Ferrocenoyl-Substituted Pyrimidine Nucleobases: An Experimental and Computational Study of Regioselective Acylation of Uracil, Thymine, and 5-Fluorouracil

Article abstract of DOI:10.1002/ejoc.201500647

Uracil, thymine, and 5-fluorouracil (5-FU) have been ferrocenoylated selectively at the N¹ position. Deprotonated pyrimidine nucleobases, prepared with sodium hydride (NaH) in N,N-dimethylformamide (DMF), reacted with either ferrocenoyl chloride (FcCOCl) or ferrocenoyl ethyl carbonate (FcCOOCOOEt), in DMF to give a single product. The regioselectivity of these reactions were analyzed in detail by using NMR spectroscopy and quantum chemical calculations. The ¹H and ¹⁹F NMR spectra of reaction mixtures, and ¹³C NMR and 2D NOESY spectra of products, confirmed the formation of the N¹-isomer only. The calculated energy barrier for acetylation at the N³-position is significantly higher (> 40 kJ/mol), which suggests that the analogous reaction at the N¹-position is kinetically controlled. The nucleophilic addition of pyrimidine bases to the carbonyl group of FcCOCl proceeds through a concerted S_N2-like mechanism with the absence of the generally assumed tetrahedral intermediate. NMR spectroscopy and quantum chemical calculations (DFT) were employed to rationalize the regioselectivity in the formation of ferrocenoyl-substituted pyrimidine nucleobases. A facile one-step method is reported for the synthesis of N¹-substituted derivatives.

Full text of DOI:10.1002/ejoc.201500647

FULL PAPER
DOI: 10.1002/ejoc.201500647
Ferrocenoyl-Substituted Pyrimidine Nucleobases: An Experimental and
Computational Study of Regioselective Acylation of Uracil, Thymine, and
5-Fluorouracil
[a]
[b]
[b]
[b]
ˇ
´
´
´
ˇ
´
Jasmina Lapic, Valentina Havaic, Davor Sakic, Kresimir Sankovic,
[a]
[b]
´
ˇ
Senka Djakovic,* and Valerije Vrcek*
Keywords: Reaction mechanisms / Density functional calculations / Acylation / Regioselectivity / Nucleobases
Uracil, thymine, and 5-fluorouracil (5-FU) have been ferro-
13C NMR and 2D NOESY spectra of products, confirmed the
cenoylated selectively at the N1 position. Deprotonated pyr- formation of the N1-isomer only. The calculated energy bar-
imidine nucleobases, prepared with sodium hydride (NaH)
in N,N-dimethylformamide (DMF), reacted with either ferro-
cenoyl chloride (FcCOCl) or ferrocenoyl ethyl carbonate
rier for acetylation at the N3-position is significantly higher
(Ͼ 40 kJ/mol), which suggests that the analogous reaction
at the N1-position is kinetically controlled. The nucleophilic
(FcCOOCOOEt), in DMF to give a single product. The re- addition of pyrimidine bases to the carbonyl group of
gioselectivity of these reactions were analyzed in detail by FcCOCl proceeds through a concerted SN2-like mechanism
using NMR spectroscopy and quantum chemical calcula- with the absence of the generally assumed tetrahedral inter-
1
tions. The H and 19F NMR spectra of reaction mixtures, and
mediate.
cological significance.[3] It is expected that the ferrocenyl
group can modulate or potentiate the biological activity of
the conjugated pharmacophores.
Introduction
The ferrocenyl fragment has been coupled to a plethora
of biologically relevant systems, such as peptides, sugars,
drugs, and nucleosides.[1] In this respect, nucleobases re-
main attractive targets for ferrocenyl conjugation. Pyrimid-
ines and purines substituted by a ferrocenyl moiety present
interesting organometallic conjugates because their struc-
tures incorporate both biologically and electrochemically
active components.[2] A series of ferrocene-substituted nu-
cleobases has been described incorporating a different set
of linkers, but no carbonyl group has been used as a chemi-
cal spacer. Carbonyl spacers enable extended conjugation
involving ferrocenyl and nucleobase aromatic moieties, and
the electron delocalization can be tuned by inserting sub-
stituents at the ferrocene ring system.
To replace the sugar part of pyrimidine nucleosides with
a ferrocene moiety, the parent nucleobases should be substi-
tuted selectively at the N1-position. In this report, synthetic
procedures for the preparation of N1-ferrocenoyl-substi-
tuted pyrimidine bases (thymine, uracil, and 5-FU) have
been developed. No protection of the N3-position is needed
to afford the N1-isomer as a single product. The regioselec-
tive N1-substitution was confirmed by NMR spectroscopic
analysis, and quantum chemical calculations were per-
formed to rationalize the kinetic preference for the N1-acyl-
ation reaction.
In this work, the Fc–C=O fragment has been linked to
“standard” pyrimidine bases uracil and thymine. In ad-
dition, the ferrocenyl moiety is attached to 5-fluorouracil
(5-FU), which is a nucleobase analogue of known pharma-
Results and Discussion
To prepare N1-ferrocenoyl-substituted uracil, thymine, or
5-fluorouracil, several bases (K2CO3, TEA, or NaH) were
probed as deprotonating agents for pyrimidines. Attempts
at the preparation with K2CO3 or TEA were less effective
in terms of both yield and efficacy. By using NaH as a de-
protonating agent, exclusive formation of the N1-isomer in
high yield was observed. The acidity of N1 and N3 protons
in uracil (or thymine) is essentially identical in solution,[4–6]
and therefore cannot account for the observed regioselectiv-
ity. Herewith, we employ NMR spectroscopy and quantum
chemical methods to rationalize the mechanism underlying
the regioselectivity of the reaction between deprotonated
nucleobases and FcCOCl (or FcCOOCOOEt) (Scheme 1).
