The complexation of rhodium(III) with acyclic diaminedithioether (DADTE) ligands

Article abstract of DOI:10.1039/c0dt00813c

¹⁰³Rhodium(III) complexes derived from seven acyclic tetradentate N₂S₂ ligands (one diaminedithiol and six diaminedithioether ligands) have been synthesized and characterized. Structural variations in the ligand include th

Full text of DOI:10.1039/c0dt00813c

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/dalton | Dalton Transactions
The complexation of rhodium(III) with acyclic diaminedithioether (DADTE)
ligands†
Zeynep Akgun,a Hendrik Engelbrecht,b Kuo-Hsien Fan,a Charles Leslie Barnes,a Cathy Sue Cutler,b
Silvia Sabine Jurissona and Susan Zemyan Lever*a,b
Received 8th July 2010, Accepted 18th August 2010
DOI: 10.1039/c0dt00813c
103Rhodium(III) complexes derived from seven acyclic tetradentate N2S2 ligands (one diaminedithiol
and six diaminedithioether ligands) have been synthesized and characterized. Structural variations in
the ligand include the length of carbon backbone between the coordinating atoms (222; 232; 323; 333),
the presence or absence of gem-dimethyl groups a to sulfur, and the nature of the organic moiety on the
sulfurs (hydrogen, p-methoxybenzyl and methyl). For each ligand, the formation of cis and/or trans
dichloro isomeric complexes was assessed. Two complexes have been further characterized by single
crystal X-ray diffraction. Preparation of the 103Rhodium(III) complexes was conducted and overall
radiochemical yields, in vitro stability and log D7.4 values were measured. From these studies, the ligand
with the 232 chain length, gem-dimethyl groups and the methyl thioether (L4) emerged as a preferred
ligand for formation of rhodium complexes with trans geometry and highest radiochemical yields.
Introduction
For the past twenty years, there has been much interest in 105Rh
as a therapeutic radioisotope due to its nuclear properties. 105Rh
decays to 105Pd by b- emissions of 566 keV (75%), 248 keV (19.7%),
260 keV (5.2%) and 133 keV (0.042%). The presence of concomi-
tant g-ray emissions [Eg = 306 keV (5.1%), Eg = 319 keV (19%)]
provides the possibility for in vivo tracking of the therapeutic dose.
105Rh has a half-life of 35.36 h, which is sufficiently long to allow
the radiopharmaceutical to be manufactured, delivered and used
clinically worldwide.1
Various chemical classes, including nitrogen and sulfur-based
cyclic and acyclic ligands (Fig. 1a),2–10 have been evaluated
for the preparation of a unique 105Rh complex with maximum
stability in vivo. Rh(III) chloride forms octahedral complexes
with tetradentate ligands where two of the three chlorides are
retained. The two chlorides can be found trans or cis to each
other, and whereas one trans isomer can be formed, multiple cis
isomers are possible (Fig. 1b). The extent to which geometric
isomers result from a selected ligand is an important design
Fig. 1 a. Stylized representation of ligands investigated for complexation
with Rh. R = –CH2CO2H, Benzyl. Actual structures are described within
the text. b. Stylized trans and one of the possible cis Rh(III) complexes
of macrocyclic (upper) and acylic (lower) nitrogen and sulfur-based
tetradentate ligands. X = S, SR, NH, or NH2.
criterion, because isomeric complexes can exhibit different bio-
logical and pharmacological characteristics. Characterization of
the Rh(III) octahedral complexes from 1,4,7,10-tetraazadecane
(222-tet) showed that the cis-dichloro configuration was strongly
favored, while the 11-membered 1,4,8,11-tetraazaundecane (232-
tet) exclusively favored the trans-dichloro- configuration.2 In a
later study, Rh(III) complexes of the 12- to 16-membered cyclic
tetraamines, 1,4,7,10-tetraazacyclododecane (cyclen), 1,4,7,10-
tetraazacyclotridecane, 1,4,8,11-tetraazacyclotetradecane (cy-
clam), 1,4,8,12-tetraazacyclopentadecane and 1,5,9,13-tetra-
azacyclohexadecane, respectively, were synthesized.3–5 According
to this study, the 12- and 13-membered N4 macrocycles formed
only cis complexes, the 14-membered structure formed both cis
and trans complexes and the 15- and 16-membered N4 macrocycles
formed only the trans complex. Clearly, the macrocyclic ring size
affected the isomeric outcome observed for the nitrogen-based
ligands.
aThe Chemistry Department, University of Missouri, Columbia, MO, 65211.
E-mail: levers@missouri.edu; Fax: 573-882-8395; Tel: 573-882-2754
bthe MU Research Reactor Center University of Missouri, Columbia, MO,
65211
† Electronic supplementary information (ESI) available: 1H and 13C NMR
spectra of ligands and complexes as well as LC-MS results for final
complexes. CCDC reference numbers 781961 (trans-[RhCl2(L3)]PF6) and
781962 (trans-[RhCl2(L6)]PF6). For ESI and crystallographic data in CIF
or other electronic format see DOI: 10.1039/c0dt00813c
In 1989, Blake and co-workers investigated Rh(III) complexation
reactions with 12-, 14-, and 16-membered macrocycles containing
This journal is
The Royal Society of Chemistry 2010
Dalton Trans., 2010, 39, 10169–10178 | 10169
©

View Article Online
sulfur as the sole type of coordinating atom.6,7 As found for the N4
macrocycles, the 12-membered S4 macrocycle formed only the cis
complexes, while the 16-membered S4 macrocycle formed the trans
complex. However, in contrast to the 14-membered N4 macrocycle,
the trans isomer was not observed in the case of S4 macrocycle.
It was postulated that the larger atomic radius of sulfur versus
nitrogen was a contributing factor in this result.
Goswami et al.8 studied the Rh(III) complexation behavior
of several acyclic tetrathioether ligands. Structural modifications
included varying the number of carbons between the sulfur atoms
and the moiety on the terminal sulfurs (–CH2CO2H “AcOH”
and –Bz). The 222-S4-diAcOH and 222-S4-diBz ligand systems
formed a cis Cl–Rh–Cl core. The larger 333-S4-diAcOH, 323-S4-
diAcOH, and 333-S4-diBz ligands made trans complexes, while
232-S4-diAcOH and 232-S4-diBz formed both cis and trans
isomers in high yield. Radiolabeling studies were performed with
222-S4-diAcOH, 232-S4-diAcOH and 333-S4-diAcOH and the
radiolabeled complexes showed high in vitro stability of 98%, 96%
and 89%, respectively, over 5 days. Biodistribution studies were
performed with the complexes derived from the 222-S4-diAcOH
(pure cis complex) and 333-S4-diAcOH (pure trans complex).
Both complexes showed effective clearance from the blood and
all tissues, primarily through the kidneys and the urine due to
their +1 charge.9
Li et al.10 further studied the Rh(III) complexation properties
with 14-membered rings where either nitrogen or nitrogen/sulfur
combinations were utilized as coordinating atoms ([14]aneNS3,
[14]aneN2S2 and [14]aneN4). The reactions were performed on the
radiochemical scale with 105Rh(III)-chloride and analyzed by radio-
TLC and paper electrophoresis. Complexation yields decreased
with an increasing number of N atoms in the macrocycle (94% NS3,
81% N2S2 and 59% N4, respectively) and the results were consistent
with formation of cis octahedral products when compared with
non-radioactive standards.
Fig. 2 Ligands synthesized and evaluated in this study.
In this study, seven acyclic N2S2 chelating ligands (one di-
aminedithiol “DADT” and six diaminedithioether “DADTE”
ligands) (Fig. 2) have been synthesized and evaluated for their
ability to incorporate Rh(III).11,12 Characterization of the non-
radioactive 103Rh(III) complexes and the radioactive 105Rh(III)
complexes was conducted. The length of carbon chain between the
coordinating atoms (222; 232; 323; 333) in the ligand backbone
was modified. The presence or absence of gem-dimethyl groups as
well as the organic moiety on the sulfur atom (p-methoxybenzyl or
methyl) were investigated. The p-methoxy group or gem-dimethyl
groups can contribute bulk and add to the lipophilicity of the
resultant complex. This increase in lipophilicity can affect the
extent of non-specific binding in fatty tissue or contribute to
greater accumulation in non-target organs, such as the liver. Key
outcomes for this study include the number of complexes formed,
the radiochemical complexation yield, and in vitro stability of the
radioactive complexes.
concentration, pH, temperature, time of reaction and percent
ethanol), were unsuccessful. Instead, compounds with very high
molecular weight were observed by LC-MS suggesting the forma-
tion of aggregates due to thiol-bridged species13 (data not shown).
