Determination of Specific Surface of Sieve-Size Powders BERNARD DUBROW and MARY NIERADKA Pyrotechnics Chemical Research Laboratory, Picatinny Arsenal, Dover, N. J. developed by Blaine is a simplification of the Lea and Nurse equipment. Another apparatus based on the air permeability prim eiple, which is suitable for rapid routine operation, wa8 developed by Gooden and Smith (4). The method described in this paper for measuring the average particle size of materials in the sievesize range is based on the Gooden and Smith method. The air permeability measurement of specific surface, which, if desired, can he expressed as average surface diameter, is based on the relation between the rate of flow of air at low pressures through packed beds and the external surface of the packing. In carrying out the measurements, three properties of the packed bed of powder are measured: its physical dimensions, the pressure drop across it, and the rate of flow of air through it. T h e air permeability apparatus uses a manometer to indicate pressure drop and a flowmeter to measure the rate of flow of air a t constant pressure. For materials having an average surface diameter between 2 and 50 microns, the precision with which differences in manmetric readings can he made with existing equipment i s higher than the over-all reproducibility of the test. This, however, is not the case for materials in the sieve particlesize range. For these comser materials the reliability of the flowmeter readings decreases and, consequently, results obtained for these samples are in doubt, owing to the inadequacy of the flowmeter. In order to extend the range of t.he existing equipment it ia necessary to incorporate a more sensitive and adjustable flowmeter. In addition to the flowmeter it appeared desirable, far measuring coarse materials, to increase the sample size and to alter the packing technique.
Instruments available for measurement of specific surface by air permeability have heen generally utilized for materials in the subsieve range. Thia work extends the range of air permeability measuvements to sievesize powders. An instrument which employs a sensitive adjustable flowmeter was developed. The accuracy and precision of the method indicate that tho iostrument can readily he adopted for rapid routine memurements of specific surface or average particle size in the range investigated.
T""
q RE are two widely adopted air permeability methods for the memurement of specific surface: a method developed by Gooden and Smith (4),and the Blaine method (1). These methods are satisfactory when applied to subsieve powders. The sensitivity of these measurements, however, decreases considerably for materials having averbge surface diameters greater than 50microns. A need existed for an air permeability method with a wider range and greater sensitivity, because for many applications of coarse powders the specific surface, rather than the mean diameter determined by microscopic count or by sieve analysis, is of fundamental importance, especidy where chemical reactions are involved.
APPARATUS
l'hotographs of the apparatus that was derreloped are shown in Figures 1 and 2. The line diagram; Figure 3, shows the essential experimental parts of the equipment.
Figure 1.
Front view of particle-size apparatus
T o meet this need, a method that extends the utility of the air permeability method to povders in the sieve particle-size range was developed. This method i s simple, rapid, and readily applicable to routine measurement. GENERAL AIR PERMEABILITY METHOD
T h e theory of the permeability method and details of ohperiments conducted with liquids have been discussed by Carman (a, 8). Based on Carman's theory, Lea. and Nurse ( 5 )developed a rapid air permeability method for measuring specific surface of subsieve powders. T h e commercially available apparatus
Figure 2.
302
Back view of particle-size apparatus
V O L U M E 27, NO. 2, F E B R U A R Y 1 9 5 5
303 A weighed sample of dry powder equal in grams to an integral multiple of the true density of the material is poured into the sample tube containing a porous plug. The sample is lightly tapped, and inserted in the apparatus (Figure 4). With the air regulated to permit approximately two or three air bubbles to escape through the standpipe per second, a calibrated glass diffusion tube is selected so that the pressure reading, F , is approximately 0.5 the initial pressure, P . The manometric deflections, F and P, are recorded. The sample tube is removed, and the height of the sample, H , is measured with a modified depth gage. CALCULATION OF PARTICLE SIZE
The average surface diameter of the powder can be calculated by means of the Gooden and Smith equation ( 4 ) . Figure 3.
