V O L U M E 25. N O . 3, M A R C H 1 9 5 3
399
Approximately 1 mg. of the protein was weighed into a 2-inch length of borosilicate glass tubing (G mm. in outside diameter, 4 mm. in inside diameter), sealed a t one end. One hundred microliters of constant boiling (20%) hydrochloric acid (glass distilled) were pipetted in, and the tube was evacuated and sealed lit ha8 been shown by Jacobsen ( 5 ) that humin formation is catalyzed by trace metals in hydrochloric acid]. The sealed tube was then heated in an oven a t 150' C. for 6 hours [it was shown with preparation E of adrenooorticotroDic hormone
,
with 100 $. of 0.38 N barium hydroxide a t 150" C.for 1.5 hours in a sealed tube, On cooling, 20 PI. of 1.9 N sulfuric acid were added, the precipitated barium sulfate was centrifuged, and the pFocedure was continued as above. It was useful for the detection of tryptophan but was not used routinely owing to extensive decomposition of serine, threonine, cystine, and arginine. PHOMGRAPHIC DETAILS
The photographs of chromatograms (Figures 2 to 8)were taken by a combination of reflected and transmitted light on Kodalith Ortho, Thin Base, Type I1 film, developed with D-85 developer (film) and printed on Eastman AZO (1-4)paper (using D-72). Since the film was relatively insensitive to the yellow-brown spots due to asparagine, proline, and tryptophan, these areas received extra treatment with Farmer's Reducer. The best photographic records were, however, obtained in color on Eastman Ektachrome, but they unfortunately cannot be reproduced in this journal. For making routine records of ohromatogram, the authors employ direct contact prints on- Kodagraph . Standard contact paper, - toprint is reproduced in Fig AC
The authors wish to express their gratitude to C . K. Li, in whose laboratory this work was carried out, for his generous encouragement and support at all stages of the investigation. They would also like to thank Leon Messier for his helpful cooperation with the photography. Thanks are due to Armour and Co. for a gift of bovine serum albumin. Figure 8. Chromatogram of Bovine Albumin Hydrolyzate 14 miomgrams of NHn-N epp1i.d;
dimon-ion
Serum LITERATURE ClTED
(1) Block, R. J.. LeStrange, R., and Zweip. G.. "Paper Chromatopraphy." New York, Academic Press, Ino., 1952. (2) Consden, R.. Gordon. A. H.. and Martin, A. J. P., Bioehem.J., 38,224 ( 1944). (3) Jaoobsen, C:.F.,Compt. rend. frav. lab. Cc&berg, S h . chim., 26, 463 (194II1 .,. (4) Kawerau, E>,,and Wielmd, T., Nature, 168,77 (1951). (5) Levy, M., 17omp. mnd. t m a . lab. C d s b e 78, S h , chim., 21, 101
run 16 hours in each
To revsel methionine spot 40 miorograms of NHI-N mnst be(app1ied
(ACTH) (6) that 3 hours a t 150" C. was sufficient to hydrolyze all peptides; after 6 hours the amino acid pattern was unchanged; after 12 hours considerable loss of serine and threonine was observed and, after 24 hours, destruction of these amino acids was almost complete]. The seal was opened and the acid was evanarated over phosphorus pentoxide and Dotassium hvdroxide
~~~~~~
~~
,..~.
._- .....
and 200 PI. used per assay) was assayed for amino nitrogen by the quantitative ninhydrin method of Moore and Stein (8). The volume of solution (5to 10 PI.), calculated to contain 1micromole (14 micrograms) of amino nitrogen, was then applied to the paper in 2-SI. portions, drying between addition6 with a current of warm air from a hair drier. Alkaline hydrolysis was carried ant by heating 1 mg. of protein
(1936).
