Synthesis of porous biomass fly ash-based geopolymer spheres for efficient removal of methylene blue from wastewaters

Synthesis of porous biomass fly ash-based geopolymer spheres for efficient removal of methylene blue from wastewaters

Accepted Manuscript Synthesis of porous biomass fly ash-based geopolymer spheres for efficient removal of methylene blue from wastewaters Rui M. Novai...

4MB Sizes 0 Downloads 1 Views

Accepted Manuscript Synthesis of porous biomass fly ash-based geopolymer spheres for efficient removal of methylene blue from wastewaters Rui M. Novais, João Carvalheiras, David M. Tobaldi, Maria P. Seabra, Robert C. Pullar, João A. Labrincha PII:

S0959-6526(18)33004-X

DOI:

10.1016/j.jclepro.2018.09.265

Reference:

JCLP 14400

To appear in:

Journal of Cleaner Production

Received Date: 10 April 2018 Revised Date:

12 September 2018

Accepted Date: 30 September 2018

Please cite this article as: Novais RM, Carvalheiras Joã, Tobaldi DM, Seabra MP, Pullar RC, Labrincha JoãA, Synthesis of porous biomass fly ash-based geopolymer spheres for efficient removal of methylene blue from wastewaters, Journal of Cleaner Production (2018), doi: https://doi.org/10.1016/ j.jclepro.2018.09.265. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT

Synthesis of porous biomass fly ash-based geopolymer spheres for efficient removal of methylene blue from wastewaters

Pullar a, João A. Labrincha a a

RI PT

Rui M. Novais a,*, João Carvalheiras a, David M. Tobaldi a, Maria P. Seabra a, Robert C.

Department of Materials and Ceramic Engineering / CICECO- Aveiro Institute of

M AN U

Portugal

SC

Materials, University of Aveiro, Campus Universitário de Santiago, 3810-193 Aveiro,

*Corresponding author: Tel.: +351234370262; fax: +351234370204

AC C

EP

TE D

E-mail address: [email protected] (Rui M. Novais)

1

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT Abstract In this work, and for the first time, fly ash (FA)-based geopolymer (d = 2.6 mm) spheres were used to extract methylene blue from synthetic wastewaters. The influence of sorption time, dye initial concentration and adsorbent amount on the dye removal

RI PT

efficiency and uptake by the porous spheres was evaluated. The adsorbents’ recyclability and their dye fixation efficiency were also considered. The initial dye concentration strongly affected the uptake and removal efficiency by the porous bodies,

SC

the former rising from 1.1 to 30.1 mg/g when the dye initial concentration jumped from 10 to 250 ppm, and the latter increasing from 82.3% to 94.3 % when the dye initial

M AN U

concentration varied from 10 to 125 ppm. Results showed a much faster (24 h) and higher (30.1 mg/g) methylene blue uptake in comparison with the other bulk-type geopolymers reported to date (30 h; 15.4 mg/g). The cumulative methylene blue uptake shown by these innovative spheres (79.7 mg/g) surpasses all other powdered

TE D

geopolymer adsorbents, being among the highest values ever reported for geopolymers. The adsorbent was successfully regenerated and reused eight times. Regeneration was found to negatively affect the MB uptake, but nevertheless, even after eight regeneration

EP

cycles a very high MB removal efficiency (83 %) was maintained. The use of these bulk-type waste-based geopolymer adsorbents is a low-cost, more eco-friendly, safer

AC C

and easier alternative to the use of powdered adsorbents in wastewater treatment systems, since these ~3 mm spheres may be used directly in packed beds, and were produced using significant amounts of waste material.

Keywords: adsorption; inorganic polymer; geopolymer; porosity; dye; waste.

1

ACCEPTED MANUSCRIPT 1. Introduction The removal of dyes from industrial wastewaters is of the utmost importance, not only because most of them are toxic and show carcinogenic properties, but also because treated industrial wastewaters may provide a vital source of clean water to mitigate the

RI PT

most pressing concern of our society – water scarcity. In the next 30 years, around 40% of the world’s population will live in areas with extreme water scarcity (OECD, 2012). Effective and low cost technologies for wastewater treatment are being eagerly pursued.

SC

Adsorption is considered the most effective, simple and universal technique for water decontamination (Gupta et al., 2012), activated carbon being the benchmark adsorbent

M AN U

material (Mestre et al., 2014). Despite the unique adsorption capacity of this material, its high production cost (Rafatullah et al., 2010) has led to extensive research for lower cost alternatives. One exciting approach could be the use of low cost and eco-friendly inorganic polymers (also known as geopolymers). Inorganic polymers are synthesised at

TE D

near-ambient temperatures (Novais et al., 2016b) by chemical activation of aluminosilicate sources, such as metakaolin (Zhu et al., 2018), fly ash (Novais et al., 2016c) and waste glass (Novais et al., 2016a). They have a negatively charged

EP

aluminosilicate network, balanced by cations such as sodium or potassium, which may in turn be exchanged with cations is solution. This feature suggests the feasibility of

AC C

using inorganic polymers as dye or heavy metal adsorbent materials. Despite this, the study of the use of geopolymers as adsorbents is fairly recent (Minelli et al., 2018; Novais et al., 2016d). Moreover, the vast majority of the investigations are focussed on the use of powdered adsorbents (Falah et al., 2016; Liu et al., 2016), which cannot be easily recovered, and so cannot be used in field applications or directly in packed beds. Powdered adsorbents may require the use of support materials (e.g. porous ceramics, polymer foams) (Zhang et al., 2016) to allow their industrial application or a separation

2

ACCEPTED MANUSCRIPT step (e.g. pressure filtration) after wastewater treatment process, both being detrimental to the wastewater treatment cost, besides increasing the process complexity. Recently, the use of porous geopolymer monoliths for methylene blue (MB) extraction from wastewaters was reported by the authors (Novais et al., 2018a). The MB removal

RI PT

capacity of the monolithic bodies reached 15.4 mg/g, while the adsorbent could be reused up to 5 times. These promising results demonstrated the feasibility of using porous bodies, not powders, to extract MB from polluted wastewaters. Nevertheless, the

SC

adsorbents’s removal efficiently dropped significantly, to around 65%, when the MB initial concentration reached 50 ppm (Novais et al., 2018a). Therefore, additional

M AN U

investigations addressing the use of bulk porous geopolymers should shed light on the most influential parameters affecting the dye adsorption by the geopolymers. One possibility to enhance the geopolymers’ MB adsorption capacity, in comparison with the use of cylindrical discs (thickness=3 mm; d=22 mm) (Novais et al., 2018a), could be

TE D

the use of porous geopolymer spheres (GS) (2-3 mm). The authors have recently reported the synthesis of waste-containing GS by a suspension-solidification approach (Novais et al., 2017), and their subsequent use as pH regulators in anaerobic digesters

EP

(Novais et al., 2018c), while other investigations have studied the use of GS as a heavy metal adsorbent (Ge et al., 2017) or as a photocatalytic material (Li et al., 2016).

AC C

Nevertheless, the use of GS as MB adsorbents has never been reported, up to now. This is the first ever investigation regarding the use of geopolymer spheres and formulations made from biomass fly ash (FA) waste as MB adsorbent materials from synthetic wastewaters. The FA waste used was obtained from a local Portuguese paper industry, thus enhancing the circular economy aspect of this work. These FA-based GS are expected to show much higher surface area than the previously reported geopolymer monoliths (Novais et al., 2018a) and, therefore, higher MB removal capacity. The

3

ACCEPTED MANUSCRIPT influence of contact time, MB initial concentration and adsorbent amount on the removal efficiency of this pollutant by the GS was studied. Desorption and regeneration tests were carried out to evaluate the MB extraction efficiency and the feasibility for multiple reuse of the adsorbent. The high removal efficiency shown by these innovative

RI PT

adsorbents, and their multiple recycling, demonstrates the adsorbents’ interesting potential for wastewater treatment systems. The present study is a significant step forward in comparison with previous investigations (Novais et al., 2018a), clearly

SC

demonstrating the performance advantage of the proposed solution (the use of

2. Experimental conditions

2.1. Materials

M AN U

geopolymer spheres instead of cylindrical discs) for wastewater treatment.

