I14 A N A L Y T I C A L E D I T I O N Vol. 5 , No. 2

Hydrogen peroxide, made u p in 1.0 N sulfuric acid solution, was found to be stable for as long as 3 weeks.

LITERATURE CITED (1) Adie and Wood, J . Chm. Soc., 77, 1076 (1900). (2) Atkinson, Pham. J., 16, 809 (1886). (3) Beckurts, “Massanalyse,” p. 515, Friedr. Vieweg und Sohn,

Braunschweig, 1913. (4) Clarke, Analyst, 36, 393 (1911). (6) Davisson, J . Am. Chem. SOC., 38, 1683 (1916). (6) Feldhaus, 2. anal. Chem., 1,426 (1862). (7) Fisoher and Steinbaoh, Z. anorg. Chem., 78, 134 (1912). (8) Fresenius-Cohn, “Quantitative Chemical Analysis,” Vol. 2,

(9) Green and Evershed, J. SOC. Chem. Ind., 5, 633 (1886); Z. anal.

(10) Griffin, “Technlcal Methods of Analysis,” p. 28, McGraw-Hill,

(11) Grossman, C h m . - Z t ~ . , 16, 818 (1892).

p. 196, Wiley, 1911.

Chem., 26, 638 (1887); Chem. News, 65, 109 (1892).

1921.

(12) Griitzner, Arch. Pharm., 235, 241 (1897). (13) Kinnicut and Nef, Am. Chem. J. , 5, 388 (1883!; (14) Kolthoff and Furman, “Volumetric Analysis,

(15) Kubel, J. prakt. Chem., 102, 229 (1867). (16) Laird and Simpson, J. Am. Chem. Soc., 41,524 (1919). (17) Lunge, Ber., 10, 1074 (1877); Chem.-Ztg., 28, 501 (1904). (18) Moir, J. S. African Assoc. Anal. Chem., 4, 3 (1921). (19) Pean de St. Gilles, Ann. chim. phys., 55, 383 (1859). (20) Phelps, Am. J . Sci., 167, 198 (1904). (21) Raschig, Ber., 38, 3911 (1905). (22) Robin, J. pharm. chim., (vi), 7, 575 (1898). (23) Rupp and Lehmann, Arch. Pharm., 249, 214 (1911). (24) Sanin, J . Russ. Phys.-Chem. SOC., 41, 791 (1909). (25) Scott, “Standard Methods of Chemical Analysis,” p. 521, Van

(26) Winkler, Chem.-Ztg., 23, 454 (1899).

Vol. 2, p. 302, Wiley, 1929.

Nostrand, 1917.

RFJCEIVED November 29, 1932. cal Laboratory, University of Virginia.

Contribution 109 from the Cobb Chemi-

Determination of Carbon Dioxide in Continuous Gas Streams

WILLIAM McK. MARTIN AND JESSE R. GREEN Department of Chemistry, Montana Agricultural Experiment Station, Bozeman, Mont.

The problem of quantitative removal qf carbon dioxide f rom continuous gas streams by absorption in barium hydroxide and its subsequent determi- nation by direct titration has been investigated. An eficient absorber for carbon dioxide or other soluble gases, in which the absorbing solution may be directly titrated, is described.

The eficiency of absorption of soluble gases f rom a continuous gas stream is determined by the design of the absorption vessel, the concentration of the ab- sorbing solution, and the rate of flow of the gas mixture. The eficiency of a n absorption vessel is

N AN at tempt to s tudy the effect of petroleum spray oils on the respiratory activity of growing plants, it was I discovered tha t the carbon dioxide in the air drawn con-

tinuously from the respiration chambers could not be com- pletely absorbed by the apparatus and methods generally used. As a foundation for subsequent work on respiration it was decided to investigate the methods for absorbing and quantitatively determining carbon dioxide in continuous gas streams.

