Synthesis and characterization of carboxymethyl cellulose from Musa paradisiaca and Tithonia diversifolia

Cellulose is the most abundant biomass in nature with properties that have enabled its application in different industrial processes. Its derivative, sodium carboxymethyl cellulose serves as an additive in food and non-food products such as desserts, detergents, paints etc. In this study, carboxymethyl cellulose (CMC) was synthesized from cellulose isolated from three ligno-cellulosic biomass, Tithonia diversifolia stalk (TDS), Musa parasidiaca stem (MPS) and unripe peel of Musa parasidiaca fruit (MPP). The isolation of cellulose was done by soda pulping and bleached using sodium hypochlorite, hydrogen peroxide, sodium hydroxide sequencing, followed by synthesis and purification of CMC. The physicochemical properties of the plant samples, isolated cellulose and bleached pulps including the synthesized CMC were determined. The effects of various processing stages on the properties of the cellulose and synthesized CMC were revealed in the study. CMC yield ranged from 62.57, 41.37 and 33.21% and the degree of substitution ranged from 0.33, 0.28 and 0.17 for TDS, MPS and MPP respectively. Further characterization of CMC using Fourier Transform Infrared (FTIR) confirmed the presence of major expected peaks that showed differences in terms of carboxymethyl substitution as compared to that of commercial CMC. The study revealed the potential of these plants for production of industrial grade CMC.


INTRODUCTION
Unutilized and underutilized agricultural plants and waste are a major environmental challenge in Nigeria because of pollution and its attendant health risk. However, most of these agricultural plants and waste are sources of lignocellulose biomass which can be converted to useful industrial raw materials Fang et al., 2016). With the current diversification policy of the government from the petroleum to the agricultural resources, vast tonnes of agricultural waste are disposed of indiscriminately and others are burnt leading to the destruction of critical infrastructures such as transformers, electricity poles, electricity sub-stations and also causing environmental pollution and its attendant health factors. Some of these unutilized and underutilized agricultural waste and plants include rice husk, plantain stem and peel, corn cob, sorghum stalk, groundnut husk, Mexican sunflower, etc. These biomasses are rich sources of lignocellulose materials which could be converted to wealth through various processing techniques into useful bio-based raw materials such as cellulose, gum, polymer, carboxymethyl cellulose etc.
Carboxymethyl cellulose is one of the major derivatives of cellulose and best renewable resource available to mankind that has received a lot of attention by researchers. Major known sources of cellulose for CMC production are wood and cotton, but researchers have discovered many other sources such as Palm kernel cake (Huang et al., 2017), Tithonia diversifolia , water hyacinth (Saputra et al., 2014), pod husk of cacao (Hutomo et al., 2012), banana waste (Arafat et al., 2008), Musa paradisiaca mid-rib (Ogunsile et al., 2006) and banana pseudo stem (Adinugraha et al., 2005). However, there are insufficient data on the use of unripe M. paradisiaca peel and fruit stem in CMC production. Most studies on M. paradisiaca peel waste have been centered on its nutritional and medicinal properties (Akubor and Ishiwu 2013;Mohammed and Saleha, 2011;Abbas et al., 2015;Akinsanmi et al., 2015).
Plantain (Musa paradisiaca) is an evergreen tropical monoherbacious plant that belongs to the musaceae family. It is reported that over 2.11 million metric tonnes of plantain are produced and consumed in Nigeria annually (Arisa et al., 2013) because it is regarded as a staple food by most families. This huge volume produced and consumed also results in the generation of large volume of waste from various parts of the plant such as the pseudo stem, fruit stem, peel and leaves. Adinugraha et al. (2005) synthesized CMC from banana pseudo stem with a degree of substitution of 0.75, viscosity of 4033 Cps, purity of 98.63% and a crystallinity of 38.33%. Plantain peel which is regarded as a waste very often discarded has been reported to be rich in minerals and vitamins (Arun et al., 2015). Several potential applications of plantain waste are in the production of biogas, local soap, starch, bio-plastics, etc (Uhuegbu and Onuorah, 2014;Rana et al., 2018;Akinyele and Agbro, 2007;Padam et al., 2014).
Tithonia diversifolia, commonly known as tree marigold or Mexican sunflower is an underutilized, herbaceous flowering plant of the Asteracea family. Though it is native to South and Central America but has now been naturalized in Asia and Africa. Researchers have reported of its several uses such as in animal feed, insecticide, poultry feed, compost and medicinal uses (Olayinka et al, 2015;Akanbi et al., 2007;Buragohain, 2016;Drechsel and Reck, 1998;Jama et al., 2000, Chukwuka andOjo, 2014;Bisht and Joshi, 2017;Tona et al., 1998). Recent studies have shown the potentials of its stalk as a source of valuable industrial chemicals for the production of lignin-base resins Friday and Muhammad, 2015) including Microcrystalline cellulose for drug formulation  and other derivatives for various industrial applications (Otoide et al., 2018).
As a result of the potentials of these agricultural plant biomasses as sources of industrial raw materials for the manufacturing sector, this study therefore aims to provide more insight on the physicochemical properties of cellulose and carboxymethylcellulose derivable from them. It will also serve as a guide for further research on these unutilized and underutilized agricultural plants and waste.

