Photochromism in spiroindolinonaphthoxazine dyes: Effects of alkyl and ester substituents on photochromic properties

Article abstract of DOI:10.1016/j.dyepig.2014.01.005

The impact of substituents on the photochromism of some spirooxazines has been explored. A series of photochromic spiroindolino[2,1-b][1,4]naphthoxazine dyes was prepared with varying alkyl substitution patterns in the 3- to 6-positions of the indoline system. It was found that placing methyl functions onto the phenyl ring produced bathochromic shifts in λ_max of the photomerocyanine as expected, but contrary to literature indications lowered photocoloration half-life. However, replacing the 3,3-dimethyl pattern with 3-ethyl-3-methyl substitution led to more intense photochromism: half-life increased without affecting λ_max. In contrast, introduction of a cyclohexyl ring to give a 3,3-pentamethylene fragment and substitution at the 5′-position of the naphthoxazine moiety with either ester or hydroxymethyl groups significantly increased the speed of fade in general contrast to literature reports for analogous dye series. Similar effects were also observed concerning alkyl substitution in a smaller synthesised set of photochromic [1,2-b][1,4]oxazines.

Full text of DOI:10.1016/j.dyepig.2014.01.005

Accepted Manuscript
Photochromism in spiroindolinonaphthoxazine dyes: effects of alkyl and ester
substituents on photochromic properties
Steven M. Partington, Andrew D. Towns
PII:
S0143-7208(14)00006-0
10.1016/j.dyepig.2014.01.005
DYPI 4234
DOI:
Reference:
To appear in: Dyes and Pigments
Received Date: 3 December 2013
Revised Date: 28 December 2013
Accepted Date: 3 January 2014
Please cite this article as: Partington SM, Towns AD, Photochromism in spiroindolinonaphthoxazine
dyes: effects of alkyl and ester substituents on photochromic properties, Dyes and Pigments (2014), doi:
10.1016/j.dyepig.2014.01.005.
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please
note that during the production process errors may be discovered which could affect the content, and all
legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT
Photochromism in spiroindolinonaphthoxazine dyes:
effects of alkyl and ester substituents on photochromic properties
Steven M. Partington and Andrew D. Towns*
Vivimed Labs Europe Ltd., PO Box B3, Leeds Road, Huddersfield, HD1 6BU,
West Yorkshire, United Kingdom
*
Author to whom correspondence should be addressed
E-Mail: andy.towns@vivimedlabs.com
Tel.: +44 (0)1484 320500
Fax: +44 (0)1484 320300
Abstract
The impact of substituents on the photochromism of some spirooxazines has
been explored. A series of photochromic
spiroindolino[2,1-b][1,4]naphthoxazine dyes was prepared with varying alkyl
substitution patterns in the 3- to 6-positions of the indoline system. It was
Page 1 of 33

ACCEPTED MANUSCRIPT
found that placing methyl functions onto the phenyl ring produced
bathochromic shifts in λmax of the photomerocyanine as expected, but
contrary to literature indications lowered photocoloration half-life. However,
replacing the 3,3-dimethyl pattern with 3-ethyl-3-methyl substitution led to
more intense photochromism: half-life increased without affecting λmax. In
contrast, introduction of a cyclohexyl ring to give a 3,3-pentamethylene
fragment and substitution at the 5'-position of the naphthoxazine moiety with
either ester or hydroxymethyl groups significantly increased the speed of fade
in general contrast to literature reports for analogous dye series. Similar
effects were also observed concerning alkyl substitution in a smaller
synthesised set of photochromic [1,2-b][1,4]oxazines.
Keywords
Photochromism; Photochromic dye; Spirooxazine; Synthesis; Absorption;
Half-life
Page 2 of 33

ACCEPTED MANUSCRIPT
1. Introduction
Spirooxazines are one of the two most industrially important classes of
organic photochromic colorant [1]. Since this family of compounds was first
explored half a century ago, a substantial amount of research has been
undertaken into their synthesis and the effects of structure on their
photochromism [2]. Much of this effort followed the discovery in the late
1970s of the relatively high resistance of spirooxazines to UV
photodegradation [3], which ultimately resulted in their successful
commercialisation as sunlight-responsive colorants in ophthalmic lenses in
the 1990s [4,5]. In addition to robustness, the spirooxazine class is prized for
synthetic accessibility and the way in which the colour and kinetics of
photochromism of its members can be fine-tuned economically by variation in
structure to give industrially valuable dyes with bluish-red to turquoise photo-
activated forms. Examples are 1 and 2 whose photomerocyanine forms are
blue [6] and turquoise [7] respectively.
Me
Me
O
Me
Me
O
N
N
N
N
NEt2
1
2
However, despite the volume of data available in the open and patent
literature illuminating such relationships, gaps remain concerning the impact
of certain structural features on both colour and kinetic properties. Although
there is continued interest in photochromic spirooxazine systems, both in
theoretical [8,9] and practical [10,11] terms, reports in the peer-reviewed
Page 3 of 33

ACCEPTED MANUSCRIPT
literature of relations between the photochromism and constitution of
commercially interesting spirooxazines are scarce. For example, while
3,3-dimethyl substitution is a common structural motif in commercial
spirooxazines, variations have been explored industrially [12,13,14,15] and
exploited successfully, but little has been published on their effects. In
addition, gleaning the impact of altering molecular features from publicly
available data can be complicated by variations in media and conditions used
in different studies preventing direct comparison of properties – this is
exacerbated by apparently small structural modifications producing significant
changes in colour and kinetics.
As part of an ongoing programme of work conducted over the last two
decades into photochromic compounds, this paper reports on some findings
of a study into substituent effects on the photochromism of spiroindolino-
naphthoxazines with potential commercial utility. The specific aim was to
discover structural modifications that would deliver enhanced photocoloration
without substantially altering hue through the preparation and examination of
the properties of a series of [2,1-b][1,4]oxazines 3 (see Table 1). Guided by
isolated reports concerning analogous systems, alkyl and ester functions were
introduced onto the indoline and oxazine moieties, but their effects did not
always tally with expectation: in some cases, modifications were found to
give photochromic compounds with weaker photocoloration than the parent
structure. In order to establish the impact of these kinds of changes in the
less well-explored [1,2-b][1,4]oxazine class, a small set of dyes 4 was
synthesised and investigated (see Table 1). Published data concerning these
types of photochromic system are sparse outside of the patent literature. We
Page 4 of 33

