to its minimum value of 49.74% at 220 • C. A larger portion of cellulose and lignin degraded in the reaction at higher temperatures [33]. Then a slight increase (5.68%) of the energy yield appeared from 220 to 240 • C. It was attributed to the greater extent of secondary hydrochar formation by hydrolysed microspheres when the temperature was increased. And higher amounts of secondary char often obtained higher HHV [15,34]. The change rate in the mass and energy yields slowed down after the temperature exceeded 200 • C. The density functional theory (DFT) could also be employed to explain the above change at the microscopic level. The energy barrier could be overcome easily with a temperature increased and the sufficient activation energy could be supplied for biomass breakup to achieve higher concentrations of free radicals by increasing the temperature. The increase in free radicals leads to accelerated free radical reactivities [35]. During the HTC process, as the residence time increased from 1 to 6 h, the mass yield of the hydrochar first increased and then decreased. The mass yield decreased from 42.99% to 40.22%. The maximum mass yield (44.46%) was observed at the residence time of 2 h. These results show that a long residence time led to a lower yield of hydrochar, though the influence of the residence time on the mass yield was not significant. The hydrolysis and dehydration reactions occurred during HTC. Therefore, liquid and volatile gas products were produced along with the hydrochar, resulting in a decreased mass yield of hydrochar. Lei et al. [27] reported the HTC of CS at 200 • C and ambient pressure (initial pressure). When the reaction time exceeded 26 h, the reaction time had no significant effect on the yield of the hydrochar.

Mass and energy yields of CS hydrochar at different residence times
The energy yields first increased and then decreased with time. The maximum value of 81.92% was observed at the residence time of 2 h. These results showed that most of the energy was retained in the hydrochar after the HTC of the CS. Although the extension of the residence time resulted in dehydration and decarboxylation of the CS components, it did not cause a significant reduction in energy. Basically, the reaction pressure and reaction temperature had significant effects on the mass and energy yields of the CS hydrochar. The reaction pressure affected the subcritical state of the reaction water. The effect of the residence time was not significant. The maximum mass yield and energy yield were obtained under the HTC conditions of 2 MPa, 200 • C, and 2 h.

Ultimate analysis and higher heating value
The ultimate analysis, H/C and O/C atomic ration and HHV are described in Table. 3. The contents of C, H, N, and S were obtained by ultimate analysis for the CS raw materials and CS hydrochar samples. The O content was obtained by calculation. The H/C and O/C atomic ratios under different hydrothermal conditions were also calculated.
The van Krevelen diagram is often used to illustrate the carbonization degree of lignocellulosic biomass. The relationship between the H/C and O/C atomic ratios of the hydrochar samples obtained under different operating conditions is shown in Figure 4.
As the temperature increased from 180 to 240 • C, the H/C ratio decreased from 1.34 to 0.93, which was a decrease of 30.6%. The O/C ratio decreased from 0.6 to 0.22, which was a decrease of 63.3%. The H/C and O/C ratios also decreased with the residence time, but the changes were not significant. It is worth mentioning that as the pressure increased from 2 to 3 MPa, the H/C and O/C ratios rapidly decreased from 1.30  [39].
The higher the reaction pressure and temperature, the closer the position of the hydrochar was to the position of Zhundong lignite in the van Krevelen diagram. This was due to the dehydration and decarboxylation reactions in the HTC process. The contents of H and O decreased, the aromatization degree of hydrochar became higher, the biomass components were converted into carbon-containing products, and the coalification degree of the biomass increased. These outcomes were also consistent with the conclusions in the literature [30].
Through the analysis of the C, H, and O contents and the heating value of the hydrochar samples obtained under different operating conditions, it can be concluded that increasing the reaction temperature and pressure can enhance the coalification degree of the biomass. The H/C and O/C ratios and the heating value of the hydrochar obtained at 2 MPa, 2 h, and 240 • C were the closest to those of lignite. Compared with raw CS, solid hydrochar after HTC is more suitable as a solid fuel.

