A study on oxygen insertion in dinuclear silver cryptates

Article abstract of DOI:10.1039/b401776e

Two novel silver(I) cryptates are reported in this paper. [Ag ₂(L¹O)](ClO₄)₂·H ₂O 1 and [Ag₂L¹](ClO₄) ₂·1.5H₂O 2 were synthesized by the condensation of tris (3-aminopropyl) amine with m-phthalaldehyde in the presence of silver(I) ion, under aerobic and anaerobic conditions, respectively. (The ligand L ¹O represents the oxygen insertion product of L¹.) Cryptates 1, 2, and their hydrogenated ligand H₁₂(L¹O) 3 and H₁₂L¹4 (obtained by reduction of the cryptates) were investigated by electrospray mass spectroscopy (ES-MS). 1 and 2 were also decomposed by HCl treatment and their products were separated and identified by HPLC and ES-MS. Our experiments show that cryptate 2 is able to activate dioxygen that results in quantitative aliphatic hydroxylation of L¹ on one of its HC=N bonds. Crystal structure analysis shows an interesting difference between 1 and 2 in that 1 is an oxygenated and 2 is a non-oxygenated cryptate. Up to date, ligand hydroxylation has not been achieved in silver(I) complex-O₂ systems.

Full text of DOI:10.1039/b401776e

View Article Online / Journal Homepage / Table of Contents for this issue
A study on oxygen insertion in dinuclear silver cryptates
Zhi-Lin Wang,a Qin-Hui Luo,*a Chun-Ying Duan,a Cheng-Yu Shenb and Yi-Zhi Li*a
a Coordination Chemistry Institute, State Key Laboratory of Coordination Chemistry,
Nanjing University, Nanjing 210093, P.R. China
b Research and Development Division, PPD Development, 8500 Research way,
Middleton, WI, 53562, USA
Received 5th February 2004, Accepted 11th February 2004
First published as an Advance Article on the web 4th March 2004
Two novel silver() cryptates are reported in this paper. [Ag2(L1O)](ClO4)2ؒH2O 1 and [Ag2L1](ClO4)2ؒ1.5H2O 2 were
synthesized by the condensation of tris (3-aminopropyl) amine with m-phthalaldehyde in the presence of silver() ion,
under aerobic and anaerobic conditions, respectively. (The ligand L1O represents the oxygen insertion product of L1.)
Cryptates 1, 2, and their hydrogenated ligand H12(L1O) 3 and H12L1 4 (obtained by reduction of the cryptates) were
investigated by electrospray mass spectroscopy (ES-MS). 1 and 2 were also decomposed by HCl treatment and their
products were separated and identified by HPLC and ES-MS. Our experiments show that cryptate 2 is able to activate
dioxygen that results in quantitative aliphatic hydroxylation of L1 on one of its HC᎐N bonds. Crystal structure
᎐
analysis shows an interesting difference between 1 and 2 in that 1 is an oxygenated and 2 is a non-oxygenated
cryptate. Up to date, ligand hydroxylation has not been achieved in silver() complex–O2 systems.
firmed that silver() in cryptate 2 was able to activate oxygen
that resulted in aliphatic hydroxylation of the ligand, namely, it
catalyzed oxygen insertion into the carbon–hydrogen bond in
Introduction
There has been growing interest in the field of dioxygen
activation1–3 because of its tremendous importance in catalytic4
one of the H–C᎐N groups to form the H–O–C᎐N group
᎐
᎐
and synthetic chemistry.5 Transition metals, especially the
transition metals in enzyme active sites or in coordination com-
plexes, are known for their ability for activating and breaking
the oxygen–oxygen bond that results in the insertion of oxygen
atoms into organic substrates.1,6,7 Among many such examples
are the non-heme oxygenase catalyzed reactions such as the
aromatic hydroxylation by tyrosinase (tyr)8 and aliphatic
hydroxylation by dopamine β-hydroxylase (dβh).9 Both tyr and
dβh contain two copper ions per catalytical unit that catalyze
the oxygen insertion into the aromatic or aliphatic C–H bond
of the substrates. Aiming at designing new catalysts for
oxidation with molecular oxygen, chemists have paid great
attention to the catalytic properties of dinuclear transition
metal complexes and studied them as models for non-heme
oxygenases.10–13 Karlin and co-workers reported the first
example of dioxygen promoted aromatic ligand hydroxylation
in dinuclear copper complexes.14 Martell and co-workers
presented the first example of arene ring hydroxylation in
macrocyclic copper() complexes.15 In contrast to the growing
examples of aromatic ligand hydroxylation observed in a series
of dinuclear copper() complex–O2 systems, aliphatic ligand
hydroxylation has been rarely reported.16–19
(Scheme 1). X-ray structure analysis confirmed that 1 is an
oxygenated cryptate while 2 is a non-oxygenated cryptate.
Although many authors have reported oxygen insertions
into carbon–hydrogen bonds catalyzed by copper, iron, etc.,
complexes, this is the first example of oxygen insertion
catalyzed by a silver() complex.
Results and discussion
Characterizations
Although cryptate [Ag2(L1O)](ClO4)ؒH2O 1 was synthesized
under atmospheric conditions, and [Ag2L1](ClO4)2ؒ1.5H2O 2
was synthesized under anaerobic condition, both have similar
spectral characteristics. Infrared spectra show the anticipated
ν(C᎐N) stretch at ca. 1640 cmϪ1, and ν(ClO4) strong band at
᎐
1100 cmϪ1 for both cryptates. Solution molar conductivity
values indicated that all four cryptates are 1 : 2 electrolytes.22
Elemental analysis data agreed with the predicted chemical
formula, that 1 is an oxygenated cryptate while 2 is a non-
oxygenated cryptate.
The silver ion was released when cryptates 1 and 2 were
reduced by NaBH4. As a result, H12(L1O) 3 and H12L1 4 were
formed. For both 3 and 4, the 1640 cmϪ1 absorption character-
Because silver and copper possess the same d10 electronic
configuration, their chemical behaviors are very similar. Silver
is an important element with abundant catalytic properties.
