Electrophoretic properties of lactose and salbutamol sulfate

Nottingham NG7 2RD, United Kingdom, and 3M Health Care, Ltd., Morley Street, ... Received July 22, 1992. In Final Form:November 4, 1992. We have studi...
1 downloads 0 Views 1MB Size
Langmuir 1993,9, 83S843

839

Electrophoretic Properties of Lactose and Salbutamol Sulfate Suspensions in Halogenated Solvents B. K. Sidhu,t C. Washington,*$tS. S. Davis,?and T. S. PurewaP Department of Pharmaceutical Sciences, University of Nottingham, Nottingham NG7 2RD, United Kingdom, and 3M Health Care, Ltd., Morley Street, Loughborough, LE11 lLP, United Kingdom Received July 22,1992. In Final Form: November 4,1992 We have studied the electrophoreticmobility of lactose and salbutamolsulfatesuspensions in chloroform and trichlorotrifluoroethane (Propellant 113) using quasielastic laser light scattering. The measured mobilities can be understood in terms of a polarity series of liquids and solids. The surfactants lecithin and Span 85 are positively charged in chloroformand adsorb to the negative lactose surface causing charge reversal; they do not appear to adsorb strongly to the positive surface of salbutamol sulfate and reduced its charge by nonspecific double-layer effects alone. The effects of traces of water were also investigated by 'drying solvents and solids rigorously, which reduced the measured mobilities, and it is possible that water mediates surfactant and surface ionization in these systems.

Introduction The stabilization and behavior of dispersions of solids or liquids in low-dielectric liquid media are of considerable interest. Systemsof this type are widely used in the paint and coatings industry, in lubrication technology, and in pharmaceuticals and agrochemicals. Unfortunately the current state of knowledge of such systemsis considerably poorer than that concerningmaterials dispersed in aqueous media, which are relatively well understood. An understanding of many aqueous dispersions is based on their electrophoretic behavior, which is of importance in determining, for example, stability, rheological properties, and coating behavior. Knowledge of the electrophoretic properties of dispersions in nonpolar liquids is much less well developed than that in aqueous dispersions for two main reasons. Firstly, there is little concensus on the mechanisms for charge generation and propagation in media of low dielectric constant. Features such as the ionization of surfaces, stabilization and energetics of charge carriers, and behavior of surface-active materials are only poorly understood. Secondly, the measurement of particle mobilities in media of low dielectric constant is experimentally more difficult than the corresponding measurement in aqueous suspensions. Morrison and Tarnawskyj' report that it is still difficult to obtain consistent microelectrophoresis results in low conductivity media and describe methods for checking the consistency of data. With the development of light s c a t t e r i n g a n d electroacoustictechniques, workershave measured zeta potentials in nonaqueous systems,24 but the techniques have numerous difficulties; trace quantities of contaminants (particularly water) can dramatically alter the measured zeta potentials,- mobilities are very small since the permittivity of organic liquids is low, and low permittivity University of Nottingham. 3M Health Care, Ltd. (1) Morrison, I. D.; Tarnawskyj, C. J. Langmuir 1991, 7,2358. (2) Labib, M. E.; Williams, R. J. Colloid Interface Sci. 1983,97,356. ( 3 ) Kosmulski, M.; Matijevic, M. Langmuir 1991, 7, 2066. (4) McGown, D.N.L.;Parfitt, G. D.;Willis, E. J . Colloid Interface Sci f

t

1966, 20, 650.

