t o 4, the iron(II1) is incompletely CC,,Jplexed. The influence of the initial pH of the solution on the granular charactzr of the precipitate suggests that the crystal nucleation for dilute iron solutions is completed prior to hydrolysis of the iron complex. If the initial concentration of iron(II1) is increased apprcciably, a hydrous oxide is precipitated quantitatively, but under these conditions the granular oxide cannot be prepared hpcause DHEG is too weak an acid to complex an appreciable fraction of the iron(II1) initially. For concentrated solutions. the formation of the granular oxide will undoubtedly be favorcd by employing an agent which more strongly complexes iron(II1) than does DHEG. The tetradentate DHEG ligand also appcars to be important in facilitating the formation of the nongelatinous oxide. While the role of the ligand has not been investigated extensively, preliminary experiments indicate that only tetradentate ligands are suitable for the preparation of a granular oxide.
Co u I o met ric
The empirical formula of the iron(II1) complex is Fe(OH)2G, where G- represents the ligand anion. If the primary nuclei of (FeO.OH), (6) condense b y a reaction of the type
+
(FeO.OH), HO-FeG-OH HO-FeG-O-Fe-(FeO.OH),l
=
+ H20
it may be recognized that tetradentate ligands are peculiarly adapted to the formation of the crystalline oxide by precluding excess water from the coordination sphere of the iron atoms. Other conditions favorable to the formation of a granular oxide have been described ( 1 ) . Attempts to alter the composition and boiling temperature of the solvent b y the addition of ethanol, ethylene glycol, and glycerol were not successful because the precipitation is much less complete in the presence of such hydroxylic solvents. ACKNOWLEDGMENT
The assistance of the Geigy Chemical
Corp. in providing the N,N-dihydroxyethylglycine used in this study is gratefully acknowledged. LITERATURE CITED
(1) Benck, R. F., M.S. thesis, Univemity
of Nebraska, 1958. (2) Duval. C., “Inorganic Thermogrsvimetric Analysis,” Elsevier, New York, 1953. (3) Gayer, K. H., Wootner, L., J. Phys. Chem. 60, 1564 (1956). (4) Hofer, L. J. E., Peebles, W. C., Dieter, W. E.. J. Am. Chem. SOC.68, 1953 (1946). (5) Kratky, D., Nowotny, H., 2. Kn’st. A100, 356 (1938). (6) Krause, A., Kolloid 2 . 7 2 , 18 (1935). ( 7 ) Toren, P. E., Kolthoff, I. M., J. Am. Chem. SOC.77, 2061 (1955). (8) Weiser, H. B., Milligan, W. O., Ibid., 57, 238 (1935). E. R. NIGHTINGALE, JR. R. F. BENCK Department of Chemistry University of Nebraska Lincoln 8, Neb. RECEIVED for review November 25,1959. Accepted February 1, 1960.
Oxidation of Boron
SIR: The method of oxidizing or reducing films on metallic surface has been extensively employed for the study of surface coatings and corrosion films (2, 3). The film is oxidized or reduced at constant current, and the potential of the metal is measured during electrolysis. The completion of the film dissolution is indicated by a rapid change in the potential. The total number of coulombs calculated from the current and time of electrolysis corresponds to the quantity of film. During the course of high-temperature studies on boron carbide, it was necessary to analyze microgram quantities of elemental boron deposited on platinum targets. Conventional wet chemical methods failed to remove the boron completely from the platinum base. Coulometric oxidation was then attempted as a method for dissolution of the boron film. Good correlation was obtained between the number of coulombs used and the amount of boron removed from the target. The results reported here are not extensive because of the secondary nature of the investigation. But in light of the increased interest recently in the chemistry of boron, the results may be of value t o the study of boron deposits and boron in alloys. This is also believed to be the first report of the oxidation of elemental boron by electrolysis. EXPERIMENTAL
Constant current was provided by
~~~~~~
Table I.
~
Results of Anodic Boron Stripping
Exposure Current, Ma. Coulombs 3-1 1.52 0.213 3-2 1.52 0.505 3-7 1.082 0.203 3-8 0.529* 3-10 1 306 1.595 3-1 1 0.998 0.734 5-16 1.065 0.416 Accuracy 1.0%. b Summation of five different current levels.
