Vibrational frequencies and structure of 2-thiouracil by Hartree-Fock, post-Hartree-Fock and density functional methods

Vibrational frequencies and structure of 2-thiouracil byHartree�/Fock, post-Hartree�/Fock and density

functional methods

M. Alcolea Palafox a, V.K. Rastogi b,*, R.P. Tanwar b,1, Lalit Mittal b

a Departamento de Quimica-Fisica (Espectroscopia), Facultad de Ciencias Quimicas, Universidad, Compultense, Madrid 28040, Spainb Department of Physics, CCS University, Meerut 250 004, India

Received 23 May 2002; received in revised form 7 November 2002; accepted 7 November 2002

Abstract

Vibrational study of the biomolecule 2-thiouracil was carried out. Ab initio and density functional calculations were

performed to assign the experimental spectra. A comparison with the uracil molecule was made, and specific scale

factors were deduced and employed in the predicted frequencies of 2-thiouracil. Several scaling procedures were used.

The geometry structure of the molecule was determined. The effect of sulfur substitution at C2 position in the uracil

molecule, on the N1�/H and N3�/H frequencies and intensities reflects changes in proton donor abilities of these groups.

Calculations with the 6-31 G** basis set with HF and DFT methods appear in general to be useful for interpretation of

the general features of the IR and Raman spectra of the molecule. Using specific scale factors a very small error was

obtained. The use of these specific scale factors resolve and correct some of the controversial assignments in the

literature.

# 2003 Elsevier B.V. All rights reserved.

Keywords: Scaling procedures; Uracil; 2-Thiouracil

1. Introduction

Extensive work has been done with the struc-

tural analogues of uracil and many have been

found to exhibit interesting biological and che-

motherapeutic properties. The nucleic acid bases

with sulfur atom instead of oxygen have been a

subject of considerable interest since they were

detected in natural tRNAs [1]. Thiouracil deriva-

tives attract attention not only because of their

unclear role in nucleic acid structures, but also

because of exhibited pharmacological activities,

such as, an increase of the hypothyroidism effect

on blood [2], or a dietary product due to the effects

on thyroid activity suppression [3,4]. For example,

6-n -propyl-2-thiouracil is a potent antithyroid

drug [5], fluorinated-2-thiouracil derivatives reveal

antitumours [6] and antithyroid activity [7],5-

* Corresponding author. Tel.: �/91-120-278-0506/11; fax: �/

91-11-2241-3388.

E-mail address: [email protected] (V.K. Rastogi).1 Permanent address: Physics Department, GGDSD

College, Palwal-121002, India.

Spectrochimica Acta Part A 59 (2003) 2473�/2486

www.elsevier.com/locate/saa

1386-1425/03/$ - see front matter # 2003 Elsevier B.V. All rights reserved.

doi:10.1016/S1386-1425(02)00409-2

mailto:[email protected]

cyano-2-thiouracils and their derivatives are a newclass of leishmanicides [8]. A series of 1-amino-5-

substituted 4-thio or 2,4-dithio-uracils analogues

were assayed for anti-conflict and anestethic

activity in rats or mice [9].

In addition to these well known medicinal

properties of thiouracils (TU), others have recently

appeared, e.g. as marine corrosion inhibitors for

steel [10,11], or as dental adhesives in the treat-ment with metal surface [12].

In the thiouracils, 2-thiouracil (2TU) offers

special importance. 2-thirouracil has been identi-

fied in t-RNA and it plays an important role in

anticancer and antiviral activity [13]. The che-

motherapeutic activity of 2TU is due to its ready

incorporation into the nucleic acid [14], impeding

the melanoma tumours growth [15,16]. 2TU alsoinduces modifications in the thyroid gland [17,18],

and thus it is known as an antithyroid drug.

2TU forms complexes with some divalent metal

ions [19�/21], such as Mn, Co, Ni, Cu, Zn, Cd and

is also used in other less known applications, e.g. it

is added as additive at different concentrations in

metal solutions for an orientation of the crystal

grown [22].Structural studies of 2TU and some thio analo-

gues have been reported at the semiempirical level

[23�/25]. Electronic absorption and fluorescence

spectra have also been investigated [26]. Tauto-

marism, hydrogen bonding, solvent effect etc in 2-

TU and its analogous have also been a subject of

many rigorous ab inito studies [27�/33].

The experimental IR spectra of 2TU have beenpreviously reported in N2 matrices 34, in Ar

matrices [35,36] and in KBr [37]. Raman spectra

of polycrystalline samples have been also regis-

tered [37,38]. An IR study of the 2TU �/HCl

complexes in Ar matrices [39], as well as the effects

of autoassociation and hydrogen bonding with

water have been reported [40]. Laser Raman and

IR spectra of complex with Co(II), [41] Mn(II) [42]and Hg [38] have been also interpreted. The

assignments proposed in these studies have mainly

based on ab initio [34,36], and DFT [35] calcula-

tions. However, some doubts appear in the assign-

ment of several bands, which are now clarified,

using a better procedure for scaling the frequen-

cies.

2. Computational methods

Ab initio calculations with wavefunction-based

HF and MP2 [43,44] and Density Functional

Theory (DFT) [45] methods were carried out

with the 6-31G** basis set. As DFT methods,

the Becke exchange functional(B)[46] and Becke’s

three-parameter exchange functional (B3) [47,48]

in combination with both the correlational func-tional of Lee, Yang and Parr (LYP) [49], and with

the P86 [50,51] were selected. These procedures are

implemented in the GAUSSIAN 94 program pack-

age [52].

The optimum geometry was determined, with

the keyword OPT, by minimising the energy with

respect to all the geometrical parameters without

imposing molecular symmetry constraints. Thekeyword FREQ was used for frequency calcula-

tions.

