FULL PAPER
WWW.C-CHEM.ORG
Quantum-Chemical Analyses of Aromaticity, UV Spectra, and NMR Chemical Shifts in Plumbacyclopentadienylidenes Stabilized by Lewis Bases Toshiaki Kawamura,[a,b] Minori Abe,[a,b] Masaichi Saito,[c] and Masahiko Hada*[a,b] We carried out a series of zeroth-order regular approximation (ZORA)-density functional theory (DFT) and ZORA-timedependent (TD)-DFT calculations for molecular geometries, NMR chemical shifts, nucleus-independent chemical shifts (NICS), and electronic transition energies of plumbacyclopentadienylidenes stabilized by several Lewis bases, (Ph)2(tBuMe2Si)2C4PbL1L2 (L1, L2 5 tetrahydrofuran, Pyridine, N-heterocyclic carbene), and their model molecules. We mainly discussed the Lewis-base effect on the aromaticity of these complexes. The NICS was used to examine the aromaticity. The NICS values showed that the aromaticity of these complexes increases when the donation from the Lewis bases to Pb becomes large.
This trend seems to be reasonable when the 4n-Huckel rule is applied to the fractional p-electron number. The calculated 13 C- and 207Pb-NMR chemical shifts and the calculated UV transition energies reasonably reproduced the experimental trends. We found a specific relationship between the 13C-NMR chemical shifts and the transition energies. As we expected, the relativistic effect was essential to reproduce a trend not only in the 207Pb-NMR chemical shifts and J[Pb-C] but also in the 13C-NMR chemical shifts of carbons adjacent to the lead C 2014 Wiley Periodicals, Inc. atom. V
Introduction
There are prior theoretical investigations on aromaticity and magnetic properties of dilithioplumbole analogs. Magnetic criteria including NICS of Silole dianions and metaloles containing Si, Ge, Sn, and Pb have been reported.[9,10] Pioneering theoretical calculations of metal NMR[11,12] were helpful for the present analyses. The relativistic correction, especially the spinorbit effect, is essentially required in heavy-element NMR chemical shifts.[13–20] We show herein the quantum-chemical calculations of several Lewis-base-stabilized plumbacyclopentadienylidenes. This work has two objectives. The first one is the verification of aromaticity. The p-orbitals of neutral plumbacyclopentadienylidenes are occupied by four electrons, whereas those of plumbole dianions are occupied by six electrons. The latter is suggested to have aromaticity.[5] We here discuss the aromaticity of neutral plumbacyclopentadienylidenes coordinated by Lewis-base ligands. A ligand donates electrons to an unoccupied p-orbital of a neutral plumbacyclopentadienylidene, which mainly consists of the 6p orbital of lead. In this case, it
Aromatic compounds have been studied for a long time principally in organic and organometallic chemistry. Although aromatic compounds have a variety of ring sizes and shapes, most of them consist of carbon, nitrogen, and oxygen atoms. Recently, we have shifted our interest to aromatic compounds containing heavy elements, such as silicon, germanium, tin, and lead, instead of carbon.[1–3] Also, theoretical aspect of silicon compounds have been already summarized.[4] Dilithioplumbole, a lead-containing five-membered ring, was reported in 2010.[5] Dilithioplumbole is an analog of dilithiocyclopentadiene CH4CLi2; a carbon atom is replaced with a lead atom, and therefore, its molecular framework is a planar cyclic form of C4H4Pb. Two lithium atoms are solvated in solution and in the crystalline state. The p-orbitals of C4H4Pb are occupied by six p-electrons as the two additional electrons are supplied by lithium atoms to the plumbole’s p-system. The H€ uckel’s (4n 1 2) p-electron rule[6] is formally satisfied in this compound, and thus, it strongly suggests that dilithioplumbole has aromaticity. Actually, the disappearance of CAC alternation is observed in this compound. The nucleus-independent chemical shift (NICS)[7] values also indicate that dilithioplumbole is an aromatic compound. Conversely, plumbacyclopentadienylidenes, where the lead atoms have divalent states, have p-orbitals containing four electrons, and it is not aromatic according to the H€ uckel’s (4n 1 2) p-electron rule. Recently, plumbacyclopentadienylidenes, stabilized by tetrahydrofuran (THF), pyridine, and N-heterocyclic carbene (NHC), were synthesized and characterized.[8] In these compounds, the total number of pelectrons is fractional, and therefore, the H€ uckel’s (4n 1 2) pelectron rule seems to be partially satisfied.
DOI: 10.1002/jcc.23556
[a] T. Kawamura, M. Abe, M. Hada Department of Chemistry, Graduate School of Science and Engineering, Tokyo Metropolitan University, Minami-Osawa 1-1, Hachi-Oji, Tokyo 1920397, Japan E-mail: [email protected] [b] T. Kawamura, M. Abe, M. Hada JST, CREST, 4-1-8 Honcho, Kawaguchi, Saitama 332-0012, Japan [c] M. Saito Department of Chemistry, Graduate School of Science and Engineering, Saitama University, Shimo-okubo, Sakura-ku, Saitama 338-8570, Japan Contract grant sponsor: JST, CREST; Contract grant sponsor: MEXT [M.S. (Grant-in-Aids for Scientific Research)]; Contract grant number: 25109510 C 2014 Wiley Periodicals, Inc. V
Journal of Computational Chemistry 2014, 35, 847–853
847
FULL PAPER
WWW.C-CHEM.ORG
seems that the H€ uckel’s (4n 1 2) p-electron rule is satisfied, although two p-electrons are shared with the plumbole and the ligand. We are interested in whether the ligandcoordinated plumbacyclopentadienylidenes have aromaticity or not. To check aromaticity theoretically, we calculate NICS values. For comparison, we also performed calculations for plumbole dianion to which two electrons are supplied by the two lithium ions, and therefore, the H€ uckel’s (4n 1 2) pelectron rule is completely satisfied. The second objective is to discuss the trend of the NMR chemical shifts and the nuclear spin–spin coupling constants J[Pb,C], and the electronic excitation energies of several ligand-coordinated plumbacyclopentadienylidenes. We examine the relationship between NMR chemical shifts and excitation energies, and discuss the ligand effect. We also discuss the relativistic effect of Pb on the NMR chemical shifts and J[Pb,C]. We note here as follows to avoid a confusion. Plumbole is generally used as a lead analog of cyclopentadiene with a tetravalent lead atom. However, in the following sections, we use the terminology, plumbole, as an abbreviation for plumbacyclopentadienylidene.
Calculation Method Figure 1 shows the molecular formulae of plumboles stabilized by Lewis-base ligands, which are calculated in the present investigation. In this article, plumbole is always defined as (SiC6H15)(C6H5)2AC4Pb. Three kinds of ligands, THF (C4H8O), pyridine (Py, C5H5N), and ((C3H7)2(CH3)2), are taken into account, as shown in Figure 1. We use the abbreviations listed below.
Plumbole 1 THF (P1THF) Plunbole 1 THF2(P1THF2) Plumbole 1 Py (P1Py) Plumbole 1 Py2 (P1Py2) Plumbole 1 NHC(P1NHC) Plumbole 1 NHC2(P1NHC2)
L1
L2
THF THF Py Py NHC NHC
THF Py NHC
THF is the weakest Lewis-base of the three ligands and NHC is the strongest. We perform calculations for six compounds (plumbole 1 THF, plumbole 1 THF2, plumbole 1 Py, plumbole 1 Py2, plumbole 1 NHC, and plumbole 1 NHC2). Plumbole 1 THF2, plumbole 1 Py2, and plumbole 1 NHC can be recrystallized, and thus, their molecular structures are determined experimentally.[8] The other three compounds cannot be recrystallized; nevertheless, we perform all calculations of them because they may exist in solution. As lead is a heavy element and its relativistic effect may be considerably large, we use the zeroth-order regular approximation (ZORA).[21–25] For comparison, we also perform a nonrelativistic Hamiltonian. Geometry optimizations, the 207Pb and 13C magnetic shielding tensors,[26] J[Pb,C],[27,28] and NICS values are calculated by ZORA-B3LYP[29,30] with the GIAO approach and the collinear functional approach. The basis functions used are TZ2P for Pb, TZP for C in a five-membered ring and 848
Journal of Computational Chemistry 2014, 35, 847–853
Figure 1. Molecular formulae of ligand-coordinated plumboles.
