Spray drying formulation of amorphous solid dispersions

Abhishek Singh1, Guy Van den Mooter1*

1 Drug Delivery and Disposition, KU Leuven, Leuven, Belgium

*Corresponding author:

Address- Drug Delivery and Disposition, Department of Pharmaceutical and Pharmacological

Sciences, University of Leuven; Campus Gasthuisberg O+N2; Herestraat 49 b921, 3000 Leuven;

BELGIUM

Tel.: +32 16 330 304 fax: +32 16 330 305 Mobile: +32 473 356 132

e-mail: guy.vandenmooter@pharm.kuleuven.be

1

Graphical Abstract

Abstract

Spray drying is a well-established manufacturing technique which can be used to formulate amorphous solid dispersions (ASD) which is an effective strategy to deliver poorly water soluble drugs (PWSD). However, the inherently complex nature of the spray drying process coupled with specific characteristics of ASD makes it an interesting area to explore. Numerous diverse factors interact in an inter-dependent manner to determine the final product properties. This review discusses the basic background of ASD, various formulation and process variables influencing the critical quality attributes (CQA) of the ASD and aspects of downstream processing. Also various aspects of spray drying such as instrumentation, thermodynamics, drying kinetics, particle formation process and scale-up challenges are included. Recent advances in the spray-based drying techniques are mentioned along with some future avenues where major research thrust is needed.

Keywords

Amorphous solid dispersions, Spray drying, Poorly water soluble drugs, Mollier diagram, Drying kinetics, Scale-up, Downstream processing, Process parameters, Carrier, Electrospraying

2

Abbreviations

AGU- Anhydro-D-glucopyranose units

API- Active pharmaceutical ingredient

ASD- Amorphous solid dispersions

ATR-FTIR- Attenuated total reflectance Fourier Transform Infrared Spectroscopy

BCS- Biopharmaceutics classification system

CAP- Cellulose Acetate Phthalate

CAAdP- Cellulose acetate adipate propionate

CMC- Carboxymethyl cellulose

CMCAB- Carboxymethylcellulose acetate butyrate

Compritol 888 ATO- Glyceryl dibehenate

CQA- Critical quality attributes

DCM- Dichloromethane

DMA- N,N-dimethylacrylamide

DOE- Design of experiments

EC- Ethylcellulose

EHEC- Ethylhydroxyethyl cellulose

Gelucire- Lauroyl polyoxyl-32 glycerides

GI- Gastrointestinal

Inutec SP1- Inulin Lauryl Carbamate

HEC- Hydroxyethyl cellulose

HME- Hot melt extrusion

HPC- Hydroxypropyl cellulose

HPCDS- Hypulcon pulse combustion dryer system

HPMC- Hydroxypropyl methylcellulose

HPMC-AS- Hydroxypropyl methylcellulose acetate succinate

HPMC-P- Hydroxypropyl methylcellulose phthalate

3

Kollicoat IR- Polyvinyl alcohol-polyethylene glycol graft copolymer

MC- Methylcellulose

MCC- Microcrystalline cellulose

Myrj 52- Polyoxyl 40 Stearate

NaCMC- Sodiumcarboxymethyl cellulose

NIR- Near infra-red

PAT- Process analytical tools

PCSD- Pulse combustion spray dryer

PDMA- Poly-dimethylacrylamide

P(DMA-grad-MAG)- Poly (N,N,-dimethylacrylamide-grad-methacrylamido glucopyranose)

PEG- Polyethylene glycol

PEP- Poly(ethylene-alt-propylene)

PEP-PDMA- Diblock copolymer of DMA and PEP

PHPMA- Poly[N-(2-hydroxypropyl)methacrylate]

Poloxamer- Poly(ethylene glycol)-block-poly(propylene glycol)-block-poly(ethylene glycol)

PVP- Poly(vinylpyrrolidone)

PVP VA64- Poly(1-vinylpyrrolidone-co-vinyl acetate)

PWSD- Poorly water soluble drugs

QbD- Quality by design

QTPP- Quality target product profile

SCM- Supercritical methods

SLS- Sodium lauryl sulphate

Soluplus- Polyvinyl caprolactam-polyvinyl acetate-polyethylene glycol graft copolymer

Sucroester 15- Sucrose monopalmitate

THF- Tetrahydrofuran

Tg- Glass transition temperature

Vitamin E TPGS- D-α-tocopheryl polyethylene glycol 1000 succinate

4

Table of Contents

1. Introduction

2. Amorphous solid dispersions (ASD)

3. Spray drying process

4. Thermodynamics of spray drying

4.1 Drying kinetics

4.2 Mollier diagram

5. Particle formation process and the effect of the process parameters

5.1 Feed variables

5.1.1 Solvent system

5.1.2 Influence of solute components in feed solution

5.1.3 Feed solution stability

5.2 Process parameters

5.2.1 Feed flow rate

5.2.2 Inlet and outlet temperature

5.2.3 Atomization and drying gas type and flow rate

5.2.4 Types of atomization device

6. Carriers for spray dried solid dispersion formulations

6.1 Poly (ethylene oxide) polymers and derivatives

6.2 Cellulosic Derivatives

6.3 Vinyl Polymers

7. Multi-component solid dispersions

8. Influence of preparation method

9. Recent advances

9.1 Electrospraying

9.2 Pulse combustion spray dryer (PCSD)

10. Scale-up challenges

5

11. Quality by design (QbD) and process analytical tools (PAT) in spray drying

12. Downstream processing and product development

13. Future outlook

14. References

6

1. Introduction

As we enter into the latter half of the current decade, the paradigm of solubility challenges faced by formulation scientists remains largely unchanged. Many of the drug molecules can be categorized under Biopharmaceutics classification system (BCS) class 2 or 4 (Figure 1) [1]. The problem is a difficult one to overcome because of multifaceted factors driving it. The use of non-aqueous (or solvent mixture) based media for screening and purification purposes in high throughput screening tend to give hits with higher molecular weight and lipophilicity [2, 3]. The quest for identification and targeting of kinase pathways, ion channels, nuclear receptors and protein-protein interactions with potent and selective agents is also motivating the choice towards lipophilic compounds [4, 5]. The presence of many low solubility compounds in the drug discovery pipeline is not good for any stakeholder of the drug development process due to high fall-out rate and associated development costs. Apart from chemistry based strategies to improve solubility, onus is on formulation scientists to provide enabling drug delivery strategies for such candidates.

Figure 1: Biopharmaceutics classification system and various approaches to overcome solubility and permeability challenges. (Adapted from [6, 7])

For a compound to reach to its target site, it should first be dissolved in the gastrointestinal (GI) fluid (in most of the cases) [7]. The rate at which this happens is given by the Nernst Brunner equation [8].

7

dCdt

=SD(C s−C t)

Vh… ..Eq .1

Here, dC/dt- dissolution rate of the drug, S- surface area of the dissolving surface, D- Diffusion coefficient of the drug, Cs- Saturation solubility, Ct- concentration at time t, V- volume of dissolution medium and h is the thickness of the diffusion layer surrounding the dissolving particle. Diffusion coefficient of the drug and dissolution medium volume are the factors which cannot be significantly modified in vivo. Thus, the enabling strategies focus on altering solubility and/or surface area.

Amorphization is an approach wherein the solid state form of the drug is changed from crystalline to amorphous. The rationale behind this approach can be understood by the following equation [9].

∆GT° Amorphous ,Crystalline=−RTln ( σT

Amorphous

σTCrystalline )… .. Eq .2

Here, ∆GT° Amorphous ,Crystalline is the energy difference between the crystalline and the amorphous

state, R is the gas constant, T is the absolute temperature of concern and σT

Amorphous

σTCrystalline is the

solubility ratio of the two forms. It follows from equation 2 that the amorphous form has a higher theoretical solubility as compared to the crystalline form due to its excess thermodynamic properties (Figure 2). In simple terms, in the amorphous state there is no energy requirement to break the crystal lattice structure so that the drug molecules can interact with solvent molecules through intermolecular interactions and become solubilized. But the excess thermodynamics properties of amorphous forms also result in their tendency to crystallize thereby negating the solubility advantage. ASD can be considered as a potential solution to this issue.

8

Figure 2: Thermodynamic descriptor-temperature diagram for the various states of a drug.

As the crystalline drug is heated, the thermal energy breaks the crystal lattice structure and at melting point (Tm) the drug gets converted into liquid state. To generate amorphous state, the liquid should be cooled at a sufficiently fast rate. This results in conversion of liquid to supercooled liquid state and subsequently the system falls out of the equilibrium at the glass transition temperature (Tg). For certain drugs such as itraconazole, formation of mesophase is observed. Tk is the Kauzmann temperature which is a hypothetical temperature at which the entropy of the supercooled liquid becomes equal to that of crystal. Spray drying process is also similar to quenching, as the time scale in which droplet to particle conversion takes place is really small and in ideal cases does not allow crystallization. (Adapted from [10])

2. Amorphous solid dispersions (ASD)

ASD consist of drug molecules dispersed in amorphous polymeric carriers. The drug stabilization is a consequence of factors such as intermolecular interactions, antiplasticization effect exerted by the polymer, physical barriers to the crystallization process (local viscosity) and the reduction in chemical potential of the drug [11]. The role of the polymeric carrier is not limited to the stabilization but also mechanisms responsible for improved dissolution rate and absorption. Hydrophilic carriers such as Poly(vinylpyrrolidone) (PVP), poly(1-vinylpyrrolidone-co-vinyl acetate) (PVP VA64) and hydroxypropyl methylcellulose (HPMC) are highly water-soluble and enhance water uptake into the solid dispersion matrix. Carriers also play a crucial

9

role in maintaining supersaturation and precipitation inhibition in vivo which is widely accepted as critical in improving solubility in the GI tract [12]. Other mechanisms responsible for improved solubility are reduced particle size resulting in increased surface area [13, 14]. In ideal case, i.e., molecular dispersion, the surface area available for dissolution is the maximum since the drug size is reduced to (almost) a single molecule. On several occasions this is not the case and the active pharmaceutical ingredient (API) distribution within the carrier matrix becomes inhomogeneous leading to drug-rich and polymer-rich regions. Since drug polymer miscibility is crucial for solid dispersions stabilization, phase separation can promote API crystallization [15-17]. Therefore, every effort should be made to produce miscible solid dispersions systems and protect them from drivers of phase separation such as high temperature, humidity and mechanical stress [13, 18, 19].

Amorphous to crystalline transition is a thermodynamically driven phenomenon due to lower free energy of the crystalline state and is bound to happen at a certain point of time (the time-scales involved can be really long in absence of external stimuli). But for crystalline to amorphous transition, external energy needs to be imparted to the system. Mechanical activation such as milling can generate amorphous forms [20]. Another way is to either dissolve in a solvent or melt the crystal form to break the crystal lattice structure. The cooling of the API from the molten state (Hot melt extrusion (HME)) or rapid solvent evaporation (Spray drying) from the dissolved state leads to amorphous forms [13]. In this review focus will be on the spray drying. The readers can refer to excellent overviews by Thiry et al. [21], Shah et al. [22], Crowley et al. [23] and Li et al. [24] for further information about HME.

3. Spray drying process

Spray drying is an energy intensive, continuous and scalable drying process [25, 26]. The process can generate nano to micron size particles that have a narrow distribution in a very short time-frame. For the context of this review, spray dryer equipment can be viewed as solid state transforming reactors where the crystalline starting material is converted into amorphous product. The first patent of spray drying process is more than 140 year old wherein it was described as a process for simultaneously atomizing and desiccating fluid and solid substances. The process was meant for exhausting moisture and to prevent destructive chemical change [27]. Historically, the spray drying process has been most extensively used in food and chemical industry. However, its use quickly expanded to other industries such as cosmetics, fabrics and electronics. The first foray of this technique in the pharmaceutical field was for the manufacture of pure API. From there on it has been used ever increasingly for various specialized applications such as microcapsules, controlled release particles, composite microparticles, nanoparticles, and liposomes [28].

The spray drying process sequentially involves several steps involving various components (1-6) as shown in figure 3. Firstly, the feed solution/suspension is pumped into the drying chamber through a nozzle (components 1, 2 and 3). During exit from the nozzle tip the droplets are atomized and come in contact with the drying fluid i.e. hot gas (often air) inside the

10

drying chamber (component 4). The residence time inside the drying chamber depends on the process parameters and the equipment dimensions and may typically last for a few milliseconds. During the transit through the drying chamber energy-mass transfer takes place at the dynamic droplet surface. Finally, the dried material is separated from the drying medium using a cyclone (component 5) and is collected in a collection device (component 6). The exhaust gases are filtered via HEPA filters (component 7). To accomplish the abovementioned steps, various hardware configurations can be used as described below.

Figure 3: Spray drying set-up

Typical spray-drying system consists of various components. Component choice and their operating parameters have crucial influence on the process output. Few of these aspects are listed below the components (1-7).

The choice of the feed pump used depends on the viscosity of the feed material and the type of atomizer system used [26]. Low pressure pumps are desirable for rotary atomizers or bi-fluid nozzles. Pressure nozzles necessitate the use of high pressure pumps. Various atomizer designs are available and use different kinds of forces input to obtain fine droplets [29]. Atomizers can be either rotary, hydraulic (pressure), pneumatic or ultrasonic nozzles. In rotary

11

atomizers, centrifugal force results in breaking down of the liquid stream into small droplets. Modification in the rotary atomizers such as straight or curved grooves provide opportunities for particle engineering [26, 30]. When using this atomization set-up care should be taken to use drying chamber of sufficient diameter. Material adhesion to the drying chamber walls can be a limiting factor for its use for expensive drugs. Bi-fluid or multi-fluid nozzles utilize the pressure energy and can be used with narrow chambers. In pneumatic nozzles kinetic energy of the compressed carrier gas is transferred to the liquid surface at a central collision point causing droplet formation. Ultrasonic nozzles use vibrational energy for atomization but still find limited use in industrial settings due to their low throughput (<50 ml/min) [31]. The vibrational energy is produced by the application of a high frequency electric signal to the two electrodes placed between the piezoelectric transducers [29]. The vibrational motion is transferred and amplified by a titanium nozzle tip. Variations in the construction and operating conditions of these established designs enables to control the particle size distribution and density.

The atomized droplets encounter the drying air in the drying chamber which commonly have a height to diameter ratio of 5:1 (tall) or 2:1 (small) chambers. The air-droplet contact system can be of cocurrent, counter-current or mixed flow type. Cocurrent contact system is most widely employed for pharmaceutical purposes [32]. The droplet size distribution generated by the atomization set-up determines the residence time needed in the chamber and its dimensions as well [33]. The nature of the gas flow (turbulent or laminar) will also have a bearing on the residence time of the droplet, and final product moisture content. Of particular importance for spray drying of amorphous systems is the need for strict inter-batch control of the temperature and humidity of the drying air.

Following drying the particles are collected using specific design features and separation devices. This is necessitated by the relatively small particle size used in pharmaceuticals. The particle collection can take place at the bottom of the drying chamber from where it needs to be scrapped. Scrapping devices such as vibratory devices, mechanical brushes and/or compressed air might be used [29]. It has been shown for lab-scale spray dryers that drug-polymer miscibility varies with location from which the product is collected [34]. Although there are no reports of such variability in large scale spray dryers, such a scenario is not impossible. Therefore, caution should be exercised while using scrapping devices for ASD as it might induce mixing of the product with different degrees of phase-behavior. Moreover, mechanical brushes employing ‘stress’ might possibly result in the change in the phase-behavior of binary solid dispersions [18, 19, 35, 36]. Such issues can be avoided by an additional design feature in the drying chamber i.e. a cone shaped bottom part which assists in flow of the product. Bag filters and cyclones are widely used as a separation device. Typical cyclones used in pharmaceutical settings are reverse-flow, gas-solid separators wherein the centrifugal force drives the separation of two phases with different mass. The centrifugal flow is generated because of the tangential entry of the gas-solid mix into the cyclone body. As the gas moves down in the swirling flow the particles experience centrifugal force and get deposited on the cyclone walls as a result of particle inertia. Particles further settle down due to gravitational force and motion

12

of the boundary layer. Finally, the gas phase hits the cyclone bottom and the direction of the glass flow reverses. The gas now moves via a reverse vortex through the central axis of the cyclone to the gas outlet at the top of the cyclone [37].

All of the above mentioned steps have an crucial impact on the drying efficiency of the process and therefore can majorly impact the solid state properties of ASD. The interplay of the machine configuration, process and formulation variables can overly complicate the process design and control whilst controlling the CQA of the spray dried amorphous system. To understand the contribution of process and formulation parameters on ASD, it is first important to understand the thermodynamics behind spray drying and the mechanisms of particle formation.

4. Thermodynamics of spray drying

In the spray drying process, both the heat and mass transfer along with their temporal and spatial aspects determine the final product features. The crucial impact of atomization is the generation of a large surface area over which heat and mass transfer takes place. The convective heat transfer from the carrier gas to the droplet surface and further inside the droplet through conduction drives this exchange.

4.1 Drying kinetics

Initially, the mass-transfer from the atomized feed solution/suspension is similar to that of a pure solvent droplet [38, 39]. This is the ‘constant drying rate period’ (Figure 4) under which the vapor transfer from the unbound solvent at the surface to the carrier gas takes place. The migration of the solvent to the droplet/solid particle surface can be mediated through molecular diffusion from high concentration in the central droplet region to the lower concentration at the surface, convection of moisture within the droplet, via evaporation within a solid and subsequent gas transport out of the solid by diffusion and/or convection and capillary flow [40]. Upon reaching a critical moisture content the ‘falling rate period’ commences. During the switch from constant to falling drying rate period over time, the droplet temperature shifts from close to the thermodynamic wet-bulb temperature to the dry-bulb temperature. Hence, for some initial part of the drying period the droplet and its content are safe from the high drying temperatures. In the constant drying rate regime, the solvent evaporation rate is driven by the heat-transfer to the droplet. This is accompanied by an increase in the humidity of the carrier gas which slows down particle formation rate. As more and more solvent evaporates, the evaporation rate into the gas medium is dominated by the solid content in the droplet. The reason for the shift from constant to falling drying period is because throughout the droplet journey in the drying chamber, its viscosity increases. At a certain point the solidification at the droplet surface takes place hindering the solvent escape from the droplet. From the stand-point of the amorphization of the dissolved drug, the drying rate can be a critical factor in determining whether the substance is completely crystallized or not, especially in the case of poor glass formers. Also, the time-period for which the drying

13

mass is exposed to dry-bulb temperature can be critical for the stability of the amorphous form which is susceptible to high temperature. Therefore, adequate care should be taken to control the drying kinetics for generating amorphous product.

Figure 4: Typical drying process of a droplet upon exposure to drying medium

Upon exposure, for a small duration the droplet experiences sensible heat till the wet-bulb temperature is reached. Further, the droplet is assumed to dry at a constant evaporation rate (𝜅) and the droplet diameter (and thus surface) decreases linearly from d0 initially to d(t) at time t. During drying, the viscosity of the droplet increases, solvent content decreases and further droplet solidification takes place. At a certain critical moisture content, the evaporation rate falls and the droplet temperature increases to the dry-bulb temperature. The characteristic droplet drying time (τD) is given by the ratio of square of the d0 and the evaporation rate 𝜅. The drying curve is affected by the solute properties and process parameters.

During the droplet drying process, there is a movement of solute from high concentration exterior to the interior of the droplet and outward movement of the solvent. Particles of various morphologies can be formed and the possibility of their formation can be estimated using a dimensionless quantity, Peclet number (Pei) of solid component i, which is the ratio of 𝜅 and diffusion coefficient (Di). When Pei<1, solid particles are likely to be formed, whereas, when Pei>1, hollow spheres or dimpled/wrinkled particles can be formed. The surface enrichment (Ei) of a solid is given by the ratio of surface concentration (cs,i) to average concentration in the droplet (cm,i) and is related to peclet number and a function, βi. (Adapted from [28, 41])

14

4.2 Mollier diagram

Two aspects which should be considered when considering the removal of the solvent from droplets is the drying temperature, the humidity and velocity of the surrounding air over the droplet surface. The thermodynamics driving force for the solvent evaporation is by virtue of the difference in the chemical potential/water activity (in case of water)/solvent potential (in case of organic solvent) between the droplet being dried and the carrier gas phase [40].

Taking the instance of drying of a droplet of pure water, the drying rate is given by:

Rate∝ ( ppuresat −pw , air)… .. Eq .3

Here, ppuresat is the vapor pressure of pure water and pw , air is the partial pressure of the water in

the gas phase. The water activity in the gas phase (awvapor¿ can be related to the ratio of partial

pressure (pw) and relative humidity (RH) by the equation 4.

awvapor=

pw

p puresat =% RH

100… .. Eq .4

The water activity in the solid (awsolid ¿ can be defined as:

awsolid=

pmixturesat

ppuresat ….. Eq .5

Here, pmixturesat is the vapor pressure of water in the mixture. The evaporation of water would take

place until thermodynamic equilibrium is established, when:

awsolid=aw

vapor=aw… ..Eq .6

Here, awsolid is the water activity in the solid phase and aw

vapor is the water activity in the vapor phase.