3. Results and discussion

3.1. Geometry optimisation

The optimised bond lengths and angles in 2TUusing HF, MP2 and DFT methods are given in

Table 1, while the labelling of the atoms is plotted

in Fig. 1. For briefness, we will use the shorthand

notations, HF, MP2, B3LYP... etc, omitting the 6-

31G** basis set. For comparison purposes, in the

last column of Table 1 the experimental data

reported by electron diffraction of the molecule

of uracil are listed [53,54].In the bond lengths, except at the HF level, the

calculated values are very close to the electron

diffraction results. It is noted that the length of the

C�/N and C�/C single bonds are intermediate

between the corresponding aromatic and the

saturated bond. Thus some aromatic character is

on the ring structure. In the angles, among the

calculated values, there was not observed appreci-able difference at different levels, neither with the

electron diffraction data of uracil. The molecule is

completely planar, so torsional angles are 0 or

1808.When the sulfur atom is introduced in the

position 2 of the uracil molecule, replacing the

M.A. Palafox et al. / Spectrochimica Acta Part A 59 (2003) 2473�/24862474

Table 1

Geometrical parameters, bond lengths (in A) and angles (in degrees) of 2TU

Parameters HF/6-31G** MP2/6-31G** BLYP/6-31G** B3P86 B3LYP uracila

Bond lengths

N1�/C2 1.3522 1.3769 1.3943 1.3730 1.3784 1.395(6)b

C2�/N3 1.3495 1.3721 1.3816 1.3639 1.3685 1.391(6)b

N3�/C4 1.3977 1.4126 1.4381 1.4117 1.4191 1.415(6)b

C4�/C5 1.4602 1.4541 1.4647 1.4531 1.4580 1.462(8)

C5�/C6 1.3285 1.3527 1.3635 1.3484 1.3506 1.343(24)

N1�/C6 1.3744 1.3745 1.3843 1.3694 1.3751 1.396

N1�/H 0.9946 1.0092 1.0183 1.0096 1.0103 l.002c

C2�/S 1.6645 1.6494 1.6778 1.6574 1.6655 �/

C4�/O 1.1913 1.2264 1.2310 1.2156 1.2177 1.211(3)

C5�/H 1.0703 1.0776 1.0876 1.0804 1.0808 �/

Bond angles

N�/C2�/N 114.37 112.48 112.89 113.31 113.28 114.6(20)

C�/N3�/C 127.64 129.02 128.42 128.31 128.25 126.0(14)

N�/C4�/C 113.70 113.01 112.98 113.25 113.23 115.5(18)

C�/C5�/C 118.95 119.66 119.97 119.56 119.64 119.7(21)

C2�/N1�/H 115.90 115.11 114.85 115.02 115.09 115.7c

N3�/C2�/S 123.37 124.48 124.84 124.38 124.41 �/

C4�/N3�/H 115.88 115.13 115.40 115.52 115.50 �/

N3�/C4�/O 120.18 120.25 119.93 120.03 120.03 120.2

C�/C5�/H 118.49 118.69 118.28 118.44 118.35 118.1c

C�/C6�/H 123.02 123.07 123.23 123.03 123.05 122.8c

a Electron diffraction results, [44,45].b Mean value.c Fixed parameter.

Fig. 1. Labeling of the atoms in (a) uracil, (b) 2-thiouracil.

M.A. Palafox et al. / Spectrochimica Acta Part A 59 (2003) 2473�/2486 2475

Table 2

Vibrntional frequencies obtained in 2-thiouracil with different methods with the 6-31G** basis set

Number HF MP2 BLYP B3LYP B3P86 Characterization

Frequency Int.a Act.b Dep.c Frequency Int.a Frequency Int.a Frequency Int.a Frequency Int.a Forced

1 3890 14 38 0.20 3704 13 3532 14 3643 14 3664 14 8.53 100%, n (N1�/H)

2 3853 12 24 0.31 3660 9 3500 10 3609 10 3629 11 8.36 100%, n (N3�/H)

3 3432 0 100 0.21 3340 0 3183 0 3265 0 3283 0 6.96 85%, n (C5�/H)�/15%, n (C6�/H)

4 3399 1 69 0.47 3298 1 3140 1 3224 1 3238 1 6.75 85%, n (C6�/H)�/15%, n (C5�/H)

5 2009 86 45 0.22 1828 57 1731 87 1816 81 1836 78 19.62 71%, n (C�/O)�/22%, d (C5�/H)

6 1845 10 50 0.16 1706 3 1611 11 1681 11 1695 12 9.76 70%, n (C�/C)�/30%, n (ring)

7 1732 100 23 0.50 1615 100 1511 100 1576 100 1589 100 3.79 32%, d (N1�/H)�/31%, n (N1�/C6)�/

20%, n (NCN)�/17%, d (N3�/H)

8 1596 9 10 0.33 1491 4 1404 3 1460 5 1474 6 5.05 27%, n (N1�/C6)�/23%, d (N3�/H)�/

20%, n (C2�/N3)�/20%, d (C�/H)

9 1543 12 2 0.30 1440 4 1347 8 1407 9 1418 6 2.95 65%, n (C2�/N)�/35%, d (ring)

10 1537 0 13 0.45 1420 6 1335 8 1389 4 1389 3 1.66 32%, d (N3�/H)�/31%, d (N1�/H)�/

20%, d (ring)�/17%, d (C6�/H)

11 1349 8 20 0.72 1276 8 1199 2 1238 2 1249 5 2.05 38%, d (C5�/H)�/25%, d (C6�/H)�/

18%, n (C6�/N1)�/10%, d (N1�/H)

12 1315 13 16 0.32 1244 20 1153 43 1212 29 1222 18 1.25 32%, d (C6�/H)�/25%, d (N1�/H)�/

25%, n (C�/N3)�/20%, d (N3�/H)

13 1249 34 9 0.21 1214 20 1109 11 1162 18 1179 19 2.55 75%, n (C�/S)�/25%, d (ring)

14 1162 0 3 0.41 1106 1 1046 2 1085 1 1093 1 1.28 40%, d (C5�/H)�/25%, d (ring)�/

20%, d (N1�/H)�/15%, d(C6�/H)

15 1088 1 4 0.19 1013 1 967 2 1003 1 1010 1 2.88 66%, d (NCN)�/34%, d (ring)

16 1110 0 4 0.75 940 0 924 0 972 0 972 0 0.71 62%, g (C6�/H)�/36%, g (C5�/H)

17 979 1 3 0.32 939 1 868 1 911 1 924 0 1.93 70%, n (N3CC)�/30%, d (ring)

18 910 12 2 0.75 800 8 783 10 816 10 818 11 0.82 38%, g (C5�/H)�/27%, g (C�/C)�/

21%, g (C6�/H)�/14%, g (C�/O)

19 828 4 2 0.75 750 11 711 6 745 5 750 4 0.52 44%, g (C5�/H)�/32%, g (N3�/H)�/

12%, g (C6�/H)�/12%, g (C�/O)

