Utilization of galactomannan from spent coffee grounds as coagulant aid to treat synthetic Congo red wastewater


 Over the last few years, there is a significant growth in research exploring natural based coagulant due to its various benefits to decrease or even substitute the usage of inorganic chemical coagulants. Polysaccharide based coagulant and coagulant aid is a promising source for this purpose, due to its abundance. In this study, we reported utilization of galactomannan extracted from spent coffee grounds as natural coagulant aid in coagulation of Congo red synthetic wastewater. The coagulation was done at fixed dosage of FeCl3 (160 mg/L) and pH of 6. The effect of galactomannan as coagulant aid was observed by varying galactomannan dosage (0-140 mg/L) and Congo red concentration (20–70 mg/L). It was found that galactomannan as coagulant aid could increase the removal of Congo red, around 30–90% removal, depends on Congo red concentration, compared FeCl3 only (0–65%). The coagulation adsorption study was also investigated using Langmuir, Freundlich, and Brunauer – Emmet – Teller (BET) isotherm models. It was found that Congo red coagulation using FeCl3 only was following Langmuir isotherm, indicating monolayer – homogenous formation during the coagulation. On the other hand, with the presence of galactomannan the coagulation was best described by BET isotherm, indicating multilayer – heterogeneous adsorption, possibly due to interparticle bridging of galactomannan during colloid aggregation. The findings in this study suggest synergistic effect of galactomannan and FeCl3 in the coagulation process and proved potential of galactomannan from spent coffee grounds as natural coagulant aid.


Introduction
Coagulation and flocculation is a water-wastewater treatment process that is still widely used due to its high efficiency and effectiveness. Inorganic coagulant such as aluminium sulphate, ferrous sulphate, and ferric chloride are commonly used in various treatment conditions. However there are some drawbacks of the previously mentioned metallic salts, such as relatively high coagulant costs, neurotoxic, corrosive and also high sludge volume generation. Therefore many efforts to substitute and reduce utilization of inorganic coagulants have been extensively studied in recent years; one of them is utilization of natural coagulant and coagulant aid.
Generally, based on the active ingredients, natural coagulant could be classified as protein, polyphenol, and polysaccharide (Kristianto 2017). Protein from Moringa oleifera (Baptista et al. 2017), Moringa stenopetala (Dalvand et al. 2016), Leucaena leucocephala (Kristanda et al. 2020;Kristianto et al. 2020), Cocos nucifera (Fatombi et al. 2013), etc. as active coagulating agent has been extensively studied. The utilization of protein based natural coagulant coagulant aid showed prospective results in treating water and wastewater. However the purification of protein extract poses some challenges to its application (Choy et al. 2014). The research polyphenol based coagulant is relatively limited. Among wide range of polyphenols, only tannin based coagulant is explored and commercialized (Kuppusamy et al. 2015). On the other hand, polysaccharide is a promising alternative coagulantcoagulant aid, due to its abundance. Polysaccharide as natural coagulant could be obtained from plant and non-plant. Non plant based polysaccharides such as chitosan and sodium alginate have been studied and used to treat water and wastewater (Saranya et al. 2014). Plant based polysaccharide comprises in various form, such as starch, pectin, gum, etc. Rice, wheat, corn, and sago starch have showed a promising result as natural coagulant (Teh et al. 2014;Aziz and Sobri 2015;Choy et al. 2016); however its utilization could pose some problems due to food issues, thus non-food alternatives should be explored to handle this problem.
Coffee beans are rich in polysaccharides and could become a promising source. With world coffee consumption 10 million tons in 2019/2020 (ICO 2020), abundant waste is generated. It is known from coffee cherry, 49 to 53%w would become spent coffee grounds that are usually discarded without further use (Vandeponseele et al. 2020). On the other hand, it is known that 50% of the beans dry weight is polysaccharides, and half of it is galactomannan fraction (Moreira et al. 2015). Galactomannan is a polymer chain of β-(1→4)mannose as main backbone that interrupted with glucose and branched with α-(1→6)-galactose or arabinose (Buckeridge et al. 2000;Moreira et al. 2015). Galactomannan from seeds such as Ipomoea dasysperma (Sanghi et al. 2006a), Cassia javahikai (Sanghi et al. 2006b), Cassia obtusifolia (Shak and Wu 2015), sesbania (Chua et al. 2020), and Guazuma ulmifolia (Tavares et al. 2020) have been utilized as coagulant aid. It was evident from the previously mentioned researches, galactomannan could assist coagulation process by increasing coagulation performance of inorganic coagulants (FeCl3, Al2(SO4)3, etc.) In this study, we focused on utilization of galactomannan extracted from spent coffee grounds as coagulant aid. Various coagulation parameters such as galactomannan dosage and Congo red concentration were explored. To the best of authors knowledge, similar study has never been done before, thus the results obtained in this study could open up the possibility for utilization of spent coffee grounds.

