Optical Ratiometric Sensor for Alcohol Measurements - CiteSeerX

Analytical Letters, 40: 715–727, 2007 Copyright # Taylor & Francis Group, LLC ISSN 0003-2719 print/1532-236X online DOI: 10.1080/00032710601017847

Downloaded By: [University of Maryland] At: 15:32 6 March 2007

CHEMICAL & BIOSENSORS

Optical Ratiometric Sensor for Alcohol Measurements Silviya Petrova, Yordan Kostov, Kim Jeffris, and Govind Rao Center for Advanced Sensor Technology, Department of Chemical and Biochemical Engineering, University of Maryland, Baltimore, MD, USA

Abstract: The need for low-cost, robust alcohol sensors has increased with the renewed interest in alternative fuels as well as high-throughput screening of biological processes involving the production of ethanol. The goal of this research was to develop a miniaturized optical ratiometric ethanol sensor to be used for in situ measurements. The sensor is based on the fluorescent dye Nile Blue Chloride. When in solution, the dye exhibits a single fluorescence peak. However, a dual emission peak is observed upon physical immobilization of the dye in the hydrogel poly(ethylene glycol) dimethacrylate. The dual emission allows for ratiometric measurements, thus circumventing drawbacks associated with fluorescence intensity measurements such as signal variations due to dye bleaching, source intensity fluctuations, etc. In developing this sensor we investigated ethanol sensitivity; alcohol selectivity; response time; and cross-sensitivity with pH, polarity, and ionic strength. We found that the sensor is sensitive to a broad range of ethanol concentrations, namely 5% to 90% v/v. Due to the hydrogel’s restrictive pore size, the sensor is sensitive to short-chain alcohols such as methanol, ethanol, and propanol, but lacks sensitivity to larger alcohols such as butanol and hexanol. We also found the sensor maintains full functionality after autoclaving. Sensor sensitivity to alcohol in solutions of varying ionic strength is negligible, whereas the solvent’s polarity must be controlled to maintain meaningful results. The sensor is most sensitive in acidic and neutral environments, indicating promising use for yeast- based alcohol fermentations.

Received 24 June 2006; accepted 20 September 2006 The support from UTDF-TEDCO grant “Optical alcohol sensor” as well as from NSF BES 0517785 is gratefully acknowledged. Address correspondence to Yordan Kostov, Center for Advanced Sensor Technology, Department of Chemical and Biochemical Engineering, University of Maryland, Baltimore County, 1000 Hilltop Circle, Baltimore, MD 21250, USA. E-mail: kostov@ umbc.edu 715

716

S. Petrova et al.

Keywords: Ethanol, optical sensor, ratiometric, fluorescence


INTRODUCTION The measurement of ethanol concentration is of great importance in clinical, industrial, and biochemical areas as well as in the beverage industry. Ethanol is a useful solvent and is utilized in the production of perfumes, paints, lacquers, and explosives. Presently, the use of ethanol as a fuel, fuel additive, or a hydrogen source in fuel cells (Arico et al. 1998) has received renewed interest. Ethanol, as a fuel, has a substantial benefit over fossil fuels because it is an environmentally friendly, renewable resource. It may be produced by the fermentation of fruits, corn, or wheat, and is synthetically derived from acetaldehyde or ethylene. With numerous industries that utilize or produce ethanol, it is apparent that reliable methods are needed for its measurement and control. Many analytical methods have been developed during the years for the measurement of ethanol and other alcohols. They can be classified in three major categories: chromatographic, enzymatic, and optical. The chromatographic method is the most accurate and sensitive (Buttler et al. 1993; Johansson et al. 1993), with a lower limit of ethanol detection on the order of 0.005% v/v (Zinbo 1994). Drawbacks of this method include high cost, as well as the necessity for sample pretreatment and long operation times. Somewhat less precise but more rapid measurements are achieved by the use of enzymes. Determination of ethanol concentration is based on either of two enzymes, alcohol oxidase or alcohol dehydrogenase, by monitoring O2 consumption or H2O2 formation (Azevedo et al. 2005; Boujitita et al. 2000). The specificity of the enzyme binding sites provides highly selective and accurate sensors. The disadvantage of the sensors lies in their instability due to protein denaturing when exposed to high temperature, pressure, or pH extremes. In recent years, attempts have been made to place the “sensing” enzymes in close proximity to a transducer (Guilbalt et al. 1983) or to immobilize them in a matrix (Belghith et al. 1987; Mitsubayashi et al. 1994). Immobilization increases the stability of the enzymes and allows for continuous monitoring. Screen-printing technology has also been used for the mass production of lowcost disposable enzymatic sensors (Boujitita et al. 2000). The simplest approach to the determination of alcohols is an optical sensor method. Optical sensors generally have the advantages of low-cost manufacturability, safety, and miniaturization and are intended for use in real-time, in situ monitoring. Recently, lifetime–based (Chang et al. 1997) and fluorescence-based (Mohr and Spichiger-Keller 1997; Mohr et al. 1997; Blum et al. 2001; Orellana et al. 1995) alcohol sensors have been introduced utilizing various alcohol-sensitive dyes. Although extremely promising, these sensors suffer from dye leaching, cross-sensitivity to pH, and low specificity. They also lack high temperature stability and are subject to interference due to autofluorescence.

