Surface composition and surface properties of water hyacinth ( Eichhornia crassipes ) root biomass : Effect of mineral acid and organic solvent treatment

The surface composition and surface properties of water hyacinth (Eichhornia crassipes) root biomass were studied before and after extraction with dilute nitric acid and toluene/ethanol (2/1, v/v) followed by ethanol, using Fourier Transform Infra-red (FT-IR) spectroscopy, thermogravimetric analysis, x-ray diffraction, scanning electron microscopy. FT-IR absorption bands were obtained at 3421, 2855, 1457 and 1035 cm -1 (O-H stretch, C-H vibration, C-H asymmetric deformation, and C-O stretch, respectively) and 1508, 1541 and 1559 cm -1 (all aromatic skeletal vibrations characteristic of lignin), as well as a C=O carboxylate stretch vibrational band at 1654 cm -1 . Scanning electron microscopy confirmed the root biomass to be amorphous and not to have a strongly structured surface. The dilute mineral acid and organic solvent treatment increased crystallinity. Thermogravimetric analysis Studies show that the treated biomass are more thermally stable than the untreated biomass. Data are presented showing that dilute mineral acid and organic solvent treatment resulted in a decrease in the amount of lignin in the biomass. The implications of the decrease in the percentage of lignin on the adsorption of volatile polar organic solvents and non-polar n-alkane hydrocarbons is discussed.


INTRODUCTION
The use of lignocellulosic biomaterials as adsorbents for various types of inorganic and organic water pollutants has been reviewed by Gupta et al. (2009), Mahamadi (2011), Hubbe et al. (2014), Priya et al. (2014) and Tran et al. (2015).The interest in bio-sorbents derives from the fact that biomaterials are environment friendly and are readily available, and therefore qualify as low-cost adsorbents.Adsorbent properties of lignocellulosic materials depend on the plant type from which they are derived, the conditions under which the plant grew, the origin of the fibre (that is, whether root, rhizome, stem or leaf), particle size, surface composition, and any physical *Corresponding author.E-mail: zaranyika@science.uz.ac.zwAuthor(s) agree that this article remains permanently open access under the terms of the Creative Commons Attribution License 4.0 International License or chemical pre-treatment the material is subjected to.Lignocellulosic biomaterials are composed of fibres that can be considered as naturally occurring composites, consisting mainly of helically wound cellulose microfibrils bound together by lignin and hemicelluloses.Cellulose is the most abundantly occurring natural polymer.Regardless of its source, cellulose is a polymer of -D glucose units.The monomer units are linked together by -(1-4)-glycosidic bonds forming straight chains.Each monomer has three hydroxyl groups.These chains are linked by hydrogen bonds as hydroxyl groups of the glucose units on one chain are able to form hydrogen bonds with an oxygen atom on the other chain.This enables the chains to be held together firmly to form microfibrils in which highly ordered (crystalline) regions exist with less ordered amorphous region (Celik and Demirbas, 2005).
Lignin is the second largest in abundance after cellulose.Unlike cellulose, lignin is a highly branched polyphenolic polymer that is three dimensional.It consists of three monomers namely p-coumary alcohol, coniferyl alcohol and sinapyl alcohols.These are found in lignin as phenylpropane units (Martone et al., 2009).Lignin is located in the spaces between cellulose and hemicelluloses and pectin components in plant cell wall, and gives the plant cell wall mechanical strength.The percentage of lignin differs from one plant to another.In soft wood and hardwood, it forms 23 to 33% and 16 to 25% respectively of the total mass, whereas it forms 26 to 33% of the total mass in plant biomass (Miretzkya and Cirelli, 2010).Lignins differ in terms of the degree of carbon-carbon crosslinking between phenyl groups (Celik and Demirbas, 2005).In hardwood it is covalently linked with xylans whereas in softwood it's linked with galactogluconannans.The chemical composition and physical composition of lignin depends on the method of isolation employed to separate it from the polysaccharide moieties.The common functional groups in lignin include aliphatic and phenolic hydroxyl groups (9 to 11%), methoxy groups (13 to 26%), and carbonyl groups (Miretzkya and Cirelli, 2010).
Hemicelluloses are short chain polymers of about 200 monosaccharide units composed of glucose, xylose, mannose, galactose, rhannose, and arabinose.Hemicelluloses and cellulose are bound to lignin by hydrogen bonds (Miretzkya and Cirelli, 2010).Lignocellulosic materials thus have carboxyl and hydroxyl functional groups that can be modified by chemical treatment to improve the selectivity of the adsorption (Bandosz, 2006).The available acidic and basic functional groups can be added or eliminated from the surface, thus governing the adsorption of either polar or non-polar species.
One plant that is receiving considerable attention as a potential low-cost bio-sorbent is water hyacinth (Eichhornia crassipes).Water hyacinth is a floating aquatic weed belonging to the pickerel weed family, Pontederiaceae.It is one of the fastest growing plants, and can tolerate a considerable variation in nutrients, temperature and pH (Rezania et al., 2015).It has flourishing roots composed mainly of cellulose and lignin (Zheng et al., 2009).It has been investigated extensively for the removal of dyes from aqueous solution (Tarawou et al., 2007;Low et al., 1995;Rajamohan, 2009;Aboul-Fetoula et al., 2010;Saltabas et al., 2012;Khan et al., 2012;El-Khalary, 2007;Modenes et al., 2013;Mahamadi and Mawere, 2013;Uddin et al., 2013;Rajamohan et al., 2013;Kaur et al., 2013;Nath et al., 2014).Water hyacinth has also been investigated for the removal of phenols from polluted water by Woverton and McKown (1976).The adsorbent properties of water hyacinth biomass were recently reviewed by Priya et al. (2014) and Mahamadi (2011), who concluded that water hyacinth biomass showed excellent biosorbent properties, and that further investigations were required on the structural properties of the material before and after modification.The aim of the present study was to evaluate the effect of dilute mineral acid and organic solvent treatment on the surface chemistry and surface properties of water hyacinth root biomass.This was achieved by studying the surface composition and surface properties of the ground dried water hyacinth root biomass powder before and after extraction with dilute mineral acid (nitric acid) and an organic solvent mixture (toluene/ethanol, 2/1v/v, followed by ethanol), using Fourier Transform Infra-red (FT-IR) spectroscopy, thermo gravimetric analysis (TGA), x-ray diffraction (XRD), scanning electron microscopy (SEM) and atomic absorption spectrometry (for elemental composition), in addition to standard chemical and physical methods for moisture, mineral matter (or ash), organic matter, organic carbon and lignin content determination.

