Physicochemical and Binder Properties of Starch Obtained from Cyperus esculentus

Materials

Tubers of C. esculentus L. and Solanum tuberosum L. (potato) were procured from an open market in Abuja, Nigeria. Sodium metabisulfite was obtained from Sigma Chemicals Ltd (St. Louis, MO, USA) while metronidazole was brought from NIPCO (Abuja, Nigeria). All reagents were analytical grade and used without further purification.

Methods

Density Measurements

The bulk and tapped density of the samples were obtained using Stampfvolumeter STAV 2003 (J. Engelsmann AG, Ludwigshafen am Rhein, Germany). A 50-g sample of starch powder was transferred to a 200-mL calibrated glass cylinder through a short-stemmed glass funnel, and the volume occupied by the powder was noted. The cylinder was then subjected to 500 taps to calculate the volume change. The true density of the samples was determined using a 50-mL specific gravity bottle with xylene as the displacement liquid. The Carr’s “percent compressibility” and the Hausner ratio were calculated as per the following equations:

1

2

Isolation

The isolation of starch from tubers of C. esculentus L. was relatively straightforward. Starch granules settled easily, and this settling was not hampered by the presence of non-starch materials from the tubers. The non-starch materials remained suspended in the supernatant liquid and were decanted off. The yield of Cyperus starch, after extraction of oil, was found to be approximately 37 g per 100 g of fresh tubers. This was considered satisfactory as yields of approximately 21% w/w basis have been previously reported in the literature (22).

Physicochemical Characterization

Starch occurs in plants as granules that are characteristic in size, shape, and morphology. As these characteristics are determined by the biological origin, a thorough investigation of the physicochemical properties is warranted. Cyperus starch is a brilliant white, odorless powder with a warm bland taste and smooth texture. As is evident from the micrographs in Fig. 1, Cyperus starch exhibits elliptical to spherical granules with a relatively smooth surface. Scanning electron microscopy indicates that Cyperus starch has uniform granular size, shape, and morphology. Size distribution of a potential excipient has been shown to affect various formulation characteristics like flowability, compactibility, water binding capacity, and drug release (23–25). Although no precise categorization of granule size is documented in the literature, starch granules have been arbitrarily classified as large (>25 μm), medium (10–25 μm), small (5–10 μm), and very small (<5 μm) (26). Cyperus starch predominantly consists of small- to medium-sized granules (percent volume basis) with a mean particle size of 8.25 μm. The particle size distribution of Cyperus starch, shown in Fig. 2, depicts granules ranging from 2 to 17 μm and is fairly consistent with the previously reported range of 3–12 μm (22). Typical particle size distributions for commercially used starches include maize starch (2–32 μm), potato starch (10–100 μm), rice starch (2–20 μm), tapioca starch (5–35 μm), and wheat starch (2–45 μm) (4).

Native starches are known to contain varying amounts of monovalent and divalent cations inherited from the parent material. Incinerating the starch sample and analyzing the residue (known as ash) can be used to estimate the metal ions present in the sample (27,28). Cyperus starch samples show a predominance of divalent cations as compared with maize starch, which has been shown to primarily contain monovalent cations like sodium and potassium (28). The ash value of Cyperus starch is slightly higher than the value recommended in the USP (21) for starch samples and may be due to the presence of magnesium and calcium. Although Cyperus starch was found to contain iron, magnesium, and calcium as indicated in Table I, no traces of heavy metals were detected. Some starches contain small amounts of lipid materials (e.g., maize starch ~0.6–0.7%), and hence require pretreatment for removal of fatty materials (29). Tubers of C. esculentus L. have been shown to contain as much as 27% fixed oils (30). Although most of this oil was removed using Soxhelet extraction, Cyperus starch retains about 0.3% of the lipid content and is probably distributed endogenously. Since fatty substances appear to be complexed with the linear fractions of native starches, their presence significantly alters properties like granule swelling, enzyme conversion, and iodine affinity (31).

Pharmacopeial Specifications

Cyperus starch conforms well to USP standards established for widely used starches like maize and potato (21). The results of the pharmacopeial specifications are indicated in Table I. A sample of Cyperus starch produced a translucent, whitish jelly in boiling water and was colored reddish violet to deep blue as indicated by the identification test in the USP (21). The total aerobic microbial count was well within the limits specified in the compendium. As indicated in Table I, Cyperus starch samples meet the specific requirement for pH values. Although starch rarely contains volatiles other than water, LOD is a good technique to determine the volatile components present in the sample. The 12.5% LOD for Cyperus starch is within the limits set by the USP (21) and is attributed to the presence of moisture.

