Anal. Chem. 2010, 82, 5002–5004
Letters to Analytical Chemistry Optical Ammonia Sensor Based on Upconverting Luminescent Nanoparticles Heike S. Mader and Otto S. Wolfbeis* Institute of Analytical Chemistry, Chemo- and Biosensors, University of Regensburg, D-93040 Regensburg, Germany The sensor exploits the phenomenon of upconversion luminescence and is based on (a) the use of upconverting nanoparticles (UCNPs) of the NaYF4:Yb,Er type that can be excited with 980 nm laser light to give a green and red luminescence and (b) the pH probe phenol red immobilized in a polystyrene matrix. Exposure of the sensor film to ammonia causes a strong increase in the 560 nm absorption of the pH probe which, in turn, causes the green emission of the UCNPs to be screened off. The red emission of the UCNPs, in contrast, remains unaffected by ammonia and can serve as a reference signal. Due to the use of 980 nm as the excitation light source, the optical signal obtained is completely free of background visible luminescence of the sample and of scattered light. This is highly advantageous in the case of sensing ammonia in complex matrixes. There is an increasing demand for robust ammonia sensing systems.1 Particularly, detection systems for continuous monitoring2 are needed because ammonia is toxic to many aquatic organisms even in very low concentrations. Large quantities of ammonia are released in the farming industries and produced and used in the chemical industry, e.g., for fertilizers or in refrigeration systems. Leakage of ammonia can lead to life-threatening situations. Therefore, there is also a need for alarm systems warning of dangerous ammonium concentrations. Ammonia sensors also are being used as transducers in biosensors where ammonia is released as a result of enzymatic activity.3,4,16 Various kinds of optical sensors have been described for ammonia.5,6 For example, it may be sensed in the gas phase via its near-infrared (NIR) absorption.7 In being a weak base, it also * To whom correspondence should be addressed. Fax: +49-941-943-4064; E-mail
[email protected]. (1) Timmer, B.; Olthuis, W.; Van Der Berg, A. Sens. Actuators, B 2005, 107, 666–677. (2) Masserini, R. T.; Fanning, K. A. Marine Chem. 2000, 68, 323–333. (3) Kar, S.; Arnold, M. A. Anal. Chem. 1992, 94, 2438–2443. (4) Elamari, A.; Gisin, N.; Munoz, J. L.; Poitry, S.; Tsacopoulos, M.; Zbinden, H. Sens. Actuators, B 1997, 38-39, 183–188. (5) Waich, K.; Borisov, S.; Mayr, T.; Klimant, I. Sens. Actuators, B 2009, 139, 132–138. (6) Orellana, G.; Haigh, D. Curr. Anal. Chem. 2008, 4, 273–295. (7) Claps, R.; Englich, F. V.; Leleux, D. P.; Richter, D.; Tittel, F. K.; Curl, R. F. Appl. Opt. 2001, 40, 4387–4394.
5002
Analytical Chemistry, Vol. 82, No. 12, June 15, 2010
may be detected via the pH changes it can induce in immobilized pH indicators.8,9 The latter approach is more sensitive and can be applied to aqueous solutions. In fact, most optical sensing approaches for ammonia are based on transduction via pH, often along with fiber optics.10-12 Usually, a pH indicator dissolved in a buffer solution13 is entrapped in a gas-permeable polymer. Ammonia, on penetrating the polymer, causes the pH to change; this results in a change in color or fluorescence of the indicator dye.1,14,15 Our approach is based on the use of upconverting nanoparticles (UCNPs) that were incorporated into a sensor film schematically shown in Figure S1 in the Supporting Information. UCNPs have gathered considerable attention due to their potential use as fluorescent labels in bioassays17-19 and for sensing pH.20 Upconversion is a process where light of low energy (usually near-infrared light) is converted to higherenergy (visible) light via multiphoton absorptions or energy transfer processes.21-23 Near infrared excitation (in our case at 980 nm) causes neither photodamage of tissue nor background luminescence in the visible spectrum, i.e., where the luminescence of the upconversion material occurs.24 The UCNPs used in this work are of the NaYF4:Yb,Er type and were synthesized according to literature procedures.25 (8) (9) (10) (11) (12) (13) (14) (15) (16) (17) (18) (19) (20) (21) (22) (23) (24)
