The Journal of Industrial and Engineering Ghemistry
Published by THE AMERIGAN GHEMIGAL SO CIETY
Volume III JU N E , 1911 No. 6
B O A R D O F E D I T O R S .
Editor: M. C. W hitaker.
Associate Editors :
Geo. P. Adamson, E. G. B ailey, H. E. Barnard, G. E . Barton, W m . B rady, W m . Campbell, F. B. Carpen
ter, Virgil Coblentz, Francis I. D upont, W . C. E baugh, W m . C. Geer, W . F. H illebrand, W . D. H om e, K a rl Langenbeck, A. D. L ittle, C. E. Lucke, P. C. M cllhiney, E. B. M cCready, W m . McMurtrie, J. M erritt M atthews, T. J. Parker, J. D. Pennock, W . D. Richardson, Geo. C. Stone, E rnst T w itchell, R obt. W ahl, W m . H. W alker, W. R! Whitney.
Published m on th ly. Subscription price to non-m em bers of the American Chemical Society, $6.00 yearly.
Foreign p ostage, seven ty-five cen ts, Canada, Cuba and M exico excepted.
Entered at th e Post-Office, Kaston, Pa., as Second-class M atter.
EDITORIALS
A SURVEY OF CYANIDATION IN 1911.
The advances made in the last ten years in the art of cyaniding ores of gold and silver have taken place so rapidly that few men are aware of the extent to which the field of application of the process has widened. Indeed, these changes and im provem ents have followed so fast on the heels of one another th at even the men engaged in the practice of cyanidation are at times ignorant of the later developm ents. The change has been gradual, and one of the most striking things about it is the chem istry of the process which is little advanced tod ay over w hat it was ten years ago. Essentially a hydrochem ical process, all the improvements have been mechanical and though we may have a better knowledge of its chem istry, th at knowledge is of little use to us in solving the problems connected with it. The essential advan tage of cyan i
dation over other hydro-m etallurgical processes is
>ts chemical simplicity. Various w orkable processes have been devised that im prove on the present cyanide process in theory, but in practice, the fact th at they require a higher degree of m etallurgical skill for their successful application, w ill alw ays prove a bar to their adoption if cyanidation w ill do the work alm ost as
"ell. There undoubtedly will be ores of gold and silver that cyanidation will not ex tra ct the values from and we can turn to other processes when th at ls so- The great va riety of ores treated b y cyanida- t]on makes one feel th at where a gold or silver ore is concerned, in at least nine cases out of ten, cyanida- t'on will make a better extraction at a lower cost than an>' other process.
Cyanidation was first practiced on the sand or granular portion of the tailing, from mills treating iuartzose, gold-bearing ores. I t was not regarded as suited to silver ores at first but soon was found to be applicable to them, though it made only a 60 per Cent' recovery, because it did so a t such a low cost.
These first plants were percolation plants and could only operate when part, or all, of the slime, or col
loidal portion of the ore, was run to waste. The decantation plants of South A frica were an effort to solve the problem of treating slime th at never fu lly succeeded. The most marked advance in cyanida
tion began when the submerged suction filter was brought to such a state of m echanical efficiency th at it was generally applicable. The two best known exponents of this principle in this country are the Moore and B utters filters, though I believe the idea was previously em ployed b y Cassel in South A frica in a less efficient form.
The submerged suction filters not only made it possible to treat all the slime b u t even made it de
sirable to reduce as much of the ore as possible to nearly a slime condition. These new conditions created new demands and a host of clever m echanical ap pliances has been perfected to meet them. It was no longer possible to bring the finely ground ore in contact with cyanide solution b y passing the solution through the ore, as in a percolation plant, so this was achieved b y m ixing the ore and solution together and keeping the ore suspended in the solution b y means of agitation. U nconsciously the tremendous im prove
ment in extraction possible through finer grinding has been learned until now cyanidation presupposes fine grinding, agitation and filtration.
A t present, gold and silver ores varyin g w idely in chem ical composition, and in value from $4 to $35 in gold, and from a few ounces to th irty ounces in silver, are being successfully cyanided. C yanidation has displaced pan-am algam ation. The cost of cyanid
ing is lower, the extraction better and pan-am al
gam ation only was suited to ores essentially of silver, whereas cyanidation extracts both gold and silver.
L ixiviation, w ith the necessity of previous roasting, costs much more to practice, makes a poorer extraction
366 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, i9„
of silver and none of the gold. Chlorination likewise requires roasting of the ore and m akes no extraction of the silver. E ven where an ore is so rebellious th at it must first be roasted, cyaniding has displaced chlori
nation and lately the ores of Cripple Creek and Cobalt, chem ically com plex and hitherto considered hard m etallurgical problems, are now successfully treated in the raw state, when low grade, b y cyanidation.
T he last field to be invaded, and one in which the pos
sibilities have not been reached, is th at of concentrate treatm ent. H itherto the concentrate produced b y gold and silver mills was alw ays shipped to the smelters, in m any cases an uneconomical proceeding— b u t now every few months adds another to the list of concen
trates th at can be successfully cyanided. Simple ore-dressing— th at is, am algam ation and concentra
t i o n — will hold its own for some tim e to come on the low-grade ores to which ■ it is suited, because the preparation of the ores is less elaborate and the whole operation can be performed for about one-third of the cost of cyanidation. E ve n then, the concentrate w ill probably be treated b y cyanidation and wherever the grade of the ore is high enough to ju stify the ex pense of m aking a higher extraction, cyanidation will be preferred.
Most of these processes require a high order of m etallurgical skill, b u t cyanidation can be success
fu lly practiced b y men of no technical education and little knowledge of chem istry.
The developm ent of the process has consisted purely in refinement. Cyanide mills are now elaborate structures in which there is a special device em ployed for every possible operation. This high efficiency has been attained a t the cost of flexibility, and smooth continuous operation is essential to low cost. The tendency in mill practice is to relieve the operator of the necessity of exercising his judgm ent and to place more responsibility in the hands of the designer.
