G /vys/ts/ScrsiG r/e
Formulation factors affecting the binding properties of Chi
nese yam ( Dioscorea oppositifolia) and corn starches
Adenike Okunlola, Oluwatoyin A. Odeku
Department o f Pharmaceutics and Industrial Pharmacy, University o f Ibadan, Ibadan , Nigeria
Received June 25 , 2009; Accepted August 2 6 , 2009
Abstract
Objective;
The quantitative effects o f formulation and processing variables affecting the binding properties of Chinese yam starch ( Dioscorea oppositifolia) in chloroquine phosphate tablet formulations have been investigated in comparison with com starch using a 23 factorial experimental design.
Methods
: Chinese yam starch, representing the "lo w " level, and co m starch, representing the "h ig h " level were used as binders at concentra
tions of 2. 5 % w /w and 10 % w /w in chloroquine phosphate tablet formulations. The mechanical properties of the tablets, measured by the tensile strength ( T ) and brittle fracture index ( B F I) as well as the release prop
erties measured by the disintegration time ( D T ) and dissolution time ( t g0- time for 80 % drug release) , were used as assessment parameters.
Results;
The ranking o f the individual coefficient values for the formulations on T was D > N » C , on BFI was N > D » C , on DT was D > N > C and on t80 was C > N > D while the ranking of the interaction coefficient on T was N-D > C-D » N -C , on BFI was N-D > N-C = C -D , on DT and t80 was N-C > N-D > C-D. Changing the binding agent from Chinese to com starch, led to a decrease in T , DT and tg0 but increase in BFI o f the tablets. There were significant ( P < 0 . 0 0 1 ) interactions between the nature o f binder, N and the other two variables, C and D.Conclusion;
The result showed that Chinese yam possessed stronger binding capacity than com starch and could be useful as an alternative binder when tablets with high mechanical strength with minimal problems o f lamination, and slow release are required.Keywords;
Starch;
B inder; Chinese yam ;Tensile strength; Brittle fracture index; Chloroquine phosphateINTRODUCTION
Starches are o f commercial importance in the food and phamiaceutical industries because o f their inert
ness, abundance and cheapness. Commercial star
ches obtained from cereals ( co rn , waxy c o m , high amylose c o m , wheat and various rice s) , tubers and roots ( particularly potato and cassava) are used as diluents, binders and disintegrants in drug formula-
Correspondence lo ; Dr. Oluwatoyin A. Odeku, Department of Phar
maceutics & Industrial Pharmacy, Faculty of Pharmacy, University of Ibadan, Ibadan, Nigeria.
Tel; 2348057320466; 23428106403 E-mail; pejuodeku@ yahoo, com
tiont l] . With the versatility o f starches in various solid dosage forms, there is the need to continue to develop new starch excipients with suitable proper
ties to meet the special needs o f drug formulators.
Tropical yam tubers (Fam ily D ioscoreaceae) are potential source o f industrial starch because o f their high starch content ( about 50 - 80 % w /w ) ^ . One o f the species cultivated in many parts o f Africa and Southeast Asia is Dioscorea oppositifolia ( Chinese yam) , which has been used in Chinese medicine as an invigorant 3 . D. oppositifolia L. , (a ls o known as D. opposita Thunb) , is a perennial twining vine which has a d e e p , persistent root-like tuber that can be 1 m long and weigh up to several kilograms.
Aboveground, it has round slender stems that twine
© 2 0 0 9 . Asian Pacific Journal of Tropical Medicine. • 1 •
UNIVERSITY OF IBADAN LIBRARY
dextrorselyupwardsandleavesthatareusuallyar rangedoppositely.Thespeciesepithet"oppositifo lia"referstothisoppositearrangementofitsleaves. Recentstudiesonthephysicochemical,materialand tabletpropertiesofChineseyam starchhaveshown thatthevariedconsiderablyfrom officialcornstarch andstarchesfrom otherDioscoreaspecies[4].Chi neseyamstarchhasanamylose,protein,ash,lipid andphosphoruscontentof21.61,0.75,0.21, 020and0.025% respectively.Whenusedasadi rectlycompressibleexcipient,Chineseyam starch wasobservedtobehighlycompressible,formingtab letsofacceptablecrushingforce[4].Thecompact ibilityofthestarchappearedtoberelatedtothe granulesize,irregularshapeandhighspecificsur faceareawhichfacilitatemoreinterparticulatecon tactandthushigherbondformation[4].Furtherstud ieshaveshownthatthestarchcouldbeusefulatcer tain concentrations as disintegrants particularly wherehighmechanicalstrengthandagoodbalance betweenbindingandreleasepropertiesofthetablets arerequired[5].However,thequantitativeeffectsof importantformulationandprocessingvariablesonthe bindingpropertiesofChineseyam starchappearto be largely uninvestigated. Thus,in the present work,thequantitativeeffectsofformulationandpro cessingvariablesaffectingthebindingpropertiesof Chineseyam starch(Dioscoreaoppositifolia)have beeninvestigatedincomparisonwithcornstarchu singa23factorialexperimentaldesign[6].Thefacto rialexperimentaldesignhasbeenfoundusefulinthe analysisofthequantitativeindividualandinteraction effectsofvariousformulationfactorsonthemechani calstrength and drug release propertiesoftab lets[79].
