Abstract

The theoretical and experimental vibrational frequencies of 3-(diacetylamino)-2-ethyl-3H-quinazolin-4-one (2) were investigated. The experimental Laser-Raman spectrum (4000–100 cm−1) and FT-IR spectrum (4000–400 cm−1) of the newly synthesized compound were recorded in the solid phase. Both the theoretical vibrational frequencies and the optimized geometric parameters such as bond lengths and bond angles have for the first time been calculated using density functional theory (DFT/B3LYP and DFT/M06-2X) quantum chemical methods with the 6-311++G(d,p) basis set using Gaussian 03 software. The vibrational frequencies were assigned with the help of potential energy distribution (PED) analysis using VEDA 4 software. The calculated vibrational frequencies and the optimized geometric parameters were found to be in good agreement with the corresponding reported experimental data. Also, the energies of the lowest unoccupied molecular orbital (LUMO), highest occupied molecular orbital (HOMO), and other related molecular energies for 3-(diacetylamino)-2-ethyl-3H-quinazolin-4-one (2) have been investigated using the same computational methods.

1. Introduction

Quinazoline derivatives are of interest because they show a variety of biological activities including anti-inflammation [1], antibacterial [2], antispasm [3], anticancer [4, 5], antiobesity [6] and reductase inhibitory properties [7]. Quinazolin-4-one derivatives have been synthesized by various methods, for example, reactions of carbon dioxide with 2-aminobenzonitrile without use of a catalyst in water [8]; reactions of N-substituted 2-bromobenzamides with formamide catalysed by CuI and 4-hydroxy-l-proline [9]; three-component reactions of benzyl halides, isatoic anhydride, and primary amines under mild Kornblum conditions [10]. We have shown that the quinazoline ring system can be easily modified via lithiation by a lithium reagent such as alkyllithium in anhydrous tetrahydrofuran at low temperature followed by reactions with electrophiles to provide access to substituted derivatives in high yields. Such derivatives might be difficult to synthesize by other means [1115].

As far as we are aware, there have been no previous reports of quantum chemical calculations or FT-IR and Laser-Raman spectral studies on 3-(diacetylamino)-2-ethyl-3H-quinazolin-4-one (2). Herein, we report experimental infrared and Raman spectra along with quantum chemical calculations, which correlate well with each other, to enable vibrational frequencies for compound 2 to be assigned.

2. Experimental Details

2.1. Characterization Techniques

FT-IR spectra over the range 4000–400 cm−1 were obtained on solid phase samples at room temperature using a Perkin-Elmer Spectrum Two FT-IR Spectrometer with 4 cm−1 resolution. Raman spectra within the range of 4000–100 cm−1 were obtained on solid phase samples using a Renishaw inVia Raman microscope (excitation line at 785 nm from a diode laser; 100 scans, resolution 1 cm−1).

Melting point determination was performed on a Gallenkamp melting point apparatus by the open capillary method. 1H (500 MHz) and 13C NMR (125 MHz) spectra were recorded on a Bruker AV500 spectrometer in deuterated dimethyl sulfoxide (DMSO-). Chemical shifts δ (ppm) are reported relative to tetramethylsilane (TMS) and coupling constants (J) are in Hz. DEPT spectra were used to detect the 13C multiplicities. Coupling patterns, integration values, and expected chemical shifts were used to assign signals. The low and high-resolution mass spectra were recorded on Waters GCT Premier and Waters LCT Premier XE instruments, respectively. A Nonius Kappa CCD diffractometer was used to record the X-ray single-crystal diffraction data by the use of graphite-monochromated Mo-Kα, ( = 0.71073 Å) radiation. The structure was solved by direct methods using SHELXS-96 [16] and refined with all data on full-matrix least squares using SHELXL-97 [17]. The full crystallographic data (CCDC 971829) for the title compound can be obtained via  http://www.ccdc.cam.ac.uk/structures.

2.2. Chemicals

Chemicals and reagents from Sigma-Aldrich were used without further purification and Fischer Scientific silica 60A (35–70 micron) was used in the purification of 2 by column chromatography.

