Scenedesmus obliquus mediated brewery wastewater remediation and CO2 biofixation for green energy purposes

Journal of Cleaner Production 165 (2017) 1316e1327

Contents lists available at ScienceDirect

Journal of Cleaner Production journal homepage: www.elsevier.com/locate/jclepro

Scenedesmus obliquus mediated brewery wastewater remediation and CO2 biofixation for green energy purposes Alice Ferreira a, Belina Ribeiro a, Paula A.S.S. Marques a, Ana F. Ferreira b, Ana Paula Dias b, Helena M. Pinheiro c, Alberto Reis a, Luisa Gouveia a, * a

LNEG, National Laboratory of Energy and Geology I.P./Bioenergy Unit, Estrada do Paço do Lumiar 22, 1649-038, Lisbon, Portugal LAETA, IDMEC, Instituto Superior T
ecnico, Universidade de Lisboa, Av. Rovisco Pais, 1, 1049-001, Lisboa, Portugal IBB, Institute for Bioengineering and Biosciences, Instituto Superior T
ecnico, Departamento de Bioengenharia, Universidade de Lisboa, Av. Rovisco Pais, 1049-001, Lisboa, Portugal

b c

a r t i c l e i n f o

a b s t r a c t

Article history: Received 7 March 2017 Received in revised form 28 July 2017 Accepted 29 July 2017 Available online 31 July 2017

Microalgae can be used for wastewater bioremediation with simultaneous CO2 biofixation producing valuable biomass. Wastewater from a brewery was treated using the Scenedesmus obliquus microalga in bubble-column photobioreactors (PBRs). The PBRs were fed with ambient air and the effect of a 10% (v/v) brewery CO2 supplement was studied. The PBRs were inoculated and a range of mean hydraulic residence time (HRT) values were tested (2.1e10.4 days). The maximum ash-free dry weight (AFDW) biomass productivity was obtained for a HRT of 3.5 days (0.29 d
1 dilution rate; 0.2 g L
1 d
1 (in terms of AFDW). The highest pollutant removal efficiencies were 92.9 and 88.5% for ammonia and total nitrogen, respectively, 40.8% for phosphorus, and 61.9% for COD. Except a dilution rate of 0.48 d
1 trial, the treated wastewater always met the Portuguese legislation quality standard for discharge into natural water bodies. Aiming to simultaneously maximize biomass volumetric productivity, CO2 biofixation rate and wastewater treatment efficiency, while minimizing residence time, 0.29 d
1 represents the optimal dilution rate value. The potential of the produced Scenedesmus obliquus biomass was evaluated for the generation of biohydrogen through dark fermentation with Enterobacter aerogenes, and of bio-oil, bio-char and bio-gas through a pyrolysis process. The yields obtained were 67.1 mL H2 g
1 (in terms of volatile solids - VS) for bioH2 and 64, 30 and 6% for bio-oil, bio-char and bio-gas, respectively (dry mass content (%) calculated over freeze dryer biomass basis). © 2017 Elsevier Ltd. All rights reserved.

Keywords: Scenedesmus obliquus Brewery wastewater Dark fermentation Biohydrogen Pyrolysis Bio-oil

1. Introduction The brewing industry holds a strategic economic position with an annual world production up to 2 billion hectoliters in 2015 (Statista Inc., 2017). The brewing process generates large amounts of wastewater (WW) and solid wastes. It is widely estimated that for 1 L of beer, about 3e10 L of waste effluent is generated (Simate et al., 2011). Thus, water management, waste treatment and disposal have become relevant cost factors and critical aspects in the operation of brewery plants. All plants aim to minimize the waste disposal costs while the legislation regarding this becomes

* Corresponding author. E-mail address: [email protected] (L. Gouveia). http://dx.doi.org/10.1016/j.jclepro.2017.07.232 0959-6526/© 2017 Elsevier Ltd. All rights reserved.

stricter (Fillaudeau et al., 2006). Wastewater, namely, should be treated according to discharge regulations set by government entities. This wastewater usually has a high chemical oxygen demand (COD) load from the organic components present (e.g. sugars, soluble starch, ethanol, volatile fatty acids) which are easily biodegradable (Simate et al., 2011). There are several conventional technologies available for the treatment of effluents from agro-food industries, such as physical, chemical, or biological processes, which allow the removal of the organic load and of inorganic nutrients, especially nitrogen (ammonia and/or nitrate) and phosphorus (Raposo et al., 2010). The first treatment steps consist mainly of physical unit operations, to remove coarse and fine suspended solid matter, leaving most of the dissolved pollutants (Simate et al., 2011). Through chemical methods, pollutant removal can be enhanced, i.e., different chemicals can be added to the brewery wastewater to alter its chemistry

A. Ferreira et al. / Journal of Cleaner Production 165 (2017) 1316e1327

(e.g. pH adjustment, coagulation/flocculation, advanced oxidation). However, while chemical and physical technologies are available to remove nutrients, they consume significant amounts of energy and chemicals, which makes them costly for both industries and municipalities (Christenson and Sims, 2011). Conventional activated sludge processes are efficient biotreatment options for COD and nutrients but they involve oxygen supply through mechanical aeration, which is energy demanding and usually accounts for 45e75% of a wastewater treatment plant's total energy costs (Razzak et al., 2013). In this context, microalgae-based biotreatment seems to be quite promising for combining biomass growth with the biological removal of wastewater contaminants in a less expensive and ecologically safer way, with the added benefits of resource recovery and recycling (Christenson and Sims, 2011). The treatment occurs via the O2 produced photosynthetically by the microalgae, which is used for the bio-oxidation of organic matter and NHþ 4 by the bacteria in the algal-bacterial consortium. N and P assimilation into this biomass also occurs, using both autotrophic and heterotrophic metabolic pathways (Razzak et al., 2013). It is also efficient in pathogen removal due to the high pH and O2 concentrations achieved in the medium, mediated by photosynthesis. The use of microalgae for wastewater treatment presents major advantages, such as: a reduced need for forced aeration, as the oxygen required for aerobic bacteria is provided by the microalgae photosynthesis, thus reducing energy demand; reduction in hazardous sludge formation (e.g., laden with heavy metals and pathogens); reduction of greenhouse gas (GHG) emissions; reduction of costs; production of useful algal biomass-energy with recycling of the nutrients present in the wastewater (Gouveia et al., 2016; Batista et al., 2015). Few studies have been published to this date concerning the potential of using microalgae for brewery wastewater treatment. Some works (Darpito et al., 2015; Mata et al., 2014; Farooq et al., 2013) proved that under certain conditions microalgae can grow in brewery wastewater. In addition, a recent study by Wu et al. (2017) showed that CO2 injection into the culture promoted algae growth in brewery wastewater. Moreover, the resulting microalgae biomass can be further upgraded for several purposes such as the production of bioenergy, animal feed, and fertilizers, among others. Scenedesmus obliquus is a versatile, very robust and fast growing organism that can be easily cultivated in different wastewaters and environmental conditions (e.g., Gupta et al., 2016; Posadas et al., 2015; Batista et al., 2015). The biomass of Scenedesmus also proved to be a good feedstock for several of the bioenergy vectors, such as biodiesel (Gouveia and Oliveira, 2009), bioethanol (e.g., Miranda et al., 2012), biohydrogen (e.g., Batista et al., 2015), biogas and bio-oil. In this context, the goals of the present work were: i) to evaluate the potential of the Scenedesmus obliquus microalga for treating brewery wastes, both liquid wastewater and gaseous CO2, from the company Sociedade Central de Cervejas e Bebidas, S.A., Portugal; ii) to study the influence of CO2 supplementation on the growth of Scenedesmus obliquus and the brewery wastewater treatment performance; iii) to assess the potential of the produced biomass for conversion to bioenergy vectors, namely biohydrogen through dark fermentation using Enterobacter aerogenes and bio-oil by pyrolysis.

