Application of Thermal Analysis to Study the Compatibility

8/9/2019 Application of Thermal Analysis to Study the Compatibility

1/12

REV. CHIM. (Bucharest) ♦ 62♦ No. 4 ♦ 2011 http://www.revistadechimie.ro 443

Application of Thermal Analysis to Study the Compatibility

of Sodium Diclofenac with Different Pharmaceutical Excipients

BOGDAN TITA,1* ADRIANA FULIAS,1 GEZA BANDUR,2 IONUT LEDETI,1DUMITRU TITA 1

1University of Medicine and Pharmacy “Victor Babeº”, Faculty of Pharmacy, Eftimie Murgu Square 2, Timiºoara, 300041, Timisoara,Romania2 Politehnica University of Timiºoara, Industrial Chemistry and Environmental Engineering Faculty, 2 Victoriei Square, 300006,Timiºoara, Romania

Thermal analysis is a routine method for analysis of drugs and substances of pharmaceutical interest.Thermogravimetry / derivative thermogravimetry (TG/DTG) and differential scanning calorimetry (DSC) are thermoanalytical methods which offer important information about the physical properties of drugs (stability,compatibility, phase transitions, polymorphism, kinetic analysis etc). In the present work, TG/DTG and DSC were used as screening techniques for assessing the compatibility between sodium diclofenac (DC) and its physical association as binary mixtures with some common excipients. Based on their frequent use in pharmacy, several different excipients as: starch, microcrystalline cellulose (PH101 and PH102), colloidal silicon dioxide, lactose (monohydrate and anhydre), polyvinylpyrrolidone (povidone K30 or PVP), magnesium stearate and talc were blended with DC. Samples were prepared by mixing the analyte and excipients in a proportion of 1:1 (w:w). In order to investigate the possible interactions between the components, the TG/

DTG and DSC curves of DC and each selected excipient were compared with those of binary mixtures, inorder to evaluate any possible solid state modification. The Fourier transformed infrared spectroscopy (FT- IR) and X-ray powder diffractometry (XRPD) were used as complementary techniques to adequately implement and assist in interpretation of the thermal results. On the basis of DSC results, confirmed by FT– IR and X-ray analyses, sodium diclofenac was found to be incompatible with lactose monohydrate, respectively anhydre, povidone K30 and magnesium stearate.

Keywords: sodium diclofenac, thermal analysis, compatibility, excipient-drug interaction

* email: [email protected]

Diclofenac sodium {2-[(2,6-dichlorophenyl)amino-

phenyl]acetate} is a potent non-steroidal anti-inflammatory drug (NSAID), therapeutically used ininflammatory and painful diseases of rheumatic and non-rheumatic origin. The anti-inflammatory activity of diclofenac and most of its other pharmacological effectsare related to the inhibition of the conversion of arachidonicacid to prostaglandins, which are mediators of theinflammatory process. Diclofenac is a potent inhibitor of cyclo-oxygenase in vitro and in vivo, thereby decreasingthe synthesis of prostaglandins, prostacyclin, andthromboxane products [1,2]. The structural formula forsodium diclofenac is shown in figure 1.

Incompatibility between drugs and excipients can alter

stability of drugs, thereby, affecting its safety and/orefficacy. Drug-excipient compatibility testing at an early stage helps in the selection of excipients that increase theprobability of developing a stable dosage-form. In particular,the low availability of drug and the time constraintsassociated with the early stages of formulationdevelopment have made such predictability particularly desirable [7–11].

Despite the importance of drug-excipient compatibility testing, there is no universally accepted protocol for thispurpose. The term thermal analysis refers to a group of techniques in which a physical property of a substanceand/or a reaction product is measured as a function of temperature whilst the substance is subjected to acontrolled temperature program [12–15].

In our previous papers we provided the importance of the thermal analysis in estimation on the thermal behaviourof different pharmaceuticals, respectively their possibleinteraction with excipients [16–24].