Li, C. H., J . Am. Chsm. Soc., 74, 2124 (1952). (7) MoFarren. E.F.,ANAL. CHEM.,23, 168 (1951). ( 8 ) Moore. S., and Stein, W. H.. J . Bid. Chhem., 176,367 (1948). (9) Partridge, S. M.. Biochm. J., 42, 238 (1948). (10) Patton, A. R.,and Chism, P., ANAL.CHEM..23, 1683 (1951). (11) Thompson, J. F., Zaeherius. R. M., and Steward, F. C., Plant Phusiol., 26, 375 (1951). (12) Woiwod. A. J., I . Gen. Micwbiol., 3, 312 (1949). (6)
RP_OEIVED for review September 12. 1952. Aceeptod December 12, 1952. Supported in part by the Rockefeller Foundation, New York, and the U. 8. Publio Health Servioe, Nhtiond Institutes of Health.
Displacement Spebctrophotometry I
ROBERT HOUSTC)N HAMILTON Temple University Schoolof Mezdicine, Philadelphia 40, Po.
I-
U T H E usual absorption photometry
of solutions, light of restricted wave length incident on the photocell or phototube is set to reed unit intensity after passing through a cell containing solvent or solvent plus the amount of impurities in reagents (blank). Then, light intensity being maintained constant, an identical cell is substituted, containing solvent plus the lighh absorbing molecules whose concentration is t o be determined. The decrease in light intensity is noted and the concentration of the salute is calculated from the absorption produced by known concentrations. The same results can be obtained by the addition to the light
".
I
. ..
"
.
patn 01 layers of solution O t constant thickness. Such addition can be accomplished by removal of a piece of plate glass immersed in the solution. For a given solution and wave length the effect of the glass plate itself on light trannsmittance will be eonstant. Either of two procedures c m be followed: (1) With the glass plate immersed a t right angles to the light beam, light intensity is set to read 100% (unity). The plate is then removed, and light intensity is read after removal. (2) Light intensity is allowed t o remain such that transmittance is close t o unity (between 80 and loo%), and transmittance is read exactly, before and again after removal of the glass plate. The difference in the
A N A L Y T I C A L CHEMISTRY
400
One of the problems encountered in laboratories using test tube type absorption cells for spectrophotometry is the securing and maintenance of sets of matched tubes of sufficiently close characteristics. In the attempt to circumvent this difficulty, glass plates were used to allow displacement and replacement of layers of the colored fluid of uniform thickness. Critical study of the use of such plates showed a higher degree of accuracy to be attainable than was expected. Results further indicate the possibility of eliminating need for transferring photometric solutions from the “working” test tubes in which previous analytical steps were carried out to an “optical” set of test tubes in which photometric readings are made. Limitations of the method are discussed, and possibilities are considered for selection of suitable displacement plates.
optical densities corresponding to the two transmittances, corrected by a similar density difference obtained with a blank, gives an optical density corresponding to that of the layer of solution displaced by the plate. It is possible to use the latter procedure because of the follo~ving mathematical relationship: Let To represent transmittance when the tube without plate contains only solvent, and T , and Ti represent transmittances of the cell containing colored solution with and without the glass plate, respectively. Then if optical density is represented by D , D1
and
= -log
Ti/To
= --log
DZ = -log T z / T o= -log
+ log To T? + log TO
An additional tube of the same size was used in some experiments as a water blank. A piece of almost colorless plate glass 6.1 mm. thick, known as “water-white,” n-as cut into sis strips 18 mm. wide by 200 mm. long. These strips were numbered 1 to 6. I n most of the esperiments plate 1 was used. REPRODUCIBLE PHOTOBlETRIC TEST SOLUTIONS OF GOOD ABSORPTION CHARACTERISTICS
Killard and Greathouse ( 7 ) showed that manganese could be kept in the heptavalent state in the presence of acid and periodate.
Ti
The optical density change produced by removing the plate is
D2 - Di = -log T (
+ log Ti
Hence, even though 2‘0 is much larger than unity (loo%), if TI and Tt fall on the scale, it is still possible, from the difference in their logarithms, t o obtain the optical density value of the displaced solution.