ACCEPTED MANUSCRIPT
therefore report the effects of such structural variations on members of series
3
and 4 in terms of their colour when photo-activated and the intensity of their
photocoloration as well as their rate of thermal fading.
Table 1 Photochromic oxazines synthesised and studied
4
4
R
R'
10'
9'
R
R'
5
6
N
5
6
N
1
0'
8
'
N
O
N
O
NEt2
7'
5
'
6'
3
4
Compound
Substituents
5'
R
R'
Phenyl ring
3
a
Me
Me
Me
Et
Me
Me
Me
Me
H
H
H
H
H
5-Me
4,5-(5,6-)Me
2
3b
3c
4,6-Me
2
3d
–
–
–
–
–
3e
-(CH ) -
2 5
3
f
Me
Et
Me
Me
Me
CO
CO
2
Me
Me
3
g
h
2
3
Me
2
CH OH
4
a
Me
Et
Me
Me
–
4,5-(5,6-)Me
2
4b
–
–
–
–
4c
2 5
-(CH ) -
Page 5 of 33

ACCEPTED MANUSCRIPT
2. Results and Discussion
2.1 Intermediate and dye synthesis
Derivatives 3a-3e were prepared using the conventional route of
condensation of Fischer’s base analogues 5 with 1-nitroso-2-naphthol 6 (X =
H) in toluene or methanol (see Scheme 1). Following recrystallisation (and
flash chromatography where necessary), the dyes exhibited sharp melting
points accompanied by the appearance of intense and irreversible coloration
just prior to, as well as during, liquefaction. The latter effect is the norm for
spironaphthoxazines as a consequence of thermally promoted ring-opening.
Purity of each dye was established by TLC and HPLC, while identity was
confirmed with proton NMR spectroscopy. A study concerning their mass
spectroscopic characterisation is in progress and will be reported separately.
Scheme 1 Synthetic route to spirooxazines 3a-3g
4
R
R'
NO
6
5
6
PhMe or
MeOH, ∆
OH
X
N
+
3a-3g
X = H or
CO Me
2
5
NaNO2 /
AcOH
X =
O
CO Me
5
a: R, R' = Me; 5-Me
2
R
PhMe, AcOH, ∆
5b: R, R' = Me; 4,5-(5,6-)Me
2
5
c: R, R' = Me; 4,6-Me2
R'
OH
X
5
3
5d: R = Et, R' = Me
e: RR' = -(CH ) -
5
2
5
4
N
NH2
Page 6 of 33

ACCEPTED MANUSCRIPT
The requisite 2-methyleneindolines 5 were prepared by Fischer synthesis [16]
involving condensation of N-neopentyl-N-phenylhydrazines with a ketone
followed by in-situ cyclisation of the resultant hydrazone. Alkyl substitution
3
patterns on the benzene ring and the sp carbon at the 3-position of 5 were
determined by choice of starting hydrazine and ketone reagent respectively.
With 3-methyl-2-butanone (R = R' = Me), N-(4-methylphenyl)-N-
neopentylhydrazine furnished 5a while its 3,5-dimethyl analogue was
converted to 5c. Variations in 3-substitution were introduced through use of
3-methyl-2-pentanone (R = Et, R' = Me) and cyclohexyl methyl ketone (RR' =
-
(CH ) -), yielding 5d and 5e, respectively, from N-neopentyl-N-
2
5
phenylhydrazine. The expectation that the 3,4-dimethyl substitution pattern of
the hydrazine precursor to 5b would furnish an isomeric mixture was
confirmed by analysis of the dye 3b prepared from it. HPLC revealed two
product peaks of near equal size corresponding to 4,5-dimethyl and 5,6-
dimethyl isomers. This was borne out by the NMR data for this dye. Although
the spectrum was complex, many of the signals could be identified as pairs of
signals of equal intensity as would be expected from a composition consisting
of two isomers, for example, two singlets with a total integral of 9H for the nine
protons of the methyl groups on the neopentyl chain (see Section 4.2.2 and
Supplementary Data).
Proton NMR data of the other dyes within the group 3a-3e were
consistent with that previously published for spirooxazines derived from
Fischer’s Base, i.e. with 3,3-dimethyl substitution [17]. However,
diastereoisomerism was detected in the case of 3d. This derivative has a
chiral centre at the 3-position of the indoline system in addition to the spiro
Page 7 of 33

ACCEPTED MANUSCRIPT
carbon. While the isomers were unresolved by TLC, two close-running peaks
of similar size considered to correspond to each isomer were evident by
HPLC. The presence of diastereoisomers was also suggested by the NMR
data. Certain signals originating from protons on the Fischer’s base and
naphthyl moieties formed pairs of equal intensity. For example, two doublet
signals of integration 0.5H each with shifts centred on 8.54ppm and 8.55ppm
were assigned to the single proton at the 10'-position (see Section 4.2.4 and
Supplementary Data).
Compounds 3f and 3g were synthesised from 3-carboxymethyl-1-
nitroso-2-naphthol 6 (X = CO Me). This intermediate was prepared by
2
addition of aqueous sodium nitrite to a chilled solution of 3-carboxymethyl-2-
naphthol in acetic acid, a method derived from a very brief outline given in the
literature [18]. Attempts to nitrosate the naphthol ester based on a literature
method claiming a 92% yield [19] in which sulphuric acid is added to a
suspension of it and sodium nitrite in aqueous pyridine furnished mixtures of
product and starting material containing less than half of the former. As was
the case with 3d, two peaks were observed upon HPLC analysis of 3g, which
can be ascribed to the resolution of diastereoisomers. The presence of
isomers was evident from NMR data, for example, by the observation of two
singlet signals, whose total integration was 1H, originating from the proton at
the 6'-position (see Section 4.2.7 and Supplementary Data). The existence of
carbonyl ester functions in 3f and 3g was confirmed by FTIR spectroscopy.
Instances of 5'-hydroxymethyl-substituted spirooxazine dyes have been
reported in the literature prepared from 3-hydroxymethyl-1-nitroso-2-naphthol
[20,21]. However, synthesis of 3h is unusual in that it was prepared by
Page 8 of 33

