Gold corroles

Article abstract of DOI:10.1142/S1088424611003045

Although a variety of conditions proved unsuccessful for gold insertion into β-unsubstituted meso-triarylcorroles, refluxing free-base β-octabromo-meso-triarylcorrole ligands, H₃[Br ₈T(p-X-P)C] (X = MeO, n-Bu, CH₃, H, and

Full text of DOI:10.1142/S1088424611003045

Journal of Porphyrins and Phthalocyanines
J. Porphyrins Phthalocyanines 2011; 15: 106–110
DOI: 10.1142/S1088424611003045
Published at http://www.worldscinet.com/jpp/
Gold corroles
Abraham B. Alemayehu and Abhik Ghosh*
Department of Chemistry and Center for Theoretical and Computational Chemistry, University of Tromsø,
N-9037 Tromsø, Norway
Received 24 November 2010
Accepted 16 February 2011
ABSTRACT:Althoughavarietyofconditionsprovedunsuccessfulforgoldinsertionintoβ-unsubstituted
meso-triarylcorroles, refluxing free-base β-octabromo-meso-triarylcorrole ligands, H3[Br8T(p-X-P)C]
(X = MeO, n-Bu, CH3, H, and CF3), with chloroauric acid in dichloromethane containing a small amount
of triethylamine for 30 min led to 54–65% yields of the corresponding gold complexes.
KEYWORDS: gold, corrole, octabromocorrole.
experiments afforded insight into the problem. With
INTRODUCTION
chloroauric acid (H3O+AuCl4-) as the gold(III) source
New ligands lead to new coordination chemistry. Por-
phyrins and related ligands amply illustrate this all too
familiar truism. Thus, syntheses of heteroporphyrins,
N-confused porphyrins and carbaporphyrins, subporphy-
and the β-unsubstituted meso-triarylcorrole ligands
H3[T(p-X-P)C] (X = MeO, CH3, H, CF3) and meso-tris
(pentafluorophenyl-corrole), mass spectrometric analy-
sis of the products indicated intractable mixtures of
rins, expanded porphyrins, porphyrazines, and corrola-
mono-, di-, tri- and tetra-chlorinated free-base corrole
zines have all led to much novel coordination chemistry
products, as well as small amounts of dimeric corroles,
[1]. The chemistry of corroles is among the most impres-
sive in this regard [2]; ever since the development of
simple one-pot syntheses by Gross [3], Paolesse [4] and
their coworkers, which were subsequently improved by
arising presumably from Au(III)-promoted oxidative
coupling. We surmised that free-base β-octabromo-
meso-triarylcorrole ligands, H3[Br8T(p-X-P)C], which
we have recently developed in our laboratory [10, 11],
Gryko [5], corrole chemistry has grown by leaps and
might lend themselves to effective gold insertion, an
bounds. Over the last few years, we have worked on a
insight that ultimately proved correct.
number of aspects of corroles, focusing particularly on
the effects of peripheral substituents on electronic [6,
7] and geometric [8, 9] structure. In the course of this
work, we recently found that free-base β-octabromo-
meso-triarylcorrole ligands, long inaccessible in their
RESULTS AND DISCUSSION
Finding an optimum solvent for gold insertion
involved much trial and error. Refluxing pyridine, which
free-base forms, could be synthesized via reductive
is commonly used for metalation of both porphyrins and
demetalation of the corresponding copper complexes
corroles, proved unsatisfactory; although gold inser-
[10, 11]. We hoped that the new ligands would lead to
tion with chloroauric acid was observed after prolonged
new coordination complexes that cannot be made from
refluxing (> 14 h), the yields were poor and isolation of
the Au corroles was correspondingly difficult. Refluxing
toluene, methanol, and DMF, with or without an added
β-unsubstituted corrole ligands. This hope has now been
realized with the synthesis of a family of gold(III) com-
plexes with systematically varying meso-aryl substitu-
base (triethylamine), all failed to result in gold insertion.
ents, as described below.
