IESISTANCE TO ICACIINC ANU UIECAY PIRPCTECTION Ul
VAIIOUUS PIRECIPITATES OIMEID IN WOOlD IY
UNITED STATES DEPARTMENT OF AGRICULTURE
FOREST PRODUCTS LABORATORY
In Cooperation with the University of Wisconsin
Digitized by the Internet Archive
RESIST:3cE 73 LEACHGIi& A:'D DECAY PRCTECTIOCT OF
VARIOUS PRECIPITATES FOBiiED I. WO.-OD BY DOUBLE DIFFUSION*
By R. H. EA.CHLER, Chemist
It has long been recognized that a water-soluble material may be
introduced into green or water-logged wood by merely steeping the wood in an
aqueous solution of the material which then tends to diffuse from the solu-
tion into the -,nter contained in the wood. This phenomenon has been utilized
in the chemical seasoning of wood' but for various reasons its application in
the field of wood preservation has been limited.2 A disadvantage of the
process is the fact that it is confined to materials that are water soluble
and, consequently, subject to leaching from wood exposed to damp surroundings.
-he experiments here described were undertaken to investigate (1)
the possibility of precipitating materials of very low solubility in wood by
a two-stage diffusion treatment, and (2) the resistance to leaching and the
toxicity to wood-destroying fungi of certain materials so formed.
Only a small number of the compounds that may be precipitated from
an aqueous system are relatively insoluble and at the same time toxic to wood-
destroying fungi. Of this small number a few are very expensive. The ex-
periments reported here were confined to the following precipitates: copper,
arsenate, copper chromate, nickel arsenate, nickel chromate, and .-.agnesium
ammonium arsenate. Parallel tests were made on the soluble salts3 used in
forming these precipitates, as well as on zinc chloride, a preservative of
Penetration of Salts
It was assumed at the start that if wet wood containing one soluble
salt were immersed in a solution of a second salt which reacted with the
first to form an insoluble precipitate, the ingress of the second salt would
,Presented at 37th Annual Leeting of the Ainer. Wood-Pres. Assn., Louisville,
Ky., Feb. 4-6, 1941; and published in the Proc. 1941, :. 23.
iChemical Seasoning of "'Thod, by W.. Loughborough. South.Lbrman.,Sept.15,136.
-2The osmose process, although not a steeping process, depends upon diffusion
as does the process of filling alternate holes with copper sulfate and
sodium arsenite, as described by J.F.Harkom in the Proc. of the A-er.
Wood-Pres. Assn., 1939, p. 306.
2-o tests were made on magnesium chloride, which is known to be practically
nontoxic to fungi.
be retarded by the precipitate as it formed in the outer parts of the wood.
This would n cessitate a longer steeping period for tho second salt, other
conditions being equal. This problematical degree of retardation, rather
than know. differences in the rates of diffusion of the various salts, made
the scl!ction of treating periods a matter of guesswork.
Preliminary tests were made in which 15-inch lengths- of peeled
gr.-ecn asxcn fence posts 4 inches in diameter were steeped in a 10 percent
solution of copper sulfate for five days and then transferred to a 20 percent
solution of sodium chromate. (A saturated solution of sodium chromate has a
conccntration about twice that of a saturated solution of copper sulfate at
the same te.peerature; the exact ratio depends upon the temperature at which
the cor.>r.rison is made.)
Pieces were removed at the end of the second five days and kiln
dried, after which discs were cut from the middle and %round up for analysis.-
It was found that the absorption of each salt amounted to roughly 3 percent
of the dry weight of the wood. This was considered satisfactory ratio of
the two chemicals since their molecular weights are nearly the same
(CuSO, = 159.S, Na2Cro = 162) and the copper and chrormate ions are both
div!.lLt. WhT.n the order of treatment was reversed, thrt is, the wood
steeped first in the sodium chromate solution and then in the copper sulfate
solution, a marked deficiency of copper sulfate was found. These tests were
repeated with other combinations of salts and similar results were obtained.
