7 4 2 T H E JOCR,V..IL O F I , V D V S T R I A L

THE FUSIBILITY OF COAL ASH I N MIXTURES OF HYDROGEN AND WATER VAPOR’ BY A. C . FIELDNER .4ND ALEX. L. FEILD

In a previous paper by Fieldner and experi- mental da ta were given on the influence of the following factors on the softening temperatures of a number of typical coal ashes:

I--Fineness of ash z-Shape and inclination of test-piece 3-Rate of heating 4--Nature of surrounding atmosphere. As a result of this work, which was undertaken with

a view t o standardizing a method of determining the “fusing” or softening temperature of coal ash, which would yield uniform results in different laboratories, the authors made the following recommendations regarding the first three factors: ( I ) The ash t o be ground t o a powder, all of which should pass through a 2 0 0 mesh sieve; ( 2 ) the ash t o be molded, with the

’ aid of dextrin solution, in the form of a Seger cone I in. high X 3/16 in. wide a t t he base, erected with one side vertical on an infusible base made of equal parts of kaolin and alumina; ( 3 ) a rate of heating of j o per minute from a t least zooo below point of initial deforma- tion until a deformation begins, and thence 2’ per minute until the cone is down. No recommendation regarding the atmosphere surrounding the cone was made, although i t was pointed out t ha t an oxidizing atmosphere of air, free from all traces of reducing gases, was a satisfactory standard as regards mere duplication in different laboratories. On t he other hand, attention was called t o the fact

t ha t much lower softening temperatures were obtained under certain reducing conditions not unlike those which may prevail in fuel beds where clinker forma- tion takes place.

The present investigation was undertaken with the object of developing a method for determining the low- est possible softening temperature of a coal ash, and of showing whether t he slag formed under the conditions imposed by such a method is comparable t o tha t found in fuel-bed clinkers.

G E N E R A L AND T H E O R E T I C A L

I t was early apparent in the v,rork of Fieldner and Hall3 tha t the difference in the softefiing temperature of a gis-en ash when operating in various atmospheres was due t o the state of oxidation or reduction of the iron content of the ash in question. I n an atmosphere of air t he original ferric oxide of the ash entered slag formation as such, unless the temperature attained was sufficient for the thermal decomposition of this oxide t o magnetite (Fe304) ; in an atmosphere of d ry hydrogen or carbon monoxide free from carbon dioxide, a large proportion of the ferric oxide was reduced t o metallic iron and thus became unavailable for slag formation; obviously both of these conditions must tend t o give higher softening temperatures t h a n a

1 Presented a t the Atlantic City Meeting of the American Society for Testing Materials, June 22, 1915: published by permission of the Director, U. S. Bureau of Mines.

2 THIS JOURNAL. 7 (1915). 399 and 474. 3 Loc. cit .

A Y D E,VGI,VEERI,VG C H E M I S T R Y Vo1. 7 , No . 9

condition of intermediate reduction favorable to the formation of the easily fusible ferrous silicates. Since this latter condition is likely t o occur in tha t part of the fuel bed where ash comes in contact with incandescent carbon, carbon monoxide, hydrogen and hydrocarbon gases, i t is reasonable t o suppose tha t this loviest soften- ing temperature should have a closer relation t o clinker formation than either of the higher results obtained in strongly reducing or oxidizing atmospheres.

The problem then resolves itself into an investigation of the effect of various mixtures of a reducing gas and its product of oxidation on the softening temperature with a L-iew to finding tha t mixture n-hich will give the lowest softening temperature.

For experimental reasons we have chosen hydrogen as the reducing gas, since any desired mixture of hydro-

FIG. 1-EQUILIBRIUM DATA OF PREUNER A N D OF CHAUDRON ,

gen and water vapor can be readily secured by bubbling the hydrogen through mater a t the proper temperature.

