Journal of Magnetic Resonance 209 (2011) 277–281

Contents lists available at ScienceDirect

Journal of Magnetic Resonance

journal homepage: www.elsevier.com/locate/jmr

Reconstruction of the first-derivative EPR spectrum from multiple harmonics

of the field-modulated continuous wave signal
Mark Tseitlin, Sandra S. Eaton, Gareth R. Eaton ⇑
Department of Chemistry and Biochemistry University of Denver, Denver, CO 80208, United States

a r t i c l e i n f o a b s t r a c t

Article history: Selection of the amplitude of magnetic field modulation for continuous wave electron paramagnetic res-
Received 6 December 2010 onance (EPR) often is a trade-off between sensitivity and resolution. Increasing the modulation amplitude
Revised 24 January 2011 improves the signal-to-noise ratio, S/N, at the expense of broadening the signal. Combining information
Available online 3 February 2011
from multiple harmonics of the field-modulated signal is proposed as a method to obtain the first deriv-
ative spectrum with minimal broadening and improved signal-to-noise. The harmonics are obtained by
Keywords: digital phase-sensitive detection of the signal at the modulation frequency and its integer multiples.
Continuous wave EPR
Reconstruction of the first-derivative EPR line is done in the Fourier conjugate domain where each har-
Digital phase sensitive detection
Harmonics
monic can be represented as the product of the Fourier transform of the 1st derivative signal with an ana-
Magnetic field modulation lytical function. The analytical function for each harmonic can be viewed as a filter. The Fourier transform
Signal reconstruction of the 1st derivative spectrum can be calculated from all available harmonics by solving an optimization
problem with the goal of maximizing the S/N. Inverse Fourier transformation of the result produces the
1st derivative EPR line in the magnetic field domain. The use of modulation amplitude greater than line-
width improves the S/N, but does not broaden the reconstructed spectrum. The method works for an arbi-
trary EPR line shape, but is limited to the case when magnetization instantaneously follows the
modulation field, which is known as the adiabatic approximation.
Ó 2011 Elsevier Inc. All rights reserved.

1. Introduction method is applicable for inhomogeneously broadened EPR spectra

that can be represented as a superposition of Lorentzian lines.
If the concentration of paramagnetic centers is low and the Modeling the effects of over-modulation can be used to estimate
acquisition time is limited, it may be difficult to achieve a signal- the linewidth that would have been observed at lower modulation
to-noise ratio (S/N) for a continuous wave (CW) EPR signal that is amplitude [7,8]. Bikineev et al. proposed reconstruction of an EPR
sufficient for reliable data analysis. Increasing the magnetic field spectrum, without prior knowledge of the lineshape, by applying
modulation is one way to improve the S/N [1,2], at the expense the maximum likelihood method to multiple harmonics of an
of line broadening. Two methods have been used to model the ef- over-modulated signal [9].
fects of over-modulation on an EPR lineshape. Pseudo-modulation In most EPR spectrometers the 1st harmonic spectrum is ob-
is a mathematical description of magnetic field modulation as a fil- tained as the output of a phase sensitive detector operating at
ter that can be applied to experimental or calculated EPR lines the magnetic field modulation frequency. In the limit of low mod-
[3,4]. Convolution of the filter with the EPR line produces a spec- ulation amplitude the 1st harmonic is a good approximation of the
trum that is broadened by over-modulation. The method is appli- first-derivative EPR signal. The 2nd harmonic (second derivative)
cable within an adiabatic approximation, which means that the has been found to be informative for resolution enhancement
magnetization follows magnetic field instantaneously. The effect [10–12] and for saturation transfer EPR [13]. Higher harmonics
of the modulation frequency on EPR lineshape is not taken into ac- have not been recorded for two reasons: (i) low S/N and (ii) if
count. The second approach is based on the solution of the Bloch phase-sensitive detection (PSD) is done in hardware, detection of
equations for a single spin 1/2 particle [5,6]. Both modulation each harmonic would require a separate phase shifter, video
amplitude and modulation frequency are taken into account. This amplifier, and digitizer channel. The first problem can be partially
solved by over-modulation. If the modulation amplitude is smaller
⇑ Corresponding author. Address: Department of Chemistry and Biochemistry,
than the linewidth, the S/N of the nth harmonic decreases rapidly
University of Denver, 2101 E. Wesley Ave., Denver, CO 80210, United States. Fax: +1
with n. However, for a strongly over-modulated line the higher
303 871 2254. harmonics have much higher S/N [2,14]. It has been shown that
E-mail address: [email protected] (G.R. Eaton). simultaneous fitting of multiple harmonics of over-modulated

