End Point of Microtitrations with Color Indicators

mission is measured at 6.5 to 6.6. Interference due toother reducingsubstances such as glutathione and cysteine is greatly inhibited by the addition o...
0 downloads 0 Views 657KB Size
828

INDUSTRIAL AND ENGINEERING CHEMISTRY

Summary

Vol. 14, No. 10

Acknowledgment

An improved microphotometric method for ascorbic acid based upon the difference in transmission of buffered 2,6-dichlorophenolindophenol before and after reduction is described. Reduction is carried out a t pH 2.5 to 2.7 and transmission is measured a t 6.5 to 6.6. Interference due to other reducing substances such as glutathione and cysteine is greatly inhibited by the addition of mercuric chloride. Frequent restandardization is unnecessary and changes in the ascorbic acid equivalency can be read directly from the transmittance-concentration curve.

The writer is indebted to V. Suntzeff for the carcinomas and for the samples of isolated epidermis used in this study.

Literature Cited (1) Baumberger, J. P., Suntseff, V., and Cowdry, E. V., J . Natl. Cancer Inst., 2, 143 (1942). (2) Bessey, 0. A,, J . Am. M e d . Assoc., 111, 1290 (1938). (3) Bessey, 0. A., J . Bid. Chem., 126, 771 (1938). (4) Evelyn, K. A., Malloy, H. T., and Rosen, C., Ibid.,126,645 (1938). (5) Kassell, B., and Brand, E., Ibid., 125, 115 (1938). (6) Menaker, M. H., and Guerrant, N. B., IND.ENG.CHEM.,ANAL. ED.,10, 25 (1938). (7) Mindlin, R. L., and Butler, A. M., J . B i d . Chem., 122, 673 (193738).

End Point of Microtitrations with Color Indicators A. A. BENEDETTI-PICHLER

AND

SIDNEY SIGGlA, Queens College, Flushing, N. Y.

Solutions of t h e concentrations customary in macroanalysis are proposed for use in microtitrimetry. A discussion of the limitations resulting from the use of color indicators in microtitrations must consider whether the color change occurs throughout the titrated solution or whether the color effect is localized in a small portion of a mixture. In the first instance i t can be show-n that in microtitrations the light must travel approximately the same distance through the titrated solution as in macroanalysis, if the indicator concentration is identical with that used in the standard procedure. I n spite of the necessity of working with volumes of the order of 0.1 ml., a 4-cm. thickness of layer can be obtained in microprocedures

C

LOSE adherence to the provisions of a well-established

analytical procedure is desirable, for it permits interpretation of the results in the light of the experience gathered in study and application of the method, For the same reason it appears desirable to reproduce the conditions of a standard procedure when it is tried on a different scale. If no heterogeneous equilibria or other surface phenomena are involved and time elements, such as rate of adding reagents or rate of stirring, have little influence, the rest of the more generally considered factors-concentration, temperature, and pressure-are easily reproduced on a smaller scale. Maintenance of the concentrations used in the standard procedure needs continuous attention, because of the universal validity of the law of chemical equilibrium. Since

by observing the end point in a coloriscopic capillary which is part of the titration vessel. This capillary may further be used to reserve some of the titrated solution for the final adjustment of the end point. Observation of color changes taking place in a portion of the titrated system is discussed with reference to the use of organic solvents for the indication of the end point in iodometric titrations. A mathematical investigation shows that this particular principle will fail when applied on a very small scale. Iodometric titrations employing a droplet of chloroform as indicator, as well as argentometric titrations with adsorption indicators, may be successfully performed on a milligram scale.

procedure. This detail of the directions of a micromethod is simply derived from the standard procedure by multiplying all specified masses and volumes by the reduction factor f

