Chapter 21
Tablets and Capsules
Solid medicaments may be administered orally as powders, pills, cachets, capsules or tablets. As these contain a quantity of drug which is given as a single unit they are known collectively as solid unit-dose forms even in the case of sustained action preparations which, technically, contain the equivalent of several normal doses of drug. The stringent formulation requirements of modern medicaments, the many advantages of tablet and capsule medication, coupled with expanding health services and the concomitant need for large-scale economic manufacture, have lead to a steady decline in the prescribing of powders and pills. A few medicaments such as the salts of para-aminosalicylic acid are administered as cachets and the production of these dose forms is described elsewhere in this volume. Tablets and capsules, on the other hand, currently account for well over half the total number and cost of all National Health prescriptions issued in the UK.
Tablets
Tablets are solid masses made by the compaction of suitably prepared medicament (granules) by means of a tablet machine. Although it is possible to manufacture tablets in a wide range of shapes, official tablets are defined as circular discs with either flat or convex faces. The British Pharmacopoeia 1973 includes monographs for over three hundred tablets.
Advantages of Tablet Medication
When correctly formulated and manufactured, tablets provide an accurate, stable dose of drug with the necessary physical and chemical properties for the required duration and intensity of therapeutic action. Modern methods are capable of large scale economic production with a high degree of tablet uniformity both within and between batches. Suitable formulations provide sustained release characteristics and coatings can be applied to improve palatability or reduce the incidence of gastric irritation. Of all dose forms, the tablet presents the medicament in the highest practicable state of compaction. This permits the use of light weight, low cost containers for packaging, while emergency supplies of the drug can be conveniently carried by the patient.
Disadvantages of Tablet Medication
Some patients, particularly children and the seriously ill, may experience difficulty in swallowing tablets. If the tablet is too large, reformulation as two smaller units each containing half the required dose may minimize the problem in some cases. Occasionally a particular property of a medicine, such as the demulcent action of a linctus, cannot effectively be obtained with a tablet. Absorbents are available which allow the production of tablets containing a small amount of liquid, but it is usually more satisfactory to formulate as a soft capsule.
Types and Uses of Tablets
To secure rapid release of drug the majority of tablets are required to break down (disintegrate) rapidly in the stomach, but there are a number of exceptions to this general rule. If the drug is inactivated at low pH, causes gastric irritation, or intended to exert its effect on the lower part of the gut, an enteric coating is applied to ensure that disintegration does not take place in the stomach but readily occurs in the small intestine. Examples are enteric coated tablets containing erythromycin or the alkali metal halides. Only part of the medicament in a sustained action tablet is released in the stomach to obtain prompt therapeutic action while the remainder is released at a controlled rate as the tablet progresses along the gastrointestinal tract. As an alternative to enteric coating for the minimization of gastric irritation, the drug, e.g. aspirin, may be formulated as a soluble tablet to be dissolved in water prior to administration; an effervescent base, as in Effervescent Potassium Tablets BPC, improves palatability. If the preferred site of absorption is the oral mucosa the tablet is directed to be dissolved in the mouth (isoprenaline), under the tongue (ethisterone) or chewed (phenolphthalein). For lozenges slow solution in the mouth, rather than disintegration, is required.
A number of preparations, not intended for oral administration, are also prepared by compression, examples being Mouth-wash Solution Tablets BPC, Testosterone Implants BP and Acetarsol Pessaries BPC. Solution tablets are dissolved in water and the solutions applied to mucous surfaces or externally. A distinctive shape and colour is used for solution tablets that contain a poison.
Granulation
The Tablet Compression Cycle
Further processing, that is granulation, is usually required before a drug can be tabletted. To a large extent the requisite physical properties of the granules are determined by the design of the tablet machine and the nature of the compaction process. Details of machine design are given later, but a brief discussion of the compression cycle, specifically that of the single-punch tablet press, is relevant at this point to a proper understanding of the objectives to be attained by granulation.
Granules flow from the hopper (Fig. 21.1) into a feedshoe which oscillates over the die to promote uniform flow of material into the die cavity. For a given granule the volume of the cavity, which is adjustable, governs the weight of the tablet. At the end of the filling stage the toe of the feedshoe is deflected so that it smooths the surface of the granules in the die and cannot foul the top punch as it is lowered to compact the granules. The penetration of the top punch into the cavity can be adjusted to regulate the degree of compaction, that is the tablet hardness. The top punch is then quickly raised and, after a short delay, the bottom punch moves upwards to eject the tablet. Finally, the bottom punch drops to the filling position and the ejected tablet is pushed towards the collection chute by the toe of the feedshoe as it moves forward to commence the filling cycle of the next tablet.
Fig. 21.1 The compression cycle for a single-punch tablet machine.
Essential Granule Properties
From the foregoing it should be clear that for the successful production of tablets the granules must fulfil a number of requirements. With both single punch and rotary tablet presses only about a fifth of a second or less is available for filling the die cavity. Rapid, reproducible, flow of granules is essential if the tablet weight is to remain constant throughout the batch. Furthermore, particles of different size or density must not separate in the hopper as a result of machine vibration if both tablet weight and composition are to be maintained. Granules of uniform size minimize difficulties due to separation, but yield tablets with an inelegant pitted surface. A broader size distribution, with a proportion of fines to fill the intergranule spaces, avoids the latter problem. The conflicting requirements for tablet uniformity and finish must be carefully balanced by rigid control of the granulation process.
During compaction intergranule bonds must be formed to allow the production of a tablet which combines the required disintegration characteristics with sufficient firmness to withstand damage due to packaging, transport and other hazards. The volume of the granules may be several times that of the tablet into which they are compacted. The reduction in volume is due to removal of air and this must readily escape for the proper development of the interparticle bonds. Compaction also forces the tablet material into very close contact with the wall of the die. Excessive die-bore wear and damage to the edges of the tablet result if the die-wall friction is not reduced, usually by the addition of a lubricant to the granules.
In addition to the above it is, of course, essential that the medicament in a tablet is stable and released at the required rate. Primarily these are matters of correct formulation and processing.
Granulation Processes
Three processes namely, moist granulation, preliminary compression or slugging and dry granulation are in common use and are described in the British Pharmaceutical Codex, 1973.
Moist Granulation
This is still the most widely used of the three processes. The mixed, powdered, tablet constituents are converted to a moist cohering mass by the incorporation of granulating fluid. Granules formed by passing the moist mass through a screen are dried, rescreened to break down agglomerates and blended with other tablet adjuvants such as lubricant and disintegrant. Although the majority of drugs are available as powders suitable for direct use in the moist granulation process, in a few cases the particle size of the powder affects the eventual compression characteristics of the granules. Smith (1950), for instance, reports that a powder finer than 80mesh should be used for the production of tablets containing either phenacetin or phenobarbitone. Often the dose of the drug is very small and an inert diluent (bulking excipient) should be added before granulation to give, as a minimum, a 50mg tablet. Some diluents will absorb small quantities of liquids (absorbent excipients); others may be used in the dry granulation process (see Table. 21.1). Although the inertness of diluents is generally assumed, there have been reports of their adverse effect on product performance. Stephenson and Humphreys-Jones (1951) examined a number of formulations for glyceryl trinitrate tablets and demonstrated the superior stability of preparations using a lactose base. Richman et al. (1965) noted that microcrystalline cellulose permits dry granulation of this during and yields a more stable product than that obtained with a lactose–sucrose diluent. A carbonyl–amine reaction causes darkening of tablets containing lactose or other reducing sugars together with a primary amine (Duvall et al., 1965). The BPC 1968 states that calcium phosphate may interfere with the absorption of calciferol.
Table 21.1 Some diluents used in tabletting |
|
| Diluent | Uses and comments |
| Calcium Phosphate BPC | Absorbent diluent. Increases granule BPC density but some grades are abrasive. |
| Calcium Hydrogen Phosphate USP | Used for direct compression formulations. |
| Colloidal silica | Small proportion used as an absorbent diluent or to improve granule flow by glidant action. |
| Anhydrous Dextrose BP | Absorbs moisture at high relative humidity—see BPC. |
| Dextrose (spray dried) | Useful for direct compression formulations but absorbs moisture at high relative humidity. |
| Lactose BP | Inert, tasteless, odourless, inexpensive and gives granules by moist granulation which have good tabletting properties. Incompatible with primary amines (Duvall et al., 1965). |
| Lactose (spray dried) | Used for direct compression formulations but tablets may assume a brown tinge on storage (Brownley & Lachman, 1963). Is also incompatible with primary amines. |
| Lactose (anhydrous) | Used for direct compression formulations. Care should be taken to prevent moisture uptake. Is also incompatible with primary amines. |
| Mannitol BP | Dissolves readily and has a ‘cooling’ effect in the mouth. |
| Microcrystalline cellulose | Mainly used for direct compression formulations; has some lubricant and disintegrant effect and may improve the behaviour of granules made by other methods. |
| Sodium Chloride BP | Oral tablets containing this substance must be dissolved in water before administration to avoid gastric irritation—see BP. Usually reserved for the formulation of solution tablets. To avoid binding during compression, dry thoroughly and compress while still Warm. |
| Starch BP | Small proportion of dried starch may be used as an absorbent excipient but its main use is as a disintegrant. |
Mixing: To ensure batch uniformity it is essential that the dry powders are thoroughly mixed before moistening. It must be stressed that mixing problems are particularly acute where one or more of the constituents forms only a small proportion of the tablet weight. Such minor components should, if possible, be dissolved in the granulating fluid or other suitable solvent and incorporated during the moistening stage. This method is included in the British Pharmacopoeia 1973 monograph on tablets and was specified for Soluble Paediatric Atropine Sulphate Tablets BPC 1968. Train (1960) pointed out that solute migration during the drying stage, analogous to separation on a chromatographic column, will lead to unequal distribution of solute within the granule mass, that is, to virtual demixing and this may be difficult to rectify in subsequent processing. With soluble dyes the effect is visually apparent in the uneven coloration of the dried granules and the mottled appearance of the finished tablet. Jaffe and Lippmann (1964) studied the migration of F.D. & C. Blue No.1 dye in lactose and noted that tragacanth, acacia and talc tended to ‘fix’ the dye so reducing its migration. The problem may be circumvented by the use of lake dyes, that is, soluble dyes bonded to an inert base. Lake dyes usually exhibit greater stability to light than the corresponding soluble colorant and are now much used, particularly for tablet coating.
A further mixing difficulty may arise if the drug is physically adsorbed on to the surface of the granules and these have a broad-size distribution. A large fraction of the drug will be adsorbed on the fines since these make a disproportionately large contribution to the total granule surface. Should separation of fines occur, this again is equivalent to demixing. The deliberate removal of fines to improve the tabletting properties of the granules is particularly suspect in these circumstances as this could lead to the production of low potency tablets. Where solvent dispersion of the minor components is not feasible a concentrated triturate may be prepared and subsequently diluted, in steps, with the remainder of the tablet ingredients. This should improve mixing, as comparable quantities of materials are blended at each stage of the process.
Batches up to a few hundred grams may be mixed in a mortar. Larger quantities of powders require power driven mixers, e.g. Y-cone, Rotocube or ribbon blenders. The transference of large quantities of material from the mixing to the moistening equipment may be obviated by the use of trough or change-pan mixers (Fig. 21.2) which are capable of performing both operations. The change-pan mixer is so called because the mixing blades and pan are detachable. Only one drive unit is required to service a number of blades and pans in which separate products may be prepared according to the needs of production. In this way the mixer is utilized more efficiently, transport of granules for further processing is facilitated and cleansing of the blades and pan simplified.
Fig. 21.2 The Beken change pan mixer with working capacity of about 0.1m3. The machine is shown with the pan lowered and the guard open (Beken Engineering Ltd.).
Moistening: This involves the gradual addition of granulating fluid to the dry mixed powders. A pestle and mortar is suitable for small batches, but power driven equipment such as the trough or change-pan mixer is essential for thorough blending action on a larger scale. Initially most of the fluid is held by capillary forces at the points where particles are in contact (Fig. 21.3, Stage 1). Further liquid builds up a continuous film on the external surfaces of the particles and at this second stage the powder is obviously damp and is very coherent. A quite small addition of fluid is now sufficient to produce a film of such thickness that the interparticle friction is lowered to an extent that close packing of the particles can take place. At this third stage of moistening, air in the voids is displaced by fluid and a wet paste, technically a very concentrated suspension, is formed. This can however be produced once the second stage of granulation has been reached, by prolonged or more efficient blending, without the further addition of granulating fluid, owing to more uniform distribution of existing moistening medium and the mechanical removal of air from the mass. For production of tablet granules the volume of granulating fluid, mixing time and efficiency must be adjusted to give a second stage mass which just coheres when lightly compressed in the hand. Often, this condition may be anticipated by observing the behaviour of the mass which breaks away from the walls of the mixer, while the powder shows obvious signs of aggregation as a result of the greater cohesiveness of the particles.
Fig. 21.3 Stages in the moistening of a powder.
In some cases moistening may be carried out with a solvent which causes partial dissolution of one or more components. With suitable materials the finished granules retain their structure, as the individual particles have been cemented together with solute deposited by the removal of solvent on drying. Reversion to the original powdered state will occur if substances of low natural cohesiveness are solvent granulated and a binding or adhesive granulating agent is required in these circumstances. The adhesive may be added as a solution or mixed with the tablet ingredients as a dry powder for subsequent activation by moistening with solvent. The latter method is sometimes less effective, as many binders are hydrophilic colloids requiring for complete hydration more solvent and mixing time than is normally used at the moistening stage. Table. 21.2 lists some of the more commonly used binders. Ethyl alcohol, isopropyl alcohol, acetone, and chlorinated hydrocarbons are useful moistening agents for water-sensitive drugs; a solvent soluble adhesive may be needed to obtain firm granules. Due to their high cost, the use of anhydrous solvents is usually restricted to those situations where other methods are not feasible. Attention must be given to fire and toxicity hazards and efficient vapour extraction is mandatory.
Table 21.2 Adhesive granulating agents
The most suitable granulating fluid, the volume to be used and the optimum operating conditions are best selected by the production of trial batches of granules. In the early stages of developing a suitable process, when even an approximate value for the volume of granulating fluid is not known, the time taken for moistening may be unduly protracted. Significant amounts of solvent may be lost by evaporation in these circumstances and should be allowed for in subsequent trials. Clearly, the volume of fluid must be sufficient to produce a second stage moistened powder mass and at the same time add sufficient binder to ensure the production of suitable granules. Powdery granules which do not compress well are obtained, if the adhesive is too ‘weak’, is employed in too Iowa concentration, or if the volume of granulating fluid is inadequate. The converse conditions lead to hard granules which produce excessive amounts of fines when dry screened. Symptoms of overmoistening are clogging of the screen and agglomeration of the wet granules. Scale-up calculations based on data obtained from small trial batches often overestimate the concentration and volume of granulating fluid on account of the higher blending efficiency of equipment used for large scale manufacture. Finally, granules must be evaluated on a press of the type used for production as their performance may to some degree be affected by mechanical features such as the method for filling the die cavity, speed of operation and compaction conditions.
Wet screening: Granules will not readily flow into a cavity with which they are comparable in size due to crowding at the entrance of that cavity. On the other hand very fine granules often exhibit poor flow characteristics. Both conditions lead to excessive weight variation in the manufactured tablets. In addition, tablets which are too thin relative to their diameter are prone to breakage when handled. For these reasons the screens used for granulation and the die size for compression are selected (Table. 21.3) according to the final tablet weight. The Table must be regarded as a guide only since, for example, a smaller die size may be required if the material to be tabletted is unusually dense, while screens somewhat larger or smaller than those recommended may give granules with better characteristics.
Table 21.3 A guide to the sieve number, punch size and maximum punch load for the production of tablets
Small batches of moistened mass may be screened by hand, the material being lightly rubbed, rather than forced, through the screen. Too vigorous manipulation at this stage is equivalent to highly efficient blending which, as already explained, may result in the formation of a third stage pasty mass and consequent difficulties due to screen clogging. Hand screening is tedious for more than a few kilograms of moist material, power driven equipment such as the Apex comminutor or the Fitz mill being employed for larger batches.
The oscillating granulator (Fig. 21.4) also finds considerable industrial use. The base of the hopper is closed by a semicylindrical interchangeable screen just above which a rotor oscillates about a horizontal axis. Blades parallel with the axis and integral with the rotor push the moist mass through the screen on to a receiver below. In a recent study, Fonner et al. (1966) found that this type of granulator gave smoother, more nearly spherical granules of a starch–lactose formulation than either hand screening or three other commonly used power-driven granulating machines.
Fig. 21.4 The oscillating granulator.
