Effect of pyrolysis temperature on Si release of alkali-enhanced Si-rich biochar and plant response

Recent studies have shown that silicon (Si) dissolution from biochar may be influenced by the pyrolysis temperature. In addition, the enhancement of biochar by treatment with alkali has been proposed to produce a Si source that can be used for environmentally friendly plant disease control. In this study, biochars from rice straw and rice husk pretreated with KOH, CaO and K2CO3 and then pyrolyzed at 350, 450 and 550 °C were prepared to evaluate the effects of pyrolysis temperature on Si release and plant uptake from alkali-enhanced Si-rich biochar. Extractable Si and dissolution Si from the prepared biochars were assessed by different short-term chemical methods and long-term (30-day) release in dilute acid and neutral salt solutions, respectively, along with a rice potting experiment in greenhouse. For both rice straw- and husk-derived alkali-enhanced biochars (RS-10KB and HS-10K2B, respectively), increasing the pyrolysis temperature from 350 to 550 °C generally had the highest extractable Si and increased Si content extracted by 5-day sodium carbonate and ammonium nitrate (5dSCAN) designated for fertilizer Si by 61–142%, whereas non-enhanced biochars had more extractable Si at 350 °C. The alkali-enhanced biochars produced at 550 °C pyrolysis temperature also released 82–172% and 27–79% more Si than that of 350 °C produced biochar in unbuffered weak acid and neutral salt solutions, respectively, over 30 days. In addition, alkali-enhanced biochars, especially that derived from rice husk at 550 °C facilitated 6–21% greater Si uptake by rice and 44–101% higher rice grain yields than lower temperature biochars, non-enhanced biochars, or conventional Si fertilizers (wollastonite and silicate calcium slag). Overall, this study demonstrated that 550 °C is more efficient than lower pyrolysis temperature for preparing alkali-enhanced biochar to improve Si release for plant growth.


Introduction
The beneficial effect of Si on plant growth has been well documented. It includes improving plant resistance to pest and disease, enhancing mechanical strength, and decreasing nutrient imbalance, as well as alleviating metal toxicity and other stresses due to soil salinity, extreme temperature and drought conditions (Meyer and Keeping 2001;Neumann and Zur 2001;Epstein 2009;Guntzer et al. 2012;Adrees et al. 2015;Islam et al. 2020). Thus, the application of Si fertilizers to plants, especially Si-accumulators such as rice, maize, sugarcane, wheat, and ryegrass is necessary to ensure plant healthy growth. Recently, several studies have shown that biochars produced by pyrolyzing Si-rich waste biomass could be also used as bioavailable Si source to Siaccumulator plants (Houben et al. 2014;Liu et al. 2014;Li et al. , 2018Abbas et al. 2017;Linam et al. 2021). Application of biochars significantly increased Si content in stem and blades of wheat, rice, and perennial ryegrass (Liu et al. 2014;Abbas et al. 2017;Wang et al. 2018aWang et al. , 2019a. Therefore, biochar could serve as a low-cost, renewable, and environmentally friendly Si source. The characteristics of biochar are influenced by various factors, including feedstock biomass types and pyrolysis conditions. Pyrolysis temperature has been considered as 1 3 one of the most important factors affecting physical, chemical and structural characteristics of biochar (Singh et al. 2012;Keiluweit et al. 2010;Nguyen et al. 2010;Kuzyakov et al. 2009). Compared with biochars prepared at low temperatures, high-pyrolysis temperature biochars tend to have higher carbon content, pH, ash content, specific surface area, aromatization and thermal stability, but exhibit lower nitrogen content, yield, and aliphatic carbon fractions Suliman et al. 2016;Ahmad et al. 2012;Kim et al. 2012;Novak et al. 2009).
Pyrolysis temperature also affects the interaction of carbon with silicon in biochar. For biomass feedstocks that are rich in amorphous silicon, it was reported that pyrolysis in the temperature range of 250 ~ 350 °C could cause a crack of carbon layer in the C-Si-C structural configuration of biomass which exposes internal Si and leads to an increasing solubility of Si in biochar (Xiao et al. 2014). However, as pyrolysis temperature is further raised to 500 ~ 700 °C, the solubility of Si is decreased due to aromatization of carbon and formation of more crystallized silica (Xiao et al. 2014). On the other hand, a new study reported generally continuous increase in accumulative Si dissolution from several biochars derived from both Si-rich and deficient biomass feedstocks as pyrolysis temperature increased from 300 to 700 °C (Wang et al. 2018b). While Si release rates for Si-rich biochars were found to be similar at different pyrolysis temperatures with different final amounts of total Si released, the Si-deficient biochars showed a high and low stages of Si release rates but with no difference in the final amount of total Si dissolved (Wang et al. 2018b). In addition, others also reported the increasing extractable Si content of rice husk and sugarcane bagasse biochars based on chemical extractions as pyrolysis temperature increased from 300 to 700 °C (Nwajiaku et al. 2018). These results suggest the solubility of Si or Si availability of biochar likely depends on the interaction between the specific feedstock and pyrolysis temperature. These biochar Si solubility characteristics have not been evaluated in regard to plant response. On the other hand, Wang et al. (2018aWang et al. ( ,2019a recently proposed the use of alkali-enhanced biochar as Si fertilizer. However, the effects of pyrolysis temperature on the properties of alkalienhanced biochars have not yet been explored. In this study, rice straw and husk were selected to prepare KOH, K 2 CO 3 and CaO-enhanced Si-biochars at different pyrolysis temperatures. The effects of temperature on physical and chemical properties as well as total and extractable Si were investigated. In addition, since the dissolution tendency of Si in Si fertilizer was highly correlated with Si content released into soil (Korndörfer et al. 2001;Sebastian et al. 2013;Xiao et al. 2014), soil available Si for crop absorption could be simulated by acid or alkali extraction due to depolymerization of non-monomeric Si (Iler 1979;NIAES 1987;Bao 2000). Moreover, soil pH changes could affect biochar phytolith Si dissolution (Fraysse et al. 2006;Wang et al. 2018a). Therefore, weak acid and neutral salt solution were chosen to evaluate Si release of alkali-enhanced biochars prepared at different pyrolysis temperatures, and the associated plant response was measured to provide a basis for the rational application of alkali-enhanced biochar as an alternative Si source.

