Surface enhanced Raman scattering based reaction monitoring of in vitro decyclization of creatinine → creatine

Abstract

Creatinine → creatine decyclization is an important reaction wherein we obtain a beneficial compound from a toxic substance. Decyclization of creatinine at basic pH has been monitored in vitro by time series surface enhanced Raman scattering (SERS) using a silver island film. NH₂ scissoring, C=N and C=O stretching modes of creatinine serve as Raman markers for monitoring the decyclization reaction. Transition state calculations using DFT have revealed the path of the formation of creatine by the cleavage of the endocyclic C=N bond of creatinine. The Raman signatures of ring opening are clearly observed after 120 min at pH 8, and further increasing the pH increases the reaction rate even more, as the signatures are observed after 60 and 30 min at pH 10 and 12, respectively; however, the reversibility of the reaction is more prominent at higher pH. Therefore, pH 8 is the most favorable among the three pH values for the decyclization reaction to be stable at room temperature. The proper understanding of this reaction where a toxic substance is converted to a beneficial compound is expected to open the scope for further extensive research.

---

1. Introduction

The well-known cyclization reaction of creatine (CR) to form creatinine (CRN) is of significant interest due to several factors. First, this is the reaction that acts as the primary source of energy during any muscular activity. Second, the final product CRN is a body waste that serves as a marker to diagnose renal dysfunction, wherein, the CRN level rises in blood. Third, the cyclization reaction that takes place by enzymatic action in the body has been successfully mimicked in vitro on several occasions by using acidic media.^1–8 Fourth, since acidic pH supports cyclization, there is a possibility that a basic pH buffer might cause the decyclization of CRN to yield CR, as this reaction has been found to be reversible at neutral and basic pH.^9–11 This reverse reaction is immensely interesting because here, a toxic waste product, CRN, gets converted to a rather beneficial substance, CR.

Being an extension of our previous study,⁸ the primary objective of the present work is to monitor the reverse reaction, i.e. decyclization of CRN → CR in vitro at the molecular level with the help of Raman spectroscopy. Jiao Gao et al.¹² have reported a detailed Raman study of creatinine protonation and tautomerism in decimolar aqueous solutions for a wide pH range. Butler et al.¹³ reported that at basic pH, CRN undergoes a ring opening reaction. However, to the best of our knowledge, the mechanism and pH dependence of this reaction have not been studied before. The reaction was initiated in aqueous solution but the Raman signals were weak and no observable changes were found with time. Linear Raman signal was therefore not sufficient to detect the decyclization of CRN in this case. Fortunately, many sensitive Raman spectroscopic techniques have been developed and employed in recent times for the precise understanding of the mechanisms of various biochemical reactions.^8,14–18 Surface enhanced Raman scattering (SERS) is one such highly sensitive technique. This technique uses the exciting optical properties of nano-sized metal clusters, which enhance the Raman signal of the analyte attached to the metallic nanostructure.^19–25 The SERS technique has been effectively used on a few occasions to monitor biological/chemical reactions.^26–29

In the present work, the course of the CRN decyclization reaction was monitored by SERS on Ag nanostructures,³⁰ due to its high sensitivity and reproducibility. Constant pH buffer solutions maintained at pH 8, 10 and 12 were added to the aqueous solution of CRN to observe the effect of basicity, leading to the decyclization process. Time series SERS spectra were recorded to examine the changes occurring in the course of the reaction. CRN is known to exist in two tautomeric forms, viz., imino and amino forms. However, in aqueous solution, CRN predominantly exists in its amino form.^13,31 As we monitored this reaction in aqueous solution, we have used the amino form of CRN for the theoretical calculations. To understand the molecular orientation on the silver surface, CR and CRN were optimized on a silver array containing nine silver molecules. Transition state calculations were performed using density functional theory (DFT) and the reaction path corresponding to the cleavage of the C–N bond of CRN leading to the formation of CR has been studied. The role of SERS in the precise understanding of the reaction path and the CRN decyclization process at the molecular level, which has been studied in this work, is expected to lead to applications in further research, especially for the monitoring of other important reactions.

