Chiral ferrocene amines derived from amino acids and peptides: Synthesis, solution and X-ray crystal structures and electrochemical investigations

Article abstract of DOI:10.1021/ic000089+

For the recognition of all but the simplest naturally occurring molecules, electrochemical sensors based on ferrocene will certainly require chiral centers. To advance the necessary chemistry, this work describes the synthesis and properties of ferrocene derivatives of enantiomerically pure amino acids, peptides, and other chiral amines. Ferrocene aldehyde is condensed with amino acid esters to yield the corresponding Schiff bases 2, which are reduced by NaBH₄ in methanol to the ferrocene methyl amino acids 3. An X-ray single-crystal analysis was carried out on the phenylalanine derivative 3a (monoclinic space group P2₁, a = 10.301(1) A, b = 9.647(1) A, c = 18.479(2) A, β = 102.98(2)°, Z = 4). Further peptide chemistry at the C terminus proceeds smoothly as demonstrated by the synthesis of the ferrocene labeled dipeptide Fc-CH₂-Phe-Gly-OCH₃ 5 (Fc = ferrocenyl ((η-C₅H₄)Fe(η-C₅H₅))). We also report the synthesis of the C,N-bis-ferrocene labeled tripeptide Phe-Ala-Leu and its electrochemical characterization. Starting from the enantiomerically pure ferrocene derivative 9, which was synthesized from ferrocene aldehyde and L-1-amino-ethylbenzene, two diastereomers 10 were obtained by peptide coupling with N-Boc protected D- and L-alanine. There was, however, only very little diastereomeric induction if 0.5 equiv of a racemic mixture of alanine were used. This suggests that amino acid activation rather than coupling is the rate-determining step. A combination of NOESY (nuclear Overhauser effect spectroscopy) spectra and molecular modeling furnished detailed insights into the solution structures of 3, 9, and 10 and was used to rationalize their different reactivity.

Full text of DOI:10.1021/ic000089+

Inorg. Chem. 2000, 39, 5437-5443
5437
Chiral Ferrocene Amines Derived from Amino Acids and Peptides: Synthesis, Solution and
X-ray Crystal Structures and Electrochemical Investigations
Alexandra Hess, Jan Sehnert, Thomas Weyhermu1ller, and Nils Metzler-Nolte*
Max-Planck-Institut fu¨r Strahlenchemie, Stiftstrasse 34-36, D-45470 Mu¨lheim/Ruhr, Germany
ReceiVed January 27, 2000
For the recognition of all but the simplest naturally occurring molecules, electrochemical sensors based on ferrocene
will certainly require chiral centers. To advance the necessary chemistry, this work describes the synthesis and
properties of ferrocene derivatives of enantiomerically pure amino acids, peptides, and other chiral amines. Ferrocene
aldehyde is condensed with amino acid esters to yield the corresponding Schiff bases 2, which are reduced by
NaBH4 in methanol to the ferrocene methyl amino acids 3. An X-ray single-crystal analysis was carried out on
the phenylalanine derivative 3a (monoclinic space group P21, a ) 10.301(1) Å, b ) 9.647(1) Å, c ) 18.479(2)
Å, â ) 102.98(2)°, Z ) 4). Further peptide chemistry at the C terminus proceeds smoothly as demonstrated by
the synthesis of the ferrocene labeled dipeptide Fc-CH2-Phe-Gly-OCH3 5 (Fc ) ferrocenyl ((η-C5H4)Fe(η-C5H5))).
We also report the synthesis of the C,N-bis-ferrocene labeled tripeptide Phe-Ala-Leu and its electrochemical
characterization. Starting from the enantiomerically pure ferrocene derivative 9, which was synthesized from
ferrocene aldehyde and L-1-amino-ethylbenzene, two diastereomers 10 were obtained by peptide coupling with
N-Boc protected D- and L-alanine. There was, however, only very little diastereomeric induction if 0.5 equiv of
a racemic mixture of alanine were used. This suggests that amino acid activation rather than coupling is the
rate-determining step. A combination of NOESY (nuclear Overhauser effect spectroscopy) spectra and molecular
modeling furnished detailed insights into the solution structures of 3, 9, and 10 and was used to rationalize their
different reactivity.
Introduction
and target discrimination. The recognition of chiral substrates
with organic molecules has been actively persued by a number
of groups.9-13 Organometallic compounds have successfully
served as markers for the detection of biomolecules with a
variety of techniques.14-21 We are interested in the synthesis
and applications of bio-organometallics, e.g., for the selective
recognition of nucleotides and DNA.22-24 In this project we have
recently reported simple achiral ferrocene amines as redox labels
Ferrocene serves as the redox active moiety in a large number
of chemical sensors with applications ranging from use in simple
anion sensors1 to the monitoring of glucose levels in the blood
of diabetes mellitus patients2,3 and use in DNA sensors.4-7 For
the binding of simple anions or cations, ferrocenes with
covalently attached amines are most often employed. Through
clever ligand design, remarkable selectivity between seemingly
very similar targets is achievable; recently, Beer and co-workers
reported the selective recognition of dihydrogenphosphate in
the presence of hydrogensulfate.8 For sensors that are targeted
to more complicated or diverse molecules of biological interest,
chiral sensors will most likely be required to optimize sensitivity
(10) Behr, J. P.; Lehn, J. M.; Vierling, P. HelV. Chim. Acta 1982, 65, 1853-
1867.
(11) Russell, K. C.; Lehn, J. M.; Kyritsakas, N.; DeCian, A.; Fischer, J.
New J. Chem. 1998, 22, 123-128.
(12) Huskens, J.; Goddard, R.; Reetz, M. T. J. Am. Chem. Soc. 1998, 120,
6617-6618.
(13) Hu, K.; Krakowiak, K. E.; Bradshaw, J. S.; Dalley, N. K.; Xue, G.;
Izatt, R. M. J. Heterocycl. Chem. 1999, 36, 347-354.
(14) Ryabov, A. D. Angew. Chem. 1991, 103, 945-955.
(15) Ryabov, A. D.; Goral, V. N.; Gorton, L.; Cso¨regi, E. Chem.sEur. J.
1999, 5, 961-967.
* To whom correspondence should be sent. New address: Institut fu¨r
Pharmazeutische Chemie, Universita¨t Heidelberg, Im Neuenheimer Feld
364, D-69120 Heidelberg. Fax: +49-(0)6227-546430. E-mail: Nils.Metzler-
Nolte@uvz.uni-heidelberg.de.
(16) Salmain, M.; Vessie`res, A.; Brossier, P.; Butler, I. S.; Jaouen, G. J.
Immunol. Methods 1992, 148, 65-75.
(1) Beer, P. D. Acc. Chem. Res. 1998, 31, 71-80.
(2) Cass, A. E. G.; Davis, G.; Francis, G. D.; Hill, H. A. O.; Aston, W.
J.; Higgins, I. J.; Plotkin, E. V.; Scott, L. D. L.; Turner, A. P. F. Anal.
Chem. 1984, 56, 667-671.
(17) Jaouen, G.; Vessie`res, A.; Butler, I. S. Acc. Chem. Res. 1993, 26,
361-369.
(18) Di Gleria, K.; Nickerson, D.; Hill, H. A. O.; Wong, L.-L.; Fu¨lo¨p, V.
J. Am. Chem. Soc. 1998, 120, 46-52.
(3) Armstrong, F. A.; Hill, H. A. O.; Walton, N. J. Acc. Chem. Res. 1988,
21, 407-413.
(19) Grotjahn, D. B.; Joubran, C.; Combs, D.; Brune, D. C. J. Am. Chem.
Soc. 1998, 120, 11814-11815.
