T H E OPTICAL SENSITIZING OF SILVER HALIDE EMULSIONS Part I. The Adsorption of Orthochrome T to Silver Bromide
BY S. E. SHEPPARD AND H. C R O U C H
Communication So. 337 from the Kodak Research Laboratories
The problem of optical sensitizing has received much attention on the practical side, and the number of valuable sensitizing dyes has been greatly extended. The theory of the action has been attacked mainly from qualitative experimental evidence, or in pure speculation, but very little quantitatively. The first questions are, how much dye is attached to silver halide and what is the nature of the combination? Earlier studies by one of the authors on the isocyanine dyes' indicated that spectrophotometric determination in non-aqueous solvents should afford a satisfactory method of estimating these dyes even in quite small amounts. A study was made therefore of the partition of a n isocyanine-orthochrome 'r-bromide between aqueous solutions and chloroform.
The partition was measured between chloroform and both M/zo (.os molar) and M/IOO(.0066molar) phthalate buffer solutions. I n these experiments cc. chloroform were shaken with 2 0 cc. of aqueous buffer and the dye determined spectrophotometrically. It was found that there was a large specific salt effect, in that the partition coefficients,
C =
concentration in aqueous layer, concentration in chloroform
mere much higher in the weaker buffer solutions of the same pH. S. E. Sheppard: Proc. Roy. Soc., 82A,256 (1909).
S. E. SHEPPARD AND H. CROUCH
752
TABLE I Partition Coefficient PH
+
Water Acid or Alkali
M/150
h1/20
I. 2.
3. 3.2 4.0
,013 ,003
4.05 5.00
. I25
5.40
89
6.00
.003
8.00
IO
.03 0.
10.00
The mechanism of this salt effect, whether a salting out effect on solubility or otherwise, requires further investigations, but is probably similar t o the depression of solubility of non-electrolytes by strong electrolytes in general. It appeared to be of interest to ascertain the effect of soluble halides on the partition between water and chloroform.
TABLE I1 Molar Ratio of Dye t o Halide
Concentration of Halide .o
Partition of Dye in Aqueous Layer NaCl
KBr
I9 '173
I :Io
.OOIN
-
.I9. 'I75 ,040
1:100
. O I S
,081
,005
I :Io00
. IN
.o j o
I :I
.O O O I N
'
-
KI '
I9
,008
-
Again 5 cc. chloroform containing 0.0010gm. dye were shaken with 20 cc. aqueous solution. These figures may represent the relative solubilities of the chloride, bromide, and iodide dye salts, on which is superposed the isohydric depression of solubility by a common ion, but require extension to determinations a t various pH values. State in Solution and Absorption Curves The isocyanines' are converted by acids into a colorless form. At any given pH there is an equilibrium between the colored and colorless forms, and this can be followed by the absorption spectra in the visible and ultra-violet.* 1
S. E. Sheppard: J. Chem. Soc., 95, 17 (1909);Phot. J.,48,300 (1908).
* Cf. also R. Brode: J. Am. Chem. SOC., 46,581 (1924).
OPTICAL SENSITIZING OF SILVER HALIDE EMULSIONS
7 53
The curves in Fig. 2 show the values obtained in aqueous solution. It will be seen that as the “colorless” form increases, the ultra-violet absorption increases. The solubility of the dye in water diminishes with increasing pH, because while the colorless form seems to be quite soluble, the colored fonn is very little soluble in true solution, but readily disperses to colloidal solution. I n
-u .I6 .I4 W
0
u .I2
L
06
0
FIG.I
acid solution the dye migrates to the cathode-corresponding to its dissociation into a large organic cation, and a halide ion. I n alkaline solution (pH 1 3 ) no migration was observed in the electric field, but the experiments require extending, as the colloid niicelles are probably charged. The tautomeric change is supposed to consist in the realignment of valencies in a conjugated chain of double bonds. In organic solvents the solubility of the colored form is greater, the location of the absorption maximum slightly affected.
