Journal of Environmental Chemical Engineering 11 (2023) 109208

Contents lists available at ScienceDirect

Journal of Environmental Chemical Engineering

journal homepage: www.elsevier.com/locate/jece

Microbial phosphorus removal and recovery by struvite biomineralisation

in comparison to chemical struvite precipitation in municipal wastewater
Yirong Leng , Ana Soares *
Cranfield University Water Sciences Institute, Cranfield MK43 0AL, UK

A R T I C L E I N F O A B S T R A C T

Editor: Dr Y Liu Microbial biomineralisation is attracting significant interest as an innovative process to recover nutrients from
wastes. Nevertheless, little is understood about the requirements to form struvite using biological pathways in
Keywords: wastewater and how does this compare with conventional chemical processes. To address this gap, Halobacterium
Bio-struvite salinarum, Bacillus pumilus, Brevibacterium antiquum, Myxococcus xanthus and Idiomarina loihiensis were grown in
Biomineralisation
wastewater to explore the relationships between cell growth, nutrients levels and properties of the recovered
Chemical precipitation
precipitates. The microorganisms were capable of removing ortho-phosphate (PO4-P) from municipal wastewater
Fertiliser value
Phosphorus removal at concentrations ranging from 5.4 to 62.4 mg PO4-P/L. Visible crystals of biological struvite (bio-struvite)
Phosphorus recovery (identified by morphology XRD and elemental analysis), were observed at PO4-P ≥ 19.7 mg/L, compared to
chemical struvite precipitation at 62.4 mg/L PO4-P (with pH adjustment). The initial nutrient concentrations
presented a strong positive correlation with bio-struvite production yields and crystal size distribution.
B. antiquum distinguished itself by relatively stable PO4-P removal (68–97%) independent of the initial nutrient
concentration, with effluents containing as low as 1 mg PO4-P/L. The recovered bio-struvite presented high
purity with low heavy metal contents, meeting regulations for inorganic fertiliser. The microbial processes for
phosphorus (P) recovery as bio-struvite presented several key advantages: bio-struvite crystals were released to
the wastewater and recoverable by filtration at PO4-P ≥ 19.7 mg/L, no need to adjust pH, bio-struvite crystals
had purity equivalent to 11.8–12.3% P and low heavy metal content, which was similar or better than that of
chemical struvite (12.6% P). This study validates bio-struvite’s relevance for low nutrient concentrations.

1. Introduction Chemical struvite is already a well-known way to recover P from

wastewater as value-added product [3]. Supersaturation is the main
P is an essential element to living organisms, and fertilisers con­ driving force of struvite precipitation: ortho-phosphate (PO3- 4 ) and
2+
taining phosphorus (P) are widely used in agriculture industry. The price ammonium (NH+ 4 ) integrated with Mg within a proper pH range to
of commercial inorganic P fertiliser has kept on increasing due to the form struvite when supersaturation condition in waste streams was
world-widely variability of phosphate rock reserves [1]. On the other achieved. A minimum of 100 mg PO4-P/L was reported to be required
hand, to meet more stringent legislation regarding P discharges (e.g. for readily spontaneous precipitation of struvite [5], which was 10 times
European Urban Wastewater Treatment Directive, Water Framework higher than typical values of raw municipal wastewater [6]. It has been
Directive), the improvement of wastewater treatment plants (WWTPs) proposed that P recovery via struvite is only economically viable at P
to remove P is imperative in many countries [2]. Processes that enable concentrations above 50 mg/L, when promoted by pH adjustment and
struvite (magnesium ammonium phosphate hexahydrate-MgNH4PO4.6 seeding for crystal nucleation [3,7]. Yet the potential P recovery from
H2O) production therefore have gained considerable attention from the WWTPs via struvite was quantified at a maximum of 30%, in relation to
water industry due to their capability to remove and recover P from the P available in raw wastewater [6]. The requirement of high P con­
wastewater streams [3]. Furthermore, owing to its high content of P and centration limited chemical struvite application at WWTPs because only
nitrogen (N), low degree of impurities and excellent slow-release prop­ P-rich waste streams (e.g. supernatant from anaerobic digester, centrate
erty [4], struvite has been recognized as a good alternative to most from sludge dewatering centrifuge) were economically rewarding [8,9].
P-rich fertilisers on market. Meanwhile, chemical usage for pH adjustment may lead to

* Correspondence to: Cranfield Water Science Institute, Cranfield University, Vincent Building, Cranfield, Bedfordshire MK43 0AL, UK.
E-mail address: a.soares@cranfield.ac.uk (A. Soares).

https://doi.org/10.1016/j.jece.2022.109208
Received 11 June 2022; Received in revised form 10 December 2022; Accepted 20 December 2022
Available online 23 December 2022
2213-3437/© 2022 The Author(s). Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
Y. Leng and A. Soares Journal of Environmental Chemical Engineering 11 (2023) 109208

environmental issue that minimised the potential benefits of struvite as 21545, German Resource Centre for Biological Material, Germany),
fertilizer. Myxococcus xanthus (CECT 422, Spanish Type Culture Collection, Uni­
Biological struvite (i.e., bio-struvite) has been proposed as an alter­ versity of Valencia, Paterna, Spain) and Idiomarina loihiensis (MAH1/
native way to recover P from wastewater. A wide spectrum of microbial CECT 5996, Spanish Type Culture Collection, University of Valencia,
strains have been reported to alter solution chemistry (e.g. pH, NH+ 4 ) by Paterna, Spain). All the microorganisms selected in this study have been
metabolic activities, promoting supersaturation conditions to form proven to form bio-struvite in streams (e.g. synthetic solution, seawater,
struvite [10]. The crystallization process of bio-struvite was also linked wastewater) [17,19].
with electrostatic interactions between cations (e.g. Mg2+, NH+ 4 ) in so­ Municipal wastewater was collected from outlet of primary lamella
lution and microbial cells/ extracellular substance secretions (e.g. clarifiers of a municipal WWTP with 2840 population equivalent (PE)
extracellular polymeric substances) with anionic nature [11,12], and (Cranfield, UK). The wastewater was filtered by a 10 µm nylon-mesh
between molecular structure of the crystal surface frameworks and filter (Plastok, UK), followed by microfiber (equivalent to 0.7 µm),
specific substances (e.g. polysaccharides) secreted by microorganisms and finally filtered sterilised (0.22 µm PVDF membrane, EMD Millipore,
[13]. Specific microorganisms (e.g. Brevibacterium antiquum, Parame­ UK). Filter sterilised (0.22 µm) concentrated solutions of magnesium
cium tetraurelia) were reported to absorb nutrients into cells for nucle­ sulphate heptahydrate (MgSO4•7 H2O), di-potassium hydrogen phos­
ation of intracellular struvite that was then assembled within cells or phate (K2HPO4), ammonium sulphate ((NH4)2SO4) were supplemented
secreted by cells, where specific regulations (e.g. controlled chemical together with bovine serum albumin (BSA) [20] as nutrient resources.
composition) in isolated compartments (e.g. membrane-bound lipid Different concentrations of PO4-P (5.4–62.4 mg/L, 4-levels) and NH4-N
vesicles) applied [14–16]. Compared with conventional chemical pre­ (35–332 mg/L, 3-levels) were applied to the filter-sterilised wastewater
cipitation, struvite biomineralisation involved various roles of micro­ (Table 1) to explore the significance of initial nutrient concentrations on
organisms to facilitate crystallization process and to exert control over the microbial capability for P removal and recovery. The concentrations
biomineral product (e.g., morphology, nucleation site), whereby it has of PO4-P, NH4-N, Mg2+, K+ and BSA were selected based on properties of
several advantages such as no pH-adjustment and recovering P at low wastewater reported suitable for bio-struvite formation and eligible
concentrations (e.g. 7.5 mg PO4-P/L) [17]. Bio-struvite production in wastewater sources at WWTPs [17,21].
wastewater were observed up to 21.5 mg P as bio-struvite/L in sludge
dewatering liquor, by pure cultures of B. antiquum [18]. 2.2. Microbial cultivation
However, the benefits of bio-struvite application to WWTPs in
comparison with standard chemical struvite precipitation, and the Starter cultures were grown in 250 mL E-flasks containing 100 mL
quality (e.g., purity, heavy metal contents) of bio-struvite recovered yeast extract solution of 4 g/L (additional 1% w/v NaCl and 1 g/L Mg2+
from wastewater are still poorly documented. This study aims to further to grow I. loihiensis), incubated on an orbital shaker (Stuart model SSL1,
investigate the importance of nutrient concentrations, including low Fisher Scientific, UK) at 150 RPM at room temperature (21–24 ◦ C) for
PO4-P, on microbial capability of P removal and recovery from waste­ 96 h. For inoculation in wastewater, the starter cultures were centri­
water, and on bio-struvite production and crystal size distribution, along fuged (Sanyo MSE Falcon 6/300 centrifuge, 2400 RCF, 4 ◦ C, 10 min) and
with an assessment of purity and heavy metal content of the recovered washed with sterile 0.9% w/v NaCl solution. The pure microorganism
bio-struvite in comparison with its chemical counterparts to reveal a pellets were re-suspended in sterile wastewater and inoculated in 40 mL
potential application of struvite biomineralisation at WWTPs. wastewater in 100 mL glass bottles (Pyrex, Fisher Scientific, UK).
Additional 0.8% w/v NaCl was added to I. loihiensis culture to ensure its
2. Material and methods growth [19]. The bottles were capped with foam stoppers and incubated
on orbital shaker at 150 RPM at room temperature for 196 h. The
2.1. Materials samples were taken at the start and at the end of incubation time. All
tests were completed in duplicate and controls were maintained under
The five microbial strains investigated in this study were wild-type identical conditions but without inoculation.
strains isolated from soil or sea water, and were purchased from com­
mercial culture collections: Halobacterium salinarum (DSM 671, German
Resource Centre for Biological Material, Germany), Bacillus pumilus
(GB43, LGC Standards, Middlesex, UK), Brevibacterium antiquum (DSM

