J. Phys. Chem. B 2007, 111, 8655-8662
8655
Inducing pH Responsiveness via Ultralow Thiol Content in Polyacrylamide (Micro)Gels with Labile Crosslinks† Michael Bajomo,‡ Joachim H. G. Steinke,*,§ and Alexander Bismarck*,‡ Department of Chemical Engineering, Polymer & Composite Engineering (PaCE) Group, Imperial College London, South Kensington Campus, London SW7 2AZ, United Kingdom, and Department of Chemistry, Imperial College London, South Kensington Campus, London SW7 2AZ, United Kingdom ReceiVed: January 19, 2007; In Final Form: April 20, 2007
Here we present the synthesis and characterization of pH responsive polyacrylamide microgels, synthesized via free radical polymerization of acrylamide and bis (acryloylcystamine) (BAC). The gels were made with ultralow amounts of thiol functional groups incorporated into the polymer. The resulting gel monoliths were mechanically chopped into microgel particles with size distributions ranging from 80 to 200 µm. The gels exhibit an interesting reversible pH-dependent rheological behavior which led to gelling of the colloidal suspension when the pH was increased, and a low-viscosity suspension was obtained when the pH was taken back to the original value. The viscosity of the colloidal system containing MBA crosslinked microgels remained insensitive to pH. This observation motivated further analysis; viscosity measurements of the highly viscous (gel-like) state of the BAC crosslinked microgel colloidal suspension were carried out to further understand the rheological behavior of the colloidal system. Electrophoretic mobility measurements as function of pH of the BAC and MBA crosslinked colloidal polyacrylamide microgel suspensions were performed. The swelling behavior of the microgels for both colloidal systems was also determined as function of pH using static light scattering. This swelling behavior was used to rationalize the observed rheological behavior. The work presented here demonstrates that free thiol groups present within a polymer gel matrix confer pH responsive behavior to the gel in solution. The viscosity of a BAC crosslinked microgel suspension was also measured under reducing conditions. The viscosity of the microgel suspension reduced with time, due to the breakage of the disulfide bonds in the crosslinkers.
Introduction The advancement of hydrogel technology has led to the development of various types of novel crosslinking methods, which result in a wide range of stimuli-responsive behavior.1 In view of environmental and biomedical applications,2 hydrogels can be synthesized to be biodegradable. This property of biodegradability is the main reason why hydrogels have found such wide and extensive use in the biomedical industry.1 However, the use of hydrogels extends beyond this industry. Hydrogels are also used as resins in affinity chromatography and electrophoresis, as well as supports for biocatalyst immobilization.3 There is a great need for water soluble polymer systems that have physical properties that can be influenced by external conditions. The motivation for developing smart crosslinking methods for hydrogels is the possibility of creating materials that have new and additionally useful properties yet maintain the advantages of non-responsive versions of hydrogels, i.e., biocompatibility, swellability, and ductility. A hydrogel with responsive characteristics is often called a “smart” gel, because its properties can be changed through a variety of environmental stimuli (chemical, physical) allowing it to respond to environmental † Part of the special issue “International Symposium on Polyelectrolytes (2006)”. * Author to whom correspondence should be addressed. E-mail:
[email protected]; fax: +44 207 5945578 (A.B.). E-mail
[email protected]; fax: +44 20 75945804 (J.H.G.S.). ‡ Department of Chemical Engineering. § Department of Chemistry.
changes. A smart gel does not necessarily have smart crosslinks. Hydrogels with smart crosslinks can be viewed as a special version of hydrogels. The word ‘smart’, when referring to a crosslink, is used because it captures the ability of the crosslinks to change the physical properties of the hydrogel in a controllable manner. For example, the crosslink might degrade (reversibly) upon encountering a particular stimulus. This stimulus might be a pH or temperature change, the presence of a chemical or a change in pressure. In cases where these crosslinks are not degradable or responsive, i.e., not smart, the hydrogel can still be designed to provide stimuli dependent behavior through the action of functional groups along the polymer backbone or the nature of the polymer backbone itself. Following previous work by Griffin et al.4 on the surfactant free synthesis of microgels using homogenization, we report the characterization of novel pH responsive polyacrylamide microgels in an aqueous medium. pH responsive microgels purely based on polyacrylamide chemistry have not yet been reported in the literature. Pure polyacrylamide gels do not exhibit any pH dependent behavior.5 Previous studies on disulfide crosslinked microgels, that are pH sensitive, have utilized polymers that were pH-sensitive, like poly(acrylic acid)6-7 and poly(methacrylic acid).8 Our use of a pH insensitive backbone creates a good opportunity to study the contribution of the crosslinker on the stimuli-responsive behavior of the microgel separate from main chain responses. Polyacrylamide microgels were synthesized from acrylamide and a crosslinker using free radical polymerization. Two
10.1021/jp070491a CCC: $37.00 © 2007 American Chemical Society Published on Web 06/06/2007
8656 J. Phys. Chem. B, Vol. 111, No. 29, 2007
Bajomo et al.