=

size of microsample size of macrosample

Concentrations and purity specifications of all reagents, standard solutions, diluents, and wash liquids must be left unchanged. Application of the outlined principle to the transposition of a standard procedure from the gram to the milligram scale is shown in Table I. As in the macroprocedure, the carbon dioxide will be eliminated by boiling after addition of a small excess of standard acid. T h e n the titration is mass concentration = - it is sufficient to change in the same finished with 0.5 N sodium hydroxide, there is no reason why volume' the color change a t the end point should not occur with a proportion all masses and volumes specified in the standard fraction of the volume of standard solution . . - listed as drop error in 'I'able 1. TABLEI. TITRATIOS OF SODIUM CARBONATE WITH 0.5 N HYDROCHLORIC ACID For the titration of milligram Methyl Red samples with standard solutions of Volume of TiVolume the customary c o n c e n t r a t i o n s , a trated Solution of 0.3 Capacity per cent Mass Reduc.It number of satisfactory burets of Mass Drop of solution At end of tion Sample Factor start point Buret taken taken Error approximately 50-cu. mm. total Macro 1 gram 25 ml. 60 ml. 5 0 ml. 0.03 ml. 90 pg. 0.03 ml. capacity are available (2, 4-7, 11, Micro 1 mg. 0,'dOl 2 5 cu. 60 cu. 50 cu. 0 . 0 3 cu. 0.09 pg. 0.03 cu. 12). The same cannot be said of titramm. mm. mm. mm. mm. tion vessels, and for this reason the

special requirements are outlined below, and ways of filling them are indicated.

Color Change throughout Titrated Solution It is obvious that the color change will take place throughout the entire solution, if the coloring matter is dissolved in the titrated solution. The design of a vessel for the titration of small samples must then satisfy the following requirements : I t s capacity must be in accord x i t h the volume of the titrated solution, and its shape must permit recognition of the color change and efficient execution of the necessary operations. These requirements are c o n t r a d i c t o r y in part. Mixing, boiling, and adding reagents call for a c o n t a i n e r of c o m p a c t shape. The perception of color in spite of the lorn c o n c e n t r a t i o n of FIGURE 1. LENGTHENIKG OF P.4TH OF LIGHTB Y 1ZEPEATED “loring matter for a depth of liquid which, REFLECTIOS b e c a u s e of t h e s m a l l volume, can be attained only by spreading the solution in a thin film or by collecting it into a long, fine thread. Neither proposal appears satisfactory from a mechanical point of view, The requirements for the perception of the color change at the end point may be stated more definitely. Beer’s law implies that the intensity of a coloration is determined b y the areal concentration of the matter causing selective absorption in the visible range of the spectrum. I n other words, the intensity of color depends upon the mass of coloring matter acting on the light per unit of the area exhibiting the coloration (1). Mika (9, IO) has determined the following limits of areal concentration t h a t assure satisfactory observation of color (micrograms per sq. cm.): methyl red, 6.5 ; bromothymol blue, 80; phenol red or phenolphthalein, 26; thymolphthalein, 5 ; alizarin yellow R, 13; and potassium permanganate, 2 (or 0.012 microgram-mole). One may assume that the titrated solution containing m units of mass of the coloring matter has a volume v a t the end point and fills length 1 of a cylindrical tube with inner diameter d. If the light used for the observation of color travels parallel to the axis of the tube, the areal concentration of the coloring matter IS given by

C, =

4m xd

=

4m - units of mass per units of volume *d21

and division of the first equation by the second gives 1

a constant of the indicator, and, consequently, the minimum length of the column of titrated solution becomes a constant of the standard procedure and must be retained, no matter how small the scale on which the titration is performed. If the standard procedure does not employ an unnecessary excess of coloring matter for the indication of the end point, this minimum length will be approximately the same as the thickness of layer of the titrated solution of the macroprocedure. The volumes of the solutions of microprocedures, however, are reduced in proportion to the decrease of the size of sample. Application to the example illustrated by Table I gives 6.5 pg./sq. cm. I = 0.09 pg./0.06 cu. em.

=

4.3 em.

I n order to see the color change, the titrated solution must be given a depth of at least 43 mm. Table I also shows that even samples as large as 1 mg. require reduction of t h e volume of the titrated solution to 60 cu. mm. The solution will fill a sphere of 5-mm. diameter, but the light must travel through i t for a distance of not less than 43 mm.

