Nov.,
T H E J O U R N A L OF 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 J L I S T R Y .
1912
Our laws are painfully strict in protecting so-called “vested property rights,” and this protection extends t o the most trivial matters. If Tom steals Dick’s two-dollar scarf pin, Dick will have no trouble in putting Tom in jail, even if Dick himself has obtained his pin by questionable methods. But when it comes t o protecting even for the short period of seventeen years, the most legitimate, the most truly personal property,
T
namely, intellectual property, with all that it involves, with enterprises depending thereon, based often on the work of a lifetime, then our law courts are wofully deficient. Any s y s t e m o j law which does izot adequately protect intellectual property so that rich avtd poor alike c a n uphold their rights, i s avt nnachrovLisiia in modern civilizatiou. L. H. BAEKELAND.
ORIGINAL PAPERS
DISTILLATION
OF
RESINOUS WOOD STEAM.’
By L. F HAWLEYA X D R. c.
BY SATURATED
PALMER.
COMMERCIAL STEAM DISTI LLATIO Pi P R O C E S S E S
.
The steam distillation process for obtaining the volatile oils from the wood of the longleaf pine has been the basis of a small industry since about 1903, being introduced apparently in a n attempt to produce wood turpentines a t temperatures lower than those used in the destructive distillation processes then in operation, and thus t o obtain a product uncontaminated by the decomposition products of wood and rosin. Quite a large number of plants have been built t o use either sawmill waste or lightwood, or both, but many have been abandoned, probably only 1 2 or ~j being in operation in 1 9 1 1 . The quality of the crude turpentine produced has usually been very good, but because this is the only Droduct obtained, or because the yield of this product is often lower than t h a t of “‘crude turpentines” from other processes, the plants have been successful only under especially favorable conditions. This process seems t o be very promising, however, when combined with other processes for the utilization of the steamed chips, as for instance the extraction of the chips with volatile solvents for the removal of the rosin. Conditions are also favorable for this process in cases where the material would be largely used as fuel or wasted, or is very cheap or so poor in quality t h a t more complicated processes would not be profitable ; these conditions are commonly realized in the case of t h a t part of the waste wood of sawmills now used as fuel a t the plant or burned on the rubbish pile. PURPOSE O F INVESTIGATION.
I n the fields mentioned above the steam distillation of resinous woods will undoubtedly expand and it was in the hope of promoting this expansion and thus increasing the utilization of a class of material now wasted that this investigation on the fundamentals of the process was undertaken. There has been no uniformity in co~nmercialpractice or in the opinions of the various operators and no experimental data have been published on the effects produced b y the different readily controlled variables, such as steam pressure, size of chips, or rapidity of distillation. I n the methods described in various patent specifications the greatest stress has been laid on the mechanPaper presented a t t h e Eighth International Congress of Applied Chemistry, Xew York, September, 1912.
789
1
ical features of charging and discharging, and of distributing the steam throughout the retort, which, although of great importance in the economy of a commercial method, throw no light on other equally important factors. There seemed t o be, therefore, a profitable field for investigation in determining the relations between the conditions under which the distillation is conducted, on one hand, and on the other hand the amount and kind of products and the readiness with which they are obtained. T H E O R E T I C A L CONSIDERATIONS.
I n the descpiption of the experimental work and the discussion of the results it will be necessary t o refer constantly t o certain theoretical principles which apply t o the distillation of volatile oils with steam, and in order t o make the future discussions clearer, a brief presentation of these principles is given a t this time. I n order t o simplify the deductions, the following assumptions are made in regard to the resinous material contained in the “lightwood” from the longleaf pine:‘ I . I t is composed only of turpentine, pine oil,* and rosin. 2. The components are all simple subst,ances completely soluble in one another. 3. None of the components are soluble in water. 4. The turpentine and pine oil are both volatile, but the turpentine has the lower boiling point. 5 . Rosin is nonvolatile. While these assumptions are not strictly true in all cases, none of them are sufficiently incorrect t o affect seriously the conclusions. Concerning the distillation with steam of the resinous material defined by the above assumptions, the following deductions can be made: 131. Vezes (Bull. SOC. Chim., 29, 470-478 (1903)) has given a very clear and complete discussion of the principles underlying the distillation of the oleoresin from the Maritime pine of France. For the simplification of the discussion the following assumptions were made: 1. That the oeloresin is composed only of essence (turpentine) and colophony (rosin). 2. T h a t these components are both simple substances completely soluble in each other. 3 . T h a t neither is soluble in water. 4 . T h a t rosin is nonvolatile. I n regard t o the oleoresin obtained by chipping the livelong leaf pine tree of the United States, the same assumption can be made, b u t the oleoresin contained in the pitchy “lightwood” from this species has another component, the heavy, high-boiling ”pine oil” which must be taken into consideration in the discussion of the steam distillation of such wood. A discussion of the occurrence of pine oil in the “lightwood” of longleaf pine is given in Forest Service Bulletin 10s (“Wood Turpentines, their Analysis, Refining and Composition,” by L. F. Hawley).
790
T H E J O U R N A L OF I A T 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 .
I . There will be a separation of the two volatile constituents, the turpentine being in greater proportion in the first part of the distillate and the pine oil in the latter part. 2 . The temperature of the distillation under normal pressure will be slightly above 95 O C . a t the beginning, the distillation,never - and will rise throughout quite reaching 100' C., however, as long as any of the turpentine or pine oil remains undistilled. 3. If the pressure a t which the distillation is carried on is increased the temperature mill be increased, the tempera-0 ture depending upon the pressure and on the concentrations of the turpentine, pine oil, and rosin; the temperature will, however, never reach the steam temperature corresponding to the pressures used as long as any turpentine or pine oil remains undistilled. 4. The proportion of mater t o oil in the distillate will increase as the distillation progresses, this proportion being influenced only b y the relative amounts of the different constituents present in the oleoresin being distilled.1 W' I n these deductions i t is considered t h a t the system is in complete equilibrium and under such conditions the behavior of this oleoresin when distilled with steam could be foretold with considerable accuracy, but in the distillation of wood containing the oleoresin, there is a disturbing factor introduced which makes necessary the investigation of the variable mentioned in the introduction. This disturbing factor is the difficulty of keeping a complete equilibrium betm-een the oleoresin and the steam, due to the fact t h a t the wood surrounds the oleoresin and tends t o keep the steam from coming in contact with it. The effects of the size of chips and the rapidity of distillation are not due t o a change of the laws governing the distillation of the oleoresin with steam, but t o the manner in which they affect the completeness of the equilibrium in the system, or in other words, the completeness of contact between the oleoresin and the steam. I n the same way the effects (other than temperature changes) of steam pressure on the results of the distillation are also due to its influence on the completeness of contact between steam and oleoresin rather than to any influence on the behavior of the steam and oleoresin when completely in equilibrium. E X P E R I M E N TA . L M ETH 0 D S .
