Chemical characterization and anaerobic biodegradability of hydrothermal liquefaction aqueous products from mixed-culture wastewater algae

Bioresource Technology xxx (2014) xxx–xxx

Contents lists available at ScienceDirect

Bioresource Technology journal homepage: www.elsevier.com/locate/biortech

Chemical characterization and anaerobic biodegradability of hydrothermal liquefaction aqueous products from mixed-culture wastewater algae Giovana Tommaso a,b, Wan-Ting Chen a, Peng Li a, Lance Schideman a, Yuanhui Zhang a,⇑ a b

Department of Agricultural and Biological Engineering, University of Illinois at Urbana-Champaign, 1304 W Pennsylvania Avenue, Urbana, IL 61801, USA Laboratory of Environmental Biotechnology, Department of Food Engineering, University of Sao Paulo, 225, Duque de Caxias Norte, Pirassununga, Sao Paulo 13635-900, Brazil

h i g h l i g h t s  Aqueous products from hydrothermal liquefaction (HTL-ap) are formed in large amounts.  HTP-ap may contain substances toxic to several organisms.  Further reuse or treatment of the HTL-ap is necessary.  Anaerobic digestion of HTL-ap could be conducted.

a r t i c l e

i n f o

Article history: Received 31 July 2014 Received in revised form 1 October 2014 Accepted 4 October 2014 Available online xxxx Keywords: Anaerobic Methane Hydrothermal liquefaction aqueous product Algae

a b s t r a c t This study examined the chemical characteristics and the anaerobic degradability of the aqueous product from hydrothermal liquefaction (HTL-ap) from the conversion of mixed-culture algal biomass grown in a wastewater treatment system. The effects of the HTL reaction times from 0 to 1.5 h, and reaction temperatures from 260 °C to 320 °C on the anaerobic degradability of the HTL-ap were quantified using biomethane potential assays. Comparing chemical oxygen demand data for HTL-ap from different operating conditions, indicated that organic matter may partition from organic phase to aqueous phase at 320 °C. Moderate lag phase and the highest cumulative methane production were observed when HTL-ap was obtained at 320 °C. The longest lag phase and the smallest production rate were observed in the process fed with HTL-ap obtained at 300 °C. Nevertheless, after overcoming adaptation issues, this HTL-ap led to the second highest accumulated specific methane production. Acetogenesis was identified as a possible rate-limiting pathway. Ó 2014 Elsevier Ltd. All rights reserved.

1. Introduction Due to superior photosynthetic efficiencies and high CO2 fixation ability, algae are regarded as an attractive biomass feedstock for renewable energy production. However, with respect to life cycle assessment, current methods of algae production have more environmental impact in terms of energy use, water use, and greenhouse gas emissions when compared to other conventional biomass sources such as switchgrass and corn (National Research Council, 2012). In order to reduce the environmental footprint of algae cultivation, the use of wastewater to cultivate algae has been suggested. Roberts et al. (2013), among others, used wastewater to cultivate algae and showed that wastewater derived algae (AW) may be an appropriate feedstock for hydrothermal liquefaction ⇑ Corresponding author. E-mail address: [email protected] (Y. Zhang).

(HTL). In fact, because algae can uptake wastewater nutrients during cultivation, it is expected that both energy production and wastewater treatment can be achieved simultaneously if mixedculture algae from wastewater treatment systems are used as a bioenergy feedstock. HTL is an attractive process to produce bio-crude oil from wet feedstocks such as manure and algae, because it reduces the need for feedstock drying and dewatering. During the HTL process, water approaches its super-critical conditions and serves as both a reactant and a catalyst (Peterson et al., 2008), and the resulting biocrude oil self-separates from the aqueous fraction. HTL typically occurs through a complex sequence of reactions, which involves converting the biomass into small reactive molecules, and then polymerizing unstable molecules, which leads to the formation of oily compounds. The main products are bio-crude oil (30–50% of dry feedstock) with a relatively high heating value (32–39 MJ/ kg), solid residues, a gas product, and an aqueous product with

http://dx.doi.org/10.1016/j.biortech.2014.10.011 0960-8524/Ó 2014 Elsevier Ltd. All rights reserved.

Please cite this article in press as: Tommaso, G., et al. Chemical characterization and anaerobic biodegradability of hydrothermal liquefaction aqueous products from mixed-culture wastewater algae. Bioresour. Technol. (2014), http://dx.doi.org/10.1016/j.biortech.2014.10.011

