A push-pull azobenzene is mercurated twice at the ring with less electron density

Article abstract of DOI:10.1016/j.jorganchem.2012.05.040

Treatment of N,N-dialkyl-4-[4-nitrophenyl)diazenyl] anilines with a mercury(II) salt in anhydrous trifluoroacetic acid results in single and double metallation of the ring with less electron density. The seemingly counterintuitive outcome of the reaction was rationalized through experimental and computational investigations of the reaction mechanism.

Full text of DOI:10.1016/j.jorganchem.2012.05.040

Journal of Organometallic Chemistry 716 (2012) 11e18
Contents lists available at SciVerse ScienceDirect
Journal of Organometallic Chemistry
journal homepage: www.elsevier.com/locate/jorganchem
A pushepull azobenzene is mercurated twice at the ring with less electron
density
*
Philip J.W. Elder, Jeff C. Landry, Anthony F. Cozzolino, Allison E.A. Chapman, Ignacio Vargas-Baca
Department of Chemistry and Chemical Biology, McMaster University, 1280 Main Street West, Hamilton, Ontario, Canada L8S 4M1
a r t i c l e i n f o
a b s t r a c t
Article history:
Treatment of N,N-dialkyl-4-[4-nitrophenyl)diazenyl] anilines with a mercury(II) salt in anhydrous tri-
fluoroacetic acid results in single and double metallation of the ring with less electron density. The
seemingly counterintuitive outcome of the reaction was rationalized through experimental and
computational investigations of the reaction mechanism.
Received 10 September 2011
Received in revised form
15 May 2012
Accepted 18 May 2012
Ó 2012 Elsevier B.V. All rights reserved.
Keywords:
Azodyes
Functionalization of azobenzenes
Mercuration
DFT calculations
Metallation reactions
Electrophilic substitution
1. Introduction
Driven to great extent by the search for pharmaceuticals and
intermediates for organic synthesis, vast variety of aryl-
a
mercurials have been obtained in over 100 years of research
[1e4]. The field has been recently reinvigorated by studies of the
unique Lewis-acid properties of perfluorinated phenylmercurials.
Their ability to form complexes with a variety of molecules [5,6],
including conjugated alkynes [7], metallocenes [8], organic esters
[9], and arenes [10] can be applied to the assembly of remarkable
supramolecular structures [11e13]. Recent work has shown that
mono- and di-functional Lewis acids containing bridging mercury
atoms exhibit affinities for fluoride ions that rival those of their
diboron counterparts, but possess greater stability in the presence
of water [14]. Many of these species have been synthesized
through transmetallation of the corresponding lithiated interme-
diates or Grignard reagents with mercuric salts (Eq. (1)); organo-
mercury compounds can also be prepared by direct metallation
(Eq. (2)). Activated aromatic molecules easily react with mercu-
ry(II) oxide, chloride, acetate or trifluoroacetate, in increasing
order of reactivity [15].
Direct mercuration of aromatic rings is strongly influenced by
the nature of the substituent groups. Such effects are usually
explicable by considering a classic electrophilic aromatic substitu-
tion (EAS) mechanism and the resonance of multiple canonical
structures [16]; in the conventional formalism, electron-donating
groups are ortho/para directors while electron-withdrawing
groups are meta-directors. However, the reactions of mercuric
salts with aromatic species occasionally yield results that e at first
glance e are unexpected. For example, the mercuration of toluene
exhibits a strong preference for para-substitution but equilibrium
between all three possible isomers will ensue over long reaction
times, especially in acidic medium [17]. This has been attributed to
the ability of mercury in the arenemercurenium ions to shift when
there is no kinetic control over the site of substitution [18]. The
* Corresponding author. Tel.: þ1 905 525 9140; fax: þ1 905 522 2509.
E-mail address: vargas@chemistry.mcmaster.ca (I. Vargas-Baca).
0022-328X/$ e see front matter Ó 2012 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.jorganchem.2012.05.040

12
P.J.W. Elder et al. / Journal of Organometallic Chemistry 716 (2012) 11e18
low yield of ortho-mercurated products obtained from some
monoalkyl-benzenes does suggest steric effects also impact these
reactions [19]. Interestingly, while the mercuration of nitrobenzene
yields predominantly meta-substituted species there is also
a significant proportion of ortho-mercuration [20]. In general,
substituents bearing electron lone pairs promote mercuration at
the ortho position, which is usually assumed to be the consequence
of an attractive interaction with the mercury ion [16].
Parallel syntheses were performed in a First-Mate Benchtop
Synthesizer (Argonaut Technologies). HPLC analytical and semi-
preparative runs were performed at room temperature on a Waters
Spherisorb 5
m
m ODS2 analytical column (4.6 ꢀ 150 mm, flow rate
2 mL minꢁ1) or a Waters Spherisorb S5 ODS2 semi-preparative
column (10 ꢀ 250 mm, flow rate 8 mL minꢁ1) using the Waters
600E Multisolvent Delivery System (Waters 600 Controller, Waters
600E Pump). The separation was monitored with a Waters 2996
Photodiode Array Detector in conjunction with the Empower
control software. Eluted fractions were recovered with a Waters
Fraction Collector II and isolated with an Eppendorf 5301 Centrif-
ugal Concentrator.
The mercuration of larger molecules with extensively conju-
gated
p-systems and a combination of donor and acceptor groups
would be a convenient tool in the preparation of precursors for
materials with a number of useful electronic and optical properties.
However, the regiochemistry of such a reaction is difficult to predict
because of the competition of different directing effects of several
substituents, as is the case with the direct mercuration of diphenyl
Schiff bases [21]. This issue greatly complicates the rational design
of molecules of practical interest and efficient synthetic routes for
their synthesis. In this context we examined the mercuration of
pushepull azodyes, the N,N-dialkylamino, nitroazobenzenes 1 and
6. The properties of such pseudostilbene chromophores have been
studied in great detail due to their applicability in liquid crystals,
electro-optical devices and nonlinear optical materials [22e27].
Preliminary results showed that the mercuration reaction occurs
predominantly on the ring that e in principle e bears less electron
density. Here we report the results of a combination of experi-
mental and computational investigations that provide a rationale
for our observations (Chart 1).