Drying: The drying temperature for granules is usually about 60°C but may be lower if thermolabile substances are involved. With tray dryers; air exchange is essential to prevent saturation of the oven atmosphere with solvent vapour. By spreading the granules as thin layers on the trays and raking these layers over from time to time, agglomeration of granules and migration of solutes is minimized and even drying of the granules promoted. If different products are to be dried in the oven at the same time care must be taken to avoid cross contamination. Up to 24 h may be required for the drying of large batches of granules.
The fluid-bed dryer (Fig. 21.5) is now widely used in the pharmaceutical industry because the drying time is only 20–30min, irrespective of batch size which may be 60kg or more, The amount of granulated solvent to be evaporated decreases as drying proceeds; thus the temperature of the air leaving the fluidized bed rises until it is constant and approximates to that of the input air stream. With experience the effluent air temperature provides a convenient means of establishing the optimum drying time. If drying is prolonged beyond this point attrition of the granules leads to an increase in the proportion of fines. The overdried granules have inferior compressing characteristics and consequently difficulties are encountered in the production of the tablets. On large scale equipment an automatic timer is provided to shut off the heaters and air flow at the end of the predetermined drying time. Generally, less fluid is needed for granules, which are to be dried in fluidbed equipment as the amount of granulating fluid suitable for tray-dried granules gives a moist mass which is too dense and cohesive for correct fluidization of the bed. On the other hand, undermoistening should be avoided as fines may clog the output air filters and impede the air flow.
Fig. 21.5 The fluid-bed drier.
Dry screening: This breaks down agglomerates and reduces them to a size compatible with the tablet diameter. Too vigorous dry screening causes granule structure to be lost with the production of excessive quantities of fines. Although a proportion of these are required to fill the voids between larger granules, giving a smooth surface to the tablet, excess causes trouble at the compression stage.
Other methods or moist granulation: The conventional process has a wide range of application but is time consuming and involves a large number of separate operations. Alternative methods of moist granulation have been developed whereby granule manufacture can be carried out as a single integrated process. Fluid-bed granulation (and drying) stems from the work of Wurster (1959, 1960) who had earlier patented a fluid-bed method for tablet coating. One equipment in current use for moist granulation has a downward-directed atomizer placed centrally above the fluidization chamber. Correct circulation of material is brought about by the special shape of the perforated base and the progressive expansion of the upper part of the chamber. The powders are mixed, granulated by the spray of warm granulating solution and dried while in a fluidized state. Lubricant, disintegrant and any other necessary tablet adjuvants are then added and blended by a further short period of fluidization.
Granule size is controlled by the volume, temperature and rate of addition of granulating fluid together with the air flow rate, which determines the expansion and circulation of the bed. Normally, these parameters are automatically controlled by presetting a programming device on the basis of data obtained from previous development batches. A granulator capable of processing up to 60kg of material in under 2h with only minimum supervision is shown in Fig. 21.6. Other methods for moist granulation are discussed in Chapter 19. The foregoing methods all show a considerable saving in time and labour costs as compared with the conventional method: the granules often possess superior flow properties due to their greater smoothness and sphericity.
Fig. 21.6 The Calmic fluid-bed granulator. The programming panel may be seen top left. This controls application of granulating agent from the pressure vessel (centre bottom) to the material in the fluidizing chamber.
Preliminary Compression or Slugging
This method is used where the tablet ingredients react chemically when moistened, degrade in the presence of water or when heated, or yield granules with poor flow or compression properties when made by other granulation techniques.
Heavy-duty tablet presses are employed for the compression of the dry, mixed powders. As these rarely flow well, vibratory feed devices may be attached to the hopper while die-cavity filling is facilitated by the use of large (2–5cm) punches and slow speed operation of the press. Flat faced punches are necessary as the high compaction pressures employed for slugging would splay the edges of a concave faced punch: the clearance between the punch and die wall is also larger than that used for ordinary tablets to allow rapid escape of air entrapped in the powders. The compaction pressure must be high enough to ensure that the rough tablets (slugs) give material with a granular structure when forced through a screen. It is usually advantageous to add some lubricant and disintegrant prior to slugging, further quantities of these being added to the granules before final compression. Talc is frequently employed as lubricant, since powder flow rather than die-wall friction is usually the major problem in the preliminary compression stage. Adhesive-containing granules, prepared by moist granulation, tend to retain their structure for long periods in contact with aqueous media. The absence of adhesives from tablets prepared by slugging however usually results in good disintegration.
Paediatric Ampicillin Tablets BPC and the official soluble analgesic tablets containing aspirin may be prepared from granules made by preliminary compression.
Dry Granulation
This is the simplest of the three granulation techniques. The material to be tabletted is screened if necessary, mixed with any additional adjuvants and is then ready for compression. Tablets made in this way are said to have been prepared by direct compression.
Formerly, the alkali metal and ammonium halides, hexamine and potassium chlorate were the only commonly used medicinal substances which could be dry granulated. More recently the list has been extended by the commercial production of suitably granular forms of Dried Yeast, Dry Cascara Extract, Aspirin, some ferrous salts and a few other substances. Provided an increase in the tablet weight can be tolerated, the addition of spray-dried lactose or anhydrous lactose to the medicament, together with lubricant and disintegrant, will often yield granules which may be directly compressed. As Lactose BP does not promote free powder flow, act as a dry binder or possess good compaction properties it would appear that these characteristics reside in the spherical shape of the spray-dried sugar. The applications of this material to tabletting technology have been reviewed by Gunsel and Lachman (1963). Tablets containing spray-dried lactose may assume a brown tinge on storage due, it has been suggested, to the presence in the sugar of 5-hydroxymethylfurfural (Brownley & Lachman 1963). This darkening should be distinguished from that resulting from the reaction of primary amines and lactose. Other excipients which may be used in direct compression formulae are listed in Table. 21.1. Of these, microcrystalline cellulose is of particular interest since, as reported by Reier and Shangraw (1966), this material also possesses marked lubricating and disintegrating properties.
In those cases where one or more ingredients are present in low concentration, efficient mixing may be more difficult with dry than with moist granulation. The problem has a two-fold origin. Firstly, solvent dispersion cannot be used and, secondly, fine powders are essential to secure efficient mixing; but fine powders lack the good flow and compaction characteristics essential to tablet manufacture. The magnitude of the problem may be appreciated by applying the methods described by Train (1960) to the example of 50mg tablets containing 1 per cent by weight of drug. If the density of the material from which the tablet is formed is 1.5g.cm−3, the particle size should not exceed 20microns and at least 107 particles per tablet are necessary to ensure that 99.7 per cent of the tablets in the batch comply with a ± 10 per cent tolerance in the content of the drug.
Dry granulation is the most economical of the three granulation processes in terms of time, labour and equipment. The availability of the drug for absorption should be good due to the absence of adhesive granulating agents while the elimination of drying and moistening stages obviates many of the stability problems associated with the moist granulation process. Against these advantages must be set the relatively high cost of some diluents used in direct compression formulae and the restricted range of materials manufactured in a suitable granular form. It seems likely however that the more widespread adoption of devices for promoting uniform die filling with difficult materials will reduce the cost and extend the applicability of the dry granulation process.
Flavours, Lubricants and Disintegrants
When moist granulation is employed, these adjuvants are normally added to the dry, screened granules, but the incorporation of internal lubricant or a proportion of disintegrant prior to moistening, as described below, are exceptions. For slugging and dry granulation the addition of these adjuvants has been discussed in the relevant sections.
Flavours
Essential oils are included in a few official tablets. They may be mixed with a small quantity of fines or absorbent excipient and the triturate thoroughly blended with the rest of the granules. Dissolution in a volatile solvent aids dispersion if the volume of flavouring agent is small or, alternatively, the solution may be directly sprayed on to the granules. In both cases solvent must be completely removed before compaction. For nonofficial tablets a wide range of flavours in powder form is available.
Flavours may not, of course, be added prior to an operation involving the use of heat.
Lubricants
These may promote granule flow (glidants), minimize die-wall friction (die-wall lubricants) or prevent the adhesion of granules to the punch faces (antiadherents). Talc, for example, is a good glidant but is inferior to magnesium stearate as a die-wall lubricant. In addition to their mode of action, lubricants may also be classified according to the way in which they are added to the granules. External lubricants are incorporated immediately before compression and function by coating the granule surface while internal lubricants are mixed with the dry powders prior to moistening with granulating fluid. Granules prepared in this manner are described as self-lubricating. One step in the moist granulation process can be eliminated by the use of an internal lubricant in conjunction with polyethylene glycol as the binding agent. In suitable cases the water solubility of the glycol permits the formulation of tablets with good disintegration properties. Commonly used lubricants are listed in Table. 21.4.
Table 21.4 Lubricants used in tabletting
Disintegrants
These assist the disintegration of the tablet by swelling (bursters), by improving the penetration of aqueous liquids (Table. 21.5), or by the liberation of gas from an effervescent base. For a number of medicaments, disintegration of the tablet is improved if part of the disintegrant is added to the powders prior to moistening; the disintegrant within the granules assists the disruption of these in contact with aqueous media. Powdered disintegrants, e.g. starch, methylcellulose, etc., are normally employed in proportions which significantly increase the fines content of the granules and in unfavourable circumstances this can give rise to compression difficulties. Additionally, premature disintegration of the tablet may occur in mouth if excessive amounts of disintegrant are employed and, apart from difficulty in swallowing the powdery mass, any unpleasant taste due to the medicament may prove objectionable. In a few cases an effervescent base is used to promote disintegration. This consists of citric or tartaric acid together with sodium bicarbonate, potassium bicarbonate or calcium carbonate. These react in contact with water to liberate carbon dioxide which disrupts the tablet. A slight trace of moisture is sufficient to initiate the reaction which produces further water and is thus potentially self-perpetuating. For stability, very dry materials, rigorous exclusion of moisture during manufacture and efficient sealing of the packages are essential—the material use of anhydrous citric acid in official preparations of this type should be noted.
Table 21.5 Disintegrants used in tabletting |
||
| Material | Concentration in the granules per cent w/w | Uses and comments |
| Alginic acid | 2–10 | Burster which may prove better than starch for ‘difficult’ materials (Berry & Ridout, 1950). May be incorporated prior to moistening to improve disintegration. If added to the dry granules, avoid finely powdered grades as these form a coherent gel on the tablet surface and may prolong disintegration time. |
| Magnesium aluminium silicate | 1–10 | Burster. Veegum is one variety |
| Methylcellulose | 2–10 | Burster. Use medium viscosity grade in a not too fine powder |
| Microcrystalline cellulose | 1–10 | Has some lubricant properties, higher concentrations used as dry-binding diluent |
| Sodium lauryl sulphate | 0.1–0.5 | Wetting agent which improves penetration of the tablet pore system by aqueous liquids. Will counteract the water-repellent effect of, for example, phenothiazine and the stearate lubricants. May minimize the increase in disintegration time which often occurs as tablets age |
| Starch | 2–10 | Burster. Potato and maize starch are usually considered superior to the other varieties. Part may be incorporated prior to moistening |
Nowadays some newer disintegrants that are effective at lower concentrations with greater disintegrating efficiency and mechanical strength are used. These substances are called superdisintegrants. The superdisintegrants on contact with water hydrate swell, change volume or form/produce a disruptive change in the tablet. These substances act by swelling and due to swelling, pressure is exerted in the outer direction or radial direction that causes tablet to burst or the accelerated absorption of water leading to an enormous increase in the volume of granules to promote disintegration. Three major groups of compounds that swell to many times their original size when placed in water while contributing minimal viscosity compared to the more traditional starch have been developed. The newer disintegrants are effective at much lower levels and comprise three groups: modified starches, modified cellulose and cross-linked povidone (Table. 21.6). Their likely mechanism of action is a combination of proposed theories including water wicking, swelling, deformation recovery, repulsion and heat of wetting. Superdisintegrants are so called because of the relatively low concentration required (2–4% w/w).
Table 21.6 List of superdisintegrants
The Compressing Weight
This is the weight of granules which must be fed into the die cavity to give a tablet containing the nominal amount of drug. The most direct method is to assay the granules and from the result calculate the compressing weight. Alternatively, this may be derived from data obtained during the manufacture of the granules and allows production of tablets to proceed without the delay due to the time taken to assay the granules. Whichever method is used, the finished tablets must be examined for medicament content, weight uniformity, etc., as discussed later.
Dry Granulation
The compressing weight is the sum of the weights of the individual ingredients in each tablet:
| Codeine phosphate | 20.00mg |
| Stearic acid, in fine powder | 0.06mg |
| Potato starch | 0.60mg |
| Spray dried lactose | 39.34mg |
| Compressing weight | 60.00mg |
As given, this formula would not be adopted for the commercial production of large batches, due to the possible darkening of the spray-dried lactose. It is suitable for the extemporaneous preparation of small batches. If a specified number of tablets must be supplied, an allowance must be made for losses incurred when adjusting the machine for correct tablet weight and hardness.
Preliminary Compression
Here, an allowance must be made for losses which occur when slugging and rescreening. Consider typical production data for 100,000 Soluble Aspirin Tablets BP:
| Per tablet (mg) | For 100,000 tablets (kg) | |
| Aspirin, in fine powder | 300 | 31.00 |
| Anhydrous citric Acid | 30 | 3.10 |
| Calcium carbonate | 100 | 10.33 |
| Saccharin sodium | 3 | 0.31 |
| Talc | 4 | 0.41 |
| Theoretical weight of ingredients | 437 | 45.15 |
| Weight of granules after slugging and rescreening | – | 44.30 |
| Weight of talc added prior to final compression | – | 0.80 |
| Total weight of granules for final compression | 45.10 |
Actual number of tablets which may be prepared from the slugged and rescreened in material is:
Compressing weight
Talc, rather than magnesium stearate, is used to avoid hydrolysis of aspirin and to improve powder flow. Disintegrant is not required as an effervescent base and finely powdered aspirin ensures rapid dissolution when the tablet is placed in water.
The use of anhydrous citric acid avoids premature reaction of the ingredients.
Moist Granulation
When calculating the compressing weight of granules prepared by this process it cannot be assumed that the starting materials are perfectly dry or that all moistening solvent is removed by drying. By keeping a record of the weight of the product at each stage of the granulating process calculation of the compressing weight is feasible. The following data might, be typical for the manufacture of, say, about 200,000 tablets each containing 0.5g of medicament:
| Medicament | 101.0kg |
| Adhesive granulating fluid | 10.6kg |
| Theoretical weight of mass | 111.6kg |
| Actual weight of moist granules | 111.1kg |
| Weight of granules after drying | 104.6kg |
| Weight of granules after screening | 104.3kg |
Formula of granules for compression:
| Screened granules | 104.3kg |
| Disintegrant | 8.0kg |
| Lubricant | 0.5kg |
| Weight of granules for tabletting | 112.8kg |
Number of tablets which could be produced from 101kg of medicament is:
Weight of moist mass per tablet:
Number of tablets that may be prepared from the actual weight of moist granules:
After drying 104.6kg of material contains the medicament present in 111.1kg of moist granules and 104.3kg of dried material is available for tabletting after screening. The number of tablets that may be produced from the screened granules is:
The final compressing weight after addition of disintegrant and lubricant is:
In the previous two examples large quantities of materials are being handled and losses can be kept to a minimum. Much greater losses are incurred when the moist granulation technique is applied to small batches; moistened mass which cannot be recovered from the granulating screen and tablets lost when adjusting the tablet press may represent a large proportion of the total batch. A loss of up to 10 per cent is possible in these circumstances and should be allowed for when calculating the weights of materials required to produce the batch.
Production of Tablets
The Single Punch Tablet Machine
Assembly
Clean the punches and die, set the machine by hand to the filling position, insert the bottom punch in its holder (Fig. 21.7), making sure that the locking notch faces the locking bolt, and lightly secure in position. Carefully fit the die, check that the die face is flush with the table and then securely tighten the locking bolt. The top punch may now be fitted, lightly secured, and this operation followed by the assembly of the feedshoe but not the connecting rod. Fill the feedshoe with granules, adjust the weight regulating collar so that the punch face is 0.5–1cm below the die table and then fill the cavity with granules. Carefully lower the top punch into the die cavity and at the same time adjust the hardness control such that the machine drive cannot be turned by hand past the compression position; maintain the load on the punches while their locking bolts are finally tightened. This procedure ensures that the punches are firmly locked in their holders and cannot work loose during tablet manufacture. Finally, reduce the compression control and eject the compact.
Fig. 21.7 Main operating parts of the single-punch tablet machine.