For the benefit of those who are confronted with the prob- lem of choosing, or devising, a method or determining carbon dioxide in other than routine studies, the results of a litera- ture review are herein presented. The methods described in the literature have been developed largely in connection with the determination of carbon dioxide derived from the oxidation of organic substances in combustion analysis, from the oxidation of carbon in steel, from carbonates, and from the respiratory activities of living organisms. Upon the basis of differences in principle the methods may be briefly classified as follows :

1. Gravimetric Methods. a. Carbon dioxide absorbed in strong alkaline solutions,

or solid absorbents, such as soda lime and ascarite (sodium

determined more by the length of the path of the gas bubbles through the absorbing liquid than by the extent of surface exposed to the gas stream.

The addition of barium chloride to barium hy- droxide solution appreciably increases its eficiency in absorbing carbon dioxide f rom continuous gas streams. However, the small amounts generally used have but little practical effect. W h e n solutions containing suspended barium carbonate are stirred with a ciarreni of carbon dioxide-free air they cannot be litrated with acid stronger than approximately 0.07 N .

hydroxide dispersed on asbestos fibers), and the increase in weight of absorbent determined.

b. Absorbed in dilute solution of sodium or potassium hydroxide and weighed as barium carbonate.

c. Absorbed in a strong solution of barium hydroxide and the resulting carbonate either weighed directly or converted to the corresponding sulfate.

a. Carbon dioxide absorbed in a dilute standard solution of sodium or potassium hydroxide and the carbonate deter- mined by double titration, using phenolphthalein and methyl orange or other suitable indicators.

b. Absorbed in a standard solution of barium hydroxide and the carbonate determined by a single titration.

c. Absorbed in a dilute standard solution of sodium or potassium hydroxide, an excess of neutral barium chloride added, and the excess of alkali titrated.

d. Absorbed in a strong solution of barium hydroxide and the resulting carbonate filtered, washed, dissolved in a stand- ard acid solution, and titrated.

a. By volumetric gas apparatus. Carbon dioxide ab- sorbed directly in strong alkali in gasometric apparatus and the decrease in the volume of the gas mixture measured a t atmospheric pressure; or absorbed in strong alkali, carbon dioxide liberated in an acid solution in a gas apparatus, and its voIume measured a t atmospheric pressure.

b, By manometric gas apparatus. Carbon dioxide ab- sorbed in strong alkaIi and Iiberated in the manometric gas

2. Titrimetric Methods.

3. Gasometric Methods.

March 15, 1933 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

apparatus by excess acid. Volume of the gas then increased to a fixed value and its pressure measured manometrically.

4. Electrometric Methods. a. Electrolytic resistance method. Carbon

dioxide absorbed in a standard solution of alkali, preferably barium hydroxide, and t,he decrease in its electrolytic conductivity measured.

b. Thermal conductivity method. The con- centration of carbon dioxide in a gas mixture is indicated by its thermal conductivity, which is determined by measuring the electrical resist- ance of a wire heated by a constant electric current and surrounded by the gases under study.

5. Colorimetric Method. Carbon dioxide ab- sorbed in a solution of the sodium salt of phenol- phthalein andits concentration indicated by color

PREVIOUS INVESTIGATIONS The chief advantage of the g rav ime t r i c

method is the relatively simple a b s o r p t i o n apparatus which is permitted by the use of s t r o n g alkaline solutions or solid absorbents. The simplicity of the method is offset, how- ever, by the precautions which are necessary in order to prevent moisture and other gases be- ing collected in the absorption vessels. The dis- advantages of the gravimetric methods have been discussed in detail by Cain (6) and Truog (55). Most of the objectionable features of the gravimetric methods have been overcome by Cain by absorbing the carbon dioxide in barium hydroxide and then weighing the barium car- bonate, or its equivalent, as the corresponding sulfate.

