I,VDUSTRIAL A S D E,VGI,VEERIXG CHEMISTRY
May, 1927
589
Effect of Particle Size on the Hydration of Lime' By F. W. Adams DEPARTMENT OF CHEMICAL ENGINEERING, MASSACHUSETTS INSTITUTE O F TECHNOLOGY, CAMBRIDGE, MASS.
hydration. At the beginning of a run the temperature of the hydrator was always brought to 43" C. and the quicklime and water were brought to a constant room temperature of about 25" C. The run was considered complete a t the end of 30 minutes, when the stirrer was stopped and the hydrate removed. The hydrate was placed in a sealed bottle and allowed to ripen for a least 3 days before any tests were made on it. Each sample of hydrate was tested for moisture content, rate of reaction with hydrochloric acid, settling rate, plasticity, and putty volume. Moisture content was determined by drying a sample to constant weight in COZ-free air a t 115-120" C. Whitman and Davis2 have used the time of reaction of a hydrate with successive portions of hydrochloric acid as a measure of the particle size or reactivity of the hydrate. Thus a weight of sample (3.7 grams) sufficient to react completely with 100 cc. of normal hydrochloric acid was titrated with successive 10-cc. portions of normal T IS well recognized that method of hydration of a acid. The time required for the return of the pink color of quicklime plays an important role in determining the phenolphthalein was noted after each addition. These properties of the hydrate. Thus, Whitman and Davis,* times were then corrected to show the time which would be using a high-calcium quicklime, have shown the effect of required for normal acid to react with the sample, by mulvarious amounts and temperatures of water and water vapor tiplying by the normality of the acid calculated a t each on some of the properties of the resulting hydrates. Briscoe addition. Thus to a 3.7-granl sample 10 cc. of water were and &fathers3 have discussed the effect of these factors, added, and after the sample was conipletely wetted 10 cc. particularly as regards plasticity, on the hydration of dolo- of acid were added and the time for the return of pink noted. mitic limes as well as the effect of adding alkali chlorides. The This time was multiplied by the normality of acid on the present commercial methods of hydration involve treatment sample a t the start-i. e., 0.5; similarly, after the next 10-cc. of quicklime with sufficient water to maintain the temper- addition of acid a factor of 0.33 was employed. From these ature of hydration around 100" C., at the same time yielding data the total corrected time of reaction was calculated for a relatively dry hydrate. This paper reports a study of the each 10 per cent of sample dissolved. Since reaction time effect of diameter of particle of quicklime, slaked with an is proportional to particle diameter, the total corrected time amount of water to yield a hydrate of low moisture content, gives a measure of the average particle size for each 10 per on the properties of the hydrate. cent of sample. The settling rates were determined on a 10-gram sample of hydrate shaken up with water to a total Experimental Procedure volume of 100 cc. and allowed to settle in a 100-cc. graduate A sample of a n eastern high-calcium lump lime was crushed of approximately 2.75 cm. diameter. Readings were taken and screened, yielding nine fractions varying in average a t frequent intervals during the settling. I n determining diameter between 10 and 0.1255 mm. Care was taken to the plasticity of the samples about 200 grams of hydrate prevent segregation as far as possible by carrying a lump of were mixed with water to a thick putty, which was allowed to lime to the desired mesh with the minimum possible pro- soak for 17 to 22 hours. The putty mas then brought to a duction of fines. The samples were placed in air-tight cans standard consistency and plasticity and putty volume determined by a Bureau of Standards method.' and sealed for subsequent hydration. The hydrator consisted of a n 8-inch capped section of Discussion of Results 4-inch Bronze pipe, set on its vertical axis, the charge being agitated by a stirrer revolving a t 21 r. p. m. It was found When quicklime is added to water with stirring, there is a that very satisfactory mixing could be obtained in this apparatus. Provision was made in the cover for the addition preliminary period during which the particles of quicklime of the quicklime and the steam generated was freely vented. absorb water into the pores with very slight hydration occurI n order to follow the progress of hydration, a thermocouple ring. This small amount of hydration raises the temperature was soldered into a hole drilled in the side of the hydrator and of the lime particles to a point where an increase in tempercalibrated by filling the hydrator with water and noting its ature causes a rapid increase in the rate of hydration and a temperature with a thermometer. A series of preliminary rupturing of the particles, the mixture of lime and water runs was made to determine the amount of water necessary jumping to the boiling point. Here hydration proceeds a t a greater or less rate dependent on the breaking up of the to obtain a low moisture content of hydrate. I n making a run 250 grams of quicklime were poured into initial particles. Following this period of constant temperathe hydrator, which contained the amount of water calculated ture, there is a period during which a small amount of hyt o produce a dry hydrate. Temperature readings were dration, of any previously unhydrated lime may take place, obtained by means of a millivoltmeter during the course of but a t which the heat losses by radiation from the hydrator exceed the heat evolution due to hydration, causing a falling 'Received March 23, 1927. off of temperature. Finally, the temperature drops more THISJOURNAL, 18, 118 (1926).
