3254 J. Agric. Food Chem.

2008, 56, 3254–3259

Influence of Flavor Solvent on Flavor Release and

Perception in Sugar-Free Chewing Gum
RAJESH V. POTINENI AND DEVIN G. PETERSON*
Department of Food Science, 327 Food Science Building, The Pennsylvania State University,
University Park, Pennsylvania 16802-2504

The influence of flavor solvent [triacetin (TA), propylene glycol (PG), medium chained triglycerides
(MCT), or no flavor solvent (NFS)] on the flavor release profile, the textural properties, and the sensory
perception of a sugar-free chewing gum was investigated. Time course analysis of the exhaled breath
and saliva during chewing gum mastication indicated that flavor solvent addition or type did not
influence the aroma release profile; however, the sorbitol release rate was statistically lower for the
TA formulated sample in comparison to those with PG, MCT, or NFS. Sensory time-intensity analysis
also indicated that the TA formulated sample was statistically lower in perceived sweetness intensity,
in comparison with the other chewing gum samples, and also had lower cinnamon-like aroma intensity,
presumably due to an interaction between sweetness intensity on aroma perception. Measurement
of the chewing gum macroscopic texture by compression analysis during consumption was not
correlated to the unique flavor release properties of the TA-chewing gum. However, a relationship
between gum base plasticity and retention of sugar alcohol during mastication was proposed to explain
the different flavor properties of the TA sample.

KEYWORDS: Flavor release; perception; flavor solvents; chewing gum; time-intensity; breath analysis

INTRODUCTION softer gum bases were reported to release flavor compounds at

a faster rate than the harder gum bases; however, he was unable
The flavor properties of chewing gum are undoubtedly a key to conclude if this was due to a higher diffusivity or if the
product attribute for consumption. Defining the mechanisms of panelists chewed the softer gum at a faster rate which facilitated
flavor release and perception in chewing gum (or foodstuffs) extraction. The gum base industry has also suggested chewing
can be better understood by combining analytical methods that gum formulated with the flavor solvent triacetin (TA) or medium
monitor the release profiles of key flavor stimuli near the chained triglycerides (MCT) are softer than if formulated with
receptors in combination with sensory evaluation. Previous propylene glycol (PG), and therefore the former two products
studies on the flavor perception of chewing gum, for example, would have a higher perceived flavor intensity (4).
have correlated perceived mint flavor intensity to the sucrose A few studies have also reported a correlation between aroma/
concentration in the saliva. Davidson et al. (1) monitored the flavor perception and textural properties of solid/semisolid food
temporal release profile of menthone and sucrose concentration systems (for example, gels, yogurt, model dessert). Baek et al.
from chewing gum while panelists recorded mint flavor intensity (5) conducted a time-intensity sensory study on gelatin gels and
over time. Panelists perceived a decrease in mint flavor intensity reported a higher maximum flavor intensity and a lower time
over time which was correlated to the decrease in sucrose to maximum flavor intensity for softer gels compared to harder
concentration rather than menthone release suggesting taste- gels which they suggested was due to a faster release of volatiles
aroma interactions were important for the overall mint flavor from a softer gel. Using a model mouth and in vivo analysis
perception. Similarly, Duizer et al. (2) reported that a longer via Proton Transfer Reaction Mass Spectroscopy (PTR-MS),
duration of peppermint flavor was perceived with a faster release Hansson et al. (6) showed the release of volatiles from a pectin
rate of sucrose in chewing gum. gel was not only dependent on the physiological properties of
The textural properties of chewing gum have also been the mouth (mastication rate and saliva), but also on the textural
suggested to influence volatile flavor release properties. de Roos properties (gel strength/structure) of the gels. In contrast, a study
(3) varied the textural properties of flavored gum bases by conducted by Weel et al. (7) using whey protein gels reported
varying the gum base composition or by adding a plasticizer that the concentration of volatiles in the breath was not
(glycerine monostearate) and measured the residual volatile influenced by gel strength; however, they did indicate that the
flavor concentration over time during consumption. Overall, the stronger gels were perceived at a lower flavor intensity. Mestres
et al. (8, 9) attempted to explain these contradictory results based
* Author to whom correspondence should be addressed (fax: 814- on the hypothesis of “first impression”, where the perceived
863-6132; e-mail: [email protected]). aroma intensity is dictated by the initial release rates rather than
10.1021/jf072783e CCC: $40.75  2008 American Chemical Society
Published on Web 04/22/2008
Influence of Solvent-Type on Chewing Gum Flavor J. Agric. Food Chem., Vol. 56, No. 9, 2008 3255