2.3. Synthesis of 3-(Diacetylamino)-2-ethyl-3H-quinazolin-4-one (2)

Acetic anhydride (Ac2O; 2.1 mL, 22.0 mmol) in anhydrous toluene (5 mL) was added to a stirred mixture of 3-amino-2-ethyl-3H-quinazolin-4-one (1; 1.89 g, 10.0 mmol) and triethylamine (TEA; 5 mL) in dry toluene (20 mL) (Scheme 1). The mixture was refluxed for 30 min, allowed to cool, washed twice with saturated aq. NaHCO3 (10 mL) and water (15 mL), and dried (MgSO4), and the solvent was evaporated under vacuum. Column chromatography (silica; Et2O-hexane, 1 : 4) was used to purify the crude product to give a white powder. Crystallization from ethyl acetate provided colorless crystals of 2 (2.58 g, 9.4 mmol; 94%); m.p. 110–111°C. 1H NMR (500 MHz, DMSO-, δ, ppm): 8.16 (dd, , 1.2 Hz, 1 H, H-5), 7.94 (app. dt, , 8.0 Hz, 1 H, H-7), 7.76 (br app. d,  Hz, 1 H, H-8), 7.61 (app. dt, , 8.0 Hz, 1 H, H-6), 2.67 (q,  Hz, 2 H, CH2), 2.40 (s, 6 H, 2 CH3C=O), 1.26 (t,  Hz, 3 H, CH3CH2). 13C NMR (125 MHz, DMSO-, δ, ppm): 170.8 (s, 2 C=O), 159.4 (s, C-4), 158.0 (s, C-2), 146.9 (s, C-8a), 136.2 (d, C-7), 128.0 (d, C-5), 127.8 (d, C-6), 127.2 (d, C-8), 120.7 (s, C-4a), 25.7 (t, CH2), 24.9 (q, 2 CH3C=O), 10.4 (q, CH3CH2). CI-MS: (m/z, %): = 274 ([MH]+, 100). HRMS (CI): Calculated for C14H16N3O3 [MH]+ 274.1192; found: 274.1201.

3. Computational Details

The use of density functional theory (DFT) calculations has increased rapidly for various applications, particularly since accurate nonlocal corrections were introduced. The methods employed in the current work have been used in many previous theoretical studies [1832].

Initial atomic coordinates can generally be taken from experimental XRD results or a database. In this work, experimental XRD data and the Gauss View software database have each been used to determine initial atomic coordinates and to optimize the input structure. The most stable structure after optimization was obtained from the initial atomic coordinates taken from the Gauss View database [33]. This most stable structure, following optimization, was used for further theoretical analysis. The calculated gas phase ground state molecular structure of the title compound was optimized by the use of DFT/B3LYP and DFT/M06-2X methods with the 6-311++G(d,p) basis set, and the calculated optimized structure was used in the vibrational frequency calculations. The calculated harmonic vibrational frequencies were scaled by 0.9614 (B3LYP) and 0.9489 (M06-2X) for use with the 6-311++G(d,p) basis set, respectively [33, 34]. Identical scale factors were used for the entire spectra. The Gauss View molecular visualization program [33] and the Gaussian 03 program [35] were used to calculate vibrational wavenumbers, optimized geometric parameters, and other molecular properties. The calculated vibrational frequencies were assigned via potential energy distribution (PED) analysis of all the fundamental vibration modes by the use of the VEDA 4 program [36, 37] that has been used previously [20, 23, 24, 32, 38, 39]. All the vibrational assignments were based on the B3LYP/6-311++G(d,p) level calculations. Consequently, some assignments could correspond to the value of the next or previous vibrational frequency at the M06-2X/6-311++G(d,p) level.

4. Results and Discussion

4.1. Synthesis

3-(Diacetylamino)-2-ethyl-3H-quinazolin-4-one (2) was synthesized, in 94% yield after crystallization, by double acetylation of 3-amino-2-ethyl-3H-quinazolin-4-one (1) with acetic anhydride (two mole equivalents) in boiling toluene for 3 h in the presence of triethylamine (Scheme 1). Various spectroscopic techniques were used to confirm the structure of 2.

4.2. Geometric Structure

The single-crystal X-ray crystallographic analysis of 2 (C14H15N3O3) showed that the crystal belonged to the monoclinic system and the P21/c space group and possessed the following cell dimensions:  Å,  Å, and  Å and °, °, °, and  Å3. The measured density of compound 2 was 1.336 mg/mm3. Table 1 shows the bond lengths and bond angles for the optimized theoretical and experimental structures along with the atom numbering scheme (Figure 1).

