ANALYTICAL CHEMISTRY
956 to 3.38 microns). Bands were observed a t 3.34 and 3.35 microns, respect.ively, for aand ,+pinene.
ACKNOWLEDGMENT
Table 111. Detection and Characterization of Double Bonds in Aromatic Compoundsa -4bsorption Bands in Vinyl Hydrogen Stretching Region Cb x c € xc t xc € xC T Codeine 0.114 3.27d 3.30 67.0 3.33 45.3 Dihydrocodeine 0.114 3.27 17.8 3.30 27.4 8.33 44.0 1-Phenylcyclopentene 0,337 3.25 29.6 3.27 87.7 3.30 34.7 3-Phenylcyclopentene 0.336 3.246 3.27 65., 3.30 47.5 Phenylcyclopentane 0.326 3.24 21.2 3.27 27.4 3.30 46.2 Eugenol 0.286 3.24 26.3 3.26 26.3 3.30 20.6 3.32 40 8 Isoeugenol 0.289 3.26 21.8 3.31 56.8 3.32d Dihydroeugenol 0.314 3.24* 6.9 3.27 20.0 3.30 84.7 3.32 36.2 a D a t a which would indicate presence and nature of double bond in structure if compound were unknown are Compound
The authors &h to thank Claudia House for assistance in the preparation of some of the m o d e l c o m p o u n d s . b Concentration in millimoles per milliliter of carbon tetrachloride. Wave length in microns a t band peak. The tetrahydropyridines and d Band was obscured b y strong adjacent band. 1 , 2 d i e t h y l p i p e r i d i n e were * KO band a t this wave length. Molar absorptivity given for comparison purposes. generously supplied by Raymond Paul, I-ethyl-2ethylidenecyclohexane by J. V. Karabinos and codeine and dihydrocodeine by 9. P. Findlay. (12) Jones. R.N.,Humphries, P., Herling, F., Dobriner, K.. J . Am. C
~~
~~~
Chem. Sac. 74, 2820 (1952). (13) Jones, R. X., Humphries, P., Packard, E., Dobriner, K., Ibid., 72,86 (1950). LITERATURE C I T E D (14) Josien, AI. L., Compt. rend. 231, 131 (1950). ~ ~ lL, Jl , , .~ , ~~ ~~spectra f ,~ ~ of complex ~ ~ d ~ l o l e c u l e s , ~ ~(15) Karabinos, J. v., Ballun, A . T., J . Am. Chem. Soc. 76, 1380 (1954). Wiley, New York. 1954. (16) iY. Gash, l'. mT., Ibid.* 767 2781 (Iga). Bladon, P., Fabian, J. SI.. Henbest, H. B., Koch, H. P., Wood, (17) Lippincott, E.R., Ibid., 73, 2001 (1951). G.W., J. Chem. Soc. 1951, 2402. (18) Nueller, G. P., Fleckenstein, J. G., Tallent. W. H., Ibid., 73, Braun, J. von, Kuhn, AI,, Ber. 60, 2551 (1927). 2651 (1951). Cole, A. R. H.. J . Chem. Soc. 1954, 3807,3810. cole, A, R,H,, R ~pure ~ and , ~ ~Chem,~ ( ~l ~ ~, t4,~111~ l i (19) ~ )Paul, R., Tchelitcheff,s., Bull. SOC. h i m . France 1954, 982. (20) Pinchas, S..ANAL.CHEM.27, 2 (1955). (1954). (21) Sheppard, N., Simpson, D. S . .Quart R e t s . (London) 6, 1 (1952). Cristol, S. J., Snell, R. L., J . Am. Chem. SOC. 76,5000 (1954). (22) Smith, L. G., Phys. Rev. 59, 924 (1941). Eddy, C. R..Eianer, -1.. .XN.AL. CHEM.26, 1428 (1954). '. L., Homing, E. C., J . A m . Chem F ~ J. ~J,,,~ ~-4,E , , ~pro,-, R~ ~ sot, ~ i ( , ~ ~ ~ ~, 208 d ~ (23) ~ )Tallent, W. H., Stromberg, 1 SOC.77, 6361 (1955). (1940). (24) Wiberley, S.E., Bunce, S.C., h-u..CHEX 24,623 (1952). ~ .~,p A , , J ,~c'hellL, p~/ l y s ,18, ~861 (1950): i 19, ~ 942 (1951). , (25) 'elinsky. N . D., Ber. 2775 (lgZ5). Johnson, D. R . . Idler. D. K.,Melorhe, V. W., Bauman, C. A , , J . A m . Chem. SOC.75, 52 (1953). R E C E I V Efor D review January 28, 195G. dccepted 1Iarch 23, 1956. Jones, R . X..Herling. F., J . O r g . Chem. 19, 1252 (1954). J.l
581
Differential Spectrophotometric Determination of Beryllium 1. C. WHITE, A. S. MEYER, Analytical Chemistry Division,
JR., and
D.
