Introduction

Dimethyl sulfide (DMS, CH3SCH3) is the most abundant biogenic volatile organic sulfur compound (VOSC) that is produced through enzymatic lysis of dimethylsulfoniopropionate (DMSP) in the oceans1,2,3. On a global scale, marine DMS plays a key role in the organosulfur cycle with an estimated annual flux of about 30 teragrams of sulfur in the atmosphere4,5. The removal of DMS under marine atmospheric boundary layer (MABL) conditions involves biological consumption, sea-atmosphere exchange, and oxidation reactions. The atmospheric oxidation of DMS to condensable products contributes to the formation of secondary sulfate aerosols that affect Earth’s climate by scattering solar irradiation and simultaneously acting as cloud condensation nuclei (CCN)6,7. Therefore, the details about the oxidation mechanism of DMS are of vital importance in understanding the interplay between atmospheric chemistry and climate change8.

According to comprehensive smog chamber experiments and theoretical modeling9,10,11,12, the oxidation of DMS in the atmosphere is rather complex, which mainly invokes the formation of methylthio radical (CH3S•) through H-abstraction and subsequent radical-initiated decomposition upon reactions with •OH, •Cl, or •NO3 via the intermediacy of elusive radicals CH3SCH2•, CH3SCH2OO•, and CH3SCH2O• (Fig. 1). Then, the oxidation proceeds by further reactions of CH3S• with atmospherically relevant oxidants (e.g., O2, O3, and •NO2) to yield a number of transient sulfur-containing radicals including sulfinyl radical CH3SO•, sulfonyl radical CH3SO2•, and sulfonyloxyl radical CH3SO3•. Eventually, CH3SO3• can either dissociate (→ •CH3 + SO3) or undergo hydrogen abstraction to furnish sulfuric acid (SO3 + H2O → H2SO4) and methane sulfonic acid (CH3SO3H, MSA)9, respectively. Both acids are key precursors to new particles formed via homogeneous nucleation and subsequent cluster growth in air masses13,14. Recently, an alternative abstraction pathway for the •OH initiated oxidation of DMS to SO2 through the intramolecular H-shift in the common peroxyl radical intermediate CH3SCH2OO• (→ •CH2SCH2OOH) has been proposed (Fig. 1), in which the formation of the key stable intermediate hydroperoxymethyl thioformate (HPMTF, HOOCH2SCHO) has been confirmed experimentally8,15,16,17,18,19. On the other hand, the atmospheric oxidation of DMS can also proceed through the OH-addition pathway (Fig. 1), resulting the stepwise formation of additional VOSCs dimethyl sulfoxide (DMSO, CH3S(O)CH3), and methanesulfinic acid (CH3S(O)OH)15,16,17,18,19,20,21,22.

Fig. 1: Reaction pathways for the atmospheric oxidation of dimethyl sulfide (DMS).
figure 1

The reaction pathways for the formation and subsequent O2-oxidation of methylthio radical (CH3S•) during the atmospheric oxidation of dimethyl sulfide (DMS). Radicals studied in this work are shown in red.

Among the O2-oxidation reactions of CH3S•, formation of three peroxyl radicals CH3SOO•, CH3S(O)OO•, and CH3S(O)2OO• has been also postulated9,10,11,12. Indeed, the sulfinylperoxy radical CH3S(O)OO• and its photodecomposition to •CH3 and SO3 via the intermediacy of CH3SO3• have been observed during the O2-oxidation of CH3SO•23,24,25. As the initial O2-oxidation product of CH3S•, CH3SOO• is less stable than CH3S(O)OO• due to a small S–OO bond dissociation energy (BDE) of ca. 10 kcal mol−1 26,27, and it has been only tentatively identified among the photolytic O2-oxidation products of CH3SSCH3 based on the observation of two transient absorptions using step-scan IR spectroscopy28. As the formal O2-oxidation product of the sulfonyl radical CH3SO229,30,31,32, CH3S(O)2OO• remains yet unobserved, although reactions between CH2SO2• and O2 via the intermediacy of CH3S(O)2OO• have been proposed during the one-electron reduction of CH3S(O)2Cl in oxygenated solutions33 and also in the pulse radiolysis of an N2O–O2 saturated solution of NaOS(O)CH334. According to the recent theoretical computations, CH3S(O)2OO• has higher stability than CH3SOO• and CH3S(O)OO• due to the highest BDE for the shortest S–OO bond35.