20 767 2 6 0.15 730 1 694 1 719 1 725 1 2.60 100%, n (ring)

21 780 7 2 0.75 701 2 692 11 716 10 718 10 0.45 50%, gCN3�/H)�/25%, n (Nl�/H)�/

20%, g (C5�/H)

22 719 4 5 0.75 629 0 609 4 643 4 649 5 1.04 60%, g (N1�/C2)�/25%, g (N1�/H)�/

15%, g (ring)

23 671 7 2 0.75 614 7 596 5 620 5 626 4 0.31 60%, g (N1�/H)�/14%, g (N3�/H)�/

14%, g (C5�/H)�/12%, g (ring)

24 581 1 3 0.25 536 1 517 8 537 I 538 1 1.47 100%, d (ring)

25 540 2 1 0.27 494 1 475 2 493 2 493 1 0.78 65%, d (C�/O)�/35%, d (ring)

26 482 3 4 0.34 465 1 433 2 452 2 456 2 1.31 70%, d (C�/S)�/30%, d (ring)

27 438 1 3 0.75 384 2 387 2 402 2 402 2 0.30 65%,g (C�/C�/H)�/35%, g (ring)

M.A

.P

ala

fox

eta

l./

Sp

ectroch

imica

Acta

Pa

rtA

59

(2

00

3)

24

73�

/24

86

24

76

oxygen atom, a small change is observed in thering structure: a slightly reduction in the N1�/C2

and N3�/C2 lengths, e.g. 0.017 A at BLYP level,

and a very slight opening of the N�/C2�/N angle,

e.g. 0.68 at BLYP and closing of the C�/C5�/C,

C4�/N3�/H and N3�/C4�/O angles.

3.2. Vibrational frequencies

The vibrational bands computed with theoreti-cal methods used by us are shown in Table 2. The

second column lists their frequencies calculated at

the HF level with their relative infrared intensities

(the third column), their Raman activities (the

fourth column), and the depolarisation ratios (the

fifth column). The relative intensities (and activ-

ities) were obtained by normalising the computed

value to the intensity of the strongest line, no 7(and 3, respectively) at the HF level. The 6�/7th

columns show the results at MP2 level, while 8�/

14th columns refer to the DFT methods of BLYP,

B3LYP and B3P86, respectively. In the last

column of this table appears the percent contribu-

tion of the different modes to a computed fre-

quency. The numbers correspond to the average

values obtained with theoretical methods used byus, except with MP2 due to their worse results.

Contributions lower than 10% were not consid-

ered.

The characterisation of all the normal modes at

B3LYP level is plotted in Fig. 2, which display the

atomic displacement vectors for each computed

frequency. These displacements are represented as

xyz coordinates, in the standard orientation,which were plotted to identify each vibration. No

significant difference in the plot was observed at

the other levels of computations. In the figure the

motion is drawn only when the displacement

vectors on the X , Y and Z axes are higher than

0.07 on the carbon atoms, 0.06 on the nitrogen

atoms. 0.05 on the oxygen atom, 0.04 on the sulfur

atom, and higher than 0.15 on the hydrogenatoms.

To improve the computed frequencies, in Table

3 the scaled frequencies also appear which are

obtained by three procedures: with an overall scale

factor, with a scaling equation, and with specific

scale factors for each mode (Table 4). With theTa

ble

2(C

on

tin

ued

)

Nu

mb

erH

FM

P2

BL

YP

B3

LY

PB

3P

86

Ch

ara

cter

izati

on

Fre

qu

ency

Int.

aA

ct.b

Dep

.cF

req

uen

cyIn

t.a

Fre

qu

ency

Int.

aF

req

uen

cyIn

t.a

Fre

qu

ency

Int.

aF

orc

ed

28

29

41

40

.70

27

61

25

82

27

02

27

11

0.4

98

0%

,d

(SC

NC

O)�

/20

%,d

(rin

g)

29

17

80

10

.75

15

60

16

30

16

80

16

80

0.0

91

00

%,g

(rin

g)

30

15

20

00

.75

12

70

12

20

13

30

13

40

0.0

87

0%

,g

(N� /

C�/

S)�

/30

%,g

(rin

g)

aR

ela

tive

infr

are

din

ten

sity

,%

.b

Rel

ati

ve

Ra

ma

na

ctiv

ity

,%

.c

Dep

ola

riza

tio

nra

tio

s.d

Fo

rce

con

stan

tin

mD

yn

e/A

.


first procedure, by using a uniform scale factornexp/vcalc the results shown in the columns 2nd�/

6th are obtained. The scale factors were from [55].

For frequencies lower than 800 cm�1 it is recom-

mended to use a new scale factor [55]. With these

predicted scaled frequencies the error obtained is

in general lower than 5%. The rms error is

approximately 25 cm�1, Table 5. However, with

the second procedure, using specific scaling equa-

tions obtained from uracil molecule instead of anoverall scale factor, a better accuracy is obtained

in the predicted frequencies, the 7�/10th columns

of Table 3. The rms error is approximately 19

cm�1, Table 5. The scaling equations used for the

6-31G** basis set were

nscaled HF �/ 6.6�/0.8867 vcalculated HF (r�/0.9992)nscaled BLYP �/ 44.7�/0.9663 vcalculated BLYP (r�/0.9993)nscaled B3P86 �/ 33.1�/0.9332 vcalculated B3P86 (r�/0.9992)nscaled B3LYP �/ 33.5�/0.9391 vcalculated B3LYP (r�/0.9993)

Again a new improvement can be carried out on

the computed frequencies with the 3rd procedure,

using specific scale factors for each mode, the 11�/

14th columns of Table 3. These specific scale

factors were determined from uracil molecule.

The scaled frequencies obtained are listed in the

15�/18th columns. The rms error appears in Table

5 with this last procedure, when the molecule of

reference is selected adequately, the most accuracy

is obtained [56]. In 2TU it was extremely accurate

with HF and DFT methods, with a rms error

approximately 8 cm�1. It is because the substitu-tion of oxygen by sulfur in uracil, results in a small

change in the positions of the bands of the

spectrum (in contrast to the conclusion reported

in [34]), except the obvious changes due to the C�/S

group, which give rise to good transferability of

the scale factors of uracil to 2TU.

