Anionic amidinourea-metal(II) tautomers: MN2O2or MN4?

Article abstract of DOI:10.1246/bcsj.20140130

Structures of amidinourea-metal complexes have been obscured by subtle tautomerism and crystal waters. To clarify the most stable form of the tautomers, bis(amidinourea)copper(II) anhydride and the alkyl derivatives were synthesized under basic conditions, and the stability of the CuN₂O₂ chromophore was suggested. From substituent effects of chelations, UV-vis spectra, IR/Raman spectra, and DFT calculations, it was suggested that the bis(amidinourea)copper-(II) has a CuN₂O₂ chromophore, not CuN₄ chromophore. The mutual exclusion rule of IR/Raman spectra supported the trans-CuN₂O₂ chromophore. Systematic calculations on bis(amidinourea)-Cu(II), -Ni(II), -Pd(II), -Zn(II), and -Cd(II) chelates revealed that the robust stability of CuN₂O₂ chromophores originates from quasi aromaticity in the six-membered chelate rings. cis- or trans-CuN₄ stability reported in the early references is suspicious.

Full text of DOI:10.1246/bcsj.20140130

Anionic Amidinourea-Metal(II) Tautomers: MN2O2 or MN4?
Masashi Hatanaka
School of Engineering, Tokyo Denki University, 5 Senju-Asahi-cho, Adachi-ku, Tokyo 120-8551
E-mail: mhatanaka@xug.biglobe.ne.jp
Received: May 2, 2014; Accepted: September 16, 2014; Web Released: September 26, 2014
Structures of amidinourea-metal complexes have been obscured by subtle tautomerism and crystal waters. To clarify
the most stable form of the tautomers, bis(amidinourea)copper(II) anhydride and the alkyl derivatives were synthesized
under basic conditions, and the stability of the CuN2O2 chromophore was suggested. From substituent effects of
chelations, UV-vis spectra, IR/Raman spectra, and DFT calculations, it was suggested that the bis(amidinourea)copper-
(II) has a CuN2O2 chromophore, not CuN4 chromophore. The mutual exclusion rule of IR/Raman spectra supported the
trans-CuN2O2 chromophore. Systematic calculations on bis(amidinourea)-Cu(II), -Ni(II), -Pd(II), -Zn(II), and -Cd(II)
chelates revealed that the robust stability of CuN2O2 chromophores originates from quasi aromaticity in the six-membered
chelate rings. cis- or trans-CuN4 stability reported in the early references is suspicious.
Amidinourea (AU (1); the left-hand side of the top of
Figure 1, often called guanylurea or 1-carbamoylguanidine)
reacts with various metal ions to form stable complexes.1-8
In early references, a tautomer shown in the right-hand side
has been widely accepted as the most stable form. Nowadays,
however, the correct form is found by X-ray analyses to be the
Schiff-base type in the left-hand side.9 Amidinourea derivatives
(AUs) with alkyl substituents also react with various metal
ions, especially in Cu2+. Recently, amidinothioureas have also
been studied as related ligands.10 AUs have been applied to
designing metal-complexing agents.11,12 Under weakly basic
or weakly acidic conditions, AU is usually neutral, and all the
hydrogen atoms combine with the branched nitrogens. In dilute
aqueous solutions or polymer matrixes, AUs react with excess
CuCl2 in a ratio 2:1 to form light-green complexes, in which
Cu2+ is combined between two central nitrogen atoms of planar
AU moieties, and two chloride ions coordinate to Cu2+ in the
perpendicular direction.11 In general, however, the coordination
structures of AU-metal complexes have wide diversity. For
example, as well as a special form above, Hubberstey et al.
studied a coordinate structure of (neutral-AU)2-copper(II)
dinitrate by X-ray analysis, and determined the chelate struc-
tures with a CuN2O2 chromophore (Figure 2a).6 In strongly
H
N
2+
H2N
N
NH2
H2N
NH2
O
C
NH
C
O
C
NH2
C
HN
O
O
NH
O
O
Cu
Cu
NH
NH
H2N
N
NH2
H2N
N
NH2
H
1
H2N
H2N
N
NH2
H2N
H2N
N
H
NH2
O
NH2
C
O
C
NH2
C
(a)
(b)
C
CH3
CH3
H2N
N
N
H2N
N
N
H
N
H
O
N
O
CH3
2
3
HN
HN
NH
NH
HN
NH2
NH
Cu
Cu
H2
C
H2N
CH3
CH3
O
N
C
O
C
HN
C
CH3
C
CH3
O
N
H
NH2
O
N
NH2
H2N
N
N
H2N
N
N
H
CH3
4
5
(c)
(d)
Figure 2. Possible
molecular
structures
of
bis-
Figure 1. Molecular structures of amidinourea (AU; 1),
N-methylamidinourea (2), N,N-dimethylamidinourea (3),
N,N¤-dimethylamidinourea (4), and N,N,N¤-trimethyl-N¤-
benzylamidinourea (5).
(amidinourea)copper(II) chelates. (a) is based on neutral
AU. (b), (c), and (d) are based on anionic (deprotonated)
AU.
162 | Bull. Chem. Soc. Jpn. 2015, 88, 162–172 | doi:10.1246/bcsj.20140130
© 2014 The Chemical Society of Japan

acidic conditions, AU is also coordinated by a proton.7,9,11
We denote it [AU + H]+. This cationic moiety forms stable
complexes with metal-containing counter anions, for example,
[CuCl4(H2O)2]2¹.7 Given that there are many possible tauto-
mers of AUs, determination of the coordination structures has
been a subtle problem. The most stable isomer of each complex
has been determined on a case-by-case basis.
ender color is investigated. The classical observation of this
complex was reported by Trimble;2 AU reacts with CuCl2 in
water in the presence of sodium potassium tartrate and NaOH
to form a lavender complex. The Cu content has been reported
to be 22.8%.2 This value suggests that AU reacts with CuCl2 in
a ratio 2:1 and the complex does not contain chlorine atoms
and/or crystal waters. Therefore, we should regard the AU
moieties in this lavender complex as deprotonated anions.
Although the complexation ratio is nearly evident from the
Cu content, the coordination structure is not clear yet. Thus, the
plausible coordination structures were investigated by means
of UV-vis, IR spectra, Raman spectra, and DFT (density func-
tional theory) calculations. The complexation ratio was recon-
firmed by ICP analysis. The complex contains six-membered
chelate rings, in which oxygen atoms of carbonyl groups and
nitrogen atoms of imino groups plausibly participate in the
coordination as depicted in Figure 2b. In addition, the IR/
Raman exclusion rule reveals the trans-type CuN2O2 chromo-
phore. Possible isomers with the carbamoyl-coordinated CuN4
chromophore (Figure 2c) or tautomers with the proton trans-
fered CuN4 chromophore (Figure 2d) are plausibly eliminated
from both experimental and theoretical points of views. The
coordination structure is also confirmed by substituent effect on
the complexation using substituted AUs; N-methylamidinourea
(2), N,N-dimethylamidinourea (3), N,N¤-dimethylamidinourea
(4), and N,N,N¤-trimethyl-N¤-benzylamidinourea (5) shown
in Figure 1. From the second-order perturbation theory and
systematic calculations on bis(amidinourea)-Cu(II), -Ni(II),
-Pd(II), -Zn(II), and -Cd(II) chelates, it is shown that the
robust stability of the metal-N2O2 chromophore is attributed to
quasi aromaticity of the 6π chelate rings, in which the branched
amino groups lower the energy of the systems through an
extended topological charge stabilization rule.
Under basic conditions, on the other hand, AUs are subject to
deprotonation by strong bases such as NaOH, and the resultant
anionic moiety serves as a robust ligand for various metals.
