Reaction of ferrocenecarboxylic acid with N,N′-disubstituted carbodiimides: Synthesis, spectroscopic and X-ray crystallographic analysis of N,N′-disubstituted N-ferrocenoylureas and identification of a one-pot coupling reagent for the formation of ferrocenecarboxamides in a non-aqueous solvent

Article abstract of DOI:10.1016/j.jorganchem.2003.12.045

N,N′-dicyclohexyl-N-ferrocenoylurea 2, N,N′-diisopropyl-N-ferrocenoylurea 3, N,N′-di-p-tolyl-N-ferrocenoylurea 4 and N,N′-di-tert-butyl- N-ferrocenoylurea 5 were obtained by reaction of ferrocenecarboxylic acid 1 with N,N′-dicyclohexylcarbodiimide (DCC), N,N′-diisopropylcarbodiimide (DIC), N,N′-di-p-tolylcarbodiimide 10 and N,N′-di-tert-butylcarbodiimide 11, respectively. Both N-tert-butyl-N′-ethyl-N-ferrocenoylurea 6 and N′-tert-butyl-N-ethyl-N-ferrocenoylurea 7 were obtained by reaction of 1 with N-tert-butyl-N′-ethylcarbodiimide 12. In all cases a small amount of ferrocenecarboxylic anhydride 8 was formed as a by-product. All compounds were characterized by ¹H NMR, ¹³C NMR, IR and MS. Single crystal X-ray structural analyses were made of 2, 3 and 4. From the consistent results, the reaction products of 1 with carbodiimides appear different from those proposed by some earlier workers. With N-(3-dimethylaminopropyl)- N′-ethylcarbodiimidehydrochloride 9 ferrocenoylurea was not isolated, but the main product was rather 8. The suitability of 8 as acylation reagent was applied by using 9 to obtain N-(3-triethoxysilyl)- propylferrocenecarboxamide in a one-pot reaction from 1 and 3-(triethoxysilyl)-propylamine.

Full text of DOI:10.1016/j.jorganchem.2003.12.045

Journal of Organometallic Chemistry 689 (2004) 1472–1480
www.elsevier.com/locate/jorganchem
Reaction of ferrocenecarboxylic acid with N,N0-disubstituted
carbodiimides: synthesis, spectroscopic and X-ray
crystallographic analysis of N,N0-disubstituted
N-ferrocenoylureas and identification of a one-pot coupling
reagent for the formation of ferrocenecarboxamides in a
non-aqueous solvent
*
Bernd Schetter, Bernd Speiser
Institut f€ur Organische Chemie, Universit€at T€ubingen, Auf der Morgenstelle 18, D-72076 T€ubingen, Germany
Received 3 November 2003; accepted 10 December 2003
Abstract
N,N0-dicyclohexyl-N-ferrocenoylurea 2, N,N0-diisopropyl-N-ferrocenoylurea 3, N,N0-di-p-tolyl-N-ferrocenoylurea 4 and N,N0-di-
tert-butyl-N-ferrocenoylurea 5 were obtained by reaction of ferrocenecarboxylic acid 1 with N,N0-dicyclohexylcarbodiimide (DCC),
N,N0-diisopropylcarbodiimide (DIC), N,N0-di-p-tolylcarbodiimide 10 and N,N0-di-tert-butylcarbodiimide 11, respectively. Both N-
tert-butyl-N0-ethyl-N-ferrocenoylurea 6 and N0-tert-butyl-N-ethyl-N-ferrocenoylurea 7 were obtained by reaction of 1 with N-tert-
butyl-N0-ethylcarbodiimide 12. In all cases a small amount of ferrocenecarboxylic anhydride 8 was formed as a by-product. All
compounds were characterized by 1H NMR, 13C NMR, IR and MS. Single crystal X-ray structural analyses were made of 2, 3 and
4. From the consistent results, the reaction products of 1 with carbodiimides appear different from those proposed by some earlier
workers. With N-(3-dimethylaminopropyl)-N0-ethylcarbodiimidehydrochloride 9 ferrocenoylurea was not isolated, but the main
product was rather 8. The suitability of 8 as acylation reagent was applied by using 9 to obtain N-(3-triethoxysilyl)-propylferro-
cenecarboxamide in a one-pot reaction from 1 and 3-(triethoxysilyl)-propylamine.
Ó 2004 Elsevier B.V. All rights reserved.
Keywords: Ferrocene; Ferrocenecarboxamide; Urea; Condensation; Coupling reaction; Activation
1. Introduction
reaction with an amine or an alcohol. Sometimes A
partly converts to an N-acylurea B (Scheme 1) which is
then obtained as an unwelcome by-product [1]. In some
special cases, the anhydride is formed in the reaction,
which is often also an effective intermediate for the re-
action to amides and esters [1].