Figure 1. NMR Spectra of BAC as received.
microgel sets were made; a first set containing the crosslinker N,N-bis(acryloyl)cystamine (BAC) was compared to a control microgel comprising of the same polyacrylamide backbone, this time crosslinked with methylene bisacrylamide (MBA). The choice of methylene bisacrylamide is based on it being the pH unresponsive, ubiquitous workhorse of biomedical research. It is used in protein analysis as the gel medium in polyacrylamide gel electrophoresis (PAGE). Experimental Materials. Acrylamide (98.5%), N,N,N′,N′-tetramethylethylenediamine (99%) and potassium persulphate (99+%), N,Nbis(acryloyl)cystamine (BAC) 98%, and methylene bisacrylamide (99+%), Ellman’s reagent (dithiobis(2-nitrobenzoic acid) (DNTB)), potassium chloride (99%), hydrochloric acid (37%) and dithiothreitol (DTT) (99+%) were purchased from SigmaAldrich. Ethanol (99.7%) was obtained from VWR International. Sodium phosphate dibasic, dodecahydrate was sourced from Fisher Scientific U.K. and sodium hydroxide pellets (98.5%) from Acros Organics. All materials were used without further purification. Analysis of Starting Materials. The assessment of the purity of a fresh batch of BAC crosslinker was first carried out semiquantitatively by 1H NMR analysis (Figure 1). The important resonances for the thiol and disulfide groups are between 2 and 3 ppm. Quantitative Ellman’s tests9-11 confirmed the presence of the reduced form of the disulfide crosslinker as this method is specific to -SH groups (Figure 2). A solution of as purchased crosslinker was combined with DNTB (Ellman’s reagent) which reacts with free thiols to give 2-nitro-5thiobenzoic acid (TNB). The formation of this colored compound was followed by UV-vis spectroscopy with the assay giving a linear response to the amount of -SH groups present. The method followed for Ellman’s titration was that given by Aliyar et al.11 The UV/vis spectrum of the test sample was measured using a Perkin-Elmer Lambda 40 UV/vis spectrometer (Perkin-Elmer, UK). The absorbance at 412 nm was used to calculate the molar concentration of free thiols via the LambertBeer law, taking the molar absorptivity of TNB9 to be 14 000 dm3 mol-1 cm-1. The amount of thiol groups (per crosslinker molecule) determined from Ellman’s titration was 0.15 mol %.
In the same manner, a suspension of microgels was reacted with Ellman’s reagent to determine the amount of free thiols that would be incorporated into the gel structure. About 2 cm3 of a suspension of microgels crosslinked with 10 wt % BAC (see Microgel fabrication) was added to a mixture of 4 mg DNTB and 40 cm3 of 0.1 M sodium phosphate in a 60 cm3 Erlenmeyer flask. The flask was covered with a septum and magnetically stirred under ambient conditions. The reaction was followed using UV-vis spectroscopy. An absorbance reading at 412 nm (the characteristic wavelength for the product) was recorded in 10 min intervals. For each measurement 3 cm3 samples from the reaction was removed with a pipette and poured directly into a cuvette to be analyzed by UV-vis spectroscopy. The result is summarized on Figure 3. The acrylamide starting material used for synthesising the microgels was titrated to determine the presence of any dissociable impurities. This was achieved by dissolving about 0.4 g of the acrylamide in 20 cm3 of deionized water and 0.1 M NaOH was added in 5µL drops using a pipette. The pH was monitored using a Corning pH Meter 240 (Corning, US). Preparation of Hydrogels. The procedure for the synthesis of our microgels is as follows: First a gel block is synthesized via free radical polymerization.10 The reaction is carried out in aqueous ethanol (30% ethanol/70% water (v/v)) by free radical polymerization of 2.8 M acrylamide together with a crosslinker. In the BAC case, 5 wt %/10 wt % (1.36 mol %/2.73 mol % of acrylamide) of this crosslinker was added and allowed to dissolve. The control group was made containing 5 wt % (2.30 mol %)/10 wt % (4.6 mol %) methylene bisacrylamide (MBA) as the crosslinker. Potassium persulphate (K2S2O8) was the initiator and N,N,N′,N′-tetramethylethylenediamine (TEMED) was used as catalyst. The reaction was carried out under ambient aerobic conditions in a beaker. After the reaction was complete (after approx. 60 min), the gel monolith was removed from the beaker and was washed with deionized water thoroughly. The gel monolith was then prepared for homogenization in water to form an aqueous dispersion of microgels. Microgel Fabrication. A 20 cm3 gel block is initially diced with a scalpel until all particles are less than 1 cm3 (by visual inspection). The diced gels were transferred to a beaker and then mixed in exactly 100 cm3 of deionized water contained in
Inducing pH Responsiveness
Figure 2. Structure of the crosslinker and its reduced form.