atoms adjacent to Pb, DZ for H, and DZP for the other atoms. Electronic excitation energies corresponding to the UV spectra are obtained by ZORA-TD-DFT, the functional of which is VWN,[31] using the same basis sets as those for the geometry optimization. All calculations are performed with ADF 2012.01a program.[32]
Result and Discussion Aromaticity Although the definition of aromaticity is not clear-cut, we can immediately show the following four criteria. The first criterion is the conventional H€ uckel’s (4n 1 2) p-electron rule. The second one is the planar ring and the disappearance of CAC bond alternation. The third one is the negative NICS value. The last one is ring stability, namely, a large heat of formation. In this discussion, the H€ uckel’s (4n 1 2) p-electron rule is not applicable as the number of p-electrons cannot be defined with certainty in the ligand-coordinated plumboles. In the quantum-chemical calculations, the second and third criteria are applicable. These criteria were used in our previous investigation.[33] In the following sections, NICS(0) and NICS(1) are used; (0) means that NICS is calculated on the ring plane of a ˚ above the molecule; (1) means that NICS is calculated on 1 A ring plane. To demonstrate and confirm the disappearance of the CAC bond alternation and the negative NICS values in a series of aromatic five-membered analogs, we perform calculations for simple five-membered rings C4H4X and C4H4X22 (X 5 Si, Ge, Si, Pb). In those compounds, we can apply the H€ uckel’s (4n 1 2) p-electron rule. The neutral compound C4H4X has been reported previously.[34] In our calculations, the lone-pair orbital of X in C4H4X is always occupied by two electrons, and therefore, C4H4X has four p-electrons, whereas C4H4X22 has six pelectrons. Table 1 shows that both the disappearance of the bond alternation and the negative NICS values are effectively equivalent in the aromaticity/nonaromaticity of five-membered C4H4X and C4H4X22 (X 5 Si, Ge, Si, Pb) rings. Actually, in neutral C4H4X, the bond lengths of CaACb and CbACb are 1.3 and WWW.CHEMISTRYVIEWS.COM
FULL PAPER
WWW.C-CHEM.ORG
Table 1. Comparison of bond lengths and NICS values of C4H4 X and C4H4X22 (X 5 Si, Ge, Sn, and Pb).
C4H4Si C4H4Ge C4H4Sn C4H4Pb C4H4Si22 C4H4Ge22 C4H4Sn22 C4H4Pb22
CaACb
CbACb
CaAX
NICS(0)
NICS(1)
1.333 – 1.330 1.329 1.431 1.422 1.416 1.409
1.519 – 1.520 1.519 1.385 1.389 1.396 1.409
1.883 – 2.183 2.247 1.825 1.936 2.119 2.183
14.0 – 10.2 9.8 27.1 26.6 27.1 26.2
3.8 – 2.2 2.7 28.1 28.0 28.0 27.3
˚ , respectively, so obvious bond alternations appear. In dia1.5 A nion C4H4X22, the bond lengths of both CaACb and CbACb ˚ . The disappearance of the bond alternation is are around 1.4 A a clear indication of the aromaticity of those compounds. Conversely, both NICS(0) and NICS(1) values of C4H4X are positive, indicating that they are not aromatic molecules. Those of C4H4X22 are negative, indicating that C4H4X22 has aromaticity. Figure 2 shows the schematic molecular structures of plumbole and six types of ligand-coordinated plumbole: plumbole 1 THF, plumbole 1 THF2, plumbole 1 Py, plumbole 1 Py2, plumbole 1 NHC, and plumbole 1 NHC2, optimized by the ZORA-B3LYP method. Table 2 lists the important geometrical parameters for the compounds in Figure 2. The ligands are always located on the Pb atom, and are almost perpendicular to the C4Pb ring. The CaACb bond lengths of plumbole 1 ˚ , respectively, whereas the experiTHF2 are 1.364 and 1.363 A
˚ .[8] Conversely, the CbACb mental ones are 1.346 and 1.347 A ˚ bond length is 1.539 A, whereas the experimental one is 1.517 ˚ . The calculated and experimental CAC bond lengths agree A well and bond alternation occurs. This trend is also found in the other ligand-coordinated compounds. Therefore, based on the criterion of bond alternation, the ligand-coordinated plumboles should not be aromatic. To confirm the third criterion, namely, the negative NICS value, we show the NICS values in Table 3. Those values are calculated by conventional nonrelativistic and ZORA Hamiltonians. Although the relativistic effect on NICS is quite small, we use the relativistic results in the following discussion. As already discussed using the data in Table 1, the NICS(0) value of neutral plumbole, which has four p-electrons, is 13.1 ppm, and therefore, neutral plumbole without ligands is not aromatic but antiaromatic. Conversely, the NICS value of plumbole22, which has six p-electrons, is 25.1 ppm, suggesting that plumbole22 has aromaticity. Note that CH4Pb in Table 1 is essentially the same as plumbole in Table 3 and Figure 2, and the difference appears on the side chain of the fivemembered ring. Conversely, the NICS(0) values of the ligandcoordinated plumboles range from 5 to 2 ppm. In comparison with plumbole22, these ligand-coordinated plumboles are not aromatic, whereas in comparison with neutral plumbole, the NICS values of these compounds are small, and we can state that the ligand-coordinated plumboles are located midway between aromatic and antiaromatic compounds. The NICS values decrease in the order of THF < Py < Py2 < THF2 < NHC < NHC2, suggesting that the NICS values correspond to the number of electrons donated by the ligands to lead. The Mulliken population of the 6p orbital of Pb for each complex is in the order of Py < THF2 < THF 5 Py2 < NHC < NHC2. This relationship between the NICS and the Mulliken population is not so clear. UV spectra
Figure 2. Schematic molecular structures of (a) plumbole, (b) plumbole 1 THF, (c) plumbole 1 THF2, (d) plumbole 1 Py, (e) plumbole 1 Py2, (f ) plumbole 1 NHC, and (g) plumbole 1 NHC2, optimized by the ZORAB3LYP method.
The UV spectra of plumbole, plumbole 1 THF2, plumbole 1 Py2, and plumbole 1 NHC2 are shown in Figure 3 as green triangles ~. The calculated excited energies and oscillator strengths are also plotted. The blue symbol 䉬 means that the spin multiplicity of the main configuration is a triplet, and the symbol 3 means a singlet. It should be noted that the pure singlet or triplet state does not exist as the spin-orbit interaction is taken into consideration in our present calculations. The calculated data
Journal of Computational Chemistry 2014, 35, 847–853
849
FULL PAPER
WWW.C-CHEM.ORG
Table 2. Comparison of bond lengths (A˚) and bond angles (degree) of plumbole 1 ligand(s). P 1 THF
CaACb
[a]
Calc.
Calc.(exptl.)
A˚
1.330 1.328 1.514 2.270 2.262 2.427
deg
78.9 89.3 86.5
1.364(1.346) 1.363(1.347) 1.539(1.517) 2.365(2.319) 2.366(2.321) 2.702(2.661) 2.697(2.663) 78.0(78.4) 82.9(80.9) 84.1(82.2) 92.4(95.1) 92.6(95.2) 174.8(175.6)
CbACb PbACa PbAL CaAPbACa CaAPbAL
P 1 THF2
LAPbAL
P 1 Py Calc.(exptl.)
P 1 Py2 [a]
1.348(1.347) 1.350(1.349) 1.517(1.525) 2.309(2.345) 2.310(2.335) 2.466(2.548) 78.8(78.2) 91.8(89.5) 85.1(85.4)
P 1 NHC
P 1 NHC2 [a]
Calc.
Calc.(exptl.)
1.351 1.350 1.527 2.352 2.352 2.758 2.758 78.0 84.3 85.8 85.4 84.0 166.7
1.365(1.363) 1.364(1.351) 1.546(1.520) 2.481(2.290) 2.487(2.292) 2.700(2.431)
Calc. 1.351 1.351 1.525 2.354 2.345 3.825 2.625 78.2 92.2 94.4 88.0 95.9 172.9
75.3(79.2) 96.5(96.5) 96.5(96.8)
[a] Experimental values are observed in the crystalline state.[8]
points and the observed spectra reasonably agree with each other, although our calculations slightly underestimate the excited energies in all the compounds. We suppose that the first absorption peak observed at around 700 nm of plumbole in Figure 3a correspond to the calculated data points located at 1000 nm. The first absorption peak is observed at around 700 nm in plumbole (a), whereas it disappears in the ligand-coordinated plumboles.[8] This trend is reproduced by the present calculations. The disappearance of the lowest transition can be explained by the orbital diagram shown in Figure 4. The first absorption peak of plumbole corresponds to the transition from the lone pair r(Pb) to the virtual p(Pb) orbital, both of which consist of 6p orbitals. In this case, orbital energy gap DE is relatively small, as shown in Figure 4. Conversely, when the ligands coordinate to plumbole, the new energy gap becomes large due to the interaction between the ligand lone pair and the virtual p(Pb) orbital.
no-ligand. This order is essentially parallel to the strength of the Lewis-base ligand. We can estimate the coordination number of THF in solution, comparing the calculated 207Pb-NMR chemical shifts of plumbole, plumbole 1 THF, and plumbole 1 THF2 with the experimental shift. In the crystalline state, we found that the coordination number is two, namely plumbole 1 THF2 is formed. However, in solution, the coordination number of THF is not clear. In the previous paper, we supposed that both plumbole, plumbole 1 THF, and plumbole 1 THF2 exist in THF solution,[8] while the ratio of plumbole 1 THF and plumbole increases in benzene solution by rapid ligand-exchange. Conversely, in plumbole 1 NHC and plumbole 1 NHC2, the calculated 207Pb chemical shift of plumbole 1 NHC2 is close to the experimental value. This proposes that plumbole 1 NHC2 is dominant in solution, although plumbole 1 NHC is formed in the crystalline state.[8] In the 13C-NMR chemical shifts, 13Cb is shifted upfield compared to 13Ca. Actually, in plumbole 1 THF2, the 13Ca and 13Cb chemical shifts (ppm) are as follows.
