Toxicological Evaluation of Biological and Electrochemical Treatments of Coal Mine-impacted Water (MIW) on Duckweed Landoltia Punctata

Two different coal mine-impacted water (MIW) treatments (biological via biostimulation of sulfate- 16 reducing bacteria (SRB), and electrocoagulation (elC)) were proposed, reaching efficiencies of up to 17 99.79% in relation to SO 42- , Fe, Mn, and Al ions, as well as acidity removals. Thus, toxicological 18 assays with duckweed Landoltia punctata were performed, in order to verify the safeness and 19 usability of the two treated waters. Therefore, duckweeds were exposed to different dilutions (0, 25, 20 50, 75, and 100% of samples) of the two treated waters, and the growth (r) and inhibition of growth 21 ( 𝐼 𝑟 ) rates were calculated, based on 50% effect concentration (EC 50 ). The water from the biological treatment (microcosm assay) presented the highest toxicity (EC 50 = 33.42%), even higher when 23 compared to the raw MIW (EC 50 = 42.78%), probably due to the hydrogen sulfide, that even after a 24 purge removal, remained in solution. The results showed that this water, despite being within the 25 standards in physicochemical terms, demonstrated risks in terms of toxicity. The water from 26 electrocoagulation (elC) treatment, in the opposite way, showed much less toxicity, even lower than 27 the control, and therefore not reaching EC 50 , also suggesting a possible nutrient function of the treated 28 water. Consequently, the treated water by elC could, for example, have a non-potable use. The study 29 made it possible to prove the efficiency of elC treatment, the importance of post-treatment 30 toxicological assessments, and the potential of the duckweeds as an option for a test organism in these 31 types of evaluations.

treatment (microcosm assay) presented the highest toxicity (EC50 = 33.42%), even higher when 23 compared to the raw MIW (EC50 = 42.78%), probably due to the hydrogen sulfide, that even after a 24 purge removal, remained in solution. The results showed that this water, despite being within the 25 standards in physicochemical terms, demonstrated risks in terms of toxicity. The water from 26 electrocoagulation (elC) treatment, in the opposite way, showed much less toxicity, even lower than 27 the control, and therefore not reaching EC50, also suggesting a possible nutrient function of the treated 28 water. Consequently, the treated water by elC could, for example, have a non-potable use. The study 29 made it possible to prove the efficiency of elC treatment, the importance of post-treatment 30 toxicological assessments, and the potential of the duckweeds as an option for a test organism in these 31 types of evaluations. AMD formation results from pyrite oxidation, through several chemical and biological processes that 38 generate an effluent which is highly acidic (pH 2-3), with high sulfate (SO4 2-) and metallic ion (e.g., 39 Fe yielding high removals of sulfate and metallic ions, as well as pH alkalization, by biostimulating 45 sulfate-reducing bacteria (SRB), using shrimp shell waste as substrate. Electrocoagulation (elC) has 46 also been tested, achieving sulfate removal efficiencies of up to 70.95%, as well as pH neutralization 47 (Rodrigues et al. 2020a).

48
The biota in aquatic environments with lower pH than its tolerance levels can die due to 49 respiratory and osmoregulatory disorders, compromising the food chain (Netto et al. 2013 added to the flasks, the MIW was submitted to a N2 purge, until it reached anoxia (DO ≤ 0.5 mg•L −1 , 101 monitored with an oximeter reading), then it was added to the flasks with the aid of a peristaltic pump 102 (to avoid oxygenation). The microcosms flasks were purged with N2 before and after the MIW was 103 inserted, sealed with a silicone stopper, kept in a dark room at 20 ± 1 °C (controlled with a wall 104 thermometer) for 41 days of incubation, being shaken manually once a day (to ensure homogeneity 105 of the flask contents).

106
As a result of the sulfate reduction by the SRB, the hydrogen sulfide (H2S) started to 107 accumulate, and to remove it as gas, at the end of the microcosm period of incubation, an N2 purge 108 system was assembled. The H2S is a weak diprotic acid, and its form depends directly on the pH (H2S, 109 HS − , and S 2-), its neutral form being partially soluble in water and toxic gas. Inside a fume hood, the 110 flasks were purged with N2, and the outgoing gas flow (H2S + N2) was bubbled into a NaOH solution, 111 generating sodium sulfide (Eq. 1), thus avoiding leakage of the toxic gas to the outside. Each flask 112 was purged for 30 minutes (four flasks per time, with a four-way manifold splitter, Fig. 1). After this 113 sulfide removal process, the microcosms contents were filtered (also inside a fume hood), 114 characterized, and submitted to toxicological assay.

