112 On a New Method for the Determination of Atmospheric Carbon Dioxide , based on the Rate of its Absorption by a Free Surface of a Solution of Caustic Alkali .
By Horace T. Brown , F.R.S. , and F. Escombe .
( Received January 9 , \#151 ; Read March 23 , 1905 .
) Tn an appendix to a paper on the static diffusion of gases , communicated to the Society in 1900 , * it was shown that when a current of air containing a constant proportion of carbon dioxide is caused to move in a turbulent stream over the free surface of a solution of caustic alkali , the rate of absorption of that gas increases with the velocity of the air-current up to a certain optimal speed , beyond which no further increase in the speed of the current influences the rate of absorption .
It was further shown that when the optimal velocity of the air-current has been reached , and the temperature is maintained practically constant , the rate of absorption then varies directly as the partial pressure of the carbon dioxide in the air .
In other words , if under the above conditions the rate of absorption per unit of area of the liquid surface is a for a partial pressure of carbon dioxide represented by and is for a partial pressure of p ' , then at similar temperatures , = a jpf .
A suggestion was also made that this principle might be found applicable to a determination of the carbon dioxide in air , and that if the method were found to be a practical one it would have the manifest advantage of not requiring any measurement of the air from which the gas was absorbed .
Since the first account of our experiments we have made a more complete investigation of the principles underlying the proposed method , and have succeeded in determining the coefficient of absorption of carbon dioxide under these conditions , and also the manner in which this coefficient is influenced by temperature .
When certain precautions are taken , this simplified method of determining atmospheric carbon dioxide gives good results , which in point of accuracy approach those obtained by the more elaborate method with the Reiset 's apparatus as described in the preceding paper .
The absorption-apparatus used in the experiments was a slightly modified form of one given at b , fig. 8 of 'Phil .
Trans. , ' B , Yol .
193 , p. 284 , which is here reproduced with its present modifications .
It consists of a glass cylinder A , about 15'5 cm .
long and 6*5 cm .
wide , closed with an india-rubber plug with a central perforation , through # See " Phil. Trans./ B , vol. 193 , p. 282 .
Determination of Atmospheric Carbon Dioxide .
which passes a short wide tube B of 2'5 cm .
diameter , opening out into the cylinder below , and having a side tube which can be connected with the aspirator , or the aspirator and meter if measurements of the air are required .
The upper part of the tube B is closed with a cork , through which passes a narrower tube C terminating below in a funnel , the width of which is only slightly less than that of the cylinder .
The mouth of the funnel is closed with a perforated porcelain disc , which is luted in with paraffin .
When air is aspirated through the side tube b it enters the apparatus down the narrow tube c , and passing through the perforated plate , impinges as a turbulent stream on the surface of the absorbing solution of caustic soda at d. In order to distribute the air more completely , a small cone of paraffin e is supported on a needle in the position shown in the Figure .
The funnel can be adjusted to any desired distance from the surface of the liquid in the cylinder , the ordinary working distance being 1 cm .
100 c.c. of a solution of caustic soda of approximately 4 per cent , concentration are introduced into the cylinder through an india-rubber cap temporarily placed over its mouth , and a similar arrangement at the close of the experiment admits of the liquid being titrated without removal to another vessel.* At the point reached in the cylinder by 100 c.c. of liquid , the area exposed was 32'557 sq .
cm .
Hart 's method of double titration was used for determining the carbon dioxide absorbed by the alkali ; this has been already fully described.f In determining the constants of the above apparatus the air-current employed was divided into two parts , one of which passed through the surface absorber , and the other through a Beiset 's tower charged with the solution of caustic soda by means of which an independent and accurate determination of the carbon dioxide was obtained .
In both cases the volumes of air passed were measured by standardized meters .
The first series of experiments , of which the results are given in the following Table , was undertaken for testing still further the accuracy of the above * Care must be taken not to wet the sides of the cylinder above the level of the liquid , t See 4 Phil. Trans. , ' B , vol. 193 ( 1900 ) , p. 289 ; also 4 Eoy .
Soc. Proc. , ' this vol. , p. 35 .
YOL .