(4) Ihara, T.; Nakayama, M.; Murata, M.; Nakano, K.; Maeda, M. J. Chem.
Soc., Chem. Commun. 1997, 1609-1610.
(20) Severin, K.; Bergs, R.; Beck, W. Angew. Chem. 1998, 110, 1722-
1743; Angew. Chem., Int. Ed. Engl. 1998, 37, 1634-1654.
(21) Fish, R. H. Coord. Chem. ReV. 1999, 186, 569-584.
(22) Hess, A.; Metzler-Nolte, N. J. Chem. Soc., Chem. Commun. 1999,
885-886.
(5) Bardea, A.; Dagan, A.; Ben-Dov, I.; Amit, B.; Willner, I. J. Chem.
Soc., Chem. Commun. 1998, 839-840.
(6) Bardea, A.; Dagan, A.; Willner, I. Anal. Chim. Acta 1999, 385, 33-
43.
(7) Bardea, A.; Patolsky, F.; Dagan, A.; Willner, I. J. Chem. Soc., Chem.
Commun. 1999, 21-22.
(23) Brosch, O.; Weyhermu¨ller, T.; Metzler-Nolte, N. Inorg. Chem. 1999,
5308, 8-5313.
(8) Beer, P. D.; Drew, M. G. B.; Smith, D. K. J. Organomet. Chem. 1997,
543, 259-261.
(24) Brosch, O.; Weyhermu¨ller, T.; Metzler-Nolte, N. Eur. J. Inorg. Chem.
2000, 323-330.
(9) Bussmann, W.; Lehn, J. M.; Oesch, U.; Plumere, P.; Simon, W. HelV.
Chim. Acta 1981, 64, 657-661.
(25) Hess, A.; Brosch, O.; Weyhermu¨ller, T.; Metzler-Nolte, N. J.
Organomet. Chem. 1999, 589, 75-84.
10.1021/ic000089+ CCC: $19.00 © 2000 American Chemical Society
Published on Web 11/01/2000

5438 Inorganic Chemistry, Vol. 39, No. 24, 2000
Hess et al.
are given in hertz. 15N spectra were referenced to the absolute frequency
of 50.696 991 0 MHz, which was the resonance frequency of neat
nitromethane under the same experimental conditions. All resonances
for the C terminus of peptides.25 We now report the synthesis
and characterization of ferrocene derivatives of chiral amines.
From the natural chiral pool, amino acids are readily available,
easy to handle but at the same time offer a variety of
functionalities that can be exploited for further chemistry. Eckert
et al. have already demonstrated that ferrocene derivatives of
glycine (the simplest and achiral amino acid) can be very useful
for the electrochemical detection of amino acids and peptides
in liquid chromatography (ECD-HPLC).26,27 However, these
authors were merely interested in reactivity studies and their
products were never comprehensively or even structurally
characterized.28 In a structural study, the complexation of Pd(II)
ions by Schiff bases derived from ferrocene and amino acids
was investigated by Beck and co-workers.29 In the following,
we present a simple, high-yielding procedure for the synthesis
of chiral ferrocene methylamines, their coupling to amino acids
and peptides, along with spectroscopic (including electrochemi-
cal) and structural (including X-ray and molecular modeling)
characterization.
were assigned by 2D NMR (H-H-COSY and H-13C-HMQC for J
and long-range couplings). Where unambiguous or proven through
spectroscopy, the following conventions are used: δ/δ′ denotes pairs
of signals originating from s-cis/s-trans isomers, integration nH/2
indicates one signal of one rotational isomer only, “δ and δ′” denotes
pairs of diastereotopic signals. 15N chemical shifts and coupling
constants were taken from the F1 projection of indirect detection 1H-
15N correlated 2D spectra with 1024/256 data points in F1/F2, processed
after applying a matched cosine function and zero filling in both
dimensions. Mo¨ssbauer data were recorded on a spectrometer with
alternating constant accelaration and a 57Co source in a 6 µm Rh matrix.
The minimum experimental line width was 0.24 mm s-1 full width at
half-maximum. The sample temperature was maintained constant in
an Oxford Instruments VARIOX cryostat. Isomer shifts are quoted
relative to iron metal at 300 K. Molecular modeling was carried out
with the SPARTAN program, version 5, on a Silicon Graphics Indigo2
workstation (Wavefunction Inc., 18401 Von Karman Avenue, Suite
370, Irvine, CA).
1
1
X-ray Structure Determination of 3a. A transparent orange single
crystal of 0.70 × 0.32 × 0.14 mm3 was sealed in a glass capillary and
mounted on a Siemens SMART CCD-detector diffractometer system
at ambient temperature. Cell constants were obtained from a least-
squares fit of 5541 reflections. A total of 17 847 intensities (7590
independent with Rint ) 0.0364) were collected by a hemisphere run
taking frames at 0.30° in ω and corrected for Lorentz and polarization
effects. The program SADABS (G. Sheldrick, University of Go¨ttingen,
Germany, 1994) was used for absorption correction. The Siemens
ShelXTL software package (Siemens Analytical X-ray Instruments, Inc.)
was used for solution and refinement of the structure. Neutral atom
scattering factors were taken from the usual sources.60 All non-hydrogen
atoms were refined anisotropically. H atoms were placed at calculated
positions and refined as riding atoms with isotropic displacement
Experimental Section
All reactions were carried out in ordinary glassware and solvents
without further precautions except where indicated. Chemicals were
purchased from Aldrich-Sigma GmbH and used as received. Enantio-
merically pure L amino acids were used except where indicated.
Elemental analyses were carried out by H. Kolbe, Analytisches
Laboratorium, Mu¨lheim. IR spectra were recorded on a Perkin-Elmer
system 2000 instrument as KBr disks, additionally in CH2Cl2 solution
where indicated. Frequencies ν are given in cm-1. UV/vis spectra were
recorded on a Perkin-Elmer Lambda 19 spectrometer; only the
wavelengths of the lowest-energy ferrocene transition are given in
nanometers, ꢀ (dm3 mol-1 cm-1) in brackets. Mass spectra were
recorded by the mass spectrometry service group, Mu¨lheim, on a MAT
8200 (Finnigan GmbH, Bremen) instrument (EI, 70 eV) or on a MAT95
(Finnigan GmbH, Bremen) instrument (ESI, CH3OH solution, positive
ion detection mode). Only characteristic fragments are given with
intensities (%) and possible composition in brackets. Analytical HPLC
on D/L amino acid derivatives were carried out on a Nucleosil-Chiral-2
column (ABIMED model 305 pump and Shimadzu SPD-M10 AV diode
array detector) with cyclohexane and 0.5% propanol at a 0.8 mL/min
flow rate. Samples were taken from the reaction mixtures, evaporated
in vacuo, and redissolved in the HPLC solvent mixture. The ratio was
determined at 215 nm, where the highest extinction coefficients were
observed, and confirmed at 440 nm, where only bands from the
ferrocene moiety are recorded. Cyclic voltammograms were obtained
with a three-electrode cell and an EG&G Princeton Applied Research
model 273A potentiostat. A Ag/AgNO3 (0.01 mol/L in AgNO3)
reference electrode, a glass carbon disk working electrode of 2 mm
diameter, and a Pt wire counter electrode were used. Square wave
voltammograms were recorded with a step height of 1 mV, a 25 mV
pulse amplitude, and a 40 Hz frequency. CH2Cl2 solutions (ca. 10-4
mol/L) contained 0.1 mol/L Bu4NPF6 as supporting electrolyte. As an
internal standard, ferrocene was added in excess as a reference. NMR
spectra were recorded in CDCl3 at room temperature on a Bruker ARX
250 (1H at 250.13 MHz and 13C), DRX 400 (1H at 400.13 MHz, 13C
(30) Osman, A. M.; El-Maghraby, M. A.; Hassan, K. M. Bull. Chem. Soc.
