THE EFFECT OF UNTREATED AND ACIDIFIED BIOCHAR ON NH 3 -N EMISSIONS FROM SLURRY DIGESTATE

,


INTRODUCTION
In recent years, there has been an increase from farming communities in using slurry conditioners.They mitigate NH3 emissions in a cost-efficient manner (Kavanagh et al., 2019) and are relatively easily applied to manure compared with other technologies that require modifications to the existing infrastructure and/or the purchase of expensive equipment (Maurer et al., 2016).Biochar (BC) is a porous material produced through pyrolysis or gasification of biomass at different temperatures with no or very low oxygen (O) availability (Sohi et al., 2009).BC has received increasing attention in recent years because of its diverse functionality in the fields of climate change mitigation, sustainable agriculture and environmental control (Xiao et al., 2018).The BC properties of high surface area, high porosity and high cation or anion exchange capacity make it a promising slurry conditioner to enhance NH4 + retention (Li et al., 2018) and reduce nitrate (NO3 − ) leaching (Saarnio et al., 2018).BC has been shown to decrease NH3 volatilization and improve N retention in poultry litter composting (Agyarko-Mintah et al., 2017) and to adsorb ammonium nitrogen (NH4 + -N) from piggery manure anaerobic digestate slurry (Kizito et al., 2015), thus enhancing the fertilizer value of manure.However, with the increasing quantity of BC addition, the alkalinity of BC is likely to increase the pH of the manure and shift the NH4 + /NH3 equilibrium toward NH3 volatilization (Sun et al., 2019).To address this issue, the acidic oxidation of BC can neutralize the alkaline pH and facilitate the adsorption of NH4 + because of the increased amount of O-containing surface functional groups (Sajjadi et al., 2019).Maurer et al., (2017) determined that untreated BC can effectively mitigate NH3 emissions from stored swine manure.At the same time, Peiris et al., (2019) stated that acid modification can affect the physicochemical properties of the BC which in turn could alter the mitigation effect.
The aim of the present study was to investigate: (i) the ability of hay BC to reduce NH3-N emissions from digestate; (ii) to what extent the suppression effect of NH3-N emissions of BC can be enhanced through acidification; (iii) which method is more effective for the application of untreated and acidified BC (mixed or surface), and (iv) whether untreated and acidified BC addition reduces total N (Ntot), NH4 + -N and nitrate-nitrogen (NO3 − -N) losses in digestate.

Experimental design
The trial was carried out under controlled laboratory conditions at the Estonian University of Life Sciences (EULS) during a period of 240 days from October 2018 until May 2019.The experiment included a total of 12 treatments and consisted of two experimental factors: (i) digestate conditioner (untreated BC, acidified BC and acid) and (ii) application method of digestate conditioner (BC on the surface -BCs; BC mixed into the digestate -BCm).The pure acids were applied in pure form mixed into digestate only (Table 1).For accuracy and logistical reasons, the experiment was conducted in two batches.In the first batch, NH3 emissions and digestate pH were measured and replicated three times from October 2018 until May 2019.In parallel, the second batch examined NH4 + -N, NO3 − -N, total nitrogen (Ntot) and total carbon (Ctot) concentrations, which were replicated four times from February 2019 until April 2019.In both batches, the replications were carried out in the same order and the digestate conditioners were applied in the same amounts, manner and laboratory conditions.

Materials
The BC was produced from reed canary (Phalaris arundinacea L.) hay pellets and torrefied at 300 °C.The quantities on a dry weight basis of total phosphorous (Ptot), total potassium (Ktot), total calcium (Catot) and total magnesium (Mgtot) were 2400 mg kg −1 , 22,800 mg kg −1 , 9300 mg kg −1 and 4700 mg kg −1 , respectively.The ash concentration was 10.4%, cumulative pore volume represented 0.0015 cm3 g −1 , and the concentration of volatile compounds was 62.9%.Physicochemical details of the acidified BC conditioners are shown in Table 2.

Experimental set-up
A multi-gas detection equipment X-am 7000 (Dräger, Lubeck, Germany) connected to an 860-CG acrylic desiccator chamber with a gas port (Plas-Labs, Lansing, MI, USA) was used to measure NH3 concentrations.A schematic of the experimental set-up is shown in Figure 1.
Figure 1.Schematic of the experimental set-up.

