neck to prevent spray loss of the solution. Adjust the oxygen flow to 5 liters per minute and allow 5 minutes to flush out carbon dioxide in the solution. Weigh a 1-gram sample and transfer to a prefired crucible. (These may be prefired in advance and stored in a desiccator.) Add tin to ensure good combustion. One-half gram is usually sufficient and many samples can be burned with no tin. Add 5 drops of bromocresol purple to the absorbing solution and adjust to colorless end point with carbonatefree alkali or acid. Flush the fritted glass dispersion tube by momentarily opening the gate of the furnace to interrupt oxygen flow and then closing. Readjust the end point. Repeat until end point is constant and note buret reading. Place the covered crucible in position and ignite the sample. Burn for 3 minutes, maintaining plate current as high as possible without overloading the induction furnace. Titrate with alkali (1 ml. = 0.0001 gram of sulfur) to colorless end point. Experiments
with samples as high as O.35y0 sulfur (NBS 133) indicate that no losses occur even if no titrant is added until the end of the combustion period. Repeat the flushing of the fritted glass dispersion tube and titrating until the end point is constant. Calculate per cent of sulfur with a correction factor of 1.09. Modifications for Low Sulfur Samples. Follow the standard procedure, using the same absorbing solution for a series of determinations. As many as 10 have been run by the authors before the absorbing solution became fogged and obscured the end point. Remove the modified apparatus and rinse with the solution from the absorption vessel. Titrate to the end point. Repeat the rinsing and titration until the end point is constant. Add a proportionate volume of alkali used for titrating the rinse to each of the individual titers. Calculate the per cent of sulfur with a correction factor of 1.09, after making suitable blank corrections for crucible and tin flux.
ACKNOWLEDGMENT
The authors are indebted to B. A. Thompson, General Electric General Engineering Laboratory, Schenectady, N. Y., for performing all of the radioassay work described. LITERATURE CITED
(1) Fossum, J. H., Markunas, P. C., Riddick, J. A., ANAL. CHEM.23, 491 (1951). (2) Fryxell R. E., Zbid., 30, 273 (1958). (3) Hale, H., Jr., Muehlberg, W. F., IND.ESG. CHEM.. ASAL. ED. 8. 317
6.
(1936). (4) Holler, -4. c., ~ A L CHEM. . 28, 1359 (1956). (5) Smith, T. B., Backhouse, A., Woodward, P., J . d p p l . Cheni. (London) 4, 75 (1954). RECEIVEDfor reviem- May 26, 1958. ilccepted October 6, 1958. Pittsburgh Conference on Bnalytical Chemistry and Applied Spectroscopy, March 1958.
Nonaqueous Determination of Acetylenic Hydrogen LUCIEN BARNES, Jr. Research laboratories, Air Reducfion
Co., Inc.,
b A combination of silver perchlorate, an amine titrant, and screened thymol blue indicator enables the direct determination of acetylenic hydrogen in nonaqueous solvents. The method is based upon the liberation of hydrogen ions when silver perchlorate reacts with monosubstituted acetylenic compounds. The acid liberated due to acetylide formation is titrated with a methanolic standard solution of tris(hydroxymethy1)aminomethane to the color change of a screened thymol blue indicator, permitting direct determinations of acetylenic hydrogen in waterimmiscible solvents and easily hydrolyzable esters. The determination may be performed in the presence of appreciable amounts of certain weak organic acids without prior neutralization. Strong acids and bases may be preneutralized so as not to interfere. Intermediate strength amines and acids interfere because of buffer action but there is no interference from halogens or aldehydes. Acetylide precipitation rarely occurs, which facilitates detection of the visual end point.
