Proposed structure of the isolated, deprotonated
Figure 4. Proposed structure of the analytical species of interest
The above elemental analysis indicates that the complex has the structure shown in Figure 3. This is a deprotonated complex of the general type R2Cu. However, a sample of this isolated species redissolved in nitrobenzene shows maximum absorption at 555 nm and not at 536 nm for that of the extracted species. Obviously, the isolated species is not the analytical species of interest. Geldard and Lions (8) have recently reported the isolation of a series of copper complexes of a closely related ligand (picolinealdehyde 2-pyridylhydrazone) and a variety of anions. They reveal a complex of the general composition, RCuOH, where R indicates the deprotonated ligand. Deprotonation readily occurs at the imine nitrogen in the hydrazone linkage (9). Qualitative tests have shown that when the 555 run species in nitrobenzene is equilibrated with a basic aqueous solution, free ligand is taken into the aqueous phase, and the
absorption maximum of the nitrobenzene phase shifts to 536 nm. Likewise, tests have shown that when the 536 nm species in nitrobenzene is treated with a dehydrating agent, such as anhydrous sodium sulfate, in the presence of excess ligand (QAQH), the absorption maximum of the nitrobenzene phase shifts to 555 nm. As a result of the above observations, the 536-nm species appears to have the structure compatible to that observed by Geldard and Lions (8), and shown in Figure 4. Attempts to isolate and identify positively the 536-nm species have been discouraging; however, work is being continued toward this end.
Figure 3. complex
(8) J. F. Geldard and F. Lions, Znorg. Chem., 4, 414 (1965). (9) J. F. Geldard and F. Lions, J. Chem. SOC.,84, 2262 (1962).
RECEIVED for review October 4, 1967. Accepted December 11, 1967. The authors thank Research Corp. and the Gustavus Research Fund for the financial support of this work. Presented at the Eighteenth Annual Pittsburgh Conference on Analytical Chemistry and Applied Spectroscopy, Pittsburgh, Pa., March 6,1967.
Determination of Vanadium and Oxygen in Vanadium Oxides by Differential Spectrophotometry and Inert Gas Fusion-Gas Chromatography Edward W. Lanning and Richard P. Weberling Bayside Laboratory, Research Center, General Telephone & Electronics Laboratories, Inc., Bayside, N . Y.
A RECENT REQUIREMENT in our laboratory for determining the stoichiometry of various oxides of vanadium necessitated precise determination of vanadium and oxygen contents. VANADIUM
Classical titrimetric methods employing either potassium permanganate ( I ) or EDTA (2) are limited by indefinite end points caused by the highly colored vanadium ions. The intense blue color of quadrivalent vanadium ion, however, suggested the possibility of utilizing a differential spectrophotometric method. In addition to eliminating the end point problem, this photometric method permits the use of tartaric acid as the reducing agent and thus eliminates the necessity for expelling excess reagents as well as the need for guarding against autoxidation. Experimental. APPARATUS.A Beckman Model DU was used for all absorption measurements.
(1) "Sutton's Volumetric Analyses," 13th ed., Butterworth Scientific Publications, London, p 457. (2) V. R. M. Kormal and S. C. Shome, Ana(. Chirn. Acta, 27, 594 (1962). 626
ANALYTICAL CHEMISTRY
REAGENTS.Fisher certified reagent grade vanadium pentoxide was used as the standard material for both the vanadium and oxygen methods. PROCEDURE. Samples containing 125 to 162.5 mg of vanadium are dissolved in 15 ml of concentrated sulfuric acid, evaporated to copious fumes of SO3 and diluted to 150 ml with distilled water. Ten milliliters of a 15% tartaric acid solution are added, and the solutions heated just to boiling. The samples are transferred to 250-ml volumetric flasks and diluted to the marks. A standard stock solution is prepared by taking sufficient reagent grade V205to give 2.500 grams of vanadium. The Vz05is dissolved in 150 ml of concentrated sulfuric acid and evaporated to copious fumes. The solution is transferred to a 1-liter volumetric flask and diluted to the mark. The standards are prepared by taking 50.0-, 55.0-, and 60.0ml aliquots of stock solution. The standards are reduced and diluted in the same manner as the samples. The absorbances of the samples and standards are then measured against the 50.0-ml standard at 765 mp using 10-cm cells. A calibration curve is prepared by plotting absorbance against milligrams of vanadium for the standards, and the amount of vanadium in the samples is read from the calibration curve.
