Trends in EPR Technology
James S. Hyde
Department of Biophysics, Medical College of Wisconsin, 8701 Watertown Plank Road, Milwaukee, WI 53226
Abstract: A personal view of trends in EPR technology is presented. It is unlikely that the fundamental structure of the field will change, but it will be strongly influenced by the rapid increase in computer power, digital storage, and signal processing capability. In the author’s laboratory current themes are resonator enhancement by electromagnetic field finite element modeling, analysis of noise, and digital detection and acquisition of data at multiple microwave frequencies. Some trends foreseen are (1) optimization of resonators for ultra- small samples; (2) step-recovery pulse EPR in which the initial conditions may be established by a step in some experimental condition such as light level or nuclear frequency irradiation; (3) blurring of the distinction between pulse and CW EPR as temporal changes in the resonant condition of a “CW”
measurement are changed in times of the order of spin relaxation times; and (4) increased use of ELDOR.
1. INTRODUCTION
The EPR field is composed of a large number of application areas, each with a small number of active participants. The systems manufacturers produce flexible general purpose spectrometers with numerous accessories in order to serve this fragmented market. Although EPR spectroscopy is a fundamental measurement tool that will remain active indefinitely, and although there will continue to be inventions, discoveries and new applications, it is unlikely that the fundamental structure of the field will change. EPR will continue to be used for research in physics, chemistry and biology to examine samples in the liquid, solid and gas phases over a range of temperature and other conditions.
409
From its earliest days, the underlying technology was based on military developments in radar. In recent years, computers, cellular telephone technology and advances in digital devices have become increasingly important in contributing to the technological foundation of EPR spectrometers.
This reality – that our field is relatively small and dependent to a considerable degree on technology from larger scale development activities – places constraints on future progress of the field. The fundamental technological event of our times is rapid increase in computing power and the performance of associated digital and mass storage devices. Our future, from a technological perspective, will be based on this fact. Advances in EPR digital detection, in data capture and storage as well as in use of advanced signal processing methods are discussed in the chapter on digital detection (see Ch. 7).
Even though EPR instrumentation is strongly dependent on technology developed in other fields, there are areas where the special constraints of EPR have led to significant technological advances. Some of these are discussed below.
2. RESONATORS
Basic contributions to EPR technology that have been developed from within the EPR discipline include resonator development. The requirements of sample access, variable microwave coupling, resonators free from impurities, wall penetrability by high frequency field modulation, temperature control, etc. place constraints on resonator design that we ourselves must face – there is little in the way of technology to borrow from other fields. It is appropriate for the EPR instrumental futurist to predict advances in those specific areas where we control the technology and an attempt is made here.
The Varian multipurpose cavity oscillating in the rectangular mode served as the default EPR resonator for many years. See Rempel et al.
(1964) and Hyde (1995) for details on the design. This structure was
designed to enable a number of specialized EPR experiments as follows: i)
light irradiation, ii) dewar insert for flowing temperature controlled gas, iii)
dewar insert for liquid nitrogen, iv) so-called “flat cells” for aqueous
samples, v) flat cells for tissue samples, vi) a mixing chamber geometry for
stopped and continuous flow EPR, and vii) an electrochemical cell. A collet
system was developed to support the dewars, flat cells, and sample tubes of
various sizes. Two cavity bodies were bolted together to form the dual
sample cavity (Hyde, 1965a) oscillating in the mode in order to
examine a reference sample and a sample of interest simultaneously. A top coupled section was added (Piette et al., 1962) to form a cavity that permitted simultaneous optical absorption and EPR experiments on a sample. The demountable sidewall construction permitted disassembly and cleaning. Dr. Robert Rempel was the leader of this initiative with substantial contributions by Dr. Lawrence Piette, both of whom were Varian scientists.
This outpouring of technology provided the basis for transfer of EPR spectroscopy from its original base in physics into chemistry and biology.
A number of special purpose X-band cavities were developed commercially. A list of some of these includes: cylindrical
“wirewound” cavity (Hyde et al., 1965b, 1966) for enhanced sensitivity in certain classes of samples, ENDOR cavities, cavities for use at cryogenic temperatures, cavity, ELDOR cavities and the cavity for aqueous samples (Hyde, 1975).
This rich array of X-band resonators is the primary reason that this microwave frequency remains dominant in EPR spectroscopy. Nothing comparable exists at either higher or lower microwave frequencies. This is a significant opportunity for future technological development. Certainly all of these X-band capabilities could be replicated at Q-band (35 GHz) and S- band (ca 3 GHz), which would greatly strengthen the concept of
“multifrequency EPR.”
In 1965, the author introduced a system of classification for EPR samples from the perspective of sensitivity and optimum microwave configuration:
namely, eight classes of samples depending on yes or no answers to three questions: i) Does it saturate? ii) Is it limited in size or availability? iii) Does it exhibit substantial dielectric loss (Varian Associates, 1965)? Of course, real samples may not correspond to one of these classes, but would lie in some intermediate category. Nevertheless, this classification system has proven helpful in thinking about resonator design in EPR spectroscopy. To a certain degree, it lies at the heart of the development of loop-gap resonators (LGR) for use in EPR spectroscopy, largely by W. Froncisz and the author (See Froncisz and Hyde, 1982; Hyde and Froncisz, 1989; and ch.
2 in the present volume). This class of resonators fundamentally improves sensitivity for those four classes of samples that correspond to a YES answer concerning question ii): Is the sample limited in size or availability? The benefits with respect to sensitivity can be quite substantial. The X-band LGR with 1 mm diameter sample access hole is one of the enabling technologies for the development of site directed spin labelling. The LGR is also a central technology for in vivo small animal imaging and spectroscopy (see ch 9 and ch. 11 in volume 23.)