1317

2. Methods 2.1. Brewery wastewater treatment 2.1.1. Feed effluent The brewery wastewater was supplied and collected from the Sociedade Central de Cervejas e Bebidas (SCC) brewery at Vialonga, Portugal, after a secondary treatment stage by anaerobic digestion. This effluent was stored in a refrigeration chamber at 10  C throughout the experimental period (about a month). The wastewater was allowed to settle for 24 h at room temperature to remove most of the large amount of sludge it contained. Only the supernatant was used as the algal culture medium to avoid clogging problems in the feeding tubes. The average composition of the collected and settled wastewater is given in Table 1. The long settling period may have also led to some extent of chemical and biological transformation, such as ammonification, as can be seen from some of the parameters in Table 1 (e.g., pH and N-NH3). 2.1.2. Microalga The microalga used was Scenedesmus obliquus (ACOI 204/07) from the Coimbra University Culture Collection, Portugal. S. obliquus was maintained in the appropriate culture medium (Bristol), at room temperature, under orbital shaking (150 rpm) and continuous artificial light at 72 W (4 lamp x 18 W/865). Bristol medium contains NaNO3 (0.250 g L
1), KH2PO4 (0.175 g L
1), K2HPO4 (0.075 g L
1), MgSO4$7H2O (0.075 g L
1), Fe-EDTA (0.060 g L
1), CaCl2$2H2O (0.033 g L
1), NaCl2 (0.025 g L
1) and trace elements solution CHU (1 mL L
1) (Vonshak, 1986). 2.1.3. Photobioreactor setup The photobioreactor system used for cultivation was composed of five cylindrical PBRs (14 cm in diameter and 40 cm in height) operating in parallel, that were continuously fed with brewery wastewater from a 20-L polycarbonate carboy containing the settled brewery effluent. Feeding was done through a series of silicone rubber tubes attached to glass tubes at both ends, one of the latter immersed in the effluent in the feed carboy and the other inserted in each PBR. A working volume of 5 L was maintained by overflow at a fixed level and the outlet streams were collected into plastic containers, through silicone rubber tubing. The supplied air was enriched with CO2, purified from the brewery process, through a ”Y” connection in the feed line fitted with a no-return valve, reaching the culture through an aquarium Elite A983 air diffuser (Hagen), placed at the center bottom of each PBR. Air was supplied at a flow rate of 0.1 vvm (L L
1 min
1), measured with an American Meter Company flow meter. The culture was continuously illuminated by 3 fluorescent lamps (Philips, 36 W) assembled at one of the lateral sides of the PBRs, with an average light intensity of 3.2 klux (measured with a Phywe Lux-meter). Fig. 1 shows a scheme of the culture system described above. 2.1.4. Photobioreactor operation Each PBR was inoculated with 120 mL of Scenedesmus culture, diluted with 2 L of brewery effluent in order to reach an optical density value (OD) of 0.2. Firstly it was operated in fed-batch mode for biomass acclimatization. This biomass included mainly Scenedesmus with other algal species (Chlorella) in symbiosis with bacteria. This fed-batch phase lasted for 17 days until the OD value reached 1 (March~ ao, 2016). The PBRs were agitated solely by the filtered compressed air stream and maintained at room temperature (23e25  C). The PBR culture volume was gradually increased with added brewery wastewater, up to a working volume of 5 L, 6 days after starting the cultivation. As soon as the culture growth

1318

A. Ferreira et al. / Journal of Cleaner Production 165 (2017) 1316e1327

Table 1 Brewery effluent average composition of the different collections. Mean values and their standard deviation are shown for at least two replicates. Effluent

pH

TSS (mg TSS L
1)

N-NH3 (mg N L
1)

TKN (mg N L
1)


1 PO3
4 (mg P L )


1 P- PO3
4 (mg P L )

P2O5 (mg P L
1)

COD (mg O2 L
1)

Raw Decanted

7.24 8.85

515 ± 55 17.5 ± 1.4

24.5 ± 2.1 29.4 ± 1.4

120.4 ± 8.4 72.8 ± 0.0

41.25 37.75

13.50 12.25

30.75 28.25

628 ± 18 226 ± 0.0

supernatant, respectively.

Removal ð%Þ ¼ 100  ðCin
Cout Þ = Cin

(1)

2.2. Microalga growth monitoring Microalgae growth was monitored twice a day, during week days, by measuring the optical density (OD) of culture samples, at 540 nm (Rocha et al., 2003), against distilled water, using a Hitachi U-2000 spectrophotometer. To overcome possible inconsistencies in this measurement due to other components of the growth medium, the biomass dry weight (DW) and ash free dry weight (AFDW) were determined at the end of the assay, when the culture reached a steady-state, detected through OD measurements. The culture pH was also measured on the same schedule using a laboratory pH meter (InoLab WTW) in order to check whether a value between 7 and 8 was maintained. Fig. 1. Schematic representation of the PBR system for the cultivation assays. Adapted from March~ ao (2016).

started to enter the stationary phase (after 17 days), the operational mode was switched from fed-batch to continuous, by starting the continuous feed of brewery effluent to the culture. At the same time, the filtered compressed air was replaced by air enriched in CO2, at 10% (v/v). The air flow rate was adjusted manually. In order to test six different hydraulic retention time (HRT) values (2.1 ± 0.5, 3.5 ± 0.4, 5.3 ± 0.3, 6.5 ± 0.4, 8.7 ± 0.2 and 10.4 ± 0.2 days), different brewery effluent feeding rates were imposed manually in each PBR, using Hoffman tubing clamps. The continuous operation regimen was maintained for a further period of time, ensuring a minimum of 3-fold the retention time value in each of the six runs, so that steady-state conditions could be attained. 2.1.5. Treatment performance monitoring The effluent from SCC was characterized in terms of pH, COD, and nitrogen (ammonia and Kjeldahl nitrogen) and phosphorus concentrations, before and after the 24-h settling step (Table 1). In order to evaluate the efficiency of the wastewater treatment by microalgae, the same analyses were performed on the different cultures at the end of the cultivation runs, after biomass settling and filtration. Ammonia nitrogen was quantified by titration after a preliminary distillation step based on standard methods 4500-NH3 B and C (Clesceri et al., 1998). The “Kjeldahl nitrogen” was determined by a modified Kjeldahl method adapted from the standard method 4500-Norg B (Clesceri et al., 1998). Phosphorus determination was done using a commercial kit with the Phosver 3 (ascorbic acid) method using Powder Pillows (Spectrophotometer HACH DR/2010). The COD determination was carried out through a closed reflux spectrophotometric method using test kits from Hach Lange, Hach 21258-51 (0e150 mg L
1 COD). The overall removal rate for each of the analyzed components was determined according to Equation (1), where Cin and Cout are the component concentrations in the feed effluent and in the filtered culture liquid

2.3. Microalga biomass harvesting and characterization 2.3.1. Biomass harvesting and general processing At the end of the assay, the collected culture volume was left to settle for 48 h at room temperature in graduated cylinders for a primary concentration of the biomass through sedimentation. The concentrated biomass was recovered by centrifugation (11 300  g) at 15  C for 10 min (Heraeus multifuge 3SRþ centrifuge, Thermo Scientific) and freeze-dried (Heto Power Dry LL3000, Thermo Scientific). The freeze-dried biomass was stored at
18  C for further characterization. The microalgal biomass was characterized in terms of its crude protein, total sugars, chlorophylls (a and b) and fatty acid contents. In general, all analyses were performed in duplicate (triplicate for total sugars) according to A.O.A.C. (2006) methods. Moisture was determined by drying in an oven at 105  C until constant mass. Total ash was determined by incineration at 550  C in a muffle furnace. The volatile solids content was determined by the difference between the mass readings at the end of the moisture and total ash measurements. 2.3.2. Protein content The Lowry method (Lowry et al., 1951) was used to measure the protein content of freeze-dried biomass samples. The latter were previously treated with NaOH (0.1 M) and diluted with water, so that the calculation was done according to Equation (2).

Protein content ð% w=wÞ ¼ 100  ðC  V  D=mbiomass Þ (2) In Equation (2), C represents the concentration result given by the Lowry method on the pre-treated biomass sample (mg L
1), V is the volume of NaOH solution used to pre-treat this sample (L), D is the dilution factor in the dilution step with water, and mbiomass is the mass of the original, freeze-dried biomass sample (mg). 2.3.3. Total sugar content The sugars present in microalgal cells (300-mg samples of

A. Ferreira et al. / Journal of Cleaner Production 165 (2017) 1316e1327

freeze-dried biomass) were first extracted by quantitative acid hydrolysis according Hoebler et al. (1989). Following extraction, the total sugar content was determined by the colorimetric method of the phenol-sulfuric reagent (DuBois et al., 1956). A calibration curve was obtained using standard sugar (glucose) solutions.

2.3.4. Chlorophyll content Microalga culture samples freshly collected from the PBR (2 mL) were first concentrated by centrifugation (2449  g) during 10 min (Sigma 2-6E, Sartorius). Then, 2 mL of acetone (99.5%, SigmaAldrich) and glass beads were added to the tube and the mixture was vortex agitated during 2 min, followed by a 2 min immersion in an ice bath. The mixture was then centrifuged (2449  g) for 20 min, and the supernatant collected. This step was repeated until a colourless supernatant was obtained. All the tubes were covered with aluminum foil during processing to prevent pigment degradation by light exposure. The total volume of the collected extract phases was quantified. Chlorophylls a and b were quantified in the extracts by spectrophotometry (Hitachi U-2000), by measuring the absorbances (A) at 630, 647, 664 and 691 nm against acetone. The chlorophyll concentrations in the extracts (Chl a and Chl b) were then calculated using Equations (3) and (4) (Ritchie, 2008).

