Molar Range Detection Based on Sideband Differential Absorption

Nov 23, 2017 - The quantitative analysis of absorption spectroscopy is based on .... Atmospheric absorption spectra consist of broadband spectral ...
0 downloads 0 Views 894KB Size
Download PDF
Article Cite This: Anal. Chem. 2017, 89, 13429−13433

pubs.acs.org/ac

Molar Range Detection Based on Sideband Differential Absorption Spectroscopy with a Concentrated Reference Liu-Chuang Zhao, Mei-Hong Guo, Xiao-Dong Li, Yu-Ping Huang, Shao-Hua Wu, and Jian-Jun Sun* Ministry of Education Key Laboratory for Analytical Science of Food Safety and Biology, Fujian Provincial Key Laboratory of Analysis and Detection Technology for Food Safety, College of Chemistry, Fuzhou University, Fuzhou, Fujian 350108, China S Supporting Information *

ABSTRACT: Conventional absorption spectroscopy (CAS) with a blank reference has only a slight capacity to detect high concentrations at characteristic wavelengths owing to the corresponding large molar absorption coefficient (ε) on the scale of 103 or 104 cm−1 M−1. To monitor concentrated analytes as high as the molar range in a plating bath and on a chemical production line, we propose a new approach using sideband differential absorption spectroscopy (SDAS). SDAS is obtained by subtracting the absorption spectra of the samples, A(λ,Cx), from that of a reference containing a concentrated standard analyte, A(λ,Cref>Cx), resulting in concave spectra with peaks at the sideband of conventional spectra with generally low ε values on the scale of 100 cm−1 M−1 or less. The negative absorbance changes linearly with the sample concentration at a certain peak wavelength, obeying Lambert−Beer’s law. In this work, SDAS was obtained and verified using inorganic and organic substances, such as chromate potassium, rhodamine B, and paracetamol.

I

solvent,14 and so on. In the reference calibration, which at least contains the solvent, some of the deviations such as fluorescence, scattering, and others resulting from the solvent can be eliminated; more importantly, any drift in the light source or photodetector characteristics can also be eliminated. The maximum detecting concentration (Cmax) with CAS can be predicted according to eq 1. The first possible approach to improve Cmax is reducing the thickness of the cuvettes (b). To achieve the molar range detection, b should be in the micrometer range in consideration of a ε value of approximately 104 cm−1 M−1. Nevertheless, the reality is that detection is adversely affected when b is too small because, in very thin films, there exists an electric field standing wave effect, which causes periodic variations of transmittance and reflectance. This interference leads to deviations from Lambert−Beer’s law.15 In addition, in actual situations, such as in a plating bath, it is difficult for sample solutions to flow into the very thin cuvettes, even when using a mechanical pump. Moreover, the interference of impurities increases as a result of the small capacity of the cuvettes. If b is fixed in the range of 0.1−1 cm and A is limited to 3 according to the present instrument technique, then Cmax cannot exceed millimolar amount when ε is approximately 104 cm−1 M−1 at the characteristic wavelength (λmax). However, if ε is approximately 10 cm−1 M−1, then the molar range of Cmax can be reached. Therefore, the second approach worth

n the plating process and in the chemical industry, monitoring the major constituents in the plating bath and in the chemical production pipe is of great concern. The concentrations of metal ions in the plating bath and the chemicals in the reactor/pipe line are generally as high as from a tenth to a few moles per liter. The common techniques for analyzing plating baths and production pipes are absorption spectroscopy,1electrochemical methods,2 high-performance liquid chromatography,3 mass spectrometry,4 and capillary electrophoresis.5 However, these techniques are generally implemented outside the plating bath and with the dilution of concentrated samples. Although conventional absorption spectroscopy (CAS) with a blank reference is economical and simple compared with the other techniques, most substances with high concentrations in the molar range cannot be detected directly, except for a few substances such as copper(II)6 and nickel(II).7 The quantitative analysis of absorption spectroscopy is based on Lambert−Beer’s law which is the combination of Lambert’s law8 and Beer’s law:9 (1)

