Uniform Field Re-Entrant Cavity for Improved Pulse EPR

Friday 20 October 2017       Electromagnetics, Published Work, Resonator Design

A recent publication in Applied Magnetic Resonance describes a new uniform field (UF) resonator that has been optimized for pulse experiments. The UF geometry is described as a re-entrant cylindrical TE01U cavity at Q-band (34 GHz). The re-entrant geometry is calculated to have the same electron paramagnetic resonance (EPR) signal intensity as a traditional TE011 cavity, a 60% increase in average resonator efficiency ╬Ťave over the sample, and has a B1 profile that is 79.8% uniform over the entire sample volume (98% uniform over the region-of-interest).

Uniform field resonators were introduced in 2002 by my colleagues at the Medical College of Wisconsin and since then a number of publications have expanded the field to Uniform Field Loop-Gap Resonators, High-Frequency LGR and Cavity Resonators, and recently Mett et al. published a study introducing non-resonant dielectric tubes to a cylindrical Uniform Field TE01U cavity. However, no studies have looked at increasing the resonator efficiency over the sample for pulse spectroscopy. My recent paper addresses this concern.

The advantage of UF resonators is the ability to extend the region-of-interest of a traditional cavity to increase concentration sensitivity by increasing volume. However, as you increase the volume of the resonator the resonator efficiency lowers. To saturate your spins, you would need more power for the same field at the sample. The re-entrant geometry overcomes this issue by introducing fins into the cavity. These fins perform three major functions:

  1. Confine the electric field, providing distributed capacitance, that lowers the cut-off frequency of the cavity. In order to maintain a cut-off frequency of 34 GHz, the overall cavity radius must be reduced, which increases the stored energy.
  2. Provide current, and magnetic field, close to the sample increasing the magnetic field concentration and minimizing magnetic field perturbations that can exist in a typical cavity. The fins can also be thought of as reducing mode coupling to perturbation modes.
  3. By not having the fins in the end section a smaller volume is needed to provide the open circuit match needed to obtain the uniform field. This reduces the volume of the sample that sees non-uniform return field in the end sections.

The end sections were designed with low-loss Rexolite plastic and provide an electromagnetic discontinuity to transfer the end-plate short to an open. Such a design is easy to manufacture with re-entrant geometries at Q- and W-band, unlike loop-gap resonators where the gaps are too small to machine plastic caps. Because of this a true uniform field can be realized.

One important aspect of the re-entrant geometry is the stability of the uniform field mode. Shown on the left is a Uniform Field Cylindrical TE01U resonator (which can include the dielectric tube) as the dielectric constant of the sample is changed from an epsilon of 1 to 5 (dark to light). The dotted lines indicate the 10 mm region-of-interest. This simulates placing different samples in the resonator, for instance changes in glycerol content. The field uniformity swings from a 'W' shape to a slightly cosine shape. And the end sections change significantly. Experimentally this would mean a change in the pi pulse needed to excite the sample. Although this is much better than the traditional TE011 cavity, it could be improved.

On the right shows the same sample change from epsilon 1 to 5 (dark to light). It is apparent that the variation is much smaller with the re-entrant geometry. This is due to the fins reducing the electric field in the sample and, consequently, reducing the frequency shift as the sample is changed. For instance, the cylindrical TE01U geometry has a 1.3 GHz shift, while the re-entrant TE01U geometry only shifts by 810 MHz. This exciting feature of the re-entrant resonator may allow it to become a general purpose resonator for pulse experiments.

Additionally, since the electric field is reduced you could even put a water sample (0.255 ID, 2.8 mm OD quartz tube) and still obtain the uniform field! This thick wall tube would be perfect for high pressure experiments.

The first experiment performed to test the re-entrant resonator was a simple nutation experiment. This experiment relies on coherent preparation pulse, which requires uniform fields. Performing the experiment in a traditional TE011 yields sub-par results. The nutations roll-off quickly, there are phase-inversions at 1200 ns 2400 ns, and there is a complex background. If this was a DEER experiment, these would all create unwanted distances due to frequency content created only by the non-uniformity of the resonator.

By placing the sample in the re-entrant TE01U cavity, these issues disappear and you are left with a clean nutation experiment with reasonable roll-off and no complex background. However, by preparing a sample that sits only in the 10 mm region of interest the nutations can be extended even further.

In the future I plan on testing this resonator using other coherent pulse spectroscopies, such as ESEEM/HYSCORE, DEER, and EDNMR. The larger bandwidth and uniform fields of the re-entrant TE01U with the same concentration sensitivity should provide a superior resonator and pave the way for uniform field resonators to be used for general purpose spectroscopy.

The design drawings and STL field can be found in the Data Section of this web-site.