The function of the mill-man now is to understand his m achinery and keep it in operation, b u t the e xtrac
tion th a t the mill m akes depends on th at m achinery being properly adjusted in the design of mill.
The cost of mills has increased enormously. This has m ade it advisable to em ploy a better standard of m aterials in their construction and more care in their design. A degree of thoroughness is exercised in testing previous to design th at was unheard of a few years ago. Then standard mill construction in
volved 850 lb. stamps, wooden m ortar blocks and fram ing, and three tons per stam p per d ay was re
garded as good practice; the rough approxim ation generally accepted for mill construction was S i , 000 per stam p. This was equivalent to a cost for con
struction of S3 3 5 per ton of d aily capacity. Later the increased cost of m aterials and labor and the in
crease to 1050 lb. in the w eight of stam ps, w ith the consequent d aily cap acity of five tons per stamp, made $2,000 per stam p the figure, equivalent to S400 per ton of d aily capacity. Now, the adven t of the fine-grinding, agitation and filtration plan t has so added to the m achinery installed in mills, th a t the role of the stam p has been altered so th a t it can no
longer be taken as a basis for estim ation and $1,000 per ton of d aily cap acity will often be found as the correct figure for m ill construction. The variation in cost is greater, however, and an estim ate must be gone into w ith more care than form erly was the case.
The standards of construction have changed. A wooden m ortar block now excites more comment than a concrete one did seven years ago. Steel framing, concrete floors, retaining walls and foundation all are common practice.
T he mills all resemble one another so that they virtu a lly conform to a type. Those th at differ from the typ e remain different not because th ey are better b u t because it would cost too much to change them.
E v e ry modern cyanide plant now involves fine-grind
ing, settling of the pulp to greater density and removal of the surplus solution, agitation, further settling and filtration. Crushing, in the m ajority of cases, is per
formed wet. The reason for the wide-spread adop
tion of this process to the exclusion of all others and of this typ e of mill is found in the cost and the extrac
tion. The cost of cyanidation seldom gets below 50 cents per ton, b u t it should never exceed Si.50 per ton. The extraction of gold should always be from go to 95 per cent., and of silver from 88 to 92*percent.
These results, which are so frequently attained in practice th at th ey can be expected, m ake cyanidation an art, w ith the wide-spread application of which every m etallurgist should be familiar.
Ar t h u r R . To w n s e n d.
TH E GENERAL MEETING IN JUNE.
Those chemists who were fortunate enough to listen to the rem arks of Charles F . Chandler during the banquet at the opening of the Chem ists’ Club last March will not soon forget the earnestness of his de
scription of the grow th of the cooperative idea among Am erican chemists. H aving been among the very first in this country to organize a small group of men interested in chem istry and having been one of the small num ber who m et around Priestley’s grave in Northum berland, Penn., A u gu st 1, 1874, w h i c h meeting resulted in the organization of the American Chemical Society, he spoke w ith authority. The chief tenor of his advice was to "g e t together,” and those who know him well know th at he as well as others of the pioneers in Am erican chem ical science have exerted their full influence throughout their lives to induce Am erican chem ists to get together for their own advancem ent and the advancem ent of the pro
fession. The point cannot be too strongly emphasized,
for its practical em bodim ent is seen in the r e m a r k a b l e
grow th of the Am erican Chemical Society in the last tw en ty years and the wonderful influence it has had upon the developm ent of chem ical industry in Amen»
and upon the increase of individual knowledge and endeavor among Am erican chemists.
The Local Sections of the Society afford means 0.
personal contact w ith chemists in limited localities, and these Local Sections have done a great and goo1
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 . 367
work in developing the spirit of helpfulness and fellowship among their members. A ltogether too few of the members of the Society, however, attend the general meetings which occur semi-annually. It is true that attendance a t these meetings is rapidly increasing, but it has not y e t reached the point th at can reasonably be expected, and it is hoped th at the members of the Society will carefully consider this point and make every endeavor to be present a t In dianapolis and future gatherings of the general Society.
No member who has not been a regular attendant at the general meetings can realize the wonderful help they may be in the developm ent of the individual.
They give opportunity of m eeting chemists in all kinds of work, thereby broadening the knowledge of the individual; th ey give opportunity of m eeting chemists who are interested in the same lines of work and thereby frequen tly add to the specific facts which are important in special industries; th ey give opportunity of m eeting the m ost successful chemists both in industrial and-educational w ork and thereby add inspiration— th ey give opportunity for visiting special industrial works, adding to the practical know ledge of the visitor; th ey give opportunity for a sum
mer or winter outing, for e very Local Section does much to entertain its guests.
Many chemists feel th at th ey cannot afford the time and expense of attending these meetings. Most of those who do attend, especially after th ey have been two or three times so th at th ey becom e acquainted, realize that they cannot afford to sta y aw ay even if the distance is great and tim e is pressing. E specially the young or isolated chem ist who has not the op
portunity of frequent intercourse w ith his fellows, who is too apt to get into one line of thought, cannot afford to stay away, for the friendships made and the inspiration gained through knowledge of the work of others will make him a better chem ist and is certain u the end to add to his m aterial as well as to his mental advancement.
It is particularly pleasant to see the changing a t
titude of many of our m ost prom inent firms tow ards attendance at the general meetings of the Society.
A few years ago chemists com plained th at th ey found it difficult even to get the necessary leave of absence to attend the meeting. Now the case is quite often reversed. Firms have found th a t th ey gain so much by the presence of their chem ists a t these meetings that those who have had .experience have alm ost invariably not only allowed the necessary tim e for attendance but in m any instances now insist upon attendance and in an increasing num ber of cases p ay all the expenses of the chosen individual. These cases are as yet altogether too few, b u t th ey are yearly increasing and it is indicative of a broader and more successful American chem ical industry. I t is certain ly true that it is the m ost progressive and successful
‘ rms that require their chem ists as a p art of their
"’ork to attend the general meetings of the Am erican Chemical Society.