[a] Faculty of Food Technology and Biotechnology, University of
Zagreb,
Pierottijeva 6, Zagreb, Croatia
E-mail: sdjakov@pbf.hr
http://www.pbf.unizg.hr/en/departments/department_of_
chemistry_and_biochemistry/laboratory_for_organic_
chemistry/senka_djakovic
[b] Faculty of Pharmacy and Biochemistry, University of Zagreb,
Ante Kovacˇic´a 1, Zagreb, Croatia
E-mail: vvrcek@pharma.hr
http://www.pharma.unizg.hr/en/about-us/staff/
valerije-vrcek,391.html
Supporting information for this article is available on the
WWW under http://dx.doi.org/10.1002/ejoc.201500647.
5424
© 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2015, 5424–5431

Ferrocenoyl-Substituted Pyrimidine Nucleobases
followed by the IRC procedure, which suggests a concerted
SN2-type mechanism.
The same mechanism is operative for the reaction be-
tween N3-deprotonated uracil (1N3, R = H) and FcCOCl,
in which a single transition state (TSN3) exists on the corre-
sponding energy surface. No tetrahedral intermediate, typi-
cal of a nucleophilic addition–elimination mechanism, was
found in either case. All starting structures of conceivable
intermediates converged, during geometry optimization, to
local minima at the reactant or product side of the reaction.
Therefore, the reaction between uracil and FcCOCl follows
a one-step mechanism in which N–C bond formation and
C–Cl bond cleavage occur simultaneously. This supports
earlier theoretical and experimental results that corroborate
that nucleophilic addition to the carbonyl group of acid
chlorides proceeds by a concerted SN2-like mechanism with
the absence of the generally assumed tetrahedral intermedi-
ate.[7–9]
The one-step mechanism is operative in the reaction be-
tween other pyrimidine bases, thymine (1N1, R = Me) or 5-
fluorouracil (1N1, R = F), and FcCOCl, as evidenced by
our computational results. However, if a poor leaving group
is incorporated in the acylating agent, then the “classical”
two-step mechanism becomes operative. According to the
computational results, in the reaction between neutral ura-
cil and ferrocenoyl ester (FcCOOCH3) and ferrocenecarb-
oxylic acid (FcCOOH), the tetrahedral intermediate was
easily located in both cases (see Figure S1 in the Supporting
Information). This finding reveals the stepwise mechanism
for nucleophilic substitution at the carbonyl carbon. Again,
this is in agreement with experimental evidence of an ad-
dition–elimination pathway, which is a preferred reaction
channel when the acylating agent contains poor leaving
groups.[10]
Scheme 1. Regioselective N1-acylation of pyrimidine nucleobases 1
(uracil, R = H; thymine, R = Me; 5-fluorouracil, R = F) with ferro-
cenoyl chloride (X = Cl) or ferrocenoyl ethyl carbonate (X =
OCOOCH2CH3). Preparation of ferrocenoyl chloride (FcCOCl)
and ferrocenoyl ethyl carbonate (FcCOOCOOEt) from ferro-
cenecarboxylic acid (FcCOOH) is presented in the inset box.
The Reaction Mechanism
To investigate the mechanism of the reaction between
pyrimidine bases and FcCOCl, DFT calculations were per-
formed. Stationary points were searched at the correspond-
ing potential energy surfaces, and characterized as either
minima (NImag = 0) or first-order saddle points (NImag =
1). For the reaction between N1-deprotonated uracil (1N1
,
R = H) and FcCOCl, only one transition state (TSN1) was
located (Figure 1). This transition structure connects reac-
tants (1N1 and FcCOCl) and the product (2N1, R = H), as
Regioselectivity of the Acylation Reaction
In the following section, we will investigate the regiose-
lectivity of the reaction between pyrimidine bases and
FcCOCl. Reports on regioselective acylations of uracil and
thymine are rather scarce.[11,12] Most of the reactions be-
tween unprotected uracil/thymine derivatives and acylating
agents (acylchlorides, esters, or anhydrides) result in the for-
mation of a mixture of N1- and N3-substituted products.[13]
It has been found that the N-acylation position depends on
the reaction temperature, base, catalyst, and acylation agent
employed.[14,15]
In case of the reaction between thymine and FcCOCl,
only one product was observed spectroscopically (Figure 2).
The reaction was followed by the sampling method in which
aliquots (0.2 mL) were taken directly from the reaction mix-
ture at selected time points. The 1H NMR spectrum of each
aliquot was measured in [D6]DMSO. Along with the disap-
pearance of signals of the reactant (the ring protons for
FcCOCl in the range of 4–5 ppm), a new set of proton sig-
nals of the product emerged (Figure 3). According to de-
tailed 1H, 13C, and 2D NOESY NMR analysis of the prod-
Figure 1. Transition-state structures TSN1 and TSN3 for the reac-
tion of N1- or N3-deprotonated uracil, respectively, with ferro-
cenoyl chloride (FcCOCl), optimized at the B3LYP/6-31G(d) level.
All distances are in angstroms.
Eur. J. Org. Chem. 2015, 5424–5431
© 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
5425

S. Djakovic´, V. Vrcˇek et al.
FULL PAPER
uct obtained (see below), we conclude that a N1-ferrocenoyl –178.3 ppm, which corresponds to deprotonated 5-fluoro-
conjugate (2N1, R = Me) was formed.
uracil, disappears in time, while a new signal at –166 ppm
shows up as the only signal in the spectrum. Again, the
detailed 2D NMR spectra (see below) confirmed that the
product formed was the N1-isomer (2N1, R = F).
Quantum chemical calculations have been employed to
rationalize the preferential reaction of ferrocenoyl chloride
with the N1-nitrogen atom in uracil. At the B3LYP/6-
31+G(d,p) level, the calculated Gibbs free energy of acti-
vation for the reaction at the N3-position is approximately
40 kJ/mol higher than the barrier for the corresponding re-
action at the N1-position (Scheme 2 and Table 1).
The calculated energy of N1-ferrocenyl-substituted uracil
(2N1, RЈ = Fc in Scheme 2) is similar to the energy of its
N3-substituted counterpart (2N3, RЈ = Fc). Therefore, the
N1-acylation reaction 1N1 Ǟ2N1 is not thermodynamically,
but kinetically favored. The comparative computational
study has been performed for the reaction between other
acyl chlorides, benzoyl chloride (RЈ = Ph) and acetyl chlor-
ide (RЈ = Me), and other pyrimidine bases (thymine and 5-
fluorouracil). In each case, the barrier for the N1-reaction
is lower than the corresponding barrier for the N3-reaction
(Table 1).
