Zero Valent Iron Significantly Enhances Methane Production from Waste Activated Sludge by Improving Biochemical Methane Potential Rather Than Hydrolysis Rate

Anaerobic digestion has been widely applied for waste activated sludge (WAS) treatment. However, methane production from anaerobic digestion of WAS is usually limited by the slow hydrolysis rate and/or poor biochemical methane potential of WAS. This work systematically studied the effects of three different types of zero valent iron (i.e., iron powder, clean scrap and rusty scrap) on methane production from WAS in anaerobic digestion, by using both experimental and mathematical approaches. The results demonstrated that both the clean and the rusty iron scrap were more effective than the iron powder for improving methane production from WAS. Model-based analysis showed that ZVI addition significantly enhanced methane production from WAS through improving the biochemical methane potential of WAS rather than its hydrolysis rate. Economic analysis indicated that the ZVI-based technology for enhancing methane production from WAS is economically attractive, particularly considering that iron scrap can be freely acquired from industrial waste. Based on these results, the ZVI-based anaerobic digestion process of this work could be easily integrated with the conventional chemical phosphorus removal process in wastewater treatment plant to form a cost-effective and environment-friendly approach, enabling maximum resource recovery/reuse while achieving enhanced methane production in wastewater treatment system.

L arge amounts of organic matter from wastewater are converted into waste activated sludge (WAS) during biological treatment processes in wastewater treatment plants (WWTPs) 1 . Anaerobic digestion of WAS has been widely applied to stabilize and reduce the volume of WAS as well as produce a renewable bioenergy resource in the form of methane [2][3][4][5] . The anaerobic digestion process generally consists four stages, i.e. hydrolysis, fermentation, acetogenesis and methanogenesis for methane production 6 . However, the application of anaerobic digestion is often limited by the slow hydrolysis rate and/or poor biochemical methane potential (or degradation extent) of the WAS [7][8][9][10][11] .
In order to effectively enhance methane production from WAS in anaerobic digestion, a number of strategies have been developed, such as thermal, chemical, and mechanical methods [12][13][14][15][16][17][18][19][20][21] . However, most of these methods are cost intensive owing to high energy input or large chemical requirements 7,22,23 . Thus, alternative cost-effective approach to improve methane production from WAS in anaerobic digestion process is highly desired.
Zero valent iron (ZVI), a cheap reducing agent, has been widely used in wastewater pretreatment, groundwater purification and soil remediation 24,25 . Recent studies have found that ZVI addition in anaerobic reactors for biological wastewater treatment could significantly improve chemical oxygen demand (COD) removal by ca. 25% [26][27][28][29] . Indeed ZVI can lower oxidation-reduction potential and serve as an acid buffer, thus helping maintain a stable and favourable condition for methanogens. Previous studies also demonstrated that ZVI addition could promote hydrolysis/acidification and optimize volatile fatty acid (VFA) compositions 30,31 . Therefore, ZVI addition to the anaerobic digester could be a potential cost-effective approach to improve methane production from WAS 32,33 .
In this work, the impacts of three different types of ZVI (i.e., iron powder, clean scrap and rusty scrap) on methane production from WAS in anaerobic digestion were evaluated systematically using both experimental and mathematical approaches. A model-based analysis was performed to reveal the mechanism of ZVI-driven enhancement of methane production from WAS. Based on the results of economic analysis, a cost-effective integrated ZVI-based anaerobic WAS digestion process was also proposed.

Results
Effects of ZVI addition on methane production. Three types of ZVI were evaluated, i.e., iron powder, clean scrap and rusty scrap. Fig. 1 presents the methane production results from the biochemical methane potential (BMP) tests in Experiment I and II (see Methods Section). In general, ZVI addition enhanced the methane production from WAS. The increased ZVI powder addition resulted in increasing methane production (Fig. 1a). For example, 4 g/L ZVI addition increased methane production by 21% as compared to the control (0 g/L ZVI powder addition) on Day 20. As shown in Fig. 1b, 10 g/L ZVI powder, 10 g/L clean iron scrap and 10 g/L rusty iron scrap led to 11%, 22% and 30% increase on methane production, respectively, compared to that of control on Day 20. The level of methane production variation between Experimental I and II is not unexpected as real sludge from a full-scale WWTP was used, with characteristics likely varying with time. Therefore, direct comparisons between Experiment I and II are not meaningful due to the possible variation of the sludge characteristics during sludge sampling. In addition, it should be noted that the methane production by utilization of ZVI as electron donors is negligible compared to the overall methane production in the system as ZVI powder or clean scrap addition produces similar content of methane with the addition of rusty scrap.
