CHEM. L. B.
|-
|-
|-
|-
|-
|-
.
|-
L
|-
|-
|-
.
.
|-
|-
|-
|-
|-
|-
|-
|-
|-
|-
|-
|-
|-
|-
|-
|-
|-
|-
|-
|-
|-
|-
|-
|-
.
|-
|-
E
|-
|-
|-
|-
.
|-
|-
|-
|-
|-
|×
January, 1916.
Vol. 8, No. 1, page 5.
Polymerization of Chinese
Wood Oil
By Carl Louis Schumann

- Chemical Library
V. P.
(2 Sº O
• ‘S Sa 2
POLYMERIZATION OF CHINESE
WOOD OIL
BY
CARL LOUIS SCHUMANN
A dissertation submitted in partial fulfilment of the
requirements for the degree of Doctor of Philosophy in
the University of Michigan.
1916
green in color. The fruit, which ripens in September
or October, changes from an apple-like green to a
dull brown color. It contains from three to five
light brown seeds, which are enclosed in a fibrous
flesh. The seeds consist of about 49 per cent shells
and 51 per cent meat. The kernels contain from 5o
to 53 per cent oil, of which 4o to 42 per cent can be
extracted by pressing. This oil is of an exceedingly
poisonous character.
The cold drawn oil obtained by pressing is pale
yellow in color, and is termed white tung oil; it has a
peculiar and characteristic nutty odor and unpleasant
taste. The oil obtained by hot pressing is dark brown
in color and is termed black tung oil. It has a stronger
smell than the light pressed oil, which precludes its
use for many purposes, and it is therefore largely
consumed in China. -
Due to the lack of efficient means of transporta-
tion throughout China the individual grower of the
tung oil tree presses the oil from the seeds, and uses
the presscake as a fertilizer or for making lampblack.
This presscake, on account of its poisonous qualities,
cannot be used as a fodder. He then carries the oil
to market in bamboo baskets lined with oiled paper.
From these local markets the excess oil gradually
finds its way to the ports of Hankow and Wuchow,
from whence the chief shipments are made to other
countries.
When we take into consideration the crude methods
employed in pressing out the oil and the opportunities
offered for adulteration, as well as possible variations
in analytical procedure, it is not to be wondered that
there has been a lack of concordant results obtained
upon the several analytical constants of commercial
samples of chinawood oil. . -
Recent investigators working upon authentic sam-
ples of this oil have been able to check much closer
in the determination of these constants. This would
indicate that adulteration may have been a decided
factor in the variation first observed. Wilson" be-
lieves that soya bean and sesame oils are chiefly
used as adulterants. g -
In 1900, 21,960 tons of the white drawn oil were
exported from Hankow, and 1627 tons from Wu-
chow; in 191o, 5o,338 tons were exported from Han-
kow and 3,465 tons from Wuchow. A few years ago
1 Loc. cit.
(2) •
comparatively little Chinese wood oil was imported
into the United States, the bulk of the oil going to
Europe. Recently, however, the consumption in
the United States has increased enormously. This
decided increase has induced the United States De-
partment of Agriculture to investigate the possi-.
bilities of starting a tung oil industry in the United
States.
In 1905 a quantity of Chinese wood oil nuts was
imported from Hankow by the United States De-
partment of Agriculture. They were planted in Cali-
fornia and a year later were sent out as one-year-
old trees. The trees proved easy to cultivate, doing
especially well in the states of Georgia, Louisiana,
South Carolina, Alabama, Florida and Mississippi.
The first fruits were collected five years later in such
quantities as to demonstrate that the tung oil industry
is feasible in the Southern states. The trees require
little attention and but little labor is involved in gath-
ering and husking the fruits. Although the freight
rate from the Southern states to New York is some-
what higher than the rate by water from Hankow
to New York the cheap lands and the use of team labor
here may offset this handicap, and that of the differ-
ence in price of labor.”
We can gather some idea as to the valuable quali-
ties of Chinese wood oil from the variety of uses to
which the Chinese have put the oil. The Occidental
nations are just beginning to realize the possibilities
of this wonderful natural varnish. In China it serves
primarily as a natural varnish for their boats, houses,
furniture, and as a waterproofing material for their
paper umbrellas. It is likewise used as an illuminant,
as a medicine, in the manufacture of silk, and as a
building material when mixed with sand, clay and
lime. In the United States and the European coun-
tries its main use is in the manufacture of paint driers
and rosin-chinawood oil varnishes. A large num-
ber of patents have been taken out in Germany and
the United States for the utilization of this oil in the
manufacture of linoleum, rubber substitutes and
waterproof paper, but the extent to which it is being
used cannot be stated because of the reticence of manu-
facturers to expose the details of their business. At
present its chief importance is as a varnish material, for
it is a recognized fact that with rosin it makes a varnish
superior to any linseed-rosin varnish on the market.
1 Fairchild, Oil, Paint and Drug Reporter, Nov. 18, 1912, p. 36
(3)
.* * *
. ... • * > * * * *
The varnish manufacturer, upon the introduction
of Chinese wood oil as a varnish material, soon observed
peculiar variations in the products he obtained upon
heat-treating tung oil, especially marked with different
oils. This lack of uniformity was thought to be due
, to excessive adulteration, and many attempts have been
made to find a standard method of analysis for this oil.
In spite of the number of investigators interested
in finding a suitable method for testing the purity of
the oil little actual progress had been made up to the
last year or so. The efforts of the “Oil Committee”
of the American Society for Testing Materialsº have
been instrumental in standardizing the methods of
analysis to a point where they are much more reliable
in the hands of different operators.
The constants for the oil were found by the
Committee to agree remarkably closely for authentic
samples of tung oil, which had not been the case with
the oils previously examined. •
In testing out in this laboratory the several pro-
posed methods, interesting results were obtained,
particularly when working with the “iodine-jelly”
test. The samples sent out by the Committee were
tested by this method and it was found that both
the pure oil and the adulterated samples jellied in
less time than we had been led to expect from work
previously" reported from other laboratories. This
acceleration of the jellying was probably due to dif-
ferences in choroform used, as it was found that cer-
tain substances act very decidedly in catalyzing
the reaction. For instance, by the addition of a very
small quantity of alcohol the jelly formation is retarded
several hours, while a slightly larger amount of alcohol
will entirely prevent it. Following the regular pro-
cedure it was possible to retard the jelly formation
to the time specified only by diluting the iodine solu-
tion to considerably below the strength specified for
the method. The above phenomenon is thus analo-
gous to the jellying of heat-treated oils and is subject
to the same retardation and prevention as is exhibited
by them. w -
- CONSTITUTION OF CHINESE . WOOD OIL .
Chinese wood oil is composed mainly of the glyc-
erides of oleic and elaeomargaric acids.” Lewko-
witsch prefers the use of the name “elaeomargaric”
* Reprint of Report of Committee D-1, American Society for Testing
Materials (1915). . . . .
* Lewkowitsch, “Chemical Technology of Oils, Fats and Waxes,” 1909.
(4)
acid to that of “elaeostearic” acid used by many in-
vestigators, and states that it is an isomer of linolic
acid. -
In this paper the term elaeomargaric acid will.be.
used for the acid of the fresh glyceride, and the term
8-elaeostearic acid for the acid of the stereoisomeride
of elaeomargaric acid, formed under the influence
of light rays.
