pH-Sensitive Holographic Sensors - Analytical Chemistry (ACS


pH-Sensitive Holographic Sensors - Analytical Chemistry (ACS...

2 downloads 95 Views 292KB Size

Anal. Chem. 2003, 75, 4423-4431

pH-Sensitive Holographic Sensors Alexander J. Marshall, Jeff Blyth, Colin A. B. Davidson, and Christopher R. Lowe*

Institute of Biotechnology, University of Cambridge, Tennis Court Road, Cambridge, CB2 1QT, U.K. Holographic sensors for monitoring H+ (pH) have been fabricated from ionizable monomers incorporated into thin, polymeric, hydrogel films which were transformed into volume holograms using a diffusion method coupled with holographic recording, using a frequency doubled Nd:YAG laser (532 nm). Unlike other optical pH sensors, it is possible to tailor the operational replay wavelength of the holographic sensor by careful control of the exposure conditions. The holographic diffraction wavelength (color) of the holograms was used to characterize their shrinkage and swelling behavior as a function of pH in various media. The effects of hydrogel composition, ionic strength, temperature, and factors influencing reversibility and response time are evaluated. Optimized holographic pH sensors show milli-pH resolution. The pH-sensing range of the holograms can be controlled through variation of the nature of the ionizable co-monomer used in polymer film construction; a series of holographic sensors displaying visually perceptible, fully reversible color changes over different pH ranges are demonstrated. A poly(hydroxyethyl methacrylate-co-methacrylic acid) holographic sensor was shown to be able to quantify the change in H+ concentrations in real time in a sample of milk undergoing homolactic fermentation in the presence of Lactobacillus casei. The measurement of pH is almost universal in the environmental, consumer, security, agricultural, food, beverage, cell culture, microbiology, biotechnology, and biomedical industries. It is conventionally achieved by the use of pH-sensitive papers, pH electrodes, or ISFETs,1 but there are many examples of devices based on other spectrophotometric, fluorimetric, or luminometric principles. An alternative approach that exploits a change in diffraction rather than absorption and which generates a visual color change has been reported.2,3 The system comprises a crystalline colloidal array of polymer spheres (∼100 nm diameter) polymerized within a hydrogel that swells or shrinks reversibly in the presence of appropriate analytes.4 The crystalline colloidal array diffracts light at visible wavelengths governed approximately by Bragg’s law3 and determined by the lattice spacing, * To whom correspondence should be addressed. Phone: (+44) 1223 334160. Fax: (+44) 1223 334162. E-mail: [email protected]. (1) Janata, J.; Huber, R. J. Ion-Sel. Electrode Rev. 1979, 1, 31-78. (2) Asher, S. A.; Flaugh, P. L.; Washinger, G. Spectroscopy 1986, 1, 26-31. (3) Rundquist, P. A.; Photinos, P.; Jagannathan, S.; Asher, S. A. J. Chem. Phys. 1989, 91, 4932-4941. (4) Holtz, J. H.; Asher, S. A. Nature 1997, 389, 829-834. Holtz, J. H.; Holtz, J. S. H.; Munro, C. H.; Asher, S. A. Anal. Chem. 1998, 70, 780-791. 10.1021/ac020730k CCC: $25.00 Published on Web 08/01/2003

© 2003 American Chemical Society

mλ ) 2nd sin θ where m is the diffraction order, λ is the wavelength of light in vacuo, n is the average refractive index of the system, d is the spacing of the diffracting plane, and θ is the glancing angle between the incident light propagation direction and the diffracting planes. The crystalline colloidal arrays were fabricated by dissolving nonionic polymerizable monomers within the suspension and photopolymerizing into a hydrogel to entrap the colloid lattice.4 A change of 0.5% in the hydrogel volume shifts the diffraction wavelength by ∼1 nm. This colloidal array has been made sensitive to Pb2+, Ba2+, and K+ by co-polymerizing 4-acryloylaminobenzo-18-crown-6 into the hydrogel,5 and more recently, the technique has been refined for Pb2+, pH, and temperature.6 While these crystalline colloidal arrays represent an attractive approach for the fabrication of visually readable sensors, we have proposed an alternative and more generic technique that exploits the concept of a simple reflection hologram as the interactive element in a truly mass-producible biochemical sensor.7-13 In this approach, the holographic element per se provides not only the analyte-specific “smart” polymer matrix, but also the optical interrogation and reporting transducer. When holographic diffraction gratings are illuminated by white light, they act as sensitive wavelength filters. Conventionally, the gratings comprise a gelatin-silver halide photographic emulsion coated onto glass or plastic substrates and are fabricated by passing a single collimated laser beam through a holographic plate backed by a mirror. Interference between the incident and reflected beams, followed by development, fixing, and bleaching creates a modulated refractive index in the form of fringes lying in planes parallel to the gelatin surface and approximately onehalf of a wavelength apart within the 10-µm thickness of the gelatin film. Under white light illumination, the developed grating acts (5) Asher, S. A.; Holtz, J.; Liu, L.; Wu, Z. J. Am. Chem. Soc. 1994, 116, 49974998. (6) Reese, C. E.; Baltusavich, M. E.; Keim, J. P.; Asher, S. A. Anal. Chem. 2001, 73, 5038-5042. Lee, K.; Asher, S. A. J. Am. Chem. Soc. 2000, 122, 95349537. (7) Lowe, C. R. Curr. Opin. Chem. Biol. 1999, 3, 106-111. (8) Blyth, J.; Millington, R. B.; Mayes, A. G.; Frears, E. R.; Lowe, C. R. Anal. Chem. 1996, 68, 1089-1094. (9) Millington, R. B.; Mayes, A. G.; Blyth, J.; Lowe, C. R. Sens. Actuators 1996, B33, 55-59. (10) Millington, R. B.; Mayes, A. G.; Blyth, J.; Lowe, C. R. Anal. Chem. 1995, 67, 4229-4233. (11) Mayes, A. G.; Blyth, J.; Millington, R. B.; Lowe, C. R. J. Mol. Recognit. 1998, 11, 168-174. (12) Blyth, J.; Millington, R. B.; Mayes, A. G.; Lowe, C. R. Imaging Sci. J. 1999, 47, 87-91. (13) Mayes, A. G.; Blyth, J.; Kyro ¨la¨inen-Reay, M.; Millington, R. B.; Lowe, C. R. Anal. Chem. 1999, 71, 3390-3396.