Electrocoagulation (elC) assay for MIW treatment 124
An electrochemical system to treat MIW was carried out in bench-scale, as performed 125 previously (Rodrigues et al. 2020a). The system consisted of duplicates of reactors (1-L plastic 126 beaker), in which flat plate electrodes of Al (anode) and stainless steel (cathode) were immersed, 127 spaced 5 cm from each other. The electrodes had the following dimension: 5.65 x 13.9 cm, with a 128 useful area of 28.76 cm 2 (anode). Magnetic stirrers were used during the process to homogenize, since 129 a chemical species concentration gradient naturally occurs. A control panel regulated the electric 130 current from the power supply (PS-A305D), providing a 65 A•m -2 current density, in continuous mode 131 of exposure (of electric current) that goes into each elC reactor (Fig. 2). In this batch assay, 1 L of 132 MIW was inserted in each beaker, the room temperature was controlled and kept at 23 ± 1 °C, and 133 the total electric current time was 5 hours. After this period, the content was filtered for 134 characterization, and submitted to toxicological assay.

151
Four different assays were performed for comparative purposes: MIW after biostimulation, 152 MIW after electrocoagulation, raw MIW, and raw MIW with corrected pH (with NaOH solution until 153 pH=7). The latter assay was included because the MIW pH is acid, and duckweed needs a pH between 154 5-9. 155 The MIW samples were diluted with culture medium, and for the analysis of toxic effects, 156 dilution factor (DF) was used, being 0% (control -only culture medium), 25% (1:4, i.e., 1 part of 157 crude sample diluted in 3 parts of culture medium), 50% (1:2), 75% (1:1.333), and 100% (gross 158 sample -1:1) (Iatrou et al. 2015), as detailed in Table 2, with a total of six replicates for each dilution. 159 The total volume was 100 mL in all cases. 160 161 All experiments were conducted in 100 mL-beakers, and each one was inoculated with 164 duckweed and incubated in a temperature-controlled incubator (25 ± 2 °C) under an 18h-continuous 165 illumination with fluorescent lamps (photoperiod). The pH was adjusted to the range of 6.5 to 7, using 166 HCl or NaOH, except for raw MIW. The test started ( 0 ) with a total of ten healthy fronds Where is the average specific growth rate of the control, and the average specific treatment 175 growth rate to each DF tested. For the EC50 determination, the values were plotted against DF, and 176 the regression of this concentration-response curve was performed. EC50 is defined as the sample 177 concentration where 50% of effect is observed, when compared to the control. In this case, the effect 178 is the growth inhibition ( ).

179
The values of the parameters (growth rate and growth inhibition) were calculated through mean 180 and standard deviations. The significant differences between the means of treatments and control 181 samples were obtained from analysis of variance according to the Tukey test, and performed through 182 Statistica (v.10, 2011) software.

MIW characterization and treatments 185
In the biological treatment (SRB biostimulation), where there was sulfate reduction activity, 186 sulfate and acidity were removed, as well as metallic ions (Fe, Al, and Mn). Similarly, for the elC 187 treatment, the same parameters were also removed, and Table 3   2018), likely being the reason that its concentration was almost seven times above the MAV (Table  210 3). It is worth mentioning that the aluminum is released continuously by the anode as a coagulant 211 agent. In this assay, for operation in a non-reducing atmosphere (differently from the microcosms 212 assay), no sulfide was formed, thus, the precipitation is caused by hydroxides, (bi)carbonates, and 213 complexed sulfates (Rodrigues et al. 2020a). 214 Subsequently, the MIW from microcosm and elC treatments were submitted to toxicological 215 assay, in order to ensure safeness and quality of the treated effluent. 216 217 both treatments (biological and elC), raw MIW, and corrected pH MIW. From these results it is 220 possible to observe that the MIW after biostimulation (Fig. 3A), raw MIW (Fig. 3C), and corrected 221 pH MIW (Fig. 3D) presented the same pattern: significant different (and lower) growth rates 222 compared to the control (0 DF), evidencing therefore, that they have considerable toxicity, since 223 growth decreased as the concentration (DF) increased. For specific cases of SBR biostimulation, the 224 growth is zero from 50% of DF. The differences in the growth rate for the four controls are attributed 225 to small differences in the medium composition, as the four experiments were carried out on different 226 days with freshly prepared solutions. 227 228 Fig. 3 Growth rate of Landoltia punctata after 7 days exposed to different effluents and dilution Otherwise, in relation to the elC treatment results (Fig. 3B), for all concentrations (except for 234 the 50% DF), they presented no significant difference from the control, as shown by the same letter 235 "a" from the ANOVA study (Fig. 3B), therefore low toxicity and good quality of this effluent is 236