Adjustments
Tablet weight: For the filling and compression stages the lifting block pulls down the weight regulating collar and holds it in contact with the main frame of the machine. The position of the bottom punch face in the bore determines the volume of the die cavity and hence the tablet weight. For preliminary weight adjustment the collar is adjusted on its thread so that a weighed quantity of granules just fills the cavity. Final weight adjustment is made with the press running normally, as die filling is affected by the dynamic flow properties of the granules under operating conditions.
Ejection: After the compression stage the lifting block moves to its top position and raises the lower punch assembly by engagement with the ejection regulating collar. This is set to bring the punch face flush with the upper surface of the die. Too low a setting will cause part of the tablet to be sliced off as the feedshoe moves forward, while the punch will foul the base of the shoe if the setting is too high. The feedshoe is fabricated from a soft metal to minimize damage in the event of machine malfunction. A locking device is provided to make sure that the regulating collars cannot move during machine operation.
Tablet hardness: The top punch is locked in a holder which moves in a guide to ensure accurate registration of the punch and die bore. During operation of the press an eccentric cam imparts a vertical movement to the punch holder. Since the amplitude of the upper punch stroke and the centre of rotation of the cam drive are fixed by the machine design, die penetration is controlled by an adjustment which, effectively, alters the distance between the centre of rotation of the earn drive and the face of the upper punch at the bottom of its stroke. If this distance is increased the punch penetrates further into the die cavity; thus, for a given fill, the granules are more highly consolidated and a harder tablet is formed. The mechanisms used to adjust tablet hardness vary considerably and are best understood by examination of actual equipment. One method is shown in Fig. 21.7; as the holder is moved down, the adjusting thread punch penetration is increased.
Feedshoe movement: A rod connects the shoe to its actuating mechanism, transverse oscillation of the feedshoe relative to the die cavity being controlled by the length of the rod. This should be adjusted to ensure that the cavity is uniformly filled with granules, the top punch does not foul the shoe and the ejected tablet is pushed well away from the vicinity of the die. Once correctly adjusted, little further attention is required for successive production runs.
Operation
Once the machine has been assembled, trial tablets may be made with the feedshoe and the press operated by hand. When soft tablets of the correct weight have been obtained by adjustment of the appropriate controls, the feedshoe operating rod is connected and further tablets produced by hand.
If these are still rather soft, but otherwise satisfactory, power drive may be applied. Make any final adjustments to the controls bearing in mind that the press will ‘settle down’ after 20–30 tablets have been made. The production of soft tablets during the preliminary stages of adjusting the machine is necessary to obviate jamming due to the more efficient filling of the die under power drive.
The optimum tablet hardness depends on the material to be compacted and the ultimate use of the tablets. Uncoated tablets must be hard enough to withstand handling and yet not so hard that the disintegration time is unduly prolonged. Firmer tablets are required to resist the abrasion normal to the film or pan coating processes, but cores for compression coating must be somewhat soft to allow further consolidation when the coating is applied. A crushing test is used to ensure compliance with predetermined specifications for tablet hardness.
There is, obviously, a limit to the compaction force which may be safely developed by a given model of tablet press. Some are fitted with an adjustable overload release whereby the weight regulating collar rests on a spring loaded stop which is deflected downwards should the preselected compacting force be exceeded. In this way jamming and machine damage is avoided. The medium duty press shown in Fig. 21.8 is provided with means for varying the production rate from 40 to 90 tablets a minute and overload release for loads up to 4 tons (4000kg).
Fig. 21.8 Power-driven single-punch tablet machine. The top punch and the bottom punch holder are enclosed by a safety screen. The value at which overload occurs is indicated by the gauge bottom left.
It is essential that precautions are taken to preclude accidental connection of the power supply during assembly and initial adjustment of the press. As large compaction forces are developed, even by hand tablet machines, it should not be necessary to emphasize that the hands must be kept clear of the compressing area during production. On some presses an interlocking switch isolates the power supply until the moving parts are enclosed by a safety screen. The machine must never be operated so that the punch faces come into contact: if it is to be left for a short period the die cavity must be filled with granules. Finally, the feedshoe connecting rod must be removed and the shoe defected away from the die area, if it becomes necessary to operate the press in the reverse direction.
The Rotary Tablet Machine
The most important factor limiting the production capacity of a tablet press is the rate at which the die cavity can be filled with granules. For the single punch machine the limit is probably of the order of 8000 tablets/h but by mounting the punches and dies in a rotating holder (turret) which passes them in sequence through the filling, compression and ejection stages much larger outputs are possible. This is the principle of the rotary tablet machine (Fig. 21.9).
Fig. 21.9 Medium duty 16-station rotary tablet press. The upper turret and punches are visible beneath the plastic dust screen. Manesty Machines Ltd.
The top and bottom punches move freely in guides bored in the upper and lower sections of the turret (Fig. 21.10) while the centre section forms the die table. Cam tracks govern the positions of the punches by engaging with the head or the under surface of the mushroom head of the punch. The die cavity is overfilled with granules as it passes below a flat feedframes (Fig. 21.11) fed from a hopper. Tablet weight is regulated by an adjustable section of the lower earn track which, towards the end of the filling period, causes the bottom punch to rise slightly. The excess granules are removed by a scraper bar and retained in the feed frame. The method gives a more consistent die fill than would be obtained by setting the bottom punch at the correct level for the whole of the filling period. Weight adjustment is followed by a slight but sudden drop in the bottom punch level both to remove air by tamping action and lower the level of the granules in the die. This facilitates subsequent compaction and at the same time obviates spillage from the die cavity due to the centrifugal forces developed by the rotating turret.
Fig. 21.10 16-station rotary tablet press. The perspex screens and hopper and lower guard have been removed to show the upper and lower turrets and die table. The upper and lower cam tracks and the upper pressure roller are also visible.
Fig. 21.11 Close-up view of rotary press. The outlet of the hopper may be seen (left) feeding granules into the feed frame. Ejected tablets are being directed towards the collection chute bolted to the end of the feed frame. The upper cam track for lifting the top punch clear of the feed frame is also visible.
At the start of the compression stage the top punch is rapidly lowered to penetrate the die cavity and then both punches pass between pressure rollers which ‘squeeze’ the punches towards one another. The position of the upper roller is fixed and it is the adjustable lower roller that controls the separation of the punch faces at the point of maximum compression, that is, the tablet hardness. The lower roller is fitted with adjustable overload protection. Following compression, the top punch is raised rapidly to clear the feed frame and allow the tablet to be ejected as the lower punch rides up the lower cam track which may be adjusted to set the correct height of the punch for tablet ejection. The lower punch is then pulled down for the filling stage on the next revolution of the turret. A scraper bar directs the ejected tablet towards the collection chute. Any granules which have seeped out of the feed frame are carried round on the die table and returned to the frame behind the scraper bar. With 16 punch and die positions (16-station machines) the output is of the order of 50,000 tablets/h. Double rotary machines produce two tablets per revolution of the turret: outputs in excess of half a million tablets/h may be obtained with 69-station double rotary machines (Fig. 21.12).
Fig. 21.12 69-station rotary tablet machine. Manesty Machines Ltd.
It is inherent in the processing of particulate solids that dust will be liberated into the environment, a problem which is particularly acute with rotary tablet machines. With highly potent medicaments toxicity hazards must be considered, while interproduct contamination is an ever present problem in all tablet production processes. Modern rotary machines are provided with clear plastic covers over the moving parts and dust extraction. In difficult cases the only complete solution may be the provision of individual dust-extracted cubicles for each machine. Finally, as discussed earlier, it should be remembered that granules which ‘run’ well on a single punch machine may need reformulation for good performance on a rotary machine (and vice versa) due to differences in the modes of die feed and compression.
Developments in the Design of Rotary Tablet Machines
Granule Flow
The flow of granules into the feed frame is controlled by the height of the hopper outlet above the die table. A number of materials flow erratically or may even block the hopper outlet. Vibratory devices for attachment to the hopper offer a partial solution to these problems but where a very high flowrate is required—this may be as much as 2kg/min for ultra high-speed presses producing half a million tablets/h—more positive feeding is essential. The Stokes metering hopper minimizes blocking by allowing the granules to flow down a wide straight section at the base of the hopper on to an inclined plate which in turn directs the granules on to a metering wheel. The form of the latter, i.e. perforated disc, mesh or paddle, its rotational speed and the height of the baffle above the wheel, control the flowrate of the granules out of the hopper and into the feed frame.
Die Filling
Even flow of granules from the feed frame into the die cavity is essential if tablets of uniform weight are to be produced. Even at low operating speeds filling of the die with ‘difficult’ materials may be erratic with conventional feed frames. A number of devices have been designed to give ‘forced’, ‘assisted’ or ‘induced’ feed of granules into the cavity. The Manesty Rotaflow feeder employs contrarotating vaned rotors in an enclosed feed frame to assist granules into the dies. Some removal of air and slight compaction of the granules in the die cavity is obtained by the use of this feeder which is so designed that excess granules are automatically returned to the die-feed area. Gold et al. (1968) have reported that with a normal 16-station press the coefficient of weight variation for 500mg tablets was 0.7–2.4 per cent when a granule with good flow properties was used and 3.4–3.9 per cent for a poor granule. Using an ultra high speed machine with induced die feed they found that the coefficient of variation fell to 2 per cent or less irrespective of the type of granule, in spite of a lower tablet weight of 200mg and a rate of production (3000–4000 tablets min) at least five times that attained with the conventional machine.
Removal of Air
The time taken for compression may not be sufficient for the complete removal of air from the granules in the die cavity. The air which remains hinders intergranule bonding and may result in capping, lamination and other difficulties. Many rotary presses are now fitted with precompression rollers for light compaction of the granules ahead of the main compressing position. In this way most of the air is removed from the granule bed which is partially compacted before final compression takes place.
The introduction of devices for improving granule flow, die filling and the removal of air has made feasible the ultra high-speed production of tablets. These devices will, undoubtedly, find wide application in the direct compression of a wider range of substances than has hitherto been possible with conventional tablet machines.
Multilayer Tablets
One further development, not directly concerned with overcoming flow and compaction problems has been the introduction of equipment for making multilayer and compression coated tablets. The latter are discussed in the section on coating techniques. Layer presses are used where chemically incompatible materials, or those which separate in a mixed granulation, must be presented in the same tablet. For sustained action products each layer may be formulated for different release characteristics. The presses employ three filling and compression stages round the turret and as each layer is filled into the die cavity in sequence the granules are lightly compressed to remove air. When the final layer has been deposited the whole is compressed into a firm tablet in the normal way. Since each layer is thinner than in normal tabletting, fine granules (16mesh or smaller) must be employed. Excessive fines and lubricants must be avoided to promote good interlayer adhesion. A sampling device is provided so that the weight of each layer may be checked without stopping the machine. For the successful production of multilayer tablets the granules must not be too dense and should compact to less than half their original volume, otherwise the depth of fill is unmanageably small and difficult to control.
Punches and Dies
These are usually fabricated from special steels, the working surface being accurately machined and highly polished to ensure proper mechanical operation and well finished tablets. For single-punch machines the punches are held, firmly seated in their holders, by wedge action of the locking notches against the tips of the locking bolts; notches or circumferential channels serve the same purpose for the dies. For the manufacture of specially shaped products, such as pressaries, registration of the upper punch and the die must be exact. In the case of rotary tablet machines a key inserted in the guide prevents free rotation of the upper punch. The clearance between the upper punch and die bore is normally 0.0013–0.004cm (0.5–1.5 thousandth of an inch) but for slugging this is increased to 0.008cm to allow rapid escape of air from the granules. Fines seeping downwards between the lower punch and die bore would become highly compacted and impede free punch movement but for a slight reduction in the shank diameter some 0.5cm below the punch face.
In most English-speaking countries and some others, punch diameters were, until recently, quoted as multiples of a thirty-second of an inch, but metric sizes have been adopted by the British Pharmacopoeia 1973 and British Pharmaceutical Codex 1973. The majority of tablets are produced with flat, flat bevelled edge or normal concave punches (Fig. 21.13) but where tablets are to be pan or film coated deep concave punches must be used to facilitate the coating process. Punch faces may be embossed so that tablets with engraved letters or other devices are produced. Diametrical scoring permits more ready subdivision of the tablet by the patient should this be required. The load which may safely be applied to punches is limited by the material from which they are fabricated and by the type of punch face. A load of 8 × 107 kg/m2 (50 tons/in2) is permissible for flat or normal concave steel punches but must be reduced to half this figure, or less, depending on the punch diameter and face contour where the edge of the punch face is very thin as is the case with deep concave, bevelled edge and ball punches. This is because the lateral forces developed during compression would splay the thin weak edges of these types of punches. The maximum loads quoted in Table. 21.3 are average figures for the special steels used in punch manufacture; where other materials are employed the manufacturer of the punches should be consulted.
Fig. 21.13 Tablet profiles and structures.
Punches and dies (tooling) must be stored, lightly oiled, in containers which prevent accidental contact. The ease of manufacture and the final appearance of the tablet depend on unblemished, highly polished working surfaces; punch edges must be sharp and free from burrs. As compaction usually occurs in the middle section of the die bore, wear ‘rings’ are formed at this point and hinder tablet ejection. Inversion of the die for alternate tablet batches minimizes ringing and extends the life of the die. Tapered dies may facilitate the compaction of difficult materials, as the progressive increase in clearance (about 0.004cm in the centre to 0.008cm at the die face) permits more ready escape of air and gradual reduction of die-wall friction as ejection proceeds. Tungsten carbide inserts, having greater resistance to wear are particularly useful for the compaction of abrasive materials. Nonmetallic tooling may be required for the compaction of corrosive substances. Agate may be employed for this purpose, for instance in the production of tablets containing mercury salts. All stations of a rotary machine must produce tablets of uniform weight and hardness. Maintenance must, therefore, ensure that all punches in a given upper or lower set are of identical length as the weight, hardness and ejection controls are common to all stations· of a rotary.
The comprehensive range of tooling required for even a modest tablet manufacturing department represents a large capital investment which rapidly depreciates in the absence of proper maintenance. Additionally, the inferior tablets produced by defective punches and dies do little to enhance the reputation of the manufacturer. Swartz et al. (1962) and Swartz and Anschel (1968) indicate that extended life and improvements in the quality of the tooling supplied by the manufacturer are positive advantages which may be expected from a well-organized punch and die quality control programme.
Tabletting Problems
Tablets may exhibit a number of defects which are immediately apparent or, as is often the case with capping, appear only after storage. It is important therefore that the appropriate tests be applied at the start of and during a production run in order that remedial action may be promptly taken.
Binding in the Die
When this occurs ejection of the tablet is difficult and is often accompanied by a characteristic ‘grunting’ noise; the tablet edges are rough or scored. The problem results from the conditions that give rise to high die-wall friction, namely poor lubrication, underdried granules and a dirty or blemished die. Poor lubrication includes inefficient blending of the lubricant with the granules as well as the wrong quantity or choice of this adjuvant. An alternative source of trouble is too large a clearance between the lower punch and the die bore as a result of excessive wear. Fines seep downwards through the gap and compact to form a tough film which hinders free movement of the punch.
Picking and Sticking
Here the problem is caused by adhesion of material to the punch faces. If localized, portions of the tablet surface are seen to be missing; this is called picking. The tablet has a dull rough appearance when ‘sticking’ occurs due to adhesion of the tablet to the whole punch face. In either case, if not corrected, the defect worsens progressively and a layer of compacted granules builds up on the punch face. This, effectively, is equivalent to an increase in punch length that causes high compaction pressure and eventual jamming of the press if not promptly rectified. The fault can usually be traced to underdried granules, poorly maintained punch faces or the use of a lubricant lacking in antiadherent properties.
Capping and Lamination
These faults can often be traced to inadequate removal of air from the granules in the die cavity before and during compression. The entrapped air interferes with granule bonding while its subsequent expansion at the ejection stage detaches the top, the ‘cap’, of the tablet. In other cases the tablet splits into a number of layers (laminates). Excessive fines, incorrectly prepared or overdried granules, or too small a top punch/die-bore clearance—all hinder escape of air from the die cavity and may be the cause of capping or lamination.
A second cause of the problem is associated with undue elastic compression of the tablet due to the use of too high a pressure at the compaction stage. During ejection, elastic recovery of that part of the tablet protruding from the die gives rise to both lateral and longitudinal forces (Fig. 21.21) which rupture the intergranule bonds and the tablet then caps or laminates. Finally, on occasion, these faults may be traced to distortion of the tablet during ejection from a ‘ringed’ die or to the use of burred punches.
