A series of F-Element chelators; diaza crown ethers functionalized with catecholate binding substituents

Article abstract of DOI:10.1016/j.jorganchem.2017.11.026

Reported here is the preparation of azacrown ethers functionalized with catechol groups. The synthetic approach was (1st) novel in that it made use of the Mannich reaction and (2nd) valuable in that it provided an improved synthesis (in terms of practical

Full text of DOI:10.1016/j.jorganchem.2017.11.026

Accepted Manuscript
A series of F-Element chelators; diaza crown ethers functionalized with catecholate
binding substituents
Benjamin W. Stein, Samantha K. Cary, John M. Berg, Eva R. Birnbaum, Stosh A.
Kozimor, Veronika Mocko, Brian L. Scott
PII:
S0022-328X(17)30675-7
10.1016/j.jorganchem.2017.11.026
JOM 20191
DOI:
Reference:
To appear in: Journal of Organometallic Chemistry
Received Date: 13 September 2017
Revised Date: 28 November 2017
Accepted Date: 29 November 2017
Please cite this article as: B.W. Stein, S.K. Cary, J.M. Berg, E.R. Birnbaum, S.A. Kozimor, V. Mocko,
B.L. Scott, A series of F-Element chelators; diaza crown ethers functionalized with catecholate binding
substituents, Journal of Organometallic Chemistry (2018), doi: 10.1016/j.jorganchem.2017.11.026.
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please
note that during the production process errors may be discovered which could affect the content, and all
legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT
A SERIES OF F-ELEMENT CHELATORS; DIAZA CROWN ETHERS
FUNCTIONALIZED WITH CATECHOLATE BINDING SUBSTITUENTS.
Benjamin W. Stein,1 Samantha K. Cary, John M. Berg,1 Eva R. Birnbaum,1 Stosh A.
Kozimor,1* Veronika Mocko,1 Brian L. Scott.1
1 Los Alamos National Laboratory, Los Alamos, New Mexico 87545, United States of America
* To whom correspondence should be addressed, stosh@lanl.gov
LA-UR-17-28162
T
ABLE OF CONTENT GRAPHIC
Dedicated to Prof. William J. Evans for his inspirational and creative
contributions to f-element redox chemistry.
ABSTRACT
Reported here is the preparation of azacrown ethers functionalized with catechol
groups. The synthetic approach was (1st) novel in that it made use of the Mannich
1

ACCEPTED MANUSCRIPT
reaction and (2nd) valuable in that it provided an improved synthesis (in terms of
practical deployment) of the known N,N’-bis(2,3-dihydroxybenzyl)-4,13-diaza-18-crown-
6, H4ChaCha. Moreover, it demonstrated potential application of the synthetic method
for accommodating a wide range of catecholate functionalities by using the synthetic
strategy to prepare N,N’-bis(2,3-dihydroxy-5-tert-butylbenzyl)-4,13-diaza-18-crown-6
t
t
(H4 Bu2ChaCha) for the first time. These H4ChaCha and H4 Bu2ChaCha macrocycles
offer exciting opportunity to expand redox chemistry for the f-elements. As “proof-of-
principle,” we isolated the unusual tetrameric cluster [La2(tBuChaCha)2]2 from reactions
t
between H4 Bu2ChaCha and La[N(SiMe3)2]3. Characterization of [La2(tBuChaCha)2]2 by
elemental analysis, single crystal X-ray diffraction, IR, and UV-vis-NIR spectroscopy
suggested that the complex represented a rare example of an f-element semiquinone. It
t
further demonstrated that the combination of La3+ and H4 Bu2ChaCha provided access
to one-electron oxidation chemistry within redox potential windows that were amenable
to mild reaction conditions.
I
NTRODUCTION
Since the initial discovery by PEDERSEN and colleagues (1967),[1] crown ethers and their
derivatives (e.g. azacrowns[2]) have emerged as one of the most fundamentally important classes
of chelating agents. For instance, crown ethers find broad use[3] in radioisotope chelation,[4]
ion transport,[5] and actinide decorporation.[6] Perhaps, the most attractive component
associated with crown and azacrown ethers is the versatility at which functional groups can be
tethered from the rings to create "lariat" crown ethers. This diversity allows chelating properties
to be tuned for highly specific functionality. For example, equipping acetate side-arms to the
2

ACCEPTED MANUSCRIPT
1,4,7,10-tetraazacyclododecane macrocycle generates one of the best lanthanide(III) chelators
known, namely 1,4,7,10-tetraazacyclododecane-N,N',N'','''-tetraacetic acid (DOTA).[7, 8] In
addition, attaching chromogenic[9, 10] and redox non-innocent pendant arms[11] opens exciting
avenues of research across the periodic table.
Herein, we have contributed to the voluminous collection of lariat azacrown ethers[12, 13]
by reporting the preparation of catechol-functionalized macrocycles via the Mannich reaction.
Previously, HOSSEINI ET. AL. prepared a catechol functionalized diaza-18-crown-6 ether, N,N’-
bis(2,3-dihydroxybenzyl)-4,13-diaza-18-crown-6 (H4ChaCha) (Scheme 1),[14] and explored
H4ChaCha’s reactivity as a ditopic binding moiety for boron and alkali metals.[14, 15]
Contributed here, is a simplification of that H4ChaCha synthesis. Compatibility of the revised
process with tert-butyl (tBu) substituted catecholate functional groups was additionally explored.
The experiments were successful and N,N’-bis(2,3-dihydroxy-5-tert-butylbenzyl)-4,13-diaza-18-
t
crown-6 (H4 Bu2ChaCha) was isolated for the first time. While more experimentation is required
to define generality for this synthetic method, it seems possible that our approach has potential to
accommodate a wide range of catecholate functionalities necessary for fine tuning metal binding
properties and ligand solubilities.
The bis-catechol azacrown ethers reported here offer exciting opportunities for expanding
redox chemistry for the f-elements. Consider the transformative impact that catechol redox
behavior has had on transition metal chemistry.[16-23] The catecholate ligand has been heavily
studied because of its ability to support two reversible oxidation reactions within redox potential
windows that are amenable to mild reaction conditions. The first electron transfer accessible to
catecholate (Cat2-) forms a semiquinone (SQ1-) and the second a quinone (Q; Scheme 2).[23]
With the exception of a few examples (particularly by RAYMOND ET. AL.),[6, 24, 25] f-element
3