APPARATUS AND MATERIALS
The general procedure was to distil charges of the same sized chips under different conditions, and of dif1 A change in the pressure a t uhich the distillation IS carried on might change the proportion of water to oil in the distillate but this change would be only slight and there is not sufficient data available t o decide even the direction of this change.
Nov., 1912
ferent sized chips under identical conditions, and to note carefully the variations in the results of the distillations. It was not practicable t o make all the runs on exactly comparable material, because a large number of charges of equal resin content could not be secured, and because the material could not be kept without some loss of volatile oil b y evaporation. The runs
/5-
FIG. 1.-Experimental
retort.
were therefore made in groups with comparable material in each group; the results from each group were made comparable t o some extent with those from other groups by a comparison of the resin content of the different groups. This was readily accomplished b y distilling the sawdust obtained in the preparation of the material for each group, as this sawdust was a good average sample of the whole group. A sketch of the retort used for the distillations is shown in Fig. I . This retort was designed for use in extraction of tannin from mood and bark, as well as for distillation experiments ; i t mas therefore made of copper and had the closed steam coils R. The perforated plate C was designed to hold up the charge while the lower head of the retort was taken off;then by revolving the plate until the slots E came opposite the supporting lugs F, the plate falls and the charge can be removed. The pressure gage was attached a t D . The other parts of the retort require no special description. An ordinary copper worm condenser was used and the distillate was caught in I-liter graduated cylinders. The J . G. Sewman Lumber Co., of Hattiesburg, Miss., furnished for the experiments a large log of pitchy light-
T H E J O U R N A L OF I l Y D L S T R I A L A N D ENGIXEERIATG C H E i W I S T R Y .
KO\-., 1 9 1 2
wood in which the pitch was distributed with unusual evenness. From this log it was possible t o prepare a number of charges of chips, of which all from one portion of the log were similar in resin content. PREPARATION OF MATERIAL.
The method of selecting material so that each charge should be comparable with the other charges in the same group is shown in Table I and Fig. 2 . The TABLE~.-EXPERIXESTAL R u s s MADE ASD JIATERIAL L-SEDIS EACH.
r:i I
i- s Y = , 1
A
2 3
A A A
4 5
I1
6 i 8 9 10 11 12 13
Sticks.
I to I to I to I to
A
A A
I
A
VII.. . VII.. VII.. \TI..
A
s x
the three runs. S a x d u s t Obtained in sawing slabs. 1 s variable 6 x same sections O f 'labs'
23, 26, 27, 3 1 ,
........................ ..........................
A
I t o VI1
.
i
)
1, 7, 8, 12, 16
Inches. of section
........................
li 18
B
' to I toVII.. .
l9 20 21
B
I to'I'' ' ' ' I toVII...
22
c
{
C
{
'
'
t
..{
1 , 4 , 7 , 1 0. . . . . 2 , 5 , 8, 11
.....
3 , 6 , 9 , 1 2. . . . . 1 , 4 , 7 , 1 0. . . . . 2 , 5 , 8 , 11 . . . . . 3 , 6 , 9 , 1 2. . . . . 1 , 4 , 7 , 10 . . . . . 2 , 5 , 8 , 11 . . . . . 3 , 6, 9, 1 2 . . . . .
2x2
t
2x2
1
4x4
.......... ..............
D
26
D
I t o VIII..
27 28 29
D
I t o VIII.. 2 , 5 , 8 . . . . . . . . I to VIII.. 3 . 7 . . . . . . . . . .
1, 4 , 6, 9 . .
....
. ., , , , ., .. .............. v. . . . . . 1 , 5 , 9 , 1 3.....
E
x I,'? Sawdust Obtained in cutting sections.
25
D
i
} All ei.en numbers
c .......... .............. I, I V , VII. 111. VI. ... 11, v . . .... 111, VI. ... 11, V... . . I, I V , 7-11. 11, V...... I, I V , VII. 111, VI. ...
t
One-half of each section used in 20, 24, 32, 33 1/2 of section eachrun. Sawdust Obtained in cutting sections. Chips from each secAll odd numbers '/ax tion divided between each run. I/a?
Sawdust Obtained in cutting sections and in slabbing blocks 1s 1 Obtained in cutting sections. I s 1 1x1 Sawdust
I I,
I I, v.. .... 11, VI... . 111, VI1 ...
31
E
32
IV.. (11. V . ..... 11, V I . . . . E 111, VII.. .
I
1
......
11-. . . . . . . .
f I, V . .. . . . 33
E
I 11, VI..... 1 11, VI1
....
1 I\-........
.. '
..
3, 7 , 11, 1 5 . . I 2, 6, 10, 1 4 . , ' . i 1, 5 . 9, 13 . . . . . j 4,8, 12, 1 6 . . . . J 2 , 6. 10. 14 . . . . 1 1, 5 , 9, 1 3 . . 4,8, 12, 16. . . . l"'x 3 , 7 , 11, 15 . . . . j 4, 8, 12, 1 6 . . . . 1' 3, 7 , 11, 15 . . . . 1 2, 6, 10, 1 4 . . l " ? s'"I 1, 5 , 9, 13.. . . .
... ..
1
DISTILLATION.
