The mechanism of different sensitivity of meso-substituted and unsubstituted cyanine dyes in rotation-restricted environments for biomedical imaging applications

Abstract

A series of cyanine dyes have been excellent environment-sensitive fluorescent probes in live cells as well as in living tissue imaging due to their low quantum yields in solution but also their large fluorescence enhancements in microenvironments. In this paper, a new mechanism of different environmental sensitivity of meso-substituted and unsubstituted cyanine dyes in rotation-restricted environments (DNA or viscosity) for biomedical imaging applications is proposed on the basis of experimental observations and quantum-chemically calculated potential energy curves for twisting around the various C–C bonds in the polymethine chain. Computational studies on these dyes reveal substituents on the bridge decrease the energy barriers for rotations around different chemical bonds, which play a major role in reducing fluorescence quantum yield in a free state and providing a high environmental sensitivity for the meso-substituted cyanine dyes. The energy barriers and energy gaps to the rotations of C–C bonds strongly depend on the choice of substituents in the meso-position of the polymethine chain. The results might provide a foundation for the interpretation of the behavior of the dyes and are useful for the future design of new cyanine fluorophores.

---

1. Introduction

Cyanine dyes, covering the spectrum from Ultraviolet (UV) to Infrared (IR) with different structures, have attracted increasing interest in biotechnology due to their unique photophysical properties and fluorogenic behavior towards macromolecules or microenvironments.^1–10 Covalently bound versions of symmetrical cyanine dyes, particularly Cy3 and Cy5,^11,12 are widely used in DNA/RNA microarray, peptide and protein fluorescent labeling. The reasons for applying indocyanines in covalent labeling can be ascribed to their relatively high fluorescence quantum yields and no associations with nucleic acids as well as proteins. The complexity of noncovalent DNA binding by symmetrical cyanines has led to very few applications. A specific example is the monocationic dye DiSC₂(5),¹³ which binds to various types of DNA by intercalation as well as by minor groove binding to be a monomer or a cofacial dimer. The symmetrical cyanine dyes are among the classic fluorescent dyes but continue to be used for applications in a variety of fields, such as biomedical imaging and microenvironmental viscosity detection.^14–19 Shieh and coworkers²⁰ reported IR-780 iodide in the near-infrared (NIR) region and labeled with the radionuclide rhenium-188 (¹⁸⁸Re) for NIR fluorescence and nuclear imaging. Luby-Phelps¹⁶ reported a ratiometric system of fluorescent Cy5 and Cy3 attached to macromolecular Ficoll 70 to measure intracellular viscosity.

An interesting class of symmetrical cyanines features substitution of the methine bridge.^21–27 For example, Yarmoluk and coworkers²⁸ synthesized Cyan 2, an analogue of Cyan 46 bearing a methyl group at the beta (i.e., central) carbon of the trimethine bridge. Under identical conditions, the unsubstituted dye exhibits a less than 5-fold increase in fluorescence in the presence of DNA while the β-methyl analogue displays ca. 100-fold fluorescence enhancement (Table 1). The difference can be traced to the background fluorescence of the two dyes in the absence of DNA. Our laboratory has demonstrated that conventional cyanine dye Cy5 is less sensitive to solution viscosity, while Cy5-CHO introducing an aldehyde group (CHO) rotor into the meso-position of the pentamethine chain has been used to measure the intracellular viscosity.²⁹ The restraining of rotation for cyanine dyes in DNA or viscous media results in strong fluorescence. Furthermore, a strong foundation of fundamental knowledge concerning how the dye structure influences their biological applications and the mechanisms for fluorogenic behavior assists in the design of new dyes with improved properties.^30–33

Table 1 Fluorescence intensity and fluorescence intensity enhancement recorded in Tris buffer and in complexes with DNA for Cyan 46 and Cyan 2 (values from ref. 28)

Dye	F (buffer)	F (DNA)	Enhancement
Cyan 46	560	2750	4.6
Cyan 2	24.5	4350	178

---

Quantum chemistry calculations have been successfully used to design the new fluorescent molecular rotors as viscosity sensors and study the photophysical properties of fluorophores and the mechanisms for fluorogenic behavior.^34–37 Our previous research measured the photophysics and fluorescence quantum yields of conventional cyanine dyes and clarified the nature of the different environmental sensitivity of symmetrical (Cy) and unsymmetrical (TO) cyanine dyes.³⁸ The results show that the symmetric dyes have substantially higher barriers to torsional motion. It also suggests that engineering the characteristics of the barrier is a useful design strategy for developing improved dyes.

Conventional symmetric cyanine dyes Cy are relatively high fluorescence quantum yields and less sensitive to the environment. The fluorescence quantum yields depend significantly on structural alterations in the polymethine chain, along with strategies through which we can improve environmental sensitivities of cyanine dyes. Meso-substituted and unsubstituted cyanine dyes are very different in environmental sensitivity and application; however the mechanism is still unclear. It is well known that the study of the nature of the different environmental sensitivity of meso-substituted and unsubstituted cyanine dyes is a starting point for further design of the dyes with required properties.

