29Si NMR Some Practical Aspects - Pascal-Man

1 29si NMR some Practical AspectsFrank UhligDortmund University Inorganic Chemistry II D-44221 Dortmund, Chr. MarsmannPaderborn University, Inorganic Chemistry, D-33095 Paderborn, , Germany1. IntroductionSilicon is in many respects one of the more important elements in both nature and chemistry. On one hand sili-cates constitute the main material of the earths crust, and on the other hand organo silicon compounds areoften used in element organic chemistry or as building blocks in material science. This is reflected also in lit-erature concerning silicon NMR. For example, one of the fastest growing sections in the last years comprisesthe application to material sciences and here especially the solid state silicon NMR. Of the naturally occurringisotopes 28Si ( ), 29si ( ) and 30Si ( ), 29si only has a spin 1/2and therefore a magneticmoment. This puts it in the same league together with the other elements of 14 group of the periodic table ofthe elements such as carbon, tin and lead.

2 All of these elements, with the exception of germanium, have atleast one isotope with a spin of 1/2(Table 1).208 Gelest, TEL: 035543-1630 FAX: 03-5543-0312 (215) 547-1015 FAX: (215) 547-2484 1 Group 14 elements, parameter of selected drel.[%] ) )to 13C 1 9/2c) ) 1 ) at constant field and equal number of nuclei; b) product of relative sensitivity and natural abundance; c) quadrupolemoment, x 10-28m; d) isotope used in most tin NMR experiments, for other relevant nuclei see literature [1] andcited 1 shows that 29si has a higher share in the isotopic mixture but the absolute value of the magneticmoment is slightly lower than of 13C. This leads to a lower resonance frequency. A complication arises fromthe fact that spin and magnetic moment are antiparallel leading to a negative sign of the gyromagnetic ratio ?

3 Concerning these facts silicon NMR had a slow start. After the first report by Lauterbur et al. in 1962 [2] therehave been a few papers per year only. However, since the beginning of the 80 s this has changed own data collection of 29si chemical shifts now contains about data sets for more than com-pounds [3]. A quick search in literature yields around compounds with a measured 29si chemical of these huge amounts of material available in silicon NMR, all discussions or reviews must be limit-ed to special and selected research fields. Using of standardsThe only magnetic isotope of silicon 29si has a natural abundance of , a spin of 1/2, a magnetic momentof and therefore a receptivity of x 10-4compared to that of 1H. It can be characterized as a mag-netically diluted isotope of medium sensitivity [4]. Similar to 1H or 13C NMR the referencing is mostly donerelatively to tetramethylsilane (Me4Si, TMS) which has the advantages of having a low boiling point, a rela-tively short relaxation time and being a chemically relatively inert substance.

4 Therefore, if nescessary, it canbe added directly to the sample. However, its resonance is in a shift range where the resonances of many otherorganosilicon compounds occur and so misinterpretations are possible. Two general strategies are usable toavoid this problem. The first one is the use of secondary standards. some secondary reference standards asknown from literature are collected in Table 2. Unfortuanately, due to their higher reactivity, in contrast toTMS, they are useful only for a limited number of Gelest, TEL: 035543-1630 FAX: 03-5543-0312 (215) 547-1015 FAX: (215) 547-2484 2 Current and historic reference compounds for silicon NMRName Formula CommonChemical shiftAbbreviation relative to TMSin ppmTetramethylsilaneMe4Si TMS (trimethylsilyl)methane(Me3Si) (Me3Si)2O (Me2 SiO) (MeO)4Si (EtO)4Si (trimethylsiloxy)silane (Me3 SiO) oila)(Me2 SiO) )

5 The use of silicone containing grease for your equipment leads normally also to impurity resonances around 22 common, except for precision measurements, is to use no standard compound at all in the sample (tubeinterchange technique). In such a case the referencing is done relative to a sample containing TMS in the samesolvent as it was used in the unknown values of the silicon chemical shift are to low frequency and high field compared to Me4Si. Specialcare must be given by using silicon chemical shift data from earlier reviews and original papers. some of thememploy the magnetic field definition of chemical shifts instead of the currently accepted frequency based one,resulting in a reversed sign for chemical shift in 29si NMR, pulse techniquesThere are a number of Aspects for running into difficulties in measuring silicon NMR. The first of it concernsthe fact that silicon containing materials such as glass and ceramics constitute a major part of the constructionmaterial of the probe head and probe tube resulting in a broad background signal at approx.