Although the drying capacity of the carrier gas would increase with increasing temperature, there is a limit to it due to product degradation and operator safety concerns. Moreover, aspects of the dynamicity of the process implies frequent change in temperature and humidity of the carrier gas within the drying chamber. This necessitates effective management of the humidity which is critical to effective drying process. Mollier’s diagram is a tool which establishes the relationship between the heat content and the water vapor of the air. As shown in the figure 5, Humidity (abscissa), temperature (ordinate) are plotted along with enthalpy (as lines sloping diagonally from top left to bottom right). For a given set of conditions, dry-bulb temperature, wet-bulb temperature, dew point, amount of solvent added or removed and quantity of heat exchanged in the drying process can be obtained by Mollier’s diagram. Practically this translates into estimating the drying capacity of the dryer, droplet/solid particle temperatures and air conditions [39]. Such information is particularly relevant for the spray

15

drying of ASD which is susceptible to both high temperature and moisture/solvent presence. ASD presents an inherently conflicting process scenario whereby on one end low residual solvent content is desirable and on the other hand exposure of the product to high temperature should be avoided. Care should be taken that the drying medium has enough drying capacity to form product with low residual solvent but does not result in a high outlet temperature. Mollier diagrams can help in creating this fine balance. Figure 6 represents a simplified example of Mollier diagram use.

Figure 5: Mollier diagram

Note that red lines indicate isotherms (the lines are not exactly horizontal), green broken lines indicate isenthalpic lines and slope diagonally. Absolute humidity is indicated by the blue vertical lines whereas relative humidity is

16

shown by the black curved lines. Absolute humidity is the weight of water vapor per kg of air (g/kg). Relative humidity is the ratio of the amount of actual water vapor to the max. amount of water vapor (%). The pressure of the humid air will affect its characteristics, hence, any mollier diagram is at single specific barometric pressure. (Modified from [40])

Figure 6: Mollier diagram example

Suppose a liquid is atomized into the drying chamber with drying gas at temperature T1 and humidity H1 (extremely low). T1 can be called the dry-bulb temperature. Since the water is added without any heat supply, the process would move along isenthalpic lines (thick broken green line). As a result of the liquid atomization, temperature of the gas decreases to T2 and humidity increases to H2. The drying gas cannot take up water indefinitely and at a certain limit the water exerts a vapor pressure equal to the partial pressure of the water vapor in the given

17

mixture. At this point, the temperature T3 is called the thermodynamic wet-bulb temperature or adiabatic saturation temperature. Note that the wet-bulb temperature is different from adiabatic saturation temperature.

5. Particle formation process and the effect of the process parameters

The formation of smooth spherical particles is just one of the numerous kinds possibly formed using spray drying. The millisecond time scale between the droplet to particle transition is fraught with competitive dynamic events, all of which impact the final morphology of the particle. The particle attributes encompassed by the term ‘particle morphology’ include size, shape, structure and surface properties. The theoretical framework behind the empirical evidence is based on equations describing the link between process parameters, material characteristics and product attributes.

A generalized view-point on the particle formation process involves three stages (Figure 7). Before stage 1 is initiated, the droplet undergoes sensible rapid heating with no mass change [42]. In stage 1, solvent evaporation starts and results in a receding droplet surface [43, 44]. Consequently there is solute mass concentration at the surface. The concentration gradient between the droplet surface and core also drives solute movement inwards from the surface [45]. Nonetheless, the solute diffusion is often not able to follow the reduction in droplet diameter and shell formation takes place in stage 2. The solvent evaporation continues from the surface and further from the interior of the droplet [42]. This added droplet surface enrichment thickens the crust thus resisting mass transfer. Any heat transfer to the droplet at this stage increases the particle temperature. Finally the drying mass can be treated as a dry non-evaporating solid sphere. The reduction in moisture level only takes up to a certain minimum level identified as equilibrium moisture content or bound solvent which cannot be removed by drying [43]. Since the shell resists the solvent evaporation after a certain stage, there is internal pressure build-up inside the particle. Depending on the strength, thickness of the shell and internal pressure build up it can explode, inflate or crack [46]. In the final stage 3, the droplet experiences sensible drying where there is no mass change.

18

Figure 7: Droplet to particle transition and effect of evaporation rate

As a particle is exposed to the drying gas, droplet core temperature increases in three stages with corresponding droplet to dry particle transition. The droplet solvent content also decreases to a certain minimum level. The droplet evaporation rate can determine the final morphology of the particles. Slow evaporation rates result in denser particles. The process gives enough time for the solute particle to migrate and come in close proximity. Strong drug-carrier interactions would determine close proximity of drug molecules to form a crystal lattice, resulting in single phase solid dispersions. Weak drug-carrier interactions can result in phase-separation. Fast evaporation rate results in skin formation at the droplet surface and generates particle of less density. Based on the permeability of the skin and within the particle, porous or hollow particles can be obtained. (Modified from [47, 48])

Various spray drying process variables significantly impact the particle formation process [47]. The process variables include inlet temperature, feed rate, drying gas flow rate and atomization parameters. Feed solution variables such as feed concentration, solution dynamics of the feed and drug/polymer ratio also significantly impact the quality attributes of the product. Since both process variables and feed solution variables determine the final properties of the spray dried product at bulk level and molecular level, they will be discussed together. The critical physical parameters which lie at the core of the process and feed variable induced changes are vapor pressure, evaporation rate, drying time, droplet size/distribution, crystallization rate, film formation rate, heat/mass transfer and outlet temperature. At a

19

particulate/bulk level, the basic properties affected are particle size, particle shape, surface smoothness, breaking strength. Related derived properties affected are bulk density, compressibility and flowability. Of particular interest with respect to the amorphous system stability is the effect on the molecular level properties, namely, miscibility and the relaxation behavior of the components.

5.1 Feed variables

Modulating feed parameters is an important influencing factor in a successful spray drying operation. Spray drying feed for preparing ASD typically contains three important components: the pure API(s), carrier(s) and other additives and the solvent(s). Under appropriate assumptions, critical factors of this multi-component system can be identified in the Stokes-Einstein equation [49].

D=KBT6 πηr

… .. Eq .7

Here, D is the diffusion coefficient, KB is the Boltzmann constant (1.38 × 10-23 m2 kg s-2 K-1), T is the absolute temperature, η is the viscosity of the solution and r is the globular radius. Below, we discuss various aspects of the spray drying feed.

5.1.1 Solvent system

Various solvents, together or in combination, have been employed to prepare feed solutions. These solvents are either aqueous, alcohols (methanol, ethanol or isopropanol) or other organic solvents such as dichloromethane (DCM), acetone, methyl ethyl ketone, dioxane, tetrahydrofuran (THF), ethyl acetate, chloroform and acetonitrile. Amongst these, DCM is the most commonly utilized system despite its toxicity potential [47]. According to ICH Harmonised tripartite guideline for residual solvents Q3C(R5), DCM is a Class 2 solvent with concentration limit of 600 ppm [50]. This can be attributed to its low boiling point (39.8°C), high volatility and excellent solubilizing power for various drug and polymers. Common solubility of feed components in a solvent is critical to obtain molecularly dispersed solid dispersions. Incomplete solubility, precipitation or inhomogeneous mixing in the blend may be frozen due to high viscosity as a result of fast drying and eventually produce inhomogeneous component distribution in solid dispersions. Some solid dispersion carriers are hydrophilic in nature and not completely soluble in organic solvents. Often, solvent mixtures such as alcohol-DCM [51], alcohol-water [52] and acetone-methanol [53] are used to overcome this issue. A water-ethanol-DCM mixture was used to solubilize itraconazole and polyvinyl alcohol-polyethylene glycol graft copolymer (Kollicoat IR) having different solubility spectrum and ASD were obtained [54]. Choice of the mixture components and their ratio is critical as some combinations can result in different morphology and reduce the drug release [55]. Naproxen-PVP solid dispersions were found to have better miscibility when DCM-acetone solvent mixture was used followed by methanol-acetone and DCM-methanol [56]. This study showed that spray drying from solvent/anti-solvent mixture resulted in solid dispersions with greater drug-polymer

20

miscibility, lesser crystallinity and higher physical stability than when using a mixture of good solvents.

Diffusion of the drug, polymer and the solvent is important from the perspective of component distribution within the particle. Surface enrichment of a particle is given by equation 6 [41, 57].

Ei=cs ,i

cm, i=e0.5 Pei

3 βi… ..Eq .8

Here, Ei is the surface concentration of component i in relation to its average concentration in the droplet, cs,i is the surface concentration, cm,i is the average concentration of the component i and β is the profile function. Important to note in equation 8 is the peclet number (Pei) which is related to the ratio of evaporation rate (K) and diffusion coefficient of component i in the liquid phase (Di), as shown in equation 9.

Pei≈κ8Di

….. Eq .9

Equations 8 and 9 indicate that the final surface enrichment and particle morphology is an interplay of solvent evaporation rate and diffusivity. If the diffusional motion of the solute is larger than the receding droplet surface (Pei<1), the solute can move quickly to equalize concentration gradients in the droplet. This leads to uniform component distribution and formation of solid particles. In case the solutes are slow to diffuse, the droplet surface recedes relatively faster (Pei>1) and there is considerable surface enrichment. In any case the evaporation rate can vary by the choice of the solvents and so can the morphology and distribution of the components. Moreover, diffusion process of both the solvent and solutes, their solubilities and surface activities of the components need to be considered [58].

One of the relatively less explored but important aspect is the impact of the solution state behavior of the drug/polymer on the final product characteristics of the solid dispersions. Polymers can exist in different conformations (extended or compact) depending on its interaction with the solvent [59]. For hydrogen bonding to occur the molecules should be favorably/specifically aligned. Different polymer conformations in solution can lead to different degree of interaction with the drug. These solution state interactions will be translated into the solid state due to fast drying kinetics of the spray drying process. Formation of nanoaggregates in solution state by hydroxypropyl methylcellulose acetate succinate (HPMC-AS) is known to be important in supersaturation maintenance [60].

Al-Obaidi et al. investigated the impact of varying the solvent system while keeping the relative ratios of the drug and polymer same [59]. Ternary solid dispersions of griseofulvin, polyvinylpyrrolidone (PVP) and Poly[N-(2-hydroxypropyl)methacrylate] (PHPMA) were more stable when spray dried from acetone-water than from acetone-methanol mix. This was explained by weaker stabilizing griseofulvin-PHPMA interactions in acetone-methanol due to

21

stronger solution state PHPMA-PVP interactions. A direct proof of the link between polymer-state assemblies in the feed solution to post spray drying structure and performance has been established [61]. The amphiphilic diblock copolymer of N,N-dimethylacrylamide (DMA) with poly(ethylene-alt-propylene) (PEP) (PEP-PDMA) forms solution state assemblies. Interestingly, the PEP block is critical here in formation of distinct micellar assemblies. Good dissolution profiles for probucol and phenytoin were obtained when spray dried with hydrophilic polymers, poly (N,N,-dimethylacrylamide-grad-methacrylamido glucopyranose) [P(DMA-grad-MAG)] and Poly-dimethylacrylamide (PDMA). But no release was obtained for >10% drug loading. When the same drugs were spray dried with PEP-PDMA, dissolution rate was significantly enhanced for upto 50% drug loading. Importantly the dissolution advantage was obtained from PEP-PDMA system only when spray dried using methanol which acted as a selective solvent for the hydrophilic PDMA block, thereby promoting micelle formation. A non-selective solvent system (THF:Methanol (15:2 v/v)) provided solid dispersions with no solubility advantage.

5.1.2 Influence of solute components in feed solution

Solvent evaporation is an energy intensive process. One of the natural consequences of the addition of a solute is that thermal efficiency increases with solid content as less solvent has to be evaporated [39]. Solute properties are equally important as lot of feed solution properties such as viscosity, pH, evaporation rate depend on the solute and its concentration in the feed. Physical properties of a solvent such as its vapor pressure, boiling point and freezing point are affected upon addition of solute [62]. Of particular importance is the entropy driven lowering of vapor pressure of the solvent in the feed solution due to the presence of solute particles. It follows that this colligative phenomenon will be affected by the presence of the drug or the carrier. Boiling point of the solvent which is directly related to the evaporation rate is elevated by addition of solute. By figure 4, it follows that through modulation of evaporation rate, the drying process and hence particle morphology can be affected by the choice of the solute and its concentration. It has been demonstrated that addition of PVP in solvent systems consisting of various volumetric combinations of methanol, DCM and acetone showed concentration dependent deviation in evaporation rate of feed solution compared to that of the pure solvent mix owing to interaction of solvents with PVP [56, 63]. Such changes in concentration are also reflected in the tap density, morphology and relaxation behavior of the final particles. The atomization of concentrated feed leads to large diameter droplets resulting in coarser particle size and high particle density [59, 64]. Interestingly, Littringer et al. found less influence of feed concentration on the droplet size at high atomizer revolution rates [65]. This is also an example of possible complex interaction between various process parameters (in this case between feed concentration and atomizer revolution rate). The changes upon varying process parameters are not limited to bulk level changes only but are also reflected at the molecular level. Change in the feed concentration resulted in the change in the amorphous content of spray dried lactose [66]. Al-Obaidi et al. showed that the denser particles obtained by using high solid content relaxed faster as opposed to the less dense particles of griseofulvin-PHPMA-PVP dispersion system thus implying higher thermodynamic instability [59]. Such changes in the solid state

22

upon feed concentration variation are perplexing but not entirely inexplicable owing to concentration dependent changes in polymer conformations in solution state {Paudel, 2012 #165}[56] and possibility of formation of drug multimers or polycondensation [67].

The addition of a tertiary constituent in addition to the drug and the carrier can alter the product. The yield, flow properties, mean particle size and moisture content of Carbamazepine-Gelucire 50/13 solid dispersions was significantly influenced by the addition of colloidal silicon dioxide (Aerosil 200) [68]. Any preferential adsorption of a surface-active component at the air-solvent interface can be preserved in the process of spray drying [69]. Low amounts of surfactant can improve apparent solubility and dissolution process by improving wetting and micellar solubilization. But addition of higher amounts of surfactant does not guarantee improved solubility as ultrafine particles might get solubilized leading to change in particle distribution and crystallization [70].

5.1.3 Feed solution stability

Feed components within a same solution can be incompatible. Omeprazole is acid labile and degrades within 10 minutes in an acidic Eudragit-L100 polymer solution [71]. Care should be taken not to mix such incompatible feed components prior to spray drying. The amount of API added in the solvent is based on either the viscosity of the resultant feed or the solvent solubilizing capacity. Typically, the API content should be <80% of the equilibrium solubility at relevant temperature range. The characteristic feed concentration of 10 to 20% (w/v) used in the preparation of ASD normally does not result in precipitation [72]. Moreover, the most important criteria for solvent choice, i.e., solubilizing power means that problems of precipitation and/or sedimentation may not occur, especially at lab-scale. During scale-up when the feed volume is larger and the process is long, feed stability becomes more relevant good dispersion of feed components into solvent should be ensured. Nevertheless, caution should be exercised as stereo-oscillatory behavior of drugs belonging to the aryl acetic acid class of non-steroidal anti-inflammatory drugs has been reported in aqueous and non-aqueous solution [73]. Since, the stereo-selective drug-polymer interactions are relevant in solid dispersions, such interaction alterations in feed solution might be translated in the solid product [74, 75]. The feed solution can be heated as well to increase feed concentration [76], solubilize components [77] or aid drug-polymer interactions [78]. Care should be taken to maintain the feed solution temperature uniformly through the process bearing in mind the flammable potential, if any, of the solvents.

5.2 Process parameters

Various spray drying variables such as feed flow rate, inlet and outlet temperature, drying and atomization gas type and flow rate and the atomization nozzle type influence the CQA of the spray dried ASD. For instance, process parameter alterations can vary the crystallinity of the spray dried material thereby affecting porosity, flowability, sorption characteristics, solubility, dissolution rate and bioavailability [79]. In the following sub-sections,

23

the influence of varying process variables is described in detail. However, one should bear in mind that there is an interactive and complex relationship of various parameters and their influence on spray dried material. If there are two interacting factors, then the value of one factor on the target value can be dependent on the set-value of the other factor. More than two factors can also be involved. For instance, particle size is an interplay of feed concentration, droplet size, drying air temperature and feed rate [65].

5.2.1 Feed flow rate

The rate at which the feed is injected influences the droplet size, its distribution and droplet velocity [72]. The impact of the feed rate on outlet temperature and process time are relevant for ASD stability. Thermodynamically, feed injection can be viewed as mass transfer into a particular drying gas volume which will have direct implication on the outlet temperature. As the feed rate is increased, the outlet temperature decreases. Feed rate also determines the duration of the process for which particles are exposed to high temperatures. This is particularly relevant in the lab-scale equipment of non-continuous nature. Feed rate can impact the particle surface topography [65]. An increase in feed flow rate caused decrease in particle size and crystallinity of the artemisinin-maltodextrin microparticles [80]. The smaller droplets were a result of higher energy supplied for breaking the droplets. Increase in feed flow rate and pressure also resulted in a decreased heat of fusion [81].

5.2.2 Inlet and outlet temperature

The inlet/drying temperature directly affects the heat and mass transfer phenomenon in the spray drying droplet. The mass and heat transfer doesn’t remain purely convective-diffusive and convective-conductive, respectively, when the drying temperature is changed from low to higher than solvent boiling point [82]. The temperature and the moisture gradients generated inside the droplet due to higher temperatures can influence the particle formation process, create moisture gradients inside the droplet [64]. This can affect the morphology of the dried particle. Temperature variation results in different shape and surface properties such as particle roughness [65, 83]. Higher inlet temperature results in increased particle diameter [80]. Skin formation at the droplet surface can take place due to high inlet temperatures leading to solvent entrapment. The outer skin can be destroyed by solvent evaporation. Increased agglomeration at higher inlet temperature has also been postulated to cause increase in particle size. A higher drying temperature leads to a faster drying due to more heat transfer into the drying droplet [68]. This has two effects on the fate of the droplet. From a solid state perspective (Figure 2), for glass-forming materials faster drying (/cooling) rate results in accelerated conversion of the equilibrium fluid state into the non-equilibrium solid state. In other words, freezing of the disordered molecular configuration of the dissolved solids takes place at higher temperatures resulting in a high Tg product. Slower drying (/cooling) rate results in delayed dynamic arrest of the amorphous solid resulting in a low Tg product. From a process perspective the probability of moist particles striking the walls of the drying chamber is reduced translating in higher yield [68].

24

The crystallization of material within the spray dryer may occur in the liquid phase when the droplet is being dried leading to increase in solute concentration above the solubility limit, or in the solid amorphous phase [79]. The solid-state crystallization rate is dependent on the difference between particle temperature and its glass transition temperature (Tg). The solvent evaporation rate is a critical factor in decreasing the extent of the nucleation [53]. Higher solvent evaporation rate implies increased transfer of thermal energy per unit time to solute molecules present in the droplet. As a result the solute molecules undergo turbulent molecular motions and faster diffusion within the droplet. This can prevent the formation of a three dimensional crystalline network. But the relationship between inlet air temperature and crystallinity of the spray dried product does not show the same trend for all the materials. Increasing the inlet air temperature (134-210°C) resulted in higher degree of crystallinity for spray dried lactose powder [84]. Furosemide spray dried at lower inlet temperature displayed lower Tg (44°C) whereas higher inlet temperature resulted in higher Tg (54°C) [85]. Intermolecular interactions were also affected by the inlet air temperature. The interaction profile changed from complete lack of intermolecular interactions to the presence of intermolecular interactions as temperature was increased. This was further translated as increased physical stability of the amorphous form prepared at higher temperature. In contrast, spray dried ursodeoxycholic acid crystallinity reduced upon increase in spray drying inlet temperature (60–200◦C) [86]. The samples prepared at higher temperatures adsorbed more water vapor. Similar crystallinity variation with inlet air temperature was observed for spray dried artemisinin-maltodextrin microparticles [80]. Not only clear solid state differences, but even the different molecular orientation of the molecules at the surface of the apparently similar spray dried product can lead to differences in the physical stability. Amorphous Cefditoren pivoxil prepared at the inlet air temperature of 40°C and 100°C showed similar Tg’s [87]. However, only samples prepared at 40°C showed crystallization during storage at 60°C and 81% relative humidity. The difference in physical stability was explained based on the different surface properties of apparently similar dry amorphous material. Paudel et al. investigated the effect of spray drying temperature on Naproxen-PVP K25 solid dispersions and found that higher inlet temperature led to formation of amorphous phase-separated solid dispersions but with superior physical stability [88]. Higher drying temperatures resulted in increased Tg, crystallization temperature and reduced molecular mobility of lactose [64]. The inlet temperature has to be optimum as to allow adequate solvent evaporation but avoiding product degradation. Extremely high inlet air temperature can lead to drying of the feed right at the exit of the nozzle resulting in clogging of the atomization device.