TE D

Metakaolin (Argical™ M1200S; Univar) and biomass FA waste were used as sources of reactive silica and alumina. The FA was sieved, and then only the portion below 63 µm was used in the geopolymer synthesis. The FA chemical composition, presented in

EP

Table 1, shows that the waste is a silica-rich material (34 wt.%). However, its low alumina content results in a SiO2/Al2O3 ratio ~ 2.5. For this reason, metakaolin was also

AC C

used in the compositions (50 wt.% regarding the aluminosilicate sources) to balance their molar ratios.

For the activation, 100 g of sodium silicate (Chem-Lab, Belgium) was mixed with 13.22 g of sodium hydroxide (ACS reagent, 97%; Sigma Aldrich), and this mixture was then used as the activator. Sodium dodecyl sulphate (pore forming agent), polyethylene glycol (medium used to promote fast curing) and MB (used as adsorbate) were all supplied by Sigma Aldrich.

4

ACCEPTED MANUSCRIPT 2.2. Geopolymer preparation The geopolymer slurry was synthesised using an approach described by the authors previously (Novais et al., 2017), in which 15 g of aluminosilicate precursors (50 wt.% metakaolin and 50 wt.% FA) were mechanically mixed with 24.38 g of alkaline

RI PT

solution, 4.15 g of water and 0.75 g of pore forming agent to produce the geopolymer slurry. Afterwards, the slurry was injected into a polyethylene glycol medium (~85 ºC), which promoted the solidification of the slurry and the formation of spheres. The FA-

SC

based GS floated in the polyethylene glycol medium, allowing their easy collection. This suspension-solidification approach has been previously reported (Novais et al.,

M AN U

2017). After collection, the spheres were then washed with distilled water to remove the excess of sodium hydroxide. Finally the specimens were cured: i) 24 h at 40 ºC and 65% relative humidity in a climatic chamber; and ii) 27 days at room temperature, before being used as MB adsorbents. The curing procedure was implemented following

TE D

previous works by the authors (Novais et al., 2018a; Novais et al., 2017).

2.3. Methylene blue adsorption tests

EP

Adsorption experiments were carried out to study the influence of contact time, MB initial concentration and adsorbent mass on the removal efficiency of this pollutant by

AC C

the FA-based GS. Various amounts (1.0–2.5 g) of spheres were added to 200 mL of a solution containing a specified dye concentration (10-250 ppm), and magnetically stirred for a predetermined period of time (1-26 h). Samples were collected from the solution at regular intervals, and the remaining dye concentration in the liquid evaluated by UV spectroscopy (Shimadzu UV-3100, JP) by measuring the absorbance at a λ = 664 nm.

5

ACCEPTED MANUSCRIPT The MB removal by the GS was evaluated by calculating the uptake ( ) and the removal efficiency percentage (E), respectively, using equations (1) and (2):  =

(  )

  

(1)

× 100

(2)

RI PT

E (%) =

×

where  quantifies the MB uptake by the GS in mg MB / g geopolymer, 

SC

corresponds to the dye initial concentration (mg/L),  is the MB equilibrium concentration (mg/L), the solution volume and  the adsorbent mass.

M AN U

Although there are many theories describing the adsorption equilibrium, those of the Langmuir and Freundlich isotherms are the most frequently adopted. The Langmuir isotherm theory foresees monolayer coverage of adsorbate over a homogenous adsorbent surface. Once the equilibrium is reached, a saturation point is attained and no

TE D

further adsorption can happen. Sorption is presumed to take place at specific homogeneous sites within the adsorbent, this meaning that, once a dye molecule occupies a site, no additional adsorption is possible at that specific site (Allen et al.,

  

AC C

 =

EP

2003):

 

(3)

In equation (3) – a non-linear equation – KL (L/mg) and qm (mg/g) are the Langmuir isotherm constants. Those can be obtained by making equation (3) linear; this will result in four different types (I, II, III and IV) of linear equations (Kumar and Sivanesan, 2007). The different linearised forms of Langmuir equation are depicted in Table 2. Once the KL value is obtained, the Langmuir isotherm can be expressed by a separation factor, RL, given by:

6

ACCEPTED MANUSCRIPT  =



(4)

 

Values of 1< RL < 0 represent favourable adsorption (Hajjaji et al., 2013).

The expression derived by Freundlich (Al Duri and Mckay, 1988), being an exponential

RI PT

equation [cf equation (5)], assumes that increasing the adsorbate concentration also increases the concentration of adsorbate on the adsorbent surface:

SC

!

 =   "

(5)

M AN U

where KF is the Freundlich constant, and n a parameter which represents the absence of linearity of the adsorbed quantity in function of Ce. Equation (5) is normally converted into an alternative linear form, thus becoming:



(6)

TE D

#$% = #$% + ' #$%

If 1 < n < 10, then there is favourable adsorption. On the contrary, larger values of n

EP

suggest a stronger interaction between the surface of the adsorbent and adsorbate. When 1/n is equal to 1, a linear adsorption has occurred, leading to identical adsorption

AC C

energies for all the sites (Febrianto et al., 2009).

2.4. Adsorbent regeneration tests

The feasibility of reusing the GS as adsorbents after initial MB adsorption was evaluated only for the specimens showing the highest MB removal capacity. Samples were heated at 400 ºC for 2 h to induce MB decomposition. Then the specimens were reused, and their adsorption ability evaluated using a 50 ppm MB solution. Eight regeneration cycles were implemented. 7

ACCEPTED MANUSCRIPT 2.5. Methylene blue desorption tests Desorption tests were performed using specimens after adsorption experiments. For this evaluation, GS which were in contact with a 50 ppm MB solution were selected. The influence of their amount (1.0–2.5 g) on the MB desorption capacity was also evaluated.

RI PT

After adsorption tests, the spheres were immersed in 200 mL of distilled water, and stirred for 24 h (section 3.2.1). Afterwards the leached MB concentration was measured

2.6. Materials characterisation and analysis

SC

by UV spectroscopy.

M AN U

The GS microstructure observation before and after MB adsorption, and after regeneration, was performed by scanning electron microscopy (SEM) using a Hitachi S4100 system coupled with energy dispersive spectroscopy (EDS - Rontec), while the GS morphology was observed using optical analysis (Leica EZ4HD microscope).

TE D

ImageJ was used to measure the spheres’ diameter and length, providing the GS size distribution. For this analysis, 120 spheres were evaluated and the average values presented.

EP

The zeta-potential of the FA-based GS was assessed with a Zetasizer Nano ZS (Malvern) at room temperature.

AC C

The aluminosilicate precursor’s chemical composition was determined by X-ray fluorescence (Philips X´Pert PRO MPD spectrometer). X-ray diffraction (XRD) was used to obtain information about the phase composition of the specimens (i.e. the amorphous and the crystalline amounts). For this purpose, full quantitative phase analyses (FQPA) were assessed using the combined Rietveld– reference intensity ratio (RIR) methods, as proposed by Gualtieri (2000) and Gualtieri and Brignoli (2004), using α-Al2O3 (NIST 676a) as an internal standard. This procedure

8

ACCEPTED MANUSCRIPT is described in detail in Novais et al. (2018a). XRD data for FQPA were recorded using a θ/2θ diffractometer (Panalytical Empyrean and X´Pert Pro3, NL), equipped with a linear PIXEL detector (PANalytical), with Cu Ka radiation(45 kV and 40 mA, 5–80 °2θ range, with a virtual step scan of 0.02 °2θ, and virtual time per step of 200 s). The

RI PT

Rietveld refinements were accomplished by means of the GSAS software suite, together with its graphical interface EXPGUI (Larson and Von Dreele, 2004; Toby, 2001).

Fourier-transform infrared spectroscopy (FT–IR) measurements of the adsorbent, before

SC

and after MB adsorption, as well as after the regeneration tests, were carried out via a Bruker Tensor 27 spectrometer, in attenuated total reflectance (ATR) mode –

3. Results and discussion

M AN U

wavenumber range 4000–350 cm–1, 4 cm–1 in resolution, 256 scans.