X-ray diffraction analysis
The composition of the CS hydrochar was characterized by X-ray diffraction analysis (XRD  Figure 5. Figure 5(a) shows the XRD spectra of the CS hydrochar samples obtained under different reaction pressures. When the reaction pressure was 0-2 MPa, the cellulose content in the hydrochar increased, and the graphitic carbon content did not change significantly compared with those in the raw CS. Under this reaction pressure, the amorphous component began to disappear, which resulted in enhanced diffraction peaks of cellulose. And the content of cellulose decreased significantly while the graphitic carbon content increased significantly when the pressure reached 3 MPa. This increase was attributed to the destruction of the crystal structure of cellulose and the formation of amorphous carbon [36]. Figure 5(b) shows the intensity changes in the typical peaks of raw CS in the range of 160-240 • C. The cellulose content in the hydrochar remained basically unchanged between 160 and 200 • C. When the temperature reached 220 • C, the cellulose content decreased rapidly until it disappeared. These results indicate that the microcrystalline structure of the CS hydrochar began to decompose rapidly at 220 • C. The diffraction peak of graphitic carbon became stronger, indicating a higher carbon content. Figure 5(c) shows the changes in the hydrochar composition from the residence time of 1-10 h at 200 • C and 2 MPa. After 1-10 h of the hydrothermal process, the diffraction peaks of the hydrochar were significantly enhanced compared with those of the raw CS. As the residence time of HTC lengthened, the composition of hydrochar did not change significantly. The residence time had no significant effect on the decomposition of lignocellulose.
The XRD patterns under different operating conditions indicated that, when the reaction pressure was increased from 2 to 3 MPa, the peak intensity of cellulose weakened, and the cellulose was greatly decomposed. When the reaction temperature reached 220 • C, the characteristic peaks of cellulose weakened, and the cellulose content decreased. As the temperature increased gradually to 240 • C, the cellulose content no longer changed significantly. Compared to the peak intensity of cellulose under different reaction pressures and temperatures, it was found that 3 MPa and 220 • C had equivalent impacts on the cellulose decomposition. The effect of the residence time on cellulose decomposition can be neglected.

SEM morphology and structure
The morphologies of the biomass and the hydrochar samples were characterized by SEM. The microscopic morphology and structural characteristics of the CS, and well as the hydrochar samples prepared under different reaction conditions are exhibited in the figures below. Figure 6 shows the SEM images (50 kV) of the CS hydrochars obtained under different reaction pressures. The reaction temperature was 200 • C, the residence time was 2 h, and the reaction pressures were 0-3 MPa. Flocculent structures gradually appeared on the surface of the CS hydrochar samples. At the initial ambient pressure, few microspheres begun to form. After the initial reaction pressure was 3 MPa, neatly arranged pores formed on the surface of the hydrochar, indicating that the structure of the CS had changed after the hydrothermal process. Figure 7 shows the SEM images of the raw CS and the CS hydrochar samples prepared at different temperatures. The surface of the raw CS exhibited a sheet-like fibre texture. When the temperature was increased to 200 • C, flocculent  structures and microspheres began to form on the surface of the hydrochar. When the temperature reached to 220 • C, a complex cross-linked network structure was formed. This was because the insoluble components of the cellulose and lignin in the raw CS were destroyed and the interface between adjacent pores was damaged at high temperatures, resulting in larger pore sizes. When the temperature reached 240 • C, the crosslinked network structure was clearer. The porous structure on the hydrochar surface allowed easier dehydration of hydrochar, which promoted the rapid drying of the hydrochar. Figure 8 shows the SEM images of the hydrochar samples obtained at residence times of 1-10 h. When the residence time was 1 h, the edge of the flake structure in the raw CS began to turn round, and flocculent structures appeared on the surface. When the residence time was 2 h, carbon microspheres began to form, but the shape of the microspheres was irregular. When the residence time was increased to 4 h, the number and average diameter of the carbon microspheres gradually increased. When the reaction time was increased to 6 h, more carbon microspheres were formed. The microspheres adhered to each other, and some even penetrated each other. The microspheres formed mainly because of the degradation of hemicellulose and the partial hydrolysis of cellulose and lignin. As the residence time increased, the hydrolysis products recondensed extensively, and the surrounding oxygen-containing functional groups formed ethers and quinones. As a result, the size of the carbon microspheres increased [13].
The three sets of SEM images reveal that hemicellulose was hydrolysed to form furfural at 200 • C. Some glycosidic bonds in the amorphous cellulose began to break, forming small molecules that dissolved in water and producing hydroxymethyl furfural. The insoluble components of cellulose experienced a process similar to pyrolysis, including molecular reorganization, dehydration, and decarboxylation, and formed a crosslinked porous network structure. Part of the lignin began to break down and underwent a homogeneous hydrolysis reaction, decomposing into phenolic derivatives. At the same time, the insoluble lignin underwent reactions similar to those of the insoluble components of cellulose to form polyaromatic hydrochar [27]. The results also verify the above XRD analysis on the diffraction peaks of cellulose and graphitic carbon.