In this article, hereafter, the anionic ligand is denoted as
¹
[AU ¹ H] (the minus symbol in the parenthesis means
deprotonation). We must clearly distinguish neutral AU (simply
denoted as AU), the cationic form [AU + H]+, and the anionic
¹
¹
form [AU ¹ H] . [AU ¹ H] forms a lavender complex with
Cu(II) in a ratio 2:1,2 and the complexes often contain apparent
crystal waters.7,8 We note that in these cases there are no counter
anions any more. For simplicity, we hereafter focus only on the
anhydride form [AU ¹ H]2Cu. The standard synthetic method
was established by Trimble.2 As well as the [AU ¹ H]2Cu com-
plex, he synthesized various amidinourea-metal and amidino-
thiourea-metal complexes under basic conditions.2 However,
the coordination structure of the complex [AU ¹ H]2Cu has
long been an open question due to insolubility to usual solvents
(slightly soluble in aprotic solvent such as DMSO or DMF). A
single crystal has not been obtained yet, and early workers
speculated the coordination structure by mean of IR, vis
(visible), and ESR spectra. An early IR study reported a CuN2O2
¹
chromophore with >C=O and >C=N groups,4 but the rare
valence of the nitrogen atoms was suspicious. Electrophilic
substitution of the complex suggests the existence of imino
groups -C=NH with quasi aromaticity,3 and some vis-spectral
studies concluded the presence of CuN4 chromophore.7,8
However, many early works including the references above
were not sensitive to elimination of apparent crystal waters. This
ambiguity has caused serious confusions for a long time. In par-
ticular, despite the fact that we cannot distinguish [AU ¹ H]2-
Cu¢2H2O and AU2Cu¢(OH)2 from the composition, the ESR
hyperfine structures was formally attributed to the same CuN4
chromophore.5,8 In early references, charges of the ligands and
the resultant change of the electronic states are often disregard-
ed. Anyway, there has been no robust confirmation for the most
stable coordination structure of the anionic AU complexes. As
chemically reasonable structures, there are three coordination
types, as shown in Figures 2b, 2c, and 2d. Figure 2b has a
CuN2O2 chromophore, and Figure 2c and Figure 2d have CuN4
chromophores. Figure 2d is the proton-transferred tautomer of
Figure 2c. Though Figure 2 shows only trans-isomers, there
exists the corresponding cis-isomers for each coordination type.
The detailed analysis of the coordination structure is important
in that AU is a hybrid analogue of biuret and biguanide, of
which Cu(II) complexes are well known in the field of bio-
chemistry. Moreover, in view of application to metal-complex-
Experimental
AU was prepared by hydrolysis of cyanoguanidine as
usual.11a The substituted AUs were kindly supplied by Shiba.11a
They were identified by NMR spectrometer (Bruker DPX 300).
1: 1H NMR (DMSO-d6): ¤ 7.11 (s, 2H), 8.66 (s, 4H);
13C NMR (DMSO-d6): ¤ 161.29, 166.68.
1
2: H NMR (DMSO-d6): ¤ 2.84 (s, 3H), 7.19 (s, 2H), 8.61
(s, 3H); 13C NMR (DMSO-d6): ¤ 27.94, 154.63, 154.92.
1
3: H NMR (DMSO-d6): ¤ 2.87 (s, 6H), 5.51 (s, 2H), 7.81
(s, 2H); 13C NMR (DMSO-d6): ¤ 35.98, 160.28, 166.47.
1
4: H NMR (DMSO-d6): ¤ 2.78 (s, 6H), 6.51 (s, 2H), 7.65
(s, 2H); 13C NMR (DMSO-d6): ¤ 27.43, 159.73, 165.85.
1
5: H NMR (DMSO-d6): ¤ 2.64, (s, 3H), 2.77 (s, 6H), 4.31
(s, 2H), 5.55 (s, 2H), 7.31 (s, 5H); 13C NMR (DMSO-d6): ¤
39.04, 39.40, 54.70, 127.68, 128.27, 128.57, 135.78, 161.34,
163.25.
Bis(amidinoureas)copper(II) chelates ([AUs ¹ H]2Cu) were
¹1
prepared referring to the literature.2 1% (7.4 © 10¹2 mol L
)
CuCl2 aqueous solution, 5% (0.24 mol L¹1) sodium potassium
tartrate (Rochelle salt) aqueous solution, 0.1 mol L aqueous
¹
¹1
ing agents, [AU ¹ H] is a promising Lewis base as well as the
neutral AUs. Indeed, we obtained anionic imino resins based on
methacryloyl AU, and it has been found that the anionic AUs
AUs solution (in 5, water/ethanol = 1:1 mixture was used as
¹1
solvent because of the low solubility), and 5% (1.25 mol L
)
have highly selective adsorption efficiency for Cu2+ 12
.
aqueous NaOH solutions were prepared using deionized and
distilled water. 2 mL of each reagent was poured into a test tube
in the named order. We note that Cu(OH)2 is precipitated if
In this paper, coordination structure of the 2:1 anhydride
complex bis(amidinourea)copper(II) ([AU ¹ H]2Cu) with lav-
Bull. Chem. Soc. Jpn. 2015, 88, 162–172 | doi:10.1246/bcsj.20140130
© 2014 The Chemical Society of Japan | 163

sodium potassium tartrate is not added. Although the mixtures
of CuCl2, sodium potassium tartrate, and aqueous AUs solu-
tions were blue, addition of aqueous NaOH solution changed
the color of the solutions into lavender. This lavender color is
due to the imino-type [AUs ¹ H]2Cu complexes. It is known
that [AU ¹ H]2Cu is insoluble in water or usual protic solvents.
However, the UV-vis spectra of these supersaturated solu-
tions could be recorded in a 10 mm cell by HITACHI U-4000
spectrophotometer.
After 3 days, lavender powder was precipitated from the
solution of AU (1). This precipitate was washed with water and
ethanol. The precipitate was filtered and dried under vacuum.
The Cu content was confirmed by ICP analysis to be 23.75%
(Calcd. 23.91%), which is consistent with that in the literature,2
and the complexation ratio was confirmed to be 2:1. The IR
spectrum was obtained by KBr-pellet using a JEOL JIR-RFX
3001 spectrometer. The Raman spectrum was recorded with a
HORIBA/Jobin Yvon LabRAM HR-800 laser-Raman spec-
trometer. No precipitate formed from the solutions of sub-
stituted-AUs 2-5 even after 6 months.
with -max = 582, 586, and 582 nm, respectively (¾ ¿ 40,
Figure 3b). The role of Rochelle salt is not clear at present,
but this probably serves as a catalyst and buffer for the
complexation. Strictly speaking, we must discount the red-shift
effects by residual Rochelle salt and Cu(II) ion. Indeed, in
DMSO, the peak of pure [AU ¹ H]2Cu appeared with -max
=
516 nm, ¾ = 40. The absorbance was essentially independent
of temperature within 20-80 °C, which indicates robust stability
of the complex. On the other hand, reaction mixtures of 4 and 5
in which the N¤ positions are substituted by methyl and/or
benzyl groups did not form lavender complex solutions. The
color of these solutions were blue with -max = 673 and 672 nm,
respectively. 4 and 5 should be neutral under the present con-
11
ditions, and subject to complexation with CuCl2 or Rochelle
salt, if any. However, they are not important for our purpose.
Thus, we can conclude from such substituent effects that
N¤-nitrogen atoms of AUs participate in the coordination. That
¹
is, OH deprotonates the N¤-amino hydrogen atoms of AUs,
and the resultant imino-type anions react with Cu2+ in a ratio
2:1 to form lavender complexes.