Analytical Chemistry, Vol. 75, No. 17, September 1, 2003 4423

as a reflector of the light for a specific narrow band of wavelengths and holographically recreates the monochromatic image of the original mirror used in its construction. The constructive interference between partial reflections from each fringe plane gives a characteristic spectral peak with a wavelength governed by the Bragg equation (λmax ) 2nd cos θ), where d is the fringe separation distance, n is the average refractive index, and θ is the angle of illumination to the normal. The peak reflectivity is dependent on the number of fringe planes and the modulation depth of the refractive index. Any physical, chemical or biological mechanism that changes the spacing of the fringes (d) or the average refractive index (n) will generate observable changes in the wavelength (color) or intensity (brightness) of the reflection hologram. For example, if a holographic grating is immersed in a test sample comprising a “wet” hydrophobic organic solvent, absorption of water by the hydrogel layer causes the grating to swell perpendicular to the plane of the polymer layer. This increases the fringe separation and causes longer wavelengths to be selected for reflection from the holographic mirror.8 Further work has shown that holograms can be fabricated in a very wide range of “smart” polymer matrixes, including natural, synthetic, or rationally designed systems containing appropriate receptor systems, to create inexpensive optical sensors that respond to a range of putative analytes.7-13 More recently, the technology has been adapted for monovalent cations using crown ether-substituted “smart” polymers,14 and we now report a further development of the technology in which pH-sensitive hydrogels have been used as the basis for pH-sensitive holograms. Hydrogels are infinite three-dimensional polymeric networks which exhibit the property of imbibing and maintaining water at >20% (w/w) within their structure without dissolution. Hydration and, hence, swelling or contraction of these materials is encouraged by the hydrophilic nature of the polymer chains but is limited by the degree of cross-linking of the network.15 Hydrogels containing an acidic or basic pendant functionality undergo changes in hydration as a function of pH of the bathing medium.16 Ionization of this functional group is governed by its dissociation constant (pKa) and leads to changes in hydration of the polymer network as a result of electrostatic and osmotic forces that draw in or expel counterions and water into or out of the gel phase.17 The increase in swelling of the hydrogel leads to a concomitant increase in its electrical conductivity.18-21 A pH-responsive crosslinked copolymer of 2-hydroxyethyl methacrylate (HEMA) and N,N′-dimethylaminoethyl methacrylate (DMAEM) has been exploited as the basis of a microfabricated pH sensor in which an interdigitated electrode array was used to detect the change in electrical conductivity of a photolithographically patterned hydro(14) Mayes, A. G.; Blyth, J.; Millington, R. B.; Lowe, C. R. Anal. Chem. 2002, 74, 3649-3657. (15) Flory, P. J. Principles of Polymer Chemistry; Cornell University Press: Ithaca, NY, 1953. (16) Kopecek, J.; Vacik, J.; Lim, D. J. Polym. Sci. 1971, 9, 2801-2815. (17) Vacik, J.; Shataeva, L. K.; Chernova, I. A.; Samsonov, G. V.; Schauer, J.; Kalal, J. Collect. Czech. Chem. Commun. 1983, 48, 3071-3078. (18) Lesho, M. J.; Sheppard, N. F. Polym. Gels Networks 1997, 5, 503-523. (19) Sheppard, N. F.; Lesho, M. J. J. Biomater. Sci. Polym. Ed. 1997, 8, 349362. Sheppard, N. F.; Lesho, M. J.; McNally, P.; Francomacaro, A. S. Sens. Actuators, B 1995, 28, 95-102. (20) Vacik, J.; Kopecek, J. J. Appl. Polym. Sci. 1975, 19, 3029-3044. (21) Beynon, R. J.; Easterby, J. S. In: Buffer Solutions: The Basics; BIOS Scientific Publishers: Oxford, 1996.

