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 E N G I N E E R I N G C H E M I S T R Y
V o l . II. JUNE, 1910. No. 6
T h e Jo u r n a l o f I n d u 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
PUBLISHED BY
T H E A M E R I C A N C H E M I C A L S O C I E T Y .
BOARD OF EDITORS.
E d ito r : W . D . R ic h a r d s o n . A ssociate E d itors.
G eo . P. A d a m s o n , E . G . B a i l e y , G . E . B ar to n , W m . B r a d y , W m . C a m p b e l l, F . B. C ar p en te r , V i r g i l C o b le n t z , F r a n c i s I. D u p o n t , W . C. E b a u g h , W m . C . G e e r , W . F . H il l e b r a n d , W . D . H o r n e , L . P. K i n n i c u t t , A . E . L e a c h , K a r l L a n g e n b e c k , A . D. L i t t l e , P. C. M c l l l i i n e y , E . B.
M c C r e a d y , W i n . M c M u r t r i e , J. M e r r i t t M a t t h e w s , T . J.
Pa rk er , J. D . P e n n o c k , G e o . C . S t o n e , F . W . T r a p h a g e n , E r n s t T w i t c h e l l , R o b t . W a h l , W m . H . W a l k e r , M . C.
W h i t a k e r , W . R . W h i t n e y .
Pu blished m onthly! Subscription price to non-m em bers o f the A m erican C hem ical S o ciety $6.00 yearly.
Vol. II. ' JU N E, 1910. No. 6
EDITORIALS.
U S E S F O R R A R E E L E M E N T S A N D S P E C I A L C O M P O U N D S .
I t frequen tly happens th a t investigators have p u t aside and regarded as im practicable, the suggested use of m any of the more difficultly obtained chem ical elements and compounds, in consequence of the fa ct th at some of these are a t present regarded as unavail
able— whether such u n availability is the result of either a v e ry lim ited present supply, or a cost -which is prohibitive for the purpose for which the material is desired.
M anufacturers are throw ing aw ay m any materials w hich are, a t present, looked upon as waste prod
ucts, and would g lad ly welcome a n y suggested use for the same. Through the cooperation of the in
vestigator and the m anufacturer, m utual difficulties m igh t be overcome. W ith a more intim ate coopera
tion, the requirements of the investigator m ight be supplied b y utilizing the w aste products of the m anu
facturer.
T h e fundam ental step tow ards obtaining this co
operation m ust be the securing of inform ation con
cerning the availab ility of the m aterials needed and the uses which m ay be m ade of them.
W hen w e consider the marvelous im provem ents recently- brought about b y the application of elec
tricity to chem ical industry,, and the accom plishm ent
thereby of results hitherto unobtained, w e m ust be- prepared to revise sçm e of our inform ation concerning, m aterials previously regarded only as “ museu specim ens,” and place m any of these in the catego of substances practical for industrial purposes.
T w enty-five years ago aluminum was not com m ercially available; to-day, millions of pounds are used in the industries. O nly a few years ago, the element silicon was regarded as a curiosity; to-day it m ay be purchased b y the ton— uses having been found for it and convenient methods devised for its preparation. The same m ight be true of m any other elements and compounds, which, a t the present time, have b u t little application.
T h e D ivision of Industrial Chemists and Chemical Engineers now has a com m ittee which is endeavoring to deal w ith this problem. This com m ittee will probably act in the cap acity of a clearing-house between the m anufacturer and consumers. The com m ittee is composed of former President W . R . W hitn ey, of Schenectady, N. Y ., M. C. W hittaker, of Gloucester, N. J., and Gellert Allem an, • Chairman, of Sw arthm ore, Pa. T h e members of this society are asked for suggestions regarding the scope of the investigations to be undertaken, and we hope that all w ill aid in this im portant work.
T h e com m ittee proposes to publish as com plete a list of the prices and availableness of the chem ical elem ents as it can obtain, and would be glad to get inform ation to add to this list. I t will later probably tak e up relatively novel or newly available com pounds. Such new compounds as come to the notice of members w ill receive attention b y the com m ittee if th ey are informed of them. G ellK R T Al l e m a n.
O R IG IN A L P A P E R S .
A COMPARISON OF THE METHODS FOR DE
POSITING IRON ELECTROLYTICALLY.
B y S. A. Tu c k e r a n d E. Sc h r a m m. R eceived April 1. 1910.
The experim ents herein described include the sul
phate and chloride baths as these have been used b y m ost investigators. O ther electrolytes have been tried w ithout much success, such as ferrous fluor- silicate, but it was considered best to restrict the experim ents to the more promising m ethods and, b y com paring them under certain conditions, to decide which would be the m ost practical for the deposition of m etallic iron.
238 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 E N G I N E E R I N G C H E M I S T R Y . June, 1910 The first solution tried was m ade b y dissolving
200 grams of F eS
0
4.(N H 4)2S0
4.6
H20
in one liter.This gives about 29 gram s of iron per liter, or a con
centration of 2 .9 per cent, of iron. T h is is the solu
tion as used b y A m berg,1 and probably b y Burgess2 and Hambuechen. The cathodes were clean sheet iron and the anodes consisted of wrought iron -bars.
The experiments wrere carried o ij a t room tem pera
ture and the results are given in T able I :
Ta b l e I.—Ir o n.
Volts. Amp. Amp.t]dcm - Amps.O'. Lbs. kw.
0 .5 0 0 .2 0 .1 7 6 1.63 4 .4
0 .5 6 0.2 5 0.22 2.02 3 .9 4
0 .5 8 0 .3 0 0 .2 6 4 2 .4 5 3 .7 6
0 .6 4 0 .3 5 0 .3 0 8 2 .9 6 3 .4 4
0.68 0 .4 0 0 .3 5 2 3 .2 7 3 .2 1
0 .7 0 0 .4 5 0 .3 9 6 3 .6 8 3 .1 2
0 .7 4 0 .5 0 0 .4 4 4 .0 9 2 .9 5
0 .9 0 0 .7 5 0.66 6 .0 4 2 .4 2
1 . 1 0 1.00 0.88 8 .2 6 1 .9 8
The current efficiency was determined b y inserting a copper coulombmeter in series and gave an effi
ciency of 95.5 per cent.
The last column in this table shows the effect of increased current density on the energy efficiency.
The current was changed tem porarily to obtain these figures. In com puting the figures for lbs.
per lew. hr., the current efficiency w as assumed to be the same throughout.