Figure 4 shows IR and Raman spectra of [AU ¹ H]2Cu.
Taking a closer look, IR peaks do not always correspond to the
Raman peaks, particularly in the fingerprint region. Differ-
ence in IR/Raman spectra appears more clearly in metal-
ligand vibrations than in ligand vibrations, because interac-
tions between the vibrations are large in the fingerprint region.
Indeed, as shown later, the mutual exclusion rule for the IR/
Raman spectra is clearly shown in the fingerprint region.
Before the detailed assignment of the metal-ligand vibra-
tions, we focus on the change of IR vibrations in the ligand
moieties.
Results and Discussion
Figure 3a shows UV spectrum of the [AU ¹ H]2Cu (solid
line), together with that of neutral AU (dashed line). For
[AU ¹ H]2Cu, the weak background of sodium potassium
tartrate was calibrated. Figure 3b shows vis spectra of the
reaction mixtures of 1-5 and CuCl2 in the presence of sodium
potassium tartrate and NaOH. While AU 1 has π-π* absorption
¹1
at 219 nm (¾ = 12600 M¹1 cm in the presence of 0.01%
NaOH), complexation with Cu2+ results in blue shift of the
π-π* absorption at 205 nm (¾ = 20700) with a broad shoulder
at 216 nm. The peak at 205 nm is consistent with that of N-
propionylamidinourea anion (203 nm, ¾ = 20900 at pH 10.9).12
If the central nitrogen coordinates with metals or proton, the
major peak for the π-π* absorption should shift to the vacuum
UV region.11,13 Therefore, the π-π* absorption observed sug-
gests that the Schiff-base moiety is conserved through the com-
plexation. 1, 2, and 3 in which the N¤ positions of AU are not
substituted by any group formed lavender complex solutions
IR
(a)
(b)
1
5
4
3
0.5
0.4
0.3
0.2
0.1
0
3250
2250
Wavenumber/cm–1
1250
250
2
2
Raman
1
0
200 220 240 260
Wavelength/nm
400
500
600
Wavelength/nm
700
800
Figure 3. (a) UV spectra of AU (dashed line) and bis-
(amidinourea)copper(II) ([AU ¹ H]2Cu, solid line). (b) vis
spectra of the reaction mixtures of 1-5 and CuCl2 in the
presence of sodium potassium tartrate and NaOH. 1, 2, and
3 form the lavender complexes. 4 and 5 do not form the
complexes.
3250
2250
Wavenumber/cm–1
1250
250
Figure 4. IR and Raman spectra of bis(amidinourea)-
copper(II) ([AU ¹ H]2Cu).
164 | Bull. Chem. Soc. Jpn. 2015, 88, 162–172 | doi:10.1246/bcsj.20140130
© 2014 The Chemical Society of Japan

are sufficiently small to be neglected (details are attached in the
Supporting Information.). We denote each structure as: trans-
CuN2O2 (6), cis-CuN2O2 (7), trans-CuN4 (8), cis-CuN4 (9),
proton-transferred trans-CuN4 (10), and proton-transferred cis-
CuN4 (11). We can see that the trans-form chelate structure 6
with the CuN2O2 chromophore is most stable. This is a nearly
planar molecule with a C2 geometry. The calculated Cu-O and
Cu-N bond lengths were 1.930 and 1.905 ¡, respectively. The
Cu-O bond length is comparable to that of typical ureas-
copper(II) complexes, for example, an X-ray analyzed com-
plex between 5,5¤-carbonylbis[1,3-dimethyl-3,4,5,6-tetrahydro-
1,3,5-triazin-2(1H)-one] and CuCl2 with 1.948 ¡ Cu-O dis-
tance.18 The C=O bond length of the complex 6 was calculated
to be 1.280 ¡, which was longer than that of neutral AU
(1.243 ¡). This is consistent with the IR-spectra data. The N-
Cu-O bond angles inside and outside the chelate ring were
calculated to be 91.0 and 92.5°, respectively. Strictly speaking,
the trans-isomer is not completely planar, as shown in the
side views in Figure 6. However, for analyzing the IR/Raman
spectra, it is convenient to regard the symmetry classification
as C2h approximately with an inversion center, as shown later.
The cis-form chelate structure with CuN2O2 chromophore was
slightly unstable (+4.1 kcal mol¹1) due to the steric hindrance
between two hydrogens of the imino groups. The cis-CuN2O2
form also has nearly a planar geometry, of which the point
group is C2 (¿ C2v). On the other hand, there are two possible
types of chelate structures with CuN4 chromophores, as shown
in the bottom of Figure 6. One is the complex of which imino
-NH and carbamoyl -NH2 groups participate in the coordina-
tion, and another is the proton-transferred tautomer of which
central nitrogens are saturated. However, these were unstable
both in trans- and cis-forms. In particular, the amino coor-
dination is very unstable. This is reminiscent of large proton
affinity of imino sites in biguanide and polyguanide.19,20
Imino-coordination preference is essentially attributed to
large delocalization energy of the π electrons, as suggested
from classical resonance structures. In the proton transferred
form, however, the saturated central nitrogen atom causes less
delocalization of the itinerant electrons in the AU moieties,
similar to [AU + H]+ or diprotonated biguanide.13 This is
clearly seen from increase of the central C-N bond lengths and
decrease of C=N bond lengths in the proton-transferred form.
That is, while in CuN2O2 chromophores the conjugation of
π electrons is realized over the whole of the chelate rings,
in CuN4 chromophores with proton transfers the conjugation
is separated at the central nitrogen. Stabilities of CuN2O2
chromophores relative to the CuN4 chromophores are quite
large (ca. 15 kcal mol¹1). Given that deviations from planarity
of each chelate ring in 6, 7, 10, and 11 are small, conforma-
tional energies of amide groups do not so much contribute to
the total stabilities. Indeed, when torsion angles of the amide
(a)
(b)
(a)
(b)
4000 3500 3000 2500
Wavenumber/cm–1
2000
1600
1200
Wavenumber/cm–1
Figure 5. IR spectra in the 4000-2500 and 2000-1200
cm regions. (a) AU (1) and (b) bis(amidinourea)copper-
¹1
(II) ([AU ¹ H]2Cu).
Figure 5 shows comparison of the zoomed IR spectra of AU
(Figure 5a) and the [AU ¹ H]2Cu complex (Figure 5b). The
peaks corresponding to N-H stretching of AU were observed
at 3361 and 3222 cm¹1. They are two broad bands. They are
broadened by intermolecular association through the carbamoyl
groups. We note that the intermolecular association through the
carbamoyl groups lowers the wavenumbers of N-H stretching.
Therefore, we should consider that the wavenumbers of N-H
stretching of nonassociated AU are higher than those of the
observed data. On the other hand, the peaks corresponding to
N-H stretching of the [AU ¹ H]2Cu complex were observed
at 3403, 3363, 3352, and 3197 cm¹1. The peaks at 3403, 3363,
¹1
and 3352 cm are attributed to carbamoyl-side-NH2 groups,
because they are more narrowed than those of AU by lifting
the intermolecular association and the complexation. Indeed,
some of these wavenumbers are higher than those of AU. The
¹1
peak at 3197 cm is attributed to the imino groups formed
by the deprotonation.
The peaks corresponding to C=O stretching were observed
¹1
¹1
at 1635 cm in AU and 1622 cm in the [AU ¹ H]2Cu com-
plex, respectively. The wavenumber of the complex was lower
than that of AU. Given that the intermolecular association of
AU lowers the wavenumber of C=O stretching and elongates
the C=O bond length, we should consider that the wavenumber
of C=O stretching of nonassociated AU is much higher than
that of the observed data. Indeed, in a polymer matrix, the peak
of C=O stretching in the nonassociated AU moiety is observed
at 1734 cm .
