Modes of xanthine complexation to dirhodium tetrakis[(R)-α-methoxy-α-(trifluoromethyl)-phenylacetate] in solution and in the solid state

Article abstract of DOI:10.1515/znb-2001-0317

It is shown by IR and NMR studies that the xanthines 1-5 prefer a side-on complexation to the chiral dirhodium tetrakis[(R)-α-methoxy-a-(trifluoromethyl)phenylacetate] (Rh*) in solution whereas carbonyl groups are involved in the solid state. For 6, at le

Full text of DOI:10.1515/znb-2001-0317

Modes of Xanthine Complexation to Dirhodium Tetrakis[(/?)-«-methoxy-«-
(trifluoromethyl)-phenylacetate] in Solution and in the Solid State
Sven Rockitta, Rudolf W artchow3, Helmut Duddeck3*, Anna Drabczynskab,
and Katarzyna Kiec-Kononowiczb
a Universität H annover, Fachbereich für Chemie, Schneiderberg IB, D-30167 Hannover,
Germany
b Departm ent of Chem ical Technology of Drugs, Collegium M edicum, Jagiellonian
University, M edyczna 9, PL 3 0 -6 8 8 Krakow, Poland
Reprints requests to Professor H elm ut Duddeck. Fax: +49 (0)511/762 4616.
E-mail: m ailto:duddeck@ mbox.oci.uni-hannover.de
Z. Naturforsch. 56b, 3 1 9 -3 2 4 (2001); received Decem ber 11, 2000
Rhodium Com plexes, Xanthines
It is shown by IR and N M R studies that the xanthines 1-5 prefer a side-on com plexation
to the chiral dirhodium tetrakis[(/?)-a-m ethoxy-a-(trifluorom ethyl)phenylacetatel (Rh*) in
solution whereas carbonyl groups are involved in the solid state. For 6, at least the carbonyl
group C-6 contributes to com plexation in solution as well. Alternating strands of 6 and Rh*
exist in the solid state as revealed by X-ray diffraction analysis described in detail. The
determ ination of enantiom eric excess of the chiral xanthine 6 can easily be accom plished by
the “dirhodium m ethod” (*H and 13C NM R in the presence of Rh*).
Introduction
solid state. In order to shed more light on this
seemingly contradictory results we extended our
NM R investigation to compound 6 (Scheme 1; it
is compound 7 in ref. [2]) and investigated the IR
spectra of all xanthines 1 -6 in the absence and
presence of Rh* and in the two aggregation states
(CDC13 solution and solid state).
For a couple of years we have been interested in
the potential of the chiral dirhodium complex Rh*
(dirhodium tetrakis[(i?)-a-m ethoxy-a-(trifluoro-
m ethyl)phenylacetate]; R h2(MTPA)4, MTPA
=
M osher acid) as an NM R auxiliary for the deter-
m ination of enantiom eric purity [1], especially in
cases where the investigated substrate molecules
contain only functional groups which do not re-
spond properly to chiral lanthanide shift reagents.
Very recently, we reported on the extension of this
Results and Discussion
Investigating the xanthines 1 -5 (as racemates
or non-racemic mixtures) by NM R in the absence
method to polyfunctional xanthines 1 -5 [2] and presence of enantiomerically pure (R )-Rh* we
(Scheme 1) where we noticed an intriguing differ-
ence in the binding modes in solution and in the
came to the conclusion that side-on complexation
of the imidazole unit (Scheme 2) should prevail
CH,R
1: R = CH3
2: R = CH2C1
3: R = CH,CH,
H,C
Schem e 1. Structures of Rh* and of
xanthine derivatives 1
-
6.