M
for the determination of acetylenic hydrogen have been reported (1, 4, 6). This nonaqueous procedure was developed for the many
Murray Hill, N. J.
mater-insoluble samples routinely received for analysis in these laboratories. Such samples range from pure acetylenic compounds to trace amounts (less than O.loJ,) contained in various waterimmiscible solvents and easily hydrolyzable esters. Because the strong bases normally employed as titrants readily saponify many esters, a weak amine, tris(hydroxymethy1)aminomethane (THAM), was used as the titrant Fyith a low pH indicator, thymol blue. Miocque and Gautier (5) published a silver nitrate procedure in which the complexing and basic properties of pyridine are used to excellent advantage. Alcoholic sodium hydroxide is used as titrant and thymolphthalein as indicator. The two methods approach the problem from opposite ends of the pH scale, and complement each other, as neither procedure is universally applicable. Silver perchlorate is appreciably soluble in many organic solvents and reacts with monosubstituted acetylenes in polar media according to R-CEC-H R-C=C--ilg.
+ 2dgC104 AgClOi
+
+ HClOi
(1)
ETHODS
The liberated perchloric acid may be quantitatively titrated with tris(hydroxymethy1)aminoniethane to the
low pH color change of a screened thymol blue indicator. REAGENTS
Silver Perchlorate, lilf. Dissolve 104 grams of anhydrous silver perchlorate (G. Frederick Smith Chemical Co.) in anhydrous methanol and dilute to 500 ml. Store in a polyethylene bottle. Xo explosions have been experienced with solutions prepared and stored in this manner. Barium Perchlorate, anhydrous (G. Frederick Smith Chemical Co.). Standard 0.1N Tris(hydroxymethy1)aminomethane (THAhI). With mechanical stirring, dissolve 12.15 grams of tris(hydroxymethy1)aniinomethane (Fisher Scientific Co.) in methanol and dilute to 1 liter. Filter any remaining insolubles. To standardize, dilute 40 ml. of this solution with 200 nil. of water and titrate \Tith aqueous acid to the methyl purple end point. Screened Thymol Blue Indicator Solution. Dissolve 0.100 gram of thymol blue and 0.025 gram of alphazurine in 100 ml. of methanol. Store in a brown bottle. Prepare fresh every week, as it decomposes on standing. Alphazurine is obtainable from the General Laboratory Supply Co., P. 0. Box 2607, Paterson, N. J., and is known as Alphazurine A or Patent Blue il, Catalog No. NB717. Standard 0.1N Perchloric Acid. Dissolve 8.5 ml. of 72% perchloric acid in VOL. 31, NO. 3, MARCH 1959
405
Table 1.
Application of Method to Various Compounds
Compound0 Alcohols 2-Propyn-1-01 2-Methyl-3-butyn-2-01 3-Phenyl-1-butyn-3-01 3-Methyl-1-pentyn-3-01 3-Methyl-1-hexyn-3-01 3,5-Dimethyl-l-hexyn-3-01 2-Et hyl- 1-heptyn-3-01 3-Ethyl-5-methyl-1-heptyn-3-01 3,6-Dimethyl-3-hydroxy-l-heptyne CEthyl-l-~ctyn-2-01 3-Methyl-1-nonyn-3-01 Hydrocarbons Acetylene (dissolved in acetone) &Methyl-1-butene-3-yne 1-Pentyne 1-Hexyne 1,f.i-Heptadiyne
Purity, yo Nona ueous Ag&O,
98.1 99.0 99.7 98.8 100.3 98.1 100.1 98.0 99.5 100.3
loo. 1
1.19 99.7, 99.1 84.9 97.5, 97.9 92.6, 92.7
Aqueous
AgNOa*
98.0 98.9 99.1 99.0
...
98.6 99.5 98.3 100,o 99.7
...
1.18 98.5 85.3 98.6 93.3
Miscellaneous 3-Met hyl- 1-pentyn-3-01dissolved in l,l,l-trichloroethane 0.057, 0.055, 0.057c Acetylene dissolved in vinyl . acetate monomer 0.023, 0.023 ... Technical grade samples of reasonable purity; no attempts made at further pur& Cation. Aqueous silver nitrate procedure ( 1 ) . 50-ml. Sam le extracted with aqueous silver nitrate prior to titration with standard alkali. Formuyated by manufacturer to contain 0.05 to 0.07% 3-methyl-1- entyn-3-01. 0.025% acetylene added to 25.0 ml. of monomeric vinyl acetate. Entire sample titrated directly. Aqueous silver nitrate value not obtained aa sample is not amenable to analysis by an aqueous procedure.