3.1. Geopolymer spheres characterisation

TE D

Fig. 1a shows a representative optical micrograph of the FA-based GS, while Fig. 1b presents their size distribution measured by image analysis. The geopolymers have an elongated spheroidal shape, with average diameter and length being 2.6±0.2 mm and

EP

2.9±0.3 mm, respectively. These results demonstrate that the production method described here is effective and reproducible, leading to a narrow size distribution. The

AC C

mm size spheres can be directly used in packed beds and easily retrieved after exhaustion, this being a crucial advantage over powdered adsorbents. Fig. S1 (Supplementary Material) shows a high magnification SEM micrograph and the corresponding EDS map of the GS surface. As observed, a geopolymeric gel composed of silica and alumina was formed, which demonstrates that the alkali activation of the FA and the metakaolin has occurred. This observation is in line with other literature studies which have shown that these two aluminosilicate sources (FA and metakaolin)

9

ACCEPTED MANUSCRIPT can be used as precursors in the production of geopolymers (Novais et al., 2016b; Novais et al., 2016c). Fig. 2 presents SEM micrographs of the spheres’ exterior surface and interior pore microstructure. The spheres present a smooth and homogeneous surface with several

RI PT

small pores being visible, which could allow the MB diffusion into the specimens, thus enhancing the adsorption ability of the geopolymers. Fig. 2b-c show an extremely porous microstructure, in which most of the pores are closed. The micrographs clearly

SC

show the presence of small-sized open pores inside the larger-sized closed pores. This is extremely relevant, as it will increase the number of active sites available for

M AN U

adsorption.

Fig. 2 also includes the EDS analysis of the spheres’ exterior surface, and two plots corresponding to the spheres’ interior. The chemical composition changes significantly within the inner part of the spheres, which is attributed to the distinct chemical

TE D

composition of the aluminosilicate precursors (Table 1), and to the heterogeneous composition of the FA particles (Novais et al., 2016c). MK is mainly composed of SiO2 and Al2O3 (account for 93.80 wt.%), while these two oxides only account for 47.50

EP

wt.% in these FA wastes. FA contains several other elements, with CaO (16.50 wt.%), K2O (5.49 wt.%), Fe2O3 (4.95 wt.%), MgO (3.07 wt.%) and SO3 (2.77 wt.%) being the

AC C

most abundant.

Fig. 2d also shows a higher sodium content in the spheres’ exterior surface, in comparison with that observed in their inner part, suggesting that the free sodium remaining in the structure after geopolymerisation migrates/diffuses to the spheres’ surface upon curing. These alkalis will be available for leaching, and are expected to significantly increase the pH of the MB-containing wastewater when immersed for long periods (Novais et al., 2018a). Considering that MB removal efficiency is known to be

10

ACCEPTED MANUSCRIPT affected by the solution pH (Qada et al., 2006), with higher pH values favouring adsorption, the expected pH increase will further promote MB uptake by the porous GS. Fig. 3 presents the zeta potential evolution with pH variation for the GS. The adsorbent’s zeta potential is negative within the studied pH interval, reaching the

RI PT

highest charge density at pH of ~ 10.5. The zeta potential then stabilises until pH values are close to 11.3, before increasing once more at higher pH values (~12). These results indicate that pH values between 10.5 and 11.3 will induce the maximum attraction

SC

between the adsorbent and the cationic adsorbate, and therefore promote higher MB removal efficiency.

M AN U

Results of FQPA analysis are reported in Table 3. An example of a graphic output of a Rietveld refinement is shown in Fig. S2. FA is composed of α–quartz (11.9 wt.%), calcite (14.1 wt.%), a mica group mineral (7.9 wt.%), microcline (14.8 wt.%), traces of anatase (0.1 wt.%), and 51.1 wt.% amorphous phase. On the other hand, specimen GS

TE D

contains a much greater amount of amorphous phase (82.5 wt.%), whilst the crystalline phases are: α–quartz (3.2 wt.%), calcite (1.7 wt.%), a mica group mineral (5.1 wt.%),

EP

microcline (7.4 wt.%), as well as traces of anatase (0.1 wt.%).

3.2. Methylene blue adsorption tests

AC C

3.2.1. Influence of contact time

The evolution with time of MB adsorption by the porous GS was evaluated using a constant GS amount (1.5 g; number of spheres = 123±3) and various MB initial concentrations. Considering the spheroidal shape of the spheres, the average volume per sphere is roughly 3.27 mm3, so the total volume of spheres used was ~402 mm3. Fig. 4 shows that in the first 60 min high MB removal occurs, with removal efficiency ranging from 62 to 77% depending on the  , while afterwards a gentler removal rate is

11

ACCEPTED MANUSCRIPT observed up to 24 h. At this point, the removal efficiency reaches 80 and 93% when the MB initial concentration is 10 and 100 ppm, respectively. Longer sorption times did not induce significant gains, and for that reason 24 h can be considered the equilibrium time.

RI PT

The equilibrium time reported here was compared with other literature studies performed using powdered and non-powdered geopolymers and activated carbons – Fig. 5. It should be highlighted that the equilibrium time is affected by several parameters

SC

such as the adsorbent concentration, the dye initial concentration and the adsorption temperature, this hindering the comparison between studies which have used various

M AN U

conditions. Anyway, results show that the equilibrium time here (24 h) is 20% smaller than the other bulk geopolymer adsorbent (30 h) reported to date (Novais et al., 2018a), demonstrating that the use of GS (d=2.6 mm) promotes a faster MB uptake by the porous geopolymers in comparison with the use of cylindrical discs (thickness=3 mm;

TE D

d=22 mm) (Novais et al., 2018a). This equilibrium time is even smaller than that reported for some powdered geopolymer (kaolin- (Yousef et al., 2009) and FA-based geopolymers (Li et al., 2006)) and activated carbon (waste rice straw) (Sangon et al.,

EP

2018) adsorbents. Not surprisingly, smaller equilibrium times have been reported for other powdered adsorbents, geopolymers (Falah et al., 2016; Khan et al., 2015) and

AC C

activated carbon (Li et al., 2018). Nevertheless, these powdered adsorbents cannot be used directly in packed beds, while the GS can be assembled without the need of any support material. Although being much larger, and therefore safer and easier to use in comparison with powdered adsorbents, the GS still show relatively fast MB sorption, further demonstrating their potential for wastewater treatment.

12

ACCEPTED MANUSCRIPT 3.2.2. Influence of MB initial concentration Fig. 6 presents the uptake and the removal efficiency of the porous GS as a function of  . The MB uptake sharply increases from 1.1 to 30.1 mg/g when the  jumped from 10 to 250 ppm. Interestingly, the removal efficiency increased from 82.3 to 94.3% when

RI PT

the  varied from 10 to 125 ppm, slightly decreasing (to 90.3%) for higher  values of 200 ppm and over. This behaviour differs from our previous investigation (Novais et al., 2018a), in which the removal efficiency dropped significantly with the increase in

SC

 . The trend observed for the removal efficiency can be explained considering the mass transfer resistance between the adsorbate and the adsorbent, which is affected by

M AN U

the  . Higher  provides an essential driving force to decrease the mass transfer resistance between the liquid and the solid part (Foo and Hameed, 2012; Zhang et al., 2011). Optical micrographs of the spheres’ microstructure, shown in Fig. 7, illustrate varying MB diffusion into the spheres depending on the  , with a more intense blue

TE D

colour observed for the GS with up to 175 ppm of MB. Fig. 7a shows that MB did not fully penetrate into the specimens’ centre when using a low  , which hinders the specimen’s maximum adsorption level. However, increasing the dye concentration

EP

decreases the mass transfer resistance, which allows a higher penetration into the

AC C

samples and, therefore, higher uptake values. Besides the distinct diffusion of the dye throughout the samples, the micrographs also show that adsorption occurs mainly on the specimen’s surface: an intense bluish hue is observed on the spheres’ surface, while a less intense blue colour is visible in the spheres’ interior. This finding suggests that the MB uptake by the GS could be further enhanced by increasing the porosity of the spheres surface and the connectivity of interior pores. The authors have recently demonstrated the possibility of tuning the superficial and interior porosity of red mudbased geopolymer spheres by changing the amount of pore forming agent (Novais et al.,

13

ACCEPTED MANUSCRIPT 2018b) or the nature of the binder (Novais et al., 2018c). Future work will address the optimisation of the spheres’ open porosity to further enhance their adsorption performance. After MB adsorption tests (1.5 g; 24 h) the solutions’ pH was measured to evaluate the

RI PT

impact of the alkalis leaching from the GS into the solution. The initial pH of the solution was dependent on the MB initial concentration, ranging from 7.4 to 5.6 when the  jumped from 10 to 250 ppm. After the tests, a substantial pH increase was