Jpn. 1975, 48, 2226.
(31) Cano, J.; Benito, A.; Mart´ınez-Ma´n˜ez, R.; Soto, J.; Paya´, J.; Lloret,
F.; Julve, M.; Marcos, M. D.; Sinn, E. Inorg. Chim. Acta 1995, 231,
45-56.
(32) Beer, P. D.; Chen, Z.; Drew, M. G. B.; Johnson, A. O. M.; Smith, D.
K.; Spencer, P. Inorg. Chim. Acta 1996, 246, 143-150.
(33) Eckert, H.; Forster, B.; Seidel, C. Z. Naturforsch., B: Chem. Sci. 1991,
46, 339-352.
(34) Kraatz, H.-B.; Lusztyk, J.; Enright, G. D. Inorg. Chem. 1997, 36,
2400-2405.
(35) Lin, L.; Berces, A.; Kraatz, H.-B. J. Organomet. Chem. 1998, 556,
11-20.
(36) Marcus, R. A.; Sutin, N. Biochim. Biophys. Acta 1985, 811, 265-
322.
(37) Therien, M. J.; Chang, J.; Raphael, A. L.; Bowler, B. E.; Gray, H. B.
Struct. Bonding (Berlin) 1991, 75, 109-129.
(38) Winkler, J. R.; Gray, H. B. Chem. ReV. 1992, 92, 369-379.
(39) Onuchic, J. N.; Beratan, D. N.; Winkler, J. R.; Gray, H. B. Annu.
ReV. Biophys. Biomol. Struct. 1992, 21, 349-377.
(40) Gray, H. B.; Winkler, J. R. Annu. ReV. Biochem. 1996, 65, 537-561.
(41) Langen, R.; Colo´n, J. L.; Casimiro, D. R.; Karpishin, T. B.; Winkler,
J. R.; Gray, H. B. J. Biol. Inorg. Chem. 1996, 1, 221-225.
(42) Fernando, S. R. L.; Kozlov, G. V.; Ogawa, M. Y. Inorg. Chem. 1998,
37, 1900-1905.
(43) Murphy, C. J.; Arkin, M. R.; Jenkins, Y.; Ghatlia, N. D.; Bossmann,
S. H.; Turro, N. J.; Barton, J. K. Science 1993, 262, 1025-1029.
(44) Holmlin, R. E.; Dandliker, P. J.; Barton, J. K. Angew. Chem. 1997,
109, 2830-2848; Angew. Chem., Int. Ed. Engl. 1997, 36, 2714-2730.
(45) Meggers, E.; Michel-Beyerle, M. E.; Giese, B. J. Am. Chem. Soc.
1998, 120, 12950-12955.
and 2D spectra), and DRX 500 (1H at 500.13 MHz, 13C, 15N, 2D). H
1
and 13C spectra were referenced to TMS using the 13C signals or the
residual protio signals of the deuterated solvents as internal standards
(CDCl3 ≡ 7.24 (1H) and 77.0 (13C), DMSO ≡ 2.49 (1H) and 39.5 (13C)).
Positive chemical shift values δ (in ppm) indicate a downfield shift
from the standard, and only the absolute values of coupling constants
(46) Giese, B.; Wessely, S.; Spormann, M.; Lindemann, U.; Meggers, E.;
Michel-Beyerle, M. E. Angew. Chem. 1999, 111, 1050-1052; Angew.
Chem., Int. Ed. 1999, 38, 996-998.
(26) Eckert, H.; Koller, M. Z. Naturforsch., B: Chem. Sci. 1990, 45, 1709-
1714.
(47) Meggers, E.; Giese, B. Nucleosides Nucleotides 1999, 18, 1317-1318.
(48) Diederichsen, U. Angew. Chem. 1997, 109, 2411-2413; Angew.
Chem., Int. Ed. Engl. 1997, 36, 2317-2319.
(27) Eckert, H.; Koller, M. J. Liq. Chromatogr. 1990, 13, 3399-3414.
(28) Eckert, H.; Seidel, C. Angew. Chem. 1986, 98, 168-170; Angew.
Chem., Int. Ed. Engl. 1986, 25, 159-161.
(29) Freiesleben, D.; Polborn, K.; Robl, C.; Su¨nkel, K.; Beck, W. Can. J.
Chem. 1995, 73, 1164-1174.
(49) Dong, T.-Y.; Chang, C.-K.; Cheng, C.-H. J. Organomet. Chem. 1999,
587, 46-48.
(50) Arduengo, A. J., III; Harlow, R. L.; Kline, M. J. Am. Chem. Soc.
1991, 113, 361-363.

Chiral Ferrocene Amines
Inorganic Chemistry, Vol. 39, No. 24, 2000 5439
parameters. The absolute structure was determined reliably (Flack
parameter -0.01(2)) in accordance with the known stereochemistry
(L) of the amino acid.
(ꢀ, dm3 mol-1 cm-1)): 437 (118). MS (m/z): 361 (92, M+•), 296 (3),
280 (6), 199 (100). CV: -3 mV. Anal. Calcd for C17H23FeNO2 (361.29
g/mol): C, 56.5; H, 6.4; N, 3.9. Found: C, 56.4; H, 6.4; N, 4.1.
1
2d. H NMR: δ 8.18 (1H, s, CHdN), 4.71 (1H, HCp,o), 4.65 (1H,
Cp,o), 4.39 (2H, HCp,m), 4.19 (5H, s, Cp), 4.04 (1H, dd, CRH), 3.73
General Synthesis of Ferrocene-Methylamine Derivatives 3 and
9. The amino acid hydrochloride, triethylamine, and ferrocene aldehyde
(1 equiv each, typical scale 2 mmol) were refluxed in dry CHCl3 for 3
h. After evaporation of the solvent on a rotary evaporator, the imine
was obtained quantitatively as a yellow solid, the purity of which was
checked by 1H NMR (see text and data below). The imine was dissolved
in dry CH3OH, and 4 equiv of solid NaBH4 were added in small portions
at 0 °C. After the mixture was stirred for 30 min, 20 mL of 1 mol/L
aqueous NaOH was added and the organic phase extracted with CHCl3
(3 × 100 mL). The combined organic phases were dried and evaporated
to dryness to afford the pure ferrocene methyl amino acid in 90-100%
yield.
H
(3H, s, OCH3), 2.54 (1H, mult, CâH), 2.42 (1H, mult, CâH), 2.21 (2H,
mult, CγH2), 2.07 (3H, s, S-CH3).
4. Compound 3a (0.38 g, 1 mmol) was suspended in aqueous 2 N
NaOH solution. After 1 h of refluxing followed by filtration, the clear
solution was cooled to 0 °C and concentrated HCl was added slowly
to it. The orange precipitate was filtered off and dried to afford the
pure acid 4 (0.34 g, 94%). 1H NMR (DMSO): δ 7.30-7.22 (5H, mult,
HPh), 4.41 (1H, HCp,o), 4.38 (1H, HCp,o), 4.23 (2H, HCp,m), 4.17 (5H, s,
Cp), 4.16 (1H, mult, CRH), 3.94 (2H, br, Cp-CH2), 3.33 (1H, dd, J )
14.0 Hz, J ) 4.7 Hz, CâH), 3.05 (1H, dd, J ) 14.0 Hz, J ) 8.7 Hz,
CâH). 13C NMR (DMSO): δ 169.5 (CO2), 135.1 (Ci), 129.6, 128.8,
127.5 (CPh,p), 76.2 (Cpi), 71.0, 70.9, 69.2 (all Cp), 69.0 (5H, Cp), 59.4
(CR), 45.2 (N-CH2), 35.2 (Câ). IR (KBr): 1623sh, 1602vs cm-1. MS
(m/z): 363 (100, M+•), 251 (4), 199 (93). Mp: 330 °C (dec).
