`
`I9!
`Vol. 4 - 1981 Pharmaceutisch Weekblad Scientific Edition
`
`944
`
`Studies on different dissolution models
`
`IV. Erosion of tablets
`
`p. DE HAAN' AND C.F.1.ERI(I'2
`
`ABSTRACT
`A comparative study of the action of erosion on non
`disintegrating theophylline tablets was performed in the
`three use xx dissolution models. the rotating basket. the
`paddle and the modified disintegration apparatus. as well
`as in a new (sandwich) model. The results show erosion to
`be dependent on the mechanical strength of the tablets and
`to be promoted by the phenomenon of softening of the
`tablets during the process of dissolution. A considerable
`and moderate erosion was found for ‘soft' tablets in the
`rotating basket and disintegration apparatus. respectively.
`whereas no or only slight erosion was observed in the
`paddle and sandwich models. The phenomenon of dissol—
`ution intensified erosion and erosion intensified dissol-
`ution is shown to affect the description of the release
`profiles. The dissolution profiles of the ‘hard' tablets in the
`models preferred to be described by the square root
`equation. whereas the profiles of the 'soft' tablets in the
`case of erosion showed a tendency to a better fit with the
`cube root of mass versus time relation.
`(Pharm. Weekbl. Sci. Ed. 4, [9I - 196)
`
`lNTRODUCTlON
`
`A fair amount of research has been published
`regarding in vitro dissolution rate of sustained
`release
`preparations
`(STEINBACH and MdLLEtt
`1977a,b; MOLLER and sramnacr-t [979; Guum et al.
`1971; JONKMAN er al. 1981). The studies often aimed
`to be correlated with in viva absorption rate studies.
`or were used for quality control.
`In vitra test
`procedures should perform relevant discrimination.
`reproducibility and closely simulate in viva dissol-
`ution mechanisms.
`
`In vitro dissolution testing of oral solid dosage
`forms is mainly performed in a stirred vessel design.
`Of the existing methods the rotating basket and the
`paddle method. as described in the use xx. are
`favoured because of world wide standardization.
`Fluid agitation is created in these models by the
`rotating basket and paddle respectively. Although
`agitation conditions should be relatively mild.
`in
`both methods mechanical stress incorporates the
`possibility of the cause of erosion of the solid dosage
`form during the dissolution process. The phenom-
`enon may have consequences on the in vitro release
`results for matrix tablets (MoLLEn and STEINBACH
`1979; MdLLt-ztt er a1. 1980) and for granules (CARSTEN-
`SEN er a1. 1978).
`This article reports a comparative evaluation of
`the action of erosion on soluble non disintegrating
`
`three use xx dissolution designs,
`tablets «in all
`apparatus 1. 2 and 3, being the rotating basket. the
`paddle and the disintegration apparatus. as well as in
`a new model, to be used as a reference. The effect of
`erosion during dissolution of the tablets on the fit
`with the cube root and the square root of mass is
`discussed.
`
`MATERIALS AND METHODS
`
`MateriaLr
`Theophylline monohydrate (Pit. Eur. grade) was ob—
`tained from acr Chemiepharma NV. Maarssen. The
`Netherlands. and was sieved to a fraction <90 urn.
`
`Methods
`
`Tablet compression. Theophylline monohydrate tablets
`were prepared by compressing 250 mg of the drug in a
`9 mm die-set. mounted between the plattens of an instru~
`mented hydraulic press (Hydro Mooi. Appingedam. The
`Netherlands). The applied compression force was 20 t i
`Mi for the ‘hard‘ tablets (poroSity t.4%) and 850 t 30 N
`for the ‘soft‘ tablets (porosity 28.4%). The loading rate
`was a kN.s ' and 85 N.s ". respectively. The ‘hard‘ tablets
`showed no change in release when stored. at least for one
`month, under normal
`laboratory conditions. This in
`contrast to the ‘soft‘ tablets. Therefore the latter were
`stored for 65 hours at 50% relative humidity before using.
`The crushing strength of the tablets was measured using a
`motorized Schleuniger 2E instrument. The data given are
`the mean values of three determinations.
`
`Drug release. The release measurements were performed
`in all
`three use xx dissolution models and in a new
`assembly. which will be referred to as the sandwich model
`(Fig.
`i). The model consisted of it Mine flat-bottom
`beaker.
`t6 cm high and 9 cm in diameter. fitted with a
`straight four bladed turbine impeller. 5 cm in diameter.
`with a width of l cm. fixed to a rotating shaft with a
`diameter of 6 mm. and a sandwich-holder in which a tablet
`was held by a spring (Fig. 1). After placing the tablet
`carefully in the holder. the holder assembly was positioned
`on the bottom of the beaker and the impeller was centrally
`placed 5.5 cm from the bottom of the beaker. The centre of
`the tablet was positioned 1 cm above the middle of the
`bottom of the beaker.
`including the rotating
`In all vessels/stirrer models.
`basket apparatus. rotation speeds were too rpm. The
`beakers were filled with 900 ml of either 0.1 N 1-10 or a
`saturated solution of theophylline in 0.1 N HCl. The
`models were thermostatically controlled at 37 i o.5°C.
`Dissolution profiles were determined by taking samples at
`
`for Pharmaceutical Technology and Dispensing. University of Groningen. Ant. Deusinglaan 2. 9713 aw
`' Laboratory
`Groningen. The Netherlands.
`1 Reprints and correspondence.
`
`ENDO - Ex. 2031
`
`Amneal v. Endo
`
`|PR2014-00360
`
`ENDO - Ex. 2031
`Amneal v. Endo
`IPR2014-00360
`
`

