Study on the interaction between emodin and ethyl violet by resonance Rayleigh scattering technique

Abstract

A novel resonance Rayleigh scattering method was developed for the determination of emodin (EMO). In pH 7.0 Britton–Robinson (BR) buffer medium, the scattering signal of ethyl violet was remarkably enhanced after adding trace amount of EMO and forming a 1 : 1 ion-association complex, which not only resulted in the change of absorption spectra, but also led to a significant enhancement of resonance Rayleigh scattering (RRS), frequency doubling scattering (FDS) and second order scattering (SOS). The maximum RRS, SOS and FDS wavelengths of the ion-association complex were located at 340 nm, 528 nm and 341 nm, respectively. The linear ranges and detection limits for RRS, SOS and FDS were 0.1–4.2 μg mL⁻¹, 0.4–4.8 μg mL⁻¹, 0.4–4.8 μg mL⁻¹ and 3.8 ng mL⁻¹, 14.2 ng mL⁻¹, 17.7 ng mL⁻¹, respectively. In this work, the optimum conditions, the influencing factors and the effects of coexisting substances on the reaction were investigated. The method can be applied to the determination of EMO in serum and urine sample and the results were satisfactory. Moreover, the reaction mechanism and reasons of the enhancement of resonance light scattering were discussed.

---

1. Introduction

Emodin (EMO), widely exists in cathartic plants such as rhubarb roots, buckthorn's root and root bark, and Cassia seeds, is a kind of an anthraquinone compound. EMO inhibits the synthesis DNA by interfering with the replication of DNA template.¹ Based on that information it has the effects such as anti-tumor,^2–4 anti-microbial growth,⁵ and immunosuppression. Clinically, pure EMO is mainly used to treat leukemia,⁶ gastric cancer and is most commonly used for bacteriostasis. Moreover, EMO also works as antitussive, diuretic, choleretic, antispasmodic, and antihypertensive. Although toxicity of emodin is small, it is worth noting that pregnant women are forbidden to take it because it may lead to miscarriage. Due to the wide application of EMO, the determination of EMO and its metabolites has become especially interesting.

At present, the methods for the determination of EMO consist of fluorescence spectrometry,⁷ micellar electrokinetic capillary chromatography,⁸ spectrophotometry,⁹ supercritical fluid extraction,¹⁰ thin layer chromatography scanning,¹¹ reversed-phase high performance liquid chromatography,¹² high-performance liquid chromatography (HPLC),¹³ and resonance light scattering technique.¹⁴ Among them, HPLC needs some complex pretreatments and 0.48–2.40 μg mL⁻¹ is a little narrow. Spectrophotometric method is simple to operate, but it has a low sensitivity. Fluorescence spectrometry method based on the fluorescence intensity of derivatives of calix[4]arene was quenched by EMO as the result of the formation of a weaker fluorescent inclusion complex sensitized in hexadecyltrimethylammonium bromide; however, its detection limit is not high enough.⁷ Therefore, to develop new, sensitive, simple, and rapid methods for trace determination of EMO makes sense.

Resonance Rayleigh scattering (RRS) and resonance nonlinear scattering (RNLS) have extensive applications^15–22 because of their sensitivity, simplicity and rapidity, and they can be achieved in a common fluorescence spectrometer as new analytical methods. In recent years, these techniques have been increasingly applied to determine DNA sequence,¹⁵ ozone,¹⁶ DNA,¹⁷ metal ions^18,19 and polysaccharides such as hyaluronic acid²⁰ and chondroitin sulfate A,²¹ and pharmaceuticals.²²

In the experiment, the interaction of EMO with some basic triphenylmethane dyes, including ethyl violet (EV), crystal violet (CV), malachite green (MG), light green (LG) and methyl green (MeG), which have a similar structure and exist as cationic in aqueous solution have been investigated. The results show that in the CV and MeG systems, the intensities of resonance Rayleigh scattering was much smaller. As for MG and LG, the intensities of EMO alone are too large. The 1 : 1 ion-association complex of EMO with EV produced a significant enhancement of RRS, frequency doubling scattering (FDS) and second order scattering (SOS), and the color fading of EV. The scattering intensities are proportional to the concentration of EMO in a wide range. The optimum conditions, the influencing factors and the effects of coexisting substances on the reaction were discussed by RRS method for its lowest detect limit of 3.8 ng mL⁻¹. The method can be applied to the determination of EMO in serum and urine sample and the results were satisfactory.