SC

observed for all compositions, the average value being ~10. Interestingly, this pH value is extremely close to that found to induce the highest charge density (10.5) for this

M AN U

innovative adsorbent (section 3.1), which explains the strong attraction between the cationic dye molecules and the negatively charged geopolymer framework. To further understand the MB adsorption by the porous GS, EDS maps from the spheres’ surface before and after MB adsorption tests were collected, and are shown in

TE D

Fig. 8. According to the XRF data, the FA contains around 3 wt.% of SO3, while this element is absent from MK composition. This explains the heterogeneous distribution of sulphur (identified in purple in the EDS map) on the spheres’ surface before

EP

adsorption. The EDS elemental mapping before adsorption also shows sodium agglomerates (identified in blue), suggesting the presence of unreacted alkalis. After

AC C

adsorption, the sodium distribution on the spheres’ surface is remarkably similar to that of the aluminium and silicon distribution (Fig. 8b). This observation suggests that the free sodium observed in the GS before adsorption has leached to the solution, while the remaining sodium in acting as a charge-balancing cation for the negatively charged aluminosilicate network, which explains the above-mentioned pH increase. After MB adsorption, the sulphur distribution also changes, with two clear features being distinguished: i) a homogeneous distribution associated with the geopolymer structure

14

ACCEPTED MANUSCRIPT (derived from the FA composition); and ii) the presence of small sulphur concentrations (not observed before) which were associated with the MB adsorption by the porous GS (MB molecular formula: C16H18ClN3S). FT-IR spectra are depicted in Fig. 9. They all show similar characteristics: the strong

RI PT

asymmetric band, having its maximum at around 3350 cm−1, is due to adsorbed molecular H2O, which also generates the feature at around 1630 cm−1, assigned to H– O–H bending (Tobaldi et al., 2010). The band located at ~980 cm−1 belongs to the

SC

asymmetric Al–O–Al/Si–O–Si stretching (Kumar et al., 2015). The band at ∼1440 cm−1 can be assigned to sodium carbonate, due to the reaction of residual sodium with

M AN U

atmospheric CO2 (Cheng-Yong et al., 2017).

After being in contact with MB (24 h), additional features appeared (cf the blue continuous line in Fig. 9). The band centred at ~1610 cm−1, belonging to the stretching vibration of aromatic rings, and features due to methyl and methylene bending

TE D

vibrations (1355 and 1335 cm−1, and ∼1410 cm−1, respectively) (Coates, 2006) occurred, providing clear evidence of MB adsorption.

EP

3.2.3. Influence of adsorbent concentration

AC C

The effect of the adsorbent amount on the MB uptake and removal efficiently by the GS was evaluated to determine the optimum adsorbent dosage, and results are shown in Fig. 10. The MB uptake and the removal efficiency were affected differently by the increase in the adsorbents amount, the uptake drops and the removal efficiency increases (exceptions to this general trend will be discussed below) when rising the GS amount. Increasing the adsorbents amount increases the number of available adsorption sites, therefore increasing the amount of adsorbed MB (i.e. removal efficiency). However, the amount adsorbed per unit of mass (i.e. uptake) decreases which is explained by the

15

ACCEPTED MANUSCRIPT presence of unsaturated adsorption sites (Aljeboree et al., 2017). The decrease in the MB uptake when rising the adsorbents amount has been previously reported (Liu et al., 2016). Results also showed that the MB removal efficiency was affected by the dye initial

RI PT

concentration: i)  = 10 ppm – the removal efficiency significantly dropped from 82 to 75% when the spheres content was above 1.5 g; ii)  = 50 ppm – the removal efficiency increased up to 2.0 g, and then slightly decreased (~1%) afterwards; iii)  =

SC

100 ppm – the removal efficiency increased steadily with the rise of the adsorbent amount, from ~88 up to ~96%, until a plateau was reached. These different trends are

M AN U

better illustrated in Fig. 11 where the removal efficiency versus contact time is shown. As depicted, the sorption kinetics are MB concentration-dependent, with higher  inducing a faster MB uptake, this being independent from the amount of adsorbents used. These differences were associated with the above-mentioned mass transfer

TE D

resistance. In any case, for the higher  values, an increase in the amount of spheres from 1.0 to 1.5 g enhanced the removal efficiency by around 5%. Increasing the quantity of spheres used to 2.0 and 2.5 g of spheres promoted even higher removal

EP

efficiencies, an increase of ~2% and ~3%, respectively, in comparison with the use of

AC C

1.5 g. Despite the slight gain in performance when using greater quantities of spheres, the use of 1.5 g of spheres seems to be the most cost-effective solution (E = 93%). This is the reason why 1.5 g of spheres was used as a proxy for the isotherm model studies, see below.

3.3. Isotherm model studies Linear regression results, for both the Freundlich and Langmuir isotherm equations, are reported in Fig. S3. As seen in Fig. S3, the linearisation resulted in a two-step trend,

16

ACCEPTED MANUSCRIPT with the turning point (for the system considered here, i.e., 1.5 g of GS as the adsorbent medium, MB as the solute, in 0.2 L of solution) being located at ~100 ppm of MB (Freundlich model) and 125 ppm of MB (Langmuir model). This is better visualised in Table 4, where the Langmuir (Type I to IV) and Freundlich

RI PT

constants are listed, considering MB concentration intervals of 10-100 and 125-250 ppm for the Freundlich isotherm model, and 10-125 and 125-250 ppm for the Langmuir model. As derived from Table 4, at relatively low MB concentrations (10-125 ppm), the

SC

data follows the Freundlich isotherm equation. The determination coefficient of the Freundlich model, at this concentration range, is 0.993; although the R2 of the Langmuir

M AN U

models (Types III and IV) are slightly better than that (i.e. 0.994), the Langmuir isotherm constants are all negative (for all of the four type of linear methods) – this being physically meaningless. On the contrary, when dealing with higher MB concentrations (i.e. ≥ 100 ppm), the Langmuir isotherm model became more applicable;

TE D

in particular the Type II has a very good R2 (0.996), with an adsorption capacity qm equal to 45.809 mg/g, this suggesting the prevalence of monolayer adsorption as the dominant sorption mechanism on that macroporous adsorbent (Rouquerol et al., 1998).

EP

This is to be expected, as transforming a non-linear equation into a linear form implies altering the error structure and the error variance (Allen et al., 2003). This might explain

AC C

the observations that linear Freundlich parameters gave isotherms which fitted the experimental data in a better way at relatively low concentrations (i.e. ≤ 100 ppm), whilst those for the Langmuir had the tendency to fit better at concentrations ≥ 125 ppm (Richter et al., 1989). These results are in good agreement with the experimental data discussed in section 3.2.3. As clearly shown by the optical micrographs (Fig. 7), at low concentration a heterogeneous MB adsorption is attained, consistent with the better fitting achieved with the Freundlich model; while at high  the MB adsorption is more

17

ACCEPTED MANUSCRIPT homogeneous, in particular at the spheres’ surface, this explaining the better fitting of the Langmuir model in this case. The differences in adsorption depending on the MB initial concentration were attributed to the decrease in the mass transfer resistance at

3.4. Geopolymer spheres recyclability

RI PT

high  , as discussed in section 3.2.2.

The feasibility of reusing the GS in multiple adsorption cycles was evaluated by using

SC

the spheres (1.5 g) that showed the highest MB uptake (30.1 mg/g) and results for the MB uptake by the GS are shown in Fig. 12. In the first adsorption cycle  was 250

M AN U

ppm. However, after the 1st regeneration cycle, the GS removal ability for such a high MB initial concentration was limited. For this reason, after 24 h adsorption the MB equilibrium concentration was still above the UV detection limit. Considering this performance, the following regeneration cycles were performed using a lower  of 50

TE D

ppm.

