Characterization of Selected Materials for Lemon Oil Encapsulation by Spray Drying W.E. BANGS and G.A. REINECCIUS ABSTRACT Viscosity Methods were developed for characterizing various polymers for their capacity to retain high amounts of encapsulated lemon oil. The testing protocol included evaluation of polymer isothermal drying curves, fluid flow characteristics, lemon oil encapsulation efficiency, emul- sion stability, maximum loading values, vapor phase diffusive flux and extractable oil. Of the encapsulationmatrices examined, the com- merical product, Capsul, was best overall, with excellent retention of volatile oil, emulsion stability and load capacity. An oxidized, phen- ylalanine-reacted starch rated best of the experimental materials. Fluid flow properties of 25% and 40% total solids aqueous disper- sions of each material base (gum arabic, Capsul, M-100, oxidized and control starches) were determined using a Haake Rotovisko Model RV2 visometer (Haake Inc., Saddle Brook, NJ). All samples were maintained at 26°C during analysis. Lemon oil encapsulation Sample preparation, drying and storage proccdurcs were followed as previously outlined (Bangs and Reineccius, 1988). Twenty percent (dty basis) cold-pressed lemon oil (compliments of Universal Flavors, Indianapolis, IN) was added to aqueous dispersions of each carrier and spray-dried in a mini-dryer [25% w/w (Bangs and Rcincccius, 1990)] or a Niro Utility Model dryer [40% w/w Niro Atomizer, Ltd., Columbia, MD)], depending on sample size constraints. The dryers were operated at 220” t 5°C inlet and 100” t 5°C exit air tcmpcra- tures. INTRODUCTION GIVEN THE IMPORTANCE of spray-dried flavorings in the food supply, published information on testing protocols for wall material selection is scarce. Other than dryer retentions (King et al., 1976; Reineccius and Coulter, 1969; Kerkhof and Thijssen, 1977; Voilley and Simatos, 1980; Leahy et al., 1983; Bangs, 1980; Bangs and Reineccius, 1981; Subramaniam, 1984), sensory evaluation (Subramaniam, 1984), Warburg O2 uptake (Subramaniam, 1984) and extractable flavoring analysis after drying (King et al., 1976; Subramaniam, 1984), no fur- ther studies have been made on wall material characterization for spray-dried encapsulationprocesses. The capacity of a pol- ymer to form stable aqueous emulsions of added flavoring materials, to create an impermeable inert barrier for flavor protection in the dried product and to result in controlled re- lease should be considerations as important as dryer retention. The purpose of our work, therefore, was to evaluate several objective criteria for selecting wall materials for encapsulation of lemon oil via spray drying. MATERIALS & METHODS Wall material Carrier solids selected for characterizing lemon oil encapsulation functionality included gum arabic (compliments of Fritzsche Dodge and Olcott, Inc., New York, NY), a maltodextrin with a dextrose equivalency of 10 (M-100; Grain Processing Corp., Muscatine, IA), a lipophilic starch (Capsul, National Starch and Chemical Co., Brid- gewater, NJ) and four corn starches, previously developed [oxidized, control, oxidized-phenylalanine and oxidized-aspartame glycoamines (Bangs and Reinccius, 1988)]. Drying curves Isothermal (60°C) drying curves for 40% (wb) total solids aqueous dispersions of Capsul, gum arabic, M-100 and oxidized starch were developed. Fifteen gram aliquots of ambient temperature (ca. 26°C) aqueous dispersions of each material were pipetted into glass petri dishes (4.90 cm diameter) in duplicate and dried in a convection oven (model 5750 gas chromatographic oven, Hewlett-Packard, Avondale, PA). Samples were periodically removed from the oven and weighed on a Sartorius Model 1213 MP (Gottingen, Germany) balance for moisture loss determinations. Author Reineccius is with the Dept. of Food Science & Nutrition, Univ. of Minnesota, 1334 Eckles Ave., St. Paul, MN 55108. Au- thor Bangs is with Campbell Soup Company, P.O. Box 578, Campbell Place, Camden, NJ 08101. Emulsion stability The effectiveness of each carrier (gum arabic, Capsul, M-100, ox- idized and control starches)in maintaining stable emulsions with 20% (db) lemon oil incorporation was monitored by an acquiescent elcc- trical resistivity measurement (Atkins, 1982). Following 30 min of mixing on a Vortex of each emulsion (1OOg of a 25% or 40% w/w solids sample in a 2.50 mL glass beaker), mixing was stopped and electrical resistivity (ohms) was measuredat 1, 5, 10, 1.5, 30, 45, and 60 min using a standard electrical multimeter (Unisound DT-1004, Radio Shack). Although resistivity measurements followed apparent first-order ki- netics for the first 15 min of analysis, the data became linear after that. The pseudozero-order linear regressionslope (resistivity vs time) of the 15 through 60 min data was, therefore, reported. In order to negate variations in initial electrical resistivity values from sample to sample, the data were always normalized prior to regressionanalysis. Lemon oil headspace content/load The maximum amount of lemon oil that could bc placed into 40% (w/w) aqueous dispersions of each carrier base (gum arabic, Capsul, M-100, oxidized and control starches) was determined. Lemon oil was added at 5%, lo%, 20% and 40% (also a 60% db loading in the Capsul emulsion) to each carrier base. These samples were used to obtain equilibrium capillary-GC headspace data as previously outlined (Bangs and Reineccius, 1990). The unit, % EXTD (cxtcrnal standard), expressesthe amount of lemon oil found in the hcadspaceabove a sample relative to the amount of lemon oil dctermincd above a pure water solution (time zero, constant temperature and equivalent lemon oil content). Maximum lemon oil taken up by each of these systems was dctcr- mined by slowly adding lemon oil into the vortex of 40% (w/w) solids aqueousdispersionsof Capsul, gum arabic, M-100, oxidized and con- trol starches, oxidized-phenylalanine and oxidized-aspartame gly- coamines and dry-blended phenylalanine/oxidized starch and phenylalanine/control starch systems (10% db substitution with phen- ylalanine). Mixing of the emulsions was completed on a magnetic stir plate (Model 4812; Cole-Parmer, Chicago, IL). Upon the appearance of a second liquid phase (i.e., lemon oil on the surface), lemon oil addition and stirring ceased and the sample beakers were covered with aluminum foil. After 24 hr at room temperature (ca. 26”C), excess lemon oil was decanted from the surface and the samples weighed. Weight increaseswere determined and recorded as maximum lemon oil incorporated. AI1 samples were tested in duplicate. 1356-JOURNAL OF FOOD SCIENCE-Volume 55, No. 5, 1990