- Email: [email protected]
Resonance-assisted hydrogen bonding in terms of substituent effect
Tetrahedron 65 (2009) 2010–2014
Contents lists available at ScienceDirect
Tetrahedron journal homepage: www.elsevier.com/locate/tet
Resonance-assisted hydrogen bonding in terms of substituent effect Tadeusz M. Krygowski *, Joanna E. Zachara-Horeglad Faculty of Chemistry, Warsaw University, Pasteura 1, 02-093 Warsaw, Poland
a r t i c l e i n f o
a b s t r a c t
Article history: Received 7 October 2008 Received in revised form 12 December 2008 Accepted 5 January 2009 Available online 8 January 2009
A quantitative description of resonance-assisted H-bond in terms of s-scale of substituent effect for proton-donating and proton-accepting groups is presented. Application of the proposed approach to malonaldehyde and ortho-hydroxybenzaldehyde shows that formation of H-bonding significantly changes the electronic properties of interacting groups expressed by s-values, which furthermore explains the increase of p-electron delocalization in such RAHB systems. The presented analyses are based on quantum-chemical modelling at B3LYP/6-311þG** level. Ó 2008 Published by Elsevier Ltd.
Dedicated to Professor Zofia Dega-Szafran and Professor Miros1aw Szafran on the occasion of their 75th birthday
Keywords: Hydrogen bond p-Electron delocalization RAHB Substituent effect
1. Introduction H-bonding belongs to the most important interactions since it is responsible for many chemical, physical and biological properties of various molecules and chemical systems.1 It also plays an important role in the development of new materials.1 The general scheme of H-bonded system(s) is as below:
R—A—H/B—R0
(1) 2
where A and B are the electronegative atoms, A–H is a protondonor whereas B a proton-acceptor. If R and R0 are bonded covalently to each other, forming one molecule, the H-bond is intramolecular. In the case when the proton-donating and protonaccepting groups are linked via p-conjugated bonds, an interplay between the H-bonding and p-electron delocalization in the linking chain is observed. For such cases, Gilli and co-workers3 introduced a concept of resonance-assisted hydrogen bonding (RAHB). Based on much experimental material, they assumed that there is a cooperative effect between H-bond strengthening and pelectron delocalization of conjugated bonds linking the protondonating and proton-accepting groups. This phenomenon was noticed for the first time in the sixties,4 and was later also
* Corresponding author. Tel./fax: þ48 22 8222892. E-mail address: [email protected] (T.M. Krygowski). 0040-4020/$ – see front matter Ó 2008 Published by Elsevier Ltd. doi:10.1016/j.tet.2009.01.006
supported by Scheiner5 and may be found in the NBO-type approaches.6 Although the idea of RAHB has been recently criticized,7 it serves as a useful qualitative concept and is widely studied.8 The electronic properties of A–H and B groups involved in intramolecular H-bonding changes in comparison with their states before interaction.9 For instance, the consequences of H-bond formation may be observed in energy changes between the H-bonded system and its ‘open’ conformation10 (see Scheme 1). It should be however emphasized that such difference in energy consists of at least three contributions: H-bond formation and changes in pelectron delocalization, which stabilizes the system in question, and deformation energy, which destabilizes it.11,12
Scheme 1.
Since the interacting groups, A–H and B, may be considered as substituents, their electronic properties may be described by the Hammett substituent constants s,13–17 which express the degree of electron-donating or electron-accepting (depending on the nature of substituent in general) power of the substituent, and in general, the higher is the absolute value, the stronger the donating/ attracting properties are. In the origins of the substituent constant concept, the electronic properties of substituents were determined
T.M. Krygowski, J.E. Zachara-Horeglad / Tetrahedron 65 (2009) 2010–2014
by changes in rate or equilibrium constants of specially designed reactions when the main reagent became substituted.18 Since the nature of the interaction between the substituent and the reaction site depends among other things on the reaction type, a variety of substituent constants are in use.16,19 For instance, sþ constants are designed for p-donors interacting with an electron-accepting reaction site, whereas s for p-acceptors interacting with an electron-donating reaction site. In general, in the RAHB systems, the p-proton-donating group A–H is an electron donor when considered as a substituent and is characterized by negative s values. Analogously, the protonaccepting group B is a p-electron-accepting substituent with positive s values. In consequence, in the H–A–R00 –B system, which may form an intramolecular H-bond, two terminal groups (substituents) A–H and B may interact via a covalent chain of bonds since they have opposite electronic properties. As the result of engagement in H-bonding, the A–H bond length should elongate and consequently its electron-donating properties as substituent should become stronger. Simultaneously, the proton approaches B, leading to the formation of B/Hdþ and as a consequence, the electron-attracting power of B in B/Hdþ should increase in comparison with B itself. Thus the magnitude of the substituent effect should consequently increase.20 The aim of our study is to show the changes in electronic properties of A–H and B groups in RAHB systems, caused by engagement in H-bonding interaction Ad–Hdþ/B, in a quantitative way, i.e., by the use of Hammett substituent constants. 2. Results and discussion 2.1. The concept First of all, it is necessary to find a way to model a relationship between the strength of H-bond associated with the position of proton in the A–H/B bridge and the electron-donating power of Ad–Hdþ and electron-accepting power of –B/Hdþ expressed by the use of the sþ and s constants, respectively. In order to describe the variable H-bonding, we apply a computational model of varying H-bond strength, already presented in the literature.11,21 This approach is based on A–H/F and B/H–F interactions for which the A/F and B/F distances are controlled and varied in order to simulate the changes of H-bonding strength. For determination of substituent constant values of Ad–Hdþ and –B/Hdþ, the method for estimation of s from exocyclically substituted derivatives of fulvene and heptafulvene, respectively, may be employed.22 This approach allows for determination of s values from p-electron delocalization in the fulvene/heptafulvene rings, which in fact is sensitive to substitution at the exocyclic carbon atom (Scheme 2). p-Electron-donating substituents increase p-electron delocalization in the fulvene ring and p-electron-accepting substituents in the heptafulvene ring. Such effects are associated with the general tendency to posses 6 p-electrons in the ring. The appropriate empirical equations where p-electron delocalization is expressed by HOMA23 index (Harmonic Oscillator Model of Aromaticity, for definition see Methodology) are presented below:22
2011
sþ p ¼ 2:401,HOMAðfulveneÞ 0:736
(2)
s p ¼ 4:208,HOMAðheptafulveneÞ 0:655
(3)
In both cases the substituent constants, sþ and s, are used for the para-position, since for this position the description of substituent interactions is most similar to that observed in cases of push–pull interactions discussed in this paper. Incorporation of both above-mentioned models allows for the estimation of s values for A–H and B groups involved in H-bonding of different strength. 2.2. Calculation of sD for proton-donating substituent Let us now consider the most common proton-donating group, which is the hydroxyl group –OH. Application of the two abovementioned approaches may be schematically presented as in Scheme 3. The experimental value of the substituent constant for the –OH group, sþ¼0.92, refers to the case of infinite distance d(O/F). If, after the proton transfer, the HF molecule is infinitely far, the other limiting value of substituent constant is sþ for the –O group, and it is equal to 2.30.16 Undoubtedly the values of sþ for intermediate situations, when the proton is between O/F, have to be between these two above-mentioned limiting sþ values. For all fixed interatomic O/F distances, optimization leads to an optimal position of proton in the O–H/F region. Thus the length of O–H and C–O bonds as well as C–C bonds in the ring are known and allow one to calculate the HOMA value for the ring, which in turn serves for estimation of sþ value from Eq. 2. Then one can correlate the C–O bond length known as a numerical measure of H-bond strength in phenols11,24 and the sþ value as presented in Figure 1, and approximate the relation by linear regression (Eq. 4).