L. MANNING
Oak Ridge National Laboratory, O a k Ridge, Tenn.
Differential spectrophotometry was applied to the determination of beryllium as the p-nitrobenzeneazoorcinol lake in basic solution. The absorbance of the complex is measured against a reference standard which contains 1.0 mg. of beryllium per 100 ml. of solution. The coefficient of variation of the method is less than 1% on duplicate determinations. The method is essentially free from interferences and is applicable to the determination of beryllium in the presence of moderate amounts of uranium and aluminum. Relatively large amounts of fluoride can be tolerated without interference.
B
ERYLLIUM and beryllium compounds have many unique
chemical and physical properties ( 5 , 1 5 )which have won for this element increasing applications and widespread interest. As the use and availability of beryllium have increased, many new problems have arisen in the analysis of this element, which have resulted in the development of new and modified methods of determination. The gravimetric methods ( I S ) most widely used for the determination of beryllium include the precipitation of beryllium as the hydroxide and ignition to beryllium oxide for weighing, and the precipitation of beryllium as the ammonium phosphate salt with
subsequent ignition to berylliuiii pyrophosphate (Bed'&,) for weighing. When beryllium is determined as the oxide, the solution from which beryllium hydroxide is to be precipitated must be free from all other cations that form insoluble hydroxides with ammonium hydroxide. The precipitation of beryllium ammonium phosphate with subsequent ignition to beryllium pyrophosphate has the advantage of almost four times more mass per equivalent beryllium content as in the beryllium oxide method. If the conditions of precipitation are not meticulously controlled, however, slight departures from the theoretical composition of beryllium pyrophosphate may occur, resulting in a loss of accuracy. A4volumetric method (8, I S ) for beryllium has been reported in xhich beryllium is first precipitated as the hydroxide at a pH of 8.5. If an excess of sodium fluoride is then added, beryllium is converted to the very stable, weakly ionized fluoride complex, and an amount of hydroxyl ions equivalent to beryllium is found as sodium hydroxide. This alkalinity is determined by titration n i t h standard acid and calculated as a measure of the concentration of beryllium present. The method, although empirical, IS rapid and precise under ideal conditions. Aluminum reacts in a similar manner to beryllium and will be included with the beryllium. Although it is possible to correct for the aluminum present, the precision of the method is seriously affected. Zirconium, hafnium, rare earths, uranium, and thorium interfere and must be separated.