6 -110 ppm. Thereare three general methods to avoid the problem:210 Gelest, TEL: 035543-1630 FAX: 03-5543-0312 (215) 547-1015 FAX: (215) 547-2484 the case of narrow signals, the smallest sweep width possible shouldbe transfer pulse programs such as DEPT otr INEPT can be used if the silicon atoms are cou-pling with protons or fluorine for broad lines are measured, you might be able to subtract it from a blanc spectrum obtained under oth-erwise identical other characteristic feature regarding spectra of organosilicon compounds is usually observed under broadband decoupling of protons. Resonances of silicon atoms containing organic substituents split into many linesby spin-spin couplings with the protons. This is mostly prevented by decoupling experiments. However, theNuclear Overhauser Effect (NOE) can then lead to zero signals, if the ( 29si ,1H) dipol-dipol contribution T1 DDto the other longitudinal relaxation paths of the silicon is close to The relaxation times depend on thecorrelation times of a molecule, therefore, the signal intensity of a 29si spectrum with NOE varies with thetemperature.

7 Again there are three ways to turn around the of shiftless relaxation reagents, for example the well-known Cr(acac)3in a concentration of~10-2 mol/l). of inverse gated decoupling. Here proton decoupling is only active during aquisition with longwaiting times (3 to 5 times the relaxation time T1) between scans. The advantage of not polluting thesample is offset by ineffective use of spectrometer time which can be alleviated somewhat by usingshorter pulses (40 ) and shorter recovery times (20s). of population transfer pulse programs such as INEPT or DEPT [5]. A third disadvantage in measuring silicon NMR spectra is related especially with pure inorganic compoundscontaining only 29si as an, for NMR experiments, useful nucleus. In such cases single pulse experiments areapplicable only. Due to the slow relaxation in such compounds, a 30 pulse is used with repetition rates ofabout state silicon-29 NMR differs from these solutions inasmuch as the spatial interactions are not averagedout by particle motion, resulting in broad lines and all the problems are related to that fact.

8 Literature about29Si solid state NMR is given in reference [6]. shiftsThe majority of 29si NMR shifts are found in a range between +50 and -200 ppm. However, as far as we knowthe current upfield and downfield world records are formed by divalent silicon compounds. The largestupfield shift is measured for the decamethylsilicocene (1) with -392 ppm [7]. The highest downfield shift isgiven for compound 2with 567 ppm [8]. 211 Gelest, TEL: 035543-1630 FAX: 03-5543-0312 (215) 547-1015 FAX: (215) 547-2484 ppm +567 ppmThe border lines for silicon(IV) compounds are given by the tetraiodosilane (3) with a chemical shift of 350ppm [9] and the central silicon atom of the dihypersilylplumbandiyl (4) with +197 ppm [10]. SiI4[(Me3Si)3Si]2Pb34-350 ppm +197 ppmThe name hypersilyl is used mainly as abbreviation for the tris(trimethylsilyl)silyl substituent, as supersilyl isused as abbreviation for the tritbutylsilyl group.

9 Derivatives containing the trimethylsilyl group, (CH3)3Si-, abbreviated also as TMS, form the largest group ofcompounds with a known 29si NMR shift. One reason to introduce one of these groups into a molecule is toobtain a certain substitution pattern in organic chemistry [11]. Another reason is for example the fact to makesubstances now containing Me3 OSi- or Me3 SiN-groups instead of labile protons in HO- or HN-groups easierto handle. For instance the volatility or the solubility in organic solvents is much better for the derivatizedcompound. The TMS group is also useful for the characterization of such compounds by NMR methods [20].The shift of TMS groups ranges from ppm for Me3Si-K+[21] up to ppm for [Me3Si]+[B(C6F5)4]-[22]. An overview about typical shift ranges of trimethylsilyl substituted compounds is given in Scheme 1. 212 Gelest, TEL: 035543-1630 FAX: 03-5543-0312 (215) 547-1015 FAX: (215) 547-2484 1: Typical shift ranges of trimethylsubstituted borderline case examples for each class of compounds are given.

10 One of the first uses of silicon-29 NMR was its application to siloxane polymers. Although it is still possiblesometimes to recognize and isolate individual molecules, the main view is to dissect the molecules into build-ing units (see Scheme 2). RRRRR R O R||||| |||R - Si - O - Si - O - Si - O - Si - O - Si - O - Si - O - Si - O - Si - O- ||||| |||RRRR O R O R||End- (M)Middle- (D) Trifunctional (T) Tetrafunctional (Q)GroupsBranching GroupsM DT(Q)R = Me37 9 ppm Me2-17 to 22Me -55 to 65-105 to -115 Me2Cl ~5Cl2~ -73HO ~ -100 MeCl2 ~ -20Me/H~ -35H~ -85Cl3~ -48Me/OH~ -55Ph~ -78Ph3~ -10Ph/OH~ -70Ph2-42 3 Scheme 2: Building units and shift range of selected siloxanes in metalMe3Si-XMe3Si-P/As/SbMe3Si-NMe3Si-S/ SeMe3Si-OMe3Si-Si/Ge/SnMe3Si-C213 Gelest, TEL: 035543-1630 FAX: 03-5543-0312 (215) 547-1015 FAX.