Swelling Power

To be potentially useful as a pharmaceutical excipient, the ability of starch granules to swell has important implications. Swelling power is not only a measure of the hydration capacity of the sample but is also indicative of the associative forces in the granules. The swelling profile of maize, potato, and Cyperus starch is shown in Fig. 3. As compared with potato and maize starch, Cyperus starch demonstrates the highest resistance to swelling. This lower swelling power could be due to the interplay of various factors: particle size and lipid content being the prominent ones. The small granule size of Cyperus starch already tends to confer compact packing, resulting in a greater degree of association and hence lower swelling. This property of Cyperus starch could be explored for its potential use as a binder in pharmaceutical formulations.

Amylose-to-Amylopectin Ratio

It is well accepted that starches contain two polysaccharide components: the linear fraction (amylose) and the branched fraction (amylopectin). Aqueous dispersions of amylose are unstable due to intermolecular attraction and association with neighboring molecules. Hence, the amylose content significantly affects many fundamental properties like pasting temperature, viscosity, gel stability, water solubility, and the degree of resistance of starch granules to digestion by amylases. Moreover, amylose plays a role in disrupting the crystalline nature of the starch granules (32). The amount of amylose tends to differ among starches from different sources, and thus, a quantitative estimation of amylose content is necessary. As indicated in Table II, the amylose content of Cyperus and maize starch are comparable, a fair indication of the similarity in some of their functional characteristics.

Density Measurements

The density of starch plays an important role for technological and engineering purposes. Several trends have been proposed that relate the bulk density of excipients to functional properties like compactibility and powder flow (33). The Carr’s index and the Hausner ratio are simple methods that use density values to predict the flow properties of powders (34,35). Carr’s index, also known as percent compressibility, describes the flow properties of powders as excellent (5–15%), good (12–16%), fair (18–21%), and poor (23–28%). Similarly, a Hausner ratio value of less than 1.20 is indicative of good flow while a value greater than 1.50 indicates poor flow properties (36). The compressibility index has also been used as measures of interparticulate friction and the likelihood of arch formation (37). The Carr’s index and the Hausner ratio for Cyperus, potato, and maize starch are given in Table II. Based on these values, it is evident that Cyperus starch has fairly comparable flow with respect to the other commonly used starches like maize and potato.

X-ray Powder Diffraction Analysis

Typically starch granules are found to exhibit diffraction patterns that have been classified as A, B, or C. The type of diffraction pattern primarily depends upon the arrangement of the double-helical amylopectin chains. The A pattern is obtained as a result of a close-packed arrangement with a water molecule between each double helix, while a more open hexagonal packing, with water molecules in the central cavity, exhibits a B pattern (38,39). The C pattern is very similar to the A pattern, except for the appearance of a peak around 5° 2θ. Although the biological origin of starch plays an important role in determining the diffraction pattern, other factors like amylopectin chain length, amylase, and moisture content are known to affect the pattern within the same parent source (40,41). Zobel noted that powder patterns may become weak at moisture contents below 10% and the discrete line pattern of B-type structure completely disappears below 1% moisture content (42). Therefore, diffractograms of all samples were obtained at appropriate moisture content (~10%). Powder X-ray diffractograms of Cyperus, potato, and maize starch are shown in Fig. 4. The Braggs reflection angle, 2θ, along with the interplanar spacing, d, and the relative intensity of the peaks are listed in Table III. The interplanar spacing has been calculated using Bragg’s equation given as nλ = 2 d sinθ, where θ is one half the angle read from the diffractogram.