Titrimetric methods are preferable to gravi- metric methods in that they are more rapid and less tedious. Un l ike the g r a v i m e t r i c methods, however, they necess i t a t e the use of dilute standard alkaline solutions which a b s o r b c a r b o n dioxide less efficiently, and i n consequence more elaborate absorption vessels must be employed. When dilute standard solu- tions of sodium and potassium hydroxides are used, the carbon dioxide absorbed is determined either by the double titration method of Brown and Escombe (5), or by pre- cipitating the carbonate with neutral barium chloride and then determining the excess alkali by a single titration. The single titration method is simplified, however, by ab- sorbing the carbon dioxide directly in a standard solution of barium hydroxide. Although some investigators, notably Humfeld (19), Willaman and Brown (45), Marsh ( 2 4 , and Newton and Anderson (%'), have obtained satisfactory results with the double titration method, the use of barium hydroxide with a single titration is more convenient. Various modifications of the single titration method have been de- scribed in the literature by Mack (%!?), Blackman and Parija (4, Neller (26), Tang (34 , Schollenberger (29), Crozier and Naves ( I O ) , Heck, ( I 7 ) , and others.

The gasometric methods for determining carbon dioxide, especially those devised by Van Slyke (36 to 41) and his co-workers, Scholl and Davis (28), Blacet and Leighton (S), Chittick (8), and Coe (Q), are capable of attaining a high degree of accuracy. They are designed, however, for de- termining the carbon dioxide in relatively small volumes of gases and therefore cannot be conveniently applied in the analysis of gas streams over extended periods.

The electrolytic resistance method was devised by Cain and Maxwell (7) for determining the carbon in steel and has been modified by Harvey and Regeumbal (16), Gane (15), and Ledebur (21). The necessity of expensive apparatus and careful temperature control has doubtless prevented this

115

FIGURE 1. CONSTANT TEMPERATURE RESPIRATION CABINET

method from coming into general use. It is especially useful for determining progressively the respiration rate of living organisms, and for studying other reactions in which it is desired to measure the rate a t which carbon dioxide is pro- duced.

The thermal conductivity method for determining the carbon dioxide in gas mixtures is based upon the principle that the thermal conductance of a gas is determined by its composition. The thermal conductivity of the gas may be determined by measuring the electrical resistance of a wire traversed by a constant electrical current and surrounded by the gas enclosed in a space, the walls of which are a t a constant temperature, Precautions are, of course, taken to prevent the dissipation of heat energy from the wire by radiation, conduction through the ends of the wire, and by thermal convection. The theoretical aspects of the method have been discussed by Daynes (12, I S ) , Shakespear ( S I , SZ), and by Weaver et al. (44). In the latter paper is presented an excellent review of the history of the development of the method, together with a valuable contribution in the way of experimental apparatus, methods, and data. Although the method has been used very largely for analysis of aliquots of gas mixtures, it has been applied satisfactorily in respira- tory studies by Waller (42) and by Stiles and Leach ($5).

A colorimetric method, based upon the relation of color to the amount of carbon dioxide absorbed by a solution of the sodium salt of phenolphthalein, has been described by Emmert (14) and Hughes ( I S ) . This method is apparently limited in its applications and has, therefore, not been used very generally.

116 A N A L Y T I C A L E D I T I O N Vol. 5 , No. 2

FIGURE 2. TITRATION APPARATUS

Under average laboratory conditions, the single titration method is, perhaps, the most satisfactory for determining carbon dioxide in continuous gas streams. It has therefore been applied in developing the methods and apparatus used in this investigation.

