INDUSTRIAL AND ENGINEERING CHEMISTRY
52
bustion was incomplete and the value obtained for carbon was low. There had been reported in the literature ( 4 ) just such difficulties with this type of compound. The following procedure has been utilized to solve them : The substance is weighed in an ordinary platinum boat and sprinkled with fine, previously heated copper oxide. Another, somewhat smaller, platinum boat is inserted into the former, so that the handles are at opposite poles. By conducting the combustion slowly, accurate results have been obtained with such compounds. This arrangement has been termed the “technic of the two boats.”
TECHNIC OF THE Two CAPILLARIES. During the last few years several thousand samples of organic liquids, having a large variety of structure, have been analyzed. Some could be burned in the usual way, while others were very volatile. The following procedure has proved useful: The sample is weighed in a regular capillary. The tip is broken off and the capillary is placed in another somewhat larger capillary about 1 cm. long which is closed at one end. The larger capillary is filled with copper oxide or, if the substance is a halide, with precipitated and previously dried silver. These two capillaries are placed in a suitable long platinum vessel. This arrangement has been termed the “technic of the two capillaries.”
VOL. 10, NO. 1
Certain organic halides (especially bromides and iodides) containing a high percentage of halogen yielded rather high values for carbon despite the rolls of silver gauze in the combustion tube. The additional silver placed in the larger of two capillaries by means of the technic described above absorbs most of the halogen and lengthens the life of the silver gauze. The two capillary arrangements also diminish the speed of vaporization of the volatile liquid sample; hence a more complete combustion can be expected.
Literature Cited Hayman, D. F., IND. ENG.CHEM., Anal. Ed., 8, 342 (1936). Hernler, F., Mikrochemie, Pregl Festschrifi (1929). Kirner, W. R . , IND.ENQ.CHEM.,Anal. Ed., 5, 363 (1933). Kolb. A.. Ann. Chem.. 291. 276 (1896). ~, Niederl, J. B., IND.ENG.CHEY.,Anal. Ed., 7, 214 (1935). (6) Niederl, J. B., and Roth, R. J., Ibid., 6, 272 (1934). (7) Pregl, F., and Roth. H., “Quantitative Organic Microanalysis,” 3rd English ed., Philadelphia, P. Blakiston’s Son & Co.. 1936. (8) Tiedcke, Carl, iWikrochemie, 16, 171 (1934). (1) (2) (3) (4) (5)
RECEIVED October 15, 1937. Presented in part before the New York-New Jersey Section of the Microohemical Society, April 11, 1937.
Turbidimetric Titration of Small Amounts of Nicotine By the Use of a Photoelectric Cell LYLE D. GOODHUE Bureau of Entomology a n d P l a n t Quarantine, U. S. Department of Agriculture, Washington, D. C.
A n inexpensive photoelectric apparatus, including a special titration cell, is used for the turbidimetric titration of small amounts of nicotine. The unknown nicotine sample is added to an excess of silicotungstic acid and the excess of the latter is titrated with standard nicotine formate. Results that check to about 5 micrograms can be obtained. Flocculation of the precipitate is prevented by the addition of Irish moss extract, and the tendency to crystallize is retarded by using formic acid instead of hydrochloric acid as in the analysis by the gravimetric method.
C
ERTAIX fumigation and dusting experiments in progress a t the laboratory of the Division of Insecticide Investigations required a rapid method for the determination of small quantities of nicotine. A survey of the numerous available methods, which include gravimetric, nephelometric, and colorimetric procedures, did not reveal a method of analysis without objection in one respect or another. Although not previously reported for nicotine, titration to maximum turbidity appeared to be a rapid and accurate method, if certain conditions, such as very low solubility and stabilization of the precipitate, could be realized. The use of the photoelectric cell to indicate the point of
maximum turbidity during the titration of SO4-- with B a + + was recently suggested by del Campo, Burriel, and Escolar (1). Large volumes (200 cc.) were used, however, and, since stabilization of the precipitate was not complete, a continuous stirring device was necessary. N o details regarding the apparatus or the type of photoelectric cell were given.