A = εbc

where A is the absorbance, ε is the molar absorption coefficient, b is the light path (the length of medium through which the light is passing; commonly given in cm), and c is the concentration of the absorbing molecule (mol L−1 or M). Although it is vital for quantitative analysis in spectroscopy, Lambert−Beer’s law is limited by certain deviations, such as fluorescence,10 scattering,11 the introduction of optical filters in the path of the beam,12 chemical equilibrium or physical interactions,13 instrumental noise, the refractive index of the © 2017 American Chemical Society

Received: September 11, 2017 Accepted: November 23, 2017 Published: November 23, 2017 13429

DOI: 10.1021/acs.analchem.7b03722 Anal. Chem. 2017, 89, 13429−13433

Article

Analytical Chemistry Table 1. Maximum Detection Concentrations Based on Lambert−Beer’s Lawa substance Cr(VI) (K2CrO4) chromium(VI)−diphenylcarbazide iron(II)−1,10-phenanthroline zinc(II)−zincon nickel(IV)−dimethylglyoxime rhodamine B paracetamol Cu(II) (CuSO4·5H2O) Ni(II) (NiSO4·7H2O)

ε/(cm−1 M−1)

wavelength/nm

Cmax

method/ref

4.53 × 104 1.8 4 × 104 1.1 × 104 2.5 × 104 2.6 × 104 4.84 × 104 76.25 0.93 × 104 64.1 13.3 2.14

372 493 540 510 620 560 553 634 249 309 810 722

66 μM to 0.66 mM 3 Mb 75 μM to 0.75 mM 0.27 to 2.73 mM 0.12 to 1.17 mM 0.11 to 1.15 mM 62 μM to 0.62 mM 0.04 to 0.4 M 0.32 to 3.2 mM 0.047 to 0.47 M 0.22 to 2.2 M 1.4 to 14 M

CASc/this work SDASd/this work CAS/ref 19 CAS/ref 20 CAS/ref 21 CAS/ref 22 CAS/this work SDAS/this work CAS/Figure S-1 SDAS/Figure S-2 CAS/Figure S-3 CAS/Figure S-4

a

A is limited to 3, and b is within 0.1−1 cm. bThe saturation concentration of K2CrO4 is 3 M at room temperature; cCAS is conventional absorption spectroscopy with a blank reference. dSDAS is sideband differential absorption spectroscopy proposed in this work with a concentrated reference that is higher than the samples for detection.

kind of chemicals as the compositions of the samples to be detected. The concentration of the reference analyte (Cref) is known and is higher than that in the detection samples (Cx). Concave spectra are obtained by subtracting the absorption spectra of the samples A(λ,Cx) from that of a reference A(λ,Cref), namely,

attempting to realize molar range detection is to obtain a small ε. Table 1 lists the Cmax of some typical organics and inorganics or their chelates, in which Cr, Fe, Zn, Ni, and Cu are common metals in plating baths and the other chemicals such as rhodamine B and paracetamol are commonly used in dyes and medicines. Using CAS, only copper and nickel with small ε at λmax can be detected in the molar range. However, for the other substances with ε of approximately 104 cm−1 M−1, Cmax does not exceed mM. Therefore, when detecting concentrated samples by using CAS, the samples must be diluted to hundredths or thousandths of their original concentration. As early as 1949, Hiskey proposed differential spectrophotometry (DS) to precisely analyze samples with high absorption.16 By changing the reference from a blank solution to a known standard sample with a concentration slightly less than that of the sample for detection, the original detected absorbance, e.g., near 3, could be shifted into a suitable range of 0.3−1 to attain a better performance for the instrument. In general, many corresponding different concentrations of references are needed when detecting a series of concentrated samples. In addition, the mathematical calculation is relatively cumbersome. DS is predominantly aimed to improve the detection accuracy of highly absorbable samples but not to improve the detection limit because the detection range is generally not increased by an order of magnitude. In 1976, Platt et al. applied optical absorption spectroscopy to the detection of OH-radicals in the lower troposphere.17 In 1979, nitrous acid in the atmosphere was detected by differential optical absorption spectroscopy.18 Since then, differential optical absorption spectroscopy (DOAS) has been developed in leaps and bounds in the field of detection of trace compositions in the atmosphere. Atmospheric absorption spectra consist of broadband spectral changes caused by Rayleigh and Mie scattering and narrowband spectral changes caused by molecule absorption. To remove the effects of Rayleigh and Mie scattering, filtering technology was used to extract the narrowband spectral changes, which were used to calculate the molecular concentration in the atmosphere. This description provides the basic idea of DOAS. Approach of SDAS. Herein, we proposed sideband differential absorption spectroscopy (SDAS) to determine a small ε and then detect concentrated samples in the molar range. SDAS is carried out by changing the reference from a blank solution to a concentrated solution containing the same

