Keywords

1 Introduction

1.1 Background

There is a dire need for an economical alternative to convert sea water into fresh water owing to prevalent methods demanding high power and large-scale infrastructures [1]. In this regard, adsorption desalination is considered as an attractive choice due to its renewable and cost-effective nature [2,3,4]. Moreover, exploration of industrial wastes as potential adsorbents for purposes of water purification have been of interest in recent times [5, 6]. In this study, banana stem waste (plantain pith) has been exploited for its capacity to efficiently remove sodium chloride from aqueous solution (sea water). Banana is widely grown across the world, with an average production of 120–150 million tons per year. Yet, only 12% of the plant’s weight is the fruit, and the rest is inedible. Banana stem comprises of cellulose (Fig. 1), lignin and a small amount of ash [7] and has many hydroxyl functional groups on its surface thus making it a potential cheap renewable adsorbent for Na+ from water [8, 9]. However, removal of Na+ remains as a challenging task because of its high solubility in water [10]. In this study, banana stem waste was bleached first to extract the cellulose nanofibers and removing hemicelluloses, lignin and other functional groups, such as carbonyls and carboxylic acids that make it an interesting adsorbent [11].

Fig. 1
figure 1

Image downloaded from https://en.wikipedia.org/wiki/Cellulose in January 2018

Structure of cellulose.

Polymers have been experimented with for various desalination methods and good salt rejection was observed [12,13,14]. In particular, polymer hydrogels have been extensively developed into draw agents for forward osmosis desalination [15, 16] and as adsorbents for metal pollutants [17, 18]. The highly electronegative atoms, such as nitrogen and oxygen, present in polymers can bind to the metal ions via electron pair sharing to form a complex, therefore removing the metal ions from the solution [18]. In the current work, cellulose-based polymer hydrogels are synthesized to increase the adsorption capacity of the extracted cellulose. The hydrogels are synthesized by chemically crosslinking the polymers using glutaraldehyde [19] (Fig. 2). The hydrogen bonds within cellulose and the polymers were broken up to form acetal bonds with glutaraldehyde [19].

Fig. 2
figure 2

(Self-drawn) Mechanism of chemical crosslinking of polymers using glutaraldehyde

1.2 Objective

The aim of this project is to synthesize cellulose-based polymer hydrogels, via chemical crosslinking using glutaraldehyde, to increase the adsorption capacity of the extracted cellulose.

1.3 Scope of Work

The adsorbents were fully characterized by Scanning Electron Microscopy (SEM) and Fourier-Transform Infrared Spectroscopy (FTIR) while Na+ concentration of samples was determined using Inductively Coupled Plasma-Optical Emission Spectrometer (ICP-OES), model Perkin Elmer Optima 5300DV. Results of batch studies depict that modified bio-waste could serve as viable second for desalination process.

2 Methodology

2.1 Bleaching Banana Stem Waste

Banana stem waste was first mashed into smaller pieces to allow larger surface area for the bleaching agent to act upon. NaOCl was used as the bleaching agent, along with HCl in water. When mixed with a material which can be oxidized, NaOCl gives out nascent oxygen which then oxidizes any impurities. After about 18 h, the bleached banana stem (BBS) would appear white. Then, BBS was filtered out and rinsed with ultra-pure (UP) water until Na+ concentration of the solution is <0.10 ppm, determined from ICP-OES. The dried samples were then characterized using SEM and FTIR (Fig. 3).

Fig. 3
figure 3

(Self-drawn) Diagram depicting the process of bleaching banana stem waste

2.2 Dialysis of Polymers

This procedure was done to remove as much Na+ as possible from the impure solid polymers before use in the synthesis of the hydrogels. This decreases the number of washes needed to reduce Na+ concentration in the gels to <0.10 ppm (Fig. 4).

Fig. 4
figure 4

(Self-drawn) Diagram depicting the process of dialysis of polymers

First, 10 g of solid polymer was dissolved in 100 ml of UP water (heated at 200 °C and stirred at 500 rpm). The necessary length of the dialysis membrane (around the height of the container used for the dialysis) was cut out. One end of the membrane was then folded two to three times and sealed with a clip to make a tube, which was filled with the polymer solution until about two-thirds full. The tube was not fully filled as water would enter the tube during the dialysis process. After all air pockets in the tube were removed, the other end was also folded two to three times and sealed with a clip. The tube was then submerged in a container filled with UP water. One end of the tube was tied to the retort stand to ensure that it remains suspended in the water. The water was continuously stirred at 250 rpm, and changed every hour to increase the efficiency of the dialysis. Once the Na+ concentration of the polymer solution was <0.10 ppm, the polymer solutions in the tubes were transferred into 50 ml centrifuge tubes and left in a freezer at −80 °C to freeze-dry before lyophilization.

2.3 Synthesis of Hydrogels

Different combination of polymers, including bleached banana stem (BBS), polyvinyl alcohol (PVA), polyacrylic acid (PAA) and chitosan, were experimented with to form a hydrogel. The protocol is as follows (Fig. 5):

Fig. 5
figure 5

(Self-drawn) Diagram depicting the process of hydrogel synthesis

  1. 1.