There are three possible explanations for this drop in performance: i) Modification of the adsorption kinetics due to the greater number of spheres used in

EP

the subsequent recycling tests. After the 1st adsorption test (24 h immersion in MB solution) the mass of the spheres was significantly reduced by ~33%, which was

AC C

associated with the loss of the alkalis released from the GS. Additionally, the 1st thermal treatment (section 2.5) further reduced the adsorbents’ weight by ~16%, associated to the extraction of physically adsorbed water, meaning that after the first adsorption/regeneration cycle the spheres had lost ~49% of their mass. Thus, to use the same amount of adsorbent (1.5 g), a significantly higher number of spheres had to be used (~241 spheres instead of 123), which altered the adsorption kinetics. Fig. 13 shows a lower sorption rate in the first couple of hours for the regenerated samples (1.5 g; 241

18

ACCEPTED MANUSCRIPT spheres) in comparison with the specimens that were solely immersed in a 50 ppm MB solution (without thermal regeneration; 1.5 g; 123 spheres). Nevertheless, after 24 h the removal efficiency between the samples (non- and regenerated) was similar, suggesting that the use of higher  for the regeneration tests could be possible, but that the

RI PT

equilibrium time would surpass 24 h.

ii) The lower pH attained after adsorption/regeneration cycles. As previously mentioned, the solution pH after the 1st adsorption test was ~10. However, the solutions’

SC

pH significantly dropped after subsequent regeneration cycles, from pH = 8.9 after the 1st, and down to pH = 7.0 after the 8th regeneration cycle. The reason for the pH drop

M AN U

when the number of regeneration cycles increases is related to the amount of alkalis available for leaching. As discussed above (see section 3.1), the geopolymers contain within their structure free alkalis which are available for leaching. However, as the number of regeneration cycles increases, the alkalis availability decreases, since most of

TE D

it has already leached into the solution, thus explaining the observed pH drop. This lower pH reduces the attraction between the cationic dye and the geopolymers’ negatively charged aluminosilicate network, thus negatively affecting the MB uptake.

EP

iii) Incomplete degradation of MB during the thermal regeneration. As demonstrated above (Fig. 7), MB adsorption occurs mainly at the spheres’ surface, and to a lower

AC C

degree in their inner part (when MB diffuses into the pores). During thermal regeneration, the MB adsorbed on the spheres’ surface is removed; however, the MB adsorbed inside the pores might not be completely removed, which would explain the performance drop. The FTIR spectrum of the regenerated sample (milled before the measurement) does not shown the presence of MB (Fig. 9), suggesting that the thermal treatment was effective. Nevertheless, it’s possible that the remaining MB moieties in the sample are below the equipment detection limit, and for that reason the possibility of

19

ACCEPTED MANUSCRIPT incomplete MB degradation cannot be ruled out. Future work will consider extending the thermal treatment to promote complete MB degradation. Furthermore, FQPA analysis of the regenerated specimen showed a decrease in the amorphous phase amount (from 82.5 to 79.4 wt.%, cf Table 3), likely due to the thermal treatment. This also

wt.% in GS to 4.2 wt.% in the regenerated specimen.

RI PT

favoured the on-going process of carbonatation, the calcite content increased from 1.7

The results shown in Fig. 12 clearly demonstrate the feasibility of reusing this

SC

innovative adsorbent in multiple MB adsorption cycles, this being a crucial advantage over benchmark activated carbons, whose recovery after use is extremely challenging.

M AN U

After the 1st regeneration cycle, seven other cycles (adsorption and thermal regeneration) were performed without any further performance compromise. At the end of the eight regeneration cycles, the GS maintained their integrity, as demonstrated by the SEM micrographs shown in Fig. 14, although an increase in porosity was observed.

TE D

A continuous and gradual mass reduction (reaching 12% after the 8th cycle) was observed throughout the tests. This means that after 9 cycles (1 adsorption + 8 adsorptions after regeneration) the sample losses ~61% of its initial mass (49% + 12%).

EP

Nevertheless, these results suggest that additional regeneration cycles could be performed. This will be considered in future work.

AC C

The cumulative MB uptake of the GS (79.7 mg/g), shown in Fig. 12, demonstrates the interesting potential of this bulk adsorbent for wastewater treatment systems. The removal efficiency throughout the nine cycles was always above 83% (results not shown by the sake of brevity), much higher than the other study addressing the use of bulk-type geopolymer adsorbent whose removal efficiency after the 1st adsorption test was only 55%, dropping to values below 50% when using the regenerated samples (up to 5 cycles) (Novais et al., 2018a).

20

ACCEPTED MANUSCRIPT The maximum (experimental) adsorption capacity observed for the GS (79.7 mg/g) was compared with that reported for other MB adsorbents, and results are shown in Table 5. The GS show very high MB removal capacity, surpassing all reported values for powdered geopolymers, such as those prepared from coal FA (Li et al., 2006; Liu et al.,

RI PT

2016), MK (Khan et al., 2015), Cu2O-modified geopolymers (Falah et al., 2016; Falah and Mackenzie, 2015), kaolin-based (Yousef et al., 2009), and acid treated zeolites (Hor et al., 2016), among others (Rida et al., 2013; Wang et al., 2015). The GS adsorption

SC

capacity is inferior to that previously reported by the authors when using geopolymer monoliths (Novais et al., 2018a). However, in this previous investigation the removal

M AN U

efficiency was 48.4% and the sorption time was 30 h (Novais et al., 2018a), while here a significantly higher efficiency (84.9%) with lower sorption time (24 h) was achieved. Moreover, the GS are synthesised through one-step, while two steps are required for the monoliths production (i) synthesis of the porous bodies; and ii) cutting of cylindrical

TE D

discs); thus, a simplified strategy is used here.

Table 5 also shows that the GS adsorption capacity is superior to that reported for Ficus carica bast (Pathania et al., 2017) and acrylic fibrous waste (Naeem et al., 2017)

EP

activated carbons, whilst being inferior to that of Fe3O4 nanoparticles (Ghaedi et al., 2015), and to palm shell (Wong et al., 2016) and waste rice straw (Sangon et al., 2018)

AC C

activated carbons. Nevertheless, the proposed strategy (use of mm size spheres instead of powders) may allow the adsorbent’s direct use in wastewater treatment facilities, and its easy recovery after exhaustion, while the recovery of nanoparticles or activated carbons after adsorption is a complex and expensive procedure. Furthermore, the GS were synthesised are near room temperature (24 h at 40 ºC), preventing the need for very high processing temperatures (700 °C) typically employed in the production of

21

ACCEPTED MANUSCRIPT activated carbons (Sangon et al., 2018), and thus being an environmental friendlier approach. The outstanding recyclability of the GS, assuring very high cumulative MB adsorption capacity, associated with the possibility of being directly used in wastewater systems

RI PT

without the need for support materials, makes this bulk-adsorbent an excellent and safer alternative to conventional powdered adsorbents for wastewater decontamination. Future work will evaluate the possibility of reusing the GS as aggregates in the

SC

production of eco-friendly mortars, after their exhaustion as adsorbent material, which would contribute towards the circular economy, besides decreasing the carbon footprint

M AN U

of the spheres.

3.5. Methylene blue desorption tests

To evaluate the MB fixation onto the GS desorption tests were performed in distilled

TE D

water. All specimens showed low leaching values, ranging from 0.21 to 0.57%, suggesting a strong fixation of the dye onto the spheres. The quantity of spheres used did not significantly affect the MB desorption level, although a slight decrease was

EP

observed when increasing the amount of spheres: 0.57% (1.0 g), 0.37% (1.5 g), 0.22% (2.0 g) and 0.21% (2.5 g). These results indicate that leaching in water medium is

AC C

ineffective, meaning that the adsorbent can be used for longer equilibrium times than those employed here (Section 3.2.1) to ensure the maximum MB removal (saturation of the spheres).

4. Conclusions In this investigation, and for the first time, FA-based geopolymer spheres were evaluated as a methylene blue adsorbent material. The MB uptake by the porous

22

ACCEPTED MANUSCRIPT geopolymer spheres is affected by the dye initial concentration, adsorbent amount and contact time. Results showed a faster (20% smaller equilibrium time) and higher MB uptake (twofold increase) in comparison with the other bulk-type geopolymer reported to date, demonstrating the performance advantage of using spheres (d = 2.6 mm) instead

RI PT

of cylindrical discs (thickness=3 mm; d=22 mm) for wastewater decontamination.

After adsorption tests, the geopolymer spheres could be regenerated and reused (up to 8 cycles) resulting in a very high cumulative MB uptake (79.7 mg/g) which surpasses all

SC

other powdered geopolymer adsorbents. These remarkable materials may be used directly in wastewater systems (e.g. packed beds), and are easily retrieved after

M AN U

exhaustion, these being crucial advantages over conventional powdered adsorbents. The proposed solution (use of mm size geopolymer spheres instead of powdered adsorbents) increases the simplicity of wastewater treatment systems, being a safer alternative to conventional powdered adsorbents, whose recovery after use is extremely challenging.