Excessive Weight Variation
This problem is associated with poor granule flow and separation of granule constituents. Thus granules which are underdried, too large (see wet screening, p. 305), too fine or contain a large proportion of fines, are incorrectly lubricated or comprise elements with widely differing densities or sizes, may all be suspected as possible causes of excessive weight variation. If the fault occurs with a rotary machine and granule flow appears to be satisfactory, then it is worth considering the possibility that one or more punches are of different length to the others due, for example, to inadvertent mixing of bottom punch sets. Occasionally, ‘difficult’ granules are produced in spite of care in processing and tablets of more uniform weight and improved appearance may be obtained at a slower machine speed that allows more time for die-cavity filling.
Fissured or Pitted Surface
If this is not due to sticking or picking the most likely cause is the presence of granules which are uniform in size and lack the fines necessary to fill the voids. Generally, the problem may be cured by the use of granules with a broader size distribution but care must be taken to see that this does not lead to other problems such as capping or lack of uniformity in the tablet weight.
Soft Tablets
Apart from the use of too low a compaction pressure this problem usually arises when granules have been inadequately dried or the intergranule bonds have been weakened either by traces of air in the granule bed (insufficient to cause capping) or by excessive proportions of fatty lubricant such as magnesium stearate.
Protracted Disintegration
In this case the tablet either rapidly breaks down to form large particles that persist for a long period, or fine particles are produced, but the overall disintegration time is excessive. The fault can usually be traced to:
1. An adhesive granulating agent which is too ‘strong’,
2. Conditions that inhibit the penetration of water; high degree of compaction, hydrophobic tablet ingredients, excessive quantities of fatty lubricant or gelling of the granulating agent, or
3. Inefficient bursting action resulting from the use of the wrong type of disintegrating agent, or insufficient of it.
If the granules are intrinsically hydrophobic or have become so due to the unavoidable use of a fatty lubricant, small amounts of surfactant may be included in the tablet formulation to facilitate penetration by aqueous fluids. Alternatively, a hard polyethylene glycol may be employed to provide ‘soluble’ points in the tablet structure. In suitable cases more efficient bursting action is obtained if part of the disintegrant is incorporated prior to the moistening or preliminary compression stages of granulation. Finally, as discussed later, it should be noted that too low a degree of compression may lead to a long disintegration time.
Mottled Tablets
These may be produced if the granule size distribution is discontinuous; the finer particles provide a background of slightly different hue which shows up the larger granules in the tablet surface. The effect is more marked with coloured tablets and is particularly noticeable if dye migration has occurred. If the problem arises when the tablets are stored, instability of one or more ingredients should be suspected.
Variation of Medicament Content
Standards will eventually be set for the drug content of individual tablets. If the content shows excessive variation and those factors contributing to weight variation can be eliminated as a cause, the most likely sources of trouble are migration of solute, physical adsorption with separation at some stage of the production process and inefficient mixing.
Drug Instability
This topic is discussed in detail elsewhere in this volume. The majority of air-dry drugs and tablet excipients contain water and this is also deliberately added, and may not be completely removed, in the moist granulation process. Moisture content must be controlled not only for purely technical reasons associated with the physical processes of tablet production, but also to ensure drug stability. The subject was reviewed by Griffiths (1969).
Tablet Coating
Tablet coatings perform one or more of the following functions. They may: mask the taste of unpalatable drugs, protect the drug from deterioration due to light, oxygen or moisture, separate incompatible ingredients, control the release of medicament in the gastrointestinal tract, and provide an elegant or distinctive finish to the tablet.
The materials used for coating may largely comprise sucrose (sugar coating), water-soluble film-forming polymers (film coating) or substances which are soluble in the intestinal secretions but not in those of the stomach (enteric coating). These types of coating can all be applied by the pan or fluid-bed processes; the compression coating technique is suitable for sugar and enteric coatings, but not for film coating.
Pan Coating
In this process the tablets are tumbled in a bowl or pan which rotates about an axis inclined about 30° to the horizontal (Fig. 21.14). With the correct pan load a three dimensional circulation is established and sufficient coating solution is added to wet the tablet surfaces. Internal baffles (Sutaria, 1968) or hand manipulation of the wetted tablets ensures that the solution is evenly distributed and a satisfactory tumbling action maintained while the coating is dried by a stream of warmed air. Small amounts of dusting powder may be applied to reduce tackiness and cohesion of the tablets during the drying stage. The volume of coating solution for each application is critical; inadequate wetting leads to irregular coating, whereas with too large a volume the tablets agglomerate and do not tumble well. The cycle of alternate wetting and drying is continued to build up a coating of the required properties. Initially, the tablets are subject to considerable abrasive action and for this reason should be more highly compressed than the corresponding uncoated tablets.
Fig. 21.14 Production model pan coating machine. The diameter of the bowl is about 1m. Courtesy: Vanguard Pharmaceutical Machinery
Sugar Coatings
This traditional coating imparts a smooth, rounded, elegant appearance to the tablet. Stephenson and Smith (1951) have given a detailed discussion on the composition of sugar coatings.
Sealing: After all dust has been removed from the tablets a dextrin, gelatin or acacia coat may be applied to ensure good coating adhesion. Where protection against the effects of water in subsequent coating solutions is required a 30–50 per cent solution of shellac in alcohol or other suitable solvent is employed for sealing, care being taken to avoid over generous application as this leads to a prolongation of the disintegration time.
Subcoating: To minimize the amount of material which must be used to round the tablet edges the cores are made on deep concave punches. The subcoat is built up in successive layers by wetting the tablets with an adhesive solution, dusting with filler and then thoroughly drying, as retained moisture may cause later cracking of the coat or deterioration of the core. The subcoating solution is usually an aqueous solution of sucrose to which is added acacia, gelatin or both, to impart adhesive properties. Talc and precipitated calcium carbonate are widely used in the subcoating filler together with some sucrose and a small proportion of acacia. A small proportion of inert filler, e.g. talc, may be added to the subcoating solution, which, if it contains gelatin, should be used warm to avoid gelling. Too heavy an application of filler must be avoided as the excess forms ‘granules’ with the coating solution and these interfere with the formation of a smooth subcoat. This stage of the process is continued until the tablets have a rounded appearance and the edges are well covered. When complete, the tablets are removed from the pan and thoroughly dried.
Smoothing and polishing: The application of a smoothing solution (60 per cent sucrose in water is satisfactory) causes limited wetting of the subcoat, which has been hardened by drying, so that it is smoothed out by the tumbling action. Soluble dye or a lake colour may be added to the smoothing solution if a coloured coating is required. As soon as the coat has become comparatively smooth the volume of smoothing solution per application should be reduced and the tablets dried without the aid of heat. In the final stages, tumbling is limited by ‘inching’ rather than rotating the bowl so that the coating is not scratched or otherwise damaged. At this stage the tablets will have a perfectly smooth but matt appearance and are thoroughly dried before polishing in a pan specially set aside for the job and coated with a beeswax–carnauba wax or similar waxy mixture.
Film Coating
Sugar coating doubles the weight and increases the size of a tablet and this is obviously undesirable if the tablet, in the uncoated state, is already large. Film coating provides an alternative means of masking the taste of the medicament and providing protection against adverse climatic conditions without significantly altering the tablet weight or size. The obvious advantages of excluding water from the coating process may be secured with film-forming polymers, such as ethylcellulose, polyvinylpyrrolidone or hydroxymethyl propyl cellulose, that are soluble in both water and anhydrous organic solvents. Some 3–10 per cent of the foregoing materials can be dissolved in an acetone–alcohol mixture together with 5–10 per cent of diethyl phthalate or other plasticizer to produce a film coating solution which may be applied by the pan technique. The plasticizer stops the film becoming brittle with age. Methylene chloride is often added to the solvent to reduce fire hazards, while undue absorption of expensive film forming agent is prevented by the prior application of a shellac sealing coat to the core tablets.
Enteric Coating
Tablets are enteric coated if the medicament is decomposed in the acid secretions of the stomach, if it causes gastric irritation, or if it is intended to exert its main effect only on the intestine. Some official tablets coated in this way include those containing biscodyl, bismuth and emetine iodide, and erythromycin. Enteric coatings resist the acid conditions of the stomach but readily dissolve in the more nearly neutral fluids of the small intestine. They are also used in the formulation of sustained action preparations as the release of medicament is delayed by the time taken for the tablet to pass from the mouth to the intestine.
Formalized gelatin, keratin, salol, shellac, sandarac, stearic acid and cetyl alcohol have all been used to produce enteric coatings but are either difficult to apply or erratic in their action. The compositions of the fluids in the gastrointestinal tract are not constant but vary with time and from person to person. It is clearly important that the enteric action shall largely be independent of such variations in compositions. Cellulose acetate phthalate is widely used for enteric coating and was reported by Antonides and DeKay (1953) to be the only cellulose derivative of the fourteen evaluated that was satisfactory for this purpose. Only one of the phthalic acid carboxyl groups is attached to the cellulose, the other being free for reaction. The polymer dissolves in a variety of solvents, gives water-soluble salts with a number of bases and forms coatings that are insoluble in, but slightly permeable to, water (Malm et al., 1951). An important feature is the solubility of cellulose acetate phthalate films in buffers having a pH as low as 5.8; thus the requirements for enteric action are a medium with a pH higher than this value which can contribute ions to the coating to form a soluble salt. Such conditions are found in the intestine but not in the stomach.
In addition to 5–15 per cent of cellulose acetate phthalate the enteric-coating solution may contain a plasticizer such as castor oil or butyl stearate together with a soluble or a lake dye if a coloured finish is required. The materials are dissolved or dispersed in a volatile solvent comprising alcohol, acetone, and, to reduce flammability, methylene chloride.
Automated Pan Coating
Successful use of the pan coating technique depends, in large measure, on the skill of the operator: it is also time consuming. Due to their higher volatility the solvents used for film coating permit rapid drying of tablets after each addition of the coating solution. As with moist granulation, spraying rather than pouring the coating solution sustained action product. Layer tablets provide an acceptable means for ingredient separation but the protective effect of the coat enveloping the tablet is lacking. It is not an easy matter to ensure a sugar coat of the same thickness for all tablets in a batch by pan coating and although reasonable variation in coat thickness and weight is not significant in plain sugar coating such variation cannot be tolerated if the coat contains a potent is a more controllable technique that relies less on the operator’s skills and which is more readily adapted to automatic control. In the case of enteric coated tablets, performance is determined by the thinnest part of the film and for reproducible characteristics between and within batches the coating must be of known uniform thickness. According to Lachman and Cooper (1963), enteric coatings complying with the foregoing criteria may be obtained by means of an automated film coating process. The tablets are tumbled in a baffled pan and a solution of cellulose acetate phthalate together with fillers in suspension is sprayed on to the tablets in repetitive ‘bursts’ alternating with longer drying periods. The spray gun for applying the coating solution and the hot or cold air flow for removing solvent are controlled by a punched tape programming device. The coating time for an 85 kilogram batch of tablets (90min) is half that required by the corresponding manual procedure, and the weight of the coating is reduced and is more uniform.
Fully automated equipment for pan coating is now commercially available. Provision is made for programmed control of the coating composition, spray application and drying air flow at each stage of the process. The different paths taken by tablets in a pear-shaped bowl leads to a lack of uniformity in the thickness of coating applied to tablets within the batch. This problem is avoided in the Manesty Autocota by the adoption of a cylindrical coating vessel rotating about a horizontal rather than inclined axis.
Fluid-Bed Coating
Wurster (1959) first described the application of sugar or film coatings to tablets suspended in an air stream. In that equipment the coating solution is introduced into the fluidizing airstream at the base of a tall vertical tube in which tablets circulate as they are coated. Evaporated solvent and air are removed at the top of the chamber. The fluid-bed moist granulation equipment described earlier has now been modified by the makers for fluid-bed coating. As the tablets circulate in the air stream they are subjected to considerable abrasive action and for this reason they are compressed with extra firmness on punches which give a well rounded profile.
Compression Coating
As noted earlier it is desirable on occasion to separate tablet ingredients to avoid incompatibilities, to facilitate manufacture or to produce a sustained action product. Layer tablets provide an acceptable means for ingredient separation but the protective effect of the coat enveloping the tablet is lacking. It is not an easy matter to ensure a sugar coat of the same thickness for all tablets in a batch by pan coating and although reasonable variation in coat thickness and weight is not significant in plain sugar coating such variation cannot be tolerated if the coat contains a potent medicament. The idea of applying granular coating materials to a preformed core was conceived by Noyes in 1896 but could not be commercially exploited until the problem of core centration and the automatic rejection of coreless tablets had been solved. The expected advantages of a compression coating process are:
1. Core tablets are not subjected to abrasive action and need not be especially hard,
2. The process is ‘anhydrous’ and sealing coats are not required,
3. The disintegration time of a coated tablet can be comparable with that of the uncoated variety, due to the above,
4. Good control of coating weight can be obtained,
5. Chemically incompatible ingredients may be separated,
6. The core and coat may be formulated for different properties, e.g. for a sustained action tablet, the coating would provide the prompt dose and the core the sustaining dose,
7. The process is continuous, completely mechanized and removes much of the element of skill from tablet coating.
Compression Coating Machines
The essential stages of coating by compression are (Fig. 21.15) deposition of the bottom fill of coating granules, transfer and centration of the core tablet, deposition of the side and top fill, and finally compression to bond the coat to the core. The machine described by Whitehouse (1954), the Kilian Prescoter, used preformed cores which were fed into holes on the periphery of a transfer disc and deposited on the lower fill of coating granules as the lower punch dropped in readiness for the top fill. The core was centred by a light tap of the top punch, the top fill deposited and the coating bonded to the core by compression. The force developed during compaction in the presence of a core was sufficient to cause slight deflection of the overload release but failed to do so in the absence of a core. A switch connected to the overload release provided an electrical signal which, when fed to a memory unit, actuated a gate on the collection chute such that coreless tablets were rejected.
Fig. 21.15 The compression coating of tablets.
The Manesty Drycota (Fig. 21.16) comprises two rotary presses coupled by a transfer unit (Fig. 21.17). The spring-loaded arms of the transfer unit engage with small collars on the upper turret to ensure accurate ejection of the core tablet into a cup fitted with a free sliding weighted plunger. Next, the cup passes over a ‘bridge’ where dust is removed to avoid contamination of the coating granules and then engages with the collar on the upper turret of the coating press. As the bottom punch drops, the tablet is pushed out of the cup on to the centre of the bottom fill of coating granules by the action of the weighted plunger. The cavity is then filled with granules for the sides and top of the coating, which are bonded to the core by compression. As the cups pass round the transfer unit the plungers are examined by feelers which operate microswitches. If a cup fails to pick up or deposit a core at the appropriate stage in the cycle the signal from the microswitches initiates action for the rejection of coreless tablets.
Fig. 21.16 Compression coating machine. The core tablet formed by the left hand unit is carried by the central transfer unit to the coating press or the right.
Fig. 21.17 Essential features of transfer mechanism of a compression coating machine (for clarity, only the die tables of the core and coating units and three transfer arms are shown).
Requirements for Core Tablets and Coating Granules
Core tablets should be prepared from rather large granules and should be lightly compressed. The surface of a soft tablet made from large granules is somewhat porous and provides a good ‘key’ for adhesion of the coating. There is a second reason for light compression of the core tablet. Residual stresses in a tablet are not instantaneously relieved when that tablet is ejected from the die. Large residual stresses induced by a high degree of compression may be sufficient to split the coating some time after manufacture has been completed.
The gap between the core and die wall should be at least 0.13cm to facilitate uniform deposition of coating granules round the edges of the cote tablet and hence to ensure adequate coating strength. For these reasons too, the granule size should not exceed a quarter of the edge coating thickness, On the other hand, excessive amounts of fines or fatty lubricants should be avoided, as these lead to poor coating strength and poor adhesion to the core.
Performance
Provided the requirements for edge thickness are met, the weight of the coating is maintained with considerable precision and may be varied over a wide range. It may be noted that the official test for weight variation applies to compression coated tablets but not to those having a film or conventional sugar coat. Centrifugal and tangential forces due to turret rotation may slant the bottom fill of coating granules and displace the core from its central position. Lachman et al. (1966) concluded that coating granules which gave minimum weight variation also tended to give less core displacement.
Coatings may be formulated for a wide range of properties but of these, enteric coatings are of particular interest. Blubaugh et al. (1958) have described the preparation of coating granules containing triethanolamine cellulose acetate phthalate which, when applied by compression, produce a coating with enteric properties, In a more recent paper Srinivas et al. (1966) have reported on the use of a number of carboxylated polymers for the production of entire coating granules.
