Environmental Science and Pollution Research https://doi.org/10.1007/s11356-018-2570-5

ENVIRONMENTAL AND SUSTAINABLE CHEMICAL ENGINEERING

Up-scaling of tannin-based coagulants for wastewater treatment: performance in a water treatment plant Kinga Grenda 1,2 & Julien Arnold 2 & José A. F. Gamelas 1 & Maria G. Rasteiro 1 Received: 29 March 2018 / Accepted: 14 June 2018 # Springer-Verlag GmbH Germany, part of Springer Nature 2018

Abstract Tannin extracts from the bark of Acacia mearnsii and wood of Schinopsis balansae, commonly known as Quebracho, were employed. These were modified at laboratory sale via the Mannich aminomethylation with formaldehyde and dimethylamine hydrochloride. Some reaction conditions were varied, namely the formaldehyde dosage and reaction time, while keeping the Mannich solution activation time constant, and their influence on the shear viscosity of the created bio-coagulants was evaluated. The effect of the final pH of the products on their shear viscosity was also analyzed. Up-scaling of the Mannich reaction for tannin from South Africa was performed and the procedure developed at 1-L scale was reproducible in upscaled conditions. One example of a modified South Africa tannin and the modified Quebracho tannin was subsequently selected for the treatment of an industrial wastewater and tested for color and turbidity reduction in jar tests. The effluent treatment was carried out in a single and dual system with cationic synthetic flocculation agents of different charge degree. Good turbidity and decoloration results (93 and 89% reduction, respectively) were obtained with the simultaneous introduction of a cationic, 40% charged polyacrylamide, with minimal dosage (5 ppm) of the latter additive. The tannin-based coagulant from Acacia mearnsii was successfully applied in dual system with cationic polyacrylamide flocculant for industrial wastewater treatment at pilot plant scale. It was shown to satisfactorily treat the water and generate less sludge. Keywords Bio-coagulant . Coagulation/flocculation . Decoloration . Wastewater treatment . Acacia mearnsii tannin extract . Quebracho tannin extract

Introduction Pollution is intensified by an expanding middle class and growing population. These drive industrial demand. Water scarcity now becomes a major issue in many cities, globally. This is exacerbated by the move to urbanization. In some arid regions, one could say that aqueous resources have been mismanaged as more water has been withdrawn from underground aquifers than can be replaced (e.g., Iran). In some coastal cities, such as Cape Town, which counts the days until Responsible editor: Bingcai Pan * Maria G. Rasteiro [email protected] 1

Department of Chemical Engineering, CIEPQPF – Chemical Process Engineering and Forest Products Research Centre, University of Coimbra, Rua Sílvio Lima, 3030-790 Coimbra, Portugal

2

AQUA+TECH Specialities, Chemin du Chalet-du-Bac 4, CH-1237 Avully, Geneva, Switzerland

it runs out of water, one could say a lack of public planning is a cause for drinking water shortages (Martínez-Huitle and Ferro 2006). Construction takes time and, in general, it costs about EUR 1000 per person equivalent to build a treatment facility. For many countries, this is an unbearable cost. The UN, in the annual water report, states that it is impossible to meet the water needs of the world’s population without recycling wastewater into potable streams, as is presently done in California (WHO and Unicef 2000). Polluted water has to go through several treatment processes in order to sufficiently purify it. In general, effluent treatment techniques are implemented based on the concentration, types of contaminants, and volume of treated wastewater (Beltran-Heredia et al. 2001), as well as the type of industry or even final purpose (Pintor et al. 2016). Various inorganic or organic (synthetic) compounds have been applied as coagulation/flocculation agents in conventional wastewater treatment (WWT) (Ebeling et al. 2003; Tzoupanos and Zouboulis 2008). However, the wastes generated from the treatment may cause several problems, due to their toxicity,

Environ Sci Pollut Res

low (or no) biodegradability, and add extra costs to sludge disposal requirements (Khandare and Govindwar 2015). Considering those issues, there is a strong need for the development and replacement of synthetic by natural-based WWT alternatives (dos Santos et al. 2018; Bouaouine et al. 2018). Among the traditional treatment strategies, a promising perspective relies in the development of plant-based natural coagulants/flocculants, which, by being mostly biodegradable, reduce the sludge generated. The transport of, generally wet, sludge is the largest cost and contributor to global warming in water treatment (Rebitzer et al. 2003). Plant-based products have in fact been used for many centuries as water purification agents (Morawetz 1993). Moreover, tannins are readily available and widely distributed in wood, bark, leaves, fruits, or even seeds (Haslam 1988). These molecules have a complex structure and chemistry that strongly depend on their origin, location, environment conditions, or even plant age, with a wide range of molecular structures and molar masses (1000–20,000 Da) (Haslam 2005). Tannins have, in the past, been used in water boilers to prevent scale formation as well as viscosity reducing agents in drilling muds (Khanbabaee and Ree 2001). Antibacterial (Akiyama et al. 2001) or antidiarrheal properties (Ashok and Upadhyaya 2012), among other medical uses, have been claimed. Their phenolic structure with an anionic character also allows application as natural coagulants for effluent treatment. However, their performance as coagulation agents strongly depends on their structure, or origin, and can be tuned based on chemical modification (SánchezMartín et al. 2010), as well as on the effluent characteristics such as the following: pH, salinity, biomass concentration, or organic matter (Roselet et al. 2017). Some studies were reported on using tannins or modified tannin products as coagulation aids in decoloration (BeltránHeredia et al. 2011a; Grenda et al. 2018), for particle removal (Özacar and Şengil 2003), or even in the treatment of surface river water or surfactant-polluted effluents (Beltrán-Heredia et al. 2010), as well as for municipal wastewater treatment (Hameed et al. 2018). Moreover, it was reported that pH may negatively influence the performance of tannin-based coagulants in wastewater treatment, while temperature was not as an important factor in obtained results (Beltrán-Heredia et al. 2011b; Hesse et al. 2017). It is also important to note that the natural materials require chemical modification to be suitable in solid-liquid separation. The production of cationic tannin-based coagulants is usually performed by a Mannich aminomethylation reaction. This reaction introduces positively charged groups (namely protonated tertiary nitrogen groups) to the complex tannin matrix (Tramontini et al. 2015); subsequent polymerization is also claimed to occur, thus, increasing the molecular weight

