Programmable Nuclear-Spin Dynamics in Ti(IV) Coordination Complexes

Article abstract of DOI:10.1021/acs.inorgchem.0c00244

Interstitial patterning of nuclear spins is a nascent design principle for controlling electron spin superposition lifetimes in open-shell complexes and solid-state defects. Herein we report the first test of the impact of the patterning principle on liga

Full text of DOI:10.1021/acs.inorgchem.0c00244

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
pubs.acs.org/IC
Article
Programmable Nuclear-Spin Dynamics in Ti(IV) Coordination
Complexes
Spencer H. Johnson, Cassidy E. Jackson, and Joseph M. Zadrozny*
Cite This: https://dx.doi.org/10.1021/acs.inorgchem.0c00244
Read Online
sı
*
Metrics & More
Article Recommendations
Supporting Information
ACCESS
ABSTRACT: Interstitial patterning of nuclear spins is a nascent
design principle for controlling electron spin superposition lifetimes
in open-shell complexes and solid-state defects. Herein we report the
first test of the impact of the patterning principle on ligand-based
1
nuclear spin dynamics. We test how substitutional patterning of H
and 79/81Br nuclear spins on ligands modulates proton nuclear spin
dynamics in the ligand shell of metal complexes. To do so, we
studied the 1H nuclear magnetic resonance relaxation times (T1 and
T2) of a series of eight polybrominated catechol ligands and six
complexes formed by coordination of the ligands to a Ti(IV) ion.
1
These studies reveal that H T1 values can be enhanced in the
individual ligands by a factor of 4 (from 10.8(3) to 43(5) s) as a
function of substitution pattern, reaching the maximum value for 3,4,6-tribromocatechol. The T2 for 1H is also enhanced by a factor
of 4, varying by ∼14 s across the series. When complexed, the impact of the patterning design strategy on nuclear spin dynamics is
amplified and 1H T1 and T2 values vary by over an order of magnitude. Importantly, the general trends observed in the ligands also
1
match those when complexed. Hence, these results demonstrate a new design principle to control H spin dynamics in metal
complexes through pattern-based design strategies in the ligand shell.
a powerful method for manipulating the dynamic field and
producing long-lived molecular electronic spin superpositions.
In this context, a set of chemical design principles to tune
the magnetic environment imposed by the nuclei in a molecule
would be extremely useful. Here, manipulation of the impact of
this proximate “edge” of the larger nuclear spin bath could be
achieved via synthetic modification of interspin distances,
relative arrangements, and identity of nuclear spins on the
molecule itself. Indeed, recent breakthroughs from our group
on V(IV) complexes19 and others on the solid-state defects in
SiC20−22 provide preliminary demonstrations of manipulating
nuclear spin position/arrangements to control nearby elec-
tronic spin superpositions. Extensive literature exists on the
fundamental mechanisms governing nuclear spin relaxa-
tion,15−18 providing a foundational backbone for translating
the aforementioned breakthroughs into broadly applicable
design principles. Yet, generalizable synthetic strategies to
tailor nuclear spin relaxation in the ligand shell of metal
complexes are still absent.
INTRODUCTION
■
Next-generation applications of open shell magnetic complexes
rely on long-lived electronic spin superposition lifetimes.
Applications for sustained molecule-based spin superpositions
span quantum sensing,1,2 quantum computation,3,4 and
dynamic nuclear polarization.5 Toward utility in these
applications, chemical design principles for lengthening
superposition lifetime are critically important.6−11 However,
the realization of useful lifetimes (>100 μs) in molecules are
still only achieved in environments engineered to be free of a
spin bath, in terms of both nuclear and electronic spins.12,13
Special environments like these are disjoint from applications,
wherein highly active local spin environments are likely to be
present. Hence, testing new design strategies to prolong spin
superposition lifetime in dynamic magnetic surroundings is a
pressing concern.
