Room-Temperature Electron Spin Relaxation of Triarylmethyl Radicals at the X- and Q-Bands

Article abstract of DOI:10.1021/acs.jpcb.5b03027

Triarylmethyl radicals (trityls, TAMs) represent a relatively new class of spin labels. The long relaxation of trityls at room temperature in liquid solutions makes them a promising alternative for traditional nitroxides. In this work we have synthesized a series of TAMs including perdeuterated Finland trityl (D₃₆ form), mono-, di-, and triester derivatives of Finland-D₃₆ trityl, the deuterated form of OX63, the dodeca-n-butyl homologue of Finland trityl, and triamide derivatives of Finland trityl with primary and secondary amines attached. We have studied room-temperature relaxation properties of these TAMs in liquids using pulsed electron paramagnetic resonance (EPR) at two microwave frequency bands. We have found the clear dependence of phase memory time (T_m T₂) on the magnetic field: room-temperature T_m values are 1.5-2.5 times smaller at the Q-band (34 GHz, 1.2 T) than at the X-band (9 GHz, 0.3 T). This trend is ascribed to the contribution from g-anisotropy that is negligible at lower magnetic fields but comes into play at the Q-band. In agreement with this, the difference between T₁ and T_m becomes more pronounced at the Q-band than at the X-band due to increased contributions from incomplete motional averaging of g-anisotropy. Linear dependence of (1/T_m - 1/T₁) on viscosity implies that g-anisotropy is modulated by rotational motion of the trityl radical. On the basis of the analysis of previous data and results of the present work, we conclude that, in the general situation where the spin label is at least partly mobile, the X-band is most suitable for application of trityls for room-temperature pulsed EPR distance measurements.

Full text of DOI:10.1021/acs.jpcb.5b03027

Subscriber access provided by NEW YORK UNIV
Article
Room-Temperature Electron Spin Relaxation
of Triarylmethyl Radicals at X- and Q-bands
Andrey A. Kuzhelev, Dmitry V. Trukhin, Olesya A. Krumkacheva, Rodion K. Strizhakov, Olga Yu.
Rogozhnikova, Tatiana I. Troitskaya, Matvey V. Fedin, Victor M. Tormyshev, and Elena G. Bagryanskaya
J. Phys. Chem. B, Just Accepted Manuscript • DOI: 10.1021/acs.jpcb.5b03027 • Publication Date (Web): 22 May 2015
Downloaded from http://pubs.acs.org on May 26, 2015
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
The Journal of Physical Chemistry B is published by the American Chemical Society.
1155 Sixteenth Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 34
The Journal of Physical Chemistry
1
2
3
4
5
Room-Temperature Electron Spin Relaxation of Triarylmethyl Radicals at
X- and Q-bands
6
7
8
9
Andrey A. Kuzhelev, †,‡,# Dmitry V. Trukhin,†,‡,# Olesya A. Krumkacheva, §, ‡ Rodion K.
Strizhakov,† Olga Yu. Rogozhnikova,†, ‡ Tatiana I. Troitskaya,†, ‡ Matvey V. Fedin, §,‡,* Victor M.
Tormyshev, †,‡,* and Elena G. Bagryanskaya†,‡,*
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
§ International Tomography Center SB RAS, Novosibirsk 630090, Russia
† N.N. Vorozhtsov Novosibirsk Institute of Organic Chemistry SB RAS, Novosibirsk 630090,
Russia
‡ Novosibirsk State University, Novosibirsk 630090, Russia
# these authors contributed equally
ABSTRACT
Triarylmethyl radicals (trityls, TAMs) represent a relatively new class of spin labels. The long
relaxation of trityls at room temperature in liquid solutions makes them a promising alternative for
traditional nitroxides. In this work we have synthesized a series of TAMs including perdeuterated
Finland trityl (D36 form) , monoꢀ, diꢀ, and triꢀester derivatives of FinlandꢀD36 trityl, deuterated form
of OX63, dodecaꢀnꢀbutyl homologue of Finland trityl, and triamide derivatives of Finland trityl
with primary and secondary amines attached. We have studied roomꢀtemperature relaxation
properties of these TAMs in liquids using pulsed Electron Paramagnetic Resonance (EPR) at two
microwave frequency bands. We have found the clear dependence of phase memory time (Tm~T2)
on magnetic field: roomꢀtemperature Tm values are ~1.5ꢀ2.5 times smaller at Qꢀband (34 GHz, 1.2
T) compared to Xꢀband (9 GHz, 0.3 T). This trend is ascribed to the contribution from gꢀanisotropy
that is negligible at lower magnetic fields but comes into play at Qꢀband. In agreement with this, the
difference between T1 and Tm becomes more pronounced at Qꢀband compared to Xꢀband due to
increased contributions from incomplete motional averaging of gꢀanisotropy. Linear dependence of
(1/Tmꢀ1/T1) on viscosity implies that gꢀanisotropy is modulated by rotational motion of trityl
radical. Based on the analysis of previous data and results of the present work, we conclude that in
general situation where spin label is at least partly mobile, Xꢀband is most suitable for application
of trityls for roomꢀtemperature pulsed EPR distance measurements.
1
ACS Paragon Plus Environment

The Journal of Physical Chemistry
Page 2 of 34
1
I. INTRODUCTION
2
3
Triarylmethyl (TAM, trityl) radicals appeared recently as a new class of spin labels having long
electron spin relaxation on the order of microseconds in liquids at room temperature.1 This
profoundly long relaxation makes trityls a promising alternative for nitroxide spin labels in
biological applications of EPR. The development of this research field requires further
improvement of tritylꢀbased labels in order to achieve even longer T1 and T2 (Tm), as well as
identification of the factors affecting relaxation in liquids at 300 K. Apart from lowꢀtemperature
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
o
EPR and general studies of TAMs,2ꢀ6 the nanoscale distances were recently measured at 4 C in
o
protein7 and at 37 C in DNA duplex8 using TAM labels and pulsed dipolar EPR spectroscopy.
Physiological temperatures were reached owing to the long phase memory time (Tm) of TAMs.1
Phase memory time is one of the most critical parameters for dipolar EPR spectroscopy determining
the range of available distances.9ꢀ10 For example, Tm=0.7 ꢁs at 4 oC in TAMꢀlabeled protein allowed
measuring distances up to 2.5 nm.7 An improvement of Tm to 1.4 ꢁs at 37 oC in TAMꢀlabeled DNA
duplex allowed extending the available distances up to 4.6 nm.8
Thus, development of strategies for further increase of Tm is a highly topical task. Note,
however, that covalent attachment of spin label to biopolymer usually leads to a decrease of roomꢀ
temperature Tm value compared to that for free radical in solution. This stems from larger rotational
correlation times and the presence of anisotropic interactions. Anisotropies of hyperfine interaction
(HFI) and gꢀfactor are the two primary sources causing the decrease of Tm when the mobility of
spin label becomes restricted. With respect to free TAM radicals, the multifrequency studies of
Finland trityl at Xꢀ, Sꢀ, Lꢀbands and 250 MHz revealed that HFI anisotropy plays dominant role and
the contribution of gꢀanisotropy is negligible.1
The available methods for measurements of long interspin distances at physiological
temperatures were recently complemented by saturationꢀrecovery EPR technique.11ꢀ12 In this
approach the spin lattice Relaxation Enhancement (1/T1) of spin labels caused by the paramagnetic
metal (Cu2+) is studied, and the range of available distances crucially depends on the values of T1.11ꢀ
16
Therefore, due to the extraordinary long T1 values (~10ꢀ20 ꢁs) trityl radicals seem to be
promising probes for this approach. The frequency dependence of T1 for Finland trityl and its
derivative OX63 was studied in the range of 250 MHz to 9.2 GHz at room temperature.1, 17 In
particular, it was found that at 250 MHz T1 is determined by interꢀ and intraꢀmolecular electron–
proton dipolar couplings modulated by molecular tumbling with the rates comparable to the
reciprocal of the resonance frequency. The contribution of this mechanism is reduced at higher
2
ACS Paragon Plus Environment