V O L U M E 28, NO. 6, J U N E 1 9 5 6 Colorimetric reagents that have been used for the deterniination of microgram quantities of beryllium include quinalizarin ( I 1 ) curcumin ( 6 ) , alkanin ( I 2 ) , p-nitrobenzeneazo-orcinol ( I S ) , and many others ( 3 , 7 , IO, 14). This investigation was undertaken t o apply the advantages of differential spectrophotometry to the determination of milligram quantities of beryllium. Hiskey ( 4 ) and Bastian ( I ) have emphasized that the accuracy and precision which can be obtained by differential spectrophotometry are equivalent to those obtainable by volumetric and gravimetric techniques. I n the differential method the absorbance scale is set at zero with a solution of a highly colored or light-absorbing substance in the place of a reagent blank. Higher concentrations of a given component are then measured against this reference solution in the usual manner. In order to obtain the increased light intensity required to adjust the zero scale by this method, the spectrophotometer slit is set a t wider apertures than normally used. For systems which obey Beer's Ian at wider slit openings, the resultant absorbance measurements are directly proportional t o the difference in concentration between the zero point standard and the solution in question. The reaction between beryllium and p-nitrobenzeneazo-orcinol to produce a highly colored solution was used for the differential spectrophotometric determination of beryllium in this study. Application of Differential Spectrophotometry to the Beryllium-p-Nitrobenzeneazo-orcinol Reaction. K h e n a solution of p-nitrobenzeneazo-orcinol ( I S ) in 0.1S sodium hydroxide (ambercolored) is added to a moderately alkaline solution n-hich contains beryllium, a red color is formed. By carefully controlling such conditions as alkalinity and the concentration of chromogenic reagent, the relationship of the absorbance of the lake to berylliiim conctentration can be fixed. Optimum alkalinity is maintained by means of a sodium borate-sodium citrate buffer. The chromogenic reagent is almost specific for beryllium; it reacts in an interfering manner only a i t h magnesium and to a much smaller extent with zinc ( I S ) by forming colored solutions with these elements. Many metals, if present in limited amounts that would ordinarily precipitate in alkaline solutions, are rendered soluble through the formation of soluble citrate or ethylenediaminetetraacetic acid ( I S ) complexes. For systems which obey Beei's law, the accuracy which is theoretically obtainable by application of differential spectrophotometry increases with increasing absorbance of the reference standard Eventually, however, because of the combined effects of broadened band widths of light which are transmitted as a consequence of the widened slit apertures and high concentrations of the complex, an optimum concentration is reached beyond which a decrease in accuracv results. These effects are discussed in detail by Bastian ( 2 ) The concentration of the reference standard used in this study was 1 mg. of beryllium in a final volume of 100 ml. The absorbance of this solution cs. a reagent blank was approximately unity. The concentration of the reference standard could h a w been increased somewhat; however, 1 mg. per 100 ml. was chosen in order that absorbance measurements could be made with a Beckman hIodel B spectrophotometer or similar type of instrument. With this solution as a reference standard, a standard beryllium curve was prepared. REAGENTS
p-Nitrobenzeneazo-orcinol, 0.02570 wJv. Keigh out 0.125 gram of p-nitrobenzeneazo-orcinol reagent, then add t o a solution of 300 ml. of water and 25 ml. of 2,V sodium hydroxide. Stir the mixture until the reagent is dissolved (4 t o 6 hours). Transfer the solution to a 500-ml. volumetric flask, and dilute t o volume with water. (Filter if the solution is not clear.) Finally transfer the reagent into an amber storage bottle. Buffer solution, p H 12.7 0.1. Dissolve 178 grams of sodium citrate dihydrate, 92 grams of sodium borate decahydrate, and 143 grams of sodium hydroxide in about 600 ml. of water by stirring with a magnetic stirrer until all the salts are dissolved.
957 Transfer the solution t o a 1-liter volumetric flask, then dilute t o volume with water. Mix thoroughly. Sodium Hydroxide, 2N. Dissolve 40 grams of sodium hydroxide in 500 ml. of water. Beryllium standard solution, 0.100 mg. per ml. Add 0.278 gram of National Bureau of Standards beryllium oxide No. 2686, ignited a t 800" C., to a solution containing 150 ml. of water and 56 ml. of concentrated sulfuric acid. Heat until the appearance of sulfur trioxide fumes, then continue heating until the oxide is in solution. If, after 1 hour of heating, all of the beryllium oxide is not in solution, add about 100 ml. of water, then heat t o sulfur trioxide fumes again. If undissolved beryllium oxide is still present, cool the mixture, add about 400 ml. of water, and filter off the undissolved beryllium oxide. Fuse the beryllium oxide with about 0.5 gram of potassium pyrosulfate, then dissolve the fusion mixture in water. Combine the solution with the filtrate. Transfer the solution to a l-liter volumetric flask, and dilute to volume with water. The final concentration is then 0.100 mg. of beryllium per ml. in a 2N solution of sulfuric acid. PROCEDURE
