A theoretical and experimental study on meridional-facial isomerization of tris(quinolin-8-olate)aluminum (Alq3)

Article abstract of DOI:10.1039/c4dt01121j

The rationale behind the stereospecific synthesis of a facial isomer of tris(quinolin-8-olate)aluminum (Alq₃) is studied by density functional theory (DFT) calculations, which predict the favourable influence of an H₃O⁺ ion on the distribution ratio between a meridional and a thermodynamically unstable facial isomer. This journal is the Partner Organisations 2014.

Full text of DOI:10.1039/c4dt01121j

Dalton
Transactions
View Article Online
View Journal | View Issue
COMMUNICATION
A theoretical and experimental study
on meridional–facial isomerization of
tris(quinolin-8-olate)aluminum (Alq3)†
Cite this: Dalton Trans., 2014, 43,
12824
Received 16th April 2014,
Accepted 30th June 2014
I. Iwakura,‡ H. Ebina, K. Komori-Orisaku§ and Y. Koide*
DOI: 10.1039/c4dt01121j
www.rsc.org/dalton
The rationale behind the stereospecific synthesis of a facial isomer emitting materials, fac-Alq3 has been largely neglected for a
of tris(quinolin-8-olate)aluminum (Alq3) is studied by density func- lack of rational synthetic methods.
tional theory (DFT) calculations, which predict the favourable
influence of an H3O+ ion on the distribution ratio between a can be isolated with an excellent yield by refluxing a water sus-
Previously, we have unexpectedly discovered that fac-Alq3
meridional and a thermodynamically unstable facial isomer.
pension of mineral boehmite [AlO(OH)] and 8-hydroxyquinoli-
nol (Hq).6 Boehmite is a naturally occurring mineral, in which
Al3+ is coordinated by five oxides and a hydroxide ion. Termi-
nal hydroxyl groups are readily replaced by various O- and
N-donor ligands to yield corresponding aluminum coordination
compounds, exemplified by the reaction between boehmite
and carboxylic acids to furnish the carboxylate-alumoxanes as
demonstrated by Barron et al.7 Thus, boehmite can be
employed as a quantitative and environmentally benign source
of Al3+ in the formation of complexes. Unlike common alumi-
num salts, such as nitrate, sulfate, isopropoxide, and ace-
tate,2b,d,8 we hypothesized that the complexation of boehmite
with Hq liberates only water-borne ions that may contribute to
the promotion of a thermodynamically unfavorable mer- to
fac-Alq3 isomerization reaction (eqn (1)).
Tris(quinolin-8-olate)aluminum (Alq3) has been referred to as
a prototypical electroluminescent material in small molecule-
based organic light-emitting diodes (OLEDs) due to its rela-
tively low operational voltage to coax its characteristic green
emission.1 Its technological importance has drawn intensive
research activities addressing synthesis,2 structural characteri-
zation,3 and physicochemical properties of Alq3. Structurally,4
three quinolin-8-olate (q−) ligands are coordinated to alumi-
num(III) ion in Alq3, existing in two isomeric forms: C3 sym-
metrical facial Alq3 ( fac-Alq3) and C1 symmetrical meridional
Alq3 (mer-Alq3). Each isomer exhibits quintessential opto-
electronic properties.4k For example, the fluorescent emission
spectrum of fac-Alq3 is blue shifted, and its quantum yield is
higher than that of mer-Alq3 in both solution and solid state.2a,c,4e,5
Different electron-transporting abilities are predicted for
two isomers.4a,b,h,j,k Thermodynamically more stable mer-Alq3
is naturally the dominant species in solution-phase synthe-
sis,2b,d,4e whereas fac-Alq3 can be obtained as a byproduct in
the sublimation purification of mer-Alq3, or by annealing
mer-Alq3 at around 400 °C.4e However, fac-Alq3 is notoriously
elusive, and it converts to mer-Alq3 at higher than −20 °C in
chloroform. Hence, little has been determined about the iso-
merization mechanism. Despite a keen interest in the blue-
AlOðOHÞ þ 3Hq ð! Hþ þ qꢀÞ ! Alq3 þ 2H2O
ð1Þ
Given that boehmite is hardly soluble in pure water,9 it is
surmised that heterogeneous reactions between boehmite and
Hq may be occurring with high efficiency and stereo selectivity
in a water medium. Analogous efficiency improvement was
detected in a stereoselective bromination reaction of stilbene
in water suspension.10
This communication addresses theoretical and experi-
mental studies on the isomerization of Alq3 observed in a reac-
tion between boehmite and Hq in a water medium. Because
authentic mer-Alq3 was completely inert even after prolonged
reflux in pure water,11 the roles of water and water-borne ions
(H2O, H+, H3O+, and OH−) were evaluated using the B3LYP
Department of Applied Chemistry, Graduate School of Engineering,
Kanagawa University, Yokohama, 221-8686, Japan
†Electronic supplementary information (ESI) available: Experimental details, level of DFT calculations.4l,12 An elegant study by Amati and
supplementary calculation results and graphics, solid-state 13C NMR spectra. See
Lelj has revealed that the fastest isomerization proceeds via
DOI: 10.1039/c4dt01121j
the Al–N bond cleavage and formation of five-coordinate inter-
‡Current address: Department of Chemistry, Graduate School of Engineering,
mediate.13 Considering their results, we performed a series of
Kanagawa University, Yokohama, 221-8686, Japan.
density functional theory (DFT) calculations to find that mer-
to fac-Alq3 isomerization is a two-state transition process with
§Current address: Department of Chemistry, School of Science, The University
of Tokyo, 7-3-1 Hongo, Bunkyo-ku, Tokyo, 113-0033, Japan.
12824 | Dalton Trans., 2014, 43, 12824–12827
This journal is © The Royal Society of Chemistry 2014