Figure 3. UV/vis absorbance data from an Ellman’s titration of a BAC crosslinked microgel.
Figure 4. Elastic modulus (G′) against frequency f for polyacrylamide gel slabs crosslinked with BAC and MBA.
the beaker. The blades of a Polytron PT 1600 homogenizer (from Kinematica) were lowered into the beaker, then the homogenizer was switched on and set to 15 000 rpm. The homogenization process was allowed to progress for 15 min. Microgel Swelling Behavior via Static Light Scattering. The swelling behavior of each microgel (both BAC and MBA) was studied via static light scattering in a background electrolyte of 5 mM KCl in water. The average particle diameter of a given microgel (in water) at different pH values was determined. These measurements were carried out using the Mastersizer Model 2000 Hydro (Malvern Ltd., Malvern, U.K.). The device determines particle sizes using laser diffraction, particle sizes are accurately determined by applying Mie theory to a map of scattering intensity versus scattering angle. Measurements cannot be made for single microgel particle but an average particle size can be found for a population of microgel particles if it
J. Phys. Chem. B, Vol. 111, No. 29, 2007 8657 can be assumed that a population of microgels in solution swell in all directions isotropically. The volume weighted average diameter of different microgel suspensions (at a specific pH) was determined as well as the particle size distribution. To make the measurement, the sample dispersion unit was filled with 500 cm3 of the background electrolyte before adding the microgel suspension. About 1 cm3 of the microgel suspension (17% (v/v) microgels in water) was poured into the sample dispersion unit of the Mastersizer 2000 for each measurement. The average particle diameter was calculated with the accompanying software package. Electrokinetic Properties. The electrophoretic mobility measurements were preformed using the Malvern Zetamaster Model ZEM (Malvern, U.K.). The microgels were initially swollen up to their equilibrium sizes in a solution of 5 mM KCl in deionized water at a specific pH under ambient conditions. The volume faction of microgels in the supporting electrolyte was 10%. The microgel suspension was then injected into the measuring cell of the Zetamaster. Rheological Characterization of Microgel Suspensions. The pH dependence of the rheology of the different microgel systems was investigated via measurements using a Paar Physica UDS 200 (Anton Paar, Austria) rotational rheometer. A concentric cylinder geometry was used to quantify the change in viscosity as a function of shear rate for the different microgel suspensions. Samples of the colloidal microgel suspension having an effective volume fraction of 17% were raised (or lowered) to the desired pH using HCl (or NaOH) solutions before placing them in the fixed cylindrical cup of the rheometer. Viscosity measurements were also made for the BAC crosslinked microgels under reducing conditions (100 mM DTT). The measurements were made at 15 h intervals. The fixed cylindrical cup requires 20 cm3 of fluid. All measurements were carried at 25 °C. Oscillatory mechanical measurements were made on thin cylinders of both MBA and BAC crosslinked gels with a TA Instruments AR1000 Rheometer, using the parallel plate geometry to measure the elastic modulus G′. The cylinders were approximately 20 mm in diameter and 3 mm in thickness. The measurements were performed 24 h after the gels were prepared. The elastic modulus was measured over the linear viscoelastic region under isothermal conditions. The frequency was scanned from 0.1 to 100 Hz. The plate diameter was 20 mm. This method is similar to that used by Fernandez et al.12 Under the assumption of a homogeneous network of Gaussian chains, the elastic modulus of the swollen gels can be related to the polymer density (F), the molecular weight of network chains (Mc), and the volume fraction of crosslinked polymer in the swollen gel (ν) through the equation given below:12,13
G′ ) A
F RT(ν0)2/3(ν)1/3 Mc
(1)
where R and T are the universal gas constant and the temperature, ν0 is the volume fraction of crosslinked polymer after preparation, A is the dimensionless front factor which is equal to (1-2/φ) for a phantom network, and φ is the functionality of the network and is equal to 4 for a tetrafunctional network. Accepting that just after preparation ν0 ) ν, so the elastic modulus just after preparation is given by.2 Mc can be obtained by measuring G′:
F G′ ) A RT(ν0) MC
(2)
8658 J. Phys. Chem. B, Vol. 111, No. 29, 2007
Bajomo et al.
