V O L U M E 26, NO. 2, F E B R U A R Y 1 9 5 4 Table 111. Comparison of Schoenheimer-Sperry and Aluminum Chloride Procedures for Free Cholesterol Recoveries from Serum Cholesterol Found, Mg./100 M1.
Procedure
399 color reagents. Since the sensitivity is several times that of existing procedures, it makes possible the determination of much lower concentrations of cholesterol than has been possible before and therefore promises to become a valuable investigative tool in the future.
Serum A AlClr
LITERATURE CITED 48 48 45 45 40 48 48 45 44
s. AlCl2 s. s. AlCla s. 8. AlCli s. s. AlCls s. s.
6.
43 Serum B
hlCI:
s. s. AlCL s. 8. .41Clz s. s. AlCli s. s.
AlClr 5. 9.
53 48 53 53 51 48
53 51 53 48
chloride techniques on the recovery of free cholesterol from aliquots of serum extract. Table I11 shows the comparison between the Schoenheimer-Sperry overnight precipitation of the digitonide and the more rapid aluminum chloride precipitation. The use of the iron(II1) chloride reagent for the quantitative determination of serum cholesterol has recently been shown to yield a more stable and sensitive end point than the LiebermannBurchard reaction ( 2 2 ) , which has several intrinsic difficulties. I t obviates the use of temperature control, rigid time control, dark chambers in which the reaction takes place, and unstable
Bernoulli, A. L., Hela. Chim. Acta, 15, 274 (1932). Bloor, W. R., J . Biol. Chem., 24, 227 (1916). Burchard, H., Chem. Zentr., 61 ( l ) ,25 (1890). Cavanaugh, D. J., and Glick, D., ANAL. CHEM.,24, 1839 (1952). Kenny, A. P., Biochem. J., 52, 611 11952). Kingsley, G. R., and Schaffert, R. K., J . B i d . Chem., 180, 315 (1949). Liebermann, C., Ber. dcufsch. chem. Ges., 18, 1803 (1885). Obermer, E., and Milton, R., Biochem. J . . 27, 345 (1933). Okey, R., J . Biol. Chem., 88, 367 (1930). Pollak, 0. J., and Wadler, B., J . Lab. Clin. Med., 39, 791 (1952). Reinhold, J. G., Am. J . Clzn. Pathol., 6, 31 (1936). Reinhold, J. G., and Shiels, E. AI., Ibid., 6 , 22 (1936). Saifer, 9., Ibid., 21, 24 (1951). Schettler, G., Arztl. Forsch.. 1, 232 (1947). Schoenheimer, R., and Rpriry, W. >I J. ., Biol. Chem., 106, 745 (1934). Sobel, d.E., Goldberg, AI., ant1 Slrtter, S. R., ANAL.CHEM.,25, 629 (1953). Sperry, W. hl., and Brand, F. C'., J . Biol. C h e m , 150, 315 (1943). Talsfant, E., Crechoslou. Chem. Communs., 15, 232 (1950) Tiinder, P., Analyst, 77, 321 (1952). Tschugaew, L., Ber. deutsch. chein. Ges., 42, 4634 (1909) Windaus, A,, Z. physiol. Chem., 65, 110 (1910). Zlatkis, .A, Zak, B., and Boyle. .L J., J . Lab. Clin. Med., 41, 486 (1953). RECEIVEDfor review June 17, 1953. Accepted August 29, 1953. Presented before the Division of Biological Chemistry at the 123rd Meeting of t h e AMERICAKCHEMICAL SOCIETY, Los Angeles, Calif. Study aided b y grants from t h e Rfichigan Heart Association and Edwin S. George Fund.
Method of Measuring Ultrafine Glass Fibers ROLF. K. LADISCH, BERNARD MCQUE, and STANLEY L. KNESBACH U. S. Army Quartermaster Corps., Philadelphia 45, Pa.