In the present paper, we compared the potential energy curves of Cyan 46 and Cyan 2 dyes by theoretical calculation and clarified the mechanism of the different environmental sensitivity of Cyan 46 and Cyan 2 (Scheme 1). Following the stepwise rational approach, we compared the photophysical properties of Cy5, Cy5-CH₃ and Cy5-CHO (Scheme 1) through experimental analysis and DFT computations, which further supports the proposed mechanism. We also have tried to synthesize meso-substituted Cy3 models (Cy3-CH₃ and Cy3-Ph, Scheme 1). Revealing photophysical properties for Cy3 and its analogues will inevitably promote research on structure and application. The rotation of polymethine chains cannot be completely suppressed in high viscosity solution. In this paper, according to the level of energy barrier, we predict that meso-substituted Cy3 models can be inhibit efficient nonradiative decay by a more constrained environment, which will guide the experimental effort towards the most promising targets. Revealing the photophysical properties for the dyes from a theoretical perspective will be useful for the design of new fluorophores.

2. Experimental section

2.1. Sample preparation

All solvents and reagents used were reagent grade. Tris base was from Promega Co., (USA). Doubly purified water, prepared using the MilliQ system, was used in all experiments. Cy3, Cy5, Cy5-CH₃, Cy5-CHO, Cy3-CH₃ and Cy3-Ph were synthesized according to the previous literature.^26,39,40 The solutions of Cy dyes were typically obtained from 1 mM stock solutions in DMSO. A buffered solution of Rhodamine B was typically obtained from 1 mM stock solution in ethanol. Fluorescence spectra were obtained with a Felix and Time-Masters PTI-C-700 system. Visible absorption spectra were determined using an Agilent HP-8453 spectrophotometer.

2.2. Viscosity detection

The solutions of dyes in different systems were prepared by adding the stock solution (1 mM) to 3 mL of solvent mixture (H₂O, 99% glycerol–H₂O, Tris–HCl buffer) to obtain the final concentration of the dye (1.0 μM). The solutions were measured in a UV-Vis spectrophotometer and a fluorescence spectrophotometer.

2.3. Quantum yield detection

Rhodamine B (Φ_F = 0.65 in ethanol) was used as the standard to calculate the quantum yields and use the following equation:

Φ_x = Φ_s(F_x/F_s)(A_s/A_x)(λ_exs/λ_exx)(n_x/n_s)²

where Φ represents quantum yield; F stands for integrated area under the corrected emission spectrum; A is absorbance at the excitation wavelength; λ_ex is the excitation wavelength; n is the refractive index of the solution (because of the low concentrations of the solutions (10⁻⁷ to 10⁻⁸ mol L⁻¹), the refractive indices of the solutions were replaced with those of the solvents); and the subscripts x and s refer to the unknown and the standard, respectively.

2.4. Computational approach

To gain a good theoretical understanding of the different nature of these fluorescent states, excited-state geometry optimization and potential energy curves with reactive coordinates are necessary. The molecular geometries (the atomic standard orientations after geometry optimization of ground state and excited state have been given in the ESI†) were investigated using density functional theory (DFT) and time-dependent density functional theory (TDDFT) calculations, respectively. For sure, one major problem of (TD-) DFT is the choice of an adequate functional. The reliability of the results depends significantly on the selected functional. It is worth recalling that a test with a large number of functionals was performed for photochemical properties for the TO-1 in our previous work (Table S1†). Only one parameter Becke's 1998 functional (HFB), which includes the Slater exchange along with corrections involving the gradient of the density,^38,41 is the most reasonable choice for good agreement with the experimental results (Table S2†), and used in both the DFT and TD-DFT methods in the sequential work. The triple-ζ valence quality with one set of polarization functions (TZVP)⁴² was chosen as basis sets throughout, which is an appropriate basis set for such an organic compound.^43–45 We have tested the stability of the DFT wave function, which is already stable for all the dyes in this work. Moreover, electronic excitation energies of low-lying electronically excited states were also computed by the TDDFT method. A continuum solvation model, COSMO (conductor-like screening model), was used for the consideration of solvent effects in aqueous solution.⁴⁶ In addition, the S₀ potential energy curves were qualitatively scanned by constrained optimizations keeping the dihedral angles fixed at a series of values. The excited-state potential energy curves were obtained by calculating the Franck–Condon transition energies for the ground-state optimized structures at fixed dihedral angles using TDDFT method. This type of approach has been successfully reported in several recent papers for studying torsional potentials.^45,47–49 All calculations on electronic structures were carried out using the Gaussian 09 program suite.⁵⁰

3. Results and discussion

3.1. Mechanism of different environmental sensitivity of meso-substituted and unsubstituted cyanine dyes

Photophysical data for Cyan 46 and Cyan 2 in both Tris buffer and DNA complex in Tris buffer solution have been reported previously,²⁸ and theoretical analysis using the DFT and the TDDFT method is presented in this section. Table 1 (values from ref. 28) depicts fluorescence intensity and fluorescence intensity enhancements of cyanine dyes, based on analysis in the unbound state and upon nucleic acid binding. In complexes with DNA, the Cyan 2 exhibits ca. 178-fold increase in fluorescence intensity, while the Cyan 46 displays ca. 4.6-fold fluorescence enhancement under identical conditions. The different fluorescence enhancement can be traced to the background fluorescence of the two dyes in the absence of DNA, which is closely related to the efficiency of nonradiative deactivation. The origin of this behavior has been attributed to a change in the stereochemistry of the polymethine chain. We have proved that the rotation about the skeleton plays a major role in reducing fluorescence quantum yield and providing a low fluorescent background in the free state for cyanine dyes. How substitution in the meso-position of the polymethine chain effects the rotation about the skeleton and promotes fluorescence enhancement binding to DNA is unclear. Experimental analysis and DFT computations were conducted to gain further mechanistic insight into the different environmental sensitivity of meso-substituted and unsubstituted cyanine dyes.