Alkali-enhanced biochar preparation
Rice straw (RS) and husk (RH) collected from Louisiana State University AgCenter Rice Research Station (Crowley, LA, USA) were utilized for biochar preparation. Collected feedstock samples were rinsed using deionized (D.I.) water and then dried at 60 °C for 24 h. The dried rice straw and husk biomass samples were ground to pass a 1-mm sieve using a cutting mill. For making biochar, rice straw and husk were mixed, respectively, with KOH, K 2 CO 3 or CaO, in powder form and at alkali:biomass ratios (on weight basis) of 0:100, 5:100, and 10:100 in a porcelain crucible followed by mixing with 100 mL of D.I. water (for 50 g biomass) for 90 min. The biomass-containing crucibles were then placed in muffle furnace under N 2 flow at 400 mL min −1 to purge air from the system for 30 min. Pyrolysis was carried out by setting muffle furnace temperature at 180 °C for dehydration for 30 min and then maintaining at 350, 450 and 550 °C, respectively, for 60 min under N 2 purge at 200 mL min −1 (Wang et al. 2018a). The resulting biochar samples were cooled and ground to pass a 1-mm sieve before characterization. For convenience of discussion, these produced biochars are referred as 0B (no alkali pretreatment), 5KB and 10KB (pretreatment with KOH at 5:100 and 10:100 ratio), 5K2B and 10K2B (pretreatment with K 2 CO 3 at 5:100 and 10:100 ratio), and 5CB and 10CB (pretreatment with CaO at 5:100 and 10:100 ratio), respectively.

Chemical characterization of biochar Si and its release
Biochar sample pH was measured based on 1:100 biochar to D.I. water ratio (Cantrell et al. 2011). Ash content was measured by furnace at 550 °C for 5 h. Total C and N analyses were carried out using an elemental analyzer (Elementar Analysen systeme GmbH, Germany). Biochar total P, K, Ca, Mg, Cu and Fe were determined by ICP-AES (SPECTRO Plasma 3200, Germany) after digestion with nitric acid and H 2 O 2 (Huang and Schulte 1985). Total Si in alkali-enhanced biochars and feedstocks was determined using a 1.5 M HF-0.6 M HCl extraction, and fertilizer soluble Si fraction in biochar samples was evaluated using 5-day sodium carbonate and ammonium nitrate (Na 2 CO 3 -NH 4 NO 3 ) extraction (referred as 5dSCAN) designated by the Association of American Plant Food Control Officials (AAPFCO) followed by light absorption spectrometry (Saito et al. 2005;Sebastian et al. 2013). Specifically, the HF-HCl extraction was performed at a 1:100 solid to solution ratio for 1 h with stirring every 10 min. The 5dSCAN extraction procedure was carried out at a 1:1000 solid to solution ratio for 1 h shaking (at 140 rpm) followed by settling for 4 days and 23 h. The samples were then filtered using a Whatman filter paper (< 2.5μ) before Si analysis. Four additional Si test methods including 0.5 M HCl, 1 M sodium acetate buffer (pH 4.0), 1-h equilibrium of 0.094 M Na 2 CO 3 -0.2 M NH 4 NO 3 mixed solution (1hSCAN), and 0.5 M ammonium acetate were also used to determine different levels of extractable Si from alkali-enhanced biochars along with commonly used mineral Si fertilizer wollastonite (WO) (W10, Vansil®, R.T. Vanderbilt Co, Norwalk, CT). Among these four methods, 0.5 M HCl is the official procedure to evaluate available Si from slag-based calcium silicate fertilizer in Japan (NIAES 1987;Ma and Takahashi 2002). The 1 M acetate buffer (pH 4.0) method is widely used to evaluate available Si in soil as well as in slags (NIAES 1987). The method could extract soluble and some exchangeable Si (Snyder 2001), although it has been shown to over-estimate Si availability in soil amended with calcium silicate slag (Takahashi 1981) and in calcareous soils (Liang et al. 1994). The 0.5 M ammonium acetate extraction method also extracts soluble and some exchangeable Si (Takahashi 1981;Kato and Owa 1997). The 1hSCAN extraction, the shorter version of 5dSCAN procedure, has been routinely used in Brazil (Korndorfer et al. 2004). For these four methods, the extractions were based on 1:1000 solid to solution ratio for 1 h of shaking at 140 rpm followed by filtration (Hallmark et al. 1982;Wang et al. 2004).