Special efforts are b ein g m ad e to h a v e th e su m m er meeting in Indianapolis a g re a t success, an d definite
assurance can now be given th a t every member who attends the m eeting will have unusual oppor
tu n ity both for enjoym ent and for professional de
velopm ent. The E ntertainm ent Com mittee is planning an interesting program for the entertainm ent of ladies, and it is certain th at they, too, will find the m eeting unusually attractive. A n especially am using program is promised for the smoker on one evening; the follow ing evening an autom obile ride to the C ountry Club w ith lunch on the grounds, law n fete, dancing, music, etc., is e x p ec ted ; on the third evening the usual banquet w ill be held. Tw o full days will probably be given up to the meetings of the Divisions, while the morning of the first d ay will be given to general addresses of interest to all chemists. One afternoon those present are to be especially entertained a t the dedication of the new laboratories of the E li L illy Com pany, one of the largest pharm aceutical m anufacturing companies in Am erica. Satu rd ay will probably be given up for the main part to visits to m anufacturing plants.
Begin to plan now for the summer meeting. R e
member the date is W ednesday, June 28th, to Saturday, Ju ly is t; the place, Indianapolis. L et us all “ get together” there! Ch a r l e s L. Pa r s o n s.
ENDOWMENT FUND.
No organization which does a great and broad public service should be w ithout the resource of available funds to prom ote and further the work.
On e very hand we see a continual and w orthy effort being made to stabilize and m ake sure of a continuous performance such endeavor.
W h y the Am erican Chemical Society should not take advantage of the ve ry apparent opportunity is hardly understandable. Here aw aiting us for action are to be found the ideal conditions under which a most successful effort can be undertaken. A n organiza
tion composed of a body of men representing all classes, working for a great common good, nam ely, the ad vancem ent of chem ical knowledge, w ith the best field for its application.
W ith an intangible valuable attrib ute to the goods to be delivered— w orking brains; a m arket for these same goods, which is boundless; an unsatisfied dem and for the best obtainable execution measured b y the ab ility of the individual and the effectiveness m ade possible b y the organization as a unit, w hat more could be asked or could be offered which would appeal to the users and consumers of such a product, nam ely, the body corporate, T he Am erican Chemical Society?
W e who are of it should individually and all together determine upon a course which will enable the In
dustrials of this country to benefit b y an open-minded, free-for-all research in all branches of chem istry, giving to the struggling young chem ist inspiration b y opportunity, and to the m anufacturers a t large an appreciation of unrestricted research, the benefit of which can be enjoyed b y all to the great advance
m ent of the industries of this country.
T h a t an endowm ent fund would properly take care of this need is v e ry apparent, and in one w a y only
3 68 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, i9n
can such a fund be made available— absolute and com plete cooperation w ith a well defined plan of action carefully thought out and made the interest of every director, councilman, division chairman, sectional chairm an and member.
In order to secure the opinion of the Society as a whole we take this opportunity to place an insert in this issue, and request your prom pt reply to the questions of the Chairman of the Endowment Com
m ittee. B. T . B. Hy d e;
ORIGINAL PA PL R 5.
TH E MICROSCOPIC EXAMINATION AND IDENTIFICATION OF CARBON.
B y G . A . Ro u s h. R eceived March 28, 1911,
The rem arkable success which has in the last fevv years attended the application of microscopic methods to the exam ination of the structure and properties of metals and alloys suggested an attem pt to ap p ly the principles of m etallography in the exam ination of carbon. The carbons of commerce are made up of small particles of one or more of the different varieties of amorphous carbon cem ented together b y means of a binder, usually tar or pitch. The object of this investi
gation was, if possible, to develop some distinguish
able difference in the appearance under the microscope of these ultim ate particles of carbon, b y which th ey could alw ays be recognized, in order th at the composi
tion of unknown m aterials m ight thus be determined.
The wide-spread uses to which carbon is being put nowadays, and the va rie ty of raw m aterials which m ay enter into the composition of carbons of different classes and grades has made desirable a method for the differentiation of various raw m aterials which m ay have been used, and an approxim ate analysis of the com position of any particular sample. The principal com mercial uses of carbon are for arc light electrodes of various kinds, electrodes in furnace and electro
lytic processes, electrodes for dry and w et batteries, brushes for dynam os and motors, contacts and resist
ance m aterial in electrical apparatus and m achinery.
Such a wide scope of uses as are shown here neces
sarily implies a considerable num ber of available raw m aterials which m ay be used to give the carbons the distinctive properties best suited for their particular uses. The list of available m aterials includes the following:
Lampblack R etort Carbon Petroleum Coke
f Powdered Natural i Flake Artificial
( B ee H ive I* Bitum inous -j B y product
Coal Coke -< Gas
( Anthracite
Coal coke is a t present used ve ry little, if any, on account of its high ash content, but it is a possible available m aterial for future use, provided some pro
cess can be devised to rem ove the ash, or for some use where the presence of the ash is not detrim ental. Lam pblack is used most in the high-grade arc carbons, and to a small exten t in some grades of m otor and dynam o brushes. Electrodes are m ainly of petroleum coke, sometimes of retort
carbon and sometimes of a m ixture of these two.
Motor and dynam o brushes m ay be composed whollv of retort carbon, petroleum coke or one of the varie
ties of graphite, or of m ixtures of tw o or more of these m aterials, w ith occasionally a small percentage oi lam pblack, as stated above.