1090-7807/$ - see front matter Ó 2011 Elsevier Inc. All rights reserved.
doi:10.1016/j.jmr.2011.01.027
278 M. Tseitlin et al. / Journal of Magnetic Resonance 209 (2011) 277–281

 
n hm u
Sn ðuÞ ¼ j J n GðuÞ ð1Þ
2
where Jn(x) is the Bessel function of the first kind, G(u) is the Fourier
transform of the absorption (not a derivative) EPR signal g(B), and
hm is the peak-to-peak modulation amplitude. Using the property
of the Fourier transform that integration in the B-domain is equiv-
alent to division by ju in the Fourier-conjugate u-domain, Eq. (1) can
be rewritten as follows:
 
ðn1Þ hm u 1
Sn ðuÞ ¼ j Jn u FðuÞ ð2Þ
2
Here F(u) is the Fourier transform of the first-derivative EPR line
f(B), where f(B) = dg(B)/dB. Using the recurrence relation for the Bes-
sel function:
 
J n1 ðzÞ þ J nþ1 ðzÞ z
J n ðzÞ ¼ ;
2n
Fig. 1. Comparison of the filter functions |D1(u)| to |D5(u)| calculated for a
modulation amplitude, hm, of 3 G (solid lines), with the Fourier transforms of first Eq. (2) can be expressed in a form that does not have u in the
derivative lineshapes with peak-to-peak linewidths of 1 G (dashed line) or 9 G (dot- denominator:
dashed line).
Sn ðuÞ ¼ Dn ðuÞFðuÞ ð3Þ
spectra could improve the accuracy of the recovered spectrum where
[15]. The second problem can be solved by using digital PSD, where       
hm ðn1Þ hm u hm u
a single detection channel suffices to obtain multiple harmonics of Dn ðuÞ ¼ j J n1 þ J nþ1 ;n > 0 ð4Þ
4n 2 2
the spectrum. The nth harmonic is denoted as sn(B). The spectrum
sn(B) can be calculated by multiplication of the digitized signal by This rearrangement is done to avoid division by zero at u = 0 when
the nth harmonic of the sinusoidal reference followed by digital using this relationship. The complex function Dn(u) is a filter that
low-pass filtering. A few labs have implemented this method converts F(u) into the nth harmonic spectrum in the u-domain. To
[15–18]. Bruker recently announced a Signal Processing Unit, make Eq. (3) applicable to experimental data, which always contain
SPU, with the capability to simultaneously digitally detect up to noise, a noise term rn(u) is added to Eq. (3):
five harmonics [http://www.bruker-biospin.com]. With this inno-
vation multiple harmonics of an EPR spectrum can be routinely
Sn ðuÞ ¼ Dn ðuÞFðuÞ þ ð1 þ jÞrn ðuÞ ð5Þ
measured with a commercial spectrometer. The standard deviations of the noise in the u-domain are assumed
In this paper a method is demonstrated to process multiple to be equal for the real and imaginary parts. The subscript n in
harmonics of an EPR spectrum recorded with over-modulation Sn(u) indicates that the noise might be different for different har-
to calculate the familiar 1st derivative EPR line with improved monics. In an experiment the bandwidth of the detection system
S/N, without prior knowledge of the lineshape. Bikineev et al. limits the maximum number of harmonics, NH, that can be mea-
solved the inverse problem for the entire digitized over-modu- sured without significant intensity attenuation.
lated signal [9]. In this paper a simpler and more intuitive way Fig. 1 compares the filter functions |Dn(u)|, n = 1, 5 for modula-