sþ ðOHÞ ¼ 15:50,dC—O 21:88 cc ¼ 0:99
(4)
As a result, the obtained relation (Eq. 4) may be used to estimate the sþ value of any –OH group involved in H-bonding by considering the neighbouring C–O bond length. 2.3. Calculation of sL for proton-accepting substituent The case where the proton in the H-bond is approaching the proton-accepting group may be treated in a similar way as described above. In order to present a general idea the protonaccepting group –CHO is chosen and thus we consider 8-formylheptafulvene interacting with HF (Scheme 4). Again, H-bonding interaction leads to changes of p-electron delocalization in the ring, which can be measured by the HOMA index. Then from Eq. 3, assigned for electron-accepting substituents, the appropriate s values for –CHO/HF substituent of fixed O/F distance may be estimated. The resultant s values may be correlated with C–O bond length of the –CHO group involved in H-bonding, as presented in Figure 2.
s ðCHOÞ ¼ 25:73,dC—O 30:54 cc ¼ 0:99
(5)
The dependence of s values on C–O bond length in the formyl group is nearly linear and may be applied to estimate the value of s for any formyl group interacting via H-bond.
Scheme 2. Stabilization of fulvene by electron-donating substituent, D, and heptafulvene by electron-accepting substituent, A.
Scheme 3.
2012
T.M. Krygowski, J.E. Zachara-Horeglad / Tetrahedron 65 (2009) 2010–2014
Figure 1. Interrelation between sþ(OH) and the C–O bond length in 6-hydroxyfulvene involved in H-bonding.
Figure 2. Interrelation between s(CHO) and the C–O bond length in 8-formylheptafulvene involved in H-bonding.
Application of both approaches described above allows one to estimate the electron properties of both proton-donating and proton-accepting substituents involved in H-bonding in RAHB systems. Two examples are presented below.
C–O bond length in the electron-donating –C–OH fragment of the H-bonded molecule of malonaldehyde (1.319 Å) it is possible to estimate the sþ(OH) value by the use of Eq. 4 and one obtains the value 1.44, which may be compared with the value of 1.08 for the open conformation. The electron-donating power (measured by sþ values) of –OH caused by engagement in H-bonding increases by 0.36 s units. In order to estimate the s value for the other substituent in malonaldehyde, the formyl group, the regression 5 is considered. This results in a s value equal to 0.75 for the formyl group in the open conformation and 1.31 for H-bonded conformation. The difference is 0.56 units of s. This results from the approach of the proton to the carbonyl group increasing its electron-accepting power as a substituent. It is well known that the substituent effect in disubstituted pelectron systems increases with an increase in the difference between electron-accepting and electron-donating power of the interacting substituents. In the case of malonaldehyde in the open conformation, the difference is 1.83 units of s, whereas for the closed conformation the difference is 2.75 units of s. The increase of the substituent effect due to H-bond formation is 0.92 units of s and is due to the increase of both electron-donating power of the –OH group and electron-accepting power of –CHO caused by a partial shift of the proton from the –OH group towards the oxygen atom of
2.4. Application to RAHB systems In order to exemplify in detail the methodology and the concept presented in this paper we have chosen two RAHB molecules with –OH and –CHO groups, malonaldehyde and o-hydroxybenzaldehyde. In both cases two conformations are considereddthe H-bonded conformation and the ‘open’ one (Scheme 1). Let us consider firstly the malonaldehyde molecule (Fig. 3). The HOMA value for the linker, i.e., the OCCCO fragment, for the open conformer is equal to 0.18, whereas for the closed one it equals 0.66. The first value corresponds to the situation where there are essentially no changes in substituent effects of the –OH and –CHO groups and thus on p-electron structure of the OCCCO fragment since there is no H-bond. The change in HOMA between the open and closed conformations, 0.46 units of HOMA, is undoubtedly caused by H-bond formation. Application of the Gilli25 parameter describing the degree of delocalization Del% (for definition see Methodology) in the linker presents a similar picture: for the open conformation Del%¼4% whereas for the closed one, Del%¼39%. Energetics of these two conformers are also distinctly different: the H-bonded conformer is more stable by 7.84 kcal/mol than the open one. Undoubtedly H-bond formation increases the stability and p-electron delocalization in the linker. The analysis of geometric parameters also leads to an important observation. The hydroxyl O–H bond length in the open conformer is shorter by 0.032 Å than in the closed one, whereas the C–O bond length next to the hydroxyl group is longer by 0.023 Å. From that
Scheme 4.
Figure 3. Optimized geometry of two conformations of malonaldehyde along with calculated values of HOMA, Del% (for definition see Methodology), and substituent constants.