As mentioned above, crystallization of material also takes place in the solid phase. Thus, the temperature to which the dried material is exposed in the drying chamber outlet and the cyclone is very important for achieving appreciable product yield, avoiding stickiness and physical instability issues. The outlet temperature is a function of solution feed rate, feed concentration, drying-gas flow rate and drying-gas inlet temperature [89]. With respect to the ASD, outlet temperature is a very critical factor. If the difference between particle temperature

25

and its Tg is increased, the solid phase crystallization rate also increases, especially when the temperature difference is greater than 30 K [79]. Outlet temperature should not be greater than the product Tg. A low yield of 2.4% w/w was obtained for spray dried Carbamazepine-PVP VA37 due to product sticking to the spray dryer walls [90]. Stickiness is a significant issue with spray dried amorphous products resulting in low yields [29, 91]. One strategy to overcome this problem is by addition of high Tg drying aids such as colloidal silicon dioxide [92].

Change in outlet temperature did not affect the particle size [93]. Particle morphology for artemisinin-maltodextrin microparticles did not change with variation in outlet temperature [80] but changed for spray dried mannitol [93]. Importantly, crystallinity of the material can be modulated by change in outlet temperature. When the outlet temperature was varied between 90°C and 157°C by varying the insulation of the cyclone, increasing temperature resulted in higher lactose crystallinity and lower yields (0.16% as compared to 47% for lower temperature) [94]. Outlet temperature can also affect the residual solvent content. Residual solvent is quite common for spray dried ASD. Even if the free solvent is removed, some bound solvent can always be present and plasticization can occur. Therefore, the drug and/or excipient stability in the presence of solvent becomes an issue worth investigating. During droplet formation the film-forming polymers quickly form skin in a few milliseconds causing solvent entrapment [57]. Solvent removal becomes progressively difficult as particle formation proceeds due to kinetic trapping of the solvent molecules in the solid matrix. Therefore, secondary drying is necessary for ASD to remove residual solvent content [95]. The drying temperature should be adequate as to remove the solvent but should not adversely impact the solid dispersion stability.

5.2.3 Atomization and drying gas type and flow rate

Atomization and drying gas type are very crucial in the spray drying process. Atomization gas type can potentially influence the droplet size, number density, and velocity, ultimately affecting the characteristics of the final product [96, 97]. Various kinds of atomization gases such as air, N2, Ar and CO2 have different physical properties which are important for the atomization process. Atomization gas property such as density and specific heat capacity are critical for the atomization process. For instance, to obtain smaller droplet sizes and higher droplet velocities, lighter gases should be employed. N2 atomizing gas (density 1.1233 kg m-3) produced smaller particles than the CO2 atomizing gas (density 1.7730 kg m-3) [96, 98]. These particles also show different morphologies. The sorption behavior of the samples atomized and dried by different gases also varies due to changes in particle size and/or shape and crystallinity variation.

When considering a gas as a drying medium, heat and mass transfer become important factors. The mass flow rate, specific heat and temperature differential of the drying gas determine the energy lost in the evaporation process [95]. CO2 provides better heat and mass transfer than air and N2 [99]. To avoid product oxidation, inert atmosphere can be obtained by using N2 as the drying gas [89]. Also process yield can vary significantly with different drying medium. Closed loop drying medium such as N2 and CO2 gases resulted in a lower yield of 40%

26

of lactose powder as compared to 70% when air was used as a drying medium due to lower absolute humidity in the latter [98]. The ability of drying gases such as CO2 to act as a plasticizer can potentially alter solid state behavior. The plasticization function depends on the solubility of the gas in the polymer which will be different for different gases. The degree of lactose crystallization was found to be highest when N2 was used as drying gas as compared to air and CO2 [98]. In this case powder morphology was altered as well.

5.2.4 Types of atomization device

Apart from the feed variables, atomizer type can affect the droplet size, size distribution, velocity and resultant spray-cone size and angle. Moreover the energy input in the different types of atomizers is significantly different. The conversion of feed to liquid droplets is a consequence of friction due to the high relative speed between the liquid stream and the surrounding air [100]. The values of the variables such as liquid feed rate can be adjusted within the constraints imposed by the atomizer type. For instance, liquid feed rate increase leads to higher droplet size for the bi-fluid nozzle and smaller droplet size for the pressure nozzle [100]. Hence, nozzle choice is important in the spray drying process optimization.

Bi-fluid nozzles (one liquid and one gas channel) are most commonly used for the pharmaceutical systems [29]. However, for some hydrophobic drug-hydrophilic polymer combinations it is difficult to be dissolved in a common solvent. Other than using solvent blends, another approach which can be employed is the multi-fluid type nozzles. In these systems the components with different solubility spectrum are dissolved separately in respective solvents and then distinctly injected into the nozzle. The atomization takes place as a result of collision of spray fluid and shear stress of the compressed air [101]. Thus, any solvent interaction effects which are possible in bulk mixing are avoided [80]. Since the liquid channels are separate, such a set-up can also be used to overcome compatibility issues [71]. 3-fluid nozzle (with two liquid and one gas channels) and 4-fluid nozzle (with two liquid and two gas channels) can be used. The variation of the relative feed rate of the different solutions and speed of feeding are important process variables for these atomization set-ups, especially when being used for encapsulation purposes. Mizoe et al. used a 4-fluid nozzle to successfully spray dry pranlukast hemihydrate-mannitol system with improved drug absorption [102]. A similar system was used for ethenzamide and flurbiprofen as poorly water-soluble drugs and lactose and mannitol as water-soluble carriers for microparticles with improved performance [103]. Achieving amorphization is an important factor to obtain enhanced dissolution rate and bioavailability of the API. Amorphization of tolbutamide was achieved with HPMC as carrier using a 4 fluid nozzle set-up [104]. A similar 4 fluid set-up was also able to produce amorphous indomethacin-HPMC systems with remarkably improved dissolution rates [105]. Such devices can also be utilized to prepare composite microparticles [106]. Flurbiprofen and sodium salicylate composite microparticles were prepared and the flurbiprofen dissolution rate was markedly improved due to increased surface area for dissolution. 3-fluid nozzle has been explored for preparing protein loaded poly(lactic-co-glycolic acid) (PLGA) microparticles [107].

27

Kondo et al. used a three-layered concentric 3-fluid nozzle with the inner and outer channels for the liquid and outermost channel for the atomization gas [101]. Ethenzamide and ethylcellulose composite microparticles prepared by this set-up were compared with those obtained by a three fluid nozzle using pre-mix solvent and a four fluid nozzle. Interestingly, most effective ethenzamide encapsulation was only obtained using a 3-fluid nozzle with separate feed solutions for drug and polymer. The particle size and size distribution obtained by the 3-fluid set-up was smaller as compared to 4-fluid nozzle. 3-fluid nozzle can also been used for spray drying of incompatible drug-polymer solutions [71]. A bi-fluid nozzle was used for spray drying omeprazole-Eudragit L100 system which resulted in microparticles which turned purple within 24 hours of storage indicating omeprazole degradation. This is expected as mixing of these two solutions prior to spray drying (in feed solution) results in omeprazole precipitation and degradation. A solution of pH 11.5 of the acid-labile omeprazole could be spray dried with an acidic Eudragit L100 solution (pH 2.4) using a 3-fluid nozzle set-up. The obtained particles were stable and showed no coloration. Microparticles obtained using bi-fluid nozzle were denser and had lower median particle size. Although external microparticle morphology did not vary with change of nozzle type, internal structure did change.

6. Carriers for spray dried solid dispersion formulations

As discussed previously, carriers are one of the most important components of the solid dispersion formulation. Their choice affects micro level properties of the ASD such as drug-polymer miscibility, intermolecular interactions and various relaxations associated with the amorphous state. The macro level properties of the ASD which have significant impact on their downstream processability are also affected by carrier choice. Typical polymer characteristics such as chemical composition, molecular weight, molecular structure, solution/melt viscosity, kinetic and thermodynamic solubility of particular API in a polymer, solubility in solvents, solubility parameter, melting point, Tg and hydrogen donor/acceptor count are taken into consideration [47]. From a regulatory perspective GRAS status of the additive is important. Based on the composition, solid dispersions can be divided into four generations with use of specific carriers for each of them [108]. Crystalline first generation solid dispersions utilized urea and sugars such as sorbitol and mannitol. Second generation solid dispersions utilized amorphous polymers which are either synthetic in origin such as poly(vinylpyrrolidone) (PVP), polyethylene glycol (PEG), crospovidone (PVP-CL), poly(1-vinylpyrrolidone-co-vinyl acetate) (PVP VA), and polymethacrylates. Cellulose derivatives such as hydroxypropyl methylcellulose (HPMC), hydroxypropylcellulose (HPC), hydroxypropyl methylcellulose phthalate (HPMCP), hydroxylpropyl methylcellulose acetate succinate (HPMC-AS) and other additives such as starch (corn starch, potato starch) and sugar glass (trehalose, sucrose, inulin) are also popular. Third generation solid dispersions utilized surfactants such as Poly(ethylene glycol)-block-poly(propylene glycol)-block-poly(ethylene glycol) (Poloxamer), Glyceryl dibehenate (Compritol 888 ATO), Lauroyl polyoxyl-32 glycerides (Gelucire), Inulin Lauryl Carbamate (Inutec SP1) and Polyvinyl caprolactam-polyvinyl acetate-polyethylene glycol graft copolymer (Soluplus). Fourth generation solid dispersions utilize ethyl cellulose, hydroxypropyl cellulose, Eudragit RL,

28

Eudragit RS, Poly(ethylene oxide) (PEO) and Poly(acrylic acid) (carbopol) to obtain controlled release. Pectin and chitosan have also been employed as carriers in spray-drying.

In recent years, more importance is being given to the supersaturation maintenance ability encompassing the concentration and duration aspects. Maintaining higher concentrations than the thermodynamic solubility after formulation dissolution in vitro/in vivo has been associated with higher absorption and hence supersaturation maintenance potential of the carrier is an important guiding factor in the carrier choice. Many carriers can inhibit precipitation by adsorbing onto the surface of nuclei and hinder crystal growth by providing steric stabilization and blocking access to the active surface [109]. In following text, we discuss in detail cellulose and vinyl based polymers which have generated tremendous amount of interest in research community for their supersaturation maintenance ability [110]. Poly (ethylene oxide) polymers and derivatives have also been discussed as they represent a different structural arrangement with respect to solid state.

6.1 Poly (ethylene oxide) polymers and derivatives

One of the most extensively studied carrier for solid dispersion formulation is Polyethylene glycol (PEG), also known as Macrogols. The molecular weight of this synthetic, semi-crystalline polymer usually lies between the range of 200- 300000, although molecular weights between 1500- 20000 only are used [111]. Particularly interesting are grades higher than molecular weight 4000 due to their less hygroscopicity and solid nature at room temperature. Their low melting point (<65°C) as well as good solubility in both aqueous and organic solvents make them good candidates for both solvent and fusion based methods. The PEG chain length, molecular weight and drug loading influences the dissolution rate of the drug. Few of the earlier studies exhibited inverse proportionality between the PEG chain length and release rate [112-114]. However, the trend is not universal as was proven by higher glyburide release from PEG 6000 as compared to PEG 4000 [115]. Similarly, different studies have pointed towards varied dependence of release rate on PEG molecular weight [112, 116, 117]. The dissolution enhancement from PEG is mediated by the conversion of the drug into the amorphous state, formation of solid solutions and reduction of the particle size. In a study where diazepam and temazepam dispersions with PEG 6000 were prepared by solvent evaporation, no advantage over physical mixtures was observed due to lack of above mentioned factors [118]. In another study, conversion of loperamide to a partially amorphous state and creation of better micro-environment by PEG 6000 for drug dissolution led to a better dissolution rate [119]. Higher release rates have also been obtained by PEG 4000, PEG 8000 and PEG 10000 [111].

PEG structure consists of seven chemical units arranged in a helical conformation with two turns [120, 121]. The PEG units arrange themselves in the lamellar form within a crystal lattice with chains existing in either extended or folded forms. The Tg of the amorphous fraction of PEG ranges from -98 to -17°C [122]. It is important to note that PEG exists as random coils in methylene chloride and chloroform but in aqueous media it exists in less ordered helical

29

configuration [123]. Moreover, PEG can crystallize into a structure with chains either fully extended or folded once or twice [124, 125]. During solid dispersion preparation PEG can crystallize. Since the crystallization phenomenon is dependent on the thermal history which in case of spray drying would be dependent on various process parameters, different PEG structures might form. This may have a significant bearing on the arrangement of the API within the PEG structure and therefore final solid state and dissolution behavior. Moreover, co-spray drying PEG with another glass former can alter the drying kinetics and effect final solid state obtained. When lactose was spray dried with 12% w/w PEG 4000, the resultant mixture was found to be crystalline. In contrast when spray dried alone, lactose formed completely amorphous material [126]. The prolonged drying time in lactose systems with PEG due to high affinity of PEG and water is apparently the reason. In another study, an apparent plasticizing effect of PEG on the spray dried lactose/PEG composites was found [127]. The extremely low Tg of the PEG means it is hard to obtain and stabilize it in completely amorphous form. In recent years there is increasing understanding about how PEG and drug influence each other’s crystallization tendency. Zhu et al. investigated this mutual influence with API’s having different physicochemical and crystallization tendencies [128]. Heterogeneous nucleation of ibuprofen and fenofibrate was promoted by PEG. The crystallization rate of benzocaine was decreased whilst the crystallization behavior of haloperidol was unaffected (negligible solubility). Entropic mixing effects and benzocaine-PEG interactions were attributed for reduced crystallization rate. On the other hand haloperidol crystallization was unaffected due to absence of favorable interactions. Benzocaine and ibuprofen also affected the crystallization tendency of PEG. Duong et al. further explored the effect of high drug loading of API on PEG 6000 crystallization kinetics in solid dispersions [129]. It was shown that indomethacin could completely transform semi-crystalline PEG to the amorphous state. Moreover, increasing drug loading had a positive effect on PEG stability. Even though these investigations were performed by a melting method, the concepts may hold true in spray dried systems as well.

Due to the low melting point, success of spray drying PEG depends on the outlet temperature during the process. Spray drying using Buchi 190 was only possible with ethanolic solutions and not from water due to higher outlet temperature in latter case [130]. Care should be taken to fix the outlet temperature below the PEG melting point to allow solidification of PEG. Lower molecular weights have slight toxicity issues which should be considered when planning formulation development.

6.2 Cellulosic Derivatives

Native cellulose consists of long chains of anhydro-D-glucopyranose units (AGU). This biopolymer is insoluble in water and most common organic solvents because of strong inter- and intra-molecular hydrogen bonding between its chains [131]. Hence, modification of the hydroxyl groups via esterification or etherification is used to modulate solubility across a spectrum of solvents at various pH conditions making them pharmaceutically relevant. Their solubility depends on the degree of substitution (maximum 3 due to presence of three hydroxyl

30

groups per glucan unit), molecular weight and uniformity of substitution. Various cellulose ether derivatives include Methyl- (MC), Ethyl- (EC), Carboxymethyl- (CMC), NaCarboxymethyl- (NaCMC), , Hydroxyethyl- (HEC), Hydroxypropyl- (HPC), Hydroxypropylmethyl- (Hypromellose/HPMC) and Ethylhydroxyethyl- (EHEC) [132]. Cellulose ester derivatives includes Ester Acetate, Acetate trimellitate, Acetate Phthalate (CAP), HPMC phthalate (HPMC-P), HPMC acetate succinate (HPMC-AS), cellulose acetate adipate propionate (CAAdP) and carboxymethylcellulose acetate butyrate (CMCAB). Acid treatment of α- cellulose also produces a partially depolymerized variant, also known as microcrystallinecellulose (MCC). Amongst these variants the most extensively employed cellulose derivatives for improving solubility of PWSD are HPMC, HPMC-AS and HPMCP. Increasing the drug to carrier ratio hastens the dissolution process due to higher hydrophilicity of the carrier. Naproxen-HPMC solid dispersions with different drug loadings were prepared by spray drying [133]. Increasing the HPMC content by a factor of 4 decreased the time required to release 50% of naproxen by 5.7 times.

Different cellulose derivatives have different potential to inhibit nucleation, crystal growth and improve dissolution rate of the PWSD [134]. This differential behavior can be attributed to different solubility of the API in the carrier, supersaturation maintenance ability, dissimilar intermolecular interactions (sometimes stereoselective) and substituent ratio. The crystallization inhibition in the solution state is well recognized as crucial in context of drug dissolution and release from amorphous drug delivery systems. Swellable controlled-release matrix tablets of butyl paraben were prepared with HPMC and HPMC-AS [135]. HPMC-AS not only maintained the supersaturation better but also turned out to be a better matrix forming agent. This was because of decreased local pH in the matrix due to carboxylic acid moieties of HPMC-AS resulting in altered swelling, very slow matrix hydration and dissolution behavior. Nielsen et al. investigated the stability of various compositions of furosemide-HPMC solid dispersions at 22°C, 33% RH and 40°C, 75% RH [136]. HPMC stabilized the spray dried system for 730 days at 22°C, 33% RH. But 80% w/w %HPMC was found to be necessary to prevent the conversion of furosemide into its polymorph under accelerated stability conditions and also dissolution studies.

Ueda et al. studied the effect of the substituent ratio on the crystallization inhibition tendency of HPMC-AS in solution [137]. HPMC-AS has different grades (L, M and H) based on different degree of substitution by methyl-, Hydroxypropyl-, acetyl- and succinoyl- moieties on the cellulose backbone [138]. The acetyl substitution increases the hydrophobic nature of the polymer whereas succinoyl substitution will increase the hydrophilic character. The acetyl:succinoyl substitution ratio (weight percent) for L, M and H grades is 8:15, 9:11 and 12:7, respectively. The solubility spectrum of these grades varies with pH, with L, M and H grades dissolving at pH ≥ 5.5, 6.0, and 6.8, respectively. The polymer can be fine (suffixed F) or granular (suffixed G). The HF solution was more effective in suppressing carbamazepine molecular mobility than the LF solution. The difference was due to the presence of hydrophobic interactions between the acetyl substituent of HPMC-AS (HF) and carbamazepine. No such

31

interactions could be detected with HPMC-AS (LF). Notably, the relation between crystallization inhibition and dissolution rate with HPMC-AS as a carrier is not favorable [139]. A series of specially synthesized HPMC-AS grades with various substitution levels were tested with various drugs such as carbamazepine, nifedipine, mefenamic acid, and dexamethasone. The crystallization inhibition was stronger but the dissolution rate was slower with lower succinoyl ratios. Recently, other than temperature and relative humidity, the degree of succinoyl substitution level of HPMC-AS has been implicated as a factor in altering the stability of the spray dried product [140]. A three term modified Arrhenius equation including the temperature, relative humidity and degree of succinoyl substitution on the polymer was needed to fit the degradation kinetics satisfactorily than the conventional two term equation.

It is important to note that enhanced solubilization does not necessarily result in high permeability of the drug [141]. Carbamazepine-Poloxamer 407 physical mixtures provided relatively higher drug concentrations (pH 7.4, 37°C) than the spray dried carbamazepine-HPMC-AS amorphous solid dispersion. However, in Caco-2 permeation assays, the permeability of the HPMC-AS amorphous solid dispersions was 3 times higher than the Poloxamer 407 physical mixtures. The micellization of carbamazepine within the hydrophobic Poloxamer 407 was attributed for reduced permeability. The authors postulated that although HPMC-AS suppressed carbamazepine self-aggregation and stabilized it in self-associated form, but, it did not incorporate the drug completely resulting in increased permeability.

HPMC and HPMCP show stereoselective interactions with nitrendipine enantiomers [75]. Solid dispersions with nitrendipine enantiomers were prepared with HPMC, HPMCP and PVP as carriers. PVP does not exhibit stereoselectivity in interacting with (+)-Nitrendipine and (-)-Nitrendipine. But HPMC and HPMCP inhibited the crystallization rate at 60°C and nucleation rate at 50-70°C more effectively for the (+)-Nitrendipine than (-)-Nitrendipine. Since the Tg of solid dispersions with both the enantiomers was the same indicating similar molecular mobility, the differences were attributed to variation in stereoselective interactions between nitrendipine and the carrier.

The carriers also have the potential to stabilize drugs degrading in certain conditions [142-144]. Quercetin and curcumin are two examples of flavonoid compounds which undergo degradation but via different mechanism. Quercetin is degraded rapidly in aqueous solutions at pH ≥ 5 mediated via a mechanism involving dioxygen addition [145]. Curcumin also undergoes alkaline degradation but via retro-aldol mechanism involving an ionic transition state [146]. The solid dispersions of these drugs with cellulose derivatives such as CMCAB, HPMC-AS, CAAdP and PVP (for comparison purposes) were prepared by spray drying [143, 144]. Cellulose ester matrices were unable to stabilize quercetin at pH 6.8 (24h, room temperature) whereas PVP offered protection against degradation. In contrast, curcumin was stabilized better by cellulose ester matrices at pH 6.8 or 7.4. The difference in stabilization profile was attributed to the fact that hydrophobic cellulose ester matrices may enhance quercetin degradation involving reaction with non-polar oxygen whereas it retards ionic state transition involving curcumin. The

32

hydrophilicity of CMCAB, HPMC-AS and CAAdP is in the order HPMC-AS>CMCAB>CAAdP. Such difference in hydrophilic nature and other properties of CMCAB, HPMC-AS and CAAdP can result in different solubilizing and stabilizing effect on the drugs. This was evident in the spray dried ASD of clarithromycin with the aforementioned structurally distinct carriers [142]. Drugs like clarithromycin provide a unique challenge to the formulator as it is better soluble in the acidic pH but also acid-labile. Clarithromycin concentration at pH 6.8 was increased in the same order as the hydrophilicity of the carriers, i.e., HPMC-AS>CMCAB>CAAdP. But when tests were conducted in more bio-relevant pH switch experiments, the higher hydrophilicity of HPMC-AS resulted in greater drug release at pH 1.2 resulting in lower drug release at pH 6.8. The weak CMCAB-clarithromycin interactions were postulated to result in high release at pH 1.2, leaving less drug to be released when pH is shifted. Interestingly the most hydrophobic polymer, i.e., CAAdP prevent acid degradation of the drug and hence resulted in maximum release at pH 6.8.