Equipment
A DR-8001 Shimadzu FT-IR Spectrophotometer (Shimadzu Corporation, Japan) was used for infrared characterization of the water hyacinth root biomass.A Brucker D2 Phase diffractometer (Brucker (Pvt) Ltd, Karlsruhe, Germany) X-ray was used to record diffractograms (XRD) of ground dried water hyacinth roots.A Jeol 6510 (Jeol, USA) Scanning Electron Microscope was used to obtain the SEM micrograms.A 6200 flame atomic absorption spectrometer (Shimadzu Corporation, Japan) was used for measuring the elemental composition for the root biomass.Sieves, 150 and 212 m (BS410/1986, Endecotss Ltd, London, England) were used to sieve the water hyacinth root biomass after grinding.

Water hyacinth root biomass
Water hyacinth plants were collected from Waerera river, Bindura, Zimbabwe, at Universal Transverse Mercator (UTM) coordinate of (318500; 8078400).At the laboratory, the plants were washed with tap water several times, and then washed with distilled water.The roots were separated from the tops, cut into pieces, and air dried for several days.The root samples were then ground using a mortar and pestle, and then sieved first through the 212 m sieve, and

Solvent extraction
Samples of dried water hyacinth root biomass, 150 to 212 μm particle size, were successively extracted for 24 h in a Soxhlet apparatus, with toluene/ethanol (2/1,v/v) followed by ethanol (Tshabalala and Han, 1999).Each sample was filtered under suction, rinsed with boiling water, and then transferred to an Erlenmeyer flask containing boiling water.The flask was placed in a hot water bath and left to boil for approximately 1 h.The extracted sample was filtered under suction and air dried for 48 h.The dry sample was stored in a sealed glass jar (Tshabalala and Han, 1999).

Acid treatment
The water hyacinth root biomass powder was acid washed by soaking in 0.1 M HNO3 for 24 h, followed by washing with deionized water, and drying at 65°C for 24 h.

Mineral matter, organic matter and organic carbon determination
About 1.0 g of the oven-dried samples was ashed in a muffle furnace at 550°C for 8 h.The ash was weighed to determine the amount of ash or mineral matter in the sample.The percentage of organic matter and organic carbon was determined using the following equations.
Where, MM is mineral matter or ash, OM is organic matter, OC is organic carbon, AW is ash weight of the sample, DW is dry weight of the sample, and 1.724 is the van Bemmelen factor (that is, organic matter contains 58% OC) (Armecin and Gabon, 2008).The results obtained are shown in Table 1.