The deposit obtained was silvery white and of a fine crystalline structure. A fter a while it became b adly pitted. The electrolyte kept in condition for a long tim e; the basic precipitate which formed b y oxidation of the air was allowed to collect in the bottom of the vessel. The electrodes were 1 1/2"
apart and the current density a t the anode w as about double that a t the cathodes.
A nother run was made w ith the solution and all conditions kep t the same as the above excep t the current density, which was raised to 1 amp. per sq. dcm. T he deposit was crystalline and silvery white like that obtained a t lower current densities, bu t showed a greater tendency to form nodules.
T h e current efficiency was nearly 100 per cent., but the voltage drop was 1 .2 , giving 1.70 5 lbs. per kw.
hour.
A third run was made with ferrous ammonium sul
phate solution, w ith the addition of a small q u an tity of glycerine. The deposit in this case was smoother and more solid.
R y ss,3 in his investigations, used a solution con
taining 200 grams ferrous ammonium sulphate, 50 gram s M gS
0
4, a n d .4 gram s N aH C0
3 per liter. R y ss claimed that such a solution gave a sm oother and thicker deposit of iron than could be obtained with ordinary ferrous ammonium sulphate.1 Zeit. E le k tr o c h e m 14, 326 (1908).
2 Trans. A m . Electrochem. Soc., 5 , 201 (1904). Electrochem. and M et.
In d ., 2 , 183 (1904).
* R yss, Z eit. Elektrochem., 12, 697 (1906).
A trial failed to show a n y im provem ent in the character of the deposits using such additions, but it did show th at the bath had a much greater tendency to form basic precipitate and the tim e in which it could be m aintained in good condition was less than
th at possible w ith the simple bath.
The n ext solution tried was one containing 30 grams F e S 0 4.7H20 , 21 gram s N a2S
0
.,.iûH20
per 100 cc.as described b y K e rn .1
A current density of 1 amp. per sq. dcm. was used.
T h e deposit w as dark gray and formed large nodules.
A t the end of 15 hours the electrolyte w as in bad condition owing to the form ation of a h eavy precipi
tate.
T h e net results of the experim ents w ith different sulphate solutions were the conclusion th at a solution containing 200 gram s per liter of M ohr’s salt was the best as regards character of deposit, perm anent good condition of electrolyte, and energy efficiency. A tte n tion was n e xt turned to the electrolysis of chloride solutions. T he first solution used contained 100 gram s FeCl,.4H 20 and 50 gram s N H ,C
1
per liter.The cathodes, as before, being of clean sheet iron and the anodes, wrought iron bar. The electrodes were 1 1/2" apart and the bath maintained a t room temperature. T he current efficiency as determined b y a copper coulom bometer was found to be 96.5 per cent, and the following data was obtained:
Tim e of run 2 1 1 / 2 hours.
A verage current during run = 0.237.
A verage voltage during run = 0.42.
W eight of iron actually deposited, 5 .1 4 grams.
W eight of iron theoretically deposited, 5.32 grams.
Current density during run = 0.304 amp. []dcm.
No. w a tt hours = 0.237 X 21 1/2 X 42 = 2 .14 . 1 w a tt hour = 2 .4 gram s i on.
1 kw . hour = 5.28 lbs.
T h e current w as varied tem porarily to obtain the following readings:
Volt. Amp. Amps. [ ] dcm - Amps. [ ] f t * Lbs. kw . hr.
0 .4 0 0.2 0 .2 6 0 2 .4 2 5 .5 5
0 .4 4 0 .2 5 0.321 2 .9 8 5 .0 5
0 .4 6 0 .3 0.3 8 5 3 .5 8 4 .8 3
0 .5 4 0 .3 6 0 .4 5 0 4 .1 8 4 .1 0
0 .5 7 0 .4 0 0 .5 1 3 4 .7 6 3 .8 9
0 .6 0 0 .4 5 0 .5 7 7 5 .3 6 3 .7 0
0 .6 0 0 .5 0.6 4 2 5 .9 7 3 .7 0
0 .7 2 0 .7 5 0 .9 6 2 8 .9 4 3 .0 8
1.00 1.00 1.280 11 .9 0 2. 2 2
In the n e xt run a solution of double the strength was used, i. e., 200 gram s FeC l2.4H20 and 100 grams N H 4C
1
per liter. T h e other .conditions were kep t much the same as before, excepting th at the length of tim e of the run was increased:Tim e of run, 141 hours.
A verage current during run, 0.323 amp.
A verage voltage during run, 0.67.
W eight of iron actually deposited, 4 7 .14 5 grams.
1 Kern, Trans. A m . Electrochem. Soc., 13, 103 (1908).
F R A N K F O R T E R O N S T U D I E S I N A S P H A L T .
239
W eight of iron theoretically deposited, 4 7 .6 grams.
Current efficiency = 99 per cent.
Current density = 0 .5 6 amp. []dcm . _ ^ 2 am ps.[]'.
No. w a tt hours = 0.323 X 141 X 0 .6 7 = 30.5.
x w a tt hour = 1 .5 5 gram s iron.
1 kw. hour = 3 .4 lbs.
T h e follow ing readings were ta k e n :
T h e following readings were ta k e n :
Current density.
]ft.
V olt. Am p. A m p s .[ ] dcm ' A m p s.fJ '. Lbs. kw . hr.
0 .4 2 0.2 0 .3 4 5 3 .2 5 .4 2
0 .5 0 0 .2 5 0 .4 3 2 4 .0 1 4 .5 5
0 .6 4 0 .3 0 .5 1 7 4 .0 8 3 .6 6
0 .7 0 0 .3 5 0 .6 0 4 5 .6 4 3 .2 6
0 .7 2 0 .4 0 .6 9 6 .4 3 .1 6
0 .8 2 0 .4 5 0 .7 2 6 6 .7 5 2 .7 8
0 .8 4 0 .5 0 .8 6 2 8 .0 2 .7 1
0 .9 0 0 .7 5 1 .2 9 1 2 .0 2 .5 3
0 .9 8 1.00 1.725 1 6 .0 2 .3 2
V olt. Am ps. Tem p. Amps.[]
0 .4 2 1 .4 M3 o o Ü 15
0 .4 6 1 .4 50° 15
0 .5 4 1 .4 4 0 ° 15
0 .6 4 1.4 3 0 ° 15
0 .3 6 1 .8 6 0° 19.3
0 .3 4 1 . 6 6 0° 11.15
0 .3 2 1.4 60° 1 5 .C
0 .3 0 1 . 2 6 0 ° 12.85
0 .2 8 1 . 0 6 0° 10.72
0 .2 6 0.8 6 0° 8 .5 7
0 .2 4 0 .6 6 0° 6 .4 3
0.22 0 .4 6 0° 4 2 .9
0.20 0.2 6 0° 2 .1 4
T aken a t start.
