Research article Received: 29 November 2014,
Revised: 31 March 2015,
Accepted: 4 May 2015
Published online in Wiley Online Library
(wileyonlinelibrary.com) DOI 10.1002/bio.2954
Potassium permanganate–glutaraldehyde chemiluminescence system catalyzed by gold nanoprisms toward selective determination of fluoride Jafar Abolhasani,* Javad Hassanzadeh and Ebrahim Ghorbani-Kalhor ABSTRACT: Gold and silver nanoparticles (NPs) are shown to exert a positive effect on the chemiluminescence (CL) reaction of permanganate aldehydes. Interestingly, between various shapes examined, Au nanoprisms have the highest beneficial effect. This effect is even more notable in the presence of sodium dodecyl sulfate (SDS) surfactant. UV-vis spectra and transmission electron microscopy were used to characterize the NP shapes and sizes. Furthermore, it was observed that iron(III) ions can slightly increase CL emission of this system. This intensification is very effective in the presence of fluoride ions (F–). These observations form the basis of the method for the high sensitive determination of F– in the 6–1200 nmol L–1 concentration range, with a detection limit of 2.1 nmol L–1. The proposed method has good precision and was satisfactorily used in the selective determination of low concentrations of fluoride ion in real samples. Copyright © 2015 John Wiley & Sons, Ltd. Keywords: chemiluminescence; permanganate; aldehydes; gold nanoprisms; fluoride
Introduction
Luminescence 2015
* Correspondence to: Jafar Abolhasani, Department of Chemistry, Tabriz Branch, Islamic Azad University, Tabriz, Iran. E-mail: [email protected] Department of Chemistry, Tabriz Branch, Islamic Azad University, Tabriz, Iran Abbreviations: CMC, critical micellar concentration; LMW, low-molecularweight; RSD, relative standard deviation; SDS, sodium dodecyl sulfate.
Copyright © 2015 John Wiley & Sons, Ltd.
1
Anions such as fluoride are very important because they play key roles in the human life (1–3). For example, fluoride (F–) at about 1 mg L 1 concentration has many favorable effects on the body, including preventing tooth decay and treating osteoporosis (3,4). However, higher levels of F– in water and food are of health concerns (5,6). Dental and skeletal fluorosis has been reported in people who consume drinking water containing high levels of fluoride (5 mg L 1) (6). In dental fluorosis, the natural integrity of the enamel is decreased and leads to tooth breakage. In skeletal fluorosis, accumulation of fluoride in the bones will result in abnormal bone formation (7). Extreme levels of fluoride can also cause various abnormalities in human health e.g. osteosarcoma, lower IQ level, inhibition of neurotransmitter biosynthesis in fetuses etc. (8,9). For adults, the lethal level is 0.20–0.35 g F– per kg body weight (10). But, more than one billion people in various countries such as China (11), Sri Lanka (12), and Estonia (13) drink fluoridated water in concentrations that exceeded the guidelines recommended by the World Health Organization (14). A simple, portable and fast method for the determination of fluoride ions will be useful for follow these ions in drinking water or other in environmental samples to which humans are exposure. Several methods have been reported for quantification of F– in drinking water, seawater, air, soils, foods, urine, serum, etc. For examples, methods such as titration, ion-selective electrodes (ISEs) (15), colorimetry (16–19), infra-red spectrometry (20), spectrophotometric (21–23), fluorescence (17,18,24–26), and chromatographic methods have been tested (27,28). Many of these methods require a trained analyst and are also expensive and time consuming. In contrast, chemiluminescence techniques provide rapid, simple and sensitive methods for fluoride determination.
For chemiluminescence (CL)-based analyses, light emission produced by oxidation of a substrate has attracted much attention in the analytical field due to their inherent sensitivity, simplicity, rapidity, relatively simple and inexpensive instrumentation. Among the wide number of CL systems, potassium permanganate (KMnO4) is the most common oxidants applied for CL emission generation. Hindson, Barnett and also Adcock et al. have presented extensive reviews indicating the wide analytical applications of this CL reagent (29,30). In many of these systems, low-molecularweight aldehydes were exploited to produce or improve CL emission (30–34). Recently, noble metals, especially gold and silver nanoparticles (NPs) have appeared in CL systems mainly as catalysts, because of their unique physical and chemical properties. Several reviews have described the positive roll of NPs on CL reactions (35,36). The most commonly NPs used in CL are spherical, and there are few reports about NPs with specific shapes. However, something is clear, the catalytic properties of NPs depend on their size and shape. It has been demonstrated that anisotropic NPs have higher catalytic effect than spherical ones in some CL reaction (37,38). Such studies have rarely been conducted, especially for the permanganate CL systems. In our previous studies, we found that the KMnO4–HCHO reaction can lead to a weak CL emission that can be catalyzed with Au and Ag nanoparticles (34). Intense emission was obtained
J. Abolhasani et al. and applied for sensitive analytical aims. The present study deals with the enhancing effect of Au and Ag NPs with different shapes on the CL reaction of KMnO4–glutaraldehyde. It is indicated that CL emission in the presence of Au prism-like NPs is the highest. Subsequently, it was found that the CL signal of the NPs catalyzed reactions could be affected in the presence of iron(III) ions. The simultaneously presence of both F– and Fe3+ made a remarkable increase in CL intensity. This enhancing effect was exploited to the selective and precise determination of F– in real samples. The method is simple, rapid and more reliable compared with other reported techniques.
