Photochemical Reactions of Azo Dyes in Solution

with Different Substrates

H. C. A . VAN BEPKand P. M. HEERTJES
Siilutioiis of azo dyes iii water eoritainirig different soluk)hbsubstrates hnvo I)O(?II exposed to ultraviulrt
aid visible radiation. It, has h ~ e n
found t h a t ma11y com1)oululs rract photochomirally 11s hyt1rogt.n ilonors
atid causc retlurt,ion of azo t l y c ~prtwmt in thc s a i n e solution. Iq’adiiig of such xolutioiie takes ))lace only
whnu tho radiation is alxm-bctl by the s u M r a t e . I‘acling is ovitlontly tht> result of iutoractioii of cxritrtl
substrate rnolncules with dyc irioleculos or cxcit,(d tljw, inoleculw. Siibstrates containing carl)oxyl,
hytlroxyl, nr aryl groups attached to methyl, nicthyloiio, 01. aliphatic ->CH- groups show high photo-
t*hciniral activity when the rompouutls tthsorl) light of lotig wavoloiigth.
Kinet,ic iiieasurciiients h a w shown t h a t t,hc failing rcactioiis arc of first ortlcr with rrspwt to the
concontration of t,he excited s u h t r a t e Inolrculw at low whst,rrttc+coriceiitrat.iona. Failing of ttzo clycs i n
solutions containing different suhstratcs is partly iiihihit,e(l by the prwwirc of oxygrii. ‘rho results iritlicato
thtit, oxygc.n ran act as a hytlrog~~ii nccyitor in t h r sarii(~way as tho azo tlyw. Kinetic cxprrimoits carriccl
wit for oiii? systrm suggest t h a t t,hta inhibition is Ilartly rausocl hy reoxitlation of rrtluct.rl t l y l rutlic*alxto
the original (lye hy peroxitle radicals. A mrcha~iisnifor t h e inhihition i s proposed.

Introduction forniation of benz(c)cinnolinc by photolgsis of

Thc fastness t o light of dyeings on tliffercnt azobenzeric 12.
fibres is a matter of great importance. Many Thus the results of investigations of the photo-
investigations have been carriccl out on the natrircx chemical decomposition of azo dyrs do not itidiciitc.
of the fading of dyes on fibrous or other substratcs, iineqiiivocall~~wlic>ther oxidative or reductive
r.g. paints and plastics, and the subject has degradation prcdominates. To try t o settlc this
recently I)ecn reviewed 1, 2 . Investigations on the point, we havc investigatctl both the primary
f d i n g of dyrs in solution. however, are rare. reactions and the ovrrall reactions occurring on
From several studies 31 it is apparent t h a t the irradiation of solutions of azo dyes. I n this papcr
interaction of dye and substrate influences the we describe the isolation of reaction products
rate of fading of dyes. This is demonstrated by tlir formed on irradiation in absen’ce of oxygen, t h e
very low light fastness of basic dges on rotton and kinetics of tlic facling reaction in presence and in
wool and thcir moderate to high fastness on absence of oxygen, and the rmctivities of different
Orlon 5 . substrates; nieasiircmetits of the wavelengths of
Investigations by Wells on the photosensitised light rcsponsible for fading are also ricwribetl. The
oxidation of aliphatic alcohols, ketoncs. and nntrire of the pritnary act iind of the inhibition of
cthcrs by oxygen in the presence of sodium fading by oxygen is discnsscd.
anthraqriinone-B-sulphonatc indicate that the
substrates used show a wide range of rcactivitic General Aspects of Fading of Dye Solutions
Very little is Iwnvn of thc nature of the decomposi- Most of the cxperimcnts were carried out on
tion of WLO dyes by light. Accorcling to Rlaisdell 7 , C.I. Acid R t d 12 (preparcd in o w Ial)oratory). It
azo dyes are reduced to the corresponding aniines is likcly that, with azo dycs containing no sub-
when their solutions in organic solvcnts are stitucnts other tliiin liydroxyl nnd slitphonic acid
irradiated in absence of oxygen. groups, only reactions involving the azo link 1% ill
Rchac+kr * showed that irradiation in presence occur. Fading was measured by means of a
of oxygen of a dyeing on cotton of a direct dye Unicam SP 500 s ~ ~ ~ ~ c t r o p l i o t o ~ nFading
eter. v as
tlcrived froin benzidine led t o forniatioii of effected in quartz vcxsels by irradiation I\ ith light
I~cnzidine. No other fission prodiicts could be from a high-pressnrc mercury-vapour lamp ( P h i l i p
isolated. As these would be rither o-diamines or HPK 125-W). Thc crnission spectrnni of this Iitmp
o-anrinophenols this is t o bc expectcd, lwcaiis(> (continuum and lines) is shonn i n Fig. 1.
thcsc compounds arc unstable in presence of The abqorption spcbctrritn of tlic (lye docs not
oxygen. Indications of the presence of 1.2- changr on irradiation. Cis-trans isonierisnr docs
naplitlroquiiiones, M liich could result froin the not occur; because this dye i an o-liytlroxyazo
oxidation of 1,2-diarninoi~aplitli~~leiieor o-amino- derivative, the trans-structure stabilised by the
naphthol, werc obtained 1)y Halter a n d Ziersch o, lo formation of an internal lrytlrogcn bond. Afore-
in tlie products of tlie fatling of the same type of over, formation of thc cis-structure is iinpossiMt.
dyw. brcause of steric Iiindrancc.
In addition to the reductive fission of azo clyrs, It was found that tlie solvent had a marked
oxidative attack has sometimes 1)ceri found t o occur. influence upon t h e fatling rate of tlic dye. Fading
For N-arylaminoazo dycs, Krollpfriffcr l1 showed in water w i ~ svery slow, hit the ratc of fading was
that triazoliiim and naplitliopheiiazinc drrivat iws incretiscd by dissolving various organic substances
ttre formed on irradiation in oxygen. Another in the dye solutions. Increastd fatling has been
example of photochemical dehydrogenation is thc found to occur in presence of methanol, acetic acid,
 662 V A N DEEK AND HEERTJK- PHOTO(’HEMTCA1, REACTIONS OF AZO IIYES I N SOLITTION .I \ I) c’ 79