Cloez' reported 24 per cent oleic acid and 72 per
cent elaeomargaric acid in Chinese wood oil. The
elaeomargaric acid he found to have a melting point
of 48°C., it being a crystalline solid. When this acid
was exposed to the light or to the action of carbon
bisulfide it was transformed into what he called a
“polymer,” B-elaeostearic, melting at 72° C. He
also mentioned that the same transformation occurred
with the glyceride of elaeomargaric acid into 3-elaeo-
stearic acid glyceride. He assigned the formula
C17H2002 to elaeomargaric acid.
Later Maquenne” repeated the experiments of Cloez:
he declared the formula to be C18H3002 instead of
C17H3002, as given by Cloez, and named it elaeostearic
acid. Upon oxidation with potassium permanganate
he obtained n-valeric and azelaic acids; the question
of the number and position of the double bonds he
left open. He disagreed with Cloez that 6-elaeo-
stearic acid is a polymer of elaeomargaric acid as he
believed it to be an isomer instead. -
Moritz Kitt” subjected the acid to oxidation with
alkaline potassium permanganate, and observed that
a part remained unchanged and also that small amounts
of dioxystearic acid, a water-soluble acid, some lower
fatty acids, and carbonic acid were formed. Kitt
likewise mentions the formation of the isomer of elaeo-
margaric acid through the action of light rays, and
ascertained that this isomer could be distilled under
a vacuum without decomposition.
Norman" and Morrel" both mention the fact that
the rate of change of elaeomargaric acid glyceride
into its stereoisomeride, under the influence of light,
is very slow. Morrel noted that this change was
accelerated by the removal of the quantity first formed
and by treatment with acetone. Meister believes
that the quick formation of a film, when Chinese
1 Compt. rend., 81 (1875), 469; 82 (1876), 501; 88 (1876), 943.
* Ibid., 185 (1903), 686; Jour. Chem. Soc., 88 (1903), 1042.
* Chem. Ztg., 28 (1899), 23 and 38.
* Chem. Zentr.-Blatt, 1 (1907), 1207. -
* Transactions Chem. Soc., 101 (1912), 2082.
(5)
wood oil dries, is due to this transformation of elaeo-
margarine into elaeostearine. - *
Kametaka" repeated the oxidation experiments of
Kitt with alkaline potassium permanganate and ob-
tained sativic and dioxystearic acids. On bromina-
tion he obtained the tetrabromide and not the hexa-
bromide, as might have been expected from the work
previously done on this acid. This indicated the pres-
ence of only two double bonds, a conclusion likewise
drawn by Walker and Warburton.”
Majima,” using the ozone method, worked out by
Harries," identified the cleavage products, azelaic
and n-valeric acids. He was not able, however, to
prove the presence of succinic acid, which should be
also formed if the formula he gave for elaeomargaric
acid was correct. This acid was later found by Schap-
ringer," using the same ozone method of Harries,
thus establishing with certainty the following formula
suggested by Majima: -
CHs.(CH2)s.C.H.C.H.(CH2)..CH.C.H.(CH2)7.COOH
Schapringer stated that this acid is a stereoisomeride
of linolic acid. Fahrion," however, pointed out that
the formula worked out by Goldsobel" is distinguished
from the elaeomargaric acid formula in the position
of the double bonds and is thus not a stereoisomeride
but an isomer of linolic acid. -
Fahrion* concluded from his work upon tung oil
that the main constituent of the fatty acids of wood
oil is elaeomargaric acid, besides about Io per cent of
oleic acid and from 2 to 3 per cent of Saturated
fatty acids. This percentage of elaeomargaric acid
is just a little higher than that calculated by Kame-
taka,” who gave 86 per cent of this glyceride and 14
per cent olein.
From the results obtained in this laboratory by crys-
tallizing out the elaeomargaric acid according to the
procedure as outlined in the experimental part, as
well as from the potassium soap and light break meth-
ods,” the writer should judge that the percentage of
elaeomargaric acid tri-glyceride present in raw Chinese
1 Jour. Coll. Sci. Tokio, 25 (1908), Art. 3, p. 1-8.
* Analyst, 27 (1902), 238.
• Ber., 34 (1909), 674. - r
4 Ber., 89 (1906), 2844, 3728; 40 (1907), 4154, 4905; 41 (1908), 1227.
* Dissertation, Karlsruhe, 1912. - -
• Farben-Zig, 18 (1913), 2418–2420.
7 Chem. Zig., 80 (1906), 825; Chem. Zentr.-Blatt, 1 (1910), 1231.
• Loc. cit.
• Jour. Coll. Sci. Tokio, 25 (1908), Art. 4, 1–4.
in Tars']ournal, 6 (1914), 806. 6 -
. - (6) .
wood oil is still higher than reported by Fahrion,
varying from 9o per cent for the darker to 94 per cent
for the light colored cold pressed oil.
GELATINATION OF CHINESE WOOD OIL
The most striking characteristic of Chinese wood
oil is the property of being transformed by heat into
a transparent, solid mass possessing considerable
elasticity in the softer condition, and then becoming
more brittle as further heat is applied. Throughout
the literature upon tung oil we find a large number of
statements regarding this ability to polymerize to
a solid product when heated to certain temperatures
for definite periods of time, yet a satisfactory explana-
tion of the causes for this transformation has not been
given, although several investigators have advanced
theories to account for the phenomenon.
Cloez, Zucker” and Norman” observed that there
is a decrease in the iodine number from 163, for the
original oil, to an iodine number of Io.7 of the polym-
erized Chinese wood oil, and that the saponifica-
tion value is subject to a slight increase.
Jean' assumed that the gelatination is due to a
sudden, unusual oxygen absorption. Moritz Kitt,"
however, opposed the views held by Jean, maintaining
that the oxygen could hardly penetrate to the lower
part of the oil, as would have to be the case in order
to explain uniform gelatination. This theory of Jean
was further disproved by Jenkins,” who found that
tung oil polymerizes to a clear, elastic solid out of air
when heated to 180° C. and finally to 25o°C
This gelatination Kitt” believed to be due to polym-
erization through the formation of intramolecular
anhydrides of the nature of lactones, these being sub-
sequently formed from the fatty acids freed by a
partial cleavage of the tung oil with liberation of
glycerine. - -
Schapringer* assumed that gelatination of Chinese
wood oil is somewhat different from the bodying of
linseed oil. From his experiments he concluded that
the gelatination proceeded in two distinct stages,
the former as a progressive change and the latter as an
1 Loc. cit.
* Pharm. Ztg., 48 (1898), 628.
* Chem. Ztg., 81 (1907), 188.
* Ibid., 21 (1898), 183.
* Ibid., 28 (1899), 38. * e
* J. Soc. Chem. Ind., 16 (1897), 194.
7 Chem. Zentr.-Blatt, 2 (1905), 1469.
* Loc. cit. .
(7)
instantaneous one. He declared it to be a “meso-
morphic polymerization,” since there is formed
during the first stage an intermediate product, soluble
in benzol, while in the second stage an insoluble final
product is formed with an evolution of heat. He did
not find the decrease in glycerine content observed by
Moritz Kitt, nor that decomposition products ap-
peared during polymerization. -
Since the saponification value remains practically
constant and the small acid number decreased to a
fraction of one per cent, Schapringer thinks that these
fatty acids alone form anhydrides, and that the union
takes place between the double bonds, as stated by
Fahrion.”
