John T. Fell

For any film coating process, a balance must be obtained between achieving a fast coating time and a high quality coated product. Film coating formulations have recently been introduced, which may significantly increase the spray rates possible in production processes. Indeed, the inclusion of formulation additives such as lactose1 and microcrystalline cellulose2 should not only increase spray application rates but may also improve product quality by promoting tablet-film adhesion. 	

Although much recent work has investigated methods of improving film coating process times and product quality by formulation techniques,3,4 relatively little work has examined the characteristics of the spray applied to the tablets.	

Twitchell5 highlighted the effect that the droplet atomization process can have on any given film coating process. Factors such as the pressure and volume (or mass) of air used; the type and design of spray gun; the spray rate; the viscosity of the coating suspension; and the distance from the spray gun to the tablet bed influence the characteristics of the spray produced. Upon arrival at the tablet bed, impingement, spreading, penetration, coalescence and adhesion may all be influenced by the characteristics of the spray produced. Coating times and uniformity may be influenced by the spray shape/area used during the coating process. Additionally, process problems, such as picking, which give rise to detrimental tablet quality, will be influenced by the spray characteristics. Unless the droplet size and velocity, the atomizing air to liquid ratio, the air exit velocity and the spray areas are considered in scale-up, problems not encountered during development may manifest themselves in production.	

The methods for characterizing sprays include laser light scattering techniques,5 droplet velocity using Phase Doppler methods6 and measurement of the spray dimensions (pattern) using a patternator device.6 	

The purpose of this study was to investigate the factors influencing atomization from two spray guns. In particular, consideration was given to spray rates and air pressures used during the coating of small development-scale batches on interchangeable drum coating equipment and how the spray conditions used here compare with those typically used in a production environment.	

Laser light diffraction. Laser light diffraction (Spraytec System; Malvern Instruments Ltd, Malvern, UK) was used to ascertain the droplet size characteristics for each of the spray guns. The median droplet size and droplet size distributions were determined for both guns.		

Figure 1:  The relationship between atomizing air pressure, fan air pressure and median droplet diameter for the Manesty spray nozzle and Figure 2:  The relationship between atomizing air pressure, fan air pressure and median droplet diameter for the Schlick 930/7-1 spray nozzle.				

A 12% w/w suspension of Opadry OY 35018 (Colorcon, Dartford, UK) was used as the spraying medium for all the droplet size measurements. This suspension was pumped and atomized using a Flowtab system (Manesty, Knowsley, UK).	

For droplet size determination, the gun was mounted on a manifold and positioned to spray through the laser beam. The gun was activated and the spray aimed through the laser. The laser/spray guns were protected with a Perspex cylinder to prevent any residual spray swirl recirculating back through the laser beam, which could produce atypical results. The spray was extracted by exhaust to prevent spray bounce back.	

A 40 g/min spray rate and a measuring distance of 15 cm was used - both intended to mimic typical process conditions used in a 610 mm diameter drum (16 L working capacity).	

Each condition tested was undertaken in triplicate to allow the reproducibility between runs to be evaluated. The spray guns investigated were a Manesty and a Schlick 930/7-1 (Coburg, Germany). Both guns were fitted with 1.2 mm orifice fluid nozzles and each was fitted with the standard air cap for the model under investigation. A range of atomizing and fan air pressures (0-200 kPa) was studied. 			

Figure 3:  Comparison of fan air mass flow rate at different fan air pressures for the Manesty and Schlick spray guns and Figure 4:  Comparison of atomizing air mass flow rate at different atomizing air pressures for the Manesty and Schlick spray guns.	

The results for droplet size are expressed as the median volume diameter (Dv 50). Droplet size distribution is expressed as the span (Equation 1).	

Phase doppler droplet velocity measurement. A phase doppler particle analyser (PDPA) from Aerometrics Inc. (Sunnyvale, California, USA) was used to measure droplet velocity from the Manesty spray gun. The gun was mounted on a manifold and the spray activated such that it was aimed directly downwards. The laser was set up so that the measuring zone (the point where the two horizontal laser beams crossed) was directly in the centre of the spray cone. The gun-to-measurement distance was set at 10 cm and the coating solution used for the droplet velocity measurements was 9% w/v Pharmacoat 606 (Shin Etsu Chemical Co., Tokyo, Japan). A range of atomizing and fan air pressures (0-200 kPa) was studied.	

Viscosity measurement. The viscosity of each coating formulation was measured prior to the laser and spray pattern studies using a Brookfield viscometer (Brookfield Viscometers Ltd, Harlow, UK) fitted with spindle no. 2 at a speed of 60 rpm.			