Carbodiimides are versatile reagents in organic syn-
thesis [1]. N,N0-dicyclohexylcarbodiimide (DCC) and
N,N0-diisopropylcarbodiimide (DIC) are particularly
well-known activators for the formation of amides and
esters directly from carboxylic acids and amines or al-
cohols, respectively, under mild conditions in good
yields. The carbodiimide and the carboxylic acid nor-
mally form an O-acylisourea A, a reactive intermediate
that can easily be transformed into the amide or ester by
Ferrocenecarboxylic acid 1 reacts in a different way
with DCC, and N,N0-dicyclohexyl-N-ferrocenoylurea 2
(structure B) is the main product, as already discussed in
the literature [2–6], but either no spectroscopic investi-
1
gation was made [2,3], the H NMR and the IR spectra
were incorrectly interpreted [4], or the substance was
erroneously identified as N,N0-dicyclohexyl-O-ferro-
cenoylisourea [5]. One report [6] about the formation of
2 from DCC and 1 in the presence of 2-(6-hydroxy-
* Corresponding author. Tel.: +49-7071-2976-205; fax: +49-7071-
295-518.
E-mail address: bernd.speiser@uni-tuebingen.de (B. Speiser).
0022-328X/$ - see front matter Ó 2004 Elsevier B.V. All rights reserved.
doi:10.1016/j.jorganchem.2003.12.045

B. Schetter, B. Speiser / Journal of Organometallic Chemistry 689 (2004) 1472–1480
1473
Scheme 2. General pathway for the formation of N-ferrocenoylureas.
Scheme 1. Conversion of O-acylisoureas to N-acylureas.
gated carbodiimides (DCC, DIC, N,N0-di-p-tolylcarbo-
diimide 10, N,N0-di-tert-butylcarbodiimide 11, and N-
tert-butyl-N0-ethylcarbodiimide 12) formed stable N,N0-
disubstituted N-ferrocenoylureas (Scheme 2) with small
amounts of ferrocenecarboxylic anhydride 8 as the by-
product. N-(3-dimethylaminopropyl)-N0-ethyl-carbodi-
imidehydrochloride 9 is an exception, which will be
discussed separately below. Apart from 4 and 5, the
formation of the N,N0-disubstituted N-ferrocenoylureas
hexyloxy)-3-methoxy-6,7,10,11-tetrahexyloxytriphenyl-
1
ene includes a crystal structure and a H NMR spec-
trum, however without interpretation. No IR
investigation was made in [6]. Furthermore, no 13C
NMR spectrum of 2 has been published yet, although
13C NMR spectroscopy should allow unequivocal dis-
tinction between O-acylisourea (A) and N-acylurea
structures (B) from the typical strong shift of the car-
bonyl carbon nuclei.
In view of the frequent use of the ferrocene unit as a
redox active building block in larger molecular struc-
tures (for particularly striking examples see e.g. [7]), or
in new medicinal drugs (e.g., [8,9]), where it is often
bound by amide linkers, the need arises to activate 1 for
the reaction with amines under mild conditions. Obvi-
ously, the conventional reagent DCC is not able to
promote the desired reaction. It appeared thus of in-
terest to investigate other carbodiimides, which might
also be used as an alternative to the active ester pathway
[10].
in dichloromethane proceeds within seconds at room
temperature, while 4 needs one day at room temperature
and 5 five minutes at 105 °C for complete conversion.
The question if the product of the reaction between
DCC and 1 is N,N0-dicyclohexyl-O-ferrocenoylisourea
13 according to [5] or 2 according to [2–4,6] as well as
the question if the product between carbodiimides and 1
in general is an N,N0-disubstituted O-ferrocenoylisourea
or an N,N0-disubstituted N-ferrocenoylurea was inves-
tigated by 1H NMR, as well as 13C NMR spectroscopy,
MS and in the case of the reaction between DCC, DIC
or 10 and 1 by a single crystal structure analysis of the
products.
2. Results and discussion
The similarity of the spectroscopic properties indi-
cates that the products 2–7 belong to the same class of
compounds. Especially the signals of the two carbonyl
carbon nuclei at about 155 and 170 ppm in the 13C
In search of a carbodiimide for the activation of 1, we
tested various derivatives of DCC. Most of the investi-

1474
B. Schetter, B. Speiser / Journal of Organometallic Chemistry 689 (2004) 1472–1480
NMR spectrum indicate the formation of N,N0-disub-
stituted N-ferrocenoylureas. Table 1 summarizes the
chemical shifts of these nuclei and compares them with
the calculated values of both the corresponding N-ben-
zoylureas and the O-benzoylisoureas. The benzoyl de-
rivatives were taken as model for the ferrocenoyl
compounds since the reliability of shift value calcula-
tions for organometallic compounds is rather low. The
comparison with values for the model structures is
supported by the good correspondence between shift
values for ferrocenoyl and benzoyl derivatives. For ex-
ample, the chemical shifts of the carbonyl carbon nuclei
of ferrocenecarboxylic acid and benzoic acid differ less
than 1 ppm. In all cases reported in Table 1 the calcu-
lated values of the N-benzoylureas fit considerably bet-
ter to the experimental data than those of the
O-benzoylisoureas.
In the 1H NMR spectrum the signal of the N–H
proton appears at d > 5 ppm. In dry CDCl3, this acidic
hydrogen atom is undergoing an exchange reaction so
slowly that the signal is split by coupling with hydrogen
atoms that are bound to the neighbouring carbon atom,
if these are present (2, 3, 6). In this way, it was possible
to differentiate between the two isomers, 6 and 7, which
were obtained by reaction of 1 with N-tert-butyl-N0-
ethylcarbodiimide.