ACCEPTED MANUSCRIPT
catecholate chemistry has lagged in comparison to the d-block elements, in part, because f-
element catecholate coordination chemistry can be difficult to control.[26] Lanthanides and
actinides have large coordination numbers, complicated hydrolysis behavior, and propensity to
be bridged by catecholates.[26] Hence, f-element catecholate reaction products tend to be
oligomeric, insoluble, or difficult to characterize. Tethering catecholates to crown – or azacrown
– ethers offers opportunities to chelate the f-element cation and limit uncontrollable
oligomerization reactions. As testament, additionally contributed here is a prototypical reaction
chemistry of the bis-catechol azacrown ethers with La3+. These efforts resulted in reproducible
formation of an intriguing f-element dioxolene cluster, specifically [La2(tBuChaCha)2]2, which is
reported for the first time.
R
ESULTS AND
Synthesis of H4ChaCha-2HCl. Two synthetic routes were developed to the
H4ChaCha 2HCl macrocycle. Both synthetic methods employed relatively benign reagents,
D
ISCUSSION
ñ
required simple chemical-handling manipulations, and were tolerant of exposure to air and
moisture. Our foray into the synthetic process was inspired by the published synthetic
methodology used to synthesize H4ChaCha previously.[14] That original published 5-step
synthetic route started from commercially available 2,3-dihydroxybenzaldehyde to give the
protected catechol-azacrown ether that was subsequently deprotected with H2/Pd.[14] Motivated
by the significant difficulty we had with reproducing several key steps, in particular the
bromination with PBr3, efforts were made to find an alternative synthesis. The 1st revision
(Scheme 3) represented marginal improvement. It was slightly less technically challenging and
provided H4ChaCha
ñ
2HCl in acceptable yields from commercially available reagents; the
4

ACCEPTED MANUSCRIPT
previous synthesis was reported to proceed in 53% overall yield (5 steps) while our route could
be carried out in 26% overall yield (5 steps). Despite the lower yield, our approach avoided the
PBr3 reaction and was initially our preferred method for making H4ChaChañ2HCl.
Our revised procedure, began by protecting both alcohol substituents of
dihydroxybenzaldehyde with the common tert-butyldimethyl silyl group using the well-
established method originally reported by COREY and coworkers (STEP A; Scheme 3).[27]
Subsequently (STEP B), the aldehyde was reduced to the benzyl alcohol with sodium borohydride
(NaBH4) in cold methanol. Then (STEP C) the alcohol was converted to a chloride using
trichlorotriazine (TCT)/DMF.[28] The resulting Int-1c was attached to the diaza-18-crown-6
backbone in STEP D in a SN2-reaction. However, this benzylic substitution represented the lowest
yielding step in the synthetic procedure (40%), likely due to the formation of an unidentified red
impurity that precipitated at low temperature. With Int-1d in hand, acidic conditions
(HCl/MeOH) were used to de-protect the alcohol substituents (STEP E), and H4ChaChañ2HCl
was isolated as a pale beige crystalline powder.
After initially isolating H4ChaChañ2HCl, and with dissatisfaction in the time consuming
multi-step procedure, a more time efficient synthetic route was developed (Scheme 4). This 2nd
procedure was based upon the Mannich condensation, which enabled H4ChaCha
prepared in three – relatively short – synthetic steps and with no tedious workup or
chromatography. Moreover, the Mannich condensation approach enabled H4ChaCha 2HCl to be
ñ2HCl to be
ñ
prepared in a 33% overall yield from commercially available and inexpensive catechol. While
the overall yield was not an improvement in comparison to the original procedure,[14] our new
synthesis was advantageous based on chemical throughput and in terms of application. For
example, our procedure did not require any refined synthetic technique and avoided particularly
5

ACCEPTED MANUSCRIPT
air or moisture sensitive reagents. It was additionally attractive because it could be carried out on
the gram scale with minimal effort in one eight-hour day.
Similar to the synthesis described in Scheme 3, the Mannich-based procedure involved
protecting the alcohol substituents of catechol with the common tert-butyldimethyl silyl group
(STEP A; Scheme 4).[27] However, the Mannich-based method differed in that only one of the
alcohol substituents was protected, not two. Mono-protection was key for STEP B, where the
Mannich reaction was used to affix Int-2b to the diaza-18-crown-6 (Kryptofix 22) reagent. The
success of STEP B seemed to rely on the presence of a single ionizable proton, as all of our
attempts to use unprotected catechol directly resulted in an intractable dark mixture of
byproducts. We speculate that the bulky tert-butyl silyl (TBS) protecting group directed
substitution for the Mannich reaction exclusively to the ortho-position of the catechol ring and
inhibited undesired catechol based oligomerization/polymerization reactions. Int-2b was easily
de-protected using a mixture of HCl and MeOH, and H4ChaCha
ñ
2HCl was isolated in high-
2HCl, we found it
yield and purity. Regardless of the method used to prepare H4ChaCha
ñ
important to isolate the macrocycle as the HCl salt. In this form, the H4ChaCha molecule was
protected against oxidative degradation and could be stored in our laboratory for long periods of
time (>6 months) on the benchtop with no detectable decomposition (1H NMR shows no
impurities) beyond a slight darkening in color.
t
Synthesis of H4 Bu2ChaCha. Motivated by the success in preparing H4ChaCha
ñ
2HCl,
attempts were made to prepare a derivative that had improved solubility in organic solvents. As
t
such, efforts were made to prepare the Bu-substituted analogue. Using similar methodology
t
described in Scheme 5, it was observed that H4 Bu2ChaCha could be prepared in a single step
6