Themechanicalpropertiesofthetabletwhichare indicatorsoftheabilityofthetabletstowithstandthe rigorsofhandlinginvolvedinmanufacture,transpor tation,dispensingandusage,hasbeenassessedby twoimportantproperties:bondstrengthandbrittle ness[7,10].Thetensilestrengthofatabletisafunc tionoftheareaofcontactbetweentheparticlesand thestrengthofbondsproducedbetweenthem[10]. Thegreaterthedegreeofbonding,thegreateristhe tensilestrengthofthetablets.Thebrittlefracturein dex(BFI)ontheotherhand,wasdevisedbyHies tandetal[11]andisobtainedbycomparingtheten silestrengthofatabletwithaholetothetensile strengthofthetabletswithoutahole,bothatthe samerelativedensity.Theholeservesasabuiltin stressconcentrator"defect".AlowvalueofBFIis
desirablefortheminimizationofcappingandlamina tionduringtabletproduction.
Chloroquinephosphate,a4aminoquinolineanti malarialcompoundwasusedasthemodeldrug. Chloroquinephosphatecannotbemadeintosatisfac torytabletsbecauseofitspoorcompressionproper tiesandthereforerequiresabinder,amongotherex cipients,toformsatisfactorytablets.
MATERIALSANDMETHODS
ThematerialsusedwerechloroquinephosphateBP andcornstarchBP(BDH ChemicalsLtd.,Poole, U.K.),polyvinylpyrrolidone,PVP Mol.Wt. 44,000(BDHChemicalsLtd.,Poole,U.K.)and tubersofDioscoreaoppositifolia(Chineseyam)ob tainedfromlocalfarmersinIbadan,Nigeriaandau thenticated.Thestarchwasextractedfromthetubers usingestablishedprocedures[12].
Preparationofgranulesandtablets
Batches(200g)ofabasicformulationofchloro quinephosphate(60%w/w),lactose(30%w/w) andcornstarch(10%w/w)weredrymixedfor5 minutesinaplanetarymixer(ModelA120,Hobart ManufacturingCo,U.K)andthenmoistenedwith appropriateamountsofstarchpastetoproducegran ulescontainingdifferentconcentrationsofthestar chesasbinder(2.510%w/w).Massingwascon tinuedfor5minutesandthewetmassesweregranu latedbypassingthem manuallythroughamesh12 sieve(1400μm).Theseweredriedinahotairov enat50℃ for18hours.Driedgranulesweresieved throughamesh16sieve(1000μm)andthen storedinairtightcontainer.Themoisturecontentof thegranuleswasdeterminedandfoundtobebetween 24% w/w.Particledensitiesweredeterminedby thepycnometermethod,usingxyleneasthedis placementfluid.
Preparationoftablets
Quantities(500 mg)ofgranuleswithinthesize rangeof7101000μmwerecompressedfor30sec ondsintotabletswithpredeterminedloadsonaCarv erhydraulicpress(modelC,Carverinc.Menom
oneeFalls,WIS)usinga10.5mm dieandflat facedpuncheslubricatedwitha1% dispersionof magnesiumstearateinacetonebeforeeachcompres sion.Afterejection,thetabletswerestoredoversil icagelfor24hourstoallowforelasticrecoveryand hardening.Theweights(w)anddimensionsofthe
· 2
·
UNIVERSITY OF IBADAN LIBRARY
2009.AsianPacificJournalofTropicalMedicine.
tabletswerethendeterminedtowithin ±1mgand 0.01mm respectively,andtheirrelativedensities, D,werecalculatedusingtheequation:
D = w/Vt.ρs (1) whereVtisthevolume,cm3,ofthetabletincluding theholewhenpresentandρsistheparticledensity, gcm3,ofthesolidmaterial.