A further chemical argument that confirms the con-
clusion regarding the structure of the reaction products
is as follows: if these isomers were O-ferrocenoylisou-
Table 1
13C NMR chemical shifts values for carbonyl carbon atoms C1 and C2
d(C1)/ppm
d(C2)/ppm
R ¼ R0 ¼ c-C6H11, 2
R ¼ R0 ¼ CH(CH3)2, 3
R ¼ R0 ¼ p-C6H4CH3, 4
R ¼ R0 ¼ C(CH3)3, 5
A
B
C
A
B
C
A
B
C
A
B
C
A
B
C
A
B
C
172.6
177.2
163.4
172.4
175.7
164.5
175.0
173.6
164.5
169.5
170.7
165.0
169.3
170.7
164.0
174.8
176.4
165.6
154.8
154.4
151.4
154.6
153.3
150.0
152.5
151.5
151.5
153.8
156.6
127.3
155.5
156.1
142.1
153.3
158.9
149.8
Scheme 4. a-Fragmentation and decarbonylation of N-ferrocenoylu-
reas.
R ¼ C(CH3)3,
R0 ¼ CH2CH3, 6
R ¼ CH2CH3,
R0 ¼ C(CH3)3, 7
A: experimental values, B: calculated values for N-benzoylureas,
C6H5–(C1@O)–NR–(C2@O)–NHR0, C: calculated values for O-ben-
zoylureas, C6H5–(C1@O)–O–C2(NR)(NHR0).
Scheme 5. MacLafferty rearrangement of the N-ferrocenoylureas.
Scheme 3. Hypothetical acid–base isomerisation between O-ferrocenoylisoureas.

B. Schetter, B. Speiser / Journal of Organometallic Chemistry 689 (2004) 1472–1480
1475
Fig. 1. Crystal structure of 3, stereoscopic picture.
reas, one should assume that they have amidine like
basicities. The take-up and release of a proton by one of
the possible isomers, e.g. 14, should lead to a mixture of
14 and 15 (Scheme 3). We do not observe such an effect
in the case of 6 and 7, which provides evidence that O-
ferrocenoylisoureas are not the products of the reaction
between 1 and carbodiimides.
Based on these results, corrections to some earlier
work should be considered. In the interpretation of the
1H NMR spectrum of compound 2 [4] the multiplet
between 2.20 and 1.00 ppm was erroneously assigned to
all 22 instead of 20 protons of the cyclohexyl rings.
Furthermore, one of the other two protons of the cy-
clohexyl rings between 3.5 and 3.6 ppm was erroneously
identified as the NH proton, while the second at 4.2 ppm
and the NH proton failed to be noticed at all.
The formula weight of 2–7 was confirmed by MS. All
these compounds show an a-fragmentation (Scheme 4)
which leads to fragment a (m=z ¼ 213). Further de-
carbonylation yields fragment b (m=z ¼ 185). Another
characteristic is the appearance of a fragment c with
m=z ¼ 228 þ mðRÞ, R being the appropriate carbodii-
mide substituent (McLafferty rearrangement, Scheme 5).
This fragment is not expected for O-acylisoureas. The
latter should form the ferrocenecarboxylic ion
(m=z ¼ 230), but this was not detected in our spectra.
The observed fragmentation is in full agreement with
reported results for other carboxy ferrocenes [11,12]. In
the case of 6 and 7 the exchange of the tert-butyl-groups
against hydrogen and in the case of 5 the exchange of
one or both tert-butyl-groups is observed and leads to
fragments with m=z ¼ M ꢀ 55 and m=z ¼ M ꢀ 110.
In the IR spectra of 2–7 the N–H vibration leads to
an absorption at 3300–3150 cmꢀ1, which is quite weak in
the case of 4. Furthermore, in the region between 2800
and 1800 cmꢀ1 no significant absorption is detectable. In
an earlier IR spectroscopic investigation [4] a band at
2100 cmꢀ1 was registered and identified as NH–CO–N–
CO stretching band. In this region of wave numbers,
however, for 2 no band is detectable and the band dis-
cussed in [4] is probably caused by DCC impurities. This
starting material gives a strong band at 2117 cmꢀ1
.
The single crystal structure analysis confirms that the
products of 1 with DCC, DIC and 10 are 2, 3, and 4,
respectively. The structure of 2 is in accordance with
1
[6]. The triclinic unit cell has the space group P1, and
dimers with the symmetry Ci are formed by two hy-
drogen bonds between the N–H and the carbonyl oxy-
gen of the ferrocenoyl residue. The structure of 3 (Fig. 1)
is also characterized by the formation of dimers by hy-
drogen bonds between N2 and O3 as well as between N4
and O1, respectively, but these units do not have a centre
of inversion. Every forth isopropyl group (one isopropyl
group per two molecules) has enough space to occupy
split positions. The monoclinic unit cell has the sym-
metry P21/c. In contrast, 4 (Fig. 2) is monomeric and
forms an intramolecular hydrogen bond between O1 and
N2. The central molecule fragment c-C5H4–(C@O1)–
N1C–(C@O2)–N2H–C6H4–C is planar, indicating that
p-electrons are delocalized over this part of the molecule
structure. Furthermore, this shape forces the approach
of O2 to the ortho-hydrogen in the tolyl residue, which is
bound to N2. The N1 tolyl substituent is twisted out of
plane by 90° and for this reason the p-electron system of
this moiety is isolated from the rest of the molecule. The
triclinic unit cell has the space group P1.