-I,
$ 1 z
r r
APPARATUS USED
Spectrophotometer used n-as Beckman hlodel DU, serial S o . 1440, fitted with a photomultiplier phototube attachment and with cell compartment for test tubes. As the cell compartment n-as not high enough to hold the 22 X 175 mm. tubes used in the Evelyn-type photometer, a circular hole about 60 mm. in diameter was cut in the lid, and 8 flange was attached. Over this was placed a tightly fitting cylindrical cap about 60 mm. high. By removal of the cap large test tubes could then be set in place. The spectrophotometer was powered by a Sorensen hlodel E-6/2-5 nobatron stabilized direct current power source working from 115-volt alternating current line. Only the “red-sensitive” phototube was used in this work, the photomultiplier tube not being used. The Beckman spectrophotometer was used in all experiments in which use of the instrument described next beloxy is not specified. The “density” or “optical density” readings given were read from the density scale on the instrument, on iyhich D = loglo
(To/T).
An Evelvn-type photometer ( 2 ) waF powered from a Sorensen voltage stabilizer through a small transformer. The light filter used was a Corning “narrow-band-pass” filter combination which gave maximum transmittance (determined) a t 515 mp. Three new cartons of Corning borosilicate glass test tubes, with lip, 22 X 175 mm. size, were opened and the first 10 tubes of the 7 2 in each carton were removed. The 30 tubes were numbered consecutively, cleaned, and used for the experiments listed below.
. 0r
I
0 440
Provided the cell is not moved during removal of the glass plate, and provided the latter is positioned in a plane perpendicular to the light beam and is larger than the light beam, the accuracy of results obtained should be dependent only on the precision of determining the change in light density produced by removal of the glass plate. Hence even scratched test tubes can be used, Furthermore] because the change in density is determined only by the characteristics of the glass plate, there is no need to select the tubes for exact uniformity of size.
O
a
.I
I
0 450
OPTICAL
l
0460
DENSITY
Figure 1. Distribution of Optical Density Values with Test Tubes Used S h o r i n g variation resulting from random selection of tubes. Beckmnn DU spectrophotometer. ? = 0.4484; u 0.00639; ( U X 100)/; = 1.43%
Stock Permanganate Solution. Exactly 200 mg. of C.P. potassium permanganate was weighed, t i ansferred to a 1-liter volumetric flask, and dissolved in n-ater. Potassium periodate (1.00 gram) was added and dissolved. Sirupy C.P. phosphoric acid ( 8 5 % ) was added in the amount of 50 ml. The mixture nas cooled, diluted to volume, and mixed. Dilute Permanganate Solutions were prepared by accurately measuring aliquots of the stock solution into volumetric flasks and diluting to volume with a solution containing per liter 1.00 gram of potassium periodate and 50 ml. of sirupy phosphoric acid. The solutions absorb light in a moderately uide, though not very smooth, band about the middle of the visible spectrum, \\here harrier layer photocell characteristics are good. They are accurately reproducible, and the dilutions are stable. VARIATION IN TEST TUBES USED
A dilute permanganate solution \$-as prepared as described above to contain 20 mg. of potassium permanganate per liter. The Beckman DU spectrophotometer (red-sensitive phototube) n-as set for 515 mp, and the sensitivity adjustment was set two revolutions from clockwise limit. Slit width employed was in the neighborhood of 0.15 mm. The slit was adjusted so that optical density was zero with a tube of water in place. Permanganate solution (about 15 nil.) was placed in each of the 30 test tubes chosen as described above. Each tube in turn was substituted, label forivard, and the optical density was ready for each. At intervals the light intensity was checked with the water tube.