One prominent origin of short-lived spin superpositions in
open-shell molecules is the dynamic nuclear spin bath.14 Rapid
and numerous fluctuations in the orientations of these nuclear
spins, when close to an electronic spin, impose a fluctuating
local magnetic field on the electron, which accelerates the
collapse of electronic spin superpositions. The extent to which
environmental nuclear spins impart this fluctuating local field
depends on their respective nuclear relaxation dynamics. In
turn, these dynamics stem (in part) from the respective
interactions between the nuclear spins.15−18 Hence, controlling
the spin−spin interactions between bath nuclear spins may be
Received: January 23, 2020
https://dx.doi.org/10.1021/acs.inorgchem.0c00244
© XXXX American Chemical Society
Inorg. Chem. XXXX, XXX, XXX−XXX
A

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
Inorganic Chemistry
pubs.acs.org/IC
Article
Toward that knowledge, we herein test the dependence of
1H nuclear-spin dynamics on substitutional patterning in a
series of brominated catechol derivatives and upon coordina-
tion to Ti(IV) (Figure 1). We hypothesized that different
Figure 1. In this manuscript, we perform the first tests of how
1
different substitutional patterns of H and 79/81Br nuclear spins on
1
ligands impact the ligand H T1 and T2 and how the effect changes
upon coordination.
patterns of Br functional groups on the ligand would enable a
significant synthetic variation in 1H spin dynamics in the
isolated catechols, as bromine and hydrogen have significantly
different nuclear magnetic moments (1H, μ = 2.79 μN, 79Br, μ
= 2.11 μN, 81Br, μ = 2.27 μN) and Larmor frequencies (at 9.4
Figure 2. Synthetic scheme and complexes. (a) The general synthetic
scheme to produce the titanium complexes is depicted. (b) The series
of ten ligands used in this study. (c) Ligands 1, 3−6, and 8 were
reacted to form six Ti(IV) coordination complexes following the
scheme shown in (a).
1
T, H = 400 MHz, 79Br = 101 MHz, 81Br = 108 MHz). We
thus reasoned that different substitutional patterns on the
ligand shell would modulate the interspin dipole−dipole
interactions, thereby impacting T1 and T2.23 The difference
in Larmor frequencies was also viewed as particularly
advantageous for suppressing the impact of nuclear Overhauser
effects on relaxation and simplifying the analysis.24 Finally, we
predicted that these trends would be reproduced when the
ligands were part of the coordination shell. This last prediction
is on the basis of the distance dependence of the dipolar
interactions between nuclei, which weaken with r−6 (where r is
the distance separating nuclei).37−39 Hence, we envisioned that
interligand interactions would be minimal relative to intra-
ligand interactions.
and that coordination enhances the relative impact of the
pattern design principle.
RESULTS
■
Isolation of the studied molecules proceeded with little
difficulty. Indeed, many of the brominated catechols are
available commercially (1, 2, 3, 8) and all others were prepared
via slight modifications of reported procedures (4−7).25−27
Overnight reactions of 3.3 equiv of the catechols 1, 3, 4, 5, 6,
and 8 with 1 equiv of Ti(NMe2)4 in THF yielded dark red
solutions containing 1a, 3a, 4a, 5a, 6a, or 8a (Figure 2), which
can be isolated as orange powders. These reactions produce
1
To test the foregoing hypotheses, we investigated the H
spin−lattice relaxation (T1) and spin−spin relaxation (T2)
times for eight catechol molecules: pyrocatechol (1), 3-
bromocatechol (2), 4-bromocatechol (3), 3,5-dibromocatechol
(4), 4,5-dibromocatechol (5), 3,4,5-tribromocatechol (6),