These results demonstrated that ZVI addition indeed enhanced methane production from WAS during anaerobic digestion. The results also showed that both the clean and the rusty iron scrap were more effective than the iron powder for improving methane production from WAS. The better performance of ZVI scrap was likely due to its better contact with sludge and liquid 33 . In particular, the addition of rusty iron scrap is the most effective ZVI form for improving methane production from WAS, likely due to the fact that Fe (III) oxides on the rusty iron scrap surface could induce dissimilatory ferric iron reduction to enhance degradation of complex substrates such as WAS 34 .
Effects of ZVI on hydrolysis rate and biochemical methane potential. The hydrolysis rate (k) and biochemical methane potential (B 0 ) were estimated using both one-substrate and twosubstrate models. Table 1 shows the estimated k and B 0 for the methane production from the WAS digestion subject to different ZVI forms and dosages using one-substrate model, while the estimated values of k, rapid , k, slow , B 0,rapid and B 0,slow in both Experiment I and II using two-substrate model are presented in Table S1 in Supplementary Material. Overall, k, rapid and B 0,rapid in two-substrate model are the same as k and B 0 in one-substrate model. Both k, slow and B 0,slow in two-substrate model are zero after fitting. These modeling results indicated that the WAS composition was homogeneous and the methane production from the WAS could be well described by one-substrate model.
The simulated methane production curves using one-substrate model are shown in Fig. 1, which matched all the experimental data from both Experiment I and II, further confirming the one-substrate model could well describe the methane production data. As can be seen from Table 1, the ZVI addition at all the levels applied achieved significantly higher B 0 than that of the control. The biochemical methane potential was enhanced by 9%-21% in Experiment I and 12%-29% in Experiment II compared to the corresponding control. In contrast, the ZVI addition has no effect on the k value and the obtained k values were constant in both Experiment I (ca. 0.083 d 21 ) and Experiment II (ca. 0.072 d 21 ) regardless of the amount of ZVI addition. Fig. 2 shows the 95% confidence regions of k and B 0 , which provide valuable information about model uncertainty and the identifiability of the obtained parameter values. The increased ZVI addition consistently resulted in better biochemical methane potential (B 0 ), and the confidence region moved rightward to the higher B 0 direction (x-axis) in Fig. 2. In contrast, the increased ZVI addition had no impact on the hydrolysis rate, with no real changes in confidence region locations on y-axis. In addition, there was no obvious increase in confidence region area in both Fig. 2a and 2b.

Discussion
ZVI addition improved biochemical methane potential of WAS rather than its hydrolysis rate. There are two key measures of sludge degradability that are relevant, the apparent first order degradation rate coefficient (k) and the biochemical methane potential (B 0 ), which represent the speed and extent of sludge conversion, respectively 35 . Model-based analysis of these two parameters and the related parameter identifiability in this work clearly showed that ZVI addition significantly enhanced methane production from WAS through improving the biochemical methane potential of WAS rather than its hydrolysis rate.
Feng et al. 32 did not look into the mechanisms for the enhanced methane production by ZVI addition and only hypothesized that the main reason might be the improved major enzyme activities related to hydrolysis and acidification. Contradictorily, this study demonstrated that the ZVI addition did not accelerate the hydrolysis rate (k) in both experiments with different types of ZVI addition. On the contrary, biochemical methane potential (B 0 ) was significantly improved by ZVI addition, indicating that ZVI increased the extent of sludge conversion and altered the sludge property 35 . It has been reported that VS destruction during anaerobic digestion of waste activated sludge generally increased with the increase of ferrous iron content in the sludge 36,37 . Indeed, ZVI can release from Fe 0 to Fe 21 (Fe 0 1 2H 1 5 Fe 21 1 H 2 ), and thus leading to a significant increase of iron content in the sludge 33,38 . As shown in Fig. 3, in this work, the released ferrous iron concentrations from ZVI also showed a good correlation with both VS reduction and the biochemical methane potential (B 0 ). Therefore, the alternation of sludge property to improve biochemical methane potential by ZVI could likely be the main reason for the enhanced performance of methane production.
A strategy to implement ZVI-based anaerobic digestion process in wastewater treatment plant. From an integrated environmental and economic perspective, nutrients source in wastewater treatment systems should be managed such that both good nutrients removal performance and high resource recovery or reuse can be achieved. Based on the findings of this work, a new strategy could be proposed to simultaneously enhance methane production from WAS and iron resource reuse through integrating the ZVI-based anaerobic digestion process of this work with the conventional chemical phosphorus removal process in WWTPs.
