The influence of intramolecular hydrogen bonding on the order parameter and photostability properties of dichroic azo dyes in a nematic liquid crystal host

Article abstract of DOI:10.1039/a901936g

It has been demonstrated that intramolecular hydrogen bonding in two classes of dichroic azo dyes has a generally detrimental effect on order parameter in a nematic liquid crystal host. Derivatives of 4-(N,N- diethylamino)-4'-nitroazobenzene with hydroxy groups intramolecularly hydrogen bonded to the azo group show lower order parameters in E7 nematic liquid crystal than the parent dye, but higher order parameters than corresponding methoxy dyes. 1-Arylazo-2-naphthols with a hydrogen bonding carboxamide group in position 3 exist exclusively in the hydrazone form and show even lower order parameters. Intramolecular hydrogen bonding also has a deleterious effect on photochemical stability.

Full text of DOI:10.1039/a901936g

View Article Online / Journal Homepage / Table of Contents for this issue
J O U R N A L O F
The in¯uence of intramolecular hydrogen bonding on the order
parameter and photostability properties of dichroic azo dyes
in a nematic liquid crystal host
John Grif®ths* and Kai-Chia Feng
C H E M I S T R Y
Department of Colour Chemistry, University of Leeds, Leeds, UK LS2 9JT. E-mail:
J.Grif®ths@leeds.ac.uk
Received 11th March 1999, Accepted 9th June 1999
It has been demonstrated that intramolecular hydrogen bonding in two classes of dichroic azo dyes has a
generally detrimental effect on order parameter in a nematic liquid crystal host. Derivatives of 4-(N,N-
diethylamino)-4'-nitroazobenzene with hydroxy groups intramolecularly hydrogen bonded to the azo group
show lower order parameters in E7 nematic liquid crystal than the parent dye, but higher order parameters
than corresponding methoxy dyes. 1-Arylazo-2-naphthols with a hydrogen bonding carboxamide group in
position 3 exist exclusively in the hydrazone form and show even lower order parameters. Intramolecular
hydrogen bonding also has a deleterious effect on photochemical stability.
regarded as true azo structures. Thus the possibilities for
conversion to hydrazone tautomeric forms are restricted to
those derivatives with hydroxy groups ortho or para to the azo
group, and even in these cases it is known that such
tautomerism is generally negligible. The methoxy-substituted
derivatives 2a,b were prepared as examples of dyes with no
hydrogen bonding and with no ability to undergo tautomerism.
The hydroxy dyes 3a,b then provided analogues with a single
intramolecular hydrogen bond involving the azo group, and
Introduction
Dyed phase change liquid crystal displays are dependent on
dichroic dyes with very speci®c characteristics, notably high
order parameter, high solubility in the host, and good
photostability.1 Consequently there are relatively few such
compounds that ®nd commercial use, and these are largely
restricted to the anthraquinones and, to a lesser extent, the azo
dyes.1,2 In general the azo dyes would be preferred to the
anthraquinones because of their better order parameter and
solubility characteristics, but the superior photostability proper-
ties of the anthraquinones have been the major factor leading to
their dominance. In the anthraquinone dyes, photostability is
in¯uenced markedly by intramolecular hydrogen bonding, and
thismay beattributed toprotectionofthe chromophorebyrapid
deactivation of the excited state by proton transfer.3
The effects of intramolecular hydrogen bonding on the
properties of azo dyes in liquid crystal media have not been
investigated. In addition to a possible in¯uence on photo-
stability, there may also be a signi®cant effect on order
parameter, although it is not clear if this would be bene®cial or
detrimental. On the one hand, a strongly hydrogen bonded dye
molecule would be more rigid and planar (as evidenced by their
generally higher extinction coef®cients and narrower absorp-
tion bands than non-bonded analogues), and thus could orient
more strongly with the mesogenic groups of the host.
Alternatively, intramolecular hydrogen bonding may block
sites of high and low electron density in the dye molecule which
would otherwise be available for intermolecular interactions
with the mesogens, or could reduce the overall dipole moment
of the dye molecule, both effects having the potential to lower
the observed order parameter.