MATERIALS AND METHODS
Sunflower stalks were obtained from the Asokoro Military Cantonment area of Abuja, Nigeria in August 2016. Plantain stem and unripe peels were obtained from traders in a local market in Nasarawa State, Nigeria in October, 2016. The plant materials were authenticated at the Department of Crop, Soil and Pest Management, Federal University of Technology, Akure, Ondo State, Nigeria. Analytical grade chemical reagents used were sodium hydroxide (BDH), acetic acid (Sigma Aldrich), ethanol (95% BDH), NaClO 2 , Monochloroacetic acid.
Sunflower stalks were harvested 10 cm above ground level. The samples were cleaned to remove dirt and contaminants. The sunflower stalk and plantain stem were debarked using knife. All plant samples were then cut into chips of about 2-3 cm and sundried for 14 days. The dried samples were milled, screened using a 325 µm screen sieve, and stored in a ziploc polyethylene bag for subsequent analysis.

Pulping
Pulping experiment was conducted in a 25-litre thermostatically controlled autoclave digester following the methods of  and Hutomo et al. (2012). The plant samples (200 g each) were initially pretreated with 1 L of water to de-lignify the materials for 30 min at 110°C. The pretreated plant materials were further digested using the alkaline sulphite pulping method with an alkaline active charge of 18% (w/w) NaOH. The pulping conditions of the plant sample materials are as follows: liquor-to-plant ratio for all the cooking was 20:1 (v/w), temperature 110°C, pressure 15 psi and cooking time of 120 min. After digestion, the pulp obtained through filtration was thoroughly washed until free of residue alkali. The pulp yield was determined after oven drying at 105°C to constant weight gravimetrically as percentage of oven-dry raw materials.

Bleaching procedure
The bleaching process was conducted using the method described by  but with a slight modification. In the bleaching procedure, 20 g of the air dried samples were placed in a 2 L Erlenmeyer flask and 500 ml of 3.5%w/v (JIK) sodium hypochlorite and 3 ml of 90%v/v acetic acid were added. The flask was covered with a watch glass and the mixture heated in water bath at 70°C for 1 h with intermittent stirring. After 1 h treatment, the sample was drained and 500 ml of hydrogen peroxide added and heated in water at 60°C for 1 h with intermittent stirring. After treatment the sample was drained followed by extraction with 500 ml of 5%w/v NaOH conducted at 70°C for 1 h. The sample was washed free of alkali after extraction using distilled water. This process sequence was conducted thrice but the sample was not washed after the third time but rather 500 ml of 3.5%w/v (JIK) sodium hypochlorite, 3 ml acetic acid and 500 ml of hydrogen peroxide were added and allowed to stand undisturbed for another 1 h. The power source was put off after final 1 h and the experimental set up left for 24 h. The pulp was filtered and washed to obtain bleached cellulose pulp of pH 7 measured using a pH meter, oven-dried to constant weight and characterized using standard method.

