ANALYTICAL CHEMISTRY
742 The rate of nitrogen flow through t h r apparatus should he sufficiently rapid to sweep the system completely free of hydrogen sulfide in a reasonable time. Sitrogen flow rat,es as high as 500 ml. per minute have been employed without loss of sulfide. Hexevcr, rates of 100 to 200 ml. per minute u e entirely arlequiate. Sloner rates require excessively long times to ensurr conipl(~tr transfer of hydrogen sulfide. LITERATURE CITED
(1) .kssoc. Offic. Agr. Chemists, ”Official and Tentative Methods of
bnalysis,” 6th ed., p. 127, 1946. ( 2 ) Brolvnlee, K. A , , “Industrial Experinlentation,” p. 33, Biooklyn, S . Y., Chemical Publishing c‘o.. 1949. (3) Field, E., and Oldach, C. S., IUD.Esc. CHEM.,A N ~ LF h . , 18, 668-9 (1946). (4) Fogo, J. F., and Popowsky, >I.,.\s.ir.
CHEM.,21, 732-4 (1949).
( 5 ) Ibid., pp. 734-7. (6) Liebhafsky, H. 8 . , and Winslow, E. H., IND.Ex-tical step for its determination. The emission spectrograph possesses definite advantages over chemical methods for the analysis of minute quantities of inorganic substances: its specificity, speed and economy of analysis, and high sensitivity in detecting all the metallic elements \There only a small amount of sample is available. Hence, the emission spectrograph present,s an attractive means whereby minute quantities of petroleum ash matter, such as is obtained by the combustion of a practical quantity of Ion-ash content cracking stock, may he analyzed. The use of a common matrix for sample excitation in the arc is a well-known technique in emission spectroscopy (1, 3-5, 7-10, 14, 16, 1 7 ) . The common matrix usually is prepared by blending a powdered sample in known weight rat y c de~ c ~ r i l ) ctielow tl has b w n found t o meet this requirement, n n t l in most inkt:inces this ~ n i a l lsmiple requirement makes possilile t,he rerording of rep!ic:ite Ppertra which enhance the p ~ w i s i o nof the
In adapting a rapid niethod t o routine use for thc deteriiiinatiori of microquantities of elements in a substance, :I conSidera1)le sacrifice in relative precision ma!- be necessary to gain the ipectl and simplicity required in numerous analyses for cont io1 use. T n connection with the prolilem of developing a rapid spectrographic method for determining metallic elements in minute quantities oi ash, both alternating and direct current arc excitations of t h e sample were studied in this laboratory to ascertain the sensitivity and preci3ion ahich these respective types of excitation would provide. In addition to describing the spectrographic technique, it is t8hepurpose of t,his paper t o present data xvhich evaluate the precision of the method employing alternating and direct current arc excitation for the det,ermination of several commonly occurring contaminant element,s in ash froin petroleuni cracking feed stock. EXPERIMENTAL
Oil Sample Combustion. Xpproximately 1 liter of oil sample, the weight of which is predetermined t o the nearest gram, is used in an analysis. T h e oil is added in portions to an open, tared platinum dish and allowed t o burn freely until combustion of the
743 total sample is complete. A few milliliters of C.P. xylene are us(’~1 for rinsing the container, and the rinsings also are added to the dish for combustion. Following combustion, the residue is ignited a t 1000° F. for 8 hours in a constant-temperature muffle furnace, cooled, and weighed. T h e ash residue is kept in a desiccator prior to preparation for spectrographic analysis. Common Matrix Preparation. The quantity of ash sample t o he employed in analysis is obtained by weighing it into a small glass vial t o the fourth significant place with a precision analytical Iialance. A camel’s-hair brush is used in transferring the ash from the platinum dish to the tared glass vial. S i n e parts by weight of special purity powdered lithium carbonate are added t o one part of the ash residue. h preliminary spectrographic test of the lithium carbonate should be made to assure the absence of contaminant elements which may interfere in measurenient of the sought elements. T h e mixture then is made homogeneous by t,horough grinding with a micromortar and pestle. Since most petroleum cracking stock ash residues are colored, this quality may serve as a visual aid for det,ecting nonhomogeneity in the niatrix mixture.
T a b l e 1. S p e c t r a l L i n e s 3 I e a s u r e d w i t h C a l i b r a t i o n R a n g e s of E l e m e n t s D e t e r m i n e d in Ash Residues in A d m i x t u r e with L i t h i u m Carbonate i,ine, A Element
v
Cr S i Si .11 Ca(I1) Fe Alp Sa
Li
.