¹1 11 Therefore, we can conclude that the C=O bond
length of the [AU ¹ H]2Cu is longer than that of AU. In short,
the oxygen atoms in carbonyl groups plausibly participate in
the coordination, and the C=O bond length is elongated by the
complexation.
In order to confirm the coordination structure described
above, we performed DFT calculations on optimized structures
of the complex. The calculations were done under unrestricted
B3LYP/6-31G(d) and B3LYP/6-31G(d,p) level of theory14-16
by using the Gaussian program.17 Though the basis set impact
on the optimized geometries and energetics was trivial,
vibration analyses, and thermal corrections succeeded only at
B3LYP/6-31G(d). Figure 6 shows optimized structures of the
possible isomers of the 2:1 complex at B3LYP/6-31G(d) level
of theory. The relative energies of each structure were also
noted. Change of the zero-point energy and thermal corrections
groups are less than 45°, the difference of conformational
¹1 21
energies should be at most 4-5 kcal mol
.
In 6, the torsion
angle of the trans-amide moiety is 15.7°, and in 10, the torsion
angles of the cis-amide moieties are 32.1° (side-H-N-C-O)
and 35.1° (central-H-N-C-O). Therefore, the large stabilities
of CuN2O2 chromophores are mainly attributed to the large
delocalization energy in the chelate rings. In the carbamoyl
-NH2 coordination 8 and 9, the delocalization energy is much
Bull. Chem. Soc. Jpn. 2015, 88, 162–172 | doi:10.1246/bcsj.20140130
© 2014 The Chemical Society of Japan | 165

6
7
89.2°
91.0°
trans-CuN2O2 (side view)
86.6°
92.5°
1.905
1.922
1.923
1.930
1.275
1.370
1.280
1.373
1.338
1.323
1.384
1.346
1.327
1.386
1.341
120.0°
120.8°
1.342
tilt angle
trans-CuN2O2 (side view 2)
cis-CuN2O2
+4.1 kcal mol–1
9
trans-CuN2O2
0.0 kcal mol–1
8
10
11
87.4°
87.9°
91.1°
91.3°
96.0°
98.7°
97.9°
95.7°
1.906
1.923
1.966
1.971
2.011
2.028
1.918
1.921
1.498
1.218
1.333
1.230
1.505
1.218
1.343
1.376
122.3° 1.341
1.307
1.387
1.336
1.231
1.305
1.385
123.6°
131.5°
131.3°
1.357
1.382
1.344
1.333
1.333
1.456
1.342
1.354
1.451
Trans CuN4
trans-CuN4
+38.6 kcal mol–1
cis-CuN4
+36.3 kcal mol–1
Cis CuN4
Proton transferred
Proton transferred
+15.2 kcal mol–1
+17.0 kcal mol–1
Figure 6. Optimized structures of bis(amidinourea)copper(II) ([AU ¹ H]2Cu) at B3LYP/6-31G(d) level of theory. The parameters
are units in ¡ and degrees.
Table 1. Summary of DFT Calculations on Bis(amidinourea)copper(II) Isomers at B3LYP/6-31G(d)// B3LYP/6-31G(d) and B3LYP/
6-31G(d,p)// B3LYP/6-31G(d,p) (in Parentheses)a)
Relative energy
/kcal mol
ZPEd)
/hartree (particle)
Isomers
Symmetry
©S2ªe)
¹1
¹1
6
7
8
9
trans-CuN2O2
C2 ¿ C2h
C2 ¿ C2v
C2
C2
C2
C2
0.0b) (0.0)c)
+4.1 (+2.2)
+38.6 (+39.1)
+36.3 (+37.0)
+15.2 (+14.6)
+17.0 (+16.4)
0.187548
0.186799
0.187308
0.188186
0.187700
0.187583
0.7521b) (0.7521)c)
0.7520 (0.7520)
0.7521 (0.7521)
0.7521 (0.7521)
0.7523 (0.7523)
0.7523 (0.7523)
cis-CuN2O2
trans-CuN4
cis-CuN4
10 trans-CuN4 (Proton transferred)
11 cis-CuN4 (Proton transferred)
a) We note that symmetry classification C2 of 6 and 7 can be approximately regarded as C2h and C2v, respectively, due to the nearly
planar geometries. b) B3LYP/6-31G(d)// B3LYP/6-31G(d), the same hereinafter. The total energy of 6 is ¹2387.359505 hartree.
c) B3LYP/6-31G(d,p)// B3LYP/6-31G(d,p), the same hereinafter. The total energy of 6 is ¹2387.393622 hartree. d) Zero-point energy
at B3LYP/6-31G(d)// B3LYP/6-31G(d). e) Spin-squared expectation value before annihilation. Since ©S2ª for a pure doublet state
should be 0.7500, all the values suggest nearly exact doublet states.
reduced due to the nonplanarity of sp3-like configuration.
Indeed, these nonplaner geometries are reflected in the large
instability in energy (over ca. 35 kcal mol¹1). Thus, we can
plausibly conclude that the lavender complex is a trans chelate
compound 6 depicted in the top of Figure 6. A summary of the
quantum-chemical calculations are shown in Table 1.
nitrogen. In the B3LYP/6-31G(d) level of theory, HOMO is
the lone-pair orbital, and HOMO¹1, HOMO¹2, LUMO are
characterized as π, π, and π* orbitals, respectively (Figure 7a).
We see that the Schiff-base moiety has a weak bonding ampli-
tude pattern, which is reflected in the π-π* UV absorption
above and weak bond alternation of AU.9 AU anion also has
non-bonding HOMO and lone-pair orbitals (Figure 7b). The
eigenvalues of frontier orbitals are much higher than those
of neutral AU due to the anionic character. In particular,
HOMO¹1 and HOMO¹2 are characterized as lone-pair
orbitals of imino nitrogen and carbonyl oxygen, which serve
as strong donor sites. The amplitudes at the amino groups are
The frontier orbitals of neutral AU (1), AU anion
¹
([AU ¹ H] ), and its trans-CuN2O2 complex [AU ¹ H]2Cu
are shown in Figures 7a, 7b, and 7c, respectively. AU has
nearly nonbonding frontier π orbitals due to the bipartite
skeleton, similar to odd alternant hydrocarbons. In addition,
there are lone pairs on the carbonyl oxygen and the central
166 | Bull. Chem. Soc. Jpn. 2015, 88, 162–172 | doi:10.1246/bcsj.20140130
© 2014 The Chemical Society of Japan

(a)
(b)
(c)
¹
Figure 7. Selected molecular orbitals of AU, deprotonated AU ([AU ¹ H] ), and bis(amidinourea)copper(II) chelates
([AU ¹ H]2Cu) at B3LYP/6-31G(d) level of theory. The eigenvalues of Kohn-Sham orbitals are unit in hartree.