When the wood was first treated with the salt of the heavy metal, magnesium,
copper or nickel, subsequent penetration by sodium arsenate or sodium
chromate v'as coruparatively rapid. However, when the wood was first treated
with the sodiumia salt, penetration by the salt of the heavy metal was much
slower. This difference may be explained on the basis of difference in pH.
Due to hydrolysis, the solutions of the salts of heavy metals are aci Lic,
whereas solutions of disodium arsenate and sodium chromate are alkaline.,
All chromc.tes and arsenates are soluble in dilute mineral acids; consequently,
when wood is first treated with the salt of the heavy metal no precipitate
would be expected to form until a sufficient amount of the alkaline sodium
salt s rrsent.
More extensive teosts have been started to study the gross absorp-
tion and distribution of different salts from varied concentrations and at
varied porio s. Results of these tests are not yet available. Up to the
present, other factors which arc known to affect penetration, especially
the initial moisture content of the wood and the terperature of the solution,
have be'n kept fairly constant but their effects have not been investigated.
T?.. question of the distribution of two salts introduced by succes-
sive stc win was then studied briefly. It would be possible for a cross
Al iccos urged in subseouent treatments were peeled and were 15 inches
Ion,;. Di-eotcrs varic.d from 3-1/2 to 6 inches.
Acknowlod ;ment is made of the valuable aid of Philip Servais, student
assistant, in the conductance of these analyses.
JI.l 7 '-2-
section of a treated piece to show two salts present in an ideal proportion,
but if one were concentrated near the outside to a greater extent than the
other, it would be in excess in that portion and the excess would be leach-
able, as would a corresponding excess of the other salt in the interior of
One piece 4 inches in diameter was treated for five days in a 7.45
percent copper sulfate solution and then for 30 days in a 14.95 percent
sodium chromate solution. A matched piece was treated for five days in a
16.75 percent sodium chromate solution and then for 30 days in an 8.38 percent
copper sulfate solution. Discs were then cut from the middle of the pieces
after which each disc was cut into three circumferential sections of approxi-
mately equal width, leaving a center core with a radius equal to the v'idth
of the sections.
Table 1 shows the copper and chromium content of'the different
sections after a total of 35 days' steeping. The total amounts are, n turally,
very different from that obtained after a total of 10 days' steeping.. -
term "ring" is used as suggestive of the shape of the sections and has no
reference to annual rings.
It may be seen that the gradation in concentration from the surface
to the center was of more or less parallel order for both salts.
Table-1. Distribution of copper and chromium in treated pieces after
35 days' steeping
Treated first with Treated first with
copper sulfate sodium chromate
Copper : Chromium : Copper : Chromium
calcu- : calcu- : calcu- : calcu-
lated as : lated as : lated as : lated .s
GuSO4 : :a2CrO4 CuSO : T CrO
- - - ------------- ---- ------------------
Outside ring............ : 3.62
Second ring..............: 1.25
Third ring.............. : .44
Center core .... .. .. ... .. : .006
To study the leaching characteristics of the various chemicals,
a preliminary series of tests was first made. Blocks 5 cm. long by 2.5 -m.
wide by 1.5 cm. thick (approximately 2 by 1 by 0.6 inches) were cut from
near the outside of treated and dried basswood pieces. About 3 inches of
each enc of the piece were discarded. Some of the blocks were ground and
analyzed for their content of chemical before leaching. FZurteen blocks
were laced in a vacuum bottle, a vacuum of about 28 inches of mercury was
drawn, annd, with vacuum pump still running, 500 ml. of distilled water was
drawn into the bottle. The vacuum was then released, cau-ir:K a rapid ab-
sorption of water by the blocks so that some of them sank while the rest
floated very low in the water. The flask was then shaken occasionally and
at the end of an hour the leaching water was poured off and saved for anal-
ysis. Another 500 ml. of distilled water was poured over the blocks (omitting
the vacuum) and, with occasional shaking, left for 2 hours after which it was
poured olff and saved for analysis. In a similar manner the blocks were
leached for period of 4 hours and then for 31 periods of 8 hours each. Only
one 8-hour leaching was made in any 24-hour period and between teachings the
blocks 'were left in the stoppered bottles without water. Analyses were made
of the leaching water from each of the first four leachings, after which
leachings were combined as follows: fifth and sixth period, seventh to tenth
period, eleventh to eighteenth period, nineteenth to thirty-fourth period.