Transfer appropriate aliquots of the standard beryllium solution into separate 50-ml. beakers so that they contain 1.0, 1.1, 1.2, 1.4, and 1.6 mg. of beryllium, respectively. iidjust the p H of the solutions to 5 to 6 by adding 22V sodium hydroxide. Transfer the solutions t o 100-ml. volumetric flasks, add 10 ml. of buffer solution, then 20 ml. of color reagent, and dilute to volume with water. Allow the color to develop for at least 10 minutes, then measure the absorbance of each of the solutions with a Beckman Model D U spectrophotometer a t 510 mfi in 1-cm. cells against the reference standard, which is the solution containing 1 mg. of beryllium. In order to obtain the proper slit width, adjust the galvanometer t o zero by setting the slit width a t a constant value. Make the necessary adjustments with the sensitivity control. Plot the absorbance as the ordinate against the concentration of beryllium in milligrams per 100 ml. as the abscissa. Since the reference standard must remain in the cell for some time, it is desirable t o place a cap on this cell t o avoid changes in concentration resulting from evaporation. RESULTS
The differential spectrophotometric method was tested by applying the method t o sulfuric acid solutions of beryllium and lithium that had been analyzed previously for beryllium by the indirect fluoride titration method (8, I S ) . The solutions were approximately 1N with respect to sulfuric acid. Aliquots of the standard beryllium solution which contained 1.0, 1.1, 1.4, 1.5, and 1.6 mg. of beryllium were treated along with the samples as described previously. Absorbance measurements of the standards and samples were made after allowing the color to develop for a period of at least 10 minutes. The beryllium content of the samples was calculated by comparison with the absorbance of the standard solutions. The following formula, which is based on linear relationship of absorbance to concentration over the range 1.1 to 1.6 mg. of beryllium per 100 ml. of solution, was used for purposes of calculation.
where
C, C, -4, AI I AI6 0.5
= concentration of beryllium in sample aliquot
concentration of beryllium in reference solution (1.0 mg. in this case) = absorbance of sample aliquot us. reference solution = absorbance of standard containing 1.1 mg. of beryllium us. reference solution = absorbance of standard containing I .6 mg. of beryllium us. reference solution = difference between concentration of standard containing 1.6 mg and that containing 1.1 mg. of beryllium
=
Because the volumetric, fluoride-complex method ( 8 )had been standardized against a standard solution of beryllium, the results by this method can be considered as reliable values (Table I ) . Hence, the excellent order of agreement between the two methods is indicative of the reliability of the differential spectrophotometric method.
958
ANALYTICAL CHEMISTRY
The particular method of calculation used was chosen because of the failure of the plot of the absorbance z's. concentration to go through the origin. Further tests were then made on thc conformity of the lake to Beer's law. The absorbance of the lake for beryllium concentrations of 1.1 to 1.9 mg. per 100 ml. \\ameasured daily over a period of IO days. The plot of the average absorbance L'S. concentration is shown in Figure 1. Excellent conformity to Beer's law is demonstrated for the range of concentration of 1.1 to 1.6 mg per 100 ml. Higher concentrations deviate sharply.
Table I.
Beryllium, M g . us.
1
2 3 4
5
Reference Standard 0.288 0.264 0,446 0.528 0 500
Weight Ratio of Metal to Berylliuni
Dilution Factor 13.3 13.3 13.3 12.5 25.0
Differential spectrophotometry, A 17.7 17.2 20.3 20.4 40.0
-~ -
~ Absorbance
~ ~ Found, xrg.
~Correction l l Factor i ~ Ratio Be present Be found
10
0 187 0.193 0.205 0,220
Uranium 1 20 1.21 1.22 1.24
I 000 0 995 0 984 n 972
0 0.05 1 5 10
0.194 0.195 0.196 o.2nB 0.230
Sluniinum 1.20 1.20 1.20 1.21 1.24
1.000 0.998 0,998 0.987 0.973
0 2.5 5
Determination of Beryllium
Absorbance Saniple
Table 11. Effect of Uranium and Aluminum on Color Intensity of Beryllium-p-Nitrobenzeneazo-orcinol Lake'
Beckman .\Iode1 D U , 510 mfi, 1-cm. cells: slit, 1.0 mm.; final volunie, 100 ml.; reference standard, 1.0 mg. per 100 ml.; beryllium present, 1.2 mg. Q
i x r 100 nil.