In addition, the total Si released during 30 days from RS, RH, K 2 CO 3 -enhanced RS and RH biochars (RS-0B/5K2B/10K2B and RH-0B/5K2B/10K2B) was evaluated in unbuffered weak acid solution (0.1 mM HCl) and neutral salt solution (0.01 M KCl), respectively. For comparison, wollastonite (WO) and industrial by-product silicate slag (SL) (SILI-CAL™, Calcium Silicate Corp., Lake Harbor, FL) samples were also investigated. In doing so, a series of 30-mg samples (ground to pass a 1-mm sieve) with each mixed with 30 ml of extractant solutions in 50-ml polyethylene centrifuge tubes were prepared. The mixtures were shaken continuously at 120 rpm under 25.0 ± 0.5 °C for 1 h, 1, 5, 9, 14, 21 and 30 days, respectively. The samples were then filtered using Whatman filter paper (< 2.5 μm) before analysis.
The Si in all the filtrates was determined using a SPEC-TRONIC 501 light absorption spectrophotometer (Thermo Scientific, Wilmington, DE) following blue silicomolybdous acid procedures (Hallmark et al. 1982;Wang et al. 2004;Sebastian et al. 2013). The spectrometer was calibrated using a standard curve for each extraction method before sample analysis. For all analytical procedures, duplicate samples were measured and a spike recovery of 95% of a known Si standard in selected samples was performed for quality control of analytical process.

Rice greenhouse potting study
A potting experiment using rice (Oryza sativa L) was conducted to assess plant response to the produced biochar Si source in the greenhouse at Louisiana State University, Baton Rouge, USA. The soil used in potting study was Crowley silt loam (fine-silty, montmorillonitic, thermic Typic Albaqualf) with pH 7.6, and its Mehlich-III soil-test extractable P and K were 10 and 69 mg kg −1 , respectively. Twelve types of Si sources including RS, RH, four alkalienhanced RS biochars produced at 350 and 550 °C (RS-0B-350, RS-10K2B-350, RS-0B-550, RS-10K2B-550), four alkali-enhanced RH biochars produced at 350 and 550°C (RH-0B-350, RH-10K2B-350, RH-0B-550, RH-10K2B-550), WO, SL along with a control (CK) without addition of amendment were used. Each Si source was applied at the rate 0. 22% on soil mass basis (equivalent to 5 t ha −1 ). All pots received a blanket application of N 0.0357 g kg −1 (80 kg N ha −1 ), P 2 O 5 0.0302 g kg −1 (68 kg P 2 O 5 ha −1 ), and K 2 O 0.0302 g kg −1 (68 kg K 2 O ha −1 ) before planting. Two rice seeds were sown in a 12-cm diameter plastic pot. A second N application of 0.025 g kg −1 (56 kg ha −1 ) was made to each pot after three weeks, and then the pot was flooded. Rice plants were harvested after 23 weeks of growth in the pots. Rice grain (husk and seed) and straw (stem and blade) were washed with deionized water and dried at 60 °C for 48 h. Plant Si content was analyzed by extracting the tissue with 1.5 M HF-0.6 M HCl, followed by detection using light absorption spectrometry (Saito et al. 2005;Sebastian et al. 2013). All treatments were replicated three times.
All statistical analyses were carried out using the Statistical Analysis Software, version 9.0 (SAS Institute, Cary, NC). Average results for the different treatments were compared using analysis of variance.

Chemical and physical properties of alkali-enhanced biochars at different preparation temperatures
In general, biochar yield was reduced with increasing pyrolysis temperature in all alkali-enhanced biochars (Tables S1 and  S2). The pH and ash content of alkali-enhanced rice straw and husk biochars increased with increasing pyrolysis temperature (Tables S1 and S2). Specifically, the pH of 350 °C prepared RS-0B and RH-0B were 6.46 ~ 6.65, 450 and 550 °C treatments increased the pH of 0B by 1.53 ~ 2. 24 and 2.12 ~ 2.95 unit, respectively. As the proportion of KOH, K 2 CO 3 and CaO increased, pH of biochars increased correspondingly. From 350 to 550 °C, ash contents of RS-0B and RH-0B increased from 32 and 34% to 45 and 44%, respectively. After alkali pretreatment, ash contents of 550 °C prepared rice straw-and husk-derived 10KB, 10K2B and 10CB were increased by 8 ~ 28%, 9 ~ 12% and 16 ~ 24%, respectively, as compared to non-enhanced 0B samples. Ash contents of 350 °C prepared biochars showed the similar trend. The carbon content of both RS-0B and RH-0B biochar samples increased with increasing temperature while C and N contents decreased as alkali proportion increased (Tables S1 and S2). At 550 °C, C contents of 550 °C prepared RSand RH-10KB, 10K2B, 10CB decreased by 22~24%, 7 ~ 11%, and 27 ~ 45%, respectively, comparing to 0B biochar samples. Increasing pyrolysis temperature also generally increased total P, K, Ca and Mg contents of RS and RH biochars, although they reached maximum concentrations at different temperature. As expected, biochar total K and Ca contents increased with increasing alkali amendment ratio of KOH, K 2 CO 3 and CaO, respectively. Overall, all three alkali pretreatments affected the elemental composition of prepared biochars at different pyrolysis temperatures.