Absorbed oxygen on a silver surface is well known in the
catalysis of epoxide formation and oxidative coupling of
methane.20 Silver complexes have also been well documented for
their catalytic abilities.21
To our best knowledge, ligand hydroxylation has not been
achieved in silver() complex–O2 systems. Here we report the
syntheses and characterization of two novel silver() cryptates:
[Ag2(L1O)](ClO4)2ؒH2O 1, [Ag2L1](ClO4)2ؒ1.5H2O 2. Cryptates
1 and 2 were synthesized by condensation of tris(3-amino-
propyl) amine (trpn) with m-phthalaldehyde in the presence of
silver() ion, under aerobic and anaerobic conditions respect-
ively. Cryptates 1 and 2 and their hydrogenated (reduced)
ligands H12L1O and H12L1 were studied by electrospray mass
spectrometry (ES-MS). The HCl decomposed products of 1
and 2 were separated and identified by high performance liquid
chromatography (HPLC) and ES-MS. All experiments con-
istic of the C᎐N band disappeared, while new bands assigned to
᎐
the N–H stretch appeared at 3283 and 3340 cmϪ1, indicating the
reduction of the C᎐N bond. Although elemental analysis data
᎐
show that neither of the hydrogenated ligands have adducted
water or methanol molecules, the IR spectrum of 3 still showed
an O–H vibration at 3427 cmϪ1 and a C–O stretch at 1513 and
1261 cmϪ1, suggesting an inserted oxygen atom in 3.
1H NMR spectra
The 1H NMR spectrum of cryptate 1 is different from that of 2
(Fig. 1(a) and (b)). For 2, the proton signals in the spectrum are
able to be assigned approximately, but the spectrum of 1 is so
complicated that its signals can not be designated one by one.
Near the imino and aromatic regions of the spectrum for 1, the
signal at 9.85 ppm has been assigned to an aromatic proton
23a
reported in the 1H NMR of [Ag2L1](NO3)2
(cryptate 2Ј,
D a l t o n T r a n s . , 2 0 0 4 , 1 1 0 4 – 1 1 1 1
T h i s j o u r n a l i s © T h e R o y a l S o c i e t y o f C h e m i s t r y 2 0 0 4
1104

View Article Online
Scheme 1 The syntheses of cryptates and cryptands (charges in the structural formulas are omitted).
Fig. 1 1H NMR spectra (in (CD3)2SO solution) of (a) cryptate 1, (b) freshly prepared solution of 2, (c) cryptate 2, after a week, 500 MHz at room
temperature. The signals at ca. 3.35 ppm and 2.5 ppm were those of (CD3)2SO and traces of water, respectively.
Ϫ
whose complex ion is the same as that of 2, but with NO3
Ϫ
instead of ClO4 in 2). However, we cannot provide evidence
for this opinion. The signal at 9.85 ppm was not found in
the spectrum of 2. The freshly prepared (CD3)2SO solution of
2 was stored in a stoppered NMR tube; after a week, its
1H NMR spectrum changed into the same spectrum as 1
(Fig. 1(c)). Therefore it is reasonable to assign the signal at
9.85 ppm to the hydroxy-proton of the HO–C᎐N group,
᎐
namely, by leaving 2 to the atmosphere, 2 was converted into
the oxygen inserted cryptate 1. Furthermore, in the spectrum
of the reduced products of 1 (cryptand H12L1O), the signal
at 9.85 ppm disappeared, and a weak hydroxy-proton signal
at 4.24 ppm appeared. These facts provide evidence for
oxygen insertion. In addition, the imine signal of 2Ј at 8.46
ppm is a duplet, while the corresponding signal of 2 (Fig. 1(b))
looks like a singlet, and that of 1 is split. The differences
between the NMR spectrum of 2Ј and that of 1 or 2 are not
able to be explained clearly at the moment. Addition work is
in progress.
Fig. 2 ES-MS spectrum of cryptate 1 in MeOH.
ES-MS spectra between 1 and 2 is informative. For cryptate 2,
the peak at m/z = 671.5 ([L1 ϩ Hϩ]ϩ) has an m/z difference of 16
(one oxygen atom) compared to the peak of m/z = 687.5 ([L1O
ϩ Hϩ]ϩ) in 1. Similarly, for 2, the fragment [AgL1 ϩ Hϩ]ϩ at m/z
= 779.4 also has a m/z difference of 15 compared to that of
[Ag(L1O)]ϩ at m/z = 794.5 in 1. All this evidence indicates that
an oxygen is indeed inserted in 1.
Electrospray mass spectra of 1 and 2
The ES-MS data of compounds 1–4 are listed in Table 1. The
ES-MS spectrum of 1 (Fig. 2) displays three major peaks at m/z
= 687.5(43), 794.5(13), and 946.1(10), assigned to the fragments
Ϫ
of [(L1O) ϩ Hϩ]ϩ, [Ag(L1O)]ϩ, and [Ag(L1O) ϩ ClO4
ϩ
ES-MS spectra of hydrogenated ligands
MeOH ϩ H2O ϩ 2Hϩ ϩ e]ϩ, respectively. This supports the
incorporation of one oxygen atom into the C–H bond of the
The ES-MS spectrum of 3 (Fig. 3) has five major peaks. Three
of them, at m/z = 699.7(75), 391.3(20) and 349.5(66), were
assigned to the proton or solvent adducts of the oxygenated
ligand [H12(L1O) ϩ Hϩ]ϩ, [H12(L1O) ϩ 2MeOH ϩ H2O ϩ
2Hϩ]2ϩ, and [H12(L1O) Ϫ 2e]2ϩ, respectively. In contrast to 3, the
ES-MS spectrum of 4 only has two peaks and neither origin-
ated from ligand oxidation. Whereas leaving cryptate 2 in an
oxygen atmosphere converts 2 to the oxygen inserted cryptate 1,
HC᎐N group. The fragmentation pattern also shows that the
᎐
oxygen in 1 is not very stable and is easy to lose under ES-MS
conditions, evidenced by other peaks assigned to the fragments
losing an oxygen or silver ion. This implies that insertion is
reversible. Fragments of proton or solvent adducts were also
observed. Cryptate 2 is quite different from 1, since no peaks
were found due to the insertion of oxygen. The comparison of
D a l t o n T r a n s . , 2 0 0 4 , 1 1 0 4 – 1 1 1 1
1105

View Article Online
Table 1 Assignments of ES-MS peaks for some compounds in MeOHa
No.