2. Experimental details

CRN purchased from Merck, Darmstadt, Germany and CR purchased from Aldrich chemicals, USA were used without further purification. The experimental details for the production of silver island film for SERS have been described in detail elsewhere.³² Fig. 1a shows the atomic force microscopic (AFM) image of the used silver island film. The AFM tip (NSG10, NT-MDT) was used to obtain the topographic image of the silver island film. The Raman spectra were collected at room temperature with a Renishaw inVia Raman spectrometer, equipped with both 532 and 785 nm lasers. To avoid heating, the 532 nm line of a diode pumped solid-state laser, delivering a power of 5 mW mm⁻² was used for excitation in acquiring the Raman spectra. The incident laser beam was focused on the sample by a 50× short distance objective attached to the Leica DM 2500M microscope. The Raman scattered light was collected in back-scattering geometry by the same objective and a 2400 groove per mm grating was used as the dispersion element. Slit width of 50 μm was maintained throughout the measurement. The resolution of the spectrometer was better than 1 cm⁻¹, with a repeat deviation of less than ±0.2 cm⁻¹. Spectrometer scanning, data collection and processing were done by a dedicated computer using Wire 4.0 software. In order to study the decyclization reaction of CRN, we prepared a 10⁻⁴ M aqueous solution of CRN. The reaction was monitored in vitro using phosphate buffered saline solutions maintained at basic pH of 8, 10 and 12, with an accuracy of ±0.1. The SERS spectra were recorded at different time intervals after adding the reaction solution to the silver substrate. Fig. 1b shows the sample solution attached to the silver substrate, as observed by the Raman microscope at 50× magnification.

Fig. 1 (a) Atomic force microscopic (AFM) image of the silver island film. (b) Sample solution attached to the silver substrate as observed by Raman microscope at 50×.

3. Theoretical details

Molecular geometries of CRN and the neutral, zwitterionic and protonated forms of CR in the gas phase were fully optimized and were reported in our previous work.⁸ Density functional theory is an extensively used technique for the optimization of molecular structures and for obtaining meaningful structure–spectra correlation.^33–36 The density functional theoretical (DFT) method, having Becke's non local three-parameter exchange along with the Lee–Yang–Parr correctional functional (B3LYP),^37,38 was employed for the calculations. We used two different basis sets: LanL2DZ, which is a double ζ quality basis set that takes into consideration the Los Alamos effective core potential for Ag,³⁹ and a 6-311++G(d,p) basis set for the carbon, nitrogen, oxygen and hydrogen atoms. To include the effect of bulk solvent, we optimized the geometries of CRN and CR in aqueous solution using the polarized continuum model (PCM),⁴⁰ as shown in Fig. 2a. For modeling the interaction of CR and CRN with the Ag surface, we constructed a one-dimensional planar array of Ag atoms and fixed the Ag–Ag bond length at 2.62 Å.^41,42 CR and CRN molecules were separately placed near the Ag array and allowed to move freely on the surface to obtain their minimum energy configurations. The geometries were optimized (Fig. 2b) at the B3LYP/6-311++G(d,p)/LANL2DZ level of theory by freezing the position of the Ag atoms. The geometry of the transition state (TS) between CRN and CR was also optimized in the gas phase (Fig. 3), just to observe the cleavage of the endocyclic C–N bond of CRN. Vibrational analysis was performed in order to make sure that each total energy extremum obtained was genuine; i.e. each minimum has all real frequencies and TS has only one imaginary frequency. For the theoretically obtained Raman spectra, Raman activity was converted to Raman intensity by the process mentioned elsewhere.^33–35 All the calculations were performed by Gaussian 03 software.⁴³ For visualization of the optimized structures and vibrational modes, Gauss View 4.1 (ref. 44) was used.