Table 1
Characteristics of the raw wastewater collected from a municipal WWTP and the wastewater with different concentrations of NH4-N and PO4-P (average ± standard
deviation of duplicates) used to investigated struvite precipitation.
Molar ratio Saturation Index of PO4-P NH4-N pH Mg2+ Ca2+ K+ (mg/ SCOD
[PO4-P]:[Mg]:[ struvite (mg/L) (mg/L) (mg/L) (mg/L) L) (mg/L)
NH4-N]

Raw wastewater 1:2:14 -1.66 5.4 ± 0.3 35.0 ± 7.8 ± 8.2 ± 0.5 39.0 ± 21.0 ± 150 ± 5
0.6 0.1 0.6 0.2
Wastewater with different molar ratio ① 1:13:14 -1.04 5.4 35 7.8–8.0 56 39 93 750a
[PO4-P]:[Mg]:[ NH4-N] ② 1:4:4 -0.43 19.7
③ 1:2:2 -0.17 33.9
④ 1:1:1 0.12 62.4
⑤ 1:13:74 -0.35 5.4 180 7.8–7.9
⑥ 1:4:20 0.21 19.7
⑦ 1:2:12 0.44 33.9
⑧ 1:1:6 0.71 62.4
⑨ 1:13:136 -0.2 5.4 332 7.8–7.9
⑩ 1:4:37 0.35 19.7
⑪ 1:2:22 0.61 33.9
⑫ 1:1:12 0.84 62.4
a
containing 0.5 g/L BSA, equivalent to 600 mg/L SCOD