Testing
Thetensilestrengthofthechloroquinephosphate tablets(T)wasdeterminedatroom temperatureby diametralcompression[13]using a hardnesstester
(KetanScientific& Chemicals,Ahmedebad,Indi a).Thetensilestrengthofthetabletwasthencalcu latedfromtheequation:
T=2L/πdh (2) whereTistheradialtensilestrength,Listheload neededtobreakthetablet,disthetabletdiameter andhisthetabletthickness[13].Resultsweretaken onlyfromtabletswhichsplitcleanlyintotwohalves withoutanysignoflamination.Allmeasurements weremadeinquadruplicatesandtheresultsgiven arethemeanofseveraldeterminations.
Thebrittlefractureindex(BFI)valuesofthetab letswerecalculatedfromtheequation:
BFI=0.5[(T/To)-1] (3) whereTisthetensilestrengthofthetabletwithouta holeandToisthetensilestrengthofatabletwitha hole.
Disintegrationanddissolutiontest
Thedisintegrationtimeofthetabletswasdetermined indistilledwaterat37±0.5℃ usingadisintegra tiontester(VeegoScientificdevices,Mumbai,Ma harashtra,India). Determinationswere done in quadruplicates.
Dissolutiontestwascarriedoutonthetabletsu singtheUSPXX IIIbasketmethod(HansonModel 72RL,U.S.A)rotatedat100rpm in900mLof 0.1M HCL,maintainedat37±0.5℃.Samples
(5mL)werewithdrawnandreplacedwithequala mountsoffreshmedium.Thesamplewasdiluted andtheamountofchloroquinephosphatereleased wasdeterminedatwavelengthof255nm,usinga UV/visiblespectrophotometer(PhillipsPye Uni
cam,PU 8610 Kinetics,Sarosescientificinstru ments,Cambridge,U.K.).Determinationswere doneintriplicates.
FactorialExperimentalDesign
Afactorialexperimentaldesignwasusedtostudythe effectsofnatureofbinder(denotedbyN),concen trationofbinder(denotedbyC)andrelativedensity
(denotedbyD)onthemechanicalandreleaseprop ertiesofchloroquinephosphatetablets[6].Thebasis
oftheexperimentaldesignwasthateachofthethree variableswasutilizedata"high"level(denotedby thesubscript,H)anda"low"level(denotedby thesubscript,L).Thenumberofexperimentsinthe designwas23=8.
Usingtheabovenomenclature,thevariouscombi nationsbetween thevariablesused in thedesign were:
NLCLDL,NLCLDH,NLCHDL,NLCHDH NHCHDH,NHCHDL,NHCLDH,NHCLDL NL= Natureofbinder(Chineseyamstarch) NH= Natureofbinder(cornstarch)
CL= Concentrationofbinder(2.5%)
CH= Concentrationofbinder(10%)
DL= Relativedensityof0.80 DH= Relativedensityof0.90
Bygroupingtheresultsintoanumberofsets,itwas possibletoassesstheeffectsthateachofthethree variableshadseparatelyonthemechanicalandre leasepropertiesofthetabletsandalsotodetermine whetherthevariableswereinteractingoractinginde pendentlyofeachother.
TheeffectsofincreasingN,from its"low"level toits"high"levelonthevariousparameterswere foundbysummingalltheresults(TorBFIorDTor t80)ofsamplescontaining"high"levelofN and subtractingthesumoftheresultsofsamplescontai ning"low"levelsofN.Thatis:
1/4[(NHCHDH+ NHCLDH+ NHCHDL+NHCLDL)
-(NLCLDL+NLCHDL+NLCLDH+NLCHDH)]
Theamountbywhichtheresultofthistreatment departedfrom zero,irrespectiveofwhetherpositive ornegativewasaquantitativemeasureoftheeffectof Nonthevaluesoftherelevantparameter.Similar expressionswereusedforfindingtheeffectsofCand D.
Todeterminewhethertherewasanyinteraction
· 3
·
UNIVERSITY OF IBADAN LIBRARY
betweentwovariables,theT(orBFIorDTort80) resultsofthecombinationsinwhichtheyappearto getherateither"high" or"low" levelswere summedandthesum ofothercombinationssubtrac tedfromthistoobtaintheinteractioncoefficient.For example,forNandC:
1/4[(NHCHDL+NHCHDH+NLCLDH+NLCLDL)
-(NHCLDH+NHCLDL+NLCHDL+NLCHDH)]
Aresultofzeroindicatesnointeraction,butifthe interactioncoefficientissignificantlyremovedfrom zero,itcanbeconcludedthatthetwovariablescon cernedareinteractingwitheachother.Theextentof removalfromzeroisameasureofinteraction.Simi larexpressionswereusedforestimatingtheinterac tionbetweenNandD,andbetweenCandD.