S. E. SHEPPARD AND H. CROUCH
754
5.0
ABSORPTION SPECTRA OF ORTHOCHROME T BR. A T pH VALUES FROM 4 8 T O 5.8
I PART IN S,OOO PARTS BUFFER 2 CM. LAYER. 4.0
3.0 Y
rv)
B n
2.0
1.0
300
SO0
400
WAVELENGTH, pp. FIG. 2
600
OPTICAL SENSITIZING OF SILVER HALIDE EMULSIONS
ABSORPTION SPECTRA OF ORTHCHROME T BR. 5.0
I PART I N 40,000 PARTS SOLV€NT. 10 CM.LAYER. x
0
9
CHLOROFORM PEAK
8
METHYL ALCOHOL
1'
- 560pp - 565pp
illli
=WATER
IO
06
02
40
300
400
500
WAVELENGTH, p p . FIQ.3
600
755
S. E. SHEPPARD AND H. CROCCH
756
It is important to note that the partitions between aqueous solutions and chloroform indicate that the colored form is very little soluble in water, and also is much less polar than the colorless form. The partition may correspond to the following equilibrium system:Water
w
3
E
H
[i] 1 [D
z
Chloroform
D
-
Colored ionized
Br
T
I
1
I I
1 1
I
i l I
Brl
Colored
-- + [n Br]
Colored form
Br solid
or a hydroxyl may replace bromide (Br) with increasing OH-ion concentration. Insolubilization of the colored form by salts, and halides, would give higher concentrations of dye in chloroform. The nature of the aqueous and chloroform solutions respectively is of great importance for the adsorption and sensitixing processes. Adsorption to Silver Bromide The method used was as follows: A suspension of silver bromide was prepared giving microscopically measurable grains. The grain-size frequency was determined by methods already described.’ To reduce interference of the gelatin with the dye adsorption, the adsorption experiments were carried out a t 5o°C, the gelatin being only I per cent in concentration. This procedure is not entirely satisfactory, and a method of enzyme stripping of the gelatin has been developed for use in later work. Five cc. of the silver bromide emulsion containing a known amount of the silver bromide were diluted to 2 5 cc. with water, adjusted to a known pH with alkali or acid, and containing a known amount of dye, were added as concentrated alcoholic solution. The reaction system was placed in stoppered amber glass bottles in a . thermostat a t jo°C and shaken for one to two hours. The silver bromide was then separated by centrifuging, a t about j o T . The residual dye in the liquid was determined by extraction with CH.Cls and spectrophotometric determination. The dye actually adsorbed to the 1 Cf.S. E. Sheppard and A . P . H. Trivelli in V?l. III., Eder’s Handbuch Phot., Part I., Luppo-Cramer, “Die Grundlagen der photographischen hegatlvverfahren.”
757
OPTICAL SENSITIZING O F SILVER HALIDE EMULSIONS
TABLE I11 Data for Adsorption Curves PH 5 . 5 Mixture used : 5 cc. precipitate (.4N AgBr, .o8N KBr - 400 sq. cm. projective area per cc.) 2 0 cc. water. I cc. Methyl Alcohol with Dye. TotalDye L sed Grns.
Final Concentration Grns.Kept in Sol.
Dye Adsorbed to 5 cc. Ppt.
,001
,00039
,0006I
,002
,0012
.00080
,004
,003I 3 ,0068 ,0084 ,0114 ,003I3 ,0065 9 ,00925
,0008 7 5
.OI .02
.04 ,004
,008 ,016
0000I 5
,003I 8 ,0116 ,0286 ,00087 5 ,00141
,0067j
Molarity of Dye left in Solution
Molecules Adsorbed per cmz
3.5 4.6
x
5.04
18.4 66.6 168. 5.04 8.I 38.8 .03
-10
,343x 2.76
6. 7.4 IO.