In an iterative fashion, six additional ligands were synthesized,
characterizedandcomplexation reactions were performed. Ideally,
the ligand identified from these studies would exhibit properties
that could be extended to conjugation with peptides for use
in radiotherapeutic applications via the bifunctional chelate
approach. These qualities include, but are not limited to, the
formation of one isomer upon complexation with rhodium(III),
suitable lipophilicity, and stability of the complex in vitro over
six days.
L2 (222-gdm-pmBz), where the free thiol groups were trans-
formed into thioethers through reaction with p-methoxybenzyl
chloride, was synthesized. The Rh(III) complexation reaction with
L2 was performed according to the general procedure and dark
orange crystals were obtained. HPLC chromatograms showed
two products eluting near to each other, in an approximately
50 : 50 ratio, presumably, the cis and trans isomers. Analysis by
LC-MS supported this assignment, since the same molecular
ions were observed for each peak with the expected isotopic
distribution. The composite product was obtained in 30% yield.
Results and discussion
Synthesis of 103Rh(III) complexes. L1
L1 (222-gdm-SH) was chosen as the initial chelating agent for
the Rh(III) complexation reaction. Several attempts to prepare
[RhCl2(L1)]PF6, using several experimental variations (ligand
10170 | Dalton Trans., 2010, 39, 10169–10178
This journal is
The Royal Society of Chemistry 2010
©

View Article Online
Scheme 1
The complexity of the 1H-NMR spectrum of [RhCl2(L2)]PF6, due
to the presence of isomers, made interpretation difficult. Before
the Rh(III) complexation, the symmetrical ligand exhibited first
order splitting; one H signal was observed for each unique –
CH2– and –CH3 group on L2. After complexation, the peaks
achieved with lithium aluminium hydride (Scheme 1). Alkylation
of the thiol groups afforded the ligand, L3.
Upon Rh(III) complexation with L3, dark orange crystals were
obtained. LC-MS analysis showed a minor product eluting prior
to one predominant product in a 5 : 95 ratio, each with the same
positive-ion peaks consistent with [RhCl2(L3)]PF6 formulation.
The 13C NMR studies were performed with the pure major product
in deuterated acetonitrile. Before the complexation, eleven 13C
signals were obtained for L3. After complexation with Rh(III), the
major isomer was no longer symmetric. Each individual carbon
atom, with the exception of the unsubstituted aromatic carbons,
was seen in the 13C NMR.
1
occurred downfield relative to the location of signals for each
1
hydrogen in the ligand and additional H signals were observed.
For example, rather than one singlet corresponding to all four –
CH3 groups in the ligand, four singlets were obtained, reflecting
the unique environment. However, the most characteristic peaks
of the 1H NMR spectrum for the ligand are the aromatic proton
signals around 7 ppm, the methoxy signal at 3.7 ppm and the
high-field methyl signal at 1.3 ppm. In addition to these signals,
the characteristic singlet –CH2– signal that was observed for the
–SCH2C6H4OCH3 on L2 became a doublet after complexation
because the two H’s become non-equivalent. It proved difficult to
identify the signals for the protons on the ethylene bridge due to the
strong geminal and vicinal couplings of these protons. The trends
observed for the 1H NMR analysis were similar to those found in
13C NMR analysis; the 13C signals occurred further downfield than
for the ligand, but overall, the 13C NMR spectrum proved to be
Single crystals of the major [RhCl2(L3)]PF6 isomer were pre-
pared by slow evaporation of an ethanol–acetonitrile solution,
and X-ray diffraction analysis confirmed the trans geometric
assignment. Fig. 3a shows the molecular structure of the cation
with the atomic labeling scheme. The ORTEP diagram shows the
complex to have slightly distorted octahedral geometry, with the
Rh(III) atom positioned in the center of the N2S2 cavity and the
two chlorine atoms in axial sites. The p-methoxybenzyl group on
the S(1) atom of the ligand lies above the N2S2 coordination plane
while the one on the S(2) lies below the N2S2 coordination plane.
Both N–H atoms are pointing toward Cl(2), and the lone pair of
the S(1) donors are pointing toward Cl(1) where the lone pair of
the S(2) donors are pointing toward Cl(2). The least-squares plane
somewhat less complicated. In the case of the ligand itself, ten 13
C
signals were observed for the twenty carbons on the symmetrical
ligand. A total of thirty 13C signals were observed for the isolated
mixture. The 13C NMR spectrum is consistent with the LC-MS
results that support formation of two isomers. 1H and 13C spectra
of ligands and complexes, as well as representative chromatograms,
are found in the supplementary information.
It was hypothesized that increasing the cavity size of the
ligand backbone would minimize the formation of geometric
isomers formed by Rh(III) complexes with DADTE ligands.
Therefore, L3 (232-gdm-pmBz), which has a larger cavity due to
the additional methylene group between the nitrogen atoms in the
backbone chain, but retains the gem-dimethyl groups and the p-
methoxybenzyl moieties on the sulfur atoms, was designed. L3 was
synthesized in a similar fashion as L1 and L2 via the macrocyclic
diiminedisulfide, formed in turn through the condensation of 2,
2¢-dithiobis(2-methylpropanal) with propylenediamine. Concomi-
tant reductions of the imine functionalities and disulfide bond were
˚
determination showed that the sulfur donor S(1) lies 0.0617 A
˚
above the plane, while S(2) lies 0.0611 A below the plane. Likewise,
˚
nitrogen donor N(1) lies 0.0689 A below the plane while N(2) lies
˚
˚
0.0683 A above the plain. The rhodium ion sits 0.0068 A above the
plane, in the direction of Cl(2). The values of the Rh–S and Rh–
N bond lengths in the equatorial ligand plane average 2.3343(7)
˚
and 2.100(2) A, respectively (Table 1). The distances between
the rhodium atom and the two axial chlorine atoms differ only
˚
slightly, (2.3358(7) and 2.3461(6) A), and compare well with the
Rh–Cl distances found in other rhodium-chloride complexes (e.g.,
9
˚
[RhCl2(232-S4-dibz)]PF6); 2.349(2) A). In this complex, there is
a central six-membered chelating ring with two five-membered
chelating rings oriented on opposite sides. All angles around the
rhodium center deviate from 90◦. The S–Rh–N angles in the
This journal is
The Royal Society of Chemistry 2010
Dalton Trans., 2010, 39, 10169–10178 | 10171
©

View Article Online
Fig. 3 a. ORTEP drawing of trans-[RhCl2(232-gdm-pmBz)]+ and b. trans-[RhCl2(323-Met)]+. Probability ellipsoids are shown at the 50% level.
◦
˚
Table 1 Selected bond lengths (A) and angles ( ) for trans-[RhCl2(232-
gdm-pmBz)]PF6 (trans-[RhCl2(L3)]PF6) and trans-[RhCl2(323-Met)]PF6
(trans-[RhCl2(L6)]PF6)
RRRS/RSSS, SSSR/SRRR, SRSS/RRSR, RSRR/SSRS and
SRRS/RSSR) and two pairs of meso complexes (SSRR/RRSS
and SRSR/RSRS) were formed for each cis and trans isomer.