In order to answer these questions, several azo and azo±
hydrazone tautomeric dyes have been synthesised containing
representative types of intramolecular hydrogen bonding, and
their photostability and order parameter properties in a nematic
liquid crystal host measured. The results have been compared
with those for analogous dyes lacking such hydrogen bonding.
Results and discussion
Dye synthesis and characterisation
(a) Derivatives of 4-(N,N-diethylamino)azobenzene. Deriva-
tives of 4-(N,N-diethylamino)-4'-nitroazobenzene 1 may be
J. Mater. Chem., 1999, 9, 2333±2338
2333

View Article Online
could be compared directly with 2. Dye 4 was also prepared
as an example of a dye with two hydrogen bonds to the azo
group. The dyes were synthesised by diazotisation of 4-nitro-,
2-methoxy-4-nitro- and 2-hydroxy-4-nitro-aniline followed
by coupling to the appropriate N,N-diethylaniline, and were
characterised by microanalysis and mass spectrometry.
(b) 1-Arylazo derivatives of 2-hydroxy-3-naphthamides. 1-
Arylazo-2-naphthols 5a are typical examples of azo dyes, which
in most cases exist predominantly in their hydrazone
tautomeric forms 5b.4 This has important implications for
the dipole moment and transition dipole moment directions of
the chromophore, both properties having a major in¯uence on
order parameters. Such dyes already contain one strong
intramolecular hydrogen bond, as shown in 5b, and by
introducing a primary or secondary carboxamide group in
the 3-position of the naphthalene ring one can introduce a
second hydrogen bond involving the 2-carbonyl group, as
shown in 6.5
Scheme 1
The dyes 7 and 8 were prepared by coupling diazotised p-
hexylaniline or p-hexyloxyaniline respectively to the relevant 2-
hydroxy-3-naphthamide in aqueous solution at pH ca. 9. The
amide couplers were prepared as summarised in Scheme 1.
Thus 2-hydroxy-3-naphthoic acid was reacted with phosphorus
oxychloride in methanol to give the methyl ester 9, and this was
then heated under re¯ux in neat amine to give 10. The dyes
were puri®ed by recrystallisation from ethanol and/or column
chromatography until they showed a single spot on TLC
analysis, and were characterised by microanalysis.
Visible absorption spectra and order parameters of the azo dyes
in E7 liquid crystal host
The host chosen for this study was E7, a commercially available
nematic eutectic mixture of 4-cyano-4@-n-pentyl-p-terphenyl, 4-
alkyl-4'-cyanobiphenyls and 4-alkoxy-4'-cyanobiphenyls. Spec-
tral data and order parameters in E7 for all the dyes studied are
summarised in Table 1. The aminoazobenzene dyes 1±4 show a
single absorption band in the visible region, and demonstrate
the usual positive solvatochromism for this type of dye. Thus
absorption maxima in acetone occur at longer wavelengths
than in the less polar solvent toluene. In the protic solvent
ethanol, however, there are anomalies, and those dyes with
intramolecular hydrogen bonding, i.e. 2b, 3b and 4, absorb at
shorter wavelengths in this solvent than in toluene, whereas the
non-hydrogen bonded methoxy counterparts show the oppo-
site effect. If one considers the spectra of dyes 2 and 3 measured
in toluene, in which speci®c solvent interactions will be
minimal, then it is apparent that the hydroxy groups in 2b
and 3b exert a somewhat greater bathochromic effect than
corresponding methoxy groups in 2a and 3a. This may be
attributed to the hydrogen bonding in the former dyes, which
will increase chromophore planarity and orbital overlap.
However the effect is not additive and in the case of dye 4
with two hydrogen bonds, the absorption maximum falls
between those of 3a and 3b.
As the additional hydrogen bond in 6 is known to improve
photochemical stability and molecular packing properties in
the azoic dyes and related pigments,6 it was of interest to
prepare representative dyes of this type in order to observe their
behaviour in a liquid crystal host. One of the main problems
with this class of dye is their low solubility in organic solvents,
and this was overcome by using N-methoxypropyl and N-
hydroxyethyl groups as side chains attached to amide nitrogen.
Long chain substituents were introduced into the para position
of the arylazo residue in order to extend the long axis of the dye
molecule and improve alignment with the host mesogens. These
groups had the added advantage of increasing further the
solubility of the dyes in the liquid crystal host. Thus two series
of dyes were prepared, namely the 4-n-hexyl series 7, and 4-n-
hexyloxy series 8.
