Introduction

In the last decades, the application of suspended microalgal technology for wastewater treatment has received considerable attention (Li et al. 2010; Abdel-Raouf et al. 2012; Bernstein et al. 2014; Han et al. 2017). The main advantages of this technology rely on the simultaneous removal of pollutants and production of lipids for biodiesel (Berner et al. 2015; Bhaiji 2016) or high value renewable fine chemicals (Spolaore et al. 2006). However, microalgal culture systems have low biomass concentrations and problems with biomass/liquid separation which constitute the main drawbacks in the scale-up and industrial application of microalgal processes (Gupta et al. 2016). Microalgal biofilm systems could provide an alternative approach to resolve the microalgae-based biomass production challenges. Due to the growth of the cells attached to a surface, biomass can be separated from the reactor mixed liquor without using additional electrical energy or chemicals (Johnson and Wen 2010; Ozkan et al. 2012), therefore allowing an increase in the cell retention time within bioreactors and a higher metabolic activity (Tam et al. 1994).

Biofilm photobioreactors (PBR) can be grouped into three categories: permanently submerged systems, intermittently submerged systems and perfused systems (Berner et al. 2015). The latter uses a porous substrate that supplies nutrients and moisture to the microalgae which grow on the outside of the membrane, exposed to the surrounding gas phase (Shi et al. 2007). Several works have studied the formation of microalgae biofilms growing on different substrates (Table 1). Attached algal culture systems such as the Algal Turf Scrubber (ATS) have been used successfully for growing filamentous macroalgae for nutrient removal from animal wastewater (Wilkie and Mulbry 2002; Kebede-Westhead et al. 2006). Pioneering studies demonstrate the feasibility of microalgal biofilms for the removal of residual nitrogen and phosphorus from wastewater (Craggs et al. 1997; Schumacher and Sekoulov 2002; Shi et al. 2007; Babu 2011; Blanken et al. 2014). However, investigations on nutrient removal from industrial wastewater by attached microalgae are still required.

Table 1 Microalgae biofilm systems using different strains and carrier materials
Full size table

The nutrient assimilation and biomass growth in a PBR strongly depend on the photosynthetic activity (Han et al. 2017). An appropriate configuration of PBRs supports an effective photosynthesis, a higher biomass yield and an optimal removal efficiency of nitrogen and phosphorus (Prajapati et al. 2013). However, the types of PBRs that are used efficiently for algae axenic biomass cultivation do not show equal results in wastewater treatment (Han et al. 2017). Since the wastewater composition is usually highly variable, it is necessary to investigate the performance of PBRs treating different types of wastewaters and optimize the operational parameters (Muñoz et al. 2009). There are few studies concerning algal biofilm bioreactors for wastewater treatment, eventually coupled to biomass production (Katarzyna et al. 2015; Han et al. 2017).

The microalgal genera Scenedesmus and Chlorella (Chlorophyta) include a large number of species with different morphological and ultrastructural characteristics (Hegewald et al. 2013; Heeg and Wolf 2015), some of which have been reported as promising species for wastewater treatment and biomass production (Mata et al. 2010; Abdel-Raouf et al. 2012). Recent studies have demonstrated the feasibility of biofilm formation by S. obliquus (Liu et al. 2013; Schnurr et al. 2013; Zheng et al. 2017; Huang et al. 2018) and C. vulgaris (Gross et al. 2013; Sirmerova et al. 2013; Gao et al. 2015; Rajendran and Hu 2016; Barros et al. 2018) and their potential role for wastewater treatment. However, few works have studied the properties of S. vacuolatus for biotechnological applications, like phytoremediation of olive-mill wastewaters (La Russa 2009) or biodiesel production (Gargano et al. 2016). The aim of this work was, therefore, to evaluate the nitrogen and phosphorous removal performance of a S. vacuolatus and C. vulgaris biofilm grown in a PBR fed with synthetic wastewater. Two sets of experiments were performed: the first under batch conditions to compare the performance of suspended versus attached cells, and the second under semi-batch conditions to study S. vacuolatus and C. vulgaris biofilm development in the PBR.

Materials and methods

Microalgae strains and medium

The green algae strains Scenedesmus vacuolatus ACUF_053 and Chlorella vulgaris ACUF_809 were selected in Moreno Osorio et al. (2018), according to their growth rate and nutrient removal performance. Seed cultures were maintained in BBM medium. The cultures were grown in 250-mL Erlenmeyer flasks, which contained 100 mL of medium and were placed on a shaking plate at 100 rpm in an incubation room controlled at 25 (± 2) °C. Fluorescence light was used to maintain a constant light intensity at 95 µmol photons m−2 s−1 (Phillips TLA, 30W/55).