4424

Analytical Chemistry, Vol. 75, No. 17, September 1, 2003

gel layer.19,20 In this report, we now disclose a further development of our holographic sensing technology in which sensors are fabricated with “smart” polymers substituted with acidic and basic functional groups. The potential of these sensors for the measurement of pH changes over defined ranges in real time is demonstrated. EXPERIMENTAL SECTION Materials. All chemicals were of analytical grade unless otherwise stated. 1,1′-Diethyl-2,2′-cyanine iodide (photosensitizing dye), 1-vinylimidazole, 2-(dimethylaminoethyl) methacrylate (DMAEM), 2-dimethoxy-2-phenyl acetophenone (DMPA), 2-hydroxyethyl methacrylate (HEMA) (ultrapure 99+%), 3-(trimethoxysilyl)propyl methacrylate, ethylene glycol dimethacrylate (EDMA), hydrochloric acid, hydroquinone, lithium bromide, methacrylamide, methacrylic acid, potassium bromide, silver nitrate (1 M, volumetric standard), and silver perchlorate were purchased from Aldrich Chemical Co., The Old Brick Yard, Gillingham, Dorset, U.K. 3-(N-Morpholino) propanesulfonic acid (MOPS), acrylamide (electrophoresis grade), sodium hydroxide, and N,N′-methylenebis(acrylamide) (MBA) (electrophoresis grade) were purchased from Sigma Chemical Co., Fancy Road, Poole, Dorest, U.K. 2-(NMorpholino) ethanesulfonic acid (hydrate) (MES), 3-[tris(hydroxymethyl)methyl] amino]ethanesulfonic acid (TES), 4-methylaminophenol sulfate (Metol), chloroactetate, and sodium chloride were purchased from Acros Organics, Janssens, Pharmaceuticalaan, 3A, 2440, Geel, Belgium. Sodium thiosulfate (hypo), acetic acid (glacial), propan-1-ol, propan-2-ol, ethanol, and methanol were purchased from Fisher Scientific Ltd, Bishop Meadow Road, Loughborough, Leicestershire, LE11 5RG, U.K. Acrylic acid (AA) was purchased from Fluka, Industriestrasse 25, CH-9471 Buchs SG, Switzerland. 2-(Trifluoromethyl) propenoic acid (TFMPA) was purchased from Fluorochem Ltd., Wesley Street, Old Glossop, Derbyshire SK13 7RY. Sodium carbonate was purchased from BDH (Merck) Ltd., Poole Dorset, U.K., and Bicine was purchased from Avocado (Research Chemicals Ltd.), Shore Road, Heysham, Lancs., U.K. Equipment. Microscope slides (Super Premium, 1-1.2 mm thick, low iron) were purchased from BDH (Merck) Ltd. Aluminized, 100-µm polyester film (grade MET401) was purchased from HiFi Industrial Film Ltd., Stevenage, U.K. A UV Exposure Unit (∼350 nm, model no. 555-279) was purchased from RS Components, Birchington Road, Weldon Industrial Estate, Corby, Northants, NN17 9RS. A frequency-doubled Nd:YAG laser (350 mJ, 532 nm, Brilliant B, Quantel) was used in hologram construction. Instrumentation. Holograms were analyzed using an LOTORIEL MS127i model 77480 imaging spectrograph in single channel mode with a 256 × 1024-pixel InstaSpec IV CCD detector and processing software. pH Meter. A bench pH meter (HI-213 pH/mV/°C was purchased from Jencons-PLS, Cherrycourt Way Ind. Est., Stanbridge Road, Leighton Buzzard, Bedfordshire, U.K. Synthesis of Polymer Films. The required volumes of monomer solutions to give a prepolymer solution with the correct molar ratio of these monomers were mixed. Monomer mixtures were then diluted with an equal volume of propan-2-ol in the case of polymer systems based on hydroxyethyl methacrylate, or 2:1 (water/solid, w/w) with deionized water in the case of systems

