Influence of ResinAbstractThirty-six resins, each with two replicates, were

factorially prepared with three formulation variables:molar ratios of sodium hydroxide to phenol of 0.4, 0.7, and1.0; levels of resin solids content of 37, 40, and 43 percent;and molar ratios of formaldehyde to phenol of 1.6, 1.9, 2.2,and 2.5. Glue bond quality decreased substantially witha change of NaOH/phenol ratio from 0.4 to 0.7. A furtherchange in ratio from 0.7 to 1.0 failed to affect it significant-ly, apparently because conversion of formaldehyde by theCannizzaro reaction became important where more thanhalf a mole of caustic was present in the reaction mix-ture. On the average, bond quality increased as solidscontent increased. Changes in the CHzO/phenol ratio af-fected delamination but not wet shear strength or wood-failure percentages; the lower molar ratios yielded glue-lines that delaminated least. Latewood-to-latewood delam-inated substantially more than earlywood-to-earlywoodgluelines. Of 72 batches of resins, two met the wood-failure standard for exterior gluelines. Three yieldedlatewood-to-latewood gluelines that averaged less than20 percent delamination in 3 months of outdoor exposurein central Louisiana.

VariablesFormulation

On Bond Quality of

Southern Pine Plywood

Chung-Yun Hse

P REVIOUS PAPERS have examined the relationships be-tween basic properties of phenolic resins and glueline

performance in southern pine plywood. Surface ten-sion, contact angle, curing properties, and alkalinityof the resins were shown to be related to bond quality(Hse 1971, 1972a, b). The analyses were chiefly interms of single resin properties. Any change in formu-lation variables, however, is likely to have interrelatedeffects on these properties. In the present researdt,therefore, changes in formulation variables were evalu-ated directly in terms of bond quality. No attempt wasmade to develop an optimum resin. Rather, the purposewas to explore a range of formulations in order to de-fine effects of three main variables. Special attentionwas paid to sodium hydroxide content, since determina-tion of the optimum catalyst ratio would be very use-ful in developing phenolic resins for commercial use.

sodium hydroxide was added, as a catalyst, in threesteps; i.e., 0.3 mole of sodium hydroxide per mole ofphenol was added at the beginning, and the balance wasdivided into two equal parts, one of which was added30 minutes after the reaction began and the other ahalf-hour later. To initiate the- reaction, the mixturewas quickly heated and maintained at 96-100'C. (refluxtemperature). When viscosity reached 3.2 stokes(ASTM Method D 1545-63) the temperature was re-duced to 80'C. When viscosity reached 9.0 stokes thereaction was terminated by rapidly cooling the mixtureto 25'C.

ProcedureEvaluation of Glue Bond Quality

Glue bond quality was evaluated by wet shearstrength and percent of wood failure in specimens sub-jected to the vacuum-pressure cycle for exterior gluelines (PS-l-66), and by delamination (expressed as per-cent of glueline area) after 3 months of exterior

exposure.Veneers were of southern pine, lIS-inch thick.

One 12- by 12-inch three-ply panel was glued up foreach of the 72 batches of resins, and 20 standard shear

The author is Wood Scientist, USDA Forest Service,Southern Forest Experiment Station, Plneville, La. Thispaper was received for publication in November 1971.

Resin Preparation

All phenol formaldehyde resins were prepared in.the laboratory; formulation variables were as follows:

1) Molar ratio of sodium hydroxide to phenol(NaOH! phenol)-O.4, 0.7, and 1.0.

2) Molar ratio of formaldehyde to phenol (CH.O!phenol)-1.6, 1.9, 2.2, and 2.5.

3) Concentration of reaction mixture (percent ofsolids by weight)-37, 40, and 43.

Thus, 36 resins were formulated; as each was replicated,72 batches of resins were prepared.