to derive expressions similar to what were obtained in Ref. [9] tion amplitude hm = 3 G, with the Fourier transforms, |F(u)|, of first-
is shown. The impact on S/N of the modulation amplitude and derivative Lorentzian lineshapes with peak-to-peak linewidths
of the number of harmonics that are used to reconstruct the spec- (DHpp) of 1 and 9 G. For the broad line, the filters overlap exten-
trum is examined. The problem is solved in the domain that is the sively with F(u), which is not the case for the narrow line. The
Fourier conjugate of the magnetic field domain. The Fourier trans- S1(u) functions for those two lines can be obtained by multiplica-
form of sn(B) can be represented as the result of multiplication of tion of the Fourier transforms of the lineshapes F1G(u) or F9G(u)
the Fourier transform of the 1st derivative spectrum by an analyt- by the filter D1(u). For the broader line (DHpp = 9 G) multiplication
ical function that depends only on the modulation amplitude and by D1(u) (for hm = 3 G) does not significantly distort F9G(u). The in-
number of the harmonic, n. Each function has an oscillatory verse Fourier transform of the product would produce a slightly
behavior and can be considered as a filter with a set of zeros broadened EPR spectrum. However, for the narrower line
(Fig. 1). The filters corresponding to different values of n have (DHpp = 1 G) multiplication of F1G(u) by D1(u) (for hm = 3 G) signif-
maxima and zeros at different positions. If only one harmonic is icantly changes F1G(u), so the Fourier transform of S1(u) is a se-
measured, the information about the Fourier transform of the first verely broadened spectrum.
derivative spectrum at the position corresponding to the filter’s Because of its oscillating behavior D1(u) not only acts as a low-
zero is lost. However, other filters have significant intensity at pass filter, it also destroys information about F1G(u) in the vicinity
that position, so that the information is preserved if multiple har- of its zeros (u  0.4, 0.75, . . .). Because of zeros in D1(u), recovering
monics of the spectrum are available. Appropriate combination of the first-derivative EPR line from the first harmonic spectrum S1(u)
the harmonics permits restoration of the undistorted Fourier by deconvolution is problematic. One may try to obtain an undis-
transform of the EPR spectrum. The reconstruction procedure is torted line by division F(u) = S1(u)/H1(u). However, since the
described in this paper and tested by numerical simulation and denominator H1(u) has zeros, small variation in the noise in S1(u)
experiment. may be amplified. This situation is known as an ill-posed problem.
It can be solved by regularization that produces a distorted but sta-
2. Theory ble solution [20,21].
Fig. 1 not only shows the problem but also suggests a solution.
It was shown in [19] that the Fourier transform of the nth har- Filter D3(u) has a maximum at the position of the first zero of D1(u),
monic EPR spectrum sn(B) can be written in the following form: which means that the over-modulated 3rd harmonic preserves
 M. Tseitlin et al. / Journal of Magnetic Resonance 209 (2011) 277–281 279