The Quality Control of Tablets
Formerly, tablets were considered satisfactory if they were elegant and firm enough to withstand handling without damage. The realization that such criteria were inadequate for the scientific control of tablet quality led to the inclusion of suitable standards and tests in all national pharmacopoeias. Although corresponding procedures differ in detail or emphasis they serve an identical purpose which may, in general, be understood by a consideration of the requirements of the British Pharmacopoeia 1973. These:
• define and describe compressed and moulded tablets, chocolate basis and tablet coatings,
• restrict flavouring agents and coatings to those products for which specific permission is given for their use in the monograph,
• standardize uniformity of tablet diameter and weight, content of medicament and disintegration time, and
• prescribe storage and labelling requirements,
• and are intended to ensure, not only a fully potent product, but also uniformity in the physical characters of official tablets regardless of source. The latter is essential as the patient quickly perceives, and is concerned about, differences in the appearances of nominally identical tablets.
Uniformity of Diameter
The standard specifies the punch diameter to be employed for a given tablet and the permitted tolerance on the specified size, thus securing dimensional uniformity between tablets of the same kind from different manufactures. The standards apply only to uncoated and compression coated tablets. They make allowance for the manufacturing tolerances in die fabrication and for tool wear by allowing a deviation about the stated diameter. Generally, the manufacturing tolerance is proportionately less for large than for small dies and the deviation is reduced from ± 5 per cent to ± 3 per cent for diameters in excess of 12.5mm.
Tests for Uniformity of Weight and for Medicament Content
The combined effect of these tests is to ensure that all tablets in a batch are, within reasonable limits, of the same potency. A perfect manufacturing procedure would yield a batch of tablets having identical weight and medicament content; in practice the values of these parameters for individual tablets deviate about the mean values for the whole batch. Such deviations will be of acceptable magnitude if correctly formulated and prepared granules are compressed on properly maintained equipment but will be larger, and probably unacceptable, if the converse conditions apply.
Ideally, the quality of a batch of tablets would be assessed by determining the potency of each tablet in a truly representative sample. The analysis of a large number of tablets by conventional means would be both costly and time consuming while automated techniques for the analysis of the majority of drugs have still to be developed. At the present time therefore the reasonable assumption is made that the variation of the weight of individual tablets is a valid indication of the corresponding variation in the drug content
Test for Uniformity of Weight
This test does not apply to sugar or enteric coated tablets. The weight of each tablet and the average weight of all tablets in the sample are determined. The standard, summarized in Table. 21.7, limits the number of tablets in the sample which may deviate from the average weight by more than specified amounts. Rogers (1956), on the basis of production data then available, showed that the coefficient of variation of tablet weight was reasonably constant for tablets weighing more than 0.2–0.3g but rose steeply for smaller tablets; a similar observation has been made by Smith et al. (1963). This is not unexpected as there is greater practical difficulty in ensuring the flow, die filling and air release characteristics necessary for low weight variation, with the finer granules which must be used for the production of small tablets. This factor is recognized by making the weight variation tolerances wider for small than for large tablets.
Table 21.7 Summary of the BP 1973 test for the uniformity of tablet weight
Twenty tablets represent a very small fraction of a manufacturing batch that will generally exceed 10,000 and may be as large as 1,000,000. If the tablet weights in a batch of, say, 0.3g tablets are normally distributed and 10 per cent deviate from the average weight by more than 5 per cent, then 0.1 per cent of the tablets in that batch could be expected to deviate by more than 10 per cent from the average weight Smith (1955) has calculated that there is a 67 per cent probability that the examination of a sample of twenty tablets by the official method would lead to the acceptance of such a batch. With a rotary machine, if the weight variation derives from a few punches of significantly different length from the others in the set, the foregoing calculations that are based on an assumed normal distribution do not apply. Nevertheless, the unsatisfactory nature of a batch of tablets produced under these conditions should be relaxed as indicated in Table. 21.8. This table may intend to detect abnormal weight variation. While the stringency of the official test could be improved, for instance, by specifying narrower weight tolerances, it is important to remember that the discrimination level must provide a reasonable balance between the risks to the patient arising from over or under dosage and the performance that can reasonably be expected with the currently available manufacturing methods. Although the manufacturer must ensure that the completed batches of tablets comply with official requirements, samples would be examined during production as an extra check on quality and for the early detection and rectification of any faults which may develop.
Table 21.8 BP 1973 Standard for the content of medicament in tablets
Content of Medicament
The standards are framed to take into account processing difficulties, variations in the purity of the drug, accuracy of the assay methods and the size of the sample relative to that of the typical manufacturing batch. The stated limits for drug content given in the individual monographs apply to the examination of a sample of 20 tablets; where less than this number is available the sample must comprise at least 5 tablets and the limits are relaxed as· indicated in Table. 21.7. This table may be used directly where the limits specified in the monograph are between 90 and 110 per cent. Thus, for a sample of 20 Promazine Tablets each containing 25mg of Promazine Hydrochloride the monograph limits are 92.5–107.5 per cent of the nominal content of drug and would be widened in the case of a sample of 5 tablets to 90.9–109.3 per cent. To comply with official requirements the tablets must therefore contain between 22.73 and 27.33mg of drug.
Where the monograph limits fall outside the 90–110 per cent range the Pharmacopoeia directs that proportionately larger allowances should be made for samples of less than 20 tablets. A method for the calculation has been suggested by Hadgraft (1945).
It is now recognized that although a sample of tablets may comply with given standards for drug content, this may not be so for individual tablets within that sample. From theoretical considerations, Train (1960) concluded that if 20 tablets are examined and as a whole comply with ±10 per cent limits, a variation as large as ±40 per cent might be expected if each tablet was separately assayed. This variation could be due to solute migration and adsorption, granule separation and other effects. As discussed earlier, such variation is more likely to occur where the weight of the medicament is only a small proportion of the total tablet weight. Even, a reduction in the proportion of medicament from 90 per cent to 23 per cent doubled the standard deviation of drug content for the tablets examined by Smith et al. (1963). On the other hand, Robinson et al. (1968) failed to detect a deviation greater than ±15 per cent when examining over 1000 imipramine and desipramine tablets and capsules. In a number of cases the United States Pharmacopoeia includes a standard for the drug content of individual tablets; Feldman (1969) has indicated that the number of products affected by this standard will be increased in the future. The BP 1973 requires that, for tablets and capsules described in its monographs, no gross deviation from the stated amount of medicament is permissible when each dose unit is individually assayed. The term gross deviation is not defined. If a formal standard for drug content per tablet is included in future editions of the Pharmacopoeia, automated analytical techniques of sufficient rapidity, sensitivity, accuracy and precision will be required. It is evident from the review by Kuzel et al. (1969) that considerable progress has been made in the field of automated analysis since 1960 when Stephenson concluded that for Digoxin tablets the methods of mixing were probably superior to the then available assay methods.
Test for Disintegration
Methods
Features common to the tests described in the literature are: an aqueous disintegration medium, agitation of the tablet or medium to simulate peristalsis and means for recognizing the end point, that is, complete disintegration. The many proposed methods may be classified in to three types:
1. The tablet, supported on a frame, is immersed in water and the time determined for a weighted wire to pass through the tablet (Berry, 1939). The method was abandoned because it gave anomalous results if the tablet disintegrated into large particles or yielded a hard core (Berry & Smith, 1944). This structure on the method might not be so relevant at the present time when, for efficient drug release, it is considered desirable that a tablet should disintegrate completely into very fine particles.
2. A tube filled with water except for a small headspace and containing the tablet, is rotated end over end (Berry & Smith, 1944; Mapstone, 1952).
3. A wire mesh cage (Hoyle, 1946) or a tube closed by a wire mesh at the lower end (Prance et al., 1946) is moved up and down in the disintegration medium so that the tablets are constantly agitated. Disintegration is complete when all particles from the disintegrated tablet pass freely through the mesh.
The reader is referred to the BP 1973 for details of the official apparatus which, essentially, is that described by Prance and coworkers. Tablets may fail the test because, for instance, they form a gummy mass in contact with water or a resistant core is left after the rest of the tablet has disintegrated. If tablets fail the test for any reason, it may be repeated with a guided disc (Fig. 21.18) placed in the disintegration tube above the tablet sample. The light percussive action of the disc as it rises and falls in the tube assists the breakdown of the tablets.
Fig. 21.18 Essential features of apparatus for the BP 1968 disintegration test showing guide disc in use.
Requirements
Uncoated and compression coated tablets: The general requirement is that these shall disintegrate within 15min. There are a number of exceptions. Tablets which are to be dissolved in water prior to administration, e.g. Sodium Citrate, Soluble Aspirin, must disintegrate within 3min. Calcium Lactate Tablets harden on storage and 30min is allowed for these. The same period is specified for Thyroid Tablets and Prepared Digitalis Tablets as these form a gummy mass which is not easily persuaded through the mesh even with the aid of a guided disc. The test does not apply to products which should dissolve slowly in the mouth, e.g. lozenges, Compound Sodium Bicarbonate Tablets, Glyceryl Trinitrate Tablets, or to those which must be masticated, e.g. Phenolphthalein, Aluminium Hydroxide, or to Effervescent Potassium Tablets BPC. The official apparatus is not used for the latter; they are required to dissolve completely within 3min when placed in water at 20°C.
Sugar and film coated tablets: As these coatings may increase the disintegration time, 1h is allowed. One exception is Quinalbarbitone tablets, which must be sugar coated to mask the bitter taste of the drug. This hypnotic acts very promptly and therefore the tablets are required to disintegrate within 15min.
Enteric coated tablets: The official apparatus is used but the test is carried out in two stages. It is first ascertained that the coating should not rupture in the stomach by subjecting the tablets to the action of 0.6 percent v/v hydrochloric acid solution for 3h. If the coating is intact at the end of this period the tablets are washed to remove adherent acid and must then disintegrate within 1h when a pH 6.8 buffer is used as the disintegrating medium. Except in the case of achlorhydric patients the pH of the stomach is normally very low but the composition of its secretions varies considerably between individuals, in any one individual throughout the day and from day to day. Similar variations are found in the contents of the duodenum and small intestine which are often not as rich in enzymes or bile salts or of as high a pH as was implied by the composition of the alkaline pancreatin solution specified by the BP 1963 for the second part of the test. A number of enteric coatings used in the past depended on enzyme action or a rather high pH and proved erratic in their action. The test solutions prescribed by the BP 1968 have been adopted with a view to the elimination of such coatings.
Comments on the Test
The choice of water as the disintegrating medium is open to criticism. Among others Abbott et al. (1959) observed that the addition of mucoid substances to the test solution had a marked effect on disintegration time, which was often doubled and on occasion increased by a factor of 16. However, as Hartley (1950) has commented, the selection of 15min for the disintegration time of uncoated tablets is itself somewhat arbitrary and may be quite unrelated to the rate at which the drug is absorbed. Generally, prompt action is required but the optional coating of, for example, Benzylpenicillin Tablets leads to the anomalous situation that the permitted maximum disintegration time may be 15 or 60min depending on whether uncoated or coated tablets have been supplied to the patient. Benzylpenicillin tablets (Ashby et al., 1954), those of Quinalbarbitone and a number of others (Howard, 1951) can be sugar coated without extending the disintegration time beyond 15min. The coating methods used for these tablets could, presumably, be more widely employed with the consequent potential advantage of more prompt therapeutic action. It is well known that although a tablet may comply with official requirements, the particles freely passing the mesh of the disintegration tube may remain intact for periods in excess of 15min. Furthermore a number of factors, such as the physical state of the medicament used to produce the tablet, may have a marked effect on therapeutic response. For these reasons it is probably the dissolution rather than the disintegration behaviour which should be tested.
Colour
Rolfe (1956) has argued the case for and against the addition of colorants to tablets. Such arguments were, presumably, involved in the decision to remove all former restrictions, in the BP 1968, but the BPC 1973 still prohibits the use of colours except for specific tablets and those which are enteric or compression coated. The colour to be employed for a given tablet is not stated in either book of standards. Even if the proportion of a named colorant were to be specified for a given tablet, considerable difficulties would still arise when attempting to standardize the final shade and depth of colour as these are markedly affected by processing conditions. The problem is obviously more acute if the same product is to be produced by a number of manufacturers.
Hardness
Although there is no official test for tablet hardness this property must be controlled during production to ensure that the product is firm enough to withstand handling without breaking, chipping or crumbling, and yet not so hard that the disintegration time is unduly prolonged. Most testers apply a load to the edges of the tablet across a diameter. The load is gradually increased until the tablet fractures, the value of the load at this point (the crushing force) giving a measure of the hardness of the tablet. In the Monsanto Hardness Tester (Fig. 21.19) the tablet is held between a fixed and a moving jaw. The force applied to the edge of the tablet is gradually increased by turning the compression screw. The body of the instrument carries an adjustable scale which can be set to zero against an index mark fixed to the compression plunger when the tablet is lightly held between the jaws. With the Strong Cobb tester the load is applied pneumatically, while a pliers mechanism is employed in the Pfizer tester. To minimize errors due to the small force required to hold the tablet in position at the start of the test and the rate at which the load is applied, the mean of several hardness determinations should be used. For tablets of fixed dimensions in an experimental series the method provides comparative estimates of hardness. On the other hand, the optimum hardness for a given tablet in the manufacturing situation can only be set by experience extending over a number of batches. A tablet which was ‘soft’ and just handleable would give a reading of l–2 kilograms on the Monsanto tester; for highly compacted tablets the reading would be 6 kilograms or more. The performance of a number of commercially available hardness testers has been reviewed by Brook and Marshall (1968) and by Ritschel et al. (1969).
Fig. 21.19 The Monsanto tablet hardness tester.
Tablets are usually subjected to a second type of test which, essentially, measures the surface hardness. The Vickers test measures the size of the indentation produced on the tablet surface by a specially shaped loaded point. More usually the tablets are tumbled in a controlled manner and the change in the average weight of the tablets determined after a specified length of treatment. Such a test provides an indication of the likely edge damage that would occur when the tablets are handled during packaging and dispensing. The equipment designed by Webster and Van Abbe (1955) uses a vertical jogging action, whereas in the Roche Friabilator the tablets drop through a fixed distance during each revolution of a horizontally disposed plastic drum. The permissible loss in weight is obviously set by production experience, but with the Friabilator is usually in the range 1–5 per cent for 10min treatment. Incipient capping and lamination usually show up on a tumbling test.
Theoretical Studies in Tabletting Technology
Although a great deal of information on the practical aspects of tabletting had appeared in the literature prior to 1952, very few papers had been concerned with theoretical studies. In that year, however, Higuchi and his associates (1952, 1953, 1954a & b) published the first in a series of papers relating to the physics of tablet compression. Subsequently, as may be seen from the extensive bibliographies of Evans and Train (1963, 1964), many workers have reported the results of studies in this field.
The Compaction Process
Both tablet machines and ram presses have been used to investigate the compaction behaviour of pharmaceutical substances. Pressures in the range 300 to 3000kg/cm2 are usual for tablet machines, but the upper limit may be higher for mechanical or hydraulic ram presses. The use of flat-faced punches simplifies subsequent treatment of the data. Where the time taken for compression is short, as with the tablet machine, transducers must be used to convert the compaction forces, punch displacements, etc. into electrical signals which can be recorded, usually by an ultraviolet oscillograph. The necessary instrumentation has been described by (Higuchi et al. 1952, 1954a), Shotton and Ganderton (1960a), Shotton et al. (1963), Goodhart et al. (1968) and by other workers. Fig. 21.20 shows typical results that are obtained using an instrumented tablet machine.
Fig. 21.20 Punch displacements and forces developed during tablet compression.