Experimental Materials Most materials applied in this study were of analytical grade and were obtained from Merck (Darmstadt, Germany; www.merckchemicals.com). They were used without further purification. Chloroauric acid (HAuCl4) was purchased from Alfa Aesar (Karlsruhe, Germany; www.alfa.com). Hexadecyl (cetyl) trimethylammonium bromide (CTAB) was purchased from Fluka (Taufkrichen, Germany; www.sigmaaldrich.com). A stock standard solution of 1 mmol L 1 potassium fluoride was prepared by dissolving the appropriate amount of KF powder (Merck) in deionized water. Double-distilled deionized water (obtained from Kasra Co. Tabriz, Iran) was used throughout.
0.1 mol L–1 ZnSO4 solution, and centrifuged at 4000 rpm for 10 min. The clear supernatant was diluted to the appropriate concentration and analyzed according to the general procedure. Water samples were used without any pretreatment according to the general procedure.
General procedure for chemiluminescence detection Chemiluminescence measurements were done in a 3 mL tube in the batch condition. 300 μL of glutaraldehyde (Glu, 12% v/v), 300 μL of SDS (0.03 mol L–1), 300 μL of H2SO4 (0.5 mol L–1), 30 μL of Fe3+ solution (0.001 mol L–1), and 600 μL of synthesized Au nanoprisms (NPrs) solution were added into the cell. Then an appropriate volume of sample or standard fluoride solution was added and the final volume was made up to 2.7 mL with deionized water. After that, 300 μL of KMnO4 (0.003 mol L–1) was injected and the CL signal was monitored versus time automatically. Maximum CL intensity was used as the analytical signal.
Apparatus The CL signals were detected by LUMAT LB 9507 chemiluminometer (Berthold; www.berthold.com). The CL spectra were measured on a Shimadzu RF-5301 PC spectrofluorimeter using a flow mode with the excitation light source being turned off. Ultraviolet–visible (UV–vis) spectra were recorded on a UV–vis UV-1800 spectrophotometer (Shimadzu). The size and shape of NPs were characterized with transmission electron microscopy (TEM, Leo 906, Zeiss, Germany). Preparation of Au and Ag nanoparticles in different shapes Synthesis of Au (8, 22, and 32 nm) and Ag (6, 18, and 38 nm) spherical nanoparticles were done according to our previous work (39). The resulting nanoparticles were characterized by TEM and statistical analysis revealed that the average diameters of the gold nanoparticles were about 8 ± 2.2, 22 ± 1.8, and 32 ± 2.1 nm and of the silver nanoparticles was 6 ± 1.2, 18 ± 1.4 and 38 ± 1 nm. Rod-like and triangular Ag NPs were prepared following published procedures (40,41). TEM images revealed that the average aspect ratios of the Ag nanorods were about 1.9 ± 0.2, and 3.2 ± 0.4 nm and average size of the Ag nanoprisms was 18 ± 1.5 nm. Prism and rod-like Au NPs in two different aspect ratios were synthesized according to Li et al. (42) and Nikoobakht et al. (43), respectively. TEM images revealed that the average aspect ratios of the Au nanorods were about 2.2 ± 0.4, and 3.4 ± 0.5 nm and average size of the Au nanoprisms was 32 ± 1.9 nm.
2
Sample preparation Human urine samples containing F were obtained from some volunteers who used fluoride-containing toothpastes. One mL of this sample was added to 1 mL of 0.1 mol L–1 Ba(OH)2 and
wileyonlinelibrary.com/journal/luminescence
Figure 1. CL profiles of Glu–KMnO4 (a, c) and Glu–KMnO4–SDS (b, d) systems in the 1 absence (a, b) and presence of Au NPrs (c, d). (Condition: 0.05 mol L H2SO4, 1 1 3 mmol L SDS, 1.2% v/v Glu, 0.3 mmol L KMnO4 and 600 μL Au NPrs.)
Table 1. Effects of some NPs and precursor reagents of Au NPrs on CL emission Added reagents – HAuCl4 Sodium citrate Sodium citrate + HAuCl4 Au NPs (average size: 8, 22, and 32 nm) Ag NPs (average size: 6, 18, and 38 nm) Au triangular NPs (average size: 32 nm) Ag triangular NPs (average size: 18 nm) Au NRs (average aspect ratio: 1.9, 3.2) Ag NRs (average aspect ratio: 2.2, 3.4)
CL intensity (a.u.) 132 139 134 145 168, 202, 241 124, 186, 209 374 197 215, 312 101, 154
NR, nanorod.
Copyright © 2015 John Wiley & Sons, Ltd.
