Improved synthesis of metformin hydrochloride-sodium alginate (MH-NaALG) microspheres using ultrasonic spray drying

Abstract

Metformin Hydrochloride (MH), an orally administered antidiabetic drug from the biguanide family, encounters issues of wide particle size distribution, inefficient dissolution rates and short half-life leading to excess dosage which can result in lactic acidosis. Novel approaches that yield smaller particle size and uniform distribution at higher yields are significant to tackle problems associated with solubility and optimum dosage levels of the administered drugs. In the current research related to microsphere synthesis, a controlled process based on pressure and ultrasonic nozzles for the atomization of liquid, was applied with an objective of optimizing particle size of microspheres of MH in sodium alginate, a biopolymer excipient matrix. The study carried out in spray dryer elucidated parameter optimization using one variable at a time approach by varying important parameters as inlet temperature of air (120°C–150 °C), rate of flow of feed (1.5 mL/min-3 mL/min), aspirator rate (800 rpm–1400 rpm) and polymer content in feed solution (1 g–8 g) using ultrasonic and pressure nozzles for comparison with the target output parameter as particle size and yield. While the particle size at optimum conditions were <10 μm for both types of atomization, ultrasound assisted spray drying exhibited narrower distribution compared to the pressure atomization. Particle characterization performed using SEM, optical microscopy, FTIR, XRD and DSC revealed slight deformation with no chemical interaction and slight decrease in crystalline nature of pure drug. Overall, an improved process based on the use of ultrasound with optimized parameters has been demonstrated for synthesis of MH containing microspheres.

1. Introduction

Metformin Hydrochloride (MH), an orally administered antidiabetic drug from the biguanide family that is used for treatment of type 2 diabetes mellitus, has low bioavailability and short half-life when following typical dose formulations. Particles having size more than 10 μm finds difficulty in penetrating mucous layer of gastrointestinal tract [1]. Hence, such drug delivery systems require the drugs to be consumed numerous times a day to maintain optimum plasma levels and achieve the desired response. The excess amount of MH consumed to maintain the dosage level in blood due to bioavailability of 40–50% [2], might lead to lactic acidosis [3,4]. Rapid increase and drop in drug level between dosage intervals and pharmaceutical wastage necessitate use of novel drug synthesis or delivery techniques that ensure controlled release to meet the dosage levels, solubility, and bioavailability requirements of administered drug in blood. Among the many advances made to address such concerns, encapsulation using microspheres [[5], [6], [7], [8]] is a recent approach.

Microspheres are spherical particles with drug mixture distributed in a polymer matrix and with size ranging from 1 to 1000 μm. Polymer matrix, that protect medications from environmental impacts such as heat, moisture, oxidation along with masking unpleasant odour and taste [9], can be natural (gelatine, collagen, lectins, alginate, pectin, chitosan etc) or synthetic (PGA, PCL, PLA, PLGA, EVA, silicones etc.) in nature [10]. The degradation of the matrix ensures precise distribution of optimum concentrations of active component at target sites and prevents wastage at sites other than the target sites thereby reducing undesirable side effects [11]. Alginate finds a wide variety of applications ranging from food additives to wound dressings because of its low cost, nontoxicity, biocompatibility, and biodegradability [12]. For example, Revaprazan, a drug used in treatment of peptic ulcer was formulated into solid lipid nanoparticles (SLNs) of sodium alginate to enhance the therapeutic efficiency which was earlier challenged by low dissolution and short half-life [13]. The SLNs with a particle size of 120 nm ensured drug incorporation of around 88%. By enhancing the gelation and buoyancy alginate SLNs potentially improved the bioavailability of Revaprazan.

Using a mucoadhesive drug delivery medium like sodium alginate provides for longer residence times, greater drug absorption, and increased bioavailability, as well as a reduction in medication dose frequency, due to their close contact with the absorption surface [14,15]. To optimize the therapeutic impact of a drug taken orally, dosage forms should have a longer residence period on the gastrointestinal mucosa, mediated by mucoadhesive polymers. The longer residence times allow for a sustained release of the drug at a specific target site. The occurrence of non-covalent bonds between the mucus gel layer and polymers, such as hydrogen bonds, ionic interactions, or physical entanglements, provides muco-adhesion. Additionally, drug delivery systems can be tailored to a specific target for the purpose of local therapy or drug release at the area where drug absorption is required. Furthermore, mucoadhesive polymers can ensure close contact with the absorption membrane, which lays the groundwork for a strong concentration gradient to act as an initiator for passive drug uptake [16].

Microsphere formulation mainly involves incorporation of drug into polymer solution followed by droplet formation and solvent removal before collection and storage of particles [17]. Popular microsphere synthesis techniques include single/double emulsion formation, spray drying, polymerization, ionic gelation and spray freeze drying methods [11,18]. Compared to the alternatives, spray drying is a swift one-step process with least dependency on active material and matrix polymer solubility [19]. The capacity of spray drying to exercise control over particle size and maintain low moisture content along with elimination of residual organic solvent, if any, are both beneficial in the synthesis of multi-particulate drug delivery systems [2]. Additionally, due to the evaporation effect, the temperature of the finished product is lower than that required for other drying methods. There have been some literature available on the use of spray drying in pharmaceutical drug synthesis.

Spray drying technique was used to encapsulate Revaprazan in gelatine to improve solubility and oral bioavailability [20]. The solubility of the drug increased by 150-fold and oral bioavailability by 2.3 times when the drug and polymer was used in the ratio 1:2 to formulate spherical and amorphous Revaprazan loaded gelatine microspheres. In another study, amorphous and crystalline microspheres of Sildenafil loaded in different polymers were spray dried using pneumatic nozzle of 0.7 mm diameter at an inlet temperature of 90 °C, air pressure of 4 kg/cm² and 100% aspiration [21]. It was reported that the size of drug powder reduced significantly from 26.6 μm to 9.3 μm for amorphous microspheres due to dissolution in ethanol whereas it remained around 27.8 μm in crystalline microspheres due to improper dissolution in water. Self-nanoemulsifying drug delivery systems (SNEDDS) composed of oil and surfactant in this study exhibited superior drug delivery characteristics compared to other microenvironments. In a continuation study, impact of carrier hydrophilicity on SNEDDS was analysed by using calcium silicate and hydroxypropyl-β-cyclodextrin as hydrophobic and hydrophilic carrier respectively [22]. Though the carriers exhibited difference in terms of appearance, molecular interaction and solubility, it was elucidated that the properties of SNEDDS were not significantly affected.