The equilibrium between hydrogen, water vapor, Fe

3Fe + 4H?O J_ FeaOd + qHz (1)

and Fe304 was first studied by Deville’ a t various :em- peratures between 2 0 0 ’ and IOOO’, and later with greater precision by PreunerZ a t three temperatures. Preuner found the following mean values for the equil- ibrium constant: Temperature of solid phases. , , , . 900’ C. 1025-1050° C. 1150’ C @H?O 0 H2

0.69 0 , i 8 0 .86

Chaudron3 has recently reported two series of equilibria of iron and its oxides in hydrogen and water vapor corresponding to the following reactions:

Fe + H2O FeO + Hg ( 2 )

( 3 ) 3FeO + H?O Jr Fe301 + Hz

His values for the equilibrium constants are as follows

Reaction (21, F e and FeO Reaction ( 3 ) , FeO and FeaOi

Temperature of @H*O Temperature of solid Bhases solid phases

C. @H2 0 c. fl H? 360 440 500 550 600 630 660 760 8 i 5

0.12 0.18 0.23 0.32 0.31 0 . 3 6 0.39 0.55 0.74

440 480 500 615 i o 0 TI0 800 860 . . .

1 Compf. r e n d . , 70 ( ISiO), 1105-1111 and 1201-1205. G. Preuner, Z. physik, Chem., 47 (1904), 385-417.

a G. Chaudron, Compl. rend. , 169 (1914). 237-239.

. .

Sept., 191j T H E J O U R N A L O F I i V D U S T R I A L

No similar da ta are available on the behavior of ferric oxide (Fen&) . Hence the above da ta for magnetite include the only information we have as to the probable or possible phases in which the iron oxide of coal ash may exist in various mixtures of hydrogen and water vapor at the sintering temperature of the ash.

In Fig. I are plotted the equilibrium da ta of Preuner, and of Chaudron. Assuming tha t FesOa reacts as regards reduction in a similar manner as Fe304 i t is evident t h a t a t temperatures up t o 11 joo no reduction to metallic iron can take place in a mixture of equal parts by volume of hydrogen and water vapor. At a temperature of 8 joo, according t o Chaudron's figures, FeO would be the stable solid phase in mixtures con- taining not less than 43 nor more than 63 per cent of water vapor, it being assumed in both cases tha t a gas mixture of constant composition is passed over the heated oxide continuously, sweeping the gaseous products of the reaction out of the apparatus. I t seems possible, therefore, t ha t a minimum softening temperature of coal ash may be found in a mixture approximating equal parts by volume of hydrogen and water vapor, provided the heating is sufficiently slow t o ailow the reaction t o approach equilibrium a t the various t e m p er a t u r e s .

E X P E R I M E S T A L

FURXACE-The arrangement of furnace and sa t - urator is shown in Figs. z and 3. -4 Heraeus platinum- foil resistance furnace which happened to be available was equipped with an extra heater of 3/loo-inch plati- num wire, I t , wound on the inside of the alundum tube e and connected in series with the platinum-foil1 heating unit of the Heraeus furnace, as described in the pre- vious paper.*

This alundum heater tube was supported on two alundum rings, leaving a 'js-in. air space between the heater tube and the Heraeus porcelain tube. Two additional alundum tubes, c and 21, were placed simi- larly, one at each end of the platinum-wound tube.

k

F I G ? - ~ E R T I C A I , LONGITL-DIS .lL S E C T I O S T H R O U G H F L R S . 4 C E

Through the furnace, modified in this way, was passed a tube, A B , of fused silica, glazed on both sides, 5 / , in. inside diameter and 1 7 in. long.

The gas mixture entered the silica tube through the brass end-piece B , which was specially designed t o make a gas-tight connection by means of the brass bushing j and the asbestos packing ~t. A removable glass window, m , permitted removal and insertion of cones in the furnace, and observation of deformation, and of temperature with the optical pyrometer. The

' A nichrome heating element could a s well be used for the outer heat-

2 Fieldner and Hall, LOG. c i f . ing unit .

A N D E N G I N E E R I N G C H E M I S T R Y 743

glass was held in place between asbestos washers by a threaded brass ring. 1

An air bath (a, Fig. 4 ) was placed around the end- piece B , t o maintain its temperature above 100' C. and thus avoid any condensation of water vapor; the exit end A was made sufficiently short so tha t its temper- ature would be above 100' C. The gas mixture passed out through the glass capillary tube inserted in the rubber stopper a t A .