Chla



 mg L
1 ¼
0:3319  A630
1:485  A647 þ 11:9442  A664
1:4306  A691

Chlb



(3)

 mg L
1 ¼
1:2825  A630
19:8839  A647 þ 4:8860  A664
2:3416  A691

(4)

Finally, the chlorophyll contents (Chl a and Chl b) in the algal cells (mg g
1) were calculated by dividing the amounts of chlorophylls (mg) in the extracts by the cell dry mass (g) in the culture samples used to obtained them. The total chlorophyll content was obtained by adding up the values for Chl a and Chl b.

2.3.5. Fatty acid content The fatty acid composition of the biomass samples was analyzed, in duplicate, by gas chromatography (GC). For this, the fatty acids were first transesterified by the method of Lepage and Roy (1986) with modifications. Portions of 100 mg of freeze-dried microalgae were added to Pyrex tubes with Teflon-sealed screw caps. Then, 2 mL of a methanol/acetyl chloride (95:5 v/v) mixture and 0.2 mL of heptadecanoic acid in petroleum benzin 60e80  C (5 mg mL
1), as internal standard solution, were also added. The mixture was heated at 80  C for 1 h and was cooled to room temperature before being diluted with 1 mL of water and 2 mL of nheptane. The tube contents were left to stand until phase separation. The upper layer, containing the fatty acid methyl ester (FAME) derivatives ready for injection in the gas chromatograph, was recovered, dried over anhydrous Na2SO4 and collected in vials. The FAMEs were analyzed in a CP-3800 GC (Varian, USA) equipped with a 30-m SUPELCOWAX 10 capillary column (film 0.32 mm) with helium as carrier gas at a constant flow rate of 3.5 mL min
1. The injector and detector (flame ionization) temperatures were 250 and 280  C, respectively. The split ratio was 1:50 for the first 5 min and 1:10 for the remaining time. The column temperature programme started at 200  C for 8 min, increased up to 240  C at a rate of 4  C min
1, and was held at that value for 16 min. Individual fatty acid contents were calculated as a percentage of the total fatty acids present in the sample, determined from the chromatographic peak areas.

1319

2.4. Determination of biomass productivity In continuous culture at steady-state, the value of m equals that of the dilution rate (D, d
1), if biomass decay can be considered negligible, and the volumetric biomass productivity (PX, g L
1 d
1) can be determined by Equation (5), where X (g L
1) is the biomass concentration at steady-state.

PX ¼ D  X

(5)

2.5. Biohydrogen production from Scenedesmus obliquus biomass 2.5.1. Bacterial strain and culture conditions The fermentative bacterium Enterobacter aerogenes, ATCC 13408 Sputum (American Type Culture Collection, Manassas, USA) was used for the production of biohydrogen in the conditions described by Batista et al. (2015). The bacterial culture was kept at 4  C in solid CASO Agar (from MERCK: 15 g L
1 peptone from casein, 5 g L
1 peptone from soymeal, 5 g L
1 sodium chloride and 15 g L
1 agareagar) and grown in a synthetic growth media (20 g L
1 peptone solution with 5 g L
1 NaCl). A preculture grown overnight at 34  C and 150 rpm orbital shaking was inoculated at 1% (v/v) and the fermentation was conducted at 34  C for 6 h, 150 rpm. The fermentation medium (FM) for the bioH2 production assays (basal fermentation medium) contained K2HPO4 (7.0 g L
1), KH2PO4 (5.5 g L
1), tryptone (5 g L
1), yeast extract (5 g L
1), (NH4)2SO4 (1.0 g L
1), MgSO4$7H2O (0.25 g L
1), CaCl2$2H2O (0.021 g L
1), Na2MoO4$2H2O (0.12 g L
1), nicotinic acid (0.02 g L
1), Na2SeO3 (0.172 mg L
1), NiCl2 (0.02 mg L
1), with a pH of around 6.8. Both media were sterilized before use. 2.5.2. Dark fermentation assays Batch fermentation experiments were performed in 159-mL glass flasks, with a volumetric phase ratio (gaseous headspace/ liquid fermentation medium) of 5. The biomass of S. obliquus harvested after the brewery wastewater treatment was used as a substrate at an initial concentration of 2.5 g L
1. The bioreactors containing the fermentation medium and the biomass were previously sterilized by autoclaving (121  C for 15 min). The reactors were purged by bubbling with N2 gas for 2 min to eliminate O2, before inoculation with an exponentially growing E. aerogenes culture at 10% (v/v), and sealed with butyl rubber stoppers. The incubation was carried out under orbital shaking (220 rpm), at 34  C, for 6 h. The assays were performed in triplicate and control fermentation assays, without microalgal biomass, were also performed for comparison. The results are expressed as averages. 2.5.3. Analysis of the gas phase from the dark fermentation The concentrations of H2 and CO2 in the reactors’ headspace were determined by GC in a Varian 430-GC gas chromatograph equipped with a TCD and a fused silica column (select Permanent gases/CO2-Molsieve 5A/Borabound Q tandem #CP 7430). The headspace (HS) samples (0.5 mL) were collected through the butyl rubber stoppers by means of a gas-tight syringe for GC analysis. The gas injector and column were at 80  C and the detector at 120  C. Argon was the carrier gas at 32.4 mL/min flow rate. The H2 produced was determined through calibration curves, previously determined for various ranges of standardized concentrations of H2 and CO2. For each peak area in the GC chromatogram the calibration curve with the most adequate range of concentrations was used to determine the volumes of H2 and CO2 produced. These volumes were then used to calculate the yield and purity of the resulting

1320

A. Ferreira et al. / Journal of Cleaner Production 165 (2017) 1316e1327

bioH2. The hydrogen production yields e mLH2 g
1 (in terms of VS) and mL H2 L
1 (in terms of FM) e were calculated by dividing the total volume of hydrogen produced by the amount of S. obliquus biomass (substrate), in terms of its volatile solids content, and by the volume of fermentation medium, respectively.

2.6. Bio-oil production from the Scenedesmus obliquus biomass 2.6.1. Thermogravimetric analysis (TGA) The microalgae biomass was characterized by TGA. This technique allowed the understanding of the behavior of this biomass in pyrolysis conditions, and for that it required only a small amount of biomass (around 20e30 mg). The pyrolysis temperature was also selected through TGA. This analysis was performed in a TGA apparatus (NETZSCH STA 409 PC/PG), simulating pyrolysis conditions, with a weighting precision of ±0.01% and sensitivity in the mass measurements of 0.1 mg. The samples were heated from 30  C to 1100  C at a heating rate of 25  C min
1 under an atmosphere of nitrogen (99.996%). The temperature was measured with an experimental uncertainty of ±1  C. The TG curve represents the evolution of the mass (loss) as a function of the temperature. The DTG is the derivative of the TG curve and represents evolution of the rate of mass variation (% min
1).

2.6.2. Pyrolysis system The pyrolysis of freeze-dried biomass was performed using a fixed bed quartz reactor (16 mm in internal diameter and 150 mm in length), under a N2 flow. The quartz reactor was filled with carborundum (SiC), as inert to reduce the void volume, and a thin layer of glass wool supported the biomass. The reactor was heated using a cylindrical oven equipped with a PID temperature controller. Fig. 2 shows a schematic diagram of the pyrolysis apparatus. More details on the pyrolysis tests procedure are given elsewhere (Silva et al., 2016).

