Materials

Metformin Hydrochloride was purchased from Macklin (Shanghai) Co., Ltd. (China). Microcrystalline cellulose (MCC), Polyvinyl alcohol (PVA), and Magnesium Stearate were procured from Sigma-Aldrich (USA). Mannitol was supplied by R&M Chemical (Malaysia), and sucralose was purchased from Nutra and Food. Quinine hydrochloride (QH) was obtained from J&K Chemical Ltd. (Shanghai). Prosolv ODT G2 was donated from JRS Pharma, and Pharmaburst 500 was gifted by SPI Pharma. Smart QD-EX 100 was gifted from Shin-Etsu Pharma, and Ludiflash was received as a gift from BASF Pharma. Fujimelt (Type-C) was donated by Fuji Chemical Industries Co., Ltd. Co-processed Met.HCl was laboratory-prepared by spray drying (as described in the Methods Sect. 2.2). A commercial Metformin Hydrochloride tablet (Glycomet® 250 mg), manufactured by USV Private Ltd. (India), was purchased and used as the comparator for tablet evaluation.

Preparation of Co-Process Met.HCl by Spray Drying

Met.HCl powder was co-processed by spray-drying to improve Met.HCl physicochemical powder properties. Aqueous polymer solution of 7% w/w PVA was prepared. Next commercial Met.HCl powder, 10% w/w mannitol, and 2.5% w/w sucralose were added and dissolved in the polymer solution. The volume was adjusted with water to a total solution of 20% w/v. The prepared solution was then dried using the Buchi Mini Spray Dryer B-290. The drying conditions for the spray were as follows: Inlet air temperature 170 °C, output air temperature 100 °C, aspirator air flow rate 700 (L/h), and peristaltic pump feed rate of 30%. Spray-drying products are collected, weighed, and stored in a desiccator over silica gel. The formulation composition and spray-drying parameters were based on previously optimized conditions established in an earlier work. The physical properties of the resulting co-processed Met.HCl powder, including flowability and moisture content, are summarized in supplementary materials Table S1.

Formulation Development of Metformin Hydrochloride Orally Dispersible Tablets (Met.HCl ODT)Formulation Design and Optimization of Met.HCl ODT by Design of Experiment

Design of experiment, D-optimal design was used as an approach for the design and optimization of Met.HCl ODT formulations. Two mixture factors, one continuous factor and one categorical factor, were applied in the design, making an experiment with a total of 30 runs (Table 1). The two mixture factors were % co-process Met.HCl and % co-process excipients. Meanwhile, one continuous factor was tablet compression pressure and co-process excipients brands as a categorical factor. The response variables studied in the design were powder properties (angle of repose, Carr’s index, and Hausner’s ratio) and tablet properties (hardness and disintegration). The statistically significant impact of main, interaction, and curvilinear effects of the chosen design variables on the target critical quality attributes was assessed with JMP® Pro Version 10.0.2 software (SAS Institute Inc., Cary, NC). The p-value, the coefficient of determination (R2), and R2 (adjusted) were utilized for evaluating the model quality. Significant p-values under 0.05 for both the model and the factors were set in the design. The utilization of the prediction profiler enabled the multicriteria optimization for the response variables. The accuracy of the optimized model was verified through the production of three replicate batches under the optimized conditions.

Table 1 Independent variable with levels and dependent variable in D-optimal designPreparation of Met.HCl ODT Formulations

The tablet formulations were designed by varying components with co-processed Met.HCl as an active ingredient. Different co-processed ready-to-mix excipients for ODT were used to formulate ODT. All components were weighed, mixed homogenously, and compressed into the tablets by direct compression in a single-punch manual tabletting machine. The mixture was compressed as a tablet using a 12 mm punch in a tablet punching machine, Enerpac GA3 Single Punch Machine (Globe Pharma, New Brunswick). Each tablet weighed 450–500 mg, which is equivalent to 250 mg Met.HCl.

Powder Mixture EvaluationsSolid-state Characterization and Compatibility a)

X-ray diffraction (XRD).

X-ray diffraction (XRD) patterns of pure Met.HCl powder and Met.HCl ODT powder mixture was recorded using MiniFlex X-ray diffractometer (Rigaku Corporation, Tokyo, Japan), with diffraction over the 5°–80° 2θ range at a 1.2 °/min speed.

b)

Differential scanning calorimetry (DSC).