As can be seen from Table 1, the optimized parameters calculated at both DFT levels differ only slightly from the experimental values, and small variations are to be expected since the calculations correspond to the gas phase rather than the solid state. For example, in the quinazoline moiety, the C4–N13 and C4–N35 bond lengths are calculated as 1.286 and 1.401 Å, respectively, by the B3LYP method, and 1.280 and 1.397 Å, respectively, by the M06-2X method, while the experimental values are 1.286 and 1.400 Å, respectively. Similarly, the N13–C4–N35 and C4–N35–C1 bond angles were calculated as 121.9 and 123.6°, respectively, by the B3LYP method and 122.3 and 123.9°, respectively, by the M06-2X method, while the experimental bond angles were 121.7 and 124.5°, respectively.

In order to quantify the level of agreement between the experimental and computational results, correlation coefficients () for linear regression analysis of the experimental and theoretical bond angles and lengths were calculated (last row in each section of Table 1). The values are 0.9820/0.9799 for bond lengths and 0.9596/0.9599 for bond angles for B3LYP/M06-2X. These values show that bond lengths calculated using the B3LYP method are slightly closer to the experimental data, but bond angles calculated using the M06-2X method are marginally closer to the data obtained experimentally.

The largest variations between the experimental and calculated (B3LYP/6-311++G(d,p)) bond lengths are 0.016 Å for the C1–N35 bond; 0.011 Å for the C3–N13 bond; 0.018 Å for the C8-C9 bond; 0.013 Å for the C15–C18 and C23–C25 bonds; and 0.015 Å for the C29-C31 bond. For bond angles the largest differences are 0.8° for N13–C4–C15; 0.9° for C23–N22–C29, C29–N22–N35, and C1–N35–C4; and 0.6° for C4–N35–N22.

The values found or calculated for 2 are consistent with values found experimentally or computationally for corresponding bonds and angles in related compounds such as 2-ethyl-3H-quinazoline-4-thione (experimental values available [40]), 3-amino-2-phenyl-3H-quinazolin-4-one (values calculated using the B3LYP/6-31, B3LYP/6-311++G(d,p), and M06-2X methods available [41]), 3-(4-methoxybenzylideneamino)-2-phenyl-3H-quinazolin-4-one (values calculated using the B3LYP/6-31 basis set available [42]), methyl 3,5-dibromo-2-diacetylaminobenzoate (experimental data available [43]), and N-(1-diacetylamino-1H-tetrazol-5-yl)acetamide (experimental data available [44]).

4.3. Vibrational Analysis

As seen from Figure 1, the molecule has 35 atoms, so there are 105 motions, 3 of which are translational, 3 of which are rotational, and 99 of which are vibration modes. The molecule has C1 symmetry. The experimental FT-IR and Laser-Raman spectra of compound 2 are compared with the selected theoretical spectra in Figures 2 and 3, respectively (identified bands are those discussed in the text and bold numbers in Table 2). The observed vibrational frequencies, scaled harmonic vibrational frequencies calculated at both B3LYP and M06-2X levels, and detailed potential energy distribution (PED) assignments are summarized in Table 2. The harmonic frequencies calculated for the title molecule relate to the gaseous phase, but the experimental ones are obtained for the solid phase. As a result, disagreement between the calculated frequencies and some of the experimental (observed) frequencies is to be expected. A PED analysis was carried out in order to introduce detailed vibrational assignments for compound 2. Within each fundamental wave number, the calculated modes are numbered downwards from the largest to the smallest frequency.

The harmony between the experimental and calculated wavenumbers is shown in Figure 4, which indicates that the experimental frequencies correlate well with the calculated ones, particularly for B3LYP. The relations between the experimental and calculated wavenumbers are linear and given by the following:

The correlation coefficients ( values) between the experimental and calculated wavenumbers were calculated as for B3LYP and for M06-2X. This indicates that the calculation methods gave reasonable agreement with the experimental values, particularly in the case of the B3LYP method. However, this masks significant differences between experimental and calculated values for a few of the vibrational frequencies, which are worthy of individual mention.