To prepare each resin, all of the phenol, formalde-hyde, and water was placed in the reaction kettle. The

SEPTEMBER1M

Table 1 - SHEAR STRENGTH, WOOD FAILURE, AND DELAMINATION

RELATED TO RESIN FORMULATION VARIABLESfSpread: 75 pounds per 1,000 square feet ofdouble gluelineOosed assembly time: 20 minutesHot press temperature: 285.F.Hot press time: 6-1/2 minutesSpecific pressure: 175 psi

Percent

solid.content

Psi

0.4 mole NaOH/mole phenol37 1.6 26637 1.9 29037 2.2 25937 2.5 286

40 1.6 30640 1.9 29040 2.2 27940 2.5 267

43 1.6 30343 1.9 28943 2.2 26843 2.5 273

0.7 mole NoOH/mole phenol37 1.6 15437 1.9 15937 2.2 15037 2.5 176

40 1.6 24540 1.9 24840 2.2 19740 2.5 213

43 1.6 24643 1.9 25843 2.2 26443 2.5 238

1.0 mo4e NaOH/mole phenol37 1.6 25537 1.9 16837 2.2 15737 2.5 160

40 1.6 21640 1.9 22340 2.2 22040 2.5 252

43 1.6 24443 1.9 23343 2.2 26243 2.5 246

Results and Discussion0.10.70.35.0

0.11.00.11.0

0.60.00.20.8

78.427.467.787.0

57.254.034.255.0

20.829.983.088.5

39.314.134.046.0

28.727.517.228.0

10.715.041.644.7

76757971

60827578

69757664

Table 1 summarizes average wet shear strength,percentage of wood failure, and percentage of delamina-tion for each combination of resin formulation variables.By all three methods of evaluation, bond quality differedsignificantly (0.05 level) with change in NaOH/phenolratio and resin solids content. The change in CHsO Iphenol ratio was significant only in the delaminationtest.

30383027

50665148

61636654

10.316.97.01.6

1.420.815.60.9

3.40.02.30.1

16.997.599.597.4

96.286.789.095.8

66.947.485.397.0

13.657.253.349.5

48.853.852.348.4

35.223.743.848.6

50252821

50655855

62455857

17.22.44.9

27.7

6.66.0

10.645.4

0.42.98.20.1

98.498.272.5

100.0

99.0100.096.987.4

83.672.388.476.3

57.850.338.763.9

52.853.053.866.4

42.037.648.338.2

As shown in Table 2, a change in NaOHI phenolratio from 0.4 to 0.7 substantially decreased bond quali.ties. Shear strength dropped from 281 to 212 psi, woodfailure declined from 73 to 49 percent, and delaminationchanged from 54 to 81 percent for L/L gluelines. Withan increase in NaOH I phenol ratio from 0.7 to 1.0,changes in bond quality were not significant by Duncan's(1955) analysis. This result suggests that there roay bea critical concentration of sodium hydroxide beyondwhich no significant improvements can be obtained.

To examine this assumption, an attempt was madeto study resin polymerization under the reaction condi-tions of the study. It is generally accepted that thepolymerization involves three reactions (Fig. 1): addi.tion of formaldehyde to the phenolic molecule to formmethylol groups (reaction 1) and a condensation reac-tion between methylol groups of one molecule and thering hydrogen (reaction 2) or methylol groups of an-other phenolic molecule (reaction 3) with the forma.tion of methylene linkage.

As the reactions proceed, the formaldehyde con.centration and methylol content of the mixture change

IF mean. forma4dehyde; P means phenol; EIE mean. eartywood-to-earlywood, L/L means latewood-to-latewood bond..

.Values are overate. of 40 .amples.

aValue. are average. of 80 .ample..

- EFFECT OF RESIN FORMULATION VARIABLES ON SHEA.STRENGTH, WOOD FAILURE, AND DELAMINATION.

Table 2.

Shear Woodstrength failure

Variables

PII PerC8Wf - -Molar ratio of NaOH t..

phenol284550

281212219

734948

19

11

548189

0.40.71.0

Concentration of reactionmixture (percent of solids byweight)

374043

434436

206246260

466162

1092

767970

specimens were cut from ead1 panel in such a manner thathalf could be pulled open and half closed. For exposurespecimens, five two-ply cross-laminated panels wereglued up for each resin and sawn into 0.5-inch squaresto yield 20 earlywood-to-earlywood (E/E) exposurespecimens and 20 latewood-to-latewood (L/L)specimens.

Details of sample preparation and testing have beendescribed previously (Hse 1968, 1971).