The results presented above confirm our hypothesis that
regioselectivity in the reaction between the acylating agent
(FcCOCl) and deprotonated pyrimidine nucleobases is ki-
netically controlled. Both reactions 1Ǟ2N1 and 1Ǟ2N3
(see Scheme 1, where R = H, Me, or F) are strongly exer-
gonic (Table 1). The calculated Gibbs free energies (ΔGr) for
Figure 2. B3LYP/6-31G(d) optimized N1- and N3-substituted prod-
ucts (2N1 and 2N3, resp.) of the reaction between thymine and ferro-
1
cenoyl chloride (FcCOCl). The double arrows indicate key H-1H
NOESY correlations observed for 2N1 (see the Supporting Infor-
mation) and the corresponding interatomic distances [Å]. The
dashed lines indicate the calculated distances [Å] between the 6-H
atom and selected ferrocenyl protons in 2N3
.
The same conclusion comes from 19F NMR spectro-
scopic measurements of the reaction between 5-fluorouracil
and FcCOCl (the small inset in Figure 3). The peak at
1
Figure 3. H NMR (400 MHz) spectra (selected resonances) of the reaction mixture aliquots (thymine/NaH + FcCOCl in DMF) taken
at several time points. The stack-plot of spectra (with an offset included) were recorded over 20 min. The small inset shows 19F NMR
(376 MHz) spectra of the reaction mixture (5-FU/NaH + FcCOCl in DMF) recorded over 15 min. All spectra were measured in [D6]-
DMSO at 25 °C.
5426
www.eurjoc.org
© 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2015, 5424–5431

Ferrocenoyl-Substituted Pyrimidine Nucleobases
Scheme 2. Schematic energy profile [B3LYP/6-31+G(d,p) + ΔGsolv] for the reaction between deprotonated uracil (at N1- or N3-position)
and a series of acylchlorides (RЈ = Me, Ph, or Fc).
Table 1. Relative[a] Gibbs free energy of the reaction (ΔGr) and acti- N3-ferrocenoylated products are almost isoenergetic pro-
vation barrier (ΔG#) (in kJ/mol, at 298.15 K) for reactions between
cesses (ΔΔGr = 0.7 kJ/mol, for the reaction uracil +
deprotonated nucleobases (at the N1- or N3-position) and acyl
FcCOCl). This shows that the calculated thermochemical
chlorides, calculated at the CPCM/UFF//B3LYP/6-31+G(d,p)/
data, for parallel reactions at N1- and N3-positions, cannot
explain the experimentally observed regioselectivity. As
stated above, the regioselectivity of this reaction is mainly
kinetically controlled.
SDD level of theory.[b]
Acylation reac- Uracil
tion at selected
Thymine
5-Fluorouracil
ΔG#
–89.3 75.5
position
ΔGr
ΔG#
ΔGr
ΔG#
ΔGr
Base +
N1
–116.4 39.4
–95.1 78.3
–98.4 55.1
–95.1 96.7
–84.3 87.8
–106.4 63.2
MeCOCl N3
–83.4 104.5 –65.4 93.3
–84.0 80.3 –66.0 77.2
–86.2 124.5 –67.9 115.8
–71.4 92.1 –53.4 97.6
Base +
PhCOCl N3
Base +
N1
FcCOCl N3
N1
NMR Analysis of the Reaction Mixture
The 13C NMR spectra of previously reported acyl-substi-
tuted uracil and thymine reveal that N1- and N3-acylated
derivatives can be distinguished on the basis of the chemical
shifts of the corresponding C5 carbon resonance.[11] The C5
resonance signals of uracil and thymine are shifted approxi-
mately 3 ppm downfield with the introduction of the N1-
–83.6 134.9 –69.1 164.0 –50.9 153.9
[a] Sum of energies for the N1-deprotonated base and acyl chloride
is set to zero. [b] All geometries optimized at the B3LYP/6-31G(d)
level, and bulk solvent effects have been calculated for DMF (ε =
37.22).
all reactions are between –50 and –120 kJ/mol. In each case, benzoyl substituent (Table 2). No such effect was observed
several product conformers are located (see the Supporting with the introduction of the benzoyl group at the N3-posi-
Information), and only the most stable are presented here- tion.[11] In our case, the reaction between uracil/thymine
with (Table 1). In reactions between acetyl chloride and nu- and FcCOCl results in the ferrocenoylated product in which
cleobases the N1-acetylated products are calculated as more the C5 signal is also shifted approximately 3 ppm downfield
stable (Ͼ 20 kJ/mol) than their N3-counterparts. However, compared with the corresponding signal of the parent nu-
in reactions between benzoyl or ferrocenoyl chloride and cleobases (Table 2). The introduction of the ferrocenoyl
nucleobases, the stability of N1- and N3-isomers is very sim- group, accompanied by the downfield shift of the C5 signal,
ilar. It comes out that the formation reactions of N1- and is therefore indicative of N1-substitution.
Eur. J. Org. Chem. 2015, 5424–5431
© 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
5427

S. Djakovic´, V. Vrcˇek et al.
FULL PAPER
Table 2. Experimental[a] and computational [GIAO-CPCM/UFF// NOESY correlations could also be a valuable tool to detect
B3LYP/6-311G(d,p)//B3LYP/6-31G(d)/ Wachter-f method][b] NMR
chemical shifts[c] for the C5 signal (and for the Fe atom, where
appropriate) in pyrimidine bases and their N1- and N3-acylated
derivatives (Bz = benzoyl, Fc = ferrocenoyl).