Simulation of electromagnetic fields in microwave cavity resonators using
finite-element computer-driven solutions of Maxwell’s equations holds great
promise for future development of EPR resonators. The author is aware of six papers in the literature where finite element modeling of electromagnetic fields was employed in a context of microwave resonators for EPR use. The earliest papers were from A. Schweiger’s group in Zurich (Pfenninger et al., 1988; Forrer et al., 1996). In Pfenninger et al. (1988) the program known as MAFIA (Solution of MAxwell’s equation by the Finite Integration Algorithm) was used as an aid to development of bridged loop gap resonators. The work of this group in this area was updated in Forrer et al., 1996. Also in that year, the MCW group used the Hewlett-Packard version of High Frequency Structure Simulator, HFSS, to develop a bimodal loop gap resonator (Piasecki et al., 1996).
Recently, the author and his colleagues wrote a series of three papers on a new class of microwave resonators that we chose to call uniform field (UF) cavities (Mett et al., 2001; Anderson et al., 2002; Hyde et al., 2002).
Transverse magnetic field (TM) modes exist, that exhibit uniformity of the fields along the axis that is perpendicular to the transverse field. This is because Maxwell’s equations allow magnetic fields that are tangential to conducting end walls to exist. The familiar cylindrical cavity widely used in EPR for aqueous samples is an example of such a structure.
However, previous to these three papers there were no known transverse electric field (TE) modes that exhibited this property of uniformity of the fields along the perpendicular axis. This is because components of the microwave electric field tangential to conducting end walls cannot exist.
The UF modes described in these three papers consisted of a central section where the fields were precisely uniform, with end sections designed to satisfy the boundary conditions at the end walls.
These three papers made extensive use of high frequency structure simulation (HFSS) finite element software. UF resonators were, in fact, discovered using HFSS. We were working on an unrelated problem and noticed some unusual microwave field patterns that we did not understand.
When the structures were built, they worked exactly as designed. Although they were largely theoretical, experimental “proof-of-principle” data were presented. Design of microwave coupling structures and sample entrance structures (so-called “stacks”) for these prototypical resonators was also carried out using HFSS.
The HP HFSS software, which is no longer available, operated in the so-
called driven mode. We found it very difficult to use for microwave cavity
development because every change in the structure required a tedious hunt
for the slightly shifted resonant frequency. When Hewlett-Packard dropped
out of this business, we shifted to Ansoft HFSS (Ansoft Corp., Pittsburgh,
PA). This software incorporates both the driven mode and the 3D modal
frequency or “eigenmode” solution method (Brauer, 1997). This method is
also available in MAFIA. We find that this method is extremely useful for EPR resonator development (Mett et al., 2001; Anderson et al., 2002; Hyde et al., 2002). Since resonators are characterized by a single resonant mode of interest, use of the eigenmode method permits focus only on that mode; a frequency sweep is not needed. Changes to the structure can be made and the resulting effect seen on the fields, resonance frequency and Q-value without having to track the changes through a frequency scan. Second, using the eigenmode method, we can focus on the design of the resonator without introducing a coupling structure. Mechanical drawing time and computation time are each reduced by one to two orders of magnitude due to the increased symmetry of the structure. Introduction of a coupling structure as well as sample access stacks is straightforward in the final stages of a design. We have also used the software to calculate the effects of sample support structures, dewar inserts, and the effect of the dielectric properties of the sample on the electromagnetic field distribution. The resonator efficiency parameter (Hyde and Froncisz, 1989) can be calculated, which permits an estimate of EPR sensitivity. Currently we run this code on a Compaq W8000 workstation with dual Xeon 1.7 GHz processors with 2 GB of RAM.
In another cavity development initiative, we gave ourselves the goal of finding a way to examine a normal X-band sample in a 3 mm i.d., 4 mm o.d.
quartz sample tube at Q-band. The problem was to determine whether the microwave sample access stacks would be beyond cutoff. As the sample diameter increases, the stack, viewed as a cylindrical waveguide propagating in either the cylindrical or cylindrical modes will first become evanescent and eventually radiate power into space. It was found that practical structures for this purpose can be constructed providing that the stack is sufficiently long.
Ansoft offers another software package known as Maxwell 3D that is
well configured for computation of the 100 kHz field modulation patterns
and eddy currents in surrounding metallic structures. Four approaches to the
design of field modulation assemblies for use in EPR can be identified from
the existing literature: i) Use of electrically thin walls – less than one skin
depth at 100 kHz but many skin depths at the microwave frequency. ii) The
wirewound technique introduced by the author for use with the cylindrical
mode (Hyde, 1965, 1966, and Fig. 1). All microwave currents must be
substantially parallel to the wires. iii) Partially cut-through slots for use with
loop gap resonators (Fig. 2). iv) Direct insertion into the resonator of wires
or rods that carry field modulation current. In addition to modeling of fields
produced by modulation coils, Maxwell 3D is well suited for modeling of
fields in a context of ENDOR and RF coils. It is apparent that these various
approaches can also affect the microwave fields, and that the final design will require iteration between both software packages.
Figure 1. Schematic diagram of wire-wound cavity and field modulation coils.
Figure 2. Outline drawing of the one-loop—one-gap resonator used for multifrequency EPR in the frequency range of 0.5 to 8 GHz. Hyde and Froncisz (1989) contains additional detail.
The horizontal slots permit penetration of the 100 kHz magnetic field modulation.