2.2. Spectroscopic instrumentation
1H and 13C{1H}
d NMR spectra were acquired in solution on
a Bruker DRX 500 (500.13 MHz) spectrometer; chemical shifts are
reported in parts per million relative to tetramethylsilane (TMS)
and were measured using the solvent as a secondary reference. 13C
{1H}
d chemical shifts assignments were confirmed using 2-D
techniques (HSQC and HMBC). Direct electron impact (DEI) mass
spectra were obtained with a Micromass GCT (GC-EI/CI TOF) Mass
Spectrometer. Infrared spectra were recorded on Bio-Rad FTS-40
spectrometer as KBr pellets. UVevisible spectra were obtained on
a Cary 50 spectrophotometer. Melting points were recorded on
a Thomas-Hoover capillary melting point apparatus and are
reported uncorrected.
2.3. Syntheses
2. Experimental
2.3.1. N-ethyl-N-pentyl aniline (C13H21N)
The aniline was prepared by a modification of a literature
procedure [28]; using the alkyl iodide [29,30] in place of the
bromide. A mixture of 1-iodopentane (20 mL, 0.153 mol) and
N-ethyl-aminobenzene (13 mL, 0.103 mol) was added to potassium
hydroxide (6.192 g, 0.110 mol) under nitrogen. The mixture was
refluxed with stirring for 6 h, and the product was isolated as
a clear, colourless liquid after distillation (68 ꢂC, 0.4 torr). Yield:
15.03 g, 78.6 mmol, 76%. 1H NMR (500 MHz, CDCl3, 7.26 ppm):
2.1. Materials and methods
The manipulation of hygroscopic materials was performed
under an atmosphere of anhydrous nitrogen with standard Schlenk
and glovebox techniques. Trifluoroacetic acid was dehydrated by
distillation over phosphorus pentoxide. All other reagents were
used as received from SigmaeAldrich (p-nitroaniline, ethyl aniline,
n-pentanol), Caledon (potassium hydroxide), Fisher (lithium chlo-
ride), Baker (iodine), and Shawinigan (sodium nitrite). The silica gel
used for column chromatography (EM Science) had a particle size of
d
¼ 0.93 (t, 3H, HC19), 1.16 (t, 3H, HC14), 1.34 (m, 4H, HC17 HC18), 1.60
(m, 2H, HC16), 3.25 (t, 2H, HC15), 3.38 (q, 2H, HC13), 6.64 (m, 2H, HC3
HC8), 6.68 (d, 2H, HC2 HC6), 7.22 (t, 1H, HC7). 13C-DEPTq NMR
(500 MHz, CDCl3, 77.1 ppm): 12.4 (s, 1C, C14), 14.1 (s, 1C, C19), 22.7
(s, 1C, C17), 27.1 (s, 1C, C16), 29.5 (s, 1C, C18), 44.8 (s, 1C, C13), 50.4
(s, 1C, C15), 111.6 (s, 2C, C2 C6), 115.2 (s, 2C, C3 C5), 129.2 (s, 1C, C4),
147.9 (s, 1C, C1). HRMS (EI, %): m/z Found: 191.1679 (Mþ, 15)
Calculated 191.1674.
40e63
mm.
Caution. Mercury and its compounds are toxic. They should be
stored, handled and disposed of in a well-ventilated facility,
wearing appropriate personal protective equipment and adhering
to local safety, labour and environmental regulations.
2.3.2. N-ethyl-4-[(4-nitrophenyl)diazenyl] N-pentylaniline
(C19H23N4O2) (1)
The compound was prepared by azo coupling from p-nitroani-
line (1.794 g, 13.0 mmol) and N-ethyl-N-pentyl aniline (2.502 g,
13.1 mmol) in 40 mL of 5 M HCl at 0 ꢂC [31]. Sodium nitrite and
acetate were used to remove excess NO2 and neutralize the acid,
respectively. A dark red solid was isolated by filtration and purified
by column chromatography (100% CHCl3). Yield: 3.888 g,
11.42 mmol, 88%. 1H NMR (CD2Cl2):
HC14), 1.39 (m, 4H, HC17 HC18), 1.67 (q, 2H, HC16), 3.38 (t, 2H, HC15),
3.51 (q, 2H, HC13), 6.68 (d, 2H, HC2 HC6), 7.93 (d, 2H, HC3 HC5), 7.95
(d, 2H, HC8 HC12), 8.31 (d, 2H, HC9 HC11). 13C-DEPTq NMR (500 MHz,
d
¼ 0.94 (t, 3H, HC19), 1.25 (t, 3H,
CD2Cl2, 54.0 ppm):
d
¼ 12.4 (s, 1C, C14), 14.1 (s, 1C, C19), 22.8 (s, 1C,
C18), 27.4 (s, 1C, C16), 29.6 (s, 1C, C17), 45.6 (s, 1C, C13), 50.9 (s, 1C,
C15), 111.6 (s, 2C, C2 C6), 122.8 (s, 2C, C8 C12), 125.1 (s, 2C, C9 C11),
126.7 (s, 2C, C3 C5),143.6 (s,1C, C4),147.6 (s,1C, C7),152.3 (s,1C, C1),
157.8 (s, 1C, C10). MP ¼ 110e112 ꢂC. UV (CH2Cl2): lmax ¼ 498 nm,
Chart 1.

P.J.W. Elder et al. / Journal of Organometallic Chemistry 716 (2012) 11e18
13
ε ¼ 19,326 cm2/mol. IR (cmꢁ1): 3732w, 3098w, 2951m, 2926m,
2869m, 1598s, 1584s, 1557m, 1515s, 1465w, 1455w, 1420w, 1404w,
1376s, 1360s, 1337s, 1307s, 1270s, 1254s, 1216m, 1189m, 1151m,
1139s, 1126s, 1102s, 1072m, 984w, 858m, 830m, 795w, 754w, 735w,
725w, 687w, 666w, 637w, 614w, 585w, 563w, 537w, 510w. HRMS
(EI, %): m/z Found: 340.1906 (Mþ, 13) Calculated 340.1899.
