A CL Mode Detector for Rapid Catalyst Selection and Environmental

DOI: 10.1021/es702845z. Publication Date (Web): April 18, 2008. Copyright © 2008 American ... Suzhou Institute of Trade and Commerce. Cite this:Envir...
0 downloads 0 Views 2MB Size
Download PDF
Environ. Sci. Technol. 2008, 42, 3886–3892

A CL Mode Detector for Rapid Catalyst Selection and Environmental Detection Fabricated by Perovskite Nanoparticles F E I T E N G , * ,†,‡ T O N G G U A N G X U , † YANG TENG,§ SHUHUI LIANG,† BULGEN GAUGE,† JIE LIN,† WENQING YAO,† RUILONG ZONG,† Y O N G F A Z H U , * ,† A N D Y O U F E I Z H E N G ‡ Department of Chemistry, Tsinghua University, Beijing 100084, P.R. China, College of Environmental Science and Technology, Nanjing University of Information Science and Technology, Nanjing 210044, P.R. China, Department of Information Science, Suzhou Institute of Trade and Commerce, Suzhou 215009, P.R. China

Received November 13, 2007. Revised manuscript received January 17, 2008. Accepted February 25, 2008.

La1-xSrxMnO3 (x ) 0, 0.2, 0.5, 0.8) nanoparticles were synthesized and their chemiluminescence (CL) and catalytic properties of CO oxidation were determined. We mainly investigated the influences of filter band length, flow rate of gas, test temperature, catalyst compositions, and particle size on CL intensities and catalytic activities of the catalysts. The catalysts were characterized by means of XRD, TEM, N2 adsorption isotherm, CO-TPD, and O2-TPD, etc. It was found that the strong CL response signals occurred over these perovskites nanoparticles; and that CL properties of the catalysts were well correlated with the reaction activities. These nanoparticles can be used to fabricate a stable gas detector due to a high activity and stability of perovskite structure. CL mode could be a rapid and effective method for the selection of new catalysts from thousands of materials, as well as for the detection of environmental deleterious gases.

Introduction Chemiluminescence (CL) is defined as the emission of electromagnetic radiation (usually in the visible or nearinfrared region) produced by a chemical reaction, in which the resulting molecular species are excited to the electronic excited states. While the exited molecular species fall to their ground states, a weak light is released. For a solid catalyst, CL generally results from the interaction between gas and solid surface, which has been studied for decades. In 1976, CL phenomenon was first reported by Breysse et al. (1). They observed that during catalytic oxidation of CO on ThO2 surface, a weak catalytic luminescence phenomenon has occurred. This luminescence mode was defined as “cataluminescence (CTL)”. At that time, however, the CTL phenomenon has not attracted attention to researchers. * Address correspondence to either author. Phone & Fax: (+86) 10-6278-7601. E-mail:[email protected](Y.Z.)[email protected] (F. T.). † Tsinghua University. ‡ Nanjing University of Information Science and Technology. § Suzhou Institute of Trade and Commerce. 3886

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 42, NO. 10, 2008

Since the 1990s, nanoparticles have attracted widespread attention because of their potential applications in various fields, such as catalysis (2, 3), chemical and biochemical sensing (4, 5), and biological imaging (6, 7). Nanoparticles have been extensively researched due to their high reactivities. Nanoparticles have provided new avenues to expand the analytical applications of this CL detection mode. McCord et al. found that the interaction of porous silicon with nitric acid or persulfate could result in an intense CL, which suggested that CL could be generated on the surface of nanoparticles (8). The direct CL reactions of gold nanoparticles with bis(2,4,6-trichlorophenyl)oxalate and hydrogen peroxide have also been examined but not yet applied in analysis (9–11). The different organic vapors can be discriminated from the different CL responses in the presence of the nanoparticles. Many investigations have indicated that CL property of nanoparticles would be promising for new applications (12, 13). So far, CL properties of noble metals-based nanoparticles have been researched extensively. Since these noble metals are scarce and expensive, it is desirable to research CL properties of inexpensive materials. Recently, CL properties of several single-oxide nanoparticles have been studied by some researchers. Zhu et al. (14) first researched CL of organic vapor over different nanoparticles, including MgO (28 nm), TiO2 (20 nm), Al2O3 (18 nm), Y2O3 (90 nm), LaCoO3:Sr2+ (50 nm), and SrCO3 (25 nm). Rakow et al. (15) reported that the nanoparticles are potentially suitable for processing as a chip-mounted sensor array, Nakagawa et al. (16) have manufactured a γ-Al2O3 sensor to detect ethanol, acetone, and carboxyl acid, which utilize the combustibility of these chemicals on this solid. Okabayashi et al. (17) have also manufactured a Dy-doped γ-Al2O3 sensor used for hydrocarbon gases. Zhang et al. (14, 18–20) have designed the highly selective CL sensors for ethanol, aldehyde, ammonia, sulfureted hydrogen, and butyl ketone. They reported that ZrO2 nanoparticles-based sensor for ethanol has no response to hexane, cyclohexane, ethylene, hydrogen, ammornia, and nitrogen oxides. While Tb was doped, the sensibility of Tb-ZrO2 sensor can be enhanced as high two level as that of the undoped one. They have also manufactured MgO-Al2O3 (4:1 mol ratio) nanosensor, which has a high selectivity to butyl ketone (21). Nevertheless, the structure–stability of the single oxides is generally not good enough to use under severe conditions. Especially at high temperatures, single oxides are easy to sinter or to transform to other crystalline phases, resulting in the decrease of sensitivity. The highly stable CL sensors are still challenged by the structure properties of materials which limit their practice applications. It is well-known that perovskites (ABO3) are one of the most active and stable catalysts for the removal of pollutants (CO, NOx and hydrocarbon), and perovskites as solid oxidants have been extensively researched in environmental catalysis (22). Up to now, nevertheless, CL properties of gases on perovskites have been scarcely researched. As we know, the catalyst selection is generally carried out by the determination of reaction activity. Generally, this determination mode is needed to construct expensive reaction equipments, and time-consumed and tedious. It is needed to design a new and rapid mode of catalyst selection. CL mode is one of the most useful methods, because the determination mode of CL can be performed within a few minutes and no expensive reaction equipment is needed. To the best of our knowledge, most of previous research about CL properties of micro- or nanosized particles have focused 10.1021/es702845z CCC: $40.75