2
Y. Leng and A. Soares Journal of Environmental Chemical Engineering 11 (2023) 109208

2.3. Biological and chemical struvite preparation, isolation, purification the beaker, and photographs were taken. This was followed by
and identification volume-based particle size distribution (Dv) analysis using a Mastersizer
(Malvern 3000, Malvern Instruments Ltd, UK) with a Hydro EV disper­
An optical microscope (Division of GT vision Ltd., UK) was used for sion unit (stirring speed of 1200 RPM). A computer application, Visual
crystal morphology in fresh culture at the end of 196 h incubation. To MINTEQ ver. 3.1 [23], was used to quantify the saturation index of
prepare enough bio-struvite for identification, additional 4 L sterile struvite (SIstruvite) as an indicator of struvite recovery potential. This
wastewater with initial PO4-P of 33.9 mg/L and NH4-N of 332 mg/L was application is based on the thermodynamic equilibrium consists of
applied for each microbial strain under the identical condition (the same Mg2+, Ca2+, PO3- 4 , NH4 , etc. with a solubility product constant (Ksp) of
+
− 13.26
as described in Section 2.2) for 196 h incubation. At the end of incu­ 10 [3]. The SIstruvite values determined by the investigated initial
bation period, precipitants were separated from the liquid by filtration nutrient levels in wastewater were presented in Table 1.
(10 µm nylon-mesh filter, Plastok, UK) and washed with deionized (DI)
water twice, passed for air-dry at 37 ◦ C for 4 h, and finally weighed to 2.5. Statistical analysis
determine the yield. Only the extracellular biological precipitants and
chemical precipitants more than 10 µm were collected for determination The statistically significant difference in terms of P removal effi­
and further analysis. ciency, crystal production and crystal size (by Dv50 - the particle median
Chemical struvite was prepared in filter-sterilised wastewater the diameter for a volume distribution) was investigated by two-tailed T-test
same source as for bio-struvite: 100 mL solution containing 0.2 M (significance level α = 0.05) in relation to the different microbial strains,
NH4H2PO4 and 0.001 M K2HPO4 was added to 200 mL solution of 0.05 concentrations of initial NH4-N and PO4-P. Statistical tests were
M MgSO47 H2O; both solutions were pre-adjusted to pH 9 by 1 M sodium considered significant at p < 0.05. The 95% confidence intervals (CIs)
hydroxide (NaOH) and mixed in a glass bottle (1 L, Duran™, Fisher were also examined to estimate the range of mean values of P removal
Scientific, UK) [20]. The mixture was agitated at 150 RPM at room efficiency, crystal production and Dv50. The statistical correlation was
temperature for 24 h for chemical struvite precipitation, followed by applied to distinguish the significant relationship between eight vari­
filtration and purification that was the same as for bio-struvite. ables, including initial parameters (pH, NH4-N and PO4-P, molar ratio of
The solid phases of purified biological and chemical precipitants initial [PO4-P] to [NH4-N], solution SIstruvite) and PO4-P removal effi­
were characterised by scanning electron microscope equipped with en­ ciency, crystal production and Dv50, where the critical significance
ergy dispersive X-ray spectroscopy (SEM-EDS, XL 30 SFEG, Phillips, The levels (R) was applied: 0.6 ≤R < 0.8 was considered strong correlation,
Netherlands) where samples were placed on carbon tabs and sputter and R ≥ 0.8 was considered very strong correlation. All the statistical
coated with gold. The X-ray powder diffractometer (XRD, D5000, analysis was performed using Microsoft Excel 2010 (Microsoft, Red­
Siemens/Bruker, Germany) analysis of precipitants were performed in mond, Washington, USA).
the range of 20–90◦ with a step size of 0.04◦ and a step time of 1 s.
Diffraction data were analysed by DIFFRAC.SUITE™ EVA software 3. Results and discussion
(Bruker AXS GmbH, Germany) and database from the International
Centre for Diffraction Data (ICDD®). The chemical characterisation of 3.1. Microbial growth and crystal identification
both biological and chemical precipitants, including PO4-P, NH4-N,
Mg2+, Ca2+, K+, Al3+ and heavy metals, was investigated by SEM-EDS or Microbial growth of all the five selected microorganisms was
crystal dissolution of 1.25 g/L (prepared in extra pure water pre- observed in each of the twelve tests (①–⑫, Table 1) with different
adjusted to pH 2 by 1 M hydrogen chloride). concentrations of PO4-P and NH4-N (Fig. 1). The microbial intact cell
count increased overall by 1–2-fold at the end of 196 h incubation period
2.4. Analytical methods (Fig. 1). As the initial NH4-N increased from 35 to 332 mg/L, the final
intact cell count of B. antiquum and B. pumilus increased by 9% and 28%,
The intact cell counts in microbial culture were estimated with a flow respectively (Fig. 1). The positive effect of NH4-N on B. antiquum growth
cytometer (BD Accuri C6, BD Biosciences, US) analyses using SYBR was in agreement with previous observation that NH4-N (510–839 mg
Green I (SG) - propodium iodide (PI) co-staining method [22]. The flow NH4-N/L) contributed to growth rate by 0.38 1/d in sludge dewatering
cytometer was also used to distinguish inorganic nanoparticles from liquors [18]. The microbial growth was correlated with SCOD reduction,
cells by SG staining method. Assuming there was no free deoxy­ which varied with the microbial species: 54–57% for B. antiquum,
ribonucleic or ribonucleic acid in microbial cultures, the inorganic 46–50% for B. pumilus, 37–43% for M. xanthus, 27–31% for H. salinarum
nanoparticles counts were estimated by using total nanoparticle counts and 26–28% for I. loihiensis. Neither intact microbial cell nor SCOD
minus total cell counts. The concentrations of PO3- 4 and NH4 were ana­
+
removal was observed in the non-inoculated controls.
lysed using a Smartchem (Smartchem200, AMS/Alliance Instruments, By the end of incubation period, the crystals recovered by each of the
France) according to the manufacturer’s instructions (Labmedics, five tested microbial strains from municipal wastewater possessed
Abingdon, UK). Magnesium and calcium (Ca2+) were analysed by dominantly the same elongated trapezoidal platy shape (Fig. 2a-e,
atomic absorption spectroscopy (AAS, Analyst 800, PerkinElmer, UK) Fig. SA) as the struvite previously described [24]. No struvite precipitate
equipped with flaming and electrothermal spectrometer. Aluminium was observed in non-inoculated controls with constant pH level
(Al), potassium (K) and heavy metals including chromium (Cr), iron (7.8–8.0). The XRD results (Fig. SB) demonstrated that the crystals
(Fe), nickel (Ni), copper (Cu), arsenic (As), cadmium (Cd), lead (Pb) and biologically recovered from wastewater (initial PO4-P of 33.9 mg/L and
mercury (Hg) were analysed by inductively coupled plasma mass spec­ NH4-N of 332 mg/L) had similar peak profiles to the standard struvite
trometry system (ICP-MS, NexION 350, PerkinElmer, UK). The pH was (pattern COD 9007674), and thus they were of the same orthorhombic
determined by digital pH-meter (Jenway 3540, Bibby Scientific, UK). crystal structure as struvite. Furthermore, the SEM-EDX and crystal so­
Soluble chemical oxygen demand (SCOD) was analysed by Spec­ lution showed the stoichiometric ratio [Mg]:[P]:[N]:[O] of the crystals
troquant® cell test kit (Merck, VWR, UK). To avoid loss of microbial cells was 1:1:1:4, corresponding to the chemical formula of struvite in
and fine particles (<10 µm) in filtration (for determination of crystal dehydrated phase (MgNH4PO4). The observed loss of oxygen (by mo­
production yield), that may affect the subsequent analysis of particle lecular water) and N (by ammonia) within crystalline framework of both
size distribution, qualitative assessment (QA) of crystal production was bio-struvite and chemical struvite in this study was due to the air-drying
applied for each test. It was completed by transferring the 40 mL method which resulted in volatilization of ammonia and molecule water
wastewater and crystal precipitates to a 500 mL glass beaker, stirring the generally on crystal surface, as previously reported [25].
wastewater clockwise to allow crystals settled at the bottom centre of Both biological and chemical struvite crystals were observed to have

3
Y. Leng and A. Soares Journal of Environmental Chemical Engineering 11 (2023) 109208

Fig. 1. Natural logarithm of intact cell counts of the H. salinarum, B. antiquum, B. pumilus, M. xanthus and I. loihiensis in municipal wastewater at initial and by the end
of 196 h incubation at different molar ratio of initial nutrient concentrations within PO4-P range of 5.4–62.4 mg/L in wastewater. Error bars represent standard
deviation obtained from duplicates.

porous surface (Fig. 2g-i), which was similar as the previously reported NH4-N correlated with PO4-P removal and removal efficiency (Fig. 3a).
surface structures of chemical struvite [26] and bio-struvite [13]. Porous Within an initial PO4-P range of 19.7–62.4 mg/L, the P removal effi­
surface originated from oriented aggregation and embedding of crys­ ciency was increased by 22% (H. salinarum), 23% (B. antiquum), 24%
talline subunits into the crystalline framework. The comparison of the (M. xanthus), 28% (B. pumilus) and 66% (I. loihiensis) as NH4-N con­
particle’s matrix on the crystal surface in this study showed slight dif­ centration increased (Fig. 3a). The variation of P removal efficiency
ference in terms of the porosity and building units: wide gap between within range of initial PO4-P between 5.4 and 62.4 mg/L were 29%,
subunits (high porosity) was observed on the bio-struvite crystal surface 38%, 65%, 71% and 77% for B. antiquum, B. pumilus, I. loihiensis, M.
structure. And the particle size of the small building units for bio-struvite xanthus and H. salinarum, respectively (Fig. 3a). No PO4-P removal was
was relatively larger (1.5–2 µm x 1–1.5 µm) (Fig. 2g-h) than the chem­ observed in non-inoculated control.
ical struvite’s (0.8 µm x 0.4 µm) (Fig. 2i). Supersaturation affected B. antiquum presented relatively constant P removal efficiency
crystal size and size distribution [27]. High supersaturation boosts (68–97%) overall the investigated nutrients’ concentrations when
nucleation rate, and thus produces large number of fine particles. In compared with the other four microbial strains (Fig. 3a), indicating that
contrast, low supersaturation is typically associated with crystals of the bacterium’s capability of P removal was less dependent on the initial
large size and lower content of fine particles [27]. The observation of PO4-P level in solution. The highest PO4-P removal efficiency by
relatively larger subunits of bio-struvite crystal surface indicated the B. antiquum (87–97%) occurred at 5.4 mg PO4-P/L, and achieved final
crystallisation process occurred under lower supersaturation where ≤ 1 mg PO4-P/L (Fig. 3a). The effluent met the EU standard for P
more component ions were used for crystal growth other than nucle­ discharge of 2 mg/L (WWTPs of 10,000–100,000 PE) or 1 mg/L
ation. Furthermore, the high porosity of bio-struvite crystals increased (WWTPs >100,000 PE) [30]. Such performance was owing to a bio­
sites where electrostatic interaction between crystal surface structure logical control mineralisation process where component ions (e.g., PO3- 4,
and charged particles (e.g., ions, cell out-layer, extracellular polymers) Mg2+) were pumped into B. antiquum cell and accumulated in com­
may occur, and therefore benefited crystal growth. partmentalized compartments such as vesicle-like structure, for inte­
Low content of Ca2+ less than 0.5 wt percent (wt%) was detected by gration with NH+ 4 to form bio-struvite under specialized regulations (e.
SEM-EDX on the bio-struvite struvite surface, whereas the Ca2+ on g., controlled chemical composition) [31]. Such regulations ensure
chemical struvite crystal surface was found 1.67–8.24 wt%. The Ca2+ species-specific crystallo-chemical properties therefore the biomineral
was reported to affect the morphology and purity of chemical struvite by type and habit [32]. In contrast, in a biological induced mineralisation
competitively combining with PO4-P at pH ≥ 9 to aggregate on the process (B. pumilus, M. xanthus and H. salinarum) where microorganisms
struvite surface [28]. A similar structure as previously reported calcium exerted limited control on bio-struvite crystallisation, microbial growth
phosphate (Ca-P) precipitation in struvite− water interfacial layer [29] and activity benefited the process mainly by heterogenous nucleation
was observed on the chemical struvite in this study (Fig. 2i). However, and metabolic activity (altering pH and NH4-N) to achieve supersatu­
such Ca-P mineral layer was not observed on the bio-struvite surface in ration condition [31]. The capability of P removal by microorganisms
this study. that were involved in bio-struvite producing via biological control
mineralisation was theocratically independent on nutrient concentra­
3.2. Phosphate removal, crystal production and size distribution tion, although the P removal efficiency by I. loihiensis varied from 16% to
82% overall the investigated PO4-P and NH4-N concentrations. It is
The capability of microorganisms to remove PO4-P from wastewater noted that the intact cell counts of I. loihiensis, halophilic only doubled
varied with microbial strains. The maximum P removal at an initial PO4- after 192 h incubation (within stationary phase of growth) (Fig. 1), and
P concentration of 5.4 mg/L was 26% by M. xanthus, and all the way up the utilisation of dissolved organic carbon was up to 210 mg/L (28%)
to 97% by B. antiquum (Fig. 3a). The initial concentrations of PO4-P and SCOD removal. Whereas in a previous study where it was grown in