Differential scanning calorimeter (DSC) techniqueinvolves the application of a heating or a cooling signal to asample and a reference, can evaluate the energy associated with various thermal events (e.g., melting, glasstransition temperature, crystallization, etc). This methodhas been extensively reported in the literature for testingcompatibility of excipients with number of drugs [13–15,25]. The use of DSC has been proposed as a rapid

method for evaluating the physico-chemical interactionbetween two components. However, the caution needs tobe exercised in the interpretation of DSC results. This is

Fig.1. The chemicalstructure of the

sodium diclofenac

The structure of diclofenac consists of a phenylaceticacid group, a secondary amino group, and a phenyl ring,both ortho positions of which are occupied by chlorineatoms. Moser et al. [3] studied 36 congeners of diclofenacas inhibitors of cyclooxygenase and the in vivo inhibition of rat adjuvant arthritis and found that both activities can be

explained by lipophilicity and twisting of the two aromaticrings. These findings allowed the rationalization of the highactivity of diclofenac [4–6].


2/12


3/12


The DTA curve of diclofenac (for β=10°C·min–1) presentsa sharp endothermic event at 84.4°C which corresponds tothe dehydration ( ∆ H

fus=491.1 J . g-1). Further, it occurs the

melting process (T pea k DTA

=285.9°C), which is anendothermic process and this process is followedimmediately by an exothermic process corresponding tothe decomposition process.

Compatibility study with excipientsThermal behaviour of the mentioned excipients is more

or less known that in this paper it was studied the thermalbehaviour of the correspondent mixtures. For this purpose,the thermal curves of DC and excipients were compared with the curves obtained for 1:1 ( w:w) physical mixtures.

Figures 3–5 show the TG, DTG and DSC curves of thesubstances used in the compatibility study. Each curveshows a specific behaviour depending on thecharacteristics of each excipient.

The TG/DTG curves of starch show a dehydrationbetween 33–120°C (∆ m=7.2%; DTG peak=65°C), followedby the process of decomposition between 295–375°C( DTG

peak=325°C; ∆ m=79.7%). Initially the DSC curve

exhibits a wide endothermic peak representing dehydration(T

peak=94°C) [9,12,29].

The thermal behaviour of microcrystalline cellulose PH-101 and respectively PH-102 is the same. Absorbed water(about 5%) is lost below 110°C, between 35 and 110°C,apparently in a single, endothermic and spread-out process

Fig.3. TG curves of all substancesused in compatibility study

Fig.4. DTG curves of allsubstances used incompatibility study

Fig.5. DSC curves of allsubstances used incompatibility study


4/12

REV. CHIM. (Bucharest) ♦ 62♦ No. 4 ♦ 2011http://www.revistadechimie.ro446

( DSC pe ak

=72°C). No other thermal phenomena areobserved before the beginning of decomposition, between307 and 385°C ( DT G

pe ak=355°C and ∆ m=88%),

respectively DSC peak

=320°C [9,11,30].In the case of the colloidal silicon dioxide, on the

thermoanalytical curves, no peak was observed in therange of 25–500°C [9,11,31].

The amorphous form of lactose was identified by thepresence of an exothermic peak at 167°C, which

represented the transformation of amorphous to crystallineform. It is followed by two endothermic peaks, one at 210and the other at 216°C. These melting peaks belong toalpha and beta-lactose respectively. It confirmed thetransformation of the amorphous form of lactose to thetwo types of crystalline form by heating [31–33].

The 100% crystalline lactose, according to XRPD,contains α and β forms.

According to the thermogram, the water-content(∆ m=4.5%) ofα-lactose monohydrate is evolved between100 and 170°C ( DT G

pe ak=161°C). The water-free

compound is stable up to about 265°C, then it decomposesup to 365°C and DTG

peak=315°C. The DSC curve shows a

first sharp endothermic peak (T peak

=145°C) correspondingto the dehydration reaction, followed by two endothermicpeaks, from the first sharp endothermic peak( DSC

peak=215°C), which corresponds to the melting of α-

lactose, the second weak peak ( DSC pe ak

=224°C)represents the melting of β-lactose [9, 11,32,33].