Statisticalanalysis
Statisticalanalysiswascarriedoutusingtheanalysis ofvariance(ANOVA)andTukeyKramermultiple comparisontestsonacomputersoftwareGraphPad Prism 4(GraphPad Software Inc. San Diego, USA).
RESULTS
Therepresentativeplotsoflogtensilestrength,brit tlefractureindex(BFI)and disintegration time
(DT)versusrelativedensity(RD)forchloroquine phosphatetabletscontaining10% w/wofthestarch bindingagentsarepresentedinFigures1,2and3 respectively,whilethedissolutionprofilesofthetab letsarepresentedinFigure4.ThevaluesofT,
BFI,DTandt80(thetimerequiredfor80% ofchlo roquinephosphatetobereleased)attwoselected relativedensityvaluesof0.8and0.9,whichare representativeoftherangeofrelativedensityvalues normallyachievedforcommercialchloroquinephos phatetablets,usedforthefactorialexperiment,are presentedinTable1.Thesevalueswereusedtocal culatetheindividualandinteractioncoefficientsfor thevariablesusingtherelevantequations.Theindi vidualandinteractioncoefficientvaluesarepresen tedinTable2.Thesevaluesprovideaclearindica tionofthequantitativeeffectsofthethreevariables studiedonthemechanicalstrengthandthedrugre leasepropertiesofthechloroquinephosphatetab lets.
Table1 Tensilestrength(T)brittlefractureindex(BFI),
disintegrationtime(DT)anddissolutiontime(t80)ofchloro quinephosphatetabletsforthefactorialexperimentaldesign.
Variablesand combination
codes
Tensile strength
(MNm-2)
Brittle fracture
index
Disintegration
time(min) t80(min)
NLCLDL 0.598 0.120 1.20 18.50 NLCLDH 1.286 0.136 9.25 36.25 NLCHDL 0.623 0.248 11.00 50.00 NLCHDH 1.967 0.100 26.25 83.50 NHCLDL 0.345 0.725 0.60 14.50 NHCLDH 0.930 0.399 6.60 31.00 NHCHDL 0.615 0.691 1.20 22.00 NHCHDH 1.155 0.307 7.30 37.00
Table2 Individualandinteractioneffectsofnatureofstarchbinder(N),concentrationofstarchbinder(C)andrelativedensity
(D),onthemechanicalanddrugreleasepropertiesofchloroquinephosphatetablets. Variables Tensilestrength
(MNm-2)
Brittle fractureindex
Disintegration time(min)
t80(min)
Independentcoefficient N -0.36 0.38 -8.00 -20.94 C 0.30 -0.01 7.03 23.06 D 0.79 -0.21 8.85 20.69 P P<0.001 P<0.001 P<0.01 P<0.01 Interactioncoefficient N-C -0.05 -0.06 -6.38 -16.31 N-D -0.23 -0.15 -2.80 -4.94 C-D 0.15 -0.06 1.83 3.56
P P<0.001 P<0.001 P<0.001 P<0.001
· 4
·
UNIVERSITY OF IBADAN LIBRARY
2009.AsianPacificJournalofTropicalMedicine.
Incomparingtheformulations,therankingofthe individual(independent)coefficientvaluesforthe formulationsonTwasD >NC,onBFIwasN > DC,onDTwasD >N >Candont80wasC>N
>D whiletherankingoftheinteractioncoefficient onTwasND> CD NC,onBFIwasND > NC =CD,onDTandt80wasNC>ND>CD.
Thereweresignificant(P<0.001)interactionsbe tweenthenatureofbinder,Nandtheothertwovari ables,CandD.
Figure1PlotofLogarithmofTensilestrengthversusrelative densityforchloroquinephosphatetabletscontaining10%w/
wofstarchbinderwithhole(……)andwithoutahole(__
__)attheircenter:■,Chineseyamstarchand▲,corn starch.
DISCUSSION
TheinfluenceofNonT,DTandt80wasnegativein dicatingthatchangingthebindingagentfrom Chi neseyam starchwhichrepresentedthe"low"level tocornstarchwhichrepresentsthe"high",ledtoa decreaseinT,DTandt80butincreaseinBFIofthe tablets.Thus,Chineseyam starchproducedtablets withhighertensilestrengthandlongerdisintegration anddissolutiontimesthanthetabletscontainingcorn starchasbindingagent.TheinfluenceofNonBFI
Figure2Plotofbrittlefractureindexversusrelativedensity forchloroquinephosphatetabletscontaining10% w/w of binder:■,Chineseyamstarchand▲,cornstarch.
