SOME STUDIES OF THE FORMULATION AND EVALUATION OF TABLETS WITH SPECIAL REFERENCE TO DIRECT COMPRESSION SUGARS

The potential of several new sugar matrices as direct compression vehicles has been investigated using a systematically organized series of tests. These matrices, produced by California and Hawaiian Sugar Company (C & H), were evaluated in a comparative manner with two commercially available products (DiPac and NuTab). In some cases, Emdex (another commercially available product) was also used. Moisture content, density, and flow properties were studied. Scanning electron micrographs were obtained t o determine the morphology of the particles . Particles size spectra were generated for all the matrices. For some of the C & H Products (AI, B, and C), a computer-interfaced instrumented tablet press was used to generate compression profiles from which the compressibility of the materials was assessed. The formulation efficiency of four C & H Products (AI, AII, B, and Brown), DiPac, and NuTab was determined by incorporating the matrices in several chewable formulations (ascorbic acid, multivitamin, dextromethorphan, antihistamine, antidiarrhea, and antacid) and one non-chewable formulation (pediatric strength aspirin). A computer-interfaced tablet press, from which compression peak heights were obtained, was used to pr~­ pare the ascorbic acid, multivitamin, and aspirin tablets . The results show that the c· & H matrices, particularly Product B, have considerable potential as direct compression vehicles. It was also demonstrated that the utility of sugar matrices is not limited to the formulation of chewable tablets. Further, the results indicate that although matrices maybe chemically alike, differences in their physical properties may indeed produce gubstantial differences in their intrinsic properties and overall formulation hehavior. The effect of aging on invitro performance of the antacid tablets was evaluated. It was found that samples stored under relatively mild stress conditions (30°C with 80% relative humidity) for three months had reduced acid neutralization rates. It was also found that these storage conditions were sufficient to induce substantial aging in sugar coated and enteric coated tablets (chlorpromazine and aspirin respectively) ove r

The effect of aging on in-vitro performance of the antacid tablets was evaluated.
It was found that samples stored under relatively mild stress conditions (30°C with 80% relative humidity) for three months had reduced acid neutralization rates. It was also found that these storage conditions were sufficient to induce substantial aging in sugar coated and enteric coated tablets (chlorpromazine and aspirin respectively) ove r a four week period.
ACKNOWLEDGMENTS I want to thank Dr. Christopher T. Rhodes for his interest, guidance and support throughout my graduate study .
understanding will always be remembered .

His unending patience and
I also want to thank the faculty members of the Department of  xx. XXI. XXII.

LIST OF FIGURES
Compressional curve .

U.S.P. disintegration apparatus
Particle size distribution of the matrices studied 26 .

LIST OF FIGURES (Continued)
Friability as a function of compressional force (multivitamin formulation) . . . .
Friability as a function of hardness (multivitamin formulation) .

I. INTRODUCTION
It is well recognized that the compressed tablet is the most widely used pharmaceutical dosage form in North America and Western Europe.
Defined by the USP as "a solid dosage form containing medicinal substances with or without suitable diluent", tablets may be classified either as molded or compressed depending on the method of manufacture (1).
The formulation of tablets has undergone rapid change and development over the last several decades. With the emergency of precompression induced die feeding, high speed (and most recently, ultra-high speed) computer controlled presses, the manufacture of tablets has become one of the most sophisticated aspects of pharmaceutical production. In addition to technological advances, recent government regulations regarding the bioavailability and efficacy of pharmaceutical products have also played a major role in the advancement of tablet production.
Tablet design and formulation as it exists today can be thought of as "the process whereby the formulator ensures that the correct amount of the drug, in the right form, is delivered at or over the proper time at the proper rate and in the desired location, while having its chemical integrity protected to that point (2) .
Tablet presses operate at varying speeds, ranging from less than a thousand to almost a million tablets per hour. In order to ensure that adequate weight uniformity is maintained, it is important, especially on high speed presses, that the material to be compressed possess acceptable flow characteristics . For smooth operation of the tablet press, the granulation or direct compression mix must have good compressibility and lubricant properties. Many of the active ingredients and excipients do not meet these criteria. and it is therefore necessary to prepare a sui t-able gr anulation prior to the compression process. Alterna t ivel y , a direct compression technique may be employed provided that the componen ts mee t the above mentioned cri t e ria. A third method, dry gr a nulation , is limited to situations where nei the r wet granulation nor direct compress ion can be used.
A. Some aspects of tabl e t manufacture 1. Methods of manufacture a. We t granulation In spite of its cos tly nature, requiring intensive l abor , conside rable material handling, and costly equipment, wet granulation continues to be the most widely used process for tablet manufacture. Although many of th e products cu· rrently being formulated by wet granulat ion could be made by direct compression, existing government regula tions would define s uch chan ges as major modifications. Thes e modifications would require changes in ingredients, or a t least, changes in forms of previously used excipients. Hence, the newly fo rmulated pr oducts would have to un<lergo additional stability, bioavai l a bility, safe t y s tudies, a nd a new submission to r egulatory agencies such as the FDA.
Addi tionally, wet granulation offers several adva nta ges over di rect com pression. These advantages include: ease of attainment of acceptab les content uniformity (par tic ularly with soluble , low dosage drugs) , abi lit y to r egulate moisture cont ent of the gr a nulation during th e dryin g cyc le, modification of flow characteristics by controlled particle si ze dis tribution, and good distribution of co lo r adclitives. Through proper selec tion of granula tion so lution and the binder, the dissolution rate of a hydrophobic drug may be improved.

We t granulation is not without i t s limitations. The primary limitation
is the high cost associated with the process in terms of labor, time, equipment, energy, and space requirements. Other limitations include soluble dye migration during the drying cycle, and a high incidence of component incompatability as a result of the close contact brought about by the use of the granulating solvent . In addition, heat and moisture sensitive drugs of ten can not be processed by this method.

b. Direct compression
Direct compression is the simplest tablet production method . In this process, tablets are compressed directly from powder blends of active ingredients and suitable excipients (including, disintegrants, and lubricants). These blends flow uniformly into the die cavities to form the compacts. The method offers several advantages over wet granulation.
First and foremost, the process is economical; a limited number of processing steps are involved in direct compression compared to wet granulation. Other manufacturing variables such as the mode of addition of the binder, drying time, etc. are limited. In tablets made from direct compression the particles do not exist in agglomerate form (granules).
Hence, upon contact with dissolution medium, a prime particle dissociation is affected. This in turn may result in faster dissolution which may indeed be the single most important factor controlling product bioavailability (3). Since no moisture is involved in the preparation of the blends for direct compression, the tablets made from this process tend to be more stable than those produced by wet granulation (4).
However, direct compression has several disadvantages . Many active ingredients have poor flow and compression characteristics. This generally requires that the active ingredient be incorporated into a matrix or a mixture of matrices to achieve adequate flow and compressibility. Sev-eral advances have been made in modifying the particle shape and form of some active ingredients and fillers to render them more suitable for direct compression processing. Such success has been achieved with aspirin, acetaminophen, and several vitamins. However, these modifications tend to increase the price of the raw material, thus negating the cost advantage of direct compression.
In most cases, the amount of active ingredient(s) that can be incorporated into a matrix (dilution potential) is quite limited (usually less than 30%). Although this level varies with different active ingredients and matrix systems, it is a major limitation when formulating high dosage drug products. On the other hand, adequate distribution is sometimes difficult to achieve when working with low dosage drugs. Also, because of lack of, or low levels of moisture in direct compression blends, development of static charges may lead to unblending. This phenomenon can also occur as a result of vast differences in particle size or density among the components of the blends. The blends are most susceptible to unblending particularly in the hopper or feed frame of the tablet press .
Another limitation of this method is the difficulty of achieving deep and uniform color distribution in tablets.