Optical Alcohol Sensor

717


Figure 1. Chemical structure of Nile Blue Chloride (NB).

This article presents an autoclavable ratiometric optical alcohol sensor that operates in the red-infrared region of the visible spectrum. It utilizes Nile Blue Chloride (NB; Fig. 1), a fluorescent dye with an excitation maximum at 650 nm. The excitation wavelength in the red region of the spectrum results in minimal background fluorescence and higher signal to noise ratio. The dye is physically entrapped in a poly(ethylene glycol) (PEG) dimethacrylate hydrogel exhibiting practically no leaching. The resulting sensor is miniature and has high temperature and pressure resistance, allowing functionality after autoclaving. Furthermore, the sensor is hydrophilic and therefore has short response times to changes in solution alcohol concentration. When entrapped in the hydrogel, the dye exhibits two emission peaks, which allows for ratiometric measurements. The ratiometric approach circumvents many of the problems of the intensity-based methods (Kermis et al. 2002; Kostov and Rao 1999), including signal variations due to dye bleaching, fluctuations in source intensity or temperature, and coloring of the media. The developed sensor was further characterized in terms of ethanol sensitivity, alcohol selectivity, response time, and cross-sensitivity to pH, polarity and ionic strength of the environment.

EXPERIMENTAL SECTION Reagents The fluorescent dye NB (90% dye content) was supplied by Aldrich (Milwaukee, WI), PEG dimethacrylate with molecular weight 1000 was supplied by Polysciences Inc. (Warrington, PA), and a photoinitiator Darocur 1173 was from Ciba Specialty Chemical Corporation (Tarrytown, NY). The effect of pH on the sensor was investigated using the following buffers: 0.1 M phosphate buffers of pH 2.6, 6.2, and 7; 0.1 M acetate buffer of pH 5.6; 0.1 M bicarbonate buffer of pH 8.5 and 9. For investigating the effect of ionic strength, NaCl solutions in water, with molarities of 0.1, 0.3, 0.5, and 0.7 M, were used. The effect of polarity was tested through

718

S. Petrova et al.


the following solvents with their respective polarity indexes: toluene (2.4), n-butanol (4), ethyl acetate (5.1), acetonitrile (5.8), dimethyl formamide (6.4). For measurements of ethanol in wine and hard liquors, no pretreatment of the sample was necessary. All reagents used in the experiments were of analytical grade or better, and de-ionized water was used throughout. Instrumentation Polymerization of the sensor was achieved using the Black-Ray Longwave UV Lamp, Model B-100A (UVP, San Gabriel, CA, USA), and all fluorescence measurements were performed using a Varian Cary Eclipse (Varian Inc., Palo Alto, CA, USA) fluorescence spectrofluorometer with a solid sample/cuvette holder. Immobilization Procedures Sensor membranes were obtained by dissolving 18 mg of NB into 0.5 ml of 70% v/v reagent alcohol (63% v/v ethanol, 3.5% v/v methanol, 3.5% v/v isopropanol, and the balance by water) to achieve a saturated NB solution. Then 600 ml of this solution, 450 mg of PEG dimethacrylate, and 12 ml of Darocur were mixed together and vortexed for 1 min to blend completely. For polymerization, the mixture was spread between glass plates in order to exclude interference from oxygen. Thickness was controlled with 5-cm wide, 0.01-cm thick aluminum spacer tape. The resulting hydrogel layers were approximately 120-mm thick. Free radical polymerization of the end methacrylate groups was initiated by exposure to a UV lamp for 7 min. A gray filter (absorption 2.00) was used to decrease the light intensity and the consequent dye bleaching. The obtained polymer was then peeled from the glass plate and washed first in water to hydrate and then soaked in 70% ethanol solution for 48 h to release unbound dye. The ethanol solution was changed several times until no coloring was observed, and then its fluorescence was measured in order to quantify dye leaching. A second immobilization procedure was also attempted in order to gain insight into dye attachment to the hydrogel. PEG dimethacrylate was polymerized to form a clear film and then boiled in saturated NB solution. After 1 h, the film was thoroughly colored to dark blue. It was then vortexed in 70% ethanol for 24 h, until the dye ceased leaching. Methods All experiments were held at room temperature. A soaking time of 12 min was adopted for each experiment. A glass cuvette with a lid was used to prevent the evaporation of the alcohol. All ethanol solutions were prepared with