We calculated the absorption maximum and emission maximum of all the molecules using the TD-DFT/HFB/TZVP method with COSMO solvation model, and compared them with the experimental data. A methyl group on the bridge leads to nearly no change in the absorption and emission spectra of the dye. The results in Table 2 show that computational absorption bands of all the dyes are consistent with the experimental data. This strong correlation between experiment and theory indicates that the calculations are useful for understanding the rotational motion about the skeleton of the cyanine dyes.^43,47 The molecular orbitals involved in the S₀ → S₁ transitions have been calculated for the two dyes in Fig. 1, and each of them proved to be a dominant π–π*-type transition. The S₀ → S₁ transition is from the highest occupied molecular orbital (HOMO) to the lowest unoccupied molecular orbital (LUMO). The meso-methyl substituent has a slight influence on the HOMO and LUMO energy levels and HOMO–LUMO gap, which is the main reason for the spectra being significantly unchanged for Cyan 2 compared to Cyan 46.

Table 2 The maximum absorption and emission wavelengths of free Cyan 46 and Cyan 2 (exptl λ from ref. 28)

Dye	Exptl λ (nm)		Calcd λ (nm)
	Abs	Flu	Abs	Flu
Cyan 46	553	577	535	563
Cyan 2	541	570	532	561

---

Fig. 1 HOMO and LUMO for Cyan 46 (left) and Cyan 2 (right).

Since there are several chemical bonds in cyanine dyes that may be involved in rotation, the application of quantum chemical calculations will be able to explain which one is the main cause of nonradiative deactivation. The angles for rotation about various bonds are defined in Scheme 1. The excited singlet state formed by twisting of the skeleton of the cyanine dyes possesses a remarkably low energy gap to the ground state, which causes mainly nonradiative deactivation.^51,52 The minimal energy conical intersection geometries are the most probable nonradiative deactivation sites.⁵³ In addition, this low yield is a consequence of an ultrafast nonradiative process which has been attributed to barrier height and position photoinduced rotational motion about the C–C bonds of the polymethine chain.^54,55 Fig. 2 presents the calculated potential energy curves for different twist dihedral angles in the ground states and low-lying electronic excited states of Cyan 46. The probability of the rotation about φ₂ should be very small due to the relatively high energy barrier of 11.00 kcal mol⁻¹ (in Table 3 and Fig. 2b) to the rotation in the excited state. Thus, the lowest energy path for the photoisomerization of the Cyan 46 is associated with the rotation about the φ₁ bond close to the benzothiazole moiety of the polymethine chain, with an energy barrier of 6.59 kcal mol⁻¹(in Table 3 and Fig. 2a). For Cyan 2, the potential energy curves (see Fig. 3c) suggest that the rotation about the φ₃ bond is easy, since there is nearly no barrier to rotation in the S₁ state. However, the energy gap between the S₁ and S₀ states of rotation about the φ₃ bond (with torsion angles of 90° at the C–C) is very large in comparison to the energy gap about the φ₁ bond and φ₂ bond (53.65 kcal mol⁻¹ for φ₃, 41.68 kcal mol⁻¹ for φ₂ and 40.35 kcal mol⁻¹ for φ₁ in Table 3 and Fig. 3), therefore, nonradiative decay through rotation about the φ₃ is not significant. From Fig. 3a and b and Table 3, the rotations about the φ₁ bond and φ₂ bond have an energy barrier of 3.87 kcal mol⁻¹ and 5.53 kcal mol⁻¹, respectively. Therefore, Cyan 2 may undergo nonradiative deactivation by rotating along with φ₁ and φ₂. Nevertheless, this former possibility is strengthened by comparing the energy barriers of the dihedral angles φ₁ and φ₂. The lowest energy path for the photoisomerization of the Cyan 2 is associated with the rotation about the φ₁. The rotation about the favored φ₁ bond drives the molecule toward low energy gap between ground state and the first excited state, which causes mainly nonradiative deactivation.

Scheme 1 Geometric structures for symmetrical and unsymmetrical dyes (see text for further details).

Fig. 2 Energy levels of the S₀ (black) and S₁ (red) states of Cyan 46 rotation along with dihedral angle φ₁ (a) and φ₂ (b) calculated at the TD-DFT/HFB/TZVP level with COSMO solvation model.

Table 3 Energy barrier E_a of S₁ and energy gap E_gap between the S₀ and S₁ states for Cyan 46 and Cyan 2 studied in this work for rotation around different dihedral angles (φ₁, φ₂ and φ₃). Energy gaps between the S₀ and S₁ states are at 90° for the two dyes

Dihedral angle	Cyan 46		Cyan 2
	Ea (kcal mol^−1)	Egap (kcal mol^−1)	Ea (kcal mol^−1)	Egap (kcal mol^−1)
φ1	6.59	42.53	3.87	41.68
φ2	11.00	41.82	5.53	40.35
φ3	—	—	0.98	53.65

---

Fig. 3 Energy levels of the S₀ (black) and S₁ (red) states of Cyan 2 rotation along with dihedral angle φ₁ (a), φ₂ (b) and φ₃ (c) calculated at the TD-DFT/HFB/TZVP level with COSMO solvation model.