ACCEPTED MANUSCRIPT
Red-Al reduction of an ester-substituted spirooxazine, namely 3f.
Spectroscopic data were consistent with a hydroxymethyl derivative: signals
associated with ester groups were absent in both NMR and FTIR spectra.
Derivatives 4 were prepared by condensation of some of the bases 5
with a 1-naphthol derivative [7] in refluxing methanol. Purity following
repeated recrystallisation was confirmed by HPLC. Like 3b, which was also
2
prepared from 5 (R = R' = Me; 4,5-(5,6-)Me ), two components were observed
by HPLC and NMR in the case of 4a in roughly equal proportion
corresponding to 4,5- and 5,6-isomers (see Section 4.2.9 and Supplementary
Data). Unlike the other two derivatives prepared from 5 (R = Et; R' = Me), i.e.
3d and 3g, the main peak observed during HPLC analysis of 4b was not
resolved into diastereoisomers, although there was evidence of their
existence in the NMR data. Certain proton signals were identified as
consisting of pairs with equal integrated intensities, e.g. two signals with an
integration totalling 3.0H attributed to the protons of the 3-methyl group (see
Section 4.2.10 and Supplementary Data).
2.2 Photochromic properties of oxazine derivatives
Values of absorption maximum (λmax), half-life (t ) and strength in toluene
½
solution for the oxazine series 3 and 4 as well as their parent compounds 1
and 2 are shown in Table 2. Solutions of the dyes developed blue to
turquoise coloration upon exposure to near UV radiation (see Supplementary
Data for example spectra). First-order rates of reversion to the colourless
forms of the dyes were observed following cessation of activation of solutions
in a photostationary state. The absorbance values tabulated give an
Page 9 of 33

ACCEPTED MANUSCRIPT
indication of strength but should be treated with caution: the figures are not
necessarily proportional at the concentration of dye employed. Nevertheless,
½
reasonable correlations between t and strength are evident.
Table 2 Absorption maxima and photochromic properties of dyes 1-4
a
Compound
λmax (nm)
t½ (s)
Strength
1
600
608
610
605
600
610
620
620
608
31
21
19
20
52
7.1
12
20
14
14
8.4
5.4
6.3
20
3a
3b
3c
3d
3e
2.5
3.5
4.8
5.6
3f
3
g
h
3
2
618
632
618
626
120
55
87
19
60
8.4
4a
4b
300
26
4c
a
2
-1
absorbance x 10 at 40mg L
Page 10 of 33

ACCEPTED MANUSCRIPT
Examples disclosed in the patent literature [22,23] indicated that methyl
substitution on the benzene ring of the indoline moiety of led to stronger,
slower-fading photocoloration (see Table 3). However, contrary to these
reports, it was found that such methyl substitution had the opposite effect:
whereas the parent dye 1 gave a reasonably intense effect with a half life of
31s, the methyl-substituted dyes 3a-3c were weaker and faded more rapidly
½
(t 19-21s). While the substituent effect on kinetics was not as anticipated
given the literature precedent, shifts in absorption maxima were in line with
expectation. Location at the 5-position tends to give the most pronounced
shift [2], which was the case here: an 8nm red-shift in absorption for 3a from
1. The influence of methyl substitution on λmax at the 4- and 6-positions was
much more modest (2-3nm assuming additivity of effect).
Previously, patent literature [22,23] had indicated that addition of
methyl groups to the indoline benzene ring increased photocoloration
intensity, for example, by an increase in ∆OD shown by 7b and 7f relative to
1
5
7
a and 7d respectively, or decreased TD values of 7c and 7h compared to
7a and 7g respectively (see Table 3). In the case of the latter pairings, half-
5
life as alluded to by R values was also lengthened by 4,5-(5,6-)dimethyl
substitution in contrast to the increase in thermal fade observed between 1
and 3b. These disparities may be a result of the variations in methyl
substitution acting in concert with other structural differences, failure to
achieve a photostationary state, a consequence of the influence of
environment (e.g. medium, UV source), or a combination of these factors.
Page 11 of 33

ACCEPTED MANUSCRIPT
Table 3 Photochromic properties disclosed by patents for methyl-substituted
spirooxazines [22,23]
4
Me
Me
O
9
'
5
6
N
N
1
∆
OD per
b
15 b
5 b
Compound
Substituents
λ
max
TD
R
a
mm
AU mm ) (nm)
-1
(
(%)
(%)
[23]
4
5
6
9'
[22]
[23]
[23]
1-Methyl
7
a
H
H
H
H
H
H
H
H
0.15
0.23
0.15
0.30
0.30
0.43
604
616
39
18
83
68
7b
Me
7c
(Me) Me (Me)
7d
H
H
H
H
H
OMe
OMe
7e
Me
7
f
(Me) Me (Me) OMe
1-(2-Methylcyclohexyl)
7
g
h
H
H
H
H
H
614
624
23
5
72
60
7
(Me) Me (Me)
a
optical density change at 23 °C of cellulose acetat e butyrate films (ca. 6mm thick) prepared
b
with 2%w/w loading of 7; in 2mm thick methacrylate-based lens at 22°C
15
5
(
TD = transmission at λmax after 15min irradiation; R = recovery of initial transmission after
5min of decolorisation)
Page 12 of 33

ACCEPTED MANUSCRIPT
The impact of methyl substitution at positions 4-6 on the indoline
moiety as discussed above for the [2,1-b] systems 3 was found to be similar in
the case of some [1,2-b] counterparts. Introduction of 4,5-(5,6-)dimethyl
substituents into 2, producing 4a, leads to bathochromism (14nm), a
significant increase in speed of fade (t 120s falls to 55s) and loss in strength.
½
Increasing crowding by attaching groups with greater bulkiness to the
nitrogen atom of the indoline moiety has long been known to cause a
reduction in fade rate with little impact on observed colour [24]. However,
very little data is available in the open literature concerning the impact of
introducing more sizeable substituents into the 3-position of the indoline
system than are present in the near-ubiquitous dimethyl pattern. Effects were
found to be mixed:
o switching one of the methyl groups of 1 for an ethyl function to give 3d
was found to bring about substantial increases in half-life and strength
while leaving the wavelength of maximum absorption unchanged;
o joining the 3,3-substituents of 1 to create a six-membered ring as in 3e
led to a red shift in absorption of 10nm and pronounced reductions in t
½
(31s to 7.1s) and strength.
The effects noted above for examples of the [2,1-b][1,4]oxazine system
were also observed in [1,2-b] systems. Extending one of the 3-methyl groups
of 2 to an ethyl chain, furnishing 4b, did not affect λmax but resulted in a much
½
slower fading derivative (t 300s compared to 120s). The spiro-aliphatic ring
at the 3-position of 4c brings about increased fading with concomitantly
weaker intensity of photocoloration.
Page 13 of 33