c. Dry granulation (Precompression or Slugging)
Dry granulation is the process whereby granules of powder blends are obtained without the use of heat or solvent. This method is usually reserved for those formulations which cannct be processed by either wet granulation or direct compression due to technical or economical reasons.
The process entails the formation of slugs from the active ingredient and one or more excipient(s). These slugs are then milled and screened into a desired particle size distribution range. The "granules" are then mix-ed with the remaining components of the formulation to obtain the final production blends.
2. Formulation of chewable tablets 5 Chewable tablets do not constitute a major portion of the pharmaceutical dosage forms . However, they offer unique convenience to young children and some geriatric patients who find swallowing difficult. Chewable tablets also provide a rapid onset of activity, such as in the treatment of hyperacidity conditions. They are also useful as target oriented dosage forms for the treatment of mouth or throat conditions.
In the production of chewable tablets, several formulation factors are taken into consideration. These factors include: the amount of active substance per tablet, flow characteristics of the production blend, compressibility, overall stability of the formulation, and organoleptic properties . Among these factors, organoleptic properties of the active ingredient(s) is perhaps the most important consideration for some formulations. It is generally accepted that, as the required amount of the active substance per tablet gets smaller and less bad-tasting, the task of achieving an acceptable formulation becomes easier (5). This is because of the variety of options available in attaining such a formulation .
However, formulating an extremely bad-tasting or high dosage drug into a chewable table t provides an enormous and difficult task. The final formula should be one with good taste, acceptable mouth-feel, bioavailability, stability, and quality. Additionally, the product should be prepared from an economical formula to ensure market success.
In order to achieve a product with the foregoing attributes, several formulation techniques and approaches are available. To solve the problem of bad mouth-feel or after-taste, the formulator oay choose to micro-encapsulate the active ingredient(s). This is a method of coating the drug particles or liquid droplets with polymeric material, thus masking the undesirable taste and forming relatively free flowing microcapsules.
The particle size of the micro-capsules ranges from 5 to 500 microns.
The technique of micro-encapsulation is heavily limited by patents. A detailed review of the various methods of micro-encapsulation has been presented by Luzzi (6). Among the various methods described in the literature ( (9) has patented an erythromycin form which is suitable fo r chewable tablet formulation. In addition to masking unpleasant taste, micro-encapsulation may eliminate or minimize potential incompatibility problems. Such success has been reported by Bakan and Sloan in the formulation of aspirin and chlorpheniramine maleate (10). It may be noteworthy that upon compression, the microcapsule integrity may be partially or completely compromised. This, in combination with the extent to which the tablet is chewed, and the length of time the drug remains in the mouth may determine the degree of taste-masking.
Adsorption of an active ingredient (with objectionable mouth-feel) onto a substrate capable of maintaining the adsorption while the tablet material is in the mouth may be an alternative method of masking undesirable taste. In this case, the release of the active ingredient from the substrate is pH dependent, and occurs in the stomach or i n the intestinal tract. A conmercially available adsorba te is that of dextromethorphan adsorbed onto magnesium trisilicate by Roche (11). This product is available in a micronized powder with an active drug content of 10% (w/w). It is possible to form adsorbates from other materials such as Veegum and silica gel . This process requires that a drug be dissolved in a solvent, mixed with the substrate, and the solvent evaporated leaving drug adsorbed upon the substrate (5).
Ionic exchange, a method analogous to adsorption, has been used not only to mask taste, but also to enhance product stability . In this method, a substrate resin (ionic in nature) cationic or anionic, possessing an affinity for oppositely charged ions on the drug molecule is used. As in adsorption , the ideal complex is one in which the drug-resin complex remains intact under the salivary pH conditions, but capable of dissociating in the intestinal environment. Such a complex involving a cationic resin, Amberlite (substrate) and vitamin s 12 is commercially available from Roche as Stablets (11). As alluded to earlier, this complexation also enhances product stability. The resin-bound form of vitamin s 12 is more stable in the presence of the acidic vitamin C (a common combination in chewable multivitamin tablets).
The most widely used techniques for taste masking and stability improvement are spray congealing and spray coating . In spray congealing, particles are suspended in a molten coating material which is then pumped into a spray dryer in which cold air is cir culated. The droplets congeal upon con t act with the cold air and are then collected as free flowing particles. On the other hand, spray coating entails the evaporation of the solvent from the solution of the coating material. This leaves a film of coating material on the particles being coated. These two techniques produce spherical particles which are usually better flowing than the original form. Sodium dicloxacillin and tetracycline have been successfully spray coated to yield free-flowing products suitable for incorpo ration into chewable formulations (12). These products are coated with a mixture of ethylcellulose and sparmaceti dissolved in methylene chloride.
The formation of different salts of derivatives of a drug may be an alternative method of circumventing an unpleasant taste. However, this approach has legal limitations in that it creates a 11 new 11 drug entity .
Thus, subject to safety, efficacy, and stability testing as required in the Investigational New Drug (IND) and New Drug Application (NDA) guidelines of the FDA .
The traditional method of wet granulation also offers an approach of masking unpleasant taste in a chewable formulation. A bad tasting drug is granulated with or without ex~ipients with the aid of a binding solution.
This method however, requires higher concentrations of binder material (compared to formulation of non-chewable formulations) to ensure adequate coating. The conventional binder materials such as povidone, cellulose derivatives, polyethylene glycols , gelatin, acacia mucilage, and corn paste are used. It is reconnnended that where possible, the drug be granulated with a sweet binder, and an intra-granular disintegrant (5). The latter ensures that the granules disintegrate in· the gastrointestinal tract following mastication. The general limitations of wet granulation discussed in Section Ia also apply to this method.
A technique that has been used successfully to counteract the unpleasant taste of penicillin. is the use of amino acids and protein hydrolysates (13). Some of the preferred amino acids are: sarcosine, alanine, taurine, glutamic acid, and glycine (5). Among these amino acids, glycine is probably the most widely used.

a. Flavoring of chewable tablets
Since a majority of the chewable tablet formulations, are available as over-the-counter products, flavor and appearance can make the difference between commerical success or failure of a product.
The trend in the pharmaceutical industry today is shifting from the traditional usage of flavors such as mint, wintergreen, clove, eucalyptus, lemon, and orange just to render the product palatable. Todays flavoring encompasses a variety of attributes such as initial impact, mouth-feel, after-effects, and olfactory sensations. These considerations are made in an effort to produce a tablet with good taste and not merely to mask a bad taste. Some of the technology of pharmaceutical flavoring is borrowed from the food industry where a considerable amount of research work has been done (14)(15)(16) . The selection of the appropriate flavor in the pharmaceutical industry today is a well-researched, and executed process.  (18,19).
The incorporation of flavors into formulation mixtures generally involves blending of the flavor powder into the final blend, usually jus t prior to the addition of the lubricant. If several flavored base-line formulations are prepared, taste panels may be used to determine which production is finally presented to the market. However, the in-put of marketing personnel carmot be over-emphasized.
As indicated in the discussion above, the aesthetic nature of a product may actually make a ~otal difference between commercial success or failure of a chewable product. It is therefore important that the tablet product contain the most suitable color. and their matching flavors has been given by Daruwala (5).

Direct compression excipients
Due to special flow and compression requirements of direct compression, only a few matrices are suitable for the process. An extensive list of the properties of an ideal direct compression excipient has been given by Khan and Rhodes (21). Some of the most widely used direct compression excipients (matrices) include several forms of lactose, starch, microcrystalline cellulose, dicalcium phosphate, and sucrose. A detailed descrption of each of these matrices is given below.

a . Lactose
Available in a spray-dried form, lactose is the earliest and the most widely used direct compression matrix. The wide availability of this material from several sources makes it an attractive product. Initial results "1th spray-dried lactose were disappointing due to browning (22).