2. Experimental

2.1. Apparatus

A Hitachi F-2500 spectrofluorophotometer (Tokyo, Japan) was used to record the RRS, SOS and FDS spectra and to measure the scattering intensities with the slits (EX/EM) of 5.0/5.0 nm for RRS and 10.0/10.0 nm for SOS and FDS. A UV-3010 UV/VIS spectrophotometer (Tokyo, Japan) was employed for noting absorption spectra and measuring absorbance. A pH-S20K meter (Shanghai, Mettler-Toledo Instruments Co., LTD.) was used for adjusting pH values.

2.2. Reagents

The working solution of ethyl violet (EV) was 1.0 × 10⁻⁴ mol L⁻¹.

A standard solution of emodin (EMO, Aladdin) at a concentration of 100 μg mL⁻¹ was prepared by weighing and dissolving suitable amount of emodin (EMO) reagent in ethanol. Then, it was further diluted with water to 20 μg mL⁻¹ as working solution.

Britton–Robinson (BR) buffer solutions with different pH were prepared by mixing the 0.2 mol L⁻¹ NaOH and the mixture of 0.04 mol L⁻¹ H₃PO₄, H₃BO₃ and HAc according to suitable proportion. Note that the pH value was adjusted by pH meter.

All reagents were of analytical reagent grade, and doubly distilled water was used throughout.

2.3. Procedure

Into a 10 mL calibrated flask were added 1.8 mL of pH 7.0 BR buffer solution, 1.2 mL EV and suitable amounts of EMO. The mixture was diluted to the mark with water and mixed thoroughly. After waiting for 10 min, the RRS spectra of the system were recorded with synchronous scanning at λ_em = λ_ex, and the SOS and FDS spectra were recorded by scanning at λ_em = 2λ_ex and λ_em = 1/2λ_ex, respectively. Then, the scattering intensity I_RRS, I_SOS and I_FDS for the reaction product and I⁰_RRS, I⁰_SOS and I⁰_FDS for the reagent blank at their maximum wavelengths were measured, ΔI_RRS = I_RRS − I⁰_RRS, ΔI_SOS = I_SOS – I⁰_SOS and ΔI_FDS = I_FDS − I⁰_FDS. Note that the absorption spectra were recorded simultaneously.

3. Results and discussion

3.1. RRS spectra

Fig. 1 shows the RRS spectra of EMO, EV and EMO-EV system at pH 7.0 (the pH value of the assay solution is 7.63). As shown in Fig. 1, it can be seen that the RRS intensities of separate EMO and EV were very weak under optimum conditions. However, when a trace mount of EMO was added to the system of EV, the new spectra of RRS appeared and their intensities were enhanced remarkably with three peaks located at 276 nm, 340 nm and 657 nm. The enhancement of RRS intensities was directly proportional to the concentration of EMO, and the maximum RRS wavelength was located at 340 nm; therefore, 340 nm was selected as the analytical wavelength.

Fig. 1 RRS spectra of EMO-EV system 1, EV (1.2 × 10⁻⁵ mol L⁻¹); 2-, EMO (1.0 μg mL⁻¹); 3–10, EV-EMO system. C_EMO: 0.1, 0.4, 1.0, 1.5, 2.3, 3.2, 3.8, 4.2 μg mL⁻¹; C_EV: 1.2 × 10⁻⁵ mol L⁻¹; pH = 7.0. Inset is the plot relationship of the EMO concentration to RRS intensity.

3.2. SOS spectra and FDS spectra

The SOS and FDS spectra of EMO-EV system were investigated (see ESI, Fig. S1 and S2†), and the results showed that the SOS and FDS intensities of EMO and EV themselves were very weak. After binding of EMO with EV to form binary complexes, SOS and FDS intensities were enhanced greatly, and their maximum wavelengths (λ_ex/λ_em) were located at 260 nm/520 nm and 680 nm/340 nm, respectively. In addition, there was also a little small SOS at 270 nm/540 nm. Due to the high signal-to-noise ratio, we select 260 nm/520 nm and 680 nm/340 nm as detection wavelengths. The enhancement of SOS and FDS intensities was directly proportional to the concentration of EMO. We all know that RRS is a special elastic scattering produced when the wavelength of Raleigh scattering (RS) is located at or close to the molecular absorption band.²³ Note that as RRS occurs, SOS and FDS may be possible, and the scattering values and sensitivities of FDS and SOS are much lower than that of RRS.