Samples of an eastern high-calcium lump lime with average diameter of particles varying between 10.0 and 0.1255 m m . were hydrated in an experimental apparatus to produce dry hydrates. After ripening, the hydrates were tested for moisture content, rate of reaction with hydrochloric acid, rate of settling, plasticity, and putty volume. I t is shown that the size of hydrate particles decreases with the diameter of the quicklime particles from which produced; a more reactive hydrate with lower settling rate is obtained from a finely ground quicklime. In the range studied, for all sizes of quicklime particle of 5.0 mm. or below a satisfactory plasticity to classify the product as a finishing hydrate is attained, the values running between 265 and 386. A 10.0-mm. diameter particle yields an inferior hydrate of plasticity 147. Putty volumes vary between 137 and 180 cc. from 100 grams of hydrate, following the variations in plasticity.
I
'l b r d ,
19, 88 (1927).
4
Bur. Standards. Cwc. 204.
INDUXTRlAL A N D ENG1NEERI;VG CHEMISTRY
590
rapidly owing to the cooling of the hydrator by radiation and convection to the surroundings. The temperature changes during the hydration of the various samples are shown in Figure 1, where temperature is plotted against elapsed time from the beginning of the addition of lime t o the hydrator. It will be seen that as the size of quicklime particle is decreased to 1.242 mm. average diameter (run D ) , the wetting of the particle takes place more rapidly and the /oo 90
80 \I
&e 70 u
8 60
50
40
0
4
8
/2
/6
20
24
38
32
T,mc mmufes
beginning of hydration is advanced. However, if we go beyond this point a decrease in the particle size causes a slower wetting and a delayed start of the hydration. An examination of the hydrates formed showed that in the runs using the larger sizes of particles (2.49 mm. and above) some particles of quicklime, which had not reacted during the hydration and subsequent ripening, remained. The presence of unhydrated particles where the smaller diameters of quicklime particles were employed was not noticeable. The hydrate produced in run D was found to contain 0.21 per cent of free lime; the remainder of the samples, however, had a low content of free moisture (Table I).