NMR chemical shifts Table 4 shows the isotropic magnetic shielding tensors and the NMR chemical shifts of 207Pb and 13C. The relativistic and nonrelativistic calculations are compared. The magnetic shielding tensor (rtot) is decomposed into three components: diamagnetic term (rdia), paramagnetic term (rpara), and spin-dependent term (rSO) which is generated from the spin-orbit interaction. The chemical shift (d X) is defined by d X 5 rref – rX, where rref is the value of tetramethylplumbane for 207Pb-NMR spectroscopy and tetramethylsilane for 13C-NMR spectroscopy. First, we check the relativistic effect from the data in Table 4, focusing on plumbole 1 THF2 and plumbole 1 NHC2. The 207PbNMR chemical shifts calculated by the nonrelativistic method are 652 and 297 ppm, whereas those calculated by the ZORA method are 3566 and 1854 ppm, and the observed values are 5539 and 1793 ppm in benzene-d6, respectively. The relativistic effect is significantly large not only on the absolute values r but also on the relative values d. The ZORA method well reproduces the trend of the 207Pb-NMR chemical shifts semiquantitatively. As shown in Table 4, the 207Pb chemical shift becomes downfield in the order of NHC2 > Py2 > NHC > Py > THF2 > THF >> 850
Journal of Computational Chemistry 2014, 35, 847–853
13
Ca Cb
13
NonR
ZORA
Exptl.
170 181
226 193
223 179
The experimental trend can be reproduced quantitatively by the ZORA method, whereas the nonrelativistic method gives the opposite result. The trend in the 13Ca and 13Cb chemical Table 3. Comparison of NICS (ppm) values of plumbole 1 ligand. Nonrelativistic
Plumbole Plumbole 1 Plumbole 1 Plumbole 1 Plumbole 1 Plumbole 1 Plumbole 1 Plumbole22
THF THF2 Py Py2 NHC NHC2
Relativistic
NICS(0)
NICS(1)
NICS(0)
10.0 5.1 4.2 4.0 3.6 3.0 1.4 26.6
3.4 0.8 0.6 0.5 1.1 0.5 20.9 27.1
13.1 6.0 5.0 5.1 5.1 3.6 2.6 25.1
NICS(1)
WWW.CHEMISTRYVIEWS.COM
4.2 0.9 1.0 1.0 1.3 0.8 0.1 27.3
Mulliken population 6p(Pb) 1.64 1.63 1.56 1.54 1.63 1.73 1.77 2.34
WWW.C-CHEM.ORG
FULL PAPER
Figure 3. Comparison of calculated and observed UV spectra of plumbole, plumbole 1 THF2, plumbole 1 Py2, and plumbole 1 NHC2, The calculation was carried out by ZORA-TD-DFT using the VMW functional. Symbol 䉬 means that the spin multiplicity of the main configuration is a triplet; symbol 3 means a singlet; and symbol ~ represents the observed spectra (Ref. [8]).
shifts of plumbole 1 Py and plumbole 1 Py2 is essentially the rSO terms of 13Ca and 13Cb are dependent on the ligands. This same as plumbole-THF and plumbole-THF2. suggests that these terms are controlled by the electron donation from the ligand to lead. In our previous calculation of acetylene iodine, IA The rSO term is the key point of the deshielding and shieldCaBCbAH, the spin-dependent term (rSO) of 13Ca is shifted 13 1 upfield, whereas that of Cb is shifted downfield, and H is ing of neighboring nuclei around lead. Figure 5 shows the correlation between 13Ca and 13Cb chemical shifts and the again shifted upfield.[35] The same trend is noted in ethylene iodine H2C@CHI; namely, the nearest-neighbor nucleus relative magnetically allowed transition energies shown in Figure 4. In the second-order perturbation theory, the rSO term primarily to the heavy atom is more magnetically shielded by the spindependent term, and the next nearest neighbor nucleus is depends on the reciprocal of the transition energy, assuming deshielded. Conversely, in the present compounds, the trend is that the Fermi-contact integral and so on are constant in any opposite to that observed in IACaBCbAH; namely, Ca is more of the compounds in this series. As expected, in the Ca deshielded than Cb. Focusing on the component analysis, in plumbole 1 THF2, the downfield shift of 13Ca is caused by the rSO term, whereas the rdia and rpara terms are almost the same in 13Ca and 13Cb. In plumbole 1 NHC2, the downfield shift of 13Ca arises from the rpara and rSO terms, whereas the rdia term is almost the same in 13 Ca and 13Cb. In plumbole, the rSO term becomes largely negative, whereas it becomes small in Figure 4. Relationship between plumbole and ligand orbitals, and HOMO-LUMO gaps (DE) of plumbole and ligandcoordinated plumboles. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.] plumbole22. The rpara and Journal of Computational Chemistry 2014, 35, 847–853
851
FULL PAPER
Table 4.
207
Pb and
WWW.C-CHEM.ORG
13
C magnetic shielding tensors and chemical shifts (ppm). Relativistic
Nonrelativistic dia
r Plumbole 1 THF
Plumbole 1 THF2
Plumbole 1 Py
Plumbole 1 Py2
Plumbole 1 NHC
Plumbole 1 NHC2
Plumbole
Plumbole22
TMPb TMS
para
r
r
tot
d
calc
dia
r
para
r
rso
rtot
dcalc
dexptl[a] 4601[b] 5539[c] 219[b] 227[c] 181[c] -[d] -[d] -[d]
Pb
10074
25532
4543
1108
9946
26856
903
3993
3885
Ca
257
2231
26
155
256
2240
243
228
210
Cb Pb Ca Cb Pb Ca Cb Pb Ca Cb Pb Ca Cb Pb Ca Cb Pb Ca Cb Pb Ca Cb Pb C
252 10074 257 252 10074 256 252 10074 256 252 10073 256 252 10073.0 256.0 251.6 10074 257 252 10078 265 251 10078 258
2245 25075 2246 2252 25228 2249 2247 2466 2258 2246 24736 2261 2248 24325 2253 2245 28813 2220 2243 25083 2240 2212 24427 276
7 4998 12 0 4845 7 5 5412 22 7 5338 25 4 5748 3 7 1262 36 9 4995 16 38 5650 182
174 652 170 181 805 175 176 239 183 175 313 187 178 298 178 175 4389 145 172 656 166 144
252 9945 256 252 9944 259 252 9946 255 253 9945 255 252 9945 255 252 9946 256 252 9949 255 250 9950 257
2244 26523 2257 2251 26570 2261 2246 26027 2269 2245 26212 2273 2248 25616 2263 2244 29556 2233 2247 26661 2258 2212 25333 276
212 890 243 213 1073 247 212 1333 233 211 1300 230 213 1695 224 210 21532 281 230 3004 216 217 3260 1
24 4312 244 211 4448 249 26 5251 247 23 5033 247 28 6025 233 22 21142 258 225 6293 219 22 7878 182
186 3566 226 193 3430 231 188 2627 229 185 2845 229 191 1854 215 184 9020 240 207 15865 202 160
217 178 -[d] -[d] 1793[c] 202[c] 194[c] -[d] -[d] -[d]
[a] Ref. [8]. [b] Measured in THF. [c] Measured in benzene-d6. [d] As the coordination number of plumboles in solution is not clear, the observed chemical shifts can be assigned clearly to neither plumbole 1 L nor plumbole 1 L2 (L 5 THF, Py, NHC).
chemical shifts, all data fall around one line, whereas in the Cb chemical shifts, the data are independent of rSO. This suggests that the rSO term of Ca is controlled by the interaction between the ligand and Pb, whereas the rSO term of Cb seems to be controlled by several mechanisms related to the spin transfer.