In E7 liquid crystal there is
a pronounced positive
solvatochromism for all the dyes and the bathochromic
effect of hydrogen bonding is signi®cantly enhanced, so
much so that dye 4 now absorbs at the longest wavelength
(lmax 545 nm). Absorption maxima calculated by the PPP-MO
method7 reproduce most of the general lmax trends observed in
toluene, although absolute agreement between theory and
experiment is only moderate.
The order parameters of the dyes 2±4 in E7 were
disappointing, in that they were all slightly lower than that
of the parent dye 1. Thus intramolecular hydrogen bonds per se
and the concomitant increase in molecular coplanarity do not
confer enhanced alignment between the dye chromophore and
the host mesogens. However, the methoxy-substituted dyes
2a,b show order parameters ca. 0.05±0.06 units lower than their
2334
J. Mater. Chem., 1999, 9, 2333±2338

View Article Online
Table 1 Visible absorption spectra and order parameters in E7 of dyes 1±4, 7 and 8
Dye
lmax(ethanol)/nm lmax(acetone)/nm lmax(toluene)/nm lmax(calc.)/nm lmax(E7)/nm Predominant tautomer in E7 Order parameter S
1
482
497
470
488
512
488
474
502
501
464
519
518
486
502
500
491
516
512
415
499
499
415
514
514
474
484
490
482
514
500
418
522
524
463
513
513
444
444
436
466
504
507
492
491
491
438
501
501
504
515
512
509
532
545
476
530
530
418
524
522
azo
azo
azo
azo
azo
azo
hydrazo
hydrazo
hydrazo
azo
0.62
0.52
0.58
0.54
0.59
0.56
0.46
0.20
0.13
0.32
0.27
0.20
2a
3a
2b
3b
4
7a
7b
7c
8a
8b
8c
hydrazo
hydrazo
hydroxy-substituted counterparts 3a,b, re¯ecting the greater
planarity and reduced rotational mobility of the latter dyes.
Thus it can be concluded that if it is necessary to introduce
electron donor groups into an azo dye chromophore in order to
produce a speci®c lmax shift, then hydroxy groups are
preferable to methoxy groups because of the higher order
parameters and greater bathochromic shifts they induce. The
order parameter for 4 lies between that for the hydroxy dyes 3a
and 3b, again demonstrating the minimal in¯uence of
intramolecular hydrogen bonding on dye alignment character-
istics in comparison with dye 1.
between the two, the poorer will be the degree of alignment. In
addition, if the transition moment direction for the visible band
of the dye chromophore differs signi®cantly from the molecular
alignment axis, the measured order parameter will be lowered
further. As the permanent p-dipole moment (mp) and the
transition moment (M) directions can be calculated readily by
PPP-MO theory,7,9 it is possible to examine if either of these
factors contributes to the low order parameters for the dyes.
Fig. 1a and 1b show calculated mp and M directions for the
two parent dyes 7a and 8a respectively. For the purposes of the
calculations a hydrazone structure was assumed in the former
case, and an azo structure in the latter (see Table 1). For both
systems the same approximate molecular alignment axis
(MAA) was assigned by considering the longest axis through
the molecule that encompassed both the dye p-chromophore
and its terminal long alkyl chain (see Fig. 1). In both the case of
7a (hydrazone) and 8a (azo), the deviation of the calculated mp
and M directions from the MAA was reasonably large, thus
accounting for the moderate order parameters for these dyes.
Interestingly, the calculations revealed that the 3-carboxamide
groups in dyes 7b,c and 8b,c had only a small effect on the M
direction, but caused a larger deviation of mp from the MAA
(Fig. 1c and 1d). This explains in part why the amido-
substituted dyes have a lower order parameter than the
unsubstituted dyes. However, another factor must be the
greater indeterminacy of the MAA in the substituted dyes. As
can be seen in Fig. 2, the methoxypropyl side chain on the
nitrogen of the carboxamide group of dye 8c provides an
elongation of the molecule in a direction almost perpendicular
to the previously assumed MAA. Thus the true MAA is less
well de®ned and the dye will be poorly aligned with the host
mesogens. This explains why the dyes 7c and 8c have even
lower order parameters than corresponding dyes 7b and 8b, as
they have a longer chain attached to the amide nitrogen.
