Syntheses and Optical Properties of Azo-Functionalized Ruthenium Alkynyl Complexes

Article abstract of DOI:10.1002/cplu.201600222

The syntheses of trans-[Ru(C≡C-1-C₆H₄-4-N=N-1-C₆H₄-4-C≡C-1-C₆H₄-4-NO₂)Cl(L₂)₂] (L₂=dppm (Ru1), dppe) (Ru2)), trans-[Ru(C≡C-1-C₆H₄-4-N=N-1-C₆H₄-4-(E)-CH=CH-1-C₆H₄-4-NO₂)Cl(dppe)₂] (Ru3), and trans-[Ru(C≡C-1-C₆H₄-4-(E)-CH=CH-1-C₆H₂-2,6-Et₂-4-N=N-1-C₆H₄-4-NO₂)Cl(dppe)₂] (Ru4) are reported, together with those of precursor alkynes. Their electrochemical properties were assessed by cyclic voltammetry (CV), linear optical and quadratic nonlinear optical (NLO) properties assayed by UV/Vis-NIR spectroscopy and hyper-Rayleigh scattering studies at 1064 nm, respectively, and their linear optical properties in the formally Ru^III state examined by UV/Vis-NIR spectroelectrochemistry. These data were compared to those of analogues with E-ene and yne linkages in place of the azo groups. Computational studies using time-dependent density functional theory were undertaken on model compounds (Ru2′–Ru4′) to rationalize the optical behaviour of the experimental complexes.

Full text of DOI:10.1002/cplu.201600222

DOI: 10.1002/cplu.201600222
Full Papers
Syntheses and Optical Properties of Azo-Functionalized
Ruthenium Alkynyl Complexes**
[a]
[b]
[b]
[b]
[a]
Dilan Wei, Mahesh S. Kodikara, Mahbod Morshedi, Graeme J. Moxey, Huan Wang,
[
b]
[b]
[a]
[b]
[a, b]
Genmiao Wang, Cristꢀbal Quintana, Chi Zhang, Rob Stranger, Marie P. Cifuentes,
[a, b]
and Mark G. Humphrey*
The syntheses of trans-[Ru(CꢀC-1-C H -4-N=N-1-C H -4-CꢀC-1-
1064 nm, respectively, and their linear optical properties in the
6
4
6
4
III
C H -4-NO )Cl(L ) ] (L =dppm (Ru1), dppe) (Ru2)), trans-[Ru(Cꢀ formally Ru state examined by UV/Vis-NIR spectroelectro-
6
4
2
2 2
2
C-1-C H -4-N=N-1-C H -4-(E)-CH=CH-1-C H -4-NO )Cl(dppe) ]
chemistry. These data were compared to those of analogues
with E-ene and yne linkages in place of the azo groups. Com-
putational studies using time-dependent density functional
theory were undertaken on model compounds (Ru2’–Ru4’) to
rationalize the optical behaviour of the experimental com-
plexes.
6
4
6
4
6
4
2
2
(
Ru3), and trans-[Ru(CꢀC-1-C H -4-(E)-CH=CH-1-C H -2,6-Et -4-
6
4
6
2
2
N=N-1-C H -4-NO )Cl(dppe) ] (Ru4) are reported, together with
6
4
2
2
those of precursor alkynes. Their electrochemical properties
were assessed by cyclic voltammetry (CV), linear optical and
quadratic nonlinear optical (NLO) properties assayed by UV/
Vis-NIR spectroscopy and hyper-Rayleigh scattering studies at
Introduction
Materials that can modify the propagation characteristics of
light (e.g. frequency, phase, path, etc) are urgently needed for
applications in current and prospective photonics technolo-
gies. Organic molecules possessing such “nonlinear optical”
phenyl to heterocyclic moieties with a lower aromatic stabiliza-
tion energy; all of these modifications favourably impact on
NLO parameters. In other studies relating structure to NLO
properties, it was noted that the key quadratic NLO coefficient
b is increased on certain modifications of the phenyl-based p-
bridge (proceeding from biphenyl to Z-stilbene, tolane, and E-
stilbene, because biphenyl exhibits unfavourable ortho-H steric
pressure, reducing bridge coplanarity, while compounds with
a Z-stilbene bridge have decreased charge separation in the
(NLO) properties that can fulfill these desired functions gener-
ally contain p-systems with highly delocalizable electron densi-
ty. Very early in the search for NLO-efficient organic molecules,
it was found that dipolar species such as p-nitroaniline (PNA)
and 4-dimethylamino-4’-nitrostilbene (DANS) were particularly
active, although subsequent research with molecules possess-
ing other multipolar charge distributions (quadrupolar, octupo-
[2]
key excited state compared to that of E-stilbene analogues).
In contrast to the aforementioned structural modifications
that improve NLO performance, the incorporation of more
electronegative heteroatoms such as nitrogen into the p-
bridge in organic NLO materials (proceeding from E-ene- to
imino- or azo-linked compounds) frequently leads to a reduc-
tion in b value due to charge localization at the electronega-
[
1]
lar) has also afforded NLO-efficient materials. Improvements
to the classical donor–p-bridge–acceptor composition of mole-
cules such as PNA and DANS have focussed on modifying all
three components. Thus, the donor group strength increases
on proceeding from amino and dimethylamino to julolidinyl
and diphenylamino, and the acceptor group strength increases
on proceeding from nitro to tricyanovinyl and 2-dicyanomethy-
len-3-cyano-4,5,5-trimethyl-2,5-dihydrofuran, while delocaliza-
tion through the p-bridge is improved on proceeding from
[3]
tive nitrogen atom(s). Nevertheless, the azo-linked com-
pounds, in particular, are of interest as archetypical examples
of photochemical switches; irradiation of azobenzenes at spe-
cific wavelengths can provide a facile means to reversibly con-
vert between the E- and Z-configured forms, although thermal
conversion of the Z form to the E form imposes constraints on
[
a] D. Wei, H. Wang, Prof. C. Zhang, Prof. M. P. Cifuentes, Prof. M. G. Humphrey
School of Chemical and Material Engineering, Jiangnan University
Wuxi, Jiangsu Province 214122 (P. R. China)
[4]
the utility of this specific molecular switch.