Luminescence 2015
Gold nanoprisms assisted CL toward determination of fluoride ion
Results and discussion Chemiluminescence reaction of potassium permanganate-aldehydes in the presence of Au and Ag nanoparticles Oxidation of low-molecular-weight (LMW) aldehydes by potassium permanganate (KMnO4) in acidic condition has been frequently
applied to produce CL emission for quantification of many compounds (29,30). Experimental results showed that Glu produce the best emission among some examined LMW aldehydes. However, this response is so weak that can be applied for analytical aims (Fig. 1a). In a previous study, we indicated that the obtained CL emission from the KMnO4–formaldehyde reaction (in SDS micellar medium) is greatly increased in the presence of Ag, Au NPs (34). In this way, the effect of these NPs was investigated fir CL intensity of the Glu system in their various sizes and shapes. As shown in Table 1, the most examined NPs have a good enhancing effect on CL intensity. However, the enhancing effect of Au prismshaped NPs (NPrs) is markedly better (Table 1). Figure 2 shows TEM image of the synthesized Au NPrs that confirms their correct synthesis and shapes. Also, slight enhancement effects were found (Table 1) for precursor solutions used for the preparation of NPrs, showing that the observed catalytic effect is due to the applied nanoparticles. A schematic display of the CL system is indicated in Scheme 1. The kinetic curve of the CL reaction was obtained in the absence and presence of Au NPrs (Fig. 1). It shows that the CL intensity reached a maximum value at 5 and 1 s after injection of permanganate solution, respectively. This result denotes the catalytic effect of NPrs with increasing effect on reaction rate. Optimization of reaction conditions
Figure 2. (a) UV–vis spectra; and (b) TEM image of applied gold nanoprisms.
Parameters influencing CL emission of the KMnO4–Glu–Au NPrs system were investigated to establish the optimal conditions for the CL reaction. The effects of type and concentration of surfactants on the CL system was studied by applying different kinds of surfactants (SDS, AOT and CTAB) at various concentrations. The presence of a surfactant in a CL system can influence the CL emission in many cases. It is obvious that at concentrations near the critical micellar concentration (CMC), the surfactant molecules tend to form micelles in aqueous solution, providing a protective environment for the Mn(II) excited state and a better contact between interacting species that leads to a significant increase in the CL quantum yield. As can be seen in Fig. 3(a), maximum CL intensity was obtained at 0.003 mol L 1 SDS concentration, which is near its CMC value (3.10 × 10 3 mol L 1 in this condition as obtained by conductometric measurements (44)). Therefore, the increasing effect of SDS in the CL system can be explained by producing micelles in CL reaction media. To study the type of acid and concentration on CL emission intensity, several acids (including HCl, H2SO4, HNO3 and H3PO4) were
3
3+
Scheme 1. KMnO4–Glu–Au NPrs CL system enhanced with Fe
Luminescence 2015
Copyright © 2015 John Wiley & Sons, Ltd.
–
and F ions.
wileyonlinelibrary.com/journal/luminescence
J. Abolhasani et al.
Figure 4. (a) CL spectra of (1) the Glu–KMnO4, and (2, 3) the Glu–KMnO4–Au NPrs sys–1 3+ 1 tems in the absence (2) and presence of 50 μmol L Fe (3) (condition: 0.05 mol L 1 1 H2SO4, 3 mmol L SDS, 1.2% v/v Glu, 0.3 mmol L KMnO4 and 600 μL Au NPr)]. (b) TEM image of reaction solution includes gold nanoprisms after CL emission.
1
4
Figure 3. Effect of (a) type and concentrations of some surfactants (0.05 mol L 1 H2SO4, 0.8% v/v Glu, 0.25 mmol L KMnO4 and 800 μL Au NPrs). (b) Type and concen1 1 trations of some acids (3 mmol L SDS, 0.8% v/v Glu, 0.25 mmol L KMnO4 and 1 1 800 μL Au NPrs). (c) Concentration of Glu (0.05 mol L H2SO4, 3 mmol L SDS, 1 1 0.25 mmol L KMnO4 and 800 μL Au NPrs). (d) Amount of Au NPrs (0.05 mol L 1 1 H2SO4, 3 mmol L SDS, 1.2% v/v Glu and 0.25 mmol L KMnO4) and (e) concentra1 1 tion of KMnO4 (0.05 mol L H2SO4, 3 mmol L SDS, 1.2% v/v Glu and 600 μL Au NPrs) 1 3+ 1 – on the CL intensity [10 μmol L Fe , 30 nmol L F ].
wileyonlinelibrary.com/journal/luminescence
applied. This parameter is very important in permanganate CL systems (29). Figure 3(b) shows that the best CL response was obtained for 0.05 mol L 1 H2SO4. Higher concentrations may demolish NPs, leading to a decrease in CL emission. The effect of Glu concentration was also examined and optimal CL intensity was obtained at 1.2% (v/v) Glu (Fig. 3c). High concentrations of Glu may increase the amount of Mn(III) produced and so the concentration of Mn(II)*, which substantially enhanced the CL intensity. At concentrations greater than 1.2% (v/v), Glu does not affect CL emission, markedly. The concentration of Au NPrs solution has a remarkable effect on the CL response of the considered system. Therefore, different amounts of prepared NPrs stock solution were added to the reaction environment. Figure 3(d) shows the result that reveals the maximum sensitivity obtain when 600 μL of NPrs solution were used. Finally, the effect of permanganate concentration was investigated over the range of 5.0 × 10 5 to 5.0 × 10 4 mol L 1. As shown in Fig. 3(e), the CL signal increased up to 0.3 mmol L 1 and then was fixed at the higher concentrations. As shown in a previous study (34), lower KMnO4 concentrations lead to a decrease in the number of excited intermediates and the CL intensity is diminished.
Copyright © 2015 John Wiley & Sons, Ltd.