4
Y. Leng and A. Soares Journal of Environmental Chemical Engineering 11 (2023) 109208

Fig. 2. SEM photo of the isolated bio-struvite crystals recovered from (a) H. salinarums, (b) B. pumilus, (c) M. xanthus, (d) B. antiquum, (e) I. loihiensis cultures in
wastewater (196 h); (f) overview of chemical struvite (scale bar - 200 µm) and single crystal (white square, scale bar - 10 µm). Crystal surface of bio-struvite produced
by B. antiquum (g), M. xanthus (h)) and abiotic struvite (i) (scale bar - 5 µm), white squares indicate the single building unit of struvite aggregated to the crystal
surface and black square indicate the nano-calcium phosphate layer.

solution containing 1% NaCl and 4 g/L yeast extract, its intact cell potential to compete with struvite formation for PO3- 4 , typically pro­
counts 20-folded after 96 h by consuming 1680 mg/L (48%) SCOD [33]. ducing amorphous calcium phosphates [3]. The presence of Ca2+ exer­
Growth of I. loihiensis required NaCl of 0.7–20%., and with optimum ted negative impact on struvite purity within pH range of 7–10.5, where
NaCl between 2% and 6% [19]. It was suggested that the low NaCl level the struvite content decreased from around 70–30% as pH increased
(0.8%) applied in this study slowed down its cell proliferation and usage from 8.0 to 9.0 [34]. Calcium was also capable to interact with organic
of carbon source, thus limited this halophilic bacterium’s capability of P molecules for biological induced mineralisation where the increasing
removal. alkalinity, electron-interaction and heterogeneous surface might benefit
The microbial strains were observed to produce bio-struvite crystal formation of Ca-species [19].
in wastewater containing initial PO4-P of 19.7–62.4 mg/L (Fig. 3b). In Positive relationships between Dv50 and initial nutrient concentra­
particularly, only B. antiquum and B. pumilus produced bio-struvite at tion were observed (Fig. 3c). Large bio-struvite crystal groups (Dv50
initial PO4-P of 19.7 mg/L, NH4-N of 35 mg/L and Mg2+ of 56 mg/L ≥100 µm) formed at initial NH4-N of 180 mg/L and initial PO4-P of
(corresponding to initial SIstruvite of − 0.43) (Fig. 3b). No precipitant was 62.4 mg/L, and at initial NH4-N of 332 mg/L and initial PO4-P between
observed in non-inoculated controls. The presence of magnesium in 19.7 and 62.4 mg/L (Fig. 3c). At the end of 196 h incubation, neither
solution is not only essential to microbial growth (especially I. loihiensis) PO4-P removal nor crystal production was observed in non-inoculated
but also significant for bio-struvite production due to the crystals’ sol­ controls. The pH of controls was then raised to around 8.5 by 1 M
ubility equilibrium. For all inoculated wastewater in this study, the final NaOH. After 72 h incubation, precipitates with orthorhombic crystal
concentrations of Mg2+ was observed between 15.0 and 54.7 mg/L, thus structure as struvite (Fig. 3b) were observed only in wastewater con­
it did not limit bio-struvite production. A considerable amount of Ca2+ taining PO4-P of 62.4 mg/L and NH4-N of 332 mg/L, with PO4-P
(19–21 mg/L) was removed from wastewater with initial molar ratio removal of 38 mg/L (61%), chemical struvite production of 2.5 QA and
[PO3- 2+ 2+
4 ]:[NH4 ]:[Mg ]:[Ca ] of 2:2:2:1, and with final pH of 8.4–8.5.
+
Dv50 of 78 µm (Fig. 3a-c). The estimated initial SIstruvite for chemical
This was corresponding to an occurrence of an abundant of inorganic struvite precipitation in this study was 1.52, which was in the range of
particles in microbial cultures (inorganic particles/total particles of previously reported values from 1.5 to 10.6 for spontaneous precipita­
13–76%) and observation of amorphous nanoparticles. Calcium had tion [27], but higher than that for bio-struvite formation (− 0.43 to 0.84)

5
Y. Leng and A. Soares Journal of Environmental Chemical Engineering 11 (2023) 109208

Fig. 3. The PO4-P removal efficiency (a), struvite

production (by qualitative assessment) (b) and Dv50
(c) at different molar ratio of initial nutrient con­
centrations and the corresponding SIstruvite values
(underlined) within PO4-P range of 5.4–62.4 mg/L in
wastewater. The five microorganisms shared same
initial SIstruvite values (underlined) for each molar
ratio of initial nutrient concentrations. Error bars
represented standard deviation obtained from du­
plicates. Dv50 represented the particle median
diameter for a volume distribution. The PO4-P
removal and chemical struvite in non-inoculated
control were only observed after raising pH to 8.5
by adding sodium hydroxide (photo with scale bar –
10.19 µm).

in this study (Fig. 3b), and the previously reported initial SIstruvite of 3.3. Importance of initial nutrient concentrations on P removal and bio-
0.6–0.8 in synthetic solution and wastewater [31,33]. struvite product
When compared the struvite biomineralisation with chemical stru­
vite precipitation in this study, the former recovered struvite by larger B. antiquum was observed to distinguish itself from the other four
crystal groups, at lower initial PO4-P and without chemical dose for pH microbial strains regarding P removal efficiency in wastewater
adjustment. Although low level SIstruvite slowed down the nucleation (Table 2). There was no significant difference among the five tested
rate, specific metabolic pathways and organic cellular substance/ microorganisms regarding crystal size (Dv50) and production, but a
structures involved in struvite biomineralisation reduced energy barrier significant difference between the inoculated and non-inoculated con­
for heterogeneous nucleation of bio-struvite and promoted its crystalli­ trols was observed (Table 2). With respect to different initial PO4-P
zation process [10]. The low SIstruvite then benefited crystal growth (e.g., concentrations, there was significant difference in terms of P removal
crystal size, morphology) by competitive for component ions to promote efficiency when initial PO4-P increased from 5.4 and 19.7 mg/L, and in
the crystals’ settleability [27]. terms of Dv50 and crystal production when initial PO4-P increased from
5.4 to 19.7 and to 33.9 mg/L (Table 2). Thus, for the investigated mi­
croorganisms, the significant initial PO4-P concentration that impacted