Size of the particles which had undergone spray drying is determined by several factors, including the atomization type, feed flow rate, aspiration rate, inlet temperature and liquid properties [23]. The uniformity in microsphere size can be attained by adjusting the spray drying conditions to produce uniform droplets. Low energy requirement for drying makes spray drying a more cost-effective method for the generation of microspheres when compared to other techniques like freeze drying. However, usefulness of spray drying in microsphere synthesis are sometimes restricted due to drawbacks such as degradation of actives at higher temperature, inability to spray viscous feed and mediocre yield for small batch processing.

Mostly spray drying employs pressure [2,19,24,25] or two-fluid atomization [26,27] for droplet formation. Even though these methods are associated with smaller particle size and adequate efficiency, the particle size distribution are often compromised along with higher energy requirements (two-fluid nozzle >centrifugal nozzle >pressure nozzle) [28,29]. Use of alternate and efficient energy source like ultrasound helps in narrowing size distribution and reducing energy requirements with the added benefit of small particle size, low spraying speed, high rate of evaporation, operation at low temperature and minor mechanical stress. Conjunction theory proposed by Bouguslavskii and Eknadiosyants by incorporating cavitation and capillary wave hypothesis associated with ultrasound explains mechanism of ultrasonic atomization and droplet size dependence on capillary wave amplitude as well on liquid vapour pressure, gas content in liquid and static pressure etc. According to this hypothesis periodic hydraulic shock, generated as an after effect of cavitation due to ultrasonic vibration induced disturbance, interacts with capillary waves having fixed amplitude, exciting them to produce droplets [30,31].

Metformin hydrochloride (MH) encapsulation to produce spray dried microparticles have gained popularity in the past years. Sander et al. [2] studied the oromucosal delivery of MH loaded chitosan particles produced via spray drying. The effect of polymer (chitosan) concentration and molecular weight on the particles obtained by pressure spray drying at an inlet temperature of 200 °C and compressed air flow rate of 0.5 m³/h using a nozzle of 0.7 mm was studied and it was reported that particles with D90 ranging from 17 μm to 33 μm could be successfully obtained accompanied by high encapsulation efficiency (82%–89%). Similarly, Szekalska et al. [25] reported that pressure drying of 9 different formulations of MH and alginate with different alginate concentration (1%, 2%, 3%) and drug:polymer ratio (1:1, 1:2, 2:1) at an inlet temperature of 200 °C and flow rate of 4.5 mL/min resulted in microparticles in the range of 1.6 μm–5.7 μm. Twin fluid atomization had also been employed for spray drying by Al-Zoubi et al. [26], for co-spray drying of MH in a variety of polymers including sodium alginate using a standard twin fluid nozzle of 406 μm. Microparticles in the size range of 6.7 μm–11 μm were successfully obtained by operating at an inlet drying temperature of 130°C–135 °C.

Analysis of the literature related to the MH microsphere synthesis revealed that ultrasonic atomization based technique for spray drying have not been explored clearly elucidating the novelty of the current work related to the use of atomization using ultrasound for synthesis of spray dried MH microspheres. The pressure atomization technique followed by drying leaves behind a wide particle size distribution which is not considered ideal for drug delivery. Hence to bridge this gap we replace pressure assisted atomization with a greener technology like ultrasonic atomization.

The overall aim of current study was to synthesize MH loaded sodium alginate microspheres using spray drying based on the use of novel ultrasonic nozzle with an objective of controlling the size distribution, which has significant effect on drug characteristics. The process was optimized based on the variations in the process parameters (feed flow rate, aspiration rate, inlet temperature) and feed composition to achieve best output in terms of particle size, distribution, yield and encapsulation efficiency. Additionally, performance and size distribution of particles produced while using a cleaner technology like ultrasound for atomization was compared with conventional pressure energy based atomization. Finally, the characteristics of the loaded particles were also quantified so as to check if any changes occur due to the use of ultrasound. The work clearly aims at elucidating the engineering improvements in the synthesis process based on the use of ultrasound and understanding into the effect of various operating conditions.

2. Materials and methods

2.1. Materials

MH (white crystalline powder, 98% purity, molecular Weight: 165.6 g mol⁻¹, IUPAC name: 1,1- Dimethylbiguanide hydrochloride) and low viscosity sodium alginate polymer (white to cream coloured powder, 30–90 mPas in 1% aqueous solution) was obtained from ZetaScientific and Alfa Aesar respectively. All other analytical grade chemicals required for experiments like monobasic potassium phosphate, dibasic potassium phosphate, sodium hydroxide, acetone and hydrochloric acid were procured from Thomas Baker Chemicals Pvt. Ltd. Mumbai. Millipore Milli-Q Gradient water purification unit was used to get deionized water needed to make standard solutions of desired concentrations for experimental studies.