Replacing the organic donor group in the classical donor–p-
bridge–acceptor molecular composition with a ligated metal
unit can confer certain advantages: the flexibility of metal, va-
lence electron count, coligands, and coordination geometry
E-mail: mark.humphrey@jiangnan.edu.cn
[
b] M. S. Kodikara, Dr. M. Morshedi, Dr. G. J. Moxey, Dr. G. Wang, C. Quintana,
Prof. R. Stranger, Prof. M. P. Cifuentes, Prof. M. G. Humphrey
Research School of Chemistry, Australian National University
Canberra, ACT 2601 (Australia)
[5]
can all impact favourably on the molecular NLO coefficients,
E-mail: mark.humphrey@anu.edu.au
while the facile reversible oxidation/reduction undergone by
many metal complexes can introduce important functionality
[
**] Part 58: Organometallic Complexes for Nonlinear Optics.
Supporting information and the ORCID identification number(s) for the
author(s) of this article can be found under http://dx.doi.org/10.1002/
cplu.201600222.
[6]
and the potential to function as molecular NLO switches. We
have previously reported that azo-containing arylalkynyl com-
ChemPlusChem 2016, 81, 1 – 9
1
ꢁ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
These are not the final page numbers! ÞÞ

Full Papers
plexes can display significant NLO performance, with the rele-
vant parameters amongst the largest for organometallic com-
(diphenylphosphino)ethane] can be highly NLO-efficient and
II/III
can undergo fully reversible formally Ru redox processes ac-
[
7]
[10]
plexes. However, the limited examples thus far generally
have short p-bridges, and thus the NLO behaviour may not be
optimized. We report herein the syntheses of several new azo-
containing arylalkyne compounds and derivative alkynyl com-
plexes, spectroscopic and in some cases X-ray structural char-
acterization of the new compounds, spectroscopic, electro-
chemical, linear and nonlinear optical, and in most cases X-ray
structural characterization of the new complexes, comparison
to precedents with shorter p-bridges or a related ene/yne-
linked composition, and theoretical studies to rationalize the
experimental observations.
companied by strong linear and nonlinear optical changes.
Complexes with this specific composition were therefore tar-
geted in the present study.
The alkynyl complexes were accessible in a straightforward
way employing existing methodology, namely reaction of the
corresponding terminal alkyne with a metal chloride com-
[11]
plex.
The syntheses of the alkynes proceeded smoothly
(Scheme 1), the products being characterized where possible
using IR and multinuclear NMR spectroscopies, high-resolution
mass spectrometry, and elemental analyses (see the Support-
ing Information). Thus, reaction of 1-(4-bromophenyl)-2-{4-[(trii-
[12]
sopropylsilyl)ethynyl]phenyl}diazene
with
trimethylsilyl-
[13]
acetylene under Sonogashira coupling conditions afforded
the difunctionalized trimethylsilylethynyl triisopropylsilylethyn-
yl compound 1 (95%), which was selectively desilylated at the
trimethylsilyl group using potassium carbonate to give 2 in ex-
cellent yield (93%). A second coupling reaction using 1-iodo-4-
Results and Discussion
Synthesis and characterization
A specific complex composition was targeted in the present
studies. Previous reports had revealed an increase in the key
NLO coefficients upon increasing the formal metal valence
[14]
nitrobenzene afforded 3 (71%), with azo and yne linkages
separating the silyl-protecting group from the nitro acceptor.
Synthesis of the analogue in which the yne linkage is re-
placed by an E-ene group was also successful, albeit in lower
overall yield. Oxidation of 4-aminobenzonitrile using potassium
[
8]
electron (VE) count, so 18 VE complexes were targeted in the
present work. The earlier studies had also shown the efficiency
series ironꢁrutheniumꢁosmium for Group 8 metal alkynyl
[
9]
[15]
complexes, with the facile synthesis of the ruthenium exam-
ples more than compensating for their slightly decreased NLO
efficiency compared to the kinetically more inert osmium ana-
logues, so the present studies focused on ruthenium com-
plexes. The precedent research had also revealed that pseudo-
octahedral complexes trans-[Ru(CꢀCR)Cl(L ) ] [L =dppm, dppe;
peroxymonosulfate afforded the corresponding labile nitro-
soarene, which was employed without further purification; re-
[16]
action with 4-[(triisopropylsilyl)ethynyl]aniline gave the nitrile
(E)-4-((4-((triisopropylsilyl)ethynylphenyl)diazenyl)benzonitrile
(4, 75%). Reduction of 4 using (diisobutyl)aluminium hydride
(DIBAL-H) afforded the aldehyde 5 (47%) which was reacted
2
2
2
[17]
dppm=bis(diphenylphosphino)methane,
dppe=1,2-bis-
with diethyl (4-nitrobenzyl)phosphonate under basic condi-
Scheme 1. Preparation of diazene-containing alkynes 3, 6 and 11.
&
ChemPlusChem 2016, 81, 1 – 9
www.chempluschem.org
2
ꢁ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Papers
tions to afford the silyl-protected terminal acetylene (E)-1-[4-
(E)-4-nitrostyryl)phenyl]-2-{4-[(triisopropylsilyl)ethynyl]phenyl}-
replacing dppe by dppm (proceeding from Ru2 to Ru1), re-
placing an E-ene by an yne linkage (proceeding from Ru3 to
Ru2) and locating the azo group closer to the nitro (proceed-
(
diazene (6, 19%) in low yield.
The preparation of (E)-1-[3,5-diethyl-4-((E)-4-ethynylstyryl)-
phenyl]-2-(4-nitrophenyl)diazene (11), in which the bridging
azo and E-ene groups of 6 are interchanged, required a se-
quence including modified Sandmeyer reactions; in contrast to
ing from Ru3 to Ru4). The n(NO ) vibration is relatively invari-
2
31
ꢂ1
ant at around 1580 cm . The P NMR spectra confirm trans
stereochemistry at the metal center, with a singlet at about
ꢂ7 ppm for the dppm-containing complex Ru1, and at about
49 ppm for the dppe-containing Ru2–Ru4.