P

light

FIGURE 2. CIRCULATION O F TITRATED SOLUTION THROUGH COLORISCOPIC CAPILLARY

The standard procedure should be transposed without change of the essential conditions. Obviously, the indicated difficulties disappear at once when this premise is waived. It is not necessary, however, to accept the disadvantages of such a short cut (flattening of titration curve, poorly defined end points, necessity of using color standards, and corrections for the blank), for it is possible to fill the contradictory requirements for the transposing of standard procedures in at least two mays: A titration vessel of simple compact shape may be used, and the path of light through the solution may be given the required length b y passing the light back and forth through the solution b y repeated reflection (Figure 1). T h e titration vessel may be designed to provide two compartments, one for the observation of color and the other for the performance of such mechanical operations as stirring, adding of reagent, etc. The titrated solution may be made to circulate (Figure 2 ) or to oscillate from one compartment to the other (Figure 3).

units of mass per units of area

The concentration of coloring matter is obviously equal t o c

829

ANALYTICAL EDITION

October 15, 1942

=

C, -units

of length

The length of the column of titrated solution is determined by the ratio of the areal concentration to the concentration by volume of coloring matter. Recalling that the concentration, c, is fixed by the choice of the standard procedure, it follows that the minimum depth of the column of titrated solution is determined by the smallest areal concentration assuring perception of the color change. The limiting areal concentration is essentially

Color Change Localized in Part of Titrated System The use of organic solvents for the detection of a small excess of iodine at the end point of iodometric titrations ( 3 ) provides a n interesting example. Assuming that the organic solvent is able to extract practically all of the iodine added in excess, one is inclined to think that the sensitivity of the end-point indication would be equal t o that of the macroprocedure, if the cross-section area of the drop of organic solvent \yere reduced in the same proportion as the size of the sample. Then the amount of iodine corresponding t o the drop error xould be divided by the same number as the area showing the coloration, and the areal concentration of iodine in the organic solvent would remain unchanged. The assumption of a practically complete extraction of the iodine is, however, not permissible. Even for the distribution of the iodine between water and chloroform the partition ratio

INDUSTRIAL AND ENGINEERING CHEMISTRY

830

has too low a value to support the premise. Furthermore, the titrated solution contains an appreciable amount of iodide ion at the end point of an iodometric determination, and the equilibrium I2 I- s I :must be taken into consideration

+

tion, which is assumed to be 0.12 molar with respect to potassium iodide. At the end point the titrated solution has a volume of 60 cu. mm., and its iodide concentration is 0.11 mole per liter because of the reaction

+

11 2

&os--

m

= vo

+ vOII~lsqu. + oli

[Izlore.

[I;

I

Using the equations for the partition ratio and for the instability constant of the triiodide ion, one calculates and [I;] as functions of [ I z ] ~and ~ , . substitutes in the above equation, which is then solved for [1210rp. Multiplication of the concentration of iodine in the organic solvent by the volume, VO, of the organic solvent gives the mass, mol of iodine dissolved in the organic solvent, and division by the area, A , of the organic solvent furnishes the areal concentration of iodine in the organic liquid:

The form of the equation indicates that the areal concentration of the iodine in the organic solvent, mo/A,must arrive at a maximum value for a certain volume, V U , of the organic solvent, provided that the other conditions are kept constant. Differentiation is made simple by setting uo equal to 53. The areal concentration of the iodine in the organic solvent becomes a maximum for

2 1-

+ &OB--

The areal concentration of iodine has been calculated for various volumes, UO, of the droplet of chloroform from the equation

mu = It becomes obvious that only a few per cent of the available iodine can be taken up by the organic solvent. The areal concentration of iodine in the drop of organic solvent is a somewhat involved function of the total amount of iodine available, the volumes of aqueous and organic phases, the partition ratio, and the iodide concentration. The organic solvent is assumed to take the shape of a sphere of volume UO, cross-section area A , and diameter d; u, represents the volume of the aqueous solution at the end point. The total mass, m, of iodine added in excess corresponds to the “drop error” or more correctly to the amount of iodine consumed in a blank titration. The calculation begins with the statement that the excess of iodine added a t the end point is partly dissolved in the organic solvent, partly dissolved in the aqueous solution, and partly converted to triiodide ion