TE D

Moreover, this innovative adsorbent also promotes waste minimisation, since biomass FA wastes were used to partially replace (50 wt.%) commercial metakaolin in the geopolymer spheres’ synthesis, which is aligned with the need to reduce the

EP

consumption of virgin raw materials and prevent landfill disposal of wastes.

AC C

Acknowledgements

This work was developed within the scope of the project CICECO-Aveiro Institute of Materials, POCI-01-0145-FEDER-007679 (FCT Ref. UID /CTM /50011/2013), financed by national funds through the FCT/MEC and when appropriate co-financed by FEDER under the PT2020 Partnership Agreement. R.C. Pullar thanks the FCT for funding under grant IF/00681/2015.

23

ACCEPTED MANUSCRIPT References Albadarin et al., 2017 A.B. Albadarin, M.N. Collins, M. Naushad, S. Shirazian, G. Walker, C. Mangwandi Activated lignin-chitosan extruded blends for efficient

RI PT

adsorption of methylene blue Chemical Engineering Journal, 307 (2017), pp. 264-272 Al Duri and Mckay, 1988 B. Al Duri, G. Mckay Basic dye adsorption on carbon using a solid-phase diffusion model The Chemical Engineering Journal, 38 (1988), pp. 23-31

SC

Aljeboree et al., 2017 A.M. Aljeboree, A.N. Alshirifi, A.F. Alkaim Kinetic and equilibrium study for the adsorption of textile dyes on coconut shell activated carbon

M AN U

Arabian Journal of Chemistry, 10 (2017), S3381-S3393

Allen et al., 2003 S.J. Allen, Q. Gan, R. Matthews, P.A. Johnson Comparison of optimized isotherm models for basic dye adsorption by kudzu Bioresource Technology, 88 (2003), 143-152

TE D

Cheng-Yong et al., 2017 H. Cheng-Yong, L. Yun-Ming, M.M. Al Abdullah, K. Hussin Thermal resistance variations of fly ash geopolymers: foaming responses Scientific

EP

Reports, 7 (2017), 45355

Coates, 2006 J. Coates Interpretation of Infrared Spectra, A Practical Approach, in:

AC C

R.A. Meyers (Ed.) Encyclopedia of Analytical Chemistry. John Wiley & Sons Ltd, Chichester (2006), pp. 10815-10837, 2000 El Qada et al., 2006 E.N. El Qada, S.J. Allen, G.M. Walker Adsorption of methylene blue onto activated carbon produced from steam activated bituminous coal: a study of equilibrium adsorption isotherm Chemical Engineering Journal, 124 (2006), 103-110

24

ACCEPTED MANUSCRIPT Falah et al., 2015 M. Falah, K.J.D. Mackenzie, J.V. Hanna, S.J. Page Novel photoactive inorganic polymer composites of inorganic polymers with copper (I) oxide nanoparticles Journal of Materials Science, 50 (2015), 7374-7383

RI PT

Falah et al., 2016 M. Falah, K.J.D. Mackenzie, R. Knibbe, S.J. Page, J.V. Hanna New composites of nanoparticle Cu (I) oxide and titania in a novel inorganic polymer (geopolymer) matrix for destruction of dyes and hazardous organic pollutants Journal

SC

of Hazardous Materials, 318 (2016), 772-782

Falah and Mackenzie, 2015 M. Falah, K.J.D. Mackenzie Synthesis and properties of

M AN U

novel photoactive composites of P25 titanium dioxide and copper (I) oxide inorganic polymers Ceramics International, 41 (2015), 13702-13708

Febrianto et al., 2009 J. Febrianto, A.N. Kosasih, J. Sunarso, Y.-H. Ju, N. Indraswati, S. Ismadji Equilibrium and kinetic studies in adsorption of heavy metals using

616-645

TE D

biosorbent: a summary of recent studies Journal of Hazardous Materials, 162 (2009),

EP

Foo and Hameed, 2012 K.Y. Foo, B.H. Hameed Dynamic adsorption behaviour of methylene blue onto oil palm shell granular activated carbon prepared by microwave

AC C

heating Chemical Engineering Journal, 203 (2012), 81-87 Ge et al., 2017 Y. Ge, X. Cui, C. Liao, Z. Li Facile fabrication of green geopolymer/alginate hybrid spheres for efficient removal of Cu (II) in water: batch and column studies Chemical Engineering Journal, 311 (2017), 126-134 Ghaedi et al., 2015 M. Ghaedi, S. Hajjati, Z. Mahmudi, I. Tyagi, S. Agarwal, A. Maity, V.K. Gupta Modeling of competitive ultrasonic assisted removal of the dyes –

25

ACCEPTED MANUSCRIPT methylene blue and safranin-O using Fe3O4 nanoparticles Chemical Engineering Journal, 268 (2015), 28-37 Gualtieri, 2000 A.F. Gualtieri Accuracy of XRPD using combined Rietveld-RIR

RI PT

method Journal of Applied Crystallography, 33 (2000), 267-278 Gualtieri and Brignoli, 2004 A.F. Gualtieri, G. Brignoli Rapid and accurate quantitative phase analyses using a fast detector Journal of Applied Crystallography, 37 (2004), 8-

SC

13

Gupta et al., 2012 V.K. Gupta, I. Ali, T.A. Saleh, A. Nayak, S. Agarwal Chemical

M AN U

treatment technologies for waste-water recycling – an overview RSC Advances, 2 (2012), 6380-6388

Hajjaji et al., 2013 W. Hajjaji, S.O. Ganiyu, D.M. Tobaldi, S. Andrejkovičová, R.C. Pullar, F. Rocha, J.A. Labrincha Natural Portuguese clayey materials and derived TiO2-

TE D

containing composites used for decolouring methylene blue (MB) and orange II (OII) solutions Applied Clay Science, 83-84 (2013), 91-98

EP

Hor et al., 2016 K.Y. Hor, J.M.C. Chee, M.N. Chong, B. Jin, C. Saint, P.E. Poh, R. Aryal Evaluation of physicochemical methods in enhancing the adsorption

AC C

performance of natural zeolite as low-cost adsorbent of methylene blue from wastewater Journal of Cleaner Production, 118 (2016), 197-209 Khan et al., 2015 M.I. Khan, T.K. Min, K. Azizli, S. Sufian, Z. Man, H. Ullah Effective removal of methylene blue from water using phosphoric acid based geopolymers RSC Advances, 5 (2015), 61410-61420

26

ACCEPTED MANUSCRIPT Kumar and Sivanesan, 2007 K.V. Kumar, S. Sivanesan Isotherms for malachite green onto rubber wood (hevea brasiliensis) sawdust: comparison of linear and non-linear methods Dyes and Pigments 72 (2007), 124-129

RI PT

Kumar et al., 2015 S. Kumar, F. Kristály, G. Mucsi Geopolymerisation behaviour of size fractioned fly ash Advanced Powder Technology, 26 (2015), 24-30

Larson and Von Dreele, 2004 A.C. Larson, R.B. Von Dreele General Structure

SC

Analysis System (GSAS) Los Alamos National Laboratory Report LAUR (2004), 86748

M AN U

Li et al., 2016 C.-m. Li, Y. He, Q. Tang, K.-t. Wang, X.-m. Cui Study of the preparation of CdS on the surface of geopolymer spheres and photocatalyst performance Materials Chemistry and Physics, 178 (2016), 204-2010 Li et al., 2006 L. Li, S. Wang, Z. Zhu Geopolymeric adsorbents from fly ash for dye

52-59

TE D

removal from aqueous solutions Journal of Colloid and Interface Science, 300 (2006),

EP

Liu et al., 2016 Y. Liu, C. Yan, Z. Zhang, Y. Gong, H. Wang, X. Qiu A facile method for preparation of floatable and permeable fly ash-based geopolymer block Materials

AC C

Letters, 185 (2016), 370-373

Li et al., 2018 Z. Li, G. Wang, K. Zhai, C. He, Q. Li, P. Guo Methylene blue adsorption from aqueous solution by loofah sponge-based porous carbons Colloids and Surfaces A, 538 (2018), 28-35 Mestre et al., 2014 A.S. Mestre, R.A. Pires, I. Aroso, E.M. Fernandes, M.L. Pinto, R.L. Reis, M.A. Andrade, J. Pires, S.P. Silva, A.P. Carvalho Activated carbons prepared

27

ACCEPTED MANUSCRIPT from industrial pre-treated cork: sustainable adsorbents for pharmaceutical compounds removal Chemical Engineering Journal, 253 (2014), 408-417 Minelli et al., 2018 M. Minelli, E. Papa, V. Medri, F. Miccio, P. Benito, F. Doghieri, E.