Spray shape/area measurement. The spray area of a series of gun configurations was determined using a 12% w/w Opadry OY35018 suspension. The appropriate spray gun was mounted on the manifold of a Premier 200 coating machine (Manesty) such that the guns sprayed directly downward. Upon activating the spray, the gun was covered with a polythene bag until the desired spray rate, atomizing and fan air pressures were achieved. At this point, the bag was removed from the gun and the spray pattern built up on a sheet of A3 paper. Spraying time was set at 2 s. The wet spray patterns were left to dry overnight and the dimensions and area of each spray pattern were determined.	

Spray gun air consumption measurement. The air consumption at a series of atomizing and fan air pressures, and liquid flow rates, was determined during the spraying of batches of tablets in an XL coater (Manesty). Air flow readings were taken with an online air flow meter (Dwyer Instruments Inc., High Wycombe, UK). Volumes were converted to air masses using appropriate pressure and temperature conversions.	

Droplet size and distribution. The viscosity value range determined for the Opadry suspension used for the laser work was 250-260 mPas for all the experiments performed. Replicate runs gave almost identical droplet size data. 				

Therefore, the technique used and the experimental set-up gave a good means of comparing the gun configurations.	

Volume diameters. Figures 1 and 2 show the median volume diameters of the spray produced by the two guns under different experimental conditions. The results confirm the findings of Twitchell,5 who showed for a range of spray guns tested, that as the atomizing air pressure increases, the droplet size decreases. The extent of the influence of fan air on droplet size differs between the spray guns. At the lowest atomizing air pressure used, the fan air influences the median droplet size for both guns. However, the effect of fan air virtually disappears for the Manesty gun at 100 kPa atomizing air pressure, whereas it still influences the droplet size from the Schlick gun at this pressure.	

In addition to the atomizing air pressure, the droplet size of any spray produced will be dependent on a number of factors. The volume or mass of air used, together with the pressure, will influence the air:liquid ratio. This, combined with the air exit velocity, must influence the efficiency of the atomization process and hence the droplet sizes produced.				

Table I:  Atomizing air annulus exit velocities for the Manesty and Schlick spray guns.	

Pressures and mass flow rates. Figures 3 and 4 show the relationship between atomizing/fan air pressures and the mass flow rates. The air mass flow rate for each spray gun is different. In particular, the atomizing air mass flow rate varies widely between the spray guns - the Manesty gun uses approximately four times more air at a given pressure than the Schlick gun. The greater amount of air used by the Manesty gun means that the air:liquid ratio is higher at the equivalent atomizing air pressure, imparting more energy into the atomization process. This is likely to be a contributing factor to the median droplet diameter (Dv 50) for the Manesty gun; at 100 kPa (atomizing) and 100 kPa (fan air), the droplet diameter is 28.3 mm, which is less than the value found for the Schlick gun (43.2 mm) at the same pressure.	

With both guns, the air intended to alter the shape of the spray (fan air) is kept completely separate from the air used to atomize the coating suspension. Nonetheless, the results obtained show that the fan air pressure has some influence on the droplet size. Previously, Twitchell5 suggested that spray shape may have an influence on droplet size and that as the spray cone is flattened (by increasing the fan air pressure) the median droplet size decreases. This finding appears to be confirmed in the present study. The main difference between the guns in terms of the influence of fan air on droplet size can again be explained by the differences in air mass flow rates.	

Although the fan air mass flow rate is very similar between the two spray guns at an equivalent pressure, the Schlick gun uses large quantities of fan air compared with atomizing air. Thus, it is not unreasonable to assume that the greater mass ratio of fan air to atomizing air may influence the characteristics of the droplets, particularly at lower atomizing air pressures. The different results observed for air mass flow rates can best be explained by considering the different designs of the spray guns.					

Figure 7:  The relationship between span and atomizing air pressure at 200 kPa fan air pressure for the Manesty and Schlick spray guns at 40g/min spray rate and Figure 8:  Air:liquid ratio for the Manesty gun at various typical development spray rates and pressures.	

Manesty spray gun. Figure 5 shows a photograph of the Manesty spray gun. The atomizing air passes from the main gun chamber through the nozzle into the atomizing air annulus via four holes set at 90Þ These holes are approximately 2.5 mm in diameter. The atomizing air then passes out of the air annulus where it exerts its atomization effect. The fan air is fed separately and exerts its effect by passing through holes (one on either side) in the horn of the air cap. These holes are approximately 1.7 mm in diameter.	