Page 3 of 34
The Journal of Physical Chemistry
1
2
3
4
frequencies and becomes negligible at Xꢀband, where spinꢀlattice relaxation of TAMs is contributed
mainly by frequencyꢀindependent local mode process with energy of 590 cmꢀ1.17
5
6
7
8
As a rule, performing pulsed dipolar EPR experiments at frequencies higher than Xꢀband
allows one to increase the S/N ratio, as was demonstrated using nitroxideꢀbased spin labels at Qꢀ
band.18ꢀ19 Therefore, one might naturally assume that the same tendency should hold for roomꢀ
temperature distance measurements on trityl radicals. Up to date, however, there was no clear
understanding of the factors affecting electron relaxation times of trityl labels attached to
biopolymers. Therefore, the data on electron relaxation times of trityls at Xꢀband and higher
frequencies should have a high practical importance for applications of pulsed dipolar EPR.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Scheme 1. Structures of studied trityl radicals.
In this work, we synthesized the broad range of triarylmethyl radicals (TAMs) with varying
number and structure of substituents (Scheme 1) and measured their relaxation times at room
temperature in liquids. In order to find optimal structures of trityl spin labels and experimental
conditions for roomꢀtemperature distance measurements in biomolecules we focused on comparison
of Tm for different trityls, as well as on field dependence of Tm in the range 9ꢀ34 GHz. In addition,
the T1 values for newly synthesized trityls also represent very useful data for applications other than
pulsed dipolar EPR, e.g. for distance measurements at physiological temperature using Relaxation
Enhancement. Note that we studied different derivatives of Finland trityl including those that were
3
ACS Paragon Plus Environment

The Journal of Physical Chemistry
Page 4 of 34
1
2
already used successfully for distance measurements at physiological conditions, namely TAM with
3
4
piperazine substituent (FP3)8 and TAM with amide substituent (FBA3).7
5
6
II. EXPERIMENTAL
7
8
9
II.1. Synthesis of Triarylmethyls
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The perdeuterated form of Finland trityl (FD, see Scheme 2 and Supporting Information) was
prepared by the twoꢀstep oneꢀpot reaction, i.e. by conversion of the corresponding diamagnenic
triarylmethanol to trityl cation with the subsequent reduction of this cation with tin(II) chloride.20
High reactivity of bromomethyl acetate toward anionic nucleophiles was used in synthesis of
esters FDAM3, FDAM2 and FDAM1. The reaction of triethylammonium salt of FD with 1.8
molar equivalents of bromomethyl acetate smoothly afforded a mixture of monoꢀ, diꢀ and triesters.
The title products were isolated in acceptable yields (27, 32 and 19 % respectively) using column
chromatography technique.
Activation of the triacid FD (Scheme 2) with carbonyldiimidazole readily afforded the triester
FDME3 in high yield. It was isolated and characterized in the form of trihydrochloride derivative.
Application of alternative methods typical of the acid activation, e.g. with
dicyclohexylcarbodiimide or conversion of initial acid to OSU ester or acyl chloride derivatives,
were found to be less useful: these reactions resulted in badly contaminated mixtures of products.
Scheme 2. Synthesis of the Finland trityl derivatives FDME3, FDAM3, FDAM2 and FDAM1.
4
ACS Paragon Plus Environment

Page 5 of 34
The Journal of Physical Chemistry
1
2
3
4
5
6
Contrary to ester FDME3, the amides FP3 and FBA3 were readily obtained by reaction of
piperidine with trisꢀchloroanhydride form of triacid IH, and by coupling of benzyl amine with trisꢀ
OSU ester FOSU3, respectively (Scheme 3).
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Scheme 3. Synthesis of trityls FP3 and FBA3.
The synthesis of deuterated analogue of OX63 is shown in Scheme 4. The initial
triarylmethanol VID21 was treated with formic acid to remove all the protecting tertꢀbutyl groups by
substituting them with formate groups. The solid material obtained after removal of volatile
components was converted to trityl cation in reaction with trifluoromethanesulfonic acid, the cation
was then reduced to intermediate TAM.
Scheme 4. Synthesis of OX63D.
To remove formate groups, the latter was hydrolyzed with excess of aqueous NaOH to give a
solution of sodium salt of trityl OX63D. The resulting solution was acidified by its slow addition to
diluted aqueous hydrochloric acid to afford a fine crystalline precipitate of the title TAM.
5
ACS Paragon Plus Environment

The Journal of Physical Chemistry
Page 6 of 34
1
2
3
4
5
6
7
8
The trityl DBT (Scheme 5) was obtained by analogy with the protocol focused on Finland
trityl.20 A reaction of nonanoneꢀ5 with 1,2,4,5ꢀtetraꢀtertꢀbutylthiobenzene in the presence of boron
trifluoride readily afforded 2,2,6,6ꢀtetraꢀnꢀbutylꢀbenzo[1,2ꢀd;4,5ꢀd’]bis[1,3]dithiole VIII.
Triarylmethyl alcohol IX was then prepared by treatment of arene VII with nꢀBuLi and the
subsequent addition of 0.32 equiv of diethylcarbonate. Purification of the crude product did not
require a lengthy and tedious column chromatography, instead it was completed by refluxing the
crude product with 1:1 mixture of hexane and carbon tetrachloride.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The arene IX was converted to a slurry of trisꢀlithium derivative by treatment with nꢀBuLi in
TMEDA/hexane solution. This organolithium intermediate readily reacted with solid carbon
dioxide to afford the triacid X in good isolated yield (57 %). The title TAM DBT was prepared in
quantitative yield by the known literature method.20 The identity of TAMs obtained in this study
was confirmed by element analyses and combination of instrumental methods: highꢀresolution
ESI/MS, EPR and FTIR spectroscopies.
Scheme 5. Synthesis of TAM substituted with nꢀbutyl groups at side positions (DBT).
II.2. EPR procedures
All used solvents were preliminary distilled. Concentrations of dissolved trityls were ~100ꢀ200 ꢀM
in water, ~50ꢀ100 ꢀM in methanol and chloroform to avoid any effects of the radicalꢀradical
collisions on electron spin relaxation. The concentration for each sample / each solvent was
adjusted experimentally, starting at 0.2 mM solution and diluting it until the spectrum shape and
relaxation times (in evacuated samples) became completely concentrationꢀindependent. The
6
ACS Paragon Plus Environment