Refluxing dichloromethane, with a small amount of tri-
Corrole researchers have long sought to synthesize
ethylamine, however, proved both highly effective and
gold complexes, without success, and our preliminary
convenient, resulting in yields of 54–65%. Using these
conditions, we succeeded in inserting the metal into five
*Correspondence to: Abhik Ghosh, email: abhik.ghosh@uit.no
different H3[Br8T(p-X-P)C] ligands (X = MeO, n-Bu,
Copyright © 2011 World Scientific Publishing Company

GOlD COrrOleS
107
analogues. Even obtaining 1H NMR
spectra proved troublesome or impos-
sible with CDCl3, CD2Cl2, benzene-d6,
DMSO-d6, DMF-d7, methanol-d4, and
acetone-d6 as solvent. Finding a solvent
for electrochemical studies proved simi-
larly troublesome [13]. Fortunately, the
complexes are modestly soluble in 1,2-
dichlorobenzene, which finally allowed
1
the determination of H NMR spectra
and cyclic voltammograms. Because the
X = CH3 complex was marginally more
soluble than the other three, i.e. X =
MeO, H, and CF3, we decided to synthe-
size the X = n-Bu analogue in the hope
of obtaining a much more soluble gold
corrole. The n-Bu complex did prove
more soluble than the others (allowing
for example cyclic voltammetry experi-
ments in dichloromethane [14]) but only
modestly so. Unfortunately, the invola-
tile character of 1,2-dichlorobenzene, the
Fig. 1. UV-visible spectra of AuIII[Br8T(p-X-P)C] in dichloromethane. In each
case, the sample concentration was 0.013 0.001 mM
CH3, H, and CF3) [10]. Unambiguous proof of purity and
composition of the Au[Br8T(p-X-P)C] complexes was
obtained from thin-layer chromatography, MALDI-TOF
mass spectrometry, 1H NMR spectroscopy and elemental
analysis.
Although our primary objective here is to focus on
details of the syntheses, certain intriguing properties of
gold complexes, relative to the well-studied copper cor-
roles, were immediately obvious. Thus, unlike copper
triarylcorroles, whose Soret maxima redshift sensitively
in response to electron-donating para substituents on the
meso aryl groups [6, 7], the Soret bands of Au[Br8T(p-
X-P)C] (Fig. 1) were essentially invariant for X = MeO,
n-Bu, CH3, H, and CF3. The constancy of the Soret maxi-
mum strongly argues for a relative lack of charge trans-
fer character in the gold Soret bands, compared with the
analogous copper complexes.
Electrochemical measurements (Fig. 2) support the
above argument. As shown in Table 1, the reduction
potential of the gold octabromocorroles are lower than
those of the analogous copper complexes by over 1.0 V
[6, 12], indicating the difficulty of adding an electron to
the LUMO, which is substantially metal(dx2-y2)-based for
all coinage metal corroles. In the same vein, the electro-
chemical HOMO-LUMO gaps (the difference between
E1/2ox and E1/2red) of about 2.3 V of the gold complexes
are far higher than those of both copper and silver cor-
roles, which are typically around 1.2–1.5 V. The apparent
lack of charge transfer character in the gold Soret bands
is thus reasonably attributed to a very high-energy metal
dx2-y2 orbital as well as to a large HOMO-LUMO gap.
WhencharacterizingtheAu[Br8T(p-X-P)C]derivatives
(X=MeO, H, andCF3), anunexpectedcomplicationarose,
involving the solubility of the compounds; the gold com-
plexes proved considerably less soluble than their copper
Fig. 2. (a) Cyclic voltammograms of Au[Br8T(p-X-P)C] in
1,2-dichlorobenzene; (b) comparison of Cu[Br8T(p-Bu-P)C
and Au[Br8T(p-Bu-P)C]
Copyright © 2011 World Scientific Publishing Company
J. Porphyrins Phthalocyanines 2011; 15: 107–110

108
A.B. AlemAyehu AnD A. GhOSh
1H NMR spectra were recorded on a 400 MHz Mercury
Plus Varian spectrometer at room temperature in 1,2-
dichlorobenzene-d4. Proton chemical shifts (δ) were ref-
erenced to residual 1,2-dichlorobenzene-d4 (δ = 7.04 and
7.30 ppm). MALDI-TOF mass spectra were recorded on
a Waters Micromass MALDI micro MX Mass Spectrom-
eter with α-cyano-4-hydroxycinnamic acid (CHCA) as
the matrix.