Schapringer based his conclusion that mesomor-
phic polymerization takes place, mainly upon the
comparative solubilities of the intermediate and final
product. He was confronted with the fact that the
changes of the Constants of wood oil are different on
heating to different temperatures, and to explain
this he assumed that at various temperatures the re-
action proceeds differently, and consequently at each
of these temperatures, different intermediate products
are formed.
Fahrion, in this earlier paper, stated that the polym-
erization of Chinese wood oil is analogous to the
thickening of linseed oil,” taking place through dis-
solution of the double bonds. This agrees with the
molecular weight determinations of Norman,” who
found a molecular weight of about 440 for the fatty
acids from the raw oil and around 8oo for the fatty
acids from the heated oil. •
The next paper by Fahrion appeared shortly after
the dissertation of Schapringer had been completed.
IIe then goes into greater detail upon the phenomena
of gelatination, and points out that the conclusion of
Schapringer, that “mesomorphic polymerization”
takes place, can hardly be correct. - -
Fahrion gives several reasons why he does not be-
lieve the polymerization of Chinese wood oil to be
analogous to the polymerization of styrolene, cited
as a typical mesomorphic polymerization by Kron-
stein and Seeligman."
1 Kronstein, Ber., 85 (1902), 4150–4153.
* Farben-Ztg., 17 (1912), 2583–2584:
* Z. für angew. Chem., 5 (1892), 174; Leeds, J. Soc. Chem. Ind., 18
(1894), 203. - . .
* Loc. cit. -
* Dissertation, Karlsruhe, 1906.
(8)
One reason he gives is that it is impossible to isolate
the intermediate product as has been done with sty-
rolene. He further states that proportions between
the changes of the constants of Chinese wood oil
are different on heating to different temperatures,
and that to explain this Schapringer must not only
assume the existence of a special intermediate product
for each temperature, but also of several of these
intermediate products when he presumes that at
different temperatures the reaction proceeds differ-
ently. Another reason he gives for not accepting
the views of Schapringer is that on heating polym-
erized styrolene a decomposition takes place with a
reformation of the initial product, while such a refor-
mation is not possible with wood oil. He further
states that the product of the polymerization of lin-
seed oil does not necessarily have to be the same as
that from tung oil, although polymerization in both
cases occurs through dissolution of a part of the double
bonds. Fahrion is inclined to believe that a transfor-
mation product, soluble in the oil up to a certain per
cent, is formed. If this per cent be exceeded, gelati-
nation, or colloidal gel formation takes place on cooling.
He goes on to say that this polymerization product is
insoluble in benzol alone, but dissolves in this medium
up to a certain per cent if an excess of the unchanged
oil be present. He also mentions the extraction ex-
periments of Schapringer, who found 17.o, 13.7 and
7.4 per cent, respectively, of benzol-soluble matter, while
Wolff" found up to 8o per cent ether-soluble matter.
Wolff, however, triturated solidified product with
sand, which Schapringer did not do.
Chinese wood oil has come to be an important
factor in the varnish industry of the world, and it is
essential, in order to maintain a more perfect control
of the operations of cooking it with other varnish
materials, that there shall exist a thorough under-
standing of its properties and of the change it under-
goes in its manipulation by the varnish maker.
In view of the differences in opinion as to the re-
action occurring on the solidification of Chinese wood
oil, as well as the fact that the data presented is not
entirely in harmony with these theories, it was de-
cided to make a study of the polymerization of this
oil, and the part polymerization plays in inducing
solidification, and to determine, if possible, the in-
1 Farben-Zig., 18, 1171 (1913). -
(9)
fluence of other materials in preventing this solidifi-
cation. º
| EXPERIMENTAL
I—PHYSICAL AND CHEMICAL constants For TUNG OLL
In the following experiments Chinese wood oil was
used which gave the following constants:
Acid number. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.7
Saponification value. . . . . . . . . . . . . . . . . . . . . . . 190.9
Iodine number (Wijs). . . . . . . . . . . . . . . . . . . . . . 167.7
Refractive index 20°C. . . . . . . . . . . . . . . . . . . . . 1.5182
Specific gravity 15.5°C. . . . . . . . . . . . . . . . . . . . 0.9402
Fatty acids. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95.1 per cent
Oxyacids. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.7 per cent
In making saponification determinations on the
solidified Chinese wood oil it was observed that the
rate of saponification is slow, requiring several days
for the hard, solid product formed by continuing the
heat treatment for some time after gelatination has
taken place. This slow rate of saponification is prob-
ably due to the insolubility of the colloid rather than
to the difficulty in the saponification of the polymerized
material. The soft gels saponified in a comparatively
short time, which may be due to the fact that the
unchanged oil saponified readily, permitting a larger
surface contact of the saponifying solution with the
polymerized tri-glyceride. In all cases complete
saponification takes place according to the regular
procedure, except that the rate is slower.
Norman" recommends the evaporation of the alco-
holic soap solution to dryness at least once, while
determining the saponification value of tung oil, as it
saponified with difficulty. Meister” followed the
above suggestion of Norman but did not state whether
he found it necessary to do so. This difficulty of ob-
taining complete saponification is not confirmed by
Fahrion,” however, and he does not recommend making
the change in the ordinary procedure as suggested by
Norman. -
In the determination of the iodine numbers of
Chinese wood oil in this laboratory no difficulty was
experienced in obtaining concordant results by the
Wijs method, when following the suggestions of Chap-
man," who likewise prefers the Wijs method to that
of Hübl.
It has been shown by Kreikenbaum" and in the Con-
tracts Laboratory" of the U. S. Bureau of Chemistry
1 Loc. cit. - o e
2 Chem. Rev. Fett, u. Hars-Ind., 1912, p. 159.
* Chemisch-Technisches Repertorium, 86 (1912), 672.
* Analyst, 87 (1912), 543.
S THIs Journal, 2 (1910), 205.
• Proceedings 7th Intern. Congr. Appl. Chem., 1909, Sect. 1, p. 89.
(Io)
that in the determination of the iodine number the
Hanus solution gives abnormally high results on tung
oil, and that the Wijs solution gives concordant re-
sults only when all the conditions of time, tempera-
ture and proportionate amounts of the weight of tung
oil to the quantity of solution are identical.
The determination of oxyacids was made according
to the method used by Fahrion," which is based on
the insolubility of the oxyacids in very light petroleum
ether, while the unoxidized acids and the polymerized
acid are quite soluble in that solvent. The writer
likewise observed that the oxyacids determined in
this manner gave a small iodine number, as earlier
reported by Fahrion.
II-METHOD TO DETERMINE PERCENTAGE of ELAEo
MARGARIC ACID .
The method used in determining the percentage of
elaeomargaric acid present in the liberated fatty acids
of Chinese wood oil is based upon the property of
elaeomargaric acid to crystallize from a dilute alcohol
solution at o° C., while the polymerized, i. e., the di-
molecular acid formed by the union of two molecules
of elaeomargaric acid through the loss of a double
bond in each chain, remains in solution under these
conditions. At o? C. the elaeomargaric acid is com-
paratively insoluble while the polymerized and oleic
acids are quite soluble in the above solvent and do
not crystallize from any organic solvent.