N1-acylation products. In the case of N1-ferrocenoyl-substi-
tuted thymine (2N1, Figure 2), the NOESY experiment pro-
duces crosspeaks (see Figure S3 in the Supporting Infor-
mation) between C6-H and ferrocenyl protons that are close
in space (the calculated distances of ca. 2.5 Å). It is ex-
pected that no cross-relaxation between C6-H and ferro-
cenyl protons is possible in the N3-isomer (2N3, Figure 2),
because the corresponding interatomic distances are larger
than 5 Å. The same cross-polarization between C6-H and
ferrocenyl protons is observed for uracil and 5-fluorouracil
(see Figure S2 and S4 in the Supporting Information),
which results in cross peaks in the corresponding 2D
NOESY spectra. To conclude, a detailed analysis of both
1H and 13C NMR spectra reveals that N1-ferrocenoylated
products are formed preferentially in the reaction between
the deprotonated pyrimidine nucleobase (using NaH as a
deprotonating agent) and FcCOCl in dimethylformamide.
C5
δexp (δcalc
C5
Fe
δcalc
[d]
)
Δδexp (Δδcalc
)
Uracil
101.1 (101.8)
103.7 (105.0)
100.1 (102.2)
102.9 (104.0)
n. a. (102.3)
107.8 (111.6)
111.5 (115.4)
107.9 (112.2)
111.6 (114.6)
n. a. (112.1)
139.9 (146.1)
–
–
–
–
N1-Bz-uracil
N3-Bz-uracil
N1-Fc-uracil
N3-Fc-uracil
Thymine
+2.6 (+3.2)
–1.0 (+0.4)
+1.8 (+2.2)
n. a. (+0.5)
–
+3.7 (+3.8)
+0.1 (+0.6)
+3.8 (+3.0)
n. a. (+0.5)
–
1970.4
1984.8
–
–
–
1919.0
1935.8
–
1935.5
1952.9
N1-Bz-thymine
N3-Bz-thymine
N1-Fc-thymine
N3-Fc-thymine
5-Fluorouracil
N1-Fc-5-fluorouracil 140.9 (147.2)
N3-Fc-5-fluorouracil n. a. (145.9)
+1.0 (+1.1)
n. a. (–0.2)
[a] Experimental NMR spectra measured in [D6]DMSO. [b] Geo-
metries were optimized and chemical shifts calculated in the model
solvent [D6]DMSO. [c] 13C NMR chemical shifts are given in ppm
downfield from tetramethysilane (TMS; 0 ppm), and 57Fe NMR
chemical shifts are relative to ferrocene (1532 ppm). [d] Difference
between chemical shifts for C5 signals δC5(acylated base) –
δC5(base).
Conclusions
A facile and selective N1-ferrocenoylation of pyrimidine
nucleobases (uracil, thymine, and 5-fluorouracil) is re-
To supplement the above experimental observation, we ported. The synthetic procedure used to obtain only the
performed GIAO-NMR calculations for N1- and N3-acyl- N1-regioisomer does not require any protection of the N3-
substituted structures and for the parent pyrimidine bases position in the nucleobase. Out of three deprotonating
(uracil and thymine). In agreement with experimental NMR agents (K2CO3, TEA, and NaH), only sodium hydride ap-
spectroscopic data, a difference of 3.2 ppm was calculated peared effective to assist the acetylation at the N1-position.
(Table 2) for the C5 signal in uracil (δcalc = 101.8 ppm) and
1D and 2D NMR spectral evidence of regioselective N1-
in N1-benzoyl-substituted uracil (δcalc = 105.0 ppm). No ferrocenoylation of pyrimidine nucleobases has been pre-
significant difference between chemical shifts (C5 signal) sented. Reactions were followed by 1H and 19F NMR spec-
was calculated in uracil and N3-benzoyl-substituted uracil troscopy and only one product was detected. The structure
(δcalc = 102.2 ppm). A similar result was found for thymine and position of the substitution were confirmed by 13C
and its N1- and N3-benzoylated products (Table 2). This NMR spectroscopy. It is unambiguously identified as the
supports earlier claims that 13C NMR spectroscopy can be N1-regioisomer. We have shown that experimental NMR
a useful tool to differentiate N1- vs. N3-acylation products. spectroscopic data analysis coupled with GIAO-NMR cal-
In the case of 5-fluorouracil, no significant chemical shift culations is a valuable tool to differentiate between N1- and
difference for the C5 signal in the parent 5-FU and N1- N3-acylated products.
ferrocenoyl-substituted 5-FU was observed either experi-
The relative acidity of N1–H vs. N3–H bonds in nucleo-
mentally or theoretically. This is in line with the recent bases, the calculated thermochemical data of the reaction
NMR study on 5-FU and its N1-analogues.[16] However, ac- products (2N1 vs. 2N3), and energy barriers for the two par-
cording to NMR analysis of the reaction mixture (the small allel reactions, 1Ǟ2N1 vs. 1Ǟ2N3, have been considered to
inset in Figure 3) we are confident that the regioselectivity explain the rationale behind the observed regioselectivity. It
in the reaction between 5-FU and FcCOCl follows the same comes out that the regioselectivity is governed by kinetic
pattern as that described for uracil and thymine.
factors only. The transition-state structures for N1-ferro-
Although the 57Fe nucleus has a large chemical shift cenoylation of nucleobases (TSN1) were calculated to be
range[17] and 57Fe NMR shift values can be very susceptible much more stable (Ͼ 40 kJ/mol) than the corresponding
to slight changes in geometry;[18] we found less than 18 ppm structures for N3-ferrocenoylation (TSN3) in each case.
relative difference between the calculated 57Fe NMR chemi-
According to our computational results at the CPCM/
cal shifts in N1- and N3-ferrocenoylated nucleobases B3LYP/6-31+G(d,p)//B3LYP/6-31G(d) level of theory, the
(Table 2). This suggests that 57Fe NMR is probably not a reaction between pyrimidine nucleobase and ferrocenoyl
sensitive tool to probe the preferred site of ferrocenoylation chloride follows a concerted SN2-like mechanism. The nu-
in pyrimidine bases.
cleophilic attack of the deprotonated base on the carbonyl
In addition to downfield shifts observed for the C5 signal group of FcCOCl is a one-step process, which implies that
in N1-acylated pyrimidine bases (ca. 3 ppm as compared no tetrahedral intermediate exists on the corresponding po-
with unsubstituted nucleobases), we found that 1H-1H tential energy surface.