6.3 Vinyl Polymers

Vinyl polymers such as polyvinyl pyrrolidone (PVP) have been known for more than 75 years. PVP is made up of N-vinylpyrrolidone monomer units and is available in different grades (K12, K17, K25, K30 and K90) [147]. The suffix ‘K’ in PVP is a value obtained from polymer solution viscosity in water. Out of the available grades, K12 to K30 (corresponding molecular weight 2500 to 50000 Da) are commonly used in solid dispersions [111]. Higher grades are poorly water soluble, prominently viscous and thus difficult to process. The universal solubility of PVP makes it easy to spray dry along with drugs with varied solubility profile. One of the severe drawbacks of PVP is its high hygroscopicity due to which it can take up more water during storage in humid environments [148]. Polymer hygroscopicity is one of the factors responsible for amorphous-amorphous phase separation of the solid dispersions [149]. Therefore, combining vinylpyrrolidone with water insoluble vinyl-acetate results in a copolymer which is still appreciably soluble in variety of solvents but slightly more hydrophobic. Poly(1-vinylpyrrolidone-co-vinyl acetate) 64 or PVP VA64 contains 1-vinylpyrrolidone and vinyl acetate in the ratio of 6:4. PVP K30 was able to considerably enhance the dissolution rate of PWSD such as temazepam [150], curcumin [151], loratidine [152], ketoprofen [153], celecoxib [154] and probucol [155]. Same carrier increased the intrinsic dissolution rate of carbamazepine by a factor of 2.7 [156]. Some other vinyl polymers which have been used rarely include poly (vinyl alcohol) [157] and polyvinylacetal diethylaminoacetate [158].

The intermolecular interactions between drug and polymer are dependent on both the drug and the carrier involved. Notably, the presence of intermolecular interactions may not be necessary for the amorphization of the materials studied. The pyrrolidone moiety of vinyl polymers has a carbonyl group which can act as hydrogen acceptor. Weuts et al. spray dried solid dispersions of loperamide (the free base) and its fragments (fragment 1 (F1): 4-dimethylamino-N,N-dimethyl-2,2-diphenyl-butyramide and fragment 2 (F2): 4-(4-chlorophenyl)-4-piperidinol) with two different vinyl polymers i.e. PVP K30 or PVP VA64 [159]. Specific drug carrier interactions were present in F2-PVP VA64 solid dispersions but not in F2-PVP K30

33

combination. This was translated in the better amorphization of F2 in PVP VA64 systems (absence of crystallinity at <80% drug loading) than PVP K30 systems (absence of crystallinity at <40% drug loading). But no H-bonding occurred between vinyl polymers and F1 or loperamide. This effect was attributed to either larger size of the molecule causing steric hindrance and/or absence of proton donor. The loperamide-polymer dispersions were stable and no crystallinity could be detected over the whole composition range. On comparing the storage stability of these solid dispersions, F2-polymer systems were found to be more stable than those containing F1 [148]. This was attributed to lower molecular mobility of the F2 molecules due to hydrogen bonding with the polymers. Loperamide-PVP K30/PVP VA64 systems were stable at lower drug concentrations owing to formation of stable glass. During the spray drying process the hydrophobic drug can be adsorbed preferentially at the liquid-air interface. Such surface composition of the solid dispersions is a crucial aspect which can affect the wetting, dissolution and the bioavailability of the final formulation [160]. It is well known that surface crystallization of the drug is orders of magnitude faster than in the bulk (even below Tg) [161]. Hence, surface enrichment by the amorphous form of the drug can impact the stability as well. Dahlberg et al. prepared solid dispersions of two drugs with PVP and HPMC and investigated the surface properties [160]. It was found that the surface enrichment of the solid dispersions was dependent on the carrier choice and hydrophobicity of the drug.

Caution should be exercised as the presence of vinyl polymers can lead to degradation of the drug in solution. This phenomenon can be influenced by temperature and polymer concentration. Temazepam undergoes degradation in the presence of PVP K30 in solution in a concentration dependent manner [150]. Therefore, care should be taken to closely monitor stability of API in presence of the vinyl based polymers.

7. Multi-component solid dispersions

Conventionally solid dispersions are made up of drug and a carrier. With increasing insight into the mechanisms responsible for the stabilization (both in dosage form and in vivo environment), solubility and bioavailability enhancement, researchers are exploring the idea of addition of a third or even fourth component to improve performance of solid dispersions. Table 1 lists examples of solid dispersions prepared with more than two components. Surfactants such as Sodium lauryl sulphate (SLS), D-α-tocopheryl polyethylene glycol 1000 succinate (Vitamin E TPGS), Sucrose monopalmitate (Sucroester 15), Poloxamer 407, Gelucire 44/14 and 50/13, Polyoxyl 40 Stearate (Myrj 52) and Inutec SP1 are added as performance aid to improve processability or in vivo performance. A quick analysis of the studies mentioned in table 1 shows that surfactants are highly effective in increasing solubility of the PWSD due to improved wetting and micellar solubilization. But higher solubilizing effect might not necessarily be translated into higher drug permeation and bioavailability [141]. In fact sometimes by increasing the solubility and reducing the surface tension of the growing crystal, surfactants might cause precipitation in vivo which is detrimental for supersaturation maintenance [162]. Addition of large quantities of surfactants might not be feasible due to miscibility and

34

processability concerns. Many surfactants are in liquid state at room temperature raising concerns for manufacture and downstream processing. Importantly, addition of a ternary substituent has the potential to alter various aspects of the spray drying process such as drying kinetics, particle formation process and resultant product attributes. Certain drugs with pH dependent solubility show low dissolution rates. Incorporation of pH modifiers can alter the microenvironment of the solid dispersions in vivo upon dissolution and hence enhance drug release. Investigators have also explored co-spraying of glidant/anti-adherent such as colloidal silicon dioxide. Such additives can act as process aid for spray drying of low Tg drugs and electrostatic powders by avoiding sticking, reduce electrostatic interactions and thus increasable process yield.

As mentioned in previous sections, polymeric carriers increase the solubility and stabilize the amorphous drug in solid dispersions. To achieve high solubility hydrophilic nature is beneficial whereas miscibility and stability is promoted by the hydrophobic nature and interactions with carrier [163]. The benefit of diverse physicochemical profiles of the polymer can be utilized by combining polymeric carriers. Polymer blends or co-carrier systems have been utilized to obtain enhanced stability, dissolution rate and bioavailability (Table 1). The combination of polymer blends can also be used to tailor drug release. One of the hurdles to obtain such ternary systems is the requirement of miscibility between multiple components and adequate attention needs to be paid.

35

Table 1: Examples of multi-component solid dispersions prepared by solvent evaporation based techniques (including both functional additives and co-carriers)

Drug Carrier Ternary component

Function Comments References

Valsartan HPMC SLS Surfactant Valsartan:HPMC:SLS at a weight ratio of 3:1.5:0.75 gave highest solubility and dissolution of valsartan.Drug solubility and bioavailability (as compared to marketed product) increased by 43-fold and 1.7- fold respectively.

[164]

Sirolimus HPCHPMCPVP VA64PEG 8000PVP K30

Vitamin E TPGSSLSSucroester 15Poloxamer 407Gelucire 44/14 and 50/13Myrj 52

Surfactant Various drug:polymer:surfactant (1:8:1) combinations were tested by solvent-casting. Drug solubilization 1h after dissolution enhanced to maximum extent by TPGS, Sucroester 15 and poloxamer 407. HPMC-TPGS and HPMC-Sucroester 15 were most effective at 24h after dissolution. Spray dried dispersions with TPGS and Sucroester resulted in significantly enhanced AUC and Cmax in Male Sprague–Dawley rats.

[165]

Carbamazepine

HPMC-AS Poloxamer 407 Surfactant Addition of poloxamer 407 increases the concentration of carbamazepine.

[141]

Carbamazepine

PVP K30 Vitamin E TPGSGelucire 44/14

Surfactant Intrinsic dissolution rate in comparison to crystalline drug is enhanced 2.9 times with TPGS and 1.7 times with Gelucire 44/14.

[156]

UC 781 HPMC PVP VA64

Vitamin E TPGS Surfactant 100% drug released in 5 minutes for UC 781-PVP VA64-TPGS and 30 minutes for UC 781-HPMC-TPGS systems. Tablet

[166]

36

formulations containing HPMC swell slowly as opposed to PVP VA64 containing tablets which disintegrate in 4 minutes.

Itraconazole PVP VA64 Inutec SP1 Surfactant Itraconazole was molecularly dispersed only in the PVP VA64 phase. No molecular level interactions of Inutec SP1 with other components.

[167]

Itraconazole PVP VA64 TPGS 1000 Surfactant Addition of TPGS results in improved itraconazole dissolution during first hour. Subsequently, precipitation takes place due to small crystalline itraconazole fraction and TPGS surfactant properties.

[168]

Itraconazole PVP VA64 Myrj 52 Surfactant Addition of surfactant expels the drug from the polymer phase indicating surfactant-drug-polymer system incompatibility.

[169]

Telmisartan PVP K30 Sodium carbonate pH modifier Drug:PVP K30:Sodium Carbonate in the weight ratio 2:0.5:3 improved dissolution rate by a factor of 3 and exhibited superior in vivo performance..

[170]

Tenoxicam PVP L-arginine pH modifier Tenoxicam and L-arginine molecules were singly ionized and situated as “amorphous salt” in ASD.

[171]

Rebamipide PVP VA64 L-lysine pH modifier Rebamipide exhibited 2.7 times higher percent inhibition in the ulcer-induced rat models than the reference product.

[51]

Celecoxib PVP Meglumine pH modifier Generation of amorphous product with enhanced performance which is stable for 3 months under accelerated storage conditions.

[172]

37

Sibutramine base

HPMC and poloxamer

Citric acid pH modifier Solid dispersion with composition Drug:HPMC:poloxamer:citric acid of 5:3:3:0.2 exhibited maximum solubility. The solid dispersions gave performance comparable to the marketed product. However, the drug was not completely amorphized.

[173]

Tacrolimus CMC-Na Citric acidSLS

pH modifierSurfactant/solubilizer

Solid dispersions with Drug:CMC-Na:SLS:citric acid in the ratio of 3/24/3/0.2 improved drug solubility (2000 fold) and dissolution rate (10 fold).

[174]

Simvastatin PVP Colloidal silicondioxide

Glidant/anti-adherent

Aerosil assisted in the spray drying of low Tg drug. Drug:PVP:Aerosil in the ratio 1:2:2 (w/w) was optimum for drug content, saturation solubility improvement (5-fold) and ease of tablet formation.

[175]

Carvedilol PVP Colloidal silicondioxide

Glidant/anti-adherent

Solid dispersions with Drug:PVP:Aerosil 200 in the ratio of 1:2:2 (w/w) was able to stabilize the low Tg drug, provide a 2-fold increase in apparent solubility.

[176]

Nitrendipine PVPPVA

Colloidal silicondioxide Tween 80

Glidant/anti-adherentSurfactant

Amorphization of nitrendipine as result of pulse combustion drying without use of organic solvents. Hydrophilicity, dispersibility of the additives aided amorphization and dissolution rate improvement. The system was also effective in maintaining supersaturation.

[177]

Curcumin Gelucire 44/14

Colloidal silicondioxide

Dryingaid/glidant/anti-adherent

Drying aid was imperative to avoid stickiness. Drug solubility was increased by 3200-fold. Dissolution test revealed

[178]

38

90% curcumin release after 10 min.Clopidogrel napadisilate

HPMC Colloidal silicondioxide

Glidant/anti-adherent

Dispersions of Drug:HPMC:colloidal silica at a weight ratio of 11:3:3.5 led to 6.5-fold improvement in solubility and enhanced stability, and bioequivalence to the commercial product.

[179]

Griseofulvin PVP PHPMA (Co-) Carrier 2:1:1 w/w combination of Drug:PVP:PHPMA ratio was spray dried. Product properties varied with the solvent system used to spray dry.

[59]

GriseofulvinProgesteronePhenindione

PVP PHPMA (Co-) Carrier Ternary solid dispersions exhibited enhanced stability as compared to binary solid dispersions. PHPMA can form H-bonds with the drug and polymer thus stabilizing the system. The stability in ternary systems and free energy of mixing was in the sequence griseofulvin > progesterone > Phenindione.

[180]

Indomethacin PVPK90 Eudragit E100 (Co-) Carrier Ternary dispersions with polymers used in low concentrations provided superior stability and dissolution rate. The drug-polymer interactions were maintained in ternary dispersions along with greater indomethacin solubility in polymer mix.

[181]

Fluconazole PVP HPMCChitosan

(Co-) Carrier Fluconazole release rate could be adapted via varying the weight ratio of PVP with HPMC or chitosan. PVP/HPMC blend was suitable for immediate release but PVP/chitosan blends provided sustained release.

[182]

Atorvastatin HPMC Nicotinamide (Co-) Carrier 4 fold increase in dissolution rate and [183]

39

calcium trihydrate

significant increase in bioavailability. Ternary solid dispersions stabilized due to H-bonding between drug and other components.

Griseofulvin HPMC-AS PHPMA (Co-) Carrier Dissolution and stability profile of binary and ternary systems were comparable. Ternary systems showed higher Tg than the binary systems. PHPMA enhanced hydrogen bonding.

[184]

Not used HPMC-ASHPMCPVPEudragit E100

PVPCMCABHPMCPVP

HPMC

(Co-) Carrier Pairwise polymer blends were investigated for potential applications in ASD. HPMC-AS/PVP, HPMC/CMCAB and PVP/HPMC were miscible in all proportions. Miscibility gap was observed for Eudragit E100/PVP and Eudragit/HPMC blends.

[163]

40

8. Influence of preparation method

Apart from spray drying, ASD can be prepared by various manufacturing methods such as co-grinding, freeze drying, hot melt extrusion (HME), supercritical methods (SCM) and electrohydrodynamic based methods. Solid dispersions are kinetically stabilized systems and therefore the choice of the manufacturing method has a significant impact not only on the external morphology but also on the intricate arrangement of the drug molecules relative to one another. This is a direct consequence of differences in the degree of disorder of the starting material, energy input, process time, drug-carrier mixing and behavior of the formulation components in response to the process induced stress. Different processing methods can result in different molecular relaxation times which implies different molecular mobility and hence stability [185]. With respect to the ASD manufacture the most critical factor is the solid state of the drug and its miscibility in the carrier system.

Various examples in literature point towards the effect of manufacturing method on the solid state characteristics, particle morphology, moisture content, dissolution rate and bioavailability. Sugimoto et al. compared these parameters for the nifedipine-polyethylene glycol 6000-HPMC system prepared by spray drying and co-grinding [186]. Unlike the spray dried systems, the co-ground systems were not amorphous but still gave a higher dissolution rate than the spray dried powder. The difference in the solid state can be attributed to the fact that in spray drying the crystal lattice structure of the drug is destroyed by dissolution in the feed solvent prior to amorphous form generation whereas co-grinding is a slow comminution process where crystal structure is gradually depleted due to mechanical stresses. Spray dried sulfathiazole-PVP and sulfadimidine-PVP solid dispersions were amorphous over a wider range of concentration than their milled counterparts [187]. Another study compared the hydrocortisone-PVP ASD prepared by spray drying and freeze drying [188]. The origin of amorphization in spray drying is a result of fast evaporation whereas in freeze drying it is mainly due to rapid freezing. The spray drying process gave spherical particles whereas freeze drying generated irregular, flake like appearance. Importantly, the spray dried solid dispersions had a lower surface area and higher moisture content than the freeze dried product. This resulted in higher hydrocortisone release from freeze-dried samples. Solvent evaporation was compared to the SCM for preparation of carbamazepine-PVP K30 solid dispersions with either Gelucire 44/14 or Vitamin E TPGS [156]. SCM gave higher intrinsic dissolution rate for carbamazepine-PVP K30 solid dispersions. But when Gelucire 44/14 or TPGS were added, there was either very little or no improvement in the dissolution process. No differences in the amorphous nature or drug-polymer interactions were reported by the authors. Won et al. also compared the solvent evaporation method with SCM using felodipine-HPMC-surfactant systems [189]. SCM resulted in improved equilibrium solubility but no major differences were obtained during in vitro dissolution. The method of preparation can affect the surface coverage of the ASD components. Surface coverage of the solid dispersion particles containing HPMC and PVP K30 as carriers prepared by spray drying and rota-evaporation was compared [160]. The surface enrichment of the drug was more for the spray dried product when compared to the slowly dried rota-

41

evaporated product. Since PWSD are hydrophobic, such preferential accumulation of the drug on the particle surface can be detrimental for in vitro dissolution.

Amongst all the techniques available, spray drying and HME are most widely used. HME is a solvent free process in which the drug-polymer mixture is melted due to combined effect of the high barrel temperature and mechanical stress employed by the mixing screw elements and extrusion through the die cavity. Since both the techniques are established on an industrial scale, they are an equally viable option for preparing solid dispersions. However, certain factors direct the choice amongst these manufacturing methods. These include the drug properties such as its solubility in the solvent or polymer, log P value and degradation temperature. Spray drying has the advantage that it can be used for thermolabile and high melting drugs. Moreover, it is useful at initial drug development stages due to less material requirement for the process. The homogeneity of the solid dispersions prepared from HME and spray drying depends on the process parameters and characteristics of the drug and the carrier. Guns et al. reported that HME provided higher kinetic miscibility between miconazole and Kollicoat IR as compared to spray drying [78]. Opposite results were obtained when felodipine solid dispersions were prepared with PVP and HPMC-AS at different drug:polymer ratios [190]. Spray dried samples were amorphous even at higher drug loading as opposed to the HME dispersions. Felodipine release rate was faster with the spray dried solid dispersions. It should be mentioned that conventional HME process often requires an additional downstream process of extrudates milling. This can be a contributing factor to ASD destabilization. Patterson et al. investigated the influence of preparation method on the physicochemical properties of the glass solutions of carbamazepine, dipyridamole, and indomethacin with PVP K30 [191]. Same drug:PVP K30 ratio of 1:2 (w/w) was used for ball-milling, spray drying and HME. Samples obtained from different techniques were amorphous and exhibited only a single Tg indicating homogeneity. But further investigations using dispersive Raman microscopy indicated the presence of carbamazepine clusters in the solid dispersions prepared by ball-milling indicating heterogeneity. Ball-milled and spray dried samples increased the dissolution rate more effectively than the spray dried powder. The residual solvent content of the products obtained from the spray drying process is a critical factor in ASD stability. In contrast to the HME systems, spray dried felodipine-PVP and felodipine-HPMC-AS dispersions were not stable upon exposure to 40°C/75% RH for 8 weeks [190]. Similar observations were obtained for compound X-PVP VA64 (1:2 w/w) solid dispersions [192]. Apart from differences in solid state, different material properties such as powder density, surface area, morphology and flow properties can be expected [192]. Such differences can affect the final dosage form performance. Antipyrine-HPMC solid dispersions were prepared by either spray drying or rota-evaporation/milling [193]. The two different processes provided differently sized particles and resulted in tablets showing different gelling behavior. The tablets prepared by larger particles obtained via rota-evaporation/milling provided an inhomogeneous gel as compared to the gel formed by the smaller size spray dried material.

9. Recent advances

42

9.1 Electrospraying

Lower particle size is a desirable property as the concomitant increase in surface area improves dissolution rate. The quest for obtaining such product characteristics has led to focus on innovative atomization systems which use different forms of energy than the conventional nozzles based on kinetic energy, pressure energy, centrifugal force or piezoelectricity for droplet generation. Electrohydrodynamic spraying or electrospraying is based on application of electrical energy for atomization of feed liquid [44, 194]. The liquid feed is injected through a capillary maintained at high potential in tune of kilovolts. As the meniscus is formed at the tip of the capillary it encounters the applied electric field resulting in reduced surface tension and droplet formation [195]. The charged droplets repel each other resulting, i.e., self-dispersing resulting in smaller particles.

Variation in the operating conditions determines the shape of the liquid meniscus formed, its motion and subsequent breakup. The different jet formation and breakup patterns or in other words electrospray modes are dripping, microdripping, spindle, multispindle, cone-jet, oscillating jet, precession and multijet type [196, 197]. Similar to spray drying, the surface tension and viscosity of the liquid, the liquid injection rate and geometry of the system affect the process. Additionally, the process is extremely sensitive to the electrical conductivity of the liquid, applied voltage and the dielectric strength of the medium.

