Experimental Study on Single Factor, Orthogonal and Characteristics of SRB Particles Fixed by Biologically Activated Lignite

： Aiming at the problems of insufficient supply of carbon source and toxic effect of heavy metal ions in the treatment of acid mine wastewater (AMD) by sulfate reducing bacteria (SRB), immobilized particles were prepared with Rhodopseudomonas spheroides, SRB and lignite as the main raw materials, and the optimal ratio of immobilized particles was determined based on single factor test and orthogonal test. The adsorption experiment of immobilized particles was carried out under the optimal ratio, the reaction kinetic process and adsorption capacity of immobilized particles for different ions were analyzed, and the action mechanism was studied. The results show that lignite not only has good adsorption performance, but also can be used as the carbon source of SRB after degradation by Rhodopseudomonas spheroides, so as to solve the problems of low removal efficiency of AMD by SRB and insufficient supply of carbon source. When the dosage of lignite, Rhodopseudomonas spheroides and SRB is 3%, 10% and 10% respectively, and the particle size of lignite is 200 mesh, the overall treatment effect is the best. The removal rates of SO 4 2- , Zn 2+ and Cu 2+ are 83.21%, 99.59% and 99.93% respectively, the pH is increased to 7.43, the release of COD is 523 mg/L, and the ORP number is -134 mV. The reduction process of SO 42-by immobilized particles conforms to the pseudo first-order kinetics, the isothermal adsorption of Zn 2+ more conforms to the Freundlich isothermal adsorption equation, and the adsorption kinetics of Zn 2+ more conforms to the pseudo second-order kinetic model.


Introduction
Acid mine wastewater is produced with mining activities. Because of its acidity and high concentration of heavy metal ions, it will have a serious impact on the environment 1,2 . At present, the most commonly used methods to deal with AMD mainly include physical method, chemical method and biological method. Physical and chemical methods have problems such as high cost or unsatisfactory treatment effect, so microbial method has been widely studied. SRB is the most commonly used, but in the actual treatment process, due to the inhibition of AMD low pH, the toxicity of heavy metal ions and the lack of carbon source in the later stage, the treatment effect of SRB is not ideal 1 . Lignite has large specific surface area and rich oxygen-containing functional groups. Heavy metal ions are easy to combine with oxygen-containing functional groups on the surface of lignite; As a coal dissolving bacterium, Rhodopseudomonas spheroides can degrade lignite for SRB for growth and metabolism. Therefore, using the biodegradable lignite as the carbon source of SRB can improve the pH of water and reduce the toxicity of acid and heavy metal ions to SRB [2][3][4] . In this experiment, biologically activated lignite was used as SRB carbon source to explore the optimal ratio of immobilized particle composition of mineral carbon source in the treatment of AMD by immobilization technology, and the reaction kinetic process and adsorption capacity of immobilized particles to SO4 2-and Zn 2+ were analyzed, so as to provide a strong basis for selecting minerals as carbon source for SRB treatment of acid mine wastewater. mesh, the removal effect of Zn 2+ is the best; When the particle size of lignite is 200 mesh, it has the best effect on increasing the pH of the solution; Lignite with particle size of 200 mesh has the best effect on stimulating SRB activity.