6
, compound 11 necessitated incorporation of alkyl groups to
ensure sufficient solubility. Diazotization of p-nitroaniline using
The identities of 3, 4, 6, 8, Ru1, Ru2, and Ru3 were con-
firmed by single-crystal X-ray diffraction studies (Figure 1 and
[
18]
isoamyl nitrite as the source of the nitronium ion followed
by reaction with 2,6-diethylaniline afforded the diazene (E)-2,6-
diethyl-4-[(4-nitrophenyl)diazenyl]aniline (7, 71%) in good
yield. The arylamino group underwent reaction with isoamyl
nitrite and then potassium iodide to give the iodo analogue,
(
E)-1-(3,5-diethyl-4-iodophenyl)-2-(4-nitrophenyl)diazene
(8,
65%), which was in turn reacted with copper(I) cyanide/ferric
chloride to give the nitrile (E)-2,6-diethyl-4-[(4-nitrophenyl)dia-
zenyl]benzonitrile (9) in 47% yield. Reduction of 9 with DIBAL-
H afforded the corresponding aldehyde (10, 47%) which was
reacted with diethyl {4-[(trimethylsilyl)ethynyl]benzyl}phospho-
nate to give the desired terminal acetylene (11, 28%).
The preparation of the new ruthenium complexes proceed-
ed by established methodologies (Scheme 2). Reaction be-
2 2
Figure 1. Molecular structure of Ru2·CH Cl , with thermal ellipsoids set at
the 40% probability level. Hydrogen atoms and the lattice dichloromethane
molecule have been omitted for clarity. Selected bond lengths [ꢂ] and
angles [8] are provided in Table S2.
Figures S1–S7, Tables S1–S3). The bond lengths and angles are
unremarkable for the organic compounds and the alkynyl com-
plexes (distortions from colinearity at the Cl1-Ru1-C1-C2-C3
vector are often seen with alkynyl complexes and commonly
attributed to packing forces), while the dihedral angles are
consistent with good coplanarity through the p-bridges in all
complexes.
Optical and electrochemical studies
The UV/Vis spectra of the silyl-protected and terminal acety-
lenes 3, 6 and 11 show absorption maxima at about 380–
Scheme 2. Syntheses of ruthenium complexes Ru1–Ru4.
400 nm, with the longest-wavelength band due to 6 (Table S5,
Figure S8). It is possible that the low-energy bands for 3 and 6
correspond to the superposition of two low-energy bands: the
lowest-energy band in the UV/Vis spectrum of 11 is only about
half as intense as the low-energy bands of 3 and 6, and the
spectrum of 11 contains a further similar-intensity band at
about 289 nm, suggesting that location of the azo group
remote from the nitro group (3, 6) red-shifts this band, result-
ing in its superposition with the existing lowest-energy band.
Coordination of 3 to the ruthenium centre results in a signifi-
cant bathochromic shift of lmax from 380 nm to 540 nm for the
dppm-coordinated complex Ru1, and to about 550 nm for the
dppe-containing complex Ru2. A similar red-shift (from 375 nm
to 550 nm) was seen on coupling 11 to the bis(dppe)rutheni-
um centre in Ru3 (Figure S9). All of these complexes contain
[
19]
tween cis-[RuCl (dppm) ] and the triisopropylsilyl-protected
2
2
acetylene 3 in the presence of fluoride afforded Ru1 (59%);
the fluoride desilylated the acetylene in situ, eliminating the
need for a further step to isolate the terminal alkyne. The 16-
[
20]
electron, five-coordinate species [RuCl(dppe) ]PF6
was the
2
precursor for the three dppe-containing complexes. It reacted
under similar in situ desilylation conditions with the triisopro-
pylsilyl-protected acetylenes 3 and 6 to give Ru2 (44%) and
Ru3 (55%), respectively, and reacted with the terminal acety-
lene 11 in the absence of fluoride to give Ru4 (88%). The vi-
brational spectra of the complexes possess bands correspond-
ing to the stretching mode of the metal-bound CꢀC at 2037–
ꢂ1
2
055 cm for Ru1–Ru4; this band moves to higher energy on
ChemPlusChem 2016, 81, 1 – 9
www.chempluschem.org
3
ꢁ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
These are not the final page numbers! ÞÞ

Full Papers
the diazene group proximal to the metal atom. A slightly
smaller shift to 503 nm is seen on coordination of 11 to give
Ru4, where the diazene group is distal to the ruthenium. Ex-
tinction coefficients are similar for Ru1–Ru3, and slightly lower
[
a]
Table 2. Summary of HRS results, and including simple and modified
two-level calculations.
[
a]
[b]
Table 1. Cyclic voltammetric and UV/Vis data for resting-state and oxi-
dized complexes (measured in CH Cl ).
2
2
1
/2
II
ꢂ1
III
ꢂ1
E
(
[V]
Ru n˜ max [cm
(e)
]
Ru n˜ max [cm
(e)
]
[
c]
[c]
DE [V], ipc/ipa
)
Ru1
0.57
0.09, 1)
18300 (2.6), 27300
(2.5), 37400 (3.4)
11500 (0.43), 20300
(sh, 1.5), 25200 sh
(
(2.2), 27000 (2.3)
Ru2
Ru3
Ru4
0.62
0.09, 1)
0.66
0.05, 1)
0.53
0.09, 1)
18200 (4.0), 27100
(4.2), 40100 sh (5.8)
18200 (3.8), 26000
(3.8), 40300 (5.2)
19800 (1.1), 26000
(4.7), 40 000 (5.5)
11100 (2.6), 21100
(3.6), 38200 (6.1)
11000 (0.68), 20900
(4.2), 38700 (5.7)
10900 (1.8), 20500
(2.1), 23500 sh (2.5),
(
[b]
[c]
[c]
[c]
[c]
Cmpd
l
max
b
1064
b
0
b
0
b
0
(
D=0
D=30 nm
D=50 nm
Ru1
Ru2
Ru3
Ru4
Ru5
Ru6
Ru7
551
552
548
503
510
466
468
2150
2700
3200
1300
1215ꢃ146
2090ꢃ66
2525ꢃ175
ꢂ115
145
190
195
125
–
185
240
260
145
–
(
ꢂ155
2
6000 (3.0), 36900 (5.7)
(
c)
ꢂ140
Ru6
0.49
–, 1)
21500 (1.5)
9700 (0.18), 10800
(0.16)
110
(
[
d]
51ꢃ6
395ꢃ12
460ꢃ32
[a] Pt disc working electrode, Pt-wire auxiliary electrode and Ag/AgCl ref-
[e]
[e]
–
–
–
–
1
/2
erence electrode, with ferrocene internal standard (E =0.56 V; [0.09, 1].