0932-0776/2001/0300-0319 $06.00 © 2001 Verlag der Zeitschrift für Naturforschung, Tübingen • www.znaturforsch.com
D
Brought to you by | University of St Andrews Scotland
Authenticated
Download Date | 6/15/15 7:16 PM

320
S. Rockitt et al. ■Modes of Xanthine Complexation
chemical shifts of 6 are collected in Table I. The
assignment methods (using ID - and 2D-correla-
tion experiments) have been described before [2]
as well as the definition of the complexation shifts
A d (in ppm) as the signal displacements due to the
addition of an equimolar am ount of Rh* and of
the dispersion effects A v (in Hz) as signal split-
tings due to the existence of diastereom eric
Rh* •••xanthine complexes.
Scheme 2. Side-on complexation of 1 to Rh* (only two
acylate residues depicted) [2].
An inspection of the NM R data showed that the
complexation mode of 6 indeed differs from that
because a clear side differentiation of diastereo-
topic methylene protons (seen from dispersion ef- of 1-5 to some extent. W hereas the latter showed
no significant C = 0 complexation in solution (Ad-
fects A v) is apparent [2]. M oreover, no *H and 13C
signal reveals a clear complexation (seen from
values of <1 ppm for C-2 and 0 -2 ppm for C-6 [2])
paramagnetic complexation shifts A d) for any par- there is a moderate deshielding of C-6 in 6 (Aö -
ticular atom. 2.41). On the other hand, clear side differentiation
In contrast, an X-ray investigation showed al- can be identified as well; very different dispersions
ternating strands of Rh* and of 6 molecules in the appear within the pairs of diastereom eric protons
crystal where each Rh* molecule complexes car- at C -l' and at C-3' (Table I). These experimental
bonyl oxygens of the same position in two dif- evidences suggest the following interpretation: in
contrast to 1-5, a solution equilibrium exists for 6
in which the C-6 carbonyl group competes with
side-on complexation to some extent. A C-2 car-
bonyl complexation cannot be identified safely.
It should be noted that the enantiom eric excess
of 6 can easily be monitored from the dispersion
of the *H NM R signals of H-2", H-3" and H-4"
(2.9, 2.2 and 6.3 Hz, respectively) caused by the
existence of diastereomeric complexes in the pres-
ence of Rh*, i.e., our “dirhodium m ethod” for chi-
ral recognition is again successful.
ferent xanthine ligands; i.e., one binds to the C-2
carbonyls on both rhodium atoms whereas the
next binds to two C-6 carbonyl groups (see
Scheme 3). Further details of the X-ray study are
given in the Experim ental Part.
In order to get a closer insight into the various
complexation modes in solution and in the solid
state we decided to inspect the IR bands of all
xanthines 1 -6 in the absence and presence of (R )-
Rh* and in both aggregation states (Table II).
The assignment of the pertinent IR bands has
been performed on the basis of IR data reported
in the literature for a series of structurally very
similar xanthines [3]. Thus, the carbonyl band with
the larger wavenumbers (v = 1693-1706 cm -1)
corresponds to C-2 whereas the other one (v =
1649-1658 cm-1) can be assigned to C-6. It should
be noted that a further band appears in close vicin-
Scheme 3. Section of a strand of Rh* and 6 as deter-
mined by X-ray diffraction; for better visibility all
Mosher acid residues are represented by the C-a atom
only. The structure of 6 is inserted for better comparabil-
ity.
In order to find out whether this is indeed a ity at v = 1600-1610 cm-1 which originates from
contradiction we decided to extend the NM R
spectral investigation to 6 using a non-racemic
mixture (S/R = 2:1, prepared by mixing appropri-
ate amounts of the pure enantiom ers) and to com-
pare the data with those of 1-5. The !H and 13C
C=C and/or C=N vibrations [3].