I00 manclelic acid, the rate as negligibly small when

the rcaction was carried out in glnss vcssc.ls instcad
80 of iii quartz vtmcblw. Radiatioii of w avrlciigt I i Iws
aQ than 280 inp was nccesqitry to induce rapid fading.
i
? 60
c
0
Primary Reaction Products of Fading in
u Absence of Oxygen
? 40
-du Solutions of the dyes I and 11 in methanol nere
L* faclcd in absence of oxygcn; c.xpriincrital tlctnilv
20 arc dcscribed latcr (sec p. T25). T h e dyes \ww
chosen because thcir probable reduction products
0
250 300 350 400 450 500 550 600
are fairly stable. The products can be srlcctivc.ly
Wavelength, r n G
convcrtcd into othcr azo dyes by simple diazotiqa-
tion and coupling. Such reactions facilitate
1’10. 1
purification by chromatography and identificcxtion
of the reaction prodrictti. The faded solutions of
acetone, succinic acid, malonic acid, tartaric acid, t h c dyes were concentrated by distilling off the
and niandclic acid. The reactivity of C-H bonds is solvent. Tho residues wcre then analysed.
increased by the prcscncc
’- ,
OH 0H
of hydroxyl groups bniintl
> -<
\ /- ,
to the same carbon atorn, N c t Q 3 S4
as has been pointed out by
Wells 8. From the present
1-7= -N - N - C o

\-Y’
/<
I
-+

h 0 NIL
NdLS NH.! -t H2N I J

Q-// 1
SOJStL
work it appears that the C I Arid llcrl 12 ( I )
reactivity of compoundr
towards azo dyes is also
increased by t h e presence
carbonyl and phenyl groups
in the a-position t o the C-H
of N~.O?S~-\N-N