5428
www.eurjoc.org
© 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2015, 5424–5431

Ferrocenoyl-Substituted Pyrimidine Nucleobases
Computational Methods
ΔΔGsol between the two Gibbs energy variations ΔG#sol and
ΔGsol for bimolecular reaction A + BǞA–B is:
Geometries were fully optimized at the B3LYP level of
theory, as implemented in the Gaussian 09 software pack-
age.[19] The basis set for optimization was standard Pople’s
6-31G(d) on non-metal atom centers, whereas the Stutt-
gart–Dresden–Bonn (SDD)[20] basis set with effective core
potential (ECP) was used for Fe, similar to previous stud-
ies.[21] Improved, single-point energies were calculated at the
B3LYP level with 6-31+G(d,p) basis set for non-metal cen-
ters, while SDD/ECP was used for Fe. This and comparable
DFT levels have proven quite successful for transition-metal
compounds and are well suited for the description of struc-
tures, energies, vibrational frequencies, and other proper-
ties.[22] Harmonic frequencies were computed from analyti-
cal second derivatives.
Gibbs energies of solvation were determined by using the
CPCM continuum solvation model at the B3LYP/6-
31+G(d,p)//B3LYP/6-31G(d) level, with the UFF atomic
radii and electrostatic scaling factor (alpha value) set to 1.1
for all atoms (default values in Gaussian09).[23–25] The UFF
cavities are selected to ensure that solvent spheres are
placed around light and heavy atoms. The solvent relative
permittivity of ε = 37.22 [N,N-dimethylformamide (DMF)]
was used.
ΔΔGsol = ΔG# – ΔGsol = RTln[(vc(A)vc(B)/vc(A-B)] – RTln[kBT/p]
sol
where vc corresponds to the cavity volume, and kB, T, and
p, correspond to the Boltzmann constant, temperature, and
pressure, respectively. The cavity volume for each species
was obtained from CPCM calculation.
Experimental Section
Methods and Materials: The syntheses were carried out under an
argon atmosphere in anhydrous solvents. Melting points were de-
termined with a Büchi apparatus. The IR spectra were recorded for
KBr pellets or CH2Cl2 solutions with a Bomem MB100 Mid FT
IR spectrophotometer. The mass spectra were acquired with a 4800
MALDI TOF/TOF-MS Analyzer. The 1H, 13C, and 19F NMR
spectra of [D6]DMSO or CDCl3 solutions were recorded with a
Varian INOVA 400 spectrometer. The spectrometer operated at
399.6 MHz (1H), 375.9 MHz (19F), and 100.5 MHz (13C). Chemical
shifts in the 1H NMR and 13C NMR spectra are expressed in parts
per million (ppm) vs. TMS as the external standard, and 19F chemi-
cal shifts are referenced to CFCl3 as the external standard.
Products were purified by column chromatography (Fluka, silica
gel, 90 Å, 70–230 mesh) using the CH2Cl2/acetone mixtures (10:1).
N,N-Diphenylferrocenecarboxamide was prepared in 54% yield
starting from ferrocene and N,N-diphenylcarbamoyl chloride.[31]
Alkaline hydrolysis of N,N-diphenylcarbamoyl chloride gave ferro-
cenecarboxylic acid in 80%.[32]
Magnetic shieldings (σ) were computed at the B3LYP
level for the CPCM/B3LYP/6-31G(d) geometries optimized
in dimethyl sulfoxide (DMSO, ε = 46.83), employing
GIAOs (gauge including atomic orbitals)[26] with the aug-
mented Wachters basis[27] on Fe (8s7p4d), and 6-311G(d,p)
basis on all other atoms. Chemical shifts were calculated
relative to tetramethylsilane (TMS; 13C and 1H magnetic
shieldings are 181.29 and 31.86 at the same level) or ferro-
Synthesis of Ferrocenoyl Chloride: To a suspension of ferro-
cenecarboxylic acid (300 mg, 1.3 mmol) and freshly distilled oxalyl
chloride (274 mL, 3.13 mmol) in anhydrous CH2Cl2 (5 mL), one
drop of pyridine was added. The reaction mixture was heated to
reflux for 2 h and the solvents were evaporated in vacuo to dryness
to give a dark residue. The crude product was repeatedly extracted
at 80 °C for 10 min with petroleum ether to give red FcCOCl crys-
1
cene (57Fe, 13C, and H magnetic shieldings are –5299.35,
110.33, and 28.11), using the experimental δ(57Fe) value of
ferrocene (1532 ppm).[28] All NMR shieldings were Boltz-
mann-averaged by selection of the most stable conformers
(see the Supporting Information).
tals (275 mg, 85%).[33] IR (CH Cl ): ν = 2958 (w, C–H, Fc), 1755
˜
2
2
(s, C=O) cm–1. 1H NMR (400 MHz, CDCl3, 25 °C): δ = 4.9 (t,
3
3JH,H = 1.9 Hz, 2 H, H-αЈ), 4.6 (t, JH,H = 1.9 Hz, 2 H, H-βЈ), 4.4
(s, 5 H, Cp) ppm. 13C NMR (100 MHz, CDCl3, 25 °C): δ = 178.3
(CO), 73.3 (Ci), 71.7 (Cp), 71.2 (C-βЈ), 70.4 (C-αЈ) ppm.