the overall amount of aroma stimulus or cognitive textural Table 1. Composition of Cinnamon-Like Aroma Mixture and Log P Values
interactions. They reported that the temporal resolution of
compound name flavor composition (g/100 g) log P value
retronasal aroma perception was influence by the opening and
closing of the velum-tongue border (passage way to the olfactory cinnamaldehyde 88.26 1.90a
L-carvone 10.16 2.87b
membrane in the oral cavity), which was controlled by the
piperitone 0.79 2.85b
texture of the gels. For example, Mestres et al. indicated that jasmone 0.79 3.55c
the velum-tongue border was found to be open during the initial
chewing phase prior to swallowing for hard gels while for soft a
Experimental value from Hansch and others (1995). b Experimental value from
gels it was only intermittently open or was closed which was Griffin and others (1999). c Estimated based on the Kow calculation program
further related to the panelist chewing pattern (opening and (Syracuse Research Corporation; http://www.syrres.com/aboutsrc/default.htm).
closing of the jaw versus a side-to-side chewing motion). If the
velum-tongue border was closed there would be no transfer of (mixed 2 min) followed by addition of about 50% of the sorbitol and
volatiles from the oral to the nasal cavity until the sample was mixed for another 2 min. At approximately 75 °C, the cinnamon-like
swallowed. These noted differences in retronasal aroma release aroma mixture (with or without flavor solvent) and the remaining
based on the velum-tongue border indicate the challenges in sorbitol was added and mixed for 2 min. Finally, the glycerin then the
investigating the influence of texture on flavor perception. remaining sorbitol syrup were added and further mixed for 1 min per
Lethuaut et al. (10, 11) investigated the influence of three each ingredient. The resultant chewing gum dough was rolled using a
Rondo dough roller (Moonachie, NJ, average thickness was 0.168 cm
flavor stimuli in combination on the overall flavor perception
( 0.005) and subsequently conditioned at room temperature at 45%
of a model dessert by varying the texture agent, sucrose, and humidity for 12 h before being cut into commercial size sticks (Package
aroma concentrations. Although the sweetness intensity was machinery cutter, West Springfield, MA). The chewing gum samples
reported not to be affected by aroma concentration, both the were wrapped in aluminum foil and stored at 21 °C at 35% ((10)
textural properties and sucrose release rates were found to impact relative humidity prior to analysis (<4 months).
the perceived aroma intensity at higher aroma concentrations. Quantification of Aroma Compounds in Chewing Gum. Twelve
However, at lower aroma concentrations, no textural effects on gum pieces per chewing gum treatment (sampled every 10th piece out
flavor perception were reported (11). of 120 pieces) were further subsampled to a 0.5 ( 0.02 g sample and
dissolved in 1 mL of hexane on a vortex shaker (Vortex Genie 2, Model
The aim of this study was to further investigate the role of CG-560, NY). The hexane mixtures were then centrifuged at 11 750
flavor solvent on the release profile of volatile (aroma) and rcf for 4 min (Brinkman Instruments Inc., Model no: 5415C, NY) and
nonvolatile (sugar-alcohol) flavor compounds as well as on the 0.3 mL of the supernatant was collected and added to 1 mL of methanol.
textural properties and how these three stimuli influence flavor The hexane-methanol extracts were then centrifuged at 11 750 rcf for
perception in chewing gum. The influence of the solvent 4 min and 1 mL of supernatant was collected. This step aided in the
properties of PG, TA, or MCT on the chewing gum matrix in precipitation of the gum base polymers. Methanol (100 µL) containing
relation to flavor perception is not well defined and therefore methyl hexanoate (as internal standard; 2500 mg/L) was then added
was investigated. to the hexane-methanol supernatant, and subsequently analyzed by
gas chromatography equipped with a flame ionization detector
(GC-FID).
MATERIALS AND METHODS
Gas Chromatography (GC). Analyses of aroma compounds was