T.M. Krygowski, J.E. Zachara-Horeglad / Tetrahedron 65 (2009) 2010–2014
the formyl group. As a result of those increases, p-electron delocalization in the spacer increases: HOMA for the OCCCO fragment increases by 0.46 unit of HOMA and Gilli’s Del% by 35%. For HOMA this magnitude is almost a half of the scale between the aromatic benzene with HOMA¼1.0 and non-aromatic Kekule´ structure of benzene with HOMA¼0.0. The same procedure may now be employed for ortho-hydroxybenzaldehyde, Figure 4 illustrates all the necessary data for such system. The HOMA values for the geometry of the linker, i.e., the OCCCO fragment of the molecule for the open conformer is equal to 0.02, whereas for the closed one it is 0.38. The Gilli Del% for the open conformer is 27% whereas for the closed one it equals 49%. If there were no changes in substituent effects due to H-bond formation, both parameters should be nearly the same for open and closed conformers. The observed changes in HOMA and Del% are undoubtedly due to H-bond formation. The energetics of these two conformers are also distinctly different: the H-bonded conformer is more stable by 10.4 kcal/mol than the open one. The analysis of geometry parameters also leads to an important observation. Again the O–H bond length in the open conformer is shorter by 0.020 Å, whereas the C–O bond length next to the hydroxyl group is longer by 0.014 Å. It should be mentioned here that in the open conformation the O–H bond length is almost the same as for phenol (0.964 and 0.963 Å, respectively) and the C]O bond length in the formyl group is almost the same as in benzaldehyde (1.209 and 1.211 Å, respectively). From the value of the C–O bond length in the –COH fragment of H-bonded molecule of o-hydroxybenzaldehyde (1.340 Å) it is possible to estimate the sþ(OH) value by the use of Eq. 4 and it equals 1.11, whereas for the open conformation it amounts to 0.89. The sþ value for the ‘free’ –OH group from experimental estimation is equal to 0.92,16 thus the substituent constant for –OH group in open conformation is nearly the same as for the ‘free’ group estimated experimentally. The change in the electron-donating power of –OH involved in H-bonding measured by the changes in sþ values amounts to about 0.2 units of s. To estimate the s value for the formyl group in o-hydroxybenzaldehyde, regression 5 is applied. The estimated values are 0.57 for the open conformation and 1.06 for the closed one. The difference is 0.49 and it results from the approach of the proton to oxygen atom of the formyl group increasing its electron-accepting power as a substituent. In summary, the difference between the substituent constants sþ(OH) and s(CHO) for the open conformation is 1.46, leading to
Figure 4. Optimized geometry of closed and open conformations of o-hydroxybenzaldehyde. HOMA and Del% values were calculated for OCCCO fragments. Substituent constants were calculated from Eqs. 4 and 5.
2013
very low delocalization estimated by HOMA¼0.02 and Gilli’s Del%¼27%. For the closed conformation the corresponding difference equals 2.17, which means an increase by 0.71 units of s. It is associated with a much stronger interaction between those substituents, revealed in increase of p-electron delocalization for the OCCCO fragment: DHOMA¼0.4 and DDel%¼22%. 3. Conclusions Our results present the first quantitative description of changes in the electronic properties of proton-donating and protonaccepting groups due to H-bonding. In the cases presented above, H-bond formation led to an increase in electron-donating power of the proton-donating substituent A–H (in both cases discussed A– H¼O–H) and electron-accepting power of proton-accepting substituent B (in both cases discussed B¼–CHO). Schematically this may be illustrated by Figure 5. In consequence, the difference in electron properties increases, which serves as a reliable explanation for increase of p-electron delocalization in the conjugated linker described in the literature as RAHB. The presented results show a cooperative effect of both groups, A–H and B, acting in two ways: (i) by means of H-bond formation, and (ii) by mesomeric effect between two substituents via the pelectron spacer OCCCO, which is undoubtedly the consequence of (i). This in turn causes an increase of p-electron delocalization in the linker. Obviously, both formally separated steps occur simultaneously. 4. Methodology Geometries of the molecules and complexes were optimized at B3LYP/6-311þG** level using Gaussian 03.26 In the H-bonded complexes the fluoride approached the 6-hydroxyfulvene molecule along the line of prolongation of the O–H bond direction. The O/F distance was controlled and varied from 4.000 Å to the distance when the proton transfer from 6-hydroxyfulvene to F was observed (i.e., 2.638 Å). The 8-formylheptafulvene complexes with HF were optimized for the same values of O/F distances as 6hydroxyfulvene with F, since for those complexes proton transfer were not observed. Moreover, linearity of O/H/F was assumed.
Figure 5. Schematic representation of changes in electronic properties of protondonor and proton-acceptor interacting via H-bond.
2014
T.M. Krygowski, J.E. Zachara-Horeglad / Tetrahedron 65 (2009) 2010–2014
HOMA values were calculated according to the following equation:23
HOMA ¼ 1
n 2 1X a R Rj n j ¼ 1 i opt;i
(8)
where n represents the total number of bonds taken into summation; ai is a normalization constant (for C–C bonds aC–C¼257.7, for C–O bonds aC–O¼157.38) fixed to give HOMA¼0 for a model nonaromatic system, e.g., Kekule´ structure of benzene and HOMA¼1 for the system with all bonds equal to the optimal value Ropt,i assumed to be realized for fully aromatic systems (Ropt,C–C¼1.388 Å, Ropt,C–O¼1.265 Å). Del% values were calculated according to the following equations:25
Del% ¼ 100,ð1 j2l 1jÞ
l ¼ 1=4,½ð2 n1Þ þ ðn2 1Þ þ ð2 n3Þ þ ðn4 1Þ
(9)
(10)
where n1, n2, n3 and n4 are the bond numbers of corresponding double and single bonds calculated from Pauling equation27 d(1)– d(n)¼c ln(n), where d(1)C–C¼1.467, d(2)C–C¼1.349, d(1)C–O¼1.367, d(2)C–O¼1.217.23 Acknowledgements We thank the Interdisciplinary Centre for Mathematical and Computational Modelling ICM (Warsaw, Poland) for providing computer time and programs. J.E.Z.-H. gratefully acknowledges financial support under grant N204 093 31/2144 and a Scholarship from the Foundation for Polish Science (FNP). References and notes 1. (a) Intermolecular Interactions: From van der Waals to Strongly Bound Complexes; Scheiner, S., Ed.; Wiley: New York, NY, 1997; (b) Desiraju, G. R.; Steiner, T. The Weak Hydrogen Bond in Structural Chemistry and Biology; Oxford University Press: New York, NY, 1999; (c) Gerlt, J. A.; Kreevoy, M. M.; Cleland, W. W.; Frey, P. A. Chem. Biol. 1997, 4, 259–267; (d) Perrin, C. L.; Nielson, J. B. Annu. Rev. Phys. Chem. 1997, 48, 511–544; (e) Desiraju, G. R. Acc. Chem. Res. 2002, 35, 565–573; (f) Meyer, E. A.; Castellano, R. K.; Diederich, F. Angew. Chem., Int. Ed. 2003, 42, 1210–1250; (g) Sobczyk, L.; Grabowski, S. J.; Krygowski, T. M. Chem. Rev. 2005, 105, 3513–3560; (h) Krygowski, T. M.; Szaty1owicz, H. Trends Org. Chem. 2006, 11, 37–53. 2. (a) Pauling, L. The Nature of the Chemical Bond; Cornell University Press: New York, NY, 1960; (b) Pure Appl. Chem. 1999, 71, 1919–1981. 3. (a) Gilli, G.; Belluci, F.; Ferretti, V.; Bertolasi, V. J. Am. Chem. Soc. 1989, 111, 1023– 1028; (b) Gilli, G.; Gilli, P. J. Mol. Struct. 2000, 552, 1–15.