ΔA(λ , Cx , Cref ) = A(λ , Cx ) − A(λ , Cref )

(2)

where λ is a representation of a range of wavelengths. The concave spectra lie in the sideband wavelengths of conventional spectra with generally low ε on the scale of 100 cm−1 M−1 or less. The negative absorbance obtained by SDAS changes linearly with the sample concentration, well obeying Lambert− Beer’s law at a certain peak wavelength.

■

EXPERIMENTAL SECTION Regents. Potassium chromate and paracetamol were purchased from Sinopharm Chemical Reagent Co. Ltd. (Shanghai, China). Rhodamine B was purchased from Aladdin Industrial Corp. (Shanghai, China). Instrumentation. Unless otherwise specified, the absorption spectra were obtained using a TU-1950 double beam spectrophotometer (Persee, Beijing), where b was fixed as 1 cm. The verification experiments were conducted in other devices, such as a Lambda 750 double beam spectrophotometer (PerkinElmer) and a Multiskan GO full-wavelength microplate reader (Thermo Scientific).

■

RESULTS AND DISCUSSION We first exhibit the use of SDAS by locating the concave spectra of a potassium chromate example and then determining its characteristics. Then, we verified the accuracy and universality of SDAS using rhodamine B as an example. Characterization of SDAS. Concentrated Cr(VI) working solutions are typically used in the metal finishing industry. To date, there is no directly effective method for monitoring concentrated chromate in the plating bath. Fortunately, we can solve this problem by using SDAS. Figure 1A shows the absorption spectra of K2CrO4 solutions in a low concentration range of 20−400 mg/L with a water reference. The peaks of the spectra appeared at 372 nm, where molar absorption coefficient ε = 4.53 × 104 cm−1 M−1. However, the spectra of samples with concentrations higher than 200 mg/L (0.9 mM) were in a superposition and were 13430

DOI: 10.1021/acs.analchem.7b03722 Anal. Chem. 2017, 89, 13429−13433

Article

Analytical Chemistry

wavelength but through the peak heights with R2 = 0.87699 (see the inset of Figure S-5). Therefore, SDAS has a unique advantage for detecting concentrated samples in the molar range. In addition, the operation process of SDAS is as simple as that of CAS. There are three characteristics that distinguish SDAS from CAS. First and the most fundamental is that the concentrated reference is higher in concentration than the samples for detection. Second, concave spectra with a negative absorbance are obtained from the differential method. Third, a low ε is obtained on the scale of 100 cm−1 M−1 or less at the sideband wavelengths of conventional spectra, which is the theoretical basis of the detection of concentrated samples. Verification of SDAS by Rhodamine B. Rhodamine B is an ordinary fluorescent organic reagent that must be monitored in the production process over a large concentration range. Figure 2A shows the absorption spectra of rhodamine B solutions over a low concentration range of 5−70 mg/L with a water reference. The characteristic wavelength was 553 nm where the spectra peaks appeared, accompanied by ε = 4.9 × 104 cm−1 M−1. The inset of Figure 2A shows that samples with concentrations exceeding 50 mg/L (0.1 mM) cannot be detected accurately at 553 nm by CAS because the large absorption exceeds the instruments’ detection sensitivity. However, when the reference was changed from water to a 20 g/L (0.04 M) standard solution, concave spectra of samples in the concentration range of 5−18 g/L (0.036 M) were found, as shown in Figure 2B. These spectra were smooth and consecutive, lying at sideband wavelengths at approximately 634 nm with ε = 76.25 cm−1 M−1. At the detection wavelength of 634 nm, a good linear relationship was found between the negative absorbance and the concentration and with R2 = 0.99924, obeying Lambert−Beer’s law (see the inset of Figure 2B). Next, we conducted experiments of the same concentrated samples of rhodamine B (5−18 g/L) by using two other UV/ vis spectrophotometers: a Lambda 750 double beam spectrophotometer (PerkinElmer) and a Multiskan GO fullwavelength microplate reader (Thermo Scientific). As shown in Figure 2C, the spectra from Lambda 750 are extremely similar to those from TU-1950. The peak wavelengths appeared around 633 nm, where a good linear relationship existed between the negative absorbance and the concentration with R2 = 0.99909 (see the inset of Figure 2C). Note that Lambda 750 and TU-1950 are both double beam spectrophotometers. As shown in Figure 2D, although the spectra obtained from the Multiskan GO were not sufficiently smooth compared with the spectra in Figure 2B,C, the peak wavelengths were constant around 629 nm. Meaningfully, the inset of Figure 2D indicates that the negative absorbance changes linearly with the concentration at 629 nm, accompanied by R2 = 0.99691. Therefore, SDAS is expected to be applied for quantitative analysis of concentrated samples using different spectrophotometers, including single and double beam spectrophotometers. Similar to analysis of inorganic chromate by SDAS, rhodamine B can also display similar particular phenomena in high concentration range concave spectra. Then, quantitative analysis can be accomplished. The concentrated rhodamine B example confirmed the accuracy and universality of SDAS. In the Supporting Information (Figures S-1 and S-2), the example of paracetamol further verified the SDAS method.

