Separation and Determination of Sodium Oxide In Presence of Lithium Oxide, Phosphorus Pentoxide, and Potassium Oxide H. Il,c’])rrcil)it:ttion of the triph, :ic.ct:rttt ~it~i~~ij)it:ite. HEAGENTS
Zinc Uranyl Acetate. This reagent is essentially that \vliicli Ii:w been described by other investigators. To 200 grams of ura11yl acetat,e dihydrate add 980 ml. of water. I n another IJcaker put 600 grams zinc acetate dihydrate, and add 030 nil. of water. Stir frequently, and heat each separately on thc hot plate for 30 to 45 minutes. Keep covered with watch glasses. .\tltl 120 ml. of 307, acetic acid t.o t,he first beaker (uranium) and GO ml. of 30y0 acetic acid to the second beaker (zinc). 111 4,5 minutes, most of the uranium salt will have dissolved. \Vhcther i t is dissolved or not, heat the zinc solution to hiling. :trid add t o the hot uranium solution (probably cloudy). Stir :ind pour back into the beaker that originally contained thc. zinc acetate. The salts should dissolve completely to give a calcar solution. Transfer to a 2-liter acid bottle vvhen suficicntly c*ool,add 50 mg. of sodium chloride, stir, and allow to stand a t rooin temperature for 24 hours. The clear supernatant liquid only is used as the reagent. Care must he used to prevcmt stirring up the precipitate when pipetting from the bottle. Ethyl Alcohol, 95%, Saturated with Sodium Zinc Uranyl Acetate. Prepare several grams of sodium-zinc uranyl ac*cbtatc by precipitating from a solution of purr sodium chloride with zinc uranyl acetate reagc.nt, collecting on a frittcd-glass filter, and washing with 95% ethvl alcohol. Transfer the precipitate. to a 2-liter ’acid I)ottle, fill with 95% ethyl alcohol, anti shake occasionally over a period of sc~veral days. For U W , filter fresh each timi., 0 1 ’ krcp previousl>.filterrd solution in a tightly stoi)pc~rcd bottle to exclutlc moisture. 1-Butanol-21 to 23Yo Hydrochloric Acid. This reagent, is a solution of dry hydrng.cn chloride cras 111 auhvdrous 1-butanol. Tho hvdrogen cilioritle gas”may come directly from “a cylinder or f m n the action of sulfuric acid 011 sodiiun cliloritl(~. All of the hydrogen chlodc. gas employed in this investigation was preparcvi b>- heating fiodium chloride in a borosilicai(~ glass flask while drops of sulfuric acid entrretl continu:tlly. The hydrogen chloride gas that evolved \vas lcvl through a water condenser anti a calcium rhloritle tower. The dried gas \vas I)uht)lcti i n t o thc I-butanol (Figure 1) to form the reapt’iit. The I-butanol was kept c.001 w i t h icc. ;Iiij. turbidity of the reagent may be sotliuni rhlori~ie:tiid must he removed by centrifuging. Apparatus for Preparation of 1-Butanol-21 to 2 3 7 ~liydroThe driirit>. (2.i”,/4°) of the reagent should lw chloric Acid hctn-rc,ri 0.!)1 and 0.93. Adjust to thcw figui~c~s
575
ANALYTICAL CHEMISTRY
576 Table 1.
Effect of Lithium Oxide on Sodium Oxide Recovery by a Single Precipitation
LizO,
NatO T a k e n , Na?O Found, Rfg. Rlg. 3.00 0.00 2.96 3 . 0 0 0.00 3.00 3.00 0.33 3.04 3.00 0.66 3.08 3.16 1.32 3.00 3.00 3.43 2.65 5.00 3.00 5.42 3.00 8.00 8.44 l5,OO 14.72 3.00 25.00 3.00 21.68 2.65 0.00 b 0.00 2.65 .\mount of precipitate above SazO calculated to I.i,O. So precipitate. Mg.
i .. ~
p r .. .. ..
Liz0 in
,
g
0 02 0.04 0.08 0 22 1.18 2 65 5 72 4 10 b 5
~
~...