EXPERIMENTaL

APPARATUS AND METHODS. The respiration apparatus with accessories is shown in Figure 1. It consists of a con- stant temperature cabinet containing six respiration chambers with carbon dioxide absorbers, and is provided with a source of carbon dioxide-free air. The cabinet is constructed from wood, lined with Celotex, and is maintained a t a constant temperature by a thermoregulator and an electric fan. The respiration chambers are formed by sealing glass bell jars to ground-glass plates with a compound prepared by melting together 16 parts of vaseline, 4 parts of raw rubber, and 1 part of paraffin. As a precaution against leakage the chambers are set in shallow trays of water. A stream of carbon dioxide-free air is supplied by drawing air from out- side the laboratory through a large soda-lime tube, 5 cm. in diameter and 62 cm. long, placed inside the cabinet in order to bring the temperature of the incoming air to that of the cabinet.

The material to be studied is placed in chamber D. Air is drawn from outside the laboratory by suction, entering the cabinet through tube A , and passing slowly upward to C through the soda-lime tube B. From the latter it is distrib- uted to each of the six respiration chambers, only one of which is shown at D in the diagram. The carbon dioxide- free air enters the top of the chamber, carrying downward with it the carbon dioxide respired by the plant or other material being studied. The carbon dioxide-laden air is drawn from the bottom of the chamber through tube E and is discharged from the jet F into absorber G, which is charged with barium hydroxide solution. The rate at which the gas mixture passes through each absorber is adjusted by the needle valve H , the rate being determined by Cali-

brating with a liquid displacement method . The needle valves are not altered after calibration.

The suction applied to the system is maintained constant by the pressure regulator J , which con- sists of a valve formed by a r u b b e r s t o p p e r seated against the end of a 25-mm. glass tube. The end of t,he tube is ground to form B tight joint, and the valve (stopper) is held in place by a spiral sp r ing , the t e n s i o n of which i s adjusted by the nut M . The reduced pressure (suction) of the system is indicated by the water m a n o m e t e r L, and the pressure inside the respiration chambers is indicated by the ms- nometer K.

When all six respiration units are in operation with the gas mixture passing through each at the rate of 20 liters per hour, the pressure in the chambers is not reduced by more than 1 cm. of water below that of the atmosphere.

In making the determinations, ample time is allowed for the temperature of the material in the respiration chambers to come to equilibrium. Atmospheric carbon d i o xi d e remaining in the apparatus after being assembled is removed by aspirating with carbon dioxide-free air for 2 hours before connecting with the absorption vessel. At the end of the respiration period the absorption vessels are removed and connected with the appa- ratus in Figure 2, and the excess barium hydroxide titrated against 0.0454 N hydrochloric acid (1 ml.

equals 1 mg. of carbon dioxide), using phenolphthalein or thymolphthalein as an indicator. The carbon dioxide-free compressed air supplied by the above apparatus not only stirs and circulates the liquid in the absorption vessel during titration, but it also prevents the diffusion of atmospheric carbon dioxide into the apparatus. Although phenolphthalein is most generally used as an indicator, it is not entirely satis- factory on account of the indefiniteness of its end point. Its transition range is from pH 8.3 to 10.0 and consequently i t shows a faint color due to the slight solubility of the suspended barium carbonate. Schollenberger (SO) has recommended thymolphthalein as a more satisfactory indicator for the titrimetric estimation of carbon dioxide. It shows a sharper end point because its transition range is narrower and above the region of reaction of the dissolved barium carbonate. Obviously, the same indicator must be used in making the blank titration.

Of the various gas absorption vessels described in the literature, those designed by Weaver and Edwards ( 4 S ) , Milligan (M), Beaumont, Willaman, and DeLong ( I ) , and Harvey and Regeumbal (16) are probably the most efficient. The absorbers described by Beaumont, Willaman, and De- Long and by Milligan are the only ones designed to permit titration directly in the vessel. Milligan absorbers, having been used in a previous investigation, were available, and were therefore used in the present experiments. A serious objection to these absorbers, however, is their large capacities for the absorbing liquid. They will not operate properly with a volume of less than 130 ml. of liquid, and, conse- quently, relatively weak alkali solutions have to be used in order not to exceed the capacity of the vessels in titrating with dilute standard acid. Strong acids cannot be used in titrating in the presence of suspended barium carbonate.