`

`
`
`192
`
`
`
`945
`
`
`VoL 4 - 1981 Pharmaceutisch Weekblad Scientific Edition
`
`
`
`FIGURE t. Impeller and sandwichholder (left). and the positioning in the flabbottom beaker (right)
`
`through membrane filters
`time intervals
`convenient
`(0.2 pm pore diameter) and analyzed. after dilution with
`0.1 N HCl. spectrophotometrically at 270 nm (Beckman
`model 25).
`in saturated theophylline solution were
`Erosions
`measured gravimetrically by drying the residual tablet at
`too-105°C to constant weight. Corrections were m'ade for
`the part of theophylline from absorbed and adhering
`medium and for the conversion of theophylline monm
`hydrate into the anhydrous form.
`The experiments were performed at least'1n triplicate
`Error barsin the figures represent the standard deviation.
`In all determinations performed. no visible disintegration
`of the tablets was'present.
`
`RESULTS AND DISCUSSION
`
`Erosion without dissolution of non disintegrating
`tablets
`
`The possibility of erosion of a tablet by a mechan-
`ical or hydrodynamic action of the model used. was
`examined by performing the ‘dissolution' test in a
`saturated solution of theophylline in 0. t N HCl. The
`results showed that no weight loss could be detected
`
`for the ‘hard’ tablets. whereas for the ‘soft' tablets a
`considerable erosion was found in the rotating
`basket apparatus a moderate erosion in the disin-
`tegration apparatus but no measurable erosion in the
`paddle and sandwich models (Table 1) Conse-
`quently. the paddle and sandwich models can be con‘
`sidered as models without (mechanical) erosion for
`the flat tablets tested. which do not move during de‘
`termination.
`
`Dissolution and erosion of non disintegrating
`tablets
`
`Figure 2 illustrates the release profiles of the hard
`tablets in all four dissolution models. The results
`
`show the highest dissolution rate in the disinte-
`gration apparatus, and the lowest dissolution rate in
`the paddle model. It should be noted. however. that
`the effective surface of a tablet for dissolution is
`
`smaller in the paddle model as compared with the
`sandwich model.
`
`The release profiles of the soft tablets are illus-
`trated in Figure'3. Compared with the hard tablets
`the dissolution rate has increased. The reproduci-
`
`raaLe 1. Mutual interference of erosion and dissolution (in percent released i so)
`
`Min.
`
`Medium
`
`
`Rotating basket
`Paddle method
`
`Sandwich method
`
`Disintegration apparatus
`
`sr'
`1-11;
`Diff.
`sr
`HT
`Diff.
`sr
`HT
`Diff.
`sr
`111'
`Diff.
`
`
`15
`30
`
`-
`—
`—
`—
`-
`-
`-
`——
`—
`—
`—
`28.83505
`saturated
`4.1
`0‘
`4 i
`0
`0‘
`o‘
`o
`o‘
`0‘
`47.4
`0‘
`47. 4+1. 1
`saturated
`1.2 38.91.24 338:” 5.1 65811.4 54. 6+1 .3 11. 2
`69714.5 394:1. 8 3o.3 34.01:3. 1 3:.8+1. 4
`dissolution
`— —1.3:3.5
`-
`-
`8 2+2 5
`-
`68 6+94
`-
`—
`0. 1:1.6
`saturated
`60
`
`
`
`- o.3to.8 — 0.211.8 —o.tto.3 — 04:12 0.0304 — 20. 8+1 .9 0.410.“)saturated120 —
`
`
`
`
`
`
`
`' Soft tablets.
`Hard tablets.
`
`‘ Difference between percentage weight 105: of the soft and the
`‘ Value estimated from t a 110 min.
`Value estimated from t a 60 min.
`
`hard tabletsto the same medium
`
`

