Origin of the Unusual Ultraviolet Absorption of Arsenicin A - The

ARTICLE pubs.acs.org/JPCA

Origin of the Unusual Ultraviolet Absorption of Arsenicin A Sundaram Arulmozhiraja, Michelle L. Coote,* Di Lu, Geoffrey Salem, and S. Bruce Wild Research School of Chemistry, Australian National University, Canberra, ACT 0200, Australia

bS Supporting Information ABSTRACT: This paper presents a combined experimental and theoretical study of the electronic spectrum of the natural adamantane-type polyarsenical Arsenicin A. Experiments reveal that this molecule strongly absorbs UV light in the absence of an obvious chromophore. The observed absorbance is supported by the time-dependent density functional (TD-DFT) calculations with B3LYP, M06-L, and M06-2X functionals combined with the 6-311þG(3df,2pd) basis set, as well as by symmetryadapted cluster/configuration interaction (SAC-CI) theory. The theoretical investigations reveal that the absorption is facilitated by through-space and through-bond interactions, between the lone pairs on the arsenic and oxygen atoms and the σ-bonding framework of the molecule, that destabilize occupied and stabilize unoccupied molecular orbitals.

1. INTRODUCTION Arsenicin A, a natural polyarsenical that has been isolated from the New Caledonian sea sponge Echinochalina bargibanti, has been shown to have potent bactericidal and fungicidal activities on certain human pathogenic strains.1 On the basis of experimental and theoretical studies, an adamantane-type structure was proposed that resembles that of arsenic(III) oxide (As4O6) in which three oxygen atoms have been replaced by methylene groups in a C2 arrangement.2,3 We recently reported the synthesis of Arsenicin A and confirmed the proposed structure of the racemate by X-ray crystallography (Figure 1).4 In the light of the chemical structure it is surprising to note that the naturally occurring compound absorbs in the UV region.1 This raises the questions: what is the chromophore in Arsenicin A, and how do we account for its UV spectrum? UV absorption in a saturated adamantane-type molecule appears to be without experimental precedent, although there have been a number of theoretical investigations of the electronic spectra of related molecular cages.5
11 For example, a study of the electronic structure of a series of poly(thiaadamantanes) indicated that interactions between sulfur lone pairs could destabilize some orbitals but stabilize others.11 Another investigation of poly(heteroadamantanes) led to the conclusion that the stabilities of the molecules depended on an interplay between hyperconjugative, electrostatic, and strain effects.12 These results suggest that similar effects could be responsible for the unexpected UV absorbance of Arsenicin A. To probe this further, theoretical studies of Arsenicin A have been performed using time-dependent density functional theory (TD-DFT).13
15 To verify the results, we have also measured the UV spectrum of the synthetic compound. Herein we report the results of our theoretical and experimental studies.

the B3LYP functional17,18 with the 6-311þG(3df,2pd) basis set and were followed by frequency calculations in order to verify that the optimized structures were true energy minima on their potential energy surfaces. Electronic structures were studied by employing TD-DFT with the B3LYP functional with the same large split-valence basis set, 6-311þG(3df,2pd), at the optimized ground-state structures. Overlap integrals and orbital interactions were calculated numerically by using natural bond orbital (NBO) analysis.19
22 For this purpose, the NBO 3.0 software incorporated into GAUSSIAN03 was used. It has been concluded in recent studies that the B3LYP functional can be used to study the electronic spectra of stable molecules.23
25 To verify further its applicability to the study of As-containing molecules, we conducted a brief assessment study. First, we studied a number of diatomic arsenic molecules whose experimental values are available in the literature. These results are given in Table S1 of the Supporting Information. Further to check the applicability of the B3LYP functional for the title compound, we compared the calculated structure and vibrational frequencies of Arsenicin A with the experimental values reported previously (Table S2 of the Supporting Information). To verify the suitability of our methodology for predicting the electronic spectrum of the Arsenicin A, we compared our calculated results with the experimental data from the present work, and also with corresponding calculations which we performed using the M06L 26 and M06-2X 27 functionals in TD-DFT and symmetryadapted cluster/configuration interaction (SAC-CI) theory.28
31 As in the case of B3LYP, the other TD-DFT calculations were carried out with the 6-311þG(3df,2pd) basis set at their respective optimized ground-state structures obtained using the same 6-311þG(3df,2pd) basis set. However, due to its greater cost, we