General gluing conditions were:Glue mix: 26 percent resin solids in the mixed

glue'EIE means eorlywood-t_rlywood, L/L means Io"wood -to-lat.-wood bonds.

lOfiFOREST PRODUCTS JOURNAL Vol. 22. No.9

Molar!

ratioF/P

mole percent of caustic has no effect on the rate ofcondensation. Furthermore, viscosity measurements(Fig. 2B) indicated that the higher caustic content re-tarded the condensation reaction. The decrease in for-maldehyde concentration is, therefore, largely attributedto the Cannizzaro reaction, i.e., the Cannizzaro reactioncompeted with the addition reaction for the formalde-hyde. Figure 2B indicates that, at NaOH/phenol ratiosof 0.7 and 1.0, the formaldehyde deficiency may Mvedeterred addition of methylol groups sufficiently toprolong the condensation reaction. The prolonged re-action of the resin may result in declining chemical re-activity and, in turn, poor bonds. This explanation is

OH

(1) 0 + CH20

OM OH

~ 0 +¥+C¥CH2

Summary of chemical reactions between phenol andformaldehyde under alkaline condition.

Figure 1

according to the reaction rates. Viscosity of the mixturerises with increases in the length of the molecular chainresulting from the condensation reaction. For thesereasons, formaldehyde concentration, methylol content,and viscosity were expressed as a function of reactiontime. Formaldehyde concentration was expressed by amethod involving titration with hydroxylamine-hydro-chloride (Walker 1964, pp. 493-494), viscosity wasmeasured by Brookfield viscometer, and methylol con-tent was estimated by a method previously described

(Hse 1972a).The sodium hydroxide catalyst caused a very rapid

addition of methylol groups; after 30 minutes of thereaction, the methylol content (as measured by absorb-ance ratio-see Fig. 2A) attained its highest level, andmore than 95 percent of the formalddlyde was con-sumed. Significant differences in the formaldehyde con-centration were detected at the end of the second sodiumhydroxide addition (i.e., at 60 minutes' reaction time): theconcentration increased at an NaOHI phenol ratio of0.4 and decreased at NaOHI phenol ratios of 0.7 and1.0. This difference seemed important and was studiedfurther.

The increase in formaldehyde concentration maysignify condensation reactions between two methylolgroups of two phenol alcohols; studies have shown thatreplacement of methylol groups occurs more rapidlythan that of ring hydrogen under alkaline conditions(Ziegler and Zigeuner 1948; Freeman and Lewis 1954).The decrease in formaldehyde concentration at NaOHIphenol ratios of 0.7 and 1.0 may result from addition ofmethylol groups, but more probably is caused by con-version of formaldehyde to methanol and formic acidby the Cannizzaro reaction.

Because methylol content-measured as absorbanceratio 1010/121o-decreased at approximately the samerate for all NaOH/ phenol ratios between 30 to 60minutes of the reaction, the ratio of methylol groupformation (by addition reaction) to consumption (bycondensation reaction) likely remained the same. If thedecreasing formaldehyde concentrations at NaOH/phenol ratios of 0.7 and 1.0 indicated a higher rate ofaddition of methylol groups, then consumption mustalso have been higher. Faster consumption, however,is not likely, for Granger (1937) has shown that an in-crease in amount of catalyst above approximately 0.05

SEPTEMBER 19721()6

OM

0CH~OM OH

Figure 3. - Relationship between delamination and CH20/phenol ratio for latewood/lotewood (L/L) and earlywood/earlywoodIE/E) bonds.

ratios of 007 and 100 substantial gains in bonding accom-panied increases in solids content, the sole exceptionbeing the latewood-to-latewood gluelines at 37 percentsolids content. These results are graphed in Figure 4;since performance at ratio 0.7 did not differ from thatat ratio 1.0, data for the two ratios are combined in thefigure.

The L/L delamination was substantially higherthan the E/E: 75 percent as compared to 7 percent(Table 1). Of the 36 resins, one with an L/L delamina-tion of 17 percent was obtained. This was significantlybetter than the average of more than 80 percent ob-tained at 15 minutes' assembly time in a previous study(Hse 1968).

On the basis of wood failure, two batches of resinsamong the 72 met the standard of vacuum-pressure soakcycle for exterior glueline (PS-I-66). Shear strengthand wood failure were positively correlated, as shownby regression analysis (Fig. 5). Delamination values,however, were correlated with neither shear strength nor

supported by the fact that the resin produced withhigher NaOH/phenol ratio had lower methylol content(Fig. lA) and longer cure time (Hse 1971).

An important aspect of the Cannizzaro reaction isits relationship to caustic level. After the secondcaustic addition, the close grouping of formaldehydeconcentration for NaOH/phenol ratios of 0.7 and 1.0seems to indicate that the Cannizzaro reaction becomessignificant when more than half a mole of sodium hy-droxide is present in the reaction mixture (i.e., causticcontent is 0.5 mole for NaOH/phenol ratio of 0.7 at theend of the second caustic addition). However, the criti-cal concentration of caustic may be any point above0.4 mole.