5. To a solution of 4, glycine methyl ester and NEt3 (1 equiv each,
1 mmol) in CH3CN was added 1 equiv of HBTU (O-(1H-benzotriazol-
1-yl)-N,N,N′,N′-tetramethyluronium hexafluorophosphate). After the
mixture was stirred for 15 min, brine was added. The solution was
diluted with ethyl acetate, and the organic phase was washed with
aqueous 2 N HCl, water, aqueous NaHCO3, and water. The organic
phase was dried and evaporated to dryness to afford 5 (0.37 g, 85%)
1
3a. H NMR: δ 7.30-7.17 (5H, mult, HPh), 4.13 (1H, HCp,o), 4.11
(1H, HCp,o), 4.06 (2H, HCp,m), 4.03 (5H, s, Cp), 3.65 (3H, s, OCH3),
3.61 (1H, t, J ) 6.8 Hz; CRH), 3.46 (1H, d, J ) 12.8 Hz, Cp-CH2),
3.34 (1H, d, J ) 12.8 Hz, Cp-CH2), 2.98 (2H, mult, CâH2), 2.0 (1H,
br, NH). 13C NMR: δ 174.9 (CO2), 137.2 (Ci), 129.2, 128.4, 126.8
(CPh,p), 86.6 (Cpi), 68.3 (5C, Cp), 68.2, 68.1, 67.7, 67.6 (all Cp), 62.4
(CR), 51.6 (OCH3), 47.1 (N-CH2), 39.7 (Câ). IR (KBr): 1728vs, 1719vs
cm-1. MS (m/z): 377 (50, M+•), 312 (2), 218 (5), 199 (100). CV: +2
mV. Anal. Calcd for C21H23FeNO2 (377.27 g/mol): C, 66.9; H, 6.1;
N, 3.7. Found: C, 66.8; H, 6.2; N, 3.5.
1
2a. H NMR: δ 7.88 (1H, s, CHdN), 4.58 (1H, HCp,o), 4.54 (1H,
1
as a yellow solid. H NMR: δ 7.81 (1H, br, t, J ) 5.1 Hz, NHGly),
HCp,o), 4.30 (2H, HCp,m), 4.07 (1H, dd, J ) 9.3 Hz, J ) 4.8 Hz, CRH),
3.96 (5H, s, Cp), 3.69 (3H, s, OCH3), 3.29 (1H, mult, CâH), 3.10 (1H,
mult, CâH).
7.30 (3H, mult, HPh), 7.23 (2H, mult, HPh), 4.10 (1H, HCp,o), 4.06 (2H,
d, J ) 5.1 Hz, CR,GlyH), 4.03 (2H, HCp,m), 3.95 (1H, HCp,o), 3.87 (5H,
s, Cp), 3.73 (3H, s, OCH3), 3.46 (1H, dd, J ) 10.6 Hz, J ) 3.8 Hz,
CR,PheH), 3.44 (1H, d, J ) 13.2 Hz, Cp-CH2), 3.27 (1H, dd, J ) 13.9
Hz, J ) 3.8 Hz, CâH2), 3.20 (1H, d, J ) 13.2 Hz, Cp-CH2), 2.68
(1H, dd, J ) 13.9 Hz, J ) 10.6 Hz, CâH2). 1H NMR (DMSO): δ 8.33
(1H, br, NHGly), 7.27 (4H, mult, HPh), 7.21 (1H, mult, HPh), 4.15 (1H,
3c. 1H NMR: δ 4.18 (1H, HCp,o), 4.12 (1H, HCp,o), 4.11 (5H, s, Cp),
4.07 (2H, HCp,m), 3.70 (3H, s, OCH3), 3.44 (1H, d, J ) 12.6 Hz, Cp-
CH2), 3.33 (1H, d, J ) 12.6 Hz, Cp-CH2), 3.31 (1H, app t, J ) 7.2
Hz, CRH), 1.68 (1H, mult, CγH), 1.45 (2H, t, J ) 7.2 Hz, CâH2), 0.90
(3H, d, J ) 6.6 Hz, CH3), 0.85 (3H, d, J ) 6.6 Hz, CH3). 13C NMR:
δ 176.5 (CO2), 86.6 (Cpi), 68.4 (5C, Cp), 67.9, 67.7 (2C), 67.6 (all
Cp), 59.5 (CR), 51.6 (OCH3), 47.2 (NCH2), 42.9 (Câ), 25.0 (Cγ), 22.7,
22.4 (both CH3). IR (KBr): 1735vs cm-1. UV (nm (ꢀ, dm3 mol-1
cm-1)): 437 (109). MS (m/z): 343 (71, M+•), 278 (5), 199 (100). CV:
+2 mV. Anal. Calcd for C18H25FeNO2 (343.25 g/mol): C, 63.5; H,
7.3; N, 4.1. Found: C, 62.5; H, 6.9; N, 4.2.
HCp,o), 4.08 (1H, HCp,o), 4.04 (2H, HCp,m), 3.99 (5H, s, Cp), 3.88 (1H,
d, J ) 6.0 Hz, CR,GlyH), 3.64 (3H, s, OCH3), 4.02 (1H, mult, CR, PheH),
3.41 (1H, d, J ) 12.4 Hz, Cp-CH2), 3.18 (1H, d, J ) 12.4 Hz, Cp-
CH2), 2.96 (1H, dd, J ) 13.8 Hz, J ) 8.3 Hz, CâH2), 2.73 (1H, dd, J
) 13.8 Hz, J ) 5.3 Hz, CâH2). 13C NMR (CDCl3): δ 174.2, 170.3
(both CO), 137.5 (Ci), 129.0, 128.9, 127.1 (CPh,p), 86.3 (Cpi), 68.3 (5H,
Cp), 67.8, 67.7, 67.5, 67.3 (all Cp), 63.8 (CR, Phe), 52.3 (OCH3), 47.7
(N-CH2), 40.8 (CR, Gly) 39.4 (Câ). 15N NMR: δ -281 (NGly). IR (KBr):
3360br, 1752vs, 1674vs cm-1. MS (m/z): 434 (42, M+•), 390 (1), 369
(2), 291 (4), 214 (18), 199 (100). Anal. Calcd for C23H26FeN2O3 (434.32
g/mol): C, 63.6; H, 6.0; N, 6.5. Found: C, 63.8; H, 6.1; N, 6.4.
6. Details of the preparation and properties of 6 will be given
elsewhere. Only preliminary data are given for comparison. 1H NMR:
δ 7.76/7.66 (1H, br, NHLeu), 7.13 (2H, mult, HPh), 7.06 (2H, mult, HPh),
5.25/4.89 (1H, mult, CR,LeuH), 4.85 (1H/2, d, CH2), 4.59 (1H/2, d, CH2),
4.46 (1H, d), 4.44 (1H/2, s), 4.38 (1H/2, d), 4.17 (1H/2, s, Cp), 4.15
(1H/2, s, Cp), 4.12 (1H, s, Cp), 4.01 (1H/2), 4.08 (1H), 4.10/4.08 (5H,
s, Cp), 4.01 (1H/2, d), 4.00 (1H, d), 3.51/3.42 (1H, quart, J ) 7.0 Hz,
CR,AlaH), 2.33/2.32 (3H, s, Ar-CH3), 1.68, 1.56, 1.43 (3H, several mult,
Câ,LeuH2, Cγ, LeuH), 1.35/1.29 (3H, d, J ) 7.0 Hz, CH3,Ala), 1.08/0.96
(3H, d, J ) 6.5 Hz, CH3,Leu), 0.79, 0.75 (6H, d, J ) 6.0 Hz, CH3,Leu).