Vol. 14, No. 10

A

0.108 130 vo

+ 4.88

The areal concentration of iodine reaches a maximum when the spherical drop of chloroform is given a volume of 18.7 cu. mm., corresponding to a diameter of 3.3 mm. Table I1 and Figure 4 show that the diameter of the drop of chloroform may vary approximately from 1 to 8 mm. without reducing the areal concentration of iodine and, consequently, the intensity of the coloration to much more than one half of the obtainable maximum. Table and graph show that the areal concentration of iodine decreases directly with the diameter of the chloroform dro when its volume becomes very small (diameter less than 2 mm.7. The areal concentration increases in inverse proportion to the crosssection area of the drop when the volume of the chloroform becomes more than ten times larger than the volume of the titrated solution ( A = 1 sq. cm. and more, or d = 11 mm. or more). Either the former or the latter will occur, depending upon which of the two items 130 00 or 4.88 can be neglected in the denominator of the equation. The corresponding macroprocedure would give a volume

of 60 ml. at the end point of the titration, and 5 ml. of carbon tetrachloride are customarily used for the extraction of the iodine ( k , = 85). Assuming approximately spherical shape for the carbon tetrachloride layer, the areal concentration of iodine at the end point will be 0.0227 m-i. e., nearly six times higher than in the microprocedure which employs the most favorable volume of a solvent giving a higher partition ratio. The appearance of the color of iodine in the organic solvent can be used without difficulty for the detection of the end

When a standard rocedure has been selected, the quantities m, k p , v., vo, and [I- fhave been assigned deiinite values, and the equation for the areal concentration may be simplified to

I n transposing to the microscale, the quantities m, VO, and va are decreased in the same proportion by multiplication by the reduction factor, f, and the areal concentration of the iodine in the organic solvent becomes

(.y)

micro

=

fl/3

W

(y)

macro

This means that the areal concentration of the iodine decreases in proportion to the third root of the reduction factor. Provided that the standard procedure employs the most favorable volume of organic solvent and does not use an unnecessary excess, m, of iodine a t the end point, it follows that transposition to a smaller scale lessens the sensitivity of the end-point indication. When proceeding from the gram scale to the milligram scale, the areal concentration of the iodine in the organic solvent is reduced ten times.

I

I

Table I1 and Figure 4 apply to a microprocedure in which chloroform is taken for the organic solvent. Thirty cubic millimeters of 0.5 N thiosulfate are titrated with approximately the same volume of 0.5 N iodine standard solu-

FIGURE 3. TITRATION BULBWITH ATTACHED COLORISCOPIC CAPILLARY

831

ANALYTICAL EDITION

October 15, 1942

point in the titration of milligram samples. However, there are serious limitations t o the application of this principle on a much smaller scale. The difficulties may be lessened b y reducing the iodide-ion concentration of the titrated solution, selecting a solvent giving a better partition ratio, and increasing the distance which t h e light travels through the organic solvent.

Experimental U-shaped burets operated by remote control (8) were employed. Millimeter scales of 30-cm. length mere placed behind the vertical capillaries of approximately 0.8-mm. bore, so as to make the total capacities of the calibrated portions of the burets equal to approximately 0.15 ml. The control of the outflow of standard solutions was sufficiently fine to warrant estimation of tenths of the millimeter scale, and 1 mm. of the scale corresponded to a volume of approximately 0.5 cu. mm. In the tables the amounts of standard solutions are given in centimeters. Conversion to units of volume would not be helpful, since only ratios were determined. A significant variation in the bore of the authors' capillary tubes has not been discovered so far.

TABLE 11. AREALCONCENTRATION O F

IODINE

As a function of volume, cross-section area, and diameter of t h e drop of chloroform. v a = 6 0 c u . mm.; kp-130; [1-]=0.11m o l e p e r l i t e r ; m(mloro)= 0.001 m UO

Cu. mm. 2120 750 266 94 33.5 27.0 18.7 10.9 5.0

4.2

0.52 0.17 0.066 0.0082 Zero

A

sq.

cm.