I
.-/ -+ (71
NkxOJs-<~>NHI i H 2 N ,

\/\I
OH
I

\\/
1
\
y’,
group of the substrate. SOINiL SOJNd
(11)
Aromatic compounds with-
out -CH,, -CH,-, or >CH- groups in an aliphatic 1)iazotiscd siil[)hanilic acid \\as adtled to the
side-chain (e.g. phthalic twid and phloroglucinol) rcmtion products of C. 1. Acitl Rod 12. An nzo t l y i
do not affect fading. The rcnction rates were not was forruecl whicli was idrnticd wit Ii that preparty1
influenced by addition of mineral acid. Moreover, from diazotisd sulphanilic acid and naphthionic.
coriversion of the organic acids into their mettxl acid. In anothcr portfor1 of tllc rcaction products,
salts was accompanied by ;ilmost, complete loss of the 2-aiiiirio-l-nnj~litl~ol-6-sulplinriic acid r x p c t d
reactivity. to be formed was convrrted into a dye by cliazo-
Oxygen-free solutions faded vcry rapidly, but tisation in neutrttl incdium, u h i g cuprica sulphatr
solutions of azo dyes containing tlic substrates as catalyst, and coiipling M itli resorcinol.
mentioned above faded slowly in prescncc of Fading productn of dyv I1 (s111ph~nilic acid-
oxygen. During fading oxygen is consumed and Nevile ant1 Winther’s (NW) acid) were similtxrly
faded solutions give a positive test with thcb converted into neu dycs, by tliazotisation in acid
potassium iodide-starch reagent, indicating that medium and coLqhig with NW-acid ant1 by
hydroperoxides are present. Similar effects have diazotination in neutral mcdiriin in prescnce of
been observed by Blaisdcll in the fading of cupric ions and coupling with rcsorcinol. The t1yc.s
derivatives of azobenzene 7 . obtained \\ c w piiritictl by I’ap(~r-cliromatogrixI,Ily.
An important observation is t h a t the wave- ‘rhc>y were itkwtifird from tlicir R,-values and
length of the irradiating light has a marlred effect absork)tion spcctra, which w r c . identical witti
U p o I i the facling rates. For all the solvcnts except thosc of compounds synthcsiscd from products of

Ta1il,lq: I
Azo dye in methanol Dorivativos of ruaction prorlurts

C.I. Acid Rcd 12 (I) Sulphanilic acid --f fading product 0.14
Sulphanilic acid -> naphthionic acid 0.14
Fatling produr:t + rosorcinolt 0.33
2-Amino-1 -1iaphthol-5-sulphonic 0-33
acid + rcsorcinolt
I k l i n p product + N W acid F
Sulphanilic acid + NW acid f
Sulphttnilic acid -> Fading product + rosorcinolt B
NW-acid (TI) 2-Aniiiio-l-na~~htho1-4-sulphonic
acid -+ I?
rosorcinolt
Snlvcnt- 0 . 5 % c t l i a n o l ; ~ ~ ~in~ iwntcr.
i~e
t Copper complcx
 D.c. IwJ V.4N BEEK AND HEERT.JES-
PHOTOCHEMICAL REACTIONS OF AZO DYES IN SOLUTION 883

I I I I I I I , I I I
250 300 350 400 450 500 550 600 250 300 350 400 450 500 550 600
Wavelength, m p Wavelength, m p
----
-- b
Fro. 2 - Idcritilleatioii of fading prdiic'tx ul DYPI Irnin absorption spectra

4-8

4.6

4.4

I; -
-0"
4.2

d 4.0

3.8

32 t 3.6

250 300 350 400 450 500 550 250 300 350 400 450 500 550
Wavelength, m p Wavelength. m p
__-_ ____
f - h

11'10. 3- Identiflcatlon of fading products of Dyo I[ from absorption spectra

known constitution. The R,-values arc given in then placed in a Littrov spectrograph (Fig. 4).
Table I, and the absorption spectra in Fig. 2 and 3. The spectrograph contains a cilsc (C) with a yiiurtz
From the results obtained it appears that the window and tubcs through whidi nitrogen can bc
azo dyes are reduced to the corresponding ainiries passed. Light from a mercury-vapour lanip
on fading. (Philips HPK 125-W) is clispcrsctl by the spectro-
graph arid focuscd on the paper strip (P). Recausc
Wavelength of Photochemically Active Light the paper is sensitive to light, fading will occur.
I n this wag a picture of t h r emission spc~ctrumis
I<nowledge of the wavelengths of the photo-
obtained in the range of wavclrngths that are
c!iemically active light is important for exact active.
niensiirrmrnts of fading rates, especially when
lamps that emit polychromatic light are used.
The cffective wavelengths were measured as
follows. Spectrograph
Strips of paper (Whatman No. 1) werr soaked
111 aqueous solutions of different acid azo dyes and l"ll+. 4
suitable non-volatile substrates. After the dye h i d
1)ec.n absorbed on the p p e r , thcl strips wrre Ccllulose also is activc. as n substratc in thc fading
rcmoved from the solution and most of the adhering process, but, it is loss activcl than tlic other sub-
liquid was removed by pressing between filter strates. Ncvcrtheless, the lines ~)roducetlon thc
paper. The strips were dried in a carrent of air and paper strips mrist I)r regardcd as the result of'
 664 VAN 13i.:~:A PHC)TOCHEMl(!AI, 1ZEACTIONS OF AZO 1lYES I N SOLl!'L'ION .I.S.I).C.79
~ N I ) HKP:HT.JKS-

