Chitosan Nanoparticles for Loading of Toothpaste Actives

and Adhesion on Tooth Analogs

Hui Liu, Bo Chen, Zhengwei Mao, Changyou Gao

Key Laboratory of Macromolecule Synthesis and Functionalization, Ministry of Education,
Department of Polymer Science and Engineering, Zhejiang University, Hangzhou 310027, China

Received 28 April 2007; accepted 28 June 2007

DOI 10.1002/app.27078
Published online 7 September 2007 in Wiley InterScience (www.interscience.wiley.com).

ABSTRACT: Delivery and sustained release of tooth- ticles showed good stability in toothpaste lixivium after
paste actives is an important but unexplored area. In this incubated at 608C for 30 days too. By contrast, the chito-
work, chitosan nanoparticles were prepared by a water-in- san/cetylpyridiniumchloride nanoparticles were easy to
oil emulsion/glutaraldehyde crosslinking method. The form occules in the toothpaste lixivium. The loaded
typical number average diameter of chitosan and tooth- toothpaste actives showed a sustained released behavior
paste active (cetylpyridiniumchloride and NaF) nanopar- for at least 10 h. All the particles could adhere onto the
ticles was within the range of 100500 nm. The particles tooth analogs such as hydroxyapatite discs and glass slides
increased their size at higher pH value. The morphology, in a simulated brushing and rinsing process. In vitro cell
adherence, and stability of these nanoparticles were inves- culture did not nd any cytotoxicity of the as-prepared
tigated by scanning electron microscopy, transmission elec- chitosan nanoparticles. 2007 Wiley Periodicals, Inc. J Appl
tron microscopy, and X-ray photoelectron spectroscopy. Polym Sci 106: 42484256, 2007
The size of the chitosan/NaF nanoparticles was doubled
after they were stored at 48C for 20 days, and then kept Key words: adhesion; chitosan nanoparticles; emulsion
constant till 251 days, the examined time so far. These par- crosslinking; toothpaste actives

INTRODUCTION sion, as shown in Eq. (2):

Decayed tooth is one of the oral diseases badly inu-


encing the daily life of people. The tooth walls are 5Ca2 3PO3

4 F @ Ca5 PO4 3 F # 2
frequently and unavoidably corroded by a great
amount of acidic substances which are produced by Thus, the uorine ions are benecial of tooth remi-
remnant food, making the teeth fragile and leading neralization, reducing the occurrence of dental
to the tooth erosion nally. The tooth surface is caries.1 On the other hand, cetylpyridiniumchloride
mainly composed of hydroxyapatite (HA), which (CPC, C21H38ClNH2O) is another multifunctional
has equilibrium in saliva as shown in Eq. (1): antibacterial drug with good attachment ability on


skin, which is suitable for persisting antibacterial
Ca5 PO4 3 OH 5Ca2 3PO3

4 OH 1 application. However, a simple addition of these
toothpaste actives has a very short-term effect on the
Besides of teeth brushing to remove the remnant teeth, i.e. during the course of teeth brushing. After
foods, toothpaste actives such as uorine ions can gargling, the actives are mostly lost. Drug delivery
harden the teeth by forming a more stable com- system is an effective way to realize the sustained
pound, which has stronger ability against acid ero- release of many kinds of drugs, yet it is scarcely
applied to the toothpaste actives.
Chitosan has been applied widely in the medical
pharmaceutical elds.2 It has abundant amino and
hydroxyl groups in its molecule, thus can bind
Correspondence to: C. Y. Gao (cygao@mail.hz.zj.cn).
Contract grant sponsor: Major State Basic Research strongly to negatively charged substances such as
Program of China; contract grant number: 2005CB623902. cell surfaces, mucus or other polymer surfaces via
Contract grant sponsor: Natural Science Foundation of electrostatic or hydrogen bonding. Therefore, chito-
China; contract grant number: 20434030. san is extremely useful as vehicle for mucoadhesion
Contract grant sponsor: National Science Fund; contract drug delivery.3 Moreover, chitosan also can acceler-
grant number: 50425311.
ate the wound healing, inhibit bacteria growth, and
Journal of Applied Polymer Science, Vol. 106, 42484256 (2007) alleviate pain.4,5 Since the tooth surface is electroneg-
V
C 2007 Wiley Periodicals, Inc. ative, we expect that the positively charged chitosan
TOOTHPASTE ACTIVES AND ADHESION ON TOOTH ANALOGS 4249