There are several consequences of these advances in simulation of
electromagnetic microwave and modulation fields. i) Design efficiency is
greatly improved, which facilitates development of resonators that are
optimized for specific applications. This, in turn, leads to an increase in
sensitivity for the particular application. Previously, a resonator that was
suitable for a broad class of applications was generally employed, which
often resulted in a compromise with respect to sensitivity. ii) Cavity
coupling structures, sample access stacks, dewar inserts and sample
geometries can now be designed on a rational basis rather than the highly
intuitive basis of the past. iii) A proliferation of resonators for use at
frequencies above and below X-band can be expected that are modeled after
the rich array of X-band designs.
It is concluded from this experience that the software and computing capability has reached the point where optimization of design is better carried out by computer than by experiment. Lower symmetry structures with a large number of design parameters can be investigated by computer modeling that would be impractical to investigate by construction of a series of actual cavities that could be experimentally probed to explore this parameter space.
The conclusion of this section is that we are truly on the verge of a paradigm shift in EPR resonator design that will enable optimization of the structure for any set of sample constraints. The new modeling tools fundamentally increase efficiency in resonator design. One can speculate that commercial design of resonators optimized for specific problems and sold in small numbers may become economically feasible. And there may well be other novel structures waiting to be discovered, in analogy to the discovery of UF modes.
3. NOISE
In the previous section, we discussed optimization of microwave resonator design to maximize the EPR signal. Now we discuss minimizing noise. Together, the goal is to identify future opportunities to improve the signal-to-noise ratio (SNR).
In continuous wave EPR, noise originating in the microwave oscillator and noise originating in the detection system as characterized by the overall receiver noise figure are of primary concern. At high microwave powers incident on the EPR resonator, noise from the oscillator dominates, while at low incident powers, receiver noise dominates. We have found that plots of spectrometer noise versus incident power give considerable insight into spectrometer performance. Six such plots in log-log format are shown in Figs. 3-5. They have been redrawn for display consistency from figures in (Hyde et al., 1982; Froncisz et al., 1986; Hyde et al., 1991; Lesniewski and Hyde, 1990; Pfenninger et al., 1995). Figures 3 and 5 are plots of noise voltage and Fig. 4 is of noise power. Note the differences in scales of the ordinate and abscissa comparing Figs. 3 and 5 with Fig. 4. At low incident powers, noise is independent of power, while at high incident power, the noise voltage varies linearly with incident power. The intercept of these two dependencies results in a “breakpoint” where the noise from each source is the same.
Figure 3a was taken from (Hyde et al., 1982), which was concerned with
phase noise 100 kHz from the carrier frequency of the oscillator when tuned
to the dispersion. The overall receiver noise figure that limits spectrometer
performance at low powers in Fig. 3a is estimated to be 11 dB. Comparison was made between a loop gap resonator with an unloaded Q-value of 650 and a standard Varian multipurpose rectangular cavity with an unloaded Q-value of 7900. Demodulation of phase noise depends on the resonator Q. One expects the ratio of break-point powers to be given by the square of the ratio of Qs, i.e., 22 dB, compared with the measured value of 25 dB, which was considered to be fair agreement. Not only is phase noise less troublesome when using low Q resonators, but the energy density in the LGR employed in this study was 65 times higher than in the cavity. It follows that studies in the dispersion mode can be carried out with the loop gap resonator over a much greater range of energy densities at the sample without encountering source phase noise – about 43 dB.
Figure 3b is similar to 3a, but displays data taken from Froncisz et al.
(1986), which was a comparison at Q-band (35 GHz) of phase noise when using a cylindrical cavity resonator versus a two-loop—one-gap resonator. Comparison of Figs. 3a and 3b suggests the presence of substantially greater receiver noise at Q-band than X-band. The breakpoints occur at higher incident powers even though one would expect about 10 dB more phase noise from the oscillator at Q-band than at X-band (Robins, 1982). This can only occur if the receiver noise figure is substantially higher at Q-band than at X-band.
In order to improve EPR performance at Q-band, a low phase noise Gunn diode oscillator was developed in the author’s laboratory by R.A.
Strangeway (Strangeway et al., 1995). A phase noise level in the range of -132 to -125 dBc/Hz at 100 kHz offset from the central frequency was achieved. This range can be compared with a measured value of -103 of the specific klystron in our Varian Q-band bridge (which was 4 dB worse than the manufacturer’s specification for a new klystron) as well as published values from the literature that fall in the range of -70 to -115. The -115 value was also from a cavity-stabilized Gunn oscillator. See Lesniewski and Hyde (1990) for additional detail. The phase noise that was achieved in Strangeway’s design remains, to the best of our knowledge, the lowest that has been reported at Q-band.
It is apparent from Figs. 3-5 that oscillator phase noise will become more evident if the overall receiver noise figure is reduced. Improvement in spectrometer performance over the widest range of conditions requires simultaneous reduction in receiver and oscillator noise. This concept was developed in detail in Hyde et al. (1991) and key results are displayed in Fig.
4. In that study, several changes were progressively made in the Q-band
microwave bridge configuration, and noise was measured in each
configuration as a function of incident microwave power while tuned to the
dispersion using a cylindrical cavity resonator.
Figure 3. Comparison of LGR and cavity resonators. a) X-band (Hyde et al., 1982) and b) Q-band (Froncisz et al., 1986).
Figure 4. Noise plots at Q-band (Hyde et al., 1991). a) Low phase noise oscillator compared with a klystron in standard bridge; b) comparison with and without an LNA.
Figure 5. Noise plots; a) Reduction of phase noise using a broadband AFC (Lesniewski and Hyde, 1990); b) Comparison of cryogenic and room temperature LNAs (Pfenninger et al., 1995).