I)) ndding anhydrous butanol or by passing more h) drogen chloride gas into the solution. The 1-butanol-21 to 2 3 7 hydrochloric acid is stored in a borosilicate glass bottle in such manner that no water is gained, and very little hJ drochloric acid is lost. The solution n a s satisfactory for 3 weeks, hut after 3 months had lost oiie third of its hydrochloric acid Therefore the reagent should be regarded as suitahle for use for not over 1 month, a t orilinarv room temperature. PROCEDURE
1)ecomposition of the sample niay be accomplished by any ordinary method, including the J. Lawrence Smith and Berzeliuj methods, or by solution in acid without separation of other eleuierits. The sodium oaide is obtained in a clear solution of 5-ml. volumts, in a 50-ml. centrifuge tube (conical bottom, Corning No. 8120) The amount of sodium oxide allowable is 0.0 to 11.0 mg., using thr. suggested volume of sample and reagent. Cool and then precipitate the sodium oxide as triple acetate, with 20 to 25 ml. of the zinc uranyl acetate reagent. Immediately stir thoroughly, particularly in the bottom cone section of the tube. Remove, police, and wash the rod with 1 to 2 ml. of reagent from a small wash bottle. Stir once again a t the end ot 10 minutes. Usually so little precipitate sticks to the rod the decond time that a very slight wash will suffice. Except when &ring, keep tube covered with a rubber cap of appropriate size. Let set for 30 minutes. K i t h cap still on, centrifuge a t a rate and time sufficient to settle thoroughly and t o pack the precipitatr. An International No. 2 centrifuge was satisfactory when operated a t 2000 r. p. m. for 10 minutes (900 X gravity). Let the centrifuge coast to a stop to prevent swirling of the liquid. Some of the precipitate invariably appears to float, so that filtration of the liquid is essential to complete recovery of the sodium oxide. Filter by decanting the liquid through a mediumporosity fritted borosilicate glass filter. Suction is necessary. Immediately add 30 ml. of the saturated 95% ethyl alcohol to the tube and swirl slightly to dissolve the water clinging to the walls. As soon as the first solution has gone through the filter, pour all the ethyl alcohol into the crucible and pull through. Reserve the crucible with its small precipitate for later solution If any salts are present that might be precipitated by thc S,i(;b ethyl alcohol of which aluminum is an example, preliminary decantation with 10 ml. of zinc uranyl acetate reagent may br made. This must be followed by the usual 30-ml. wash with 9 5 5 ethyl alcohol. Pour into the centrifuge tube 12 to 30 ml. of 1-butanol-21 to 2Jyo hydrochloric acid. The amount to add is judged by the volume of the precipitate. Immediately stir to solution of the original triple acetate precipitate, cover the tube TT ith the dry rubber cap, and place a small beaker over tube and cap. To exclude moisture, the tube should then be placed in a beaker and covered n i t h tin foil, aluminum foil, or paper. A precipitate of sodium chloride will form simultaneously with solution of the original triple acetate. Although not proved necessary, setting (It the tubes and beaker in a refrigerator is desirable to reduce the solubility of sodium chloride. Let sit 1 hour. To gather the sodium chloride in the bottom of the tube, rotate in a centrifuge until a clear solution is obtained (2000 r.p.m. for 20 minutes proved satisfactory). Decant the liquid into a dry beaker, and let the tube drain for a few seconds. Examine the liquid in the beaker for clearness and, if clear, discard. Dissolve the small precipitate in the fritted-glass crucible with 10 drops of concentrated hydrochloric acid; then filter by suction into a small test tube placed under the funnel. Wash with 5 ml.
.
of witer, also catching this in the test tube. Mix and pour the contents of the t,est tube into the centrifuge tube holding the bulk of the sodium chloride. Set the centrifuge tube in a 100ml. beaker containing 100 ml. of hot water, and boil on the hot plate for 10 minutes. Transfer the liquid from the centrifuge tube to a 30-ml. beaker. Evaporate the liquid to 5-ml. volume and cool. For small amounts of sodium oxide, evaporate to 2 to 3 ml. Prreipitation of sodium oxide should be at or near the same temperature a t which the reagent solution was saturated Lvith the triple salt. A corollary is that the zinc uranyl acetat,e reagent should not be allowed to become excessively cool, as by standing overnight in an unheated laboratory. Precipitate the sodium oxide as triple acetate by adding 20 to 25 ml. of zinc uranyl acetate reagent. Stir immediately. Cover a i t h a watch glass and let, set 30 minutes, stirring twice in the memtimr. Filter by suct,ion through a Jveighed, medium-porosity, fritted-glass crucible (borosilicate glass with nominal maximum pore size of 14 microns). Stir the precipitate and pour the entire contents of the beaker into the crucible. Immediately wash down the sides of the beaker with a small quantity of alcoholic wash solution. As soon as all the liquid in the filter has been pulled through, wash the sides, and precipitate once with the alcoholic solution. Police the beaker, and transfer all the precipitate to the crucible. Again wash down t,he sides of the crucible. Stop the suction and thoroughly stir the precipitat'e from the bottom of the crucible with a stream of the alcoholic wash. Suck all the liquid through, and wash twice from the inside top all the way around. Suck away excess alcohol. Remove the crucible from the suction and Tvipe the crucible, including the top and boJtom flange, with a damp cloth. Dry for 20 minutes a t 80 C., desiccate until cool, and weigh. The sodium oxide factor is 0.02015. DATA AND DISCUSSION
Lithium Oxide. The extent of interference from lithium oxide by one precipitation only is s h o m in Table I. Here the sodiurii oxide is precipitated once only n-ith zinc uranyl acetate in the presence of various amounts of lithium oxide, and is weighed as triple acetate without the separat,ion with l-hutanol21 to 23% hydrochloric acid. Lithium oxide in amounts oirer 4 mg. cannot be eliminated by simple reprecipitation of the triple acetate. Less than 4 mg. of lithium oxide probably could be eliminated successfully by reprecipitation. In lithium minerals, salts, and unknown ,samples, lion-ever, the lithiuni oxide is or may be too high for successful analysis by such method.