IRC calculations (intrinsic reaction coordinate as im-
plemented in Gaussian 09) were performed at the corre-
sponding level to identify the minima connected through
the transition state. The initial geometries used were that of
the corresponding transition structures, and the paths were
followed in both directions from that point. This method
verified that a given transition state structure indeed con-
nected the presumed energy minimum structures.[29]
To account for the entropic effect of the presence of sol-
vent molecules around a solute, the cell model presented by
Ardura et al. was used.[30] This model is proposed in order
to explicitly evaluate the effect of the loss of translation
degrees of freedom in solution on the Gibbs activation en-
ergy in bimolecular (or higher order of molecularity) reac-
tions. It has been shown that the standard implementation
of the continuum model is not capable of adequately esti-
mating the increase in Gibbs energy corresponding to the
constriction of the translation motion of the species along
the reaction coordinate when passing from the gas phase
Synthesis of Ferrocenoyl Ethyl Carbonate: A suspension of ferro-
cenecarboxylic acid (500 mg, 2.175 mmol) in water (0.41 mL) was
dissolved by the addition of acetone (7.4 mL) and then cooled to
0 °C. Triethylamine (7.4 μL) in acetone (4.5 mL) is added dropwise,
followed by the addition of ethyl chloroformate (5.8 μL) in acetone
(1.15 mL). After stirring for 30 min, the reaction mixture was evap-
orated, and the ferrocenoyl ethyl carbonate crude product (orange
solid) was used for reactions with pyrimidine bases. IR (CH2Cl2):
ν = 2932 (w, C–H, Fc), 1770 (s, C=O), 1713 (s, C=O), 1675 (s,
˜
1
C=O), 1047 (s, C–O) cm–1. H NMR (400 MHz, CDCl3, 25 °C): δ
3
3
= 4.9 (t, JH,H = 1.8 Hz, 2 H, H-αЈ), 4.6 (t, JH,H = 1.8 Hz, 2 H,
3
H-βЈ), 4.4 (s, 5 H, Cp), 4.3 (q, JH,H = 7.2 Hz, 2 H, CH2), 1.4 (t,
3JH,H = 7.2 Hz, 3 H, CH3) ppm. 13C NMR (100 MHz, CDCl3,
25 °C): δ = 167.7 (FcCO), 161.8 (CO), 72.7 (C-αЈ), 70.8 (C-βЈ), 70.2
(Cp), 69.3 (Ci), 61.2 (CH2), 14.3 (CH3) ppm.
General Procedure for the Preparation of N1-Ferrocenoylated Pyrim-
idine Bases: Sodium hydride (NaH; 1.5 mmol), was added portion-
wise to pyrimidine bases (1 mmol) suspended in DMF (3 mL). Af-
to the solution. According to the cell model, the difference ter stirring at room temperature for 30 min, FcCOCl (1 mmol) (or
Eur. J. Org. Chem. 2015, 5424–5431
© 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
5429

S. Djakovic´, V. Vrcˇek et al.
FULL PAPER
ana-Douki, E. Mouray, T. Bousquet, S. Pellegrini, P. Grellier,
F. S. T. Ndouo, J. Lebibi, L. Pelinski, New J. Chem. 2011, 35,
2412–2415; c) S. I. Kirin, H.-B. Kraatz, N. Metzler-Nolte,
Chem. Soc. Rev. 2006, 35, 348–354; d) M. Patra, G. Gasser, M.
Wenzel, K. Merz, J. E. Bandow, N. Metzler-Nolte, Organome-
tallics 2010, 29, 4312–4319; e) J. Lapic´, S. Djakovic´, I. Kodrin,
Z. Mihalic´, M. Cetina, V. Rapic´, Eur. J. Org. Chem. 2010, 2512–
2524; f) B. Adhikari, R. Afrasiabi, H.-B. Kraatz, Organometal-
lics 2013, 32, 5899–5905; g) H.-B. Kraatz, J. Inorg. Organomet.
Polym. 2005, 15, 83–106; h) T. Moriuchi, T. Hirao, Acc. Chem.
Res. 2010, 43, 1040–1051.
a) A. N. Patwa, R. G. Gonnade, V. A. Kumar, M. M.
Bhadbhade, K. N. Ganesh, J. Org. Chem. 2010, 75, 8705–8708;
b) H. V. Nguyen, A. Sallustrau, J. Balzarini, M. R. Bedford,
J. C. Eden, N. Georgousi, N. J. Hodges, J. Kedge, Y. Mehellou,
C. Tselepis, J. H. R. Tucker, J. Med. Chem. 2014, 57, 5817–
5822; c) A. A. Simenel, G. A. Dokuchaeva, L. V. Snegur, A. N.
Rodionov, M. M. Ilyin, S. I. Zykova, L. A. Ostrovskaya, N. V.
Bluchterova, M. M. Fomina, V. A. Rikova, Appl. Organomet.
Chem. 2011, 25, 70–75; d) H. V. Nguyen, Z. Zhao, A. Sallus-
trau, S. L. Horswell, L. Male, A. Mulas, J. H. R. Tucker, Chem.
Commun. 2012, 48, 12165–12167; e) K. Kowalski, J. Skiba, L.
Oehninger, I. Ott, J. Solecka, A. Rajnisz, B. Therrien, Organo-
metallics 2013, 32, 5766–5773; f) A. Houlton, C. J. Isaac, A. E.
Gibson, B. R. Horrocks, W. Clegg, M. R. J. Elsegood, J. Chem.
Soc., Dalton Trans. 1999, 3229–3234; g) P. Brázdilová, M.
Vrábel, R. Pohl, H. Pivonˇková, L. Havran, M. Hocek, M.
Fojta, Chem. Eur. J. 2007, 13, 9527–9533.
D. B. Longley, D. P. Harkin, P. G. Johnston, Nature Rev. 2003,
3, 330–338.
M. A. Kurinovich, J. K. Lee, J. Am. Chem. Soc. 2000, 122,
6258–6262.
A. Zhachkina, J. K. Lee, J. Am. Chem. Soc. 2009, 131, 18376–
18385.
V. Verdolino, R. Cammi, B. H. Munk, H. B. Schlegel, J. Phys.
Chem. B 2008, 112, 16860–16873.
J. M. Fox, O. Dmitrenko, L. Liao, R. D. Bach, J. Org. Chem.
2004, 69, 7317–7328.
T. W. Bentley, G. Llewellyn, J. A. McAlister, J. Org. Chem.
1996, 61, 7927–7932.
F. Ruff, O. Farkas, J. Phys. Org. Chem. 2011, 24, 480–491.
J. March, in: Advanced Organic Chemistry, 4th ed.; Wiley: New
York, 1992, p. 330–335.