Materials. Cinnamaldehyde, L-carvone, and jasmone were purchased performed using a Hewlett-Packard 5890 Series II GC equipped with
from Aldrich (Sigma Aldrich, Milwaukee, WI). Piperitone was from a split/splitless injector, flame ionization detector (FID), autosampler
the Penta Manufacturing (Livingston, NJ). Methanol was from Fisher (HP 7673), and a fused-silica capillary column (DB-wax, 30 m, 0.32
Scientific (Fairlawn, NJ). Hexane and formic acid were from EMD mm inner dia, 0.32 µm film thickness, Agilent Technologies, CA). The
Chemicals (Gibbstown, NJ). N-Caproic acid methyl ester was purchased GC operating conditions were as follows: 1 µL of sample was injected
from TCI America (Portland, OR). VH1 gum base was obtained from in split mode (1:20); inlet temperature was 200 °C, oven program was
Hersheys Foods (Hersheys, PA). Sorbitol was from SPI polyols 35 °C for 2 min, then increased at 10 °C/min to 230 °C and held for
(Wilmington, DE). Glycerine was from Univar (Bedford Park, IL). 3 min; constant flow rate of 1.0 mL/min (He).
Hydrogenated glucose syrup was from Roquette Americas (Lycasin Measurement of Sorbitol/Hydrogenated Glucose Syrup Release
80/55; Keokuk, IA). Lecithin was from Solae (St. Louis, MO). Titanium from Chewing Gum during Mastication. The concentration of sorbitol
dioxide was from Sensient (St. Louis, MO). Propylene glycol and and hydrogenated glucose syrup (HGS) were determined by HPLC from
triacetin were from Givaudan Flavors (Cincinnati, OH), whereas the expectorated saliva for three panelists while chewing a 2.5 g piece
medium chain triglycerides (Neobee-80) was from Stepan Company of chewing gum sample over a 12 min time period. The panelists were
(Northfield, IL). trained to follow a defined chewing and expectorate saliva (C/E)
Chewing Gum Models. The chewing gum ingredient formulation protocol: chew at 60 chews/min (used a metronome) and expectorate
consisted of gum base (25.57 g/100 g), sorbitol crystals (54.48 g/100 saliva into 20 mL cups with lids at 0, 10, 30, 50, 70, 120, 240, 420,
g), hydrogenated glucose syrup (11.80 g/100 g), glycerine (2.95 g/100 660 s. One-half a gram of saliva was immediately transferred into a
g), saturated sorbitol solution (1.97 g/100 g), flavor mixture (0.98 1.5 mL centrifuge tube and mixed with 1 mL solution of nanopure
g/100 g), flavor solvent (0.66 g/100 g; PG, TA, or MCT), titanium water containing 0.1 g/100 g formic acid. The samples were then
dioxide (0.49 g/100 g), and Lecithin (0.10 g/100 g); the gum base centrifuged at 12 000 rpm for 3 min and the supernatant was transferred
composition consisted of polyisobutylene (7-10 g/100 g), styrene into 2 mL amber bottles with lids prior to HPLC analyses. Each chewing
butadiene rubber (3-5 g/100 g), polyvinyl acetate (15-20 g/100 g), gum sample was analyzed in triplicate. The sugar alcohol concentration
wood rosin (11-15 g/100 g), polyethylene (0-2 g/100 g), filler (25-35 was determined using an external standard curve at 6, 13, 25, 38, 50
g/100 g; CaCO3), BHT (0-0.1 g/100 g), waxes (4-8 g/100 g), g/L for sorbitol and 0.16, 0.31, 0.63, 1.25, 2.5, 5, 10 mg/L for HGS
emulsifier (0-2 g/100 g), and softeners (10-17 g/100 g; hydrogenated plotted versus peak area (r2 > 0.99).
confectionery fat) (12). The cinnamon-like aroma mixture compositions High Performance Liquid Chromatography (HPLC). Analyses
is reported in Table 1. Chewing gum samples were made by initially of sorbitol and HGS was performed using a Shimadzu HPLC system
melting the gum base (raised to 98-104 °C) in a Littleford Day gum consisting of pump (LC-10ATvp), degasser (DGU-14A), an auto sampler
mixer (Florence, KY). Lecithin and titanium dioxide were then added (SIL-10Ai), column heater (CTO-10ACvp), and refractive index detector
to the molten gum base during mixing, and after 2 min, the heat was (RID; RID-10A). Separations were performed on a LC column Supelco-
shut off and the mixer was cooled by circulation of room temperature gel-H (5 µm, 250 × 4.6 mm i.d.,) using an isocratic run with 0.1 g/100 g
water. During cooling, the hydrogenated glucose syrup was added formic acid in water as the mobile phase maintained at 40 °C. The flow
3256 J. Agric. Food Chem., Vol. 56, No. 9, 2008 Potineni and Peterson