The downwards thrust of the top punch in the die bore during compaction causes movement of material relative to the wall of the die. The movement is resisted by frictional effects, in part within the bed of material but mainly between that material and the die wall. As a result, the maximum force transmitted to the lower punch (Fb) is less than the maximum force (Fa) applied by the top punch to the contents of the die cavity by an amount Fd, that is the force ‘lost’ to the die wall. As compaction also squeezes the compacting material into close contact with the wall of the die this latter experiences a maximum radial force Fr. Even when the top punch has been removed from the die, residual stresses within the tablet give rise to a load normal to the die wall and therefore an ejection force, maximum value Fe, must be applied to the compact by the lower punch to overcome friction and initiate ejection. The extent to which the material in the die has been compacted may be evaluated from the relative volume (Vr), relative density (ρr) or porosity (ε) of the tablet. By loading the edges of the tablet across a diameter the minimum crushing force (Fc), just sufficient to rupture the tablet, may be found. If, as is often the case, the tablet dimensions are maintained in an experimental series, Fc provides a comparative measure of tablet hardness. The foregoing parameters, together with a number of others commonly used in tabletting theory, are defined in Table. 21.9
Table 21.9 Definitions of some parameters used in tabletting theory |
||
| Symbol | Definition | Derivation |
| A | Area of tablet in contact with die wall | πDL |
| Ap | Area of punch face | πD2/4 |
| D | Diameter of punch face | |
| Fa | Maximum upper punch force applied to compact | |
| Fb | Maximum force transmitted to lower punch by compact | |
| Fc | Crushing force applied across a diameter to edges of compact, just sufficient to cause rupture | |
| Fd | Force lost to the die wall | Fa – Fb |
| Fe | Maximum force applied by lower punch to initiate ejection | |
| Fr | Maximum radial force at die wall during compaction | |
| Fr | Force normal to die wall during ejection | |
| kd | Die wall coefficient of friction during compaction | |
| ke | Die wall coefficient of friction during | |
| L | Observed length of compact | |
| L0 | Length compact would assume if compressed to zero porosity | W/Apρ0 |
| Pa | Maximum upper punch pressure | Fa/Ap |
| Pb | Maximum lower punch pressure | Fb/Ap |
| Pm | Mean compaction pressure | (Pa + Pb)/2 |
| Pr | Maximum radial pressure during compaction | Fr/A |
 |
Radial pressure during ejection of compact |  |
| R | Punch force ratio | Fb/Aa |
| T | Transmission ratio | Fb/A |
| V | Observed volume of tablet | Ap/La |
| V0 | Volume compact would assume if compressed to zero porosity | Pr/Pa |
| Vr | Relative volume of compact | Ap/La |
| W | Weight of tablet | W/ρ0 |
| ε | Porosity ratio of void to total volume | V/V0 |
| ρ | Observed density of compact | (V−V0)/V, 1 − ρr |
| ρ0 | True density of material used in compaction study | W/V |
| ρr | Relative density of tablet | ρ/ρ0, 1/Vr |
It is assumed that the punches are flat faced and circular in section. |
||
By finding the values of the parameters described in the previous paragraph at a number of preselected maximum upper punch pressures the compaction behaviour of a substance may be studied. Higuchi et al. (1953) determined these values for sulphathiazole compacts made in a ram press. In each case the upper punch pressure (Pa) was raised to its preselected maximum value over a period of 20–60 s. While the logarithm of the disintegration time was found to be proportional to Pa, the crushing force (Fc) was linearly related to log Pa. Thus, the compaction pressure had a much greater effect on the disintegration time than on hardness as is shown in Table. 21.9. The porosity of the tablets progressively decreased as the pressure was increased but the specific area rose to a maximum at about 1600kg/cm2 and subsequently decreased. This was attributed to fragmentation of granules during the initial stages of compaction followed by bonding of the freshly formed surfaces at higher pressures. Further studies with aspirin, lactose and sulphadiazine (Higuchi et al., 1954b) gave results qualitatively similar to those obtained with sulphathiazole.
Train (1956) compacted alternate layers of normal and coloured magnesium carbonate. When the die wall was lubricated with graphite the layers compressed evenly, whereas, in the absence of the lubricant, the layers curved downwards from the wall of the die towards the centre of the compact indicating that friction had restricted movement of the particles near the die wall. Due to intense local shear action a hard ‘skin’ of distorted strongly bonded particles was formed where the compact had been in contact with the die. This ‘skin’ is often formed on tablets and provides resistance to abrasive damage. The degree of compaction was assessed by measuring the thickness of the layers in each region and was found to be higher in a peripheral ring near the top punch and in a lower central region (Fig. 21.21). The latter may explain why the ‘core’, often seen towards the end of a disintegration test, breaks down more slowly than the rest of the tablet. A plot of Vr as a function of log Pa was not a smooth curve but showed several abrupt changes of slope (Fig. 21.22). The initial decrease in Vr was due to closer packing of the magnesium carbonate but this was eventually limited by the available space in the die cavity. Thereafter, the slope increased abruptly and during the second stage the load was supported by temporary structures of compacting material. At still higher pressures the structures collapsed and the material failed by crushing and plastic flow (third stage) until; in the fourth stage, there was sufficient rebonding of the freshly created surfaces to give a compact with enough strength to involve the normal compression characteristics of solid magnesium carbonate in any further reduction of volume. The data of (Higuchi.et al. 1953, 1954b) and Shotton and Ganderton (1960a), when plotted in the form used by Train, show hat the third and fourth stages of compaction also occur with aspirin, lactose, sulphadiazine, sulphathiazole and sodium chloride. In the cases of the first four substances the abrupt changes in the slope of the log Pa/Vr plot occur at approximately the same compaction pressures as the maxima in the specific surface area curves. This would be expected if, as Train suggested, fragmentation and rebonding are competitive processes with the latter predominating in the fourth stage of compaction.
Fig. 21.21 Physical conditions during tabletting.
Fig. 21.22 Stages in the compaction of magnesium carbonate.
Tablet Hardness
Effect of Compaction Pressure
The behaviour of granules during compaction, the extent to which they bond together and the strength of the intergranule bonds relative to the strength of the granules determine tablet hardness. Strickland et al. (1956), who used activated carbon to delineate the granule boundaries, found that the carbon did not penetrate the interior of the granules even though these were laterally deformed by compaction. Shotton and Ganderton (1960b) applied a vividly coloured sucrose coating to preformed spherical granules of that material. The coating remained intact and the granules largely retained their shape at low compaction pressure but as this was increased the granules distorted progressively until, at the highest compaction pressures, the coating had ruptured to such an extent that it appeared as coloured fragments against the white underlying sucrose. The strength of the intergranule bonds formed by light compression was less than that of the granules with the result that the tablets failed along, rather than across, the granule boundaries when submitted to a crushing force. When this was applied to the more heavily compressed tablets the bond strength was large enough to permit the development of faults within the granules which spread across the boundaries and led to failure of the tablet. The rebonding of surfaces formed by fragmentation, rather than granule interlocking, was apparently the more important factor determining the hardness of these tablets.
Effect of Materials Used and Tablet Dimensions
For given compaction conditions the intergranule bond strength is highly dependent on the nature of the base material. Strong bonds offer resistance to the propagation of faults across the grain boundary in which case, according to Orewan (1949) and others, the strength of the compact formed at a particular compaction pressure should be greater with fine than with coarse granules. This effect was noted with sodium chloride and hexamine compacts (Shotton & Ganderton, 1961). Aspirin granules, however, formed weak bonds and softer tablets and, since fault propagation was not involved, granule size had little effect on tablet hardness. Coating sodium chloride and hexamine granules with stearic acid to weaken the intergranule bonds, reduced the hardness of the compacts, and abolished the particlesize effect observed with hexamine, but in the case of sodium chloride the hardest tablets were formed by the coarsest granules. Shotton and Ganderton suggested that the particle to particle contact pressure was greater in coarse sodium chloride grains than in fine ones since, for a given weight of that substance, fewer contacts were available to support the applied load. The higher contact pressure with the coarsest granules allowed penetration of the stearic acid film, the formation of strong bonds and, consequently, the production of the hardest tablets.
The majority of tabletting studies have used a constant die size for an experimental series and the thickness of the tablet, over a range of compaction pressures, has been maintained at a particular value by varying the compressing weight of the base material. Rees and Shotton (1969) reported preliminary studies on the effect of tablet dimension on the crushing strength. For sodium chloride their data indicated a relation of the form:
(21.1)
where D is the tablet diameter, Lo the length the compact would assume if compressed to zero porosity and kc and Cc are constants.
Capping and Lamination
Granule Bonding and Relaxation Stresses
Occasionally, at the start of a disintegration test, small radiating cracks appear on the upper face of the tablet and small bubbles of air are released. If air does not freely escape from the granules in the die cavity, for example, where the fines content is high, the force created by the expansion of entrapped air may be sufficient to disrupt the bonds between the ‘cap’ and the rest of the tablet when the top punch is removed at the initiation of the ejection stage although, as will be seen later, capping and lamination can occur under conditions which preclude the implication of air as the causative agent. It is well known that too high a compaction pressure may lead to capping and lamination and has been shown to occur, for example, with hexamine (Shotton & Ganderton, 1961) and sodium chloride (Rees & Shotton, 1969).
In 1956, Train pointed out that on removal of the compacting force the elastic recovery of the dense peripheral ring (Fig. 21.20) would be larger than that of the adjacent, less dense, part of the tablet. The differential stress in this region is aggravated by both longitudinal and radial relaxation of that part of the tablet extruded from the die during the initial stages of ejection (Fig. 21.20). Both effects would increase the tendency to capping and lamination. It may be noted that the increase in tablet dimensions that is due to elastic recovery is usually sufficient to prevent re-insertion of the tablet into the die from which it has just been ejected. Clearly, the strength of the granule bonding may be critical in determining whether the tablet will cap or laminate. According to Shotton and Ganderton (1961) the bonds in sodium chloride compacts were strong enough to support relaxation without laminar failure although this can occur if extreme compaction pressures have been employed (Rees & Shotton, 1969). The weaker bonds in aspirin compacts allowed the relaxation stresses to be dissipated by bond weakening and partial separation of the granules within the tablet. Undoubtedly, some loss of hardness must have occurred but the bonding was still sufficient to give a firm tablet. Hexamine tablets capped and laminated by fault propagation across the granule boundaries. This was not due to entrapped air as compacts prepared in vacuo behaved in a similar manner. Coating the hexamine granules with stearic acid eliminated the fault entirely and this was attributed to the bond weakening effect of the fatty acid which permitted dissipation of the relaxation stresses by the mechanism postulated for aspirin.
Residual Die-Wall Pressure
Practical experience has shown that some tablets do not cap or laminate immediately, but may do so on storage. Higuchi et al. (1965) studied the decay of die-wall pressure for a number of materials. In addition to an immediate elastic recovery in the die there was also a residual die-wall pressure (Table. 21.11) which decayed slowly over a period of a minute or more, during which time the tablet ‘flowed’ in a direction opposite to that in which the compacting pressure had been applied. Addition of lubricant lowered the residual die-wall pressure due, presumably, to the ‘softer’ nature of the tablets (see the values for crushing strength) and to reduced die-wall friction. The fact that a radial pressure persisted at the end of the decay period showed that there was an elastic deformation which could only be relaxed by ejection of the tablet from the die. Internal strains within the tablet might also provide a further source of stress energy which could be dissipated at any time during the life of the tablet. For single punch and rotary tablet machines, upper punch pressure release and ejection take only a fraction of a second. Even if a tablet with a tendency to cap survived the rapid elastic recovery stage on extrusion from the die, the subsequent slower relaxation may cause capping when the tablets are handled and packed. Incipient capping and lamination of this type is usually apparent if a few tablets are shaken in the cupped hands, dropped on to a hard surface or subjected to a friability test.
Table 21.10 Effect of compaction pressure on the properties of sulphathiazole compacts
Table 21.11 The relation of the hardness of some pharmaceutical substances to their tabletting behavior
Hardness of Materials and Tablet Properties
Prior to compression, a granule porosity exceeding 60 per cent would not be unusual. As shown in Table. 21.10, porosity is reduced to about 20 per cent with light compression and decreases to 5 per cent or less at much higher compaction pressures. The reduction in porosity is due to granule fragmentation giving smaller particles which may be more closely packed, and plastic deformation which allows the granules to ‘flow’ into the void spaces. The contribution of each process depends on the hardness of the substance being compressed. If it were possible to compress a liquid in a tablet machine Pa, Pb and Pr would be equal, i.e. the transmission ratio would be unity. Rubber and silicone putty show ‘liquid’ behaviour and are used for die-wall pressure calibration. For other materials the transmission ratio will be less than unity. When sodium bicarbonate or sulphathiazole were compressed, Nelson (1955) observed that about a third of the upper punch pressure was transmitted to the die wall. Windheuser et al. (1963) distinguished three distinct stages in the development of die-wall pressure with increasing values of upper punch pressure. The initial steep rise of Pr was due to the reduction of void space by the removal of air. This was followed by a less rapid rise in Pr during which granule fragmentation and consolidation took place. In the third stage, Pr increased more rapidly than in the second stage and represented the compression of an essentially void-free material, the ‘yield value’ of which was estimated by extrapolating the third portion of the curve to the abscissa (Fig. 21.23). Generally, the first ascending portion of the curve was steeper, the second stage more extended and the yield value higher for the harder materials examined. The reduced hardness of potassium chloride and sulphathiazole tablets containing magnesium stearate (Table. 21.11 and 21.12) was reflected in the smaller yield value obtained when that lubricant was employed. For phenacetin and acetanilide a measurable radial pressure could not be obtained until a significant upper punch pressure had been applied: the second stage was virtually absent. These materials produced poor tablets due to the fact that the crystals were plate like and tended to compact into layers with poor interlayer bonding. Although there were considerable experimental difficulties, Ridgeway et al. (1969a) obtained values for Young’s modulus and the surface hardness for a number of pharmaceutical materials; these values were then related (Ridgeway et al., 1969b) to the transmission ratio and to the surface hardness of the compact. The results of these studies (together with data from other sources) are summarized in Table. 21.11 where it will, be seen that, as anticipated, the transmission ratio was large for soft materials such as aspirin, and vice versa. Overall, the data in the table generally support the view that the hardness of a substance is an important factor determining its behaviour both during and after compression.
Fig. 21.23 Radial pressure as a function of upper punch pressure. after Windhesuer et al., 1963
Table 21.12 The effect of lubricants on tabeletting properties
Lubrication
Lubricants are used to improve granule flow (glidants), minimize die-wall friction (die-wall lubricants) and prevent adhesion of the granules to the punch faces (antiadherents). The extent to which they possess these properties varies; talc is a better glidant than magnesium stearate but is an inferior die-wall lubricant. Excessive proportions of the stearate may actually impede granule flow.
Glidants
The effect of a glidant on the flow properties of a granule may be assessed by measuring the angle of repose, flow rate through an orifice or by the other methods outlined in Chapter 17. A number of mechanisms have been proposed to account for glidant action. The flow properties of smooth, nearly spherical particles are known to be better than those of more irregular shape: it has been suggested that glidant may fill surface depressions and thereby reduce roughness. If the coefficient of friction of the glidant is less than that of the granules then interparticle friction may be lowered. Alternatively, the glidant may physically separate the solid particles so that intermolecular attractive and capillary adhesion forces are reduced. The latter derive from the thin film of moisture normally present on solid surfaces. A proportion of fines will often improve granule flow, presumably by a combination of the foregoing effects although, as discussed earlier, an excess of fines often has an adverse effect on tabletting properties. Some materials acquire a frictional electrostatic charge when handled. The mutual repulsion of particles due to this effect is sufficient to impede proper die filling during the production of hexoestrol implants. Although the solution to this specific problem resides in recrystallization from a suitable solvent, glidant action may in some cases be attributed to the reduction of electrostatic charges. A detailed review of glidant action is given by Jones (1969).
Die-wall Lubricants
The effects of die-wall lubricants on tabletting behaviour may be seen by inspection of Table. 21.11, 21.12 and 21.13. They reduce the force lost to, the die wall, the ejection force, the crushing strength and the residual die-wall pressure. The relative density, punch force ratio, transmission ratio and disintegration time are increased. In part, the lubricant effect appears to reside in the formation of a tough film on the die wall. If a die ‘conditioned’ by the production of tablets from lubricated granules is subsequently used to compact unlubricated granules, easy tablet ejection persists for a short period. Strickland et al. (1956) expressed the view that the reduction of force lost to the die wall as a result of lubrication (ΔFd) may be related to the fractional coverage of the die-wall–compact interface. At low concentrations (c) of lubricants they reported an apparent agreement with a Langmuir-type adsorption equation, i.e.
(21.2)
Table 21.13 Effect of lubricants on the disintegration time of sodium bicarbonate tablets compacted at approximately 1300kg/cm2 |
||
| Lubricant concentration per cent | Disintegration time (s) Stearic acid | Magnesium stearate |
| 0 | 200 | 200 |
| 0.25 | 370 | 1500 |
| 0.50 | 520 | 2000 |
| 0.75 | 680 | 2200 |
| 1.00 | 900 | 2300 |
| 2.00 | 1400 | 2900 |
| 4.00 | 2500 | 3800 |
Materials used as lubricants must shear readily even when highly compressed. In this connection it may be noted that the slip and orientation of layers within the crystal lattice, on which the action of laminar lubricants such as talc depends, are inhibited by large applied loads and degree of compaction. Under these conditions the shear strength of talc is high (Train & Hersey, 1960) which undoubtedly accounts for the poor die-wall lubricant action of that substance.