2

G. Tommaso et al. / Bioresource Technology xxx (2014) xxx–xxx

20–50% of dry feedstock solids that is rich in organics (Chen et al., 2014a; Pham et al., 2013). Among several parameters that may impact the HTL process, reaction time and reaction temperature are generally viewed as two of the key parameters (Yu et al., 2011). According to Marcilla et al. (2013), reaction times usually range between 5 and 60 min, and have a strong effect on the composition of products and overall conversion rates. Peterson et al. (2008) showed that HTL is typically carried out at temperatures ranging from 200 to 360 °C and pressures high enough to keep water in the liquid phase (15– 25 MPa). Due to the existing competition between hydrolysis, fragmentation and repolymerization, intermediate temperatures are expected to yield higher amounts of bio-crude oil. Nevertheless, the suitable temperature for oil production also depends upon the feedstock type. In this way, several feedstock types including lignocelluloses, manure and algae have been carefully investigated under hydrothermal liquefaction (He et al., 2000; Peterson et al., 2008; Vardon et al., 2011). For example, it was found that the structure of lignocelluloses (e.g. lignin and cellulose) would be greatly degraded when HTL reaction temperatures were beyond 300 °C (Peterson et al., 2008). Another example is that swine manure generally requires reaction temperature between 275 and 315 °C for bio-crude oil production (He et al., 2000). Vardon et al. (2011) also pointed out that both the feedstock’s organic content and its nutritional composition can greatly affect HTL bio-crude oil yields and quality. For example, the bio-crude oil obtained via HTL of low-lipid microalgae typically contains more nitrogen content than those from manure and sludge (Chen et al., 2014a; Vardon et al., 2011), which indicates that the lowlipid microalgae-based bio-crude oil may require further denitrification processes for transportation fuel applications. In addition, it is generally believed that under HTL, the solid residue yield is positively correlated to the amount of cellulose in the feedstock (Chen et al., 2014a; Demirbasß, 2000). Chen et al. (2014a) found that protein derivatives may react with crude fat derivatives when the crude fat was drastically increased in the liquefaction system. Overall, feedstock with lower cellulose content is suggested for lower solid residue production and the separation of protein derivatives from crude fat derivatives may be needed in order to reduce the nitrogen content in the bio-crude oil. Nonetheless, because bio-crude oil is the most desired product from the HTL reaction, relatively little attention has been paid to the HTL-ap, even though it accounts for a significant fraction of the organics. In the work conducted by He et al. (2000), for example, the HTL-ap represented up to 82% of the total mass (including water), and it contained at least 25% of the original organic mass supplied to the system as well as most of the nitrogen, phosphorus and potassium. Marcilla et al. (2013) also stated that the aqueous product should be recycled to improve the overall economic viability of the HTL process. HTL-ap also contains substances toxic to several organisms and it can be classified as a petrochemical refinery wastewater (Appleford, 2005). Recently, Pham et al. (2013) also observed that Spirulina-derived HTL-ap was highly cytotoxic to mammalian Chinese hamster ovary cells, causing a 50% decrease in cells when present in a relative volume of 7.5% in the growth media. Several nitrogenous organic compounds were identified and a synergistic cytotoxicity effect among most of these compounds was also observed. In conclusion, the authors stated that HTL-ap should be treated before discharging it into the environment. According to Razo-Flores et al. (2006), toxic wastewater are likely difficult to treat aerobically and anaerobically. However, Moreno-Andrade and Buitrón (2004) stated that the impact of the toxicant is obviously related to the amount of the compound and to the amount of biomass (sludge) inside the reactors. In addition, Razo-Flores et al. (2006) mentioned that biomass

concentration in modern high-rate anaerobic reactors is 10–20 times higher than in conventional activated sludge. This leads to the conclusion that anaerobic reactors would be a good choice for toxic wastewater treatment. Furthermore, Chen et al. (2008) stated that the concepts of toxicity and inhibition are intrinsically related to the process conditions. Changes in the anaerobic microflora in relation to prevailing methanogenic species result in adaptation of biomass, which is a preponderant condition in the anaerobic degradation of various toxic compounds. Thus, the present work aims to correlate the key HTL operating parameters (reaction time and temperatures) with the aqueous products characteristics, including its anaerobic biodegradability. Therefore, the possibility of combining these two promising renewable energy production technologies (HTL and anaerobic digestion) is evaluated. This study is expected to contribute to the establishment of a sustainable system with an independent and secure energy production, which is referred as EnvironmentEnhance Energy (E2-Energy) here and elsewhere (Yu et al., 2011; Chen et al., 2014b). This proposed sustainable system aims to integrate waste treatment, water purification and carbon capture into the energy production process, so that it can maximize the economic value of bio-wastes while at the same time minimizing their negative impacts on the environment. 2. Methods 2.1. Hydrothermal liquefaction (HTL) process The Hydrothermal liquefaction conversions were performed according to Yu et al. (2011), using a stainless steel cylinder reactor of 100 ml capacity with a magnetic drive stirrer and moveable vessel (Model 4593, Parr Instrument Co., Moline, IL, USA) in batch mode. Reaction temperatures were 260, 280, 300 and 320 °C. Reaction temperatures were selected based on previous studies (Yu et al., 2011) about hydrothermal liquefaction (HTL) of microalgae. When the temperature was 300 °C, four reaction times of 0, 0.5, 1.0 and 1.5 h were studied. The reaction time variation study was performed only at the temperature of 300 °C because it was demonstrated that the reaction temperature of 300 °C typically can lead to a higher bio-crude oil yield as well as energy recovery to biocrude oil (Zhang et al., 2013; Chen et al., 2014b; Yu et al., 2011). Consequently, reaction temperature of 300 °C was designated for reaction time variation study in this case. The reaction time was considered zero when the experiments were conducted during the heating and cooling period without any maintenance at the chosen temperature. 30 g slurry feedstock containing 25% total solid content by weight was placed into the reactor for each HTL test. The composition of AW is available in Chen et al. (2014b). 2.2. Aqueous phase characterization Organic matter concentration expressed as chemical oxygen demand (COD) was performed based on the methods described in Standard Methods for Examination of Water and Wastewater (APHA, 1998), methods 5220 and 5540 respectively. The pH value was measured with a calibrated potentiometer, total Nitrogen (TN) content was measured by the Persulfate Digestion Test ‘N Tube, Hach Method 10072 (range: 10–150 mg L1). Total ammonia (TAN) content was measured as NH+4-N using Salicylate Test ‘N Tube, Hach Method 10031 (range: 0.4–50.0 mg L1). 2.3. GC–MS analyses The chemical compositions of the HTL-ap were analyzed as described by Chen et al. (2014b) using a GC–MS (7890A, Agilent

Please cite this article in press as: Tommaso, G., et al. Chemical characterization and anaerobic biodegradability of hydrothermal liquefaction aqueous products from mixed-culture wastewater algae. Bioresour. Technol. (2014), http://dx.doi.org/10.1016/j.biortech.2014.10.011