Figure 5. Viscosity as a function of shear rate for suspensions of polyacrylamide microgel suspensions (17 vol %) crosslinked with 5 wt % BAC at 25 °C in water.
Figure 6. Viscosity as a function of shear rate for suspensions of polyacrylamide microgel suspensions (17 vol %) crosslinked with 10 wt % BAC at 25 °C in water.
ν0 was calculated from the starting concentration of acrylamide monomers Co (2.8 M) to be
ν0 ) 10-3CoV hR
(3)
where V h R is the average molar volume of polymer repeat units taken to be 52.6 m3 mol-1 (assuming a pure arcylamide polymer network,9 which is a reasonable assumption given that the network comprises of 95-97 mol % acylamide). Results and Discussion Quantifying the Amount of Thiol within the BAC Microgels. The results of the reaction between microgels crosslinked with 10 wt % BAC and DNTB are shown in Figure 3. The maximum absorbance is reached after 60 min; the reaction can be assumed to be complete after this time. The concentration of thiols within the microgel particles was calculated using the Lambert-Beer law. The mole fraction of free thiol groups incorporated into the microgel structure was determined to be 88% of that present in the starting batch of crosslinker. Given the oxidizing environment in which the reactions have been carried out, it is not likely that more free thiol groups have been formed through reduction of the incorporated BAC crosslinker during the polymerization or at the microgel preparation stage. Mechanical Properties and Viscosity of Microgel Suspensions. The elastic modulus of cylindrical slabs of marcogels comprising 5 wt % and 10 wt % of both crosslinkers were carried out over a range of frequencies (0.1 to 100 Hz). The results are presented in Figure 3. The effective network densities can be calculated in terms of the molecular weight of network chains (Mc) using eq 2 and the measured value of G′. Using the value of G′ at 1 Hz (F ) 1.35 g cm-3, ν0 ) 0.15), the calculated Mc for the MBA crosslinked gel is 4181 g mol-1 (at 5 wt %) and 1792 g mol-1 (at 10 wt %) and for BAC Mc is 3136 g mol-1 (5 wt %) and 1394 g mol-1 (10 wt %). This would suggest that the BAC gel has more effective crosslinks than the MBA gel, and on this basis alone the BAC gel should swell less because it has fewer effective crosslinks. However, this is not what is observed. The BAC gel achieves a higher equilibrium swelling size than the MBA gel at neutral pH. There are other factors to consider when the swelling behavior of hydrogels is to be understood, these factors will be discussed later using the Flory-Rehner theory of swelling equilibria. Viscosity measurements of BAC and MBA crosslinked microgel systems were carried out with samples containing 5 and 10 wt % of both crosslinkers over a pH range of 7-10 at low shear rates. The viscosity increases by about 2 orders of
Figure 7. Viscosity as a function of shear rate for suspensions of polyacrylamide microgel suspensions (17 vol %) crosslinked with 5 wt % MBA at 25 °C in water.
magnitude in the case of the BAC-crosslinked microgels (Figures 5-8). The effect is quite different for the MBA crosslinked microgels. The viscosity increases as the pH increases. At high shear rates all the MBA microgels converge to the same plateau viscosity. The BAC microgels are about 3 orders of magnitude more viscous than the MBA crosslinked suspension even at high shear rates. The difference in viscosity would seem counterintuitive if no pH-responsive groups are present. BAC or polyacrylamide if taken notionally, do not possess functional groups that are pH responsive. This indeed is confirmed in the behavior of the MBA system (results are shown in Figures 7 and 8). To systematically investigate the rheology of the microgels, the flow curves were fitted to the power law model
η ) κ‚γm-1
(4)
η 1 ) η0 1 + C‚γn
(5)
and the Cross model
where η is the viscosity of the suspension, ηo the viscosity of the solvent, γ the shear rate, and C and n are constants. In the power law model the index m can be found as the gradient of the plot log(η) against log(γ). The Cross model was fit to the data using Origin 7 (OriginLab Corporation, US). The different constants are summarized in Table 1. The Cross model is known to give a good fit for colloidal suspensions.14,15 The Cross model provided a fit to the MBA crosslinked microgel suspensions while the power law model gave a fit to the flow curves of the BAC crosslinked microgel suspensions. The solid lines in Figures 3-6 are the model fits; the models fit the flow curves
Inducing pH Responsiveness
J. Phys. Chem. B, Vol. 111, No. 29, 2007 8659
Figure 8. Viscosity as a function of shear rate for suspensions of polyacrylamide microgel suspensions (17 vol %) crosslinked with 10 wt % MBA at 25 °C in water.