Page 7 of 34
The Journal of Physical Chemistry
1
2
3
4
5
6
7
8
samples were sealed off in the glass capillary tubes (OD 1.5 mm, ID 0.9 mm, max. volume 50 µl).
In order to prevent the line broadening and relaxation enhancement caused by oxygen,
deoxygenation was performed by repeated “freezeꢀpumpꢀthaw” cycles for at least ten times. For
each sample we ensured that the further evacuation does not increase the observed relaxation times
T1 and T2. The obtained T1 and T2 values for FD and FH in H2O perfectly agree with those reported
previously,1, 17 thus confirming the efficiency of deoxygenation.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Pulsed and low temperature CW EPR experiments were carried out at Xꢀband (9 GHz) and Qꢀ
band (34 GHz) using commercial Bruker ELEXSYS E580 spectrometer equipped with an Oxford
flow helium cryostat and temperature control system.
Measurements of electron spin relaxation (Tm~T2, T1) were performed at Xꢀ and Qꢀbands at
T=300 K. Tm was measured using twoꢀpulse Electron Spin Echo (ESE) sequence, T1 was measured
using inversionꢀrecovery technique with inversion πꢀpulse and detecting twoꢀpulse ESE sequence.
These experiments were carried out using 2ꢀstep phase cycling to eliminate a signal of FID. The
pulse lengths were 100/200 ns at Xꢀband and 120/240 ns at Qꢀband for π/2 and π pulses,
respectively. ESE was measured at field position corresponding to the maximum of the EPR
spectrum. Obtained relaxation curves were fitted well using monoexponential function. All
measured relaxation times were reproduced for at least 3 times for independently prepared new
sample each time. All pulse EPR experiments were performed with repetition time of 300 ꢁs and
3000 ꢁs at 300 K and 80 K, respectively. FIDꢀdetected EPR spectra were recorded at Qꢀband at 80
K; the pulse length was 1200 ns.
Experimental CW EPR settings at 140 K were as follows: sweep width 2.5 mT; microwave
power 6.142·10ꢀ7 W; modulation frequency 100 kHz; modulation amplitude 0.1 mT; time constant
81.92 ms; sweep time 83.89 s; number of points 1024; number of scans 4.
CW EPR experiments at room temperature were performed at Xꢀband using commercial
spectrometer Bruker EMX. Experimental CW EPR settings at room temperature were as follows:
sweep width 0.1 – 0.5 mT; microwave power 1ꢀ6 ꢀW; modulation frequency 100 kHz; modulation
amplitude 2 – 5 ꢀT; time constant 327.68 – 1310.72 ms; sweep time 167.77ꢀ671.09 s; number of
points 512; number of scans 1.
In order to determine isotropic gꢀfactor of TAM (giso) at 300 K, we simultaneously recorded
Xꢀband CW EPR spectra of two degassed samples placed in separate sample tubes (FD as a
reference and the investigated TAM), and thus obtained target giso value relative to the known giso of
FD (2.0026 22).
7
ACS Paragon Plus Environment

The Journal of Physical Chemistry
Page 8 of 34
1
III. RESULTS
2
3
4
III.1. CW EPR study
5
6
7
8
Before studying electron spin relaxation of series of TAMs, we investigated all these radicals using
continuous wave (CW) EPR in order to identify their purity and HFI couplings. Figure 1 shows Xꢀ
band CW EPR spectra of studied trityl radicals obtained at 300 K in different solvents. The
experimental spectra were simulated using Easyspin software (see SI).23 The obtained spectroscopic
parameters are summarized in Table 1.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Fig. 1. Xꢀband CW EPR spectra of studied trityl radicals at 300 K in different solvents.
CW EPR spectrum of FD radical consists of single narrow EPR line of Gaussian shape due to
the unresolved hyperfine constants on methyl deuterons of the aryl rings. Similarly, CW EPR
spectra of OX63D and DBT are also simulated by single line, but taking into account both Gaussian
and Lorentzian broadenings. The spectra of other radicals FDME3 and FDAM1-3 demonstrate
additional splitting due to the hyperfine interaction with methylene protons –OꢀCH2ꢀ of side ester
8
ACS Paragon Plus Environment

Page 9 of 34
The Journal of Physical Chemistry
1
2
3
4
5
6
7
moieties. EPR spectra of FDME3 and FDAM3 show 7 well or partially resolved lines that were
simulated using two groups of three equivalent protons with slightly different hyperfine constants.
The observed difference between hyperfine constants should be assigned to the effect of chirality.21,
24ꢀ25
Due to steric hindrance caused by bulky tetrathioaryl groups attached to central carbon atom,
8
9
trityl radicals adopt a propellerꢀlike conformation in which all rings have the same direction of
twist. This geometrical arrangement allows two conformations to occur, i.e. the rightꢀhanded
propeller (Pꢀconformation) and leftꢀhanded propeller (Mꢀconformation). The propellerꢀshape
conformation of TAM associated with a slow interconversion of conformers is responsible for the
diastereotopic character of geminal hydrogen atoms of methylene groups bonded to carboxyl
moieties. In addition, the steric hindrance caused by adjacent sulfur atoms creates severe obstacles
for free rotation of carboxylꢀderived groups.6 The combination of hindered rotation of paraꢀ
substituents and general chirality of TAM core make the two protons of each methylene group
inequivalent. Thus, the total of 6 protons is subdivided into two diastereotopic groups with three
protons in each. The Aiso(1H) values of methylene proton of FDAM3 are two times smaller
compared to FDME3. EPR spectra of FDAM2 and FDAM1 consist of 5 and 3 partially resolved
lines indicating hyperfine interaction with 4 and 2 methylene protons, respectively, with hyperfine
constants close to those in FDAM3. In addition to the main EPR signal of FDAM1, a weak singlet
line was observed at gꢀvalue corresponding to the signal of FD radical. It was assigned to the
product of ester solvolysis, whose amount does not exceed 13% in freshly prepared samples in
water (see SI).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The EPR spectrum of triamide derivative of Finland trityl FBA3 in chloroform consists of 11
partially resolved lines. This spectrum was simulated using 3 equivalent 14N HFI constants and two
groups of three equivalent protons with slightly different hyperfine constants related to protons of
methylene groups bonded to amine. The hyperfine splitting on proton of amine group was not
observed; presumably, the hyperfine constant for this group is smaller than the linewidth. In
methanol only poorly resolved lines of three 14N HFI splittings were observed, with all other
hyperfine constants being hidden within a linewidth.
The EPR spectra of triamide FP3, which represents the derivative of secondary amines, do
not show any resolved splitting in both methanol and chloroform indicating that nitrogen constant is
smaller compared to FBA3; according to performed estimation, it does not exceed 0.01 mT.
The radicals FDME3, FDAM2ꢀ3, FBA3 (that can be dissolved in MeOH and CHCl3) show
the dependence of hyperfine constant Aiso(1H) (arising from methylene protons –OꢀCH2ꢀ for
FDME3, FDAM2ꢀ3 and –NHꢀCH2ꢀ for FBA3) on solvent polarity: the Aiso(1H) value decreases as
solvent polarity grows, indicating a decrease of spin density on methylene protons.
9
ACS Paragon Plus Environment