LXXYI.\#151 ; B. I Table I.\#151 ; Showing the Relation of the Surface-absorption of Carbon Dioxide by a 4 per cent. NaHO Solution to the Partial Pressure of that Gas in the Air-Stream when the Velocity of the Air-Current exceeds the Optimum for Maximal Absorption .
experiment which is taken as unity for the ratios of absorptions and of partial pressure is one in which the C02 content is practically that of normal air .
1 .
No. of experiment .
2 .
Mean temperature , Centigrade .
3 .
Rate of flow of air-current i hnmi rrB o.banvnf , inn armaratus .
4 .
C02 in vols .
per 10,000 of dry air determined by Reiset 's apparatus .
5 .
Partial pressure of C02 in moist air in terms of tottoo atmosphere .
6 .
C02 absorbed per hour by 32 557 sq .
cm .
of surface of NaOH ( in cubic centimetres of C02 at If .
T. P. ) .
7 .
CO2 in cubic centimetres absorbed by 1 sq .
cm .
of surface per hour for one part in 10,000 of C02 .
8 .
Ratios .
Litres per hour .
Mean linear velocity in metres per hour .
( a ) Of partial pressures of co2 .
( f\gt ; ) Of rates of absorption of C02 .
1 14'-1 192 9 328 0-040 0 039 0 -076 0 -0583 0-012 0-013 2 12'-3 187 '2 319 0-189 0-186 0-324 0 -0526 0 -056 0-058 3 14''4 163 1 278 0-568 0-558 0-982 0 -0531 0-168 0-177 4 14'-8 146-2 249 1 -080 1-062 1-870 0 -0531 0-320 0 337 5 12''4 162 -3 276 1 -152 1135 1 -984 0 -0553 0-342 0-357 6 ll'-2 167 *3 285 3-216 3 -174 5-252 0 0501 0-958 0 " 946 7 13 ' 6 140 -1 238 3-364 3-313 5-547 0 -0506 1 -ooo 1 -ooo 8 15'3 156 -0 266 4-588 4-511 7-725 0 -0517 1 -361 1 -392 9 13''8 164-1 279 4-893 4-817 8 -329 0 -0522 1 -453 1-501 10 13''5 155 -3 264 6-759 6 " 656 11 -276 0 -0512 2-009 2-032 11 13''2 143 -7 245 7-653 7 -540 12 -492 0 -0501 2-275 2 -250 12 13'-7 155 -3 264 8-547 8-416 13 -863 0 *0498 2*540 2 -499 13 13''7 167-1 284 12-532 12 -341 20 -873 0 -0511 3-724 3-763 14 15'-2 156-8 267 17 -949 17 -645 30 -905 0 -0528 5-325 5 -571 114 Dr. H. T. Brown and Mr. F. Escombe .
1905 .
] Determination of Atmospheric Carbon Dioxide .
proposition that under the conditions postulated , the rate of surface-absorption of carbon dioxide is proportional to the partial pressure of that gas in the moving air-current .
A preliminary series of experiments , which need not be quoted here , had already established the fact that with this special form of apparatus , and with amounts of carbon dioxide not exceeding about 14 parts per 10,000 of air , a velocity of current of about 150 litres per hour was sufficient to ensure a maximum rate of absorption .
With the funnel adjusted to a distance of 1 cm .
from the surface of the liquid , this corresponds to an average forward movement in the turbulent air-stream of about 260 metres per hour .
In those cases where it was required to employ air containing less than the normal amount of carbon dioxide , part of the air was previously passed through a tower containing soda-lime , the necessary admixture with ordinary air being made on emergence from the tower before it was divided between the two forms of apparatus .
When , on the other hand , air richer in carbon dioxide than ordinary air was required , the air stream was passed through a tower containing fragments of marble , on which there dropped a graduated flow of one-tenth normal hydrochloric acid .
In the last two columns of the above Table are given side by side the ratios of the partial pressures of the carbon dioxide in the air employed , and the ratios of the rates of absorption by the surface of the alkaline solution .
It will be seen that these series of values are practically identical , thus clearly showing the direct relation between the partial pressures and the rates of absorption .
The values given in the seventh column were obtained by dividing the number of cubic centimetres of carbon dioxide ( measured at N. T. P. ) absorbed per square centimetre of liquid surface per hour by the number of volumes of carbon dioxide contained in 10,000 volumes of dry air .
The results represent the coefficient of absorption of carbon dioxide , stated in the form of cubic centimetres per square centimetre per hour , corresponding to a uniform partial pressure of 1/ 10000 of an atmosphere .
