Colloid Polym Sci (2003) 281: 614–623
`DOI 10.1007/s00396-002-0808-7
`
`O R I G I N A L C O N T R I B U T I O N
`
`Jong-Yun Kim
`Jun-Yeob Song
`Eun-Joo Lee
`Seung-Kyu Park
`
`Rheological properties and microstructures
`of Carbopol gel network system
`
`Received: 17 May 2002
`Accepted: 30 September 2002
`Published online: 1 February 2003
`Ó Springer-Verlag 2003
`
`J.-Y. Kim (&) Æ J.-Y. Song Æ E.-J. Lee
`S.-K. Park
`LG Household & Healthcare Research
`Park, 84 Jang-dong, Yusong-gu,
`Taejon 305-343, Republic of Korea
`E-mail: jykimx@lgcare.co.kr
`
`Abstract Carbopol gel systems
`have been studied using steady,
`oscillatory rheology, and cryoscan-
`ning electron microscopy (cryo-
`SEM) analysis in order to elucidate
`the nature of the different micro-
`structures of the gel in relation to
`polymer concentration as well as
`triethanolamine (TEA) content. The
`effect of changing the concentration
`of Carbopol (0.1–4 wt%) for 0, 1,
`and 10 wt% TEA has been investi-
`gated. Cryo-SEM revealed that
`honeycomb structures were observed
`in the gel system depending on the
`amount of TEA and Carbopol while
`the irregular fibrous three dimen-
`sional gel network systems were seen
`at the lower level of polymer content
`even in the high concentration of
`TEA. In addition to that, as the
`amount of polymer was increased,
`strings of fibrous network became
`thicker and of honeycomb-like
`structure. Shape of storage modulus-
`shear stress curve in the dynamical
`
`rheometric study was significantly
`changed as a result of variation in
`the microstructures while frequency
`sweep curve and yield values ob-
`tained from the model fitting in the
`steady rheological measurements
`couldn’t reflect the structural differ-
`ence of Carbopol gels. Two distinct
`relaxation phenomena were ap-
`peared with increase in polymer
`concentration as well as TEA con-
`centration. Temperature dependence
`of the stress sweep experiment was
`measured and shown that the effect
`of temperature (1–80 °C) on the
`shape of the curve was the similar
`trend with that of TEA and polymer
`concentrations, although the tem-
`perature dependency on the incre-
`ment was much weaker than TEA
`concentration.
`
`Keywords Rheology Æ Carbopol Æ
`Cryoscanning electron microscopy Æ
`Temperature effect
`
`Introduction
`
`Poly(acrylic acid) polymers as an anionic hydrogel are
`widely used to improve the rheological properties of
`thickening systems. Those cover a wide range of appli-
`cations from cosmetics to pharmaceutical uses for
`emulsification,
`stabilization and rheological control
`[1, 2, 3, 4, 5]. They are also used to control the release of
`medicaments from time-release tablets or from en-
`trapped systems [6, 7, 8, 9] as well as known to be a
`highly efficient
`thickener by forming a networked
`
`microgel structure in aqueous solutions [10]. Concept of
`microgel has been utilized in many hydrogel systems and
`applications as a vehicle for the drug and other active
`ingredient [11, 12, 13]. Microgel structure with intersti-
`tial spaces can help the suspended particles entrapped
`and stable for a sufficiently long time. There are various
`types of acrylic acid-based polymers commercialized
`[14]. In general, these polymers have poor ionic surfac-
`tant and electrolyte tolerance. Carbopol insensitive to
`electrolyte and ionic surfactant were therefore chosen in
`this study.
`
`1
`
`AMN1029
`
`