Luminescence 2015
Gold nanoprisms assisted CL toward determination of fluoride ion Possible mechanism for chemiluminescence reaction
Figure 5. CL profiles of (a, b) the Glu–KMnO4 systems in the absence (a) and pres3+ – ence of Fe and F , (c-e) the Glu–KMnO4–SDS–Au NPrs system in the (c) absence 3+ 3+ – 1 and presence of (d) Fe and (e) Fe and F (Condition: 0.05 mol L H2SO4, 1 1 3 mmol L SDS, 1.2% v/v Glu, 0.3 mmol L KMnO4 and 600 μL Au NPrs).
Various studies have been carried out to discover the possible mechanism for CL emission from the KMnO4–aldehyde reaction, (29). In our previous study, we investigated the acidic permanganate–formaldehyde system mechanism and found that the oxidation of a substrate by Mn(VII) generates Mn(III), which then produces Mn(II)* in the excited state (emission at about 700 nm) (34). Researchers have confirmed this mechanism, they related red CL emission from permanganate CL reactions to the relaxation of the Mn(II) 4 T1 excited state to the 6A1 ground state (29,30). The CL emission spectra for the Glu–KMnO4–Au NPrs system (Fig. 4a) shows the characteristic red emission (λem = 708 nm) of permanganate CL. Thus, emitting species in this condition are the same and NPrs only act as catalysts to enhance the net reaction rate. The TEM image obtained after the reaction (Fig. 4b) showed no nanoparticles in solution, showing that the nanoparticles oxidized with MnO–4. Thus, the oxidation of Au NPrs by permanganate was investigated by UV–vis spectra in acidic conditions (34). The oxidation rate is low and cannot be responsible for enhancement of the CL emission. Conversely, Glu and SDS do not affect NPrs structures. Therefore, it can be said that NPrs do not participate directly in the CL reaction and just act as catalysts. As illustrated in a previous study, NPs may facilitate the oxidative cleavage of permanganate ester and accelerate Mn(III) generation leading to a higher CL intensity (34). Triangular Au NPs have higher surface-to-volume ratios and more active surface sites compared with spherical Au NPs, which facilitated the active intermediates generation on the surface of Au NPrs (45). Therefore, Au NPrs displayed greater catalytic activity than spherical ones. Moreover, the triangular Au NPs are more stable at a wider pH range (45), and thus they can be relatively unchanging in acidic media applied to the CL reaction, leading to their high catalytic activity.
Analytical application of the CL system
3+
Figure 6. Effect of the concentration of Fe on CL intensity under optimum condition.
It was found in the initial experiments that the addition of low amounts of iron(III) ions (Fe3+) to the presented CL system increased the CL intensity (Fig. 5). This increase is linear in the concentration range 0.01–15 μg mL–1 (0.18–268 μmol L–1) of Fe3+
5
3+
Figure 7. CL profiles of the KMnO4–Glu–SDS–Au NPrs–Fe calibration graph.
Luminescence 2015
–1
–
system in the presence of various concentrations (nmol L ) of F at optimum conditions and (inset) corresponding
Copyright © 2015 John Wiley & Sons, Ltd.
wileyonlinelibrary.com/journal/luminescence
J. Abolhasani et al. with a detection limit of 2.4 ng mL–1 (0.043 μmol L–1) (Fig. 6). Conversely, simultaneous existence of both Fe3+ and F– ions led to a significant intensification that was proportional to each ion concentration. These effects are not clearly understood, but it may be related to the catalyzing effect of Fe3+ and its complex with F–. These ions can increase CL intensity of the system in the absence of Au NPrs (Fig. 5), and also, they do not affect NPrs absorption spectra. Thus, the enhancing effect of Fe3+ and F– cannot be related to these NPs. Conversely, it is well known that metal ions can interact with aldehydes (46). Fe3+ or its complexes may interact with Glu or oxidation products, and this interaction may be responsible for their catalyzing effect. This phenomenon was exploited to develop a sensitive method for the determination of the F– ion. In the presence of 10 μmol L–1 Fe3+, the wide linear dynamic range was obtained for F– determination. Under the optimum conditions described and in the presence of 10 μmol L–1 Fe3+, the linear range of the calibration graph is 6–1200 nmol L 1 with a detection limit (3 sec) of 2.1 nmol L 1. The equation for the regression line was ΔI = 0.879C + 0.417(R2 = 0.9997), where ΔI = I0 – I is the difference between the CL intensity in the absence (I0) and presence of F– (I), and C is the concentration of F– ion in nmol L 1 (Fig. 7). The relative standard deviation (RSD) was calculated to be 0.81, 0.40, and 0.31% for five determinations of 72, 300 and 840 nmol L 1 F–, respectively. Results demonstrated tgat this CL system has good precision, relatively high sensitivity and a wide linear range. A comparison between the presented method and some other reported analytical methods for the determination of F– ion is summarized in Table 2. The characteristics of presented method are comparable with most other methods. Moreover, it has potential applications in the determination of some other compounds.
from these species in F– determination. In addition, Fig. 8 show the response of our CL system to different ions. As can be seen, the method was selective for the F– ion.
