Crude Khaya senegalensis gum obtained from Khaya senegalensis trees, Purified Khaya senegalensis gum (KSG), Ethanol 97% denatured with 3% IPA, Disolol^® (Chem.-Lab NV, Belgium), Potassium sorbate Ph. Eur (SA Aca Pharma NV). All other materials were of analytical grade.

Gum exudates were obtained dry from Khaya senegalensis trees grown in Zaria town of Kaduna State. The crude gum exudate was authenticated in the herbarium unit of the Department of Biological Sciences of Ahmadu Bello University Zaria where a voucher number 900181 was assigned and specimens were kept. The crude KSG was coarsely size reduced and a 500 g quantity was weighed on a Metler Toledo balance (AB 204, Switzerland) and dispersed whilst stirring intermittently in 5L of deionised water, preserved with 0.1% potassium sorbate and kept at room temperature for 24 h to allow for complete hydration. The dispersion was then homogenised with a Silverson mixer (model L4S-R, Switzerland) and 15 L of deionised water was further added to obtain an even dispersion which was stored at 6ºC for 24 h. Thereafter, a total of 12.5 L of ethanol was used to precipitate the gum from the aqueous dispersion. At this point two layers formed and particulates settled at the bottom, settling was allowed for 30 min after which KSG was filtered through Whatman filter paper no.1 and then spread on absorbent paper prior to drying in a vacuum oven (WTC Binder, Tuttlingen, Germany) at 40ºC for 48 h, the gum obtained was referred to as KSG-VD (vacuum dried). The gum was again extracted as described above and then dried in a freeze dryer (Telstar Bomba RD-18, Spain). The freeze-drying cycle was operated at shelve temperature - 5°C for 25 min then freezing at - 45º C for 150 min and there after drying at a low vacuum pressure of 0.100 mbar for 30 min at -15º C for 13 h 30 min and then at 0ºC for 15 min and finally at 10ºC for 9 h. The resulting gum from this drying process was labelled Khaya senegalensis gum Freeze dried (KSG- FD).

The crude (w1) and the dried, precipitated and purified gum (w2) samples were weighed and the percentage yields were computed as given in the equation below. The colour, odour and texture were observed physically.

1% yield =w1/w2×100

A 1.0% w/v concentration of KSG-VD and KSG–FD were separately prepared in deionized water and the pH determined using a pH meter. Aqueous solubility of the gum samples was determined by gravimetric analysis as described by Nep and Conway 20.

The method of CHN with calibration based on K factors was employed, a 2.361g of KSG was used for this experiment. The percent composition of Carbon, Hydrogen and Nitrogen in the sample were determined using Eager 300 elemental analyzer.

A 5mg quantity of KSG-VD and KSG- FD were individually encapsulated in hermetic pans and covered with hermetic lids (A39817-03/ USA) in order to measure their degradation, glass transition (Tg) temperatures and melting points (Tm)using Differential scanning calorimeter (DSC Q2000 V24.10 Build 122) connected to a Universal V4.5A TA instrument software. A heating rate of 10° C/ min was used and ramping was performed from-20 °C to 250 °C using nitrogen as purge gas 21.

A 4.74 g quantity of KSG-VD was placed on the DVS sample pan of an ultra-microbalance with a mass resolution of ± 0.1µg and dried under a stream of dry Nitrogen gas at 21°C. Humidity was increased in 10% RH steps to 100% RH for the sorption phase and then decreased in a similar fashion for desorption phase. This same procedure was performed on KSG-FD powder (3.29 g). The nature of the gum after drying led to the variation in weights of gum sample used for the experiment.

