combustion-gas chromatographic technique. Carbon and hydrogen were determined on benzoic acid, anthracene, and acetanilide by the gas chromatographic measurement of carbon dioxide and acetylene, which are obtained as combustion products from sample combustions (Table 111). Although the smallest sample size used in this work was 1.5 mg. i t should be possible to extend the determination of carbon and hydrogen to the decimilligram level b y combining the high sensitivity of gas chromatography with sample ~ ~ e i g h i n gon a quartz fiber balance. This technique eliminates the weighing of the absorption tubes, which are too heavy to be weighed on a quartz fiber balance with the accuracy required for very small differences in m i g h t (g). The combustion-gas chromatographic method described for determination of carbon and hydrogen represents the first phase of this technique. Recent work indicates that nitrogen can be determined simultaneously with carbon and hydrogen by means of gas chromatography. A report of this work will be submitted for publication in the near future.
Table I.
Calibration Data for Carbon-Hydrogen Obtained by Combustion of Benzoic Acid Peak Areas, Sq. Inches Corresponding Wt., Mg. Taken, Mg. C H c02 C2Hz 1.02 0.073 1.05 1.40 1.48 2.38 0.172 2.40 2.98 3.46 4.02 0.287 4.10 4.80 5.84
Table II.
Carbon-Hydrogen Determinations in Helium Atmosphere by Conventional Gravimetric Procedure in Absence of Oxygen
96 Theory Compound C H Naohthalene 93.68 6.29 Di;henylamine 8 5 . IS 6.55 14.38 Cyclohexane 85.62 6.37 Pyridine 75.91 Anthraquinone 80.76 3.87 6.48 Sucrosea 42.10 2.91 Bromoacetamido AQa 55,83 a Premixed with copper oxide in platinum sample boat.
LITERATURE CITED
( 1 ) Kirsten, W.J., Chim. anal. 40, 25368 (1958). ( 2 ) Kuck, J. A., Altieri, P. L., Towne, A. K., dfikrochim. Acta 1954, 1-16.
H-
93.55 85.45 85.24 76.14 81.01 42.46
6.16 6.34 14.31 6.73 3.98 6.43 3.08
56.06
Table 111. Carbon-Hydrogen Results Obtained by Combustion-Gas Chromatographic Technique
Calcd., C
YG
Compound Benzoic acid
68.85
4.92
Anthracene Acetanilide
94.33 i l ,09
5.66 6.70
Found, yo C H
H
69.4 68.1 68.8 69.5 68.0 94.9 70,66
ACKNOWLEDGMENT
The authors acknowledge the assistance of H. C. Lawrence, R. A. Hofstader, J. S. Parsons, J. H. Haurand, and J. F. Cannelongo in various phases of this investigation.
yc Found C
( 3 ) Kyryacos, G., Boord, C. E., “Gas
Diff., % C
H
5.27 4.87 4.61 4.83 4.66 5.90 6.86
$0.55 -0.75 0.00 +0.65 -0.85 ~ 0 . ~ 5 7 -0.43 Mean deviation 1 0 . 5 4
+0.35 -0.05 -0.31 -0.09 -0.26 $0.24 +0.16 10.21
F. J., IKD. ENG.CHEIII.,ANAL. ED.
Absorption Chromatography in the Analysis of Cool-Flame Combustion Products,” Division of Analytical Chemistry, 13ist Meeting, ACS, lliami, Fla., April 1957. (4) Niederl, J , B., \$-hitrnan, B,, ~fik~,,-
14,79-82 (1942). ( 6 ) Royer, G. L., Sundberg, 0. E., Zbid., 12,688-90 (1940). ( 7 ) Schoniger, IT., ikfikrochim. Acfa 1957, 545-52. (8) UnterZaucherj J.9 Ber. 7-39 391 (1940).
chemie 11,274-300 (1932). ( 5 ) Royer, G. L., Sorton, -4.R.,Foster,
RECEIVEDfor review August 24, 1959. Accepted November 16, 1959.
Determination of Micro Quantities of Boron in Steel by a Solvent Extraction Method LASZLO PASZTOR and
J. DANIEL BODE
Graham Research laboratory, Jones & laughlin Steel Corp., Pittsburgh 30, Pa. QUINTUS FERNANDO Department o f Chemistry, University o f Pittsburgh, Pittsburgh 7 3, Pa.
b On the basis of the work of Ducret,
of boron in 0.1 00-gram samples of
a solvent extraction procedure has been worked out in which the boron in steel i s converted to BFI-, extracted as a colored complex with methylene blue into dichloroethane, and determined spectrophotometrically. The method has been successfully applied to the determination of 0.2 to 25 y
steel.