rate was 0.17 mL/min, and the injection volume was 10 µL. Experimental chord value after 24 h (mm)
Plasticity Index (PI) )
Measurement of Volatile Release from Chewing Gum during Theoretical chord value (mm)
Mastication. Breath-by-breath analysis was performed with an atmo- (1)
spheric pressure chemical ionization-mass spectrometer (APcI-MS) as
Sensory Analyses. Time-intensity analysis was conducted with a
previously described by Schober and Peterson (13). Three panelists (1
trained panel of 9 people (3 males and 6 females; age range: 21-35
male and 2 female) were trained to chew the samples using a defined years) recruited from the Department of Food Science, The Pennsyl-
chewing and swallow (C/S) protocol; chew at 60 chews/min (using a vania State University. Three attributes of the chewing gum were rated:
metronome) and to swallow at 0, 10, 30, 50, 70, 120, 240, 420, 660 s. sweetness, cinnamon-like aroma, and effort to chew. The panelists were
The whole experiment was conducted over a 3 day period, with 4 gums trained on the attributes and time intensity procedures during 12 1-h
per panelist per day, in random order. To minimize carry over effects sessions. References for each attribute were developed by panel
and fatigue, each panelist rinsed with water and waited at least 30 min consensus. Panelists used references for sweet (sucrose solutions 2, 4,
between sample analyses. 6, 9, 12 g/100 g w/w corresponding to 2, 4, 6, 9 and 12 reference value),
The breath from the nose was directly and continuously sampled aroma (solutions of cinnamon mixture with 4 compounds as in Table
via an interface set at 80 °C into the retrofitted Micromass ZMD 1; 25, 50, 100 µg/L of water corresponding to 3, 6, and 9 reference
4000-mass spectrometer (Waters, Milford MA) at 0-4 min, 6-8 min, value), and texture (Jet-Puffed marshmallows, Gaint orange slices,
10-12 min time intervals. The APcI operating conditions are as follows: Swedish fish candy, Tootsie Roll chocolate corresponding to 1, 4, 7
SIM mode; sampling rate was 200 mL/min; block temperature is 100 and 9 reference values, respectively). For each attribute, a 15 point
°C; transfer line 90 °C; corona discharge was 3.5 kV, cone voltage scale was used for intensity measurement. Data was collected in
was 15 V. Ions monitored were 133 [M + H]+ for cinnamaldehdye, individual testing booths using Compusense software (Compusense Inc.,
Guelph, Canada). Panelists rated one attribute at a time while chewing
151 [M + H]+ for carvone, 153 for piperitone, and 165 [M + H]+ for
the gum (2.5 g) for 4 min. For the aroma and effort to chew attributes,
jasmone. Day to day variation in instrumental response was corrected
panelists were instructed to breathe normally through their nose with
for based on the peak heights obtained sampling a known amount of
their mouths closed. For the sweetness attribute, panelists evaluated
L-carvone (1 µL of 1000 mg/L solution in pentane) which was injected the samples wearing a nose-clip. For the final evaluations, 2.5 g of
into an airtight water-jacketed 1.1 L deactivated glass vessel maintained chewing gum was cut, wrapped in a wax paper and placed into small
at 40 °C and held for 5 min with constant stirring (200 rpm). Ziploc bags coded by a three-digit number 1 h before each session.
Quantification of aroma compounds from the breath was determined Four treatments were evaluated in triplicate over 12 sessions, using a
via standard calibration curve methodology. Different quantities of each complete balance block design. Two sessions were conducted per day
compound dissolved in 50 µL pentane (0.009, 0.018, 0.089, 0.179, with at least 3 h between same-day sessions. During each session,
0.536, 0.837, 1.768 µg) were injected into the deactivated glass vessel panelists evaluated all 3 attributes for a given treatment with 2 min
as described above. The peak height (ion intensity) versus µg weight interval between each attributes.
of each compound per L of air was plotted (all compounds reported an Statistical Analysis. ANOVA and Tukeys Pairwise Comparisons
r2 > 0.99). using SAS statistical software (V. 8.0, SAS Institute, Cary, NC) were