1
The reference mentions that results of the single crystal structure
analysis of 2 were deposited at the Cambridge Crystallographic Data
Centre, but the deposition number was not reported. Following the
advice of one of the reviewers, the lattice parameters of 2 from our
determination are given in the following: a ¼ 1015:19ð9Þ pm,
b ¼ 1068:50ð8Þ
pm,
c ¼ 1177:53ð9Þ
pm;
a ¼ 104:369ð9Þ°,
b ¼ 108:628ð9Þ°, c ¼ 108:307ð9Þ°.

1476
B. Schetter, B. Speiser / Journal of Organometallic Chemistry 689 (2004) 1472–1480
Fig. 2. Crystal structure of 4.
Fig. 3. Selected crystallographic bond lengths for compounds 2, 3, and
4; plain and parenthesized values for 3 refer to the two independent
molecules of the compound within the crystal.
Some selected crystallographic bond lengths of the
urea residue are shown in Fig. 3 for species 2, 3, and 4.
The urea unit is asymmetric. Especially the N1–CO2
bond is remarkably long in contrast to the N2–CO2
bond. This could be a reason for the facile McLafferty
rearrangement observed in the mass spectra. The N–R
bond length depends on the nature of R.
tautomeric guanidinium ions [13,14]. The reaction
product between 1 and these ions is not able to convert
to an N-acylurea as easily as O-acylisoureas do owing to
the absence of a double bond and the presence of hy-
drogen at both nitrogen atoms (Scheme 6).
Of course, the stable N-ferrocenoylureas are not
suitable as coupling reagents for 1. On the other hand, 9
was found to form anhydride 8 with 1 (Scheme 6). The
latter compound is indeed activated with respect to
amide formation (see below). We explain this different
reaction by the fact that 9 converts in solution to cyclic
To demonstrate the activating ability of 9 in a
non-aqueous solvent, we used this carbodiimide as a cou-
pling reagent to form an amide from 3-(triethoxysilyl)-
propylamin 16 and 1 (Scheme 7) in a one-pot reaction.
Indeed, we could prepare the new compound N-(3-tri-
ethoxysilyl)-propylferrocenecarboxamide 17, in reason-
Scheme 6. Hypothetical pathway for the formation of 8 from 1 and 9.
Scheme 7. One-pot reaction of 1 with 16 under activation by 9 to form amide 17.

B. Schetter, B. Speiser / Journal of Organometallic Chemistry 689 (2004) 1472–1480
1477
able yield. Owing to the sensitivity of the silyl moiety to
hydrolysis, other pathways, e.g. condensation of an
amine with an acyl chloride with liberation of HCl, is
not advisable in this case. None of the other carbodii-
mides, especially DCC and DIC, usually employed in
non-aqueous solution, was successful in this case. It is
not necessary to prepare the chlorocarbonylferrocene [4]
or another activated ferrocene derivative [10] in a sep-
arate step. We are currently using 17 as a synthetic
building block for the modification of silica nanoparti-
cles [15].
JHH ¼ 7:1 Hz, 1H, NH); 13C NMR (CDCl3): d 24.6,
25.4, 25.5, 26.2 (c-CH–CH2–CH2–CH2–CH2–CH2),
31.0 (c-CH(NH)–CH2–CH2–CH2–CH2–CH2), 32.5 (c-
CH(N)–CH2–CH2–CH2–CH2–CH2), 49.8 (CH(NH)),
56.6 (CH(N)), 70.2, 70.3, 70.4 (c-C5H5, c-C–CH–CH–
CH–CH), 77.8 (c-C–(CH)5), 154.8 (N–(C@O)–N), 171.6
(C–(C@O)–N); MS (EI, 70 eV): m=z 436.2, 311.1, 229.0,
213.0, 185.0. Anal. Calc. for FeC24H32O2N2 (436.37): C,
66.06; H, 7.39; N, 6.42. Found: C, 65.77; H, 7.47; N,
6.17.
ꢀ1
~
3: IR (Atr): m/cm 3230 (NH), 3150–3000 (CH, fer-
rocene), 3000–2850 (CH, CH3CHCH3), 1700 (C@O); 1H
NMR (CDCl3): d 1.02–1.05 (d, JHH ¼ 7:5 Hz, 6H,
NHCH(CH3)2), 1.38–1.41 (d, JHH ¼ 7:9 Hz, 6H,
(C@O)2NCH(CH3)2), 3.83–3.87 (m, JHH ¼ 6:8 Hz, 1H,
NH–CH–(CH3)2), 4.17 (s, 5H, c-C5H5), 4.32 (s, 2H, c-C–
CH–CH–CH–CH), 4.62 (m, 1H, (C@O)2NCH–(CH3)2),
4.67 (s, 2H, c-C–CH–CH–CH–CH), 6.5 (b, 1H, NH);
13C NMR (CDCl3): d 21.1 (NH–CH–(CH3)2), 22.3
((C@O)2NCH–(CH3)2), 42.7 (NH–CH–(CH3)2), 48.9
((C@O)2NCH–(CH3)2), 70.1, 70.2, 70.3 (c-C5H5, c-C–
CH–CH–CH–CH), 77.8 (c-C–(CH)4), 154.6 (N–(C@O)–
N), 172.4 (C–(C@O)–N); MS (EI, 70 eV): m=z 356.1,
271.1, 229.1, 213.0, 185.0. Anal. Calc. for
FeC18H24O2N2 (356.24): C, 60.69; H, 6.79; N, 7.86.