TAI1LT: I1
])YO Structure Mnxiinuiii wavttltmgtli of w t i v e light ( n i p )
for wlutioiis corit,airiing
maloiiic ~ n w i r i i r tart,arin ptioiiylaretic maiitIdic
ni,i(l ariil acid acid acid
,-

(2.1. Acid Itcil 12

<-,
, \
\

\
/. H (?
I
' '
"\ I '75s 2ti5 205 313 9ti5
(Prcparcd by a u t h o r s ) NllOaSi<- /''-N=N
\'V
SO3Nt&

313 3ti5

25s 436

-
2.53 265 265 3I 3 3fi5

fatling rcactions of both t h e substrate i d t h c

cc~llulose.
From the measiirmienth it w a s clcar t h a t t h r
maximum wavelength of tlie photochemical1,y
activc light is the same for t h e five different azo dyw I t
4
in prcscncc of any one snbstrate. This is true also
for otlicr t m i dyes riot discnssrd in this p a ~ ~ c Tlic.
r.
important conclusion can t hcrcforcb be drawn that
the. wavelengths of the photochemically activc
light arc d e t e r m i n d by t h e substrate and not Iiy
thc. dyes. Tthle I1 siiminitriscs tlie results obtainrcl.
Thc results indicatc that fading of azo dycs is
clue to interaction of excited substrate moleculw
with unexcitctl or exc,itcd dye molrculcs. 220 230 240 250 260 270 280
Excitation of the substratc molecules is caused Wavelength, in+

by absorption of light by the substrate, as was dye

shown by the following experiment with (2.1. Acitl --_-_-- acetic acid

Red 14. Fading experiments were carried out in ------_ malonic acid
--_-- - wccinlc acid
aqueous solutions of the dye containing each of the tartaric acid
substrates in turn in the absence of oxygen. The - - - - - phenylacetic acld
solutions fade rapidly M hen irradiated ~ i t hlight -m-m-D-m-m- mandelic acid

from thc mercury vapour lump. However, fitding 1 10 5- .4bsorptiou sprctl"~of siibztrater studied
is inhibitrd if a quartz ciivettc, filled with a
conccntratctl aqLicoiis solution of the same reaction a ith the substrate as a rrsiilt of r c d u a l
substrate, is placcd between the lamp and the clyr absorbance may alro play a part.
solution. The wavelengtlih of the active light arc,
thcrcforc, in the first instance restricted to the Kinetics of Fading of C.I. Acid Red 12
region in which t h e suhstmte ahsorbs light. The in Absence of Oxygen
absorption spcctra of the wbstratcs are shown in Fading a a s carried out in the apparatur qhou 11
Fig. 5 . in Fig. 6. Dye solutionr \ v c w irradiated in the
In onc or two installers in T a l k 11, light of the quartz cuvcttc (V) 11 ith light from the high-prcssnrc
maximum wavelength is not absorbed, or is onlp mrlcriry-vspoiir Innip ( I , ) (Philips H PI< 125-W )
slightly absorbctl, by the siibhtrate. This may placed 16 cm away from the cuvrttc I'iirifirrl
mean that photochemical rcvmtion bctween crllnlose nitrogen (t\fo \fashing liottleb containing alkalinr
a n d the clyc occurs, although photochcniical sodium tlithionite bolutionr aiid tn o wasliing
 DPC. 1963 VAN BEEE AND HEEXTJE-PHOTOCHEMICAL REACTIONS OF AZO DYES IN SOLUTION 666

bottles containing alkaline pyrogallol) was passed Atherton and Seltzer 3 . Fluctuations of the light
through the dye solution. During the experiment output of the lamp in the range 200-280 m p were
the temperature of the solution was kept a t 25°C thereby accounted for. This arrangement enabled
with the aid of ventilators and by thermostatic reproducible measurements of fading rate t o be
adjustment of the room temperature. obtained for the reaction of C.I. Acid Red 12 with
acetic, malonic, succinic, and tartaric acids.
During the experiments samples could be taken
B from the cuvette (V) via the valve (B) by reversing
the flow of gas. The dye concc~ntration was
measurcd spectrophotoinetrically with a Unicam
Sip 500 spectrophotometer at the wavelength of

El maximuni absorption of the dye (522mp). The

amounts of dye faded were calculatcd in arbitrary
units (D), which are related to the concentration
(C) of the dye in solution by the equation
FIQ. 6 u = v log (I,/I)= EC'V
where C = 4.261 x lW510g (Zo/I)nimole/ml, log Z,/I
The intensity and the spectral composition of the was measured in a 1-cm cuvette, E is the extinction
light depend on the temperature of the lamp. cocflicient a t 522 m p and v is the cuvette volume
A measure of the intensity of the photochemically 65.0 ml). I n the calculation of fading rates,
active light was obtained by passing part of the corrections were made for the samples taken from
light emittcd by the lamp through a mono- the solutions. The fading rates havc bcen expressed
chromator (M). Behind this monochromator a as amount of dye faded per 100 or 1000 (for the
photocell with slides was mounted in such a way acetic acid system) counts (n)on the light in-
that only light of wavelengths 200-280 m p fell on tegrator. The amount of fading is lincarly related
the photocell. The flux of the light was counted on to dye concentration, so the values given can be
a light-flux-time integrator, as described by used in the study of the kinetics of the reaction.