A comparison of the two runs given above w ith the run on ferrous ammonium sulphate solution shows th at w hile the current efficiency is practically the same in both cases, the energy efficiency a t any given current density is much higher for the chloride solutions. Com paring the two chloride solutions, we see th at the more concentrated solution gives a higher energy efficiency a t a n y given current density.
T h e physical character of the deposit from the chloride solution was som ewhat different from that deposited from the sulphate solutions, being more crystalline than the latter excepting that to which glycerine had been added.
T h e n e xt electrolyte tried was the ferrous sodium chloride solutions described b y E . F . K e rn .1 This w as m ade b y dissolving 285 gram s FeC l2.4H20 and 102 gram s N aCl per liter, giving 8 per cent. F e and 4 per cent. Na. T h e solution m ay also be m ade b y dissolving iron in the equivalent w eight of H C
1
. The cathode used was thin sheet iron plate 3 3 / 4 " X 3 1/2".T h e anodes were bars of wrought iron 3" X 2" X 1/2".
T h e electrolysis w as carried on in a beaker placed on an asbestos pad w ith a hole cu t in it, and heated b y placing an incandescent lam p underneath. This k ep t the tem perature between 500 and 700 C.
Previous runs having all shown a high current efficiency of over 95 per cent., it was not considered necessary to use a copper coulom bm eter in this run.
T he o bject of the run was t o :
(1) Compare the voltage required b y this process w ith the others described.
(2) T o find the effect of heating the electrolyte on the energy required.
(3) T o find the effect of heating on the physical character of the deposit.
(4) T o secure a thick deposit in a short time.
T he cathode area was 1 3 .1 5 square inches.
T h e electrolyte was k ep t in condition b y adding a little H C
1
to the bath every d a y to dissolve the basic precipitate which formed. W ater was fed to the b ath autom atically to m ake up for evaporation.1 Trans. A m . Elecirochem. Soc., IS , 103 (1908).
9 days after start.
T he above tables show: that the voltage decreases w ith increased tem perature; and that the voltage decreases as the run proceeds.
Comparison of this process w ith the others tried shows the best energy efficiency for the ferrous so
dium chloride electrolyte w orking a t 50 “-7 0 ° C.
T his bath also gave the best and thickest deposits, the iron being gray and non-crystalline.
El e c t r o c h e m i c a l La b o r a t o r y, Co l u m b i a Un i v e r s i t y.
STUDIES IN ASPHALT, I.
B y C. J . Fr a n k f o r t e r. R eceived March 23, 1910.
W hile engaged in testing asphalt for the C ity E n gineer of Lincoln, N ebraska, it was observed th at the tem perature of the asphaltic m ixture, as brought to the street, varied as much as xoo° C. m axim um to minimum. I t was also noticed that the per cent, of bitum en was b y no means constant. This seemed a little unusual, and the author decided to experim ent and determ ine if possible the effect of such varia
tions on the paving material.
Literature was consulted, b u t nothing bearing on the w ork in mind was found. P av in g specifications from Chicago, Brooklyn, N ew Orleans, and several other large cities, were obtained and consulted. W ith out exception, these specifications condemned the overheating of an asphaltic cement. Ju st how hot such m aterial m ight be heated w ithout injury, or w h at the harm ful effects of overheating really were, w as not stated. In all cases the per cent, of bitum en was to be reasonably constant, usually about xo per cent, of the surface m ixture was called for.
Sam ples of Trinidad, Texas, Cuban R ock, Obispo and Sarco asphalts were collected. R esults on two samples only, Obispo and Trinidad, will be presented, since the d ata on the other brands are as y e t insuffi
cient to w arrant definite conclusions.
In this paper the author wishes to be distin ctly understood as neither recommending nor condem n
ing the asphalts mentioned.
Obispo refined asphalt was a hard substance re
sembling coal tar in general appearance, but it did n ot have the “ ta r r y ” odor nor w as it “ s tic k y ” a t
240 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 E N G I N E E R I N G C H E M I S T R Y . June, i g i o ordinary temperatures. The physical properties of
this m aterial varied w ith the tem perature as recorded b elow :
A t 0 ° C.
From 5 -1 0 ° C.
From 2 0 -2 5 ° C.
From 5 5 -6 0 ° C.
From ’9 0 -9 6 ° C.
From 115-120° C.
From 135-140° C.
From 155-160° C.
From 180-190° C, A t 2 2 2 -2 2 4 ° C.
A t 2 3 5 -2 3 8 ° C.
Very brittle, broke like glass Slightly less brittle
Less brittle, enough ten acity to prevent crumbling
Softer, very tough B egan to m elt Consistency of molasses More mobile, luster dim inished Very mobile
A pparently no change Flashed
Burned
X o odor.
N o odor.
N o odor.
F ain t oily odor.
Stronger odor.
Sam e odor.
Stronger odor.
V ery strong odor.
Strong rancid odor.
T h e Obispo flux w as a ve ry viscous greenish-black oil having an odor similar to crude petroleum. Its specific g ra vity w as 0.9894. I t flashed at 13 5 -13 7 0 C. and burned a t 161-162 0 C.
A11 asphaltic cement is the m aterial which, when m ixed with approxim ately 90 per cent, b y weight of sand, constitutes the wearing surface of an asphaltic pavem ent. In the case of a natural asphalt contain
ing incorporated native mineral m atter a larger proportion would have to be used in order that the m ixture m ight still contain the required 10 per cent, of bitumen.
A sphaltic cem ent is prepared b y fluxing a refined asphalt with some oil, either natural or the petroleum residues from oil refineries. T h e fluxing is carried on until the asphaltic cem ent reaches the desired con
sistency, which is determined b y testing its ductility and penetration. A fter consulting the various p a v
ing specifications, a penetration of 40 to 45 at 250 C. was decided 011 as a standard.
In all these experim ents a standard D ow penetra
tion machine was used for the penetration tests.
F or determ ining ductility, the author constructed a device w ith which the material was tested under w ater at any desired temperature. This apparatus consisted of a zinc-lined w ater-tight trough, 2 meters long, 13 centim eters wide, and 7 centim eters deep.