Figure3PlotofDisintegrationtimeversusrelativedensity forchloroquinephosphatetabletscontaining10% w/w of binder:■,Chineseyamstarchand▲,cornstarch.
waspositiveindicatingthatChineseyamstarchbind erproducedtabletswithlowerBFIvaluesthancorn starch.Thisindicatesthatthetabletscontaining Chinesestarchhasbetterabilitytorelievestressand preventcappingandlaminationduringtabletproduc tionthantabletscontainingcornstarch.Thus,Chi neseyamstarchappearstobeastrongerbindinga
· 5
·
UNIVERSITY OF IBADAN LIBRARY
Figure4Dissolutionprofilesofchloroquinephosphatetablets containing10% w/wofstarchesasbindersatrelativedensi tyof0.80:■,Chinese;▲,Corn.
gentthancornstarch.However,allthetabletscom
pliedwiththeBP[14]requirementondisintegration
(i.e.disintegrationwithin15minutes)exceptfor mulationscontaining10%w/wChineseyam starch which had a disintegration time of26 minutes. Thus,Chineseyamstarchcouldbemoreusefulasa bindingagentwhenstrongermechanicalproperties andlongerdisintegrationanddissolutiontimeisre quired.
TheeffectsofCindicatethatincreasingthecon centrationofbinderfrom2.5to10% w/wledtoan increaseinthetensilestrength,disintegrationand dissolutiontimesofthetablet.However,theeffect ofContheBFIwasnegative,indicatingthatthere wasadecreaseinthetabletbrittleness.Ithasbeen shownthattheheatproducedduringcompression wouldcausemeltingoftheasperitiesand ofthe binderswhich on cooling would solidify to form strongsolidbondsbetweentheparticles[15].Thea
mountofbondingwoulddependontheconcentration ofbindingagentpresent.ThedecreaseinBFIvalues withincreaseinCindicatesthatthepresenceofa binderatinterparticulatepointsfacilitatesplasticde formationforthereliefoflocalizedstresses.Thus, theconcentrationofthestarchbinderhastobecare fullyselectedtoobtaintabletswithadequatestrength and with desired disintegration/dissolution times. Thisisinagreementwiththeresultsofotherwork
ers[15,16].
Therelativedensity(D)hadthehighesteffecton tensilestrengthanddisintegrationtimebuttheleast effectondissolutiontime.Thus,increasingtherela tivedensityfrom0.80to0.90ledtoanincreasein T,DTandt80,butadecreaseinBFI.Formost pharmaceuticalmaterial,therelativedensityofthe tabletsgenerallyincreaseswithincreaseincompres sionpressure.Thiswillleadtoadecreaseinporosi tyorreductionincapillarymicrostructureofthetab lets[17]resultingincloserpacking.Thus,moresolid bondswillbeformedbetweentheparticlesleadingto anincreaseinthemechanicalstrengthofthetablets. Particlerearrangement,fragmentationanddeforma tionmayalsoresultintheclosureoftheinterandin tragranularporespaces,therebyreducingthecapil larymicrostructureoftabletsandthereductioninthe sizeofthecapillaryspacesbetweentheparticleswill reducetherateofpenetrationofwater,whichwillin turnleadtoanincreaseindisintegrationanddissolu tiontimes. Thereweresignificant(P<0.01) differencesintheeffectsoftheindividualvariables onthemechanicalandreleasepropertiesofchloro quinephosphatetablets.
Theinteractioncoefficientvalues(Table2)indi catethatallthevariablesinteractedwitheachother toalterthemechanicalandreleasepropertiesofthe tablets.TheinteractionbetweenN andD hadthe largesteffectonmechanicalpropertiesofthetablets whiletheinteractionbetweenN andChadthelar gesteffectondisintegrationanddissolutiontimesof thetablets.Thus,thenatureandconcentrationof bindingagentusedintabletformulationsneedtobe carefullychoseninordertoobtaintabletsofdesired mechanicalanddrugreleaseproperties.Ingeneral, therewerestatisticallysignificant(P<0.001)in teractionsbetweenthenatureofbinder,N andthe othervariables,CandD.TheeffectofNontablet propertieswereshowntobestronglyrelatedtothe concentrationofthestarchbinder(C)andtherela tivedensityofthetablet(D)employedintheformu lation[8].Thus,thereappearstobeastronginterac
tionbetweenChineseyam starchandtheotherfor mulation and processing variables. Furthermore, changingthetypeofstarchbinderfromChineseyam starchtocornstarch,andincreasingthebindercon centration(C),led to an increase in tensile strength,disintegrationanddissolutiontimes,buta
· 6
·