On the average, glue bonds were bast with 43 per-cent resin solids and poorest with 37 percent resin solids(Table 2). It is generally recognized that resins for-mulated with lower solids content have higher molecularweights, if all other conditions are equal. The high-weight resins, with their fast cure, are usually desirablefor southern pine plywood, but they create problems ofglueline dry-out. The present data therefore indicatethat the effect of dry-out on bond quality is more im-portant than that of cure rate. Ripley (1967) hasestimated that glueline dry-out causes approximately 8;percent of the poor bonds in southern pine plywood.

The effects of CHtO/phenol ratio on del3minationare plotted in Figure 3. The resins formulated withratios of 1.6 and 1.9 yielded stronger glues than wereobtained from ratios of 2.2 and 2.5. This result is inagreement with industrial experience. Booth (1958)reported that firms producing phenol-formaldehyderesins have in general favored a ratio of 2, since addi-tional formaldehyde will not react effectively.

Although the resin with a ratio of 1.6 performed aswell as that with a ratio of 1.9, it tended to yield aproduct having a longer cure time (Hse 1971).

Resin solids content interacted with NaOHI phenolratio to affect bond quality. At a ratio of 0.4, theeffect of sodium hydroxide predominated and bondsvaried only slightly as solids content increased. But at

1001

(~ '~O NaOH

~/L ~;;80;~..

-60Z0~C40ZiC~~ 20a

"' :~.Fleure 4. - Interactions of wetshear strength IAI. wood failureIBI. and delamination 1'1 withNaOH/phenol ratio and salldscontent.

""

.,.- ~w~

OA~

37 40 43RESIN SOLIDS CONTENT

(PERCENT)

37 40 43RESIN SOLIDS CONTENT

(PERCENT)

37 40 43RESIN SOLIDS CONnNT

(PERCENT)

107FOREST PRODUCTS JOURNAL Vol. 22, No. .

of various methods of adding sodium hydroxide is inprogress and may provide a basis for optimizing theratio.

WOOD FAILURE (PERCENT)

- Wood failure as related to wet shear strength.Figure 5.

wood failure. Probably extreme wood vari4tion be-tween 111 and E/E specimens obscured significantrelationships.

Of the three formulation variables, the NaOHIphenol ratio gave the most surprising results. The ratioof 0.4 is considerably lower than any used in industry,but resins thus catalyzed gave consistently better bondsthan those with higher proportions of caustic. Thisresult seems to indicate the importance of controllingthe Cannizzaro reaction in resin formulation. A study

Literature Cited

BOOTH, C. C. 1965. Adhesive mix formulation for southernpine plywood. Paper presented at Southern Pine Seminar,The Borden Chern. Co., Meridian, Miss., Jan. 12-13.

DUNCAN, D. B. 1955. Multiple range and multiple F tests.Biometric 11(1):1-42.

FREEMAN, J. H., and C. W. LEWIS. 1954. Alkaline-catalyzedreaction of formaldehyde and the methylols of phenol.J. Amer. Chem. Soc. 76:2080-2087.

GRANGER, F. S. 1937. Condensation of phenols with for-maldehyde. 11. Resinification of phenol alcohols. Ind.Eng. Chem. 29:860-866.

HSE, C.- Y. 1968. GluabilitY of southern pine earlywoodand latewood. Forest Prod. J. 18(12):32-36.

. 1971. Properties of phenolic adhesives as relatedto bond qualitY of southern pine plywood. Forest Prod.

J.21(1):44-52.. 1972a. Surface tension of phenol formaldehyde

wood adhesive. Hoizforschung (in press).. 1972b. Wettability of southern pine veneer by

phenol fornlaldehyde wood adhesives. Forest Prod. J.22(1):51-55.

RIPLEY, R. H. 1967. Gluing faults and causes. In A summaryof papers presented at meetings of southern pine pro-ducers in January 1967, pp. 5-6. Amer. Ply. Assoc.

WALKER, J. F. 1964. Formaldehyde, 3rd Ed. 510 pp. Rein-hold Publishing Corp., N.Y.

ZIEGLER, E., and G. ZIGEUNER. 1948. Phenol alcohols. V.Behavior toward diazonium compounds. Monatsh. 79:42-54.

SEPTEMBER 1972108