relatively small. The frontier orbitals of the trans-CuN2O2
complex consist of d orbital of Cu2+ and the nonbonding
fragments of AU (Figure 7c). The singly occupied molecular
orbital (SOMO) consists of antibonding interactions between
the dxy component and lone pairs of the carbonyl oxygens and
imino nitrogens. The amplitudes on the remaining nitrogens are
also not trivial. Therefore, ESR hyperfine structures reported
by early workers8 are attributed to not CuN4 chromophores, but
the nitrogen atoms in the trans-CuN2O2 chromophore and the
second nearest-neighbor nitrogens, contrary to bis(biguanide)-
copper(II),22 bis(O-alkylamidinourea)copper(II).23,24 The other
occupied frontier orbitals also come from the d components
and nonbonding moieties of AU. However, nonplanarity of the
complex allows them to interact with each other to afford π-σ
mixed orbitals. The vis spectra are attributed to the transition
from these π-σ mixed orbitals to SOMO. Indeed, TDDFT
(time-dependent DFT) calculations at the same level predicted
the transition from the π-σ mixed orbitals (for example,
SOMO-2) to SOMO with 552.9 nm, which is nearly consistent
with the experimental data. It is noteworthy that the absorption
coefficients (obs. ¾ = 40) in visible region is relatively smaller
than those of general Schiff-base complexes.25
symmetric and symmetric modes with respect to the copper
atom, respectively. In particular, large peaks (for example,
ring-breathing at 952 and 956 cm¹1) approximately satisfies
the mutual exclusion rule, in which IR and Raman peaks do
not coexist at the same wavenumber. In the fingerprint region
(below 1300 cm¹1), we see good consistency between the
observed and calculated profiles within the scale factor errors
(0.96 for B3LYP/6-31(d)).26 NH2 rocking peaks around 1132
¹1
and 1134 cm are probably coupled between the adjacent
modes, and the order of the wavenumber is slightly changed in
the observed data. Above 1300 cm¹1, deviations from theoret-
ical are slightly large due to perturbation of molecular fields.
¹1
For example, NH2 scissoring peaks 1469 and 1474 cm
(calcd.) are considerably shifted to low-energy due to the edge
effects. Nevertheless, as a whole, most of the peaks were
reasonably assigned. The local molecular fields are probably
asymmetric to violate the selection rule to some extent, and the
IR/Raman intensities slightly deviate from theoretical predic-
tion. As for the high wavenumber region, each peak is subject
to molecular fields grossly, because the vibrations are localized
at specific functional groups or terminal residues. Thus, the
vibrations do not so much interact all over the molecule, and
the mutual exclusion rule is also not so important. However, we
All the optimized structures of the trans- and cis-chelates
belong to C2 point group, as discussed above. However, if the
degree of nonplanarity is small, the trans-chelates are regarded
as C2h molecules with an inversion center, and cis-chelates are
regarded as C2v molecules. Therefore, we can expect the mutual
exclusion rule for IR/Raman spectra for the most stable trans-
isomer. Figure 8 shows IR/Raman spectra of the complex
emphasize the remarkable IR-band shift of the nonassociated
¹1 11
carbonyl stretching: before coordination, obs. 1734 cm
,
calcd. 1719 cm¹1; after coordination, obs. 1622 cm¹1, calcd.
1643 cm¹1. For reference, the details of IR/Raman simulations
of [AU ¹ H]2Cu for all the wavenumbers are summarized in
the Supporting Information.
¹1
within 1800-700 cm and the corresponding calculated data.
Remarkable data associated with the metal-ligand bonds
appeared below 700 cm¹1. Figure 9 shows comparison of IR
IR and Raman peaks are approximately classified into anti-
Bull. Chem. Soc. Jpn. 2015, 88, 162–172 | doi:10.1246/bcsj.20140130
© 2014 The Chemical Society of Japan | 167

2
2
IR Obs.
ρ
ω
IR
2
s
a
ν
τ
ω
1236
ρ
ω
IR Calcd.
2
2
ρ
τ
ω
ρ
ω
s
ν
Raman
ν
ρ
ω
σ
ρ
700
600
500
400
ω
ρ
1665
Wavenumber/cm-1
σ
(
C
=
O
,
C
=
N
)
ρ
ρ
ν
ν
Figure 9. Mutual exclusion rule for IR/Raman spectra
of bis(amidinourea)copper(II) ([AU ¹ H]2Cu) at 700-400
¹1
cm
.
1500
1000
(cm-1)
Raman
IR
σ
Raman Calcd.
Raman Obs.
Obs. 496 cm-1
Calcd. 507 cm-1
Obs. 482 cm-1
Calcd. 502 cm-1
N
N
N
O
O
N
N
O
O
N
Figure 8. Experimental and theoretical IR/Raman spectra
of bis(amidinourea)copper(II) ([AU ¹ H]2Cu) at 1800-700
cm¹1. Each approximate assignment corresponds to the
pair of antisymmetric/symmetric (IR/Raman) modes. The
Raman peak 1648 cm (calcd.) is -NH2 scissoring rather
than C=O stretching. Notation: ½; wagging, ¸; twisting,
μ; rocking, ·; scissoring, ¯; stretching.
Cu
Cu
¹1
N
N
ν
ν s
as
and Raman spectra of the complex within 700-400 cm¹1. We
see that the wavenumber of each Raman peak never coincides
with that of the IR peak. That is, the exclusion rule for IR/
Raman intensities exactly holds. In particular, symmetric and
antisymmetric Cu-O (partially containing Cu-N) stretching
modes ¯s (Cu-O) and ¯as (Cu-O) are clearly separated into
Raman and IR activities, respectively. This suggests that the
complex is the trans-isomer with an inversion center.
Figure 10. Symmetric and antisymmetric Cu-O (Cu-N)
stretching modes of bis(amidinourea)copper(II)
([AU ¹ H]2Cu).
6-31G(d)) are also shown. All the observed peaks were well
reproduced and attributed as stretching or bending modes.
Though we can use the scaling factor for B3LYP calculations if
necessary, the present simulation is quite adequate for assign-
ment of each peak.
Figure 10 shows schematic representation of the symmetric
and antisymmetric Cu-O stretching modes obtained by the
vibration analysis. The observed wavenumbers corresponding
to these stretching modes are very close to those of typical
complexes bis(glicinate)Cu(II)27 or tetraamminecobalt(III)
salts.28 In particular, our results are comparable to ¯as and ¯s in
Theoretical Aspects of the Electronic States
Quasi Aromaticity in Anionic Amidinourea-Cu(II) Com-
plexes. The origin of stability of the CuN2O2 chromophore
is not so evident from DFT calculations only. Given that
the amino-coordinated complexes are evidently eliminated as
above, we here focus on the energetics of trans-CuN2O2 (6) and
proton-transferred trans-CuN4 (10) chelates. From the view-
point of valence bond (VB) descriptions, the CuN2O2 chelate
¹1
trans-bis(glicinate)Cu(II)¢2H2O:27 IR; 483 cm (¯as), Raman;
¹1
462 cm (¯s) (Though the approximate assignment in the
original paper is Cu-N,27 these are probably coupled with Cu-O
to a large extent.). Table 2 shows a summary of the IR/Raman
spectra at 700-400 cm¹1. The DFT-calculated data (B3LYP/
168 | Bull. Chem. Soc. Jpn. 2015, 88, 162–172 | doi:10.1246/bcsj.20140130
© 2014 The Chemical Society of Japan

Table 2. Summary of IR and Raman Spectra at 700-
400 cm
of the amplitude is ca. 0.1-0.2. The nontrivial coefficients
result from the residual 4p component after the dsp2 hybrid-
ization. The HOMOs of TMP3¹ are shown in the middle of
Figure 11. These are triply degenerate nonbonding molecular
orbitals (NBMOs),33 in which the amplitudes of branched
methylene groups are relatively large. 6 and 10 consists of
two TMP3¹ fragments and Cu(II), and their MOs are also
approximately constructed by the Hückel method with het-
eroatom parameters. Now, we consider an isoelectronic hydro-
carbon corresponding to the complexes as the zero-th order
approximation. The skeleton is shown in Figure 11. We note
that the central site is formally linked to the four adjacent sites.