Figure 1. UV/vis absorption spectra of potassium chromate aqueous solutions with different references. (A) CAS, reference: water. Sample concentrations: 20, 50, 100, 200, 300, 350, and 400 mg/L (from bottom to top). (B) SDAS, reference: 300 g/L (1.4 M). Sample concentrations: 50, 100, 150, 200, and 250 g/L (from bottom to top). The insets show the linear relationship between absorbance and concentration at 372 and 493 nm, respectively.

relatively misshapen. The inset of Figure 1A indicates that higher concentration samples of more than 200 mg/L cannot be detected at 372 nm by CAS because, at 372 nm with a large ε, the absorbance of a concentrated sample easily exceeds the upper limit of detection for most spectrophotometers. However, when the reference was changed from water to a 300 g/L (1.4 M) standard solution, unexpected concave spectra of concentrated K2CrO4 solutions in the range of 50−250 g/L (1.17 M) were found (Figure 1B). Concave spectra lay at the sideband wavelengths of conventional spectra, away from 372 nm. At 493 nm with ε = 1.8 cm−1 M−1, the standard working curve of the negative absorbance and concentration was plotted. The degree of fitting, R2, was 0.99371 (see inset of Figure 1B). SDAS detected a Cmax value that was 3 orders of magnitude higher than that detected by CAS. The differential method in which the samples for detection (50−250 g/L) were subtracted from a higher concentrated reference (300 g/L) led to concave spectra; therefore, the method was called sideband differential absorption spectroscopy. Note that the concentrated reference, with a higher concentration instead of a lower concentration than that of the samples for detection, is selected to accomplish precise quantitative analysis. If the reference was 50 g/L and the sample concentrations were 100−300 g/L, then the resulting spectra shown in Figure S-5 were partially overlapped and the peak wavelengths were clearly red-shifted from 482 to 492 nm. Thus, quantitative analysis cannot be performed at a consistent 13431

DOI: 10.1021/acs.analchem.7b03722 Anal. Chem. 2017, 89, 13429−13433

Article

Analytical Chemistry

Figure 2. UV/vis absorption spectra of rhodamine B aqueous solutions with different references and using different spectrophotometers. (A) CAS, reference: water. Sample concentrations: 5, 10, 15, 30, 50, and 70 mg/L (from bottom to top). (B−D) SDAS, reference: 20 g/L (0.04 M). Sample concentrations: 5, 9, 12, 15, and 18 g/L (from bottom to top). (B) By TU-1950, (C) by Lambda 750, and (D) by Multiskan GO. The insets show the linear relationship between absorbance and concentration, at (A) 553, (B) 634, (C) 633, and (D) 629 nm.