ꢂ1
4
ꢂ1 ꢂ1
[b] cm ; [10 cm m ]. [c] Ref. [11b].
[
a] Measured at 1064 nm. [b] nm, Ru1–Ru4, Ru6 in CH
THF. [c] 10 esu in THF. [d] Ref. [7a]. [e] Ref. [11b].
2
Cl
2
, Ru5, Ru7 in
ꢂ30
for Ru4 (Table 1, Figure S9); this difference in the absorption
band intensity is consistent with observations with the precur-
sor alkynes but, in contrast to the alkynes, no clear additional
band is apparent at higher energy (although the broad absorp-
tion profile makes such analysis difficult). Comparison of the
thin-layer electrochemical (OTTLE) cell. Using the redox data
obtained from the CV experiments, oxidation potentials slightly
higher than indicated on the CV scans were applied to di-
lmax values to that of trans-[Ru(CꢀC-1-C H -4-CꢀC-1-C H -4-N= chloromethane solutions of Ru1–Ru4, affording spectral pro-
6
4
6
4
[
7a]
III
N-1-C H -4-NO )Cl(dppm) ] (Ru5, Table 2) confirms the red-
gressions indicating the clean formation of the Ru species
6
4
2
2
shift in proceeding from complexes with a distal diazene to
complexes with a proximal electron-withdrawing diazene
group.
(Figure 2 and Figures S11–S14); subsequent application of a re-
ducing potential to the oxidized species resulted in a return to
the starting spectra, with isosbestic points in each case. Fre-
quency maxima and extinction coefficients are collected in
Table 1 and show the emergence of a new low-energy band at
Cyclic voltammetry (CV) studies of solutions of the metal-
containing complexes in dichloromethane show a fully reversi-
II/III
+
ble Ru oxidation process at 0.62 V (cf. FcH/FcH couple at
.56 V) (Table 1, Figure S10) for Ru2, with a significant shift to
0
a higher potential on replacing the yne linkage with the E-ene
linkage proceeding to Ru3 (we recognize that the relevant
HOMOs in such ruthenium alkynyl complexes have an alkynyl
ligand component, but they certainly have significant metal-
centred content, and the redox processes have been labelled
as such for simplicity). Changing the coordinated diphosphine
ligand from dppe (Ru2) to dppm (Ru1) results in a slight shift
to lower potentials. Oxidation of the metal centre is most
easily achieved for Ru4 (0.53 V), for which the electron-deplet-
ing effect of the azo group is reduced due to its increasing dis-
tance from the metal. The reduction processes are nitro-cen-
tred, occurring at about ꢂ0.9 V for Ru1–Ru3, and ꢂ0.8 V for
Ru4, again consistent with the varying location of the azo
group (the electron-withdrawing azo proximal to the nitro-
phenyl group in 4 results in more facile reduction).
Figure 2. Progressive changes to the UV/Vis-NIR spectra on application of an
oxidizing potential of 0.7 V to a solution of Ru4. Measured in CH Cl at
2 2
Spectroelectrochemistry allows the measurement of spectral
data, in this case UV/Vis-NIR data, whilst a redox potential is
applied to a solution of the complex in an optically transparent
ꢂ258C. Pt gauze working electrod, Ag wire reference electrode and Pt wire
auxillary electrode.
&
ChemPlusChem 2016, 81, 1 – 9
www.chempluschem.org
4
ꢁ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Papers
ꢂ1
about 11000 cm for Ru1–Ru4, with a progression to lower
model, while for Ru1–Ru3, for which lmax values are close to
550 nm, the modified two-level model yields a greater discrep-
ancy; nevertheless, the trend in experimental maximal values
(Ru3>Ru2>Ru1>Ru4) is maintained for the various two-
level approximations.
frequency on replacement of coligand dppm by dppe (pro-
ceeding from Ru1 to Ru2), replacement of the yne linkage
with the E-ene linkage (proceeding from Ru2 to Ru3), or loca-
tion of the azo group further from the metal centre (proceed-
ing from Ru2 to Ru4).
Computational studies
Hyper-Rayleigh scattering studies
Time-dependent density functional theory (TD-DFT) calcula-
tions were undertaken on a set of model complexes (Figure 3)
The quadratic nonlinearities of Ru1, Ru2, Ru3, and Ru4 were
determined at 1064 nm using the hyper-Rayleigh scattering
(HRS) technique and 20 ns pulses; the results are given in
Table 2, together with the two-level-corrected values, and data
for related complexes collected under similar conditions. Prob-
lems with the two-level model have been discussed previous-
[
10a]
ly.
While it is not generally considered adequate for donor–
bridge–acceptor organometallics such as those herein, it may
be useful in comparing closely related complexes and, with
this caveat in mind, one can compare the data. Both experi-
mental and two-level-corrected nonlinearities are consistent
with replacement of coligand dppm by dppe leading to an in-
crease in quadratic nonlinearity (proceeding from 1 to 2). Re-
placement of the yne linkage by the E-ene linkage (proceeding
from 2 to 3) affords data that are within the ꢃ10% error mar-
gins. Location of the azo group distal rather than proximal to
the metal (proceeding from Ru2 to Ru4 or from Ru1 to Ru5)
leads to a decrease in quadratic NLO parameters. Further com-
parison to our previously reported data for Ru6 and Ru7 re-
veals that the frequency-dependent b values for the azo-con-
taining complexes from the present study are comparatively
large, but the proximity of the linear optical absorption bands
to the second-harmonic wavelength of the Nd:YAG laser em-
ployed in the present studies (532 nm) inevitably results in fre-
quency-independent data for the azo-containing compounds
that are significantly smaller than those of the E-ene-contain-
ing analogues.