based on acrylamide/methacrylamide. The free radical initiator 2-dimethoxy-2-phenyl acetophenone (DMPA) was added to a final concentration 1% (w/v), and the mixture was vortexed briefly. A 70-µL droplet of monomer solution was pipetted onto the polyester side of an aluminized polyester sheet resting on a clean flat surface. A glass microscope slide, presubbed with methacryloxypropyltriethoxysilane as described previously11 was then gently lowered, silane-treated side down, onto the solution, and any trapped air bubbles were gently squeezed out. The films were polymerized by UV-initiated free radical reaction at room temperature under UV light for 30 min. Polymerized films were carefully peeled off the aluminized polyester while submerged in deionized water. To remove any unpolymerized material and byproducts of the polymerization reaction, films were washed several times in ethanol/water 1:1 (v/v). Prior to hologram construction, the edges of each film were cleaned with a scalpel blade to remove any excess polymer material and give a smooth, flat polymer surface. Hologram Construction. It was necessary to use slightly different protocols to construct holograms in different copolymer systems. All photosensitization protocols were based on the principle of the diffusion method as outlined previously.12 All photosensitization, exposure, and development work was carried out under red safe lighting. Photosensitization of Hydroxyethyl Methacrylate-Based Copolymer Films. Under red safe lighting, 200 µL of silver perchlorate solution (0.3 M in propan-1-ol/water 4:1 (v/v)) was pipetted as an elongated blob onto a clean glass sheet, and the preformed polymer film was lowered onto it, polymer side down. This was left for 3 min to allow the solution to soak into the polymer before excess surface liquid was removed using a soft rubber squeegee. The film was dried briefly under a tepid air current and then submerged, polymer side uppermost, in a continuously agitated bath of 3% (w/v) lithium bromide solution (40 mL, in methanol/water 3:1 (v/v)), cyanine dye (1 mL, 0.1% (w/v) 1,1′-diethyl-2,2′-cyanine iodide in methanol) and sensitizer (5 mg/L ascorbic acid) for 2 min. Films were removed to a continuous stream of water and were thoroughly washed before being dried and stored polymer side uppermost in a light tight box. The construction of holograms in copolymers of poly-HEMA containing basic co-monomers (i.e., DMAEM and VI) was slightly different from that outlined above. Slides were treated for 1 min with an acidified silver perchlorate solution (0.3 M in propan-1ol/water 4:1 (v/v) with 4% (v/v) acetic acid; 200 µL) and immediately after drying were immersed in the LiBr bath for 1 min. Prior to exposure, the photosensitive films were immersed in a sensitizer bath (ascorbic acid, 2% (w/v) titrated to pH 5.5 with NaOH) for 2 min. This was necessary to prevent chemical fogging22 of the hologram caused by the presence of basic groups within these films. Photosensitization of Acrylamide-Based Copolymer Films. Under red safety lighting, 300 µL of silver nitrate (0.3 M in water) solution was pipetted as an elongated blob onto a clean glass sheet, and the preformed polymer film was lowered onto it, polymer side down. This was left for 5 min to allow the solution to soak into the polymer before excess surface liquid was removed using a (22) Saxby, G. In: Manual of Practical Holography; Butterworth-Heinemann Ltd.: Oxford, 1991.

soft rubber squeegee. The film was dried briefly under a tepid air current and submerged, polymer side uppermost, in a continuously agitated bath of 4% (w/v) potassium bromide solution (40 mL, in methanol/water 1:1 (v/v)), photosensitizing cyanine dye (1 mL, 0.1% (w/v) 1,1′-diethyl-2,2′-cyanine iodide in methanol) and sensitizer (ascorbic acid 5 mg/L) for 1 min. Films were removed to a continuous stream of water and washed thoroughly before being dried and stored polymer side up in a light-tight box. Exposure and Development. Photosensitive films were placed polymer side downward in the exposure bath. This consisted of a plastic trough containing a front surface mirror with a spacer positioned at one end to hold the slide at an angle of approximately 4° in relation to the surface of the mirror. This small displacement from the horizontal position prevents the finished hologram from diffracting the incident light at exactly the same angle as specularly reflected light. To ensure that the holograms replayed over the visible spectrum films containing acidic groups (i.e., MAA, TFMPA, and AA) were exposed in 2% (w/v) ascorbic acid, while those containing basic groups (i.e., VI and DMAEM) were exposed in 3 M sodium nitrate. Films were exposed to a single 10-ns pulse from the Nd:YAG laser, except those constructed with basic monomers which were given four separate pulses of the same length. After exposure, the films were washed in deionized water and immersed, polymer side up, in developer solution. For films constructed in copolymers of hydroxyethyl methacrylate, this was an equal volume of 5% (w/v) sodium hydroxide in 50/50 methanol (v/v) and 2% (w/v) hydroquinone in 50/50 methanol. In the case of films constructed in copolymers of acrylamide, 40 mL of ascorbic acid 20 g/L, 4-methylaminophenol sulfate (Metol) (5 g/L), sodium carbonate (20 g/L), and sodium hydroxide (6.5 g/L) was used.22 The time spent in the developer varied according to the individual copolymer/development system, but was typically around 20 s. In all cases, films were removed to a stream of water before immersion into a bath of stop solution (5% (w/v) acetic acid) for ∼1 min to stop development. Excess dye and undeveloped silver bromide were removed by immersion in agitated 10% (w/v) sodium thiosulfate for 4 min, whence slides were rinsed under running tap water and then in methanol for a further 5 min to remove all traces of the cyanine dye. Completed holographic gratings were rinsed in deionized water and air-dried before being stored in sealed polythene bags until use. Hologram Interrogation and Testing. Holographic devices were interrogated using an in-house-built reflection spectrophotometer, as described previously.11,13 Strips of hologram, ∼8-9 mm wide, were placed, polymer side facing inward, into 4-mL cuvettes to which a test solution was added and kept at a constant temperature with a water bath and stirred at a constant rate with a microflea/stirrer arrangement. Holographic pH sensors were tested using the following pH buffer systems: chloroacetate (pH 2.5-3.75), acetate (pH 4-5.5), MES (pH 5.75-6.75), MOPS (pH 7-8), and bicine (pH 8.259.5). The concentration of buffer component varied according to the individual investigation, and the ionic strength was fixed to the required level with sodium chloride. The apparent pKa values of the holographic sensor pH-response curves were determined from the point of inflection of curves fitted using SigmaPlot Analytical Chemistry, Vol. 75, No. 17, September 1, 2003