Instrumental Texture and Gum Volume Analysis of Chewing used to compare all treatments.
Gum during Mastication. Three panelists (1 male and 2 female) were Instrumental Data. For breath analysis, evaluations were conducted
asked to chew the chewing gum samples using the c/s protocol described on a 6 s moving average concentration values at 30, 70, and 150 s,
above. In two different experiments at the time intervals of 30, 60, while for sorbitol release, texture analysis, gum volume analysis, and
PI values all analyses were conducted at times 30, 70, and 120 s.
120, 240, 420, 720 s, the panelists expectorated the chewing gum
Sensory Data. Time Intensity curves for each panelist were generated
samples for texture or volume measurement. A new piece of chewing
using Compusense software. For the attributes sweetness and cinnamon-
gum was used for each time point (clock restarted at 0 s). After each
like aroma, the parameters Maximum intensity (Imax) and the Time-at-
sample, panelists rinsed their mouth with water to clear the palate, and
Maximum intensity (Tmax) were analyzed while for the attribute “effort
at least 20 min breaks were taken between samples to minimize the to chew” the Minimum Intensity (Imin) and time to reach minimum
effects of fatigue. intensity (Tmin) were examined.
For textural analysis the chewing gum samples were placed in 3.5
mL caps (1.5 cm diameter and 1.5 cm length) and analyzed with a
TA-XT2 Texture Analyzer (Godalming, Surrey, UK) equipped with a RESULTS AND DISCUSSION
cylindrical stainless steel probe (1 cm diameter and 5 cm length). The The concentrations of the four aroma compounds in the
probe penetrated the first 2 mm of the product at 2 mm/s and the total chewing gum samples made with PG, MCT, TA, or NFS were
work done was recorded from the area of the force-distance curve for
determined prior to the analytical and sensory studies. The
chewing gum chewed at a given interval. Five samples were evaluated
average quantities per gram of chewing gum for cinnamalde-
for each treatment.
hyde, L-carvone, piperitone, and jasmone were reported to be
For gum volume analysis, the samples were initially placed onto a
5.2, 0.5, 0.1, and 0.1 mg, respectively. No significant differences
Kimwipe absorbent napkin (Kimtech, Ontario, Canada), to remove
(R ) 0.05) were reported for each compound between these
additional saliva and then placed into a 10 mL volumetric cylinder
samples (coefficient of variation was <10%).
containing 10 mL of water. The volume displaced by the addition of
chewing gum was then measured using a magnifying glass. Each sample
To investigate the influence of flavor solvent type or addition
was measured in triplicate. on aroma release in chewing gum, the release profiles of
cinnamaldehyde, carvone, piperitone, and jasmone from samples
Plasticity Index Analysis of Gum Base with Flavor solvent.
Molten gum bases (100 g) containing 2.7 g of TA, PG, MCT, or NFS
made with PG, MCT, and TA or with NFS over a 12 min
were poured into an aluminum weighing pan (7.5 cm diameter and 1.5 consumption period were determined. The release curves for
cm depth), cooled for 24 h, and analyzed with an Instron 4444 universal cinnamaldehyde and carvone are illustrated in Figure 1 (pip-
machine (Instron Corp., MA) equipped with a spherical probe (diameter: eritone and jasmone not shown). For each compound, similar
12.6 mm). The probe penetrated the gum bases to distance of 1.016 release patterns were observed among these chewing gum
mm at a speed of 2.54 mm/min. The plasticity index was calculated samples for all three panelists. No statistically differences (R
based on the equation listed below. Theoretical chord value was ) 0.05) were reported in the concentration of each compound
obtained from trigonometric calculations based on the dimensions of (based on 6 s moving average value) detected in the breath when
the spherical probe. After 24 h of indentation, an experimental chord comparing the different flavor solvent formulated chewing gums
value (ECV) was measured using Vernier calipers for each sample. at three time points (30, 70, and 150 s); the average coefficient
Influence of Solvent-Type on Chewing Gum Flavor J. Agric. Food Chem., Vol. 56, No. 9, 2008 3257