Mills, A.; Wild, L.; Chang, Q. Microchim. Acta 1995, 121, 225–236. Giuliani, J. F.; Wohltjen, H.; Jarvis, N. L. Opt. Lett. 1983, 8, 54–56. Wolfbeis, O. S. Anal. Chem. 2008, 80, 4269–4283. Malins, C.; Butler, T. M.; Maccraith, M. D. Thin Solid Films 2000, 368, 105–110. Rhines, T. D.; Arnold, M. A. Anal. Chem. 1988, 60, 76–81. Lobnik, A.; Wolfbeis, O. S. Sens. Actuators, B 1998, 51, 203–207. Preininger, C.; Mohr, G. J.; Klimant, I.; Wolfbeis, O. S. Anal. Chim. Acta 1996, 334, 113–123. Raimundo, I. M.; Narayanaswamy, R. Sens. Actuators, B 2001, 74, 60–68. Kuswandi, B.; Andres, R.; Narayanaswamy, R. Analyst 2001, 126, 1469– 1491. Chen, Z.; Chen, H.; Hu, H.; Yu, M.; Li, F.; Zhang, Q.; Zhou, Z.; Yi, T.; Huang, C. H. J. Am. Chem. Soc. 2008, 130, 3023–3029. Kumar, M.; Guo, Y.; Zhang, P. Biosens. Bioelectron. 2009, 24, 1522–1526. Mai, H. X.; Zhang, Y. W.; Si, R.; Yan, Z. G.; Sun, L. D.; You, L. P.; Yan, C. H. J. Am. Chem. Soc. 2006, 128, 6426–6436. Sun, L.-N.; Peng, H.; Stich, M. I. J.; Achatz, D.; Wolfbeis, O. S. Chem. Commun. 2009, 5000–5002. Auzel, F. Chem. Rev. 2004, 104, 139–173. Rantanen, T.; Pa¨kkila¨, H.; Ja¨msen, L.; Kuningas, K; Ukonaho, T.; Lo¨vgren, T.; Soukka, T Anal. Chem. 2007, 79, 6312–6318. Wang, F.; Liu, X. Chem. Soc. Rev. 2009, 38, 976–989. Li, Z.; Zhang, Y. Angew. Chem., Int. Ed. 2006, 45, 7732–7735. 10.1021/ac1007283 2010 American Chemical Society Published on Web 05/19/2010
Figure 1. A, B, C: Absorption spectra of phenol red in aqueous solutions of pH 5, 7, and 9, respectively. D, E, F: Luminescence emission (green and red) of the upconverting nanoparticles in aqueous solution following 980 nm laser excitation. The inset shows a TEM image of the upconverting nanoparticles (scale bar: 200 nm).
They are known26,27 to display excellent photostability and thermal stability. Additionally, their luminescence intensity is independent of pH in the range from 3 to 10 (see Figure S2 in the Supporting Information). The particles are uniform, with a diameter of 60-90 nm, as can be seen from the transmission electron microscopy (TEM) image in Figure 1, inset. Following excitation with a (low-cost) 980 nm diode laser (∼5 W), the UCNPs display green and red emission peaks (see Figure 1). The pH probe used in this work (phenol red; PR) is nontoxic, undergoes a large spectral shift (from 435 to 560 nm), and a distinct color change (from yellow to pink) with increasing pH. PR is not fluorescent by itself, and its pKa value at 25 °C is reported28 to be 7.9. Absorption spectra at three typical pH values and the overlap with the green emission of the nanoparticles are shown in Figure 1. If PR is present in its (pink) base form, the dye can be expected to exert an inner filter effect on the green emission on the nanoparticles. In its (yellow) acidic form, it can be assumed not to affect the intensity of the emission of the UCNPs. Initial experiments were performed with PR and UCNPs in buffered aqueous solution. On changing their pH, they undergo the color changes shown in Figure S3 (see the Supporting Information). In the presence of PR, the intensity of the green emissions of the UCNPs (bands D and E in Figure 1) is reduced with increasing pH, even though the nanoparticles by themselves have an emission that is independent of pH (Figure S2 in the Supporting Information). The red emission of the UCNPs, in contrast, is not affected by the filter effect exerted by the pH probe. Next, a sensor film was designed that is composed of (a) the NaYF4:Yb,Er nanoparticles; (b) the pH indicator probe PR; and (c) a proton impermeable polystyrene matrix that hosts the NPs and withholds the lipophilic pH indicator. The resulting “cocktail” (a solution of all components in chloroform; see the Supporting Information) was spread onto an inert and optically transparent polyester support made from poly(ethylene tere(25) Yi, G.; Lu, H.; Zhao, S.; Ge, Y.; Yang, W.; Chen, D.; Guo, L.-H. Nano Lett. 2004, 4, 2191–2196. (26) Li, Z; Zhang, Y; Jiang, S Adv. Mater. 2008, 20, 4765–4769. (27) Wu, S.; Han, G.; Milliron, D. J.; Aloni, S.; Altoe, V.; Talapin, D. V.; Cohen, B. E.; Schick, P. J. Proc. Natl. Acad. Sci. U.S.A. 2009, 106, 10917–10921. (28) Berthois, Y.; Katzenellenbogen, J. A.; Katzenellenbogen, B. S. Proc. Natl. Acad. Sci. U.S.A. 1986, 83, 2496–2500.