The inside was given a h eavy coating of white enamel paint. This w hite background enabled the operator to determine the ex a ct moment of rupture of the asphalt. In one end of this trough ■ a small hook was fastened, in the other a small pulley. A reel, m anipulated b y an ordinary crank handle and ca rry
ing a light strong cable, was set up over the trough on a framework near the pulley. This cable ran over the pulley and reached to the opposite end. A standard Dow mold w as used in this machine, one^
end hooked to the end of the box, the other to the cable. W hen the cable was wound up on the reel the mold was pulled apart. On the bottom of this trough two strips, 1 centim eter high, were fastened ju st far enough apart to perm it the mold to slide between them. T h ey were graduated in millimeters and Served as a means of guiding the m oving erid of the mold, thereby ensuring a straight-aw ay pull.
T h ey also perm itted the result of a test to be read off directly. W hen a test was to be made, the trough was nearly filled w ith w ater of the desired tempera-, ture, o°, -25° or 46° C. The mold containing the asphalt was placed in position and allowed to remain tw enty or th irty minutes. T h e w ater was stirred constantly in order to m aintain a uniform tem pera
ture in all parts of the trough and to bring the asphalt to the tem perature of the water. T h e reel was then turned until the asphalt pulled apart and the dis
tance it stretched read off directly.
Penetration is recorded in degrees Dow, equivalent to hundredths of a centim eter. D u ctility is recorded in centim eters. B oth d u ctility and penetration were determined according to D ow ’s directions.
H eating the asphalt to a high tem perature causes a loss of volatile m atter which is accompanied b y a decrease in d uctility and penetration, though in no definite ratio.
T he three tem peratures a t which these tests are made are supposed to represent approxim ately the different tem peratures a pavem ent would be obliged to withstand.
W hen the Obispo refined asphalt was heated for 15 hours a t a temperature of 16 3 -16 50 C. in an air bath, as prescribed b y the paving specifications, it lost 2 .12 per cent, in weight. The samples weighed 20.0526 and 20.0812 grams, respectively; There w as no moisture in this asphalt.
T able I gives the d uctility and penetration figures on Obispo refined asphalt, showing the per cent, of loss due to the heating which caused the above-m en
tioned loss of w eight:
Ta b l e I . — Ob i s p o Re f i n e d As p h a l t.
D u ctility. Penetration.
A t 0 ° C. A t 2 5 ° C. A t 4 6 ° C. A t 0 ° C. A t 2 5 ° C. A t4 6 ° C.
B efore heating 0 .2 5 .2 2 0 .0 2 .0 9 .0 3 3 .0 4 .2 0 .0 0 .5 5 .0 ' 19.0% 100 .0 £ 9 4 .4 4 # 8 4 .8 4 # A fter heating 0 .0 0 .2
L o s s . . . . 100 ,0 # 9 6 .1 5 #
B y fluxing 102 grams of Obispo refined asphalt w ith 42 gram s of Obispo flux; an a sp h a ltic cemfent was obtained which had a penetration of 45 a t 25°
C. T h e d u ctility could n o t ' be accurately deter
mined, as the m aterial persistently adhered to • the am algam ated brass p late upon which the mold was poured.
F or heating the asphaltic cem ent an open-air bath arrangem ent was used. I t was constructed in a w ay w hich'allow ed the asphalt container to rest on a false bottom , and this served to keep it outr o f direct con
tact w ith the flame of the burner. I t was possible to stir the m ixture a t all times. Since a thermom
eter was used for this purpose, i t w a s e a sy to keep a ve ry close w atch on the temperature.
T able II shows the variation of d u ctility and pene
tration of the Obispo asphaltic ceineiit after being heated to different tem peratures: . ■
F R A N K F O R T E R O N S T U D I E S I N A S P H A L T . 241
Ta b l e t I I .-t- Ob i s p o As p h a l t i c Ce m e n t.
D u ctility . Penetration.
A t A t A t At A t. . A t
0 ° C. 2 5 ° C. 4 6 ° C. O o O
25°>C. 4 6 ° C.
H eated to 130° C . . . 0 .9 3 3 .5 6 4 .6 7 .0 3 9 .0 106.0 H eated to 150° C . „ . 0 .7 , 2 7 .2 4 7 .8 6 .0 2 9 .0 102.0 H eated to 175° C . . . 0 .4 1 9 .0 4 3 .0 5 .0 2 5 .0 7 3 .0 H eated to 2 0 0 ° C . . . 0 .1 1 5 .0 4 0 .0 4 . 0 2 1 . 0 68.0
The ductility here recorded 011 Obispo is relatively very high. Trinidad and Berm udez, the so-called best asphalts, did not show such high figures. There was this difference. In the case of the two latter asphalts the “ b o d y ” of the mold stretched and seemed to be “ rubbery.” W ith Obispo the m aterial pulled out to a fine thread, and- this thread stretched out until it was so fine it could scarcely be seen even against the white background. T h e “ b o d y ” in this case showed no tenacity or ductility.
Trinidad refined asphalt w as a fairly hard grayish black m aterial w ith no luster and an odor like freshly plowed earth. A t different tem peratures it behaved as follow s:
A t 0 ° C. B rittle, crumbled rather easily N o odor.
From 5 -1 0 ° C . ' S lightly brittle, n o t “crum bly” F ain t earthy odor.
From 2 0 -2 5 ° C. Less brittle, n o t “ crum bly’' Stronger odor.
From 4 5 -5 0 ° C. N o apparent change Sam e odor.
From 6 5 -7 0 ° C. B egan to soften Sam e odor.
From 102-1 0 8 ° C. E ntirely m elted Disagreeable odor.
From 120-125° C. Consistency of m olasses, showed
stringy or fibrous when stirred Sam e odor.
From 150-155° C. Sligh tly m obile, odor similar to ' a m ercaptan; (Later qu alita
tiv e test showed the presence of b o th nitrogen and sulphur.)
From 170-1 7 5 ° C. V ery m obile V ery strong odor.
(I t is very probable th a t th e in corporated mineral m atter pre
ven ted this asphalt from show ing th e same degree o f m ob ility as the Obispb, at approxim ately th e same tem perature.) A t 2 0 8 -2 1 0 ° C. Flashed
A t 2 3 3 -2 3 5 ° C. Burned
F or fluxing this refined asphalt a h eavy petroleum residue was used. I t was a viscous greenish black oil having an odor much like cylinder oil. Its specific g ra vity was 0.9782.
