High-Frequency 13C and 29Si NMR Chemical Shifts in Diamagnetic

Jan 28, 2016 - Relativistic Spin–Orbit Heavy Atom on the Light Atom NMR Chemical Shifts: General Trends ... Mendeleev Communications 2018 28 (1), 1-...
0 downloads 0 Views 3MB Size
Download PDF
Article pubs.acs.org/IC

High-Frequency 13C and 29Si NMR Chemical Shifts in Diamagnetic Low-Valence Compounds of TlI and PbII: Decisive Role of Relativistic Effects Jan Vícha,*,† Radek Marek,‡ and Michal Straka§ †

Center of Polymer Systems, University Institute, Tomas Bata University in Zlín, Třída T. Bati, 5678, CZ-76001, Zlín, Czech Republic CEITEC - Central European Institute of Technology, Masaryk University, Kamenice 5/A4, CZ-62500 Brno, Czech Republic § Institute of Organic Chemistry and Biochemistry, Academy of Sciences, Flemingovo nam. 2, CZ-16610, Prague, Czech Republic ‡

S Supporting Information *

ABSTRACT: The 13C and 29Si NMR signals of ligand atoms directly bonded to TlI or PbII heavy-element centers are predicted to resonate at very high frequencies, up to 400 ppm for 13C and over 1000 ppm for 29Si, outside the typical experimental NMR chemical-shift ranges for a given type of nuclei. The large 13C and 29 Si NMR chemical shifts are ascribed to sizable relativistic spin− orbit effects, which can amount to more than 200 ppm for 13C and more than 1000 ppm for 29Si, values unexpected for diamagnetic compounds of the main group elements. The origin of the vast spin−orbit contributions to the 13C and 29Si NMR shifts is traced to the highly efficient 6p → 6p* metal-based orbital magnetic couplings and related to the 6p orbital-based bonding together with the low-energy gaps between the occupied and virtual orbital subspaces in the subvalent TlI and PbII compounds. New NMR spectral regions for these compounds are suggested based on the fully relativistic density functional theory calculations in the Dirac−Coulomb framework carefully calibrated on the experimentally known NMR data for TlI and PbII complexes. interactions, particularly for the compounds with TlI−TlI bonds.23 The plumbylenes, PbIIR2 (R = σ-bonded alkyl,24 aryl,25 or silyl26) are used in transition-metal chemistry27,28 and for preparation of exotic heterometallic compounds, metalloplumbylenes, with direct Pb−M (M = Pb, Ti, Zr, Hf, Cr, Mo, W) bonds.26,29,30 Metalloplumbylidynes, alkyne analogues featuring a triple bond between Pb and M (M = Mo, W) with C−Pb−M angle ∼180°,31,32 were synthesized as well;29 for an example, see compound 12 in Figure 1. Nuclear magnetic resonance (NMR) spectroscopy is used as a main tool for identification of subvalent TlI and PbII organometallic compounds in solution. For a historical overview, see the Supporting Information. However, the NMR signals of the light atoms (LA) directly bonded to the heavy atoms (HA) are often missing from the spectra, especially for the TlI compounds.10,17 The reported signals of vicinal LAs, mostly for PbII compounds, are notably deshielded, with NMR chemical shifts at LA > 100 ppm for 13C signals in alkyls,24,33 over 250 ppm for 13C signals in aryls,34,35 and up to 200 ppm for 29Si signals in silyls.27,36 To our best knowledge, the highest-frequency 13C NMR resonance of LA reported in a

1. INTRODUCTION The relativistic stabilization of low oxidation states, the inert lone-pair effect,1 has a strong influence on the chemistry of heavier group 13 to 15 elements, particularly that of thallium, lead, and bismuth.2 In the present work we focus on TlI and PbII. The lower oxidation states (TlI and PbII) are preferred in inorganic compounds of Tl and Pb, while their organometallic chemistry is dominated by the full-valence TlIII and PbIV species,3,4 with an exception of the remarkably stable π-bonded cyclopentadienyl complexes.5,6 The first stable organolead(II) compound was prepared already in 1970s,7 but only the development of sterically crowded ligands, such as terphenyls,8 allowed for the preparation of stable organothallium(I) species more than two decades later.9−12 The subvalent TlI and PbII organometallics with unsaturated coordination sphere found interesting applications in chemical synthesis. Organothallium(I) compounds are used as mild reagents for synthesis of unique, otherwise inaccessible compounds of the main group13 and transition-metal14 elements. They also act as reactive intermediates15,16 in various chemical reactions. Thallium clusters with TlI−TlI,9,17 and TlI− M bonds (M = Pt, Au)18,19 bound by the weak metallophilic 6s2−6s2 forces,20−22 are of the particular interest due to their unusual optical properties given by the metallophilic © XXXX American Chemical Society