The culture medium for the assays in both sections was a synthetic wastewater (modified BBM), giving an initial molar N:P ratio of 10:1 that was prepared to simulate a phosphate-rich wastewater (Moreno Osorio et al. 2018) by mixing the following solutions in distilled water: (i) nutrients and salts; (ii) 1 M of alkaline EDTA solution; (iii) 1 mL of acidified iron solution; (iv) 1 mL of boron solution and (v) 3 mL of trace metals solution. The composition of each solution (mg L−1) was: (i) CaCl2 (25), NaCl (25), NaNO3 (250), MgSO4·7H2O (75), KH2PO4 (175), K2HPO4 (175); (ii) EDTA (5) and KOH (3); (iii) FeSO4·7H2O (4980) and H2SO4 (96%) (10 mL L−1); (iv) H3BO3 (114.2) and (v) MnCl2·2H2O (0.6), CoSO4·7H2O (0.5), CuSO4·7H2O (1.6), Na2MoO4·2H2O (0.6) and ZnSO4·7H2O (8.8). The growth medium was autoclaved at 121 °C for 20 min.

Cell growth and nutrient removal assays

Algal growth in the liquid phase was monitored by optical density (OD) in a spectrophotometer (Secomam s.250) at 680 nm. In order to determine dry cell weight (DCW) (mg L−1), a calibration curve was employed as described by Moheimani et al. (2013). For dry weight (DW) determination, culture samples at different optical densities were measured by filtering algal samples on pre-weighed glass fibre filter paper with a pore size of 0.45 μm. The filters were then dried at 105 °C in an oven for 12 h, then algal biomass DW was determined by the difference of the two weights. 2 mL samples from each flask were taken daily to determine phosphorus and nitrogen concentrations after filtration (0.45 μm).

Evaluation of PBR system performance under batch operation

The experimental design was divided in two sets of experiments. The performance of the nutrient removal and biofilm formation of the microalgae strains was followed under two cultivation modes as: set 1.1 compared attached and suspended growth and set 1.2 the nutrient removal efficiency in semi-batch mode of S. vacuolatus. The set 2.1 compared attached and suspended growth and the set 2.2 the algal biofilm formation and nutrient removal efficiency in semi-batch mode of C. vulgaris. The first set consisted of a batch period of 7 days where the attached growth in the PBR was tested and compared to the suspended growth. In the second set, the biofilm PBR was operated in semi-batch mode for 41 days. Different nutrient concentrations were tested in order to analyse the performance of the biofilm system (Table 2). Daily changes in biofilm morphology of both microalgae strains were recorded by digital photography.

Table 2 Initial nitrogen and phosphorus concentrations used in the experiment of this study on S. vacuolatus and C. vulgaris biofilm performance
Full size table

Suspended cell culture

For microalgal suspended cultures, the cells were grown in triplicate in 100 mL medium inside 250 mL Erlenmeyer flasks and incubated at 25 (± 3) °C with horizontal shaking at 100 rpm supplied with constant illumination (90 µmol m−2 s−1) by a cold-white directional LED light source (Beghelli cold white lamps, 12 W).

Attached cell culture in pilot PBR

The PBR design was a stationary biofilm reactor design (Fig. 1). The system is a vertical bioreactor made of plexiglass with the dimensions of 0.145 × 0.10 × 0.185 m (wide, long, height). In these reactors, the biofilm develops on the flat surface of the PBR (Fig. 1a) and a thin layer of liquid is flown over it, recycled by a pump (Gross et al. 2015). Cotton fabric was tested as supporting material for algae attachment. This carrier was selected because of the promising results obtained in previous biofilm attachment assays (Carbone et al. 2017). Moreover, as reported in previous studies, this material presents other advantages such as local availability, low cost, reusable and easy production (Johnson and Wen 2010).

Fig. 1
figure 1

Biofilm PBR design: a PBR front view, b PBR side view and c PBR photograph. (1) Cotton carrier (for attachment of microalgae), (2) synthetic wastewater, (3) carrier holder, (4) dropper, (5) reservoir of synthetic growth medium, (6) pump, (7) PBR side view and (8) LED lamp

Full size image

The PBR design facilitated the establishment and growth of the microalgal biofilm. The vertical position of the carrier allowed its homogeneous illumination and flow of the synthetic wastewater over the biofilm. The synthetic wastewater circulates in the system from the reservoir bottle through plastic pipes to the PBR where the flow was split in 4 dropper devices. The medium flows as a thin layer on the carrier and then falls to the bottom of the reactor where it is collected in a sink that directs the medium to the output valve. From the outlet, the medium runs through a plastic pipe back to the reservoir bottle (Fig. 1a).