This reaction is catalyzed by tartrate, citrate, and acetate ions (23).
Advances in the spray-drying technique have eliminated the extent of browning considerably. Spray-dried lactose which has about 5% moisture content has less than satisfactory compressibility. However, studies have shown that tablets made from this material are not significantly affected by elevated temperature, high humidity or s unlight (24). The product's (lactose) poor compressibility and low dilution potential prompted

b. Starch
In its natural state, starch lacks the compressibility and flowability essential for a direct compression matrix. The modification that has received considerable industrial acceptance is that of Sta-Rx-1500. This is a partially hydrolyzed starch that has better flow properties than Starch USP. Although Sta-Rx-1500 can be compressed into tablets (25,26), its major application is as a disintegrant (27).

c. Microcrystalline cellulose
This form of cellulose was introduced as a direct compression matrix in the early 1960's following the isolation of cellulose fiber chains .
Under the trade name Avicel, microcrystalline cellulose is available in several grades, differing in particle size and flow characteristics.
Due to the presence of hydrogen bonding among the product's particles, excellent compressibility and high capacity can be achieved . Other favorable properties of this product include low bulk density, broad particle size range, and low coefficient of friction. Several studies concerning the use of microcrystalline cellulose in tableting have been reported in the literature (28)(29)(30)(31)(32)(33)(34)(35)(36)(37)(38)(39) Company. Its excellent flow and compressibility make it very suitable for use in high speed tablet presses. This product is relatively inexpensive, and possesses an acceptable degree of physical and chemical stability. Studies have shown that the material will lose water of hydration at elevated temperatures resulting in surface hardening (22).
Due to the product's poor solubility in neutral or alkaline media, its use in the formulation of low water soluble and alkaline sensitive drugs is limited. Accelerated stability studies have shown that this product is unstable in formulations containing acidic compounds, ascorbic acid and thiamine hy drochloride (41). Khan (44).

e . Sucrose
Sucrose in its raw form has poor flow, although it can be compressed into a solid mass. The lack of adequate flow has necessitated modifications to render the product adaptable to ·tablet production by direct

compression. The material ' s universal availability and abundance has led
to recent depression in its world market prices. The depression in s ugar prices has in turn led to an intensive effort to find new methods of using the product.
Successful modification of the basic s ucrose crystals is exemplified by DiPac and NuTab . These are co-crystallized products consisting of 97% sucrose with 3% modified dextrins, and processed sucrose with 4% inv.ert sugar; 0.1 to 0.2% corn starch, and magnesium stearate respectively (45,46). It may be noteworthy that the process of sugar refining involves It is thought that the DiPac agglomerate have a 'lacy' porous cluster of very small individual syrup coated crystals bonded together at their interface by points of contact. It is this porous nature that provides co-crystallized products with their tableting characteristics (47). The essential steps involved in the patented process of manufacturing DiPac are illustrated in Appendix 1.

The use of Tablet Presses in Powder Characterization
Instrumented tablet presses have been in use in the pharmaceutical industry for about three decades. The use of these instruments in the industry represents part of a continuing effort to convert tablet production from an emperical state into a systematic, and scientifically quantifiable process. St udies of tablet press instrumentation have encompassed both basic research, and the physics of tablet production.
In order to understand how instrumented tablet presses can be of value to a pharmaceutical researcher, an understanding of the process of manufacturing tablets by compaction is essential. The production of tablets involves the compaction of a powder system with an external force.
The powder is contained, and confined within the die cavity and the force is applied through the punches (upper and lower). The compression of the powder by the punches consolidates the powder material while displacing air from the void spaces. When force is applied to the powder in the die(s), bulk volume decreases through a number of mechanisms.
These mechanisms include: repackaging of the particles, elastic deformation (a reversable process), plastic deformation and or brittle fracture (when the elastic limit of the system is exceeded), and further compression.
The instrumentation of tablet presses is an attempt to measure the applied force. A commonly used instrument to determine such a quan tity is the transducer. The transducers used in tablet presses are those that convert force measurements to electrical signals in the form of voltage.
The two types of force measuring devices currently used are the st rain gauges and piezoelectric transducers. A strain gauge consists of a coil of high resistance wire mounted on a paper backing. This gauge shows a change in electrical resistance in direct p r oporation to its change in length. Piezoelectric transducers contain quartz crys t als which develop electrical charges proportional to the applied force . Strain gauges may be bonded to the punches or any other part of the press subject to similar changes. During compression , the applied force produces a small elastic deformation in the punches . This deformation in turn produces a change in the resistance of the strain gauges.
Piezoelectric transducers are available in many sizes, and may be placed in several locations of the tablet press . However, piezoelectric transducers offer the advantages of being less sensitive to temperature effects, and require no bonding to the tablet press. The major limitation of this type of transducers is that the change developed in them (as a result of applied force) dissipates over time, and thus· are not useful in static measurements.
The signals produced during the various cycles of the tablet press may be recorded by various devices such as oscilloscopes, recording oscillographs, and high speed chart recorders. The measurements recorded during a given run are compared to some standard or expected patterns, and may be used to control a particular phase of the tableting process .
Schwartz has described how instrumented tablet press data can be used as 'finger-prints' of tableting ingredients (53).
Whether a formulator uses a single punch or rotary instrumented tablet press, the data generated allows one to see how variables, such as specific surface area, apparent density, porosity. hardness (crushing strength), volume, thickness, disintegration time, dissolution, lowerpunch force, ejection force, etc., relate to compressional force. This information would indicate the force required to produce acceptable tablets. It would also be valuable in assessing how various tablet properties vary with different pressure levels, reproducibility of the process in use, materials' batch to batch variability and scale-up.
In order to facilitate rapid data collection on a large number of tablets, and to determine various compression profile parameters, instrumented tablet presses are interfaced with computers. The vast amount of data collected enables the generation of the entire compression profile.
Such profiles can provide a useful tool for trouble-shooting work.
Although a considerable amount of data has been generated with the use of instrumented tablet presses, there is no apparent uniform approach in its interpretation. Krycer and Pope (59) have demonstrated the existence of many discrepancies in the literature. These authors have also observed that there are serious conflicts in the conclusions drawn by researchers that employ similar methods. Nevertheless, recognizing that during tablet compression, the powder particles will und~rgo plastic deformation and/or brittle fracture before a compact is formed, the relative proportions of these changes in a formulation may be evaluated with the aid of 'Heckel plots' (60). The log of 1 / (1-d), where d is the relative density o f th2 compact, is plotted as a function of applied pressure . Heckel plots generated for variously sized fractions of a given material become superimposed when brittle fracture is the major effect,due to the rapid destruction of the original particle size.
These curves however, remain discrete when plastic deformation is the predominant process. The radial force (force transmitted radially to the die wall) to applied force profiles are also useful in assessing plastic deformation and brittle fracture. Lubricating properties are evaluated by measuring the residual force between the die wall and the tablet .

The generation of applied force versus punch displacement curves enables
the calculation of the work involved in the tableting process . This technique has been applied to evaluate lubricating efficiency (61, 62).  and (v) polishing-the finishing touch. A comprehensive review of the process of sugar coating, and some of its improvements has been made by Sutaria (64).

Tablet Coating
It is apparent that two factors threaten the utility of sugar coating in the pharmaceutical industry. First, the difficulty of automation of the process has severely limited its expansion (65,66). Second, only very few skilled individuals are able to perform the technique adequately.
The stability of sugar coated tablets has been shown to be less than satisfactory (67).
In some instances, this may be due to cracking of the coating as the tablet changes shape with variations in environmental conditions, o r when water migrates into the core . However, in some cases, the coating may harden considerably (wi thout cracking), thereby impeding prompt disintegration of the tablet core.

b. Film coating
Film coating is a relatively new te chnique compared to sugar coating. The first commercial (non-enteric) film coated product was introduced onto the market in 1954 (2). Film coating may be divided into two categories, namely enteric film coating and non-enteric; depending on the solubility of the coating material in gas trointestinal fluids.
The rapid development of this process may be attributed to the 20 vast interest not only in its pharmaceutical application, but also in other areas where surface coating is of concern . A number of considera-

tions have been cited as advantages of film coating over sugar coatinG·
These advantages include the reduction of coating time and material; lack of significant increase in tablet weight; no undercoat or water-proof coat required; durability and resistance to chipping and c racking; tablet can be monogrammed for identification; effective protection to light, air, and moisture; lack of adverse effects on disintegration time; nonaqeous coating solutions can be used; and feasibility of standardization of the process and materials.
In contrast to non-enteric film coating, en teric coating has been in existence for quite a long time (74). Enteric coating protects acid labile drugs from gastric fluids. It also protects gastric mucosa from irritating compounds. Courveur and co-workers (75), and Kanig (76), have described an ideal enteric coating material as one that is impermeable to gastric juices, susceptible to intestinal juices, non-reactive, stable during storage, provides a continuous coating, non-toxic, inexpensive, and easy to apply with minimum equipment.
Although the USP XX has disintegration specifications, and USP XXI will contain both disintegration and dissolution specifica tions for enteric coated tablets, these products will continue to be troublesome. Studies