Pioneering Research Laboratories,
F
IBEItG with ultrafine diameters around 1 micron or less :ire the result of a new development in glass fiber production. Ent,angled masses of these fihers are characterized by low bulk densities and large surface areas of the glass. The volume of free space in these masses often approaches 100%. These unique properties are desirahle in lightweight heat and sound insulation, filtering of submicroscopic dusts, and flotxtion equipment. Most recently, thermally stable glass paper for condenser and cable dielectrics has been produced from such fihers (3'). To characterize the physical :illpemance of these superfine glass fiber masses, e l e c t r o n m i c r o scopic studies and other test methods are commonly employed, which arp based on the resistance an air stream encounters when being blown or drawn through a pad of fibers. The present study was undertaken to determine whether surface area determinations would be adequate for the same purpose, since they, in addition, would F~~~~~ indicating provide useful information n ith respect to finishing such fiber Manometer
nia~sesby coating, varnishing, or impregnating. This latter in formation cannot be obtained from the above mentioned conventionally used methode. EXPERIMENTAL
The Brunauer, Emmett, Teller method of adsorption (1, 4 ) was used to determine surface areas b means of nitrogen. An adsorpwas modifiedslightly. Thegas tion apparatus of standard design buret was replaced by a straight buret of 50-cc. capacity with subdivisions of 0.1 cc. An indicating manometer (Figure 1) was employed to maintain a fixed pressure in the system, the contact point being connected over a relay to a light signal. Both the buret and manometer were kept thermostated at 25" f: 0.1" C. The adsorption of nitrogen, a t the boiling point of liquid nitrogen, on the glass fibers was measured a t 39, 115, and 190 mm. of mercury pressure, respectively. The pressure was kept constant during the course of the adsorption by raising the mercury in the gas buret slowly, in cooperation with the light signal attached to the zero contact point (Figure 1). Ordinarily, equilibrium wa8 established within 15 minutes. Dried and purified tank nitrogen waB used. Dead space measurements were made with dried and purified tank helium. The gas contained in the dead space around the fibers was corrected in each case for deviation from the perfect gas laws. The correction made was 5% for nitrogen a t - 195.8' C. and 760 mm. of mercury pressure, the deviation being assumed to vary linearly with pressure. The volume of monomolecular adsorption was calculated in each case according to Brunauer et d. (2) from the intercept l/V,C and slope (C - l)/V,C of straightline plots. Surface areas, in square meters per gram, were calculated by multiplying the Vm values with the factor 4.38. This factor is obtained by assuming an area value of 16.2 sq. A. for the nitrogen molecule (4, 5 ) . The surface of the glass was considered
6)
400
ANALYTICAL CHEMISTRY Table I. Dead Space and Nitrogen Adsorption Values of Glass Fiber Samples Pressures of Nitrogen, Mm. Hg Pressures of Helium, hlm. Hg 39 115 190 39 115 190 39 115 190 Nitrogen Adsorbed, Dead Space, Cc. S.T.P. Total Nitrogen, Cc. S.T.P. Cc. S.T.P./Gram
Weight, Grams
Sample I 2.2032
0.760 2.229 3 . 6 6 3 0.757 2.223 3.653 0 617 1.810 2.975 0.617 1 . 8 1 0 2.975 0.792 2.324 3.823 0.792 2.324 3.823
1.8766 2,3487
2,4484 2,7900
2.3106 2.5879 2.0518
0.805 0.800
0.638 0.632 0.637 0.642
0.957 0.950 0.941 0.923
0.796 0.923 0 . 7 9 4 0.921
0 . 7 1 0 2.083 3 . 4 2 8 0 . 7 1 0 2.083 3.428 0.840 2.463 4.048 0.840 2.463 4.048
Sample I1 1 . 9 2 6 3.645 1.926 3.631 2.250 4.240 2.264 4.254
5.273 5.259 6.130 6.144
0.497 0,497 0,505 0,510
0.821 0.821 0.940 0.940 0.717 0.717
Sample 111 1.276 3 . 0 2 2 1.281 3.021 1 . 4 4 8 3.444 1,455 3 . 4 6 4 1.126 2 . 6 3 0 1.126 2 . 6 2 2
4.705 4.704 5.392 5.422 4.085 4.087
0.197 0.266 0 323 0,199 0.266 0 323 0.196 0.265 0 331 0.199 0.273 0.343 0.199 0.256 0 30.5 0.199 0 . 2 5 2 0.300
2.407 3 . 9 5 8 2.407 3.958 2.758 4.635 2.758 4.535 2.105 3.460 2.105 3.460
RESULTS
Table 11. Surface Areas and Average Fiber Fineness of Glass Fiber Samples Calculated from B.E.T. Adsorption Values Sample No. Intercept Slope
r
Vm
0.808 0.803
0.641 0.638 0.628 0.635 0.634 0.635
A , sq. meters/gram Specific gravity of glass Specific volume of glass Average fiber fineness, microns
I
I1
111
0.015 1.38 93.00 0.7168 3.14 2.55 0.39
0.021 1.73 83,38 0.5711 2.50 2.50 0.40
0.080 3.85 49.13 0.2645 1.11 2.45 0.41
0.64
1.48
0.50
i’
Previously boiled water was admitted through a buret in the top of the desiccator into the evacuated sample. Three types of ultrafine calcium borosilicate glass fibers I, 11, and 111, obtained from commercially available experimental quantities, were tested. The fibers were heated overnight a t 400” C. to remove adsorbed water and volatile impurities, then closely packed into the adsorption tube. Outgassing was accomplished by heating a t 450” C. for 2 hours in situ under a vacuum of 5 X 10-j mm. of mercury or better.