0 160 200 300 400

Acetic acid concn.. mmolell. Malonic acid concn., mmoleil
(0) (b)

25

20

1
0 50 I00 I50 200 0 50 loo 15
Succinic acid concn.. mmole/l. Tartaric acid concn., rnmole/l.
(C) (4
FIa. 7- Initial rates of fading of dye solutions (concn. c&. 0.1 mmoldl.) in presence of (a)acctie aci& ( b ) malonic add,
(e) succinic acid, and (ct) tnrtnric acid

F2
 868 VAN BEERAND HEERTJES-PHOTOCHEMICAL REACTIONS O F AZO DYES IN SOLUTION J.8.D.C. 79

60 - @
MC
/a

20

L 1 1 I

0 I 2 3 4 5 6 7 8 9 1 0 0 2 4 6 8 10 12 14 I6 18 20

(4 (d)
FIO.8- Rclation hetween rate of fading and substrate concentrntion for solutions of C.I. Acid Red 12 (concn. ca. 0.1 ininole/l.) in
prcxence of ( a )acetic: acid. ( b ) malonlc acid, (c) succinic acid. nnd ( d ) tartaric acid

In the following graphs, initial fading rates have (where Q = dye, HX = substrate, HX* = excited
been plotted; these were independent of dye con- substrate, and QH, = hydrazo compound).
centration for solutions of high optical density. The irradiated solutions had a high optical
Plots of fading rate against substrate concentration density for photochemically active light, as can be
are given in Fig. 7 . seen from the absorption spectra in Fig. 5, so that
Fig. 7 shows that, as the substrate concentration all the incident photochemically active light was
increases, the fading rate first increases and then absorbed. The fraction absorbed by the substrate
becomes constant. Because fading is initiated by (F,) is given by-
interaction of activated substrate molecules with
dye molecules, the following sequence of reactions
will be assumed to occur-
k where cg and Cd are thc averagc extinction
HX + hv 1 3 HX* . . . . (1) coefficients of substrate and dye, respectively.
HX* A+-
k HX . . . . (3) From the reaction scheme, and assuming that
the rate of fading is determined by the fraction of
HX* + HX >-+
k
2HX . . . . (3) the light absorbed by the substrate, i t can be
HX*
QH.
+
+ HX
Q
k
A+QH.
J-+
k
$- X.

QH, -1- X.
. . . . (4)
. . , . (6)
shown that the rate of fading
be expressed as-
(- %I) can
 DW. 1 9 6 ~VAN BEEKAND HEERTJES-PHOTOCHEMICAL REACTIONS O F AZO DYES I N SOLUTION 667

microammeter. With this electrode, a linear

relation between oxygen concentration and the
measured current was obtained.
The reaction temperature was slightly higher
(26-28'C) than in the experiments carried out in
absence of oxygen, because of the longer times of
irradiation necessary.
-1 The variation in initial fading rate with oxygen
It follows that, for constant I , by plotting -~
d[Ql /dt concentration is shown in Fig. 9 . Even small
as a function of l/[HX], a non-orthogonal amounts of oxygen depress the fading rate appreci-
hyperbola would result. At low substrate con- ably. The reciprocal of the fading rate was found
centrations, k,[HX]<k, + k4[Q]. To a first to vary linearly with the square root of the oxygen
1 concentr at'ion.
approximation, a linear relation between ___
dCQlldt
and l/[HX] is then obtained. It also follows, from
Eqn. (ii) and (iii), that at higher substrate con-
centrations -d[Q]/dt tends to a maximum.
-1
In Fig. 8, the measured values of ___ have
d[Ql idn
been plotted against l/[HX] for the different sub-
strates used. The experimental results for tartaric,
malonic, and succinic acids a t low concentration8
are in agreement with the proposed scheme. At
-1
d,Q,/dn should pass
high substrate concentrations, -~
through a minimum, as predicted from Eqn. 8,
'
0
L 1 2 3 4 5 6
but measurements a t extremely high substrate Square root of oxygen concn , (mgil )f