Synthesis of sodium carboxymethyl cellulose
The synthesis of carboxymethyl cellulose was conducted according to the method described by Ambjornsson et al. (2013) but with a slight modification. A sample of 5 g of oven dried cellulose pulp was placed in a 500 ml flask with 64 mL of ethanol and 5.8 ml of distilled water and covered with aluminium foil to avoid evaporation. The flask was placed in an automated mechanical shaker rotating at 120 rpm and at a temperature of 20±5°C. After 15 min, a solution of 6.7 g of NaOH and 10.2 ml of distilled water was added to the mixture maintained under the mechanical agitation for 2 h. In the next step, a solution of 7.3 ml of 87% ethanol and 7.3 g of monchloroacetic acid was added to the reaction mixture and the temperature of the mechanical shaker gradually increased from 20 to 25°C within 30 min and then maintained for 2 h at 60°C. The reaction was terminated by neutralization with the addition of 20 ml of 90% (v/v) acetic acid. The suspension was filtered and the filtrate washed repeatedly with 200 ml of 87% ethanol and 200 ml of 70% methanol, and finally washed with 250 ml of absolute methanol to remove all sodium containing by-products (NaCl and C 2 H 3 NaO 3 ), until the filtrate gave negative response to silver nitrate test of chloride. The slurry obtained was suspended in acetone, stirred for 30 min and dried in an oven at 50±5°C for 12 h to constant weight. The percentage yield of carboxymethyl cellulose synthezised was calculated based on the weight of oven dried sample.

Determination of the properties of carboxymethyl cellulose
The properties such as ash and moisture contents were determined by T221om-93 and T550om-03, pulp viscosity (TAPPI T230om-99). pH by method described by JECFA (1998), bulk and tapped densities were determined by a modification of the method of Kumar and Kothari (1999), true density by method of Itiola (1991), swelling capacity by Iwuagwu and Okoli (1992), degree of substitution by ASTM (2005) and Sodium Chloride content by ASTM (1995)

Pulp viscosity
The kinematic viscosity of the CMC was determined using a modified capillary viscometer method (TAPPI T230om-99). The viscosity (centipoise) was calculated using the following formulae: Alabi et al. 11 Where [ᶯ] is the intrinsic viscosity (cP), ᶯ sp is the specific viscosity [ ] ᶯ solution is the product t solution x ρ solutiont, ᶯ solvent is t solvent x ρ solvent , and ᶯr is the relative viscosity (t solution /t solvent ), C is the concentration of the sample (1.052g/cm 3 ), ρ solution is the density of the solution (g/cm 3 ), ρ solvent is the density of the solvent (g/cm 3 ), t solution is the solution flow time (s) and t solvent is the solvent flow time (s).

Bulk density and tapped densities
The bulk and tapped densities of the CMC powder were determined by a modification of the method of Kumar and Kothari, 1991. The densities were calculated as follows,

Swelling capacity
The swelling capacity was determined according to the method described by Iwuagwu and Okoli, (1992) with a slight modification. The swelling capacity was calculated as: Where: S = Percentage Swelling capacity, Vs = Volume of swollen material, Vt = Tapped volume of sample material.

True density
The true density, (D t ) of the CMC was determined by the Pycnometer method using liquid displacement technique with xylene as the immersion fluid (Itiola, 1991) and the sample density calculated as follows: Where: w = weight of the sample, SG = specific gravity of the solvent (xylene), a = weight of the bottle + solvent, b = weight of the bottle + solvent + sample.

Degree of substitution
The standard method (ASTM, 2005) was used to determine the degree of substitution of the prepared CMC samples. The Degree of Substitution (DS) was calculated as follows: Where, A = milli-equivalent of consumed acid per gram of specimen, B = volume of Sodium hydroxide added, C = concentration in molarity of sodium hydroxide added, D = volume of consumed hydrochloric acid, E = concentration in molarity of hydrochloric acid used, F = weight of sample used (g),

Sodium chloride content
The sodium chloride content of the synthesized CMC was determined using the standard method of ASTM, 1995 and JECFA 1998 and the NaCl content was calculated as follows: Where: a = ml of the silver nitrate utilized, b = dry weight of the sample (g) The actual NaCl content was then obtained by subtracting the blank value from the sample value.