(5) Kandori, K.; Kon-no, K.; Kitahara, A. Bull. Chem. SOC.Jpn. 1984, 57, 3419.

causes additional problems of cell capacitance and discontinuous electric fields when a driving potential is applied.6 discussed the instrumental requireKornbrekke et menta for the measurement of electrophoretic mobilities in liquids of low dielectric constant. The electrophoresis cells normally used with light scattering instrumentssuch as the Malvern Zetasizer or the Coulter Delsaare optimized for use with aqueoussuspensions. They consist of aliquid channel which is several times longer than ita width in order to increase ita electrical resistance and minimize electrical heating during the experiment. When filled with a low-dielectric liquid, the electric field does not readily pass down the liquid channel and will take any alternative low impedanceroutes between the electrodes via adjacent conducting paths; in particular the water bath surrounding a thermostated cell may appear almost as a direct short circuit between the electrodes or to the earthed cell body. In order to obtain a uniform field between the electrodes, a parallel plate cell should be used, in which the electrode plates are large compared to their separation. An electrophoresis cell which conforms to these specifications is commercially available for the Malvern Zetasizer series of light scattering instruments, and we have used this cell in the present work. We are particularly interested in the mobility of particles of organic solids suspended in chlorofluorocarbon media; these are widely used for delivery of drugs in inhalation therapy, e.g. for asthmatics.' The drug, finely ground by fluid energy milling, is suspended in the chlorofluorocarbon propellant of an aerosol which is inhaled by the patient. The performance of this device is intimately related to the dispersion of the drug in the propellan%in general the drug must be well dispersed or at most weakly flocculated in order to be uniformly delivered in a predictable dose when the metered dose inhaler is actuated. Consequently a knowledge of the electrophoretic properties of the drug suspensions in the aerosol propellant is extremely valuable when designing such delivery systems. There has been additional interest recently due to the need to reformulate many aerosol products with non-chlorofluorocarbon propellants such as alkanes. (6) Kornbrekke, R. E.; Morrison, I. D.; Oja, T. Langmuir 1992,8,1211. (7) Ganderton,D.;Jones,T. M.DrugDeliuery to theRespiratory Tract; Ellis Horwood: Chicheater, 1987.

Q743-7463/93/24Q9-Q839$04.oO/o0 1993 American Chemical Society

Sidhu et al.

840 Langmuir, Vol. 9, No. 3, 1993

We have studied the electrophoretic properties of two model suspension systems, those of lactose, a typical hydrophilic model material, and salbutamol sulfate, a potent bronchodilator. These were suspended in either chloroform (as a model system for chlorinated solvents) or trichlorotrifluoroethane (Propellant113). These liquids were chosen since they were among the lowest-boiling propellant-like materials that could be used in the electrophoresisinstrument without excessiveevaporation. It is unfortunate that the widely used propellants 11and 12 (trifluorochloromethaneand dichlorodifluoromethane) are too volatile for study in the present apparatus, and studies involving these materials will have to await the development of an electrophoresiscell that can be pressurized to severalatmospheres. We have also investigated the effect on particle mobilities of varying amounts of two surfactants, Span 85 (an impure preparation of sorbitan trioleate) and egg lecithin, a mixture of phosphatides containing a r h g e of unsaturated fatty acid ester chains of 12-18 carbon atoms. Both of these materials are widely used to control flocculation in aerosol suspensions.

,---

Compression Screw

_ , - - * ' S ~ p p ~Block rt

Photomultiplier

---_---. (Transparent Prism Tin Oxide Electrode Coating) Cell Body

'-------

Figure 1. Malvern 5126 low-conductivity electrophoresis call (plan view, not to scale). The sample is contained in the central cavity which is approximately 1mm wide. Table I. Zeta Potentials of Lactose and Salbutamol Sulfate Suspensions

