RPT ACC.

NO

. THIS COPY FOR

L J

TECHNICAL REPORT
I

DIVISION
KODAK RESEARCH LABORATORIES
Photoscience Research Division
ROCHESTER, N.Y.

STABILIZATION OF INFRARED
SPECTRAL
SENSITIZING
DYES IN SILVER
HALIDE I
I EMULSIONMELTS. 5105
I
ALTHGD5 CONTRIBJTORS

J. R. Lenhard and B. R. Hein

Infrared-lightabsorbing dyes that are used for the spectral sensitization

of photographic
silver halide materialsare prone to oxidative decomposition whenin an emulsion-melt
environment. This instability of the sensitizing dyemay lead to variability in the speed and fog
levels of the coated photographic emulsion. Infrared diffuse reflectance spectrophotometry
was used to show that cationic tricarbocyanine sensitizing dyes undergo a one-electron
oxidation in AgX melts to yield the corresponding dye radical dications. Correlations from
spectra obtainedfor dye radical dicationsin solution were usedto aid in identificationof the dye-
related species presentin silver halide emulsions. The rateof sensitizing dye oxidationin an
emulsion at40 O C depends on the dye redoxpotential and on the aggregation stateof the
unadsorbed dye. Sensitizing-dye stabilitycan be greatly enhancedby the introduction of a mild
reducing-agent/surfactantcombination to the emulsion meltprior to the addition of dye.
to improved manufacturabilityof
Improved melt stability of the sensitizing dye should lead
infrared-sensitizedphotographic materials.

PROPERTY OF EASTMAN KODAK

COMPANY NOT TOR EBPER O D U C E D
 Stabilizatlon of InfraredSpectralSensitizingDyesinSilver
HalideEmulsionMelts

I. Introduction 1

II. Experimental Section 4

111. Results and Discussion 6

A. Oxidation of Tricarbocyanine Dyesin Solution 6

6. Oxidation of EnamineTricarbocyanine Dyesin Solution 12

C. Decompositionof Dyes in Silver Halide Emulsions 16

D. Dye Structure- Reactivity Relationships 21

E. Mechanistic Aspects 28

F. SurfactanUReducing Agent Combinations 31

G. Coated Emulsions Containing ReducingAgent and Surfactant 32

IV. Conclusions 34

References and Notes 36

I. Introduction
Interest in infrared-sensitive silver halide materials has greatly expanded
in recent years

due, for the most part, to the development of exposing devices equipped with infrared-emitting

laser diode sources. Photographic materials with sensitivityin the spectral regionof 700 nm to

900 nm have been actively pursued for commercial application

in the areas of graphic arts,

business systems, and medical imaging.’

Because of the energy requirementsfor spectral sensitization of AgX, the searchfor

in an
infrared sensitizers that exhibit both thermodynamic stability and high photoefficiency

emulsion presents a formidable challenge. Thesensitizingdye properties whichare most

important in this regard are the one-electron oxidation and reduction

potentials of the dye.

Relationships between the electrochemical, spectral, and photographic properties of an

infrared sensitizer areillustratedin Scheme I :

Scheme 1. Energy Diagram for Infrared Spectral Sensitizationof Silver Halide

+ I
Conduction

1
>
8
W
Band

A
hv
E LV --C Correlates with
Reduction
Dye
Potential

E HO -” Correlates with
t
-I

2
5
5 Dye Oxidation
I Potential

SILVER SENSITIZING
+
HAL1D E DYE

EHO molecular orbitalsof

where the energies of the lowest vacant ELV and highest occupied

the dye are described by the dye reductionpotential Ered and oxidationpotential Eox

respectively, and hv is the spectral transition energy.

1
 According to Scheme I, sensitization efficiency fora given infrared dye is expected to

correlate with its reduction potential, sincethis measurement reflects both the electron-injection

ability of the dyes' excited state and the electron-trapping ability

of the dyes' ground state.

Cyanine dyes that have been considered for

use as infrared sensitizers, particularly those
from

to have reduction
the tricarbocyanine and enamine tricarbocyanine class,2 are reported

potentials that range between-0.70 V and -0.90 V vs the Ag/AgCI referencee l e ~ t r o d e .The
~

energy of the silver halide conductionband edge will depend slightlyon the halide composition

of the crystal. A previous calibrationof AgBrl indicates the energyof the conduction bandto be

equivalent to that of a dye having anEred of -0.92 V.4 The conduction band for AgCl is

expected to be at a slightly higher energy (more negative potential)than that for AgBrl. Given

these electrochemical data, the lowest vacant molecular orbitals

of infrared dyes lie just below

the level of the silver halide conductionband edge. Consequently, spectral sensitization

efficiency should be very sensitiveto changes in dye reduction potential.

Contrary to expectation, results obtainedon AgCl and AgClBrl emulsions indicate that no

clear correlation exists between infrared photographic speed and dye reduction
potential.3a

Instead,photographicperformancedepended on both Ered Eox. As shown in Figure 1, an

approximately linear relationship exists between relative

infrared speed and the absolute

difference (Eox - Ered) in the redox potentials of a dye. Since the spectral absorption energy hv

of a dye is directly relatedto the quantity (Eox - Ered), minimum infrared speed and maximum

blue desensitizationare associated with sensitizationsintended at the longer infrared

wavelengths.

This dependency of photographic performanceon Eox is thought to be the result of the

correlation of this measurement witha dyes' propensity toward oxidation

in the silver halide

emulsion melt. Depending on the type and extent of substitution, tricarbocyanine and enamine

in the range of 0.3 to 0.5 V vs Ag/AgC1.3 Dyes

tricarbocyanine dyes have oxidation potentials

0.5 V are thermodynamically unstable with respect

with oxidation potentials lower than to silver

ion and are, furthermore, proneto aerial oxidation. Chemical reactivity of these dyesin emulsion

rneltS may lead to destruction of dye and creationof desensitizing species which reduce

spectral sensitizing efficiency and loweroverall photographic speed.6 Increasedfog and higher

2
incubation instability is also expected for photographic systems containing dyes
of such low

oxidation p ~ t e n t i a l . ~ ? ~

3.0

2.5

2.0

1.5

1.o
- I .45 -1.35 -1.25 -1.1 5 -1.05

(Eox- E red) 9 "

Figure 1. Plot of relative infrared speed versus absolute difference
in redox potentials for dyeson a 0.35 pm cubic AgCl emulsion.

This paper describes the results

of spectroscopic and electrochemical experiments

designed to study the oxidative decompositionof infrared sensitizing dyesin silver halide

emulsion systems. Absorption spectra obtainedin acetonitrile for mono-oxidized, radical

dication formsof various tricarbocyanine and enamine tricarbocyanine dyes

are presented, and

these data areused to facilitate identificationof the dye oxidation products formed
in an

emulsion environment. Diffuse reflectance spectrophotometry is used to explore the important

factors controlling the rateof dye oxidationin silver halide emulsion melts. The resultsfrom

these experiments suggesta model that relates the observed decomposition rate
in an

emulsion meltto the redox, aggregation, and adsorption properties

of the dye. Finally, a

3
practical method for stabilizinginfrared spectral sensitizing dyesis presented which should lead

to improved manufacturability of infrared-sensitizedphotographic materials.