Fig. 2 (a) DFT/B3LYP/6-311++G(d,p) optimized structures of CRN and neutral, zwitterionic and protonated forms of CR in aqueous solution showing a schematic representation of the in vitro CRN → CR decyclization reaction. (b) DFT/B3LYP/6-311++G(d,p)/LANL2DZ optimized adsorption geometries of CR and CRN on the Ag surface.

Fig. 3 Decyclization of CRN to CR by C=N bond cleavage through the transition state (TS), using 1 water molecule in the gas phase obtained at the DFT/B3LYP/6-311++G(d,p) level of theory.

4. Results and discussion

4.1 Geometry optimizations and structural calculations

4.1.1 Molecular orientation of CRN and CR on the silver surface. The optimized structures of CRN and CR molecules on the Ag array, containing nine Ag molecules, using DFT at the B3LYP/6-311++G(d,p) level of theory for O, N, C and H atoms, and B3LYP/LANL2DZ for silver atoms, suggest that the molecules are attached perpendicular to the Ag surface via the NH₂ and CH₃ groups, as shown in Fig. 2b. The endocyclic C=N bond, which is cleaved during the decyclization reaction does not adsorb to the Ag surface and thus ring opening can be detected by the SERS technique.

4.1.2 Transition state calculations of the CRN → CR decyclization reaction. Transition state calculations were performed in order to verify the ring opening scheme of CRN using one water molecule. Fig. 3 shows the structures of reactant complex (CRN), transition state (TS) and product (CR) for the reaction studied in the gas phase. In going from CRN to TS, the C5–N4 bond length increased from 1.364 to 1.626 Å. The activation energy for this reaction step predicted at the B3LYP/6-311++G(d,p) level is 54.1 kcal mol⁻¹ (Fig. 3). The activation energy is high due to the absence of a solvent medium. Activation energies were calculated from the Gibbs free energy differences between the optimized TS and the optimized CRN. The transition state calculations show that during the decyclization reaction, the water molecule breaks into H⁺ and OH⁻ ions, and at the TS, the moving proton H17 of water is located between the O15 and N4 atoms (Fig. 3). The O15–H17 and N4–H17 distances are 1.367 and 1.166 Å, respectively. This reaction was predicted to be exothermic, as shown by the ZPE-corrected barrier (54.1 kcal mol⁻¹), and released energy of 16.2 kcal mol⁻¹. The transition state calculations give the idea of C–N bond cleavage and suggest the reaction path for the decyclization reaction.

4.2 SERS monitoring of CRN → CR decyclization reaction

The time series surface enhanced Raman spectra at three different pH values (8, 10 and 12) provide signatures of decyclization of CRN to CR at the molecular level. In this section, the significant changes in spectra with time have been discussed for the reaction at pH 8 in two different wavenumber ranges and the observations have been compared to those of the reaction monitored at pH 10 and 12. It is worth mentioning that the SERS spectra of CRN in urine^45–47 and CR in food additives⁴⁸ were previously reported using silver colloid, where the reported CRN spectra showed enhancement in the lower wavenumber range (<1100 cm⁻¹), but weak signal at higher wavenumbers (>1100 cm⁻¹), and CR showed very weak SERS signals. The SERS spectra in the present work resemble the previously reported spectra and also present a few additional bands for CRN in the higher wavenumber range. They also have better quality and reproducibility for both CRN and CR, due to the use of silver island films as SERS substrates.

4.2.1 Decyclization of creatinine to creatine monitored at pH 8. The time series SERS spectra of CRN at pH 8 in the wavenumber range 400–1800 cm⁻¹ are shown in Fig. 4. The decyclization reaction is very sensitive to pH and was initiated after the pH 8 buffer was added. The basic pH buffer provides OH⁻ ions to the medium. These OH⁻ ions catalyze the reaction and participate in the cleavage of the C–N bond, resulting in ring opening of the CRN. Additionally, due to the catalytic action of the OH⁻ ions, which also influences the reversibility of the reaction, the initial, final as well as the intermediate products are present in different proportions at any instant during the reaction. The amount of conversion of CRN to CR can be controlled by maintaining the concentration, pH and temperature of the solution.