Line diagram of equipment
Air is introduced into the apparatus through a reducing valve a t approximately 2-pound gage pressure, then passes through a drying bottle which is connected to both a standpipe and a sample tube. The standpipe, filled with water, serves to fix the pressure of the air a t the point of introduct,ion t o the sample. The escess p r e s s u r e is released a s bubbles escaping through the water. The air a t this c o n s t a n t p r e s s u r e flows downward through the sample tube in which the powder to be tested is contained. A porous plug supports the powder in the tube. After leaving the s a m p l e t,ube, the air is divided along two paths: Figure 4. sample tube through an air resistance assembly t u b e , and into a water manometer. The air resistance tube is a fritt,ed-glass diffusion tube for which the rate of air passing through it has been determined by means of a wet-test meter. The manometer and the resistance formed by tht. air resistance tube toget,her constitute a flowmeter. If a snmple of powder is added to the sample tube, the difference in pwssure as read on the manomet,er, with and without the sample in the apparatus, will indicate the resistance of the powder sample to a constant air flow. Before the particle size of a sample can be determined, calibration is necessary.
where p
K H A P F A-
=
surface weighted diameter, microns
= permeability constant, ~ m 2, .of ~air resistance tubes = sample height, cm. = cross-sectional area, eq. cm., of sample tube =
initial air pressure, grams per sq. em.
= exit air pressure, grams per sq. em. = weight of sample per density
Nomograms, based 011 the Gooden and Smith equation, have been constructed to simplify tlte calculations (Figures 5, G , and
'
F
0
- 50
F1 50
b
145 r
'lSE
L 45
140
35
F
F30
F40
E
CALIBRATION PROCEDURE
Calibration of the apparatus consists of determination of the permeability constants of coarse, medium, and fine glass diffusion tubes used as flowmeter resistances. The calibration procedure is schematically illustrated in the line diagram (Figure 3). The wet-test meter is connected to the bot,tom end of the air resistance tube. With the empty sample tube inserted in the air permeability apparatus, air is passed through the system. When air is observed to pass through the wet-t,est meter a t a constant rate, the initial volume reading, VI, and the time of t'he reading, [I, are recorded. The deflection of the water in the manometer indicating the air pressure, P, in the system is read directly. After 1 hour the volume reading and the time of the second reading, T-2 and tz, respectively, are recorded. The permeability constants, Ii,of the air resistance tubes are calculated by means of Equation *.
1;
[(:;It,;;]
K = 1.263 X IO3
c CALCULATION OF AVERAGE PARTICLE DIAMETER (FA. AIR PERMEABILITY APPARATUS)
~
T E S T PROCEDURE FOR DETERMIh-ING AVERAGE PARTICLE SIZE
To determine the surface diameter of a sample the procedure is as follom:
P5 Figure 5.
Nomogram for determination of P
ANALYTICAL CHEMISTRY
304 Table I. Particle Size of Atomized Magnesium Determined by Three Methods Magnesium Screened Fractions, Mesh 16/20 20/30 30/40
40/50
50/70 70/100 100/140 140/200 200/325
Screen .4nalysis,
Microscopic, p
Air Permeability,
P
P
1000 705 498 353 260 177 125 88 57
990 703 510 350
742 620 451 346 233 170 116
360 112
370 112
255
182 118 85 61
79
Production Samples so. 1
s-0 2
379 101
____~ i ) . P, equated to d F / P - F , is obtained using nomogram I
(Figure 5 ) . 01, equated to K H S i ( 9 H - 5 ) 3 /where 2 , S is 10, can be determined by means of nomogram I1 (Figure 6). 01 and p , when aligned in nomogram I11 (Figure i ) ,give a value in microns corresponding t o the average particle size of the material tested. DI SCU SSlOiY
The accuracy and precision of the permeability method depend to a great extent on the preparation of a uniform bed of material. For coarse samples it was found necessary to pour and tap the material in the sample tube. This technique gives uniform packing, nhich eliminates the possibilit? of fracturing coarse friable salts. A larger sample of pon-der than is ordinarily