C28H37FeN3O2 (503.47 g/mol). MS (m/z): 503 (100), 438 (66), 360
(61), 199 (77).
7. 7 was prepared analogously to 5 (70% yield). 1H NMR: δ 7.84/
7.76 (1H, d, J ) 7.9 Hz, NHAla), 6.82/6.73 (1H, d, J ) 8.6 Hz, NHLeu),
7.32-7.03 (10H, several mult, HPh), 5.22/4.89 (1H, mult, CR,LeuH), 4.83
and 4.06, 4.60 and 4.02, 4.37 and 4.03 (1H each, all pairs of
diastereotopic N-CH2), 4.53 (1H, s, N-CH2), 4.45 (1H, CR,AlaH), 4.40
(1H, Cp), 4.15-3.94 (7H, Cp), 4.09/4.08 (5H, Cp), 3.91/3.89 (5H, Cp),
3.46 (1H, mult, CR,PheH), 3.44/3.20 (2H, Phe-CH2-Cp), 2.33 (s, Ar-
CH3), 1.67/1.52 (1H, mult, Cγ,LeuH), 1.63 and 1.45/1.50 and 1.30 (2H,
mult, Câ,LeuH2), 1.36/1.31 (3H, d, J ) 7.0 Hz, Câ,AlaH), 1.03, 0.91,
0.76, 0.71 (3H/2 each, d, J ) 6.4 Hz, 6.5 Hz, 6.4 Hz, 6.2 Hz, Cδ,LeuH3).
13C NMR: δ 173.6/173.5 (COPhe), 172.2/171.8 (COLeu), 171.7/171.4
(COAla), 137.4 (Ci,Phe), 137.3/137.0, 133.9/133.1 (Cquart,Ar), 129.5, 129.3,
129.0, 128.8, 127.9, 127.0, 126.8 (aromatic C), 86.2/86.1 (Phe-CH2-
Cpi), 82.6/82.0 (Cpi), 68.8/68.6, 68.1/68.2 (10H, Cp), 69.6-67.6 (all
1
2c. H NMR: δ 8.11 (1H, s, CHdN), 4.65 (2H, HCp,o), 4.35 (2H,
Cp,m), 4.15 (5H, s, Cp), 3.89 (1H, dd, J ) 5 Hz, J ) 9.3 Hz, CRH),
H
3.69 (3H, s, OCH3), 1.80 (1H, mult, CâH), 1.70 (1H, mult, CâH), 1.56
(1H, mult, CγH), 0.90 (3H, d, J ) 6.7 Hz, CH3), 0.86 (3H, d, J ) 6.4
Hz, CH3).
3d. 1H NMR: δ 4.18 (1H, HCp,o), 4.12 (1H, HCp,o), 4.11 (5H, s, Cp),
4.06 (2H, HCp,m), 3.71 (3H, s, OCH3), 3.46 (1H, d, J ) 12.6 Hz, Cp-
CH2), 3.41 (1H, dd, CRH), 3.34 (1H, d, J ) 12.6 Hz, Cp-CH2), 2.57
(2H, mult, CγH2), 2.07 (3H, s, S-CH3), 1.92 (1H, mult, CâH), 1.83
(1H, mult, CâH). 13C NMR: δ 175.5 (CO2), 86.6 (Cpi), 68.4 (5C, Cp),
68.3, 67.8, 67.7, 67.6 (all Cp), 59.7 (CR), 51.8 (OCH3), 47.2 (NCH2),
32.9 (Câ), 30.5 (Cγ), 15.4 (SCH3). IR (KBr): 1735vs cm-1. UV (nm
(51) Arduengo, A. J., III; Dias, H. V. R.; Dixon, D. A.; Harlow, R. L.;
Klooster, W. T.; Koetzle, T. F. J. Am. Chem. Soc. 1994, 116, 6812-
6822.
(52) Denk, M.; Green, J. C.; Metzler, N.; Wagner, M. J. Chem. Soc., Dalton
Trans. 1994, 2405-2410.
(53) Roberts, G. C. K. NMR of Macromolecules - A Practical Approach;
Oxford University Press: Oxford, 1993; Vol. 134.
(54) Meissner, A.; Haehnel, W.; Vahrenkamp, H. Chem.sEur. J. 1997, 3,
261-267.
(55) Knipp, B.; Mu¨ller, M.; Metzler-Nolte, N.; Balaban, T. S.; Braslavsky,
S. E.; Schaffner, K. HelV. Chim. Acta 1998, 81, 881-888.
(56) Turner, D. L. Prog. Nucl. Magn. Reson. Spectrosc. 1985, 17, 281-
358.
(57) Perrin, C. L.; Dwyer, T. J. Chem. ReV. 1990, 90, 935-967.
(58) Neuhaus, D.; Williamson, M. The Nuclear OVerhauser Effect in
Structural and Conformational Analysis; VCH Publishers: New York,
1989.
(59) Tropp, J. J. Chem. Phys. 1980, 72, 6035-6043.
(60) International Tables for Crystallography; Kynoch Press: Birmingham,
1974; Vol. 4.

5440 Inorganic Chemistry, Vol. 39, No. 24, 2000
Hess et al.
Scheme 1
Table 1. Summary of Crystallographic Details for 3a
Cp), 63.6/63.5 (CR,Phe), 48.4/48.3 (CR,Ala), 47.8/47.6 (CR,Leu), 47.7/47.6
(Phe-CH2-Cp), 49.0, 46.5, 45.4, 44.0 (all remaining CH2), 43.5/42.2
(Câ,Leu), 39.4 (Câ,Phe), 24.8/24.6 (Cγ,Leu), 23.6, 23.4, 21.6, 21.6 (Cδ,Leu),
21.1/21.0 (Ar-CH3), 18.4/18.0 (Câ,Ala). 15N NMR: δ -262, -263. IR
(KBr): 3092w, 3026vw, 2956m, 2927m, 2868w, 1638vs,br cm-1. MS
(m/z): 848 (19, M+•), 783 (10), 705 (17), 199 (100). CV: +60 mV.
Mp: 72-73 °C. Anal. Calcd for C48H56Fe2N4O3 (848.69 g/mol): C,
67.9; H, 6.7; N, 6.6. Found: C, 67.7; H, 6.7; N, 6.7.
chemical formula
fw
C21H23FeNO2
377.25
space group (No.)
a, Å
b, Å
c, Å
P21
10.301(1)
9.647(1)
18.479(2)
102.98(2)°
1789.4(3)
4
â, deg
V, Å3
1
9. Brown oil. H NMR: δ 7.33 (4H, mult, HPh), 7.25 (1H, mult,
Z
HPh), 4.13 (1H, mult, HCp,o), 4.12 (1H, mult, HCp,o), 4.07 (2H, HCp,m),
4.05 (5H, s, Cp), 3.80 (1H, quart, J ) 6.6 Hz, CH), 3.55 (1H, d, J )
13.0 Hz, CH2), 3.25 (1H, d, J ) 13.0 Hz, CH2), 1.32 (1H, d, J ) 6.6
Hz, CH3). 13C NMR: δ 145.6 (Ci), 128.4, 126.8 (CPh,p), 126.6, 87.2
(Cpi), 68.4, 68.3 (5C), 68.1, 67.7, 67.6 (all Cp), 57.5 (CH), 46.6 (N-
CH2), 24.6 (CΗ3). IR (KBr): 3084m, 3024, 2962m, 2924, 1492m,
1450m cm-1. MS (m/z): 390 (100, M+•), 253 (18), 200 (25), 199 (32).