As can been seen from Table II, both carbonyl
bands in the pure substrates are hardly affected in
their wavenumbers by the transition from the li-
quid to the solid phase (compare Fig. 1 and Ta-
Brought to you by | University of St Andrews Scotland
Authenticated
Download Date | 6/15/15 7:16 PM

S. Rockitt et al. • Modes of Xanthine Complexation
321
Table I. *H and 13C chemical shifts d (in ppm), complexation shifts Ad (in ppm), and dispersions (in Hz) of 6. For
experimental details and the methods of signal assignments see ref. [2],
<5 [ppm]
in presence of Rh*
Ad [ppm]
in the presence of Rh*
<5 [ppm]
pure sample
Av [Hz]
5
R
S
R
H-l'qe
H-l'qa
H-2'
4.294
4.167
4.238 ddd
4.136 ddd
ca 2.04 m
ca 1.98 m
2.961 ddd
3.214 ddd
3.374 s
3.537 s
1.614 d
5.869 q
7.355 dm
7.355 dm
7.295 m
151.99
4.273
4.167
+0.06
+0.03
+0.03
+0.03
10.6
0-1
ca 1.80a
ca -0.2a
-0.15
n.d.b
H-3'qe
H-3'qa
H-l"
H-2"
H-3"
H-4"
H-6"
H-7"
H-8"
C-2
-0.15
n.d.b
+0.17
+0.05
-0.03
0
17.7
0-1 or 8-9c
3.7
2.849
3.067
3.551
3.582
1.589
5.859
2.813
-0.11
overlap
3.544
3.587
1.584
5.873
+0.18
+0.04
-0.03
+0.02
2.9
2.2
6.9
ca 7.34
ca 7.34
n.d.b
-0.02
-0.02
n.d.b
n.d.b
n.d.b
n.d.b
3.4
152.26
150.42
152.23
150.45
+0.27
+1.36
+0.24
+1.39
C-4
149.06
3.1
0
C-5
102.99
104.63
156.29 brd
+1.64
+2.41
C-6
153.88
n.d.b
1.9
C-8
151.79
152.50
41.83
20.94
38.29
152.52
41.82
20.90
38.22
+0.71
+0.06
-0.42
-0.05
+0.73
+0.05
-0.46
-0.12
C-l'
41.77
1.7
C-2'
C-3'
C-l"
C-2"
C-3"
C-4"
C-5"
C-6"
C-7"
C-8"
21.36
4.6
38.34
8.9
27.58
28.53
+0.95
0
29.72
30.05
15.78
54.15
139.97
127.14
128.58
30.06
15.65
54.10
139.94
127.16
128.60
+0.33
+0.09
+0.11
-0.06
-0.04
-0.01
+0.34
-0.04
+0.06
-0.09
-0.02
+0.01
0-1
17.0
5.5
15.69
54.04
4.1
140.03
127.18
2.2
128.59
1.7
127.66
127.64
-0.02
0-1
a Not resolvable; b n.d.: not detectable due to signal complexity; c due to signal overlap it cannot be decided safely
whether Av is either 0-1 or 8-9 Hz; d broadened signal.
ble II, column III) so that any significant effect in
the presence of Rh* has to be attributed to com-
plexation. However, the corresponding data of the
xanthine-Rh* complexes (Table II, column III) are
significantly higher, particularly for the C-6 car-
bonyl indicating a difference of the binding modes
in the two aggregation states. Conversely, if band
shifts originated by complexation are calculated
(Table II, columns IV and V) it turns out that, in
the case of the substrates 1 -5 , the complexation
shifts in solution (Table II, column IV) are close to
zero so that we have to assume that no significant
complexation at any of the two C = 0 occurs here.
This is different for 6 (Fig. 2, top) where the band
shifts are significantly larger ( - 4 for C-2 and -11
cm -1 for C-6). Here, some C = 0 complexation -
particularly at C-6 - exists, a fact which has al-
ready been noticed from the NM R spectral eval-
Wavenumbers v (cm'1)
Fig. 1. Sections of the IR spectra of 1, (a) in CDC13solu-
tion, (b) in the solid state.