Only a general description of the procedure will be given here; the same general methods were used in all cases and the details are given later in the tabulated record of the runs. REGULATION.
..............
15 A 16 B
VI1
....
Chips from each sec-
IV
VI
1x 1
I t o VI1
4 A
I t o S'II..
24
...
10, 22, 2 8 . . 6, 18, 2 9 . . . . . . 4, 17, 3 0 . . 2, 14, 2 5 . . . . . .
. . .
runs of a group had been longer cut than those of the first runs, and although they were kept in covered cans, or if without cover were piled close together, there was still some chance of loss by volatilization, especially when some time intervened, as in the large number of runs in Group I. To lessen the loss by volatilization during the chipping process, the chips were covered with water as fast as they were made. The sawdust charges were put into the retort without any delay after collection, except in runs 2 and 3, which were prepared and collected but not distilled at the same time with run I .
Remarks.
Inches. Sawdust Obtained in cutting sections.
........................ ........................
14 A
23
Size of chips.
..............
I11
T
Sections.
791
'"
1
1
sticks were sawed into sections a t one time, but usually the sections were chipped only as required; t h a t is, just before each distillation. Thus the sections for the last
The speed and pressure of the distillations vere regulated simply b y the inlet valve a t the bottom and by the outlet valve a t the top of the retort. During distillation a t atmospheric pressure the outlet valve was left open and the speed regulated by the inlet valve alone ; but during distillation a t higher pressure i t was, of course, necessary t o use both valves in order to keep both speed and pressure a t the required values. The variation in speed was never more than one-half minute per liter of distillate, and in pressure never more than 2 pounds per square inch from the required values. Data Obtained.--A typical data sheet giving a n example of the records obtained in each distillation is shown in Table 11. The distillate was caught in oneliter fractions and the time and pressure were recorded with every fraction. The amount of oil in each liter of distillate was determined as accurately as possible while in the original receiver (one-liter cylinders graduated to I O cc.) ; the oil was then separated from the water in a separatory funnel and its specific gravity determined. When the amount of oil in the fractions was so small t h a t a specific gravity determination could not readily be made on a single fraction, the oil from a sufficient number of fractions mas combined t o make the determination possible. As a check on the total amount of oil computed from the rough measurements on the separate fractions, the combined oil from all the fractions or from the fractions of different portions of the distillation was accurately measured. The detailed records for each distillation will not be given in this report, but instead the essential part of this data will be tabulated together with other factors computed from the data. The discussion of the data from all groups will be taken up together after the experimental work has been described. End Point.-It will be noticed in Table I1 that the amount of oil in one liter of distillate gradually decreased and decreased very slomly toward the end of a distillation. O n this account it became necessary t o arbitrarily choose a ratio of oil to water 1%-hich might be considered a s the end of a distillation under one set of conditions in order to make the results of differ-
792
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 ENGINEERING C H E M I S T R Y .
ent distillations comparable and a t the same time to keep the time required within practicable limits. I n the first five runs the end point was too high and too variable, since it was not recognized t h a t the end point must be very carefully regulated in order to obtain comparable results. These first five runs are, therefore, not comparable with those which follow, in which the end point was carefully regulated a t somewhat lower values. It was also found that after distilling a charge under one set of conditions until a certain end point was reached, if the distillation was
Nov., 1 9 1 2
as important until after Run 5 was finished, so t h a t €or another reason Runs I t o 5 are not comparable with those which follow. The proper end point for any distillation after Run 5 was considered to be reached only when the required ratio of oil to water in the distillate had been attained both before and after interruption of the distillation. DISCUSSION O F R E S U L T S .
The data obtained by the distillation of the various runs are given in Table 111, together with the condi-
SECTIONS CUT PROM STICKS OF BLOCK
A
SECTIONS CUT FROM STICKS O F BLOCK
c
METHOD OF CUTTING ALL BLOCKS INTO STlCKS
FIG.?.-Diagram
SECTIONS CUT FROM SECTIONS CUT FROM STICKS OF BLOCKS BmE STICKS OF BLOCK showing method of cutting up material. (In thichess the sections are drawn to scale, the amallest being 1 inch thick.)
interrupted for a n hour or more and then continued under the same conditions a s before, a further supply of oil could be obtained before the same end point was reached again. This additional amount of oil obtained as a result of interrupting the distillation amounted to from 2 per cent. t o 18.8 per cent. of the oil already obtained before the distillation was interrupted. The highest increases in yields due t o this manipulation were those runs where there was still considerable oil present in the wood when the distillation was interrupted (although, of course, all the oil possible had been removed under the prevailing conditions). For instance, in Runs 14 and 15, after all the oil possible had been distilled a t 5 0 pounds pressure, interruption of the distillations over night made i t possible t o distil respectively 1 8 . 8 per cent. and 15. I per cent. more oil under the same conditions; in both these cases there was still considerable oil present in the wood a s shown by further distillation a t increased pressures. I n Runs 11 and 2 1 , however, after all the oil possible had been distilled a t atmospheric pressure, interruption of the distillations made i t possible t o obtain only 2.9 per cent. and 4.0 per cent. more oil under the same conditions; in these cases there were much smaller quantities of oil left in the wood than in Runs 14 and 15. This effect also was not recognized
tions under which the distillations were made. In the table the size of the chip a s given in column two is expressed in inches with the length parallel t o the grain given first. The values given in the columns headed “yield” are expressed in cc. of oil per pound of wood. The possible error in these determinations of yields is apparently about 6 t o 7 per cent. and is due t o difficulties in sampling, in regulating evaporation during the preparation of material, and in obtaining comparable end points in different distillations. An example of results which must be due t o such errors is seen in Runs 23 and 2 4 . The chips in Run 24 are larger than those in Run 2 3 and the yield should be perhaps less and certainly not greater from the larger chips, and yet the yields obtained from Run 23 are 5 . 0 per cent. less than those from Run 24. Another similar example is shown in Runs 30 and 31. I t might be thought t h a t some of these variations in yields were due t o incomplete distillation caused by the “channeling” of the steam through the charge in such a way that part of the wood was never touched by the steam, but in several runs after all the oil possible had been distilled under some one set of conditions the top of the retort was removed, the charge well stirred, and the distillation continued under the
Nov.,
T H E J O U R N A L OF I N D U S T R I A L A N D EAVGIiVEERING C H E X I S T R Y .