3

G. Tommaso et al. / Bioresource Technology xxx (2014) xxx–xxx

Technologies, Santa Clara, CA, USA). Gas chromatography was performed on a 15 m ZB-FFAP column with 0.25 mm inner diameter (I.D.) and 0.25 lm film thickness (Phenomenex, Torrance, CA, USA) with injection temperature of 250 °C, MSD transfer line of 250 °C, and the ion source adjusted to 230 °C. The spectra of all chromatogram peaks were evaluated using the HP Chem station (Agilent, Palo Alto, CA, USA) and AMDIS (NIST, Gaithersburg, MD, USA) programs and compared with electron impact mass spectrum from NIST Mass Spectral Database (NIST08) and W8N08 library (John Wiley & Sons, Inc.). To allow comparison between samples, all data were normalized to the internal standard (3-methyl butanoic acid, 0.1 lM). 2.4. Biomethane potential (BMP) assays BMPs were assembled using inoculum from an anaerobic reactor treating secondary sludge provided by the Urbana Sanitary District (Urbana, Illinois, USA). The sludge presented total volatile solids of 12.8 g L1 and had specific methanogenic activity (SMA) measured according to recommendations present at Angelidaki et al. (2009), using glucose and acetate separately as substrates. The glucose initial concentration was 1.0, 3.5 and 5.0 g L1. The acetate initial concentration was 1.0 and 3.5 g L1. When glucose was used, the average SMA was 0.028 g CODCH4 gTVS1 d1; when acetate was used the SMA value was 0.013 g CODCH4 gTVS1 d1. Before the BMP assay assembly, the sludge was acclimatized with a synthetic substrate, composed by glucose and urea, with COD/Nitrogen ratio of 1/0.15. That rate value was similar to the one verified by Appleford (2005), in the liquid phase from swine manure HTL conversion. With this ratio, the COD was raised from 0.5 to 3.5 g L1 in three weeks. When biogas production was verified consistently, as well as the organic matter removal, BMPs were assembled based on the COD of each HTL-ap. Table 1 presents the volumes and specific added COD values. The specific loading rate for this assay averaged 0.28 g COD gTVS1. Blank flasks were assembled without external carbon source. Control flasks fed with glucose and urea (COD/Nitrogen ratio of 1/0.15) were also assembled with the same specific loading rate in order to guarantee that the biomass was active at the time. The assays were assembled in 50 mL flasks with 20 mL of reaction volume from HTL processes occurred at 260, 280 and 300 °C, and 13 mL of reaction volume for samples from HTL processes occurred at 320 °C, maintaining the same rate between substrate (COD) and microorganisms (STV). The dilutions were performed in basal medium prepared according to Angelidaki et al. (2009). The pH value was adjusted with the addition of 0.1 mL of 10% HCl. The flasks were sealed with butyl rubber septa and aluminum screw caps, and the headspace was flushed with N2 before incubation. After anaerobic digestion, samples for volatile fat acids (VFA) concentration analysis were centrifuged at 5000g for 10 min and filtered using 0.22 lm filters (ANOW Microfiltration co., Ltd., Hangzhou, China). VFAs were quantitatively analyzed using high-performance liquid chromatography (Shimadzu

Scientific Instruments, USA) equipped with an SPD-M20A UV detector and an Aminex HPX-87H column (300 mm  7.8 mm and 9 lm particle size) (Bio-RAD, California, USA). 2.5. Data analysis The cumulative methane production curves, obtained from the BMP tests were fitted by a modified Gompertz equation (Eq. (1)), as previously done by Chen et al. (2006):

   ke PCH4 ðtÞ ¼ PCH4 e e ðk  tÞ þ 1 ; PCH4

ð1Þ

PCH4 ðtÞ is the cumulative specific methane production (mmol/ SSV) at time t; PCH4 is the specific methane production potential (mmol/SSV), k is the specific methane production rate; k is the duration of the lag phase and e is Euler’s number, the mathematical constant (2.71828). The model was fitted using Levenberg–Marquardt method (Microsoft Origin 8.0). The percentage of methanogenesis was calculated using Eq. (2), where MCH4 is the cumulative methane production from external carbon sources (expressed in COD). It is calculated by subtracting the average cumulative production observed in blank flasks from the average cumulative production observed in the flasks fed with HTL-ap as carbon source; and MCOD is the mass of organic matter provided in each trial (expressed in COD).

Methanogenesis ð%Þ ¼

MCH4 MCOD

ð2Þ

In order to verify the presence of methane overproduction, the maximum theoretical methanogenic potential was calculated according to Speece (1996), who stated that 0.395 L of CH4 is generated per gram of removed COD. 3. Results and discussion 3.1. Aqueous phase characterization Fig. 1 shows the influence of the reaction temperature and reaction time on the COD measured in the aqueous phase from HTL conversion of AW, comparing it with carbon recovery data (Chen et al., 2014b). In Fig. 1a, it is found that from 0 to 1 h, there is a negative correlation between reaction time and COD. In contrast, the carbon recovery distributed to the bio-crude oil was increased, which indicated that the organic matter partitioned from the liquid phase to the oil phase. With 1.5 h of reaction time, the COD value showed a sharp increase inferring a migration of organic matter to the aqueous products. Valdez and Savage (2013) also found similar results for the hydrothermal liquefaction of Nannochloropsis sp. and stated that as the retention time increased, light crude oil tended to be converted into aqueous products. A similar trend was found when considering the effect of reaction temperature in the HTL process (Fig. 1b). It was possible to see that, from 260

Table 1 Conditions applied to the BMP assay where the carbon source was provided by the addition of the aqueous phase from the HTL of mixed-culture algae from wastewater treatment systems (AW).

a b c

HTL condition (temperature °C)

CODadda (mg)

pH

TANaddb (mg)

TVSc (g)

260 280 300 320 Blank

0.78 0.72 0.85 0.81 0

7.50 7.61 7.67 7.60 7.66

6.2 2.1 5.1 1.8 0

0.103 0.097 0.083 0.058 0.084

CODadd – mass of organic matter expressed in COD added to the flasks. TANadd – total ammonia nitrogen added to the flasks. TVS – total volatile solids.