Heretofore there has been no means of even approxi
m ately estim ating the composition of products of this kind, excep t b y judging from appearance, and com
paring w ith trial m ixtures. Hence the determination of the composition, or the duplication, of a sample of unknown composition was a m atter of considerable difficulty, which, it is hoped, will be somewhat lessened b y the results of this investigation. The results here described are not as full and com plete as might be desired, bu t those obtained are given as a step in the right direction, w ith the hope of adding to them in the future.
Methods.— The methods are sim ply those used in the exam ination of a sample of a m etal or alloy to determine its internal structure. These can be found in full in any standard w ork on metallography, and so will be given only briefly here. The method in general consists in selecting a suitable sample of the material in question, grinding on one side a flat surface by means of an em ery wheel, and then bringing this flat surface to a smooth polish b y rubbing it successively on em ery papers of increasing fineness. After a per
fectly smooth polished surface is obtained, the struc
ture of the m aterial is developed b y etching the sur
face b y means of some suitable etching medium, and the prepared surface is then ready for examination under the microscope, the object being, as stated before, to develop some distinguishable difference in the appear
ance of the particles of which the m aterial is composed, b y which th ey can be recognized and distinguished.
Preparation of the Samples.— The samples s h o u l d be selected so as to be as representative as possible, and several different sections should be prepared so that the results obtained m ay represent a fair average o!
the m aterial in question. In order to accomplish thss, the samples exam ined should include sections, th e pol
ished surfaces of which were originally p erp en d icu la r
to each of the three principal axes of the original sample, since the orientation of the particles of carboa varies w ith the shape of the particle and with the shape of the die through which th ey are forced.1 The samples to be exam ined are best rem oved from the entire mas b y means of a hack saw. The surface to be prepared is first flattened b y touching to an emery wheel or dressing w ith medium rough file. The further polish-
1 Those not familiar with the m anufacture of carbon are referred'0 the writer’s article on this subject in Th i s Jo u r n a l, May, 1909.
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 .
ingof the sample then consists in rem oving the scratches left on the surface b y one grinding medium w ith a suc
ceeding one of finer grit. The em ery or file can be followed by rubbing on an em ery paper of No. i or No. o grade, and this is followed b y successive grades of Crocus paper from No. o to No. oooo. The sample should be rubbed on each paper so th at the scratches produced are a t right angles to those produced b y the preceding paper, since b y this means it can be most easily determined when all the scratches of the pre
ceding paper have been eliminated. A few rubs over the surface of the No. oooo Crocus paper usually suffices to fill the grit of the paper w ith finely divided carbon from the sample in hand so th at the surface is given a more or less mirror-like polish, which is the result desired. The steps necessary to obtain such a polish vary widely with the different m aterials. In some cases only two or three of the different polishing papers are required, and w ith others one m ay have diffi
culty in obtaining the desired results a t all. Some samples give satisfactory results a t first trial, while others require repeated attem pts before th ey finally give a surface that can be exam ined w ith a n y satis
faction. It is a m atter th at can be learned only b y experience, since every individual piece requires its own individual treatment.
After the specimen is polished, it has to be given a further treatment in order th a t the structure m ay be distinguished. The surface m ust be treated in such a manner that the softer parts of the m aterial are worn away more than the harder, so th at in this m an
ner the structure of the m aterial is left in outline.
This may be accomplished in one of two w ays: the sample may be heated to a red heat so that the air attacks and oxidizes the polished surface, leaving the denser and more resistant parts in outline, or the sur
face may be still further polished on a soft, yielding material such as chamois skin or broadcloth, so th at the softer parts are worn aw ay more rapidly than the hard, thus leaving the outlines in relief. In the fur
ther descriptions the former is known as “ heat etch in g ” and the latter as “ relief polish.” These tw o different ways of finishing the specimen give a som ewhat dif
ferent final appearance, and the question as to which is the more suitable for a given sample can be decided only by trial. It is often of benefit to prepare samples in both ways and compare them w ith each other. No set rule can be given as to the am ount of etching nec- i cssaiT> as this varies with the different samples, but usually, if heat-etched, it should not be more than just enough to destroy the gloss of the polished surface.
The specimen should be heated to a red heat in a blast amp and allowed to oxidize in the open air until the gloss from the polished surface has ju st disappeared,
•'hen the entire sample should be cooled below a red eat as quickly as possible b y holding it in a blast of c°ld air. It is often more difficult to get good results
"ith the relief polish than w ith the heat etching. The Wimen should be exam ined frequen tly w ith the mi
croscope during the polishing operation, an d as soon as a good relief is obtained the polishing should be
■topped, for a relief once developed can be easily de-
.169 stroyed b y polishing a little too long. In some cases the relief polish m ight be followed to good advantage b y the heat etching, giving b etter results than could be obtained from either used alone.
A fter a specimen has been etched sufficiently, and has been shown b y a prelim inary exam ination under the microscope to be in the condition in which it is desired, it should be m ounted perm anently on a slide.
The etched surface is very delicate, and if it is attem pted to preserve the specimens loose, even though th ey are packed in cotton, the liability of dam age to the etched surface is great. T h ey are best preserved b y m ount
ing them on a regular microscope slide b y means of sealing w ax. In order th at the entire surface of the specimen m ay be in focus, it is necessary in m ounting th at its surface be kept perfectly parallel w ith the surface of the slide on which it is mounted. This' is done b y
Fig. 1.— M ounting d ev ice for s p e cim en s.
made in an y m achine shop. B y turning the circular ring A, the surfaces B and C can be separated any dis
tance desired and y e t alw ays remain parallel to each other, if the instrum ent has been properly made. The specimen D is laid polished surface down on the face C, the ring A is turned until a slide laid on B ju st clears the specimen, a few drops of hot w ax are then dropped on the back of the specimen or on the center of the slide, and the slide is pressed down against the face of B until the w ax is cold.