Extraction of galactomannan
Spent coffee grounds were collected from coffee shops in Bandung, Indonesia. The coffee grounds were then washed using distilled water and then dried using tray dryer at 50°C until the water content below 10%w. The dried spent coffee grounds were then sieved using standard sieve to obtained grounds size 80-10 mesh. The extraction process was done by mixing spent coffee grounds with distilled water at feed to solvent ratio 1:20 (w/v). The extraction was done at 70°C and neutral pH for 2h. After extraction, the extract was then separated and then ethanol (90% v/v) was added to separate the galactomannan from the solvent at ethanol to extract ratio of 3:1 (v/v). The galactomannan would be obtained as whitish precipitate after 48h. The precipitate was then separated using muslin cloth and then dried using tray dryer at 50°C until the water content below 10%w. The dried precipitate was stored in a desiccator for further analysis and coagulation study. The obtained galactomannan was characterized using Scanning Electron Microscope -Energy Dispersive X-Ray Spectroscopy (SEM-EDS; Hitachi SU3500), X-ray Diffraction (XRD; Bruker D8 Advance with Cu K-α radiation) and Fourier Transform Infrared (FTIR) Spectroscopy (Prestige 21 Shimadzu Instruments) by KBr pellet method.

Jar test study
The coagulation test was done using a standard jar test apparatus. Congo red solution was used as a model dye for the coagulation study. A stock solution of 1g/L was prepared and further diluted to obtain solution with desired concentration. The coagulation study was done at various galactomannan dosage and Congo red concentration, presented in Table 1, at fixed pH 6.0, which is known as the optimum pH for FeCl3 (Boughou et al. 2018;Bakar and Halim 2013), and FeCl3 dosage of 80 mg/L. The coagulation was done by mixing Congo red solution, FeCl3 and galactomannan at rapid mixing of 200 rpm for 3 min, followed by slow mixing (60 rpm, 30 min) and settling (1 h). The initial (Ci) and final concentration (Ce) of Congo red was measured using a visible spectrophotometer (Genesys 20) at its maximum wavelength of 500 nm. Congo red removal was calculated using Equation 1. The sludge volume formed after coagulation was measured after 1 h settling by using Imhoff cone, and calculated by following Equation 2. The obtained sludge with and without galactomannan was characterized using SEM-EDS to observe the morphology and atomic composition.

Coagulation isotherm adsorption study
Investigation of the possible coagulation mechanism in this study was done by approach of isotherm adsorption models, namely the Langmuir, Freundlich and Brunauer -Emmett -Teller (BET) isotherm models. Coagulation adsorption capacity (qe; mg/mg) was calculated from the data of various Congo red concentration coagulation study by following Equation 3, where m is the coagulant dosage (mg/L) and V is the total solution volume (L).