 2008 American Chemical Society

Published on Web 04/18/2008

FIGURE 1. XRD patterns of La1-xSrxMnO3 catalysts prepared by the coprecipitation method. on CL applications in analysis (8, 18, 23). CL applications in catalysis have not been reported, and few studies have revealed the essence correlations between catalytic reaction and CL mechanism (24). In this research, we reported, for the first time, CL properties for CO over perovskite La1-xSrxMnO3 (x ) 0, 0.2, 0.5, 0.8) nanoparticles. The effects of filter’ band length, flow rate of gases, test temperature, catalyst compositions, and particle size on CL intensity of CO were investigated, and the essence correlation between CL and catalytic reaction of the catalyst was elucidated.

Experimental Section Chemicals. In this work, all chemicals, including La(NO3)3 · 6H2O (A.R.), Sr(NO3)2(A.R.), Mn(NO3)2 · 2H2O (A.R.), KOH (A.R.) and Citric acid (CA) (A.R. C6H8O7 · H2O, MW ) 210), were purchased from Beijing Chemicals Company of China and used as received. Synthesis of La1-xSrxMnO3 Nanoparticles (50 nm). La1-xSrxMnO3 (x ) 0, 0.2, 0.5, 0.8) nanoparticles were prepared by a coprecipitation method, in which the molar ratios of Sr/La/Mn were kept at (1 - x)/x/1. Under ultrasonic agitation, the mixed solution of metals nitrates was mixed with KOH, while the final pH value of the mixture was adjusted to pH 9. The solids were separated by centrifugation, washed with deionized water, and dried at 80 ° overnight in an oven. The solids were then calcined at 700 °C for 5 h under flowing air. Synthesis of La0.8Sr0.2MnO3 Nanoparticles (20 nm). La0.8Sr0.2MnO3 nanoparticles were prepared by a citrate method (25). The mixture solution was prepared by dissolving nitrates and citric acid (CA) in deionized water, in which the molar ratios of Sr/La/Mn/CA were kept at 0.8/0.2/1/4. The resulting solution was evaporated at 80°, and completely dried at 80° overnight in an oven. The obtained spongy material was crushed and calcined at 600° for 5 h under flowing air. Synthesis of La0.8Sr0.2MnO3 Particles (g100 nm). The large La1-xSrxMnO3 particles were obtained by a solid state reaction. The metals nitrates were ground and mixed homogeneously and then calcined at 1000° for 2 h under flowing air. Characterization of the Nanoparticles. The sample was characterized by XRD (Rigaku D/MAX-RB X-ray powder diffractometer), using graphite monochromatized Cu KR radiation (λ ) 0.154 nm), operating at 40 kV and 50 mA. The patterns were scanned from 10° to 70° (2θ) at a scanning rate of 1° min-1. A nitrogen adsorption isotherm was performed at -196° on a Micromeritics ASAP2010 gas adsorption analyzer. The sample was degassed at 250° C for 5 h before the measurement. Surface area was calculated by the BET (Brunauer–Emmett–Teller) method. The morphology of the