RI PT

Landi Characterization of novel geopolymer - zeolite composites as solid adsorbents for CO2 capture Chemical Engineering Journal, 341 (2018), 505-515

Naeem et al., 2017 S. Naeem, V. Baheti, J. Wiener, J. Marek Removal of methylene

SC

blue from aqueous media using activated carbon web The Journal of the Textile Institute, 108 (2017), 803-811

M AN U

Nasrullah et al., 2018 A. Nasrullah, A.H. Bhat, A. Naeem, M.H. Isa, M. Danish High surface area mesoporous activated carbon-alginate beds for efficient removal of methylene blue International Journal of Biological Macromolecules, 107 (2018), 17921799

TE D

Novais et al., 2016a R.M. Novais, G. Ascensão, M.P. Seabra, J.A. Labrincha Waste glass from end-of-life fluorescent lamps as raw materials in geopolymers Waste

EP

Management, 52 (2016), 245-255

Novais et al., 2018a R.M. Novais, G. Ascensão, D.M. Tobaldi, M.P. Seabra, J.A.

AC C

Labrincha Biomass fly ash geopolymer monoliths for effective methylene blue removal from wastewaters Journal of Cleaner Production, 171 (2018), 783-794 Novais et al., 2016b R.M. Novais, L.H. Buruberri, G. Ascensão, M.P. Seabra, J.A. Labrincha Porous biomass fly ash-based geopolymers with tailored thermal conductivity Journal of Cleaner Production, 119 (2016), 99-107.

28

ACCEPTED MANUSCRIPT Novais et al., 2016c R.M. Novais, L.H. Buruberri, M.P. Seabra, D. Bajare, J.A. Labrincha Novel porous fly ash-containing geopolymers for pH buffering applications Journal of Cleaner Production, 124 (2016), 345-404

RI PT

Novais et al., 2016d R.M. Novais, L.H. Buruberri, M.P. Seabra, J.A. Labrincha Novel porous fly ash-containing geopolymer monoliths for lead adsorption from wastewaters Journal of Hazardous Materials, 318 (2016), 631-640

SC

Novais et al., 2018b R.M. Novais, J. Carvalheiras, M.P. Seabra, R.C. Pullar, J.A. Labrincha Innovative application for bauxite residue: red mud-based inorganic polymer

M AN U

spheres as pH regulators Journal of Hazardous Materials 358C (2018) 69-81 Novais et al., 2018c R.M. Novais, T. Gameiro, J. Carvalheiras, M.P. Seabra, L.A.C. Tarelho, J.A. Labrincha, I. Capela High pH buffer capacity biomass fly ash-based geopolymer spheres to boost methane yield in anaerobic digestion Journal of Cleaner

TE D

Production, 178 (2018), 258-267

Novais et al., 2017 R.M. Novais, M.P. Seabra, J.A. Labrincha Geopolymer spheres as

EP

novel pH buffering materials Journal of Cleaner Production, 143 (2017), 1114-1122 OECD (Organisation for Economic Co-operation and Development), 2012 OECD

AC C

(Organisation for Economic Co-operation and Development) OECD Environmental Outlook to 2050: the consequences of inaction, OECD Publishing, Paris (2012), dx.doi.org/10.1787/9789264122246-en Pathania et al., 2017 D. Pathania, S. Sharma, P. Singh Removal of methylene blue by adsorption onto activated carbon developed from Ficus carica bast Arabian Journal of Chemistry, 10 (2017), S1445-S1451. doi:10.1016/j.arabjc.2013.04.021

29

ACCEPTED MANUSCRIPT Rafatullah et al., 2010 M. Rafatullah, O. Sulaiman, R. Hashim, A. Ahmad Adsorption of methylene blue on low-cost adsorbents: a review Journal of Hazardous Materials, 177 (2010), 70-80

RI PT

Richter et al., 1989 E. Richter, W. Schütz, A.L. Myers Effect of adsorption equation on prediction of multicomponent adsorption equilibria by the ideal adsorbed solution theory Chemical Engineering Science, 44 (1989), 1609-1616

SC

Rida et al., 2013 K. Rida, S. Bouraoui, S. Hadnine Adsorption of methylene blue from aqueous solution by kaolin and zeolite Applied Clay Science, 83-84 (2013), 99-105

M AN U

Rouquerol et al., 1998 F. Rouquerol, J. Rouquerol, K.S.W. Sing Adsorption by powders and porous solids, principles, methodology and applications, 1st edition. Academic Press, New York (1998), ISBN: 978-0-08-052601-0 Sangon et al., 2018 S. Sangon, A.J. Hunt, T.M. Attard, P. Mengchang, Y. Ngernyen, N.

TE D

Supanchaiyamat Valorisation of waste rice straw for the production of highly effective carbon based adsorbents for dyes removal Journal of Cleaner Production, 172 (2018),

EP

1128-1139

Tobaldi et al., 2010 D.M. Tobaldi, A. Tucci, A.S. Škapin, L. Esposito Effects of SiO2

AC C

addition on TiO2 crystal structure and photocatalytic activity Journal of the European Ceramic Society, 30 (2010), 2481-2490 Toby, 2001 B.H. Toby EXPGUI, a graphical user interface for GSAS Journal of Applied Crystallography, 34 (2001), 210-213 Wang et al., 2015 S. Wang, Y. Boyjoo, A.A. Choueib A comparative study of dye removal using fly ash treated by different methods Chemosphere, 60 (2015), 14011407 30

ACCEPTED MANUSCRIPT Wong et al., 2016 K.T. Wong, N.C. Eu, S. Ibrahim, H. Kim, Y. Yoon, M. Jang Recyclable magnetite-loaded palm shell-waste based activated carbon for the effective removal of methylene blue from aqueous solution Journal of Cleaner Production, 115

RI PT

(2016), 337-342. doi:10.1016/j.jclepro.2015.12.063 Yousef et al., 2009 R.I. Yousef, B. El-Eswed, M. Alshaaer, F. Khalili, H. Khoury The influence of using Jordanian natural zeolite on the adsorption, physical, and mechanical properties of geopolymers products Journal of Hazardous Materials, 165 (2009), 379-

SC

387

M AN U

Zhang et al., 2011 W. Zhang, L. Dong, H. Yan, H. Li, Z. Jiang, X. Kan, H. Yang, A. Li, R. Cheng Removal of methylene blue from aqueous solutions by straw based adsorbent in a fixed-bed column Chemical Engineering Journal, 173 (2011), 429-436 Zhang and Liu, 2013 Y. Zhang, L. Liu Fly ash-based geopolymer as a novel

TE D

photocatalyst for degradation of dye from wastewater Particuology, 11 (2013), 353-358 Zhang et al., 2016 Z. Zhang, L. Li, D. He, X. Ma, C. Yan, H. Wang Novel self-

EP

supporting zeolitic block with tunable porosity and crystallinity for water treatment Materials Letters, 178 (2016), 151-154.

AC C

Zhu et al., 2018 W. Zhu, X.H. Rao, Y. Liu, E.-H. Yang Lightweight aerated metakaolin-based geopolymer incorporating municipal solid waste incineration bottom ash as gas-forming agent Journal of Cleaner Production, 177 (2018), 775-781.

31

RI PT

ACCEPTED MANUSCRIPT

Fig. 1 a) Optical characterisation of the FA-based geopolymer spheres and their b)

SC

cumulative size distribution (measured using image analysis). Horizontal lines

AC C

EP

TE D

M AN U

corresponding to cumulative values of 10, 50 and 90 % were included.

32

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

Fig. 2 SEM micrographs of the FA-based geopolymer spheres: a) surface and (b, c) interior. Fig. 2d presents the EDS spectra of the geopolymer spheres surface and interior (at two different positions). 33

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

Fig. 3 Geopolymer spheres zeta potential variation with pH.