`

`Rheological analysis of this hydrogel polymer is im-
`portant, especially in the pharmaceutical applications,
`since its mucoadhesive performance is closely related to
`the rheological properties [15, 16, 17]. Rheological
`properties of Carbopol gels are extensively investigated
`and this polymer is known to form dispersion and show
`weakly viscoelastic properties, instead of dissolving in
`aqueous solution [18, 19, 20, 21, 22]. Early studies on
`Carbopol concluded that
`their unique rheological
`properties resulted from the interchain entanglements of
`the high molecular weight polyacrylate [23]. The
`Herschel-Bulkley model can be successfully applied to
`such a viscoplastic test media to analyze their rheologi-
`cal behavior [24]. High viscosity can only slow down the
`settling rate, and yield stress value is the predominant
`factor to provide a permanent suspension. The shear
`modulus (G°) represents interaction between swollen
`particles and should be proportional to the cross-link
`density [10]. Rheology of network dispersions was cor-
`related to network structure and cross-link density de-
`termined from swelling ratio measurement of microgel
`networks [10]. More precise analysis of network gel
`structure, however, has not been fully understood.
`Carbopol 941, one of the most popular and commer-
`cially available poly(acrylic acid), was chosen as a
`thickener to study the rheological behavior in the pres-
`ence of different amount of neutralizing agent,
`i.e.
`triethanolamine (TEA) [25].
`The purpose of this study was to understand visco-
`elastic properties of Carbopol correlated to the direct
`analysis of microscopic network structure by cryo-SEM.
`
`Experimental
`
`Materials. Carbopol 941 was provided by BF Goodrich as a white
`powder form, which is synthetic, high molecular weight of
`(21±3)·106 [26], nonlinear polymers of acrylic acid cross-linked
`with a polyalkenyl polyether [27, 28] containing up to 20% soluble
`linear polyacrylate [20]. The cross-linking agents for all the Carbo-
`pol 900 series are allyl ethers of either sucrose or pentaerythritol [29],
`although the crosslinking agent used in Carbopol 941 is not known.
`All other reagents are of analytical grade. Chemical compositions of
`Carbopol used in this experiment are shown in Table 1. All polymer
`solutions were prepared using de-ionized water (Barnstead E-pure
`System, Barnstead/Thermolyne Co., Dubuque, Iowa). The water
`had an electrical conductance of 18.2 MW cm.
`
`Preparations of polymer gel. For the gel state suspensions, pre-
`mixed aqueous solution of 4 wt% of poly(acrylic acid) polymer was
`prepared with a four-blade marine impeller, as recommended by
`the manufacturer, at room temperature, 400 rpm (Heidolph RZR
`2030, Germany) and diluted to the final concentrations. Neutral-
`ization of the poly(acrylic acid) polymer was accomplished with
`TEA. Gels produced at the final state have excellent clarity after
`being neutralized by TEA. Bubbles in a clear gel system must be
`controlled to satisfy marketing goals. The two major sources of
`bubbles are from mechanical entrapment and chemical generation.
`Mechanical bubble formation was minimized by careful dispersion
`of the polymer resins. In simple systems this can be handled by
`
`615
`
`Table 1 Chemical composition of Carbopol 941 in white powder
`form
`
`Impurities
`
`Water
`Benzene
`Propionic acid
`Acetic acid
`Acrylic acid
`Heavy metals
`Iron
`Arsenic
`Lead
`
`Typical values
`
`<0.5%
`1800 ppm
`1200 ppm
`600 ppm
`80 ppm
`10 ppm
`1 ppm
`<1 ppm
`<0.3 ppm
`
`allowing the dispersion to stand. In sophisticated systems vacuum
`de-aeration (or even vacuum mixing) readily eliminates bubbles.
`Another method to control bubble inclusion is to avoid disturbing
`the gel-air interface. A frequent cause of bubbles, especially myri-
`ads of tiny ones, is neutralizer that has become carbonated upon
`exposure to air. When added to the acidic acrylic acid resin dis-
`persion, the CO2 is liberated. To avoid this, store the neutralizing
`agents in closed containers. Here, simply, the centrifugation force
`was exerted to remove the entrapped gas bubbles.
`
`Rheometric measurement. The rheology of aqueous polymer so-
`lutions was characterized using a Parr Physica UDS 200 mechan-
`ical rheometer (Stuttgart, Germany) at room temperature. Cone-
`and-plate geometry was used with a gap size of 0.05 mm, a radius
`of 50 mm, and an angle of 2° for steady state and frequency
`measurements. Another cone-and-plate geometry (0.05-mm gap
`size, 2°, 25-mm radius cone) was utilized to obtain dynamic storage
`(G¢) and loss (G¢¢) moduli as a function of torque. In frequency
`sweep tests, G¢ and G¢¢ were always determined in the regime of
`linear viscoelasticity where the material functions are only func-
`tions of the angular frequency. All measurements to study tem-
`perature effect were carried out fitted with a temperature control
`and a thermostated concentric-cylinder adaptor to avoid possible
`leakage of water by evaporation. Samples showing two relaxation
`behaviors in the stress sweep experiment were selected and heated
`from 25 to 80 °C while the samples with only one relaxation pro-
`cess were cooled down from 25 to 1 °C. The data reported here
`were reproducible within ±15%. Before each dynamic experiment,
`a steady pre-shear was applied at a shear rate of 1 s–1 for 60 s,
`followed by a 120-s rest period. This procedure was necessary to
`erase any previous shear histories on the sample and to ensure that
`the sample establishes its equilibrium structure [30].
`
`Cryogenic scanning electron microscopy (Cryo-SEM). Samples
`were mounted on a brass specimen holder and plunged into liquid
`nitrogen slush at –210 °C. Rapid freezing supercools water and
`elevates its viscosity, thereby delays nucleation and the growth of
`ice crystals, and produces frozen samples with negligible freezing
`artifacts. The frozen specimen was fractured with a remotely con-
`trolled push rod in a Bio-Rad E7450 cryotransfer system (Bio-Rad,
`Hercules, CA) at –120 °C under vacuum. Frozen Poly(acrylic acid)
`polymer/H2O mixture in the specimen was partially sublimed away
`at –60 °C and 2·10–9 bar for 10–90 min to expose the structure
`beneath the vitrified media. The mounted sample was transferred
`into a precooled Balzers MED 010 sputtering device (Balzers Un-
`ion, Balzers, Liechtenstein) against a counter flow of dry nitrogen
`gas. Gold was sputtered on the specimen surface differentially at
`–180 °C for 5 min. The specimen was then transferred to and
`examined by JEOL-JSM 840A (Japan Electron Optics Co., Japan)
`SEM with a modified oil-free evacuation system. All specimens
`were imaged under low acceleration voltage and low specimen
`current conditions, 5 keV, and 0.65·10–11 A, respectively, to reduce
`
`2
`
`