information about the function F1G(u) at u  0.4. In that region F(u) oyl-2,2,5,5-tetramethylpyrrolin-1-yloxy-d12 (mHCTPO) in water in
can be obtained by division S3(u)/D3(u). Since the denominator is a 4 mm quartz tube was degassed and sealed. The CW EPR spec-
not zero, the solution would be stable. In the same manner other trum of mHCTPO has two widely separated lines as a result of
D1(u) zeros can be ‘patched’ by the higher harmonics. Thus, the hyperfine interaction of the unpaired electron with the 15N nu-
first-derivative EPR signal can be reconstructed based on process- cleus. Each nitrogen hyperfine line consists of two partially over-
ing of multiple harmonics of the EPR spectrum in the u-domain. lapping lines with about 0.27 G peak-to-peak linewidth, due to
The solution of that problem is sought in the form of a about 0.5 G hyperfine splitting by the proton at the 4-position of
combination: the ring [24]. The spectrum of the low-field nitrogen hyperfine line
vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi was measured in this study.
uN
X
NH uX H X
NH
S¼ an Sn ¼ FD þ ð1 þ jÞt jan j2 r2n ; D¼ an Dn ð6Þ
3.2. Spectroscopy
n¼1 n¼1 n¼1

Here (and in subsequent equations) the argument u of the functions Experiments were done with a Bruker E500T spectrometer
S, D, an and r is omitted. The coefficients an ¼ jan jexpðjuan Þ are the using 10 kHz magnetic field modulation. The resonator was a crit-
complex weighting factors and uan are the phases. The fact that the ically-coupled EN4118X-MD5, which has Q  2000 for the empty
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffir1
sum of two sets of random numbers having standard deviations resonator. The SPU unit, although designed to measure up to five
and r2 results in a set with standard deviation equal to r21 þ r22 is harmonics, currently only outputs the 1st and 2nd harmonics. To
used to derive Eq. (6). obtain the higher harmonics, phase-sensitive detection was done
The function F(u) can be found from Eq. (6) as follows: numerically in a PC. The field-modulated EPR signal was digitized
ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
sP with a LeCroy Wave Runner 44Xi-A oscilloscope. The input to the
NH 2 2
S n¼1 jan j rn scope or SPU was filtered with a 4th order Butterworth low-pass
¼ F þ ð1 þ jÞErr; Err ¼ PNH ð7Þ
D n¼1 an Dn
filter (Krohn-Hite 3955). The cut-off frequency of the filter was
100 kHz, so up to 10 harmonics of the modulation frequency could
Division of S(u) by D(u) (Eq. (7)) gives the desired function F(u) plus be measured. The signal was over-sampled at a rate of 2 MS/s to
a term that originates from the noise, which is (1 + j) Err(u). A set of compensate for the low 8 bit vertical resolution and minimize
weighting factors an(u) is sought that minimizes function Err(u), at the quantization error. An HP 3110A function generator was used
each value of u. Although the numerator of Err(u) depends only on to synchronize data acquisition by the LeCroy and the spectrome-
|an(u)|, the denominator depends on an(u), which are complex. For a ter’s software, XEPR. The output of the function generator was con-
given set of an(u), Err(u) is minimized when the denominator is nected to the ‘EXT TRIG’ port on the SPU and to the trigger input of
maximized. This can be achieved when all the terms in the sum the LeCroy. After the trigger, XEPR initiated a sweep of the external
are real and positive, which requires magnetic field and measurement of the 1st harmonic EPR spec-