To unveil the mechanism for the oxidation of CH3S•, a practical method for efficient generation of this thiyl radical is desirable. Typically, thiyl radicals (RS•) can be generated through homolytic cleavage of the S–S or S–H bonds in disulfides (RS–SR) or thiols (RS–H) under photolysis or pyrolysis conditions36. However, the associated large BDEs (>60 kcal mol−1) in these compounds render the efficiency of thermal fragmentation relatively low in the absence of catalyst37. Hence, UV-laser photolysis of CH3SSCH3 and CH3SH has been frequently used in generating CH3S• in the gas phase38,39,40. Recently, the formation of CH3S• from the decomposition of CH3SSCH3 was observed on metal surfaces by using visible light irradiation41,42, in which the photo-induced plasmon serve as the catalyst.

Herein, we report an optimized method for facile generation of CH3S• by high-vacuum flash pyrolysis (HVFP) of S-nitrosothiol CH3SNO in the gas phase, which enables us to study the mechanism for the O2-oxidation reactions of CH3S• and the first-time unambiguous identification of the two important intermediates CH3SOO• and CH3S(O)2OO• that are critical to the validation of the DMS oxidation process occurring in the atmosphere (Fig. 1).

Results and discussion

Generation of CH3S•

S-nitrosothiols are endogenous sources of nitric oxide (•NO) in biological systems43 due to easy breakage of the S−N bonds with BDEs less than 30 kcal mol−1 44,45. Particularly, the S−N bond energy in CH3SNO (1) is about 20 kcal mol−1 46, implying facile fragmentation under pyrolysis conditions. A typical IR spectrum for the pyrolysis (400 °C) products of CH3SNO (Fig. 2a) isolated in N2-matrix at 10 K shows the formation of CH3S• (2, 1398.3, 1053.4, and 783.0 cm−1, Fig. 2b)39, •CH3 (3, 611.1 cm−1)47, and •NO (4, 1874.9 cm−1)46. Owing to the moderate BDEs for C–S (70 kcal mol−1) and C–H (49 kcal mol−1) bonds in 238, further increase of the pyrolysis temperature to ca. 650 °C leads to complete dissociation of 2 to 3 and H2CS (5, 1438.7, 1062.7, and 994.9 cm−1)48 through the elimination of sulfur and hydrogen atoms, respectively. When using 13C-labeled CH3SNO as the precursor, the isotopically labeled 13CH3S• can be generated, and noticeable 12/13C-isotopic shifts of 2.7, 6.2, and 4.8 cm−1 for the aforementioned three IR fundamental modes of 2 have been determined for the first time.

Fig. 2: IR spectra showing the generation and photochemistry of CH3S• (2).
figure 2

a IR spectrum of CH3SNO in N2-matrix at 10 K. b IR spectrum of the high-vacuum flash pyrolysis (HVFP) products of CH3SNO in N2-matrix. c IR difference spectrum reflecting the change of the HVFP products upon UV-light irradiation (365 nm, 15 min). The IR bands for CH3SNO (1), CH3S• (2), •CH3 (3), •NO (4), H2CS (5), SO2 (6), N2O2 (7), CO (8), H2O (9), and N2S (10) are labeled.

Upon irradiation with UV-light emitting diode (LED, 365 nm), the corresponding IR difference spectrum (Fig. 2c) reflecting the change of the matrix-isolated pyrolysis products of 1 shows primary depletion of 2 with concomitant formation of N2S (10, 2047.8 and 737.5 cm−1)49 and •CH3 (3, 628.1 cm−1). Therefore, the thiyl radical 2 acts as an effective sulfur atom transfer (SAT) reagent50 by reacting with the matrix material (N2) under the photolytic excitation at 365 nm. The identification of 10 is confirmed by the observation of large 14/15N-isotopic shifts of 67.8 and 13.1 cm−1 for the two bands in the experiment using 15N2 as the matrix material. The noticeable shift (Δν = 17.0 cm−1) of the IR band at 628.1 cm−1 for the newly formed 3 during the photolysis implies strong interactions with neighboring N2S in the same N2-matrix. The generation of CH3S• by pyrolysis of CH3SNO is reproducible when using Ar or Ne as carrying gas (Supplementary Fig. 1), and its photodecomposition to H2CS (5) was observed under similar UV-irradiation conditions (365 nm). It should be noted that the photoactivity of 2 coincides with the observed absorption at 375 nm for the radical in Ar-matrix (vide infra).