`

`
`
`946
`
`
`
`I93
`Vol. 4 - 1982 Phannacezuisch Weekblad Scientific Edition
`10'
`
`
`
`PercentageDISSO‘VOd
`
`\
`
`0
`
`so
`
`TIME (mus)
`
`FIGURE 2. Dissolution profiles of the hard tablets. O rotating basket model;
`El disintegration apparatus; 0 paddle model; A sandwich model.
`
`bility of the dissolution profiles. as shown by the
`error bars in the figures 2 and 3, decreased except for
`the disintegration apparatus.
`To obtain a clear idea of the increase in drug
`release. the differences in release between the soft
`and the hard tablets. measured in the same model at
`the same time, are plotted in Figure 4. From this
`
`that no significant
`figure it can be concluded,
`(P <0.05) increase in release rate was exhibited by
`the soft tablets. compared with the hard tablets.
`when tested in the paddle model; in the sandwich
`model only 'a slight increase was observed. whereas a
`considerable increase was shown in the disinte-
`g'ration and the rotating basket apparatus. Con-
`
`.”
`
`I
`
`/"
`
`
`
`. ./"
`
`so
`
`i,
`
`3
`
`E5?
`
`Ez
`
`?
`
`0
`
`so
`
`so
`
`co
`
`120
`
`TIME (mitts)
`
`FIGURE 3. Dissolution profiles of the soft tablets. Symbols as in Figure 2
`
`