    Lyophilized polymer was dissolved in UP water (heated when necessary). Note that chitosan powder was used as it is insoluble in water and was dissolved in an acidic solution instead of UP water.

  2. 2.

    All polymer solutions that were required for the particular combination were mixed and allowed to stir until the solution was homogenous.

  3. 3.

    BBS, if needed, was then added into the solution and allowed to stir until the mixture was homogenous.

  4. 4.

    A few drops of 50% glutaraldehyde in water was added dropwise into the mixture for cross-linking.

  5. 5.

    A few drops of concentrated hydrochloric acid were then added as a catalyst.

  6. 6.

    The mixture was left to evaporate while being stirred at 750 rpm until a gel formed.

  7. 7.

    UP water was added to the gel and replaced every 3 h until Na+ concentration in the gel is <0.10 ppm, determined from ICP-OES. This is to ensure that as much Na+ from the BBS is removed as possible before each batch adsorption studies (Table 1).

    Table 1 Amount of polymers used to synthesize the hydrogels

BBS only hydrogel could not be formed but instead remains as a thick white mixture. Therefore, BBS in its fibrous form was used as a control instead of a hydrogel. Other polymers were tried as well and were concluded to be unsuitable as adsorbents. The observations were as follows:

Polyvinyl Alcohol (PA): A pale yellow and relatively soft hydrogel, compared to PVA hydrogel, was formed. When added into water in a 5 ml centrifuge tube and left to shake at 150mot for a few minutes, the hydrogel began breaking off into smaller pieces. Therefore, it was not suitable as an adsorbent.

Polyethyleneimine (PEI): Glutaraldehyde reacts too quickly with PEI, resulting in patches of red hydrogel suspended in the solution while glutaraldehyde was added in dropwise. Glutaraldehyde was mixed with a small amount of water to dilute it before attempting the procedure again. While no red patches were formed immediately upon adding the diluted glutaraldehyde, a red and paste-like mixture was formed. A hydrogel could not be obtained therefore PEI was not suitable as an adsorbent.

Polyethylene Glycol (PEG): The PEG solution remained as a liquid even after adding a large amount of glutaraldehyde. Since a hydrogel could not be obtained, PEG was not suitable as an adsorbent.

Starch: The gel formed was relatively soft and, breaks off easily into smaller when added into water in a 5 ml centrifuge tube and left to shake at 150mot for a few minutes. Therefore, it is not suitable as an adsorbent.

2.4 Batch Adsorption Studies

The adsorption of NaCl onto the synthesized cellulose-based polymer hydrogels was studied by a series of batch adsorption experiments. The efficiency of the synthetic adsorbents for the removal of sodium chloride ions from aqueous solutions has been determined at the different adsorbent dosage (100, 300, 500 mg) in water (10 ml), fixed Na+ concentration (10 and 2500 ppm), agitation speed (150 rpm) and temperature (25 °C). The solutions after adsorption for a duration of 24 h were filtered and the Na+ concentrations quantified using ICP-OES.

3 Results and Discussion

3.1 Characterization

Raw and bleached banana stem were characterized via various techniques such as SEM and FTIR. After bleaching treatment, the banana stem is significantly smoother with less residues on it as shown in the SEM photographs (Fig. 6). The fibrous structure obtained is representative of the vast cellulosic network present in banana stem. The SEM image has shown that the bleaching agent has removed the non-cellulosic matter, which are dissolved and washed out with water during the treatment.

Fig. 6
figure 6

a SEM of raw banana stem b SEM of bleached banana stem

Difference between the FTIR spectra of raw and bleached banana stem can be observed below (Fig. 7). Different absorption bands characteristic for cellulose were observed in raw banana stem, which became less pronounced after bleaching. The broad band from 3000 to 3700 cm−1 is characteristic of –OH stretching. The sharp peak at around 2900 cm−1 is due to C–H symmetrical stretching and that at around 2800 cm−1 is due to aliphatic C–H stretching. The small peak at around 1750 cm−1 is assigned to C=O stretching vibration which disappeared in the BBS sample. The sharp peak at about 1600 cm−1 is representative of –OH bending of absorbed water. The peak at about 1400 cm−1 which disappeared in the BBS sample is representative of –HCH and –OCH in-plane bending vibration. The broad absorption band at 800–1200 cm−1 may be due to C–C vibration within the cellulose backbone [20, 21]. As cellulose does not contain any carbonyl groups, it can be concluded that most of the non-cellulosic organic matter has been washed away by bleach.

Fig. 7
figure 7

FTIR of raw banana stem and BBS

Photographs of a piece of the hydrogels and cellulose were taken (Fig. 8). It was observed that the hydrogel becomes whiter than transparent, and also more solid, as the BBS:PVA ratio increases.