2.6.3. Pyrolysis process The biomass pyrolysis temperature was selected based on a previous thermogravimetric analysis, as described in the previous section. Following the results obtained, the chosen pyrolysis reference temperature was 475  C. The pyrolysis test was started by placing a certain amount of microalgae (about 2.5 g) in the quartz reactor. Nitrogen was used as the carrier gas and its reference flow rate was monitored using a mass flowmeter. The N2 flow rate was set to 200 mL min
1. The pyrolysis was performed for 15 min. After that the reactor was removed from the oven and the bio-char was collected. The bio-oil was recovered by washing the reactor with acetone followed by evaporation of the acetone under reduced pressure in a rotary evaporator. The acetone used in this process could thus be recycled for further use. The yields of pyrolysis products (bio-char, bio-oil and bio-gas) were quantified. The weighted masses of bio-char and bio-oil were the basis for calculating the yields of these products. The bio-gas yield (wet) was computed from a mass balance. The bio-oil was characterized by infrared spectroscopy, as explained in the next section. However, bio-char and bio-gas were not further characterized. 2.6.4. Characterization of the bio-oil by infrared spectroscopy (HATR-FTIR) The bio-oil produced from pyrolysis was further characterized by infrared spectroscopy (HATR-FTIR) in reflectance mode. Spectra were obtained with a resolution of 16 cm
1, using a spectrophotometer from BOMEN (FTLA200-100, ABB). This equipment has a horizontal total attenuated reflection accessory (HATR), from PIKE Technologies, with a ZnSe crystal. The infrared spectra were recorded at room temperature in the range of 3725e725 cm
1. Sixty-four scans were accumulated for each spectrum to obtain an acceptable signal-to-noise ratio. Additionally, the reflectance signal (R) was corrected using the Kubelka-Munk (KM) function (Miljkovi
c et al., 2012). The functional group identification was made taking into account previously published results. Annex I shows the FTIR bands and functional group attribution. 3. Results and discussion 3.1. Brewery wastewater treatment

Fig. 2. Schematic diagram of the fixed bed pyrolysis apparatus (Silva et al., 2016).

The wastewater bioremediation capacity of the present system was assessed by measuring the nutrient (N and P) contents and organic load (COD) remaining in the brewery effluent at the end of the microalgae growth trials. The results are shown in Table 2. In this Table, the emission value limits (EVLs) imposed by the Portuguese legislation (Decree-Law No. 236/98, 1998) are also given, in order to assess the possibility of discharging the effluents, after microalgae separation and harvesting, into a natural water body without potential harmful consequences to the environment. Maximum removal efficiencies were also calculated for ammonia nitrogen, total Kjeldahl nitrogen, phosphorus and COD (Table 3). According to Tables 2 and 3, the treatment at the lowest dilution rate, i.e. 0.10 d
1, allowed the highest removal efficiencies for most pollutants such as N and P. On the other hand, it showed the lowest removal value for organic load (COD), i.e., 50%. However, if the aim is to achieve efficient wastewater treatment but also optimal biomass production, one should select the most favorable conditions for both requirements. This means that the treatment at a D value of 0.29 d
1 represents the best compromise, since it achieved the highest volumetric productivity (217 ± 6 mgAFDW L
1 d
1) and allowed high removal efficiencies for all the nutrients. Removal

A. Ferreira et al. / Journal of Cleaner Production 165 (2017) 1316e1327

1321

Table 2 Characterization of the brewery wastewater after biological treatment with Scenedesmus obliquus for all 6 continuous mode trials. The first row corresponds to the initial effluent (decanted) fed to the culture. For all the parameters, mean values and their standard deviation are shown for at least two data points at the steady-state condition. HRT (d)

D (d
1)

Feed brewery effluent (decanted) 2.1 0.48 3.5 0.29 5.3 0.19 6.5 0.15 8.7 0.11 10.4 0.10 Legal EVLs (DecreeLaw No. 236/98, 1998)

N-NH3 (mg N L
1)

TKN (mg N L
1)


1 PO3
) 4 (mg P L


1 P-PO3
4 (mg P L )

P2O5 (mg P L
1)

COD (mg O2 L
1)

pH

29.4 ± 1.4

72.8 ± 0.0

37.75

12.25

28.25

226 ± 0

8.85

± ± ± ± ± ±

18.2 ± 1.4 8.4 ± 1.4 11.2 ± 0.0 14.0 ± 0.0 14.0 ± 2.8 8.4 ± 2.8

35.00 29.25 28.75 30.75 23.50 22.25

11.50 9.50 9.50 10.0 7.75 7.25

26.25 21.75 21.50 23.00 17.75 16.75

100 ± 4 86.2 ± 0.0 88.4 ± 2.2 94.8 ± 0.0 100 ± 4 113 ± 0

8.74 8.72 8.81 8.62 8.55 8.60

15

e

10

e

150

6e9

8.4 2.8 4.2 5.6 2.8 2.1

1.4 1.4 0.0 1.4 1.4 0.7

10

Table 3 Nutrient maximum removal efficiency. HRT (d)

2.1 3.5 5.3 6.5 8.7 10.4

D (d
1)

0.48 0.29 0.19 0.15 0.11 0.10

Maximum removal efficiency (%) N-NH3

TKN

P

COD

71.4 90.5 85.7 81.0 90.5 92.9

73.1 88.5 84.6 80.8 80.8 88.5

6.1 22.4 22.4 18.4 26.7 40.8

55.8 61.9 60.9 58.1 55.7 50.0

rates were 91 and 89% for ammonia and total Kjeldahl nitrogen, respectively, 22% for P and 62% for COD. This corresponds to the final values of 2.8 ± 0.0 mg N L
1 N-NH3 (from an initial value of 29.4 ± 1.4 mg N L
1), 8.4 ± 1.4 mg N L
1 TKN (from an initial value of 72.8 ± 0.0 mg N L
1), 9.5 mg P L
1 P-PO34 (from an initial value of 12.3 mg P L
1) and 86.2 ± 0.0 mg O2 L
1 (from an initial value of 226 ± 0 mg O2 L
1). For all dilution rate values, there was an efficient removal of both forms of nitrogen, with removal values ranging from 71 to 93% and 73e89%, respectively, for ammonia and total Kjeldahl nitrogen (Table 3). These values suggest that the biological treatment performed by S. obliquus was very efficient for nitrogen removal. These results are higher than those reported by Mata et al. (2012) (11e24.4% for N removal), but are comparable to the ones reported by Raposo et al. (2010) and Darpito et al. (2015) (above 85%), in batch mode brewery wastewater treatment using Scenedesmus and Chlorella, respectively. The present results are also in accordance with those obtained by Gouveia et al. (2016) for domestic wastewater treatment using Scenedesmus (95%). Concerning P removal, efficiencies were lower, which reveals that, unlike nitrogen, the treatment was not very efficient in the removal of this nutrient (6e40.8%) (Table 3). These values are significantly lower than those achieved by Raposo et al. (2010), Darpito et al. (2015) and Gouveia et al. (2016), 54e66%, 90% and 92% respectively. For continuous trials performed by McGinn et al. (2012) using S. obliquus for domestic wastewater treatment, near complete removal of both nitrogen and phosphorus was achieved, even after a retention time of only 1.33 days (which corresponds to a dilution rate of 0.75 d
1). Regarding COD removal, the efficiency values are between 50 and 62%. These values are contained in the range of values obtained by Mata et al. (2012) (13.3e66.8%), but are slightly higher, in general. Also, these values are in accordance with the 63% value achieved by Gouveia et al. (2016). Lastly, comparing the present results with those presented by

~o (2016), who run a similar experiment but with no CO2 Marcha supplementation, it is clear that a higher removal efficiency was obtained for total nitrogen, but not for COD. It is also important to note that, in general, the present work was able to achieve higher removal efficiencies for phosphorus. Nonetheless, there were no major differences between the removal efficiencies obtained in the two studies. The results presented in Table 2 show that, excluding the D ¼ 0.48 d
1 condition, all the other runs resulted in treated effluents that can be discharged into natural water bodies as they meet the legal discharge requirements, revealing an effective biological wastewater treatment capacity in S. obliquus. The higher dilution rate corresponds to the lowest retention time, clearly exceeding the culture's nutrient removal capacity, thus explaining the lower removal efficiencies. However, when developing a technology for treating large amounts of wastewater which are continuously being generated, it is important to use the lowest retention time possible, thus reducing bioreactor volumes, to achieve the EVLs imposed by law. Nowadays, one major obstacle that still hinders the widespread application of the algal treatment process is the relatively long HRT it requires to achieve efficient pollutant removal, when compared to traditional activated sludge processes, which can achieve efficient overall reduction of COD, nitrogen, and phosphorus within a much shorter process time (4e6 h) (Wang et al., 2010). 3.2. Microalga growth In the continuous mode trials, 6 mean HRT values were tested, namely, 2.1, 3.5, 5.3, 6.5, 8.7 and 10.4 days, and the trial for each one ended when the steady-state was established. The steady-state condition was identified through biomass concentration measurements, i.e., when these values did not vary significantly in successive days. The microalgae growth was followed by measuring the ODl¼540nm on broth samples collected throughout the operational time of the 6 different continuous assays until each reached its steady-state. Table 4 summarizes the Scenedesmus obliquus growth results for each mean HRT in terms of AFDW, biomass productivity (Px), operational time up to the steady-state and pH. The steady-state times for all the cultures are in accordance with literature values, which observed that a microalgal culture reaches its steady-state after 3e5 residence times in continuous operation (Fonseca and Teixeira, 2007). At the end of the trial, the biomass concentration and volumetric productivity were calculated (Table 4), and they are represented as a function of the dilution rate in Fig. 3.