is one of the best candidate materials as delivery ve- for the toothpaste actives, exemplied here with NaF
hicle for toothpaste actives. and CPC. Different from the reported methods for
Nowadays, chitosan and their derivative materials, the chitosan nanoparticle fabrication, we adopt here
in particular with a format of particles, have been a water-in-oil emulsion dispersion technique, which
diversely employed in the eld of drug delivery. produces more stable chitosan nanoparticles against
Various methods such as emulsion crosslinking,6,7 agglomeration. The adhesion, drug loading, and
ionotropic gelation,810 emulsication/solvent evapo- release of the chitosan nanoparticles shall also be
ration,11 spray drying,12 and coacervation/precipita- presented.
tion13,14 have been developed to prepare the chitosan
microparticles. For example, Thanoo et al. prepared
chitosan microspheres (425600 lm) by glutaralde- MATERIALS AND METHODS
hyde crosslinking of an aqueous acetic acid disper- Materials
sion of chitosan in parafn oil.6 The highly cross-
linked microspheres showed a slower release rate Chitosan (Mn 5 620 kDa, degree of deacetylation
whereas the less crosslinked ones showed a faster 5 90%, viscosity 5 115 cps) was a commercial prod-
rate observed in simulated gastric and intestinal u- uct of Haidebei Halobios. (Jinan, China). Cetylpyri-
ids. He et al. prepared both uncrosslinked and cross- diniumchloride (CPC), NaF, and Crest toothpaste
linked chitosan microparticles (210 lm) by a spray (containing NaF and SiO2 particles besides toothpaste
drying method for H2-receptor antagonists.12 The basic components) were kindly provided by P&G.
positively charged microspheres are benecial of (Guangzhou, China). Glutaraldehyde (25% solution)
enhancing the mucoadhesion, enabling them suitable and lysozyme were of biological grade. HA discs
for delivery of drugs via the gastrointestinal or nasal were from Hitemco Medical Application, USA. Bo-
routes of delivery. The mucoadhesive properties of vine serum albumin (BSA) and fetal calf serum (FCS)
the microspheres can be mediated by the prepara- were purchased from Sijiqing Biotech (Hangzhou,
tion parameters. For example, at a lower crosslinking China). Dulbeccos minimum essential medium
degree, both the particle size and the mucoadhesive (DMEM) was purchased from GibcoBRL. All other
properties were improved. Chitosan microparticles chemicals were of analytical grade and used as
(2030 lm) have also been prepared by a water-in- received. The water used in all experiments was tri-
vegetable oil emulsion coalescence technique, using ple distilled.
metal ions as the chelating reagents.15
Because of the special smooth characteristic of
Preparation of the chitosan nanoparticles
tooth surface, the microparticles show very poor ad-
hesion property. More recently, nanoparticles have The chitosan nanoparticles, either loaded with drugs
brought much attention in virtue of their large drug or not, were prepared by an emulsion dispersion
loading capacity, good adsorption performance, and technique. In a typical experiment, in a 250 mL
long shelf life. Several techniques have been devel- round-bottom ask 100 mL olive oil was mixed with
oped to prepare chitosan nanoparticles. On the basis 1 mL Span 80 and 0.25 g magnesium stearate at
of the ionic gelation of chitosan with tripolyphos- room temperature. Four milliliter of 1% chitosan
phate anions, the chitosan nanoparticles and copper- solution in 3 vol % acetic acid was diluted to 20 mL
loaded nanoparticles were prepared.16 Antibacterial with water (nal concentration 2 mg/mL), which
studies showed that the nanoparticles could in- was then poured into a ask. The mixture was
hibit the growth of various bacteria, most probably stirred at 450 rpm for 1 h with a mechanical stirrer
following a mechanism of membrane disruption to form a water-in-oil emulsion. 0.5 mL of 3.1% glu-
and leakage of the cellular proteins. Ultrane chito- taraldehyde solution was injected into the mixture to
san nanoparticles were also prepared in an AOT crosslink the chitosan molecules and stabilize the
(sodium bis(ethylhexyl) sulfosuccinate)/n-hexane chitosan nanoparticles. Two hours later, the tempera-
reverse micellar system and crosslinked by glutaral- ture was improved to 408C by a water bath. Under
dehyde.17 After intravenous injection the nanopar- agitation at 400 rpm the reaction was continued for
ticles could circulate in blood and distribute in the 6 h. After the oil layer was separated, the pH value
bone marrow. of the water phase was adjusted to 8 by 0.1M
Up to present, however, the chitosan nanoparticles NaOH under vigorous agitation. The water phase
are scarcely applied as delivery vehicle for tooth- was then centrifugated at 3000 rpm to remove the
paste actives. Besides the physical adsorption of the unreacted chitosan and bigger particles. Finally, the
chitosan molecules on the tooth surface, the very supernatant was adjusted to neutral pH and stored
small size of the chitosan nanoparticles benets the at 48C. The drug-loaded nanoparticles were similarly
adhesion too. Therefore, in this study, we make use prepared by using a chitosan solution containing the
of the chitosan nanoparticles as the delivery vehicle drugs.