On the DSC curve, the β-lactose presents a smallendothermic peak (T

peak=145°C) with an insignificant

mass loss on the TG curve, followed by two peaks, the firstlight corresponds to the melting ofα-lactose (T

peak=215°C),

respectively the second represent the melting of the β-lactose (T

peak=224°C). The decomposition process takes

place in the temperature range of 275 and 365°C(DTGpeak=312°C), accompanied by an endothermic event

on the DSC curve (T peak=318°C) [9,11,33,34].The TG/DSC curves of PVP, below 150°C display on initialmass loss of ≈9%. This mass loss is accompanied by abroad endothermic phenomena ( DSC

peak=82°C) over an

ill-defined baseline which makes evaluation of thedehydration enthalpy quite uncertain. The sample readily dehydrates and its initial mass depends upon the moisturecontent of the atmosphere. Apparently, dehydration iscompleted at 110°C ( DTG

peak=164°C) in N2. However, a

second loss stage (≈2%) begins past 150°C and completesaround 250°C. Thermal analysis, SEM and XRPD all showthat the compound is in a vitreous phase with glasstransition near 200°C. Decomposition begins around 384°C( DTG

peak=442°C, ∆ m=86%) up to 485°C [30,35–37].

Simultaneous TG/DSC curves of magnesium stearateshow several dehydration stages below 110°C. The firstendothermic effect is due to the release of a small amountof surface water. Around 50°C begins the first dehydrationstage of structural water, which partially overlaps with asecond stage at higher temperature. The overall mass lossdue to surface water and to the first stage is ≈3%, whilethe amplitude of the second stage is≈1.5% of the initialmass. DSC curve of magnesium stearate initially show

wide endothermic effec t (T pea k=75°C), representingdehydration. Melting begin at ≈110°C and produce anendothermic peak with a shoulder in the high temperatureside which is caused by melting of magnesium palmitateor high-melting polymorphs. The decomposition of thesample begin around 311°C ( DTG

peak=362°C) and to 480°C,

92.5% of sample mass is lost. Corresponding to thedecomposition process, the DSC curve presents a sharpendothermic with T

max=372°C [9,30,33,34,37].

The TG/DTG and DSC curves of talc present any significant events under the conditions in the present work[9,11,33,34,37].

The experimental data obtained for each excipientcorrespond to those from the speciality literature[9,11,12,29–37], which confirms the purity of substancesand the correctness of the used methods.

TG, DTG and DSC curves of the pure diclofenac and the1:1 drug:excipient physical mixture are shown in figures6–8.

In the 1:1 physical mixtures when there is no any interaction between drug and excipient the T

peak value of

melting event (DSC curve) and the first stage of thedecomposition (T

onset and T

peak of TG/DTG curves) should

remain practically unchanged, similarly when the drug isalone. In this case the thermal profiles of the mixture canbe considered as a superposition of the curves of thediclofenac and excipients. Thus, in the DSC curve of DC

and mixtures, the T peak value of melting, a referenceconstant, is the same. According to the thermal curves (fig.6–8), especially

DSC curves that provide the most complete information, itis found some smaller or larger differences (the case of the mixtures with α-lactose, β-lactose, PVP and MS)regards to the melting temperature values and those of the dehydration, respectively of the thermal decompositionranges. Basically, all the other excipients present somedifferences, however small, on the melting temperature,dehydration temperature, respectively the value of thedehydration enthalpies (table 1). These differences may be due to the small interactions that have not beenconfirmed by FTIR spectroscopy and X-ray diffraction

patterns.

Fig.6. TG curves of diclofenac and its 1:1

physical mixtures


5/12


6/12


Table 1

THERMOANALYTICAL DATA OF SODIUM DICLOFENAC AND DRUG:EXCIPIENT PHYSICAL MIXTURES

- a strong C=O stretching vibrations. The bands from1453 and 1393 cm–1 correspond to the scissoring vibrationof the CH2 group adjacent to the carbonyl. The doublet from1305 and 1284 cm-1 corresponds to C–O stretching band,as well as to C–N stretching absorption, together with C–H

bend (in plane) from aromatic ring. The strong bands from769, respectively 747 cm-1 correspond to the out-of-planeC–H bend.