1307s,1262s,1233s,1196w,1139s, 1104s,1073m,1024m, 992w,
892w, 869m, 834w, 818m, 798w, 747w, 727w, 691w, 634w,
566w, 539w, 518w. HRMS (EI, %): m/z Found: 466.0866
(Mþ, 65) Calculated: 466.0866.
2.3.5. [(2-Iodo-4-nitrophenyl)diazenyl]N,N-dimethylaniline
(C14H13N4O2I) (6)
2.3.3. Mercuration of N-ethyl-4-[(4-nitrophenyl)diazenyl] N-
pentylaniline
The compound was prepared using the above mercuration method
from 4-nitroaniline (1.02 g, 7.39 mmol) and N,N-dimethylaniline
(0.887 g, 7.33 mmol), followed by mercuration of a small amount
of material (0.0317 g, 0.117 mmol) with Hg(CF3CO2)2 (0.120 g,
0.281 mmol) and treatment with iodine (0.203 g, 0.798 mmol),
yielding a dark red powder (0.245 g, 0.906 mmol, 48%). Crystals
suitable for X-ray diffraction were grown by slow evaporation of
a chloroform/toluene solution. Yield ¼ 1.616 g, 4.082 mmol, 56%. 1H
In a typical experiment, 1 (0.064 g, 0.19 mmol) and mercury
trifluoroacetate (0.159 g, 3.72 mmol) were combined with anhy-
drous trifluoroacetic acid (0.13 mL) under nitrogen. The mixture
was heated with stirring for 48 h in an oil bath at 68 ꢂC. A
concentrated solution of sodium chloride (0.055 g, 0.942 mmol)
and sodium acetate (0.190 g, 2.31 mmol) was added to the reaction
flask and the entire sample was treated with ultrasound for 10 min.
The crude material was extracted with dichloromethane, dehy-
drated and used immediately for the iodination step.
NMR (500 MHz, CD2Cl2, 5.32 ppm):
d
¼ 3.13 (s, 6H, N(CH3)2), 6.78
(d, 2H, H2 H6), 7.66 (d, 2H, H9), 7.94 (d, 2H, H5 H3), 8.22 (dd, 2H, H8),
8.79 (d,1H, H11). 13C-DEPTq NMR (500 MHz, CD2Cl2, 54.0 ppm): 40.6
(s, 2C, N(CH3)2), 99.8 (s, 1C, C7), 112.2 (s, 2C, C2 C6), 117.5 (s, 1C, C8),
124.8 (s,1C, C9),127.4 (s, 2C, C3 C5),135.4 (s,1C, C11),145.2 (s,1C, C4),
148.4 (s, 1C, C12), 154.5 (s, 1C, C1), 156.0 (s, 1C, C10). UV (CH2Cl2):
lmax ¼ 502 nm, ε ¼ 25,294cm2/mol. IR (cmꢁ1): 3732w, 3087w, 2901w,
2854w, 2816w, 1610s, 1571m, 1553m, 1518s, 1506s, 1441w, 1416m,
1409m, 1357s, 1327s, 1306s, 1258m, 1238m, 1196w, 1139s, 1105s,
1064m, 1028w, 995w, 940w, 899w, 886w, 832w, 820m, 747w, 725w,
699w, 690w, 634w, 549w, 537w, 520w, 511w. HRMS (EI, %):
m/z Found: 396.0094 (Mþ, 100) Calculated: 396.0083.
2.3.4. Iodination of the mercuriated azodye
Iodine (0.202 g, 0.80 mmol) was added to the crude product of
mercuration in chloroform. This solution was then stirred for 48 h
and washed with concentrated sodium bicarbonate. Aqueous
sodium thiosulfate was added to the mixture with stirring. After
5 min, the organic layer was separated and dehydrated with
sodium sulfate then evaporated to dryness. The dark residue was
treated with a mixture of acetonitrileewater (86% v/v) and sepa-
rated by semi-preparative HPLC in 100 mL portions. The method
employed an 8 mL/min flow rate with a stepwise elution profile
that began isocratic 86% v/v for 14 min, and was followed by a linear
gradient to 100% acetonitrile over 1 min. The two major fractions
were collected, which were in order of elution:
2.4. Timed experiments
In a typical experiment, compound 1 (0.064 g, 0.19 mmol) and
mercury trifluoroacetate (0.159 g, 3.72 mmol) were combined with
anhydrous trifluoroacetic acid (0.13 mL) in each of six reaction
tubes of a parallel synthesis reactor (Argonaut FirstMateÔ). The
samples were maintained at 68 ꢂC with vigorous stirring under
nitrogen. Each mixture was quenched at a prescribed time by
addition of 2 mL of a solution of sodium chloride (0.47 M) and
sodium acetate (1.16 M). Each sample was treated with ultrasound
for 10 min, washed repeatedly with aqueous sodium bicarbonate,
centrifuged, and the supernatant was removed by pipette. The
solid residue was treated with a solution of iodine in chloroform
for 48 h. After removing the solvent, the solid samples were
washed with aqueous sodium thiosulfate and water to remove
residual iodine.
(i) N-ethyl-4-[(2,6-diiodo-4-nitrophenyl)diazenyl] N-pentyla-
niline (C19H22N4O2I2) (5). tr ¼ 12.4 min 1H NMR (CD2Cl2):
d
¼ 0.94 (t, 3H, HC19), 1.25 (t, 3H, HC14), 1.39 (m, 4H, HC17 HC18),
1.69 (q, 2H, HC16), 3.41 (t, 2H, HC15), 3.51 (q, 2H, HC13), 6.68 (d,
2H, HC2 HC6), 7.91 (d, 2H, HC3 HC5), 8.76 (s, 2H, HC9 HC11).