■

■

CONCLUSION In general, for organics and inorganics, low-concentration samples can be detected by CAS and high-concentration samples as high as the molar range can be determined by SDAS, which is as convenient as CAS. The particular concave spectra are obtained with a concentrated reference (with a higher concentration than that of the samples for detection) at sideband wavelengths of conventional spectra. Then, quantitative analysis of these samples can be performed. A concentrated reference not only functions as the blank reference but also eliminates the error originating from the interactions between different compositions. In addition, the common saturation solution set as the reference is stable and simple to use in an experimental process. Importantly, dilution of the original concentrated samples is no longer required using the proposed method. Therefore, the practicality and accuracy of SDAS can be guaranteed. SDAS has provided a novel avenue for real-time detection, especially in the plating process and in the chemical industry.

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Tel./Fax: +86 591 22866136. ORCID

Jian-Jun Sun: 0000-0002-7987-5242 Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS We are thankful for the financial support from the National Science Foundation of China (Grant No. 21475023) and the Program for Changjiang Scholars and Innovative Research Team in University (Grant No. IRT_15R11).

■

REFERENCES

(1) Inoue, F.; Philipsen, H.; Radisic, A.; Armini, S.; Civale, Y.; Shingubara, S.; Leunissen, P. J. Electrochem. Soc. 2012, 159, D437− D441. (2) Xie, B.-G.; Sun, J.-J.; Chen, X.-B.; Chen, J.-H.; Xiang, T.-L.; Chen, G.-N. J. Electrochem. Soc. 2007, 154, D516−D519. (3) Sun, Z. W.; Yu, C.; Dixit, G. U.S. Pat. Appl. 10/051,611, Jan. 18, 2002. (4) Khosla, M. U.S. Pat. 6,726,824, Apr. 27, 2004. (5) Oehrle, S. A. J. Chromatogr. A 1996, 739, 413−419. (6) Costas, L. P. U.S. Pat. 3,925,168, Dec. 9, 1975. (7) Mack, H. S. U.S. Pat. 4,699,081, Oct. 13, 1987. (8) Lambert, J. H. Photometria sive de mensura et gradibus luminis, colorum et umbrae; Klett, 1760.

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.analchem.7b03722. UV/vis spectra of paracetamol ethanol solutions, of CuSO4 aqueous solutions, of NiSO4 aqueous solutions, and of K2CrO4 aqueous solutions (PDF) 13432

DOI: 10.1021/acs.analchem.7b03722 Anal. Chem. 2017, 89, 13429−13433

Article

Analytical Chemistry (9) Beer, A. Ann. Phys. 1852, 162, 78−88. (10) Ungnade, H. E.; Kerr, V.; Youse, E. Science 1951, 113, 601−601. (11) Rose, H. E. Nature 1952, 169, 287−288. (12) Galli, C. J. Pharm. Biomed. Anal. 2001, 25, 803−809. (13) Buijs, K.; Maurice, M. J. Anal. Chim. Acta 1969, 47, 469−474. (14) Garcia-Rubio, L. H. Macromolecules 1992, 25, 2608−2613. (15) Mayerhöfer, T. G.; Mutschke, H.; Popp, J. ChemPhysChem 2017, 18, 2916−2923. (16) HISKEY, C. F. Anal. Chem. 1949, 21, 1440−1446. (17) Perner, D.; Ehhalt, D. H.; Pätz, H. W.; Platt, U.; Röth, E. P.; Volz, A. Geophys. Res. Lett. 1976, 3, 466−468. (18) Perner, D.; Platt, U. Geophys. Res. Lett. 1979, 6, 917−920. (19) de Andrade, J. C.; Rocha, J. C.; Pasquini, C.; Baccan, N. Analyst 1983, 108, 621−625. (20) Saywell, L.; Cunningham, B. Ind. Eng. Chem., Anal. Ed. 1937, 9, 67−69. (21) Platte, J.; Marcy, V. Anal. Chem. 1959, 31, 1226−1228. (22) Haar, K. T.; Westerveld, W. Recl. Trav. Chim. Pays-Bas 1948, 67, 71−81.

13433

DOI: 10.1021/acs.analchem.7b03722 Anal. Chem. 2017, 89, 13429−13433