3,4,6-tribromocatechol (7), and tetrabromocatechol (8)
(Figure 2). We also selected six specific titanium complexes
from this set of ligands for investigation: [Ti(C6H4O2)3]2−
(1a), [Ti(C6H3-4-BrO2)3]2− (3a), [Ti(C6H2-3,5-Br2O2)3]2−
(4a), [Ti(C6H2-4,5-Br2O2)3]2− (5a), [Ti(C6H-3,4,5-
Br3O2)3]2− (6a), and [Ti(C6Br4O2)3]2− (8a) (Figure 2), all
+
Me2NH2 counterions, suggesting that the Me2N− ligands of
the Ti(IV) starting material Ti(NMe2)4 are deprotonating the
catechol ligands during reaction progression. This same
counterion formation by ligand deprotonation characterizes
the synthesis of the analogous V(IV) complexes.28,29
Single-crystal X-ray diffraction of crystals resulting from the
reactions of 1 and 5 with Ti(NMe2)4 reveals an octahedral
geometry for the TiO6 coordination environment (Figures 3
and S1, Tables S1−S3). These experiments reveal a general 3-
to-1 ligand-to-metal stoichiometry and two Me2NH2+ counter-
ions. All other characterization data for the metal complexes
match this general stoichiometry, though isolating high-quality
single crystals of 3a, 4a, 6a, and 8a was significantly more
challenging (see SI). Bond metrics taken from the obtained
structures match literature expectations. For example, the
average Ti−O distances in 1a and 5a are 1.9640(3) and
1.993(5) Å, respectively, close to those of known Ti-
catecholate analogues.30−32 Furthermore, C−O bond distances
in the ligands are 1.345(8) Å which match those for the fully
+
isolated as the Me2NH2 salts. We found that substituting
79/81Br in a 1H spin system extends 1H T1 and T2 by a factor of
4, depending on the pattern of the substitution. When
complexed with Ti(IV), the ligand protons mimic the
relaxation trends of the uncomplexed ligands. Yet, the relative
variation is considerably more dramatic in the studied
1
complexes, as H T1 and T2 are enhanced by an order of
magnitude depending on substitutional pattern. These data
provide the first evidence of a synthetic design principle for
modifying nuclear spin dynamics via substitutional patterning,
B
https://dx.doi.org/10.1021/acs.inorgchem.0c00244
Inorg. Chem. XXXX, XXX, XXX−XXX

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
Inorganic Chemistry
pubs.acs.org/IC
Article
Figure 3. Molecular structures of [Ti(C6H4O2)3]2− (left) and
[Ti(C6H2Br2O2)3]2− (right), as determined in the crystal structures
+
of 1a and 5a. Hydrogen bonding contacts with the Me2NH2
counterions are omitted for clarity. Orange, maroon, red, gray, and
white spheres represent titanium, bromine, oxygen, carbon, and
hydrogen atoms, respectively.
reduced catecholate, suggesting closed-shell ligands (as
opposed to possible semiquinones, where C−O is 1.302(11)
Å).33 These data establish these molecules as closed-shell
Ti(IV) complexes.
Proton NMR spectroscopy (400 MHz, 9.4 T) was applied to
the ligands and complexes to locate and identify the chemical
shifts of the nuclei of interest. All ligands and complexes depict
1H peaks in the aromatic region (save for 8 and 8a, which have
only aromatic 79/81Br nuclear spins). Here, the peaks appear to
be shifted downfield by proximity to either hydroxy or bromine
groups. Furthermore, these peaks are split by J-coupling, where
J is largest for 1,2-coupling between ortho protons (5.86−8.43
Hz) and decreasing as the protons are spaced apart on the
catechol ring (2.05−3.57 Hz for 1,3-coupling between meta
34
1
1Hs). The 1,4-coupling strength between para Hs was too
small to observe for 1, 3, and 5 in our spectrometer. Peaks
corresponding to OH groups for the free ligand show up from
1
+
Figure 4. Example H NMR analyses of a catechol molecule, here 1.