Please cite this article in press as: Tommaso, G., et al. Chemical characterization and anaerobic biodegradability of hydrothermal liquefaction aqueous products from mixed-culture wastewater algae. Bioresour. Technol. (2014), http://dx.doi.org/10.1016/j.biortech.2014.10.011

4

G. Tommaso et al. / Bioresource Technology xxx (2014) xxx–xxx

Fig. 1. Effects of (a) reaction temperature and (b) reaction time on carbon recovery distributed in various liquefaction products and COD of HTL-ap (HTL aqueous products).

to 300 °C, COD values of HTL-ap decreased, but when the reaction temperature was increased to 320 °C, the COD value slightly increased. Chen et al. (2014b) studied the oil yield from the HTL processes that originated the HTLap here studied. The best oil yield was obtained when 1 h and 300 °C were used as HTL conditions. Accordingly, this condition originated the HTLap with lowest COD. Tables 2 and 3 present the main classes of molecules identified in this study when residence time and temperature were varied, respectively. Major compounds identified in aqueous products by GC–MS are given on Figs. S1 and S2 (Supplementary data). It is noticed that the percentage of the relative peak area (%RPA) of organic acids increased with reaction time and then decreased when the reaction lasted for 1.5 h (Table 2). In contrast, cyclic amines and amides, which may be degraded from proteins, appeared at the very beginning of the HTL reaction and greatly decreased when the reaction lasted for more than 0.5 h. The above observation indicated that the decomposition of proteins was more complete when reaction time lasted for more than 0.5 h. Besides, it has to be pointed out that when reaction time was increased from 1 h to 1.5 h, the %RPA of fatty acid derivatives increased, whilst glycerol decreased (Supplementary data). This infers that fatty acids were synthesized in this case. Considering the effect of reaction temperatures (Table 3), the RPA of acetic acids decreased as the temperature increased from 260 to 300 °C. Nevertheless, when temperature was further increased to 320 °C, the relative pick area (RPA) of propanoic acid increased (Sup-

Table 2 Effect of retention time on aqueous products composition when temperature was set at 300 °C. Molecules classes

Short chain organic acid (C2-C4) Long chain organic acid (C>4) Fatty acid & Fatty alcohols Amino acid Benezoic acid Derivatives Cyclic hydrocarbons Phenols Straight amides derivatives N-heterocyclic cmpds Oxygenates (cyclic & straight) Ketones Total

HTL – residence time (h) 0 (%)

0.5 (%)

1 (%)

1.5 (%)

4.73 2.24 5.31 2.04 1.75 0.00 1.58 12.1 54.7 0.47 6.21 91.2

24.5 2.81 7.17 2.90 1.06 2.54 0.61 6.48 35.5 0.73 2.33 86.6

34.6 4.26 4.01 4.88 2.70 0.39 0.57 2.74 36.7 1.82 1.83 94.5

24.2 3.35 16.2 1.53 1.30 2.95 0.76 13.3 20.5 2.01 0.89 87.0

Table 3 Effect of temperature on aqueous products composition when retention time was set at 60 min. Molecules classes

Short chain organic acid (C2-C4) Long chain organic acid (C5-C6) Fatty acid & fatty alcohols Amino acid Benezoic acid Derivatives cyclic Hydrocarbons Phenols Straight amides Derivatives N-heterocyclic cmpds Oxygenates (cyclic & straight) Ketones Total *

HTL temperature 260 °C

280 °C

300 °C

320 °C

21.1% 1.89% 3.78% 7.07% 2.03% Nd* 0.21% 6.31% 31.6% 6.70% 11.6% 92.2%

26.7% 2.98% 7.13% 4.91% 1.38% Nd 0.36% 5.89% 31.0% 10.1% 4.94% 95.4%

34.6% 4.26% 4.01% 4.88% 2.70% 0.39% 0.57% 2.74% 36.7% 1.82% 1.83% 94.5%

9.10% 8.74% 1.78% 0.14% 5.55% Nd 8.44% 16.3% 36.8% 1.48% 4.98% 93.4%

Nd = not detected.

plementary data). This also infers that two competitive reactions, the polymerization and decomposition of volatile compounds, may take place at the same time. It is found that the %RPA of cyclic amines and amides decreased when the reaction temperature increased. In addition, when reaction temperature reached 320 °C, the area of peaks relative to nitrogen-heterocyclic compounds significantly decreased (Supplementary data), deducting that those molecules might be further decomposed at relatively high reaction temperatures. It is important to point that the RPA of urea sharply increased when temperature was 320 °C. Because there are limitations (e.g. the oven temperature of GC) by using GC–MS to analyze aqueous products, more characterizations such as pyrolysis GC– MS or LC–MS are recommended to further better understand these aqueous products in terms of molecular structures. Fig. 2 presents data on the nitrogen concentrations of the aqueous products. It is possible to notice that both total nitrogen (TN) and total ammonia nitrogen (TAN) presented the same pattern. There is no obvious trend in these parameters related as a function of residence time, but when the temperature was raised, the nitrogen concentrations generally decreased. It was reported by Yenigün and Demirel (2006) that above threshold concentrations (1100 mg L1), free ammonia nitrogen is a powerful inhibitor for anaerobic digestion, and it can easily cause process instability. This was observed by the decrease in both biogas and methane yields, eventually leading to failure of the reactor. According to Speece

Please cite this article in press as: Tommaso, G., et al. Chemical characterization and anaerobic biodegradability of hydrothermal liquefaction aqueous products from mixed-culture wastewater algae. Bioresour. Technol. (2014), http://dx.doi.org/10.1016/j.biortech.2014.10.011

5

G. Tommaso et al. / Bioresource Technology xxx (2014) xxx–xxx

Fig. 2. Influence of residence time and temperature on the nitrogen concentrations in the aqueous product from HTL using AW biomass as feedstock: total nitrogen (j), total ammonia nitrogen ( ).