Table 3 shows the results for the energy gap between the S₀ and S₁ states and energy barrier E_a of S₁ for the different rotation dihedral angles of Cyan 46 and Cyan 2. Obviously, the energy barrier decreases from 6.59 kcal mol⁻¹ for Cyan 46 to 3.87 kcal mol⁻¹ for Cyan 2 for rotation along the dihedral angle φ₁. Similar results at the same time appear in dihedral angle φ₂ (11.00 kcal mol⁻¹ for Cyan 46 and 5.53 kcal mol⁻¹ for Cyan 2). It has been implicit in our discussion of the C–C bond rotation of the two dyes that substitution on the bridge is expected to decrease the energy barriers for rotations around different chemical bonds, which play a major role in reducing fluorescence quantum yield in a free state and providing a high environmental sensitivity for the meso-substituted cyanine dyes. The excited-state barriers calculated in this work are slightly overestimated, due to the fact that the molecule was kept rigid except for rotation about a particular bond. But this strategy allows us to separate the effect of a single rotation from other rotations. It was found that allowing the molecule to relax along the remaining internal coordinates made little difference to the qualitative conclusions.^{30,33–37,56}

3.2 Support for the proposed mechanism of different environmental sensitivity of meso-substituted and unsubstituted cyanine dyes

With the help of quantum chemical calculations, we have investigated the nature of the different environmental sensitivities of Cyan 2 and Cyan 46. The introduction of the methyl decreases the energy barrier of the rotation for the polymethine chain. In a similar fashion, cyanine dyes as fluorescent molecular rotors for viscosity sensors are extremely popular. These sensors show enhanced fluorescence emission in viscous environments compared with those in fluid solutions. In a viscous environment, the environmental factors restricting the internal molecular rotation of these dyes could lead to an evident increase in fluorescence quantum yields and first excited state lifetime. Furthermore, the proposed mechanism of different environmental sensitivity of meso-substituted and unsubstituted cyanine dyes were confirmed by similar structural changes in cyanine dyes such as Cy5, Cy5-CH₃ and Cy5-CHO.

Our previous research has shown that dyes with a low quantum yield corresponding to a low non-signal background will provide highly sensitive viscosity detection when the fluorescence enhancement is in the same target range. Table 4 depicts fluorescence quantum yields and fluorescence enhancements (fluorescence quantum yield enhancements) of cyanine dyes, based on analysis in the unbound state and in 99% glycerol. In 99% glycerol, Cy5 exhibits ca. 2.5-fold increase in fluorescence enhancements, while Cy5-CH₃ and Cy5-CHO display ca. 3.7-fold and ca. 6.7-fold fluorescence enhancements under identical conditions, respectively. Introduction of a methyl group into the meso-position of the pentamethine chain decreases fluorescence quantum yield in the free state and promotes fluorescence enhancement, which is very similar to Cyan 2 and supports the proposed mechanism of different environmental sensitivity of meso-substituted and unsubstituted cyanine dyes. For the cyanine dye Cy5-CHO with the aldehyde group (CHO) as substituent in the meso-position of the polymethine chain, a significant effect of substituent in the meso-position of the polymethine chain, i.e. a 6.7-fold increase in fluorescence enhancement. Cy5 and its derivatives with substituents in the meso-position of the polymethine chain have different background quantum yields.

Table 4 Fluorescence quantum yield (Φ_f) and fluorescence enhancement (fluorescence quantum yield enhancement) recorded in ethanol and 99% glycerol

Dye	Φf (ethanol)	Φf (99% glycerol)	Enhancement
Cy5	7.837 × 10^−2	1.932 × 10^−1	2.5-fold
Cy5-CH3	5.345 × 10^−2	1.965 × 10^−1	3.7-fold
Cy5-CHO	2.853 × 10^−2	1.894 × 10^−1	6.7-fold

---

The spectra of Cy5 and Cy5-CH₃ are similar (Table 5), the absorption/emission maxima of Cy5-CH₃ are slightly blue-shift. Cy5-CHO has a significant blue-shift of the maximum absorption peak (λ_abs = 600 nm) and emission peak (λ_em = 638 nm). The Stokes shift of Cy5-CHO is larger than that of Cy5 and Cy5-CH₃. The molecular orbitals involved in the S₀→S₁ transitions have been illustrated in Fig. 4 for all the molecules. DFT calculations (HFB/TZVP) showed that such wavelength shifts are ascribed to the effects of the substituents into the meso-position of the polymethine chain on the highest-occupied molecular orbital (HOMO) and lowest-unoccupied molecular orbital (LUMO) levels.

Table 5 The maximum absorption and emission wavelengths of free Cy5, Cy5-CH₃ and Cy5-CHO

Dye	Exptl λ (nm)		Calcd λ (nm)
	Abs	Flu	Abs	Flu
Cy5	637	660	591	636
Cy5-CH3	635	654	590	631
Cy5-CHO	600	638	580	622

---

Fig. 4 HOMO and LUMO for Cy5 (left), Cy5-CH₃ (middle) and Cy5-CHO (right).