Alundum stops, b , f and g, were placed a t intervals back of the middle of the inner heater t o provide a suit-

f '

able background FIG. FRONT ELEVATION OF FCJRNACE A N D

with the optical pyrometer. These stops were made of thin-walled alundum tubes closed a t one end, fitting snugly in the silica tube. T o avoid rapid devitrification of the fused silica tube, temperatures were not carried

for taking readings VERTICAL S E C T I O N T H R O U G H S A T U R A T O R

above 1400~. ( i S ~ T ~ R ~ T o ~ " - T h e apparatus by which various mix-

tures of hydrogen and water vapor were formed de- pends upon the saturation of pure hydrogen gas with water vapor a t temperatures ranging from zero t o within a degree or two of the boiling point. It is shown in Fig. 3 with furnace connections as used, except the hot-air ba th surrounding the brass end- piece m and side tube 1. The saturator consists of a glass tube, e, 1 3 / ~ in. in diameter and 9 in. long, closed a t the ends with rubber stoppers. Through the upper stopper k, passes the thermometer d ; the outlet tube Y, communicating with the furnace through the brass tube 1; and a compact, adjustable vapor-pressure thermo-regulator, with electrical connections controll- ing through a relay the current in the heating coil f. The thermo-regulator is essentially tha t described by Feild,' without the refinement of being independent of atmospheric pressure. I t consists of a bent glass tube , in the short a rm of which is a small quantity of liquid u-ether, alcohol, or water, depending on the temperature desired-resting on a mercury column, c. Above the liquid u , is a small quantity of its saturated vapor, t , the pressure of which depends only upon the temperature and determines the height of the mercury column c. This mercury column. c , makes and breaks contact with the platinum wire 0. The temperature control, with a continuous stream of hydrogen bubbling around the glass beads f and through the water e , was within 0.1 t o 0.2' C.. which was sufficiently ac- curate for the purpose of the experiments. Hydrogen stored in a high-pressure cylinder was introduced a t i,

1 Alex. L. Feild, an electrical contact vapor pressure thermo-reEiilator. 1. A m . Chem. SOL., 36 (1914), 72-76

744 T H E ' J O U R N A L OF I h T D C S T R I A L A N D E N G I N E E R I N G C H E M I S T R Y Vol. 7 , NO. 9

sists of the sampling bulb a, jo mm. in diameter and 2 jo mm. long, communicating with a mercury manom- eter, b. closed a t one end. The gas stream from the heated furnace was passed for 30 minutes through the sampling cylinder t o remove all traces of air before a determination was t o be made. In order to prevent condensation of water from the mixture, the entire apparatus was immersed in a jacket of boiling water, h , heated by the coil e. The outlet tube of the sampling cylinder passes through the rubber stopper in the bottom of the water jacket, and terminates in a capillary, f . The mixture is prevented from condensing in the inlet tube, by h-

means of a heating coil, c. This tube also contains a portion drawn out t o a capillary. When a sample was to be taken, both capillaries were sealed off with a blowpipe. The boiling water was then removed, replaced by water a t room temperature, and the pres- sure indicated by the manometer observed, as vel1 as the tempera- ture of the water-jacket, after FIG, 5--APPARATUS POR

temperature equilibrium had been ANALVZING HYDROGEN obtained. Since a portion of the AND W A T E R V A P O R

water vapor condensed t o liquid water a t room temperature, the partial pressure of hydrogen a t the same temperature was calculated b y subtracting from the observed manometric pressure the vapor pressure of water for t ha t temperature. The partial pressure of the hydrogen in the original mixture a t 100' C. was calculated by multiplying the result by the ratio of the respective absolute tempera- tures. The partial pressure of the water vapor was then found by subtracting the partial pressure of the hydrogen a t 100' C., from the barometric pressure ob- served a t the time of taking the sample.

0,

MIXTURES

ANALYSES O F HYDROGEX-IVATER V A P O R M I X T U R E S

The analyses made by the method described above appear in Table I and the results are considered satis- factory.

Since these analyses, made upon samples containing large quantities of water vapor, differed from the theoretical by amounts probably within the limits of experimental error. it was not considered necessary to make further analyses upon samples containing small percentages of water vapor. Here without doubt the probability of securing a practically theoretical mixture is greater t han where the hydrogen is passed through the water in the saturator a t temperatures near the boiling point.