K. A. Cruickshank, J. Jiricny, C. B. Reese, Tetrahedron Lett.
1984, 25, 681–684.
M. N. S. Rad, S. Behrouz, Z. Asrari, A. Khalafi-Nezhad,
Monatsh. Chem. 2014, 145, 1933–1940.
S. A. Lahsasni, Nucleosides Nucleotides Nucleic Acids 2013, 32,
439–452.
T. Kametani, K. Kigasawa, M. Hiiragi, K. Wakisaka, S. Haga,
Y. Nagamatsu, H. Sugi, K. Fukawa, O. Irino, J. Med. Chem.
1980, 23, 1324–1329.
FcCOOCOOEt in selected cases) was added dropwise to the clear
solution. The mixture was stirred for 10–30 min and then neutral-
ized with 10% aqueous solution of citric acid and extracted with
CH2Cl2. The organic layer was washed with water and the solvents
were evaporated under vacuum. Subsequent purification by column
chromatography afforded N1-ferrocenoylated pyrimidine bases.
N1-Ferrocenoyluracil (2N1, R = H): Red-orange crystals (269.6 mg,
64.6% yield); m.p. Ͼ 200 °C. C15H12FeN2O3 (324.112): calcd. C
55.59, H 3.73, O 14.81, N 8.64; found C 55.61, H 3.75, O 14.84, N
8.67. IR (CH Cl ): ν = 3374 (w, NH), 3099 (w, CH, Fc), 1700 (s,
˜
2
2
[2]
C=O), 1633 (w, C=O) cm–1. 1H NMR (400 MHz, [D6]DMSO,
3
25 °C): δ = 11.5 (s, 1 H, N3-H), 8.0 (d, JH,H = 8.0 Hz, 1 H, C6-
3
3
H), 5.7 (d, JH,H = 8.0 Hz, 1 H, C5-H), 4.9 (t, JH,H = 2.0 Hz, 2
3
H, H-βЈ), 4.7 (t, JH,H = 2.0 Hz, 2 H, H-αЈ), 4.3 (s, 5 H, Cp) ppm.
13C NMR (100 MHz, [D6]DMSO, 25 °C): δ = 173.6 (FcCO), 163.8
(C4), 149.9 (C2), 141.3 (C6), 102.9 (C5), 74.1 (C-αЈ), 72.6 (Ci), 71.9
(C-βЈ), 71.1 (Cp) ppm. HRMS (MALDI-TOF/TOF): m/z calcd. for
C15H12O3N2Fe [M + H]+ 325.1203; found 325.1205.
N1-Ferrocenoylthymine (2N1, R = Me): Orange crystals (245.79 mg,
72.7% yield); m.p. Ͼ 200 °C. C16H14FeN2O3 (338.139): calcd. C
56.83, H 4.17, O 14.19, N 8.28; found C 56.85, H 4.18, O 14.21, N
8.29. IR (CH Cl ): ν = 3375 (w, NH), 2927 (w, CH, Fc), 1730 (s,
˜
2
2
C=O), 1700 (s, C=O) cm–1. 1H NMR (400 MHz, [D6]DMSO,
4
25 °C): δ = 11.5 (s, 1 H, N3-H), 7.9 (q, JH,H = 1.3 Hz, 1 H, C6-
3
3
H), 4.8 (t, JH,H = 2.2 Hz, 2 H, H-αЈ), 4.7 (t, JH,H = 2.2 Hz, 2 H,
[3]
[4]
[5]
[6]
[7]
[8]
H-βЈ), 4.3 (s, 5 H, Cp), 1.8 (d, 4JH,H = 1.3 Hz, 3 H, CH3) ppm. 13
C
NMR (100 MHz, [D6]DMSO, 25 °C): δ = 173.7 (FcCO), 164.6
(C4), 149.9 (C2), 136.6 (C6), 111.6 (C5), 73.8 (C-αЈ), 72.9 (Ci), 71.9
(C-βЈ), 71.1 (Cp), 12.3 (CH3) ppm. HRMS (MALDI-TOF/TOF):
m/z calcd. for C16H14O3N2Fe [M + H]+ 339.1469; found 339.1470.
N1-Ferrocenoyl 5-Fluorouracil (2N1
,
R
=
F): Purple crystals
(203.04 mg, 59.4% yield); m.p.
Ͼ 200 °C. C15H11FFeN2O3
(342.108): calcd. C 52.66, H 3.24, O 14.03, N 8.19; found C 52.62,
H 3.22, O 14.00, N 8.16. IR (CH Cl ): ν = 3366 (w, NH), 2924 (w,
˜
2
2
C–H, Fc), 1720 (s, C=O), 1648 (s, C=O) cm–1. 1H NMR (400 MHz,
3
[D6]DMSO, 25 °C): δ = 12.0 (s, 1 H, N3-H), 8.4 (d, JF,H = 6.4 Hz,
[9]
[10]
3
3
1 H, C6-H), 4.9 (t, JH,H = 2.3 Hz, 2 H, H-αЈ), 4.7 (t, JH,H
=
2.3 Hz, 2 H, H-βЈ), 4.3 (s, 5 H, Cp) ppm. 13C NMR (100 MHz,
[D6]DMSO, 25 °C): δ = 172.7 (FcCO), 158.1 (d, JF,C = 27.1 Hz,
[11]
[12]
[13]
[14]
3
2
3
C4), 148.6 (C2), 140.9 (d, JF,C = 234.5 Hz, C5), 125.7 (d, JF,C
=
F
=
36.1 Hz, C6), 73.9 (C-αЈ), 72.5 (Ci), 71.9 (C-βЈ), 71.1 (Cp) ppm. 19
3
NMR (376 MHz, [D6]DMSO, 25 °C): δ = –167.4 (d, JH,F
6.4 Hz, C5-F) ppm. HRMS (MALDI-TOF/TOF): m/z calcd. for
C15H11FO3N2Fe [M + H]+ 343.1108; found 343.1111.
Supporting Information (see footnote on the first page of this arti-
cle): Cartesian coordinates and calculated energies for all computed
structures and additional NMR spectroscopic data for products.