Interplanar spacing and relative intensities of characteristic peaks by which the three patterns can be differentiated are well documented in the literature (42). Visual inspection of the diffraction data obtained for the three samples (Fig. 4) indicates that Cyperus and maize starch have their strongest peaks at approximately 15°, 17°, and 23° 2θ. Moreover, the peak at 18° 2θ is unresolved and has been converted into a shoulder, a feature characteristic of an A-type pattern. Potato starch has its strongest diffraction peak at around 17° 2θ, relatively medium peaks at around 5°, 15°, 22°, and 24° 2θ and a couple of weak peaks scattered around 10° and 19° 2θ. The pattern of peaks seen in the diffractogram of potato starch is characteristic of a B-type crystalline structure (42). Although tuber starches usually give a B-type powder pattern, the distinguishing peak for B pattern at 5° 2θ is not seen for Cyperus starch. Thus, it was concluded that the crystalline structure of Cyperus starch is very similar to that of maize starch. Irrespective of the powder pattern, it is generally accepted that a starch granule involves alternating regions of amorphous and crystalline lamellae. The degree of crystallinity was estimated utilizing the technique described by Nara and Komiya (43). The degree of crystallinity of the three samples was in the order of Cyperus starch > maize starch > potato starch with the numerical values being 39.4%, 38.3%, and 20.9%, respectively. Potato starch with the highest amylose content has the lowest degree of crystallinity, thus providing more evidence that amylose seems to disrupt the crystalline nature of starch granules. Cyperus and maize starch have similar amylose content and correspondingly exhibit similar crystallinity.

Moisture Sorption Analysis

Moisture sorption by starch has been attributed to the interaction between the hydroxyl groups of the hexose moiety and water molecules (44). Water molecules form hydrogen bonds with both amylose and amylopectin, but the amylopectin structure has been shown to physically trap water molecules. On this basis, it was hypothesized that starch granules high in amylopectin would have a higher moisture sorption potential (45). However, another factor that seems to supersede this hypothesis is the crystallinity of the polymer. Crystalline polymers have been proposed to have extensive secondary intermolecular bonding. This secondary bonding causes the hydroxyl groups on adjacent glucose units to interact with each other and hence reduces the available sites for adsorption of water molecules (46). As a result, the higher degree of crystallinity could reduce the moisture sorption, as seen in Fig. 5. No significant difference is noted in the ability of Cyperus and maize starch to interact with moisture. This is further confirmed by fitting the Guggenheim, Anderson, and de Boer (GAB) equation (47) to the moisture sorption data. It has also been shown that most of the water associated with polymers is localized in the amorphous region (48). Thus, the value of the so-called monolayer coverage obtained from the GAB equation should be corrected by dividing the W_m value by the fraction of the non-crystalline region and then referred to as the W^corr_m. The GAB parameters for the three starch samples are given in Table IV. It should be of interest to note that, upon accounting for the degree of crystallinity, potato starch has the lowest value of monolayer coverage. The high equilibrium moisture content values exhibited by potato starch are primarily due the localization of moisture in the amorphous region at high relative humidity.

Diffraction studies performed to elucidate the structural characteristics of starch granules have clearly indicated the effect of water content on powder pattern. It was further shown that granules exhibiting a B-type pattern (e.g., potato) can hold more water molecules as compared with an A-type pattern (e.g., maize and Cyperus) (38,39). These studies clearly designate that water associated with starch is not merely adsorbed at the surface but also penetrates the granules. Young and Nelson derived a set of equations that differentiate the moisture associated with biological materials as “bound” or adsorbed monolayer on the surface, “normally condensed” or external moisture, and “absorbed” or internal moisture (49–51). The amount of internally absorbed moisture tends to decrease the “free” water and thus alter the potential of an excipient to participate in bond formation. The analysis of moisture sorption data according to the Young and Nelson equations is illustrated in Fig. 6a–c. As explicated by this analysis, Cyperus starch has the lowest amount of absorbed water and most of the water associated with Cyperus starch is externally adsorbed. The high amount of internally absorbed water for potato starch is in good agreement with its internal structure, i.e., more open and hydrated hexagonal crystallites. The amount of externally adsorbed water is important as it can assist in the formation of liquid bridges between particles. Soluble components present in the formulation tend to associate with these liquid bridges and upon drying; they convert to solid bridges that help maintain the binding efficiency. In cases where the internally absorbed moisture is high, the amount of water available for bridge formation decreases and may result in formulation problems.