CV: -6 mV.
D (calcd), g/cm-3
temp (K)
1.40
298
λ, Å
0.71073
0.856
0.0599
0.1648
7287/3/457
1.062
abs coeff (mm-1
R1a (I > 2σ(I))
wR2a (all data)
)
data/restraints/parameters
GOF on F2
a R1 ) (∑Fo| - |Fc|/∑Fo|); wR2 ) [∑(Fo2 - Fc2)2/∑Fo ]1/2; GOF )
4
General Synthesis of Ferrocene Amino Acids 10. The amino acid
(1.15 equiv), DhbtOH (1.15 equiv), and 9 (1 equiv, typical scale 1
mmol) were dissolved in 25 mL of DMF. After addition of 25 mL of
CH2Cl2 and cooling to 0 °C, DCC (1.25 equiv) was added, and stirring
continued for 2 h and an additional 12 h at room temperature. A white
precipitate of dicyclohexylurea was removed by filtration and washed
twice with CH2Cl2. The combined organic phases were washed with
saturated NaHCO3 (3 × 50 mL), brine (50 mL), KHSO4 (3 × 50 mL),
and brine (50 mL). The organic phase was dried over MgSO4 and the
solvent removed on a rotary evaporator. The oily residue was
chromatographed over a silica column with ether/pentane (5:1).
D-10b. 1H NMR: δ 7.40-7.15 (5H, mult, HPh), 5.79/5.12 (1H, quart,
J ) 6.8 Hz, C*H-Ph), 5.44/5.26 (1H, d, J ) 8.0 Hz, NHBoc), 4.66
(1H, mult, CRH), 4.44/4.05 (1H, d, J ) 6.8 Hz, CH2), 4.40/3.81 (1H,
d, J ) 8.8 Hz, CH2), 4.18 (1H/2 see text, HCp), 4.11 (1H/2, HCp), 4.05/
4.02 (5H, s, Cp), 3.96 (2H/2, HCp), 3.92 (2H/2, HCp), 3.85 (2H/2, HCp),
3.81 (1H/2, HCp), 1.55/1.53 (3H, d, J ) 6.8 Hz, C*H-CH3), 1.43/1.41
(9H, s, C(CH3)3)), 1.28/1.12 (3H, d, J ) 6.8 Hz, CâH3). 13C NMR: δ
174.3/173.1, (COAmide), 155.1/155.0 (COBoc), 140.6/139.9 (Ci), 128.7,
128.4, 127.7, 127.3, 127.2, 127.1 (CPh, p), 84.8/84.4 (Cpi), 79.5
(C(CH3)3), 68.8/68.5 (5H, Cp), 70.2, 69.9, 69.5, 69.1, 68.2, 67.8, 67.7,
67.2 (all Cp), 55.1/52.0 (C*H), 46.7/46.6 (CR), 44.0/42.1 (N-CH2),
19.9/18.9, 19.3/17.2 (both Câ). 15N NMR: δ -288/-287. IR (KBr):
3340br, 1706vs, 1639vs cm-1. UV (nm (ꢀ, dm3 mol-1 cm-1)): 436
(113). MS (m/z): 490 (100, M+•), 434 (52), 416 (28), 369 (41), 199
(66). CV: +20 mV. Anal. Calcd for C27H34FeN2O3 (490.43 g/mol):
C, 66.1; H, 7.0; N, 5.7. Found: C, 66.0; H, 7.1; N, 5.6.
2
[∑(Fo - Fc2)2/(no. of reflns - no. of params)]1/2
.
(C*H), 47.0/46.6 (CR), 43.2/42.4 (N-CH2), 30.3/29.7 (C*H-CH3),
28.4/28.3 (CBoc), 19.6/18.4, 19.3/16.7 (Câ). 15N NMR: δ -287.5. IR
(KBr): 1709vs, 1638vs cm-1. UV: 445 (115). MS (m/z): 490 (98,
M+•), 434 (44), 416 (81), 369 (36), 199 (61), 105 (100). CV: +15
mV. Anal. Calcd for C27H34FeN2O3 (490.43 g/mol): C, 66.1; H, 7.0;
N, 5.7. Found: C, 66.2; H, 7.0; N, 5.7.
10c. 1H NMR: δ 7.40-7.23 (5H, mult, HPh), 5.78/5.22 (1H, quart,
J ) 7.0 Hz, C*H-Ph), 5.2 (1H, br, NHBoc), 4.87 (1H, mult, CRH),
4.58 and 3.65 (1H, d, J ) 14.4 and 14.6 Hz, N-CH2), 4.35 and 3.88
(1H, d, J ) 16.0 and 16.3 Hz, N-CH2), 4.11-3.91 (4H, HCp), 4.06/
4.01 (5H, s, Cp), 1.58/1.36 (3H, d, C*H-CH3), 1.45/1.37 (9H, s,
C(CH3)3)), 1.52/1.18 (2H, mult, CâH2), 1.70 (1H, mult, CγH), 0.96/
0.89 (3H, d, J ) 6.4 Hz/6.7 Hz, CδH3), 0.85/0.82 (3H, d, J ) 6.8 Hz,
CδH3). 13C NMR: δ n. o. (COAmide), n. o. (COBoc), 140.6 (Ci), 128.6,
128.5, 127.9, 127.6, 127.5/127.2 (CPh,p), 84.9/84.6 (Cpi), 79.5 (C(CH3)3),
68.8/68.6 (5H, Cp), 70.1, 70.0, 69.9, 69.8, 69.7, 68.9, 68.8, 68.1 (all
Cp), 54.7/52.3 (C*H), 49.7/48.9 (CR), 43.4/42.5 (N-CH2), 42.9/42.8
(Câ), 25.0 (Cγ), 28.4/28.3 (CBoc), 23.6, 23.4, 21.9, 21.8 (all Cδ), 18.6/
16.7 (C*H-CH3). IR (KBr): 1708 vs, 1636 vs cm-1. MS (m/z): 532
(100, M+•), 476 (41), 458 (52), 411 (33), 199 (67). Mp: 50 °C. Anal.
Calcd for C30H40FeN2O3 (532.51 g/mol): C, 67.7; H, 7.6; N, 5.3.
Found: C, 67.5; H, 7.5; N, 5.1.
Results and Discussion
L-10b. 1H NMR: δ 7.39-7.21 (5H, mult, HPh), 5.82/5.11 (1H, quart,
J ) 7.0 Hz, C*H-Ph), 5.47/5.38 (1H, d, J ) 7.6 Hz, NHBoc), 4.73
(1H, mult, CRH), 4.54/and 3.70 (1H, d, J ) 14.4 Hz, CH2), 4.31/4.04
(1H, d, J ) 16.0 Hz, CH2), 4.14 (2H/2 see text, HCp), 4.09 (2H/2,
Synthesis. Ferrocene aldehyde 1 reacts readily with amino
acids to form Schiff bases 2.29,30 These imines were subsequently
reduced by NaBH4 in methanol to yield ferrocene methyl
derivatives 3 of the corresponding amino acids in good yield
(phenylalanine 3a, alanine 3b, leucine 3c, methionine 3d,
Scheme 1).29,31,32 In fact, formation of the Schiff base and
subsequent reduction may be carried out in a one-pot reaction
in methanol without isolation of the Schiff base. To our surprise,
HCp), 4.04/4.00 (5H, s, Cp), 3.98 (2H/2, HCp), 3.62 (2H/2, HCp), 1.57/
1.40 (3H, d, J ) 7.0 Hz, C*H-CH3), 1.44/1.40 (9H, s, C(CH3)3)),
1.19/1.12 (3H, d, J ) 6.4 Hz, CâH3). 13C NMR: δ 173.8/172.9,
(COAmide), 155.1/155.0 (COBoc), 140.6/139.9 (Ci), 128.6, 128.5, 127.9,
127.7, 127.6/127.0 (CPh, p), 84.8/84.4 (Cpi), 79.4 (C(CH3)3), 68.8/68.5
(5H, Cp), 70.0, 69.8, 69.7, 68.8, 68.6, 68.0, 67.3, 67.2 (all Cp), 52.1

Chiral Ferrocene Amines
Inorganic Chemistry, Vol. 39, No. 24, 2000 5441
Scheme 2
Scheme 3
aldehyde and enantiomerically pure L-1-amino-ethylbenzene
(Scheme 3). Compound 9 reacted smoothly with amino acids
such as alanine and leucine and in an equally facile manner
with both enantiomers of alanine. The ferrocene-labeled amino
acids 10 were obtained by our standard protocol in good yield
(Scheme 3).