`

`616
`
`the excessive charging and electron beam damaging to the speci-
`mens. The specimen temperature was kept lower than –140 °C
`during the entire imaging.
`
`Results and discussion
`
`Steady-state measurements of polymer gel
`
`Typical shear stress-shear rate data for the aqueous poly
`(acrylic acid) polymer gels of different compositions are
`presented in Fig. 1 for shear rates ranging from 10–3 to
`103 s–1. As shown in Table 2, the data were analyzed by
`using conventional flow equations such as Bingham,
`
`Fig. 1A,B Shear stress as a function of the shear rate for samples:
`A without TEA; B neutralized by TEA at 25 °C
`
`Casson, and Ostwald equations as well as the Herschel-
`Bulkley equation known as the model describing the flow
`behavior of the poly(acrylic acid) polymer very well. The
`yield stress value determination depends on the rheolog-
`ical method and model used [31]. Regression coefficients
`of each equations and parameters in Herschel-Bulkley
`model are shown in Tables 3, and 4, respectively. Ostwald
`and Herschel-Bulkley model were more satisfactory to
`describe the flow behavior while the Bingham and Casson
`model were digressed from the actual flow curves. Yield
`stress of poly(acrylic acid) polymers without TEA de-
`creased with polymer concentrations. However, yield
`stress decreased with TEA content in the concentration
`ranges tested in this study. Yield stress values obtained
`from the equation were substantially low, compared with
`the literature reported previously [24, 32]. The difference
`comes from the fact that the extrapolation of the flow
`curves from _cc  0:1 s1 to zero shear rate were done in the
`literature while entire shear rate ranges were fitted using
`models in this study to obtain the yield stress values. When
`the simple Bingham equation was applied over the entire
`shear rate ranges, yield stress values were well consistent
`with trends reported in the above literature (Table 4). In
`other words, the Bingham model was more suitable in
`terms of yield stress values in spite of the low regression
`coefficient values.
`As shown in the flow curves re-plotting the viscosity
`as a function of the shear rate neutralizing agent, it was
`difficult to determine the zero-shear-rate viscosity since
`there was no clear Newtonian viscosity plateau region in
`the flow curves (Fig. 2). The power law can’t generate
`the theoretical zero-shear-rate viscosity value (Table 2).
`The exponential form of flow model could determine the
`zero-shear-rate viscosity value at _cc ! 0, but the model
`was not reliable according to regression coefficients for
`all samples. The Cross model [33] in order to fit rhe-
`ological behavior of Carbopol 941 (Table 5) should be
`modified in that the g1 and p are assumed to be 0 and 1,
`respectively, in the following equation:
`g g1
`1
`g0 g1
`1 þ K _ccð Þp
`
`where g0 is viscosity at _cc ! 0, g1 is viscosity at _cc ! 1,
`and p is the Cross exponent. g0 of both samples with and
`without TEA increased clearly with polymer concentra-
`tions while it made only a slight difference with TEA
`concentration (Table 6, Fig. 3). Zero-shear-rate viscosity
`of all samples fell in the range of 103 due to the gel net-
`work. K values also showed the similar tendency. In this
`case, K value decreased dramatically with polymer con-
`centrations when neutralizing agent doesn’t exist in the
`solution. Steady-state measurements were not sufficient
`to analyze the overall rheological behavior of aqueous
`polymer gel system, considering a micro-structural point
`of view.
`
`ð1Þ
`
`¼
`
`3
`
`