Stripping 8 0
18.9 7.8 19.7 59.6 27.4 15.6
Boron Found, Titrations 7.3
16.5 7.9 14.9 51.2 25,l 13.2
y
Residual. 2.8 2.2 0.5
3.5 29.3 18.7 10.2
Q
four 45-volt B batteries connected in series and dropped across a selector m i t c h bank of large standard resistances for different current levels. The current levels were determined by mezsuring the voltage drop across a standard IO-ohm precision resistance with a Leeds & Northrup portable precision potentiometer. The samples analyzed were targets coated with boron. The boron was deposited in vacuum from a Knudsen cell containing boron carbide at about 2300O K. The target itself was a platinum disk ll/s inches in diameter by 0.005 inch thick. The boronplated disk was placed, face up, in a 30-ml. quartz beaker containing 5.0 ml. of 0.10A’ sodium hydroxide. A threeholed rubber stopper covered the beaker. It also supported two saturated calomel electrodes (S.C.E.) and a platinum wire probe. The probe was used for making electrical contact Rith the disk. The saturated electrodes were of the type used by Adams et al. ( 1 ) ; one served as a working cathode and the other a reference electrode.
The potential of the platinum disk us. S.C.E. was monitored by a Leeds & Xorthrup Speedomax recorder adjusted for a full scale deflection corresponding to 2.0 volts. The recorder operated a t a chart speed of 2 inches per minute. After the coulometric oxidation end point was reached, the electrodes and target were removed and the resulting solution was analyzed for boric acid by coulometric titration of the mannitol complex (4). Residual boron left on the targets after the coulometric stripping was determined by burning the targets in air, dissolving the resulting boron oxide (B203)in 0.10A‘ sodium hydroxide, and titrating as above. Spectrographic analysis proved that the burning process completelv removrd all boron from the targets (4). RESULTS
A well defined S-shaped anodic potential-time curve was obtained for the boron dissolution. The initial potential of 0.30 volt us. S.C.E. corresponds to
+
V O L . 32, NO. 4, APRIL 1960
567
the oxidation of elemental boron in the 0.10N sodium hydroxide solution. The inflection point in the curve is the end point for the determination. The background oxidation for platinum in the 0.10X sodium hydroxide solution is about 0.60 volt us. S.C.E. The amount of boron in micrograms calculated from the quantity of electricity consumed agrees well with the chemical determination of the boron electrolyzed into solution if a 3-electron oxidation process is assumed. The product of the oxidation under these conditions is, therefore, the stable-plus-three oxidation state, probably the orthoborate ion. Table I summarizes the data for the determination of boron by coulo-
+
metric oxidation (stripping) and coulometric titration of the resulting borate solution. The agreement for most cases is within 20% for this level of boron. Column 1 gives exposure numbers for the Knutlsen experiment reported elsewhere (6). l s the original amount of boron on the target increases, the amount of residual boron left on the targets becomes larger. Small sections on the target similar in color to the original boron deposit suggest that electrical contact was not established between the boron and the platinum base in these cases. One possible reason for this may be the presence of impurity vapors in the vacuum system.
LITERATURE CITED
(1 Adam?, R. K., McClure, J. H I d o r r i s J J , ANAL.CHEM.30, 471 (1958). (2) Delahaj-, P., "Sew Instrumental 6
Methods in Electrocheniistp," Chap. 14, Section YIII, Interscience, Sew York, 1954. (3) Furman, S . H., J . Electrocizetri. SOC.
101, 19c (1994). (4) Robson, H. E., Ph.D. thesis, University of Kansas, 1958. HARRY ROBSOX Esso Research Laboratories
Eseo Standard Oil Co. Baton Rouge, La.
THEODORE KVYm.4 iierojet-General Corp. P. 0.Box 296 Azusa, Calif. RECEIVED for review Sovember 2, 1Y59. Accepted February 12, 1960. K o r k done a t the Department of Chemistry, Cniversity of Kansas, Lan-rence, Kan.