In Fig. 4a, phase noise levels at the microwave bridge output utilizing a klystron and the low phase noise Gunn diode oscillator are compared.
Breakpoints differed by 22 dB, which was in fair agreement with the value of 26 dB that was carefully measured on a microwave test bench. The data for this figure were obtained using a 1N53D point contact detector diode.
The Gunn oscillator experiment of Fig. 4a was repeated using a more modern balanced mixer detector, and the results are shown in the lower graph of Fig. 4b. (Note that the results of Fig. 4a and 4b are not directly comparable because the gain of the preamplifier following microwave detection was reduced in the experiment of Fig. 4b.) A low noise microwave amplifier (LNA) with a specified noise figure of 4.3 dB and a gain of 25.5 dB was then inserted into the microwave bridge. Results are shown in the upper graph of Fig. 4b.
Notably, at low power the noise increased by only 14 dB, although the gain increased by 25.5 dB. A value for the noise figure of the balanced mixer 100 kHz from the carrier is readily calculated: 25.5 – 14 + 4.3 = 15.8 dB. This is not a very good number relative to X-band standards, but it was a considerable improvement relative to the 1N53D. At high powers, the two graphs in Fig. 4b are displaced by 25.5 dB, the gain of the LNA, as expected.
The breakpoints are separated by 13 dB, in good agreement with the low power separation of 14 dB, demonstrating consistency of the data.
Figure 5a is derived from Lesniewski and Hyde (1990). It shows data at 19 GHz using a microwave bridge with a varactor-tuned Gunn oscillator.
An AFC circuit was developed where the feedback loop had high gain at 100 kHz from the carrier. This enabled reduction of phase noise by 23 dB. This is viewed as an interesting and relatively unexplored strategy for reduction of microwave phase noise.
Pfenninger et al. (1995) reports data obtained at X-band using a
cryogenic microwave preamplifier cooled to the temperature of liquid
nitrogen with a 0.5 dB noise figure corresponding to 35 K equivalent noise
temperature. As expected, phase noise originating from the microwave
oscillator becomes more evident as shown in Fig. 5b. Data in this figure
were obtained using a loop gap resonator observing either the dispersion or
the absorption. The cryogenic amplifier had 15 db higher gain and 3 dB
lower noise figure than the room temperature LNA with which it was
compared, resulting in about 12 dB separation of the noise levels, as
observed. The separation of the breakpoints for dispersion and absorption
was about 35 dB. This value is dependent on the resonator Q, reference arm
power, amplifier gains, and other details of the bridge setup and is not
necessarily the same for the two amplifiers. Excess oscillator noise was
clearly observable in the absorption mode configuration when using the
cryogenic amplifier, and it would have been more apparent if a high Q cavity
resonator had been utilized. Detailed analysis presented in this paper as well as in Rinard et al. (1999) predicted further improvement if the microwave circuit elements in the bridge were to be cooled. This paper is cast not only in the context of CW EPR, but also pulse EPR.
From the perspective of “Future Trends,” modern LNAs operating at room temperature seem adequate. Cooling of the amplifier combined with cooling of microwave bridge components may be desirable for selected experiments, but seems unlikely to become the general practice. However, improvement in microwave oscillator phase noise in CW EPR is feasible and desirable for two reasons: i) With an LNA in place, simultaneous detection of dispersion and absorption followed by a Hilbert transform of the dispersion and then addition of the two signals results in a highly desirable 3 dB improvement in SNR. Inspection of Figs. 3-5 indicates that phase noise remains excessive under practical conditions. Further reduction in oscillator phase noise is required in order to attain this improvement at higher levels of incident power. ii) As seen from the discussion concerning resonators, high Q cavities yield optimum performance for many classes of samples. In this case, reduced phase noise will be of value both for dispersion and absorption.
The Gunn diode itself is inherently noisy: the noise figure is on the order of 25 dB. Modern microwave transistors that could be used in an oscillator circuit have a noise figure on the order of a few dB. Continued improvements in transistor-based microwave oscillators have led to the recent development of low noise Yttrium-Iron-Garnet (YIG) oscillators. For example, standard manufactured YIG oscillators with a ±1 GHz electronic tuning range, 14 dBm power output and a center frequency anywhere in the 3 - 1 0 GHz range have a typical phase noise of -125 dBc/Hz at 100 kHz offset (Korber et al., 2002). Further improvement in future years can be anticipated.
Microwave frequency translation is employed in an increasing number of modern EPR microwave bridges. Examples include multiquantum EPR (Hyde, 1998a), time-locked sampling (ch. 7 in this volume), and translation of complex irradiation patterns formed at X-band to Q-band and also W- band for sample irradiation and then back to X-band for detection as employed in current Bruker spectrometer systems. Frequency translation can introduce additional sources of noise. The appendix of Strangeway et al.
(1995) discusses several different approaches to frequency translation. One is to mix the output of a microwave oscillator with the output of a frequency synthesizer. Noise aspects of frequency synthesizers in the context of EPR spectroscopy are discussed in ch 7 of this volume.
As noise originating in the EPR spectrometer decreases, it is possible that
noise originating in the environment will become dominant. Environmental
noise includes microphonics in the audio frequency range and below, line frequency noise, instabilities in connections to ground, radio frequency interference, and temperature instabilities. The author has encountered each of these. It seems likely that future EPR spectrometer installations will be in shielded rooms with independent connections to ground, much as in magnetic resonance imaging installations. Mass storage devices, for example RAID arrays, permit storage of raw data with no signal averaging or filtering. This will allow inspection of the data for spurious environmental noise. It will also permit development of optimum filters of environmental noise.