Fig. 4 Room temperature time series SERS spectra of 10⁻⁴ M aqueous CRN solution at pH 8 at different time intervals in the range (a) 1100–1800 cm⁻¹ and (b) 400–1100 cm⁻¹, showing the signature of decyclization of CRN to CR after ∼120 min.

For comparison, and in order to understand the changes observed during the reaction, we also recorded the SERS spectra of aqueous solutions of CRN and CR (Fig. 5). The SERS spectra are quite different from the Raman spectra of the powder form, due to the difference in geometry, which results in the difference in polarizability between bulk CRN and CR, and when adsorbed on the Ag surface. Raman bands have been assigned by carefully studying the band assignments from reported literature^45–51 and cross-referencing with the theoretically derived spectra of Ag–CRN and Ag–CR complexes (Fig. S1, ESI†).

Fig. 5 Comparison between the SERS spectra of CRN and CR in 10⁻⁴ M aqueous solution, and the Raman spectra of CRN and CR in powder form. The SERS spectra are quite different from the Raman spectra of the powder form.

Range 1: 1100–1800 cm⁻¹. The bands at 1603 and 1633 cm⁻¹ (Fig. 4a) have a contribution from the endocyclic C=N and C=O stretching modes of CRN, and the band at 1653 cm⁻¹ corresponds to the NH₂ scissoring mode of CRN. The two main observed differences in the vibrational features of CR and CRN are the absence of the C=N stretching mode of the CRN ring in CR due to ring opening, and the appearance of the C=O stretching and NH₂ scissoring modes in CR at lower wavenumbers, compared to CRN (Fig. 5). These features have been considered as the main Raman signatures for visualizing the presence and conversion of CRN to CR during the decyclization reaction.

During the reaction we observed that the intensities of the C=O stretching and NH₂ scissoring modes (1603 and 1653 cm⁻¹) of CRN vary with time and show shifts towards lower wavenumber, up to 120 min. The C=N stretching mode (1633 cm⁻¹) of the CRN ring disappears in the spectra after 120 min. Disappearance of the Raman band at 1633 cm⁻¹ gives evidence of the breaking of the C=N bond of creatinine, leading to the formation of creatine, i.e. the SERS signature of decyclization. The band at 1362 cm⁻¹ corresponds to the C–N stretching of CRN, which is associated with N–CH₃ and CH₂ groups. This band shows a lower wavenumber shift as the reaction proceeds. This lower wavenumber shift by ∼40 cm⁻¹ after 120 min can be explained by considering the charge distribution over 5C–6C–2N–8C atoms. As the dihedral angle 5C–6C–2N–8C changes from 7.23° in CRN to 89.29° in CR, the charge cloud shifts from 8C towards the 2N site. This suggests the intramolecular rotation about the C–N bond during ring opening. Another band at 1262 cm⁻¹ corresponds to the combination of the C=N stretching associated with the NH₂ group and CH₂ wagging vibrations. The weakening of this bond is evident from the observed lower wavenumber shift in the 1262 cm⁻¹ band with time, suggesting the cleavage of the endocyclic C=N bond at basic pH. We compared the spectrum observed after 120 min with the SERS spectrum of CR that shows close resemblance to it, which suggests that the changes observed after 120 min are due to the decyclization of CRN to form CR.

The Raman spectra recorded at later stages suggest the reversibility of the reaction as the C=O and NH₂ scissoring vibrations again shift towards higher wavenumbers, characterizing the formation of CRN. The band at 1450 cm⁻¹, which corresponds to CH₃ deformation, shows the appearance of a doublet structural feature during the reaction with a shift to high wavenumber that may be due to the coexistence of CR and CRN in the solution. Several other changes in this range suggest the decyclization of CRN to form CR up to 120 min; however, after 120 min both CR and CRN are found to coexist, due to the reversibility of the reaction.