In a preliminary investigation the respiration rates of standard bean seedlings were determined, using Milligan absorbers with the respiration cabinet shown in Figure 1. The absorbers were charged with 60 ml. of 0.0454 N barium hydroxide and 70 ml. of distilled water neutralized with barium hydroxide, using phenolphthalein as an indicator.

March 15, 1933 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 117

The absorptions extended over 2-hour periods and the titra- tions were made with 0.0454 N hydrochloric acid, as already described.

An analysis of the data obtained in these experiments indicated that the carbon dioxide was not being completely absorbed. Consequently, the Milligan absorbers were abandoned and some time devoted to the development of apparatus better suited to the problem undertaken.

The bead absorber shown in Figure 3 was first developed, but was found to be unsatisfactory in that it necessitated the use of not less than 60 ml. of absorbing liquid. Barium hydroxide solutions stronger than 0.105 N could not be used in these absorbers on account of the excessive volume of standard acid required in titration.

The spiral absorber shown in each of the three figures was finally developed and was found to be highly satisfactory. It will operate efficiently on 30 ml. of absorbing liquid, thus permitting relatively concentrated solutions of barium hydroxide to be titrated without using excessive volumes of standard acid. The spiral is 40 mm. in diameter and is made from 4-mm. (inside bore) tubing. The reservoir has a capacity of 200 ml.

+

FIGURE 3. ABSORPTION VESSELS

EFFICIENCY OF ABSORBERS. The efficiency of each of the absorbers was determined by passing through it a current of air a t the rate of 20 liters per hour, into which was liberated a known amount (29.1 mg.) of carbon dioxide. This was accomplished by slowly dropping 50 ml. of 0.0264 N sodium carbonate solution into 100 ml. of 1.0 N sulfuric acid. The sodium carbonate solution was run into the sulfuric acid in a closed flask from a pipet, the orifice of which was adjusted to deliver 50 ml. in 20 minutes. A current of carbon dioxide- free air was bubbled through the sulfuric acid solution carry- ing the carbon dioxide liberated to the absorber. The stream of carbon dioxide-free air was passed through the system for 40 minutes after adding the sodium carbonate solution in order to sweep into the absorber any carbon dioxide re- maining in the system. Three absorbers were inserted in the chain so that the carbon dioxide not absorbed in the first would be caught in the other two.

The efficiency of absorption, as influenced by the concen- tration of barium hydroxide in the absorbing liquid, was determined for each of the three absorbers. Concentrations of barium hydroxide greater than 0.0525 and 0.105 N could not be used with the Milligan and bead absorbers, respec- tively, on account of the large volumes of standard acid required in titration. An attempt was made to titrate the

large volumes of standard barium hydroxide solutions in these absorbers by first adding a measured volume of strong standard acid and then titrating to the end point with weak acid. However, these studies showed that when the strong acid was added in the presence of suspended barium car- bonate, carbon dioxide was liberated by local concentrations of the acid. The carbon dioxide was lost by being aspirated from the solution by the carbon dioxide-free air used to stir and circulate the liquid in the absorber during titration. With this method it was shown that acid more concentrated than approximately 0.07 N could not be used in titrating barium hydroxide solutions containing suspended barium carbonate without losing carbon dioxide. The effect of rate of flow of the gas mixture on the efficiency of absorption, using 0.0167 N and 0.105 N barium hydroxide, was also investigated. The results of these studies are presented in Tables I and 11.