(Graham et al. 2008). The dual nature of the obtained products, cationic (tertiary ammonium) and anionic (ionized phenols), is beneficial in a coagulation system (Beltrán-Heredia et al. 2010). Typically, the positively charged modified tanninbased products are able to destabilize anionic impurities in wastewaters. The process of coagulation is followed by settling, enabling separation, and removal of contaminants (Beltrán-Heredia et al. 2010). This study’s main focus was on the production of naturalbased cationic coagulation agents for wastewater treatment using different origins and sources of tannin. The modifications were performed on bark tannin extracts of A. mearnsii either from South Africa (Mimosa ME-SA) or provided from New Zealand (Mimosa ME-NZ), where the concentrate is classified as condensed tannin and consists of polymerized robinetinidol, fisetinidol, and gallocatechin units (Covington et al. 2005). The Mannich aminomethylation was also performed on a wood tannin extract from Schinopsis balansae tree, typically known as Quebracho tannin (a class of condensed tannin), from Argentina. Note that Quebracho tannin is predominantly composed of fisetinidin while Mimosa tannin shows relevant amounts of robinetinidin units which exist only in minor amounts in Quebracho tannins (Pasch et al. 2001). As a consequence of the source diversity, some adaptations in the developed production procedure according to the raw material (type and source of tannin) had to be applied. The optimized process for the modification of South African tannin was then up-scaled. The performance of obtained ecocoagulants and their biodegradability were evaluated in a wastewater treatment plant for the treatment of an industrial effluent. The results were compared with commercial blends of inorganic and polymeric coagulants typically used for this application.

Materials and methods Materials for bio-coagulant production A. mearnsii bark tannin extracts from two different sources, one provided by Fontis, New Zealand, and another from South Africa (provided by Extract Dongen B.V., Netherlands), were used without any further treatment. A batch of Schinopsis balansae wood tannin extract, commonly known as Quebracho (under the commercial name Tupafin ATO), was provided by INDUNOR S.A. Argentina, and was also used without any further treatment. The dimethylamine hydrochloride (DMA∙HCl, powder, 99%) was purchased from Sigma-Aldrich (Steinheim, Germany). Formaldehyde (FA, aqueous solution of 37 wt%) was supplied by Acros Organics (Geel, Belgium). All of the other chemicals were purchased from Sigma-Aldrich (Steinheim, Germany).

Environ Sci Pollut Res

Synthesis of the tannin-based coagulants Tannin-based coagulants were synthesized via Mannich aminomethylation using as amine and aldehyde sources DMA∙HCl and FA, respectively. Figure 1 shows the modification strategy developed for the Mimosa ME-SA tannin in a two-step reaction. Briefly, a mixture of DMA∙HCl/FA/water with a mass ratio of 1:1:1 (Mannich solution, MS) was prepared (left resting overnight) and added to a tannin aqueous solution under agitation. To force the reaction, an extra amount of FA was added suddenly to the reaction mixture. After cooking at desired temperature and period of time, a viscous cationic tannin-based product was obtained. The modification of the Mimosa ME-SA tannin was based on the procedure previously developed in our team and described in the literature (Grenda et al. 2018). A mass of 50 g of the tannin powder was dissolved in 117.5 g of water for 1 h at 30 °C. Following this, 40.5 g of MS was added and the reaction mixture was heated to 85 °C. At this temperature, a shot of 37 wt% aqueous FA was injected. Two different amounts of FA were tested: 6 g (FA/tannin ratio of 0.04) and 3.5 g (FA/ tannin ratio of 0.02) of 37 wt% aqueous FA. Four different reaction times were evaluated (45–120 min). The reaction mixture was then diluted with 100 g of deionized water (conductivity below 0.3 μS/cm), and subsequently acidified with hydrochloric acid, and thus the effect of pH reduction to different final values was evaluated. At the end of the procedure, the shear viscosity of the obtained tannin-based coagulant was determined with a Brookfield Viscometer (LVDV-II+ Pro, Stoughton, USA) at 30 rpm and a temperature of 24.5 ± 0.5 °C using the appropriate spindle (1 to 6 depending on the product viscosity). The modification procedure developed for Mimosa MESA was successfully applied for Mimosa ME-NZ bio-coagulant production, without significant adjustments. The FA/ tannin ratios of 0.02 and 0.04 mentioned above, as well as three different reaction times (60, 90, and 120 min), were tested. Similarly, the progress of the Mannich reaction was monitored based on measurements of viscosity. Applying the cationization procedure optimized for A. mearnsii bark tannin extracts directly to the S. balansae wood Fig. 1 Schematic of the modification of the Mimosa ME tannin

tannin did not lead to obtain the desired product, and insoluble tannin powder was still present at the bottom of the reactor. It was, therefore, necessary to optimize the reaction conditions for the latter tannin. In this case, a complete dissolution of tannin (50 g) in water (117.5 g) for 1-h agitation at 60 °C was achieved. Subsequently, by the MS (40.5 g) and FA addition, followed by dilution and acidification, viscous tanninbased products were obtained. The FA/tannin ratios of 0.02 and 0.04, as well as the reaction times of 60, 100, and 135 min, were evaluated in this case.

Mannich reaction: up scaling The synthesis was up-scaled to the 5- and 75-L sizes. The work on the 75-L scale is described elsewhere (Grenda et al. 2018). In a 2-L beaker, 150 g of DMA∙HCl was dissolved in 150 g of water (pH = 6). After full dissolution, 150 g of 37 wt% aqueous FA solution was added. The MS mixture was stirred with a magnetic stirrer for 20 min at 200 rpm, and then left resting overnight. In a 5-L pressurized vessel manufactured by ACME, 316SS, Mannich modification of the Mimosa ME-SA tannin was carried out. In this reactor, temperature was controlled using a steam-water mixture in an external jacket. A PID controller was used to regulate the steam input to the cooling water loop. Temperature was measured with a PT-100 thermocouple in the reactor and in the heating loop. Agitation was achieved with a two-stage impeller, one four-pitched blade impeller at the bottom, and a marine propeller at 50% of the reactor height. The speed was set to the minimum of 100 rpm. For the reaction, 500 g of the tannin powder was dissolved in 1175 g of water under stirring for 1 h at 30–34 °C. Then, 405 g of MS was added, and the mixture was stirred for 20 min before being heated to 85 °C ± 2 °C. Subsequently, 35 g of 37 wt% FA was added. The reaction was then kept and monitored for 1 h 45 min; after that period, the obtained product was diluted with 1000 g of water. Finally, acidification of the solution with hydrochloric acid to reduce the pH to 1.6 was performed. The heating profile of the production of the tannin-based coagulant is presented in Fig. 2. Reacon: 1h at 85°C Diluon and acidificaon