ligand-coordinated plumboles did not have aromaticity, the NICS values indicated that they were located midway between
Nuclear spin–spin coupling constants The calculated and observed J [Pb,C] are shown in Table 5. In plumbole 1 THF2, the J values calculated by the nonrelativistic DFT and the ZORA-DFT methods are 2395 and 21164.9 Hz, respectively, whereas the observed value is 1049 Hz. The plus– minus sign of observed J values in Table 5 should be corrected referring to the calculated values. The calculated result is significantly improved by the relativistic effect. The relativistic correction is approximately 60%. The sign of J[Pb,C] is determined to be negative in our calculations. Unfortunately, the calculated results cannot reproduce the difference between plumbole 1 Py2 and plumbole 1 NHC2
Conclusions We performed calculations for a series of lead-containing plumboles and ligand-coordinated plumboles using the ZORADFT method and the ZORA-TD-DFT method with B3LYP. Although, from the viewpoint of molecular structure, the 852
Journal of Computational Chemistry 2014, 35, 847–853
Figure 5. Correlation between rSO terms of 13Ca and 13Cb chemical shifts and reciprocal of the magnetically allowed transition energies in plumbole and ligand-coordinated plumboles. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]
WWW.CHEMISTRYVIEWS.COM
FULL PAPER
WWW.C-CHEM.ORG
Table 5. Nuclear spin–spin coupling constants (J[Pb,C]) of plumbole 1 ligand (Hz). Nonrelativistic Hamiltonian
Plumbole Plumbole Plumbole Plumbole Plumbole Plumbole
1 1 1 1 1 1
THF THF2 Py Py2 NHC NHC2
ZORA
dia
para
FC
tot.
dia
para
FC 1 SO
tot.
exptl.[a]
0.3 0.2 0.1 0.2 0.1 0.2
9.9 5.1 21.7 1.9 25.9 23.8
2358.6 2400.3 2283.3 2346.2 2330.1 2298.5
2348.5 2395.1 2284.9 2344.2 2335.9 2302.2
0.3 0.2 0.1 0.2 0.1 0.2
8.5 6.5 24.5 1.5 29.8 27.4
21495.8 21171.6 21106.1 21107.8 21106.4 21098.6
21487.1 21164.9 21110.4 21106.2 21116.0 21105.9
1049 1049 735 735 754 754
[a] Ref. [8].
aromatic and antiaromatic compounds. In the UV spectra, the present method reproduced the trends of plumbole and ligand-coordinated plumboles. The lowest excited state of plumbole disappeared in the ligand-coordinated plumboles due to the interaction between the 6pp orbital of Pb and the lone-pair orbital of the ligand. The trends of 207Pb- and 13CNMR chemical shifts and J[Pb,C] were also well-reproduced by the present method. The relativistic effect was essential for those properties. The nearest-neighbor Ca from Pb was more shielded than Cb. This trend was opposite to the previous heavy-metal effect observed in IACaBCbAH. We discussed this point based on the results of decomposition into rSO, rdia, and rpara terms, and the electron transition energies. Keywords: lead Pb NMR ZORA chemical shift nuclear spin–spin coupling plumbole relativistic effect aromaticity five-membered ring
How to cite this article: T. Kawamura, M. Abe, M. Saito, M. Hada. J. Comput. Chem. 2014, 35, 847–853. DOI: 10.1002/ jcc.23556
[1] M. Saito, Coord. Chem. Rev. 2012, 256, 627. [2] M. Saito, R. Haga, M. Yoshioka, K. Ishimura, S. Nagase, Angew. Chem. Int. Ed. 2005, 44, 6553. [3] M. Saito, M. Yoshioka, Coord. Chem. Rev. 2005, 249, 765. [4] M. Karni, Y. Apeloig, J. Kapp, P. von R. Schleyer, In The Chemistry of Organic Silicon Compounds, Vol. 3, Chapter 1; Z. Rappoport, Y. Apeloig, Eds.; Wiley: Chichester, 2001; pp. 1–163. [5] M. Saito, M. Sakaguchi, T. Tajima, K. Ishimura, S. Nagase, M. Hada, Science 2010, 328, 339. [6] E. H€ uckel, Z. Phys. 1931, 70, 204. [7] P. v. R. Schleyer, C. Maerker, A. Dransfeld, H. Jiao, N. J. R. v. E. Hommes, J. Am. Chem. Soc. 1996, 118, 6317. [8] M. Saito, T. Akiba, M. Kaneko, T. Kawamura, M. Abe, M. Hada, M. Minoura, Chem. Eur. J. 2013, 19, 16946. [9] B. Goldfuss, P. von R. Schleyer, F. Hampel, Organometallics 1996, 15, 1755. [10] B. Goldfuss, P. von R. Schleyer, Organometallics 1997, 16, 1543. [11] H. Nakatsuji, K. Kanda, K. Endo, T. Yonezawa, J. Am. Chem. Soc. 106, 4653, 1984 [12] H. Nakatsuji, In Nuclear Magnetic Shieldings and Molecular Structure; J. A. Tossell, Ed.; Kluwer: Dortrecht, 1993; p. 263. [13] H. Nakatsuji, H. Takashima, M. Hada, Chem. Phys. Lett. 1995, 233, 95.
[14] H. Kaneko, M. Hada, T. Nakajima, H. Nakatsuji, Chem. Phys. Lett. 1996, 261, 1. [15] M. Hada, H. Kaneko, H. Nakatsuji, Chem. Phys. Lett. 1996, 261, 7. [16] M. Hada, R. Fukuda, H. Nakatsuji, Chem. Phys. Lett. 2000, 321, 452. [17] R. Fukuda, M. Hada, H. Nakatsuji, J. Chem. Phys. 2003, 118, 1015. [18] J. Seino, M. Hada, J. Chem. Phys. 2010, 132, 174105. [19] W. Nakanishi, S. Hayashi, K. Narahara, D. Yamaki, M. Hada, Chem. Eur. J. 2008, 14, 7278. [20] V. G. Malkin, O. L. Malkina, D. R. Salahub, Chem. Phys. Lett. 1996, 261, 335. [21] E. v. Lenthe, A. E. Ehlers, E. J. Baerends, J. Chem. Phys. 1999, 110, 8943. [22] E. v. Lenthe, E. J. Baerends, J. G. Snijders, J. Chem. Phys. 1993, 99, 4597. [23] E. v. Lenthe, E. J. Baerends, J. G. Snijders, J. Chem. Phys. 1994, 101, 9783. [24] E. v. Lenthe, J. G. Snijders, E. J. Baerends, J. Chem. Phys. 1996, 105, 6505. [25] E. v. Lenthe, R. v. Leeuwen, E. J. Baerends, J. G. Snijders, Int. J. Quantum Chem. 1996, 57, 281. [26] S. K. Wolff, T. Ziegler, E. van Lenthe, E. J. Baerends, J. Chem. Phys. 1999, 110, 7689. [27] J. Autschbach, T. Ziegler, J. Chem. Phys. 2000, 113, 936. [28] J. Autschbach, T. Ziegler, J. Chem. Phys. 2000, 113, 9410. [29] C. Lee, W. Yang, R. G. Parr, Phys. Rev. B 1988, 37, 785. [30] A. D. Becke, J. Chem. Phys. 1993, 98, 5648. [31] S. H. Vosko, L. Wilk, M. Nusair, Can. J. Phys. 1980, 58, 1200. [32] E. J. Baerends, T. Ziegler, J. Autschbach, D. Bashford, A. B erces, F. M. Bickelhaupt, C. Bo, P. M. Boerrigter, L. Cavallo, D. P. Chong, L. Deng, R. M. Dickson, D. E. Ellis, M. v. Faassen, L. Fan, T. H. Fischer, C. Fonseca Guerra, A. Ghysels, A. Giammona, S. J. A. v. Gisbergen, A. W. G€ otz, J. A. Groeneveld, O. V. Gritsenko, M. Gr€ uning, S. Gusarov, F. E. Harris, P. v. d. Hoek, C. R. Jacob, H. Jacobsen, L. Jensen, J. W. Kaminski, G. van Kessel, F. Kootstra, A. Kovalenko, M. V. Krykunov, E. v. Lenthe, D. A. McCormack, A. Michalak, M. Mitoraj, J. Neugebauer, V. P. Nicu, L. Noodleman, V. P. Osinga, S. Patchkovskii, P. H. T. Philipsen, D. Post, C. C. Pye, W. Ravenek, J. I. Rodrıguez, P. Ros, P. R. T. Schipper, G. Schreckenbach, J. S. Seldenthuis, M. Seth, J. G. Snijders, M. Sola, M. Swart, D. Swerhone, G. te Velde, P. Vernooijs, L. Versluis, L. Visscher, O. Visser, F. Wang, T. A. Wesolowski, E. M. v. Wezenbeek, G. Wiesenekker, S. K. Wolff, T. K. Woo, A. L. Yakovlev, ADF2012 SCM Theoretical Chemistry; Vrije Universiteit: Amsterdam, The Netherlands, 2012. [33] H. Tanimura, Y. Honda, K. Sugiura, M. Hada, Bull. Chem. Soc. Jpn. 2011, 84, 845. [34] M. Z. Kassaee, S. Arshadi, M. Acedy, E. Vessally, J. Organomet. Chem. 2005, 690, 3427. [35] S. Fukawa, M. Hada, R. Fukuda, S. Tanaka, H. Nakatsuji, J. Comput. Chem. 2001, 22, 528.