concentrations have not been made. The results 1m. 0

given in Fig. 8 do, however, show the expected
deviations a t low values of l/[HX]. The results I n another srries of experiments, the relation
obtained with acetic acid show deviations that may between fading rate and initial dye concentration
be due to the existence of side reactions. The was studied, the substrate concentration being
reactivity of acetic acid is much lower than that of kept constant a t 150mmole/l. The solution was
the other acids, for which side reactions, if any kept saturated with air, in order to obtain a
occur, cannot be detected. constant oxygen concentration. The results are
It can be concluded that the reaction scheme given in Fig. 10, in which thc fading rate is shown
proposed can be used to describe the reactions that to increase with initial dye concentration.
occur when azo dyes are faded in absence of The results so far described can be explained by
oxygen. More accurate measurements of fading extension of the series of reactions already given
reactions, preferably initiated by monochromatic in the following way.
light, are, however, necessary to enable the +
mechanism to be established quantitatively.
HX* k'l + HO,.+X
0, ~
. . . . (6)
k
Fading of C.I. Acid Red 12 in
QH. 4- HO,. l + Q+H,O, . . . . (7)
k
Presence of Oxygen 2H0,. --1_-+ H,O,+O, . . . . (8)
Fading was carried out in the same way as k
before. Succinic acid was used as substrate.
2X. &+ end products . . . . (9)
Fading rates were measured for systems containing Sincc it has been found that the fading rate is
dye a t a constant initial concentration, a constant inversely proportional to [O,]', it must be con-
substrate concentration of 150 mmole/l., and cluded that peroxy radicals are involved. The rate
differing oxygen concentrations. of formation of HO,. radicals is linearly dependent
Oxygen concentration was varied by circulating on oxygen concentration, but thcir deactivation
mixtures of oxygen and nitrogen through the may procecd predominantly by reaction (8)(k,>k,).
irradiated solutions via a membrane pump and a The stationary-state concentration of these radicals
buffer vessel containing a gas mixture of known will then be a function of [o,]', i.e.
composition. The concentration of oxygen in
solution was calculated from the composition of
the gas mixture by means of Henry's Law. To
[HO,.] =
k,
* LO,] . . . . (iv)
control the oxygen concentration during an If i t is assumed that reaction ( 5 ) i8 rate-deter-
experiment, an oxygen electrode was mounted in mining for fading, a relation betwrcn rate of
the vessel containing the irradiated solution. The fading and oxygen concentration can be derived,
electrode was constructed as suggested by Butler 13, viz.--
but using glass instead of the epoxy resin. The
cathode was a spiral and the anode a silver ring.
The electrodes were connected to a battery giving
0.7 V. Readings of current were made on a
 668 VAN BEEKAND HEERTJES-PHOTOCHEMICAL REACTIONS OF AZO DYES IN SOLUTION J.S.D.C. 79

TABLEI11
Exporiment Composition of
solution*
Malonic Initial Malonic Counts
E acid in [O,.] acid n
(ingil.)
0 0 p/ water
(9)
concn.
(gmole/l.)
(x

0 I 96.73 31.50 3.1,

6 0 I1 97.84 17.84 1.3
0 e / I11 9747 5.69 0.5
// O IV 97.83 24.44 2.1,
0’ O
V 97.55 19.90 1.6,
4 VI 97.86 4.81 0.4
VII 97.96 13.90 0.123 1.1
VIII 97.81 10.88 1.2
IX 97.79 10.10 1.0,
2 x 97.37 7.69 0.7,
XI 97.72 32.82 3.0,
XI1 97.53 7.24 0.7

For dye concentrations flee Fig. 11 (purity of the dye 38%).

Equation (v) can explain the experimcntally

determined relation between the fading rate
measured in presence of various amounts of oxygen
and in absence of oxygen, if oxygen has no appreci-
able influence upon the stationary-state con-
centration of the excited substrate molecules.
The concentration of excited substrate molecules
1s- L
0 1000 2000 3000
Number of counts (n)