EPR spectra are often obtained as a function of one or more additional experimental parameters, for example, temperature or time. Almost always in modern spectrometers the information is available in digital format, permitting the use of a variety of digital filters. Ideally, such filtering should be performed in multidimensional parameter space making use of all available information. An example is provided in ch 6 of volume 23, by N.
Hogg. The EPR literature in this area is sparse, but will undoubtedly increase in the years ahead. In the author’s view this is a particularly promising opportunity for further improvement of the SNR in EPR spectroscopy.
4. MULTIFREQUENCY EPR
In 1963, Varian initiated a project to develop a commercial EPR bridge at Q-band (35GHz) headed by the author. The bridge circuit was the first from Varian to employ a circulator and reference arm. It was a modest commercial success – about 100 units were sold over the ensuing years.
Analysis of the business opportunity presented by development of a bridge operating at a third microwave frequency, S-band, was not favorable.
After establishment of the National Biomedical EPR Center at the Medical College of Wisconsin, funds became available to develop an S-band bridge for use in collaborative projects at the center. The project was headed by Wojciech Froncisz, visiting from Jagiellonian University, Krakow, Poland. A commercially available tunable 2 to 4 GHz (octave bandwidth) microwave oscillator was used in the bridge. Although microwave cavities were initially used, the ungainly size of these structures led directly to the development of loop gap resonators.
In the ensuing years, octave bandwidth bridges were developed from 0.5
to 1, 1 to 2, and 4 to 8 GHz, resulting in continuous coverage from 0.5 to 8
GHz in coaxial transmission line configurations. An additional bridge was
developed at 18 GHz, resulting in EPR capabilities at discrete frequencies in
three narrow bands centered at 9.5, 18 and 35 GHz. This multifrequency capability was viewed as a national resource available to all EPR spectroscopists.
The 18 GHz bridge was designed to be suitable both for CW and saturation recovery (SR) EPR. An S-band module was developed for use with Froncisz’s S-band bridge, resulting in multifrequency capabilities for saturation recovery at S-, X-, and K-band (see ch. 1 in this volume, by Eaton and Eaton).
Bruker has developed an approach that may well simplify the problem of achieving pulse multifrequency capabilities. The firm’s W-band (94 GHz) bridge is based on an X-band bridge with advanced pulse capabilities. The output of the X-band bridge is mixed with the output of an 84 GHz oscillator to form a W-band microwave source, and the reflected signal from the W- band resonator is translated back to X-band for detection. The same strategy has recently been developed at Q-band, translating again from X-band, Tradeoffs made by this approach have not yet been analyzed in the literature.
Several initiatives were undertaken at MCW to arrive at multifrequency ENDOR and ELDOR capabilities in the S-band range. Newton and Hyde (1991) observed that the ENDOR enhancement effect increases inversely with the applied static magnetic field, which makes S-band a favorable microwave frequency for ENDOR of low moment nuclei such as They demonstrated ENDOR of the P1 center in diamond. Pace et al. (1993) used S-band ENDOR to observe hyperfine interaction of and nuclei in trinitrophenylmethylnitroxide. Christidis et al. (1994) describe a probehead with both interchangeable LGRs and RF coils for multifrequency ENDOR in the range of 1 – 10 GHz. An ELDOR capability at S-band using two separate S-band bridges and a bimodal resonator was described by Piasecki et al. (1996).
Extensions of X-band multiquantum EPR to MQ ENDOR (Mchaourab et al., 1993) and electron-nuclear-electron triple resonance (Christidis et al., 1996) were reported. MQ ELDOR is described by Mchaourab et al. (1991).
The center in Milwaukee has recently completed a Q-band bridge with
conventional CW and ELDOR capability, MQ and MQ ELDOR capability
and saturation recovery. Use of all five displays for ENDOR seems
straightforward. The sample and resonator need not be changed when
switching between modes of operation. Historically, each of these advanced
capabilities was developed one at a time; it is now possible to combine them
in a single microwave configuration. Translation to another microwave
frequency following Bruker’s approach seems feasible, although
multiquantum EPR presents a special problem. The translation processes
may generate unacceptable levels of intermodulation sidebands. A possible
solution to this problem would be to translate the two closely spaced
irradiating microwave frequencies independently and combine them after translation.
The continuous range of microwave frequencies that is available from 0.5 to 8 GHz presents an opportunity for future instrumental development. To date, the number of resonators available in this range in the author’s laboratory is small – two in the S-band range and one in each of the other three octave bandwidths. Development of a multiplicity of resonators over this entire range seems impractical. What would suffice, as an example, is four resonators each with octave bandwidth resonator tunability in order to arrive at a true overall four-octave capability. Figures 6 and 7 are cartoons of possible resonator designs that indicate how this capability might be achieved. Figure 6 illustrates a re-entrant cavity. The re-entrant rods lie parallel to the electric field in regions of electric field maxima. They short out the field, reducing the resonant frequency. Rotation of the central section, as indicated in the figure, will dramatically shift the resonant frequency. Figure 7 illustrates a tunable bridged loop gap resonator (Pfenninger et al., 1988; Symons, 1995). As the plastic tuner is rotated, the capacitance of the resonator (gap in parallel with the two bridge capacitances in series) changes. Automated collection of spectra over an octave bandwidth at closely spaced microwave frequencies would appear to be possible, providing that the scientific rationale for the capability is sufficiently strong to merit the technological development effort
A general weakness in the multifrequency approach is that the wide range of resonator configurations presently available at X-band seems unlikely to be developed over the full range of frequencies of interest.
However, it does seem feasible, as mentioned above, to develop a range of resonators at one or two additional frequencies, for example, S- and Q-band.