Figure 10. Electrophoretic mobility µ as a function of pH for polyacrylamide microgels suspensions (10 vol % microgels in water) crosslinked with either BAC or MBA.
Figure 9. Viscosity (at a shear rate of 100 s-1) as a function of pH for polyacrylamide microgels suspensions (17 vol % microgels in water) crosslinked with either BAC (5 and 10 wt %) or MBA (5 and 10 wt %).
Figure 11. ζ-potential for polyacrylamide microgels suspensions (10 vol % microgels in water) crosslinked with either BAC or MBA.
TABLE 1: Parameters Obtained Form the Power Law and Cross Model Fitted to the Viscosity Data of the Different Polyacrylamide Microgel Suspensionsa 10 wt % BAC
5 wt % BAC
10 wt % MBA
5 wt % MBA
pH
κ
m
κ
m
C
n
C
n
7 8 9 10
0.49 12.89 38.23 118.81
0.60 0.30 0.34 0.28
0.46 7.14 16.41 66.00
0.70 0.40 0.30 0.41
14.00 14.00 8.72 6.44
2.53 2.52 2.19 1.44
19.44 11.44 9.86 4.25
1.68 1.61 1.39 1.36
a κ and m (the parameters for the power law model) for the BAC crosslinked microgel suspension; C and n (the parameters for the Cross model) for the MBA crosslinked microgel suspension.
well. In the power law case, a linear plot is obtained with m values between 0.7 and 0.3. An m value of 1 indicates Newtonian behavior while m < 1 indicates shear thinning behavior. The model suggests that at pH 7 the BAC microgel suspensions (BAC 10 wt % and 5 wt %) behave more Newtonian. In other words the microgels have less of an impact on the flow properties of the suspension. At higher pH the value of m decreases significantly suggesting that the microgel suspension becomes more shear thinning, although the suspension is more viscous. The increase in shear thinning behavior suggests that the phenomenon that gives rise to the increase in viscosity becomes less significant at high shear rates. In the Cross model, which was fitted to the data for the MBA microgel suspension, the parameter C is a measure of how quickly, in terms of shear rate, a fluid approaches its plateau viscosity. It is thus a measure of how resistant a fluid is to high shear rates. The value of C halves between pH 7 and 10 for the MBA crosslinked microgel suspension (MBA 10 wt % and 5 wt %) indicating an increasingly shear thinning behavior. Both BAC and MBA crosslinked microgel suspensions are shear thinning,
which is the expected flow behavior for microgel suspensions.12 The viscosity decreases at higher shear rates because the mechanical shear overpowers any interactions between the particles. In order to understand the difference in the viscosities of the BAC and MBA microgel suspensions electrokinetic measurements were performed. Electrophoretic Mobility. Electrophoretic mobility measurements were carried out for both BAC and MBA crosslinked microgels over a pH range of 2-11 (Figures 10 and 11). This covers the pH range where rapid changes in viscosity are observed. The electrophoretic mobility of microgel systems crosslinked with 5 and 10 wt % BAC and MBA are presented in Figure 10. The ζ-potential (assuming the theoretical model of hard spheres) is plotted in Figure 11. The ionic strength was kept constant at 5 mM of KCl. As the pH increases the electrophoretic mobility becomes more negative for both types of microgel systems. A very similar system to the one studied here was presented by Wei et al.16 They measured the ζ-potential of polyacrylamide covered silicon-carbide particles. In their system the isoelectric point (iep) occurs around pH 7 and tails off with increasing pH. This is qualitatively similar to the profile found for our MBA crosslinked microgels. However, all the microgels presented here have an iep, where ζ is zero, of about 4.2, which is rationalized as being caused by residual amounts of acidic species present in the acrylamide monomer used to prepare the microgels. This however cannot be responsible for the observed difference in the viscosity of suspensions containing microgels crosslinked with BAC as the MBA ones an iep of about 4.2 but do not show a proportional viscosity increase. Moreover, the amount and source of acrylamide used to synthesize both BAC and MBA microgels are the same so the same effect should be seen in both cases. The BAC crosslinked microgels on the other hand show a gradual decrease in ζ-potential as the pH is increased, from electrophoretic mobility
8660 J. Phys. Chem. B, Vol. 111, No. 29, 2007
Figure 12. Average particle diameter of polyacrylamide microgels crosslinked with either BAC or MBA as a function of pH.