Lignin content determination
The amount of lignin in the untreated water hyacinth biomass and the treated water hyacinth root biomass was calculated from the remaining solid after the hydrolysis of the water hyacinth root biomass with 1.25% H2SO4 for 2 h followed by 72% H2SO4 for 4 h.The residues obtained were filtered and then washed with Distilled water in order to remove the sulphuric acid.The residues were then dried at 105°C to constant weight.The lignin in the samples was calculated using Equation 4 (Irfan et al., 2011).
The results obtained are shown in Table 2.

Elemental composition
Approximately 0.5 g of the ground water hyacinth root biomass were weighed into a 100 ml Teflon beaker and digested with a mixture of 1 ml of 65% perchloric acid, 5 ml of 55% nitric acid and 0.5 ml 98% sulphuric acid, all GR grade, Merck, Germany.Heating was continued until the volume was reduced to approximately 2 ml, before transferring to a 100 ml volumetric flask, and diluting to the mark with deionised water for analysis by flame atomic absorption spectrometry (Muramoto et al., 1989).The levels of Fe, Ni, Cu, Pb, Cr and Zn in the root biomass were determined before and after acid treatment.The results obtained are shown in Table 3.

FT-IR spectrophotometry
Transmission mode FT-IR spectra of the KBr pellets of ground water hyacinth root biomass were obtained within the 4000 to 400 cm -1 wave number range at 4 cm -1 resolution.50 scans were run for each sample and averaged.Figure 4 shows the FT-IR spectra obtained.

Thermal gravimetric analysis (TGA)
Thermal characteristics of the natural and treated samples were studied using TGA (STA 6000, Perkin Elmer, Massachusetts, USA) at a range of 34 to 900°C in a nitrogen atmosphere at a heating rate of 20°C/min.The data for derivative of the thermogravimetric analysis were also obtained from the same instrument and was used to plot derivative thermogravimetric analysis curves (DTA) (Figures 2 and 3).

X-Ray diffraction (XRD) and scanning electron microscopy (SEM) analysis
X-ray diffractograms (XRD) of ground dried water hyacinth roots were recorded in a Brucker D2 Phase diffractometer (Brucker (Pvt) Ltd, Karlsruhe, Germany) using 30 kV, 10 mA and CuK -radiation, a step size of 0.01°, and a step scan of 1.05 s were used for the entire reading range (10 to 65°). Figure 4 shows the XRD spectra obtained.The SEM micrograms were obtained using a Jeol 6510 (Jeol, USA) Scanning Electron Microscope.The samples were clamped in such a way that they were presented to the analysing beam.Figures 5 and 6 show the typical SEM microgram obtained.