The conditions for GPC experiment were; Pump: 610 fluid units (Waters) + 600 Controller (Waters) Column setup: Guard column SB-G (Shodex) + 2x Ohpak SB806M-OH (Shodex) and a Detector: 410 RI detector (Waters) set at a flow rate of 1ml /min. The column and detector temperatures were set at 80°C and 50°C respectively. For the purpose of SEC, a 5.0 g quantity of KSG-VD was dispersed in distilled water to dissolve. The upper soluble part of this dispersion was pipetted onto petri dishes and then freeze dried as previously describe above in order to get a purer form of the gum. A 10 mg sample of gum was separately dissolved in 2 ml phosphate buffer solution (pH 7) and allowed to stand for 1h, then conditioned in warm water bath at 50°C for 30 min and further sonicated for 15 min to ensure complete dissolution. The solution was filtered using a syringe filter 0.45μm in size before injection into the GPC system (only a partial filtration of the gum dispersion was achieved). The retention time (RT), number average molecular weight (Mn), weight average molecular weight (Mw) and polydispersity index (PI) were determined.

This experiment was performed on purified KSG-VD solid sample as described by 21 with slight modification. Spectra were recorded using an ATR-FTIR spectrometer (Thermo Fisher Scientific, Nicolet iS5 ATR-FTIR spectrometer, Erembodegem-Aalst, Belgium) was repeated twice. A diamond ATR crystal was pressed against the sample and the spectra were collected in the range 4000 - 550 cm-1 range with a resolution of 4 cm-1 and averaged over 128 scans. An identity to the polysaccharide depending on its chemical structure and chain conformation was determined.

Images of the surface appearances of the gum samples were viewed at x100 and x1000 magnifications and recorded using a Scanning electron microscope Quanta 200F (FEI, Eindhoven, The Netherlands) operated at an acceleration voltage of 20 kV.

The unmilled KSG powders were directly fed into the dispersion funnel and the optical concentration was maintained at a range 5.0-8.3%. The average median particle size (D50), the specific surface area and span for KSG powders were determined using a laser diffraction (Master sizer, Malvern, Worcestershire, UK) at a feed rate of 3.0 g and jet pressure of 1.3 bars. The mean value of three determinations was used to plot a size distribution curve.

The bulk volume (V0) of 30 g of gum samples were recorded in a 100 ml measuring cylinder as well as the volumes after 500 (V500) and 1250 (V1250) taps in an automated tapping machine (J. Englesman, Ludwigshafen, Germany). The experiment was performed in triplicate according to USP specifications 22. The bulk and tapped densities, Hausner ratio and Compressibility index were calculated as given below;

2 Bulkdensity (Bρ)= mass of powder (g)/ bulk volume (ml) 3 Tappeddensity (Tρ)= weight of powder (g)/tapped volume (ml)

where tapped volume at V500 and V1250 were computed separately for the tapped volumes at V500 and V1250.

4 Hausner's ratio = tapped density / bulk density 

5 Carr ' s index = Tapped density-Bulk density  Tapped density X100

The physicochemical properties of the gums obtained either by vacuum or freeze drying are summarized in Table 1 .

The yield of KSG was unaffected by either of the drying techniques employed. The colour of gum powder was affected by the drying methods. Vacuum drying produced gum powder with a brown colour while the freeze-dried gum appeared cream coloured with more aesthetic quality which can be applicable where formulation is not required to be coloured.