an Dn ¼ jan Dn j ð8Þ trum. The same signal was sent to the SPU and the LeCroy. Data
from the LeCroy were transferred to the PC and the harmonics of
This is achieved when uan = -uDn, where uan(u), uDn(u) are the the spectra were calculated using a Matlab program. The numerical
phases of the complex functions an(u) and Dn(u), respectively. Since demodulation method described in [18] was implemented. The 1st
the phases of the weighting functions are defined by Eq. (8), the harmonic spectra measured with the XEPR software and by digital
goal is to find the set of cn(u) = |an(u)| for which the error term is demodulation of the digitized modulated signal were in good
minimized. This occurs when all derivatives of Err(u) with respect agreement, which confirmed that digital phase-sensitive detection
to cn(u) are equal to zero. is equivalent to the analog data processing.
2v ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
u NH , 3, Spectra of the LiPc sample were recorded and harmonics calcu-
uX XNH
@4 t cn rn
2 2 cn jDn j5 @ ck ¼ 0; k ¼ 1; NH ð9Þ lated for 21 modulation amplitudes between 0.2 and 4 G. For each
n¼1 n¼1 modulation amplitude, data were recorded three times. Harmonics
were processed as described above to obtain the 1st derivative EPR
The solution of Eq. (9) is spectra. The result was three reconstructed EPR spectra at each hm.
cn ¼ CjDn j=r2n ð10Þ Spectra of mHCTPO were recorded and harmonics calculated for
nine values of the modulation amplitudes between 0.13 and
where C is an arbitrary constant. The result, that the weighting fac- 1.12 G. Measurements were repeated three times.
tor of a particular harmonic in the sum Eq. (6) is proportional to
|Dn(u)| and inversely proportional to the noise variance, is quite rea- 3.3. Results
sonable. The larger the amplitude of the filter and the smaller the
noise level, the better information about F(u) that is preserved in For comparison of the signals obtained by the standard method
Sn(u). If the noise level is the same for all harmonics, Eq. (10) sim- and the proposed reconstruction method, parameters were se-
plifies to Eq. (11). lected that broadened the signal to the same extent. In the stan-
cn ¼ CjDn j ð11Þ dard method, broadening and S/N are affected both by the
modulation amplitude and the characteristics of the low-pass fil-
ter. In the proposed method, broadening and S/N are determined
3. Experimental only by the filtering that is selected in the post-processing.
For the traditionally detected 1st harmonic EPR spectra s1(B) of
3.1. Samples LiPc the modulation amplitude hm was 377 mG (about 50% of
DHpp), which caused about 6% increase in DHpp. The experimental
Lithium phthalocyanine (LiPc) prepared electrochemically fol- data were filtered with a low-pass 4th order Butterworth filter in
lowing procedures in the literature [22,23] was provided by Prof. the PC software to suppress high frequency noise. The cut-off fre-
Swartz, Dartmouth University. A single very small crystal was used quency was selected so that the line was just slightly broadened.
for the measurements. A sample in a 4 mm tube was open to the In the proposed method f(B) were filtered with the same low-pass
air. The EPR spectrum of the LiPc crystal had a Lorenzian lineshape filter but with a lower cut-off frequency. It was adjusted so that
with DHpp = 0.77 G. An 0.1 mM solution of 15N-4-hydro-3-carbam- over-filtering produced 6% broadening, analogous to the effect of
280 M. Tseitlin et al. / Journal of Magnetic Resonance 209 (2011) 277–281