T able I I I shows the results obtained b y treating the Trinidad refined asphalt in the same manner as the Obispo refined asphalt. The samples weighed 20.2276 and 20.1102 grams, respectively. There was no moisture in this asphalt:
Ta b l e III.—Tr i n i d a d Re f i n e d As p h a l t.
D u ctility. Penetration.
A t A t A t A t A t A t
0 ° C . 2 5 ° C. 4 6 ° C. 0 ° C . 2 5 ° C. 4 6 ° C.
Before h eat
ing, 0 .1 1 .8 8 . 0 . 0 .7 5 1 .5 1 1.0
A fter heatin g 0 .0 0 .2 0 .7 5 0 .0 0 .2 1 .0 L o s s . . . 10010# 8 8 . 88# 9 0 .6 3 # 10 0 .0 # 8 6 .66# 9 0 .9 #
The low penetration figures here recorded are ac
counted for, in p art a t least, b y the presence of min
eral m atter in the asphalt. This would prevent the needle of the machine from sinking into the m aterial as far as it would otherwise.
M ixing 135 grams of Trinidad refined asphalt w ith 19 gram s of the fluxing oil, produced an asphal
tic cem ent w ith ^ penetration of 40 and a d uctility of 24 a t 250 C. I t was necessary to heat this asphaltic cem ent to 130° C. in order to get a thorough m ixture.
T able IV shows the results on this asphalt, when heated to variou s tem peratures:
Ta b l e I V . — Tr i n i d a d As p h a l t i c Ce m e n t.
D u ctility Penetration.
A t A t A t A t A t A t
o o p 2 5 ° C. 4 6 ° C. o Ü
25° C. 4 6 ° C.
H eated to 130° C. . 1.3 2 4 .0 2 9 .3 4 .0 4 0 .0 1 6 5 .0 H eated to 150° C . . . 0 .2 20.8 2 9 .0 4 .0 2 4 .0 1 0 4 .0 H eated to 175° C . . . 0 . 1 1 5 .6 2 8 .8 3 .0 20.0 8 5 .0 H eated to 2 0 0 ° C . . . 0 .1 10. 6 2 3 .0 2 . 0 14 .0 6 2 .0
D u ctility and penetration are the necessary char
acteristic properties of an asphalt which m ake it valuable as a paving m aterial. I t is these proper
ties which prevent a good asphalt pavem ent from cracking up in the winter or from becom ing so soft in the sum m er-that traffic tends to wear it off the crown of the street into the gutters. So, if these properties are decreased, the efficiency of the asphalt as a paving m aterial would be seriously impaired.
There are certain streets in Lincoln, Nebraska, where the asphalt pavem ent has deteriorated much more rapidly than it should. A t the time these streets were paved the writer was testing the asphalt for the city engineer. T h e m ixing tanks were known to have been heated to a tem perature which caused the asphaltic cem ent to flash and bum . A t other times the m ixture of asphaltic cem ent and sand, even after it was dumped into the street from the wagons, showed a tem p eratu re' above 200° C. The hottest load was 220° C., or 428° F . T h e ex a ct location of these overheated portions were noted and th ey are now the w orst spots in the street. So in this case the results obtained in the laboratory were confirmed b y actual outside observation.
Ta b l eV .—So l u b i l i t yo f Ob i s p oa n d Tr i n i d a d As p h a l t s. Per cent, asphalt Mineral m atter Per cent, organic
soluble. in asphalt. m atter soluble.
Trini Trini Trini
Solvent used. Obispo. dad. Obispo. dad. Obispo. dad.
A c eto n e ... 4 5 .8 3 14.1 4 0 .4 3 3 4 .9 3 4 6 .0 2 2 1 .7 3 B e n z o l... 99 .4 5 65 .0 1 0 .4 3 3 4 .9 3 99 .8 7 9 9 .9 Carbon b isu lp h id e.,. 99.5 3 6 5 .0 5 0 .4 3 3 4 .9 3 9 9 .9 5 9 9 .9 6 Chloroform... . 9 9 .5 5 60 .7 7 0 .4 3 3 4 .9 3 99.9 7 9 3 .3 9 Ether (e th y l)... . not det. 4 4 .8 7 0 .4 3 3 4 .9 3 n o t det. 6 8 .9 5 Ligroin ( 0 . 7 1 0 ) . .. ., 68.88 4 0 .4 0 .4 3 3 4 .9 3 6 9 .1 7 6 2 .0 8
The testing of the solubility of asphalts in various organic solvents was taken up as a prelim inary to an attem p t to determine the constitution of the so-called
“ A sp h alten e” and “ P etrolene." Com paring the lest two columns on the right, it would seem th at there is a great difference in. the constitution of the organic m atter of there two asphalts. The, per cent, of m in
eral m atter here recorded as constant is the average obtained b y various methods both direct and indi*
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 E N G I N E E R I N G C H E M I S T R Y . June, 1910 rect w ith and w ithout oxidizing agents on m any
samples of each asphalt.
All the common methods of extracting the solu
ble portions of an asphalt were tried and set aside.
Some were accurate, b u t too slow; others were fairly rapid but finally divided m ineral m atter would get aw ay. The following extraction apparatus was found to work better than any of the others, consider
in g both time and accuracy.
Glass tubes 6 centim eters long and 2 centimeters in diam eter were obtained. A t one end of each of these tubes was a slight constriction which produced a rim (see illustration). O ver this rimmed end a C. S. & S. filter paper, No. 589 “ Blue R ibbon,” was placed. O ver this, a C. S. & S. filter paper, No. 575, was laid. T h e two papers were then firmly and carefully pressed down over the end of the tube and securely fastened with platinum wire. (Copper wire was later found ( to be just as satisfactory.) The superfluous paper w as then clipped off close and the capsule thus formed extracted w ith the solvent which was to be used on the asphalt. Three or four of these capsules could be prepared a t a time and the extraction allowed to proceed while the attention was directed elsewhere.
T h e tube w ith its weighed sample was then placed in a Soxhlet extractor having ground glass joints.
In the lower part of the Soxhlet a piece of sealed and weighted glass tubing was placed. This tube served a double purpose. I t supported the capsule so that about half of it was above the top of the curve of the siphon tube of the extractor, and it displaced a large volum e of the solvent thereby causing the ap
paratus to siphon over more often. The small am ount of solvent needed enabled the operator to control easily the volatilization and re-condensation on the sample.
