Determination of Sodium and Potassium, Employing Ion-Exchange

Note: In lieu of an abstract, this is the article's first page. Click to increase image size Free first page. View: PDF | PDF w/ Links. Citing Article...
1 downloads 10 Views 466KB Size
582

ANALYTICAL CHEMISTRY Table l V .

Amount Present, Compound Carbazole

9- Ethylvnrbasole

1)- Rutylcarbazule

of t h c two phases to the optinium v:ilue according to Bush and

Analysis of Carbazole Mixture llg.

4.0

2.0

1.0

Amount Tube Found, SO. Xg. 12 4.15 13 4.12 14 4.06 .iY. 4 , l l :3 4 35

36

2 14 2 07 2 12 .iv. 2 . 1 1

48

0.84

+!I

,,o

0.99 0.90 0.91

Error,

(;t

llensen ( 2 ) ,followed by the "completion of the square" method of ir:ic*tionation, is another possiblc niritns of increasing the efficiency of scparation of the homologs. ACK\IOW L E I K M E V T

+2.8

The author wishes to ackno\r-lcdge thc interest of hIilton Oicliin this work, the advice on qm-trnl measurements given by Robert Friedcl, and the trc~1iiiic:ilassistance of George Goldbach. 111

+5.5

LITEH.4Tl'Ht: CITED

L. C., J . B i d . Cheni.. 174, 221 (1948). ! 2 ) Bush, M. T., and Denscn. 1'. X I . , .\x.t!,. C h x . , 20, 121 (19481. 18) ( Y a k . L. C . . J . Bid. C'hcw.. 155. 519 119441. 14) ('raig, L. C., Golunibic, ('., Mi~hton,H., anf'l Titus, E., S ' c i ~ r ~ w . 103,2680 (1946). 6 )('raig, 1,. C., and Post, O., :is.\!.. ('HEM., 21, 500 (1949). 1 0 ) C;olumhic, C., J . A m . C'hrm. ,Sot., 71, 2627 (1949) 17) Sato, I-., Barry, G . T.,and Craig, L. C.. J . B i d . Chem., 170, 301 (1947). 1s) \\-arshowsky, B., m i d S:c,h:intz. I.zed by means of i l l i t h i . titmtioiis to find the total chloride. The amount of alkali chloritic. N W cil~tninedfrom the difforenco \ i c . t \ v c ~ ~the ii t\vo titrations.

’IY

JI

= molarity of alkali cation in any frwtioii

I.

=

.lI, = C*m = TI = = =

p

C

I .o =

Figure 1.

c- n - h c ~ r l .I[

-1I

= 2”

The dashed graphs of‘ Figure 1 ~verecalculated by Equation 3 wliere C is 2.37 arid 3.32, arid p is 1.21 X lo3and 1.19 X lo3 for sodium and potassium, respectively. These values were obtained by solving Equations 1 and 2 from the data of Figure 1. The close agreement betvieen the solid and dashed curves indicates the validity of Equation 3 and the close approach to equilibrium conditions during the elution. From these data, it is readily calculated that the average height of a theoretical plate is 0.048 cm. The average height per plate was not found to be constant as the column length was changed, but varied randomly from 0.04 to 0.12 cm. as the column length varied from 15 to 59 cm. Good agreement was not always obtained between the number of plates calculated from the sodium and potassium curves of a given run. Elution of Sodium and Potassium

( h l u m n , 3.80 sq. o m . X 59.0 c m . Exchanger, 59.5 g r a m s of colloidal Dowex 50 (oven-dried basis) Elutrient solution, 0.701 m o l a r HCI Flow rate, 0.60 rnl./min./bq. o m . E l u t i o n of 1.00 m i l l i m o l e of s o d i u m a n d 0.50 m i l l i m o l e of p o t a s s i u m Temperature, 32’ C. One V unit = 186 m l .

*\bout 5 hours are required f o i :III average elution (flow r‘ite = 0.60), from the time of introduction 01 the mixture into the column until the completion of the potawium elution. However, very little of the analyst’s attention is required during this period. I n the course of this investigation, it became desirable to hnown how closely equilibrium cwnditions were approached during elution, The symmetry of the elution curves was used :LS a criterion of equilibrium. ?’he,e curves are obtained by plotting the concentration of ulLali metal in the eluate as ordinate and volume of eluate :is ahciisa. The concentration of d k a l i metal was determilid ( ~ i i8-nil. fractions with a Pcrkin1,:lnirr flame photometer.