Received: 21 November 2013 Revised: 23 January 2014 Accepted: 24 January 2014 Published online on 7 February 2014
Journal of Computational Chemistry 2014, 35, 847–853
853
WWW.C-CHEM.ORG
Quantum-Chemical Analyses of Aromaticity, UV Spectra, and NMR Chemical Shifts in Plumbacyclopentadienylidenes Stabilized by Lewis Bases Toshiaki Kawamura,[a,b] Minori Abe,[a,b] Masaichi Saito,[c] and Masahiko Hada*[a,b] We carried out a series of zeroth-order regular approximation (ZORA)-density functional theory (DFT) and ZORA-timedependent (TD)-DFT calculations for molecular geometries, NMR chemical shifts, nucleus-independent chemical shifts (NICS), and electronic transition energies of plumbacyclopentadienylidenes stabilized by several Lewis bases, (Ph)2(tBuMe2Si)2C4PbL1L2 (L1, L2 5 tetrahydrofuran, Pyridine, N-heterocyclic carbene), and their model molecules. We mainly discussed the Lewis-base effect on the aromaticity of these complexes. The NICS was used to examine the aromaticity. The NICS values showed that the aromaticity of these complexes increases when the donation from the Lewis bases to Pb becomes large.
This trend seems to be reasonable when the 4n-Huckel rule is applied to the fractional p-electron number. The calculated 13 C- and 207Pb-NMR chemical shifts and the calculated UV transition energies reasonably reproduced the experimental trends. We found a specific relationship between the 13C-NMR chemical shifts and the transition energies. As we expected, the relativistic effect was essential to reproduce a trend not only in the 207Pb-NMR chemical shifts and J[Pb-C] but also in the 13C-NMR chemical shifts of carbons adjacent to the lead C 2014 Wiley Periodicals, Inc. atom. V
Introduction
There are prior theoretical investigations on aromaticity and magnetic properties of dilithioplumbole analogs. Magnetic criteria including NICS of Silole dianions and metaloles containing Si, Ge, Sn, and Pb have been reported.[9,10] Pioneering theoretical calculations of metal NMR[11,12] were helpful for the present analyses. The relativistic correction, especially the spinorbit effect, is essentially required in heavy-element NMR chemical shifts.[13–20] We show herein the quantum-chemical calculations of several Lewis-base-stabilized plumbacyclopentadienylidenes. This work has two objectives. The first one is the verification of aromaticity. The p-orbitals of neutral plumbacyclopentadienylidenes are occupied by four electrons, whereas those of plumbole dianions are occupied by six electrons. The latter is suggested to have aromaticity.[5] We here discuss the aromaticity of neutral plumbacyclopentadienylidenes coordinated by Lewis-base ligands. A ligand donates electrons to an unoccupied p-orbital of a neutral plumbacyclopentadienylidene, which mainly consists of the 6p orbital of lead. In this case, it
Aromatic compounds have been studied for a long time principally in organic and organometallic chemistry. Although aromatic compounds have a variety of ring sizes and shapes, most of them consist of carbon, nitrogen, and oxygen atoms. Recently, we have shifted our interest to aromatic compounds containing heavy elements, such as silicon, germanium, tin, and lead, instead of carbon.[1–3] Also, theoretical aspect of silicon compounds have been already summarized.[4] Dilithioplumbole, a lead-containing five-membered ring, was reported in 2010.[5] Dilithioplumbole is an analog of dilithiocyclopentadiene CH4CLi2; a carbon atom is replaced with a lead atom, and therefore, its molecular framework is a planar cyclic form of C4H4Pb. Two lithium atoms are solvated in solution and in the crystalline state. The p-orbitals of C4H4Pb are occupied by six p-electrons as the two additional electrons are supplied by lithium atoms to the plumbole’s p-system. The H€ uckel’s (4n 1 2) p-electron rule[6] is formally satisfied in this compound, and thus, it strongly suggests that dilithioplumbole has aromaticity. Actually, the disappearance of CAC alternation is observed in this compound. The nucleus-independent chemical shift (NICS)[7] values also indicate that dilithioplumbole is an aromatic compound. Conversely, plumbacyclopentadienylidenes, where the lead atoms have divalent states, have p-orbitals containing four electrons, and it is not aromatic according to the H€ uckel’s (4n 1 2) p-electron rule. Recently, plumbacyclopentadienylidenes, stabilized by tetrahydrofuran (THF), pyridine, and N-heterocyclic carbene (NHC), were synthesized and characterized.[8] In these compounds, the total number of pelectrons is fractional, and therefore, the H€ uckel’s (4n 1 2) pelectron rule seems to be partially satisfied.
DOI: 10.1002/jcc.23556
[a] T. Kawamura, M. Abe, M. Hada Department of Chemistry, Graduate School of Science and Engineering, Tokyo Metropolitan University, Minami-Osawa 1-1, Hachi-Oji, Tokyo 1920397, Japan E-mail: [email protected] [b] T. Kawamura, M. Abe, M. Hada JST, CREST, 4-1-8 Honcho, Kawaguchi, Saitama 332-0012, Japan [c] M. Saito Department of Chemistry, Graduate School of Science and Engineering, Saitama University, Shimo-okubo, Sakura-ku, Saitama 338-8570, Japan Contract grant sponsor: JST, CREST; Contract grant sponsor: MEXT [M.S. (Grant-in-Aids for Scientific Research)]; Contract grant number: 25109510 C 2014 Wiley Periodicals, Inc. V
Journal of Computational Chemistry 2014, 35, 847–853
847
FULL PAPER
WWW.C-CHEM.ORG
seems that the H€ uckel’s (4n 1 2) p-electron rule is satisfied, although two p-electrons are shared with the plumbole and the ligand. We are interested in whether the ligandcoordinated plumbacyclopentadienylidenes have aromaticity or not. To check aromaticity theoretically, we calculate NICS values. For comparison, we also performed calculations for plumbole dianion to which two electrons are supplied by the two lithium ions, and therefore, the H€ uckel’s (4n 1 2) pelectron rule is completely satisfied. The second objective is to discuss the trend of the NMR chemical shifts and the nuclear spin–spin coupling constants J[Pb,C], and the electronic excitation energies of several ligand-coordinated plumbacyclopentadienylidenes. We examine the relationship between NMR chemical shifts and excitation energies, and discuss the ligand effect. We also discuss the relativistic effect of Pb on the NMR chemical shifts and J[Pb,C]. We note here as follows to avoid a confusion. Plumbole is generally used as a lead analog of cyclopentadiene with a tetravalent lead atom. However, in the following sections, we use the terminology, plumbole, as an abbreviation for plumbacyclopentadienylidene.
Calculation Method Figure 1 shows the molecular formulae of plumboles stabilized by Lewis-base ligands, which are calculated in the present investigation. In this article, plumbole is always defined as (SiC6H15)(C6H5)2AC4Pb. Three kinds of ligands, THF (C4H8O), pyridine (Py, C5H5N), and ((C3H7)2(CH3)2), are taken into account, as shown in Figure 1. We use the abbreviations listed below.
Plumbole 1 THF (P1THF) Plunbole 1 THF2(P1THF2) Plumbole 1 Py (P1Py) Plumbole 1 Py2 (P1Py2) Plumbole 1 NHC(P1NHC) Plumbole 1 NHC2(P1NHC2)
L1
L2
THF THF Py Py NHC NHC
THF Py NHC
THF is the weakest Lewis-base of the three ligands and NHC is the strongest. We perform calculations for six compounds (plumbole 1 THF, plumbole 1 THF2, plumbole 1 Py, plumbole 1 Py2, plumbole 1 NHC, and plumbole 1 NHC2). Plumbole 1 THF2, plumbole 1 Py2, and plumbole 1 NHC can be recrystallized, and thus, their molecular structures are determined experimentally.[8] The other three compounds cannot be recrystallized; nevertheless, we perform all calculations of them because they may exist in solution. As lead is a heavy element and its relativistic effect may be considerably large, we use the zeroth-order regular approximation (ZORA).[21–25] For comparison, we also perform a nonrelativistic Hamiltonian. Geometry optimizations, the 207Pb and 13C magnetic shielding tensors,[26] J[Pb,C],[27,28] and NICS values are calculated by ZORA-B3LYP[29,30] with the GIAO approach and the collinear functional approach. The basis functions used are TZ2P for Pb, TZP for C in a five-membered ring and 848
Journal of Computational Chemistry 2014, 35, 847–853
Figure 1. Molecular formulae of ligand-coordinated plumboles.