Analysis of The Results of Orthogonal Experiment
The orthogonal test was carried out on the basis of the optimal dosage of each factor of immobilized particles obtained from the previous single factor test (the dosage of Rhodopseudomonas spheroides was 30%; the dosage of lignite was 5%; the dosage of SRB bacteria was 30%; and the particle size of lignite was 200 mesh).
The range of a factor in the orthogonal test results is positively correlated with the effect of affecting the test results. The greater the range, the greater the influence of the influencing factor on the test results [7][8][9] . According to Table 1, the order of influence of various factors on SO4 2-removal effect is: A > D > C > B. The analysis of variance is mainly carried out with item B, as shown in Table 2 and Table 3. The P value of the four factors and their influence in this experiment is very low (P < 0.05), indicating that the four factors have no significant correlation with the removal effect of SO4 2-. According to the mean value of four factors 10 , when the combination type of immobilized particles is A1B1C1D3, the removal effect of SO4 2-is the best.
According to Table 1, the order of influence of various factors on Cu 2+ removal effect is: A > D > C > B. The analysis of variance based on item B is shown in Table 2 and Table 3. The results show that the removal effect of factor A on Cu 2+ is significant, and factors B, C and D are not significant. According to the mean value of the four factors, when the combination type of immobilized particles is A1B2C2D3, the removal effect of Cu 2+ is the best.
According to Table 1, the order of influence of various factors on Zn 2+ removal effect is: D > A > B > C. The analysis of variance based on item C is shown in Table 2 and Table 3. The results show that the removal effect of factors A, B and C on Zn 2+ is not significant, and factor D is significant. According to the mean value of the four factors, when the removal effect of Zn 2+ is the best, the optimal type of immobilized particle combination is A1B3C1D3.
According to Table 1, the order of the influence of various factors on the pH enhancement effect is: A = C > D > B. The analysis of variance based on item B is shown in Table 2 and Table 3. The results show that the four factors have no significant correlation with the effect of pH improvement. According to the mean value of the four factors, the optimal combination type of immobilized particles with the best effect on pH improvement is A3B3C3D3.
According to Table 1, the order of influence of various factors on COD release is: A > D > B > C. The analysis of variance based on item C is shown in Table 2 and Table 3. The results show that factors A, B and D show significant differences in COD release, while factor C shows no significant differences. According to the mean value of the four factors, the optimal combination type of immobilized particles affecting cod release is A3B1C2D3.
According to Table 1, the order of influence of various factors on ORP value change is: A > D > B > C. The analysis of variance is mainly carried out with item C, as shown in Table 2 and Table 3. The experimental results show that factors A and D show a significant level on the change of ORP value, while factors B and C show a non significant level. According to the mean value of the four factors, the best combination type of immobilized particles affecting the change of ORP value is A1B1C1D1.

Physical Properties of Particles
The immobilized particles were prepared according to the best composition ratio obtained by orthogonal test. The particle morphology before and after AMD treatment is shown in Supplementary Figure S1. Before the treatment of AMD with immobilized particles, the appearance was spherical with smooth surface, hard texture, no obvious agglomeration, and the inner black core was faint under the white coat; After AMD treatment, the appearance of the immobilized particles presents a full spherical shape, but the texture becomes softer and more moist, and the black core of the particles is clearly visible 11 .

