Journal of Cleaner Production 323 (2021) 129195

Contents lists available at ScienceDirect

Journal of Cleaner Production

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

Biowaste-based biodegradable flocculants for clean and sustainable tailings

management in industrial mining and mineral processing
Yeling Zhu a, 1, Michael Chae a, Jie Wang a, Birendra Adhikari a, Paolo Mussone b, David
C. Bressler a, *
a
Department of Agricultural, Food & Nutritional Science, University of Alberta, 116 St. and 85 Ave., NW, Edmonton, Alberta, T6G 2P5, Canada
b
Applied BioNanotechnology Industrial Research Chair, Industry Solutions, Northern Alberta Institute of Technology, 10210 Princess Elizabeth Ave., NW, Edmonton,
Alberta, T5G 0Y2, Canada

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

Handling editor: Cecilia Maria Villas Bôas de Massive fluid tailings discharged from industrial mining and mineral processing operations pose numerous
Almeida economic and environmental challenges. To remove water and condense tailings into sediment, we developed a
novel bioflocculant from specified risk materials, a particular type of waste biomass from the animal rendering
Keywords: industry. By properly controlling key synthesis conditions, the bioflocculant successfully met industrial re­
Wastewater treatment
quirements for fluid tailings treatment, including enhanced short-term tailings settling and improved ultimate
Clean mining
water removal extent. Settling tests using real industrial fluid tailings showed that the synthesized bioflocculant
Clean mineral processing
Biodegradable enabled an initial settling rate 4 magnitudes higher than the control case and achieved 93.2% the dewatering
Specified risk materials efficiency of a commercial flocculant (hydrolyzed polyacrylamide; HPAM). In contrast to HPAM, our bio­
Land reclamation flocculant can achieve tailings settling without gypsum, a common processing aid that can hinder downstream
industrial operations. More importantly, growth tests using two common microorganisms (Saccharomyces cer­
evisiae and Escherichia coli) suggested that the bioflocculant can be easily degraded, indicating reduced negative
environmental impacts after its use. Taken together, this study demonstrates the great potential of implementing
a biodegradable bioflocculant from proteinaceous wastes for clean and sustainable management of mining and
mineral processing fluid tailings.

1. Introduction removal/recycling for clean and sustainable management of these

intractable tailings (Asr et al., 2019).
Industrial mining and mineral processing operations are generating Aggregating suspended fine solids through flocculation has been
huge quantities of fluid tailings (Northey et al., 2016; Yan et al., 2019). widely recognized as an effective strategy to densify fluid tailings
Due to the lack of efficient and economically-viable dewatering tech­ (Gosselin et al., 2010; Qi et al., 2018). A flocculant is able to bridge
niques, billions of tonnes of fluid tailings are currently impounded in multiple solid particles along its molecular chains. Such bridging leads
tailings ponds worldwide and subject to time-consuming natural settling to the generation of large solid clusters with greater settling speed
(C. Wang et al., 2014; Y. Zhu et al., 2020). This practice not only makes (Gregory and Barany, 2011), which accelerates the separation of fluid
the entrapped water inaccessible, but also brings numerous concerns tailings into clear water that can be recycled and solid sediment that can
about the economy and environment, such as the seepage of fluid tail­ be stacked in post-mining areas. Unfortunately, since most commercial
ings to surface/ground water (Palmer et al., 2010) and the leakage of flocculants, such as polyacrylamide (Alamgir et al., 2012), are derived
toxic organics, e.g., polycyclic aromatic hydrocarbons and synthetic from petrochemicals (Vajihinejad et al., 2019), implementing such
surfactants (Gosselin et al., 2010; Redman et al., 2018). Therefore, green synthetic-polymer flocculants intensifies the consumption of
and inexpensive dewatering techniques are urgently demanded to non-renewable petroleum resources. In addition, certain
simultaneously reduce the mobility of fluid tailings and achieve water synthetic-polymer flocculants contain unreacted monomers that are

* Corresponding author.
E-mail addresses: [email protected] (Y. Zhu), [email protected] (M. Chae), [email protected] (J. Wang), [email protected] (B. Adhikari), pmussone@
nait.ca (P. Mussone), [email protected] (D.C. Bressler).
1
Present Address: Department of Wood Science, University of British Columbia, 2900-2424 Main Mall, Vancouver, BC, Canada V6T 1Z4.

https://doi.org/10.1016/j.jclepro.2021.129195
Received 11 May 2021; Received in revised form 30 August 2021; Accepted 26 September 2021
Available online 27 September 2021
0959-6526/© 2021 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
Y. Zhu et al. Journal of Cleaner Production 323 (2021) 129195