ACCEPTED MANUSCRIPT
from commercially available starting materials: diaza-18-crown-6 (Kryptofix-22), 4-tert-
t
butylcatechol, and paraformaldehyde. This reaction proceeded directly to H4 Bu2ChaCha via the
Mannich condensation reaction,[29] which surprisingly formed only the N,N’-bis(2,3-dihydroxy-
5-tert-butylbenzyl)-4,13-diaza-18-crown-6 isomer. Additionally, unlike the syntheses for
H4ChaChañ2HCl (the one reported previously[14] and the two described herein) protection of
the alcohol substituents on the catecholate functional groups was not required to achieve
synthetic success. Finally, this reaction showed greatly improved yield relative to the synthesis
t
of H4ChaCha
ñ
2HCl, at 84%. The relative ease with which H4ChaCha
ñ
2HCl and H4 BuChaCha
were prepared was consistent with previous reports that used Mannich condensations to prepare
substituted lariat crown ethers.[13, 29]
t
Comparisons of H4ChaCha–2HCl with H4 Bu2ChaCha. The isolation of H4ChaCha–
t
2HCl and H4 Bu2ChaCha represent a new family of readily accessible aza-crown-ethers
functionalized with catechol groups. The subtle differences in catechol substitution imparted
substantial differences in solubility, such that H4ChaCha-2HCl was soluble in polar solvents,
t
like H2O, acetonitrile, methanol, and DMSO. Meanwhile, H4 Bu2ChaCha favored less polar
solvents. It readily dissolved in chloroform and ethyl acetate, had slight solubility in toluene and
tetrahydrofuran, and was insoluble in H2O.
t
The 1H NMR spectra from both H4ChaCha-2HCl and H4 BuChaCha were easily interpreted,
as both compounds were diamagnetic. The spectra were additionally simplified by fast (on the
NMR time scale) fluxional and rotational processes, which rendered equatorial and axial protons
as single peaks. As shown in Figure 1, the spectrum from H4ChaCha-2HCl contained five
resonances. There were a series of poorly resolved peaks associated with the azacrown
7

ACCEPTED MANUSCRIPT
backbone. A singlet from the benzylic protons was present at 4.49 ppm, and two similarly poorly
resolved peaks associated with the aromatic rings were seen at 7.0 and 6.9 ppm. A similar
t
spectrum was obtained for H4 Bu2ChaCha (Figure 1). Subtle differences associated with the
t
benzyl resonance were observed, such that the benzyl peak from H4 Bu2ChaCha was shifted
slightly downfield at 3.80 ppm in comparison to H4ChaCha-2HCl. This was likely associated
with the nitrogen atoms in H4ChaCha-2HCl being protonated while the amines were basic in
t
t
H4 Bu2ChaCha. Additionally, the H4 Bu2ChaCha spectrum was dominated by a large peak
(18H) at 1.25 ppm attributable to the tert-butyl substituents on the catechol groups. Interestingly,
features seen in the spectrum of H4ChaCha-2HCl were substantially broadened relative to
t
H4 Bu2ChaCha. Perhaps the steric crowding from the large tert-butyl substituents or stronger
t
catechol-amine hydrogen bonding limited fluxional processes in H4 Bu2ChaCha and provided a
more rigid solution-phase structure, which (in turn) accounted for the sharper resonances.
t
t
Structure of H4 Bu2ChaCha. Single crystals of H4 Bu2ChaCha readily formed by slow
evaporation of ethanolic solutions (Figure 2). The X-ray diffraction data was best modeled in the
1
P-1 space group and the structure was consistent with the H NMR results, showing methylene
bridged catechols tethered from each amine. The 18-membered ring orientation of the diaza-18-
crown-6 macrocycle was analogous to the “chair” confirmation of cyclohexane. Each amine
group was pyramidal and had catechol groups occupying the equatorial ring position (not axial).
The C–C, C–O, and C–N distances for the diaza-18-crown-6 macrocycle were consistent with
single bonds and averaged 1.502(7), 1.426(4), and 1.467(5), respectively (error reported as the
standard deviations from the mean, 1σ). This orientation placed the catechol groups on opposite
sides of the macrocycle, providing a geometry that minimized destabilizing steric interactions.
8

ACCEPTED MANUSCRIPT
Interestingly, each catechol groups had one -OH substituent oriented toward the diaza-18-crown-
6 macrocyclic ring, likely reflecting intramolecular hydrogen bonding with the amine. The bond
distances were consistent with the formulation of neutral catechols, showing C–C (catechol ring)
double bonds and C–OH (catechol substituent) single bonds, averaging 1.392(9) and 1.37(1),
t
respectively. Hence, the structural results and diamagnetic behavior of the H4 BuChaCha
t
suggested that no unexpected oxidation occurred during the synthesis. Thus, the H4 BuChaCha
compound was best described as containing two catechol (not semiquinone or quinone) groups.
t
Synthesis of [La2(tBu2ChaCha)2]2.
The reaction of H4 Bu2ChaCha with tris(bis
trimethylsilyl)amide lanthanum(III), La[N(SiMe3)2]3 in dry tetrahydrofuran, generated green
solutions from which green solids were isolated (Scheme 6). This observation was intriguing,
given that most complexes containing diamagnetic LaIII are weakly colored (off white or yellow).
Attempts to characterize this product using NMR spectroscopy were unsuccessful because the
spectra were complicated and difficult to interpret. Elemental analysis of the non-crystalline
green powder suggested that the identity of the product was La(tBu2ChaCha). The one-to-one
La3+-to-tBu2ChaCha ratio was consistent with single crystal X-ray diffraction studies carried out
on blue single crystals obtained by vapor diffusion of diethyl ether into toluene solutions. The X-
ray diffraction experiment revealed the crystalline product as tetrameric [La2(tBu2ChaCha)2]2
(Figure 3). Formulation of the [La2(tBu2ChaCha)2]2 reaction product helped explain the
complicated NMR data. For one, if the [La2(tBu2ChaCha)2]2 structure was retained in solution,
protons associated with the macrocycles, the arene rings, and potentially the methylene bridges
on the side-arms would likely be inequivalent. Assigning proton resonances to the >200 protons
associated with this cluster is indeed daunting. Moreover, dissociation reactions to form dimers
9