O2-oxidation of CH3S•

When the pyrolysis of CH3SNO (1) was performed in presence of oxygen (1/O2/Ar, 1:50:1000), the IR spectrum of the products at 10 K (Fig. 3a) shows complete disappearance of •CH3 and CH3S• by forming CH3OO• (13, 1447.8 and 1180.4 cm−1)51 and a new species (11) with strong IR bands at 1392.2 and 1102.2 cm−1 (Fig. 3a). These frequencies are close to the two transient absorptions at 1397 ± 1 and 1110 ± 3 cm−1 that were tentatively assigned to CH3SOO• in the previous gas-phase study on the photolytic (248 nm) O2-oxidation of CH3SSCH328. In order to distinguish the IR bands for the most likely candidate (11), the matrix was subjected to a red-light LED irradiation (830 nm). The resulting IR difference spectrum shows exclusive depletion of 11 (Fig. 3b), and sulfonyl radical CH3SO2• (1413.9, 1274.2, 1074.5, 915.6, 631.3, and 460.2 cm−1, 14)29 forms, indicating isomerization of 11 under the irradiation conditions. Similar photoisomerization has been found for PhSOO• (→ PhSO2•)52 and OSOO (→ SO3)53. The previously28 proposed secondary oxidation of 11 with O2 to form CH3SO• in the gas phase was not observed under the matrix-isolation conditions.

Fig. 3: IR spectra showing the formation and photochemistry of CH3SOO• (11) and CH3S(O)2OO• (15).
figure 3

a IR spectrum of the HVFP products of CH3SNO/O2/Ar (1:50:1000) at 10 K. b IR difference spectrum reflecting the change of the HVFP products upon red-light irradiation (830 nm, 20 min). c IR difference spectrum reflecting the change of the HVFP products upon subsequent blue-light irradiation (440 nm, 40 min). d IR difference spectrum reflecting the change of the HVFP products upon further UV-laser irradiation (266 nm, 10 min). The IR bands for CH3SNO (cis: 1; trans: 1′), •NO (4), H2CS (5), SO2 (6), N2O2 (7), CO (8), CH3SOO• (11), NO2• (12), CH3OO• (13), CH3SO2• (14), CH3S(O)2OO• (15), CH3S•···•ON (16), SO3 (17), OCS (18), CH2O (19), and CH3SSCH3 (*) are labeled.

The selective conversion (11 → 14) allows unambiguous identification of all the remaining weaker IR fundamental bands for 11 (Table 1). The assignment is supported by the agreement with the CCSD(T)/aug-cc-pV(T + d)Z computed IR spectrum for 11 (Table 1) in a favorable syn-conformation between the CH3 moiety and the terminal oxygen atom with respect to the S–O bond. According to the 18O-isotope labeling experiment (Fig. 4), the strong band in CH3SOO• at 1102.2 cm−1 (calc. 1135.9 cm−1) is reasonably assigned to the O–O stretching mode (ν(OO)) due to a large 16/18O isotopic shift of 61.3 cm−1 (calc. 71.5 cm−1). In contrast, no shift occurs to this band in the 13C-isotope labeling experiment (Fig. 4). The frequency is close to the ν(OO) mode in other peroxyl radicals such as PhSOO• (1173 cm−1)52 and CH3S(O)OO• (syn/anti: 1100.3/1081.3 cm−1)23 with comparable 16/18O-isotopic shifts of 64 and 61.0/58.3 cm−1, respectively. The bands at 547.6 and 443.7 cm−1 correspond to the S–O stretching (calc. 571.1 cm−1) and SOO bending modes (calc. 406.0 cm−1) with large 16/18O-isotopic shifts of 27.0 (calc. 29.0 cm−1) and 15.3 cm−1 (calc. 10.7 cm−1) but small 12/13C-isotopic shifts of 0.8 (calc. 1.0 cm−1) and 0.5 cm−1 (calc. 0.3 cm−1), respectively. The C–S stretching mode locates at 722.5 cm−1 as it displays a large 12/13C-isotopic shift of 15.2 cm−1 (calc. 14.4 cm−1, Table 1), and it is close to the same mode in CH3SO3• (757.6 cm−1, Ar-matrix) and CH3S(O)OO• (syn/anti: 686.4/676.7 cm−1)23.