Analysis of real samples The considered method was easily applied to the determination of F– ion in urine and environmental water. The preparation step was
Study of interferences In order to test the interference effect of some potentially interfering substances, increasing amounts of these species were added into a standard solution of 50 nmol L 1 F– ion. The tolerable concentration ratios for interferences at a relative error of
Revised: 31 March 2015,
Accepted: 4 May 2015
Published online in Wiley Online Library
(wileyonlinelibrary.com) DOI 10.1002/bio.2954
Potassium permanganate–glutaraldehyde chemiluminescence system catalyzed by gold nanoprisms toward selective determination of fluoride Jafar Abolhasani,* Javad Hassanzadeh and Ebrahim Ghorbani-Kalhor ABSTRACT: Gold and silver nanoparticles (NPs) are shown to exert a positive effect on the chemiluminescence (CL) reaction of permanganate aldehydes. Interestingly, between various shapes examined, Au nanoprisms have the highest beneficial effect. This effect is even more notable in the presence of sodium dodecyl sulfate (SDS) surfactant. UV-vis spectra and transmission electron microscopy were used to characterize the NP shapes and sizes. Furthermore, it was observed that iron(III) ions can slightly increase CL emission of this system. This intensification is very effective in the presence of fluoride ions (F–). These observations form the basis of the method for the high sensitive determination of F– in the 6–1200 nmol L–1 concentration range, with a detection limit of 2.1 nmol L–1. The proposed method has good precision and was satisfactorily used in the selective determination of low concentrations of fluoride ion in real samples. Copyright © 2015 John Wiley & Sons, Ltd. Keywords: chemiluminescence; permanganate; aldehydes; gold nanoprisms; fluoride
Introduction
Luminescence 2015
* Correspondence to: Jafar Abolhasani, Department of Chemistry, Tabriz Branch, Islamic Azad University, Tabriz, Iran. E-mail: [email protected] Department of Chemistry, Tabriz Branch, Islamic Azad University, Tabriz, Iran Abbreviations: CMC, critical micellar concentration; LMW, low-molecularweight; RSD, relative standard deviation; SDS, sodium dodecyl sulfate.
Copyright © 2015 John Wiley & Sons, Ltd.
1
Anions such as fluoride are very important because they play key roles in the human life (1–3). For example, fluoride (F–) at about 1 mg L 1 concentration has many favorable effects on the body, including preventing tooth decay and treating osteoporosis (3,4). However, higher levels of F– in water and food are of health concerns (5,6). Dental and skeletal fluorosis has been reported in people who consume drinking water containing high levels of fluoride (5 mg L 1) (6). In dental fluorosis, the natural integrity of the enamel is decreased and leads to tooth breakage. In skeletal fluorosis, accumulation of fluoride in the bones will result in abnormal bone formation (7). Extreme levels of fluoride can also cause various abnormalities in human health e.g. osteosarcoma, lower IQ level, inhibition of neurotransmitter biosynthesis in fetuses etc. (8,9). For adults, the lethal level is 0.20–0.35 g F– per kg body weight (10). But, more than one billion people in various countries such as China (11), Sri Lanka (12), and Estonia (13) drink fluoridated water in concentrations that exceeded the guidelines recommended by the World Health Organization (14). A simple, portable and fast method for the determination of fluoride ions will be useful for follow these ions in drinking water or other in environmental samples to which humans are exposure. Several methods have been reported for quantification of F– in drinking water, seawater, air, soils, foods, urine, serum, etc. For examples, methods such as titration, ion-selective electrodes (ISEs) (15), colorimetry (16–19), infra-red spectrometry (20), spectrophotometric (21–23), fluorescence (17,18,24–26), and chromatographic methods have been tested (27,28). Many of these methods require a trained analyst and are also expensive and time consuming. In contrast, chemiluminescence techniques provide rapid, simple and sensitive methods for fluoride determination.
For chemiluminescence (CL)-based analyses, light emission produced by oxidation of a substrate has attracted much attention in the analytical field due to their inherent sensitivity, simplicity, rapidity, relatively simple and inexpensive instrumentation. Among the wide number of CL systems, potassium permanganate (KMnO4) is the most common oxidants applied for CL emission generation. Hindson, Barnett and also Adcock et al. have presented extensive reviews indicating the wide analytical applications of this CL reagent (29,30). In many of these systems, low-molecularweight aldehydes were exploited to produce or improve CL emission (30–34). Recently, noble metals, especially gold and silver nanoparticles (NPs) have appeared in CL systems mainly as catalysts, because of their unique physical and chemical properties. Several reviews have described the positive roll of NPs on CL reactions (35,36). The most commonly NPs used in CL are spherical, and there are few reports about NPs with specific shapes. However, something is clear, the catalytic properties of NPs depend on their size and shape. It has been demonstrated that anisotropic NPs have higher catalytic effect than spherical ones in some CL reaction (37,38). Such studies have rarely been conducted, especially for the permanganate CL systems. In our previous studies, we found that the KMnO4–HCHO reaction can lead to a weak CL emission that can be catalyzed with Au and Ag nanoparticles (34). Intense emission was obtained
J. Abolhasani et al. and applied for sensitive analytical aims. The present study deals with the enhancing effect of Au and Ag NPs with different shapes on the CL reaction of KMnO4–glutaraldehyde. It is indicated that CL emission in the presence of Au prism-like NPs is the highest. Subsequently, it was found that the CL signal of the NPs catalyzed reactions could be affected in the presence of iron(III) ions. The simultaneously presence of both F– and Fe3+ made a remarkable increase in CL intensity. This enhancing effect was exploited to the selective and precise determination of F– in real samples. The method is simple, rapid and more reliable compared with other reported techniques.
