Extraction and Characterization of Cellulose and Microcrystalline Cellulose from Teff Straw and Evaluation of the Microcrystalline Cellulose as Tablet Excipient

ABSTRACT Teff (Eragrostis tef), a grass which belongs to the Family Poaceae, is widely cultivated for its starch-rich grains in Ethiopia, generating large amounts of agricultural byproduct, teff straw. The aim of this study was to isolate and characterize cellulose and microcrystalline cellulose (MCC) from teff straw and evaluate MCC as directly compressible tablet excipient. Cellulose was extracted from teff straw and partially depolymerized to obtain MCC. The physicochemical properties of cellulose and MCC powders were characterized. Yields of cellulose and MCC powder from the raw material were 35.2% and 27.2%, respectively. The samples exhibited type-I crystal lattice and similar infrared spectra with that of Avicel PH-101. The degree of polymerization (DP) and crystallinity index of cellulose were 594.51 and 72.26%, respectively. Whereas, MCC powders showed DP of 241.09–257.38 and crystallinity indexes of 76.45–84.52%. Spray dried MCC was found to be more porous and poorly flowable and had higher moisture content when compared with the oven-dried MCC powder. Tablets prepared from both MCC powders fulfil most of the pharmacopoeial requirements. The spray-dried MCC powder also showed superior compactibility to oven-dried MCC. Teff straw can, therefore, be considered to be an alternative source of cellulose and MCC.

Cellulose is a linear homo-polysaccharide composed of Anhydro-D-glucopyranose units (AGUs), which are linked together by β-1,4-glycosidic bonds.The chain length of cellulose can be defined by its average degree of polymerization (DP), which expresses the number of AGUs in the molecule.Average DP values vary with the type and nature of source and treatment methods employed (Klemm et al. 1998a).
Cellulose has been isolated with different pretreatment processes, commonly categorized into physical, chemical, physicochemical, and biological pretreatments (Sun et al. 2016).Formic/Acetic acid delignification rate is reported to be effective cellulose isolation from agricultural byproducts (Jahan et al. 2014).Hydrogen peroxide has become the preferred bleaching agent since the use of hypochlorite gives rise to the formation of a number of chlorinated hydrocarbons such as the carcinogenic trichloromethane (Lacasse and Baumann 2012).
Microcrystalline cellulose (MCC) is prepared by mineral acid catalyzed partial hydrolysis of native cellulose.The main difference between cellulose and MCC is that the later has lower molecular size (DP ≤ 350) but higher degree of crystallinity (Battista et al. 1956).MCC is widely used in the pharmaceutical, cosmetics, food, and other industries.It is one of the most widely used filler/binders, disintegrant, and lubricant in direct tablet compression.It is also used as a diluent in tablets prepared by wet granulation as well as a filler for capsules and microspheres (BP 2009).
The most common natural sources of industrial cellulose are wood pulp (90%) and cotton linter (6%) (Jahan et al. 2014).However, deforestation and the acceleration of greenhouse effects are nowadays becoming global concerns.Also, competition among many industries such as furniture, pulp, textile, and paper become challenging to provide all sectors sufficient quantities of wood and cotton at a reasonable price.
Hence, interest is increasingly growing on agricultural products and byproducts as alternative industrial cellulose sources since they possess versatile attributes.An appreciable amount of cellulose can be obtained from these highly renewable (non-wood) sources.Moreover, agricultural byproducts usually demand less energy and chemical for bleaching because of lower lignin component than wood (Trache et al. 2016).Teff (Eragrostis tef), a grass which belongs to the family Poaceae, subfamily Eragrostoidae, and genus Eragrostis (Seyfu 1997), is the most widely cultivated cereal crop in Ethiopia.In a single "meher" season (June -August) of 2015/2016, 2.87 million hectares (22.95%) of land was covered solely with Teff.From this massive farming, the country's annual production of teff straw will reach as much as 9.42 million tons (Mottaleb 2018).Though there are few pharmaceutical manufacturing companies in Ethiopia, they import literally all pharmaceutical excipients (including MCC) from abroad.To the investigators' knowledge, the physico-chemical properties of MCC from Teff straw have not yet been studied and evaluated as directly compressible excipient.
Therefore, this study aims to isolate and characterize native cellulose and MCC from Teff straw and evaluate MCC as directly compressible pharmaceutical excipient.