2.2. Microsphere formulation

200 mL of 1% (w/v) MH aqueous solution was prepared and subsequently solid content was adjusted to 1.5% (w/v) by adding sodium alginate. The mixture was stirred at 700 rpm for 30 min using an overhead stirrer to get a uniform mixture of MH and sodium alginate. The resulting mixture was then pumped through a spray dryer unit obtained from JISL instruments, Mumbai, involving atomization using an ultrasonic nozzle of diameter 1 mm equipped with titanium probe of 40 kHz and maximum power output of 80 W (Oscar Ultrasonics, model no – PR40 N). For understanding the temperature effect, studies were performed using 120 °C, 130 °C, 140 °C and 150 °C as hot air inlet temperatures. After fixing the optimum temperature, feed flow rate studies were carried out at 1.5 mL/min, 2 mL/min, 2.5 mL/min, and 3 mL/min. Aspirator rate study was carried out after fixing the optimum temperature and flow rate and by varying the aspirator rate with different values as 800 rpm, 1000 rpm, 1200 rpm and 1400 rpm. After optimizing the aspirator rate, polymer content in the feed was altered as 1 g, 2 g, 4 g, and 8 g, with drug content fixed as 2 g, to study the effect of polymer concentration. Studies were also performed using a pressure nozzle at 2 bar pressure to compare the efficiencies of both nozzles. The product obtained in the collecting flask was then removed, weighed, and desiccated for future analysis.

The operating equipment operating conditions have been selected on the basis of the available range of parameters and not selecting too low or too high values in the given equipment specified range. Preliminary studies indicated that above the maximum flow rate selected in the current work, simultaneous atomization and drying was challenging because of which most of the feed was collected in residue flask. Similarly, use of the speed above 1400 rpm resulted in improper drying. The minimum of range was decided as the minimum values attainable in the equipment. The minimum and maximum amount of the polymer used in the study was also decided based on literature and the preliminary study for deciding efficient equipment operation. Sosnik and Seremeta [32], stated in the review paper that one of the challenges associated with spray drying is the inability to handle highly viscous solutions and also much lower encapsulation efficiency at significantly lower polymer content. Preliminary studies indicated that when the polymer content of higher than 8 g was used in the work, problems in pumping and atomization of the feed was observed.

All the experiments related to understanding the effect of operating conditions were performed two times to check the possible experimental errors in the measurements. It was observed that the obtained data was quite reproducible with the observed experimental errors within the range of ±2% of the reported average value.

2.3. Particle characterization

Samples of commercial MH, Na-ALG and microspheres synthesized using both ultrasonic and pressure atomization were characterised using different techniques for understanding morphology and variations in crystal structure and functional groups.

2.3.1. Morphology

The morphology of pure MH and Na-ALG were compared with the microspheres using Scanning Electron Microscopy. Quanta 200 - SEM from FEI Company, with a tungsten emitter was used for the purpose. SEM images between 500× and 5000× were produced for microspheres synthesized by both methods. Additionally, Olympus 51X-TF optical microscope was used to analyse images of each sample at 45× and 100× resolution.

2.3.2. Particle size distribution

Bettersizer 2600E particle analyser (Bettersize Instruments Ltd., China), which operates based on the concept of laser diffraction, was used for particle size measurement using acetone as solvent medium in wet module. A 0.1% (w/v) solution was sonicated for about 4 min before analysis to eliminate any possibility of agglomerations. All analysis were carried out at an obscurity of 0.5 using the software, Bettersize Laser Particle Size Analyzer System Ver 8.0, to give a particle size distribution curve with detailed information on particle diameter and span. For the current study, Saute Mean Diameter or D (3,2) is considered appropriate for comparison since the active surface area of the particle is important for proper encapsulation and to ensure uniform particle size distribution.

2.3.3. Encapsulation efficiency

10 mg of MH loaded alginate microspheres were made up to 10 mL using phosphate buffer at 7.4 as the pH [33]. The solution was sonicated for 15 min followed by centrifugation for 15 min to collect supernatant devoid of the polymer debris [34] and diluted 100 fold. The supernatant collected was evaluated for drug content using UV–Visible spectrophotometer by measuring the absorbance at ƛ_max of 232 nm. The absorbance values were compared with the calibration plot for pure MH to obtain the actual amount of drug present in microspheres.

The drug encapsulation efficiency or EE (expressed in %) of alginate loaded MH microparticles was determined using formula given in equation (1).

$$EE\left(\%\right)=\frac{\text{Actual}\text{amount}\text{of}\text{drug}\text{in}\text{microspheres}}{\text{Theoretical}\text{amount of}\text{drug}\text{used}\text{in}\text{microspheres}}\times 100$$ (1)

2.3.4. Yield

Yield was calculated to compare the efficiency of both types of atomization. All weights were recorded at a temperature of 25 °C. The yield of the obtained product was calculated using equation (2) as give below:

$$\text{Yield}=\frac{\text{Weight}\text{of}\text{product}\text{obtained}}{\text{Weight}\text{of}\text{solids}\text{initially}\text{taken}}\times 100$$ (2)

2.3.5. XRD analysis

To compare the crystallinity of the pure sample and microspheres using XRD analysis, Rigaku Smartlab SE Powder X-ray diffractometer (Rigaku Corporation, Japan) was used. Monochromatic CuKα radiation of wavelength 0.15419 nm was used to record diffractograms across 5° to 80° at a scan speed of 15° per minute.

2.3.6. FTIR analysis

The functional groups in pure sample and MH microspheres were compared using the Bruker Alpha II Compact FTIR (Bruker lab, India) equipment. KBr pellet method was employed for studying chemical interaction between drug and polymer, with wavenumber for scanning ranging from 400 to 4000 cm⁻¹.

2.3.7. DSC analysis

DSC thermograms were captured using a TGA -55 thermal analyser from TA instruments and used to examine the behaviour of the formulations under heat. Under a nitrogen purge, samples were heated continuously in airtight aluminium pans across a temperature range of 35 °C–600 °C. Thermal peaks of all samples were compared to identify physicochemical interactions.

3. Results and discussion

3.1. Morphology

Microspheres subjected to SEM analysis displayed spherical shaped particles with dimpled and shriveled outer surfaces as shown in Fig. 1a–d. Vehring et al. [35] in his study explained the shriveling of surfaces using Peclet number (Pe) based analysis. Peclet number is defined as the rate of advection of a physical quantity by flow to the rate of diffusion of the same quantity. Vehring et al. [35], in his work defined Pe number as:

$$Pe=\frac{k}{8D_i}$$

where k is the evaporation rate and Di is the diffusion coefficient of solute.