atoms adjacent to Pb, DZ for H, and DZP for the other atoms. Electronic excitation energies corresponding to the UV spectra are obtained by ZORA-TD-DFT, the functional of which is VWN,[31] using the same basis sets as those for the geometry optimization. All calculations are performed with ADF 2012.01a program.[32]
Result and Discussion Aromaticity Although the definition of aromaticity is not clear-cut, we can immediately show the following four criteria. The first criterion is the conventional H€ uckel’s (4n 1 2) p-electron rule. The second one is the planar ring and the disappearance of CAC bond alternation. The third one is the negative NICS value. The last one is ring stability, namely, a large heat of formation. In this discussion, the H€ uckel’s (4n 1 2) p-electron rule is not applicable as the number of p-electrons cannot be defined with certainty in the ligand-coordinated plumboles. In the quantum-chemical calculations, the second and third criteria are applicable. These criteria were used in our previous investigation.[33] In the following sections, NICS(0) and NICS(1) are used; (0) means that NICS is calculated on the ring plane of a ˚ above the molecule; (1) means that NICS is calculated on 1 A ring plane. To demonstrate and confirm the disappearance of the CAC bond alternation and the negative NICS values in a series of aromatic five-membered analogs, we perform calculations for simple five-membered rings C4H4X and C4H4X22 (X 5 Si, Ge, Si, Pb). In those compounds, we can apply the H€ uckel’s (4n 1 2) p-electron rule. The neutral compound C4H4X has been reported previously.[34] In our calculations, the lone-pair orbital of X in C4H4X is always occupied by two electrons, and therefore, C4H4X has four p-electrons, whereas C4H4X22 has six pelectrons. Table 1 shows that both the disappearance of the bond alternation and the negative NICS values are effectively equivalent in the aromaticity/nonaromaticity of five-membered C4H4X and C4H4X22 (X 5 Si, Ge, Si, Pb) rings. Actually, in neutral C4H4X, the bond lengths of CaACb and CbACb are 1.3 and WWW.CHEMISTRYVIEWS.COM
FULL PAPER
WWW.C-CHEM.ORG
Table 1. Comparison of bond lengths and NICS values of C4H4 X and C4H4X22 (X 5 Si, Ge, Sn, and Pb).
C4H4Si C4H4Ge C4H4Sn C4H4Pb C4H4Si22 C4H4Ge22 C4H4Sn22 C4H4Pb22
CaACb
CbACb
CaAX
NICS(0)
NICS(1)
1.333 – 1.330 1.329 1.431 1.422 1.416 1.409
1.519 – 1.520 1.519 1.385 1.389 1.396 1.409
1.883 – 2.183 2.247 1.825 1.936 2.119 2.183
14.0 – 10.2 9.8 27.1 26.6 27.1 26.2
3.8 – 2.2 2.7 28.1 28.0 28.0 27.3
˚ , respectively, so obvious bond alternations appear. In dia1.5 A nion C4H4X22, the bond lengths of both CaACb and CbACb ˚ . The disappearance of the bond alternation is are around 1.4 A a clear indication of the aromaticity of those compounds. Conversely, both NICS(0) and NICS(1) values of C4H4X are positive, indicating that they are not aromatic molecules. Those of C4H4X22 are negative, indicating that C4H4X22 has aromaticity. Figure 2 shows the schematic molecular structures of plumbole and six types of ligand-coordinated plumbole: plumbole 1 THF, plumbole 1 THF2, plumbole 1 Py, plumbole 1 Py2, plumbole 1 NHC, and plumbole 1 NHC2, optimized by the ZORA-B3LYP method. Table 2 lists the important geometrical parameters for the compounds in Figure 2. The ligands are always located on the Pb atom, and are almost perpendicular to the C4Pb ring. The CaACb bond lengths of plumbole 1 ˚ , respectively, whereas the experiTHF2 are 1.364 and 1.363 A
˚ .[8] Conversely, the CbACb mental ones are 1.346 and 1.347 A ˚ bond length is 1.539 A, whereas the experimental one is 1.517 ˚ . The calculated and experimental CAC bond lengths agree A well and bond alternation occurs. This trend is also found in the other ligand-coordinated compounds. Therefore, based on the criterion of bond alternation, the ligand-coordinated plumboles should not be aromatic. To confirm the third criterion, namely, the negative NICS value, we show the NICS values in Table 3. Those values are calculated by conventional nonrelativistic and ZORA Hamiltonians. Although the relativistic effect on NICS is quite small, we use the relativistic results in the following discussion. As already discussed using the data in Table 1, the NICS(0) value of neutral plumbole, which has four p-electrons, is 13.1 ppm, and therefore, neutral plumbole without ligands is not aromatic but antiaromatic. Conversely, the NICS value of plumbole22, which has six p-electrons, is 25.1 ppm, suggesting that plumbole22 has aromaticity. Note that CH4Pb in Table 1 is essentially the same as plumbole in Table 3 and Figure 2, and the difference appears on the side chain of the fivemembered ring. Conversely, the NICS(0) values of the ligandcoordinated plumboles range from 5 to 2 ppm. In comparison with plumbole22, these ligand-coordinated plumboles are not aromatic, whereas in comparison with neutral plumbole, the NICS values of these compounds are small, and we can state that the ligand-coordinated plumboles are located midway between aromatic and antiaromatic compounds. The NICS values decrease in the order of THF < Py < Py2 < THF2 < NHC < NHC2, suggesting that the NICS values correspond to the number of electrons donated by the ligands to lead. The Mulliken population of the 6p orbital of Pb for each complex is in the order of Py < THF2 < THF 5 Py2 < NHC < NHC2. This relationship between the NICS and the Mulliken population is not so clear. UV spectra
Figure 2. Schematic molecular structures of (a) plumbole, (b) plumbole 1 THF, (c) plumbole 1 THF2, (d) plumbole 1 Py, (e) plumbole 1 Py2, (f ) plumbole 1 NHC, and (g) plumbole 1 NHC2, optimized by the ZORAB3LYP method.
The UV spectra of plumbole, plumbole 1 THF2, plumbole 1 Py2, and plumbole 1 NHC2 are shown in Figure 3 as green triangles ~. The calculated excited energies and oscillator strengths are also plotted. The blue symbol 䉬 means that the spin multiplicity of the main configuration is a triplet, and the symbol 3 means a singlet. It should be noted that the pure singlet or triplet state does not exist as the spin-orbit interaction is taken into consideration in our present calculations. The calculated data
Journal of Computational Chemistry 2014, 35, 847–853
849
FULL PAPER
WWW.C-CHEM.ORG
Table 2. Comparison of bond lengths (A˚) and bond angles (degree) of plumbole 1 ligand(s). P 1 THF
CaACb
[a]
Calc.
Calc.(exptl.)
A˚
1.330 1.328 1.514 2.270 2.262 2.427
deg
78.9 89.3 86.5
1.364(1.346) 1.363(1.347) 1.539(1.517) 2.365(2.319) 2.366(2.321) 2.702(2.661) 2.697(2.663) 78.0(78.4) 82.9(80.9) 84.1(82.2) 92.4(95.1) 92.6(95.2) 174.8(175.6)
CbACb PbACa PbAL CaAPbACa CaAPbAL
P 1 THF2
LAPbAL
P 1 Py Calc.(exptl.)
P 1 Py2 [a]
1.348(1.347) 1.350(1.349) 1.517(1.525) 2.309(2.345) 2.310(2.335) 2.466(2.548) 78.8(78.2) 91.8(89.5) 85.1(85.4)
P 1 NHC
P 1 NHC2 [a]
Calc.
Calc.(exptl.)
1.351 1.350 1.527 2.352 2.352 2.758 2.758 78.0 84.3 85.8 85.4 84.0 166.7
1.365(1.363) 1.364(1.351) 1.546(1.520) 2.481(2.290) 2.487(2.292) 2.700(2.431)
Calc. 1.351 1.351 1.525 2.354 2.345 3.825 2.625 78.2 92.2 94.4 88.0 95.9 172.9
75.3(79.2) 96.5(96.5) 96.5(96.8)
[a] Experimental values are observed in the crystalline state.[8]
points and the observed spectra reasonably agree with each other, although our calculations slightly underestimate the excited energies in all the compounds. We suppose that the first absorption peak observed at around 700 nm of plumbole in Figure 3a correspond to the calculated data points located at 1000 nm. The first absorption peak is observed at around 700 nm in plumbole (a), whereas it disappears in the ligand-coordinated plumboles.[8] This trend is reproduced by the present calculations. The disappearance of the lowest transition can be explained by the orbital diagram shown in Figure 4. The first absorption peak of plumbole corresponds to the transition from the lone pair r(Pb) to the virtual p(Pb) orbital, both of which consist of 6p orbitals. In this case, orbital energy gap DE is relatively small, as shown in Figure 4. Conversely, when the ligands coordinate to plumbole, the new energy gap becomes large due to the interaction between the ligand lone pair and the virtual p(Pb) orbital.
no-ligand. This order is essentially parallel to the strength of the Lewis-base ligand. We can estimate the coordination number of THF in solution, comparing the calculated 207Pb-NMR chemical shifts of plumbole, plumbole 1 THF, and plumbole 1 THF2 with the experimental shift. In the crystalline state, we found that the coordination number is two, namely plumbole 1 THF2 is formed. However, in solution, the coordination number of THF is not clear. In the previous paper, we supposed that both plumbole, plumbole 1 THF, and plumbole 1 THF2 exist in THF solution,[8] while the ratio of plumbole 1 THF and plumbole increases in benzene solution by rapid ligand-exchange. Conversely, in plumbole 1 NHC and plumbole 1 NHC2, the calculated 207Pb chemical shift of plumbole 1 NHC2 is close to the experimental value. This proposes that plumbole 1 NHC2 is dominant in solution, although plumbole 1 NHC is formed in the crystalline state.[8] In the 13C-NMR chemical shifts, 13Cb is shifted upfield compared to 13Ca. Actually, in plumbole 1 THF2, the 13Ca and 13Cb chemical shifts (ppm) are as follows.