1 liter of methanol. Standardize against 0.1N tris(hydroxymethy1)amino methane using screened thymol blue as indicator.
acetylide precipitation, with the exception of acetylene and l,g-heptadiyne, and it may be eliminated in these instances by increasing the concentration of silver perchlorate three- to fourfold.
PROCEDURE
Acids Present. If the sample contains a strong acid (indicator purple), neutralize with 0.1N tris(hydroxymethy1)aminomethane t o the green of the indicator prior to the addition of silver perchlorate and titration with tris(hydroxymethy1)aminomethane. The determination may also be performed directly in the presence of certain weak acids without neutralization when the Ka of the acid is in the range of loF6 or smaller (weaker acids) and if a soluble silver salt of the acid is formed. In the presence of these acids, increase the amount of solvent (methanol) to 50 ml. to minimize silver salt precipitation which causes high results. In addition, increase the amount of indicator to 15 to 20 drops. After addition of neutral silver perchlorate, titrate with 0.1N tris(hydroxymethy1)aminomethane. The end point becomes more difficult to detect as the amount of weak acid present is increased because of the buffering action caused by the weak acid which becomes more pronounced as the equivalence point is approached. To avoid high results, the end point should be regarded as that point during the titration a t which the purple tinge of the indicator has just disappeared and the solution is pale green.
Acids and Basic Substances Not Present. Add 3 drops of screened thymol blue indicator t o 10 ml. of 1 M methanolic silver perchlorate contained in a 50-ml. beaker. Neutralize any free perchloric acid with 0 . 1 , ~tris(hydroxymethy1)aminomethane to the green of the indicator. Accurately weigh a sample containing 1 to 3 meq. of acetylenic hydrogen into a 250-ml. Erlenmeyer flask containing 5 t o 10 ml. of neutral methanol and 5 to 10 drops of indicator. Pour the neutralized silver perchlorate into the sample flask and titrate the liberated perchloric acid to a permanent green end point with 0.1N tris(hydroxymethyl) aminomethane. If the visual end point cannot be obtained because of the color of the sample, a potentiometric titration may be performed. A glass indicating electrode and a Beckman double junction reference electrode (Catalog No. H3960) may be used. Remove the sodium nitrate solution from the double junction electrode and replace with 3M sodium perchlorate in a 2 to 1 methanolwater solution. The application of the procedure to various neutral acetylenic compounds is illustrated in Table I. There was no
.406
ANALYTICAL CHEMISTRY
Table I1 shows the results of the
analysis of a methanolic stock solution of 3-methyl-1-pentyn-3-01 in the presence of various weak acids which were neutral to the indicator. Recoveries exceed 100% after a practical maximum amount of weak acid (about 0.5 gram) has been reached. Bases Present. If strong bases, such as the alkali hydroxides, are present (indicator blue), neutralize with 0.1N perchloric acid to the purple of the indicator before adding silver perchlorate. If amines with K B values of 1 X 10-6 or larger (stronger bases) are present, the indicator will remain green. These should also be neutralized with 0.1N perchloric acid in methanol before addition of silver perchlorate and subsequent titration with 0.1N tris(hydroxymethy1)aminomethane. Table I11 shows the results obtained for the analysis of a methanolic stock solution of 3-methyl-1-pentyn-3-01 in the presence of various moderately strong amines and one which is extremely weak. DISCUSSION