ACCEPTED MANUSCRIPT
and monomers or association reactions that increase oligomerization would additionally
complicate the NMR spectra.
Although the diffraction from these crystals was weak, it was of sufficient quality to
determine connectivity. The model of the diffraction patters showed that [La2(tBu2ChaCha)2]2
was best described as containing two lanthanum dimers. These dimers were related to each other
by a C2 rotational axis. Hence the coordination environment for the left lanthanum dimer in
t
Figure 4 was equivalent to the right dimer. Each dimer was swaddled by a single H4 Bu2ChaCha
t
ligand that folded around the La2 unit. These dimers were bridged by two H4 Bu2ChaCha
ligands that were essentially unfolded, and contained no metal atom within the diaza crown ether
t
pocket. For each swaddling BuChaCha ligand, there were four oxygen atoms associated with
the dioxolene groups. Two oxygen atoms bridged two La3+ cations in a µ2-fashion, one bound
only one La3+ metal, and one was terminal. For the each of the unfolded tBuChaCha ligands, two
oxygen atoms were bridging and two bound only a single La3+ cation (no terminal oxygens).
While tempting to evaluate bond distances within this structure, we refrain, as the 0.0679 R-
value (wR2 = 0.2108; Table S-1) was quite high. As a result, the structural data could not be
used to reliably distinguish catecholate functional groups from singly protonated catecholates
(HCat1-), semiquinones (SQ1-), or quinones (Q) within the [La2(tBu2ChaCha)2]2 cluster. To
provide a neutral cluster, four reasonable dioxolene scenarios existed (Scheme 7). Because the
t
cluster contained four La3+ cations (which gave a +12 metal charge) and had four BuChaCha
ligands (which contained eight dioxolene groups), it was possible that four of the eight
catecholates were singly protonated giving four HtBuChaCha3- ligands (4 Cat2- and 4 HCat1-).
Equally likely was the case where two catecholates had been singly protonated while two
catecholates underwent a single electron oxidation to form two semiquinones (4 Cat2-, 2 HCat1-,
10

ACCEPTED MANUSCRIPT
and 2 SQ1-). It was also possible that four single electron oxidation reactions occurred. In this
situation, each tBu2ChaCha chelate would contain one catecholate and one semiquinone (4 Cat2-
and 4 SQ1-). Finally, two catecholate groups could have undergone two 2e1- oxidation reactions
to form two quinones. In this last case two tBu2ChaCha ligands would contain only catecholates,
meanwhile two would contain a catecholate and a quinone functional group (6 Cat2- and 2 Q). To
characterize which possible scenarios were most likely, a series of spectroscopic experiments
were conducted, as follows.
It is well established that UV-vis spectroscopy provides a fingerprint for identifying
catecholate oxidation products, specifically semiquinones and quinones.[30] For instance,
catecholates have absorption bands in the UV (>30,000 cm-1; <333 nm), quinones in the blue
(approximately 20,000 to 25,000 cm-1; 500 to 400 nm), and semiquinones in the red to NIR
(~12,000 to 20,000 cm-1; ~833 to 500 nm).[30] Spectra from [La2(tBu2ChaCha)2]2 showed a
strong absorbance at high energy (>20,000 cm-1, 500 nm; Figure 4), which was consistent with
the presence of catecholate functional groups. In addition to this large peak, and at lower energy,
was a less-intense transition that began at 20,000 cm-1, terminated around 12,000 cm-1 (833 to
625 nm), and was centered near 16,000 cm-1 (625 nm). It was this low energy feature that was
responsible for the product’s green color. Moreover, this absorption band provided unique
t
insight into the fate of H4 Bu2ChaCha during the reaction with La[N(SiMe3)2]3. The absorption
near 16,000 cm-1 provided strong evidence that some degree of oxidation of the catechol
moieties has occurred during the reaction. Note, however, that the measured extinction
coefficient was rather low (~20 M-1cm-1), suggesting that complete conversion to the
semiquinone had not occurred, as semiquinones typically contain extinction coefficients in the
range of ~500-1000 M-1cm-1. This supports the notion that only a few of the catecholate groups
11

ACCEPTED MANUSCRIPT
were converted to semiquinones during the formation of [La2(tBu2ChaCha)2]2; however, one
cannot rule out the possibility that the spectral signature has alternative origins nor can the
possibility of dissociation (or further oligomerization) of the tetramer to form a mixture of
complicated species in solution be ignored.
The absence of intense absorbances in the blue spectral region (20,000 to 25,000 cm-1; 500 to
400 nm) suggested that two electron oxidation reactions to form quinones had not occurred.
However, there was a weak peak near 23,000 cm-1. To eliminate the possibility that this subtle
feature was associated with quinone formation, FT-IR spectra were taken (Figure 5). Quinones
exhibit characteristic stretching frequencies in the infrared spectral region at ~1650 cm-1.[31] In
the spectrum from [La2(tBu2ChaCha)2]2 there was clearly no feature that could be correlated with
quinone functionalities. Combined with the UV-vis data, these results provided evidence that
new product, [La2(tBu2ChaCha)2]2, contained a mixture of only catecholates and semiquinones
t
side-arms. Given that the La[N(SiMe3)2]3 reaction with H4 Bu2ChaCha was carried out with
rigorous exclusion of air and moisture, the best description of [La2(tBu2ChaCha)2]2 was having 4
Cat2- and 4 SQ1- ligands; although, at this time the possibility of catecholate protonation cannot
be exclusively ruled out. Hence, the 4 Cat2-, 2 HCat1-, and 2 SQ1- scenario was equally
reasonable.
OUTLOOK
Many recent advances have revealed extraordinary reaction chemistry when strongly Lewis
acidic f-block elements are paired with redox active ligands.[32-41] Among the many
researchers contributing to this area is Prof. WILLIAM J. EVANS, to whom this special addition is
dedicated. EVANS has been a leader in bringing reduction chemistry to the entire f-block.[42-51]
12