Compound
Peaks (m/z)
Assignments
1
[Ag2(L1O)](ClO4)2
946.1 (10)
794.5 (13)
778.3 (32)
687.5 (43)
671.5 (100)
443.3 (41)
336.3 (32)
[Ag(L1O) ϩ ClO4Ϫ ϩ MeOH ϩ H2O ϩ 2Hϩ ϩ e]ϩ
[Ag(L1O)]ϩ
[AgL1]ϩ
[(L1O) ϩ Hϩ]ϩ
[L1 ϩ Hϩ]ϩ
[Ag2L1]2ϩ
[L1 ϩ 2Hϩ]2ϩ
2
[Ag2L1](ClO4)4
985.1 (37)
811.4 (9)
779.4 (48)
671.5 (18)
459.4 (8)
443.5 (100)
389.5 (11)
336.3 (37)
[Ag2L1 ϩ ClO4Ϫ Ϫ H]ϩ
[AgL1 ϩ MeOH ϩ H]ϩ
[AgL1 ϩ H]ϩ
[L1 ϩ Hϩ]ϩ
[Ag2L1 ϩ MeOH]2ϩ
[Ag2L1]2ϩ
[AgL1 ϩ Hϩ]2ϩ
[L1 ϩ 2Hϩ]2ϩ
3
4
H12(L1O)
699.7 (75)
683.7 (100)
398.3 (7)
391.3 (20)
349.5 (66)
342.5 (96)
[H12(L1O) ϩ Hϩ]ϩ
[H12L1 ϩ Hϩ]ϩ
[H12(L1O) ϩ 3MeOH ϩ 2Hϩ]2ϩ
[H12(L1O) ϩ 2MeOH ϩ H2O ϩ 2Hϩ]2ϩ
[H12(L1O) Ϫ 2e]2ϩ
[H12L1 ϩ 2Hϩ]2ϩ
H12 L1
683.7 (83)
342.5 (100)
[H12L1 ϩ Hϩ]ϩ
[H12L1 ϩ 2Hϩ]2ϩ
a Relative abundances are given in parentheses.
Fig. 3 ES-MS spectrum of hydrogenated ligand H12(L1O).
leaving H12L1 under an oxygen atmosphere for one week did not
produce oxygen insertion products, indicating a critical role of
Ag() in such an oxygen insertion reaction.
Investigation of acid decomposed products
To further confirm that 1 and 2 have different compositions, we
treated both complexes with HCl and followed their decom-
position with ES-MS and HPLC. The decomposition products
of cryptate 1 are more complicated. Its two fractions obtained
from a silica gel column were measured by ES-MS spectro-
scopy. The ES-MS spectrum of the second fraction (Fig. 4(a))
displays two peaks. The dominant one is at m/z = 151.1(100)
which corresponds to [C8H6O3 ϩ Hϩ]ϩ, the proton adduct of
m-formyl phenyl formic acid. A second peak at m/z = 301.2(38)
was assigned to [(C8H6O3)2 ϩ Hϩ]ϩ, the proton adduct of
dimeric m-formyl phenyl formic acid. For the first fraction,
the ES-MS spectrum (Fig. 4(b)) displays the base peak of
m-phthalic aldehyde [C8H6O2 ϩ Hϩ]ϩ at m/z = 135.0(100) and
its dehydrated product [(C8H6O2 Ϫ H2O) Ϫ e]ϩ at m/z =
116.0(29). The two solids obtained from two fractions have mp
89–91Њ and 174Њ which are in accordance with those of m-
phthalic aldehyde and m-formylbenzoic acid respectively. For
cryptate 2, only one fraction was obtained and its ES-MS
spectrum was the same as Fig. 4(b). It is the base peak of
m-phthalic aldehyde [C8H6O2 ϩ Hϩ]ϩ at m/z = 135.0(100) and
Fig. 4 ES-MS spectra of acid decomposed products: (a) cryptate 1, (b)
cryptate 2.
its dehydrated product [(C8H6O2 Ϫ H2O) Ϫ e]ϩ at m/z =
116.0(29).
The acid decomposed products of 1 and 2 were also injected
into HPLC for separation and characterization. The chromato-
gram of 1 displayed two bands with retention times of 2.39 and
4.34 min. They were identified as m-phthalic aldehyde (2.39
min) and m-formyl phenyl formic acid (4.34 min) by standard
D a l t o n T r a n s . , 2 0 0 4 , 1 1 0 4 – 1 1 1 1
1106

View Article Online
Table 2 Comparison of crystallographic data for some dinuclear silver cryptatesa
Cryptate
No.
[Ag2(L1O)](ClO4)2ؒH2O
1
2[Ag2(L1)](ClO4)4ؒ3H2O
2
[Ag2(L1)](NO3)2
2Ј
Formula
M
Crystal system
Space group
C42H56Ag2Cl2N8O10
1119.59
Triclinic
P(Ϫ1)
C84 H114 Ag4 Cl4 N16 O19
2225.19
Triclinic
P(Ϫ1)
C42H54Ag2N10O6
1010.69
Rhombohedral
R3c
T /K
a/Å
b/Å
c/Å
α/Њ
β/Њ
293(2)
293(2)
293(2)
15.82(3)
15.82(3)
34.58(1)
10.381(6)
11.929(5)
19.735(9)
83.37(2)
89.08(4)
81.58(4)
2401(2)
2
10.417(2)
11.857(3)
19.818(6)
83.56(19)
89.20(19)
81.774(15)
2407.3(11)
1
γ/Њ
V/Å3
Z
6
Dc/g cmϪ3
1.549
1.535
F000
1144
0.989
1138
0.985
µ/mmϪ1
0.835
Range for data collection
Number of data measured
Number of data with I > 2σ(I )
R(int)
Goodness of fit
R1 (I > 2σ(I ))
1.92 < θ < 25
9976
8456
0.013
1.013
0.0479
0.1428
1.93 < θ < 25
10000
8475
0.019
1.06
0.0563
0.1414
3309
2957
0.0530
1.094
0.0635
0.1777
wR2 (I > 2σ(I ))
2
2
a R1 = ͚||Fo| Ϫ |Fc||/͚|Fo|; wR2 = [͚w(Fo2 Ϫ Fc2)2/͚w(Fo )2]1/2, w = 1/[σ2(Fo ) ϩ (aP)2 ϩ bP], where P = [Fo2 ϩ 2Fc2]/3.
