Isolation and evaluation of brackish diatoms for the photobiological treatment of reverse osmosis concentrate

Reverse osmosis (RO)-based brackish water desalination and potable water reuse are the key solutions for solving the global challenge of water scarcity due to population growth and prolonged and more frequent droughts (Fritzmann et al. 2007; Greenlee et al. 2009; Pérez-González et al. 2012). However, 10–25% of the feed water becomes a waste stream, called RO concentrate or brine, presenting a disposal issue, especially for the inland community. Additional water recovery from the RO concentrate is hindered by membrane scaling due to silica, phosphate, and calcium, and other hardness metals, such as strontium and barium. Currently, advanced water purification facilities (AWPFs) for indirect and direct water reuse and brackish water desalination facilities (BWDFs) utilize ocean/surface water discharge, sewer discharge, land application, deep well injection, and evaporation pond to dispose of the RO concentrate often without any treatment (Xu et al. 2013; Hobbs et al. 2016; Adams et al. 2022). Several technologies have been proposed to improve the freshwater recovery and reduce the volume of RO concentrate, including the removal of scaling constituents by ion exchange or chemical softening (Venkatesan & Wankat 2011), variations of RO processes with vibration and precise control of concentrate discharge, including closed-circuit RO (Efraty et al. 2011; Gu et al. 2021), and the combination of non-RO desalination processes such as nanofiltration (NF) (Park et al. 2017), forward osmosis (Cath et al. 2006; Hancock et al. 2013), membrane distillation (Kim et al. 2016), and electrodialysis reversal (Kawahara 1994). However, these technologies tend to be highly chemical and/or energy-intensive and produce secondary waste streams such as spent resins. Furthermore, they often require the precise process control of multiple pumps and valves, which vastly complicates the operation of the process and prohibits their integration into full-scale RO facilities for concentrate treatment. There has been an urgent need for more cost-effective, environmentally friendly, and sustainable scalant removal technologies for AWPFs and BWDFs to enable higher water recovery using RO while reducing the volume of concentrate and its environmental impact.

Recently, the use of brackish diatoms for the treatment of RO concentrate from AWPFs and BWDFs has been proposed (Ikehata et al. 2017, 2018b, 2018c). In these studies, brackish diatoms Gedaniella flavovirens (formerly Pseudostaurosira trainorii (Li et al. 2018)) PEWL001 and Nitzschia communis PEWL002 isolated from agricultural drainage water demonstrated the rapid removal of aqueous silica and nutrients including orthophosphate, nitrate, and ammonia from the brackish agricultural drainage water (Ikehata et al. 2017) and RO concentrate from AWPFs and BWDFs (Ikehata et al. 2018c; Ikehata et al. 2019; Kulkarni et al. 2019). More than 95% removal of aqueous silica, phosphate, and ammonia, as well as >60% removal of calcium carbonate, were possible with this photobiological treatment with G. flavovirens PEWL001 and N. communis PEWL002 within 3–4 days (Ikehata et al. 2018c). After the photobiological treatment, additional water recovery from the treated RO concentrate with a secondary RO becomes feasible (Ikehata et al. 2018b). Concentrate samples from four AWPFs and seven BWDFs in the Southwestern United States have been successfully treated by this photobiological process using G. flavovirens PELW001 (Ikehata et al. 2018a, 2019). The use of this photobiological technology in the RO concentrate treatment appears to have potential advantages over conventional physical–chemical processes including the utilization of solar energy, efficient nutrient removal, inorganic carbon and certain divalent and trivalent metal removal, and production of useful algal biomass. However, only two brackish diatom isolates (G. flavovirens and N. communis) have been tested so far for this treatment process (Ikehata et al. 2018c), and the utility of other species is still unknown.

Therefore, in this study, 11 strains of brackish diatoms representing five different genera, including Gedaniella, Nitzschia, Anomoeoneis, Halamphora, and Encyonema, that were culturable in the RO concentrate sample were tested in this study to compare their silica and nutrient uptake, as well as the removal of other dissolved constituents from a silica- and nutrient-rich RO concentrate from a full-scale AWPF. Among these 11 strains, five strains were newly isolated from evaporation ponds in this study, while four strains were previously isolated, including G. flavovirens PEWL001 and N. communis PEWL002 (Ikehata et al. 2017, 2018c), and two were obtained from a brackish lake (Sato et al. 2011). Nine strains were tested for the first time.