3.3. Absorption spectra

Under the optimum conditions, the absorption spectra of EMO, EV and their complexes were researched (see ESI, Fig. S3†). The results demonstrated that the absorption of EMO is located at 305 nm, and the maximum absorption wavelength of EV was at 590 nm. When EV reacted with EMO to generate ion-association, its maximum absorption at 590 nm decreased, and the absorption intensity was directly proportional to the concentration of EMO in certain ranges (Fig. 2). Note that a new method for the determination of trace EMO based on fading reaction of EV was advanced.

Fig. 2 Absorption spectra of EV-EMO system. C_EMO(1–6): 0.8, 1.6, 2.4, 3.2, 4.0, 4.8 μg mL⁻¹; C_EV: 1.2 × 10⁻⁵ mol L⁻¹; pH = 7.0.

3.4. Optimum reaction conditions

3.4.1. Effect of acidity. The influences of solution acidity on the RRS intensity of the reaction system were tested. The results showed that the optimum pH range was 6.6–7.6. If pH was lower than 6.6, the RRS intensity of EMO increased rapidly, and if pH was higher than 7.6, the RRS intensity of EMO-EV decreased significantly (see ESI, Fig. S4†). Therefore, pH 7.0 was chosen as the acidity for the reaction, and the appropriate volume of buffer solution was 1.8 mL.

3.4.2. Effect of the EV concentration. The experiment results indicated that the concentration of EV had little effect on the determination of EMO. The RRS intensity reached the maximum when EV concentration is 1.2 × 10⁻⁵ mol L⁻¹ (see ESI, Fig. S5†). If EV is not enough, the reaction would be incomplete; hence, we chose 1.2 × 10⁻⁵ mol L⁻¹ as a suitable EV concentration.

3.4.3. Effect of ionic intensity. The effect of ionic strength on the intensity of RRS was investigated using 1.0 mol L⁻¹ NaCl solution (see ESI, Fig. S6†). The experiment results showed with the concentration of NaCl increasing, the intensities of RRS remained almost stable. Therefore, the reaction between EMO and EV, which may be electrostatic attraction, hydrogen bond and hydrophobic force played an important role together.

3.4.4. Effect of ethanol concentration. The effect of different volume ratio of ethanol on the RRS spectra of the EV and EMO-EV system was studied, and the results showed with increase in the ethanol concentration, ΔI_RRS was slightly affected, and decreased dramatically when the volume ratio of ethanol was beyond 0.04% (see ESI, Fig. S7†). Note that too much ethanol would reduce the hydrophobic interface between EMO anion and EV cation or even make the interface between water molecules and association complexes disappear, which results in a rapid decrease in RRS.^24,25 Therefore, ethanol should be controlled in a low concentration in the determination system.

3.4.5. Reaction speed and the stability. At room temperature, the reaction could be completed in 10 min and RRS intensity was stable in an hour.

4. Reaction mechanism and reasons for RRS enhancement

4.1. Formation of ion-association reaction

In the experiment, the composition ratio of EMO with EV in ion-association was researched by using Job's method of continuous variation and molar ratio method. When the concentration of both EMO and EV was 7.4 × 10⁻⁵ mol L⁻¹, the total volume of EMO and EV was 3.0 mL in the Job's method of continuous variation, and the volume of EMO was constant 1.0 mL in the molar ratio method. The RRS intensity was determined at λ_em = λ_ex = 340 nm, and the results showed that the ratio of EMO : EV was 1 : 1 (see ESI, Fig. S8†).

Using conductor-like Polarizable Continuum Model (CPCM)²⁶ processing solution and Mo62X/CCPVCTZ level of density functional theory (DFT) method,²⁷ total energy and charge distribution of EMO were optimized to define the binding site of EMO and EV. The results showed EMO has three conformational isomers, and the most stable conformation of EMO can be seen in Fig. 3. As shown in Fig. 3, the charge densities of 1,8α O and 3β O of EMO were −0.711, −0.710 and −0.704. However, α-OH interacts with near ketonic oxygen to form a stable intramolecular hydrogen bond and make its acid weakened; therefore, the most acidic phenolic hydroxyl group of emodin is at the 3β position of the molecule. For instance, in sodium hydroxide solution, its balance is reversible. With the change in pH, 3-OH firstly dissociates, followed by 1,8-OH.²⁸ The structure of the ion-association complex is shown in Fig. 4.

Fig. 3 Optimized conformation and its electronic structure.

Fig. 4 The structure of binary ion-association.