toxic (e.g., carcinogenic acrylamide monomers in commercial poly­ 2.5 g glucose, 17 g pancreatic digest of casein, 3 g papaic digest of
acrylamide products (Guezennec et al., 2015)), which could bring soybean, 5 g sodium chloride, and 2.5 g dipotassium phosphate), Fish­
long-term hazards to the environment. erbrand™ agar, and Whatman filter papers (diameter: 11 cm; pore size:
A promising technological alternative toward clean and sustainable 20–25 μm) were purchased from Fisher Scientific, PA, USA. According to
industrial tailings management can be achieved by using flocculants data obtained using a CHNS/O Analyser Series II (PerkinElmer Inc., MA,
derived from biomass sources such as cellulose (Kemppainen et al., USA), the tryptone contains a nitrogen content of 13.5 ± 0.1%.
2016), proteins (Piazza et al., 2011), and alginate (Maruyama et al., MgCl2∙6H2O (99%) was purchased from EMD Millipore, MA, USA.
2020), due to their renewability, availability, and inherent biodegrad­ Calcium sulfate dihydrate (gypsum, 98%) was purchased from Acros
ability (Vajihinejad et al., 2019). In particular, to create greater eco­ Organics, NJ, USA. Flopam A3338™, an anionic hydrolyzed poly­
nomic and ecological benefits, specified risk materials (SRM), a type of acrylamide (HPAM) was purchased from SNF Canada Ltd., AB, Canada.
waste biomass produced in massive quantities by the animal rendering Industrial oil sands mature fine tailings (MFTs) were purchased from
industry and disposed of at a high cost, have been investigated as InnoTech Alberta, AB, Canada, and were stored in a 4 ◦ C fridge.
feedstock for the production of bioflocculants. The feasibility of using
SRM-derived polymers as a bioflocculant has been demonstrated in 2.2. Hydrolysis of SRM and recovery of peptides
previous proof-of-concept studies, which demonstrated that unmodified
SRM-derived peptides could flocculate both synthetic fluid tailings as Rendered SRM was treated using a CFIA approved thermal hydro­
well as industrial tailings (Y. Zhu, Gong, et al., 2021) and that cross­ lysis protocol for disinfection and peptide production; detailed proced­
linking of SRM-derived peptides with polyamidoamine epichlorohydrin ures were described elsewhere (Mekonnen et al., 2013). The protein
(PAE) led to better flocculation of synthetic fluid tailings (Y. Zhu, Chae, fragments (referred to as “peptides” hereafter) after hydrolysis were
et al., 2021). recovered following a standard protocol detailed previously (Adhikari
In this study, three knowledge gaps will be bridged to push forward et al., 2016) and stored in a 4 ◦ C fridge before use. The peptides contain a
the technology development for scale-up. Firstly, how synthesis factors nitrogen content of 15.3 ± 0.1% (C. Zhu, 2019).
impact the product’s performance will be investigated to facilitate
production of bioflocculants that meet different tailings settling goals. 2.3. Synthesis of peptides-PAE bioflocculant
For instance, a bioflocculant that enables greater ultimate dewatering
efficiency is of great value for mining operations that are heavily For the synthesis of peptides-PAE materials, the reactants’ concen­
dependent on the amount of water that can be recycled, e.g., mining oil tration (on dry basis) in suspension, PAE% (with respect to the total
sands industry; alternatively, a bioflocculant that facilitates rapid tail­ amount of peptides and PAE), reaction temperature, and reaction time
ings settling in a short time is important for facilities with limited tail­ were varied at three levels using the Taguchi method as given in Table 1.
ings processing capability. Secondly, based on the successful application The selection of these factors, as well as their variance levels, were
of peptides-PAE for the flocculation of synthetic tailings (Y. Zhu, Chae, determined based on previous studies (Adhikari et al., 2016, 2019) and
et al., 2021), this bioflocculant will be employed for tailings treatment of preliminary experimental trials. Particularly, the reactants’ concentra­
industrial streams to validate its effectiveness, as it may behave differ­ tion (Factor 1) was found to affect the gelation behaviors of product,
ently than observed in synthetic process water. Thirdly, although the which is important to the structure of the final product; PAE% (Factor 2)
SRM-derived peptides coming from cattle are biodegradable, it is dominated the crosslinking behavior between SRM peptides and PAE;
necessary to study whether the peptides-PAE bioflocculant can still reaction temperature and time (Factors 3 and 4) were critical in deter­
retain this characteristic. mining the extent of condensation between the two reactants. Synthesis
Hence, this study focused on developing an SRM-derived bio­ of the materials were then carried out systematically according to the
flocculant through adjusting key synthesis factors, investigating its experimental design that was generated based on the Taguchi-L9
performance in treating industrial fluid tailings, and studying the po­ method, as detailed in Table 2. Specifically, peptides were diluted with
tential of biodegradability. The Taguchi method (Antony and Kaye, Milli-Q water to reach certain reactants’ concentrations (peptides and
2000) was implemented to screen for the best condition tested with PAE combined; Table 2). To maintain a thermostatic reaction, a coiled
reduced experimental trials, and two characteristic functions were used condenser and an oil bath with a feed-back thermocouple were
to evaluate the settling of synthetic fluid tailings. Settling of industrial deployed. After the specified temperature was reached, a predetermined
tailings stream was carried out using the developed bioflocculant and amount of PAE resin (Table 2) was added drop by drop. Mixing using a
compared with a commercial synthetic flocculant. The biodegradability magnetic bar at a speed of 300 rpm was implemented throughout the
of the bioflocculant was examined through degradation tests, quantified experiment. After the reaction, the product was transferred to a 50 mL
by microbial growth experiments conducted using two common centrifuge tube and cooled down in a − 20 ◦ C freezer. The concentration
microorganisms. of solid materials in the final products was quantified by drying 2.00 ±
0.05 g of vortexed suspension at 80 ◦ C until the weight became stable
2. Experimental over time, based on which all the products were diluted with Milli-Q
water to an identical concentration of 1.43 wt% as working suspensions.
2.1. Materials and chemicals
2.4. Tailings settling
Rendered SRM, provided by a large rendering company in western
Canada, were handled following the Canadian Food Inspection Agency
Bench-scale cylinder settling tests were performed to evaluate the
(CFIA) recommended protocol for safe handling and disinfection
flocculation behaviors of the synthesized peptides-PAE materials
(Mekonnen et al., 2013). PAE resin with the chemical formula HO
(C13H24N3O3Cl)nH and molecular weight of the repeating unit of 305.8
g/mol, was purchased in the form of a 12.5 wt% aqueous solution from Table 1
Selected synthesis factors and their variation levels.
Solenis, DE, USA, and stored at 4 ◦ C. According to the manufacturer’s
information, the resin has a molecular weight in the range of 750, Synthesis factor Level 1 Level 2 Level 3
000–850,000 g/mol. PolyGloss-90 kaolinite was purchased from KaMin Reactants’ concentration (wt%) 8 10 12
Performance Minerals, GA, USA. CaCl2 (98%), KCl (99%), NaCl (99%), PAE% (wt%; dry basis) 11.4 22.8 34.2
Na2CO3 (99%), CH3OH (99.9%), hexane (99.9%), MP Biomedicals™ Temperature (◦ C) 60 75 90
Reaction time (h) 0.5 2 20
tryptone, BD Bacto™ tryptic soy broth (TSB; 1-liter medium contains