Experimental
Section.
Materials. 3,3'-Diethylthiatricarbocyanineiodide was commercially available (Kodak

Laboratory Chemicals). Other infrared sensitizing dyes were obtained from samples stored

under refrigeration. Dyes obtained with halide counterions were converted to perchlorate
or

tetrafluoroboratesalts for electrochemical studies; however, emulsion studies were

carried out

on bromide and iodide salts as well. Acetonitrile (CH3CN, MCB spectrograde) wasdried over 4

A molecular sieves (Kodak Laboratory Chemicals, baked

at 400 OC). Tetrabutlylammonium-

tetrafluoroborate (TBABF4, Kodak Laboratory Chemicals) was recrystallized three times from

ethanolhvater, vacuum dried at50 0 C, and stored in a desiccator. Ferric Chloride hexahydrate

(FeC13, Kodak Laboratory Chemicals), ascorbic acid (Kodak Laboratory Chemicals),

hydroquinone (Kodak Laboratory Chemicals), methanol (Kodak Laboratory Chemicals,

spectrograde), Triton X-100 (Aldrich),Tween 80 (Aldrich),glycerol (Aldrich), and 10G were used

as received.

The silver halide emulsion most usedin these studies consisted of0.35 pm edge-length

AgCl cubes (1.8 kg/ mol Ag, pH=4.8, vAg=125 mV). In some experiments a 0.28pm edge-

length, cubic AgCI(90%)Br(8%)1(2%) emulsion (0.83 kg/ mol Ag) was used. Each emulsion was

pretreated withtriazinylstilbenesulfonic acid (TSSA,500 mg/mol Ag), 1 -(3-acetomidophenyl)-5-

mercaptotetrazole (APMT, 150 mg/mol Ag), and1 mole oo/ potassium bromide before additionof

reducing agent, surfactant, or dye. Gelatin solutions,4.3% , pH 5.5, were prepared from

crystallizedtype IV deionized gel(class 088).

Electrochemistry. Electrochemical experiments were performed using a Princeton

Applied Research Corp. (PAR) 173 potentiostat in conjunction with PAR models 175 universal

programmer, PAR 179 digital coulometer, and 124A Lock-in amplifier. A Hewlett-Packard model

239A low-distortion oscillator was usedin AC measurements. Formal oxidation and reduction

potentials for dyes were obtained via phase-selective second-harmonic

AC voltammetry

(quadrature component) at an applied frequency

of 400 Hz (16 mV peak-to-peak signal)as

4
described previously.8 Solutions forAC voltammetric examination contained0.1 M TBABF4

and ca. 5 x 10-4 M dye and were deaerated with argon prior to examination. The working

electrode was aPt disk (ca. 0.02 cm2) that was polished with1-pmdiamond paste (Buhler

Metadi), rinsed with water, and dried before each experiment. Redox
potentials for reducing

agents were obtained by cyclic voltammetry using a potential scan rate of 0.1 VIS. Current-

voltage curves were recordedon a Hewlett-Packard model 7045AX-Y recorder. All potentials

were measured vs the NaCl saturated calomel electrode

at 22 OC and converted to the Ag/AgCI

reference by adding 40 mV.

TransmissionSpectra of Radicals andAggregates. Solutions of dye radical

dication were prepared by chemicalor electrochemical one-electron oxidation. Chemical

oxidations were performed by adding a slight excess

of a ferric chloride/methanol solutionto a

dilute (ca. 10-6 M) solution of dye in methanol. Electrooxidized dye solutions were prepared by

controlled-potentialcoulometry with a PAR 337A (three compartment) coulometriccell system

equipped with a 67 cm2 area

pt gauze working electrode (electrolysis time for >98% oxidation
is

ca. 2 min). In some experiments, thecell used was a tubular electrochemical

flow-cell of design

similar to that reported by Miner and Kissingerg except that the working electrode was

compressed Pt gauze (Fisher). In either experiment the working electrode was poised at a

potential 200 mV beyond the reversible oxidation potential of the dye. The electrolyzed

solutions were rapidly pumped through a 1 mm10

ormm quartz spectralflow-cell (Helma) for

absorption measurement.

Solutions containing a mixture

of dye in the monomeric and aggregate form were prepared

at a constant total dye concentration by careful dilution of methanol solutionsof dye with

deionized water.

or
Transmission spectra were measured with a Varian 2400 UV-vis-IR spectrophotometer

with a Hewlett-Packard 8450A UV-vis diode array spectrophotometer.

Diffuse Reflectance Spectroscopy. Stabilities of the dyes in the liquid emulsions

(melts) were assessed by diffuse reflectance spectrophotometry using a Varian 2400

spectrophotometer fitted with an integrating sphere. Melt samples were prepared by adding 1

mL dye stock solution (in methanol) to 4 mL of stirred emulsion(2.2 mmol Ag). Stock solutions

5
of each dye wereused to achieve a concentrationof 0.03 mmol dye/mol Ag . The dyed melt-

concentrate washeld for a minimumof 3 minat 40 OC to facilitate adsorption of the dye

to the

silver halidegrain surface. The emulsionwas subsequently diluted with 9.5 mL of 4.3% gelatin

or, in some experiments, phosphate buffer,pH 5.5. Dye stability results were largely

independent of the diluent used. Dilution with gelatin gave a

final emulsion that contained270

g gel /mole Ag. Capped "Opticlear" vials (Kimble, 4 dram) served

as convenient, disposable

cells for incubation and for recording reflection spectra

of turbid emulsion melt samples. Dyed

samples were incubatedin the dark at 40 OC using a LauddBtinkman MS-20 constant

temperature bath. Spectra were recordedas a functionof incubation time and were measured

vs an electronically stored spectrum

of an undyed emulsion. In general, there was no changein

reflectance spectrum of emulsions after several successive spectral recordings,

indicating that

any photoexposure incurred during routine measurementcontributed little to the observed

spectral changes.

Photographic Evaluation. Coatings of dyed emulsion that were incubated for 3.5 hr

at 40 OC were made on Estar support at a level of200 mg/ft2 Ag and 483 mg/ft2 of gelatin. AgCl

emulsion samples were doctored

as described above and were prepared with and without

added surfactant and reducing agent. Relative spectral sensitivity was determined
by

exposures (10 sec) using a wedge spectrographic instrumentfitted with a Kodak Wratten 89B

filter. The degree of desensitization was obtainedby exposures using a Corning #3850 cutoff

filter. All coatings were processed for2 min in Dektol.

Ill Results and Discussion.

A. Oxidation of Tricarbocyanine Dyes in Solution. The one-electronoxidation

of a cationictricahcyanine dye initially yields the corresponding radical dication. The electron

density distributionin the singly occupied molecular orbital (SOMO)

of the radical was calculated

using MOPAC1 and is schematically depicted in Figure 2. The odd-electron densityat a given

atom is proportional to the square of the orbital coefficient andis represented in Figure 2 by the

size of the circle. For a thiatricaftmcyanine radical dication, the electron density
in the SOMO is

6
seen to be symmetrically distributed and foundto reside primarilyon the even-alternant carbon

atoms of the polymethine chain(c8, Cqo, C12, and C14) andto a lesser extenton the nitrogen

atoms.

Figure 2. Diagram for electron density distribution in SOMO of a

thiatricarbocyanine radical dication.