13C-DEPTq NMR (500 MHz, CD2Cl2, 54.0 ppm):
d
¼ 12.6 (s, 1C,
C14), 14.2 (s, 1C, C19), 23.0 (s, 1C, C18), 27.7 (s, 1C, C16), 29.9 (s,
1C, C17), 46.1 (s, 1C, C13), 51.4 (s, 1C, C15), 87.8 (s, 2C, C8 C12),
111.8 (s, 2C, C2 C6), 127.2 (s, 2C, C3 C5), 135.9 (s, 2C, C9 C11),
145.7 (s, 1C, C4), 147.0 (s, 1C, C7), 152.9 (s, 1C, C1), 159.4 (s, 1C,
C10). UV (CH2Cl2): lmax ¼ 460 nm, ε ¼ 13,080 cm2/mol. IR
(cmꢁ1): 3732w, 3363w, 3186w, 3088w, 3071w, 2957m, 2923s,
2852m, 1733w, 1646w, 1632w, 1605m, 1570w, 1556w, 1524w,
1508w, 1462w, 1410w, 1371w, 1333m, 1311w, 1275w, 1260w,
1216w, 1196w, 1184w, 1137m, 1114w, 1073w, 1043w, 995w,
946w, 914w, 893w, 879w, 822w, 795w, 749w, 739w, 720w,
703w, 525w, 503w. HRMS (EI, %): m/z Found: 591.9828
(Mþ, 100) Calculated: 591.9832.
2.5. X-ray crystallography
A single crystal of 7 was mounted on a Siemens P4 four-cycle
diffractometer with a Bruker 1000 CCD detector and a rotating
ꢀ
(ii) N-ethyl-4-[(2-iodo-4-nitrophenyl)diazenyl] N-pentylani-
anode utilizing Mo-K
a
radiation (
l
¼ 0.71073 A, graphite mono-
line (C19H23N4O2I) (4). tr ¼ 14.2 min 1H NMR (CD2Cl2):
d
¼ 0.94
chromator) equipped with an OXFORD cryosystem. A hemisphere
(t, 3H, HC19), 1.25 (t, 3H, HC14), 1.39 (m, 4H, HC17 HC18), 1.69
(q, 2H, HC16), 3.41 (t, 2H, HC15), 3.51 (q, 2H, HC13), 6.82 (d, 2H,
HC2 HC6), 7.68 (d, 1H, HC12), 7.94 (d, 2H, HC3 HC5), 8.24 (dd, 1H,
HC11), 8.81 (d, 1H, HC9). 13C-DEPTq NMR (500 MHz, CD2Cl2,
of reciprocal lattice was scanned in 0.36ꢂ steps in
u
with a crystal-
to-detector distance of 4.97 cm. Preliminary orientation matrices
were obtained from the first frames using SMART [32]. The
collected frames were integrated using preliminary orientation
matrices which were updated every 100 frames. Final unit cell
parameters were obtained by the refinement of the position of
54.0 ppm):
d
¼ 12.7 (s, 1C, C14), 14.4 (s, 1C, C19), 23.1 (s, 1C,
C18), 27.8 (s, 1C, C16), 29.8 (s, 1C, C17), 46.1 (s, 1C, C13), 51.4
(s, 1C, C15), 100.0 (s, 1C, C12), 112.0 (s, 2C, C2 C6), 117.5 (s, 1C,
C8),124.8 (s,1C, C9),127.7 (s, 2C, C5 C3),135.4 (s,1C, C11),144.0
(s, 1C, C4), 147.7 (s, 1C, C7), 153.7 (s, 1C, C1), 156.3 (s, 1C, C10).
UV (CH2Cl2): lmax ¼ 520 nm, ε ¼ 27,310 cm2/mol. IR (cmꢁ1):
3186w, 3086w, 2953m, 2921s, 2851m, 1734w, 1645w, 1600s,
1570m, 1555m, 1515s, 1436w, 1400m, 1357m, 1325s, 1314s,
reflections with I > 10s(I) after integration of all data using SAINT
[32]. The data sets were empirically corrected for absorption and
other effects using SADABS [33]. The structure was solved by direct
methods and refined by the full-matrix least squares method on all
F2 data using SHELXTL [34]. All non-H atoms were refined aniso-
tropically; H atoms were constrained to idealized positions using

14
P.J.W. Elder et al. / Journal of Organometallic Chemistry 716 (2012) 11e18
appropriate riding models. Molecular graphics were produced
using Mercury (Version 2.3) [35].
solubility of the product hampered optimization of the method.
Longer alkyl groups were placed on the amino group in order to
increase the solubility of the azodye. Satisfactory results were
obtained for the N-ethyl-[(4-nitrophenyl)diazenyl] N-pentylaniline
(1), therefore this compound was used for a full study of the
reaction. After mercuration, the trifluoroacetate anion was
exchanged for chloride in order to facilitate isolation of the prod-
ucts. The crude mercurated materials were usually obtained as
dark-maroon solids which could not be thoroughly purified on
their own because of their low solubility in common solvents.
Instead, reaction with iodine in chloroform was used to produce the
more soluble halogenated species; such a treatment indeed affor-
ded a soluble material that was shown by NMR and HPLC to be
a mixture. This analysis was accomplished using a custom HPLC
separation method developed that combined gradient and isocratic
water/acetonitrile regimes in analytical scale. Three major bands
were observed, and the separation method was scaled up in order
to collect samples large enough for spectroscopic characterization.
Given the scale in which the purification could be efficiently carried
out, identification of the components relied on NMR and mass
spectrometry; combustion elemental analyses would have required
large amounts of solvent, a bigger column and a higher throughput
pump. Up to 7% of the material obtained from the iodination
reaction eluted in 12 small fractions which could not be conclu-
sively identified. Also from the chromatographic separation, 10% of
starting material was recovered.
2.6. Computational details
All structures were fully optimized using the ADF DFT package
(versions 2007.01e2010.2) [36e38]. The adiabatic local density
approximation (ALDA) was used for the exchange-correlation
kernel [39,40], and the differentiated static LDA expression was
used with the VoskoeWilkeNusair parameterizations [41]. The
calculation of models of ground and transition state geometries
was gradient-corrected with the exchange and correlation func-
tionals of the gradient correction proposed in 1991 by Wang and
Perdew [42,43]. Geometry optimizations were conducted using
triple-z all-electron basis sets with one polarization function each
and applying the zero order relativistic approximation (ZORA)
[44] formalism with the specially adapted basis sets.
