Products
B. W. S t o r y a n d Y. A. K a l i c h e v s k y
Research and Development Laboratories, Socony-Vacuum Corporation, Paulsboro, N . J.
I
N TH E petroleum industry it is necessary to measure what is usually called the“color” of various oils, both for purposes of controlling refinery operations and for the final pro
duction of oils sold under a color specification. Many m e t h o d s have been p ro p o s e d for color measurement and s e v e ra l are used for this purpose. Most of them are based on visual sensa
tion wherein the attem pt is made to m a t c h the c o lo r of the oil with a colored glass, or possibly with a s o l u t i o n of iodine or other c o l o r ed substance. The oil is then given a color v a l u e in terms of the scale of colored
glasses or solution strength which seems most closely to match it. These values are arbitrary, depending on the method.
In use, such methods are simple and rapid but generally in
accurate, because the oil rarely matches the colored glass or solution with respect to the three recognized factors in color sensation—hue, saturation, and brilliance. Hence the re
sult reported depends to a considerable extent on which of the three components of color sensation has the greatest appeal to the operator.
Petroleum products vary from water-white to black, pass
ing through various shades of yellow, orange, and red. No definite absorption bands appear in the visible spectrum. In passing from water-white, the first loss of transmittance ap
pears in the violet, which, as the color increases, moves pro
gressively toward the red. When the per cent transmittance is plotted against the wave length of the light, the slope of the curve is not very steep as compared to similar curves for many of the dyes. An oil may or may not approach 100 per cent transmittance a t the longer wave lengths, depending on its source and refinement, and the slope may vary for the same reasons. I t is evident, then, th at for equal brilliance, oils may vary considerably in hue and saturation. For this reason it has been practically impossible to measure color visually as a single value representing sensation with any accuracy, because different operators do not react the same to variations in hue and saturation, nor does a good solution of the problem seem probable as long as there is a personal equation in the determination. Therefore, the measurement of color of petroleum oils in the past has in reality been an attem pt to measure brilliance, which was then ar
bitrarily converted to a so-called “color” scale.
Ph o t o e l e c t r i c Ce l l
The photoelectric cell has often been proposed as a means of eliminating the personal equation, but for various reasons no practical method based on use of such a cell has been
adopted in the petroleum indus
try. However, a photoelectric cell has recently been put on the market (photronic cell, model 594, m a n u f a c t u r e d b y t h e Weston E l e c t r i c I n s t r u m e n t Co.), which combines to a re
markable degree the characteris
tics required to evaluate color s e n s a t i o n or, more accurately speaking, brilliance. The most important consideration here is th at the relative sensitivity for hue should be approximately the same as the human eye. Figure 1 illustrates the r e m a r k a b l e similarity for this cell and the normal eye. This means that, regardless of the spectral dis
tribution of the light transmitted by two substances, if the eye perceives one to be darker than the other, the cell will grade them in the same order. The somewhat higher sen
sitivity in the region of 4500 A. will, of course, cause the cell to grade certain light-colored oils slightly different from the eye, but for practical purposes this is not serious with pe
troleum oils. Darker oils have little or no transmittance in this region, so th a t what appears to be a substantial dif
ference in response is in reality of small consequence with most of the petroleum products.
This photoelectric cell possesses other characteristics which greatly simplify its adoption to color measurement. I t generates its own current in amounts sufficient to be measured directly, the output being about 1.4 microamperes per foot- candle. Its response is quick and it stabilizes rapidly without fatigue. I t is small, rugged, and inexpensive.
One of the essential features in a colorimeter is the repro
ducibility of the results obtained. Not only should a single instrument reproduce itself, but for its wide applicability it must give concordant results with all others of the same type.
According to the claims of the manufacturers of the above photoelectric cells, the spectral response curve of the cells can be reproduced within 1 per cent of the average, which is more than sufficient for practical color measurements. The results obtained in the authors’ laboratories with different photoelectric cells of the kind described are in substantial agreement with these claims.
If a light source of suitable and uniform intensity is placed a t a short distance from a photoelectric cell, the cell will generate some current. If the intensity of the light falling on the cell is changed by increasing or decreasing the distance between the cell and the lamp, the current will be respectively less or greater. If the light remains stationary but a glass cell containing oil or some other transparent colored sub
stance is introduced between the photoelectric cell and the light, the current will decrease because some light will be Commercial colorimeters fo r measuring the
color o f petroleum products depend on visually matching the unknow n w ith some standard scale o f color. A s the standards rarely match the unknow n with respect to all the attributes o f color—hue, saturation, and brilliance— a large personal equation is introduced, which leads to erratic results. A sim ple photoelectric colorimeter has been constructed which com
pletely removes the hum an equation and still grades the colors closely in accordance with nor
m al visual sensation. Petroleum refining opera
tions involving the measurement o f color m ay now be studied more precisely because o f the greater accuracy' o f this method.