Electrospraying is an attractive strategy since it can produce quasi-monodisperse low particle size (nanometer range) products [197]. Notwithstanding the advantages, there are still few drawbacks which need to be addressed before this technique can be widely used for pharmaceutical manufacturing. The throughput of the electrospraying set-up is limited by the low feed rates which are typically used in to attain the most-preferred cone-jet mode [44]. The process yield can be increased via use of multiple nozzles or spray through uniform holes in a metal plate [198, 199]. More development is warranted to overcome this issue and make the technique truly scalable to GMP production levels. Another disadvantage is that the constraints imposed on the physical properties of the liquid such as surface tension, conductivity and dipole moment which reduces the favorable design space [200-202].

Electrospraying has been utilized to fabricate particles with core-shell structure to improve solubility of PWSD [203]. Acyclovir was incorporated into the PVP K30 core along with SDS and sucralose. MDSC and XRD results indicated the crystalline to amorphous conversion of the microparticle components. The final performance of the microparticles viz. in vitro acyclovir dissolution and permeation were improved. Roine et al. prepared griseofulvin-Eudragit L100-55 micromatrix particles by a dual-capillary electrospraying device [204]. The process could successfully convert ca. 90% of griseofulvin to the amorphous form which was otherwise difficult using conventional spray drying [205]. The crystalline griseofulvin had dimensions of ≤ 8 nm. Thus, the process provided a double advantage, i.e., conversion of the majority of griseofulvin to the amorphous state and the remaining crystalline fraction was also of nano dimensions. This ultimately translated into improved dissolution rate at enteric pH and

43

enhanced permeability. The presence of nanocrystallites did not adversely affect the particles which were stable for 6 months (RH = 67%, T = 30 °C). However, the process yield was only 15%. Nguyen et al. prepared darunavir-HPMC solid dispersions encapsulated with in a Eudragit L100 shell by electrospraying [206]. Although the obtained core-shell structured nanoparticles were not completely amorphous, they provided the advantages of a high encapsulation efficiency (90%) and reduced darunavir release (<20%) in the acidic medium. Electrospraying can also be combined with electrospinning to provide dual drug release properties [207]. Amorphous nanocomposites of ketoprofen were prepared with PVP as the sheath material and Eudragit L100 forming the core using a concentric spinneret. For the sheath fluid alone the electrospraying process was observed whereas for the core fluid single fluid electrospinning process took place. On simultaneous pumping of the sheath and core fluids, straight jet was emitted from a compound taylor cone at a semi-vertical angle of 57°. The core sheath fibers provided dual drug release in vitro at acidic (35.1 % release) and enteric pH (62.2 % release).

9.2 Pulse combustion spray dryer (PCSD)

Spray dryers are not ideal adiabatic systems and a substantial part of the input energy is wasted [208]. Pulse combustion drying is a promising drying technology with the aim of process intensification [209]. Unlike electrospraying, the atomization of the feed takes place due to sound waves produced by the combustor [210]. The major difference between a conventional spray dryer and PCSD lies in the drying function. Instead of a constant drying air flow, PCSD employs intermittent (pulse) high-temperature shock waves repeating itself at a frequency of 50-1000 Hz [211]. The hot high pressure gases and resulting shock waves are generated as a result of back-to-back combustion of fuel-air mixture in the pulse combustion chamber. The feed liquid is atomized into the drying gas flow resulting in particle formation. Drying takes place via actions of shock waves, ultrasonic waves (> 155 dB), gas flow and gas temperature (> 200°C) in the drying chamber.

Pulse combustion dryers can improve the drying rate by 1.2 to 3 times, reduce air consumption and air emissions [212, 213]. One of the major constraints associated with the PCSD is the level of noise generated. Also, the particle formation process, effect of various process variables on the macro- and micro-scale properties of the solid dispersions is largely unknown. Detailed studies investigating various process variables need to draw attention of the pharmaceutical community.

Wang et al. prepared nitrendipine-Aerosil-Tween 80 solid dispersions using the Hypulcon pulse combustion dryer system (HPCDS) and a conventional spray dryer [177]. The dispersion particles prepared with the HPCDS equipment exhibited no agglomeration and had smaller particle size with narrow size distribution. Importantly, the drug was in the amorphous state. The dissolution of nitrendipine was significantly improved as well. HPCDS was also used to manufacture ibuprofen dispersions with four different polymers, viz., PVP VA64, PVP K25, PVP K30 and crosslinked PVP [214]. The dispersions particles were found to be amorphous using MDSC and XRPD analysis. The dissolution properties of ibuprofen was markedly improved,

44

even when compared to solid dispersion particles prepared by conventional spray drying. The Ibuprofen-Kollicoat IR solid dispersions showed a 50 fold increase in ibuprofen dissolution rate ̴ when prepared using HPCDS [215].

10. Scale-up challenges

Scalability refers to the possibility of increasing the throughput of the manufacturing process in an energy efficient manner without compromising the CQA of the product. The natural consequence of different API availability and demand projection at different stages of drug development is the marketing of equipment with different capacities. Three different scales of spray drying equipment available are laboratory, pilot and production scale. To understand the challenges associated with scaling up it is imperative to know the specific changes which occur in various stages of spray drying when scaled up.

As a shift from lab to production scale is made the configuration and dimensions of the spray drying equipment should be adjusted. Such adjustments involve but are not limited to a.) change in feeding system; b.) atomization device type, location and conditions; c.) drying gas dispensing system; d.) chamber dimensions; e.) exhaust gas duct and; f.) shift from single pass mode to the multiple pass mode. These alterations can change the droplet trajectories, evaporation rate, increase the drying time, solvent mass in the drying air and alter the wall deposition profiles [31]. Consequently, CQA such as particle size and its distribution and residual solvent content are affected.

Upon scale-up of the process, the amount of feed solution to be prepared and its hold time is increased. Adequately sized spray-solution tanks are needed to support the throughput and desired batch size. Degradation kinetics of the feed solution should be taken into account and adequate storage temperature should be determined to ensure solid solubility, feed stability and avoid chemical degradation or precipitation. Atomization set-up and conditions is a critical aspect in spray dryer scale-up as it has direct influence on the particle size which effects flowability, compressibility and dissolution performance. The smaller dimensions of the drying chamber at small scale necessitates use of external mixing bi-fluid nozzles (narrow spray angle, 20°) resulting in small particle size (d̴ V50<10 µm ) and low bulk density (0.15g/ml) [29, 95]. The resultant poor flowability is detrimental for capsule filling and/or tableting. The change in production scale does not necessarily result in different homogeneity of the solid dispersions as was observed for the acetaminophen-PVP system [216]. Despite of using different orifice diameters of the bi-fluid nozzle for pilot and production scale, similar solid state of the dispersions were obtained. Nonetheless, the quest for obtaining a homogenous amorphous product at lab-scale often overrides the downstream processability which is addressed during scale-up [31]. Laboratory scale dryers are generally open-loop systems. But process costs and environmental concerns result in application of closed loop systems for many large pilot- and production-scale equipment. The process conditions need to be optimized if a switch is made between one of these modes [89]. The residual solvent content should be brought within acceptable range by use of secondary drying unit operations. Usually vacuum dryers or

45

convection tray-dryers are employed [95]. The solvent is trapped within the solid dispersion matrix and should diffuse out. When drying at large scale, the diffusion of solvent through the powder bed becomes a significant factor and needs to be considered [95]. The residence time of the product in the spray dryer at high outlet temperatures can be critical for the stability of ASD. One outcome of scale-up is the increased residence time resulting in chemical degradation, phase-separation or crystallization of amorphous products. Product age-chemical degradation profiles can be generated. Accordingly, process optimization to change the residence time and/or residual solvent content should be performed.

Development of a stable and robust lab-scale method is a prerequisite for any scale-up [95]. This process acts a guideline for modeling of thermodynamics, atomization and particle formation. Nevertheless, the design space at different scales might not necessarily be the same and requires verification, optimization and validation. Thermodynamics modelling aids in calculating relative saturation of the drying gas. It also provides a useful way to compare thermodynamics spray drying parameters of different scales on the same plot. For this purpose, specific drying ratio is used [95]. Atomization modelling aids in droplet size and process throughput estimation. It also takes into account any change in the atomizer type. Particle formation modelling accounts for drying kinetics and final morphology of the particles.

11. Quality by design (QbD) and process analytical tools (PAT) in spray drying

The success of a spray drying process scale-up is vested in the detailed knowledge and understanding of the parametric response and inter dependability of variables. This makes QbD and PAT extremely useful with respect to process transfer from a lab to production plant. Additionally, these tools also aid in avoiding material/product wastage due to batch failure, increasing process efficiency and continuous process improvement within the design space. QbD is described by United States Food and Drug Administration (U.S. FDA) in its Guidance for Industry: Q8 (R2) Pharmaceutical Development [217] as “A systematic approach to development that begins with predefined objectives and emphasizes product and process understanding and process control, based on sound science and quality risk management”. QbD involves defining CQA and critical process parameters (CPP) through risk analysis and subsequent identification of their relationship through design of experiments (DOE) approach [218]. DOE has been used extensively in spray drying to understand the complex relationship between various process parameters and its effect on product quality [219-226].

Statistical tools such as DOE and response surface analysis help to create a design space. According to the Current guidance on pharmaceutical development ICH Q8(R2) guidelines, design space is the multidimensional combination and interaction of formulation variables and process parameters resulting in product with assured quality [227]. The thermodynamic design space for a spray drying process for ASD is limited by some product specific constraints [89, 95]. The maximum inlet temperature is limited by the relevant thermal degradation temperature of the feed constituents in relevant time scale as well as the possibility of instantaneous drying of the droplet at atomizer openings. The minimum inlet temperature should be sufficiently high to

46

dry the droplets before they strike the spray dryer walls. The outlet temperature is directly related to the inlet air temperature and is constrained by the solid state stability of the ASD. Of crucial importance is the residual moisture content and its effect on the product Tg. This would have a dual impact on the product stability and product yield. If the outlet temperature is higher than the product Tg at a particular residual solvent content, stickiness would occur resulting in low process yields. The inlet and outlet temperatures are also constrained by the product attributes such as density which are crucial for downstream processing. The process efficiency is constrained by the feed rate. Low solution feed rates can result in better atomization and smaller droplet size but reduce process efficiency. The design space is crucial in post approval manufacturing as if the formulation and process variables are varied within the design space then no regulatory post-approval change is warranted [227].

In order to obtain enhanced control over the process, PAT tools are employed. In U.S. FDA guidance for Industry regarding PAT [228], PAT is defined as a “A system for designing, analyzing, and controlling manufacturing through timely measurements (i.e., during processing) of critical quality and performance attributes of raw and in-process materials and processes with the goal of ensuring final product quality”. Typically in the lab-scale spray drying equipment, temperature, relative humidity and pressure probes are employed. PAT tools for spray drying in small scale equipment are not so widely employed until now except for research purposes. However, at pilot and industrial scale they are extensively used and encompass various aspects of the process. Consistency of the raw material is important from regulatory perspective and handheld near infra-red (NIR) devices can be used for raw material characterization. Process turbidimetry, viscosimetry and laser diffraction based devices ensure that feed has optimum physical characteristics before being sprayed. Particle sizing of solute in feed is extremely important not only to avoid nozzle clogging but to ensure that no solute crystallization takes place in the solution itself. During the process PAT tools are employed for real-time spray pattern and particle size distribution analysis. Spray pattern analysis can prevent deposition on drying chamber and determine if droplet size distribution is appropriate. Particle size distribution is critical for dissolution and tableting of the powder [229, 230]. Specifically for the spray drying of ASD, NIR and Raman probes are useful to analyze the polymorphic changes in the product and exhaust gas analysis for solvent content. In order to reduce residual solvent content to appropriate level secondary drying step is required [95]. Process mass spectrometry is useful to monitor the exhaust gas and NIR probes can perform end-point solid state characterization. All of these PAT tools can be employed in-line and help maintain process feedback loops which leads to process correction by analyzing the real-time data thereby saving time and money.

12. Downstream processing and product development

Tablets and capsules are the most popular final dosage forms to deliver drugs orally as they provide advantages of patient compliance, low cost and formulation related factors. Their popularity as a dosage platform for solid dispersions is evident by the fact that most (if not all) of the marketed products are either tablets or capsules [108]. Spray dried powders as such are

47

not be ideal for conversion into tablets or capsules and require downstream processing. The CQA of spray dried ASD which are important with respect to their downstream processing are residual solvent, particle size, bulk density, flowability, compressibility and compactability, disintegration and stability. Various aspects of the downstream processing are depicted in the figure 8. For large-scale manufacturing, powder flowability is a relevant CQA irrespective of the intended final dosage form. Spray dried powders generally have a relatively small particle size and low bulk density resulting in poor flowability. Depending on the droplet drying rate and temperature, the particles may have hollow sphere morphology with a low bulk density (<0.2 g/cm³) or shriveled raisin morphology having high bulk density (0.2-0.4 g/cm³) [95]. Other morphologies are also possible, however, large sized spherical particles generally give better flow properties and compressibility which can be achieved via spray drying by process optimization [89]. Pressure nozzle is useful to achieve such particles which might even be suitable for direct compression. A pre-compaction step can be employed for low bulk-density solid dispersions to improve their flowability [231]. Furthermore, dry and wet-granulation are viable strategies to improve solid dispersions handling.

48

Figure 8: Downstream processing of the spray dried ASD

The spray dried powder should have certain CQA (particle size and its distribution, bulk density, flowability etc.) for successful downstream processing. The powder can be subjected to a pre-densification step and/or granulation to improve its formulatibility. The powder/granules with improved flowability are either compressed or filled in capsules and packaged. The thermal, moisture and mechanical stresses involved in the unit operations have the potential to effect the amorphous solid dispersion adversely. The final product should adhere to the quality target product profile.

While making a choice for any downstream approach it should be kept in mind that amorphous systems are already thermodynamically unstable. Even a minor mechanical activation such as scratching with a pin has been implicated in the formation of stable crystalline polymorph of bicalutamide in frozen glass [232]. The fact that antiplasticization due to high molecular weight polymers, enhanced local viscosity and the drug-polymer interactions are critical for maintaining the drug in amorphous state implies that any downstream process affecting these stabilization mechanisms may have detrimental effects on the solid state stability. Mechanical and heat stresses involved in dry granulation via roller compaction can affect ASD stability undesirably. Roller compaction has not only been found to alter the solid fraction, tensile strength and flexural modulus of the excipients but also higher degree of crystallinity [233, 234]. Notably, roller compaction reduced the % bioavailability of the drug achieved from ASD

49

from 81% to 52% due to roller compaction [235]. Wet granulation has been reported for making tablets of glimepiride-crospovidone [236] and nimodipine-Eudragit EPO/PVP VA64 systems [237]. Exposure to solvent during wet granulation can cause plasticization and result in physical instability of the amorphous systems. This was observed for nimodipine-Eudragit EPO/PVP VA64 tablets which showed reduced drug release due to crystallization upon storage for 2 months at 40°C and 75% RH [237]. Direct compression involves less number of steps and has been shown to be favorable with respect to stability. This is particularly relevant when the dosage of the active ingredient is <30% of the tablet mass [238].

Upon obtaining a powder with optimum flow properties it is converted into the final dosage form via capsule filling or tableting. Carriers used in solid dispersions such as PVP and PVP VA64 undergo high degree of plastic deformation. Tableting step imparts compressive and thermal stress to the ASD which might induce plastic deformation induced alteration in molecular mobility and/or intermolecular interactions of the ASD resulting in phase behavior changes. Naproxen-PVP K25 ASD with 20, 30 and 40% (w/w) drug loading were spray dried and subjected to 188, 376, 565 and 1130 MPa compression pressure for a dwell time of one minute [35]. Compression resulted in the demixing of the ASD above the threshold composition and compression pressure of 30% w/w naproxen and 376.7 MPa, respectively. The demixing phenomenon was attributed to the compression induced dynamic transition of dihedral angles defined by the vectors along the pyrrolidone ring, planar vinyl side chains, and between pyrrolidone and vinyl side chain backbones. In another study, spray dried naproxen-PVP VA64 solid dispersions with 10, 20, 25, 30, 40, and 50% (w/w) drug loading were compressed at 188, 753 and 1130 MPa for 10 seconds [239]. Although no specific conclusive trends could be drawn from the thermal markers, attenuated total reflectance Fourier Transform Infrared Spectroscopy (ATR-FTIR) data indicated enhanced specific interactions between the hydrogen of the carboxylic acid functional group of naproxen and the amide carbonyl of PVP VA64. Upon storage for 21 days at 75% RH at ambient temperature, the samples compressed at highest compression pressure showed better homogeneity due to enhanced weak drug-polymer interactions. Singh et al. studied the effect of compression on the phase-behavior of non-interacting spray dried miconazole-PVP VA64 solid dispersions [18]. The solid dispersions with 10, 20, 30 and 40% (w/w) of miconazole were subjected to 188, 564 and 1,129 MPa for a uniform dwell time of 60 seconds. The lower drug loaded (10, 20% drug content) dispersions showed no variation in the Tg or Tg width. But the 30 and 40% compositions underwent mixing which was indicated by the reduction in number of Tg’s upon compression. The authors attributed compression induced increase in the molecular mobility to be the causative factor for observed mixing of drug-rich and polymer-rich phase. It is mentionable that the compression pressures and dwell times studied in above mentioned studies were relatively higher than what is normally encountered in pharmaceutical tableting. Therefore, Singh et al. studied the compression effects on miconazole-PVP VA64 systems at relatively lower compression pressures (60, 120 and 200 MPa) and dwell times (1, 20 and 60 seconds) [19]. Phase-separated systems became more homogeneous upon compression even at low

50

compression pressure and dwell time combinations. However, mixing was favored at high compression pressures and dwell times. The X-ray diffraction studies showed halo patterns with two broad peaks at 12.31° (I1) and 22.15° (I2) 2θ values, the ratio (I1/I2) of which decreased upon compression. These changes point towards polymer component mediated changes in phase-behavior due to compression. From the above mentioned studies it is clear that a compressive step can induce phase behavior changes in a compression pressure and dwell time dependent manner. During large scale-tablet manufacture, friction between tool parts moving at high-speeds can result in rise in temperatures higher than 50°C. This is relevant for pharmaceutical tableting as compression temperature can modify the microstructures and mechanical behavior of the tablets, especially for materials characterized by low transition temperatures [240, 241]. In comparison to tableting, capsule filling process seems to be less damaging to the ASD. But it should be noted that hard gelatin capsules may contain 13-16% (w/w) moisture content to impart flexibility and avoid brittleness of the capsule shell [242]. Solid dispersions made of a hygroscopic carrier such as PVP can possibly take up moisture from the capsule shell during its shelf-life which may have disastrous consequences for both formulation stability and dosage form intactness.

Irrespective of the method of manufacture and final dosage form, modulation of powder properties is required and additives are added to make the final formulation. These additives act as binder, filler, lubricant, glidant and disintegrant. Two major issues encountered with ASD formulations is the rapid crystallization of the amorphous form during dissolution [243] and formation of gelling polymer networks (GPN) [244], and choice of the additives can be used to mitigate these problems. Many of the vinyl and cellulose based carriers used for solid dispersions act as binders due to their plastically deforming nature. The resultant hard tablets can result in formation of GPN which warrants the addition of sufficient disintegrant. GPN formation can result in slower release kinetics as shown for UC781 from HPMC2910 tablets [166]. Caution should be exercised while adding disintegrants as specific sites on the glassy network of the disintegrants can interact with water molecules resulting in drop of Tg [245]. This can be detrimental for both, the ASD stability and tablet properties. Fillers such as mannitol, lactose or MCC are used to increase formulation bulk in both tablets and capsules. Capsules of amorphous BMS-48804 containing MCC filler resulted in less release than lactose filler [243]. The lowered release was due to formation of a hard plug as a result of deposition of crystallized drug on the surface of the capsules. However, MCC has been found to provide better stability of amorphous ibipinabant due to its cushioning effect providing protection to the amorphous particles [233]. Care should be taken while coating ASD tablets as exposure to moisture and heat involved in the process can lead to crystallization [233].

The ultimate goal of all of the above mentioned downstream processing steps is to generate a final product which confers to the Quality target product profile (QTPP). Thus the product should release the drug according to desired kinetics, provide appropriate in vivo performance (release kinetics, supersaturation maintenance, bioavailability) and should be stable for its intended shelf-life at planned storage conditions. For further information about downstream

51

processing of solid dispersions prepared by other manufacturing methods the readers are referred to recent excellent publications by Demuth et al. [244] and Page et al. [231].

13. Future outlook

Spray drying is a versatile technique which has been employed extensively for generating amorphous forms of drugs, particle engineering and as an efficient drying technique. Notwithstanding the several strengths which have been summarized in the SWOT analysis for spray drying of ASD (Figure 9), the process has certain weaknesses such as poor flowability of the resulting powder making downstream processing challenging, and the need for a secondary drying step to remove residual solvent. Particularly relevant for ASD is the stickiness due to low Tg of the product and instability induced due to downstream processes. Furthermore, the most important challenge which lies ahead is the trend of new chemical entities being poorly soluble in both aqueous and organic solvents. It will be interesting to see if solvent free processes such as spray-congealing would be able to provide a solution.