Total Si content of alkali-enhanced biochar at different preparation temperatures
Total Si in the alkali-enhanced biochar changed with biomass feedstock, pyrolysis temperature, and type and proportion of alkali pretreatment (Fig. 1). For 0B samples, total Si was maximized at 450 °C for rice straw and at 550 °C for rice husk. Increasing alkali pretreatment proportion tended to decrease total Si content in biochar samples with KOH, K 2 CO 3 and CaO pretreatment. On the other hand, for the rice straw biochar, total Si content of 10KB prepared at 350 °C was 21.8% and decreased to 18.9% with increasing temperature. When pyrolysis temperature was raised to 550 °C, total Si content of 10CB was 64% higher than that of rice straw biochar prepared at 350 °C (Fig. 1A). A similar trend was also observed for rice husk biochars, although 0B samples had generally higher total Si compared to that of rice straw biochars (Fig. 1B), emphasizing the impact of the feedstock.

5dSCAN-extractable Si content of alkali-enhanced biochar at different preparation temperatures
Extractable Si evaluation of produced biochar Si sources using the designated 5dSCAN method by AAPFCO (Sebastian et al. 2013) for fertilizer Si is presented in Fig. 2. Although total Si was higher in rice husk biochar ( Fig. 1), rice straw biochar showed generally higher 5dSCAN-Si (Fig. 2). For all pyrolysis temperatures, 5dSCAN-Si content of non-enhanced 0B samples, regardless of feedstock source, was higher than that of the feedstock sample. The non-enhanced 0B samples prepared at 350 °C exhibited the maximum Si content, but it decreased with pyrolysis temperature (Fig. 2). For biochar samples, increasing pyrolysis temperature generally shifted the 5dSCAN-Si concentration higher at 550 °C, particularly in biochars with greater alkali pretreatment. For example, in the KOH-enhanced rice straw biochar, the 10KB prepared at 550 °C yielded 5dSCAN-Si content of 4.67%, which was significantly higher than the 10KB prepared at 350 and 450 °C by 110 and 61%, respectively. The 5dSCAN-Si content of the 10K2B and 10CB treatments prepared at 550 °C was also 61 and 62% higher than that of those prepared at 350 °C ( Fig. 2A). A similar trend was seen for the rice husk biochars (Fig. 2B). Interestingly, the extractable Si content of non-enhanced rice husk 0B by 5dSCAN method was less than that of rice straw 0B samples. In addition, the 5dSCAN-Si content of biochars pretreated with KOH and K 2 CO 3 increased significantly with increasing pyrolysis temperature; however, the 5dSCAN-Si content of 5CB and 10CB biochars pretreated with CaO remained relatively stable. Figure 3 shows the comparison of extractable Si level in alkali-enhanced biochar sources by four additional extraction methods. Since high alkali pretreatment yielded higher Si extractability, only 10KB, 10K2B, and 10CB were selected in this comparison along with feedstock, 0B and wollastonite (WO), which was used as a standard Si fertilizer. The 0.5 M HCl method extracted significantly more Si from wollastonite than that from biochars produced at all three pyrolysis temperatures, regardless of which biomass feedstocks were used. On the other hand, the 1hSCAN method extracted significantly more Si from alkali-enhanced biochars made from both feedstocks, especially rice straw, than that from wollastonite. The 1hSCAN method yielded equivalent or even more extractable Si from alkali-enhanced biochars made at 350 and 550 °C than that from the 0.5 M HCl extraction of wollastonite (Fig. 3). This was particularly evident for the 10K2B, 10KB, and 10CB treatments prepared at 550 °C, which showed the increase of 1hSCAN-Si content of 2163, 1675, and 903% higher than wollastonite, respectively. Furthermore, both 1 M sodium acetate buffer and 0.5 M ammonium acetate methods generally extracted less Si compared to the 1hSCAN extraction procedure but they extracted more Si from alkali-enhanced biochars than that from wollastonite as well as from non-enhanced 0B samples. These four additional test methods also showed that alkali-enhanced biochars tended to have greater extractable Si concentrations when produced at higher pyrolysis temperatures, and extractable Si concentrations were generally significantly higher than those of non-enhanced 0B biochars. Overall, these methods measure various fractions of extractable Si from different Si sources. Nonetheless, it was found that the Si fraction extracted by 0.5 M HCl was closely related to that extracted by the buffered sodium acetate (pH 4.0) (r = 0.967, P < 0.01), whereas the 1hSCAN extraction method was related to 0.5 M ammonium acetate method (r = 0.978, P < 0.01). The 1 h and 5 day Na 2 CO 3 -NH 4 NO 3 methods were only weakly related (r = 0.626, P < 0.05).