2.76 5.8 8.15 ,0134 ,0134 ,0178 ,064
. 00002 . 0000j
.OOOOI~
,000035
,0001
.00002
.000080
,0002
,000074
.ooos
.000017
,001
,000556
.ooo126 ,000330 ,000444
.00002
.00000~
.00001~
.09
,005
.00005
. 0000I4
,000036
.2I
,012
,0001
.000020
. o00080
,0004
,000 II
,0008
,0003I
.00029 ,000488
,001
,00045
,00055
,001
.00040 ,0045 ,0039 ,0036 ,0035 ,000460 ,000473 ,000475
.00060 ,0035 2
,008
,008 .OI .OI ,001
,001
,001
I
.00000~
.2
.46 ,73 I .96 2.55
.46 1.67 2.81 3.16 3.46 20.2.
,004 I 8
23 ' 7
,0074
42.6 43 ' 5 3.06 3.03
.00755 .000532
.ooo527 ,000525
3.02
88.j X 92.5 89.
,008 ,008 ,008
,00445
.OI
,00552
,00448
110.
.OI
,005IO
.OI
,0049 ,004
.02
,008
.OI200
.02
,0069
.OI310
97.5 80. 161. 137'
,00465 ,00448
.00355 ,00335 ,00052
.00600
0-4
I .06
. 15 .49
,018 ,097 '274
.39 .35 3.96 3.42 3.2
3.1
.41 .42
.42 3.9 X 4.1 4.0
4.9 4.35 3.6 7.1
6.I
IO-(
758
S. E. SHEPPARD AND H. CROUCH
TABLE 111 (Continued) Data for Adsorption Curves pII 3.0 (HCl added) Mixture used: 5 cc. precipitate (.4 iK AgBr, .08 S KBr per cc.) 20 cc. water. I cc. hlethyl alcohol with Dye. Total %zd Gms
Final Concentration Gms. Kept in Sol.
Dye Adsorbed to 5 cc. Ppt .
-
400 sq. cm. projective area
.0000333 .0000762 .0000763 ,0001638 ,0003476 ,000348
.oooo167 .ooooz38 .0000237 .0000362 .0000524
,001
,0008I 2
.oo01880
,001
,000828
.000172
,002
,001668 ,001667 ,003572 ,00357 5
,000332 ,000338 .000428 .000425
2.45
.005522
,000478
2.75
.005523 .007485 ,007492
,000477
2.75
,0005I j
2.97 2 ' 93
.00005 ,0001 ,0001
.0002
,0004
,0004
,002
,004 004
,006 ,006 ,008 ,008
pH 3.8 (HBr added) .00005 ,0000305 . ooai .0000728 ,000r6.i ,0002 .000358 ,0004 ,002 ,001726 .02 .0165000 ,008 ,007632 .OI ,009689 '01 ,009609 ,012 . 01 I47 2 . or4 ,01269 ,016 ,01402 ,018 ,01572 ,020
,01772
,020
,01776 .003724
,004
. 0000 5 2 0
.000508
.0000195 . ooooz 7 2 .0000337 .0000413 . 0002 740
.0034800 ,000368 .0003 I I ,000393 .000528
.0013I ,00198 .ooz85 ,00286 ,00238 .ooo2 7 6
Molarity of Dye left in Solution
Molecules Adsorbed per cmz
,096 X , I37 '137
10-l~
,029 X IO-^ .06 j ,067
,208
'i44
,302 ,300 1.08 .99 1.915 1.92
,305 ,306 ,715 .73 1.47
1.47 3.14 3.14 4.85 4.85 6.6 6.6
2.47
. I I3 x
10-10
.02 j
'I57
,064
19s .238
.147
'
1.58 20. 2.12
1.79 2.26 3.04 7.55 11.4 16.4 16.4 13.7 I ' 59
,315 1.51 14.5
6.7 8.6 8.6 IO.
11.2 12.3 13 ' 7 15.1 15.1
3.3
X
IO-(
OPTICAL SENSITIZING OF SILVER HALIDE EMULSIOXS
759
silver bromide was also directly determined by dissolving the silver halide in thiosulfate solution, extracting the dye with chloroform and determining with the spectrophotometer. The sum of dye extracted from silver bromide and residual dye should equal the amount of original dye, and this was fairly closely fulfilled.