(a) Inversion−recovery spectrathe offset spectra represent different
delay times between inversion and measurement, with the shortest
7.76 to 9.22 ppm. The Me2NH2 counterions show peaks at
2.60 and 8.09−8.60 ppm for the −CH3 and NH2 protons,
respectively, in the complexes. Upon complexation to Ti(IV),
1
1
delays at the bottom in green. (b) Depiction of H peak assignment.
the catecholate H aromatic peaks move slightly upfield but
(c) Peak intensity as a function of delay following inversion. The
black line is a fit to an exponential recovery to yield T1 for a given
peak. (d) CPMG experimental resultsthe offset spectra represent
different numbers of applied π pulses: the green spectrum at the
bottom was collected with one π pulse and the red spectrum at the
otherwise retain the same relative δ values and J couplings. The
exception to the spectral similarity is complex 3a, which
exhibits rearranged peaks due to the removal of the shifting
effect from the hydroxyl protons (Figures S3 and S9 in the SI).
Spin−lattice relaxation times were determined by standard
1H NMR inversion−recovery experiments at room temper-
ature (see Figures 4 and S2−S7 and the SI) for all ligands and
complexes. These experiments proceeded with π pulses of 22
μs, and all catechol solutions were 200 mM in acetone-d6. All
Ti(IV) complexes were measured at 100 mM in DMSO-d6
(solubility challenges precluded the use of acetonesee SI).
NMR spectra were then collected following the initial π pulse,
a variable-delay period, T, and a final π/2 pulse. For all ligands
and complexes, the initial spectra following the π pulses are
inverted, eventually recovering to the expected positive signal
intensity. All recoveries were successfully fit with mono-
exponential functions, suggesting that only a single relaxation
process dominates for each peak. Importantly, the time scale of
this recovery is peak-dependent (Figures 4 and S2−S8).
Our results show that, with minor exceptions, the
1
top was collected with 12 π pulses. (e) H peak intensity versus total
time after the π/2 pulse. The black line is a fit to an exponential decay
to yield T2 for the indicated peak.
times of 10.8(4) and 11.9(2) s, respectively, the shortest of 1−
8. This general trend holds true for the adjacent protons in 1
and 3, which exhibit aromatic 1H T1 values only slightly longer
compared to 2, ca. 13−14 s. Second, the presence of a Br atom
1
adjacent to a H (versus another proton) appears to have a
slight lengthening effect on T1. Indeed, in 2, the proton in the
4-position (which has one adjacent proton and one adjacent
bromine) exhibits a slightly longer T1 (17.9(4) s) than the
other two protons of this ligand, which lack a neighboring Br
atom (10−12 s). Third, T1 is further lengthened when the Br
atom isolates a given proton from other protons. For example,
the 1H in the 3-position of 3, which is separated from the other
protons, exhibits a relaxation time of 27(3) s, nearly twice as
long as the proton in the corresponding position on 1.
Similarly, for 4 and 5, which have no adjacent protons, the 1H
T1 values are extended to 28−31 s. Finally, fourth, when only
1
substitution of Br atoms for H atoms engenders longer H
T1 times (Table 1), based on four key observations. First, the
shortest T1 values are observed for protons that are adjacent to
other protons. For example, the adjacent protons of 2 have T1
C
https://dx.doi.org/10.1021/acs.inorgchem.0c00244
Inorg. Chem. XXXX, XXX, XXX−XXX

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
Inorganic Chemistry
pubs.acs.org/IC
Article
1
Table 1. Tabulated Aromatic H Spin−Lattice Relaxation
Times (T1) Determined via Inversion Recovery
a b c
, ,
Experiments
Ring Position
Ligands
3
4
5
6
1
2
3
4
5
6
7
14.8(7)
13.3(6)
17.9(4)
13.3(6)
10.8(4)
13.7(6)
14.8(7)
11.9(2)
13.4(7)
31(5)
28(4)
32(4)
27(3)
30(5)
28(4)
43(5)
Ring Position
Complexes
3
4
5
6
1a
3a
4a
5a
1.32(2)
0.48(2)
2.11(6)
2.11(6)
0.74(8)
1.32(2)
0.38(3)
6.8(3)
6.0(5)
7.1(2)
6.3(5)
6.0(5)
6a
a
b
All T1 values given in units of s. There are no aromatic protons for
c
8 and 8a, hence the omission. T1 values for OH groups and
Me2NH2 counterions can be found in the SI.