Si release from alkali-enhanced biochar prepared at different pyrolysis temperatures
Total Si released over 30 days from biochar prepared at different temperatures in unbuffered weak acid (0.1 mM HCl, pH 4.0) and neutral salt (0.01 M KCl, initially adjusted to pH 7.0) solutions are shown in Figs. 4 and 5, respectively. Since both total Si and 5dSCAN-Si contents of biochars were similar between KOH and K 2 CO 3 pretreatments and were generally greater than those of the CaO pretreatment, only K 2 CO 3 -enhanced rice straw biochars were selected for this evaluation. In this assessment, the alkali-enhanced biochar was compared with 0B (non-enhanced), feedstock (RS) as well as with wollastonite and silica slag. For both acid and neutral salt solutions, the 10K2B biochar showed the greatest cumulative Si release, followed by 5K2B biochar. The Si release from the both was higher than that from the 0B biochar as well as that from the raw feedstock, wollastonite and silica slag over the 30-day incubation. The 30-day cumulative Si release in the unbuffered weak acid and neutral salt solutions was highly correlated (r = 0.927, P < 0.05 for rice straw biochars and r = 0.995, P < 0.01 for husk biochars, respectively). A higher pyrolysis temperature generally led to greater Si release, with 450 and 550 °C-prepared biochars yielding greater Si release than 350 °C prepared biochar. The Si release characteristics suggested that these samples could be divided into three groups: (1) the RS, WO and SL group that released the least Si over 30 days and for which Si release curves were close to each other or mostly overlapped, (2) the K 2 CO 3 -enhanced biochar (5K2B and 10K2B) group, which showed a rapid Si release during the first 2 days, continued to release gradually with time and had the greatest Si release among all treatments, and (3) the 0B biochars, which yielded Si release curves between those of the other two groups. It is interesting to note that the Si release curves and the cumulative release from the rice straw 0B biochars were closer to those of the K 2 CO 3 -enhanced biochars for rice straw feedstock (Figs. 4A and 5A), while the rice husk 0B samples were not to their K 2 CO 3 -enhanced biochars for rice husk (Figs. 4B and 5B), indicating a difference in the nature of Si release between the two biochar feedstocks in responding to alkali pretreatment.

Effect of alkali-enhanced biochar application on rice Si uptake and yield
Rice growth in soils treated with alkali-enhanced biochars along with non-enhanced biochar (0B) produced at different temperatures was evaluated in a greenhouse potting study, and the results are presented in Figs. 6 and 7. Based on the amount of Si extracted by different chemical methods and the cumulative Si release studies (Figs. 2, 3, 4 and 5), the potting trial focused only on a comparison of K 2 CO 3 -enhanced biochar with non-enhanced 0B biochar, wollastonite, and silica slag. Application of biochars significantly increased the Si content of rice straw compared to that of the control (CK) treatment. The Si uptake under the treatments of rice straw biochars, RS-10K2B-350 and RS-10K2B-550, was 31 and 29% higher than that of the control treatment, respectively (Fig. 6A). The Si uptake under the treatment of rice husk biochar RH-10K2B-550 was the highest among all Si treatments and was significantly higher than wollastonite and calcium silicate slag by 34 and 24%, respectively (Fig. 6B). In addition, for the rice husk-derived biochar treatments, the Si uptake contents in straw associated with the RH-0B-550 and RH-10K2B-550 treatments were significantly higher than those of the corresponding RH-0B-350 and RH-10K2B-350 treatments by 9 and 33%, respectively. These results indicated that alkali-enhanced biochar Si sources prepared at high pyrolysis temperatures tended to promote higher Si uptake by rice straw, particularly in the case of the rice husk-derived biochar products. Furthermore, the alkali-enhanced biochar Si sources also increased rice grain yields. Applications of both RS-and RH-10K2B-550 biochars produced significantly higher rice grain yields than those of nonenhanced RS-and RH-0B-350 biochars (Fig. 7). The RH-10K2B-550 application had significantly higher rice yield than the RH-0B-550 treatment, while RS-10K2B-550 had greater yield than RS-0B-550 (Fig. 7). In addition, RH-10K2B-350 treatment led to a rice yield that was numerically higher than the yield of the RH-0B-350 treatment. Rice yield associated with the RH-10K2B-550 treatment was even significantly greater than that of wollastonite and silica slag amendments. Grain yield was significantly correlated with Si uptake in rice straw (r = 0.808, P < 0.01) for biochar Si sources of RS/RH-0B-350/550 and RS/RH-10K2B-350/550 treatments (Fig. S2). These results indicate that both pyrolysis temperature and alkalipretreatment are important in producing biochar-based Si fertilizer source. These results were consistent with greater cumulative dissolution of Si in both weak acid and unbuffered neutral salt solutions from the alkali-enhanced   0B, 5K2B and 10K2B indicate the biochars prepared at the proportion of K 2 CO 3 to feedstock of 0:100, 5:100 and 10:100; WO wollastonite, SL calcium silicate slag. The error bar indicates plus and minus of one standard deviation