4. (a) Williams, D. E.; Dunke, W. L.; Rundle, R. E. Acta Crystallogr. 1962, 15, 627– 635; (b) Ogoshi, H.; Yoshida, Z. J. Chem. Soc. D: Chem. Commun. 1970, 176– 177. 5. Scheiner, S. Hydrogen Bonding. A Theoretical Perspective; Oxford University Press: Oxford, 1997. 6. (a) Glendening, E. D.; Badenhoop, J. K.; Weinhold, F. J. Comput. Chem. 1998, 19, 628–646; (b) Weinhold, F.; Landis, C. R. Valency and Bonding; Cambridge University Press: Cambridge, 2005. ˜ ez, M.; Del Bene, J. E. Mol. Phys. 2004, 7. (a) Alkorta, I.; Elguero, J.; Mo´, O.; Ya´n ˜ ez, M.; Del Bene, J. E. 102, 2563–2574; (b) Alkorta, I.; Elguero, J.; Mo´, O.; Ya´n ˜ ez, M.; Elguero, J. Chem. Phys. Lett. 2002, 411, 411–415; (c) Sanz, P.; Mo´, O.; Ya´n ˜ ez, M.; Elguero, J. Chem.dEur. J. 2008, 14, 4225–4232; (d) Sanz, P.; Mo´, O.; Ya´n ChemPhysChem 2007, 8, 1950–1958. 8. Hydrogen BondingdNew Insights; Grabowski, S. J., Ed.Series: Challenges and Advances in Computational Chemistry and Physics; Springer: Dordrecht, 2006; Vol. 3. 9. Grabowski, S. J. J. Phys. Org. Chem. 2003, 16, 797–802. 10. (a) Grabowski, S. J. J. Mol. Struct. 2001, 562, 137–143; (b) Cuma, M.; Scheiner, S.; Kar, T. J. Mol. Struct. (THEOCHEM) 1999, 467, 37–49. 11. Krygowski, T. M.; Zachara, J. E.; Szaty1owicz, H. J. Phys. Org. Chem. 2005, 18, 110–114. 12. Grabowski, S. J. J. Phys. Org. Chem. 2004, 17, 18–31. 13. Hammett, L. P. Physical Organic Chemistry; McGraw-Hill: New York, NY, 1940. 14. Exner, O. In Advances in Linear Free Energy Relationships; Chapman, N. B., Shorter, J., Eds.; Plenum: London, 1972; pp 1–69. 15. (a) Charton, M. Prog. Phys. Org. Chem. 1973, 10, 81–204; (b) Charton, M. Prog. Phys. Org. Chem. 1981, 13, 119–251. 16. Hansch, C.; Leo, A.; Taft, W. R. Chem. Rev. 1991, 91, 165–195. 17. Krygowski, T. M.; St˛epien´, B. T. Chem. Rev. 2005, 105, 3482–3512. 18. Jaffe, H. H. Chem. Rev. 1953, 53, 191–261. 19. (a) Shorter, J. Pure Appl. Chem. 1994, 66, 2451--2468; (b) Shorter, J. Pure Appl. Chem. 1997, 69, 2497–2510. 20. Krygowski, T. M.; Zachara-Horeglad, J. E.; Palusiak, M.; Pelloni, S.; Lazzeretti, P. J. Org. Chem. 2008, 73, 2138–2145. 21. Krygowski, T. M.; Zachara, J.; Szaty1owicz, H. J. Org. Chem. 2004, 69, 7038–7043. ´ , B. T.; Cyran´ski, M. K.; Krygowski, T. M. Chem. Phys. Lett. 2001, 350, 22. Stepien ˛ 537–542. 23. Krygowski, T. M. J. Chem. Inf. Comput. Sci. 1993, 33, 70–78. 24. (a) Krygowski, T. M.; Szaty1owicz, H.; Zachara, J. E. J. Chem. Inf. Model. 2005, 45, 652–656; (b) Krygowski, T. M.; Szaty1owicz, H.; Zachara, J. E. J. Chem. Inf. Comput. Sci. 2004, 44, 2077–2082. 25. Bertolasi, V.; Pretto, L.; Ferretti, V.; Gilli, P.; Gilli, G. Acta Crystallogr., Sect. B 2006, 62, 1112–1120. 26. Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.; Montgomery, J. A., Jr.; Vreven, T.; Kudin, K. N.; Burant, J. C.; Millam, J. M.; Iyengar, S. S.; Tomasi, J.; Barone, V.; Mennucci, B.; Cossi, M.; Scalmani, G.; Rega, N.; Petersson, G. A.; Nakatsuji, H.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.; Nakai, H.; Klene, M.; Li, X.; Knox, J. E.; Hratchian, H. P.; Cross, J. B.; Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev, O.; Austin, A. J.; Cammi, R.; Pomelli, C.; Ochterski, J. W.; Ayala, P. Y.; Morokuma, K.; Voth, G. A.; Salvador, P.; Dannenberg, J. J.; Zakrzewski, V. G.; Dapprich, S.; Daniels, A. D.; Strain, M. C.; Farkas, O.; Malick, D. K.; Rabuck, A. D.; Raghavachari, K.; Foresman, J. B.; Ortiz, J. V.; Cui, Q.; Baboul, A. G.; Clifford, S.; Cioslowski, J.; Stefanov, B. B.; Liu, G.; Liashenko, A.; Piskorz, P.; Komaromi, I.; Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham, M. A.; Peng, C. Y.; Nanayakkara, A.; Challacombe, M.; Gill, P. M. W.; Johnson, B.; Chen, W.; Wong, M. W.; Gonzalez, C.; Pople, J. A. Gaussian 03, Revision C.02; Gaussian: Wallingford, CT, 2004. 27. Pauling, L. J. Am. Chem. Soc. 1947, 69, 542–553.