2.0 1.0 0.50 0.25

..

0 : 685

0.037

.. 0.00365 ... . Zero

d

Mm. 16.0 11.3 8.0 5.6 4.0

3.7 3.30 2.75 2.16 2.00 1 .oo 0.70 0.50 0.25 Zero

mo/A Mass/sq. cm. 0.00049 m 0.00096 m 0.00176 m 0,00288m 0.00378 m 0.00385 m 0.00391 m 0.00380 m 0.00333 m 0.00321 m 0.00175 m 0.00123 m 0.00089 m 0.00045 m Zero

For stirring, a stream of small gas bubbles was passed through the titrated solutions by means of a capillary of 0.1- to 0.2-mm. bore. A satisfactory supply of gas may be obtained by com ressing air in a 2.5-liter bottle by blowing with the mouth. &nce 16,000 bubbles of 0.5-mm. diameter have a total volume of only 1 ml., one filling of the bottle should suffice for a number of titrations. The outlet of the bottle is connected through a U-tube with Ascarite to the fine capillary, and a screw clamp on the rubber tubing permits regulation of the rate of the gas stream. Most of the titrations were performed in centrifuge cones of 2-ml. capacity and of the shape commonly used in qualitative work. The bulb with attached colorisco ic capillary (Figure 3) facilitates recognition of the end point. $he capillary is given a length of 4 to 6 cm. Its capacity should not exceed 0.25 c, if c is the total capacity of the buret used. This then always leaves a sufficient volume of liquid in the bulb, which is given a capacity of 2 c or more. A T-tube with a window of thin glass at W is placed over the coloriscopic capillary. Its open arm, T , is connected by means of flexible rubber tubing to an Ascarite tube provided with a mouthpiece. Light is sent in the direction of the arrow, L, through the bulb into the coloriscopic capillary. A shield, S, made of black cardboard, may be placed around the T-tube. For use with burets operated by remote control, titration vessel and fine capillary, M , are mounted on an elevating stand, so that the tip of the buret can be touched with the meniscus of the titrated solution by raising the titration vessel by means of the rack-and-pinion motion of the elevating stand. A slow stream of gas through ill is first started, and then the solution to be titrated is introduced into the bulb. Part of the solution enters the coloriscopic capillary, and, if necessary, slight suction with the mouth may be applied through T to draw the meniscus u into the opening which faces W. This portion of the solution is fept in the capillary for use in the final adjustment of the end point. The flow of gas is adjusted to give a steady stream of bubbles from the opening of M , and then standard solution is added from the buret. By observation of the color of the liquid

e

4

6

8

la

16 m.

FIGURE 4. AREALCONCENTRATION OF IODINE AS A FUNCTION OF THE DIAMETER OF THE DROPOF CHLOROFORM in the bulb it is possible to shut off the flow of standard solution soon after the end point has been overstepped. The solution in the coloriscopic capillary is now transferred to the bulb by cautiously blowing with the mouth through 2'.' The tip of the coloriscopic capillary cannot be drawn out very fine, since this would delay the back-and-forth transfer of the titrated solution. Thus, care must be exercised, since a violent stream of lar e air bubbles may throw some of the solution out of the bulb. d s soon as the contents of the bulb have become homogeneous, the original color of the titrated solution is restored, and then the coloriscopic capillary is made to fill up. Standard solution is now added more slowly until the end point is again overstepped in the bulb. The contents of the bulb are once more back-titrated as described, and the whole procedure may be repeated, if the ratio of the capacity of the coloriscopic capillary to the total volume of titrated solution permits. The final adjustment of the end point requires observation of the color change by means of the coloriscopic capillary. The buret is adjusted to give a low rate of outflow. The standard solution is added in small portions, and after each addition the contents of coloriscopic capillary and bulb are mixed. The color of the resulting solution is observed after it has again filled the whole length of the capillary. The openin of the coloriscopic capillary is viewed through W with the aid of a magnifying glass. Thus, the end point is recognized before a change of color can be seen in the bulb of the titration vessel. After some practice has been acquired, a titration can be performed within 10 minutes. This time will even allow the elimination of carbon dioxide b y blowing steam on the outside of the titration vessel while a stream of gas is bubbled through the titrated solution. NEUTRALIMETRIC TITRATION WITH METHYL RED. Half normal sodium hydroxide was titrated with hydrochloric acid of approximately the same normality. The customary concentration of indicator was obtained by adding 1 drop of methyl red solution to 200 ml. of each standard solution. The base was run from a buret into the titration vessel, and acid was added from another buret until the color changed to pink. Carbon dioxide was expelled by heating before the final adjustment of the end point. Table I11 indicates the superiority of the titrating bulb with attached coloriscopic capillary. Use of centrifuge