catalyst was observed on a TEM (JEOL 200CX) instrument with an accelerating voltage of 200 kV. The powders were ultrasonically dispersed in ethanol, and then deposited on a thin amorphous carbon film supported by a copper grid. Chemiluminescence (CL) Experiments. The CL detection system employed in this work is shown in the literature (14). The detect system consists of a CL-based sensor, a digital programmable temperature controller of the sensor, and an optical detector. The CL sensor was made by sintering a 0.2 mm thick layer of the catalyst powder on a cylindrical ceramic heater of 5 mm in diameter. Typically, 0.03 g of catalyst powders were mixed with absolute ethanol to prepare a paste, and the paste was coated on the surface of heating tube; and then it was heated in an oven at 110° for 24 h and heated at 500° C for 1 h in air to form a coating. In order to accurately control the thickness, the same procedure was repeated. The obtained sensor was set in a quartz tube of 12 mm (i.d.) through which air at atmospheric pressure flows at a constant rate. A certain volume pulse of CO was injected into the air flow. The sample gas can flow only through the outside of the ceramic heater because this ceramic tube is solid. The temperature of the sensor was controlled by a digital temperature controller. The CL intensity at a certain wavelength was measured by a photon-counting method with a BPCL ultraweak chemiluminescence analyzer (BPCL, Chemiluminescence analyzer made by Biophysics Institute of the Chinese Academy of Science). In the experiment, the optical filters with different wavelength (from 400 to 750 nm) were used. Before each test, the catalyst sensor was heated at 500° C for 1 h in air to avoid the influence of previous absorbates. Catalytic Reaction Experiments. The oxidation of CO was carried out in a conventional flow system at atmospheric pressure. A 0.1 g aliquot of catalyst was loaded in a quartz reactor (inner diameter: 5 mm), with quartz beads packed at both ends of the catalyst bed. The thermal couple was placed in the catalyst bed to monitor the reaction temperature since CO oxidation is an exothermic reaction. Before each run, the catalyst was flushed with air (200 mL min-1) at 500° for 1 h in order to remove the previous absorbates from the catalyst surface, and then cooled to 30°. A gas mixture of 2 vol.% CO and 98 vol.% air was fed to the catalyst bed at a certain flowing rate of 200 mL min-1. The inlet and outlet gas compositions were analyzed by an online gas chromatograph with a GDX-403 GC-column (1.5 m × 4 mm) at 100° and a hydrogen flame ionization detector (FID).

Results XRD, TEM, and BET Results of La1-xSrxMnO3 Nanoparticles. Figure 1 gives XRD patterns of La1-xSrxMnO3 samples by the coprecipitation method. All the diffraction peaks can VOL. 42, NO. 10, 2008 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

3887

be well indexed to a pure perovskite structure with a high crystallinity, and no peak shift can be observed for all samples. All samples can be indexed to a primitive pseudocubic unit cell (space group I4/mcm, a ) 0.54806 (3) nm, c ) 0.77206 (6) nm), in agreement with the lattice parameter of bulk material. The observation with TEM (not given) confirmed that all the samples consisted of 50 nm nanoparticles, and their size differences were not obvious. However, the particle size distributions of the catalysts determined by the light scattering method were slightly different (see Figure S1 of the Supporting Information). The sample at x ) 0.2 has a narrowest particle size distribution. Calculated by Scherrer equation, the average crystal sizes of these samples are similar, which are in the range of 26–27 nm. Besides, their BET areas are also similar (about 14 m2 g-1) (see Table S1, Supporting Information). This is because these catalysts were prepared by the same preparation in which the particles sizes are easy to be controlled. Effect of Test Conditions on CL Intensities of La0.8Sr0.2MnO3 Nanoparticles. Typically, we determined CL intensities of La0.8Sr0.2MnO3 nanoparticles at different test conditions, such as flowing rate of gases (F), test temperature (T), CO concentration (C), and band length of filter (λfilter). The effect of flowing rate of gases on CL intensity of La0.8Sr0.2MnO3 catalyst was investigated (see Figure S2, Supporting Information). It can be observed that CL intensity first increased with flowing rate (F e 320 mL min-1), and then it reached a maximum at F g 320 mL min-1. This directly means that the catalytic reaction, which produces the luminescent molecules, is controlled by a diffusion process at lower flowing rates; but the catalytic reaction is controlled by a surface reaction process at higher flowing rates. In the diffusioncontrolled process, the adsorbates on the catalyst surface reacted rapidly once the reacting molecules reached the catalyst surface, and the mass-transfer process dominated the whole process. In the surface reaction-controlled process, nevertheless, the mass transfer of reacting molecules from bulk to catalyst surface is quick and the surface reaction dominates the whole process. The effect of CO concentration on CL intensity of La0.8Sr0.2MnO3 catalyst was researched (see Figure S3 and Table S2, Supporting Information). A good linear correlation between CL intensity and CO concentration can be observed in the concentration range (5.0–320.0 µg mL-1), and the detection limit of CL sensor for CO has been determined as low as 0.5 µg mL-1. This means that CL properties of the catalyst should be determined at a constant concentration of CO. Figure S4 in the Supporting Information shows the temperature dependent CL intensity over La0.8Sr0.2MnO3 catalyst. CL intensity increased with test temperature and reached a high value. It is clear that the test temperature has a significant influence on CL intensity. This is because CO conversion increased with the temperature. It is important for CL determination to maintain a constant test temperature. Figure S5 in the Supporting Information gives CL intensities of La0.8Sr0.2MnO3 nanoparticles, which were tested using eight filters with different band lengths (400–750 nm). La0.8Sr0.2MnO3 nanoparticles show the highest CL intensity with a 640 nm filter, indicating that the generated weak lights in CO oxidation centered near 640 nm. From above, CL properties of the catalysts should be determined under the same conditions in order to obtain the correct results. CL Intensities of La1-xSrxMnO3 Catalysts. Figure 2 shows the effect of Sr doption of La1-xSrxMnO3 catalysts on their CL intensities. While x ) 0, 0.2, 0.5, and 0.8, CL intensity ratios of La1-xSrxMnO3 catalysts are 1.5/4.8/3.5/2.9. At x ) 0.2, La0.8Sr0.2MnO3 catalyst has the strongest CL intensity. The result shows that CL intensity of the catalyst was enhanced significantly by the Sr doption. It has been concluded that luminescent efficiencies are dependent on the kinds of 3888