have shown that even for those samples that pass in-vitro tests, success-
ful in-vivo performance can not be assured (75,78,79). This variability in performance is, at least in part, due to inter. and intra-subject gastric and intestinal pH variations.
Among the vast number of materials that have been evaluated or used for enteric coating, the most commonly used products are shellac and its cellulose acetate phthalate derivative (78). Studies have shown the superiority of cellulose acetate as an enteric coating product (80,81).
However, high hygroscopicity, susceptibility to hydrolytic breakdown upon storage, and permeability to ionic solutions are the major l imitations of this material (76). Other materials that have been successfully used recently are: copolymers of maleic anhydride and e thylene compounds (82),

and ionic polymers synthesized from methacrylic acid and methacrylic
esters (83) . These two parameters are often dependent on tablet hardness. In the evaluation of antacid tablets, acid neutralizing rates and acid neutralizing capacity (efficiency) are used in place of dissolution rates (86)(87)(88)(89). Some studies ·have shown a correlation between disintegration time and physiological activity (90,91) . However, other studies have shown that such a correlation does not exist (92). In r ecent years more emphasis has been placed on dissolution as the better in-vivo performance indicator. The correlation between dissolution rate and drug absorption has been documented (93)(94)(95) . Tests such as tablet thickness, hardness, friability, and resistance to impact s tr ess are unofficial.

Variability in tablet thickness is often indica tive of inadequate flow c haracteristics of a formulation. The other unofficial tests are indicators of the ability of a formulation to maintain its integrity
during packaging, shipping, and patient handling.

B. Accelerated stab ili t y testing
The single major purpose of accelera t ed stability testing is to permit one to predict the shelf life of a product at some label storage condition, usually "room temperature". While kinetic evaluation of stability data may be of value in most dosage forms, this approach is limited when applied to tablets . Due to the multip licity of excipients in tablet formulations, the use of rate constants through Arrhenius equation is impractical. Several approaches have been used to investigate the stability of pharmaceutical tablets. Chafetz has sugges t ed the use of stability indicating assays (96). Thes~ are procedures which afford the selective determination of a drug substance in the presence of its decomposition and reaction products.

Interactions between ingredients in a solid dosage form may lead to changes in physical properties such as equilibrium solubilities, partition coefficients, and dissolution. Guillory
and co-workers (97) have suggested control procedures to identify such interaction through the use of differential thermal analysis (98). Diffuse reflectance has also been shown to be of value in solid-solid interaction studies (98) . This technique enables the determination o f chemi-sorption due to surface chelation.

The literature reveals several publications of an empirical or practical nature describing incompatabilities, instabilities or other changes
in the solid state (99)(100)(101)(102)(103)(104)(105)(106)(107)(108)(109). In these studies, changes in parameters (non-chemical) as a function of time are reported. The moisture sorption and volume expansion of tablets made from various forms of lactose has been reported by Otuska and co-workers (110) . The moisture sorption and tablet expansion was shown to occur more readily with « -lactose leading to the formation of the monohydrate.
The effect of the container on the stability of pharmaceutical products is of considerable importance. Inter-tablet migration, a phenomenon which involves capillary condensation, exemplifies the importance of containers in product stability (111). The influence of environmental conditions on tablet properties is perhaps the most widely investigated area of tablet stability. The effect of light on color stability of tablet formul ations has been reported (112)(113)(114). Alans and Parrott (115) have shown a relationship between the changes occurring in the dissolution rate of hydrochlorothiazide tablets at elevated temperature and those occurring after prolonged storage at room temperature. The decrease in the dissolution rates of sodium salicylate (116), and acetaminophen (94) tablets due to aging has been shown. A study on phenylbutazone tablets showed a gradual decrease in the dissolution rate with aging (94).
Other valuable studies on the effect of aging on the physical properties of tablets include those of Rhodes and co-workers (117-120), Lachman (121), and Sangekar et al (122) .

C. Objectives and Significance of the Study
Although the production of tablets by direct compression has been used successfully for a number of years, the understanding of the process is still based substantially on empiricism. In particular, little has been done to rationalize the use of sugars as direct compression matrices.
Secondly, little consideration has been given to the effect of cyclic environmental changes on the stability of pharmaceutical products.
The prime objective of this study was to evaluate comparatively the potential utility of several oew direct compression sugar matrices .

Additionally, the effects of cyclic environmental changes on some fundamental physical properties of compressed tablets have heen investigated~
This study provides a rational approach to tablet formulation evaluation and stability analysis . The study is of both theoretical and practical importance not only to the pharmaceutical industry, but also to the food and chemical industries where packaging of raw materials and finished products is of paramount conc~rn.

II. EXPERIMENTAL
A. Materials

Intrinsic powder properties characterization
The following intrinsic properties of the matrices were determined: bulk and tap densities, moisture content (loss on drying), particle size distribution, and flow. Bulk volume was measured by pouring a 50g sample of powder into a 100 ml graduated cylinder from a height of approximately 2.5 cm. Three volume measurements were obtained for each matrix, and the values were averaged. Bulk density (gm/ml) was calculated as: weight (50g)

Formulation and evaluation of tablets made from sugar matrices
Several blends containing a matrix (C&H A, B, C, NuTab, and DiPac) and half a percent magnesium stearate were compressed into placebo tab - lets. An instrumented Stokes model F single punch press 8 , located at Based on data from the above mentioned study , formulation work was carried out using the following matriceS: C&H A, B, DiPac, and NuTab .
A brown matrix manufactured by C&H was also included in this series.
The formulas used to evaluate the formulation efficiency of the above matrices were : aspirin (pediatric), ascorbic acid, multivitamin (con- :.: The tablet press was instrumented with two piezoelectric tran sducers (one for each punch holder, upper and lower) and interfaced with an Apple IIElO The software to this computer enabled the calculation of peak height (compression curve) by picking signals off the press at a rate of 200 data points per second. A mean of peak height from data on ten tablets was calculated. Since the lower transducer was nonfunctional, no ejection force parameters were measured. In addition to the tablets made on the instrumented tablet presses, antacid , antitussive, and antidiarrhea tablets were prepared on a non-instrumented single punch press (located at the research laborato~ies of the Department To the above premix, ascorbic acid, flavor, microcrystalline cellulose, and matrix were added. The ingredients were blended for five minutes, and passed through the Fitzmill with #1 plate.
The lubricants were incorporated into a small portion of the above premix, and passed through a #20 mesh screen, and added to the remainder of the premix. The blend was then mixed for ten minutes and comp ressed using ~ in . standard concave tooling.