2 0 0 mesh was molded with the aid of dextrin solution into a slender triangular pyramid or Seger cone and mounted with one side vertical, in a refractory base according t o the method described in a previous re-

P R E P A R A T I O X O F TEST-PIECE-The ash ground t o

and its flow SO adjusted as t o give the desired current of hydrogen and water vapor through the fused silica furnace tube. The water vapor present in the mixture of hydrogen and water vapor was prevented from condensing t o liquid water on its passage t o the furnace by two heating coils. a and b , constructed of nichrome wire wound on alundum cores and heated, similarly t o the immersed coil f , by a current from a low-voltage transformer operating a t about 11 volts. An air ba th of asbestos paper (Fig. 4, a), heated by a small electric hot-plate, was placed around the brass end-piece wz, and side tube I .

ing complete saturation, the composition of the mixture of hydrogen and water vapor obtained from the satura- tor depends upon the temperature of the water in the saturator. The relative proportion of hydrogen and water vapor present is proportional t o their respective partial pressures. The partial pressure of mater vapor

T E S T I X G THE EFFICIESCY O F T H E SATURATOR-hSUm-

F I G . 4-ARR.4NGZslERT O F FURRACE A X D ACCESSORIES

a t various temperatures is known. Hence, t ha t of hydrogen can be calculated-the sum of the two being equal t o atmospheric pressure plus a negligibly small pressure required to force the mixture through the furnace tube.

The apparatus' shown in Fig. j was designed for the analysis of the mixture of hydrogen and water rapor . These analyses were made t o determine how closely the composition of the mixture corresponded t o the theoretical, calculated from the temperature of the saturator water. The device was attached to the outlet end of the vitreosil furnace tube d (Fig. j ) . It con-

1 Suggested by G. A. Burrell.

Sept., 191 j T H E J O U R N A L O F I N D U S T R I A L A N D ElVGINEERING C H E M I S T R Y 745

TABLE I-ANALYSES OF HYDROGEN-WATER VAPOR MIXTURES AVERAGE VOLUME

PERCENTAGES OF HYDROGEN A N D WATER TEMPERATURE OF WATER IN SATURATOR Mano- Vapor VAPOR AS DETERMINED B Y CAI,CUI.ATION FROM

CORRECTION FOR Barometer meter Temp. pressure Experimental da ta Temperature of saturation Uncor- Superheated reading reading ‘C. of water Water

Exp. No. rected Scale steam Corrected Mm. Mm. (a ) a t (a)”C. Hydrogen Water vapor Hydrogen vapor

i:i z g iz:!] 5 2 . l W 4 7 . 9 ( a ) 50.6 49.4

15 .4 84.6

” 1 9 . 0 91 .0 10.9 89 .1

I1 81.0 -0.16 -0.19 80.65 736.9 I2 . .

I11 9 5 : O . . . . , . 94:48 735:O iii $: $:::] 19.2 80.8 112 1111 sJj:o . . . . 96:48 7 s i : 2 84 29 1 1 1 2 , . . . . . . . . . 84 29 . . I

(a ) CALCULATION F O R 1 1 : 323-19.. . . . . . . . . . . . . . . . . . . . . . . . . . 22‘ c. Then 304 X (3 i3 f 295) . . . . . . . . . . . . . . . . . . 1 0 0 c C .

... 384 + 737 . . . . . . . . . . . . . . . . . . . . . . . . Vapor

port.’ On account of the small diameter of the fused temperatures of final deformation. The initial defor- silica tube, it was possible t o use only the upper half mation points are plotted in a similar manner in Fig. 7 . of the standard 3,’16 inch by I inch cone which was It will be noted tha t both sets of curves have well- adopted in the former experiments. The dextrin defined minima in the neighborhood of 4 0 t o i o per cent binder was burned off before placing the cone in the water vapor. The final deformation points, which were furnace by ignition a t a red heat in a muffle. more accurately determined, show a fairly uniform

XEASURELIEKT OF TEMPERATURE-The temperatures minimum in mixtures ranging from 3 0 t o 70 per cent TABLE 11-DESCRIPTION AND ORIGIX OF SAMPLES AND ANALYSIS OF THEIR ASHES

LOCATIOX O F hlINE LAB. No. DESCRLPTION Bed Seares t town County

15848 Bituminous Thompson Marvel Bibb 16586 Semi-bituminous Paca. h-0. 3 Big Four McDowell 15844 Bituminous Minga Fork Ridge Claiborne 16243 Bituminous N o . 5 Boonrille a‘arwick 15845 Bituminous Coal Creek Oliver Morgan

were read with a Wanner optical pyrometer, which was sighted on the thin alundum disk just back of the cone. Correction for absorption of the glass window was made as described in the previous paper.’