Experimental I n the turbidimetric titration of nicotine with silicotungstic acid, i t was found that more accurate results were obtained when the nicotine solution was added to the silicotungstic acid than when the reverse procedure was used. A deiinite quantity of silicotungstic acid in excess of the unknown nicotine sample was placed in the titration cell and the excess titrated with standard nicotine solution. The quantity of nicotine solution equivalent to the silicotungstic acid was determined by a blank titration. The end point of the titration was obtained by plotting the scale readings of the photoelectric apparatus against cubic centimeters of standard nicotine. DESCRIPTION OF APPARATGS. The circuit shown in Figure 1 employs a highly sensitive gas-filled photoelectric cell, 918, with one power amplifier, 1F4. Because of its instability, a gas-filled cell is not usually recommended for an instrument of this kind, but if a separate battery is used to supply the potential across the photoelectric cell, and the grid of the amplifier is connected with the anode, a nearly constant plate current is obt’ained. In this way the effects caused by the small changes in resistance of the vacuum tube and the photoelectric cell pract’icallyneutralize each other. The over-all sensitivity of the set can be varied within wide limits by varying the load resistance, RI. About 40 megohms have been found t o give good sensitivity without sacrificing stability. The sensitivity can be increased by using an indicator,
JANUARY 15, 1938
ANALYTICAL EDITION
M A , with a range of 1 milliampere in connection with a compensating current so that only the change in the plate current is measured. Variable resistors, RB and Rs, are used t o adjust the indicator on the scale and to return it to zero after t'he scale has been traversed once. A 10-to-1 shunt, R1, can be used to protect the milliammeter when necessary. Also, in this circuit the greatest sensitivity is obtained when the screen grid and the plate of the vacuum tube are connected and operated at 90 volts. A constant source of light is obtained from a 60-milliampere radio pilot light, L, operated by a 2-volt storage cell. About 1 per cent of the light from the filament is utilized. This small amount will change the plate current 10 milliamperes. The design of the apparatus may be adapted to t.he needs of the individual. The author has found it convenient to use a steel cabinet 20 X 30 X 17.5 cm. (8 by 12 by 7 inches). All instruments are mounted on a Bakelite panel, which is dropped 2.5 cm. (1 inch) below the top of the cabinet to allow the lid to be closed when the set is not in use. The titration cell shown in Figure 1 is made by welding a portion of a small (12-mm.) inverted test tube in the bottom of a larger (25-mm.) test tube. This indentation accommodates the small pilot light. As little as 5 cc. or as much as 40 cc. can be titrated. The heating effect of the light is so slight that no perceptible rise in temperature is noted for an hour or more. A holder for the titration cell consists of a braw tube mounted below the panel 6.5 cm. (2.5 inches) from the photoelectric cell. The light is held a t t'he loxer end while the upper end opens above the panel. A 4-mm. hole in the side of the brass tube opposite the filament allows more than enough light to fall on the photoelectric cell. A black paper on the photoelectric cell in which a small opening is cut serves to reduce the light further and to exclude radiation and reflection from the vacuum tube. A set of this type should be protect,ed from the capacity effects of the operator by a shield under the panel over the vacuum tube and photoelectric cell. The grid lead should be very short. Some kind of spring socket' should be used for the vacuum tube, since a sharp jar may change the position of the elements slightly and ruin a determination. A 5-cc. microburet graduated to 0.02 cc. is used in the titrations, and is held by a removable stand built into the panel of the apparatus. Only pipets having a delivery that varies less than 0.005 cc. from the indicated volume are used. A stirrer for use with the titration cell is made by bending the end of a glass rod into a nearly complete circle in a plane perpendicular to the rod. The opening in the circle is held in such a position that the stirring motion does not interrupt the beam of light falling on the photoelectric cell. REAGENTS.A silicotungstic acid solution made by dissolving 5 grams of silicotungstic acid (4HzO~Si0~~12W0~~22HzO) in 1 liter of water. (Three samples of silicotungstic acid were tried, but no significant difference in the results was noted.) The turbidity is removed by settling and decanting. A standard nicotine formate solution made by adding enough formic acid to neutralize both basic groups in 0.5 gram of nicotine and diluting to 1 liter. The exact amount of nicotine was determined by the official -4. 0. A. C. gravimetric method. This solution is stable for at least 2 months. This concentration ap-
FIGCRE2. TITRATION CURVES,SHOWINGMETHODOF LoC.4TING ENDP O I N T Two or three points a t maximum turbidity are always noted. Some points are t o o far t o t h e right t o be included on this graph.
