extremes of temperature ; indeed, on the Continent air-tight tile stoves
are in general use.

UkI/Ined Digesters. — The Schenk.

The Schenk digester is a stationary, upright cylinder, 7 feet in diam-
eter by 22 feet height, and is made in sectional castings of deoxidized
bronze with planed flanges, which are bolted together and lead-jointed
in assembling. This alloy the designer assumes is suiBciently acid-
proof for the purpose, without the protection of other resistant lining.

Absolutely homogeneous kalchoid alloys are not of usual occurrence
in manufactures ; liquation, the presence of copper oxides, and the oc-
clusion of gas, militate against the production of constant results. I
take it that "deoxidized" bronze is the outcome of a process that
holies for uniform, reliable castings.

The value of phosphor-bronze in the arts is well established, and it
is used in all the lead-lined digesters for the steam inlets that are im-
mersed in the acid solution. In every case these inlet pipes are in time
consumed.

I am not prepared to admit, after siich examination as I have been
able to make, that deoxidized bronze is superior as an acid resistant
to phospor-bronze. From the little that I have gleaned, it is claimed
that it casts more surely in large masses. Even from this standpoint,
aluminium-bronze will probably be found to be more reliable.

I mention phosphor and aluminium bronze because there is no pat-
ent per se upon the Schenk digester ; the only protection claimed lies
in the fact that the company enjoys the exclusive right of making
digesters of deoxidized bronze ; hence, should any other copper alloy



MICHAELIS ON LIME SULPHITE FIREK. 2'^5'

be found equally available for the purpose, no Impediment exists to its
free use.

It is acknowledged that the deoxidized bronza is acted upon by the
acid solution, and observation confii-ms this conclusion.

The upper portion of the digester, the only part that could be ex-
amined on account of the chip filling, was coated with black oxide of
copper, and both liquid and fiber hold it in solution or suspension.

It is claimed that this erosion is so slight that the longevity of the
digester is not threatened thereby ; this, of course, is a matter of indi-
vidual opinion.

The digester we entered at Appleton showed extensive and deep
honeycombing iu the throat ; we were told that this was due to bad
casting. Without this information I should have come to a very differ-
ent conclusion.

As I understood, one of the claimed valuable attributes of deoxi-
dized bronze is that it makes sound castings.

It hardly seems judicious to use so unsound a section in an entirely
new apparatus, designed as the foundation of a great manufacture. The
I^ittings looked to me like the usual erosions manifested when corroding
liquids or gases move over bronze surfaces.

Owing to liquation, more soluble alloys occur in spots, and the eat-
ing-out of these produces the typical pitting.

The base section of this digester either had been, or was to be, re-
placed ; the top section I would call in unserviceable condition ; at both,
these i^laces the bronze is exposed, in blowing-off, to moving liquid
and gas.

Copper-tin alloys are " kittle-cattle," all authorities agree iu testify-
ing to their sensitiveness. (See Note 2.) Fluctuations in temijerature
affect their strength, and their constant exhibition under active circum-
stances in the presence of sulphur compounds, cannot fail to produce
deleterious effects. (See Note 3.)

It certainly is not customary iu permanent engineering construc-
tions, bull J to withstand continuing, constant strains, to use material
that is subject to uncontrolled, cumulative, indefinite wear or abrasion.

Acid Process.

The manufacture of the bisulphite solution may be classified under-
three heads : The Vacuum Process, the Modified Tower Process, the
Tower Process.



276 MICHAELTS ON" LIME SULPHITE FIBER.

I will dwell but briefly upon tliis matter, confining myself to evident
features that would impress the average intelligent observer.

The vacuum system is used in connection Avitli the Partington, the
Sehenk and the Graham processes. It requires large exhaust pumps, a
series of tanks arranged vertically in echelon, a lime mixer, etc.,
and uudoul)tedly yields with certainty the high solution required. It
can unquestionably be used for all the processes. The best plant I saw
was at Birmingham, copied from the English Partington, and costing
here about six thousand dollars for the entire installation.

The modified tower system iu use with the Eitter-Kellner process at
Cornwall is a sort of cross between the Mitscherlich tower and vacuum
method. The solution is pumped by a battery of pumps into a series of
low towers under cover, filled with limestone. It did not strike us as
possessing merits over the vacuum method.

The Mitscherlich tower process is in a measure automatic and is cer-
tainly the most economical.

