Clear signals from surfaces

Article metrics

Nuclear magnetic resonance is a versatile analytical technique, but acquiring well-resolved NMR spectra of chemical surfaces has been hard. The coming of age of a spectral enhancement method should change all that.

For many decades, nuclear magnetic resonance (NMR) studies of surfaces have promised to provide detailed information about reaction mechanisms involving solid catalysts, but many of the results have been limited by the low signal-to-noise ratio of the experiments. With the recent development of advanced dynamic nuclear polarization (DNP) techniques to enhance NMR sensitivity, however, this situation has changed dramatically. Reporting in the Journal of the American Chemical Society, Lesage et al.1 describe an outstanding example of DNP in which the NMR signals of molecules attached to silica surfaces were enhanced approximately 50-fold. Such an increase in sensitivity could revolutionize NMR studies of surfaces.

The physicist Albert Overhauser first proposed2 the idea of DNP in NMR experiments in 1953, and the concept was demonstrated experimentally by Charles Slichter and co-workers3 shortly thereafter. NMR spectroscopy involves the use of radio-frequency electromagnetic radiation to excite polarized (aligned) nuclear spins in a magnetic field. But the spin polarizations achieved — and therefore the signal-to-noise ratio of the resulting NMR spectra — are low. However, the spin polarization of electrons in paramagnetic compounds (such as stable free radicals) is hundreds or thousands of times larger than that of nuclei. The DNP technique therefore involves transferring spin polarization from electrons in a paramagnetic compound to the nuclei in a surface sample. This is accomplished by irradiating the electron paramagnetic resonance (EPR) spectrum, the electronic analogue of the NMR spectrum, with microwaves, thereby exciting electron–nucleus transitions and transferring polarization.

In the 1980s, DNP was combined with magic-angle spinning (MAS), an NMR technique used to obtain high-resolution spectra of solids, to enhance the sensitivity of NMR for studying polymers and other materials4,5,6. These were 'low-field' experiments — they used a relatively low magnetic field (1.5 tesla), low radio frequencies (60 megahertz for NMR of 1H nuclei) and low microwave frequencies (40 GHz for EPR). But at that time, MAS was rapidly moving towards using higher magnetic fields (5–20 T) and radio frequencies (200–850 MHz for 1H), which offer greater resolution and sensitivity. To obtain large signal enhancements from DNP in such high-field experiments requires microwave sources operating at 130–600 GHz. Such sources were not readily available at the time, and so DNP–MAS failed to take off as a solid-state analytical technique. DNP therefore resumed its former position as an interesting intellectual curiosity.

The 1990s witnessed the development of several pieces of instrumentation that altered this landscape considerably. For instance, 1993 saw the introduction of a class of microwave oscillators known as gyrotrons7 for DNP. The first gyrotrons8,9,10 provided continuous microwave power at 140 and 250 GHz, but more recent devices11,12 do so at up to 460 GHz (which could be used for 1H DNP–NMR at 700 MHz). Microwave sources corresponding to the highest available NMR frequencies (1,000 MHz) are now on the horizon. The fact that DNP functions optimally at low temperatures has also necessitated the development of a new generation of cryogenic MAS probes that operate at temperatures of 90 kelvin and below13,14.

Another major development was the discovery of innovative polarizing agents. In the first 50 years of DNP experiments, researchers used single electrons (mainly from organic free radicals) as polarizing agents. These act through a mechanism known as the solid effect, which involves two spins — the electron's spin and the spin of the nucleus to be polarized. But a more efficient process involving three spins was demonstrated in 2004, with the development of biradical polarizing agents15. Compared with experiments in the absence of DNP, 250-fold signal enhancements have been observed16 using such biradicals. Currently, the favourite polarizing agent is a water-soluble biradical17 known as TOTAPOL, which typically yields 170-fold enhancements in model systems and in proteins.

Lesage and colleagues' exciting NMR experiments1 bring together all of the state-of-the-art developments in DNP-enhanced NMR. They used an NMR spectrometer equipped with a 263-GHz gyrotron, operating at 90 K with TOTAPOL as the polarizing agent, to enhance the spectra of organic groups covalently attached to the surface of porous silica. Surface-modified silica is commonly used in many applications, and is a good case study for various other systems whose surface chemistry is ripe for DNP–NMR studies. The authors observed a 50-fold enhancement of the carbon-13 NMR signals for the silica-bound groups (Fig. 1), which allowed the acquisition of 13C spectra in approximately 30 minutes. In the absence of DNP, these experiments would have taken about 70 days. Refinements to the technique could eventually yield approximately 500-fold signal enhancements.