Figure 6. Variable frequency re-entrant rectangular cavity.
Figure 7. Variable frequency bridged loop gap resonator.
5. EPR FOR ROUTINE ANALYSIS
The use of EPR spectroscopy in routine commercial applications began as early as 1961, when Saraceno, Fanale and Coggeshall of Gulf Oil (Saraceno et al., 1961; Saraceno, 1963) developed techniques for using EPR to measure vanadium in crude oil. The vanadium poisoned catalysts used in refineries.
Under contract with Gulf, Varian Associates developed a portable EPR spectrometer with various automatic calibration features that permitted its reliable operation by unskilled personnel in the environment of a refinery (Nelson and Baker, 1967).
Another effort occurred in 1971 when President Nixon requested and Congress agreed to the establishment of a Special Action Office of Drug Abuse, primarily for compulsory treatment and rehabilitation of addicted Vietnam veterans. Syva Corporation, a cooperative venture between Syntex and Varian, developed the Frat™ EPR spin immunoassay systems to monitor morphine in the urine of addicts. The EPR instrument was highly automated and approximately 50 spectrometers were built. Many were used in extremely difficult conditions in Vietnam, making several million examinations. See Hyde (1998b) for additional details.
In 1973, Hwang and Pusey developed an oil well logging process for determining paleotemperatures (Pusey, 1973; Hwang and Pusey, 1973).
They measured the EPR free radical signal that occurs naturally in kerogen.
The process was made commercially available by Robertson Research International, Ltd.
One of the most promising applications for routine EPR analysis is for
radiation dosimetry using the tissue-equivalent alanine dosimeter. The
National Institute of Standards and Technology (NIST) offers a number of
services based on EPR analysis of the free radical content in alanine pellets exposed to high doses of ionizing radiation, including calibration services.
Bruker offers an automated spectrometer, the e-scan, specially configured for this measurement. In 2002, a news release issued jointly by Kodak and Bruker describes a system for alanine dosimetry based on coating of microcrystalline alanine on a film, which is sold under the trademark Biomax. The system includes a bar code and alphanumeric identifier for sample tracking. The news release describes the system as an unmatched method for high-volume verification of radiation in processes ranging from the sterilization of medical devices to decontamination of food.
Another industrial application of EPR was introduced by Uchida and Ono in 1996 (Uchida and Ono, 1996). They employed a spin trapping strategy to measure the quantity of anti-oxidants in beer (Uchida et al., 1996;
Uchida and Ono, 2000). This measurement correlated with shelf-life of the product. A patent was issued in 1998 (Ono and Uchida, 1998) for the process. Bruker has also configured the e-scan spectrometer for this application.
Modern approaches to EPR spectrometer design can be expected to result in enhanced performance and reduced cost for commercial applications of EPR, providing that the markets are sufficiently large. Using finite element modeling of microwave fields, resonator design and sample geometry can be optimized efficiently for each application. In addition, microwave circuits can be miniaturized and optimized by computer design. It is highly desirable to use the smallest magnet gap possible. In a patent entitled
“Narrow Cavity Low Cost EPR Spectrometer” (Hyde, 1987), the author used a or cavity just 2 mm wide in order to reduce the mass of the magnet. Loop gap resonators provide ample opportunity to reduce magnet size in process-control or medical diagnosis applications. Modern PCs plus digital detection permit the equivalent of phase sensitive detection in the computer (ch. 7, this volume), manipulation of the gain and bandwidth of automatic frequency control circuits, and incorporation of quite sophisticated signal processing software.
Process control applications of EPR have been slow to develop, but the success that Bruker has been experiencing in recent years indicates progress.
Newer computer-based engineering design tools, the introduction of digital detection, and modern computer-based signal analysis tools can be expected to provide the technological basis for further progress.
Development of medical diagnosis applications of EPR that are widely
used has been a long-standing goal, but has remained elusive. It has become
clear that success can only be achieved if there is a sound scientific rationale
for EPR spectroscopy in comparison with other more familiar spectroscopic
methods. Spin trapping in a context of free radical chemistry offers one
opportunity, since it can be argued that EPR is generally a preferred spectroscopy when free radicals are involved.
6. DISCUSSION
From the author’s own laboratory, several “mini-trends” that may flourish in the years ahead can be identified:
1) Optimization of resonators for ultra-small samples, i.e., submicroliter. In a context of loop gap resonators, electric discharge machining (EDM) permits cutting of very small gaps. In a poster (Camenisch et al., 2002), a Q-band resonator with a sample volume as small as 30 nl containing 1 picomole of protein was successfully used in a multiquantum experiment.
2) Step recovery pulse EPR. The concept is to generalize the step of incident microwave power used in saturation recovery to include other steps in experimental conditions including light level, temperature, pH value, stop flow mixing of reacting species, applied electric field, nuclear irradiation level in ENDOR, pressure, etc., and observe the time-dependent evolution to a new equilibrium condition. Observation of the transient response using either spin echo methods or CW methods is appropriate. The technological challenge lies in appropriate delivery of the step. Change in sample composition or temperature can result in changes in microwave resonance frequency or coupling, that require dynamic adjustment. Changes in the EPR field/frequency relationship during approach to a new equilibrium may require attention.
3) Blurring of the distinction between CW EPR and pulse EPR. High frequency field modulation is a form of pulse EPR, since it involves inversion of the magnetization during each sweep of the instantaneous modulation field through resonance. Multiquantum EPR is a temporal modulation of the saturation conditions that can be described equally well in the time domain or the frequency domain. MQ in the presence of high frequency field modulation has been demonstrated in unpublished work in the author’s laboratory.