Fig. 8
figure 8

a PVA hydrogel b BBS-PVA 1:10 hydrogel c BBS-PVA 1:5 hydrogel d BBS-PVA 1:2 hydrogel e BBS-PVA 1:1 hydrogel f BBS-PVA 2:1 hydrogel g BBS-PVA 5:1 hydrogel h BBS-PVA 10:1 hydrogel i BBS

3.2 Batch Adsorption Studies

Batch adsorption studies were conducted with initial Na+ concentration of 10 ppm (or 0.001%). The average adsorption capacity of BBS was 0.1027 mg/g which is not shown in the graph because the measured adsorption capacity of BBS fibers cannot be used for comparison with the hydrogels directly. This is because the hydrogels comprise water which makes the concentration of adsorbent different. To overcome this problem, we had lyophilized the synthesized hydrogels so that we could use its dry weight for a more accurate experiment, but the lyophilized hydrogel had an increased Na+ concentration which made them unsuitable for use. Another solution we attempted was to synthesize a BBS only hydrogel, which we failed to obtain as cellulose is insoluble in water. We are still looking into other experimental methods to quantify the adsorption capacity of BBS, on top of the method of using varying ratios of BBS and another polymer to estimate its adsorption capacity which we will discuss next.

The synthesized hydrogels shown in Fig. 9 do not show a general trend in their adsorption capacity with the increasing BBS:PVA ratio. Note that for every cellulose monomer, which has a molecular weight of 324.3 g/mol, there are five oxygen atoms while for every PVA monomer, which has a molecular weight of 44.05 g/mol, there is one oxygen atom. This means that for the same mass of BBS and PVA (we used synthetic PVA that was about 90% hydrolyzed), there would be around the same number oxygen atoms. Assuming that all the oxygen molecules in both BBS and PVA have equal attraction force for Na+ ions, we can approximate the adsorption capacity to be that of PVA. This also explains the results we obtained in Fig. 9 (Figs. 10 and 11).

Fig. 9
figure 9

Comparison of adsorption capacities of PVA and BBS-PVA hydrogels in a 10 ppm NaCl solution

Fig. 10
figure 10

Image downloaded from https://en.wikipedia.org/wiki/Polyvinyl_alcohol in May 2018

Structure of Polyvinyl Alcohol (PVA).

Fig. 11
figure 11

Comparison of adsorption capacities of PVA, PPA and chitosan hydrogels in a 2500 ppm NaCl solution

Batch adsorption studies were conducted with initial Na+ concentration of 2500 ppm (or 0.25%). The increase in Na+ concentration used was done to observe more significant differences between the adsorption capacities of the hydrogels and to minimize the possibility that any differences is due to chance.

PVA-PAA 1:1 hydrogel was synthesized instead of a PAA only hydrogel because PAA alone would remain as a viscous liquid rather than a gel. The addition of the highly viscous PVA solution could help solidify the PAA solution into a gel (Fig. 12).

Fig. 12
figure 12

Comparison of estimated adsorption capacities of hydrogels in a 10 ppm NaCl solution

By scaling, we can make a rough estimation of the differences between all the gels. The data shows that PVA-PAA 1:1 hydrogel has an average adsorption capacity of 12.2833 mg/g which is only slightly greater than that of all PVA hydrogels. This may be due to the fact that PAA has twice the number of oxygen atoms of PVA but its molecular weight (72.06 g/mol) is also about twice as much, therefore having similar adsorption capacities (Fig. 13).

Fig. 13
figure 13

Image downloaded from https://en.wikipedia.org/wiki/Polyacrylic_acid in May 2018

Structure of polyacrylic acid (PAA).

We also noted that chitosan has a similar adsorption capacity to all the other tried polymers. Nevertheless, the need to use acid to dissolve the chitosan powder may result in negative side effects such as degradation of BBS when we attempt to synthesize BBS-chitosan hydrogels. Hence, chitosan is highly unsuitable as an adsorbent.

The mechanism of adsorption is based on the interaction of electronegative atoms on the cellulose fiber and salt in water, as shown below cartoon (Fig. 14) and appears to be consistent with the data we have obtained.

Fig. 14
figure 14

(Self-drawn) Cartoonistic representation of the sodium complexation of hydrogels

4 Conclusion

Desalination via adsorption is a promising field of research but remains a challenging task due to the high solubility of Na+ in water. In this study, we experimented using cellulose-based polymer hydrogels as renewable adsorbents and concluded that BBS, PVA, PAA and chitosan show the equal potential as adsorbents of salt due to the similar number of electronegative atoms present in the polymer hydrogels which is in line with our hypothesized mechanism of sodium complexation in hydrogels. From our experiments, we can also conclude that PVA is the most suitable polymer to form a hydrogel with BBS as it is the most stable and can achieve similar adsorption capacities as the other tried synthetic polymers.

5 Future Work

We can look into increasing the concentration of the polymers or cellulose extracted from other bio-wastes for optimization. We can also be more selective in choosing the polymers to work with by looking at the structures now that we are more certain of the properties of the hydrogels. We can also try using other methods to synthesize the hydrogels. The hydrogels will be used for batch adsorption studies before proceeding with kinetic studies to decide whether cellulose in hydrogel form has any advantage over cellulose in solid form.