4425

(version 8; SPSS Science, Woking, Surrey, U.K.). The pKa of ionizable monomers was determined using standard titration methods.21 Isolation of Lactobacillus casei and Fermentation of Whole Milk. Actimel fermented milk drink (Danone) was purchased from a local supplier. The active strain in the drink (L. casei immunitas) was isolated by streaking on MRS agar (Oxoid) and incubating at 37 °C for 48 h. Colonies were picked at this time and subcultured on MRS agar. Cells of L. casei from an overnight 37 °C culture in MRS broth (Oxoid) were centrifuged at 10000g for 15 min, resuspended in water, and recentrifuged before being resuspended in 20% (v/v) glycerol and frozen at -20 °C in identical 0.5-mL aliquots. The number of cells in an aliquot was determined by a viable count. Fresh, pasteurized, full-fat milk was frozen at -20 °C in 15-mL aliquots. Prior to use, aliquots were defrosted and warmed to 37 °C, whence 104 cells mL-1 were added, and the sample was cultured statically at 37 °C for 17 h. This culture (100 µL) was added to 2 mL of freshly defrosted milk, warmed to 37 °C in a cuvette containing a poly-HEMA-co-MAA (4 mol %) hologram, which had previously been washed with whole milk until a stable baseline λmax was obtained. The pH of the milk solution was monitored at regular intervals at 37 °C for 200 min using a small-volume (needle tipped) micro combination pH electrode and the holographic pH sensor. RESULTS AND DISCUSSION pH-Sensitive Hologram Fabrication. The fabrication protocol described herein involves coating a thin layer (∼10 µm thick) of unsensitized polymer film on a silanized glass slide and then immersing the slide sequentially in solutions of silver perchlorate and lithium bromide containing a photosensitizing dye, 1,1′-diethyl2,2′-cyanine iodide. Following exposure of the film to laser light14 and a conventional photographic development step, an interference fringe pattern comprising ultrafine (8 is approximately linear as a function of the mol % ionizable monomer up to 3 mol %, but then shows signs of saturation, presumably because the matrix backbone becomes more hydrophilic as the mol % of MAA increases. It is known that the use of copolymers of methacrylic acid, butyl methacrylate and hydroxyethyl methacrylate influence the structure of the polymer backbone and, thus, the swelling as a function of pH, with polymers containing more hydrophilic backbones having a lowered pH sensitivity.29 The swellability as a function of pH of MAA-containing poly-HEMA hydrogels appears to be dependent on a complex relationship between the number of ionizable functionalities introduced into the polymer backbone and the hydromechanical properties of the resulting gels.30 Nevertheless, changes in the replay color of the poly-co-MAA-HEMA holograms are readily visible to the naked eye. Figure 2c shows the replay color of a poly-HEMA hologram containing 4 mol % MAA as a function of pH over the range 3-9, with the most obvious visual change occurring as the pH is increased from 4 to 7. Similar pH responses within the range 3-9 were obtained with holographic sensors constructed in copolymers of acrylamide and methacrylamide (2:1 mol %) cross-linked with 5 mol % N,N′methylene-bis(acrylamide) and containing increasing amounts of acrylic acid (AA) within the range 0-12 mol % (Figure 3). The reproducibility of holographic sensor manufacture was good; the error bars in Figure 3a represent the standard deviation of the average diffraction wavelength of the responses from three holographic pH sensors. The acrylamide-based holographic sensors displayed no response to pH over the range 3-7 in the absence of ionizable co-monomer, but as the mol % acrylic acid co-monomer increased, a series of sigmoidal curves were obtained in the pH range 4-6 that were analogous to those obtained for the poly-HEMA-based holographic sensor. The response of the acrylamide-based sensors reached saturation at pH values >6. The apparent pKa of the matrix-bound acidic group was 4.58 (at 6 mol %), as compared to 4.22, when the monomer acrylic acid was titrated in solution. The difference between these two values (0.36 units) is smaller than that observed for the HEMA-based pHsensitive gels (e1.51). This can be attributed to the fact that the HEMA system is more hydrophobic than the acrylamide system, and polymer hydrophobicity has been shown previously to decrease the strength (i.e., the pKa) of acidic or basic comonomers.27,28 (30) Brøndsted, H.; Kopecek, J. In: Polyelectrolyte Gels - Properties, Preparation and Applications; Harland, R. S., Prud’homme, R. K., Eds.; American Chemical Society: Washington, DC, 1992.