Results fromcyclic voltammetry experiments indicate that tricarbocyanine radical

dication

stability is markedly dependent on the type and extentof substitution in the polymethine

chain. Radical dications derived from dyes which contain alkyl substituents
at the even-

numbered carbonsof the methine chain which stabilize the radical and/or sterically inhibit

radical-radical coupling reactions were foundto exhibit a modest degreeof stability. Chemical

structures of various dyesfor which solutions of radical dication could be preparedin methanol

or acetonitrile by electrochemical one-electron oxidation

at a Pt anode or by chemical oxidation

using FeC13 are shown in Table I.

The thiatricarbocyanine dye 1, which contains hydrocarbon bridges at 8the

, 10, 12, and

14 positions of the methine chain, exhibits reversible cyclic voltammetry and forms the most

stable radical dication. Results from spectroelectrochemical experiments are also consistent

with formationof a persistent radical. Figure3a shows, for example, absorption spectra of a 17

pM solution of dye 1 obtained before and after exhaustive one-electron oxidation

at 0.65 V vs

Ag/AgCI. The radical dication of dye 1 exhibits an absorption maximum at 542 nmin acetonitrile

and a molar extinction coefficient&,axrad of 9.5 x 1O4 M-l cm-l. Immediate one-electron

reduction of the solution of Figure 3a, at 0.2 V, regenerates ca. 95% of the parent dye1

7
( L a x d y e = 808 nm, &,,dye =20.7x l o 4 M-l cm-l). The shape of the principal absorption

band for the radical closely resembles that for the parent dye, except for (ca.170 cm-1)
a small

shift in the positionof the vibronic transition which appears as a shoulder on the short-

wavelength side of the dyes'O-->O transition. The similarity in spectral band shapes supports

the conclusion that there are

no gross differences in the overall structure (bond lengths,

molecular conformation, etc.) between dye and radical.

2-6 was found to depend on

The stability of the radical species derived from the dyes

molecular structure and concentrationof radical. The radical dications are susceptible to

radical-radical dimerizationat the unsubstituted, even-numbered methine carbon atoms as

depicted in Scheme 11.l1 Under cyclic voltammetry conditions, where the concentration of dye

is ca. 5 x 1O4 M, the radical dications havea half-life of about 1s. More stable solutions of

(10-6 M) via chemical

radical dication could, however, be prepared at extreme dilution

oxidation using ferric chloride. Spectral absorption data are listed

in Table I.

Scheme II. Proposed Dimerization Mechanism for Tricarbocyanine Dyes.

Et

Et I Et

8
 Table I. Spectral Data for Tricarbocyanine Dyes and Radical Dicationsa.

1 *
q -p 808 20.7
542 9.5

2 785 20.0 554 13.2

3 q+ CH3 9CH3
708 26.8 528 10.7

4 q2WL,
\ / Et s - p 804 27.7 568 13c

5 824 19.3 600,580

6 762 29.6 604 13.3d

R R
~~ ~~~

a- Radical dications generated by chemical oxidation with FeC13/methanol or by

electrochemical oxidation in CH3CN/O.l M TBABF4. b. E m a is the optical extinction at
theabsorptionmaximum, x 1O-4; h a d y e , and hmarad, Emaxfad arevalues
obtained for the dye and corresponding radical dication, respectively.c. Radical dication
instabilitypreventsanaccurateassessment of d. R is sulfobutyl

9
 b. iI
! ',

C.

I I I I I I
300 400 500 600 700 800

Wavelength, nm

(- - - - - - -)
Figure 3. Absorption spectra for tricarbocyanine dyes recorded before
andafter ( ) oxidation.Spectrum of oxidationproductaftertreatment with
base (.................. ). Absorbance scale is arbitrary, see text for extinction values.

10
The radical dications of dyes 2-6 exhibit absorption maximain the range of 528nm to 604 nm

and have molar extinction coefficients that are slightly less than one-half that measured
for the

parent dye. UV-vis spectra for the radical dications

of dyes 2,3,4 and6 are very similarto that

shown for dye 1 (Figure 3a). Dyes containing the 4,5-benzo (dye 5)or the 5,6-dithiomethyl

substituent, on the other hand, exhibit an unusual double-banded radical dication absorption

feature as shown in Figure 3b. These radical dications could be converted backto dye by

electrochemical one-electron reductionor by chemical reduction with aqueous ascorbic acid.12

In contrast to dyes 1-6, tricatbocyanine dyes

7-9 (shown below) that contain no alkyl

substituents on the even-methine carbon atoms exhibit completely irreversible, one-electron

cyclic voltammograms in acetonitrile. The electrochemical data are consistent with a rapid and

irreversible dimerization of the corresponding radical dications. The majority of tricarbocyanine

radicals with this patternof chain substitution exhibit lifetimes under 50 msec
in CH$N/O.l M

TBABF4 with structurally dependent rates of decomposition.13

Dye 7
Et Et

q=@ Et Et
I
Dye 8

Dye 9
Et Et

(FeC13) oxidation of a 10-6 M solution of

Absorption spectra recorded during the chemical

dye 7 (hmaxdye
= 764 nm, &,dye = 18.1 x l o 4 M-l cm-l), which is bridged across the9 , l l
(hmaxrad
carbons of the heptamethine chain, also indicate the radical dication = 540 nrn) to be

transient. The oxidation product of dye 7, which can be isolated as a solid by counter-ion

11
exchange precipitation, exhibits an absorption maximum at 391 nm
in acetonitrile (Figure 3c)

and may be assignedto the dimer formof dye 7 (&mmdimer= 2 .9 x lo4 M-l cm-l). Similarly,
chemical oxidationof dye 8 gives a transient radical(hmaxrad = 542 nm) anda stable dimer

product (hmaXdimer = 410 nm). This pattern of reactivity observedfor tricahcyanine radical

dications is completely analogousto that reported by Parton and Lenhardfor dicarbocyanine

dyes.l

Depending on reaction conditions, dimeric products formed during the one-electron

oxidation of tricarbocyanine dyes are susceptibleto further reaction. For example, treatmentof

the UV-absorbing dimerof dye 7 with triethylamine gives a species whose spectral properties

(Figure 3c) are similarto those of the parent dye. This behavioris consistent with

deprotonation of the methine carbonsof the dimerat the sites of radical couplingto yield the

corresponding bisdye,,X,( =764 nm). Parton similarly isolated the products of the chemical

,,X,(
oxidation of dye 9 and identifiedthem as the bis-dye = 745 nm) and oxidized bis-dye

species (Amax = 520 nm).I5

B. Oxidation of Enamine Tricarbocyanine Dyes in Solution. The structural

features which control the reactivity of radical dications derived from enamine tricarbocyanine

dyes parallel those discussed above for conventional tricarbocyanine dyes. Dye 10, for

instance, which contains alkyl bridging at each of the even-numbered carbonsof the

heptamethine chain, exhibits reversible one-electron cyclic voltammetry. The absorption

spectra recorded during the oxidationof enamine tricarbocyanine dyes are, however, distinctly

different from those obtained fortheir tricarbocyanine analogs. As shown in Figure 4a , the UV-

vis spectrum obtained for a solution

of dye 10 (hmaxdye = 777 nm, Emaxdye =13.2x 1O4 M-l

cm-l) after exhaustive electrolysis at0.5 V exhibits ~JQabsorption bands at 660 nm