A
analytical methods have not been completely satisfactory for the determination of microgram quantities of boron. especially in alloy steels. The solrcnt extraction method VAIL.4BLE
of Ducret ( Z ) , though time-consuming because of its slow BF4- formation, seemed most promising and IIXS investigated for the deterniination of boron in steel. APPARATUS
Spectrophotometric
measurements
VOL. 32, NO. 2, FEBRUARY 1960
277
After 2 hours' standing at room temperature add 0.1M potassium permanganate dropwise until the solution just turns pink, then add 2.0 ml. of 4% ferrous ammonium sulfate solution. If a calibrated polyethylene flask is not available, read the volume to determine the amount of rinse water needed, then transfer the solution to a 100-ml. polyethylene flask or bottle and rinse the graduate with enough water to bring the solution to 50 ml. Add 10.0 ml. of 0.001M methylene blue solution and 25.0 ml. of dichloroethane.
0 80
w U
060
4
m
0
2 04C. 020
0
I
5
I
1
10
15
MICROGRAMS
Figure 1.
, 20
I
25
rinsing out the platinum crucible make up the solution to 20.0 ml. Add 5.0 ml. of 5% hydrofluoric acid to the solution and then proceed as before. This gives the acid-insoluble boron. Adding the acid-soluble and insoluble boron gives the total boron. If only total boron has t o be determined, mix the soluble Fortion with the solution from the insoluble residue fusion in a 200-nil. polyethylene flask and make to 100 ml. before extraction. Extract and dilute as before, but use a calibration curve based on 100 ml. of aqueous solution.
I
30
car
OF B O R O N
Calibration curve
Micrograms of boron per 50 ml. of sample solution 40
W
were made with a Bcckrnan Model DU spectrophotometer and a Cary recording spectrophotometer using 1-em. quartz cells. A Model B Beckman spectrophotometer mas used for thP routine part of this work. Dissolution of the samples was carried out in boron-free glassware. For all other purposes, polyethylene laboratory ware was used. Pipets were made from polyethylene tubing and calibrated. REAGENTS
A standard boron solution containing 10 y of boron per ml. was prepared by dissolving reagent grade boric acid in water and diluting appropriately. Methylene blue (3.739 grams National Formulary) was dissolved in 1 liter of water and this solution was diluted ten times with water t o give a 0.001M solution. Commercial purified grade 1,2-dichloroethane was used for the extractions. All other chemicals were of reagent grade. A standard solution containing 1 y of boron per ml. as BFI- was prepared by adding 5 ml. of 5% hydrofluoric acid t o 50 ml. of standard boron solution. After 24 hours, the solution was diluted to 500 ml. with water. All reagents containing hydrofluoric acid were kept below 5' C. and all reagents except dichloroethane were stored in polyetliylene bottles. PROCEDURE
Soluble Boron. Dissolve 0.1000 gram of iron or steel sample in exactly 5.0 ml. of 2.5N sulfuric acid and 5.0 ml. of 2.5N phosphoric acid in a 100ml. boron-free flask on a n electric heating unit at 95' t o 98' C. under reflux (boron-free glass or polyethylene should be used for t h e reflux condensing tube). F o r stainless steel use 10.0 ml. of 2.5N sulfuric acid. Transfer the solution t o a polyethylene graduated cylinder and by washing the flask with small quantities of water dilute to 20.0 ml. Add 5.0 ml. of 5% hydrofluoric acid. Stopper the cylinder and shake it.
278
ANALYTICAL CHEMISTRY
z
4
P 4
zo
'1
0
- - - - -- -- - -- - 4Y)
\
,.q, '\.