2. THEORETICAL METHODOLOGY All calculations were performed using the GAUSSIAN03 suite of programs.16 Geometry optimizations were carried out using

Received: January 28, 2011 Revised: March 14, 2011 Published: March 24, 2011

r 2011 American Chemical Society

4530

dx.doi.org/10.1021/jp200956b | J. Phys. Chem. A 2011, 115, 4530–4534

The Journal of Physical Chemistry A

Figure 1. Arsenicin A. Atoms: in violet, As; in red, O; in black, C; in white, H.

Figure 2. Comparative excitation energies (eV) obtained in all four theoretical methods with their relevant experimental peaks.

employed the SAC-CI, singles and doubles (SD)-R method with LevelTwo accuracy in conjunction with the smaller 6-311þG(d, p) basis set, using the B3LYP/6-311þG(3df,2pd) optimized geometry. The calculated results, along with the experimental values, are given in Tables S3
S7 of the Supporting Information. A summary of the principal results is provided in Figures 2 and 3, which represent the comparison of the experimental and various theoretical predictions of the excitation energies and normalized peak intensities, respectively. Tables S1 and S2 reveal clearly that the B3LYP functional can be used to study the ground-state properties of As-containing molecules including Arsenicin A. Further, it is evident from Figure 2 and Tables S3
S7 that the excitation energies calculated using all of the selected theories correlate well with the experimental values and thus these theories could produce reliable excitation energies. Since the experimental absorbance is proportional to oscillator strengths, we can compare the relative intensities of the experimental and theoretical data (Figure 3). Results reveal that the experimental trends are well-produced qualitatively and quantitatively by almost all of the theoretical methods. The values calculated using B3LYP show the best overall correlation with experiment, and we therefore use B3LYP results for the remainder of our study. It should be noted here that SAC-CI is a higher level of theory and would normally be

ARTICLE

Figure 3. Comparison between the normalized experimental absorbance of the various peaks with the normalized oscillator strengths, as obtained using different levels of theory.

Figure 4. UV/vis spectrum of Arsenicin A.

expected to provide superior performance to TD-DFT procedures. However, due to its higher computational cost, we were unable to apply it under its most accurate conditions (i.e., using the LevelThree accuracy and applying it with a large basis set incorporting Rydberg functions) and this is presumably why the TD-DFT methods out-performed it on this occasion. It is worth mentioning that there is a good agreement between the various functionals, especially B3LYP and M06-L, concerning the order and even excitation origin of some of the lowest excited states. However, we note that TDDFT methods may not always be suitable for higher energy excited states.

3. RESULTS Arsenicin A was prepared by the published four-step procedure from methylenebis(phenylarsinic acid).4 The UV/vis spectrum of the compound, mp 182
184 °C, in dichloromethane (0.0424 g dm
3) was measured in a quartz cell with use of a Shimadzu UV-2450 UV
visible spectrometer. The spectrum shown in Figure 4 contains peaks at λmax/nm [CH2Cl2 (ε/(dm3 mol
1 cm
1))]: 314 (2399), 288 (1599), 257 (3971), and 230 (11 830). The data are consistent with those reported for naturally occurring Arsenicin A except that the high-energy band was reported at 240 nm with a significantly lower molar absorptivity.1 4531

dx.doi.org/10.1021/jp200956b |J. Phys. Chem. A 2011, 115, 4530–4534

The Journal of Physical Chemistry A

ARTICLE

Table 1. Excitation Energies (ΔE) and Oscillator Strengths (f) Corresponding to the Most Important Excited States of Arsenicin A