Journal of Applied Polymer Science DOI 10.1002/app

4250 LIU ET AL.

Morphology of the chitosan nanoparticles 20 mL nally. The concentration was quantied by

referring to a calibration curve recorded from known
Scanning electron microscopy (SEM; SIRION-100,
amount of NaF at the same condition.
FEI), transmission electron microscopy (TEM; JEOL
The loading capacity was calculated by the drug
JEM-200CX, Japan), and scanning force microscopy
weight/the particle weight 3 100%.
(SFM; SPI3800N Probe Station and SPA400 SPM
Unit, Seiko) were used to observe the chitosan nano-
particles. The samples were prepared as follows. Release of the toothpaste actives
SEM: A drop of the nanoparticle suspension was
applied onto a HA disc or a H2O2/H2SO4 (3/7, v/v) The CPC release was performed by putting 20.2 mg
treated (1 h) glass slide, dried at 408C and xed on nanoparticles into a dialytic-bag (cut-off molecular
an aluminum stub with a conductive adhesive tape, weight Mw: 814.4 kDa). The bag was then
then sprayed with a gold layer under vacuum with immersed into 30 mL PBS (pH 6.8) at 378C under
a Sputter Coater Edwards S 150A. The nanoparticles continuous shaking. The dialytic PBS was exchanged
were observed also after freeze-dried. The accelera- every 1 h. The release of NaF was performed by dis-
tion voltage was 5 kV. persing 4 mg freeze-dried nanoparticles in 3 mL PBS
TEM: A drop of the nanoparticle suspension was (pH 6.8) at 378C under continuous shaking. The
applied onto a carbon-coated copper meshwork, nanoparticle suspension was periodically centrifuged
dried at room temperature. The acceleration voltage and 1.5 mL supernatant was exchanged with the
was 100 kV. same volume of freshly prepared PBS every 2 h. The
SFM: A drop of the particle suspension was released amount of CPC or NaF in the supernatant
applied onto newly cleaved mica, dried at 408C. Sili- was quantied as aforementioned.
con tips with a resonance frequency f0 of 150 kHz
and a spring constant of 16 N/m were utilized. The Adhesion property of the chitosan nanoparticles
scanning frequency was 1 Hz. mixed in toothpaste
The adhesion property of the chitosan nanoparticles
Size of the chitosan nanoparticles was performed by mixing the particles with tooth-
paste, and under a condition simulating the tooth
The size of the nanoparticles was measured by a brushing process. The chitosan nanoparticle suspen-
dynamic light scattering particle size analyzer (DLS) sion was mixed with toothpaste for 1 h. The HA
(90 Plus/BI-MAS). Each measurement was taken for disc or glass slide was immersed in simulated saliva
1 min with a 908 xed angle detector. The diameter (BSA/PBS 1/4 (v/v), pH 5 6.8) for 40 min, then
was averaged from ve parallel measurements. brushed with the chitosan nanoparticle/toothpaste
mixture. Finally, the HA disc or glass slide was
rinsed with water for 3 times to remove the non-
Loading amount of the toothpaste actives
tightly adhered components. After dried, the HA or
The nanoparticles containing the drugs were centri- glass surface was observed under SEM. To detect
fugated at 8000 rpm for 10 min to obtain the nano- the adhesion more sensitively and precisely, CdCl2
particle precipitate, which was washed with water was incorporated into the chitosan nanoparticles.
twice, and freeze-dried. 5.1 mg freeze-dried chito- Following the same mixing and brushing process,
san/CPC nanoparticles were then dispersed in the surface elements of the HA disc were deter-
10 mL ethanol under continuous shaking for 12 h to mined by X-ray photoelectron spectroscopy (XPS) on
dissolve the CPC. After centrifugation, the absorb- an ESCA LAB Mark II spectrometer employing
ance of the supernatant at 259 nm was recorded by a mono X-ray Al Ka (hm 5 1486.6 eV), 150 w, 15 kV
UVvis spectrophotometer (Shimadzu UV-2550). The excitation radiation. The base pressure was 2 3 1029
concentration was quantied by referring to a cali- mbar. The charging shift was referred to the C(1S)
bration curve recorded from known amount of CPC line emitted from the saturated hydrocarbon. The
at the same condition. take off angle of the XPS was 908. The constant angle
For determination of the NaF amount, the chito- energy was 20 eV.
san/NaF nanoparticle precipitate was dispersed in
3 mL lysozyme solution (10 mg/mL) at 378C for
5 days to degrade the chitosan molecules completely. Stability of the chitosan nanoparticles in
toothpaste lixivium
After centrifugation, the supernatant was collected
and the potential of uorine was measured by a Toothpaste was mixed with certain amount of water,
uorine ion-selective electrode, with 10 mL citrate which was centrifugated to get the lixivium. For a
sodium as whole regulating regent of ion intensity, speeding up experiment, the nanoparticle suspen-
7.5 mL water and certain amount of PBS to maintain sion/lixivium mixture was kept at 608C for 30 days.