ProceDURE—Five grams of chinawood oil, raw or heat-treated,
are saponified and the fatty acids liberated. These mixed fatty
acids, amounting to approximately 4.760 g., are then dissolved
in 40 cc. absolute alcohol and the solution is cooled to o' C.
From 14 to 14.5 cc. water are added, drop by drop, shaking con-
tinuously. The amount of water used is dependent upon the
relative concentrations of the polymerized and unpolymerized
fatty acids present. This solution is then warmed to about
5°C., and allowed to cool slowly and to stand over night at
o°C. The elaeomargaric acid crystallizes out in rhombic .
flakes, which may be filtered out upon paper held in a specially
constructed, ice-cooled copper Buchner funnel. The alcohol
is then distilled from the filtrate under reduced pressure at the
temperature of a steam bath. The remaining fatty acids may
be dried by the addition of absolute alcohol and redistilling the
alcohol, after which they may be weighed, and the elaeomargaric
acid computed by difference. The amount of elaeomargaric
acid remaining soluble in the filtrate under these conditions is
about o. 132 g., a correction for which is applied to the weight
of elaeomargaric acid as computed. When the percentage of
* Forben-zig. 17 (1912), 2635, 2636. -
(11)
elaeomargaric acid present in the sample under analysis is less
than o. 132 g., O.5 g. of raw tung oil, which is known to be pure,
may be added before saponification. This procedure permits
the estimation of smaller amounts of elaeomargaric acid in a
mixture.
This method was also found to give excellent results
in determining the purity of samples of chinawood
oil. It has a decided advantage over the methods
not founded upon a distinct property of elaeomargaric
acid. It further allows examination of the fatty acids
from the adulterated oil, contaminated by only a small
percentage of oleic and elaeomargaric acids, and is
not subject to the extreme precautions necessary
in working with the insoluble potassium soap method.”
III–IEFFECT OF OXIDATION AND CATALYZERS
The avidity with which tung oil takes up oxygen
from the air is striking, our finding being in contrast
to the results of Schapringer," who reports that the
oxygen merely induces polymerization and does not
produce oxidation products during the gelatination
experiments. In all cases studied in this laboratory,
as has been reported,” there have been formed pro-
gressively increasing amounts of oxidized products,
whenever Chinese wood oil or wood oil mixtures have
been heated in air (see Fig. II).
oxIDATION.—It has been further observed that the
oxygen does not act directly as a catalyzer on the rate
of polymerization, as stated by Schapringer, but in-
directly in the formation of oxidation products which
then act as positive catalyzers. Blowing air through
heated wood oil had a decided effect upon the time
required for gelatination to take place. Without a
doubt the products formed by autoxidation had some
bodying effect upon the wood oil, but not in proportion
to the actual quantity present, as indicated by the con-
sistency of the mixture in the second of the following
experiments. It was evident that there was a decided
catalyzing influence, however, on the rate of polym-
erization. -
Chinawood oil solidified when blown at 1.5o" C. in
2 */2 hours, and at 200° C. in less than 1 hour. Heated
in an open dish at 150° C., but not blown, it did not
gelatinize before the end of 30 hours. The following
experiments were carried out to determine whether
the oxygen or oxidation products were responsible for
the catalyzing influence.
1 Loc cit.
* This Journal, 7 (1915), 572.
(12)
1—200 grams of Chinese wood oil were heated at 235° C. un-
der an atmosphere of nitrogen, in a glass container, connected
to a mercury seal to permit volatile products to escape. A trans-
parent, solid gel was formed at the end of 51/2 hours; no gases
were given off except at the moment of gel formation.
2–200 grams of a mixture of chinawood oil, containing 15
per cent blown wood oil, were heated under the same conditions
as in the above experiment. Identical results took place at the
end of 1°/2 hours. -
3–200 grams of Chinese wood oil were heated at the above
temperature, open to the air. The same sesult took place at the
end of 4!/2 hours. . º
It is apparent from these experiments that the oxi-
dation products, and not the oxygen itself, cause the
increased rate of polymerization.
REDUCING METALs—Several experiments were car-
ried out in this laboratory to observe the influence
of the presence of a quantity of strongly reducing
powdered metal in Chinese wood oil upon heat treatment.
Five 200 gram masses of tung oil were mixed with 1 per cent
finely powdered zinc, 1 per cent magnesium, I per cent lead,
and 5 per cent zinc, respectively, and heated under an atmos-
phere of nitrogen at 235° C. Gelatination took place in every
case within 2 hours compared to 5"/2 hours for the wood oil
heated alone. Two like quantities of wood oil were then taken;
to the first 2 per cent sodium amalgam was added, and to the
second 2 per cent litharge, and heated under the same conditions
given above. Again gelatination took place at the end of 2
hours.
The results in the first series of experiments with
the powdered metal are hardly what we should expect
from the process for gel preventation, as given by
Weinschenk." - -
IV—CONSTITUTION OF POLYMER
The explanation for the gelatination of chinawood
oil is that there are formed molecular complexes of the
dimolecular tri-glyceride, through the dissolution
of one-half the unsaturated linkages in the tri-glyceride
of elaeomargaric acid, while the insolubility of this
jellied product is largely due to its colloidal condition.
Several patents have been taken out for processes
to prevent the gelatination of wood oil, incorrectly
called polymerization preventation. The German,
Patent No. 211,405 is of particular interest from a
theoretical standpoint. The claim of this patent is
the formation of hard-drying product obtained by
heating the raw chinawood oil alone, for a short time,
s * German Patent No. 219,715.
(13)
above the “polymerizing” temperature, while keeping
the oil in constant motion. &
The inventor of this process recognizes that the gel
formation is also dependent upon the inertness of the
separate mass particles, and that with the help of purely
physical means gel formation may be prevented with-
out the addition of a resin. He does not, however,
distinguish between polymerization and gelatination,
and makes an error when he states that the Chinese
wood oil can be heated above the polymerizing tem-
perature, and that since a certain length of time is
required for the polymerizing reaction to go to com-
pletion by rapid heating, this condition can be elimi-
nated. Had he determined the iodine number of
the superheated product, he would have noticed his
error. w
The constants upon several superheated Chinese
wood oils are as follows:
Acid number. . . . . . . . . . . . . . . . . . . . . . . . . . 12.5 to 16. I
Iodine number. . . . . . . . . . . . . . . . . . . . . . . . 80.0 to 84.6
Saponification value. . . . . . . . . . . . . . . . . . . 187 to 194
Unsaponifiable. . . . . . . . . . . . . . . . . . . . . . . . 8.2 to 11.9 per cent
With the exception of the saponification value a pro-
nounced change has taken place in the constants.
The glycerides were separated from the superheated oil by
repeated extraction with hot absolute alcohol and finally with
extra light petroleum ether. A glyceride to the extent of about
75 per cent of the original oil was found to be insoluble in the
latter solvent under these conditions, while the ordinary elaeo-
margaric acid tri-glyceride is quite soluble in that solvent.
The constants and molecular weight for this separated tri-
glyceride were as follows: - .