V O L U M E 2 5 , NO. 3, M A R C H 1 9 5 3
401
Density readings obtained are shown in Figure 1. For this series, 2 = 0.4484 k 0.00639 (D.F. = 29).
e o = 1.43% f DESSlTIES
OBTAINED BY DISPLACEMENT WITH GLASS PLATE
OF
SOLUTION
Beckman Spectrophotometer. A dilute permanganate solution was prepared as described above to contain 60 mg. of potassium permanganate per liter and about 15 ml. of it was placed in each test tube. The spectrophotometer (red-sensitive tube) was set for 515 mg, and the sensitivity adjustment was set two revolutions from clockwise limit. Slit widths employed were in the neighborhood of 0.44 mm. They were so chosen that when the glass plate was immersed at right angles t o the light beam, the transmittance would lie between 80 and 100%. The same test tubes were used as in the preceding esperiment.
Evelyn-Type Photometer. d dilute permanganate solution was prepared as described above to contain 60 mg. of potassium permanganate per liter, and about 15 ml. of it was placed in each test tube. The Evelyn-type photometer had a light filter with maximum transmittance a t 515 mp. The same test tubes were used as in the preceding experiments. Each tube in turn was placed in the tube holder, with its label forward. The glass plate (the same one used above), sitting in the tube, was lined up by eye with a spot on the opposite wall of the room, so that its surface appeared to be normal t o the light beam. Voltage to the exciter lamp in the photometer was adjusted so that transmittance lay between 80 and loo%, and was not altered thereafter, unless it was necessary to do so to bring the readings within these limits. The transmittance was read and recorded. Care being taken not to move the test tube, the glass plate was removed and, without being rinsed or wiped, was placed in the next tube. Transmittance of the tube without glass plate was then read and recorded. Transmittancies were translated to optical density values. Differences between respective densities without the plate and with it are plotted for all test tuhes in Figure 3. For this series 3 = 0.3701 =I=0.00186 (D.F. = 20).
l-2U-l
w m
0-
3
2 0.400
_-L
0.390
0.380
OPTICAL
Khen transmittance with rvater in a tube was adjusted to be the tubes containing permanganate had tranvniittancies in theneighborhood of 5.5% ( D = 1.26).
loo%,
DENSITY
Figure 2. Distribution of Optical Density Differences without and with Displacement Plate Same test tubes as in Figure 1. Beckman DU spectrophotometer. z = 0.3886; v = 0.00135; ( U X 10O)lZ = 0.35%
Table I. Variation in Light Transmittance with i n g l e of Incidence, Due to Reflection, R , and Length of Light Path, L Angle of Incidence, i 00 10
C Each tube in turn was placed in the cell holder, label forward,
20 30
with the glass plate lined up by eye with a spot on the opposite wall of the room, so that its surface appeared to be normal to the light beam. Optical density was read. Then, care being taken not to move the test tube, the plate was removed and placed in the next tube without rinsing or wiping. The optical density was read without the plate.
,A,
U
0 5 -
0..
1 0.350
0:::. 0.03700.
OPTICAL
:
A 0.380
DENSITY
Figure 3. Distribution of Optical Density Differences without and with Displacement Plate Same test tubes as i n Figure 1. Evelyn-type photometer. 2 = 0.3701; u = 0.00186; ( u X loo)/.? = 0.50%
Differences between the respective densities without and n-ith the plate are plotted for all test tubes in Figure 2 . For this series 2 = 0.3886 =k 0.00135 (D.F. = 29). u
x P
100
0.50%
f
- 0.35%
When slit width was set with a tube containing water for D = 0.000, the permanganate tubes had optical densities of the order of 1.3 to 1.4.