rg12
same conditions as before without any indications t h a t the stirring had discovered undistilled material in the charge. I t seems probable, therefore, that with a retort of the shape and size used in these distillations, the effect of incomplete distillation due to inTABLEII.-TYPICAL DATASHEET,ILLUSTRATING THE RECORDS TAXEXOF THE DISTILLATIONS. Project 123-Run No. 23. Chips 1” X 2“ X 2”. Weight of can water160 Ibs. Weight of can water chips215 Ibs. Weight of chips55 lbs. Steam turned on 9.23 9.28.30 Distillate began to flowCc. Cc. of oil Liters of oil Time. total CC. of Specific 7 of dis- in dis- com- oil total gravity tillate. tillate. puted. determined. of oil. Pressure. Hrs . .Min. 9 1 88 36 88 0.8810 70 2 154 48 242 9 0.8794 72 9 3 110 59 % 352 362 0.8851 71 4 73 425 8 10 0.8944 j1 17 5 60 485 10 68 10 6 55 29 % 540 0.8980 71 10 40 H 7 50 590 70 10 8 44 51 H 634 0.8981 11 9 39 673
+ +
+
} ). }
H
10 8 54 11 18 %. 12 28 13 38 47 % 14 15 57 % 7 16 17 17 18 27 % 19 38 % 50 20 0 21 22 9 23 18 % 24 27 % 25 39 48 26 27 0 28 11 % 29 23 % 30 34 % 31 45 !4 32 56 !4 33 7% 34 17 35 26 %. 37 36 37 43 3 38 45 Distillation interrupted 9 23% 39 9 35 40 9 46 41 9 46 41 9 57 42 10 8 43 10 19 44 10 30 45 10 43 46 10 53 47 11 4 48 11 6 49 No. of liters. 16 25 38 49 11 11 11 11 11 11 12 12 12 12 12 1 1 1 1 1 1 2 2 2 2 2 2 3 3 3 3 3
24 697 27 724 26 750 19 769 22 791 17 808 17 825 16 84 1 16 85 7 18 875 19 894 15 909 13 922 11 933 10 943 12 955 10 965 11 976 10 986 12 998 8 1006 8 1016 10 1026 13 1039 6 1045 5 1050 4 1054 1 105 i 1 1056 over night. 22 1078 17 1095 15 1110 15 1110 10 1120 9 1129 8 1137 7 1143 15 1158 8 1166 6 1172 2 1174 Yield cc. lb. 15.3 17.8 19.9 22.4
/-
.... 839
1
1 }
0.9010
69 72 0.9016
0.8987
i
979
0.8965
J
1
I
0.9858
1047
J
1095
1235
1
0.8936
}
0.8934
1
0.9003
I
I1
zi 69 69
1
-I
ii ;i
;i
64 71 72 71 69 68 70 70 68 72 73 72 63 45 32 20 12 4 0 j2
73 71 71
:i
72 58 34 0.8935
’i 0
Efficiency. 0.95 0.71 0.51 0.46
complete contact between the steam and the surface of the chips is negligible. The values given under “efficiency” are obtained by dividing the yields per pound of wood by the number of liters of total distillate, the efficiency factor being cubic centimeters of oil per pound of wood per liter of distillate
793
cubic centimeters oil pounds wood-x liters distillate..
It might be thought that this “efficiency factor” would have more significance if it represented only the relation between oil and total distillate, but, as will be seen later,’ this relation would be affected by the amount of wood distilled. Of course, the effect may not be in exact proportion to the amount of wood distilled as represented in the factor used, but it is thought that more nearly comparable efficiency factors are obtained by including the amount of wood as above. These factors represent approximately the relative amounts of oil obtained in the different runs per unit of steam consumed, exclusive of the steam which supplies the heat lost by radiation. Or
EFFECT OF SIZEOF CHIP O N YIELDA N D EFFICIENCY.
In general, other conditions being the same, the smaller the chip, the larger the yields and the higher the efficiency. This is shown in Table IV, which contains selected data from Table 111. Four groups of distillations are given, in each of which all the other conditions except size of chip are as nearly as possible the same, and in every case the smaller-sized chips show the larger yield and higher efficiency. The effect on yield is not so marked in the case of Runs 26 and 2 7 (and some of the other runs given in Table 111), but this is accounted for by the fact that the pressure was high enough so that nearly all the oil was removed even from the larger-sized chips. I n the case of two runs in which all the oil was removed even from the larger chips, the yields would of course be the same, but the efficiency would probably be higher with the smaller-sized chips. EFFECTOF PRESSUREO N YIELDA N D EFFICIENCY.
I n general, other conditions being the same, higher pressures give larger yields without lowering the efficiency. This is shown in Table V, which gives three groups of runs, in each of which all conditions except steam pressure are as nearly as possible the same. I n all cases the higher steam pressure produced the larger yield and with the same or higher efficiency. The effect of pressure on yields is also shown in another way in many of the runs in Table 111, in which, after obtaining all the oil possible by distilling under one pressure, a further yield of oil was obtained by continuing the distillation under a higher pressure. EFFECTOF SPEEDO F DISTILLATION ON YIELDA N D
EFFICIENCY.
Other conditions being the same, increased speed of distillation decreases both the yield and the efficiency. This is shown clearly in Table VI, which gives the results of two sets of two runs each, all the conditions except the speed being the same in each set. The more rapid passage of the steam through the charge probably causes it to be less saturated with the oil vapors, thus directly decreasing the efficiency. The yield is decreased probably because the same end point is reached sooner when the steam is less completely saturated. This is indicated b y the more nearly equal total yields obtained in each set of runs by finishing up the distillations a t the same pressure 1 The same reasoning as is given on page 794 regarding the effect of the size of retort on the efficiency applies also to the effect of the amount of wood distilled on the efficiency.