719

100% ethanol and de-ionized water unless otherwise noted. An excitation wavelength of 580 nm was used for all fluorescence measurements. Downloaded By: [University of Maryland] At: 15:32 6 March 2007

RESULTS AND DISCUSSION The sensor’s performance is based on the sensitivity of NB to short-chain alcohols. The dye is immobilized in the hydrogel during cross-linking of the PEG methacrylate. It became obvious from the experiments that different forces are responsible for encapsulating the dye prior to and after polymerization. We hypothesize that, if the dye is added after polymerization, similar to staining DNA (Yang et al. 2000; Chen et al. 1999) and other proteins and macromolecules such as RNA and BSA (Lee et al. 2003), hydrogen bonding, hydrophobic interactions, and Van der Waals forces are responsible for keeping it in the PEG. In addition, the positively charged amino group of the dye could be attracted to the partial negative charge of the ketone group in the polymer. However, these forces are not strong enough to keep NB encapsulated, and 98% of it is washed away in 24 h. Apparently, a different mechanism takes place when the dye is added to the PEG solution prior to polymerization. Most probably, the PEG methacrylate macromers conform to the NB molecules during cross-linking in a mechanism similar to molecular imprinting (Andersson 2000a, 2000b).

Calibration, Sensitivity, and Response Time The ability to detect a ratiometric response from the alcohol sensor could arise from alcohol binding to free and “trapped” dye within the hydrogel. Free NB in solution has only one fluorescence emission peak at 670 nm (Fig. 2, insert), whereas NB inside PEG dimethacrylate hydrogel has two fluorescence peaks at approximately 630 nm and 700 nm (Fig. 2). The peak at 700 nm is assigned to the dye aggregate, and the peak at 630 nm is assigned to the monomer. This assignment is based on the following observation: when the sensor was prepared with a highly concentrated solution of NB (0.16% w/v), the emission peak at 700 nm was dominant. When the sensor was prepared with dilute solutions of NB (0.06% w/v), the emission peak at 630 nm became dominant. However, in both cases, the increase in fluorescence intensity was uniform for the entire spectrum for ethanol concentrations ,40% v/v. For higher ethanol concentrations, only the peak at 630 nm increased, whereas the peak at 700 nm shifted to a lower wavelength and, for 90% v/v ethanol, became a shoulder of the dominant peak. Because the dye is much more soluble in ethanol than in water, it is plausible that the dye forms aggregates in the polymer when exposed to a solvent in which it has low or no solubility. Accordingly, when the film is exposed to a solvent in which

720

S. Petrova et al.


Figure 2. Dependence of emission spectra of NB suspended in PEG on ethanol concentration (0%, 15%, 30%, 45%, 60%, 75%, 90% ethanol v/v). Insert: Dependence of emission spectra of NB in solution on ethanol concentration (0%, 20%, 40%, 60%, 80% ethanol v/v). In both cases, the excitation was at 580 nm.

the dye has a high solubility, the observed spectra is mostly that of the monomer. As mentioned earlier, when NB is embedded in a PEG dimethacrylate hydrogel, two emission peaks are observed at approximately 630 nm and 700 nm. With NB being a solvatochromic dye, a blue shift of both peaks is noted with increase of ethanol concentration. This is a typical behavior for most dyes (Yang et al. 2000), and thus a fluorescence intensity ratio (688 nm/700 nm) (625 nm/640 nm) accounting for the degree of the shift was adopted in building the calibration curve. A calibration function in the concentration range of 0 –90% v/v was established. The resultant calibration curve was found to fit a second-order polynomial of the form y ¼ 6E – 05x2 þ 0.0015x þ 0.6882, where x is the ethanol concentration, with r2 ¼ 0.9979 (n ¼ 6), and standard deviation of 1.7% (Fig. 3). Based on this calibration curve, the sensor was used to measure ethanol concentration in wines and hard liquors (Table 1). The measured values corresponded to the reported values with a relatively small standard deviation. The response time was also measured utilizing the developed ratio relationship (Fig. 4). The forward response time, t95 (for 95% of the total signal change to occur), was in the range of 4 – 8 min, whereas the reverse response time was in the range of 10 – 12 min. This response time is sufficient for the monitoring of most biological processes.