Results. A spectral scan of 125 mg of V4+ per 250 ml in dilute sulfuric acid showed an absorption maximum at 765 mp. Variation of the sulfuric acid content from 4 to 1 2 x produced no significant change in the absorption of V4+ ion nor did the variation of the tartaric acid content from 0.1 to 2.0 grams. Stability tests indicated that the reduced vanadium solutions were stable for at least 96 hours. The slope-concentration method of Bastian ( 3 ) and Hiskey ( 4 ) for determining the optimum concentration of absorbing species was employed. The slope for the absorption us. concentration curves was determined for increasing amounts of vanadium in the reference solutions (Table I). As the amount of vanadium in the reference is increased, wider slit settings are required in order to provide enough light to penetrate the solution. However, wider slits result in increased stray light, which in turn produces a decrease in slope. The slope-concentration product is a direct index of the precision attainable. Table I shows that an increase in vanadium concentration above 150 mg per 250 ml results in a loss of precision. Table I1 shows the analytical recoveries obtained for four synthetic samples which were determined by two analysts on successive days. The precision approaches one part per thousand. Table I11 compares spectrophotometric results with those obtained by the permanganate titration. The accuracy for the spectrophotometric results is at least comparable with the permanganate titration.
Table I. Optimum Concentration of Vanadium in Reference Vanadium concn in reference, Slope X mg/250 cc Slope, abs/mg V concentration 0 25 50 75 100 125 150 175
munication.
0.700 1.035 1.320 1.513 1.575 1.470
z
125.1 124.9 125.0 125.3
100.16 100.00 100.08 100.32
u = 1.4ppt
z
125.0 124.7 124.8 125.0 u =
100.08 99.84 99.92 100.08
1.3ppt
Table III. Comparison of Photometric and KMn04 Methods KMnOa Spectrophotometric Sample method method 1
(3) R. Bastian, ANAL.CHEM., 23, 582 (1951). (4) C. F. Hiskey and I. G. Young, Zbid., 23, 1196 (1951). (5) W. 0. Smiley, Ibid.,27, 1098 (1955). (6) V. A. Fassel and R. W. Tabeling, Spectrochim. Acta, 8, 201 (1956). (7) R. K. Winge and V. A. Fassel, ANAL.CHEM., 37, 67 (1965). (8) L. Singer, IND.ENG.CHEM., ANAL.ED.,12, 127 (1940). (9) H. L. MacDonell, R. J. Prosman, and J. P. Williams, ANAL. CHEM., 35, 599 (1963). Q (10) H. R. Grady, Vanadium Corp. of America, private com-
0.353
Table 11. Recovery of Vanadium from Synthetic Samples (124.9 mg of V taken) Recoverv Analyst A, day 1 Analyst B, day 2 Mg Mg
OXYGEN
Our next problem was to determine oxygen with an accuracy approaching 1%. In the past, vacuum fusion (9, emission spectrometry (6), dc arc extraction (3, and inert gas fusion (8) techniques have been used for the determination of minor or trace amounts of oxygen in refractory materials, Recently, MacDonell, Prosman, and Williams (9), by modifying the Leco inert gas fusion titrimetric apparatus and employing a concentrated Ba(OH)2 solution, were able to determine major amounts (0 to 55%) of oxygen in glasses, refractories, and refractory oxides. The precision attainable was 2%, which was inadequate for our problem. Moreover, the heavy BaC03 precipitate formed in the titration cell necessitated frequent flushing with hydrochloric acid. Grady (IO) employed the Leco inert gas fusion apparatus, but replaced the conductometric titration unit with a COz absorption train. While the precision approached 1%, the reduced argon gas flow (150 ml per minute), necessary to prevent the loss of COz,extended the analysis to 20 minutes. More recently, commercial apparatus (Leco Nitrox-6) for the determination of small amounts of oxygen has become available. This apparatus combines inert gas fusion and gas chromatography. The CO and Nz formed in the reaction crucible are swept by the helium carrier gas directly into a trap. The gases are then released to the chromatograph.