This is a hypothetical pentaanion of which NBMOs are fivefold
degenerate. The MO levels of the isoelectronic hydrocarbon
with 20π electrons are also shown in Figure 11. Let us consider
change of the Coulomb integrals to evaluate the energy of 6
and 10. The energy change is estimated by the perturbation
theory without ab initio calculations. Change of σ-bond energy
is trivial, because the tautomerism is formally regarded as
proton transfer between two lone pairs on the same pyridine-
type nitrogen atoms. Within the second-order perturbation, the
change of the Coulomb integrals ¤¡r (in unit of the standard
resonance integral ¢) on the r-th site is reflected in the total
energy E.34
¹1
Observed
/cm
Calculateda),b)
Assignment
Notationc)
¹1
¹1
/cm
IR
418
457
496
421
471
507
-NH2 wagging
-NH2 rocking
Cu-O stretching,
antisymmetric
-NH2 twisting
-NH2 wagging
>N-H rocking
>N-H wagging
>N-H wagging
½ (NH2)
μ (NH2)
¯
(Cu-O)
as
513
528
634
690
706
542
553
638
669
689
¸ (NH2)
½ (NH2)
μ (NH)
½ (NH)
½ (NH)
Raman
478
482
471
502
-NH2 rocking
Cu-O stretching, ¯s (Cu-O)
symmetric
-NH2 twisting
-NH2 wagging
>N-H rocking
>N-H wagging
>N-H wagging
μ (NH2)
ca. 505d)
520
665
ca. 690?
695
539
552
638
669
693
¸ (NH2)
½ (NH2)
μ (NH)
½ (NH)
½ (NH)
X
X
1
2
2
ꢀ
a) Calculated under B3LYP/6-31G(d) level of theory. No scal-
E
E þ
qr¤¡r þ
³r;rð¤¡rÞ
¼
0
¹1
r
r
ing factors are used. b) Calcd. 638 and 669 cm are very
X
weak, and not completely mutual-excluded due to the devia-
tion of symmetry. c) Corresponding to Figure 9. d) Shoulder.
þ
¤Ebondð¤¢r;sÞ
ð1Þ
r;s
E0 is the π-electron energy of the reference hydrocarbon.
¤Ebond(¤¢r,s) is energy change by deviation ¤¢r,s from the reso-
nance integral ¢. However, this term is small, and almost
cancelled out when only energy difference between the two
isomers is sought. qr is π electron density on the r-the site, and
ring has 6π electrons, and the proton-transferred CuN4 chelate
ring has 7π electrons. Two fragments of chelate rings bind
through Cu(II) to form 20π electron systems. Figure 11 shows
one of the representative resonance structures for the CuN2O2
(for 6) and CuN4 (for 10) chelate rings. In the CuN4 chelate
rings, all the coordinated nitrogens are imino type. Each ligand
with 10π electrons is approximately described by Hückel
molecular orbitals of tetramethylenepropane trianion (TMP3¹).
In view of molecular orbital (MO) theory, 6π systems form
a bonding “shell” under the nonbonding levels, and each
branched amino nitrogen has lone-pair electrons due to the
nonbonding levels. In 7π systems, on the other hand, π-electron
densities on the branched carbonyl oxygens will become close
to unity. If π-electrons on Cu are nontrivial, the 6π systems
are thought to be much stable than 7π systems due to quasi
aromaticity. In the meantime, we use the term “quasi aroma-
ticity” for describing apparently cyclic compounds of which
the most plausible resonance structure contains (4n + 2)π
electrons in the ring moieties. Theoretically speaking, quasi
aromaticity is rationalized by Gimarc’s topological charge
stabilization rule;29-32 electronegativities of the constituent
atoms match the pattern of electron density distribution of
the isoelectronic hydrocarbon. However, as shown below, the
present 6π-7π energetics is concisely solved by perturbation
theory.
³
is self-atom polarizability in the reference hydrocarbon:
r,r
occ unocc
2
2
X X
Cjr Ckr
³
¼ 4
r;r
¾j ꢁ ¾
k
j
k
NBMOs LUMOs
2
2
X
X
Cjr Ckr
ꢀ
4
ð2Þ
¼
¾
NBMO ꢁ ¾
LUMO
j
k
where Cjr is the coefficients of the j-th Hückel MO on the r-th
site. Within the first-order perturbation, the gap E(CuN4) ¹
E(CuN2O2) is too roughly estimated. Through the second-order
perturbation, many components of the excited configurations
are mixed, and the estimation is highly improved. Then, only
NBMO-LUMO interactions are important for our purpose,
because they mainly contribute to eq 2. The π-electron den-
sity qr of the reference hydrocarbon is shown in Figure 11.
Supposing that ¤¡r(amino N) = 1.5¢, ¤¡r(carbonyl O) = 1.0¢,
¤¡r(imino N) = 0.5¢, and the resonance integral for benze-
noids ¢ = ¹1.0 eV,34
EðCuN4Þ ꢁ EðCuN2O2Þ
¼ 0:208j¢j ð1st orderÞ þ 0:184j¢j ð2nd orderÞ
ꢁ1
ꢀ
9:0 kcal mol
ð3Þ
¼
Given that the unpaired electron on SOMO has σ character,
we separate it from the present analysis. Mulliken population of
the DFT results shows that 4pz orbital of Cu participates in
forming the frontier orbitals to an extent, in which magnitude
The order of the energy difference is consistent with the DFT
results. Magnitude of the Coulomb integral on Cu(II) is proba-
bly half as large as that of carbon. However, this is not
Bull. Chem. Soc. Jpn. 2015, 88, 162–172 | doi:10.1246/bcsj.20140130
© 2014 The Chemical Society of Japan | 169

2+
2+
H
M
H
O
M
H
O
C
N
C
N
C
N
C
6π
7π
N
H
H
H
H
N
H
N
N
N
H
H
1
1
2
1
1
2
1
1
2
−
−
−
−
8
2
8
1
2
1
2
1
2
1
2
−
−
−
1
1
2
−
8
8
8
ε
ε
=
=
α
α
−
− 2
6
β
β
1.604
(.267)
1.604
(.267)
1.417
(.147)
ε
ε
= α − 2β
1 (0)
1(0)
LUMO
Non-bonding level
1.438
(.193)
1.438
(.193)
1
=
=
α
α
(0)
1.438
(.193)
1.438
(.193)
1
1
(0)
ε
+ 2β
(0)
1.417
(.147)
1.604
(.267)
1.604
(.267)
ε
ε
=
=
α
α
+ 2
β
6
+
β
M=
σ π
4pz or 5pz
Figure 11. Construction of molecular orbitals of bis(amidinourea)metal(II). Dots in the molecular structures represent π electrons.
The isoelectronic hydrocarbon for the ligand is tetramethylenepropane trianion (TMP3¹), of which NBMOs are threefold
degenerate. Isoelectronic hydrocarbon for the complex has fivefold degenerate NBMOs. The numerals on the skeleton of the
complex are the π-electron densities qr and approximate self-atom polarizabilites ³ (in parentheses, unit in ¢¹1).
r,r
necessary explicitly due to cancelation. It is essential that qr
and ¤¡r of the branched amino group are relatively significant,
and thus, contribution to the first-order stability becomes large.
This situation corresponds to the conventional topological
charge stabilization rule. The absolute value of self-atom
Systematic Calculations on Other Metal’s Complexes.
Amidinourea anion forms complexes with various metals.