1
0.8
0.6
0.4

Intensity, a.u.
0.2
0
-0.2
-0.4
-0.6
-0.8
-1
3425 3430 3435 3440 3445
Magnetic Field, G
Fig. 3. Comparison of the traditional first harmonic EPR spectrum s1(B) of LiPc
measured with hm/DHpp  0.5 (red line) and f(B) reconstructed from NH = 10 with
Fig. 2. Ratio of the S/N for f(B) to S/N for s1(B) as a function of the ratio of the hm/DHpp  2.4 (blue dashed line).
modulation amplitude (hm) to the peak-to-peak first derivative linewidth (DHpp) for
the LiPc sample. Blue solid lines with error bars represent experimental results for
three measurements and dashed red lines are results of numerical simulations. The
S/N comparison was made for f(B) spectra reconstructed from NH = 5 (a) and NH = 10 A
(b) harmonics and s1(B) obtained at constant low modulation amplitude. 1 s1: hm=0.13 G
s1: hm=1.12 G
0.8
C f: hm=1.12 G
0.6
over-modulation on the traditionally-detected signal. This ap-
proach was used to permit comparison of S/N for s1(B) and f(B) 0.4
Intensity, a.u.

spectra that are broadened by the same amount. 0.2

The ratio of the S/N for reconstructed f(B) as a function of 0
hm/DHpp for LiPc to that of s1(B) traditionally obtained with modu-
-0.2
lation amplitude that gives only 6% broadening is shown in Fig. 2
for three independent experimental data sets (blue1 solid curves -0.4
with error bars) and compared with numerical simulations (red -0.6
dashed). The noise level was assumed to be the same for all harmon- B
x 0.33
-0.8
ics. For hm < 0.37DHpp the modulation is smaller than selected for D
s1(B) so the S/N obtained by reconstruction from multiple harmonics -1
is poorer than for s1(B). However, it is the same as would have been 3416.5 3417 3417.5 3418 3418.5 3419
obtained by traditional detection at the same modulation amplitude. Magnetic field, G
To benefit from the multiple harmonics, higher modulation ampli-
tudes are needed. The simulations show that for NH = 5 the [S/ Fig. 4. Comparison of s1(B) measured with hm = 0.13 G (solid green) with f(B)
(dashed blue) recovered from spectra at harmonics n = 1–10, measured with
N(f)]/[S/N(s1)] ratio has a maximum of 1.7 at hm/DHpp  1.2 and de-
hm = 1.12 G (s1(B), solid black). To facilitate comparison of the lineshapes the
creases slowly for increasing modulation amplitudes. The decreasing amplitude of the over-modulated s1(B) spectrum measured with hm = 1.12 G is
S/N at higher modulation amplitudes arises from noise amplification decreased by a factor of 3.
if higher harmonics are not included. For NH = 10 [S/N(f)]/[S/N(s1)] is
relatively constant at about 1.75 for hm/DHpp between 1.5 and 3.
Experimental results are in a good agreement with the simulations
(Fig. 2). The comparison of s1(B) and f(B) shown in Fig. 3 demon- O2 that is not sufficient for reliable data analysis. Fig. 4 demon-
strates the S/N improvement that can be obtained by using multiple strates that with the method described in this paper the minimally
harmonics. broadened lineshape (f(B), dashed blue) can be recovered from a se-
Recovery of f(B) from multiple harmonics of the over-modulated verely over-modulated spectrum (s1(B), black solid). The recovered
EPR spectrum of the proton hyperfine-split low-field 15N hyperfine spectrum is in excellent agreement with s1(B) recorded at low mod-
line of mHCPO also was tested (Fig. 4). mHCTPO was synthesized to ulation amplitude.
be used for EPR oxymetry [24]. The ratio K = (A  D)/(C  B)
between inner and outer peak-to-peak heights (see Fig. 4) is sensi-
4. Discussion
tive to the oxygen concentration. Parameter K increases when
oxygen broadens the overlapping lines. Over-modulation causes
In a conventional CW EPR experiment only the 1st harmonic
broadening of the lines that is indistinguishable from that due to
signal is measured. In this paper a method is described that utilizes
oxygen. Thus O2 measurements can be compromised if the modula-
information that is contained in multiple harmonics of the field-
tion amplitude is too large. Since in the general case the concentra-
modulated EPR signal. The 1st derivative spectra reconstructed
tion of O2 is not known, the modulation amplitude in a conventional
by using the proposed algorithm not only show improved S/N
experiment must be set low enough that it does not broaden the
but also are not broadened by over-modulation. It is also important
oxygen-free spectrum, which may result in S/N in the presence of
that the method is applicable to an EPR signal of an arbitrary shape,
provided that the magnetization instantly follows the magnetic
1
For interpretation of color in Figs. 1–4, the reader is referred to the web version of field, which is typical for CW in the absence of power saturation
this article. or passage effects. This approach frees a spectroscopist from the
 M. Tseitlin et al. / Journal of Magnetic Resonance 209 (2011) 277–281 281