3. Results and discussion
Most of what is already known about the mercuration of azo-
benzenes derives from early studies which meticulously examined
the reaction of the parent compound and several ortho-substituted
derivatives [45]. In general, azobenzenes are not very reactive and
do require stringent mercuration conditions. For example, 22 h of
reflux in methanol with one equivalent of mercury acetate yields
just 40% of ortho-monomercurated and 3% of each o,o- and o,o0
dimercurated products. Addition of a small amount of perchloric
acid increases the corresponding yields to 55%, 8% and 9%. Similarly,
reflux with mercury trifluoroacetate salt in neat trifluoroacetic acid
yields 57% of ortho-monometallated product. The effect of the
substituent groups is complicated. Ortho-methylazobenzene
experiences 71% monomercuration on the substituted ring after
22 h of reflux in methanol with one equivalent of mercury acetate.
o,o0-Dimethylazobenzene gives 51% monomercuration, and 4%
o,o0-disubstitution under the same conditions, while o-methox-
yazobenzene yields 63% monomercuration and 1% and 4% of o,o-
and o,o0-dimercurials respectively [45]. Interestingly, electron-
withdrawing groups in the ortho position appear to promote
substitution of their own ring and noticeably decrease the yields
(eI: 22%, eNO2: 6%, eCN: 1%) of monomercurials. In most cases, the
characterization/identification of the mercurated products is
complicated by their limited solubility in common solvents; such
a problem can be circumvented by working with the more soluble
iodinated analogues. These are conveniently prepared by treatment
of the organomercurials with elemental iodine; the reaction
replaces the HgR group for an iodine atom without isomerization
(Eq (3)) [45].
3.1. Structures of the isolated compounds
Each species was initially identified using two-dimensional 1H
NMR spectroscopy, taking advantage of the changes in the patterns
of the aromatic region. The first band to elute (t ¼ 8.78 min) cor-
responded to the unreacted azodye (1). Its NMR spectrum displays
characteristic pairs of doublets of the amino [
d
¼ 6.82, 7.93 ppm
(3JHeH
¼
9.02 Hz)] and nitro rings
[
d
¼
7.95, 8.31 ppm
(3JHeH ¼ 8.85 Hz)] with equal integrations. The second band
(t ¼ 12.37 min) belongs a di-substituted species giving three
equally intense resonances in the aromatic region: two doublets at
6.68 ppm (3JHeH ¼ 9.11 Hz) and 7.90 ppm (3JHeH ¼ 9.11 Hz) which
would belong to the amino ring and one singlet at 8.76 ppm which
indicated that two iodine atoms are in equivalent positions, likely
ortho to the azo bridge. The third band (t ¼ 14.21 min) corresponds
to a product of monosubstitution. While the chemical shifts of the
protons in the amino ring are essentially unchanged [
d
¼ 6.82,
7.94 ppm (3JHeH ¼ 9.00 Hz)] the nitro ring protons produce two
doublets at
d
7.68 ppm (3JHeH
¼
9.00 Hz) and 8.81 ppm
(3JHeH ¼ 2.52 Hz) which couple to the resonance centred at
8.24 ppm. On the basis of their 1H NMR spectra the second and
third major bands in the chromatogram were assigned to
compounds 5 and 4 respectively (see Scheme 1 for 1H NMR
assignments).
3.2. X-ray crystallography
The observations summarized above do provide an excellent
overview of the influence of individual substituents on the
mercuration reaction of azobenzene but molecules that contain
two substituents exerting competing effects, as would be the case
of pushepull azobenzenes, are notably absent from that list. To fill
in this gap, we conducted preliminary studies of the mercuration of
the azodyes on [(4-nitrophenyl)diazenyl]N,N-dimethylaniline (6).
Derivatization attempts with mercury (II) chloride and acetate
were unsuccessful. Chromatographic evidence of partial mercura-
tion of 6 was obtained with the trifluoroacetate in anhydrous tri-
fluoroacetic acid [45]. However, the yield was low and the limited
Although the NMR characterization of the species in the second
band does show functionalization of the nitro ring, there remains
Scheme 1.

P.J.W. Elder et al. / Journal of Organometallic Chemistry 716 (2012) 11e18
15
Table 1
found. Despite these limitations, the structural determination did
establish the position at which 6 underwent metallation. Given the
similarities of their NMR spectra, 7 and the mono iodo derivative of
1 must have analogous connectivities, which confirms the identi-
fication of 4 as the species eluting at 14.2 min. This observation
confirms by extension the identification of 5 and implies that the
products of the mercuration reaction are 2 and 3.
Summary of crystal data, collection and refinement conditions for 7.
7
Formula
Formula weight
Radiation (wavelength, A)
Temperature
Crystal system
C14H13N4O2I
396.18
0.71073
173 (2)
Monoclinic
I2/m
13.447 (4)
6.69 3(2)
16.799 (6)
90.00
99.083 (3)
90.00
1492.9 (8)
4
1.763
2.155
0.069
ꢀ
Space group
ꢀ
3.3. Temporal distribution of products
a (A)
ꢀ
b (A)
ꢀ
Once the 1H NMR spectra of the predominant species were
assigned, it became possible to use the 8.76 ppm singlet of 5, the
8.81 ppm doublet of 4 and the 8.31 ppm doublet of 1 to probe the
composition of the reaction mixture over time (after treatment of
an aliquot with iodine). This method permitted an examination of
the evolution of the system in the form of the plot presented in
Fig. 2. The consumption of the monomercurated species to form the
dimercurated product is only significant after an induction period of
2 h. However, the reaction does not appear to proceed to comple-
tion, with 10% of starting material remaining unreacted after 24 h.
These features suggest that the mercuration steps might have
a reversible character; alas the high concentrations and especially
the formation of multiple by-products precluded the execution of
a proper kinetic study in this case. In spite of the limitations, this
information is useful to determine the reaction time that maximizes
the yield of the monosubstituted species (Table 2).
c (A)
a
b
g
(ꢂ)
(ꢂ)
(ꢂ)
3
ꢀ
Volume (A )
Z
dcalcd. (g/cm3)
m (mmꢁ1
)
a
R1
b
WR2
0.113
P
P
a
R1
¼
jjFoj ꢁ jFcjj/ jFoj.
P
P
b
2
2
WR2 ¼ ( wjjFoj ꢁ jFcjj / wjFoj )1/2
.
ambiguity about the exact position of metallation. Both the nitro
group and the azo bridge are known to have an ortho-directing
effect. An assignment based only on the 1H NMR chemical shifts
would be uncertain because the aromatic protons are subject to
both the inductive effect of the substituents and the anisotropic
shielding from the p-electron cloud. Single-crystal X-ray diffraction
3.4. Computational modelling
would be the best way to make a definitive structural assignment.