T H E J O U R S A L OF I , V D U S T R I A L AA’D ESGI-\-EERISG
794
CHEXISTRY.
Nov.,
1912
TABLEIII.-SU&lMARY 0 lbs. pressure.
Group Run
No.
170.
Yield, cc. Effi- Yield cc. ciencv. per lb. per lb. Size of chips. 1. i l 20.6 Sawdust .... .... .... Sawdust 1.56 .... 12.5 Sawdust 2“ 1” X 1” .... .... 211 I/ X l/8Jf 0.61 5.8 6.4 5.1’ 8.0 0.55 2” X X“ x 1 / P 0.63 15.7 4.61 1” x 1 / 4 ” x 1/8” 1” 52’’ X 1/4” 6.51 0.46 10.2 . . . . 22.42 11, X 95.;” x ‘/a“ .... .... .... 1“ x X’’x 1/4“ 27.7 1 .46 Sawdust 1” X variable .... .... .... .... .... 6“ X variable 1 I x 5“ X 8” .... .... 1” X Y2’’x 1 %” .... 25.2 0.97 .... Sawdust 1” x X l/8” .... 24.52 .... 1” x x ‘/a” .... ... .... .... 1” x %” X %“ .... 1” X %” x W” .... 24.6 1 .05 .... Sairdust 1” X 2” x 2” .... .... 1” x 2” X 2(’ 1” x 4” x 4” .... 24.0 0.89 .... Sawdust 1” x 1” x 1” .... .... .... .... 2” x 1” x 1” 3” x 1” x 1” 19.3 0.84 .... Sawdust .... 18.8 0.94 Shavings 0.6 .... .... 19.4 .... 0.85
....
1
2 3 4 I. 5 6 7 8 9 10 11 11. 12 13 111. 14 15 16 17 IV. 18 19 20 21 v. 22 23 24 25 VI. 26 27 28 29 30
....
x x x
.... ....
....
....
....
....
VII. 31
Shavings
1” X %“ X
32
33
1”
.... ....
....
16.0 1.4 17.4
0.84
.... ....
X“
.... ....
x X“ x X“
.... .... ....
....
....
.... ....
0.70
....
.... .... ....
....
.... .... ....
= 20 lbs.
2
-
....
9.93 32.73 8.23
....
....
0.45 0.28 0.29 0.25 0.46
3.93 6.13 5.33 62 3 3.04 27.24 2 .03
....
.... ....
....
.... ....
.... ....
Run S o . Size of chip. 7 1” X ’/a’’ X ]/a‘‘
14
1” X 5” X 8”
26 27 2s
1” X 1” X 1” 2“ X 1“ X 1” “ x 1” x 1”
OF
....
......
....
..,.
....
.... . . .
....
20 3 9.1
0.33 0.33
0.33 0.24 0.14 0.19 0.19 0.46 0.25
.,.. 19.74 24.54 4.03
0.35 0.41 0.31
....
4 5 1.5
0.21 0.12
....
....
25.7 12.3 5.9 4.7
0.49 0.29 0.17 0.15
.... .... ....
..,.
26.84 25.34
0.50 0.55
.... ....
....
....
29.8
0.23
....
0.63
....
23.1 22.4 23.6
0.26
....
0.39 0.46 0.44
.... ....
22.8 21.4 12.9
0.45 0.31 0.29
....
.... ....
.... .... ....
1 .i3
.... ....
.... 2.63
....
.... .... ....
....
.... ....
....
.... ....
....
....
.... .... ....
3.13
0.39
.... ....
....
.... .... .... .... ....
....
....
1.7=
.... ....
....
.... 3.83
.... .... .... .... ....
.... ....
= 30 lbs
0.li
.... .... ....
;:;;).
10
10
10
10
10
10
If, as seems probable, the effects of speed are due t o the variations in the time during which the steam is in contact with the wood, then the size of the retort would have a similar effect, that is, a speed of I O minutes per liter in a certain sized retort would be equivalent t o 5 minutes per liter in a retort twice as
.... ....
.... ....
.... .... ....
....
.... ....
.... .... ....
20.1 0.7 . . . . 20.8 . . . . 16.1 .... 2.2 20.3 3 = 40 Ibs
....
....
....
....
....
.... ....
29.8 26.3 23.1 22.4 23.8 26.6 22.8 21.4 12.9 21 .o
0.63 0.82 0.39 0.46 0.44 0.70 0.45 0.31 0.29 0.64
.... .... .... ....
.... ....
0.90 1.26 0.74 0.33 0.38 0.34 0.40 0.29 0.34 0.40 1.1 0.49 0.29 0.28 0.32 0.75
.... ....
.... ....
0.38
....
.... ....
.... .
I
.
.
22.5
0.73
.... ....
....
....
.... ....
21.2
0.60
0.39 0.39
20.8
0.39
....
.... ....
0.33
20.3
0.33
.... .... ....
....
...
.... ....
0.43
....
....
....
....
....
....
4
4 4 4 5 4 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 6 10 6 and 10 6 and 3 10 3 and 10 3 and 6 10 6 and 3 10 3 and = 50 Ibs.
E n d point, cc. per liter. 25 and 12 18 17 and 12 12 12 10 10
10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 5 7 10 5 and 7 10 10 5 a n d 7 5
10 10
10
10
7 5 and 7 10 5 and7 5 7 5 and 7 5 7 5 and 7
large, since a unit of steam would be in contact with a unit of wood for the same length of time in either case. RELATIOSS BETWEEN ENDPOINT, YIELIIAND EFFICIENCY.
As shown in Table 11, the amount of oil in a liter of total distillate is greatest a t the beginning of the distillation and decreases steadily as the distillation
End point, Speed, cc. min. oil per Effi- per literdisYields. ciency. liter. tillate.
0.45 0.31) 0.29
....
....
....
....
30.5 32.7 20.7 20.3 25.2 19.2 25.6 23 .O 29.9 28.7 29.7 25.7 12.3 25.6 29.2 29.1
....