(1996), TAN toxicity may present a problem with feedstocks that contain high ammonium concentrations or ammonium precursors. In the present work, it was possible to observe high TAN concentration and even higher TN concentrations. This indicates the presence of ammonium precursors. The ionization of NH3 is controlled by pH, thus pH control, is important to maintain neutral pH values, in order to avoid un-ionized ammonia (which is more inhibitory) to reach inhibitory levels. In addition, considering the high concentrations obtained both for TN and TAN in the HTL-ap, the necessity of further dilution are essential if anaerobic digestion is considered as a tool for this aqueous product stabilization. 3.1.1. BMP assays Fig. 3 depicts the influence of HTL reaction temperature on cumulative methane production as measured in BMPs assays. Table 4 presents the parameters estimated through the nonlinear curve fitting using Eq. (1). It is possible to see that when the flasks were fed with HTL-ap obtained at the lowest temperatures, 260 °C and 280 °C, the methane production occurred without a noticeable lag phase. Nevertheless, those were the conditions in which smallest specific cumulative productions were observed. This occurrence may be related to the presence of relatively small amounts of inhibitory compounds, such as cyclic hydrocarbons and phenol, but it also could signify the presence of recalcitrant molecules, such as ketones (Speece, 1996) and cyclic amines. In general, cyclic

Fig. 3. Influence of the temperatures on the accumulated specific methanogenic production. Experimental values: 260 °C (j), 280 °C (d), 300 °C (N), 320 °C (.) and model (—).

Table 4 Values predicted by the modified Gompertz equation to specific methane production potential (PCH4 ), specific methane production rate (k), and duration of the lag phase (k). Temperature (°C)

PCH4 (mmol/SSV)

k (mmol/SSV h1)

k (h)

260 280 300 320

8935.6 8159.0 13464.2 16304.8

30.6 26.6 20.6 108.9

– – 175 109.6

amines and their derivatives are recalcitrant in terms of anaerobic treatability (Johnson et al., 2003) and are therefore undesirable in this case. In addition, the presence of Octahydro-1H-indene, must be highlighted (Supplementary data), because it does not have oxygen in its carbon skeleton. According to Field et al. (1995) such kind of molecules are not indicated for anaerobic digestion. The longest lag phase (k) was observed in the process fed with HTL-ap obtained at 300 °C. According to Table 3, HTL-ap obtained at 300 °C presented higher RPA associated with cyclic amines, amides, organic acids, and cyclic hydrocarbons such as Benzene derivatives (e.g. fluorobiphenyl), when compared to other HTL-ap obtained at the same retention time but at different temperatures. Benzene is one of the hydrocarbon petroleum constituents. It is considered the most toxic and persistent of all petroleum components. In addition, the aromaticity of benzene makes it exhibit a stronger stabilization and thus it is recalcitrant to oxidization and degradation (Johnson et al., 2003). Its derivatives, such as fluorobiphenyl (FBP) in this case, are very toxic molecules, which are highly recalcitrant and difficult to degrade. Although recalcitrant, FBP degradation is thermodynamically possible under anaerobic conditions, and previous literature revealed that FBP could only be biotransformed with specific metabolites such as special enzymes or bacteria (Parsons et al. (2008)). Nevertheless, the removal efficiency of FBP may be improved by adding electron 2 acceptors (e.g. NO 3 or SO4 ) to increase the reducing power of anaerobic systems (Selesi and Meckenstock, 2009; UribeJongbloed and Bishop, 2007). Only HTL-ap at 300 °C, 1-h retention time contained cyclic hydrocarbons identified in a small, yet still important RPA. On the other hand, among the retention times higher than zero (Table 2), 1 h was the condition that resulted in the smallest RPA associated with hydrocarbons. This suggests that the retention time should be adjusted to this value. According to Heider et al. (1999) the enzymes required for the anaerobic metabolism of hydrocarbons are more substrate-specific when compared to enzymes involved in aerobic metabolism of such compounds. In addition, enzymes required for the anaerobic

Please cite this article in press as: Tommaso, G., et al. Chemical characterization and anaerobic biodegradability of hydrothermal liquefaction aqueous products from mixed-culture wastewater algae. Bioresour. Technol. (2014), http://dx.doi.org/10.1016/j.biortech.2014.10.011