Dissoluon: 1h at 30°C

H2O

+

Tannin powder

MS

Reacon mixture

FA

Tannin based coagulant

Environ Sci Pollut Res

microbeads (about 10 μm in diameter) of different colors made from polyurethane and poly(methyl methacrylate) in a suspension process. Wastewater from this production was left outside in the open waste reservoirs and, therefore, was diluted slightly by rain or concentrated by water evaporation over the year, as is legal in Switzerland. The industrial colored effluent was collected, at a certain date, in containers and stored, and was then used for the jar test as well as in the wastewater treatment plant. Its average characteristics are given in Table 2.

Coagulation-flocculation experiment: jar test

Fig. 2 Heating profile of the reaction of Mimosa ME-SA tannin with MS and FA, in a 5-L reactor. The modification steps were as follows: (1) dissolution of the tannin powder in water, (2) addition of the MS, (3) heating, (4) activation of the MS, (5) FA addition, (6) cooking, (7) dilution, and (8) acidification

Characterization of the raw material and modified tannins To characterize the tannins prior to, and after, modification, the Fourier transform infrared (FTIR) spectra were recorded with a Bruker Tensor 27 spectrometer (Karlsruhe, Germany) over the range of 400 to 4000 cm−1 using 128 scans and a resolution of 4 cm−1. The samples were analyzed in KBr pellets. Previously, gels of the modified tannins were oven-dried at 50 °C. The elemental analyses of C, H, and N were performed for all tannin raw materials using an EA 1108 CHNSO element analyzer from Fisons Instruments (Milano, Italy). The chemical 2,5-bis(5-tert-butyl-benzoxazol-2-yl)thiophene was used as a standard. The elemental characterization of the unmodified tannins is provided in Table 1.

Raw wastewater The industrial wastewater was collected from a company, Achemis AG, located in Geneva, Switzerland, that produces essentially a cosmetic raw material and, as well, Table 1

The jar test was applied to evaluate the coagulationflocculation performance of the modified tannins (Mimosa ME-SA with a shear viscosity of 41 cP and Quebracho with a viscosity of 40 cP), as well as to develop the most effective procedure for the effluent treatment. A modified tannin solution at a 1% (w/w) concentration was prepared by dissolution in water under stirring at 500 rpm for 60 min. For the flocculation-coagulation experiment, 200 mL of the raw wastewater was placed in a beaker, and the pH was adjusted to the target value by adding 10 wt% phosphoric acid. A suitable amount of tanninbased coagulant was added dropwise, followed by the addition of cationic (BHMW with 80% charge or FHMW with 40% charge) polyacrylamides (PAM) at a 0.1 wt% concentration. The polyacrylamides used, distributed under the brand name SnowFlake, were kindly provided by aquaTECH (Geneva, Switzerland). The effluent was mixed with all added components for 30 s, and then, flocs were allowed to settle. Supernatant samples of approximately 2 mL were collected from the surface, at about 75% height from the bottom of the beaker, to determine the color removal and turbidity reduction over time (30 min, 1 h, and 24 h). However, only the samples obtained from the treatment procedure after 1 h of settling were considered as the ones with significant meaning and valuable for application in WWT. The decoloration was calculated based on the measurement of the absorbance of the supernatant at the optimized/universal wavelength of 350 nm, as reported in (Grenda et al. 2018). The turbidity was measured using a turbidity meter (Hanna LP 2000, Ronchi di Villafranca, Italy), and each result is the average of three replicate Table 2 Raw water characterization data

Elemental analysis characterization data

Tannin raw material

C (wt%)

H (wt%)

N (wt%)

Mimosa ME-SA Mimosa ME-NZ Quebracho

52.15 53.49 56.24

5.49 5.67 5.68

0.66 0.67 0.43

Parameter

Value

pH COD Turbidity Absorbance Color

8.0 406 mg O2 L−1 70 NTU 1.743 at 350 nm Pink

Environ Sci Pollut Res

measurements. Turbidity reduction was calculated using the following equation: Turbidity reduction ð%Þ ¼

T 0 −T f *100 T0

where T 0 and Tf (NTU) are the turbidity of the initial wastewater at time zero and the turbidity of the supernatant of the treated sample at a given time, respectively. Comparisons were made with the commercial cationic (at 80 and 40% charge) PAMs.

Coagulation-flocculation experiment: wastewater treatment plant The wastewater treatment plant (Fig. 3) consists of a coagulation-flocculation/biology unit of 10 L and a 10 L decanter with open outlet for water sample collection and controlled recirculation/removal of settled sludge. The WWTP was running 24 h, 7 days per week. The coagulants could be added every 4 h, during 12 h; then, overnight, the system was running without addition of any coagulant. Three times per day (every 4 h during the 12 h shift), purified water samples were collected, and the turbidity, absorbance at 350 nm, and pH were measured. The chemical oxygen demand (COD) was measured just for the samples with the best turbidity and color removal results. The pH of inlet water was reduced to pH 7.0 by using phosphoric acid. Urea (at 40%), used as nutrient, was also added to the inlet water container, due to biological requirements. Additionally, the biodegradability when using the applied tannin-based coagulants was evaluated, based on the amount of produced sludge and compared with WWT results obtained using a commercially available inorganic and polymeric blend, based on polyaluminium chloride, iron trichloride, and polyamine, Amber RF05, which is well suited for

decanter-settling. Amount/volumes of produced wastes in both cases, either by using procedure with Amber RF05 or tannin-based coagulants, were evaluated during 2 weeks and then compared. Furthermore, during 3 days, the recycle mode was tested, where no water was coming in, but wastewater treatment plant continued to function—this may simulate longer holiday periods. It was noticed that purified water after that period presented better properties comparing to the last water sample measured before applying the recycle mode.