...
0.0141 0.0141 0.0140 0.0138 0.0132 0.0121 0.0105 0.0084
2
3
4
80.18 80.37
79.86 80.54
Av. 80.28
80.20
75.04 74.68
74.63 74.73
Avo 74.86
74.68
72.21 72.39 72.25
72.37 72.41
Av. 72.28
72.39
71.30 71.42 71.18
71.11 71.18
Av. 71.30
71.15
The time for an analysis is about 10 minutes. We decided to attempt to extend this method to the analysis of major amounts of oxygen. Experimental. APPARATUS.A Leco inert gas fusion furnace (537-300) and a Research Specialties gas chromatograph were fitted together in the following manner. The exhaust side of the furnace combustion tube was connected to an 8-inch high U tube trap through a diverter valve. The “out” side of the valve was connected to the inlet leg of the gas chromatograph column. Both the trap and the column were packed with Linde Molecular Sieve 5A (40 to 60 mesh). The sample-loading device (Leco 734100) was modified with an optical flat and a reflecting prism so that temperature measurements could be made with an optical pyrometer. Graphite capsules similar to those suggested by Beck and Clark (11) were employed for transferring samples to the furnace. The use of these capsules prevents the pacification of the reaction crucible which is encountered when vanadium oxide is reacted by the usual bath technique. Each capsule is (11) E. J. Beck and F. E. Clark, ANAL.CHEM., 33, 1767 (1961). VOL 40, NO. 3, MARCH 1968
627
in effect a small reaction chamber and serves to retain most of the vanadium after combustion. PROCEDURE.The Leco furnace and the gas chromatograph are set to the conditions listed in Table IV. The furnace is outgassed at 2450" C for 2 hours. The temperature is then dropped to 2400" C and a capsule blank is run. A value of less than 0.05 mg of oxygen is satisfactory. If the blank is higher, continued outgassing is indicated. ~~
Table IV. Furnace and Chromatograph Operating Conditions Leco furnace Furnace temperature 2400" c Helium flow 350 mljmin
Gas chromatograph Column diameter, i.d. Column length Packing material
'la inch 12 ft Linde Molecular Sieve 5A,40-60 mesh 150" C 150" C 170 ma 50 ml/min
Column temperature Detector temperature Detector bridge current Helium flow rate
TIME (minutes)
Figure 1. Typical gas chromatogram Table V. Total Analysis of Vanadium Oxides Std %0 2 dev, % % V Total Oxide Theor Found Av V208 32.03 32.0 32.4 31.8 0.5 67.6 99.4 31.0
vo
23.90
VOa
38.56
24.55 24.63 24.44 23.84 40.5 40.2 39.3 40.0 39.8 40.3 39.3 41.6
24.33
0.7
40.0
0.4
59.2
99.3
Table VI. Total Analysis of Miscellaneous Materials V % Cr %0 Total 3.71 32.17 95% VzOa64.51 64.58 3.85 32.06 5% C~ZOP Av 64.55 3.78 32.06 100.47 %V
50% v20850% Thoa
33.88 33.89 Av 33.89
%Ti 33.34 33.26 33.26
%O
32.68 32.51 32.67
99.82
%Ti %O 34.46 65.69 34.43 65.31 Av 34.45 65.50 99.95 % Ti %0 2 TiOt 66.57 33.14 66.67 33.09 Av 66.62 33.12 99.74 a Vanadium determined after separation from chromium via cupferron precipitation. Vanadium determined in presence of titanium. TiaOa
628
ANALYTICAL CHEMISTRY