The support material was cut into 0.10 × 0.10 m pieces and positioned on the carrier holder in the centre of the chamber (Fig. 1). The system was operated in semi-batch mode and at room temperature (25 ± 3 °C). The PBR was incubated with 100 mL of a microalgae suspension pre-cultivated in a shake flask with a cotton plug to a container (200 mL Pirex bottle covered with aluminium foil) with the synthetic wastewater. The algal cell suspension had an OD680 of 0.2 and 0.25 for S. vacuolatus and C. vulgaris, respectively, for a concentration of ≈ 3.5 × 107 cells mL−1. The medium was injected at a flow rate of 4.45 mL min−1 without air insufflation.

For the entire duration of the experiments, the light intensity was set at 5920 lux (90 µmol m−2 s−1). Only one side of the carrier was illuminated by a cold-white directional LED light source (Beghelli cold white lamps, 12 W) (Fig. 1b). Cool white LED lights have a wider-spectrum wavelength (380–760 nm) which encompasses the blue, red and green light spectrum (Yan et al. 2013). Since the PBR was transparent, aluminium foil was used to cover the bottom of the reactor (4 cm) and only the upper portion of the reactor was illuminated. This selective pressure allowed minimizing the suspended microalgae growth in the medium.

Evaluation of PBR system performance under semi-batch operation

The operation of the PBR was continued after the conclusion of the comparative periods as described above. The reactor was fed with new medium after the complete removal of one or both nutrients (phosphate and nitrate). Three increasing NO3−-N and PO43−-P concentrations were tested with S. vacuolatus ACUF_053 according to the consumption of these nutrients over time (Table 2). On other hand, only one concentration was tested for C. vulgaris ACUF_053 in the synthetic wastewater with NO3− as nitrogen source due to the different nutrient removal trends observed. Algal biomass was not harvested to avoid external contamination of the biofilm system.

Analytical methods

Medium samples from the synthetic wastewater reservoir were collected daily in triplicate for absorbance, pH, NO3− and PO43− analysis. The pH values were measured with a pH meter Bench model AD-1030 (Adwa Instruments Inc., Szeged, Hungary). Phosphorus was measured as orthophosphate (PO43−-P) and nitrogen as nitrate (NO3−-N). Analyses of nitrate and phosphate were performed by ion chromatography using a 761 compact IC analyzer (Metrohm, Herisau, Switzerland) as described by APHA (2005).

The nutrient removal efficiency Re (%) was calculated with the equation Re = (ΔS·S −1 i )*100, where Re represents the nutrient removal efficiency of the substrate (NO3−-N or PO43−-P), ΔS is the variation of nutrient concentration (mg L−1) and Si is the initial concentration of the nutrient. Mean and standard deviation were calculated. One way ANOVA for statistical examination of significant differences in the nutrient removal efficiency between and within suspended and attached cultures was carried out using OriginPro 2017 (OriginLab, Northampton, MA) with a confidence level of 95%. The null hypothesis stated that the removal efficiencies were equal.

Results

S. vacuolatus ACUF_053 performance

Evaluation of PBR system performance under batch operation

Suspended and attached S. vacuolatus ACUF_053 cells cultivated under the same conditions in synthetic wastewater as described above were able to grow in both culture conditions. The absorbance measurements showed that the biomass growth in the suspension was linear with a short lag phase and no stationary phase was reached (Fig. 2a). No biomass growth was observed in the medium of the PBR system, suggesting that cells grew exclusively on the carrier material (Fig. 2a).

Fig. 2
figure 2

a Biomass growth, b pH changes in liquid, c phosphorus and d nitrogen removal kinetics of S. vacuolatus ACUF_053 cultivated in suspension and as attached biofilm in PBR. The error bars correspond to the standard deviation

Full size image

During the first 3 days, pH values of the cell culture in suspension increased reaching an alkaline pH of 10 (Fig. 2b). On day seven, a drop of the pH of the medium was observed. In the PBR system, the pH of the medium dropped to 7.27 at day seven, when complete removal of nitrate and phosphate was observed in the medium. Then the reactor was fed with a new phosphorus solution with the double phosphorus concentration (6.5 mg PO43−-P L−1).