However, attempts to grow single crystals of the second band
species were unsuccessful and only dendritic microcrystals were
obtained. Instead the product of monosubstitution of 6 did yield
crystals of good enough quality for structural determination. Data
from the final refinement is provided in Table 1 (CCDC # 840135,
included in supplementary). The structure of 7 (Fig. 1) does show
the iodine atom ortho- to the azo bridge on the nitro-substituted
ring. The molecule exhibits a perfectly planar structure, which is
somewhat unexpected. Although there are no structurally charac-
terized ortho-iodo azobenzenes, the structures of similarly ortho-
halogenated azodyes [46] feature the halogenated ring rotated
between 2.05ꢂ and 8.60ꢂ from the average plane and, in one notable
extreme, 20-bromo-4-(N-(2-cyanoethyl)-N-(2-phenylethyl)amino)-
40-nitroazobenzene has rings which are rotated 38.19ꢂ from each
other [47]. The crystal structure of 7 does show displacement
ellipsoids somewhat elongated along b. This fact suggests that there
might be some disorder in the orientation of the rings with respect
to each other and the observed flat structure might be the average;
however, a satisfactory model for such a disorder could not be
The most intriguing characteristic of the azodye mercuration
process is the definite preference for reaction at the nitro-
substituted ring. Invoking a conventional EAS mechanism, the
reaction should occur on the amino-substituted ring because it is
more electron rich. DFT calculations were carried out in an attempt
to rationalize the experimental observations. In order to simplify
modelling of the system, the calculations were performed on
derivatives of the N,N-dimethyl dye, 6, and chloride was used in
lieu of the trifluoroacetate anion. The relative thermodynamic
stabilities of all possible products of monomercuration were
assessed by means of the comparison of the total bonding energies
of their minimized structures; the result is graphically presented in
Fig. 3. According to the DFT results, substitution at the positions
ortho to the amino (C2, C6) and nitro (C9, C11) groups is less
favourable than ortho to the azo bridge (C3, C5, C8, C12). There is
little preference (<7 kJ/mol) for functionalization at C5 and C8 over
C3 and C12 but overall the calculations do suggest there is no
preference for reaction on either aromatic ring. Therefore, if the
reaction were under thermodynamic control an approximately
equimolar distribution of all the products of metallation ortho to
the azo bridge would be obtained. Thermal rotation of the rings
would exchange positions C3 and C5 as well as C8 and C12, but at
least two isomer products of monometallation would be observed.
As the experimental observations appear to be in conflict with
the thermodynamic preferences, attention was devoted to the
reaction mechanism. In the initial stages of the reaction, the
mercuric salt would dissociate, generating a HgXþ cation which
would add to the azodye, forming a transient species of general
structure 8. The most likely sites for electrophilic attack could be
identified by the electrophilic Fukui function (Eq. (4)) [48,49],
which is graphically approximated as the projection of the squared
HOMO on the total electron density in Fig. 4. In this map, the darker
areas indicate that the carbon atoms C2, C4, C6, C8, C10, and C12 are
the most favourable points for an initial cation binding. However,
steric repulsion would hinder substitution at C2 and C6; similarly
C4 and C10 are unreactive because the nitro group and the azo
Fig. 1. ORTEP of 4-nitro, 2-iodo, 4-dimethylamino azobenzene, 7 (50% displacement
ꢀ
ellipsoids). For clarity, all hydrogen atoms are displayed as spheres of B ¼ 0.15 A.
ꢂ
ꢀ
Selected bond lengths (A) and angles ( ): Avg. N4eO 1.212 (12), Avg. N1eC (methyl)
1.443 (16), N4eC10 1.474 (13), C8eI1 2.092 (10), C7eN3 1.426 (12), N3eN2 1.257 (11),
N2eC4 1.382 (12), C1eN1 1.337 (12), O2eN4eC10 117.6 (9), C9eC8eI1 117.1 (7),
C8eC7eN3 116.6 (9), C7eN3eN2 114.1 (8), N3eN2eC4 114.6 (9), N2eC4eC5 125.4 (9),
C6eC1eN1 122.1 (11), C1eN1eC13 119.7 (10), C13eN1eC14 120.1 (10).

16
P.J.W. Elder et al. / Journal of Organometallic Chemistry 716 (2012) 11e18
Fig. 2. Temporal evolution of the main components of the mixture used for mercuration and subsequent iodination of 1. Spline lines highlight the trends but were not obtained by
actual fitting to rate laws.
Table 2
Composition of the mixture used for mercuration and subsequent iodination of 1.
Time (h)
Molar fractions
1
5
4
0
1
2
4
8
12
24
1.00
0.54
0.25
0.15
0.12
0.12
0.10
0.00
0.45
0.71
0.79
0.76
0.73
0.67
0.00
0.00
0.03
0.06
0.11
0.15
0.23
Fig. 5. Relative bonding energies of the transition states that would lead to each
possible product of metallation. (The carbon atoms and the positions of substitution
are numbered as in Fig. 1.)
bridge cannot be displaced. Although this result appears to agree
with the experiment, a more rigorous analysis requires examina-
tion of the activation barrier. In this respect the mechanism of EAS
reactions, including metallation, is suitable for quantum mechan-
ical modelling because the barriers can be approximated by
calculating the relative energies of 8, which in most cases can be
optimized [50e52]. In the present case, we calculated the geome-
tries of the transition states which would lead to the substitution of
each hydrogen on the aromatic rings. The relative energies of the
structures are presented graphically in Fig. 5. In this case the results
do show preference for the structure that would lead to metallation
at C8, as experimentally observed. The next structure in order of
energy is 14.7 kJ/mol above and would lead to mercuration at C2
(Chart 2).
Fig. 3. Relative bonding energies of the possible products of metallation of 6. (The
carbon atoms and the positions of substitution are numbered as in Fig. 1.)
f
ꢁðrÞ ¼ rNðrÞ ꢁ rNꢁ1ðrÞ
(4)
The results thus far discussed suggest that the substitution is
directed by the combined effect of the substituents on reactivity of
Fig. 4. Squared HOMO projected onto the total electron density of
isosurface).