solution injections. The area ratio of the two peaks is measured
to be 66.9/33.02, in agreement with the value of 66.7/33.3
calculated from the proposed chemical formula of 1, confirm-
ing the quantitative ligand hydroxylation. By contrast, the
chromatogram of 2 treated with HCl only displayed one band
with a retention time of 2.39 min. We also treated m-phthalic
aldehyde with Agϩ ion and no m-formylphenyl acid was
formed. This shows that the formic acid in the final products
is not the oxidation product by Agϩ. Indeed, it is the product of
oxygen insertion in the cryptate. The mechanism for acid
decomposition reaction is suggested in Scheme 2.
Ag(2) and N(4) in cryptate 2 or 1. The average bond length of
Ag–N (imino) is 2.299(5) Å for 1 and 2.300(4) Å for 2. The
average Ag–N (bridge) bond lengths in 1 and 2 are 2.425(4) and
2.439(4) Å respectively. All three cryptates of 1, 2, and 2Ј retain
a trigonal pyramidal N4 cap derived site for Ag() coordination,
but the crystal cell parameters of 1 and 2 are different from 2Ј.
In 1, Ag() and Ag() ions deviated from the plane composed of
three imino nitrogen atoms by an average of 0.0894(2) Å. In 2,
Ag() and Ag() ions deviated from the plane composed
of three imino nitrogen atoms by an average of 0.0873(2) Å.
However, in 2Ј, the Ag() lies almost coplanar with the imino
nitrogen (0.08 Å). The separation of Ag ؒ ؒ ؒ Ag for 1 (3.974(2)
Å) is the largest among the three analogues with the three
m-xylyl-spaced cryptates (Table 3). Also, the separations of
N(br) ؒ ؒ ؒ N(br) for 1 and 2 are larger than that of analog 2Ј.
For cryptate 1, this large separation favors the attack of the
dioxygen on Ag() and the activation of oxygen that leads to
oxygen insertion into the C–H᎐N group. It is noted that oxygen
᎐
insertion resulted in the formation of C(32)–O(2) and C(25)–
O(1) disordered bonds with an average distance of 1.475(2) Å,
slightly longer than normal C–O bond. The C–O bond lengths
and angles in complex 1 are listed in Table 4. The occupancy
probability of oxygen atoms O1 and O2 at each of the bonds is
0.502(11) and 0.498(11) Å, respectively.
CCDC reference numbers 201193 and 201194.
See http://www.rsc.org/suppdata/dt/b4/b401776e/ for crystal-
lographic data in CIF or other electronic format.
Interestingly, slow evaporation of the MeOH solution of
non-oxygenated cryptate 2 in atmospheric conditions produced
crystals with the same X-ray structure as the oxygenated
cryptate 1. It shows that cryptate 2 absorbs oxygen and its
conversion to cryptate 1 is easy in solution. In general, trans-
Scheme 2 The mechanism for acid decomposition.
ition metal ions are able to catalyze the hydrolysis of the C᎐N
᎐
Crystal structures of 1 and 2
bond via the addition of a water molecule across one imine
Crystal structure data of cryptates 1 and 2 are summarized in
Table 2. Table 2 also includes cyptate 2Ј, the nitrate salt of
[Ag2L1]2ϩ crystallized from EtOH–MeCN after condensation
of trpn with m-phthalic aldehyde in the presence of AgNO3.23b
It was reported that no oxygen was inserted into the ligand for
cryptate 2Ј. The structures of 1 and 2 are shown in Fig. 5. They
have similar crystal cell parameters and similar triple-helical
structure. A C3 or pseudo-C3 axis passes through N(1), Ag(1),
bond in Schiff-base ligands,24 but complex 1 is not formed by
partial hydrolysis of the C᎐N bond. The reasons are: (1) if
᎐
complex 1 was formed according to the hydrolysis mechanism,
the acid decomposed product of 1 would be only m-phthalic
aldehyde, not the mixture of m-formyl benzolic acid and
m-phthalic aldehyde.24 (2) Hydrolysis of C᎐N would form
᎐
a single HOC–NH bond by water addition. In complex 1,
however, bond lengths of C(32)–N(7) and C(25)–N(6) are
D a l t o n T r a n s . , 2 0 0 4 , 1 1 0 4 – 1 1 1 1
1107

View Article Online
Table 3 Comparison of average bond distances (Å) or angles (Њ) in dinuclear silver cryptates
Cryptate
No.
[Ag2(L1O)](ClO4)2ؒH2O
1
[Ag2(L1)](ClO4)2ؒ1.5H2O
2
[Ag2(L1)](NO3)2
2Ј
Ag(1)–Ag(2)
N(br)–N(br)
Ag–N(im)
Ag–imine plane
Ag–N(br)
3.974(2)
8.822(2)
2.299(5)
0.0894(2)
2.425(4)
3.954(2)
8.830(2)
2.300(4)
0.0873(2)
2.439(4)
3.775
8.610
2.310
0.08
2.42
N(im)–Ag–N(im)
119.85(17)
119.86(15)
119.9
Table 4 Bond lengths (Å) and angles (Њ) of C–O bond in cryptate 1
C(32)–O(2)
1.500(10)
C(25)–O(1)
1.450(10)
N(7)–C(32)–O(2)
N(7)–C(32)–C(34)
C(24)–C(25)–O(1)
126.0(7)
121.2(5)
109.3(6)
C(34)–C(32)–O(2)
N(6)–C(25)–O(1)
N(6)–C(25)–C(24)
112.5(6)
129.0(6)
121.5(5)
Ϫ
the difference of the anions in the cryptates (NO3Ϫ in 2Ј, ClO4
in 1 and 2).