[15]
[16]
A. Novácˇek, I. Hedrlín, Collect. Czech. Chem. Commun. 1967,
32, 1045–1050.
G. S. Abdrakhimova, M. Y. Ovchinnikov, A. N. Lobov, L. V.
Spirikhin, S. P. Ivanov, S. L. Khursan, J. Phys. Org. Chem.
2014, 27, 876–883.
Acknowledgments
[17]
[18]
P. S. Pregosin, in: Transition Metal Nuclear Magnetic Reso-
nance (Ed.: P. S. Pregosin), Elsevier, Amsterdam, 1991.
a) M. Bühl, F. T. Mauschick, F. Terstegen, B. Wrackmeyer, An-
gew. Chem. Int. Ed. 2002, 41, 2312–2315; Angew. Chem. 2002,
114, 2417; b) B. Wrackmeyer, O. L. Tok, A. Ayazi, H. E. Mai-
sel, M. Herberhold, Magn. Reson. Chem. 2004, 42, 827–830; c)
The research was supported by the University of Zagreb (grant
number KFPI 1.1.1.8). The authors thank the Computing Centre
SRCE for allocating computer time on the Isabella cluster. The
authors thank Professor Branka Zorc for the generous donation of
selected chemicals, and Ivana Amerl for the technical assistance.
The NMR instrument donation from Symrise AG is gratefully ac-
knowledged.
ˇ
A. Saric´, V. Vrcˇek, M. Bühl, Organometallics 2008, 27, 394–
ˇ
401; d) J. Lapic´, M. Cetina, D. Sakic´, S. Djakovic´, V. Vrcˇek, V.
Rapic´, ARKIVOC 2010, 9, 257–268.
[19]
M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria,
M. A. Robb, J. R. Cheeseman, G. Scalmani, V. Barone, B.
Mennucci, G. A. Petersson, H. Nakatsuji, M. Caricato, X. Li,
H. P. Hratchian, A. F. Izmaylov, J. Bloino, G. Zheng, J. L. Son-
[1] a) V. Sai Sudhir, C. Venkateswarlu, O. T. Muhammed Mus-
thafa, S. Sampath, S. Chandrasekaran, Eur. J. Org. Chem.
2009, 2120–2129; b) G. Mwande-Maguene, J. Jakhal, J.-B. Lek-
5430
www.eurjoc.org
© 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2015, 5424–5431

Ferrocenoyl-Substituted Pyrimidine Nucleobases
nenberg, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hase-
gawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai,
T. Vreven, J. A. Montgomery Jr., J. E. Peralta, F. Ogliaro, M. [24] V. Barone, M. Cossi, J. Tomasi, J. Chem. Phys. 1997, 107, 3210–
[23] M. Cossi, G. Scalmani, N. Rega, V. Barone, J. Chem. Phys.
2002, 117, 43–54.
Bearpark, J. J. Heyd, E. Brothers, K. N. Kudin, V. N. Starov-
erov, R. Kobayashi, J. Normand, K. Raghavachari, A. Rendell,
J. C. Burant, S. S. Iyengar, J. Tomasi, M. Cossi, N. Rega, J. M.
Millam, M. Klene, J. E. Knox, J. B. Cross, V. Bakken, C. Ad-
amo, J. Jaramillo, R. Gomperts, R. E. Stratmann, O. Yazyev,
A. J. Austin, R. Cammi, C. Pomelli, J. W. Ochterski, R. L. Mar-
tin, K. Morokuma, V. G. Zakrzewski, G. A. Voth, P. Salvador,
J. J. Dannenberg, S. Dapprich, A. D. Daniels, O. Farkas, J. B.
Foresman, J. V. Ortiz, J. Cioslowski, D. J. Fox, Gaussian 09,
revision D.01, Gaussian, Inc., Wallingford CT, 2009.
3221.
[25] V. Barone, M. Cossi, J. Phys. Chem. A 1998, 102, 1995–2001.
[26] J. R. Cheeseman, G. W. Trucks, T. A. Keith, M. J. Frisch, J.
Chem. Phys. 1996, 104, 5497–5509.
[27] a) A. J. H. Wachters, J. Chem. Phys. 1970, 52, 1033–1036; b)
P. J. Hay, J. Chem. Phys. 1977, 66, 4377–4384.
[28] E. Haslinger, W. Robin, K. Schloegel, W. Weissensteiner, J. Or-
ganomet. Chem. 1981, 218, C11.
[29] C. Gonzales, H. B. Schlegel, J. Phys. Chem. 1990, 4, 5523–5527.
[30] D. Ardura, R. López, T. L. Sordo, J. Phys. Chem. B 2005, 109,
23618–23623.
[20]
M. Dolg, U. Wedig, H. Stoll, H. Preuß, J. Chem. Phys. 1987,
86, 866–872.
[31] a) W. F. Little, R. Eisenthal, J. Am. Chem. Soc. 1960, 82, 577–
580; b) J. Lapic´, A. Pezerovic´, M. Cetina, S. Djakovic´, V. Rapic´,
J. Mol. Struct. 2011, 990, 209–216.
[21] a) T. Romero, A. Caballero, A. Espinosa, A. Tárraga, P. Mol-
ina, Dalton Trans. 2009, 2121–2129; b) W.-Y. Wang, N.-N. Ma,
S.-L. Sun, Y.-Q. Qiu, Phys. Chem. Chem. Phys. 2014, 16, 4900–
4910.
[22] W. Koch, M. C. Holthausen, in: A Chemist’s Guide to Density
Functional Theory; Wiley-VCH, Weinheim, Germany, 2000,
and the extensive bibliography therein.
[32] A. Sonoda, I. Moritani, J. Organomet. Chem. 1971, 26, 133–
140.
[33] H. J. Lorkowski, R. Pannier, A. Wende, J. Prakt. Chem. 1967,
35, 149–158.
Received: May 19, 2015
Published Online: July 17, 2015
Eur. J. Org. Chem. 2015, 5424–5431
© 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
5431