Figure 3. Model compounds used in computational studies. Pꢂ
2 2 2 2
P=Me PCH CH PMe .
to rationalize the experimental UV/Vis spectra. Calculated low-
ꢂ1
lying (<30000 cm ) singlet excited-state transitions and oscil-
lator strengths are listed in Table S7, alongside the experimen-
tal data. Frontier molecular orbitals (FMOs) involved in the
main transitions for the model compounds Ru2’ and Ru4’ are
given in Figures S15 and S16, respectively; the FMOs obtained
for Ru2’ and Ru3’ are very similar in their composition.
Calculated transitions for all model complexes in the lower-
ꢂ1
energy region (<30000 cm ) can be divided into two differ-
ent groups; each model complex exhibits an intense transition
ꢂ1
ꢂ1
in the region 15000 cm to 17000 cm and one or two tran-
ꢂ1
sitions with large oscillator strengths in the region 21500 cm
ꢂ1
to 25000 cm . The calculated data are consistent with the ex-
Various modified two-level models have been proposed to
account for the proximity of the second-harmonic wavelength
to the optical absorption maximum, and specifically the effects
of absorption line broadening and vibrational modes in both
perimental data, particularly for model compounds Ru2’ and
Ru3’. The second spectral band for Ru2 and Ru3 is broader
than the lowest-energy band; this can be attributed to the
presence of several transitions close in energy in this region,
according to the calculated data (Table S7). The experimental
lowest-energy band is slightly more intense than the second
lowest-energy band, and this behaviour is nicely reproduced
by the computational outcomes of the corresponding model
compounds (Ru2’/Ru3’). In the case of Ru4’, these calculated
transitions are substantially red-shifted compared to the UV/Vis
data. However, the incorporation of ethyl substituents in the
laboratory complex Ru4 to overcome the solubility problems
may induce steric crowding around the E-ene linkage in the
coplanar conformation, and thus distort the complex from the
ideal coplanar arrangement of the phenylene groups which is
assumed in these calculations. Further calculations reveal that
distortion from a coplanar structure blue-shifts the low-lying
transitions, in particular the lowest-energy band, and decreases
their intensity.
[
21]
the ground and excited states. We have therefore examined
our experimental data utilizing the modified two-level model
[
21b]
of Wang
[Eq. (1)]:
<
b > HRS ¼< b > Fðu,dÞ
ð1Þ
0
2
2
where
F(u,d)=(1/3){[2/(1+2u)-1/(1-u )+2(1–2u)/((1–2u) +
2
2
2 2 1/2
d)] +[2d/((1–2u) +d )] } , u=lmax/l, d=D/lmax, D is a damp-
ing factor associated with the broadening of the absorption
peak, and l is the fundamental light wavelength (1064 nm in
this case). Note that, if D=0, the expression above simplifies
to that of the usual two-level model. Using this equation and
assuming that D is 30–50 nm (estimated from the UV/Vis spec-
tra: D likely differs from sample to sample), the recalculated
<
<
b0 > values are listed in Table 2. For Ru4, the recalculated
b0 > is close to that calculated with the simple two-level
ChemPlusChem 2016, 81, 1 – 9
www.chempluschem.org
5
ꢁ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
These are not the final page numbers! ÞÞ

Full Papers
Although the highest occupied molecular orbital (HOMO) for
all model compounds is broadly similar (i.e. the HOMO is
mainly located on the ClRuCꢀC-1-C H -4- moiety; Figures S15
when two spectral bands relatively close in energy exhibit
MLCT character; this is particularly the case with Ru2 and Ru3,
for which the high-energy bands may also contribute to the
overall b. It should also be noted that although we considered
only the lowest-energy band when deriving the experimental
static value for Ru4, a contribution from the more intense
second-lowest-energy band (which is predominantly MLCT in
character) to the b0,exp should also be taken into account.
Ru2’ and Ru3’ show similar linear optical behaviour in the
low-energy region, possessing almost identical lowest-energy
transitions. Moreover, the degree of MLCT character of these
bands, on the basis of calculations, is almost the same, and
thus it is not surprising that replacing an yne linkage with an
E-ene group on proceeding from Ru2’ to Ru3’ has only
a minor effect on the b value. In contrast, the calculated UV/
Vis spectrum of Ru4’ contains a significantly red-shifted
lowest-energy transition compared to the lowest-energy transi-
tions for Ru2’ and Ru3’, which may account for the large btot
calculated for model compound Ru4’.
6
4
and S16), the lowest unoccupied molecular orbitals (LUMOs) in
Ru2’ (and Ru3’) are delocalized on the N=N-1-C H -4-CꢀC-1-
6
4
C H -4-NO group, whereas in Ru4’ electron density is primarily
6
4
2
located at the N=N-1-C H -4-NO group, particularly at the azo-
6
4
2
!
phenylene unit. Overall, the lowest-energy LUMO (L) HOMO
H) transition involves a mixture of metal-to-ligand charge
(
transfer (MLCT) and intraligand CT (ILCT) character, together
with some contribution from ligand-to-ligand CT (LLCT).
The second- and third-lowest energy transitions, which lead
to a somewhat broad band in the experimental UV/Vis spec-
!
!
trum of Ru2 (and Ru3), originate from L+1 H and L Hꢂ2
electronic transitions. Hꢂ2 differs from H in that density leaks
into the central part of the bridge [that is, the N=N-1-C H -4-
6
4
CꢀC unit]. p-bonding within the azo and yne groups (or the E-
ene group in Ru3’) also contributes to Hꢂ2. L+1 is primarily
N=N (p*) and 1-C H -4-NO C-N (p) based. The second-lowest-
6
4
2
energy transition for the model complex Ru4’ is characterized
as promotion from Hꢂ2 (dominated by electron density
around the ClRuCꢀC-1-C H -4-CꢀC-1-C H -4 unit) to L.