Received: November 21, 2015

A

DOI: 10.1021/acs.inorgchem.5b02689 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry

Figure 1. Structures of SnII (2a,24 4102), SnIV (1a),24 TlI (6,5 7,11 18103), PbII (2b,24 3,104 5,33 8,105 9,33 10,33 11,34 12,31 13,106 14,106 15,34 16,27 1727,36), PbIV (1b)24 compounds. Experimental (δEXP) and 4c-PBE calculated (δCALC and δSO) 13C and 29Si NMR chemical shifts for the LA (C or Si) directly bonded to the metal center M (M = Sn, Pb). Light atoms in question are highlighted. All values are given in parts per million. Me = Methyl, iPr = isopropyl, tBu = tert-butyl, Ph = phenyl.

diamagnetic PbII compound is 357 ppm in a cationic PbII complex (isoelectronic with TlI).37 Though similar chemical shifts of LAs could be expected in analogous known TlI compounds,10,17 the NMR chemical shifts of LAs in these compounds were not reported, despite the fact that the rest of NMR signals was readily observed. The possible reason may be unexpectedly large downfield NMR chemical shift of LA in TlI compounds, well outside standard 13C NMR chemical shift region, as discussed below. The deshielding observed in PbII compounds (as well as in other low-valent tetrels) was previously rationalized by the sizable paramagnetic effects38 arising from the small singlet− triplet energy gaps24 found in the diamagnetic divalent compounds of the group 14 elements, as compared to their tetravalent analogues.39 However, this is in contradiction with the known fact that the singlet−triplet energy splitting in MII species (M = C, Si, Sn, Pb) rises considerably with increasing proton number of the group 14 element,40 particularly between SnII and PbII, for instance, from 23.8 kcal/mol in SnH2 to 41 kcal/mol in PbH2.41 In the present work we demonstrate that the large deshielding effects at the vicinal LA nuclei in PbII and TlI species originate primarily from the large relativistic effects of the Heavy-Atom on the Light-Atom (HALA).42,43 HALA effect

induces a spin−orbit contribution to light-atom NMR chemical shift, δSO(LA), via mixing the triplet states in the ground-state wave function due to the spin−orbit coupling at HA.44−46 For a more detailed discussion of the HALA mechanism, see corresponding section in the Supporting Information and references within. The electronic structure effects on the δSO(LA) have been extensively studied in the past.45,47−52 In brief, the correlations were found, for example, between the size of the δSO(LA) and the s-character of the LA in the HA−LA bond53 and/or the contribution of the HA atomic orbitals (particularly 5d and 6p AOs in Pt and Au compounds) to MOs via the MO → MO* magnetic couplings.54,55 Also the smaller the energy denominators (see eq S1) related to highest occupied molecular orbital−lowest unoccupied molecular orbital (HOMO− LUMO) gap, the larger the δSO(LA) can be expected.56,57 As shown recently, these correlations can be conveniently simplified by a quantitative correspondence between the δSO(LA) and the covalence of the LA−HA bond.58 The typical size of δSO for the LA, such as 13C or 29Si, covalently bound to a heavy element of the fifth or sixth period is several tens of parts per million (usually not more than 50 ppm).59−61 The δSO(LA) should be rather marginal in systems with lighter elements. Therefore, the size of the δ SO B

DOI: 10.1021/acs.inorgchem.5b02689 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry

PBE40 level, the ΔδSO may be easily calculated;95 see Table S4. For the sake of clarity (to avoid doubling of all values and corresponding figures), NMR chemical shift contributions of the spinors arising from the scalar-relativistic molecular orbitals due to the spin−orbit splitting on 2c-PBE40 level are summarized and reported as contributions of the parental scalar-relativistic MOs. The contributions from the individual spinors thus correspond to a half of the reported value. The molecules 32−34 were analyzed also using the sum-over-state finite-field third-order perturbation theory96,51 as implemented in MAG 2.1 suite97 interfaced with Gaussian 09.98 Gaussian/MAG calculations were done at the PBE level with def2-TZVPP basis set66 and the scalar-relativistic ECPs with 60 electrons in core (ECP-MWB60) for Pb and Tl atoms.67,98 As shown previously,54,58 this approach allows for a direct MO analysis of δSO. The obtained results were in a good agreement with those from 2c-PBE40 level. Because of better treatment of relativistic effects by the SO-ZORA approach, only 2cPBE40 obtained data are presented. The composition of molecular orbitals (NLMO) and M−L (L = C, N) bonds was determined from natural bond orbital (NBO) analysis99 as implemented in the NBO 6.0 module (Gennbo)92 interfaced to ADF code.

contribution to 13C (>200 ppm) and 29Si (>1000 ppm) NMR chemical shifts predicted for PbII and particularly TlI complexes studied in this work is unparalleled by any of those found in the diamagnetic p- or d-element compounds. It is only comparable with the recently predicted62 and later proven63 giant δSO(LA) effects found for the LAs in the 5f-element uranium complexes. The δSO(LA) is predicted to be so large in certain TlI compounds, that the LA signals are shifted outside the standard NMR chemical-shift ranges, suggesting an explanation for the lack of experimental data for the NMR signals of vicinal light atoms in these complexes.10,17 On the basis of the state-of-theart fully relativistic density functional theory (DFT) calculations, origin of the δSO(LA) is rationalized, and appropriate 13 C and 29Si NMR spectral regions are suggested in this work.

2. METHOD SECTION Molecular Structures. These are based on the published X-ray data, where applicable, or optimized using the PBE0 functional64,65 and def2-TZVPP basis set66 for all atoms, with corresponding relativistic effective core potentials (ECPs)67 for the metal centers (ECP substituting 60 electrons for all heavy elements Pt−At). The D3 dispersion correction by Grimme68 was employed during the optimizations. This level of theory has been justified by the previous studies of metal complexes.69−71 The implicit conductor-like screening model72 solvent model was adopted throughout the geometry optimizations of the experimentally prepared compounds, with solvent parameters selected according to the experimental setup. The structures of theoretically predicted model compounds were optimized in vacuo, that is, without implicit solvent treatment. Calculations of Nuclear Magnetic Resonance Chemical Shifts. These were performed using the four-component (4c) matrix Dirac−Kohn−Sham (DKS) relativistic calculations within the Dirac− Coulomb framework73,74 as implemented in the ReSpect 3.42 code,75 with PBE functional76,77 and uncontracted Dyall’s valence triple-ζ basis set78−80 for all atoms. For the sake of the computational time, Dyall’s valence double-ζ basis set was used for distant atoms, at least three bonds away from the NMR spectator atom. The four-component results were compared with those obtained by two-component (2c) spin−orbit zeroth-order regular approximation (SO-ZORA) as implemented in the ADF package,81−83 using standard TZP basis set from ADF library for all atoms84,85 and either the PBE or PBE40 functional, which is standard PBE0 functional with exactexchange admixture increased to 40%.86 The PBE40 functional has proven to provide good results in previous studies of transition-metal complexes71,87 as well as actinide complexes.62,63 Referencing. To reduce the systematic errors in calculations of the NMR chemical shifts, the following secondary references were used: For 13 C, benzene in benzene (δ = 127.8 ppm relative to tetramethylsilane (TMS));88 for 15N, pyridine in dimethyl sulfoxide (δ = 316.9 ppm relative to liquid ammonia);89,90 for 29Si, hypersilane, (Me3Si)4Si, (δ = −135.5 ppm relative to TMS).91 The chemical shifts δi were obtained based on the equation δi = σref − σi + δref

3. RESULTS AND DISCUSSION 3.1. Calibration of the Theoretical Levels. The computational approach for the NMR calculations was carefully calibrated on the NMR data of the experimentally characterized SnII, SnIV, PbII, Pb,IV and TlI compounds 1−18 (Figure 1). To the best of our knowledge, only TlI compounds with small or moderately sized experimental δEXP(LA) have been reported; hence, the method calibration largely relies on the experimental NMR data for a series of PbII compounds with sizable δEXP(LA). Unless otherwise stated, only the NMR chemical shifts (δ) of light spectator atoms (LA) directly bound to the HA center are presented and are noted as δ(LA), where LA = 13C or 29Si. The spin−orbit contribution to the calculated total NMR chemical shift, δCALC or δCALC(LA), is noted as δSO or δSO(LA). Experimental shifts are referred to as δEXP or δEXP(LA). The overall performance of the tested methods versus the experimental data is illustrated in Figure 2. The performance of the two-component (2c) SOZORA,81−83 using PBE0 functional64,65 with 40% of exactexchange admixture (denoted as 2c-PBE40) was evaluated initially. This approach was recently thoroughly tested on