2
Y. Zhu et al. Journal of Cleaner Production 323 (2021) 129195

Table 2 synthesis factor was predicted based on the S/N analyses of the char­
Taguchi-L9 orthogonal array of experimental design. acteristic functions. New peptides-PAE materials were prepared
Experimental Reactants’ PAE Temperature Reaction accordingly and, to validate the Taguchi method, their performance on
group concentration proportion (◦ C) time (h) actual tailings settling was compared with the theoretical values. The
(wt%) (wt%; dry supernatant at the end of 72 h was collected and homogenized prior to
basis)
turbidity measurement using an Orion™ AQ4500 turbidimeter (Thermo
E1 8 11.4 60 0.5 Fisher Scientific, MA, USA).
E2 8 22.8 75 2 Industrial tailings, prepared by diluting the original MFTs with
E3 8 34.2 90 20
E4 10 11.4 75 20
synthetic process water at a 1:2 mass ratio to achieve a solid content of
E5 10 22.8 90 0.5 8.9 wt%, were tested for tailings settling. The select bioflocculant and its
E6 10 34.2 60 2 two reactants (SRM peptides and PAE) were tested at a dosage of 10 kg/
E7 12 11.4 90 2 tonne tailings solid equivalent, similar to the case of synthetic kaolinite
E8 12 22.8 60 20
tailings. HPAM was tested in parallel at the optimum dosage (i.e., 0.14
E9 12 34.2 75 0.5
kg/tonne) identified by Kaminsky and Gong (Kaminsky and Gong,
2018).
(Gumfekar and Soares, 2018; Vajihinejad et al., 2019). The synthetic
process water contained 28.3 ppm KCl, 545 ppm NaCl, 39.3 ppm MgCl2, 2.5. Biodegradation measurements
41.5 ppm CaCl2, 895 ppm NaHCO3, and 443 ppm Na2SO4, with pH
adjusted to 8.4 using 5 mol/L NaOH; detailed preparation of the syn­ Two microorganisms, Superstart™ Saccharomyces cerevisiae (Lalle­
thetic process water is described elsewhere by Y. Zhu, Chae, et al. mand Inc., WI, USA) and Invitrogen™ Escherichia coli (DH5α) (in the
(2021). Synthetic process water carrying 4 wt% kaolinite clays was used form of glycerol stock, Thermo Fisher Scientific, CA, USA), were selected
as model fluid tailings. Tailings settling was carried out by mixing 243 as the yeast and bacterial strain, respectively, for biodegradation tests. A
mL of synthetic kaolinite tailings with 7 mL flocculant working sus­ yeast glycerol stock was prepared according to J. Wang et al. (2020).
pensions (carrying flocculant at a dosage of 10 kg/tonne tailings solid in Both glycerol stocks were stored at − 80 ◦ C before use.
equivalent), after which the normalized mudline height, ̂ h (equals to the All test agar plates contained the following chemicals (final amount
actual mudline height divided by that of the whole tailings), was in 1 L of medium): 8 g glucose, 15 g agar, 6.78 g Na2HPO4, 3 g KH2PO4,
recorded at various timepoints over a period of 72 h. To obtain char­ 0.5 g NaCl, 0.24 g MgSO4, 11 mg CaCl2, 2.4 mg FeCl3⋅6H2O, 100 μg
acteristic functions, the normalized mudline height was fitted in CuSO4⋅5H2O, 50 μg MnSO4⋅5H2O, 50 μg ZnSO4⋅7H2O, 10 μg CoCl2, 7 μg
response to time using OriginPro 2017 (non-linear fitting) with a hy­ H3BO3, and 10 μg NaMoO4. The above medium was autoclaved and
perbolic equation (Equation (1)) as the fitting equation (Y. Zhu et al., mixed with various nitrogen-bearing chemicals (final concentration): (a)
2018). none (negative control); (b) 2 g/L tryptone (positive control); (c) 1.76 g/
( ) L SRM-derived peptides (contained the same nitrogen amount as (b));
αt
̂
h= + 1 × 100% (1) and (d) 2.28 g/L peptides-PAE materials (contained the same peptides
β+t
amount as (b)). To prepare agar plates, the mixture was poured into
Fisherbrand™ petri dishes (100 mm × 15 mm in size, Fisher Scientific,
where, α (%) and β (min) are two fitting parameters. To identify a bio­
CA, USA) in a biosafety cabinet.
flocculant with enhanced tailings settling performance, two character­
For cultivation, microorganisms were first streaked onto TSB agar
istic functions were used to evaluate the settling behaviors of synthetic
plates (containing 30 g/L TSB and 15 g/L agar) and incubated for 48 h,
kaolinite tailings, i.e., v0 (normalized initial settling rate) and hULT (ul­
at 30 ◦ C for S. cerevisiae or 37 ◦ C for E. coli. A single colony isolate was
timate sediment height). These two characteristic functions are associ­
then selected and transferred to a 250 mL shaking flask containing 50
ated with the short-term tailings settling and the ultimate water removal
mL TSB medium (30 g/L TSB) and incubated in shakers (New Bruns­
extent, respectively, which are two important considerations for in­
wick™ Innova 44, Eppendorf Canada, ON, Canada) for 36 h at 200 rpm,
dustrial operators. These two characteristic functions can be calculated
30 ◦ C for S. cerevisiae (37 ◦ C for E. coli). The broth was then subject to
following Equations (2) and (3) (Y. Zhu et al., 2018).
serial dilution using sterile phosphate buffer (pre-autoclaved and con­
( )
dh α taining 6.78 g/L Na2HPO4, 3 g/L KH2PO4, and 0.5 g/L NaCl; pH:
v0 = − =− (2)
dt t=0 β 6.9–7.0) to prepare inoculum to make the final colony count falling
between 30 and 200 per plate. After this, 100 μL of the diluted culture (i.
hULT = lim h = 1 + α (3) e., 30–200 CFU) was applied to plates using a Fisherbrand™ L-shape
t→+∞
spreader (Fisher Scientific, PA, USA). The agar plates were then wrapped
In other words, v0 and hULT refer to the normalized settling rate at t = with Parafilm™ and incubated at 30 ◦ C for S. cerevisiae (37 ◦ C for E. coli).
0 and the thickness of sediment at t→ + ∞, respectively. The signal-to- Cell counts were performed for each plate every 24 h until 96 h and were
noise (S/N) was employed to indicate the extent of changes in the two carried out using 5 replicates. Both the TSB medium and phosphate
characteristic functions due to variations of the selected synthesis fac­ buffer were sterilized before use following J. Wang et al. (2017).
tors with respect to errors, as given in Equation (4) (Antony and Kaye,
2000).
2.6. Statistical analysis
S / N = − 10 log(MSD) (4)
All experiments were conducted using at least 3 replicates and results
∑n
(1/y2i ) were presented in the form of “mean value ± standard deviation”. The
where, MSD (mean square deviation) equals to i=1
n for the Larger-
∑n 2 statistical analyses of the data were carried out using Minitab (version
y
the-Better case (i.e., v0 ) and i=1
n
i
for the Smaller-the-Better case (i.e., 15), a statistical software package. Single factor Analysis of Variance
hULT ); n refers to number of observations; yi refers to the value of (ANOVA) was applied to the data populations examined to identify
characteristic function in each experimental trial. The best level for each significant differences among mean values, based on the Least Signifi­
cant Difference (LSD) criteria at a 95% confidence level.

