been corrected for the impurities present, in each nitroparaffin. These corrections amount to an increase of 0.5% for the data obtained using nitromethane, 401, for nitroethane, and 1.7% for l-nitropropane. However, when the observed absorbances for 2-nitropropane are corrected for impurities, they are reduced to only 20% of their original values. After correction, the absorption of the product of the 2-nitropropane and p diazobcnzcne sulfonic acid a t 395 mp is only 1 or 2Oj, of the absorbances a t 395 mp for the rcactions involving nitroethane and 1-pitropropane. Even this small amount of absorbance may be due to too small a correction for impurities, as the alpha carbon in a secondary nitroparaffin should not be capable of coupling t o form a hydrazone. In terms of specific extinction coefficients in absorbance units per microgram per milliliter for a 1-cin. optical path length, the following values are obtained for the product of the reaction of p-tiiazobenzenesclfonic acid with nitro-
methane, 0.042; with nitroethane, 0.023; with 1-nitropropane, 0.016; and with 2nitropropane, 0.0003. The corresponding molar absorption coefficients are 9.0 X lo3 for 1-nitroformaldehyde p sulfophenylhydrazone, and 3.9 X lo3 for 1-nitropropionaldehyde psulfopheny lhydrazone. It was not possible to resolve the product peaks when a mixture of 75 y of 1-nitropropane and 30.6 of nitromethane reacted with diazotized sulfanilic acid. Instead, a broad absorption maximum a t about 412 mp mas observed. However, it would seem possible to determine nitromethane in the presence of smaller quantities of higher nitroparaffins. A solution containing 100 y of 2-nitro2-methyl-1-propanol did not seem to react with pdiazobenzenesulfonic acid to produce any appreciable absorption in the region above 350 mp. However. 100 y of 2-nitro-2-ethyl-1, 3-propanediol had an absorbance of 1.05 at an absorption maximum near 390 mp. Consequently, it appears that the 2-nitro-2-
alkyl-] ,3-alkanediols would interiere appreciably if present in a sample with 1nitroparaffins. ACKNOWLEDGMENT
The authors thank Clarence Clemons for obtaining the gas chromatographic data on the nitroparaffins, which permitted them to calculate the amount of impurity in each nitroparaffin. LITERATURE CITED
(1) Altshuller, A. P., Cohen, I., I d . Eng. Chem. 51, 776 (1959). (2) Browning, L. C., Watts, J. O., ANAL. CHEM.29, 24 (1957). (3) Hass, H. B., Riley, E. F., Chem. Revs. 32, 373 (1943). (4) Jones, L. R., Riddick, J. A , , ANAL. CHEM.24, 1533 (1952). (5) Rosie, D. M., Grob, R. L., Zbid., 29, 1263 (1957). (6) Scott, E. W., Treon, J., IND.ENG. CHEM.,ANAL.ED.12, 189 (1940). (7) Turba, F., Haul, R., Uhlen, G., Angew. Chem. 61, 74 (1949).
RECEIVED for review January 21, 1959. Accepted June 8, 1959.
Spectrographic Analysis of Molybdenum Metal Powder RUDOLPH DYCK and THOMAS J. VELEKER Chemical and Metallurgical Division, Sylvania Electric Products Inc., Towanda, Pa.
b The following elements in molybdenum metal powder are determined spectrographically: aluminum, barium, calcium, chromium, copper, iron, potassium, magnesium, manganese, sodium, nickel, lead, silicon, tin, strontium, and tungsten. Samples are buffered with graphite mixtures containing the internal standard. A high voltage alternating current arc i s used for all the elements except the alkalies and tungsten, for which direct current is used. The combined effect of graphite buffering and high voltage alternating current excitation produces an extremely refractory matrix during arcing, depressing molybdenum and enhancing the medium volatility elements. Tungsten is enhanced by buffering with zinc oxide. The method is far superior to wet chemical analysis in simplicity, speed, and sensitivity. It is reasonably accurate and has been applied to the control of molybdenum purity.
T
currrnt need for heat-resistant materials is creating much interest in molybdenum, not only as an alloying constituent, but also as a metal in its own right. It is a highly refractory metal that is now finding many applications in aircraft and missile technology. It is widely used for electrodes, filament HE
1640
ANALYTICAL CHEMISTRY
wire, filament supports, and high temperature heating elements. There are several major problems associated with molybdenum fabrication, such as protection from catastrophic oxidation, its relatively high ductileto-brittle transition temperature, and its microstructure after fabrication (3). The two latter problems are greatly influenced by the impurity level in the molybdenum. For example, surprisingly small additions of various impurities make sintered molybdenum rods unworkable (6). Certain impurities that have relatively high vapor pressures cause blistering on the surface of rods during sintering. Although the role of many trace impurities is not fully understood ( 5 ) , there is ample evidence that the estimation of trace elements is of paramount importance in the processing of molybdenum. Chemical determination of 16 trace elements in a matrix such as molybdenum would be difficult t o perform routinely, from the standpoint of the time and analytical skill required. The emission spectrograph is ideal for such a comprehensive analysis; however, there are problems inherent in the spectrochemical analysis of molybdenum. Molybdenum, like other refractory metals, emits an extremely complex
spectrum as well as an intense continuum under ordinary arc excitation. This results in line interferences and an unfavorable line to background ratio, and thus poses a serious problem for the spectrographer. A technique must be sought that will depress the matrix spectrum without depressing the spectra of the impurity elements. A technique in which this is accomplished is presented here. There is an analogy between this technique and a method reported by the authors for the spectrochemical analysis of tungsten ( 2 ) . The molybdenum metal powder is mixed intimately with high purity graphite (except for the tungsten determination) and arced using high voltage alternating current. For the tungsten and alkali determination, direct current is used. Excitation parameters are so selected that molybdenum carbide is formed in situ as soon as the arc is initiated. This results in a marked depression of the molybdenum spectrum while the more volatile impurity elements are enhanced. The metal powder matrix is particularly advantageous in the case of molybdenum, because the trioxide is extremely volatile and will result in poor spectral sensitivities for most elements, even if the trioxide is diluted with graphite. Other advantages in analyzing the metal powder are its greater density, and the
O!& c
-
I
1
I
1
i
Figure 1. 4
Working curves
r
I-
!