. . . . (Vi) lwi, 11

Estimation of the extent t o which this condition is

satisfied is not yet possible. Quantum efficiency
measurements of fading reactions initiated by
monochromatic light are necessary t o dctermine
whether the proposed mechanism is, in fact,
operative. Moreover, similar investigations would
have to be carried out on other active substrates.
The mechanism also explains the relation between
the dye concentration and the rate of fading.
An increase in rate of fading with increasing dye
concentration in presence of oxygen would follow
from equation (w).
Fading can also be carried out in a different way.
If dye solutions having a constant substrate
concentration and varying initial oxygen con-
centrations are irradiated without introduction of
fresh oxygen, the oxygen concentration will
decrease during fading. When the oxygen originally
present is nearly exhausted, there will be a sharp
increase in the fading rate. This was demonstrated
by the following experiments. Solutions having
different initial oxygen concentrations were faded
in the apparatus previously described. The
substrate was malonic acid (123 mmole/l.). The
compositions of the solutions are shown in Table
111. The dye concentration was measured spectro-
photometrically and then plotted as a function of
the number of counts ( n ) registcred by the liglit- Number 0 1 counts (n)

flux-time integrator (Fig. 11). FIG. 12

 Dre. 1963 VAN BLEK AND HEERTJES-PHOTOCHEMICAL REACTIONS OF AZO DYES I N SOLUTION 669

It can be seen that the dye concentration is caused by reoxidation, by peroxide radicals, of
decreases only slowly a t the start of each experi- partially reduced dye radicals to the original dye.
ment. After a certain amount of irradiation has Measurement of the wavelengths of light that is
occurred, there is a sharp increase in rate of photochemically active has shown that, to deter-
fading. Ultimately, a high and constant rate is mine the reactivities of different substrates, it is not
obtained, which is identical with that in absence sufficient t o measure rates of fading by mono-
of oxygen. The number of counts ( n ) (see Table chromatic or polychromatic light, but that
111) corresponding to the intersection of the quantum efficiencies must be measured as a
extrapolated lines (dotted lines, Fig. 11) have been function of the wavelength of the radiation.
plotted in Fig. 12 against the amount of oxygen It is to be expected that the fading of azo dyes on
originally present in the solution. Apparently, a n solid substrates, e.g. fibres, will follow the same
increase in the initial oxygen concentration results general course as fading in solution. To ensure
in an increase of the amount of irradiation needed that dyeings have high light fastness, the fibres
to induce rapid fading, which indicates that the should not contain substances that are active as
sharp increase in the rate of fading is due to the hydrogen donors in the sense described in this
fact that all the oxygen originally present has been paper. Free access of oxygen should be permitted,
consumed. in order to retard fading.
It can be concluded that equation (v),for the
* * *
rate of fading of dye solutions containing high We thank Mrs. T. M. Brummer-Einthoven for
concentrations of Substrates, explains very well the assistance with the experimental work.
fading of oxygen-containing solutions in closed LABORATORY OB CKEMICALTEOHNOLOOY
vessels. TF:CANICAL UNIVERSITY
DELBT
HOLLAND
Conclusions (MS.received 8th April 1965)
From the results obtained i t can be concluded References
that the primary reaction occurring in the fading, 1 Zollinger, H., “Azo and Diazo Chemistry” (Interscience
in absence of oxygen, of azo dyes in aqueous Publishers, New York & London, I961), Chap. 12.
solutions containing various substrates is 9 van Beek. H. C. A., The&, Delft, 1960.

abstraction of hydrogen from the excited substrate 3 Atherton, E., and Peters, R. H., J.s.D.c..68, 64 (1952);
Atherton, E., and Seltzer, E., ibid., 65. 629 (1949).
by the dye molecule. The stable end-products in 4 Cuniming, J. W., Giles, C. H., and McEachran, A. E.,
this reduction process are the corrcsponding ibid., 72, 373 (1956).
aniines, which have been isolated and identified. Wegmann, J., Melliand Textilber., 39, 408 (1958).
The primary reaction of the dye with the Wells, C. F., Trana. ParadaySoc., 57, 1703, 1718 (1961).
7 Blaisdell, B. E., J.s.D.c., 65, 618 (1949).
substrate in presencc of oxygen is the same * Schaeffer, A.. S . V.P. Fachargun, 11. 106 (1956).
hydrogen-abstraction process. This follows from O Haller, R . , and Ziersch, G., Melliand Textilber., 10, 951
the following observations- (1929).
lo Idem, Angew. Chem., 43, 209 (1930).
(1) Substrates that increase the rate of 11 Krollpfeiffer, F., Miihlhausen, C., and Wolf, G., ,4nn.,
fading in absence of oxygen also increase i t in 508, 39 (1934).
presence of oxygen.
. - la Lewis, G . E., Tetrahedron Lettera, 9, 12 (1960);
Hugelshofer, P., Kalvoda, J., and Schaffner, K.,
(2) The rate of fading in absence of oxygen Helv. Chim. Acta, 43, 1328 (1960).
can be related to the rate in presence of various Butler, R. A., Nunn, J. F., and Askill, S., Nature, 196.
amounts of oxygen by recognising that inhibition 781 (1962).