Results and discussion This study may be divided into two sections: the synthesis and characterization of tannin-based coagulants from different sources at different scales and the application of the obtained products in industrial wastewater treatment. The first part of the work examines the influence of synthesis variables on the physical-chemical properties of the polymer produced. The application and performance of the biocoagulants in the treatment of an industrial wastewater included the use of the tannin itself, as well as in combination with flocculants.

Synthesis and characterization of the tannin-based coagulants As a first step, an optimization study of tannin modification at laboratory scale using Mimosa ME-SA tannin was carried out. Cationic tannin-based coagulants are products of condensation/polymerization reactions, where the chain length and crosslinking between tannin molecules, formaldehyde, and Mannich base progresses over time. This leads to a thickening of the product. However, the final pH of the polymer, in solution, can influence the extent to which the condensation reaction continues during storage. It is thus important to determine the relationship between the pH of the final product and its viscosity (Fig. 4). 350

Shear viscosity (cP)

300 250 200 150 100 50 0 3.4

Fig. 3 Wastewater treatment plant. (1) Inlet wastewater; (2) biological unit; (3) decanter; (4) purified collected water; (5) timer; (6) pump

2.0

1.8 pH

1.6

1.5

Fig. 4 Influence of pH on the final viscosity of the modified Mimosa ME-SA tannin (after acidification step)

Environ Sci Pollut Res

A tannin-based coagulant (Mimosa ME-SA) modified and after Bdilution^ though without acidification had a pH of 3.4 and a viscosity (measured at a shear rate of 30 rpm and a temperature of 24.5 ± 0.5 °C) of 339 cP (Fig. 4). Following reduction of pH (with hydrochloric acid) to 2.0, the viscosity drastically decreased to 131 cP. It was also observed that viscosity gradually decreased with the increase of acidity of the final product. It was, therefore, decided to acidify all the obtained products of tannin aminomethylation to pH 1.6. In all the modification trials, A. mearnsii tannin powder from South Africa was dissolved at 30 °C during 1 h; then, the MS was added. The MS activation time was fixed at 10 min and the reaction was carried out at 85 °C. The influence on modification, of reaction time and amount of FA added to the reaction mixture, was evaluated through the changes in shear viscosities of the final products (Fig. 5). Increasing the FA/Tannin ratio from 0.02 till 0.04 led to an increase in the viscosity of the obtained products (A compared with B and C with D). During the modification, the formaldehyde and Mannich base propagated crosslinking between the flavonoid units of tannin (Sánchez-Martín and BeltránHeredia 2014), which led to an increase in the molecular weight and, correspondingly, to a higher shear viscosity of the final modified tannin. These changes in viscosity were much more significant when working with tannin supplied from New Zealand, especially at higher FA values. If the reaction time is longer than a certain threshold, or if simultaneously FA dosage is relatively high (long reaction time and high FA/tannin ratio), then crosslinking occurs and tannin gelation takes place. In this case, the product is useless as a coagulation agent for wastewater treatment. Additionally, it was reported (Grenda et al. 2018) that if the initial viscosity of the obtained tannin product is above 200 cP, the latter acquires a gel-type structure by aging very fast, in approximately 14 days, which results in the insolubility of the product. The desired initial viscosity, in order to guarantee a long shelf life

of more than 6 months of the product, is below 50 cP. For a shelf life of 1 year or more, stored at 20–25 ºC, an initial viscosity of 20 cP would be reasonable. From Fig. 5, it was inferred that products obtained from South Africa tannin (trials A and B) had lower molecular weights (lower reactivity), compared to those from New Zealand tannin (trials C and D). The differences in the viscosities obtained while using a FA/tannin ratio of 0.02 and 60 min of reaction time, between Mimosa ME-SA (29 cP) and Mimosa ME-NZ (37 cP), were approximately 28%, as reaction time increased to 120 min, the increase was of a factor 1.8 (80 cP for Mimosa ME-SA and 145 cP for Mimosa ME-NZ). The process developed for the Mimosa ME tannins was directly applied to the Quebracho tannin raw material. However, some modifications to the procedure needed to be applied, as already described previously, since this type of tannin does not dissolve at 30 °C. Note that Quebracho tannin differs from Mimosa ME-SA tannin, since it comes from a different source (Schinopsis balansae), and it is a tannin extracted from wood which, additionally, has a different origin (Argentina). Thus, having a different chemical composition, it was foreseen to obtain products with different viscosities compared to those from the Mimosa ME trials. A comparison of results is shown in Fig. 6. The obtained Quebracho-based products were less viscous than the Mimosa ME-SA ones (trials A compared to E and B to F). Since, to have an adequate balance between shelf life and the efficiency in wastewater treatment, viscosities in the range of 40 and 50 cP are desirable, as discussed below, Quebracho-based coagulants with higher viscosities could be interesting. To obtain Quebrachobased products with a slightly higher viscosity, it would be important to work with higher dosages of FA to intensify the crosslinking between the FA, Mannich base, and tannin molecules, in order to obtain products with higher molecular weights. The bio-coagulants TB2 (41 cP) and TF4

Fig. 5 Shear viscosity (at 25 °C) of the modified Mimosa ME-SA and Mimosa ME-NZ tannin products for different FA/tannin ratios as a function of the reaction time

A B C D

1000 900

Shear viscosity (cP)

800 700 600 500

Test

FA/Tannin Ratio

A

0.04

B

0.02

C

0.04

D

0.02

160

200

400

300 200 100 0 0

40

80 120 Time (min)

Mimosa ME SA

NZ

Environ Sci Pollut Res 120

A

Test

FA/Tannin Ratio

Type of Tannin

E

A

0.04

F

B

0.02

Mimosa ME-SA

E

0.04

F

0.02

B

Shear viscosity (cP)

100 80 60

Quebracho

TTBA22

40

TF4

20 0 0

40

80 120 Time (min)

160

200

Fig. 6 Shear viscosity (at 25 °C) of the modified Mimosa ME-SA and Quebracho tannin products for different FA/Tannin ratios, as a function of the reaction time

(40 cP) were selected for further evaluation in wastewater treatment, considering their similar viscosities and amount of added FA. The raw materials, as well as the modified tannins, were characterized by FTIR spectroscopy (Fig. 7). The spectra of the two Acacia mearnsii tannins (South Africa and New Zealand origins) were similar to each other as was previously reported (Grenda et al. 2018). The FTIR spectrum of Quebracho tannin was slightly different, reflecting a unique source (wood from Schinopsis balansae). The spectrum of this tannin (Fig. 7) showed high intensity bands at 1615, 1519, and 1450 cm−1 from the –C=C stretching vibrations. Other bands were observed at 1371, 1284, 1243, 1204, 1160, 1114, 1036, 975, and 778 cm−1. These bands were related to O-H deformation, C–O stretching in =C–O–H and =C–O–C bonds (asymmetric and symmetric modes), CH2 deformation, in-plane and out-of-plane C–H deformations, as the most relevant vibrations. The presence of minor amounts of polysaccharides and sugars (Pasch et al. 2001) can also contribute to some of the absorption features observed in the spectrum of Quebracho tannin.