Sucrose Acetate Isobutyrate as a New Ester Liquid Phase for Gas-Liquid Partition Chromatography SIR: Liquid phases such as polyesters have been useful as partitioning agents for polar compounds in gasliquid partition chromatography. However, there have been difficulties because of nonuniformity (4) and thermal instability. Decomposition products tend to foul the detector, and the gradual loss of the liquid phase leads to eventual ineffectiveness of the column (1, 5 ) . -4 strongly polar ester phase which avoids these defects is sucrose diacetate hexaisobutyrate (SAIB). This material is a monomeric compound of uniform properties now produced in large quantities a t a low price for use as a plasticizer ( 2 ) . SL41B decomposes a t temperatures above 200' C., but the only volatile decomposition products are acetic and isobutyric acids. These acids, unlike polymer fragments. do not char in hot wire type detector systems. The apparatus used was a Burrell
Table
I.
Data in Tahle I compare rctvntion volumes of mint oil constituents. Ternis, symbols, and units arr those reconi-
Relative Retention Volumes of Mint Oil Constituents with Ester Liquid Phases
Phase Temperature, "C. =k 2" Flow rate, ml. He/min. Mint Oil Constituents a-Pinene (3-Pinene I-Limonene Cineole Menthofuran I-Menthone d-Neomenthol Menthyl acetate I-Menthol Piperitone I-Carvone .Relative t o I-menthol, corrected
568
Kromotog K-2 nith a thermal conductivity detector cell operated a t 200" C. TKO column packing were used and their performances were evaluated under conditions which gave the best separation in a reasonable time (30 minutes). For the first column packing, 10 parts of SAIB nere deposited from ethanolic solution on 100 parts of support [aqua regia-washed firebrick, 30 to 60 mesh (671. The second packing was Resoflex LAC-IR-296 (an adipate polyester of diethylene glycol) on the support material (Celite 30 to 60 mesh) at a 257, concentration. Depending on column temperature and on the relative concentrations and positions of the major constituents, samples of from 2 to 10 pl. were used. Samples were injected with a Hamilton microliter hypodermic springe through silicone seal. into the inlet of the column.
ANALYTICAL CHEMISTRY
SAIB 120 i8
145 77
LAC-296 170 i4
120 41
mended by Section L on Gas Chromatography of Research Division IV of $STM Committee D-2 (3). The reference standard used was l-menthol. Table I1 shows the specific retention volumes (V,) and the partition coefficients ( H ) for l-menthol. These parameters ( V , and H ) were calculated from observed values of flow rate, p,lp,, Vg and V L under the reported conditions. The data given permit the calculation of T', and H for all constituents in Table I lvith both phascq a t the temprratures used. SAIB gave a sharper separation of limonene and cineole than LAC 296 by visual inspection. SAIB separated menthofuran from melithone and neomenthol from menthyl acctate a t all temperatures, n hereas LAC 296 did not separate these components a t an! temperature. Seither liquid phase used separated piperitone from carvone. Results n.ith r c o n (a polyether) w r e
Table II. Specific Retention Volumes and Partition Coefficients for /-Menthol
145 32
170 10
0.07
0.09 0.14 0.19 0.23
Phase SAIB
0.65 0.84 0.78 1.00 1.78 1.85
LAC-296
Retention Volumesa 0.08
0.11 0.16 0.19 0.53 0.62 0.86 1.09 1.00 1.50 1.55
0.10 0.15 0.19 0.23 0.58
0.68 0.87
1.11
1.00 1.55 1.54
0.13 0.19
0.23
0.27 0.63 0.74
0.89
1.11 1.00 1.57 1.50
0.05 0.09 0.12 0.14 0.53 0.58 0.81
0.80
1.00 1.86
2.00 to correspond to ( Vg)' values in ( 3 ) .
0.12 0.16
0.18 0.62 0 68 0.82 0.82 1.00 1.85 1.90
0.69
Specific RetenTemperation Partition ture, Volume, Coefficient, "C. + 2" V, H 120 145 170
850 320 130
1280 505 210
120 145 170
230 84 39
360
145 72
Deneity (120' to 1iO"C.): for SAIB, 1.150 (1 - 7.4 X 10-4t); for LAC-296, 1.248 (1 - 7.5
x
10-4t)