Conclusion
6
4
6
4
Static first hyperpolarizability values, b , were calculated for
The new complexes Ru1–Ru4 and the previously reported
Ru5–Ru7 afford a suite of complexes that permit comment on
the effect of azo group introduction and location on electronic
tot
all model complexes (Table S8). Replacing an yne linkage with
an E-ene unit, on proceeding from Ru2’ to Ru3’, has little
II/III
effect on b , an outcome that is consistent with laboratory
and optical properties of metal alkynyl complexes. The Ru
tot
observations (Table 2 and Table S8). The position of the azo
linkage exerts a significant influence on the b value; the calcu-
lated findings clearly indicate that the complex in which the E-
azo linkage is distal from the metal exhibits a much larger non-
linear response than the proximal isomer (proceeding from
Ru3’ to Ru4’ leads to an approximately twofold increase in cal-
culated b), although this is contrary to experimental observa-
tions. Given the discrepancy with experimental data for the
latter, the performance of calculations in reproducing these
structure–property outcomes was examined as a function of
the method used (PBEPBE (the principal method used in these
studies), CAM-B3LYP, and LC-PBEPBE: see the Supporting Infor-
mation); all methods suggest that b increases significantly
when the E-azo linkage changes its position from A to B
oxidation potentials shift to higher potential on replacing yne
linkage with E-ene linkage (proceeding from Ru2 to Ru3),
changing the coordinated diphosphine ligand from dppm
(Ru1) to dppe (Ru2), and locating the azo group proximal to
the metal centre (proceeding from Ru4 to Ru1, Ru2, and Ru3);
the last-mentioned observation is consistent with previous
studies of ruthenium alkynyl complexes for which the electron-
withdrawing influence of groups is particularly pronounced
when they are appended to the proximal (to ruthenium)
[10a]
ring.
The wavelength of the UV/Vis lmax band can be tuned
by bridge variation, red-shifting on proceeding from com-
plexes bearing E-ene linkages (e.g. trans-[Ru(CꢀC-1-C H -4-(E)-
6
4
CH=CH-1-C H -4-(E)-CH=CH-1-C H -4-NO )Cl(dppe) ]
(Ru7):
6
4
6
4
2
2
468 nm) to the complex with a distal azo group (Ru4: 503 nm)
to complexes with a proximal azo group (Ru1–Ru3: 540–
550 nm). The frequency-dependent b values for the azo-con-
taining complexes from the present study are comparatively
(Figure 3), whereas replacing an yne group with an E-ene link-
age has a minor effect.
The simple two-level model (TLM) developed by Oudar and
[
22]
ꢂ30
Chemla
is widely used for extrapolating the experimental
large (ca. 3000ꢃ10 esu for Ru2 and Ru3) but strongly reso-
hyper-Rayleigh scattering (HRS) data (frequency-dependent
b value) to the static first hyperpolarizability, but this simple
model is not applicable when the scattered photon frequency
nance enhanced; experimental values for Ru2 and Ru3 are
similar to that of Ru7, but the two-level-corrected data for the
azo-linked compounds are much smaller than that of the ene-
linked Ru7 and, while being mindful of the problems associat-
ed with this approximation, we conclude that the ene-linked
alkynyl complexes are more efficient NLO materials. The for-
[21b,23]
approaches the resonance frequency of the CT transition
see above), and thus the simple TLM is not suitable for evalu-
(
ating the static hyperpolarizability of laboratory analogues of
model complexes Ru2’ and Ru3’ (in particular) as they exhibit
transitions very close to the second harmonic at 532 nm.
Hence, Equation (3) in the Supporting Information, which ex-
III
mally Ru complexes exhibit low-energy bands at about
ꢂ1
11000 cm for Ru1–Ru4, and the location of this LMCT band
can be tuned by molecular modification; it moves to lower fre-
quency on replacement of coligand dppm by dppe (proceed-
ing from Ru1 to Ru2), replacement of yne linkage with E-ene
linkage (proceeding from Ru2 to Ru3), or location of the azo
group further from the metal centre (proceeding from Ru2 to
[21b]
tends the simple two-level model into the resonant regime,
was used to calculate the static values. However, Equation (3)
assumes that only two electronic states contribute to the first
CT
hyperpolarizability, b . It should be used with great caution
&
ChemPlusChem 2016, 81, 1 – 9
www.chempluschem.org
6
ꢁ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Papers
II/III
Ru4). Thus, Ru1–Ru4 are in principle electrochemically—(Ru
)
Proc. SPIE-Int. Soc. Opt. Eng. 1994, 2143, 20–29; g) C. R. Moylan, R. J.
Twieg, V. Y. Lee, S. A. Swanson, K. M. Betterton, R. D. Miller, J. Am. Chem.
Soc. 1993, 115, 12599–12600; h) F. Wꢄrthner, F. Effenberger, R. Wort-
mann, P. Kraemer, Chem. Phys. 1993, 173, 305–314; i) E. M. Breitung, C.-
F. Shu, R. J. McMahon, J. Am. Chem. Soc. 2000, 122, 1154–1160; j) M.
Matsui, M. Suzuki, M. Hayashi, K. Funabiki, Y. Ishigure, Y. Doke, H. Shio-
zaki, Bull. Chem. Soc. Jpn. 2003, 76, 606–607; k) M. Barzoukas, M. Blan-
chard-Desce, D. Josse, J. M. Lehn, J. Zyss, Chem. Phys. 1989, 133, 323–
as well as photochemically—(EQZ azo) switchable optical ma-
terials.
Computational studies were undertaken on models Ru2’–
Ru4’ to rationalize this optical behaviour. Calculations agree
well with the experimental linear optical data, although we
found a mismatch between the calculated and experimental
lowest-energy bands for Ru4 as a consequence of employing
a fully planar structure for Ru4’ to represent the laboratory
compound Ru4. In Ru4’, the ethyl substituents used as an aid
to mitigate solubility problems in Ru4 are removed because
they may cause distortions in the bridge, and thus adversely
affect both empirical peak position and intensity of the lowest-
energy band. On the basis of calculations, the HOMO for all
three model systems is broadly similar and mainly located on
the ClRuCꢀC-1-C H -4- unit, while the LUMO for Ru4’ is more
329; l) B. K. Spraul, S. Suresh, T. Sassa, M. Angeles Herranz, L. Echegoyen,
T. Wada, D. Perahia, D. W. Smith, Tetrahedron Lett. 2004, 45, 3253–3256;
m) T.-G. Zhang, M. J. Therien, Polym. Prepr. Am. Chem. Soc. Div. Polym.