I
I
I
I
I
I
0.01
z z: w
%
E I2 LLi
u a W
0 0.001
/' /'
pY'i -2
-3
-I
0
i
I
I + i
' - I ti!
7 +3
LOG INTENSITY R A T I C
Figure 2. -3
-2
-I
0
*I
+2
+3
LOG INTENSITY RATIO
convenience of analysis at this stage of the process, thereby avoiding bothersome conversions.
200 mesh). One gram of metal powder is mixed with 0.28 gram of this graphitenickel buffer. For the determination of tungsten on the Dual grating spectre
Working curves
graph, the molybdenum metal powovr is mixed 1 to 1 with zinc oxide powdci bv volume. All samvles are mixed on L: 'Pi'ig-LBug using i-inch polystyrcni vials and 3/rinch Plexiglas balls The prepared samples are then tamped into the graphite electrodes for arcing. A smooth cup flush with the toy edge of the crater is desirabie. Sampies
EQUIPMENT AND PROCEDURE
Table I. Excitation Conditions The following equipment is used: a Bausch & Lomb large Littrow quartz Excitation Spectrograph spectrograph and a dual grating specLittrow Spectrograph trograph; high voltage Duffendacktype alternating current source, 2200 Cr, Fe, Mn, Ni, Pb, Si, and Sn volts, 2.5 to 15 amperes, built by Wave length region, 2400-3200 A. High voltage alternating current arc Sylvania Electric Products Inc.; Na%micron slit width Voltage, 2200 tional Spectrographic Laboratories Amperea, 6 I.%mm. slit height power source 110-25; National SpectroRotating sector, none Analytical gap, 1 mm. 62.5 cm., source to slit distance graphic Laboratories projection microExposure time, 45 seconde No preburn EK33 spectrographic plate photometer; Jarrell Ash developing Triplicate exposurea per sample machine; and a Spex Industries, Inc. Crescent Wig-LBug. Al, Ca, Cu, and Mg The spectral plates are processed in a Wave length. reeon, 3200-4100 A . High voltage alternating current arc Jarrell-Ash developing machine at 21' C. Voltage, 2200 20-micron slit width for 2.5 minutes using D-19 developer. l.%mm. slit height Amperes, 4 Chrome alum solution is used for 1.25 Rotating sector, a / ~o.pen" Analytical gap, 1 mm. minutes to condition the emulsion for 62.5 cm., soiirre of slit distance Exposure time, 60 seconds fast drying; thon the plates are fixed No preburn EK33 spectrographic plate in hypo solution until clear. The Triplicate exposures per sample plates are washed for 3 minutes and Dual Grating Spectrograph dried rapidly with heat. Ba, K, Ea, and Sr Procedure. SANPLE PREPARATION. All the determinations are performed Gratings 15,000-5725-7625 A. Direct current arc on the molybdenum in the metal Voltage, 300 30,00--4o(wT5000 A. Grating srptirrition, 0, superimpose spectra powder form. Three buffers are reAmperes, 7.5 Analytical gap, 4 mm. Biprism, 0.6 density on 30,000 grating quired for all 16 elements. For the Primary aperture, 0.5 mm. Exposure time, 45 seconds elements using cobalt as internal Green filter, Coming No. 3-73 between slit and No prebum standard, position 6 on the Littrow Triplicate exposures per sample biprism spectrograph, and alkali elements on 1.5-mm. slit height the dual grating spectrograph, 25 20-micron slit width parts of National SP-2 graphite 1L spectrographic plate powder are mixed with 1 part of high w purity Matthey cobalt sponge (at least 200 mesh) by weight. One gram of Grating 15,000-7oo(t9000 A. Direct current arc Voltage, 300 Biprism, 0.6 density molybdenum metal powder is mixed Primary aprture, 0.5 mm. Amperes, 5 with 0.28 gram of this buffer. Glass filter between slit and biprism Analytical gap, 3 mm. For the determination of the position 1.5-mm. slit height Exposure time, 60 aeconds 4 elements on the Littrow spectrograph, No preburn 20-micron slit width where nickel is the internal standard, SA-1 spectrographic plate Duplicate exposurea per Rsmple 27 parts of National SI?-2 graphite a Rotating sector at full open allows Myoof the incident radiation to reach the slit. powder are mixed with 1 part of Matthey high purity nickel sponge (at least VOL. 31, NO. 10, OCTOBER 1959
1641
1
I
1
I
30-60
60-90
90-120
120-150
1001 0-30
150-180
TIME- SECONDS Figure 3.
Moving plate study
901 100
0-30
Figure 4.
arc arced in triplicate for all determinations except tungsten which is run in duplicate. A set of electrodes is needed for each spectral region. The conditions for determination are giwn in Table I. I