Unfortunately, it was essential to include these side chains in
order to ensure adequate solubility of the dyes in the host.
The 1-arylazo-2-naphthol dyes 7 and 8 are more complex
systems than 1±4, as they can show signi®cant azo±hydrazone
tautomerism. Thus their visible absorption spectra sometimes
show two absorption bands attributable to the two tautomers,
and the ratio of the tautomers at equilibrium depends both on
substituents in the dye and the nature of the solvent. In the case
of parent dye 7a, roughly equal amounts of the two tautomers
could be detected in toluene, the azo form dominating slightly
[lmax (azo) ca. 418 nm, lmax (hydrazone) ca. 476 nm, relative
absorbance values ca. 1.2 : 1]. In ethanol and E7 the hydrazone
form was much more predominant, demonstrating the known
stabilising effect of strong solvent±solute interactions on
hydrazone tautomers.8 The p-hexyloxyphenylazo dye 8a
showed a greater propensity to exist as the azo tautomer,
which was consistent with the known effects of electron
donating groups in the arylazo residue,8 and in toluene the azo
tautomer was the dominant species [lmax (azo) 463 nm, lmax
(hydrazone) ca. 510 nm, relative absorbance values ca. 1.0 : 0.4].
Even in ethanol and E7 the azo tautomer was predominant.
However, the dyes 7b,c and 8b,c which have a 3-carboxamide
substituent hydrogen bonded to the hydrazone carbonyl group,
showed exclusive hydrazone formation in all solvents. Thus
they showed a single relatively narrow band in the 520±530 nm
range in E7. In this medium the 3-carboxamide substituent had
a pronounced effect on the colour of the dyes, and whereas 7a
and 8a were orange in E7, dyes 7b,c and 8b,c were red±purple.
X-Ray crystallographic studies have also shown that 3-
carboxamide substituents contribute to a bifurcated hydrogen
bond and favour the hydrazone tautomeric form.5
Although 7b,c and 8b,c exist as discrete tautomers in E7, and
have planar strongly hydrogen bonded structures, their order
parameters were very low, and signi®cantly less than for the
parent dyes 7a and 8a. The reason for this was ®rst sought by
application of molecular orbital theory. A dye chromophore
will try to align itself with respect to the host mesogenic groups
so that its permanent dipole moment is parallel to that of the
mesogen, and its spatial overlap with the mesogen is maximal.
The E7 host used in this study comprises a mixture of 4-n-
pentyl-4@-cyano-p-terphenyl, 4-alkyl-4'-cyanobiphenyls and 4-
alkoxy-4'-cyanobiphenyls, all such mesogens having a strong
permanent dipole moment as well as a rigid rod-like shape. A
compromise situation will hold if the dipole moment direction
and the long axis of the dye differ, and the greater the deviation
Photostability properties in E7
It is well known that intramolecular hydrogen bonding in a dye
molecule will often lead to a marked improvement in
photochemical stability. For example, in the anthraquinone
dyes, hydroxy, primary and secondary amino groups in the 1-
position of the anthraquinone ring will provide a strong
hydrogen bond to the adjacent carbonyl group, and thus they
have high light fastness.2 For this reason, all commercial
anthraquinone dyes have such a substitution pattern. It is
believed that the improvement in stability is due at least in part
to facile proton transfer in the photoexcited state, leading to a
reduction in the excited state lifetime. As photostability is one
of the most important features of dichroic dyes used in liquid
crystal displays, it was of interest to see if the hydrogen bonding
present in dyes 3, 4, 7 and 8 could also confer improved
resistance to photochemical fading.
An additional advantage of the intramolecular hydrogen
J. Mater. Chem., 1999, 9, 2333±2338
2335

View Article Online
Fig. 1 Assumed molecular alignment axes (MAA) and MO calculated p-dipole moment (mp) and transition moment (M) directions for (a) 7a; (b) 8a;
(c) 7b,c; (d) 8b,c.
bond in azo and hydrazone dyes is that photochemical EAZ
isomerisation is not a complicating factor, as the rate of
thermal ZAE reversion is so high that it is not possible to
detect any of the Z isomer under steady state illumination
conditions at ambient temperatures.10±12
slow fading rates. The t1/2 values and initial ®rst order rate
constants are summarised in Table 2.