.... .... .... ....
....
::;;I
.... ....
....
SIZEOF CHIP O N YIELDAND EFFICIENCY.
Pressures.
....
.... ....
....
but a t lower speeds. The variation in efficiency is not, however, as might be expected, exactly proportional t o the speed, since doubling the speed decreases the efficiency by only about I O per cent. from 0 . 9 4 to 0.84 in Runs 30 and 31, and from 0.43 to 0.39 in Runs 3 2 and 33. TABLEIV.-EFFECT
0.45 1 26 0 41
0.50
.... ....
.... .... ....
....
....
....
....
.... 1
RECORD OF EXPERIZlESThL RCXS. 40 and 50 Ibs. pressure. 7 0 lbs. pressure. Total. ? 7----,--Effi- Yield, cc. Effi- Yield, cc. Effi- Yield. cc. Effi- Speed, min. ciency per Ib ciency.. per lb. ciency. per liter. ciency. per lb.
20 and 30 lbs. pressure.
TABLEV.-EFFECT
Run No. 8 9 10 17 18 19 20
1” 1” 1“ 1“ 1” 1” 1”
OF
Size of chip. X %” X ., . X 54’‘ X X %” X I,’*‘ . . . . X 1/4‘‘ X >/ a”..
.. .
X I,”‘’ X * / 8 “ . . . X 3” X %”. , . , X 3.i” X %”. , , ,
PRESSURE O N YIELDA N D EFFICIEWY. End point, Speed, CC. min. oil per Effiper liter disPressures. Yields. ciency. liter. tillate. Atmospheric 10.2 0.46 10 10 22.4 0.46 10 10 30 pounds 50 pounds 27.2 0.46 10 10 0.50 10 10 30 pounds 24.5 26.8 0.50 10 10 50 pounds 50 pounds 25.3 0.55 10 10 0.63 10 10 io pounds 29.8
progresses, except when the conditions of distillation are changed, and then the increase in the amount of oil per liter is usually only slight and temporary. This was true in all the distillations and it is evident, therefore, t h a t the efficiency factor will decrease steadily throughout the distillation and its final value will
depend on the end point used. Therefore, the efficiency can be increased b y stopping the distillation before all the oil possible has been obtained and thus decreasing the total yield of oil. For the same reason the efficiency will be decreased by continuing the disTABLEVI-EFFECT
30
Shavings.. . . .
31
Shavings.. . . .
32
1” X
7;’’ X :4”..
33
1” X
::”X %” ...
{
J\
OF
EFFECT OF
SPEEDo s YIELDA X D EFFICIESCS.
Atmospheric Atmospheric 4 0 pounds Atmospheric Atmospheric 4 0 pounds 70 pounds 7 0 pounds
6 10 10 3
10 10
{ {
5
7 10 5
i 10
5 1;
i
1;
5 5
18.8 0.6 3.1 16.0 1.4 3.8 20.1 0.7 18.1 2.2
18.8 19.4 22.5 16.0 17.4 21.2 20.1 20.8 18.1 20 3
0.94
0.94 0.89 0.39 0.73 0.84 0.84 , ,, , 0.70 0.38 0.60 0.43 0.43 _ _ . .0 . 3 9 0.39 0.39 , . . . 0.33
....
tillation until all the oil possible has been obtained and thus increasing the total yield of oil. For instance, in Run No. 2 3 (Table 111) if the distillation had been stopped with a n end point of 1 2 cc. per liter, a t the 25th liter the yield would have been only I j .3 cc. per pound while the efficiency would have been 0 . 9 5 . By continuing the distillation until the end point (after a n interruption of the distillation) was I O cc. per liter, a much larger yield, 2 2 . 4 cc. per pound, was obtained, but only a t the expense of a much decreased efficiency ( 0 . 4 6 ) . THEPRESSURES REQUIRED
larger than 2 inches with the grain and 4 by 4 inches across the grain and 24). It is probable t h a t 80 to oil could be removed from chips 2 ” tillation a t 7 0 pounds pressure.
TO
DISTIL COMPLETELYDIFFERENTSIZES OF MATERIAL.
Samittst.-The volatile oil cannot be completely distilled a t atmospheric pressure even from a material as finely divided as sawdust. This can be seen from Runs 1 1 , 16, 21, 2 5 and 29 (Table 111) in which, after removing all the oil possible a t atmospheric pressure, a further distillation a t 40 pounds pressure removed from 6 . 6 t o 1 5 . 8 per cent. more oil. I t was found t h a t after distilling a t 40 pounds pressure a further distillation a t 70 pounds was without appreciable effect. I t can be safely stated that 40 pounds pressure is sufficient for the removal of all the volatile oil from material as small as sawdust. I t is possible that lower pressures might give almost as good results, but this point cannot be determined from the data on hand. Chips I” X I / / X */,”.-This size material cannot be completely distilled a t 30 pounds pressure (Run I 7) and probably not a t 50 pounds pressure (Run I 8 ) . I n Run 18 the yield obtained from chips I” X * / 4 f f x I / 8 f f is almost the same, within the limit of possible variation, as from the sawdust of Run 16, but apparently not quite all the oil has been removed. Chips I ” X X I/,”.--Chips of this size cannot be completely distilled a t jo pounds pressure (Run I S ) , but can a t 7 0 pounds (Runs 2 0 , 3 2 , and 3 3 ) . Chips larger than I” X I,’,~~ X =/,“.-At the maximum pressure used, 70 pounds, chips larger than I ” x I/~“ X cannot be completely distilled, but as the size of chips is increased there is no sudden drop in the yields obtainable a t this pressure until sizes
(Runs 2 7 and 2S)I are used (Runs 14 8 5 per cent. of the X 4” X 4’f by dis-
P R E S S U R E O S COMPOSITION O F O I L .