The columns 15th and 16th of Table 3 collect

the infrared data of 2TU in N2 [34] and in Ar [35]matrices. In parenthesis appears the relative in-

tensity, in percentage. The columns 17th and 18th

list the IR and Raman data reported in KBr [37]

and in polycrystalline sample [38], respectively.

Many discrepancies with the data of columns 15th

and 16th are observed, especially in the assign-

ments. Thus they were not considered in the

present discussion. For identification purposes,

the last column shows briefly the characterisationgiven in the last column of Table 2.

A mislay is observed in [35] in the correspon-

dence established between three scaled frequencies

and the experimental values. Thus the scaled

wavenumbers at 1431, 1379 and 1362 cm�1 by

B3LYP (1431, 1382, and 1363 cm�1 by MP2, and

at 1437, 1389 and 1383 cm�1 by HF), unrelated to

experimental values in [35], should be correlatedwith the IR bands at 1430, 1376 and 1363 cm�1,

respectively.

Table 5 shows the absolute errors obtained in

the calculated and in the scaled frequencies by the

three procedures. Large errors were determined in

the calculated wavenumbers by HF although they

were remarkably reduced with the use of a scaling

procedure. Thus the HF results should be scaledalways. By contrast, BLYP and B3LYP, except for

wavenumbers higher than 2000 cm�1, give good

calculated wavenumbers, although obviously, the

use of a scaling procedure reduces the rms error.

It is also noted that the results of MP2 calcula-

tions were always somewhat less accurate than the

HF results. Thus in Tables 3 and 5 they were in

general omitted. The only purpose of introducingthe columns of MP2 for the prediction of the

frequency with an overall scale factor, is to

compare the results obtained using the scale

factors recommended in [55], the 7th column of

Table 5, with those obtained using the value of

0.96 utilised [35] in the prediction of 2TU, the 12th

column. Although a slightly lower error is ob-

tained by MP2 for the prediction of frequenciesbelow 2000 cm�1 using the scale factor [35] of

0.96, it fails for higher frequencies. Thus we

obtained a lower rms error in the predicted

frequencies than in [35].

The calculated frequency with BLYP was the

most accurate, with a rms error of 32 cm�1.

However, the error was not reduced with the use

of an overall scale factor, and slightly with the useof a scaling equation. Only the use of specific scale

factors remarkably reduce the rms error to the

very low value of 8, analogously to other DFT

methods. It can be attributed to the fact that there

is not a systematic error in the BLYP procedure

that can be corrected using an overall scale factor

or a scaling equation.


Theoretical infrared and Raman spectra of 2-

TU are shown in Fig. 3. The wavenumbers

correspond to those scaled using specific scale

factors for each mode, except those marked with

(*) which were scaled with the scaling equation.

It is observed a good agreement between the

theoretical IR and Raman spectra with the corre-

sponding experimental ones, especially in the

wavenumbers. In the intensity it is also noted a

good concordance with the experimental spectra,

Fig. 2. Characterization of the normal modes in 2-thiouracil by B3LYP/6-3IG**.


Table 3

Scaled frequencies by several procedures and methods in 2TU

No. With an overall scale factora With a scaling equation With specific scale factors Experimental Characterization

HF�/

0.8992

MP2�/

0.9370

BLYP�/

0.9945

B3LYP�/

0.9614

B3P86�/

0.9558

HF BLYP B3LYP B3P86 HF BLYP B3LYP B3P86 IR in Ar

matrixg

IR in N2

matrixh

IR in

KBr1

Ramanj

1 3498 3471 3513 3502 3502 3456 3458 3455 3452 3474 3470 3471 3468 3457(24) 3439(58) �/ �/ n (N1�/H)

2 3465 3429 3481 3470 3469 3423 3427 3423 3420 3426 3426 3425 3422 3415(13) 3399(41) �/ �/ n (N3�/H)

3 3086 3130 3165 3139 3138 3050 3120 3100 3097 3087 3091 3085 3086 �/ �/ 3086 s 3084 w n (C5�/H)�/n (C6�/H)i

4 3056 3090 3123 3100 3095 3020 3079 3061 3055 3081 3081 3079 3080 �/ �/ 2928 s �/ n (C6�/H)�/n (C5�/H)i

5 1806 1713 1721 1746 1755 1788 1717 1739 1746 1753 1749 1749 1747 1738(87) 1732(49) 1686 vs 1682 s n (C�/O)�/d (C5�/H)i

6 1659 1599 1602 1616 1620 1643 1601 1612 1615 1641 1634 1634 1634 1634(5) 1635(5) 1628 m 1627 m n (C�/C)�/n (ring)i

7 1557 1513 1503 1515 1519 1542 1505 1514 1516 1548 1539 1540 1539 1534(100) 1541(100) 1566 vs 1555 m d (N1�/H)�/n (N1�/C6)

8 1435 1397 1396 1404 1409 1422 1401 1405 1409 1434 1430 1436 1444 1430(2) 1432(7) 1421 m �/ n (N1�/C6)�/d (N3�/H)

9 1387 1349 1340 1353 1355 1375 1346 1355 1356 1379 1394 1389 1390 1376(6) 1396(13) 1395 m 1397 m n (C2�/N)�/d (ring)i

10 1382 1331 1328 1335 1328 1369 1335 1338 1329 1366 1372 1367 1361 1363(1) 1372(6) �/ d (N3�/H)�/d (N1�/H)

11 1213 1196 1192 1190 1194 1203 1203 1196 1199 1219 1223 1224 1226 1223(5) 1225(10) 1215 vs 1222 vs d (C5�/H)�/d (C6�/H)

12 1182 1166 1147 1165 1168 1173 1159 1172 1173 1192 1207 1198 1196 1191(24) 1197(32) 1177 s �/ d (C6�/H)�/d (NI�/H)

13 1123 1138 1103 1117 1127 1114 1116 1125 1133 �/ 1148(31) 1148(47) �/ �/ �/ 1160 s �/ n (C�/S)�/d (ring)i

14 1045 1036 1040 1043 1045 1037 1055 1052 1053 1061 1070 1067 1068 1060(1) 1062(1) 1073 m �/ d (C5�/H)�/d (ring)i

15 978 949 962 964 965 971 979 975 976 993 994 995 998 986(2) 994(4) 1003 m �/ d (NCN)�/d (ring)i