A fter the preparation and mounting of the specimen it is ready for exam ination under the microscope. For this purpose any microscope can be used which can be supplied w ith an apparatus for vertical illum ination, descriptions of which can be obtained from an y stand
ard work on m etallography. The instrum ent used in this investigation was the L eitz m etallograph, supplied w ith eye-pieces and objectives giving m agnifications of 65, 200, 325 and 550 diameters. P ractically all of the w ork was done with the tw o lower magnifications, the larger ones being used only occasionally when a sample was found in which the raw m aterials had been ground exceedingly fine.
W henever possible the samples should be filed aw ay for future reference, b u t where this is ' impossible, or where records are w anted for an y other purpose, recourse m ay be had to photography, which will repro
duce the appearance of the samples in all of their detail.
In some w ays the photom icrograph is of more value
3 7° 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, 19,i than the original sample as a record, since it m akes Retort Carbon.— R etort carbon is the carbon deposited easy the comparison of one sample w ith another, or in the flues leading from the retorts in a coal gas bench, different m agnifications of the same sample. The Most retort carbon is found as a h eavy steel-gray mas details of this ‘ process can be obtained from some w ith a granular structure. The mass m ay vary con- book on photography or m etallography.1 siderably in density and porosity in different parts, and
A carbon treated in this manner and view ed under m ay be contam inated w ith more or less soot, but it the microscope has an appearance sim ilar to Figs. 2 is of a fairly uniform granular structure throughout, and 3. In these the individual grains can be distin- A second form is found in the shape of nodules, which guished readily, the former being magnified 65 diameters have been deposited in successive layers, and the and the latter 200 diameters. The problem now grains of carbon are usually smaller than in the gran- before us is to find some means of distinguishing be- ular variety. Fig. 12 shows a section of granular tween the different individual grains when some are retort carbon under a m agnification of 65 diameters;
of one m aterial and some are of another. This was Fig. 13 shows a section of nodular retort carbon under accomplished b y studying pure samples of the different the same magnification. In the former the uniform raw m aterials th at enter into the m ake-up of the car- body of the m aterial is shown, while in the latter the bons. A s has been shown in the list of raw m aterials, successive layers from which the nodules are built up the m ost im portant of these are petroleum coke, are readily distinguished. U nder a higher magnifi- retort carbon, artificial graphite and natural graphite, cation both varieties show p ractically the same struc- varyin g in im portance in the order named. These w ill ture, a som ewhat roughened, finely pitted surface, ap- be taken up and discussed in this order. parently made up of v e ry fine granules, as shown in
Petroleum Coke.— Petroleum coke is the residue left Fig. 14, where the same sample as shown in Fig. 12 is in the. still in the refining of crude oil. A s obtained shown under a m agnification of 200 diameters. This from the oil refineries, it is a black, porous carbona- structure is not changed a t a still higher magnification ceous mass, still containing a considerable percentage of a s 's shown in Fig. 15, which is a sample of retort car- oil, which must be rem oved b y calcining before the bon under a magnification of 325 diameters. This is coke can be used in the m anufacture of carbons. W hile the characteristic structure of retort carbon. As in the mass is visib ly porous, from its nature and origin the case w ith petroleum coke, wide variation may be one would suspect this porosity to extend beyond the met in shape of the individual grains and the structure lim its of visib ility, and as a m atter of fact, it does. the body of the carbon, as can be seen in Figs. 16, Fig. 4 is a section of petroleum coke magnified 65 1§ an(i 2°> but under the higher magnification all diameters. This shows to good advan tage the large show the characteristic pitted surface, as seen in Figs.
pores in the mass, and the cellular nature of the finer 1
7
> *9 and 21-part of the structure. Fig. 5 shows the same section Graphite.— The presence of graphite in a carbon, if under a magnification of 200 diameters. Here it can to an y appreciable extent, can be detected by the be seen th at the porosity of the m aterial still persists general appearance of the carbon and b y the specific up to the lim its of this magnification, givin g the sur- g ravity. Carbons- made from petroleum coke and re
face a peculiar striated appearance, due to the numer- tort carbon are not liable to have a specific gravity ous fine pores throughout the mass of the m aterial, higher than 2.05, and m any do not go above 2.00, This striated appearance is characteristic of petroleum while graphite has a specific g ra v ity of 2.25-2.30. The coke, and can easily be detected and recognized where- am ount of ash in the sample is also an indicator. A ever this m aterial is found. Numerous samples of carbon m ade from a good grade of petroleum coke petroleum coke have been exam ined from different should not e xceed 0.25-0.75 per cent, of ash, the latter sources, th at had received all sorts of different treat- being a high figure. A good grade of retort carbon ment, and none were found in which this peculiar will run from 1.00 to 2.00 per cent., w ith occasionally characteristic striated appearance could not easily be 2.50 per cent. If artificial graphite be present in the developed and recognized. carbon, these figures m ay not be changed very much.
Fig. 6 shows a section of a carbon rod known to be for the better grades of artificial graphite do not have made from petroleum coke, magnified 65 diameters, a v e ry high ash content. I t is probable, though, showing the structure of the body of the carbon, th at one of th e lower grades would be used, which wcula Fig. 7 is the same section under a m agnification of result in an appreciable increase in the ash content.