(Emax =

6.27 x lo4 M-l cm-l)and 542 nm (Emax = 4.1 x 104 M-1 cm-1). The appearance oftwo

to have two allowed spectroscopic

absorption bands may indicate the dye 10 radical dication

transitions in the visible region, oneof which involves electron density associated with the

piperazinylgroup. Alternatively, the spectral data may be interpretedin terms of two distinct

chemical species, i.e., dye 10 radical dication and an additional oxidation product. Thesolution

Of oxidized dye 10 was relatively stable( 1 1 of

~ ca. 30 min) and electroreduction of the solution

12
of Figure 4a at 0.0 V resulted in regeneration of dye at >90% yield. The reversibility of the

oxidation processis consistent with formationof a stable radical dication and supports the

conclusion that both absorption bands

are associated with the radical.16

YCOZEt

Dye 10

/Co2Et

Dye 1 1

Interestingly, the oxidationof dye 11 (hmaxdye

= 731 nm, Emaxdye
= 16.5 x lo4 M-l cm-l)

was found to be irreversiblein cyclic voltammetry experiments, yet the absorption spectrum

obtained foroxidized dye 1 1 (Figure 4b) also showstwo absorption featuresin the visible

region (648 nm and 528 nm). The absorption bandat 648 nm was transient,in contrast to the

long wavelengthband in Figure 4a, and absorbance increases at 528 nm were concomitant

with the lossof absorbance at 648 nrn. In further contrastto dye 10, dye regenerationdid not

accompany electroreductionat -0.1 V of the solutionof Figure 4b . The radical dicationof dye

1 1 could conceivably undergo radical-radical couplingat the unsubstituted even-numbered

illustrated in Scheme
methine carbon atoms (8 and14) to yield the corresponding dimer as

111.11

13
 a
/co2E'

b.
(*ZE'

i '\
* \
/

/
/
/
'\
\

-""""" """""-4@

400 500 600 700 800

Wavelength, nm

Figure 4. Absorption spectra for enamine tricarbocyanine dyes recorded before

(..: ............... ) andafter ( ) electrochemicaloxidationin CH3CN/O.1 M TBABF4.

14
 Scheme 111. Proposed Dimerization Mechanism for Enamine Tricarbocyanine Dyes.
PE'

FIE'
Q-J$+"--
/ -/ -
..,.-.,,-. ..,.'
+"../"
YEl . I
Et

Due to the presence of the piperazinyl group at thecentral methine carbon in the dye, the

dimeric oxidation product of dye 1 1 will contain a hemicyanine chromophore. This chromophore

consists of piperazinyl and benzothiazole nuclei that are connected by a dimethine linkage.As

such, dimers derived from enamine tricarbocyanine can be expected to exhibit spectral

absorption in the visible region. A mass balance analysisof the data of Figure 4b

indicates the species absorbing at528 nm to have anextinction coefficient (per chromophore)

of ca. 8 x 1 O4 M-l cm-l. Nearly identical spectra characteristics are reported for an authentic

sample of the similarly structured hemicyanine dye12 (Amaxdye= 522 nm, Emaxdye = 7.2x lo4
M-l cm-l).17

Dye 12

Overall, these data suggest that the spectral absorption bands

at 648 nm and528 nm of

Figure 4b correspond to dye 1 1 radical dication and dimer oxidation product, respectively.

15
 C. OxidativeDecomposition of DyesinSilverHalideEmulsions.

The reactivity of various infrared sensitizing dyes

in silver halide emulsions was measured

by diffuse reflectance spectrophotometry using

a 0.35 prn AgCl or AgClBrl cubic emulsion. For

the most part, these emulsions were handled

in much the same manner asif to make coatings

gel contents were adjusted slightlyto

for photographic applications, except that dye and

provide an optimum reflection density for measurement.Some emulsions were chemically

sensitized (Au~S)and all contained the antifoggantAPMT. As described in the experimental

section, the emulsions werepretreated with potassium bromide,to enhance the adsorptionof

the dye, and with the supersensitizer TSSA, to inhibit aggregationof the dye on the AgX

surface.18 Dyes were addedto the pretreated emulsionin its concentrated form,at 40 O C ,

from a methanolic solutionof dye at a level to achieve a surface coverageof ca. 10%. This low

surface coverage of dye, together with the presence

of TSSA, ensured that the sensitizing

dye was adsorbed primarily in the monomeric state. Dimer-aggregate and H-aggregate forms of

halide^.^ After
infrared dyes are avoided because they are strong desensitizers for silver

to yield a sample that closely

addition of dye, the concentrated emulsion was diluted

resembled an emulsion suitable for preparing photographic coatings.

Experimental data were obtained

as plots of reflection absorbance(- Log R) vs

wavelength. For quantitative analysisof dye spectra in these turbid emulsion systems the

Kubelka-Munk function (Eqn.1) was utilized;

(1-Roo)*
ws = 2R, = csE

where K and S are absorption and scatter coefficients

of the silver halide substrate,,R is the

reflectivity of an infinitely thick sample, andC and E are the molar concentration and molar

absorptivity coefficientsof the dye, respectively.19 Experimental- Log R values were used to

calculate R, and K/Sratios according to Equation 1 so that changes in reflectivity could be

proportionally relatedto changes in the surface concentrationof dye. The calibration plot of

Figure 5 demonstrates this proportionality between K/S and dye concentration and verifies

that conditions of infinite reflectivity are adequately satisfiedin these experiments.20

16
 0.2

0.1
1L

w
0.0 0.1 0.2 0.3 0.4

[ Dye 1, mmoles dye/mole Ag

Figure 5. Calibration plot of the Kubelka-Munk function K/S

versus total dye concentrationin a 0.35 pm cubic AgCl emulsion.

Figure 6a shows reflectance spectra periodically recorded for the AgCl emulsion
(pH 6), which

was incubatedat 40 OC and contained the tricarbocyanine dye 1. Theinitial spectrum shows two

strong (reflection)absorption bands at 860 nm,and 590 nm, and a relatively weak band at 460 nrn.

The spectral bandat 860 is due to the monomerically adsorbed dye andis hypsochromically shifted

(52 nm) and broadened relativeto the spectrum of the dye in solution. This hypsochromic shift is

related to the refractive indexof the silver halide and also depends on the spectral
transition

energy of the dye. Most infrared sensitizers exhibit a50 nm to 60 nm shift on adsorption. Having a

very low oxidation potential, ,E, = 0.18 V, dye 1 was especially reactive in the melt environment,

and the rapid loss of the dye absorption band at 860 nm was accompanied by increases
in

absorption at 590 nm. The latter spectral band occurs 48 nm hypsochromic of the absorption

maximum of the dye 1 radical dication

in solution. Spectral changes similar to that displayedin

Figure 6a could also be obtained by deliberate chemical oxidation of the AgCl emulsion containing

dye 1. Treating the emulsion sample with

FeC13 resulted in bleaching of the dye and

17
0.3

0.2

0.1

0.2

0.1

500 600 700 800 900 1000

Wavelength, nm
Figure 6. Reflection spectra recorded for dyed emulsionsas a function of time at
40 O C ; a)tricarbocyaninedyeon AgCl at ( ) O h r , ( - - - - - - - ) hr,(-
l -- -)5hr;
and b) enamine dye on AgClBrl at ( ) 0 hr, (- - - - - - -) 2.5hr, (- -- -) and 18 hr.