We previously demonstrated that the nonradiative deactivation of Cy5 is dominated by the rotation of φ₁ in the first excited state (Fig. 5). The high-energy barrier to rotation around the φ₁ (11.78 kcal mol⁻¹) is the cause of the insensitivity to solution viscosity of Cy5. Cy5 with a long chain fixed rigidly by polymethylene bridges, has a higher background quantum yield of fluorescence and sharp decrease in the efficiency of sensitization. For Cy5-CH₃, the potential energy curves suggest that the rotation of the methyl group should be very easy since there is no energy barrier to rotation in the S₁ state (Fig. 6b). However, the energy gap between the S₁ and S₀ states of the methyl group rotation (with a torsion angle of 90° at the C–C) is 48.26 kcal mol⁻¹ and rather large, compared with 38.35 kcal mol⁻¹ for rotation about the φ₁ bond (Table 6), therefore, nonradiative decay through the channel is not significant. The energy barrier to rotation around the φ₁ is 8.51 kcal mol⁻¹ which is lower than that of Cy5 (Fig. 6a and Table 6). The changes in the excited energy barrier are similar to those discussed previously (in Section 3.1). Introduction of the methyl decreased the energy barrier around rotation at φ₁, which facilitated efficient nonradiative internal conversion and decreased background quantum yield. However, for Cy5-CH₃ with a longer chain, the methyl provides a slightly smaller influence on the energy barrier of the polymethine chain than that of Cyan 2.

Fig. 5 Energy levels of the S₀ (black) and S₁ (red) states of Cy5 rotation along with the dihedral angle φ₁ calculated at the TD-DFT/HFB/TZVP level with COSMO solvation model.

Fig. 6 Energy levels of the S₀ (black) and S₁ (red) states of Cy5-CH₃ rotation along the dihedral angle φ₁ (a) and φ₄ (b) calculated at the TD-DFT/HFB/TZVP level with COSMO solvation model.

Table 6 Energy barrier E_a of S₁ and energy gap E_gap between the S₀ and S₁ states for Cy5, Cy5-CH₃ and Cy5-CHO studied in this work for rotation around different dihedral angles (φ₁ and φ₄). Energy gaps between the S₀ and S₁ states are at 90° for the three dyes

Dihedral angle	Cy5		Cy5-CH3		Cy5-CHO
	Ea (kcal mol^−1)	Egap (kcal mol^−1)	Ea (kcal mol^−1)	Egap (kcal mol^−1)	Ea (kcal mol^−1)	Egap (kcal mol^−1)
φ1	11.78	36.31	8.51	38.35	6.28	40.30
φ4	—	—	0.12	48.26	5.53	45.55

---

These findings imply that low environment-sensitivity for Cy5-CH₃ is related to its longer chain and meso-substituent. Fig. 7 shows the calculated potential–energy curves for φ₁ and φ₄ twisting dihedral angles in the ground state and the first electronic excited state of Cy5-CHO. Introduction of an aldehyde group substantially decreases the energy barrier to rotation around φ₁. In addition, the energy barrier to rotation around φ₄ is quite low and the energy gap between the ground state and the first excited state is also small. The energy barrier is 6.28 kcal mol⁻¹ for φ₁ and 5.53 kcal mol⁻¹ for φ₄ (in Table 6). It should also be said that the aldehyde group can also reduce the energy barrier to rotation around φ₁ more and provide a high environmental sensitivity. Specifically, the aldehyde group is also an excellent rotor and gives rise to internal conversion by a nonradiative process. It has been implicit in our discussion of the C–C bond rotation of the two dyes that substitution on the bridge is expected to decrease the energy barriers for rotation around different chemical bonds, which play a major role in reducing fluorescence quantum yield in the free state and providing a high environmental sensitivity for the meso-substituted cyanine dyes. The substitution regioselectivity plays a major role in determining background fluorescence quantum yield and environmental sensitivity.

Fig. 7 Energy levels of the S₀ (black) and S₁ (red) states of Cy5-CHO rotation along with dihedral angle φ₁ (a) and φ₄ (b) calculated at the TD-DFT/HFB/TZVP level with COSMO solvation model.

3.3. Design and predict meso-substituted Cy3 dyes based on the mechanism

A detailed theoretical understanding of the substituent effects of the polymethine chain in the cyanine dyes on the environmental sensitivity is greatly desired to design and predict new dyes for the improvement of environmental sensitivity. In this section, we further use this method to predict environmental sensitivity of meso-substituted Cy3 models. We compare potential energy curves for Cy3, Cy3-CH₃ and Cy3-Ph. We have proved that the nonradiative deactivation of Cy3 is dominated by the rotation of φ₁ in the first excited state (Fig. 8). The results for Cy3-CH₃ and Cy3-Ph in Fig. 9 and Fig. 10 show that the energy barrier of S₁ for the rotation of φ₁ is smaller than Cy3. Table 7 lists the energy barrier of S₁ and energy gap between the S₀ and S₁ states for Cy3, Cy3-CH₃ and Cy3-Ph. The energy barrier is 8.54 kcal mol⁻¹ for Cy3. For Cy3-CH₃, the energy barrier to rotation around the φ₁ is very low (only 1.07 kcal mol⁻¹). It has been shown that the introduction of methyl has a strong influence on the energy barrier around rotation around the φ₁, and predictably, the low background fluorescence quantum yield for Cy3-CH₃. For Cy3-Ph, the energy barriers to rotation around the φ₁ and φ₃ are quite low and the energy gaps between the ground state and the first excited state are also small. Similarly, Cy3-Ph should have low background fluorescence quantum yield. This conformation is so crowded that it probably cannot assume a completely planar geometry. The benzene ring must necessarily be highly distorted from planarity. We calculated the absorption and emission maxima of the two dyes using the TD-DFT/HFB/TZVP method with the COSMO solvation model. The oscillator strengths in the lowest singlet excitation state of meso-substituted Cy3 models are very low, which indicates that the first excitation state becomes a dark state that decays nonradiatively to the ground state. However, the nonradiative deactivation results of Cy3-CH₃ and Cy3-Ph were only drawn from a calculation, and no experimental evidence has been reported yet.