Figure 1: Enhanced NMR signals from surfaces.

Lesage et al.1 report an impressive signal enhancement for NMR spectra of organic groups attached to a silica surface using a technique known as dynamic nuclear polarization (DNP). a, Without DNP, the authors observed no sharp peaks in the carbon-13 spectrum of the surface groups. b, With DNP, they obtained a 50-fold improvement in the NMR signal. The spectra in a and b are shown at the same scale. Peaks at different chemical shifts correspond to different carbon atoms in the surface groups. Spectral data are from ref. 1; p.p.m. denotes parts per million.

The mechanism by which the enhanced NMR signals1 are most frequently generated involves polarizing the solvent in which the silica material is suspended. This polarization is transferred to the surface of the material (and probably deeper than that) through spin diffusion processes. Other experiments18,19 on nanocrystals and membranes have shown that polarization can be transferred over distances of around 1,000 ångströms, so it may be possible to examine the structure of a material not only at the surface, but also in the layers immediately below. This is one of the many possibilities envisaged by Lesage and colleagues1.

The authors' experiments represent a huge step forward in the study of reactions that occur at the surfaces of bulk solids, including many scientifically and industrially important reactions on solid catalysts. The application of DNP–NMR to such systems will undoubtedly stimulate many new avenues of research. More generally, the authors' work1 represents another example of the application of DNP to heterogeneous systems (those that involve more than one phase of matter), whose structures are often difficult to determine.


  1. 1

    Lesage, A. et al. J. Am. Chem. Soc. 132, 15459–15461 (2010).

  2. 2

    Overhauser, A. W. Phys. Rev. 92, 411–415 (1953).

  3. 3

    Carver, T. R. & Slichter, C. P. Phys. Rev. 92, 212–213 (1953).

  4. 4

    Wind, R. A., Duijvestijn, M. J., van der Lugt, C., Manenschijn, A. & Vriend, J. Prog. Nucl. Magn. Reson. Spectrosc. 17, 33–67 (1985).

  5. 5

    Afeworki, M. & Schaefer, J. Macromolecules 25, 4092–4096 (1992).

  6. 6

    Singel, D. J., Seidel, H., Kendrick, R. D. & Yannoni, C. S. J. Magn. Reson. 81, 145–161 (1989).

  7. 7

    Felch, K. L. et al. Proc. IEEE 87, 752–781 (1999).

  8. 8

    Becerra, L. R., Gerfen, G. J., Temkin, R. J., Singel, D. J. & Griffin, R. G. Phys. Rev. Lett. 71, 3561–3564 (1993).

  9. 9

    Gerfen, G. J. et al. J. Chem. Phys. 102, 9494–9497 (1995).

  10. 10

    Bajaj, V. S. et al. J. Magn. Reson. 189, 251–279 (2007).

  11. 11

    Matsuki, Y. et al. Phys. Chem. Chem. Phys. 12, 5799–5803 (2010).

  12. 12

    Torrezan, A. C. et al. IEEE Trans. Plasma Sci. 38, 1150–1159 (2010).

  13. 13

    Barnes, A. B. et al. J. Magn. Reson. 198, 261–270 (2009).

  14. 14

    Thurber, K. R. & Tycko, R. J. Magn. Reson. 195, 179–186 (2008).

  15. 15

    Hu, K.-N., Yu, H., Swager, T. M. & Griffin, R. G. J. Am. Chem. Soc. 126, 10844–10845 (2004).

  16. 16

    Matsuki, Y. et al. Angew. Chem. Int. Edn 48, 4996–5000 (2009).

  17. 17

    Song, C., Hu, K.-N., Joo, C.-G., Swager, T. M. & Griffin, R. G. J. Am. Chem. Soc. 128, 11385–11390 (2006).

  18. 18

    van der Wel, P. C. A., Hu, K.-N., Lewandowski, J. & Griffin, R. G. J. Am. Chem. Soc. 128, 10840–10846 (2006).

  19. 19

    Barnes, A. B. et al. Phys. Chem. Chem. Phys. 12, 5861–5867 (2010).

Download references

Author information

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Griffin, R. Clear signals from surfaces. Nature 468, 381–382 (2010) doi:10.1038/468381a

Download citation

Further reading


By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.