Surprisingly it was found that intramolecular hydrogen
bonding had a signi®cant deleterious effect on photochemical
stability. This can be seen most dramatically with the doubly
hydrogen bonded azo dye 4, which has a t1/2 of 125 hours, in
comparison with the parent dye 1, which is more than 30 times
more stable with a t1/2 of 3900 hours. To place these stabilities
in context, it should be noted that 1,4-bis(methylamino)an-
thraquinone, a blue dye belonging to a general group of
anthraquinone dyes used in LC displays, was found to have a
t1/2 of ca. 5000 hours under identical exposure conditions. Thus
dyes 1 and 7a may be considered to approach the minimum
technical requirements for LC display applications, whereas all
the other derivatives have unacceptably low photostabilities.
The source of the low photostability in the hydrogen bonded
derivatives may be attributed in part to azo±hydrazone
Glass cells (gap 25 mm) containing aligned solutions of the
dyes in E7 were exposed to the radiation from a phosphor-
coated 500 W mercury±tungsten ®lament lamp, designed to
simulate accelerated daylight fading. The cells were unsealed,
so ensuring aerobic conditions, and fading was measured
spectrophotometrically at the lmax of the dye at intervals over a
period of 1500 hours, and at an ambient temperature of
40¡5 ³C. It was found that fading kinetics were approximately
®rst order over the initial stages of the fading process but then
became more complex. A useful empirical measure of the
photostability was provided by the time for 50% fade (t1/2),
which could be measured directly from the fading curve, or
determined by extrapolation in the case of those dyes with very
Table 2 Photofading parameters for dyes 1±4 and 7, 8 in E7
Dye
Initial ®rst order rate constant/1027 s21
t1/2/ha
1
0.5
1.5
1.2
2.7
1.7
1.6
0.4
3.2
1.7
1.9
5.4
3.4
3900
1100
1600
600
1100
125
4700
700
1100
1000
350
2a
2b
3a
3b
4
7a
7b
7c
8a
8b
8c
570
aTime for 50% fade.
Fig. 2 Two potential molecular alignment axes for 8c.
2336
J. Mater. Chem., 1999, 9, 2333±2338

View Article Online
tautomerism. It is well known that hydrazone tautomers are
more susceptible to photooxidative degradation than azo
tautomers.13 The degradation process is initiated by peroxide
radical hydrogen abstraction from the hydrazone N±H group,
or by an ``ene'' reaction between the hydrazone and singlet
oxygen, both pathways producing the same unstable hydro-
peroxide which thermally degrades to a quinone and a
diazonium species.13 In the aminoazobenzene series 3 and 4,
although such dyes exist predominantly in the azo form, there
will be a small amount of a hydrazone structure present in
equilibrium, and this will undergo such a photodegradation
process, leading ultimately to a rapid fading of the azo dye. The
importance of this process can be inferred by comparing the
fading rates of the hydroxy dyes 3 with those of the
corresponding methoxy dyes 2, which are unable to produce
a hydrazone tautomer. Considering empirical t1/2 values, it is
found that 2a is almost twice as stable as 3a, and 2b is ca. 40%
more stable than 3b. However, hydrazone formation is not the
whole story, as the methoxy dyes themselves are signi®cantly less
photostable than the parent dye 1. This may re¯ect the general
oxidative nature of azo dye photodegradation, when electron
donating groups will be expected to facilitate such a process.
With the exception of 7a, all the dyes in the series 7 and 8
exist in their hydrazone forms in E7. It is signi®cant therefore
that only 7a, present predominantly in the azo form, shows a
high photostability in this medium. Thus it appears that the
hydrogen bonding 3-carboxamide group in dyes 7b,c and 8b,c
serves principally to force the dyes to adopt the hydrazone
structure and consequently cause a decrease in photostability.
Any bene®cial effects that the additional hydrogen bonding
from this group might have on excited state lifetimes are more
than offset by this factor.
hydrazone form. The known sensitivity of hydrazone tauto-
mers to photooxidation thus explains the observed decrease in
photostability.