200 diameters, showing the characteristic striated If natural graphite were used, the difference would be appearance developed in the particles of petroleum still more m arked, unless an exceptionally pure graphite coke. Figs. S and 10 com pared w ith Fig. 6 show were used.
w hat varied appearances m ay be m et in the shape of Artificial Graphite.— Artificial graphite, being £ the individual grains and the structure of the b o d y of m anufactured product obtainable from several di*- the carbon, yet all, under the higher m agnification, ferent sources, does not show a distinctive structure show the same characteristic appearance of petroleum such as characterizes petroleum coke and r e to rt car- coke, as seen in Figs. 7, 9 and 11. bon. No samples of artificial graphite have been
1 For greater detail concerning tlie apparatus and operations men- f o u n d t h u s f a r t h a t p O S S e S S a n y d i s t i n c t i v e ch a fi tioned consult any of th e following books: Gulliver. M etallic A llo ys; Law. i s t i c m a r k i n g s of t h e i r OWH. F i g . 2 2 s h o w s & SCCtlO
A llo ys and T h eir In d u stria l A p p lica tio n ; Osmond and Stead, T he M icro - f r o m & b l o c k Q£ A c h e S O n g r a p h i t e
scopic A n a lysis o f M etals; Howe, Iro n , Steel an d Other A lle y s; Ruer-Mathew- .
son, Introduction to M etallography. lVOITI p e t r o l e u m C O k e. F i g . 2 3 i s E
, presum ably niaa sectio n from another
F ig. 3. - 200.
Carbon sa m p le sh o w in g stru ctu re and in d ivid u al grain s, lle a t-etch ed .
Fig; 5. 200.
S ection o f p etroleu m c o k e . R e lie f p o lish .
F ig . 7. X 200.
F ig . 6 under h ig h er m agn ification . Ilea t-etch ed . Fig. 2. X 65.
Carbon sam ple sh o w in g structure and in d ivid u al grain s.
H ea t-etch ed .
F ig. 4- X 65.
Section o f petroleum c o k e . R elief p o lish .
F ig. 6. X 65.
Carbon rod s h o w in g ind ividu al g ra in s o f p etroleu m cok e.
H eat-etched.
F ig-. 8. X 65.
Carbon rod s h o w in g in d ivid u al g ra in s o f p etroleu m co k e.
H ea t-etch ed .
F ig . 10. X 65.
Carbon rod sh o w in g in d iv id u a l g ra in s o f p etroleu m cok e.
R elief polish.
F ig. 12. X 65.
Section o f gra n u la r retort carbon. R e lie f p o lish .
F ig. 9. X 200.
Fig. S under h ig h er m agn ification . H eat-etch ed .
Fig. 11. X 200.
F ig. 10 under h ig h er m a g n ifica tio n . R elief p olish .
F ig . 13. X 65.
Section of nodular retort carbon. R elief polish.
Fig. 14. 200.
Fig. 12 under h ig h er m agn ification . R e lie f polish R e lie f polish,
F ig . 16. X 65
Carbon sh o w in g in d ivid u al grai us o f retort carb on. R e lie f p olish .
F ig. 17. X 200.
Fig. 16 u n d er h ig h er m agn ification . R elief polish
F ig. 18 X 65. F ig . 19. X 200.
Carbon sh o w in g in d ivid u al g r a in s o f retort carb on. H eat- F ig. 18 under h ig h er m agn ification . H eat-etched, etch ed .
F ig. 24. 200. 25 2°°-
s a m p le c o n ta in in g p etroleu m co k e and a rtificia l g ra p h ite. Sam p le c o n ta in in g retort carbon and artificial graphite.
H eat-etch ed , R elief p olish .
F ig. 20. 65.
Carbon sh o w in g ind ividu al g r a in s o f retort carbon. Heat etched.
Fin. 23. 200.
Section o f A cheson g r a p h ite. H ea t-etclied . Fig. 2 1. X 200
Fig. 20 u n d er h ig h er m agn ification . H eat etched,
F ig. 22. X 200.
S ection o f A ch eso n grap h ite. H eat-etch ed .
F ig . 26. X 65.
Carbon from fine natural g ra p h ite. H eat-etch ed
F ig. 27. X 200.
Fig. 26 under h ig h er m a g n ifica tio n . H eat-etched,
F ig. 28. X 65.
Carbon from larger natural g r a p h ite. H eat-etched.
F ig . 29. 200.
Fig. 2S u n d er h ig h er m a g n ifica tio n . H cat-etch ed .
F ig. 30. 3*5- F ig . 31 . < 65.
Carbon from larger natural g ra p h ite, k e lie f p o lish . Carbon from c o a rse natural g r a p h ite. R e lie f p o lish .
June 1 9 1 i
piece of Acheson graphite th at p lain ly shows its origin by its characteristic striated appearance. The question immediately presents itself, W h y should one of these samples show no distinctive m arkings, and the other show them so plainly? Is this second sample a case of incomplete conversion', or is this retention of form analogous to a pseudomorphous crystal? Sam ples similar to the latter are not a t all hard to fin d ; as a matter of fact, th ey seem to be more numerous than the former.
So far, artificial graphite has shown no characteristics by which it can be distinguished from other m aterials under the microscope. The samples shown in Figs. 24 and 25 contain more or less graphite along w ith the other constituents, bu t one is not able to pick out the individual particles as can be done w ith the petroleum coke and retort carbon, and since, as is stated above, artificial graphite can be made from any of the different forms of amorphous carbon, we are liable to find sam ples of artificial graphite which still retain more or less of the appearance of any of the possible original materials. A full and careful study of the graphites obtained from the different com mercial varieties of amorphous carbon will be required before one can say with any assurance th a t these forms of graphite can be identified, and distinguished one from another.
Natural Graphite,— The natural graphites are as a rule more or less of a flaky nature, w ith the flakes varying in size from an alm ost im palpable powder up to almost an eighth of an inch across. W hile the natural graphites do not have any v e ry distinctive markings, after a little experience one can generally recognize them by the shape of the flakes. Figs. 26 and 27 show a sample of very fine natural graphite under magnifi
cations of 65 and 200 diameters, respectively. Figs.
28 and 29 show a graphite w ith a slightly larger flake, under the same magnifications. Figs. 30 and 31 show flakes of still larger size, the form er being enlarged 325 times, and the latter only 65 times. These last two samples are finished w ith the relief polish, while the other two are heat-etched. The different results produced by the two methods on sim ilar m aterials is shown clearly.