- Molecular
WEIGHT
§§ § 1692 and 1730
Iodine number. . . . . . . . . . . . . . . . . . . . . . iš: b) 1752 and 1814
§ #: point method using benzol as the solvent.
reezing-point-method using naphthalene as the solvent.
These results show the tri-glyceride has double the
molecular weight of the original elaeomargaric acid
tri-glyceride and that one-half of the unsaturated
linkages have disappeared. - -
The fatty acids were then liberated and their molecular weight
and iodine number determined. The freezing- and boiling-point
methods were used in the determination of the former, and were
found to be 57o and 558, respectively, using benzol as the solvent.
The iodine number was found to be 89.8. * -
Upon heating a portion of the di-molecular tri-glyceride
separated from the superheated oil by the above process
gelatination resulted almost immediately. Further portions
were then dissolved in chloroform and benzol and allowed to
(14)
stand for 24 hours out of contact with air. A gel formed more
readily with the former solvent: this gel did not redissolve upon
the addition of further solvent, nor upon heating the solution.
These gels were then saponified, the acids liberated, and the
molecular weight as well as iodine number determined. The
molecular weight by the freezing-point method with benzol as .
the solvent was found to be 55o and 570, and the iodine number
90. I. - -
A residue benzol solution of the di-molecular tri-glyceride
remaining from a molecular weight determination was allowed
to stand over night, stoppered tightly. The molecular lowering
was then again determined and it was found to be but a fraction
of the original lowering. - - -
It is evident that the above gel product is not the
same as the intermediate product, but must consist
of complexes formed from the intermediate product.
Since the iodine number and molecular weight of the
fatty acids from this gel remain the same, and the re-
formation of the dimolecular tri-glyceride, i. e., the
intermediate product, is possible by heating with rosin
or oleic acid, the writer is inclined to believe that this
formation of complexes is not accompanied by a de-
cided change in the molecule.
The absence of gelatination in the superheated
product upon heating at polymerizing temperatures
must be due to the presence of decomposition products
formed during this superheating process, for the
polymerization to the intermediate product is complete.
Upon the addition of 25 per cent raw tung oil to a
superheated oil, the mass will readily gelatinize at
polymerizing temperatures. It is thus apparent that
there is a minimum amount of decomposition products
that must be present to prevent gel formation, i.e.,
to prevent the formation of complexes of the di-
molecular tri-glyceride. -
An experiment was carried out to see if the polymer-
ized tri-glyceride of elaeomargaric acid formed as
readily when the raw Chinese wood oil was dissolved
in a solvent such as naphthalene.
150 grams of Chinese wood oil were heated with so grams of
naphthalene at 200° C., out of air for 24 hours under a reflux
condenser. Gelatination took place, forming an insoluble
colloidal gel, similar to that formed when the oil is heated alone.
Equal quantities of Chinese wood oil and naphthalene were
then heated under the conditions as outlined above; gelatina-
tion did not take place and the mass remained soluble in the º
ordinary organic solvents. From the latter mixture Ioo grams
were taken and an attempt made to distil off the naphthalene.
Upon the removal of about 15 per cent of the naphthalene
(15)
g gelatination took place. This gel product was apparently as
difficultly soluble in the ordinary organic solvents as that un-
influenced by a solvent. e
From the above experiments it is shown that the
action of the naphthalene must be that of a diluent,
and that the minimum percentage which must be
present to prevent gel formation is exceedingly large.
A further quantity of this heat-treated mixture of 50 per cent
tung oil and 50 per cent naphthalene was taken and the polym-
erized tri-glyceride separated from the mass by extraction
with hot absolute alcohol and finally with petroleum ether,
boiling point of 55° C. The molecular weight of this insoluble
glyceride was found to be 1711 and 1755, by the freezing-point
method with benzol as the solvent. The iodine number was
found to be 86.o. *
The iodine number of the fattº
*acids liberated from this
glyceride was found to be 90.1 and the molecular weight 544 and
560, determined by the boiling-point method with acetone as the
solvent.
Two samples of Chinese wood oil were then heated out of air
at 150° C. and 225°C., respectively, the heating being dis-
continued immediately upon solidification. The gelatinized
products were then saponified, the fatty acids liberated and their
molecular weights determined by the freezing-point method,
using benzol as the solvent, and were found to be 483 and 495
for the former and 452 and 460 for the latter fatty acids. The
iodine numbers were found to be 114 and 128. A superheated
tung oil that had gelatinized, due to an insufficient amount of
decomposition products being formed, was then saponified and
the fatty acids liberated. The molecular weight upon two
samples by the above method was found to be 533 and 553,
and the iodine number 87.3.
The percentage of elaeomargaric acid glyceride remaining
unpolymerized was then determined in the above three samples
of heat-treated, gelatinized wood oils, and assuming IO per cent
olein present in the oil used the amount was calculated to be
20 per cent, 28 per cent, and 4 per cent, respectively, of the
total amount present in the original oil.
Complete polymerization to the intermediate product takes
place in 15 minutes at 350° C., while at 150° C. it requires at
least 30 hours to solidify when heated out of air.
ACTION OF HEAT ON CHINESE WOOD OIL
What actually occurs on the transformation of
Chinese wood oil by heat from the raw state to the
solid product is as follows: Upon heating at suffi-
ciently high temperatures two elaeomargaric acid tri-
glyceride molecules unite through dissolution of one-
half of their double bonds, as evidenced by their
molecular weights and iodine number. Each elaeo-
margaric acid chain in the tri-glyceride loses one

(16)
unsaturated linkage through the union with the other
elaeomargaric acid chain. This union is shown by
the molecular weight, and iodine number, and by the
entire loss of the elaeomargaric acid. This inter-
mediate product formed by polymerization of the elaeo-
margaric acid tri-glyceride is formed in 'every case.
of heat treatment of Chinese wood oil whether the
wood oil is heated alone or in the presence of a solvent,
and does not vary with the temperature. The amount
formed is dependent upon several factors to be men-
tioned later. This soluble intermediate product has
the power of forming, under favorable conditions,
molecular complexes that constitute the colloidal
gel. The intermediate product and colloidal mass
are not identical, as thought by Fahrion. The col-
loidal state of the latter complexes is apparently the
main cause of the insolubility of this product, as shown
by the extraction experiments. º
The formation of these complexes is largely depen-
dent upon the concentration, and is influenced by the
presence of certain substances acting as accelerating
or retarding catalyzers, the presence of a solvent
acting as a diluent as well as upon the inertness of the
separate mass particles, as pointed out by the German
Patent No. 21 1,405. -
The rate of formation of this intermediate product
is also dependent upon the temperature, and is in-
fluenced by the presence of certain substances acting
as catalyzers, and the freedom of the separate mass
particles. This product is the same, however, in every
Ca,SC.
The solid gel may be dissolved by simply heating
with rosin, oleic acid or other solvents of the same type
for a time at 3oo° C. By this treatment the di-
molecular tri-glyceride is again formed. Since the
initial polymerization occurs through the unsaturated
linkages, the writer sees no reason why Fahrion should
expect a complete depolymerization. e
Fahrion stated that gel formation takes place when
a certain concentration of the polymerized glyceride
in the unchanged oil is reached. In other words,
he assumes that the percentage of polymerized glyc-
eride is the same in every case when gel formation
takes place. He then concludes that since changes
for the constants differ on heating to different tempera-
tures, the polymerization of Chinese wood oil and styro-
lene are not analogous. -
The reasons given by Fºon are untenable and not
(17) -
in keeping with actual facts. Not only is an interme-
diate product actually formed, but it can be isolated.