4.3.1. Quinazolinone Carbonyl Group Vibrations

The carbonyl group has a strong C=O stretching vibration around 1850–1550 cm−1 [45]. Differences of 20–30 cm−1 are used to help recognize particular functional groups such as ester, ketone, or amide groups. For compound 2, the quinazolinone νC=O mode is observed as a strong band at 1693 cm−1 in the IR spectrum and at 1694 cm−1 in the Laser-Raman spectrum. The calculated values are 1683 cm−1 (B3LYP) and 1721 cm−1 (M06-2X). This demonstrates that although the correlation coefficients are high, individual calculated frequencies can differ significantly from each other and from the experimental values, and in the case of the M06-2X method the calculated carbonyl stretching frequency differs from the experimental value by a sufficiently large margin that it could not be relied on to identify the functional group. In other cases, however, the level of agreement may be greater. For example, γONCC (O14–N35–C1–C2) is seen at 774 cm−1 in the IR spectrum and calculated at 774 cm−1 by B3LYP and 767 cm−1 by M06-2X. Bending modes δOCC (O14–C1–C2) were not seen in either the IR spectrum or the Raman spectrum, but two such modes were calculated to be at 394 (B3LYP)/400 (M06-2X) and 391 (B3LYP)/393 (M06-2X) cm−1. For similar quinazoline derivatives [46, 47] the νC=O mode was reported at 1652 (3-(2-hydroxyphenyl) methylene]amin-2-phenylquinazolin-4(3H)-one) and 1673 cm−1 (3-(4-fluorophenyl) methylene]amin-2-phenylquinazolin-4(3H)-one) in the IR spectrum and at 1656 and 1661 cm−1, respectively, in the Raman spectrum. For 3-(4-methoxybenzylideneamino)-2-phenyl-3H-quinazolin-4-one this mode was seen as a strong band at 1683 cm−1 in the IR spectrum and at 1682 cm−1 in the Raman spectrum and calculated to be 1690 cm−1 using the B3LYP/6-31 basis set [42].

4.3.2. C15H16, 17 and C18H19, 20, 21 Group Vibrations

The CH2 group has 6 vibrational modes (one asymmetrical and one symmetrical stretching, one scissoring, one wagging, one twisting and rocking). The asymmetric stretch CH2, the symmetric stretch CH2, the scissoring vibrations δCH2, and the wagging vibration ωCH2 appear at , , , and  cm−1, respectively [48, 49]. For compound 2, agreement between the experimental values and those calculated by B3LYP was generally good but there was less good agreement with those calculated by M06-2X (see Table 2). For example, the CH2 asymmetrical stretching mode was seen at 2934 cm−1 in the IR spectrum and 2938 cm−1 in the Raman spectrum. This asymmetrical stretching mode was calculated to be at 2939 (B3LYP)/2922 (M06-2X) cm−1. The CH2 symmetrical stretching mode was observed at 2915 cm−1 in the IR spectrum and 2914 cm−1 in the Raman spectrum. The symmetrical stretching mode was calculated to be at 2911 cm−1 by B3LYP, but 2888 cm−1 by M06-2X. Similarly, the CH2 scissoring and wagging modes were seen at 1409 and 1328 cm−1 in the IR spectrum, and 1406 and 1333 cm−1 in the Laser-Raman spectrum, respectively. These modes were calculated to be at 1406 (B3LYP)/1389 (M06-2X) and 1337 (B3LYP)/1331 (M06-2X) cm−1, respectively. CH2 twisting and rocking modes appeared at and  cm−1, respectively [49]. For compound 2, the CH2 twisting and rocking modes were seen at 1240 and 803 cm−1 in the IR spectrum, 1247 and 794 cm−1 in the Laser-Raman spectrum, respectively, and calculated to be at 1246 (B3LYP)/1246 (M06-2X) and 782 (B3LYP)/784 (M06-2X) cm−1, respectively.