Table 1. Half-wave potentials (E1/2, V vs. SCE) for Cu and Ag
corroles in CH2Cl2 and for Au corroles in 1,2-dichlorobenzene,
all in presence of 0.1 M TBAP
Corrole
X
E1/2ox E1/2red
Reference
6
Cu[T(p-X-P)C]
CF3
H
0.89 -0.08
0.76 -0.20
0.70 -0.23
CH3
Cyclic voltammetry experiments were carried out at
298 K using an EG & G Model 263A potentiostat with
a three-electrode system consisting of a glassy carbon
working electrode, a platinum wire counterelectrode,
and a saturated calomel reference electrode (SCE). Tetra-
n-butyl ammonium perchlorate (TBAP), recrystallized
from ethanol and dried in a desiccator for at least one
week, was used as the supporting electrolyte. Distilled
1,2-dichlorobenzene (and, in a couple of cases, dichlo-
romethane)wasusedasthesolventforcyclicvoltammetry
experiments. The reference electrode was separated from
the bulk solution by a fritted glass bridge filled with the
solvent/supporting electrolyte mixture. Pure argon was
bubbled through the solution containing the gold corrole
complexes for at least 2 min prior to the cyclic voltam-
metry experiments and the solutions were also protected
from air by an argon blanket during the experiments. All
half-wave potentials reported here are referenced to the
SCE.
OCH3 0.65 -0.24
Cu[Br8T(p-X-P)C]
Cu[F8T(p-X-P)C]
Ag[T(p-X-P)C]
CF3
H
1.24
1.14
1.12
0.25
0.12
0.07
0.04
0.35
0.22
0.18
0.17
6
CH3
OCH3 1.1
CF3
H
1.24
7
1.15
1.12
CH3
OCH3 1.06
CF3
H
0.91 -0.78
Manuscript
0.73 -0.86 in preparation
0.69 -0.88
CH3
OCH3 0.66 -0.91
Au[b-Br8-T(p-X-P)C] CF3
1.41 -0.90
1.29 -1.02
1.27 -1.00
This work
H
CH3
C4H9 1.26 -1.06
OCH3 1.25 -1.09
Chloroauric acid, known commercially as gold(III)
chloride hydrate (99.999%), and 1,2-dichlorobenzene-d4
(98 atom-% D) were obtained from Sigma-Aldrich and
used as received. Column chromatography was per-
formed on Matrex 35–70 silica from Millipore. The free-
base H3[Br8T(p-X-P)C] corrole ligands were synthesized
as described previously [10].
only effective solvent found so far, has frustrated our
efforts to obtain X-ray quality crystals.
In the absence of a crystal structure, relativistic spin-
orbit DFT calculations with ZORA basis sets (BP86-D/
ZORA-STO-TZP; ADF 2009 [15]) were used to optimize
the molecular structures of Au[TPC] and Au[Br8TPC].
In sharp contrast to distinctly saddled Cu[TPC] [8], the
calculations predicted a much flatter geometry for (the
experimentally unknown) Au[TPC]. As shown in Fig. 3,
the optimized geometry of Au[Br8TPC] was also found to
be much less saddled than that of Cu[Br8TPC] [9]. Using
a combination of X-ray crystallography and DFT calcu-
lations [8, 9], we have shown elsewhere that saddling in
copper corroles is driven by a Cu(dx2-y2)-corrole(p) inter-
action, which is symmetry-forbidden for a planar corrole
but becomes allowed under saddling. Copper corroles are
thus best described as CuII-corrole•2-. In contrast, the high
energy of theAu 5dx2-y2 orbital and the relative lack of sad-
dling in the gold case suggest a true Au(III) description.
Common procedure for gold insertion
A 100-mL round-bottomed flask was charged with
free-base H3[Br8T(p-X-P)C] (X = MeO, n-Bu, CH3, H,
and CF3; 20 mg) and triethylamine (60 mL) dissolved in
dichloromethane (15 mL). To the well-stirred solution
was added chloroauric acid (10 equiv.) and the reaction
mixture was refluxed for 0.5 h. The deep green solution of
the free-base corrole turned reddish-brown within 10 min
after the addition of the gold reagent. The reaction mix-
ture was cooled to room temperature and rotary evapo-
rated to dryness. The crude product was passed through
a short silica gel column with dichloromethane as elu-
ent (despite the limited solubility in this solvent). The
resulting solution was rotary evaporated to dryness and
the resulting solid was washed repeatedly with n-hexane,
affording 54 to 65% of the pure Au[Br8T(p-X-P)C] prod-
uct. Yields and analytical details for the different com-
plexes are as given below.
EXPERIMENTAL
Au[Br8T(p-CF3-P)C]. Yield 54%. UV-vis (CH2Cl2):
General
λ
max, nm (ε × 10-4, M-1.cm-1) 429 (15.88), 541 (2.31), 578
1
Ultraviolet-visible spectra were recorded on an HP
8453 spectrophotometer with dichloromethane as solvent.