Range 2: 400–1100 cm⁻¹. The time series SERS spectra at pH 8 in this range are shown in Fig. 4b. This range covers the vibrations of the CRN ring and many twisting, wagging and deformation vibrations of CH₃, CH₂ and NH groups of CRN in the beginning of the reaction. As the reaction proceeds, the band at 666 cm⁻¹ corresponding to ring vibrations almost disappears after 120 min, confirming the ring opening of CRN to form the linear structure of CR. The other bands at 753, 814 and 843 cm⁻¹ that involve NH₂ group vibrations show changes during the reaction. The disappearance and decrease in the intensities of the ring vibrations after 120 min provide the evidence of ring opening, and their re-appearance at later stages suggests the reversibility of the reaction. Additionally, the change in the charge distribution in the system due to ring opening is responsible for the wavenumber shifts.

4.2.2 pH dependence of the decyclization of CRN → CR (at pH 8, 10 and 12). Fig. 6 shows the comparative time series SERS spectra of an aqueous solution of CRN at (a) pH 8, (b) pH 10 and (c) pH 12. The peak positions of the bands at three pH values may differ due to the SERS effect and experimental conditions. When we compare the SERS spectra of CR and CRN in aqueous solution, we can clearly observe changes in the positions of C=O stretching, NH₂ scissoring and C=N stretching modes, due to the ring opening of CRN. We have considered the lower wavenumber shifts in NH₂ scissoring and C=O stretching vibrations as the Raman signatures of decyclization of CRN to CR. The endocyclic C=N stretching vibration of CRN also disappeared due to ring opening. These signatures are observable in the SERS spectra after 120 min at pH 8, 60 min at pH 10 and just 30 min at pH 12. This suggests that CRN is converted to CR faster, when the basicity of the medium is increased. As we have already discussed, the conversion of CRN to CR is reversible. This is confirmed by the re-appearance of the signature bands corresponding to CRN with low intensity at some later stages of the reaction. By comparing the spectra in the entire spectral range for all three basic pH values (pH 8, 10, 12), we may infer that the CR formed as a result of decyclization is comparatively more stable at pH 8, where CR and CRN both coexist after 120 min, while in the case of pH 10 and 12, no signature peak regarding the coexistence of both is observed. At higher pH, the signatures of CRN appear very soon after decyclization. At pH 8, which is closest to body pH, decyclization is significant, with CR having greater stability despite the low conversion rate.

Fig. 6 Comparative room temperature time series SERS spectra of 10⁻⁴ M aqueous CRN solution at (a) pH 8, (b) pH 10 and (c) pH 12 at different time intervals in the range 400–1800 cm⁻¹. The signature of decyclization is observed after ∼120 min at pH 8, ∼60 min at pH 10 and ∼30 min at pH 12.

5. Conclusion

The present study shows that a basic pH buffer solution causes in vitro decyclization of CRN to CR, which is a reaction of significant interest. The surface enhanced Raman scattering (SERS) technique using silver island film has led to the precise monitoring of the in vitro decyclization process with respect to time. The changes in several Raman bands as the reaction proceeds have been discussed in detail by comparing the SERS spectra of CRN and CR. The increase in bond length of the endocyclic C=N bond of CRN and the formation of CR by cleavage of this C=N bond was observed by transition state calculations using DFT. The decyclization process is marked by the disappearance of the endocyclic C=N stretching mode and lower wavenumber shifts in NH₂ scissoring and C=O stretching Raman modes of CRN after ring opening. At pH 8, the Raman signatures for decyclization are observed after 120 min at room temperature, while the rate of decyclization is much greater at higher pH values (pH 10 and 12, where the signatures appear after 60 and 30 min, respectively). However, since the reaction is reversible, traces of CRN will be present throughout the reaction, along with CR. At higher pH values, although decyclization occurs faster, the reversibility is also more pronounced. Thus, pH 8, which is also closest to the pH of human blood, is the most favorable amongst the three pH values for the decyclization reaction to be stable at room temperature. Whether the reversibility of the reaction can be reduced by modifying certain physical parameters to get stable conversion of CRN to CR is a topic for further research. The proper understanding of this reaction, where a toxic substance, viz. CRN, is decyclized back to its linear precursor CR, which is a beneficial compound acting as an energy provider for the body, is expected to open the scope for further extensive research in this field.