In the suspended culture system, complete PO43−-P removal was obtained after 4 days, while the complete depletion of nitrogen (13.5 mg NO3−-N L−1) required 6 days (Fig. 2c, d). Different behaviour was observed in the algal biofilm system, a complete phosphorus removal (4 mg PO43−-P L−1) was achieved from the first day of the experiment (Fig. 2c), although no significant difference was observed between the phosphorus removal efficiency of the suspended and attached systems (P > 0.05). Due to the fast and complete phosphate depletion, the nitrate removal ceased. Indeed, 28% of the nitrate concentration was consumed the first day in the PBR and no uptake was observed for the subsequent 2 days. At day three, the system was fed with phosphate only (10 mg PO43− L−1 or 3.2 mg PO43−-P L−1), after which the NO3−-N removal started again until its complete removal.

Nitrogen (NO3−-N) removal was constant and regular in the suspended cell cultures until its complete starvation on day 6. 28% of the nitrate concentration was consumed the first day in the PBR and no uptake was observed until day three, when the system was re-fed phosphorus (3.2 mg PO43−-P ·L−1). Complete NO3−-N removal in the suspended cell culture and PBR was obtained on the 6th day of the assay (Fig. 2d). No significant difference was observed between the nitrogen removal efficiency of the suspended and attached systems (P > 0.05). No biomass growth was observed in the synthetic wastewater in the PBR system, confirming that the growth was exclusively on the fabric carrier. After 7 days of the batch comparative assay, semi-batch conditions were applied to the PBR system.

Evaluation of PBR system performance under semi-batch operation

Scenedesmus vacuolatus ACUF_053 biofilm formation was tested under semi-batch conditions for 41 days in a new PBR. The PBR was fed with new synthetic wastewater when NO3−-N depletion was achieved. Figure 3 shows the daily changes in the biofilm morphology.

Fig. 3
figure 3

Biofilm development phases observed in the biofilm growth of S. vacuolatus ACUF 053 in the PBR

Full size image

Five phases were identified in the development of the biofilm from the morphological changes observed: (1) attachment of the microalgae cells inoculated in the medium onto the carrier; (2) biofilm formation, the attached cells grew on the carrier and cover its full area; (3) maturation I, microalgae biomass on the carrier became ticker; (4) adaptation; microalgae biomass partially detached from the carrier and (5) maturation II, microalgae biomass on the carrier increased in thickness.

After the initial biofilm maturation phase, the biofilm showed a detachment period that was monitored, reason why the biofilm biomass was not harvested. After this maturation I and detachment phase, the biofilm could re-grow in 10 days from the biomass that remained on the carrier surface (adaptation phase); comparable to the time required for the biofilm initial attachment phase (Fig. 3).

The NO3−-N contained in the medium was efficiently consumed by the microalgal biofilm, as shown by its concentration in the synthetic wastewater (Fig. 4a). It should be noted that the five periods show some differences in nitrate removal rates. As expected, the lower rate occurs during the biofilm formation in phase 1. During phases 2 and 3, it is possible to see similar removal rates despite the change in nitrogen concentration (from 27 to 54 mg NO3−-N L−1) in the PBR influent. During phase 4 the rate of the nitrate removal decreased, but a higher NO3−-N removal rate was detected in the last period.

Fig. 4
figure 4

S. vacuolatus ACUF_053 a nitrogen (NO3−-N) and phosphorus (PO43−-P) removal kinetics in the PBR under semi-batch conditions; b pH changes in the medium of the biofilm photobioreactor and, c microalgal biomass growth in the medium of the biofilm PBR. The numbers define the development phases: (1) attachment, (2) biofilm formation, (3) maturation I, (4) adaptation and (5) maturation II

Full size image

By analysing the PO43−-P concentration in the system influent (Fig. 4a), the phosphorus removal rate was high during the attachment phase and the second biofilm maturation phase, with removal efficiencies of 72.9% and 57.8%, respectively, after 1 day of refreshing the medium. However, in the second maturation phase the system did not reach a complete phosphorus removal before nitrogen limitation occurred in the medium. After the attachment phase, the phosphorous demand decreased inversely to the nitrogen demand, as shown by the faster NO3−-N consumption. For this reason, the system was fed with new synthetic wastewater when complete NO3−-N removal was recorded, although 2.5–3 mg PO43−-P L was still present in the medium.