Journal of Applied Polymer Science DOI 10.1002/app

TOOTHPASTE ACTIVES AND ADHESION ON TOOTH ANALOGS 4251

Figure 1 (a) SEM, (b) TEM, and (c) SFM images of chitosan nanoparticles. (d) Line prole recorded at the place shown
in (c).

The morphology of the nal products was observed cultured with the nanoparticles for 48 h. After
under SEM. washed with PBS to remove the remaining glutaral-
dehyde, the samples were dehydrated with a graded
series of ethanol. Then the samples were further
Biocompatibility of the chitosan nanoparticles dehydrated with acetone and treated with isoamyl ac-
In vitro cell culture was used to determine the bio- etate. After dried by the critical point dry method, the
compatibility of the as prepared chitosan nanopar- samples were coated with ultrathin gold layers and
ticles. Pure chitosan nanoparticles were sterilized by observed under SEM (Cambridge stereoscan 260).
UV irritation before cell culture. Human dermal
broblasts were isolated from foreskins and rou-
tinely expanded.18 The broblasts were incubated in RESULTS AND DISCUSSION
a culture medium consisting of 10% (v/v) FCS and
Morphology of the chitosan nanoparticles
90% (v/v) DMEM, supplemented with 100 U/mL of
penicillin and 100 cg/mL of streptomycin in humidi- Because of the highly sticky property of chitosan
ed air containing 5% CO2 at 378C. About 200 lL molecules, the chitosan particles, in particular in a
human dermal broblast suspension was seeded in a nanometer range, are rather easy to combine with
96-well polystyrene plate, with a nal cell number of each other. For example, the as-prepared chitosan
7.5 3 104 per well. After 24 h, 20 lL nanoparticle nanoparticles without the presence of magnesium
solution was added into each well. The nal nano- stearate formed larger particle clusters. By contrast,
particle concentration was high enough for close addition of the magnesium stearate can largely alle-
contact between the cells and the nanoparticles. The viate the agglomeration of the particles,20 resulting
culture medium was changed every 2 days. The cell in well dispersed chitosan nanoparticles [Fig. 1(a)].
proliferation was measured using methylthiazolete- The nanoparticles show more condense structure
trazolium (MTT) method.19 The absorbance was with smoother surface morphology too. Here the
recorded at a wavelength of 570 nm by a microplate nonionic emulsier, i.e. Span 80, also takes an im-
reader (Bio-Rad model 550). Three parallel experi- portant role for the particle dispersion during the
ments were conducted and data were expressed as fabrication process. On the water/oil interface, the
mean 6 standard deviation. nonpolar groups of span-80 and the lipophilic group
To observe the cell morphology under SEM, (stearyl) of magnesium stearate dip into the oil
the cells were washed with PBS and xed with 2.5% phase, while the three hydroxyls of span-80 and
glutaraldehyde in PBS at 48C for 48 h after co-

COO2 of magnesium stearate insert into the aque-

Journal of Applied Polymer Science DOI 10.1002/app

4252 LIU ET AL.

Figure 2 (a) SEM and (c) TEM images of chitosan/NaF nanoparticles. (b) is higher magnication of (a).