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 42, NO. 10, 2008

FIGURE 2. CL intensities of CO over La1-xSrxMnO3 catalysts (x ) 0, 0.2, 0.5, 0.8) prepared by the coprecipitation method: C (CO concentration in the mixed gases of CO and air) ) 200 µg mL-1, F (flowing rate of the mixed gases of CO and air) ) 200 mL min-1, λfilter (the band length of filter) ) 640 nm, T (testing temperature) ) 280°. reactants and catalysts (14, 26–31). It is generally accepted that the partial substitution of Sr for La can improve the catalytic activity of the catalyst. We further investigated CL properties of La0.8Sr0.2MnO3 samples with different sizes. La0.8Sr0.2MnO3 samples were prepared by citrate, coprecipitation, and solid state reaction methods, respectively. TEM results confirm that their particle sizes are 20, 50, and larger than 100 nm, respectively (Figure 3). Figure 4 shows their CL spectra. CL intensities of the nanoparticles are as high 5 times as that of large particles (g100 nm). It is clear that the smaller the nanoparticles, the stronger its CL intensity. This is due to the high reactivity of nanoparticles. Reactivity of La1-xSrxMnO3 Catalysts. The reaction activities of CO oxidation over La1-xSrxMnO3 catalysts are shown in Figure 5 and summarized in Table S3 of the Supporting Information. At x ) 0.0, 0.2, 0.5, and 0.8, the complete conversion of CO was achieved at 365, 225, 245, and 280°, respectively. At x ) 0.2, the catalyst shows the highest activity. It has been confirmed that the catalytic activity of catalyst can be enhanced by the partial substitution of Sr for La. This is because the doption of Sr leads to the formation of defects (anion or oxygen vacancies), which significantly affect the adsorption and activation of reacting molecules (23, 32–34). This will be confirmed and discussed in the latter section. Most importantly, the activity order of the catalysts is well consistent with that of CL intensity. This may indicate that CL mode can be an effective means to select the new catalysts from thousands of materials, since the determination of CL spectra can be fulfilled within a few minutes.

Disscussion Correlation of CL Properties with Catalytic Activities of the Catalysts. Concerning CO oxidation mechanism, a socalled suprafacial catalytic process has been proposed by Voorhoeve et al. (35) They considered that the crystal structure and the surface properties of the catalyst seem to be very important. Further, Tascon et al. proposed the following catalytic mechanism of CO oxidation on LaCoO3 (36): O2(g) f O2(ads) f 2O(ads)

(i)

Co(g) f CO(ads)

(ii)

CO(ads) + 2O(ads) f CO3(ads)

(iii)

CO3(ads) f CO2(ads) + O(ads)

(iv)

CO2(ads) f CO2(g)

(v)

FIGURE 3. Typical TEM micrographs of La0.8Sr0.2MnO3 samples prepared by the different methods: (a) citrate, (b) coprecipitation, (c) solid state reaction.

FIGURE 4. Effect of particle size on CL intensity of La0.8Sr0.2MnO3 catalyst: C ) 200 µg mL-1, F ) 200 mL min-1, λfilter ) 640 nm, T ) 280°. The 20, 50 and g100 nm La0.8Sr0.2MnO3 particles were prepared by citrate, coprecipitation, and solid state reaction methods, respectively.

FIGURE 5. The activities of CO oxidation over La1-xSrxMnO3 catalysts prepared by the coprecipitation method: F (Flowing rate of the mixed gases of CO and air) )200 mL min-1, C (CO concentration in the mixed gases of CO and air) ) 2 vol. %. Here, (g) and (ads) refer to gas phase and adsorbed states. The authors suggested that Co and O sites act as adsorption and activation centers for CO and O2, respectively. It is considered that the lattice oxygen does not participate in complete combustion and that the adsorbed oxygen species