40
50 10” 15; 20
R
L
0.00581 0.00581 0.00.581 0.00581 0.0058l 0 . 00581
0.0 0.000086 0.00044: 0.00101 0.00180 0 00281 0.0113
0.00381 0.00583 0.00588
0.02Z6 0.0461
The smooth curves in Figures 1, 2, and 3 are theoretical distribution curves drawn to fit the data in each case. The area is the same under each of the three curves. EFFECT OF ROTkTIOY OF GLkSS PL4TE OV ERROR
Ideally the plate should be in a plane perpendicular, or normal, to the light beam. Rotation of the plate will produce (1) variation in the reflectance a t entrance and exit surfaces, and (2) variation in the length of solution displaced due to the longer slanting path of the light beam through the plate. Factors concerned in these tn-o sources of error are (1) refractive index of the liquid, nil, ( 2 ) refractive index of the displacement plate, n2, and (3) angle of incidence, i. Take n = n2’nl
Also take E = .\/n2 and
R=
- sin?i
reflected light incident light
Frewel’s formula can be modified to give R as a function of n and of the angle of incidence:
E - cos i R = (-i)*((E
E2+ sinti tan%
+ sin i t a n
i)2)
402
ANALYTICAL CHEMISTRY
hssuming for n1 the value 1.33, and for nz the value 1.55, n becomes 1.165, and the above formula gives values for fraction of light reflected a t the entrance surface, a t varying values of i, as shoan in Table I under E. As will be seen, reflection does not change appreciably for small deviations in the position of the glass plate. The situation is different, however, with respect to the change in length of the path of the refracted ray through the glass plate with rotation. Taking D as the optical density of a layer of the displaced liquid equal in thickness to the glass plate set normal or perpendicular to the light ray, and taking Df as the apparent or false optical density of a thicker layer of liquid corresponding to the longer path of the refracted ray through the rotated plate, we can define the resulting positive error, L, as
ordinary good grade of machinist’s micrometer was used, calibrated to 0.001 inch, and readable by interpolation to one tenth of this interval. Thickness was measured a t each end of each strip-twelve measurements in all (Table 11).
Table 11. Variation in Thickness of Consecutive Glass Strips Cut from Single Piece of Plate Glass Plate No.
Top End, Inch
Bottom End, Inch 0.2381 0.2383 0.2410 0.2379 0.2414 0.2409 2 = 0.2396
0.2403 0 2395 0 2398 0 2382 0.2436 0 2418 2 = 0.2405 u = 0.00190
L=- D / - D
c = 0.00166
= 0,79%
D
= 0.69%
It can be s h o m , using the notations given above in the determination of reflected light, that
L = - -Df - D - n
D
-E E
Again assuming n = 1.165, the values for L corresponding to various angles of incidence are given in Table I. It m i l l be noted, if T is taken as the angle of refraction of the light ray corresponding to the angle of incidence, i, that
L =seer - I
Khen the observed optical density values were corrected for variations in thickness of the glass, the standard deviations, expressed in percentage values, ( u X lOO)/a, fell to acceptabls figures. For sets of six observations each, made a t varioue times, values before and after correction for variations in plate thickness are given in Table 111. What can be done in the x-ay of duplication of results vith one tube of solution and one glass plate was s h o m by making 12 successive readings of density differences, the plate being transferred between pairs of readings to another tube of permanRanate. u x 100 For these observations, 3 = 0.3871, and -= 0.19%. ^c
D1 SCUSSION
DEVIATION OF INCIDENT RAY FROM PERPENDICULAR
Displacement photometry offers the opportunity for making accurate relative optical density measurements in unselected test tubes, which can be calibrated for certain volumes, used in the water bath, employed for dilution to volume without transfer, and then placed directly in the photometer. Displacement plates have been employed previously in optical cells to decrease thickness of the liquid layer, for use Tvith liquids of high optical density.