Thermal analysis was conducted on Met.HCl ODT powder mixtures with excipients using a differential scanning calorimeter (DSC) DSC 6000 (PerkinElmer, Waltham, MA, USA). 1–3 mg of sample powder was weighed and placed in an aluminium sample pan. The sample was analyzed at 10 °C/min over a 30–270 °C temperature range with purging nitrogen gas at a 50 L/min flow rate to maintain an inert atmosphere.

c)

Fourier transform infrared (FTIR).

FTIR spectra were recorded using a FTIR spectrometer equipped with an attenuated total reflectance (ATR) accessory, Spectrum Two (PerkinElmer, Waltham, MA, USA), was used for compatibility evaluation. The sample was scanned over a 4000 –400 cm− 1 wave number range with a 4 cm− 1 resolution.

Particle Morphology and Size Distribution a)

Scanning electron microscopy.

The Met.HCl ODT powder mixture and compressed tablet were mounted on stubs with double-sided sticky tape of conductive carbon. Samples were coated with gold using a LEICA EM SCD005 sputter coater (Leica Microsystems, Vienna, Austria). Scanning electron microscopy (SEM) imaging was performed using a Quanta FEG 650 field emission scanning electron microscope (FEI Company, Hillsboro, OR, USA) at an acceleration voltage of 10 kV and a magnification of ×250, ×500, ×1000, and ×2000.

b)

Particle size distribution.

The particle size distribution was analyzed using a laser diffraction instrument (Mastersizer 2000, Malvern Panalytical, UK) to determine the median diameter (Dv50), the distribution width (Dv90/Dv10), and the interquartile coefficient of skewness (IQCS). The width of the particle size distribution, referred to as the span, was determined to evaluate the polydispersity of the powder. The span was calculated according to Eq. (1) [22].

$$\:Span\:=\:\left(Dv90\:-\:Dv10\right)\:/\:Dv50$$

(1)

Where Dv10, Dv50, and Dv90 are the particle diameters at which 10%, 50%, and 90% of the sample volume is smaller than that diameter, respectively.

Powder Flowability and Density a)

Angle of repose.

The powder mixture was passed through a funnel that was fixed to the nozzle stand at a fixed height of 4 cm. The diameter of the powder mixture was measured, and the slope angle was calculated using Eq. (2) [23].

$$\:\theta\:=}^\left(\:h/r\:\:\right)\:\:$$

(2)

Where h is the height of the pile of powder mixture and r is the radius of the pile of powder mixture.

b)

Hausner’s ratio and Carr’s index.

The bulk density (ρ₀) and tapped density (ρₜ) of the co-processed powder were measured using a tap density meter, BKDT-100 C (Biobase, Jinan, China). A 25 mL glass cylinder was filled with 20 cm³ of the powder sample, and its weight was recorded. The cylinder was then placed on the tapped density tester, and the bulk density was measured prior to tapping. The Carr’s index and Hausner’s ratio were calculated using Eq. (3) and Eq. (4) [23].

$$\:Car^}s\:index=\:(t-b/t\:)\:\times\:\:100\:\%$$

(3)

$$\:Hausne^}\:ratio=\:\text\:/\text$$

(4)

Where ρt represent tapped density and ρb resresent bulk density.

c)

True density.

True density (ρT) was determined by the Helium pycnometer (AccuPyc 1330, Micromeritics, USA) (Kibbe, 2000). The powder sample was loaded into the sample cell and weighted. The sample volume was calculated by measuring the pressure observed by filling the sample chamber with ultra-high-purity helium gas and then discharging the gas into a second empty chamber. The measurement was repeated for ten of these cycles.

Compressed Tablet EvaluationsHardness and Friability Test a)

Hardness test.

A hardness test was conducted using the hardness tester, EBT-2PL (Torontech, Beijing, China). 10 prepared tablets from each formulation were selected and tested with a hardness tester. The force (N) required to break the tablets was recorded.

b)

Friability test.