ACCEPTED MANUSCRIPT
The isolation of [La2(tBu2ChaCha)2]2 is well within the spirit of EVANS’ innovative achievements.
For example, these results demonstrated that the unique redox properties associated with
t
H4 Bu2ChaCha offers new opportunities in electron transfer for lanthanides and actinides. Our
approach, inspired by EVANS, used catecholate substituents tethered to an azacrown ether, in an
attempt to bring oxidation chemistry (not reduction chemistry) to the “primarily” redox inactive
f-elements metals.
Contributed herein is the first step to achieve this goal. We report an improved synthetic
procedures for H4ChaCha-2HCl, whose solubility properties offer opportunity to access
reactions in polar solvents. Additionally contributed is the preparation and characterization of
t
H4 Bu2ChaCha for the first time, which represented a hydrocarbon soluble analogue of the
H4ChaCha-2HCl macrocycle. Isolation of [La2(tBu2ChaCha)2]2, further demonstrated, as
t
“proof-of-principle,” that H4 Bu2ChaCha was capable of binding metals and forming novel
dioxolene complexes. The isolation of [La2(tBu2ChaCha)2]2 is additionally exciting as it was
formed from an unusual oxidation reaction. Characterization of [La2(tBuChaCha)2]2 by UV-vis
spectroscopy showed that the complex contained both catecolate and semiquinone functional
groups. The UV-vis data further suggested that there were no quinone functional groups present,
a claim substantiated by the IR-spectra. As such [La2(tBuChaCha)2]2 represents a rare example
of an f-element semiquinone complex. The first two examples were prepared by prepared by
GATTESCHI and coworkers: a monometalic gadolinium tris(pyrazolyl)borate-semiquinone
complex[52] and its analogous dimer.[53] The only other reported examples of lanthanide
semiquinone complexes – of that we are aware – involved the reduction of benzoquinone by
either strongly reducing Ln(II) complexes[54, 55] or Ln metals.[56] Overall, our results
demonstrated that oxidation chemistry under mild reaction conditions could be accessed by La3+
13

ACCEPTED MANUSCRIPT
t
and H4 BuChaCha. With the synthetic procedure that provided high-purity [La2(tBu2ChaCha)2]2
in hand, we are currently defining the electronic properties of this unusual lanthanum cluster.
Our goal is to unambiguously characterize the ability of [La2(tBu2ChaCha)2]2 to access
catecholate vs. semiquinone vs. quinone functionality. In addition, we have commenced
t
exploration of H4ChaCha-2HCl and H4 Bu2ChaCha as f-element oxidants, chelators, molecular
magnets, and in terms of chromogenic detection of low concentrations of metal ions.
EXPERIMENTAL
General Considerations. The catechol (Acros), imidazole, 4-tert-butylcatechol (Alfa-Aesar),
4,13-diaza-18-crown-6 (kryptofix 22, EMD Millipore), tert-butyldimethylsilyl chloride (Sigma-
Aldrich), magnesium sulfate (Sigma-Aldrich), tris[N,N-bis( trimethylsilyl)amide] lanthanum(III)
(Alfa Aesar), imidazole (Sigma-Aldrich), and paraformaldehyde (Sigma-Aldrich) reagents were
obtained from commercial sources and used as received. Solvents (DMF, Et2O, MeOH, toluene,
tetrahydrofuran, and acetonitrile) were obtained commercially (HPLC grade) and stored over
molecular sieves (3 Å; 24 h) prior to use. Solvents for use in an argon glovebox (tetrahydrofuran,
toluene, and pentane) were purchased at the anhydrous grade and stored over molecular sieves (3
Å). All H2O used in the following experiments were purified by means of a Millipore Milli-Q
system (18 MΩ). Elemental analyses (C, H, N) were performed by Midwest Microlab Inc.,
Indianapolis, IN via combustion analysis. The NMR experiments were conducted using a Bruker
Avance 400 MHz NMR and processed with MestReNova v10. All NMR spectra were referenced
to either TMS or residual solvent peaks. Samples for UV-Visible-NIR spectroscopy were made
as tetrahydrofuran solutions (8.5 mM) in an inert atmosphere glovebox and loaded into screw top
quartz cuvettes (Starna cells). Spectra were collected on a Cary 6000i with a constant bandwidth
14

ACCEPTED MANUSCRIPT
setting of 1 nm. FT-IR spectra were collected on a Thermo Scientific Nicolet iS5 spectrometer
with an diamond ATR accessory.
Synthetic Route 1 to N,N’-Bis(2,3-dihydroxybenzyl)-4,13-Diaza-18-Crown-6
Dihydrochloride Monohydrate (H4ChaCha-2HCl).
2,3-Bis(tert-Butyldimethylsilyl Ether)Benzaldehyde (Int–1a).
The first intermediate in Scheme 3 was prepared on the benchtop without exclusion of air and
moisture by adding dry dimethyl formaldehyde (DMF; 25 mL) to a round bottom flask (50 mL)
charged with 2,3-dihydroxybenzaldehyde (1.00 g; 7.24 mmol) and imidazole (1.97 g; 28.96
mmol). The mixture was stirred at room temperature until the solids dissolved (~5 minutes).
Then, tert-butyldimethylsilyl chloride (2.29 g; 15.20 mmol) was added in one portion and the
solution was stirred under argon. After one hour, the room temperature solution was transferred
to a separatory funnel and diethyl ether (Et2O; 20 mL) and water (H2O; 20 mL) were added and
the mixture shaken. The aqueous layer (bottom) was discarded, the organic layer (top) washed
with brine (3 x 20 mL), and the organic layer dried with magnesium sulfate, MgSO4. The solvent
was removed under reduced pressure using rotary evaporation leaving a colorless oil. Slight
heating of this oil (30 °C) under vacuum for 24 h ensures removal of any silyl ether byproducts,
1
and Int–1a was isolated as a pale yellow oil (2.26 g; 6.16 mmol; 85% crude yield). H NMR
(400 MHz, CDCl3: 7.26 ppm) δ 10.39 (d, J = 0.8 Hz, 1H), 7.42 (dd, J = 7.8, 1.7 Hz, 1H), 7.10
(dd, J = 7.9, 1.7 Hz, 1H), 6.92 (td, J = 7.8, 0.8 Hz, 1H), 1.05 (s, 9H), 0.99 (s, 9H), 0.25 (s, 6H),
0.17 (s, 6H). Please note that in the exposing the crude mixture for shorter times to vacuum does
not remove all of the DMF and bis(tert-butyldimethylsilyl)ether contaminants. However, these
15