Had there been an exact correspondence between partial pressure and rate of absorption in all these experiments , this coefficient would have been a constant .
The variations which occur are largely due to differences of temperature ranging from ll'-2 to 15a3 C. In order to investigate the influence of temperature on the rate of surface-absorption of carbon dioxide by soda-solutions , a separate series of experiments was undertaken , in which the absorption cylinder was immersed in Dr. H. T. Brown and Mr. F. Escombe .
[ Jan. 9 , water , the temperature of which was observed at frequent intervals.* Six experiments of this nature were made between 130,7 and 23a7 C. , ordinary-air being used , of which the content of carbon dioxide was independently-observed with the Reiset 's apparatus .
Table II.\#151 ; Showing the Influence of Temperature on the Coefficient of Absorption of C02 by the Surface of a 4 per cent. NaOH Solution .
Temperature .
13'-7 15'-3 15'*7 19'-5 21'-8 23'-7 Coefficient of absorption of CO2 per square centimetre of liquid surface per hour , for 1 vol. of C02 in 10,000 vols .
of air ( dry ) .
00519 0-0537 0-0537 0-0583 0-0623 0-0635 On plotting out these results , it may be seen that within the experimental limits the coefficient of absorption varies directly with the temperature , and that an increase of temperature of 1 ' C. corresponds to an increase of O'OOIS in the absorption coefficient .
Hence the value of this coefficient for any temperature t ' within the above limits will be represented by 0-0356 -f t ' 0"0018 , 0 0356 being the coefficient of absorption at 0 ' C. , as determined by extra polation .
Assuming now that the relation of the partial pressures to absorptions is of the simple nature already postulated , then the volume of carbon dioxide contained in 10,000 volumes of dry air should be given by the formula A 0-0356 + t'x 0-00118 ' where A represents the carbon dioxide absorbed per square centimetre of surface per hour , at temperature t ' , from an air-stream which is drawn over the absorbing liquid at a sufficient rate to ensure the limit of maximal absorption being reached ; A being stated in terms of cubic centimetres at normal temperature and pressure .
With the knowledge thus gained we are now in a position to compare the estimations of carbon dioxide made by means of the method of surface-absorption with those made with the Eeiset 's apparatus .
* In order to avoid evaporation and consequent cooling of the absorbing solution , the stream of air was previously saturated with moisture before being passed into the absorption apparatus .
1905 .
] Determination of Atmospheric Carbon Dioxide .
In Table III we have taken the results obtained with the surface-absorption method as given in Table I , in which very variable amounts of carbon dioxide were present in the air-stream , and have calculated the amount A present in volumes per 10,000 by means of the formula q-q35q + f\gt ; x'q~qq2 ] g Alongside these results we have given the amount of carbon dioxide found by an independent determination with the Reiset 's apparatus .
Table III.\#151 ; C02 in Parts per 10,000 of Dry Air .
No. of experiment , j By Reiset 's apparatus .
By surface absorption with NaHO solution .
A 0-0356 + t ' x 0-00118 ' Difference .
1 0-04 0-04 0^00 2 0-19 0-19 0-00 3 0-56 0-57 + 0-01 4 1-08 1-08 0-00 5 1*15 1-21 + 0-06 6 3*21 3-30 + 0-09 7 3-36 3-30 \#151 ; 0-06 8 4-58 4-41 -0-17 9 4-89 4-93 + 0-04 10 6-75 6-72 -0-03 11 7-65 7-50 -0-15 12 8-54 8-23 -0-31 13 12-53 12-40 -0-13 14 17-94 17-74 -0-20 A further comparative test of the two methods was made with ordinary air , and gave the following results :\#151 ; Table IV.\#151 ; C02 in Parts per 10,000 of Dry Air .
No. of experiment .
Tempera- rure .
By Reiset 's apparatus .
By surface absorption with NaHO .
A 0-0356 + t ' x 0-00118 ' | Difference .
1 15'-3 3-29 3-29 \#151 ; o-oo 2 15'-7 3-33 3-32 -o-oi 3 21'-8 3-05 3-10 + 0-05 4 23'-7 3-15 3-16 + 0-01 Jo dr ell Laboratory } Royal G , Ke\lt ; io .