Fig. 8 Energy levels of the S₀ (black) and S₁ (red) states of Cy3 rotation along with dihedral angle φ₁ calculated at the TD-DFT/HFB/TZVP level with COSMO solvation model.

Fig. 9 Energy levels of the S₀ (black) and S₁ (red) states of Cy3-CH₃ rotation along with dihedral angles φ₁ (a) and φ₃ (b) calculated at the TD-DFT/HFB/TZVP level with COSMO solvation model.

Fig. 10 Energy levels of the S₀ (black) and S₁ (red) states of Cy3-Ph rotation along with dihedral angles φ₁ (a) and φ₃ (b) calculated at the TD-DFT/HFB/TZVP level with the COSMO solvation model.

Table 7 Energy barrier of S₁ and energy gap between the S₀ and S₁ states for Cy3, Cy3-CH₃ and Cy3-Ph studied in this work for rotation around different dihedral angles (φ₁ and φ₃). Energy gaps between the S₀ and S₁ states are at 90° for Cy3 and Cy3-CH₃ and the energy gap between the S₀ and S₁ state is at 180° for the φ₃ of Cy3-Ph

Dihedral angles	Cy3		Cy3-CH3		Cy3-Ph
	Ea (kcal mol^−1)	Egap (kcal mol^−1)	Ea (kcal mol^−1)	Egap (kcal mol^−1)	Ea (kcal mol^−1)	Egap (kcal mol^−1)
φ1	8.54	43.22	1.07	37.11	1.42	34.31
φ3			0.17	49.61	4.44	43.86

---

Cy3-CH₃ and Cy3-Ph were synthesized according to the previous literature. As is shown in Table 8, spectral data allow for a comparison with spectroscopic characteristics determined by DFT. As expected, we measured the fluorescence quantum yields of Cy3-CH₃ and Cy3-Ph in water and glycerol. The introduction of a substituent in the polymethine chain of Cy3 leads to a significant low background fluorescence quantum yield. Cy3-CH₃ is specified by a low fluorescence quantum yield in water (Φ_F = 0.0015). The low fluorescence quantum yield of the Cy3-CH₃ in water closely resembles that predicted by DFT. Indeed, our calculation indicated the dark state in the first excited state and the very low rotation energy barrier. Φ_F ranges from 0.0015 in water to 0.0134 in 99% glycerol for Cy3-CH₃. As glycerol was gradually added to water, the viscosity of the solutions increased from 1.0 cp (water) to ca. 950 cp (99% glycerol), and paralleling fluorescence enhancements of the Cy3-CH₃ increased continuously (Fig. 11). Cy3 displays some sensitivity to the viscosity of glycerol. However, there is rather low fluorescence quantum yield in 99% glycerol. Due to the low energy barrier for Cy3-CH₃, 99% glycerol is not possible to restrain the rapid rotation of polymethine chains. The above discovery revealed that Cy3-CH₃ with very weak fluorescence in a free state is not suitable for quantifying viscosity for biomedical imaging applications because of low brightness. The more constrained environment will enhance the fluorescence quantum yield for Cy3-CH₃. The result of Cy3-Ph is similar to Cy3-CH₃. However, substituents strongly affect the energy barrier and fluorescence quantum yield for Cy3. No or low energy barriers for meso-substituted Cy3 do not favor viscosity measurements in microenvironments. We expect that this consideration presented in this study will help the exploration and development of a special group of fluorescent viscosity-sensitive probes in biological studies.

Table 8 The maximum absorption and emission wavelengths of free Cy3, Cy3-CH₃ and Cy3-Ph (The dark states of the first excitation states predicted by TDDFT for Cy3-CH₃ and Cy3-Ph)

Dye	Exptl λ (nm)		Calcd λ (nm)
	Abs	Flu	Abs	Flu
Cy3	541	560	534	587
Cy3-CH3	535	558	538	Dark state
Cy3-Ph	580	635	579	Dark state

---

Fig. 11 The emission spectra changes of Cy3-CH₃ (excitation at 530 nm) with the increase of the viscosity of water–glycerol systems.

4. Conclusions

In this article, we have clarified the mechanism of the different sensitivities of meso-substituted and unsubstituted cyanine dyes in rotation-restricted environments for biomedical imaging applications. Cyan 2 has a relatively low level of intrinsic fluorescence while the fluorescence of Cyan 46 in an unbound state is rather significant. The different environmental sensitivity of Cyan 46 and Cyan 2 is due to methyl on the bridge which decreases the energy barriers for rotation around different chemical bonds and reduces the fluorescence quantum yield in a free state. Cy5, Cy5-CH₃ and Cy5-CHO have different solution viscosity sensitivity. Introduction of a methyl group and an aldehyde group decreased the energy barrier of rotation around φ₁, which facilitated efficient nonradiative internal conversion and decreased background quantum yield. The aldehyde group is also an excellent rotor and gives rise to internal conversion by a nonradiative process. The introduction of methyl in the shorter polymethine chain of Cy3 has a strong influence on the energy barrier of rotation around φ₁, which leads to a significant low background fluorescence quantum yield. We believe that the proposed mechanism for different environmental sensitivity of meso-substituted and unsubstituted cyanine dyes lays the foundation for further investigation into new cyanine dyes using highly sensitive environmental detection and biomedical imaging applications. Ongoing work is directed toward investigating how other substituents on the polymethine chain influence the barrier height and environmental sensitivity.