TABLE I. COMPARATIVE EFFICIENCIES OF SPIRAL, BEAD, AND MILLIGAN ABSORBERS

(Gas mixture paseed through absorber at 20 liters per hour. 29.1 mg. carbon dioxide present)

ABBORBING SPIRAL ABSORBER BEAD ABSORBER MILLIGAN ABSORBER LIQUID COzab- Effi- COzab- Effi- (%ab- Effi- Ba(OHh sorbed ciency sorbed ciency sorbed ciency

MQ. % .WQ. % MQ. % 0.1050 N 29.2 100.1 27 .6 9 5 . 1 . . . . . . . . 0.0780 N 2 9 . 1 100 .0 2 7 . 3 94 .7 . . . . . . . . 0.0525 N 28 .6 9 8 . 6 2 5 . 7 88.6 2 7 . 7 9 5 . 5 0.0350 N 2 8 . 6 9 8 . 6 2 4 . 3 8 3 . 8 2 6 . 1 8 9 . 8 0.0262 N 27.9 9 6 . 2 2 3 . 9 8 2 . 5 2 5 . 9 8 9 . 3 0 ,0210 N 2 7 . 4 9 3 . 8 2 3 . 7 8 1 . 8 2 5 . 9 8 9 . 3 0.0175 N 2 5 . 7 8 8 . 6 2 3 . 0 7 9 . 4 2 4 . 1 8 3 . 2

TABLE 11. EFFICIENCY OF SPIRAL ABSORBER WITH GAS MIX- TURE FLOWING AT DIFFERENT RATES

(Absorber charged with 0.067 N and 0.105 N Ba(0H)d CARBON DIOXIDE ABSORBBD EFFICIENCY

RATE OF FLOW 0.067 N 0.105 N 0 067 N 0.105 N OF GAS MIXTURE Ba(OH)s Ba(OH)z Ba(OH)z Ba(OH)z

Literdhr. Mu. Mu. % % 20 2 9 . 2 2 9 . 2 100.1 100: 1 25 2 8 . 4 28 .1 9 7 . 8 9 7 . 0 30 2 8 . 2 27 .7 9 7 . 2 96.6 40 2 5 . 6 2 6 . 2 8 8 . 3 9 0 . 3 50 2 2 . 1 2 4 . 9 7 6 . 3 85.8

EFFECT OF BARIUM CHLORIDE ON THE EFFICIENCY OF ABSORPTION OF CARBON DIOXIDE BY BARIUM HYDROXIDE SOLUTIONS. A number of investigators recommend the addition of barium chloride t o the barium hydroxide solution. As observed by Mack (IS), the mass-action effect of barium chloride reduces the solubility and hydrolysis of the barium carbonate. In the present work this would not only increase the efficiency of absorption of carbon dioxide, but would tend to prevent the decomposition of barium carbonate by local concentration of the acid added in titration.

TABLE 111. EFFICIENCY OF ABSORPTION OF CARBON DIOXIDE BY 0.0175 N BARIUM HYDROXIDE IN SPIRAL ABSORBER, AS INFLUENCED BY DIFFEREXT CONCENTRATIONS OF BARIUM

CHLORIDE COS RECOVERED

IN FIRST AB- CONCENTRATION CARBON DIOXIDE ABSORBED SORBER

OF BaClz Absorber 1 Absorber 2 Absorber 3 Efficiency M a MQ . Mg. %

0.00 N 0 .01 N 0 . 1 0 N 0 . 2 0 N 0 40 N 0 80 A'

2 6 . 4 2 6 . 5 2 7 . 0 2 7 . 5 2 7 . 7 2 8 . 0

3 . 3 3 . 3 2 . 6 2 . 3 2 . 4 1 . 8

0 . 0 0 . 0 0 . 0 0 . 3 0 . 2 0 . 0

8 9 . 5 8 9 . 8 9 1 . 5 9 3 . 2 9 3 . 9 9 4 . 8

A series of determinations were, accordingly, carried out, using various concentrations of barium chloride. To show the influence of the latter more effectively, a very dilute solution (0.0175 N ) of barium hydroxide was used. The concentrations of barium hydroxide and barium chloride

118 A N A L Y T I C A L E D I T I O N Vol. 5, No. 2

in the solutions were adjusted by adding measured amounts of 1.0 N barium chloride and conductivity water to a measured volume of 0.105 N barium hydroxide. The results of these studies are presented in Table 111.