The sulphurous acid gas is drawn up the high towers, filled with
limestone, by atmosi^heric draft, and therein meets water trickling
through the filling. Its main disadvantage is the assurance of proper
draft. It strikes me that the application of a little American engineer-
ing skill would speedily bring about reliable and satisfactory results.

The consumption of sulphur varies from 200 pounds per ton of fiber
in the Mitscherlich, up to nearly 600 pounds iu the others. In none of
the others is it less than 350 to 400 pounds.

Mechaxical Pkepakation of the Wood.

All the isrocesses, except the Mitscherlich, use chips. In this latter,
discs cut from the log, li inches deep, are iised.

Dr. Mitscherlich claims that these discs afford a stronger fiber, and
that more bulk can be put into the digester, than if loosely piled chips
were used. One great disadvantage connected with the use of discs, is
that even if the finest cross-cut saws are used, circular or band, at least
ten per cent, of the wood is wasted in sawdust. Further, more time
must be consumed in charging the digester. All this is a mere question
of dollars and cents. There is no absolute objection to using chijjs in
the Mitscherlich process; wood would be saved, labor economized, and
time of boiling probably shortened. But, it is claimed by the Professor,
the outjiut is not so good. Hence, if the better product earns more,



MICHAELIS ON LIME SULPHITE FIBER. 2?7

even at the expense of material, labor and time, than the so-called infe-
rior, why of course it would pay to follow Mitscherlich's recommenda-
tions. It would not be a very significant item of expense to be prepared
to use both discs and chips.

Cooking.

"Cooking," or the treatment of the wood with calcic bisulphite to
obtain cellulose, is fully explained by Mr. Griffiu. I will merely dwell
long enough u^jon this branch of the subject, therefore, to make my own
impressions clear.

There are two methods pursued, a quick and a slow jDroct ss, ajiply-
ing the words merely to time, and not to output. The first process,
which is not under patent protection, is liursued in all the digesters
except the Mitscherlich ; the second, covered by an American patent, is
peculiar to the Mitscherlich.

The quick process requires for a boil strong liquor, high pressure,
short time, and yields a small product. The slow process requires a
weak liquor, low pressure, long time, but yields from four to eight times
greater a product.

The Mitscherlich, aside from every other consideration, api^ealed to
me as being logical and precise ; the other struck me as inconsequent
and unscientific.

In the Mitscherlich i^rocess, the cooking is done by indirect heat,
steam pipes ; in the other, by direct heat, the condensation of steam.
Hence the former method uses a liquor of uniform, the latter one of
constantly varying, strength. To illustrate, 2 600 gallons strong liquor
entered the Graham digester, a "quick" boiler, 2 000 gallons of con-
densed steam are added during cooking.

At one quick method plant we learned that it was frequently neces-
sary to add strong sulphite solution while the boiling was actually
going on.

If warm baths were prescribed for a patient, I imagine he would
protest vigorously against being immersed in boiling water, even if
assured that the cold water had been turned on. The application is
readily seen.

The quick process, as we gleaned from inquiries at the various plants
where it is carried on, requires from fifteen to twenty -four hours for
boiling ; the Mitscherlich, as shown by the Alpena experience, from



"278 MICHAELIS OX LIME SULPHITE FIBER.

iorty-five to seventy-two liours. It will thus be observed that the range
of variation, 60 per cent., is precisely the same for both processes. It
was noticeable that in certain cases, the best possible conditions of
time, temperature, etc., of the quick process were compared wuth the
most unfavorable showing of the slow.

This matter of time is apt to be misleading. The best claimed show-
ing was made by the Graham expert, nine complete "turns " per week,
or a single turn in about eighteen hours, producing 1^ tons of fiber.

The best figures for the Mitscherlich are a complete turn in seventy-
two hours, producing nearly 10 tons of fiber, four times as long, with
over six times the output.

Iq the Mitscherlich procedure, one operation impressed me as gro-
tesquely un-American. The cooked fiber is washed for ten hours iu the
digester. This great money-coining apparatus is compelled, for over
one-tenth of its time-capacity, to function as an ordinary washing
•engine. This is simjily an accidental part, in my judgment an evidence
of bad business management; yet the opponents are not slow in arguing,
from the condition imposed by this ill-judged operation, against the
whole method.

As cold water is used in this washing, they claim that the tempera-
ture of the Mitscherlich digester fluctuates from 40 to 270 degrees
Fahr., a range of 230 degrees, while in their process the temperature
never falls below 150 degrees, thus giving a range of only 150 degrees.