1322

A. Ferreira et al. / Journal of Cleaner Production 165 (2017) 1316e1327

Table 4 Summary of growth related values for Scenedesmus obliquus grown in brewery wastewater at different HRT. For ash-free dry weight (AFDW) and biomass productivity (PX), mean values and their standard deviation are shown for at least two data points at the steady-state condition. HRT (d)

D (d
1)

tsteady-state (d)

AFDW (g L
1)

2.1 3.5 5.3 6.5 8.7 10.4

0.48 0.29 0.19 0.15 0.11 0.10

9 14 24 22 29 32

0.26 0.76 0.88 0.58 0.95 0.81

± ± ± ± ± ±

0.00 0.02 0.14 0.08 0.07 0.05

PX (g L
1 d
1)

pH

173 ± 0 217 ± 6 167 ± 27 83.9 ± 11.6 109 ± 8 77.9 ± 4.8

7.09 8.22 7.93 7.17 6.85 7.08

According to Fig. 3, AFDW shows strong dependence on the dilution rate. For higher dilution rates, the AFDW value is markedly reduced, from 0.88 g L
1 at 0.19 d
1 to 0.36 g L
1 at 0.48 d
1. For the lower D range a fluctuation in the values of AFDW was registered. This reduction of biomass concentration at high D values is reported in the literature for general chemostat operation (Fonseca and Teixeira, 2007). The highest biomass concentration was obtained for a dilution rate of 0.11 d
1 (0.95 ± 0.07 g L
1). Regarding biomass productivity, an increase in its value was observed up to a maximum at D ¼ 0.29 d
1 (217.1 ± 5.7 mg L
1 d
1), decreasing for higher D values. Thus, D ¼ 0.29 d
1 would be the optimal dilution rate in terms of biomass productivity. This behavior is also reported in Tang et al. (2012), with a decrease in biomass productivity for dilution rates both lower and higher than the optimal value. These results, although lower, are comparable to the volumetric productivities achieved by McGinn et al. (2012) when cultivating Scenedesmus in secondary municipal wastewater in continuous mode at a dilution rate of 0.75 d
1 (234 and 267 mg L
1 d
1). Also, ~o the present results are comparable to those obtained by Marcha (2016) (between 80.5 ± 3.0 and 224.3 ± 22.9 mg L
1 d
1 (in terms of AFDW).

3.3. Biomass characterization The biochemical composition of Scenedesmus obliquus biomass, cultivated in brewery wastewater with different HRT values, is shown in Table 5. The protein content of the obtained S. obliquus (37e40%) biomass is in accordance with those reported in the literature, and slightly higher than some, namely 20.4% for S. obliquus grown in Bristol medium (Batista et al., 2014), and 32.7% for S. obliquus used for domestic wastewater treatment (Gouveia et al., 2016). This could be explained by the higher content of nitrogen present in the brewery wastewater, when compared to Bristol medium and domestic wastewater, which could be predominantly used by the microalga for the synthesis of proteins (Beuckels et al., 2015). Regarding the content in total sugars (20e26%), the obtained values are higher than those reported by Becker (2007) (10e17%) and the 11.7% measured in S. obliquus grown in domestic wastewater (Batista et al., 2015), but lower than the 30.7 and 31.8% reported for S. obliquus grown in Bristol medium (Batista et al., 2014; Miranda et al., 2012, respectively). In relation to chlorophyll contents, all values are within a range between 26 and 40 mg/g (11e32 mg L
1 for Chla). The values achieved, although lower, are comparable to those achieved by Raposo et al. (2010) for Chlorella cultivated in brewery wastewater (around 45 mg L
1). According to Veloso et al. (1991), the Chla/AFDW ratio is a clear indicator of the physiological state and “health” of algal cells. A value lower than 1% means that the population is at risk of crashing, which can be due to predators or the lack of available nutrients. In this context, the values shown in Fig. 4 are significantly higher than 1%, which suggests that nutrient requirements are fulfilled by the brewery effluent. These high ratio values can also be explained by the supplementation of the culture with CO2 providing an additional feed of nutrients, which has a positive effect on the stability of the microalgae culture, also mentioned by Veloso et al. (1991). The fatty acid profile was also assessed for all cultures and the results are presented in Fig. 5 and Table 6. All microalgal lipids are mainly composed of unsaturated fatty acids (45e66%), but a

Fig. 3. Biomass concentration (AFDW) and volumetric productivity (PX) of the Scenedesmus obliquus cultures at steady-state as a function of dilution rate in the 6 PBRs. Mean values and their standard deviation (error bars) are shown for at least two data points at the steady-state condition.

A. Ferreira et al. / Journal of Cleaner Production 165 (2017) 1316e1327

1323

Table 5 Composition of Scenedesmus obliquus biomass grown in brewery wastewater with CO2 supplementation. For all the parameters, mean values and their standard deviation are shown for at least two data points at the steady-state condition. HRT (d)

D (d
1)

Chlorophyll a (mg L
1)

2.1 3.5 5.3 6.5 8.7 10.4

0.48 0.29 0.19 0.15 0.11 0.10

11.8 21.9 19.9 18.6 17.6 22.2

± ± ± ± ± ±

0.4 0.7 2.8 1.4 0.0 0.5

Chlorophyll aþb (mg L
1) 31.1 36.7 28.8 40.0 27.0 36.0

± ± ± ± ± ±

Crude protein (%)

2.8 1.2 3.4 2.0 0.7 1.0

39.3 34.2 40.4 37.7 37.6 28.9

± ± ± ± ± ±

0.6 2.0 4.5 1.8 0.7 2.5

Total sugars (%) 23.4 24.0 20.1 23.1 26.1 23.2

± ± ± ± ± ±

0.1 3.2 0.8 0.7 1.3 0.3

significant percentage of palmitic acid (C16:0) is also present (14e25%). In general, the dominant fatty acids are palmitic and linoleic acids, which is in accordance with the results achieved by previous studies (e.g., Gouveia and Oliveira, 2009). 3.4. Effect of supplementation with brewery CO2 on Scenedesmus obliquus growth and treatment performance

Fig. 4. Chl a/AFDW ratio of steady-state cultures of Scenedesmus obliquus grown in brewery effluent at different dilution rates. Mean values and their standard deviation (error bars) are shown for at least two data points at the steady-state condition.

Fig. 5. Fatty acids profile for Scenedesmus obliquus grown in brewery wastewater at different dilution rates.

Since the C:N ratio in wastewaters, such as those from brewery processes, is generally lower than that exhibited by microalgal biomass, the microalgae cultivation and wastewater treatment performances can be enhanced by CO2 supplementation (Park and Craggs, 2011). In fact, an elevation of the CO2 levels made available to the cell culture is known to improve the specific growth rate and photosynthetic activity of microalgae. Thus, S. obliquus was grown in brewery wastewater, with (present work) and without CO2 ~o, 2016). The comparison between these supplementation (Marcha two sets of results provides insights regarding the effect of CO2 on this microalgae culture system. Table 7 presents the removal efficiencies obtained in these two studies for the pollutants present in the brewery effluents. The results indicate a positive effect on treatment efficiency when a CO2 supplement is added to the culture. Additionally, ~o (2016) (around 9), it is observing the pH ranges shown in Marcha important to take into account pH mediated nutrient removal processes. According to Heubeck et al. (2007), for pH values higher than 8, most of the pollutant removal occurs by pH mediated processes such as phosphate precipitation and ammonia volatilisation. These processes are more rapid than microalgal assimilation, and so they act first on the removal of pollutants. The addition of CO2 promotes the decrease in culture pH to value ranges in which the pH mediated processes are inhibited. This can be seen as an advantage, since both phosphate precipitation and ammonia volatilisation represent losses of important nutrients, which, in microalgae-based wastewater treatment, can be incorporated into microalgae biomass and thus enhance their biochemical value. Moreover, feeding CO2 to the culture overcomes both carbon limitation and pH inhibition of microalgal growth (Heubeck et al., 2007). Regarding microalgae growth, higher biomass productivities were expected with CO2 supplementation since the levels of dissolved carbon in the culture would be higher, ensuring no risk of

Table 6 Fatty acids composition of Scenedesmus obliquus grown in brewery wastewater at different retention times. The percentage of saturated, unsaturated and monounsaturated fatty acids corresponds to the proportion if each one relatively to the total amount of fatty acids. HRT (d)