Experimental Section Lactose, salbutamol sulfate, and trichlorotrifluoroethanewere kindly donated by 3M Health Care, Ltd., Loughborough, U.K. Lactose and salbutamol sulfate were ground by fluid energy milling to mass mean diameters of 2.6 and 3.3 pm, respectively (measuredby laser diffraction in heptane suspension by Malvern 2600 sizer, Malvern Instruments, U.K.). Chloroformwas HPLC grade from Prolabo, Paris, France; Span 85 was obtained from Sigma (Poole, Dorset, U.K.) and egg lecithin was Ovothin 180 (Lucas-Meyer, Hamburg, Germany), consisting of 90 96 unsaturated fatty phosphatides, approximately 80% of which were phosphatidylcholine. This material is purified from crude egg lecithin by chromatographic means and is typical of the many well-characterized egg phosphatidylcholine products available. Measurementswere performed using both chloroformassupplied and chloroform dried over a grade 3A molecular sieve (BDH Chemicals, Poole, England) for 8 days; in this case lactose and salbutamol sulfate were dried under vacuum for 18 h (Edwards Modulyo Freeze Drier). Dispersions of solid powders in propellants were prepared by adding a small quantity of the solid to the liquid and then sonicatingin a Decon FS 100ultrasonic bath (DeconUltrasonics, Ltd., Hove, England) for 5 min. The effect of surfactants was investigated by preparing a solution of the surfactant in the liquid phase and then adding a small quantity of solid powder and sonicating as previously described. All concentrations are expressed as percent weight to volume instead of molarity, since the molecular weights of the surfactants used (which are heterogenous materials) are not precisely known. Electrophoretic measurements were made using the wellestablished quasielastic Doppler shift light scattering method, with the Malvern Zetasizer 4 (Malvern Instruments, Malvern, Worm, England) using the ZET 5126 nonaqueous cell (Figure l),which has a path length of 1 mm, an electrode area of approximately 8 mm X 8 mm, and electrodes of conductive tin oxide coated onto glass prisms through which the laser beams illuminate the sample. Initial studies demonstrated that the measured zeta potentials were independent of the applied electric field in the range 40-120 V/cm, and so a value of 100 V/cm was maintained for the present experiments, this is rather lower than the 400 V/cm used by Kornbrekke et al.6 The totalmeasurement time was 10 s, during which time the field was alternated at 1 Hz to avoid polarization and particle charging effects? This instrument has the facility to delay data collection for a fixed period after the cell voltage is changed, in order to allow field stabilization. Six measurements were taken for each sample and at least two samplesof suspension were prepared for each system studied. The sample cell was washed thoroughly (5-10 separate rinses) with the suspending liquid before injection of the next

solid lactose

liauid zeta mtential/mV chloroform -23 f 3 trichlorotrifluoroethane -33 f 2 dry chloroform -5 f 2 salbutamolsulfate chloroform 34 f 5 trichlorotrifluoroethane -18 3 dry chloroform 55 f 2

*

sample. Zeta potentials were calculated from electrophoretic mobilities using the Hiickel equation.8 Experiments using highly dried materials were performed as quickly as possible and measurements were made within about 10 min after removing the powder samples from the drying chamber. Moisture was reduced to below 1.6 ppm as measured by the Karl Fischer method. "Water-free" studies with surfactants were not conducted due to the heterogeneous nature of these materials, which themselves contain 0.1-1 76 water.

Results Table I shows the zeta potentials of lactose and salbutamol sulfate in chloroform and trichlorotrifluoroethane and of the dried solidsin dried chloroform. Lactose had a negative surface charge in all three liquids, -5 mV in dry chloroform, increasing to -23 mV in undried chloroform (water content approximately 100 ppm), and -33 mV in trichlorotrifluoroethane. Salbutamol sulfate had a positive charge in dry chloroform (+55 mV) and in undried chloroform (+34 mV), but a negative charge in trichlorotrifluoroethane (-18 mV). In all cases the addition of surface-active materials caused monotonic changes in zeta potential. Figure 2 showsthe effectof adding lecithin to the lactose/chloroform system;initiallythe solid was negatively charged,but small quantities of lecithin rapidly caused this charge to be neutralized and a positive charge to develop. A maximum zeta potential of +64 mV was attained at a lecithin concentration of 0.01 % , and further increase in the surfactant concentration to 1%(w/v) did not lead to a correspondingincrease in the zeta potential. The effects of lecithin on the salbutamol sulfate/chloroform system (Figure3) were rather different; concentrationsof lecithin in the range studied for the lactose/chloroformsystem (8)Hunter, R.J. Zeta Potential in Colloid Science; Academic Press: London 1981.

Electrophoretic Mobility

Langmuir, Vol. 9, No. 3, 1993 841

I

60

>

10''

10.'

10.'

10"

1w5

10.~

Figure 2. Effect of lecithin on the zeta potential of micronized lactose suspended in chloroform.