Experimental Materials Most materials applied in this study were of analytical grade and were obtained from Merck (Darmstadt, Germany; www.merckchemicals.com). They were used without further purification. Chloroauric acid (HAuCl4) was purchased from Alfa Aesar (Karlsruhe, Germany; www.alfa.com). Hexadecyl (cetyl) trimethylammonium bromide (CTAB) was purchased from Fluka (Taufkrichen, Germany; www.sigmaaldrich.com). A stock standard solution of 1 mmol L 1 potassium fluoride was prepared by dissolving the appropriate amount of KF powder (Merck) in deionized water. Double-distilled deionized water (obtained from Kasra Co. Tabriz, Iran) was used throughout.
0.1 mol L–1 ZnSO4 solution, and centrifuged at 4000 rpm for 10 min. The clear supernatant was diluted to the appropriate concentration and analyzed according to the general procedure. Water samples were used without any pretreatment according to the general procedure.
General procedure for chemiluminescence detection Chemiluminescence measurements were done in a 3 mL tube in the batch condition. 300 μL of glutaraldehyde (Glu, 12% v/v), 300 μL of SDS (0.03 mol L–1), 300 μL of H2SO4 (0.5 mol L–1), 30 μL of Fe3+ solution (0.001 mol L–1), and 600 μL of synthesized Au nanoprisms (NPrs) solution were added into the cell. Then an appropriate volume of sample or standard fluoride solution was added and the final volume was made up to 2.7 mL with deionized water. After that, 300 μL of KMnO4 (0.003 mol L–1) was injected and the CL signal was monitored versus time automatically. Maximum CL intensity was used as the analytical signal.
Apparatus The CL signals were detected by LUMAT LB 9507 chemiluminometer (Berthold; www.berthold.com). The CL spectra were measured on a Shimadzu RF-5301 PC spectrofluorimeter using a flow mode with the excitation light source being turned off. Ultraviolet–visible (UV–vis) spectra were recorded on a UV–vis UV-1800 spectrophotometer (Shimadzu). The size and shape of NPs were characterized with transmission electron microscopy (TEM, Leo 906, Zeiss, Germany). Preparation of Au and Ag nanoparticles in different shapes Synthesis of Au (8, 22, and 32 nm) and Ag (6, 18, and 38 nm) spherical nanoparticles were done according to our previous work (39). The resulting nanoparticles were characterized by TEM and statistical analysis revealed that the average diameters of the gold nanoparticles were about 8 ± 2.2, 22 ± 1.8, and 32 ± 2.1 nm and of the silver nanoparticles was 6 ± 1.2, 18 ± 1.4 and 38 ± 1 nm. Rod-like and triangular Ag NPs were prepared following published procedures (40,41). TEM images revealed that the average aspect ratios of the Ag nanorods were about 1.9 ± 0.2, and 3.2 ± 0.4 nm and average size of the Ag nanoprisms was 18 ± 1.5 nm. Prism and rod-like Au NPs in two different aspect ratios were synthesized according to Li et al. (42) and Nikoobakht et al. (43), respectively. TEM images revealed that the average aspect ratios of the Au nanorods were about 2.2 ± 0.4, and 3.4 ± 0.5 nm and average size of the Au nanoprisms was 32 ± 1.9 nm.
2
Sample preparation Human urine samples containing F were obtained from some volunteers who used fluoride-containing toothpastes. One mL of this sample was added to 1 mL of 0.1 mol L–1 Ba(OH)2 and
wileyonlinelibrary.com/journal/luminescence
Figure 1. CL profiles of Glu–KMnO4 (a, c) and Glu–KMnO4–SDS (b, d) systems in the 1 absence (a, b) and presence of Au NPrs (c, d). (Condition: 0.05 mol L H2SO4, 1 1 3 mmol L SDS, 1.2% v/v Glu, 0.3 mmol L KMnO4 and 600 μL Au NPrs.)
Table 1. Effects of some NPs and precursor reagents of Au NPrs on CL emission Added reagents – HAuCl4 Sodium citrate Sodium citrate + HAuCl4 Au NPs (average size: 8, 22, and 32 nm) Ag NPs (average size: 6, 18, and 38 nm) Au triangular NPs (average size: 32 nm) Ag triangular NPs (average size: 18 nm) Au NRs (average aspect ratio: 1.9, 3.2) Ag NRs (average aspect ratio: 2.2, 3.4)
CL intensity (a.u.) 132 139 134 145 168, 202, 241 124, 186, 209 374 197 215, 312 101, 154
NR, nanorod.
Copyright © 2015 John Wiley & Sons, Ltd.