This objection is met by an obvious practical remedy, saving time,
and consequently money — the introduction of separate washing en-
gines.

In confirmation of Dr. Mitscherlich's claim that slow boiling affords
a better fiber, I have been told by an experienced soda pulja manufac-
turer that it was a well-known fact that in this latter process slow boil-
ing yielded a better output.

It is singular that while the sulphite process was the invention of an
American, Germany made it a practical success. We are now begin-
ning to take hold of this American idea, but burdened with German
methods, and iucrusted with Teutonic barnacles.

The most comislete quick process plant I saw was at Cornwall ; here
everything, including even the assembled digesters and the gauges re-
cording pressures in atmospheres, was inijiorted. The superintendent
told me that he thought the best course had been jiursued, that they



MICHAELIS OK LIME SULPHITE FIBER. 279

Lad the necessary experience abroad, and so on, and yet I noticed he
was modifying his apparatus about as ra^jidly as he couhL

The German adjuncts and methods at Alpena had to be entirely
thi'own aside, and apparatus and operations developed in our own paper-
making industry have bean introduced. There is a reason for all this;
in Germany labor is phlegmatic, inexpensive. I have been told that
women sort and inspect chips for twelve and a half cents per day, and
of course they have none but wooden ideas. As a consequence, the Ger-
man manufacturer, absorbing the phlegm, of his employees, has both
perseverance and patience.

We are quick-witted, we have abundance of jaerseverance in attain-
ing our ends, but precious little patience in waiting for results. To this
I ascribe, in a measure, the allurement which the quick process has for
us. We hug the delusion to our hearts that we can produce ten tons
just as quickly as one and a half, and mentally eliminate all intervening
insurmountable difficulties.

The most German of all the processes is the Mitscherlich, yet I
firmly believe that with American business management and American
engineering skill, it can be rehabilitated as a genuine American
method.

It is idle to catalogue the ponderous, painful steps that still charac-
terize it; there is not one that could not be practically improved. The
discs, for instance, are really fed, piece by piece, into the digester; they
should be automatically shoveled in bv the cart-load.

Astonishing as it may seem, the German engineers turned over the
first completed plant and acquiesced in its being run by an energetic
American, utterly without experience, who at the same time superin-
tended two saw-mills producing annually some forty million feet! It
recalls poor Colonel Sellers' " little side speculations."

To attain the best results, the head of the plant should be a sterling
business man, interested pecuniarily in its success, aided by competent
assistants in the technical, chemical and engineering departments.



Note 1. — It may be of interest briefly to investigate the stress to
which the lead coating in the Graham digester is subjected by change of
temperature.

The linear co-efficient of expansion of lead is 0.0000158

Its modulus of elasticity 720 000



280



:michaelis on lime sulphite fiber.



A range of 200 degrees is a very moderate assumption for tliis i^ro-
cess, for it boils at over 300 degrees.

We have, then, the stress p= .0000158 X 720 000 x 200

=- 2 275 i^ountls.

The elastic limit of lead is given by Trautwine as 1 100 ; we see.
therefore, that, even after making due allowance for the expansion of
the iron shell, the fixed lead is strained beyond its elastic limit. Every
plumber, from his own experience, can tell what will result.

Note 2. — Mr. Griffin's analysis of the Schenk boiler metal gives:

Copper 91.28

Tin 7 . 68

Zinc 0.89

In 1877 the British Admiralty determined the effect of heat upon the
kalchoid alloys of coj^per (see Engineering, October 5th, 1877).

*' The metal was cast in the form of rods one inch in diameter, and
composed of five diflfeient alloys as follows:

****** *******

"No. 2. — Copper, 91; Tin, 7; Zinc, 2. (The nearest apin-oach to
the Schenk mixture.)

*************

"The specimens were heated in an oil bath near the testing machine,
and the operation of fixing and breaking was rapidly and carefully per-
formed, so as to prevent, as far as possible, loss of heat by radiation.
At 100 degrees Fahrenheit the strength and diictility of the above test
piece was 525 ijounds and 15.5 per cent. At 300 degrees Fahrenheit
the strength was 265 pounds, the ductility nil."

I cite this as confirmatory of my statement regarding the sensitive-
ness of these alloys.