ACCEPTED MANUSCRIPT
byproducts do not affect any subsequent synthetic steps and the crude product can be carried
through the synthetic steps described below without further purification.
2,3-Bis(tert-Butyldimethylsilyl ether)Benzyl Alcohol (Int–1b).
The second intermediate described in Scheme 3 was prepared on the benchtop without exclusion
of air and moisture by charging a round bottom flask (100 mL) with Int–1a, (2.26 g; 6.16 mmol)
dissolved in methanol (MeOH; 25 mL). This solution was cooled using an ice-water bath. Then,
four equivalents of solid sodium borohydride (NaBH4; 0.206 g; 5.45 mmol) was added slowly
(30 s). The flask was equipped with a glass stopper and its contents stirred (3 h; 0 °C → room
temp). Subsequently, the solution was transferred to a separatory funnel, and diethyl ether (Et2O;
20 mL) and water (H2O; 20 mL) that had been saturated with ammonium chloride, NH4Cl, were
added slowly, taking care to relieve any pressure buildup. The aqueous layer (bottom) was
discarded, the organic layer (top) was washed with water saturated with NH4Cl (2 x 20 mL), and
the organic layer dried with magnesium sulfate, MgSO4. The solvent was removed under
reduced pressure using rotary evaporation (1 h) and Int–1b was isolated as a colorless oil (1.95
g; 5.29 mmol) in 86% yield. 1H NMR (400 MHz, CDCl3: 7.26 ppm) δ 6.95 (dd, J = 7.2, 2.1 Hz,
1H), 6.86 – 6.78 (m, 2H), 4.67 (s, 2H), 1.02 (s, 9H), 0.96 (s, 9H), 0.22 (s, 6H), 0.17 (s, 6H).
2,3-Bis(tert-Butyldimethylsilyl Ether)Benzyl Chloride (Int–1c).
The third intermediate in Scheme 3 was prepared on the benchtop without exclusion of air and
moisture by adding trichlorotriazine (1.07 g; 5.82 mmol) to a round bottom flask (50 mL)
charged with DMF (5 mL) that had been slightly heated (30 °C). The mixture was stirred (2 h)
and solids that precipitated were occasionally broken up with a glass stirring rod.
16

ACCEPTED MANUSCRIPT
Dichloromethane (CH2Cl2; 25 mL) was added followed by a solution of Int–1b (1.95 g; 5.29
mmol) that had been dissolved in CH2Cl2 (5 mL). The reaction was monitored by NMR
spectroscopy and diethyl ether (Et2O; 30 mL) was added when the starting material was
consumed (~3 h). This caused a yellow solid to precipitate. The mixture was filtered through
Celite and the filter cake was washed with Et2O (2 x 10 mL). The filtrate was transferred to a
separatory funnel. The aqueous layer (bottom) was discarded, the organic layer (top) was washed
with brine (1 x 30 mL), and the organic layer dried with magnesium sulfate, MgSO4. The solvent
was removed by rotary evaporation and Int–1c was isolated as a yellow oil (1.64 g; 4.23 mmol)
1
in 80% yield. H NMR (400 MHz, CDCl3: 7.26 ppm) δ 6.95 (dd, J = 7.2, 2.1 Hz, 1H), 6.87 –
6.77 (m, 2H), 4.67 (s, 2H), 1.02 (s, 9H), 0.96 (s, 9H), 0.22 (s, 6H), 0.17 (s, 6H). If desired, minor
impurities can be removed by silica gel chromatography (10% Et2O:hexanes). However, we
found it more convenient to carry the crude product through the following steps without further
purification.
N,N’-[2,3-Bis(tert-Butyldimethyl Silyl Ether)Benzyl]-4,13-Diaza-18-Crown-6 (Int–
1d).
The fourth intermediate shown in Scheme 3 was prepared on the benchtop without exclusion of
air and moisture by combining 4,13-diaza-18-crown-6 (Kryptofix 22, 501 mg; 1.91 mmol) and
N,N-diisopropylethylamine (920 uL; 5.29 mmol) with anhydrous acetonitrile (MeCN; 25 mL) in
a round bottom flask (50 mL) equipped with a magnetic stirrer. To this solution, Int–1c (1.64 g;
4.23 mmol), also dissolved in anhydrous MeCN (10 mL), was added. The mixture was heated to
reflux under argon. After 3 h, the resulting mixture was cooled to room temperature and white
crystals precipitated over 12 hours. Note: cooling the mother liquor below room temperature
17

ACCEPTED MANUSCRIPT
causes precipitation of a red impurity that is difficult to remove. The white solid was collected
by vacuum filtration, recrystallized from hot cyclohexane and washed with water and
acetonitrile. After drying the solids under vacuum (overnight), Int–1d was isolated as a white
crystalline solid (684 mg; 0.71 mmol; 37% yield). 1H NMR (400 MHz, CDCl3: 7.26 ppm) δ 7.09
(dd, J = 7.7, 1.8 Hz, 2H), 6.78 (t, J = 7.8 Hz, 2H), 6.72 (dd, J = 8.0, 1.8 Hz, 2H), 3.66 (s, 4H),
3.59 (s, 8H), 2.79 (d, J = 5.8 Hz, 4H), 1.02 (s, 18H), 0.94 (s, 18H), 0.20 (s, 12H), 0.12 (s, 12H).
N,N’-Bis(2,3-dihydroxybenzyl)-4,13-Diaza-18-Crown-6
Dihydrochloride
Monohydrate (H4ChaCha-2HCl) From Int–1d.
The H4ChaCha–2HCl product was prepared from Int–1d (Scheme 3) on the benchtop without
exclusion of air and moisture by adding hydrochloric acid (1 mL; 12 M) to a scintillation vial
that contained a stirred solution of Int–1d (684 mg; 0.71 mmol) dissolved in methanol (MeOH;
10 mL). After gently heating the solution (40 °C; 3 h) under air, the solvent was removed by
rotary evaporation. The resulting sticky colorless oil was exposed to vacuum (40 °C; 24 h),
which enabled H4ChaCha-2HCl to be isolated as a fluffy beige hygroscopic powder (418 mg;
1
0.70 mmol; ~99% yield). H NMR (400 MHz, D2O: 4.79 ppm) δ 7.02 (2H, m), 6.92 (4H, m),
4.49 (4H, s), 3.87 (8H, br s), 3.53 (18H, m). Anal. Calcd for C26H42Cl2N2O9: C, 52.26; H, 7.09;
N, 4.69. Found: C, 52.01; H, 7.01; N, 4.87.
Synthetic Route 2 to N,N’-Bis(2,3-dihydroxybenzyl)-4,13-Diaza-18-Crown-6
Dihydrochloride
Monohydrate
(H4ChaCha-2HCl)
via
Mannich
condensation.
18