3
Y. Zhu et al. Journal of Cleaner Production 323 (2021) 129195

3. Results and discussion in larger-the-better and smaller-the-better modes of Taguchi method,

respectively (Antony and Kaye, 2000).
3.1. Formulation optimization using synthetic kaolinite tailings Effect of reactants’ concentration. The reactants’ concentration
(on a dry basis) was the second major influential factor (ranked 2) for v0
3.1.1. Taguchi method-based optimization (Fig. 2a) and was the third influential factor (ranked 3) for hULT (Fig. 2b).
SRM feedstock was processed using a CFIA-approved thermal hy­ Both v0 and hULT appeared to be positively correlated with the reactants’
drolysis protocol to destroy any potential prions and to obtain peptides concentration in the test range (Fig. 2). Such a trend is likely associated
for bioflocculant development. Syntheses of peptides-PAE materials with the density of crosslinking nodes in the resultant bioflocculant.
were carried out based on Taguchi-L9 design (Table 2) and were then According to previous literature (Adhikari et al., 2016), the synthesis of
used for settling synthetic kaolinite tailings. According to the fitting peptides-PAE materials depends on the reactions between the azetidi­
results of using the formulated peptides-PAE materials to settle the nium group of PAE resin (molecular weight in the range of 750–850 kDa
synthetic kaolinite tailings (Tables S1 and S2) (Y. Zhu et al., 2018), the (Y. Zhu, Chae, et al., 2021)) and the functional groups (-NH2, –COOH,
results of v0 and hULT were obtained and presented in Fig. 1 as functions etc.) of the peptide molecules (molecular weights mainly in the range of
of each individual formulation. Except for the blank reference where 10–50 kDa (Y. Zhu, Chae, et al., 2021)). Thus, with a greater reactants’
tailings settling barely took place (Tables S1 and S2), a high coefficient concentration, crosslinking nodes between PAE and peptides could be
of determination (r2 > 0.992) was obtained from the fittings of all the densified, which increased the molecular weight of the resultant bio­
other groups, suggesting that the settling kinetics were well described by flocculant. According to Vajihinejad et al. (2019), a flocculant with a
the two characteristic functions. As shown in Fig. 1, all formulated large molecular weight is beneficial for aggregating the fine particles in
bioflocculants gave much greater v0 than the blank reference, which suspensions, leading to the formation of larger aggregates with a greater
recorded a v0 of (8.1 ± 1.3) × 10− 3%/min, suggesting that the settling rate, which agreed with the greater v0 at higher concentration in
peptides-PAE materials triggered a rapid phase separation at the early Fig. 2a. However, a high concentration of reactants also increased the
settling stages. Meanwhile, all hULT values observed in these experiments system viscosity. This could hinder the movement of molecular chains
were much smaller than the blank reference (94.1 ± 0.2%), indicating and reduce the crosslinking extent, which might account for the pla­
that peptides-PAE materials facilitated a more complete water recovery. teaued v0 at a reactant concentration greater than 10%. A similar
Both trends suggested the advantages of using peptides-PAE bio­ observation was reported by Adhikari et al. where a highly viscous
flocculants for settling clay-enriched tailings, which was in line with a binder was produced at a reactant concentration of 20% (Adhikari et al.,
previous study (Y. Zhu, Chae, et al., 2021). More importantly, a dramatic 2019). On the other hand, larger aggregates could trap more water
difference in characteristic function values was observed among (intra-aggregate water), which is more difficult to drain compared to the
different individual formulations (Fig. 1). For instance, E8 gave a v0 water between aggregates (inter-aggregate water) and has been recog­
(76.8 ± 11.1%/min) that was ~2.5 folds of the E1 group (29.6 ± nized as the major contributor for the swelling of sediment solids
1.2%/min); E3 indicated a much denser sediment (hULT of 11.3 ± 0.1%) (Costine et al., 2019; Grabsch et al., 2013). Such a difference could ac­
with the volume being only 61.7% of the E6 group (18.3 ± 0.1%). Such count for the lower compactness of the final sediment (a higher hULT ) at
variations demonstrated that the factors selected in this study do have a higher reactants’ concentration in Fig. 2b.
significant effect on the desired tailings settling properties. Effect of PAE proportion. The PAE proportion with respect to the
total reactants (dry basis) was the most influential factor (ranked 1) for
3.1.2. Analysis of the S/N ratios and the effect of select synthesis factors v0 (Fig. 2a). The main effect plots showed that the S/N ratio of v0
S/N ratio represents the sensitivity of the select characteristic func­ exhibited a sharp increase as the PAE proportion increased from 11.4%
tion to the noise factors, which reflects both the average and variation to 22.8%, and a sharp decrease with PAE proportion reaching 34.2%.
trend of the quality characteristics in the Taguchi method (Antony and This observation may be explained by the effect of PAE proportion, or
Kaye, 2000). The obtained S/N ratios are presented in the form of main specifically, the ratio between the reactive functional groups of PAE and
effect plots in Fig. 2, to identify the best level of each synthesis factor for peptides, on the reaction types and hence the product. According to
bioflocculant development (Table 2). The variation of S/N ratio in Adhikari et al. (2016), the reactive functional group (azetidinium group)
response to each synthesis factor was ranked in the order of relative in PAE was present at 3.27 mmol/g; in contrast, the occurrence of major
importance, with a complete dataset given in Table S3 and results reactive functional groups in peptides were reported to be: carboxylic
denoted in Fig. 2. Note that a greater v0 and a smaller hULT were acid groups – 1.61 mmol/g, and primary amine – 0.53 mmol/g (C. Zhu,
preferred in this analysis; hence, S/N ratios of v0 and hULT were studied 2019). At a low PAE proportion of 11.4%, the azetidinium groups of PAE

Fig. 1. v0 (panel a) and hULT (panel b), derived from the settling of synthetic kaolinite tailings using peptides-PAE materials (Table 2). As the benchmark, tailings
settling without peptides-PAE materials gives v0 of (8.1 ± 1.3) × 10− 3%/min and hULT of 94.1 ± 0.2%. The synthesis factors used to generate each of the experimental
groups can be found in Table 2.

4
Y. Zhu et al. Journal of Cleaner Production 323 (2021) 129195

Fig. 2. Main effect plot of the key synthesis factors for v0 (panel a) and hULT (panel b). The four synthesis factors, including reactants’ concentration, PAE proportion
with respect to total reactants in weight, reaction temperature, and reaction time, are listed at the horizontal axis, each presented at three variation levels.