Luminescence 2015
Gold nanoprisms assisted CL toward determination of fluoride ion
Results and discussion Chemiluminescence reaction of potassium permanganate-aldehydes in the presence of Au and Ag nanoparticles Oxidation of low-molecular-weight (LMW) aldehydes by potassium permanganate (KMnO4) in acidic condition has been frequently
applied to produce CL emission for quantification of many compounds (29,30). Experimental results showed that Glu produce the best emission among some examined LMW aldehydes. However, this response is so weak that can be applied for analytical aims (Fig. 1a). In a previous study, we indicated that the obtained CL emission from the KMnO4–formaldehyde reaction (in SDS micellar medium) is greatly increased in the presence of Ag, Au NPs (34). In this way, the effect of these NPs was investigated fir CL intensity of the Glu system in their various sizes and shapes. As shown in Table 1, the most examined NPs have a good enhancing effect on CL intensity. However, the enhancing effect of Au prismshaped NPs (NPrs) is markedly better (Table 1). Figure 2 shows TEM image of the synthesized Au NPrs that confirms their correct synthesis and shapes. Also, slight enhancement effects were found (Table 1) for precursor solutions used for the preparation of NPrs, showing that the observed catalytic effect is due to the applied nanoparticles. A schematic display of the CL system is indicated in Scheme 1. The kinetic curve of the CL reaction was obtained in the absence and presence of Au NPrs (Fig. 1). It shows that the CL intensity reached a maximum value at 5 and 1 s after injection of permanganate solution, respectively. This result denotes the catalytic effect of NPrs with increasing effect on reaction rate. Optimization of reaction conditions
Figure 2. (a) UV–vis spectra; and (b) TEM image of applied gold nanoprisms.
Parameters influencing CL emission of the KMnO4–Glu–Au NPrs system were investigated to establish the optimal conditions for the CL reaction. The effects of type and concentration of surfactants on the CL system was studied by applying different kinds of surfactants (SDS, AOT and CTAB) at various concentrations. The presence of a surfactant in a CL system can influence the CL emission in many cases. It is obvious that at concentrations near the critical micellar concentration (CMC), the surfactant molecules tend to form micelles in aqueous solution, providing a protective environment for the Mn(II) excited state and a better contact between interacting species that leads to a significant increase in the CL quantum yield. As can be seen in Fig. 3(a), maximum CL intensity was obtained at 0.003 mol L 1 SDS concentration, which is near its CMC value (3.10 × 10 3 mol L 1 in this condition as obtained by conductometric measurements (44)). Therefore, the increasing effect of SDS in the CL system can be explained by producing micelles in CL reaction media. To study the type of acid and concentration on CL emission intensity, several acids (including HCl, H2SO4, HNO3 and H3PO4) were
3
3+
Scheme 1. KMnO4–Glu–Au NPrs CL system enhanced with Fe
Luminescence 2015
Copyright © 2015 John Wiley & Sons, Ltd.
–
and F ions.
wileyonlinelibrary.com/journal/luminescence
J. Abolhasani et al.
Figure 4. (a) CL spectra of (1) the Glu–KMnO4, and (2, 3) the Glu–KMnO4–Au NPrs sys–1 3+ 1 tems in the absence (2) and presence of 50 μmol L Fe (3) (condition: 0.05 mol L 1 1 H2SO4, 3 mmol L SDS, 1.2% v/v Glu, 0.3 mmol L KMnO4 and 600 μL Au NPr)]. (b) TEM image of reaction solution includes gold nanoprisms after CL emission.
1
4
Figure 3. Effect of (a) type and concentrations of some surfactants (0.05 mol L 1 H2SO4, 0.8% v/v Glu, 0.25 mmol L KMnO4 and 800 μL Au NPrs). (b) Type and concen1 1 trations of some acids (3 mmol L SDS, 0.8% v/v Glu, 0.25 mmol L KMnO4 and 1 1 800 μL Au NPrs). (c) Concentration of Glu (0.05 mol L H2SO4, 3 mmol L SDS, 1 1 0.25 mmol L KMnO4 and 800 μL Au NPrs). (d) Amount of Au NPrs (0.05 mol L 1 1 H2SO4, 3 mmol L SDS, 1.2% v/v Glu and 0.25 mmol L KMnO4) and (e) concentra1 1 tion of KMnO4 (0.05 mol L H2SO4, 3 mmol L SDS, 1.2% v/v Glu and 600 μL Au NPrs) 1 3+ 1 – on the CL intensity [10 μmol L Fe , 30 nmol L F ].
wileyonlinelibrary.com/journal/luminescence
applied. This parameter is very important in permanganate CL systems (29). Figure 3(b) shows that the best CL response was obtained for 0.05 mol L 1 H2SO4. Higher concentrations may demolish NPs, leading to a decrease in CL emission. The effect of Glu concentration was also examined and optimal CL intensity was obtained at 1.2% (v/v) Glu (Fig. 3c). High concentrations of Glu may increase the amount of Mn(III) produced and so the concentration of Mn(II)*, which substantially enhanced the CL intensity. At concentrations greater than 1.2% (v/v), Glu does not affect CL emission, markedly. The concentration of Au NPrs solution has a remarkable effect on the CL response of the considered system. Therefore, different amounts of prepared NPrs stock solution were added to the reaction environment. Figure 3(d) shows the result that reveals the maximum sensitivity obtain when 600 μL of NPrs solution were used. Finally, the effect of permanganate concentration was investigated over the range of 5.0 × 10 5 to 5.0 × 10 4 mol L 1. As shown in Fig. 3(e), the CL signal increased up to 0.3 mmol L 1 and then was fixed at the higher concentrations. As shown in a previous study (34), lower KMnO4 concentrations lead to a decrease in the number of excited intermediates and the CL intensity is diminished.
Copyright © 2015 John Wiley & Sons, Ltd.