D (d
1)

Saturated (% w/w)

Unsaturated (% w/w)

Monounsaturated (% w/w)

Esters (%)

2.1 3.5 5.3 6.5 8.7 10.4

0.48 0.29 0.19 0.15 0.11 0.10

33.0 31.4 27.7 30.2 27.0 24.3

46.3 48.1 64.3 57.4 65.7 63.3

13.2 14.9 21.2 18.4 40.9 15.7

9.4 6.0 3.5 6.4 6.1 7.9

1324

A. Ferreira et al. / Journal of Cleaner Production 165 (2017) 1316e1327

Table 7 Maximum removal efficiency achieved after biological treatment with Scenedesmus obliquus, grown in brewery wastewater with and without CO2 supplementation. The results ~o (2016). obtained for the growth with no CO2 supplementation were taken from Marcha HRT (d)

D (d
1)

Maximum removal efficiency (%) TKN

N-NH3

2.1 3.5 5.3 6.5 8.7 10.4

0.48 0.29 0.19 0.15 0.11 0.10

P

COD

No CO2

10% (v/v) CO2

No CO2

10% (v/v) CO2

No CO2

10% (v/v) CO2

No CO2

10% (v/v) CO2

67 87 91 96 97 97

71 91 86 81 91 93

65 76 59 76 76 76

73 89 85 81 81 89

14 23 17 26 9 13

6 22 22 18 27 41

57 74 66 66 55 66

56 62 61 58 56 50

The bold values refer to the highest removal efficiency, thus was the choosen HRT and Dilution rate.

carbon limitation. However, the results in Table 7 do not show enough evidence to support this observation, since not all PBR in the present study achieved higher volumetric productivities than ~o (2016). This was probably due to some those obtained by Marcha inhibition of the microalgal photosynthesis at low pH and nonoptimal environmental conditions as a result of the acidification of the stromal compartment of the chloroplast (Solovchenko and Khozin-Goldberg, 2013). In the CO2-supplemented runs of the present study, occasional periods of low pH were observed, possibly due to CO2 overfeeding. Regarding chlorophyll contents, it is clear that, for the microalgae grown in the medium supplemented with CO2, there was a higher synthesis of this pigment. This observation was corroborated by Sutherland et al. (2015), who studied the effect of CO2 addition on microalgae used for wastewater treatment. The higher the percentage of CO2 in the air fed to the culture, the lower were the measured pH values, and consequently the higher was the content of Chla in the biomass (Sutherland et al., 2015). 3.5. Scenedesmus obliquus biomass grown in brewery wastewater as feedstock for hydrogen production by dark fermentation In order to evaluate the potential of the S. obliquus biomass as a substrate in dark fermentation, the process yields (mL H2 g
1 (in terms of VS) and mL H2 L
1 (in terms of FM)) and gas purity (H2/ CO2) were evaluated: 67.1 and 167.8 for brewery wastewater, respectively, with a purity of 1.0 (against the control, using Bristol medium of 57.6 and 135.6, respectively, and a purity of 1.2 (Batista et al., 2014)). The specific bioH2 production yields were around 1.2-fold higher when the microalga was grown in the brewery wastewater, which has environmental (pollution control, renewable energy) and economical (resources recovery, lower costs of waste management) benefits. Moreover, the purity of the produced biogas (CO2 þ H2) was very similar in the two runs, which is a further advantage. 3.6. Bio-oil production from Scenedesmus obliquus biomass grown on brewery wastewater 3.6.1. Thermogravimetric analysis Fig. 6 shows the results of thermogravimetric analysis for the microalga S. obliquus biomass grown in brewery wastewater under different aeration conditions (air and air with CO2). From Fig. 6, the TG/DTG curves show three stages of decomposition for the biomass. Stage 1 corresponds to the removal of moisture, which occurs up to a temperature of 150  C. Stage 2 includes three devolatilization zones from 200 to 450  C: the first zone refers to the intrinsic lipid decomposition (e.g. aldehydes and ketones) at 200e250  C, which is perceptible for the microalgae

grown without the CO2 supplement; the second zone corresponds to the decomposition of proteins and carbohydrates between 250 and 350  C; the last zone is related to lipids decomposition, mainly associated to the break-down of hydrocarbon chains of fatty acids, taking place at temperatures from 350 to 500  C. Finally, stage 3 corresponds to the oxidation of the bio-char in the temperature range between 500 and 750  C. These observations are in agreement with the studies reported by Ferreira et al.
pez-Gonza
lez et al. (2014) and Kebelmann et al. (2013). (2015), Lo Moreover, a final decomposition stage appears between 1000 and 1050  C, which can be assumed to be related to the volatile metal loss and carbonate decomposition (Ferreira et al., 2015). The microalgal biomass cultivated under air has the highest peak in comparison to microalgal biomass cultivated under air with CO2. This means that it presents the highest contents in protein and carbohydrates. It is also possible to conclude that the microalgae grown in brewery wastewater present lower contents of both components than the microalga grown in Bristol medium. This can be explained by the lower C:N ratio value, which is common for wastewaters. 3.6.2. Characterization by HATR-FTIR (infrared spectroscopy) Fig. 7 shows the FTIR spectra of the microalgae examined in this study. It can be seen that all microalgae show small peaks at around 1740 cm
1. The more significant peaks appear at 1700-1500 cm
1 and 1150-950 cm
1. Following the results in Annex I, the peaks located at 1740 cm
1 correspond to the lipids, the peak located at 1025 cm
1 corresponds to the carbohydrates and the peaks located at 1650 and 1540 cm
1 correspond to the proteins (Ferreira et al., 2015). For the S. obliquus biomass cultivated in brewery effluent there is a very high peak at 700-600 cm
1 which represents the S-O stretching vibration of sulphonic components (Pugazhendy, 2012). 3.6.3. Pyrolysis products yields According to the methodology developed by Silva et al. (2016) the potential of S. obliquus biomass, after brewery effluent treatment, was investigated for the production of bio-oil, bio-char and bio-gas. The S. obliquus biomass grown in brewery wastewater and in Bristol medium showed significant differences in the pyrolysis behavior. The pyrolysis process yields (gproduct/gbiomass) obtained for brewery wastewater were 64, 30 and 6 for bio-oil, bio-char and biogas, respectively (dry mass content (%) calculated over freeze dryer biomass basis). Using the Bristol medium, the yield of bio-oil and bio-char production decreased to 58 and 25% and the yield of biogas increased to 17% (Silva et al., 2016). The data for brewery wastewater growth under air with CO2 were not reproducible due to the small amount of sample used in

A. Ferreira et al. / Journal of Cleaner Production 165 (2017) 1316e1327

1325

Fig. 6. TG (%) and DTG (% min
1) curves for the microalga Scenedesmus obliquus biomass grown in different wastewaters (brewery and brewery with CO2).

Fig. 7. FTIR spectra of Scenedesmus obliquus biomass grown in Bristol medium, in brewery wastewater under air, and in brewery wastewater under air supplemented with CO2.

the pyrolysis tests, however the bio-oil was analyzed. In this work, only the bio-oil was analyzed qualitatively. Besides it being the product with the greatest yield, the objective was to assess its properties to verify whether it has the potential for application as, for example, a biofuel. A previous study has shown that the higher heating value (HHV) of the bio-oil obtained from microalgae is around 25e30 MJ/kg and the lower heating value (LHV) is around 23e28 MJ/kg (Silva et al., 2016). The bio-char and bio-gas were not analyzed. Bio-char is widely used for generation of heat and power and as an amendment to soils, serving as a fertilizer and carbon sequestration agent. Also in the form of activated biochar, it holds significant value in various adsorption applications (Amin et al., 2016). The bio-gas can contain valuable low molecular

mass of alkanes or alkenes in addition to CO, CO2, CH4 and H2. 3.6.4. Bio-oil characterization by HATR-FTIR (infrared spectroscopy) Fig. 8 shows the FTIR spectra of the bio-oil produced from the microalgae examined in this study. The bio-oils obtained from S. obliquus should be acidic since the features of alkyl groups in the spectra are weak and thus the bands in the range 1735e1705 cm
1 belonging to carboxylic acids are more intense. This bio-oil shows a marked peak in the range of 1100 cm
1 corresponding to carbohydrates. It is possible to verify that the bio-oil from the microalgae grown in brewery wastewater shows a higher content of some compounds, like carbohydrates and proteins, in relation to the Bristol-

1326

A. Ferreira et al. / Journal of Cleaner Production 165 (2017) 1316e1327

Fig. 8. FTIR spectra of bio-oil produced from the biomass of Scenedesmus obliquus grown in Bristol medium, in brewery wastewater under air, and in brewery wastewater under air supplemented with CO2.

grown biomass bio-oil. Comparing the two brewery-grown microalgae, the bio-oil from the biomass grown without CO2 addition shows a higher content of lipids (1740 cm
1) and proteins. FTIR bands were assigned to specific molecular groups based on biochemical standards and published studies (Stehfest et al., 2005). FTIR bands were attributed to y(C]O) stretching of amides from proteins (amide I, ~1655 cm
1); d(NeH) bending of amides from proteins (amide II, ~1545 cm
1); das(CH2) and das(CH3) bending of methyl from proteins (~1455 cm
1); and ds(CH2) and ds(CH3) bending of methyl and (CeO) stretching of COOe groups (~1380 cm
1) and >P]O stretching, associated with phosphorus compounds (~1260 cm
1). Two bands were of particular interest, the band at 1740 cm
1 which was associated with y(C]O) of ester groups, primarily from lipids and fatty acids, and the region from 1200 to 950 cm
1 associated with y(CeOeC) stretching of polysaccharides (Dean et al., 2010). The bio-oil obtained from brewery biomass grown with and without CO2, showed similar compositions presenting analogous FTIR spectra.