721


Figure 3. Calibration curve of the sensor for measurement of ethanol concentration (SD ¼ 1.7%, n ¼ 6).

The threshold of ethanol sensitivity was found to be 1% v/v for NB in solution and 5% v/v for NB suspended in a PEG methacrylate hydrogel. The accuracy of the sensor for ethanol was found to be within 1% v/v. Iimmobilizing the dye in the hydrogel seems to quench the fluorescence signal by imposing sterical hindrance on the trapped dye molecules. Selectivity to Various Alcohols Due to selective diffusivity within the PEG dimethacylate hydrogel, the sensor is only sensitive to short-chain alcohols such as methanol, ethanol, and propanol. Even though sensitivity of the sensor toward methanol and isopropanol is acknowledged, the emphasis of this article is on the sensor’s ethanol sensitivity. The selectivity of the sensor to various alcohols is represented in Fig. 5. As anticipated, the sensor can be used with equal accuracy for measurements of methanol, ethanol, and isopropanol. The sensor is unable to detect changes in the concentration of longer chain alcohols such as n-butanol and hexanol. Table 1. Measurements of ethanol concentration in wines and liquors using the developed sensor Sample Chablis blanc (White Wine) Smirnoff vodka (liquor) Seagram’s extra dry gin (liquor) Beefeater dry gin (liquor)

Measured (v/v), (n ¼ 3)

Reported (v/v)

9.9 (SD ¼ 0.7%) 35.1 (SD ¼ 0.5%) 39.7 (SD ¼ 0.6%) 46 (SD ¼ 1.6%)

10.5 35 40 47

722

S. Petrova et al.


Figure 4. Forward and reverse response time measured by the fluorescence intensity ratio (688 nm/700 nm) (625 nm/640 nm) for different ethanol concentrations.

Longevity and Durability Longevity of the sensor was tested using the sensing film to measure a 40% ethanol solution daily for a week. The standard deviation for the measurements was found to be 3.0%, which shows an excellent reproducibility of results and minimal leaching of the dye.

Figure 5. Selectivity of the sensor to various alcohols based on the fluorescence intensity ratio: V – methanol (SD ¼ 0.9%, n ¼ 3), B – ethanol (SD ¼ 0.7%, n ¼ 3), O – isopropanol (SD ¼ 4.2%, n ¼ 3), - n-butanol (SD ¼ 1.41%, n ¼ 3), D – hexanol (SD ¼ 0.41%, n ¼ 3).


723


The poly(alkyl acrylate) family of polymers typically has high thermostability (Lozinsky et al. 1995), allowing for a fully autoclavable sensor. As a test of durability, the sensor was positioned in a glass vessel, covered with de-ionized water, and autoclaved. A volume transition at the elevated temperature resulted in a reversible shrinking of the gel matrix. Upon rehydration, the sensing film did not resume its original smoothness and remained somewhat crinkled. Despite the structural change, the sensitivity and function of the sensor to alcohol were preserved. As can be seen in Fig. 6, the calibration curves of the sensor before and after steam sterilization are comparable. Nevertheless, a recalibration after autoclaving is needed, and the sensor can only be used under extreme conditions, such as high temperature or pressure, a limited number of times. However, the low cost of the sensor allows it to be disposable.

pH Dependence At present, NB is widely used as a pH indicator and pH stain due to its color change in response to a change in pH. The optical properties of the dye are altered due to the reversible protonation and deprotonation of the primary amine functional group (pKa 11.6). In acidic conditions, when the primary amine is fully protonated, the observed color is blue. It gradually shifts to purple with increasing pH and consequent deprotonation of the amine. Therefore, sensitivity of the sensor to pH was anticipated. However, to evaluate the pH interference with ethanol sensitivity, two sets of

Figure 6. Calibration curve for ethanol measurement based on the fluorescence intensity ratio: D – before autoclaving, A – after autoclaving.