Long resident time and temperature differences are two causes in producing dark and red coloured flax seed gum powder 19. The change in colour of KSG-VD can be attributed to temperature differences of the drying methods. Freeze drying imparts light weight, increases stability, surface area and solubility as well as porosity to materials subjected to the process 23, 24, 25 .These effects were observed as the drying method was changed from vacuum to freeze drying. A similar occurrence on plant gums experimented have been reported 14, 16. The pH of KSG in this work was found to be mildly acidic. This is in agreement with previous works 9, 15. Although a slight shift in pH increased solubility, a decrease in particle size increased the effective surface area of the gum particles in contact with aqueous milieu, this increased solubility and ionization of ionizable components of the excipient in water. The solubility of drug is often intrinsically related to drug particle size; as a particle becomes smaller, the surface area to volume ratio increases. The larger surface area allows greater interaction with the solvent which causes an increase insolubility 26. Basic excipients promote oxidation of susceptible drugs when used for their formulations 15. Therefore, acidic and neutral hydrocolloids are more widely used for pharmaceutical formulations 9. A significantly higher absorption was observed for the more amorphous of the two gums, KSG-FD changing faster from glassy to rubbery state. The moisture sorption kinetic profile showed the highest mass absorption (% change in mass) at RH 98 for both gum powders. The mass increase been 44.66% for KSG-VD and the most absorbed 57.93% KSG- FD. The mechanism of moisture uptake can be explained as initial surface adsorption at low RH and subsequently bulk absorption when exposed over a long period of time at RH 70%. Although, the measurement parameters for both samples was the same (2 cycles of 0-98 % RH, water). Water caused the amorphous content to crystallize at RH 30 point during desorption (cycle 2) due to change in crystal lattice (Figure 1 ) whereas an early recrystallization (RH 98, cycle 1) was observed for KSG-FD due to its hygroscopic nature (Figure 2 ). It is presumed that water in amorphous material is absorbed which therefore results in a higher water uptake and the high affinity for water may be as a result of instability of the amorphous content 27, 28. A gradual increase in water uptake created a hysteresis that remained throughout RH 0 to 98% (Figure 3 ). This is a typical hysteresis shape for many pharmaceutical and food materials. An obvious characteristic is that the gap remains over the entire partial pressure range. The hysteresis was between 0 to 90% RH in Figure 4, it revealed a good agreement between the adsorption and desorption branch at 90-98% RH with no significant hysteresis because the second cycle was identical to the first cycles at this point confirming that there was no irreversible effect.

Figure 1

Figure 2

Figure 3

Figure 4

Hysteresis is the gap or difference between adsorption and desorption isotherms in cases where they do not meet 29. Partially amorphous materials tend to exhibit considerable bulk absorption of water which is a slow, diffusion-controlled process and therefore non-equilibrium regime 30. The high percent moisture content of KSG-FD was expected due to increase solubility and porosity imparted by its drying method. However, this value still remained lower than the USP 2007 specification of 15% w/w for natural gums. Other authors,16, 17, 19 also noticed that freeze drying increased solubility as well as hygroscopicity on gums. Nonetheless, whichever method is employed in drying KSG, it is required to be stored in air tight containers to improve stability during storage and economic importance. The particle size distribution (PSD) curves of the gum samples (Figure 5 ) appeared positively skewed that is, the tail elongated towards the larger particles and therefore not a normal distribution where the central values (mean, median and mode) is equal.

The D10 and D90 values of the PSD curve showed that 10% of the particle population of KSG-VD are composed of fine and 90% of the population consisted of coarse particles. Similarly, D90 values of KSG-FD revealed 90% of particles are below 380.6 ± 48.3 µm and 10% (D10) of the population comprised of particles ≤ 20µm in diameter. The median particle diameter D50, revealed half of the particle population of KSG-FD to be much smaller than that obtained with KSG-VD (Table 1,Figure 5 ). During the extraction of KSG, two portions resulted, a fine soluble and a coarse insoluble layer. This large particles in KSG-FD are the insoluble coarse portion of the gum which was unaffected by freeze drying while increase in surface area was the result of the effect of freeze drying on the soluble particles. The coefficient of variation (COV) for the three replicate determinations of D10 and D90 were ≤ 15 % for the both gums dried by vacuum and freezing while D50 for KSG-VD was within the 10 % USP <429> specification, indicating excellent method repeatability, that of KSG-FD failed the requirement.

The SEM images (Figure 6 ) revealed differences in shape and crystal habits in the surface features of KSG-VD and KSG-FD powders and defined clearly the fineness and coarseness of the gums by a distinct difference in particle size and shape. The SEM images of KSG-VD particles revealed particles larger than 100 µm, almost twice the size estimate for 50% of the powder population of KSG-FD. An increase in specific surface area (SSA) of KSG-FD was noticed with the appearance of flaky irregular small particles in contrast to a more defined rectangular multifaced discreet particles of KSG-VD agreeing with its gritty texture (p<0.05), (Table 1 ). This implies that the SEM images represented well the data obtained via laser diffraction technique. Works performed were able to correlate data obtained from laser diffraction and SEM images of different grades of microcrystalline cellulose wherein fine and coarse particles were well estimated by both physical techniques and supported one another 31. The effects of drying methods on textural characteristics of gums was also reported 32.