Effect of pyrolysis temperature and alkali treatment on biochar chemical and physical properties
As pyrolysis temperature increased from 350 to 550 °C, the yield of both non-enhanced (0B) and alkali-enhanced biochars from rice straw and husk sharply decreased (Tables S1 and S2), due to increasing decomposition of lignocellulosic materials in biomass (Paris et al. 2005). This result is similar to studies of wood, rice husk and sugarcane bagasse biochars produced at different pyrolysis temperatures (Kim et al. 2012;Keiluweit et al. 2010;Jeong et al. 2016;Suliman et al. 2016;Nwajiaku et al. 2018). In addition, alkali-enhanced biochars from both rice straw and rice husk increased the pH and ash content as pyrolysis temperature increased from 350 to 550 °C, consistent with non-enhanced biochar (0B). The latter was attributed to the conversion of alkali and alkaline earth cations (Ca, Mg, K, Na) to alkaline substances such as oxides, hydroxides, chlorides, and carbonates and the loss of acidic functional groups (Novak et al. 2009;Van Zwieten et al. 2010;Ahmad et al. 2012). Higher ash content was generally observed with alkali pretreatment of larger proportion of KOH, K 2 CO 3 and CaO. On the other hand, the effect of pyrolysis temperature on the C content of alkali-enhanced biochar was less consistent, although C concentration in the non-enhanced biochar (0B) generally increased, which has been widely reported along with increase in concentrations and husk biochars prepared with the proportion of K 2 CO 3 to feedstock at 0 and 10 at 350 °C, respectively; RS/RH-0B-550 and RS/ RH-10K2B-550 are rice straw biochars prepared with the proportion of K 2 CO 3 to feedstock at 0 and 10 at 550 °C, respectively; WO wollastonite, SL silicate calcium slag. The error bar indicates plus and minus of one standard deviation of inorganic elements (Novak et al. 2009;Laird 2008;Jeong et al. 2016). The increase in alkali pretreatment proportion generally decreased biochar C content. Between 350 and 450 °C pyrolysis temperature, alkali pretreatment showed inconsistent changes in biochar C content. However, at 550 °C, C content of KOH-enhanced biochars generally decreased compared to 0B biochar samples (Tables S1 and  S2). This was likely due to a removal of more non-fixed carbon in feedstock biomass by KOH during pyrolysis (Huang et al. 2017) and a relative increase in the mass of alkali agent as ash components (Agrafioti et al. 2013). In addition, as expected, biochar total K and Ca contents were significantly elevated with specific amendment ratio of KOH, K 2 CO 3 and CaO, respectively. Overall, similar to non-enhanced biochar, all three alkali pretreatments affected the output yield, ash content, and pH as well as part of elemental composition of prepared biochars at different pyrolysis temperatures. Also, the significantly enhanced total K content of KOH and K 2 CO 3 -enhanced biochars could function as K source in soil biochar amendment. Several studies have investigated the influence of alkali agents on biomass C transformation during pyrolysis (Azargohar and Dalai 2006;Chen et al. 2020;Yang et al. 2020;Zhao et al. 2020). In general, decarboxylation reaction was known as the main process to remove C from feedstock during pyrolysis and it had similar preference for biomass as dehydration reaction under low temperature pyrolysis (Liu and Han 2015). In addition, decarboxylation reaction was shown to dominate at pyrolysis temperature above 350 °C, whereas decarbonylation reaction was at above 450 °C (Zhang et al. 2020). At 550 °C, biochar exhibited stable characteristic atomic ratios of H/C < 0.4 and O/C < 0. 2 (Jeong et al. 2016) and significant fusion into aromatic ring structure (Zhang et al. 2020). The higher pyrolysis temperature mainly promoted the decomposition of the volatiles into the gaseous products while did not promote the decomposition of biochar significantly (Zhang et al. 2020). In our study, the clear decrease in C contents of alkali-enhanced biochars at 550 °C compared to 0B biochars indicated the significant role of alkali pretreatment (Tables S1 and S2). Alkali agent such as KOH was found to promote dehydration, decarboxylation, and decarbonylation reactions with biomass during pyrolysis Zhao et al. 2020). The KOH treatment could enhance the decomposition of volatile intermediate, strengthen the reaction with stable C fragments, and facilitate the overall reaction rate with C matrix of biomass (Azargohar and Dalai 2006;Chen et al. 2020).