The synthetic wastewater in the PBR showed low variations in the pH (7.3–8.1) from the beginning until day ten. Higher pH values (9.7) were recorded between days 13 and 21 of the experiment, during the maturation I period, after the increase of NO3−-N and PO43−-P doses from 13.5 mg NO3−-N L−1 and 6.5 mg PO43−-P L−1 to 27 mg NO3−-N L−1 and 13 mg PO43−-P L−1, respectively. Consequently, the concentrations were further increased to 54 mg L−1 of NO3−-N and 13 mg L−1 of NO3−-N after day 21, but the system reached a pH stability similar to that observed during the first 10 days. The maximum pH value during the adaptation and maturation II phases was 8.5 (Fig. 4b).

The algal biomass concentration in the reactor medium was measured daily (Fig. 4c). During the initial 6 days, no or a very low biomass concentration (0.012 mg L−1, day 6) was detected. After day 7, the beginning of an exponential growth was observed when the PBR was fed for the second time with synthetic wastewater with 13.5 mg NO3−-N− L−1 and 6.5 mg PO43−-P L−1. At day ten, a biomass concentration of 0.42 mg L−1 was measured in the medium and the reactor was fed with 27 mg NO3−-N L−1 and 13 mg PO43−-P L−1. Between day 10 and 13, a linear increase of the biomass in the medium was observed and 0.26 mg L−1 was achieved. On day 17 of the PBR run, the reactor was fed with 54 mg NO3−-N L−1 and 13 mg PO43−-P L−1 and a new exponential biomass growth phase was observed in the recirculation medium. This rise was recorded until the 21th day, when the biomass concentration was 0.17 mg L−1 and the PBR was fed with the same nutrient concentrations (end of formation phase).

Higher biomass concentrations in the medium (0.53 mg L−1) were observed on day 24 of the experiment, 3 days after the exponential biomass increase (Fig. 4c). This higher value is probably due to biofilm detachment. From day 25, an adaptation phase was observed and the biomass concentration in the medium decreased although the values recorded in this phase (0.28 mg L−1) were still higher than in the attachment phase, probably due to the detachment of death cells from the biofilm. The second maturation phase began on day 35 when the start of a gradual increase in the biofilm biomass concentration was observed (Fig. 4c).

C. vulgaris ACUF_809 performance

Evaluation of PBR system performance under batch operation

Chlorella vulgaris ACUF_809 was able to grow in both culture conditions. No lag phase was observed during biomass growth in the suspension assay and a decrease was evident from day 3 to 6 (Fig. 5a). On the other hand, a low biomass concentration was observed in the medium of the PBR system, suggesting that—like for S. vacuolatus ACUF_053—the cells grew mainly on the carrier material (Fig. 5a).

Fig. 5
figure 5

a Biomass growth, b pH changes in liquid medium, c phosphorus and d nitrogen removal kinetics of C. vulgaris ACUF_809 cultivated in suspension and as attached biofilm in the PBR. The error bars correspond to the standard deviation

Full size image

A steep increase of the pH values was observed in both suspended and attached cultivations during the first day, reaching an alkaline pH of 9.4 and 9.9, respectively (Fig. 5b). A slightly increased pH in the suspended cultivation was observed until day 3 and kept constant till the end of the experiment. The pH of the PBR medium dropped to 6.8 at day two, when the reactor was fed with a new phosphate dose (6.5 mg PO43−-P L−1). In the suspended culture system, complete PO43−-P and NO3−-N removal was obtained after 3 days (Fig. 5c, d). However, in the algal biofilm system almost complete phosphorus removal (6.5 mg PO43−-P L−1) was achieved at day two (Fig. 5c). A significant difference was observed between the phosphorus removal efficiency of the suspended and attached systems (P < 0.05). The nitrate removal rate decreased between day one and two, it increased again after day two. When almost complete NO3−-N and PO43−-P removal was observed in the medium, then the reactor was fed with a new phosphorus solution with equal initial dose (6.5 mg PO43−-P L−1).

A lower NO3−-N removal efficiency was observed during the first day of cultivation in the suspended cell cultures. However, a steep decrease was recorded from day one to three, until its complete depletion. A significant difference was observed between the nitrogen removal efficiency of the suspended and attached systems (P < 0.05). After 6 days of batch comparative assay, semi-batch conditions were applied to the PBR system.

Evaluation of PBR system performance under semi-batch operation

Chlorella vulgaris ACUF_809 biofilm formation was followed under semi-batch conditions for 41 days in the same PBR configuration system. The PBR was fed with synthetic wastewater with a higher concentration of nitrogen (13.5 mg NO3−-N L−1) and phosphorus (6.5 mg PO43−-P L−1), and reloaded with new synthetic wastewater whenever the complete removal of both or one of the nutrients was achieved. Figure 6 shows the daily changes in the biofilm morphology. A gradual but constant development of the biofilm was observed along the 41 days of the experiment. No abrupt morphological changes were observed in the biofilm formation and no association with the nutrient consumption could be inferred.