Brought to you by | University of St Andrews Scotland
Authenticated
Download Date | 6/15/15 7:16 PM

322
S. Rockitt et al. • Modes of Xanthine Complexation
Table II. IR wavenumbers v (section between 1550 and 1750 cm !) of 1-6 in the absence and presence of Rh*,
each in CDC13 solution and in the solid state.
Av:
Solution
Solid state
Solution —*
solid state
(Column III)
Soluation
Solid !
(Colui
(Column IV)
(Column I)
(Column II)
1704 (m)
1697 (m)
1652 (s)
1638 (w)
1703 (m)
1698 (m)
1650 (s)
1638 (w)
1704 (m)
1701 (m)
1651 (s)
1640 (s)
1699 (s)
1702 (m)
1650 (s)
1633 (w)
1707 (m)
1698 (m)
1650 (s)
1647 (w)
1695 (m)
1671 (w)
1644 (s)
1635 (w)
1
2
3
4
5
6
C-2
C-2
C-6
C-6
C-2
C-2
C-6
C-6
C-2
C-2
C-6
C-6
C-2
C-2
C-6
C-6
C-2
C-2
C-6
C-6
C-2
C-2
C-6
C-6
pure
1704 (m)
1704 (m)
1657 (s)
1659 (s)
1706 (m)
1706 (m)
1658 (s)
1660 (s)
1704 (m)
1704 (m)
1656 (s)
1658 (s)
1703 (m)
1703 (m)
1658 (s)
1657 (m)
1701 (m)
1700 (m)
1656 (s)
1655 (m)
1693 (m)
1689 (m)
1649 (s)
1638 (m)
0
-7
+Rh*
pure
0
+2
0
-7
-14
-5
-5
+Rh*
pure
-21
-3
-8
-8
-22
0
-3
-5
-18
-4
-1
-8
-24
+6
-2
+Rh*
pure
+Rh*
pure
+2
0
-12
-3
+Rh*
pure
+Rh*
pure
+2
0
-11
+3
+Rh*
pure
+Rh*
pure
-1
-1
-1
-4
-11
-17
-9
+Rh*
pure
-6
-8
+Rh*
pure
-3
+2
-18
-5
+Rh*
pure
-24
-11
+Rh*
-3
a Bruker DRX-500 spectrom eter at 500.1 MHz
(!H) and 125.8 MHz (fH J-B B decoupled 13C) at
uation (see above). Finally, an inspection of the
complexation shifts in the solid state (Table II, col-
umn V) shows large and significant values for all
compounds (compare Fig. 2, bottom ) which is in
agreem ent with the crystal structure result of 6 .
We expect that the same or a very similar com-
plexation mode exists for the other xanthines as
well; however, it was not possible to obtain crystals
of sufficient quality for X-ray diffraction. Proba-
ambient tem perature. In
a typical experiment
40 mg Rh* (3.5-10-2 mmol) were dissolved in
0.5 ml CDC13 containing 17.2 mg of acetone-d6 for
better solubility (7.7 m olar relative to Rh*). All
chemical shifts are referenced to internal tetra-
methylsilane ((3 = 0). Standard Bruker software
was used for all one- and two-dimensional experi-
ments. E l mass spectra were obtained on a Finni-
bly, this is due to the fact that these compounds gan MAT 312 (70 eV) with direct inlet. IR spectra
were taken on a Bruker Vector 22 (attenuated to-
tal reflection mode, ATR) in the solid state and in
CDCI3 solutions identical to those prepared for
the NM R studies [2].
were not available in enantiomerically pure form.
In conclusion, the xanthines 1 -5 strongly prefer
the side-on binding to the dirhodium complex Rh*
in the liquid state (CDC13 solution) whereas C-6
carbonyl complexation competes in the case of 6
to a certain extent. The IR data in the solid state
indicate that here C = 0 complexation is predomi-
nant as shown by crystal structure determination
of Rh*-6.