6
Y. Leng and A. Soares Journal of Environmental Chemical Engineering 11 (2023) 109208

Table 2 initial NH4-N concentration presented a positive correlation with bio-

Statistical significance (two-tailed T-test) and 95% confidence intervals with struvite crystal size for all the tested microorganisms.
respect to PO4-P removal, Dv50 and crystal production at different concentra­ The 95% confidence intervals (CIs) of P removal efficiency (80–91%)
tions of initial NH4-N and PO4-P in the five selected microbial cultures. by B. antiquum showed a relatively high PO4-P removal within a narrow
Two-tailed T-test range, when compared with that of the other four microbial strains
Between groups p values (68–84% of B. pumilus, 51–82% of M. xanthus, 50–83% of H. salinarums
and 41–69% of I. loihiensis) (Table 2). This indicated a relative stable
PO4-P removal Dv50 Crystal
efficiency production
high P removal efficiency by B. antiquum within the investigated
nutrient concentrations, which was in agree with the observation of P
Microbial B. pumilus and 0.233 0.980 0.937
removal efficiency by B. antiquum in Section 3.2. The 95% CIs of Dv50 at
strain M. xanthus
B. pumilus and 0.244 0.761 0.795 initial 332 mg NH4-N/L, 33.9 mg PO4-P/L and 62.4 mg PO4-P/L were of
H. salinarum 72–141, 36–118 and 55–133 µm, respectively (Table 2). This indicated
B. pumilus and 0.048 0.834 0.681 potential formation of large bio-struvite crystal groups (Dv50 >100 µm)
B. antiquum at initial NH4-N of 332 mg/L combined with initial PO4-P ≥ 33.9 mg/L.
B. pumilus and 0.010 0.754 0.970
I. loihiensis
The highest 95% CIs of crystal production occurred at initial 62.4 mg
B. pumilus and 0.000 0.033 0.008 PO4-P/L, followed by initial nutrient concentrations of 33.9 mg PO4-P/L
control and then 332 mg NH4-N/L (Table 2).
M. xanthus and 0.991 0.776 0.849 Initial PO4-P, NH4-N and SIstruvite were identified to have significant
H. salinarum
positive correlations with P removal efficiency, bio-struvite production
M. xanthus and 0.024 0.850 0.729
B. antiquum and Dv50 for the five tested microorganisms (Table 3). Compared with
M. xanthus and 0.234 0.768 0.969 the initial PO4-P and NH4-N, the SIstruvite was observed to have stronger
I. loihiensis correlations with Dv50 and bio-struvite production (Table 3). The P
M. xanthus and 0.000 0.026 0.005 removal efficiency by B. antiquum was of less dependence on these three
control
H. salinarum and 0.028 0.930 0.884
initial parameters than the other four microorganisms; whereas the P
B. antiquum removal efficiency by B. pumilus highly depended on the initial SIstruvite
H. salinarum and 0.250 0.986 0.828 (Table 3).
I. loihiensis The significant initial PO4-P concentration of 19.7 mg/L for P
H. salinarum and 0.000 0.018 0.006
removal and 33.9 mg/L for bio-struvite production and Dv50 suggested
control
B. antiquum and 0.001 0.917 0.715 wider range of wastewater sources at municipal WWTPs for struvite
I. loihiensis biomineralisation than struvite chemical precipitation that frequently
B. antiquum and 0.000 0.026 0.004 sourced from P-rich liquids with P above 50 mg/L [9]. A potential to
control produce large bio-struvite crystal groups with Dv50 more than 100 µm
I. loihiensis and 0.000 0.022 0.009
control
for readily settlement in wastewater at initial nutrient concentrations of
Initial PO4- 5.4 mg/L and 0.007 0.014 0.000 high levels (NH4-N = 332 mg/L, PO4-P ≥ 33.9 mg/L) added benefits to
P 19.7 mg/L struvite biomineralisation. The application of struvite biomineralisation
19.7 mg/L and 0.336 0.046 0.024 at WWTPs was further discussed in Section 3.5.
33.9 mg/L
Significant positive correlations were also observed among P
33.9 mg/L and 0.952 0.532 0.073
62.4 mg/L removal efficiency, Dv50 and bio-struvite production, especially be­
Initial NH4- 35 mg/L and 0.068 0.001 0.068 tween Dv50 and bio-struvite production (Table 3). This indicated a po­
N 180 mg/L tential of high P removal efficiency integrated with P recovery by bio-
180 mg/L and 0.606 0.003 0.085 struvite of more production and larger size, as shown in a 3D scatter
332 mg/L
95% confidence intervals (CI)
plot, where all the large 3D spheres (crystal production, QA≥3) were
Range of mean values filled with colour representing high Dv50 (≥80 µm), and appeared with
PO4-P removal Dv50 Crystal P removal efficiency more than 75% (Fig. SC).
efficiency (%) (μm) production
(QA)
Microbial B. pumilus 68–84 4–88 0.4–2.1 3.4. Struvite purity and heavy metal content
strain M. xanthus 51–82 7–87 0.5–2.1
H. salinarum 50–83 11–98 0.5–2.3 No significant difference was observed regarding the purity between
B. antiquum 80–91 7–96 0.6–2.3
I. loihiensis 41–69 10–100 0.4–2.1
bio-struvite recovered from wastewater with PO4-P of 33.9 mg/L and
Initial PO4- 5.4 mg/L 34–63 1 0 chemical struvite recovered from PO4-P more than 2000 mg/L (Table 4).
P 19.7 mg/L 63–83 8–54 0.6–1.4
33.9 mg/L 71–87 36–118 1.2–2.3
Table 3
62.4 mg/L 73–85 55–133 1.8–3.3
Significant correlations between the initial parameters and P removal, Dv50 and
Initial NH4- 35 mg/L 50–70 2–4 0.2–0.6
N 180 mg/L 62–84 22–64 0.9–1.9 crystal production (R ≥0.6) in • B. pumilus, ▴ M. xanthus, ◆ H. salinarum, ■
332 mg/L 68–85 72–141 1.5–2.8 B. antiquum, ★ I. loihiensis cultures.
P removal Dv50 (μm) Crystal production
efficiency (QA)
P removal efficiency and bio-struvite production and crystal size were
R 0.6 ≤R R 0.6 ≤R R ≥ 0.8 0.6 ≤R
19.7 mg/L and 33.9 mg/L, respectively. Initial nutrient concentrations
≥ 0.8 < 0.8 ≥ 0.8 < 0.8 < 0.8
above the significant values did not significantly change the P removal
Initial NH4-
efficiency, bio-struvite production and size. Moreover, Dv50 varied ▴◆★ ●▴◆■ ◆
N (mg/L) ★
significantly as the initial NH4-N increased from 35 to 180 and to Initial PO4- ● ■ ●▴■★
332 mg/L, but the changes of initial NH4-N concentrations did not P (mg/L)
significantly correlate with the P removal efficiency and bio-struvite Initial ● ▴◆★ ▴◆■ ● ●▴◆
production (Table 2). Therefore, within the investigated range, the SIstruvite ★ ■★
Dv50 (μm) ●★

7
Y. Leng and A. Soares Journal of Environmental Chemical Engineering 11 (2023) 109208

Table 4
Quantification of macronutrients and heavy/toxic metal in bio-struvite and chemical struvite (average ± standard deviation of triplicates).
B. antiquum B. pumilus H. salinarum M. xanthus I. loihiensis Chemical Theoretical EU permissible levels
struvite struvitea