pears to be the optimum for the most precise results with this apparatus. A 1.5 A4 formic acid solution. An extract of Irish moss made by soaking 2 grams in 200 cc. of water overnight, decanting, washing, and boiling 30 minutes in 100 cc. of water. The cellular material was strained out with cheesecloth but not pressed, because this increases the turbidity of the extract. A crystal of thymol must be added to prevent the growth of mold. This solution is stable for a month or more.
General Procedure
:1
TITRATION CELL
T
12v 1
FIGURE 1. WIRING DIAGRAM
AND TITR.4TION
CELL
R I , 40-megohm load resistance R2. 2000-ohm radio potentiometer Rr, ZOO-ohm radio potentiometer H I , one-tenth the resistance of milliammeter M A , milliammeter with 0 t o 1 range L , 60-milliampere, 2-volt light 018, sensitive gas-filled photoelectric cell 1F4,power amplifier with high amplification factor
53
For samples of from 0.05 to 0.75 mg. of nicotine, 2 cc. of the silicotungstic acid solution are pipetted into the titration cell, and 4 drops of the formic acid solution, 4 drops of Irish moss extract, and 10 cc. of water are added. The cell is placed in the apparatus, the indicator adjusted to zero, and titration made with the standard nicotine solution. After about half the titration has been done, readings should be taken about every 0.1 cc. The contents of the cell must be well stirred after each addition and about 30 seconds alloived for complete precipitation. The titration is continued until the maximum reading remains constant for 0.2 to 0.3 cc., after which a small dilution effect becomes apparent. To determine the end point the scale reading in milliamperes is plotted against cubic centimeters of standard nicotine (Figure 2). A straight line is obtained over the last, one-third to one-half of the titration up to within less than 0.1 cc. of the end point. The end point is taken as the intersection of this projected line with a horizontal line passing through the point of maximum turbidity. For the analysis of a n unknown sample of nicotine, which should be contained in dilute formic acid solution, the same procedure should be carried out, except for the substitution of the unknown nicotine solution for a portion of the water.
54
INDUSTRIAL AND ENGINEERING CHEMISTRY TABLEI. Sicotine Solution Taken
TITRATIONOF NICOTINEBY
Silicotungstic Acid Taken
Titrations
cc.
Nicotine Taken Gram
cc.
cc.
0.0
0.0
1
... ... . . ...
... ... ... ...
...
... ... ...
0 0
0.0
...
1.14 1.143
Ar.
1 14 1 155 1.16 1,l5 1.13 1 14 1,145
CC.
Nicotine Taken Gram 0.066
Silicotungstic Acid Taken
cc. 1
...
1.0
0.132
1
2 0
0 263
1
... . .
... ...
,j
...
AV.
TURBIDIMETRIC METHOD Titrations
cc. 1.01 1.02 1.015 0.535 0,520 0.535 0.915 Av. 0.526 0.68
Xcotine Found Gram 0.075
...
Av.
, . .
0
... ... . . ... ...