18
generation of a spectral band at590 nm. This correspondence in the spectral and reactivity

features between solution and emulsion systems permits assignment

of the absorption band at

590 nm to the adsorbed, radical-dicationform of dye 1.

As the reflectance spectraof Figure 6a indicate, the oxidative decomposition of dye

1 to

yield radical dicationis more than 90% complete after5 hr. A kinetic analysisof dye 1 oxidation

of equation 1. As illustrated in Figure 7, a plot of In [dye] versus timeis

was obtained with the aid

linear, indicating that the oxidation reaction proceeds

in a manner thatis first order with respect to

the concentration of dye.

0 1 2 3 4 5 6
Time , hrs.

Figure 7. Kinetic plot of In [dye] versus time for the oxidative

decomposition of a dye 1 in a AgCl emulsion meltat 40 OC.

Good correspondence between solutionand emulsion spectral features was also obtained

for a AgClBrl emulsion containing a

for enamine tricarbocyanine dyes. Spectra recorded

substituted enamine dyeare shown in Figure 6b. As was found for tricarbocyanine dye
1,

considerable decomposition occurswithin minutes of introduction of the enamine dyeto the

emulsion melt. The initial spectrum shows reflection-absorption bands

at 810 nm, 720 nm, and

19
560 nm. These spectral bands, when corrected for the spectral shift due to adsorption
to silver

halide, correlatewell with the absorption bands determined

in solution for the dye,at 752 nm,

radical dication,at 664 nm and 540 nm.

Although the data of Figure6 clearly demonstrate the formation

of radical dication species,

conclusions regarding the participation

of radical-radical couplingin the overall mechanism of

dye decomposition in the emulsion environment are less obvious. For enamine tricarbocyanine

dyes, identificationof the dimer is made difficult becauseof the stabilityof the radical and by the

fact that both the enamine radical-dication and the dimer product exhibit spectral absorption

near 540 nm. Proof of dimer product formation during the oxidative decomposition
of

tricatbocyanines in an emulsion was equally difficult. Dimer products for tricarbocyanine dyes

can be expected to absorb in the 390 - 450 nm range depending on the

specific dimer structure

and strengthof adsorption to ASCI. Due to spectral absorptionand light scattering by the

emulsion at wavelengths lessthan 450 nm, formationof tricarbocyanine dimer products cannot

easily be verified by reflectance spectrophotometric methods.

To identify the important emulsion variables that influence spectral sensitizing dye stability,

reflectance spectra analogous

to those of Figure 6 were qualitatively comparedas a function of

emulsion composition. Dyes 2 and 11 were chosen as representative tricarbocyanineand

enamine tricarbocyanine dyes for this analysis because they gave well-defined radical dication

spectra in the emulsion. Higher emulsion pH, lower temperature, and the presence
of added

bromide ion decreased sensitizing dye reactivity.21Among these factors,pH was the dominant

emulsion variable controlling dye stability. Substantial increasesin dye oxidation rate were

associated with lowering emulsion

pH from 7.2 to 4.8. This sensitivity to pH is most likely related

to the hydrogen ion dependence of the oxygen electrode reaction22(Equation 2) and

suggests that dye decompositionin an emulsion meltis largely driven by the aerial oxidation

reaction represented by Equation 4:23,24

@ + 2H+ + 2e- ------> Hz@ (2)

2 Dye+ ------> 2 Dye++ + 2e- (3)

Overall: 02 + 2 Dye+ + 2H+ ------> 2 Dye++ + Hz@ (4)

20
 D. Dye
Structure-Reactivity
Relationships.

The relative melt-stabilitiesof approximately 20 infrared sensitizers were measured by

reflectance spectroscopy in an effort to examine the relationship between dyeEox and dye

stability. Attempts to obtain quantitative data for the rate

of sensitizing dye oxidationin the

emulsion melt using the kinetic analysisas shown in Figure 7 , however, were largely

unsuccessful. Complications dueto the appearance of broad and overlapping spectral bands,

together with the presence of the various aggregate bands observed for many dyes, prevented

the use of equation1 to obtain the necessary dye surface concentrations. The
formation of

aggregated dyein the solution (aqueous) phaseof the emulsion was predominant during the

initial stages of the dye adsorption process, and furthermore, played

an integral rolein the dye

oxidation mechanism. Compared to cyanine dyes of shorter methinechain length,

tricarbocyanines are very insolublein water and are knownto rapidly form H-aggregates when

introduced to an aqueous solution.

An increasein the concentrationof dye or an increasein the water/solvent ratioof the

medium causesa shift in the aggregate equilibrium away from the monomer and toward the

dimer, trimer, and higher H-aggregate forms.25 Figure

8 shows the progressionin the

absorption maximumof the aggregate, from750 nm to ca. 600 nm, that resultsfrom a gradual

increase in the water contentof a solution of dyes 2 and 5. A solution compositionof 80%

water/20% methanol most closely corresponds

to the methanoVwater content of the

concentrated emulsion melts described here. For each dye there exists
a characteristic

water/methanol ratio wherein one-half

of the dye is in an aggregated state. These half-

aggregation solvent ratios can

be used as a measure of aggregation propensity for various dyes.

A comparison of such data showed that tricarbocyanines containing the 4,5,4',5'-benzo

or the

5,6,5'6'-thiomethyl substituents aremore prone to H-aggregationthan similar dyes thatdo not

contain these functional groups.

21
 a.
0.6

0.4

0.2

0.0
b.

dt

0.4
I \

0.2

0.0 I I I I
550 650 750
Wavelength, nm

Figure 8. Absorption Spectra for tricarbocyanine dyes in methanol/water

mixtures. Methanol content is20°h (- -- -), 50% (- - - - ), and 100%
&"+; [dye] = 4 m.
22
 These aggregation phenomena influence the dynamics
of dye adsorptionto ASCI, as

illustrated by the data

of Figure 9. Reflectance spectra are shown for dyes
2 and 5 wherein the

time allowed for dye adsorption, that is, period

the for which the dyed emulsion was
held in the

to 5 min. The data obtained for 4,5,4',5'-benzo-

concentrated form, was vaned from 1 min

substituted dye5 show that in the earliest stagesof adsorption, distinct spectral bands are

observed for the adsorbed dye monomer (875 nm) and H-aggregate
(640 nm). Continued

adsorption resultsin a decreasein the aggregate spectralband and a concomitant growth of the

5 minutes,
bands associatedwith the adsorbed monomer and radical dication (670 nm). After

the adsorption processis essentially complete. In comparison, H-aggregationis less obvious

of adsorbed monomer is relatively high. Small changesin the

for dye 2 and the initial amount

relative amountsof adsorbed dye 2 monomer (845 nm) and radical dication (625 nm) are

observed during these first 3 minutes.The data of Figure 9 and other results indicate that,for

most cationic tricarbocyanines, a maximum

of adsorbed monomeric dyeis obtained during a6-

min melt-hold at40 OC.