`

`
`
`947
`
`
`
`Vol. 4 - r982 Phannaceutisc‘lt Weekblad Scientific Edition
`I94
`C.
`
`
`
`
`
`Dilerereepercentagereleased
`
`3
`
`
`
`l (mates)
`
`noun: 4. Differences between dissolution profiles of soft
`and hard tablets. Symbols as in Figure 2
`
`coming the rotating basket apparatus, it should be
`noted, however,
`that the increased release rates
`were found to be reproducible in the same appar-
`atus. but not reproducible for different baskets and
`different motor and shaft assemblies. This agrees
`with the results of camenseu er a1. (1980), who
`reported significant differences in release results.
`when shaft length was varied in a basket and a paddle
`design. investigating a system inwhich both disin-
`tegration and dissolution were concerned.
`Moreover,
`it
`is noted that particles, dislodged
`from the tablet, may be subject to different agitation
`intensities in the model. This is especially the case in
`the rotating basket apparatus, where particles may
`fall out of the basket and may consequently be
`submitted to lower agitation intensities (TINGSTAD er
`a1. 1973). In our study this effect will not result in a
`significant change in dissolution rate of the tablets.
`because it has been shown that 250 mg of the
`theophylline powder (<90 pm) dissolved within
`2 minutes in all the models used.
`
`To differentiate between the phenomena of
`erosion and dissolution, the percent weight loss in
`saturated solution and the percent released in 0.1 N
`
`TABLE 11. Effect of wetting on crushing strength
`
`Porosity (9b)
`
`Crushing strength i so (kg)
`
`
`dry
`after wetting
`
`10.3 $1.2
`17.6 $0.6
`1.40
`7.7 $0.3
`14.7 11.7
`2.85
`6.4 to;
`12.3 $0.4
`4.98
`3.4 $0.2
`11.4 1-0.:
`10.25
`2.1 10.2
`7.1 10.1
`16.91
`—
`3.1 $0.:
`25.63
`
`
`2.1 $0.128.40 -
`
`I-ICl are given in Table l for both the soft and hard
`tablets after a ‘release' time of 30 minutes in all four
`dissolution models.
`Evaluation of the data found in the rotating
`basket. 1‘. e. no mechanical erosion of the hard tablets
`but 47.4% erosion for the soft tablets when tested in
`saturated solution, compared with release rates of
`39.4% and 69.7% for the hard and soft tablets.
`respectively, indicate the action of erosion intensi‘
`fied dissolution for the soft tablets.
`
`Concerning the paddle and sandwich models, both
`the hard and soft tablets exhibited no mechanical
`
`erosion in saturated medium. and showed at most a
`small increase in dissolution rate for the soft tablets
`
`when compared with the hard tablets. In the disin-
`tegration model a mechanical erosion is found of
`about 4% for the soft tablets, together with an
`increase in release, as compared with the hard
`tablets. of about 11%.
`These results point to the occurrence of dissolu-
`tion intensified erosion, which phenomenon gener-
`ally has to be considered as contributing to an
`increase in release rate. Both physical abrasion of
`the tablets by the solvent flow and the phenomenon
`of mechanical erosion of the (soft) tablets in the
`dissolution models are intensified by the softening of
`the solids when exposed to a dissolution medium.
`Table 11 shows the effect ofwetting on the crushing
`strength of theophylline tablets with a different
`porosity (after storage for 65 h at 50% relative
`humidity). Wetting was performed over.a period of
`30 minutes with a quantity of water approximately
`sufficient to fill all the pores of the tablet. The results
`show a considerable decrease in crushing strength of
`the tablets after wetting. Tablets with a porosity of
`about 25% and higher showed no mechanical re-
`sistance after wetting. Dissolution of the contacts
`between the particles in the tablet may be a factor
`contributing to the process of tablet softening.
`or. BLAEY and VAN 01-211 GRAAFF (1977) reported a
`definite failure of the linear relationship between
`retreat rate of ali‘near dimension and the reciprocal
`tablet density with porosities of about 20% and
`higher. This result might be explained by the
`phenome'non of softening with consequent erosion
`of the sodium chloride disks in the centrifugal stirrer
`design.
`In conclusion. softening and erosion have to be
`recognized as phenomena that can strongly affect the
`release of drugs from solid dosage forms.
`
`Comparative evaluation of the dissolution patterns
`Frequently dissolution profiles are presented as
`Hixson-Crowell cube root plots. 1'. e. (Wo'B—W'”) is
`plotted versus time, where Wu and W are the
`amounts undissolved at time zero and at time t
`respectively. In those cases. however, where the
`data do not fit the cube root equation. many other
`plotting models for tablets are used including the
`square root plot.‘ which is not based on a physical
`
`

`

`948
`
`ama"bum)
`
`I95
`
`
`
`Hm")
`
` Jn
`
`s.
`
`m
`
`' ("”‘l
`
`a
`
`m
`
`V out
`
`.l.
`
`a 4 4 .
`
`.
`
`n
`
`s 4 a l 1 I
`
`-l
`-:
`-:|
`.4
`.,
`
`I
`
`art-«na.mud)
`
`nun-muMI!)
`
`and.1-IIIn.hull
`
`1 (nut.
`
`l (n...)
`
`2 5
`
`'.
`
`M0
`
`
`
`awn-n{‘mvmn)
`
`
`
`.mmu-(sun-4M4”
`
`FIGURE 5. Deviation of the dissolution profiles from regression lines. both calculated according to the square root (closed
`symbols) and the cube root expression (open symbols). 0.0: rotating basket model; ID: disintegration apparatus;
`0,0: paddle model; A.A: sandwich model. To the left: hard tablets; to the right: soft tablets
`
`