(5.09). H NMR (C6Cl2D4): δ, ppm 8.26 (d, J = 7.6 Hz,
4H, 5,15-o or m-phenyl), 8.23 (d, J = 7.6 Hz, 2H,
Copyright © 2011 World Scientific Publishing Company
J. Porphyrins Phthalocyanines 2011; 15: 108–110

GOlD COrrOleS
109
Fig. 3. Highlights of BP86/STO-TZP optimized geometries (Å, deg) of Cu and Au corroles, with key saddling dihedrals indicated
in red. The Au geometries were obtained with relativistic spin-orbit calculations with ZORA basis sets, whereas nonrelativistic
calculations were used for the Cu complexes
10-o or m-phenyl), 8.16 (d, J = 7.2 Hz, 4H, 5,15-m or
o-phenyl), 8.12 (d, J = 7.6 Hz, 2H, 10-m or o-phenyl).
MS (MALDI-TOF, major isotopomer): m/z [M]+ 1555.83
(expt.), 1555.73 (calcd. for C40H12N4F9Br8Au). Elemen-
tal analysis: C 30.10, H 1.14, N 3.93 (calcd. 30.86, 0.78,
3.60, respectively).
Au[Br8T(p-CH3-P)C]. Yield 62%. UV-vis (CH2Cl2):
max, nm (ε × 10-4, M-1.cm-1) 430 (11.60), 541 (1.61), 578
λ
(3.78). 1H NMR (C6Cl2D4): δ, ppm 8.02 (d, J = 8 Hz, 4H,
5,15-o or m-phenyl), 7.99 (d, J = 7.6 Hz, 2H, 10-o or
m-phenyl), 7.70 (d, J = 7.2 Hz, 4H, 5,15-m or o-phenyl),
7.66 (d, J = 7.6 Hz, 2H, 10-m or o-phenyl), 2.77 (s, 9H,
5,10 and 15-p-CH3). MS (MALDI-TOF, major isoto-
pomer): m/z [M]+ 1394.15 (expt.), 1393.82 (calcd. for
C40H21N4Br8Au). Elemental analysis: C 34.81, H 1.29, N
3.53 (calcd. 34.45, 1.52, 4.02, respectively).
Au[Br8TPC]. Yield 56%. UV-vis (CH2Cl2): λmax, nm
(ε × 10-4, M-1.cm-1) 429 (12.23), 540 (1.41), 464 (9.41),
1
578 (3.39). H NMR (C6Cl2D4): δ, ppm 8.04–8.13 (bm,
7H, 5,15-o or m-phenyl, 10-m or o-phenyl and 10-p-
phenyl overlapping), 7.75–7.90 (bm, 8H, 5,15-m or
o-phenyl, 10-o or m-phenyl and 10-p-phenyl overlap-
ping). MS (MALDI-TOF, major isotopomer): m/z [M]+
1351.59 (expt.), 1351.74 (calcd. for C37H15N4Br8Au).
Elemental analysis: C 33.24, H 1.25, N 3.94 (calcd.
32.86, 1.12, 4.14, respectively).
Au[Br8T(p-Bu-P)C]. Yield 54%. UV-vis (CH2Cl2):
λ
max, nm (ε × 10-4, M-1.cm-1) 430 (16.58), 542 (2.23), 578
1
(5.34). H NMR (C6Cl2D4): δ, ppm 8.03 (d, J = 7.6 Hz,
4H, 5,15-o or m-phenyl), 8.01 (d, J = 7.6 Hz, 2H, 10-o or
m-phenyl), 7.70 (d, J = 8.4 Hz, 4H, 5,15-m or o-phenyl),
7.66 (d, J = 8.4 Hz, 2H, 5,15-m or o-phenyl), 3.03 (m,
Copyright © 2011 World Scientific Publishing Company
J. Porphyrins Phthalocyanines 2011; 15: 109–110

110
A.B. AlemAyehu AnD A. GhOSh
6H, 5,10 and 15-p-butyl), 1.95 (m, 6H, 5,10 and 15-p-
butyl), 1.63 (m, 6H, 5,10,15-p-butyl), 1.14 (m, 9H, 5,10
and 15-p-butyl). MS (MALDI-TOF, major isotopomer):
m/z [M]+ 1520.05 (expt.), 1520.06 (calcd. for C49H39N4-
Br8Au). Elemental analysis: C 38.53, H 2.37, N 3.76
(calcd. 38.70, 2.56, 3.69, respectively).
4. Paolesse R, Mini S, Sagone F, Boschi T, Jaquinod
L, Nurco DJ and Smith KM. Chem. Comm. 1999;
1307–1308.