Table 11.
Effect of Lithium Oxide on Sodium Oxide Recovery, Using Directed Procedure
IizO, Mg.
1.00 3.00 5 00 10.00 15.00 20.00 25.00 30.00 30.00
50,oo
100.00 100.00
Na?O Taken, Mg. 3.00 3.00 3 00 3.00 3.00 3.00 3.00 0.30 9.00 3.00 3.00 3.00
NazO Found, Rig. 3.00 2.96 2 - 94 .. 2.98 3.10 2.98 3.07 0.27 9.14 3.03 3.04 3.01
To determine the upper limit of lithium oxide allowable when the analysis is made as directed in procedure, a series with constant sodium oxide and variable lithium oxide was prepared and analyzed. The data are summarized in Table 11. Lithium oxide was added as lithium carbonate. When the lithium content was high, a correction had to be made for the sodium oxide content of the lithium carbonate. As the data show, recovery of 3.00 mg. of sodium oxide, from amounts of lithium oxide up to 200 mg., was complete and accurate within experimental error. The data of Table 111 present a reverse view of the same problem. Here the data are presented when lithium oxide is held constant a t 15 mg., an amount to be expected from a 200-mg.
V O L U M E 22, NO. 4, A P R I L 1 9 5 0
571
sample of lithium-containing mineral, The sodium oxide was varied widely. Again, within experimental error, the recovery of sodium oxide was complete in all samples. Eleven to 12 nig. of sodium oxide are about the upper limit of sodium oxide allow able with 20 ml. of zinc uranyl acetate reagent. To determine the upper limit of recovery of sodium oxide in the presence of large quantities of lithium oxide, a series was prepared with lithium oxide held constant a t 100 mg. and the sodium oside varied from 0.3 to 10.0 nig. The data are in Table ITr. llecovery of sodium oxide was complete with 0.30 and 1.00 nig. of sodiuin oxide but T Y ~ Sslightly low Kith 7.00 and 10.00 mg. L-ntloubtedly the proper approach for correct results in such an unusual ronihination would be merely to increase the amourit of reagent a n d sample volume used, t=vm t o as murh as 50 ml. of reagent, arid 10 ml. of sample volunlt..
Table 111. Effect of Constant Lithium Oxideon Recoiery of Wide Limits of Sodium Oxide, Using Directed Procedure LizO, Blg.
S a 2 0 Taken.
hlg.
15.0 15.0
0 0 0 1 3
15.0 15.0 1.5.0 15.0 15.0 13.0 15.0
TS
7 9 II
Xaz0 Found,
hlg.
LO
0.19 0.37 0.76 I . 54
40 80 60 00 00 00 011 00
z.10 0.01 7.02 9.01 11.12
The data of Table I indicate that an error of 0.1 nig. of sodium xide is found when 1.4 mg. of lithium oxide are present with 3.0 nig. of sodium oxide. To determine if a simple increase in volume of both reagent and sample would extend the threshold of lithium oside interference, two samples of 3.00 mg. of sodium oxide and cj.ti nig. of lithium oxide in 25 ml. of Yater were precipitated with 100 ml. of reagent. The results are tabulated in Table 5’. Evi~dc~iitly, increased volumes will tolerate more lithium oside. Howvbw,the larger volumes will make the sodium determination iiiricli less sensitive and will assume previous knowledge of the :iniount of litbium oxide present. In short, preventing lithium ositle interference by increased volumes alone is applicable only i n special samples and has not the virtually universal application ~)i’t,!ie outlined procedure. Phosphorus Pentoxide. To test the effectiveness of the direct,ed procedure on phosphate interference, sodium oxide was precipitated in the presence of 100 mg. of potassium dihydrogen ptiospliate. Even to the eye, it was evident that, in the first odium oxide precipitate with zinc uranyl acetate, the phosphate iittti formed a huge precipitate. Attempts to filter this mass led t o plugged crucibles which effectively stopped filtration. The resn!ts obtained by use of the directed procedure on such a disisouraging mass of precipitate are listed in Table 1-1. Sodium oside was recovered accurately and iitterfererire from phosphorus pentoside was eliminated.