214
I N D U S T R I A L A N D E N G I N E E R I N G C H E M I S T R Y 215 was necessary to move the light to attain the desired effect.
/ \ measured by the photoelectric cell, might be expressed either as a function of the current with a stationary source of light, or as a function of the distance, if the light is movable. The latter arrangement was preferred in constructing the colorime
ter, as it does not require a calibrated microammeter, and as it permits compensations for slight changes in the sensitivity of different photoelectric cells, lamps, or ammeters.
Se t-Up o f C o l o r i m e t e r
The colorimeter, which has been successfully used in the authors’ laboratories for a number of months, is shown dia- grammatically in Figure 2. I t consists essentially of a photo
electric cell connected with a microammeter, an oil cell, and a uniform source of light. The photoelectric cell has already been discussed, and does not require further description.
The oil cell is 5.2 mm. in depth and 50 mm. in diameter.
The diameter is immaterial, providing it completely covers the face of the photronic cell. While the depth may he varied also, a standard depth must be established eventually in order to provide complete agreement with different instruments on all oils.
The light source is a 100-watt, 115-volt Mazda projection lamp, manufactured by the General Electric Company.
Deviations from the above specifications are allowable with
out greatly changing the results, provided the spectral energy distribution of the light source remains about the same.
Theoretically, the spectral energy distribution of daylight is d e s i r a b l e in a light made stationary and attached to the frame of the instrument.
The oil cell was fastened directly in front of the photoelec
tric cell, also in a fixed position. The lamp was made mov
able, in order to vary a t will the intensity of light falling on the oil cell. A special shield was provided in front of the photo
electric and oil cells, to eliminate the possibility of errors due to the reflection of light from the walls of the room or from the nearby objects. In later designs, it was found preferable to make the lamp stationary and the cells movable and to later designs. By properly adjusting the depth of the oil cell, this shortening was accomplished without decreasing the working range of the colorimeter. The reading of the microammeter connected to the photoelectric cell is then taken and is used as the standard reading. This reading includes compensations for variations in sensitivity of dif
ferent photoelectric cells, lamps, or ammeters and also for reflections a t the glass-air and glass-oil interfaces. Theo
retically, this last compensation is altered somewhat by the intensity of the light falling on the cell and therefore varies with the position of the lamp. However, in practice this is of no importance. The amount is small and it cancels out in the final results because they are comparative and not ab
solute.
After obtaining the standard microammeter reading, the oil cell is filled with oil of unknown color. The lamp is then moved towards the cell until the microammeter gives a reading equal to the standard. The distance between th e photoelectric cell and the lamp is determined and taken as the measure of the color intensity of the unknown oil.
I t is evident th a t the “color” indicated by the photoelectric colorimeter is in reality not the measure of the color or hue darker and lighter ones as they appear to the eye. This can
not be accomplished successfully by the present methods of measuring the color of oils. An oil which appears darker
216 A N A L Y T I C A L E D I T I O N Vol. 5, No. 3 medicinal clear oil, although it was found th a t the read
ings which were obtained in presence or absence of the oil cell were identical, provided the standard ammeter read
ings were properly adjusted. Table I shows a very good agreement between the calculated and experimental values.
The agreement is believed to be within the limits of the ex
perimental error involved in these measurements and more
Fi g u r e 3 . Ca l i b r a t i o n o p Ph o t o e l e c t r i c Co l o r i m e t e r w i t h Fl i c k e r Wh e e l s
b y flicker-w heel) C a lc u la te d O b served
% C m . C m .
100 1 0 0 .0 1 0 0 .0
9 0 9 4 .8 9 3 .6
80 8 9 .4 8 7 .7
7 0 8 3 .7 8 1 .8
60 7 7 .5 7 5 .7
50 7 0 .7 6 9 .0
45 6 7 .1 6 5 .2
40 6 3 .2 6 1 .0
35 5 9 .2 5 7 .2
30 5 4 .8 5 3 .0
25 5 0 .0 4 7 .8
20 4 4 .7 4 3 .5
15 3 8 .7 3 7 .0
10 3 1 .6 3 0 .3
6 2 4 .5 2 4 .0
4 2 0 .0 1 9 .5
2 14. 1 1 4 .0
ciprocal brilliance. In practice this is easily accomplished by reversing the scale of the colorimeter and placing the one hundred and not the zero point a t the photoelectric cell.