Figure 9: SWOT analysis of a spray drying process to generate ASD

52

The microstructure of ASD is extremely critical for their performance. Since a plethora of formulation and process parameters affect it, it is not trivial to establish which conditions are most useful for generating a homogeneous solid dispersion without doing an extensive set of (trial and error) experiments. The use of experimental design strategy is a huge reprieve. But future research needs to establish a well-defined link between the formulation and process parameters and ASD microstructure along with its performance. One important avenue which needs to be explored further is the role of solution state drug-carrier interactions, factors affecting them and how they are related to the strength of the solid state interactions between components. It is a well-known fact that drug-carrier H-bonding is very important for ASD stability and that liquid state interactions can be frozen into solid state. But there is a knowledge gap regarding the impact of various nozzle types, atomization rate and drying temperature on the interactions in the liquid state and whether it can be translated into differences in solid-state interactions. Some research has been done relating the transition of solution state interactions to solid state during spray drying but further knowledge would aid in tweaking the process to enhance the stabilizing intermolecular interactions. Furthermore, there is an increasing attention to the effect of downstream processing on ASD stability. However, not much is known about the susceptibility of the ASD to the compressive stress and the role of drug polymer interactions. Such information would aid in better informed downstream processing decisions for susceptible systems. Obtaining ASD product which requires minimum downstream processing is an ideal scenario and can be explored further.

Acknowledgements

The authors acknowledge the financial support from FWO Vlaanderen (G.0764.13). Abhishek Singh acknowledges the financial support via an OT grant (OT/12/077) from KU Leuven.

53

14. References

[1] G.L. Amidon, H. Lennernäs, V.P. Shah, J.R. Crison, A theoretical basis for a biopharmaceutic drug classification: the correlation of in vitro drug product dissolution and in vivo bioavailability, Pharm. Res., 12 (1995) 413-420.

[2] C.A. Lipinski, F. Lombardo, B.W. Dominy, P.J. Feeney, Experimental and computational approaches to estimate solubility and permeability in drug discovery and development settings, Adv. Drug Deliv. Rev., 64 (2012) 4-17.

[3] C.A. Lipinski, Drug-like properties and the causes of poor solubility and poor permeability, J. Pharmacol. Toxicol. Methods, 44 (2000) 235-249.

[4] G.M. Keserű, G.M. Makara, Hit discovery and hit-to-lead approaches, Drug Discov. Today, 11 (2006) 741-748.

[5] J.A. Wells, C.L. McClendon, Reaching for high-hanging fruit in drug discovery at protein–protein interfaces, Nature, 450 (2007) 1001-1009.

[6] Y. Kawabata, K. Wada, M. Nakatani, S. Yamada, S. Onoue, Formulation design for poorly water-soluble drugs based on biopharmaceutics classification system: basic approaches and practical applications, Int. J. Pharm., 420 (2011) 1-10.

[7] H.D. Williams, N.L. Trevaskis, S.A. Charman, R.M. Shanker, W.N. Charman, C.W. Pouton, C.J.H. Porter, Strategies to Address Low Drug Solubility in Discovery and Development, Pharmacol. Rev., 65 (2013) 315-499.

[8] A. Dokoumetzidis, P. Macheras, A century of dissolution research: from Noyes and Whitney to the biopharmaceutics classification system, Int. J. Pharm., 321 (2006) 1-11.

[9] B.C. Hancock, M. Parks, What is the true solubility advantage for amorphous pharmaceuticals?, Pharm. Res., 17 (2000) 397-404.

[10] B.C. Hancock, G. Zografi, Characteristics and significance of the amorphous state in pharmaceutical systems, J. Pharm. Sci., 86 (1997) 1-12.

[11] B.T. Surikutchi, S.P. Patil, G. Shete, S. Patel, A.K. Bansal, Drug-excipient behavior in polymeric amorphous solid dispersions, J. Excip. Food Chem., 4 (2013) 70-94.

[12] J. Brouwers, M.E. Brewster, P. Augustijns, Supersaturating drug delivery systems: The answer to solubility limited oral bioavailability?, J. Pharm. Sci., 98 (2009) 2549-2572.‐[13] A. Singh, Z.A. Worku, G. Van den Mooter, Oral formulation strategies to improve solubility of poorly water-soluble drugs, Expert Opin. Drug Deliv., 8 (2011) 1361-1378.

[14] A. Singh, Z. Worku, G. Van den Mooter, Polymeric Delivery Systems for Poorly Soluble Drugs, in: J. Swarbrick (Ed.) Encyclopedia of Pharmaceutical Science and Technology, fourth ed., Taylor and Francis, 2013, pp. 2763-2772.

54

[15] P.J. Marsac, T. Li, L.S. Taylor, Estimation of Drug–Polymer Miscibility and Solubility in Amorphous Solid Dispersions Using Experimentally Determined Interaction Parameters, Pharm. Res., 26 (2009) 139-151.

[16] F. Qian, J. Huang, M.A. Hussain, Drug–polymer solubility and miscibility: stability consideration and practical challenges in amorphous solid dispersion development, J. Pharm. Sci., 99 (2010) 2941-2947.

[17] A.C. Rumondor, I. Ivanisevic, S. Bates, D.E. Alonzo, L.S. Taylor, Evaluation of drug-polymer miscibility in amorphous solid dispersion systems, Pharm. Res., 26 (2009) 2523-2534.

[18] A. Singh, J. Van Humbeeck, G. Van den Mooter, A New Twist in the Old Story-can Compression Induce Mixing of Phase Separated Solid Dispersions? A Case Study of Spray-Dried Miconazole-PVPVA64 Solid Dispersions, Pharm. Res., 31 (2014) 3191-3200.

[19] A. Singh, C. De Bisschop, H. Schut, J. Van Humbeeck, G. Van Den Mooter, Compression Effects on the Phase Behaviour of Miconazole–Poly (1-Vinylpyrrolidone-Co-Vinyl Acetate) Solid Dispersions—Role of Pressure, Dwell Time, and Preparation Method, J. Pharm. Sci., 104 (2015) 3366-3376.

[20] T. Feng, R. Pinal, M.T. Carvajal, Process induced disorder in crystalline materials: differentiating defective crystals from the amorphous form of griseofulvin, J. Pharm. Sci., 97 (2008) 3207-3221.

[21] J. Thiry, F. Krier, B. Evrard, A review of pharmaceutical extrusion: Critical process parameters and scaling-up, Int. J. Pharm., 479 (2015) 227-240.

[22] S. Shah, S. Maddineni, J. Lu, M.A. Repka, Melt extrusion with poorly soluble drugs, Int. J. Pharm., 453 (2013) 233-252.

[23] K. Crowley, A. Gryczke, Hot Melt Extrusion of Amorphous Solid Dispersions, in: Pharmaceutical Sciences Encyclopedia, John Wiley & Sons, Inc., 2015, pp. 1-23.

[24] S. Li, D.S. Jones, G.P. Andrews, Hot Melt Extrusion: A Process Overview and Use in Manufacturing Solid Dispersions of Poorly Water-Soluble Drugs, in: D. Douroumis, A. Fahr (Eds.), Drug Delivery Strategies for Poorly Water-Soluble Drugs, John Wiley & Sons Ltd, Oxford, UK, 2013, pp. 325-358.

[25] I.C. Kemp, Fundamentals of Energy Analysis of Dryers, in: E. Tsotsas, A. S. Mujumdar (Eds.), Modern Drying Technology: Energy Savings, Volume 4, Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim, Germany, 2011, pp.1-45.

[26] A.S. Mujumdar, Handbook of industrial drying, fourth ed., CRC Press, Florida, 2014.

[27] S. R. Percy, Improvement in drying and concentrating liquid substances by atomizing, US Patent US125406, 1872.

[28] D.M. Parikh, Handbook of pharmaceutical granulation technology, third ed., CRC Press, New York, 2009.

55

[29] K. Cal, K. Sollohub, Spray drying technique. I: Hardware and process parameters, J. Pharm. Sci., 99 (2010) 575-586.

[30] C.G.J. Baker, Industrial drying of foods, Springer Science & Business Media, New York, 1997.

[31] F. Gaspar, J. Vicente, F. Neves, J.-R. Authelin, Spray Drying: Scale-Up and Manufacturing, in: N. Shah, H. Sandhu, D.S. Choi, H. Chokshi, A.W. Malick (Eds.), Amorphous Solid Dispersions, Springer, New York, 2014, pp. 261-302.

[32] I. Zbicinski, C. Strumillo, A. Delag, Drying kinetics and particle residence time in spray drying, Drying Technol., 20 (2002) 1751-1768.

[33] D.E. Oakley, Spray Dryer Modeling in Theory and Practice, Drying Technol., 22 (2004) 1371-1402.

[34] Z.A. Worku, J. Aarts, A. Singh, G. Van den Mooter, Drug–Polymer Miscibility across a Spray Dryer: A Case Study of Naproxen and Miconazole Solid Dispersions, Mol. Pharm., 11 (2014) 1094-1101.

[35] Z. Ayenew, A. Paudel, G. Van den Mooter, Can compression induce demixing in amorphous solid dispersions? A case study of naproxen–PVP K25, Eur. J. Pharm. Biopharm., 81 (2012) 207-213.

[36] Z. Ayenew, A. Paudel, P. Rombaut, G. Van den Mooter, Effect of compression on non-isothermal crystallization behaviour of amorphous indomethacin, Pharm. Res. , 29 (2012) 2489-2498.

[37] A.J. Hoekstra, Gas flow field and collection efficiency of cyclone separators (Doctoral dissertation), in: Applied Sciences, TU Delft, Netherlands, 2000, pp. 188.

[38] H.E. Snyder, Pharmaceutical spray drying: solid-dose process technology platform for the 21st century, Ther. Deliv., 3 (2012) 901-912.

[39] K. Masters, Spray drying handbook, fifth ed., Longman Scientific & Technical, New York, 1991.

[40] D.W. Green, Perry's chemical engineers' handbook, eighth ed., McGraw-hill, New York, 2008.

[41] R. Vehring, W.R. Foss, D. Lechuga-Ballesteros, Particle formation in spray drying, J. Aerosol Sci., 38 (2007) 728-746.

[42] M. Farid, A new approach to modelling of single droplet drying, Chem. Eng. Sci., 58 (2003) 2985-2993.

[43] M. Mezhericher, A. Levy, I. Borde, Spray drying modelling based on advanced droplet drying kinetics, Chem. Eng. Process., 49 (2010) 1205-1213.

56

[44] A. Bohr, J. Boetker, T. Rades, J. Rantanen, M. Yang, Application of Spray-drying and Electrospraying/Electrospinning for Poorly Water soluble Drugs: A Particle Engineering Approach, Curr. Pharm. Des., 20 (2014) 325-348.

[45] E.H.J. Kim, X. Dong Chen, D. Pearce, On the mechanisms of surface formation and the surface compositions of industrial milk powders, Drying Technol., 21 (2003) 265-278.

[46] S. Nešić, J. Vodnik, Kinetics of droplet evaporation, Chem. Eng. Sci., 46 (1991) 527-537.

[47] A. Paudel, Z.A. Worku, J. Meeus, S. Guns, G. Van den Mooter, Manufacturing of solid dispersions of poorly water soluble drugs by spray drying: formulation and process considerations, Int. J. Pharm., 453 (2013) 253-284.

[48] M. Mezhericher, A. Levy, I. Borde, Theoretical models of single droplet drying kinetics: a review, Drying Technol., 28 (2010) 278-293.

[49] J.T. Edward, Molecular volumes and the Stokes-Einstein equation, J. Chem. Educ., 47 (1970) 261.

[50] I.C.H.H.T. Guideline, Impurities: Guidelines for residual solvents Q3C (R5), Step 4 (2009).

[51] N.T. Tung, C.W. Park, T.O. Oh, J.Y. Kim, J.M. Ha, Y.S. Rhee, E.S. Park, Formulation of solid dispersion of rebamipide evaluated in a rat model for improved bioavailability and efficacy, J. Pharm. Pharmacol., 63 (2011) 1539-1547.

[52] E.J. Lee, S.W. Lee, H.-G. Choi, C.-K. Kim, Bioavailability of cyclosporin A dispersed in sodium lauryl sulfate–dextrin based solid microspheres, Int. J. Pharm., 218 (2001) 125-131.

[53] J.X. Wu, M. Yang, F. van den Berg, J. Pajander, T. Rades, J. Rantanen, Influence of solvent evaporation rate and formulation factors on solid dispersion physical stability, Eur. J. Pharm. Sci., 44 (2011) 610-620.

[54] S. Janssens, M. Anné, P. Rombaut, G. Van den Mooter, Spray drying from complex solvent systems broadens the applicability of Kollicoat IR as a carrier in the formulation of solid dispersions, Eur. J. Pharm. Sci., 37 (2009) 241-248.

[55] K. Rizi, R.J. Green, M. Donaldson, A.C. Williams, Production of pH responsive microparticles‐ by spray drying: Investigation of experimental parameter effects on morphological and release properties, J. Pharm. Sci., 100 (2011) 566-579.

[56] A. Paudel, G. Van den Mooter, Influence of solvent composition on the miscibility and physical stability of naproxen/PVP K 25 solid dispersions prepared by cosolvent spray-drying, Pharm. Res., 29 (2012) 251-270.

[57] R. Vehring, Pharmaceutical particle engineering via spray drying, Pharm. Res., 25 (2008) 999-1022.

[58] S. Wang, T. Langrish, A review of process simulations and the use of additives in spray drying, Food. Res. Int., 42 (2009) 13-25.

57

[59] H. Al Obaidi, S. Brocchini, G. Buckton, Anomalous properties of spray dried solid ‐dispersions, J. Pharm. Sci., 98 (2009) 4724-4737.

[60] D.T. Friesen, R. Shanker, M. Crew, D.T. Smithey, W. Curatolo, J. Nightingale, Hydroxypropyl methylcellulose acetate succinate-based spray-dried dispersions: an overview, Mol. Pharm., 5 (2008) 1003-1019.

[61] M.C. Dalsin, S. Tale, T.M. Reineke, Solution-State Polymer Assemblies Influence BCS Class II Drug Dissolution and Supersaturation Maintenance, Biomacromolecules, 15 (2014) 500-511.

[62] Nonelectrolytes, in: P.J. Sinko (Ed.) Martinʼs physical pharmacy and pharmaceutical sciences, fifth ed., Lippincott Williams & Wilkins, New Delhi, 2006, pp. 119-141.

[63] M. Bercea, J. Eckelt, B.A. Wolf, Vapor pressures of polymer solutions and the modeling of their composition dependence, Ind. Eng. Chem. Res., 48 (2009) 4603-4606.

[64] M.G. Abiad, O.H. Campanella, M.T. Carvajal, Effect of Spray Drying Conditions on the Physicochemical Properties and Enthalpy Relaxation of α-Lactose, Int. J. Food Prop., 17 (2014) 1303-1316.

[65] E.M. Littringer, A. Mescher, S. Eckhard, H. Schröttner, C. Langes, M. Fries, U. Griesser, P. Walzel, N.A. Urbanetz, Spray drying of mannitol as a drug carrier—the impact of process parameters on product properties, Drying Technol., 30 (2012) 114-124.

[66] O.C. Chidavaenzi, G. Buckton, F. Koosha, R. Pathak, The use of thermal techniques to assess the impact of feed concentration on the amorphous content and polymorphic forms present in spray dried lactose, Int. J. Pharm., 159 (1997) 67-74.

[67] D.L. Tomasko, M.T. Timko, Tailoring of specific interactions to modify the morphology of naproxen, J. Cryst. Growth, 205 (1999) 233-243.

[68] R.M. Martins, S. Siqueira, L.A. Tacon, L.A. Freitas, Microstructured ternary solid dispersions to improve carbamazepine solubility, Powder Technol. , 215 (2012) 156-165.

[69] A. Millqvist-Fureby, M. Malmsten, B. Bergenståhl, Spray-drying of trypsin—surface characterisation and activity preservation, Int. J. Pharm., 188 (1999) 243-253.

[70] S.V. Biradar, A.R. Patil, G.V. Sudarsan, V.B. Pokharkar, A comparative study of approaches used to improve solubility of roxithromycin, Powder Technol., 169 (2006) 22-32.

[71] T. Sunderland, J.G. Kelly, Z. Ramtoola, Application of a novel 3-fluid nozzle spray drying process for the microencapsulation of therapeutic agents using incompatible drug-polymer solutions, Arch. Pharmacal Res., 38 (2013) 566-573.

[72] D.A. Miller, M. Gill, Spray-drying technology, in: R.O. Williams III, A.B. Watts, D.A. Miller (Eds.) Formulating poorly water soluble drugs, Springer Science & Business Media, New York, 2011, pp. 363-442.

58

[73] M. Sajewicz, M. Matlengiewicz, M. Leda, M. Gontarska, D. Kronenbach, T. Kowalska, I.R. Epstein, Spontaneous oscillatory in vitro chiral conversion of simple carboxylic acids and its possible mechanism, J. Phys. Org. Chem., 23 (2010) 1066-1073.

[74] I.T. Ivanov, Z. Tsokeva, Effect of chirality on PVP/drug interaction within binary physical mixtures of ibuprofen, ketoprofen, and naproxen: A DSC study, Chirality, 21 (2009) 719-727.

[75] T. Miyazaki, Y. Aso, S. Yoshioka, T. Kawanishi, Differences in crystallization rate of nitrendipine enantiomers in amorphous solid dispersions with HPMC and HPMCP, Int. J. Pharm., 407 (2011) 111-118.

[76] D. Engers, J. Teng, J. Jimenez Novoa, P. Gent, S. Hossack, C. Campbell, J. Thomson, I. ‐Ivanisevic, A. Templeton, S. Byrn, A solid state approach to enable early development ‐compounds: Selection and animal bioavailability studies of an itraconazole amorphous solid dispersion, J. Pharm. Sci., 99 (2010) 3901-3922.

[77] C. Duret, N. Wauthoz, T. Sebti, F. Vanderbist, K. Amighi, Solid dispersions of itraconazole for inhalation with enhanced dissolution, solubility and dispersion properties, Int. J. Pharm., 428 (2012) 103-113.

[78] S. Guns, A. Dereymaker, P. Kayaert, V. Mathot, J.A. Martens, G. Van den Mooter, Comparison Between Hot-Melt Extrusion and Spray-Drying for Manufacturing Solid Dispersions of the Graft Copolymer of Ethylene Glycol and Vinylalcohol, Pharm. Res., 28 (2011) 673-682.

[79] T.A.G. Langrish, Assessing the rate of solid-phase crystallization for lactose: the effect of the difference between material and glass-transition temperatures, Food. Res. Int., 41 (2008) 630-636.

[80] N.G. Sahoo, A. Abbas, Z. Judeh, C.M. Li, K.H. Yuen, Solubility enhancement of a poorly water soluble anti malarial drug: Experimental design and use of a modified multifluid nozzle ‐ ‐pilot spray drier, J. Pharm. Sci., 98 (2009) 281-296.

[81] K. Ståhl, M. Claesson, P. Lilliehorn, H. Lindén, K. Bäckström, The effect of process variables on the degradation and physical properties of spray dried insulin intended for inhalation, Int. J. Pharm., 233 (2002) 227-237.

[82] B. Adhikari, T. Howes, B. Bhandari, V. Troung, Surface stickiness of drops of carbohydrate and organic acid solutions during convective drying: experiments and modeling, Drying Technol., 21 (2003) 839-873.

[83] E.M. Littringer, M.F. Noisternig, A. Mescher, H. Schroettner, P. Walzel, U.J. Griesser, N.A. Urbanetz, The morphology and various densities of spray dried mannitol, Powder Technol., 246 (2013) 193-200.

[84] D. Chiou, T. Langrish, R. Braham, The effect of temperature on the crystallinity of lactose powders produced by spray drying, J. Food Eng., 86 (2008) 288-293.

[85] Y. Matsuda, M. Otsuka, M. Onoe, E. Tatsumi, Amorphism and Physicochemical Stability of Spray dried Frusemide, J. Pharm. Pharmacol., 44 (1992) 627-633.‐

59

[86] Y. Ueno, E. Yonemochi, Y. Tozuka, S. Yamamura, T. Oguchi, K. Yamamoto, Pharmaceutics: Characterization of Amorphous Ursodeoxycholic Acid Prepared by Spray drying, J. Pharm. ‐Pharmacol., 50 (1998) 1213-1219.

[87] M. Ohta, G. Buckton, A study of the differences between two amorphous spray-dried samples of cefditoren pivoxil which exhibited different physical stabilities, Int. J. Pharm., 289 (2005) 31-38.

[88] A. Paudel, Y. Loyson, G. Van den Mooter, An investigation into the effect of spray drying temperature and atomizing conditions on miscibility, physical stability, and performance of naproxen–PVP K 25 solid dispersions, J. Pharm. Sci., 102 (2013) 1249-1267.

[89] D.E. Dobry, D.M. Settell, J.M. Baumann, R.J. Ray, L.J. Graham, R.A. Beyerinck, A model-based methodology for spray-drying process development, J. Pharm. Innov., 4 (2009) 133-142.