(ELAC2014/BEE0357) GREENBIOREFINERY-Processing of brewery wastes with microalgae for producing valuable compounds, COST Action 1408 EUALGAE- European network for algal-bioproducts and through the funding received by iBBdInstitute for Bioengi~o para a Cie ^ncia e a Tecnoneering and Biosciences from Fundaça logia, Portugal (UID/BIO/04565/2013), and from Programa Operacional Regional de Lisboa 2020 (Project No. 007317). The authors would like to thank to the supported by Fundaç~ ao para a ^ncia e a Tecnologia (FCT), through IDMEC, under LAETA-UID/ Cie EMS/50022/2013. Ana F. Ferreira is pleased to acknowledge the FCT for the Post-Doctoral financial support through grant no. SFRH/ BPD/95098/2013. The authors would like to thank Sociedade Central Cervejas e Bebidas (SCC), Portugal, to the access of brewery effluents, Ana Cristina Oliveira (LNEG) for fatty acid analysis assistance and Graça
rcia Santos (LNEG) for microalgae culture Gomes (LNEG) and Nate maintenance and laboratory assistance. Appendix A. Supplementary data

4. Conclusions This work proved the effectiveness of Scenedesmus obliquus to treat the brewery effluents (liquid and gaseous), with a production of valuable biomass useful for bioenergy production, such as an energy carrier BioH2 (67.1 mL H2 g
1 (in terms of VS)), or the three products bio-oil (64%), bio-char (30%) and bio-gas (6%) from a single pyrolysis process. This work proved that it is possible to integrate several sustainable processes, resulting in a closed loop system that allowed an efficient nutrient recycle (wastes) and the full use of the energy content of algal biomass. Acknowledgments The study was supported by the Project ERANETLAC/0001/2014

Supplementary data related to this article can be found at http:// dx.doi.org/10.1016/j.jclepro.2017.07.232. References Amin, F.R., Huang, Y., He, Y., Zhang, R., Liu, G., Chen, C., 2016. Biochar applications and modern techniques for characterization. Clean. Technol. Environ. Policy 18 (5), 1457e1473. http://dx.doi.org/10.1007/s10098-016-1218-8. A.O.A.C., 2006. Official Methods of Analysis, eighteenth ed. Association of the Official Analytical Chemists, Gaithersburg MD, USA. Batista, A.P., Ambrosano, L., Graça, S., Sousa, C., Marques, P.A.S.S., Ribeiro, B., Botrel, E.P., Neto, P.C., Gouveia, L., 2015. Combining urban wastewater treatment with biohydrogen production e an integrated microalgae-based approach. Bioresour. Technol. 184, 230e235. http://dx.doi.org/10.1016/ j.biortech.2014.10.064. Batista, A.P., Moura, P., Marques, P.A.S.S., Ortigueira, J., Alves, L., Gouveia, L., 2014. Scenedesmus obliquus as a feedstock for bio-hydrogen production by Enterobacter aerogenes and Clostridium butyricum by dark fermentation. Fuel 117,

A. Ferreira et al. / Journal of Cleaner Production 165 (2017) 1316e1327 537e543. http://dx.doi.org/10.1016/j.fuel.2013.09.077. Becker, E., 2007. Micro-algae as a source of protein. Biotechnol. Adv. 25 (2), 207e210. http://dx.doi.org/10.1016/j.biotechadv.2006.11.002. Beuckels, A., Smolders, E., Muylaert, K., 2015. Nitrogen availability influences phosphorus removal in microalgae-based wastewater treatment. Water Res. 77, 98e106. http://dx.doi.org/10.1016/j.watres.2015.03.018. Christenson, L., Sims, R., 2011. Production and harvesting of microalgae for wastewater treatment, biofuels, and bioproducts. Biotechnol. Adv. 29 (6), 686e702. http://dx.doi.org/10.1016/j.biotechadv.2011.05.015. Clesceri, L.S., Greenberg, A.E., Eaton, A.D. (Eds.), 1998. Standard Methods for the Examination of Water and Wastewater, twentieth ed. American Public Health Association, Washington DC. Darpito, C., Shin, W.-S., Jeon, S., Lee, H., Nam, K., Kwon, J.-H., Yang, J.-W., 2015. Cultivation of Chlorella protothecoides in anaerobically treated brewery wastewater for cost-effective biodiesel production. Bioprocess Biosyst. Eng. 38 (3), 523e530. http://dx.doi.org/10.1007/s00449-014-1292-4. Dean, A.P., Sigee, D.C., Estrada, B., Pittman, J.K., 2010. Using FTIR spectroscopy for rapid determination of lipid accumulation in response to nitrogen limitation in freshwater microalgae. Bioresour. Technol. 101 (12), 4499e4507. http:// dx.doi.org/10.1016/j.biortech.2010.01.065. Decree-Law No. 236/98, 1998. Decree-law No. 236/98 of the Portuguese Ministry of the Environment of 1 August Establishing Water Quality Standards. Di
ario da
rie. República, pp. 3676e3722. I Se DuBois, M., Gilles, K.A., Hamilton, J.K., Rebers, P.A., Smith, F., 1956. Colorimetric method for determination of sugars and related substances. Anal. Chem. 28 (3), 350e356. http://dx.doi.org/10.1021/ac60111a017. Farooq, W., Lee, Y.-C., Ryu, B.-G., Kim, B.-H., Kim, H.-S., Choi, Y.-E., Yang, J.-W., 2013. Two-stage cultivation of two Chlorella sp. strains by simultaneous treatment of brewery wastewater and maximizing lipid productivity. Bioresour. Technol. 132, 230e238. http://dx.doi.org/10.1016/j.biortech.2013.01.034. Ferreira, A., Dias, A.S., Silva, C., Costa, M., 2015. Evaluation of thermochemical properties of raw and extracted microalgae. Energy 92 (Part 3), 365e372. http:// dx.doi.org/10.1016/j.energy.2015.04.078. Fillaudeau, L., Blanpain-Avet, P., Daufin, G., 2006. Water, wastewater and waste management in brewing industries. J. Clean. Prod. 14 (5), 463e471. http:// dx.doi.org/10.1016/j.jclepro.2005.01.002. da Fonseca, M.M., Teixeira, J.A., 2007. Design and operational modes. In: Biological Reactors e Fundaments and Applications (In Portuguese). LIDEL, Lisboa, pp. 27e63. Gouveia, L., Graça, S., Sousa, C., Ambrosano, L., Ribeiro, B., Botrel, E.P., Neto, P.C., Ferreira, A.F., Silva, C.M., 2016. Microalgae biomass production using wastewater: treatment and costs. Scale-up considerations. Algal Res. 16, 167e176. http://dx.doi.org/10.1016/j.algal.2016.03.010. Gouveia, L., Oliveira, A.C., 2009. Microalgae as a raw material for biofuels production. J. Ind. Microbiol. Biotechnol. 36 (2), 269e274. http://dx.doi.org/10.1007/ s10295-008-0495-6. Gupta, S.K., Ansari, F.A., Shriwastav, A., Sahoo, N.K., Rawat, I., Bux, F., 2016. Dual role of Chlorella sorokiniana and Scenedesmus obliquus for comprehensive wastewater treatment and biomass production for bio-fuels. J. Clean. Prod. 115, 255e264. http://dx.doi.org/10.1016/j.jclepro.2015.12.040. Heubeck, S., Craggs, R., Shilton, A., 2007. Influence of CO2 scrubbing from biogas on the treatment performance of a high rate algal pond. Water Sci. Technol. 55 (11), 193e200. http://dx.doi.org/10.2166/wst.2007.358. Hoebler, C., Barry, J.L., David, A., Delort-Laval, J., 1989. Rapid acid hydrolysis of plant cell wall polysaccharides and simplified quantitative determination of their neutral monosaccharides by gas-liquid chromatography. J. Agric. Food Chem. 37 (2), 360e367. http://dx.doi.org/10.1021/jf00086a020. Kebelmann, K., Hornung, A., Karsten, U., Griffith, s. G., 2013. Intermediate pyrolysis and product identification by TGA and Py-GC/MS of green microalgae and their extracted protein and lipid components. Biomass Bioenergy 49, 38e48. http:// dx.doi.org/10.1016/j.biombioe.2012.12.006. Lepage, G., Roy, C.C., 1986. Direct transesterification of all classes of lipids in a onestep reaction. J. Lipid Res. 27 (1), 114e120.
pez-Gonza
lez, D., Fernandez-Lopez, M., Valverde, J., Sanchez-Silva, L., 2014. KiLo netic analysis and thermal characterization of the microalgae combustion process by thermal analysis coupled to mass spectrometry. Appl. Energy 114, 227e237. http://dx.doi.org/10.1016/j.apenergy.2013.09.055. Lowry, O.H., Rosebrough, N.J., Farr, A.L., Randall, R.J., 1951. Protein measurement with the folin phenol reagent. J. Biol. Chem. 193 (1), 265e275. March~ ao, L., 2016. Brewery Wastewater Treatment by Microalgae Using Closed
cnico. Bubble-column Photobioreactors. Master's thesis. Instituto Superior Te Lisbon University. ~es, M., Caetano, N.S., 2012. Parametric study of a Mata, T.M., Melo, A.C., Simo brewery effluent treatment by microalgae Scenedesmus obliquus. Bioresour. Technol. 107, 151e158. http://dx.doi.org/10.1016/j.biortech.2011.12.109. Mata, T.M., Santos, J., Mendes, A.M., Caetano, N.S., Martins, A.A., 2014. Sustainability