Compression and ejection processes involve movement of material relative to the die-bore surface. The frictional force (F) resisting motion of one surface over another is proportional to the load (L) applied normally to those surfaces, that is:
(21.3)
where k is the coefficient of friction applicable to the particular conditions. For consolidation of granules the maximum force resisting further reduction in the length of the compact is related to Fd and the normal load is the die-wall force (Fr). During ejection the resistance to motion of the compact is related to the ejection force (Fe) whilst 
, the radial force operating as ejection takes place, constitutes the normal load. As Fr = Pr A and 
, by analogy with Eq. 21.3:
, the radial force operating as ejection takes place, constitutes the normal load. As Fr = Pr A and , by analogy with Eq. 21.3:
(21.4)
(21.5)
If it is assumed that Pr and 
are proportional to a generalized compaction pressure (P) the appropriate substitution in Eqs. (21.4) and (21.5) yields expressions with a common factor (PA), which implies that Fe and Fd are linearly related. This has been confirmed by, for instance, Train (1956), Lewis and Shotton (1965a) and by other workers. Clearly, this is a simplification of a very complex process and could not be expected to apply directly to all materials and compaction conditions. Train’s data for the ejection of magnesium carbonate compacts was best represented by a power law, but Lewis and Shotton used linear expressions:
are proportional to a generalized compaction pressure (P) the appropriate substitution in Eqs. (21.4) and (21.5) yields expressions with a common factor (PA), which implies that Fe and Fd are linearly related. This has been confirmed by, for instance, Train (1956), Lewis and Shotton (1965a) and by other workers. Clearly, this is a simplification of a very complex process and could not be expected to apply directly to all materials and compaction conditions. Train’s data for the ejection of magnesium carbonate compacts was best represented by a power law, but Lewis and Shotton used linear expressions:
(21.6)
(21.7)
to describe the compression and ejection of aspirin, hexamine, sodium chloride and sucrose compacts.
In these equations kd and ke were coefficients which depended on the frictional conditions at the die wall. They found that the values of kd (Table. 21.12) were different for each of the unlubricated substances and assumed (except in the case of hexamine granules) a virtually identical and lower value when 2 per cent magnesium stearate was added to the base materials. The values of ke showed a similar but less marked trend. In these circumstances the lubricant governed the frictional characteristics of the die-wall-compact interface regardless of the properties of the material being compacted or ejected. The absolute magnitudes of Pr and 
are determined by the transmission ratio (T) and this would not be identical for all lubricated substances which explains why these produced different values of Fd and Fe at a given mean compaction pressure.
are determined by the transmission ratio (T) and this would not be identical for all lubricated substances which explains why these produced different values of Fd and Fe at a given mean compaction pressure.In common with the investigations described above, most compaction studies have employed a constant compact diameter and thickness in an experimental series. Sometimes, while tabletting ‘difficult’ granulations, it is found that there is an optimum die size which minimizes, e.g. Fd, Fe, capping, lamination, etc. The effect of thickness and diameter on tabletting properties is therefore of considerable technical interest. From Table. 21.9 it can be seen that:
(21.8)
Also, if the assumption is made that 
is proportional to Pa:
is proportional to Pa:
(21.9)
Eqs. 21.8 and 21.9 fitted the results of Rees and Shotton (1969) for the compaction of sodium chloride. As these workers pointed out, these equations are not applicable when the length of the compact is sufficient to prevent proper consolidation of granules near the bottom punch—a condition which should rarely be encountered in normal pharmaceutical tabletting operations. Further studies will be necessary to show whether the expressions are valid for a wider range of materials.
It is implicit in the reduction of residual die-wall pressure and improvement of transmission ratio (Table. 21.11) by die-wall lubricants that they facilitate the movement of material in the die cavity both during and after compression. The more ready elimination of voids is reflected in the higher relative density of the compacts (Table. 21.12). Die-wall lubricants are not, however, without adverse effects on other tablet properties such as disintegration time (Table. 21.13). Also, as has been noted earlier, they may reduce the strength of the bonds between particles in a tablet. While this has the beneficial effect of reducing capping and lamination in the case of hexamine, with other substances an undesirable reduction in tablet hardness may occur. Clean surfaces are known to produce the strongest bonds but, as hard tablets can be prepared with lubricated granules, it seems reasonable to suppose that enough of the clean surface produced by fragmentation during compression remains uncontaminated by lubricant and capable of forming strong bonds. If this is so, lubricants should affect the hardness of tablets produced from crystals more than those produced from granules. This is because the crystals, being harder than granules, would tend to fragment to a lesser extent under pressure. In confirmation of this, Shotton and Lewis (1964) found that the strength of hexamine and sucrose tablets prepared from crystalline material was reduced by 76 per cent and 50 per cent respectively by the addition of 2 per cent magnesium stearate. The corresponding reduction in hardness was 28.1 per cent and 18.4 per cent respectively for tablets formed from granulated hexamine and sucrose.
Disintegration
In the context of tablet technology, disintegration implies penetration of the tablet by an aqueous liquid, disruption of internal bonds and the subsequent breakdown of the tablet. It is reasonable to suppose that rapid penetration of liquid is an essential requirement for rapid disintegration of conveniently formulated tablets.
Theory
A pressure differential P is required to cause liquid of viscosity η to flow with a velocity dL/dt in a horizontal pore of hydraulic radius m at a time t when the meniscus has moved a distance L along the pore.
(21.10)
Here, k0, a pore shape factor and m, the hydraulic radius, are used because the pore in a compact has an irregularly shaped rather than a circular cross section. In all other respects the equation is of the same form as that used to describe the flow of liquids in circular pipes. The pressure differential derives from the drop in pressure across a meniscus:
(21.11)
where γ is the surface tension of the liquid and θ the contact angle with the pore wall (cf. the corresponding expression for a liquid meniscus in a capillary of circular section). From the foregoing:
(21.12)
If L = 0 when t = 0 on integration
(21.13)
If the pore is part of a network that is uniform throughout the compact, for penetration into one face of that compact, the volume of liquid in the pores (V) is proportional to L and at any time t:
(21.14)
k being a coefficient representing the properties of the liquid and of the pore network within the compact.
Effect of Granule Properties
For a compact of given porosity the void space may comprise a few large diameter pores or many fine ones. Fine pores are most likely to be formed from soft friable granules that readily fragment into much smaller particles during compression. In these circumstances the size of the pores should be largely independent of the size of the granules used to prepare the compact. Conversely, hard granules do not collapse to the same extent during compaction and thus the pore network will be coarser and dependent on the original granule size. According to Eq. 21.13, such a network would be more readily penetrated by liquids than that resulting from the compaction of soft materials. Ganderton and Selkirk (1969) made granules from lactose and sucrose by massing with water and found that the hardness and bulk density increased with the amount of water used for granulation. They suggested that this was due to a reduction in intragranule pore volume. Sucrose granules, which were harder and less friable than those prepared with lactose, gave compacts with a coarser pore structure that was more readily penetrated by liquid.
Initially, the imbibition rate for all compacts conformed to Eq. 21.14, but on many occasions slowed down or stopped abruptly before the theoretical volume of liquid had been absorbed. This was more frequently observed with sucrose compacts, particularly in those circumstances where the porosity was high and the largest, hardest, granules had been used. These conditions were essential for the effect to occur with lactose compacts. It was postulated that compacts showing this type of behaviour had a broad and perhaps discontinuous distribution of pore sizes. Rapid penetration of the wider pores isolated some regions of the compact, occluded air prevented subsequent penetration of the isolated regions and, as a result, a lower than theoretical degree of saturation was attained. By compressing the compacts to a lower porosity the size and lack of uniformity in size distribution was reduced and higher fractional saturation obtained.
In the foregoing experiments liquid penetrated one face of the compact. However, complete immersion of the tablet during disintegration exposes the whole surface to penetrant liquid. In this situation the geometry of the system demands a reduction in the total pore cross-sectional area available for flow as liquid approaches the centre of the tablet. Furthermore, as the pore entrances on all surfaces of the tablet will be sealed by liquid, air entrapment will occur. Liquid will penetrate until the increase in the pressure of the occluded air is equal to the capillary pressure. A larger fraction of the total pore length will be penetrated with fine rather than coarse pores, but this need not necessarily imply good disintegration characteristics. It is quite possible that the back pressure of the occluded air as it compresses in the pore and the decreasing area available for flow could act in concert with the normally operative viscous forces so that the rate of penetration was too low for a short disintegration time.
Effects of Die-Wall Lubricants
Some medicaments, e.g. phenothiazine, and most die-wall lubricants are hydrophobic and impart this property to the pore walls in the tablet. The resultant increase in the contact angle reduces the value of cos θ in Eq. 21.13 and therefore the pore is less readily penetrated by an aqueous liquid. This is reflected in the longer disintegration time which is found when, for instance, the stearate lubricants are used (see Table. 21.13). For magnesium carbonate compacts, there was a rough proportionality between the concentration of magnesium stearate and the time taken for the uptake of a given amount of water (Ganderton, 1969). This time was extended, other conditions being constant, by increasing the efficiency of the process whereby the lubricant was blended with the granules. When 1 per cent of magnesium stearate was incorporated during the moist granulation stage penetration of water was completely inhibited: the pores were so well coated that the contact angle was greater than 90° and probably approached that of the solid lubricant. As the cosine of an angle greater than 90° is negative, V2 in Eq. 21.14 is also negative and penetration cannot occur. Where such a situation arises in tablet manufacture, due either to the intrinsic water repellency of the medicament or the lubricant, a water soluble surfactant may be added to the granules to reduce the contact angle. Surfactants also facilitate the displacement of air from blind pores in a tablet under the conditions of a dissolution test. Studies by Wurster and Seitz (1960) with benzoic acid compacts into which 1mm diameter artificial holes had been formed by drilling, demonstrated that the total pore surface was not available for attack by simple aqueous dissolution fluids, due to occluded air. This effect was abolished when air was removed from the pores by vacuum treatment or by addition of a surfactant to the dissolution medium.
Mode of Disintegrant Action
Starch has been used for many decades as a disintegrating agent. Since instrumented tablet machines were not available to Berry and Ridout (1950) they used the compression ratio, that is the ratio of the weight of a tablet to its thickness, as a measure of the degree of compaction in their studies. The disintegration time of tablets containing 15 percent of starch, dry mixed with phenacetin granules, fell linearly from 15min at a compression ratio of 0.57, to 6min as that ratio was raised to 0.7; further compaction to 0.75 brought about a steep rise in disintegration time to 24min. They argued that the disintegrating action of starch was due to swelling and that this would be most efficient when the granules were in close contact with the starch grains in a tablet of low porosity. At still lower porosities (higher compression ratios) imbibition of water was restricted by the small pore size. Similar effects have been noted by Higuchi et al. (1954b) with sulphadiazine tablets. Berry and Ridout compared the effect of starch added as described above with that of alginic acid incorporated during the moist granulation stage. This latter substance also swells in contact with aqueous liquids, but, since it was part of the granule structure, disintegration occurred by erosion and breakdown of the granules into the original particles from which they had been prepared. A low disintegration time of approximately 3min was maintained up to a compression ratio of 0.65 but then, as with starch, further compaction caused a marked increase in the time taken for the tablet to disintegrate. Burlinson (1950) observed that there was little to choose between aliginic acid and starch if part of the latter was incorporated into the granules during moist granulation of the medicaments.
The assumption that starch facilitates tablet disintegration by swelling action (bursting) is based on the known increase in the dimensions of starch grains exposed to aqueous fluids. Patel and Hopponen (1966) found a 70 per cent increase in volume, but Ingram and Lowenthal (1968) quote a lower figure which, they suggest, is not sufficient to account for the disintegrant action of starch. The latter workers suggest that the irreversible swelling of damaged starch grains is much greater but could not establish a correlation between the degree of damage and the tablet disintegration time. As an alternative to the ‘swelling’ theory it has been argued that starch may derive its disintegrant properties from capillary action in the intergranule pore system. It is possible that it may also prevent the complete collapse of the pore system during compaction or provide a secondary pore surface of low contact angle against aqueous media.
It seems likely that the disintegration time of tablets is determined by the complex interaction of a wide variety of factors. The rate at which liquid penetrates a tablet, the nature and method of incorporation of lubricants, the action of disintegrants, the degree of compaction and the reduction of interparticle bond strength in the presence of water, are all clearly of major importance. In practice, granules often contain hydrophilic colloids which, apart from their effect on the surface tension, viscosity and contact angle of the penetrant liquid, may dissolve or gel and this may have a profound effect on the rate at which a tablet disintegrates. Finally, some tablets disintegrate by continuous erosion of an outer wetted layer in which case the complete rapid saturation of the pore system by aqueous media may not be an essential step in the disintegration process.
Capsules
A capsule consists of a dose of drug enclosed in a water-soluble shell or envelope. Although capsules are predominantly used for oral medication, suppositories, consisting of the drug and a suitable dispersion medium enclosed in a flexible gelatin shell, are available (Senior, 1969). There are two types of capsule, the hard variety which is intended for the administration of particulate solids and the soft or flexible capsule used for powders, nonaqueous liquids or pastes. As the shell takes several minutes to dissolve, the taste of unpalatable drugs is effectively masked. Officially, capsules may be spherical (perles), ovoid or cylindrical with hemispherical ends—only the last of these descriptions applies to hard capsules. The shell may include an antimicrobial preservative, can be coloured as an aid to identification and may be rendered opaque by dyes or fillers if the drug, e.g. some vitamins and chlorodiazepoxide, is photosensitive. For official preparations, diluents may not be added to the medicament unless this is specifically permitted in the relevant monograph: a similar restriction also applies to the use of coloured shells for capsules described in the Codex. It is possible to coat capsules with cellulose acetate phthalate or a mixture of stearic acid and butyl stearate if enteric action is required, but it is more usual in these circumstances to formulate if possible as an enteric coated tablet.
Soft Capsules
Formerly, these were extemporaneously prepared by filling and sealing shells which had been made by dipping metal formers or ‘olives’ into a molten aqueous base containing glycerin and gelatin (glycogelatin). The glycerin ensures that a flexible shell is obtained. The first step in mechanizing the production of soft capsules was the development of a process whereby a sheet of glycogelatin was covered with a measured amount of anhydrous fluid drug. A second sheet was then placed over the drug and the ‘sandwich’ compressed between plates into which had been set hollow elliptical or hemispherical dies. The pressure forced the glycogelatin sheets and the drug into the dies and formed individual capsules by sealing and cutting at the rims. The capsules were then separated from the residual glycogelatin matrix, washed in solvent to remove traces of oil used as a release agent and dried to give a tough though flexible shell. The process was used until recently but could not complete with the rotary die machine (Figs. 21.24 and 21.25) in terms of economic, continuous, high speed production. In this machine the dies are set in a pair of contrarotating cylinders. Two continuous glycogelatin ribbons, previously warmed by the filling wedge to make them plastic, are fed between the cylinders. At the moment when opposing dies converge, a measured fill of drug is forced between the ribbons by a metering pump and simultaneously the edges of the dies seal and cut out a complete capsule. These are separated from the matrix, washed in solvent and dried (Pharmaceutical Journal, 1960). It is claimed that the fill of drug per capsule can be maintained to within 1 per cent of the mean fill for the batch.
Fig. 21.24 The rotary die soft capsule filling machine. R.P. Scherer Ltd.
Fig. 21.25 Method of filling and sealing soft gelatin capsules in a rotary die machine (diagrammatic).
Depending on machine capacity and capsule size, production rates in excess of 30000 capsules/h are possible while production lines are available for the formation of the ribbons, sealing of the capsules, and their subsequent washing and drying as a continuous integrated process. Rotary die machines have also been developed whereby powders as such, may be directly encapsulated in a flexible gelatin shell. The ribbon is made to conform to the contour of the die by application of reduced pressure, a ‘plug’ of powder is forced between the ribbons and the capsules sealed, washed, and dried as described above. Capsules of some tetracycline antibiotics are made by this method.
Virtually any nonaqueous liquid drug or a powdered solid formed into a suspension of suitable consistency by the addition of oil may be presented as a soft capsule. Vegetable oils, Liquid Paraffin and Soft Paraffin are used when necessary to dilute or suspend the drug. Vegetable oils are specified for the dilution of the oil-soluble vitamins, as the paraffins are known to interfere with the absorption of these medicaments from the gut. As noted in the BPC 1968, drugs containing water or other gelatin solvent cannot be filled into soft capsules due to the resultant pitting, softening and eventual breakdown of the shell; the water should be removed and the residue mixed with an oil prior to encapsulation. Conversely, anhydrous but hygroscopic liquids may withdraw water from the shell which then hardens and has reduced solubility. Glycogelatin softens markedly as the temperature is raised and it is usual to increase the proportion of gelatin for capsules used for tropical climates. Due to changes in the equilibrium moisture content of the shell, the flexibility of soft capsules varies with ambient humidity. They are hard and inflexible at 20 per cent, taut but flexible at 50 per cent, and soft and flaccid at 80 per cent relative humidity. In the latter condition the relatively high moisture content of the shell may encourage the diffusion of any water-soluble substance from the fill into the shell and as a result there may be unwanted reactions between the constituents or changes in the appearance and solubility of the shell.