[90] J.E. Patterson, M.B. James, A.H. Forster, T. Rades, Melt extrusion and spray drying of carbamazepine and dipyridamole with polyvinylpyrrolidone/vinyl acetate copolymers, Drug Dev. Ind. Pharm., 34 (2008) 95-106.

[91] M. Ameri, Y.F. Maa, Spray drying of biopharmaceuticals: stability and process considerations, Drying Technol., 24 (2006) 763-768.

[92] K. Sollohub, K. Cal, Spray drying technique: II. Current applications in pharmaceutical technology, J. Pharm. Sci., 99 (2010) 587-597.

[93] S.G. Maas, G. Schaldach, E.M. Littringer, A. Mescher, U.J. Griesser, D.E. Braun, P.E. Walzel, N.A. Urbanetz, The impact of spray drying outlet temperature on the particle morphology of mannitol, Powder Technol. , 213 (2011) 27-35.

[94] M. Islam, T. Langrish, An investigation into lactose crystallization under high temperature conditions during spray drying, Food. Res. Int., 43 (2010) 46-56.

[95] D.E. Dobry, D.M. Settell, J.M. Baumann, Spray Drying and Scale-Up, in: Pharmaceutical Sciences Encyclopedia, John Wiley & Sons, Inc., 2015, pp.1-26.

[96] R. Aftel, A.K. Gupta, C. Cook, C. Presser, Gas property effects on droplet atomization and combustion in an air-assist atomizer, in: Symposium (International) on Combustion, Elsevier, 1996, pp. 1645-1651.

[97] S. Özbilen, Influence of atomising gas on particle characteristics of Al, Al–1 wt-% Li, Mg, and Sn powders, Powder metallurgy, 43 (2000) 173-180.

[98] M.I.U. Islam, T.A.G. Langrish, The effect of different atomizing gases and drying media on the crystallization behavior of spray-dried powders, Drying Technol., 28 (2010) 1035-1043.

[99] T. Kudra, M. Poirier, Gaseous carbon dioxide as the heat and mass transfer medium in drying, Drying Technol., 25 (2007) 327-334.

60

[100] S. Mandato, E. Rondet, G. Delaplace, A. Barkouti, L. Galet, P. Accart, T. Ruiz, B. Cuq, Liquids' atomization with two different nozzles: Modeling of the effects of some processing and formulation conditions by dimensional analysis, Powder Technol. , 224 (2012) 323-330.

[101] K. Kondo, T. Niwa, K. Danjo, Preparation of sustained-release coated particles by novel microencapsulation method using three-fluid nozzle spray drying technique, Eur. J. Pharm. Sci., 51 (2014) 11-19.

[102] T. Mizoe, T. Ozeki, H. Okada, Preparation of drug nanoparticle-containing microparticles using a 4-fluid nozzle spray drier for oral, pulmonary, and injection dosage forms, J. Control. Release, 122 (2007) 10-15.

[103] T. Mizoe, S. Beppu, T. Ozeki, H. Okada, One-step preparation of drug-containing microparticles to enhance the dissolution and absorption of poorly water-soluble drugs using a 4-fluid nozzle spray drier, J. Control. Release, 120 (2007) 205-210.

[104] R. Chen, M. Tagawa, N. Hoshi, T. Ogura, H. Okamoto, K. Danjo, Improved dissolution of an insoluble drug using a 4-fluid nozzle spray-drying technique, Chem. Pharm. Bull., 52 (2004) 1066-1070.

[105] R. Chen, H. Okamoto, K. Danjo, Particle design of indomethacin using a four-fluid-nozzle spray-drying technique, J. Drug Deliv. Sci. Technol., 17 (2007) 129-135.

[106] T. Ozeki, S. Beppu, T. Mizoe, Y. Takashima, H. Yuasa, H. Okada, Preparation of two-drug composite microparticles to improve the dissolution of insoluble drug in water for use with a 4-fluid nozzle spray drier, J. Control. Release, 107 (2005) 387-394.

[107] F. Wan, M.J. Maltesen, S.K. Andersen, S. Bjerregaard, C. Foged, J. Rantanen, M. Yang, One-step production of protein-loaded PLGA microparticles via spray drying using 3-fluid nozzle, Pharm. Res., 31 (2014) 1967-1977.

[108] C.L.N. Vo, C. Park, B.J. Lee, Current trends and future perspectives of solid dispersions containing poorly water-soluble drugs, Eur. J. Pharm. Biopharm., 85 (2013) 799-813.

[109] D.E. Alonzo, S. Raina, D. Zhou, Y. Gao, G.G. Zhang, L.S. Taylor, Characterizing the impact of hydroxypropylmethyl cellulose on the growth and nucleation kinetics of felodipine from supersaturated solutions, Cryst. Growth Des., 12 (2012) 1538-1547.

[110] S. Xu, W.G. Dai, Drug precipitation inhibitors in supersaturable formulations, Int. J. Pharm., 453 (2013) 36-43.

[111] C. Leuner, J. Dressman, Improving drug solubility for oral delivery using solid dispersions, Eur. J. Pharm. Biopharm., 50 (2000) 47-60.

[112] J.L. Ford, A.F. Stewart, J.L. Dubois, The properties of solid dispersions of indomethacin or phenylbutazone in polyethylene glycol, Int. J. Pharm., 28 (1986) 11-22.

[113] J.C. Shah, J.R. Chen, D. Chow, Preformulation study of etoposide: II. Increased solubility and dissolution rate by solid-solid dispersions, Int. J. Pharm., 113 (1995) 103-111.

61

[114] W.L. Chiou, S. Riegelman, Preparation and dissolution characteristics of several fast‐release solid dispersions of griseofulvin, J. Pharm. Sci., 58 (1969) 1505-1510.

[115] G. Betageri, K. Makarla, Enhancement of dissolution of glyburide by solid dispersion and lyophilization techniques, Int. J. Pharm., 126 (1995) 155-160.

[116] A. Kassem, S. Zaki, N. Mursi, S. Tayel, Chloramphenicol solid dispersion system I, Pharm. Ind., 41 (1979) 390-393.

[117] P. Mura, M. Faucci, A. Manderioli, G. Bramanti, P. Parrini, Thermal behavior and dissolution properties of naproxen from binary and ternary solid dispersions, Drug Dev. Ind. Pharm., 25 (1999) 257-264.

[118] S. Verheyen, N. Blaton, R. Kinget, G. Van den Mooter, Mechanism of increased dissolution of diazepam and temazepam from polyethylene glycol 6000 solid dispersions, Int. J. Pharm., 249 (2002) 45-58.

[119] I. Weuts, D. Kempen, G. Verreck, A. Decorte, K. Heymans, J. Peeters, M. Brewster, G. Van den Mooter, Study of the physicochemical properties and stability of solid dispersions of loperamide and PEG6000 prepared by spray drying, Eur. J. Pharm. Biopharm., 59 (2005) 119-126.

[120] Y. Matsui, T. Kubota, H. Tadokoro, T. Yoshihara, Raman spectra of polyethers, J. Polym. Sci. A, 3 (1965) 2275-2288.

[121] D.Q.M. Craig, A review of thermal methods used for the analysis of the crystal form, solution thermodynamics and glass transition behaviour of polyethylene glycols, Thermochim. acta, 248 (1995) 189-203.

[122] B. Read, Mechanical relaxation in some oxide polymers, Polymer, 3 (1962) 529-542.

[123] J. Koenig, A. Angood, Raman spectra of poly (ethylene glycols) in solution, J. Polym. Sci. A-2 Polym. Phys., 8 (1970) 1787-1796.

[124] C. Buckley, A. Kovacs, Melting behaviour of low molecular weight poly (ethylene-oxide) fractions, Prog. Colloid Polym. Sci., 58 (1975), pp. 44-52.

[125] C. Buckley, A. Kovacs, Melting behaviour of low molecular weight poly (ethylene-oxide) fractions, Colloid Polym. Sci., 254 (1976) 695-715.

[126] O.C. Chidavaenzi, G. Buckton, F. Koosha, The effect of co-spray drying with polyethylene glycol 4000 on the crystallinity and physical form of lactose, Int. J. Pharm., 216 (2001) 43-49.

[127] K. Mosén, K. Bäckström, K. Thalberg, T. Schaefer, A. Axelsson, H.G. Kristensen, The apparent plasticizing effect of polyethylene glycol (PEG) on the crystallinity of spray dried lactose/PEG composites, Eur. J. Pharm. Biopharm., 64 (2006) 206-211.

[128] Q. Zhu, M.T. Harris, L.S. Taylor, Modification of crystallization behavior in drug/polyethylene glycol solid dispersions, Mol. Pharm., 9 (2012) 546-553.

62

[129] T.V. Duong, J. Van Humbeeck, G. Van den Mooter, Crystallization Kinetics of Indomethacin/Polyethylene Glycol Dispersions Containing High Drug Loadings, Mol. Pharm., 12 (2015) 2493-2504.

[130] D.O. Corrigan, A.M. Healy, O.I. Corrigan, The effect of spray drying solutions of polyethylene glycol (PEG) and lactose/PEG on their physicochemical properties, Int. J. Pharm., 235 (2002) 193-205.

[131] S. Kamel, N. Ali, K. Jahangir, S. Shah, A. El-Gendy, Pharmaceutical significance of cellulose: a review, Express Polym. Lett., 2 (2008) 758-778.

[132] F.D. Marques-Marinho, C.D. Vianna-Soares, Cellulose and Its Derivatives Use in the Pharmaceutical Compounding Practice, (2013), DOI: 10.5772/56637.

[133] K. Adibkia, M. Barzegar-Jalali, H. Maheri-Esfanjani, S. Ghanbarzadeh, J. Shokri, A. Sabzevari, Y. Javadzadeh, Physicochemical characterization of naproxen solid dispersions prepared via spray drying technology, Powder Technol., 246 (2013) 448-455.

[134] I. Ghosh, J. Snyder, R. Vippagunta, M. Alvine, R. Vakil, W.Q.T. Tong, S. Vippagunta, Comparison of HPMC based polymers performance as carriers for manufacture of solid dispersions using the melt extruder, Int. J. Pharm., 419 (2011) 12-19.

[135] F. Tajarobi, A. Larsson, H. Matic, S. Abrahmsén-Alami, The influence of crystallization inhibition of HPMC and HPMCAS on model substance dissolution and release in swellable matrix tablets, Eur. J. Pharm. Biopharm., 78 (2011) 125-133.

[136] L.H. Nielsen, T. Rades, A. Müllertz, Characterization during storage and dissolution of Solid dispersions containing furosemide and hydroxypropyl methylcellulose, J. Drug Deliv. Sci. Technol., 23 (2013) 409-415.

[137] K. Ueda, K. Higashi, K. Yamamoto, K. Moribe, Inhibitory effect of hydroxypropyl methylcellulose acetate succinate on drug recrystallization from a supersaturated solution assessed using nuclear magnetic resonance measurements, Mol. Pharm., 10 (2013) 3801-3811.

[138] M. Fukasawa, S. Obara, Molecular weight determination of hypromellose acetate succinate (HPMCAS) using size exclusion chromatography with a multi-angle laser light scattering detector, Chem. Pharm. Bull., 52 (2004) 1391-1393.

[139] K. Ueda, K. Higashi, K. Yamamoto, K. Moribe, The effect of HPMCAS functional groups on drug crystallization from the supersaturated state and dissolution improvement, Int. J. Pharm., 464 (2014) 205-213.

[140] A. Patterson, A.P. Ferreira, E. Banks, K. Skeene, G. Clarke, S. Nicholson, C. Rawlinson-Malone, Modelling drug degradation in a spray dried polymer dispersion using a modified Arrhenius equation, Int. J. Pharm., 478 (2015) 348-360.

[141] K. Ueda, K. Higashi, W. Limwikrant, S. Sekine, T. Horie, K. Yamamoto, K. Moribe, Mechanistic differences in permeation behavior of supersaturated and solubilized solutions of

63

carbamazepine revealed by nuclear magnetic resonance measurements, Mol. Pharm., 9 (2012) 3023-3033.

[142] J.M. Pereira, R. Mejia-Ariza, G.A. Ilevbare, H.E. McGettigan, N. Sriranganathan, L.S. Taylor, R.M. Davis, K.J. Edgar, Interplay of degradation, dissolution and stabilization of clarithromycin and its amorphous solid dispersions, Mol. Pharm., 10 (2013) 4640-4653.

[143] B. Li, S. Konecke, L.A. Wegiel, L.S. Taylor, K.J. Edgar, Both solubility and chemical stability of curcumin are enhanced by solid dispersion in cellulose derivative matrices, Carbohydr. Polym., 98 (2013) 1108-1116.

[144] B. Li, S. Konecke, K. Harich, L. Wegiel, L.S. Taylor, K.J. Edgar, Solid dispersion of quercetin in cellulose derivative matrices influences both solubility and stability, Carbohydr. Polym., 92 (2013) 2033-2040.

[145] J. Zhu, Z. Yang, X. Chen, J. Sun, G. Awuti, X. Zhang, Q. Zhang, Preparation and physicochemical characterization of solid dispersion of quercetin and polyvinylpyrrolidone, J. Chinese Pharm. Sci., 16 (2007) 51.

[146] S. Onoue, H. Takahashi, Y. Kawabata, Y. Seto, J. Hatanaka, B. Timmermann, S. Yamada, Formulation design and photochemical studies on nanocrystal solid dispersion of curcumin with improved oral bioavailability, J. Pharm. Sci., 99 (2010) 1871-1881.

[147] V. Bühler, Polyvinylpyrrolidone excipients for pharmaceuticals: Povidone, crospovidone and copovidone, Springer, Heidelberg, Germany, 2005.

[148] I. Weuts, D. Kempen, A. Decorte, G. Verreck, J. Peeters, M. Brewster, G. Van den Mooter, Physical stability of the amorphous state of loperamide and two fragment molecules in solid dispersions with the polymers PVP-K30 and PVP-VA64, Eur. J. Pharm. Sci., 25 (2005) 313-320.

[149] A.C. Rumondor, L.S. Taylor, Effect of polymer hygroscopicity on the phase behavior of amorphous solid dispersions in the presence of moisture, Mol. Pharm., 7 (2010) 477-490.

[150] G. Van den Mooter, P. Augustijns, N. Blaton, R. Kinget, Physico-chemical characterization of solid dispersions of temazepam with polyethylene glycol 6000 and PVP K30, Int. J. Pharm., 164 (1998) 67-80.

[151] A. Paradkar, A.A. Ambike, B.K. Jadhav, K. Mahadik, Characterization of curcumin–PVP solid dispersion obtained by spray drying, Int. J. Pharm., 271 (2004) 281-286.

[152] F. Frizon, J. de Oliveira Eloy, C.M. Donaduzzi, M.L. Mitsui, J.M. Marchetti, Dissolution rate enhancement of loratadine in polyvinylpyrrolidone K-30 solid dispersions by solvent methods, Powder Technol., 235 (2013) 532-539.

[153] P.S. Yadav, V. Kumar, U.P. Singh, H.R. Bhat, B. Mazumder, Physicochemical characterization and in vitro dissolution studies of solid dispersions of ketoprofen with PVP K30 and D-mannitol, Saudi Pharm. J., 21 (2013) 77-84.

64

[154] A. Homayouni, F. Sadeghi, J. Varshosaz, H.A. Garekani, A. Nokhodchi, Comparing various techniques to produce micro/nanoparticles for enhancing the dissolution of celecoxib containing PVP, Eur. J. Pharm. Biopharm., 88 (2014) 261-274.

[155] P. Thybo, B.L. Pedersen, L. Hovgaard, R. Holm, A. Müllertz, Characterization and physical stability of spray dried solid dispersions of probucol and PVP-K30, Pharm. Dev. Technol., 13 (2008) 375-386.

[156] S. Sethia, E. Squillante, Solid dispersion of carbamazepine in PVP K30 by conventional solvent evaporation and supercritical methods, Int. J. Pharm., 272 (2004) 1-10.

[157] E.A. Fouad, M. EL-Badry, G.M. Mahrous, F.K. Alanazi, S.H. Neau, I.A. Alsarra, The use of spray-drying to enhance celecoxib solubility, Drug Dev. Ind. Pharm., 37 (2011) 1463-1472.

[158] J.Y. Jung, S.D. Yoo, S.H. Lee, K.H. Kim, D.-S. Yoon, K.H. Lee, Enhanced solubility and dissolution rate of itraconazole by a solid dispersion technique, Int. J. Pharm., 187 (1999) 209-218.

[159] I. Weuts, D. Kempen, A. Decorte, G. Verreck, J. Peeters, M. Brewster, G. Van den Mooter, Phase behaviour analysis of solid dispersions of loperamide and two structurally related compounds with the polymers PVP-K30 and PVP-VA64, Eur. J. Pharm. Sci., 22 (2004) 375-385.

[160] C. Dahlberg, A. Millqvist-Fureby, M. Schuleit, Surface composition and contact angle relationships for differently prepared solid dispersions, Eur. J. Pharm. Biopharm., 70 (2008) 478-485.

[161] L. Zhu, L. Wong, L. Yu, Surface-enhanced crystallization of amorphous nifedipine, Mol. Pharm., 5 (2008) 921-926.

[162] N. Rodríguez hornedo, D. Murphy, Significance of controlling crystallization mechanisms ‐and kinetics in pharmaceutical systems, J. Pharm. Sci., 88 (1999) 651-660.

[163] J.A. Marks, L.A. Wegiel, L.S. Taylor, K.J. Edgar, Pairwise Polymer Blends for Oral Drug Delivery, J. Pharm. Sci., 103 (2014) 2871-2883.

[164] Y.D. Yan, J.H. Sung, K.K. Kim, D.W. Kim, J.O. Kim, B.J. Lee, C.S. Yong, H.-G. Choi, Novel valsartan-loaded solid dispersion with enhanced bioavailability and no crystalline changes, Int. J. Pharm., 422 (2012) 202-210.

[165] M.S. Kim, J.S. Kim, W. Cho, K.H. Cha, H.J. Park, J. Park, S.J. Hwang, Supersaturatable formulations for the enhanced oral absorption of sirolimus, Int. J. Pharm., 445 (2013) 108-116.

[166] C. Goddeeris, T. Willems, G. Van den Mooter, Formulation of fast disintegrating tablets of ternary solid dispersions consisting of TPGS 1000 and HPMC 2910 or PVPVA 64 to improve the dissolution of the anti-HIV drug UC 781, Eur. J. Pharm. Sci., 34 (2008) 293-302.

[167] S. Janssens, J. Van Humbeeck, G. Van den Mooter, Evaluation of the formulation of solid dispersions by co-spray drying itraconazole with Inutec SP1, a polymeric surfactant, in combination with PVPVA 64, Eur. J. Pharm. Biopharm., 70 (2008) 500-505.

65

[168] S. Janssens, S. Nagels, H.N. De Armas, W. D’autry, A. Van Schepdael, G. Van den Mooter, Formulation and characterization of ternary solid dispersions made up of Itraconazole and two excipients, TPGS 1000 and PVPVA 64, that were selected based on a supersaturation screening study, Eur. J. Pharm. Biopharm., 69 (2008) 158-166.

[169] X. Wang, A. Michoel, G. Van den Mooter, Solid state characteristics of ternary solid dispersions composed of PVP VA64, Myrj 52 and itraconazole, Int. J. Pharm., 303 (2005) 54-61.

[170] N. Marasini, T.H. Tran, B.K. Poudel, H.J. Cho, Y.K. Choi, S.C. Chi, H.G. Choi, C.S. Yong, J.O. Kim, Fabrication and evaluation of pH-modulated solid dispersion for telmisartan by spray-drying technique, Int. J. Pharm., 441 (2013) 424-432.

[171] J.R. Patel, R.A. Carlton, F. Yuniatine, T.E. Needham, L. Wu, F.G. Vogt, Preparation and structural characterization of amorphous spray dried dispersions of tenoxicam with enhanced ‐dissolution, J. Pharm. Sci., 101 (2012) 641-663.

[172] P. Gupta, A.K. Bansal, Spray drying for generation of a ternary amorphous system of celecoxib, PVP, and meglumine, Pharm. Dev. Technol., 10 (2005) 273-281.

[173] D.X. Li, K.Y. Jang, W. Kang, K. Bae, M.H. Lee, Y.K. Oh, J.P. Jee, Y.J. Park, D.H. Oh, Y.G. Seo, Enhanced solubility and bioavailability of sibutramine base by solid dispersion system with aqueous medium, Biol. Pharm. Bull., 33 (2010) 279-284.

[174] Y.J. Park, D.S. Ryu, D.X. Li, Q.Z. Quan, D.H. Oh, J.O. Kim, Y.G. Seo, Y.I. Lee, C.S. Yong, J.S. Woo, Physicochemical characterization of tacrolimus-loaded solid dispersion with sodium carboxylmethyl cellulose and sodium lauryl sulfate, Arch. Pharmacal Res., 32 (2009) 893-898.

[175] A.A. Ambike, K. Mahadik, A. Paradkar, Spray-dried amorphous solid dispersions of simvastatin, a low Tg drug: in vitro and in vivo evaluations, Pharm. Res., 22 (2005) 990-998.