The evaluation of friability was conducted through the friability tester, TAR II (Erweka, Heusenstamm, Germany). A selection of tablets equivalent to ± 6.5 g was weighed. The total weight was denoted as the initial weight, Wi. Subsequently, the tablets were placed in the friability tester and subjected to 100 rotations at a speed of 25 rpm. Following this, the weight of the tablets was recorded as the final weight, Wf. The percentage of weight reduction was then computed utilizing the equation specified as Eq. (5) [24].

$$\:Percentage\:of\:weight\:loss=\left(\right(Wi-Wf)/Wi\:\:\:)\:\:\times\:\:100\:\%$$

(5)

Disintegration, Wetting and Moisture Content Test a)

Disintegration test.

The tablet disintegration test was performed using a tablet disintegration tester (Pharma Chem Machineries, Mumbai, India). The test was conducted in 900 mL of phosphate buffer, pH 6.8, maintained at a temperature of 37 ± 2 °C. Six tablets were randomly selected and individually placed in each of the six tubes on the basket rack. Discs were then placed on top of the tablets, and the basket rack was immersed in the water. The time required for complete disintegration, with no visible residue remaining, was recorded.

b)

Wetting test.

The wetting test was carried out by adding 2 mL of phosphate buffer (pH 6.8) into a petri dish containing a piece of Whatman filter paper. Tablets were then placed on the moistened filter paper, and the time taken for each tablet to become fully wetted was recorded.

c)

Moisture content.

The content of moisture residue in the compressed tablet was determined using a halogen moisture analyzer (HB43-SSD, Mettler-Toledo GmbH, Giessen, Germany).

Uniformity of Weight and Content a)

Uniformity of weight.

The twenty tablets from the batch were randomly selected. Subsequently, the cumulative weight of these selected tablets and the mean weight were calculated. Next, the mass of each individual tablet among the twenty was recorded, and the difference in mass from the average mass was calculated.

b)

Uniformity of drug content.

Ten tablets were randomly chosen and weighed. The tablets were then crushed and mixed uniformly in a mortar, using a pestle. The crushed tablet powder corresponding to 100 mg of drug was dissolved in 100 ml of phosphate buffer with a pH of 6.8 within a volumetric flask. 1 mL of the mixture was diluted with pH 6.8 phosphate buffer, followed by the determination of the absorbance of the diluted mixture using a UV–Vis spectrometer, Cary 60 (Agilent Technologies, Santa Clara, CA, USA) at 233 nm. The proportion of drug concentration was then computed using Eq. (6).

$$\begin \\&\:Percentage\:of\:drug\:content=\\&Amount\:of\:drug\:\left(mg\right)/Average\:drug\:content\:\left(mg\right)\times\\&\:100\:\% \end$$

(6)

The consistency of the drug distribution was evaluated using the Relative Standard Deviation (RSD) and the Acceptance Value (AV), calculated using Eq. (7) and Eq. (8) [25].

$$\:RSD\:=\:\left(s/\overline\right)\times\:\:100\:\%$$

(7)

$$\:AV=\left|M-\overline\right|+ks$$

(8)

Where \(\:\overline\) is the mean of individual contents expressed as a percentage of the label claim. s is the sample standard deviation. k is the acceptability constant (2.4 for n = 10). M is the reference value (typically 100% if the mean is within 98.5%–101.5%).

Dissolution Test

Dissolution apparatus Type 2, DIS8000 (Copley, Nottingham, UK) was used. 900 mL of phosphate buffer at pH 6.8, 37 ± 5 °C, was utilized as the dissolution medium. Six tablets were inserted into the six vessels of the equipment, and the paddles were activated at a velocity of 70 rpm for a duration of thirty minutes. Every two minutes, 5 mL of the dissolution sample was extracted from each vessel and analyzed using a UV-spectrophotometer at 233 nm. The amount of drug from each sampling will be calculated based on the percentage of drug content.

The dissolution profiles of the test (Met.HCl ODT) and reference formulations were compared using the difference factor (f1) and similarity factor (f2). The f1 was calculated using Eq. (9), and f2 was calculated using Eq. (10) [26].

$$\:f1=\left(\frac_^\left|_-_\right|}_^_}\right)\times\:100$$

(9)

$$\begin\\& \:f2=50cdot\:log\\&\left\\sum\:_^_-_\right)}^\right]}^cdot\:100\right\} \end$$

(10)

Where Rt​ and Tt​ represent the percentage of drug dissolved at time t for the reference and test products, respectively. n is the number of time points considered.