0.655
0.675
Results To test the accuracy of the method a known solution of nicotine was used. By gravimetric analysis it was found to contain 0.1318 mg. per cc. The gravimetric analysis of the standard nicotine solution on two different days showed 0.5802 and 0.5795 mg. per cc. For the smaller amounts of nicotine only 1 cc. of silicotungstic acid was used in half the 12-cc. volume indicated above. A summary of the titrations and a comparison of the results with those obtained by gral-imetric analysis are given in Table I.
Discussion of Method and Results The precision of the method depends primarily on how accurately the end point can be determined. About 0.01 cc., which is equivalent to 0.0058 mg. of nicotine, seems to be the lower limit. The average of several volumetric determinations appears to check the gravimetric results with about the same precision.
Av.
0.683
Av.
1.14 1.14 1.140
2
0.659
...
2 ,277
The same 7-olume in the titration cell should be reached a t the end point of each titration. The difference in the two titrations multiplied by the concentration (milligrams per cubic centimeter) of nicotine in the standard solution gives the nicotine in the unknown sample. Duplicate determinntions should be run on both the blank and the unknown. The blank titration remains constant from day t o day and serves as a check on the apparatus. The titration cell must be washed with dilute sodium hydroxide a t least once a day to remove any seed crystals that might form on the malls of the container.
0.127
... ... ...
0.268
... ...
0.660
...
The final volume affectsthe end point of the titration. This effect has been determined by titrating 2 cc. (10 mg.) of silicotungstic acid in four different volumes. The results are presented graphically in Figure 3. If the volume a t the end point is constant t o 5 per cent, very little effect will be noted. To increase the accuracy, a correction can be read from the curve and applied or a second aliquot can be titrated in which the exact volume of water is used. The results recorded in Table I were obtained with volumes constant to 2 per cent. The concentration of formic acid was varied from 0.01 t o 0.03 AT without affecting the end point. At concentrations below 0.01 X the end point was indefinite. Concentrations above 0.03 111 were not investigated, but it is known that the precipitate is more soluble with increasing concentration of acid. Hydrochloric acid cannot be used, because it promotes the formation of large crystals. Acetic acid was tried, but it apparently exerts a solvent action on the precipitate of nicotine silicotungstate. Among the stabilizers tried were agar-agar, gum karaya, gum tragacanth, gum arabic, saponin, gelatin, and Irish moss extract. Irish moss was the only one that gave any protection. -4very small amount of this will prevent flocculation for several days. The equipment required for a microdetermination of this type is inexpensive. A photoelectric apparatus such as the one described above can be built in any laboratory for about $25. KOexpense need be added for the titration cell, since it can easily be made where glass-blowing equipment is available. For the analysis of small quantities of nicotine, this turbidimetric titration appears to be a rapid and accurate method. It eliminates errors due to solubility of the precipitate and avoids an indefinite end point. For extremely accurate results the Spies gravimetric method ($?)is preferable, but where a large number of analyses must be run and results to 5 or R micrograms are sufficiently accurate, this method has many advantages. This method of titration appears to be applicable in the determination of other substances whenever a good precipitation reaction can be found. Experimental work on its application to the analysis of other materials is in progress.
Literature Cited
U
0 2
Nicotine Solution Taken
0 5 ...
... ...
1.15
...
. . . . . .
Sicotine Found Gram
THE
VOL. 10, NO. 1
'
12
I3 14 IS 16 VOLUME CCC.)IN TITRATION CELL AT
17
END
I8
IS
POINT
FIQURE 3. EFFECTOF VOLUMEON END POIKT Two cubic centimeters (10 mg.) of silicotungstic acid were titrated in four different total volumes. T h e relationship shown is equivalent t o a rectilinear variation of titer with concentration of silicotungstic acid.
(1) Campo, A. del, Burriel, F., and Escolar, L. G., Bol. acad. cienc.
(Madrid), 2 (7), 10-13 (1936). (2) Spies, J. R., IND. EX. CHEM.,Anal. Ed., 9, 46-7 (1937). RECEIVED Sovember 8 , 1937.