In these tests small samples, never more than 1 gram, were used, as it was found the best results were obtained w ith samples of such weight and less time was required for extraction. The samples were not pulverized, those powdered showing a tendency to “ c a k e ” and clog the filter. This made it more difficult to dissolve out the incorporated organic m atter.
W ith this apparatus very accurate duplicate re
sults were obtained. The am ount of mineral m at
ter recovered b y evaporating the solution and ignit
ing the residue w as not sufficient to cause any ap
preciable change in the percentage figures.
The writer is indebted to the city engineer for the samples of asphalt and for the paving specifications;
also to the Globe A sphalt Com pany, of Pittsburg, for samples of their products.
The author wishes to take this opportunity of e x pressing his most sincere and hearty thanks to Chan
cellor Samuel A very, of the U niversity of Nebraska, for his kind advice which was offered a t the most op
portune moments and which proved of so much value.
Re f e r e n c e s.
Jour. A m tr. Chem. Sac.. 16, 809; 17, 55, 275; 27 , 293; 28 , 648.
Th i s Jo u r n a l. 1 , 7 5 1 .
A . W. Dow , “ D irections for Using D ow Penetration M achine." P u b
lished b y D ow and Sm ith, o f N ew York.
A. W . D ow , “T he T estin g of B itum en s for Paving Purposes.” From Proceedings Am er. Soc. for Testing M aterials, Vol. I l l , 1903.
Byrne, “ Insp ection of M aterials and W orkm anship Em ployed ^ C o n struction.”
Byrne, “ H igh w ay Construction.”
Richardson, “ Modern A sphalt P avem en ts.”
Stillm an, “ Engineering Chem istry.”
Ulzer and Fraenkel, “ Chemical Technical A nalysis.”
Thorpe, "O utlines of Industrial Chem istry.”
Brannt, “ Petroleum and I ts Products.” . Sadtler, “ Industrial Organic Chem istry.”
Allen, Vol. II, P t, 2: "Commercial Organic A nalysis.”
Un i v e r s i t yo f Ne b r a s k a, Ch e m i c a l La b o r a t o r y.
ON THE FORMATION OF CARBENES.
B y D . B a s i l W. A l e x a n d e r . R eceived March 2, 1910.
Since the appearance of the note on a constant- level reservoir,1 the w riter has received sundry in
quiries regarding the last paragraph; it seems there
fore advisable to give the investigations th at pro
duced that statem ent. This article is not to be con
sidered exhaustive b y any means, bu t only gives results and hints in the hope th at other workers m ay take up the problem, and give us results that m ay help to a better understanding of this subject.
A s early as May, 1905, it was noticed b y the writer th at although carbenes in an asphalt are represented b y the difference between the am ounts soluble in CS2 and CC
1
4, nevertheless when the carbenes are separated and dried, only a portion is then soluble in CS2, showing th at a change takes place which renders them insoluble in CS2, and only slightly more so in CH C1
3; this change being brought about either b y drying, or some other unconsidered cause, such as the free action of daylight. I t is stated b y A. S.Cooper, form erly S tate M ineralogist for California,2 th at b y the action of d ayligh t asphalt undergoes polym erization, and certain portions are rendered less easily dissolved than before the action of light.
Some work was done along these lines, but was tem porarily abandoned on account of other and more pressing business, so it was not until early in 1909 that it was taken up again. A t this time it was noticed th at certain other laboratories w ith which the writer had been able to com pare results had not agreed with his own results when deter
mining carbenes b y Clifford R ichardson’s carbon tetrachloride method, while other laboratories es
tablished and controlled b y the wrriter were in fair agreement. This led to investigation, and it was
1 J . A m . Chem. Soc.. 31, 1052.
2 B ull. 16, Cal. Sta te Mining Bureau, p. 18.
A L E X A N D E R O N T H E F O R M A T I O N O F C A R B E N E S . 243
found th at the chem ist consistently finding a larger am ount of carbenes than he, was in the habit of stand
ing his solutions of bitum en in CC
1
4 in beakers, on a glass shelf, placed half-w ay up a window having a northern exposure.T h e following experim ents were then m ade: A kettleful of Durango when melted, was thoroughly stirred and portions taken from the top, middle and bottom , and dissolved in CS2 and CC
1
„ then filtered:Top. Middle. B ottom .
Per cent. Per cent. Per cent.
B itum en b y CS2 9 9 . 1 9 9.3 9 9 .6 B itum en b y CC14 9 4 .8 9 4 .4 9 4 .3
A nother and larger portion was taken from the same kettle, and from this samples were subm itted to several chemists, w ith the following results:
Per cent.
A ssociated results:
Soluble in carbon tetrachloride... 9 4 .8 Soluble in carbon tetrachloride... ... 9 5 .1 Outside results:
Insoluble in carbon tetrachloride... 10.8 Insoluble in carbon tetrachloride... 11.2 Insoluble in carbon tetrachloride. ... 11.9 Insoluble in carbon tetrachloride... 13.6
From our own sample (which gave 98.9 per cent, soluble in CS2), two portions, of about a gram each, were covered w ith 100 cc. of CC
1
4 in an Erlenm eyer flask, one being set on a window ledge and the other in a cupboard in the photographic dark room. These were filtered n ext day, and am ounts soluble obtained:(а) Exposed in window, 85.3 per cent., or 13 .6 per cent, carbenes.
(б) Stood in cupboard, 94.9 per cent., or 4 .0 per cent, carbenes.
Showing an increase of 9.6 per cent, carbenes due to the light.
I t was considered possible th at tem perature m ight have had some effect, so two more fresh portions were covered with 100 cc. of CC
1
4 and both stood together on the window ledge, b u t one was fitted with a close-fitting sheath of b lack paper. T h e results next d a y were:
(c) U nsheathed flask, 8 3 .7 per cent., or 15 .2 per cent, carbenes.
(d) Sheathed flask, 93.8 per cent., or 5 .1 per cent, carbenes.
H ere w e have a difference of 10 .1 per cent, car
benes due to light.
The filtrates from these two samples .w'ere allowed to stand, both unsheathed, on the regular CC
1
4 shelf in the analysis room in subdued light for a day, and then filtered:(e) F iltrate from (c) gave additional carbenes, 0 .5 per cent, of original am ount taken.
(/) F iltrate from (d) gave additional carbenes, 3 .3 per cent, of original am ount taken.
T his seems to show th at the carbenes were con
tained in the sheathed sample and only needed the light to bring them into evidence.
T h e filtration was done w ithout an y precautions against light, and the filtrates were allowed to stand for three days in the dark and again filtered:
(g) F iltrate from (e) gave additional carbenes, 0 .6 per cent, of original am ount taken.