Because it was believed that the number of blocks available from
each treatment was too small to correct for variability in the blocks, the
results of the analyses were considered only indicative and will not be re-
ported in detail. It was found that the depletion of metallic ions was
much slower than from wood given a single treatment. For example, under the
conditions described, 77 percent of the copper was leached from wood treated
'with c per sulfate alone. 3nly 2.8 percent of the copper was leached from
wood treated with about the same amount of copper sulfate, followed by an
excess of sodium arsenate, and 3.2 percent of the copper was leached from
copper sulfate-sodium chromate treated wood with the latter chemical in ex-
cess. In other treatments in which a deficiency of either second chemical
was fouid, the leaching of copper, as would be expected, was much faster at
first because f" the presence of some uncombined copper sulfate. The leachingr
of nickel and magnesiumm in the presence of an excess 'f either of the sodium
salts was likewise very slow.
'Iith the acid ion, the leaching of chromate was very slow, as has
been found by various investigators. The leaching of arsenate was alw
containi; sodium arsenate and an excess of nickel sulfate as compared with
J1I percent fro!. blocks treated with sodium arsenate alone. The true picture
of ehe l;achirn of arsenic from an arsenate formed under favorable conditions
is di-'Licult to discern. An excess of the metallic salt creates an acidic
condition which increases the solubility of the arsenate. An excess of
sodium rscrn tc furnishes an alkaline condition which is desirable for its
deprcssant flfect on the solubility of the precipitated arsenate, but the
l;achin: of this excess sodium arsenate masks the le. chin of the precipitated
arsenate. When slightly soluble salts of weak bases are exposed to the action
of water, some hydrolysis takes place; the solution contains relatively .ore
of the acid ion than the basic ion and the remaining undissolvcd matui:'1
becomes progressively more basic. :Tur..:rous tests have shown this ch"n to
take place in treated wood. For this reason on, would expect the arsenic
to be leached more rapidly than the metal from wood treated with an insol-,1lI
arscnr.e andc also that the remaining material would, in time, become too 1.-
soluble to be toxic. This rate of leaching might still be very slow" co,;-
pared with that of a highly soluble salt. ::u.ch work woulc be required to
determine the most desirable ratio of any two salts from the standpoint of
toxicity, permanence, corrosion, and cost. At present it is believed that
a slight excess of the alkaline reacting salt is preferable to stoichio-
A second series of treatments was made, using sections of :rcen
maple fence posts; these posts were about 6 inches in diameter with a 2-inch
band of sapwood and had an overage moisture content of 92 percent. As a
result of experience gained in the first treatments, a more satisfactory
ratio of the reacting salts was introduced into the wood. Weaker trc-tin
solutions were used than in the first series in an effort to .rrivc at -
sorptions of salt in the region of those employed in commercial pressure
treatments. The absorptions -ctu.lly obtained were higher thn desired. On
the basis of averrge specific grr'vity, the -,bsorptions of totl chcmic, he'n
calculatold to pounds per cubic foot, ranged from 0.58 m'-gnesium chlo-ice +
sodium -.maoniuim rsenFtc to 1.78 nickel sulfate + sodium chromt,. A l rr
amount of m. teria.l was treated so that sever-l hundred blocks were obtrinXd
from c.ch treatment. One-third of the blocks from c-ch treatment ,orc
ground together "nd -nlyzed to show the concentration of chemic-ls b-for:
leching. Ti, rcm'ining blocks ,were loeched by first water loggin ae,
then 1p!-.cina them in wire ba-skets and keeping them submerged in bath
which wa.s constantly being changed by stre-am of w'ter. Th. block '.'ere
lec'ched continuously, day rnd night, for one month so th-t conditions of
le-chin, vere more severe than in the first series. Ono-hrif of the 1,c chcd
blocks, co.-iprising one-third of the ori inal trc-ted blocks, were dri. d -.d
ground for '.nrlysis. `h- rem.inin, blocks were set '-side for decoy te.sts.