16 998 881 919 934 929 991 938 946 940 963 964 965 964 �/ �/ �/ �/ g (C6�/H)�/g (C5�/H)i

17 890b 961c 922d 912e 917f 875 883 889 895 891 903 904 908 907(1) 906(1) 912 m �/ d (N3CC)�/(ring)i

18 827b 818c 832d 817c 812f 813 801 800 796 �/ �/ �/ �/ 806(9) 811(18) 837 m 839 w g (C5�/H)�/n (C�/C)

19 753b 767c 755d 746c 744f 741 732 733 733 740 728 734 736 727(1) 744(5) 736 m �/ g (C5�/H)�/(N3�/H)

20 697b 747c 737d 720d 719f 687 715 709 710 702 711 707 705 710(1) 717(6) 710 m 718vs d (ring)

21 709b 717c 735d 717d 7l2f 698 713 706 703 707 688 690 686 694(11) 714(5) �/ �/ g (N3�/H)�/g (N1�/H)i

22 653b 643c 647d 644d 644f 644 633 637 639 �/ �/ �/ �/ 643(3) 655(12) 648 m �/ g (Nl�/C2)�/g (Nl�/H)’

23 610b 628c 633d 621d 621f 602 621 616 617 629 605 607 604 604(5) 635(3) 580 m �/ g (Nl�/H)�/g (N3�/H)’

24 528b 548c 549d 538e 534f 522 544 538 535 534 532 534 535 530(2) 531(4) 548 vs 539m d (ring)

25 491b 505c 504d 494e 489f 485 504 496 493 488 491 488 487 491(2) 492(4) �/ �/ d (C�/O)�/d (ring)i

26 438b 476c 460d 453e 452f 434 463 458 459 �/ �/ �/ �/ 451(2) �/ 454 m 454 m d (C�/S)�/d (ring)i

27 398b 393c 411d 403e 399f 395f 419 411 408 420 419 417 416 395(2) 402(3) 413 m 416 w g(C�/C�/H)�/g(ring)i

28 267b 282c 274d 270e 269f 267 294 287 286 270 274 274 275 269 (2) 271(3) �/ 285 m d (SCNCO)�/d (ring)i

29 162b 160c 173d 168e 167f 164 202 191 190 184 183 183 182 �/ �/ �/ �/ g (ring)

30 138b 130c 130d 133e 133f 141 163 158 158 �/ �/ �/ �/ �/ �/ �/ �/ g (N�/OS)�/g (ring)i

a [46].b With the scale factor of 0.9089 recommended for the prediction of low-frequency vibrations [46].c �/1.0229 [46].d �/1.0620 [46].e �/1.0013 [46].f �/0.9923 [46]g [28].h [27].i Low contribution of this mode.j [30,31].

M.A

.P

ala

fox

eta

l./

Sp

ectroch

imica

Acta

Pa

rtA

59

(2

00

3)

24

73�

/24

86

24

80

being the most intense ones those calculated withhigher relative intensity.

It is noted that the degree of coupling of the

modes in 2TU is slightly higher than in uracil. An

analysis of the different modes in 2TU is as

follows:

3.2.1. Bands due to NH vibrations

Two bands due to stretching vibrations of N1�/

H and N3�/H are predicted accurately at 3450�/

3470 and 3420�/3427 cm�1, respectively, by using

a scale equation or specific scale factors and with

all of our methods. The band at higher wavenum-

ber and intensity corresponds to the N1�/H group,

in agreement with the experimental, and the lower

wavenumber and intensity to the N3�/H group.

Comparing the spectral positions of these vibra-tions in 2TU with the IR spectrum of uracil [56], it

is noted that the replacement of an oxygen atom

by sulfur leads to a shift of the n (N�/H) bands to

lower wavenumber, 28 cm�1 for the N1�/H mode

and 20 cm�1 for the N3�/H mode. This slightly

larger change in the N1�/H mode than the N3�/H

mode is in accordance with a smaller proximity

(ca. 0.03 A) of H7 to the sulfur atom than the H9atom. This slightly larger attraction of the sulfur

atom to H7 than H9, give rises to a slight lower

positive charge on H7 than H9, table 6.

The bending modes of N1�/H and N3�/H groups

are more coupled than in uracil. This feature can

be related with the fact that in 2TU the most

intense band of the whole spectrum, experimen-

tally observed at 1534 cm�1, is characterised as d

(N1�/H)�/n (N1�/C6)�/d (N3�/H) with an almost

equal contribution of the bending and stretching

modes, while in uracil it is n (C�/N)�/d (N1�/H)

with a remarkable lower contribution of the N�/H

bending modes. That is to say, the main contribu-

tion to the intense band at approximately 1550

cm�1 comes from N�/H bending vibrations in

2TU, and from C�/N or from ring vibrations in thecase of uracil and N-methylated derivatives[34].

This feature, rather unusual for a carbonyl com-

pound, is properly predicted by our theoretical

calculations. In 2TU derivatives, different ring

atoms contribute strongly to the intensity of this

vibration [34].

The out-of-plane wagging vibrations of the N1�/

H and N3�/H groups appear in the 750�/600 cm�1

range. Our theoretical calculations, even the MP2

method, reproduce well the spectral positions and

relative intensities of the bands that involve these

vibrations, allowing their clear correspondence

with the experimental spectra.

It is noted that the stretching and bending

vibrations of the N1�/H group appear at higherfrequencies than the N3�/H group (i.e. in the

stretching is 42 cm�1 in Ar matrix, and 34 cm�1

by B3LYP), while the N3�/H out-of-plane vibra-

tion appears at higher wavenumber than the N1�/

H (i.e. 90 cm�1 in Ar matrix, and 96 cm�1 by

B3LYP). This fact is also observed in uracil,

although with slightly larger values, i.e. in the

stretching 50 and 38 cm�1, in Ar and by B3LYP,respectively, and in the out-of-plane 111 and 124

cm�1, respectively. These features are also com-

mon in other thiouracils [34�/36] although with

slightly larger values, and may be due to the fact

that the N1�/H group interacts with only one of

the C�/S (or C�/O in uracil) group, while the N3�/

H group has the combined influence of a C�/O and

C�/S (C�/O in uracil) groups.