`

`Table 2 Flow models in shear
`rate-shear stress curve and
`shear rate-viscosity used in
`this study
`
`Model
`
`Bingham
`
`Casson
`
`Ostwald
`
`Herschel-Bulkley
`
`Š
`
`
`Model equation s ¼ s _ccð Þ½
`s ¼ s0 þ g _cc
`
`s0:5 ¼ s0:50 þ g0:5 _cc0:5
`s ¼ K _ccn
`s ¼ s0 þ K _ccn
`
`Parameters
`
`s0
`g
`s0
`g
`K
`n
`s0
`
`K
`n
`
`Model
`
`Exponential Decay
`
`Ostwald
`
`Cross
`
`Š
`
`Model equation g ¼ g _ccð Þ½
`
`ð
`Þ
`g ¼ g0 exp K _cc
`g ¼ K _ccn1
`g ¼ g0
`1þK_
`
`Parameters
`
`g0
`K
`K
`n
`g0
`K
`
`617
`
`Bingham yield stress (Pa)
`Bingham viscosity (PaÆs)
`Casson yield stress (Pa)
`Casson viscosity (PaÆs)
`Consistency (PaÆsn)
`Power law index
`Herschel-Bulkley yield
`stress (Pa)
`Consistency (PaÆsn)
`Power law index
`
`Zero shear viscosity (PaÆs)
`Consistency (s)
`Consistency (PaÆs1–n)
`Power law index
`Zero shear viscosity (PaÆs)
`Consistency (s)
`
`Table 3 Regression coefficient
`(R2) for various flow models
`in shear rate-shear stress curve
`
`Concentration (wt%)
`
`Bingham
`
`Casson
`
`Ostwald
`
`Class
`
`Carbopol
`
`TEA
`
`a
`b
`c
`d
`e
`f
`g
`h
`i
`j
`k
`
`0.1
`1.0
`2.0
`4.0
`0.1
`0.1
`0.1
`1.0
`1.0
`2.0
`2.0
`
`–
`–
`–
`–
`0.1
`1.0
`10.0
`1.0
`10.0
`1.0
`10.0
`
`0.986
`0.962
`0.962
`0.947
`0.957
`0.941
`0.943
`0.915
`0.924
`0.927
`0.886
`
`0.989
`0.958
`0.963
`0.935
`0.956
`0.957
`0.961
`0.935
`0.933
`0.943
`0.906
`
`0.999
`0.995
`0.995
`0.990
`0.999
`0.999
`0.999
`0.999
`0.999
`0.999
`0.999
`
`Herschel-
`Bulkley
`
`0.999
`0.997
`0.997
`0.994
`0.999
`0.999
`0.999
`0.999
`0.999
`0.999
`0.999
`
`Table 4 Parameters of
`Herschel-Bulkley model in
`steady state experiments
`
`Concentration (wt%)
`
`s0 (Pa)
`
`K (PaÆsn)
`
`n
`
`Class
`
`Carbopol
`
`TEA
`
`a
`b
`c
`d
`e
`f
`g
`h
`i
`j
`k
`
`0.1
`1.0
`2.0
`4.0
`0.1
`0.1
`0.1
`1.0
`1.0
`2.0
`2.0
`
`–
`–
`–
`–
`0.1
`1.0
`10.0
`1.0
`10.0
`1.0
`10.0
`
`0.75 (2.47)a
`3.67 (10.51)
`7.72 (22.80)
`16.07 (41.31)
`1.94 (10.16)
`2.23 (11.41)
`2.08 (12.48)
`2.58 (54.83)
`1.59 (65.59)
`6.00 (57.09)
`2.65 (73.77)
`
`0.16
`1.23
`2.53
`5.66
`1.50
`2.12
`2.07
`13.28
`24.60
`11.51
`47.25
`
`0.78
`0.62
`0.58
`0.58
`0.62
`0.59
`0.59
`0.51
`0.48
`0.53
`0.46
`
`a( ) Bingham yield stress
`
`Oscillatory rheological behavior
`
`The dynamic rheology provides a more direct correla-
`tion with microstructure than steady rheology since the
`materials can be examined in their at-rest state without
`causing any disruption of their underlying structures
`
`[33]. The viscoelastic behavior of the polymer solution is
`illustrated in Figs. 4 and 5 using frequency and stress
`sweep test at 25 °C.
`The frequency spectrum for a given system offers a
`signature of the microstructure existing in the system
`[34]. The real images of microstructure were obtained by
`
`4
`
`