View Article Online
Dalton Transactions
Communication
a high-energy intermediate species, and the pathway trans- higher than mer-Alq3, which is comparable to the values esti-
forms to a single-state transition process upon H3O+ complexa- mated by other studies.4j,k,13 Subsequently, solvent effects were
tion. The role of H3O+ was experimentally verified by refluxing included by performing the single-point polarizable continuum
reactants in acidic water that drastically accelerated the reaction. model (PCM) calculations on each model in the reaction coordi-
The integrity of calculations were assessed using different nate to find a relatively substantial drop of the potential energy
basis sets (LANL2DZ, 6-31G*, 6-31+G*) and compared the for the facial isomer (2.1 kcal mol−1).17
resulting geometries with corresponding X-ray crystallographic
The association models are created by placing H+, H3O+,
data.14 Although equally satisfactory values are obtained for OH−, or H2O at the sterically least-congested sites around the
both 6-31G* and 6-31+G* level calculations (Table S1 and Al3+ centre of each isomer. Calculations reached convergence,
Fig. S1 in ESI†), we employed the one with diffuse functions except for the OH− association model (see Fig. S2 in ESI† for
(6-31+G*) throughout this research in the expectation of better optimized coordination geometries). The smallest potential
handling the anticipated intermediate species with elongated energy difference between mer- and fac-Alq3 (3.3 kcal mol−1) is
bonds.
obtained for H3O+ association models (Alq3/H3O+), in which
Fig. 1 shows the calculated reaction coordinate for isomeri- two out of three hydrogen atoms are bonded to the oxygen
zation of Alq3. The initial guess for the five-coordinate tran- atoms in mer-Alq3/H3O+, and the third hydrogen atom is point-
sition state (TS) geometry was prepared by breaking the Al–N ing away from Al3+. For symmetrically coordinated fac-Alq3/
bond in a ligand L1 (shown in green) and twisted 90° about H3O+, H3O+ occupies the centre of the triangular face defined
the Al–O bond. A converged calculation shows that the rate- by the three ligand oxygen atoms. This geometry can accom-
determining transition state geometry (TS1) has its imaginary modate a relatively large interstitial space around L1, which is
frequency vectors (yellow arrows) pointing the direction of responsible for isomerization through rotation about the Al–O
ligand rotation.15 Intrinsic reaction coordinate (IRC) calcu- bond.
lations were performed by perturbing the coordinates of TS1
The search for the reaction coordinate of Alq3/H3O+ started
to create a stable intermediate (INT) along the reaction coordi- from the preparation of the five-coordinate TS geometry of
nate going forward (mer- to fac-Alq3). Further, TS geometry mer-Alq3/H3O+ by cleaving the Al–N bond of L1 and twisted
optimization between the INT and fac-Alq3 discovered that one 90°; therefore, the pyridine moiety moves away from the Al3+
of the imaginary frequency vectors of TS2 pointed the direction center, and H3O+ was fixed between the oxygen atoms. A suc-
of ligand rotation to furnish the fac-Alq3 configuration,16 the cessfully converged geometry (TS3) shows that three equatorial
potential energy of which was estimated to be 5.4 kcal mol−1 oxygen atoms are slightly puckered to bring them closer to
each other (Fig. 2). The imaginary frequency vectors on L1 are
Fig. 1 Reaction coordinate for isomerization of Alq3. The potential
energy values after applying single-point PCM calculations are shown in
parentheses. Yellow arrows on the calculated molecular structures
indicate the direction of imaginary frequency vectors. Colour scheme:
Al = purple, O = red, N = blue, and C = grey. Hydrogen atoms are
omitted for clarity.
Fig. 2 Reaction coordinate for isomerization of Alq3/H3O+. The poten-
tial energy values after applying single-point PCM calculations are
shown in parentheses. Yellow arrows on the calculated molecular struc-
tures indicate the direction of imaginary frequency vectors. Color
scheme: Al = purple, O = red, N = blue, and C = grey. Hydrogen atoms
on the ligands are omitted for clarity.
This journal is © The Royal Society of Chemistry 2014
Dalton Trans., 2014, 43, 12824–12827 | 12825