1327

evaluation of biodiesel produced from microalgae Chlamydomonas sp grown in brewery wastewater. Chem. Eng. Trans. 37, 823e828 doi:0.3303/cet1437138. McGinn, P.J., Dickinson, K.E., Park, K.C., Whitney, C.G., MacQuarrie, S.P., Black, F.J., Frigon, J.-C., Guiot, S.R., O'Leary, S.J., 2012. Assessment of the bioenergy and bioremediation potentials of the microalga Scenedesmus sp. AMDD cultivated in municipal wastewater effluent in batch and continuous mode. Algal Res. 1 (2), 155e165. http://dx.doi.org/10.1016/j.algal.2012.05.001. Miljkovi
c, M., Bird, B., Diem, M., 2012. Line shape distortion effects in infrared spectroscopy. Analyst 137 (17), 3954e3964. http://dx.doi.org/10.1039/ c2an35582e. Miranda, J., Passarinho, P., Gouveia, L., 2012. Pre-treatment optimization of Scenedesmus obliquus microalga for bioethanol production. Bioresour. Technol. 104, 342e348. http://dx.doi.org/10.1016/j.biortech.2011.10.059. Park, J.B.K., Craggs, R.J., 2011. Algal production in wastewater treatment high rate algal ponds for potential biofuel use. Water Sci. Technol. 63 (10), 2403e2410. http://dx.doi.org/10.2166/wst.2011.200.
mez, C., Acie
n, F.G., Mun ~ oz, R., 2015. Influence of pH Posadas, E., Morales, M.M., Go and CO2 source on the performance of microalgae-based secondary domestic wastewater treatment in outdoors pilot raceways. Chem. Eng. J. 265, 239e248. http://dx.doi.org/10.1016/j.cej.2014.12.059. Pugazhendy, K., 2012. Fourier transform infrared FT-IR spectoroscopic analysis of Spirulina. Int. J. Pharm. Biol. Arch. 3 (4), 969e972. Raposo, M.F.D.J., Oliveira, S.E., Castro, P.M., Bandarra, N.M., Morais, R.M., 2010. On the utilization of microalgae for brewery effluent treatment and possible applications of the produced biomass. J. Inst. Brew. 116 (3), 285e292. http:// dx.doi.org/10.1002/j.2050-0416.2010.tb00433.x. Razzak, S.A., Hossain, M.M., Lucky, R.A., Bassi, A.S., de Lasa, H., 2013. Integrated CO2 capture, wastewater treatment and biofuel production by microalgae culturing - a review. Renew. Sustain. Energy Rev. 27, 622e653. http://dx.doi.org/10.1016/ j.rser.2013.05.063. Ritchie, R.J., 2008. Universal chlorophyll equations for estimating chlorophylls a, b, c, and d and total chlorophylls in natural assemblages of photosynthetic organisms using acetone, methanol, or ethanol solvents. Photosynthetica 46 (1), 115e126. http://dx.doi.org/10.1007/s11099-008-0019-7. Rocha, J.M.S., Garcia, J.E.C., Henriques, M.H.F., 2003. Growth aspects of the marine microalga Nannochloropsis gaditana. Biomol. Eng. 20, 237e242. Silva, C., Ferreira, A., Dias, A.S., Costa, M., 2016. A comparison between microalgae virtual biorefinery arrangements for bio-oil production based on lab-scale results. J. Clean. Prod. 130, 58e67. http://dx.doi.org/10.1016/j.jclepro.2015.09.053. Simate, G.S., Cluett, J., Iyuke, S.E., Musapatika, E.T., Ndlovu, S., Walubita, L.F., Alvarez, A.E., 2011. The treatment of brewery wastewater for reuse: state of the art. Desalination 273 (2e3), 235e247. http://dx.doi.org/10.1016/ j.desal.2011.02.035. Solovchenko, A., Khozin-Goldberg, I., 2013. High-CO2 tolerance in microalgae: possible mechanisms and implications for biotechnology and bioremediation. Biotechnol. Lett. 35 (11), 1745e1752. http://dx.doi.org/10.1007/s10529-0131274-7. Statista Inc, 2017. Worldwide Beer Production, 2015jStatistic [WWW Document]. https://www.statista.com/statistics/270275/worldwide-beer-production/. (Accessed 30 May 2017). Stehfest, K., Toepel, J., Wilhelm, C., 2005. The application of micro-FTIR spectroscopy to analyze nutrient stress-related changes in biomass composition of phytoplankton algae. Plant Physiol. Biochem. 43 (7), 717e726. http://dx.doi.org/ 10.1016/j.plaphy.2005.07.001. Sutherland, D.L., Howard-Williams, C., Turnbull, M.H., Broady, P.A., Craggs, R.J., 2015. The effects of CO2 addition along a pH gradient on wastewater microalgal photo-physiology, biomass production and nutrient removal. Water Res. 70, 9e26. http://dx.doi.org/10.1016/j.watres.2014.10.064. Tang, H., Chen, M., Simon, N.K., Salley, S.O., 2012. Continuous microalgae cultivation in a photobioreactor. Biotechnol. Bioeng. 109 (10), 2468e2474. http:// dx.doi.org/10.1002/bit.24516. Veloso, V., Reis, A., Gouveia, L., Fernandes, H., Empis, J., Novais, J., 1991. Lipid production by Phaeodactylum tricornutum. Bioresour. Technol. 38 (2), 115e119. http://dx.doi.org/10.1016/0960-8524(91)90141-6. Vonshak, A., 1986. Laboratory techniques for the cultivation of microalgae. In: Richmond, A. (Ed.), Handbook of Microalgal Mass Culture. CRC Press, Boca Raton, pp. 117e143. Wang, L., Min, M., Li, Y., Chen, P., Chen, Y., Liu, Y., Wang, Y., Ruan, R., 2010. Cultivation of green algae Chlorella sp. in different wastewaters from municipal wastewater treatment plant. Appl. Biochem. Biotechnol. 162 (4), 1174e1186. http:// dx.doi.org/10.1007/s12010-009-8866-7. Wu, X., Yan, C., Zheng, H., Luo, S., Liu, Y., Li, W., Cheng, Y., Addy, M., Zhou, W., Chen, P., Ruan, R., 2017. Fixing CO2 and treating wastewater from beer brewery using microalgae. J. Biobased Mater. Bioenergy 11 (2), 101e105. http:// dx.doi.org/10.1166/jbmb.2017.1654101.