Figure 4. Error Produced by Rotation of Displacement Plate from Perpendicular to Light Beam Table 111. Decrease in Standard Deviations after Correction for Variations in Thickness of Plates
The values of L are plotted as percentage error against i in Figure 4 (curve). Experimentally determined points are also plotted (as circles). Measurements of the angles of deviation from normalcy were not made with precision. Furthermore, instrumental and observational sources of error contributed to the determined errors as the glass plate was rotated. The value of D used was 0.388. Thus an error in the density difference of 0.001, corresponding to one tenth, or a t the most to one fifth, of a scale division, was equivalent to 0.258%. VARIATIONS IN DISPLACEMENT PLATES CUT FROM A SINGLE PIECE OF PLATE GLASS
When attempts were made to secure equal density differences by using a series of glass plates cut in sequence from a single piece of plate glass, they were repeatedly unsuccessful, the standard deviations from the mean being considered too large to be acceptable. Plotting of results gave recurring patterns of errors; so measurements of the thickness of the glass strips were made a t the location through which the light beam passed when the strips were positioned in test tubes in the spectrophotometer. An
Before Correction (T x 100 ,775 2 0 60 0.3878 0.3910 0.58 0.3880 0.88 0,3883 0.87 0,3887 0.66 0.3902 0.96 ~
Biter Correction 0f
0 3885 0 3917 0.3887 0.3890 0.3893 0.3903
,% 0.31 0 05 0.24 0.27 0.33 0.17
Their use in ordinary test tubes as described here may make desirable readjustment of the concentrations of light-absorbing molecules to secure an optimum optical density range. This change can be produced by using a larger aliquot of the sample, by lesser dilution of the final colored solution, or in some cases by using a different n a v e length of light a t which the absorption is greater. The maximum thickness of displacement plate that can profitably be used will be determined by several considerations. The plate should be enough wider than the light beam used that stray light will not escape around the edges. Furthermore, the
V O L U M E 2 5 , NO. 3, M A R C H 1 9 5 3 nidth of the plate should be enough to avoid reflection from the edges. Within the limits of the circular cross section of the tube, the thicker the plate is, the narrower it will have t o be. Because the plate displaces solution, care will have to be taken that liquid is not caused to run over vihen it is inserted. If dilution to a mark is made, part of the contents can be spilled into the sink after mixing, to reach a safe level. Thickness of the plate will determine the amount of displacement, and hence the amount of liquid that can safely be left in the tube. The errors in displacement photometry include in general all those of other types, \Tith one or two exceptions. Thus the eirors in dark current (zero setting) and in initial light intensity setting, as analyzed by Hamilton ( S ) , make their contributions to the scale error function developed by Twyman and Lothian (6). >lost of the instrumental errors mentioned by Caster ( 1 ) remain. An exception is the contribution to error made by surface dirt or scratches on the absorption cell surface. These do not, M ithin limits, affect the accuracy of displacement photometry. Errors peculiar to displacement photometry include dirt, qcratches, or bubbles on the displacement plate. These imperfections may cause a loss of uniformity if several plates are used. Khen a single plate is used surface imperfections that are constant in effect, suc-h as scratches, should produce little error, whereas changing imperfections such as bubbles will cause errors even when only a single plate is used. T a b l e 1 and Figure 4 shom- that it is not necessary to achieve precision in setting the plate at right angles to the light beam, but that good accuracy can be attained by approximating the proper position. Sufficient accuracy can be attained by inspection, without elaborate precautions. Out of twenty consecutive times in which the plate was set in place by inspection and in which the angle of deviation of the plate from normalcy was then measured, in no instance did the deviation exceed 2”. At such values of 2, the error from this source is negligible. A more serious source of error is the presence of inequalities in the displacement plates used. I t was shown above that ordinary polished plate glass can vary more than 2% in thickness over very short distances. Glass plates and even polished narrowband-pass filters may deviate 1% or more in the parallelism of the t n o surfaces (4). Such lack of parallelism will affect the integrated thickness of the plate within the light beam. According to Peckham ( 5 ) , the limits of this variability in commercial plate glass differ with the method of manufacture, and should be least with drawn glass finished n-ith larger polishing blocks, Peckham also called attention to the possibility of error arising from local variations in the index of refraction of glass not prepared for opticd purposes. These errors can be minimized