Experimental
Preparation of intermediates and dyes
General procedure for synthesis of 4-(N,N-diethylamino)-4'-
nitroazobenzenes 2 and 3. Diazotisation of nitroanilines. The
appropriate nitroaniline (0.02 mol) was dissolved in a mixture
of concentrated hydrochloric acid (6 cm3) and water (6 cm3)
and poured onto crushed ice (ca. 20 g) with stirring. A solution
of sodium nitrite (0.02 mol) in water (20 cm3) was added
immediately and the mixture stirred at 0±5 ³C for 30 minutes.
The resultant diazonium solution was then used directly for the
second step.
Coupling. The appropriate N,N-diethylaniline (0.02 mol)
was dissolved in a mixture of acetic acid (12 cm3) and water
(6 cm3), and sodium acetate trihydrate (15 g) was added with
stirring, followed by suf®cient acetic acid to give a clear
solution. The diazo solution was added to the stirred solution at
0±5 ³C over 30 minutes, maintaining the pH between 4 and 7 by
addition of sodium hydroxide solution. The mixture was
allowed to rise to room temperature over 3 hours and the
precipitated product ®ltered off, washed with water and dried.
The following puri®cation procedures were used: 2a: recrys-
tallisation from cyclohexane; 2b: recrystallisation from petro-
leum spirit (bp 80±100 ³C); 3a: column chromatography
(neutral alumina/toluene±dichloromethane) followed by recrys-
tallisation from toluene; 3b: column chromatography (silica/
dichloromethane) followed by recrystallisation from toluene.
Yields and characterisation data are summarised in Table 3.
Conclusions
Although 2- and 2'-hydroxy groups in the 4-(N,N-diethyl-
amino)-4'-nitroaminoazobenzene chromophoric system will
intramolecularly hydrogen bond to the azo group, so affording
increased planarity and reduced rotational mobility of the
molecule, this does not lead to enhanced order parameters in
comparison with the parent non-hydrogen bonded dye. It
appears that the additional hydroxy substituents inhibit dye±
4-(N,N-Diethylamino)-2,2'-dihydroxy-4'-nitroazobenzene 4.
2-Hydroxy-4-nitroaniline (0.02 mol) was dispersed in water
(50 cm3) containing sodium nitrite (0.02 mol) and cooled to
v5 ³C by addition of ice. The suspension was poured into a
solution of concentrated hydrochloric acid (6 cm3) and water
(6 cm3) and stirred at 0±5 ³C for 2 hours. The diazo solution
was then added with stirring at 0±5 ³C to a solution of 3-
hydroxy-N,N-diethylaniline (0.02 mol) in water (60 cm3) con-
taining sodium hydroxide (0.8 g, 0.02 mol) and sodium
carbonate (2.12 g, 0.02 mol) over a period of 90 minutes.
Stirring was continued for 3 hours, allowing the temperature to
rise to ca. 25 ³C, and after adjusting the pH to 7 with
hydrochloric acid the product was ®ltered off, washed and
dried (98% yield). Puri®cation was effected by recrystallising
twice from toluene. Characterisation data are given in Table 3.
mesogen interaction, presumably by
a combination of
increased lateral molecular size and prevention of speci®c
interactions between the azo group and the host mesogens.
However, when the hydroxy derivatives are compared with
their methoxy-substituted analogues, they do show slightly
enhanced order parameters.
In the 1-arylazo-2-hydroxynaphthalene system the situation
is complicated by azo±hydrazone tautomerism, and the
introduction of a hydrogen bonding carboxamide group in
the 3-position causes the dyes to adopt exclusively the
hydrazone form, whereas in the absence of this substituent
the azo form is present in signi®cant amounts. The carbox-
amide-substituted dyes showed low order parameters, and this
was attributed partly to the unfavourable change in the
permanent p-dipole moment direction caused by the carbox-
amide group, and partly to the lateral extension of the molecule
introduced by the side chain attached to the amide group.
These side chains were essential, however, to provide adequate
dye solubility in the liquid crystal host.