Volumetry,
B
17.7 17.5 19.7 20.5 39.7 A v . difference
Beryllium Standards (Reference Standard, 1.0 Mg. Be/100 RI1.) Absorbance Be, mg./100 ml. 1.o 0 . oon 1.1 0 105 1.2 0.195 1.4 0,354 1.6 o 5no
Difference. A - B 0.0 -0 3 +0.6 -0.1 +o 3 0 3
reagent in a volume of 100 nil. The results are shovrn in Table 111. The data indicate that a fivefold excess of either uranium or aluminum can be tolerated 15 ithout causing an error greater than 2%. By means of the appropriate correction factor, however, the interference of larger amounts of uranium and aluminum can be corrected for by multiplying the apparent amount of beryllium found by the factor corresponding to the metal-beryllium I\ eight ratio
Table 111. Effect of Fluoride on Color Intensity of Optimum Wave Length for Absorbance Measurements. Thc Beryllium-p-Nitrobenzeneazo-orcinolComplex" partial absorption spectrum of the beryllium-p-nitrobenzeneazoWeight Ratio F/Be Absorbance orcinol lake, shown in Figure 2, exhibits a broad mavimum fiom 1.45 0.0 505 to 515 mp with the center point at about 510 nip, the point 1. o 1.45 which was chosen as the optimum wave length for the absorbance 1.45 5.0 1.43 50 measurements. The existence of the broad maximum make- it 1.40 110 possible to apply a Beckman Model B spectrophotorneter, 01 its * Beryllium, 1.0 mg. per 100 nil.; buffer, 10 nil.; color reagent. 20 i d . ; reference solution, reagent hlank. equivalent, in these measurements of absorbance. Effect of Time on the Absorbance of the Beryllium-#-Nitrobenzeneazo-orcinol Lake. The color of the lake is a t its maximum intensity, as seen in Figure 3, within 10 minutes after the addition of the color reagent and is stable over a period of at least 3 hours. Subsequent tests showed that the absorbance of the lake diminished about 5% after 15 hours. Interferences. It was of interest to ascertain the extent of interference of uranium, aluminum, and fluoride ions. The effect of uranium and aluminum R as determined by adding varying amounts of the metals to two series of solutions containing a known amount of beryllium. T h r lake was formed, and absorbance measurements were made us. the reference standard that did not contain any interfering metal. The data are shown in Table I1 The interference of fluoride \vas determined by adding known amounts of sodium 10 I I 12 I 3 14 15 16 1.7 1 8 19 B E R Y L L I U M , m g p e r 100 rnl fluoride to several solutions containing 1 mg. of beryllium. The lake was Figure 1. Standard curve for determination of berjllium with p-nitrobenzeneformed, and its absorbance was measazo-orcinol by differential spectrophotometry ured us. a reagent blank containing Beckman Model DU. 510 m p . 1-cm. cells. Slit width 1.0 mm. Referenoe standard, beryllium. 1.0 mg. per 100 nil. 10 nil. of buffer and 20 ml. of color
~
V O L U M E 28, NO. 6, J U N E 1 9 5 6
959
The tests indicate that rather large amounts of fluoride can be tolerated without affecting the color of the beryllium lake. The fact that moderate amounts of fluoride produced no fading leads to the conclusion that the colored lake is more stable than the beryllium fluoride complex.
Table 17'.
Determination of Beryllium by Differential Spectrophotometry
(Calculation of coefficient of variation for each sample from difference between duplicates) Beryllium,
cz
Sainple 1
0.70
1
2 R 4 3
(i i
8 9 10 11
12 480
I
I
I
I
1
13
490
500
510
520
530
14
WAVELENGTH, m p
1:
Figure 2. Partial absorption spectrum of berglliump-nitrobenzeneazo-orcinollake
P
=
Beckman Model DU.
D
=
1-cm. cells. Final volume 50 ml. lium, 4 y per ml.
Beryl-
V = ,2
-
1- =
Precision of Method. The precision of the method was determined by analyzing a number of solutions containing about equal concentrations of beryllium and uranium. From the data obtained, the coefficient of variation was calculated for each sample type from the difference b e k e e n duplicates (9). The data are shown in Table IV.