Brackish diatom strains used in this study were isolated from several surface water sources as shown in Table 1. As described earlier in Ikehata et al. (2017, 2018c), the first four strains namely G. flavovirens PEWL001, N. communis PEWL002, Anomoeoneis sphaerophora cf. f. costata PEWL003, and Halamphora sydowii PEWL004 were isolated from agricultural drainage water samples collected from an evaporation pond complex in Kings County, CA in the summer of 2010 (PEWL001) and in November 2015 (PEWL002, 003, and 004) by incubating unfiltered water samples with or without the sediment in a clear plastic container at room temperature (∼25 °C) under continuous illumination with a 13-W compact fluorescent light or a 9-W light-emitting diode (LED) bulb, followed by serial dilution, agar pour plate, and micropipette techniques (Andersen & Kawachi 2005). G. flavovirens PEWL001 was further purified by single-colony isolation by a micropipette. Nitzschia sp. PEWL008 was isolated from the sediment and water sample was collected from a hypersaline evaporation pond for RO concentrate at a drinking water treatment plant in Dateland, AZ in a similar manner to PEWL001-004. Two strains of G. flavovirens, Psetr3 and Psetr7, were isolated from bottom sands at Obuchi-numa Lake in Aomori Prefecture, Japan (Sato et al. 2011).

Halamphora sp. PEWL011, Encyonema sp. PEWL012, Anomoeoneis sp. PEWL013, and Anomoeoneis sp. PEWL014 were isolated directly from the sediment and sand samples were collected from the agricultural drainage water evaporation pond complex in Kings County, CA in September 2017. A total of 46 single diatom cells or colonies were picked up by the micropipette technique and transferred to a 0.2-μm filter-sterilized, silica- and nutrient-rich brackish water medium (see Brackish water medium and RO concentrate sample) in non-treated 12-well cell culture plates. The culture plates were incubated at ∼27 °C under continuous illumination with four 9-W LED bulbs. After about 1 week of incubation, the active growth of nine isolates representing four different diatom strains was confirmed. Isolates with brown-colored colonies were transferred to larger culture tubes (15- and 50-mL VWR SuperClear polypropylene centrifuge tubes, VWR International, Radnor, PA) and maintained until use.

Once the active growth of greenish brown/brown diatom biomass was confirmed by visual observation with naked eyes, as well as with an inverted microscope, each isolate was transferred to and maintained in 50-mL culture tubes containing a 0.2-μm filter-sterilized RO concentrate from an advanced water purification facility (see Brackish water medium and RO concentrate sample). The filtered RO concentrate was replaced every 3–4 days to obtain enough biomass for the photobiological treatment experiment. In addition, subcultures were maintained by transferring an aliquot of biomass into new culture tubes containing the brackish water medium every 3–4 weeks.

A Hach DR-2800 spectrophotometer (Loveland, CO) was used for the colorimetric analysis of dissolved constituents. A Hach ISENa38101 sodium ion selective electrode combined with an HQ40d portable meter was used for sodium analysis. Appropriate Hach methods were used for general chemical parameters including molybdate-reactive silica, orthophosphate, ammonia-N, nitrate-N, chloride, sulfate, hardness, alkalinity, iron, manganese, color, and chemical oxygen demand (COD), as described earlier (Ikehata et al. 2017, 2018c). In this study, molybdate-reactive silica (hereafter referred to as ‘reactive silica’) was used as a measure of dissolved silica in the aqueous samples.

A silica- and nutrient-rich brackish water medium was prepared from a groundwater RO concentrate sample obtained from a brackish groundwater desalination facility in Alamogordo, NM (collected in January–March 2017) spiked with sodium phosphate monobasic and sodium nitrate (ACS reagents; Sigma-Aldrich, St. Louis, MO) or an f/2 medium concentrate (f/2 Algae Food Part B; Fritz Aquatics, Mesquite, TX). Another RO concentrate sample was collected from the third stage of a full-scale RO unit at the Groundwater Replenishment System Advanced Water Purification Facility of the Orange County Water District in Fountain Valley, CA on September 26, 2017, and used in the photobiological treatment experiments. The RO concentrate sample was analyzed for water quality parameters upon arrival and refrigerated until use. The water quality of the brackish water medium and the RO concentrate sample are shown in Table 2.

Two bench-scale photobiological treatment experiments, namely the preliminary screening and time-series experiments, were carried out to evaluate the suitability of diatom strains for the treatment of RO concentrate. In both experiments, the photobiological treatment was conducted in VWR SuperClear 50-mL centrifuge tubes with screw caps as described previously (Ikehata et al. 2018c). The RO concentrate was filter-sterilized by 0.2-μm syringe filters and statically incubated at 27 ± 2 °C under continuous illumination with four 9-W LED bulbs (light temperature 5,000 K, 800 lm each).

In the preliminary screening experiment, 11 different diatom strains were evaluated for their silica uptake in two cycles of 4-day incubation. After the first 4 days of incubation, the supernatant was removed by decantation and the new RO concentrate was added into the tubes containing biomass to initiate the second cycle. Reactive silica concentrations were measured at the end of each cycle. After the second 4 days of incubation, the biomass was harvested and used in the subsequent time-series experiment. The biomass concentration was not quantified in the preliminary screening experiment. The preliminary experiment was conducted without replication.