Spectroscopic Characterization and Discussion
The reaction of 1 with amines proceeds smoothly in refluxing
CHCl3 or CH3OH with quantitative yield. The purity of the
1
Schiff bases was established by the disappearance of the H
NMR signal of the aldehyde proton at 10 ppm and the
appearance of a new signal at 7.9 ppm for the Fc-CHdN-R
imino proton. After reduction with NaBH4, this signal disappears
again, while a broad signal at 2.0 ppm (in CDCl3 at 300 K) for
the amino proton and a signal for the newly formed methylene
group are observed. As a consequence of the chiral CR atom of
the attached amino acid, the two protons of this CH2 group are
diastereotopic and give rise to an AB signal. For the same
reason, all protons and carbon atoms of the substituted Cp ring
become nonequivalent and four methine 13C NMR signals are
observed from this ring. Even at 500 MHz, not all signals in
the 1H NMR spectra are equally well resolved and usually only
three broad signals with unresolved couplings and intensities
1:1:2 are observed. Hydrolysis of the methyl ester proceeds
cleanly in very good yield, as does the coupling of the free
acid 4 to glycine methyl ester and dipeptide 6. The resulting
tripeptide 7 was purified by HPLC and its identity established
by electron-spray ionization mass spectrometry (ESI-MS).
A chiral Schiff base was also prepared by stirring a mixture
of ferrocene aldehyde 1 and L-1-amino-ethylbenzene in CHCl3
over solid NaHCO3. This Schiff base was readily reduced to
the ferrocene-methyl-N-(1-amino-ethylbenzene) (9) by NaBH4.
No racemization was observed in this reduction or the formation
of 3a-d. We have previously reported a simple coupling
reaction of secondary ferrocene amines 8 to the C terminus of
amino acids and peptides (Scheme 4). The same reaction could
be applied to 9 to yield amides 10. It should be noted that a
diastereomeric pair is formed after reaction with D- and
L-alanine. In addition, the amides exist as two isomers due to
rotation about the tertiary amide bond as indicated in Scheme
this reaction did not yield the corresponding glycine derivative
3e despite various efforts. This corresponds to an observation
by Eckert et al., who obtained 3e by a slightly different route.26,33
Also, the alanine derivative 3b could not be obtained analytically
pure, although the corresponding Schiff base 2b formed in
quantitative yield. The NaBH4 reduction appears to work better
for the larger amino acids such as leucine and phenylalanine,
but we were unable to isolate even one defined reaction product
from the glycine or alanine reductions. Compounds 3 are orange-
yellow oils or solids, and crystalline material for a single-crystal
X-ray analysis was obtained for the phenylalanine derivative
3a (Table 1). For further derivatization, the methyl ester in 3a
was hydrolyzed by aqueous NaOH to yield the free acid 4.
Coupling of glycine methyl ester to 4 proceeds smoothly via
an HBTU mediated coupling scheme to yield the dipeptide Fc-
CH2-Phe-Gly-OMe 5 (Fc ) ((η-C5H4)Fe(η-C5H5))) (Scheme
1). By the same method, the bis-ferrocene tripeptide 7 can be
prepared from 4 and the dipeptide H-Ala-Leu-N(CH2-
pCH3-C6H4)-CH2-Fc 6 in good yield (Scheme 2). In an
earlier study, we prepared a number of N-terminally function-
alized amino acids and peptides by reaction of activated amino
acids with the secondary ferrocene amine Fc-CH2-NH-CH2-
pCH3-C6H4 (8).25 To our surprise, we were unable to couple
activated amino acids to the secondary amino group in 3a. To
test whether steric congestion at the amino center might be the
reason for this unexpected result, we prepared the slightly
smaller methylamino-ferrocene derivative 9 from ferrocene
1
3. Both facts are nicely illustrated by the H NMR spectra of
D-10b and L-10b in Figure 1. In previous work we have
Scheme 4

5442 Inorganic Chemistry, Vol. 39, No. 24, 2000
Hess et al.
Figure 2. ORTEP plot of one molecule of 3a with numbering scheme,
selected bond lengths and angles: Fe-CCp (av) 2.047 Å; Fe-Cp-
(centroid) 1.653 and 1.652 Å (unsubstituted Cp); C(10)-C(11) 1.514-
(7) Å; Cp(centroid)-Fe-Cp(centroid) 178.9 Å.
and 9 is the chiral center in 9. In fact, 9 was used to investigate
whether a chiral induction would occur in the coupling reaction
of 9 to alanine as a model amino acid. To this end, the pure
diastereomers D-10b and L-10b were prepared independently
and characterized spectroscopically and by HPLC. Then, a
racemic mixture of alanine was activated with DhbtOH and
DCC and 0.5 equiv of 9 were added. Two parallel experiments
were carried out at 0 and 22 °C in DMF, and samples from
both runs were investigated by HPLC after 15 min and after 6
h. At both temperatures, there was only a slight excess of the
diastereomer originating from the D enantiomer (58% at 22 °C
and 59% at 0 °C), which did not change with time. The HPLC
findings could be confirmed by 1H NMR by careful integration
of characteristic signals of both diastereomers (see Figure 1) in
the reaction mixtures. This result is consistent with previous
electrochemical experiments in which activation of the amino
acid was found to be the rate-determining step in DhbtOH/DCC
mediated coupling reactions. DMF is the preferred solvent for
peptide couplings. However, the above findings were reproduced
in THF, suggesting little if any solvent-dependence of the rate-
determining step.