Contents lists available at ScienceDirect
Tetrahedron journal homepage: www.elsevier.com/locate/tet
Resonance-assisted hydrogen bonding in terms of substituent effect Tadeusz M. Krygowski *, Joanna E. Zachara-Horeglad Faculty of Chemistry, Warsaw University, Pasteura 1, 02-093 Warsaw, Poland
a r t i c l e i n f o
a b s t r a c t
Article history: Received 7 October 2008 Received in revised form 12 December 2008 Accepted 5 January 2009 Available online 8 January 2009
A quantitative description of resonance-assisted H-bond in terms of s-scale of substituent effect for proton-donating and proton-accepting groups is presented. Application of the proposed approach to malonaldehyde and ortho-hydroxybenzaldehyde shows that formation of H-bonding significantly changes the electronic properties of interacting groups expressed by s-values, which furthermore explains the increase of p-electron delocalization in such RAHB systems. The presented analyses are based on quantum-chemical modelling at B3LYP/6-311þG** level. Ó 2008 Published by Elsevier Ltd.
Dedicated to Professor Zofia Dega-Szafran and Professor Miros1aw Szafran on the occasion of their 75th birthday
Keywords: Hydrogen bond p-Electron delocalization RAHB Substituent effect
1. Introduction H-bonding belongs to the most important interactions since it is responsible for many chemical, physical and biological properties of various molecules and chemical systems.1 It also plays an important role in the development of new materials.1 The general scheme of H-bonded system(s) is as below:
R—A—H/B—R0
(1) 2
where A and B are the electronegative atoms, A–H is a protondonor whereas B a proton-acceptor. If R and R0 are bonded covalently to each other, forming one molecule, the H-bond is intramolecular. In the case when the proton-donating and protonaccepting groups are linked via p-conjugated bonds, an interplay between the H-bonding and p-electron delocalization in the linking chain is observed. For such cases, Gilli and co-workers3 introduced a concept of resonance-assisted hydrogen bonding (RAHB). Based on much experimental material, they assumed that there is a cooperative effect between H-bond strengthening and pelectron delocalization of conjugated bonds linking the protondonating and proton-accepting groups. This phenomenon was noticed for the first time in the sixties,4 and was later also
* Corresponding author. Tel./fax: þ48 22 8222892. E-mail address: [email protected] (T.M. Krygowski). 0040-4020/$ – see front matter Ó 2008 Published by Elsevier Ltd. doi:10.1016/j.tet.2009.01.006
supported by Scheiner5 and may be found in the NBO-type approaches.6 Although the idea of RAHB has been recently criticized,7 it serves as a useful qualitative concept and is widely studied.8 The electronic properties of A–H and B groups involved in intramolecular H-bonding changes in comparison with their states before interaction.9 For instance, the consequences of H-bond formation may be observed in energy changes between the H-bonded system and its ‘open’ conformation10 (see Scheme 1). It should be however emphasized that such difference in energy consists of at least three contributions: H-bond formation and changes in pelectron delocalization, which stabilizes the system in question, and deformation energy, which destabilizes it.11,12
Scheme 1.
Since the interacting groups, A–H and B, may be considered as substituents, their electronic properties may be described by the Hammett substituent constants s,13–17 which express the degree of electron-donating or electron-accepting (depending on the nature of substituent in general) power of the substituent, and in general, the higher is the absolute value, the stronger the donating/ attracting properties are. In the origins of the substituent constant concept, the electronic properties of substituents were determined
T.M. Krygowski, J.E. Zachara-Horeglad / Tetrahedron 65 (2009) 2010–2014
by changes in rate or equilibrium constants of specially designed reactions when the main reagent became substituted.18 Since the nature of the interaction between the substituent and the reaction site depends among other things on the reaction type, a variety of substituent constants are in use.16,19 For instance, sþ constants are designed for p-donors interacting with an electron-accepting reaction site, whereas s for p-acceptors interacting with an electron-donating reaction site. In general, in the RAHB systems, the p-proton-donating group A–H is an electron donor when considered as a substituent and is characterized by negative s values. Analogously, the protonaccepting group B is a p-electron-accepting substituent with positive s values. In consequence, in the H–A–R00 –B system, which may form an intramolecular H-bond, two terminal groups (substituents) A–H and B may interact via a covalent chain of bonds since they have opposite electronic properties. As the result of engagement in H-bonding, the A–H bond length should elongate and consequently its electron-donating properties as substituent should become stronger. Simultaneously, the proton approaches B, leading to the formation of B/Hdþ and as a consequence, the electron-attracting power of B in B/Hdþ should increase in comparison with B itself. Thus the magnitude of the substituent effect should consequently increase.20 The aim of our study is to show the changes in electronic properties of A–H and B groups in RAHB systems, caused by engagement in H-bonding interaction Ad–Hdþ/B, in a quantitative way, i.e., by the use of Hammett substituent constants. 2. Results and discussion 2.1. The concept First of all, it is necessary to find a way to model a relationship between the strength of H-bond associated with the position of proton in the A–H/B bridge and the electron-donating power of Ad–Hdþ and electron-accepting power of –B/Hdþ expressed by the use of the sþ and s constants, respectively. In order to describe the variable H-bonding, we apply a computational model of varying H-bond strength, already presented in the literature.11,21 This approach is based on A–H/F and B/H–F interactions for which the A/F and B/F distances are controlled and varied in order to simulate the changes of H-bonding strength. For determination of substituent constant values of Ad–Hdþ and –B/Hdþ, the method for estimation of s from exocyclically substituted derivatives of fulvene and heptafulvene, respectively, may be employed.22 This approach allows for determination of s values from p-electron delocalization in the fulvene/heptafulvene rings, which in fact is sensitive to substitution at the exocyclic carbon atom (Scheme 2). p-Electron-donating substituents increase p-electron delocalization in the fulvene ring and p-electron-accepting substituents in the heptafulvene ring. Such effects are associated with the general tendency to posses 6 p-electrons in the ring. The appropriate empirical equations where p-electron delocalization is expressed by HOMA23 index (Harmonic Oscillator Model of Aromaticity, for definition see Methodology) are presented below:22
2011
sþ p ¼ 2:401,HOMAðfulveneÞ 0:736
(2)
s p ¼ 4:208,HOMAðheptafulveneÞ 0:655
(3)
In both cases the substituent constants, sþ and s, are used for the para-position, since for this position the description of substituent interactions is most similar to that observed in cases of push–pull interactions discussed in this paper. Incorporation of both above-mentioned models allows for the estimation of s values for A–H and B groups involved in H-bonding of different strength. 2.2. Calculation of sD for proton-donating substituent Let us now consider the most common proton-donating group, which is the hydroxyl group –OH. Application of the two abovementioned approaches may be schematically presented as in Scheme 3. The experimental value of the substituent constant for the –OH group, sþ¼0.92, refers to the case of infinite distance d(O/F). If, after the proton transfer, the HF molecule is infinitely far, the other limiting value of substituent constant is sþ for the –O group, and it is equal to 2.30.16 Undoubtedly the values of sþ for intermediate situations, when the proton is between O/F, have to be between these two above-mentioned limiting sþ values. For all fixed interatomic O/F distances, optimization leads to an optimal position of proton in the O–H/F region. Thus the length of O–H and C–O bonds as well as C–C bonds in the ring are known and allow one to calculate the HOMA value for the ring, which in turn serves for estimation of sþ value from Eq. 2. Then one can correlate the C–O bond length known as a numerical measure of H-bond strength in phenols11,24 and the sþ value as presented in Figure 1, and approximate the relation by linear regression (Eq. 4).