Cellulose isolation from teff straw
Cellulose was isolated with formic/acetic acid pretreatment with three successive treatments.Initially, an air-dried straw was boiled with 85% formic acid and 99.5% acetic acid (70:30) at 1:8 fiber-to-liquor ratio in water bath at 90°C for 90 min.After straining through nylon cloth, the residue was washed with distilled water and further delignified in water bath for 90 min along with 10% H 2 O 2 , 85% formic acid, and acetic acid (2:1:1 volume ratio).Following filtration and washing with distilled water, the residue was bleached with 10% H 2 O 2 for 60 min in boiling alkaline medium (4% NaOH) with 1:10 fiber-to-liquor ratio.Finally, it was repeatedly washed with distilled water, strained by nylon cloth, and oven-dried to constant weight at 50°C (Golbaghi, Khamforoush, and Hatami 2017;Padmadisastra and Gonda 1989).

Preparation of MCC from the isolated cellulose
MCC was prepared following the method developed by Battista (Battista 1950).Briefly, the isolated cellulose fibers were treated with 2.5N HCl (1:20 fiber-to-liquor ratio) by boiling at 105°C for 30 minutes.After completion of the reaction time, the mixture was washed, at room temperature, with distilled water, followed by 5% ammonium hydroxide solution.It was then washed repeatedly with distilled water till the filtrate became clear and odorless (Battista and Smith 1962).The resulting MCC was dried in hot air oven (Kottermann® 2711, Germany) at 100°C until constant weight was achieved.Finally, it was ground and sieved through mesh number 224 µm and labeled as MCC-0.Other MCC powders (MCC-OD & MCC-SD) were prepared with similar conditions of hydrolysis as MCC-0 but, in this instance, the hydrolyzed cellulose was milled with juice blender for 2 min as a slurry in water.In case of MCC-OD, the material was dried in hot air oven till constant weight was achieved and then pulverized.Whereas, MCC-SD powder was prepared by spray drying (Tall Form Spray Drier (FT80), England) of the milled slurry at 16% consistency at inlet and outlet temperatures of 175°C & 120°C, respectively, and air pressure of 1 bar (Battista and Smith 1962).

Identification tests
Reactions with iodinated zinc chloride and DP (using cuprammonium hydroxide (Cuam) as solvent) were determined (Klemm et al. 1998a(Klemm et al. , 1998b)).The average viscosity molecular weights were calculated as the product of DP value and the molecular weight of monomer units (162 g/mole).Percent yield was calculated based on the dry weight of raw material.

Morphological study
Morphological analysis of cellulose and MCC samples was done using scanning electron microscope (JEOL/MP-JSM-IT300) with gold coating.All images were taken at an accelerating voltage of 20 kV and magnifications of 500 and 1000 × .

FTIR spectroscopy
Dried samples were analyzed with Fourier transform infrared (FTIR) spectrophotometer (FTIR-8400S, SHIMADZU, Japan) following the standard KBr pellet technique in transmittance mode.Each FTIR spectrum was collected with 20 scans and spectral resolution of 8 cm −1 in the region 4000 to 400 cm −1 .

X-Ray diffraction (XRD)
Diffraction patterns of cellulose and MCC samples were measured with an automated powder X-ray diffractometer (XRD-7000S, SHIMADZU, Japan).The XRD data were generated by a diffractometer with Cu-Ka radiation (λ = 1.542°A) at 40 kV voltage and 30 mA current over 2θ angle range of 10°-40°, angle step of 0.02°, a time step of 0.4 seconds, and scan speed of 3°/minute.The crystalline indices (CrI) of samples were calculated from the XRD patterns based on the peak height method developed by Segal et al. (1959) (Eqn. 1).
Where, CrI represents relative degree of crystallinity (%), I 002 is the maximum intensity (in arbitrary units) of the 002-lattice diffraction at a 2θ angle between 22-23 degrees, and I am is the intensity of diffraction of amorphous portion in the same units at 2θ around 18°. Intensities of both I am and I 002 were taken above the baseline at their respective positions.
Crystal size was estimated using the Scherrer equation (Eqn.2) where, L is the crystal dimension (in nanometers) perpendicular to the diffracting planes with Miller indices of hkl, λ is the wavelength of X-ray radiation (λ = 1.542Ǻ),Scherrer constant (K = 0.94) represents the shape factor (depends on the shape of the crystallites), β 1/2 is the full width at half maximum (FWHM) of the diffraction peaks, in radians, at a height half-way between background and the peak maximum, and θ is half of the (002) Bragg diffraction peak position (2θ max position) (Trache et al. 2014).