Macronutrient PO4-P (g/ 122.6 ± 3.3 121.9 ± 3.9 121.7 ± 1.9 118.1 ± 2.8 121.9 ± 3.3 127.2 ± 3.6 12.6% P 2% (total P2O5)bd
kg) (12.3% P) (12.2% P) (12.2% P) (11.9% P) (12.2% P) (12.7% P)
NH4-N (g/ 57.7 ± 0.2 56.2 ± 1.1 57.1 ± 0.4 53.94 ± 0.8 55.6 ± 0.3 52.42 ± 0.3 5.7% N 2.5% (total N)bd
kg) (5.8% N) (5.6% N) (5.7% N) (5.4% N) (5.6% N) (5.2% N)
Mg (g/kg) 97.7 ± 0.9 96.2 ± 0.3 97.1 ± 0.7 93.1 ± 0.6 94.6 ± 0.6 96.0 ± 1.2 16.4% MgO NYD
(16.2% (16.0% (16.2% (15.5% (15.8% (16.0% MgO)
MgO) MgO) MgO) MgO) MgO)
K (g/kg) 0.8 ± 0.0 0.7 ± 0.0 0.7 ± 0.0 0.7 ± 0.0 0.6 ± 0.0 0.2 ± 0.0 0 2% (water-soluble
K2O)bd
Ca (g/kg) 0.2 ± 0.0 0.2 ± 0.0 0.2 ± 0.0 0.2 ± 0.1 0.1 ± 0.0 0.8 ± 0.0 0 NYD
Heavy/toxic Al (mg/ 18.5 ± 0.7 12.0 ± 0.4 22.9 ± 0.8 24.8 ± 0.5 26.3 ± 0.3 31.2 ± 0.9 0 NYD
metal kg)
Fe (mg/ 2.65 ± 0.23 2.19 ± 0.62 <LOD 3.25 ± 0.45 4.18 ± 0.07 8.69 ± 0.38 0 NYD
kg)
Cr VI (mg/ 0.17 ± 0.02 0.23 ± 0.03 0.15 ± 0.01 0.26 ± 0.02 0.23 ± 0.01 0.38 ± 0.00 0 2 mg/kgce
kg)
Ni (mg/ 0.11 ± 0.00 0.08 ± 0.04 0.07 ± 0.01 0.11 ± 0.03 0.14 ± 0.01 1.19 ± 0.10 0 100 mg/kgcd
kg)
Cu (mg/ 0.33 ± 0.19 0.53 ± 0.18 0.18 ± 0.02 1.57 ± 0.26 1.24 ± 0.03 0.42 ± 0.03 0 200 mg/kgce
kg)
As(mg/kg) 0.38 ± 0.01 0.30 ± 0.02 0.32 ± 0.01 0.27 ± 0.04 0.39 ± 0.04 0.07 ± 0.02 0 40 mg/kgcd
Cd (mg/ 0.06 ± 0.05 0.66 ± 0.01 <LOD 0.04 ± 0.02 0.26 ± 0.02 0.00 ± 0.00 0 60 mg/kg (if P2O5
kg) >5%)cd
Pb (mg/ 0.18 ± 0.03 0.32 ± 0.04 0.11 ± 0.01 0.38 ± 0.16 0.36 ± 0.02 0.40 ± 0.02 0 120 mg/kgcd
kg)
Hg (mg/ 0.03 ± 0.01 0.11 ± 0.05 0.03 ± 0.01 0.03 ± 0.00 0.02 ± 0.01 0.00 ± 0.00 0 1 mg/kgcd
kg)

LOD - limit of detection; NYD – not yet defined

a
- Theoretical struvite with chemical formula: MgNH4PO4.6 H2O; b- Minimum nutrient content in solid inorganic fertilisers; c- maximum heavy metal content in
inorganic fertilisers; d - The European parliament and the council of the European union [36]; e - EU Fertiliser Regulation revision proposal 2014 [35]

Bio-struvite contained nutrients equivalent to 11.8–12.3% P, 5.4–5.8% [37,38]. Thus, the bio-struvite recovered from primary settled munic­
N and 15.5–16.2% MgO, and the chemical struvite contained nutrients ipal wastewater in this study is of better-quality regarding P and heavy
equivalent to 12.6% P, 5.7% N and 16.4% MgO (Table 4). The purity of metal content. The impurity of bio-struvite may due to the influence of
both products was within the proposed acceptable purity range of the heavy metal-concentrated environment. Verma and Kulia, 2019 re­
struvite for fertiliser usage, with requirement of nutrients’ content ported a great potential of EPS to bind heavy metal ions (e.g., Cd, Cu, Pd,
equivalent to 10.0–13.9% P, 4.6–6.3% N and 13.1–18.1% MgO [35]. Hg, Fe, Cr, Ni) by metabolic pathway. Such microbial biomineralisation
Both P and N content of the bio-struvite met the EU minimum require­ may lead to precipitation of mineral containing inorganic pollutants
ment as inorganic fertiliser, thus it has a potential use as P and N fer­ under proper conditions [39]. Moreover, microorganism may possess
tiliser alternative. Bio-struvite was also observed contain K resistance or tolerance to specific heavy metal (e.g. Cd) by accumulating
(0.6–0.8 g/kg) and Ca (0.1–0.2 g/kg) that were recognized as essential and concentrating them within cells [40]. These microbial cells have
macronutrients for organism growth (Table 4). Although the K content potential to release the heavy metal to the mineral surface during
in bio-struvite was relatively higher than chemical struvite (0.2 g/kg), it struvite biomineralisation, via molecular interaction between cells and
was still far below than the EU regulated level of K in inorganic fertiliser the crystal surface [10].
[36].
Settled wastewater contained various heavy/toxic metals (Table SA), 3.5. Implication for wastewater industry
although the amount of heavy metal Cr, Ni, Cu, As, Cd, Pb and Hg in
both bio-struvite and chemical struvite products (Table 4) met the EU Most of the previously reported struvite process were applied to
maximum limitation [35,36]. Compared with the chemical struvite, the waste streams rich of PO3- 4 and NH4 , such as side streams and liquors
+

bio-struvite contained lower amounts of Ca, Al, Fe, Cr, Ni and Pb, but from anaerobic digestion [6,9,41]. Chrispim et al., 2019 suggested
relatively higher amount of Cu, As, Cd and Hg (Table 4). In particular, several preferred wastewater sources for struvite precipitation at
bio-struvite produced by M. xanthus and I. loihiensis was observed con­ WWTPs, mostly digester supernatant and sludge dewatering liquors, and
tained higher Al (24.8–26.3 mg/kg), Fe (3.25–4.18 mg/kg), Cr highly dependent on the performance of sludge digestion to generate
(0.23–0.26 mg/kg), Ni (0.11–0.14 mg/kg) and Pb (0.36–0.38 mg/kg) enough nutrient constituents. In this study, P removal and P recovery by
than the other three microbial strains (Table 4). The highest Cd content biomineral forming microorganisms were observed at relatively low
(0.66 mg/kg) was found in the B. pumilus bio-struvite, compared with PO4-P of 5.4 and 19.7 mg/L, respectively. Thus, compared with chem­
the other four bio-struvite (0–0.26 mg/kg) and the chemical struvite ical struvite precipitation, struvite biomineralisation could be applied to
(<1 μg/kg) (Table 4). wastewater streams with a much lower and wider range of PO4-P con­
The bio-struvite obtained in this study was found of similar purity as centrations. The eligible wastewater sources at WWTPs for bio-struvite
previously reported bio-struvite (11.9% P) which was recovered from formation could range from raw municipal wastewater (4–15 mg
sludge dewatering liquors, but lower of heavy metal contents, especially PO4-P/L and NH4-N > 35 mg/L) to settled wastewater (7.5 mg PO4-P/L)
the Hg, As and Ni [21]. Furthermore, when comparing with the and to digester supernatant (20–400 mg PO4-P/L) [6,17]. However,
bio-struvite produced from sewage sludge ash (SSA), which was also aeration and nitrogenous organic compounds are also required to pro­
proposed as a P-fertiliser alternative recovered from WWTPs, it was vide enough dissolved oxygen and carbon sources for microbial growth
observed that the previously described SSA bio-struvite contained lower and specific metabolic pathways [20]. In particular, B. antiquum may
P content (9.4% P, 10.56% P) but higher toxic/heavy metal contents have potential use to achieve high PO4-P effluent quality due to the