The pyrometer was compared daily with the standard amyl acetate flame, and once each week a thin piece of Kahlbaum’s pure copper foil was substituted for the ash cone and its melting point read in an atmosphere of hydrogen. The results were between 107 j and 1 0 9 0 ~ C. (31. P. = 1083’C.)

E X P E R I 51 E N T A L R E S U LT S

Softening-temperature determinations were made upon five different samples of’coal ash (see Table 11) containing from 7 to 69 per cent of ferric oxide (Fe703) , in various mixtures ranging from pure dry hydrogen t o pure water vapor. The temperature was raised a t the rate of I O to I j” per minute t o approximately goo“. then jc per minute until deformation began, and z o per minute thereafter until the cone was down. The dry hydrogen was obtained by bubbling the hydro- gen through a wash bottle of concentrated sulfuric acid. The IOO per cent water-vapor atmosphere was obtaincd by cutting off the hydrogen from the saturator and heating the water to boiling.

All the results obtained are given in Table 111, in which the softening temperature is given as a temperature interval, the IoTver figure before the hyphen being the temperature a t which deformation (rounding of the apex or bending of the cone) began, and the higher figure following the hyphen being the temperature at which the cone had bent over so tha t the apex touched the base, or, failing t o bend, had softened t o a lump. This latter final deformation point is meant when refer- ence is made t o the “fusion,” “softening” or “cieforma- tion” point of an ash.

The final deformation points are plotted in Fig. 6 , in which the abscissae represent the percentages of water vapor in the mixture and the ordinates the

1 Fieldner and Hall, LOC. c i f .

State SiOz A1203 Fe?Os TiOz CaO LIgO NaL) Kz0 SO1 Ala. 54.8 27.0 7 . 0 1 . 3 4 . 3 1 . 7 0 . 3 3 . 1 1.4 W . V a . 3 7 . 2 25.5 11 .8 1 .5 12.6 1 . 9 1 . 4 0 . 4 5 6 Tenn. 4 2 . 2 3 0 . 6 1 9 . 0 1 . 2 1 . 3 1.0 1 . 3 2 . 9 0 . 2 Ind. 3 7 . 1 1 7 . 6 3 5 . 9 0 . 7 3 . 2 0 . 9 0 . 4 1 . 8 2 . 1 Tenn. 12 .3 12.2 69.7 0 .4 3 . 9 0 , 7 0 . 3 0 . 6 0 . 2

water vapor. This is a fortunate circumstance. as i t enables the formulation of a practical method of de- termining the lowest softening temperature of an ash

TABLE 111-SOFTEKING TEMPERATURES IX VARIOUS ~ I I X T U R E S OF HYDROGEN A N D WATER X-APOR

Lab. No. 15848 16586 15844 1624.3 15845 YG FerOs 7.0 11.8 19.0 35.9 69.7 Atmos. ‘3 Hz0 SOFTENING TEMPERATURES, OC.

0 . 0 1322(a)-1360(b) 1390-1402 Above 1400 1304-1370 Above 1400 0 . 7

6 . 3 . . . 7 . 6 . . .

1 0 . 7 . . . 12.4 . . . 15.8 . . . 18.1 . . . 1 9 . 9 . . , 29.6

4 1 . 7 , . . 51.2

7 7 . 2 . . .

5 . 7 i27Gi290

31 .2 117511271

57.5 i2201i292

1313li360 A b o i e 1400 . . . . . . . . . . . .