Analyses were made by the method described in Forest Service Bulletin 105 of part or all of the oil from each of the runs, but there was not enough difference between the various samples so that all of the analyses will need to be given. A few distillation curves showing the main points of interest will be given. Pitze Oil.-The proportion of pine oil in the crude turpentine did not vary except in cases which could be explained b y variation in other factors besides pressure and, therefore, so far as the results show, the pressure has no influence on the proportion of pine oil except the influence due to increasing the total yields. I n all cases where the total oil obtained was analyzed, and where the oil was nearly completely removed from the wood, the percentage of pine oil by weight varied only between 48 per Cent. and 5 2 per cent.2 I n cases where only part of the oil was removed b y the distillation, as in Run 4 , the proportion of pine oil was less. Dipeiztem.-The detection of small differences in the proportion of dipentene present cannot be made b y the method of examination used, especially n-hen such large proportions of pine oil are present. There seemed t o be, however, more dipentene in the crude turpentines produced a t higher pressures. Figs. 3 and 4, representing the distillation curves obtained in the analyses of the oils from Runs 2 I and 23, respectively, illustrate this point. The oil obtained from sawdust mostly a t atmospheric pressure (Fig. 3 ) apparently contains less dipentene than the oil distilled entirely a t 6 0 pounds pressure (Fig. 4), the specific gravity values being lower and the proportion of the oil boiling between 165’ and 180’ larger in the latter case. I t had formerly been thought that the dipentene which had been found in wood turpentines was caused by the temperature used in distilling the oil from the wood, but indications of dipentene were found in all the samples of oil obtained in this investigation, even in those produced a t atmos1
I n determining the percentage of total oil obtained in the runs of Group
VI it must he remembered t h a t the sawdust of Run 25 was not exactly a representative sample of t h e material of t h a t group b u t t h a t i t was mixed with the sawdust obtained in cutting the slabs from the blocks. I n several cases similar samples of sawdust obtained in cutting t h e slabs had been distilled and found t o contain more volatile oil than the sawdust obtained in cutting t h e blocks. It is probable, therefore, t h a t the proportion of volatile oil in the mixed sample of sawdust was somewhat greater than in the rest of the material in this group. This point is further indicated by a comparison of the yields obtained from the sawdust runs in the different groups. With the exception of Group VI (Run 25) the yields from the sawdust decrease, as might be expected, for the reason t h a t the material for the groups was cut from the same log in order of the group number. beginning a t the b u t t end and the content of volatile oil in the b u t t end was higher than in the upper portions of the log. It is probable, therefore, t h a t a value of 24 t o 25 cc. oil per pound of wood would more nearly represent t h e volatile oil content of the group. 2 Exceptions were found t o this in the oils from Group I1 which contained about 28 percent. pine oil. B u t the material for this group did not represent a complete cross section of the log, being composed instead only of t h e outside pieces, the slabs. The outer layers of this log evidently contained a smaller proportion of pine oil than the rest of the n-ood.
T H E J O U R N A L OF 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 .
796
pheric pressure, and it is very probable t h a t dipentene was present as such in the wood distilled. I n order t o make sure t h a t this material with low specific gravity and high boiling point was dipentene and not some other terpene with similar physical properties, a chemical examination was made of the fractions 165' t o 185' from some of the turpentines produced at atmospheric pressure and dipentene was identified by means of the tetrabromide, m. p. 125'-126~.
80
W
Nov.,
1912
at the temperature to which i t is subjected. The first fractions from the analyses which contained this substance were slightly yellow and had a peculiar odor, different from the rest of the fractions. A treatment with caustic soda reduced the gravity of these fractions b u t increased the yellow color. It was found, however, t h a t b y the treatment of a turpentine like that shown in Fig. 4 with caustic soda followed by a distillation i t was possible to prepare a refined turpentine which showed no abnormality of the first fraction in color, odor, or gravity. The presence of this substance should not, therefore, introduce any difficulty in the refining process. Another test for the presence of decomposition products was made on several of the samples produced a t different pressures by treating the oil with concen-, trated hydrochloric acid; a red color produced in this
70
Bo
I
1 Icl
10
30
20
IO
155
so
ms
SOLMC PMNTS
FIG.J.-Boiling
I
170
175
1 CORRECTED]
mo
86
87
.ss
.ss .m
.si
.ge
.I .SA
SPECIFK CRAVlTl 41 6 ' C .
point and specific gravity curves for oil from Run 21.
I n order t o determine the possibility of the transformation of pinene into dipentene under the condition of steam distillation the sawdust from Run 29 was air-dried and moistened with 1 1 7 5 cc. of gum turpentine (an amount equal to the total volatile oil originally present) and distilled a t atmospheric pressure. The oil, on analysis, showed no indications of dipentene. The experiment was repeated, making the distillation at j o pounds pressure, but with the same result. These results preclude the possibility of formation of dipentene from pinene under the conditions of steam distillation pressures below 50 pounds and indicate very strongly that dipentene occurs as such in lightwood. Light Oils.-Pigs. 3 and 4 also illustrate another effect of pressure on the composition of the crude turpentines. I n these analyses, as in many others, the crude turpentines produced at pressures as high a s 70 pounds show a considerably higher value for the specific gravity of the first fraction than do the turpentines produced a t lower pressures. This indicates t h a t some substance with low boiling point and high gravity (above 0.870 a t 15' C.) is produced a t the higher pressures ; this substance might come from the incipient decomposition of some portion of the resin
BOlLRC WlNTS
FIG 4.-Boiling
(CORRECTED)
SPECIFIC CRAVlTI AT R'C
point and specific gravity curves for Run 23.
oil from
way is supposed t o indicate the presence of rosin oil. There was only a very slight coloration of the oils produced a t atmospheric pressure, but this coloration increased with the pressure, becoming very marked in the oils produced at j o pounds and 7 0 pounds pressure. FRACTIONATION OF THE OIL DURING DISTILLATION.
Some very interesting conclusions regarding the details of the manner in which the volatile oil leaves the wood can be obtained by comparing the values of the specific gravity of various portions of the distillate. As was previously stated, the specific gravity of the oil was determined from each liter of distillate, or from as many liters as were necessary t o furnish the
T H E JOURA‘AL OF 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 .