Note 3.— "It is but fair to add here that Professor Langley has
observed that acids and the fumes of his laboratory will change the very
structure of metals. He had in his laboratory a frame of the following
form :




GRIFFIN ON" CHEMISTRY OF LIME SULPHITE FIBER. 381

"The wires w, of copper, brass and German silver, were run through
the crossbars a, Avhich were of wood, about 2 inches wide. After three
years he noticed the wires breaking, and upon examination he found
them to be coarsely crystalline, brittle — in fact, rotten, and entirely
changed in structure. He found also that the jDarts that were in the
wood, and so protected from the fumes, were soft, ductile, and entirely
unaffected. All of the exposed wires were affected similarly, and all
the protected parts were equally unaffected, except the copper wire,
which was stiffened, but not materially changed in structure."* ,



SOME EEMARKS ON THE CHEMISTRY OF THE PROCESSES
OF LIME SULPHITE FIBER MANUFACTURE.



By Maetin L. GEiFriN, M.A.



I propose in this paper to give an outline simply of the chemistry of
the suljihite fiber industry. This will be concerned only with the
material for digesters, the bisulphite solution and the fiber.

By suljihite fiber is meant, fiber manufactured from wood for paper
making by the use of a solution of bisulphite of lime or magnesia. As
you are doubtless aware, the chief difficulty to be overcome in this pro-
cess is the securing of a digester which will stand the action of the acid
solution used to reduce the wood to fiber. I will first explain the com-
position and nature of this acid liquid. Sulphurous acid gives two
classes of salts, like carbonic, the bisulphite and the neutral or normal
sulphite. The normal sulphite of calcium has this formula Ga SO-^,
while the bisulphite requires two molecules of the acid radical giving
twice as much sulphur and expresses thus:

^>^ = 0,

The former is only soluble in 800 parts of water, though freely soluble
in aqueous sulphurous acid, thereby becoming the bisulphites. To

* steel: its Properties; its Use in Structures and in Heavy Guns. William Metcalf, M. Am.
Soc. C. E., Trans. Am. Soc. C. E., Vol. XVI, p. 291, June, 1887.



282 GRIFFIN ON CHEMISTRi" OF LIME SULPHITE FIBER.

obtain the solution desired for the mxnufacture of sulphite fiber we
should require for each part of calcium (the ratio of calcium to its
oxide, lime, being as 5 to 7) 3.2 parts of sulphurous acid gas or 1.6 parts
of sulphur.

This acid solution is mads in two principal ways and in all cases
the sulphui-ous gas is j)roduced from burning sulphur in retorts to
which only a limited supply of air is admitted. The gases are either
received into tall towers, as in the Mitscherlich system, where it is
absorbed by porous limestone over which water trickles downward,
meeting the ascending gases; or it is absorbed directly into a solution of
milk of lime by the aid of a vacuum pump. This is known as the
vacuum system. In the former case the sulphurous gas displaces the
carbonic, thus: Ca COj + 'SO.. = Ca SO 3 + CO 2, and we have the bi-
sulphite solution flowing off. In the latter there is no gas to be dis-
placed, the reaction being Ca {OH). + SO. = Ca SO-^ 4- H.O. By
this process a solution of any desired strength can be made.

It may seem to be a very simple operation to burn sulphur to sul-
jjhurous oxide; but in reality great care is required, since by the
admission of too much air to the retorts sulphuric oxide may be
formed. This uniting with the lime forms insoluble sulj^hate of lime,
which in the Mitscherlich towers would form a coating over the lime-
stone and cause a variety of troubles. In the vacuum system it would
form an inert deposit, causing a waste of lime and suljihur.

An excess of air will cause the sulphur to burn more rapidly, thus
developing too much heat, which facilitates the formation of first lower,
then higher, polythionic acids, then sublimed sulphur. These acids are
all sulpho-acids having an increasing proportion of sulphur. They not
only cause a waste of material, but exert injurious effects upon the
quality of the fiber. They also cause a fictitious strength of the liquor
which cannot be detected by the ordinary workman.