1
L71
8200” 1 2R8.5 A 8391 0 2881 A 2iiCi0. 4 ? 1 i Y :3 2iXX 1 Si79 X
xme R ‘.it52
.i
Ranee. 7; I 1-10 1 2-10 1.3-10 0.3-10 1 5-26 2 7-13 1 4-35 0.3-6 0.1-22 Intensity control
Calibration Blends. The tilends for spectrographic calibration are prepared by admixing the several test elements in the t’oi,ms of their osides, sulfates, or carbonates. Fresh C.P. Uakei,’~ analyzed reagents n-ere used in this laboratory for prrpai,ing calibration blends and each compound was predried a t 105” C. for several hours prior to blending. The compounds are blended in such proportions as to cover the concentration ranges of the elements normall>- encountered in cracking stock ash residues. Calibration blends are thoroughly ground in preparation. Ranges in concentration of the respective elements for covering the calibrations may be provided by varying the ratios of the elements in the several hlends. The calibration working curves used are established hy analyzing synthetic blends prepared in the foregoing nianrier and followed is the same for calibration Apparatus. An Xpplied Resear source unit, comparator-densitom spectrograph are employed for spectrochemical analysis. Electrodes. Special high purity ‘,/,-inch graphite rods supplied by the Sational Carbon Co., Xiagara Falls, S . Y., are used for electrodes. I n the loner (anode) electrode a crater l/16 inch in diameter by ‘/E inch in depth is drilled to contain the sample. -1l/*-inch rod is used for the upper electrode. The electrodes are positioned for a 2-mm. arc gap with the optical axis a t the niidpoint of the gap. Excitation. DIRECTCURREST11 THOD. For direct current arc excitation a current of 5 ampere is employed for an arcing period of 30 seconds. -1 rotating shutt,er is adjusted to allow passage of 15Tcof the incident radiation, and a slit-rvidth setting of 40 microns is made. .%LTERXATING CURREST ~ I E T H OFor D . alternating current arc excitation a current of 3.8 amperes with an output voltage of 800 volts is employed. T h e nominal resistance is 100 olinis, th; capacitance 60 microfarads, and the discharge point control 90 , The sample is arced for 60 seconds n.ith the same spectrographic settings as are used for direct current arc excitation. If sufficient sample is available, portions of the common matrix blend normally are arced in triplicate t o record three spectra. Tn-o to 3 mg. of ash matter are an adequate sample quantity to prepare the matrix blend for three spectrograms in an analysis. Photography and Photometry. Eastman Spectrum Analysis No. 1 emulsion is used for recording the spectra and the development is carried out in a mechanically oscillating tray a t 68’ F. T h e esposed emulsion is developed for 3 minutes in D-19 developer, placed in a stop bath for 30 seconds, and then in an Eastman acid fixing solution. -1fter washing, the emulsion is dried for densitometer measurements.
ANALYTICAL CHEMISTRY
744 Film emulsion calibration is established through use of the rotating step srctor (factor of 2) by measurement of the iron emission The analytical spectral lines select,ed for n ~ e ~ s urement and the c a l i b r a h n range of each element in per cent as employed in the analysis of ash matter in this laboratory are shown in Table I. S o correction is made for background, as the background level normally is below detection in the regions of the spectrum where measurements are made.
Table 11.
Interpretation, The concentration of each element in an ash i~ found by referring intensity ratios of the spectral measurements to the above-mentioned working curves. About 2 to 2.5 hours’ time is required for the complete spectroof an ash residue for the ,Ienlents shon.rl in graphic Table I
Analysis of Synthetic Inorganic Powder Blends in Lithium Carbonate IIatriu
( A x . arc excitation-intensity ratio from each spectrum with corresponding per cent of element measured) Blend B
Blend A SpecMe- t r u m menr N o . E‘e 1 7
3 4 5 6 .4v
&-a
1 2 3 4 5 6
5
I.R.
