Dynamics and the interaction effects of trace elements and volatile fatty acids concentrations on methanization processes during thermophilic anaerobic digestion

Volatile fatty acids (VFA) and trace elements (TEs) influence methane production and VFA metabolism in positive and negative ways during methanization. There is insufficient knowledge on why the interactions could go either way and this has generated unpleasant consequences for biogas operators. To fill the knowledge gap, statistically designed thermophilic batch experiments were conducted with VFA mixture as substrate to investigate the dynamics and interaction effects of TEs including Nickel (Ni), Cobalt (Co), Selenium (Se) and Molybdenum (Mo) on the methanization processes of CH production and VFA degradation rate. Response surface model and desirability 4 functions were used to determine TEs and VFA interaction effects and the dynamics of the optimum TEs configurations for CH production and VFA degradation rate at different 4 VFA concentrations. The results showed that TEs supplementation influenced Y CH4 production by -14% to 11%: whereas the positive interaction effects of VFA and TEs were by Co*Mo, VFA*Se and VFA*Mo; the interaction effects of Se*Mo and VFA*Co were negative. -15% to 45% increase in Y was obtained with TEs supplementation: Ni*Co VFA-DR interaction produced the positive influences; and Co*Se produced negative effects. The optimum TEs configuration for CH production and VFA degradation rate varied with 4 changes in VFA concentrations.


Introduction
VFA metabolism and CH production 4 Methanization involves anaerobic could be improved by supplementing trace digestion (AD) of biomass to produce energyelements (TEs) to a methanization digester or rich methane (Ch ) (Appels et al., 2008).
performing methanization at thermophilic 4 Common substrates used for methanization temperature (Appels et al., 2008; Mata-Alvarez, include grease trap residue and wastewater; bio-2003). TEs are micro-nutrients that influence the waste from restaurants and households; and expression and dominance of VFA degradation agricultural residues. Irrespective of the pathways, rate of VFA metabolism and methane substrates used, volatile fatty acids (VFA) production (Halarnkar and Blomquist, 1989;Osuna including formate, acetate, butyrate and et al., 2003;Pobeheim et al., 2011). TEs are widely propionate are produced as intermediate reported in literature as being associated with products in the acidification stage of AD. During the enzymology of the acetyl CoA pathway, a acetogenesis and methanogenesis, VFA are dominant pathway during VFA metabolism and metabolized to produce Ch (Appels et al., 2008;4 CH production. Commonly reported TEs used in 4 Mata-Alvarez, 2003). Weak VFA metabolism is a co-supplementation during methanization challenge in methanization and results in VFA include Nickel (Ni), Cobalt (Co), Selenium (Se), accumulation and low CH production (Tiantao et 4 Tungsten (W), Molybdenum (Mo), and Iron (Fe) . al., 2010) . (Gustavsson et al., 2011;Hinken et al., 2008; Experimental design and test system Pobeheim et al., 2010;Zitomer et al., 2008).
The experimental design included five The challenge with the use of TEs for factors consisting of Ni, Co, Se, Mo and VFA enhancing methanization is that the metallomixture, which were expressed in three levels enzymes (MEs) whose biocatalytic potentials are namely low, medium and high concentrations enhanced by TEs supplementation are (Ezebuiro, 2014;Ezebuiro and Koerner, 2017). temperature and TEs specific. For example, 30 implementable experimental runs were anaerobic MEs such as Carbon monoxide selected from the 243 possible experimental d e h y d r o g e n a s e / A c e t y l C o A s y n t h a s e runs following earlier reported statistical (CODH/ACS) complex and methyl transferase procedures and assumptions (SAS Institute Inc., (MeTr) contain TEs in their active sites and are 2012a). The 30 implementable experimental dysfunctional without TEs or when in operation runs were grouped as follows: 3 control units outside the optimum temperature range (R7, R29 and R30) and 27 treatment units (R1 - (Dobbek et al., 2001;Menon and Ragsdale, R27). The control units contained the inoculum, 1999;Svetlitchnyi et al., 2001) . This study is basic nutrient medium and VFA mixture embarked upon because limited knowledge comprising sodium salts of acetic-, propionicabout the interaction effects of TEs, VFA and and butyric acids.  (Ezebuiro, 2014;Ezebuiro and Koerner, 2017). VFA degradation rate (Y ) and CH VFA-DR 4 production (Y ); and CH4 Sample collection and analyses b) Dynamics of the Optimum TEs Liquid (suspension) and gas samples were configuration for co-optimization of Y VFA-collected once in 3 or 4 days during the period of and Y production.
DR CH4 the experiment. The standard methods used for analyses include total VFA concentration (DIN Materials and Methods 38414 -19); pH (DIN 38404 -5); and biogas The materials and methods adopted for volume . CH concentration was measured using 4 the realization of the objectives afore mentioned a GeoTech 5000 gas analyser, Geotechnical are the same as those published for the Instrument, UK, Ltd. mesophilic investigation (Ezebuiro and Koerner, 2017). Highlights of the published materials and The experimental responses methods, and statistical procedures that are The VFA degradation rate (Y ) was measured VFA-DR pertinent to this paper are provided as the change in the concentration of the total subsequently. The inoculum adaptation and VFA on every third day. The average value of the duration of the experiment have not been 3 highest VFA degradation rates for each reactor published earlier and are presented.
was taken to represent its VFA degradation rate. CH production (Nml) (Y ) was measured as 4 CH4 the CH produced every day under standard (D) for maximizing single response (Y or Y ) is shown in Eq. 2a; and the desirability function conditions . For direct comparison of the results (D ) for co-maximizing Y and Y is shown in from the treatments with those from the 1-k VFA-DR CH4 controls, the relative values (Y ) of the Eq. 2b. Details of the desirability functions can relative, r be found in the JMP 10 documentation(SAS responses (Y ) were determined as the ratio of i Institute Inc., 2012b). the treatment value for a response to the control value for the same response. Relative value greater than 1 was considered beneficial; relative value equal to 1 was equivalent to the control; and relative value less than 1 was Where Yi is the response (relative Y or Y VFA-DR CH4 inhibitory or non-beneficial. production) predicted with the RSM shown in Eq. 1; A, B are the lowest and highest values Determination of the optimum TEs configuration respectively of the relative Y or Y