Based on crystal structure analysis, it is found that trans-
[RhCl2(L3)]PF6 represents an SRSS stereoisomer. Since the
structure is centrosymmetric, the mirror image is also present but
not shown. Complete data and structure refinement for trans-
[RhCl2(L3)]PF6 are found in supplementary information.
trans-[RhCl2(232-gdm-
pmBz)]PF6 trans-
[RhCl2(L3)]PF6
trans-[RhCl2(323-Met)]-
PF6 trans-[RhCl2(L6)]PF6
Rh–N1
Rh–N2
Rh–S1
Rh–S2
Rh–Cl1
Rh–Cl2
2.101(2)
2.099(2)
2.3259(7)
2.3427(7)
2.3461(6)
2.3358(7)
2.1071(15)
2.0955(15)
2.3339(5)
2.3329(5)
2.3489(5)
2.3353(5)
Shifting the ratio of cis:trans isomers found in the prod-
uct distribution from 50 : 50 to 5 : 95 in [RhCl2(L2)]PF6 and
[RhCl2(L3)]PF6, respectively, supported the hypothesis that a
larger cavity was required for formation of a sole product; however,
L3 was not deemed the optimum ligand. L4 (232-gdm-Met) was
synthesized as an analog of L3 where compound 4 was alkylated
with methyl iodide rather than p-methoxybenzyl chloride to afford
a less bulky Rh(III) complex (Scheme 1). It was anticipated that
this structural modification would further improve the possibility
of obtaining only the trans isomer. The complexation reaction
with L4 was performed according to the general procedure, and
dark orange crystals were obtained in 42% yield. The product was
analyzed by HPLC and LC-MS. The LC-MS spectrum displayed
only one peak with the expected isotopic distribution, supporting
that a single isomer of [RhCl2(L4)PF6] is formed. For the solid
N2–Rh–N1
N2–Rh–S1
N1–Rh–S1
N2–Rh–Cl2
N1–Rh–Cl2
S1–Rh–Cl2
N2–Rh–S2
N1–Rh–S2
S1–Rh–S2
Cl2–Rh–S2
N2–Rh–Cl1
N1–Rh–Cl1
S1–Rh–Cl1
Cl2–Rh–Cl1
S2–Rh–Cl1
94.83(9)
176.93(7)
84.75(6)
88.24(7)
84.72(7)
94.78(2)
85.35(6)
176.27(7)
95.27(2)
88.56(2)
89.81(7)
89.24(7)
87.14(2)
176.24(2)
94.48(2)
84.44(6)
174.86(4)
92.17(4)
87.88(4)
84.12(4)
95.633(17)
96.48(4)
179.07(4)
86.918(17)
95.815(19)
89.19(4)
95.43(5)
87.275(17)
177.069(17)
84.689(18)
product obtained from the complexation reaction with L4, the 13
C
NMR lent support to the isolation of a single isomer.
Based upon these results, the preferred backbone appeared
to be the 232 ligand system. Replacement of the bulky p-
methoxybenzyl groups with methyl groups also achieved one of
the major objectives of the study; i.e., synthesis of a ligand that
would form only one Rh(III) complex. However, the presence
of the gem-dimethyl groups on the backbone contributes to the
lipophilicity of the ligand. We were interested in reducing the
lipophilicity of the final complex without jeopardizing formation
of a sole trans product; thus, L5 (232-Met), a direct analog of
L4 without the gem-dimethyl groups, was synthesized. The route
of synthesis (Scheme 2) required the step-wise construction of the
N2S2 core. 3-bromopropionyl chloride was used as the bifunctional
five-membered rings are close to the expected value of about 82◦
(range 84.75◦ and 85.35◦), whereas the N–Rh–N angle formed by
the six-membered ring is 94.83◦. These large values allow the metal
atom to lie in the plane of the N2S2 ligand.
The two nitrogen and two sulfur atoms on the N2S2 ligand
system become chiral stereocenters S(1), S(2), N(1), and N(2)
upon coordination with Rh(III). Each stereogenic center has
two possible stereoisomers, R or S. Thus, all possible combi-
nations of the four centers should have produced eight pairs
of enantiomers; however, due to the plane of symmetry in the
N2S2 Rh(III) complex, 6 pairs of enantiomers (RRRR/SSSS,
10172 | Dalton Trans., 2010, 39, 10169–10178
This journal is
The Royal Society of Chemistry 2010
©

View Article Online
ligand to attach two molecules of 2-(methylthio)ethylamine to
form intermediate 3, followed by reduction of the adduct by the
borane tetrahydrofuran complex to yield L5. The non-radioactive
rhodium complex was prepared and the reaction mixture was
analyzed by HPLC and LC-MS, 1H NMR and 13C NMR
spectroscopy. A single or at least one predominant, geometric
isomer was anticipated upon complexation, since L5 has the same
backbone chain as L3 and L4. However, according to the LC-MS
results, the formation of at least two isomers was observed, each
consistent with the molecular formulation of [RhCl2(L5)]PF6. This
complex was not pursued further.
six signals were observed corresponding to the methyl groups on
the complex (two signals for the trans-isomer and one signal for
each cis-pair). This result was consistent with the 13C findings
confirming the observation of five isomers. For example, six unique
signals were observed for the S-methyl group.
Based on the LC-MS and 13C NMR results it was found that
Rh(III) forms at least 5 isomers with L6; however, only the trans
isomer of [RhCl2(L6)]PF6 was crystallized. Crystals of complex
L6 suitable for an X-ray diffraction study were grown by diffusion
of methanol into an acetonitrile solution of the complex. The
molecular structure of trans-[RhCl2(L6)]PF6, showed a slightly
distorted octahedral geometry (Fig. 3b). Crystal structure analysis
confirmed the expected octahedral coordination geometry of
the complex cation, trans-[RhCl2(L6)]+. The tetradentate ligand
coordinates the rhodium atom in a planar fashion, with two
chloride atoms occupying the axial positions of the complex
(Table 1). The octahedral coordination polyhedron is slightly
distorted. The least-squares plane determination showed that the
To understand if the unexpected multiple isomer formation with
L5 occurred due to the absence of the gem-dimethyl groups on the
backbone, and if it can be overcome by increasing the length of
the backbone chain even further, ligands L6 (323-Met) and L7
(333-Met) were synthesized. In the case of L6, the number of
carbons between nitrogen and sulfur was increased to three, with
two carbons between the nitrogens; in the case of L7 (333-Met),
three carbons were placed between all coordinating atoms. These
two ligands were synthesized by analogy to the synthesis of L5;
viz., reaction of the appropriate haloalkylated acyl halide with 3-
(methylthio)propylamine (Scheme 2). Rh(III) complexation reac-
tions with L6 and L7 were performed according to the general pro-
cedure and the dark orange [RhCl2(L6)]PF6 and [RhCl2(L7)]PF6
crystals were obtained with 44% and 40% yield, respectively.
In each case, LC-MS spectroscopy indicated the presence of at
least five isomers with two major and three minor positive-ion
peaks with the correct isotopic distribution, approximately in a
47 : 47 : 2 : 2 : 2 ratio for both [RhCl2(L6)]PF6 and [RhCl2(L7)]PF6.
1H and 13C NMR studies were performed with the isolated
products [RhCl2(L6)]PF6 and [RhCl2(L7)]PF6 in deuterated ace-
tonitrile. Other than the central carbon signal contained in the
propylene bridge, all 13C signals for [RhCl2(L6)]PF6 occur further
˚
sulfur donor S(1) lies 0.0352 A above the plane, while S(2) lies
˚
0.0352 A below the plane. Likewise, nitrogen donor N(1) lies
˚
˚
0.0413 A below the plane, while N(2) lies 0.0401 A above the plane.
˚
The rhodium ion sits 0.0359 A above the plane, in the direction of
Cl(2). Both methyl groups of the ligand lie on the same side of the
N2S2 coordination plane, with the N–H atoms pointing toward
Cl(2) and the lone pair on each of the S donors pointing toward
Cl(1). Thus, the donor stereocenters S(l), S(2), N(l), and N(2) in
the complex have an RSRS absolute configuration, respectively.
The values of the Rh–S and Rh–N bond lengths in the equatorial
˚
ligand plane are 2.3339(5) and 2.1071(15) A, respectively. The
distances between the rhodium atom and the two axial chlorides
˚
differ only slightly (2.3353(5) and 2.3489(5) A) and compare well
with the Rh–Cl distances found in other rhodium(III)-chloride
5
˚
complexes (e.g., [RhCl2(232-S4-dibz)]PF6) (Rh–Cl) 2.349(2) A).