Acknowledgements

RKS, DG, SKS and VD are thankful to German Research Foundation (DFG) (Grant No. GZ: DE 1376/6-1 AOBJ: 595508), Germany for financial support. RKS is grateful to DST, India for DST-FIST program. DG is thankful to CSIR, India and SKS is grateful to DRDO, India for providing research fellowships. The authors are thankful to Prof. P. C. Mishra for his valuable suggestions.

References

1. M. Wyss and R. Kaddurah-Daouk, Physiol. Rev., 2000, 80, 1107.
2. T. Wallimann, M. Wyss, D. Brdiczka, K. Nicolay and H. M. Eppenberger, Biochem. J., 1992, 281, 21.
3. G. Edgar and R. A. Wakefield, J. Am. Chem. Soc., 1923, 45(2), 2242.
4. C. Lillocci and C. Vernon, Gazz. Chim. Ital., 1981, 111, 23.
5. A. Witkowska, Acta Aliment. Pol., 1985, 9(2), 263.
6. X. Wang and Q. Yin, J. Chem. Eng. Chin. Univ., 2003, 17(5), 569.
7. B. J. Diamond, Theses, Dissertations and Capstones, Paper 564, M.S., College of Science, Marshall University, Huntington, WV 25755, USA, 2005, http://mds.marshall.edu/etd/564.
8. D. Gangopadhyay, P. Sharma, S. K. Singh, P. Singh, N. Tarcea, V. Deckert, J. Popp and R. K. Singh, Chem. Phys. Lett., 2015, 618, 225.
9. G. Edgar and H. E. Shiver, J. Am. Chem. Soc., 1925, 47, 1179.
10. R. K. Cannan and A. Shore, Biochem. J., 1928, 22(4), 920.
11. N. J. Fuller and M. Elia, Clin. Chim. Acta, 1988, 175, 199.
12. J. Gao, Y. Hua, S. Li, Y. Zhang and X. Chen, Chem. Phys., 2013, 410, 81.
13. A. R. Butler and C. Glidewell, J. Chem. Soc., Perkin Trans. 2, 1985, 9, 1465.
14. P. Sharma, D. Gangopadhyay, P. Singh, P. C. Mishra, V. Deckert, J. Popp and R. K. Singh, Chem. Phys. Lett., 2014, 613, 127.
15. O. Svenson, M. Josefson and F. W. Langkilde, Chemom. Intell. Lab. Syst., 1999, 49, 49.
16. M. Lee, H. Kim, H. Rhee and J. Choo, Bull. Korean Chem. Soc., 2003, 24(2), 205.
17. P. D. I. Fletcher, S. J. Haswell and X. Zhang, Electrophoresis, 2003, 24, 3239.
18. M. P. Houlne, C. M. Sjostrom, R. H. Uibel, J. A. Kleimeyer and J. M. Harris, Anal. Chem., 2002, 74, 4311.
19. M. Sackmann and A. Materny, J. Raman Spectrosc., 2006, 37, 305.
20. L. Delfino, A. R. Bizzarri and S. Cannistraro, Biophys. Chem., 2005, 113(1), 41.
21. K. Kneipp, H. Kneipp and J. Kneipp, Acc. Chem. Res., 2006, 39(7), 443.
22. L. Baia, M. Baia, J. Popp and S. Astilean, J. Phys. Chem. B, 2006, 110(46), 23982.
23. K. K. Strelau, T. Schüler, R. Möller, W. Fritzsche and J. Popp, ChemPhysChem, 2010, 11(2), 394.
24. L. Yang, P. Li and J. Liu, RSC Adv., 2014, 4, 49635.
25. E. L. Keller, N. C. Brandt, A. A. Cassabaum and R. R. Frontiera, Analyst, 2015, 140, 4922.