Thermal properties
Thermal properties, TGA and DTA, of samples were investigated on a simultaneous thermal analyzer (DTG-60 H, SHIMADZU, Japan).Each sample was heated from room temperature to 700°C using platinum cell at a heating rate of 10°C/min.Nitrogen gas was used as the carrier gas to provide inert atmosphere at a flow rate of 50 mL/min.
Background reading was done with the dispersing medium (distilled water).Obscuration was kept at 17 to 19%.Determinations were done in triplicates, and average values with corresponding standard deviations were calculated.

Density and related properties
Bulk, tapped density, Hausner ratio, and Carr's index of MCC powders were determined according to the methods described in the United States Pharmacopeia (USP-616 2021).True density was determined by liquid displacement method using xylene as immersion fluid.

Other physicochemical properties
Solubility in Cuam, moisture content, ether-soluble fraction, ash value, and pH were determined using the methods described in the British Pharmacopeia (BP 2009).Moisture content, hydration capacity, particle size, and density and related properties of MCC powders were also studied.

Preparation and evaluation of tablets
Tablets containing MCC powder were prepared to study compaction property by compressing at targeted crushing strengths of 50, 75, 100, 125, and 150 N (adjusted using Avicel PH-101) and 11 mm diameter with rotary tablet machine (EC0 Press, 7891, India).To study dilution potential, MCC powders were compressed along with different concentrations (30, 45, 60, and 75%) of paracetamol (Kuentz and Leuenberger 2000) at a fixed compression pressure which is adjusted to give Paracetamol/Avicel (3:7) tablets with diametrical crushing strength of 80 N. Paracetamol and MCC powders were thoroughly mixed for 10 min in a Turbula ® mixer at 45 rpm.Tablets weighing 400 mg and containing either plain MCC powder or MCC with magnesium stearate (0.5%) were also prepared at a fixed compression pressure to study lubricant sensitivity of the filler/binders.Dilution potential, lubricant sensitivity ratio, tablet weight variation, tablet strength, and disintegration time were determined for MCC tablets prepared at all compression forces and all levels of paracetamol content following pharmacopoeial methods.Dissolution test was conducted according to the USP specification using dissolution apparatus Type II (Paddle Method) with phosphate buffer (pH 5.8, 900 ml) as the dissolution medium.

Statistical analysis
Statistical analysis was performed using one-way Analysis of Variance (ANOVA) with Origin 7 statistical software.Tukey multiple comparison test was used to compare the individual differences in physicochemical properties of MCC powders and their tablet properties.Powder and tablet properties were analyzed at 95% confidence level (α = 0.05); hence, P-values less than 0.05 were considered statistically significant.The results are reported as mean and standard deviation (SD) of sufficient number of measurements (specified in each task).

Yield and physicochemical properties
The percentage yield of cellulose was about 35.2%, which is comparable with a study reported by Gabriel et al. (2020).Whereas, the yield of MCC was found to be 77.4% (±1.1) and 27.2% from isolated cellulose and raw material, respectively.Slightly higher and lower yields were reported by previous studies from brown algae (32.5%) (Tarchoun, Trache, and Klapötke 2019) and date seeds (25%) (Abu-Thabit et al. 2020), respectively.Both the isolated cellulose fiber and the prepared MCC powder had white color with no odor and taste.In addition, both cellulose and MCC samples were completely soluble in Cuam solvent and gave violet-blue color in iodinated zinc chloride solution.
The DP values of all the prepared MCCs and Avicel PH-101 powders (Table 1) are consistent with those commonly reported values; 140-350 (Battista et al. 1956).All prepared MCC powders showed significantly higher DP value than Avicel PH-101 (P < .05).

Morphology
Figure 1 shows the scanning electron micrographs of cellulose fibers and MCC-SD.The isolated cellulose displayed distinct elongated rod-shaped fibers.The surface of cellulose fibers has flattened appearance.Whereas, MCC-SD powder exhibited irregular-shaped morphology in its aggregates.There are also few well-defined MCC particles with rod shape but shorter length.Fiber length is obviously shortened with organic acid hydrolysis of the cellulose fibers (El-Sakhawy and Hassan 2007).