The Journal of Physical Chemistry
Page 10 of 34
1
2
3
4
5
Table 1. Spectroscopic properties of deoxygenated trityl radicals at 300 K. The values of isotropic hyperfine
constants (Aiso) and peakꢀtoꢀpeak line widths (LW) are given in µT.
a
a
LWa
giso
6
7
8
9
Trityl
Solvent
Aiso
1H (nb)
14N (nb)
(G; L)c
MeOH
CHCl3
MeOH
CHCl3
MeOH
H2O
5.0 (3), 4.5 (3)
6.1 (3), 4.9 (3)
5.1 (2), 4.5 (2)
6.1 (2), 4.8 (2)
4.7 (1), 4.3 (1)
4.6 (1), 5.5 (1)
9.4 (3), 9.1 (3)
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
5.1; 0
5.1; 0
5.2; 0
5.0; 0
2.00277
2.00282
2.00271
2.00281
FDAM3
FDAM2
FDAM1
FDME3
FD
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
4.4; 1.0 2.00266
5.1; 0
5.2; 0
5.6; 0
4.5; 0d
4.6; 0d
2.00266
2.00274
2.00278
2.00260
2.00260
MeOH
CHCl3 10.6 (3), 10.6 (3)
MeOH
H2O
MeOH
H2O
MeOH
MeOH
ꢀ
ꢀ
ꢀ
ꢀ
6.0; 2.3 2.00267
7.5; 0 2.00268
10.0; 3.9 2.00266
OX63D
ꢀ
ꢀ
DBT
FBA3
<6.0(6)
18.2 (3)
20.0; 0
13.5; 0
<10.7(3) 39.5; 0
<10.7(3) 38.3; 0
2.00255
2.00261
2.00252
2.00253
CHCl3 15.0 (3), 13.3 (3) 15.3 (3)
MeOH
CHCl3
ꢀ
ꢀ
FP3
a
The simulated spectra were calculated in MATLAB using the EasySpin software package.
Isotropic gꢀfactors are close to the literature values of FD (giso=2.0026)
b n ꢀ number of equivalent nuclei
c G and L correspond to the Gaussian and Lorentzian contributions into the linewidth
d See Supporting Information for comparison with literature data.
III.2. Measurements of electron spin relaxation times T1 and Tm
III.2.1. Dependence on magnetic field (X/Q-bands)
We have studied relaxation properties of several perspective trityls including FD and its
derivatives. In all cases we have observed clear dependence of Tm on magnetic field by passing
from Xꢀ to Qꢀband (9 to 34 GHz). Figure 2 exemplifies the observed trend using deuterated Finland
trityl. The phase memory time Tm remains closely constant below 9 GHz, but it decreases by a
factor of ~2.5 at 34 GHz; at the same time, T1 remains roughly fieldꢀindependent between 3 and 34
GHz.
In order to understand the underlying mechanisms of Tm and its magnetic field dependence,
we have performed a lot of additional measurements of T1 and Tm relaxations for the series of trityls
depending on solvent viscosity, depending on deuteration of radical and/or solvent, depending on
polarity of the solvent and, finally, on radical structure (see summary in Table 2). Despite variation
10
ACS Paragon Plus Environment

Page 11 of 34
The Journal of Physical Chemistry
1
2
3
4
of Tm values on radical structure, the decrease of Tm by a factor of 1.5ꢀ2.5 at Qꢀband compared to
Xꢀband was observed for all studied radicals.
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Fig. 2. The dependence of relaxation times Tm and T1 on mw frequency for deuterated Finland trityl (FD) in
H2O at 300 K. The data at 0.25ꢀ3 GHz were taken from Ref.1.
Table 2. Electron spin dephasing time, Tm, and electron spinꢀlattice relaxation time, T1 for
investigated radicals at 300 K at Xꢀ and Qꢀbands. The values of Tm and T1 are given in µs. Accuracy
of measurements is about 10%.
Trityl
Solvent
X-band
Q-band
Tm
T1
Tm
T1
MeOH
CHCl3
MeOH
CH2Cl2
CHCl3
8.3
8.4
8.1
9.6
8.6
6.0
6.0
4.0
7.6
11.1
9.9
5.6
6.0
4.9
6.7
5.6
4.3
3.1
1.4
2.8
4.7
5.4
5.4
3.8
4.5
4.5
2.0
4.6
4.0
9.7
9.6
FDAM3
FDAM2
11.6
12.3
11.0
10.5
10.1
11.7
14.3
14.5
12.2
11.4
15.6
17.1
17.7
16.0
15.0
18.0
11.6
11.3
10.0
10.3
10.0
12.2
12.6
CH2ClCH2Cl
CH3CH2OH
Tertꢀbutanol
MeOH
H2O
FDAM1
FDME3
FD
10.1
8.5
9.1
MeOH
CHCl3
MeOH
H2O
D2O
MeOH
H2O
11.2
11.2
15.6
16.2
17.0
15.3
14.4
16.8
9.2
10.8
13.6
6.3
9.2
10.4
FH
D2O
11
ACS Paragon Plus Environment

The Journal of Physical Chemistry
Page 12 of 34
1
2
3
4
5
6
7
8
MeOH
H2O
D2O
MeOH
MeOH
CHCl3
MeOH
CHCl3
5.8
7.3
7.6
5.0
6.5
7.2
5.0
4.5
16.5
16.0
16.1
14.9
19.5
18.0
23.0
26.4
1.8
2.2
2.0
2.1
3.6
4.5
2.8
2.3
15.6
15.3
16.1
14.3
19.2
17.6
23.0
25.7
OX63D
DBT
FBA3
FP3
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
III.2.2. Dependence on solvent viscosity
One of the first factors to be established is what motion does induce the observed relaxational (Tm)
behavior. Tm values for studied samples are smaller than T1 at 9 and especially 34 GHz indicating
the presence of additional contributions to spinꢀspin relaxation apart from spinꢀlattice relaxation. In
order to extract these contributions, the values (1/Tmꢀ1/T1)≈(1/T2ꢀ1/T1) were obtained that depend
both on magnetic field and on solvent viscosity (Fig.3).
Fig. 3. The dependence of (1/Tmꢀ1/T1) values on solvent viscosity for FDAM2 at 300 K at Xꢀ (black squares)
and Qꢀbands (blue circles). The solid red lines represent the fits obtained using function (1/Tmꢀ1/T1)=a⋅η+b.
Viscosity values26 for used solvents are given in SI.
FDAM2 is soluble in broad range of solvents, therefore we obtained the (1/Tmꢀ1/T1) values as
a function of solvent viscosity and found linear dependence (Fig. 3). Since rotational correlation
time of radical c is proportional to , the data of Fig.3 also implies linear dependence of
η
τ
η
(1/Tmꢀ1/T1) on
to anisotropic interaction (V) modulated by rotational diffusion: (1/Tmꢀ1/T1)∝ V 2
proportionality factor depends on magnetic field and increases ~4 fold at Qꢀband compared to Xꢀ
band, i.e. <
2> turns out to be approximately proportional to magnetic field. Previous studies on FD
τc . This observation agrees well with the Redfield theory’s result for relaxation due
τ
c . The
V
at Xꢀ, Sꢀ, Lꢀbands and 250 MHz demonstrated that dominant contribution stems from HFI
12
ACS Paragon Plus Environment