4.2. Reasons for RRS enhancement

4.2.1. Effect of absorption spectra on RRS. When Raleigh scattering spectrum is located at or close to the molecular absorption band, the scattering can resonate with absorption light leading to the significant enhancement of RRS intensity; therefore, RRS spectra should be closely related to the absorption spectra.²⁹ From the comparison of RRS and absorption spectra (Fig. 5), it is seen that three RRS peaks at 276 nm, 340 nm and 657 nm are close to the corresponding absorption peaks at 255 nm, 305 nm and 590 nm of the EMO-EV system; thus, RRS intensity is remarkably increased. Therefore, the resonance-enhanced Raleigh scattering effect is a critical reason of scattering enlargement.

Fig. 5 Comparison of RRS and absorption spectra. 1, RRS spectrum; 2, absorption spectrum. C_EMO: 1.0 μg mL⁻¹, C_EV: 1.2 × 10⁻⁵ mol L⁻¹ pH = 7.0.

4.2.2. Enlargement of molecular volume. According to the simplified formula of Raleigh scattering I = KCMI₀,³⁰ where molecular volume is replaced by molecular weight, RRS intensity is proportional to the molecular weight of aggregation of the particle. When EMO reacts with EV to form ion-association complex, it results in the increase of molecular weight, which is in favor of RRS enhancement. Moreover, the aggregation of the cationic dye on EMO is another important reason for the RRS enhancement.

4.2.3. Formation of the hydrophobic interface. Under the experimental conditions, EMO and EV exist in the form of anion and cation, respectively. They are water soluble, and can form hydrates easily in water, and the intensity of RRS is relatively low. When EMO and EV react with each other to form a neutralized ion-association complex, a hydrophobic liquid–solid interface appeared owing to the presence of the hydrophobic aryl framework of the binary complex. The formation of the hydrophobic interface is conducive to the enhancement of RRS signal.³¹

5. Selectivity, sensitivity and analytical application

5.1. Sensitivity

Under the optimum conditions, the RRS, SOS, FDS and absorbance spectral intensities of the system were measured at their own maximum wavelengths. The calibration graphs of ΔI_RRS, ΔI_SOS, ΔI_FDS and ΔA versus the concentration of EMO were constructed. The regression equation, linear range, correlation coefficient (r) and detection limit are listed in Table 1. It can be seen from the table that RRS method had highest sensitivity, and spectrophotometry sensitivity was far below the RRS, SOS and FDS method. The sensitivity of RRS method was higher than one to two orders of magnitude than those of common spectrophotometry, fluorimetry, micellar electrokinetic and capillary chromatography (shown in Table 2). Therefore, RRS method can be applied for the trace determination of EMO.

Table 1 Related parameters of the calibration graphs for the determination of EMO

Method	Linear regression equation (μg mL^−1)	Linear range (μg mL^−1)	Correlation coefficient (r)	Detection limit (3σ, ng mL^−1)
RRS	ΔI = 1272.38 + 1241.12c	0.1–4.2	0.9986	3.8
SOS	ΔI = 214.37 + 123.07c	0.4–4.8	0.9930	14.2
FDS	ΔI = 31.65 + 16.7675c	0.4–4.8	0.9951	17.7
SP	ΔA = −0.1327 − 0.07682c	0.8–4.8	0.9920	125.5

---

Table 2 Comparison of the sensitivities of some methods for the determination of EMO^a

Method	Medium condition	Detection wavelength (λex/λem, nm)	Linear range (μg mL^−1)	Detection limit (ng mL^−1)	Reference
a SP: spectrophotometry; FL: fluorimetry; TLCS: thin-layer chromatography scanning; MECC: micellar electrokinetic capillary chromatography.b μg.
HPLC		254	0.48–2.40		8
FL	pH 6.0	λex = 270	1.17–23.40	340	2
SP	pH > 12.00	530	1–200	500	4
TLCS		445	0.1084–1.084		6
MECC	pH 9.5		5.2–260	290	3
RP-HPLC		254	0.01584–0.1584^b		7
RRS	pH 6.5	350	0.54–9.72	10.3	9
RRS	pH 7.0	340	0.1–4.2	3.8	This work
SOS	pH 7.0	528	0.4–4.8	14.2	This work
FDS	pH 7.0	341	0.4–4.8	17.7	This work

---

5.2. Selectivity

Based on the potential interference composition in human urine and serum, the effects of common metal ions, common inorganic anions, proteins, amino acids and sugars on the determination of 1.0 μg mL⁻¹ EMO were investigated, and the results were given in Table 3. This showed that (the relative standard deviation was within ±5%) the larger amounts of common metal ions NH₄⁺ and Na⁺, inorganic anions Cl⁻¹ and NO₃⁻, urea, surfactants and saccharides, as well as the large numbers of some amino acids were allowed in high concentrations, whereas tolerable amounts of divalent metal ions may be small. The interferences can be diminished by diluting with doubly distilled water. Therefore, the method had a good selectivity and could be applied to real samples.

