VCITABIt[ (Sitarch) CLU[
Ieviscd May 1950
,,b~ lb ,.'L 0111i X ,Jl '
UNITED STATES DEPARTMENT OF AGRICULTURE
<. FOREST SERVICE
FOREST FPJCTS ---X R A T 0 R Y
Madison 5, Wisconsin
In Cooperation with the University of Wisconsin
Digitized by the Internet Archive
-cl ,TABLE (..:r .}?) GLUE
F'rect productss Laboratory,- Torest Service
U. S. Department of Agriculture
Vegetable glues are used in woodworking mainly for making plywood
and other veneered products for interior use, but they have found some
application in edge gluing and in certain assembly o rations. These glues
are usually prepared by the user by heating the vegetable-starch powder with
water and then cooling the mixture to room tem-erature before use. Some
;r':.ared vegetable glues for woodworking have been made available in liquid
fnrm, ready for use. Vegetable glues are room-temperature-setting and
develop their strength by loss of water. They produce joints of high dry
.trcngth but with low water and moisture resistance.
The information 'resented here on the composition of vegetable glue
has been obtained largely from the literatui-re, from -atents, and from manu-
facturers and users of such glues, while the information concerning the
methods of using vegetable glue for joining wood is based in part on experi-
ments at this Laboratory and in pe-rt on observations of its use in woodwork-
Source of Vegetable Glue
The raw material for vegetable glues is starch, a carbohydrate
occurring abundantly throughout the vegetable world. The use of starch in
the food industry is much .-reater than is its use in woodworking glues. It
is sometimes found isolated in larg0- quantities and in a il'hi- d&:rree of
purity in certain plant _r^-ns, such as the stems of the sago palm, the fruit
of the '.,,nan-, the sd w of' w'., at and corn, atnd the roots of the potato,
arrowroot, and cassava. Starch is the Trinci al form in which plants store
food for future requirements.
-Maintained at Madison, Wis., in cooperation with the University of
Agriculture-" 'd scn
Report No. E30
In spite of its great variety and wide distribution in nature, there
are comparatively few sources of commercial starch.2 By far the greatest
amount manufactured in the United States is made from corn, which averages
about 55 percent starch. Considerable potato starch is also made in this
country, as well as some wheat starch. Tapioca (cassava) and sago starches
are imported to some extent from the Far East. The latter is used narticu-
larly in the manufacture of envelope gums. Cassava starch is noted for the
uniformity of its paste. Of these starches, cassava is most often used in
the production of vegetable glues for woodworking, although other starches
can be used for this pur-oose.
The general principles- of starch manufacture are: (a) Disintegrating
the plant tissue in such a way that the starch grains are set free but not
ruptured; (b) separating the starch from the gluten by diluting with water
the disintegrated mixture, which has been previously treated with chemicals
or subjected to fermentation, and by then settling out the heavy starch by
subsidence; (c) washing the starch by agitating with water in tanks,
"running," or decantation; (d) recovery of the starch by draining in cloth-
bottom boxes or in deep frame filter presses; and (e) drying the starch in
The starches from different sources will, of course, vary in certain
properties; and, in addition, there may be variations in the starch from
any one source depending upon growth conditions and on the methods and the
care used in the production. Some of the starches, such as cassava, are
marketed in different grades, and it is reported that the grade of the
starch influences the properties of the adhesive.
Development of Adhesives from Starch
The use of adhesives from starches or sugars is supposedly a very
ancient practice, although in woodworking industries it is a comparatively
recent one. In connection with the early history of paper making it is said
that the Egyptians used thin sheets of papyrus bound together with an adhe-
sive made from a starch base. It is further reported that Chinese,
centuries ago, used starch in some form for gluing wood. In more recent
times attempts were made to use sugars as a base for an adhesive. In 1891 a
patent was granted to Higgins on an adhesive of dextrin and borax, and
another in 1894 to Wagner on an adhesive of dextrin, copper sulphate, sugar,
and nitric acid. An early United States patent on the use of starch as an
adhesive was issued to Murphy in 1896. Bloede re'oorts that a French patent
was issued to Gerard in 1874 on a glue consisting of 1 part potato starch,
5 parts water, and 0.08 part caustic soda. It is probable that these
earlier adhesives did not have sufficient strength to warrant their use as
glues for woodworking. In 1908 a patent covering vegetable glue was issued
to Frank G. Perkins. Under this and later patents the Perkins company
-Rogers, Allen. Industrial Chemistry; A Manual for the Student and
Manufacturer. 6th ed. Van Nostrand, New York City. 1942.