Instrumental analysis
Fourier Transform Infrared (FTIR) Spectroscopy was conducted using ThermoNicolet Avatar 370 FT-IR Spectrometer operating in the attenuated total reflection (ATR) mode (SmartPerformer, ZnSe crystal)

Statistical analysis
Data obtained in triplicate were analyzed using Duncan's Multiple Range Test (DMRT) and Analysis of Variance (ANOVA)

Physicochemical properties of raw materials
The results of the physicochemical properties of the plant raw materials samples are shown in Table 1 and sample of raw materials is shown in Figure 1a to c.

Holocellulose and alpha cellulose
The holocellulose content of the sample materials as shown in Table 1

Cold and hot water solubility
The cold water solubility ranged from 16.82 to 29.86%. M. paradisiaca (Stalk) has the highest solubility when compared to M. paradisiaca and T. diversifolia. However, the hot water solubility of M. paradisiaca (Stalk), M. paradisiaca (unripe peel) and T. diversifolia (stalk) was 37.70, 30.76 and 19.24 % respectively.

Ethanol-Benzene solubility
The result indicates that M. paradisiaca (unripe peel) had the highest content of 6.22% followed by T. diversifolia (stalk) 2.56% and M. paradisiaca (stalk) having the lowest value of 2.33%.

Physicochemical properties of pulp cellulose samples
The results of the physicochemical properties of the pulp samples are shown in Table 2 and pulp samples are shown in Figure 2a to c.

Ash and silica contents
The ash and silica contents of the cellulose pulp samples ranged from 6.71 to 11.19%, with Musa paradisiaca (stem) the lowest at 6.71% which was followed by T. diversifolia with 8.67% and M. paradisiaca (peel) the highest with 11.19%. The silica content ranged from 2.13 to 3.36% with T. diversifolia recording the lowest at

Kappa number
The kappa number is an index of lignin content (Solange et al., 2008). The result of the kappa number indicates that M. paradisiaca (peel) had the highest (40.21%) followed by T. diversifolia (29.30%) and M. paradisiaca (stalk) the lowest (24.61%).

Physicochemical properties of bleached pulp cellulose samples
The physicochemical properties of the bleached pulp cellulose samples are indicated in Table 3.

Yield
There was a general decrease in the yield of the bleached cellulose pulp compared to the unbleached pulp. The bleached pulp yield ranged from 27.11 -52.01% with T. diversifolia recording the highest and M. paradisiaca (peel) the lowest.

Ash and silica contents
The ash content of the bleached cellulose pulp ranged from 1.82 to 3.21% with M. paradisiaca (stalk) the lowest at 1.82%, followed by M. paradisiaca (peel) at 2.07% and T. diversifolia the highest at 3.21%. The result for the silica content showed M. paradisiaca (peel) having the lowest at 0.97%, followed by M. paradisiaca (stalk) at 1.14% and T. diversifolia the highest at 1.31%.

Kappa number
The kappa number ranged from 7.15 to 11.06% with T. diversifolia the lowest at 7.15%, followed by M. paradisiaca (stalk) while M. paradisiaca (peel) recorded the highest at 11.06%.

Bulk and tap densities
The bulk and tap densities of the bleached pulp samples followed the same order in their decrease from M. paradisiaca (peel) to T. diversifolia stem to peel. The bulk densities of the bleached pulp samples were 0.38, 0.43 and 0.58% for M. paradisiaca (peel), M. paradisiaca (stalk) and T. diversifolia respectively. While the tap densities were 0.47, 0.54 and 0.69% for M. paradisiaca (peel), M. paradisiaca (stalk) and T. diversifolia respectively.

Physicochemical properties of synthesized carboxymethyl cellulose
The physicochemical properties of synthesized carboxymethyl cellulose are presented in Table 4 and samples of synthesized CMC are presented in Figure 3.