%

.ni

1

.1

1

Span 85

Figure 5. Effect of Span 85 on the zeta potential of micronized salbutamol sulfate suspended in chloroform.

-18 5

.1

.01

.001

% lecithin

r

Lactose

r

Salbutamol sulphate

1

I

t Trichlorotrifluoro 1 methane

% lecithin

Figure 3. Effect of lecithin on the zeta potential of micronized salbutamol sulfate suspended in chloroform.

salbutamol sulfate in chloroform and trichlorotrifluoroethane.

40 1

""

.ooni

All zeta potentials in mV

Figure 6. Relative zeta potential diagram for lecithin and

.om

.01

.1

1 %

Span 85

Figure 4. Effect of Span 85 on the zeta potential of micronized lactose suspended in chloroform.

(lO-9-lO-2%) did not significantly alter the zeta potential, and it was only after the addition of relatively large amounts of lecithin (0.01-13' % ) that a change in the zeta potential was observed. The zeta potential was reduced from +34 to +12 mV in the presence of 1%lecithin. Broadly similar trends were observed in the lactose/ chloroform (Figure 4) and salbutamol/chloroform (Figure 5 ) systems in the presence of Span 85. Span 85 concentrations from lWto 1% caused the zeta potential of lactose in chloroform to become less negative and ultimately reached a maximum of +30 mV with 0.1 % Span 85, with no significant increase at higher surfactant concentrations. The zeta potential of salbutamol sulfate in chloroform was reduced from +35 to +8 mV at 0.1% Span 85 concentration, with no measurable change at higher surfactant concentrations.

Discussion Solid/Liquid Systems without Surfactant. Examination of Table I reveals that it is possible to arrange the solids and liquids into a polarity series, in which each component, solid or liquid, is negatively charged with respect to the materials below it (Figure 6). This suggests that the particle charge is determined by the relative polar nature of the solvent and solid. Although this method successfully predicts the sign of the charge observed in a particular system, it does not unfortunately predict the magnitude; the diagram is not an additive one, as can be seen from the measured zeta potentials indicated in Figure 6. For this purpose we have used the potentials as measured in the dried materials; in the undried solvent the potential is moved down the scale away from lactose, without changing the order of the items; thus the lactosechloroform potential becomes more negative and the salbutamol sulfate-chloroform potential becomes less positive. This is a highly empirical description but it does appear to have some predictive value, albeit in a rather small set of materials. Many theories have been proposed regarding the origin of charge in nonaqueous media. Hypotheses that charge transfer occurs via electrons, hydroxyl ions or protons as charge carriers have been proposed since the 1940s. Verwey and Overbeekg believed that protons were the charge carriers in systems comprising inorganic oxides suspended in organic media. Lyklema'O considered the case of aprotic organic solvents and suggestedthat protons (9) Verwey, E.J. W.; Overbeek, J. Th. G. Theory of the stability of lyophobic colloids; Elsevier: New York, 1948. (10) Lyklema, J. Adu. Colloid Interface Sci. 1968,2, 67.

842 Langmuir, Vol. 9, No. 3, 1993

Sidhu et al.