ous phase, stabilizing the droplets. Meanwhile, the 100200 nm. Normally, chitosan nanoparticles with
charged interface can also stabilize the droplets this size tend to aggregate, leading to formation of
because of the repulsion of the same charge.21 After larger aggregates because of strong interparticle
crosslinking by glutaraldehyde, the shape and size interaction. However, as shown in Figures 1 and 2,
of the chitosan droplets are xed. Therefore, by this the chitosan nanoparticles prepared by the present
process the chitosan nanoparticles with good disper- process have rather good dispersivity, which is cru-
sivity are successfully obtained. In the following cial for delivery of drug by a mucoadhesive manner.
experiments, all the chitosan particles were fabri-
cated under the existence of magnesium stearate.
The chitosan nanoparticles were further subjected Size variation of the chitosan/toothpaste active
nanoparticles
to TEM and SFM characterizations [Fig. 1(b,c)]. Again
spherical and well dispersed chitosan nanoparticles Measured by DLS, all the chitosan nanoparticles
can be observed, whose size is consistent with the present double peak distribution regardless of the
SEM observation (100500 nm). No severe agglomera- existence of drugs. For example, at pH 8 DLS meas-
tion of the particles was found. The line prole in ured two peaks centered at 140 and 484 nm, while
Figure 1(d) shows that the particles in dry state were the sizes increased slightly after toothpaste active
collapsed severely, leading to smaller value in height loading. Because of this nature, in the next discus-
( 40 nm) than that in width. This phenomenon is sion, we shall use only the number average diameter
very common for many soft polymeric particles, to represent the whole samples.
conveying a hint that the chitosan nanoparticles Since the charge density of the chitosan molecules
should have a rather loose inner structure. This is rea- is highly inuenced by environmental pH, the parti-
sonable since the theoretical volume in the particles cle size may also be varied accordingly. In the pres-
reaches >99% if no condensation occurs during the ent case, when the pH value was increased from pH
particle preparation. 7.18.3, the average diameter was sharply increased
The drugs loaded chitosan nanoparticles can be from 29 to 237 nm. This phenomenon is understood
similarly prepared, as shown in Figure 2. Here SEM as a result of particle coagulation because at high
[Fig. 2(a,b)] and TEM [Fig. 2(c)] images of freeze- pH the charge degree of the chitosan molecules is
dried chitosan/NaF nanoparticles are present as a reduced. The diminishing charge repulsion between
typical example. Most of the chitosan/NaF nanopar- particles will then result in particle aggregation to
ticles are separated from each other, with a size of decrease the surface free energy. Detail study of the

Journal of Applied Polymer Science DOI 10.1002/app

TOOTHPASTE ACTIVES AND ADHESION ON TOOTH ANALOGS 4253

was controlled at 48C to avoid the environmental

temperature uctuation. Figure 3 shows that the par-
ticle size was quickly doubled after 20 days, then
reached an equilibrium oscillating between 60 and
80 nm until the examined time so far (251 days).
This result demonstrates that the chitosan/NaF
nanoparticles are rather stable at neutral pH, at least
at low temperature.