724

S. Petrova et al.


Figure 7. pH dependence of the sensor measured by the fluorescence intensity ratio: O – plain buffer (SD ¼ 0.66%, n ¼ 3), B – buffer þ 20% ethanol v/v (SD ¼ 1.51%, n ¼ 3).

experiments were conducted with plain buffers and buffers to which 20% ethanol v/v was added. It was found that the spectra in both cases were comparable, particularly for low pH. The trends based on the fluorescence intensity ratio are compared in Fig. 7. For pH from 2.6 to 7, fluorescent intensity spectra exhibited peaks at 630 nm and 700 nm, whereas, for pH . 7, there was a single peak at 640 nm. Consequently, the jump in fluorescence intensity ratio corresponds to the switch from neutral to basic conditions. It is also notable that, despite the dye’s pH sensitivity, pH had negligible effect on the sensor performance below pH 7. It can be concluded that optimal sensor sensitivity occurs in the pH range of 2– 7. Coupled with the fact that the dye is nontoxic, the sensor can be used safely for in situ measurements of fermentation processes, such as yeast-based alcohol fermentations (pH 4).

Effects of Polarity and Ionic Strength In order to assess the influence of the ionic strength on the sensor’s sensitivity, aqueous NaCl solutions were used. It was found that ionic strength has a relatively small effect on the fluorescence intensity spectra (Fig. 8). However, the fluctuation of the signal maximum is compensated by ratiometric analysis (Fig. 8, insert). The standard deviation was found to be 1.51% (n ¼ 3). It can be concluded that ionic strength has negligible effect on sensor performance.


725


Figure 8. Effect of ionic strength on the sensor performance. Insert: Ionic strength dependence based on the fluorescence intensity ratio (SD ¼ 0.64%, n ¼ 3).

For testing the effects of polarity, six solvents with polarity indexes in the range from 2.4 to 6.4 were used. Since NB is a polarity-sensitive dye, an optical response to the polarity of its microenvironment was expected. NB is a fluorophore with a large excited-state dipole moment, resulting in fluorescence spectral shifts to longer wavelengths in polar solvents. Consequently, with increasing polarity index of the solvent, gradual intensity decrease of the peak at 700 nm and increase of the peak at 630 nm were noticed (Fig. 9). The intensity peak at 700 nm completely disappeared and the peak at 630 nm reached the highest intensity when the sensor was soaked in dimethyl formamide (polarity index of 6.4). Therefore, polarity of the environment has to be controlled to avoid interference with ethanol sensitivity.

CONCLUSION A ratiometric ethanol sensor was developed by physically immobilizing NB in a PEG dimethacrylate hydrogel. Using this immobilization technique, we found the sensor to be sensitive through a broad range of ethanol concentrations, from 5% to 90% v/v. Furthermore, the sensor exhibits measurement reproducibility (standard deviation of 1.7%), and it has a response time of 4– 8 min to 20% step changes in ethanol concentration. The sensor is also

726

S. Petrova et al.


Figure 9. Effect of polarity on the sensor performance (SD ¼ 3.8%, n ¼ 3). The following solvents with their respective polarity indexes were used for the experiment: toluene (2.4), n-butanol (4), ethyl acetate (5.1), acetonitrile (5.8), dimethyl formamide (6.4).

sensitive to other short-chain alcohols, such as methanol and isopropanol. Due to the high thermal stability of both NB and PEG dimethacrylate hydrogels, this immobilization also allowed for the physical preservation and proper functioning of the sensor even after autoclaving. Additionally, the preparation of the sensor is relatively simple and low cost. We found that the performance of the sensor is not affected by the ionic strength of the environment, but it is affected by highly polar solvents, in which a spectral shift to longer wavelengths is observed. We found that pH has an effect on the sensor, and greatest sensitivity occurs at acidic to neutral conditions. Through the characterization of this sensor, we believe we have achieved our goal of producing a robust, low-cost, disposable sensor that can be used for in situ, real-time measurements of alcohol.

REFERENCES Andersson, L.I. 2000a. Molecular imprinting: Developments and applications in the analytical chemistry field. J. Chromatogr. B, 745: 3 – 13. Andersson, L.I. 2000b. Molecular imprinting for drug bioanalysis. A review on the application of imprinted polymers to solid-phase extraction and binding assay. J. Chromatogr. B, 739: 163– 173. Arico, A.S., Creti, P., Antonucci, P.L., and Antonucci, V. 1998. Comparison of ethanol and methanol oxidation in a liquid-feed solid polymer electrolyte fuel cell at high temperature. Electrochem. Solid-State Lett., 1: 66 – 68. Azevedo, A.M., Prazeres, D.M.F., Cabral, J.M.S., and Fonseca, L.P. 2005. Ethanol biosensors based on alcohol oxidase. Biosens. Bioelectron., 21: 235– 247. Belghith, H., Romette, J-.L., and Thomas, D. 1987. An enzyme electrode for on-line determination of ethanol and methanol. Biotech. Bioeng., 30: 1001 –1005.