Although quantitative rate data were difficult to extract from reflectance spectra, valuable

structure-reactivity information for dyes was obtained from a qualitative analysis

of spectral

results. Figure 10 shows,for example, that significant stability differences exist between a family

of similarly structured dyes thatvary in oxidation potential and aggregation propensity. In

each

of the spectra(a-c), reflection-absorption bandscan be distinguished that correspond

to the

adsorbed dye monomer(800-900 nm), aggregated dye (680-800nm), and radical dication (500-

680 nm). As given by the relative magnitudeof the spectral bands for dye and radical
, the

4,5,4',5'-benzo substituted dye 5 (Eox = 0.315 V) is much less stable than the 5,6,5',6"

thiomethyl substituted dye or the unsubstituted parent dye 2. The thiomethyl substituted dye

(Eox= 0.35 V ) appears slightly more stablethan the unsubstituted dye(Eox = 0.382 V).

Interestingly, the 4,5,4',5'-benzo substituted dye and the 5,6,5',6'-thiomethyl substituted dye

posses similar radical dication characteristics, and aggregation tendencies

in solution. Clearly,

dye stability is a complex functionof not only the dyes' oxidationpotential and aggregation

propensity, but alsoon the adsorptivity to silver halide.

23
 a.
0.2

0.1

b.

0.2

0.1

500 600 700 800 900 1000

Wavelength, nm

Figure 9. Reflectance spectra for aAgCl emulsion recorded at various times

after addition of dye. a) dye 2 at ( )1 minand (- - -) 3 min, pH 4.8; b) dye
5 at ( ) 1 minand (--- ) 5 min,pH 7.5. Eachemulsionsamplemeltwas
diluted with phosphate buffer before measurement.

24
 a

0.2

0.1

b.
0.2

0.1

C.
0.2
Et

0.1

500 600 700 800 900

Wavelength, nm
Figure 10. Reflectance spectra for AgCl emulsions at pH 4.8 recorded after
melt-holdtimes of 0 hr ( ), 0.5 hr (- ”), and 2.0 hr (..................
).