Martell and co-workers15 reported the activation of dioxygen
and subsequent hydroxylation of the arene ring in an analogous
planar macrocyclic copper() complex derived from m-phthal-
aldehyde with diethylenetriamine. Silver() and copper() have
the same d10 electronic configuration and the absorbed oxygen
on a silver surface is well known in catalytic reactions. An
analogous triple-helical Ag() cryptate derived from trpn and
pyridine-2,6-dicarboxyaldehyde has been found to be able to
catalyze the hydrolysis of acetonitrile.25a In cryptate 2, the
triple-helical structure of the complex and better flexibility of
the ligand, both play an important role in the oxygen activation
and subsequent aliphatic ligand hydroxylation. In its crystal
structure, the distance between Ag() and the carbon of the
C᎐N bonds (3.249–3.294 Å) is shorter than those between
᎐
Ag() and the phenyl carbon atoms (C(5), C(9), and C(33),
3.294–3.408 Å), making it favorable for C᎐N bond oxygen
᎐
insertion instead of arene hydroxylation.
As a result of the above, undoubtedly, the silver() in complex
2 is able to activate oxygen resulting in aliphatic hydroxylation.
The reaction mechanism is as follows (Scheme 3)
Fig. 5 Crystal structures of (a) [Ag2(L1O)]2ϩ, (b) [Ag2L1]2ϩ
.
about 1.246(7) and 1.246(7) Å, respectively, shorter than the
average single bond C–N length of 1.43 Å. Therefore these
bonds are more reasonably designated as a double bond. (3)
The average bond angles of 117Њ on both C(32) and C(25) are
characteristic of an sp2 hybridised configuration and not
of an sp3 configuration for a single bond, indicating the
existence of a double bond between carbon and its neigh-
boring imino nitrogen atom on both carbon atoms. The
structural difference between 2Ј and 1 or 2 maybe results from
Scheme 3 The oxygen insertion mechanism.
By step I, in anaerobic conditions the non-oxygenated
complex 2 was synthesized and then by step II, the reaction of
complex 2 with dioxygen in anaerobic conditions resulted in the
formation of a new silver cryptate with silver–oxo bond as an
intermediate. The formation of an Ag᎐O bond is a result of
᎐
D a l t o n T r a n s . , 2 0 0 4 , 1 1 0 4 – 1 1 1 1
1108

View Article Online
silver promoted activation and breakage of dioxygen. Finally,
by step III the oxygen located by the silver atom was activated
by the silver, resulting in insertion of the oxygen into the C–H
bond (cryptate 1). In this reaction, the silver–oxo intermediate
can not be isolated, because it is unstable. However, its silver–
oxo analog synthesized using p-phthalaldehyde instead of
m-phthalaldehyde has been obtained under aerobic conditions
and structurally characterized by X-ray analysis.25b Recently,
the existence of a silver–oxo analog has been confirmed again
by a comparative study on their ES-MS, ES-MS-MS spectra
with a non-oxygen counterpart. These results not only confirm
a Siemens P4 four-circle diffractometer with monochromated
Mo Kα radiation (λ = 0.71073 Å) using θ/2θ scan mode with
variable scan speed 5.0–50.0Њ minϪ1 in ω. Samples of size
0.30 × 0.20 × 0.20 mm3 for 1 and 0.30 × 0.25 × 0.20 mm3 for
2 were mounted on glass fibers. Reflection data were collected
in the range 1.92 < θ < 25Њ. The data were corrected for
Lorentz and Polarization effects during data reduction using
XSCANS.27
The structure was solved by weight atom method and refined
on F 2 by full matrix least squares methods using SHELXTL
version 6.10.28 All the non-hydrogen atoms were refined
anisotropically. Hydrogen atoms were located and added
to the structure. Factor calculations were carried out using
SHELXTL-PC Program Package.28
the presence of Ag᎐O species, but also explain the mechanism
᎐
of oxygen insertion.
Conclusions
Preparations of [Ag2(L1O)](ClO4)2ؒH2O 1 and [Ag2L1](ClO4)2ؒ
Two cryptates, [Ag2(L1O)](ClO4)2 H2O 1, [Ag2L1](ClO4)2ؒ
1.5H2O 2, were synthesized under aerobic (1) and anaerobic (2)
conditions. The ES-MS spectra of 1 and its hydrogenated
ligand H12(L1O) display fragments corresponding to oxygen
insertion products, while the ES-MS spectra of 2 and H12L1
do not display these peaks. Leaving H12L1 under atmospheric
conditions did not produce an oxygen insertion product, while
slowly evaporating the solution of 2 under atmospheric con-
ditions produced single crystals of an oxygen insertion product,
indicating a critical role of Ag() in such oxygen insertion
reactions.
1.5H2O 2
To a 15 mL absolute methanol solution of AgNO3 (0.085 g,
0.5 mmol) and m-phthalic aldehyde (0.1005 g, 0.75 mmol) was
added dropwise a 15 mL absolute methanol solution of trpn
(0.095 g, 0.5 mmol) for over 30 min. The reaction mixture was
then stirred in the dark at 25–30 ЊC for 8 h. After the reaction
was complete, the solution was filtered to remove small amount
of brown silver residue. Then, a 5 mL absolute methanol
solution of anhydrous sodium perchlorate (0.4 g, 4 mmol) was
added to the filtered solution. White microcrystals of 1 formed
immediately and could be obtained after filtration. Yield:
The acid decomposed products of 1, m-formylphenyl benzoic
acid and m-phthalic aldehyde, were identified by HPLC and
ES-MS. The m-formylphenyl formic acid was not formed
from oxidation by the Agϩ ion, but was formed from oxygen
0.208 g, 72%. IR (KBr): ν = 3452 (br, OH); 1644 (s, C᎐N);
1107 (s, ClO4 ). H NMR ((CD3)2SO): δ: 9.85 (s, ∼1H); 8.61
᎐
Ϫ
1
(d, H–C᎐N); 7.90, 7.77 (m, H-Ph); 2.92, 2.54 (m, H C–N); 1.84
᎐
2
(m, H CN᎐); 1.67 (m, –CH –). Λ (MeOH, 293 K): 318 S cm2
᎐
2
2
m
insertion into the C᎐N bond of the ligand. The crystal
᎐
molϪ1. Elemental analysis: Calculated (%) for C42H56Ag2Cl2-
N8O10: C 45.06, H 5.04, N 10.00, Ag 19.27; found: C 44.58, H
5.38, N 9.58, Ag 18.86. Crystals suitable for X-ray structure
determination were obtained by slow evaporation of the filtered
final solution.
structural difference between 1 and 2 was also caused by oxygen
insertion.