NH3 concentration and digestate pH
The NH3 concentrations were measured in two separate sessions.The first session lasted for 48 h (2880 min), immediately after the digestate was exposed to ambient conditions.In the second session, the gas concentrations were monitored for 24 h (1440 min) from the 7th to 8th day after the beginning of the experiment.The pH of the digestate was measured with an HD 2156.2 pH meter (Delta OHM, Padua, Italy) at the beginning of the experiment and after 48, 72, 96 and 120 h.In the untreated and acidified BCs treatments, the pH measurements were recorded under the BC layer (approximately 3-5 cm deep) and in the middle of the vessel (approximately 7-9 cm deep).In the untreated and acidified BCm as well as in the control and acid treatments, the pH was measured only in the middle of the vessel.

Nutrient concentration
The Ntot, Ctot, NH4 + -N and NO3 − -N concentrations were measured in the digestate at the beginning of the experiment and after 30 days.The Ntot and Ctot concentrations were determined after dry combustion with a varioMAX CNS elemental analyzer (Elementar Analysensysteme GmbH, Langenselbold, Germany).The NH4 + -N and NO3 − -N were determined after steam distillation (Bremner and Keeney, 1965) using a UDK 126D distillation unit (VELP Scientifica, Usmate Velate, Italy).The N-containing species were calculated by subtracting the concentration in the BC from the total concentration retained in the digestate.

Statistical analysis
Statistical analysis was conducted using the R programming software (R Development Core Team, 2019).Data were analyzed using analysis of variance (ANOVA) to study the effect of BC treatment (untreated and acidified), time and the interaction between them on NH3 emissions.When significant differences existed, a post-hoc Tukey's honest significant difference (HSD) test was conducted to study the differences between groups (Tukey, 1977).The different N compounds were tested between treatments at the end of the 30-day period.Linear regression and Pearson correlation coefficient were used to analyze the relationship between NH3-N emissions and pH or mineral N concentration.

Effect of untreated BC on NH3-N emissions
The suppression effect of untreated BCs was not different from untreated BCm during the first 3.5 h (p = 0.80).In the untreated BCm treatment, a sharp increase in NH3-N emissions was detected during the first 1.5 h (Figure 2B), which was followed by a decrease between 1.5 and 3.5 h.However, in untreated BCs, no emissions were recorded during the first 1.5 h, and a steep increase in NH3-N emissions was recorded from 1.5 until 3.5 h.From 3.5 h onward, the NH3-N emissions in untreated BCs were higher than untreated BCm, but the differences were not significant (p > 0.05).The short-term effectiveness (<1.5 h) of untreated BCs is likely to be related to the formation of a physical barrier on the digestate surface.The physical cover of untreated BCs particles prevented the digestate from being exposed to the surrounding environment and slowed down the transfer of NH3-N gases at the digestate-air interface.The effectiveness of untreated BCs declined after 1.5 h possibly due to its strong water repellent characteristic.The hydrophobic property prevented the BCs particles from submerging deeper into the digestate, and the resulting BC cover was less resistant to NH3-N emissions.This strong physical property was probably induced by the large quantity of aliphatic functional groups on the BC surface (Figure 3).Previous studies have shown a strong correlation between hydrophobicity and the presence of aliphatic functional groups (-C-H, CH2) on the BC surface (Kinney et al., 2012;Mao et al., 2019).Our findings are in accordance with those of Holly and Larson (2017) in which BC, made from wood and maize cob pyrolyzed at 400 °C, effectively mitigated NH3-N emissions from the dairy manure storage.The effectiveness of BC in their experiment was due to the action of the physical barrier on the manure surface provided by the BC cover thickness of 3.1 and 2.5 cm.NH3-N volatilization in BCm treatment occurred probably because of the pH increase in the digestate caused by aeration when the BC was mixed with the digestate.In the BCs treatment, the NH3-N emissions and pH under the BC cover were notably lower than in control because the digestate was not in direct contact with ambient air.The reduced effect of the BCs treatment might also be because the relative change of digestate pH only occurred below the BC cover.According to García-González et al. (2015), manure aeration stimulates OH− groups release and increases manure pH, which in turn increases the conversion rate of NH4 + -N to NH3-N.In the BCm treatment, however, the aeration increased NH3-N emissions in the short-term (<3.5 h).As the BCm particles started to migrate to the upper part of the vessel, they formed, like the BCs treatment, a physical barrier against NH3 loss.However, not all BCm particles floated to the digestate surface after mixing to provide resistance to NH3-N emissions.The BCm particles located closer to the center of the vessel decreased the digestate pH in the middle of the vessel and diminished the formation of NH3-N.The pH reduction was furthermore exacerbated by greater contact between BCm particles and digestate.