Loading and release of toothpaste actives

The loading capacity of CPC and NaF within the chi-
tosan nanoparticles was measured as 14 and 10%,
respectively, which are only about 1/4 and 1/5 of
Figure 3 Alteration of average size of chitosan/NaF the theoretical value. However, these values are not
nanoparticles along with the storage time at 48C. low considering that both drugs are highly water
soluble, which may cause the loss during the rinsing
steps. The slight higher loading capacity of CPC
pH inuence was performed to the chitosan/tooth- may result from its relatively larger molecular size,
paste active nanoparticles, since they are the practi- and the existence of long aliphatic hydrocarbon
cally applicable delivery systems. Along with the pH chains.
increase, both kinds of the nanoparticles increased Both the loaded drugs can be released in a sus-
their sizes with similar alteration tendency. At the tainable manner, as shown in Figure 4. An initial
same pH value, the size of the chitosan/CPC nano- burst release was observed for both kinds of nano-
particles is always larger than that of the chitosan/ particles, which is a common phenomenon for many
NaF nanoparticles. For example, at pH values of 4.5, drug release systems.23,24 However, the initial burst
7.0, and 8.1, the sizes of the chitosan/CPC nanopar- release is not severe compared with other sys-
ticles were measured as 92, 127, and 194 nm, tems.12,25,26 The burst release is known as the result
whereas the sizes of the chitosan/NaF nanoparticles of quick release of the surface adsorbed drugs. Com-
were measured as 4, 46, and 140 nm, respectively. pared with NaF, the CPC was released with a very
When the pH value was decreased from 8.1 to 7.1, slow rate. For example, during the rst 3 h, 17% of
the size of the chitosan/CPC nanoparticles (137 nm) the loaded CPC and 50% of the loaded NaF were
decreased to their initial value again, demonstrating released, respectively. These values reached to 33%
that the aggregation or dispersion of the chitosan for CPC and 88% for NaF after 10 h, respectively.
nanoparticles is indeed governed by the charge The larger molecular size and the hydrophobic
interaction. It has to mention that here the particle hydrocarbon chains should account for the slower
aggregation must be taken into consideration, release rate of CPC.
because theoretically the chitosan particles should
decrease their size at high pH as a result of reducing
intramolecular charge repulsion. This phenomenon
has been observed for the ionically crosslinked chito-
san particles, in which the particles have smaller size
at higher pH.22 Different crosslinking mechanism
may result in different ionic intensity in the particles,
leading to the reverse alteration tendency of the par-
ticle size in response to pH stimulus.
Furthermore, we observed the sediment in all the
chitosan nanoparticle solutions with a pH value
higher than 8 after the solutions were stored for a
couple of days. This phenomenon reminds us that
the chitosan at higher pH value is intrinsically unsta-
ble. Because both the toothpaste lixivium and the
oral cavity environment are almost neutral, a detail
study was performed at pH 7.0 to explore the stor- Figure 4 Release proles of the toothpaste actives from
age stability, using the chitosan/NaF nanoparticles chitosan nanoparticles in PBS (pH 6.8) at 378C as a func-
as a typical example. Here the storage temperature tion of time.

Journal of Applied Polymer Science DOI 10.1002/app

4254 LIU ET AL.

Figure 5 SEM images of chitosan nanoparticles with different actives adhered on HA discs (upper row) or glass slides
(lower row). (a,b) HA disc surface, (c) chitosan nanoparticles on HA disc, and (d) chitosan/CPC nanoparticles on HA
disc; (e) glass slide surface, (f) chitosan nanoparticles, (g) chitosan/CPC nanoparticles, and (h) chitosan/NaF nanoparticles
on the glass slides. Arrows indicate the particles. (i) XPS spectrum to show Cd3d5 and N1s scan of HA surface adhered
with CdCl2 stained chitosan nanoparticles.

Adhesion property of the chitosan nanoparticles of the glass slide is very smooth [Fig. 5(e)]. As
shown in Figure 5(fh), a large number of nanopar-
As a drug delivery vehicle for toothpaste actives, the
ticles can be observed on the glass slides, regardless
adhesion property of the drug-loaded particles on
of loading of the drugs.
the tooth surface is of great importance. In situ drug
To unambiguously demonstrate the adhesion of
release can be achieved only when the particles are
the chitosan nanoparticles on the tooth analogs, the
stably adhered. Here we used tooth analogs as the
surface chemistry was analyzed by XPS. The survey
substrates to qualitatively check the adhesion of the
scan spectrum of the HA disc presents the P, Ca,
chitosan nanoparticles. As shown in Figure 5(a,b),
and O. After brushing with the CdCl2 stained chito-
the surface texture of the HA disc is very inhomoge-
san nanoparticles, extra Cd and N elements can be
neous, with rough morphology and micron-sized
observed as highlighted in Figure 5(i). Their atomic
particulates. After brushing the nanoparticle/tooth-
ratios are between 1.24 and 2.31%, respectively.
paste mixture and sufcient rinsing, tiny chitosan
These results demonstrate the chitosan particles are
and chitosan/CPC nanoparticles can be identied as
denitely adhered onto the HA disc surface.
pointed by the arrows [Fig. 5(c,d)]. To further con-
rm this adhesion and avoid the interference of the
rough surface of the HA disc, planar glass slide was Stability of the chitosan nanoparticles in
toothpaste lixivium
used as a model substrate, since both kinds of mate-
rials have plenty of

OH groups and electronega- The stability in the toothpaste lixivium and adhesion
tive charge. Compared with the HA disc, the surface ability after storage in the toothpaste lixivium was