3.2.2. Bands due to C�/N vibrations

The C�/N stretching mode appears in general

coupled with N�/H and C�/H bendings, e.g. the

calculated bands by B3LYP at 1576, 1460 and

1407 cm�1, and slightly at 1212 cm�1, Table 2,

and related to the IR bands in Ar matrix at 1534,

1430, 1376 and 1191 cm�1, respectively, Table 3.

3.2.3. Bands due to C�/C vibrations

If specific scale factors from uracil are used, our

theoretical methods predict accurately the frequen-

cies of C�/C group, especially in the stretching.

Compared with uracil, in 2TU a small change

towards a lower wavenumber, approximately 10

cm�1 in the stretching and approximately 15

cm�1 in the out-of-plane, is observed.The C�/C stretch is very sensitive to sulfur

substitution. Thus for 2TU and all its derivatives,

the band is relatively weak, similar to uracil; but it

is almost as strong as the C2�/O stretch for 4TU

and its derivatives [34], because in this case the C�/

C motion no longer couples with the C4�/S stretch.


3.2.4. Bands due to C�/O and C�/S vibrations

The typical pattern of the absorption bands due

to C�/O stretching vibrations of uracil and its

derivatives is in general very complex [35,36]. In

the present study the band appearing at 1738

cm�1 in IR matrix is close to our prediction, and

therefore, is related to the C�/O stretching mode.

Similarly, the band observed at 1148 cm�1 is

assigned to the n (C2�/S) mode. In both cases, the

stretching modes are clearly identified in the

spectrum by their strong intensity, in agreement

with our calculations. It is observed that the ring

breathing and Kekuley stretching mode for 2TU

have lower magnitudes as compared with those for

uracil, which could be due to the mass effect of the

S atom in place of the oxygen atom.The out-of-plane g (C�/S) vibration was char-

acterised at 130�/160 cm�1, in agreement with the

spectral region of this mode [57]. This assignment

corrects that reported in [34,45], which identified

partially the motions of g (N1�/H) and g (C�/S),

but without a view of the whole motion of thestructure. The normal mode at 643 cm�1 (Fig. 2)

in which appear the g (N1�/C2) mode is in the

range of this vibration [58,59], and it represents the

60% of contribution (Table 2) that characterises

this band.

4. Summary and conclusions

The effect of sulfur substitution on the N1�/H

and N3�/H frequencies and intensities reflects

changes in proton donor abilities of these groups.

This is expected to affect the strength of hydrogen

bonding in which they participate, particularly

that formed by the biologically significant N3�/H

group.The C�/S stretching mode appears as a relatively

strong band in the 1070�/1150 cm�1 range, and

coupled with the vibrations of other groups as in

the case of the C�/O stretches.

Without scaling, the DFT methods are remark-

ably better (especially BLYP) than HF and MP2.

The use of a scale factor or a scale equation,

remarkably reduces the error (except by BLYP)approximately three times. In HF the reduction is

approximately eight times. The use of a scale

factor or a scale equation with the BLYP results

is not recommended, and only specific scale factor

should be used to improve the frequencies.

Calculations with the 6-31 G** basis set with

HF and DFT methods appear in general to be

useful for interpretation of the general features ofthe IR and Raman spectra of the molecule.

Using specific scale factors a very small error

was obtained, in general lower than 5%, being

quite similar to HF and DFT methods. The results

performed at the MP2 level were somewhat worse.

The use of these specific scale factors resolve and

correct some of the controversial assignments in

the literature.

Acknowledgements

One of the authors (V.K. Rastogi) is grateful to

Professor Ramesh Chandra, Vice Chancellor, CCS

University, Meerut, India, for the encouragement

Table 4

Specific scale factors used for scaling the wave numbers of

uracil derivatives

Number HF BLYP B3LYP B3P86

1 0.8931 0.9825 0.9527 0.9465

2 0.8892 0.9789 0.9489 0.9429

3 0.8994 0.9710 0.9449 0.9400

4 0.9063 0.9812 0.9550 0.9511

5 0.8727 1.0104 0.9629 0.9514

6 0.8896 1.0142 0.9723 0.9642

7 0.8937 1.0187 0.9774 0.9684

8 0.8985 1.0182 0.9838 0.9797

9 0.8938 1.0350 0.9872 0.9802

10 0.8889 1.0280 0.9841 0.9798

11 0.9035 1.0201 0.9886 0.9815

12 0.9066 1.0469 0.9883 0.9785

14 0.9132 1.0229 0.9835 0.9772

15 0.9126 1.0283 0.9919 0.9879

16 0.8676 1.0433 0.9928 0.9918

17 0.9098 1.0402 0.9927 0.9826

19 0.8941 1.0243 0.9849 0.9809

20 0.9156 1.0243 0.9832 0.9718

21 0.9068 0.9940 0.9636 0.9553

23 0.9371 1.0147 0.9787 0.9650

24 0.9198 1.0299 0.9942 0.9942

25 0.9039 1.0328 0.9908 0.9871

28 0.9178 1.0625 1.0156 1.0156

29 1.0335 1.1212 1.0882 1.0819


Table 5

Absolute errors, D (nscaled�/nexperimental), obtained in the calculated and in the scaled wavenumbers of 2TU

Calculated frequencies With an overall scale factor With a scalig equation With specific scale factors

Number HF BLYP B3LYP B3P86 HF MP2 BLYP B3LYP B3P86 HFa MP2a B3LYPa HF DLYP B3LYP B3P86 HF BLYP B3LYP B3P86

1 433 75 186 207 41 14 56 45 45 44 99 113 �/1 1 �/2 �/5 17 13 14 11

2 438 85 194 214 50 14 66 55 54 52 99 122 8 12 8 5 11 11 10 7

5 271 �/7 78 98 68 �/25 �/17 8 17 70 17 41 50 �/21 1 8 15 11 11 9

6 211 �/23 47 61 25 �/35 �/32 �/18 �/14 27 4 14 9 �/33 �/22 �/19 7 0 0 0

7 198 �/23 42 55 23 �/21 �/31 �/19 �/15 25 16 11 8 �/29 �/20 �/18 14 5 6 5

8 166 �/26 30 44 5 �/33 �/34 �/26 �/21 7b 1b 1b �/8 �/29 �/25 �/21 4 0 6 14

9 167 �/29 31 42 11 �/27 �/36 �/23 �/21 13b 6b 3b �/1 �/30 �/21 �/20 3 18 13 14

10 174 �/28 26 26 19 �/32 �/35 �/28 �/35 20b 0b �/1b 6 �/28 �/25 �/34 3 9 4 �/2

11 126 �/24 15 26 �/10 �/27 �/31 �/33 �/29 �/9 2 �/10 �/20 �/20 �/27 �/24 �/4 0 1 3