sþ ðOHÞ ¼ 15:50,dC—O 21:88 cc ¼ 0:99
(4)
As a result, the obtained relation (Eq. 4) may be used to estimate the sþ value of any –OH group involved in H-bonding by considering the neighbouring C–O bond length. 2.3. Calculation of sL for proton-accepting substituent The case where the proton in the H-bond is approaching the proton-accepting group may be treated in a similar way as described above. In order to present a general idea the protonaccepting group –CHO is chosen and thus we consider 8-formylheptafulvene interacting with HF (Scheme 4). Again, H-bonding interaction leads to changes of p-electron delocalization in the ring, which can be measured by the HOMA index. Then from Eq. 3, assigned for electron-accepting substituents, the appropriate s values for –CHO/HF substituent of fixed O/F distance may be estimated. The resultant s values may be correlated with C–O bond length of the –CHO group involved in H-bonding, as presented in Figure 2.
s ðCHOÞ ¼ 25:73,dC—O 30:54 cc ¼ 0:99
(5)
The dependence of s values on C–O bond length in the formyl group is nearly linear and may be applied to estimate the value of s for any formyl group interacting via H-bond.
Scheme 2. Stabilization of fulvene by electron-donating substituent, D, and heptafulvene by electron-accepting substituent, A.
Scheme 3.
2012
T.M. Krygowski, J.E. Zachara-Horeglad / Tetrahedron 65 (2009) 2010–2014
Figure 1. Interrelation between sþ(OH) and the C–O bond length in 6-hydroxyfulvene involved in H-bonding.
Figure 2. Interrelation between s(CHO) and the C–O bond length in 8-formylheptafulvene involved in H-bonding.
Application of both approaches described above allows one to estimate the electron properties of both proton-donating and proton-accepting substituents involved in H-bonding in RAHB systems. Two examples are presented below.
C–O bond length in the electron-donating –C–OH fragment of the H-bonded molecule of malonaldehyde (1.319 Å) it is possible to estimate the sþ(OH) value by the use of Eq. 4 and one obtains the value 1.44, which may be compared with the value of 1.08 for the open conformation. The electron-donating power (measured by sþ values) of –OH caused by engagement in H-bonding increases by 0.36 s units. In order to estimate the s value for the other substituent in malonaldehyde, the formyl group, the regression 5 is considered. This results in a s value equal to 0.75 for the formyl group in the open conformation and 1.31 for H-bonded conformation. The difference is 0.56 units of s. This results from the approach of the proton to the carbonyl group increasing its electron-accepting power as a substituent. It is well known that the substituent effect in disubstituted pelectron systems increases with an increase in the difference between electron-accepting and electron-donating power of the interacting substituents. In the case of malonaldehyde in the open conformation, the difference is 1.83 units of s, whereas for the closed conformation the difference is 2.75 units of s. The increase of the substituent effect due to H-bond formation is 0.92 units of s and is due to the increase of both electron-donating power of the –OH group and electron-accepting power of –CHO caused by a partial shift of the proton from the –OH group towards the oxygen atom of
2.4. Application to RAHB systems In order to exemplify in detail the methodology and the concept presented in this paper we have chosen two RAHB molecules with –OH and –CHO groups, malonaldehyde and o-hydroxybenzaldehyde. In both cases two conformations are considereddthe H-bonded conformation and the ‘open’ one (Scheme 1). Let us consider firstly the malonaldehyde molecule (Fig. 3). The HOMA value for the linker, i.e., the OCCCO fragment, for the open conformer is equal to 0.18, whereas for the closed one it equals 0.66. The first value corresponds to the situation where there are essentially no changes in substituent effects of the –OH and –CHO groups and thus on p-electron structure of the OCCCO fragment since there is no H-bond. The change in HOMA between the open and closed conformations, 0.46 units of HOMA, is undoubtedly caused by H-bond formation. Application of the Gilli25 parameter describing the degree of delocalization Del% (for definition see Methodology) in the linker presents a similar picture: for the open conformation Del%¼4% whereas for the closed one, Del%¼39%. Energetics of these two conformers are also distinctly different: the H-bonded conformer is more stable by 7.84 kcal/mol than the open one. Undoubtedly H-bond formation increases the stability and p-electron delocalization in the linker. The analysis of geometric parameters also leads to an important observation. The hydroxyl O–H bond length in the open conformer is shorter by 0.032 Å than in the closed one, whereas the C–O bond length next to the hydroxyl group is longer by 0.023 Å. From that
Scheme 4.
Figure 3. Optimized geometry of two conformations of malonaldehyde along with calculated values of HOMA, Del% (for definition see Methodology), and substituent constants.