Journal of Applied Polymer Science DOI 10.1002/app

TOOTHPASTE ACTIVES AND ADHESION ON TOOTH ANALOGS 4255

Figure 6 SEM images to show the morphology change. (a) The toothpaste lixivium and (b) its adherence on glass slide;
(c) chitosan nanoparticles and (d) their adherence on glass slide after mixed with lixivium and stored at 608C for 30 days;
chitosan/CPC nanoparticles after mixed with the toothpaste lixivium for (e) 10 min and for (f) 30 days (608C); (g) chito-
san/NaF nanoparticles and (h) their adherence after mixed with the toothpaste lixivium and stored at 608C for 30 days.

qualitatively studied by SEM. As shown in Figure at 608C for 30 days [Fig. 6(f)]. These results demon-
6(a), submicron particles can be found after the lixi- strate that unlike the pure chitosan nanoparticles,
vium was dried on glass slide, which may result the chitosan/CPC nanoparticles in toothpaste lixi-
from the added salt, tiny CaCO3 particles, and other vium are unstable. CPC is a quaternary ammonium
unidentied substances. After rinsing with water, no salt having positive charge. It can chelate with sac-
particulate materials can be found [Fig. 6(b)], indicat- charides and electronegative surfactants, which are
ing that the lixivium itself cannot adhere on the exclusively included in the toothpaste. The chelation
glass slide in a particulate format. When the chitosan will then cause formation of the occules, and nally
particles were suspended in the lixivium at 608C, no precipitation of the chitosan/CPC/lixivium mixture.
precipitation was observed until 30 days. SEM also This was further conrmed by directly mixing CPC
shows no occules in the chitosan particles/lixivium and the lixivium, where precipitation occurred im-
solution [Fig. 6(c)]. After brushing and rinsing, mediately. In contrast to the chitosan/CPC nanopar-
tightly adhered particles can be still found [Fig. ticles, the chitosan/NaF nanoparticles were very
6(d)]. When the chitosan/CPC nanoparticles were stable. No detectable occules or precipitates were
mixed with the lixivium, however, occules were observed macroscopically and microscopically [Fig.
observed within 10 min, as shown in Figure 6(e). 6(g)]. Similar to the chitosan nanoparticles, the chito-
The agglomeration became even severe after storage san/NaF nanoparticles could adhere on the glass

Figure 7 (a) Cytoviability of broblasts measured by MTT assay as a function of culture time. Blank control was made
by culture medium. And TEM images of broblasts cultivated for 24 h without (b) and with (c) the chitosan nanoparticles.

Journal of Applied Polymer Science DOI 10.1002/app

4256 LIU ET AL.

slide rmly too after brushing and rinsing [Fig. respect to long term storage in both pure water and
6(h)]. Together with the loading and release prop- toothpaste lixivium. Therefore, one can predict that
erty, one can conclude that the chitosan/NaF nano- the chitosan nanoparticles have great potentials to be
particle delivery system possesses a good potential used for delivery of toothpaste actives and for in situ
for application in toothpaste. release of the actives in a sustained manner.