3. Experimental
All reactions and operations were carried out under
argon. All solvents were distilled, dried and degassed
with Ar.
N,N0-dicylohexyl-N-ferrocenoylurea (2), N,N0-diiso-
propyl-N-ferrocenoylurea (3), and N,N0-di-p-tolyl-N-fer-
rocenoylurea (4) (see Table 2). Acid 1 and the
corresponding carbodiimide were suspended in 10 ml
dichloromethane in a Schlenk tube with the reactant and
solvent by means of an ultrasonic bath and then stirred
at room temperature. The clear brown solution was
concentrated to 4 ml and separated by column chroma-
tography (CC) with silica gel as the stationary phase and
a mixture of acetone and dichloromethane as the eluent.
The first coloured fraction was 8 [16] (yields see Table 2),
the second was the ferrocenoylurea. The solvent was
removed by vacuum and the resulting solid was kept at
0.1 mbar to removeꢀtr1ace amounts of carbodiimide.
Found: C, 60.23; H, 6.75; N, 7.45.
ꢀ1
~
4: IR (Atr): m/cm 3230–3190 (NH), 3150–3000 (CH,
ferrocene, tolyl group), 3000–2850 (CH, CH3), 1712
(C@O); 1H NMR (CDCl3):
d
2.33 (s, 3H,
NHC6H4CH3), 2.47 (s, 3H, (C@O)2NC6H4CH3), 4.22 (s,
5H, c-C5H5), 3.92, 4.28 (s, each 2H, c-C–CH–CH–CH–
CH), 7.12–7.15 (d, JHH ¼ 8:1 Hz, 2H, (C@O)2N–(c-C–
CH–CH–C(CH3)–CH–CH)), 7.21–7.23 (d, JHH ¼ 8:1
Hz, 2H, HN–(c-C–CH–CH–C(CH3)–CH–CH)), 7.31–
7.33 (d, JHH ¼ 7:8 Hz, 2H, HN–(c-C–CH–CH–C(CH3)–
CH–CH)), 7.52–7.54 (d, JHH ¼ 8:3 Hz, 2H, (C@O)2
N–(c-C–CH–CH–C(CH3)–CH–CH)), 11.82 (b, 1H,
NH); 13C NMR (CDCl3): d 20.8, 21.3 (CH3), 70.2, 71.8,
72.4 (c-C5H5, c-C–CH–CH–CH–CH), 74.2 (c-C–(CH)4),
120.0, 129.4, 129.8, 130.0 (c-C–(CH)2–C–(CH)2), 133.3,
~
2: IR (Atr): m/cm 3240 (NH), 3150–3000 (CH, fer-
rocene), 3000–2830 (CH, c-C6H11), 1700 (C@O); 1H
NMR (CDCl3): d 1.00–1.33 (m, 8H, c-C6H11), 1.54–1.65
(m, 4H, c-C6H11), 1.75–1.83 (m, 6H, c-C6H11), 1.97–2.06
(q, JHH ¼ 9:6 Hz, 2H, cc-C6H11), 3.5–3.65 (q, JHH ¼ 7:6
Hz, 1H, NH–CH), 4.20 (s, 5H, c-C5H5), 4.20–4.26 (m,
1H, (C@O)2NCH), 4.33 (s, 2H, c-C–CH–CH–CH–CH),
4.71 (s, 2H, c-C–CH–CH–CH–CH), 6.20–6.21 (d,
Table 2
Reaction conditions and properties of products 2, 3, and 4
Ferrocenoylurea
2
3
4
Amount of carbodiimide
0.625 mmol
(129 mg)
0.5 mmol
(115 mg)
15
0.625 mmol
(78 mg)
0.625 mmol
(220 mg)
1 mmol
Amount of 1
0.5 mmol
(115 mg)
5
(230 mg)
24
1/60
Stirring time/h
Eluent composition CH2Cl2/acetone
1/30
1/30
Carbodiimide removal time and temperature
Yield
48 h/50 °C
122 mg (58%)
7.6 mg (5%)
Dark orange
170 °Ca
20 h/30 °C
94 mg (55%)
6.3 mg (4%)
Dark orange
121 °C
20 h/30 °C
280 mg (64%)
13.7 mg (4%)
Brick red
149 °Cb
Yield of 8
Product color
Mp.
a Above slow thermal decomposition.
b Above fast thermal decomposition.

1478
B. Schetter, B. Speiser / Journal of Organometallic Chemistry 689 (2004) 1472–1480
135.4, 135.5, 139.1 (c-C–(CH)2–C–(CH)2), 152.5 (N–
(C@O)–N), 175.0 (C–(C@O)–N); MS (EI, 70 eV): m=z
319.1, 213.1, 185.1, 129.2, 91.2, 77.2, 66.2; MS (FAB):
m=z 452.1, 375.1, 319.1, 213.0, 185.0. Anal. Calc. for
FeC26H24O2N2 (452.38): C, 69.04; H, 5.35; N, 6.19.
Found: C, 68.69; H, 5.33; N, 5.91.
N,N0-dicylohexyl-N-ferrocenoylurea 2, in analogy to
Wang and Huang [5]. 0.32 g (140 mmol) of 1 and 0.43 g
DCC were dissolved in 15 ml THF and stirred for 9 h.