Ingredient
Ascorbic acid 90% 2   Aspirin was added to the matrix, and blended for five minutes. Starch and talc were added to the premix and mixed for five minutes .
Stearic acid was added to the above blend and mixed for five minutes.
Compression was carried out using 11/32 in. standard concave tooling.
(    was screened through #1 plate on the Fitzmill .
Stearic acid was added to the above blend and mixed for five minutes.
The tablets were comp ressed at 4 to 7 kg, using 3/4 inch standard concave t oo ling . Tablets were compressed at 4-7 kg, using 3/4 inch standard concave tooling.
(  The tablet sample to be tested was place in lOOml of one-ten t h normal hydrochloric acid (preheated to 37°C) contained in 250ml beaker. The 30°0 with 80% relative humidity-cyclic , and 30°0 with 80% relative humidity-constant. The samples exposed to the cyclic conditions were kept in a desiccator at 80% relative humidity in a 30°c oven for twelve hours. The bottles containing these samples were removed from the desiccator at the end of the twelve hours, and kept in room conditions for twelve hours. This cycle was repeated daily fo r three months. Those samples exposed to 30°C/80% relative humidity-constant , were kept in the desiccator (within the 30°c oven) throughout the test period .
At the end of the test period, the tablets were analyzed fo r both physical properties and acid neutralizing efficiency. Samples of tablets, twelve to a bottle , were stored in amber safety-lock~d bottles and exposed to one of the following : A -room temperature, approximately 20°C B -cyclic (12 hours 30°c , 12 hours room temperature) C -300C isothermal D -cyclic (12 hours 30°c with 90% relative humidity, 12 hours room temperature) E -30°C isothermal with 90% relative humidity .
The samples were kept in these conditio ns for up to four weeks.
The temperatures were controlled to+ i 0 c and relative humidity was controlled to :t 5% .
At the end of the four week storage period, disintegration times were determined for six tablets using the USP method, with discs. Dissolution profiles (plots of percent drug dissolves as a function of time) were obtained by the use of a six unit USP paddle dissolution apparatus in conjunction with an automatic ultra-violet spectrophotometer 21    sys t ems, s uch as was observed here, a change i n particle size spectra will result in changes in density , flow characteristics, and compressibility. Figure 3 is a plot of the particle s ize distribution for the matrices studied. The percentage (weight) of powder accumulating at the specified sieve aperture was plo t ted on a probability graph. This approach yields a linear plot compared to the sigmoidal curves that would result from the plot of cummulative amounts versus aperture size on linearlinear scale . Figure 4 shows the particle size dist r ibution for the two batches of C & HA (I and II) . C & HAI showed a mean particle size (corresponding to 50% point on the ordinate) of 250µm, whereas C & H All had a mean particle size of 150µm . This difference in particle size distribution may account for differences observed in both flow and compressibility of the two matrices.
As has been alluded to, particle size distribution of a direct compression system is o f profound importance. This is exemplified when the flow data obtained in this study are examined (Table X). C & HAI with a mean particle size of 250µm had a flow rate of 170 gm/sec, whereas C & H All, with a mean particle size of 150µm, had a flow rate of only 75 . 7 gm/sec.
In addition to intrinsic flow properties (unlubricated matrices alone), formulation mixtures as outlined in Tables I through VII were    C & H B shows the best overall flow characteristics.
In direct compression, the capacity, i . e. the amount of an active ingredient that can be incorporated into a matrix, without causing significant formulation problems, is of considerable importance. One of the properties of a formulation or powder system that may change with increase in percent active, is the flow behavior. Figure  However, an increase over thirty-five percent did not produce further reduction in flow rate. It may be noteworthy that these levels were chosen arbitarily around thirty percent (regarded as the average capacity of most direct compression vehicles) . Nevertheless, this level may be different with different active ingredients.

Morphology of bulk powders
Samples of the matrices were photographed using a scanning electron microscope.     SCANNING ELECTRON MICROGRAPH -EMDEX pressibility, while maintaining adequate flow.

Compression parameters and tablet properties
One of the objectives of this study was to use a computerized, instrumented tablet press to screen a series of new direct compression matrices.
In order to achieve this, the following compression curve parameters were generated: height (peak), inflection point ordinate, maximum slope, area under the curve, center, and width (at mid height). Three of the investigational matrices, C & H Al, B, and C, were lubricated with one-half a percent of magnesium stearate and compressed to obtain the above mentioned parameters. Figure 1 shows a typical compression curve. The parameters used to define the curve have been illustrated. The tablets were compressed at four compressional force levels.   (4) 0.99 (4) 0.99 (4)

Smax: H values. The three varieties of C & H products did not show
significant difference in Smax:H values (Table XIII). Chilamkurti (123) has also shown that there was no significant difference in this parameter for two types of dicalcium phosphate dihydrate. It is conceivable that this parameter is inherent of the chemical properties of a system.

Inflection point ordinate, slope (maximum), and area under the curve
varied linearly with applied force, Figures 12, 13, and 14 respectively.
Disintegration time appeared to be independent of tablet hardness ( Figure   15), probably due to the fact that these water soluble matrices form

This phenomenon of erosion has been compared to the peeling of an onion, one layer at a time.
Physical properties of the tablets made from the lubricated matrices (C & H AI, B, and C) are shown in Tables XIV-XVI. As expected from satisfactory flow data, weight uniformity for all the three matrices was excellent. Tablet thickness decreased with increase in applied force.

Both C & H AI and C & H B showed increases in friability when tableted
at high compressional forced (Table XV). This increase in friability was observed although hardness appeared to have increased with higher compressional forces . The hardness values used in Figure 16 were obtained immediately after compression, while those hardness values used in Figure 17 were obtained forty five days after manufacture. There appears to be an aging process in which a reduction in tablet harndess occurs.
All the matrices investigated except the Brown sugar,. were incorpo-   . c H c 10 20 35 rated into a pediatric aspirin formulation . The aspirin used was a 10% starch granulated type for direct compression. This type of aspirin was chosen over the crystalline material to enable the study of the formulation variables, such as compressional forces, on the release of an active ingredient from a sugar matrix (water soluble) where tablet disintegration occurs through a combination of matrix erosion, and actual particle disintegration (starch-aspirin granules). Table XVII and XVIII show the physical properties of the aspirin tablets made at different compressional forces . Figure 18 represents the effect of compressional forces on tablet hardness. It was observed that C & H AI produced the hardest tablets at all compressional forces .                     Figure 37 shows the effect of compressional force on tablet thickness. As expected, tablets thickness decreased with increasing compressional force. A special feature with a plot such as one presented in Figure 36 is that it enables one to visualize the compressibility of a powder system . Here, in a relative manner, it can be observed that at compressional forces greater than ten thousand Newtons, C & H All compressed the most, i.e.
had the greatest decrease in tablet thickness. The friability of the multivitamin tablet decreased with increase in compressional forces (Figure 38). At compressional levels below ten thousand Newtons, there was considerable difference in the friability of the tablets obtained from various matrices. However, when the tablets were compressed at compressional force greater than ten thousand New tons, no significant difference was observed among the friability values. Figure 39 shows the relationship between hardness and friability of the multivitamin tablets. The differences in friability were only observed at hardness levels below 4 kg. At levels where optimum hardness for most chewable formulations, approximately 7 kg, no significant difference in friability was observed. conditions. However, these products are bulky and may be inconvenient to carry. Tablets, particularly, chewable forms, offer an advantage over suspension (especially for pediatric and generiatric groups who find liq-

uids awkward or inconvenient). However~ it has been shown that tablets
are not equal to liquid antacids on a milligram-for-milligram basis (125) .
Although both the USP (1) and BP (85)  offers an example of a dynamic method where the pH of an acid medium is measured as a function of time.
In the analysis of acid neutralizing efficiency of the antacid tablets manufactured in the study, a method similar to the British Pharmacopeal guidelines for the analysis of aluminum hydroxide tablets was used.
This method is simple, and does not require elaborate equipment. Although it is recognized that as the tablet matrix erodes within the acid medium (intact tablet were used) a suspension-like system forms, and may affect the pH readings, the overall effect is negated as similar test conditions were used for samples tested before, and those tested after aging. Although coating processes such as the application of enteric coating to tablets have been used in industry for a long time, little consideration has been given to the adaptability of such coating to tropical-like con-

ditions.
A study was designed to evaluate the effect of moder ate stress conditions on the disintegration and dissolution of several coated tablet products (commercially available). Tables XXIII through XXVI show the disintegration of the tablets studied, before and after aging. A comparison of the dissolution data, evaluated by two methods, i . e . dissolution efficiency approach, and absorbance at various t"dime points, is shown in Tables .X.XVII through .XXX for prednisone (plain, film coated chlorpromazine, sugar coated chlorpromazine, and enteric coated aspirin respectively.          In comparing the dissolution profiles for the four products (before and after exposure to the various stress conditions), two methods were used . Firstly, the mean dissolution efficiency (six replicates), as defined by Khan and Rhodes (12 ) was determined, and the change in this parameter evaluated statistically using Student ' s t -test. Although this approach offers a convenient method whereby the entire dissolution profiles can be evaluated, "dissolution efficiency" is a derived value, rather than a raw experimental datum . Therefore, a second method of evaluating dissolution data was used. The second method involved a direct comparison of the percent d~ug dissolved at specific time points using Student's t-test (128).
Both disintegration and dissolution data (Table XXIII and Figure   49) for prednisone showed no significant change. The dissolution profiles (before and after exposure)were virtually superimpossible. This finding suggests that the product is reliable as a dissolution standard.
Results for the film coated chlorpromazine did not indicate any significant change in disintegration and dissolution characteristics. Although the Student's t-test usin g dissolution efficiency showed a significant change in dissolution for samples stored under 30°c constant, and 30°C/RH constant for two weeks, this difference was not detected when raw data were compared.