4428 Analytical Chemistry, Vol. 75, No. 17, September 1, 2003

Figure 3. (a) The effect of pH over the range 3-9 on the replay wavelength (λmax) of a polyacrylamide hologram (5 mol % N,N′methylene-bis(acrylamide) cross-linker) containing increasing amounts (0-12 mol %) of ionizable co-monomer, acrylic acid (AA) at 30 °C. (b) A plot of ∆λmax at pH 9 for polyacrylamide holograms containing 5 mol % N,N′-methylene-bis(acrylamide) cross-linker and various mol % of co-monomer, AA.

Figure 4a shows the effect of increasing the mol % of the cationic co-monomer, dimethylaminoethyl methacrylate (DMAEM), on the pH response of a poly-HEMA-based holographic sensor. The error bars in Figure 4a represent the standard deviation of the average diffraction wavelength of the responses from three holographic pH sensors. As anticipated from the pKa of the solution monomer, 8.22 (Figure 1b), the apparent pKa of DMAEM in the matrix phase was downshifted by ∼1.47 units at a co-monomer concentration of 6 mol %. The presence of the pendant DMAEM functionality reduces the overall response of the holographic sensor by ∼2-fold, as compared to the more hydrophilic methacrylic acid (MAA)-containing polymers (Figure 4b). The Effect of Cross-Linker Concentration. Cross-linking density is known to influence pH-dependent swelling of ionizable hydrogels, since an increase in mol % cross-linker restricts the equilibrium degree of swelling.25,30 Figure 5 shows the effect of increasing the mol % cross-linker (ethylene glycol dimethacrylate) on the pH response in the range 4-9 in a poly-HEMA holographic system containing 6 mol % methacrylic acid (MAA). It is apparent that color shifts (∆λmax) well in excess of 500 nm are achievable with these sensors with low (2.5 mol %) cross-linker density, but that as the density of cross-linker increases, the increased mechanical rigidity of the hydrogels reduces their relative swellability at any given pH value. The data also confirm other observations that the effect of cross-linking density on the swelling of poly(hydroxyethyl methacrylate-co-methacrylic acid) hydrogels was more pronounced at high pH than at low pH. Ionization of the carboxylic function at high pH interrupts the hydrophobic

Figure 6. The effect of increasing ionic strength (mM NaCl) at a fixed buffer concentration (20 mM) on the replay wavelength (λmax) of a poly-HEMA-co-MAA hologram at 5 mol % EDMA cross-linker density and 2 mol % MAA within the pH range 3-9 at 30 °C.

Figure 4. The effect of pH over the range 3-9 on the replay wavelength (λmax) of a poly-HEMA hologram (5 mol % EDMA crosslinker) containing increasing amounts (0-6 mol %) of ionizable comonomer, DMAEM at 30 °C. (b) A plot of ∆λmax at pH 3 for polyHEMA holograms containing 5 mol % EDMA cross-linker and various mol % of co-monomer, DMAEM.

Figure 5. The effect of increasing cross-linker density (2.5-15 mol %) on the replay wavelength (λmax) as a function of pH within the range 4-9 at 30 °C for a poly-HEMA hologram containing 6 mol % MAA co-monomer.

interactions and hydrogen bonding inside the gel that exists at low pH values.24 The Effect of Buffer Composition and Ionic Strength. Buffer composition and ionic strength affect the swelling of polyelectrolyte hydrogels.26 As the ionic strength increases at any

given pH value, the swelling decreases due to the increased concentration of counterions, the shielding of charges on the polymer backbone and the high ion concentration in the bulk solution. As the latter increases, the concentration of ions within the gel phase and the bulk solution equalizes and the osmotic pressure in the gel phase decreases. Figure 6 shows the combined effect of pH and ionic strength, obtained by varying the concentration of NaCl, at a fixed buffer concentration of 20 mM, on the replay wavelength of a poly-HEMA-co-EDMA (5 mol %)-co-MAA (4 mol %) hologram, was observed when the hologram was subjected to solutions of pH values within the range 3-9. Increasing the ionic strength from 25 to 500 mM reduces the ∆λmax by ∼2-fold at pH values >8. Similar results were obtained at 2 and 6 mol % MAA in the poly-HEMA-based holographic sensor and when KCl replaced NaCl as the salt added to adjust the ionic strength. Furthermore, reduced swellability as a function of pH was also observed with 2, 4, and 6 mol % DMAEM co-monomer in poly-HEMA and with 4, 6, and 8 mol % acrylic acid co-monomer in a polyacrylamide holographic matrix. Control experiments with base hydrogels containing no co-monomers showed that the replay color (λmax) of polyacrylamide holograms changed by only +10 nm as the ionic strength increased from 0 to 500 mM, while under the same conditions, the poly-HEMA-based holograms changed by -30 nm. Poly-HEMA contracts at high ionic strength, since hydrophobic interactions with the matrix backbone are promoted. There was little effect of the concentration of the buffer within the range 5-100 mM on the holographic response to pH, irrespective of the mol % of MAA or DMAEM in poly-HEMA or of acrylic acid in polyacrylamide, when the total ionic strength was maintained constant at 200 mM with NaCl. The monovalent buffer salts chloroacetate (pH 3) and acetate (pH 4.5) and the zwitterionic buffer salts MES (pH 6), MOPS (pH 7.5), and bicine (pH 9) were used for this study. The Effect of Temperature. Responsive hydrogels are known to undergo changes in hydration in response to temperature, which affects the solvation of the polymer.31 pH-sensitive holo(31) Hirokawa, Y.; Tanaka, T.; Matsuo, E. S. J. Chem. Phys. 1984, 81, 63796380.