The solvent was removed by vacuum and the solid res-
idue was redissolved with a mixture of petrol ether (60–
90 °C) and ethyl acetate 5/1 and separated by CC with a
mixture of petrol ether (60–90 °C) and ethyl acetate 7/1.
The first fraction obtained was 8, and the second was 2,
which was purified by recrystallisation from petrol ether
(60–90 °C) and ethyl acetate 7/1. Differing from [5], no
acetic acid was added to destroy the unreacted carbo-
diimide.
N,N0-di-tert-butyl-N-ferrocenoylurea (5). 1 mmol (230
mg) 1 and 1 mmol (222 mg) N,N0-di-tert-butylcarbo-
diimide were suspended in 5 ml dichloromethane by
application of ultrasound and stirred for 5 min at 105 °C
under microwave heating. CC separation with silica gel
as the stationary phase and a mixture of ethyl acetate
and n-hexane 1/6 as the mobile phase yielded a first
coloured fraction [8, yield: <1 mg (<1%)], and a second
one (5), closely followed by N-tert-butylferrocenecarb-
oxamide [17,18] (22.7 mg, 8%). Because of the insuffi-
cient separation, the purification by CC had to be
repeated twice. The solvent was removed by vacuum
and the orange solid was kept for 5 h at 40 °C at 0.1
mbar to remove any trace amounts of carbodiimide.
ꢀ1
~
Yield: 64 mg ¼ 17%; Mp.: 145 °C; IR (Atr): m/cm
3314 (NH), 3150–3000 (CH, ferrocene), 3000–2850 (CH,
1
CH3), 1699 (C@O); H NMR (CDCl3): d 1.12 (s, 9H,
NHC(CH3)3), 1.49 (s, 9H, (C@O)2N–C(CH3)3), 4.18 (s,
5H, c-C5H5), 4.21–4.22 (t, JHH ¼ 1:6 Hz, 2H, c-C–CH–
CH–CH–CH), 4.77–4.78 (t, JHH ¼ 1:7 Hz, 2H, c-C–
CH–CH–CH–CH), 4.99 (s, 1H, NH); 13C NMR
(CDCl3): d 27.9 (NHC(CH3)3), 28.6 ((C@O)2N–C
(CH3)3), 51.9 (NHC(CH3)3), 57.7 ((C@O)2N–C(CH3)3),
69.5, 70.0, 70.4 (c-C5H5, c-C–CH–CH–CH–CH), 79.7
(c-C–(CH)4), 153.8 (N–(C@O)–N), 169.5 (C–(C@O)–
N); MS (EI, 70 eV): m=z 384.1, 329.1, 285.1, 255.0,
230.1, 213.1, 185.1, 154.1, 136.1. Anal. Calc. for
FeC20H28O2N2 (384.29): C, 62.51; H, 7.34; N, 7.29.
Found: C, 62.76; H, 7.56; N, 7.20.
N-tert-butyl-N0-ethyl-N-ferrocenoylurea (6) and N0-
tert-butyl-N-ethyl-N-ferrocenoylurea (7). 1 mmol (230
mg) 1 and 1 mmol (126 mg) N-tert-butyl-N0-ethylcar-
Table 3
Crystal data for 3 and 4
Compound
3
4
Empirical formula
Formula weight
Temperature (°C)
Wavelength (pm)
Crystal system
Space group
C18H24N2O2Fe
356.24
C26H24N2O2Fe
452.32
25
)60
71.073
71.073
Triclinic
Monoclinic
P21/c
P1
Unit cell dimensions
a ¼ 1406:77ð8Þ pm
b ¼ 1419:02ð10Þ pm
c ¼ 1828:18ð12Þ pm
a ¼ 90°
a ¼ 1030:97ð11Þ pm
b ¼ 1062:04ð11Þ pm
c ¼ 1150:18ð11Þ pm
a ¼ 69:083ð11Þ°
b ¼ 79:927ð12Þ°
c ¼ 65:890ð11Þ°
1.07313(19)
2
b ¼ 99:990ð7Þ°
c ¼ 90°
Volume (nm3)
Z
3.5941(4)
8
1.317
Density (calculated) (g/cm3)
Absorption coefficient (mmꢀ1
F (0 0 0)
1.400
1.457
944
)
0.849
1504
Crystal size (mm3)
Theta range
Index ranges
0.2 ꢁ 0.2 ꢁ 0.4
2.47 to 25.79°
ꢀ17 6 h 6 16
0 6 k 6 17
0 6 l 6 22
6115
0.2 ꢁ 0.2 ꢁ 0.4
3.34 to 24.71°
ꢀ13 6 h 6 12
ꢀ12 6 k 6 12
ꢀ13 6 l 6 12
7025
Reflections collected
Independent reflections
Completeness
6115
to h ¼ 25:79°: 98.8%
3443
to h ¼ 24:71°: 93.7%
Refinement method
Data/restrains/parameters
Goodness-of-fit on F 2
Final R indices ½I P 2rðIÞꢂ
R indices (all data)
Full-matrix least-squares on F 2
6115/0/582
0.731
3443/0/376
1.073
R1 ¼ 0:0379; wR2 ¼ 0:0815
R1 ¼ 0:0694; wR2 ¼ 0:0856
0.452
R1 ¼ 0:0400; wR2 ¼ 0:0997
R1 ¼ 0:0460; wR2 ¼ 0:1028
0.707
ꢀ
Largest diffraction peak (eA)