Figure 13. Particle size distribution of polyacrylamide microgels crosslinked with 5 wt % MBA. The individual particle size distribution curves are offset by 1% to ease visualization.
theory of colloidal particles17-22 this is characteristic of passing a second pKa at a pH of about 8.5 and the formation of negatively charged functional moieties in or on the microgel particle. The ζplateau values of 20 and 50 mV are due to an increased number of dissociated functional groups. The charged groups suspected to be responsible for the observed difference in the BAC microgel case are deprotonated thiol groups (thiolates). This is consistent with electrophoretic mobility becoming more negative as more crosslinker, hence thiol, is incorporated into the microgel. Particle Size Distribution. The average particle diameter at equilibrium for BAC and MBA crosslinked microgels are compared across a pH range of 7-10 in Figure 12. With increasing pH the BAC crosslinked microgels become significantly larger than the MBA crosslinked ones, clearly illustrating their greater relative response to pH change. The MBA crosslinked microgels show almost no change in size as the pH increases. For a quantitative comparison, the microgels crosslinked with 5 wt % MBA have the same average diameter at pH 7 (128 µm) and pH 10 (130 µm), while those crosslinked with 5 wt % BAC increase from about 150 µm at pH 7 to about 210 µm at pH 10. The particle size distribution for the MBA crosslinked microgels (Figures 13 and 14) does not change with pH. For the BAC crosslinked microgels the particle size distribution narrows (Figures 15 and 16), indicating that smaller particles swell more than the larger ones. This reduces the skew in the distribution and results in a more uniform particle size distribution. The pH dependent swelling that is observed for the BAC microgels would support the explanation that the increase in viscosity is due to an increase in the particle size (as a result of microgel swelling). The fact that microgel particle size affects viscosity is established in the literature.23-26 Viscosity is empirically related to the effective volume fraction of the microgel in its solvent. The effective volume fraction of
Bajomo et al.
Figure 14. Particle size distribution of polyacrylamide microgels crosslinked with 10 wt % MBA. The individual particle size distribution curves are offset by 1% to ease visualization.
Figure 15. Particle size distribution of polyacrylamide microgels crosslinked with 5 wt % BAC. The individual particle size distribution curves are offset by 1% to ease visualization.
Figure 16. Particle size distribution of polyacrylamide microgels crosslinked with 10 wt % BAC. The individual particle size distribution curves are offset by 1% to ease visualization.
the microgel in solution increases as the microgel particle size increases. Actually the viscosity is directly proportional to the self-diffusion coefficient of one particle in solution. Hence the viscosity is dependent on the particle size as described by the Stokes-Einstein equation. The Effect of Osmotic Pressure on the Swelling Behavior of Microgels. The factors that govern the swelling behavior of hydrogels are captured by the Flory-Rehner theory of swelling equilibrium. The theory explains that the osmotic pressure π is determined by three factors:
π ) πmix + πel + πion
(6)
where πion, πel, and πmix are the osmotic pressures due to the nonuniform distribution of mobile counterions between the gel and the surrounding liquid, the pressure due to the deformation of the polymer chains, and the pressure due to polymer solvent
Inducing pH Responsiveness
J. Phys. Chem. B, Vol. 111, No. 29, 2007 8661
mixing. All these contributions are given below:
πmix )
RT (ln(1 - ν) + ν + χν2) V1
(7)
where χ is the polymer solvent interaction parameter and V1 is the molar volume of solvent. ν is the volume fraction of crosslinked polymer in the swollen gel:
πel ) -
RTF 1/3 2/3 (ν) (Vo) 2Mc
(8)
fi νRT V1
(9)
πion )
where fi is the effective charge density or the mole fraction of charged units that are effective in swelling the polymer. Moreover, ν can be calculated from the results of swelling by
ν)
( )
D0 3 ν D o
(10)
where Do is the diameter of microgel once it has been made and D is the diameter of the swollen microgel at equilibrium (for our calculations the average diameters were used). Equations 6-9 can be combined and used to calculate the effective charge on the microgels at equilibrium (at equilibrium π ) 0). These values will then be compared to that determined experimentally by a combination of titration and an Ellman’s assay. The amount of charged groups in the MBA microgels was determined from the titration of the acrylamide starting material (assuming that all the charged species present in the starting material are incorporated into the microgel). In addition to this, the microgels crosslinked using BAC were subjected to an Ellman’s assay (see Analysis of starting material). Since the same backbone polymer is used for both the BAC and MBA microgels, it is assumed that the BAC will contain a background charge equal to that determined on the MBA gels, in addition to that determined by the Ellan’s assay. Following the analysis given by Okay et al.,12 the effective charge density can be determined theoretically from the equation below:
ln(1 - ν) + ν + χν2 +
F V (ν)1/3(νo)2/3 - fiν ) 0 2Mc 1
(11)
An alternative charge density fexp can be determined experimentally from eq 12 by taking x to be mole the fraction of charged species present in the microgel as determined by titration (and Ellman’s assay):
fexp )
() V1
x
Figure 17. The theoretical effective charge density of polyacrylamide microgels crosslinked with BAC (5 wt %, 10 wt %) and MBA (5 wt %, 10 wt %) in water.