Characterization of galactomannan
The characteristics of galactomannan extract from spent coffee grounds are presented in Figure 1. Based upon the observation of galactomannan morphology using SEM (Fig 1.a), it could be observed that the galactomannan extract possessed irregular shape with some cavities on the surface. Similar shape was reported by previous researchers (Niknam et al. 2020;Prashanth et al. 2006). Furthermore from the EDS analysis (Figure 1.b), the extract consisted of C and O atom, indicated the obtained extract was relatively pure. From the XRD spectra ( Figure   1.c), it could be observed that the galactomannan was in amorphous phase. A sharp crystalline peak was observed at 2θ of 20°, similar with galactomannan from guar (Mudgil et al. 2012;Liyanage et al. 2015), fenugreek seeds (Niknam et al. 2020), and Leucaena leucocephala seeds (Mittal et al. 2016). The FTIR spectra is presented in Figure 1.c with several notable peaks discussed as followed. It could be observed that peak at 3388.9 cm -1 due to O-H stretching that [ Figure 1]

Effect of galactomannan dosage
The effect of galactomannan addition in the coagulation system is presented in Figure 2.
It could be observed that without any coagulant aid, the FeCl3 could give 27.5% removal of Congo red. Addition of galactomannan increased the %removal with the highest removal of 40% was obtained at galactomannan dosage of 80 mg/L. This increase was possible due to interparticle bridging mechanism of galactomannan that played an important role in aggregation and flocs formation that resulted in higher %removal (Sanghi et al. 2006b). Further addition of galactomannan did not give any increase to the coagulation performance, instead slight decrease of %removal was observed. This was possible due to overdosing of galactomannan resulted in gave higher sludge volume. At removal of 27.5%, no sludge volume could be observed due to low removal. With addition of galactomannan, the sludge volume also increased, confirming the role of galactomannan as particle bridges that helped formation of flocs. At overdosing condition, the sludge volume was slightly higher that possibly due to repulsion forces between flocs that making the sludge become more voluminous (Kristianto et al. 2019).

Effect of Congo red concentration
The study of effect of Congo red concentration was done at pH 6.0 and galactomannan dosage of 80 mg/L which gave the highest Congo red removal. The profile of % removal and sludge volume at various Congo red concentrations without and with galactomannan are presented in Figure 3. It could be observed that high removal was obtained at low Congo red  (Mercê et al. 2001).
Synergistic effect between natural active ingredient with metal ions has been reported before to increase coagulation performance (Okuda et al. 2001;Moral et al. 2016;Devrimci et al. 2012).
[ Figure 3] The sludge obtained from the coagulation process at Congo red concentration of 20 mg/L was furthermore observed using SEM-EDX, presented in [ Figure 4]

Isotherm adsorption
It is known that both charge neutralization and interparticle bridging coagulation mechanisms are preceded by adsorption which the rate determining step of the coagulation process. Based on this assumption, utilization of isotherm adsorption models is justified. The The parameters of Langmuir and BET isotherm models are presented in Table 2. It could be seen based on the high R 2 values, the coagulation process could be described by Langmuir isotherm model. Furthermore for coagulation using FeCl3 the RL values were in range 0 < RL < 1, indicating a favorable adsorption. However the Langmuir isotherm model could not be used to describe the adsorption-coagulation process with the presence of galactomannan, depicted with negative KL and RL value. The negative value of KL could indicate the coagulation mechanism was multilayer in a non-uniform or random distribution (Hossain et al. 2019). The Freundlich constant (1/n) for FeCl3 coagulation showed a negative value, confirming that the adsorption process was monolayer. As for the coagulation in the presence of galactomannan, the adsorption was favorable, as shown in the 1/n value between 0 and 1. Furthermore, it could be said that the adsorption was heterogeneous as the value was close to 0 (Foo and Hameed 2010). However the low R 2 value for both coagulation conditions might implicate that Freundlich isotherm model was not suitable to describe the adsorption-coagulation process in this study. The BET isotherm model could be moderately fitted to the experimental data, as shown in R 2 value around 0.88.
Negative value of A could indicate the surface of coagulant was already saturated with the dye and the adsorption was in multilayer (Hossain et al. 2019;Ngteni et al. 2020). Based on suitability of the isotherm models, it could be considered that the Congo red coagulation using FeCl3 was monolayerhomogeneous adsorption. Similar observation was reported when using FeCl3 as coagulant (Agbovi and Wilson 2017;Kastl et al. 2004). On the other hand, under the presence of galactomannan as coagulant aid, the adsorption was multilayer and heterogeneous, indicating an interparticle bridging mechanism of polymer (galactomannan) during colloid aggregation (Szilagyi et al. 2014).