ACCEPTED MANUSCRIPT
These findings are in contrast to the limited amount of data reported for
spirooxazines and spiropyrans pertaining to these kinds of substituent
changes at the 3-position. For example, replacement of one of the 3-methyl
substituents of either 8a or 9a with an ethyl group to give 8b and 9b makes
little difference to thermal fade rate (K) or intensity of photocoloration as
0
expressed by colourability (A ) [25] (see Table 4).
Table 4 Effect of 3-substitution in pyrimidino and thiazolo spirooxazines 8
-
5
and 9 in toluene solution at a concentration of 2.5x10 M and temperature of
5 °C [25]
2
Me
R
Me
R
N
N
N
N
S
N
N
O
N
O
8
9
-
1
a
b
Compound
R
λmax (nm)
K (s )
A0
8
a
Me
Et
595
598
0.13
0.10
0.61
0.53
8b
9
a
Me
Et
621
622
0.09
0.08
0.24
0.24
9b
a
b
rate constant of decolorisation colourability (see [25])
Page 14 of 33

ACCEPTED MANUSCRIPT
As was observed in the case of spirooxazines 1-4, a study of spiropyrans that
included 10 also found that replacement of the gem-dimethyl function with a
cyclohexyl ring shortened half-life [26]. However, the difference was less
½
marked between 10a and 10b (t 2min and 1min, respectively, in acetonitrile
and 0.65min to 0.5min in dichloroethane at 21 °C ).
X
1
0a, X = C(Me)
2
N
O
NO2
10b, X =
Ester substitution in the naphthol moiety adjacent to the oxazine ring, i.e. at
the 5'-position, has been previously disclosed to slow down fade rate in the
case of derivatives 11 [13]. Not only is 11b claimed to be more bathochromic
compared to its unsubstituted analogue 11a, but also its photocoloration is
more intense and persistent after removal of the UV stimulus.
11a, 5'-H
N
2
11b, 5'-CO Me
N
O
5'
Broadly similar observations have been reported for adamantyl-derived
spironaphthoxazines [19]. Table 5 shows that introduction of a
5'-methoxycarbonyl group into 12a to give 12b reduces the rate of thermal
fade (K) while colourability (A
0
) increases, i.e. half-life lengthens and intensity
rises.
Page 15 of 33

ACCEPTED MANUSCRIPT
Table 5 Effect of 5'-substituents in homoazaadamantane spirooxazines 12 in
-
5
toluene solution at a concentration of 2.5x10 M and temperature of 25 °C [27]
N
N
O
1
2
Y
-
1
a
b
Compound
2a
2b
2c
Y
H
λmax (nm)
K (s )
A
0
1
580
590
580
0.092
0.044
0.037
0.96
1.09
0.80
1
CO Me
2
1
2
CH OH
a
b
rate constant of decolorisation colourability (see [27])
However, introduction of a 5'-methoxycarbonyl function into 1 to create 3f
does neither. The latter is considerably faster fading (t 12s compared to
1s) and weaker colouring than the former. The significant bathochromic shift
½
3
of 20nm is however in agreement with theory, the ester function being a
moderate electron acceptor, although the analogous shift in the case of
derivatives 12 is smaller (10nm). Similar changes in behaviour were
observed with another pairing: 5'-ester substitution in 3g confers
photocoloration that is much faster fading, less intense and greener compared
to that of unsubstituted 3d. The shift is again 20nm, while the decreases in t½
and strength in the case of this pair are roughly proportional to those
observed with the pair of dyes 1 and 3f. Absorption maxima are consistent
with literature data: the analogue of 1 with N-methyl substitution (7a) has a
Page 16 of 33

ACCEPTED MANUSCRIPT
maximum at 594nm in toluene [19], a hypsochromic shift of 6nm compared to
the neopentyl derivative as a consequence of reduced electron donation. A
similar shift (8nm) was seen between the λmax of 3f and the maximum (612nm
in toluene) reported for its N-methyl analogue [28].
Swapping the 5'-ester function of 3f for the more weakly electron-
accepting hydroxymethyl group produces 3h whose absorption maximum is
less red-shifted (8nm cf. 1 as opposed to 20nm for 3f) as expected. While 3h
has a similar half-life to 3f, it is stronger, albeit still considerably weaker and
faster fading than the parent derivative 1. These findings are in contrast to
those reported for the same changes in adamantyl derivatives (see Table 5),
where replacement of the ester group of 11b with a hydroxymethyl substituent
to give 11c leads to a slower-fading but weaker photocoloration.
3. Conclusions
The effects of structural variation upon the photochromic properties of
members of a series of spiroindolino[2,1-b][1,4]naphthoxazine analogues 3 of
the commercially interesting parent dye 1 have been examined. The
influences of alkyl substituents in the 3- to 6-positions of the indoline system
were not found to correspond with behaviour previously reported in patents
and peer-reviewed literature. While the placement of methyl functions onto
the 4- to 6-positions of the phenyl ring of 1 produced bathochromic shifts in
the absorption maxima of the photo-activated forms of the resultant
derivatives in a cumulative manner as expected, thermally-driven ring-closure
was accelerated contrary to disclosures related to other spirooxazine-based
series.
Page 17 of 33

ACCEPTED MANUSCRIPT
Replacing one of the substituents of the 3,3-dimethyl pattern of dye 1
with an ethyl group led to more pronounced photochromism without affecting
the position of the long wavelength absorption maximum. Both half-life and
strength increased significantly whereas literature results for related
spirooxazine series indicated only modest changes. In sharp contrast,
introduction of a cyclohexyl ring to give a 3,3-pentamethylene fragment led to
a substantial drop in intensity and half-life.
It was discovered that reduction of an ester-substituted spirooxazine to
the corresponding hydroxymethyl derivative is possible. However,
substitution at the 5'-position of the naphthoxazine moiety of 1 with ester or
hydroxymethyl groups significantly increased speed of fade in general
contrast to literature reports for analogous dye series.
Similar effects were also observed concerning alkyl substitution at the
3- to 6-positions in a smaller synthesised set of [1,2-b][1,4]oxazines 4 derived
from the industrially useful dye 2. For example, dimethyl substitution on the
benzene ring of the indoline moiety, or swapping the gem-dimethyl pattern at
the 3-position with pentamethylene functionality, red-shifted the absorption of
the photo-activated form, but considerably reduced half-life and intensity.
Page 18 of 33