[176] V.B. Pokharkar, L.P. Mandpe, M.N. Padamwar, A.A. Ambike, K.R. Mahadik, A. Paradkar, Development, characterization and stabilization of amorphous form of a low Tg drug, Powder Technol. , 167 (2006) 20-25.

[177] L. Wang, F.D. Cui, H. Sunada, Improvement of the dissolution rate of nitrendipine using a new pulse combustion drying method, Chem. Pharm. Bull., 55 (2007) 1119-1125.

[178] R. Araújo, C. Teixeira, L. Freitas, The preparation of ternary solid dispersions of an herbal drug via spray drying of liquid feed, Drying Technol., 28 (2010) 412-421.

[179] Y.I. Kim, K.S. Kim, K.H. Suh, S. Shanmugam, J.S. Woo, C.S. Yong, H.G. Choi, New clopidogrel napadisilate salt and its solid dispersion with improved stability and bioequivalence to the commercial clopidogrel bisulphate salt in beagle dogs, Int. J. Pharm., 415 (2011) 129-139.

[180] H. Al-Obaidi, P. Ke, S. Brocchini, G. Buckton, Characterization and stability of ternary solid dispersions with PVP and PHPMA, Int. J. Pharm., 419 (2011) 20-27.

[181] D. Prasad, H. Chauhan, E. Atef, Amorphous stabilization and dissolution enhancement of amorphous ternary solid dispersions: combination of polymers showing drug–polymer interaction for synergistic effects, J. Pharm. Sci., 103 (2014) 3511-3523.

66

[182] G.Z. Papageorgiou, D. Bikiaris, F.I. Kanaze, E. Karavas, A. Stergiou, E. Georgarakis, Tailoring the release rates of fluconazole using solid dispersions in polymer blends, Drug Dev. Ind. Pharm., 34 (2008) 336-346.

[183] S.K. Paidi, S.K. Jena, B.K. Ahuja, N. Devasari, S. Suresh, Preparation, in vitro and in vivo ‐ ‐evaluation of spray dried ternary solid dispersion of biopharmaceutics classification system ‐class II model drug, J. Pharm. Pharmacol., 67 (2015) 616-629.

[184] H. Al-Obaidi, G. Buckton, Evaluation of griseofulvin binary and ternary solid dispersions with HPMCAS, AAPS PharmSciTech, 10 (2009) 1172-1177.

[185] J.W. Lubach, D. Xu, B.E. Segmuller, E.J. Munson, Investigation of the effects of pharmaceutical processing upon solid-state NMR relaxation times and implications to solid-state formulation stability, J. Pharm. Sci., 96 (2007) 777–787.

[186] M. Sugimoto, T. Okagaki, S. Narisawa, Y. Koida, K. Nakajima, Improvement of dissolution characteristics and bioavailability of poorly water-soluble drugs by novel cogrinding method using water-soluble polymer, Int. J. Pharm., 160 (1998) 11-19.

[187] V. Caron, L. Tajber, O.I. Corrigan, A.M. Healy, A comparison of spray drying and milling in the production of amorphous dispersions of sulfathiazole/polyvinylpyrrolidone and sulfadimidine/polyvinylpyrrolidone, Mol. Pharm., 8 (2011) 532-542.

[188] R. Dontireddy, A.M. Crean, A comparative study of spray-dried and freeze-dried hydrocortisone/polyvinyl pyrrolidone solid dispersions, Drug Dev. Ind. Pharm., 37 (2011) 1141-1149.

[189] D.H. Won, M.S. Kim, S. Lee, J.S. Park, S.J. Hwang, Improved physicochemical characteristics of felodipine solid dispersion particles by supercritical anti-solvent precipitation process, Int. J. Pharm., 301 (2005) 199-208.

[190] O. Mahmah, R. Tabbakh, A. Kelly, A. Paradkar, A comparative study of the effect of spray drying and hot melt extrusion on the properties of amorphous solid dispersions containing ‐felodipine, J. Pharm. Pharmacol., 66 (2014) 275-284.

[191] J.E. Patterson, M.B. James, A.H. Forster, R.W. Lancaster, J.M. Butler, T. Rades, Preparation of glass solutions of three poorly water soluble drugs by spray drying, melt extrusion and ball milling, Int. J. Pharm., 336 (2007) 22-34.

[192] A.M. Agrawal, M.S. Dudhedia, A.D. Patel, M.S. Raikes, Characterization and performance assessment of solid dispersions prepared by hot melt extrusion and spray drying process, Int. J. Pharm., 457 (2013) 71-81.

[193] C. Dahlberg, A. Millqvist-Fureby, M. Schuleit, I. Furó, Relationships between solid dispersion preparation process, particle size and drug release–an NMR and NMR microimaging study, Eur. J. Pharm. Biopharm., 76 (2010) 311-319.

[194] A. Jaworek, Micro-and nanoparticle production by electrospraying, Powder Technol., 176 (2007) 18-35.

67

[195] L. Rayleigh. On the equilibrium of liquid conducting masses charged with electricity, Philos. Mag., 14 (1882) 184-186.

[196] M. Cloupeau, B. Prunet-Foch, Electrostatic spraying of liquids: main functioning modes, J. Electrostat., 25 (1990) 165-184.

[197] A. Jaworek, A. Sobczyk, Electrospraying route to nanotechnology: an overview, J. Electrostat., 66 (2008) 197-219.

[198] R. Bocanegra, D. Galán, M. Márquez, I. Loscertales, A. Barrero, Multiple electrosprays emitted from an array of holes, J. Aerosol Sci., 36 (2005) 1387-1399.

[199] J. Almekinders, C. Jones, Multiple jet electrohydrodynamic spraying and applications, J. Aerosol Sci., 30 (1999) 969-971.

[200] M. Mutoh, S. Kaieda, K. Kamimura, Convergence and disintegration of liquid jets induced by an electrostatic field, J. Appl. Phys., 50 (1979) 3174-3179.

[201] D.P.H. Smith, The electrohydrodynamic atomization of liquids, IEEE Trans. Ind. Appl., (1986) 527-535.

[202] V.G. Drozin, The electrical dispersion of liquids as aerosols, J. Colloid Sci., 10 (1955) 158-164.

[203] Z.P. Liu, L. Cui, D.G. Yu, Z.X. Zhao, L. Chen, Electrosprayed core–shell solid dispersions of acyclovir fabricated using an epoxy-coated concentric spray head, Int. J. Nanomedicine, 9 (2014) 1967.

[204] J. Roine, M. Kaasalainen, M. Peurla, A. Correia, F. Araújo, H.A. Santos, M. Murtomaa, J. Salonen, Controlled dissolution of griseofulvin solid dispersions from electrosprayed enteric polymer micromatrix particles: physicochemical characterization and in vitro evaluation, Mol. Pharm., 12 (2015) 2254-2264.

[205] S. Wong, I. Kellaway, S. Murdan, Enhancement of the dissolution rate and oral absorption of a poorly water soluble drug by formation of surfactant-containing microparticles, Int. J. Pharm., 317 (2006) 61-68.

[206] D.N. Nguyen, L. Palangetic, C. Clasen, G. Van den Mooter, One step production of ‐darunavir solid dispersion nanoparticles coated with enteric polymers using electrospraying, J. Pharm. Pharmacol., (2015), DOI: 10.1111/jphp.12459.

[207] D.G. Yu, G.R. Williams, X. Wang, X.K. Liu, H.L. Li, S.A. Bligh, Dual drug release nanocomposites prepared using a combination of electrospraying and electrospinning, RSC Advances, 3 (2013) 4652-4658.

[208] C. Baker, K. McKenzie, Energy consumption of industrial spray dryers, Drying Technol., 23 (2005) 365-386.

68

[209] S. Khalloufi, C. Almeida-Rivera, A. Mudaliar, Modern Drying Technology, Volume 5, Process Intensification: E. Tsotsas and A.S. Mujumdar (Eds.), Wiley-VCH Verlag GmbH & Co. KgaA: Weinheim, Germany, 2014, 359 pp, Drying Technol., 32 (2014) 2017-2020.

[210] I. Zbicinski, I. Smucerowicz, C. Strumillo, C. Crowe, Application of pulse combustion technology in spray drying process, Braz. J. Chem. Eng., 17 (2000) 441-450.

[211] D.N. Bikiaris, Solid dispersions, part I: recent evolutions and future opportunities in manufacturing methods for dissolution rate enhancement of poorly water-soluble drugs, Expert Opin. Drug Deliv., 8 (2011) 1501-1519.

[212] T. Kudra, Pulse-combustion drying: Status and potentials, Drying Technol., 26 (2008) 1409-1420.

[213] D.J. Lee, S. Jangam, A.S. Mujumdar, Some recent advances in drying technologies to produce particulate solids, KONA Powder Particle J., 30 (2013) 69-83.

[214] L. Xu, S.M. Li, H. Sunada, Preparation and evaluation of ibuprofen solid dispersion systems with kollidon particles using a pulse combustion dryer system, Chem. Pharm. Bull., 55 (2007) 1545-1550.

[215] L. Xu, S. Li, Y. Wang, M. Wei, H. Yao, H. Sunada, Improvement of dissolution rate of ibuprofen by solid dispersion systems with Kollicoat IR using a pulse combustion dryer system, J. Drug Deliv. Sci. Technol., 19 (2009) 113-118.

[216] P. Thybo, L. Hovgaard, J.S. Lindeløv, A. Brask, S.K. Andersen, Scaling up the spray drying process from pilot to production scale using an atomized droplet size criterion, Pharm. Res., 25 (2008) 1610-1620.

[217] U.S. Department of Health and Human Services, Guidance for industry: Pharmaceutical Development Q8 (R2), Revision 2 (2009).

[218] A.S. Rathore, H. Winkle, Quality by design for biopharmaceuticals, Nature Biotechnol., 27 (2009) 26-34.

[219] A. Billon, B. Bataille, G. Cassanas, M. Jacob, Development of spray-dried acetaminophen microparticles using experimental designs, Int. J. Pharm., 203 (2000) 159-168.

[220] V. Nekkanti, T. Muniyappan, P. Karatgi, M.S. Hari, S. Marella, R. Pillai, Spray-drying process optimization for manufacture of drug–cyclodextrin complex powder using design of experiments, Drug Dev. Ind. Pharm., 35 (2009) 1219-1229.

[221] T. Mihajlovic, S. Ibric, A. Mladenovic, Application of design of experiments and multilayer perceptron neural network in optimization of the spray-drying process, Drying Technol., 29 (2011) 1638-1647.

[222] M.J. Maltesen, S. Bjerregaard, L. Hovgaard, S. Havelund, M. Van De Weert, Quality by design–Spray drying of insulin intended for inhalation, Eur. J. Pharm. Biopharm., 70 (2008) 828-838.

69

[223] M.J. Maltesen, M. Van de Weert, H. Grohganz, Design of experiments-based monitoring of critical quality attributes for the spray-drying process of insulin by NIR spectroscopy, AAPS PharmSciTech, 13 (2012) 747-755.

[224] P. Tewa-Tagne, G. Degobert, S. Briançon, C. Bordes, J.Y. Gauvrit, P. Lanteri, H. Fessi, Spray-drying nanocapsules in presence of colloidal silica as drying auxiliary agent: formulation and process variables optimization using experimental designs, Pharm. Res., 24 (2007) 650-661.

[225] J.D. Ormes, D. Zhang, A.M. Chen, S. Hou, D. Krueger, T. Nelson, A. Templeton, Design of experiments utilization to map the processing capabilities of a micro-spray dryer: particle design and throughput optimization in support of drug discovery, Pharm. Dev. Technol., 18 (2013) 121-129.

[226] A. Baldinger, L. Clerdent, J. Rantanen, M. Yang, H. Grohganz, Quality by design approach in the optimization of the spray-drying process, Pharm. Dev. Technol., 17 (2012) 389-397.

[227] I.C.H.H.T. Guideline, Pharmaceutical Development Q8 (R2), Step, 4 (2009).

[228] U.S. Department of Health and Human Services, Guidance for industry: PAT—a framework for innovative pharmaceutical development, manufacturing, and quality assurance (2004).

[229] M. Jivraj, L.G. Martini, C.M. Thomson, An overview of the different excipients useful for the direct compression of tablets, Pharm. Sci. Technol. Today, 3 (2000) 58-63.

[230] Z. Sun, N. Ya, R.C. Adams, F.S. Fang, Particle Size Specifications for Solid Oral Dosage Forms: A Regulatory Perspective, Am. Pharm. Review, 13 (2010) 68.

[231] S. Page, R. Maurer, Downstream Processing Considerations, in: N. Shah, H. Sandhu, D.S. Choi, H. Chokshi, A.W. Malick (Eds.), Amorphous Solid Dispersions, Springer, New York, 2014, pp. 395-417.

[232] Z. Német, J. Sztatisz, Á. Demeter, Polymorph transitions of bicalutamide: a remarkable example of mechanical activation, J. Pharm. Sci., 97 (2008) 3222-3232.

[233] M.M. Leane, W. Sinclair, F. Qian, R. Haddadin, A. Brown, M. Tobyn, A.B. Dennis, Formulation and process design for a solid dosage form containing a spray-dried amorphous dispersion of ibipinabant, Pharm. Dev. Technol., 18 (2013) 359-366.

[234] R.M. Iyer, S. Hegde, J. DiNunzio, D. Singhal, W. Malick, The impact of roller compaction and tablet compression on physicomechanical properties of pharmaceutical excipients, Pharm. Dev. Technol., 19 (2014) 583-592.

[235] S.R.K. Vaka, M.M. Bommana, D. Desai, J. Djordjevic, W. Phuapradit, N. Shah, Excipients for Amorphous Solid Dispersions, in: N. Shah, H. Sandhu, D.S. Choi, H. Chokshi, A.W. Malick (Eds.), Amorphous Solid Dispersions, Springer, New York, 2014, pp. 123-161.

[236] T. Kiran, N. Shastri, S. Ramakrishna, M. Sadanandam, Surface solid dispersion of glimepiride for enhancement of dissolution rate, Int. J. Pharm. Tech. Res., 1 (2009) 822-831.

70

[237] F. Jijun, X. Lishuang, W. Xiaoli, Z. Shu, T. Xiaoguang, Z. Xingna, H. Haibing, T. Xing, Nimodipine (NM) tablets with high dissolution containing NM solid dispersions prepared by hot-melt extrusion, Drug Dev. Ind. Pharm., 37 (2011) 934-944.

[238] N.A. Armstrong, Tablet manufacture by direct compression, in: J. Swarbrick (Ed.), Encyclopedia of Pharmaceutical Technology, 3rd ed., Informa Healthcare, New York, 2007, pp. 3673-3683.

[239] Z.A. Worku, J. Aarts, G. Van den Mooter, Influence of Compression Forces on the Structural Stability of Naproxen/PVP-VA 64 Solid Dispersions, Mol. Pharm., 11 (2014) 1102-1108.

[240] E. Roueche, E. Serris, G. Thomas, L. Périer-Camby, Influence of temperature on the compaction of an organic powder and the mechanical strength of tablets, Powder Technol., 162 (2006) 138-144.

[241] M. Cespi, G. Bonacucina, L. Casettari, S. Ronchi, G.F. Palmieri, Effect of temperature increase during the tableting of pharmaceutical materials, Int. J. Pharm., 448 (2013) 320-326.

[242] R.J. Buice, T.B. Gold, R.A. Lodder, G.A. Digenis, Determination of moisture in intact gelatin capsules by near-infrared spectrophotometry, Pharm. Res., 12 (1995) 161-163.

[243] M.G. Fakes, B.J. Vakkalagadda, F. Qian, S. Desikan, R.B. Gandhi, C. Lai, A. Hsieh, M.K. Franchini, H. Toale, J. Brown, Enhancement of oral bioavailability of an HIV-attachment inhibitor by nanosizing and amorphous formulation approaches, Int. J. Pharm., 370 (2009) 167-174.

[244] B. Démuth, Z. Nagy, A. Balogh, T. Vigh, G. Marosi, G. Verreck, I. Van Assche, M. Brewster, Downstream processing of polymer-based amorphous solid dispersions to generate tablet formulations, Int. J. Pharm., 486 (2015) 268-286.

[245] D. Faroongsarng, G.E. Peck, The swelling & water uptake of tablets iii: moisture sorption behavior of tablet disintegrants, Drug Dev. Ind. Pharm., 20 (1994) 779-798.

71

Captions to illustrations

Figure 1: Biopharmaceutics classification system and various approaches to overcome solubility and permeability challenges. (Adapted from [6, 7])

Figure 2: Thermodynamic descriptor-temperature diagram for the various states of a drug.

As the crystalline drug is heated, the thermal energy breaks the crystal lattice structure and at melting point (Tm) the drug gets converted into liquid state. To generate amorphous state, the liquid should be cooled at a sufficiently fast rate. This results in conversion of liquid to supercooled liquid state and subsequently the system falls out of the equilibrium at the glass transition temperature (Tg). For certain drugs such as itraconazole, formation of mesophase is observed. Tk is the Kauzmann temperature which is a hypothetical temperature at which the entropy of the supercooled liquid becomes equal to that of crystal. Spray drying process is also similar to quenching, as the time scale in which droplet to particle conversion takes place is really small and in ideal cases does not allow crystallization. (Adapted from [10])

Figure 3: Spray drying set-up

Typical spray-drying system consists of various components. Component choice and their operating parameters have crucial influence on the process output. Few of these aspects are listed below the components (1-7).

Figure 4: Typical drying process of a droplet upon exposure to drying medium

Upon exposure, for a small duration the droplet experiences sensible heat till the wet-bulb temperature is reached. Further, the droplet is assumed to dry at a constant evaporation rate (𝜅) and the droplet diameter (and thus surface) decreases linearly from d0 initially to d(t) at time t. During drying, the viscosity of the droplet increases, solvent content decreases and further droplet solidification takes place. At a certain critical moisture content, the evaporation rate falls and the droplet temperature increases to the dry-bulb temperature. The characteristic droplet drying time (τD) is given by the ratio of square of the d0 and the evaporation rate 𝜅. The drying curve is affected by the solute properties and process parameters.

During droplet drying process, there is a movement of solute from high concentration exterior to the interior of the droplet and outward movement of the solvent. Particles of various morphologies can be formed and the possibility of their formation can be estimated using a dimensionless quantity, Peclet number (Pei) of solid component i, which is the ratio of 𝜅 and diffusion coefficient (Di). When Pei<1, solid particles are likely to be formed, whereas, when Pei>1, hollow spheres or dimpled/wrinkled particles can be formed. The surface enrichment (Ei) of a solid is given by the ratio of surface concentration (cs,i) to average concentration in the droplet (cm,i) and is related to peclet number and a function, βi. (Adapted from [28, 41])

72

Figure 5: Mollier diagram

Note that red lines indicate isotherms (the lines are not exactly horizontal), green broken lines indicate isenthalpic lines and slope diagonally. Absolute humidity is indicated by the blue vertical lines whereas relative humidity is shown by the black curved lines. Absolute humidity is the weight of water vapor per kg of air (g/kg). Relative humidity is the ratio of the amount of actual water vapor to the max. amount of water vapor (%). The pressure of the humid air will affect its characteristics, hence, any mollier diagram is at single specific barometric pressure. (Modified from [40])

Figure 6: Mollier diagram example

Suppose a liquid is atomized into the drying chamber with drying gas at temperature T1 and humidity H1 (extremely low). T1 can be called the dry-bulb temperature. Since the water is added without any heat supply, the process would move along isenthalpic lines (thick broken green line). As a result of the liquid atomization, temperature of the gas decreases to T2 and humidity increases to H2. The drying gas cannot take up water indefinitely and at a certain limit the water exerts a vapor pressure equal to the partial pressure of the water vapor in the given mixture. At this point, the temperature T3 is called the thermodynamic wet-bulb temperature or adiabatic saturation temperature. Note that the wet-bulb temperature is different from adiabatic saturation temperature.

Figure 7: Droplet to particle transition and effect of evaporation rate

As a particle is exposed to the drying gas, droplet core temperature increases in three stages with corresponding droplet to dry particle transition. The droplet solvent content also decreases to a certain minimum level. The droplet evaporation rate can determine the final morphology of the particles. Slow evaporation rates result in denser particles. The process gives enough time for the solute particle to migrate and come in close proximity. Strong drug-carrier interactions would deter close proximity of drug molecules to form a crystal lattice, resulting in single phase solid dispersions. Weak drug-carrier interactions can result in phase-separation. Fast evaporation rate results in skin formation at the droplet surface and generates particle of less density. Based on the permeability of the skin and within the particle, porous or hollow particles can be obtained. (Modified from [47, 48])

Figure 8: Downstream processing of the spray dried ASD

The spray dried powder should have certain CQA (particle size and its distribution, bulk density, flowability etc.) for successful downstream processing. The powder can be subjected to a pre-densification step and/or granulation to improve its formulatibility. The powder/granules with improved flowability are either compressed or filled in capsules and packaged. The thermal, moisture and mechanical stresses involved in the unit operations have the potential to effect the amorphous solid dispersion adversely. The final product should adhere to the quality target product profile.

73

Figure 9: SWOT analysis of spray drying process to generate ASD

74