Chem. 2002, 43, 570–571; n) R. Twieg, V. Lee, R. D. Miller, C. R. Moylan,
Polym. Prepr. Am. Chem. Soc. Div. Polym. Chem. 1994, 35, 200–201; o) S.
Priyadarshy, M. J. Therien, D. N. Beratan, J. Am. Chem. Soc. 1996, 118,
1
504–1510; p) M. Moreno-MaÇas, R. Pleixats, R. Andreu, J. Garin, J.
Orduna, B. Villacampa, E. Levillain, M. Sallꢅ, J. Mater. Chem. 2001, 11,
74–380; q) S. R. Marder, B. G. Tiemann, A. Friedli, E. Yang, L.-T. Cheng,
3
Nonlinear Opt. 1995, 9, 213–221; r) A. K. Y. Jen, T.-A. Chen, V. P. Rao, Y.
Cai, Y.-J. Liu, L. R. Dalton, Adv. Nonlinear Opt. 1997, 4, 237–249; s) M.
Barzoukas, A. Fort, G. Klein, C. Serbutoviez, L. Oswald, J. F. Nicoud,
Chem. Phys. 1992, 164, 395–406.
3] T. Tsunekawa, K. Yamaguchi, J. Phys. Chem. 1992, 96, 10268.
4] V. Guerchais, H. Bozec, Molecular Organometallic Materials for Optics in
Top. Organomet. Chem., Vol. 28, Springer, Berlin, Heidelberg, 2010,
pp. 171–225.
6
4
localized compared to that for Ru2’ and Ru3’. The LUMOs for
Ru2’ and Ru3’ are delocalized on the N=N-1-C H -4-CꢀC-1-
[
[
6
4
C H -4-NO group, and in Ru4’ primarily located at the distal
6
4
2
N=N-1-C H -4-NO group. The lowest-energy bands associated
6
4
2
!
with L H excitation can be ascribed to a mixture of MLCT and
ILCT character with a LLCT contribution. Although the decrease
in b value on proceeding from Ru2–Ru3 to Ru4 is not repro-
duced computationally, due to a distorted non-coplanar struc-
ture for laboratory compound Ru4, as discussed earlier, both
the experimental and calculated data reveal that replacing an
yne linkage with an E-ene group on proceeding from Ru2 to
Ru3 (or Ru2’ to Ru3’) has only a minor effect on the b value
due to the similar linear optical properties they possess in the
low-energy region.
[5] a) J. P. L. Morrall, G. T. Dalton, M. G. Humphrey, M. Samoc, Adv. Organo-
met. Chem. 2007, 55, 61–136; b) M. G. Humphrey, T. Schwich, P. J. West,
M. P. Cifuentes, M. Samoc in Comprehensive Inorganic Chemistry II, Vol. 8
(
Eds.: J. Reedijk, K. Poeppelmeier), Elsevier, Oxford, 2013, pp. 781–835.
[
[
[
[
6] a) B. J. Coe, S. Houbrechts, I. Asselberghs, A. Persoons, Angew. Chem. Int.
Ed. 1999, 38, 366–369; Angew. Chem. 1999, 111, 377–380; b) B. J. Coe,
N. R. M. Curati, Comments Inorg. Chem. 2004, 25, 147–184; c) I. Assel-
berghs, K. Clays, A. Persoons, A. M. McDonagh, M. D. Ward, J. McClever-
ty, Chem. Phys. Lett. 2003, 368, 408–411.
7] a) T. N. Fondum, K. A. Green, M. D. Randles, M. P. Cifuentes, A. C. Willis,
A. Teshome, I. Asselberghs, K. Clays, M. G. Humphrey, J. Organomet.
Chem. 2008, 693, 1605–1613; b) A. M. McDonagh, N. T. Lucas, M. P. Ci-
fuentes, M. G. Humphrey, S. Houbrechts, A. Persoons, J. Organomet.
Chem. 2000, 605, 184–192.
8] a) I. R. Whittall, M. P. Cifuentes, M. G. Humphrey, B. Luther-Davies, M.
Samoc, S. Houbrechts, A. Persoons, G. A. Heath, D. Bogsanyi, Organome-
tallics 1997, 16, 2631–2637; b) I. R. Whittall, M. G. Humphrey, S. Hou-
brechts, A. Persoons, D. C. R. Hockless, Organometallics 1996, 15, 5738–
Acknowledgements
We thank the Australian Research Council (ARC), the National
Natural Science Foundation of China (51432006), and the Chi-
nese Government Ministry of Education and State Administration
of Foreign Experts Affairs (111 Project: B13025) for financial sup-
port. M.P.C. thanks the ARC for an Australian Research Fellow-
ship.
5745.
9] C. E. Powell, M. P. Cifuentes, A. M. McDonagh, S. Hurst, N. T. Lucas, C. D.
Delfs, R. Stranger, M. G. Humphrey, S. Houbrechts, I. Asselberghs, A. Per-
soons, D. C. R. Hockless, Inorg. Chim. Acta 2003, 352, 9–18.
[
10] a) S. Hurst, M. P. Cifuentes, J. P. L. Morrall, N. T. Lucas, I. R. Whittall, M. G.
Humphrey, I. Asselberghs, A. Persoons, M. Samoc, B. Luther-Davies, A. C.
Willis, Organometallics 2001, 20, 4664–4675; b) R. H. Naulty, A. M.
McDonagh, I. R. Whittall, M. P. Cifuentes, M. G. Humphrey, S. Hou-
brechts, J. Maes, A. Persoons, G. A. Heath, D. C. R. Hockless, J. Organo-
met. Chem. 1998, 563, 137–146; c) R. H. Naulty, M. P. Cifuentes, M. G.
Humphrey, S. Houbrechts, C. Boutton, A. Persoons, G. A. Heath, D. C. R.
Hockless, B. Luther-Davies, M. Samoc, J. Chem. Soc. Dalton Trans. 1997,
4167–4175.