All angles around the rhodium center deviate from 90◦. The S–
Rh–N angles in the equatorial plane are larger than 90◦ (96.48 and
92.17◦), and the S–Rh–S and N–Rh–N angles are smaller than 90◦
1
downfield than the ligand. The H spectrum was very difficult
to interpret, since the hydrogens on each methylene become
nonequivalent and adopt an individual chemical shift. However,
Scheme 2
This journal is
The Royal Society of Chemistry 2010
Dalton Trans., 2010, 39, 10169–10178 | 10173
©

View Article Online
Table 2 Selected results of coordination reactions with 103/105Rh(III)
Chemistry with 103Rh(III)Cl3
Radiochemistry with 105Rh(III)Cl3
% Overall Labeling Efficiency
Compound
103/105RhCl2(L)]Cl
[
Number of isomers
cis:trans HPLC peak ratios
Log D7.4
L2
L3
L4
L5
L6
L7
2
2
1
5
5
5
50 : 50
5 : 95
0 : 100
~ 47 : 47:[cis -2:2 : 2]
~ 47 : 47:[cis -2:2 : 2]
~ 47 : 47:[cis -2:2 : 2]
46
46
61
53
56
48
ND
1.90 0.07
0.46 0.48
-0.26 0.50
ND
ND
(86.92 and 84.44◦, respectively). In contrast to L2, this complex
has a central five-membered chelating ring with two six-membered
chelating rings oriented on the opposite sides. We note that smaller
central bite angles are observed only when the complex has a
five-membered chelate ring. The bite angle for the five-member
chelating ring is 84.44◦ and for six-member chelating rings are
92.17◦ and 96.48◦.
Before Rh(III) complexation with L7 six 13C signals were
observed, but after coordination to the Rh(III), the 13C NMR
spectrum indicated the formation of five isomers. A similar trend,
as seen with the complex derived from L6, was observed from the
13C NMR spectrum, where six peaks were observed for each of
the symmetric carbon groups and five peaks were observed for the
central methylene group.
These results were surprising, because increasing the distance
between coordinating atoms was expected to decrease the forma-
tion of multiple isomers. However, a possible rationale for the
number of isomers could be due to increased structural fluidity
from the absence of the gem-dimethyl groups on the backbone
chain.
percent intact complex in peak 2 decreased from 99% to 76%, in
168 h. This result suggests that both isomers, were ~76% stable
over 7 days. In the case of [105RhCl2(L4)]Cl, the percent intact
complex decreased from 98% to 88% in 144 h, indicating that the
radiolabeled complex was 90% stable over 6 days.
In the case of [105RhCl2(L5)]Cl, the percent intact complex in
peak 1, decreased from 84% to 77%, and the percent intact complex
in peak 2 decreased from 82% to 75%, in 72 h. This result suggests
that both peaks, cis/trans isomers, were 91% stable over 3 days.
Overall, greater than 90% stability was observed
for
105RhCl2(L5)]Cl, while greater than 76% stability was observed for
[
105RhCl2(L4)]Cl and the isomeric formulations of
[
each isomer of [105RhCl2(L3)]Cl. From the previously published
literature, it has been shown that p-methoxybenzyl groups on
the ligand are not as stable as alkyl groups. In addition, they
are bulkier, which may increase the chances of exposure to
self-radiolysis. Therefore, these structural features may result in a
less stable radiolabeled complex.
The distribution coefficients (log D)
The distribution coefficients between octanol and water at pH 7.4
Radiolabeling reactions with 105Rh
were determined with the shake-flask method for [105RhCl2(L3)]Cl,
[
105RhCl2(L4)]Cl and [105RhCl2(L5)]Cl. These 105Rh complexes
To investigate the similarities and the differences between the
macroscopic and the radiotracer level in Rh(III) complexation
reactions, the radiolabeling reactions with 105Rh and L1–L7 were
performed parallel to the Rh(III) complexation reactions. The
identity of [105RhCl2(L)]Cl was verified by co-injecting radiola-
beled with the non-radioactive complexes onto the radio-HPLC
system. Determination of the yields were based on the area of
the intact [105RhCl2(L)]Cl complexes detected on the radio-HPLC
chromatogram. Similar to the results obtained in the Rh(III)
complexation reaction at the macroscopic level with L1, no 105Rh
radiolabeling was observed at the tracer level with L1. However,
in the case of ligands L2–L7 the 105Rh labeling was successfully
accomplished. The labeling efficiencies for the DADTE ligands
ranged from 46–61% (Table 2).
were found to be very hydrophilic due to the overall charge on
the complex (Table 2). As expected, for complexes [105RhCl2(L4)]Cl
and [105RhCl2(L5)]Cl, a decrease in log D values was found relative
to [105RhCl2(L3)]Cl. Substituting the p-methoxy group with a
methyl group decreased the log D value by 1.44 units, and in
the case of [105RhCl2(L5)]Cl, removal of the gem-dimethyl groups
on the backbone chain decreased the log D value by an additional
0.72 units. Therefore, an effect on the resulting log D values was
observed as moieties were removed and replaced with smaller
groups or hydrogen atoms. However, these complexes all proved to
be quite hydrophilic; thus, the presumption that the gem-dimethyl
groups on the ligand would adversely contribute to the overall
lipophilicity of the complexes proved less important.
Stability studies of [105RhCl2(L3)]Cl, [105RhCl2(L4)]Cl
[
Experimental
105RhCl2(L5)]Cl
Materials and methods
The stability of HPLC-purified [105RhCl2(L3)]Cl (peak 1 (cis) and
peak 2 (trans)) [105RhCl2(L4)]Cl and [105RhCl2(L5)]Cl (peak 1 (cis)
and peak 2 (trans)) were performed in phosphate-buffered saline
(PBS) buffer (pH = 7.5) over 3 to 7 days at 37◦ C in the presence
of gentisic acid (GA). In the case of [105RhCl2(L3)]Cl, the percent
intact complex in peak 1, decreased from 97% to 76%, and the
All chemicals were obtained from commercial sources (Aldrich
Chemical Co.) and used as received unless otherwise stated.
Trifluoroacetic acid (TFA) was obtained from VWR Scientific
Products (St. Louis, MO). Tetrahydrofuran was freshly distilled
from sodium benzophenone ketyl.
10174 | Dalton Trans., 2010, 39, 10169–10178
This journal is
The Royal Society of Chemistry 2010
©

View Article Online
HPLC analysis of the non-radioactive complexes was performed
on a Beckmann Coulter System Gold HPLC equipped with a 168
diode array detector, a 507e autoinjector and the 32 KARAT
software package (Beckmann Coulter, Fullerton, CA). A C-18
Kromosil column (150 ¥ 4.6 mm, 5 mm, 100 A, Keystone Scientific
Inc, San Jose, CA) was used for the analytical HPLC.
mixture was cooled to -78 ◦C, quenched with 20 mL of saturated
aqueous Na2SO4, diluted with 50 cm3 of diethyl ether, and then
allowed to warm to room temperature with vigorous stirring. The
pH of the resulting paste was lowered to ~9 by the addition of 1 M
HCl. The remaining white precipitate was removed by filtration,
and the filtrate was washed with water (2 ¥ 100 cm3), dried over
anhydrous Na2SO4, and then concentrated to afford 2 as a pale
yellow oil (2.12 g, 60%). dH(300 MHz; CDCl3) 2.75 (4 H, t, J 7,
NCH2–CH2–CH2N), 2.60 (4 H, s, 2 X C(CH3)2CH2), 1.69 – 1.75
(8 H, m, J 7, 2 X NCH2–CH2–CH2N, SH, NH), 1.38 (12 H, s,
2 X C(CH3)2; dC(75 MHz; CDCl3) 63.34, 48.60, 44.97, 30.41 and
28.33. This material was immediately used in the next reaction.
˚
ESI-MS analyses were performed on a Finnigan TSQ7000
mass spectrometer (Thermo Finnigan, San Jose, CA). A NovaPak
column (3.9 ¥ 300 mm, Waters Corp, Milford, MA) was used for
the LC-MS analysis. In all HPLC experiments, eluents used in all
runs were water (A), and acetonitrile (B), each containing 0.1%
TFA at 1 cm3 min-1 while being monitored at 214/280 nm.
