Solid-state DNP NMR experiments have been carried out on the two classes of microporous compounds investigated in this work, zeolites and MOFs, to better understand their surface interactions and evaluate potential spectroscopic enhancement for challenging nuclei, such as 17O.
Solid-state NMR characterisation of surface sites, while of fundamental importance for the development of new materials with porous and catalytic properties, is affected by sensitivity problems related to the very small concentration of surface sites compared to bulk sites. The introduction of DNP helped to overcome these problems with the transfer of polarisation from unpaired electrons, intrinsic to samples or from exogenous radical solutions, to the nucleus of interest in the sample, by microwave Table 2.3 Spin and magnetic quantum number dependent coefficients for spin I = 3/2 and I = 5/2.62 For the corresponding negative values of mI the sign of the coefficients is
irradiation at the electron paramagnetic resonance (EPR) frequency. The resulting theoretical enhancements (ε) are dependent on the ratio of the gyromagnetic ratios of the electron (e) and the polarised nucleus (x), γe/γx. Theoretically developed in the 1950s by Overhauser,63 DNP was experimentally observed in 195364 and then further investigated over the following decades with experimental applications to thin films of diamonds with intrinsic polarisation sources in the 1990s.65 The introduction of high-power microwave sources, gyrotrons,66 cryogenic MAS probes for experiments at 100 K67 and external polarisation sources in the form of efficient biradicals68 subsequently opened the way for DNP experiments to achieve high enhancements. Indeed, as a result of the better saturation of the EPR transitions, more efficient polarisation mechanisms and with no magnetic field strength limitation, the range of applications was widened to challenging systems, including those of biological origin.
If the materials under analysis are characterised by the absence of intrinsic sources of polarisation, radical molecules need to be introduced in solution without diluting the sample or affecting its chemistry. This has been achieved in this work using optimised stable biradicals, as shown in Figure 2.22, in water-based solutions, e.g. TOTAPOL and AMUPOL, or organic-based solutions, e.g. TEKPol in 1,1,2,2- tetrachloroethane (TCE), depending on the chemistry of the sample. The simplest method used to impregnate porous materials is known as the incipient wetness impregnation (IWI) technique69 and typically involves the use of ∼10 µl of radical solutions (∼15-20 mM) mixed with 10-20 mg of porous sample.
The two main types of mechanisms in DNP experiments of insulating solids are the solid effect (SE), involving one electron and one nuclear spin, and the cross effect (CE), involving a pair of electrons, as in the biradicals shown in Figure 2.22, and a nuclear spin. A third mechanism known as thermal mixing involves multiple dipolar coupled electrons and is not commonly observed as it would experimentally require high radical concentrations. Depending on the homogenous linewidth, δEPR, and the inhomogenous breadth, ΔEPR, of the EPR spectrum of the radical compared to the nuclear Larmor frequency, ω0I, DNP occurs through the solid effect, if δEPR, ΔEPR<ω0I, or the cross effect, if δEPR<ω0I< ΔEPR. The solid effect is less efficient at high magnetic field since it relies on one electron and one nucleus interacting through electron-nuclear coupling and subsequent mixing of nuclear states with microwave irradiation (ω!!) of forbidden transitions at
ω!! =ω!"±ω!" , (2.46) where ω!" and ω!" are the electron and nuclear Larmor frequencies. The applicability of this mechanism is also restricted by the fact that it requires a relatively narrow EPR spectrum of the polarising agent. The cross effect can be
N •O N O• OH N H O TOTAPOL AMUPOL TEKPol N •O O O • O O O N O N H N (CH2CH2O)4Me N O• O• N O O O O
Figure 2.22Biradical molecules used as polarising agents in DNP experiments carried out in this work. TOTAPOL and AMUPOL are used in water-based solution, whereas TEKPol is employed in TCE solutions.
treated as a three-spin system with two dipolar coupled interacting electrons and a nuclear spin achieving maximum polarisation transfer when the Larmor frequency separation of the electrons satisfies the relation
ω!!!−ω!!! =ω!", (2.47) where ω!!! and ω!!! are the Larmor frequencies of the dipolar coupled electrons, S1 and S2, and ω!" is the Larmor frequency of the nucleus I. With this mechanism the best enhancements are obtained using biradical molecules, such as TOTAPOL, with the possibility of fine tuning the electron-electron dipolar coupling on the basis of the length of the alkyl chain of the radical.70-72
Typical experimental temperatures for DNP experiments are in the range of 90-100 K, leading to the freezing of the impregnated solid sample. Polarisation disperses in the bulk of the frozen matrix, containing the radical solution and the sample, through spin diffusion and can either be transferred directly (e−→13C/17O) or more usually indirectly (e−→1H→13C/17O) to the nucleus of interest in the sample. In the latter case, polarisation transfer relies on 1H spin diffusion across the frozen matrix and is transferred to the sample through cross-polarisation (CP), as schematically shown in Figure 2.23 for a MOF sample. Results obtained from the DNP NMR investigation carried out for Al MIL-53 samples will be presented in Section 6.4.
The success of this experiment strongly depends on the efficiency of the polarisation transfer in the final CP step, but has the advantage of relying on the 1H fast polarisation build-up times, characterised by an efficient spin diffusion process. In case polarisation is transferred directly to nuclei (13C,17O) other than 1H, higher enhancements can be potentially obtained, as a result of the higher theoretical γe/γx ratios, and the spectrum obtained is not edited by proximity to 1H. However, a significant disadvantage of this last method is that polarisation build-up times become much longer because of a much less efficient spin diffusion process. This results from the lower gyromagnetic ratios and concentrations of nuclei such as 13C or 17O compared to 1H, and hence a smaller dipolar coupling mediating the process.73 The observed enhancement resulting from DNP experiments can be directly evaluated by comparison of the intensities of spectra recorded with and without microwave irradiation (εon/off). However, this comparison does not take into account other experimental conditions, such as the presence of radical species and the cryogenic temperatures involved, to quantify DNP gains in sensitivity compared to room temperature experiments. Indeed, reduction or quenching of the signal, resulting from the presence of paramagnetic species, and changes in the longitudinal proton relaxation times, determined by thermal and paramagnetic effects, need to be taken into account, as well as a thermal enhancement factor at the typical experimental temperatures (100 K), to determine the overall sensitivity enhancement.74-75
The instrumentation required for the implementation of DNP experiments includes, in addition to the NMR spectrometer, a stable microwave source and a waveguide for the transmission of microwaves to the NMR probe. Considering the microwave power requirements of DNP experiments, gyrotron sources are typically used. A gyrotron is an electronic device that operates under vacuum in static magnetic field and is able to generate high microwave power at high frequencies. Electrons are emitted from an electron gun, accelerated by high voltage and, in the presence of an external magnetic field, they start to gyrate following the magnetic field lines with increasing rotational energy as the magnetic field gets stronger until they reach the cavity region, where emission of microwave radiation occurs by conversion of the electron kinetic energy. Microwaves emitted are then directed to the NMR sample
through the waveguide and the spent electrons are collected within the gyrotron. Probes need to be connected to the waveguide to allow irradiation of the sample and are designed to achieve MAS conditions at temperatures of ∼100 K, since DNP processes are most efficient at these low temperatures. An example of DNP instrumentation is shown in Figure 2.24. Moreover, a sweep coil is normally added to the magnet since, to achieve conditions of maximum enhancement, the frequency of the microwave source or the magnetic field need to be adjusted, but gyrotrons typically operate at fixed frequencies.72,76
Figure 2.24Picture of the UK DNP MAS NMR Facility at the University of Nottingham, where the DNP measurements reported in this work have been carried out.