Fig. 6
figure 6

Photographic record of C. vulgaris ACUF 809 biofilm development on cotton fabric in the PBR

Full size image

The NO3−-N and PO43−-P contained in the medium were efficiently consumed by the microalgae biofilm, as shown by the concentration in the synthetic wastewater (Fig. 7a). Interestingly, removal of phosphorus and nitrogen was mainly constant during the cultivation time. A low removal efficiency of both nutrients was observed from day 6 to 10, followed by a higher nutrient removal efficiency after 1 day of feeding with new medium, i.e. 62–73% of NO3−-N during the period from day 10 to 21 and 80.5–85.7% of PO43−-P from day 14 to day 18. The synthetic wastewater in the PBR showed a fast increase of the pH during the first day after each medium renewal, i.e. 7 (± 0.4) to 10.2 (± 0.2) along the experiment. Higher pH values (10.6) were recorded at days 17 of the experiment (Fig. 7b).

Fig. 7
figure 7

C. vulgaris ACUF_809, a nitrogen (NO3−-N) and phosphorus (PO43−-P) removal kinetics in the PBR under semi-batch conditions; b pH changes in the medium of the biofilm photobioreactor and c microalgal biomass growth in the medium of the biofilm PBR

Full size image

The algal biomass concentration in the reactor medium was measured daily (Fig. 7c). A low, but constant biomass concentration (8.2–57 mg L−1) was recorded during the 41 days of the experiment. Five peaks of biomass concentration were measured on days 8 (52.8 mg L−1), 9 (50.1 mg L−1), 11 (46.8 mg L−1), 27 (46.2 mg L−1) and 38 (57.6 mg L−1). On the contrary, two periods of low biomass concentration were observed from day 13 to 25 (12.9–23 mg L−1) and from day 30 to 35 (9.3–30 mg L−1) (Fig. 7c).

The performance of biofilms of both strains in the PBR showed significant differences (P < 0.05) in pH: the pH of the synthetic wastewater in PBR with S. vacuolatus displayed a less variation between pH 7 and 8. On the other hand, the pH in synthetic growth medium in PBR with C. vulgaris increased up to 10 and kept constant at this pH value. Biomass growth between the liquid phase of the PBR system with S. vacuolatus ACUF_053 and C. vulgaris ACUF_809 were significantly different (P < 0.05). Higher biomass concentrations were recorded in the synthetic wastewater in the PBR with S. vacuolatus biofilm compared to the medium in the PBR with C. vulgaris biofilm.

Discussion

Suspended vs. attached microalgae growth

This study showed that the microalgae strains S. vacuolatus ACUF_053 and C. vulgaris ACUF_809 grew both under suspension and attached conditions in the new PBR configuration. In other studies, the surface colonization by microalgae has been reported within 24 h in horizontal systems (Irving and Allen 2011). In this study, the algae colonization lasted 7 (± 1) days. Long term (≥ 3 months) stability has been demonstrated in continuous studies despite the dynamic nature of algal biofilms (Boelee et al. 2011). During the 7 days of the comparative experiments, the algae strains showed adhesion onto cotton in the PBR (Figs. 4a, 7a). This period could present an adaptation phase of the algae cells to the medium flow and the surface features of the carrier material. Suspended algal systems are designed to minimize attachment or settlement of cells onto bioreactor surfaces. In contrast, attached algal systems, as used in this study, promote cell adhesion to a surface favoured by a polymeric matrix on the cotton fabric material (Kesaano and Sims 2014). Cotton fabric not only supports excellent attachment for algal growth, but it’s also durable and low cost (Johnson and Wen 2010; Gross et al. 2013).

The biomass concentration obtained in the suspended culture was 0.37 g L−1 and 0.2 g L−1, respectively, for S. vacuolatus ACUF_053 and C. vulgaris ACUF_809; which falls in the range of 0.1–1 g L−1 as reported by Jonson and Wen (2010). Similar biomass concentrations were reported for Chlorella sp., i.e. 0.1 g L−1 (Nayak et al. 2013) and for C. vulgaris, i.e. 0.39 g L−1 (Matos et al. 2015) grown in BBM. However, this concentration is low compared with the one obtained for S. vacuolatus ACUF_053 (2.9 g L−1) by Gargano et al. (2016) and for Scenedesmus sp. (0.78–0.87 g L−1) by Nayak et al. (2013).