Table 1 Observed and calculated IR data (>400 cm−1) for CH3SOO•.
Fig. 4: IR difference spectra for the isotope-labeling experiments.
figure 4

a Expaned IR difference spectra (1600–400 cm−1) for the photo-induced (830 nm) conversion of CH3SOO• (11) to CH3SO2• (14) and CH3S(O)2OO• (15). The spectra for the experiments with the 18O- and 13C-labeling samples are shown in b and c, respectively.

In addition to CH3SO2• (14), another species (15) with distinct IR bands at 1435.3, 1402.8, 1320.9, 1204.2, 1078.2, and 767.7 cm−1 also forms after the red-light irradiation of CH3SOO• (11) in the O2-doped Ar-matrix (Fig. 3b). Further irradiation of the same matrix with blue-light LED (440 nm) leads to specific conversion of 14 to 15 (Fig. 3c) together with the previously observed photodissociation of the two conformers of CH3SNO (cis: 1; trans: 1’) to the metastable caged radical pair CH3S•···•ON (16, 1820.9 cm−1) in the cryogenic matrix46. In the 18O-labeling experiment (Fig. 4), the two bands at 1435.3 and 1204.2 cm−1 shift to 1382.1 and 1156.5 cm−1, corresponding to isotopic shifts of 53.2 and 47.7 cm−1, respectively, whereas, only very small shifts occur to the two bands in the 13C-isotope labeling experiment. The frequencies are close to the two SO2 stretching modes (νas(SO2) and νs(SO2)) in methane sulfonic acid CH3SO3H (1403 and 1202 cm−1)24 and sulfonyl nitrene FS(O)2N (1426.4 and 1206.5 cm−1)54. In contrast, they are significantly higher than the νas(SO2) and νs(SO2) modes in 14 at 1274.2 and 1074.5 cm−1, for which the 18O-isotopic shifts are 38.9 and 44.1 cm−1, respectively. Additionally, the weak band at 1078.2 cm−1 displays a large 18O-isotopic shift of 59.6 cm−1. The frequency and associated isotopic shift are similar with the ν(OO) mode in CH3SOO• (1102.2 cm−1, Δν(16/18O) = 61.3 cm−1), strongly suggesting the assignment of this new species to the peroxyl radical CH3S(O)2OO• (15), which is formed from CH3SO2• (14) by further combination of molecular oxygen in the O2-doped Ar-matrix. The efficient isomerization of 11 to 14 is consistent with the theoretically predicted higher stability of the latter by a free energy difference (ΔG) of –65 kcal mol−1 (M06-2X/6-311++G(3df,3pd). In contrast, the subsequent conversion of 14 to 15 is a near thermodynamically neutral process with a calculated free energy change of –1.5 kcal mol−1.

The assignment of the IR bands for 15 is also supported by the agreement with the CCSD(T)/aug-cc-pV(D + d)Z computations (Table 2). For instance, the computed frequency for the ν(OO) mode in 15 is 1061.5 cm−1 (obs. 1078.2 cm−1), and the ν(SO) frequency at 621.2 cm−1ν(16/18O) = 25.1 cm−1) matches the observation at 631.3 cm−1ν(16/18O) = 24.0 cm−1). The experimentally observed 16/18O-isotopic shift of 53.2 cm−1 for the νas(SO2) mode at 1435.3 cm−1 (cal. 1436.0 cm−1) agrees with the predicted 16/18O-shifts of 58.3 cm−1, and it mixes with the δ(CH3) mode at 1402.8 cm−1 (cal. 1367.9 cm−1) as evidenced by the 16/18O-isotopic shift of −7.5 cm−1 (cal. −14.9 cm−1). The distinguishment of the νas(SO2) mode from the three δ(CH3) modes in the range of 1500–1350 cm−1 can be also acertained with their distinct 12/13C-isotopic shifts (Table 2). The generation of 11 and its photoisomerization to 14 with further oxidation to 15 is reproducible in N2- and Ne-matrixes (Supplementary Figs. 2 and 3).

Table 2 Observed and calculated IR data (>400 cm−1) for CH3S(O)2OO•.