Toxicological assay for growth and inhibition rate 218
inferred. It is important to note a slight increase in growth rate for the elC effluent in the 100% DF 237 (gross sample). The hypothesis that this sample has some element that may have stimulated the 238 growth of duckweed is raised. The ability of an Mn ion to act like a micronutrient when in adequate 239 concentrations, stimulating the chlorophyll formation, and intervening in the production of enzymes 240 is well known. Enzymes play an important role in protein metabolism and cellular division 241 (Chatzistathis et al. 2011;Soiltech 2021). 242 In relation to the inhibition of growth rate, most of the results (Fig. 4A, C and D) showed the 243 same toxicological trend: as the DF increased, the inhibition also increased, especially for the SRB 244 biostimulation treatment (Fig. 4A), which increased at a higher rate than the other samples. The raw MIW (Fig. 4C) also presented considerable toxicity as the concentration increased, 254 with significant difference between the inhibition rates at the different DF. In the case of corrected 255 pH MIW (Fig. 4D), its inhibition rate was also raised at a higher DF, but in a smoother way. Between 256 these two last treatment results (raw MIW and corrected pH MIW), the effect of a simple pH 257 correction (to 7) in toxicity is highlighted, revealed by their EC50 value (42.78% and 92.37% of DF, 258 Table 4). This is coherent with the ideal conditions for duckweed development (minimum pH of 6.5) 259 (OECD 2006; ISO/DIS 20079 2010). The main factor associated with toxicity was evidenced by pH 260 correction. This fact was already expected, as coal mines in the region impact water resources. 261 262 duckweed (Lemna minor) when exposed to AMD. 269 In the opposite way, for the elC toxicological experiment (Fig. 4B), no statistical differences 270 were observed in the inhibition of growth rates with increasing DF, even presenting a slightly negative 271 inhibition of growth rate for the gross sample (100% of DF), corroborating the earlier graph (Fig. 3).

272
As it did not reach a 50% inhibition rate, it was not possible to determine the 50% effect concentration In the present study, since a slight stimulus in the growth of plants exposed to the effluent 277 from the elC treatment was observed, a sufficient removal of metal ions ceasing to be toxic and 278 behaviour like nutrients can be inferred. As macrophytes need nutrients for their development (Teles 279 et al. 2017), the results suggest that the treated effluent may have a nutrient function for the duckweed. 280 The SRB biostimulation treatment showed toxicity in Landoltia punctata, reaching 100% 281 inhibition with only 50% of DF (Fig. 4A)  Due to the pH of the microcosm (7.19) being close to the 1 (6.9), the 2 ( ) remained 291 partly in solution in equilibrium with H + (aq) + HS − (aq) (Eq. 4), making its escape to the gaseous 292 phase difficult (Eq. 5) even with the use of N2 as a purge gas. 293 Conventional treatment using electrocoagulation proved to be efficient in reducing toxicity, and 294 the treatment performed by biostimulation, under the conditions of the experiment, proved to be 295 ineffective. However, the correction of MIW pH proved to be efficient in reducing toxicity. Despite 296 the low efficiency of biostimulation, the results show that it presents a good opportunity for studies 297 and that further research can be carried out to improve the method. 298 Duckweed species are sensitive to extreme environments (Wang 1990), and the results obtained 299 in the present study demonstrate this, probably due to the presence of hydrogen sulfide, known to be 300 toxic even to human beings (APHA 2017). It should also be noted that the odor resulting from this 301 treated effluent is very pungent, detracting therefore from a non-potable secondary reuse application, 302 since it would be impracticable to use it (for example, for garden irrigation, sidewalk washing, etc). 303 304

Conclusions 305
According to the toxicological evaluation performed, among the proposed MIW treatments, 306 the BRS biostimulation (microcosm) and the elC, the latter evidenced no toxicity, even presenting 307 nutrient potential for duckweed. However, the effluent from biological treatment, which in 308 physicochemical terms showed even greater removal of sulfate and metallic ions, presented high 309 toxicity, even higher than raw MIW. This is probably due to the residual hydrogen sulfide that 310 remained, even after purging, because of the pH. In this sense, the quality of MIW treated in the elC 311 assay was superior to the assay treated by biostimulation. This supported the fact that the elC treated 312 effluent does not present odor and requires treatment of only a few hours, showing its potential for 313 non-potable use purposes, conferring a use for a water initially highly polluted, for instance for 314 irrigation, due to its suggested nutrient function. The findings also led to the conclusion that despite 315 the physicochemical parameters of the effluent from biological treatment having met the requirements 316 of the legislation for the evaluated parameters, the effluent does not meet the toxicity standards, which 317 reveals the importance of these assessments. In addition, the macrophytes proved to be an interesting 318 organism in these types of evaluations. 319 Furthermore, in future tests, scanning electron microscopy (SEM) and transmission electron 320 microscopy (TEM) images from the duckweed organelles tissues will be carried out, in order to 321 visualize possible damage from the exposure, and complement the toxicological evaluation, 322 providing more information about the mechanism of the toxicity observed. 323

324
Declarations 325 326 Conflict of interest: The authors declare that they have no conflict of interest. 327 Ethical approval: Not applicable. 328 Consent to participate: Not applicable. 329 Consent to publish: Not applicable. 330 Availability of data and materials: Not applicable. 331 332