4) Increased use of ELDOR. Intense saturation transfer mechanisms
are one of the distinguishing characteristics of EPR spectroscopy. The
technological challenge in exploiting this opportunity lies largely in
resonator development. Frequency sweep of the pumping microwaves is
highly desirable, but technically challenging. These methods were
pioneered by the author in the 1960s. The hypothesis can be made that
modern technology creates opportunities for improved bimodal resonators
with frequency sweep of the pumping mode. ELDOR is a powerful method that remains underutilized because of technological complexities.
The material in this chapter has been largely derived from the author’s experience and must be considered to be a personal view of technological trends in EPR. Several significant instrumental trends, including developments in pulse spin echo EPR, in vivo EPR and ultra-high frequency EPR, are not discussed. However, the themes of this article, viz, resonator enhancement by electromagnetic field finite element modeling, analysis of noise, digital detection and acquisition of data at multiple microwave frequencies, are relevant to pulse, in vivo and ultra-high frequency EPR.
The chapter by Freed and Peisach in this volume gives additional information on pulse EPR and high frequency EPR, and the chapters by Swartz (ch.10) Halpern (ch. 11), and Krishna (ch. 12) in volume 23 provides an overview of in vivo EPR. The recent monograph on pulse EPR by Schwieger and Jeschke (2001) provides a foundation for future progress in that field.
7. REFERENCES
Anderson, J.R., Mett, R.R. and Hyde, J.S. (2002). Cavities With Axially Uniform Fields For Use In Electron Paramagnetic Resonance. II. Free Space Generalization. Rev. Sci.
Instrum. 73, 3027-3037.
Brauer, J.R. (1997). IEEE Trans. Microwave Theory Tech., 45, 810.
Camenisch, T.G., Klug, C.R., Ratke, J.J., Hubbell, W.L. and Hyde, J.S. (2002). Multiquantum EPR of Arrestin K267C-MT SL at 35 GHz. Proc. Rocky Mountain Conf. on Analytical Chemistry, p. 43.
Christidis, T.C., Froncisz, W., Oles, T. and Hyde, J.S. (1994). Probehead with Interchangeable Loop-Gap Resonators and RF Coils for Multifrequency EPR/ENDOR. Rev. Sci. Instrum. 65, 63-67.
Christidis, T.C., Mchaourab, H.S. and Hyde, J.S. (1996). Hyperfine Selectivity Using Multiquantum Electron-Nuclear-Electron Triple Resonance. J. Chem. Phys. 104, 9644- 9646.
Forrer, J., Pfenninger, S., Sierra, G. Jeschke, G. and Schweiger, A., Wagner, B. and Weiland, T. (1996). Probeheads And Instrumentation For Pulse EPR And ENDOR Spectroscopy With Chirped Radio Frequency Pulses And Magnetic Field Steps. Appl. Magn. Reson. 10, 263-279.
Froncisz, W. and Hyde, J.S. (1982). The Loop-Gap Resonator: A New Microwave Lumped Circuit ESR Sample Structure. J. Magn. Reson. 47, 515-521.
Froncisz, W., Oles, T. and Hyde, J.S. (1986) Q-Band Loop-Gap Resonator. Rev. Sci. Instrum.
57, 1095-1099.
Hwang, P.T.R. and Pusey, W.C. (1973) U.S. Patent 3,740,641.
Hyde, J.S. (1995). Electron Paramagnetic Resonance in Handbook of Microwave Technology, Volume 2, Ishii, T. K., ed., pp. 365-402, Academic Press, New York.
Hyde, J.S. (1965a). Gyromagnetic Resonance Spectroscopy, U.S. Patent 3,197,692.
Hyde, J.S. (1965b). ENDOR Of Free Radicals In Solution. J. Chem Phys. 43, 1806-1818.
Hyde, J.S. (1966). Microwave Cavity Resonator, U.S. Patent 3,250,985.
Hyde, J.S. (1975). EPR Spectrometer Resonant Cavity, U.S. Patent 3,878,454.
Hyde, J.S. (1987). Narrow Cavity Low Cost EPR Spectrometer, U.S. Patent. 3,931,569.
Hyde, J.S. (1998a). Multiquantum EPR. in Foundations of Modern EPR, G.R. Eaton, S.S.
Eaton and K.M. Salikhov, eds., pp. 741-757, World Scientific, New York.
Hyde, J.S. (1998b). EPR at Varian: 1954-1974. in Foundations of Modern EPR, G.R. Eaton, S.S. Eaton and K.M. Salikhov, eds., pp. 695-716, World Scientific, New York.
Hyde, J.S. and Froncisz, W. (1989). Loop Gap Resonators. In Advanced EPR: Applications in Biology and Biochemistry, A.J. Hoff, ed., pp. 277-306, Elsevier, Amsterdam.
Hyde, J.S., Froncisz, W. and Kusumi, A. (1982). Dispersion Electron Spin Resonance with the Loop-Gap Resonator. Rev. Sci. Instrum. 53, 1934-1937.
Hyde, J.S., Newton, M.A., Strangeway, R.A., Camenisch, T.G. and Froncisz, W. ( 1991).
Electron Paramagnetic Resonance Q-Band Bridge with GaAs Field-Effect Transistor Signal Amplifier and Low-Noise Gunn Diode Oscillator. Rev. Sci. Instrum. 62, 2969- 2975.
Hyde, J.S., Mett R.R. and Anderson, J.R. (2002). Cavities With Axially Uniform Fields For Use In Electron Paramagnetic Resonance. III. Re-Entrant Geometries. Rev. Sci. Instrum.
73, pp. 4003-4009.