Similarities and Differences.— O ccasionally cases are found where particles of one m aterial show an appear
ance similar to one of the other m aterials, and require considerable study and m anipulation before th ey can be definitely identified as one or the other. The prep
aration of several specimens from each sample exam ined usually will enable one to avoid false conclusions in a
«ase of this kind.
A nodular retort carbon, such as is shown in Fig. 13, m ay under some conditions appear very sim ilar to the striated surface of petroleum coke. This can usually be de
tected by a higher magnification, which will develop to a greater extent the characteristic structure of the retort carbon, and will also aid in the identification b y bringing out more clearly the alm ost perfect parallel-
■sni of the lines on the surface, which is much more perfect than is found in the striations of petroleum coke.
The striated appearance of the petroleum coke is
3 7 1 due to the presence of m inute pores extending through
out the body of the m aterial. If the section is cu t so that its surface is perpendicular, rather than parallel, to the course of these pores, the surface w ill appear more or less pitted, sim ilar to the characteristic struc
ture of retort carbon, p articu larly uncler the lower m agnifications. This can usually be detected w ithout m uch difficulty b y a higher magnification and speci
mens cu t from other parts of the sample under exam ination.
A natural graphite in large flakes, if etched b y heat, m ay show striations sim ilar to petroleum coke, but this m aterial can usually be detected b y the shape of the flakes, as shown in Figs. 29, 30 and 31.
W ith a little experience, one can becom e fam iliar w ith these sim ilarities and differences, so th at the iden
tification of the different m aterials can be accomplished w ithout much difficulty.
Optical A nalysis.— The n ext step after the identifi
cation of the constituents of the carbon is the deter
m ination of the quantities in which the different con
stituents are present. W hen the m aterials have been ground ve ry fine, this is a difficult m atter, but if the particles are not too small, so th at their outlines can be definitely determined, fairly accurate results can be obtained. F or this purpose, the methods of opti
cal analysis used in petrography for the analysis of rocks can be used.
It has been determ ined1 th at the sum of the diameters of a large num ber of adjacent particles, measured in a straight line, bears the same ratio to the total length of the line from outside to outside of the first and last particles, as the volum e of these particles is to the total volum e of the m aterial; in other words, the quotient ob
tained b y dividing the total length of the line into the sum of the diameters of the particles touching the line and measured on the line, gives the percentage b y volum e of the whole, occupied b y those particles. F or this purpose the sample in question is p u t under the m icro
scope a t a m agnification sufficient to clearly distin
guish all of the individual particles, using a m icrom eter eye-piece, preferably one w ith cross hairs unless the scale of the microm eter has a line drawn lengthwise through it, as some do have. This w ill give the field an appearance like th at shown in Fig. 32. The hair line serves as the straight line across the surface of the specimen, along which the measurements are to be made. Those portions of the line are measured which lie on the particles whose percentage is sought.
W h atever m ay be the size or shape of the particles encountered, th at portion of the line is measured th at lies on the surface of all the particles under the scale of the micrometer, and note is m ade of the total length of the m icrom eter scale, and of the sum of all the diam eters of the particles as measured along the straigh t line. O ther portions of the specimen are then m eas
ured in the same w a y until the total distance measured over is a t least 100 tim es the diam eter of the largest particle encountered. The greater the total length measured, the greater the accuracy of the results, and,
1 Rosiwal, Verh. W ie n . Geol. Reichsanstalt, 32, 143 (1898); W illiams, A m . Geol., 35, 4 0 -4 3 (1905); Q uantitative Classification of loneous Rocks, b y Cross, Iddings, Pierson and W ashington, 204.
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 IN E E R IN G C H E M I S T R Y .
372 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 9i, if possible, several different specimens of the same m a
terial should be used in order to get a better average.
W hen the desired num ber of measurements have been made, the sum of the diameters measured, times 100,
F 'ft 32- X 65-
S p ecim en under m icrom eter e y e-p ie c e read y for m easu re
m ent.
divided b y the total distance measured over gives the percentage b y volum e occupied b y the particles meas
ured. B y this method, several different constituents m ay be determined at one tim e sim ply b y having sep
arate columns to add the diameters of the different particles measured. W hen the m aterial is made up of tw o or more constituents, it requires more work to get the desired results, because it usually requires a higher magnification to distinguish one constituent from another b y their characteristic m arkings than it does to sim ply distinguish the individual particles from each other; for this reason the apparent diameters of the particles as measured will be greater, and a greater num ber of measurements will be required in order to secure the desired accuracy.
W hen the percentage b y volum e has been determined the percentage b y weight, if desired, can be calculated from the density of the whole mass and th at of the m aterial in question.1
The accuracy of this method as applied to materials of this kind is not as great as might be desired, but w ith care, results can be obtained th at have sufficient accuracy to make them of considerable value. The first precaution to be observed is to secure as fair an average as possible b y m aking a large num ber of meas
urements from as m any different samples of the m ate
rial in question as can be obtained. The fineness to which the constituents entering into the carbon have been pulverized has considerable effect on the accuracy of the results, for if the particles are exceedingly small, it is difficult to obtain a polished section in which the outlines of the grains are cleai* and definite. For this reason, the results as obtained are alw ays liable to be more or less below the true values, owing to the fact th at it is difficult, in m aking the measurements, to find and include all of the very fine particles. L ow results
1 Methods for the determ ination of the density of carbon can be found in the article referred to in N ote 1.
m ay also be expected to a certain exten t if the section has been prepared b y heat etching rather than by the relief polish, since the burning destroys to a greater extent the original outline of the grains.
Specimens of known composition have shown results differing from the calculated contents b y only one per cent., while in other cases the difference was as high as four or five per cent., depending on the success w ith which the sections measured were prepared, the size of the grains, and the definiteness of their outline.