`

`
`
`Vol. 4 - 1982 Pharmaceutical! Weekblad Scientific Edition
`I96
`
`
`
`
`
`949
`
`model (DE 31.11151 and VAN DER GRAAFF 1977).
`The hard and soft tablets tested did not meet the
`requirements under which the cube root law applies.
`The shape of the tablets in the models was retained
`fairly well after a dissolution of about 50% of the
`mass. With proceeding dissolution. however, a
`visible change in shape of the tablets was observed in
`all models. except the paddle design. In the sandwich
`model the tablets became oval (the long axis in a
`direct line with the impeller axis) during dissolution.
`as could be expected from the flow patterns in a
`stirrer/vessel design (LAGAS and LERK 1977). The
`tablet tumbled out of the sandwich holder after a
`weight loss of about 94%.
`Figure 5 shows the mean of the deviant behaviour
`of each dissolution profile with respect to its re-
`gression line for a dissolution up to 50% of mass.
`both calculated according the cube root and the
`square root respectively, expressed as a percentage
`of the original tablet weight. The data presented in
`this way permit tracing of a systematically deviant
`behaviour from both kinetic expressions applied.
`Except for the paddle model, a tendency to a better
`description of the dissolution kinetics in the models
`for the hard tablets by the square root and the soft
`tablets by the cube root expression. is present. The
`change in systematically deviant behaviour of the
`plots in relation to its regression lines from one type
`of tablet to the other is most obvious for the erosive
`
`dissolution apparatus tested, can roughly be grouped
`into the non- or at most slightly erosive paddle and
`sandwich designs. and the moderate to highly eros-
`ive disintegration and rotating basket models. More-
`over it is shown. that the release profiles are strongly
`affected by the phenomenon of erosion during
`dissolution.
`
`REFERENCES
`
`31.1.1511. c.1. DE. and 11. VAN DER GMAFF (1977) J. Pharm.
`Sci. 66. 1696.
`CARSTENSEN. 1.1.. 1.1.. wmcm. K. BLESSEL and 1. snmmu
`(1978) J. Pharm. Sci. 67. 982.
`cxnmnsen. 1.1.. a. normal. v.11. 11on and 1. 51-115mm»:
`(1980) J. Pharm. Sci. 69. 290.
`GUMMA. A.. 1-1. mass and ILA. RAMSAY (1971) Pharm. Ind.
`33. 291-
`JONKMAN. 1.11.o.. 11. scnoeumxu. N. numeric and 11.11.
`DE zeeuw (1981) Intern. J. Pharm. 8. 153.
`14:28. 11.. and c.1=. LERK (1977) Pharm Weekbl. 112.
`5.
`Mount. 11.. and o. srsmaacn (1979) Pharm. 213. 124.
`1207.
`MOLLER. 11.. 5.1.. ALI and o. STEINBACH (1980) Intern. J.
`Pharm. 7. 157.
`1111111011. 5.1... D.E. wunsren and r. 111ouc111 (1955) .I. Am.
`Phann. Assoc, Sci. Ed. 44. 269.
`snowmen. 1).. and 11. MOLLER (1977a) Pharm. Ztg. 122.
`507: Ibidem (1977b) 122. 2067.
`111165110. 1.. E. 01101121511. 1.. LACHMAN and [-2. 5mm (1973)
`J. Pharm. Sci. 62. 293.
`
`basket design. For the paddle model. on the contra-
`ry. no change is perceptible.
`The results presented indicate.
`
`the four
`
`that
`
`Received April 1982.
`Accepted for publication September 1982.
`
`Drug release from non-aqueous suspensions. 1. Release of phenobarbital
`and phenobarbital sodium from paraffin suspensions. Erratum
`
`1.6. FOKKENS AND CJ. DE BLAEY
`
`In the paper ‘Drug release from non-aqueous
`suspensions. 1. Release of phenobarbital and pheno—
`barbital sodium from paraffin suspensions‘, pub-
`lished in Pharm. Weekbl. Sci. Ed. (1982)4. 117-121.
`
`the two photographs forming Figure 4 (page 120)
`were printed upside down. The dark part of the
`photographs. indicating the paraffin layer. should
`have appeared On top.
`
`