Experimental section
[Ag2L1](ClO4)2ؒ1.5H2O 2. Synthesis of 2 was performed with
Schlenck equipment under an argon atmosphere. Methanol was
strictly dried and deoxygenated before use. Freeze–thaw
degassed AgNO3 (0.085 g, 0.5 mmol) and m-phthalic dialde-
hyde (0.1005 g, 0.75 mmol) were dissolved in 15 mL of MeOH.
Under stirring, to this solution was added dropwise a methanol
solution of trpn (0.095 g, 0.5 mmol) over 30 min. After a few
minutes, a small amount of brown gel appeared. The mixture
was kept stirring in the dark at 25–30 ЊC for 8 h. After reaction
was completed, the solution was filtered on a Schlenck line. 5
mL methanol solution of soldium perchlorate (0.4 g, 3 mmol)
was added to the filtered solution. A white precipitation formed
immediately. The precipitates were filtered, washed, and dried
in the Schlenck equipment. Yield: 0.21 g, 76.1%. IR (KBr):
Materials
Schlenk technique was used to prepare cryptates 2 and 4 under
a dry argon atmosphere. Argon (99.99%) was first dried over
4 Å molecular sieves, and then deoxygenated by silver sieves
and K–Na sieves sequentially. All glassware was first dried and
deoxygenated in a vacuum oven, and then purged with argon for
at least 20 min. Solvents and reagents were deoxygenated on
Schlenkequipmentbythefreeze–thawmethodatleastthreetimes.
Tris(2-aminopropyl)amine (99%) and m-phthalic aldehyde
were synthesized using literature methods.26 Solvents were dried
and purified by standard methods and distilled under argon
prior to use.
Physical measurements
Ϫ
1
ν = 3412 (br, OH); 1640 (s, C᎐N); 1110 (s, ClO ). H NMR
᎐
4
((CD ) SO, 500 MHz): δ: 8.45 (s, 6H, H–C᎐N); 7.26 (s, 12H,
Elemental analyses were performed on a Perkin-Elmer 240
analytical instrument. The molar electrical conductivity (10Ϫ4
mol LϪ1) was measured in MeCN at 26 0.1 ЊC using a BSD-A
conductometer. Infrared spectra were recorded using KBr discs
on a Nicolet 5DX FTIR spectrometer. Electrospray mass
spectra were recorded on a Finigan LCQ mass spectrometer
with sample concentrations of approximately 1.0 µmol LϪ1 in
50 : 50 methanol–water. The solutions were electrosprayed at a
flow rate of 5 × 10Ϫ6 L minϪ1 with needle voltage of ϩ4.5 kV.
The acid decomposed products were identified by a Varian
SY-5000 HPLC on a µ-C18 column, and the mobile phase was
70 : 30 acetonitrile–water for 1 at a flow rate of 0.6 mL minϪ1.
1H NMR spectra were obtained in (CD3)2SO solution on a
Brucker AM 500 spectrometer. Proton chemical shifts are
reported in ppm using TMS as internal reference.
᎐
3
2
H-Ph); 3.85 (s, 12H, H C–N); 2.61 (m, 12H, H CN᎐); 1.94 (m,
᎐
2
2
12H, –CH2–). Λm (MeCN, 293 K): 310 S cm2 molϪ1. Elemental
analysis: Calculated (%) for C42H57Ag2Cl2N8O9.5: C 45.34, H
4.89, N 10.07, Ag 19.39; found: C 45.58, H 5.18, N 10.48, Ag
19.56.
Crystal growth: Colorless crystals suitable for structure
determination were obtained by storing the filtered final
solution in an argon atmosphere for 48 h. The crystals obtained
are stable in air.
Preparation of H12(L1O) 3
The reduction of cryptand L1 was performed in Schlenck
equipment in order to avoid the oxidation of L1. A suspension
of cryptate 1 (0.552 g, 0.5 mmol) in 30 ml pretreated methanol
was reduced by sodium borohydride in an ice-bath. When the
reaction was complete, the brown precipitates were filtered and
then dried in vacuum. 10 mL water was added to the filtrate to
X-ray structure determinations
Single-crystal X-ray diffraction data were collected at 293 K on
D a l t o n T r a n s . , 2 0 0 4 , 1 1 0 4 – 1 1 1 1
1109

View Article Online
(c) L. Jr. Que and W. B. Tolman, Angew. Chem. Int. Ed., 2002, 41,
1114–1137.
make the solution more basic. The product was extracted with
180 mL chloroform in three batches from aqueous solution.
The extraction layer was dried with anhydrous Na2SO4. After
Na2SO4 was removed, a white wax-like substance was obtained
by evaporating the filtrate under reduced pressure. Yield: 0.205
g, 60%. IR (KBr): ν = 3427 (br, OH); 3283 (w, NH); 2924, 2798
(s, CH), 1615, 1458 (m, benzene ring), 1513, 1261 (m, CO). 1H
NMR (CD3SO, 500 MHz): δ: 7.15 (s, 12H, H-Ph); 4.45 (m, 1H,
HO–C); 3.55 (s, 11H, H-benyzl); 2.43 (m, 12H, benzyl-N–CH2);
3 (a) M. A. Kopf and K. D. Karlin, in Biomimetic Oxidations, ed.
B. Meanier, Imperial College Press, London, 1999, ch. 7; (b) D. Lee,
B. Pierce, C. Krebs, M. P. Hendrich, B. H. Huynh and S. J. Lippard,
J. Am. Chem. Soc., 2002, 124, 3993–4007; (c) J. M. Rowland,
M. M. Olmstead and P. K. Mascharak, Inorg. Chem., 2002, 41,
2754–2760.