6 (0.03 a.u.
Chart 2.

P.J.W. Elder et al. / Journal of Organometallic Chemistry 716 (2012) 11e18
17
Hierarchical Academic Research Computing Network (SHARCNET:
www.sharcnet.ca).
Appendix A. Supplementary material
CCDC # 840135 contains the supplementary crystallographic
data for this paper. These data can be obtained free of charge from
The Cambridge Crystallographic Data Centre via www.ccdc.cam.ac.
uk/data_request/cif.
Appendix B. Supplementary material
Supplementary material associated with this article can be
found, in the online version, at http://dx.doi.org/10.1016/j.
jorganchem.2012.05.040.
References
[1] N.N. Greenwood, A. Earnshaw, Chemistry of the Elements, second ed. But-
terworth Heinemann, Oxford, 1997, pp. 1222e1224.
[2] G.E. Coates, Organometallic Compounds, Wiley, London, 1960, pp. 72e87.
[3] R.C. Larock, Organomercury Compounds in Organic Synthesis, Springer-Verlag,
New York, 1985.
Fig. 6. Projections of the squared HOMOs onto the total electron densities (0.03 a.u.
isosurfaces) of two isomers of [Hg(Cl)(6)]þ coordinated by: a) N2 and b) N3.
each carbon. However, it has been suggested that the azo bridge
itself may play an active role in the mercuration mechanism as the
initial point of cation binding, from which HgXþ would migrate
onto the position of substitution [45]. Indeed the map of the elec-
trophilic Fukui function of 6 (Fig. 4) identifies the bridge as a reac-
tive site but only through N3, which is ortho to the position of actual
substitution. The effect of attachment of the cation on the reactivity
of the carbon atoms was then investigated with the maps of the
electrophilic Fukui function on optimized models of 6 binding
[HgCl]þ by N3 and, for comparison, N2 (Fig. 6). In both instances
binding to the azo bridge causes rotation of the aromatic rings, to
interplanar angles greater than 55ꢂ; in each case, the mercury atom
is in the plane of the adjacent benzene. The electrophilic Fukui
functions of the mercury bound 6 show little change with respect to
the free molecules, i.e. the reactivity of the C atoms is preserved,
this also suggests that mercury migration and substitution would
preferentially happen onto C8 and C12, in agreement with the
experiment.
[4] J.L. Wardell, Mercury, in: G. Wilkinson, F.G.A. Stone, E.W. Abel (Eds.), Compre-
hensive Organometallic Chemistry, vol. 2, Pergamon Press Ltd., Toronto, 1982,
pp. 863e978.
[5] M. Tschinkl, A. Schier, J. Riede, F.P. Gabbaï, Organometallics18(1999) 2040e2042.
[6] M. Tschinkl, F.P. Gabbaï, J. Chem. Crystallogr. 33 (2003) 595e598.
[7] T. Taylor, O. Elbjeirami, C. Burress, M. Tsunoda, M. Bodine, M. Omary,
F.P. Gabbaï, J. Inorg. Organomet. Polym. 18 (2008) 175e179.
[8] M.R. Haneline, F.P. Gabbaï, Angew. Chem. Int. Ed. 43 (2004) 5471e5474.
[9] A.A. Yakovenko, J.H. Gallegos, M.Y. Antipin, A. Masunov, T.V. Timofeeva, Cryst.
Growth Des. 11 (2011) 3964e3978.
[10] T.J. Taylor, C.N. Burress, L. Pandey, F.P. Gabbaï, Dalton Trans. (2006) 4654e4656.
[11] T.J. Taylor, C.N. Burress, F.P. Gabbaï, Organometallics 26 (2007) 5252e5263.
[12] G.B. Deacon, C.M. Forsyth, P.C. Junk, E.I. Izgorodina, T.J. Ness, G. Meyer,
I. Pantenburg, CrystEngComm 13 (2011) 88e92.
[13] G.B. Deacon, C.M. Forsyth, P.C. Junk, T.J. Ness, E. Izgorodina, J. Baldamus,
G. Meyer, I. Pantenburg, J. Hitzbleck, K. Ruhlandt-Senge, Eur. J. Inorg. Chem.
(2008) 4770e4780.
[14] C.L. Dorsey, P. Jewula, T.W. Hudnall, J.D. Hoefelmeyer, T.J. Taylor, N.R. Honesty,
C.-W. Chiu, M. Schulte, F.P. Gabbaï, Dalton Trans. (2008) 4442e4450.
[15] K.A. Kobe, T.F. Doumani, Ind. Eng. Chem. 33 (1941) 170e176.
[16] Y. Ogata, M. Tsuchida, J. Org. Chem. 20 (1955) 1637e1643.
[17] H.C. Brown, C.W. McGary, J. Am. Chem. Soc. 77 (1955) 2300.
[18] G.A. Olah, I. Hashimoto, H.C. Lin, Proc. Natl. Acad. Sci. U.S.A. 74 (1977) 4121.
[19] H.C. Brown, C.W. McGary, J. Am. Chem. Soc. 77 (1955) 2310e2312.
[20] Y. Ogata, M. Tsuchida, J. Org. Chem. 20 (1955) 1644e1649.
[21] H.B. Singh, W.R. McWhinnie, J. Chem. Soc. Dalton Trans. (1985) 821.
[22] Z. Li, W. Wu, Q. Li, Y. Gui, L. Xiao, Y. Liu, C. Ye, J. Qin, Z. Li, Angew. Chem. Int. Ed.
49 (2010) 2763.
4. Summary and outlook
[23] J. Garcia-Amorós, H. Finkelmann, D. Velasco, Chem. Eur. J. 17 (2003) 6518.
[24] A.G. Chen, D.J. Brady, Appl. Phys. Lett. 62 (1993) 2920.
[25] M. Lee, H.E. Katz, C. Erben, D.M. Gill, P. Gopalan, J.D. Heber, D.J. McGee, Science
298 (2002) 1401.