Crystal structure analysis o f Rh*-6
Cii7H107Cl3F24N10O28Rh4, M = 3075.12 g/mol,
green
crystal
of
irregular
shape,
size
0.26 x 0.17 x 0.07 mm, monoclinic, space group
P21? (No. 4), a = 19.517(2), b = 14.080(1), c =
23.918(2) A, a = 90.00°, ß = 106.04(1)°, y = 90.00°,
V = 6316.8(9) Ä 3, Z = 2, D x = 1.617 g/cm3, T =
300(2) K, Stoe IPDS diffractometer, A (M o-K a) =
Experimental
The NM R spectra of compounds 1 to 5 were
published before [2], those of 6 were recorded on
Brought to you by | University of St Andrews Scotland
Authenticated
Download Date | 6/15/15 7:16 PM

S. Rockitt et al. • Modes of Xanthine Complexation
323
Scheme 4. One Rh* complex (see Scheme 3 and text)
without ligands as determined by X-ray diffraction.
in Scheme 4 without its axial ligands. It is easily
seen that the conformation of the ligands is such
that the complex has nearly the symmetry of point
group 4 (C4). The second Rh* complex (center of
gravity approximately in position 0.75, 0.18, 0.25)
is similar, but the torsion angle of one phenyl
group is quite different. The two symmetrically in-
dependent molecules of compound 6 are almost
identical in their conformation. Axial Rh-O bond
lengths are 2.235(10) to 2.307(12) Ä. These dis-
tances are significantly larger than Rh-O bond
lengths for acylate residues: 1.990(9) to 2.093(11)
Ä. Rh-Rh o bond lengths are 2.3804(14) and
2.3772(15) A. It is interesting to note that the dir-
hodium tetraacylate skeletons adopt chiral confor-
mations (Scheme 4); O -Rh-Rh-O torsion angles
for the acylates are 0 to 2.6° for one and 3.1 to
4.6° for the other Rh* entity.
Wavenumbers v (cm'1)
Fig. 2. Sections of the IR spectra of 6; top: in CDC13
solution, (a) pure 6 and (b) in the presence of Rh*; bot-
tom: in the solid state, (a) pure 6 and (b) in the presence
of Rh*.
0.71073 A, <9max = 24.15°, 34069 measured, 19523
unique (/?int = 0.0636) and 7851 observed (/ >2 oj)
reflections, 604 refined parameters, Rgt(F) = 0.068,
wi?(F2) = 0.121. Full crystallographic details with-
out structure factors have been deposited at
CCDC, no. 154420.
There are two Rh* complexes, two molecules of
compound 6, and one disordered solvent moleule
(CHC13) in the asymmetric unit. Rh* and 6 are
bonded in an alternating sequence forming an infi-
nite chain in the [100] direction (see Scheme 3). A
second chain in the same direction is related to the
first one by the crystallographic twofold screw
Syntheses
l,3-Dim ethyl-2,4-dioxo-9- (1-phenylethyl) -
l,3,6,7,8,9-hexahydropyrim ido[2,l-f]purine (6)
Compound 6 has been described previously [4]
but its structure was confirmed only by elemental
analysis and UV spectra. Different starting m ateri-
als, modified reaction conditions and isolation
methods provided 6 in purer form. The starting
material, 7-(3-chloropropyl)-8-bromotheophylline
axis. The first Rh* complex (center of gravity ap- [5], was obtained by a modified two-phase chloro-
proximately in position 0.25, 0.34, 0.25) is shown alkylation in acetone in the presence of anhydrous
Brought to you by | University of St Andrews Scotland
Authenticated
Download Date | 6/15/15 7:16 PM

324
S. Rockitt et al. • Modes of Xanthine Complexation
K2C 0 3 and benzyltriethylammonium chloride
(TEBA) as catalyst: a mixture of 8-bromotheo-
rated after cooling and recrystallized from 70%
ethanol. Yield 95%; mp. 142-3 °C (lit. [4]:
phylline (5.16 g, 0.02 mol) [6], anhydr. K2C 0 3 137 °C); TLC: Kieselgel 60F254, Rf - 0.56 (ben-
(2.8 g, 0.02 mol) TEBA (0.30 g), l-bromo-3-chlo-
ropropane (4 ml, 0.04 mol) in of acetone (40 ml)
was heated at reflux for 10 h with stirring. The pre-
cipitate was filtered off from the hot mixture,
mixed with 40 ml of 15% N aO H (to remove unre-
acted 8-bromotheophylline), washed with water
(to remove inorganic salts) and recrystallized from
ethanol. The acetone filtrate was cooled and the
main crop of 7-(3-chloropropyl)-8-bromotheo-
phylline was separated and recrystallized from
ethanol. Total yield 86%; m.p. 132-133 °C (lit. [5]:
131 °C). The substance was used to synthesize
compound 6.