12 124 �/38 21 31 �/9 �/25 �/44 �/26 �/23 �/8 3 �/3 �/18 �/32 �/19 �/18 1 16 7 5

13 101 �/39 14 31 �/25 �/10 �/45 �/31 �/21 �/24 17 �/8 �/34 �/32 �/23 �/15 �/ �/ �/ �/

14 102 �/14 25 33 �/15 �/24 �/20 �/17 �/15 �/14 2 3 �/23 �/5 �/8 �/7 1 10 7 8

15 102 �/19 17 24 �/8 �/37 �/24 �/22 �/21 �/7 �/14 �/3 �/15 �/7 �/11 �/10 7 8 9 12

17 72 �/39 4 17 �/17 54 15 5 10 �/26 �/6 �/14 �/32 �/24 �/18 �/12 �/16 �/4 �/3 1

18 104 �/23 10 12 21 12 26 11 6 14 �/38 �/6 7 �/5 �/6 �/10 �/ �/ �/ �/

19 101 �/16 18 23 26 40 28 19 17 19 �/7 3 14 5 6 6 13 1 7 9

20 57 �/16 9 15 �/13 37 27 10 9 �/20 �/9 �/5 �/23 5 �/1 0 �/8 1 �/3 �/5

21 86 �/2 22 24 15 23 41 23 18 8 �/21 84 19 12 9 13 �/6 4 �/8 �/

22 76 �/34 0 6 10 0 4 1 1 4 �/39 �/13 1 �/10 �/6 �/4 �/ �/ �/ �/

23 67 �/8 16 22 6 24 29 17 17 0 �/15 4 �/2 17 12 13 25 1 30 �/

24 51 �/13 7 8 �/2 18 19 8 4 �/7 �/15 �/4 �/8 14 8 5 4 2 4 5

25 49 �/16 2 2 0 14 13 3 2 �/7 �/17 �/8 �/6 13 5 2 �/3 0 �/3 �/4

26 31 �/18 1 5 �/13 25 9 2 1 �/17 �/5 �/8 �/17 12 78 �/ �/ �/ �/ �/

27 43 �/8 7 73 �/2 16 8 4 �/1 �/26 �/1 0 24 16 13 �/ �/ �/ �/ �/

28 25 �/11 1 2 �/2 13 51 0 �/5 �/4 �/5 �/2 25 18 17 1 5 5 6 �/

Rms 163 32 59 68 23 26 32 23 21 24 32 35 18 21 20 14 11 8 7 8

a With the scaled frequency of [27], using a scale factor of 0.90 for HF, 0.96 for MP2, and 0.98 for B3LYP.b With our reassignments.

M.A

.P

ala

fox

eta

l./

Sp

ectroch

imica

Acta

Pa

rtA

59

(2

00

3)

24

73�

/24

86

24

83

Fig. 3. Theoretical infrared and Raman spectra of 2-TU. The wavenumbers correspond to those scaled using specific scale factors for each mode, except those marked

with (*) which were scaled with the scaling equation.

M.A

.P

ala

fox

eta

l./

Sp

ectroch

imica

Acta

Pa

rtA

59

(2

00

3)

24

73�

/24

86

24

84

and useful discussions during the course of thiswork.

References

[1] M. Yaniv, W.R. Folk, J. Biochem. 250 (1975) 3243.

[2] T. Padro, C.M. Van der Hoogen, J.J. Emeis, Blood

Coagul. Fibrinolysis 4 (1993) 797.

[3] E.D. Peebles, H. Miller, C.R. Boyle, J.D. Brake, M.A.

Latour, Poult. Sci. 73 (1994) 1829.

[4] E. Peebles, E.H. Miller, C.R. Boyle, J.D. Brake, M.A.

Latour, J.P. Thaxton, Poult. Sci. 76 (1997) 236.

[5] M.B. Mattammal, V.M. Lakshmi, T.V. Zenser, B.B.

Davis, J. Pharm. Biomed. Anal. 8 (1990) 151.

[6] M. Bretner, T. Kulokowski, J.M. Dzik, M. Balinska, W.

Rode, D. Shugar, J. Med. Chem. 36 (1993) 3611.

[7] H.Y. Aboul-Enein, N.M. Al-is, J. Enzyme Inhib. 7 (1993)

197.

[8] V.J. Ram, A. Goel, M. Nath, P. Srivastava, Bioorg. Med.

Chem. Lett. 4 (1994) 2653.

[9] M. Inazumi, F. Kano, S. Sakata, Chem. Pharm. Bull. 40

(1992) 1808.

[10] N. Al-is, E. Khamis, A. Al-Mayouf, H. Aboul-Enein,

Corros. Prev. Control 42 (1995) 13.

[11] A.B. Tadros, El-Nabey, Polym. Mater. Sci. Eng. 74 (1996)

120.

[12] M. Kimura, M. Aizawa, Eur. Pat. Appl. EP 802194 A2222

(1997) 55.

[13] W. Saenger, in: C.R. Cantor (Ed.), Principles of nucleic

acid structure, Springer�/Verlag, New York, 1984, pp.

129�/158references cited therein..

[14] M.Y.W. Yu, J. Sedlak, R.H. Lindsay, Arch. Biochem.

Biophys. 155 (1973) 111.

[15] A. Palumbo, A. Napolitano, L. De Martino, W. Vieira,

V.J. Hearing, Biochim. Biophys. Acta 271 (1994) 1200.

[16] A. Napolitano, A. Palumbo, M. d’Ischia, G. Prote, J. Med.

Chem. 39 (1996) 5192.

[17] I. Roman, R. Giurgea, Z. Uray, Stud. Cercet Biochim. 35

(1992) 121.

[18] G.G. Skellern, C.D. Bates, D.G. Watson, R.J. Mairs, S.

Martin, Pharm. Sci. 1 (1995) 451.

[19] M. Gupta, M.N. Srivastave, Synth. React. Inorg. Met-

Org. Chem. 26 (1996) 305.

[20] M.S. Masoud, O.H.A. El-Hamid, Z.M. Zaki, Transition

Met. Chem. 19 (1994) 21.