(h) F iltrate from (/) gave additional carbenes, 1 .6 per cent, of original am ount taken.
These carbenes m ight have been produced during filtration, and probably were, so the- filtrates from (g) and (h) were stood in full sunlight for three hours, and 011 filtering:
(i) F iltrate from (g) gave 2.3 per cent, additional carbenes.
(j) F iltrate from (h) gave 7 .9 per cent, additional carbenes.
Thus it m ay be seen that the total am ount df car
benes obtained from the two original samples, (c) and (d), differ to the am ount of only 0 .7 per cent.
T o further investigate the possible effect of tem perature the mode of procedure was slightly changed;
three samples were taken from the original can of supply, covered w ith 100 cc. of CC
1
., iri three E rlenm eyer flasks; the flask sheathed in black paper was further covered w ith w hite paper, another w ith white paper only, and the third left uncovered. A ll three flasks were placed on a board and set in full sunlight for 6 hours, and filtered:
(k) Unsljeathed flask, soluble in CC
1
4, 76.2 per cent, or 2 2 .7 per cent, carbenes.(I) W hite paper only, soluble in CC
1
4, 84.4 per cent., or 14 .5 per cent, carbenes.(m) W hite and black paper, soluble in CC
1
4, 93.5 per cent., or 5 .4 per cent, carbenes.O nly the filtrate from (m) was saved, and that was set in the dark for two months, and on filtering gave:
(n) A dditional carbenes 2 .5 per cent, of am ount taken for (m).
A further standing for two months, in the dark, of the filtrate from (n) gave:
(0) A dditional carbenes 0 .4 per cent.; b u t on al
low ing the filtrate from (o) to stand in full sunlight for only twro hours we obtained:
A dditional carbenes 10 .4 per cent.
E ven a t this the total am ount of carbenes did not come up .to the am ount obtained in one setting in the sun, as in (k).
I t was considered possible th at the effect m ight not be due entirely to the CC
1
4, and so a fresh sample of the m aterial was spread as thinly as possible on a piece of glass and exposed to full sunlight for ’] l/2 hours. A bou t a gram of this w as weighed out, and also the same am ount of the fresh m aterial, and subjected to the regular operations:
R egular method, soluble in CC
1
4, 93.4 per cent., or 5 .5 per cent, carbenes.Exposed asphalt, soluble in CC
1
4, 9 3 .7 per cent., or 5 .2 per cent, carbenes.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 E N G I N E E R I N G C H E M I S T R Y . June, 1910 I t is evident the CC
1
4 is partly, at least, responsiblefor carbenes.
A further exam ination b y fusing some of the car
benes obtained w ith a fusing m ixture showed the presence of chlorine in them, a blank determ ination w ith the same chemicals showing none.
H avin g reached this point, we desired to know the reliability of results form erly obtained, so we used two samples that had been reported thus:
Per cen t. Per cent.
T est N o. 10790. 10792.
Bitum en b y carbon b isu lp h id e.. ... 9 9 .5 9 9 .5 Bitum en b y carbon tetrachloride 9 9 .7 9 9 .5 Penetrom eter at 77° F ... 3 4 ° 54°
W eighed portions of these two samples were dis
solved in io o cc. of CC
1
4 and allowed to stand for 6 hours in full sunlight and then filtered:10790 bitum en b y CC
1
4, 89.4 per cent., or 10.1 per cent, carbenes.10792 bitum en b y CC
1
4, 9 3 .7 per cent., or 5.8 per cent, carbenes.I t was here that the effect of penetration was noticed, as the above results show that the harder asphalt contained the larger am ount of carbenes, the other conditions being identical, or nearly so.
Carbenes are generally supposed to be produced b y overheating asphaltic oils during their condensa
tion into asphalt, and experiments were conducted along these lines. For this purpose three samples of an asphaltic oil were heated a t 400° F. until they came to asphalt:
Sol. in CCIV Penetrometer. Per cent.
Sam ple N o. 1 ... 51 9 9 .8 Sam ple No. 2 ... 25 9 9 .4 Sam ple N o. 3 . . . ... 21 99 .5
Fresh portions of these three samples were dis
solved in CC
1
4 and exposed to direct sunlight for 6 hours and filtered:Per cent.
Sam ple N o. 1... 9 8 .9 Sam ple No. 2 ... 9 8 .2 Sam ple No. 3 ... 9 8 .9
T he filtrates were treated the same w ay, and gave:
Per cent.
Sam ple N o. 1 ... 9 7 .0 Sample N o. 2 ... 9 4 .8 Sam ple N o. 3 ... 9 4 .1 *
It is apparent that asphalt made b y evaporation a t a com paratively low tem perature does not suffer the decomposition that produces carbenes, and tends to confirm the theory th at they are the result of
“ crack in g” in the stills; but the action of light will still produce them, Especially when it is run down to a low penetration.
Three samples of Durango th at had been dissolved in CS2( decanted, and the bitum en recovered from solution and heated to 300° F ., were examined for their solubility in CC
1
4, w ith the following resu lts:Before After
treatm ent. treatm ent.
Per cent. Per cen t.
Test N o. 10615. Soluble in CCl* 9 9 .2 9 9 .4 T est N o. 10615. Soluble in CCl4 9 9 .2 9 9 .4 T est N o. 10706. Soluble in CCU 9 2 .3 92 .8
I t will be noticed that the second sample is a dupli
cate of the first.
I t does not appear th at the solubility of an asphalt in CC
1
4 is affected b y dissolving it in CS2, and subsequently recovering the bitumen.
Satisfactory results were not obtained from re
fined Trinidad asphalt, on account of the extrem ely fine mineral m atter, so a sample of the Trinidad asphalt was dissolved in CS2, filtered, and the filtrate swung in a centrifuge, the solvent evaporated, and the' resi
due heated to 300 ° F ., this being the same treatm ent given to the above samples. T est Nos. 10615 and
10706.
■ The following results were obtained:
Bitum en b y carbon bisulphide, 9 9 .7 per cent.
(a) Soluble in CC
1
4 after exposure to sunlight for i 5 Y 2 hours, 7 3 .9 per cent., or 25.8 per cent, carbenes.(b) Sam e as (a) excep t flask sheathed in w hite over black paper, 99.3 per cent., or 0.4 per cent, carbenes.
(c) F iltrate from (a) after 6 months in subdued light, carbenes, 2 .5 per cent.