832

INDUSTRIAL AND ENGINEERING CHEMISTRY

Vol. 14, No, 10

and then increasing the rate of the gas stream so that the chloroform droplet is completely evaporated. This does not require much time, whereas the extraction of iodine from In scopic with Attached Capillary Coloria chloroform drop of more than 2-mm. diameter is extremely In Alicrocone Cm. NaOH c m . ~ & oslow. ~ The excess of thiosulfate takes care of the iodine NaOH HC1 Cm.HC1 XaOH HCI Cm. HCI liberated during the evaporation of the chloroform. FiCm. Cm. Cm. Cm. nally, a fresh droplet of chloroform is added, and the end 11.99 12.00 0.9993 1 10.15 point is reached by adding iodine standard solution. 10.94 11.10 0,9858 10 2 .. 12 23 12.26 10.02 10.15 0.9872 18.21 18.21 l . 0000 The ratio of centimeter of iodine to centimeter of thio21.35 21.43 0.9962 25.99 26.08 0.9965 sulfate calculated as the mean of six determinations was Mean 0,991 0.9974 1.119 * 0.002. The average devia10,004 ~0.0005 tion of the single determinations was equal to +0.005 or i.5 parts per TABLE Ip. PERMANGANATE TITRATION WITH 0.1 N SOLUTIONs thousand. K h e n the titration was carried out in the same cones with I n Microcone I n Bulb with Attached Coloriscopic Capillary Cm. KIInOl Cm. KIlnOd the customary concentration of starch C m . FeSOA KlInOa (NHd)nFe(SOa)n Khfn0.r Cm. FeSOa for the indication of the end point, Cm. Cm. Cm . the ratio 1.130 * 0.003 was obtained 10.41 12.12 1.166 10.57 12.14 1.147 as the mean of three determinations. 12,26 15.59 1.166 13.62 14.30 1.145 1 . I62 15.96 19.15 16.77 1.142 18.55 ARGEXTOMETRIC TITRATION L‘SIXG 22,27 1,163 19.30 19.76 1,154 22.98 22.08 24 22 1.166 20.97 25.75 1. I55 ADSORPTIOKINDICATOR. The color .\lean 1 165 1.149 change takes place a t the surface of the =tO.OOl -0 002 precipitate formed during the titration, If it sufficed to watch the color shom-n by one particle of the precipitate, no matter how small, this method of end-point indication should cones cannot be considered by any experimenter who is be suitable for titrations on an extremely small scale. There is, unable t o recognize slight differences of hue in faint colorahovever, reason to assume that a satisfactory color effect is tions. obtained b y a repeated reflection of the light from the surTITRATION WITH PERMANGAXBTE. Tenth normal acid ferrous faces of a large number of particles. This view is supported ammonium sulfate was measured from a buret into the titration by the experience in microtitrations, which showed that it vessel and permanganate solution of approximately the same normality was added from another buret until a pink coloration is preferable to prevent flocculation of the silver chloride b y of the titrated solution could be perceived. adding the standard solution rather fast while agitating by a moderate stream of air bubbles. The results are compiled in Table IV. Again it is obvious

TABLE111. DETERMINATION O F BASE-ACID R.4TIO SOLUTIOKS

WITH

0.5 N

::::;:

that the end points were not recognized in proper time when the titrations were performed in centrifuge cones.