Kinetics of SO4 2-Reduction
As the reaction continues, the overall change trend of SO4 2-removal rate is gradually increased at first and then stabilized, and the SO4 2-concentration in the system gradually decreases to a stable level. When the reaction lasted for 1 day, the reduction rate of SO4 2-was 12.283 mg/(L·h); The average reduction rate of SO4 2-in 1-4 days was 5.839 mg/(L·h); When the reaction reaches 5d, the reduction rate of SO4 2-is only 0.840 mg/(L·h)(Supplementary Figure S2(a)). SRB is a microorganism that uses sulfur oxide such as sulfate as electron receptor to obtain energy supply 12 . The test results show that SRB immobilized particles have strong adaptability to the environment, and SRB can quickly enter the state of high growth rate and vigorous metabolism 13 . The fundamental reason is that at the initial stage of the reaction, Rhodopseudomonas spheroides degraded lignite into organic matter that can be utilized by SRB [14][15] . The rich organic matter in the system promoted the growth of SRB, enhanced microbial SRB activity and promoted the removal of SO4 2-5 ; Lignite can also adsorb heavy metal ions and H + in the system, reduce environmental load and reduce the stimulation of metal ions and peracid conditions on SRB. According to the pH change curve of the solution, the environment of the system in the early stage of the reaction is too acidic. When the pH is between 7-8, the SRB activity is the best, and both acid and alkali will inhibit the SRB activity. Therefore, the growth and activity of SRB in the early stage of the reaction are inhibited, which is consistent with the results of this test; When the pH value gradually increases to 7-8, the SRB activity is the best. However, with the continuous occurrence of the reaction, the nutrients in the system are gradually consumed, and the SO4 2-content is gradually reduced due to the reduction, so that the SRB growth rate and activity are slowed down in the later stage of the reaction, and the removal efficiency of SO4 2-is reduced 6 , and the sediment produced by the early reaction is attached to the particle surface, It blocks the internal channel of the particles and affects the ion exchange reaction between the particles and the outside 16 , so the SRB activity is inhibited and the removal effect of SO4 2-is reduced.
The adsorption kinetic process of SO4 2-by immobilized particles was analyzed by zero order kinetic model and first-order kinetic model respectively, and See equations (1) and (2) for the expression: In the formula: 0 C --Initial concentration of SO4 2-, mg/L; t C --The concentration of SO4 2-at t , mg/L; 0 k --Zero order reaction rate constant, mg/L -1 ·h -1 ; 1 k --First order reaction rate constant, h -1 .
According to the fitting results( Fig. 5(a, b)), we can know the SO4 2-reduction kinetic parameters. According to the SO4 2-reduction kinetic fitting parameters, for the linear correlation coefficient R 2 , the first-order reduction kinetic model is greater than the zero order reduction kinetic model, indicating that the first-order reduction kinetic model can more accurately describe the SO4 2-reduction reaction process than the zero order reduction kinetic model, The linear correlation coefficient of the first-order reduction kinetic model was R 2 =0.98821. It shows that the reduction process of SO4 2-is mainly electron receptor 17 .
According to Fig. 5(c), with the progress of the reaction, the pH first increased and then stabilized, and finally showed weak alkalinity, pH=8.1. The initial pH of the system is 4.0. The peracid reaction conditions are not conducive to the growth and metabolism of SRB, and the SRB activity is low. At this time, the pH rise of the solution is mainly due to the rich ionizable groups (RCOOH⇌ RCOO -+H + ) in the lignite inside the immobilized particles, which has a good two-way pH regulation ability 18 . Therefore, the high concentration of H + in the system before the reaction will inhibit the dissociation of active groups, the dissociation equilibrium moves reversely, and the pH value of the solution gradually increases. At this time, the adsorption is mainly physical adsorption, mainly ion exchange 19 ; However, the ability of lignite to adjust pH is limited, so when the pH value rises to a certain extent and the physiological activity of SRB is no longer inhibited, SRB begins to actively carry out SO4 2-reduction reaction. At this time, the root cause of the decrease of H + concentration in the system is the biochemical reaction of SRB. The H + in the solution provides electrons for SRB to reduce SO4 2-and improves the pH of the solution to a great extent. Until the later stage of the reaction, although the pH of the system is very suitable for the growth and metabolism of SRB, the concentration of SO4 2-in the system is low and the content of electron receptor is insufficient, resulting in the ability of SRB to reduce SO4 2-was reduced and the demand for H + , so the pH value of the system finally tends to be stable.
The overall change trend of COD release is that it increases gradually with the reaction, decreases after a period of reaction, and then increases gradually. The final release of COD is 700 mg/L. The above changes in COD release are related to the leakage of organic matter released from lignite and SRB biological metabolites in immobilized particles 16 . At the initial stage of the reaction, Rhodopseudomonas spheroides degraded the lignite, so that the lignite was degraded into organic matter, and a large amount of organic matter leaked out. At the initial stage, SRB was not enough to make use of such rich organic matter, so the organic matter in the solution gradually accumulated and the COD content increased. With the progress of the reaction, the activity of SRB and the ability to use organic matter are enhanced. The lignite degraded into organic matter provides carbon source for the growth of SRB, promotes the growth and metabolism of SRB, and the organic matter in the solution is gradually consumed, so the organic matter content shows a downward trend. When the reaction proceeds to the later stage, more sediments are attached to the lignite surface, which makes the lignite degradation process more difficult, the organic carbon source available for SRB growth and metabolism in the system is reduced, the SRB activity and the ability to utilize organic matter are limited, and the COD in the system begins to accumulate again.
The overall change trend of ORP value is: It gradually decreases and then tends to be stable, and the final ORP value of the system was -136 mV. The lower the ORP value, the better the bacterial activity in the system. The test results showed that the SRB activity in the reaction system gradually increased until the SRB activity remained unchanged. At the initial stage of the reaction, lignite is both a carbon source and an electron donor of SRB 20 . Lignite degraded by Rhodopseudomonas spheroides provides nutrients for SRB growth and metabolism and enhances SRB activity; Lignite can also be used as an electron donor of SRB to further enhance SRB activity, so the ORP value decreases. In the later stage of the reaction, due to the limitation of the content of carbon source and terminal electron receptor, the proliferation activity of SRB is inhibited and the activity is low. Therefore, the ORP value tends to be stable, and the gradual adsorption saturation of lignite to H + is also one of the reasons.  Figure S2(b)), and the adsorption rate of Zn 2+ by particles in the early stage is fast, and then the rate gradually slows down. When the equilibrium concentration is 8.6175 mg/L, the equilibrium adsorption capacity of Zn 2+ by particles is 0.759 mg/g. Langmuir and Freundlich 21 isothermal adsorption equations are used to fit the isothermal adsorption process of Zn 2+ respectively. See equations (4) and (5) for isothermal adsorption equation: In the formula: b --Constant, which is related to the enthalpy of adsorption reaction, expressed as reciprocal of concentration; k --Freundlich adsorption coefficient; n --Constant, which indicates the relative size of surface heterogeneity and adsorption strength of adsorbent, is usually greater than 1; max q --Maximum theoretical adsorption capacity of Zn 2+ on particles, mg/g.