The dissolution efficiency (DE %) was calculated using Eq. (11) [27].

$$\:DE=\left(\frac_^y\:dt}\right)\times\:100$$

(11)

Where ? is the percent dissolved at each time point, ? is the final sampling time, and ?100 is the theoretical 100% dissolution.

Compressibility and Mechanical Properties

250 mg of powder was punched by the Enerpac GA3 Single Punch Machine (Globe Pharma, New Brunswick) at different compression pressures, ranging from 400 to 2000 psi. The tablet’s dimensions and weight at different compression pressures were recorded.

a)

Heckel analysis.

The compression properties of tablets were evaluated using Heckel Eq. (12). A Heckel plot of ln(1/1-D) as a function of compression pressure, P, was plotted, and the constants from the plot were determined. Relative density calculated based on Eq. (13) [28].

$$\:ln\:\left(1/1-D\right)=kP+A\:\:$$

(12)

Where D, the relative density of powder compacts at its compaction pressure, k, and A are constants from the slope of the heckle plot.

Where ρA is the apparent density of the compact at pressure P and ρt is the powder tapped density.

b)

Kawakita analysis.

Next, to further evaluate the compressibility of the powder, the Kawakita model was employed. This model describes the relationship between the applied pressure (P) and the degree of volume reduction (C). The Kawakita equation is expressed as Eq. (14)

$$\:P/C\:=\:P/a+\:1/ab$$

(14)

Where a represents the compressibility (the capacity of the powder to reduce in volume) and b is a constant related to the cohesiveness/plasticity of the particles. The degree of volume reduction (C) was determined by comparing the initial bulk density ρ0 to the apparent density ρA of the compact at pressure P, according to Eq. (15).

A plot of P/C against P was used to derive these constants from the slope (1/a) and the intercept (1/ab).

c)

Mechanical properties.

The force (N) required to break the tablets was determined using EBT-2PL (Torontech, Beijing, China). Next, the plot of porosity as a function of compression pressure, tensile strength as a function of porosity, and tensile strength as a function of compaction pressure was plotted. Tensile strength was calculated using Eq. (16), and porosity (ɛ) was calculated using Eq. (16)

\(\:\:T=\:2F/\pi\:dt\)(16) Where F, d, and t represent the hardness, diameter, and thickness of the compact, respectively.

$$\:\:=\:1-\left(\text/\text\right)$$

(16)

Where \(\:\), tablet porosity \(\:\text\) is the apparent density at compaction pressure P, and \(\:\text\) is the true density of the powder.

Bitterness Evaluation

The taste masking of the Met.HCl ODT was assessed by a six-member panel using the method outlined in the European Pharmacopoeia 7.0. This study has been reviewed and approved by the Medical Research Ethics Committee UMMC (202479 − 13897).

Individual panelists’ sensitivity to bitterness was determined using serial dilutions of a standard quinine hydrochloride solution, from which a correction factor (k) was calculated to standardize perception differences Eq. (17). Met.HCl ODT powder mixture stock solutions were prepared and serially diluted, and panelists identified the lowest concentration provoking a bitter sensation (solution C(x)), which was further tested in a second dilution series (D1–D6) to determine the minimum volume eliciting bitterness. The bitterness value (BV) of the co-processed Met.HCl was calculated using the panelists’ threshold (k), the dilution factor of C(x) (Yd), and the smallest volume of C(x) used in the secondary dilution (X′) according to the established equation, providing a quantitative measure of taste masking efficiency. The bitterness value was calculated using Eq. (18).

Where n is the number of millilitres of the stock solution in the dilution of the lowest concentration that is judged to be bitter.

$$\:Bitterness\:value\:=\frac^}\times\:0.1}\:\:$$

(18)

Where Yd is the dilution factor for solution C(x), X’ is the lowest volume of solution C(x) used to make solution D, and k is the correction factor based on the taste evaluation [29].

Accelerated Stability Study

Accelerated stability study was conducted on the optimized Met.HCl ODT formulation using stability chamber, ClimaCell® 222 ECO Line (BMT, Germany). The tablet stability was tested at 40 ± 5° C/ 75% Relative humidity (RH) ± 5% condition in the stability chamber for six months as recommended by ICH guidelines. The tablet was then evaluated for dimensions, friability, hardness, content, disintegration time and dissolution.