Figure 1. 1H NMR spectra of D-10b (top) and L-10b (bottom) between
3.5 and 6.0 ppm (see text).
unsuccessfully tried to induce a preference for one rotational
isomer by hydrogen bonding or steric bulk.25 It is noteworthy
that in diastereomers 10b, one rotational isomer is in slight
1
excess as indicated by integration of the H NMR signals in
Figure 1. We have used molecular modeling to gain further
insight into the geometry of both isomers (named s-cis and
s-trans conformers, vide infra). For diastereomeric compounds
such as 10b, two different values for the barrier of rotation about
the central amide bond are expected. In D6-DMSO, coalescence
for the two methyl groups in both diastereomers of 10b was
easily observed between 20 and 60 °C, giving activation energies
of 70.5 kJ/mol (D-10b) and 70.7 kJ/mol (L-10b). These values
are equal within the accuracy of determination ((1 kJ/mol) and
match the value previously determined for the achiral N-Boc-
glycine amide 12e (70.5 ( 0.5 kJ/mol, Scheme 4).25
All new compounds show reversible electrochemical waves
in the cyclic voltammograms around 0 mV vs ferrocene (see
Experimental Section) which are attributed to oxidation of the
iron center. For compounds 3, the shift potential difference does
not exceed 5 mV and is thus unsuitable for electrochemical
differentiation of the attached amino acid.34,35 Interestingly, a
small difference of 5 mV is also observed for the two
diastereomers of 10b. Comparable to the previously investigated
pair of 8 and 12e, the electrochemical potential of the ferrocene
moiety shifts more than 20 mV between amine 9 and amides
10. This shift difference is large enough to be analytically useful
in monitoring the process of the peptide coupling reaction. The
electronic interaction between metal centers at a defined
distance36 has raised considerable interest among chemists, e.g.,
in the question of electron transfer through peptides37-42 or the
stacked bases of helical DNA.43-48 The square wave voltam-
mogram of 7 is not significantly different from the sum of the
square wave voltammograms of 3a and 6, with its peak potential
comparable to that of 3a and 6 and a width at a half-potential
X-ray Analysis of 3a and Structural Investigations by
Molecular Modeling
An X-ray single-crystal analysis was carried out on 3a at 100
K. The result shows two independent molecules per asymmetric
unit. A graphical representation of molecule 1 is shown in Figure
2. The main difference from molecule 2 is a clockwise rotation
about the C(15)-C(16) bond by 36°. In the solid state, 3a forms
discrete molecules with the amino acid pointing away from the
ferrocene moiety and the phenyl ring being almost perpendicular
to the cyclopentadienyl rings. Both Cp rings are almost eclipsed
as concluded from an average dihedral angle H-C-C-H of
0.3°. All bond lengths and angles are well within the range
usually observed for simpler achiral ferrocene amines.25,31,32
Eckert et al. reported the synthesis of some related ferrocene
amino acid derivatives.28 However, these compounds were never
structurally characterized. The crystal structure of 3a presents
unambiguous proof for the proposed constitution of the chiral
ferrocene derivatives 3. To rationalize their reactivity, a more
detailed insight into their conformation in solution structures is
required.
w
1/2 of 230 mV (compared to a w1/2 of 118 and 119 mV for 3a
and 6, respectively). This result does not support a direct
electronic interaction between the ferrocene centers.49 Indeed,
molecular modeling suggests a minimum distance between the
metal centers of about 10 Å even in a helical structure which
places the metal centers in proximity.
We were surprised to find that coupling of activated amino
acids to ferrocene amines was possible for 9 but not for 3a.
Unfortunately, X-ray quality crystals of 9 could not be obtained.
We have therefore used molecular modeling (SYBYL force
field; see Experimental Section) to compare the steric congestion
Chiral Induction
The most significant difference between ferrocene amines 8

Chiral Ferrocene Amines
Inorganic Chemistry, Vol. 39, No. 24, 2000 5443
Figure 4. Molecular modeling structures of the s-cis and s-trans
isomers of L-10b (see text for details).
NOE cross-peaks58 are observed and confirm the overall
molecular structure of the molecule but are of little diagnostic
value with regard to the s-cis/s-trans problem. However, a few
diagnostic NOE cross-peaks confirm the molecular modeling
structures for both rotational isomers qualitatively; for the s-trans
isomer, a medium NOE is observed between the alanine methyl
group and one of the diastereotopic protons of the CH2 group.
From molecular modeling, the distance Câ,Ala (a good mean
value for the three hydrogen atoms on a freely rotating methyl
group)53,59 to the CH2 atoms is 2.9 and 4.3 Å. These distances
are 4.9 and 5.8 Å in the s-cis isomer, corresponding to a very
small NOE cross-peak only. On the other hand, the alanine
methyl group gets closer to the hydrogen center at the chiral
carbon atom of the phenylethylamine moiety in the s-cis isomer
(4.2 Å vs 5.1 Å), and indeed, a weak NOE cross-peak between
these groups is detectable for the s-cis isomer only. In the s-cis
isomer, the protons at both chiral centers are only 2.2 Å apart
(4.4 Å in the s-trans isomer), and indeed, a strong NOE in the
s-cis isomer is observed.
These data demonstrate that a combination of NMR spec-
troscopy and molecular modeling may be successfully used to
elucidate the solution structures even of relatively flexible bio-
organometallics.53 For a comprehensive treatment, a molecular
dynamics simulation would certainly be appropriate. However,
the agreement between crystallographic data and the minimum
structure for 3a in solution is convincing. A qualitative
interpretation of NOE data further supports the proposed
structure, and this is used in turn to rationalize the fundamentally
different reactivity of the two superficially very similar mol-
ecules 3a and 9 toward activated amino acids.
Figure 3. Comparison of the molecular structure of 3a derived from
an X-ray single-crystal structure (A, left) and molecular modeling (B,
middle). On the right, the molecular modeling structure of 9 is also
shown (C). See text for discussion of the structures and details on the
molecular modeling.
around the secondary nitrogen center in 9 and 3a (Figure 3).
Figure 3 compares the X-ray solid-state structure of 3a (left,
A) with the calculated minimum solution structure of 3a (middle,
B). Given the limitations of the force field approach, especially
when dealing with transition metal compounds, the overall
similarity of the two structures is quite convincing. The
calculated solution structure of 9 is shown on the right (C). A
comparison between B and C reveals that the ester group in B
points in the same direction as the lone pair of the nitrogen
atom in B and in C. This offers a compelling explanation for
the different reactivity of 3a and 9 toward activated amino acids
in that the carbonyl oxygen atom O(1) shields the lone pair on
the nitrogen from attack of the carbon electrophile. This
electronic shielding is not present in the simpler ferrocene
methylamine 9, and thus, amide bonds are as readily formed,
as in the sterically less demanding achiral ferrocene methylamine
8. Among others, a related reasoning has been used to explain
the stability of Arduengo-type carbenes.50-52
We have used a combined NMR and molecular modeling
approach to elucidate the structures of the s-cis and s-trans
conformers of 10b (Scheme 3).53 While structural analysis of
this kind is very common for biopolymers such as proteins and
oligonucleotides, the analysis of small molecules such as 10 is
complicated by their high degree of flexibility.54,55 To search
the conformational space, 72 starting structures were generated
by stepwise rotation around the N-CH bond and around the
C-Ph bond and subsequently optimized without constraints.
Figure 4 shows the minimum energy structures for the two
families of molecules. Both structures exhibit a similar orienta-
tion of the phenylethylamine substituent relative to the ferrocene
group and a voluminous Boc group pointing far away from the
rest of the molecule. The NMR situation is further complicated
by chemical exchange between the related protons sites of the
two rotational isomers. This gives rise to cross-peaks of opposite
sign relative to NOE (nuclear Overhauser effect) cross-peaks
between proton sites in spatial proximity.56,57 To avoid extensive
secondary NOE as a consequence of interconversion between
the two rotational isomers, a mixing time of 300 ms at 300 K
sample temperature was chosen. Under these conditions, many
This work shows a simple and high-yielding route to
ferrocene derivatives of highly functionalized biomolecules with
well-defined and predictable conformation. The development
of biosensors based on ferrocene peptide conjugates presented
in this work is in progress in our group.
Acknowledgment. The technical assistance of Agnetha
Schmidt is gratefully appreciated. The authors are also grateful
to Kerstin Sand and Jo¨rg Bitter for running numerous NMR
spectra and to Manuela Trinoga for expert HPLC services.
N.M.N. thanks Prof. K. Wieghardt for his support of our work.
Supporting Information Available: An X-ray crystallographic file
in CIF format for the structure determination of 3a. This material is
available free of charge via the Internet at http://pubs.acs.org.
IC000089+