(2)Analysis of fitting results
According to the fitting results ( Fig. 5(d, e)), the isotherms of the two isothermal adsorption types are regular, non negative and concave to the concentration axis, and the adsorption reaches equilibrium at high concentration 17 . For the linear correlation coefficient R 2 , both Freundlich isothermal adsorption equation and Langmuir isothermal adsorption equation are greater than 0.95, indicating that the adsorption of Zn 2+ on the particle surface is mainly chemical adsorption, and for the adsorption correlation coefficient, Freundlich isothermal adsorption equation is greater than Langmuir isothermal adsorption equation. Therefore, Freundlich isothermal adsorption equation can more accurately describe the adsorption process of Zn 2+ in AMD wastewater than Langmuir isothermal adsorption equation. The detailed analysis of the parameters of Freundlich isothermal adsorption equation shows that Freundlich isothermal adsorption equation is used to describe the heterogeneity of adsorbent surface and the adsorption strength of adsorbent 1/n=0.41701, 0.1＜1/n＜0.5. The results show that the immobilized particles have strong heterogeneity 22 and the Zn 2+ adsorption process on lignite surface is easy to occur 23-24 .

Adsorption Kinetics of Zn 2+
Langergren equation (pseudo first-order kinetic model) and McKay equation (pseudo second-order kinetic model) are used to study the adsorption kinetic process of Zn 2+ by immobilized particles. See equations (6) and (7) for the expression: According to Fig. 5(f, g), the adsorption capacity of Zn 2+ by immobilized particles gradually increases with the progress of the reaction until it is stable. When the adsorption reaches equilibrium, the equilibrium adsorption capacity of Zn 2+ by immobilized particles is 0.2 mg/g. The kinetic model is fitted based on the adsorption curve model, and the primary and secondary kinetic fitting diagrams of ln(qe/(qe-qt))~t and t/qt~t are drawn respectively ( Fig. 5(h, i)).
In this experiment, in the reaction stage with fast adsorption rate of Zn 2+ , chemical precipitation and biological adsorption exist at the same time, and the first process is chemical precipitation; In the later stage, the adsorption reaction of Zn 2+ is mainly biosorption. Chemical precipitation works through lignite in the particle. Lignite adsorbs a large number of positively charged metal ions in the solution by virtue of its developed specific surface area and rich oxygen-containing functional groups; SO4 2-is reduced to sulfide by SRB and combined with the remaining Zn 2+ not adsorbed by lignite in the solution to form precipitation, which is removed.
The fitting results show that the pseudo second-order kinetic regression coefficient is greater than the pseudo first-order kinetic regression coefficient, and the pseudo second-order kinetic regression coefficient R 2 =0.99992, and the correlation is very high. In this case, the pseudo second-order kinetics can more accurately describe the adsorption behavior of immobilized particles on Zn 2+ 25 . It shows that the adsorption process of Zn 2+ by immobilized particles is dominated by chemical adsorption, and the influence of chemical adsorption is higher than that of mass transfer step control [26][27][28] .
The pH value of the solution increased first and then tended to be stable as a whole (Fig. 5(j)). Lignite in the particle contains rich active groups. The dissociation of active groups (RCOOH⇌ RCOO -+H + ) makes the lignite surface negatively charged. At the initial stage of the reaction, the pH of the solution is 4, indicating that there is a large amount of H + in the solution, and the negatively charged lignite on the surface is very easy to attract positively charged H + , resulting in the decrease of H + content in the solution, so the pH value of the solution increases; Moreover, as the pH of the solution gradually approaches neutral, the SRB activity in the system also increases. In addition to the adsorption of lignite on H + , the process of SRB reducing SO4 2-has also become a way to consume H + , and the pH value of the solution gradually increases with the decrease of H + concentration. As shown in the test results, the pH of the solution can be increased rapidly when the two reactions act at the same time. However, with the continuous occurrence of the reaction, the spare sites on the lignite surface are gradually occupied by metal ions and H + , resulting in the reduction of its adsorption capacity. Moreover, the precipitates produced by the early reaction adhere to the lignite surface, blocking the internal channels of lignite and further affecting the adsorption capacity of lignite. Therefore, the pH value of the later system basically does not change. In addition, in the early stage of the reaction, the pH of the solution has been raised to neutral, the H + concentration in the solution is low, and the ability to provide electrons for SRB is weak. Therefore, the H + concentration in the system basically does not change significantly, and the pH of the solution tends to be stable.