Quebracho tannin

Transmiance (a.u.)

Modified Quebracho tannin

2000

1615

1519 1450

1284

1114

1717 935

1521 1609 1800

1600

1470

1288

1400

1200

1000

800

600

400

Wavenumber (cm-1)

Fig. 7 FTIR spectra of the raw Quebracho tannin and modified Quebracho tannin (TF4)

The FTIR spectrum of the Quebracho tannin modified by the Mannich reaction (Fig. 7) was significantly different from that of the raw material. In particular, two new bands were seen at 1717 and 935 cm−1, assigned to the C=O stretching of the carbonyl bonds and the C–N stretching in the alkylammonium groups, respectively, that formed during the reaction. The band formerly at 1450 cm−1 was shifted to 1470 cm−1 in the modified tannin spectrum, indicating overlapping with absorption coming from asymmetric bending of CH3 groups. The expected reactions of the raw tannins with FA and amine, where positively charged groups were added to the tannin backbone, were thus confirmed. The modification process of Mimosa ME bark tannin (from South Africa), optimized at laboratory scale in a 1-L reactor, was up scaled to a 5-L reactor. The Mannich modification was performed using MS at 85 °C for 105 min with an extra dosage of 37 wt% aqueous formaldehyde (FA/tannin ratio of 0.02). The reaction procedure resulted in a bio-coagulant with a shear viscosity of 53 cP (after dilution and acidification till pH 1.6), measured at 25 °C and 30 rpm. Therefore, it was possible to reproduce the 1-L reactor procedure with a minor adjustment, which yielded a bio-based coagulant with a viscosity of 49 cP (FA/tannin ratio of 0.02, 85 °C, 100 min of reaction time).

Wastewater treatment Preliminary screening of an industrial wastewater Several experiments aimed at industrial wastewater clarification were carried out using two selected tannin-based coagulation agents: samples TB2 and TF4. (Fig. 6). Dual systems with synthetic PAM of different charges: BHMW and FHMW at 80 and 40 wt% charge, respectively, were also tested. The results for the turbidity and color removal using only tannin or PAM (single system) are presented in Figs. 8 and 9, respectively. As shown in Fig. 8, quite similar turbidity reductions were obtained for the set of experiments using the synthetic flocculants, with no strong influence of the flocculant charge, and again for the experiments using the bio-coagulant independently of which one was being used. Better performance was obtained with synthetic PAM, with up to 95% of turbidity removal after 1 h of settling while using 200 ppm BHMW which performed slightly better than FHMW. Increasing the dosage of either BHMW or FHMW led to a slight increase in their performance. On the other hand, using 200 ppm of the modified Mimosa ME-SA tannin coagulant allowed a 68% of turbidity reduction after 1 h of settling. Moreover, increasing the dosage of the used bio-coagulant from 50 to 100 ppm and then to 200 ppm provided significant improvements in the clarification results. In all tested cases,

Environ Sci Pollut Res 100 50ppm BHMW

90

100ppm BHMW 200ppm BHMW

80 Turbidity reducon (%)

Fig. 8 Preliminary screening in industrial effluent clarification based on turbidity reduction using single systems of coagulant/ flocculant, with dosages of 50– 200 ppm, at room temperature and neutral pH

50ppm FHMW 100ppm FHMW

70

200ppm FHMW

60

50ppm Mimosa ME-SA 100ppm Mimosa ME-SA

50

200ppm Mimosa ME-SA 50ppm Quebracho

40

100ppm Quebracho

30

200ppm Quebracho

20 25

30

the coagulant from Mimosa ME-SA showed a slightly better performance over the one from Quebracho at similar dosages. Furthermore, Fig. 9 shows, complementary to turbidity reduction (Fig. 8) results, the influence on color removal of using single systems, either only natural-based coagulants (Mimosa ME-SA or Quebracho) or synthetic flocculants of different charge content (BHMW or FHMW). It shows the better efficiency in color removal while using 200 ppm of BHMW over the FHMW or natural-based coagulation agents. Moreover, as it was observed in Fig. 8 for the turbidity reduction, also in color removal, increasing the dosage of the used wastewater treatement agent from 50 ppm through 100 ppm up to 200 ppm provided significant improvements in the clarification results, both when using tannin-based coagulants or PAMs. In all tested cases of color removal, the coagulant from Mimosa ME-SA showed a slightly better performance over the one from Quebracho at similar dosages. However, overall, obtained results using only natural-based coagulants were still not sufficient for water discharge or reuse for industrial purposes.

40

45 Time (min)

50

55

60

65

A dual system comprising a bio-coagulant/synthetic flocculant (FHMW or BHMW) was evaluated with the aim to provide good clarification and reduce the sludge. Figure 10 shows the results of the color removal and turbidity reduction of the industrial effluent using dual systems. Five different parameters were tested: two different dosages of tanninbased coagulant (100 and 200 ppm) and as well two different cationic flocculants (40 wt% charged and 80 wt% charged) at one concentration (5 ppm). Using a bio-coagulant (200 ppm) together with 5 ppm of synthetic flocculant with 40% charge (FHMW) provided the best color removal (91%) and the highest turbidity reduction (92%). The latter is higher than the turbidity removal achieved when using only the bio-coagulant. These results show the positive impact of combining a bio-coagulant with just a small amount of a traditional synthetic polyelectrolyte, allowing to reduce dramatically the need for the synthetic polymer, with, of course, an overall positive environmental benefit. Using the same amount of coagulant but with the 80 wt% charged flocculant (BHMW) enabled to reduce the color in 73% and