Evaluation of Binding Efficiency

Cyperus starch was employed as a binder in concentrations of 5%, 7.5%, and 10% w/w while a similar formulation containing 10% w/w potato starch as binder was used for comparison. To study the effect of binder concentration, all other factors, i.e., the amount of drug, lubricant, and endo-disintegrant were maintained constant across formulations. The details of the four formulations namely C1, C2, and C3 for Cyperus starch and P1 for potato starch are provided in Table V. Both granules and subsequently tablets, obtained using the technique of wet granulation and compression, were characterized. The moisture content, density, friability, porosity, and compressibility data for the granules are presented in Table VI. The average particle size of granules from all formulations ranged between 1,050 and 1,080 μm. Particle size is an important parameter since particle size distribution affects the average tablet weight, disintegration time, granule friability, and granule flowability. Similarly, the moisture content of the granules did not show significant differences. For Cyperus starch, the concentration of binder seemed to have a direct effect on the compressibility index. Increasing the concentration of Cyperus starch as binder seemed to lower the compressibility index of the granules. This indicated that the flow properties of the granules improved significantly with an increase in binder concentration. A similar trend was also observed for the porosity and the friability of the granules. A comparison of formulations containing the same amount of Cyperus and potato starch as binder (10% w/w) indicates that granules obtained using Cyperus starch have better flow properties, lower porosity, and lower friability.

Selected properties of metronidazole tablets are listed in Table VII. For formulations with 5% and 7.5% w/w Cyperus starch (formulations C1 and C2, respectively), no significant difference in the hardness of the tablets was noted, but the disintegration time increased significantly. Increasing the concentration of Cyperus starch paste to 10% w/w (formulation C3) leads to a considerable increase in both hardness and disintegration time. It is interesting to note that the hardness of the tablets employing 5% w/w of Cyperus starch as binder is similar to that obtained using twice the amount of potato starch (10% w/w). When Cyperus starch is used at 10% w/w (formulation C3), the hardness of the tablets was found to be significantly higher while the friability appeared negligible with respect to the potato starch formulation at the same binder concentration (formulation P1). The disintegration time, however, did increase with increasing concentration of Cyperus starch. At a similar binder concentration (10% w/w), the disintegration time for tablets containing Cyperus starch (formulation C3) was twice that of tablets containing potato starch (formulation P1). The increase in disintegration time can be attributed to two factors, viz., the lower porosity of the granules formulated using Cyperus starch and the inherent ability of the two starches to swell. Porosity values reflect the network of pores through which water can penetrate a compact, causing it to swell and consequently disintegrate. Additionally, as discussed earlier, the swelling power of Cyperus starch is inherently lower than that of potato and maize starch (Fig. 3). Interplay between the lower porosity of formulated granules and the intrinsic resistance to swelling appears to manifest itself in the form of increased disintegration time for tablets containing Cyperus starch as binder.

Although differences were noted in the disintegration times of the various formulations, tablets from all four formulations disintegrated within 5 min or less. Hence, a significant impact of binder concentration on the dissolution profile of either formulation was not expected. The dissolution profiles of all four formulations are illustrated in Fig. 7. Visual inspection of the dissolution profiles does not indicate considerable differences between them, and all formulations release 85% or more of the drug in less than 15 min. To further verify this, a model independent mathematical approach to compare the dissolution profiles proposed by Moore and Flanner (52) was used. Using the mean dissolution values from curves at each time point, the mathematical treatment calculates the difference factor (f₁) and the similarity factor (f₂). The f₁ value is 0 when the mean profiles are identical and increases proportionately as the difference between the mean profiles increases. Similarly, the f₂ value is 100 when the profiles are identical, and an average difference of 10% at all measured time points results in a f₂ value of 50. A value of f₁ between 0 and 15 while a value of f₂ between 50 and 100 implies sameness or equivalence of the dissolution profiles (53). This method has been widely adopted by the Food and Drug Administration in various guidance documents, with more emphasis on the use of f₂ values. It should be noted that, for products which are rapidly dissolving, i.e., more than 85% in 15 min or less, a profile comparison is not necessary. As expected, the f₂ values used to measure closeness between the various dissolution profiles of the four formulations were greater than 75, clearly establishing similarity. It is noteworthy that insignificant difference is noted between the dissolution profiles of formulation C3 and formulation P1 (f₂ = 92.4), despite the fact that the disintegration time for tablets from formulation C3 is almost twice that of those from formulation P1. These results insinuate that Cyperus starch may provide relatively better binding efficiency without compromising drug release characteristics.