[32.9] 1.72 30.3 1.66 29.0 160 27.3 166 29.0 1.53 25.9 1.68 29.5
0.93 0.82 1.13 0.65 1 07 0.97
1 2
3 4 5 6
Blend C
;
;.4;
7.6 3.2 7.0 6.0
0.97 08-1 0 0 I ,
71
2.0 5.0
6.0 4.8 3.8 4.2
I19 31 107 14.8 1.00 13.2 1.28 19.4 Ll5 164 1.20 17.5 1.24 185
0.53
2 4 3.0 2.6 3.4 3,.5 4.0
0.61 0.56 0.67 0 68
0.75
-
[5.6] 0.38 5.7 0 33 5.0 5.1 0.34 0.43 6.5 0 31 4.7 5.4 0.36
l5.01 0.34 5,l 0.33 5.0 0.32 4.7 0.33 5.0 0 37 5 6 0 33 5.0
13.21 0.31 4.7 0 20 3 0 0.22 3.3 0.29 4.3 0.23 3.4 0.16 2.3
3 1
3.5
r1.51 1 2 3 4 3 6
1.7 1.3 1.5 2.1 1.4 1 5
Ar.
1.6
n.41 0.20 0.17 0.16 0.17 0.14 0 17
lO.71 0.13 0.09 0.11 0.12 0.11 0.12
0.78 0.73 1.31 073 0.98 1.03
017 0.8 0.7 0.9
1.3
0.8
%
___Blend H
I.R.
%
7.9 8 0 7.4 6.7 8.0
4.2
10.6
15.9
14.2 14.2 18.0 18.6 18 2 19.9
0 19 0.5 0.81 4.6 S o detection 0.15 0.3 0 35 I.? iYo detertion
O i H
9.3 4.0
6.3 14.5 14.1 7.7 11.8 13.Y
1.00 1.70 1.68 1.13 1.48 1.6;
1.66 175
1.65 1.19 1.33 1 19
11.4
..
1.0
I.R.
[14.0] 1.10 I52 1.16 16.3 1.13 16.0 1.17 16.8 1.10 15.2 1.13 16.0
078 0.71 0.72 0.68 0.64
Synthesis, 9; CO.91 f4.71 Sodetection 0 32 0.38 0.32 0.34 0.35 0 30
D.91
---__ Blend G
[11.2] 091 11.5 078 9.2 089 11.: 081 9.l 091 11.5 083 10.3
,.
4.3 3.9, 9.< 3.9 6.1 6.7
%
I.R.
[& 21 047 4 2 043 3 9 0.30 4.6 048 4 3 031 4 7 0.41 3.4
0.G 0.5 0.5 0 5 0.5
_ _Blend __- F
7
17.01
5.8
.~
1.6 1.3 1.2 1.3 1.0 1.3
%
0.5
8.1
7
Blend E
I.R. Synthesis, % o/o
I.R.
16.6
4.3
5.4
Blend D
%
I.R.
2s.:
5.8
Ar.
‘41
%
[29.01 . 1.92 36.0 171 30.3 1.63 27.0 1.56 26.4 1.50 25.0 1.54 26.0
28.5
Ar.
Ck
I.R.
4 8 5.7 4.8 5.1
5,3
4.5
14.0 15.1 13.9 8.2 10.0 8.2 11.6
[l4.7] 0.84 13.0 0.70 10.9 1.05 16.4 0 80 12,2 1 17 18.1 0.73 11.2
5.0
13.6
Synthesis. Yo [ 8.81 [26.51 23.9 13 3 2.20 20.5 7.8 1.93 21.0 1.99 6.8 22 1 6.0 2.07 11.0 19 5 1 87 2.01 21 5 6.9
13.21 0.43 3.7 0.35 2.9 0.4.5 3.Y 0.42 3.6
8.7
1.69 1.68 1.93 1.99 1.98 2.06
0.30
4.4
0.34
2.9
21.4
3.6
7.8 5.7
[0 31 0 4 0.24 0.27 0.4 0.3 0.18 0.2 0.16 0.2 0 16 0.11 0.1
L1.71 097 2 0 0.90 1.8 1.21 2.6 1.03 2.2 0 94 1.9 0.91 1.9
0.3
2.1
..
17.2
( 5 . 9) 0.40 0.27 0.34 0.36 0.36 0.29
--
6 0 4.2 5 1 5.4 5.4 4.3
[14.71 ~. 0.76 11.8 098 15.1 0.76 11 8 1.00 15.5 0 94 14.6 0 7g 12 1
5 1
13.5 61
[0.5] S o detection
0 22 0 20 0 21 0.21 0.23 0.21
1.7 1.5 I # 1 7
.,s
1 l.,
.,
1. 7
Synthesis, ?&
DIE
1
2 3 4 5 6
12.21 0.96 0.88 1.08 1.03 0.97 0.94
BY.