DISCUSSION The data show clearly that the efficiency of absorption of

carbon dioxide from a continuous gas stream is determined by the concentration of the absorbing liquid, by the rate of flow of the gases, and by the design of the absorption appa- ratus. Although it is generally assumed that a bead column is highly efficient in absorption, on account of the large surface of the absorbent exposed to the gases, these studies show it to be less efficient than either the spiral or Milligan absorbers. The average of the values for the efficiency of absorption for the five concentrations of barium hydroxide, ranging from 0.0175 N to 0.0525 N, is 83.22, 89.42, and 95.16 per cent for the bead, the Milligan, and the spiral absorber, respectively. These differences are apparently due to the differences in the length of the path of the gas through the three vessels. The gas bubbles travel approximately 39 em. in the bead, 76 cm. in the Milligan, and 142 cm. in the spiral absorber. The higher efficiency of the spiral and the Milligan absorbers may also be attributed to the stirring action of the walls of the tubes as the alternate columns of liquid and gas pass upward through them. As indicated by the work of Ledig and Weaver (22) and by Davis and Crandall ( I I ) , the layer of liquid a t the gas-liquid interface reaches equi- librium with the gas very quickly, and unless the dissolved gas is carried into the interior of the liquid by convection, the rate of absorption of the gas is determined by its diffusion rate in the liquid. The surface or stationary film is, of course, not stirred by the movement of the bubble through the tube, but the liquid adjacent to it is continuously renewed, thus removing the dissolved gas into the interior of the column of liquid following each gas bubble.

The method of titrating directly in the absorption vessel in the presence of suspended barium carbonate raises the question as to whether or not carbon dioxide is lost through the decomposition of the carbonate by the acid added during the titration. Truog (35), Bergman (a), Mack (23), and others have investigated this problem by the conventional method of shaking the solution during titration, and have concluded that the barium carbonate in the barium hydroxide solution does not affect the results. In this investigation, however, the solutions were stirred and circulated in the absorption vessels during titration by a current of carbon dioxide-free air, which would obviously remove by aspiration any carbon dioxide formed by local concentrations of acid. A series of titrations was, therefore, made with standard acid of various concentrations. The effect of adding the acid a t different rates was also studied. These studies showed that when solutions containing suspended barium carbonate are stirred with a current of carbon dioxide-free air during titration, carbon dioxide is lost if titrated with a standard acid stronger than approximately 0.07 N.

As shown by the data in Table 111, the addition of barium chloride to 0.0175 N barium hydroxide solution increased quite markedly its efficiency in absorbing carbon dioxide. It may be noted, however, that an equivalent weight or more of barium chloride per liter of barium hydroxide solution must be used in order to be of any practical value; -and that the small amounts generally used have but slight effect. Kostychev (20)) for example, recommends the addition of 1

gram, or 0.0096 equivalent weight of barium chloride to each liter of barium hydroxide solution. Mack (23) used 4.5 grams, or 0.0432 equivalent weight per liter. Obviously the mass action of such low concentrations could have but slight practical effect; and this was quite apparent in these studies. The effect would be even less with the more concen- trated barium hydroxide solutions which are generally used.

According to Mack (23) the barium chloride reduces the solution and hydrolysis of the barium carbonate during titration. In this work, however, the standard acid used in titrating the barium hydroxide solution was so dilute (0.0454 N) that no carbon dioxide was liberated from the suspended barium carbonate by local concentrations of the acid. The increase in the carbon dioxide recovered must, therefore, have been due to the mass-action effect of the barium chloride in increasing the absorption of carbon dioxide by the barium hydroxide solution, and not to its effect in reducing the loss of carbon dioxide during titration.