Nov., 1912
amount of oil required for a determination. Figs. 5 , 6 and 7 show these values of the specific gravity obtained in Runs 16, 2 0 and 7 , respectively, plotted against the percentages of the total oil obtained. Fig. j shows the changes in the specific gravity of
.E6
.E7
FIG. S.-Curve
.BE .89 .SO 31 SPECIFIC GRAVITY AT IS’ C.
.92
.93
797
the interruption of the distillation followed by a further distillation under the same pressure produced a small further yield of oil with a lower gravity, and on increasing the steam pressure, still more oil was obtained with a still lower gravity. This indicates that both the interruption of the distillation and the increase in steam pressure brought more oil into contact with the steam and t h a t this oil contained some of the low-gravity turpentine material. A very different behavior is shown in Fig. 6 , which represents the distillation of chips I” X I / ~ ” X I/2” a t a pressure of 7 0 pounds. I n distillation under these conditions there was much less tendency for the oil t o be separated as it is distilled, the gravity of the very first fraction being higher than t h a t of pure tur-
.94
showing fractionation of oil during Run 16.
the oil obtained during the distillation of a charge of sawdust, first a t atmospheric pressure and then at 40 pounds pressure. The first portions of the oil were nearly pure turpentine, but after about 44 per cent. had been distilled the gravity increased rapidly, indicating the presence of pine oil in increasing quantities; when the distillation was from about 67 per cent. t o 83 per cent. completed the oil was nearly pure pine oil. The part of the curve up t o 83 per cent. resembles very closely the distillation curve which could be obtained from the distillation with steam of a crude turpentine; that is, the presence of the wood seems t o have no effect on the manner in which the volatile oils are distilled. A difference is seen, however, in t h a t portion of the curve beyond 83 per cent. ; after practically all the oil possible had been removed by a continuous distillation a t atmospheric pressure,
SPECIFIC GRAVITY AT 15’ C FIG. &-Curve
showing fractionation of oil during Run 20.
pentine and the gravity of the later fractions never reaching t h a t of pure oil; that is, the turpentine and pine oil distilled together throughout the run. This indicates that new supplies of volatile oils were brought into contact with the steam more or less continuously throughout the distillation, since otherwise the turpentine would have distilled first and the last fractions would have been nearly pure pine oil. A still more striking picture of the variation in the
7 98
T H E J O U R N A L OF I N D U S T R I A L AiVD E-\-GISEERISG
gravity of the distillate due to changes in the conditions of distillation is shown in Fig. 7 ; this represents the gravities of different parts of the oil obtained during the distillation of chips I” X I/~’’ X I/~’’ a t atmospheric, a t 20 pounds, and then a t 40 pounds pressure. During the first of the run a t atmospheric pressure the gravity gradually increased, but never quite reached t h a t of pure pine oil. By a continuous distillation a t atmospheric pressure only about 50.5
CHEJIISTRY.
Nov., 1912
Here again the effect of interrupting the distillation and of increasing the pressure are very plainly shown, viz.,increased yield of oil with gravity lower than the last fraction obtained before the conditions were changed. This effect of the increased pressure in increasing the yields is due then t o bringing more steam and oil into contact with each other than is possible a t lower pressures. This could result either from a penetration of the steam further into the wood a t the higher pressures or from a flow of resin toward the surface of the wood due to the decreased viscosity a t the higher temperatures. It seems probable that both these have some influence, but the effect of the latter is quite certain, since it was noticeable t h a t in the distillations made a t high pressures a considerable amount of rosin would collect in the bottom of the retort or the outside of many of the chips would be coated with thin layers of rosin. The effect due t o the interruption of the distillation and continuing it again under the same conditions cannot be explained so readily, but it is probably due t o a slow flow of rosin toward the surface or t o the diffusion of the volatile oils in the rosin from the interior of the chip to the rosin a t the surface from which the oil has been removed. APPLICATION O F RESULTS.
The foregoing discussions have considered the effects of the different variables ( I ) size of chip, (2) pressure of steam, (3) speed of distillation, and (4) end point a t which distillation is stopped, on ( a ) the yield of total oil, ( b ) the composition of the oil, and (c) the amount of steam required rto remove the oil. It can be seen that there should be a certain combination of values for these variables which would give a most economical method of operation for a steam distillation plant; but there are other factors which must be taken into consideration in determining the proper combination of values. For instance, the best size of the chip will not be determined entirely by the effect of size on yield and efficiency, but also b y the relative costs of preparing different-sized chips and the use to which the chips are t o be put after steaming; the best pressure of steam will not be determined entirely by the effect of pressure on yield and efficiency, FIG.7 . - C u r v e showing fractionation of oil during Run 7 . but also by the relative costs of high and low pressure per cent. of the total oil could be removed, but on in- steam and of apparatus designed for use with differterrupting the distillation for about fourteen hours ent pressures; the best speed for the distillation will and continuing again a t atmospheric pressure, 7 per not depend entirely upon the effect of speed on the cent. more oil was obtained, the gravity of the fir& yield of products and on the amount of steam repart of this 7 per cent. being much lower and of the quired, but also upon the cost of steam and the overlast part only slightly lower than t h a t of the last frac- head charges; the best end point a t which t o stop tion of the continuous run. On increasing the pres- the distillation will not depend entirely upon the sure t o 2 0 pounds, about 16.8 per cent. more oil was effect of end point on yield and efficiency, but also obtained, the gravity suddenly dropping and then upon the cost of the raw material, the value of the gradually rising during the distillation of this 16.8 products, etc. per cent. On increasing the pressure t o 40 pounds Sufficient experimental data have been given, howand distilling continuously, a further yield of 19.5 ever, so that a knowledge of the various cost factors per cent. was obtained, the gravity of this 19.5 per mentioned above (which would naturally vary a great cent. dropping suddenly a t first and then gradually deal in different plants) wouldmake it possible to decide rising. A similar additional yield was obtained b y readily on the most economical methods for operating. another interruption of the distillation, about 6 per FOREST PRODUCTS LABORATORS, cent. more oil being obtained. MADISOX,\VXSCONSIN.