Table 3 Effects of coexisting substances (EMO was 1.0 μg mL⁻¹)^a

Coexisting substance	Times	Relative error (%)	Coexisting substance	Times	Relative error (%)
a SLS: sodium lauryl sulfate; SDBS: sodium dodecyl benzene sulfonate; SDS: sodium dodecyl sulfate; CTAB: cetyl trimethyl ammonium bromide.
CoCl2	10	4.6	SDBS	348.5	0.7
CuCl2	5	−4.9	SDS	288.4	−2.8
CaCl2	8	−4.0	CTAB	218.7	0.9
NaCl	2435	2.1	Cellulose	400	4.3
NH4NO3	2000	0.4	Glucose	200	3.0
MnSO4	10	4.5	HSA	40	4.8
ZnSO4	6	−4.8	L-Aspartic acid	160	5.0
MgSO4	10	2.1	L-Cysteine	100	3.8
BSA	100	3.6	L-Histidine	600	4.2
D-Tyrosine	40	4.9	Glycine	200	3.9
D-Phenylalanine	100	1.5	Sucrose	100	4.4
D-Fructose	100	−4.7	Starch	200	2.3
SLS	544.8	−3.4	Urea	500	4.9

---

5.3. Analytical application

5.3.1. Determination of EMO in human urine. Fresh urine sample (healthy human) centrifuged at 6000 rpm for 10 min, and then 1.0 mL aliquot of supernatant fluid was pipetted and diluted to 10 mL. A 1.0 mL aliquot of this solution was added into a 10.0 mL volumetric flask, and EMO was determined according to the general procedure. The recovery and relative standard deviation (R.S.D) were tested by using the standard addition method, and the results are listed in Table 4.

Table 4 Results for the determination of trace EMO in serum and urine sample

Sample	Found (μg mL^−1)	Added (μg mL^−1)	Total found (μg mL^−1, n = 5)	Recovery (%)	RSD (%)
Serum 1	ND	0.8	0.812, 0.817, 0.786, 0.821, 0.772	101.7	1.5
Serum 2	ND	1.6	1.66, 1.58, 1.65, 1.61, 1.63	101.9	4.7
Serum 3	ND	2.4	2.45, 2.33, 2.43, 2.35, 2.48	101.2	3.8
Urine 1	ND	0.8	0.809, 0.812, 0.792, 0.822, 0.831	101.5	4.5
Urine 2	ND	1.6	1.64, 1.62, 1.56, 1.64, 1.69	102.6	5.0
Urine 3	ND	2.4	2.45, 2.44, 2.36, 2.42, 2.47	101.8	5.6

---

5.3.2. Determination of EMO in human serum. Fresh serum sample (healthy human) was treated with suitable amounts of trichloroacetic acid and centrifuged for 10 min at 6000 rpm to exclude proteins. A 1.0 mL aliquot of the supernatant fluid was diluted to 20 mL, and 1.0 mL of this solution was pipetted into a 10.0 mL volumetric flask and EMO was determined according to the general procedure. R.S.D. and recovery were examined by using the standard addition method, and the results are listed in Table 4.

6. Conclusion

In the paper, EMO can interact with EV via electrostatic attraction, hydrogen bond and hydrophobic force to form a 1 : 1 ion-association complex, resulting in the fading of absorption and the significant enhancement of RRS, FDS and SOS. Based on this, a novel and simple RRS assay with high sensitivity and good selectivity for the determination of trace EMO has been developed. The proposed method was successfully applied to the determination of EMO in serum and urine sample.

Acknowledgements

This work was financially supported with the National Science Foundation of China (no. 21175015), and all authors here express their deep gratitude.