Page 13 of 34
The Journal of Physical Chemistry
1
2
3
4
5
6
7
8
anisotropy rather than from gꢀanisotropy.1 The results obtained in our work, however, clearly show
that in the range 9ꢀ34 GHz the fieldꢀdependent contributions play a significant role. As will be
discussed below, the fieldꢀdependent contributions stem from pure gꢀanisotropy ([g:g] contribution)
as well as from inner product of gꢀ and HFI anisotropies [g:A], and these anisotropies are
modulated by radical tumbling, as follows from viscosity dependence in Fig. 3.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
III.2.3. Role of protons
Protons (of both radical and the solvent) are known to largely contribute to the phase memory time
of nitroxides as well as trityls.1, 27 We have studied T1 and Tm at X/Qꢀbands for Finland trityl
depending on (i) deuteration of the radical (compare deuterated FD and its protonated analogue FH)
and (ii) deuteration of the solvent (H2O vs. D2O); in addition, we have studied T1 and Tm of OX63D
depending on deuteration of the solvent (Table 2 and Fig.4). In all cases T1 remains nearly the same
at Xꢀ and Qꢀbands (variations <10%). Contrary to that, Tm noticeably changes depending on
deuteration. For Finland trityl deuteration of radical and solvent increases Tm by 10ꢀ25% at Xꢀband,
but at the same time the effect at Qꢀband is negligible.
III.2.4. Influence of solvent polarity and hydrogen bonding
Methanol and chloroform differ in polarity, but they have almost the same viscosity and thus the
same rotational correlation time for the same radical. Therefore, comparison of the relaxation times
in methanol and chloroform allows one to study the influence of solvent polarity on electron
relaxation. For all trityls that are soluble in both methanol and chloroform (FDME3, FDAM3,
FDAM2, FBA3, FP3), the obtained (1/Tmꢀ1/T1) values were found nearly the same within the
accuracy (Table 2 and Fig.5), thus excluding the significant dependence of relaxation time on
solvent polarity.
The hydrogen bonding is known to be more effective for atoms with large electronegativity.
Therefore, one anticipates two trends to be found: (i) hydrogen bonding between methanol and ester
derivatives of TAM (FDAM1-3, FDME3) should be stronger in comparison with triamide
derivatives (FBA3, FP3); (ii) hydrogen bonding between TAM and methanol should be stronger
than that between TAM and chloroform. Comparison of the (1/T2ꢀ1/T1) values in methanol and
chloroform obtained for all these radicals (mentioned in (i)) did not reveal any noticeable influence
of hydrogen bonding within experimental accuracy of 20%, both at Xꢀ and Qꢀbands (Fig.5). Further
detailed studies in series of suitable solvents are required to find out the effects of hydrogen
bonding on relaxation rates of TAMs (if any).
13
ACS Paragon Plus Environment

The Journal of Physical Chemistry
Page 14 of 34
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Fig. 4. The dependence of relaxation times Tm and T1 for FD, FH and OX63D at Xꢀ and Qꢀbands in different
solvents (indicated on the plot) at 300 K. Accuracy of measurements is about 10%.
III.2.5. Influence of radical structure
All studied trityls are soluble in methanol that allows obtaining the field dependence of (1/Tmꢀ1/T1)
as a function of trityl structure. OX63D and DBT (the derivatives of Finland trityl with methyl
groups substituted by heavier groups) demonstrate the larger (1/Tmꢀ1/T1) values compared to FD.
Fig. 5. (1/Tmꢀ1/T1) values (contributions to 1/T2 from incomplete motional averaging of anisotropic interactions)
for studied trityls in different solvents at 300 K at Xꢀ and Qꢀbands. Accuracy of (1/Tmꢀ1/T1) values is about 20%.
14
ACS Paragon Plus Environment

Page 15 of 34
The Journal of Physical Chemistry
1
2
3
4
5
6
7
8
FDAM1-3 radicals include ester substituents attached to carboxyl moieties of FD; therefore
they have additional hydrogen hyperfine constants ~5 µT resolved in CW EPR (Fig.1). FDAM1-3
show Xꢀband (1/Tmꢀ1/T1) values close to those of FD in methanol. At Qꢀband the (1/Tmꢀ1/T1) value
for FDAM3 (that has 3 ester substituents) is about 2 times lower than that for FD. At the same time,
trityls FDAM3 and FDME3 (differing by HFI constants of CH2 groups of the ester moiety) have
nearly the same (1/Tmꢀ1/T1) values at Xꢀ and Qꢀbands, respectively.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The amide derivatives of Finland trityl FBA3 and FP3 have the additional nitrogen hyperfine
constants ~0.01ꢀ0.02 mT. They demonstrate 2 and 3 times higher (1/Tmꢀ1/T1) values compared to
FD, respectively. At the same time, the difference between (1/Tmꢀ1/T1) values of amide derivatives
(FBA3, FP3) and FD at Qꢀband is reduced, resulting in practically the same (1/Tmꢀ1/T1) values for
FBA3 and FD.
We have found that T1 is sensitive to the substituent introduced into carboxyl moiety of trityl
FD. There are two trends which were found: the decrease of T1 upon introduction of the “ester
groups”, and increase of T1 upon introduction of “amide groups” (resulting ratio of T1 for FDAM3
and FBA3 is about 2). Replacement of the methyl groups of FD by more bulky groups (OX63D
and DBT) does not affect the observed T1 values.
III.3. EPR spectra in frozen solution
The lineshapes of EPR spectra of radicals in frozen solution yield information on anisotropic
interactions that are partly (or fully) averaged by motion in liquids. In order to separate fieldꢀ
dependent and fieldꢀindependent anisotropies (gꢀ and HFI, respectively), we performed lowꢀ
temperature EPR measurements at Xꢀ and Qꢀbands in glassꢀforming methanol. We measured CW
EPR spectra at Xꢀband at 140 K and FIDꢀdetected EPR spectra at Qꢀband at 80 K. It was
technically difficult to employ the same approach at both mw bands due to insufficient sensitivity of
FID detection at 80 K at Xꢀband and, at the same time, easy saturation of CW EPR spectra at Qꢀ
band for all temperatures below melting point of methanol. For one of the samples (FD) we verified
that FIDꢀdetected and integrated CW EPR spectra coincide at Xꢀband (see SI).
15
ACS Paragon Plus Environment

The Journal of Physical Chemistry
Page 16 of 34
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Fig. 6. EPR spectra of studied trityls in frozen methanol: (top) First integral of CW EPR spectra at Xꢀband at 140
K and (bottom) FIDꢀdetected EPR spectra at Qꢀband at 80 K.
The obtained EPR spectra at Xꢀband are similar for all trityls (Fig. 6). The Qꢀband spectra of
trityls exhibit partially resolved gꢀanisotropy that slightly depends on radical structure. For semiꢀ
quantitative comparison of spectra at Xꢀ and Qꢀbands we introduced the value of γ1/3, defined as the
width of the spectrum at the 1/3 height of the maximum. The more common value, FWHM, does
not account for the shoulders in Qꢀband spectra. We obtained nearly the same values γ1/3≈0.25 mT
at Xꢀband and ~2.5 times higher values at Qꢀband, indicating fieldꢀdependent contribution of gꢀ
anisotropy to the lineshape of EPR spectra in frozen solution. Note that the obtained increase of γ1/3
by a factor of 2.5 at Qꢀband is smaller than the corresponding increase of the microwave frequency
(~3.5). This indicates that anisotropy of hyperfine interaction might also contribute to the line
width, especially at Xꢀband.
The obtained spectra were simulated using rhombic gꢀtensor (see SI). Similar values of gꢀ
tensor were found for all trityls (except for OX63D) that agree well with the previously published
data on FD.22 The rhombic gꢀtensor of FD was previously explained by distortion from 3ꢀfold
symmetry occurring in frozen solution.22
16
ACS Paragon Plus Environment