(0.37 mmol/g with respect to overall reactants) were more likely to time reaching 2 h can be attributed to the formation of a larger bio­
react with the relatively small peptide molecules that carried abundant flocculant. In contrast, at an even higher temperature of 90 ◦ C or a
functional groups (1.89 mmol/g of –COOH and –NH2 combined). With sufficiently long reaction time of 20 h, a substantial amount of the
increasing PAE proportion, more azetidinium groups were introduced to functional groups in peptides (-COOH, –NH2, etc.) would be consumed
the reactions, where both intramolecular and intermolecular in the reaction. This would increase the hydrophobicity of the product
homo-crosslinking of PAE was more likely to take place (Adhikari et al., and reduce molecular stretching in aqueous environments, which could
2016). For example, at a higher PAE proportion of 22.8%, the intensity lead to a lower settling rate (lower v0 in Fig. 2a). The increased hydro­
of azetidinium groups (0.75 mmol/g) in reactants was closer to the 1.65 phobicity of the bioflocculant also explained the decreasing S/N ratio of
mmol/g of –COOH and –NH2 combined of peptides, which increased the hULT with increasing reaction temperature or time (Fig. 2b), as its
probability of homo-crosslinking of PAE. In particular, the intermolec­ reduced affinity with water could facilitate drainage of water from
ular homo-crosslinking of PAE dramatically improved the molecular sediment, leading to sediment with a greater compactness (lower hULT ).
weight of the resultant peptides-PAE materials, which incurred faster Taken together, bioflocculants prepared at 75 ◦ C and 2 h facilitated
flocculation of the suspended clay particles as suggested by the sub­ crosslinking levels that benefited rapid aggregation of tailing solids,
stantially greater v0 at a PAE proportion of 22.8% in Fig. 2a. However, while the one prepared in a more severe condition (90 ◦ C and 20 h)
the intensified homo-crosslinking of PAE not only elevated the molec­ allowed the ultimate release of more water.
ular weight of the product but also led to the formation of rigid, Interaction of factors. Results of the Taguchi method provided in­
three-dimensional matrix of polymers interlinked via covalent bonds. formation about the interaction between the factors examined. A matrix
This effect could become dominating at high initial PAE proportion of interaction plot for each pair of the four key synthesis factors (6 two-
(34.2%), which could hamper the relaxation of the bioflocculant’s chain factor interactions in total) on the means of v0 and hULT are provided in
in an aqueous solution. Hence, the decrease of v0 at a PAE proportion of Figs. S1 and S2, respectively. Since neither the curves in Fig. S1 nor in S2
34.2% in Fig. 2a can be attributed to the reduced flexibility of the bio­ shows any parallel pattern, a significant degree of interaction was pre­
flocculant’s polymeric chains and restrained ability to bridge the sus­ sent between the factors within the test range.
pended clay particles in tailings. An optimum PAE proportion of ~23%
was also reported by Adhikari et al. in developing peptide-PAE adhe­ 3.1.3. Validation of the optimized formulation
sives (Adhikari et al., 2019). Based on analyzing the effects of the synthesis factors on the S/N
Similarly, the PAE proportion was the most influential factor (ranked ratio of v0 (Fig. 2a) and hULT (Fig. 2b), the corresponding variation level
1) for hULT (Fig. 2b). Specifically, the S/N ratio of hULT versus PAE of each synthesis factor was predicted for the theoretically largest v0
proportion followed a similar trend to the v0 case (Fig. 2a), i.e., it (Case 1) and smallest hULT (Case 2). To achieve the predicted largest v0 ,
increased as the PAE proportion increased from 11.4% to 22.8% and the bioflocculant needed to be formulated at a reactants concentration of
then decreased with PAE proportion up to 34.2%. Such variation may 12 wt%, a PAE proportion of 22.8 wt%, a reaction temperature of 75 ◦ C,
have resulted from the dewatering of intra- and inter-aggregate water in and a reaction time of 2 h; in contrast, the bioflocculant corresponding
sediment as well, as the bioflocculant with greater molecular weight and to the theoretically smallest hULT needed to be formulated at a reactants
more flexible polymeric chain incurred faster aggregation and settling of concentration of 8 wt%, a PAE proportion of 11.4 wt%, a reaction
suspended solid particles (greater v0 ), and it simultaneously led to a temperature of 90 ◦ C, and a reaction time of 20 h. Tailings settling ex­
greater content of intra-aggregate water that resulted in less compact periments using these select conditions were conducted, with their
sediment (greater hULT ). corresponding characteristic function values presented in Fig. 3;
Effect of reaction temperature and time. As shown in Fig. 2a, re­ meanwhile, the predicted characteristic function values of the optimized
action temperature and time were ranked fourth and third, respectively, formulations were calculated using Minitab software and are also pre­
with regard to their influence on v0 . Interestingly, their effects on v0 sented in Fig. 3. It was observed that the experimental values were in
exhibited a high similarity in trend; though the overall variation was good agreement with those predicted; the v0 presented in Fig. 3a was the
marginal, the S/N of v0 was found to increase and then decrease with highest compared to all the groups in Fig. 1a, and the hULT given in
increasing reaction temperature in the range of 60–90 ◦ C, while a similar Fig. 3b was much lower than any of the 9 tests in Fig. 1b. This validated
trend was observed with regard to reaction time in the range of 0.5–20 h. the Taguchi method-based formulation optimization.
As demonstrated by Adhikari et al., 2016, 2019, reaction temperature
and time affected the shear strength of peptides-PAE solution in a binder 3.2. Treatment of industrial fluid tailings
application, which was associated with the extent of chemical reactions
between the two multi-functional reactants. The increase of tailing Mining oil sands mature fine tailings (MFTs) were studied as the
settling rate (Fig. 2a) with temperature increasing to 75 ◦ C or reaction target stream in this study to characterize the performance of peptides-