Tahle VI. Effect of Phosphorus Pentoxide on Recovery of Sodium Oxide, Using Directed Procedure Pzos (as KHzPOI), Mg. 52 2 52 2 52 2 62.2
Kz0, hlg. 34.6 34.6 34.6 34.6
S a 2 0 Taken, hlg. 1.00 3.00 3.00 i.00
KazO Found, 1Ig.
0.94 2.93 2.99 6.95
Potassium Oxide. Potassium is not effectively eliminated by the 1-butanol-21 to 23% hydrochloric acid reagent, because potassium chloride also is insoluble. Hov-ever, double precipitation of the sodium oxide as triple acetate did eliminate the potassium oxide interference. The fact is sho5v-n in Table VI. The presence of 34.6 mg. of potassium oxide in each sample caused no difficulties. To explore further the upper limits of allowable potassium oxide, sodium oxide was precipitated and carried through the directed procedure in the presence of 200 mg. of potassium chloride (126.3 mg. of potassium oxide). The results appear in Table VII. The large quantity of potassium oxide did not cause interference. After addition of the zinc uranyl acetate reagent, the solution should be stirred immediately. General. Application of the method is easy; no particular manipulative skill is required. Results may be ohtained in one day.
Table VII.
Effect of Potassium Oxide on Sodium Oxide Recovery, CTsing Directed Procedure
K20, Llg.
S a 2 0 Taken,
3Ig.
S a 2 0 Found,
Mg.
126.3 126.3
1.00 4.00
1.02 4.04
Cse of a centrifuge is mandatory. With the centrifuge tubes, moisture may be successfully excluded, the precipitate may be handled without dissolving and redrying, and the sodium chloride, which separates as a very fine grain, may be quantitatively collected with ease. CONCLUSION
A procedure is presented for eliminating lithium oxide, phosphorus pentoxide, and potassium oxide interference in the zinc uranyl acetate method for the determination of sodium. The procedure is rapid and easily applied; it should give wider application to the uranyl acetate method for determining sodium oxide. LITERATURE CITED
Adams, ,J. I., Benedetti-Pichler, A. A., and Bryant, J. T., Alikrochemie, 26, 29-35 (1939).
Barber, H. H., and Kolthoff, I. lf., J . Am. Chem. SOC.,51, Table I\’. Effect of Large Quantities of Lithium Oxide on Recovery of Sodium Oxide, Using Directed Procedure LizO, hlg. 100 100 100 100
Ka20 Taken, hlg. n 30
S a 9 0 Found,
hI g
n
i.00
RI
1.04 6.88 9.70
7.00 10.00
____ __ Table V. Effect of Simple Increase i n Volume of Reagent and Sample, with Single Precipitation as Triple Acetate ~
LipO, hlg. 6.60 6.60
Sample Volume, hl 1 25 25
Sa20 Taken, blg. 3.00 3.00
Reagent Volume, nr 1 100 100
Sat0 Found, hk 2.96 3.08
3233-7 (1929).
Caley, E. R., and Foulk, C. W.,Ibid., 51, 1664-74 (1929). Collins, T. T., Jr., “Determination of Sodium with Uranyl Acetate Reagents,” Document 1798, American Documentation Institute, 1719 N St., ?i.T7’., Washington, D. C. Koenig, E. W.. J . Am. Ceram. SOC.,22, 24-31 (1939). Kolthoff, I. M., Chem. Weekblad, 26,294-8 (1929). Kolthoff, I. RI., 2.anal. Chem., 70, 397-400 (1927). Nydahl, F., Ann. Agr. Coll. Sweden, 6,37-57 (1937). Smith, G. F., and Ross, J. F., “Perchloric Acid,” T’ol. I , 4th ed., pp. 54-61, Columbus, Ohio, G. Frederick Smith Chemical Go., 1940. Sumuleanu, C., and Botezatu, M., Mikrochemie, 21, 65-74 (1936).
RECEIVED Optoher 11, 1949. Work by Bureau of ;\lines in cooperation with Tennessee Valley Authority.