View Article Online
Communication
Dalton Transactions
3 (a) M. Utz, C. Chen, M. Morton and F. Papadimitrakopoulos,
J. Am. Chem. Soc., 2003, 125, 1371–1375; (b) M. Utz,
M. Nandagopal, M. Mathai and F. Papadimitrakopoulos,
Appl. Phys. Lett., 2003, 83, 4023–4025; (c) M. Coelle,
S. Forero-Lenger, J. Gmeiner and W. Bruetting, Phys. Chem.
Chem. Phys., 2003, 5, 2958–2963; (d) M. Brinkmann,
G. Gadret, M. Muccini, C. Taliani, N. Masciocchi and
A. Sironi, J. Am. Chem. Soc., 2000, 122, 5147–5157.
Table 1 Comparison of the amount of Alq3 isomers dissolved in pure
and acidic water (2.0 × 10−2 M HCl) at 25 °C and 100 °C, respectively.
A sample (500 mg) was stirred in 60 mL of water and collected after
60 min (25 °C) or 10 min (100 °C). The amounts of the isomer dissolved
were calculated from a dry weight of the retrieved solid
mer-Alq3 (mg)
fac-Alq3 (mg)
H2O (25/100 °C)
75.5/110
110/130
20.0/73.5
18.0/70.5
2.0 × 10−2 M HCl (25/100 °C)
4 (a) A. Fuchs, T. Steinbrecher, M. S. Mommer, Y. Nagata,
M. Elstner and C. Lennartz, Phys. Chem. Chem. Phys., 2012,
14, 4259–4270; (b) J. J. Kwiatkowski, J. Nelson, H. Li,
J. L. Bredas, W. Wenzel and C. Lennartz, Phys. Chem. Chem.
Phys., 2008, 10, 1852–1858; (c) J. E. Knox, M. D. Halls,
H. P. Hratchian and H. B. Schlegel, Phys. Chem. Chem.
Phys., 2006, 8, 1371–1377; (d) M. Coelle and C. Gaerditz,
Appl. Phys. Lett., 2004, 84, 3160–3162; (e) M. Muccini,
M. A. Loi, K. Kenevey, R. Zamboni, N. Masciocchi and
A. Sironi, Adv. Mater., 2004, 16, 861–864; (f) M. Amati and
F. Lelj, J. Phys. Chem. A, 2003, 107, 2560–2569; (g) M. Amati
and F. Lelj, Chem. Phys. Lett., 2002, 358, 144–150;
(h) G. G. Malliaras, Y. Shen, D. H. Dunlap, H. Murata and
Z. H. Kafafi, Appl. Phys. Lett., 2001, 79, 2582–2584;
(i) M. Sugimoto, S. Sakaki, K. Sakanoue and M. D. Newton,
J. Appl. Phys., 2001, 90, 6092–6097; ( j) R. L. Martin,
J. D. Kress, I. H. Campbell and D. L. Smith, Phys. Rev. B:
Condens. Matter Mater. Phys., 2000, 61, 15804–15811;
(k) A. Curioni, M. Boero and W. Andreoni, Chem. Phys. Lett.,
1998, 294, 263–271; (l) F. Papadimitrakopoulos,
X. M. Zhang, D. L. Thomsen III and K. A. Higginson, Chem.
Mater., 1996, 8, 1363–1365.
5 (a) M. M. Levichkova, J. J. Assa, H. Frob and K. Leo, Appl.
Phys. Lett., 2006, 88, 201912; (b) H. Kaji, Y. Kusaka,
G. Onoyama and F. Horii, Jpn. J. Appl. Phys., Part 1, 2005,
44, 3706–3711; (c) F. Li, Z. Chen, W. Wei, H. Cao, Q. Gong,
F. Teng, L. Qian and Y. Wang, J. Phys. D: Appl. Phys., 2004,
37, 1613–1616.
6 R. Katakura and Y. Koide, Inorg. Chem., 2006, 45,
5730–5732.
7 (a) Y. Koide and A. R. Barron, Organometallics, 1995, 14,
4026–4029; (b) C. C. Landry, N. Pappe, M. R. Mason,
A. W. Apblett, A. N. Tyler, A. N. MacInnes and A. R. Barron,
J. Mater. Chem., 1995, 5, 331–341.
in accordance with the ligand rotation toward the facial con-
figuration. IRC calculations on the TS3 found two local
minima,18 one each going forward and backward. The former
configuration coincides with that of fac-Alq3/H3O+, whereas the
latter is practically identical to the starting mer-Alq3/H3O+.
Although the single-point PCM calculations predicted lower
potential energies for TS3 and fac-Alq3/H3O+, the forward reac-
tion remained thermodynamically uphill. Muccini et al. attri-
butes the driving force of this uphill reaction to the higher
lattice free energy of fac-Alq3,4e which will precipitate out as a
crystalline solid. Indeed, fac-Alq3 is confirmed to be less
soluble in water than mer-Alq3, and disparities grow larger in
acidic water (Table 1).