Fig. 1.

SEM images of ultrasonically spray-dried microspheres at a) 1000×, b) 2000× magnification and pressure-dried microspheres at c) 1000×, d) 2000×.

In the case of polymers, which typically have higher Pe, diffusional motion of solute is slower resulting in surface enriched with solute. Depending on the polymer, different solidification mechanisms are initiated resulting in the formation of a shell which might have dimpled or shriveled surface depending on the rate at which the shell forms.

The SEM images also confirmed that the particles existed mostly as small evenly sized particles or as smaller particles adhered to crevices of a single larger sphere. Comparison of SEM images of ultrasonic (Fig. 1a and b) and pressure (Fig. 1c and d) dried microspheres concluded that size distribution was more uniform and less agglomerated in the case of ultrasonic assisted approach. It was clearly elucidated that the conversion of cylindrically shaped pure MH crystals to spherical shaped microspheres on spray drying is observed which increases the surface area available for absorption in subsequent applications.

3.2. Particle size analysis

3.2.1. Effect of inlet temperature

Temperature studies for particle size optimization were carried out using the inlet temperature range of 120 °C–150 °C. Other process and flow parameters including feed flow rate, aspirator rate and feed polymer content were kept constant at 2.5 mL/min, 1200 rpm and 2 g respectively. The temperature difference between droplet wet-bulb temperature (T) and gas temperature ($T_{\infty })$ determines the heat transfer rate and for a fixed relative humidity of the gas, $\Delta \left(T~T_{\infty }\right)$ rises with $T_{\infty }$, increasing the evaporation rate and subsequently Pe [36]. Higher Peclet number corresponds to formation of wrinkled microspheres with a hardened crust.

At higher temperatures, due to rapid drying, the surface is coated with a crust of soluble materials. The crust formed prevents further shrinkage of microparticles and evaporation of solvent thereby increasing the size of microparticles from 120 °C to 150 °C as reported in Table 1, Table 2 (also depicted in Fig. 2a). Since the boiling point of the solvent (water) was 100 °C, drying temperature setting below 120 °C resulted in incomplete drying along with large amount of residue collection and hence were not considered. An inlet temperature of 120 °C was thus shown to achieve minimum particle size in ultrasonic (10 μm) as well as pressure (4.6 μm) nozzle. Studies conducted by Aghbashlo et al. [37]; Chegini and Ghobadian [38]; Chegini and Ghobadian [39] and Jumah et al. [40] showed similar results. Aghbashlo et al. [37] studied fish oil encapsulation, using two-fluid nozzle spray drying and reported that a change in the temperature from 140°C to 180 °C resulted in slight increase in particle size from 4.12 μm to 6.6 μm because of crust formation at higher temperatures which prevented further shrinkage. Similar explanation was given by Chegini and Ghobadian [38], while discussing the increasing particle size trend of dried orange juice with and without additives like maltodextrin and glucose. In the presence of additive, when temperature was raised from 110 °C to 190 °C, particle size increased from around 6 μm–9 μm and from 20 μm to 30μm in absence of additive [39]. Jumah et al. [40] also observed an increase in particle size of pressure spray dried jameed from 23.1 μm to 27.16 μm when temperature was raised from 80 °C to 150 °C. It was also reported that increased inlet temperature favoured aroma and feed solid retention through increased initial drying rate.

Table 1.

Particle size variation for metformin chloride coated with sodium alginate with inlet temperature, aspiration rate, flow rate, and drug: polymer ratio along with the yield and encapsulation efficiency data using an ultrasonic nozzle.

Parameter Studied	Temp (°C)	Flow rate (mL/min)	Asp Rate (rpm)	Drug: Polymer	D (3,2) (μm)	Yield	EE
Inlet Temperature	120	2.5	1200	2	10.00	19.83	53.90
	130	2.5	1200	2	11.37	15.77	56.14
	140	2.5	1200	2	12.04	13.97	51.81
	150	2.5	1200	2	16.23	11.13	45.98
Flow rate	120	1.5	1200	2	7.57	46.17	58.20
	120	2	1200	2	8.11	31.10	62.91
	120	2.5	1200	2	10.00	19.83	53.90
	120	3	1200	2	13.25	11.20	49.77
Aspiration Rate	120	1.5	800	2	7.50	48.27	57.22
	120	1.5	1000	2	7.57	46.17	58.20
	120	1.5	1200	2	8.14	31.10	62.41
	120	1.5	1400	2	8.37	34.93	61.01
Weight of Polymer	120	1.5	1000	1	7.57	46.17	58.20
	120	1.5	1000	2	7.26	48.20	56.14
	120	1.5	1000	4	8.66	39.84	61.96
	120	1.5	1000	8	11.72	32.40	64.26

Table 2.

Particle size variation for metformin chloride coated with sodium alginate with inlet temperature, aspiration rate, flow rate, and drug: polymer ratio along with the yield and encapsulation efficiency data using a pressure nozzle.

Parameter Studied	Temp (°C)	Flow rate (mL/min)	Asp Rate (rpm)	Drug: Polymer	D (3,2) (μm)	Yield	EE
Inlet Temperature	120	3	1200	2	4.62	37.50	37.02
	130	3	1200	2	5.71	34.67	40.16
	140	3	1200	2	7.64	29.80	34.88
	150	3	1200	2	7.11	14.37	28.79
Flow rate	120	2	1200	2	3.93	55.20	44.75
	120	2	1200	2	3.99	46.80	42.19
	120	3	1200	2	4.62	37.50	37.02
	120	3	1200	2	5.73	28.07	30.69
Aspiration Rate	120	2	800	2	3.70	56.57	40.43
	120	2	1000	2	3.93	55.20	44.75
	120	2	1200	2	4.36	54.80	42.19
	120	2	1400	2	4.47	54.10	44.40
Weight of Polymer	120	2	1000	1	3.93	55.20	44.75
	120	2	1000	2	4.06	54.61	48.42
	120	2	1000	4	4.96	51.47	50.31
	120	2	1000	8	6.61	43.83	55.16

Fig. 2.