Yield
Yield is a function of the amount of materials lost during dialysis step. The yield of the synthesized CMC ranged from 33.21-62.57%, with T. diversifolia recording the highest, closely followed by M. paradisiaca stem, and M. paradisiaca peel recording the lowest.

pH
The pH which is a measure of the acidity or alkalinity of the CMC ranged from 6.51 -6.74. M. paradisiaca had the highest pH of 6.74, closely followed by M. paradisiaca peel with 6.61, and T. diversifolia the lowest having 6.51.

Degree of substitution (DS)
The DS is one of the most important properties of CMC. It does not only influence the solubility of the CMC  molecules but also affects the solution characteristics (Barba et al., 2002). The DS of the CMC samples ranged from 0.17 to 0.33. T. diversifolia recorded the highest (0.33), followed by M. paradisiaca stem (0.28), and the least was M. paradisiaca unripe peel (0.17).

Viscosity
The DS also influences the viscosity, a higher DS results in better viscosity and cation exchange ability. Additional carboxyl groups provide more sites for cross-linking by multivalent cations. The viscosity of CMC samples ranged from 26 -32%. T. diversifolia had the highest viscosity (32%), followed by M. paradisiaca stem (30%), and M. paradisiaca unripe peel the lowest (26%).

Swelling capacity
The swelling capacity of prepared CMC in this work ranged from 350.14 -687.01 with T. diversifolia having the highest (687.01) which was followed by M. paradisiaca stem (553.37), and the unripe peel recording the lowest (350.14).

Bulk and tap densities
There is no much significant difference in the bulk and tapped densities as the bulk densities ranged from 0.65 -0.82% and the tap densities ranged from 0.73 -0.89% for CMC derived from M. paradisiaca peel, stem, and T. diversifolia, respectively. This means that M. paradisiaca stem and T. diversifolia bleached cellulose powders have better flow than M. paradisiaca peel.

Fourier transform infrared spectroscopy (FTIR)
Fourier transform infrared (FT-IR) spectroscopy was used to verify the successful etherification of cellulose. The FT-IR spectral of synthesized and commercial CMC are shown in Figures 3 to 6 and the spectral data analysis in Table 5.

Holocellulose and alpha cellulose
The percentage of holocellulose in the plant materials is in the mid-range holocellulose content. The quality of end product depends on the content of holocellulose; high content increases pulp and quality of end product (Zawawi et al., 2013). The holocelluose content obtained in this research work compared well with those obtained by other researchers.  had reported 71.60% for T. diversifolia, 72.60 and 73.40% for M. parasidiaca and M. sapentium, respectively. Ibrahim et al. (2010) reported 67.80, 57.46, and 53.89% for corn cob, banana plant and cotton gin waste, respectively. The alpha cellulose content of raw materials gives an indication of pulp yield (Sezgin and Serhat 2018). Studies by other researchers have reported comparable results on non-woody lignocellulosic materials.  reported 54.00% for T. diversifolia, 55.00 and 55.33% for M. parasidiaca and M. sapentium respectively. Saelee et al. (2014) reported 44.5% for Sugarcane baggase.

Ash and silica content
Ash content represents different metal salts such as carbonates, silicates, oxalates and potassium phosphates, magnesium, calcium, iron and manganese as well as silicon. From the results in Table 1, M. paradisiaca (unripe peel) has the highest ash and silica content when compared with the other two lignocellulosic materials. Jaramogi et al. (2016) reported higher ash content (9.1%) for M. paradisiaca (stalk) and T. diversifolia (stalk), but lower than the ash content of M.  paradisiaca (unripe peel). The values of the ash were indicative of the presence of high mineral (especially the macrominerals) content in the lignocellulosic materials. The higher the ash content, the higher the mineral composition.

Cold and hot water solubility
From the results in Table 1, M. paradisiaca (Stalk) has the highest solubility when compared to M. paradisiaca and T. diversifolia. Water solubility removes a part of extraneous components, such as inorganic compounds, tannins, gums, sugars and colouring matter present in the lignocellulosic plant and hot water removes, in addition, starches (Shakhes et al., 2011). It can therefore be inferred that M. paradisiaca (Stalk) was more prone to the removal of extraneous components.