ACCEPTED MANUSCRIPT
2-(tert-Butyldimethylsilyl Ether)phenol (Int–2a).
The first intermediate in Scheme 4 was prepared on the benchtop without exclusion of air and
moisture by adding catechol (5.00 g; 45.4 mmol) to a round bottom flask (50 mL) containing
cold DMF (0˚ C, 25 mL) and was stirred until the catechol was fully dissolved (~5 min).
Imidazole (6.49 g; 95.3 mmol) and then tert-butyldimethylsilyl chloride (6.84 g; 45.4 mmol)
were added, and stirring was continued (3 h; 0 °C → room temp). The reaction was quenched
with water (20 mL), stirred (5 min) and the cloudy mixture was transferred to a separatory funnel
with the assistance of Et2O (30 mL). The aqueous layer was separated and discarded. The
organic layer was washed with water (25 mL x 3), then brine (25 mL x 1), and dried with sodium
sulfate. The solution was filtered through a pad of silica gel, and solvent was removed by rotary
1
evaporation. The product was isolated in 79% crude yield (8.04 g; 35.9 mmol). H NMR (400
MHz, CDCl3: 7.26 ppm) δ 7.02 (dd, J = 7.9, 1.7 Hz, 1H), 6.92 (m, 2H), 6.81 (td, J = 7.7, 1.7 Hz,
13
1H), 5.60 (s, 1H), 1.09 (s, 9H), 0.34 (s, 6H). C NMR (101 MHz, CDCl3: 77.2 ppm) δ 147.26,
142.33, 122.13, 119.94, 117.85, 114.91, 25.69, 18.14, -4.39. Although vacuum distillation
improves purity, the compound can be used in subsequent synthetic steps without further
purification.
N,N'-[2-hydroxy-3-(tert-Butyldimethylsilyl Ether)]-4,13-Diaza-18-Crown-6 (Int–2b).
The second intermediate in Scheme 4 was prepared on the benchtop without exclusion of air and
moisture by combining kryptofix-22 (1.5 g; 5.72 mmol), paraformaldehyde (435 mg; 14.5
mmol), and Int–2a (2.82 g; 12.6 mmol) in a round bottom flask (100 mL) charged with toluene
(50 mL) and equipped with an air-cooled condenser. The cloudy mixture was stirred (40 ˚C;
overnight) and the solids dissolved generating a clear and colorless solution. The solvent was
19

ACCEPTED MANUSCRIPT
removed by rotary evaporation leaving a colorless oil. After 3 h at room temperature, a white
solid precipitated from the oil. The oil was removed with repeated cold (-10 ˚C) hexane washes
(10 mL x 3, warm hexane or excess washes will reduce yield) and the white solid was isolated by
vacuum filtration. Drying the product under reduced pressure provided Int–2b in 42% yield
1
(1.75 g; 2.38 mmol). H NMR (400 MHz, CDCl3: 7.26 ppm) δ 6.77 (dd, J = 7.2, 2.2 Hz, 2H),
6.64 – 6.56 (m, 4H), 3.77 (s, 4H), 3.66 (t, J = 5.5 Hz, 8H), 3.62 (s, 8H), 2.84 (t, J = 5.5 Hz, 8H),
13
1.01 (s, 18H), 0.20 (s, 12H). C NMR (101 MHz, CDCl3: 77.2 ppm) δ 149.17, 143.76, 123.38,
121.55, 120.54, 118.39, 70.87, 69.15, 58.39, 53.43, 25.83, 18.49, -4.48. This crude product was
of sufficient purity to be carried on in subsequent synthetic steps.
N,N’-Bis(2,3-dihydroxybenzyl)-4,13-Diaza-18-Crown-6
Dihydrochloride
Monohydrate (H4ChaCha-2HCl) From Int–2b.
The H4ChaCha–2HCl product was prepared from Int–2b (Scheme 4) on the benchtop without
exclusion of air and moisture by adding hydrochloric acid (1 mL; 12 M) to a scintillation vial
that contained a stirred solution of Int-2b (1.75 g; 2.38 mmol) dissolved in methanol (MeOH; 10
mL). After gently heating the solution (40 °C; 3 h) under air, the solvent was removed by rotary
evaporation. The resulting sticky colorless oil was dried under vacuum (40 °C; 24 h), which
enabled H4ChaCha-2HCl to be isolated as a pale beige hygroscopic powder (1.35 g; 2.33 mmol;
98% yield). See above for characterization details.
t
N,N’-[2,3-Dihydroxy-5-tert-Butylbenzyl]-4,13-Diaza-18-Crown-6 (H4 Bu2ChaCha).
t
The H4 Bu2ChaCha product was prepared in a single step (Scheme 4) on the benchtop without
exclusion of air and moisture by dissolving 4,13-diaza-18-crown-6 (500 mg; 1.9 mmol) and
20