The resulting scale, which is designated as the color intensity scale, is shown on Figure 3. I t is evident th a t it bears the same simple relationship to the distance as the brilliance scale and is also plotted as a straight line on double logarith
mic paper, because the only difference introduced in the above formula is the substitution of (100 — D )1 for Z)s. In order to express the color intensities in some convenient units, the color intensity corresponding to the theoretical position of the lamp in direct contact with the photoelectric cell was assigned a value of 1000.
T a b l e II. Co l o r In t e n s i t i e s o f Io d i n e- Po t a s s i u m Io d i d e So l u t i o n s'*
(S o lu tio n s co n ta in p o ta ssiu m io d id e d o u b le b y w e ig h t o f io d in e) Di s t a n c e f r o m La m pt o Ph o t r o n i c
than sufficient for practical purposes. I t has also been found th a t the variation of the correction factor for the reflection of light from the oil-glass and glass-air interfaces a t various dis
tances of the lamp from the photoelectric cell is too small to be of importance and can, therefore, be neglected.
Ta b l e I . Th e o r e t i c a l a n d Ex p e r i m e n t a l l y De t e r m i n e d Di s t a n c e s o f t h e Ph o t o e l e c t r i c Co l o r i m e t e r Co r r e
s p o n d i n g t o Eq u a l Tr a n s m i t t a n c e s o f Li g h t Di s t a n c e
Io d i n e Co n c e n t r a t i o n Ce l l
M g . / 10 cc. Cm.
1 .6 0 9 3 . 0
3 .2 0 9 0 . 5
3 .4 0 9 0 . 8
4 .4 S 8 8 . 5
6 .5 0 8 6 .0
1 0 .0 5 8 3 . 0
1 3 .0 8 0 .2
1 3 .8 7 9 . 5
1 5 .0 7 8 . 5
5 0 .0 6 1 .7
6 0 .0 5 9 .2
210 4 1 .5
281 3 7 .0
282 3 8 .2
30 5 3 6 .7
581 2 9 .4
663 2 8 . 4
1270 2 3 . 0
° S p e cific g r a v ity o f th e se s o lu tio n s is fa ir ly c lo se ly r ep resen ted b y th e form u la: S ** 0 .9 9 7 ■+* 0 .0 0 0 2 3 1 2 /, w here S ■* sp . gr. a t 2 5 ° C . a n d I “ io d in e co n ce n tra tio n in m illig ra m s per 10 cc . o f so lu tio n .
An attem pt was made to calibrate the photoelectric colorimeter in terms of iodine solutions of known concentra
tions. Solutions containing iodine and double its amount by weight of potassium iodide were used. Iodine was se
lected, as the iodine colors are more or less familiar to pe
troleum chemists and as this substance is easily obtained in
(MTAWX F BOM PHOTDCUCTltlC U lL TO UfeMT 5CD9U CM.
0 0 8 . 5 0 40 3 0 6 0 7 0 »
On the basis of these findings it is simple to calibrate the instrument in terms of visual brilliance, by taking the square of the distance between the photoelectric cell and the posi
tion of the lamp a t the point a t which the reading is taken.
For this purpose, however, the distance should be expressed as a per cent of the maximum distance of the lamp from the photoelectric cell. I t is therefore preferable to divide this maximum distance into one hundred equal parts instead of measuring it in generally accepted units of length, such as centimeters, inches, etc. The resulting equation connecting the distance D and brilliance B
B = AD 1
(where A is a constant depending on the units chosen for ex
pressing brilliance or distance) can be easily expressed graphi
cally as a straight line on double logarithmic paper.
The calibration of the instrument on the basis of the above scale, which might be designated as the brilliance scale, is not suitable for industrial use, as refinery men are generally accustomed to speak of the color of the oil and not its
re-\
0 0 2 0 5 0 4 0 50 60 70 « 9 0 0 0
OttTANCC niOH PH0T0QXCTRJC COL TO U6MT SOURCE CM.
Fi g u r e 4 . Ca l i b r a t i o n o p Ph o t o e l e c t r i c Co l o r i m e t e r w i t h Io d i n e So l u t i o n s
(2 :l[r a tio of K I to Ii in so lu tio n )
pure state. The strength of diluted solutions was ascertained by titration with sodium thiosulfate, according to the standard analytical method. The calibration curve for the colorimeter is shown in Figure 4.
From the shape of this calibration curve, it is apparent th at the color intensities of iodine solutions cannot be proportional to iodine concentrations throughout the whole experimental
range (1, 2, 3, 5). I t is of interest, however, th at for the major portion of the curve the logarithms of iodine concentra
tions are a straight-line function of distances to which the lamp must be moved to obtain the standard ammeter read
ings. Because of the questionable value of the iodine colors as indicators of the color intensities of petroleum oils, no further work along these lines has been done.