Figure 2. Representative luminescence spectra of the sensor film at ammonia concentrations between 0 and 20 mM, following photoexcitation at 980 nm. Spectra are normalized to the peak at 655 nm, whose intensity is not affected by ammonia.
phthalate) to form a film with a thickness of ∼12 µm after evaporation of the solvents. This film, a cross section of which is shown in Figure S1 (see the Supporting Information), is insensitive to changes in the pH of solutions in contact with the sensor membrane because the polystyrene matrix is impermeable to protons. Ammonia, on the other hand, is capable of penetrating polystyrene and inducing an increase in the pH, and thus, the (expected) color changes from yellow to pink. Diffusion of ammonia into the sensor membrane causes the pH of the water layer around the probe to increase. This, in turn, causes the indicator PR to change color from yellow to pink. The intensities of neither the green nor the red emission of the UCNPs are affected by low pH (i.e., in the absence of ammonia) because the pH probe in the sensor film does not display any absorbance between 500 and 700 nm. On increasing the concentration of dissolved ammonia, the local pH is increased so that the absorbance of the 560 nm band of PR in the film increases substantially, thus screening off the green emission of the UCNPs. Figure 2 shows the decrease in the intensity of the green emission (bands D and E) in the presence of different concentrations of dissolved ammonia. The ratio of the intensities of the green emissions (bands D and E) and the red emission (band F) of the UCNPs changes from 2.6 (in absence of ammonia) to 1.2 (saturation with ammonia) (see Figure S4 in the Supporting Information). This corresponds to a 54% decrease in the optical signal. This effect is not due to any absorption of light of the 980 nm laser because neither PR nor the polystyrene has measurable absorbance at this wavelength. The color of the UCNP emission can be easily detected by the naked eye and changes from bright green to yellow (the chromatic mixture of green and red; see Figure S5 in the Supporting Information). The red emission (band F) of the UCNPs is not affected by ammonia. Therefore, this peak can serve as an internal standard to allow for ratiometric measurements. By determining the ratio of the green to the red peak (I541/I655), the signal of the sensor becomes independent of inhomogeneities in the sensor membrane, of the concentration of the UCNPs, and of fluctuations of the intensity of the laser. The sensor responds to aqueous solutions of ammonia in concentrations between 1 and 20 mM, corresponding to 40-800 ppm. The limit of detection (LOD) was determined as 400 µM, assuming that the luminescence can be measured with a precision Analytical Chemistry, Vol. 82, No. 12, June 15, 2010
5003
Figure 3. Relative signal change, reversibility, and response time of the response of a 6 µm sensor film to 50 mM solution of ammonia in water at 22 °C. The ratio of the intensities of the green to the red emission serves as the analytical signal.
of ±1%. Sensors with lower limits of detection have been reported5,29 which is due to the use of indicator dyes of lower pKa values. Scorsone et al.30 report on a fiber-optic sensor for gaseous ammonia also based on the use of PR and give a dynamic range from 100 to 10 000 ppm. The difference may be due to the use of another matrix material and the addition of UCNPs to our sensor matrix. The response of a 6 µm sensor film is shown in Figure 3. Repetitive cycling between a 50 mM ammonia solution and water revealed good reversibility, with virtually no leaching of the indicator. The ratios of the intensities of the green and the (29) Werner, T.; Klimant, I.; Wolfbeis, O. S. J. Fluoresc. 1994, 4, 41–44. (30) Scorsone, E.; Christie, S.; Persaud, K. C.; Simon, P.; Kvasnik, F. Sens. Actuators, B 2003, 90, 37–45. (31) Wolfbeis, O. S. Fiber Optic Chemical Sensors and Biosensors; CRC Press: Boca Raton, FL, 1991; Vol. 2, Chapter 11, pp 55-82. (32) Werner, T.; Klimant, I.; Wolfbeis, O. S. Analyst 1995, 120, 1627–1631. (33) Mcdonagh, C.; Burke, C. S.; Maccraith, B. D. Chem. Rev. 2008, 108, 400– 422. (34) Borisov, S. M.; Wolfbeis, O. S. Chem. Rev. 2008, 108, 423–461.
5004
Analytical Chemistry, Vol. 82, No. 12, June 15, 2010
red peak slightly differ from the calibration plot due to a slightly different experimental setup. Response times in either direction (forward and reverse) were determined (by alternatingly exposing the sensor membrane to plain water and to 50 mM aqueous solutions of ammonia, respectively) and were found to be 5 and 7 min, respectively. It is noted that the system is rather sensitive to temperature and requires thermostatization to ±0.1 °C. Like in all other sensors for ammonia, the thickness of the sensor film has a considerable effect on response and recovery times due to the slow diffusion of ammonia through the membrane. In addition, response times are much longer for lower concentrations of dissolved ammonia but much faster for gaseous ammonia.31-33 The sensor is surprisingly stable. It was stored for 3 months in a humid atmosphere, after which it still was fully responsive. The preparation of much thinner sensor layers is conceivable and will shorten response times, and the LOD may be lowered by choosing an indicator of a lower pKa. In conclusion, we note that this new type of ammonia sensor has attractive features including NIR laser excitation, complete lack of background luminescence, use of easily accessible materials, and wide scope in that various other combinations of sensor materials (in terms of pH probes, polymers, and additives) may be used in the future to adapt such sensors to the specific needs of sensing and (enzymatic) biosensing.34 SUPPORTING INFORMATION AVAILABLE Additional information as noted in text. This material is available free of charge via the Internet at http://pubs.acs.org. Received for review March 21, 2010. Accepted May 16, 2010. AC1007283