The sample is placed in a tared graphite capsule, weighed, plugged, and transferred into the loading device. It is flushed with helium for 21/2minutes and the plunger is then pushed and locked in place directly over the graphite furnace crucible. The diverter valve is turned to the collect position, and a Dewar of liquid nitrogen is raised over the trap. Any nitrogen lost is replaced. The sample is dropped into the furnace and a timer is set for 6 minutes. After the required time has elapsed, the diverter valve is turned to direct the flow of the trapped gases into the gas chromatograph. The liquid nitrogen container is lowered and a beaker of hot water (75" C) is raised over the trap and held there for 1 minute, releasing the gases into the gas chromatograph. An analytical curve is obtained by plotting the weight of oxygen against CO peak area for Vz05standards. The amount of oxygen in the sample is read from this curve. Results. To minimize the weighing error it was necessary to take enough sample to provide 15 to 20 mg of oxygen. This necessitated the use of a 12-foot column in the chromatograph, in order to distinguish the nitrogen peak from the carbon monoxide peak. A typical chromatogram is seen in Figure 1. The area under the CO peak is proportional to the amount of oxygen in the sample. This relationship was linear over a range of 12 to 25 mg of oxygen. However, if significant amounts of nitrogen (1 to 27,) were present, no more than 20 mg of oxygen could be taken. This was necessary to maintain good resolution between carbon monoxide and nitrogen. As samples are added and the graphite furnace crucible becomes full, the temperature of the capsule during the combustion reaction is reduced because of the inefficient heat transfer between graphite furnace crucible and the sample capsule. A capsule temperature in excess of 1950" C was necessary to ensure complete reduction of the oxides to CO. After eight or nine samples have been processed, the system is usually incapable of raising the new sample capsule to temperature. If a capsule temperature of 1950" C is not attained, the furnace is opened and the spent capsules are removed from the crucible. The furnace is reassembled and the run continued after a 30-minute outgassing.
Tables V and VI show summation analyses for various materials. Vanadium results were obtained using differential spectrophotometry and nitrogen values were obtained by means of the micro-Kjeldahl method. Table V shows results obtained for Vz03, VO, and VOz. For vzo3 and VO the mean oxygen value is within 0.7 and 1 . 8 z , respectively, of theoretical. For VOz the oxygen content found is higher than theoretical but is validated by the summation analysis. In addition, the precision for the VOz results is about 1%. Table VI summarizes the results obtained for the total analyses of samples of other oxide matrices. The summations indicate good accuracy for both vanadium and oxygen
methods. Both methods of analysis are at least as accurate as prior methods, and the time for analyses is markedly reduced. In the case of the oxygen method the dovetailing of samples reduces the analysis time per sample to about 10 minutes. ACKNOWLEDGMENT
The authors thank Frank V. Durkin and William D. Shelby for their assistance in performing the analyses. RECEIVED for review October 24, 1967. Accepted December 29,1967.