Figure 3. Time course of chewing gum volume made with NFS (0), PG
(square with dots), TA (square with horizontal stripes), or MCT (square
with vertical dotted lines) consumed over 4 min; average of triplicates
(95% Confidence interval for one representative panelist

Figure 1. Cinnamaldehyde (top) and L-carvone (bottom) release profile

from chewing gum samples made with NFS (-), PG (- -), TA (- - -),
and MCT (- - -) consumed over 12 min; each curve represents the mean Figure 4. Time course of chewing gum compressibility made with NFS
of three replicates subsequently smoothed by a 1.5 s moving average (0), PG (square with dots), TA (square with horizontal stripes), or MCT
trendline for one representative panelist. (square with vertical dotted lines) consumed over 12 min; average of five
replicates ( 95% confidence interval for one representative panelist.

sample mass, the TA containing samples would also be predicted

to have a higher volume during this initial time period, if the
sugar alcohol phase was released at a lower rate. Time course
analysis of the chewing gum volume during mastication for these
samples was determined and the data for one panelist is reported
in Figure 3 (similar findings were reported for the other two
panelists, data not shown). As predicted, the TA formulated
chewing gum sample had the largest volume during the initial
mastication period (up to 90 s).
We also observed, although not as part of the research
objectives of this study, that the release profile of cinnamalde-
Figure 2. Sorbitol release profile from chewing gum samples made with
hyde was correlated to the release of the sugar alcohol phase
NFS (-[-), PG (-9-), TA (-2-), or MCT (- × -) consumed over 8
(see Figures 1 and 2) which would not be anticipated based on
min; average of triplicates for one representative panelist.
the estimated log P value of cinnamaldehyde (see Table 1). In
of variation for each panelist was <30% for each compound. a subsequent study, Potineni and Peterson (14) reported that
Therefore, flavor solvent addition or type had no apparent affect cinnamaldehyde and sorbitol generated transient hemiacetal
on the release properties of the select aroma compounds (log P reaction products during chewing gum manufacture/storage
values ranged from 1.90 to 3.55, Table 1) in these chewing which were hydrolyzed back to free “cinnamaldehyde” and
gum samples. sorbitol in the oral cavity during mastication. Because these
In a parallel study, the release profile of sorbitol from the transient hemiacetal reaction products of cinnamaldehyde were
same samples over a 12 min consumption time period was more polar than the free “cinnamaldehyde”, a more rapid release
monitored and is reported in Figure 2. Overall, the triacetin- of cinnamaldehyde was observed.
containing chewing gum had a lower initial sorbitol release rate To establish if the “softness” of chewing gum, as influenced
in the first 70 s compared to the rest of gum samples for all by flavor solvent, was related to the unique sugar alcohol release
panelists. As anticipated, analogous release patterns were also profile of the TA-chewing gum (Figure 2), the compressibility
observed with hydrogenated glucose syrup (HGS), the second of these samples over a 12 min mastication period was
most abundant sugar alcohol in the chewing gum (data not monitored. The data for one panelist is illustrated in Figure 4.
shown). Statistical analyses of the sorbitol (or HGS) concentra- Similar findings were also reported for the other two panelists,
tions in the saliva were found to be significantly lower for the data not shown. Overall, chewing gums made with TA or MCT
triacetin gums at 30 and 70 s (see Table 3) but not statistically were found to be statistically softer (based on “total work done”)
different at 120 s. Considering that the sugar alcohol phase of than the chewing gum samples made without a solvent or with
these samples made up approximately 65 g/100 g of the total PG. There was no statistical difference in softness between the
3258 J. Agric. Food Chem., Vol. 56, No. 9, 2008 Potineni and Peterson

Table 2. Plasticity Index Values of Gum Base Samples Made with PG,
TA, MCT, or NFS

treatments plasticity index (PI)a (range 0-1)

NFS 0.97 A
PG 0.83 B
TA 0.76 C
MCT 0.98 A

a
Average of six replicates; different letters (A–C) indicate a statistically significant
difference between samples (R ) 0.05).

Table 3. Sorbitol Concentration in the Saliva during Mastication of

Chewing Gum Samples Made with PG, TA, MCT, or NFS at the 30 and
70 s Time Points

sorbitol concentration (mg/mL saliva)a

treatment 30 s 70 s
NFS 103.5 A 96.8 A
PG 99.8 A 96.2 A
TA 81.0 B 120.7 B
MCT 109.6 A 101.1 A

a
Different letters (A-B) indicate a statistically significant difference between
samples (R ) 0.05); average of triplicate for all three panelists.