AAEM content and removal efficiency
The major constituents of AAEMs involve calcium (Ca), potassium (K), magnesium (Mg) and sodium (Na). The AAEM contents of the raw CS and resultant hydrochar samples which obtained under different operating conditions were illustrated in Figure 9. The AAEM contents was reduced significantly after the HTC for all the cases. The remaining content of AAEMs in the hydrochar of the CS decreased first and then increased when the temperature increase from 180 to 240 • C. It was verified that a porous  solid left after all the hemicelluloses and much of the cellulose has reacted well at 240 • C. Porous structure might absorb some AAEMs, which might explain the increase of AAEMs [37,40,41]. The remaining content of Ca in the hydrochar samples decreased sharply from 23.07% to 4.65% with the increase of pressure. The remaining content of K and Mg had a slight change with increase of operating conditions. It should be noted that the Na content of hydrochar was very low, only 0.066 mg/g. Thus, the change in Na content after HTC could not be used for reference because it could be easily disguised by errors. In addition, the influence of such a low content on the AAEM analysis was minimal.
The removal efficiency of AAEMs were listed in Table 4. The removal efficiency of K and Mg were promoted by increasing above process parameters. The 97-99% of K and Mg in the hydrochar of the CS can be removed after HTC. The removal efficiency of Ca also increased with the reaction temperature. The removal efficiency of Ca was overall lower than that of K and Mg. These results are probably attributed to the reason that potassium salts in ash are water-soluble while calcium salts are insoluble during hydrothermal process [33,42].

CONCLUSIONS
This work studied the changes in the composition and structure of CS hydrochar under a wider range of reaction pressure values, a longer residence time, and under different temperatures than in past studies. The migration of AAEMs, which is essential for biomass combustion, was also studied. The reaction pressure and temperature had significant effects on the mass and energy yields, while the residence time had little effect. The maximum mass yield (44.46%) and energy yield (81.92%) were obtained under the HTC conditions of 2 MPa, 200 • C, and 2 h.
As the reaction pressure and temperature increased, the H/C and O/C ratios decreased. The HHV of hydrochar was greatly increased. The H/C and O/C ratios and the heating value of the hydrochar obtained at 240 • C were closer to those of lignite.
The XRD patterns showed that as the reaction pressure increased from 2 to 3 MPa, the cellulose was greatly decomposed. When the reaction temperature reached 220 • C, the cellulose content was reduced. It was found that 3 MPa and 220 • C had equivalent impacts on the cellulose decomposition.
The SEM results showed that as the reaction temperature reached 200 • C, the hydrothermal process caused flocculation on the surface of the raw CS, producing various numbers of carbon microspheres and pores. The number of carbon microspheres and the crosslinking degree of pores were determined by different reaction conditions.
The AAEMs in the CS were removed to a great degree by HTC. The removal efficiency of Ca increased together with the temperature. The overall removal efficiency of Ca was lower than that of K.
In conclusion, the mass and energy yields, HHV, proximate and ultimate analyses, crystal composition, microscopic morphology, and AAEM removal efficiency of the CS hydrochar were all greatly improved after the HTC treatment. The hydrochar prepared in these conditions has prospects for utilization as a clean solid fuel.