Effect of acidification on BC characteristics and NH3-N emissions
BC acidification improved the suppression effect of BCs but not of BCm (Figure 2B,C).Acidification with H2SO4, H2O2 and H3PO4 prolonged the reduction effect of BCs by 41.5, 38.5 and 31.5 h, respectively.
Of the pure acids, only H2SO4 and H3PO4 significantly reduced NH3-N emissions.During the 48-h time period, the reduction effect of both acids was statistically significant during the first 5 h and from 28 to 33 h (p < 0.01) (Figure 2A).
The physicochemical characteristics of BC that affected NH3-N emissions were positively influenced by acidification.SSA increased the most in BC+H2O2 and BC+H2SO4, whereas the SSA increase was slightly less in BC+H3PO4.Maximum peaks of hydroxyl (-OH) and carbonyl (-C=O) functional groups were observed in the BC+H2SO4.The peak of aliphatic functional groups was highest in BC+H3PO4, but it was also high on the surface of BC+H2SO4 (Figure 3).The amount of previously mentioned functional groups was a little lower in BC+H2O2.Vithanage et al. (2015) and Cibati et al. (2017) showed that BC treatment with H2SO4 increased the O/C ratio twofold, whereas treatment with H2O2 only increased it 1.5-fold.
These findings indicate the capacity of different acids to generate O-containing functional groups on the BC surface (Shi et al., 2019).The addition of O-H, C=O, C-O and N-H functional groups on the BC surface can decrease its hydrophobicity (Ahmed et al., 2016;Rechberger et al., 2017).Shen et al. (2008) found that acid treatment of BC with HNO3 and H2SO4 enhanced the hydrophilic surface of the BC.This means that BC acidification is an effective method to cause the hydrophobic surface of the BC to become slightly more hydrophilic.Acidification remarkably reduced the pH and alkalinity of BC.The effect of acid-treated BC on digestate pH was comparable, except for H2O2, to the respective pure acid applied to the digestate alone.Our results contrast with those of Huff and Lee (2016), who found that the pH of BC was only slightly reduced by H2O2 because of the weak ability of this acid to create acidic functional groups on the BC surface.
Acidification improved the effect of BCs to suppress NH3-N emissions.As a result of the developed hydrophilic property, acidified BC particles could form a thicker cover on the digestate surface that withstood the pressure exerted by the gases in the digestate for a longer period than untreated BCs.The ability to reduce NH3-N emissions of BCs+H2SO4 or BCs+H3PO4 was slightly greater than that of both acids when applied in the pure form in the first 48 h only (Figure A2).However, the capacity of BCm to reduce NH3-N emissions was not improved by acidification because most of acidified BCm particles could not float to the digestate surface as they did in the untreated BCm treatment.Although the effect of acidified BCm to influence digestate pH was comparable to that of the pure acids, their impact on NH3-N emissions was smaller than that of the pure acids.Such results might be because of the slower capacity of acidified BCm to alter the digestate pH suggested mostly by the great differences in pH and NH3-N emissions between acid and acidified BC treatments observed in the early stage of the experiment (<3.5 h) (Figure 2A,B).The NH3-N emission difference later diminished when digestate pH in acidified BC and acid treatments leveled off.The results of the current study show that the effect of acidified BC on digestate pH remained the same as that of pure acid for at least five days.The pH of digestate where acidified BC was added could increase at a certain moment in time due to the salts present in the ash of acidified BC.Thus, further studies are needed to analyze the effect on pH in longer periods.