Page 17 of 34
The Journal of Physical Chemistry
1
2
3
4
IV. DISCUSSION
The rotational correlation time of FD in water at 300 K is about 0.8 ns,1 thus the motion regime
5
6
7
8
corresponds to ω0τc ≥1 at both Xꢀ and Qꢀbands, where T1>Tm was experimentally found.
9
Therefore, the values of (1/Tmꢀ1/T1) describe all contributions to spinꢀspin relaxation apart from
spinꢀlattice relaxation. The obtained linear dependence of (1/Tmꢀ1/T1) on solvent viscosity at Xꢀ and
Qꢀbands indicates incomplete motional averaging of anisotropic interactions, and different slopes in
these two mw bands imply significant contribution of fieldꢀdependent anisotropies. In agreement
with this, for frozen solutions we have found that the EPR linewidth at Qꢀband is 2ꢀ3 times larger
compared to that at Xꢀband.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The only fieldꢀdependent anisotropic interaction present in studied systems is Zeeman
interaction with anisotropic gꢀtensor. At the same time, fieldꢀindependent anisotropic HFI
interactions are also unavoidable. Not aiming at comprehensiveness, below we would like to
demonstrate that our experimentally obtained trends can be rationalized using these two plausible
relaxation mechanisms and reasonable sets of parameters. One should consider the following
contributions to the phase memory time, taken in a model of radical with one magnetic nucleus
I
=1/2 of effective anisotropy and neglecting all terms proportional to 1 (1+ω02τc2
mentioned above, ω0τc ≥1):28
)
(because, as was
1
1
2
β
2H2
4
πβ
H
7
π 2
30
−
≈
[
g
:
g
]τc
±
[
g
:
A]τc
+
[
A
:
A]τc
(1)
Tm
T
15h2
15h
1
Here [g:g], [g:A] and [A:A] are the inner products of corresponding gꢀ and Aꢀtensors, generally
defined as [T1 : T 2 ] = T1Tk2i . The first term of eq.(1) is proportional to the square of magnetic
∑
ik
ik
field (H), the second one to the first power of magnetic field, and the third one is fieldꢀindependent.
We do not observe the quadratic dependence of (1/Tmꢀ1/T1) on magnetic field (Fig.5), thus it is
most reasonable to assume that the observed field dependence stems from an interplay of [g:g] term
and crossꢀanisotropy term [g:A]. Given that previous observations indicate negligible contribution
of gꢀanisotropy and dominant role of HFI anisotropy at νmw<10 GHz,1 one anticipates that going to
higher and higher magnetic field will first reveal the contribution of [g:A] term, and only then the
term [g:g] might become dominant. Thus, the observed tendency is qualitatively coherent with
previous observations.
Based on the available data, we can only estimate the contribution of the first [g:g] term in
eq.(1) and its dependence on magnetic field. The values for FD g=[2.0028 2.0020 2.0030] obtained
17
ACS Paragon Plus Environment

The Journal of Physical Chemistry
Page 18 of 34
1
2
from simulation agree well with literature data22 and give [g:g]≈6⋅10ꢀ7 (details in SI). Introducing
3
4
5
6
7
8
this [g:g] value into the first term of eq.(1),
(1/Tmꢀ1/T1)~0.45⋅105 sꢀ1 in water, i.e. very close to the experimental value measured for FD (Fig.5).
For
=1.2 T (Qꢀband) the same estimation yields (1/Tmꢀ1/T1) ~7⋅105 sꢀ1, that is even higher than the
τ
c =0.8 ns, and H=0.3 T (Xꢀband) we obtain
H
9
experimental value (see Fig.5), despite the second and the third terms of eq.(1) are not taken into
account. Since [g:g] term is known to play a minor role at Xꢀband,1 the obtained value
(1/Tmꢀ1/T1)~0.45⋅105 sꢀ1 is certainly overestimated. This might be ascribed to the overestimation of
[g:g] parameter obtained on the basis of frozen solution data, since structural distortions (and thus
gꢀanisotropy) of trityls in frozen solution were assumed to be larger compared to fluid solutions.22
Summing up, the account of [g:g] term using reasonable parameters is already sufficient (and even
excessive) to explain the experimentally observed field dependence of Tm and (1/Tmꢀ1/T1).
Estimation of [g:A] and [A:A] terms is much more difficult, since there are many small HFI
constants whose anisotropies are not known, as well as relative directions of the corresponding
tensors. In addition, our experiments show that interactions of electron spin with nuclei of the
solvent are not completely negligible. Therefore, we cannot provide meaningful estimation for these
two terms, but, as was reasoned above, contribution of [g:A] term is expected to be higher than that
of [g:g] term. Overall, the combination of the first and the second terms of eq.(1) can readily
provide necessary mechanism to drive noticeable shortening of Tm when passing from Xꢀ to Qꢀ
band.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The following observations summarize the “Pros” for the proposed above mechanism of Tm in
studied trityls:
(i) Clear field dependence of Tm at 0.3ꢀ1.2 T (X/Qꢀband) is found for all studied radicals;
(ii) Viscosity dependence of (1/Tmꢀ1/T1) shows that anisotropy is modulated by radical
tumbling (not by any kind of local modes or internal motions of radical fragments);
(iii) The 1/Tm and (1/Tmꢀ1/T1) values at 9 and 34 GHz are greater for OX63D and DBT than
for FD, as one would expect because of the slower tumbling of OX63D and DBT that have
more bulky substituents than the methyl groups;
(iv) Effect of deuteration of radical and/or solvent on Tm is noticeable at Xꢀband, but an
increase of the relaxation time Tm is very moderate (<50%), even when all protons are
substituted by deuterons. This clearly means that other interactions add to HFI already at Xꢀ
band. At Qꢀband the effect of deuteration becomes nearly indistinguishable, meaning that at
this frequency contributions from HFIs become negligible, and thus gꢀanisotropy dominates
the relaxation process.
18
ACS Paragon Plus Environment