10
+
-
0
8
8
I
’$
3
)
I
A D S O R P T I O N OF ORTHOCHROME T Br T O SILVER
BROMIDE
~
I
I
I
L
II
7 -
b
CURVES COMPUTED FROM LANGMUIRS ADSOIPTIOU EQUATION
RESIDLUL CONCENTRATION OF D Y E I N SOLUTION M x 1 0 . ~
Fro. 4
Adsorption Function The adsorption data obtained did not fit any modification of the (Freundlich) adsorption isotherm. The limit values in the horizontal part of the curves were taken as saturation values, and Langmuir’s theory’ of unimolecular layers applied in the form: TW
? =
+
I.( a/v1 gram molecules adsorbed per cm2 average life of adsorbed molecule (time of relaxation) fi = gram molecules impinging on surface per cm2 per sec. a = fraction of molecules which adhere v1 = rate of evaporation (or solution) of molecules from surface. Assuming a/v1 = K, is a constant for a given solid and adsorband, then 7 = n,/X K, where no = number of atomic (molecular or ionic) spaces which can be filled 1
where
7 = 7 =
N
= Avogadro’s number = 6.06 X 1oZ3
‘ 3 . Am. Chem. SOC.,40, 1368 (1919).
760
9. E . S H E P P A R D A N D H . C R O U C H
and
M
I 2 K
iv where
C
M T R
= = = = =
KThZ
.c
X IO-7C concentration (.: osmotic pressure p molecular weight of adsorband absolute ternperature gas constant 83.2 X IO^. 1.2
The equation was used in the final form:
From X-ray data’ the distance between Ag and Br atoms in the lattice of AgBr is 2.89 A.U. an$ the distance between centers of Br ions in an octahcdral face is d T X 2.89 A.V. and each ion occupies 14.45 sq. A.V. I n I cm2surface of AgBr there are 6.93 X 1oi4Rr ions. From the integrated area-frequency curve the projective area per cc. of emulsion was 400 cm2,and the total surface 800 cm2.2 Hence for I cc. AgBr emulsion we have 800 X 6.93 X 1014, i.e., 5 . 5 5 X 1oL7Br ions on the surface. And a t a pH 5 . 5 there were 5 x I O - ~ O gram mols dye adsorbed a t saturation per cm2. This gives cules dye per Br ion, or
I
molecule dye per
2.3
.
5X
1 0 - l ~X
6.9 X
6 X 1oZ3 1oL4
mole-
Br ions.
Discussion The adsorption curves obtained and the saturation values calculated are a t first sight in agreement with the following theory. We may suppose that to bromide adsorption is due to the electrostatic attraction of a dye cation [D] anions of the AgBr surface. It has been shown by Trivelli and Sheppard3 that in AgBr crystals found in gelatin emulsions the octahedral surfaces constitute much the largest proportion. These surfaces will consist either of silver ions or bromide ions; in the presence of slight excess of bromide, of the latter.3 The limiting adsorption density would then be I dye cation or molecule to every bromide ion of the lattice surface. The increase of adsorption density with pH may then be interpreted as due to increased concentration of the colored dye form, as cation. The second rapidly ascending portion of the adsorption curves, on this view, corresponds to the adsorbed dye (from molecular solution) acting a t a certain density as nuclei for the precipitation of colloidally dispersed dye. According to this, the position of the second ascending portion should approximately correspond to the true solubility of the dye in the aqueous solution of given pH. ’R. B. Wilsey: Phil. Mag., 42, 262 ( 1 9 2 1 ) . *Assuming the grains to be flat tablets. 8 “Silver Bromide Grain of Photo raphic Emulsions,” 8 . P. H. Trivelli and S. E. Sheppard, Monographs of the Theory of Piotography, No. I , Eastman Kodak Company ( 1 9 2 1 ) .