Effect of pyrolysis temperature and alkali treatment on biochar phytolith-Si release
Biochar total Si content generally increased with increasing pyrolysis temperature. However, total Si content of some treatments such as RS-5KB and RH-5K2B decreased from 450 to 550 °C. The loss was probably due to the formation of trace silica scale sticking to the wall of crucible during the production of alkali-enhanced biochar. Similar trend of lower total Si content was also observed when biochar pyrolysis temperature increased from 300 to 700 °C . Non-enhanced biochar (0B) samples showed the higher 5dSCAN-Si content at 350 °C pyrolysis temperature, and it had lower extractable Si as pyrolysis temperature was increased to 550 °C (Fig. 2). This result was consistent with the research of Xiao et al. (2014), who reported higher Si dissolution release from pristine rice straw biochar of 350 °C than that of biochars prepared at lower (150-250 °C) and higher (500-700 °C) pyrolysis temperatures. Nguyen et al. (2014) also reported higher Si dissolution from rice straw ashing at 400 °C than that at 600-800 °C. Based on various spectroscopic analyses of structure, morphology and species, Xiao et al. (2014) proposed a C-Si-C structure model to explain this phenomenon. According to these authors, polymerization of monosilicic acid dominated at < 250 °C through dehydration to result in a dense C-Si-C structure in biochar. As pyrolysis temperature was increased to 350 °C, the outermost C layer began to crack, exposing the inner Si layer of C-Si-C structural configuration and generating a large amount of dissolved Si. At 500-700 °C, the C layer of C-Si-C structure was further detached. However, due to crystallization of Si, the cumulative release of dissolved Si was less than biochar produced at 350 °C (Xiao et al. 2014;Wang et al. 2018b). The C-Si coupling interaction in response to pyrolysis temperature change was further suggested as the mechanism for mutual protection between Si and C from dissolution and responsible for biochar C stability (Guo and Chen 2014;Wang et al. 2019b). The enhanced BC stability at pyrolysis temperature at above 500 °C was attributable to the increased aromaticity along with Si encapsulation of C and Si components changing from amorphous to crystalline and from metastable α-quartz to stable β-quartz phase, which is intertwined with C (Guo and Chen 2014). In our study, through alkali pretreatment, 5dSCAN-Si content of resulting biochars was clearly increased as evidenced in 10KB, 10K2B and 10CB samples as pyrolysis temperature was increased from 350 to 550 °C (Fig. 2). In a previous study, we explored the release mechanism of phytolith-Si in alkali-enhanced biochar at 550 °C and showed that alkali such as KOH pretreatment could significantly damage the dumbbell structures of phytoliths that are stacked in layers by exposing the outermost layers of phytoliths for Si release (Wang et al. 2018a). Liu et al. (2012) also indicated that KOH could be embedded into the interior of phytolith and react with SiO 2 to generate less stable silicate as: It is well known that K 2 SiO 3 dissolution is much greater than SiO 2 due to the change in silicate structure from threedimensional tectosilicate to single-chained inosilicate (Sparks 2003). While increased pyrolysis temperature could accelerate the burst of C layer of C-Si-C structural configuration and expose Si in the inner layer (Xiao et al. 2014), the enhanced contact of alkali reagents with inner Si layer accelerated the Si dissolution rather than Si crystallization in biochar (Fig. S1). The SEM-EDX image and elemental mapping clearly showed bleaching out of photolith-Si with KOH treatment while CaO treatment had less impact due to weak base nature and that CaO tended to stay on phyolith structural surface (Fig. S1). The latter likely explained the trend reversal in 5dSCAN-extractable Si of 10KB, 10K2B and 10CB as compared to 0B samples over increasing pyrolysis temperature.
The higher 5dSCAN-extractable Si content in alkalienhanced biochars compared to non-enhanced biochars as influenced by pyrolysis temperature was also supported by cumulative release of Si over 30 days in unbuffered weak acid and neutral salt solutions (Figs. 4 and 5). Among rice husk and straw biochars prepared at different pyrolysis temperatures, alkali-enhanced biochars (5K2B and 10K2B) had the highest Si release in unbuffered weak acid and neutral salt solutions, and was also higher than feedstocks (RS and RH) and conventional Si fertilizers (WO and SL). The increase in pyrolysis temperature generally significantly increased Si release of alkali-enhanced 5K2B and 10K2B samples while it had less effect on non-enhanced 0B samples. This was consistent with the result of 5dSCAN-extractable Si content of alkali-enhanced biochars prepared at different temperatures (Fig. 2). These results further indicate the importance of alkali pretreatment, especially K 2 CO 3, and appropriate higher pyrolysis temperature (550 °C) in preparing biochar Si fertilizer.

Effect of pyrolysis temperature on Si uptake from alkali-enhanced biochar
The unique Si dissolution and release characteristics of biochar with higher Si dissolution at 350-400 °C than the temperature out of this range noted by Xiao et al. (2014) and Nguyen et al. (2014) suggested that pyrolysis temperature could be an important factor to control Si release of biochars in soil. This phenomenon was, however, never examined previously through plant response, although several studies demonstrated that biochar made from Si-rich waste biomass could be used as a Si fertilizer source (Houben et al. 2014;Liu et al. 2014;Li et al. , 2018Abbas et al. 2017).