NMR chemical shifts Table 4 shows the isotropic magnetic shielding tensors and the NMR chemical shifts of 207Pb and 13C. The relativistic and nonrelativistic calculations are compared. The magnetic shielding tensor (rtot) is decomposed into three components: diamagnetic term (rdia), paramagnetic term (rpara), and spin-dependent term (rSO) which is generated from the spin-orbit interaction. The chemical shift (d X) is defined by d X 5 rref – rX, where rref is the value of tetramethylplumbane for 207Pb-NMR spectroscopy and tetramethylsilane for 13C-NMR spectroscopy. First, we check the relativistic effect from the data in Table 4, focusing on plumbole 1 THF2 and plumbole 1 NHC2. The 207PbNMR chemical shifts calculated by the nonrelativistic method are 652 and 297 ppm, whereas those calculated by the ZORA method are 3566 and 1854 ppm, and the observed values are 5539 and 1793 ppm in benzene-d6, respectively. The relativistic effect is significantly large not only on the absolute values r but also on the relative values d. The ZORA method well reproduces the trend of the 207Pb-NMR chemical shifts semiquantitatively. As shown in Table 4, the 207Pb chemical shift becomes downfield in the order of NHC2 > Py2 > NHC > Py > THF2 > THF >> 850
Journal of Computational Chemistry 2014, 35, 847–853
13
Ca Cb
13
NonR
ZORA
Exptl.
170 181
226 193
223 179
The experimental trend can be reproduced quantitatively by the ZORA method, whereas the nonrelativistic method gives the opposite result. The trend in the 13Ca and 13Cb chemical Table 3. Comparison of NICS (ppm) values of plumbole 1 ligand. Nonrelativistic
Plumbole Plumbole 1 Plumbole 1 Plumbole 1 Plumbole 1 Plumbole 1 Plumbole 1 Plumbole22
THF THF2 Py Py2 NHC NHC2
Relativistic
NICS(0)
NICS(1)
NICS(0)
10.0 5.1 4.2 4.0 3.6 3.0 1.4 26.6
3.4 0.8 0.6 0.5 1.1 0.5 20.9 27.1
13.1 6.0 5.0 5.1 5.1 3.6 2.6 25.1
NICS(1)
WWW.CHEMISTRYVIEWS.COM
4.2 0.9 1.0 1.0 1.3 0.8 0.1 27.3
Mulliken population 6p(Pb) 1.64 1.63 1.56 1.54 1.63 1.73 1.77 2.34
WWW.C-CHEM.ORG
FULL PAPER
Figure 3. Comparison of calculated and observed UV spectra of plumbole, plumbole 1 THF2, plumbole 1 Py2, and plumbole 1 NHC2, The calculation was carried out by ZORA-TD-DFT using the VMW functional. Symbol 䉬 means that the spin multiplicity of the main configuration is a triplet; symbol 3 means a singlet; and symbol ~ represents the observed spectra (Ref. [8]).
shifts of plumbole 1 Py and plumbole 1 Py2 is essentially the rSO terms of 13Ca and 13Cb are dependent on the ligands. This same as plumbole-THF and plumbole-THF2. suggests that these terms are controlled by the electron donation from the ligand to lead. In our previous calculation of acetylene iodine, IA The rSO term is the key point of the deshielding and shieldCaBCbAH, the spin-dependent term (rSO) of 13Ca is shifted 13 1 upfield, whereas that of Cb is shifted downfield, and H is ing of neighboring nuclei around lead. Figure 5 shows the correlation between 13Ca and 13Cb chemical shifts and the again shifted upfield.[35] The same trend is noted in ethylene iodine H2C@CHI; namely, the nearest-neighbor nucleus relative magnetically allowed transition energies shown in Figure 4. In the second-order perturbation theory, the rSO term primarily to the heavy atom is more magnetically shielded by the spindependent term, and the next nearest neighbor nucleus is depends on the reciprocal of the transition energy, assuming deshielded. Conversely, in the present compounds, the trend is that the Fermi-contact integral and so on are constant in any opposite to that observed in IACaBCbAH; namely, Ca is more of the compounds in this series. As expected, in the Ca deshielded than Cb. Focusing on the component analysis, in plumbole 1 THF2, the downfield shift of 13Ca is caused by the rSO term, whereas the rdia and rpara terms are almost the same in 13Ca and 13Cb. In plumbole 1 NHC2, the downfield shift of 13Ca arises from the rpara and rSO terms, whereas the rdia term is almost the same in 13 Ca and 13Cb. In plumbole, the rSO term becomes largely negative, whereas it becomes small in Figure 4. Relationship between plumbole and ligand orbitals, and HOMO-LUMO gaps (DE) of plumbole and ligandcoordinated plumboles. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.] plumbole22. The rpara and Journal of Computational Chemistry 2014, 35, 847–853
851
FULL PAPER
Table 4.
207
Pb and
WWW.C-CHEM.ORG
13
C magnetic shielding tensors and chemical shifts (ppm). Relativistic
Nonrelativistic dia
r Plumbole 1 THF
Plumbole 1 THF2
Plumbole 1 Py
Plumbole 1 Py2
Plumbole 1 NHC
Plumbole 1 NHC2
Plumbole
Plumbole22
TMPb TMS
para
r
r
tot
d
calc
dia
r
para
r
rso
rtot
dcalc
dexptl[a] 4601[b] 5539[c] 219[b] 227[c] 181[c] -[d] -[d] -[d]
Pb
10074
25532
4543
1108
9946
26856
903
3993
3885
Ca
257
2231
26
155
256
2240
243
228
210
Cb Pb Ca Cb Pb Ca Cb Pb Ca Cb Pb Ca Cb Pb Ca Cb Pb Ca Cb Pb Ca Cb Pb C
252 10074 257 252 10074 256 252 10074 256 252 10073 256 252 10073.0 256.0 251.6 10074 257 252 10078 265 251 10078 258
2245 25075 2246 2252 25228 2249 2247 2466 2258 2246 24736 2261 2248 24325 2253 2245 28813 2220 2243 25083 2240 2212 24427 276
7 4998 12 0 4845 7 5 5412 22 7 5338 25 4 5748 3 7 1262 36 9 4995 16 38 5650 182
174 652 170 181 805 175 176 239 183 175 313 187 178 298 178 175 4389 145 172 656 166 144
252 9945 256 252 9944 259 252 9946 255 253 9945 255 252 9945 255 252 9946 256 252 9949 255 250 9950 257
2244 26523 2257 2251 26570 2261 2246 26027 2269 2245 26212 2273 2248 25616 2263 2244 29556 2233 2247 26661 2258 2212 25333 276
212 890 243 213 1073 247 212 1333 233 211 1300 230 213 1695 224 210 21532 281 230 3004 216 217 3260 1
24 4312 244 211 4448 249 26 5251 247 23 5033 247 28 6025 233 22 21142 258 225 6293 219 22 7878 182
186 3566 226 193 3430 231 188 2627 229 185 2845 229 191 1854 215 184 9020 240 207 15865 202 160
217 178 -[d] -[d] 1793[c] 202[c] 194[c] -[d] -[d] -[d]
[a] Ref. [8]. [b] Measured in THF. [c] Measured in benzene-d6. [d] As the coordination number of plumboles in solution is not clear, the observed chemical shifts can be assigned clearly to neither plumbole 1 L nor plumbole 1 L2 (L 5 THF, Py, NHC).
chemical shifts, all data fall around one line, whereas in the Cb chemical shifts, the data are independent of rSO. This suggests that the rSO term of Ca is controlled by the interaction between the ligand and Pb, whereas the rSO term of Cb seems to be controlled by several mechanisms related to the spin transfer.