T.M. Krygowski, J.E. Zachara-Horeglad / Tetrahedron 65 (2009) 2010–2014
the formyl group. As a result of those increases, p-electron delocalization in the spacer increases: HOMA for the OCCCO fragment increases by 0.46 unit of HOMA and Gilli’s Del% by 35%. For HOMA this magnitude is almost a half of the scale between the aromatic benzene with HOMA¼1.0 and non-aromatic Kekule´ structure of benzene with HOMA¼0.0. The same procedure may now be employed for ortho-hydroxybenzaldehyde, Figure 4 illustrates all the necessary data for such system. The HOMA values for the geometry of the linker, i.e., the OCCCO fragment of the molecule for the open conformer is equal to 0.02, whereas for the closed one it is 0.38. The Gilli Del% for the open conformer is 27% whereas for the closed one it equals 49%. If there were no changes in substituent effects due to H-bond formation, both parameters should be nearly the same for open and closed conformers. The observed changes in HOMA and Del% are undoubtedly due to H-bond formation. The energetics of these two conformers are also distinctly different: the H-bonded conformer is more stable by 10.4 kcal/mol than the open one. The analysis of geometry parameters also leads to an important observation. Again the O–H bond length in the open conformer is shorter by 0.020 Å, whereas the C–O bond length next to the hydroxyl group is longer by 0.014 Å. It should be mentioned here that in the open conformation the O–H bond length is almost the same as for phenol (0.964 and 0.963 Å, respectively) and the C]O bond length in the formyl group is almost the same as in benzaldehyde (1.209 and 1.211 Å, respectively). From the value of the C–O bond length in the –COH fragment of H-bonded molecule of o-hydroxybenzaldehyde (1.340 Å) it is possible to estimate the sþ(OH) value by the use of Eq. 4 and it equals 1.11, whereas for the open conformation it amounts to 0.89. The sþ value for the ‘free’ –OH group from experimental estimation is equal to 0.92,16 thus the substituent constant for –OH group in open conformation is nearly the same as for the ‘free’ group estimated experimentally. The change in the electron-donating power of –OH involved in H-bonding measured by the changes in sþ values amounts to about 0.2 units of s. To estimate the s value for the formyl group in o-hydroxybenzaldehyde, regression 5 is applied. The estimated values are 0.57 for the open conformation and 1.06 for the closed one. The difference is 0.49 and it results from the approach of the proton to oxygen atom of the formyl group increasing its electron-accepting power as a substituent. In summary, the difference between the substituent constants sþ(OH) and s(CHO) for the open conformation is 1.46, leading to
Figure 4. Optimized geometry of closed and open conformations of o-hydroxybenzaldehyde. HOMA and Del% values were calculated for OCCCO fragments. Substituent constants were calculated from Eqs. 4 and 5.
2013
very low delocalization estimated by HOMA¼0.02 and Gilli’s Del%¼27%. For the closed conformation the corresponding difference equals 2.17, which means an increase by 0.71 units of s. It is associated with a much stronger interaction between those substituents, revealed in increase of p-electron delocalization for the OCCCO fragment: DHOMA¼0.4 and DDel%¼22%. 3. Conclusions Our results present the first quantitative description of changes in the electronic properties of proton-donating and protonaccepting groups due to H-bonding. In the cases presented above, H-bond formation led to an increase in electron-donating power of the proton-donating substituent A–H (in both cases discussed A– H¼O–H) and electron-accepting power of proton-accepting substituent B (in both cases discussed B¼–CHO). Schematically this may be illustrated by Figure 5. In consequence, the difference in electron properties increases, which serves as a reliable explanation for increase of p-electron delocalization in the conjugated linker described in the literature as RAHB. The presented results show a cooperative effect of both groups, A–H and B, acting in two ways: (i) by means of H-bond formation, and (ii) by mesomeric effect between two substituents via the pelectron spacer OCCCO, which is undoubtedly the consequence of (i). This in turn causes an increase of p-electron delocalization in the linker. Obviously, both formally separated steps occur simultaneously. 4. Methodology Geometries of the molecules and complexes were optimized at B3LYP/6-311þG** level using Gaussian 03.26 In the H-bonded complexes the fluoride approached the 6-hydroxyfulvene molecule along the line of prolongation of the O–H bond direction. The O/F distance was controlled and varied from 4.000 Å to the distance when the proton transfer from 6-hydroxyfulvene to F was observed (i.e., 2.638 Å). The 8-formylheptafulvene complexes with HF were optimized for the same values of O/F distances as 6hydroxyfulvene with F, since for those complexes proton transfer were not observed. Moreover, linearity of O/H/F was assumed.
Figure 5. Schematic representation of changes in electronic properties of protondonor and proton-acceptor interacting via H-bond.
2014
T.M. Krygowski, J.E. Zachara-Horeglad / Tetrahedron 65 (2009) 2010–2014
HOMA values were calculated according to the following equation:23
HOMA ¼ 1
n 2 1X a R Rj n j ¼ 1 i opt;i
(8)
where n represents the total number of bonds taken into summation; ai is a normalization constant (for C–C bonds aC–C¼257.7, for C–O bonds aC–O¼157.38) fixed to give HOMA¼0 for a model nonaromatic system, e.g., Kekule´ structure of benzene and HOMA¼1 for the system with all bonds equal to the optimal value Ropt,i assumed to be realized for fully aromatic systems (Ropt,C–C¼1.388 Å, Ropt,C–O¼1.265 Å). Del% values were calculated according to the following equations:25
Del% ¼ 100,ð1 j2l 1jÞ
l ¼ 1=4,½ð2 n1Þ þ ðn2 1Þ þ ð2 n3Þ þ ðn4 1Þ
(9)
(10)
where n1, n2, n3 and n4 are the bond numbers of corresponding double and single bonds calculated from Pauling equation27 d(1)– d(n)¼c ln(n), where d(1)C–C¼1.467, d(2)C–C¼1.349, d(1)C–O¼1.367, d(2)C–O¼1.217.23 Acknowledgements We thank the Interdisciplinary Centre for Mathematical and Computational Modelling ICM (Warsaw, Poland) for providing computer time and programs. J.E.Z.-H. gratefully acknowledges financial support under grant N204 093 31/2144 and a Scholarship from the Foundation for Polish Science (FNP). References and notes 1. (a) Intermolecular Interactions: From van der Waals to Strongly Bound Complexes; Scheiner, S., Ed.; Wiley: New York, NY, 1997; (b) Desiraju, G. R.; Steiner, T. The Weak Hydrogen Bond in Structural Chemistry and Biology; Oxford University Press: New York, NY, 1999; (c) Gerlt, J. A.; Kreevoy, M. M.; Cleland, W. W.; Frey, P. A. Chem. Biol. 1997, 4, 259–267; (d) Perrin, C. L.; Nielson, J. B. Annu. Rev. Phys. Chem. 1997, 48, 511–544; (e) Desiraju, G. R. Acc. Chem. Res. 2002, 35, 565–573; (f) Meyer, E. A.; Castellano, R. K.; Diederich, F. Angew. Chem., Int. Ed. 2003, 42, 1210–1250; (g) Sobczyk, L.; Grabowski, S. J.; Krygowski, T. M. Chem. Rev. 2005, 105, 3513–3560; (h) Krygowski, T. M.; Szaty1owicz, H. Trends Org. Chem. 2006, 11, 37–53. 2. (a) Pauling, L. The Nature of the Chemical Bond; Cornell University Press: New York, NY, 1960; (b) Pure Appl. Chem. 1999, 71, 1919–1981. 3. (a) Gilli, G.; Belluci, F.; Ferretti, V.; Bertolasi, V. J. Am. Chem. Soc. 1989, 111, 1023– 1028; (b) Gilli, G.; Gilli, P. J. Mol. Struct. 2000, 552, 1–15.