OPTICAL SENSITIZIXG O F SILVER HALIDE EMULSIONS
761
Although this interpretation appears reasonable, there are certain facts not in harmony with it. From the data on partition of the dye between aqueous solutions and chloroform, it appears as if the colored form is very slightly, if a t all, soluble, in water, and also is very little ionized. As a nonpolar, or less polar form, it shows correspondingly greater solubility in chloroform. If the dye is first adsorbed to the silver bromide as ionized, colorless’ cation, and then transformed to the colored form a t the prevailing alkalinity, it is difficult to see why the adsorption should increase with increasing pH, which lowers the concentration of the colorless form. An alternative hypothesis is that the colloidally disp-rsed colored form is “salted out” on to the silver bromide by the surface excess of bromide ions, and that this first proceeds to a surface saturation, as the concentration of dye is increased, and then passes over into mechanical adsorption of the colloid dye at higher concentrations. I n this case also it must be supposed that preliminary “patches” of adsorbed dye act as nuclei for further precipitation and adhesion. A more complete investigation of the aqueous solutions of the dyes, and of the adsorption process is necessary before decision between ionic-molecular and colloid adsorption can be effected. Provisionally, the former hypothesis appears the more probable. Sensitizing Diffusion experiments into gelatin jellies a t different pH values show that the diffusion rate increases with decreased pH. Experiments were run with dye solutions adjusted to a given p H and diffusing into cylinders of gelatin jelly at the same pH. After given periods, the jelly cylinders were cut in sections, and the total dye determined in each section. The plot of dye concentration against depth gives a measure of the diffusion. The results showed that the actual amount of dye passing the cross-section of gelatin jelly surfuce was approximately constant, as measured by total area of curve, but the slope was lower, the lower the pH, indicating that the diffusivity diminished with rising pH. This agrees with the tendency of the colored form to form colloidal micelles of low diffusivity. Since it is the colored form which actually sensitizes, sensitizing by bathing with these dyes can sometimes be most uniformly effected by bathing in an acidulated solution of the dye, then adjusting with alkali to a higher pH (cf. absorption curves). The diffusion experiments indicate that a t a p H 5 to 8 there is some dye present in the colored form in molecular solution, but the proportion relative to colloidally dispersed dye diminishes as pH is increased. From rough estimates, the amount of dye adsorbed a t saturation appears to be very considerably higher than the amount taken up by silver halide for optimum sensitizing. Provisionally, the mechanism of optical sensitizing, on the basis of the theory of adsorption proposed within, is as follows: Supposing that colored dye cation is electrostatically held to bromide ion, but that this original electrostriction passes into homopolar combination, in agreement with the con1
That is, visibly.
762
S . E. SHEPPARD AND H. CROGCH
clusion that the colored form is notably less polar tkan the colorless, on absorption of light by this in its own absorption region’ an electron is freed, possibly from the bromide ion, and a silver ion reduced, or indirectly by the “reduced” dye cation. This mechanism would give only one Ag atom for each dye molecule adsorbed to the silver halide. Recently2 Leszynski has published evidence that (with erythrosin) up to 2 0 silver atoms may be photocherhically reduced per dye molecule acting as sensitizer. He suggests that the photoelectron may travel some distance through the silver halide crystal, and effect a chain reaction of rather high efficiency, or that the reduced silver continues to act as an optical sensitizer. As an alternative to this, it may be suggpsted that in the photodecomposition of adsorbed dye on silver bromide, the dye molecule is practically exploded with release both of several free electrons, and also of very active free radicals. The photochemical efficiency might then be considerably greater than unity, but would probably be a pronounced function of the intensity of the illumination. That the optical sensitizing is connected with the photodecomposition of the dye is supported by the fact that the addition of silver ions to aqueous solutions of the dye greatly accelerates its decomposition (bleaching) by light. Our experiments on this indicated that below a molar ratio of about 1.5 A i to I mole dye, little or no acceleration of decomposition was produced, while from this point the acceleration was approximately proportional to the silver concentration. Khether the apparent threshold is significant or not has not yet been determined. Rochester, N . Y., January 9 , 198’8.
Modified by the deformation effect of adsorption to silver bromide. Cf.Fajans: Z. Electrochemie, 28, 499 (1922). * Z. wiss. Phot., 24, 261 (1926).