Reagents. Silver perchlorate, in contrast to silver nitrate, possesses a remarkably high solubility in many organic solvents and the perchloric acid liberated upon acetylide formation is the strongest of the mineral acids. Tris (hydroxyniethy1)aminom e t h a n e (9) is a stable, nonhygroscopic, crystallinesolid available as a primary standard. It does not absorb atmospheric carbon dioxide, and methanolic standard solutions have proved t o be exceptionally Tris(hydroxymethy1)aminostable. methane is a moderately weak base (Ks= 1.20 X 10-6) which complexes silver ions, and it may be added in excess of the perchloric acid present without precipitating silver, saponifying esters, or causing silver to be reduced by aldehydes, if present. Tris(hydroxymethy1)aminomethane is, however, a strong enough base in nonaqueous media to produce a relatively sharp end point under the conditions prescribed. Perchloric acid in methanol is used as a standard acid to neutralize basic impurities, if present. It does not form insoluble precipitates with silver as do hydrochloric and sulfuric acids. Thymol blue indicator exhibits three colors in both aqueous and certain nonaqueous solutions-Le., pH in water 1.2 t o 2.8, red to yellow, pH 2.8 to 8.0, yellow, pH 8.0 t o 9.6, yellow to blue. I n nonaqueous media, perchloric acid may be titrated quantitatively with
tris(hydroxymethy1)aminomethane from the red to yellow change of thymol blue. Under these conditions, moderate amounts of many of the weak acids, such as acetic acid, are neutral to the
indicator (yellow) and hencenot titrated. Homver, when appreciable amounts of neutral salts such as silver perchlorate are present, change from red to yellow is not perceptibly sharp. The addition of a small amount of the blue dye alphazurine, as a screening agent, to thymol blue makes the visual end point much more distinct. The red of thymol blue is thus changed to purple, and the yellow to green. The change from purple t o green is readily detected. Strong bases such as sodium hydroxide cause the screened indicator to turn blue, weaker bases such as amines and weak acids such as acetic produce a green. Strong and intermediate strength acids produce a purple. Solvents. I n general, ionizing or polar solvents should be used if possible because the reaction between silver perchlorate and the acetylenic hydrogen function is ionic. Nonpolar solvents such as benzene and carbon tetrachloride may be used with methanol, but the reaction towards the end point is somewhat sluggish. Once obtained, however, the end point is permanent. Methanol has proved to be the best general solvent. Successful determinations (99% or better recoveries) have been performed in methanol. ethyl alcohol, acetone, dioxane, benzene, chlorobenzene, chloroform, and l,l,l-trichloroethane. Ten milliliters of methanol were first added to all nonpolar solvents to ensure proper functioning of the indicator. Effect of Sulfuric Acid. Sulfuric acid is anomalous, in that in acidic solvents such as methanol, only one hydrogen is titrated by tris(hydroxymethy1)aminomethane. This is in accordance with potentiometric findings of Critchfield and Johnson ( 2 ) . When silver perchlorate is added, perchloric acid is liberated in proportion to the acetylenic hydrogen present plus an amount equal to the second hydrogen of the sulfuric acid which was initially not neutralized by tris(hydrosymethy1)aminomethane. Equations which are believed t o represent the over-all stoichiometry (not necessarily the mechanism) of this reaction are given below:
+
(HOCHZ)3CNHz HzSOc * (HOCHl)&SHz. H$Oa ( 2 ) (H0CHz)sCNHz HzSOc (X~)SgClO + ~ Ag2S04 (HOCHz)3CNH2 HClOd HC104 (3)
+
+
+
The free perchloric acid liberated in Equation 3 requires an amount of tris(hydroxymethy1)aminomethane for neutralization equal to the amount required for Equation 2. Thus if the titration required for Equation 2 is subtracted from that required for Equation 3, any additional amount represents that due to acetylenic hydrogen. However, sulfuric acid can
II.
Table
Analysis of Stock Solution of 3-Methyl-1 -pentyn-3-ol in Presence of Various W e a k Acids
Acid Formic
KA 1.77 x 10-4
Acetic
1.8 X
Propionic
1.34 X 10-6 2
Butyric
1.5 X 1 x 10-6
Stearic
Q
0.60
1.20 0.118 Succinic K , 6 . 4 X lo-’ K , = 2 . 7 X 10-8 0.236 0.474 Benzoic 6.3 X 0.5 Cinnamic 3 . 7 x 10-6 0.5 1.0 No end point obtainable due to silver salt precipitation.