The overall change of COD shows a trend of first increasing, then decreasing, and then increasing. Finally, the cumulative release of COD is 1090 mg/L. The organic matter released from lignite and the leakage of SRB biological metabolites in immobilized particles are the reasons affecting the change of COD release. The gradual increase of COD content in the initial stage of the reaction is due to the degradation of Lignite by Rhodopseudomonas spheroides, which releases a large amount of organic matter. At the initial stage, SRB activity is low and all organic matter can not be used quickly, resulting in the overflow of some organic matter, the increase of organic matter content in the solution and the gradual increase of COD. As the reaction proceeds, SRB begins to adapt to the current living environment and uses the degraded lignite as the carbon source for its own proliferation. Sufficient carbon source in the system can accelerate the growth and metabolism of SRB and enhance the activity of SRB, while organic matter can be utilized by SRB, so the release of COD in the solution is reduced. In the later stage of the reaction, the growth, metabolism and activity of SRB are affected by the content of carbon source and terminal electron receptor, resulting in the limited utilization of COD by SRB. Therefore, COD accumulates again and the release increases.
The ORP value first decreased and then stabilized with the increase of reaction time. The final ORP value of the system was -106 mV. The results showed that the SRB activity in the reaction system increased gradually and remained unchanged. Combined with the pH change diagram, it can be seen that due to the dissociation of active groups, the lignite surface inside the particles is negatively charged, and has obvious adsorption on positively charged heavy metal ions and H + , so it reduces the toxic effect of acid and heavy metal ions on SRB in AMD, enhances SRB activity, and the ORP value of the system decreases; Moreover, under the degradation of Rhodopseudomonas spheroides, lignite also acts as a carbon source for SRB proliferation, provides nutrients for SRB growth and metabolism, further enhances SRB activity and reduces ORP of the system. At the later stage of the reaction, the solution environment has been neutral and lignite has basically reached adsorption saturation, and the change of H + concentration in the solution is not obvious, indicating that there are less electrons in the system for SRB reduction of SO4 2-process. At this time, SRB activity is low, and ORP value basically does not change. The change trend of ORP value in the experimental results is consistent with the change law of pH.

Conclusion
(1)The single factor experiment shows that when the dosage of Rhodopseudomonas spheroides is 30%, the dosage of lignite is 5%, the dosage of SRB is 30% and the particle size of lignite is 200 mesh, the biological activated lignite fixed SRB particles has the best effect on AMD treatment.
(2) Based on the single factor experiment of the ratio of immobilized particles in the early stage, based on the orthogonal experiment, combined with the changes of SO4 2-, Cu 2+ , Zn 2+ removal rate, pH improvement effect, COD release and ORP value in the solution, it is finally determined that the best combination of biologically activated lignite immobilized SRB particles is A1B1C1D3. That is, the dosage of Rhodopseudomonas spheroides in the immobilized particles is 10%, and the dosage of lignite is 3%, SRB dosage is 10% and lignite particle size is 200 mesh. Under this ratio, the removal rates of SO4 2-, Zn 2+ and Cu 2+ by immobilized particles were 83.21%, 99.59% and 99.93%, the pH was raised to 7.43, the COD release was 523mg/L and the ORP value was -134mV.    Fitting curves

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download.