However, there are few structural studies except for early
speculations. Let us consider complexes of Ni(II), Pd(II),
Zn(II), and Cd(II) as comparative examples of Cu(II). The
Ni(II) complexes are known as orange materials, of which
structures have been speculated to have CuN4 chromophores
assuming uncertain tautomerism.1 The Pd(II) complexes are
also reported in the early literature as light yellow materials,1
but the structures are unknown. Zn(II) complexes are known as
white powder with some crystal waters,1 though the structures
are unknown. A Cd(II) complex is known as white powder.2
Cd(II) is weakly adsorbed by methacryloyl AU resins in basic
solutions,12 which suggests similar chelation. The Cd(II) com-
plexes are of interest in view of environmental chemistry, but
the structure is also unknown. Though Zn(II) and Cd(II) are not
polarizabilities ³ on the branched positions also becomes
r,r
large due to significant NBMO amplitudes on the branched
positions. This contributes to the second-order perturbation,
and stabilizes the CuN2O2 chromophore with significant ¤¡r of
amino groups. This implies that the topological charge stabili-
zation rule can be modified by the second-order perturbation.
The applicability of the extended rule to other systems will be
discussed elsewhere by further studies on other quasi-aromatic
molecules. Thus, as a whole, we can plausibly conclude that
the bis(amidinourea)copper(II) formed under basic conditions
has a trans-CuN2O2 chromophore, not CuN4 chromophores.
170 | Bull. Chem. Soc. Jpn. 2015, 88, 162–172 | doi:10.1246/bcsj.20140130
© 2014 The Chemical Society of Japan

Table 3. Summary of DFT Calculations on Bis(amidinourea)metal(II) Isomers
Metal M(II) Isomers Symmetry Charge on M
Tilt angle
/degree
Relative energy
/kcal mol
©S2ªa)
¹1
Cu(II) 6
Cu(II) 10
trans-CuN2O2
trans-CuN4 (Proton transferred) C2
C2 ¿ C2h
+0.840 (+0.836) 28.3 (34.4)
+0.811 (+0.802) 44.2 (40.6)
0.0 (0.0)
0.7521 (0.7521)
+15.2 (+14.6) 0.7523 (0.7523)
[B3LYP/6-31G(d)// B3LYP/6-31G(d) (in parentheses, B3LYP/6-31G(d,p)//
B3LYP/6-31G(d,p))]
Ni(II)
Ni(II)
trans-CuN2O2
trans-CuN4 (Proton transferred) C2 ¿ C2h
C2 ¿ C2h
+0.725 (+0.723)
+0.718 (+0.712)
2.8 (2.7)
9.8 (10.1)
0.0 (0.0)
0.0000 (0.0000)
+14.6 (+14.1) 0.0000 (0.0000)
[B3LYP/6-31G(d)// B3LYP/6-31G(d) (in parentheses, B3LYP/6-31G(d,p)//
B3LYP/6-31G(d,p))]
Pd(II)
Pd(II)
trans-CuN2O2
trans-CuN4 (Proton transferred) C2 ¿ C2h
C2 ¿ C2h
+0.520
+0.567
0.63
1.21
0.0
+8.3
0.0000
0.0000
[B3LYP/LANL2DZ// B3LYP/LANL2DZ]
Zn(II)
Zn(II)
trans-CuN2O2
trans-CuN4 (Proton transferred) C2
C2
+0.938 (+0.933) 87.4 (87.4)
+0.885 (+0.875) 85.3 (85.3)
0.0 (0.0)
0.0000 (0.0000)
+15.1 (+14.5) 0.0000 (0.0000)
[B3LYP/6-31G(d)// B3LYP/6-31G(d) (in parentheses, B3LYP/6-31G(d,p)//
B3LYP/6-31G(d,p))]
Cd(II)
Cd(II)
trans-CuN2O2
trans-CuN4 (Proton transferred) C2
C2
+1.072
+1.027
85.1
83.3
0.0
+20.7
0.0000
0.0000
[B3LYP/LANL2DZ// B3LYP/LANL2DZ]
a) The Ni(II) and Pd(II) complexes are thought to be diamagnetic.1,8
transition metals, their chelates are interesting as comparative
examples with the Cu(II) complex.
the Zn(II) and Cd(II) are larger than those of the Cu(II), Ni(II),
and Pd(II) complexes (Table 3), two perpendicularly twisted
resonance structures with the metal’s vacant σπ orbitals and
NBMOs of TMP3¹ are probably suitable for the nonperturbed
structures. Then, the total energies of these complexes are
similarly calculated as:
If quasi aromaticity plays a dominant role in the tautom-
erism, these complexes should also have stable CuN2O2
chromophores. Thus, our interest is geometries and the relative
energies of the metal (M)-N2O2 and M-N4 chromophores.
Table 3 is the summary of the calculations. We see that trans-
MN2O2 isomers are stable for all the metal ions. The tilt angle
between the two chelate planes suggests dsp2 hybridization for
Ni(II) and Pd(II) complexes, and sp3 hybridization for Zn(II)
and Cd(II) complexes. In the actual computations, LANL2DZ
basis set was used for Pd and Cd atoms. Then, valence of
Cd(II) is described by sd3 hybridization using 5d atomic orbital
components, rather than sp3 hybridization. In Ni(II) and Pd(II)
complexes, we can expect large quasi-aromaticity effects,
because deviation from planarity is very small. Then, the
vacant 4pz and 5pz atomic orbitals are responsible for π con-
jugations in the Ni(II) and Pd(II) complexes, respectively. The
mechanism of the quasi-aromaticity stabilities of these com-
plexes is similar to that of Cu(II) (eq 3).
In the Zn(II) and Cd(II) complexes, on the other hand, quasi
aromaticity is realized by hyperconjugation of σπ orbitals
consisting of sp3 or sd3 components. The σπ orbitals have π
symmetry, as depicted in the bottom of Figure 11, and con-
jugates with the ligands. This kind of interaction is similar to
those of alkyl benzenes or polyphenols, in which σπ orbitals
consisting of sp3 orbitals on carbon and 1s on hydrogen partic-
ipate in hyperconjugation.35 In general, whether a given π sys-
tem is stabilized or destabilized by hyperconjugation depends
on the situation. In the present cases, the interactions always
stabilize the nonperturbed structures, because the active orbitals
on metal ions are vacant. In the cases of Zn(II) and Cd(II), the
choice of the zero-th order structures is different from those of
the Cu(II), Ni(II), and Pd(II) complexes. Given that charges on
EðMN4Þ ꢁ EðMN2O2Þ
¼ 0:250j¢j ð1st orderÞ þ 0:442j¢j ð2nd orderÞ
¼
ꢁ1
ꢀ
16:0 kcal mol ðM = Zn and CdÞ
ð4Þ
Within the first-order perturbation, the branched amino groups
stabilize the system to a large extent, because qr on branched
carbons are relatively large and ¤¡r of amino nitrogen is
also significant. In addition, it can be easily shown that self-
atom polarizability on the central carbon of the TMP3¹ is
smaller than those of the branched positions, because the
coefficient in the LUMO is zero. Therefore, the second-order
perturbation mainly comes from self-atom polarizabilities on
the branched positions. Given that ¤¡r of amino nitrogen is
more significant than those of inimo nitrogen or oxygen, con-
tributions to the first- and second-order perturbation stability is
mainly due to branching of the amino groups. Thus, we can
conclude that branching of amino groups is more preferable
than those of imino groups or carbonyl oxygen. This is the
essential origin of MN2O2 preference.
As a whole, discussions by the perturbation theory above
are fully consistent with the experiments and DFT results, and
magnitude of the stability of trans-MN2O2 chromophores is
robust. From the estimated energy gaps between the MN2O2
and MN4 chromophores, the equilibrium constants [MN4]/
[MN2O2] becomes less than 10¹6 (in particular, 10¹15 order for
M = Cu) at room temperature. The thermal stability of the
trans-isomer relative to the cis-form is estimated to be
ca. 4 kcal mol¹1, which is large enough to separate it at room
Bull. Chem. Soc. Jpn. 2015, 88, 162–172 | doi:10.1246/bcsj.20140130
© 2014 The Chemical Society of Japan | 171

Res., Synop. 1985, 380.