alone are responsible for the total oxidation reaction (37). However, Rao et al. (38) have proposed that the lattice oxygen of perovskite participate in CO oxidation. A full consensus has not been reached so far on the main aspects which determine the catalytic behavior of perovskites. In our opinion, we think that the adsorbed oxygen could play a key role in CO oxidation because CO oxidation proceeded in a low temperature in which the lattice oxygen may less active. This needs further research. In our catalyst, Mn ions act as active centers for CO chemisorption. CO and Mn sites function as Lewis base and Lewis acid sites, respectively (39). The interaction between Mn ions and CO molecules can be modeled by crystal field theory. CO 5σ electrons and empty 2π* orbitals can coordinate with d orbitals and d electrons of Mn ions, respectively. CO 5σ electrons donate electrons to coordinate with MnO6 species possessing empty eg (dz2) orbitals. The empty 2π* orbitals of CO molecules show some P orbital properties. Therefore, the t2g (dxy, dyz, or dxz) electrons of Mn ions can hold to the empty 2π* orbitals of CO, forming feedback π bond. It not only increases the coordination bond strength (MnsCO) but also decreases the bond strength of CsO. In conclusion, CO chemisorbed on Mn site reacted with the adsorbed oxygen to form CO2. We could assume that the catalyst with a larger number of Lewis acid sites may favor for CO chemisorption, leading to a faster oxidation rate. The adsorption of reacting molecules on the catalyst played in an important role in CO oxidation. Therefore, it is necessary to research the adsorption behaviors of reacting molecules on the catalysts. Figure S6 (see Supporting Information) shows CO-TPD profiles of CO on La1-xSrxMnO3catalysts. A significant portion of CO desorbed as a form of CO2 in the processes, so the adsorption amount of CO on the catalyst amounts to the total desorption amounts of both CO2 and CO. In the range of 100–600 °, five to six desorption peaks of CO2 peaks, as well as two to five peaks of CO, can be observed. It is clear that the intensity, number and distribution of peaks over these catalysts are different. The CO or CO2 peak intensity of the Sr-doped catalysts is stronger than that on LaMnO3. The quantitative analysis results (see Table S4, Supporting Information) showed that the total adsorption amount of CO on the parent LaMnO3 is smaller than those on the Srdoped ones. At x ) 2, the total CO adsorption amount is largest. The results indicate that there may be more active centers on La1-xSrxMnO3 catalysts due to the Sr doption. It has been reported that the partially substituting of lowvalence ion for La can result in the formation of anion or oxygen vacancies (23, 32–34). The Sr doption may improve the adsorption properties of reacting molecules on the VOL. 42, NO. 10, 2008 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

3889

catalysts. The sorption is directly related to CO catalytic reaction over La1-xSrxMnO3 nanoparticles. La0.8Sr0.2MnO3 catalyst is expected to have more defects, corresponding to a larger number of Lewis acid sites. The oxygen vacancies can be confirmed by O2-TPD profiles (see Figure S7 and Table S5, Supporting Information). It is generally accepted that the desorption peaks of oxygen at lower temperatures result from the oxygen adsorbed on the catalyst surface (designated R-oxygen), and the desorbed amount of R-oxygen can be referred to surface oxygen vacancies; and the desorption peaks of oxygen at high temperatures can be referred to lattice oxygen or bulk oxygen (related to the reduction of Mn ion) (designated β-oxygen). As shown in Figure S6. There are two oxygen-desorption peaks, respectively. R-oxygen peak at about 480 °C corresponds to the desorption of the oxygen chemically adsorbed on oxygen vacancies. β-oxygen peak in the range of 700-800° C might be attributed to the lattice oxygen associated with the redox of B ions. Compared with the undoped LaMnO3, the desorbed amounts of R-oxygen increase slightly, implying that oxygen vacancies increase, but the peaks areas of β-oxygen varied inconsistently. Based on the above information, we could infer that the oxygen defects increased due to the Sr doption. These results confirmed that the Sr doption increased the adsorpotion of oxygen. In our experiment, La0.8Sr0.2MnO3 had the maximum oxygen vacancies. It may be that x ) 0.2 is the appropriate doping amount. This needs research further. Breysse et al. have proposed a mechanism of recombination radiation based on ThO2 (1). CO is chemisorbed to form the CO+ ion; O2 is dissociated and chemisorbed to form the O- ion. This means that the chemisorbed CO forms a positively ionized surface state (the surface donor state) accompanied by the localized hole, and the chemisorbed oxygen forms a negatively ionized surface-state (the surface acceptor state) accompanied by the localized electron. Surface reaction between these chemisorbed species will occur to form the chemisorbed CO2. The chemisorbed CO2 accompanied by the localized electron and hole is a chemisorption surface state bound to an exciton. The desorption of CO2 is accompanied by annihilation of the exciton, and luminescence is emitted by recombination of electron and hole. Over our perovskite catalysts, it is possible gases from the adsorption and catalysis reaction may be more complicated and different from the former. However, many researchers have proposed that the CL mechanism is luminescence from the excited species produced in catalytic oxidation (40, 41). The exothermal chemical reaction produces energy enough to induce the transition of an electron from its ground-state to an excited electronic state. This electronic transition is often accompanied by vibrational and rotational changes in the molecule. CL is observed when the electronically excited product relaxes to its ground-state with emission of photons (42). Based on the above (40–44), the CL reaction can be represented with the following formula. CO2(ads) + energy f *CO2(ads) f CO2 + hv

(vi)