Tenth normal sodium chloride was run from the microburet into a centrifuge cone of 2-ml. capacity. There the standard solution was treated with one tenth of its volume of 1 M acetic acid, IODOMETRIC TITRATION USISG CHLOROFORM -4s INDICATOR. two tenths of its volume of 1 M ammonium acetate, and a small Tenth normal thiosulfate solution was titrated with iodine soludrop of dichlorofluorescein indicator solution. Then 0.1 N silver tion of approximately the same strength in a centrifuge cone of nitrate was added in a rather rapid stream until the end point was 2-ml. capacity. The approximate volumes of standard solutions near. Most of the precipitate remained suspended in the solurequired were found by a preliminary titration, and then the exact tion. The titration was finished by adding small portions of silver ratios were determined as follows: The thiosulfate solution in nitrate until the color change took place throughout the mixture. the centrifuge cone was continuously stirred by a stream of small Back-titration was not feasible, since the pink coloration of the air bubbles and treated with such a volume of iodine solution precipitate was not completely discharged on adding sodium as to approach the end point closely. A small drop of propyl chloride. Thus, the approximate location of the end point had to alcohol was then added t o the titrated solution, and a small drop be established by a preliminary titration. of chloroform was allowed to fall into the titration vessel from a height of about 5 cm. above the surface of the titrated solution. The ratio centimeter of silver nitrate to centimeter of This procedure and the lowering of the surface tension by the sodium chloride was found as the mean of 4 titrations t o addition of propyl alcohol enabled the drop of chloroform t o penetrate the surface of the aqueous solution and to sink to the bottom be 1.051 * 0.003. The average deviation of the single of the container. In spite of its greater density, the chloroform determination was equal to * 0.005 or * 5 parts per thousand. would otherwise have floated on the surface. The titration was finished by adding small portions of iodine solution until t’he end Literature Cited point was indicated by the appearance of a yellow coloration in the drop of chloroform. Since a stream of gas bubbles serves for agitation, the chloroform evaporat’es rapidly. It is not practical, therefore, to use a very small drop of chloroform or to add the chloroform at a n early stage of the titration. The rate of evaporation may be considerably decreased, however, b y passing the gas used for agitation through a mashing bottle with chloroform before it is led t’o the fine capillary. Drops of chloroform of 1- to 2-mm. diameter have been found most satisfactory for the indication of the end point. If the droplet never gets close t o the tip of the buret, iodine will not be absorbed as long as thiosulfate is present in the solution. Even drops of 3- to 4-mm. diameter may be used, if t h e opening of the buret is kept far away from the chloroform. mhenever an excess of iodine solution is added, backtitration is best performed by adding a n excess of thiosulfate

Benedetti-Pichler, A., 2. anal. Chem., 73, 205 (1928). Brandt-Rehberg, P., Biochem. J., 19,270 (1925). Hall, W. T . , “Textbook of Quantitative Analysis”, 3rd ed., p. 128, New York, John Wiley & Sons, 1941; and other textbooks. Heatley, N. G., Biochem. J., 29,626 (1935). Heatley, N. G., Mikrochemie, 26, 147 (1939). Kirk, P. L., Ibid., 14, 1 (1933). Linderstr#m-Lang, K., and Holter, H., Compt. rend. trav. lab. Carlsberg, 19, No. 14 (1933). Llacer, A. J., and Sossi, J. A, see -4.4 . Benedetti-Pichler, “Introduction to the Microtechnique of Inorganic Analysis”, New York, John Wiley & Sons, 1942. Mika, J., Mikrochemie, 9, 143 (1931). Mika, J., 2. and.Chem., 78, 268 (1929). Schwars, K., -2fikrochemie, 13, 1 (1933); 18, 309 (1935). Wermuth, S., see F. Emich, “Urnsetsungen sehr kleiner Stoffmengen”, in A. Stahler, E. Tiede, and F. Richter, “Handbuch der Arbeitsmethoden in der anorganischen Chemie”, Vol. 2, p. 655, Berlin, W. de Gruyter, 1925.