Variation of particle size with (a) Inlet Temperature, (b) Feed Flow Rate, (c) Aspiration Rate and (d) Weight of Polymer.

While raising the inlet air temperature, it is also important to note that the optimum temperature should never cause thermal decomposition of product and needs to be chosen specifically for the system in question.

3.2.2. Effect of feed rate

Experiments to study dependance of particle size on varying feed flow rate was conducted by increasing flow rate from 1.5 mL/min to 3 mL/min. The inlet temperature, aspirator rate and feed polymer content were kept constant at 120 °C, 1200 rpm and 2 g respectively. Feed rate also had a visible effect on spray drying since higher flow rates resulted in highly agglomerated product along with residue formation and particle size varied as shown in Fig. 2b. Wu et al., in his study stated that when the dimensionless Ohnesorge number (Oh) and Reynolds number (Re) values are low, homogeneous droplet sizes will be produced, whereas high numbers produce a non-uniform spray [41]. Higher flow rate results in larger Re leading to non-uniform spray and improper drying, resulting in denser particles that aggregate because of the high moisture content [31,42,43]. In addition to the atomized droplet size, the microparticle size is determined by the extent of removal of moisture during drying, and if there is incomplete removal of moisture, it is expected that particle size increases as the feed flow rate rises. In the current study, a minimum particle size of 7.5 μm and 3.9 μm were reported for ultrasonic and pressure nozzle respectively at 1.5 mL/min feed flow rate. Similar results were reported by Desai and Park [44] and Jumah et al. [40]. Desai and Park [44] studied feed flow rate dependence during encapsulation of Vitamin C in chitosan using tripolyphosphate crosslinking. Feed was pumped at 2, 4 and 6 mL/min and an increase in particle size was observed from 4.5 μm to 7.3 μm when flow rate was increased from 2 mL/min to 6 mL/min. Similarly, Jumah et al. [40] observed an increase in particle size from 25.1 μm to 27.84 μm when feed flow rate (w/w%) was increased from 7.3% to 36.5% during study of jameed spray drying.

3.2.3. Effect of aspirator rate

While studying the effect of aspirator rate on size of microparticles, inlet temperature (120 °C), feed flow rate (1.5 mL/min) and feed polymer content (2 g) were kept constant whereas spray dryer aspirator rate was varied from 800 rpm to 1400 rpm in increments of 200 rpm. Increased air velocity because of elevated aspiration rate improves convective mass transfer, which in turn increases surface evaporation. As a result, constant drying rate period diminishes and majority of drying takes place during the period of falling rate. Here, the amount of moisture lost to the drying air is restricted by the absence of moisture movement from within the droplet to the surface due to crust formation, and drying is constrained by internal factors. As a result, increasing the aspirator rate from 800 rpm to 1400 rpm had little impact on the size of the finished product as shown in Fig. 2c. The slight increase in size is attributed to moisture trapped inside the microsphere crust. Lowest particle size of 7.5 μm and 3.7 μm were reported at 800 rpm for ultrasonic and pressure atomization respectively. Prabhuzantye et al. [43], while studying recovery of whey proteins considered aspiration rate as parameter and reported a result like the current one in the case of pressure nozzle whereas a dip in particle size was reported while using ultrasonic nozzle. When aspiration rate increased from 1350 rpm to 2800 rpm, particle size increased from 278.3 nm to 300.7 nm. However, while using ultrasonic nozzle a dip was reported in particle size from 178.6 nm to 152.8 nm for the same range and this was attributed to the increasing cavitation effect at higher aspiration rates.

While studying the effect of aspiration rate, the hydrodynamic effects of a higher air flow rate appear to be countered by the significant quantity of air entrainment caused by the expanding spray. Greater moisture removal is achieved with longer residence times, and product recovery from the drying chamber is aided by lower air velocity. Though longer residence time and lowest velocity were achieved at 800 rpm, more exposure to hot air resulted in rupture and decrease in encapsulation efficiency. Readings at 1000 rpm reported a desired particle size (7.57 μm for US and 3.93 μm for pressure atomization) at increased encapsulation (1% and 4% increase for US and Pressure atomization respectively). Taking these factors into consideration and to provide a suitable drying process, the air flow rate is restricted to aspirator rate of 1000 rpm after the investigation of the effect of the atomizer aspirator rate.

3.2.4. Effect of weight of polymer

After fixing the inlet temperature, feed flow rate and aspirator rate at 120 °C, 1.5 mL/min and 1000 rpm respectively, weight of polymer in feed was changed from 1g to 8 g per 2 g of pure drug. In the case of the polymer concentration study, as the weight of polymer increased, feed viscosity increased leading to an increasing trend for the particle size as shown in Fig. 2d [44,45]. Due to increase in viscous energy dissipation with increased viscosity, more energy is needed to break the liquid into droplets. Therefore, greater viscosity liquid would generate a thicker layer on the surface at the same liquid flow rate, resulting in larger droplet sizes and low yield because of difficulty in overcoming viscous energy [31,46]. A rise in particle size from 7.255 μm to 11.72 μm was observed when the polymer content in solution was raised from 1 g to 8 g while operating an ultrasonic nozzle. The same study while using pressure nozzle resulted in a particle size increase to 6.605 μm from 3.928 μm over the same change in polymer quantity. The obtained results could be justified using Pe and Stokes-Einstein equation according to which dynamic viscosity of solution is inversely related to diffusion coefficient [47]. Greater solute concentrations led to higher effective viscosities, which in turn resulted in lower diffusion coefficients as per the equation. Hence crust growth is accelerated at higher concentrations because of bigger values of Pe number leading to larger particles. Banerjee et al. [48], while studying efficacy of pectin in MH encapsulation studied effect of concentration of pectin on the obtained size distribution. An increase in particle size from 34.56 μm to 71.34 μm was observed when the concentration of pectin was raised from 0.5 parts to 4 parts per part of MH. Similarly, Desai and Park [44], observed an increase in particle size from 4 μm to 4.4 μm when concentration of chitosan was elevated from 0.5% to 1% while encapsulating Vitamin C in tripolyphosphate and chitosan.