)
Both the disintegration and dissolution <lata for sugar coated chlorpromazine ( Table .XXV and Figure 51) showed significant changes as a result of storage. Examining the dissolution profile of this product, it appears that the aging changes are probably associated with the sugar coating rather than the tablet core .
The results of the evaluation of the enteric coated aspirin (Table   XXX and Figure 52) showed that the product was susceptable to significant aging. The dissolution of this product is extremely slow even for nonstressed samples. Although the only really conclusive proof that a bioavailability problem exists with enteric coated aspirin would be data from an in-vivo blood level study, these data suggest that enteric coated aspirin has potential for aging problems .

CONCLUSIONS
The following are believed to be the major points of this work as reported and discussed in the previous section.

The new sugar matrices produced by California and Hawaiian Sugar
Company (C & H) sh""' considerable potential as direct compression vehicles.
In particular, C & H B exhibits the t ype of flow and compressibility that are essential for high speed tablet production . This study demonstrates the importance of raw material specification.

Currently, the use of sugar matrices is mainly in the formulation
When two batches of C & H A (differing only in their particle size distribution spectra) were evaluated, significant differences were observed, both in their intrinsic properties and in their tableting characteristics.
Such differences are particularly critical in the production of tablets by direct compression, since little or no change in the material is af- a drug product could be exposed to . A commonly overlooked fact is that in no geographical region of the world do constant temperature and humidity prevail. The more natural situation, even in the tropics, is one in which temperature .and humidity may be high during the day, but drop con- A crys1alli zed sugar producl contai ning a heat-sensitive, acidic, o r high invert sug:ir suhs1ance is prepa red by admi xing 1he heat-sensitive, acid ic, o r hig h inve rt suga r substa nce with a dry suga r base to form a premix, concen trating a suga r syrup containing at least about 85% by weight sucrose to a solids conten t of abou t 95% to aboul 98% by heating 10 a tempc:rature of about 255• F . to abo ut Joo• F., mi xi ng the p rem ix wilh the concentrated sugar sy rup to fo rm a mix1ure, "ubjec ti ng the mixture 10 impact be:ui ng wi1hin a crystallization zone until a d ry c rystallized suga r product is formed, and recove ring the sugar produc t fro m !he cryst:1 lli za 1io n zone. The resulting sugar product compri~cs aggrega1es of fon dant-size sucrose crysla ls inlimately assoc ia led with the he3t-sensitive, acid ic, o r high invert sugar substance. The sugar product is d ry, gran ular, freefl o wing, non-caking, and readily dispersible in water. By mcan!i or the prc!.Cnt inventio n, a crys1a lli zed sugar producl is produced whic h incorporates a heats sensitive, acidic, o r high invert sugar subslance. The product is d ry, granular, free-flowing. and non-caking.