[21] B.N. Singh, U. Singh, R. Ghose, A.K. Ghose, Asian J.

Chem. 5 (1993) 262.

[22] R.L. Sarma, J. Cryst. Growth 102 (1990) 542.

[23] P.K. Tripathi, M.M. Husain, K. Singh, Asian J. Phys. 3

(1994) 89.

[24] K. Singh, D.K. Rai, R. Rai, J. Mol. Struct. (Theochem.)

235 (1991) 15.

[25] K. Singh, D.K. Rai, J.S. Yadav, J. Mol. Struct. (Theo-

chem.) 231 (1991) 103.

[26] C. Parkanyi, C. Boniface, J.-J. Aron, M.D. Gaye, L. Von

Szentpaly, R. Ghose, K.S. Raghuveer, Struct. Chem. 3

(1992) 277.

[27] J. Marino, N. Russo, E. Sicilia, M. Toscamo, Int. J.

Quantum Chem. 82 (2001) 44.

[28] H. Yekeler, J. Comput. Aid Mol. Des. 149 (2000) 243.

[29] Y.V. Rubin, Y. Morozov, D. Venkateshwarlu, J. Lesz-

cynski, J. Phys. Chem. A 102 (1999) 9489.

[30] J. Sponer, J. Leszcynski, P. Hobza, J. Phys. Chem. A 101

(1997) 9489.

[31] J. Leszczynski, J. Phys. Chem. 96 (1992) 1649.

[32] A. Les, L. Adamowiez, J. Am. Chem. Soc. 112 (1990)

1504.

[33] A.R. Katritzky, M. Szafran, J. Stevens, J Chem Soc Perk

Trans II, (1989) 1507.

[34] H. Rostkowska, K. Szczepaniak, M.J. Nowak, J. Leszc-

zynski, K. KuBulat, W.B. Person, J. Am. Chem. Soc. 112

(1990) 2147.

[35] L. Lapinski, H. Rostkowska, M.J. Nowak, J.S. Kwiat-

kowski, J. Leszczynski, Vib. Spectrosc. 13 (1996) 23.

[36] M. Graindourze, T. Grootaers, J. Smets, T. Zeegers-

Huyskens, G. Maes, J. Mol. Struct. 237 (1990) 389.

[37] R.A. Yadav, P.N.S. Yadav, J.S. Yadav, Proc. Indian

Acad. Sci. (Chem. Sci.) 100 (1988) 69.

[38] J.S. Casas, E.E. Castellano, M.S. Garcıa-Tasende, A.

Sanchez, J. Sordo, E.M. Vazquez-Lopez, J. Zukerman-

Schpector, J. Chem. Soc. Dalton Trans. 9 (1996) 1973.

[39] J. Smets, M. Graindourze, T. Zeegers-Huyskens, G. Maes,

J. Mol. Struct. 318 (1994) 37.

[40] M. Graindourze, T. Grootaers, J. Smets, T. Zeegers-

Huyskens, G. Maes, J. Mol. Struct. 243 (1991) 37.

[41] C.B. Arora, R. Rastogi, U. Awasthi, Krishna, Asian J.

Phys. 5 (1996) 231.

[42] J.K. Gupta, S. Kumar, S. Kumar, S.D. Kaushik, Asian J.

Phys. 5 (1996) 245.

[43] W.J. Hehre, L. Radom, P.V.R. Schleyer, J.A. Pople, Ab

Initio Molecular Orbital Theory, Wiley, New York, USA,

1986.

[44] J.A. Pople, R. Krishnan, H.B. Schlegel, J.S. Binkley, Int. J.

Quant. Chem. Quant. Chem. Symp. 13 (1979) 225.

[45] J.M. Seminario, P. Politzer (Eds.), Modern Density Func-

tional Theory: a Tool for Chemistry, vol. 2, Elsevier,

Amsterdam, 1995.

[46] A.D. Becke, Phys. Rev. A38 (1988) 3098.

[47] A.D. Becke, J. Chem. Phys. 97 (1992) 9173.

[48] A.D. Becke, J. Chem. Phys. 98 (1993) 5648.

[49] C. Lee, W. Yang, R.G. Parr, Phys. Rev. B37 (1988) 785.

[50] J.P. Perdew, Phys. Rev. B33 (1986) 8822.

[51] J.P. Perdew, Phys. Rev. B34 (1986) 7406.

[52] M.J. Frisch, G.W. Trucks, H.B. Schlegel, P.M.W. Gill,

B.G. Johnson, M.A. Robb, J.R. Cheeseman, T. Keith,

G.A. Petersson, J.A. Montgomery, K. Raghavachari,

M.A. Al-Laham, V.G. Zakrzewski, J.V. Ortiz, J.B. Fores-

man, J. Cioslowski, B.B. Stefanov, A. Nanayakkara, M.

Challacombe, C.Y. Peng, P.Y. Ayala, W. Chen, M.W.

Wong, J.L. Res, E.S. Replogle, R. Gomperts, R.L. Martin,

D.J. Fox, J.S. Binkley, D.J. Defrees, J. Baker, J.P. Stewart,


M. Head-Gordon, C. Gonzalez, J.A. Pople, GAUSSIAN 94,

Gaussian, Inc, Pittsburgh PA, USA, 1999.

[53] G. Ferenczy, L. Harsanyi, B. Rozsondai, I. Hargittai, J.

Mol. Struct. 140 (1986) 71.

[54] L. Harsanyi, A. Csaszar, P. Csaszar, J. Mol. Struct.

(Theochem.) 137 (1986) 207.

[55] A.P. Scott, L. Radom, J. Phys. Chem. 100 (1996)

16502.

[56] M. Alcolea Palafox, J. Phys. Chem. (Submitted for

publication).

[57] A. Les, L. Adamowicz, M.J. Nowak, L. Lapinski, Spectro-

chim. Acta 48A (1992) 1385.

[58] N.P.G. Roeges, A Guide to the Complete Interpretation of

Infrared Spectra of Organic Structures, references cited

therein, Wiley, UK, 1994.

[59] M. Alcolea Palafox, F.J. Melendez, J. Mol. Struct.

(Theochem.) 459 (1999) 239.


Vibrational frequencies and structure of 2-thiouracil by Hartree-Fock, post-Hartree-Fock and density functional methods

Documents