100 50ppm BHMW

95

100ppm BHMW 200ppm BHMW

90 Colour removal (%)

Fig. 9 Preliminary screening results in industrial effluent clarification based on color removal, using single systems of coagulant/flocculant, with dosages of 50–200 ppm, at room temperature and neutral pH

35

50ppm FHMW 100ppm FHMW

85

200ppm FHMW 50ppm Mimosa ME-SA

80

100ppm Mimosa ME-SA 200ppm Mimosa ME-SA

75

50ppm Quebracho 100ppm Quebracho

70

200ppm Quebracho

65 25

30

35

40

45 Time (min)

50

55

60

65

100 90 80 70 60 50 40 30 20 10 0

100 90 80 70 60 50 40 0

50 100 150 200 Tannin-based coagulant (ppm)

Turbidity reducon (%)

Colour removal (%)

Environ Sci Pollut Res

30 250

Colour removal aer addion of 5ppm BHMW

demonstrated in the jar test. Increasing the bio-coagulant dosage from 100 to 200 ppm only slightly improved the results. Applying the dual system of natural-based coagulant with a synthetic flocculant (trials 3–6, Table 3) significantly improved color removal; however, results for the turbidity and COD reduction did not always improve compared to trials 1–2 (FHMW did not lead to improvements but BHMW led to some improvements in turbidity removal). In fact, the procedure developed in the jar test (200 ppm of tannin-based coagulant followed by the addition of 5 ppm of 40% charged FHMW) did not work as expected in the WWTP and only 14% (89% in jar test) of color and 29% (93% in jar test) of turbidity were removed. Nevertheless, using 200 ppm of tannin-based coagulant followed by the addition of 20 ppm of cationic (80% charge) PAM showed reasonably good clarification results: removals of 67 and 57%, in color and turbidity, respectively. Furthermore, during 3 days, recirculation mode (RCM) was tested (with procedure from trial 4—Table 3), where no water was coming in but wastewater treatment plant continued to function. It was noticed that purified water after that period presented better properties compared to the last water sample measured before applying this mode. The color removal increased from 48% (before RCM) till 67% (after RCM), and an improvement in turbidity removal was also observed from 46% (before RCM) to 65% (after RCM). However, the COD reduction decreased from 36% (before RCM) to 29% (after RCM). The results of the effluent treatment while applying dual system to decanter and to the biology unit are shown in Table 4, where four different conditions, regarding the dosage of the synthetic flocculant, were tested at this stage. It was demonstrated in trial 2 that using low amount of BHMW (2 ppm) in the biology unit together with 50 ppm of tanninbased coagulant followed by the addition of 50 ppm of the modified tannin and 10 ppm of BHMW in the decanter was the best procedure for this effluent treatment in the WWTP. The COD was reduced to 123 mg/L (70% reduction), while color and turbidity were reduced in 85 and 79%, respectively. Repeating this procedure with the FHMW synthetic agent with

Colour removal aer addion of 5ppm FHMW Turbidity reducon aer addion of 5ppm BHMW Turbidity reducon aer addion of 5ppm FHMW

Fig. 10 Preliminary screening in industrial wastewater clarification using dual systems: tannin-based coagulant Mimosa ME-SA at 100 and 200 ppm, with synthetic flocculant BHMW or FHMW at 5 ppm. Results obtained at neutral pH, 20 °C and after 1 h of settling. The best color and turbidity reduction was highlighted

turbidity in 45%. Moreover, at the lower dosage of biocoagulant (100 ppm), when using 5 ppm of BHMW, better results were obtained (73% of turbidity reduction and 87% of color removal.). The opposite trend was observed when using 40 wt% charged polyacrylamide-based flocculant (FHMW). For pilot testing, the tannin-based coagulant followed by FHMW was selected.

Industrial WWTP—wastewater treatment plant The effectiveness of the 20-L wastewater treatment plant process was lower than that of the jar test trials. Table 3 shows the results for the reduction of color, turbidity, and COD while applying the effluent treatment procedure with modified Mimosa ME-SA tannin-based coagulant (TB2 with 41 cP) to industrial wastewater, with properties described in Table 2. The WWTP (Fig. 3) includes a biology unit, which requires application of necessary nutrients and provides neutral pH. To meet those requirements, urea and phosphoric acid were added to the inlet water. Table 3 shows the results obtained when treatment was applied in the decanter. Using modified tannin alone (trials 1–2, Table 3) provided poor clarification, as it was already Table 3 Trial

1 2 3 4 5 6

Procedures and results of industrial wastewater treatment in WWTP, while applying treatment in decanter Used procedure in decanter Tannin-based coagulant (ppm)

Synthetic flocculant (ppm)

100 200 200 200 200 200

_ _ 5 (FHMW) 20 (FHMW) 5 (BHMW) 20 (BHMW)

pH

7.7 7.3 7.9 8.7 7.8 8.1

Color removal (%)

2 3 14 48 62 67

Turbidity reduction (%)

40 49 29 46 50 57

COD Real value (O2 mg L−1)

Reduction (%)

198 192 347 258 244 243

51 53 15 36 40 40

Environ Sci Pollut Res Table 4 Procedures and results of industrial wastewater treatment in the WWTP, while applying treatment (dual system) in decanter as well as in biology unit Trial

Used procedure Biology

1 2a 3*

pH Decanter

Color removal (%)

Turbidity reduction (%)

COD Real value (mg O2/L)

Reduction (%)

C

F

C

F

50 50

4 2

50 50

20 10

8.8 6.9

67 85

65 79

289 123

29 70

50

2

50

10

6.9

82

74

161

60

C corresponds to tannin-based coagulant, in ppm, F corresponds to synthetic flocculant of 80 wt% charge (BHMW) in ppm *Trial performed with FHMW flocculant of 40 wt% cationic charge a