.irerage I
8.18 8.24 8 29 6.80 K8fi 6.91 8.02 8.00 7.97 5.69 5 . Rii 5.64 6.71 6.69 6 68 8 53 8 611 8 79 8 47 8 43 8 43 8 00 8 00 8 00 8 09 8 16 8 23 8 96 9 00 8 93 3 66 3 R5 3 65 7 26 7 25 7 24 7 1.5 7 14 7 13 8 31 8 30 8 29 8 23 8 24 8 27 number of pairs of duplicates coefficient of variation, To difference between duplicates, % ZD z = -18.87 = 0.c,28 2r 30 v'\io.628 = 0.8%
Difference. 7 C
1.34 1.GO
0 02 0 88 0 43
3 00 0 47
0 00 1 72 0 78 0 27 0 27 0 28
0 24 0 36
tion shown above. There are obviously alternate methods for calculating the beryllium content. An obvious one is to refer to the standard curve, as shonn in Figure 1, which is a composite curve compiled over a period of several days. As the method is dependent for its high degree of precision on the reliability of the reference solution and the differences in absorbance of standard solutions from this reference solution, the standard curve should be re-established each time the method is used in order to achieve maximum precision. For routine usage, calculation from a standard curve which is rechecked periodically is satisfactory. LITERATURE CITED
0
I
I
I
I
IO
20
30
40
T I M E , MINUTES
"
"
180
Figure 3. Effect of time on color development and stability of beryllium-p-nitrobenzeneazoorcinol lake Beokman Model DU. 510 m p . 1-cm. cells. Slit width 1.0 m m . Beryllium, 1.5 m g . per 100 m l . Reference standard. beryllium, 1.0 m g . per 100 m l .
The coefficient of variation of the differential spectrophotometric method is less than 1yofor duplicate determinations. This order of precision compares favorably to that which can be obtained by either gravimetric or volumetric methods, and, more important, this precision is achieved n ith far greater ease and rapidity than by other methods. The results in Table IV Tyere calculated by the use of the equa-
Bastian. It.,.%SAL. CHEM.21, 972 (1949). Ibid., 23, 580 (1951). Cucri, AI. W., Neuman, W. F., Blulryan. B. J., U. S. Atomic Energy Commission, AECD-1990 (1948). Hiskey, C. F., ASAL. CHEY.21, 1440 (1949). Kjellgren, B. R. F., in "Encyclopedia of Chemical Technology." vol. 11, Interscience. S e w York, 1948. Kolthoff, I. 3I.,J . Am. Chem. SOC.50, 393 (1928). Luke, C. L., Campbell, AI. E., ANAL.CHEY.24, 1056 (1952). JIcClure, J. H., Banks, C. V., C . S.Atomic Energy Commission. AECU-812 (Sovember 1950). Oak Ridge Xational Laboratory, Oak Ridge, Tenn., ORSL Master Manual, "Statistical Treatment of Data," RIethods 10060-14,90050-14.
Sandell, E. B., IKD.ENG.CHEM.,ANAL.ED. 12, 674 (1940). Siemens-Konzern, Berlin, Germany, "Beryllium, Its Production and Application," from German by Richard Rimbach and A . J. Alichel, Reinhold, New York, 1932. Underwood, A. L., Keuman, W. F., ANAL. CHEM.21, 1348 (1948). T'inci, F. d.,Ibid.,25, 1580 (1933). White. c. E., Lowe, c. s., IND. ESG. C H E M . , .%SAL. ED. 13, 809 (1941). White, D. W., Jr., Burke, J. E., "The Metal Beryllium." American Society for AIetals, Cleveland, 19.55. R w x r v L n f o r review November 1 4 , 1955. Accepted March 23, 1986. Southeastern Regional Division, 4CS, Columbia, S. C . , Sovemher 3 t o 5 , 1955. Work carried o u t under Contract S o . W-7405-eng-26 a t Oak Ridge Sational Laboratory, operated b y Union Carbide r u c l e a r Co., division of I-nion Carbide and Carbon Corp.. for -4tomir Energy Commission.