(12)
VR
The solvent interaction parameter is taken to be independent of the crosslinker used as the network is largely composed of acrylamide. A constant value of 0.48 ( 0.02, which was given by Okay et al.12 for a network of polyacrylamide chains, is used. Taking V1 ) 18 g mol-1 and F ) 1.35 g cm-3 and knowing that Mc has already been determined from elasticity measurements, fexp can be obtained directly. As can be seen in Figure 17, the Flory-Rehner theory shows a significant difference between the effective charge densities on the two different microgels as pH increases. The titration of the MBA mircogels yielded fexp values of 1.36 × 10-5 (5 wt %) and 2.56 × 10-5 (10 wt %), which is a fraction of that predicted by the Flory-
Figure 18. Viscosity as function shear rate for a suspension of polyacrylamide microgel suspensions (17 vol %) crosslinked with 10 wt % BAC at 25 °C over a period of 75 h.
Rehner model. However, the model does reflect the pH independence of the MBA microgel swelling. In the case of BAC, it is expected that the highest charge density is achieved when all thiol groups in the polymer microgel are deprotonated, this would be achieved at the highest pH. Hence comparisons between the theoretical and experimental values can only be made at pH 10. The combination of the titration and Ellman’s assay yielded fexp values of 4.75 × 10-4 (5 wt %) and 1.06 × 10-3 (10 wt %) but the Flory-Rehner model gives a much higher charge density for this case as well. The fact that the Flory-Rehner model overestimates the effective charge, effectively suggesting that a far greater charge density is required to achieve the observed swelling, might expose a limitation of the Flory-Rehner model when treating swollen highly crosslinked polyelectrolyte systems. Nonetheless, these results show that a difference in charge density does exist between the two microgels and can explain their different swelling behavior. Also, one has to remember that a constant polymer-solvent parameter χ was used in the Flory-Rehner model. While it is reasonable to assume that χ is independent of the crosslinker used (as the matrix is largely composed of acrylamide), it might not hold strictly under changing pH. The effective charge densities (experimental and theoretical) on the BAC microgels are at least thirty times higher than on the MBA microgels but the magnitude of the difference begs the question: how large an effective charge density would be required to achieve pHdependent swelling. What has been shown in this work is that the presence of a relatively low amount charged groups (in this case thiols) is sufficient to cause pH-dependent swelling, it would be worth knowing how low a concentration would suffice.
8662 J. Phys. Chem. B, Vol. 111, No. 29, 2007 Viscosity of BAC Crosslinked Microgel Suspensions under Reducing Conditions. The viscosity of a BAC crosslinked microgel suspension (10 wt % BAC, 17 vol % microgel in water) was measured under reducing conditions (100 mM DTT) (see Figure 18). It was observed that the viscosity decreases with time, which agrees with the reduction of the disulfide bond in the crosslinker to thiols, resulting in the breaking down of the microgels. As the disulfide bonds break the flow behavior of the suspension approaches that of a solution of linear polymers. After 15 h the viscosity of the suspension at low shear rates is observably different but it still has the same general profile at the start of the experiment. It is believed that the microgels are not fully degraded after 15 h and are mostly in a swollen state. At low shear rates, the microgels still have an effect on the viscosity of the suspension but as the shear rate is increased the mechanical force deforms the gels and viscosity decreases. As time proceeds the microgels become more degraded and viscosity falls faster as shear rate is increased. After 75 h the microgels are assumed to be fully degraded. Conclusion We have demonstrated that the presence of low amounts of thiol functional groups confer unique pH dependent rheological properties to a suspension of polyacrylamide microgels crosslinked with BAC. The motivation for this work was to study the sharp variation in viscosity that was observed when the pH of this specially synthesized polyacrylamide microgel solution was changed. The viscosity increased by about 3 orders of magnitude above a control suspension containing microgels crosslinked with MBA. By comparing a BAC crosslinked system with one crosslinked with MBA, it has been shown that the observed pH responsive properties are not arising from the polymer main chain but are associated with the reduced thiol form of the crosslinker BAC. For the rheological properties of different microgel suspensions we have shown that the increase in viscosity of the BAC crosslinked microgels is due to an increase in particle solvent uptake with increasing pH. Systems like these provide a means of creating systems that with labile bonds that are also responsive. We are now investigating the redox trigger behavior of the gels crosslinked with BAC by varying the crosslink density and testing the reversibility of the physical properties of the gels.