VFA-DR CH4
for thermophilic Y and Y production  production; and w is the value that shows the Response surface methodology (RSM) importance of Y production in relation to Y . procedure can be found in the statistical software Inoculum adaptation and experimental duration (JMP 10) used for the experimental design and The inoculum was the source of microbes analyses (SAS Institute Inc., 2012b). Eq. 1 is the for the experiments. A thermophilic inoculum JMP 10-generated RSM for the factors Ni, Co, Se, was prepared by adapting a mesophilic inoculum Mo and VFA.
as earlier described (Ahn and Forster, 2002). The test system for the adaptation of the inoculum from mesophilic to thermophilic condition was adopted from a standardized procedure (Verein Deutscher Ingenieure, 2006). The duration of the investigation was 39 days in the high and medium VFA levels and 20 days in the low VFA Where, Yi is the predicted response (Y or Y level. optimum TEs configuration produced process production efficiency of 99% (desirability of 0.99) in both Fig. 1a and 1b show the optimum TEs low and medium VFA levels. The VFA profile configuration that maximized thermophilic Y CH4 suggests that 0.2 -1.5 mg/L Ni could maintain production at different VFA levels. Fig. 1a shows the desirability of 0.99 up to VFA concentration of the prediction profiles for the relative Y CH4 175 mmol/L. The Ni profile suggests that production and the influences of the TEs in the changes outside the beneficial range of Ni would low VFA levels. The profiles also apply to the result in decline in Y production. The profiles CH4 medium VFA level (125 mmol/L, not shown). The for Co, Se and Mo suggest that concentrations up range of beneficial TE is 0.2 -1.5 mg/L Ni. The to 0.3 mg/L Co, 0.5 mg/L Se and 0.6 mg/L Mo are optimum TE configuration is 0.9 mg/L Ni and tolerable for 20 -175 mmol/L VFA concentration. resulted in a relative Y production of 1.11 (low VFA-DR TEs in the high VFA level. The ranges of which corresponds to 32% improvement in VFA beneficial TEs are 0.5 -1.5 mg/L Ni and 0.6 -1.5 metabolism and an associated process efficiency mg/L Se. The desirability profiles suggest that of 99% (desirability of 0.99). Whereas changes the TEs configuration of 1.1 mg/L Ni and 1.2 in Mo concentrations are weakly influential, Ni concentrations outside 0.8 -1.6 mg/L and Co mg/L Se is optimum for VFA concentrations concentration > 0 mg/L show antagonism with 200 mmol/L. The TEs profiles for relative Y CH4 VFA concentration 55 mmol/L and induce production suggest that Ni concentrations > 1.5 decline in Y . Conversely, Se concentrations mg/L would result in significant decline in Y VFA-DR CH4 0.5 mg/L show synergy with VFA concentration production; concentrations of Co and Mo > 0 mg/L are antagonistic to Y but up to 1.5 mg/L 55 mmol/L and enhance Y . configurations that simultaneously maximize TEs configurations for simultaneous optimization both responses. Table 2 shows the output of the of thermophilic Y production and Y CH4 VFA-DR multi-response optimization in the different The influence of TEs on Y production CH4 levels of VFA. and Y were co-evaluated to derive the TEs VFA-DR  Furthermore, as shown in Fig. 2a, b and  production, a decrease in Co (3.73 -0.03 during thermophilic methanization. Changes in mg/L) concentration is required for an increase VFA*Co, VFA*Mo, VFA*Se, Se*Mo, Co*Se, in Se (0 -1.24 mg/L) so as to reverse the Co*Mo and Ni*Co shaped the optimum TEs negative interaction from Co*Se. This is requirements for co-optimization of Y and evident in the Co and Se concentration VFA-DR Y production. As a result of VFA*Co, VFA*Mo dynamics in the VFA levels of 10 -100 mmol/L CH4 and 150 -250 mmol/L in Table 2. and Co*Mo, the optimum TEs configurations for enhanced Y production and Y require CH4 VFA-DR Acknowledgement decreasing Co (3.73 -0.03 mg/L) and Mo The scholarship to enable the author to (1.24 -0.21 mg/L) concentration as VFA study and conduct this research at the increased (10 -250 mmol/L). Conversely, the Hamburg University of Technology (TUHH), requirement for Se increased (0 -1.24 mg/L) Germany, was granted by the German in response to the interaction effects of Academic Exchange Programme (DAAD). The VFA*Se, Se*Mo and Co*Se following VFA author is thankful to TUHH for funding the increase (10 -250 mmol/L). laboratory analyses during this investigation.

The dynamics of the interaction effects
The author is also thankful to the Herrling between Co, Se, Mo and VFA are consistent Group Hamburg for providing the inoculum with VFA*Co and Se*Mo having significant used for the investigations. PD Dr. habil. Ina negative influence on Y production.

CH4
Körner and Prof. An-Ping Zeng are highly Specifically, the Se*Mo effect necessitated a appreciated for supervising and co-supervising decrease in Mo concentration with increase in the authorduring the investigations. Se concentration (Table 2). This implies that based on the Se*Mo and VFA*Se effects, Mo bears an inverse relationship with Se and VFA