According to Notton (1956) the oxygen permeability also increases with the moisture content of the glycogelatin envelope. It will be noted that the official storage requirements for both varieties of capsule are framed to prevent moisture uptake and other forms of deterioration as well as crushing and breakage. Sufficient preservative must be used to ensure effective antimicrobial action even where the moisture content of the envelope is high. Additionally, the possibility that preservative may partition between the shell and the capsule content cannot be ignored. Studies by Patterson and Lerrigo (1947) showed that beta naphthol, used extensively at that time as a mould inhibitor, partitioned to such an extent that after approximately 18months storage the amount of preservative in the capsule contents was more than twice that in the shell. It is unlikely, however, that partitioning would be a serious problem with the ionized antimicrobials, such as potassium sorbate, that are now employed as preservatives for capsule shells.
Hard Capsules
These are made by filling the drug as powder, granules or pellets into a preformed cylindrical shell or body, the contents being retained by a shorter shell or cap fitted over the body. Wood (1965) has reported that on occasion the cap may disengage from the body during the act of swallowing. This, clearly, is undesirable, and may be avoided by securing the shell components with a narrow sealing strip. Hard capsules provide an alternative to powders, cachets and tablets for the administration of solid medicaments. They are at least as durable as tablets and correct formulation permits the rapid release and absorption of drug in the gastrointestinal tract.
Hard capsule shells are made by dipping moulds into a gelatin solution, the film on the mould dried, the shell cut to length and then stripped from the mould (Pharmaceutical Journal, 1965 & 1966). The shells are made in a range of 8 sizes to accommodate 50–1000mg doses of drug but for veterinary preparations three larger shells are available for 5–30g doses. Particularly where high-speed filling equipment is used there must be no difficulty in removing and replacing the cap, yet the junction of this with the body must form an efficient seal. The necessarily small tolerances on the dimensions of the cap and body are achieved by rigid control of all stages of the manufacturing process. Peck et al. (1964) have described a device utilizing the absorption of β-radiation from a chlorine-36 source for the control of shell thickness.
A description of the small scale filling of hard capsules is given in Chapter 26 for large scale roduction the capsules are sorted and positioned body downwards and sucked into holes in upper and lower rings which, when separated, retain the cap and body respectively (Fig. 21.26). The lower ring is then presented to a head which fills the body with medicament composition, as either a powder or a lightly compacted pellet. Subsequently the rings are recombined, the body pushed into the cap, the capsule ejected and any dust adhering to the outer capsule surface removed. In semiautomatic equipment the rings are separated, transferred to the filling head and recombined by hand, but with fully automatic machines these operations are accomplished by mechanical means.
Fig. 21.26 Stages in the filling of hard capsules.
In common with many other pharmaceutical materials such as bacitracin, magnesium trisilicate and starch (Strickland, 1962), the moisture uptake–ambient humidity curve for gelatin exhibits marked hysteresis (Strickland & Moss, 1962). Thus, the moisture content of the shell, normally 9–12 per cent, is determined by both its previous history and the ambient humidity; it may vary between 4 and 16 per cent under extreme conditions. Strickland and Moss concluded that the diffusion of moisture into the capsule contents occurs mainly through the shell surface with only a small contribution due to the junction between the body and cap. Moisture will be absorbed from the environment via the capsule wall if the medicament is more hygroscopic than the shell and for this reason such substances should not be dispensed in capsules. The preparation of the drug and its subsequent filling into the body poses many of the problems which have already been discussed in the context of tabletting, for instance, care must be taken to ensure the efficient mixing of minor constituents with other drugs or diluents. The weight of drug in both tablets and capsules is governed by the volume available for filling and the bulk density of the drug formulation under production conditions, while for low weight variation the powder must flow in a reproducible manner. Unlike the die cavity of the tablet machine, the volume of the capsule body cannot be adjusted to accommodate a given weight of powder and therefore, with hard capsules, the bulk density of the powder must be adjusted by the addition of an inert diluent such as lactose. Care is necessary in the selection of the diluent as, apart from its effect on powder flow properties there is some evidence, at least in the case of chloramphenicol capsules (Withey & Mainville, 1969), that high concentrations of lactose may interfere with drug dissolution. In addition to diluent, a lubricant such as magnesium stearate (about 0.5 per cent) may be added for correct operation of the mechanism in the filling head and up to 5 per cent talc may be added to improve the flow properties of the powder. Reier et al. (1968) have reported on preliminary studies to elucidate the effect of medicament type, production rate, capsule size, presence or absence of talc, particle size, specific volume and powder flow on the semiautomatic filling of hard capsules. These studies showed that, for a given capsule size and production rate, the filling weight was largely determined by the specific volume of the powder, while the action of talc in reducing weight variation was attributed to more reproducible, rather than more rapid, flow of the powder. Computer analysis of the data gave equations for filling weight and filling weight variation which included all possible interactions of the factors listed above, except that terms for medicament type and particle size were not required as their effect was reflected in the flow properties of the powder and these were adequately defined by the flowability and specific volume. As the capsule size is set by the specific volume of the powder, only that parameter and the flowability need be measured, using a small sample, to predict the performance of a given filling composition. Thus, production difficulties may be anticipated and corrected prior to filling trials with large and potentially expensive batches of medicament. Reier and coworkers found excellent agreement between the predicted and observed behaviour of a number of model filling compositions.
Quality Control of Capsules
The reasons for controlling the disintegration time, weight variation and medicament content of tablets also apply to capsules and have been discussed in considerable detail earlier in this chapter.
Disintegration
The tablet disintegration apparatus of the Pharmacopoeia is specified for the test but the guided disc may not be used. With capsules, it is clearly necessary to distinguish, during the course of the test, between particles of shell and drug; the BP 1973 requires that no portion of drug shall remain after 15min which does not freely pass the 10mesh screen at the base of the disintegration tube. Aguiar et al. (1968) found that hard capsules containing chloramphenicol disrupted sufficiently to release the drug after a few minutes immersion in simulated gastric fluid, but several hours were required for complete solution of the shell. These authors emphasized that both rapid deaggregation (disintegration of powder aggregates into primary particles) and rapid dissolution were, necessary conditions for good biological availability of the drug.
Uniformity of Filling Weight
This cannot be determined by simply weighing the filled capsules as there will be a variation in weights of the individual shells. For this reason the weight of the content of an individual capsule is defined as the difference in weight of the intact dose unit and the empty shell. Removal of dry particulate solid from the capsule merely involves detaching the cap or slitting the shell (Method A) but oily contents are squeezed out of the split shell and the remainder removed by washing in solvent (Method B). In all other respects the principle of the test is the same as that applied to tablets: Table. 21.14 summarizes official requirements. Only one set of tolerances is specified when Method B is used to determine the weight of the capsule contents as this method applies to soft capsules and it is not usual to fill these with less than about 0.3g of material.
Table 21.14 Summary of the BP 1973 test for uniformity of filling weight of capsules
Content of Medicament
A sample from the bulked contents of the capsules used in the test for weight uniformity is assayed and the dose of medicament per capsule calculated. This must comply with the limits set out in the relevant monograph which is framed to take into account such factors as drug purity, assay accuracy, assay precision and manufacturing difficulties. The general requirement that the drug content of individual dose units shall not show a gross deviation from that stated or prescribed also applies to capsules.
The available literature, see for example Aguiar et al. (1968), Robinson et al. (1968) and Stock (1962), suggests that, provided the capsule contents are correctly formulated and the manufacturing process adequately controlled, there should be few instances where difficulties will be encountered in the large scale production of capsules complying with official requirements.
References
Abbott DD, Packman EW, Rees EW, Harrisson JWE. J. Am. Pharm. Ass. (Sci. Edn). 1959;48:19.
Aguiar AJ, Wheeler LM, Fusari S, Zelmer JE. J. Pharm. Sci. 1968;57:1844.
Antonides HJ, DeKay GH. Drug Standards. 1953;21:205.
Ashby FW, Muggleton PW, Taylor F, Woodward WA. J. Pharm. Pharmacol. 1954;6:1048.
Berry H. Quart. J. Pharm. 1939;12:501.
Berry H, Ridout CW. J. Pharm. Pharmacol. 1950;2:619.
Berry H, Smith AN. Quart. J. Pharm. 1944;17:248.
Blubaugh FC, Zapapas JR, Sparks MC. J. Am. Pharm. Ass. (Sci. Edn). 1958;47:857.
Brook DB, Marshall K. J. Pharm. Sci. 1968;57:481.
Brownley CA, Lachman L. J. Pharm. Sci. 1963;52:86.
Burlinson H. J. Pharm. Pharmacol. 1950;2:638.
Castello RA, Mattocks AM. J. Pharm. Sci. 1962;51:106.
Duvall RN, Koshy KT, Pyles JW. J. Pharm. Sci. 1965;54:607.
Evans AJ, Train D. Bibliography of the Tabletting of Pharmaceutical Substances. London: Pharmaceutical Press; 1963.
Evans AJ, Train D. Supplement to Bibliography of the Tabletting of Pharmaceutical Substances. Pharmaceutical Press: London; 1964.
Feldman EG. J. Pharm. Pharmacol. 1969;21:134.
Fonner DE, Banker GS, Swarbrick J. J. Pharm. Sci. 1966;55:181.
Ganderton D. J. Pharm. Pharmacol. 1969;21:9S.
Ganderton D, Selkirk AB. Symposium on Powders. Dublin: Society of Cosmetic Chemists of Great Britain; 1969.
Gold G, Duvall RN, Palermo BT, Slater JG. J. Pharm. Sci. 1968;57:2153.
Goodhart FW, Mayorga G, Mills N, Ninger FC. J. Pharm. Sci. 1968;57:1770.
Griffiths RV. Manufacturing Chemist and Aerosol News. 1969;40:29.
Gunsel WC, Lachman L. J. Pharm. Sci. 1963;52:178.
Hadgraft HG. Pharmaceut. J. 1945;154:183.
Hartley F. J. Pharm. Pharmacol. 1950;2:628.
Higuchi T, Arnold RD, Tucker SJ, Busse LW. J. Am. Pharm. Ass. (Sci. Edn). 1952;41:93.
Higuchi T, Rao AN, Busse LW, Swintosky JV. J. Am. Pharm. Ass. (Sci. Edn). 1953;42:194.
Higuchi T, Nelson E, Busse LW. J. Am. Pharm. Ass. (Sci. Edn). 1954;43:344.
Higuchi T, Elowe LN, Busse LW. J. Am. Pharm. Ass. (Sci. Edn). 1954;43:685.
Higuchi T, Shimamoto T, Eriksen SP, Yashiki T. J. Pharm. Sci. 1965;54:111.
Howard WR. J. Pharm. Pharmacol. 1951;3:777.
Hoyle H. Quart. J. Pharm. 1946;19:279.
Ingram JT, Lowenthal W. J. Pharm. Sci. 1968;57:393.
Jaffe J, Lippmann I. J. Pharm. Sci. 1964;53:441.
Jones TM. Symposium on Powders. Dublin: Society of Cosmetic Chemists of Great Britain; 1969. 1969
Kornblum SS, Zoglio MA. J. Pharm. Sci. 1967;56:1569.
Kuzel NR, Roudebush HE, Stevenson CE. J. Pharm. Sci. 1969;58:381.
Lachman L, Cooper J. J. Pharm. Sci. 1963;52:490.
Lachman L, Sylwestrowicz HD, Speiser PP. J. Pharm. Sci. 1966;55:958.
Lewis CJ, Shotton E. J. Pharm. Pharmacol. 1965;17:71S.
Lewis CJ, Shotton E. J. Pharm. Pharmacol. 1965;17:82S.
Malm CJ, Emerson J, Hiatt GD. J. Am. Pharm. Ass. (Sci. Edn). 1951;40:520.
Mapstone TJ. Pharmaceut. J. 1952;168:291.
Nelson E. J. Am. Pharm. Ass. (Sci. Edn). 1955;44:494.
Notion HEF. Phaimaccut. J. 1956;177:69.
Orewan E. Rec. Prog. Phys. 1949;12:185.
Patel NR, Hopponen RE. J. Pharm. Sci. 1966;55:1065.
Patterson SJ, Lerrigo AF. Quart. J. Pharm. 1947;20:83.
Peck EG, Christian JE, Banker GS. J. Pharm. Sci. 1964;53:607.
Pharmaceut. J. (1960) 185, 372.
Pharmaceut. J. (1965) 194, 475.
Pharmaceut. J. (1966) 197, 169.
Prance HP, Stephenson D, Taylor A. Quart. J. Pharm. 1946;19:286.
Rees JE, Shotton E. J. Pharm. Pharmacol. 1969;21:731.
Reier GE, Shangraw E. J. Pharm. Sci. 1966;55:510.
Reier GE, Cohn R, Rock S, Wagenblast F. J. Pharm. Sci. 1968;57:660.
Ribeiro D, Stevenson D, Samyn J, Milsovich G, Mattocks AM. J. Am. Pharm. Ass. (Sci. Edn). 1955;44:226.
Richman MD, Fox CD, Shangraw RF. J. Pharm. Sci. 1965;54:447.
Ridgeway K, Shotton E, Glasby J. J. Pharm. Pharmacol. 1969;21:19S.
Ridgeway K, Glasby J, Rosser PH. J. Pharm. Pharmacol. 1969;21:24S.
Ritschel WA, Skinner FS, Schlumpf R. Pharm. Acta Helv. 1969;44:547.
Robinson MJ, Grass GM, Lantz RJ. J. Pharm. Sci. 1968;57:1979.
Rogers AR. J. Pharm. Pharmacol. 1956;8:1103.
Rolfe HG. Pharmaceut. J. 1956;176:178.
Senior N. Pharmaceut. J. 1969;203:703.
Shotton E, Ganderton D. J. Pharm. Pharmacol. 1960;12:87T.
Shotton E, Ganderton D. J. Pharm. Pharmacol. 1960;12:93T.
Shotton E, Ganderton D. J. Pharm. Pharmacol. 1961;13:144T.
Shotton E, Deer JJ, Ganderton D. J. Pharm. Pharmacol. 1963;15:106T.
Smith AN. J. Pharm. Pharmacol. 1950;2:627.
Smith CD, Michaels TP, Chertkoff MJ, Sinotte LP. J. Pharm. Sci. 1963;52:1183.
Smith KL. J. Pharm. Pharmacol. 1955;7:875.
Srinivas R, DeKaY HG>, Banker GS. J. Pharm. Sci. 1966;55:225.
Stephenson D. Pharmaceut. J. 1960;185:137.
Stephenson D, Humphreys-Jones JF. J. Pharm. Pharmacol. 1951;3:767.
Stephenson D, Smith DS. J. Pharm. Pharmacol. 1951;3:547.
Stock FEH. Pharmaceut. J. 1962;188:453.
Strickland WA. J. Pharm. Sci. 1962;51:310.
Strickland WA, Nelson E, Busse LW, Higuchi T. J. Am. Pharm. Ass. (Sci. Edn). 1956;45:51.
Strickland WA, Moss M. J. Pharm. Sci. 1956;51:1002.
Sutaria RH. Manufacturing Chemist and Aerosol News. 1968;39:37.
Swartz CJ, Weinstein S, Windheuser J, Cooper J. J. Pharm. Sci. 1962;51:1181.
Swartz CJ, Anschel J. J. Pharm. Sci. 1968;57:1779.
Train D. J. Pharm. Pharmacol. 1956;8:745.
Train D. Pharmaceut. J. 1960;185:129.
Train D, Hersey JA. J. Pharm. Pharmacol. 1960;12:97T.
Webster AR, van Abbe NJ. J. Pharm. Pharmacol. 1955;7:882.
Whitehouse RC. Pharmaceut. J. 1954;172:85.
Windheuser J, Misra J, Ericksen SP, Higuchi T. J. Pharm. Sci. 1963;52:767.
Withey RJ, Mainville CA. J. Pharm. Sci. 1969;58:1120.
Wood JH. J. Pharm. Sci. 1965;54:1207.
Wurster DE. J. Am. Pharm. Ass. (Sci. Edn). 1959;48:451.
Wurster DE. J. Am. Pharm. Ass. (Sci. Edn). 1960;49:82.
Wurster DE, Seitz JA. J. Am. Pharm. Ass. (Sci. Edn). 1960;49:335.