Analytical Chemistry, Vol. 75, No. 17, September 1, 2003

4429

Figure 7. The effect of two cycles of step changes in pH (0.5 unit) within the range 4-6 on the replay wavelength response of a polyacrylamideco-AA (8 mol %) hologram in 50 mM buffer at 30 °C.

graphic sensors comprising 2, 3, and 4 mol % MAA in poly-HEMA and 4, 6, and 8 mol % acrylic acid in polyacrylamide were subjected to solutions of buffers prepared to the correct pH at the desired temperature and made up to an ionic strength of 200 mM with NaCl at temperatures of 20, 30, 37, and 45 °C. The poly-HEMA holograms were more temperature-sensitive at low co-monomer concentrations but became less sensitive as the hydrophilicity of the backbone increased. Exposure to a temperature change of 20-45 °C altered the pH response (λmax) at pH 9 by 29, 10, and 3%, respectively, for 2, 3, and 4 mol % MAA substitution levels and by less at other points in the pH titration curve. In contrast, the more hydrophilic polyacrylamide-based holographic pH sensors containing 4, 6, and 8 mol % acrylic acid displayed negligible response to temperature over the range 20-45 °C at any point in the pH range 3-9. Reversibility of the Holographic pH Sensors. A key issue in the application of pH-sensitive holographic sensors for realtime measurements of pH in biological samples is the swelling and deswelling kinetics of the hydrogel.18 The characteristic response times of poly-HEMA-co-methacrylic acid sensors to reach a new equilibrium λmax following a step pH change were proportional to the concentration of ionizable monomer (mol %) within the polymer, were a function of the magnitude of the pH step, and were inversely proportional to the buffer concentration. This behavior was similar to that noted for cross-linked polymers of poly-HEMA containing up to 20 mol % dimethylaminoethyl methacrylate.18 The buffering capacity of the sample solution influenced the response time of poly-HEMA-co-MAA (4 mol % MAA; 5 mol % EDMA cross-linker) holograms to a pH step from 5.5 (acetate) to 6 (MES) at 30 °C. The buffer concentration (5200 mM) was adjusted to 200 mM with NaCl in each case, and the initial replay wavelengths (λmax) of the holograms were within the range 530-541 nm. The response times for the holographic sensors were fastest (∼250 s) at high buffer concentrations (200 mM) and markedly slower (∼2000 s) at low buffer ion concentrations (5 mM). Furthermore, at low buffer concentrations, the replay color does not return to the basal level, suggesting a residual buffering capacity within the polymer matrix that overcomes the externally added buffer ions and that is not apparent 4430

Analytical Chemistry, Vol. 75, No. 17, September 1, 2003

at high buffer concentrations. This effect manifests itself as a hysteresis effect when pH-sensitive holograms are titrated up and down the pH scale within the range 4-9 at low buffer ion concentrations. At 5 mM buffer concentration, poly-HEMA holograms containing 4 mol % methacrylic acid show a hysteresis effect with a ∆pH of 0.15 at the approximate pKa (6.01) of the polymerbound acid group, while at 200 mM buffer, no hysteresis effect and ∆pH is observed. At the higher buffer ion concentrations, the hydrogels respond rapidly and reversibly to step changes in the pH. Figure 8 shows the effect of two cycles of sequential 0.5 pH unit step changes from pH 4 to pH 6 and back down to pH 4 on the replay wavelength (λmax) of polyacrylamide holograms containing 8 mol % acrylic acid co-monomer in 50 mM buffers at a total ionic strength of 200 mM and 30 °C. It is clear that under these conditions of buffer concentration and ionic strength, the ∆λmax changes are freely reversible and linearly related to the pH values, irrespective of whether the pH step changes are being monitored in the upward or downward directions. Under these experimental conditions, there is no evidence of hysteresis in the holographic response. Use of pH-Sensitive Holograms for Monitoring the Fermentation of Whole Milk. Actimel fermented milk drink was used as a source of the active strain of L. casei immunitas. The organism carried out homolactic fermentation and generated lactic acid as the main product of the fermentation of lactose32,33 was added to freshly defrosted milk and warmed to 37 °C in a cuvette containing a poly-HEMA-co-MAA (4 mol %) hologram, which had previously been washed with whole milk until a stable baseline λmax was obtained. The pH of the milk solution was monitored at regular intervals at 37 °C for 200 min using a small-volume pH electrode and the holographic pH sensor. Figure 9 shows the replay wavelength of the poly-HEMA-co-MAA hologram coplotted with the response from a standardized glass pH electrode. The pH of the milk solution fell nonlinearly from a starting pH of ∼6.3 to a value of ∼4.8 over a period of 200 min as the L. casei fermented lactose to lactic acid. The inset shows that the (32) Litchfield, J. H. Adv. Appl. Microbiol. 1996, 42, 45-95. (33) Singleton, P.; Sainsbury, D. Dictionary of Microbiology; John Wiley & Sons: Chichester, 1981.