Biocompatibility of the chitosan nanoparticles

References
Since the nanoparticles can be uptaken by cells upon
contact, it is necessary to assess the inuence of the 1. Wang, B. B.; Zheng, B. S.; Zhai, C.; Yu, G. Q.; Liu, X. J. Envi-
chitosan nanoparticles on the cytoviability. Here ron Int 2004, 30, 1067.
only the chitosan nanoparticles were assessed, since 2. Illum, L. Pharm Res 1998, 15, 1326.
3. Sinha, V. R.; Singla, A. K.; Wadhawan, S.; Kaushik, R.; Kumria,
the delivery vehicle is the main concern, while the R.; Bansal, K.; Dhawan, S. Int J Pharm 2004, 274, 1.
actives themselves have been well studied. For this 4. Howling, G. I.; Dettmar, P. W.; Goddard, P. A.; Hampson, F.
purpose, in vitro culture of human broblasts in the C.; Dornish, M.; Wood, E. J. Biomaterials 2001, 22, 2959.
presence of the chitosan nanoparticles was per- 5. Okamoto, Y.; Kawakami, K.; Miyatake, K.; Morimoto, M.;
formed. Figure 7(a) shows that no signicant differ- Shigemasa, Y.; Minami, S. Carbohydr Polym 2002, 49, 249.
6. Thanoo, B. C.; Sunny, M. C.; Jayakrishnan, A. J Pharm Phar-
ence of the cytoviability between the control (cells macol 1992, 44, 283.
only) and the samples supplemented with the nano- 7. Genta, I.; Costantini, M.; Asti, A.; Conti, B.; Montanari, L.
particles through all the culture period. The cells in Carbohydr Polym 1998, 36, 81.
both cases show typical and normal S-shaped 8. Pan, Y.; Li, Y. J.; Zhao, H. Y.; Zheng, J. M.; Xu, H.; Wei, G.;
growth behavior. Observed under SEM [Fig. 7(b,c)], Hao, J. S.; Cui, F. D. Int J Pharm 2002, 249, 139.
9. Berger, J.; Reist, M.; Mayer, J. M.; Felt, O.; Peppas, N. A.;
all the cells show elongated shape, which is typical Gurny, R. Eur J Pharm Biopharm 2004, 57, 19.
for the broblasts. These results demonstrate that the 10. Ko, J. A.; Park, H. J.; Hwang, S. J.; Park, J. B.; Lee, J. S. Int J
crosslinked chitosan nanoparticles have not brought Pharm 2002, 249, 165.
detectable cytotoxicity, hence can be safely used as 11. Genta, I.; Perugini, P.; Conti, B.; Pavanetto, F. Int J Pharm
the delivery vehicle for the toothpaste actives. 1997, 152, 237.
12. He, P.; Davis, S. S.; Illum, L. Int J Pharm 1999, 187, 53.
13. Bayomi, M. A.; Al-Suwayeh, S. A.; El-Helw, A. M.; Mesnad, A.
F. Pharm Acta Helv 1998, 73, 187.
CONCLUSIONS
14. Vandenberg, G. W.; Drolet, C.; Scott, S. L.; De La Noue, J.
Chitosan nanoparticles with an average number di- J Control Release 2001, 77, 297.
15. Kofuji, K.; Qian, C. J.; Murata, Y.; Kawashima, S. React Funct
ameter of 100500 nm, either loaded with toothpaste
Polym 2005, 62, 77.
actives or not, were prepared by a water-in-oil emul- 16. Qi, L. F.; Xu, Z. R.; Jiang, X.; Hu, C. H.; Zou, X. F. Carbohydr
sion technique, following by GA crosslinking. SEM, Res 2004, 339, 2693.
TEM, and SFM characterizations revealed that the 17. Banerjee, T.; Mitra, S.; Singh, A. K.; Sharma, R. K.; Maitra, A.
chitosan molecules and the loaded drugs were Int J Pharm 2002, 243, 93.
18. Mao, Z. W.; Ma, L.; Gao, C. Y.; Shen, J. C. J Control Release
evenly distributed within the particles, instead of
2005, 104, 193.
forming a core-shell structure. The chitosan nanopar- 19. Liu, Y. Y.; Yoshikoshi, A.; Wang, B. C.; Sakanishi, A. Colloid
ticles showed good stability at neutral pH, while Surf B Biointerfaces 2003, 27, 287.
precipitate quickly at alkaline condition with 20. Bogataj, M.; Mrhar, A.; Grabnar, I.; Rajtman, Z.; Bukovec, P.;
increased particle size at the initial stage. The loaded Srcic, S.; Urleb, U. J Microencapsul 2000, 17, 499.
21. Lim, L. Y.; Wan, L. S. C. J Microencapsul 1998, 15, 319.
drugs could be sustained released at least for 10 h,
22. Lopez-Leon, T.; Carvalho, E. L. S.; Seijo, B.; Ortega-Vinuesa,
with a release percentage of 33% for CPC and 88% J. L.; Bastos-Gonzalez, D. J Colloid Interface Sci 2005, 283, 344.
for NaF, respectively. The nanoparticles can adhere 23. Ozbas-Turan, S.; Akbuga, J.; Aral, C. J Pharm Sci 2002, 91,
onto tooth analogs such as HA discs and glass 1245.
slides, which are conrmed by direct SEM observa- 24. Janes, K. A.; Fresneau, M. P.; Marazuela, A.; Fabra, A.; Alonsoa,
M. J. J Control Release 2001, 73, 255.
tion and XPS analysis. Floccules were formed when
25. Zhou, S. B.; Deng, X. M.; Li, X. H. J Control Release 2001, 75,
the chitosan/CPC nanoparticles were mixed with 27.
toothpaste lixivium. By contrast, the chitosan/NaF 26. Wu, Y.; Yang, W. L.; Wang, C. C.; Hu, J. H.; Fu, S. K. Int J
nanoparticles showed very good stability with Pharm 2005, 295, 235.

Journal of Applied Polymer Science DOI 10.1002/app