The change in the color intensity of oils on diluting them with benzene was investigated in order to determine the additive character of the proposed scale. The experimental data presented in Table II I include also the Lovibond colors and the “true” colors (4) of the same diluted samples. The Lovibond colors were taken by several experienced operators and are given as average readings in order to obtain their best possible approximation. I t is believed th at the color in
tensities of diluted solutions obtained by means of the photo
electric colorimeter are as additive as might be expected theoretically, being superior in this respect to many other color scales now in use, but still unsatisfactory.
T a b l e III. Co l o r s o f Lu b r i c a t i n g Oi l So l u t i o n s i n Be n z e n e
Vo l u m e o r Lo v i b o n d
St o c ki n Co l o r 5 0 0
Be n z e n e Io d i n e Co l o r Am b e r Se r i e s Tr u e Co l o r So l u t i o n In t e n s i t y 0 . 2 5 -i n. Ce l l Co l o r In t e n s i t y
% M g ./10 cc.
X. M I D - C O N T I N E N T C Y L I N D E R S T O C K ( T R E A T E D )
1 0 0 3 1 8 3 6 0 7 0 0 4 0 5
5 0 1 1 6 2 3 0 3 5 0 2 4 5
2 5 4 4 1 5 0 1 8 0 1 3 0
1 2 . 5 1 9 . 2 7 7 8 3 6 2
6 . 2 5 9 . 4 3 8 3 8 2 8
3 . 1 2 5 3 . 8 1 5 1 5 1 2
S. C O A S T A L C Y L I N D E R S T O C K ( T R E A T E D )
1 0 0 1 2 2 2 2 0 3 2 5 2 5 0
5 0 4 1 1 2 0 1 4 5 1 2 5
2 5 1 6 . 6 5 9 6 1 5 2
1 2 . 5 7 . 8 2 9 2 9 2 2
6 . 2 5 3 . 3 1 5 1 5 1 0
3. P E N N S Y L V A N I A C Y L I N D E R S T O C K ( P E R C O L A T E D )
1 0 0 3 5 7 3 5 0 6 6 0 4 2 5
5 0 1 1 3 2 1 0 3 0 5 2 4 5
2 5 4 1 1 2 0 1 4 5 1 2 5
1 2 . 5 1 8 . 3 6 1 6 3 5 9
6 . 2 5 9 . 0 3 4 3 4 2 6
3 . 1 2 5 3 . 7 1 3 1 3 11
4. P E N N S Y L V A N I A C Y L I N D E R S T O C K ( P E R C O L A T E D )
1 0 0 2 2 . 8 1 1 2 1 2 5 7 4
5 0 1 2 . 9 6 4 6 8 3 8
2 5 6 . 5 4 0 4 0 1 9
1 2 . 5 3 . 4 2 2 2 2 1 0
Ac k n o w l e d g m e n t
The authors wish to thank W. V. Betts of this laboratory for making some of the measurements included in this paper.
Li t e r a t u r e Ci t e d
(1) D ossios and W eith, Z. Chem., 1869, 379.
(2 ) Jakovkin, Z . physik. Chem., 1 3 , 539 (1894); 20, 19 (1896).
(3) N o y es and Seidensticker, Ib id ., 27, 357 (1898).
(4) Parsons and W ilson, J . I n d . E n o . C h e m ., 14, 269, 1169 (1922).
( 5 ) R oioff, Z . ph ysik. Chem., 13, 327 (1894).
Re c e i v e d F ebruary 1, 1933. P r esen te d before th e D iv is io n o f P etro leu m C h em istry a t th e 8 5 th M e etin g of th e A m erican C h em ica l S o c ie ty , W a sh in g ton , D . C ., M arch 2 6 to 3 1 , 1933.
In making dilutions it has been found th at the same color intensities of diluted samples are obtained, irrespective of whether benzene or a “colorless” medicinal oil was used as diluent.
The proposed method for evaluating colors of lubricating oils is a means of measuring only the color intensity and does not take into consideration other attributes of color—i. e., hue and saturation. The determination of these attributes, particularly the hue of the oil, might be possible by means of the same photoelectric colorimeter but equipped with color filters. The information now available on this subject is not sufficient, however, to warrant its further discussion.
The 5.2-mm. oil cell has been used almost entirely in the development work with this colorimeter. For all oils which are darker than a bright yellow (2 NPA), a cell of this thickness is very satisfactory. As the trend is now toward lighter colored oils, the use of a thicker oil cell is desirable in order to obtain greater precision in these oils. Therefore, the use of one-inch cell and a shorter working distance (50 cm.) has been adopted for the new design. These changes do not affect the principles or operation of the instrument ex
cept th a t dilution of very dark oils may be required. This is considered to be justified in view of the greater precision throughout most of the scale and especially w ith pale yellow oils below 2 NPA.