difficult trade-off between S/N and lineshape distortion in selecting [9] V.A. Bikineev, E.A. Zavatskii, V.V. Isaev-Ivanov, V.V. Lavrov, A.V. Lomakin, V.N.
Fomichev, K.A. Shabalin, Methods of correcting modulation-induced lineshape
modulation amplitude.
distortions in electron paramagnetic resonance spectroscopy, Tech. Phys. 40
Insensitivity to the modulation amplitude comes with a cost. (1995) 619–625.
The S/N of the recovered 1st derivative EPR spectrum becomes [10] W.E. Antholine, Low frequency EPR of Cu2+ in proteins, Biol. Magn. Reson. 23
more sensitive to the characteristics of the low-pass filter. This (2005) 417–454.
[11] T. Halpern, W.D. Phillips, Third-derivative detection of EPR signals, Rev. Sci.
sensitivity occurs because D(u), which is in the denominator of Instrum. 41 (1970) 1038–1039.
the calculation of F(u) (Eq. (7)), decreases with increasing u, result- [12] B. Smaller, E.L. Yasaitis, Paramagnetic resonance detection using dual field
ing in amplification of the noise at higher frequencies. Selection of modulation, Rev. Sci. Instrum. 24 (1953) 991–992.
[13] D. Marsh, L.I. Horváth, T. Páli, V. Livshits, Saturation transfer spectroscopy of
a lower cut-off frequency of the filter efficiently reduces noise, but biological membranes, Biol. Magn. Reson. 24 (2005) 309–367.
over-filtering can broaden the EPR spectrum. In comparison with [14] C.P. Poole, Magnetic field scanning and modulation, in: C.P. Poole (Ed.),
the broadening caused by over-modulation, the effect of over-fil- Electron Spin Resonance, Wiley, New York, 1983, pp. 212–258.
[15] M.P. Tseitlin, O.A. Tseitlin, Using of digital demodulation of multiharmonic
tering is reversible, since it is done in software. The filter character- overmodulated EPR signals to improve EPR oximetry reliability, Appl. Magn.
istics can be selected to optimize S/N or resolution, without doing Reson. 36 (2009) 25–34.
another measurement. [16] R. Ahmad, S. Som, E. Kesselring, P. Kuppusamy, J.L. Zweier, L.C. Potter, Digital
detection and processing of multiple quadrature harmonics for EPR
spectroscopy, J. Magn. Reson. 207 (2010) 322–331.
Acknowledgments [17] J.S. Hyde, T.G. Camenisch, J.J. Ratke, R.A. Strangeway, W. Froncisz, Digital
detection by time-locked sampling in EPR, Biol. Magn. Reson. 24 (2005) 199–
222.
Funding for this work from NIH NIBIB P41 EB002034 (Howard J.
[18] M. Tseitlin, V.S. Iyudin, O.A. Tseitlin, Advantages of digital phase-sensitive
Halpern, PI) is gratefully acknowledged. detection for upgrading an obsolete CW EPR spectrometer, Appl. Magn. Reson.
35 (2009) 569–580.
References [19] P.A. Berger, H.H. Gunthart, The distortion of electron spin resonance signal
shapes by finite modulation amplitudes, Z. Angew. Math. Phys. 13 (1962) 310–
323.
[1] G.R. Eaton, S.S. Eaton, D.P. Barr, R.T. Weber, Quantitative EPR, Springer-Verlag/ [20] F. Momo, S. Colacicchi, A. Sotgiu, Limits of deconvolution in enhancing the
Wein, New York, 2010. resolution of ESR imaging experiments, Measur. Sci. Technol. 4 (1993) 60–64.
[2] G.V.H. Wilson, Modulation broadening of nuclear magnetic resonance and [21] Y. Deng, G. He, P. Kuppusamy, J.L. Zweier, Deconvolution algorithm based on
electron spin resonance line shapes, J. Appl. Phys. 34 (1963) 3276–3285. automatic cutoff frequency selection for EPR imaging, Magn. Reson. Med. 50
[3] J.S. Hyde, M. Pasenkiewicz-Gierula, A. Jesmanowicz, W.E. Antholine, Pseudo (2003) 444–448.
field modulation in EPR spectroscopy, Appl. Magn. Reson. 1 (1991) 483–496. [22] P. Turek, J.J. Andre, A. Giraudeau, J. Simon, Preparation and study of a lithium
[4] R.D. Nielsen, E.J. Hustedt, A.H. Beth, B.H. Robinson, Formulation of Zeeman phthalocyanine radical: optical and magnetic properties, Chem. Phys. Lett. 134
modulation as a signal filter, J. Magn. Reson. 170 (2004) 345–371. (1987) 471–476.
[5] B.H. Robinson, C. Mailer, A.W. Reese, Linewidth analysis of spin labels in liquids [23] V.O. Grinberg, A.I. Smirnov, O.Y. Grinberg, S.A. Grinberg, J.A. O’Hara, H.M.
I. Theory and data analysis, J. Magn. Reson. 138 (1999) 199–209. Swartz, Practical conditions and limitations for high spatial resolution of
[6] B.H. Robinson, C. Mailer, A.W. Reese, Linewidth analysis of spin labels in multi-site EPR oximetry, Appl. Magn. Reson. 28 (2005) 69–78.
liquids. II. Experimental, J. Magn. Reson. 138 (1999) 210–219. [24] H.J. Halpern, M. Peric, T.D. Nguyen, D.P. Spencer, B.A. Teicher, Y.J. Lin, M.K.
[7] C. Mailer, B.H. Robinson, B.B. Williams, H.J. Halpern, Spectral fitting: the Bowman, Selective isotopic labeling of a nitroxide spin label to enhance
extraction of crucial information from a spectrum and a spectral image, Magn. sensitivity for T2 oxymetry, J. Magn. Reson. 90 (1990) 40–51.
Reson. Med. 49 (2003) 1175–1180.
[8] Y. Deng, R.P. Pandian, R. Ahmad, P. Kuppusamy, J.L. Zweier, Application of
magnetic field over-modulation for improved EPR linewidth measurements
using probes with Lorentzian lineshape, J. Magn. Reson. 181 (2006) 254–261.