Elemental analyses were determined by Atlantic Microlab
N,N¢-bis(2-(4-Methoxybenzylthio)-2-methylpropyl)propane-1,
3-diamine, L3 (232-gdm-pmBz). To a stirred solution of 2 (2.0 g,
8.0 mmol) in ethanol (50 cm3), aqueous NaOH (50 cm3, 2.5 M)
was added followed by neat p-methoxybenzyl chloride (4.3 cm3,
32 mmol). Stirring was continued for 1 h and most of the
ethanol was distilled under reduced pressure. The mixture was then
extracted (3 ¥ 20 cm3) with diethyl ether. The combined ethereal
solutions were dried with anhydrous Na2SO4. After filtration, the
diethyl ether was evaporated and ethanol saturated in HCl(g) was
added until the pH was 2–3. The warm mixture was cooled to
room temperature and the product was precipitated with ether.
The precipitate was filtered and washed with ether to yield L3 as
a white di-HCl salt (6.99 g, 76%), mp 285 ◦C. Found: C, 57.2; H,
7.9; N, 4.9. Calc. for C27H42N2O2S2·2HCl: C, 57.5; H, 7.9; N, 5.0%.
dH(300 MHz; D2O) 7.28 (4 H, d, J 8.7, aromatic H), 6.88 (4 H, d, J
8.7, aromatic H), 3.76 (4 H, s, 2 X SCH2), 3.70 (6 H, s, 2 X OCH3),
3.26 (4 H, s, 2 X C(CH3)2CH2), 3.01 (4 H, m), 2.62 (2 H, m), 1.31
(12 H, s, 2 X C(CH3)2); dC(75 MHz; D2O) 156.0, 130.6, 130.1,
114.4, 55.8, 55.3, 45.1, 44.0, 31.5, 25.7 and 21.6. ESI-MS: m/z
490.27 (M+, 100%), 491.27 (31.6); found: (M+H)+ 491.2, 492.3.
1
(Norcross, GA). H NMR (250, 300 and/or 500 MHz) and 13C
NMR spectra were obtained with Bruker instruments (Bruker
BioSpin, Westmont, IL). J values are given in Hz. Radiolabeling
analyses were performed on a Varian Pro Star HPLC (Model
240 pump) equipped with a Phenomenex Jupiter C-18 (4.1 ¥
250 mm, 5 mm) column with water (A), and acetonitrile (B)
eluents, each containing 0.1% TFA at 1 cm3 min-1 while being
monitored at 280 nm.
Sep-Pak Vac (0.1 g) cartridges were obtained from Waters Inc,
Milford, MA. Thin layer chromatography plates (MKC-18F Silica
˚
gel 60 A 2.5 ¥ 7.5 cm, 200 mm thickness) were obtained from
Whatman International Ltd, Madison, England. Quantitation
of the radioactivity on the TLC strips was performed using a
radiochromatographic strip scanner (BioScan System 200); the
radioactivity in the extracts was quantitated by counting aliquots
of the solutions in a Na(Tl)I well counter. Paper chromatographic
strips were developed with acetonitrile containing NH4PF6.
Chemistry
Ligand synthesis. 2,2,9,9-tetramethyl-4,7-diaza-1,10-decan-
edithiol, (L1, 222-gdm-SH) and N,N¢-bis-(2-(4-methoxy-
benzylthio)-2-methylpropyl)ethane-1,2-diamine (L2, 222-gdm-
pmBz) were synthesized as previously described.14
N,N¢-bis(2-Methyl-2-(methylthio)propyl)propane-1,3-diamine,
L4 (232-gdm-Met). L4 was prepared from 2 (1 g, 4 mmol) and
methyl iodide (0.75 cm3, 12 mmol) using the procedure described
for the preparation of L3. After purification, L4 was obtained
as the white di-HCl salt (0.6 g, 44%). Found: C, 44.4; H, 9.2;
N, 8.0. Calc. for C13H30N2S2·2HCl: C, 44.1; H, 9.15; N, 7.9%.
dH(500 MHz; D2O) 3.18–3.14 (4 H, m, NCH2–CH2–CH2N), 3.11
(4 H, s, 2 X C(CH3)2CH2), 2.18–2.14 (2 H, m, NCH2–CH2–CH2N),
1.98 (6 H, s, 2 X SCH3), 1.30 (12 H, s, 2 X C(CH3)2); dC(125 MHz;
D2O) 54.74, 45.30, 41.70, 25.19, 21.62 and 9.86. ESI-MS: m/z
278.19 (M+, 100%); found: (M+H)+ 279.10.
1,2-Dithia-5,9-diaza-3,3,11,11-tetramethyl-cycloundeca-4,9-
diene (1). 2,2¢-dithiobis(2-methylpropanal) 1114 (10.0 g,
0.0485 mol) and 1,3-diaminopropane (4.05 cm3, 0.0485 mol) were
refluxed in benzene (100 mL) in a reflux apparatus equipped
with a Dean Stark trap. After ~ 0.57 cm3 (0.097 mol) of H2O
was collected (8 h), the reaction was cooled and concentrated
to afford a yellow oil. The oil was dissolved in ethyl acetate and
passed through silica. Removal of the solvent gave the crude
product, which was recrystallized to yield 1 (4.0 g, 34%) as
white crystals: mp 103 ◦C (from EtOAc). Found: C, 54.1; H,
8.2; N 11.6. Calc. for C11H20N2S2: C, 54.05; H, 8.25; N, 11.5%.
dH(250 MHz; CDCl3) 7.54 (2 H, s, 2 X N CH), 3.80–3.73 (2 H,
m, NCH2–CH2–CH2N), 3.60–3.52 (2 H, m, NCH2–CH2–CH2N),
2.07 (2 H, apparent p, NCH2–CH2–CH2N), 1.39 (6 H, s, 2 X
CH3); 1.30 (6 H, s, 2 X CH3); dC(62.5 MHz; CDCl3) 168.41,
60.81, 50.71, 24.75, 24.65 and 24.16.
N-(2-(methylthio)ethyl)-3-(2-(methylthio)ethylamino)propana-
mide (3). To a stirred solution of 3-bromopropionyl chlo-
ride (4.6 g, 27 mmol) in CH2Cl2 (35 cm3) at -78 ◦C, 2-
(methylthio)ethylamine (4.9 g, 54 mmol) and triethylamine
(19 mL, 136 mmol) in 35 cm3 of CH2Cl2 were added dropwise
over 15 min. After the addition, the reaction mixture was warmed
to room temperature and stirring was continued for 24 h. Water
was added (20 cm3) and the layers were separated. The organic
portion was washed with 1M HCl, saturated NaHCO3 and brine
(20 cm3 each). The organic layer was dried over anhydrous Na2SO4,
filtered, and the solvent was removed under reduced pressure. For
purification, the crude product was loaded onto a short-path silica
column and washed with CHCl3. The pure product was eluted
with MeOH to afford a white solid (1.27 g, 23%). dH(250 MHz;
2,2,10,10-Tetramethyl-4,8-diaza-1,11-undecanedithiol (2). The
reduction was modeled after a literature procedure15 for a similar
compound. Excess LiAlH4 (51.2 mmol) was added carefully to
a solution of 1 (3.2 g, 12.8 mmol) in dry THF (100 cm3), and
the mixture was heated at reflux for 12 h. The resulting reaction
This journal is
The Royal Society of Chemistry 2010
Dalton Trans., 2010, 39, 10169–10178 | 10175
©

View Article Online
MeOD) 3.35 (2 H, t, J 6.9, NC(O)CH2CH2N), 2.97 (2 H, t, J 6.6,
C(O)NCH2CH2S), 2.91 (2 H, t, J 6.7, NCH2CH2S), 2.66 (2 H, t,
J 6.7, NCH2CH2S), 2.57 (2 H, t, J 6.9, NC(O)CH2CH2N), 2.46
(2 H, t, J 6.6, C(O)NCH2CH2S), 2.07 (3 H, s, SCH3), 2.06 (3 H, s,
SCH3); dC(62.5 MHz; MeOD) 172.5, 47.1, 45.0, 37.4, 35.6, 34.2,
33.7, 15.1, 14.8.