Composition and Structure of Acrylonitrile-Methacrylic Acid-Methyl Methacrylate System Ferencz Denes, Nicolae N. Asandei, and Cristofor I . Simionescu Institute of Macromoleculare Chemistry “ P . Poni,” Jassy, Rumania
DETERMINATION of the composition of copolymers presents many difficulties, stemming from the nature of the functional groups, the number of monomers taking part in the reactions, and the limited sensitivity of many physicochemical methods of analysis to a relatively small number of functional groups present in the macromolecule. To avoid these difficulties we suggest a titration system, using solvents and organic bases that react within the homogeneous phase both with the low molecular weight compounds, and polymers or copolymers with acid functions (1-8). In this study the composition of the acrylonitrile (AN)methacrylic acid (MAC)-methyl methacrylate (MMA) ternary copolymer was determined using potentiometric titration, with sodium hydroxide and tetraethyl ammonium hydroxide (TEAH) in different solvents. EXPERIMENTAL
Titration with NaOH. In preliminary determinations, dimethylformamide (DMF) and dimethylsulfoxide (DMSO) were used as polymer solvents; the titration was carried out potentiometrically in an aqueous solution of sodium hydroxide. The copolymer starts precipitating on addition of 0 . W aqueous NaOH solution and the greater the number of acid groups in the polymer chain, the faster the precipitation takes place. Use of organic solvents for dissolution of copolymers and for preparation sf base solutions permits titration of the carboxylic groups. Titration of these groups in the ternary copolymer dissolved in DMSO takes place more satisfactorily than the titration of the carboxylic groups of the acid homo~
~~~
(1) G. A. Harlow, C. M. Noble, and Garrard E. A. Wyld, ANAL. CHEM., 28,787 (1956). (2) Robert H. Cundiff and Peter C. Markunas, Zbid., p 792. (3) J. E. Burleigh, 0. F. McKinney, and hl. G. Barker, ANAL. CHEM.31, 1684 (1959). (4) A. H. Beckett and E. H. Timley, “Titration in Nonaqueous Solvents,” 3rd ed., BDH Laboratory Chemicals Division, Poole, Dorset, England. ( 5 ) T. Iasinski and S . Krystyna, Chem. Anal. (Warsaw), 10, 211 (1965). (6) Mitsuru Nagasawa, J. Phys. Chem., 11,4005 (1965). (7) S. J. Fritz and N. 31.Lisicki, ANAL.CHEM., 23, 589 (1951). (8) Ichiro Sakurada and Yukihiko Osurni, Kobunshi Kuguhu, 19, 620 (1962).
polymer where precipitation takes place near the equivalence point (Figure 1). However, the NaOH solutions in D M F and DMSO are unstable; NaOH solutions in D M F decompose faster. Titration with TEAH. The difficulties in using NaOH for the titration of carboxylic groups in the ternary copolymers of acrylonitrile can be overcome by using organic bases. Results in satisfactory agreement with the total concentration of carboxylic acid groups present, have been obtained for the titration of carboxylic groups of the AN-MAC-MMA copolymer with tetraethylammonium hydroxide in DMSO. Results of potentiometric titration of carboxylic groups by TEAH both in polyacid, and in the physical mixture of polymethacrylic acid and binary copolymer AN-MMA, indicate that in both cases half of the number of carboxylic groups existing in the polymers were titrated (Figure 2). The other functional groups of the copolymer (-CN and -0-CH3) are not affected in this reaction. Because of the large size of the quaternary ammonium ions as compared to the sodium ions steric hindrance occurs. Gregor and Frederick (9) titrating polymethacrylic acids with quaternary ammonium bases in aqueous solutions, noticed a decrease of acid strength with increasing size of the basic molecule. These authors concerned themselves primarily with the qualitative aspects of these phenomena. Based on these results, we found it possible to determine the structure of the polyacids and the acrylic copolymers with functional acid groups. Potentiometric titration with sodium hydroxide of the polymethacrylic acid in water gave a base consumption equivalent to carboxylic groups existing in the polymer. When the same polyacid was titrated under similar conditions with TEAH in water, only half the carboxylic groups were titrated. It thus appears that DMSO does not interfere, and that steric factors prevent TEAH from reacting with all polyacid carboxylic groups. The carboxylic groups are likely to be arranged in a certain configuration in the chain. By reacting the carboxylic groups of the ternary AN-MAC-MMA copolymer with a 0.05N solution of TEAH in DMSO, the same results were obtained as by titrating the copolymers in a 0.05N solution of sodium hydroxide in DMSO. Thus the same amount of organic base and NaOH is consumed by the carboxylic groups of the copolymer (Figure I). (9) Harry P. Gregor and hIichael Frederick, J . Polymer Sci., 23, 451 (1957). VOL 40, NO. 3, MARCH 1966
629