The drug to polymer ratio is an important parameter which affects the drug release profiles and hence the effect of drug to polymer ratio studied in the current work gives an indirect idea into the possible behaviour during the actual usage of drugs. For example, swelling index which affects the muco-adhesion as well as the drug release profiles of polymeric drug delivery system is strongly dependent on the drug to polymer ratio. Szekalska et al. [25] reported a decrease in swelling ratio when the drug: polymer ratio changed from 1:2 to 2:1. Though the swelling index studies could not be done in the current work, an indirect confirmation can be obtained on the basis of the results of particle size at varying polymer content.

3.3. Comparison of ultrasonic and pressure nozzle results in terms of size and distribution

Lower particle size was exhibited mostly by the conventional nozzle in comparison to the ultrasonic nozzle (Fig. 2a), which can be a result of the slightly better efficiency exhibited by pressure energy in atomizing particles in comparison to ultrasonic vibrations. However, it is clear from the studies that a narrower distribution of particle size is obtained using ultrasound as compared to the pressure nozzle as shown in Fig. 3. In each case, span of curve exhibited by particles were higher for the pressure nozzle as compared to ultrasonic nozzle, which can be attributed to the more uniform spray induced by ultrasonic vibration in comparison to more uneven one for conventional approach as per results. Span of particle size distribution rose to a maximum of 1.655 and 2.31 in the case of ultrasonic and pressure atomization respectively and at the optimum condition they were observed to be 1.437 (US) and 1.95 (Pressure). The microscopic images of particles obtained using pressure nozzle (Fig. 1c and d) showed higher agglomeration whereas in the case of ultrasonic nozzles (Fig. 1a and b) they were much more uniform sized as expected.

Fig. 3.

Particle size distribution curve of microspheres obtained using the conventional approach and ultrasound-assisted approach.

The particle size distribution observed were mostly bimodal for both ultrasonic and pressure nozzles as shown in Fig. 3. Similar results were obtained in other works including Goula and Adamopoulos [49] and LeClair et al. [42]. Penetration of relatively small particles to crevices of larger one along with collapse and shattering of thin-walled spray dried particles could result in varied distribution forms [42]. Another reason for the bimodal distribution could be the different droplet-droplet interactions. The extent of dry particle agglomeration is heavily governed by a variety of critical events that take place within the spray drier in addition to the droplet interactions. The influence of moisture evaporation and glass-transition temperature along with turbulent dispersion effects, which result in relative velocities between particles, allow these particles to collide resulting in varied particle size and bimodal distributions.

One of the main outcomes of the current work is achieving uniform size distribution of the obtained particles based on the use of ultrasonic nozzle and optimization of the important operating parameters to minimize the mean particle size. The results are important in terms of the expected benefits during the actual application of the drugs. Achieving uniform size distribution have been reported to give better invitro release profile and swelling studies when compared to pure metformin hydrochloride drug [24,25]. As a specific example, Sander et al. [2], after his ex vivo flow retention studies on porcine mucosa observed that microparticles of metformin with uniform particle size distribution gave high encapsulation efficiency and low moisture content. The 32 factorial design study done on Metformin hydrochloride by Mokale et al. [19], concluded that sustained release profile of uniform sized MH loaded ethyl cellulose particles could bring about a reduction in dosage and reduce the side effects caused by the drug. Also the compactability and tabletability of the drug was reported to improve because of spray drying with a polymer [26] resulting into uniform size distribution.

3.4. Yield

Product recovery exhibited by lab-scale spray dryer is often low due to sticking of droplets to walls of dryer and formation of fine particles that is carried away by aspirator. Yield of product ranged from around 11.13% to a maximum of 48.27% during ultrasonic atomization. Pressure atomization reported yield between 14.37% and 56.57% under similar conditions. In the current study, product yield primarily depended on inlet temperature (Fig. 4a), flow rate (Fig. 4b) and concentration of the formulation (Fig. 4d). Aspirator rate in the range considered had little to no effects on the yield (Fig. 4c). At very low temperatures below the boiling point of solvent, particles can stick to walls due to high moisture content. However, in the range of temperature considered (120°C–150 °C) the particles started sticking to the wall only beyond a level of temperature, due to melting of particles, resulting in decreased yields from 120 °C to 150 °C for both ultrasonic and for pressure spray drying as shown in Table 1, Table 2. Chegini and Ghobadian [39] reported that when inlet temperature was increased from 130 °C to 150 °C, at a constant flow rate of 15 mL/min, a decrease in yield from 35% to 25% and 85% to 35% were observed with maltodextrin and liquid glucose as additives respectively. The increased flow rate resulted in improper drying since the feed flowrate was not in proportion to the drying time. This eventually led to the collection of excess undried liquid feed in residue flask of spray dryer and hence low yield. Formation of product with higher moisture content due to viscous feed at higher polymer concentration also led to agglomeration, making it difficult to collect and store the product. Desai and Park [44] reported a decrease in yield (from 59.2% to 54.5%) of encapsulated Vitamin C with an increase in flow rate (2 mL/min to 6 mL/min). Similarly, Chegini and Ghobadian [39] reported a reduction in the yield from 25% to 18% when flow rate was raised from 15 mL/min to 30 mL/min at a constant inlet temperature of 150 °C.

Fig. 4.

Variation of product yield with (a) Inlet Temperature, (b) Feed Flow Rate, (c) Aspiration Rate and (d) Weight of Polymer.