ACCEPTED MANUSCRIPT
4. Methods and Materials
All starting materials were of commercial origin and used as supplied. TLC
and HPLC were employed to follow reaction progress and confirm product
purity. Unless stated otherwise, TLC was undertaken using mixtures of
petroleum ether (b.p 60-80 °C ) and acetone as eluent in ratios of 9:1, 6:1 or
5:2 by volume on silica (TLC Silica gel 60 F254 aluminium sheets, Merck).
HPLC was performed in reverse phase mode in which samples were diluted
with, or dissolved in, acetonitrile prior to injection. The same type of column
was employed throughout: LiChrospher 60 RP-Select B (5µm) of dimensions
250mm x 4mm (Merck). A mobile phase of acetonitrile and water was used in
a ratio of 80:20v/v for compounds 3 and 4a. A ratio of 85:15v/v was employed
for compounds 4b and 4c. An isocratic flow rate of 1ml/min was maintained
in all cases and monitoring performed at a wavelength of 254nm. RT denotes
retention time of the product peak(s). Uncorrected melting points were
determined on Büchi B-540/B-545 equipment, employing a temperature
gradient of 0.5 °C /min. Proton NMR (300MHz) and FTI R spectra were
recorded in deuterochloroform at 27 °C and KBr, resp ectively, using Bruker
instrumentation.
4.1 Synthesis of intermediates
Fischer’s base analogues 5a-e were prepared by condensation of N-
neopentyl-N-phenylhydrazine or a derivative with a ketone and in situ
cyclisation as illustrated by the synthesis of 5d given below (see Section
4.1.1.); product was isolated cleanly as its hydroiodide salt, which was readily
Page 19 of 33

ACCEPTED MANUSCRIPT
converted to its free base immediately prior to use. N-(4-Methylphenyl)-N-
neopentylhydrazine furnished 5a while the 3,4-dimethyl and 3,5-dimethyl
analogues were converted to the respective 5b and 5c. Cyclohexyl methyl
ketone was used to generate 5e from N-neopentyl-N-phenylhydrazine. All
were prepared as per 5d except isolation of their free base form was
performed by vacuum distillation rather than precipitation of the hydroiodide
salt followed by basification. The hydrazine precursors needed were
prepared via nitrosation and reduction of the parent N-neopentylaniline or an
analogue [6].
4.1.1. Preparation of 3-ethyl-3-methyl-1-neopentyl-2-methyleneindolenine 5d
To a solution of N-neopentyl-N-phenylhydrazine in toluene (0.51M, 236mL)
was added 3-methyl-2-pentanone (39.6g, 0.387mol) and glacial acetic acid
(61mL). The mixture was stirred up to reflux under a Dean-Stark trap for 3h
by which time TLC (silica/5:2 petrol:ethyl acetate) indicated completion of
reaction. The reaction mass was stirred down to ambient temperature before
the addition of ice (225g). Sodium hydroxide (25%w/w, 115mL) was added
gradually <10 °C to give a positive test for alkalin ity against phenolphthalein,
prior to screening to remove a small amount of precipitate. The organic layer
was collected and dried over sodium sulphate. After removal of drying agent
and solvent, the resultant darkish yellow-orange oil was dissolved in
isopropanol (75mL) to which was added hydroiodic acid (57%w/w, 19mL) with
stirring over 30min at <15 °C . A suspension resulte d which was stirred at
0-10 °C for 1h. The solid was collected by filtrati on and washed with
isopropanol (75mL), furnishing the hydroiodide salt of 5d as an off-white
Page 20 of 33

ACCEPTED MANUSCRIPT
powder (33.2g, 74% yield) after air-drying. TLC (as above) confirmed that the
material was pure. The free base form of 5d was liberated as a pale pink oil
in quantitative yield by basification in a mix of water (1.5L/mol) and ethyl
acetate (1L/mol) with sodium hydroxide (25%w/w) until alkaline to
phenolphthalein, followed by isolation of the organic layer, drying over sodium
sulphate and removal of solvent under vacuum. It was then used immediately
without purification.
4
.1.2. Preparation of 3-carboxymethyl-1-nitroso-2-naphthol 6 (X = CO
2
Me)
3-Carboxymethyl-2-naphthol (6.00g, 30mmol) was dissolved in acetic acid
(99%, 200mL) to give pale yellow solution after screening off a small amount
of fine white insoluble material. Water (20mL) was added to the stirred filtrate,
which was then cooled to 5 °C (no precipitation occu rred). A solution of
sodium nitrite (6.2g, 90mmol) in water (28mL) added dropwise at 4-6 °C over
30min. The resultant dark yellow clear solution was stirred at 0-5°C for
50min, during which time a fine bright orange-yellow solid separated out. TLC
(2:2:1v/v/v petrol:chloroform:acetone) showed a high level of conversion of
starting material. Over 20min, water (100mL) was added to the suspension at
a temperature of less than 10 °C followed by ice (10 0g) and then further water
(
200mL) whilst maintaining temperature below 5 °C . After stirring for 15min,
the suspension was filtered and collected solid washed free of acid with water
700mL) before being allowed to air dry, furnishing 6 (X = CO Me) as a dark
yellow powder (4.09g, 59% yield). TLC (as above) indicated that the product
(
2
(R
f
f
0.25) contained only a trace of starting material (R 0.56) and a small
Page 21 of 33