FTIR spectra
Figures 2 and 3 show the FTIR Spectra of cellulose and MCC obtained from Teff straw, respectively.The absorbance bands at 1377 cm −1 , 1315 cm −1 , 1157 cm −1 , 1057 cm −1 , and 895 cm −1 are associated with the typical values of MCC spectrum (Sun et al. 2005).Strong peaks of O-H stretching vibrations are reflected with broad band in the region 3150-3400 cm −1 .The absorption bands around 2854 cm −1 and 2923 cm −1 originate from symmetric and asymmetric C-H 2 stretching vibrations, respectively (Klemm et al. 1998c).
Symmetric C-H 2 deformations are reflected by a sharp peak at 1458 cm −1 (Adel et al. 2010), whereas asymmetric deformations are seen at 1377 cm −1 Click or tap here to enter text (Sun et al. 2005).The band at 1313 cm −1 is associated with CH 2 rocking vibration at C-6 position.The absorption band at 1157 cm −1 corresponds to C-O-C asymmetrical stretching of β-1,4-glycosidic linkage.The band at 1057 cm −1 is observed due to C-O stretching vibrations.Out of plane in position ring C-H stretching, which are specific for β-glucoside linkages between glucose units in cellulose, are also observed with weak band at 895 cm −1 (Klemm et al. 1998c).
Since characteristic bands of lignin (in the region 1509-1609 cm −1 and around 1433 cm −1 ) are absent in both cellulose and MCC spectra (Figure 3), it is possible to argue that the polymer is  effectively removed by treatments applied.Moreover, there are no peaks around 1730 cm −1 which is found in spectra of hemicelluloses and lignin (Sun et al. 2005)

X-Ray diffraction analysis
The crystallinity index (CrI) and crystal sizes of samples are summarized in Table 2. Amorphous (I am ) and crystalline (I 002 ) intensity peak positions were located in the angle (2-θ) range of 18.04°-18.94°and 22.30°-22.68°,respectively (Figure 4).Differences in peak height and width of samples are  indicative of differences in their degree of crystallinity and crystal size (Trache et al. 2014).The highest CrI value of MCC-0 (85%) is attributed to the removal of the amorphous regions of cellulose by acid hydrolytic cleavage of 1,4-glycosidic bonds (Battista 1950).Crystal size also increased from 2.64 nm (cellulose) to 3.82 nm (MCC-0) (Table 2).
There is no doublet peak at the 002 plane (Figure 4) which indicates that all the samples comprised cellulose-I polymorph (Haafiz et al. 2013).This is also supported by the FTIR spectrum of MCC (Figure 3) where absorbance at 1315 cm −1 is seen, which corresponds with CH 2 wagging of cellulose-I (Klemm et al. 1998c).Closely related infrared spectra have been reported by various studies done on MCC isolated from different agricultural residues (Abu-Thabit et al. 2020; Tarchoun, Trache, and Klapötke 2019).

Thermal properties
The TGA and DTA thermograms displayed three major weight loss stages with different rate and extent (Figure 5).The main processes involved in cellulose degradation are identified to be dehydration, decarboxylation, depolymerization, and decomposition of glycosyl units (Trache et al. 2014).
MCC-SD and Avicel PH-101 showed better thermal stability on heating up to 325°C, while teff straw was highly thermo-resistant above 345°C.Removal of hemicellulose during cellulose isolation and subsequent reduction of amorphous portions during hydrolysis (higher degree of crystallinity thereof) are the main reasons for enhanced stability of MCC-SD and Avicel PH-101 (Gabriel et al. 2020).Initial weight loss of Teff straw (9.17%), cellulose(7.63%),MCC-SD (5.36%), and Avicel PH-101 (5.98%) was observed with endothermic peaks on DTA curves (Figure 5) at respective temperatures of  56.84°C, 64.91°C, 67.54°C, and63.59°C.This weight loss is reported to be associated with the evaporation of loosely bound moisture on the surface of the samples and evolution of absorbed water (Adel et al. 2010;Tarchoun, Trache, and Klapötke 2019).Final residue at 700°C was highest for teff straw (26.25%), whereas Avicel PH-101 exhibited the lowest (4.99%) value.The analogous values for cellulose and MCC-SD were, respectively, 11.167% and 9.58%.Presence of lignin and its high DP polymers could have contributed for higher temperature resistance at elevated temperature and the high residue seen in case of teff straw (Abu-Thabit et al. 2020).