Acknowledgements

This work was supported by the National Science Foundation of China (21136002, 21076032, 20923006), the National Basic Research Program of China (2009CB724706 and 2012CB733702) and National High Technology Research and Development Program of China (863 program 2011AA02A105).

Notes and references

1. B. A. Armitage, Top. Heterocycl. Chem., 2008, 14, 11–29.
2. A. Mishra, R. K. Behera, P. K. Behera, B. K. Mishra and G. B. Behera, Chem. Rev., 2000, 100, 1973–2012.
3. T. G. Deligeorgiev, S. Kaloyanova and J. J. Vaquero, Recent Pat. Mater. Sci., 2009, 2, 1–26.
4. S. Murudkar, A. K. Mora, P. K. Singh and S. Nath, Chem. Commun., 2012, 48, 5301–5303.
5. J. W. Park, Y. Kim, K.-J. Lee and D. J. Kim, Bioconjugate Chem., 2012, 23, 350–362.
6. N. I. Shank, K. J. Zanotti, F. Lanni, P. B. Berget and B. A. Armitage, J. Am. Chem. Soc., 2009, 131, 12960–12969.
7. C. Zhang, S. Wang, J. Xiao, X. Tan, Y. Zhu, Y. Su, T. Cheng and C. Shi, Biomaterials, 2010, 31, 1911–1917.
8. Y. Xu, A. Malkovskiy, Q. Wang and Y. Pang, Org. Biomol. Chem., 2011, 9, 2878–2884.
9. X.-D. Liu, R. Sun, J.-F. Ge, Y.-J. Xu, Y. Xu and J.-M. Lu, Org. Biomol. Chem., 2013, 11, 4258–4264.
10. D. Zhu, G. Li, L. Xue and H. Jiang, Org. Biomol. Chem., 2013, 11, 4577–4580.
11. R. B. Mujumdar, L. A. Ernst, S. R. Mujumdar, C. J. Lewis and A. S. Waggoner, Bioconjugate Chem., 1993, 4, 105–111.
12. H. Brismar and B. Uifhake, Nat. Biotechnol., 1997, 15, 373–377.
13. J. O. Smith, D. A. Olson and B. A. Armitage, J. Am. Chem. Soc., 1999, 121, 2686–2695.
14. X. Peng, T. Wu, J. Fan, J. Wang, S. Zhang, F. Song and S. Sun, Angew. Chem., Int. Ed., 2011, 50, 4180–4183.
15. S. Luo, E. Zhang, Y. Su, T. Cheng and C. Shi, Biomaterials, 2011, 32, 7127–7138.
16. K. Luby-Phelps, S. Mujumdar, R. B. Mujumdar, L. A. Ernst, W. Galbraith and A. S. Waggoner, Biophys. J., 1993, 65, 236–242.
17. B. Soroushian and X. M. Yang, Biomed. Opt. Express, 2011, 2, 2749–2760.
18. V. Sundström and T. Gillbro, Chem. Phys., 1981, 61, 257–269.
19. H. S. Muddana, T. T. Morgan, J. H. Adair and P. J. Butler, Nano Lett., 2009, 9, 1559–1566.
20. C.-L. Peng, Y.-H. Shih, P.-C. Lee, T. M.-H. Hsieh, T.-Y. Luo and M.-J. Shieh, ACS Nano, 2011, 5, 5594–5607.
21. R. Allmann, H.-J. Anis, R. Benn, W. Grahn, S. Olejnek and A. Waśkowska, Angew. Chem., Int. Ed. Engl., 1983, 22, 876–877.
22. A. Gretchikhine, G. Schweitzer, M. Van der Auweraer, R. De Keyzer, D. Vandenbroucke and F. C. De Schryver, J. Appl. Phys., 1999, 85, 1283–1293.
23. W. West, S. Pearce and F. Grum, J. Phys. Chem., 1967, 71, 1316–1326.
24. V. Khimenko, A. K. Chibisov and H. Görner, J. Phys. Chem. A, 1997, 101, 7304–7310.
25. A. K. Chibisov, G. V. Zakharova and H. Gorner, J. Chem. Soc., Faraday Trans., 1996, 92, 4917–4925.
26. A. M. K. F. A. Mikhailenko, Russ. Chem. Rev., 1987, 56, 466–488.
27. V. B. Kovalska, K. D. Volkova, M. Y. Losytskyy, O. I. Tolmachev, A. O. Balanda and S. M. Yarmoluk, Spectrochim. Acta, Part A, 2006, 65, 271–277.
28. S. M. Yarmoluk, V. B. Kovalska, S. S. Lukashov and Y. L. Slominskii, Bioorg. Med. Chem. Lett., 1999, 9, 1677–1678.
29. X. Peng, Z. Yang, J. Wang, J. Fan, Y. He, F. Song, B. Wang, S. Sun, J. Qu, J. Qi and M. Yan, J. Am. Chem. Soc., 2011, 133, 6626–6635.
30. K. J. Zanotti, G. L. Silva, Y. Creeger, K. L. Robertson, A. S. Waggoner, P. B. Berget and B. A. Armitage, Org. Biomol. Chem., 2011, 9, 1012–1020.
31. F. Yu, P. Li, G. Li, G. Zhao, T. Chu and K. Han, J. Am. Chem. Soc., 2011, 133, 11030–11033.