ACCEPTED MANUSCRIPT
amount of yellow baseline impurity. No purification was undertaken prior to
use of the product.
4.2 Synthesis of photochromic dyes
4
1
7
5
.2.1. Preparation of dye 3a
-Nitroso-2-naphthol (3.64g, 21mmol) in toluene (12mL) was stirred up to
0 °C . To the resulting intense yellow-brown soluti on was added a solution of
a (3.65g, 15mmol) in toluene (4mL) over 75min at 60-65 °C . The reaction
mass was stirred at 60-65 °C for 90min after which t ime TLC(9:1) showed that
all of the indoline derivative had been consumed. Activated charcoal (0.28g)
was added and stirring continued for 15min. The mixture was screened and
the collected solid rinsed with toluene (1.5mL) before being discarded. The
product was precipitated from the stirred combined filtrate by dropwise
addition of methanol (53mL) after the filtrate had been cooled to ambient
temperature. Following the end of the addition, the suspension was stirred for
1
h. The crude product was filtered off, washed with methanol (3 x 8mL) and
air-dried, furnishing a cream-coloured crystalline powder (3.53g, 59% yield) of
5.9%area purity by HPLC. Recrystallisation (toluene/methanol) afforded
9
pure 3a as an off-white powder (3.12g, 52% yield). HPLC: 99.7%area (RT
1
9
.7min). Melting point: 173-174 °C . H NMR: 0.97 (s, 9H, CMe
3
); 1.33 (s,
); 2.92 (s, 2H,
); 6.58 (d, J = 7.8Hz, 1H, H-7); 6.89 (s, 1H, H-4); 6.97 (d, J = 8.0Hz, 1H,
3
3 3 3
H, >C(Me)CH ); 1.39 (s, 3H, >C(Me)CH ); 2.33 (s, 3H, Ar-CH
N-CH
2
H-6); 6.99 (d, J = 9.0Hz, 1H, H-5'); 7.38 (t, J = 7.5Hz, 1H, H-8'); 7.56 (t, J =
Page 22 of 33

ACCEPTED MANUSCRIPT
7
.7Hz, 1H, H-9'); 7.64 (d, J = 9.0Hz, 1H, H-6'); 7.73 (d, J = 8.1Hz, 1H, H-7');
.80 (s, 1H, N=CH); 8.54 (d, J = 8.7Hz, 1H, H-10').
7
4.2.2. Preparation of dye 3b
A mixture of 5b (2.57g, 10mmol), 1-nitroso-2-naphthol (2.40g, 14mmol) and
methanol (31mL) was stirred up to reflux. As the mixture was warmed, the
yellow-brown mixture became greener. TLC(6:1) showed the appearance of
a blue product spot (R
f
0.57). The dark green solution was held at reflux for
2h after which time no trace of indoline compound was evident by TLC. The
reaction mass was allowed to cool overnight and the dark, slightly sticky solid
filtered off and washed with methanol (23mL) before drying at 50 °C in an air
oven to give a brown solid (1.73g, 42% yield). The crude product was
dissolved in toluene, purified by flash chromatography (silica, 95:5v/v
petroleum ether:ethyl acetate) and recrystallised repeatedly
(toluene/methanol) to furnish 3b as a pale yellow crystalline powder (0.78g,
19% yield). Chromatography indicated the product was pure. However, while
TLC(6:1) revealed only a single major blue component, two peaks (52.9%area
at RT 11.3min; 47.1%area at RT 11.6min) were observed by HPLC,
presumed to correspond to the 4,6-dimethyl and 5,6-dimethyl isomers or vice
1
versa. Melting point: 140-141 °C . H NMR: 0.95 (s, 4.5H, CMe ); 0.97 (s,
3
4
1
1
.5H, CMe
.5H, >C(Me)CH
.5H, Ar-CH ); 2.27 (s, 1.5H, Ar-CH
); 6.45 (d, J = 7.8Hz, 0.5, H-7), 6.49 (s, 0.5H, H-7); 6.84 (s, 0.5H, H-4);
3
); 1.32 (s, 1.5H, >C(Me)CH
); 1.48 (s, 1.5H, >C(Me)CH
); 2.88 (s, 1H, N-CH
3
); 1.39 (s, 1.5H, >C(Me)CH
); 2.24 (s, 3H, Ar-CH
); 2.92 (s, 1H, N-
3
); 1.46 (s,
3
3
3
); 2.26 (s,
3
3
2
CH
2
6.97 (d, J = 7.5Hz, 0.5H, H-6); 7.00 (d, J = 8.7Hz, 1H, H-5'); 7.37 (t, J = 7.4Hz,
Page 23 of 33

ACCEPTED MANUSCRIPT
1H, H-8'); 7.55 (t, J = 7.5Hz, 1H, H-9'); 7.64 (d, J = 9.0Hz, 1H, H-6'); 7.73 (d, J
=
8.1Hz, 1H, H-7'); 7.79 (s, 0.5H, N=CH); 7.80 (s, 0.5H, N=CH); 8.54 (d, J =
8.7Hz, 1H, H-10').
4.2.3. Preparation of dye 3c
Synthesis and purification was performed as per 3b except from 5c, furnishing
pure 3c as a bluish-white powder in 8% yield. HPLC: 99.4%area [RT
f
12.0min]. TLC(6:1): one component (R 0.52) only. Melting point: 85-86 °C .
1
H NMR: 0.96 (s, 9H, CMe
3
); 1.42 (s, 3H, >C(Me)CH
3
); 1.45 (s, 3H,
>
C(Me)CH ); 2.31 (s, 6H, Ar-CH ); 2.91 (s, 2H, N-CH ); 6.36 (s, 1H, H-7);
3
3
2
6.48 (s, 1H, H-5); 7.01 (d, J = 8.7Hz, 1H, H-5'); 7.37 (t, J = 7.5Hz, 1H, H-8');
7.55 (t, J = 7.5Hz, 1H, H-9'); 7.64 (d, J = 9.0Hz, 1H, H-6'); 7.73 (d, J = 8.1Hz,
1H, H-7'); 7.78 (s, 1H, N=CH); 8.55 (d, J = 8.4Hz, 1H, H-10').
4.2.4. Preparation of dye 3d
Synthesis from 5d in free base form was performed as per 3a but flash
chromatography (silica / 95:5v/v petroleum ether:ethyl acetate) was required
followed by repeated recrystallisation (toluene/methanol), giving 3d as a white
powder in 32% yield. Purity was 99.0%area by HPLC as two main
components (54.7%area of RT 9.4min and 44.3%area of RT 9.7min) that
f
were unresolved by TLC(9:1) which showed a single blue spot (R 0.52).
1
Melting point: 96-97 °C . H NMR: 0.85-0.99 (m, 12H, CMe
3
and >C(Me)-
); 1.40 (s, 1.5H, >C(Et)CH ); 1.73-1.97 (m,
); 2.83-2.97 (m, 2H, N-CH ); 6.62 (d, J = 8.1Hz, 0.5H, H-
CH
2
-CH
3
); 1.29 (s, 1.5H, >C(Et)CH
-CH
3
3
2
7
H, >C(Me)-CH
2
3
2
); 6.69 (d, J = 8.1Hz, 0.5H, H-7); 6.83-6.90 (m, 1H, H-5); 6.96-7.07 (m, 2H,
Page 24 of 33