In the time-series experiment, selected diatom species were tested for the reactive silica and nutrient uptake in a single-batch experiment in duplicate. The initial biomass concentrations ranged from 0.3 to 0.5 g L⁻¹. An aliquot of sample was collected every day to follow the silica uptake until at least 90% reactive silica removal was achieved, then the supernatant was separated from biomass and analyzed for calcium, bicarbonate, iron, and manganese. The biomass was washed with ultrapure water at least three times and then dried to determine the final dry weight.

Halamphora sp. PEWL011, Encyonema sp. PEWL012, Anomoeoneis sp. PEWL013, and Anomoeoneis sp. PEWL014 were isolated directly from sediment samples collected from agricultural drainage water evaporation ponds. As described above, these four strains were among 46 single-cell isolates that were culturable in both brackish water medium and the RO concentrate sample. The salinity of their habitat ranged from relatively low (TDS = 3 g L⁻¹, PEWL013 and 014) and moderate (TDS = 6 g L⁻¹, PEWL011) to high (TDS = 20 g L⁻¹, PEWL012). All the strains formed a thin brown or greenish brown biofilm like Nitzschia sp. PEWL008.

In this study, the suitability of nine brackish diatom strains collected from evaporation ponds in the Southwestern United States along with two strains from a brackish lake in Japan was evaluated for the photobiological treatment of silica- and nutrient-rich RO concentrate from an advanced water purification facility. In addition to the two strains previously described, namely G. flavovirens PEWL001 and N. communis PEWL002, all the brackish diatoms that represent five different genera, including Gedaniella, Nitzschia, Halamphora, Anomoeoneis, and Encyonema, were able to remove reactive silica from the brackish wastewater (Figure 2).

Among them, three strains of G. flavovirens showed an excellent ability to remove reactive silica (Figure 2), as well as the removal of nutrients, calcium, bicarbonate, iron, and manganese (Figures 3,⁴–5), from the RO concentrate. These strains of G. flavovirens originated from two entirely different geographical areas with different types of climates and ecosystems. As described earlier (Ikehata et al. 2017), G. flavovirens PEWL001 was isolated from an agricultural drainage water evaporation pond, which is a typical man-made brackish waterbody in the semi-arid Central Valley of California with a hot and dry Mediterranean climate. The drainage water is generated by periodical flushing of farmlands, whose soil contains high levels of dissolved solids due to repeated irrigation and evapotranspiration (Woltemade 2000; Skaggs et al. 2009). The fresh drainage water typically contains a high concentration of reactive silica (typically 30–40 mg L⁻¹) along with other dissolved minerals, nutrients, and organic matter (Woltemade 2000; Ikehata et al. 2017). The availability of silica and nutrients in the drainage water from the evaporation ponds most likely makes it an ideal habitat for brackish diatoms.

On the other hand, two G. flavovirens strains, Psetr3 and Psetr7, from Japan were isolated from a natural brackish lake, Obuchi-numa, in the northernmost part of Japan's main island (Sato et al. 2011). The area is classified as an oceanic climate with cooler summers and year-round precipitation. This lake is connected to both freshwater river and the Pacific Ocean, which creates an ecosystem with a wide variation of salinity ranging from 5 to 30 g L⁻¹ depending on the locations, time of the day, and seasons (Ueda et al. 2006). According to Ueda et al. (2006), the concentrations of ammonia-N, nitrate-N, and phosphorus were relatively low (<3.6, <11, and 0.17 μmol L⁻¹, respectively, which correspond to less than 0.2 mg L⁻¹ of nitrogen and less than 0.16 mg L⁻¹ of orthophosphate) in the shallow areas (<2 m deep) of Obuchi-numa. The very similar behaviors of these three G. flavovirens strains from two distant geographic locations observed in the photobiological treatment of RO concentrate are very interesting. This may warrant additional studies on this or other brackish Gedaniella species.

After G. flavovirens, two Halamphora species (PEWL004 and PEWL011) and two Nitzschia species (PEWL002 and PEWL008) were good silica removers (Figures 2 and 3). However, two Anomoeoneis species (PEWL013 and PEWL014) showed less promising silica removal ability. The poor growth of Anomoeoneis sp. might be due to the relatively low salinity of their original habitat (3 g L⁻¹), which was the lowest among the sources of brackish diatoms evaluated in this study (Table 1). Although those strains obtained from the environment with salinity similar or higher to the RO concentrate to be treated (5.1 g L⁻¹ as TDS) performed well in the photobiological treatment, those from lower salinity might be more sensitive to the changes in osmotic pressure. Encyonema sp. PEWL012 also showed a relatively poor ability to remove reactive silica (Figure 3), although the nutrients, calcium, bicarbonate, iron, and manganese removal were comparable to the other brackish diatom strains tested (Figures 4 and 5). The reason for the lower silica removal/uptake efficiency of this strain is unclear.