The spectra of lactose monohydrate, respectively lactose anhydre, are virtually identical with the observation

that the lactose monohydrate shows a sharp medium bandat 3528 cm -1 due to the vibration of O–H bond of crystallization’ water. The main bands appear at:

- 3380 and 3343 cm-1, as strong and large band (3500–3000 cm-1) attributed to O–H stretch: intermolecular

hydrogen bonded;- a triplet at 2977, 2933 and 2900 cm-1 that correspondsto the C–H stretch: methylene;

- the range of 1420–1330 cm-1 corresponding to the O–H bending vibrations (in plane).

Fig.9. IR spectra of lactose monohydrate, DCand 1:1 blend as simple mixture of DC and

lactose monohydrate

Fig.10. IR spectra of lactose anhydre, DCand 1:1 blend as simple mixture of DC and

lactose anhydre


7/12


- the range of 1260–100 cm-1 which characterises thestretching vibrations C–O (in fact C–O–C).

- 892 and 876 cm-1 that corresponds to out-of-plane C–H bend.

In respect of the povidone, it presents the followingbands, at:

- 3460 cm–1 – a large band attributed to the OH groupfrom the crystallization water;

- 2977 cm–1 – that corresponds to the C=O bending;- 1669 cm–1 – that corresponds to the carbonyl amidic

group;- 1495; 1465; 1422 cm –1 – these correspond to

asymmetrical vibration (δas

CH3);- 1291 cm–1 – that corresponds to the in-plane C–H

bending.Magnesium stearate presents a weak and large band in

the region 3600-3100 cm-1 (the maximum at 3421 cm-1).

At 2918 and 2850 cm-1, there were observed two sharpbands with maximum absorption due to the CH2–CH3 vibrations. In the 1570–1468 cm-1 region, it showed anasymmetric stretch corresponding to the carboxyl anion.

Other bands that must be maintained have their peaksat 2956 cm-1 corresponding to the asymmetrical vibrationof C–H bond in methyl group, respectively those at 721cm-1 which correspond to “rocking” deformation (H–C–H)n; n>3.

For the binary mixture with α-lactose, there wereshowed the following differences:

- the disappearance of the bands from 3528, 2978, 2933and 2900 cm-1, for the α-lactose spectrum.

- the bands at 3464 cm-1 (DC), respectively 3380-3340cm-1 (α-lactose) are greatly enlarged, corresponding to the3650-2500 cm-1 range.

- the bands from 1652–452 cm-1 corresponding to DC,respectively those from 1655–403 cm-1 corresponding to

Fig.11. IR spectra of PVP, DC and 1:1 blend assimple mixture of DC and PVP

Fig.12. IR spectra of MS, DC and 1:1 blend assimple mixture of DC and MS

α-lactose, are grouped into three areas: 1653-1250 cm-1;1250-915 cm-1 and 915-404 cm-1, from which the last twoin particular are in the form of two wide bands with a lownumber for so-called maximum and a lot of shoulders. Formost of the maximums, the intensity is not significantly reduced.

In the case of the mixture with β-lactose, were foundsimilar differences as for α-lactose’s mixture. Thus:

- the broad bands for DC (3464 cm -1) and for β-lactose(3454–3293 cm-1) are greatly enlarged, corresponding tothe range of 3650–2500 cm-1.

- the triplet bands at 2977; 2901 and 2878 cm-1 from thespectrum of β-lactose disappears.

- the bands from the range: 1652–452 cm-1 (DC) and1600–418 cm-1 (β-lactose) are grouped into three areas:1600-1200 cm-1; 1200-870 cm-1 and 870-418 cm-1. In thiscase, the first area is in the form of broad bands too, with

less of so-called maximum than the summation of thetwo spectra (DC and β-lactose).For the mixture of DC with PVP, it is found that:- the broad bands at 3464 cm-1 (DC) and 3447 cm-1 (PVP)

are more wider than those corresponding to the range of 3700–2700 cm-1.

- the relatively large band corresponding to themaximum 2955-2886 cm-1 from the spectrum of PVPdisappears.

- the intense band at 1662 cm-1 from the PVP’s spectrumdisappears.

- the bands from the range: 1604–1284 cm -1 (DC) and1496–1293 cm-1 (PVP) are reduced in number and they are in the form of a wide band with multiple maximums.