The small decreasing effect that aspirator rate has on yield (a fall of 3.34% and 2.47% respectively for ultrasonic and pressure nozzle), might be due to lesser time spent in the chamber along with the excess drag force experienced around the exhaust region making it difficult for the particles to sediment down in the cyclone separator. Prabhuzantye et al. [43] also observed that as aspirator rate increased from 1350 rpm to 2800 rpm, the yield of dried whey protein reduced from 1.61 g to 0.6 g for pressure drying. However, ultrasonic spray drying carried out in the same study showed an increase in yield from 0.5 g to 0.948 g for same range of aspiration rate. The slight increase was attributed to the small particle size which combined with low aspiration rate resulted in a very small portion of dried feed reaching the cyclone separator and thereby affecting the expected yield. An increase in polymer concentration leads to increased viscosity and hence improper drying since increased viscosity means a higher energy is required to break up the fluid into spray.

In the case of US atomization, the particle yield was reported at a maximum of 48.27% at inlet temperature of 120 °C, flow rate of 1.5 mL/min, aspirator rate of 800 rpm and polymer content of 2 g. However, the particle size was slightly higher (7.50 μm) compared to that observed at 1000 rpm (7.26 μm) with a yield of 48.2% and hence it was chosen as optimum condition. Similarly for pressure atomization though highest yield (56.57%) was obtained at an aspirator rate of 800 rpm, yield of 54.61% at 1000 rpm was chosen as optimum because of the much better encapsulation efficiency exhibited at that condition (difference of around 8%).

3.5. Encapsulation efficiency

Larger particles, formed at higher temperature, leads to better core retention (lesser surface to volume ratio), and hence result in increased encapsulation for a temperature change from 120 °C to 130 °C [49]. However, with escalated exposure to higher temperature and flow rate, number of irregularly shaped particles with cracks and shrinkage also increases which leads to reduction in encapsulation [37] due to leakage of drug through the cracks in some cases (Fig. 5a and b). The competing effect of slightly higher particle size and ballooning effect of microspheres results in an irregular trend of EE with the aspiration rate [50] as shown in Fig. 5c.

Fig. 5.

Variation of encapsulation efficiency (EE) with (a) Inlet Temperature, (b) Feed Flow Rate, (c) Aspiration Rate and (d) Weight of Polymer.

Higher polymer concentration however led to better encapsulation for the same drug loading as shown in Fig. 5d. Ultrasonic and pressure atomization showed a maximum encapsulation efficiency of 64.26% and 55.16% respectively, when feed constituted 8 g of polymer and 2g of pure drug. As viscosity increased, denser and wrapped particles that protect the drug were synthesized rather than the weaker shell obtained at lower concentrations [7,40]. According to Reineccius and Coulter [50], there is an ideal feed solid content required for optimum encapsulation, beyond which inclusion of more polymer, transcends its solubility and these undissolved components are unable to provide any useful encapsulating effect. Higher polymer content also rises challenge of elevated viscosity and slower generation of droplets during atomization [44,45].

In ultrasonic atomization, experimental condition with inlet temperature of 120 °C, feed flow rate of 1.5 mL/min, aspirator rate of 1000 rpm at 2 g polymer content was chosen as optimum despite the slightly lesser EE (56.14%) compared to that observed at 8 g (64.26%). The lowest particle size (7.26 μm) and highest yield (48.2%) available with 2 g of polymer in feed compensated for the dip in encapsulation efficiency. Similarly for pressure atomization, the slight difference between size at 1 g (3.93 μm) and 2 g (4.06 μm) polymer content was made up by higher encapsulation for 2 g polymer (48.42%) when compared to 1 g of polymer (44.75%) with nearly equal yield values. With these factors in mind and analysing the results, choosing equal parts of drug and polymer resulted in small particle size with satisfactory yield accompanied by an increase in the EE. For all the physical and chemical characterization studies obtained samples from optimum conditions using ultrasonic and pressure atomization were considered. It is important to note that the best parameters selected in the present study is a typical optimization exercise balancing between multiple output requirements considered as particle size, yield and encapsulation efficiency. The observed EE values also need to be tested for actual anti-diabetic effects so as to confirm the drug usability. Credence to good effects at the obtained levels of EE in the current study can be obtained from the literature illustration. Kotha et al. [51] studied the application of synthesized Metformin hydrochloride (HCl) microspheres and nanoparticles for reducing the sugar content using the classical hypoglycemic activity tests. It was reported that obtained microspheres with EE values over the range similar to that obtained in the current work showed good activity for reducing sugar as well as better dissolution characteristics.

3.6. XRD analysis

The results for the XRD characterization of pure MH and sodium alginate along with microspheres synthesized using ultrasonic and pressure nozzles are shown in Fig. 6. X-Ray diffraction plots clearly depict the crystalline and amorphous nature of pure MH and sodium alginate respectively. The characteristic drug peaks observed around the 2θ values of 17°, 22°, 24°, 28° and 31° were also seen in the microsphere formulations, albeit with a lower intensity.

Fig. 6.

XRD plot for Pure MH (MH), Sodium Alginate (Na- ALG), (c) sample obtained using Ultrasonic spray drying (US SD) and (d) sample obtained using Pressure spray drying (Conv SD).

Similar peaks were recorded by Jagdale et al. [52] and Mokale et al. [19]. Jagdale et al. [52] also concluded that dissolution rate and bioavailability of drug is influenced heavily by polymorphic structure of the pharmaceutical compound. Al-Zoubi et al. [26], also obtained PXRD peaks in the range of 20°–40° and observed that some of the low-intensity peaks of pure MH are retained during normal spray drying and they disappear if co-spray drying was employed.

The observed trend in intensity reduction could be because of lower concentration of drug in formulation compared to that in pure drug or because of presence of polymer which is highly amorphous in nature. This implies that the drug does not change into an amorphous state in the formulations but rather stays in its crystalline state but has lower overall crystallinity in the microsphere formulation as compared to pure drug [27].