Discussion
Dr. W. INOAMELLS: Has the author considered causing fading, even for fading reactions carried
the possible effects of viscosity on the mode of out in presence of oxygen. I n the investigations
decomposition of the dye! concerned with measurements of the Bccquerel
It has been demonstrated (P. J. Hillson and effect for azo dyes (Hillson and Rideal), such
E. Rideal, Proc. Roy. Soc., A 216, 468 (1953) ) that substrates were not present. Since light of wave-
both oxidation and reduction of azo dyes can take length longer than 300 mp was used, the collodion
place during illumination, the predominat’ing and palmitic acid used to obtain homogeneous
reaction depending upon the experimental layers on the electrode were presumably not
conditions. It may be, therefore, that on a rigid activated.
medium, freely accessible to oxygen (e.g. fibres), It has been made clear that the mechanisms of
the lifetime of the photochemically excited dye fading reactions of systems containing dye,
molecule may be lengthened and diffusion- substrate, and oxygen are not yet completely
controlled oxidation reactions would become more elucidated. I n this connection, the influence of
important on those fibres less able to donate oxygen in absence of substrates with high
hydrogen to the dye. hydrogen-donor activities has yet to be determined.
Dr. VAN BEEK: Comparison of the present Theoretically, the possibility exists that other
investigations with those of Hillson and Rideal photochemical reactions than those described in
is, in our opinion, not possible, because the systems this paper will be predominant in such a system.
studied are entirely different. In the present work With regard to the effect of the viscosity of the
j t has been shown with certainty that activated solvent (or fibre), it can be stated that, if more
substrate molecules are the active intermediates than one type of reaction can occur, the ratio of
 670 V A N BWEXA N D HICERTJE-
PHOTOCHEMICAL REACTIONS OF AZO DYES IN S O r m m N J.R.D.O. 79

the reaction rates is determined by the reactivity to oxygen and can be reoxidiscd to the correspond-
of the active agents and by their concentration. ing azo compound. Therefore it is very likely t h a t
The last factor can be influenced by diffusion tlie phenomenon observed can be ascribed to the
processes. I agree, therefore, with the suggestion oxidation of Mmall amounts of hydrazo compounds
put forward by Dr. Ingamclls. For systems present in the solution. The occurrence of cis-trans
consisting of azo dyes absorbed on cellulosic isomerism can be excluded, not only for theoretical
fibres, however, it seems probable t h a t the reasons, but also because no changes in t h e
mechanism of the fading reaction will resemble the absorption spectruni of the dye (measured between
one proposed in this paper. The substrate (cellulose) 200 and 600 mp) have been observed after short
has a closc structural similarity to some of the periods of illumination in absence of active
substrates used and the experiments carried out to substrates. Moreover, the deepening of the
dctcrniine the wavelength region of the photo- colour then should also occur without introduction
chemically active light point in the same direction. of oxygen.
Mr. E. COATES*: The photochemical reductive
Dr. L. PETERS:There is no reference in the fission of azo dyes gives rise to amine products
paper to the statement mado by Dr. van Beek ill which have been identified by diazotisation,
his introduction that bubbling of oxygen rifter coupling with well-known intermediates, and
fading has occurred causes the colour to reappear. chromatographic separation of the resulting dyes,
Is it possible that c*is-trans isomerism could account followed by spectrophotometric studies. This
for this effect, although in the paper i t is claimed procedure should be capablc of providing
that such isomerism does not take place? quantitative data. Has any information been
Dr. VAN BEEK: The statement referred t o tlie obtained as to the amounts of ttmiurs produced
observation that solutions of the azo dye used, from a given amount of dye degraded?
which were partly faded in nitrogen, show a Dr. VAN BEEK*: It is difficult t o obtain
deepening of colour when oxygen is introcluced. quantitative data on the amounts of' amines
This phenomenon has not yet been thoroughly produccbd by photochemical reduction of azo
investigated. However, it is well known that dyes. The diseotisation and coupling processes
hydrazo compounds are intcrmediates in the do not proceed quantitatively, especially with
reduction of azo compounds to the corresponding the aminonaphtholsulphonic acid derivatives.
amines. For photochemical reductions, indications Chromatographic separation also gives a low yield.
of the presence of hydrazo compounds have also Investigations on a larger scale, however, are
been obtained by Blaisdell7. Secondly, it is well being carried out in our laboratory, the results of
known that many hydrazo compounds are sensitive which might throw light on this matter.
Collllllunlcntcd