6

G. Tommaso et al. / Bioresource Technology xxx (2014) xxx–xxx

metabolism of hydrocarbons are only produced by the organisms in response to the presence of the compound. In this way, after a long period of adaptation (k = 175 h), the anaerobic consortium was able to overcome such difficult conditions, and it showed accumulate specific methane production 57% higher than the average found when the substrate was composed by the HTL-ap obtained at 260 and 280 °C. This result is probably also related to the presence of the highest %RPA associated to organic acids, from which 34.63% was considered short chain fatty acids. Although short chain fatty acids are known to be easily degradable compounds, high concentrations of acetate can cause thermodynamic infeasibility for methanogenesis (Speece, 1996). For instance, biodegradability of organic acids from the petrochemical industry can be as low as 6% (Razo-Flores et al., 2006). In addition, some organic acids can inhibit acetoclastic methanogens for a period of time before recovery. For example, acrylic acid can be further metabolized throughout long periods of time, such as 40 days, even with concentrations as low as 100 mg L1 (Stewart et al., 1995). These statements allow the inference that the higher %RPA related to organic acids, at first, probably contribute to the enlargement of the lag phase. However, after biomass adaptation, the higher %RPA also contributes to the increase of the cumulative methane production. When the most drastic temperature condition (320 °C in this study) was applied to the HTL process, HTL-ap presented the higher RPA associated to phenol derivatives (Table 3). Nevertheless, Fig. 3 shows that such compounds caused a moderate lag phase (k = 109.6 h) in the methanogenic production. In addition, this condition resulted in higher values both for PCH4 and specific methane production rate (k). According to Chen et al. (2008), many organic chemicals that are toxic to anaerobic systems at higher concentrations can serve as a source of food to anaerobic microorganisms at lower concentrations. Phenol is one such chemical that could be removed in plug flow reactors, even in sequencing batch reactors (Rosenkranz et al., 2013), which configuration provides microbial exposition to high concentrations at the beginning of the cycles. Although difficult to be treated, phenol was found to be biodegradable both in mesophilic and thermophilic municipal solid waste anaerobic digestion. Its biodegradation led to its mineralization as revealed by the CH4 and CO2 13C enrichment measured in the produced biogas (Limam et al., 2013). The presence of a moderate lag phase and the highest cumulative methane production when HTL-ap obtained at 320 °C was digested may be also justified due to the fact that the organic matter partitioned from bio-crude oil to aqueous phase was increased in this HTL condition, which was revealed by the increase of the COD values (Fig. 1). In such a situation, easily degradable organic matter could be produced. Previously, Deguchi et al. (2006) found that cellulose undergoes crystalline-to-amorphous transformation in water at around 320 °C and 25 MPa, and, recrystallization was not observed when the system was cooled. As the AW composition had 14.4% of cellulosic material (Chen et al., 2014b), its structure could be degraded due to the high temperature (320 °C) used in the HTL process. Another occurrence that has to be pointed out is the higher RPA associated with urea, found in HTL-ap produced at such condition. Urea is an easily usable organic nitrogen source, which is commonly utilized in the composition of culture medias for anaerobic biomass cultivation and synthetic substrates for biomass acclimatization. It is also important to mention that the inoculum used at the present work was previously adapted with a synthetic substrate containing urea as a nitrogen source. Thus, the presence of urea instead of cyclic nitrogenous compounds in the HTL-ap, verified by GC–MS analysis, certainly was a benefit to the anaerobic consortium. As a consequence of the results obtained, there is a need to refine the reaction conditions. It is suggested to fine tune the HTL

reaction temperature in the range of 300–320 °C, so the HTL-ap can be more favorable for anaerobic digestions and generate more methane. According to Chen et al. (2014b) the HTL process conducted with a 1.0 h reaction time at 300 °C resulted in the highest oil yield, while the bio-crude oil converted at 320 °C contained more light components, which is more advantageous for future upgrades to transportation fuels. This fact reveals that higher temperature may be favored in terms of both biodegradability of aqueous products and bio-crude oil quality. Table 5 displays values of important metabolite concentrations at the end of the BMP assays, the COD conversion rates, and the data on methane production as well. It is possible to see that TAN concentrations ranged from 2.0 to 5.7 times higher than the concentration at the beginning of the processes, which indicates organic nitrogen conversion took place during the anaerobic digestion process. Important to comment that according to Speece (1996) and Yenigün and Demirel (2006), TAN concentrations did not reach inhibitory levels (commonly shown around 1700–1800 mg L1) at the practice pH values. Despite this fact, the COD removal rates ranged from 44% to 61% after 45 days. The values found for the anaerobic biodegradability consistently showed the same pattern observed for COD removal. It is important to highlight that, considering the COD removal rates, all the cumulative methane productions were below the theoretical methane potential (Speece, 1996), indicating that no overproduction was registered. From the acids observed at the samples GC–MS characterization, propionic acid was no longer observed in the digested HTLap samples (Table 5). This result indicates that propionic acid may have been used as a precursor of the acetogenesis process. According to Speece (1996), this reaction is thermodynamically favorable only at low values of the partial pressure of H2, which indicates that the hydrogenotrophic methanogenesis occurred, even at very low rates. Acetate was verified in the flasks fed with HTL-aps obtained at 260, 300 and 320 °C. In the flask fed with HTL-ap produced at 280 °C, as no substance eluted at the same time as acetate (Table 5), it is possible to infer that all the acetate added or produced was consumed, which indicates that at this condition acetoclastic methanogenic production occurred. On the other hand, a residual concentration present at all the other conditions may be indicative that this process was somehow impaired or limited. Speece (1996) stated that acetate accumulation in anaerobic reactors probably occurs due to kinetic limitations and due to the absence of optimal conditions for the methanogenic microorganisms. Table 5 also shows that butyrate was observed when higher HTL temperatures (300 and 320 °C) were practiced. It is well known that butyrate is also a common substrate for the acetogenic step (Speece, 1996). According to Kim et al. (1996), butyrate oxidizers are sensitive to halogenated aromatics, and acetogenesis may become the rate-limiting step in such situations. Elevated levels

Table 5 Removal efficiencies, anaerobic biodegradability, pH, and concentrations of important metabolites at the end of the Biomethane potential assays. Parameters

COD removal (%) pH NH3 (mg L1) Acetate (mg L1) Butyrate (mg L1) Valerate (mg L1) Relative CH4 production (%) Anaerobic biodegradability (%)