1340 1 z 74-i 3 1 3 I 1 5 5- 1 2 i 1

1331-1355

122C-1270 . . . Above’ 1400

i248Li263 . , ,

. . . . . . . . . 130413 330

i i o 4 1 i i 9 8 . , ,

83.2 095 83.9 1 1 7 i I i 2 9 j i 2 4 i l i z 4 9 i 2 z i l i 2 6 3 1206-i 350 9 0 . 3 . . 1227-1251 1206-1295 (cjlio95 , , , 9 4 . 8 1048-1080

100.0 1356-1i370 13131i322 1400 1304-1313 .Abo;e’ 1402 (a ) Beginning of deformation. ( b ) End of deformation or “down” point of cone. (c) Starting point not observed.

in a mixture of approximately equal parts of hydrogen and water vapor. Apparently the gas mixture need be controlled only between limits of 30 and 7 0 per cent water vapor to insure the lowest softening temperature within the unavoidable experimental del-iation in determinations of this character.

The pronounced effect of viscosity on the rate of deformation of the ash cones is shown in the relative positions of the curves of initial deformation in Fig. 7 and the curves of final deformation in Fig. 6. Sample 16 j 8 6 has the highest initial-deformation temperature and next t o the lowest final-deformation temperature in a j o : 50 mixture of hydrogen and water vapor. This is due to the low viscosity of the melting eutectic formed from this ash, which contains 12.6 per cent of calcium oxide, a larger percentage of this basic oxide than occurred in any of the other samples.

The sharp rise of the softening-point curves at both the oxidizing and reducing end confirms the reasons

746 T H E J O U R N A L OF I N D U S T R I A L A N D ENGINEERING C H E M I S T R Y Vol. 7 , NO. 9

ATMOSPHERE - PER CENT

FIG. 6-EFFECT OF VARIOUS MIXTURES OF HYDROGEN A N D XvATER VAPOR O N THE SOFTENING POIXT (FIKAL DEFORMATION) OF COAL ASH

advanced in a previous paper1 for the large variations in results obtained in different furnaces. I t is evident t ha t no concordant results can be obtained in oxidizing atmospheres containing small proportions of reducing gas or similarly in essentially reducing atmospheres with small admixtures of oxidizing gases, since in these regions the curves have a sharp inclination.

R E L A T I O P : O F S O F T E K I K G T E M P E R A T U R E IS ATMOS-

P H E R E S O F H Y D R O G E N A K D W A T E R VAPOR

T O C L I K K E R F O R M A T I O N U N D E R F U R N A C E

C O X D I T I O S S

The general form of the softening-point curves in hydrogen and water vapor suggests a choice of three atmospheres under which uniform softening points may be determined, v i z . : ( I ) IOO per cent hydrogen; ( 2 ) jo per cent hydrogen and 50 per cent water vapor; ( 3 ) IOO per cent water vapor or IOO per cent air, which gives nearly the same result as is shown in Table IV. TABLE 1x7-COMPARISON OF SOFTENINO POINTS IN AIR AYD WATER VAPOR

Softening point in air Water vapor LAE. KO. 3/16 in. X I in. cones 8 / 3 1 in. X I/? in. cones

35845 1458’ C. Above 1400’ C. 15844 1490 Above 1400° C. 15848 1425 1370 16243 1335 1313 16586 1322 1322

The atmosphere t o be selected for a standard soften- ing temperature test should be tha t one which produces a slag with the iron in the same state of oxidation as is usually found in fuel-bed clinkers. Determinations were therefore made of the relative percentages of ferrous, ferric and metallic iron in the glassy portions of furnace clinkers and in the ash cones fused in the three above-mentioned atmospheres. F E R R O U S , F E R R I C A S D M E T A L L I C I R O S I N F U S E D A S H

C O N E S A K D C L I K K E R SLAGS

Twelve t o fourteen ash cones, aggregating about z grams, were placed in a platinum boat, and heated in the silica tube of the furnace a t the usual rate up t o the softening temperature of the particular ash in the particular atmosphere in question. The boat and its contents were held a t this temperature for 30 minutes t o allow the softening reactions t o approach equilibrium.