TA- or MCT-chewing gums during the initial phase (at the 30

or 60 s time point) for all three panelists. Consequently, a softer
textured chewing gum did not necessarily result in the faster
release of the sugar alcohol (e.g., the MCT formulated chewing
gum; Figure 2). This suggested the softness of the sample as
predicted by compression tests was not a good indicator of the Figure 5. Sensory time-intensity analysis of (i) cinnamon-like aroma, (ii)
flavor release potential of chewing gum at constant chewing sweetness, and (iii) effort to chew of chewing gum samples made with
rates. Based on the similar textural properties yet different sugar NFS (s), PG (zig-zag line), TA (- -), or MCT (- - - · ); scale 0-15,
alcohol release rates of the TA and MCT chewing gum samples, average of 9 panelists.
it was hypothesized that TA was primarily plasticizing or and likely is softening the lipid or discontinuous phase in the
softening the gum base polymeric continuous phase (polyvinyl chewing gum, whereas TA softened the polymeric continuous
acetate and so on) whereas MCT was mainly softening the lipid phase of chewing gum. The more plasticized (flowable) gum
or discontinuous phase of the gum base material. Therefore, base for the TA sample was therefore correlated to the slower
based on the solubility properties of these two solvents, two release properties of the sugar alcohol phase during mastication
different mechanisms of softening the chewing gum were in chewing gum.
proposed. Triacetin would be predicted to be more soluble with To further investigate the role of the instrumental findings in
polyvinyl acetate (similar structureslike dissolves like) and this study on flavor perception, a time-intensity sensory analysis
likewise MCT (more lipophilic) would be more soluble with profile was determined for the sweetness, cinnamon-like aroma,
the lipid/wax component of the gum base. We suspected that a and “effort to chew” attributes from the different solvent-type
more plasticized continuous phase would be softer or more formulated chewing gum samples. The results are shown in
flexible and would be anticipated to entrap the discontinuous Figure 5. The perceived sweetness of the TA chewing gum
phase sugar alcohol more efficiently during consumption which was significantly lower in maximum intensity (Imax) whereas
resulted in the delayed release of the sorbitol phase as reported the time to maximum intensity (Tmax) was found to be
in Figure 2. significantly higher (Table 4) in comparison to the other
To support the hypothesis that TA primarily softened the chewing gum samples. This is in agreement with the slower
continuous polymeric phase, whereas MCT softened the dis- sorbitol release profile reported from the instrumental analysis
continuous lipid phase of the chewing gum, the influence of (Figure 2). However, the maximum concentration of sorbitol
flavor solvent on gum base plasticity was determined. The released from the TA chewing gum was not observed to be
plasticity index value for the gum base samples mixed with an lower than other chewing gum samples (Figure 2), which
equivalent load of flavor solvent, in comparison to the chewing indicated the lower rate of sorbitol release was related to the
gum composition, are reported in Table 2. The TA-gum base decreased sweetness perception and not to the absolute con-
statistically had the lowest plasticity index value indicating the centration of the sorbitol reported in the mouth. The Imax for
addition of TA resulted in a continuous phase which was more the cinnamon-like aroma of the TA chewing gum sample was
elastic and more flexible for deformation when compared to statistically lower than the other samples, whereas the Tmax for
the other solvents. In contrast, the addition of MCT did not the TA chewing gum sample was significantly higher than
influence the gum base plasticity as this sample was not the MCT chewing gum. These findings are in contrast to the
statistically different from the plasticity value for the gum base instrumental breath analysis which found flavor solvent-type
with NFS. This supported the premise that, although both TA or addition did not influence the aroma release profile these
and MCT resulted in softer chewing gum (Figure 4), the samples (Figure 1). This indicated that for the TA chewing
mechanisms each solvent altered the chewing gum texture was gum, the noted slower sugar alcohol release profile was
unique. MCT did not plasticize the continuous polymeric phase correlated to the suppression of the cinnamon-like aroma
Influence of Solvent-Type on Chewing Gum Flavor J. Agric. Food Chem., Vol. 56, No. 9, 2008 3259