In line with the computed lowest-energy vertical transition at about 220 nm for CH3S(O)2OO• (Supplementary Table 2), no change occurs to this peroxyl radical under visible-light irradiations. In contrast, it can be partly depleted by a 266 nm laser with unspecific decomposition (Fig. 3d) to SO2 (6), CO (8), CH3SO2• (14), SO3 (17), OCS (18), and CH2O (19). The formation of OCS in the oxidation of CH3S• is consistent with the previous discovery of the tropospheric oxidation of DMS as a potent source of OCS55,56, which serves as a key tracer for the global carbon cycle. The formation of SO3 indicates the possible involvement of CH3SO3• (CH3S(O)2OO• + O2 → CH3SO3• + O3), which can decompose (CH3SO3• → •CH3 + SO3) upon UV-light irradiation23,24. As a further step, photofragmentation of CH3OO• (13) to CO (8), CH2O (19), and CO2 occurs upon the laser irradiation.

Electronic spectra

The stepwise O2-oxidation reactions of CH3S• (2) via the intermediacy of peroxyl radicals were also followed with matrix-isolation UV–vis spectroscopy. Thanks to the efficient production of 2 in the gas phase, a full UV–vis absorption spectrum for this simplest organosulfur radical isolated in an Ar-matrix at 10 K has been obtained (Fig. 5). Note that only weak absorptions in the range of 220–200 nm were assigned to 2 in previous gas-phase studies57,58,59. In sharp contrast, 2 isolated in Ar-matrix displays two absorptions that completely differ from its precursor (1, 340 and 215 nm). The weak absorption band (λmax) of 2 at 375 nm exhibits pronounced vibrational fine structures with onset at ca. 450 nm. This assignment coincides with the previous MRCI computed energy of 373 nm for the Ã(2A1) ←  2E transition60, and it also reasonably explains the aforementioned photochemistry of 2 by the irradiation at 365 nm (Fig. 2c). The second stronger band of 2 at 270 nm contains superimposed vibrational fine structures for the byproduct S2 that is generated via fragmentation of 2 (→ •CH3 + S) followed by immediate aggregation during the same deposition process. The assignment of the strongest band at 205 nm is unclear since other accompanied decomposition products (S2, •NO, and •CH3) of undecomposed CH3SNO in the same matrix also contain absorptions at around 200 nm.

Fig. 5: UV–vis spectra of CH3S• and its O2-oxidation products in Ar-matrix.
figure 5

UV–vis spectra of Ar-matrix isolated CH3SNO, HVFP products of CH3SNO, HVFP products of CH3SNO/O2, and the photolysis products of the HVFP products of CH3SNO/O2 at 10 K. Inset: the expanded spectra in the range of 800–320 nm.

In the UV–vis spectrum of the matrix-isolated HVFP products of CH3SNO/O2, the absorption bands for CH3S• (375 and 270 nm) disappear while two new overlapping bands occur at 295 and 255 nm with onset near 400 nm. Subsequent irradiation with 830 nm light causes partial depletion of the broad absorption, implying the contribution of the absorption from the highly photolabile peroxyl radical CH3SOO• (11). In the same time, absorptions for the products CH3SO2• (14) in the range of 330–400 nm27,30 and CH3S(O)2OO• (15) with predicted intense absorption at 221 nm should appear by referring to the corresponding IR spectrum (Fig. 3b). The byproduct CH3OO• (13)61 formed in the pyrolysis of CH3SNO/O2 also contributes to the broad band, and it remains unchanged during the successive visible light irradiations (830 and 440 nm, Fig. 3). The absorption of 11 in the range of 400–250 nm is consistent with the computed vertical transitions at 428, 354, 315, and 280 nm at the EOM-CCSD/aug-cc-pVDZ level of theory (Supplementary Table 2). Additionally, a very weak band in the range of 750–550 nm also belongs to 11, as it corresponds to the computed transition at 859 nm and consequently explains its sensitivity to the red-light irradiation (830 nm). In line with the observation in the IR spectrum (Fig. 3d), the bands of 13 and 15 in the range of 230–400 vanish upon subsequent 266 nm laser irradiation. As a result, a broad band at 285 nm with onset at about 400 nm becomes discernible, and it associates with the complex mixture of the photolysis products SO2 (6), CH3SO2• (14), SO3 (17), and OCS (18). Concurrently, the characteristic absorptions at 670 and 630 nm for •NO362,63 forming from the O2-oxidation of •NO in the matrix appear. Further irradiation of the matrix with UV-light (365 nm) results in the formation of unknown species with weak absorption at 500 nm in the UV–vis spectrum.