As presented in Fig. 4, waste iron scrap (the most efficient ZVI as demonstrated in this work) can be freely obtained from machinery factory and then transported to the WWTP. The obtained iron scrap (ZVI) can be added to the anaerobic digester in order to enhance the methane production by increasing the biochemical methane potential. In anaerobic digester, ZVI can be released from Fe 0 to Fe 21 , and thus eliminated the potential sulfide production/accumulation issues as well as the possible H 2 S emission in the biogas in traditional anaerobic digester through iron sulfide precipitation 39 . This in turn could further enhance the performance of WAS digestion without additional chemical cost from external ferrous/ferric iron dosing 40 . With regard to the generation of organic sulfur odors from the dewatered sludge cakes, iron could also reduce odor-causing gases, resulting in better quality of dewatering sludge. More importantly, the Fe (II) in anaerobic digestion liquor can be recycled to bioreactor and further oxidized to Fe (III), which can be used for chemical    phosphors removal via the generation of FePO 4 41 . This strategy would not only represent a significant process cost reduction (further discussed below), but also improve the sludge and wastewater treatment efficiency, enabling maximum resource (iron) reuse while achieving improved methane production. In addition, from a network-wide view, commonly used ferric iron dosing in sewers for H 2 S control 42 might also be useful for CH 4 production enhancement during anaerobic digestion and phosphors removal in the WWTP.
Potential economic feasibility of ZVI-based technology for enhancing biological methane production. It has been demonstrated that the estimated lab-scale BMP results are more conservative or comparable to full-scale test results 43 . Thus, the estimated values obtained in the current study are used for a conservative assessment of the potential economic feasibility of the proposed ZVI-based anaerobic digestion technology. This was carried out by a desktop scaling-up study on a full-scale WWTP with a population equivalent (PE) of 400,000 and with an anaerobic sludge digester at a hydraulic retention time (HRT) of 20 days. 10 g/L rusty iron scrap was chosen for the following economic evaluations.
From the Fe 21 released (41 mg/L), theoretically, the iron scrap could be recycled for approximately 243 batches (10*1000/41) if the loss of iron solid through effluent is ignored. With a 29% increase in methane production at this level of ZVI addition, the net economic benefit is estimated to be around $231,000 per annum compared with the system without ZVI addition (see Table S2 in Supplementary Material). The net benefit arises from the enhanced methane production associated benefit (i.e., its conversion to heat and power) ($150,000 per annum) and decreased WAS transport and disposal costs ($90,000 per annum) overweighing the additional costs for ZVI transport and ZVI chamber ($9,000 per annum). The advantages of ZVI addition on sulfide control in digester, phosphors removal through anaerobic digestion liquor recycle and better dewa-tering sludge have not been considered. Therefore, the ZVI-based technology is potentially economically attractive indeed. However, the benefit and cost values presented should be considered indicative only. In particular, they may vary from region to region and from country to country, depending on the local conditions. In addition, the direct quantitative economic and performance comparison with other available technologies are difficult at this stage since the results depend on many factors including the WAS characteristics among others 23 , which remains further investigations in the future. This could and should be done in future studies by performing experiments using the same WAS and under similar operating conditions. Moreover, it should be noted that there is no environmental consequence of the proposed chemical-free ZVI-based technology based on CO 2 emission, revealing this approach being environmental friendly. In comparison, some other WAS pretreatment technologies (i.e., thermal and alkaline pretreatment) might cause negative environmental effect 44 . Different from temperature phased anaerobic digestion and mechanical pretreatment which generally increase k 45,46 , this ZVI-based approach improved B 0 , thus potentially allowing more methane production in terms of performance improvements for anaerobic digesters. Since it does not improve the degradation rate, this requires the same amount of HRT in order to achieve maximized sludge reduction. It should be noted that labscale batch tests were performed in our study. Full-scale system may behave differently in terms of k and B 0 as demonstrated in Bastone et al. 43 . Therefore, full-scale trials are needed to further evaluate this technology.
In summary, the effects of three different types of ZVI (i.e., iron powder, clean scrap and rusty scrap) on methane production from WAS in anaerobic digestion were investigated by using both experimental and mathematical approaches. The results demonstrated that both the clean and the rusty iron scrap were more effective than the iron powder for improving methane production from WAS. ZVI addition significantly enhanced methane production from WAS through improving the biochemical methane potential of WAS rather than its hydrolysis rate. The alternation of sludge property by ZVI resulted in improved biochemical methane potential and thus the enhanced methane production. The ZVI-based anaerobic digestion process could be potentially implemented and integrated with the conventional chemical phosphorus removal process in wastewater treatment plant to form a cost-effective and environmentfriendly technology, enabling maximum resource recovery/reuse while achieving enhanced methane production in wastewater treatment system.