by using a single displacement plate, TT hich will have to be wiped dry between tubes containing solutions with appreciable differences in concentration. Cleaning would be facilitated by treatment of the surface with a silicone, or by making the plate from a transparent nonwetting plastic of sufficient resistance to chemicals. Care xould have to be taken that air bubbles did not cling to such a surface on immewion. I t should not, hoxever, be difficult to prepare plates of glass, plastic, quartz, or other transparent material, of sufficiently uniform thickness and optical characteristics to be interchangeable. Caster ( 1 ) states that under ideal conditions the variation between duplicate readings made with the Beckman DU spectrophotometer is u = 0.1 to 0.2% Tvhen the optical density reading is around 0.4 to 0.5, and that for relative values a variation of the order of u = 0.570 is reasonable. These variations are based on the use of optical absorption cells. The author’s errors, using unselected glass test tubes with displacement plates, are comparable to these. The question whether to adjust the slit width with plate in place to correspond to a density reading of 0.000 or to read the initial density value and subtract it from the value obtained
403 after the plate is removed, is one to be decided by convenience or personal preference. Light absorption of the “unused” or undisplaced solution will act as an additional light filter, and may accentuate stray light effects, especially with wide band width. Thus, marked changes in the geometry of the containing vessel may affect the apparent density values obtained (4). The lesser precision obtained 15 ith displacement plates in the Evelyn-type photometer is associated with a general lesser precision of this type of instrument, as compared with the more expensive one, as well as possibly with a qider light beam. The difference in mean observed density of the displaced layer of solution (0.3701 in the Evelyn-type instrument versus 0.3886 in the Beckman) can be accounted for by the aider band of light wave lengths transmitted by the “light filter,” and consequent stray light effects. Possibly concerned also was the difference in geometry of the light rays, the incident beam in the Evelj-n-type instrument being divergent, nhereas in the Beckman it is collimated, or a t least deviates less from parallel. For use in the Evelyn-type photometer, tubes will have to be selected to the degree that they will fit into the instrument. Use of the plates in an Evelyn-type instrument offers certain advantages in addition to those found with the Beckman. Although one has to line up the plate in a position perpendicular t o the light beam, this is easily done, and no attention need be paid to orientation of the test tube except to see that it is not changed while the plate is being removed. Removal of the plate is easier than is substitution of the tube containing colored sample for the one containing the blank, as is usually done. The setting or reading made with the plate in place corrects for variation? in the photocell output. Density differences obtained as described are not proportional to concentration of solute. Plotting of density difference values against concentration gives a straight line which in general does not go through the origin. The same statement applies to the usual type of photometry, unless a blank is used, because of the constant contribution to each value by the “blank” absorption. It is customary to correct for this blank value, either automatically in the setting of the initial light intensity (setting the instrument for T = 10070 with a blank), or by subtracting the density value of the blank from each of the other values. In displacement photometry this constant value is the resultant of the reagent blank plus the contribution of the plate itself (absorption by the glass, reflectance, and other factors). It can easily be determined by measuring the density difference produced by immersion of the plate in a “reagent blank” such as is usually prepared. Density differences produced by immersion in water of plates used in experiments above, a t 515 mh, were of the order of 0.006. SUMMARY,
Ordinary, unselected test tubes can be used for quantitative absorption photometry by making density readings before and after removal of a flat plate of glass or other transparent material set in the tube to displace a constant depth of solution. Studies of the nature and magnitude of errors that may be encountered in the procedure show that with moderate precautions good accuracy can be anticipated. Advantages are obvious. LITERATURE CITED
(1) Caster, W. O., ANAL. CHEM.,23, 1229-36 (1951). (2) Evelyn, K. d., J . Biol. Chem., 115, 63-75 (1936). (3) Hamilton, R. H., ISD. ENC.CHEM.,ANAL.ED., 16, 123-6 (1944).
(4) Hare, George, personal communication. (5) Peckham, R. H., personal communication. (6) Twyman, F., and Lothian, G. F., Proc. Phys. SOC.(London),45, 643-62 (1933).
( 7 ) Willard, H. H., and Greathouse, L., J . Am. Chem. SOC.,39, 236677 (1917).
RECEIVED for review September 18, 1952. Accepted December 17 1952.