The impact of salinity on the photobiological treatment is still inconsistent and requires further investigation. In a separate study, while G. flavovirens PEWL001 was able to treat RO concentrate samples with a wide range of salinity from 3.6 to 18 g L⁻¹ (Ikehata et al. 2018a, 2019), two marine strains including Amphora sp. CCMP 129 and Melosira nummuloides CCMP 481 were unable to grow in brackish wastewater with TDS of ∼6 g L⁻¹ in our previous attempt (Ikehata et al. unpublished data). Clavero et al. (2000) reported a wide range of salinity tolerance (0.5–17.5% wt/vol) of several Amphora, Nitzschia, and Entomoneis species isolated from thalassic hypersaline environments including tidal channels and evaporation ponds. The salinity tolerance and insensitivity of the diatom strains such as Halamphora and Nitzschia species observed in the current study are consistent with the reported observation. The euryhalinity of certain brackish diatoms was probably due to the adaptation to the environments of changing salinity (Carpelan 1978; Clavero et al. 2000). Additional investigation is suggested to explore the wider varieties of euryhaline water diatoms and to determine any relationships between the silica removal efficiency and biological characteristics of diatoms.

The performance of brackish diatoms in the photobiological treatment may vary depending on the water quality (e.g., salinity, pH, and nutrient availability) and other factors (e.g., temperature and light conditions). Therefore, it is probably advisable to test several promising diatom strains to find the best one for the RO concentrate to be treated in a bench or pilot-scale testing based on the specific treatment goals (e.g., silica, calcium carbonate, and/or nutrient removal) of the project. The use of a consortium of brackish diatoms (Ikehata et al. 2017) may also be considered.

In this study, 11 different strains of brackish diatoms that represent five different genera were evaluated for their ability to remove reactive silica, nutrients, calcium, bicarbonate, iron, and manganese from a silica- and nutrient-rich RO concentrate obtained from a full-scale advanced water purification facility in Southern California. Although all the brackish diatoms evaluated here were able to remove reactive silica in the photobiological treatment, two strains of G. flavovirens PEWL001 and Psetr7 from California, USA and Aomori, Japan, respectively, exhibited the fastest silica uptake rate of approximately 33 mg L⁻¹ day⁻¹, followed by Halamphora sp. PEWL011, H. sydowii PEWL004, and Nitzschia sp. PEWL008. Three strains of Anomoeoneis species and one strain of Encyonema species were either less efficient in silica uptake and removal or unable to grow rapidly in the RO concentrate tested. The difference in the silica uptake kinetics among these strains might be due to the disparity between the water quality of RO concentrate, such as salinity and nutrient compositions, to be treated and their optimum growth conditions in their habitats. More research will be needed to explore the possible relationships and correlations. More than 95% of orthophosphate and ammonia, as well as a majority of the calcium, bicarbonate, iron and manganese, were removed from the RO concentrate with six diatom strains tested. The outcome of this study showed that the evaporation ponds in the Southwestern United States contained euryhaline brackish diatoms that could be useful for the treatment of RO concentrate. This will possibly lead to the development of more sustainable brackish water desalination and water purification processes for potable reuse.

The authors would like to thank Mr Dustin Fuller from Tulare Lake Drainage District, Corcoran, CA and Ms Linda Stevens from Dateland Public Service Company, Dateland, AZ for granting access to the evaporation ponds for water and sediment sampling. The authors would also like to thank Dr Kenneth P. Ishida, Ms Jana Safarik, and Dr Megan H. Plumlee from Orange County Water District for providing brackish wastewater samples. We thank Dr Eduardo Morales from Herbario Criptogámico, Universidad Católica Boliviana San Pablo, Cochabamba, Bolivia for his assistance on four diatom identifications (PEWL001-PEWL004). The assistance of Ms Yuan Li and Mr Steve Sanchez from Pacific Advanced Civil Engineering, Inc. in Fountain Valley, CA was also gratefully acknowledged. The materials presented in this paper are based upon work supported by the National Science Foundation under the Small Business Innovation Research Program (Award #: 1648495, PI: KI) and the United States Bureau of Reclamation Desalination and Water Purification Research Program (Award #: R21AC10106, PI: KI). Any opinions, findings, and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of the National Science Foundation. Pacific Advanced Civil Engineering, Inc. holds a U.S. patent (#9416036) relating to this photobiological treatment technology.

All relevant data are included in the paper or its Supplementary Information.

Some of the authors’ previous employer Pacific Advanced Civil Engineering, Inc. holds a U.S. patent (#9416036) relating to this photobiological treatment technology.