`

`618
`
`Fig. 2A,B Viscosity as a function of the shear rate for samples:
`A without TEA; B neutralized by TEA at 25 °C
`
`Table 5 Regression coefficient(R2) for three different flow models
`in shear rate-viscosity curve
`
`Concentration (wt%)
`
`Class
`
`Carbopol
`
`TEA
`
`Exponential
`decay (two
`parameter)
`
`Ostwald Cross
`
`a
`b
`c
`d
`e
`f
`g
`h
`i
`j
`k
`
`0.1
`1.0
`2.0
`4.0
`0.1
`0.1
`0.1
`1.0
`1.0
`2.0
`2.0
`
`–
`–
`–
`–
`
`0.1
`1.0
`10.0
`1.0
`10.0
`1.0
`10.0
`
`0.949
`0.971
`0.989
`0.972
`0.963
`0.965
`0.968
`0.953
`0.952
`0.949
`0.971
`
`0.999
`0.997
`0.991
`0.999
`0.997
`0.995
`0.996
`0.986
`0.989
`0.998
`0.998
`
`0.993
`0.996
`0.999
`0.997
`0.995
`0.992
`0.995
`0.979
`0.981
`0.993
`0.996
`
`Š used to
`Table 6 Parameters of Cross model g ¼ g0= 1 þ K _cc½ ð
`Þ
`
`
`predict the rheological behavior in steady state experiments and
`shear modulus (G¢ at 1 Hz) from frequency sweep measurement
`data
`
`Concentration (wt%)
`
`Class Carbopol TEA
`
`g0
`(Pa Æ s · 103)
`
`K
`(s · 102)
`
`Go
`(Pa)
`
`a
`b
`c
`d
`e
`f
`g
`h
`i
`j
`k
`
`0.1
`1.0
`2.0
`4.0
`0.1
`0.1
`0.1
`1.0
`1.0
`2.0
`2.0
`
`–
`–
`–
`–
`0.1
`1.0
`10.0
`1.0
`10.0
`1.0
`10.0
`
`0.31
`1.18
`1.69
`2.41
`0.11
`0.13
`0.15
`1.19
`1.06
`1.53
`1.95
`
`13.89
`3.57
`1.76
`0.10
`0.52
`0.38
`0.45
`1.92
`1.33
`1.75
`1.22
`
`4.11
`40.00
`74.50
`128.00
`12.60
`15.40
`15.00
`47.10
`50.60
`65.90
`91.00
`
`cryo-SEM and will be discussed in the following section.
`Storage modulus (G¢) and loss modulus (G¢¢) are plotted
`in Fig. 4. G¢ was always larger than G¢¢, i.e., elastic
`component is dominant over viscous component. The
`Carbopol polymer gels without neutralization exhibited
`a frequency-independent elastic modulus G¢
`that
`is
`about one order of magnitude higher than the G¢¢ over
`the entire frequency range (Fig. 4A). The shear modulus
`was determined as the high-frequency limiting G¢ at the
`highest possible frequency to remove the inertia effects
`[35]. This type of dynamic response is a characteristic of
`gel-like materials [36] in good agreement with the pre-
`vious results [10, 37]. The elastic modulus (G¢) of a gel
`system correlates with the rigidity (stiffness) of the net-
`work where G¢ is independent of the frequency. Thus,
`we expect that more rigid structures were formed with
`increasing polymer concentrations as shown in Table 6
`and Figs 3 and 4A. However, when the Carbopol poly-
`mers were neutralized by TEA, both G¢ and G¢¢ were no
`longer independent of the frequency showing polymer-
`like viscoelasticity (Fig. 4B). G¢ and G¢¢ increased lin-
`early with frequency from 10–2 to 10–1 Hz as shown in
`Fig. 4B. G¢ of unneutralized samples was lower than
`that of neutralized ones in low frequency ranges while
`higher in high frequency ranges as expected for samples
`containing the same amount of Carbopol [21]. It has
`also been suggested that a more rigid structure con-
`tributes to this behavior of unneutralized gel systems
`[21]. The constant G¢ at low frequencies means that it is
`sufficient to form the entanglement or cross-links at even
`low polymer concentration (0.1%) with or without
`TEA. Here an interesting feature could be seen at very
`low polymer concentrations of neutralized samples. G¢
`of 0.1% Carbopol 941 neutralized by TEA showed
`almost constant values in low frequency ranges, and
`then increased with frequency while G¢ of the neutral-
`ized samples (1.0 and 2.0% Carbopol) monotonically
`increased. It is possible that monotonic increase in G¢ at
`
`5
`
`