Table I.

so.

Typical data for the separation and recovery of sodium and potassium are listed in Table I. Typical elution graphs are shown in the solid curves of Figure 1, which illustrates both syinmetry and good clear separation of the alkalies. The graphs indicate a better seprtratiori t l i a i i has hren attained previously by i o i i exchange ( 2 , 3 ) . Magnesium, which was vwied in these mixtures from 0.5 t o 0.0 me., consistently made its first :ippcarance in the eluate a t n1)out 1100 nil. DISCUSSION

13y combining the equations of hiayer and Tompkins (4)with tlic probability equation, i t can be shown that

= 2

cyrn= C’V c ( T+) 1 7’. -

(1)

I;,

Separation and Recovery of Sodiurn and Potassium

(All weights expressed in milligram) Alkali Chloride HCl Alkali Taken in Molar Ratio Found in Determined Nixture of Xa to K Residue

1

Na K

2

Na

3

K Sa

4

Na

5

Na

K K

IC

0

Na

7

K Xa

8

RESULTS

P

t lir

(if

eluate volume of eluate in milliliters masinium value of .If 7’whc.n .If = .Iffn ”frec. volumo” or intcwtitial volunic~01‘ t l i ( * c+olriiiin, ml. number of‘ theoretical plates “distribution ratio” of the alkali catioii-i.e., ratio of the quantity of alkali cation in the resin in any plate to the quantity of alkali cation in the liquid in this platti at equilibrium

K

h’a

K

351.2 149.3 350.9 149.1 349,R 148.1 58.1 37.1

3

330. ii 15.1 35.3 149.4 352.8 1.7 3.5 149.1

30

3

3 2

0.3 265 0.03

0.4 0.4 0.4 0.3 0.2 0.2 0: 1 0.1

0.8 0.3 0.1 0.5 0.5 0.2 0.2 0.4

Alkali Chloride Found in Residue 350.9 149.4 350.7 149.5 349.4 147.9 58.7 37.7 350.1 15.7 35.9 149.2 352.4 1.7 3.3 149.2

Error -0.3

+0.1 -0.2

$0.4

-0.5 -0.2

+0.2 0.0 -0.5 4-0.6 +0.6 -0.2 -0.4 0.0 -0.2 +O.l

The effect of varying the flow-rate from 0.36 to 0.66 ml. per minute per sq. cm. can be seen from the results given in Table 11. Variations of this order have little effect on the position and size of the fractions.

Table 11.

Effect of Flow Rate

Column. 3.80 8q. cm. X 58 cm. Exchanger. 59.5 grams of colloidal Dowex 50 (oven-dried basis) Elutrient solution. 0.709 molar hydrochloric acid Fractions analyzed at 8-ml. intervals Range of Eluate Volume in Which Cation Was Detected, Quantity Taken, Millimoles Flow Rate, M1. Ka K lIl./Min./Sq. Cm. Ka K 544 t o 688 0.36 384 t o 496 2.00 6.00 392 to 512 552 to 704 0.56 5.99 2.00 392 t o 512 544 t o 696 0.66 6.00 2.00

ANALYTICAL CHEMISTRY

584

was not separated from potassium. Tlic sizes of the respective fractions n-cw~ Column. 3.80 sq, om. X 14 em. diminished with higher temperatrirr. Exchanger. 25 grams of oven-dried Dowex 50, 200 to 270-mesh The higher temperature brings t l i e Elutrient solution. 1.00 molar hydrochloric acid Range of Eluate Volume in Slae of sodium and potassium closer togethrr, Which Cation Was Detected, Fraction, M . , indicating an upper temperature limit for 111. Containing: Flow Rate Temp., hfg Mg these conditions. Rfl./Rlin./Sq. 'cm. C. Na K 24 172-288 399-719 426-715 116 320 289 For a given column, the products of 0 42 0.37 70 139-244 260-494 543-775 105 234 232 U , and the molarity of elutrient acid ~are nearly constant for sodium and potassium; the product of U , and the square of the molarity is nearly constant for magnesium. These relationships can be a useful guide in predicting the relative positions of the elution curve maxima after values of the constants have been evaluated from at least one run. Table 111. Effect of Temperature on Separation

Quantity Taken, .\fillimoles Na K bfg 1.8 1.8

1.7 2.1 1.7 2.0

__

0

~~

~

~

~~

The equilibrium constant for the reaction

HR

+ Sa+ e SaR + H+

is

where A' denotes the mole fraction of the appropriate ion in the resin and the brackets denote molarities in the liquid phase. The distribution coefficient of sodium may be defined as

where LR and' Ls are the millimoles of sodium in the resin and liquid phases, respectively, when m grams of hydrogen resin are equilibrated with v ml. of solution. Substitution in Equation 4 of Figure 2.