Effect of conditioner treatments on N concentration in digestate
Thirty days from the beginning of the experiment, the NH4 + -N concentration was significantly higher (p < 0.01) than that of the control in all conditioner treatments, except H2O2.NO3 − -N concentration was higher than control in the BC+H2O2, BC+H2SO4 and the H2SO4 and H3PO4 treatments (p < 0.001) and Ntot in the H2SO4 and H3PO4 treatments only (p < 0.01) (Figure 4).The Ntot concentration in digestate was not significantly correlated with NH3-N emissions recorded in the first 48 h and between the 7th and 8th days.In addition to NH3-N emissions, the Ntot concentration could be affected by other N-containing gas emissions.Previously, Chadwick et al. (2011) showed that the crust or permeable cover formed on the slurry surface during the storage period reduced NH3 but increased N2O emissions.In addition, in our experiment, the BC cover might have stimulated N2O emissions because of the anaerobic conditions.The conditioners used in the current experiment affected digestate pH and probably also gaseous exchange at the liquid-air interface by the cover formed on the digestate surface.As digestate pH increases, the concomitant increase of NH3 concentration may inhibit the activity of nitrifying bacteria.As a result, NH4 + is converted into N2O and dinitrogen (N2) gases bypassing the NO3 − -N phase (Soliman and Eldyasti, 2018).
In the BCs treatments, N2O emissions were probably not the main factor influencing the NO3 − -N concentration in digestate.The NO3 − -N concentration was higher in acidified BCs than in untreated BCs, even though a longer-lasting acidified BCs cover and slightly lower digestate pH should promote denitrifying bacteria activity (Chadwick et al., 2011;Kupper et al., 2020;Šimek et al., 2002).It is possible that NO3 − -N was adsorbed by acid-treated BC, and its degradation by microorganisms became more difficult.Lan et al. (2017) noted that BC adsorption of NO3 − in soil can decrease its availability for denitrifiers.The NO3 − -N concentration between acidified BCs and BCm with H2SO4 or H2O2 was not different, although the cover was not formed in acidified BCm.NO3 − -N sorption could be promoted by surface basic functional groups such as primary and secondary amine groups (Figure 3) via electrostatic interactions.This is in accordance with the findings of Wu et al. (2019) that NO3 − can interact with amine groups through electrostatic interactions during the adsorption process.In BC+H3PO4, the NO3 − -N could be limited because of the presence of competitive phosphate anions (PO4 3− ) bound by sorption sites during acidification, which could remain immobilized because of their highly negative charge.A small amount of NO3 − -N could be adsorbed to positively charged cations (K + , Ca 2+ and Mg 2+ ) present in the ash of acidified BC by bridge bonding.Fidel et al. (2018) mentioned that some NO3 − -N sorption to acidified BC can occur via cation bridging.However, NO3 − -N could not be bound by untreated BC primarily because of the lower peaks of basic functional groups.Alsewaileh et al. (2019) found that low-temperature BC (300 °C) exhibited minimal adsorption efficiency of NO3 − -N because of the reduced total basicity (surface basic functional groups) and limited surface area.

CONCLUSION
In this study, H2SO4 and H3PO4 acids, untreated BCm and acidified BCs treatments were found in reducing NH3-N emissions from digestate.Acids reduced the digestate pH and reduced the rate of conversion from NH4 + -N to NH3-N.Untreated BCm formed a physical barrier on the digestate surface that isolated the digestate from the atmosphere.Acidification of BC increased its specific surface area and number of O-containing surface functional groups and decreased the pH, alkalinity and the hydrophobic property.The impact on NH3-N emissions of acidified BC was dependent on the application method.Compared with untreated BC, the ability of BC to reduce NH3-N emissions was greater when it was acidified with H2SO4 and applied to the digestate surface.The novel finding of our study is that acidified BC applied on digestate surface could have an effective application potential to reduce NH3 emissions from slurry storage tanks.
Our results suggest also that BC cover might stimulate N2O emissions because of the anaerobic conditions.In addition, BC+H2SO4 and BC+H2O2 might adsorb NO3 − -N present in digestate and decrease its availability to denitrifying bacteria.These new hypotheses need to be tested in future research.

Figure 2 .
Figure 2. NH3-N emissions (mean ± standard error) during 0-48 h and from the 7th to 8th day after the application of pure acids (A), untreated and acidified BC mixed in the digestate (B) and untreated and acidified BC applied on the digestate surface (C).Control without conditioner.

Figure 3 .
Figure 3. FTIR spectra of untreated and acidified BC with identified functional groups.
Figure A1.Cumulative NH3-N emissions for 0-48 h from conditioner application (mean ± standard error (n = 3).Results marked with different lowercase letters are statistically different.

Table 1 .
Experimental treatment abbreviations and description.

Table 2 .
Characteristics of digestate, untreated and acid-treated BC.