Synthesis of 1-arylazo-2-naphthol dyes 7a and 8a. Diazotisa-
tion of 4-hexyl- and 4-hexyloxy-aniline. A solution of the
arylamine (6 mmol) in a mixture of water (25 cm3) and
hydrochloric acid (20 mmol) was cooled to 0±5 ³C, and a
solution of sodium nitrite (6 mmol) in water (25 cm3) was
added over 30 minutes with constant stirring, maintaining the
temperature below 5 ³C . After stirring at this temperature for a
further 30 min. the clear solution of the diazonium chloride was
used immediately for preparation of the appropriate azo dye.
In addition to providing no order parameter advantages,
hydrogen bonding groups in both the aminoazobenzene and 1-
arylazo-2-naphthol series proved to have a pronounced
accelerating effect on photochemical fading. This could be
attributed to the hydrogen bonding groups in¯uencing azo±
hydrazone tautomerism. Thus in the aminoazobenzenes, the
hydroxy groups permitted the existence of hydrazone forms,
which were otherwise absent. In the 1-arylazo-2-naphthols the
carboxamide group enforced exclusive adoption of the
Coupling. 2-Naphthol (6 mmol) was dissolved in a solution
of sodium hydroxide (20 mmol) in water (150 cm3) and cooled
to 0±5 ³C. The appropriate diazo solution was added dropwise
with vigorous stirring over 30 min, maintaining the tempera-
ture below 5 ³C. The pH of the solution was then adjusted to
ca. 9 by addition of sodium hydroxide and stirring was
continued at this temperature for a further hour. The
suspension was then warmed to room temperature and after
J. Mater. Chem., 1999, 9, 2333±2338
2337

View Article Online
Table 3 Yields and characterisation data for dyes 2±4, and 7,8
Microanalysis
Found/%
Calculated/%
C
Dye
Yield/%
Mp/³C
Formula
C
H
N
H
N
2a
2b
3a
3b
4
7a
7b
7c
8a
8b
8c
93
33
95
19
98
58
77
61
93
71
74
138±139a
140±141b
212±214
190±200
268±270
65±67
145±148
69±72
78±80
C17H20N4O3
C17H20N4O3
62.6
62.2
60.9
61.3
58.1
79.6
71.5
71.9
75.7
68.5
69.8
6.1
6.0
6.0
5.8
5.6
7.8
7.2
6.9
6.7
7.0
7.0
17.1
17.1
17.6
17.8
16.8
8.0
10.2
9.6
8.0
9.5
62.2
62.2
61.1
61.1
58.2
79.5
71.6
72.5
75.9
69.0
70.0
6.1
6.1
5.7
5.7
5.5
7.2
6.9
7.4
6.9
6.7
7.1
17.1
17.1
17.8
17.8
17.0
8.4
10.0
9.4
8.1
9.7
C
16H18N4O3
C16H18N4O3
16H18N4O3
C22H24N2O
25H29N3O3
C27H39N3O3
22H24N2O2
C
C
C
160±163
106±108
C25H29N3O4
C27H39N3O4
8.5
9.0
aLit.,15 138±140 ³C. bLit.,16 139±140 ³C.
30 minutes the product was ®ltered off, washed with water and
recrystallised from ethanol. Yields and characterisation data
are summarised in Table 3.
Apar Aper
S ˆ
ꢀ1†
2Aper ‡ Apar
where Apar and Aper are the absorbance values for light
polarised parallel and perpendicular respectively to the
alignment direction of the liquid crystal host.
General procedure for 7b and 8b
Methyl 2-hydroxy-3-naphthoate (561022 mol) was heated
until melted and ethanolamine (0.1 mol) was added. The
mixture was then heated under re¯ux for 15 minutes and then
poured into warm water (200 cm3) at 70 ³C with thorough
stirring. Hydrochloric acid was added to adjust the pH to ca. 1,
and stirring was continued until the suspension had cooled to
room temperature. The solid was ®ltered off, washed with
water, dried and recrystallised from a mixture of petroleum
ether and toluene (1 : 1) to give 2-hydroxy-3-[N-(2-hydro-
xyethyl)]naphthamide 10b in 35% yield. A solution of 10b
(6 mmol) in aqueous sodium hydroxide (20 mmol in 150 cm3)
was then coupled to the appropriate diazo solution following
the procedure outlined for dyes 7a and 8a. The dye was ®ltered
off and recrystallised repeatedly from ethanol until pure by
TLC. Characterisation data are given in Table 3.