The absorptions and losses of chemicals by leaching are _,ive_ in
table 2. Absorptions, as well as content of chemical after leaching, '"re
calculated on the basis of the original chemicals used in the treatment, no
specific che.iical reaction being assumed. The data on the absorptions o -
tained with different combinations of chemicals are not strictly comp rabli.
Since these experiments were exploratory in nature, the increased ti.. -
expense necessary to control closely all the conditions affecting a.bsorp-
tion were not considered justified. Solutions of about 6 percent wcr used.
The concentrations were checked by analysis but were not adjusted to any
exact fi ,'urc. Because of checks which developed in ma'nesium-ammonirn.-
Frson.ate-treoated pieces, it was necessary to cut blocks further from the sur-
facc than was otherwise done.
With allowances made for differences in conditions, th res-lts
obtained in the second series of leaching tests are in general rgreceant
with those obtained in the first series, The rr-ther high loss of nickel
from the nickel sulfate-sodium arsen-te treated wood is difficult to explain,
Ps is the unusually high loss of chromate from the copper sulfate-sodium
chromate treated wood. The loss of copper from the latter is not surprising
if the excess of copper sulfate is considered.
Decay Resistance of Treated Blocks After Leaching
The leached blocks were tested for their resistance to decay by a
method used extensively in the Division of Plant Pathology, U. S. zuircau of
Plnt Industry. It is a modification of the Kolle flask method, square
bottles with metal screw tops being used in place of the flasks. After a
vigorous growthh of the fungus has covered the agar, a small glass rod bent
into angular shape is laid upon the mycelium to support the test block.
Only one block is placed in a bottle. The screw top is closed loosely to
permit access of air and, in order to avoid drying of the agar, the bottles
are kept in a. room maintained at a relative humidity of 80 percent "nd a
temperature of 80 F.
T. leached blocks were first conditioned to constant weiht in a
65 percent relative humidity chamber. Extra blocks were used to determine
the avera e moisture content by drying at 105c. Thc oven-dry weight of each
block -t the strrt of thc test ws obtained by calculation to avoid any
chemical changes th-t might be produced through oven dryir.-. The test
blocks were immersed ii boiling water for about 10 seconds before they were
placed over the fuin:is ;m-t.. After four months' exposure they were removed,
inspected for i,ians of fuiimus attack, dried, and weighed.
Four test funi were used, na-mely, Madison 517, Coniephor" cere-
bella 658, Pori- incr Ta ta- 563, and Lenzites trabca 617, T\i cultures
were obtain: o'rig2in'LU x.':o the Division of Forest Pathoo U. S. -urau
of Pl-nt Ii stry. c. olo: : from eac.h treatment were tec, againstt each
fungus. Th loss in weig'ht reported is the average for ten blocks except
th't 'an occ-,sion!'l block was discarded for contamination of the fungus i-.t.
As h, s booeen st-ted, the number of blocks obtained in the first
scri s of treatments was inadequ-te in view of the differences between
indiviui-l blocks. In some decay tests made with 1c .-it. il. .r'l- only, the
leached blocks that had been treated with the insoluble nickel and copper
compounds showed practically no loss due to decay, whereas those treated with
single soluble salts showed considerable loss in weight. Th< absorptions
were -bove t .ose commonly used in commercial practice.