10 A. W. Trochimczuk, B. N. Kolarz, Eur. Polym. J. 2000, 36,
2359.
11 a) M. Hatanaka, M. Hoshi, R. Shiba, Kobunshi Ronbunshu
2006, 63, 745. b) M. Hatanaka, Non-bonding properties of ureas in
Urea: Synthesis, Properties and Uses, ed. by C. M. Muñoz, A. M.
Fernández, Nova Science Publishers, 2012, New York, Chap. 8,
pp. 213-231.
temperature. The experimental and theoretical results are
fully consistent. Thus, stability of the CuN4 chromophores
that have been supported by early studies is very suspicious.
Despite many studies, the single crystals of [AU ¹ H]2Cu and
the related compounds have not been obtained yet at present.
However, complexes of substituted AUs, say, long-alkyl AUs,
may be crystallized from usual solvents more easily. It is to be
desired that crystallographic studies are performed in the future
by specialists of X-ray or neutron-diffraction analyses to
establish the MN2O2-MN4 energetics.
12 S. Tanaka, M. Hatanaka, T. Shimizu, R. Shiba, Kobunshi
Ronbunshu 2009, 66, 43.
13 R. C. Hirt, R. G. Schmitt, Spectrochim. Acta 1958, 12, 127.
14 C. Lee, W. Yang, R. G. Parr, Phys. Rev. B 1988, 37, 785.
15 A. D. Becke, J. Chem. Phys. 1993, 98, 1372.
16 A. D. Becke, J. Chem. Phys. 1993, 98, 5648.
17 M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria,
M. A. Robb, J. R. Cheeseman, V. G. Zakrzewski, J. A.
Montgomery, Jr., R. E. Stratmann, J. C. Burant, S. Dapprich,
J. M. Millam, A. D. Daniels, K. N. Kudin, M. C. Strain, O. Farkas,
J. Tomasi, V. Barone, M. Cossi, R. Cammi, B. Mennucci, C.
Pomelli, C. Adamo, S. Clifford, J. Ochterski, G. A. Petersson, P. Y.
Ayala, Q. Cui, K. Morokuma, N. Rega, P. Salvador, J. J.
Dannenberg, D. K. Malick, A. D. Rabuck, K. Raghavachari,
J. B. Foresman, J. Cioslowski, J. V. Ortiz, A. G. Baboul, B. B.
Stefanov, G. Liu, A. Liashenko, P. Piskorz, I. Komaromi, R.
Gomperts, R. L. Martin, D. J. Fox, T. Keith, M. A. Al-Laham,
C. Y. Peng, A. Nanayakkara, M. Challacombe, P. M. W. Gill, B.
Johnson, W. Chen, M. W. Wong, J. L. Andres, C. Gonzalez, M.
Head-Gordon, E. S. Replogle, J. A. Pople, Gaussian 98 (Revision
A.11.2), Gaussian, Inc., Pittsburgh PA, 2001.
Conclusion
Coordination structures of bis(amidinourea)copper(II) che-
lates were suggested by substituent effects of complexation,
UV-vis spectra, IR/Raman spectra, and DFT analysis. Con-
trary to early speculations, the anionic amidinourea-Cu(II)
complex plausibly has a trans-CuN2O2 chromophore, not CuN4
chromophores. DFT calculations revealed relative stabilities of
the possible isomers. Vibration analysis and the exclusion rule
for IR/Raman spectra were consistent with the trans-CuN2O2
preference. Systematic DFT calculations on bis(amidinourea)-
Cu(II), -Ni(II), -Pd(II), -Zn(II), and -Cd(II) also supported the
metal-N2O2 preference, and the essential origin of the stability
was clarified by the perturbation theory. The branched amino
groups contribute to the metal-N2O2 preference to a large
extent through the topological charge stabilization.
Thanks are due to Prof. R. Shiba and Dr. S. Tanaka of Tokyo
Denki University for valuable discussion on chemistry of ureas.
Thanks are also due to KOBELCO Research Institute Inc., for
the Raman spectroscopy measurement.
18 T. Ebisuno, M. Takimoto, M. Takahashi, R. Shiba, Bull.
Chem. Soc. Jpn. 1989, 62, 651.
19 Z. B. Maksić, B. Kovačević, J. Org. Chem. 2000, 65, 3303.
20 M. Hatanaka, Comput. Theor. Chem. 2011, 971, 58.
21 S. Ilieva, B. Hadjieva, B. Galabov, J. Mol. Struct. 1999,
476, 151.
Supporting Information
For the bis(amidinourea)copper(II) complexes, fully opti-
mized Cartesian coordinates, details of the energetics, and IR/
Raman simulation results. This material is available electroni-
cally on J-STAGE.
22 R. K. Ray, G. B. Kauffman, Inorg. Chim. Acta 1990, 174,
237.
23 J. R. Wasson, C. Trapp, J. Phys. Chem. 1969, 73, 3763.
24 O. I. Singh, M. Damayanti, N. R. Singh, R. K. H. Singh, M.
Mohapatra, R. M. Kadam, Polyhedron 2005, 24, 909.
25 H. Ünver, A. Elmali, A. Karakaş, H. Kara, E. Donmez,
J. Mol. Struct. 2006, 800, 18.
References
1
2
3
P. Rây, Chem. Rev. 1961, 61, 313.
R. H. Trimble, Anal. Chem. 1962, 34, 1633.
C. R. Saha, D. Sen, S. Guha, J. Chem. Soc., Dalton Trans.
26 M. W. Wong, Chem. Phys. Lett. 1996, 256, 391.
27 J. R. Kincaid, K. Nakamoto, Specrtochim. Acta, Part A
1976, 32, 277.
1975, 1701.
28 K. Kanamori, K. Kawai, Bull. Chem. Soc. Jpn. 1982, 55,
764.
4
P. V. Babykutty, C. P. Prabhakaran, R. Anantaraman,
C. G. R. Nair, J. Inorg. Nucl. Chem. 1977, 39, 913.
29 B. M. Gimarc, J. Am. Chem. Soc. 1983, 105, 1979.
30 B. M. Gimarc, J. J. Ott, J. Am. Chem. Soc. 1986, 108, 4298.
31 J. J. Ott, B. M. Gimarc, J. Am. Chem. Soc. 1986, 108, 4303.
32 J. Aihara, Bull. Chem. Soc. Jpn. 1987, 60, 2268.
33 M. Hatanaka, R. Shiba, J. Comput. Chem., Jpn. 2005, 4,
101.
5
A. Syamal, Indian J. Chem., Sect. A: Inorg., Phys., Theor.
Anal. 1985, 24, 946.
6
M. J. Begley, P. Hubberstey, C. H. M. Moore, J. Chem.
Res., Synop. 1986, 120.
7
M. J. Begley, P. Hubberstey, S. P. Martindale, C. H. M.
Moore, N. S. Price, J. Chem. Res., Synop. 1988, 2.
34 A. Streitwieser, Jr., Molecular Orbital Theory for Organic
Chemists, John Wiley & Sons, New York, 1961.
35 M. Hatanaka, Tetrahedron Lett. 2013, 54, 1452.
8
R. K. Ray, G. B. Kauffman, Inorg. Chim. Acta 1990,
173, 207.
9
M. J. Begley, P. Hubberstey, C. H. M. Moore, J. Chem.
172 | Bull. Chem. Soc. Jpn. 2015, 88, 162–172 | doi:10.1246/bcsj.20140130
© 2014 The Chemical Society of Japan