3.7. FTIR analysis

FTIR analysis conducted for Pure MH, Na-ALG and microspheres synthesized using ultrasonic and pressure spray drying showed the typical peaks of MH, as well as stretching regions [48,[52], [53], [54]] as shown in Fig. 7. Two distinctive MH bands were seen at 3365 cm⁻¹ and 3291 cm⁻¹ representing N–H primary stretching vibration. The bands at 1625 cm⁻¹, 1543 cm⁻¹ and 1163 cm⁻¹ were pertained to C C, C NH and C N respectively while typical bands at 1622 cm⁻¹ and 1568 cm⁻¹ represented C–N stretching and the band at 3173 cm⁻¹ was due to the N–H secondary stretching. C–N stretching vibrations were responsible for the band at 1058 cm⁻¹, while the bands at 935 cm⁻¹ and 632 cm⁻¹ were results of NH₂ rocking and C–H bending vibrations, respectively. The presence of these peaks in the microspheres supports the required MH purity. No interaction between the functional groups even during the ultrasound-based processing was observed since there was no visible shift in the peaks. There was a slight decrease in the magnitude of transmittance because of the stretching of carboxylate salt ion and mannuronic acid in Na-ALG and C– N and N–H stretching and deformation vibrations corresponding to pure MH [2].

Fig. 7.

FTIR plot for Pure MH (MH), Sodium Alginate (Na- ALG), (c) sample obtained using Ultrasonic spray drying (US SD) and (d) sample obtained using Pressure spray drying (Conv SD).

3.8. DSC analysis

The appearance or disappearance of thermal peaks along with shift in thermograms produced by DSC helps in detection of potential interactions between a drug and a polymer. Fig. 8 shows the plots of pure MH, Na-ALG, and microspheres synthesized using ultrasonic and pressure spray drying. Pure MH thermogram detected a sharp endothermic peak at 232.1 °C, which corresponds to its melting point. The peak for DSC thermograms in a study conducted by Al-Zoubi et al. [26], indicated 229.2 °C as melting point of pure MH which is very close to the results obtained in current study. The peak of the MH did not alter considerably even in the microspheres, as per the thermogram of microspheres [25,52,55] with the melting points reported as 228.8 °C for ultrasonically-dried microspheres and 225.2 °C for pressure dried microspheres. The slight decrease, howsoever, in melting point from 232.1 °C to 228.8 °C (US) and 225.2 °C (Pressure dried) could be due to Na-ALG introduction in the microsphere formulation. Observation made in the work of Al-Zoubi et al. [26], that co-spray dried MH-Na ALG particles shifted peaks from 227.4 °C to 225.1 °C for an increase in polymer concentration from 2.5% to 5% supported results of the current study.

Fig. 8.

DSC plot for Pure MH (MH), Sodium Alginate (Na- ALG), (c) sample obtained using Ultrasonic spray drying (US SD) and (d) sample obtained using Pressure spray drying (Conv SD).

Kulkarni et al. [27] also observed that for an increasing polymer concentration the intensity of peaks decreased. Analysis of thermograms of MH and Eudragit obtained by preparing solutions of different drug:polymer ratio concluded that as concentration of polymer increased from 1:1 to 1:7, the peak fell from 236.66 °C to 228.44 °C. This result also backs up the notion that amorphous areas were formed and that the degree of MH crystallinity was slightly lowered when it is incorporated into the microspheres. Na-ALG showed a strong exothermic peak at 242.50 °C, which corresponded to the polymer degradation and a minor endothermic peak at 109.30 °C, which was associated with the dehydration process. The absence of the melting point peak of Na-ALG (129.27 °C) in microparticles possibly indicates that Na-ALG dehydrated during the spray drying process. In both formulations, no formation or any small shift in the peak was seen as shown in Fig. 8, which indicates that there was no interaction between the excipient polymer and the drug. Thus, following encapsulation the drug stayed crystallized inside the polymer covering, dispersed as a solid or molecular dispersion.

4. Conclusions

The current research was aimed at optimizing size distribution of MH loaded Na-ALG particles produced through ultrasonic and pressure assisted spray drying. Under best conditions, spherical microspheres with a wrinkled outer crust and slight reduction in crystallinity due to encapsulation were obtained. Further optimum conditions as inlet air temperature (120 °C) and flow rate (1.5 mL/min) yielded smallest particle size without compromising encapsulation efficiency and yield. Aspiration rate did not have a distinguishable effect on particle size in the range considered for study, however an aspiration rate of 1000 rpm was deduced to be optimum after comparing the yield and encapsulation efficiency values. For all parameters other than polymer content as particle size increased, EE decreased as a result of cracks formed on microsphere surface that cause leakage of MH. However, an increase in polymer content in feed solution was accompanied by lager particle size and better encapsulation because of the viscous feed. In ultrasonic atomization, 2 g was chosen as optimum weight of polymer despite the slightly lesser EE (56.14%) compared to 64.26% (at 8 g). The lowest particle size (7.26 μm) and highest yield (48.2%) available compensated for the dip in encapsulation at 2 g of polymer. For pressure atomization, same ratio was chosen as optimum since 2 g of polymer in feed offered small particle size (4.06 μm) along with high enough encapsulation (48.42%) and nearly equal yield values.

Comparison between performance of ultrasonic and pressure atomization concluded that while both methods produced particles sized less than 10 μm, the former offered narrow distribution and better encapsulation because of uniform spray produced by ultrasonic vibration. Though the yield in pressure atomization was higher due to more robust pressure energy, the wider distribution gave indications of polydisperse spray. It was clearly elucidated that using ultrasound is a viable alternative to conventional spray drying for synthesizing similar sized spherical particles of encapsulated MH along with achieving better control over particle size distribution and higher encapsulation efficiency, necessary in pharmaceutical applications.

Data availability statement

Data will be made available by the authors on reasonable request.

CRediT authorship contribution statement

Anjana M: Writing – original draft, Methodology, Investigation. Parag R. Gogate: Writing – review & editing, Supervision, Conceptualization.

Declaration of competing interest

The authors declare the following financial interests/personal relationships which may be considered as potential competing interests.