In our work, it was accepted that CO2 was the luminescence species. While CO molecules were oxidized on the catalyst surface, an amount of energy was released, which would be absorbed by CO2 molecules. As a result, CO2 molecules would jump from ground-state up to electronic excited state (*CO2). While the electronic excited *CO2 molecules decayed to the ground state, a CL was generated. CL intensity is linearly proportional to the produced CO2 concentration. It is clear that CL spectra are closely correlated with the catalytic reaction, in which the conversion of CO into CO2 is directly related to the catalytic properties of the catalysts. The easier the catalytic reaction, the more CO2 molecules were formed. As a result, the CL intensity is stronger. The activity and CL represent the same process. It 3890

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 42, NO. 10, 2008

FIGURE 6. Schematic diagram of CL sensor with multiregions for catalyst selection. is reasonable that CL spectra can be used to quantitatively evaluate a given catalytic reaction. CL Detector for Environmental Combustible Gases and Catalyst Selection. We have found that luminescent efficiencies and spectral shapes of the CL are dependent on the kinds of reactants and catalysts. Even if the same luminescent species could be produced from the different combustible gases on a given catalyst, the amounts of the generated luminescent species by different combustible gases are different. Therefore, the same nanomaterial exhibits different intensities of CL upon exposure to different gases. This enables us to discriminate the kind of combustible gas. The compounds within a given chemical class can be still discriminated effectively because the mechanisms and rates of catalytic reactions are dependent on temperature, which leads to different luminescence efficiencies and spectral shapes (26). Even if similar shape were recorded with the sensor for two gases at one temperature, they may be differentiated by different CL intensities at another temperature. The CL sensor system may be useful for analyzing various environmental toxic gases because of the linear characteristics of CL intensity as a function of gas concentration (40). Besides, a variety of highly active catalysts have been pursued for a series of important industrial reactions. Therefore, it has become crucial in the field of catalysis to develop innovative approaches allowing rapid evaluation of the activity by screening a large diversity of catalyst candidates for a specific catalytic reaction. The CL spectral shapes of a compound are different on different nanomaterials. The good correlation between CL and activity indicates that the CL intensity could be applied for the screening of the catalytic activity. Based on the good correlation between CL and catalysis reactions, we have developed a new CL-based detector with multiregions (Figure 6). The new detector can be used for the selection of multicatalysts at one time. The different catalysts can be coated onto the different regions. CL spectra can be acquired only by adjusting the location of region and the flowing region of reacting gases. The multiregion CL detector is simple and rapid for screening a large number of catalysts. This technique would be potentially applied to explore the new catalysts. As we know, the catalysts are generally selected by testing their activities, in which the complicate and expensive equipment is needed and the test needs a long duration. Especially, while selecting new catalysts from thousands of materials, the procedures are expensive, time-consuming, and laborious. The catalyst selection is still a challenge to researchers. CL determination of catalyst is facile, rapid, and low-cost since CL spectra can be obtained within a few minutes and no complicated and expensive equipment is needed. The CL determination mode can be used as an effective means to select catalysts. Because the perovskite has a high catalytic activity and structurestability, many combustible molecules may be oxidized completely into CO2. It could be expected that such CL sensor of perovskite nanoparticles could potentially be used to justify other high-temperature combustion reaction (e.g., catalytic combustion of methane). We have also observed that the CL intensity is linearly proportional to the concentration of methane in our research range of 0–2.5 vol.% in air with a detection limit below 0.15 vol.% at 600°; and that the CL sensor made of La0.8Sr0.2MnO3 nanoparticles maintained a high stability after running for 48 h at 600°. However, the CL sensor of single-oxide nanoparticles was not suitable for

high-temperature reaction due to their unstable structure. CL mode is useful to select excellent catalysts for catalytic oxidation reactions. The new catalysts could be useful to eliminate effectively the environmentally toxic gases. The CL responses on catalytic nanomaterials provide abundant optical information which motivates the fabrication of chemical sensor arrays. An intensive effort is currently devoted toward the development of high-throughput screening approaches. The CL reaction is of practical importance for detecting environmental pollutants. We intend to apply this technique to the analysis of explosives and to volatile organic compounds. In summary, a long-lifetime CL gas sensor can be fabricated using perovskite nanoparticles due to stable crystal structure and high activity. The good correlation between CL and activity indicates that the CL intensity could be applied for the screening of the catalytic activity. CL mode could be a facile and rapid means for the detection of environmentally deleterious gases and the selection of excellent catalysts from thousands of materials.

Acknowledgments We gratefully acknowledge the research funding supported by National Basic Research Program of China (2007CB613303) and Chinese Postdoctoral Science Foundation (no. 20060390057).

Supporting Information Available CO- and O2-TPD experiments; CL intensities of La0.8Sr0.2MnO3 at different flowing rate, CO concentration, temperature, band-length filters, and particle sizes; BET areas, particle sizes and size distributions, crystal sizes, CO- and O2-TPD results of La1-xSrxMnO3 catalysts. This material is available free of charge via the Internet at http://pubs.acs.org.