BACKGROUND OF THE INVENTION
The product is composed of agglomerates o r aggregates This invent io n rclate5 10 3 process o f producing 3 or minute, fondan1 -siie sucrose crys1als o r part icles intimately associa1ed wilh the active ingredicnl. Due to granula r, free-flowing, non-caking sugar inco rporalcd 10 its porous structure, the cryslallized sugar product is product. Mo re !.pcciri1.:ally, this invention relates to a readily dispersed or dissolved in water. crySl:dllizcd sugar product which incorpora les a heat- The crysta llized s ugar product of the prese nt invenscm.i1ive, acidic, or high invert sugar cont ent substance lion is prepared in a two-stage process. In the first stage, and 10 a process for making the sugar incorporated a premix is prepared by mixing a dry granular o r lransproduct.
15 formed sugar base wil h a hcaH•c nsitive, acidic, or high In !he manufacture of su gar products, a process inve rt sugar substance. In !he second o r coc rysta llizakn ow n as 1he tr:rnsformi~g process is. used 10 rroducc a tion stage, the crystallized sugar product is prepared by dry.' gr~nula r •. fre~-.nowmg: no n-ca king . sugar prod~ct concentrati ng a sugar syrup 10 about 95-98% by weigh t whic h 1s readily d1spcr~ed 1~ wale r. This transfo rming solids by heating at a lemperature in the range from pr~ess has been de~nbed m. U.S. Pat. Nos. 3,149,682 20 about i55" -30Ct F ., mixing the concentrated sugar (Tippens ct al.), 3,~65,331 (Miller et al.), and 4 ,159,~JO sy rup with a predetermined amoun t of the premix, sub-(Chen el a~.). In Tippens et al.. the method comp~ises jecting the new mixture ro impact healing withi n a co1~ccnt ra11ng_ a sugar sy rup 10 about 95-97% by "'.eight crysta lli zation zo ne until a crysta lli zed sugar product !>ollds by heat mg the sugar syrup to a temperature 1n the made u o f aggregates o f fondant-!>ize sucrose crysta ls ~ange of abou_t 250"-2b5" F., and immedia1ely subject-zs and th/heat-se nsitive, acid ic, o r high invert sugar sub-1~g .the . resuh1ng ~u pe~salurated sugar . syrup to a he~t stance is formed, lhe crys1allizt!d sugar producl having d1ss1pa11on opcralJOn s1mu l1aneously with vigorous ag1-a moisture conlen t of less than 2.5% by weight, and lat.ic:>"· The method produces a ~ry sugar pr~uct co m-recovering the crystallized sugar product from the crysprismg aggregat~ of fondant-s~ze (3-. 50. microns) s~tallization zone. If desired, the resulting crystallized cro~ c~ystals. M11l~r el _ al. describes a s1m1.lar process m JO suga r product may be dried to a moisture content ofless wh ich impact beat mg 1s used 10 crystallize the sugar than I% by weight, followed by screening to a uniform produc1 from lhe supersaturated sugar syrup.
size and packaging. The feed syrup which is used in the processes of Tippens et al. and Milh:r ct al. has a purity in the range DETAILED DESCRIPTION OF THE of 85-97% by weight sucrose. Thus, invert sugar (equal 35 INVENTIO N portions of g lucose and fru close), which has a tendency The accompanying drawing is a now chart illustratto cake, may not comprise more than about 15% by ing a preferred process or sc heme for preparing a crys· weight of the feed sugar syrup.
tallized sugar product in accordance wi1h 1he presen t The 1: ; ugar products prepared in accordance with the invention. processes of Tippens ct al. and Miller ct al. arc usefu l as 40 Referring 10 the now chart, the process oft he present ca rriers for food additives, such as colorants, n avoran ls, invention comprises two stages. In the first sta ge, a and pharmaceuticals. The food additives may be intro-premix containing an active ingredient is prepared. The duced into the suga r sy rup al either the concen tration active ingredic:nt 10 in a dry srnle is blended wi th a dry o r the crystalliz.ation stage of the processes, depending suga r base 12 suc h as a granular or transformed sugar, o n the nature of the additive. However, the high tern· 45 to form a dry premix 14. The ac1ive ingredient compe ratures used in the transforming process (about prises a heat se nsitive, acidic, or high invert suga r sub-250•-265• F.) restrict the nature of the food addit ives stance. For example, the active ingredient may be a heat which may be incorporated into 1he final sugar product.
sensitive substance, such as a volatile navor or an cn-Heat-sensitive ingredients, suc h as volatile navors o r zyme, or an acidic substance, such as Vitamin C (ascorenzymes. ca nnot be incorporn1cd in10 the sugar product 50 bic acid) o r a fruit juice concentrare, or a high invert by the methods described . Further, acidic ingredients, sugar substance, !>uch as honey or molasses. The dry such as Vitamin C o r fruit juices, ch:rnge sucrose into sugar base may be pure suc rose or may contain up to invert sugar by the reaction known as suga r inversion. about 15% by weight of non-sucrose so lids comp rising A further restriction in these processes is that the feed addiiional monosaccharides, disaccharides, o r modified syrup must conta in less than 15% by weigh t invert 55 dextrins. For e:tamp lc, the non -sucrose solids may comsugar.
prise invert sugar, dextrose, fructose, corn sy rup, malto-Accordingly. it is an object of this in ve ntion to pro-dcxtrins, or mixtures thereof. The amoun l and type of vide a sugar product which incorporates an edib le heat-sugar base which is used may va ry depending upon the sensi tive, acidic, or high invert sugar substance. amount and nature of the :ictive ingredient. llle active It is also an objec1 of this invention to provide this 60 ingredient is blended with lhe suga r base, fo r e>.amp lc, sugar product in granular, free -nowing, noncaking by means of a Hobart Olendcr, unti l the desired degree form.
of homogeneity of the premix is achieved. It is also an object 10 provide this sugar product in a In the second stage o f the ope ration, cocrys1allization form which is readily dispersed o r di!>solved in water.
of sugar wi th the active ingredie nt is achieved. A sugar It is a further object of this invenlion 10 provide a 65 sy rup 20 containi ng at leas! 85% sucrose is conce nmcthod of preparing this sug:ir product. trated by evaporation 22, under vacuum or under almo· These and o ther objects arc accomplished by means spheric pressure, at a temperature in the range of :lbout of 1he present invention dt"scribed below.
255°-Joo· F ., depending upon the nalure o f the aclive material, until the solids content of the concenlra ted dc1crioration of the active ingredient by !he high temsugar syrup exceeds about 95%. The non-sucrose so lids pcra1ure. The premix become!. thoroughly mixed in the ) in the feed syrup may comp rise additional monosaccha-earlier stages of the crystallization step as the concenrides, di!.accharides, or modified dextrins, for example, trated sy rup is transformed from the liquid slate 10 a invert sugar, dextrose, fructose, corn sy rup, maltodex-5 se mi-so lid state. Consequently, when the syrup reaches trins, or mixtures thereof. the relatively dry agglomerated stale, the resulting The resulting supersaturated suga r syrup 24 having a product is a homogeneous blend of 1he cocrystallized solids con1e n1 exceeding abou1 95% by weighr is main· sugar ;ind ac1ive ingredie nt. tained at a 1empcra1ure not less than aboul 240· F . in The phy'iical structure of the crystallized sugar prodorder to prevent pri:mature cryslallizatio n. A predetcr· JO uct is highly dl!pendent o n the raie and tcmpera1ure o f mined amoun t of the premix prepared in the first stage impac1 bea1ing and crys1a ll ization, and on the deg ree o f o f 1he proce~s is added 26 to the conccnlrated syrup suga r transformation. The optimu m time for the con· with vigorous mechanical agitalion or impac1 beating ccntraled sy rup mix1ure to spend in the crys1allit.ation 28 wi1hin a suitable crystalliza1ion zo ne , such as a J{o.
zone during impacl beating depends o n seve ral factors hart Mixer or Turbuli ze r. Alternatively, the concen-IS including: (a) the na1ure of the non-sut.:rose solids (such tra1ed syrup may be added to a predetermined amount as invert sugar and ao;h) in the syrup; (b) the nature and of lhe premix and mixed in a similar manner.
c haracteristics of the active ingredien t (such as moisture Impact beating is continued until the resulting super· conten t, invert sugar content p H, etc.); (c) the concen· saturated syrup is transformed, crystallized 30, and ag· !ration of the active ingredient in the premix: and (d) the glomcrnted 32. A crystalline sugar incorporated prod· 20 temperature used for concentration of 1he feed S)' rup. uct 34 is recovered from the crysta lli zat io n zone. The Jn structu re, the crys1a ll ized ~ugar products of the latent hc:at of crystalliza tio n is sufficie nt to evaporate present invention is comprised ofaggrega1es or agglom-1he moislure so that the product is substa ntially dry, i.e., cra tes of fond:lnt-l:iize sucrose crystals, e.g., in !he range has a moisture content o f less than about 2.5%. If de· o f about 3-SO microns, intimaidy associated with the si red, the crystalli zed sugar product 34 may be further 2S non-sucrose solids. The agglomerates form a loose. dried to a moisture conte nt of less than I%, followed by lacey network bonded together at their in terfaces by screening and packaging 36. point con tact. Accordingly, aqueous liquid can rapidly During crystalliz.at ion, it is desirable 10 remove the pcne1ra1e the porous cluster of agglomcr31C:S and free heat of crys1allization to prevent overheati ng within the each of the particles making up the agglomerates. The c rystall ization zone. The heat of crystallizatio n can be JO particles thus become readily dispersed and/or dis· removc:d o r dissipated by indirect heat excha nge e.g., by so lved in the aqueous liquid. surrounding the crystallizat ion zone with a water In 1he crystallized suga r product oft he present inven· jacke1, o r, preferably, by forced air now through the tion, the active ing redient is incorpor;;it ed as an in1 egral ) healer-crystallizer, e.g., with 3 va po r sepa rator. part o f the sugar matrix and !here is no tendency for the Suitable apparat us for car rying o ut the process of the JS active ingredien t 10 separate o r se ll le ou t during han· present invention is described in U.S. Pat. No. 3,365,331 dling, packaging, or storage. The resuhing product is (Miller et al.). granula r, free-newing, non -caking, and is readily dis· In o rder to ensu re maximum homogeneity in the final persed o r dissolved in water. Data from a typical analyproduct, it is desirable to introduce the premix into the sis of three different sugar incorpora1ed products pre· concent rated syrup as early in the process as practical. 40 pared in accordance with the present invention are However, in most cases, the premix is introduced dur· presented in Tab le I. ing the suga r crystallization step in order to prevent  A wide va riety of produc ls may be made in acco rdance wi1h the present invention. The following examples illust rate some embodimenls of this invention bUl are not meant in any way to limit the scope 1hereof.
A navored sugar product may be prepared by incorporating a Oavorant into a crystalli ne sugar matrix. The navorants include vo)a lil e na vors, such as acct.a ldehyde or diacctal, nonvolatile na vo rs, such as natural navor sugar solution at 65• Brix was heated 10 285• F. 10 form a ~upersa1urated sy rup of approximately 98% solids con1ent. The supersaluraled suga r solution was added 10 10 1he premix with impact beating. Impact beating was continued and crystallization proceeded unlil a dry powdered product was formed.
EXAMPLE 6 e"tracts or artificia l flavorings, and essen tial oils, suc h IS 150 grams o f grape juice concentrate (68" Brix) was as lemon oi l or peppermint oil. The product made in this mixed wi1h 350 gra ms o f sugar (Bakers Special Grade) manner provides a fast flavor-releasing character due to to form a slurry. The process continued as in Example the c rystalline sugar matrix.
5. The grape juice incorporaled producl ca n be used in EXAMPLE I a grape jelly mix formulation by dry blending wi1h IO. 7 20 grams of pecti n. 100 g rams of natural peppermint oil in a dry slate was ln another embodi men t, a vitamin, such as oxidative blended with 300 grams of granu lar sugar (Bake rs Spe-vitamin A,C,D,E, o r K, is incorporated into a sugar cia l Grade) using a Hobart Blender. At 1he same time, matrix. The resulting product is a homogeneous mixture 700 grams of a 65• Brix sugar solution wa~ concentra ted with high stabi lity. It can be used to fortify ot her foods. a1 260· F. to 95% by weight solid s conten1. 300 grams of 25 the peppe rmi nt oil-s ugar premix was added to the su-EXAMPLE 7 persa1urated, hot syrup with mechanical agi1a1ion by an JO grams of Vitamin A pa lmit a1e (Type 250-SD Ho ffimpact beater. Impact bea1ing conlinued until crys1a ll i-mann-LaRoche) was admixed wi1h 390 grams of transz.ation occ urred and a dry sugar product incorporating formed suga r (Di-Pac@) to form a premix. 6CX) grams the peppermint oil was produced .
JO of a heated, ~upe rsalurated sugar syrup. prepared as in EXAMPLE 2 Example I. was added to the premix wi th mechanical agitation. Stirring was continued until the sugar was A map le flavored sugar prod uct was prepared ac-transformed and agglomerated into a dry sugar product. cording 10 the process described in Examp le 1. 100 One g ram of this incorporated product providesexac1ly grams o f artificial maple fl avor (contain ing 2.5% maple JS 2,500 1.U. of Vitamin A. flavor, FMC6829) was dry-b ll!nded wi1h 300 grams of In ano1her embodimen t, a c hemical having beneficial granular sugar (Bakers Special Grade). 300 grams of the properties, such as ferrous sulfate, dicalcium phosphate, premix was added to the hot, supersa1 ura1ed syrup wit h sodium bicarbonate, or a trace mine ral is inco rporated impact healing until a dry product was formed.
into a sugar matrix. The product is a homogeneous In anot her embodiment, a high inve rt sugar sub-40 mixture of the ingredients and can be used 10 fo n ify slance, such as honey or molasses, is incorporated into a other foods. crystalline sugar matrix. The product made in this manner possi:sscs free-flowing and non-caki ng prope rties while retaini ng a natural delicate fla vor. EXAMPLE 3 EXAMPLE 8 2.08 grams o f stanno us nuoride was mixed with 4S 297 .92 g rams of iransfo rmed suga r (Di-Pac@) to form a premix. 600 grams o f a healed supersa1ura1ed sugar 200 grams of transformed suga r (D i-Pac@) was syrup, prepared as in Example I, was added to 1he preblcnded wi1h 200 grams of pure honey to form a slurry. mix with impact beating. Impact beating was con tinued 600 grams of a supersa1urated sugar syrup, prepared in and crystalliza1ion proceeded. eventually resulting in Examp le t. was 1hen added to rhe premix with agita-SO rhe formation ofa dry powdered product. Jn spi1e of the rion. Stirring was conlinucd un1il 1he mixture was trans-high chemical activity and acidi ty of 1he fluoride, this formed into a dry sugar product. The producl possesses c hemical was success fully incorpo rated inlo the sugar free-flowing charac1eristics and has a delicate honey matrix and the resulting product provides exact ly 1,()(X) taste.