5. Koszarna B and Gryko DT. J. Org. Chem. 2006; 71:
3707–3717.
6. Wasbotten IH, Wondimagegn T and Ghosh A.
J. Am. Chem. Soc. 2002; 124: 8104–8116.
7. Steene E, Wondimagegn T and Ghosh A. J. Am.
Chem. Soc. 2003; 125: 16300–16309.
8. AlemayehuAB, Gonzalez E, Hansen LK and Ghosh
A. Inorg. Chem. 2009; 48: 7794–7799.
9. Alemayehu AB, Conradie J, Hansen LK and Ghosh
A. Inorg. Chem. 2010; 49: 7608–7610.
10. Capar C, Thomas KE and Ghosh A. J. Porphyrins
Phthalocyanines 2008; 12: 964–967.
11. Capar C, Hansen L.-K, Conradie J and Ghosh A.
J. Porphyrins Phthalocyanines 2010; 14: 509–512.
12. Ou Z, Shao J, Zhao H, Ohkubo K, Wasbotten IH,
Fukuzumi S, Ghosh A and Kadish KM. J. Porphy-
rins Phthalocyanines 2004; 8: 1236–1247.
13. The poor solubility of the Au[Br8T(p-X-P)C] deriv-
atives in CH2Cl2 may be viewed as somewhat para-
doxical, given that the complexes were synthesized
in that solvent.
14. Gold corroles appear to be rather stable, consider-
ably more so than silver corroles, which demeta-
late easily, even by traces of acid typically present
in dichloromethane: Stefanelli M, Shen J, Zhu W,
Mastroianni M, Mandoj F, Nardis S, Ou Z, Kadish
KM, Fronczek FR, Smith KM and Paolesse R.
Inorg. Chem. 2009; 48: 6879–6887.
Au[Br8T(p-MeO-P)C]. Yield 65%. UV-vis (CH2Cl2):
λ
max, nm (ε × 10-4, M-1.cm-1) 433 (14.70), 542 (1.06), 580
(3.56). 1H NMR (C6Cl2D4): δ, ppm 8.02 (d, J = 8 Hz, 4H,
5,15-o or m-phenyl), 7.99 (d, J = 8.4 Hz, 2H, 10-o or
m-phenyl), 7.46 (d, J = 8 Hz, 4H, 5,15-m or o-phenyl), 7.42
(d, J = 8 Hz, 2H, 10-m or o-phenyl), 4.08 (s, 6H, 5,15-p-
OCH3), 4.07 (s, 3H, 10-p-CH3). MS (MALDI-TOF, major
isotopomer): m/z [M]+ 1441.96 (expt.), 1441.82 (calcd.
for C40H21N4O3Br8Au). Elemental analysis: C 33.71, H
1.58, N 3.88 (calcd. 33.30, 1.47, 3.89, respectively).
CONCLUSION
In concluding, we would like to emphasize that the
successful synthesis of gold corroles owes critically to the
ease and generality of our reductive demetalation-based
synthesis of free-base β-octabromo-meso-triarylcorrole
ligands [16]. Electronic absorption, electrochemical and
DFT studies already indicate fascinating variations in
electronic structure among the coinage metal corroles.
Acknowledgements
This work was supported by the Research Council of
Norway.
15. a) Becke AD. Phys. Rev. 1988; A38: 3098–3100.
b) Perdew JP. Phys. Rev. 1986; B33: 8822. Erratum:
Perdew JP. Phys. Rev. 1986; B34: 7406. c) Dis-
persion corrections: Grimme S. J. Comput. Chem.
2006; 27: 1787.
16. Aside from gold corroles, iridium corroles are the
only other examples of third-row transition metal
corroles synthesized to date: Palmer JH, Day MW,
Wilson AD, Henling LM, Gross Z and Gray HB.
J. Am. Chem. Soc. 2008; 130: 7786–7787.
REFERENCES
1. For reviews on the coordination chemistry of
porphyrins and related ligands, see: Handbook of Por-
phyrin Science, Vol. 2, Kadish KM, Smith KM and
Guilard R. (Eds.) World Scientific: Singapore, 2010.
2. Aviv-Harel I and Gross Z. Chem. Eur. J. 2009; 15:
8382–8394.
3. Gross Z, Galili N and Saltsman I. Angew. Chem. Int.
Ed. 1999; 38: 1427–1429.
Copyright © 2011 World Scientific Publishing Company
J. Porphyrins Phthalocyanines 2011; 15: 110–110