25
 Table II. Electrochemical and Reflectance Spectral Data for Tricarbocyanine Dyes.
~~~

No. Eox,StructureDye hdy8v bad, Stability

Relative
v nm m

8
Q - 9 0.42 790 600 Stable
A dt

7 0.422 81 0 600 Stable

13 0.374 860 Stable

9 q“.1p / 0.450 81 0 590 Less Stable

dl dt

14 0.407 825 61 0 Less Stable

2 0.382 845 625 Less Stable

15 0.43 720 565 Less Stable

1 0.183 860 590 Unstable

5 0.315 875 670 Unstable

16 0.37 860 670 Unstable

17 0.39 835 650 Unstable

4 0.322 860 - Unstable

a. Relative stability was determinedon an AgCl emulsion atpH 4.8 for emulsion concentrates that
were diluted with buffer solution. Redox potential measured in CH3CN/O.1 M TBABF4. b. byeand
brad are the wavelengthof maximum reflection-absorption of the dye and radical dication,
respectively.

26
 Table I I lists reflectance spectral data obtained for other variously structured

tricarbocyanine dyeson a AgClBrl emulsionat pH 4.8. Each dye is categorizedon a relative

basis as being stable, less stable, and unstable, as gauged by similarities

in emulsion spectra

with those of Figures 6 and 10. These data lead to the following conclusions:

i. Other factors being constant, the

more reactive dyes are those with the

lowest electrochemical oxidation potential.

ii. Dyes which exhibit the greater tendencyto form H-aggregatesin

solution are more reactive than those thatdo not readily aggregate. In

particular, dyes containing 4,5,4',5'-dibenzo or 6,7,6,7'-dibenzo group fit

into this category.

iii. For dyes of similar redox potential, those dyes which contain

functional groups (eg. thiomethyl) that facilitate adsorption to silver halide

are more stable than analogous dyes

in which the groupis absent.

iv. The dyes (1,4,5,16,17) of Table II that were unstablein the emulsion

melt are also reported to perform poorly

in photographic

experirnents.lC~2a~3a

27
 E. Mechanistic
Aspects.

Guided by the above generalizations, we propose Scheme IV

to describe the Processes

associated with theaddition of a methanolic solutionof an infrared Spectral Sensitizing dye to

an aqueous silver halide emulsion melt.

Scheme IV. Oxidation and Adsorption Equilibria for Infrared Dyes in an Emulsion.

DYE-MONOMER
(methanol)

DYE-MONOMER
DYE-MONOMER
(adsorbed) (solution)
dye
AgX ;TSSA
+ 1 add to
emulsion

(solution)
+
Kaggreg
* DYE-AGGREGATE
ads

+/kOx3

ASX
tl ko:

RADICAL- + RADICAL-MONOMER
MONOMER 4 rad (solution)
(adsorbed) ads

According to this scheme, the initial state of the sensitizing dye is thatof a dissolved,

monomeric species. Upon addition to the aqueous emulsion, the dye monomeris susceptible

to aggregation to yield unadsorbed dimer-aggregate and H-aggregate complexes. For

simplicity, these equilibria are considered as one process and are described Kaggreg.
by The

aggregation reactions compete with the silver halide adsorption process for locally available

dye monomer. Since the silver halide grain surfaceis pre-treated with TSSA, the dyeis

essentially adsorbed in the monomeric state. The factor KadSdYerepresents the dye adsorption

equilibrium constant.

28
 Sensitizing dye moleculesin each of the defined physical states, i.e. solution monomer,

solution aggregate, and adsorbed monomer, can undergo one-electron

oxidation to yield the

corresponding dye radical ion.The individual ratesof oxidation are given by bxl,kox2, and

kox3. It is assumed in Scheme IV that during the oxidation

of a dye-aggregate speciesin

solution, the aggregate undergoes a rapid dissociation

to give radical-ion monomers.

Dissociation of the oxidized aggregateis expected for coulombic reasons andis supported by

the fact that experimental efforts to deliberately prepare or measure radical-aggregates have

been unsuccessful. Finally, radical ions that are formed via processes
in the dissolved state

may adsorb to the silver halide grain according to the equilibrium designated
&&lad.

In the ideal emulsion melt, the adsorption

of infrared spectral sensitizing dyes to the

surfaces of silver halide crystals would proceed without the complication

of aggregation and

oxidation of the dye moleculesin the solution phase. We attempted to individually control the

processes of oxidation and aggregationin this emulsion system through the use
of reducing

agents and surfactant compounds, respectively.

For the most part, the additionof an organic reducing-agent, suchas ascorbic acid, to the

emulsion prior to addition of dye had no major influence on the processes

of dye oxidation.

Adding 20-200 mg ascorbic acid/ mol Ag did lower the relative amountof radical dicationinitially

of radical dication formation measured

detected in a freshly prepared melt. However, the rate in

by the presence of
melts subjected to a prolonged melt-hold was only nominally affected

ascorbic acid. The use of ascorbic acid alone did eliminate radical dication formation.

Likewise, separate additionof various levelsof a surfactant, such as Phorwhite, TritonX-1 00 or

1OG,to the meltprior to dye addition (no ascorbic acid present) had little effect
on the kinetics

of dye oxidation.

The spectraof Figure 11 demonstrate, however, that a dramatic enhancementof dye

stability could be obtained if the ascorbic acid and surfactant were used
in combination. Figure

1l a compares the reflectance spectra measured after a 3melt-hold

hr period for an emulsion

containing dye5 with and without added ascorbic acid (200 mg/mol
Ag) and surfactant10G

29
 a

L”-

b.

- I I I I I
500 600 700 800 900

Wavelength, nm
Figure 11. Reflectance spectra for dyes 2 and 5 on a AgCl emulsion at 40 O C ;
a) comparisonafter 3 hrhold of emulsionwith ( ), andwithout (- - -) added
ascorbic acid1OG. b) cornparison of emulsion with added ascorbic acid/TX-100 after
Ohr(-”--),and4hr(--- ) melt-hold.

30
(0.14% by vol after melt dilution). When ascorbic acid and 10G are present
in the emulsion at

the timeof dye addition, only a5 o/o loss of dye occurs during the 3 hr incubation40
atOC.

In contrast, if these reagentsare excluded from the emulsion, ca

90% of the dyeis

consumed by oxidation over thesame period of time. For tricarbocyanine dyes with slightly

higher oxidation potential, additionof a reducing agenffsurfactant combination practically

prevents oxidative decomposition of dye. Figure11b shows reflectance spectra obtainedafter

melt-hold timesof 0 and 4 hr for a companion emulsion 40

at OC containing dye2.

Overall, these datashow that slight improvementsin dye stability couldbe obtained by

the single addition of ascorbic acidto an emulsion-melt. Substantial gainsin sensitizing dye

stability canbe achieved, however,if a surfactantis added together with the antioxidant
for the

purpose of minimizing dye aggregation and oxidation.

These experimental observationsmay be mechanistically interpreted

in terms of kinetic or

thermodynamic differences between monomeric, adsorbed, and aggregated dye. With

respect to SchemeIV, factors that promote adsorption

of the dye to the grain surface, suchas

the presence of added bromideion in the emulsion, and substitutionof dye with thiomethyl

groups, have a stabilizing influenceon the dye. This stabilizing effect suggests that the

AgX as compared
oxidation rateis lower, or oxidation potentialis higher, for a dye adsorbed to

to the same dye in aqueous solution, Le., koxl > kox3. Furthermore, the stability

improvements notedby the presence of the deaggregating surfactant implies that the rate of

oxidation of the aggregated dye is than that for the monomeric dye, Le., kox2 > kox 1. In

be interpreted to mean that the

terms of electrochemical potential, these data may, likewise,

effective one-electron oxidation potentialof a dye aggregateis Jesspositive than that of the

corresponding dye monomer.

F. OptimumReducing Agent / SurfactantCombinations.

In principle, a varietyof reducing agents and surfactants could be employed

to minimize

oxidative decompositionof an infrared dye provided the combinationof reagents possessed

the proper redox potential, deaggregating property, and water solubility. An effective

reducing agentmust have an oxidation potential that

is less than or equal to theEox of the

31
infrared sensitizing dye,(I0.5 V vs Ag/AgCI). If the redox potential of the reducing agent is

lower than about 0.2 V, however, the reducing agent itselfwill not be sufficiently stablein an

emulsion environment and can cause formation

of photographic fog. Table Ill lists chemical

structures andredox potentials for various organic reducing agents that were examined
at

equal levels(200 mg/mol Ag) in combination with a fixed level

of Triton X-100. The oxidation

reactions in 0.1 M phosphate buffer(pH = 5.5) for ascorbic acid,dihydroanhydropiperidino-

hexose reductone (DAPHR), piperidinohexose reductone (PHR), 4-methyl-4-hydroxymethyl-

1-phenyl-3-pyrazolidone (MOP) and hydroquinone were irreversible and the oxidation

potential values reported are the peak potentials

Ep obtained from the jrreversible cyclic

voltammetric waves. As such, Ep serves as a rough estimate of the thermodynamically

significant oxidationpotential and may be usedto rank order reducing powerof these agents at

pH 5.5. Because Ep of the reducing agents and EOx

of the dyes were determinedin different

solvents, thetwo sets of oxidation potentials cannot be directly compared. The stabilizing

activity of each reducing agent was measured by monitoring the melt stability of dye
in the
4

AgCl emulsion over 4a hr period. Results indicate that the reducing agents
of a Table Ill with

the lowest valuesof Ep to be the most effectiveat preventing oxidative decompositionof dye.

At these concentrations the relative stabilizing ability

was found to be in the order: ascorbic

acid = DAPHR >> PHR = MOP >> hydroquinone.

Also shown in Table Ill are chemical structures for several surface active agents that were

examined (at a levelof 0.14% by vol after melt dilution) for use