given size; for example, hydrate A contains 30.5 per cent of particles which are incompletely dissolved in 0.10 minute, while sample D contains only 16.3 per cent of these particles. It is also possible to compare the average particle size of the various hydrates, as particle size will be proportional to total corrected time of reaction. Thus, hydrate A is 90 per cent dissolved in 0.495 minute, while hydrate D is 90 per cent dissolved in 0.332 minute. The ratio of these times, 0.495: 0.332, is the ratio of the average particle size of 90 per cent of hydrate A to the average particle size of 90 per cent of hydrate D, and is equal to 1.49. The total corrected time at which 90 per cent of the sample is dissolved is shown in Table I and is plotted in Figure 3 against the logarithm of the average diameter of quicklime particle. These points fall on a reasonably smooth curve in which reaction time decreases with decrease in size of quicklime particle. Thus, to produce a hydrate of small average diameter of particle a finely divided quicklime should be employed. It will be noted from Figure 2 that . not only i s the size of the largest particle considerably reduced, but the percentage of particles of a size so smaIl that the reaction time is less than 0.005 minute
smallest particles present, and here one must turn to settling rates where one may study the particle distribution of the smaller sizes. Table I Av. DIAMETER OF
RUN
QUICKLIME
A B
C D E F
G
H
J
Son+-
DJssdvrd
The distribution and magnitude of particle size of the hydrates may be measured by obtaining their reaction time with hydrochloric acid as outlined above. The total corrected time has been plotted against the per cent of sample dissolved in Figure 2. It has been shown5 that the rate of solution of a solid is inversely proportional to the diameter of particle; hence, as particle size increases the time required for solution is correspondingly increased. The rate of reaction curve gives a measure of the percentage of particles above a Haslam, Adams, and Kean,
THISJOURNAL, 18, 19 (1926).
hfOISToRE IN HYDRATE
90% REACTION
DISTANCE
Per cent 5.20 1.21
PUTrY FROM in0
SETTLED IN
15
TIME &KIN. Minutes Cm. 72.8 0,495 10.00 0.416 54.0 5.00 2.49 2.40 0.269 56.2 59.2 --0.21~ 0.332 1.242 35.7 0.149 0.76 0.625 38.2 0.132 1.29 0.366 45.0 0.110 0.92 0.252 41.2 0.097 2.99 0.1773 34.0 0,072 5.36 0.1255 Indicates 0.21 per cent of free CaO in hydrate. PARTICLES
Mm.
Pcrcrnt of
Vol. 19, No. 5
PLASTICITY
GRAMS
HYDRATE
cc.
147 312 265 338 370 279 386 272 265
137 175 160
172 171 164 180 155 156
The distance settled in a given time by a solid particle decreases as the diameter of the particles decreases, and thus a measure of particle size may be obtained by plotting height of sludge against time of settling, as in Figure 4. As the larger particles of hydrate settle very quickly they are removed in less than a minute from the dividing line between clear supernatant liquor and sludge. This enables one t o study the size of the smallest particles with ease. It will be seen that the curves group themselves in a manner similar to the grouping in Figure 2, although the order is not identical in any group. The distance settled in 15 minutes is given in Table I. It will be seen that in genera1 the finer particles of quicklime give finer hydrate particles. To produce a satisfactory finishing hydrate a plasticity well over 200 is desirable. It will be seen from Table I that with a particle diameter of quicklime of 10 mm., the low plasticity renders the hydrate unsuitable as a finishing lime. I n all the other runs a plasticity figure of over 265 was obtained, although the plasticity figure apparently bears no direct relation to the particle size of the quicklime from which made (Figure 5 ) . The higher plasticities occur a t a particle size between
INDUSTRIAL AND ENGINEERING CHEMISTRY
May, 1927
591
are much coarser than those in run J, as shown by the relative distances settled in fifteen minutes. Nevertheless, the plasticity of a putty from hydrate B is 312 compared with a similar figure of 265 for hydrate J . Conclusion
01,
'
'
/o
'
m
'
'
30
Tim,
mrnvtrs
4 10
" 50
" 60
0.252 and 1.24 mm. diameter. As might be expected, putty volume follows plasticity, showing that increase in the latter is largely due to the increased water-carrying capacity of the hydrate. The two runs on the smaller sizes of particles, although yielding satisfactory finishing hydrates, do not attain the highest plasticity figures. It is evident that a high plasticity is not the result entirely of fineness of particle, but that there is undoubtedly a distribution of particle size in the dry hydrate which after soaking produces a putty of maximum plasticity. Thus hydrate B has about twice the percentage of particles of reaction time greater than 0.07 minute that hydrate J possesses; moreover, its finest particles
It has been found that in the hydration of a high-calcium lime the size of hydrate particles may be decreased by decreasing the diameter of the quicklime particles. Thus, a finely ground quicklime will yield a more reactive hydrate, possessing a lower settling rate. While a quicklime particle of 10 mm. average diameter yields a hydrate, with the low plasticity figure of 147, indicating an inferior hydrate, by , i reducing the size of g, quicMiGe particle o,ol I /a? 200 300 I 40 I 0 I =m I I t o 5.0 mm. a n d ' below. the Droduct may be classified as a finishing hydrate, plasticity values running between 265 and 386. The putty volume is found generally to follow the plasticity figure, varying between 137 and 180 CG. from 100 grams of hydrate.