Page 19 of 34
The Journal of Physical Chemistry
1
2
3
4
5
6
7
8
(v) Introduction of amide substituents (FBA3 and FP3) leads to increase of (1/Tmꢀ1/T1)
values compared to FD by a factor of 2ꢀ3, most likely due to the two effects: (i) the slower
tumbling rate and (ii) additional hyperfine constants, which may well be anisotropic. At Qꢀ
band the (1/Tmꢀ1/T1) for FBA3 and FD become nearly the same, meaning again that at this
frequency contributions from HFIs become negligible, and thus gꢀanisotropy dominates the
relaxation process.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Although the central observation of this work, namely the pronounced decrease of Tm at Qꢀband
compared to Xꢀband, is valid for all studied trityls and therefore is general, several smaller details
remain unanswered and need to be addressed in the future. The open questions are:
ꢀ
Introduction of ester into carboxyl moiety of FD increased τc, yet the decrease of Tm for
FDAM3-1 compared to FD was not observed. Furthermore, the (1/Tmꢀ1/T1) value for
FDAM3 at Qꢀband is 2 times smaller than that for FD. Why does the increase of the number
of ester substituents decrease (1/Tmꢀ1/T1) value and the magnitude of its change with
magnetic field is currently unclear.
ꢀ
ꢀ
The viscosity of water (~1 cP) is two times higher than the viscosity of methanol (~0.5 cP).26
However, the 1/Tm and (1/Tmꢀ1/T1) values for FD, OX63D, FDAM1 are greater in methanol
than in water at both Xꢀ and Qꢀbands.
OX63D has lower gꢀanisotropy compared to FD (as follows from EPR spectra at Qꢀband in
frozen methanol). However, the increase of Tm at Qꢀband is nearly the same for both radicals.
Note, that the above peculiarities arise, first of all, from the very long relaxation times of trityls, that
makes them sensitive to the subtle influences of structure and environment. These specifics of
trityls, which at the same time are their most important advantage for EPR, certainly need to be
studied in the future; at the same time, the conclusions concerning field dependence of relaxation
and optimum conditions for EPR measurements, e.g. distance measurements, are already in hand.
We stress that one of the radicals (monoamide analogue of FP3) has already been used by us as spin
label for roomꢀtemperature distance measurements at Xꢀband.8 An attractive progress would seem
to be using higher magnetic fields (e.g. mw frequencies 17ꢀ94 GHz), smaller sample volumes,
better sensitivity. However, the present work clearly shows that this will be done at the expense of
7
shortened phase memory time. Possibly, shorter Tm values at 17 GHz compared to those at 9
GHz 8 found previously have the same origin as the observations of the present work.
19
ACS Paragon Plus Environment

The Journal of Physical Chemistry
Page 20 of 34
1
V. CONCLUSIONS
2
3
In this work, we describe convenient methods for synthesis of broad variety of triarylmethyl
radicals (TAMs) with varying polarity, with varying number and nature of substituents attached to
carboxyl moieties located in paraꢀpositions of TAM aryl rings, and finally with varying structure of
the TAM core. The primary aim of these developments is the future application of trityls with
optimum relaxation properties as spin labels for roomꢀtemperature distances measurements in
biomolecules. As the first step, in this work we investigated relaxation of this new series of
prospective TAMs in solutions at room temperature. Certainly, the next step is the attachment of
most promising TAMs to biomolecules, that might shorten their relaxation times. Because of this,
an understanding of mechanisms governing Tm is vital. This work shows that Tm shortens at Qꢀband
compared to Xꢀband, due to the contribution of gꢀanisotropy terms. Although different SDSL
strategies, different linkers and immobilization procedures might affect ultimate Tm times of spinꢀ
labeled biomolecules, we anticipate that the dependence on magnetic field similar to the present
work will be manifested for partly mobile TAM labels. At the current state of research, we believe
that, unless spin label is strongly immobilized, Xꢀband is most suitable for roomꢀtemperature
distance measurements using TAMs.
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ASSOCIATED CONTENT
Supporting Information available: Synthesis of TAMs; Simulation of Xꢀband CW EPR spectra at
room temperature; Solvolysis of FDAM1; Comparison of linewidth of FD with literature data;
Properties of solvents; Comparison of FIDꢀdetected and CW EPR spectra; Simulation of Qꢀband
FIDꢀdetected EPR spectra at 80 K; Estimation of contribution from gꢀanisotropy; Primary
relaxation data. This material is available free of charge via the Internet at http://pubs.acs.org.
AUTHOR INFORMATION
Corresponding Authors:
mfedin@tomo.nsc.ru (M. V. F.), torm@nioch.nsc.ru (V.M.T), egbagryanskaya@nioch.nsc.ru
(E.G.B)
Notes
The authors declare no competing financial interests.
20
ACS Paragon Plus Environment

Page 21 of 34
The Journal of Physical Chemistry
1
2
3
4
ACKNOWLEDGMENT
5
6
7
8
All pulse EPR studies performed in this work were supported by Russian Science Foundation (No.
14ꢀ14ꢀ00922). Synthesis and preliminary characterization of studied TAMs were supported by
RFBR (Nos. 13ꢀ04ꢀ00680, 14ꢀ03ꢀ93180, 14ꢀ03ꢀ31839) and The National Institute of Biomedical
Imaging and Bioengineering (grant No. 5P41EB002034). We thank the personnel of the Collective
Service Center of SB RAS for recording of IR and NMR spectra and performing element and HPLC
analyses.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
REFERENCES
1.
Owenius, R.; Eaton, G. R.; Eaton, S. S. Frequency (250 MHz to 9.2 GHz) and Viscosity Dependence of
Electron Spin Relaxation of Triarylmethyl Radicals at Room Temperature. J. Magn. Reson. 2005, 172 (1), 168-
175.
2.
Bobko, A. A.; Dhimitruka, I.; Zweier, J. L.; Khramtsov, V. V. Trityl Radicals as Persistent Dual Function
pH and Oxygen Probes for in Vivo Electron Paramagnetic Resonance Spectroscopy and Imaging:ꢀ Concept
and Experiment. J. Am. Chem. Soc. 2007, 129 (23), 7240-7241.
3.
Kunjir, N. C.; Reginsson, G. W.; Schiemann, O.; Sigurdsson, S. T. Measurements of Short Distances
between Trityl Spin Labels with CW EPR, DQC and PELDOR. Phys. Chem. Chem. Phys. 2013, 15 (45), 19673-
19685.
4.
Nanometer-Distance Measurements. Chem-Eur. J. 2012, 18 (43), 13580-13584.
5. Trukhan, S. N.; Yudanov, V. F.; Tormyshev, V. M.; Rogozhnikova, O. Y.; Trukhin, D. V.; Bowman, M. K.;
Reginsson, G. W.; Kunjir, N. C.; Sigurdsson, S. T.; Schiemann, O. Trityl Radicals: Spin Labels for
Krzyaniak, M. D.; Chen, H.; Martyanov, O. N. Hyperfine Interactions of Narrow-Line Trityl Radical with
Solvent Molecules. J. Magn. Reson. 2013, 233 (0), 29-36.
6.
Magn. Reson. 2005, 172 (2), 254-267.
7. Yang, Z. Y.; Liu, Y. P.; Borbat, P.; Zweier, J. L.; Freed, J. H.; Hubbell, W. L. Pulsed ESR Dipolar
Bowman, M. K.; Mailer, C.; Halpern, H. J. The Solution Conformation of Triarylmethyl Radicals. J.
Spectroscopy for Distance Measurements in Immobilized Spin Labeled Proteins in Liquid Solution. J. Am.
Chem. Soc. 2012, 134 (24), 9950-9952.
8.
Shevelev, G. Y.; Krumkacheva, O. A.; Lomzov, A. A.; Kuzhelev, A. A.; Rogozhnikova, O. Y.; Trukhin, D.
V.; Troitskaya, T. I.; Tormyshev, V. M.; Fedin, M. V.; Pyshnyi, D. V., et al. Physiological-Temperature Distance
Measurement in Nucleic Acid Using Triarylmethyl-Based Spin Labels and Pulsed Dipolar EPR Spectroscopy. J.
Am. Chem. Soc. 2014, 136 (28), 9874-9877.
9.
Electron Paramagnetic Resonance. Phys. Chem. Chem. Phys. 2007, 9 (16), 1895-1910.
10. Saxena, S.; Freed, J. H. Theory of Double Quantum Two-Dimensional Electron Spin Resonance with
Application to Distance Measurements. J. Chem. Phys. 1997, 107 (5), 1317-1340.
11. Yang, Z. Y.; Jimenez-Oses, G.; Lopez, C. J.; Bridges, M. D.; Houk, K. N.; Hubbell, W. L. Long-Range
Jeschke, G.; Polyhach, Y. Distance Measurements on Spin-Labelled Biomacromolecules by Pulsed
Distance Measurements in Proteins at Physiological Temperatures Using Saturation Recovery EPR
Spectroscopy. J. Am. Chem. Soc. 2014, 136 (43), 15356-15365.
12.
Electron Spin Resonance Distance Measurements. Biochemistry 2006, 45 (38), 11666-11673.
13. Razzaghi, S.; Brooks, E. K.; Bordignon, E.; Hubbell, W. L.; Yulikov, M.; Jeschke, G. EPR Relaxation-
Jun, S.; Becker, J. S.; Yonkunas, M.; Coalson, R.; Saxena, S. Unfolding of Alanine-Based Peptides Using
Enhancement-Based Distance Measurements on Orthogonally Spin-Labeled T4-Lysozyme. ChemBioChem
2013, 14 (14), 1883-1890.
21
ACS Paragon Plus Environment