Luminescence 2015
Gold nanoprisms assisted CL toward determination of fluoride ion Possible mechanism for chemiluminescence reaction
Figure 5. CL profiles of (a, b) the Glu–KMnO4 systems in the absence (a) and pres3+ – ence of Fe and F , (c-e) the Glu–KMnO4–SDS–Au NPrs system in the (c) absence 3+ 3+ – 1 and presence of (d) Fe and (e) Fe and F (Condition: 0.05 mol L H2SO4, 1 1 3 mmol L SDS, 1.2% v/v Glu, 0.3 mmol L KMnO4 and 600 μL Au NPrs).
Various studies have been carried out to discover the possible mechanism for CL emission from the KMnO4–aldehyde reaction, (29). In our previous study, we investigated the acidic permanganate–formaldehyde system mechanism and found that the oxidation of a substrate by Mn(VII) generates Mn(III), which then produces Mn(II)* in the excited state (emission at about 700 nm) (34). Researchers have confirmed this mechanism, they related red CL emission from permanganate CL reactions to the relaxation of the Mn(II) 4 T1 excited state to the 6A1 ground state (29,30). The CL emission spectra for the Glu–KMnO4–Au NPrs system (Fig. 4a) shows the characteristic red emission (λem = 708 nm) of permanganate CL. Thus, emitting species in this condition are the same and NPrs only act as catalysts to enhance the net reaction rate. The TEM image obtained after the reaction (Fig. 4b) showed no nanoparticles in solution, showing that the nanoparticles oxidized with MnO–4. Thus, the oxidation of Au NPrs by permanganate was investigated by UV–vis spectra in acidic conditions (34). The oxidation rate is low and cannot be responsible for enhancement of the CL emission. Conversely, Glu and SDS do not affect NPrs structures. Therefore, it can be said that NPrs do not participate directly in the CL reaction and just act as catalysts. As illustrated in a previous study, NPs may facilitate the oxidative cleavage of permanganate ester and accelerate Mn(III) generation leading to a higher CL intensity (34). Triangular Au NPs have higher surface-to-volume ratios and more active surface sites compared with spherical Au NPs, which facilitated the active intermediates generation on the surface of Au NPrs (45). Therefore, Au NPrs displayed greater catalytic activity than spherical ones. Moreover, the triangular Au NPs are more stable at a wider pH range (45), and thus they can be relatively unchanging in acidic media applied to the CL reaction, leading to their high catalytic activity.
Analytical application of the CL system
3+
Figure 6. Effect of the concentration of Fe on CL intensity under optimum condition.
It was found in the initial experiments that the addition of low amounts of iron(III) ions (Fe3+) to the presented CL system increased the CL intensity (Fig. 5). This increase is linear in the concentration range 0.01–15 μg mL–1 (0.18–268 μmol L–1) of Fe3+
5
3+
Figure 7. CL profiles of the KMnO4–Glu–SDS–Au NPrs–Fe calibration graph.
Luminescence 2015
–1
–
system in the presence of various concentrations (nmol L ) of F at optimum conditions and (inset) corresponding
Copyright © 2015 John Wiley & Sons, Ltd.
wileyonlinelibrary.com/journal/luminescence
J. Abolhasani et al. with a detection limit of 2.4 ng mL–1 (0.043 μmol L–1) (Fig. 6). Conversely, simultaneous existence of both Fe3+ and F– ions led to a significant intensification that was proportional to each ion concentration. These effects are not clearly understood, but it may be related to the catalyzing effect of Fe3+ and its complex with F–. These ions can increase CL intensity of the system in the absence of Au NPrs (Fig. 5), and also, they do not affect NPrs absorption spectra. Thus, the enhancing effect of Fe3+ and F– cannot be related to these NPs. Conversely, it is well known that metal ions can interact with aldehydes (46). Fe3+ or its complexes may interact with Glu or oxidation products, and this interaction may be responsible for their catalyzing effect. This phenomenon was exploited to develop a sensitive method for the determination of the F– ion. In the presence of 10 μmol L–1 Fe3+, the wide linear dynamic range was obtained for F– determination. Under the optimum conditions described and in the presence of 10 μmol L–1 Fe3+, the linear range of the calibration graph is 6–1200 nmol L 1 with a detection limit (3 sec) of 2.1 nmol L 1. The equation for the regression line was ΔI = 0.879C + 0.417(R2 = 0.9997), where ΔI = I0 – I is the difference between the CL intensity in the absence (I0) and presence of F– (I), and C is the concentration of F– ion in nmol L 1 (Fig. 7). The relative standard deviation (RSD) was calculated to be 0.81, 0.40, and 0.31% for five determinations of 72, 300 and 840 nmol L 1 F–, respectively. Results demonstrated tgat this CL system has good precision, relatively high sensitivity and a wide linear range. A comparison between the presented method and some other reported analytical methods for the determination of F– ion is summarized in Table 2. The characteristics of presented method are comparable with most other methods. Moreover, it has potential applications in the determination of some other compounds.
from these species in F– determination. In addition, Fig. 8 show the response of our CL system to different ions. As can be seen, the method was selective for the F– ion.
Analysis of real samples The considered method was easily applied to the determination of F– ion in urine and environmental water. The preparation step was
Study of interferences In order to test the interference effect of some potentially interfering substances, increasing amounts of these species were added into a standard solution of 50 nmol L 1 F– ion. The tolerable concentration ratios for interferences at a relative error of