ligand-coordinated plumboles did not have aromaticity, the NICS values indicated that they were located midway between
Nuclear spin–spin coupling constants The calculated and observed J [Pb,C] are shown in Table 5. In plumbole 1 THF2, the J values calculated by the nonrelativistic DFT and the ZORA-DFT methods are 2395 and 21164.9 Hz, respectively, whereas the observed value is 1049 Hz. The plus– minus sign of observed J values in Table 5 should be corrected referring to the calculated values. The calculated result is significantly improved by the relativistic effect. The relativistic correction is approximately 60%. The sign of J[Pb,C] is determined to be negative in our calculations. Unfortunately, the calculated results cannot reproduce the difference between plumbole 1 Py2 and plumbole 1 NHC2
Conclusions We performed calculations for a series of lead-containing plumboles and ligand-coordinated plumboles using the ZORADFT method and the ZORA-TD-DFT method with B3LYP. Although, from the viewpoint of molecular structure, the 852
Journal of Computational Chemistry 2014, 35, 847–853
Figure 5. Correlation between rSO terms of 13Ca and 13Cb chemical shifts and reciprocal of the magnetically allowed transition energies in plumbole and ligand-coordinated plumboles. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]
WWW.CHEMISTRYVIEWS.COM
FULL PAPER
WWW.C-CHEM.ORG
Table 5. Nuclear spin–spin coupling constants (J[Pb,C]) of plumbole 1 ligand (Hz). Nonrelativistic Hamiltonian
Plumbole Plumbole Plumbole Plumbole Plumbole Plumbole
1 1 1 1 1 1
THF THF2 Py Py2 NHC NHC2
ZORA
dia
para
FC
tot.
dia
para
FC 1 SO
tot.
exptl.[a]
0.3 0.2 0.1 0.2 0.1 0.2
9.9 5.1 21.7 1.9 25.9 23.8
2358.6 2400.3 2283.3 2346.2 2330.1 2298.5
2348.5 2395.1 2284.9 2344.2 2335.9 2302.2
0.3 0.2 0.1 0.2 0.1 0.2
8.5 6.5 24.5 1.5 29.8 27.4
21495.8 21171.6 21106.1 21107.8 21106.4 21098.6
21487.1 21164.9 21110.4 21106.2 21116.0 21105.9
1049 1049 735 735 754 754
[a] Ref. [8].
aromatic and antiaromatic compounds. In the UV spectra, the present method reproduced the trends of plumbole and ligand-coordinated plumboles. The lowest excited state of plumbole disappeared in the ligand-coordinated plumboles due to the interaction between the 6pp orbital of Pb and the lone-pair orbital of the ligand. The trends of 207Pb- and 13CNMR chemical shifts and J[Pb,C] were also well-reproduced by the present method. The relativistic effect was essential for those properties. The nearest-neighbor Ca from Pb was more shielded than Cb. This trend was opposite to the previous heavy-metal effect observed in IACaBCbAH. We discussed this point based on the results of decomposition into rSO, rdia, and rpara terms, and the electron transition energies. Keywords: lead Pb NMR ZORA chemical shift nuclear spin–spin coupling plumbole relativistic effect aromaticity five-membered ring
How to cite this article: T. Kawamura, M. Abe, M. Saito, M. Hada. J. Comput. Chem. 2014, 35, 847–853. DOI: 10.1002/ jcc.23556
[1] M. Saito, Coord. Chem. Rev. 2012, 256, 627. [2] M. Saito, R. Haga, M. Yoshioka, K. Ishimura, S. Nagase, Angew. Chem. Int. Ed. 2005, 44, 6553. [3] M. Saito, M. Yoshioka, Coord. Chem. Rev. 2005, 249, 765. [4] M. Karni, Y. Apeloig, J. Kapp, P. von R. Schleyer, In The Chemistry of Organic Silicon Compounds, Vol. 3, Chapter 1; Z. Rappoport, Y. Apeloig, Eds.; Wiley: Chichester, 2001; pp. 1–163. [5] M. Saito, M. Sakaguchi, T. Tajima, K. Ishimura, S. Nagase, M. Hada, Science 2010, 328, 339. [6] E. H€ uckel, Z. Phys. 1931, 70, 204. [7] P. v. R. Schleyer, C. Maerker, A. Dransfeld, H. Jiao, N. J. R. v. E. Hommes, J. Am. Chem. Soc. 1996, 118, 6317. [8] M. Saito, T. Akiba, M. Kaneko, T. Kawamura, M. Abe, M. Hada, M. Minoura, Chem. Eur. J. 2013, 19, 16946. [9] B. Goldfuss, P. von R. Schleyer, F. Hampel, Organometallics 1996, 15, 1755. [10] B. Goldfuss, P. von R. Schleyer, Organometallics 1997, 16, 1543. [11] H. Nakatsuji, K. Kanda, K. Endo, T. Yonezawa, J. Am. Chem. Soc. 106, 4653, 1984 [12] H. Nakatsuji, In Nuclear Magnetic Shieldings and Molecular Structure; J. A. Tossell, Ed.; Kluwer: Dortrecht, 1993; p. 263. [13] H. Nakatsuji, H. Takashima, M. Hada, Chem. Phys. Lett. 1995, 233, 95.
[14] H. Kaneko, M. Hada, T. Nakajima, H. Nakatsuji, Chem. Phys. Lett. 1996, 261, 1. [15] M. Hada, H. Kaneko, H. Nakatsuji, Chem. Phys. Lett. 1996, 261, 7. [16] M. Hada, R. Fukuda, H. Nakatsuji, Chem. Phys. Lett. 2000, 321, 452. [17] R. Fukuda, M. Hada, H. Nakatsuji, J. Chem. Phys. 2003, 118, 1015. [18] J. Seino, M. Hada, J. Chem. Phys. 2010, 132, 174105. [19] W. Nakanishi, S. Hayashi, K. Narahara, D. Yamaki, M. Hada, Chem. Eur. J. 2008, 14, 7278. [20] V. G. Malkin, O. L. Malkina, D. R. Salahub, Chem. Phys. Lett. 1996, 261, 335. [21] E. v. Lenthe, A. E. Ehlers, E. J. Baerends, J. Chem. Phys. 1999, 110, 8943. [22] E. v. Lenthe, E. J. Baerends, J. G. Snijders, J. Chem. Phys. 1993, 99, 4597. [23] E. v. Lenthe, E. J. Baerends, J. G. Snijders, J. Chem. Phys. 1994, 101, 9783. [24] E. v. Lenthe, J. G. Snijders, E. J. Baerends, J. Chem. Phys. 1996, 105, 6505. [25] E. v. Lenthe, R. v. Leeuwen, E. J. Baerends, J. G. Snijders, Int. J. Quantum Chem. 1996, 57, 281. [26] S. K. Wolff, T. Ziegler, E. van Lenthe, E. J. Baerends, J. Chem. Phys. 1999, 110, 7689. [27] J. Autschbach, T. Ziegler, J. Chem. Phys. 2000, 113, 936. [28] J. Autschbach, T. Ziegler, J. Chem. Phys. 2000, 113, 9410. [29] C. Lee, W. Yang, R. G. Parr, Phys. Rev. B 1988, 37, 785. [30] A. D. Becke, J. Chem. Phys. 1993, 98, 5648. [31] S. H. Vosko, L. Wilk, M. Nusair, Can. J. Phys. 1980, 58, 1200. [32] E. J. Baerends, T. Ziegler, J. Autschbach, D. Bashford, A. B erces, F. M. Bickelhaupt, C. Bo, P. M. Boerrigter, L. Cavallo, D. P. Chong, L. Deng, R. M. Dickson, D. E. Ellis, M. v. Faassen, L. Fan, T. H. Fischer, C. Fonseca Guerra, A. Ghysels, A. Giammona, S. J. A. v. Gisbergen, A. W. G€ otz, J. A. Groeneveld, O. V. Gritsenko, M. Gr€ uning, S. Gusarov, F. E. Harris, P. v. d. Hoek, C. R. Jacob, H. Jacobsen, L. Jensen, J. W. Kaminski, G. van Kessel, F. Kootstra, A. Kovalenko, M. V. Krykunov, E. v. Lenthe, D. A. McCormack, A. Michalak, M. Mitoraj, J. Neugebauer, V. P. Nicu, L. Noodleman, V. P. Osinga, S. Patchkovskii, P. H. T. Philipsen, D. Post, C. C. Pye, W. Ravenek, J. I. Rodrıguez, P. Ros, P. R. T. Schipper, G. Schreckenbach, J. S. Seldenthuis, M. Seth, J. G. Snijders, M. Sola, M. Swart, D. Swerhone, G. te Velde, P. Vernooijs, L. Versluis, L. Visscher, O. Visser, F. Wang, T. A. Wesolowski, E. M. v. Wezenbeek, G. Wiesenekker, S. K. Wolff, T. K. Woo, A. L. Yakovlev, ADF2012 SCM Theoretical Chemistry; Vrije Universiteit: Amsterdam, The Netherlands, 2012. [33] H. Tanimura, Y. Honda, K. Sugiura, M. Hada, Bull. Chem. Soc. Jpn. 2011, 84, 845. [34] M. Z. Kassaee, S. Arshadi, M. Acedy, E. Vessally, J. Organomet. Chem. 2005, 690, 3427. [35] S. Fukawa, M. Hada, R. Fukuda, S. Tanaka, H. Nakatsuji, J. Comput. Chem. 2001, 22, 528.
Received: 21 November 2013 Revised: 23 January 2014 Accepted: 24 January 2014 Published online on 7 February 2014
Journal of Computational Chemistry 2014, 35, 847–853
853