[26] P. Gopalan, H.E. Katz, D.J. McGee, C. Erben, T. Zielinski, D. Bousquet, D. Muller,
J. Grazul, V. Olsson, J. Am. Chem. Soc. 126 (2004) 1741e1747.
[27] M.M.M. Raposo, M.C.R. Castro, A.M.C. Fonseca, P. Schellenberg, M. Belsley, 67
(2011) 5189e5198.
[28] A.R. Lepley, W.A. Khan, J. Org. Chem. 33 (1968) 4362e4368.
[29] E.C. Ashby, J. Oswald, J. Org. Chem. 53 (1988) 6068e6076.
[30] H. Finkelstein, Chem. Ber. 43 (1910) 1528e1532.
[31] A.I. Vogel, A Textbook of Practical Organic Chemistry, Longman, London, 1957,
p. 1188.
[32] SMART and SAINT, Siemens Analytical X-ray Instruments Inc, Madison, Wisconsin,
USA, 1995.
[33] SADABS, Bruker AXS Inc., Madison, Wisconsin, USA, 2001.
[34] (a) G.M. Sheldrick, SHELX97 [Includes SHELXS97, SHELXL97, CIFTAB].
Programs for Crystal Structure Analysis. Release 97-2, Institüt für Anorgani-
sche Chemie der Universität, Göttingen, Germany, 1997;
The combination of electron donating and withdrawing groups
reduces the reactivity of azobenzene towards direct mercuration,
thus very aggressive conditions such as mercury(II) trifluoroacetate
in anhydrous trifluoroacetic acid are necessary to achieve the
functionalization of the chromophore. Even then, the reaction does
not proceed quantitatively, leaving behind a significant proportion
of starting material. The two major products are the result of
sequential metallation of the nitro-substituted ring, at the positions
ortho to the azo bridge. The DFT study of the reagent and models of
the possible isomers of the first product and the corresponding
transition states points to a system under kinetic control and the
electrophilic Fukui function correctly identifies the most reactive
positions. The results and methods of this study constitute a tool for
the preparation of intermediates in the synthesis of functional
molecules and materials derivatized with organic chromophores.
(b) G.M. Sheldrick, Acta Crystallogr. A64 (2008) 112e122.
[35] C.F. Macrae, I.J. Bruno, J.A. Chisholm, P.R. Edgington, P. McCabe, E. Pidcock,
L. Rodriguez-Monge, R. Taylor, J. van de Streek, P.A. Wood, J. Appl. Crystallogr.
41 (2008) 466e470.
Acknowledgements
[36] G. te Velde, F.M. Bickelhaupt, E.J. Baerends, C. Fonseca Guerra, S.J. A. van
Gisbergen, J.G. Snijders, T. Ziegler, J. Comput. Chem. 22 (2002) 931e967.
[37] C. Fonseca Guerra, J.G. Snijders, G. te Velde, E.J. Baerends, Theor. Chem. Acc. 99
(1998) 391e403.
[38] E.J. Baerends, J. Autschbach, A. Bérces, C. Bo, P.M. Boerrigter, L. Cavallo, D.P.
Chong, L. Deng, R.M. Dickson, D.E. Ellis, M. van Faassen, L. Fan, T.H. Fischer, C.F.
This project was supported by grants from the Natural Sciences
and Engineering Research Council of Canada (NSERC), the Canada
Foundation for Innovation (CFI), and the Ontario Innovation Trust
(OIT). This work was made possible by the facilities of the Shared

18
P.J.W. Elder et al. / Journal of Organometallic Chemistry 716 (2012) 11e18
Fonseca Guerra, S.J.A. van Gisbergen, J.A. Groeneveld, O.V.Gritsenko, M.
[44] E. van Lenthe, A. Ehlers, E.J. Baerends, J. Chem. Phys. 110 (1999) 8943e8953.
[45] P.V. Roling, D.D. Kirt, J.L. Dill, S. Hall, C. Hollstrom, J. Organomet. Chem. 116
(1976) 39e53.
[46] X.-L. You, Y. Zhang, Y. Zhang, D.-C. Zhang, Acta Crystallogr. Sect. E. Struct. Rep.
Online 61 (2005) o2656eo2658.
Grüning, F.E. Harris, P. van der Hoek, H. Jacobsen, G. Jensen, G. van Kessel, F.
Kootstra, E. van Lenthe, D.A. McCormack, A. Michalak, V.P. Osinga, S. Patch-
kovskii, P.H.T. Philipsen, D. Post, C.C. Pye, W. Ravenek, P. Ros, P.R.T. Schipper,
H.G. Schreckenbach, J.G. Snijders, M. Solà, M. Swart, D. Swerhone, G. te Velde,
P. Vernooijs, L. Versluis, O. Visser, F. Wang, E. van Wezenbeek, G. Wiesenek-
ker, S.K. Wolff, T.K. Woo, T. Ziegler, ADF 2010.01, 2010.
[47] M.L. Glowka, Z. Olubek, Acta Crystallogr. C. Cryst. Struct. Commun. 50 (1994)
458e460.
[39] S.J.A.vanGisbergen,J.G.Snijders,E.J.Baerends,Phys.Rev.Lett.78(1997)3097e3100.
[40] S.J.A. van Gisbergen, J.G. Snijders, E.J. Baerends, J. Chem. Phys. 109 (1998)
10644e10656.
[41] S.H. Vosko, L. Wilk, M. Nusair, Can. J. Phys. 58 (1980) 1200e1211.
[42] J.P. Perdew, Phys. Rev. B 33 (1986) 8822e8824.
[48] W. Yang, R.G. Parr, R. Pucci, J. Chem. Phys. 81 (1984) 2862e2863.
[49] R.G. Parr, W. Yang, J. Am. Chem. Soc. 106 (1984) 4049e4050.
[50] S. Kolboe, S. Svelle, B. Arstad, J. Phys. Chem. A 113 (2009) 917e923.
[51] S. Kolboe, J. Phys. Chem. A 115 (2011) 3106e3115.
[52] J.S. Kadavalli, D.R. Boyd, D. Coyne, J.R. Keeffe, D.A. Lawlor, A.C. MacCormac,
R.A.M. O’Ferrall, S.N. Rao, N.D. Sharma, Org. Lett. 12 (2010) 5550e5553.
[43] J.P. Perdew, Y. Wang, Phys. Rev. B 45 (1992) 13244e13249.