The difficulties encountered w ith ve ry small grains can be only p artially avoided b y higher magnification.
This will of course m ake the grains more easily dis
tinguished, but th ey will still lack definiteness of out
line. A s w ith all the other operations, a little experi
ence will give one the ab ility to make the measure
ments fairly readily.
Conclusion.— These methods of identification and analysis should prove of some value to the m anufacturers and users of carbons of various kinds, particularly so of dynam o and motor brushes, the m anufacture of which must be very closely followed in order to secure a uniform prod
uct, and a system atic stu d y of the structure and body of the different materials should lead to a marked im
provem ent in their qu ality and wearing properties.
The outline of this subject as presented here still leaves much to be desired, b u t it is hoped that sufficient results have been shown to prove themselves of use, and it is hoped th at the field m ay soon be broadened, and further results added to those already obtained.
Mo r g a n t o w n, \ V . Va.
[C O N T R IH U T IO N FROM R E S E A R C H D E P A R T M E N T , T H E A M E R IC A N ROLLING Mi l l Co., Mi d d l e t o w n, « Oh i o.]
TH E DETERMINATION OF OXYGEN IN IRON AND STEEL.
B y Al l e r t o n S. Cu s i i m a n, Director of In stitu te of Industrial Research.
R eceived April 22, 1911.
The determ ination of oxygen in iron and steel has not received sufficient attention in the United States.
The text-books on steel works analyses, which are in common use in this country, do not include methods of analysis for oxygen content of irons and steels.
This is probably due to the fa ct th at in the ordinary steel-making processes ferro-manganese can be freely used, so th at it is assumed th a t the percentage of oxygen in the steel rarely reaches or exceeds the danger point. A s a m atter of fact, steel often carries much more oxygen than it should, as can be seen by re
ferring to Table II.
In the m anufacture of iron of ve ry high purity in basic open-hearth furnaces the oxygen content has to be very carefully w atched, or the product may be overburned and contain an excess of oxide and dissolved oxygen.
A num ber of m ethods for the determination of oxygen have been proposed, am ong which may be m entioned: ( i ) H eating the sample in a stream of dry chlorine; (2) dissolving the sample in special solvents such as copper sulphate or bromine; (3) com bustion of the sample in the form of borings in pure dry hydrogen.
June, 1911 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 .
The latter method, which is due to L edebu r,‘ is the onlv one that has proved reliable. In Ledebur's original method, the sample is given a prelim inary combustion in pure nitrogen in order to b u m off the last traces of impurities and to get rid of all hydro
carbons, as well as adsorbed oxygen on the surface of the particles of the sample. If the prelim inary heating in nitrogen is dispensed with, the results will be slightly higher, but it is probable th a t for general work sufficiently accurate results can be obtained if the sample is carefully prepared for the com bustion in hydrogen.
S A M P L E S .
The samples should consist of fine borings or sh av
ings from a milling machine. The drill or machine tool should be scrupulously clean and free from all traces of oil or dirt, and should be geared to run slowly so as not to heat the sample while it is being cut. Lack of careful attention to this po nt will lead to high results, owing to surface oxidation of the fine particles of the drillings.
A P P A R A T U S .
The apparatus used in m aking the oxygen determ ina
tion is shown in illustration B-740.
A one-gallon K ip p generator is used for generating the hydrogen. It should be charged w ith drillings
3 7 3 as shown in the figure. It passes first over stick potash, and n e xt through a 30 per cent, potash solu
tion. This solution in the second bottle should be renewed as soon as it shows a tinge of yellow, due to the appearance of sulphides. The hydrogen n e xt passes through concentrated sulphuric acid to dry it, and then enters a silica tube w ith 11 " bore, 30"
in length, which contains a 6 " roll of platinum gauze.
The l / " tube lies on top of a 1" X 30" fused silica tube contained in a suitable 12" gas blast furnace.
The object of the prelim inary heating over platinum foil is to free the hydrogen from the sm all q u an tity of oxygen which it alw ays contains. If this precau
tion is not taken, the results will be too high. T h e w ater formed in the small-bore silica tube is caugh t in a “ U ” tube shown in the figure, which contains phosphoric anhydride opened up w ith glass wool.
T his drying tube has rubber stoppers. The con
nection is made with pure gum tubing and is perm anent, the sample being introduced from the opposite end of the com bustion tube. '
B lanks should be run from time to. tim e to make sure that the apparatus is in good order and e very
thing w orking properly. Samples should not be introduced into or rem oved from the com bustion tube when it is more than hand-hot, b u t silica tubes m ay be
D en otation s:
K—K ipp g en era to r.
P t—P otassiu m h yd rate stic k s.
P 2—P otassiu m h yd rate solution . S — S u lp h u ric acid conc.
Tj— yx n b ore silica tube.
Tjj— 1" bore silica tube.
°f pure iron or mossy zinc, and dilute hydrochloric ac>d (r:i). Steel turnings should not be used in the generator, as the object is to generate the purest pos- S1ble hydrogen. H ydrochloric is preferable to sul
phuric acid. A fter its form ation the hydrogen is purified and dried h y passing through the usual train
Leitfaden fur K isenhuttenlaboratonen,” Viewcg und S oh n, Braun-
sd i'vieg, 6 A u f l a g e , 1 9 0 3. S . 1 2 2.
A i—U -tube p h osp h oric acid.
Ao—U -tu be p h o sp h o ric acid •weighed.
A 3—U-tube p h osp h oric acid trap.
B— P latin u m boat.
H — H ook o f J4" co p p er w ire.
qu ickly cooled w ith perfect safety b y turning off the gas and allow ing the cold air blast to p lay on the tube.
M E T H O D .
T w e n ty to thirt}' gram s of finely divided borings are weighed into a platinum or silica b o at !/ / X 1 /V X 6". The boat with its charge is q u ick ly in