The calculated vertical excitation energies, oscillator strengths, principal configurations of each transition, and their proposed assignments, along with the experimental values, are given in Table S3. The assignments were made on the basis of the calculated oscillator strengths. In most cases, the assigned state had an oscillator strength that was at least 1 order of magnitude greater than the other possible assignments. However, in the case of the 4.83 eV experimental peak, the assigned state 6 1B had an oscillator strength that was approximately twice that of the alternative assignment 5 1B, and it is possible that this latter state contributes. The important molecular orbitals of Arsenicin A are also provided in the Supporting Information (Figure S1); the principal results are summarized in Table 1. Throughout this study, the molecular orbitals are labeled according to their irreducible representations a and b, and within each irreducible representation are numbered according to their increasing energy. For clarity, the key orbitals are also labeled in terms of their relationship to the HOMO (highest occupied molecular orbital) or LUMO (lowest unoccupied molecular orbital). Interestingly, there are at least 35 excited states present within the 7 eV energy range. It should be noted that some of the transitions have moderate to large oscillator strengths. This suggests that some of the peaks in the experimental absorption spectra originate from the combination of many electronic transitions. The calculated excitation energies for the different states agree well with our experimental values (maximum error is about 0.35 eV) and the experimentally observed UV absorbance therefore can be reproduced theoretically. Taken together, the experimental and theoretical results in Table 1 mutually confirm that the UV spectrum of Arsenicin A is indeed associated with the chemical structure depicted in

Figure 1. In what follows we show that the origin of this UV absorbance is due to the presence of the lone pairs, especially on arsenic, and the through-space and through-bond interactions between them and the As
C and As
O framework that help to destabilize the occupied orbitals and stabilize the unoccupied orbitals, thereby reducing the energy of the electronic transitions.

4. DISCUSSION Arsenicin A lacks a functional group that might be normally recognized as a chromophore; however, it has some interesting aspects that might help to account for its UV absorbance. For example, it has four arsenic and three oxygen atoms, and each arsenic atom has one p-type nonbonding orbital, denoted as As(p) in the present study, and each oxygen has two p-type nonbonding orbitals, referred to as O(p). By interacting with themselves, these 10 nonbonding orbitals can split into more orbitals and it is expected that these nonbonding orbitals play an important role in the electronic excitations of Arsenicin A. Additionally, Arsenicin A has C2 symmetry, and therefore all transitions are symmetry allowed on both one- and two-photon absorption, and it is therefore expected to have many mixed excited states. Also, for symmetry reasons, the σ
π partitioning of the molecular orbitals is not strictly valid, and though the As(p) and O(p) lone pairs are primarily centered on these respective atoms, the As(p) lone pairs in particular extend along As
C and As
O bonds and interact with these σ-bonds. The cage-like structure of Arsenicin A also plays a role by holding the orbitals in close proximity, allowing them to interact strongly. Given these features, intramolecular interactions 4532

dx.doi.org/10.1021/jp200956b |J. Phys. Chem. A 2011, 115, 4530–4534

The Journal of Physical Chemistry A

ARTICLE

Figure 6. Structures of the Arsenicin A models obtained at the B3LYP/ 6-31G(d) level of theory. Atom colorings are as in Figure 1.

Figure 5. Model systems along with their calculated HOMO and LUMO stabilities (Hs and Ls, respectively), calculated relative to those of AsH3 at the B3LYP/6-311þG(d,p) level of theory. All values are in electronvolts. A positive sign represents “stabilized” compared with AsH3, and the negative sign refers to the opposite. Atom colorings are as in Figure1.