1 Fieldner and Hall. LOC. cit .

1400

1350 + 5 g 13w

5 1250

0

Y

c

2 ; 1200

g 1150

E 1100

+ z

B

i 5 1050

loo0 watervaWro 10 20 30 40 50 60 70 80 90 100 nyarOgen IM) go 80 70 60 M 40 30 20 IO o

ATMOSPHERE-PER CENT FIG. ?-EFFECT OF VARIOUS MIXTURES OF HYDROGEX A N D WATER VAPOR

O S THE TEMPERATVRE O F I N I T I A L DEEORXATIOP OF COAL As11

As in the case of softening-temperature determinations, the cones were heated to redness previous t o the ex- periment t o oxidize and remove the organic binder. At the end of the go-minute heating period the elec- trical current v a s cut off and the furnace allowed t o cool rapidly, without interrupting the gas stream. I t mould have been better to avoid any possibility of change during cooling t o have quenched the fusion suddenly. However, the viscosity of these ash melts was so great t ha t little change was probable a t the rate of cooling used. The fused material appeared glassy.

Fusions n-ere made of sample 16243 in IOO per cent hydrogen, jo : j o per cent mixture of hydrogen and water vapor, and IOO per cent water vapor; also of sample I j 8 1 8 in 5 0 : jo hydrogen-water vapor, and sample 13629 in air.

-4fter cooling the furnace, the fusions were removed from the boat and were pulverized, first b y impact in a diamond mortar of hardened steel, and then in an agate mortar, to IOO mesh, and analyzed for ferrous, ferric and metallic iron by the method of Mathesius.’ The results given in Table 1’ show well-defined dif- ferences in the state of oxidation of the iron content of the fusions in the various atmospheres. I n hydrogen, 78 per cent of the iron was reduced t o metal; in steam and air approximately 7 0 per cent remained as ferric iron; in the j c : 5 0 per cent hydrogen-water vapor mix- ture, 80 per cent or more was present as ferrous iron.

Together with these analyses of fused ash cones are given similar analyses of the fused and glassy portions of several clinkers produced under working conditions in the fuel beds of different boiler furnaces operating with Pittsburgh coal. The coal, however, came from different mines. Clinker 20137 was obtained from the grate of a small hand-fired experimental furnace which was used for combustion investigations. Clinkers 20145 and 2 0 4 5 2 were kindly furnished by Mr. A. A. Straub, steam engineer of the Duquesne Light Compaqy of Pittsburgh. These clinkers were dumped from the fuel bed while hot and were quenched with water. Both of these clinkers were of t he particularly trouble-

1 Ludwig Mathesius, “Untersuchungen iiber die Redusierbarkeit van Eisenerzen in Stromenden Gasen.”

747

Percentage Max- 01 total iron imum Total occurring as

temp. of Fe Metal- Fer- F?r- I A B . iurion Per lie f0"S .r*c hTo DBSCPIiPTrON 0 C. cent iron Iron nron CH***CTEL os_ SLAC 15848) Ash fused in atmosphere of SO% hydro- 1280 4 . 7 6 94 Trace 1.1glii-groy Color Not magnetic.

16243 i ( 1080 2 7 . 4 Tncc 82 18 Glairy black. Not magnetic. ih243 Ash fused in atmosphere of hydrogen. 1370 2 8 . 6 78 21 I Bisck,mcialliclu~tter:contained mrgneticpartieles of metalliciron. 36243 Ash fused in atmosphere of water vapor. 1300 2 7 . 9 Trace 28 72 CIlasry black. suitace had reddish tint; strongly magnetic. 13629 Ash fused in atmosphere of air. 1400 I S .S Trace 30 7 0 Dark red to ahnost black. glarny. highly, magnetic. 20137 Clinker fmm a hand-tired furnace. . . 14.5 12 79 9 Dark gloss: some piece3 slightly magnetic. 20145 Clinker from a boiler furnace equipped

204.52 Ilitto, Brunof's Island station. . . 8 . 9 9 75 16 Black glass, slightly mametic.

sen : 50% water YhPOT. 1

with Roney stoker. 20th Street sintion. . . 12.7 6 75 19 Brown to black glass: some pieces slightly mr~nelic.