26. L. Cui, P. Wang, Y. Fang, Y. Li and M. Sun, Sci. Rep., 2015, 5, 11920.
27. H. W. Cheng, W. Q. Luo, G. L. Wen, S. Y. Huan, G. L. Shena and R. Q. Yu, Analyst, 2010, 135, 2993.
28. G. Qi, C. Fu, G. Chen, S. Xua and W. Xu, RSC Adv., 2015, 5, 49759.
29. Z. Q. Tian and B. Ren, Annu. Rev. Phys. Chem., 2004, 55, 197.
30. A. Rasmussen and V. Deckert, J. Raman Spectrosc., 2006, 37, 311.
31. S. Du Pre and H. Mendel, Acta Crystallogr., 1955, 8, 311.
32. R. Stöckle, V. Deckert, C. Fokas and R. Zenobi, Appl. Spectrosc., 2000, 54, 1577.
33. A. Singh, D. Gangopadhyay, J. Popp and R. K. Singh, Spectrochim. Acta, Part A, 2012, 99, 136.
34. S. Mishra, D. Gangopadhyay, A. Singh, J. Popp and R. K. Singh, Vib. Spectrosc., 2013, 65, 116.
35. P. Sharma, D. K. Singh, V. Gupta, B. P. Asthana, P. C. Mishra and R. K. Singh, Spectrochim. Acta, Part A, 2013, 116, 74.
36. P. Sharma, D. Gangopadhyay, S. Umrao, S. Kumar, A. K. Ghosh, P. C. Mishra and R. K. Singh, J. Raman Spectrosc., 2016, 47, 468.
37. A. D. Becke, J. Chem. Phys., 1993, 98, 5648.
38. C. Lee, W. Yang and R. G. Parr, Phys. Rev. B: Condens. Matter Mater. Phys., 1988, 3, 785.
39. J. C. Cook, C. M. P. Cuypers, B. J. Kip, R. J. Meier and E. Koglin, J. Raman Spectrosc., 1993, 24, 609.
40. J. Tomasi, B. Mennucci and R. Cammi, Chem. Rev., 2005, 105, 2999.
41. S. Bhunia, S. Forster, N. Vyas, H. C. Schmitt and A. K. Ojha, Spectrochim. Acta, Part A, 2015, 151, 888.
42. A. Campion and P. Kambhampati, Chem. Soc. Rev., 1998, 27, 241.
43. M. J. Frischet al., Gaussian 03, Revision A.1, Gaussian, Inc., Pittsburgh, 2003.
44. R. Dennington II, T. Keith, J. K. Eppinnett, W. L. Hovell and R. Gilliland, GaussView 4.1, Semichem, Inc., Shawnee Mission, KS, 2003.
45. R. Stosch, A. Henrion, D. Schiel and B. Güttler, Anal. Chem., 2005, 77, 7386.
46. T. L. Wang, H. K. Chiang, H. H. Lu and F. Y. Peng, Opt. Quantum Electron., 2005, 37, 1415.
47. M. Li, Y. Du, F. Zhao, J. Zeng, C. Mohan and W. C. Shih, Biomed. Opt. Express, 2015, 6(3), 849.
48. E. Podstawka, M. Światłowska, E. Borowiec and L. M. Proniewicz, J. Raman Spectrosc., 2007, 38, 356.
49. D. Gangopadhyay, P. Sharma and R. K. Singh, Spectrochim. Acta, Part A, 2015, 150, 9.
50. K. Vikram, S. Mishra, S. K. Srivastava and R. K. Singh, J. Mol. Struct., 2012, 1012, 141.
51. D. Gangopadhyay, S. K. Singh, P. Sharma, H. Mishra, V. K. Unnikrishnan, B. Singh and R. K. Singh, Spectrochim. Acta, Part A, 2016, 154, 200.

---

Footnotes

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c6ra03674k

‡ Main workers with equal contribution.

---