Literature Cited (1) Breysse, M.; Claudel, B.; Faure, L.; Guenin, M.; Williams, R. J. J.; Wolkenstein, T. J. Chemiluminescence during the catalysis of carbon monoxide oxidation on a thoria surface. J. Catal. 1976, 45, 137–144. (2) Daniel, M.-C.; Astruc, D. Gold nanoparticles: Assembly, supramolecular chemistry, quantum-size-related properties, and applications toward biology, catalysis, and nanotechnology. Chem. Rev. 2004, 104 (1), 293–346. (3) Burda, C.; Chen, X.; Narayanan, R.; El-Sayed, M. A. Chemistry and properties of nanocrystals of different shapes. Chem. Rev. 2005, 105, 1025–1102. (4) Aslan, K.; Lakowicz, J. R.; Geddes, C. D. Nanogold plasmon resonance-based glucose sensing: 2. Wavelength-ratiometric resonance light scattering. Anal. Chem. 2005, 77, 2007–2014. (5) Elghanian, R.; Storhoff, J. J.; Mucic, R. C.; Letsinger, R. L.; Mirkin, C. A. Selective colorimetric detection of polynucleotides based on the distance-dependent optical properties of gold nanoparticles. Science 1997, 277, 1078–1081. (6) Santra, S.; Dutta, D.; Walter, G. A.; Moudgil, B. M. Fluorescent nanoparticle probes for cancer imaging. Technol. Cancer Res. Treat. 2005, 4, 593–602. (7) El-Sayed, I. H.; Huang, X.; El-Sayed, M. A. Surface plasmon resonance scattering and absorption of anti-EGFR antibody conjugated gold nanoparticles in cancer diagnostics: applications in oral cancer. Nano Lett. 2005, 5, 829–834. (8) McCord, P.; Yau, S. L.; Bard, A. J. Chemiluminescence of anodized and etched silicon: Evidence for a luminescent siloxene-like layer on porous silicon. Science 1992, 257, 68–69. (9) Zhang, Z.; Cui, H.; Shi, M. Chemiluminescence accompanied by the reaction of gold nanoparticles with potassium permanganate. Phys. Chem. Chem. Phys. 2006, 8, 1017–1021. (10) Cui, H.; Zhang, Z.; Shi, M. Chemiluminescent reactions induced by gold nanoparticles. J. Phys. Chem. B. 2005, 109, 3099–3103. (11) Cui, H.; Zhang, Z.; Shi, M.; Xu, Y.; Wu, Y. Light emission of gold nanoparticles induced by the reaction of bis(2,4,6-trichlorophenyl) oxalate and hydrogen peroxide. Anal. Chem. 2005, 77, 6402– 6406. (12) Poznyak, S. K.; Talapin, D. V.; Shevchenko, E. V.; Weller, H. Quantum dot chemiluminescence. Nano Lett. 2004, 4, 693– 698.

