NMR SHIELDING CONSTANTS IN BF3 AND MAGNETIC DIPOLE MOMENTS OF 11B AND 10B NUCLEI

Anna Szyprowska1, Wodzimierz Makulski1, Karol Jackowski1, Andrej Antuek2, Micha Jaszuski3, and Jonas Juslius4 Laboratory of NMR Spectroscopy, Department of Chemistry,University of Warsaw, Pasteura 1, 02-093 Warsaw, Poland 2 Slovak University of Technology in Bratislava, Faculty of Materials Science and Technology in Trnava, Paulinska 16, 917 24 Trnava, Slovak Republic 3 Institute of Organic Chemistry, Polish Academy of Sciences, Kasprzaka 44, 01-224 Warsaw, Poland 4 Centre for Theoretical and Computational Chemistry, Department of Chemistry, University of Troms, N-9037 Troms, Norway
1

The increasing accuracy of ab initio calculations of nuclear magnetic resonance (NMR) shielding constants, combined with NMR frequency measurements, brings forth an interesting new application - it enables the determination of accurate nuclear magnetic dipole moments. To determine the magnetic dipole moment of a bare nucleus, the effects of chemical shielding must be isolated. NMR experiments in the gas phase can provide the zero-pressure extrapolated values, not affected by intermolecular forces and thus suitable for a direct comparison with theoretical values computed for a single molecule, so a correct description of the chemical shielding can be obtained. Our previous results for a series of nuclei, important in chemical applications of NMR, are discussed in a recent review [1]. Trifluoroborane, BF3, is a stable and structurally simple compound of boron. At room temperature BF3 is a gas-phase substance, which makes it a good candidate for the determination of boron magnetic moments by applying NMR spectroscopy. The boron atom has only six electrons (sp2 hybridization) in its outer shell in BF3 so this molecule is planar and can act as a Lewis acid. On the other hand, the large number of electron lone pairs of fluorine atoms in this molecule makes the calculation of highly accurate NMR shielding constants a challenging task. In this work we have first calculated the total shielding constant at the molecular equilibrium geometry. The shielding constant was obtained as the sum of a nonrelativistic electroncorrelated contribution, and a relativistic correction computed at a lower level of theory. Furthermore, vibrational and temperature corrections to the shielding constants were added. The nonrelativistic calculations were carried out using the ACES II program. The relativistic corrections were estimated using the Dalton program and a 2008 version of the DIRAC package. We obtained the total ab initio shielding constants: 10B = 97.879; F = 331.949 ppm for 10 BF3 and 11B = 97.882; F = 332.009 ppm for 11BF3. The reference standard for boron NMR chemical shifts is liquid boron trifluoride-diethyl etherate. We observed in the experiment that 11 B nuclei in BF3OEt2 are 12.91 ppm more shielded than in isolated BF3 molecules. This allows us to establish the absolute shielding scale for boron; the shielding constant in liquid BF3OEt2 is 110.9 ppm for 11B (and similarly 110.8 ppm for 10B). In the present study we have observed the 11B, 10B and 19F resonance frequencies of BF3 in gaseous samples with natural abundance of boron isotopes. The gaseous samples contained BF3 at various densities and a small constant amount of helium-3 (~1 microgram), which was used for the measurement of 3He NMR frequency. For this purpose we used BF3 (99.5%, Aldrich) and 3He (99.993%, Eurisco-Top, France) from lecture bottles without further purification. 4 mm o.d. gas samples were prepared as described earlier [1]. The gas samples were fitted into standard 5 mm o.d. thin-walled NMR tubes (Wilmad 528-PP) with liquid toluene-d8 in the annular space. The deuterium CD3 signal from toluene-d8 was used for the

lock system. The NMR measurements for 11B, 10B, 19F and 3He were carried out on a Varian INOVA-500 spectrometer at 300 K (a detailed description of our experimental work with helium-3 is given in Ref. [2]). At a stable external magnetic field the resonance frequencies were linearly dependent on BF3 density. The experiments were carried out at constant temperature (300K), with decoupling of the boron and fluorine nuclei. For the resonance frequencies of 11B and 10B nuclei we obtained 011B(BF3) = 160.6167935(15) MHz and 010B(BF3) = 53.7842298(16) MHz, respectively. Similar 19F NMR experiments were carried out for the same gaseous samples and after the extrapolation to zero-density we determined the 0 fluorine frequencies for the 11B and 10B isotopomers equal to 470.9782739(20) and 470.9783039(20) MHz, respectively. We also measured the resonance frequency of helium-3 present in the gaseous samples and obtained 0(3He) = 381.3585401(3) MHz. The determination of the nuclear magnetic moment Y from NMR spectra is based on the equation:

Y =

Y (1 X ) X (1 Y ) X

(1)

where Y/X is the experimentally measured gas-phase NMR frequency ratio and X and Y are the computed theoretical shielding constants of nuclei X and Y, respectively. We first applied Eq. 1 to derive the magnetic moments of 11B and 10B (Y) from 3He (X) data, using the discussed above resonance frequencies and shielding constants and the literature values (3He) = -2.12762523466 N [3] and He = 59.937ppm [4]. We obtained 11B = 2.6883781(11)N and 10B = 1.8004636(8)N while the literature values are 11B = 2.6886489(10)N and 10B = 1.80064475(57)N [3]. To check the consistency of the results, we also derive the boron magnetic moments using fluorine data (Y = 11B, 10B; X = 19F). In this case we used as the reference an improved value of the fluorine nuclear magnetic moment determined in our previous work F = 2.6283214(132)N [5] and we obtained 11B = 2.6883642(318)N and 10B = 1.8004543(213)N. As shown the differences between the boron magnetic moments derived using different reference data are 10-5 N for both isotopes, whereas the old literature values differ by more than 10-4 N. References: [1] K. Jackowski, M. Jaszuski, Conc. Magn. Reson. A 30 (2007) 246. [2] K. Jackowski, M. Jaszuski, B. Kamieski, M. Wilczek, J. Magn. Reson. 193 (2008) 147. [3] P. Raghavan, Atomic Data and Nuclear Data Tables 42 (1989) 189. [4] G.W.F. Drake, in Springer Handbook of Atomic, Molecular and Optical Physics, edited by G.W.F. Drake (Springer, 2006), p. 199. [5] A. Antuek, K. Jackowski, M. Jaszuski, W. Makulski, M. Wilczek, Chem. Phys. Lett. 411 (2005) 111. Ackowledgement: This work was partially supported by the Ministry of Science and Higher Education as the research grant No. N204 244134 (2008-2011).