Nutrient removal trends in the suspended cell experiments showed a singular biomass growth pattern for S. vacuolatus ACUF_053: low uptake percentages were observed at the beginning of the growth phase (Fig. 2), increased uptake occurred in parallel with the growth peaks and the stationary phase was characterized by a complete nutrient removal from the medium (Boelee et al. 2011; Kesaano and Sims 2014). Low nutrient removal efficiencies at the beginning of the growth phase of microalgae biofilms have been reported (de Godos et al. 2009). This behaviour was attributed to the insufficient establishment and acclimation of the algal biofilm community. In contrast, a high nutrient removal percentage, especially phosphorus, was observed during the establishment of the S. vacuolatus ACUF_053 biofilm (Fig. 4a). C. vulgaris ACUF_809 showed similar removal efficiencies in its suspended and attached growth (Fig. 5b).

Microalgae can store phosphorus within the biomass as polyphosphates that can be actively involved in the metabolism of the cell or being stored for when the external phosphate concentration becomes limiting (Powell et al. 2009). However, the difference in the PO43−-P removal efficiency between suspended and attached cell conditions could be related to polymeric substances (EPS) formation. EPS are mainly composed of polysaccharides, proteins, extracellular DNA (eDNA) and lipids (Singha 2012). Previous studies reported a high EPS to biomass ratio in young algal biofilms. Hence, the authors suggested that this probably occurs due to EPS production as a survival mechanism (Barranguet et al. 2005), although EPS on the algal biofilms was not measured in this study to avoid contamination and destruction of the biofilm during the analysis. A maximum EPS production during the early stationary growth phase (about 8 days) was reported by Wolfstein and Stal (2002), which fits the period of the highest phosphorous demand by S. vacuolatus ACUF_053 (Fig. 2). A possible contribution of EPS to the initial process of cell attachment in young biofilms has been suggested (Romaní et al. 2008). On the other hand, microalgal EPS production and composition can vary depending on different factors, i.e. the strain, stress conditions, nutrient concentrations and community diversity (El-Sheekh et al. 2012; Matsumoto et al. 2014; Xiao and Zheng 2016).

Development of S. vacuolatus ACUF_053 and C. vulgaris ACUF_809 biofilms in PBR

Biofilm-based photobioreactors are typically designed with high surface area to volume ratios for an efficient light penetration. However, immobilized algal cultures on the reactor surfaces are vulnerable to photodamage from constant exposure to high irradiation (Muñoz et al. 2009). An irradiation range between 12 and 88 µmol m−2 s−1 is a limited condition for algal biofilms (Hill and Fanta 2008). In this study, the irradiation used was over this range (90 µmol m−2 s−1) and it was constant during the experiment. The biofilm PBR configuration tested has several advantages: (1) a lower footprint than horizontal systems, (2) medium dropping ensures a simple but constant contact of biofilm with the growth medium, (3) the thin medium layer and slow flow rate to which the algal biofilm is exposed allows a reduction in the required volume of the medium (Boelee et al. 2011; Katarzyna et al. 2015) and (4) biofilm detachment/regrowth phases allow serial growth/harvesting cycles.

Biofilm accumulation is determined by the balance of attachment, growth and detachment processes (Hunt et al. 2004; Kesaano and Sims 2014; D’Acunto et al. 2015; Mattei et al. 2018). Five phases were defined according to the observed colonization, nutrient removal and the green intensity of the cotton fabric carrier by S. vacuolatus ACUF_053 (Fig. 3). On the contrary, the marked distinction in time of biofilm morphology and nutrient removal by C. vulgaris ACUF_809 (Figs. 6, 7a) was not observed with the S. vacuolatus ACUF_053 biofilms. This could be due to differences in the EPS production: with a low EPS production the microalgae cells take longer to establish cell to cell interactions, which increase the biomass content (thickness) of the biofilm (Ozkan and Berberoglu 2013; Matsumoto et al. 2014). This suggests a longer biofilm maturation period that was not totally comprised during the experimental time. This could be the cause for the constant microalgae biomass concentration measured in the synthetic wastewater circulating in the PBR (Fig. 7c).

Scenedesmus vacuolatus ACUF_053 biofilm exhibited a second maturation phase at day 35, the following days a gradual decrease of the biomass concentration in the synthetic wastewater was observed (Fig. 4c) accompanied by the re-greenering of the carrier. In general, the attachment of algal cells to the free surface of the materials is crucial and can be time consuming. Once the initial colonization occurs, attachment of additional algal cells to the existing algal biofilm layer is relatively easier (Ozkan and Berberoglu 2013; D’Acunto et al. 2015).