Particle size and size distribution
Particle size and size distributions of MCC powders are presented in Table 3. Milling of the hydrolyzed cellulose in case of MCC-OD and MCC-SD and mechanical size reduction after drying MCC-OD could be the cause for their lower mean particle size compared with MCC-0.
Avicel PH-101 had significantly narrower particle size distribution (lower span value) than both MCC-OD (P = .002)and MCC-SD (P = .007)(Table 3).But there was no statistically significant difference between particle size distributions of MCC-OD and MCC-SD.Whereas, MCC-0 displayed the broadest particle size distribution of all the samples.
The specific surface area of all MCC powders found in this study are lower than the reported value for Avicel PH-101 (1.0 and 1.5 m 2 /g).This difference could be associated with the limitation of the  method used (Laser Diffraction), which underestimates specific surface area because it assumes particles to be spherical in shape and non-porous (Malvern 1999).Moreover, the higher true density obtained in the current study compared with the value reported (1.521 g/ml) by Pesonen, Paronen, and Puurunen (1989) could have resulted in lower surface area since the two variables are inversely related.

Density and related properties
Density and related properties of the prepared MCC and Avicel PH-101 powders are summarized in Table 4. MCC-OD.and Avicel PH-101 displayed comparable values of both bulk and tapped densities.MCC-0 had significantly lower bulk density than the other MCC powders.The decrease in bulk and tapped density in case of MCC-SD, compared with MCC-OD, could result from the fact that porosity increases with spray drying (Thoorens et al. 2014).Based on the criteria set by USP Pharmacopoeia, both MCC-OD and Avicel PH-101 have passable flow, whereas MCC-SD showed poor flow property (USP-1174 2021).MCC-SD showed significantly higher porosity value than both MCC-OD and Avicel PH-101.However, this increased porosity can improve compactibility, resulting in tablets with higher strength (Kumar, De Luz Reus-Medina, and Yang 2002).Almost similar densities were reported for MCC powders isolated from corn cobs by Azubuike and Okhamafe (2012).

Lubricant sensitivity ratio
The LSR of MCC powders was in the order of Avicel PH-101 (0.19 ± 0.010) > MCC-OD (0.16 ± 0.008) > MCC-SD (0.14 ± 0.012).LSR is largely determined by the completeness of the lubricant surface film formation which depends on both flow property and surface area of MCC powder (Doelker et al. 1995).Particle size is reported to affect flow and surface area available for the lubricant, whereas moisture content affects flow and the magnitude and nature of the attractive forces between excipient and the lubricant.The poor flow property of MCC-SD could hinder mixing with the lubricant and hence reduced its LSR (Vromans, Bolhuis, and Lerk 1988).The higher specific surface area of Avicel PH-101, in addition to its better flow property, could make it to be the most sensitive for magnesium stearate lubrication.

Tablets compressed from plain MCC powders Crushing strength and friability
Statistically significant (CI = 95%) difference in crushing strength was observed between MCC-OD and MCC-SD tablets at low (CF2) and high (CF5) compression forces (Figure 6).Tablets prepared from plain MCC-OD and MCC-SD at all compression forces (CF1-CF5) showed significantly lower crushing strength than those of Avicel PH-101 (P < .05).The higher specific surface area of MCC-SD and Avicel PH-101, compared to MCC-OD, could have improved their compactibility.Additionally, the higher degree of crystallinity of MCC-SD and Avicel PH-101 powders (Section 3.4) could have resulted in higher tablet strength than MCC-OD, which had significantly lower degree of crystallinity (Suzuki and Nakagami 1999).With the exception of those prepared at CF1, all the MCC-OD tablets had friability values within the acceptable range (<1%).

Disintegration time
Although Avicel PH-101 tablets showed higher crushing strength than both MCC-SD and MCC-OD tablets, it exhibited faster disintegration.This faster disintegration could be ascribed to the higher specific surface area of Avicel PH-101 powder.Since disintegration of MCC tablets is attributed to the penetration of water into the hydrophilic tablet matrix by means of capillary action of the pores (Bolhuis and de Waard 2011), the relatively higher amount of residual wax found in MCC-OD and MCC-SD could have slowed their swelling.Moreover, the lower hydration capacity, which results in lower swelling capacity, could also have increased disintegration times of MCC-OD and MCC-SD (Rojas and Kumar 2012).