ACCEPTED MANUSCRIPT
H-4 and H-5'); 7.13-7.21 (m, 1H, H-6); 7.35-7.42 (m, 1H, H-8'); 7.56 (t, J =
7.7Hz, 1H, H-9'); 7.66 (t, J = 8.9Hz, 1H, H-6'); 7.72-7.85 (m, 2H, H-7' and
N=CH); 8.54, 8.55 (two superimposed d, J = 8.4Hz, 1H, H-10').
4.2.5. Preparation of dye 3e
Preparation and purification was undertaken as per 3b except from 5e as
starting material. Crude material was obtained in 51% weight yield (HPLC
91.7%area of RT 12.7min) that, when purified, afforded 3e as a pale yellow-
orange powder of purity 98.4%area in 22% yield.
4.2.6. Preparation of dye 3f
Synthesis from 3-carboxymethyl-1-nitroso-2-naphthol and 5 (R = R' = Me) was
undertaken as per 3b, except that flash chromatographic purification of crude
product was not required. Instead, residual naphthol was readily removed
from the light yellow crude material (61% weight yield; HPLC purity 91%area)
by means of a single recrystallisation (toluene/methanol), giving pure 3f as an
off-white crystalline solid in 21% yield. HPLC: 99.6%area (RT 7.6min).
TLC(5:2): one component (R
f
0.51) only. Melting point: 118-119 °C . FTIR:
733cm (ester C=O). H NMR: 0.95 (s, 9H, CMe ); 1.39 (s, 3H,
C(Me)CH ); 1.42 (s, 3H, >C(Me)CH ); 2.89 (s, 2H, N-CH ); 3.63 (s, 3H,
-1
1
1
3
>
3
3
2
2 3
CO CH ); 6.64 (d, J = 7.9Hz, 1H, H-7); 6.86 (t, J = 7.4Hz, 1H, H-5); 7.05 (d, J
=
7.2Hz, 1H, H-4); 7.15 (t, J = 7.5Hz, 1H, H-6); 7.42 (t, J = 7.2Hz, 1H, H-8');
7.63 (t, J = 7.7Hz, 1H, H-9'); 7.79 (s, 1H, H-7' or N=CH); 7.82 (s, 1H, H-7' or
N=CH); 8.27 (s, 1H, H-6'); 8.54 (d, J = 8.1Hz, 1H, H-10'). A second crop of
Page 25 of 33

ACCEPTED MANUSCRIPT
impure product (94.9%area, 18% yield) was collected and used in the
preparation of 3h (see Section 2.2.8).
4.2.7. Preparation of dye 3g
A suspension of 3-carboxymethyl-1-nitroso-2-naphthol (1.25g, 5.4mmol), the
hydroiodide salt of 5d (1.90g, 5.1mmol) and methanol (12mL) was prepared.
Tripropylamine (1.07g, 7.3mmol) was added and the stirred mixture warmed
up to reflux over 15min and held for 3h, furnishing a bluish-green reaction
mass. TLC(5:2) indicated that no naphthol remained. The reaction mass was
allowed to stir down to room temperature overnight, producing a light grey-
brown suspension. The product was filtered off and washed with methanol
(10ml), affording a light greenish-yellow solid (1.39g, 60% weight yield). Two
incompletely resolved main peaks were observed by HPLC (total area 99.3%
centred around 9.8min). Recrystallisation (toluene/methanol) furnished 3g as
a pale greenish-yellow powder (1.24g, 53% yield) of HPLC purity 99.9%area
(combined area for peaks of RT 9.4min and 10.0min). Melting point: 114-
-
1
1
1
15 °C . FTIR: 1729cm (ester C=O). H NMR: 0.85-0.95 (m, 12H, CMe
C(Me)-CH -CH ); 1.35 (s, 1.5H, >C(Et)CH ); 1.40 (s, 1.5H, >C(Et)CH );
.75-2.00 (m, 2H, >C(Me)-CH -CH ); 2.80-2.94 (m, 2H, N-CH ); 3.60, 3.71 (s,
H, CO CH ); 6.58 (d, J = 7.8Hz, 0.5H, H-7); 6.65 (d, J = 7.8Hz, 0.5H, H-7);
3
and
>
2
3
3
3
1
3
6
6
7
2
3
2
2
3
.81-6.88 (m, 1H, H-5); 7.03 (d, J = 7.2Hz, 1H, H-4); 7.15 (t, J = 7.7Hz, 1H, H-
); 7.38-7.45 (m, 1H, H-8'); 7.62 (t, J = 7.7Hz, 1H, H-9'); 7.78-7.85 (m, 2H, H-
' and N=CH); 8.26, 8.30 (s, 1H, H-6'); 8.54 (d, J = 8.8Hz, 1H, H-10').
Page 26 of 33