Korber, MR., Teuthorn, D. and Lockie, D. (2002). YIGs Tune High-Speed Millimeter-Wave Radios. Microwaves & RF, pp. 92-100 (July).
Lesniewski, P. and Hyde, J.S. (1990). Phase Noise Reduction of a 19 GHz Varactor-Tuned Gunn Oscillator for Electron Paramagnetic Resonance Spectroscopy. Rev. Sci. Instrum.
61, 2248-2250.
Mchaourab, H.S., Christidis, T.C., Froncisz, W., Sczaniecki, P.B. and Hyde, J.S. (1991).
Multiple-Quantum Electron-Electron Double Resonance. J. Magn. Reson. 92, 429-433.
Mchaourab, H.S., Christidis, T.C. and Hyde, J.S. (1993). Continuous Wave Multiquantum Electron Paramagnetic Resonance Spectroscopy. IV. Multiquantum Electron-Nuclear Double Resonance. J. Chem. Phys. 99, 4975-4985.
Mett, R.R., Froncisz, W. and Hyde, J.S. (2001). Axially Uniform Resonant Cavity Modes For Potential Use In Electron Paramagnetic Resonance Spectroscopy. Rev. Sci. Instrum.
72, 4188-4200.
Nelson, F.A. and Baker, G.A. (1967), Gyromagnetic Resonance Apparatus Utilizing Two Sample Signal Comparison, U.S. Patent 3,348,136.
Newton, M.E. and Hyde, J.S. (1991). ENDOR at S-Band (2-4 GHz) Microwave Frequencies.
J. Magn. Reson., 95, 80-87.
Ono, M. and Uchida, M. (1998). Analytical Method For Evaluating Flavor Stability Of Fermented Alcoholic Beverages Using Electron Spin Resonance, U.S. Patent 5,811,305.
Pace, M.D., Christidis, T.C. and Hyde, J.S. (1993). S-Band ENDOR of Hyperfine Interactions of and Nuclei in Trinitrophenylmethylnitroxide. J. Magn. Reson., 102A, 101-104.
Pfenninger, S., Forrer, J. and Schwieger, A. (1988). Bridged Loop-Gap Resonator: A Resonant Structure For Pulsed ESR Transparent To High-Frequency Radiation. Rev. Sci.
Instrum. 59, 752-760.
Pfenninger, S., Froncisz, W. and Hyde, J.S. (1995). Noise Analysis of EPR Spectrometers with Cryogenic Microwave Preamplifiers. J. Magn. Reson. 113A, 32-39.
Piasecki, W., Froncisz, W. and Hyde, J.S. (1996). Bimodal Loop-Gap Resonator. Rev. Sci.
Instrum. 67, 1896-1904.
Piette, L.H., Ludwig, P., Adams, R.N. (1962). EPR And Electrochemistry. Studies Of Electrochemically Generated Radical Ions In Aqueous Solution, Anal. Chem., 34, 916.
Pusey III, W.C. (1973). How To Evaluate Potential Gas And Oil Source Rocks. World Oil, April
Rempel, R.C., Ward, C.E., Sullivan, R.T., St. Clair, M.W., Weaver, H.E. (1964).
Gyromagnetic Resonance Method and Apparatus, U.S. Patent 3,122,703.
Rinard, G.A., Quine, R.W., Song, R., Eaton, G.R., Eaton, S.S. (1999). Absolute EPR Spin Echo and Noise Intensities. J. Magn. Reson. 140, 69-83.
Robins, R.J. (1982). Phase Noise in Signal Sources, Peregrinus Ltd. on behalf of the Institute of Electrical Engineers, Herts, England.
Saraceno, A.J., Fanale, D.I. and Coggeshall, N.D. (1961). An Electron Paramagnetic Resonance Investigation of Vanadium in Petroleum Oils. Anal. Chem 33, 500.
Saraceno, A.J. (1963). Determination of Vanadium Content of Hydrocarbon Oils by Electron Paramagnetic Resonance Spectrometry, U.S. Patent 3,087,888 (reissued 1967).
Schwieger, A. and Jeschke, G. (2001). Principles Of Pulse Electron Paramagnetic Resonance, Oxford University Press, New York.
Strangeway, R.A., Mchaourab, H.S., Luglio, J., Froncisz, W. and Hyde, J.S. (1995). A General Purpose Multiquantum Electron Paramagnetic Resonance Spectrometer. Rev. Sci.
Instrum. 66, 4516-4528.
Symons, M.C.R. (1995). Whole Body Electron Spin Resonance Imaging Spectrometer, In Bioradicals Detected By ESR Spectroscopy, H. Ohya-Nishiguchi and L. Packer, eds., pp.
93-102, Birkhäuser Verlag, Basel, Switzerland.
Uchida, M. and Ono, M. (1996). Improvement For Oxidative Flavor Stability Of Beer – Role Of OH-Radical In Beer. J. Am. Soc. Brew. Chem 54, 198-204.
Uchida, M., Suga, S. and Ono, M. (1996). Improvement For Oxidative Flavour Stability Of Beer – Rapid Prediction Method For Beer Flavour Stability By Electron Spin Resonance Spectroscopy. J. Am. Soc. Brew. Chem 54, 205-211.
Uchida, M. and Ono, M. (2000). Technological Approach To Improve Beer Flavor Stability:
Analysis Of The Effect Of Brewing Processes On Beer Flavor Stability By The Electron Spin Resonance Method. J. Am. Soc. Brew. Chem. 58, 8-13.
Varian Associates Technical Information Bulletin: Signal Amplitudes in Electron Paramagnetic Resonance, Fall 1965, pp. 10-13.