`

`619
`
`Fig. 3 Zero-shear-rate viscosity and shear modulus as a function
`of polymer concentrations at 25 °C for samples containing different
`TEA concentrations: 0% (circles); 1% (squares); 10% (triangles).
`Filled symbols correspond to zero-shear-rate viscosities. Open
`symbols represent shear modulus
`
`the
`higher frequencies means the partial breakage of
`interconnected network, inferred from the existence of
`the plateau region at lower frequencies, which represents
`a true cross-linked polymer gel network. It is, therefore,
`concluded that the gel network is retained at low fre-
`quencies and, on the other hand, destroyed by the more
`frequent changes of the displacement at higher fre-
`quencies due to the formation of too rigid and brittle
`structures in the presence of TEA. Theoretically, plateau
`regions will also appear if the frequency is extended to
`much lower values for the samples of 1.0 and 2.0%
`Carbopol neutralized by TEA.
`The stress sweep experiment was performed to find
`the resistance to deformation at a constant frequency of
`1 Hz. For a very low shear stress, gels without TEA
`seem to behave linearly and G¢¢ is constant but a sub-
`sequent increase in shear stress above a critical value
`causes an increase of G¢¢ before it drops further, showing
`typical characteristics of a Maxwell fluid (Fig. 5A,C). As
`can be observed, increasing polymer concentration re-
`sults in both a shift to higher values of critical shear
`stress and an increase of the G¢¢ maximum. Increase in
`shear stress induces breakage of the initial network
`structure. Rigidity of the microgel structure can be
`clearly identified by the shape of stress vs G¢ curve, while
`it is only qualitatively estimated in frequency vs G¢ curve
`
`c
`Fig. 4A–C Elastic modulus (G¢) and loss modulus (G¢¢) as a
`function of frequency for samples: A without TEA; B neutralized by
`TEA; C complex viscosity at 25 °C
`
`6
`
`

`

`620
`
`as mentioned above. A more detail description will be
`given in the following section.
`Temperature effects on the dynamic rheological be-
`havior of Carbopol have not been widely studied since
`water is evaporated at the elevated temperature and, in
`addition, no distinct trend in the frequency sweep test is
`shown with temperature [21]. However, we found that
`two distinct relaxation phenomena disappeared as the
`temperature increased as shown in Fig. 6. In Contrast, it
`appeared again as the temperature decreased. Effect of
`temperature on the behavior in the stress sweep experi-
`ment was found to be equivalent to that of Carbopol
`and TEA content.
`
`Direct microstructures analysis
`
`Direct cryo-SEM imaging provides a three-dimensional
`topographical contrast of poly(acrylic acid) polymer
`hydrogels consisting of different compositions. Figure 7
`shows the representative images of the samples tested.
`The gel network structures were classified largely into
`two categories, i.e., honeycomb and irregular fibrous
`network structure. Samples showing two distinct relax-
`ation behavior in the stress sweep experiment had
`a honeycomb structure as shown in Fig. 7C,F. This
`structure appeared either by increasing polymer, or by
`neutralizing agent content. At a very low polymer con-
`centration (0.1% Carbopol), very thin fibers were
`formed and entangled physically (Fig. 7A). As the poly-
`mer concentration increased (Fig. 7A–C), cross-linking
`density increased, and subsequently, polymer hydrogel
`formed the honeycomb-like structure possessing a wall.
`Compared with a firm and strong honeycomb network
`structure (Fig. 7F),
`it showed less cross-linked and
`smaller pore volumes. On the other hand, at very low
`polymer concentrations, this fibrous structure couldn’t
`grow to be a wall, though each fiber became thicker and
`was interconnected to form the cross-linking by adding
`TEA as a neutralizing agent (Fig. 7A,D). The sample
`containing 2% Carbopol and 1% TEA also showed
`honeycomb structure. In this case, it showed only one
`relaxation process since the wall
`is so thin that the
`structure is not sufficient to resist the stress (Fig. 7E).
`The wall grew in thickness with TEA contents without
`changing inner pore size. This thin wall structure can be
`identified by the very small kink in shear stress ranges
`around 102 Pa in the stress sweep rheometry (Fig. 5B)
`after the structure was collapsed by the excess defor-
`mation. This reproducible kink for the sample with 2%
`Carbopol and 1% TEA is more prominent than that
`
`c
`Fig. 5A–C Elastic modulus (G¢) and loss modulus (G¢¢) as a
`function of shear stress at a frequency of 1 Hz. for samples:
`A without TEA; B,C neutralized by TEA at 25 °C
`
`7
`
`