Paracetamol loaded MCC tablets crushing strength and friability
At all levels of paracetamol concentration, tablets prepared from MCC-OD showed significantly lower crushing strength than those of Avicel PH-101 and MCC-SD (except at 60%).
Though not statistically significant, apparently MCC-SD tablets showed lower crushing strength than Avicel PH-101 at 30 & 45% paracetamol.But at higher paracetamol concentrations (60 and 75%), MCC-SD and Avicel PH-101 tablets exhibited comparable crushing strengths (Figure 7).The friability of paracetamol tablets containing MCC-SD and Avicel PH-101 were within the acceptable range (<1%) (USP-1216 2021) with the exception of tablets with 75% drug content.Whereas, tablets compressed from MCC-OD with only 30 or 45% paracetamol content had friability values within the acceptable range.

Dilution capacity
Dilution capacity of Avicel PH-101, using paracetamol as model drug, was reported to be 79.9% (w/w) (Kuentz and Leuenberger 2000).A lower dilution capacity of Avicel PH-101 (65%) was reported by a previous study using similar model drug (Habib et al. 1996).Looking at the tablet strength values found in this study, it can be inferred that MCC-SD and Avicel PH-101 have similar dilution capacity for the model drug used.Hence, equivalent amount of paracetamol reported for Avicel PH-101 in the literatures can be loaded with MCC-SD provided that compression pressures are adjusted.But, MCC-OD exhibited lower dilution potential compared with both MCC-SD and Avicel PH-101 (Figure 7).

Disintegration time
All paracetamol loaded MCC tablets had disintegration time of less than 15 min, which is recommended for conventional dosage forms.Tablets prepared from MCC-SD showed delayed disintegration than MCC-OD and Avicel PH-101 tablets with equivalent hardness.The variations in disintegration times of these tablets could be attributed to the differences in particle specific surface area and hydration capacity (Bolhuis and de Waard 2011).

Dissolution study
The absorbance (at 243 nm) versus concentration plot prepared from paracetamol solutions provided a linear regression equation Y = 0.5702X-0.02881(where Y is the absorbance and X is the concentration in µg/mL) and correlation coefficient, R 2 = 0.999.All the formulations displayed dissolution profile which is in agreement with the pharmacopoeial limit set for conventional tablets (USP-711 2021).There was no significant difference in dissolution profiles of paracetamol tablets prepared from all MCC powders (Figure 8a-c)).Dissolution test was not performed for tablets prepared from MCC-OD at paracetamol content of 75% as they were found to have crushing strength below the acceptable hardness level for immediate release tablets.

Conclusion
Teff straw was delignified with formic/acetic acid-based treatment followed by hydrogen peroxide bleaching.The spray-dried MCC had comparable degree of crystallinity with Avicel PH-101.The flow property of MCC-OD was comparable with Avicel PH-101 while MCC-SD was inferior to both MCC-OD and Avicel PH-101.MCC-SD showed lower compressibility than Avicel PH-101 when compressed alone; their paracetamol formulations had comparable tablet properties.But paracetamol loaded MCC-OD tablets showed inferior tablet tensile strength and hence lower dilution potential.The physicochemical properties of the tablets depended on the technique of drying.Spray drying significantly improved different properties of tablets as compared to oven drying technique.Based on the findings, it may be concluded that teff straw can be considered as promising alternative source of cellulose and MCC for various industrial applications.

Highlights
• The X-ray diffraction pattern and thermograms of MCC powders prepared in our study are closely related with that of the commercial MCC (Avicel PH-101).• The spray-dried MCC powder (MCC-SD) and Avicel PH-101 have similar dilution capacity and comparable compressibility using paracetamol as the model drug.• Tablets prepared from MCC-SD powder had comparable mechanical strength with those prepared from Avicel PH-101.
• Disintegration time and dissolution profiles of paracetamol loaded tablets prepared from both MCC-OD and MCC-SD met pharmacopoeial standards.

Figure 2 .
Figure 2. Infrared spectrum of cellulose isolated from Teff straw.

Figure 3 .
Figure 3. Infrared spectrum of MCC prepared from teff straw cellulose.
*Values in parenthesis indicate accepted values according to British Pharmacopoeia [10].

Table 3 .
Volumetric mean particle size and size distribution of MCC powders.
D(v, x) = represents the size of particles (volume %) below which "x" fraction of the sample lies.*The values in parenthesis indicate acceptable range according to manufacturer's specification for Avicel pH-101 [10].

Table 4 .
Density and related properties of MCC powders.