[11] a) D. Touchard, P. Haquette, N. Pirio, L. Toupet, P. H. Dixneuf, Organome-
tallics 1993, 12, 3132–3139; b) L. Rigamonti, B. Babgi, M. P. Cifuentes,
R. L. Roberts, S. Petrie, R. Stranger, S. Righetto, A. Teshome, I. Assel-
berghs, K. Clays, M. G. Humphrey, Inorg. Chem. 2009, 48, 3562–3572.
[12] T.-T.-T. Nguyen, D. Tꢄrp, D. Wang, B. Nçlscher, F. Laquai, K. Mꢄllen, J. Am.
Chem. Soc. 2011, 133, 11194–11202.
Keywords: alkene ligands · alkyne ligands · electrochemistry ·
nonlinear optics · transition metals
[
1] a) P. N. Prasad, D. J. Williams, Introduction to Nonlinear Optical Effects in
Molecules and Polymers, Wiley, New York, 1991; b) J. Zyss, Molecular
Nonlinear Optics: Materials, Physics and Devices, Academic Press, Boston,
1
994; c) D. M. Roundhill, J. P. Fackler Jr., Optoelectronic Properties of Inor-
ganic Compounds, Plenum, New York, 1999.
[
2] a) V. P. Rao, A. K. Y. Jen, K. Y. Wong, K. J. Drost, J. Chem. Soc. Chem.
Commun. 1993, 1118–1120; b) L. Qiu, Y. Shen, J. Hao, J. Zhai, F. Zu, T.
Zhang, Y. Zhao, K. Clays, A. Persoons, J. Mater. Sci. 2004, 39, 2335–
[13] K. Sonogashira, Y. Tohda, N. Hagihara, Tetrahedron Lett. 1975, 16, 4467–
4470.
2
340; c) J.-M. Raimundo, P. Blanchard, N. Gallego-Planas, N. Mercier, I.
Ledoux-Rak, R. Hierle, J. Roncali, J. Org. Chem. 2002, 67, 205–218;
d) A. J. Moore, A. Chesney, M. R. Bryce, A. S. Batsanov, J. F. Kelly, J. A. K.
Howard, I. F. Perepichka, D. F. Perepichka, G. Meshulam, G. Berkovic, Z.
Kotler, R. Mazor, V. Khodorkovsky, Eur. J. Org. Chem. 2001, 2671–2687;
e) S. R. Marder, J. W. Perry, B. G. Tiemann, C. B. Gorman, S. Gilmour, S. L.
Biddle, G. Bourhill, J. Am. Chem. Soc. 1993, 115, 2524–2526; f) M. Blan-
chard-Desce, V. Bloy, J.-M. Lehn, C. Runser, M. Barzoukas, A. Fort, J. Zyss,
[14] A. Yassin, P. Jimenez, B. Lestriez, P. Moreau, P. Leriche, J. Roncali, P. Blan-
chard, H. Terrisse, D. Guyomard, J. Gaubicher, Chem. Mater. 2015, 27,
4057–4065.
[15] A. Zarei, A. R. Hajipour, L. Khazdooz, Synthesis 2009, 941–944.
[16] S. Zarwell, K. Rueck-Braun, Tetrahedron Lett. 2008, 49, 4020–4025.
[17] N. G. Anthony, D. Breen, J. Clarke, G. Donoghue, A. J. Drummond, E. M.
Ellis, C. G. Gemmell, J.-J. Helesbeux, I. S. Hunter, A. I. Khalaf, S. P. Mackay,
ChemPlusChem 2016, 81, 1 – 9
www.chempluschem.org
7
ꢁ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
These are not the final page numbers! ÞÞ

Full Papers
J. A. Parkinson, C. J. Suckling, R. D. Waigh, J. Med. Chem. 2007, 50,
116–6125.
18] T. Fehrentz, C. A. Kuttruff, F. M. E. Huber, M. A. Kienzler, P. Mayer, D. Trau-
ner, ChemBioChem 2012, 13, 1746–1749.
[22] J. L. Oudar, D. S. Chemla, J. Chem. Phys. 1977, 66, 2664–2668.
[23] a) M. P. Robalo, A. P. S. Teixeira, M. H. Garcia, M. F. Minas da Piedade,
M. T. Duarte, A. R. Dias, J. Campo, W. Wenseleers, E. Goovaerts, Eur. J.
Inorg. Chem. 2006, 2175–2185; b) I. R. Whittall, M. P. Cifuentes, M. G.
Humphrey, B. Luther-Davies, M. Samoc, S. Houbrechts, A. Persoons,
G. A. Heath, D. C. R. Hockless, J. Organomet. Chem. 1997, 549, 127–137.
6
[
[
[
19] B. Chaudret, G. Commenges, R. Poilblanc, J. Chem. Soc. Dalton Trans.
1984, 1635–1639.
20] H. Zhao, P. V. Simpson, A. Barlow, G. J. Moxey, M. Morshedi, N. Roy, R.
Philip, C. Zhang, M. P. Cifuentes, M. G. Humphrey, Chem. Eur. J. 2015, 21,
1
1843–11854.
Manuscript received: April 27, 2016
Accepted Article published: May 23, 2016
Final Article published: && &&, 0000
[
21] a) A. M. Kelley, J. Opt. Soc. Am. B 2002, 19, 1890–1900; b) C. H. Wang, J.
Chem. Phys. 2000, 112, 1917–1924.
&
ChemPlusChem 2016, 81, 1 – 9
www.chempluschem.org
8
ꢁ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

FULL PAPERS
D. Wei, M. S. Kodikara, M. Morshedi,
G. J. Moxey, H. Wang, G. Wang,
C. Quintana, C. Zhang, R. Stranger,
M. P. Cifuentes, M. G. Humphrey*
&
& – &&
It takes alkynes: Dipolar organometallic
complexes with a ligated ruthenium
donor, nitro acceptor, and a p-delocaliz-
able bridge incorporating an azo group
exhibit large quadratic optical nonli-
nearities (see figure). The potentials of
their reversible oxidation processes and
the wavelengths of their absorption
maxima in both the resting and oxi-
dized forms are tuned by systematic
structural modification.
Syntheses and Optical Properties of
Azo-Functionalized Ruthenium Alkynyl
Complexes
ChemPlusChem 2016, 81, 1 – 9
www.chempluschem.org
9
ꢁ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
These are not the final page numbers! ÞÞ