2.42 (2 H, t, J 5, N(CO)CH2CH2N), 2.11 (3 H, s, SCH3), 2.10
(3 H, s, SCH3), 1.82 (4 H, m, NCH2CH2CH2S); dC(62.5 MHz;
CDCl3) 172.26, 49.95, 49.65, 47.87, 45.27, 37.92, 34.80, 31.65,
31.36, 31.23, 28.52, 28.30, 15.27, 15.23.
N,N¢-bis(2-(Methylthio)propyl)propane-1,3-diamine, L7 (333-
Met). The same procedure as described for the preparation
of L5 was applied to 5 (1.2 g, 5.16 mmol). The product was
obtained as a colorless oil (0.75 g, 58%). Found: C, 39.8; H, 8.4;
N, 8.4. Calc. for C11H26N2S2·2HCl: C, 40.85; H, 8.7; N, 8.7%.
dH(300 MHz; CDCl3) 2.85 (2 H, br s, NH), 2.64 (8 H, t, J 6,
2 X NCH2CH2CH2S, NCH2CH2CH2N), 2.47 (4 H, t, J 6, 2 X
NCH2CH2CH2S), 2.03 (6 H, s, 2 X SCH3), 1.73 – 1.64 (6 H, m,
NCH2CH2CH2S, NCH2CH2CH2N); dC(75 MHz; CDCl3) 48.47,
48.25, 31.78, 29.58, 28.93, 15.23. ESI-MS: m/z 250.15 (100.0%):
found: (M+H)+ 250.91.
N,N-bis(2-(Methylthio)ethyl)propane-1,3-diamine, L5 (232-
Met). Borane-THF complex (1M, 20 mL) was added to a
solution of 3 (0.67 g, 2.86 mmol) in anhydrous THF (20 cm3).
The reaction ◦was refluxed overnight under N2. The reaction was
cooled to 0 C and 10 cm3 of concentrated HCl were added
dropwise. The THF was removed by distillation under atmospheric
pressure, and 20 cm3 of water were added to increase the solvent
volume. NaOH pellets were added to adjust the pH to >12,
and the resulting mixture was extracted with diethyl ether (3 ¥
5 cm3). The combined ethereal extracts were dried with anhydrous
Na2SO4 and filtered. The ether was evaporated under reduced
pressure to afford the product as a white solid (0.38 g, 60%).
dH(300 MHz; D2O) 3.23 (4 H, t, J 6.8, 2 X NCH2CH2S), 3.10 (4
H, m, NCH2CH2CH2N), 2.76 (4 H, t, J 6.8, 2 X NCH2CH2S), 2.06
(2 H, m, NCH2CH2CH2N), 2.04 (6 H, s, 2 X SCH3); dC(75 MHz;
D2O) 45.83, 44.27, 28.88, 22.45, 13.85. ESI-MS: m/z 222.96 (M+,
100%); found: (M+) 222.12.
General procedure for the preparation of 103Rh(III) complexes
[
103RhCl2(L)]PF6. A solution of 0.15 mmol of 103Rh(III)Cl3·xH2O
(Rh(III) 40% by weight) in 10 cm3 of 20% aqueous EtOH was added
to a stirring solution of 0.15 mmol of ligand in 40 cm3 of distilled
water. The pH of the reaction mixture was adjusted to ~5 by
dropwise addition of 2.5 M NaOH. The resulting deep red solution
was stirred and heated at 80 ◦C for 1 h. During this period of time,
the color of the reaction mixture turned to bright yellow. Any
insoluble material was removed by centrifugation, and the volume
of the solution was reduced by distillation under reduced pressure.
A solid product with a light yellow color was precipitated from
the solution with the addition of saturated aqueous NH4PF6. The
precipitate was isolated by centrifugation and washed several times
with water, EtOH and diethyl ether. In each case, the color of
the isolated product turned dark orange after drying under high
vacuum.
N-(3-(Methylthio)propyl)-2-(3-(methylthio)propylamino)aceta-
mide, (4). In a procedure identical with that used in the prepa-
ration of 3, 4 was prepared from 3-(methylthio)propylamine (3 g,
28.5 mmol) and 2.9 g (14.3 mmol) of bromoacetyl bromide. After
purification, the desired product was obtained as a colorless oil
(3.2 g, 89%). dH(250 MHz; CDCl3) 7.81 (1 H, br s, C(O)NH),
3.39 (1 H, t, J 6.6, C(O)NCH2CH2CH2S), 3.36 (1 H, t, J 6.6,
C(O)NCH2CH2CH2S), 3.31 (2 H, s, NCH2C(O)N), 3.17 (< 1
H, s, NH), 2.75 (2 H, t, J 7, NCH2CH2CH2S), 2.58 (2 H, t, J
7, NCH2CH2CH2S), 2.54 (2 H, t, J 7, NCH2CH2CH2S), 2.10 (6
H, s, 2 X SCH3), 1.83 (4 H, m, NCH2CH2CH2S); dC(62.5 MHz;
CDCl3) 172.93, 52.05, 48.50, 37.79, 31.63, 31.43, 28.67, 28.49,
15.32, 15.30.
cis/trans-[Dichloro
(N,N-bis-(2-(4-methoxybenzylthio)-2-
methylpropyl)ethane-1,2-diamine) rhodium(III)] hexafluorophos-
phate. cis/trans-[RhCl2(222-gdm-pmBz)]PF6, [RhCl2(L2)]PF6.
Yield: 30% (36 mg). A gradient mobile phase protocol (98% to
80% in 54 min, and from 80% to 20% in the last 6 min) was
utilized for analysis by HPLC and LC-MS. ESI-MS: m/z for
C26H40Cl2N2O2RhS2: 649.1 (100%), 650.1 (30) and 651 (70);
found: (M)+ 649 (100%), 650 (30) and 651 (70).
N,N¢-bis(3-(Methylthio)propyl)ethane-1,2-diamine, L6 (323-
Met). The same procedure as described for L5 was applied to
starting material 4 (3.2 g, 12.8 mmol). The product was obtained
as a colorless oil (2.3 g, 75%). Found: C, 38.95; H, 8.25; N, 9.0.
Calc. for C10H24N2S2·2HCl: C, 38.8; H, 8.5; N, 9.1%. dH(300 MHz;
CDCl3) 2.72 (4 H, s rising out of m, NCH2CH2N), 2.71 (4 H, t, J
7, 2 X NCH2CH2CH2S), 2.55 (4 H, t, J 7, 2 X NCH2CH2CH2S),
2.10 (6 H, s, 2 X SCH3), 1.78 (4 H, apparent p, J 7, 2 X
NCH2CH2CH2S), 1.56 (2H, s, 2 X NH); dC(75 MHz; CDCl3)
48.74, 48.34, 31.59, 28.88, 15.05. ESI-MS: m/z 236.14, 237.14;
found: (M+H)+ 236.9, 238.9.
cis/trans-[Dichloro
(N,N-bis-(2-(4-methoxybenzylthio)-2-
methylpropyl)propane-1,3-diamine) rhodium(III)] hexafluorophos-
phate. cis/trans-[RhCl2(232-gdm-pmBz)]PF6, [RhCl2(L3)]PF6.
Yield: 35% (43 mg, 0.052 mmol). Found: C, 40.4; H, 5.3; N, 3.5.
Calc. for C27H42Cl2F6N2O2PRhS2: C, 40.1; H, 5.2; N, 3.5%. A
gradient mobile phase protocol (98% to 80% in 5 min, 80% for
5 min then from 80% to 60% in the next 20 min and from 60%
to 20% in the last 5 min) was utilized for analysis by HPLC and
LC-MS. ESI-MS: m/z 663.1 (100%), 664.1 (30) and 665 (70);
found: (M)+ 663.1 (100%), 664 (30) and 665 (70). Single crystals
were prepared by slow evaporation of an ethanol–acetonitrile
solution, and analyzed by X-ray diffraction methods (vide infra).