cc-_-M'
aao
100
400
8%
700
WAVE LENGTH -MILLIMICRONS
Figure 2. Spectral absorbance curve of methylene blue-boron complex in dichloroethane Molar absorption at 660 m p is of the order of 65,000 mole-' 1 cm-1 1 0 y of 6 in 1 2 5 ml. of solvent Reagent blank
--
----
Stopper the flask and shake it well for 1 minute. After the layers separate completely, pipet out 5.0 ml. of the organic layer and dilute to 25.0 ml. with dichloroethane in a volumetric flask. Read the absorbance of the solution at 660 mp against dichloroethane. Run two reagent blanks. Subtract the reagent blank average from the absorbance value of the solution and convert the result to micrograms of boron, using a curve made up with standard boron solutions. This gives the soluble boron (Figure 1). Insoluble and Total Boron. If both soluble and insoluble boron are t o be determined, follow t h e procedure as before, b u t when transferring t h e solution t o t h e polyethylene graduate, filter t h e solution through a small S. & S. 589 Blue Ribbon filter paper using a polyethylene funnel and wash i t with as little water as possible (ca. 4 ml.). Place the filter paper in a platinum crucible and ignite in a muffle furnace a t 250' C.and finally a t 550' C. Add exactly 1.OO gram of solid sodium carbonate to the platinum crucible and fuse the contents. Cool t o room temperature and dissolve carefully in 5.0 ml. of 2.5N sulfuric acid and 5.0 ml. of 2.51V phosphoric acid. Transfer the contents to a polyethylene graduated cylinder and by
0
IO
20 30 M L OF 2 . 5 N
50
40 AC D
Figure 3. Effect of acids on extraction 2.5N acids added to 10 ml. of standard 6 s o b tion, containing 0.05% HF, and 10 ml. of 0.001M methylene blue, diluted to 60 ml. then extracted with 2 5 ml. of dichloroethane and subsequently diluted 1 to 5. Absorbance measured at 660 m u in 1 -cm. cell
L
0 60
0
1
lor _--
2
3
4
PH
BORON REAGENT B L A N K
Figure 4. tion
Effect of pH on extrac-
6 added as BFd-. H2S04 solutions. pH above 2 adlusted with N o z H P O ~ . Absorbonce at 660 m N
Mix and treat the blanks in the same ray. For low boron contents and especially with low insoluble boron, i t is advisable t o take 0.2000- t o 1.0000-gram samples and t o use correspondingly higher quantities of all the inorganic reagents except sodium carbonate, where it is not necessary t o use more than 2.00 grams for any sample. Before color development and extraction, take aliquots corresponding t o 0.10 gram of sample, dilute them t o 50.0 ml., and treat as before.
Table I. Effect of Time after Oxidation Stage Adjustment y B Found ( 1 . 9 y
Extracted, Min.
Added) Oxidized Oxidized after before HF HF treatment treatment
Immediately After 3 min. 6
12
1.9 1.9 1.9 2.0
2.6
RESULTS AND DISCUSSION
Preliminary extractions were carried out using 50 ml. of aqueous phase containing the standard BIT4- solution, acid or mixed acids, and 10 ml. of methylene blue solution (0.OOlM). This was shaken for 1 minute with 25 ml. of dichloroethane, then 5 ml. of the separated organic phase were diluted to 25 ml. with dichloroethane and the absorbance \\as measured a t 660 mp in a 1em. cell with a spectrophotometer. Figure 2 shows the absorbance curves of the methylene blue-BF4- complex (0.05M in HF) and the reagent blank, against dichloroethane. The peak a t 660 mp was used in all subsequent measurcmcnts because the difference in the absorbance between the methylene blue complex and reagent blank is greatest a t this Lqave length. This maximum for both the methylene blue complex and the reagent blank did not shift with change in acidity or type of acid. The effect of various acids on the extraction of the complex is shown in Figure 3. Sulfuric acid, phosphoric acid, or sulfuric-phosphoric acid mixtures give satisfactory results even a t p H 1.0. Nitric, perchloric, and hydrochloric acids are unsuitable and so is hydrofluoric acid a t high concentrations. Fortunately, iron and most steel samples can be dissolved in sulfuric acid or sulfuric-phosphoric acid mixtures. If the p H of the solution containing the dissolved steel is raised above 1.5 or 2, some of the metals present may be precipitated and there is a serious danger of coprecipitating boron ( I , S, 6). Hence, the possibility of working in strongly acid solutions was investigated. Figure 4 s h o w the effect of p H on the extraction of the methylene blue-BF4- complcx. The extractability of methylene blue itself increases, while that of the complex decreases slightlv with decreasing p H in sulfuric and phosphoric acid solutions. Therefore the difference in absorbance betn een the reagent blank and the methylene blue-BF4- complcx decreases significantly. However, the difference a t pH 1, or slightly less, is still large enough for the method to be applicable. Because the concentration of hydrofluoric acid has a profound effect on the extraction of the colored complex, it was
cal because the absorbance difference between the methylene blue-BF4- complex and reagent blank decreases rapidly (Figure 3). These results were confirmed when dilute solutions of dissolved steel were used instead of standard boron solutions. Investigation of the effect of shaking time on the extraction of a standard boron solution containing 10 y of boron as BF4- and 1.0, 4.0, and 10.0 nil. of 2.5M hydrofluoric acid in the aqueous phase, which had a total volume of 60 ml., showed a 1-minute shaking time was sufficient. Diethyl ether, chloroform, carbon tetrachloride, 1, 2-dichloroethane, di-
Table ll.
Results Using Direct Solvent Extraction and Other Methods" Curcumin QuinalizNew after ann Sample Method UNIARCb CuF2b Fordc Distil. Directd RH1093 0.0005