used in permeability measurement of particle size was selected in order t,o provide a larger total surface and, thereby, increase the sensitivity of the method. Filter paper disks, which usually are employed in the commercial Blaine and Gooden and Smith methods to contain the sample, were not used, because they x e r e found to offer significant resistance to the air flow, compared with the resistance of the coarse sample. In the present design, plugs covered with wire mesh were used in place of filter paper. I n order to determine the accuracy of the method, a series of comparative test,s was made for sieve-size atomized magnesium powder having R mean length diameter ranging from approximat,ely 50 to 1000 microns. Results of the average particle size as determined by screen analysis, microscopic count, and t,he air permeability method, are given in Table I. As the magnesium particles are generally spherical, the diameter of the individual particles could be measured microscopically viith a filiar micrometer. The geometric mean of 100 particles for each samplc is recorded in Table I. The sieve analysis geometric mean values of the production samples were obtained by plot,ting on logarithm probability paper the per cent of the sample greater than a particle size corresponding to a sieve opening against the size of the sieve opening. The mean value was obtained by reading the SOYo size. The geometric mean of each sieve fraction was obtained by averaging the openings of the upper and lower sieves. The experimental trend that Tvas established indicates that t,he surface diameters were generally equal to or lower than the geometric mean length diameters. This is in agreement with the expected theoretical trend. If the average mean length diameter of the magnesium par-
cc
K i300
E
:1t
2cc
i.
j
.:t
:
~ , O O i90 js0 i-70
n
!
~
a:POROSITY p :PRESSURE
C O i F F I C I E N T: .C
RATIO C0EF"ClENIT p = A V E R A G E P4RTICLE S I Z E MICRONS
1
I 7
5~
Figure 6 . Nomogram for determination of
Figure 7 . 01
for N = 10
Nomogram for determination of average particle diameter
V O L U M E 27, NO. 2, F E B R U A R Y 1 9 5 5 ticles is considered constant, any deviation of the particles from sphericity would increase the surface per unit weight of material. Thus, the addition of irregular particles t,o a batch of perfectly spherical particles would increase the average surface of the batch, and consequently decrease t,he average surface diameter per unit weight. I n view of this, the air permeability geometric mean and the mean obtained from sieve analysis and microscopic count should be equal if t.he particles are perfect spheres. If the particles are not spherical, the air permeability geometric mean should be less than the mean length diameter. Visual examination of the coarse fract,ions indicated that the particles had indentations and were not as smooth as the fine fractions. The coarse fractions, as expected, gave lo^ air permeability values. The discrepancy in values of surface and mean length diameters for the coarse fractions is not an indication that the air permeability method is less accurate than the microscopic or sieve analysis methods. It only emphasizes the fact that different physical dimensions are considered. As chemical reactions are more dependent on the surface of powders than on diameters, the air permrahilitl- values, which consider the surface of materials, are preferred. Table I1 compares the values of average diameter obtained by the three cxperimental methods for barium nitrate. Since barium nitrate is crystalline, the air permeability values are expected to he lower t,han the screen anal!-& and microscopic vdueq. Sonteclinicd persoiltiel could he readill- trained to tieternline the average size of a pondered sample. The tinic required for a single det,erniination is less than 10 minutes. Thc avcrage reproducitiility of the results is less thnti 5% over
305 the range of 50 to 1000 microns. Three times the standard deviation was taken as a measure of reproducibility. Below 50 microns the reproducibility decreases.
Table 11. Particle Size of Barium Nitrate Determined by Three Methods Production Samples No. 1 No. 2
Screen Analysis,
Microscopic,
P
P
Air Permeability, P
580 315
650 340
49 5
303
ACKNOWLEDGMENT
The authors are indebted to David Hart for his cooperation and guidance, and to Henry Eppig and Garry Weingarten for their contribut,ions. George Kottler, Howard Keyes, Paul Gallien, George Sichols, and Ray Carlstrom aided in the preparation of the nomograms and diagrams.
*
LITERATURE CITED
(1) Blaine, I