Figure 6

The bulk density of a powder is primarily dependent on particle size, particle size distribution and particle shape 33. The density of gum powders was affected by the drying methods employed. The bulk and tapped densities of KSG-VD were higher than values obtained for KSG-FD (p<0.05) (Table 1 ). This implies that KSG-VD is denser and more consolidated and thus, will occupy less space during packing. When poured into cylinder coarser particles of KSG-VD were more consolidated and had excellent free flow compared to fines and flaky particles of KSG-FD powder that had poor flow as a result of the drying method. The values of Hausner ratio and Compressibility index further confirmed this observation. The Compressibility index and Hausner ratio are measures of the propensity of a powder to be compressed. As such, they are measures of the powder ability to settle and they permit an assessment of the relative importance of interparticulate interactions 34. According to Hausner, powders with low interparticle friction, such as coarse spheres, have ratios of approximately 1.2, whereas more cohesive, less free-flowing powders such as flakes have ratios greater than 1.6. Similarly, powders with 12-16% compressibility index are said to have good free flowing property whereas powders with Carr’s index of 23-28% are very fluid with poor flow properties 35. Again, the flow properties of these gum powders agree with the SEM and PSD results.

Carbon 37.83 %, Hydrogen 5.74 % and Nitrogen 0.85 % were elemental composition found to be present in KSG-VD. The nitrogen content indicated the presence of amino acids which has been found in plant gums in small and relatively large quantities ranging from 0.04 % in Acacia leucoclada to 5.6 % in Azadirachta indica gum. A nitrogen content of 0.56 % was reported in Khaya grandifolia gum, another species of Khaya 36, 37, 38. The higher the nitrogen content in the plant gum the more insoluble gel it forms in aqueous dispersions 39. The partial insolubility of KSG may also be attributed to the nitrogen content found in this gum. Other elements including magnesium, sodium, calcium, copper (II) ions and a high content of potassium ions were reported to be present in KSG 15.

Partially crystalline polymers give rise to very broad melting peaks because of the size distributionof the crystallites which can melt over a wide range of temperature 40, 41. The thermal curves were characterized by broadness and this was more pronounced in the DSC curve of KSG-FD (Figure 7 ), although sealed in a hermetic pan and lid, it melted due to bound moisture acquired from its drying process and showed an endothermic peak at 43.54°C and 196.30°C related to loss of moisture or dehydration and a somewhat melting peak 42, 43 In (Figure 8 ), the endotherm is comparatively less broad than that of KSG-FD (Figure 7 ). A glass transition (Tg) marked at 56.14°C and a melting point at 182.35°C with no evidence of degradation at the end of the transition was revealed. The Tg indicates the presence of amorphous content while the characteristic melting curve depicts crystallinity. Melting points are increasingly sharper with smaller sample sizes and purer substances due to crystalline rearrangement, fusion or solid-state transition. Impure substances often show several peaks 41 and such peaks were not seen in all the curves obtained for both KSG-VD and KSG-FD powders and could be an indication for a level of purification in the extraction stage. The gum powder will remain stable at room temperatures and above due to the high Tg. The slight differences in Tg and Tm of the KSG-VD and KSG-FD gum samples suggest structural and functional group differences in the samples, an effected of the drying methods. It should be noted that this is the first time DSC scans of KSG obtained by two drying methods are displayed.

Figure 7

Figure 8

The possible functional groups associated with band widths of ATR-FTIR spectra of purified KSG-VD (Figure 9 ) are shown in Table 2 . The presence of carboxylic acids supports the assumption of CO2 evolution as a mechanism of thermal transition of gums in the DSC thermograms 44. Similarly, the presence of carboxylic acid at 1726 cm-1 frequency in cashew gum was an indication of its presence in cellulose containing compounds 45. The results of ATR-FTIR corrobrated with elemental analysis as both showed the presence of amide groups. Further, the possible presence of Azo (-N=N-) groups at 1416 cm-1in KSG-VD suggests that they can be degraded in the small intestine where the enzyme diazo reductase is present and can be considered biodegradable 4, 46. This implies that KSG-VD can be useful in modified release especially in controlling drug release along the GI segments. In relation to sugar components of the gum, the hydroxyl, carbonyl and methyl functional groups represent the backbone structure of carbohydrates while C-O stretch at 1239 cm-1 represent aromatic compounds of rhamnose, galactose and galacturonic acid and the medium peak at 1725 cm-1 due to C=O may be related to galacturonic acid. Galacturonic acid was reported as a constituent of Khaya grandifolia gum 47, 48, 49.