4 (a) K. D. Karlin and A. D. Zuberbühler, in Bioinorganic Catalysis,
ed. K. J. Reedij and E. Bouwman, Marcel Dekker, New York, 1999,
ch. 14; (b) B. de Bruin, T. P. J. Peters, S. Thewissen, A. N. J. Blok,
J. B. M. Wilting, R. de. Gelder, J. M. M. Smits and A. W. Gal,
Angew. Chem. Int. Ed., 2002, 41, 2135–2137.
5 (a) D. H. R. Barton, A. E. Martell, D. T. Sawyer, The Activation
of Dioxygen and Homogenous Catalytic Oxidation, Plenum,
New York, 1993; (b) M. A. Kopf and K. D. Karlin, in Biomimetic
Oxidations Catalyzed by Transition Metal Complexes, ed.
B. Meunier, Imperial College Press, London, 2002, pp. 309–362; (c)
N. W. Aboelella, E. A. Lewis, A. M. Reynolds, W. W. Brennessel,
C. J. Cramer and W. B. Tolman, J. Am. Chem. Soc., 2002, 124,
10660–10661.
6 (a) Bois J. Du, T. J. Mizoguchi and S. J. Lippard, Coord. Chem. Rev.,
2000, 200–202, 443–485; (b) D. A. Rockcliffe and A. E. Martell,
Inorg. Chem., 1993, 32, 3143–3144.
7 (a) W. B. Tolman, Acc. Chem. Res., 1997, 30, 227–237; (b)
B. J. Wallor and J. D. Lipscomb, Chem. Rev., 1996, 96, 2625–2655.
8 (a) K. D. Karlin, Science, 1993, 261, 707–708; (b) D. Ghosh and
R. Mukherjee, Inorg. Chem., 1998, 37, 6597–6605.
9 (a) J. Klinman, Chem. Rev., 1996, 96, 2541–2561; (b) E. I. Solomon,
U. M. Sundaram and T. E. Machonkin, Chem. Rev., 1996, 96,
2563–2605.
10 (a) D. Ghosh and R. Mukherjee, Inorg. Chem., 1998, 37, 6597–6505;
(b) S. Ryan, H. Adams, D. E. Fenton, M. Becker and S. Schindler,
Inorg. Chem., 1998, 37, 2134–2140.
11 (a) G. Alzuet, L. Casella, M. L. Villa, O. Carugo and M. Gullotti,
J. Chem. Soc., Dalton Trans., 1997, 4789–4794; (b) K. D. Karlin,
M. S. Nasir, B. I. Cohen, R. W. Cruse, S. Kaderli and
A. D. Zuberbühler, J. Am. Chem. Soc., 1994, 116, 1324–1336.
12 (a) O. J. Gelling and B. L. Feringa, J. Am. Chem. Soc., 1990, 112,
7599–7604; (b) C.-Y. Shen, M.-F. Hu, Q.-H. Luo and M.-C. Shen,
J. Inorg. Biochem., 1997, 68, 195–200.
13 (a) A. E. Martell, R. J. Motekaitis, D. A. Rockcliff and A. Llobert,
J. Mol. Catal. A: Chem., 1997, 107, 205–213; (b) K. Chen and L. Jr.
Que, J. Am. Chem. Soc., 2001, 123, 6327–6337; (c) A. Bassan,
M. R. A. Blomberg, P. E. M. Siegbahn and L. Jr. Que, J. Am. Chem.
Soc., 2002, 124, 11056–11063.
14 (a) K. D. Karlin, Y. Gultneh, J. P. Hutchinson and J. Zubieta,
J. Am. Chem. Soc., 1982, 104, 5240–5242; (b) K. D. Karlin,
J. C. Hoyes, R. W. Cruse, J. W. Mckown, J. P, Hutchinson and
J. Zubieta, J. Am. Chem. Soc., 1984, 106, 2121–2128.
15 (a) R. Menif and A. E. Martell, J. Chem. Soc., Chem. Commun.,
1989, 1521–1523; (b) R. Menif, A. E. Martell, P. J. Squattrito and
A. Clearfeid, Inorg. Chem., 1990, 29, 4723–4729.
16 (a) I. Blain, P. Bruno, M. Giorgi, E. Lojou, D. Lexa and M. Reglier,
Eur. J. Inorg. Chem., 1998, 1297–1299; (b) S. Mahapatra,
J. A. Halfen and W. B. Tolman, J. Am. Chem. Soc., 1998, 118,
11575–11586.
17 (a) V. Mahadevan, Z. Hou, A. P. Cole, D. E. Root, T. K. Lai,
E. I. Solomon and T. D. P. Stack, J. Am. Chem. Soc., 1997, 119,
11996–11998; (b) W. E. Allen and T. N. Sorrell, Inorg. Chem., 1997,
36, 1732–1734.
2.22 (s, 12H, CH N᎐); 1.46 (m, 12H, –CH –). Elemental
᎐
2
2
analysis: Calculated for C42H66N8O: C 72.23, H 9.45, N 16.05;
found: C 72.08, H 9.25, N 15.80.
Preparation of H12L1 4
The preparation of 4 was the same as for H12(L1O) 3. A white
wax-like substance was obtained as the product. Yield: 0.205 g,
60%. IR (KBr): ν = 3443 (w, NH); 2954, 2795 (s, CH), 1615,
1
1457 (m, benzene ring). H NMR (CD3SO, 500 MHz): δ: 7.48
(s, 12H, H-Ph); 4.21 (s, 12H, H-benzyl); 3.09 (s, 12H, benzyl-
N–CH ); 2.91–2.94 (m, 12H, CH N᎐); 1.94 (m, 12H, –CH –).
᎐
2
2
2
Elemental analysis: Calculated for C42H66N8: C 73.92, H 9.67,
N 16.41; found: C 74.08, H 9.85, N 16.40.