727


Blum, P., Mohr, G.J., Matern, K., Reichert, J., and Spichiger-Keller, U.E. 2001. Optical alcohol sensor using lipophilic Reichardt’s dyes in polymer membranes. Anal. Chim. Acta, 432: 269– 275. Boujitita, M., Hart, J.P., and Pittson, R. 2000. Development of a disposable ethanol biosensor based on a chemically modified screen-printed electrode coated with alcohol oxidase for the analysis of beer. Biosens. Bioelectron., 15: 257– 263. Buttler, T., Johansson, K.A.J., Gorton, L.G.O., and Marko-Varga, G.A 1993. On-line fermentation process monitoring of carbohydrates and ethanol using tangentialflow filtration and column liquid chromatography. Anal. Chem., 65: 2628– 2636. Chang, Q., Lakowicz, J.R., and Rao, G. 1997. Fluorescence lifetime-based sensing of methanol. Analyst, 122: 173– 177. Chen, Q.Y., Li, D.H., Yang, H.H., Zhu, Q.Z., Xu, J.G., and Zhao, Y. 1999. Interaction of a novel red-region fluorescent probe, Nile Blue, with DNA and its application to nucleic assay. Analyst, 124: 901– 906. Guilbalt, G.G., Danielsson, B., Mandelus, C.F., and Mosbach, K. 1983. Enzyme electrode and thermistor probes for determination of alcohols with alcohol oxidase. Anal. Chem., 55: 1582– 1585. Johansson, K., Jonsson-Petterson, G., Gorton, L., Marko-Varga, G., and Csoregi, E. 1993. A reagentless amperometric biosensor for alcohol detection in column liquid chromatography based on co-immobilized peroxidase and alcohol oxidase in carbon paste. J. Biotech., 31: 301– 316. Kermis, H.R., Kostov, Y., Harms, P., and Rao, G. 2002. Dual excitation ratiometric fluorescent pH sensor for noninvasive bioprocess monitoring: development and application. Biotechnol. Prog., 18: 1047– 1053. Kostov, Y. and Rao, G. 1999. Low cost device for fluorescent measurements. Rev. Sci. Instrum., 70: 4466– 4470. Lee, S.H., Suh, J.K., and Li, M. 2003. Determination of bovine serum albumin by its enhancement effect on Nile Blue fluorescence. Bull. Korean Chem. Soc., 24: 45 – 48. Lozinsky, V.I., Solodova, E.V., Zubov, A.L., and Simenel, I.A. 1995. Study of cryostructuration of polymer systems. XI: The formation of PVA cryogels by freezingthawing the polymer aqueous solutions containing additives of some polyols. J. Appl. Polymer Sci., 58: 171–177. Mitsubayashi, K., Yokoyama, K., Takeuchi, T., and Karube, I. 1994. Gas-phase biosensor for ethanol. Anal. Chem., 66: 3297– 3302. Mohr, G.J., Lehmann, F., Grummt, U-.W., and Spichiger-Keller, U.E. 1997. Fluorescent ligands for optical sensing of alcohols: Synthesis and characterization of p-N,N-dialkylamino-trifluoroacetylstilbenes. Anal. Chim. Acta, 344: 215– 225. Mohr, G.J. and Spichiger-Keller, U.E. 1997. Novel fluorescent sensor membranes for alcohol based on p-N,N-dioctylamino-4’-trifluoroacetylstilben. Anal. Chim. Acta, 351: 189– 196. Orellana, G., Gomez-Carneros, A.M., de Dios, C., Garcia-Mertinez, A.A., and MorenoBondi, M.C. 1995. Reversible fiber-optic fluorescing of lower alcohols. Anal. Chem., 67: 2231– 2238. Yang, Y., Hong, H.Y., Lee, I.S., Bai, D.G., Yoo, G.S., and Choi, J.K. 2000. Detection of DNA using a visible dye, Nile Blue, in electrophoresed gels. Anal. Biochem., 280: 322– 324. Zinbo, M. 1994. Determination of one carbon to three-carbon alcohols and water in gasoline/alcohol blends by liquid chromatography. Anal. Chem., 56: 244– 247.