8
Y. Leng and A. Soares Journal of Environmental Chemical Engineering 11 (2023) 109208

observation of its capability in this study to reduce PO4-P from 5.4 mg/L Data Availability
to less than 1 mg/L. The application of microbial strains for P removal
and P recovery may vary with wastewater characteristics. In some Data will be made available on request.
WWTPs as the one described by Jaffer et al. (2002) for example, all the
wastewater streams presented conditions suitable for microbial growth Appendix A. Supporting information
of the tested microbial strains [20], B. antiquum and B. pumilus can be
applied for P removal from raw and settled wastewater (PO4-P of Supplementary data associated with this article can be found in the
5.7–14.2 mg/L, NH4-N of 16.1–23.9 mg/L, Mg2+ of 8–8.9 mg/L), and online version at doi:10.1016/j.jece.2022.109208.
additional Mg2+ source might be required. In sludge dewatering liquors
(PFT liquor, centrifuge liquor) containing 33.2–94.9 mg PO4-P/L, References
129–615 mg NH4-N/L and 13–44 mg Mg2+/L, all the five tested mi­
croorganisms can be applied for P recovery by bio-struvite. In particular, [1] L. Egle, H. Rechberger, J. Krampe, M. Zessner, Phosphorus recovery from
municipal wastewater: an integrated comparative technological, environmental
the centrifuge liquor was higher in pH, PO4-P, NH4-N and Mg2+, and and economic assessment of P recovery technologies, Sci. Total Environ. 571
lower in Ca2+ [41], thus it is anticipated with better performance than (2016) 522–542, https://doi.org/10.1016/j.scitotenv.2016.07.019.
PFT liquor regarding crystallization process and crystal purity. [2] A.T.W.M. Hendriks, J.G. Langeveld, Rethinking wastewater treatment plant
effluent standards: nutrient reduction or nutrient control? Environ. Sci. Technol. 51
In this study the concentrations of nutrients in the wastewater were (2017) 4735–4737, https://doi.org/10.1021/acs.est.7b01186.
manipulated to investigate the relationships between cell growth, [3] N. Krishnamoorthy, B. Dey, Y. Unpaprom, R. Ramaraj, G.P. Maniam, N. Govindan,
wastewater composition in relation to nutrients, and the property of the S. Jayaraman, T. Arunachalam, B. Paramasivan, Engineering principles and process
designs for phosphorus recovery as struvite: a comprehensive review, J. Environ.
precipitates recovered by filtration. Overall, bio-struvite production Chem. Eng. 9 (2021), 105579, https://doi.org/10.1016/j.jece.2021.105579.
presented important advantages, in comparison with chemical struvite, [4] V.G. Le, D.V.N. Vo, N.H. Nguyen, Y.J. Shih, C.T. Vu, C.H. Liao, Y.H. Huang,
and the process should be further developed for implementation at pilot Struvite recovery from swine wastewater using fluidized-bed homogeneous
granulation process, J. Environ. Chem. Eng. 9 (2021), 105019, https://doi.org/
and full scale where the organic matter and suspended solids in waste­
10.1016/j.jece.2020.105019.
water, in addition to its complicated ion components, exerted influence [5] B.E. Rittmann, B. Mayer, P. Westerhoff, M. Edwards, Capturing the lost
on the biomineralisation process. Further work will need to focus on the phosphorus, Chemosphere 84 (2011) 846–853, https://doi.org/10.1016/j.
design of reactors/processes, operational conditions to ensure prolifer­ chemosphere.2011.02.001.
[6] L. Egle, H. Rechberger, M. Zessner, Overview and description of technologies for
ation of the selected microorganisms, and even out-compete other mi­ recovering phosphorus from municipal wastewater, Resour. Conserv. Recycl. 105
croorganisms in mixed cultures, for eventual enhanced P recovery by (2015) 325–346, https://doi.org/10.1016/j.resconrec.2015.09.016.
bio-struvite from waste streams. [7] C.M. Mehta, W.O. Khunjar, V. Nguyen, S. Tait, D.J. Batstone, Technologies to
recover nutrients from waste streams: a critical review, Crit. Rev. Environ. Sci.
Technol. 45 (2015) 385–427, https://doi.org/10.1080/10643389.2013.866621.
4. Conclusions [8] A.F. Santos, P.V. Almeida, P. Alvarenga, L.M. Gando-Ferreira, M.J. Quina, From
wastewater to fertilizer products: alternative paths to mitigate phosphorus demand
in European countries, Chemosphere 284 (2021), 131258, https://doi.org/
This study demonstrated the capability of selected microorganisms 10.1016/j.chemosphere.2021.131258.
to produce visible bio-struvite crystals in wastewater containing PO4- [9] M.C. Chrispim, M. Scholz, M.A. Nolasco, Phosphorus recovery from municipal
P ≥ 19.7 mg/L, much less than the PO4-P (62.4 mg/L, with pH adjust­ wastewater treatment: critical review of challenges and opportunities for
developing countries, J. Environ. Manag. 248 (2019), 109268, https://doi.org/
ment) for struvite chemical precipitation. B. antiquum distinguished it­ 10.1016/j.jenvman.2019.109268.
self from the other four microbial strains by relatively stable P removal [10] D. Arias, L. Cisternas, M. Rivas, Biomineralization mediated by ureolytic bacteria
efficiency (68–97%) independent on initial nutrient concentrations and applied to water treatment: a review, Crystals 7 (2017) 345, https://doi.org/
10.3390/cryst7110345.
SIstruvite, and reduced PO4-P from 5.4 to less than 1 mg/L. Within the
[11] Z. Wang, J. Su, A. Ali, R. Zhang, W. Yang, L. Xu, T. Zhao, Microbially induced
investigated range, a significant initial PO4-P concentration for P calcium precipitation based simultaneous removal of fluoride, nitrate, and calcium
removal efficiency was 19.7 mg/L, and was 33.9 mg/L for bio-struvite by Pseudomonas sp. WZ39: mechanisms and nucleation pathways, J. Hazard.
production and crystal size. Large bio-struvite crystal groups with Mater. 416 (2021), 125914, https://doi.org/10.1016/j.jhazmat.2021.125914.
[12] R.B. Frankel, D.A. Bazylinski, Biologically induced mineralization by bacteria, Rev.
Dv50 above 100 µm had most potential to occur at initial NH4-N of Mineral. Geochem. (2003) 95–114, https://doi.org/10.2113/0540095.
332 mg/L and PO4-P above 33.9 mg/L. Compared with chemical stru­ [13] R.R. Sadowski, J. Prywer, A. Torzewska, Morphology of struvite crystals as an
vite, bio-struvite recovered from wastewater with much lower PO4-P evidence of bacteria mediated growth, Cryst. Res. Technol. 49 (2014) 478–489,
https://doi.org/10.1002/crat.201400080.
was of higher surface porosity, and similar high purity and extremely [14] J.E. Grover, A.F. Rope, E.S. Kaneshiro, The occurrence of biogenic calcian struvite,
low heavy metal content. It met the inorganic fertilisers regulation (Mg,Ca)NH4PO4.6H2O, as intracellular crystals in Paramecium, J. Eukaryot.
proposed by Fertilisers Working Group, which enable its potential use as Microbiol. 44 (1997) 366–373.
[15] A. Smirnov, N. Suzina, N. Chudinova, T. Kulakovskaya, I. Kulaev, Formation of
a good alternative of P and N fertiliser. Biomineralisation benefited bio- insoluble magnesium phosphates during growth of the archaea Halorubrum
struvite production by low initial PO4-P requirement, large crystal size distributum and Halobacterium salinarium and the bacterium Brevibacterium
and no pH adjustment, whereby PO4-P recovery through bio-struvite antiquum, FEMS Microbiol. Ecol. 52 (2005) 129–137, https://doi.org/10.1016/j.
femsec.2004.10.012.
presents interesting benefits and opportunities to be applied to WWTPs. [16] S. Weiner, P.M. Dove, An overview of biomineralization processes and the problem
of the vital effect, in: Rev. Mineral. Geochemistry, 2003, pp. 1–29, https://doi.org/
CRediT authorship contribution statement 10.2113/0540001.
[17] A. Soares, M. Veesam, F. Simoes, E. Wood, S.A. Parsons, T. Stephenson, Bio-
Struvite: a new route to recover phosphorus from wastewater, Clean. Soil Air Water
Yirong Leng: Conceptualization, Investigation, Writing – original 42 (2014) 994–997, https://doi.org/10.1002/clen.201300287.
draft. Ana Soares: Project administration, Supervision, Conceptualiza­ [18] F. Simoes, P. Vale, T. Stephenson, A. Soares, Understanding the growth of the bio-
tion, Writing – review & editing. struvite production Brevibacterium antiquum in sludge liquors, Environ. Technol.
39 (2017) 1–10, https://doi.org/10.1080/09593330.2017.1411399.
[19] M.T. González-Muñoz, C. De Linares, F. Martínez-Ruiz, F. Morcillo, D. Martín-
Declaration of Competing Interest Ramos, J.M. Arias, Ca-Mg kutnahorite and struvite production by Idiomarina
strains at modern seawater salinities, Chemosphere 72 (2008) 465–472, https://
doi.org/10.1016/j.chemosphere.2008.02.010.
The authors declare that they have no known competing financial [20] Y. Leng, R. Colston, A. Soares, Understanding the biochemical characteristics of
interests or personal relationships that could have appeared to influence struvite bio-mineralising microorganisms and their future in nutrient recovery,
the work reported in this paper. Chemosphere 247 (2020), 125799, https://doi.org/10.1016/j.
chemosphere.2019.125799.
[21] F. Simoes, R. Colston, C. Rosa-Fernandes, P. Vale, T. Stephenson, A. Soares,
Predicting the potential of sludge dewatering liquors to recover nutrients as