Further, this intermediate product is soluble in all
the ordinary, -organic solvents with the exception
of extra light petroleum ether, as well as in unchanged
wood oil, rosin, and like solvents. Its solubility in
benzol is not dependent upon the presence of a certain
minimum amount of unchanged wood oil as claimed
by Fahrion. - -
Schapringer, in order to explain this difference
in constants, presumes, as mentioned before, that at
different temperatures different intermediate products
are formed, also assuming, as did Fahrion, that for
gels of the same consistency, the proportion of polym-
erized oil present is the same. He concluded that
the course of polymerization cannot be followed
by the iodine number alone, but requires a correlation
with the refractive index and specific gravity, since
the iodine number varies with the different gels, being
greater with the gels formed at the higher tempera-
tures than those formed at lower temperatures.
These assumptions of Fahrion and Schapringer as
to the lack of variation of the different gels in the
percentage of polymerized glyceride, caused them to
draw wrong conclusions. The writer has given the
various conditions under which gels are subject to
variation, stating that the concentration of the inter-
mediate product alone is not the factor determining
gel formation, as assumed by Fahrion and Schap-
ringer. Wolff' stated that the gel formation was de-
pendent upon the concentration and temperature,
and drew the conclusion that the theory, assuming
the solidification of wood oil to be due to the formation
of a homogeneous polymerization product, as well as
the theory that it is due to a chemical reaction, must
be abandoned.
The course of polymerization in Chinese wood oil,
inasmuch as it takes place only through the union of
double linkages in the elaeomargaric acid tri-glyceride,
may be followed by the iodine number alone, and does
not require correlating with the specific gravity and
refractive index as stated by Schapringer. This
would apply only to mixtures heated out of air, for
upon heating in air, the oxyacids must be determined
and the drop in iodine number allowed for, due to oxi-
dation. - *
We may have complete polymerization of the tri-
* Farben-zus., 18 (1913), 1171 to 1173. -
(18)
glyceride of the elaeomargaric acid to the di-molecular
intermediate product without gel formation taking
place. This is possible when a small quantity of de-
composition products are present, or in the presence
of a diluting solvent. -
v–POLYMERIZATION IN VARNISH MANUFACTURE
ACTION OF Rosſ N–It was originally supposed that
addition of rosin to Chinese wood oil acted to prevent
polymerization in a heat-treated mixture. This was
found not to be the case, the rosin acting rather in
preventing gel formation, while the polymerization
of the elaeomargaric acid tri-glyceride took place
even more rapidly than when raw wood oil was heated
alone. -
Upon using the potassium soap method' for the de-
tection of Chinese wood oil in rosin-chinawood oil
varnish, it was found that only a very small percentage
of the tung oil could be precipitated out. The reason
for this was first thought to be due the solvent action
of the rosin present, preventing quantitative precipi-
tation. This was not the case, however, for, with
mixtures of rosin and wood oil that had not been
treated, the potassium soap of elaeomargaric acid
could be quite completely separated out. The heat
treatment interfered with the separation to an extent
that was dependent upon the temperature and time
of heating. On prolonged heat treatment it was
impossible to separate out any potassium elaeomar-
garate. -
Since the heat treatment gave such varying re-
sults it was thought advisable to make a thorough
study of the rate of polymerization with mixtures
of rosin and Chinese wood oil at the temperatures
15o, 175, 200, 225, 25o and 275° C. -
As stated in the preliminary paper,” the ordinary
procedure for the making of a rosin-chinawood oil
varnish is to heat alone until the oil will gel slightly
on cooling. It is then mixed with rosin and heated at
higher temperatures for a short time.
In order to obtain more uniform and comparable
results it was thought advisable to permit the rosin
to exert its influence during the whole heat treatment,
even at the expense of color, which was considered
an unimportant factor. -
The rosin used throughout the following experi-
ments graded water-white and gave the following
Constants: -
* Loc. cit.
(19)
Saponification value Acid number Unsaponifiable
- 180.8 156. 1 5.9 per cent
It was found that 25 per cent rosin was not sufficient to pre-
vent gel formation, and in Series A mixtures of 5o per cent
Chinese wood oil and 50 per cent rosin were heated together.
In Series B, 5o per cent limed rosin (limed with 4 per cent
hydrated lime) was heated with 50 per cent Chinese wood oil.
In Series C, 49 per cent rosin and 50 per cent Chinese wood
oil, to which 2 per cent litharge had been added, were heated at
only 200° C., a temperature found to show a typical rate of
reaction for the various mixtures.
Heat treatments were carried out in air and with exclusion
of air. The container used for heat treatment out of air con-
sisted of a 250 cc. distilling flask sealed at the mouth with the
side arm bent downward at an angle of 90° dipping into mer-
cury, acting as a trap to allow volatile products to escape while
preventing the entrance of air. The mixtures to be heated in
the air were treated in uncovered porcelain beakers of 400 cc.
capacity. The heating was carried out in an electric furnace
especially constructed for this purpose, in which constant
temperatures could be readily maintained.
Rosin SEPARATION.—The method found to work
satisfactorily for separating the rosin from the china-
wood oil is based upon the fact that all drying oils
as well as most semi-drying oils give a comparatively
insoluble sodium soap in cold absolute alcohol, while
the sodium soap of rosin is comparatively soluble in
that menstruum. When the proper conditions are
observed as to temperature and concentration of the
absolute alcoholic saponifying solution, the soaps
resulting from the saponification of the rosin-chinawood
mixture will precipitate in a sufficiently granular
form to permit of ready filtration.
Procedure—Ten grams of the rosin-chinawood oil mixture are
saponified in 200 cc. of N/2 absolute alcoholic sodium hydroxide
solution in the cold, the operation requiring from 12 hours, for
mixtures containing but little polymerized elaeomargaric acid
tri-glyceride, to several days, for those containing a large per-
centage of the polymerized product. This solution is then
warmed and allowed to cool slowly to o' C. It is then filtered
upon the ice-cooled copper Buchner funnel, and the sodium
soaps are washed with cold absolute alcohol. They are then
removed, dissolved in warm water and the acids liberated in a
separatory funnel by the addition of hydrochloric acid. The
fatty acids on standing will separate out on top, and the watery
solution may be drawn off; the fatty acids are then taken up in
ether and transferred to a weighed flask; the weight is then
determined by evaporation of the solvent and drying by the
addition of absolute alcohol and redistillation of the alcohol.
In every case except at the temperatures above
25o° C. the fatty acids amounted to approximately
(20)
the same weight, 4.760 g. This is what we should
expect, assuming that the rosin does not combine with
the Chinese wood oil but exhibits the same behavior
toward the oil as did naphthalene in the experiments
in which that material was used as the diluting sol-
vent. - • * -
FATTY ACIDs—As mentioned above, the mixtures
heated at the higher temperatures did not give the
same weight of fatty acids, but less in every case,
the loss in fatty acids being a function of the tempera-
ture and length of heating. Further investigation
revealed that this loss was due to decomposition.