could play an important role in the electronic structure of Arsenicin A. Hoffmann et al.32
34 introduced and classified intramolecular interactions as either through space (TS) or through bond (TB). The former result from direct interactions between nearby orbitals and the latter involves a coupling through the connecting σ-bond framework. Arsenicin A has both of these elements: (1) extension of p-type nonbonding orbitals into the interior of the cage and (2) separation of these orbitals by only two σ-bonds. Thus, both TS and TB interactions are expected to be significant. As noted above, the Houk group recently characterized such intramolecular interactions, and their resultant orbital splitting, in poly(thiaadamantanes).11 Such interactions might be expected to be further enhanced in Arsenicin A, due to the presence of the heavier arsenic atoms. In Arsenicin A, the possible intramolecular interactions are due to the interactions among the lone-pair orbitals and the interactions between the lone pairs with the neighboring σAs
C and σAs
O bonds. While TS interactions can be understood as the direct interaction between nearby orbitals, there are two types of TB interactions: lone-pair
σ repulsive interactions between a lone pair and a neighboring σ-bond, and lone-pairfσ* hyperconjugation interactions between a lone pair and adjacent σ*As
C and σ*As
O orbitals. Both types of TB interaction are important: calculated overlap integrals confirm that As(p) lone pairs mix significantly with their neighboring As
C and As
O σ-bonds (see Table S8 of the Supporting Information), while a number of calculated lone-pairfσ* interactions are of the order of 0.2 eV (see Table S9 of the Supporting Information). To study the effect of these intramolecular interactions on the orbital energies of Arsenicin A, we compared its HOMO and LUMO energies with those of five model systems (Figure 5), as calculated at the same level of theory. Figure 5 reveals that the HOMO of the Arsenicin A is destabilized by 1.41 eV when compared with that of AsH3 and its LUMO is stabilized by 1.28 eV compared to that of the latter. As seen in Figure S1, the HOMO (44b) involves a substantial mixture of TS interactions and TB interactions, in which As(p) orbitals mix with the adjacent σAs
C and σAs
O bonds. The calculations on the model systems indicate that the As
C bonding network plays an important

role in the destabilization of the Arsenicin A HOMO. TS As(p) interactions also destabilize the HOMO; however, there is no contribution from the oxygen atom to this destabilization. It is possible that the stabilization of HOMO by O(p) might be compensated by the destabilization of As
O bonds. Figure 5 also indicates that the LUMO of Arsenicin A is stabilized by As
O bonding networks and by As(p) TS interactions, but it is not influenced by the carbon networks. To further understand the role of the heteroatom lone pairs, we also compared the electronic structure of Arsenicin A with its parent cage molecule, adamantane. It was found that the HOMO of Arsenicin A is destabilized by 1.38 eV and its LUMO is stabilized by 1.57 eV when compared with that of the respective orbitals of adamantane. The molecular orbitals in adamantane are mixed and many orbitals contribute to the transitions in complicated ways. Three triply degenerate T2 states (formed through the transitions with noticeable oscillator strengths among the calculated 20 low-lying excited states) have excitation energies of 6.653, 7.374, and 7.834 eV, respectively. The excitation with the largest oscillator strength occurs at a substantially higher energy (7.374 eV) than in Arsenicin A (5.333 eV), highlighting the important role of intramolecular interactions between the heteroatom lone pairs in the latter compound. To quantify the relative contributions of TS versus TB intramolecular interactions in Arsenicin A, we performed the following TB and TS interaction analysis. Since oxygen atoms are bonded with the arsenic atoms, separating TB and TS interactions in Arsenicin A could not be accomplished directly, so we instead considered an Arsenicin model, model A, in which of the oxygen atoms are replaced by CH2 groups (Figure 6). We optimized model A at the B3LYP/6-31G(d) level of theory. At first orbital splitting due to TS interactions was determined by removing the carbon skeletons from model A and replacing each arsenic atom with an AsH3 group. This modified model structure, e.g., model B, is optimized at the B3LYP/6-31G(d) level with the fixed arsenic coordinates and with constrained bond angles and dihedrals; i.e., only As
H bonds were allowed to optimize. Now the orbital splitting in the model system is caused by TS interactions alone; splitting due to the TB interactions can be derived from the net orbital splitting of the model A. Both model structures have Td symmetry, and their HOMOs are triply degenerate. So orbital splitting is calculated between these degenerate orbitals and the next neighboring occupied orbitals. The calculated TS is 0.44 eV, and the net orbital splitting is 2.90 eV, affording a TB contribution of 2.46 eV. These values indicate that TB interactions dominate over TS in the model system studied. It should be noted however 4533