Schlick spray gun. Figure 6 shows the Schlick spray gun. The atomizing air enters the atomizing air annulus via six holes positioned at 60Þ These holes (approximately 1.5 mm in diameter) are smaller than those on the Manesty gun. The fan air is supplied from two holes of differing diameter (1.7 and 1.0 mm) positioned on either side of the air cap. The diameter of the atomizing air annulus is different between the guns - 4.0 mm for the Manesty and 4.5 mm for the Schlick. The outside diameter of the fluid nozzle also differs (2.7 mm for the Manesty, 2.9 mm for the Schlick). Thus, the area through which the atomizing air leaves the guns varies considerably (6.8 mm2 and 9.3 mm2 for the Manesty and Schlick spray guns, respectively). This, in turn, will influence the exit velocity of the air. Table I shows the theoretical atomizing air annulus velocity for the two different guns.	

Air annulus. Twitchell5 stated that a major factor affecting the atomizing process is the area of the atomizing air annulus because this influences the volume and mass flow rate, and the velocity of the atomizing air. The air mass flow rates of the two guns will differ considerably and be significantly higher for the Manesty gun. This can be explained by the different number and dimensions of the holes that feed the air to the annulus and the area of the atomizing air annulus itself. The pressure measured and quoted in most film coating processes for atomizing air is normally the pressure in the line feeding the guns. Thus, although the atomizing air pressure in the lines feeding the spray guns in the current study is the same, because of differences in design, the actual conditions at the atomizing point may be vastly different for each gun.				

Figure 9:  The velocity of droplets of a 9% Pharmacoat 606 solution (viscosity of 202 mPas) sprayed from a Manesty gun at 10 cm distance at various atomizing and fan air pressures.	

Droplets. Another important consideration with respect to the coating process is the uniformity of droplet size. Coating theory suggests that the homogeneity of the film may be related to the properties of the droplets as they impinge on the tablet core. Uniform droplet distribution would be desirable because it would suggest similar droplet behaviour at the tablet drop interface. One method of considering uniformity of droplet size is to consider the span of the size frequency profile.	

Figure 7 shows the change in the span for the two spray guns at different atomizing air pressures. The trend for droplet size distribution is that more uniform droplet sizes arise as the atomizing air pressure is increased. One major factor that will determine the droplet size distribution is the air:liquid ratio. When this ratio reaches the required level it appears that the uniformity of droplets is optimized. This occurs at a lower pressure for the Manesty gun, presumably because it uses higher atomizing air mass flow rates and because the air exit velocities are higher.	

The findings with respect to air mass flow rates and droplet size are particularly important when one considers the advent of coating increasingly small batches of tablets in interchangeable drum coaters. In such machines, batches between 250-3000 g can be coated. For small batches, such as these, the pressures used range between 50-200 kPa at spray rates of 10-40 g/min. The results from development work are then scaled up onto production equipment where pressures of 200-400 kPa are commonly used together with spray rates of 80-120 g/min/gun. 	

Air:liquid ratios. The air:liquid ratios for the Manesty gun at typical development spray rates and pressures are shown in Figure 8. Generally, a ratio of greater than 4 has been considered necessary to optimize the atomization process.5 For a gun with a high air mass flow rate and a high air exit velocity at low spray rates (10-20 g/min) it is possible to use low atomizing air pressures and still have an air:liquid ratio greater than 4. However, as the spray rate is increased, it is necessary to increase the atomization pressure to maintain an air:liquid ratio greater than 4. For the Manesty gun, the droplet size reaches a minimum at approximately 150 kPa (Figure 1) at 40 g/min. At this pressure, the air:liquid ratio is less than 4, suggesting that the exit air velocity is contributing to the atomization process. Another indication that the air:liquid ratio is not the only factor influencing atomization is that at typical production spray rates (100-150 g/min) and pressures (300-350 kPa) the ratio can be less than 4. It is apparent that different spray guns will have different optimal air:liquid ratios dependent on other factors such as atomizing air exit velocity, air annulus area and suspension viscosity. This is an important scale-up consideration. 				

Table II:  Spray characteristics at atomizing air pressure 250 kPa; spray rate 80g/min; gun distance 25 cm.	

For spray guns of low air mass flow rates, such as the Schlick, increasingly high pressures must be used to ensure maximum atomization. Thus, at a spray rate of 40 g/min, droplet size continued to decrease at all the pressures tested (Figure 2). Presumably the combination of low exit velocity and low air mass flow rate means that higher pressures are necessary to achieve optimal atomization. Therefore, when scaling up, the same air:liquid ratio and droplet size profile should ideally be used on the production scale that gave the appropriate results during development.	