Table 4. Sensory Parameters of Chewing Gum Samples Made with PG, LITERATURE CITED
TA, MCT, or NFS
(1) Davidson, J. M.; Linforth, R. S. T.; Hollowood, T. A.; Taylor,
sensory parametersa A. J. Effect of sucrose on the perceived flavor intensity of chewing
cinnamon-like gum. J. Agric. Food Chem. 1999, 47, 4336–4340.
sweetness aroma intensity effort to chew (2) Duizer, L. M.; Bloom, K.; Findlay, C. J. Dual-attribute time-
intensity measurement of sweetness and peppermint perception
time time time
of chewing gum. J. Food Sci. 1996, 61, 636–638.
max. to max. max. to max. min. to min.
intensity intensity intensity intensity intensity intensity (3) de Roos, K. B. Effect of texture and microstructure on flavor
treatments (Imax)b (Tmax) [s] (Imax)b (Tmax) [s] (Imin)b (Tmin) [s] retention and release. Int. Dairy J. 2003, 13, 593–605.
(4) Anonymous Gum base-flavor interactions; Cafosa Gum C.A.:
NFS 9.06 A 39.56 A 8.83 A 49.44 AB 4.43 A 62.11 A
PG 9.06 A 37.41 A 8.67 A 50.04 AB 3.76 B 68.89 A
Barcelona, Spain, 2007; www.cafosa.com.
TA 8.34 B 47.74 B 8.15 B 59.44 A 2.69 C 69.11 A (5) Baek, I.; Linforth, R. S. T.; Blake, A.; Taylor, A. J. Sensory
MCT 9.07 A 39.11 A 8.69 A 43.85 B 3.15 C 48.85 B perception is related to the rate of change of volatile concentration
in-nose during eating of model gels. Chem. Senses 1999, 24, 155–
a
Different letters (A–C) indicate a statistically significant difference between 160.
samples (R ) 0.05); average of nine panelists b a 15 cm line scale was used for (6) Hansson, A.; Giannouli, P.; van Ruth, S. The influence of gel
evaluations (0 ) none, 15 ) very high). strength on aroma release for pectin gels in a model mouth and
in vivo, monitored with proton-transfer-reaction mass spectrom-
perception. This observation was consistent with previous
etry. J. Agric. Food Chem. 2003, 51, 4732–4740.
research that reported the perception of mint flavor intensity in
(7) Weel, K. G. C.; Boelrijk, A. E. M.; Alting, A. C.; van Mil, P.;
a minted flavored chewing gum was correlated to the sucrose
Burger, J. J.; Gruppen, H.; Voragen, A. G. J.; Smit, G. Flavor
concentration, suggesting taste-aroma interactions or that sweet- release and perception of flavored whey protein gels: Perception
ness level influenced the perception of the aroma intensity (1). is determined by texture rather than by release. J. Agric. Food
The influence of flavor solvent-type on the perceived chewing Chem. 2002, 50, 5149–5155.
gum textural properties (effort to chew, Figure 5) was also (8) Mestres, M.; Kieffer, R.; Buettner, A. Release and perception of
analyzed and was in general agreement with the analytical ethyl butanoate during and after consumption of whey protein
textural measurements (Figure 4). Overall, chewing gum with gels: Relation between textural and physiological parameters. J.
TA and MCT were found to be softer or required less effort to Agric. Food Chem. 2006, 54, 1814–1821.
chew versus with PG or NFS (Imin and Tmin, Table 4). Although (9) Mestres, M.; Moran, N.; Jordan, A.; Buettner, A. Aroma release
the analytical textural measurements reported that the chewing and retronasal perception during and after consumption of flavored
gum samples made with TA and MCT were comparable in whey protein gels with different textures. 1. In vivo release
softness (Figure 4), the sensory data showed the level at which analysis. J. Agric. Food Chem. 2005, 53, 403–409.
the softness was perceived in the first 2 min were statistically (10) Lethuaut, L.; Weel, K. G. C.; Boelrijk, A. E. M.; Brossard, C. D.
different (“effort to chew” in Figure 5). Gums with MCT had Flavor perception and aroma release from model dairy desserts.
the lowest ratings for “effort to chew” around 40 s whereas for J. Agric. Food Chem. 2004, 52, 3478–3485.
the gums with TA had the lowest ratings around 70 s, again (11) Lethuaut, L.; Brossard, C.; Meynier, A.; Rousseau, F.; Llamas,
suggesting different plasticizing affects of these flavor solvents G.; Bousseau, B.; Genot, C. Sweetness and aroma perceptions in
on the gum base (Table 4). The “effort to chew” analysis was dairy desserts varying in sucrose and aroma levels and in textural
not, however, correlated to the perceived cinnamaldehyde flavor agent. Int. Dairy J. 2005, 15, 485–493.
intensity. (12) Carroll, T. VH1 gum base composition Personal communication:
In summary, the flavor properties of chewing gum were Hershey Foods Corporation, Hershey, PA, 2006.
influenced by the addition of TA but not by the addition of the (13) Schober, A. L.; Peterson, D. G. Flavor release and perception in
hard candy: Influence of flavor compound-compound interactions.
other flavor solvents, PG or MCT. The TA formulated chewing
J. Agric. Food Chem. 2004, 52, 2623–2627.
gum had a lower sugar alcohol release rate during mastication
(14) Potineni, R.; Peterson, D, G. Mechanisms of flavor release in
and likewise was reported to have a suppressed sweetness and
chewing gum: Cinnamaldehyde. J. Agric. Food Chem. 2008, 56,
aroma intensity. Although both TA and MCT resulted in softer 3260–3267.
chewing gum, the TA sample uniquely plasticized the continu-
ous phase of the gum base. Therefore, the influence of flavor
Received for review September 18, 2007. Revised manuscript received
solvent on the flavor properties of chewing gum was not
February 15, 2008. Accepted February 16, 2008.
correlated to the chewing gum softness but rather to plasticiza-
tion of the polymeric phase of the gum base. JF072783E