B. Schetter, B. Speiser / Journal of Organometallic Chemistry 689 (2004) 1472–1480
1479
bodiimide were suspended in 10 ml dichloromethane by
application of ultrasound and stirred for 15 h at room
temperature. The resulting clear brown solution was
concentrated to 4 ml and separated by CC as above. The
first coloured fraction was 8 (6.4 mg, 4%), the second
was 7 and the third was 6. Fraction 3 was purified again
by CC to remove any traces of 7. The solvent was re-
moved by vacuum and the red (7) and orange (6) solids,
respectively, were kept for 20 h at 30 °C and 0.1 mbar to
remove any trace amounts of carbodiimide. Yields: 104
2H, Si–CH2), 1.18–1.25 (t, JHH ¼ 7:1 Hz, 9H, CH3),
1.67–1.70 (quint, JHH ¼ 7:6 Hz, 2H, CH2–CH2–CH2),
3.35–3.36 (q, JHH ¼ 5:5 Hz, 2H, NCH2), 3.78–3.83 (q,
JHH ¼ 6:8 Hz, 6H, O-CH2–CH3), 4.16 (s, 5H, c-C5H5),
4.28 (s, 2H, c-C–CH–CH–CH–CH), 4.65 (s, 2H, c-C–
CH–CH–CH–CH), 6.0 (t, JHH ¼ 5:3 Hz, 1H, NH); 13C
NMR (CDCl3): d 7.7 (Si–CH2), 18.2 (CH3), 23.1 (CH2–
CH2–CH2), 41.7 (NCH2), 58.3 (CH3–CH2–O), 68.0,
69.6, 70.1 (c-C5H5, c-C–CH–CH–CH–CH), 76.4 (c-C–
(CH)5), 169.9 (C@O); MS (EI, 70 eV): m=z 433.1, 388.1,
322.0, 280.0, 229.0, 213.0, 185.0. Anal. Calc. for Fe-
SiC20H31O4N (433.38): C, 55.43; H, 7.21; N, 3.23.
Found: C, 55.19; H, 7.20; N, 2.93.
X-ray crystallography: The crystals of 2–4 were ob-
tained by cooling down the solutions of 2, 3 or 4 from 50
to 10 °C within 2 days. The structures were solved by
direct methods using SHELXTL and refined by SHELX-97
[19]. All non-hydrogen atoms were refined anisotropi-
cally using full-matrix least squares to give the final R
values. Some data are presented in Table 3.
mg of 6 (29.2%), 123 mg of 7 (34.6%).
ꢀ1
~
6: Mp.: 173 °C; IR (Atr): m/cm 3270 (NH), 3150–
3000 (CH, ferrocene), 3000–2850 (CH), 1645 (C@O); 1H
NMR (CDCl3): d 0.93–0.97 (t, JHH ¼ 7:4 Hz, 3H,
NHCH2CH3), 1.52 (s, 9H, (C@O)2NC(CH3)3), 3.08–
3.15 (quint, JHH ¼ 7:1 Hz, 2H, NHCH2CH3), 4.19 (s,
5H, c-C5H5), 4.25–4.26 (t, JHH ¼ 1:8 Hz, 2H, c-C–CH–
CH–CH–CH), 4.73–4.74 (t, JHH ¼ 1:8 Hz, 2H, c-C–
CH–CH–CH–CH), 5.30 (t, JHH ¼ 5:8 Hz, 1H, NH); 13
C
NMR (CDCl3):
d
13.9 (NH–CH2–CH3), 28.6
((C@O)2NC–(CH3)3), 36.2 (NHCH2–CH3), 48.9 (N–C–
(CH3)3), 69.9, 70.0, 70.1 (c-C5H5, c-C–CH–CH–CH–
CH), 77.9 (c-C–(CH)4), 155.5 (C–(C@O)–N), 169.2
(C–(C@O)–N); MS (FAB): m=z 356.1, 301.1, 285.1,
255.0, 230.1, 213.0, 185.1, 154.1, 136.1. Anal. Calc. for
FeC18H24O2N2 (356.24): C, 60.69; H, 6.79; N, 7.86.
Found: C, 60.33; H, 6.59; N, 7.86.ꢀ1
Crystallographic data for the structural analyses re-
ported in this paper have been deposited with the
Cambridge Crystallographic Data Centre CCDC and
have been allocated the deposition numbers 223438 (3)
and 223439 (4). Copies of this information can be ob-
tained free of charge from: The Director, CCDC, 12
Union Road, Cambridge CB2 1EZ, UK (Fax: (int code)
+44(1223)336-033, E-mail: deposit@ccdc.cam.ac.uk, or
www: http://www.ccdc.cam.ac.uk).
~
7: Mp.: 101 °C; IR (Atr): m/cm 3271 (NH), 3150–
3000 (CH, Ferrocene), 3000–2850 (CH), 1699 (C@O);
1H NMR (CDCl3): d 1.20–1.24 (t, JHH ¼ 6:9 Hz, 3H,
(C@O)2N–CH2CH3), 1.37 (s, 9H, NHC(CH3)3), 3.94–
4.00 (q, JHH ¼ 7:1 Hz, 2H, (C@O)2N–CH2CH3), 4.23 (s,
5H, c-C5H5), 4.38 (s, 2H, c-C–CH–CH–CH–CH), 4.69
(s, 2H, c-C–CH–CH–CH–CH), 9.00 (s, 1H, NH); 13C
NMR (CDCl3): d 15.7 ((C@O)2N–CH2CH3), 28.8
Chemical shift calculations: ACD/ChemSketch
(Advanced Chemistry Development, Inc.).