The most effective procedure

lower cationicity (40% charge), in trial 3, allowed also for reasonable results, even if slightly lower removal of COD, turbidity, and color (60, 74, and 82%, respectively) were achieved. Using a higher dosage of the BHMW synthetic flocculant (4 ppm in the biology unit and 20 ppm in the decanter) while maintaining the addition of natural-based coagulant constant (trial 1), led to worse results. It must be referred that pH could not be kept at the same level as in trial 2 and was slightly different. It was noticed that using dosages higher than 2 ppm of the synthetic flocculant in the biology unit and 10 ppm in the decanter highly increased the size of flocs, but that did not lead to better clarification results due to poorer settling. It is known that tannin-based products work as coagulation agents. Thus, in the described process of wastewater treatment, they allow to destabilize the impurities in the effluent, but are not enough, on their own, to guarantee good reduction of color and turbidity (see Table 3), even if COD reduction is reasonable. Moreover, destabilization of particles and dye molecules in the effluent occurred mainly due to charge neutralization, after adsorption of the cationic tannin additive on the small particle surface. As soon as the charge was neutralized, the small suspended elements were capable of gathering into fragile flocs and the surrounding water becomes clearer. Additionally, coagulants with low or medium molecular weight (as in the case of, e.g., tannin-based coagulants) would require adequate long contact time to allow slow flocs formation which would lead to maximum removal of suspended solids (turbidity reduction). Thus, addition, at this stage, of a minor amount of synthetic flocculant (PAM), with higher molecular weight and charge density compared to modified tannin, to the effluent system containing fragile flocs composed of effluent impurities/tannin coagulant, helped to bridge, and strengthen the flocs, further increasing floc size and corresponding settling rate, which results in higher clarification efficiency. Improvements in color removal as well as turbidity reduction were observed, even if not in COD reduction (Table 3). The short retention time in the decanter did not allow to obtain high clarification results while using wastewater treatment procedures only in the decanter unit.

Nevertheless, by simultaneous addition of small amount of PAM together with the tannin-based coagulant, in the biology unit and in the decanter, significant improvements in the color removal/turbidity reduction were observed. The bio-coagulant induces de-stabilization of the effluent particles, which, followed by the addition of very small amounts of the synthetic PAM leads to the formation of larger, more resistant, and denser aggregates, due to bridging. This treatment has to be applied both in the biology unit and in the decanter (Table 4) to allow enough residence and contact time. Thus, with the proposed methodology, it was possible to substitute the usual large amounts of synthetic polymers used in wastewater treatment, for this particular effluent, by a natural bio-coagulant, combined with only very small amount (12 ppm for optimal conditions) of synthetic polymer, to achieve extremely good treatment efficiency, especially regarding COD reduction. Moreover, the procedure developed at lab scale could be tuned to be applied in a WWTP. Dosing of the bio-coagulant and flocculant took place over a period of 3 weeks. As a comparison, for an additional 3 weeks, the station was run only using a commercial blend of polyaluminium chloride with iron trichloride and polyamine (Amber RF05). It was noticed that the amount of obtained sludge while using only Amber RF05 was significantly higher compared to the sludge obtained while using tannin-based coagulant with slight addition of synthetic flocculant. The suggested procedure with use of natural-based wastewater treatment agent with a slight addition of synthetic flocculant allows to reduce significantly the amount of used oil-based, harsh to the environment wastewater treatment agent. Moreover, suggested strategy also has a positive impact in reduction of produced sludge since it is decomposed by the bacteria in biology unit.

Conclusions Cationic tannin-based coagulants were synthesized via the Mannich aminomethylation with formaldehyde and

Environ Sci Pollut Res

dimethylamine hydrochloride, from bark tannin extract of Acacia mearnsii (from different origins: New Zealand and South Africa) as well as from wood tannin extract of Schinopsis balansae. Different experimental conditions, namely the formaldehyde /tannin ratio and reaction time, were evaluated at a laboratory scale, and their influence on the physico-chemical properties of the final products was confirmed. The effect of the final acidification step (the final pH) on the shear viscosity of the resulting products was also determined. With the acidity increase, a significant viscosity reduction was observed. It was possible to control the biocoagulant viscosity and tune the reaction conditions for all tested raw materials. The optimized modification conditions (0.02 FA/tannin ratio, 85 °C, 100 min of reaction time) for A. mearnsii tannin from South Africa at laboratory scale were up-scaled successfully with a resulting product with shear viscosity of 53 cP while at 1-L scale, it was 49 cP. The modified tannins from Acacia mearnsii and Schinopsis balansae, with a shear viscosity of ca. 40 cP, were evaluated for an industrial effluent treatment. However, an acceptable performance was obtained only using a dual system with cationic polyacrylamide. The turbidity removal could increase from 68% (while using 200 ppm of bio-coagulant alone) to 93% (200 ppm of bio-coagulant followed by 5 ppm of cationic PAM with 40% charge). The dual system approach showed to be effective for the clarification (color, turbidity, and COD reduction) of the industrial effluent in a (20 L) wastewater treatment plant, using only small amount of the synthetic flocculant as a complementary additive. Acknowledgments The authors would also like to express gratitude to Extract Dongen B.V., Netherlands, for kindly providing the Mimosa ME tannin, and to Dr. Peter Urben and Dr. Alberto Venica for their advice and assistance in the laboratory. Funding information The authors are grateful for the financial support of the Marie Curie Initial Training Networks (ITN)–European Industrial Doctorate (EID) (Grant agreement FP7-PEOPLE-2013-ITN-604825) and FCT/MCTES (PIDDAC) (Pest/C/EQB/UI0102/2013), which is cofinanced by the European Regional Development Fund (ERDF) through the program COMPETE (POFC).

References Akiyama H, Fujii K, Yamasaki O, Oono T, Iwatsuki K (2001) Antibacterial action of several tannins against Staphylococcus aureus. J Antimicrob Chemother 48:487–491. https://doi.org/10.1093/ jac/48.4.487 Ashok P, Upadhyaya K (2012) Tannins are astringent. J Pharmacogn Phytochem 1:45–50. https://doi.org/10.11648/j.ijnfs.20140304.18 Beltran-Heredia J, Torregrosa J, Dominguez JR, Peres JA (2001) Comparison of the degradation of p-hydroxybenzoic acid in aqueous solution by several oxidation processes. Chemosphere 42:351– 359. https://doi.org/10.1016/S0045-6535(00)00136-3 Beltrán-Heredia J, Sánchez-Martín J, Gómez-Muñoz MC (2010) New coagulant agents from tannin extracts: preliminary optimisation