- also the bands from 770–452 cm-1 range (DC) and 747–406 cm-1 (PVP) do not return to baseline in the case of themixture, forming a band almost as wide at the top as thebase.


8/12


In the case of DC-MS mixture, the main change is the

fact that the doublet at 2918 and 2850 cm -1 of MS spectra with maximum absorption disappears. At the same time,the two broad bands of 3464 cm -1 (DC), respectively 3421cm-1 (MS) are greatly enlarged (3650–2700 cm-1) and they include the doublet which was mentioned above, but theirintensity is reduced only with 10-15%. Also, the absorptionbands at 1571 and 1468 cm -1 in the MS spectrum(absorption significantly as approx. 75%) together with the1576–1393 cm-1 (MS) are found as a broad band with asmall number of maximums. In the same way, the bandsat 770-408 cm-1 are of the form of a band with the base wide as the top. The FT–IR spectra from the figures 9–12indicate some chemical interactions between DC and

mentioned excipients.To investigate the possible interaction of diclofenac withα-lactose, β-lactose, povidone and magnesium stearate,besides the FT-IR spectroscopy which is a qualitativeanalysis technique, the X-ray powder diffraction has been

Fig.13. X-ray diffractogram of lactose monohydrate, DC and 1:1blend as simple mixture of DC and lactose monohydrate.

Fig.16. X-ray diffractogram of MS, DC and 1:1 blend as simplemixture of DC and MS

Fig.15. X-ray diffractogram of PVP, DC and 1:1 blend as simplemixture of DC and PVP

Fig.14. X-ray diffractogram of lactose anhydre, DC and 1:1 blend assimple mixture of DC and lactose anhydre.

used for qualitative and quantitative identification of

crystallinity [31,32,35,36,39,41]. The X-ray diffractionpatterns of sodium diclofenac, α-lactose, β–lactose,povidone, magnesium stearate and of the binary mixturesare shown in figures 13–16.

The additional prominent DSC peaks in the mixtures of the drugs and excipients are a positive indication of chemical interaction of the drugs with excipients. Suchinteraction should result in the partial or completedisappearance of the reactant phases and appearance of new phases, which can be inferred from X-ray diffractionpatterns. X-ray diffraction patterns of the mixture, preparedat room temperature, when compared with those of itsindividual components showed appearance of new linesand disappearance of some of the lines present in theindividual components.

The X-ray patterns of diclofenac–α-lactose mixtureprepared at room temperature shows the lines in additionto those present in patterns of the individual components


9/12


Table 2

X-RAY DIFFRACTION DATA FOR DICLOFENAC,LACTOSE MONOHYDRATE AND

DICLOFENAC– LACTOSE MONOHYDRATE (1:1)MIXTURE

Table 3

X-RAY DIFFRACTION DATA FORDICLOFENAC, LACTOSE ANHYDRE ANDDICLOFENAC– LACTOSE ANHYDRE (1:1)

MIXTURE


10/12


(table 2). However, the number of lines present in the XRDpatterns of the individual components was found missingin the similar pattern recorded for the mixture. Thesignificant difference in the X-ray patterns of the drug-excipient mixtures compared to those of individual drugsand excipient indicates possible incompatibility of thedrugs with the excipient, even at room temperature. Thepresence of majority of the lines of the parent substancesin the thoroughly ground mixture prepared at room

temperature, however, suggests the interaction of the drug with the excipient at room temperature, which couldincrease with the increased temperature.

Al so , fo r th e binary mixt ure: DC–β -lactose, thediffractogram (fig.14) and the X-ray diffraction data (table3) show the interaction of these two substances. Thenumber of new lines appeared in DC–PVP, respectively DC–MS mixtures are shown in tables 4 and 5. The sametables indicate disappearance of some of the diffractionlines of higher, moderate and lower intensities in themixture which are originally present in the X-ray diffractionpatterns of the individual components which indicates the

interaction of DC with PVP and MS.

Table 4

X-RAY DIFFRACTION DATA FOR DICLOFENAC,MAGNESIUM STEARATE AND DICLOFENAC–MS (1:1)

MIXTURE


11/12


12/12

Application of Thermal Analysis to Study the Compatibility

Documents