ACCEPTED MANUSCRIPT
4.2.8. Preparation of dye 3h
A solution of 3f (0.40g, 95% purity, 0.90mmol) in anhydrous THF (5mL) was
cooled below 15 °C . A toluene solution of sodium bi s(2-
methoxyethoxy)aluminium hydride (0.53g at 70%, 1.8mmol) was diluted with
THF (2mL) immediately prior to addition over 40min at 10-20 °C . TLC(5:2)
indicated that around half the conversion had taken place by the end of the
addition, which remained the case after 20min stirring at 20 °C . A second
portion of sodium bis(2-methoxyethoxy)aluminium hydride (0.58g at 70%,
2.0mmol) in THF (2mL) was added over 30min at 20-25°C. TLC(5:2)
revealed that only a trace of starting material remained. After stirring for a
further 30min at 20-25 °C , ethyl acetate (5mL) was a dded cautiously below
25 °C , followed by a solution of ammonium chloride ( 0.25g) in water (2.5mL).
The quenched mixture was stirred for 30min and filtered to remove solid,
which was washed with ethyl acetate (3 x 5mL). The combined filtrate and
2 4
wash was dried (Na SO ) and the solvent removed to give an impure dark red
oil (0.7g, HPLC 63%area). Flash chromatography (silica, 90:10v/v petroleum
ether:ethyl acetate) followed by refrigeration under methanol afforded 3h as a
pale blue-grey powder (0.07g, 19% yield). HPLC: 97%area (RT 6.2min).
1
Melting point: 148-149 °C . H NMR: 0.97 (s, 9H, CMe ); 1.35 (s, 3H,
3
>
C(Me)CH
3
); 1.42 (s, 3H, >C(Me)CH
3
); 1.87 (t, J = 6.5Hz, 1H, OH); 2.92 (s,
); 6.67 (d, J = 7.8Hz, 1H, H-7);
2
6
1
H, N-CH ); 4.66 (t, J = 6.2Hz, 2H, Ar-CH
2
2
.89 (t, J = 7.0Hz, 1H, H-5); 7.06 (d, J = 7.3Hz, 1H, H-4); 7.17 (t, J = 7.8Hz,
H, H-6); 7.40 (t, J = 7.5Hz, 1H, H-8'); 7.56 (t, J = 7.7Hz, 1H, H-9'); 7.70 (s,
Page 27 of 33

ACCEPTED MANUSCRIPT
1H, H-6'); 7.75 (d, J = 8.1Hz, 1H, H-7'); 7.85 (s, 1H, N=CH); 8.52 (d, J =
8.4Hz, 1H, H-10').
4.2.9. Preparation of dye 4a
A suspension of 5b (2.09g, 8.1mmol), 4-(4'-diethylaminophenyl)-2-nitroso-1-
naphthol (2.66g, 8.3mmol) [7] and methanol (26mL) was stirred and heated to
reflux for 64h. TLC(9:1) indicated no Fischer’s base derivative remained. The
resultant olive-grey suspension was stirred down to room temperature. The
solid was filtered off, washed with methanol (6 x 5mL) and air-dried to give a
khaki-coloured material (3.86g, 85% weight yield). As expected, HPLC
revealed two main components (44.4%area of RT 26.8min and 50.1%area of
RT 27.8min) that were unresolved by TLC(9:1). Repeated recrystallisation
(toluene/methanol) gave 4a as a pale pea-green powder, 2.18g (48% yield).
HPLC: 98%area with two main peaks in ratio 50:47 that were unresolved by
TLC(6:1) which showed a single blue-green spot (R 0.69). Melting point:
f
1
2
7
1
20-221 °C . H NMR: 0.95 (s, 4.5H, CMe
3
); 0.98 (s, 4.5H, CMe
); 1.38 (s, 3H, >C(Me)CH ); 1.50 (s, 1.5H, >C(Me)CH
.52 (s, 1.5H, >C(Me)CH ); 2.25-2.28 (m, 6H, Ar-CH ); 2.86-3.03 (m, 2H, N-
); 3.43 (q, J = 7.0Hz, 4H, N-CH CH ); 6.44-6.48 (m, 1H, H-7); 6.78-6.80
3
); 1.23 (t, J =
.1Hz, 6H, N-CH
2
CH
3
3
3
);
3
3
CH
2
2
3
(m, 2H, Ar-H); 6.84 (s, 0.5H, H-4); 6.98 (d, J = 8.1Hz, 0.5H, H-6); 7.34-7.41
(m, 4H, Ar-H); 7.45 (s, 1H, H-10'); 7.68 (br s, 1H, N=CH); 7.95-7.98 (m, 1H,
Ar-H); 8.05-8.08 (m, 1H, Ar-H).
4.2.10. Preparation of dye 4b
Page 28 of 33

ACCEPTED MANUSCRIPT
Product was obtained in the same manner as 4a except using 5d in free base
form. Crude material was obtained as a green powder (69% weight yield;
HPLC 94%area) which when purified gave 4b as a pale lime-green powder in
33% yield. HPLC: 98%area (RT 12.6min). TLC(5:2): one component (R
f
1
0
.50) only. Melting point: 156-157 °C . H NMR: 0.87-0.98 (m, 12H, CMe3
and >C(Me)-CH
C(Et)CH ); 1.77-2.07 (m, 2H, >C(Me)-CH
.43 (q, J = 7.0Hz, 4H, N-CH CH ); 6.63 (d, J = 7.9Hz, 0.5H, H-7); 6.71 (d, J =
.9Hz, 0.5H, H-7); 6.78-6.80 (m, 2H, Ar-H); 6.84-6.92 (m, 1H, H-5); 7.06 (d, J
7.3Hz, 1H, H-4); 7.15-7.23 (m, 1H, H-6); 7.31-7.43 (m, 4H, Ar-H); 7.46 (s,
2
-CH
3
); 1.23 (t, J = 7.0Hz, 6H, N-CH
2
CH
3
); 1.36, 1.39 (s, 3H,
>
3
2
-CH ); 2.90-3.06 (m, 2H, N-CH
3
2
);
3
7
2
3
=
1H, H-10'); 7.62, 7.79 (s, 1H, N=CH); 7.94-8.09 (m, 2H, Ar-H).
4.2.11. Preparation of dye 4c
Synthesis was performed as per 4a except using 5e to give crude product as
an olive green solid in 73% weight yield. Repeated recrystallisation
(THF/methanol) furnished 4c as a cream-coloured powder in 41% overall
1
yield. HPLC: 98%area (RT 16.1min). Melting point: 194-195 °C . H NMR:
0
.97 (s, 9H, CMe
(CH -); 2.88-3.04 (m, 2H, N-CH
d, J = 7.8Hz, 1H, H-7); 6.78-6.80 (m, 2H, Ar-H); 6.87 (t, J = 7.2Hz, 1H, H-5);
.19 (t, J = 7.7Hz, 1H, H-6); 7.34-7.41 (m, 5H, Ar-H); 7.45 (s, 1H, H-10'); 7.72
br s, 1H, N=CH); 7.96-8.05 (m, 2H, Ar-H).
3
); 1.23 (t, J = 7.0Hz, 6H, N-CH
2
CH
3
); 1.30-1.95 (m, 10H,
-
(
2
)
5
2
); 3.43 (q, J = 7.0Hz, 4H, N-CH
2
CH ); 6.69
3
7
(
4.3 Determination of colour and photochromic properties
Page 29 of 33