5
Y. Zhu et al. Journal of Cleaner Production 323 (2021) 129195

substantially increased v0 to a level (~101%/min) comparable to the

industrially applied HPAM flocculant (with gypsum) (Y. Zhu, Chae,
et al., 2021) and was 4 magnitudes faster than the blank reference
(~10− 3%/min; Fig. 4b). Meanwhile, hULT was found to be 48.1 ± 0.6%
through implementing the bioflocculant, which indicates a dewatering
efficiency as high as 93.2% of the HPAM case (Fig. 4c). More impor­
tantly, unlike HPAM, the bioflocculant enabled a satisfactory tailings
settling performance without the use of gypsum coagulant, which agrees
with a previous publication (Y. Zhu, Chae, et al., 2021). Hence, the
SRM-derived bioflocculant opens up the potential of a gypsum-free
approach to settle industrial fluid tailings, which is critical for in­
dustries where gypsum use is unfavored, for instance, the mining oil
sands industry (Kasongo et al., 2000; Liu et al., 2002).
On the other hand, the use of either SRM peptides or PAE alone was
unable to achieve an industrial fluid tailing dewatering as fast (reduced
v0 in Fig. 4b) and complete (greater hULT in Fig. 4c) as that of the
peptides-PAE bioflocculant. Such observations agree with previous
literature for synthetic kaolinite tailings (Y. Zhu, Chae, et al., 2021).
Taken together, the peptides-PAE bioflocculant demonstrated improved
Fig. 3. Comparison of the experimental and predicted characteristic function
dewatering capability in treating complicated industrial fluid tailings.
values of the optimized formulations. Panel a and b refers to the v0 in Case 1
(largest v0 ) and the hULT in Case 2 (smallest hULT ), respectively. Future studies will assess the performance of the peptides-PAE bio­
flocculant at larger scale in order to validate the findings of our
bench-scale work.
PAE bioflocculant in settling industrial fluid tailings using bench-scale
cylinder settling test (Gumfekar and Soares, 2018; Vajihinejad et al.,
2019). As the developed bioflocculant is expected to be applied at the 3.3. Biodegradation
stage of thickener prior to the tailings’ discharge at the tailings pond
(Masliyah et al., 2010), MFTs pre-diluted with process water at a 1:2 Since SRM-derived peptides are hydrolyzed proteins from biological
mass ratio with a solid content of 8.9 wt% was used as the sample stream sources, one of their major advantages is the potential of biodegradation
to better represent the raw tailings at this stage (Gosselin et al., 2010). and low toxicity. This makes peptides and their derivatives of particular
The industrial mining oil sands operation has been challenged by limited interest in terms of reduced environmental impacts compared to HPAM,
freshwater intake during the wintertime and is heavily dependent on the a commercial flocculant that is not easily broken down in nature (Sun
amount of water that can be recovered (Masliyah et al., 2010). There­ et al., 2017). As peptides and peptides-PAE materials were soluble in
fore, a complete water removal (i.e., a small hULT ) is more important in water and impractical to be quantified after tailings treatment, their
this case compared to a fast tailing settling (i.e., a large v0 ). In this study, biodegradation behaviors were quantitatively investigated using a col­
the developed bioflocculant associated with the smallest hULT for syn­ ony count method. Due to the high nitrogen content in SRM-derived
thetic kaolinite tailings was evaluated in settling industrial fluid tailings. peptides (i.e., 15.3 ± 0.1% C. Zhu (2019)), the biodegradation of the
A preliminary examination of the dosage effects showed that a complete peptides-PAE bioflocculant was tested by evaluating its potential to
solid sedimentation can be achieved at a threshold dosage of 10 serve as the sole nitrogen source for the growth of two common mi­
kg/tonne (Fig. S3), which was further investigated for comparison with croorganisms, S. cerevisiae (yeast) and E. coli (bacteria), on agar plates.
the industrial practice. As given in Fig. 4a, the settling of the industrial For comparison, agar plates containing all essential nutrients except
fluid tailings was substantially improved with the use of the nitrogen were used as the negative controls, while the same plates
SRM-derived bioflocculant, which agrees with the observations for supplemented with tryptone, a commonly used organic nitrogen source
synthetic kaolinite tailings. In particular, the use of the bioflocculant in microbiology, were used as the positive controls. Furthermore, during
inoculum preparation, broth containing live cells was diluted using

Fig. 4. Settling of prediluted mature fine tailings (MFTs) from the mining oil sands industry using the select peptides-PAE bioflocculant at a dosage of 10 kg/tonne
tailings solids in equivalent (panel a) and the corresponding v0 (panel b) and hULT (panel c). The inlet of panel a shows the settling behaviors within the initial
120 min.

6
Y. Zhu et al. Journal of Cleaner Production 323 (2021) 129195

phosphate buffer (at least 105 folds), to minimize nitrogen levels in the ternary bioflocculant with the inclusion of a thermo-responsive agent
inoculum. (with a lower critical soluble temperature, or LCST) as the third reactant
The results of the colony count experiments are summarized in may be of interest (Vajihinejad et al., 2019). The bioflocculant with the
Fig. 5a (S. cerevisiae) and 5b (E. coli). In both cases, no colonies were thermo-responsive segment becomes more relaxed and flexible at a
observed throughout the 4-day cultivation on the negative control higher operation temperature (T > LCST), which results in rapid solid
plates, which contained no nitrogen source (Fig. 5a and b). Conversely, particles aggregation; as temperature ramps down during tailing settling
visible microbial colonies were observed on the positive control plates (T < LCST), the bioflocculant tends to shrink and become less soluble in
within the first 24 h (Fig. 5), suggesting that addition of tryptone to the water, which facilitates the drainage of intra-aggregate water and leads
media would facilitate cell growth. to a sediment with greater compactness. Hence, the ternary bio­
SRM-derived peptides were also analyzed as a potential nitrogen flocculant has a chance to achieve optimum v0 and hULT simultaneously
source. Compared to the positive controls, yeast growth on the peptides and will be investigated in future work.
plates (Fig. 5a) exhibited slightly lower colony counts before 48 h, but
reached similar levels by 72 h; meanwhile, no statistical difference in 4. Conclusions
bacterial colony counts was observed relative to the positive control
plates throughout the entire experiment (Fig. 5b). Taken together, these This is the first study demonstrating the development of peptide-PAE
data suggested that nitrogen present in the SRM-derived peptides is bioflocculant for real industrial fluid tailings treatment. By properly
easily accessible to microbes. Since the only source of nitrogen in these controlling key synthesis factors, bioflocculants with different perfor­
plates was from the SRM-derived peptides, these results suggest that the mance can be synthesized to meet particular industrial purposes, which
nitrogen must be easily accessible to microbes, clearly indicating that can be enhanced short-term tailings settling or improved ultimate water
the bioflocculant is biodegradable. removal. Compared to the commercial HPAM flocculant, the developed
The peptides-PAE plates (Fig. 5) were found to support the growth of SRM-derived bioflocculant not only achieved a comparably fast and
both microorganisms as well, albeit at lower levels. Interestingly, no complete densification (93.2% the dewatering efficiency of the HPAM
yeast colonies were observed on the peptides-PAE plates before 24 h, case) of real industrial fluid tailings, but also allowed tailings settling in
with the quantity of visible colonies gradually increasing until they a gypsum-free condition. In addition, the bioflocculant can be degraded
reached levels similar to the positive control plates by 72 h (Fig. 5a). On and used as a nitrogen source by two common microorganisms. Overall,
the other hand, the number of bacterial colonies plateaued by 24 h, bioflocculants derived from negative-value SRM feedstocks shows a
reaching values that were ~1/3 that of the positive controls (Fig. 5b). great potential for clean and sustainable management of industrial fluid
One possible explanation is that E. coli is unable to break down certain tailings, as well as certain biodegradability that minimizes environ­
bonds within the peptide-PAE materials, resulting in an incomplete mental impacts.
usage of the materials. Nevertheless, the data in Fig. 5 demonstrated that
the peptides-PAE materials can still serve as a nitrogen source for the test CRediT authorship contribution statement
microorganisms, indicating that the bioflocculant is biodegradable.
It should also be noted that common laboratory yeast and bacterial Yeling Zhu: Conceptualization, Methodology, Validation, Formal
strains were used in these studies; mixed populations of environmental analysis, Investigation, Data curation, Writing – original draft, Writing –
strains would likely be much more adept at breaking down the peptides- review & editing, Visualization. Michael Chae: Conceptualization,
PAE materials (Naranjo et al., 2020). Regarding actual industrial fluid Methodology, Validation, Resources, Writing – review & editing, Su­
tailings, diverse microorganisms from the environment can be easily pervision, Project administration, Funding acquisition. Jie Wang:
introduced to the fluid tailings through the open-pit mining operations, Conceptualization, Methodology, Validation, Formal analysis, Writing –
open-air tailings ponds, and the dykes impounding the tailing ponds original draft, Writing – review & editing. Birendra Adhikari: Meth­
(Foght et al., 2017). Therefore, if applied for industrial fluid tailings odology, Writing – review & editing. Paolo Mussone: Conceptualiza­
dewatering, this peptide-PAE bioflocculant has great potential to be tion, Resources, Validation, Writing – review & editing, Supervision,
biodegraded by environmental strains. Future work will investigate the Funding acquisition. David C. Bressler: Conceptualization, Methodol­
effect of the bioflocculant on microorganism activities in the actual ogy, Validation, Resources, Writing – review & editing, Supervision,
tailings, as well as evaluate the impact on the landscape reclamation of Project administration, Funding acquisition.
the tailings pond areas. In addition, although it is most desirable to
achieve both rapid settling and minimum sediment height simulta­ Declaration of competing interest
neously in tailings treatment, the conditions for largest v0 do not
necessarily match those for smallest hULT in this study. In this case, a The authors declare that they have no known competing financial