/
0.734
0,748
0.746 0.751
Table I shows the amounts of nitrogen adsorbed on the three types of glass fiber along with the dead space. Figure 2 is a straight-line plot of the a ’sorption values obtained from these samples. The monomolecular adsorption values, V,, were calculated from the slopes and intercepts from this figure (see Table 11).
DISCUSSION
Basically, the B.E.T. method permits one to measure the surface area of a finely divided material. The values found for various glass fiber sampl’es express absolute quantities in terms of surface areas per weight, and they may be directly used for an intelligent classification of these fibers. Designations such as square meters per gram or square yards per pound characterize a given fibrous material with respect to its over-all degree of fineness. A rigid quality scale can be established in this manner. Table I and Figure 2 show the excellent reproducibility of the adsorption values. The precision is better than 2% for the samples I and IT, both of which are composed of very fine fibers. As the volume ratio of dead space to nitrogen adsorption becomes appreciably greater than 1, the precision decreases somewhat. This is an inherent characteristic of the B.E.T. method. HOTever, the values are still precise t o better than 570 for fibers below 2 microns (sample I11 in Table I). On the other hand, small deviations in the degree of averzge fiber diameter between different grades of fibers are readily detected (Table 11),since the surface area of a given weight of sample increases inversely with decreasing fiber diameter. For the same reason, “shot” or relatively coarse fibers in the batch will be revealed immediately. These latter features combined with the high precision and the relatively simple and speedy technique make the B.E.T. method an excellent tool for routine and specification testing. Once a fiber type has been classified on the basis of a complete absorption run the method for routine testing becomes simple. I t is then only necessary to determine the volume of nitrogen adsorbed a t the predetermined partial presssure of monomolecular adsorption and the dead space. These two measurements take less than 1 hour. LITER4TURE CITED
0.2
0.1
P/ Po
Figure 2. Straight-Line Plot of Nitrogen Adsorption on Glass Fiber Samples I, 11, and 111
Brunauer, S.,“Adsorption of Gases and Vapors,” Vol. I, Princeton, N. J., Princeton University Press, 1943. (2) Brunauer, S.,Emmett, P. H., and Teller, E., J . Am. Chem. Soc., (1)
60, 309 (1938). (3)
Callinan, T. D . , et al., “Electrical Properties of Glass Fiber Paper,” Naval Research Laboratory, Washington, D. C., hIay
(4) (5) (6)
Emmett, P. H., Advances in Colloid ELL.,1 , 1-36 (1942). Harkins, W. D., and Jura, G., J . Chem. Phys., 11, 431 (1943). Razouk, R. I., and Salem, A. S., J . Phys. & Colloid Chem., 52,
1951.
to be essentially nonporous. The specific gravity of the glass fibers was determined by means of a pycnometer. The standard procedure was modified in so far as the water displacing the air around the sample was admitted to the evacuated pycnometer containing the fibers ( 6 ) . For this purpose, the pycnometer was placed in a desiccator and evacuated by a high vacuum pump.
1208 (1948). RECEIYED for review March 28, 1953.
Accepted September 10. 1953.