Pi
1 2 3 4
5 6
hv.
h-i
1
2 3 4 5 6 AV.
Cr
2.0 1.8 2.3 2.2 2.0 1.9
[1.3] 059 1.1 0.57 1.0 0.65 1.2 0.67 1.3 0.75 1.5 0.72 1.4
2.0
2.0
1.2
6.1
0.2 0 2 0.1
Toodense for measurernent
0.25 0.38 0.40 0.33 0.29 0.38
0.2 0.4 0.4 0.3 0.3 0.4
0.38 0.25 0.24 0.44 0.41 0 35
0.4 0.2 0.2 0.5 0.4 0.3
-
0.3
0.4
11.61 1.4 0.18 1.8 0.23 0.17 1.3 0.26 2.0 1.4 0.18 1.3 0.16
Synthesis, % 13.91 0.78 5.7 0.61 4.7 0.48 3.7 0..50 3.8 0.76 5.7 0.58 4.4
1.5
4.7
1
0.26 0.21 0.19 0.23 0.31 0.27
1.3
3.0
2.51 2.22 2.07 2.25 2.92 2.28
0.2 0.3 0.2
6.5 5.6 5.1
5,7
0.83 1.11 1.38 1.13 1.08 0.79
1.3 2.0 2.8 2.1 1.9 1.2
0 2
--. 111.81 1.17 1.86 2.33 1.62 1.54 1.37
8.6 13.5 17.0 12.0 11.4 10.0 12.1
8.9 6.4 7.1 6.3 8.6 5.9
2 3 4
5 6 Av.
I 5 71
P.01 1,ll 2.4 1.00 2.1 0.88 1.8 1.00 ,2.1 0.89 1.8 0.92 1.9
7.2
0.80 0.65 1.00 0.58 0.71 0.64
1 9 Element I .
Spectrum KO.
,-
1
0.17 0.17 0.li 0.21 0.19 0 14
2 3 4 5 6 Av.
1.2 0.9 1.7 0.7 1.0 0.8
10.91 O..jO 0.38 0.39 0.36 0.44 0.41
0 9 0 6 0.7 0.6 0.8 0.7
3 4 3.9 5.4 4.6 5.3 2.9
1.0
4.2
1.2 1.2 1.2 1.6 14 0.9
Synthesis, % i2 81 0.34 2.9 0.32 2.7 0.37 3.3 0.31 2.6 0.31 2.6 0.29 2.4
1.2
2.7
11 11
12.91 1.22 1.36 1.21 1.38 1.29 1.2Y.
0.7
1.57 1.71 2.10 1.89 2.07 1.40
-
-
2.7 3.0 2.6 3.1 2 9 2 9 2.6
2.17 2.64 2.49 3.02 2.25 2.83
5 6 7.8 7.0 9., 6.0 8.7 7 5 7
[8 51
0.57 0.70 0.43 0.70 0.72 0.77
5.8 7.6 4.0 7.6 7.9 8.6 6.9
V O L U M E 24, NO. 4, A P R I L 1 9 5 2
745
DISCUSSION O F PRECISION
Efforts \yere directed to\\-nrd determining the order of precision that could he espected in using the spectrographic technique routinely for the quantitative determination of metal elements in niinut,e amounts of ash matter. T o olitain estimates of the precision of the common niatris technique, eight synthetic inorganic powders mere analyzed, eniploying both alternating and direct current arc escitation, and the resulting dat,a then were treated st.atistically t,o obtain indeses of precision for each element. The synthetic blends, similar in preparation and composition t o those used for calibration and containing varying amounts of the test elements, were analyzed hy the procedure and types of escitation described in the preceding section. The blends vere analyzed by the operator as unknown?, six spectra being obtained for each of the eight synthetic blends malyzed. The data presented in Table I1 shov.- the synthetic elemental content of each blend, the intensity rat,io obtained from the measurements of each single spectrum, and the rquivalent value in per cent of each element obt,ained b y reference of the intensity ratio to the appropriate calibration curve. The per cent values for each element obtained on the respective ztandards were used for statistical treatment. h similar tahulation of data (not shown) was obtained for the direct, current arc type escitation as applied t o the anal+ of the same synthetic blends. The method of statist,ical treatment applied to the data in a-hich the alternating current arc was employed also was applied t o the data obtained by direct current arc escitat'ion. Each set of data for R given elenlent was treated st~atisticallyas
Table 111. Precision of Spectrochemical Analyses Using Direct and Alternating Current I r c Excitation _ _ _ _ _ ~ _Concentration _ ~ _ _ ~ _Le\-el _______ 1%
SO.