Conclusion

In conclusion, we presented an optimized method for efficient gas-phase generation of the simplest organosulfur radical CH3S• (2) in the gas phase, opening the door to further studies on its structure and reactivity, particularly on its diverse reactions involving in Earth’s atmosphere and also the potent involvement in the astrochemistry of merthyl mercaptan (CH3SH) that has been recently detected in the interstellar medium (ISM)64,65,66. In addition to the first time identification of the characteristic absorption at 375 nm in the UV–vis spectrum of 2, its photo-induced (365 nm) sulfur atom transfer SAT to molecular nitrogen has been observed in an N2-matrix. Furthermore, two important peroxyl radicals CH3SOO• (13) and CH3S(O)2OO• (15) involving in the atmospheric oxidation of dimethyl sulfide have been generated by reacting 2 with molecular oxygen and characterized using IR and UV–vis spectroscopy in cryogenic Ar-, N2-, and Ne-matrixes. The assignment of all the IR-active fundamental modes in the range of 4000–400 cm−1 for both species is supported by 18O- and 13C-isotope labeling and quantum chemical computations. The spectroscopic characterization of the sulfur-containing radical species (CH3SOx, x = 0–4) and their photochemistry in the laboratory contribute to understanding the complex mechanism for the atmospheric oxidation of dimethyl sulfide.

Methods

Sample preparation

S-Nitrosothiol (CH3SNO) was prepared by reacting CH3SH with ClNO according to the published protocol46. Ar (≥99.999%, Messer), N2 (≥99.999%, Messer), O2 (≥99.999%, Messer), 15N2 (98 atom %, Aldrich), 18O2 (97 atom %, Aldrich) gases were used without further purification. For the 13C-labeling experiments, 13C-MeOH (99.5%, Eurisotop) was used for the synthesis of 13CH3SH (Supplementary Methods).

Matrix-isolation spectroscopy

Matrix IR spectra were recorded on a FT-IR spectrometer (Bruker 70 V) in a reflectance mode by using a transfer optic. A KBr beam splitter and MCT detector were used in the mid-IR region (4000–400 cm−1). Typically, 200 scans at resolution of 0.5 cm1 were co-added for each spectrum. Matrix UV–vis spectra were recorded on a UV–vis spectrometer (Lambda 850+, spectral range of 800–190 nm) in a transmission mode, and a scanning speed of 2 nm s−1 at resolution of 1 nm was used for each spectrum. For the preparation of the matrix, the gaseous sample (CH3SNO) was mixed by passing a flow of N2 or noble gas (Ar and Ne) through a cold U-trap (−110 °C) containing ca. 20 mg of the CH3SNO. Then the mixture (1:1000, estimated) was passed through an aluminum oxide furnace (o.d. 2.0 mm, i.d. 1.0 mm), which can be heated over a length of ca. 30 mm by a tantalum wire (o.d. 0.4 mm, resistance 0.4 Ω). The pyrolysate was deposited (2 mmol h1) in a high vacuum (~10−6 pa) onto the gold-plated copper block matrix support for IR or CaF2 window for UV–vis (3 K for Ne or 10 K for N2 and Ar) using closed-cycle helium cryostat (Sumitomo Heavy Industries, SRDK-408D2-F50H) inside the respective vacuum chambers. Temperatures at the second stage of the cold head were controlled and monitored using an East Changing T290 digital cryogenic temperature controller a Silicon Diode (DT-64). Photolysis experiments were performed using light emitting diodes (LED) (830/440 nm, 100 mW), UV flashlight (365 nm, 100 mW), and Nd3+:YAG laser (266 nm, MPL-F-266, 10 mW)

Computational details

Structural optimizations and IR frequencies were computed using both DFT M06-2X/6-311++G(3df,3pd)67 and CCSD(T)/aug-cc-pV(X + d)Z (X = D and T)68,69,70 methods. Local minima were confirmed by vibrational frequency analysis. EOM-CCSD/aug-cc-pVDZ71 computations were performed for the prediction of vertical excitations. All calculations we used default threshold, for CCSD(T) calculations we use the default active space i.e., the inactive space consists of all inner-shell orbitals, and the active space of all valence orbitals which are obtained from the atomic valence orbitals (full valence active space). The DFT computations were performed using the Gaussian 09 software package72. The ab initio computations were performed with MOLPRO program73.