(1)

where σref is the calculated NMR shielding of reference nucleus, σi is the calculated NMR shielding of the studied atom, and δref is the experimental NMR chemical shift of the secondary reference relative to the primary standard. Analysis of the Nuclear Magnetic Resonance Chemical Shifts and Electronic Structure. Analysis of δSO was also done at the SO-ZORA PBE40 (referred to as 2c-PBE40 in the paper) level using NMR shielding decomposition into canonical molecular orbitals as implemented in ADF2014 package, using NBO 6.0 module (Gennbo)92 interfaced to ADF code.93,94 Because only the sum of paramagnetic (δpara) and spin−orbit (δSO) NMR chemical shift contribution of individual MOs, noted as δs+p, is available using this approach, analogous analysis was done on scalar ZORA 1c-PBE40 level (without spin−orbit contribution). As the δs+p = δpara at the 1c-

Figure 2. Correlation of experimental and calculated chemical shifts at LA (ppm). C

13

C NMR

DOI: 10.1021/acs.inorgchem.5b02689 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry various transition-metal complexes with δSO(LA) < 50 ppm, and a very good agreement between the theory and experiment was found.87,100,101 The method was also used in the relativistic calculations of the NMR chemical shifts in actinide complexes with very large δSO(LA) > 300 ppm,62,63 although substantial deviations from the experimental data were observed in some cases.63 The 2c-PBE40 provides a good agreement between predicted and experimental data for δ(13C) in compounds with small and medium-sized (in context of this work) δSO(13C) < 50 ppm; see, for instance, 1a−4 in Figure 1. However, the deviations of the calculated data from the experimental ones increase considerably in cases where δSO(13C) amounts 100 ppm and more, leading to a poor RMSD value of ∼17 ppm for 2c-PBE40 level versus experiment as seen in Figure 2 and Table S1 in the Supporting Information. This behavior can be assigned to the recently highlighted approximations of response kernel in the current implementation of the SO-ZORA approach,107 which were found to be responsible for deviations as large as 20% of calculated total δSO(13C).95,101 For smaller δSO(13C), these deviations can be partially compensated for by the increased exact-exchange admixture in the functional.87,95,100,101 To avoid the above-mentioned approximations, we tested the performance of the four-component (4c) DKS approach74,108 as implemented in the ReSpect code,75,109 using PBE functional76,77 (denoted as 4c-PBE) on the experimental test set in Figure 1. Results were compared with the SO-ZORA calculations at PBE level (denoted as 2c-PBE), to demonstrate the importance of the missing kernel while using the same functional for both relativistic levels. Although some minor differences between 4c-PBE and 2cPBE approaches can be expected, for example, from the different types of basis sets (Gaussian vs Slater basis sets in ReSpect and ADF, respectively), our comparison revealed rather striking differences between the two- and fourcomponent levels; see Figure 2 and Table S1 for numerical values. The 2c-PBE results quickly deteriorate with the increasing size of the δSO(LA), resulting in errors of over 50 ppm for δ(13C) and root-mean-square deviation (RMSD) ≈ 40 ppm, Table S1. The tested 2c-PBE implementation cannot be recommended for calculations of studied systems and their analogues. In contrast, the 4c-PBE approach provides very good δ(13C) results given the large size of δSO(13C) with RMSD = 7 ppm. Thus, the 4c-PBE level was adopted in the production calculations of δSO(13C). The number of NMR-characterized PbII compounds with Sibased ligands is rather limited,27,35,36 while the majority of reactions attempted between silicon-based ligands and thallium yields bis-substituted (SiR)2Tl−Tl(SiR)2 dimers instead of expected TlI complexes.103,110,111 Accordingly, the number of reference compounds for calibration of δ(29Si) calculations is limited. The RMSDs for the calculated versus experimental 29Si NMR data are larger than those for the 13C NMR; ∼100 ppm for 2c-PBE, ∼45 ppm for 2c-PBE40, and ∼35 ppm for 4c-PBE approach. Note that similar deviations were observed in the previous DFT studies of lighter metallosilyl compounds, with considerably smaller spin−orbit effects involved.61,112 Several other levels (e.g., 2c-B3LYP and 4c-BP86) were tested during calibration, as well as different secondary references [(Me3Si)3SiH, (Me3Si)4Sn] for δ(29Si) (data not shown), but the 4c-PBE results could not be improved. Hence, 4c-PBE level is also used for discussing the 29Si NMR.