of Element Spectra Fe 1 3 sa 1 3
D.C. 0 3 0.3 1.6 0 9
_
_
_
_
~
10% 30% 3% .i.C. D.C. A.C. D.C. A.C. D.C. -4.C. Standard deviarion from mean, % __
0.6 0.4 2.6 1.5
0.4
1.4 0.8
0.9
0.2 1.1 0.6
6.4 3.7
,.
0.3
3.0
0.07 0.05 0 6 0.4
3 7 2 2
2 9 1 7
2 1 1.2
..
..
1 0
7 2
0.:
Ca
1 3
, .
,.
1 3 0.7
0.2
1.7
llg
1 3
0.3 0.2
0.2 0.1
0.6 0.4
0.6 0.3
.. ..
.
AI
1
0.2
3 1
1.4 0.8 1 2
0 5
3 7 2 1 3 3 2.0
1.6 0 Y
Si
0.7 0.4 0 5 0.3
0.2 0.1
0.1 0.5 0.3
v
3 1 3
Cr
1 3
0.5 0.3
0.2 0.1 0 1 0.06
Xi
1 3
0 6 0 3
0.3 0.2
0.7 0.4 0.2 1 1
0.3 0.8 0.5 0.5 0.3
0.7
0.9 0.5
0.g 0.5
0.6 0.3
.. .. .. .. ..
..
0 h
I 7 1 0
4 2
..
.. ..
. . . 2 9 1.7
.. ..
..
follows: The variance was calculated and plottcd against the concentration mean for each of the several blends. From the plotted points a smooth curve relating concentration t o variance then \?as drawn and, in so far as the concentration range of the calibration permitted, standard deviation values were derived for selected concentration levels of 1, 3, 10, and 30%. The results thus obtained for both alternating current and direct current arc type excitation are summaiized in Table 111. For either type of excitation the precision of the method varies from element t o element, with sodium apparently shoFing the lowest precision of the nine elements studied. T h e precision for every element, evcept vanadium, is better for alternating current arc than for direct current arc ewitation. REPLlC4TE AY4LYSES O I C4TALYTIC CR4CKI1G FEED STOCK OILS
80,
I
I
I
l
l
I
l
l
-I
too LL 60 / I !
20
*O
40
6o
loo
11
TERMINAL TIMES O F IO-SECOND
\
eo
I
I
I
i
A
I
j
-yq-+&+-- - ___ 40
6o
loo
EXPOSURES
Figure 1. Time-Exposure Relationships for Spectral Lines of Elements in Lithium Carbonate Matrix
--
-A.C. arc excitation -D.C. arc excitation
An estimate of the over-all precision of the combined steps involved in reduction by burning, recovery, and spectrographic component analysis of the ash from a given oil is furnished by replicate results obtained on several samples of the same feed stock. The data appearing in Table IV are illustrative of those obtained in replicate analyses. Three spectra u ere measured t o obtain each analysis value shown. The amount of ash sample in every analysis waq more than adequate for at least three spectra, 13 t o 14 mg. of ash being available on each oil portion burned. The spectrographic data incorporate, of course, the aggregate errors of the several steps involved, which may include one or more of the following: errors of sampling due to possible nonhomogeneity or t o the introduction of extraneous contamination in the oil; errors in combustion and reduction of the oil t o ash, such as preferential loss either by entrainment or volatility, or errors due to differences in the final forms of combination of t h e metals; and the errors inherent in the spectrographic method itself. Certain components were determined chemically on a composite of portions of the ash residues. A qualitative test indicated sulfate t o
ANALYTICAL CHEMISTRY
746
Marked differences ill symmetry between the plots for a1tr.rnating current and direct current arc excitation of the respective elements are noted only for magnesium. The maximum exposure for direct current arc excitation of Mg 2780 is a t 10 seconds, whereas for the alternating current arc it is at, 30 seconds.
Table IV. Replicate Ash and Spectrographic -4nalyses on Portions of a Catalytic Cracking Feed Stock Oil (Unit 1 , March 2 , 1951) 2
1
784
784
0 0134
0 0137
Ele- 0 0051
0 0051
v
2.1