9
Y. Leng and A. Soares Journal of Environmental Chemical Engineering 11 (2023) 109208

struvite biominerals, Environ. Sci. Ecotechnol. 3 (2020), 100052, https://doi.org/ [32] S. Mann, Biomineralization: principles and concepts in bioinorganic materials
10.1016/j.ese.2020.100052. chemistry, Oxford University Press, 2001.
[22] A. Nocker, R. Cheswick, P.M. Dutheil de la Rochere, M. Denis, T. Léziart, P. Jarvis, [33] Y. Leng, A. Soares, Understanding the mechanisms of biological struvite
When are bacteria dead? A step towards interpreting flow cytometry profiles after biomineralisation, Chemosphere 281 (2021), 130986, https://doi.org/10.1016/j.
chlorine disinfection and membrane integrity staining, Environ. Technol. 38 chemosphere.2021.130986.
(2017) 891–900, https://doi.org/10.1080/09593330.2016.1262463. [34] X. Hao, C. Wang, M.C.M. Van Loosdrecht, Y. Hu, Looking beyond struvite for P-
[23] J.P. Gustafsson, Visual MINTEQ, (2013). 〈https://vminteq.lwr.kth.se/visual-mint recovery, Environ. Sci. Technol. 47 (2013) 4965–4966, https://doi.org/10.1021/
eq-ver-3–1/〉. es401140s.
[24] P. Kemacheevakul, S. Chuangchote, S. Otani, T. Matsuda, Y. Shimizu, Effect of [35] ESPP, Proposed E.U., fertiliser regulation criteria for recovered struvite, (2015)
magnesium dose on amount of pharmaceuticals in struvite recovered from urine, 1–8. https://phosphorusplatform.eu/images/download/ESPP struvite FR criteria
Water Sci. Technol. 72 (2015) 1102–1110, https://doi.org/10.2166/wst.2015.313. proposal sent 24–4-15.pdf (accessed August 8, 2022).
[25] B. Tansel, G. Lunn, O. Monje, Struvite formation and decomposition characteristics [36] EUR-Lex, Regulation (EU) 2019/1009 of the European Parliament and of the
for ammonia and phosphorus recovery: a review of magnesium-ammonia- Council of 5 June 2019 laying down rules on the making available on the market of
phosphate interactions, Chemosphere 194 (2018) 504–514, https://doi.org/ EU fertilising products and amending Regulations (EC) No 1069/2009 and (EC) No
10.1016/j.chemosphere.2017.12.004. 1107/2009 and repealing Regula, Off. J. Eur. Union. (2019). 〈https://eur-lex.eur
[26] L. Wei, T. Hong, X. Li, M. Li, Q. Zhang, T. Chen, New insights into the adsorption opa.eu/eli/reg/2019/1009/oj〉 (accessed August 8, 2022).
behavior and mechanism of alginic acid onto struvite crystals, Chem. Eng. J. 358 [37] H. Herzel, O. Krüger, L. Hermann, C. Adam, Sewage sludge ash - a promising
(2019) 1074–1082, https://doi.org/10.1016/j.cej.2018.10.110. secondary phosphorus source for fertilizer production, Sci. Total Environ. 542
[27] S. Shaddel, S. Ucar, J.P. Andreassen, S.W. Sterhus, Engineering of struvite crystals (2016) 1136–1143, https://doi.org/10.1016/j.scitotenv.2015.08.059.
by regulating supersaturation - correlation with phosphorus recovery, crystal [38] M.M. Thant Zin, D.J. Kim, Struvite production from food processing wastewater
morphology and process efficiency, J. Environ. Chem. Eng. 7 (2019), 102918, and incinerated sewage sludge ash as an alternative N and P source: optimization
https://doi.org/10.1016/j.jece.2019.102918. of multiple resources recovery by response surface methodology, Process Saf.
[28] A. Rabinovich, J.R. Heckman, B. Lew, A.A. Rouff, Magnesium supplementation for Environ. Prot. 126 (2019) 242–249, https://doi.org/10.1016/j.psep.2019.04.018.
improved struvite recovery from dairy lagoon wastewater, J. Environ. Chem. Eng. [39] S. Verma, A. Kuila, Bioremediation of heavy metals by microbial process, Environ.
9 (2021), 105628, https://doi.org/10.1016/j.jece.2021.105628. Technol. Innov. 14 (2019), 100369, https://doi.org/10.1016/j.eti.2019.100369.
[29] L. Qin, C.V. Putnis, L. Wang, Facet-specific dissolution–precipitation at [40] D.M. Voica, L. Bartha, H.L. Banciu, A. Oren, Heavy metal resistance in halophilic
struvite–water interfaces, Cryst. Growth Des. 21 (2021) 4111–4120, https://doi. Bacteria and Archaea, FEMS Microbiol. Lett. 363 (2016) 1–9, https://doi.org/
org/10.1021/acs.cgd.1c00400. 10.1093/femsle/fnw146.
[30] UWWTD, Phosphorous Standard for Wastewater Treatment Works, (2013). [41] M.M. Rahman, M.A.M. Salleh, U. Rashid, A. Ahsan, M.M. Hossain, C.S. Ra,
〈http://www.epa.ie/licences/lic_eDMS/090151b280328428.pdf〉 (accessed August Production of slow release crystal fertilizer from wastewaters through struvite
8, 2022). crystallization - a review, Arab. J. Chem. 7 (2014) 139–155, https://doi.org/
[31] Y. Leng, A. Soares, The mechanisms of struvite biomineralization in municipal 10.1016/j.arabjc.2013.10.007.
wastewater, Sci. Total Environ. 799 (2021), 149261, https://doi.org/10.1016/j.
scitotenv.2021.149261.

10