The steep increase in pH values in the synthetic wastewater observed during the C. vulgaris ACUF_809 biofilm formation could represent an impact on the integrity of the biofilm, which resulted in the detachment of biomass from the biofilm (Qureshi et al. 2005). Freshwater chlorophytes are capable of tolerating a wide range of pH, due to an acquired tolerance to both acidic and alkaline pH as an adaptive response to widely fluctuating pH levels that occur commonly in many productive freshwater environments (Goldman et al. 1982). It was reported that Scenedesmus obliquus (Turp.) Kutz. and Chlorella vulgaris Beij. can grow up to pH 10.6 although C. vulgaris is more adversely affected by alkaline pH than was S. obliquus (Goldman et al. 1982).

The difference observed in the pH behaviour for S. vacuolatus and C. vulgaris in the PBR biofilm formation can be due to the ability of some blue-green algae as well as green algae strains to alter the pH of the culture media (Wildman et al. 1974). As it was observed, the pH values of the synthetic wastewater went up to about 10 and 10.5, respectively, in S. vacuolatus ACUF_053 and C. vulgaris ACUF_809 suspended cultivations (Figs. 1b, 5b), respectively. However, the pH in the S. vacuolatus ACUF_053 biofilm cultivation was about 8 (Fig. 4b). This might be explained by a higher production of EPS, which can protect the cell from pH damage (Bethmann and Schönknecht 2009). It is known that pH plays an important role in both microalgal growth and flocculation. Most algae species achieve higher biomass productivity as the pH rise from 6 to 9 (Molina Grima et al. 2003). However, since the cell membrane and EPS of microalgae are mainly composed of polysaccharides, proteins and lipids, the components of extracellular metabolites can vary with different pH (Shen et al. 2014). Alternatively, the algae can excrete acids to neutralize the basic conditions to protect their microenvironment (Bethmann and Schönknecht 2009). According to Zhang et al. (2011), when the pH is above 9, microalgal cells may release EPS to protect themselves from the environment.

The increase in pH values over 8 was recorded during the maturation phase of S. vacuolatus ACUF_053 biofilms (Fig. 4b), when the higher photosynthetic activity occurred due to the high biomass concentration growing on the cotton. Algal photosynthetic activity raises the pH of the medium (Wei et al. 2008). Therefore, the results suggest a different EPS production between S. vacuolatus ACUF_053 and C. vulgaris ACUF_809; this should be demonstrated quantitatively to propose the application of biofilm technologies of these strains for different biotechnological processes, e.g. treatment of different types of wastewater.

During the 41 days of the experiment, different nutrient uptake rates were displayed by S. vacuolatus ACUF_053, which correlate with the biofilm development phases (Fig. 3). C. vulgaris ACUF_809 showed high maximum removal efficiencies, i.e. nitrate (89%) and phosphate (91%). Previous studies have shown that microalgal biofilm systems can achieve good nitrogen and phosphorus removal efficiencies from wastewater. Removal capacities of over 90% were measured for nitrate (NO3−) and over 80% for phosphate (PO43−) (Craggs et al. 1997; Shi et al. 2007; de Godos et al. 2009).

Conclusions

This study showed biofilms of S. vacuolatus ACUF_053 and C. vulgaris ACUF_809 have a high capacity for nutrient removal. Attached growth of S. vacuolatus had a higher phosphate removal efficiency than the suspended cultures. Complete removal of nitrate (60–240 mg L−1) and phosphate (20–40 mg L−1) was recorded during the S. vacuolatus biofilm formation and maturation phases, and C. vulgaris ACUF_809 achieved a nutrient removal efficiency of about 90% during the semi-batch process in the PBR. It was observed that the depletion of one of the nutrients limited the removal of the other.

The biofilm PBR design using S. vacuolatus and C. vulgaris showed potential for wastewater treatment due its higher removal rates. S. vacuolatus biofilms showed two advantages for the potential application in wastewater treatment: first, the possibility of biomass harvesting at the end of the maturation phases, and second the higher nutrient removal efficiency compared with suspended cell cultures under batch conditions. The phases of the S. vacuolatus biofilm development showed differences in nutrient removal efficiency. The biofilm showed a higher demand of phosphate during the initial adhesion and formation phases, while it has a higher nitrate demand during the subsequent maturation phase.