3.2. The Relativistic Effects on the Calculated Nuclear Magnetic Resonance Chemical Shifts in the Experimentally Known SnII, SnIV, PbII, and PbIV Compounds. In this Section, the NMR chemical shifts are analyzed and compared among the structurally related SnII, SnIV, PbII, and PbIV compounds from the previous studies of Wrackmeyer et al.24,38 (1a, 1b, 2a, and 2b in Figure 1), Eichler et al.102 (4 in Figure 1), and Pu et al.33 (5 in Figure 1). Differences between the full- and low-valence species (MIV and MII) as well as differences between PbII and SnII compounds are discussed. Important role of the δSO(13C) contribution to the LA NMR chemical shifts in PbII compounds as compared to their less relativistic SnII analogues is highlighted. Both the experimental and calculated δ(13C) at the LA in PbIV and SnIV compounds 1a and 1b are close to 0 ppm, with a negligible δSO contribution. As expected,39 the δ(13C) resonances in analogous SnII and PbII compounds (2a and 2b) are considerably more deshielded, δ(13C) is 50 and 109 ppm in 2a and 2b, respectively. Whereas the deshielding in the SnII compound 2a is caused predominantly by an increase in the scalar-relativistic paramagnetic contribution to the shielding tensor (cf. above) with only a small δSO(13C) of +10 ppm, the δSO(13C) in PbII analogue 2b is 5 times larger, +52 ppm, and is thus mostly responsible for the difference in δ(13C) between 2a and 2b (50 vs 109 ppm). With increasing s-character of the LA in the bonding with the metal center,53 the δSO(13C) values in PbII compounds can reach up to 150 ppm in 14, Figure 1. However, for SnII compound 4 the δSO(13C) is only 24 ppm, while analogous PbII compound 5 has δSO(13C) = 126 ppm. With scalar-relativistic δCALC(13C) (4c-PBE level without the spin−orbit contribution, i.e., 1c-PBE) in 4 and 5 almost identical, ∼160 ppm, the ∼100 ppm larger δ(13C) in 5 is almost entirely caused by the difference in δSO(13C). Similar differences in NMR chemical shifts can be found for a number of SnII and PbII analogues.106 It may be of interest to briefly discuss the possible correlation between paramagnetic contribution to the NMR chemical shift of heavy element, δ(HA), and δSO(LA). The paramagnetic deshielding contribution is known to be the determining factor for 207Pb NMR chemical shifts in PbII compounds,113 which can be as large as +19 500 ppm.113 Considerable size of the paramagnetic effects was traced to the low-energy gap between occupied and vacant Pb 6p orbitals and their efficient coupling in the magnetic field,26,39 factors related also to δSO(LA), see below.57,58 Therefore, a correlation between the size of δ(207Pb) and δSO contributions at LA can be expected. Indeed, according to recent study of Wilfling et. al,106 the δ(13C) in the series of PbII compounds with aromatic ligands increases from ∼260 ppm (8 in Figure 1) to ∼330 ppm (14 in Figure 1), while δ(207Pb) is increasing accordingly, from 4900 ppm to ∼9600 ppm at the same time.106 Note that the difference in δ(13C) between 8 and 14 is ∼70 ppm, while difference in calculated δSO(13C) for these two compounds is ∼60 ppm (Figure 1). Thus, only a fraction of the total LA NMR chemical-shift increase in these PbII compounds can be attributed to the increased scalar-relativistic paramagnetic contribution at LA. This conclusion is valid for the above-discussed cases as well as for 29Si results presented in Section 3.5. Analogous correlation between δ(119Sn) and δ(13C) in certain aromatic ligands was found as well.106 Although correlation between the δ(HA) and the δ(LA) in MII compounds (M = Sn, Pb) was found for number compounds with similar ligand type, it should not be directly D

DOI: 10.1021/acs.inorgchem.5b02689 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry

The δ(13Cipso) and δSO(13Cipso), calculated on 4c-PBE level (including spin−orbit) and without spin−orbit correction, essentially on the scalar-relativistic level (1c-PBE), are shown in Figure 3. Numerical data and chemical formulas of individual species are given in Table 1.

generalized.106 Factors such as type of the LA (aliphatic or aromatic carbon), presence of electron donating/withdrawing groups in the ligand, or size of the ligand side chains (widening the LA−Pb−LA angle) were found to affect both the HA and the LA NMR chemical shifts.106 Nevertheless, large 119Sn or 207 Pb NMR chemical shift with significant paramagnetic contribution in subvalent compounds of heavier group 14 elements may indicate high-frequency δ(13C) at LA. To summarize, the δ(13C) difference between SnII and SnIV compounds is caused mainly by the paramagnetic (deshielding) contribution to the shielding tensor, as expected.38 However, the high-frequency δ(13C) at LA in PbII compounds, which are often more than 100 ppm deshielded compared to the SnII analogues,106 are significantly influenced by the δSO contribution induced by the heavy Pb atom. The δSO(LA) in PbII compounds can be as large as ∼150 ppm for δSO(13C) in 14 and ∼210 ppm for δ(29Si) in 17. 3.3. Trends in the Relativistic Effects on the Nuclear Magnetic Resonance Chemical Shifts at the Light Atom along a Series of Complexes of the Sixth Row Elements Pt−At. A comparison that was discussed above for the complexes of the group 14 elements Pb and Sn is not possible for the group 13 elements (Tl and In) due to the lack of the experimental NMR data in structurally related complexes, particularly for TlI. Instead, a series of model complexes featuring phenyl ligand bonded to a heavy-metal center was used to map the trends in the LA relativistic NMR chemical shifts along the sixth period of the periodic table, from platinum to astatine. This allows us to see the previously58,88,101 and presently studied HALA effects (δSO) as well as the total NMR chemical shifts at LA in a broader perspective. The δ(13Cipso) of the phenyl ligand is calculated for each model compound given in Table 1. Compounds are noted as

Figure 3. 1c-PBE (without δSO) and 4c-PBE (with δSO) calculated δ(13Cipso) for Pt(II) through At(I) complexes. For chemical formulas, see Table 1. The differences between 1c-PBE and 4c-PBE values correspond to δSO contribution to the total 13C NMR chemical shift at Cipso, see Table 1.

The δ(13Cipso) values along the series calculated on the 1cPBE level (without δSO, yellow columns in Figure 3) are comparable for all systems within ∼50 ppm. Similar results are obtained for the δ(13Cipso) values calculated at the 4c-PBE level (δSO included, orange columns in Figure 3) in the transitionmetal complexes Pt(II)−Hg(II) and in the full-valence Tl(III), Pb(IV), and Bi(V) model compounds. The δSO(13Cipso) in these complexes is relatively small, 10−30 ppm in absolute size. Note that the negative (shielding) δSO(13Cipso) turns positive (deshielding) between Au(III) and Hg(II), with maximum δSO(13Cipso) = +17 ppm for Tl(III) and +11 ppm for Pb(IV), while from Bi(V) on it becomes of the shielding nature again. The change of the sign of δSO(LA) after Au complexes was rationalized only recently,58 and it is discussed in more detail in Section 3.6. The most important observation in Figure 1 is strikingly large δSO(13Cipso) of +270, + 175, and +54 ppm for the subvalent species of groups 13 to 15 resulting in the highfrequency total δ(13Cipso) shifts of 460, 360, and 215 ppm for Tl(I), Pb(II), and Bi(III), respectively. The δSO(13Cipso) values in subvalent species are more than 15 times larger than those in the full-valence species Tl(III), Pb(IV), and Bi(V), while decreasing along the sixth period from Tl(I) to Bi(III). The calculated δ(13Cipso) of ∼460 ppm in Tl(I) model system supports our assumption about the large δSO(LA) effects to be present, but not yet reported, in the monocoordinated TlI compounds.10,17 The δ(13Cipso) in Tl(I) is ∼100 ppm larger than that in Pb(II) and lies in the highestfrequency region of the 13C NMR resonances of diamagnetic systems (>400 ppm), although not near the most deshielded 13 C NMR resonance reported so far, 675 ppm in ruthenium carbide.114 Nevertheless, assuming the size of the standard 13C NMR regions to be