`

`621
`
`b
`Fig. 6A–C Elastic modulus (G¢) as a function of temperature at a
`frequency of 1 Hz for samples containing: A 4% Carbopol; B 2%
`Carbopol, 10% TEA; C 2% Carbopol, 1% TEA
`
`with 1% Carbopol and 1% TEA. This very important
`finding is only observable in the stress sweep test. Gen-
`erally, frequency sweep measurements give good infor-
`mation on microstructures of the gel system. However,
`in this study it was found that the frequency sweep test
`was not enough to differentiate the structure changes
`from string-like entangled structure to honeycomb-like
`rigid structure. The frequency sweep test cannot detect
`this transformation since the measurement
`is done
`within linear response ranges, which mean no structure
`breakage during measurement. On the other hand, the
`stress sweep test may induce destructive effect on the
`network system, especially, at higher shear stress region.
`Rigid structures such as honeycomb structure in this
`study can contribute to upward shift of stress sweep
`curve. This wall structure were also changed with tem-
`perature as expected. Figure 8 shows the wall became
`thinner and pore size increased as the temperature in-
`creased from 25 to 80 °C. Temperature also affected the
`rigidity of the structure in the same way with polymer
`concentration and neutralizing agent.
`Among the four fundamental attractive interactions
`in hydrogel
`that act
`to shrink network, hydrogen
`bonding is the dominant force in the unneutralized
`systems, which is easily destroyed by shear [38, 39, 40].
`After neutralization, apparent viscosity increased due
`to uncoiling of the polymer molecules to form rigid
`chains [38, 41] or swollen gel particles [42]. Changes
`in the interaction between polymers and the solvent
`after neutralization are attributed to the forming of
`thick wall above the certain polymer concentration.
`Although the Carbopol gels are composed of either a
`high molecular weight amorphous polymer with long
`side chains or a lightly cross-linked amorphous poly-
`mer or a highly crystalline polymer [43], it has been
`proposed that the elasticity is mainly incurred by the
`side chain entanglements and the possible hydrogen
`bonding while dissipated viscous flow is caused by
`inter-chain movement [21]. It can be proposed, how-
`ever, that the rheological characteristics mainly come
`from the cell-like wall structure, based on the Cryo-
`SEM results. Neutralization of Carbopol gel by TEA
`brought about the equivalent structural change in the
`microstructure of gel network to decreasing temperature.
`
`Conclusion
`
`The aqueous poly(acrylic acid) polymer gel system
`was investigated by static and dynamic rheometry. The
`emphasis in these studies was on the profile of the elastic
`
`8
`
`

`

`622
`
`Fig. 7 Representative cryo-
`SEM images (at 5000· magni-
`fication) of samples at 25 °C
`with: A 0.1% Carbopol; B 2%
`Carbopol; C 4% Carbopol;
`D 0.1% Carbopol, 1% TEA;
`E 2% Carbopol, 1% TEA;
`F 2% Carbopol, 10% TEA.
`Arrows indicate cross-linking
`interconnections thickened by
`TEA. White scale bar represents
`5 lm
`
`Fig. 8A,B Comparison of cryo-
`SEM images (at 2000· magni-
`fication) of samples comprising
`2% Carbopol and 10% TEA:
`A at 25 °C; B at 80 °C. White
`scale bar represents 20 lm
`
`modulus G, which is a measure of gel rigidity. Since the
`frequency sweep measurements are done under the
`conditions showing a linear viscoelastic behavior, which
`means there is no structure breakage, the strength of the
`network given by the honeycomb structure cannot be
`properly analyzed. High shear stress in the stress sweep
`measurement may induce a destruction of the cross-
`linked network depending on the network strength.
`Strength of the gel network resistant to deformation is
`
`related to its typical network structure. Correlation has
`been observed between the microstructure determined
`by cryo-SEM and rheological properties. Cryo-SEM
`may be a valuable tool for the study of polymer hydrogel
`microstructure as a function of polymer and TEA con-
`centration. The microscopic characterization shows ev-
`idence of the appearance of a honeycomb structure of
`polymer hydrogel in the presence of neutralizing agent
`and high concentration of the poly(acrylic acid) poly-
`
`9
`
`