Effect of mineral acid and organic solvent treatment on the chemical composition of water hyacinth root biomass
Table 3 shows the elemental composition of the roots of water hyacinth from Waerera River, Bindura, Zimbabwe, before and after mineral acid treatment.Data for untreated water hyacinth root biomass was found to be 2399, 172, 427, 177, 245, and 617 g/g for Fe, Ni, Cu, Pb, Cr and Zn, respectively.The data in Table 3 shows that up to 86.8, 78.5, 75.1, 75, 51.9 and 7% of Fe, Cr, Ni, Zn, Cu and Pb respectively can be removed by soaking the water hyacinth root powder in 0.1 M nitric acid for 24 h, followed by washing with deionized water, and oven drying at 65°C for 24 h.Thus, while acid washing does not remove any of the metals completely, Fe, Ni, Cu, Cr and Zn are removed to a greater extent than Pb, suggesting that the major portions of Fe, Ni, Cu, Cr and Zn are loosely bound in the root biomass while the greater portion of Pb is more strongly bound.The ash content of untreated, mineral acid treated and organic acid treated water hyacinth root biomass was found to be 18.43, 17.49 and 17.41 respectively (Table 2).Abdelhamid and Gabr (1991) reported an ash content of 27.7, Poddar et al. (1991) reported 16.29% andPatel et al. (1993) 20.2% for water hyacinth.An ash content of 7.91% was reported by Carvalho et al. (2015) for straw.The reduction in the ash content for the mineral acid treated biomass is in agreement with elemental analysis data for the acid treated sample which show that heavy metal content decreased after acid treatment (Table 3).Organic solvent treatment also caused a 5.5% reduction in the mineral content, most probably as a result of removal of extractable metal organic complexes in the water hyacinth root biomass.
The lignin content for the untreated, acid treated and organic solvent treated root biomass was found to be 23, 20 and 17% respectively.These values are within the range of those reported in literature for water hyacinth (Table 2).The lignin content of water hyacinth has been found to vary from 3 to 28% (Ahn et al., 2012), while land plants can have 20 to 40% hemicelluloses, 30 to 50 cellulose and 15 to 30% lignin (Bhattacharya and Kumar, 2010).Table 2 shows that organic solvent and mineral acid treatment lead to a reduction in the lignin content of 26 and 13% respectively.
Table 4 summarizes the peaks observed in the FT-IR spectra of untreated water hyacinth root biomass.The bands at 3691.62 and 3400 cm -1 are attributed to O-H  stretch vibration (Sundari and Ramesh, 2012;Parikh, 1974).The band at 2920.30 cm -1 is attributed to C-H vibration for alkanes (Liu et al., 2006).The band at 2361.04 cm -1 was probably due to P-H or C=N (Parikh, 1974).The root biomass also shows a band at 1654.09 cm -1 attributed to C=O stretch vibration of carboxylate (Tshabalala et al., 2005;Liu et al., 2006;Southichak et al., 2006).Sundari and Ramesh (2012) reported a similar peak on water hyacinth nanofibers.The bands between 1500 cm -1 and 1600 cm -1 are attributed to the C=C aromatic skeletal vibration of lignin (Cordeiro et al., 2011a).The bands at 1457.81 and 1035.94cm -1 are attributed to C-H deformation and C-O stretch respectively (Tshabalala et al., 2005).The presence of C-O stretch indicates the presence of alcoholic hydroxyl groups (Lim et al., 2008).FT-IR data thus suggest that the water hyacinth root biomass is mainly lignocellulosic in composition.
The FT-IR spectrum shows that untreated water hyacinth root biomass contains a greater percentage of cellulose than lignin or hemicelluloses.This is demonstrated by the fact that the most intense bands in the FT-IR of untreated water hyacinth root biomass are 3421.70,2920 and 1035.94cm -1 attributed to O-H and C-H stretch vibrations typical of cellulose (Cordeiro et al., 2011a).

FT-IR bands: Effect of acid washing and solvent extraction
The peak at 3691.62 cm -1 , O-H stretch, was present in the solvent extracted water hyacinth root biomass (c)     , while that for acid washed sample is shifted downwards to 3440.52 cm -1 . The band at 2920.30 cm -1 , C-H vibration, is also present in both samples, but with a slight high energy shift to 2920.30cm -1 .On the other hand, the band at 2361.04 cm -1 , was present in the solvent extracted sample, but absent in the acid washed sample.The loss of this peak in the acid washed sample was however accompanied by the appearance of a new peak at 2856.61 cm -1 .We attribute this new peak to C-H stretch (Parikh, 1974).The band at 1654.09 cm -1 , attributed to asymmetric and symmetric C-H stretch of the aliphatic groups (Parikh, 1974), is present in both mineral acid and organic solvent treated samples.The solvent extracted had a peak at 1637.08 cm -1 due to C=C and possibly adsorbed water.Peaks around that value are also due to C=O stretch vibration of carboxylate (Tshabalala et al, 2005;Liu et al., 2006;Southichak et al., 2006).The natural and the acid treated water hyacinth root biomass also showed the presence of the C=O stretch.Sundari and Ramesh (2012) reported a similar peak on the water hyacinth nanofibers (2012).Peaks between 1500 cm -1 and 1600 cm -1 are for the C=C aromatic skeletal vibration of lignin (Cordeiro et al., 2011a).These were observed in the natural biomass but not in the treated samples, suggesting that in the treated samples some lignin was removed by the treatment process.Peaks for C-H, as well as for C-O stretch, deformation was observed in all the samples (Tshabalala et al., 2005).The presence of C-O stretch indicates the presence of alcoholic hydroxyl groups (Lim et al., 2008).A summary of the peaks observed is shown in Table 4.