References

1. M. Ionov, I. Tukfatullina, B. Salakhutdinov, N. Baram, M. Bryszewska and T. Aripov, Cell. Mol. Biol. Lett., 2010, 15(1), 98–117.
2. Y. C. Chen, S. C. Shen, W. R. Lee, F. L. Hsu, H. Y. Lin, C. H. Ko and S. W. Tseng, Biochem. Pharmacol., 2002, 64(12), 1713–1724.
3. X. Jing, N. Ueki, J. Cheng, H. Imanishi and T. Hada, Cancer Sci., 2002, 93(8), 874–882.
4. G. Srinivas, R. J. Anto, P. Srinivas, S. Vidhyalakshmi, V. P. Senan and D. Karunagaran, Eur. J. Pharmacol., 2003, 473(2), 117–125.
5. H. H. Wang and J. G. Chung, Curr. Microbiol., 1997, 35(5), 262–266.
6. W. Chun-Guang, Y. Jun-Qing, L. Bei-Zhong, J. Dan-Ting, W. Chong, Z. Liang and W. Yan, Eur. J. Pharmacol., 2010, 627(1), 33–41.
7. L. Ma and X. Zhu, Spectrochim. Acta, Part A, 2012, 95, 246–251.
8. X. Wu, G. H. Chen, K. Wang, J. Shi and C. Q. Wu, J. Hebei Univ. Sci. Technol., 2011, 32(5), 426–430.
9. X. M. Sun and C. Ma, J. South-Cent. Univ. Natl., Nat. Sci. Ed., 2001, 20(4), 13–16.
10. K. B. Gou, H. L. Yuan, X. Y. Li, C. Zhang, Z. L. Liu and C. Guo, Chin. Pharm. J., 1999, 34(9), 619–621.
11. L. B. Wei, M. Meng and L. Z. Xia, Chin. J. Inf. Tradit. Chin. Med., 2009, 16(10), 47–48.
12. X. Xia, Chin. J. Pharm. Anal., 2008, 28(8), 1310–1312.
13. D. Guo, N. N. Chen, Y. Yan and L. B. Hou, China Pharm., 2003, 14(5), 296.
14. H. Xiang, Q. Luo, K. Dai, W. Duan, Y. Fan and Y. Xie, Spectrochim. Acta, Part A, 2012, 96, 874–881.
15. Z. G. Chen, S. H. Qian, X. Chen, J. H. Chen, Y. J. Lin and J. B. Liu, RSC Adv., 2012, 2, 2562.
16. C. Lin, G. Wen, A. Liang and Z. Jiang, RSC Adv., 2013, 3(18), 6627–6630.
17. L. Kong, Z. Liu, S. Liu and D. Wang, Anal. Methods, 2012, 4(12), 4346–4352.
18. Y. Luo, L. Xu, A. Liang, A. Deng and Z. Jiang, RSC Adv., 2014, 4(37), 19234–19237.
19. S. P. Liu, Z. F. Liu and H. Q. Luo, Anal. Chim. Acta, 2000, 407(1), 255–260.
20. T. Li, S. Liu, Z. Liu and X. Hu, Arch. Pharm., 2005, 338(9), 427–432.
21. H. Q. Luo, N. B. Li and S. P. Liu, Biosens. Bioelectron., 2006, 21(7), 1186–1194.
22. F. Tian, W. Huang, J. Yang and Q. Li, Spectrochim. Acta, Part A, 2014, 126, 135–141.
23. Y. Cui, F. Cui, L. Wang, Q. Zhang, W. Xue, F. Jing and J. Sun, J. Lumin., 2008, 128(10), 1719–1724.
24. S. P. Liu, S. Chen, Z. F. Liu, X. Hu and T. Li, Anal. Chim. Acta, 2005, 535(1), 169–175.
25. J. D. Peng, S. P. Liu, Z. F. Liu and Y. Shi, Acta Chim. Sin., 2005, 63(8), 745–751.
26. Y. Takano and K. N. Houk, J. Chem. Theory Comput., 2005, 1(1), 70–77.
27. M. Orio, D. A. Pantazis and F. Neese, Density functional theory[J], Photosynth. Res., 2009, 102(2–3), 443–453.
28. X. D. Hou, R. C. Shi and L. P. Ye, China Trop. Med., 2005, 5(8), 1738–1740.
29. J. Anglister and I. Z. Steinberg, J. Chem. Phys., 1983, 78, 5358.
30. Edition of Editorial Board of Chinese Macropaedia, Chinese Macropaedia Biology (II), Chinese Macropaedia Press, Beijing, 1991, p. 1374.
31. S. Liu and L. Kong, Anal. Sci., 2003, 19(7), 1055–1060.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c4ra04613g

---