N-(3-(Methylthio)propyl)-3-(3-(methylthio)propylamino)propan-
amide, 5. Compound 5 was prepared from 3-(methylthio)-
propylamine (4.69 g, 44.5 mmol) and 3-bromopropionyl chloride
(4.50 g, 22.3 mmol) in a procedure identical with that used in
the preparation of 4. The product was obtained as a clear oil
(1.20 g, 23%). dH(250 MHz; CDCl3) 7.81 (1 H, br s, N(CO)H),
3.34 (1 H, t, J 6, SCH2CH2CH2C(O)N), 3.32 (1 H, t, J 6,
SCH2CH2CH2C(O)N), 3.19 (1 H, br s, NH), 2.92 (2 H, t, J 5,
N(CO)CH2CH2N), 2.77 (2 H, t, J 6, NCH2CH2CH2S), 2.57 (2
H, t, J 6, NCH2CH2CH2S), 2.54 (2 H, t, J 6, NCH2CH2CH2S),
cis/trans-[Dichloro
propane-1,3-diamine) rhodium(III)] hexafluorophosphate. cis/trans-
[RhCl2(232-gdm-Met)]PF6, cis/trans-[RhCl2(L4)]PF6. Yield:
(N,N-bis-(2-methyl-2-methylthio)propyl)
42% (38 mg, 0.063 mmol). A gradient mobile phase protocol (98%
to 57% in 5 min, 57% for 5 min then from 57% to 52% in 30 min
10176 | Dalton Trans., 2010, 39, 10169–10178
This journal is
The Royal Society of Chemistry 2010
©

View Article Online
◦
3
˚
and from 52% to 10% in the last 2 min) was utilized for analysis by
HPLC and LC-MS. ESI-MS: m/z for C13H30Cl2N2RhS2: 451.03
(100%), 452.03 (16.7) and 453.03 (65.2); found: (M+) 451 (100%),
452 (17) and 453 (70).
90 , V = 1905.4(3) A , T = 173(2) K, Crystal system, Monoclinic,
-3
˚
space group P21/c, l source = 0.71073 A, Z = 4, Dc = 1.935 g cm ,
F(000) = 1112, dimensions 0.55 ¥ 0.50 ¥ 0.25 mm3, semi-empirical
absorption correction from equivalents, maximum and minimum
transmission factors of 0.70 and 0.53, respectively, data collection
range 1.68 to 27.11◦, -15 £ h £ 15, -12 £ k £ 12, -20 £ l £ 20,
cis/trans-[Dichloro (N,N-bis-(methylthio)ethyl) propane-1,3-
diamine) rhodium(III)] hexafluorophosphate. cis/trans-[RhCl2(232-
Met)]PF6, cis/trans-[RhCl2(L5)]PF6. Yield: 30% (24 mg,
0.045 mmol). A gradient mobile phase protocol (98% to 76% in
32 min and 76% to 20% in the last 3 min) was utilized for analysis
by LC-MS. ESI-MS: m/z for C9H22Cl2N2RhS2: 394.97 (100.0%),
396.96 (73.0); found: (M+) 394.75, 396.73.
12833 reflections were measured, and 4168 were unique (Rint
=
0.0169), R(F) = 0.0203, RW (F)2 = 0.0507, GoF = 1.094.
Radiolabeling reactions with 105Rh. Caution! 105Rh is a moder-
ate energy beta emitter with a low abundance gamma emission.
Proper radiation safety procedures and adequate shielding were
used when handling this radionuclide in laboratories approved for
radioisotope use.
cis/trans-[Dichloro (N,N-bis-(3-methylthio)propyl) ethane-1,2-
diamine) rhodium(III)] hexafluorophosphate. cis/trans-[RhCl2(323-
Met)]PF6, cis/trans-[RhCl2(L6)]PF6. Yield: 44% (36 mg,
0.66 mmol). mp 295 ◦C (dec). Found: C, 21.9; H, 4.3; N 5.2. Calc.
for C10H24Cl2N2RhS2·PF6: C, 21.6; H, 4.4; N, 5.05. ESI-MS: m/z
for C10H24Cl2N2RhS2 408.98 (100%), 410.98 (73), 412.98 (10.5);
found (M+) 408.82, 410.8, 412.81. Crystals suitable for an X-ray
diffraction study were grown by diffusion of methanol into an
acetonitrile solution of the complex (vide infra).
105Rh production. 105Rh was prepared at the University of
Missouri Research Reactor Center (MURR) by bombardment
of an enriched 104Ru metal target encapsulated in T21 Quartz
for 155 h at a thermal neutron flux of 3 ¢ 1014 n/cm2 s-1. The
Ru metal was oxidized to RuO4 and separated from 105Rh by
the method previously described.19 This involved breaking the
quartz vial and transferring the irradiated Ru metal into an
impinger containing NaOH (3 mL, 2 M) through which Cl2 gas
(50 s @ 30 cm3 min-1) had been bubbled to generate NaOCl
in situ. The reaction mixture was heated for 1 h at 40 ◦C with
air flowing through the impinger followed by the bubbling of
additional Cl2 gas (12 min @ 30 cm3 min-1), followed by heating
at 90 ◦C for 1 h with airflow, during which time the RuO4 was
distilled off and trapped in HCl (3 M, 400 cm3). The reaction
mixture was filtered to remove any undissolved Ru and heated for
30 min at 90 ◦C, and then treated with HCl (1 M, 0.25 cm3) and
heated at 90 ◦C without airflow for 30 min. Approximately 20–
25 mCi (740–925 MBq) of 105Rh(III)-chloride in an HCl solution
were typically obtained at the end of processing. The 105Rh(III)-
chloride reagent is considered to be a mixture of primarily
cis/trans-[Dichloro (N,N-bis-(3-methylthio)propyl) propane-1,3-
diamine) rhodium(III)] hexafluorophosphate. cis/trans-[RhCl2(333-
Met)]PF6, cis/trans-[RhCl2(L7)]PF6. Yield: 40%. ESI-MS: m/z
for C11H26Cl2N2RhS2 423.00 (100.0%), 424.99 (73.0); found: (M+)
422.87, 424.88.
X-ray structure determinations and refinements
Intensity data were obtained at minus 100 ◦C on a Bruker SMART
CCD Area Detector system using the w scan technique with
Mo-Ka radiation from a graphite monochromator. Intensities
were corrected for Lorentz and polarization effects. Equivalent
reflections were merged, and absorption corrections16 were made
using the multi-scan method. Structures were solved by direct
methods with full-matrix least-squares refinement, using the
SHELX package.17,18 All non-hydrogen atoms were refined with
anisotropic thermal parameters. The hydrogen atoms were placed
at calculated positions and included in the refinements using a
riding model, with fixed isotropic U, except those except those
on the carbons of a disordered ethanol of crystallization in the
structure of [RhCl2(L3)]PF6, which were omitted from the final
model. Final difference maps contained no features of chemical
significance.
three or four different chloro-aquo species (i.e., 105RhCl3(OH2)3,
1-
105RhCl4(OH2)2
,
105RhCl5(OH2)2-, and 105RhCl63-),1 although
electrophoresis analysis generally shows the presence of only the
three anionic species.20
105Rh radiolabeling to form [105RhCl2(L)]Cl complexes. An
aliquot of stock ligand solution (1 ¥ 10-2 M, 0.100 cm3) was
added to a vial containing 1.6 mg of gentisic acid (GA), and
the mixture was vortexed. After addition of 0.100 cm3 of 105Rh
(~ 3.2 mCi), the pH of the solution was adjusted to ~ 4.5
with the addition of 2.5 M and 0.1 M NaOH consecutively.
After the pH adjustment, 0.100 cm3 of EtOH was added, and
the volume of the reaction mixture was brought to 1 cm3 by
addition of water. Reaction mixtures were heated at 80 ◦C for
1h. After radiolabeling, the resulting solution was analyzed by
radio-HPLC, and fractions were collected as a function of peaks
observed by the radiodetector. The radioactivity of the collected
fractions was compared to the amount of radioactivity injected
to confirm that, in all cases, none of the activity was retained
on the column. The [105RhCl2(L)]Cl peak was purified by radio-
HPLC and recovered from a reverse-phase SepPak light cartridge.
The SepPak cartridge was first activated with 4 cm3 of EtOH and
washed with 2 ¥ 4 cm3 of water. In the cases of [105RhCl2(L2)]Cl,
(trans-[RhCl2(232-gdm-pmBz)]PF6, trans-[RhCl2(L3)] PF6).