Condition (°C) 260

280

300

320

61 7.43 713 87 Nd 118 72 45

56 7.53 991.7 Nd Nd 151 68 35

44 7.71 851 142 223 311 83 37

61 7.45 721 91 236 253 89 84

Please cite this article in press as: Tommaso, G., et al. Chemical characterization and anaerobic biodegradability of hydrothermal liquefaction aqueous products from mixed-culture wastewater algae. Bioresour. Technol. (2014), http://dx.doi.org/10.1016/j.biortech.2014.10.011

G. Tommaso et al. / Bioresource Technology xxx (2014) xxx–xxx

of butyrate were reported by Pind et al. (2003) during elevated levels of valerate, and those were the conditions in which the highest final concentration of valerate was found. Actually, in the present study, valerate was present in all flasks, and since one of its main degradation products, propionate (Pind et al., 2003), was completely absent, it is possible to infer that valerate oxidizers were impaired. As observed by Pham et al. (2013) for Chinese hamsters ovary cells, this study demonstrates that HTL-ap presents toxic or inhibitory effects on the anaerobic consortium. Nevertheless, given the volumes of HTL-ap produced (Chen et al., 2014b), and the methane production potential found here, the energy production through methane generation is feasible. From these experiments, it was clear that anaerobic digestion of HTL-ap can be successfully conducted, and it has significant advantages in terms of energy generation and organic matter removal. The process, however, requires optimization since the conversion rates were small, which would lead to relatively long digestion times and large reactor sizes. More studies are recommended in order to find possibilities to promote the reduction of the inhibitory effect caused by HTL-ap. Microbial adaptation, microbiological enrichment, co-digestion, biofilm formation or adsorbents addition are recommended by the literature (Chen et al., 2008), and in this way, interesting results were found by Zhou (2011), who observed positive effects of activated carbon in the anaerobic stabilization of HTL-ap from swine manure HTL conversion. 4. Conclusions Considering minimal COD values for the HTL-ap, the best HTL operating condition would use 1 h of reaction time and 300 °C. This condition produced HTL-ap that caused both the highest lag phase for methane production and the smallest production rate. Nevertheless, after overcoming adaptation issues, this HTL-ap led to the second highest accumulated specific methane production. Cyclic hydrocarbons and phenol were observed in HTL-ap samples that occasioned longer lag phases in BMP assays. Butyrate oxidation was likely impaired when BMP assays were conducted using HTL-ap produced at higher temperatures, while valerate was observed at the end of all BMP assays. Acknowledgements The first author is grateful to FAPESP (Fundação para o Amparo à Pesquisa do Estado de São Paulo) for the fellowship (Process n. 2012-18064/2). The second author appreciates the financial support from Ministry of Education of Republic of China (Taiwan) and ERM (Environmental Resource Management) foundation. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.biortech.2014.10. 011. References Angelidaki, I., Alves, M., Bolzonella, D., Borzacconi, L., Campos, J.L., Guwi, A.J., Kalyuzhnyi, S., Jenicek, P., van Lier, J.B., 2009. Defining the biomethane potential (BMP) of solid organic wastes and energy crops: a proposed protocol for batch assays. Water Sci. Technol. 59, 927–934. APHA, 1998. American Public Health Association, Standard Methods for Examination of Water and Wastewater, 20th ed. Washington DC. Appleford, J.M., 2005. Analyses of the products from the continuous hydrothermal conversion process to produce oil from swine manure (M.S. thesis). in: Agricultrual and Biological Engineering, Vol. M.S., University of Illinois at Urbana-Champaign, Urbana.