Figure 10

Table 2

Bandwidth frequency	Possible functional group
1725 cm-1	C=O: stretch, different types: saturated aliphatic ketones, saturated aliphatic aldehydes, aryl esters, formates, sat aliphatic carboxylic acid dimers
1600 cm-1	N-H: primary amide, primary amine NH2+: amine salt CO2-: asymmetric stretch, carboxylic acid salts NH3+: amine salt (C=C: aromatic structure)
1416 cm-1	N=N: aromatic azo compound C-N: primary amide, amide III band C-O stretch and OH carboxylic acids O-H: primary, secondary alcohols
1239 cm-1	C-O: carboxylic acid dimers N-H: secondary amides, amide III C-O-C: saturated aliphatic esters C-O: acetic acid ester CH3COOR Cyclobutanes R-O-Ar (alkylthicketones)
1100-950 cm-1	Thioamides N-C=S CH3-C: aliphatic CH3-Ar Saturated aliphatic ethers
633 cm-1	N-H primary amides C=C-H bending, terminal acetylene groups N-C=O bending, primary saturated aliphatic amides, NO2, nitroalkane, Vinyl compounds

Table 3

Parameters	RT (min)	Mn	Mw	Mp	Polydispersity
Values (Da)	18,485	27,494	59,536	35718	2.17

Key: RT = retention time, Mn = number average molecular weight, Mw = weight average molecular weight, Mp= molecular weight of highest peak

Polymers are composed of repeated units of monomers chemically bonded together to form a long chain. Knowledge of the chain length is used in understanding their physical properties such as mechanical strength, solubility and brittleness. Chain length is usually expressed in relation to relative molecular mass of monomers and the number of monomers in a chain. Although all synthetic polymers polydisperse, they have unequal chain length and therefore the molecular weight is not a single value and has to be expressed as an average 50. Natural polymers on the other hand are often monodisperse i.e. they have molecules of uniform mass in a single chain 51, 52. A monodisperse curve with a peak molecular weight (Mp) of 36K Daltons due to Dextran as a calibration material was obtained for KSG (Figure 10 ,Table 3 ). The broadness of a molecular weight distribution of a polymer is expressed by its polydispersity index (Mw/Mn)is important to any research on cellulose and its derivatives because they influence dissolution, regeneration and accessibility or reactivity of the biopolymer 53. A dispersity value of 2.17 is indicative of step polymerization reaction. Polymerization usually occurs as a reaction between two functional groups e.g. carboxyl and hydroxyl groups while a step polymerization reaction involves a reaction between monomers with two or more functional groups. This finding supports the ATR-FTIR results with several functional groups present at an absorption peak. The peak molecular weight (Mp) is quoted for very narrowly distributed polymers such as those used for standard calibration. The Mp of 36 K Dalton and the retention time obtained in this experiment for KSG indicates that it composes of a Dextran -like polymer. The presence of Dextran-like material may be an indication for the non-gelling property of KSG and also its good disintegrant action that has been earlier mentioned. Polysaccharides as a group have wide range of properties from insoluble (cellulose) to soluble and high swellable (guar, starch); low viscous to highly viscous (acacia and guar gum), gelling agar to non-gelling dextran 54 To this end, KSG standing alone cannot be used to formulate extended release formulations but may act as a release modifier to hasten drug release in combination with other polymers.

Prof. Jean Paul Remon and Dr. Jurgen Vercruysse of the Laboratory of Pharmaceutical Technology Ghent University, Belgium are acknowledged for their support in the conduct of this research.