Alkali solubility
The alkali solubility of sample indicates an extent of cellulose degradation during processes and has been related to strength and other properties of the further pulp of the sample. M. paradisiaca (unripe peel) has the highest solubility when compared to M. paradisiaca (stalk) and T. diversifolia (stalk). This indicates that M. paradisiaca (unripe peel) has higher cellulose degradation than the other two lignocellulosic materials.

Ethanol-Benzene solubility
The ethanol-benzene extractive consists of soluble materials not generally considered part of the plants substance and is primarily the waxes, fats, resins and some gums as well as some water soluble substances. The result in Table 1 indicates that M. paradisiaca (unripe peel) had the highest ethanol-benzene solubility content. These results obtained in this study were comparable with that obtained for some non-woody plants, corn stalk, 3.5% (Barbash et al., 2012), canola stalk, 2.5% (Enyati et al., 2009) and cotton stalk, 2.93-3.03% (Ali et al., 2001) Physicochemical properties of cellulose

Pulp yield
The results of the yield as presented in Table 2 for T. diversifolia, M. parasidiaca stalk and unripe peel compared favourably with that reported for extracted banana waste (EBW) and waste banana fibre (WBF) using soda pulping method (Arafat et al., 2018); the results ranged from 46.7 to 66.8% for EBF and 29.3 -38.8% for WBF. The result also compared favourably with report by Ogunsile et al. (2006) for M. paradisiaca Mid-Rib using soda pulping and ranged from 25.80 -49.13%.

Ash and Silica contents
The results of the ash and silica content of the pulp showed a reduction in their content compared to the plant raw materials. This trend correlates with that of other reports on the effect of pretreatment and pulping on the reduction of ash and silica content of lignocellulosic materials (Ainun et al., 2017;Serzgin and Serhat, 2018). The higher silica and ash content of T. diversifolia has been attributed to its grass nature (Jones and Handrick, 1967).

Kappa number
From the results in Table 2, M. paradisiaca peel recorded the highest kappa number and this could be attributed to its fruit covering duty which might have built in much lignin as plant glue which in turn assists in fruit covering.

Yield
From the results indicated in Table 3, there was a reduction in the yield of the bleached cellulose pulp.
Higher and lower yield values on non-woody biomass have been reported by other authors (Shirkolaee, 2009;Mohsen et al., 2015). Reduction in yield of bleach compared to non-bleached pulp could be attributed to the sequencing, type and concentration of bleaching agents used. However,  attributed the reduction as a result of removal of some residual lignin and other oxidizable compounds.

Ash and silica contents
There was also a reduction in the ash and silica content of the bleached cellulose pulp as indicated in Table 3. Lower values of 0.06, 0.57 and 0.77% have been reported for Musa sapentium, T. diversifolia and M. paradisiaca  while higher value of 50.6% for rice straw using atmospheric acetic acid pulping and bleaching has also been reported (Xuejun et al., 1999). Reduction in ash and silica content has been attributed to the removal of lignin and other oxidizable compounds which might have contained both ash and silica .

Kappa number
Kappa number of the bleached cellulose pulp also recorded a reduction compared to the unbleached pulp as depicted in Table 3. The reduction could be as a result of the washing and squeezing action during bleaching which might have caused removal of more lignin.

Yield
Yield is a function of the amount of materials lost during dialysis step. Higher and lower yields have been reported. The yield of the synthesized CMC in this research work from Table 4 (Saputra et al., 2014;Hutomo et al., 2012). The difference in yield could be attributed to temperature of the reaction and concentration of NaOH and MonoChloroacetic acid (MCA) applied during synthesis.

pH
The pH of the synthesized CMC in this work as depicted in Table 4 indicates that the samples are in a very weak acidic medium. Lower CMC pH values could indicate a lower purity of the product with non-reacted reagents such as monochloroacetic acid and reaction by-products. Saputra et al. (2014) reported pH range of 7 -14. The pH of this research compared with the report of Bono et al.
(2009) (6.5) for CMC from Palm kernel cake. The variation in properties of the different CMC could be as a result of the source of cellulose used, plant species, age and source which affect the cellulose content compositions (Chandra et al., 2007;Carere et al., 2008).