mm
4
~
1
1
0
1
P/CCfIC@
High-Temperature Whitewash' By Edwin P. Arthur, W. B. Mitchener, a n d James R. Withrow THE OHIO STATEUNIVERSITY,COLUMBUS, OHIO
PPLICATION of whitewash to the clay work of industrial furnaces has not been a general practice. Whitewash, however, improves appearances and facilitates cleaning. The diffusion of light from the white surfaces materially aids in illuminating dark corners. Whitewash on the ports and flues to the regenerators reduces the infiltration of cold air. Surface temperature of some types of furnaces may approach 350-400" I?. and higher. Lime and water alone give a paint which dusts readily, has no gloss, and shows a tendency to flake off if applied thick enough to give a good color. Several washes were made and tested before applying on a large scale in an effort to find one of good color which would not crack or scale. Finely ground lime was slaked and made into a thin slurry with water. The weight ratio of lime and water to make a good paint appears variable. Other materials were added with stirring. The wash was applied immediately to the face of clean standard fire bricks with an ordinary paint brush and allowed to dry for several days. The specimen bricks were then subjected on the face opposite the painted surface for 2 hours to full heat of a Bunsen burner, or greatly in excess of service conditions. After cooling the specimens to room temperature, the surface was examined. The results are shown in the accompanying table. Washes numbered 5, 7, and 8 were good. Addition of common salt apparently improves the spreading quality of the paint, and plaster of Paris improves the covering power. Unfortunately, no opportunity has been offered to try these paints on surfaces other than fire brick, or with spray and air brush rather than the brush.
A
1
Received March 2 5 , 1927.
C o m p a r a t i v e Tests of W h i t e w a s h i n g TEST MATERIALS~ RESULTS Cracked badly 1 Lime 84, silicate of soda 16 parts b y weight
Lime 90, silicate of soda 10 parts b y weight Cracked Lime 95 silicate of soda 5 parts b y weight Scaled Lime 77: silicate of soda 8, plaster of Paris 15 parts Cracked b y weight 5 5 gallons slurry, 5 pounds salt, 5 pounds plaster Good in thin of Paris, half pint silicate coats 6 Fire-resistant paint according t o Bulletin 304B Cracked, scaled, National Lime Association (borax-casein mixoff color ture) 7 Limewater slurry, coated when dry with solution Good, slightly offcolor of silicate of sodab Lime 85, silicate of soda 5 , magnesia 10 parts b y Very good 8 weight Commercial silicate soda 40° BC., alkali ratio not determined. b Recommended by Philadelphia Quartz Company for lime kiln shells. 2
3 4
(I
RECOMMENDED WASH-Prepare a thin slurry by stirring finely ground lime into 5 gallons of water. Slowly add, with stirring, 5 pounds of salt, 5 pounds of plaster of Paris, and one-half pint of silicate soda. Apply immediately with a paint brush to brick work. Addition of the silicate too quickly or in excess will cause the slurry to become thick and useless as a paint. Such a wash was made and applied to the side walls and ports of an 800-ton regenerative glass furnace with excellent results. It was found much more desirable to apply two thin coats rather than one single heavy coat. The amount of wash required depends on the character of the clay surfaces. Acknowledgment Acknowledgment is due J. G. Callinan and R. S. Graetz for assistance in carrying out the experiments and permission of the Adamston Flat Glass Company of Clarksburg, W.Va., for the large-scale experiment mentioned.