T.. results of the second series of decay tests are given in table
2. The fun ;i are placed in the order of the intensity of attack on untreated
wood, but it is apparent that this bears no relation to their resistance to
different chemicals. As has been found by others, lcntLL, tr-bera is shown
to be ".rscnic tolerant; it is susceptible to copper but does not seem to be
highly suv ceutible to nickel. In the first series of tests both nickel
chromate nid nickel rrsenate protected the wood 'C-ainst I. tr-*ba.- but they
were -rL,(.t in bout double the final "concentration of the second series.
Table 2.--Leachin a4ind decay tests on treated maple blocks
(Copper sulfate, 5 days.....:
(Sod..um arsenate, 20 days...:
(Nickel sulfate, 5 days.....:
(Sodium arsenate, 20 days...:
(4a~nesium c iloride, 5 days.
(Sodiu-i aI.,ionium .rsenate-l. .
( 20 days
(Copper sulfate, 5 days.....
(Sodium chroiate, 20 days...:
(.Tickel sulfa te, 5 days..... :
(Sodium chromrte, 2 d;,'s...:
Copper sulfate, 5 days......:
Nickel sulfate, 5 da3 s ......:
Sodium arscnate, 5 days.....
Sodium chro e, days.
Zinc chloride, 5 days ....... :
:Concentration:Loss of : Loss of weight due to decay
of chemical :chemical: during four months oX-osure
(Based on : by : to pure cultures of
oven-drywood) leaching: -------------------------------
----- ---: :Lenzites:Xodison:Conio-: Pcria
:Before: After: trabea : 517 :phora : in-
:leach-:leach- : : : cere- :crassata
ing : ing : : :bell-:
- - _ -. - -: _ :-- - ^ _. - _
Untreated con rols....... ...:. : . ...........
Solution contained 5 percent amaonixmia hydroxide.
Coniophora cerebella, on the other hand, is susceptible to arsenic but rather
tolerant to copper. None of the chemicals in the amounts present and undCer
the conditions of the test gave decisive protection against Poria incrassata.
In any comparison between a single treatment and a double treatment, allow-
ance should be made for the fact that higher absorptions of total ch-',mial
were present in the latter.
The lack of protection afforded by zinc chloride might see; to
indicate that these tests were extremely severe. It must be remembered thpt,
of the factors affecting permanence, only leachin-- and, to some extent,
hydrolysis ,were accelerated. The absence of gradual deterioration by checi-
cal changes which might take place in service remains to be proven. How-Cver,
it appears that the insoluble materials investigated are sufficiently 3rol.i-
ising to merit further study. Losses in weight combined with appeLranc'cs
of the blocks at the end of thc test point to copper arsenate as the .iost
effective of the materials studied. However, it may be suspected of prccnt-
ing a greater corrosion hazard than any of the other matcri',ls. In this
respect magnesium ammoniumJ arsenate seems less likely to attack metal first i-
ings than the others; moreover, either of the solutions used in this trertmcnt
may be hOlC' in an iron container, which is not true of copper sulfate or
Shorter treating periods than were eploycd in these experiments
might be combiined with higher concentration of solution to yield c desired
absorption of chemical. This would result in a steeper concentration
gradient in the treated wood which, while undesirable in th, prcp ration of
material for experimental purposes, might be desirable in practice.
1. Compounds of low solubility may be deposited in green woo.' b'
successively steeooping it in solutions of different salts.
2. The chromates Pnd arsentes of copper and nickel .nd ma ncsiui
ammoniiul.i rsenatc, when formed in wood by this method, rre leached more
slowly than "re soluble salts.
3. The lerch( d blocks show considerable rcsistancc to rtt, c b"y
three woo(-,'cstroying fungi and some rcsist'-nce to a fourth.
UNIVERSITY OF FtORIDA
3 1262 08927 3238 I1
3 1262 08927 3238