Table
a
x 10-6
Crotonic
Acid Added, Recovery of 3-Methyl-l1-pentyn-3-01, yo Gram 0.27 102.0 111.3 0.55 99.6 0.02 99.6 0.05 99.7 0.10 0.25 99.7 101.1, 99.7 0.50 103.4, 102.9 1.0 99.8 0.5 1.0 102.6 101.6 0.5 103.2 1.0 0.48 99.3 0.96 101.5
111.
100.2
102.7 99.8 100.1 0
101.0 100.4 104.1
Analysis of Stock Solution of 3-Methyl-1 -pentyn-3-ol in Presence Various Amines
of
Amine Recovery Added, of %MethylAmine KB Gram l-pentyn-3-ol, % Diethylamine 1.26 X 10-3 0.14 99.6 Butylamine 4.1 X lo-‘ 0.15 100.0 Monoethanolamine 2.77 X 10-6 0.20 98.6 Ethylenediamine 8 . 5 X 10-6 0.10 100.0 N-Phenyl-%naphthylaminea about 10-13 0.40 101.5 0.20 99.8 Sample neutral to indicator, and did not require neutralization.
esterify methanol very rapidly resulting in a loss of hydrogen ions according to: H,SO,
+ CHIOH
+
H2O
+ CHaOS020H
(4’1
Therefore, the titrations required for Equations 2 and 3 must be performed promptly or results will be low. To guard against obtaining low results when sulfuric acid is known or suspected to be present perform the analysis by an alternative method ( 4 ) or an aqueous silver nitrate procedure ( I ) , after first neutralizing the sulfuric acid present with alkali. If the proposed nonaqueous procedure is preferred for reasons which preclude the use of other methods, precipitate the sulfuric acid by the addition of excess neutral barium perchlorate dissolved in methanol. Without filtering, neutralize the liberated perchloric acid with tris(hydroxymethyl)aminomethane, add neutral silver perchlorate, and titrate with
tris(hydroxymethy1)aminomethane. Interfering Substances. WATER. Water interferes with the analysis, causing low results. For practical purposes, however, 0.5 to 1.0 gram of
Table IV. Effect of Water on Nonaqueous Determination of 3-Methyl1 -pentyn-3-ol
Water Added, Grams 0.0
0.5 1 .o 2.0
Recovery of 3-Methyl-lpentyl-3-01, 7 0 100.0
99.4, 99.8 98.7, 98.9 94.9
water may be present without seriously affecting the results as Table IV illustrates. SOLVENTS CONTAINIIiG SULFUR AIiD
NITROGEN.Certain solvents such as dimethyl sulfoxide, N,N-dimethylforrnamide, acetonitrile, and probably others, apparently by virtue of their sulfur or nitrogen groups, can complex silver ions in nonaqueous media. If sufficient complexing occurs, results are low. Under the prescribed conditions (2 grams of silver perchlorate) the maximum amounts of these substances present which still permit quantitative recoveries are 1 gram of acetonitrile, 2.5 of N,N-dimethylformamide, and 5.0 VOL. 31, NO. 3, MARCH 1959
407
of dimethylsulfoxidc. By increasing the amount of silver perchlorate, the amount of interfering solvent m-hich can be tolerated may be increased. INTERMEDIATE STRENGTH ACIDSAND BASES. Acids with K a values of to 10-I are acidic to screened thymol blue but are not quantitatively titrated by tris(hydroxymethy1)aminomethane. The solution is thus strongly buffered and the end point cannot be obtained. Aromatic amines, with Kg values of 10-9 to 10-12 are only partially neutralized by perchloric acid in methanol. These interfere and cause erroneous results. The effect of amines with
Kg values in the range of to 10-8 has not been investigated. If the amine is so weak that it is neutral to the indicator (KB = lO-I3), results are quantitative.
this manuscript and t o the Air Reduction Co., Inc., for permission to publish. LITERATURE CITED
(1) Barnes,
Lucien, Molinini, L. J., CHEW 27, 1025-7 (1955). (2) CritcMeld, F. E., Johnson, J. B., ANAL.