The unsaponifiable matter thus formed passed, into
the filtrate, and the acid formed as one of the products
of decomposition gave a soluble sodium soap in ab-
solute alcohol. This loss in fatty acids, gave a fair
approximation of the percentage decomposition,
assuming that the samples of oil should ordinarily
give 4.760 g., if no decomposition had taken place.
When rosin is heated alone out of air at the higher
temperatures, the acid number drops, the decrease
depending upon the temperature used for heat treat-
ment and the time of heating at those temperatures.
The saponification number, however, remains the same.
Apparently ordinary anhydrides are formed. -
When Chinese wood oil is heated below decompo-
sition temperatures the acid number drops, and at
the point of gelatination the acid number has prac-
tically disappeared. When heated to higher tempera-
tures the acid number increases and unsaponifiable
matter is formed, but not in proportion to the actual
amount of decomposition. This is due to the fact
that an acid volatile at these temperatures is formed
which is not indicated in the acid number. Gelati-
nized wood oil heated at 250° C. for several days, out
of air, again becomes liquid, as mentioned by Schap-
ringer" and was found to have an acid number of 76.5.
A very large per cent of decomposed products had been
formed, which in turn had caused a de-gelatination of the
remaining complexes of the di-molecular tri-glyceride.
Since the acid number of the rosin falls and that of
the Chinese wood oil rises it is impossible to follow
the rate of decomposition of the oil from the acid
number of the mixture. - -
color—When rosin is heated, out of air, at tempera-
tures as high as 275° C. for as long as 24 hours it re-
mains as clear as the original rosin, in striking contrast
* Dissertation. - .
(21)
to heating in air at the same temperature, or even at
considerably lower temperatures. When heated for
24 hours in air at the above temperature it is so badly
decomposed and dark colored as to be valueless as
a varnish material. r - *.
Like results occur in the heat treatment of Chinese
wood oil, especially when heated in metal vessels.
Chinese wood oil driers may be made with but little
change in color if heat-treated out of air.
This was found to be due to the fact that the oxida-
tion products are quite susceptible to decomposition
by heat, and the resulting decomposition products
darken the mass. -
When chinawood oil or rosin-chinawood oil mixtures
were heated at high temperatures around 275° C.
out of air they remained as clear as the original mass,
and upon being blown at 150° C. for hours they dark-
ened only to a very slight degree, although there were
considerable oxidation products formed. It is apparent
that the oxidation products are not much darker
than the original oil. The manufacture of varnishes
out of air is to be recommended since the matter of
color is important. This is especially true in the manu-
facture of driers that tend to be highly colored by the
ordinary procedure.
The results obtained upon the rosin-chinawood oil
mixtures are given in Table I, from which the curves
in Figs. I and II have been plotted.
TABLE I-PolyMERIzarron AND OxIDArron of MIxºrures of CHINEs):
WOOD OIL AND ROSIN
SERIES A SERIEs B
CoNDITIONs Chinese wood oil 50% Chinese wood oil 50%
OF Rosin 50% Limed rosin (4%) 50%
ExPERIMENTs % PolyMER % % PolyMER O
Temp. Time Out of In OxyacIDs Out of In Oxyacids
•C. Hrs. air air In air air air In air
150 58/4 23.2 41 .. 5 5.2 26. 1 43.0 6. I
16?/s 35. O 48. 1 15.0 tº gº e • . . © tº
24 42.7 52. 1 21.8 43.4 53. 1 23.6
175 58/4 44.3 53. 6 7. 9 46.3 55. 1 8.6
162/s 64.4 71 .. 2 20. O 67.9 72. 4 21.5
24 77.7 82.8 27.7 80. 7 83.3 29.4
200 l 37.9 35.8 l l .9 52. 1 53. 4 12.8
4. 67. 1 73.6 15.6 82.4 77. 6 16.7
8 79.2 93. 1 19.4 100.0 100.0 20.6
24 100.0 100. O 25.7 e & & gº º º 26.8
225 I 85. 6 80. 4 9. 1 87.6 83. 2 8.8
4. 100.0 98.7 11.3 100.0 100.0 1 1.0
8 tº ſº º 100.0 10.8 tº e ſº . . . . . .
250 1 100.0 97.8 3.0 100.0 100.0 3.4
3 © tº e 100.0 9.6 tº e º tº e º © tº
275 I 100.0 100.0 6.3 100.0 100.0 3.8
200 l 53.4 55. 1 14.6 SERIES C. . .
4 84.7 79.6 18.2 | Chinese wood oil... . . . . 50
8 100.0 100.0 22.6 ſ Rosin...... . . . . . . . . . . . 49
24 © tº e tº e º 30.3 || Litharge. . . . . . . . . . . . ... 1%
FILM—The pure polymerized tri-glyceride dries
exceedingly slowly, which is to be expected from its
small iodine number. It gives, however, an excellent
(22)
film, very resistant to saponifying action as compared.
with the elaeomargaric acid tri-glyceride, and does.
not turn white as is the case with the latter glyceride,
but remains clear and glossy. This change in color
and sudden setting of the film in the case of the raw
chinawood oil, is as suggested by Meister,” beyond .
doubt due to the transformation of elaeomargarine
into its stereoisomeride, 6-elaeostearine, as shown by
the following experiments:
A small quantity of Chinese wood oil was spread over the
interior of a thin-walled, round-bottomed flask, making as thin.
a film as possible. The flask was then evacuated and filled with
hydrogen; this was done several times and then placed in direct.
sunlight. At the end of two days the oil had turned white,
and had been transformed into a soft, greasy film, somewhat.
softer than a like film formed in the air, which may be due
to the lack of an exterior tungoxyn film. It was found to be
largely insoluble in light petroleum ether, as is the case with the
light break, but soluble in the other organic solvents, showing
that oxidation had not occurred in sufficient quantity to affect.
the solubility of the film.
IOO
----- "A
4.5670.9
OOOOOO
3
O
2
O
HOURS
FIG. I–PolyMERIzarron CURves
Solid Lines Refer to “Out of Air” Treatments
Broken Lines Refer to “In Air” Treatments
These results indicate that considerable elaeomar-
garic acid tri-glyceride is transformed into its stereo-
isomeride, 3-elaeostearic acid tri-glyceride during
the drying process, and is responsible for the change
in color of the film, as suggested by Meister. This
is further shown by the drying of the heat-treated
wood oil to a clear film, for the polymerized tri-glyceride
exists in a sufficient quantity to mask the partial
1 J. Soc. Chem. Ind., 80 (1911), 95.

(23)
change of the remaining elaeomargaric acid tri-glyc-
eride into its stereoisomeride, if not to prevent it
entirely. The initial setting must also be due to this
change, and does not take place with the heat-treated
wood oil containing considerable polymerized tri-
glyceride.
RATE of oxid ATION.—The rate of formation of oxi-
dation products is slow when blowing the polymerized
material. Superheated oils, subjected to blowing
at 1.5o to 160° C., exhibit the same tendency as does
3O
.
O
O
Hours
Frg. II—OxIDArron Curves-ALL TREATMENTs “IN AIR”
the pure dimolecular tri-glyceride. It is possible
that the unsaturated linkages which have been neu-
tralized through polymerization were originally more
susceptible to oxidation than the remaining double
bonds, though the slow oxidation of the polymerized
tri-glyceride may be due to a certain rate of formation
of oxy-products dependent upon the extent of un-
saturation. This slow rate of drying characterizes
rosin-chinawood oil varnishes that contain consider-
able polymerized oil.