34

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

Fig. 4 Influence of contact time and methylene blue initial concentration on the MB removal efficiency by the FA-based geopolymer spheres (adsorbent dose: 1.5 g). Dotted

AC C

EP

TE D

lines are a guide for the eye.

35

SC

RI PT

ACCEPTED MANUSCRIPT

Fig. 5 Equilibrium time of various methylene blue adsorbents reported in literature.

M AN U

When available the size of the adsorbents was included above the columns (in blue

AC C

EP

TE D

colour) for comparison.

36

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

Fig. 6 Influence of MB initial concentration on the uptake and removal efficiency of

AC C

EP

TE D

this dye by the porous geopolymer spheres (adsorbent dose: 1.5 g; contact time: 24 h).

37

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

Fig. 7 Optical micrographs of the inner part of the geopolymer spheres after MB adsorption illustrating the influence of the MB initial concentration on the MB diffusion throughout the spheres (real colours are shown). MB initial concentration being: a) 10, b) 25, c) 75, d) 100, e) 175 and f) 200 ppm. 38

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

Fig. 8 EDS elemental mapping of FA-based geopolymer spheres’ surface a) before and b) after methylene blue adsorption tests.

39

ACCEPTED MANUSCRIPT

GS GS+MB GS Reg

80

1610

20

90 85 1335 80 1355 75 1800

0

1410

95

1700

1600

1500

1400 -1

4500

4000

3500

1300

1200

M AN U

Wavenumber (cm )

SC

40

RI PT

60

Transmittance (%)

Transmittance (%)

100

3000

2500

2000

1500 -1

1000

500

0

Wavenumber (cm )

Fig. 9 FTIR spectra of the geopolymer spheres before, after MB adsorption, (contact time = 24 h; adsorbent dose: 1.5 g;  = 200 ppm), as well as being regenerated.

AC C

EP

TE D

Note: the noise in the 2450–2000 cm–1 region is due to atmospheric CO2.

40

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

Fig. 10 Influence of the adsorbent amount on a) the MB uptake and b) removal efficiency by the FA-based geopolymer spheres for various

AC C

EP

concentrations of MB (contact time: 24 h).

41

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

Fig. 11 Influence of the adsorbent amount and MB initial concentration on the removal efficiency kinetics. 42

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

Fig. 12 Influence of the number of regeneration cycles on the MB uptake by the FA-

AC C

EP

TE D

based geopolymer spheres.

43

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

Fig. 13 Influence of the number of regeneration cycles on the MB removal efficiency by the geopolymer spheres. Note: the data in the first hours for the “1st adsorption (200

AC C

EP

TE D

ppm)” was above the UV-Vis spectrometer detection limit.

44

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

Fig. 14 SEM micrographs of the spheres’ surface (a) and inner part (b) after eight regeneration cycles illustrating the microstructural changes induced in the geopolymer

TE D

spheres by the multiple adsorption and thermal regeneration cycles. Fig. 14c presents

AC C

EP

the EDS spectrum of the geopolymer sphere surface.

45

ACCEPTED MANUSCRIPT Table 1 Chemical composition from XRF of metakaolin (MK) and fly ash (FA).

MK

FA

SiO2

54.40

34.00

TiO2

1.55

0.65

Al2O3

39.40

13.50

Fe2O3

1.75

4.95

MgO

0.14

3.07

CaO

0.10

16.50

MnO

0.01

0.45

Na2O

-

K 2O

1.03

SO3

-

SC 5.49 2.77

0.06

1.11

2.66

14.30

AC C

EP

TE D

LOI

1.52

M AN U

P2O5

RI PT

Oxides (wt.%)

46

ACCEPTED MANUSCRIPT

Plot

Type I

Ce/qe versus Ce

Type II

1/qe versus 1/Ce

Type III

qe versus qe/Ce

Type IV

qe/Ce versus qe

AC C

EP

TE D

M AN U

SC

Linear regression

RI PT

Table 2 Linearised forms of the Langmuir equation.

47

ACCEPTED MANUSCRIPT

Table 3 Refinement parameters and phase composition of the specimens.

Rwp (%)

χ2

α– quartz

calcite

18.08 13.16 11.14

4.62 3.55 3.43

3.19 2.21 1.86

11.9(2) 3.2(1) 4.7(1)

14.1(3) 1.7(1) 4.2(1)

mica

7.9(3) 5.1(1) 2.6(2)

microclin e

anatase

amorphou s

14.8(5) 7.4(3) 8.6(5)

0.1(1) 0.1(1) 0.5(1)

51.1(7) 82.5(3) 79.4(5)

TE D

M AN U

SC

R(F2) (%)

EP

FA GS Regenerated GS

Number of variables 38 44 44

AC C

Sample

Phase composition (wt.%)

RI PT

Agreement factors

48

ACCEPTED MANUSCRIPT Table 4 Isotherm constants for MB adsorption onto GS (1.5 g of adsorbent in 0.2 L MB solution). Type I

Type II

Type III

Type IV

Langmuir constants, MB concentration range: 10-125 ppm –4.028

–3.619

–4.070

KL (L/mg)

–0.110

–0.116

–0.110

R2

0.954

0.991

0.994

–4.131

RI PT

qm (mg/g)

–0.109 0.994

0.276

1/n

1.895

R2

0.993

M AN U

KF (mg/g)(L/g)n

SC

Freundlich constants, MB concentration range: 10-100 ppm

Langmuir constants, MB concentration range:125-250 ppm 45.434

45.809

43.562

46.960

KL (L/mg)

0.076

0.074

0.084

0.072

0.822

0.822

R2

TE D

qm (mg/g)

0.959

0.996

Freundlich constants, MB concentration range: 125-250 ppm

1/n

AC C

R2

6.227

EP

KF (mg/g)(L/g)n

0.495 0.946

49

ACCEPTED MANUSCRIPT

Table 5 Maximum methylene blue adsorption capacity of various adsorbents reported in literature. Material Coal fly ash geopolymer

a

(Zhang and Liu, 2013)

3.0

(Khan et al., 2015)

Not given

3.8

(Wang et al., 2015)

95.3

4.0-4.5

100

Coal fly ash Acid treated zeolite

a

8.0

(Wang et al., 2015)

Acrylic fibrous waste activated carbon

Not given

8.8

(Naeem et al., 2017)

Not given

14.8

(Falah et al., 2015)

Not given

14.8

(Falah and Mackenzie, 2015)

19.7

(Falah et al., 2016)

25.6

(Yousef et al., 2009)

Not given

33.5

(Rida et al., 2013)

88

36.3

(Albadarin et al., 2017)

38.4

(Li et al., 2006)

SC

Not given

M AN U

Cu2O /TiO2 geopolymer Cu2O /TiO2-CTAB geopolymer

Not given

Kaolin geopolymer

90

Zeolite Activated lignin-chitosan

80

Ficus carica bast activated carbon Coal fly ash geopolymer Fe3O4 nanoparticles

EP

Palm shell activated carbon

TE D

Coal fly ash geopolymer

Waste rice straw activated carbon

a

a

Not given

48

25 a

50.7

(Liu et al., 2016)

Not given

91.9

(Ghaedi et al., 2015)

Not given

163

(Wong et al., 2016)

Not given

528

(Sangon et al., 2018)

64.8

15.4 (1st adsorption)

48.4

109.0 (6th adsorption)

90.3

30.1 (1st adsorption)

84.9

79.7 (9th adsorption)

AC C

Biomass FA-geopolymer monoliths

FA-based geopolymer spheres a

(Hor et al., 2016)

Acid treated coal fly ash

Cu2O-geopolymer

Bulk-type

Reference

0.7

90

Phosphoric acid MK-based geopolymer

powder

qe (mg/g)

Removal efficiency (%)

RI PT

Adsorbent shape

(Pathania et al., 2017)

(Novais et al., 2018a)

This work

maximum removal efficiency values extrapolated from the experimental data

50

ACCEPTED MANUSCRIPT Highlights •

For the first time FA-based geopolymer spheres were evaluated as MB adsorbents. Highest ever reported MB uptake (30.1 mg/g) by bulk-type geopolymer

RI PT



adsorbents.

Bulk-type adsorbents are a safer and easier alternative to powdered adsorbents.



These novel adsorbents can be regenerated and reused in multiple adsorption

SC



cycles.

Cumulative MB uptake (79.7 mg/g) is among the highest ever reported for

M AN U



AC C

EP

TE D

geopolymers.