dx.doi.org/10.1021/jp200956b |J. Phys. Chem. A 2011, 115, 4530–4534

The Journal of Physical Chemistry A that the presence of oxygen atoms in Arsenicin A could influence these results. In a nutshell, the HOMO (44b) of the Arsenicin A is destabilized by As(p)
σAs
C repulsive interactions and by As(p) TS interactions while its LUMO (47a) is stabilized through the delocalization of lone-pair electrons mainly into vacant σ*As
O orbitals, resulting in an orbital, denoted π*[As(p)] TS, with a π*-like wave function and density distribution.35 This destabilization of the HOMO (and other occupied orbitals) of Arsenicin A, and the concurrent stabilization of the LUMO (and other unoccupied) orbitals due to the intramolecular interactions, explains its UV absorbance. The strongest observed peak (5.39 eV; calculated at 5.33 eV) is mainly due to the As(p)fπ*[As(p)] TS transition, but the As(p) lone-pair orbital is mixed with the σAs
C bond and (to a lesser extent) with σAs
O and O(p). All four peaks observed in the experimental spectrum were assigned (see Table 1); most are mainly due to either As(p)fπ*[As(p)] TS or As(p)fσ*As
C or As(p)fσ*As
C þ σ*As
C transitions.

5. CONCLUSION In this work, the photoelectronic spectrum of Arsenicin A has been measured and calculated using TD-DFT theory. The calculated vertical excitation energies agree well with the experimental values. The presence of lone pairs, especially on arsenic, and the through-space and through-bond interactions between them and the As
C and As
O framework account for the strong UV absorption of Arsenicin A. ’ ASSOCIATED CONTENT

bS

Supporting Information. Further computational details, including orbital diagrams, geometries, electronic spectra, overlap integrals, and orbital interaction energies. This information is available free of charge via the Internet at http://pubs.acs.org.

’ AUTHOR INFORMATION Corresponding Author

*Phone: þ61-2-6125-3771. Fax: 61-2-6125-0750. E-mail: mcoote@ rsc.anu.edu.au.

’ ACKNOWLEDGMENT We gratefully acknowledge support from the Australian Research Council and generous allocations of computing time on the National Facility of the Australian National Computational Infrastructure. M.L.C. acknowledges receipt of an ARC Future Fellowship. ’ REFERENCES (1) Mancini, I.; Guella, G.; Frostin, M.; Hnawia, E.; Laurent, D.; Debitus, C.; Pietra, F. Chem.—Eur. J. 2006, 12, 8989–8994. (2) Tahtinen, P.; Saielli, G.; Guella, G.; Mancini, I.; Bagno, A. Chem. —Eur. J. 2008, 14, 10445–10452. (3) Guella, G.; Mancini, I.; Mariotto, G.; Rossi, B.; Viliani, G. Phys. Chem. Chem. Phys. 2009, 11, 2420–2427. (4) Lu, D.; Rae, A. D.; Salem, G.; Weir, M. L.; Willis, A. C.; Wild, S. B. Organometallics 2010, 29, 32–33. (5) Litvinyuk, I. V.; Zheng, Y.; Brion, C. E. Chem. Phys. 2000, 253, 41–50. (6) Litvinyuk, I. V.; Zheng, Y.; Brion, C. E. Chem. Phys. 2000, 261, 289–300.