RATE of PolyMERIzATION.—Upon examination of
the curves representing the rate of polymerization
at 150 and 17.5°C. (Fig. I), the polymerization is found
to be decidedly less than for 200° C.; further, the rate
is about the same for heat treatment in air and out of
air, the oxidation products having but a small influence.
The rate of polymerization for the mixture containing
the limed rosin is slightly higher than for the rosin-
chinawood mixture alone.
At 200° C. the rate of polymerization for the rosin- .
chinawood oil mixture containing litharge and the
mixture containing limed rosin have the same approxi- . .
mate rate in and out of air. The rosin-chinawood oil
mixture without the addition of any material exhibits
considerable difference from the rate of the mixtures
above. Here the oxidation products have consider-

(24)
able catalyzing influence, and the rate for the mass
heated in the air is decidedly greater than for the mix-
ture heated out of air. The former has a rate approach-
ing that of the other two séries. w
At 235° C. the rate for the limed rosin-chinawood
oil mixture and for the 5o : 5o mixture is very nearly
the same, in or out of air. A reverse occurs, however,
in the rate in and out of air, the latter having a higher
rate of polymerization than the former, which the
writer believes is due to the cooling effect of the air
in contact with the former mass predominating at
this temperature over the catalyzing influence (see
Fig. I). w
At the temperatures of 25o and 27.5°. C., although
we should have more rapid oxidation at these tem-
peratures than at the lower ones, we have pro-
nounced decomposition, and the oxy-products appear
to be decomposed as fast as they are formed. Com-
plete polymerization takes place at the end of 1 hour,
in and out of air, in both series. At the end of 4 hours
the decomposition at the former temperature, and at
the end of 1 hour the decomposition at the latter tem-
perature was about the same, amounting to nearly
27 per cent. *
The effect of the lime and litharge appear to be the
same as the effect of oxidation products. They are
decided catalyzers at 200° C., somewhat less at the
lower temperatures, but at the higher temperatures
the rate of polymerization is so rapid that they have
no pronounced influence (see Fig. I).
It is not surprising that the varnish maker finds
great variations in the rosin-chinawood oil varnishes
... since a slight variation in the temperature changes the
rate of polymerization considerably; also the rate is
subject to catalyzing action by the presence of cer-
tain materials. - s
The ordinary procedure for making rosin-chinawood
oil varnishes has no apparent advantage over heating
the two materials together throughout the heat treat-
ment, except to permit a shorter period for heating
the rosin which, as mentioned before, is subject to
darkening when heated for a time in the air. The
method of heating together has decided advantages
in giving uniform results, by permitting the control
of the amount of polymerization by a judicious selec-
tion of temperature and length of heating, as well as
permitting the heating out of air, in which case the
mass remains perfectly clear at normal polymerizing
(25)
temperatures. With a range of rate of polymeriza-
tion of 25.7 per cent for 5 hours and 72.4 per cent for
'66 hours at 1.5oº C. to complete polymerization in
1 hour at 250° C., chinawood oil shows decidedly
greater tendency for polymerization than linseed oil,
which requires considerably higher temperatures.
This fact must be taken into account in the use of china-
wood oil at temperatures at which linseed oil works
the most satisfactorily.
SUMMARY
I—The polymerization of Chinese wood oil takes
place by the union of double bonds of the (wood oil)
fatty acids, analogous to the polymerization occurring
on heating linseed oil. This polymerization takes
place through the dissolution of one unsaturated .
linkage in each elaeomargaric chain in the tri-glyc-
•eride, forming a tri-glyceride composed of di-molec-
ular fatty acids. This intermediate product, or
polymerized tri-glyceride, has the power of uniting,
forming molecular complexes under favorable condi-
tions, giving an insoluble colloidal mass, not, however,
accompanied by a further loss of double linkages, as
-evidenced by the iodine number.
II—The polymerization is “mesomorphic” and
the intermediate product can be isolated. It has
been found to be the same in all cases whether the
wood oil is heated alone or in the presence of a sol-
vent, and does not vary with the temperature. The
intermediate polymerization product is soluble in
all the ordinary organic solvents, with the exception
of extra light petroleum ether, as well as in unchanged
wood oil, rosin, and like solvents; its solubility is not
dependent upon the presence of unchanged wood oil. . .
III—The percentage of polymerized tri-glyceride
in gels of the same apparent consistency is not neces-
.sarily the same, and the gel is not entirely dependent
upon the concentration of the polymer or upon heat
for its formation. -
IV—A satisfactory method for the examination
of wood oil may be based upon the property of elaeo-
margaric acid to crystallize from a dilute solution of
alcohol, while the fatty acids from the polymerized
tri-glyceride of wood oil as well as those from the other
'drying and semi-drying oils do not have this property.
V—Wood oil can be completely polymerized upon
heating at abnormally high temperatures for a short
time, such as at 350 °C. for 15 minutes, due to the
(26)
presence of decomposition products formed during
the superheating process; it will not gelatinize under
this treatment. -
VI—The rate of polymerization for wood oil is
considerably faster than for linseed oil at corresponding
temperatures, taking place completely in one hour at
25o° C. At 150° C. only about 72.4 per cent polym-
erization takes place in 66 hours. g •.
VII—Wood oil is subject to decided oxidation
when heated in air. Progressively increasing amounts
of oxidized materials are formed whenever wood oil
or wood-oil mixtures are heated in the air.
VIII–Lime, litharge and oxidation products cata-
lyze the polymerization of the elaeomargaric acid tri-
glyceride into the intermediate product. Oxygen does
not catalyze the rate of polymerization except indi-
rectly through the oxidation products. A large num-
ber of metals also act as positive catalyzers at the
higher temperatures.
- IX—The course of polymerization can be followed
by the iodine number alone and does not require
correlating with the specific gravity and refractive index.
X—Dissolving the solid gel in rosin or continually
heating it above 250 ° C. constitutes a reformation of
the soluble intermediate products. In the latter case
considerable decomposition occurs, the products of
which are responsible for the breaking down of the
molecular complexes of the polymerized tri-glyceride.
This decomposition is considerable on heating wood
oil or wood oil mixtures above 250° C. for any length
of time, and is accelerated by the presence of certain
metals. .
XI—The darkening of rosin or rosin-chinawood
-oil mixtures on heating in air is primarily due to the
-decomposition products from the oxidized oil, which
are very sensitive to temperatures above I 75° C.
. This decomposition may be largely avoided by carrying
-out the heat treatment out of air. This is also appli-
cable to the manufacture of chinawood oil driers.
XII—The pure polymerized product of chinawood
oil dries very slowly, as might be expected from the
small iodine number. This slow rate of drying char-
acterizes such rosin-chinawood oil varnishes as con-
tain large percentages of polymerized tri-glyceride.
It is with great pleasure that acknowledgment is
made to Professor E. E. Ware for his 'interest in this
work and for his valuable suggestions and criticisms.
UNIversrry of Michigan, ANN ARBox -
(27)

2,70ſºo 7604.3 ouſ 7