Fig. 5. Cell counts of (a) S. cerevisiae and (b) E. coli, on agar plates containing various media. Data points sharing the same letter within the same time groups
represent no statistical difference at a confidence interval of 95%.

7
Y. Zhu et al. Journal of Cleaner Production 323 (2021) 129195

interests or personal relationships that could have appeared to influence Liu, J., Zhou, Z., Xu, Z., Masliyah, J., 2002. Bitumen-clay interactions in aqueous media
studied by zeta potential distribution measurement. J. Colloid Interface Sci. 252 (2),
the work reported in this paper.
409–418. https://doi.org/10.1006/jcis.2002.8471.
Maruyama, H., Seki, H., Igi, A., 2020. Flocculation of quartz and kaolin by alginate-
Acknowledgements protamine complex. Biochem. Eng. J. 162, 107713. https://doi.org/10.1016/j.
bej.2020.107713.
Masliyah, J., Czarnecki, J., Xu, Z., 2010. Handbook on Theory and Practice of Bitumen
This work was supported by the Alberta Prion Research Institute Recovery from Athabasca Oil Sands. Vol. I. Theoretical Basis. Kingsley Knowledge
(APRI) of Alberta Innovates [No. 201700003, 2017]; Alberta Agricul­ Publishing. https://doi.org/10.7939/r3-nr43-8t34. Published in Canada.
ture and Forestry/Results Driven Agriculture Research (RDAR) [No. Mekonnen, T., Mussone, P., Stashko, N., Choi, P., Bressler, D., 2013. Recovery and
characterization of proteinacious material recovered from thermal and alkaline
2017A004R, 2017]; and Natural Sciences and Engineering Research hydrolyzed specified risk materials. Process Biochem. 48 (5–6), 885–892. https://
Council of Canada (NSERC) [No. RGPIN-2019-04184, 2019]. doi.org/10.1016/j.procbio.2013.04.015.
Naranjo, D.R., Callanan, M., Thierry, A., McAuliffe, O., 2020. Superior esterolytic activity
in environmental Lactococcus lactis strains is linked to the presence of the SGNH
Appendix A. Supplementary data hydrolase family of esterases. JDS Communications 1 (2), 25–28. https://doi.org/
10.3168/jdsc.2020-0003.
Supplementary data to this article can be found online at https://doi. Northey, S.A., Mudd, G.M., Saarivuori, E., Wessman-Jääskeläinen, H., Haque, N., 2016.
Water footprinting and mining: where are the limitations and opportunities?
org/10.1016/j.jclepro.2021.129195. J. Clean. Prod. 135, 1098–1116. https://doi.org/10.1016/j.jclepro.2016.07.024.
Palmer, M., Bernhardt, E., Schlesinger, W., Eshleman, K., Foufoula-Georgiou, E.,
References Hendryx, M., Wilcock, P., 2010. Mountaintop mining consequences. Science 327
(5962), 148–149. https://doi.org/10.1126/science.1180543.
Piazza, G., McAloon, A., Garcia, R., 2011. A renewable flocculant from a poultry
Adhikari, B., Appadu, P., Kislitsin, V., Chae, M., Choi, P., Bressler, D.C., 2016. Enhancing
slaughterhouse waste and preliminary estimate of production costs. Resour. Conserv.
the adhesive strength of a plywood adhesive developed from hydrolyzed specified
Recycl. 55 (9–10), 842–848. https://doi.org/10.1016/j.resconrec.2011.04.004.
risk materials. Polymers 8 (8). https://doi.org/10.3390/polym8080285, 285, 281-
Qi, C., Fourie, A., Chen, Q., Tang, X., Zhang, Q., Gao, R., 2018. Data-driven modelling of
213.
the flocculation process on mineral processing tailings treatment. J. Clean. Prod.
Adhikari, B., Chae, M., Zhu, C., Khan, A., Harfield, D., Choi, P., Bressler, D.C., 2019.
196, 505–516. https://doi.org/10.1016/j.jclepro.2018.06.054.
Pelletization of torrefied Wood using a proteinaceous binder developed from
Redman, A., Parkerton, T., Butler, J., Letinski, D., Frank, R., Hewitt, L., Parrott, J., 2018.
hydrolyzed specified risk materials. Processes 7 (4), 229. https://doi.org/10.3390/
Application of the target lipid model and passive samplers to characterize the
pr7040229.
toxicity of bioavailable organics in oil sands process-affected water. Environ. Sci.
Alamgir, A., Harbottle, D., Masliyah, J., Xu, Z., 2012. Al-PAM assisted filtration system
Technol. 52 (14), 8039–8049. https://doi.org/10.1021/acs.est.8b00614.
for abatement of mature fine tailings. Chem. Eng. Sci. 80, 91–99. https://doi.org/
Sun, J., Peng, H., Alharbi, H.A., Jones, P.D., Giesy, J.P., Wiseman, S.B., 2017.
10.1016/j.ces.2012.06.010.
Identification of chemicals that cause oxidative stress in oil sands process-affected
Antony, J., Kaye, M., 2000. Analysis and Interpretation of Data from Taguchi
water. Environ. Sci. Technol. 51 (15), 8773–8781. https://doi.org/10.1021/acs.
Experiments. Springer, Boston, MA, USA. https://doi.org/10.1007/978-1-4615-
est.7b01987.
5293-2_9.
Vajihinejad, V., Gumfekar, S.P., Bazoubandi, B., Rostami Najafabadi, Z., Soares, J.B.,
Asr, E.T., Kakaie, R., Ataei, M., Mohammadi, M.R.T., 2019. A review of studies on
2019. Water soluble polymer flocculants: synthesis, characterization, and
sustainable development in mining life cycle. J. Clean. Prod. 229, 213–231. https://