(13) Powe, A. M.; Fletcher, K. A.; Luce, N. N.; Lowry, M.; Neal, S.; McCarroll, M. E.; Oldham, P. B.; McGown, L. B.; Warner, I. M. Molecular fluorescence, phosphorescence, and chemiluminescence spectrometry. Anal. Chem. 2004, 76, 4614–4634. (14) Zhu, Y. F.; Shi, J.; Zhang, Z.; Zhang, C.; Zhang, X. R. Development of a gas sensor utilizing chemiluminescence on nanosized titanium dioxide. Anal. Chem. 2002, 74, 120–124. (15) Rakow, N. A.; Suslick, K. S. A colorimetric sensor array for odour visualization. Nature 2000, 406, 710–714. (16) Nakagawa, M.; Yamamoto, I.; Yamashita, N. Am al. Sci. 1998, 14, 209–215. (17) Okabayashi, T.; Fujimoto, T.; Yamamoto, I.; Utsunomiya, K.; Wada, T.; Yamashita, Y.; Yamashita, N.; Nakagawa, M. A new method for chemiluminescence sensor. Sens. Acatuators, B 2000, 64, 54–58. (18) Zhang, Z. Y.; Zhang, C.; Zhang, X. R. Development of a chemiluminescence ethanol sensor based on nanosized ZrO2. Analyst. 2002, 127, 792–796. (19) Shi, J. J.; Li, J. J.; Zhu, Y. F.; Fan, W.; Zhang, X. R. Nanosized SrCO3-based chemiluminescence sensor for ethanol. Anal. Chim. Acta 2002, 466, 69–78. (20) Rao, Z. M.; Shi, J. J.; Zhang, X. R. Study of cataluminescence characteristics of NH3 on the surface of nanosized materials. Acta. Chim. Sin. 2002, 60, 1668–1671. (21) Zhou, K. W.; Zhang, X. R. Determination of butanone in workshop air utilizing chemiluminescance on nanosized materials. Chin. J. Anal. Chem. 2004, 32 (1), 25–28. (22) Seiyama, T. Total oxidation of hydrocarcons on perovskite oxides. Catal. Rev. Sci. Eng. 1992, 34 (4), 281–300. (23) Corralles, T.; Peinado, C.; Allen, N. S.; Edge, M.; Sandoval, G.; Catalina, F. J. Synthesis and photochromism of polymer-bound phenoxyquinone derivatives. Photochem. Photobiol. A 2003, 156, 151–157. (24) Ye, Q.; Gao, Qi.; Zhang, X. R.; Xu, B. Q. Cataluminescent characteristics of catalytic ethanol oxidation over Ce1-xMnxO2 catalysts. Acta. Chim. Sin. 2006, 64 (8), 751–755. (25) Baythoun, M. S.G.; Sale, F. R. Production of strontiumsubstituted lanthanum mangantite perovskite powder by the amorphous citrate process. J. Mater Sci. 1982, 17, 2757–2769. (26) Na, N.; Zhang, S.; Wang, S.; Zhang, X. A catalytic nanomaterialbased optical chemo-sensor array. J. Am. Chem. Soc. 2006, 128, 14420–14421. (27) Sun, Z.; Yuan, H.; Liu, Z.; Han, B.; Zhang, X. A highly efficient chemical sensor material for H2S: R-Fe2O3 nanotubes fabricated using carbon nanotube templates. Adv. Mater. 2005, 17, 2993– 2997. (28) Ye, Q.; Gao, Q.; Zhang, X.; Xu, B. Cataluminescence and catalytic reactions of ethanol oxidation over nanosized Ce1-xZrxO2 (0 E × E 1) catalysts. Catal. Commun. 2006, 7, 589–592. (29) Lv, Y.; Zhang, S.; Liu, G.; Huang, M.; Zhang, X. Development of a detector for liquid chromatography based on aerosol chemiluminescence on porous alumina. Anal. Chem. 2005, 77, 1518– 1525. (30) Huang, G.; Lv, Y.; Zhang, S.; Yang, C.; Zhang, X. Development of an aerosol chemiluminescent detector coupled to capillary electrophoresis for saccharide analysis. Anal. Chem. 2005, 77, 7356–7365. (31) Liu, G.; Zhu, Y.; Zhang, X.; Xu, B. Chemiluminescence determination of chlorinated volatile organic compounds by conversion on nanometer TiO2. Anal. Chem. 2002, 74, 6279–84. (32) Voorhoeve, R. J. H.; Johnson, D. W.; Remeika, J. P.; Gallagher, P. K. Perovskite oxides: Materials science in catalysis. Science. 1977, 195, 827–833. (33) Royera, S.; Alamdarib, H.; Duprezc, D.; Kaliaguinea, S. Oxygen storage capacity of La1-xA′xBO3 perovskites (with A′ ) Sr, Ce; B ) Co, Mn)-relation with catalytic activity in the CH4 oxidation reaction. Appl. Catal., B 2005, 58, 273–288. (34) Barnard, K. R.; Forger, K.; Turney, T. W.; Williams, R. D. Lanthanum cobalt oxide oxidation catalysts derived from mixed hydroxide precursors. J. Catal. 1990, 125, 265–275. (35) Voorhoeve, R. J. H.; Remeika, J. P.; Tribble, L. E. Ann. N. Y. Acad. Sci. 1976, 272, 3–5. (36) Tascon, J. M. D.; Fierro, J. L. G.; Tejuca, L. G. Phys. Chem. 1981, 124, 249–253. (37) George, S.; Viswanathan, B. catalytic oxidation of carbon monoxide on LnCoO3 perovskite oxides. J. Colloid Interface Sci. 1983, 95 (2), 322–326. (38) Om, P.; Ganguly, P.; Rama Rao, G.; Rao, C. N. R.; Rajoria, D. S.; Bhide, V. G. Mater. Res. Bull. 1974, 9, 1173–1177. VOL. 42, NO. 10, 2008 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

3891

(39) Zhou, H.; Shen, Y. F.; Wang, J. Y.; Chen, X.; O’Young, C. L.; Suib, S. L. Studies of decomposition of H2O2 over manganese oxide octahedral molecular sieve materials. J. Catal. 1998, 176, 321– 328. (40) Nakagawa, M.; Yamashita, N. Cataluminescence-based gas sensors. Springer Ser. Chem. Sens. Biosens. 2005, 3, 93–132. (41) Wang, X.; Na, N.; Zhang, S.; Wu, Y.; Zhang, X. Rapid screening of gold catalysts by chemiluminescence-based array imaging. J. Am. Chem. Soc. 2007, 129, 6062–6063. (42) Huang, X. Y.; Ren, J. C. Trends Anal. Chem. 2006, 25, 155–161.

3892

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 42, NO. 10, 2008

(43) Shi, J. J.; Zhu, Y. F.; Zhang, X. R.; Baeyens, W. R. G.; GarciaCampana, A. M. Recent developments in nanomaterial optical sensors. Trends Anal. Chem. 2004, 23, 351–360. (44) Nakao, K.; Ito, S.; Tomishige, K.; Kunimori, K. IR chemiluminescence probe of the vibrational energy distribution of CO2 formed during steady-state CO oxidation on Pt(111) and Pt(110) surfaces. J. Phys. Chem. B 2005, 109, 24002–24007.

ES702845Z