(id) F iltrate from (i>) after 6 m onths in cupboard, carbenes, o . 5 per cent.
F iltrate from (c) after sunlight for 7 1/., hours, car
benes, 4 .0 per cent.
F iltrate from (d) after sunlight for 7 ‘/ 2 hours, car
benes, 17 .0 per cent.
These results appear to subvert the “ form ation b y cracking” theory, and narrow the question down to one of penetration and. actinic light. T o verify the latter, two fresh . portions of the bitum en e x tracted from Trinidad refined asphalt were dissolved as usual in 100 cc. of CC
1
4, and the flasks p u t in two tin boxes, made for the purpose, having one side, respectively, of red and blue glass; these were ex posed to full sunlight for 15 l/2 hours and w atched so th at the sun’s rays fell on the solutions through the glass a t all tim es; filtered:(a) Exposed to red rays gave carbenes, 0.8 per cent.
(b) Exposed to blue rays g ave carbenes, 16.6 per cent.
In the case of (6) the flask, after exposure, was stained b y the solution on the side n ext to the light to a dull brown, b u t hardly stained on the opposite side at a ll; it was nearly all removed b y CC
1
4 with a feather; the flask (a) was not stained, and washed clean. The filtration of these samples was done in the dark room under red light, and there the filtrates were allowed to stand in their respective colored boxes for six m onths nearly, when they were again filtered in the dark room:A L E X A N D E R O N T H E F O R M A T I O N O F C A R B E N E S . 245
(c) F iltrate from (a), additional carbenes, 0.8 per cent, of original am ount taken.
(d) F iltrate from (b), additional carbenes, 3 .2 per cent, of original am ount taken.
T he filtrates were exposed to sunlight for 7 hours and filtered:
(e) F iltrate from (c), additional carbenes, 15 .9 per cent, of original am ount taken.
(/) F iltrate from (d), additional carbenes, 6 .7 per cent, of original am ount taken.
T he filtrates were again exposed to sunlight for 7 hours, b u t the sun was obscured nearly all the time;
filtered:
F iltrate from (e), additional carbenes, 2 .4 per cent, of original am ount taken.
F iltrate from (/), additional carbenes, 2 .0 per cent, of original am ount taken.
These results cannot be called satisfactory, inas
much as the total am ount of carbenes in each case are n ot the same, as one would logically expect them to be. H ow ever, more w ork will be done on these lines.
I t has been noted above th at a sample of asphalt showing no carbenes b y our regular method, would display them is exposed to ligh t; and it was wondered if the reverse would be true, so a sample that showed 82.2 per cent, soluble in CC
1
4, though the.penetration was 8 1 0 Penr., was dissolved in 100 cc. of CC1
., and allowed to stand for the regular time of 18 hours, all operations, after weighing, being conducted in the dark room. A fter filtering in the dark we ob tained :Soluble in CC
1
„ 92.8 per cent.So even light cannot produce carbenes unless the conditions áre favorable. Investigation along these lines were discontinued for the present, but continuing the subject of penetration, experim ents were con
ducted on Gilsonite, using 100 cc. of CS2 and CC
1
4, and exposing to full sunlight for 7 hours:Sol. in CS2. Sol. in CC14.
(a) Screened from lig h t... 9 9 .9 9 9 .6
• ( 6) Unscreened from lig h t... 9 9 .9 4 8 .0
T h e filtrates from the CS2 solutions were further exposed under the same conditions for 20 hours, b u t there was no change, excep t that the flask above the solution was stained brown. The filtrate from CC
1
., solution (a) was exposed to full sunlight for 7 hours and filtered:Per cent.
t
Soluble in CCU... 7 6 .7
The carbenes in these cases appear gelatinous, and apparently quite different from the carbenes from Durango grade of asphalt; this qu ality makes them ve ry hard to filter. During the process of investigation of this material, carbenes of a brown color were obtained, b u t these results are not y e t ready for publication.
Some observations relating to the CC
1
, used m aybe of interest, and lead to results throwing more light on this subject. During the distillation of solutions of bitum en in CC
1
4, to recover the solvent, it has been noticed that IICl is evolved if w ater is present, the residue in the flask becomes quite brittle when the solvent is evaporated, and the sides of the flask used for distillation show the characteristic “ carbene xing.’ ' The evolution of IICl is more marked when a copper still is used. A piece of blue litm us paper suspended over a solution of bitum en in CC1
4, especially when in the sunlight, is quickly reddened; again, if a sample of the CC
1
, itself is exposed for a length of time in a stoppered bottle to sunlight, it will separate into two layers, the upper one about i per cent., yellowish, w atery, and acid; and blue litm us paper is reddened. This layer, however, is not IIC l, as it gives a buff, flocculent, not curdy, precipitate w ith silver nitrate.SUM M ARY AND CONCLUSIONS.
I t cannot be affirmed that carbenes are altogether formed during the production of asphalt from petro
leum in a still, b u t that they are due:
(i) T o a concentration, or hardness, produced under circum stances about which we can only theorize until w e acquire greater knowledge; and,
•(2) T o a condition effected in the asphalt which when a suitable agent (in this case chlorine) is introduced, renders b y com bination a portion insoluble and b rit
tle; the action is probably analogous to the form a
tion of “ acid sludge” in the refining of distillates b y sulphuric acid; and,
(3) T o the action of actinic light on solutions of bitum en in carbon tetrachloride, or on the tetra
chloride itself, when chlorine is liberated, and is taken up b y the unsaturated hydrocarbons form ing an in
soluble combination.
LA T E R NO TE.
We are glad to see from Mr. M ackenzie’s excellent article in the A pril number of Th i s Jo u r n a l th at he has been able to verify in a great measure our own re
sults, and hope th at this pu b licity m ay lead to some method of knowing and determ ining accurately these vexatious carbenes.
W e would like to further state th at following a sug
gestion of Mr. R alph M artin it is now our practice to filter our re-distilled CC
1
., through a layer of CaCl2, which seems to sweeten it considerably.A nother hint we have recently received from Mr. E . O. Heinrich, C ity Chemist of Tacom a, W ash., is that the form ation of H C
1
is possibly due to the presence of CHCI3 in the CC1
,, derived during m anufacture,1 a reaction being set up under the influence of sunlight, in accordance with the equation:CC
1
( and CH C1
3 — H C1
and C2C1
0.1000 Da te St., Los Ang el es. Cal. Feb. 24, 1910.
* R ichter’s Org. Chem., V ol. 1, p. 386.