performance assessment. Macromol. Mater. Eng. 304 (2), 1800521–1800543.
doi.org/10.1016/j.jclepro.2019.05.029.
https://doi.org/10.1002/mame.201800526, 1800526.
Costine, A.D., Vajihinejad, V., Botha, L., Fawell, P.D., Soares, J.B., 2019. Aggregate
Wang, C., Harbottle, D., Liu, Q., Xu, Z., 2014. Current state of fine mineral tailings
structures formed by hyperbranched functionalized polyethylene (HBfPE) treatment
treatment: a critical review on theory and practice. Miner. Eng. 58, 113–131.
of oil sands tailings. Can. J. Chem. Eng. 97 (1), 99–102. https://doi.org/10.1002/
https://doi.org/10.1016/j.mineng.2014.01.018.
cjce.23287.
Wang, J., Chae, M., Bressler, D.C., Sauvageau, D., 2020. Improved bioethanol
Foght, J.M., Gieg, L.M., Siddique, T., 2017. The microbiology of oil sands tailings: past,
productivity through gas flow rate-driven self-cycling fermentation. Biotechnol.
present, future. FEMS (Fed. Eur. Microbiol. Soc.) Microbiol. Ecol. 93 (5) https://doi.
Biofuels 13 (1), 14. https://doi.org/10.1186/s13068-020-1658-6.
org/10.1093/femsec/fix034.
Wang, J., Chae, M., Sauvageau, D., Bressler, D.C., 2017. Improving ethanol productivity
Gosselin, P., Hrudey, S., Naeth, M., Plourde, A., Therrien, R., Van Der Kraak, G., Xu, Z.,
through self-cycling fermentation of yeast: a proof of concept. Biotechnol. Biofuels
2010. Environmental and Health Impacts of Canada’s Oil Sands Industry. Royal
10 (1), 1–11. https://doi.org/10.1186/s13068-017-0879-9.
Society of Canada Ottawa, Ontario, Canada.
Yan, B., Han, L., Xiao, H., Zhang, J., Huang, J., Hu, W., Zeng, H., 2019. Rapid dewatering
Grabsch, A., Fawell, P., Adkins, S., Beveridge, A., 2013. The impact of achieving a higher
and consolidation of concentrated colloidal suspensions-mature fine tailings via self-
aggregate density on polymer-bridging flocculation. Int. J. Miner. Process. 124,
healing composite hydrogel. ACS Appl. Mater. Interfaces. https://doi.org/10.1021/
83–94. https://doi.org/10.1016/j.minpro.2013.04.011.
acsami.9b05692.
Gregory, J., Barany, S., 2011. Adsorption and flocculation by polymers and polymer
Zhu, C., 2019. Thesis: Development of a Binder from Hydrolyzed Specified Risk Materials
mixtures. Adv. Colloid Interface Sci. 169 (1), 1–12. https://doi.org/10.1016/j.
for Pelletization of Torrefied Wood. University of Alberta, Edmonton, Alberta,
cis.2011.06.004.
Canada. https://doi.org/10.7939/r3-wh3h-sj55. https://era.library.ualberta.ca/ite
Guezennec, A.-G. n. l., Michel, C., Bru, K., Touzé, S., Desroche, N., Mnif, I., Motelica-
ms/962fd09c-d7f8-4552-aa8c-b58a9ea68854.
Heino, M., 2015. Transfer and degradation of polyacrylamide-based flocculants in
Zhu, Y., Chae, M., Adhikari, B., Khatri, V., Kaminsky, H., Mussone, P., Bressler, D.C.,
hydrosystems: a review. Environ. Sci. Pollut. Control Ser. 22 (9), 6390–6406.
2021a. Valorizing biowaste for wastewater treatment: dewatering sludges using
https://doi.org/10.1007/s11356-014-3556-6.
specified risk material-based flocculants for industrial sustainability. ACS Sustain.
Gumfekar, S., Soares, J., 2018. Polymer reaction engineering tools to design
Chem. Eng. 9 (5), 2037–2046. https://doi.org/10.1021/acssuschemeng.0c06841.
multifunctional polymer flocculants. Chemosphere 210, 156–165. https://doi.org/
Zhu, Y., Gong, Y., Kaminsky, H., Chae, M., Mussone, P., Bressler, D.C., 2021b. Using
10.1016/j.chemosphere.2018.06.175.
specified risk materials-based peptides for oil sands fluid. Fine Tailings Management.
Kaminsky, H., Gong, Y., 2018. Assessment of the Flocculation Potential of Specified Risk
Materials 14 (7), 1582. https://doi.org/10.3390/ma14071582.
Materials (SRM). Northern Alberta Institute of Technology-Centre for Oil Sands
Zhu, Y., Lu, Y., Liu, Q., Masliyah, J., Xu, Z., 2020. Comprehensive study on cleaner
Sustainability, Edmonton, Alberta, Canada, pp. 24–26.
production of heavy oil from Athabasca oil sands using chemical additives in
Kasongo, T., Zhou, Z., Xu, Z., Masliyah, J., 2000. Effect of clays and calcium ions on
biodiesel-assisted ambient-aqueous bitumen extraction process. J. Clean. Prod. 277
bitumen extraction from Athabasca oil sands using flotation. Can. J. Chem. Eng. 78
https://doi.org/10.1016/j.jclepro.2020.122940, 122940, 122941-122910.
(4), 674–681. https://doi.org/10.1002/cjce.5450780409.
Zhu, Y., Yan, C., Liu, Q., Masliyah, J., Xu, Z., 2018. Biodiesel-Assisted ambient aqueous
Kemppainen, K., Suopajärvi, T., Laitinen, O., Ämmälä, A., Liimatainen, H., Illikainen, M.,
bitumen extraction (BA3BE) from athabasca oil sands. Energy Fuel. 32 (6),
2016. Flocculation of fine hematite and quartz suspensions with anionic cellulose
6565–6576. https://doi.org/10.1021/acs.energyfuels.8b00688.
nanofibers. Chem. Eng. Sci. 148, 256–266. https://doi.org/10.1016/j.
ces.2016.04.014.