32. H.-J. Yoon, M. Dakanali, D. Lichlyter, W. M. Chang, K. A. Nguyen, M. E. Nipper, M. A. Haidekker and E. A. Theodorakis, Org. Biomol. Chem., 2011, 9, 3530–3540.
33. V. Ediz, J. L. Lee, B. A. Armitage and D. Yaron, J. Phys. Chem. A, 2008, 112, 9692–9701.
34. J. Cao, C. Hu, F. Liu, W. Sun, J. Fan, F. Song, S. Sun and X. Peng, ChemPhysChem, 2013, 14, 1601–1608.
35. F. Zhou, J. Shao, Y. Yang, J. Zhao, H. Guo, X. Li, S. Ji and Z. Zhang, Eur. J. Inorg. Chem., 2011, 4773–4787.
36. F. Liu, T. Wu, J. Cao, S. Cui, Z. Yang, X. Qiang, S. Sun, F. Song, J. Fan, J. Wang and X. Peng, Chem.–Eur. J., 2013, 19, 1548–1553.
37. M. Koenig, G. Bottari, G. Brancato, V. Barone, D. M. Guldi and T. Torres, Chem. Sci., 2013, 4, 2502–2511.
38. J. Cao, T. Wu, C. Hu, T. Liu, W. Sun, J. Fan and X. Peng, Phys. Chem. Chem. Phys., 2012, 14, 13702–13708.
39. G. Chen, F. Song, X. Wang, S. Sun, J. Fan and X. Peng, Dyes Pigm., 2012, 93, 1532–1537.
40. D. J. Owen, D. VanDerveer and G. B. Schuster, J. Am. Chem. Soc., 1998, 120, 1705–1717.
41. A. D. Becke, Phys. Rev. A: At., Mol., Opt. Phys., 1988, 38, 3098–3100.
42. O. Treutler and R. Ahlrichs, J. Chem. Phys., 1995, 102, 346–354.
43. G.-Y. Li, G.-J. Zhao, Y.-H. Liu, K.-L. Han and G.-Z. He, J. Comput. Chem., 2010, 31, 2109–2125.
44. G.-J. Zhao and K.-L. Han, J. Phys. Chem. A, 2009, 113, 14329–14335.
45. G.-J. Zhao and K.-L. Han, Phys. Chem. Chem. Phys., 2010, 12, 8914–8918.
46. A. Klamt, J. Phys. Chem., 1996, 100, 3349–3353.
47. G. L. Silva, V. Ediz, D. Yaron and B. A. Armitage, J. Am. Chem. Soc., 2007, 129, 5710–5718.
48. K. A. Willets, P. R. Callis and W. E. Moerner, J. Phys. Chem. B, 2004, 108, 10465–10473.
49. R. Improta and F. Santoro, J. Chem. Theory Comput., 2005, 1, 215–229.
50. M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria, M. A. Robb, J. R. Cheeseman, G. Scalmani, V. Barone, B. Mennucci, G. A. Petersson, H. Nakatsuji, M. Caricato, X. Li, H. P. Hratchian, A. F. Izmaylov, J. Bloino, G. Zheng, J. L. Sonnenberg, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, T. Vreven, J. A. Montgomery Jr, J. E. Peralta, F. Ogliaro, M. Bearpark, J. J. Heyd, E. Brothers, K. N. Kudin, V. N. Staroverov, R. Kobayashi, J. Nor-mand, K. Raghavachari, A. Rendell, J. C. Burant, S. S. Iyengar, J. Tomasi, M. Cossi, N. Rega, J. M. Millam, M. Klene, J. E. Knox, J. B. Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R. E. Stratmann, O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. W. Ochterski, R. L. Martin, K. Morokuma, V. G. Zakrzewski, G. A. Voth, P. Salvador, J. J. Dannenberg, S. Dapprich, A. D. Daniels, O. Farkas, J. B. Foresman, J. V. Ortiz, J. Cioslowski, and D. J. Fox, Gaussian, Inc., Wallingford CT, 2009, The Gaussian 09 package refer to Gaussian 09, Revision A.02.
51. R. Englman and J. Jortner, Mol. Phys., 1970, 18, 145–164.
52. M. Dekhtyar, W. Rettig and M. Sczepan, Phys. Chem. Chem. Phys., 2000, 2, 1129–1136.
53. A. Muñoz-Losa, M. E. Martín, I. F. Galván, M. L. Sánchez and M. A. Aguilar, J. Chem. Theory Comput., 2011, 7, 4050.
54. F. Dietz and S. K. Rentsch, Chem. Phys., 1985, 96, 145–151.
55. A. Sanchez-Galvez, P. Hunt, M. A. Robb, M. Olivucci, T. Vreven and H. B. Schlegel, J. Am. Chem. Soc., 2000, 122, 2911–2924.
56. X. Cao, R. W. Tolbert, J. L. McHale and W. D. Edwards, J. Phys. Chem. A, 1998, 102, 2739–2748.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c3ra46612d

---