`

`mer, confirming the two relaxation process of swollen
`hydrogel. Thermal energy plays a destructive role in the
`firm and strong honeycomb microstructures. Decreasing
`temperature makes the gel network restore its firmness
`and strength. In the final analysis, increasing polymer
`and neutralizing agent concentration is equivalent to
`decreasing temperature in terms of a network structure.
`
`It is theoretically possible to prepare the hydrogels with
`the same microstructure by changing polymer, TEA
`concentration, and temperature, independently.
`
`Acknowledgment We are grateful to Yeo-Kyeong Yoon, Director
`of LG Household & Personal Care Research Park, for his kind
`permission to publish this work. We also thank Dr. Giyoong Tae
`for helpful comments and discussions.
`
`623
`
`References
`
`1. David LS (1973) Cosmet Perfum
`88(3):51
`2. Bremecker KD, Koch B, Kranse W,
`Neuenorth L (1992) Pharm Ind 54:182
`3. Hariharaputhiran M, Subramanian RS,
`Campbell GA, Chhabra RP (1998) J
`Non-Newtonian Fluid Mech 79(1):87
`4. Sjo¨ berg M, Bergstro¨ m L, Larsson A,
`Sjo¨ stro¨ m E (1999) Colloids Surf 159:197
`5. Gutirrez CA, Moreno R (2000) J Mater
`Sci 35:5867
`6. Choulis N (1975) J Pharm Sci
`64(6):1033
`7. Mortazavi AS, Smart JD (1994)
`J Pharm Pharmacol 46:86
`8. Muramatsu M, Kanada K, Nishida A,
`Ouchi K, Saito N, Yoshida M, Shim-
`oaka A, Ozeki T, Yuasa H, Kanaya Y
`(2000) Int J Pharm 199:77
`9. Ozeki T, Yuasa H, Kanaya Y (2000)
`J Control Release 63:287
`10. Carnali JO, Naser MS (1992) Colloid
`Polym Sci 270:183
`11. Kiser PF, Wilson GW, Needham D
`(1998) Nature 394:459
`12. Antonietti M, Bremser W, Schmidt M
`(1990) Macromolecules 23:3796
`13. Berli CLA, Quemada D (2000)
`Langmuir 16:7968
`14. Clarke MT (1993) In: Laba D (ed)
`Rheological properties of cosmetics and
`toiletries. Marcel Dekker, New York,
`p 55
`
`15. Tamburic S, Craig DQM (1997) Eur J
`Pharm Biopharm 44:159
`16. Hagerstro¨ m H, Paulsson M, Edsman K
`(2000) Eur J Pharm Sci 9:301
`17. Riley RG, Smart JD, Tsibouklis J,
`Dettmar PW, Hampson F, Davis JA,
`Kelly G, Wilber WR (2001) Int J Pharm
`217:87
`18. Cho YI, Hartnett JP, Lee WY (1984)
`J Non-Newtonian Fluid Mech 15:61
`19. U¨ nlu¨ N, Ludwig A, van Ooteghem M,
`Hincal AA (1991) Pharmazie 46:784
`20. Barry BW, Meyer MC (1979) Int J
`Pharm 2:1
`21. Barry BW, Meyer MC (1979) Int J
`Pharm 2:27
`22. Dolz M, Herrez M, Gonza´ lez M, Dı´ ez
`O, Delegido J, Herna´ ndez MJ (1998)
`Pharmazie 53:2
`23. Testa B, Etter J-C (1973) Pharm Acta
`Helv 6(7):378
`24. Atapattu DD, Chhabra RP, Uhlherr
`PHT (1990) J Non-Newtonian Fluid
`Mech 38:31
`25. Bremecker KD (1989) Pharm Ind
`51(2):199
`26. Testa B (1970) Thesis, University of
`Lausanne
`27. Martin EW (1965) Remington’s phar-
`maceutical sciences, 13th edn. MACK
`Publishing, Pennsylvania
`28. Saito S, Taniguchi T (1973) J Am Oil
`Chem Soc 50:276
`
`29. Tamburic S, Craig DQM (1995)
`J Control Rel 37:59
`30. Raghavan SR, Khan SA (1995) J Rheol
`39:1311
`31. Nguyen QD, Boger DV (1985) J Rheol
`29:335
`32. Nguyen QD, Boger DV (1983) J Rheol
`27:32
`33. Cross MM (1965) J Colloid Sci 20:417
`34. Russel WB, Saville DA, Schowalter WR
`(1989) Colloidal dispersions. Cambridge
`University Press, Cambridge
`35. Steeneken PAM (1989) Carbohydr
`Polym 11:23
`36. Imae T (1994) In: Herb CA,
`Prud’homme KR (eds) Structure and
`flow in surfactant solutions. ACS Sym-
`posium Series 578. American Chemical
`Society, Washington, DC, p 140
`37. Ketz RJ Jr, Prud’homme RK, Graessley
`WW (1988) Rheol Acta 27:531
`38. Dittmar CA (1957) Drug Cosmet Ind
`81:447
`39. Lang W (1972) Drug Cosmet Ind 52:54
`40. Lang W (1972) Drug Cosmet Ind 52:127
`41. Fischer WH, Bauer WH, Wiberley SE
`(1961) Trans Soc Rheol 5:221
`42. Taylor NW, Bagley EB (1975) J Polym
`Sci Polym Phys Ed 13:1133
`43. Ferry JD (1970) Viscoelastic properties
`of polymers, 2nd edn. Wiley, New York
`
`10
`
`