some taffy-like variety t h a t flow over the grat,e bars hydrogen and water vapor, ranging from io0 per cent and shut off tbe air suppiy for combustion. Fig. 8 is hydrogen t o 100 per cent water vapor. a photograph of clinker z o 4 j z . The straight-line edge 11-These results piotted in the form of curves is the side which adhered t o the corrugated grate of the showed t h a t for each of t he ashes tested, there was a new model type 1) Roney stoker. T h e matrix of this high softening temperature in pure hydrogen on one clinker was a black glass which could be readiiy chipped end, due t o reduction of iron oxide t o metallic iron; a out in quant i ty for analysis without including a n y similar high softening' temperature in water vapor or piece of cokc or unfuscd material. ,4t t h e contact air on thc other end, due t o the iron oxide remaining surface of coke and slag, particles of metallic iron coul(j for the most par t in the form of ferric iron or mag- be iilentified. A microscopic examination ol t he pow- netite; and a more or less lower softening temperature derrd sla,g from this ciinker as well as from the others, in the middle portion ranging from 3 0 t o 70 per cent by MI. A. A . Klcin of the Bureau of Standards, shoxed water vapor, due t o the reduction of iron t o the fer- the material t o lx: essentially a riass with sillimanite rous s ta te in which i t combined t o form the readilv

I ---

fusible ferrous silicates. 111-Analyses of actual clinkcr slags from two dif-

ferent boiier furnaces and one experimental furnace showed tha t fuel-hcd conditions are such as t o favor tlie iormation of clinkers containing iron principally in the ferrous state.

IT'---A new method oi determining the minimiim softening temperature of coal ;ish has been devised, in which the ash is heated i n an atmosphere of approxi- matcly j o per cent hydrogen and j o per cent mater vapor, whereby the iron oxide is caused t o combine principally in the ferrous s ta te , as actuaily found in fuel-bed clinkers.

Fir; ~ . - - - C L ~ N X ~ K A.0. 20.452 FHOM GnalE or' R O S B U S T c ~ a E i i

A C L K O W L S D G I E N ' T S (A1103, SiOl) as the oniy crystalline phase present,

iron mineral, probably magnetite. analyses of the fused ash cones,

and of the fucl.hed clinkers gi-i.en in ~ ~ b l ~ 17,

sho.rvs that of the (lificrent in the iaijorzitory furnace, jo : jo per cent i,ydrogcn- water vapor mixture produced a slag which corrc- sponded most closely in the s ta te of oxidation of its iron content t o the glassy portions of the furnace clinkers. In both cases approximately 80 per cent of the iron appeared as ferrous iron, the form in which it imparts the maximum fluxing action on the silicate mixture.

except sampic 2 o r 3 7 wllici, aiso co,ltained Bn opaque The authors take pleasure in acknowledging their indebtedness t o Dr. G. 4. Hulett, chief consulting chemist, for many heipful suggestions and criticisms; t o A. E. Hall for assistance in building the furnace and in taking the preliminary observations; and t o H. N. Hill for thc careful analysis of the slags and clinkers.

,I comparison of

C I I T M ~ C A L I,*BORATOBY, R u n Z ~ i i OF MzNBJ FllTS.IIRCII. P A .

.

PONTiANAK (JELLUTONG) RUBBER RESIN

Reccivcd April I?. 1915 ny C I ~ I . E I O N Z L L ~ AND A. A. WT.LI.S

Wc may, thcrefore, conclude t h a t t he minimum soft- Pontianak ox Jellutong rubber resin has appeared eiiing temperature of a coal ash, as determined in a jo : 50 on the market in relatively large quantities during the per cent atmosphere of hydrogcn and water vapor, is past few years and no little effort has been made t o more representative of the temperature of clinker find profitable outlets lor this material. A t t he formation under furnace conditions than the more or present t ime, for various reasons, the available supply less higher results obtained in strongly reducing atmos- of the resin has been much reduced. pheres of hydrogen, or carbon monoxide, on the one Jellutong is obtained l age ly from Sarawak, Dutch hand, and oxidizing atmospheres of air, water vapor Borneo, Sumatra and the Maiaysian Peninsuia. The or carbon dioxidc on the other. product from Saravak was the first t o be termed

S U Id H A R Y Pontianak and the name subsequently was applied t o I-Softening temperatures of five diffcrent coal substantially similar products f rom the other localities

ashes have been determined in various mixtures of mentioned.