ARTICLE

(7) Tian, S. Xi.; Kishimoto, N.; Ohno, K. J. Phys. Chem. A 2002, 106, 6541–6553. (8) Ando, S.; Ueda, M. J. Photopolym. Sci. Technol. 2003, 16, 537–544. (9) Lehtonen, O.; Sundholm, D. Phys. Rev. B 2006, 74, No. 045433. (10) Jankowiak, A.; Kaszynski, P. J. Org. Chem. 2009, 74, 7441–7448. (11) Norton, J. E.; Briseno, A. L.; Wudl, F.; Houk, K. N. J. Phys. Chem. A 2006, 110, 9887–9899, and references therein. (12) Wang, Y.; Wu, J. I-C.; Li, Q.; Schleyer, P. v. R. Org. Lett. 2010, 12, 1320–1323. (13) Stratmann, R. E.; Scuseria, G. E.; Frisch, M. J. J. Chem. Phys. 1998, 109, 8218–8224. (14) Bauernschmitt, R.; Ahlrichs, R. Chem. Phys. Lett. 1996, 256, 454–464. (15) Casida, M. E.; Jamorski, C.; Casida, K. C.; Salahub, D. R. J. Chem. Phys. 1998, 108, 4439–4449. (16) 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.; Bakken, V.; 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 E.01; Gaussian: Wallingford, CT, 2004. (17) Becke, A. J. Chem. Phys. 1993, 98, 5648–5652. (18) Lee, C.; Yang, W.; Parr, R. G. Phys. Rev. B 1998, 37, 785–789. (19) Reed, A. E.; Curtiss, L. A.; Weinhold, F. Chem. Rev. 1988, 88, 899–926. (20) Reed, A. E.; Weinstock, R. B.; Weinhold, F. J. Chem. Phys. 1985, 83, 735–746. (21) Reed, A. E.; Weinhold, F. J. Chem. Phys. 1985, 83, 1736–1740. (22) Reed, A. E.; Weinhold, F. J. Chem. Phys. 1983, 78, 4066–4073. (23) Silva-Junior, M. R.; Schreiber, M.; Sauer, S. P.; Thiel, W. J. Chem. Phys. 2008, 129, No. 104103. (24) Dierksen, M.; Grimme, S. J. Chem. Phys. 2004, 120, 3544–3554. (25) Jacquemin, D.; Perpete, E. A.; Ciofini, I.; Adamo, C.; Valero, R.; Zhao, Y.; Truhlar, D. G. J. Chem. Theory Comput. 2010, 6, 2071–2085. (26) Zhao, Y.; Truhlar, D. G. J. Chem. Phys. 2006, 125, No. 194101. (27) Zhao, Y.; Truhlar, D. G. Theor. Chem. Acc. 2008, 120, 215–241. (28) Nakatsuji, H.; Hirao, K. J. Chem. Phys. 1978, 68, 2053–2065. (29) Nakatsuji, H. Chem. Phys. Lett. 1978, 59, 362–364. (30) Nakatsuji, H. Chem. Phys. Lett. 1979, 67, 329–333; 1979, 67, 334–342. (31) Nakatsuji, H. Acta Chem. Hung. 1992, 129, 719.Computational Chemistry—Review of Current Trends; World Scientific: Singapore, 1997; Vol. 2, pp 61
124. (32) Hoffmann, R.; Imamura, A.; Hehre, W. J. J. Am. Chem. Soc. 1968, 90, 1499–1509. (33) Hoffmann, R. Acc. Chem. Res. 1971, 4, 1–9. (34) Hoffmann, R.; Heilbronner, E.; Gleiter, R. J. Am. Chem. Soc. 1970, 92, 706–707. (35) Because of the same mixing and π*-like wave function and density distribution reasons, LUMOþ1 (MO 48a) is also denoted as π*[As(p)] TS. LUMOþ2 (MO 45b) on the other hand, looks like a σ*type orbital due to the mixing of σ*As
O As(p) in a σ*-like fashion, while LUMOþ4 (MO 50a) is a kind of σ*-type orbital due to the mixing of both σ*As
O and σ*As
C with As(p). We refer these orbitals as σ*As
O and σ*As
C þ σ*As
C, respectively. Most of the lone-pair orbitals in the Arsenicin A are not the pure nonbonding orbitals because of symmetry reasons. 4534

dx.doi.org/10.1021/jp200956b |J. Phys. Chem. A 2011, 115, 4530–4534