zene - acetone, 7:3). - IR (solid state): v = 3541,
3476, 2945, 1695, 1644, 1616, 1568, 1535, 1478,
1453, 1431, 1400, 1372, 1225, 1204, 1177, 1070,
1037, 977, 914, 884, 765. - MS (El, 70 eV): m /
Z (% ) = 340 (15) [M +H]+, 339 (36) [M+], 236 (15)
[M+-C8H 7], 235 (100) [M+-C8H 8], 234 (12) [M+-
C8H 9], 207 (8) [235-CO]+, 191 (13), 178 (7), 159
(14), 149 (8), 133 (9) [C9H n N+], 105 (48)
[C8H 9+], - C18H 21N50 2 (339.40): calcd. C 63.70,
H 6.24, N 20.64; found C 63.84, H 6.57, N 21.08.
The preparation of Rh* has been reported be-
fore [la]; for the xanthines 1 -5 see ref. [2],
A mixture of 7-(3-chloropropyl)-8-bromotheo-
phylline (3.3 g, 0.01 mol), racemic 1-phenylethy-
lamine (4 ml, 0.03 mol) and butanol (6 ml) was
heated at reflux for 10 h. Then, butanol was re-
moved by distillation under reduced pressure and
the excess of amine by steam distillation. Com-
pound 6 was precipitated in water solution, sepa-
Acknowledgem ents
This work was supported by the Deutsche
Forschungsgemeinschaft (Du 98/14), the Fonds der
Chemischen Industrie and by the Polish State
Committee for Scientific Research (KBN) grant
No 4 P05F 014 18.
[3] A. Rybär, K. Antoa, Coll. Czech. Chem. Commun.
35, 1415 (1970).
[1] a) K. Wypchlo, H. Duddeck, Tetrahedron: Asymm.
27, 5 (1994); b) K. Wypchlo, H. Duddeck, Chirality
9, 601 (1997); c) S. Hameed, R. Ahmad, H. Dud-
deck, Magn. Reson. Chem. 36, S47 (1998); d) S.
Hameed, R. Ahmad, H. Duddeck, Heteroatom
Chem. 9, 471 (1998); e) C. Meyer, Ph. D. Thesis,
Universität Hannover (1999); f) C. Meyer, H. Dud-
deck, Magn. Reson. Chem. 38, 29 (2000).
[4] M. Eckstein, W. tosori; Dissertationes Pharmaceuti-
cae et Pharmacologicae 20, 35 (1968).
[5] V. B. Kalcheva, T. M. Apostolova, V. Z. Anakieva,
J. Prakt. Chem. 327, 165 (1985).
[6] M. Eckstein, M. Gorczyca, A. Zejc, Acta Pharm.
Jugosl. 22, 133 (1972).
[2] S. Rockitt, H. Duddeck, A. Drabczynska, K. Kiec-
Kononowicz, Eur. J. Org. Chem. 2000, 3489.
Brought to you by | University of St Andrews Scotland
Authenticated
Download Date | 6/15/15 7:16 PM