in conjunction with ascorbic acid.

Among these surfactants onlyTriton X-100 and 10G exhibited suitable deaggregating

properties. Despite similaritiesin structure, IGEPAL was considerably lesseffective than 10G

when used at comparable levels. Tween 80 and glycerol were not suitable under the

described experimental conditions .

32
TABLE 111. Reducing-Agent and Surfactant Compounds.

Reducing Agent Structure Ep, V a

Reductant

Ascorbic Acid OH OH 0.45

DAPHR 0.23

PHR 0.53

MOP 0.68

Hydroquinone 0.66

Surfactant

1OG

IGEPAL

GlyceroI

at a
a. Ep is the voltammetric peak potential measured versus Ag/AgCI
scan rate of 0.1 V/s in pH 5.5 buffer.

33
 G. CoatedEmulsionsContainingReducing Agent andSurfactant.

The use of reducing agentlsudactant combinations to stabilize infrared dyes

in practical

emulsion systems would have many advantages, particularly

in manufacturing processes

where extensive (up to 8 hr) melt-hold times may be encountered. Photographic

coatings of stabilized emulsions were prepared

to ascertain if the noted stability gains are

modified by the processof physically coating the emulsion, and

to determine if the stabilizing

reagents adversely affect photographicfog levels. Evaluations were madeon the same AgCl

emulsions describedin Figure 1l a using dye5. The emulsion containing ascorbicacid/lOG,

and a companion emulsion

in which these reagents were absent, was held for 3.5athr
40 OC

and then coated on anESTAR support at 200 mg/ft2 Ag and 483 mg/ft2 gelatin.

Diffuse reflectance spectra recorded for the coated emulsions were nearly identical to the

spectra recorded for the originating liquid emulsion (Figure

la).
1 This result indicates that the

process of coating and drying the emulsion does not substantially influence the dynamics
of

dye 5 oxidation and showsthat the expected stability gains

are indeed realizedin the coated

material.

A preliminary photographic evaluation was carried

out by obtaining wedge spectrograms

(10 s with the Wratten898 filter) of the coated emulsions.A comparison of the sensitometric

data for the emulsions reveals a relative speed gain

of 0.45 log E at the wavelengthof

maximum dye5 sensitivity (860 nm) for the emulsion containing ascorbic
acid/lOG. This

substantial gainin relative spectral speed that results from the presence
of the dye-stabilizing

reagents was accompaniedby very minor relative changesin fog and desensitization levels.

Additional photographic experiments are necessary

to fully quantify these sensitometric

advantages.

IV. Conclusions.
Infrared spectral sensitizing dyes from the tricarbocyanine and enamine tricarbocyanine
class are prone to oxidative decomposition
in silver halide emulsion meltsto give the
corresponding monooxidized radical dication formsof the dye. Although both silver ion and
molecular oxygenmay serve as oxidizing agentsin the emulsion environment, the observed
of dye
dependence of dye oxidation rate on pH suggests that the predominant mechanism

34
decay involves aerial oxidation. The important dye features that influence stability were defined
by comparing results for a series
of variously structured tricarbocyanines dyes.
Dye stability in
AgCl and AgClBrl emulsions was a complex function
of the dyes' oxidation potential,
aggregation propensity, and adsorptivity
to silver halide. The most reactive dyes were those
with the lowest oxidationpotential that had the greatest tendencyto form H-aggregates. A
comparison of the dye stability results with available photographic data indicates that dyes
demonstrated to be unstable in these studies also gavepoor photographic performance. Of
most significance was the discovery that substantial improvements
in the emulsion stabilityof an
infrared sensitizing dyecould be made by the additionof a mild reducing-agent/surfactant
combination to the emulsion melt prior
to the additionof dye. Initial experiments to examine the
practical utility of these reagents indicates that enhancements
in dye stabilitycan be obtained
without adverse effects on photographic fog levels. Use of the procedures describedin this
report to improve themelt stability of sensitizing dyes should lead
to improved manufacturability
of infrared-sensitizedphotographic materials.

Acknowledgements. We wish to thank A. Adin, A. A. Muenter, and R. L. Parton for many

helpful discussions andA. Cameron for computer calculations.

35
References and Notes.
1. (a) A. D. Gingello and V. N. Purushotham, "Photographic Film System for Laser Diode

Imaging", TR212603P, 1986. (b) A. Adin, "Infrared Sensitization of a Silver Chlorobromide

Emulsion with Monomerically AdsorbedSensitizing Dyes", TR210152N,1985. (c) A. A.

Muenter, R. L. Parton, and J. L.Krieger, "Synthesis and Evaluation of Infrared Sensitizing

Dyes for False-Sensitized Color Paper", TR230263V, 1987. (d) A. Adin and A. T. Wyand,

"Formulation Effects on the Performance of Infrared-Sensitized Color Paper Coatings

Utilizing

AgCl Emulsions", TR230284C, 1987. (e) R. Rader, "Infrared Spectral SensitizationStudies

of Several Enamine tricarbocyanine Dyes and Their Associated Addenda", TR166441 M,

1983.

2. (a) G. S. Proehl, "The Synthesis of New Sensitizing Dyes for the Infrared",TR196146L,

1985. (b) R. L. Parton, "Infrared Dyes for a Medical imaging Film", TR262552F, 1990.

3. (a) J. R. Lenhard and A. A. Muenter, "Electrochemical and Photographic Data forInfrared

Spectral Sensitizing Dyes", TR243482U, 1988. (b) J. R. Lenhard, "Relationships Between

Dye Structure and Reversible Electrochemical Potentialsfor Cyanine Dyes" TR210713H,

1985.

4. A. A. Muenter, P. B. Gilman, J. R. Lenhard, and T. L. Penner, "Mechanistic Consequences

of Anomalously Efficient Spectral Sensitization by Desensitizing Dyes", TR190576K, 1984

5. C. R. Berry, Photogr. Sci. Eng., 1975, 19, 93.

6. R. L. Parton and J. R. Lenhard, "Oxidation Chemistry of Dicarbocyanine Dyes - Implications

for Red Speed Keeping", TR1905057, June, 1984.

7. (a) A. A. Muenter and J. S. Kittle, TR157542, 1982. (b) P. B. Gilman andT. Kozselak,

"Spectral Sensitization of a Very Fined Grain Emulsionin the Infrared 'I, TR181067Q 1984.

36
8. J. Lenhard, J. Imag. Sci. 1986, 30, 27

9. D. J. Miner and P. T. Kissinger, Biochem. Pharmacol., 1979, 28,3285.

10. Dye geometries were optimized using MOPAC/pm3. Electron densiities were obtained

of these calulationswill be
for the radical dications using the INDOhohf approximation. Details

reported separately.

11. This scheme outlines a general oxidation route

for this class of dyes and shows only the

8,lO' isomer. The particular isomer or isomer distribution obtained for a given dye
will depend

on its chain substitution pattern.

12. Cyanine radical dications which contain N-C or C-C five-membered hydrocarbon rings

undergo oxidative dehydrogenation to yield stable products. Dye 4 is dehydrogenated in the

solid state to give a product that absorbsat 684 nmin CH3CN; see J. R. Lenhard, "Oxidative

Decomposition of Bridged Cyanine Dyes", TR230208G, 1986. Other tricarbocyanine dyes

discussed in this report, including the enamine tricarbocyanines, do not appearto undergo

dehydrogenation.

13. Electroreduction of some tricarbocyanine dimers at negative applied potentials (I

-0.3 V)

results in partial regeneration of dye.

14. R. L. Parton and J. R. Lenhard, J. Org. Chem., 1990, 55,49.

15. R. L. Parton, unpublished results.

16. Alternatively, these data may indicate that the radical-radical dimerization reaction
is rapid

and reversible. Similarly substituted dicarbocyanines are known to undergo reversible

37
dimerization in acetonitrile (see ref.6 ) . Further studies are requiredto elucidate the role of

dimerization.

17. The reduction potential of -1.1 V vs Ag/AgCI measuredfor dye 12 suggeststhis species

to be an effective green sensitizer for silver halide. The green safelight sensitivity of certain
may be attributed,in part, to formation of
infrared films that contain enamine sensitizing dyes

dimer oxidation-productsthat incorporate the hemicyanine chromophore.

18. TSSA readily adsorbsto silver halide andis thought to form an association complexwith

the dye on the AgX surface. Association complexes between TSSA and enamine

tricarbocyanines in solution hasbeen reported; see R. Rader and J. M. Harbison, "Studies of

Various Substituted Triazinostilbenesulfonates as Supersensitizers for Enamine

Tricarbocyanine Infrared Sensitizer" TR190575W, 1984.

19. (a) W. W. Wendlandt and G. G. Hecht, Reflectance Spectroscopy; Interscience, New

York, 1966, pp 46-90. (b) A.Herz, R. Danner, and G.Janusonis, Adsorption from Aqueous

Solution, Advancesin Chemistry Series No.79, Amer. Chem. SOC., Washington,D. C., 1968,

p. 173.

20. The calibration curvewas obtained usingdye 2. Ascorbic acid and surfactant were added

to the emulsion to prevent oxidationof the dye.

to the presence of
21. Dye reactivityon the AgCl emulsion was much more sensitive

potassium bromide and

APMT than wasthat on the AgClBrl emulsion.

New York, 1975, p 220.

22. H. A. Laitinen andW. E. Harris, Chemica/Ana/ysis, McGraw-Hill,

23. The effect of AgX morphology and composition on infrared dye stabilitywill be addressed

in a separate report.

38
24. When compared at similar concentrations and temperatures, tricarbocyanine and

enamine tncarbocyanine dyes were,in general, much more reactivein the emulsion thanin

methanol or methanol (80%)/water(20%) solutions. This result may reflect an increase in dye

reactivity becauseof adsorption to AgX, but it ismost probably due to simple

pH differences

between the emulsion and solvent systems.

25. A. H. Herz, Adv. Colloid Interface Sci., 1977, 8,237

39
Accession Numbers
186255
1671 12
57351 7
135795
573519
571 700
58 1 874
579226
571 703
567277
33871 3
555246
084555
908441
535542
598671
605254
559623
176434
543464

40
Stabilization of InfraredSpectralSensitizing Dyesin SilverHalide
EmulsionMeits.

'd. R. Lenhard

B. R. Hein

41