ACCEPTED MANUSCRIPT
paraformaldehyde (145 mg; 4.8 mmol) in anhydrous benzene (25 mL) in a round bottom flask
(50 mL). To this solution, 4-tert-butylcatechol (790 mg; 4.9 mmol) was added. The solution was
roughly degassed by exposure to vacuum followed by argon (3x) and heated at reflux under an
argon atmosphere (12 h). Next, the solvent was removed using rotary evaporation. The resulting
white solid was ground using a mortar and pestle, washed with diethyl ether, and dried under
vacuum (0.1 Torr; 30 °C; 12 h) to give a free-flowing white powder (1.02 g; 1.6 mmol, 84%
yield). Single crystals suitable for single crystal X-ray diffraction were obtained by slow
t
1
evaporation of an ethanolic solutions (EtOH) of H4 Bu2ChaCha. H NMR (400 MHz, CDCl3:
7.26 ppm) δ 6.90 (t, J = 2.3 Hz, 2H), 6.50 (d, J = 2.3 Hz, 2H), 3.79 (s, 4H), 3.66 (t, J = 5.3 Hz,
8H), 3.61 (s, 8H), 2.81 (t, J = 5.3 Hz, 8H), 1.25 (s, 18H). Anal. Calcd for C34H54N2O8: C, 65.99;
H, 8.80; N, 4.53. Found: C, 65.83; H, 8.79; N, 4.24.
[La2(tBu2ChaCha)2]2
In an argon filled glovebox, tris[N,N-bis(trimethylsilyl)amide] lanthanum(III), La[N(SiMe3)2]3
(367 mg; 0.644 mmol), was dissolved in tetrahydrofuran, THF (10 mL). The colorless solution
t
was added to a vial charged with H4 Bu2ChaCha (398 mg; 0.643 mmol). After stirring the
mixture overnight, all of the solids dissolved to generate a dark green solution. The solution was
filtered through glass wool and the solvent was removed under vacuum to give a sticky green
solid. For spectroscopic and analytical analyses this sticky green solid was washed with pentane
(3 x 1 mL) and dried at elevated temperature (40 ˚C) under reduced pressure (~0.1 Torr) to give a
free flowing pale green solid (443 mg, 0.146 mmol, 91% yield). For single crystal diffraction
studies, aforementioned sticky green solid was dissolved in a minimal amount of dry toluene (~1
mL), gently heated (40 °C), and filtered again through glass wool. Slow diffusion of diethyl
21

ACCEPTED MANUSCRIPT
ether, OEt2, into this solution generated a pale green precipitate (non-crystalline compound) and
dark blue block crystals of [La2(tBuChaCha)2]2 that were suitable for single crystal X-ray
Diffraction. Anal. Calcd for C136H204N8O32: C, 54.11; H, 6.81; N, 3.71. Found: C, 53.87; H,
7.07; N, 3.61. UV-vis 14,830 cm-1 (ε=18.5). The NMR spectrum was uninterpretable, owing to a
very large number of unresolvable peaks.
Crystallographic Studies.
Single crystals of [La2(tBuChaCha2)]2 and H4tBuChaCha were mounted on nylon loops with
Paratone-N oil (Hampton Research) and immediately cooled in a cold nitrogen stream on a D8
Bruker QUEST diffractometer. No corrections for crystal decay were necessary. Standard Apex
II software was used for determination of the unit cells and data collection control. The
intensities of the reflections of a sphere were collected by combining four sets of exposures
(frames), which totaled to 1464 frames with an exposure time of 10-60 seconds per frame,
depending on the crystal. Apex II software was used for data integration including Lorentz and
polarization corrections. Disorder associated with solvent molecules was accounted for by using
the SQUEEZE program in PLATON[57]. All of the atoms were refined anisotropically with the
acception of a few badly disordered carbon atoms in [La2(tBuChaCha2)]2. Attempts to model this
disorder anisotropically resulted in on going splitting of the carbon atoms. The data was
corrected for absorption using redundant reflections using the SADABS program. Structure
solution and refinement were performed using SHELX. The hydrogen atoms in H4tBuChaCha
were generated using HFIX function in the SHELXL program. The CIF files used in this
manuscript are available through the Cambridge Crystal Data Centre (CCDC 1573878 and
1573879).
22

ACCEPTED MANUSCRIPT
ACKNOWLEDGEMENTS
We gratefully recognize the Heavy Element Chemistry Program at LANL by the Division of
Chemical Sciences, Geosciences, and Biosciences, Office of Basic Energy Sciences, U.S.
Department of Energy (Kozimor, Mocko, Scott) and the LANL LDRD program (Berg,
Birnbaum). Los Alamos National Laboratory is operated by Los Alamos National Security, LLC,
for the National Nuclear Security Administration of U.S. Department of Energy (contract DE-
AC52-06NA25396). Portions of this work were supported by postdoctoral fellowships from the
Glenn T. Seaborg Institute (Stein) and the LDRD office, named fellowship program; Marie Curie
Distinguished Postdoctoral Fellowship (Cary).
Supporting Information
t
The CIF file for [La2(tBuChaCha)2]2 and H4 Bu2ChaCha are available through the Cambridge
Crystal Data Centre (CCDC 1573878 and 1573879).
23

ACCEPTED MANUSCRIPT
t
Scheme 1: The structure of H4ChaCha-2HCl and H4 Bu2ChaCha.
24

ACCEPTED MANUSCRIPT
Scheme 2: Electron transfer chemistry available to the catecholate ligand.
25

ACCEPTED MANUSCRIPT
Scheme
3:
Synthesis
of
N,N’-[2,3-dihydroxybenzyl]-4,13-diaza-18-crown-6
dihydrochloride (H4ChaCha
ñ
2HCl) via benzylic substitution of 2,3-bis(tert-
Butyldimethylsilyl Ether)benzyl chloride (Int-1c) and 4,13-Diaza-18-Crown-6.
26

ACCEPTED MANUSCRIPT
Scheme 4: Synthesis of N,N’-[2, 3-dihydroxybenzyl]-4,13-diaza-18-crown-6
dihydrochloride (H4ChaChañ2HCl) via Mannich condensation of 2-hydroxy-3-(tert-
Butyldimethylsilyl Ether) and 4,13-Diaza-18-Crown-6
27

ACCEPTED MANUSCRIPT
Scheme 5: Synthesis of N,N’-[2, 3-dihydroxy-5-tert-butylbenzyl]-4,13-diaza-18-crown-6
t
(H4 Bu2ChaCha).
28

ACCEPTED MANUSCRIPT
Scheme 6: Synthesis of [La2(tBu2ChaCha)]2.
29