Report No. R50
started the comrcnrcial production of vegetable glue for wocdwcrkin,-. More
recently, still other -atents on improvements in vegetable glue have been
issued to Tunnell, Bloede, and others.
Sources of Cassava
Cassava belongs to a class of woody shrubs and is related, botanically.
to the rubber tree. It will attain a height of 12 to 16 fcct. Two-year old
plants are generally used, however, and at this time they arc ordinarily 6 tc
10 feet in height. The roots are tuberous, resembling a giant Sweet potato,
and it is from these roots that the starch is obtained. Cassava -robably
originated in Brazil, but it has reached its greatest production in Java; and
it is from Java that the bulk of the present su ply is received.
The., raw starch m"y be converted directly into an adhesive, or it may
receive an intrm.diate treatment known as ",rocessing." The bulk of the
vegetable glue used in woodworking has been made from so-called processed
starch. The relative im'ints of the -orocessfd and raw starches used at the
present time are unknown.
Starch is often processed before its final conversion into glue for
the purposes of: (1) Making it more soluble by modifying the outer wall of
the starch grain, (2) reducing its water-absorbing quality and thus producing
a glue of low water content, and (5) producing a uniform product.
The processing may consist of dig-sting the starch in hot water
acidified with sulphuric acid and of later neutralizing the acid, or of
treating the starch with sodium peroxide and later neutralizing the alkali.
The processed starch is then drained and dried. These are only two of
several methods of recessingg, most of which are covered by -atents or care -
fully guarded as trade secrets.
Pri-paration of the Adhesive
'hTethcr the basic material is a raw rtarch, a nrocesscd starch, or a
starch to which chemical r'n-r.--dients havk. ben added, the preparation of an
adhesive ordinarily follows one of three general methods. T.- e are treat-
rrint with (1) caustic soda and heat, (2) caustic soda alone, or (5) heat
If both caustic soda and heat are used, the amount of caustic is com-
monly in the neighborhood of 3 Iercent ((based on the weight of the dry' starch),
alti:u.i'. it r'jay be varied to met special r quircmnts,. A temperature in
the general neighborhood of 15' F. is required to convert such a mixture.
Report i1o. R50
When the conversion depends on the action of caustic soda alone, about
7 percent caustic is customary, although variations of 6 percent to 10 per-
cent are not uncommon.
A temperature of about 150 F. will convert the starch without the
addition of caustic soda.
The amount of water used depends upon the variety and previous treat-
ment of the starch. Processed starches or those containing added chemicals
usually require a starch-water ratio of 1:1.2 to 1:2.5. If the raw starch
is used, the proportion of water may be somewhat higher (such as 1:3.5). The
use of a starch-water ratio of 1:14 to 1:20 is reported for the preparation
of a vegetable-glue size.
The caustic soda in the mixture produces the stringiness that is desir-
able in this class of glues. The glue without caustic or other chemicals
resembles paste, but good joint strengths can be obtained with it. The higher
percentages of caustic cause the glue to set somewhat more slowly, and this
allows the use of longer a-ssembly times but requires a somewhat longer time
in the clamps. Furthermore, the higher is the percentage of caustic, the
greater is the danger of staining such woods as oak, maple, cherry, elm, ash,
birch, and beech. For these reasons a compromise is usually made, under
which the first method is most generally used.
The equipment commonly used for the preparation of vegetable glue con-
sists of a mechanical mixer surrounded by a steam-heated water jacket. It is
possible, but not at all convenient, to prepare small batches of these glues
over a water bath with hand stirring. The general procedure for conversion
of starch into an adhesive is as follows:
The required amount of water is poured into the mixer, the starch is
added, and the mass is stirred until any lumnr's have been broken. The stirring
is continued while the caustic soda (previously dissolved in about three times
its weight of water) is slowly added. Steam is now turned into the water
jacket, and the temperature is increased gradually (the stirring being con-
tinuous) until the water in the jacket is somewhere between 190 and 200 F.