The CH3 group has nine vibrational modes (two asymmetrical and one symmetrical stretching modes, two antisymmetrical deformations, one symmetrical deformation, two rocking modes, and 1 twisting mode). The CH3 asymmetric stretching vibrations are expected in the 2950–3050 cm−1 region and CH3 symmetric vibrations in the 2900–2950 cm−1 region [48, 49]. For compound 2 the asymmetric stretching modes were calculated to be at 3003 (B3LYP)/2988 (M06-2X) and 2989 (B3LYP)/2977 (M06-2X) cm−1, and the symmetric mode at 2928 (B3LYP)/2904 (M06-2X) cm−1. Asymmetric stretching modes were observed at 3003 cm−1 in the IR spectrum and at 2993 cm−1 in the Laser-Raman spectrum (the two bands appearing within a single envelope in each case), while symmetric stretching modes were observed at 2934 cm−1 in the IR and 2914 cm−1 in the Laser-Raman spectrum. Again, therefore, the B3LYP method is a better predictor of the experimental values than the M06-2X method. The same trend is observed for some of the other vibrational modes. For example, the symmetrical bending deformation, CH3, is expected at  cm−1 [48]. The calculated value was 1366 (B3LYP)/1353 (M06-2X) cm−1, while the observed value was 1364 cm−1 in the IR spectrum and 1363 cm−1 in the Laser-Raman spectrum. Similarly, aromatic molecules display a methyl rocking in the neighborhood of 1045 cm−1, and a second rocking in the region of  cm−1, which is more difficult to find among the C–H out-of-plane deformations [48]. For compound 2, these ρCH3 modes were observed at 1038 and 944 cm−1 in the IR spectrum; the first of these bands was observed at 1026 cm−1 in the Laser-Raman spectrum, but the second rocking mode was not observed. These bands were calculated at 1035 (B3LYP)/1038 (M06-2X) and 948 (B3LYP)/955 (M06-2X) cm−1, respectively.

Although the M06-2X calculated frequencies of some vibrational modes were a little closer to the observed values than the ones calculated by B3LYP, in the overwhelming majority of cases B3LYP provided calculated values that were closer to the observed values (see Table 2). This perspective is more easily seen in the graphical experimental and calculated spectra shown in Figure 2 (IR) and Figure 3 (Raman). Figure 3 also reveals that the calculated intensities of some of the bands in the Raman spectra were substantially different from the experimentally observed intensities.

4.3.3. Diacetylamino Group Vibrations

For compound 2, νC=O (C23–O24 and C29–O30) stretching modes were observed at 1735 (IR)/1743 (Raman) and 1725 (IR)/1725 (Raman) cm−1, and calculated at 1720 (B3LYP)/1764 (M06-2X) and 1718 (B3LYP)/1721 (M06-2X) cm−1. For 3-nitroacetanilide this mode was observed at 1676 cm−1 in the IR spectrum and calculated at 1774 cm−1 by using the HF/6-31 method [50]. For compound 2, two C–N stretching modes (N22–C23 and N22–C29) were observed at 1240 (IR)/1230 (Raman) and 1188 (IR)/1192 (Raman) cm−1, respectively, but the latter band contained two C–N stretching frequencies within the envelope; all three frequencies were calculated, at 1235 (B3LYP)/1227 (M06-2X), 1190 (B3LYP)/1198 (M06-2X), and 1180 (B3LYP)/1195 (M06-2X) cm−1.

In the acetyl group, the two CH3 groups have 18 vibrational modes (4 asymmetrical and 2 symmetrical stretching modes, 4 antisymmetrical and 2 symmetrical deformations, 4 rocking modes, and 2 twisting modes). The CH3 asymmetric stretching vibrations are expected to be at 2950–3050 cm−1 and the CH3 symmetric vibrations to be at 2900–2950 cm−1 [48, 49]. For compound 2, the symmetric stretching modes were seen at 2949 and 2934 cm−1 in the IR spectrum and at 2938 cm−1 in the Laser-Raman spectrum. They were calculated to be at 2943 (B3LYP)/2929 (M06-2X) and 2934 (B3LYP)/2916 (M06-2X) cm−1, again showing that the B3LYP method gives better agreement with experimental values.