EXAMPL E 9
A product was prepared as in Example 3 using molas-Example 8 was repeated except that JOO grams o f ses instend of honey.
ferrous sulfate was blended with 300 grams of suga r to In another embodiment, a dehydrated fruit juice form 1he premix. Sugar inversion by the sulfate was product is prepa red by inco rporating a fruit juice con-60 avoided due 10 the present process. The homogeneous ce ntra1e into a c ryslalline sugar matrix. The resulting iron product can be used to forti fy other foods. product is a free-flowing, nonperishable dry powder Jn another embodiment, a dry enzyme product o r an which can be used in dry blendi ng formulations. ac1ive culture is produced by incorpo rating an enzyme, EXAMPLE 5 such as invertasc, ce llulose, glu cose, isomerase, amy-6S la~e. catalase, glucose oxidase, lac1ase, o r pectinasc, o r 100 grams o f nalu ral apple juice co ncentrate an active culture, in10 a sugar matrix. Notwi1hstanding (65.Brix) was admixed with 400 grams of granular sugar 1he high 1empcrature of the process, the enz.y mc rc-(Bakers Special Grade) to form a slurry. 777 grams of mains in its active form.  8 Special Grade) to form a premix. 800 grams o f a healed, supersatu rated sugar syrup, prepared as in Example I, was added to the premix wilh impact heating. Impact bc:uing was continued and crys talli1.ation proceeded, t it@, Wallerstei n C om pany) wa3 mixed with 450 grams of transformed suga r (Di-Pac@) to form the premix. 600 grams of a hot supersaturated sugar sy rup, prepa red as in Example I. was added 10 1he premix wi th mechani-S eventua ll y resuhing in the formation o f a dry powdtrcd product.
ca l agi1a tio n. The agita tion was continued until the o; ugar was tran sfor med, crysta lli zed and agglomera ted . The incorpo rated prod uct ( 10 grams) was evaluat ed 10 wit h re3pect 10 its inverting ca pabilities by blend ing wi th various co ncentrations of liquid sugar {10-40 gra ms per 100 mis.) and incubated at 30• C. and 55" C. for 1. 5 hours. In spite of the hig h temperature used in the process, the experimen tal rcsu h s indica te tha t a IS significant po rtio n of the invc=rtase remained act ive. In anolhcr embodimen t, a nalura l coloran t, !<ouc h as annatto extracts, bee t juice conce ntrates, bcta-caro1ene, grnpc skin extracts, o leo resin paprika, or tumeric ex-trac1s, is incorporated inlo a suga r matrix. The incorpo-20 ralcd product is a homogeneo us, s1able, dry powder w hich shows no loss of color s1reng1h or hue and whic h ca n be used in dry blend formulalions.
EXAMP LE II 25 100 grams of tumeric color ( PT 8-S, Hansen Labora-1ory) was blended w ith 400 grams o f granular suga r (Bake rs Special Grade) to form a p rem ix. 500 grams of hea led, supersatura1ed sugar solution, prepared as in Examp le 1, was add ed to the prem ix with vigorous JO agi tation. The agitation was con1 inucd until all the sugar was cryii1a lli zed. The incorporated product was eva luated for color hue and for color strc:ngth (Bexin con tent). Resu lts showed no signi fii.;an t change in bot h c haracleristics despite exposure to lhe high lemperature JS cocrys1allizu1ion process.
Jn anot her embod iment, an acidu lcnt substance, suc h as malic acid , fumaric ac id , adipic acid, tartaric acid , cit ric acid, and sodium citrate, is incorporated into a suga r matrix . The resulting prod uct is a free-nowing 40 homogeneous powder w hich ca n be used in dry blend formulations.
In spi te o f 1hc high kmpcratu rc emp loyed in the presen t process, the resulting products are frcc-Oowing, non-caking, dry, homogeneous, stable, non-perishable, and are read il y d ispe rsed o r disso lved in wa ter.
While the invention has been dco;cribcd wit h re ference to specific embodimenls, these were for the pu rposes o f illustralion only and should nol be construc1ed to limit 1he scope o f the present invention.
We clai m: 1. A method for prepa ri ng a cocrystallized sugar produc1 co ntaining an active ingredienl ~elected from the group consisting of heat-sensitive, acidic, and high invert sugar substances, comprising: (a) adm ixing the active ingredient with a d ry sugar base to fo rm a prem ix; (b) concentrating a sugar syrup al a temperature in the range o f abou t 255• F. to abou t 3C~Y F . to a solids con lent o f abo ut 95% to 98% by weight, said sugar syrup con t:iinin g no more than abou t 15% by weigh t non-sucrose so lids; {c) d irectly adm ixing the concen 1rated sugar syr up at a lemperature 255• -300• F. wi th suid premix to fo rm a mixlure; (d) su bjecting sa id resul tin g mixture upon admixing said premix to impact beating wi thin a cryst<illization z.one until a crys1allized sugar product is formed, sa id c rystallized suga r product made up of aggregates of fondant-siz.e sucrose crystals and the active ingredient and having a moisture conten t of less tha n about 2.5% by weight ; and (e) recovering sa id c r ystallized sugar product from sa id c rys1alliza1ion zone. EXAMPLE 12 2. The method of claim 1 fu rlher comprising drying The process of Example 11 was repeated using cit ric 4S said crystallized sugar p roduct to a mois1ure con tent of acid instead of tumeric coloran t. less than about I% by weight. In another embodiment, an emulsifier, suc h as leci-3. A c rystalli zed sugar product made in accordance thin, mono-and dig lycericides, propylene glyco l esters, with the method of clai m 1. sorbi1an esters, polysorbate eslers, polyoxycthylene 4. The c rystallized suga r product of claim 3 wherei n 30rbitan es1ers, o r lac1yla1ed esters. is incorporated into SO !he ac1ive ingredient is a volat ile navor, a nonvolatile a sugar matrix . The c r ysta ll ized Product permits rapid fl avor. o r an essential oi l. dispersion of the emu lsifier in emu lsi fi ca tio n npp lica-5. The crysta lli zed sugar product o f c lai m 3 wherei n tio ns. The c rys1alli zed product, when added lo cake mix the ac1ive ingredient is honey. or icing mix, provides eJ;ce llent emulsion characteris-6. The crysla ll ized suga r product o f claim 3 wherein tics. For e>.amp le, ca ke volume. porosity, and appear-SS the :lCtive ingredient is molasses. ance; and icing stability and density arc imp roved wit h 7. The cryslallized sugar produc t o f claim 3 wherein the sugar inco rporated emulsi fi er as compared wit h an the active ingredient is a frui l juice. emulsifier added in the conventional manner.