Acid decomposed products
Cryptate 1 (0.577 g, 0.507 mmol) or 2 (0.552 g, 0.50 mmol) was
dissolved in 15 mL of 1 mol LϪ1 HCl and the solution was
stirred for 1 h at 40–50 ЊC. White precipitates were removed by
filtration. The filtered solution was extracted three times in
chloroform (3 × 20 mL). The chloroform layer was then dried
with anhydrous Na2SO4 before chloroform was slowly evapor-
ated under reduced pressure. A white mixture of decomposition
products (0.186 g for 1, 0.183 g for 2) was obtained. Two
methods were used to determine its composition. In the first
method, a methanol solution of the white mixture was chrom-
atographed on a silica gel column of 15 × 1.5 cm (grade
62 special). The first fraction was eluted with a mobile phase of
60 : 40 v/v of petroleum (bp 90–110 ЊC) and MeCN. The second
fraction was eluted with 100% ethanol. The solvents of the
collected fractions were slowly evaporated and the products
were recrystallized in chloroform. A white powder of m-
phthalic aldehyde C8H6O2 was obtained from the first fraction.
mp 89–91 ЊC (lit. 89–90 ЊC)26b; ES–MS (50 : 50, MeOH–H2O):
m/z = 135.0 (100) [C8H6O2 ϩ Hϩ]ϩ, 116.0 (29) [(C8H6O2 Ϫ H2O)
Ϫ e]ϩ. A pale yellow powder of m-formylphenyl benzoic acid
(C8H6O3) was obtained from the second fraction. mp 174 ЊC
(lit. 175 ЊC)29; ES-MS (50 : 50, MeOH Ϫ H2O): m/z = 151.1
(100) [C8H6O3]ϩ, 301.2 (38) [(C8H6O3)2 ϩ Hϩ]ϩ.
In the second method, the composition of the mixture was
determined by HPLC analysis using a µ-C18 column. The
chromatogram displayed two bands with retention times of
4.34 min (m-formylphenyl benzoic acid) and 2.39 min
(m-phthalic aldehyde) each. For cryptate 2, only one product
identified as C8H6O2 was obtained as the decomposition
product.
18 S. Itoh, H. Nakao, L. Berreau, T. Kondo, M. Komatsu and
S. Fukuzumi, J. Am. Chem. Soc., 1998, 120, 2890–2899.
19 S. Itoh, T. Kondo, M. Komatsu, Y. Ohshiro, C. Li, N. Kanchisa,
Y. Kai and S. Fukuzumi, J. Am. Chem. Soc., 1995, 117, 4714–4715.
20 (a) J. G. Serafin, A. C. Liu and S. R. Seyedmonir, J. Mol. Catal. A:
Chem., 1998, 131, 157–161; (b) D. M. Minahan and G. B.
Hoflund, J. Mol. Catal. A: Chem., 1996, 158, 109–115; (c)
W. S. Epling, G. B. Hoflund and D. M. Minahan, J. Catal., 1997,
171, 490–497; (d ) M. A. Barteau and R. J. Madix, in the chemical
physics of solid surfaces and heterogenenous catalysis, ed. D. A. King
and P. Woodraff, Elsevier, Amsterdam, 1982, vol. IV, p. 995; (e)
A. J. Nagy, G. Mestl and R. Schlogl, J. Catal., 1999, 188, 58–68.
21 (a) H. M. J. Wang and I. J. B Lin, Organometallics, 1998, 17, 972–
975; (b) M. Ohkouchi, D. Masui, M. Yamaguchi and T. Yamagishi,
J. Mol. Catal. A: Chem., 2001, 170, 1–15.
Acknowledgements
This project is supported by the National Science Foundation
of China.
References
1 (a) B. Meunier, Metal-Oxo and Metal-Peroxo Species in Catalytic
Oxidation, Springer-Verage, New York, 2000; (b) M. Merkx,
D. A. Kopp, M. H. Sazinsky, J. L. Blazyk, J. Müller and
S. J. Lippard, Angew. Chem. Int. Ed., 2001, 40, 2782–2807; (c)
E. I. Solomon, P. Chen, M. Metz, S.-K. Lee and A. E. Palmer,
Angew. Chem. Int. Ed., 2001, 40, 4570–4702.
22 W. Geary, J. Coord. Chem. Rev., 1971, 7, 81–114.
23 (a) O. W. Howarth, G. G. Morgan, V. Mckee and J. Nelson, J. Chem.
Soc., Dalton Trans., 1999, 2097–2102; (b) M. G. B. Drew, D. Farrell,
G. G. Morgan, V. Mckee and J. Nelson, J. Chem. Soc., Dalton Trans.,
2000, 1513–1519.
2 (a) K. D. Karlin, S. Kaderli and A. D. Zuberbühler, Acc. Chem. Res.,
1997, 30, 139–147; (b) L. Dong, Acc. Chem. Res., 1996, 29, 190–196;
D a l t o n T r a n s . , 2 0 0 4 , 1 1 0 4 – 1 1 1 1
1110

View Article Online
24 (a) E. C. Constable, Metals and ligand reactivity, VCH, Weinheim,
1996, pp. 72–78; (b) A. B. Blake, E. Sinn, A. Yavari, B. Moubaraki
and K. S. Murray, Inorg. Chim. Acta, 1995, 229, 281–290.
Soc., 1989, 111, 186–187; (b) H. E. Baumgarten, Org. Synth., vol. V,
John Wiley & Sons, New York, 1973, pp. 668–669.
27 XSCANS (Version 2.1), Siemens, Analytical X-ray Instruments,
Inc., Madison, USA, 1994.
25 (a) R.-S. Luo, X.-A. Mao, Z.-Q. Pan and Q.-H. Luo, Spectrachim.
Acta, 2000, 56, 1675–1680; (b) S.-Y. Yu, Q.-H. Zuo, M.-C. Shen and
X.-Y. Huang, Inorg. Chem., 1994, 33, 1251–1252.
28 SHELXTL (Version 6.10), Siemens, Industrial Automation, Inc.,
Analytical Instruments, Madison, USA, 1995.
29 H. Simonis, Ber. Dtsch. Chem. Ges., 1912, 45, 1584–1585.
26 (a) J. Chin, M. Banaszczyk, V. Jubian and X. Zou, J. Am. Chem.
D a l t o n T r a n s . , 2 0 0 4 , 1 1 0 4 – 1 1 1 1
1111