(This is, of course, considerably in excess of the t.:mpr-.-.ture required to
convert the starch, but a considerable temperature gradient will exist
between the water jacket and the center of the inner container, so that the
average te-mpe-rature of the starch sus-ension is not greatly in excess of
150 F.) About one-half hour is required to bring the tem'.rature of the
water jacket from room temperature to 200 F. At the end of this time it will
be noticed that the starch has changed in appearance from a white suspension
to an amber-colored stringy paste. If picked up on a paddle and held to the
light, it will appear almost transparent. When it reaches this stage, the
conversion is stopped by turning off the steam and by replacing the hot water
in the water jacket with cold. The stirring of the glue is continued until
the temperature has been reduced to about that of the room. The glue is now
ready for use. Details of the general procedure given above may be varied
under the instructions of the glue manufacturer according to the chemical
additions and to the grade or previous treatment of his starch.
Report No. B30
When a starch suspension is overheated, there suems to be some dan;'.r
of an undesirable chemical ch.n':.e in the nature of the paste. This ch'.ng.
has been called :ar-m lization, and it is characterized by a reddish-brown
coloration in the glue. No att-m -t will be made to explain the exact nature:
of this change; the result is said to be a reduction in the adhesive -owcr
of the glue. In the preparation of a starch adhesive, this caramelization
is guarded against by: (1) Adding the caustic soda to the starch suspension
before applying heat, (2) controlling the temperature of the water bath
during the mixing operation, and (5) stopping the conversion when the adhe-
sive has reached the tr".nsyarent, amber-colored, stringy stage.
Application of Vegetable Glue.
Vc'!table glue is very viscous and difficult to spread by hand, but
on a mechanical spreader it works very well. The mixtures commonly used dry
slowly, and their working life is very long. In plant o orations, the glue
rom'_ining from a day's run my be left in the spreader, the s:preader be
cor.'ard with a damp cloth at night, and fresh gluo be added in the morning.
Joints have been made in the laboratory with glue that has stood for 42 days
after mixing, and the resulting strength has been entirely satisfactory.
This slow rate of chnn.c in viscosity enables the manufacturer to use a long
assembly t-m. in gluing, which is particularly desirable in the production
of plywood. However, this same characteristic of slow setting is sometimes
a disadvantage in joint work where the stock is often under pressure for
only a short time.
Tests based on the more common types of vegetable glues have shown
that the strength of joints increases slowly during the first 2 hours after
pressure is applied. During the p-riod of 2 to 6 hours after applying the
pressure, the rate of increase is somewhat more rapid, and after about 8
hours it becomes more gradual again. Joints of animal glues, casein glues,
and some of the newer vegetable glues will increase in strength more rapidly.
In thick stock, the joint will be strong enough for machining after about
7 hours, but 2 hours in the press :s sufficient, provided that the joint is
allowed to condition further for a day before machining. In veneer work,
the best results may be expected if the joints are left under pressure over-
night and are then removed and allowed to dry or to condition thoroughly.
Although most of the vegetable glues arec slow in setting and all that
have come to our attention are low in moisture resistance, attempts are being
made constantly by the producers to improve their glues in these respects.
Manufacturers have prepared starch glues that show a more rapid increase in
joint strength than do the ordinary starch-caustic mixtures. The attempts to
produce a water-resistant starch glue have not, so far as is known, met with
Experiments conducted at the Forest Products Laboratory with a good
grade of glue have shown that a spread of about 45 square feet of single glue
line per pound of dry glue can be defended uron to produce very satisfactory
Report No. P30
results with normal conditions of pressure and assembly time. The spread
may, of course, be varied, but a decrease in the strength of the joint will
probably be noticed if the amount of glue is cut so as to get more than 55
or 60 square feet of glue line per pound of dry glue. On the other hand,
there is some danger in increasing the amount of glue spread too much be-
cause of the amount of water added during the gluing operation. Vegetable
glue, after being spread on the wood, ordinarily remains in a condition
satisfactory for pressing for periods up to about 25 minutes. A pressure
of about 200 pounds per square inch gives good results when the assembly
time varies over a rather wide range. In commercial operations, however,
a pressure of from 100 to 150 pounds per square inch is commonly used with
success. If the assembly time is short, say from 1 to 5 minutes, a pressure
of about 100 pounds per square inch may give slightly stronger joints than
one of 200 pounds per square inch. Since vegetable glues generally contain
large proportions of water, it is usually necessary to recondition glued
stock after removal from pressure before further machining or finishing.
Glued stock for interior use should have a moisture content of 6 to 8 per-
cent before final machining.