LITERATURE CITED (1) Beaumont, Willaman, and DeLong, Plant Physiol., 2, 487

(2) Bergman, Am. J. Botany , 12,641 (1925). (3) Blacet and Leighton, IND. EXG. CHmr., Anal. Ed., 3, 266 (1931). (4) Blackman and Parija, Proc. Roy . SOC. (London) , B103, 446

(5) Brown and Escombe, Trans . Roy . SOC. (London) , B193, 223

(6) Cain, J. IND. ENQ. CHEM., 6, 465 (1914). (7) Cain and Maxwell. Ibid.. 11. 852 (1919).

(1927).

(1928).

(1900).

(8j Chittick, J. Assoc.’O$cial A&. Chem., 6, 453 (1923). (9) Coe, Ibid. , 14, 99 (1931).

(10) Crozier and Navez, J. Gen. Physiol. , 14, 617 (1931). (11) Davis and Crandall, J. Am. Chem. Soc. ,S2, 3757 (1930). (12) Daynes, Proc. Roy . SOC. (London) , A97,273 (1920). (13) Daynes, Proc. Phys . SOC. (London) , 33, 165 (1921). (14) Emmert, J. Assoc. Oficial Agr . Chem., 14, 386 (1931). (15) Gane, Dept. Sci. Ind. Research (Brit.), Rept. Food Investiga-

tion Board, pp. 103-4 (1930). (16) Harvey and Regeumbal, Plan t Physiol., 1, 205 (1926). (17) Heck, Soil Sci., 28, 225 (1929). 118) Huehes. Chem. Eno. Minino Rev.. 24. 179 (1932’1

I , , I (igj Humfeld, soil S~i . r30, I (1930). (20) Kostychev, “Plant Respiration,” p. 31, Blackiston, 1927. (21) Ledebur, Mikrochem. Pregl Festschr., 1929,253. (22) MLedig and Weaver, J . Am. Chem. Soc., 46, 650 (1924). (23) Mack, Plant Physiol., 5 , l (1930). (24) Marsh, Soil Sci., 25,253 (1927). (25) Milligan, IND. ENQ. CHEM., 16, 889 (1924). (26) Neller, Soil Sci., 5 , 225 (1918). (27) Newton and Anderson, Can. J. Research, 5 , 337 (1931). (28) Scholl and Davis, IXD. ENG. CHEM., Anal. Ed., 3, 276 (1931). (29) Schollenberger, J. IND. ENG. CHEM., 8,427 (1916). (30) Schollenberger, Ibid. , 20, 1101 (1928). (31) Shakespear, Proc. Phys . SOC. (London) , 33, 163 (1921). (32) Shakespear, Electrician, 89, 543 (1922). (33) Stiles and Leach, Ann. Botany , 45,461 (1931). (34) Tang, J. Gen. Physiol., 15,87 (1931). 135) Truos!. J. IXD. ENG. CHEM.. 7. 1045 (1915). i36j Van giyke, J . Biol. Chem., 30,347 (1917). ‘ (37) Ibid. , 73, 121 (1927). (38) Van Slyke and Stadie, Ibid. , 49, 1 (1921). (39) Van Slyke and Sendroy, Ibid., 73, 127 (1927). (40) Ibid. , 95, 509 (1932). (41) Van Slyke, Sendroy, and Lin, Ibid. , 95, 531 (1932). (42) Waller, N e w Phytologist, 25, 109 (1926). (43) Weaver and Edwards, J. IND. ENG. CHEM., 7, 534 (1915). 144) Weaver, Palmer, Frantz, Ledig, and Pickering, Ibid., 12. 359

(45) Willaman and Brown, Plan t Physiol., 5, 535 (1930).

RWEIVED September 10, 1932. Presented before the Division of Agrioul- tural and Food Chemistry a t the 84th Meeting of the American Chemioal Sooiety, Denver, Colo., August 22 to 26. 1932. Contribution from Montana State College Agricultural Experiment Station, Paper 24, Journal Series.

(1920).