7

Chen, Y., Cheng, J.J., Creamer, K.S., 2008. Inhibition of anaerobic digestion process: a review. Bioresour. Technol. 99, 4044–4406. Chen, W.-H., Chen, S.-Y., Khanal, S.K., Sung, S., 2006. Kinetic study of biological hydrogen production by anaerobic fermentation. Int. J. Hydrogen Energy 31, 2170–2178. Chen, W.-T., Zhang, Y., Zhang, J., Schideman, L., Yu, G., Zhang, P., Minarick, M., 2014a. Co-liquefaction of swine manure and mixed-culture algal biomass from a wastewater treatment system to produce bio-crude oil. Appl. Energy 128, 209– 216. Chen, W.-T., Zhang, Y., Zhang, J., Yu, G., Schideman, L.C., Zhang, P., Minarick, M., 2014b. Hydrothermal liquefaction of mixed-culture algal biomass from wastewater treatment system into bio-crude oil. Bioresour. Technol. 152, 130–139. Deguchi, S., Tsujii, K., Horikoshi, K., 2006. Cooking cellulose in hot and compressed water. Chem. Commun. 2006, 3293–3295. Demirbasß, A., 2000. Effect of lignin content on aqueous liquefaction products of biomass. Energy Convers. Manage. 41, 1601–1607. Field, J.A., Stams, A.J.M., Kato, M., Schraa, G., 1995. Enhanced biodegradation of aromatic pollutant in coculture of anaerobic and aerobic bacterial consortia. A. Van Leeuw. J. Microbiol. 67, 67–77. He, B.J., Zhang, Y., Yin, Y., Funk, T.L., Riskowski, G.L., 2000. Operating temperature and retention time effects on the hydrothermal process of swine manure. Trans. ASABE 43, 1821–1826. Heider, J., Spormann, A.M., Beller, H.R., Widdel, F., 1999. Anaerobic bacterial metabolism of hydrocarbons. FEMS Microbiol. Rev. 22, 459–473. Johnson, S.J., Woolhouse, K.J., Prommer, H., Barry, D.A., Christofi, N., 2003. Contribution of anaerobic microbial activity to natural attenuation of benzene in groundwater. Eng. Geol. 70, 343–349. Kim, I.S., Young, J.C., Tabak, H.H., 1996. Impact of chloroanilines on hydrogenotrophic methanogenesis in ethanol-enriched cultures. Water Res. 30, 601–612. Limam, I., Mezni, M., Guenne, A., Madigou, C., Driss, M.R., Bouchez, T., Mazéas, L., 2013. Evaluation of biodegradability of phenol and bisphenol A during mesophilic and thermophilic municipal solid waste anaerobic digestion using 13C-labeled contaminants. Chemosphere 90, 512–520. Marcilla, A., Catalá, L., García-Quesada, J.C., Valdés, F.J., Hernández, M.R., 2013. A review of thermochemical conversion of microalgae. Renew. Sust. Energy Rev. 27, 11–19. Moreno-Andrade, G., Buitrón, G., 2004. Influence of the origin of the inoculum and the acclimation strategy on the degradation of 4-chlorophenol. Bioresour. Technol. 94 (2), 215–221. National Research Council, 2012. Sustainable Development of Algal Biofuels in the United States. The National Academies Press, Washington, DC. Parsons, J.R., Sáez, M., Dolfing, J., de Voogt, P., 2008. Biodegradation of perfluorinated compounds. Rev. Environ. Contam. Toxicol. 196, 53–71. Peterson, A.A., Vogel, F., Lachance, R.P., Fröling, M., Antal Jr., M.J., Tester, J.W., 2008. Thermochemical biofuel production in hydrothermal media: a review of sub and supercritical water technologies. Energy Environ. Sci. 1, 32–65. Pham, M., Schideman, L., Scott, J., Rajagopalan, N., Plewa, M.J., 2013. Chemical and biological characterization of wastewater generated from hydrothermal liquefaction of spirulina. Environ. Sci. Technol. 47, 2131–2138. Pind, P.F., Angelidaki, I., Ahring, B.K., 2003. Dynamics of the anaerobic process: effects of volatile fatty acids. Biotechnol. Bioeng. 82, 791–801. Razo-Flores, E., Macarie, H., Morier, F., 2006. Application of biological treatment systems for chemical and petrochemical wastewaters. In: Cervantes, F.J., Pavlostathis, S.G., Van Haandel, A.C. (Eds.), Advanced Biological Treatment Processes for Industrial Waste-waters: Principles & Applications. IWA Publishing, London, pp. 267–297. Roberts, G.W., Fortier, M.-O.P., Sturm, B.S.M., Stagg-Williams, S.M., 2013. Promising pathway for algal biofuels through wastewater cultivation and hydrothermal conversion. Energy Fuels 27, 857–867. Rosenkranz, F., Cabrol, L., Carballa, M., Donoso-Bravo, A., Cruz, L., Ruiz-Filippi, G., Chamy, R., Lema, J.M., 2013. Relationship between phenol degradation efficiency and microbial community structure in an Anaerobic SBR. Water Res. 47, 6739–6749. Selesi, D., Meckenstock, R.U., 2009. Anaerobic degradation of the aromatic hydrocarbon biphenyl by a sulfate-reducing enrichment culture. FEMS Microbiol. Ecol. 68, 86–93. Speece, R.E., 1996. Anaerobic Biotechnology for Industrial Wastewater. Archae Press, Nashville, p. 394. Stewart, J.M., Bhattacharya, S.K., Madura, R.L., Mason, S.H., Schonberg, J.C., 1995. Anaerobic treatability of selected organic toxicants in petrochemical wastes. Water Res. 29, 2730–2738. Uribe-Jongbloed, A., Bishop, P.L., 2007. Comparative study of PAH removal efficiency under absence of molecular oxygen: effect of electron acceptor and hydrodynamic conditions. J. Environ. Eng. Sci. 6, 367–376. Yu, G., Zhang, Y., Schideman, L., Funk, T., Wang, Z., 2011. Hydrothermal liquefaction of low lipid content algae into bio-crude oil. Trans. ASABE 54, 239–246. Valdez, P.J., Savage, P.E., 2013. Reaction network for the hydrothermal liquefaction of Nannochloropsis sp.. Algal Res. 2, 416–425. Vardon, D.R., Sharma, B.K., Scott, J., Yu, G., Wang, Z., Schideman, L., Zhang, Y., Strathmann, T.J., 2011. Chemical properties of biocrude oil from the hydrothermal liquefaction of Spirulina algae, swine manure, and digested anaerobic sludge. Bioresour. Technol. 102 (17), 8295–8303.

Please cite this article in press as: Tommaso, G., et al. Chemical characterization and anaerobic biodegradability of hydrothermal liquefaction aqueous products from mixed-culture wastewater algae. Bioresour. Technol. (2014), http://dx.doi.org/10.1016/j.biortech.2014.10.011

8

G. Tommaso et al. / Bioresource Technology xxx (2014) xxx–xxx

Yenigün, O., Demirel, B., 2006. Ammonia inhibition in anaerobic digestion: a review. Process Biochem. 48, 901–911. Zhang, J., Chen, W.-T., Zhang, P., Luo, Z., Zhang, Y., 2013. Hydrothermal liquefaction of Chlorella pyrenoidosa in sub- and supercritical ethanol with heterogeneous catalysts. Bioresour. Technol. 133, 389–397.

Zhou, Y., 2011. Improving algal biofuel production via a novel integrated biofuel production-wastewater treatment process with nutrient recycling (M. S. thesis). in: Agricultrual and Biological Engineering, vol. M.S., University of Illinois at Urbana-Champaign, Urbana, p. 75

Please cite this article in press as: Tommaso, G., et al. Chemical characterization and anaerobic biodegradability of hydrothermal liquefaction aqueous products from mixed-culture wastewater algae. Bioresour. Technol. (2014), http://dx.doi.org/10.1016/j.biortech.2014.10.011