L

Elution Curves Showing Cross Contamination

Column 3.80 sq. c m . X 21.1 c m . Exchangkr, 14.7 grams of colloidal Dowex 50 (oven-dried basis) Elutrient solution, 0.407 molar HCI Flow rate, 0.40 m l . / m i n . l s q . om. Elution of 1.80 millimoles of sodium and 1.82 millimoles of potassium Temperature, 28' C. One V u n i t = 70 m l .

I n developing satisfactory conditions for separation, some elutions were obtained in vhich the two alkalies were not completely separated-Le., a certain port,ion of the eluate contained both cations. One such elution is taken for consideration here to show the conformity with the method of Mayer and Tompkins ( 4 )for calculating the point a t which maximum cross contamination occurs. The elution considered is shown in Figure 2. The number of theoretical plates calculated from Equation 2 is 244 and 222 for the sodium and potassium curves, respectively. The mean, 233, was used in the calculations that follow. The volume of liquid in each theoretical plate is designated by v, and n, is the number of L"s that have entered into a plate up to any given time. C:, is 200 ml. for sodium, and 2T2 ml. for potassium. 200 272 Therefore, CN&= - = 2.86 and CK = - = 3.88. Substitu70 70 tion in Equation 8A of LIayer and Tompkins' article gives a value of n = 800 a t the point of maximum cross contamination. As v = V / p = 0.30, the volume a t which the maximum cross contamination occurs is 800 X 0.30 = 240 ml. This is in good agreement with the experimental value of 23i ml. This investigation included the study of separations with Dowex 50 of 200- to 270-mesh. It was found that satisfactory separations, in which the eluate volumes and time required for elutions were comparable with those employing colloidal Dowex 50, could be achieved only a t elevated temperatures. Elutions a t elevated temperatures were performed by housing the column in a thermostatically controlled air bath. Table I11 shows the effect of increasing the temperature from 24" to 70 O C. Table I11 shows that a separation of sodium, potassium, and magnesium was achieved at 70", while a t 24" the magnesium

[sa+] = 2 anti

ir-here Q is tlir capacity of the resin in milliequivalents per gram, yields

Combination of this equation with Equation 5 yiel I-.

If the hydrogen ion concentration is sufficiently high .YHRis almost unity,

SO

that

Mayer and Tompkins (4)hare reported the following relation Tvhich applies to column behavior:

where Jf, is the mass of rr,sin in the column. Therefore, iii)

and analogously,

The capacity of the exchanger used for Figure 1 was determined as 4.77 millimoles per gram of oven-dried resin. Substitution of this value along with 0.700 for the hydrogen ion concentration and Bauman and Gichhorn's ( 1 ) values of 1.20 for K N and ~ 1.50 for K K in Equations 6 and 7 yields C N = ~ 2.66 and CK = 3.26. These are in fair agreement with the values

V O L U M E 22, NO. 4, A P R I L 1 9 5 0

585

2.37 and 3.32, respectively, from the data of Figure 1. The mean values for C obtained from five other elutions are 2.39 for sodium and 3.28 for potassium, with mean deviations of 0.03 and 0.05, respectively. CONCLUSION s

precipitation techniques, indicates possible application 01 this: method to the determination of these alkalies in silicates such as glasses, clays, and feldspars.

A simple, accurate method for the separation and deterinination of sodium and potassium is described. The ion-exchange elution curves of sodium and potassium, under the recommended conditions, follow closely the theoretical curves based on the normal curve of error. The effects of some variables such as flow rate and temperature are shov-n. The relation t)etn-een distribution coeffirient’s and column behavior and the observed and calculated points of maximum cross contamination have been found to check the predictions of other authors. The success of these separations, dorig with t,he absence of tedious

The authors expms:: thrir inc1el)tcdness t o the Office of S:iv;il Research for fin:incial support during the latter part of this i r i vestfigation.