Bajomo et al. Acknowledgment. The authors would like to thank the Faraday Plastics Partnership (U.K.) and Halliburton Energy Services (Duncan, OK, U.S.A.) for funding and Dr. Ian Robb (Halliburton Energy Services) for his most valuable comments. References and Notes (1) Hennink, W. E.; van Nostrum, C. F. AdV. Drug DeliVery ReV. 2002, 54 (1), 13-36. (2) Rosiak, J. M.; Yoshii, F. Nucl. Instrum. Methods 1999, B 151, 5664. (3) Cong, L.; Dissing, U.; Kaul, R.; Mattiasson, B. J. Biotechnol. 1995, 42, 75-84. (4) Griffin, J. M.; Robb, I.; Bismarck, A. J. Appl. Polym. Sci. 2007, 104, 1912-1919. (5) Yezek, L. P.; van Leeuwen, H. P. J. Colloid Interface Sci. 2004, 278, 243-350. (6) Bromberg, L.; Temechenko, M.; Alakov, V.; Hatton, T. A. Langmuir 2005, 21, 1590-1598. (7) Chiu, H. C.; Wang, C. H. Polym. J. 2000, 32, 574-582. (8) Hiratani, H.; Alvarez-Lorenzo, C.; Chuang, J.; Grosberg, A. Y.; Tanaka, T. Langmuir 2001, 17, 4431-4436. (9) Ellman, G. L. Arch. Biochem. Biophys. 1959, 82, 70-77. (10) Riddles, P. W.; Blakeley, R. L.; Zerner, B. Anal. Biochem. 1979, 94, 75-81. (11) Aliyar, H. A.; Remesen, E. E.; Hamilton, P. D.; Ravi, N. J. Bioact. Compat. Pol. 2005, 20, 169-181. (12) Okay, O.; Durmaz, S. Polymer 2002, 43, 1215-1221. (13) Fernandez, E.; Lopez, D.; Lopez-Cabarcos, E.; Mijangos, C. Polymer 2005, 46, 2211-2215. (14) Tan, B. H.; Tam, K. C.; Lam, Y. C.; Tan, C. B. J. Rheol. 2004, 48 (4), 915-926. (15) Senff, H.; Richtering, W. J. Chem. Phys. 1999, 111 (4), 17051711. (16) Wei, M.; Zhang, G.; Wu, Q. Ceram. Int. 2004, 30, 125-131. (17) Fernandez-Nieves, A.; Fernandez-Barberot, A.; de las Nieves, F. J.; Vincent, B. J. Phys. 2000, 12, 3605-3614. (18) Ohshima, H. Coll. Surf., A 1995, 103, 249-255. (19) Hermans, J. J.; Fujita, H. Koninkl. Ned. Akad. Wetenschap. Proc. 1995, B58, 182. (20) Cohen, E. G. D.; Verberg, R.; de Shepper, I. M. Phys. A 1998, 251, 251-265. (21) Duval, J. F.; Ohshima, H. Langmuir 2006, 22, 3533-3546. (22) Ohshima, H. AdV. Colloid Interface Sci. 1995, 62, 189-235. (23) Mewis, J.; Frith, W. J.; Strivens, T. A.; Russel, W. B. AIChE J. 1989, 35, 415-422. (24) Meeker, S. P.; Poon, W. C. K.; Pusey, P. N. Phys. ReV. E 1997, 55, 5718-5722. (25) Ballauff, M. Macromol. Chem. Phys. 2003, 204, 220-234. (26) Horn, F. M.; Richtering, W.; Bergenholtz, J.; Willenbacher, N.; Wagner, N. J. J. Colloid Interface Sci. 2000, 225, 166-178.