X-Ray diffraction and scanning electron microscopy analysis
In all the samples, Figure 4, there is a peak around the 22° reflection which corresponds to the (200) crystallographic plane (Tserki et al., 2005).A shoulder at around 16° is probably due to one of the 100 or (100 ) planes (Thiebaud and Borredon, 1995).For materials that have high cellulose content like cotton, flax or other fibers, one may observe two peaks around 16°, but in cases where the content of amorphous materials such as lignin, hemicelluloses and amorphous cellulose is high, the two peaks are smeared and may appear as one shoulder.Thus the appearance of one peak at 16° in all the XRD diffractograms for the water hyacinth root biomass samples, suggests the presence of a high content of such amorphous materials in untreated, mineral acid treated and organic acid treated water hyacinth root biomass.The relative intensities of the peaks for the treated samples are higher than for the untreated sample suggesting the treatment procedures increased the crystallinity of the biomass.This could be as a result of removal of some amorphous constituents of the root biomass during treatment.Zafeiropoulos et al. (2002) reported a similar increase in crystallinity after acetylation of flax fiber.Similarly, Cordeiro et al. (2011b) reported an increase in crystallinity for sisal fiber after alkaline treatment.
The SEM microgram, Figure 5, shows that the root powder does not have a thick epiticuticular wax layer and it does not have a strongly structured surface, in agreement with the XRD results.The organic solvent and   5 and 6).

Thermogravimetric analysis
The initial loss in mass in the TGA curves (Figure 2, 50 to 200°C) is due to loss of adsorbed moisture in the water hyacinth root biomass.The residues that remain in untreated and treated biomass after 500°C constitute carbonaceous materials.The treatment increased the stability of the biomass.From the DTA curves (Figure 3), the main decomposition temperatures for the untreated, acid treated and organic solvent treated were found to be 270, 282.96 and 301.14°C respectively.The lower decomposition temperature for the untreated biomass probably is due to the presence of higher content of pectin, lignin and hemicelluloses which have lower decomposition temperature (Chen et al., 2011).The decomposition of the polymer structure of lignin start at low temperature of 200 to 600°C (Brebu and Vasile, 2010;Kubo and Kadla, 2008).

Effect of surface composition on adsorbent properties
Recently, Mukaratirwa-Muchanyereyi et al. (2015) reported that the enthalpy of adsorption of volatile n-alkane hydrocarbons (n-hexane to n-decane) by water hyacinth root biomass increases following dilute mineral acid or organic solvent treatment, whereas that for volatile polar organic solvents (diethyl ether, dichloro-methane, acetone, ethyl acetate) increases after similar treatment.Figure 7 shows the correlation of enthalpy for the adsorption of the volatile polar solvents on water hyacinth root biomass and % lignin of the root biomass.All the plots show negative slope suggesting there is a positive correlation between enthalpy of adsorption and % lignin in the root biomass.Mukaratirwa-Muchanyereyi et al. (2015) attributed the decrease in the enthalpy of adsorption following mineral acid and organic solvent treatment to the removal of lignin and hemicelluloses on treatment.This interpretation is consistent with the trend in Figure 7. On the basis of these results we conclude that the adsorption of polar solvents by water hyacinth root biomass occurs mainly on lignin functional groups.For the adsorption of non-polar n-alkanes n-hexane, n-heptane and n-octane (Figure 8), the opposite trend is observed, suggesting negative correlation between enthalpy of adsorption and % lignin.The regression curves for n-hexane, n-heptane and n-octane show high linearity, with R 2 values of 0.8368, 0.8331 and 0.956 respectively.The increase in the enthalpy of following mineral acid and organic solvent treatment, suggests that the adsorption of the three nalkanes by water hyacinth root biomass occurs mainly on cellulose chains.N-nonane shows a very low negative correlation coefficient (R 2 = 0.0286), suggesting little or no correlation between enthalpy of adsorption and % lignin for n-nonane.

Conclusions
From the foregoing discussion we conclude that the surface of the water hyacinth root biomass is composed of several different functional groups which include aliphatic groups, carboxylates, aromatic and alcoholic moities.Dilute mineral acid and organic solvent treatment lead to, not only changes in the surface compsition, but also a reduction in the level of lignin in the biomass, with the concequential increase in crystallity and thermal stability of the biomass.

Figure 8 .
Figure 8. Plot of ∆Ha versus lignin percentage in the biomass a) Hexane b) Heptane c) Octane, d) Nonane.
a %DM b %MM c %OM d %OC e
then the 150 m sieve, retaining the 150 to 212 m fraction.Potassium bromide (AR grade, Sigma -Aldrich, Germany) was used to prepare KBr pellets for FT-IR analysis.

Table 3 .
Elemental analysis of untreated and acid treated biomass.

Table 4 .
Summary of FT-IR bands in untreated water hyacinth root biomass.