Spray patterns. Although fan air has been shown to influence droplet size, its main influence is in determining the dimensions and shape of the spray zone produced. Tables II and III show dimensions, areas and shapes of the various patterns obtained from the Schlick and Manesty guns at different pressures, spray rates and gun distances. The tables confirm that there is a difference in the dimensions of the spray pattern produced by the two guns. 	

The Schlick gun produces a much wider spray pattern than the Manesty at both atomizing air pressures. This can again be explained in terms of the differences in air consumption and design of the air cap of the different guns. The relatively large amount of fan air in the Schlick will have the effect of flattening the spray to a greater extent. Thus, the spray produced by the Schlick under production conditions will cover a larger area of the tablet bed than the Manesty gun under equivalent conditions. 	

In addition to fan air consumption, another factor that will affect the shape of the resultant spray is the design of the air cap, and there are clear differences between the Schlick and Manesty air caps. As well as the difference in the number and dimensions of the holes used to deliver the fan air, the angle of the air cap horns differ. In the Manesty gun, the air cap horns deliver the air at a steeper angle. This must influence the point on the spray cone at which the cone is flattened and has implications on the dimension of the spray pattern. The steeper angle of the Manesty gun produces spray patterns of a smaller area. 				

Table III:  Spray characteristics at atomizing air pressure 100 kPa; spray rate 40g/min; gun distance 15 cm.	

Droplet velocity. Another factor worthy of consideration is the velocity of the droplets themselves as they leave the spray gun and impinge on the tablet bed. The atomizing and fan air pressures, air consumption, air exit velocity and distance of the spray gun from the tablet bed will not only influence the droplet size but will also determine the velocity of the droplets during impingement. If the atomizing air pressure (or air exit velocity) is too high or if the gun is too close to the tablet bed, then a hole is blown in the tablet bed. It is normal practice when setting up a coating process to ensure that this hole in the tablet bed is kept to a minimum. 	

Figure 9 shows the velocity of droplets of a 9% Pharmacoat 606 solution (202 mPas viscosity) sprayed from a Manesty gun at a distance of 10 cm at various atomizing and fan air pressures. The droplet velocity measured ranged from 11-32 m/s - within the range quoted by Scattergood et al.6	

As atomizing air pressure increases, the volume of air passing through the gun increases, as does the air exit velocity. This results in an increase in droplet velocity. As would be expected, an increase in fan air pressure widens the spray pattern and, as a result, decreases droplet velocity at the measuring point. Thus, by increasing fan air pressure, the hole blown in the tablet bed will be minimized. However, this increase in fan air pressure will also influence the spray pattern and also have some effect on the droplet size, and potentially the droplet size distribution. 	

Implications. These results have important implications if one considers the film coating process. Processes scaled up to equipment in which the spray gun, air:liquid ratio or atomizing and fan air pressures are different from those used for the development phase will have widely different characteristics. These differences will certainly result in processes that are fundamentally different to those produced in development. 	

A variation in the properties of the atomized coating suspension produced from the spray gun will lead to droplets behaving differently as they impinge on the tablet bed. Larger droplets will have a proportionally lower surface area, meaning less evaporation will occur as they travel to the tablet bed. Smaller droplets will tend to evaporate to a greater extent. Certain film defects, such as picking, twinning, mottling and orange peeling are caused by inappropriate droplet or spray properties at the production scale. Efficient processes that are well controlled require optimal droplet size and velocity as well as a tight droplet size distribution. This ensures the balance between over wetting and spray-drying is maintained. Differences in the spray pattern, such as the dimension and area of the spray zone, will occur between different spray guns set up at the same pressures, which may influence coating times and coat uniformity. 	

More crucially, significant differences in the physical properties of the film may occur as a result of differing atomizing conditions as the coalescence process may be affected. When scaling up, the atomizing conditions should be considered - if the atomization conditions differ at the tablet bed, then the physicochemical properties of the droplets will also differ. This will influence spreading, wetting, penetration and coalescence phenomena. Differences here may significantly influence the homogeneity of the polymeric film produced. Thus, the potential for differences in the roughness and porosity of the films exists. This has important implications, particularly when the film is to be used for a sustained release or enteric application. 	

Articles

Consideration of the key process variable will define the ease with which coating processes can be transferred from development to production. This study investigates those factors influencing atomization from two spray guns and examines how development-scale procedures on interchangeable drum coating equipment compare with those typically used in a production environment.