Most biochars used as soil amendment for agronomic application were prepared at temperatures greater than 450 °C to (1) 2KOH + SiO 2 → K 2 SiO 3 + H 2 O enhance C stability and to improve soil water and nutrient retention (Laird 2008;Spokas et al. 2012;Jeong et al. 2016).
On the other hand, Wang et al. (2018a) showed that alkalienhanced biochar could significantly increase extractable Si content of phytolith-rich biochars using AAPFCO-designated 5dSCAN extraction method for Si fertilizer evaluation. In this study, the greenhouse potting trial clearly showed that amendments of all biochars increased rice straw Si content. Treatment of the soil with alkali-enhanced rice husk biochar prepared at 550 °C (RH-10K2B-550) led to the highest Si uptake, greater than that of wollastonite and slag (Fig. 6). Most importantly, the same biochar treatment also showed the highest grain yield increase among all biochar amendments and had comparable or higher grain yield than the wollastonite and slag treatments (Fig. 7). The higher grain yield was correlated with the greater cumulative Si release in unbuffered weak acid and neutral salt solutions from alkalienhanced biochars at 550 °C than from those prepared at 350 °C or from non-enhanced biochars (0B) (Figs. 4 and 5).
In addition, the difficulty in indexing fertilizer Si sources should be pointed out. Although different chemical methods have been used to evaluate fertilizer Si sources, various studies showed no single extraction that could predict Si availability (Haynes et al. 2013;Zellner et al. 2015). Therefore, the specific response of plant growth is especially important for evaluating different Si sources, particularly when it comes to biochar-based Si fertilizer. A recent study indicated that the chemical methods that could extract more Si quantitatively from both mineral fertilizers and biomassbased thermal products as well as could reproduce soil root conditions by buffering pH close to neutral likely do better in correlating with plant Si uptake (Duboc et al. 2019). In our study, we used unbuffered weak acid (0.1 mM HCl) and neutral salt solution (0.01 M KCl) to assess Si release over 30 days. The 30-day cumulative Si release in the unbuffered 0.1 mM HCl and 0.01 M KCl solutions showed good correlations with rice straw Si uptake (r = 0.879 and 0.881, P < 0.05, respectively). In addition, they were strongly correlated with rice grain yields (r = 0.924 and 0.890, P < 0.01, respectively) (Figs. S3 and S4). Among the four extraction methods, 1hSCAN and ammonium acetate extractions were found most closely correlated with rice yield (r = 0.708, P < 0.05). Nonetheless it was clear that alkali-enhanced biochar at 550 °C, especially K 2 CO 3 pretreatment, facilitated better rice Si uptake and grain yield than the non-enhanced biochars or biochars prepared at lower temperatures.
It should be pointed out while the enhanced phytolith-Si solubility of biochar by alkali pretreatment likely primarily contributed to the greater rice Si uptake and higher yield, the potentially stronger interaction between biochar and rice roots could play a role. Recent work suggested that biochar amendment could cause an increase in electrical potential difference between rice root membrane and soil due to increased soil E h and pH, which improves nutrient uptake (Quin et al. 2015;Sun et al. 2017;Chew et al. 2020). Biochar particles were found to be embedded into the plaque layer formed on rice roots and root hairs could enter the pores of biochar (Joseph et al. 2013). According to Zeng et al. (2014) and Haruta et al. (2018), O 2 , K + and pH are some of key environmental factors influencing plant root membrane potentials. Although we did not measure rhizosphere E h change in potting study, the generally high pH and K content of alkali-enhanced biochars used in the study (Table S1) would be expected to increase root membrane/soil potential difference due to increased soil E h with increasing pH, which drives nutrient uptake (Chew et al. 2020;Zeng et al. 2014). Sufficient K nutrient could ensure K + -signaling that mediates plant-adaptive response to environment including those of Si-specific transporters (Hosseini et al. 2017). Both could help improve Si uptake as monosilicic acid uptake besides solubilizing phytoliths-Si.

Conclusion
This study demonstrated the influence of pyrolysis temperature on the Si release of alkali-enhanced biochars. Different from unenhanced biochar, which generally showed higher chemically extractable Si when prepared at 350 °C, alkalienhanced biochar, especially with KOH or K 2 CO 3 , tended to have higher extractable Si when it was prepared at 550 °C. In addition, biochar prepared with K 2 CO 3 pretreatment at 550 °C released more Si to weak acid and unbuffered neutral salt solutions over 30 days. The greenhouse potting trial also showed that rice plants responded, with greater Si uptake and grain yield, to alkali-enhanced biochar. Overall, 550 °C was the better pyrolysis temperature to prepare alkali-enhanced biochar that could increase Si release for plant growth as compared to 350 or 450 °C. Additionally, KOH-or K 2 CO 3 -enahnced biochar can provide K source besides enhanced C stability.