˚
◦
C28H45N2O2.50S2Cl2Rh·PF6, M = 832.56, a = 17.8256(11) A, b =
◦
˚
˚
10◦.4632(6) A, c = 18.7671(11) A, a = 90 , b = 94.3150(10) , g =
3
˚
90 , V = 3490.4(4) A , T = 173(2) K, Crystal system, Monoclinic,
-3
˚
space group P21/c, l source = 0.71073 A, Z = 4, Dc = 1.584 g cm ,
F(000) = 1708, dimensions 0.50 ¥ 0.30 ¥ 0.05 mm3, semi-empirical
absorption correction from equivalents, maximum and minimum
transmission factors of 0.96 and 0.64, respectively, data collection
range 2.18 to 27.14◦, -22 £ h £ 22, -10 £ k £ 13, -24 £ l £ 24,
24240 reflections were measured, and 7684 were unique (Rint
=
0.0271), R(F) = 0.0325, RW (F)2 = 0.0819, GoF = 1.030.
(trans-[RhCl2(323-Met)]PF6,
trans-[RhCl2(L6)]
PF6).
˚
◦
C10H24N2S2Cl2Rh·PF6, M = 555.21, a = 12.4677(12) A, b =
[
105RhCl2(L3)Cl and [105RhCl2(L4)]Cl, the collected HPLC eluent
◦
˚
˚
9.5972(9) A, c = 16.3961(15) A, a = 90 , b = 103.783(2) , g =
(~2–3 cm3) was diluted with water to attain a 1 : 9 CH3CN–H2O
This journal is
The Royal Society of Chemistry 2010
Dalton Trans., 2010, 39, 10169–10178 | 10177
©

View Article Online
ratio (total 10 cm3), and then was passed through the SepPak.
Acknowledgements
The column was then washed with 2 ¥ 4 cm3 of water and the
105RhCl2(L)]Cl was eluted with 0.02–0.04 cm3 mL of EtOH. In the
We are grateful for a University of Missouri Research Reactor
(MURR) Research Fellowship Award (MURFA) to Z.A. We also
acknowledge grant awards from the National Science Foundation
(US) and the National Institutes of Health (US) (NSF CHE-89-
08304, CHE-95-31247 CHE-92-2183526 and NIH 1S10RR11962-
01) which supported the purchase of these spectrometers.
[
case of [105RhCl2(L5)]Cl, due to its high solubility in water, instead
of Sep-Pak purification, evaporation of the solvent under argon
gas was performed.
Stability studies for [105RhCl2(L)]Cl (L = 3, 4, 5). Purified
105Rh complexes were first prepared according to the optimized
reaction conditions. To the vials containing 0.02 cm3 (0.5 mCi) of
purified [105RhCl2(L3)]Cl (cis and trans), [105RhCl2(L4)Cl (trans)
and [105RhCl2(L5)]Cl (cis and trans), respectively, 0.500 cm3 of
phosphate-buffered saline (PBS) and 10 mg of GA were added.
References
1 S. S. Jurisson, A. R. Ketring and W. A. Volkert, Transition Met. Chem.,
1997, 22, 315–317.
2 B. Bosnich, R. D. Gillard, E. D. McKenzie and G. A. Webb, J. Chem.
Soc. A, 1966, 1331–1339.
3 J. P. Collman and P. W. Schneider, Inorg. Chem., 1966, 5, 1380–
1384.
4 P. K. Bhattacharya, J. Chem. Soc., Dalton Trans., 1980, 810–812.
5 N. F. Curtis and D. F. Cook, J. Chem. Soc., Dalton Trans., 1972, 691–
697.
6 A. J. Blake, G. Reid and M. Schro¨der, J. Chem. Soc., Dalton Trans.,
1989, 1675–1680.
7 A. J. Blake, G. Reid and M. Schro¨der, Polyhedron, 1992, 11, 2501–
2506.
Distribution coefficient (log D). The log D values were
determined by the classic shake-flask method.21,22 100 mCi
(0.001–0.005 cm3) of the purified radiolabeled complexes
([105RhCl2(L3)]Cl, [105RhCl2(L4)Cl and [105RhCl2(L5)]Cl) were
added to a pre-mixed suspension of 3.5 cm3 of buffer-equilibrated
octanol and 3.5 cm3 of octanol-equilibrated PBS buffer at pH 7.4.
The resultant solution was vortexed for two min and then
centrifuged for 2 min at 2000 rpm. A 0.10 cm3 aliquot of the
octanol layer and a 1.0 cm3 aliquot from the aqueous layer were
counted in a NaI(Tl) scintillation well counter. The log D was
calculated according to the following equation:
8 N. Goswami, R. Alberto, C. L. Barnes and S. Jurisson, Inorg. Chem.,
1996, 35, 7546–7555.
9 N. Goswami, C. Higginbotham, W. Volkert, R. Alberto, W. Nef and S.
Jurisson, Nucl. Med. Biol., 1999, 26, 951–957.
10 N. Li, C. M. Eberlein, W. A. Volkert, L. Ochrymowycz, C. Barnes and
A. R. Ketring, Radiochim. Acta, 1996, 75, 83–95.
⎡
⎢
⎤
⎥
(10×averageof organiccounts-BKG )
LogD = log
⎢
⎣
⎥
⎦
11 Preliminary reports of a portion of these results were presented:
Z. Akgun, H. Engelbrecht, F. Gallazzi, C. Cutler, C. L. Barnes,
S. S. Jurisson, and S. Z. Lever, ABSTRACTS OF PAPERS OF THE
AMERICAN CHEMICAL SOCIETY 231:135-NUCL March, 2006.
12 Z. Akgun, H. Engelbrecht, C. S. Cutler, C. L. Barnes, S. S. Jurisson and
S. Z. Lever, J. Labelled Compd. Radiopharm., 2005, 48, S299.
13 T. Konno, Bull. Chem. Soc. Jpn., 2004, 77, 627–649.
14 K. E. Baidoo, Technetium-99m Labeling of Proteins. Ph.D. Thesis, The
Johns Hopkins University, Baltimore, 1988.
[average buffer count-BKG]
Then 3.0 cm3 of the remaining octanol phase was transferred
to another tube containing 0.5 cm3 of octanol and 3.5 cm3 of
PBS. The last three steps were repeated until multiple, reproducible
measurements of constant log D were obtained.
15 J. J. D’Amico and W. E. Dahl, J. Org. Chem., 1975, 40, 1224–1227.
16 S. Fox, R. T. Stibrany, J. A. Potenza, S. Knapp and H. J. Schugar, Inorg.
Chem., 2000, 39, 4950–4961.
17 G. M. Sheldrick, SADABS v2.05, Bruker AXS Inc., Madison, WI,
USA, 2001.
Conclusion
Of the seven ligands investigated for the coordination of
rhodium(III), one emerged as a preferred ligand. The ligand with
the 232 chain length, gem-dimethyl groups and methyl thioether
(L4) forms trans rhodium complexes with highest radiochemical
yields and in vitro stability. One caveat from the results obtained
is that the temperature used for radiolabeling (80 ◦C) may not be
compatible with all biological constructs. If a construct were stable
to the radiolabeling conditions, this ligand could serve as a suitable
carrier for 105Rh in biological targets designed for radiotherapeutic
studies.
18 G. M. Sheldrick, Acta Crystallogr., Sect. A: Found. Crystallogr., 2008,
64, 112–122.
19 B. Grazman and D. E. Troutner, Int. J. Radiat. Appl. Instrum., Part A,
1988, 39, 257–260.
20 F. A. Cotton and G. Wilkinson, in Advanced Inorganic Chemistry, John
Wiley and Sons Publishing Co. Inc., New York, 4th edn, 1980, 944.
21 R. B. Del Rosario, Y.-W. Jung, K. E. Baidoo, S. Z. Lever and D. M.
Wieland, Nucl. Med. Biol., 1994, 21, 197–203.
22 S. Z. Lever, K. E. Baidoo, A. Mahmood, K. Matsumura, U. Scheffel
and H. N. Wagner, Jr., Nucl. Med. Biol., 1994, 21, 157–64.
10178 | Dalton Trans., 2010, 39, 10169–10178
This journal is
The Royal Society of Chemistry 2010
©