Figure 8. The progress of fermentation of whole milk at 37 °C with L. casei monitored over 200 min with a poly-HEMA-co-MAA (4 mol %) hologram ([) and a standard glass pH electrode (0). The inset shows the hologram response plotted against the calibrated pH value of the milk suspension.

holographic replay wavelength varied linearly with the glass pH electrode response over the pH range investigated. Three independent fermentation runs produced almost superimposable data and attested to the validity of using holographic sensors to monitor real-time pH changes in whole milk samples. CONCLUSIONS A variety of different hydrogel compositions containing pendant ionizable functionalities have been synthesized as thin films (∼10 µm thick) and transformed into pH-sensitive reflection holograms. The replay wavelength of such sensors has been used to characterize the volume changes that occur when the hydrogels are exposed to solutions of different pH values. The effects of comonomer and cross-linker concentration, buffer composition, ionic strength, and temperature on the pH responses have been studied, whence optimized film compositions containing pendant ionizable groups showed fully reversible responses to changes in the pH of the sample solutions. These characteristics are similar to those reported for sensors based on intelligent polymerized crystalline colloidal arrays (IPCCA).2-6 However, in the IPCCA technique, the polyacrylamide hydrogel was partially hydrolyzed to convert pendant amides into carboxyls, although the mol % of these groups within the polymer was not measured. The diffraction behavior of the gels showed a monotonically red shift from 500 to 677 nm between pH 2 and 9.6 and a blue shift from 677 to 590 nm between pH 9.6 and 11.11.6 The system displayed a linear diffraction wavelength response up to pH 9.6, with a sensitivity corresponding to ∼23 nm/pH unit. Since it is possible to measure wavelength to a resolution of 1 nm, the authors claim to determine pH with 0.05 pH unit resolution. In contrast, the holographic pH sensor reported here displays a classical pKa-type sigmoidal wavelengthversus-pH curve, with measured pKa values close to what would be predicted for the matrix-bound ionizable monomers they contain. A poly-HEMA-co-MAA (6 mol %) holographic system is able to measure pH over the linear range (pKa ( 1; pH 6.01 ( 1; Figure 1) with a sensitivity of ∼165 nm/pH unit, equivalent to a

pH resolution of 0.006 units. This is a 7-fold higher sensitivity than the IPCCA system and could readily be enhanced by increasing the mol % of ionizable co-monomer or reducing the mol % crosslinker. Under these conditions, sensitivities of (1 milli-pH unit could easily be achieved. Both IPCCA8 and holographic systems exhibit very low temperature sensitivity over the range 20-45 °C. The IPCCA system does not appear to have been tested with crude samples to evaluate the practical utility of the device. In the work described in this report, a poly-HEMA-co-methacrylic acid holographic sensor was shown to be able to quantitate H+ concentrations in real time in a sample of milk undergoing fermentation in the presence of a L. casei. The hologram was exposed to fermenting whole fat milk at 37 °C, and the replay wavelength was monitored from the nonsample side as a function of time for 200 min. Calibration against a conventional glass pH microelectrode showed a linear relationship between the replay λmax (nm) and pH over the range investigated (pH 6.3-4.8). It is clear that these holographic sensors may find application in monitoring cell cultures, process streams, consumer products and clinical diagnostics. A particular advantage is that the sensors are planar and operate in a reflection format and are thus highly suitable for miniaturization in an array. Current work is aimed at demonstrating the utility of holographic sensors for monitoring the fermentation behavior of a number of microorganisms in nanowells. ACKNOWLEDGMENT The authors thank the Biotechnology and Biological Research Council (BBSRC) for a studentship (A.J.M.) and A. P. James, S. Kabilan, B. Madrigal-Gonzalez, D. Young, E. Taylor, and A. R. Lowe for useful discussions. Received for review November 25, 2002. Accepted June 5, 2003. AC020730K

Analytical Chemistry, Vol. 75, No. 17, September 1, 2003

4431