+
1
one aromatic H is present (as in the triply brominated 6 and
7), T1 can become even longer (32(4) and 43(5) s). In
summary, these data demonstrate that T1 can be varied by a
factor of approximately four depending on the substitutional
pattern.
1
The aromatic H signals for the ligands display shorter
overall T1 values when coordinated to Ti(IV) (Figures 5, S9−
S13, and Table 1). However, (and importantly), the same
trends in 1H T1 values as a function of Br substitution patterns
are seen for both the complexes and the lone ligands. For
example, the adjacent protons in 3a display the shortest T1
values among the studied complexes, 0.48(2) and 0.38(3) s. In
contrast, 4a−6a, which have H nuclear spins isolated from
one another by Br atoms, exhibit T1 values that are an order of
magnitude greater (6.0(5) to 7.1(2) s).
Figure 5. Example 1H NMR analyses of a Ti(IV) catecholate complex
molecule, here 1a. (a) Inversion−recovery spectrathe offset spectra
represent different delay times between inversion and measurement,
1
1
with the shortest delays at the bottom in green. (b) Depiction of H
peak assignment. (c) Peak intensity as a function of delay following
inversion. The black line is a fit to an exponential recovery to yield T1
for a given peak. (d) CPMG experimental resultsthe offset spectra
represent different numbers of applied π pulses: the red spectrum at
the bottom was collected with one π pulse, and the fluorescent green
Tests of the effects of substitutional pattering on 1H T2 were
performed through application of Carr−Purcell−Meiboom−
Gill (CPMG) experiments (Figures 4, 5, S2−S7, and S9−
S12).35,36 This experiment is similar to a Hahn-echo
experiment; however, the time separating sequential π pulses
is fixed, and the number of π pulses is varied, such that the
total delay time between the initial pulse and the observed
echo is the sum of repeated pulses and fixed delays. Application
of the CPMG sequence to obtain T2 removes field
inhomogeneities as a potential impact on spin−spin
relaxationensuring any determined trends are attributable
to the studied complexes and not extrinsic effects. As with T1,
the time-dependent data set could be effectively modeled with
a single exponential function, implying a single relaxation
process is dominant. Critically, as with T1, T2 is peak-
dependent.
1
spectrum at the top was collected with 10 π pulses. (e) H peak
intensity versus total time after the π/2 pulse. The black line is a fit to
an exponential decay to yield T2 for the indicated peak.
trends. Furthermore, protons in 3 and 5 that are isolated from
the others by bromination show similar T2 values. The doubly
brominated 5’s T2 values fall within the range of 3’s relaxation
times. In contrast, the compositional isomer to 5, with
1
relatively closer H spins, 4, shows >60% relative lengthening
of T2. The two triply brominated ligands also display radically
different T2 values. 6 was found to have a T2 of 4.30(4) s,
shorter even than 1. In contrast, 7 exhibits the longest T2 of all
of the ligands, 18(1) s, exceeding the T1 of pyrocatechol. In
summary, these data show substantial variation in T2 with
substitutional patterning that mimics the trends in T1.
The T2 times of the aromatic protons in 1−8 broadly follow
the trends seen in the T1 datagreater degrees of bromination
1
As with T1, the spin−spin relaxation times of the aromatic
produce longer H T2 values (see Table 2). For example, the
1H ligand peaks all decrease upon complexation and generally
adjacent ligand 1Hs on 1 were found to have T2 parameters of
approximately 5 s, which are the shortest of the series (second
to 6, see below). The singly brominated ligands 2 and 3 both
have T2 values that are larger than 1, again consistent with T1
1
follow the trends in T1. For example, the ligand Hs in 1a
exhibit the second-shortest T2 values (∼0.9 s), generally
smaller than complexes where 1H spins are separated by
D
https://dx.doi.org/10.1021/acs.inorgchem.0c00244
Inorg. Chem. XXXX, XXX, XXX−XXX