studies. Chem Eng J 162:1019–1025. https://doi.org/10.1016/j.cej. 2010.07.011 Beltrán-Heredia J, Sánchez-Martín J, Jiménez-Giles M (2011a) Tanninbased coagulants in the depuration of textile wastewater effluents: elimination of anthraquinonic dyes. Water Air Soil Pollut 222:53– 64. https://doi.org/10.1007/s11270-011-0803-z Beltrán-Heredia J, Sánchez-Martín J, Martín-Sánchez C (2011b) Remediation of dye-polluted solutions by a new tannin-based coagulant. Ind Eng Chem Res 50:686–693 Bouaouine O, Bourven I, Khalil F, Baudu M (2018) Identification of functional groups of Opuntia ficus-indica involved in coagulation process after its active part extraction. Environ Sci Pollut Res 25: 11111–11119. https://doi.org/10.1007/s11356-018-1394-7 Covington AD, Lilley TH, Song L, Evans CS (2005) Collagen and polyphenols : new relationships and new outcomes. Part 1. Flavonoid reactions for new tanning processes. J Am Leather Chem Assoc 100: 325–335 dos Santos JD, Veit MT, Juchen PT, da Cunha Gonçalves G, Palácio SM, Fagundes-Klen M (2018) Use of different coagulants for cassava processing wastewater treatment. J Environ Chem Eng 6:1821– 1827. https://doi.org/10.1016/j.jece.2018.02.039 Ebeling JM, Sibrell PL, Ogden SR, Summerfelt ST (2003) Evaluation of chemical coagulation–flocculation aids for the removal of suspended solids and phosphorus from intensive recirculating aquaculture effluent discharge. Aquac Eng 29:23–42. https://doi.org/10. 1016/S0144-8609(03)00029-3 Graham N, Gang F, Fowler G, Watts M (2008) Characterisation and coagulation performance of a tannin-based cationic polymer: a preliminary assessment. Colloids Surf A Physicochem Eng Asp 327:9– 16. https://doi.org/10.1016/j.colsurfa.2008.05.045 Grenda K, Arnold J, Hunkeler D et al (2018) Tannin-based coagulants from laboratory to pilot plant scales for coloured wastewater treatment. BioResources 13:2727–2747. https://doi.org/10.15376/ biores.13.2.2727-2747 Hameed YT, Idris A, Hussain SA, Abdullah N, Man HC, Suja F (2018) A tannin–based agent for coagulation and flocculation of municipal wastewater as a pretreatment for biofilm process. J Clean Prod 182:198–205. https://doi.org/10.1016/j.jclepro.2018.02.044 Haslam E (1988) Plant polyphenols (syn. vegetable tannins) and chemical defense-a reappraisal. J Chem Ecol 14:1789–1805. https://doi.org/ 10.1007/BF01013477 Haslam E (2005) Practical polyphenolics: from structure to molecular recognition and physiological action. Cambridge University Press, Cambridge Hesse MCS, Santos B, Selesu NFH, Corrêa DO, Mariano AB, Vargas JVC, Vieira RB (2017) Optimization of flocculation with tanninbased flocculant in the water reuse and lipidic production for the cultivation of Acutodesmus obliquus. Sep Sci Technol 52:936–942. https://doi.org/10.1080/01496395.2016.1269130 Khanbabaee K, van Ree T (2001) Tannins: classification and definition. Nat Prod Rep 18:641–649. https://doi.org/10.1039/b101061l Khandare RV, Govindwar SP (2015) Phytoremediation of textile dyes and effluents: current scenario and future prospects. Biotechnol Adv 33: 1697–1714. https://doi.org/10.1016/j.biotechadv.2015.09.003 Martínez-Huitle CA, Ferro S (2006) Electrochemical oxidation of organic pollutants for the wastewater treatment: direct and indirect processes. Chem Soc Rev 35:1324–1340. https://doi.org/10.1039/ b517632h Morawetz H (1993) Polyelectrolytes, science and technology. J Polym Sci A Polym Chem 31:2413–2413. https://doi.org/10.1002/pola. 1993.080310927 Özacar M, Şengil IA (2003) Evaluation of tannin biopolymer as a coagulant aid for coagulation of colloidal particles. Colloids Surf A Physicochem Eng Asp 229:85–96. https://doi.org/10.1016/j. colsurfa.2003.07.006

Environ Sci Pollut Res Pasch H, Pizzi A, Rode K (2001) MALDI-TOF mass spectrometry of polyflavonoid tannins. Polymer (Guildf) 42:7531–7539. https://doi. org/10.1016/S0032-3861(01)00216-6 Pintor AMA, Vilar VJP, Botelho CMS, Boaventura RAR (2016) Oil and grease removal from wastewaters: sorption treatment as an alternative to state-of-the-art technologies. A critical review. Chem Eng J 297:229–255 Rebitzer G, Hunkeler D, Jolliet O (2003) LCC—the economic pillar of sustainability: methodology and application to wastewater treatment. Environ Prog 22:241–249. https://doi.org/10.1002/ep. 670220412 Roselet F, Vandamme D, Roselet M, Muylaert K, Abreu PC (2017) Effects of pH, salinity, biomass concentration, and algal organic matter on flocculant efficiency of synthetic versus natural polymers for harvesting microalgae biomass. Bioenergy Res 10:427–437. https://doi.org/10.1007/s12155-016-9806-3

Sánchez-Martín J, Beltrán-Heredia J (2014) Nature is the answer: water and wastewater treatment by new natural-based agents. In: Advances in water treatment and pollution prevention. pp 337–375 Sánchez-Martín J, Beltrán-Heredia J, Solera-Hernández C (2010) Surface water and wastewater treatment using a new tannin-based coagulant. Pilot plant trials. J Environ Manag 91:2051–2058. https://doi.org/ 10.1016/j.jenvman.2010.05.013 Tramontini M, Roman G, Abedini-Torghabeh J et al (2015) Mannich reaction. Eur J Med Chem 16:303–17681. https://doi.org/10.3390/ ijms160817668 Tzoupanos ND, Zouboulis A (2008) Coagulation-flocculation processes in water/wastewater treatment: the application of new generation of chemical reagents WHO, Unicef (2000) Global water supply and sanitation assessment 2000 report. Water Supply 87. http://www.who.int/water_ sanitation_health/monitoring/globalassess/en/