4.3.4. Quinazolinone Group Vibrations

In quinazoline compounds, C=N stretching bands are generally observed at 1627–1566 cm−1 [5153], while for conjugated azines the νC=N mode has been reported to appear around 1553 cm−1 [54]. For compound 2, the (C4=N13) bands were seen at 1605 (IR)/1609 (Raman) and 1570 (IR)/1578 (Raman) cm−1 and calculated to be at 1593 (B3LYP)/1623 (M06-2X) and 1580 (B3LYP)/1601 (M06-2X) cm−1. Again, the B3LYP method provides the closer fit with the experimental values. Other C–N stretching bands [55] generally occur around 1100–1300 cm−1, while aromatic C–H stretching typically exhibits several weak to moderate bands above 3000 cm−1 [56]. Aromatic C–H in-plane bending vibrations usually occur in the region of 1390–990 cm−1 and are very useful for characterization purposes [57], while C–H out-of-plane deformations generally occur around 1000 to 700 cm−1, with bands having higher wavenumbers being weaker than those absorbing at lower wavenumbers [58]. Table 2 shows the experimental and calculated values for compound 2.

Ring C=C stretching vibrations usually occur around 1625–1430 cm−1 [59]. For benzenoid compounds there are two or three bands due to skeletal vibrations; the strongest band is at ca. 1500 cm−1. The observed and calculated values for compound 2 are shown in Table 2.

4.4. HOMO-LUMO Analysis

Chemical stability is mainly influenced by the frontier orbitals (HOMO and LUMO). The HOMO represents electron-donating capability, while the LUMO represents electron accepting capability [60]. The HOMO and LUMO energies of 2 were calculated by the B3LYP/6-311++G(d,p) (Figure 5; positive phase represented in red and negative phase represented in green) and M06-2X/6-311++G(d,p) methods. A large HOMO-LUMO gap implies a chemically “hard” molecule and a small HOMO-LUMO gap implies a “soft” molecule. The chemical reactivity of a molecule is also related to its “hardness,” molecules with lower HOMO-LUMO gaps being more reactive [61]. The frontier molecular orbital energy gap therefore helps in understanding the kinetic stability and reactivity of molecules [62, 63].

The HOMO-LUMO energy gap calculated for compound 2 is 5.119590424 e.V by B3LYP/6-311++G(d,p) and 7.459788024 e.V by M06-2X/6-311++G(d,p). This energy gap reflects the chemical activity of the molecule and influences its biological activity. As is evident from Figure 5, the HOMO is located on the quinazoline rings, the acetylamino group, and over the C1=O14 carbonyl group. The LUMO is more focused on the quinazoline ring and the C1=O14 carbonyl group and partially over the C29 atom.

The ionization energy () and electron affinity () can be expressed by HOMO and LUMO orbital energies as and . The global hardness . The ionization energy, along with electron affinity, can be used to give electronic chemical potential, , the global electrophilicity index , and the softness [64]. Such parameters were calculated and are tabulated in Table 3. The ionization potentials given by the B3LYP and M06-2X methods for 2 are 6.938958 eV and 8.084022128 eV, respectively. The HOMO and LUMO energies of 2-quinazolinone and 4-quinazolinone calculated by B3LYP/6-311G(d), 6-311+G(d,p), and 6-311++G(d,p) sets have been reported previously [58]. For both molecules, LUMOs are distributed uniformly over the entire molecules and show antibonding character. For both molecules, the HOMOs are shifted mainly towards the pyrimidinone and some part of the benzene ring [58].

5. Conclusion

In this work, the vibrational modes of newly synthesized 3-(diacetylamino)-2-ethyl-3H-quinazolin-4-one (2) were studied experimentally by use of FT-IR and Laser-Raman spectra and computationally using DFT/B3LYP and M06-2X methods. The vibrational harmonic frequencies, optimized geometric parameters, molecular orbital energies, and other properties related to HOMO and LUMO energy values of compound 2 were calculated using DFT/B3LYP and M06-2X methods with the 6-311++G(d,p) basis set. The assignments of the vibrational frequencies were made with the help of potential energy distribution (PED) analysis using VEDA 4 software. The theoretical optimized geometric parameters and vibrational frequencies have been found to be in good agreement with the corresponding experimental data and results in the literature. The calculated HOMO and LUMO orbitals and their energies have been used to gain understanding of charge transfer within compound 2.

Conflict of Interests

The authors declare that there is no conflict of interests regarding the publication of this paper.

Acknowledgment

The authors would like to extend their appreciation to the Deanship of Scientific Research at King Saud University for funding this research through the research group project RGP-VPP-239.