The Journal of Physical Chemistry
Page 22 of 34
1
2
3
4
14.
Jager, H.; Koch, A.; Maus, V.; Spiess, H. W.; Jeschke, G. Relaxation-Based Distance Measurements
between a Nitroxide and a Lanthanide Spin Label. J. Magn. Reson. 2008, 194 (2), 254-263.
15. Kulikov, A.; Likhtenstein, G. The Use of Spin Relaxation Phenomena in the Investigation of the
5
Structure of Model and Biological Systems by the Method of Spin Labels. Adv. Mol. Relax. Int. Pr. 1977, 10
6
(1), 47-79.
7
8
9
16.
Lueders, P.; Jager, H.; Hemminga, M. A.; Jeschke, G.; Yulikov, M. Distance Measurements on
Orthogonally Spin-Labeled Membrane Spanning WALP23 Polypeptides. J. Phys. Chem. B 2013, 117 (7), 2061-
2068.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
17.
Yong, L.; Harbridge, J.; Quine, R. W.; Rinard, G. A.; Eaton, S. S.; Eaton, G. R.; Mailer, C.; Barth, E.;
Halpern, H. J. Electron Spin Relaxation of Triarylmethyl Radicals in Fluid Solution. J. Magn. Reson. 2001, 152
(1), 156-161.
18.
Radical Biol. Med. 2001, 30 (5), 489-499.
19. Polyhach, Y.; Bordignon, E.; Tschaggelar, R.; Gandra, S.; Godt, A.; Jeschke, G. High Sensitivity and
Versatility of the DEER Experiment on Nitroxide Radical Pairs at Q-Band Frequencies. Phys. Chem. Chem.
Phys. 2012, 14 (30), 10762-10773.
Berliner, L. J.; Khramtsov, V.; Fujii, H.; Clanton, T. L. Unique in Vivo Applications of Spin Traps. Free
20.
Rogozhnikova, O. Y.; Vasiliev, V. G.; Troitskaya, T. I.; Trukhin, D. V.; Mikhalina, T. V.; Halpern, H. J.;
Tormyshev, V. M. Generation of Trityl Radicals by Nucleophilic Quenching of Tris(2,3,5,6-
Tetrathiaaryl)Methyl Cations and Practical and Convenient Large-Scale Synthesis of Persistent Tris(4-
Carboxy-2,3,5,6-Tetrathiaaryl)Methyl Radical. Eur. J. Org. Chem. 2013, 2013 (16), 3347-3355.
21.
Tormyshev, V. M.; Genaev, A. M.; Sal'nikov, G. E.; Rogozhnikova, O. Y.; Troitskaya, T. I.; Trukhin, D. V.;
Mamatyuk, V. I.; Fadeev, D. S.; Halpern, H. J. Triarylmethanols with Bulky Aryl Groups and the Noesy/Exsy
Experimental Observation of a Two-Ring-Flip Mechanism for the Helicity Reversal of Molecular Propellers.
Eur. J. Org. Chem. 2012, (3), 623-629.
22.
Appl. Magn. Reson. 2005, 28 (3-4), 231-238.
23. Stoll, S.; Schweiger, A. Easyspin, a Comprehensive Software Package for Spectral Simulation and
Analysis in EPR. J. Magn. Reson. 2006, 178 (1), 42-55.
24. Driesschaert, B.; Robiette, R.; Le Duff, C. S.; Collard, L.; Robeyns, K.; Gallez, B.; Marchand-Brynaert, J.
Configurationally Stable Tris(Tetrathioaryl)Methyl Molecular Propellers. Eur. J. Org. Chem. 2012, (33), 6517-
6525.
25.
Tetrathiatriarylmethyl Spin Probes. Chem. Commun. (Cambridge, U. K.) 2011, 47 (16), 4793-4795.
26.
27.
Electron Spin Relaxation Times and Rapid Scan EPR Imaging of pH-Sensitive Amino-Substituted Trityl
Radicals. Magn. Reson. Chem. 2015, 53 (4), 280-284.
Fielding, A. J.; Carl, P. J.; Eaton, G. R.; Eaton, S. S. Multifrequency EPR of Four Triarylmethyl Radicals.
Driesschaert, B.; Robiette, R.; Lucaccioni, F.; Gallez, B.; Marchand-Brynaert, J. Chiral Properties of
Lide, D. R. CRC Handbook of Chemistry and Physics. 85th ed.; Boca Raton, Fla, 2004.
Elajaili, H. B.; Biller, J. R.; Tseitlin, M.; Dhimitruka, I.; Khramtsov, V. V.; Eaton, S. S.; Eaton, G. R.
28.
Hawkins, C. L.; Davies, M. J. Detection and Characterisation of Radicals in Biological Materials Using
EPR Methodology. Biochim. Biophys. Acta-Gen. Subj. 2014, 1840 (2), 708-721.
TOC graphic
22
ACS Paragon Plus Environment

Page 23 of 34
The Journal of Physical Chemistry
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
85x132mm (300 x 300 DPI)
ACS Paragon Plus Environment

The Journal of Physical Chemistry
Page 24 of 34
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
81x83mm (300 x 300 DPI)
ACS Paragon Plus Environment

Page 25 of 34
The Journal of Physical Chemistry
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
81x60mm (300 x 300 DPI)
ACS Paragon Plus Environment

The Journal of Physical Chemistry
Page 26 of 34
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
73x109mm (300 x 300 DPI)
ACS Paragon Plus Environment

Page 27 of 34
The Journal of Physical Chemistry
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
83x56mm (300 x 300 DPI)
ACS Paragon Plus Environment

The Journal of Physical Chemistry
Page 28 of 34
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
82x135mm (300 x 300 DPI)
ACS Paragon Plus Environment

Page 29 of 34
The Journal of Physical Chemistry
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
46x106mm (300 x 300 DPI)
ACS Paragon Plus Environment