When we reflect what a peculiar substance sulphur is, and how easily
its chemical and physical properties are changed, we cannot wonder that
it should cause difficulties in any process in which it is used. At the
ordinary temperature sulphur is a light colored brittle solid; at 115
degrees C. it melts to a thin amber liquid; heated to 250 degrees
it boL'omes dark colored and thick; at 450 degrees it boils and passes
off as a dark colored vapor. These changes in its state of aggregation
are due to changes in the molecule, which varies in the number



GRIFFIN ON" CHEMISTRY OF LIME SULPHITE FIBER. 283"

of atoms it contains according to the temperature. At 1 000 degrees
and above it consists of two atoms, at 500 degrees it consists of six
atoms, and at lower temperatures it probably contains a still greater
number. This iDeculiarity makes possible a great variety of sulpho-
acids and salts. Sulphur is one of the most useful, powerful and
peculiar elements with which chemistry deals. Its gases are ijungeut
and penetrating, and its acids the most jjowerful and corrosive, forming
a most stable class of salts. A very small fraction of one per cent, in
metals renders them useless and it exerts injurious effects even in the
srnallest proportions.

Is it any wonder, then, that we should meet with difficulties in finding
a material for suljihite digesters ? Practically there is only one metal
yet discovered w^hich withstands the action of sulpho-acids fairly well;
it is lead. A substance very rich in silica is the only mineral. These
two kinds of material are all that have accomi)lished much so far in the
construction of suli^hite digesters. There are, however, digesters of
bronze in operation for which success is claimed. I will now briefly
describe those which are doing a representative business.

The Mitscherlich is 14 feet in diameter and 40 feet long, holding
about twenty -five cords of wood when cut up into discs li inches thick
and jjroduces at each "turn" 12^ tons of dry fiber. It is composed
of a wrought-iron or steel shell J inch thick, which is first coated
with pitch or tar within. Upon this is laid a continuous layer of
very thin sheet lead; next comes a course of bricks si^ecially made for
this i^urpose, each having a tongue and groove, which are laid flatwise
in the best Portland cement. Upon these is laid another course of
bricks edgewise, having their tongues and grooves arranged accordingly.
Each digester is stationary and placed horizontal.

The Schenk, a bronze digester, has the following composition:

Per Cent.

Tin 7.68

Copper 91 .28

Zinc 0.89

The specific gravity of a small casting was 8.5579; the computed
gravity is 8.76. The specific gravity of copper is 8.94. These figures
will indicate the difficulty in making perfect castings of copper alloys
free from combined and occluded oxygen. Hence it follows that the
larger or more numerous the molecular interstices, the more surface is
exposed and the more permeable is the metal by gases, and so more



284 GRIFFIN ON CHEMISTRY OF LIME SULPHITE FIBER.

affected. All metals are susceptible to this influence to a greater or less
extent. A very marked illustration of the absorjjtion of gas by a metal
is observed in the case of jjalladium, which absorbs 376 volumes of
hydrogen at the ordinary temperature.

According, therefore, to the nature of the gas, the metal and the
temijerature, will various results be observed. I am personally cog-
nizant of a case where lead, under the influence of sulphurous acid gas,
has been changed to a salt of this acid. Similar instances are recorded
of coj^per and other metals. Copper when plunged into molten sulphur
is immediately changed to the sxili^hide of the metal.

At the isresent time we see several manufacturers of bronzes claiming
that they have an acid-proof alloy, but no reputable chemist can enter-
tain such an idea for a moment. Neither is any copper alloy proof
against the action of sulphurous and sulphuric acids in any proportions.
The effect of these acids on digesters of copper alloys is, first, to dis-
solve a little from the surface, which is immediately reduced to the oxide
and sulphide of the metal by the reducing j)ower of the organic matter
dissolving from the wood. This soon forms a black coating on the
interior, which is continually crumbling off by the expansion and con-
traction of the metal, thereby contaminating the fiber with black sijecks.
As regards the life of such digesters, it is simply a question of the
rapidity of the action of the chemicals and the thickness of the metal.
It is not improbable, also, that the scale affords some slight protection
to the metal.

We have now left for our consideration digesters lead lined in dif-
ferent ways and of different shapes.

The Ritter-Kellner digester is vertical and stationary, about 10 feet
in diameter and 28 feet high, made of wrought-iron or steel, 1 inch thick.
The interior is lined with sheet lead half an inch thick. This is attached
to horizontal and vertical tenons of lead and antimony dovetailed inta
the seams. This gives an opi^ortunity for expansion of the lining in
sections. The ratio of the expansion of iron to lead is a little less thaa
1 to 3. As the digester is rejieatedly heated and cooled, there
must be a recurring movement back and forth from the shell, or
else the lining must "buckle." This movement, in time, is sufficient
to jiroduce cracks, which then must be rej^aired. This coincides with
experience.

The Partington digester is spherical and rotates upon an axis. It is.