Methods
Waste activated sludge. The waste activated sludge used in this work was collected from the sludge treatment unit at a full-scale municipal wastewater treatment plant in Dalian, China. The sludge was stored at 4uC before use. The volatile solids (VS) to total chemical oxygen demand (TCOD) ratios of the sludge used for methane production ranged between 0.60 and 0.67. ZVI sources. Three types of ZVI were evaluated, i.e., iron powder, clean scrap and rusty scrap. The ZVI powder has a diameter of 0.2 mm with BET surface area of 0.05 m 2 /g and purity .98%. The rusty scrap (about 8 mm * 4 mm * 0.5 mm, purity . 95%) was obtained from a machinery workshop in Dalian, China. The clean scrap was acquired through a pretreatment of the rusty scrap to remove the rusty cover. The difference between the two scraps is that the rusty scrap had a corrosion layer covering the surface of the scrap 33 .
Anaerobic biochemical methane potential tests. In order to evaluate the effect of different forms of ZVI on methane production in anaerobic digestion, methane production from the WAS with different types of ZVI addition was assessed using anaerobic batch BMP tests 32 . The inoculum for the BMP tests was collected from an anaerobic digester 33 . Two types of batch experiments were performed. In Experiment I, 0, 1.0, and 4.0 g/L of ZVI powder were added into three identical sets of BMP test vials, respectively. In Experiment II, 10 g/L ZVI powder, 10 g/L clean scrap and 10 g/ L rusty scrap were used as ZVI sources and dosed to three identical sets of BMP vials for comparison, with a control test in which no ZVI was added.
In each test, WAS, ZVI and the inoculum obtained from the anaerobic digester were added into serum vials for BMP tests. After that, the vials were capped with silica Figure 4 | A proposed strategy to integrate ZVI-based anaerobic digestion process of this work with the conventional chemical phosphorus removal process in wastewater treatment plant. ZVI addition in anaerobic digester can enhance methane production from WAS. The sulfide produced in anaerobic digester can be precipitated by ferrous iron that produced from ZVI addition, resulting in enhanced sulfide-free biogas (methane) production. The anaerobic digestion liquor containing Fe (II) can be reused and fed into bioreactors, in which the Fe (II) can be oxidized to Fe (III). The generated Fe (III)-containing effluent can then be used for chemical phosphorus removal process, to form a cost-effective and environment-friendly technology, enabling maximum resource recovery/ reuse while achieving enhanced methane production in wastewater treatment system. www.nature.com/scientificreports SCIENTIFIC REPORTS | 5 : 8263 | DOI: 10.1038/srep08263 gel stoppers. The oxygen was removed from the headspace by exchanging it with nitrogen gas for at least 10 min. All BMP tests were conducted at 35 6 1uC for 20 d. The biogas (methane) production in BMP vials was collected and monitored by using gas chromatograph (Shimadzu, GC-14C) equipped with a thermal conductivity detector. More details of the BMP tests can be found elsewhere 32,33 .
Model-based analysis. The hydrolysis rate (k) and biochemical methane potential (B 0 ) are the two key parameters associated with methane production from WAS 7,8,10 . In this work, these two parameters were used to evaluate and compare the methane production kinetics and potential of the WAS at different ZVI levels or with different types of ZVI. They were estimated by fitting the methane production data from the BMP tests to a first-order kinetic model using a modified version of Aquasim 2.1d with sum of squared errors (J opt ) as an objective function 43 . The uncertainty surfaces of k and B 0 , based on a model-validity statistical F-test with 95% confidence limits, were also estimated by using Aquasim 2.1d 43 .
Two models were applied. The first one considered a single substrate type (i.e., onesubstrate model) in the first-order kinetic model 22,43 , as shown in Equation (1): where B(t) (L CH 4 /kg VS) is the cumulative methane production at time t (d). In the second model, the WAS samples comprised a rapidly biodegradable substrate type and a slowly biodegradable substrate type (i.e. two-substrate model) 47 . The equation of the two-substrate model is shown below: where B 0,rapid and B 0,slow (L CH 4 /kg VS) are biochemical methane potentials of the rapidly biodegradable substrates and slowly biodegradable substrates, respectively; k rapid and k slow (d 21 ) are hydrolysis rates of the rapidly biodegradable substrates and slowly biodegradable substrates, respectively. Based on the determined B 0 , the degradation extent (Y) of WAS could then be calculated using Equation (3): where 380 (L CH 4 /kg TCOD) is theoretical biochemical methane potential of WAS under standard conditions (25uC, 1 atm); R WAS is the measured VS to TCOD ratio in the WAS.