and hydroxyl ions were the charge carriers in suspensions containingthese solvents. These ions would arise through ionization mediated by water impurity in the dispersion. In the present experiments, lactose in dry solvents carries a small negative charge, possibly due to loss of a proton. This weak ionization may become more pronounced in the presence of trace amounts of water, as the potential is observed to become more negative. Similarly, salbutamol sulfate could become positively charged by dissociation of sulfate groups from the crystal lattice, in a similar manner to the extensively studied charge generation processes in ionic materials (typified by silver iodide1'). Labib12considered nonaqueous solvents and proposed mechanisms of charge transfer based on the relative electron donicity of the disperse and continuous phases. He suggested that electrons were the charge carriers in these systems and that the solvents acted as Lewis bases. The donicity of many different solids in the presence of a variety of different aprotic liquids was measured by electrophoresis. Each liquid was assigned a donor and acceptor number depending upon the relative electron donating nature of the liquid, and the acceptor number was a measure of the tendency of the liquid to accept an electron. This approach is qualitatively similar to the polarity series discussed above. Solid/Liquid Systems with Surfactant. Both surfactants had a marked effect on the mobility of the solids. The main problem in the interpretation of the data is that the surfactants are relatively heterogeneous;for example, lecithin consists of a number of phosphatides, of which the major component (80%)is phosphatidylcholine (PC), with smalleramounts of phosphatidylethanolamine (PE), phosphatidylserine (PS),and phosphatidylglycerol(PG). The behavior of these materials in aqueous systems is well understood;13they are zwitterionic surfactants, PC and PE being uncharged at pH 7, while PG and PS are negatively charged. Unfortunately it is difficult to predict the degree of ionization (if any) that these materials will display in nonaqueous solvents; however, since very small amounts of lecithincause a positivechargeto be developed on lactose in chloroform, it is reasonable to suggest that this surfactant carries a positive charge and adsorbs to the negative lactose surfaceby electrostatic attraction, or even a specific adsorption process, ultimately causing charge reversal. Electrostatic processes are likely to be among the most important in nonaqueous systems, since the low dielectric constant of the continuous phase causes free ionsto have relativelyhigh free energies, and the electrical double layers are extended considerablyover comparable aqueous systems. The major component of lecithin (PC) becomes positively charged in aqueous media below pH 4.5,14and it is reasonable to suggest that in nonaqueous solvents,any free protons (e.g. arising from traces of water impurity) would be scavenged by the surfactant. The salbutamol sulfate surface is positively charged in the absence of surfactant, and so lecithin does not adsorb to it; the reduction of mobility in the presence of relatively high surfactant concentrations is almost certainly due to double-layer shielding effects by the ions generated by the ionization of lecithin. It should be noted that the concentration of lecithin required to significantly change

the mobility in the salbutamol-chloroform system is several orders of magnitude higher than in the lactosechloroform system, which is consistent with this interpretation. The behavior of the Span 85 surfactant is more difficult to interpret. Although it appears to be more strongly surface-active in the lactose-chloroform system, there is not as pronounced a difference between its effects in this system and the salbutamol-chloroform system, and so it could be assigned either a positive or a negative charge. However, the fact that it causes charge reversal in the lactose-chloroform system suggests that it is specifically adsorbing to the negatively charged lactose surface and thus may be positively charged. It is then possible that the reduction in mobility of salbutamol is due to nonspecific screeningeffects, in the same manner as observed for lecithin. In aqueous systems the span series are not normally thought of as ionic surfactants, and they could change the mobility in the present systems by a number of mechanisms. Since they are relatively heterogenous materials, it is possible that small amounts of impurities could contribute to the surface charges, similarly to the known behavior of lecithin in oil-in-wateremulsions,14in which the entire surface charge is contributed by 2-3% of acidic phosphatides in the surfactant. It is possible that if adsorption of charged species was the charging mechanism in the presence of both types of surfactant, then the lower charge in Span 85 suspensions was caused because there was asmaller quantity of charged impurities present in Span 85 than in the lecithin preparations. Koelmansand Overbeek15reported that the positivecharge developed by Span 40 and 80 on alumina and barium sulfate may have been due to impurities in these surfactants, since methanol recrystallization caused a decrease in the conductivity of the surfactant solutions. Alternatively Span 85 could exist as a positively charged species in apolar media since it may be protonated by traces of water present. Mobility data can normally be correlated with suspension stability using DLVO approaches. In the present case this is difficult to perform quantitatively because the extended double layers present in nonaqueous media are highly sensitive to the composition of the system. However, Morrison16has suggested that mobility is a reliable indicator of stability in nonaqueous media; thus it is reasonable to expect the data presented here to correlate with the stability of the suspensions. We have previously reported studies of suspension stability of lactose in chl~roform'~ using sedimentation and rheological techniques, and reasonable agreementis found with the present mobility data. Lecithin was found to stabilize lactose suspensions in chloroform at very low (