Durability of Vegetable-glue Joints
Vegetable glues are generally low in water resistance and are very
similar to animal glues in this res-ect. Conventional shear-test specimens
of yellow birch plywood (three-ply, 3/16-inch) bonded with a typical
vegetable glue were exposed under each of several repeating cycles. The
cycles consisted of exposure, at a temperature of 80 F. for each :hase,
for 14 days at a high relative humidity or of soaking in water for 2 days,
followed by exposure for 12 or 14 days at 30 percent relative humidity.
Such cycles caused swelling and shrinking of the wood and subsequent mechan-
ical stresses on the joints. At some of the higher relative-humidity
conditions softening of the glue also occurred. The exposure cycles are
given in table 1.
Specimens subjected to test No. 6 (table 1) separated completely,
either in the first cycle or in the drying that immediately followed.
Specimens exposed to test No. 5 showed at the end of the first complete
cycle an average joint strength of less than 40 percent of the original.
All specimens failed before the end of the fourth cycle, or in somewhat
less than 16 weeks. A relative humidity of 97 percent is sufficiently high
to permit development of molds and to bring the moisture content of the
wood to about 28 percent during the wet half of the cycle.
Specimens exposed to test No. 4 failed less rapidly than those ex-
posed to test No. 5. They had decreased in strength, however, by some 50
percent by the end of the first test cycle, and all specimens had failed
by the end of 36 weeks. Moisture content of the wood during the wet half
of this cycle probably reached about 22 percent.
In test No. 3, more than 20 weeks of exposure were required before
the average test values had fallen to 50 percent of the original. By the
Report No. R3B50
end of the 81 -,'ecks, however, all specimens had failed, which indicated that
vegetable-glue joints cannot be depended upon to remain permanently durable
when exosed to conditions under which the wood may at times exceed about 17
percent moisture content, which is alpr',ximately the equilibrium condition
for :C percent relative lar-iidity. Furniture and other productss bonded with
v,-ctablrl glues often s-rv. satisfactorily in spite of occasional txporures
to relative humidities in excess of &0 percent, but in those cases -rotec-
tion afforded by the finish may pr.v'ent the moisture content of the' wood
from reaching equilibrium values, particularly if the exposure to damoness
is not prolonged.
In tests 1:.. 1 and 2, no evidence of significant loss in strength
could be detected during the 160 weeks that the tests were in progress.
Test No. 2 a&.roxhmates the chani-,cs in moisture content that can be expected
in interior woodwork in normal use in heated buildings in the- northern
part of the United States. In this type of service, propmrly designed and
well-made joints of '-,'ctable glue should prove ermanently durable, and
experimental evidence, other than thes(; tests, supports that belief.
Uses of Vegetable Glue
Vegetable glue is utilized chiefly in plywood and other veneered
products manufactured for interior use, such as furniture of all kinds,
pianos, and interior trim. Because of their slow setting the ordinary
vegetable glues are not extensively used in edge-joint gluing, but it is
r,-ported that the increased setting rate of some of the newer types has
allowed vegetable glues to compete successfully in this field.
The lack of water resistance prevents the use of vegetable glue in
products that will be subjected to moist conditions.
Although research on starch has been extensive, no simple chemical or
mechanical test has been devised that will permit the user to test starch
for quality and for suitability for glULmtk.mng.
If it is desired, the starch.-s may be distinguished from each other
by such tests as the temperature of i..tinization, by the form and size of
the grain, by the appearance under polarized light, by the refractive index,
by the character of the paste, by the color reactions, and probably by X-ray
determinations of the crystal structure.
Consistency tests at different starch-water ratios may give useful
information, but the user of the glue must, for the mcst part, rely upon
joint tests in determining the quality of starch for gluemaking.
Report Tlo. R30
Table l.--Exposure cycles of yellow birch plywood bonded with
: First stage Second stage _--=
Test Type of ----------------- ------------------
exposure cycle : Time : Relative : Time : Relative
: humidity : : humidity U
Days Percent Days : Percent
1 : Continuous (controls) :..... : 30 .... .......
2 Alternating : 14 : 60 : 14 50
5 : .........do.......... 14 : 80 : 14 : 50
4 ......... do ....... 14 : 90 14 : 30
5 : ......... do.......... : 14 : 97 : 14 : 30
6 : ......... do ......... 2 : ( ) : 12 : 30
-Soaked in water at room temperature.
Report Ho. B350