.A Cli\lOWLEI)C, A 1ENT

LITERATURE CITED

(1) Raunian, XT. C., and Eichhom, ,J., J . A m . Chem. Soc., 69, 2k30 (1947). (2) Cohn, IT.E., and Kohn. H. IV..Ihid., 70, 1986 (1938). (3) Kayas, 11.G., C o m p t . rend., 228, 1002 (1949). (4) Mayer, S. IT.,and Tompkins, E. R.. J . Am. Chem. Soc., 69, 2859, 2866 (1947); J . Chcm. Education, 26, 32, 92 (1949).

RECEIVEDAuguEt 19, 1949.

lodometric Determination of Resorcinol lIOB.-IKT 13. WlLL-IKD

AND

A . I,. WOOTEN’, Z’nirersity of Michigun. Ann Arbor, Mich.

This ~olumetriemethod f o r resorcinol is based on selective iodination in a buffered solution. Kelatibely few other phenols are reactive under the conditions chosen. An iodination time of 1 minute at pH 5.0 is used. Fift? millip r a m s of resorcinol ma: he determined with an accuracy of 5 parts per 1000 or 1 nip. with an accuracj of 10 parts per 1000.

ETHODS ior t l w mnc.rodeterIiiiriation of resorcinol have been based principalll- on bromination ( 2 , S ,5-7) or iodination reactions (1, 4 , 6), but :is used do not distinguish resorcinol from any of the other phenols. Methods based on the oxidation of resorrinol by slkaline permanganate are even less selective. The resinous precipitate formed by the reaction of resorcinol and furfural in an acid solution forms the basis of a selective method (9). Phloroglucinol, pyrogallol, cresol, sj-lenol, and orcinol interfere. The method is not adapted to rapid determinations and requires an empirical factor. This determination may be concluded gravimetrically or volumetrically. Resorcinol may be titrated in very dilute solutions with nitrous acid. Phenol does not interfere. The titration is continued until a permanent (30-minute) end point is obtained with starchiodide paper. The product is 2,4dinitroresorcinol. The present procedure is based on the selective iodination of resorcinol in a buffercd solution. Soper and Smith (8) have investigated the iodination of phenol as a function of pH, and conclude that the phenate ion is very much more readily iodinated than the nonionized phenol molecule. It is probable that the present method is possible because of the large ionization constant of resorcinol. Very fe\v of the common phenols interfere with this procedure. m-Dihydroxy phenols, such as phloroglucinol and orcinol, also absorb iodine and thus interfere. o-Dihydroxyphenols, such as catechol, interfere by the formation of exceedingly dark flocculent precipitates which obscure the end point. Catechol may be removed by the lead acetate method of Jones et al. ( 5 ) . Resorcinol samples of 50 mg. may be determined with an error of less than 5 parts per 1000 and samples of 1 mp. n-ith an error of less than 10 parts per 1000. REAGENTS

Iodine. .4 0.1 N iodine solution was prepared by weighing 6.5 grams of iodine and 10 grams of potassium iodide into a beaker and adding 20 ml.of water. The mixturo was st,irred until solution was complete and then was diluted to 1 liter. The iodine solution was standardized against arsenious oxide using starch as an indicator. 1

Present address, Reichhold Cheiniralr;, Inc., I‘erndale, hlich.

Sodium Thiosulfate. A 0.1 sodium thiosulfate solution wa.; prepared by dissolving 24.85 grams of sodium thiosulfate i n 1 liter of freshly boiled water containing 0.1 gram of sodinin carbonate. It was allowed to stand in a stoppered bottle for 2 days before being standardized against pot,assiuni iodate. Starch Solution. A 1% starch solution in distilled water coiltaining 2y0 potassium iodide. Buffer. The buffer was acetic acid-sodium acetate and wa5 1molar in acetate ion. A solution of 120 nil. of acetic acid in about 1700 ml. of water was neutralized with concentrated sotliuin hydroxide until the p H rose to about 1.5. The solution \ \ a \ cooled to room temperature and the neutralization wab (wntinued with dilute sodium hydroxide until the p H rose to 5.0. The buffer is stored in a rubber-stoppered bottle and the pH should be frequently rccliecked. PROCEDURE

-1saniple containing about 0.05 gram of resorcinol is dissolved

i n a little water and 50 1n1. of buffer are added. Fifty milliliters of 0.1 Ai iodine are added from a pipet. After 1 minute the excess iodine is titrated with st,andard 0.1 &Vsodium t.hiosulfatr :iiid starch. . CALCULATIONS

Because the reartion is

+

CsH,(OH)> + 3 I ~ + C ~ s l l I ~ ( O F I ) ~ 3HI the equivalent \wight of re