Degree of substitution (DS)
Since degree of substitution (DS) is the average number of hydroxyl groups replaced by the substituent in every anhydroglucose unit in the chain; therefore the result in  (Kimani et al, 2016). Adinugraha et al. (2005) reported DS range of 0.26 -0.76 for banana pseudo stem. The normal DS range for commercially available CMC is approx. 0.5 -1.5 (Karatas and Arslan, 2016). When the DS is below 0.4, the CMC is swellable but insoluble, while above this value, then, CMC is fully soluble with its hydro-affinity increasing with increasing DS value (Arshney et al., 2006). Since the DS of the prepared CMC are below 0.4, they are insoluble in water but swellable and therefore a good material for superabsorbent biopolymers.

Viscosity
Viscosity of CMC greatly influences the DS, a higher DS results in better viscosity and cation exchange ability. Additional carboxyl groups provide more sites for crosslinking by multivalent cations. From the results in Table 4, T. diversifolia possesses higher swelling capacity. Higher and lower viscosities for non-woody biomass have been reported by several authors. Higher viscosity (66.6 cP) was reported for CMC from PKC (Bono et al., 2009) while lower viscosity (14.0 cP) was reported for CMC from Orange mesocarp.

Swelling capacity
The swelling capacity of the synthesized CMC in this study as shown in Table 4 followed the same trend as the DS. This suggests that swelling capacity is a function of the degree of substitution since T. diversifolia with the highest swelling ability still has the highest DS value of 0.33. Higher and lower swelling capacity of synthesized CMC on non-woody plants have been reported by several authors. Kimani et al. (2016)

Sodium chloride content
The sodium chloride level in CMC is an important parameter; it is a reaction minor-product, considered a contaminant. From the results in Table 4 M. paradisiaca (peel) had more minor bye-products than M. paradisiaca (stalk) and T. diversifolia. The sodium chloride values obtained in this work compared favourably with sodium chloride values of 0.15 and 0.19 % reported for cotton linters (Latif et al., 2007)

Bulk and tap densities
Bulk and tap densities provide an estimate in the ability of a material to flow and be packed into a confined space. Generally, the higher the bulk and tap densities, the better the potential for a material to flow and to rearrange under compression (Azubuike et al., 2012). From the results in Table 4, there is no much significant difference in the bulk and tapped densities. This means that M. paradisiaca stem and T. diversifolia bleached cellulose powder have better flow than M. paradisiaca peel.

Fourier transform infrared spectroscopy
The spectral data analysis is shown in Table 5. The introduction of strong peaks at 1597, 1588 and 1585 cm -1 could be attributed to the presence of carbonyl group (C=O) stretching, confirming the presence of the -COO group and the successful etherification. This suggests that the cellulose from T. diversifolia, M. paradisiaca stem, and unripe peel were successfully modified into CMC (Huang et al., 2017;Asl et al., 2017;Bono et al., 2009). Strong absorption band at 3398, 3372 and 3347 cm -1 was due to the stretching frequency of the -OH group and another band at 2345, 2360 and 2364 cm -1 was due to C-H stretching.

Conclusion
The result of the present work provided an insight into the physicochemical properties of synthesized cellulose and carboxymethylcellulose from M. paradisiaca stem and unripe peel and T. diversifolia. The high alpha cellulose content of T. diversifolia makes it a potential source of sustainable industrial grade cellulose production. The yield from M. paradisiaca peel was significantly low compared to the fruit stem and T. diversifolia. Major functional groups present in the commercial (Fidelo) CMC were also identified in T. diversifolia CMC whereas functional groups such as C-O of the ether and O-C-O stretching of ether were not identified in M. paradisiaca stem and unripe peel CMC. Furthermore, higher DS and viscosity of T. diversifolia CMC when compared to M. paradisiaca stem and unripe peel CMC makes it superior as a useful bio-polymer for industrial applications.