ACKNOWLEDGMENT
The author is indebted to C. A. Wamser for his aid in the selection of tris-(hydroxymethy1)aminomethane as the amine titrant and alphazurine as the screening agent for the thymol blue indicator. Acknowledgment is also extended to A. H. Taylor and C. A. Kamser for their critical examination of
Zbid., 26, 1803-6 (1954). (3) Fossum, J. H., Markunas, P. C., Riddick, J. A., Ibid., 23, 491-3 (1951). (4) Koulkes, ilfichel, Marszak, Israel, Bull. SOC. chim. France 1952, 556-7.
(5) Miocque, Marcel, Gautier, J. A4., Zbid.. 1958. 476-9. (6) Siggia, S'idney, AXAL. CHEhI. 28,
1481-3 (1956).
RECEIVEDfor review June 13, 1958. Accepted October 13, 1958.
WESLEY W. WENDLANDT Department of Chemistry and Chemical Engineering, Texas Technological College,
,The thermal decomposition of the hydrated oxalates of terbium, dysprosium, thulium, ytterbium, and Iutetium was studied on the thermobalance. The oxalates were precipitated from solution as the 10- or 5-hydrates and when subjected to pyrolysis, began to lose weight in the 45" to 60" C. temperature range. After various stages of intermediate hydrate formation and decomposition, the metal oxide levels were obtained in the 71 5 " to 745' C. temperature range.
T
thermal decomposition of the scandium, yttrium, and the lighter rare earth metal oxalates has been described (2, 6). This report is concerned with the thermal decomposition of the heavier metal oxalates-dysprosium, terbium, ytterbium, thulium, and lutetium. Except for degrees of hydration, little is known about the chemistry of the heavier rare earth metal oxalates. According t o Vickery (s), even the optimum conditions for their preparation have never been defined. T o circumHE
Table
Compound Terbium oxalate," curve A Dysprosium oxalate, curve B
Formula Tb,( Cz04)~. lOHzO
Lubbock, Tex.
I.
Wt.First Loss,
c.
45
vent this question, the metal oxalates were prepared by homogeneous precipitation with methyl oxalate. EXPERIMENTAL
Reagents. The source of the dysprosium, terbium, and ytterbium oxides was the St. Eloi Corp., Newtown, Ohio. The thulium and lutetium oxides were obtained from the Lindsay Chemical Co., West Chicago, Ill. The purity of these compounds was listed as 99.9% by the supplier. The methyl oxalate was prepared as
Thermal Decomposition Data
Breaks in Curve C. Formula 140b Tbz(C204)3.5HzO 265 Tbz(Cz04)3. lHzO O
Anhydrous Oxalate C. Formula O
435
Loss of CO and C o t ' C. Formula
Tbz(CzO4)3
Horizontal Wt. Level c. Formula 220-295 730 TmZ03 T ~ z ( C Z O 5~H ) I~. 0 55 195-335 Tmz(C204)3. 2H20 Thulium oxalate, curve C 730 Ybz03 Yb?(Cz04)~. 5HzO 60 175-325 Ybz(C20r)a.2HzO Ytterbium oxalate, curve D 715 LUzo3 Lu2(Cz04)3. 6Hz0 55 190-315 Luz(CzO4)r.2Hz0 Lutetium oxalate, curve E u Thermal decomposition curve similar to that found for gadolinium oxalate ( 5 ) . Because horizontal weight levels were not obtained, amount of hydrate water may he fortuitous. If a 1-hydrate is actually formed, terbium oxalate is the only rare earth that shows such behavior. c This amount of hydrate water may be fortuitous.
408
0
ANALYTICAL CHEMISTRY