Photofading measurements
Unsealed cells were irradiated in a Mark 1 MBTL Microscal
Fadometer, using a phosphor-coated 500 watt mercury±
tungsten ®lament lamp, and were situated at a distance of
15 cm from the lamp outer surface. The lamp output consisted
of broad weak emission between 300 and 700 nm, with major
lines at 348, 395, 438, 545, 585 and 666 nm, thus simulating
high intensity daylight exposure. The cells attained an
equilibrium temperature of ca. 40¡5 ³C throughout the
exposure period. Absorption spectra were measured at
appropriate intervals and the absorbance of the dye solution
at the lmax determined.
Acknowledgements
General procedure for 7c and 8c
We thank the EPSRC Mass Spectrometry Service Centre,
Swansea, for provision of mass spectrometry services.
Following the method detailed for 10b, methyl 2-hydroxy-
naphthoate (561022 mol) was reacted with 3-methoxypropy-
lamine (0.1 mol) to give 3-[N-(3-methoxypropyl)]-2-hydroxy-3-
naphthamide 10a in 52% yield. This was coupled to the
appropriate diazo solution following the general procedure
outlined for 7b and 8b. Yields and characterisation data for 7c
and 8c are summarised in Table 3.
References
1
2
G. W. Gray, Dyes Pigm., 1982, 3, 203.
P. Gregory, High Technology Organic Colorants, Plenum, New
York, p. 991.
3
N. S. Allen and J. F. McKellar, Developments in Polymer
Photochemistry, ed. N. S. Allen, Applied Science, London, 1980,
vol. 1, p. 191.
Measurement of order parameters
The liquid crystal host E7 (BDH-Merck), a eutectic mixture
with a stable nematic mesophase over the range 210 to 60 ³C,
was used for the measurements. Cells were constructed
according to the procedure detailed by Jones and Reeve,14
using microscope slides coated with oriented polyvinyl alcohol.
A cell gap of 25 mm was maintained with aluminium foil spacers,
and the cells were sealed along three sides with epoxy resin. Cells
were ®lled by capillary action with solutions of the dyes in E7,
the concentration being adjusted to give an absorbance reading
in the range 0.5±1.5 at the lmax of the dye (corresponding to dye
concentrations in the range 0.5±1.0%). After ®lling, the cells
were heated to 15 ³C above the N±I transition temperature and
cooled slowly to room temperature before measurements were
made. The cells were inspected microscopically to ensure that
they were free from undissolved dye particles. The absorption
spectra of the cells in the parallel and perpendicular orientation
modes were measured using polarised light, and the order
parameters S were calculated using eqn. (1):
4
5
6
J. Kelemen, Dyes Pigm., 1981, 2, 73.
A. Whitaker, J. Soc. Dyers Colour., 1978, 94, 431.
K. Venkataraman, The Chemistry of Synthetic Dyes, Academic
Press, New York, 1952, vol. 1, p. 650.
7
8
J. Grif®ths, Dyes Pigm., 1982, 3, 211.
A. Burawoy, A. G. Salem and A. R. Thompson, J. Chem. Soc.,
1952, 4793.
C.BlackburnandJ.Grif®ths,Mol.Cryst.Liq.Cryst.,1983,101,341.
9
10 G. Wettermark, M. E. Langmuir and D. G. Anderson, J. Am.
Chem. Soc., 1965, 87, 476.
11 G. Gabor and E. Fischer, J. Phys. Chem., 1962, 66, 2478.
12 T. Kobayashi, E. O. Degenkolb and P. M. Rentzepis, J. Phys.
Chem., 1979, 83, 2431.
13 J.Grif®thsandC.Hawkins,J.Chem.Soc.,PerkinTrans.2,1977,747.
14 F. Jones and T. J. Reeve, Mol. Cryst. Liq. Cryst., 1981, 78, 201.
15 P. Gregory and D. Thorpe, J. Chem. Soc., Perkin Trans. 1, 1979,
1990.
16 C. H. Haessner and H. Mustroph, J. Prakt. Chem., 1987, 329, 493.
Paper 9/01936G
2338
J. Mater. Chem., 1999, 9, 2333±2338