In order to experimentally verify the role of H3O+, boehmite
and Hq are reacted in a dilute HCl solution (2.0 × 10−2 M). A
bright yellow precipitate formed within 1 h, and the color then
faded to a yellowish cream in 3 h, which was substantially
faster than in pure water, in which it took at least 24 h. The
13C NMR spectrum of the solid product confirmed the for-
mation of fac-Alq3 (Fig. S3†).6 More significantly, even though
a suspension of mer-Alq3 in pure water was never converted to
the facial isomer that occurred in acidic water after refluxing
for 48 h.
In conclusion, we have conducted theoretical studies on the
isomerization mechanisms of Alq3 in an aqueous medium and
showed that there are two transition states in addition to a
high-energy intermediate species along the reaction coordi-
nate. The studies also found that H3O+ association modifies
the two-state TS to a single TS process. Theoretical predictions
were experimentally confirmed by performing the reaction
under acidic conditions, which lead to a great increase in the
reaction rate.
8 X. Ma, G. K. Lim, K. D. M. Harris, D. C. Apperley,
P. N. Horton, M. B. Hursthouse and S. L. James, Cryst.
Growth Des., 2012, 12, 5869–5872.
9 P. Benezeth, D. A. Palmer and D. J. Wesolowski, Geochim.
Cosmochim. Acta, 2001, 65, 2097–2111.
Notes and references
1 C. W. Tang and S. A. VanSlyke, Appl. Phys. Lett., 1987, 51,
913–915.
2 (a) M. Colle, J. Gmeiner, W. Milius, H. Hillebrecht and
10 K. Tanaka, R. Shiraishi and F. Toda, J. Chem. Soc., Perkin
Trans. 1, 1999, 3069–3070.
W. Brutting, Adv. Funct. Mater., 2003, 13, 108–112; (b) H. Li, 11 Chirality inversions, which happens significantly faster
F. Zhang, Y. Wang and D. Zheng, Mater. Sci. Eng., B, 2003, than isomerization, is not considered in our study.
B100, 40–46; (c) M. Braun, J. Gmeiner, M. Tzolov, 12 K. A. Higginson, X.-M. Zhang and F. Papadimitrakopoulos,
M. Coelle, F. D. Meyer, W. Milius, H. Hillebrecht, Chem. Mater., 1998, 10, 1017–1020.
O. Wendland, J. U. von Schutz and W. Brutting, J. Chem. 13 M. Amati and F. Lelj, Chem. Phys. Lett., 2002, 363, 451–457.
Phys., 2001, 114, 9625–9632; (d) A. K. Saxena, Synth. React. 14 All calculations were carried out using the GAUSSIAN 09
Inorg. Met.-Org. Chem., 1999, 29, 1747–1767.
program package. M. J. Frisch, et al., Gaussian 09 (Revision
12826 | Dalton Trans., 2014, 43, 12824–12827
This journal is © The Royal Society of Chemistry 2014

View Article Online
Dalton Transactions
Communication
A.02), Gaussian, Inc., Wallingford CT, 2009; M. J. Frisch, 16 Note that there are two sets of imaginary frequency
et al., Gaussian 03, (Revision D.02), Gaussian, Inc., Walling-
ford, CT, 2004. Parameters are set to the default value.
15 Imaginary rotation vectors have relatively low intensities for
vectors for TS2 that substantially diminishes the reliability.
Lines in Fig. 1 are shown in lighter color to admonish this
point.
all TS structures presumably because the ligands are 17 B. Mennucci, J. Tomasi, R. Cammi, J. R. Cheeseman,
directly coordinated to the metal center. Thus, the result-
ing structures may not necessarily reflect those of TSs. The
M. J. Frisch, F. J. Devlin, S. Gabriel and P. J. Stephens,
J. Phys. Chem. A, 2002, 106, 6102–6113.
calculated activation energies are considered qualitative 18 (a) K. Fukui, J. Phys. Chem., 1970, 74, 4161–4163;
rather than quantitative.
(b) K. Fukui, Acc. Chem. Res., 1981, 14, 363–368.
This journal is © The Royal Society of Chemistry 2014
Dalton Trans., 2014, 43, 12824–12827 | 12827