Acknowledgements
(NHC(CH3)3),
40.4
((C@O)2N–CH2CH3),
50.8
€
The authors wish to thank Markus Strobele, Institut
(NHC(CH3)3), 70.1, 70.3, 70.8 (c-C5H5, c-C–CH–CH–
CH–CH), 78.3 (c-C–(CH)4), 153.3 (C–(C@O)–N), 174.8
(C–(C@O)–N); MS (FAB): m=z 356.0, 301.0, 257.0,
213.0, 185.1, 136.1, 121.1. Anal. Calc. for
FeC18H24O2N2 (356.24): C, 60.69; H, 6.79; N, 7.86.
Found: C, 60.41; H, 6.45; N, 7.54.
€
€
€
fur Anorganische Chemie, Universitat Tubingen, for
collecting the X-ray data sets for compounds 2–4, and
Klaus Eichele for discussion of the analysis results. We
€
also acknowledge Michael Barth, Jorg Bauer and Steffen
Weik for the provision of and technical help with the
microwave heater and the IR-spectrometer.
N-(3-Triethoxysilyl)-propylferrocenecarboxamide (17).
2 mmol (460 mg) 1 and 2.2 mmol (420 mg) N-(3-
dimethylaminopropyl)-N0-ethylcarbodiimidehydrochlo-
ride were suspended in 20 ml dichloromethane by
application of ultrasound and stirred for 4 h at room
temperature. Then the clear brown solution was con-
centrated to 8 ml and separated by CC with silica gel as
the stationary phase and a mixture of dichloromethane
and acetone ratio 1/8 as the eluent. After removal of the
solvent by vacuum an orange solid was obtained. Yield:
323 mg ¼ 37%; Mp.: 80 °C, above slow thermal de-
composition; IR (Atr): ~m/cmꢀ1 3300 (NH), 3130–3050
(CH, ferrocene), 3000–2850 (CH), 1625 (C@O), 1520,
References
[1] P. Nuhn, Naturstoffchemie, S. Hirtzel, Stuttgart, 1997, pp. 157–
158.
[2] H.H. Lau, H. Hart, J. Org. Chem. 24 (1959) 280–281.
[3] E.M. Acton, R.M. Silverstein, J. Org. Chem. 24 (1959) 1487–1490.
[4] A.F. Neto, J. Miller, V.F. de Andrade, S.Y. Fujimoto, M.M. de
Freitas Alfonso, F.C. Archanjo, V.A. Darin, M.L. Andrade e
Silva, A.D.L. Borges, G. Del Ponte, Z. Anorg. Allg. Chem. 628
(2002) 209–216.
[5] Q.M. Wang, R.Q. Huang, J. Chem. Res. (S) (2001) 246–247.
[6] G. Cooke, A. Radhi, N. Boden, R.J. Bushby, Z. Lu, S. Brown,
S.L. Heath, Tetrahedron 56 (2000) 3385–3390.
1
1050; H NMR (CDCl3): d 0.65–0.68 (t, JHH ¼ 7:8 Hz,
[7] D. Astruc, Pure Appl. Chem. 75 (2003) 461–481.

1480
B. Schetter, B. Speiser / Journal of Organometallic Chemistry 689 (2004) 1472–1480
[8] B. Pradines, T. Fusai, W. Daries, V. Laloge, C. Rogier, P. Millet,
E. Panconi, M. Kombila, D. Parzy, J. Antimicrob. Chemoth. 48
(2001) 179–184.
[14] I. Yavari, J.D. Roberts, Org. Chem. 43 (1978) 4689–4693.
[15] K. Albert, G. Fischer, F. Novak, B. Schetter, B. Speiser,
unpublished results.
[9] R. Kovjazin, T. Eldar, M. Patya, A. Vanichkin, H.M. Lander, A.
Novogrodsky, FASEB J. 17 (2003) 467–469.
[16] E.J. Kupchik, R.J. Kiesel, J. Org. Chem. 31 (1966) 456–461.
[17] R.L. Cox, T.W. Schneider, M.D. Koppang, Anal. Chim. Acta 262
(1992) 145–159.
[10] H.S. Mandal, H.-B. Kraatz, J. Organomet. Chem. 674(2003)32–37.
[11] A. Mandelbaum, M. Cais, Tetrahedron Lett. (1964) 3847–3852.
[18] M. Oberhoff, L. Duda, J. Karl, R. Mohr, G. Erker, R. Frohlich,
M. Grehl, Organometallics 15 (1996) 4005–4011.
[19] G.M. Sheldrick, Structure Determination and Refinement with
ꢁ
ꢁ ꢁ
[12] V. Rapic, N. Filipovic-Marinic, Org. Mass Spectrom. 20 (1985)
104–109.
[13] T. Tenforde, R.A. Fawwaz, N.K. Freeman, N. Castagnoli Jr., J.
Org. Chem. 37 (1972) 3372–3374.
€
the Program System SHELX-97, Universitat Gottingen,
€
1997.