4. (a) Williams, D. E.; Dunke, W. L.; Rundle, R. E. Acta Crystallogr. 1962, 15, 627– 635; (b) Ogoshi, H.; Yoshida, Z. J. Chem. Soc. D: Chem. Commun. 1970, 176– 177. 5. Scheiner, S. Hydrogen Bonding. A Theoretical Perspective; Oxford University Press: Oxford, 1997. 6. (a) Glendening, E. D.; Badenhoop, J. K.; Weinhold, F. J. Comput. Chem. 1998, 19, 628–646; (b) Weinhold, F.; Landis, C. R. Valency and Bonding; Cambridge University Press: Cambridge, 2005. ˜ ez, M.; Del Bene, J. E. Mol. Phys. 2004, 7. (a) Alkorta, I.; Elguero, J.; Mo´, O.; Ya´n ˜ ez, M.; Del Bene, J. E. 102, 2563–2574; (b) Alkorta, I.; Elguero, J.; Mo´, O.; Ya´n ˜ ez, M.; Elguero, J. Chem. Phys. Lett. 2002, 411, 411–415; (c) Sanz, P.; Mo´, O.; Ya´n ˜ ez, M.; Elguero, J. Chem.dEur. J. 2008, 14, 4225–4232; (d) Sanz, P.; Mo´, O.; Ya´n ChemPhysChem 2007, 8, 1950–1958. 8. Hydrogen BondingdNew Insights; Grabowski, S. J., Ed.Series: Challenges and Advances in Computational Chemistry and Physics; Springer: Dordrecht, 2006; Vol. 3. 9. Grabowski, S. J. J. Phys. Org. Chem. 2003, 16, 797–802. 10. (a) Grabowski, S. J. J. Mol. Struct. 2001, 562, 137–143; (b) Cuma, M.; Scheiner, S.; Kar, T. J. Mol. Struct. (THEOCHEM) 1999, 467, 37–49. 11. Krygowski, T. M.; Zachara, J. E.; Szaty1owicz, H. J. Phys. Org. Chem. 2005, 18, 110–114. 12. Grabowski, S. J. J. Phys. Org. Chem. 2004, 17, 18–31. 13. Hammett, L. P. Physical Organic Chemistry; McGraw-Hill: New York, NY, 1940. 14. Exner, O. In Advances in Linear Free Energy Relationships; Chapman, N. B., Shorter, J., Eds.; Plenum: London, 1972; pp 1–69. 15. (a) Charton, M. Prog. Phys. Org. Chem. 1973, 10, 81–204; (b) Charton, M. Prog. Phys. Org. Chem. 1981, 13, 119–251. 16. Hansch, C.; Leo, A.; Taft, W. R. Chem. Rev. 1991, 91, 165–195. 17. Krygowski, T. M.; St˛epien´, B. T. Chem. Rev. 2005, 105, 3482–3512. 18. Jaffe, H. H. Chem. Rev. 1953, 53, 191–261. 19. (a) Shorter, J. Pure Appl. Chem. 1994, 66, 2451--2468; (b) Shorter, J. Pure Appl. Chem. 1997, 69, 2497–2510. 20. Krygowski, T. M.; Zachara-Horeglad, J. E.; Palusiak, M.; Pelloni, S.; Lazzeretti, P. J. Org. Chem. 2008, 73, 2138–2145. 21. Krygowski, T. M.; Zachara, J.; Szaty1owicz, H. J. Org. Chem. 2004, 69, 7038–7043. ´ , B. T.; Cyran´ski, M. K.; Krygowski, T. M. Chem. Phys. Lett. 2001, 350, 22. Stepien ˛ 537–542. 23. Krygowski, T. M. J. Chem. Inf. Comput. Sci. 1993, 33, 70–78. 24. (a) Krygowski, T. M.; Szaty1owicz, H.; Zachara, J. E. J. Chem. Inf. Model. 2005, 45, 652–656; (b) Krygowski, T. M.; Szaty1owicz, H.; Zachara, J. E. J. Chem. Inf. Comput. Sci. 2004, 44, 2077–2082. 25. Bertolasi, V.; Pretto, L.; Ferretti, V.; Gilli, P.; Gilli, G. Acta Crystallogr., Sect. B 2006, 62, 1112–1120. 26. Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.; Montgomery, J. A., Jr.; Vreven, T.; Kudin, K. N.; Burant, J. C.; Millam, J. M.; Iyengar, S. S.; Tomasi, J.; Barone, V.; Mennucci, B.; Cossi, M.; Scalmani, G.; Rega, N.; Petersson, G. A.; Nakatsuji, H.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.; Nakai, H.; Klene, M.; Li, X.; Knox, J. E.; Hratchian, H. P.; Cross, J. B.; Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev, O.; Austin, A. J.; Cammi, R.; Pomelli, C.; Ochterski, J. W.; Ayala, P. Y.; Morokuma, K.; Voth, G. A.; Salvador, P.; Dannenberg, J. J.; Zakrzewski, V. G.; Dapprich, S.; Daniels, A. D.; Strain, M. C.; Farkas, O.; Malick, D. K.; Rabuck, A. D.; Raghavachari, K.; Foresman, J. B.; Ortiz, J. V.; Cui, Q.; Baboul, A. G.; Clifford, S.; Cioslowski, J.; Stefanov, B. B.; Liu, G.; Liashenko, A.; Piskorz, P.; Komaromi, I.; Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham, M. A.; Peng, C. Y.; Nanayakkara, A.; Challacombe, M.; Gill, P. M. W.; Johnson, B.; Chen, W.; Wong, M. W.; Gonzalez, C.; Pople, J. A. Gaussian 03, Revision C.02; Gaussian: Wallingford, CT, 2004. 27. Pauling, L. J. Am. Chem. Soc. 1947, 69, 542–553.