~ lend ori&iD81 m8Duscr1pts (four copies) to:

EAitorial Office HolzforschungWalter de Gruytcr GmbH &:. Co. KGPostf8Cb 30 34 21D- I 0728 Berlin I F.R. of GennanyFax: ++ 3O-2600S-32.5/TcI. ++ Jo-2~-219e-mail: Holzfmlcl.1lll.editorial. deGruyter .de

tables and figures. Short DOtes do n« ~ a swnmary and will have

priority in publication. .;-

IntJ'oduetioa. This should define die poblem and if possible die frame

of existing knowledge. PIe8Ie ~ ,tb8I people n« wcxkina in your

particular field are able to uncIeraand yoor iDteotioos.

Materials and Methods. Please be u precise u possible to enable

odIer scientists to repeat your ~

Genen1 R-.hL Only material peI1i_t to die 1Ubp:t must be ilM:llMIed. Data

. . must n« be repeated in figures and tables.

Only unpublished original papen of expenmental content. corre- . .. .sponding to die scientific requirement of HOLZPORSCHUNG. are Disc-'OD and. Con.eI~ should .mterpret die raults ID VI~ of die

KCepted fcx consideration. Papers submjtted to symposia fcx publica- problem u outlined In ~ I.n~ and of ~ obIe;n'a.tions by

tion in a proceedings-volwne are to be considered u pre-published. you~f IX" odIerS. I~oos fcx fuI'dIa' .studies cx app~~ may

Theref~. it is not possible to accept these contributions unless they be discussed. A conclusion should be Idded 1f results and d1scusSion are

differ and are enriched by new findings and results. With die ~e combined.

of the manuscript fcx publication. die copyrip is automatically AckBOWled.-ta may be used to aedit ~

assigned to the publisben and the author may n« plblish the same . .. .paper elsewhere not even in a foreign country. Refere_. Relevant Uterature. m~ be C1~ .In the ~t :WIth the

. author's name and year of pubhcanon. In addition, ~ blbhograpbyOnly papers written in the English language will be accepted fcx must be listed at die end of the text in alphabetical order of the authors'

publication. names. togedIer with the title of the paper and die fun quotaIion of the

To elimlDate u_ry, time-coDSWDiDg corl'e5poOOeoce, the bibliographical ref~, fcx example:

authon are requsted to submit only manuscripts which are ready

for printing. TbJs includes draftina and orpnlzing of the text in JollmaLs:

accordance ,nth the rules adopted by HOLZFORSCH~G for Ralph J.. S. Hatfield. S. Quidcau. R.F. Helm. J.H. Grabber and H.-J.

typesetting. AU manuscripts are subject to a thorough reVIew by Jung. 1994. Pathway of p-coumaric Kid illCOfP(lratiOl1 into maiu lignin

selected referees; the authon themselves are Invited to recommend as revealed by NMR. J. Am. Chem. Soc. 116, 9448-94S6.

competent experts for reviewing. - - - ~--~ SUDJD1WDl a manuscript. dif iofiowtuc ruls .lIuuld bf.-d 8m",:~---: - --

carefuDy. Manuscripts that differ from the speciftcatiOJ15 and do not Pan!'hin. AJ. and C. de Zeeuw. 1980. TexttMJOk of Wood Technology.

comply with our formal requirements wiD be returned to the McGraw.Hill. New York. pp. 9448-~~9.

autbon for correctiou before review. Please use a recent Issue of

this journal as a guideline In the preparation of your manuscript. M.,lti-tUIIlto, booAs:

For further Information you can visit our homepage www.d~ Lai. Y.Z. and K. V. Sarkanen. 1971. IIOlation and ~ studies. III:

Gruyter.de. Occurrence. Fonllution. SInICtwe and function. Eda. K. V. Sarkanen,

C.H. Ludwig. Wiley IntencielK:e. New York. W. 150-163.

After the references full JIOItal aIkIres8eS of all aIIdKJn Ihoold be

mentioned.

The numbers of FI&1II'eI must be limited to the absolute r.' Micropbotograpbl must be ~ and rich in CCXlIrast. Dnwillll

must be suitable fcx reproduction (no ~) and reduction to a

width of 8.1 cID. Do not use blnck 1etten in the lettering of fiJures and

drawings. The legends fIX" the figures must be concise and self-expla-

natory 10 that they are intelligible even without refere",-"C to the texL

The key to the symbols in die figures Ihoold be included in the legelM1,

-- not fonn part of the figure itlelf, Each figure muat be plVaGnted on a ,~- - -~gle shee~usinB conseelltt\le-8fabic-numet'8ls~latt.-app'Ies~'::'~

Tabla with concise and self-explanatory heading!'. ll1C teaenda of aUficgre8 should be listed on one separate ~.

Dimensions and units. The metric sy!'tem mull' I'C ulied. SI units arerecommended. Compound units are itiven with 1M proper cApI)nent

0 I d h-1Wtu t a point, e. a. g 2'- W .

Only those manuscripts which confcxm t(l Ihr alMWC' ,uidelillel will beconsidered for publication. The EditlM' reoCrw~ ttlc riJht to suggestabbreviations and text improvement".Gane, pr-. are forwarded.o the .uthln in dUplic.te. The immediatereturn of one corrected proof to lhe "Ait.. i. ~uc.ccd. In view of thehigh costs. correction!' mu!'1 I'C limit~" hI ptintin, errors; amendmentsto or changes in the lex' which exceed 4 aill,le I:llITCCtions. or the cor-rection of a sin~le KenleOCe. will M l:\I8rJed.

20 reprints of each 'lI"i,inal pa~r will M rllrW.nk-d to the author freeof ch8Jie. If funhcr rr:!Jfinl~ .~ reqlll~1. lhe.e can he ordered whenreturning Ihc ~"lI"re~'lcd I",.of 11M! lklihlr. 1~ Ipproximate price perreprint i!' given.tn Ilk' 'lfdrcl hlfnl "1..~n)'ln. ttlc ,alley proofs.

SpedftC8d08 for MaD1*ripUSubmitted manuscripts must be both concise and precise, and consisl ofDO - than 8-10 single-sided typewritten sheets of 30 double-spocedlines. The length of the IeXt should be in keeping with the resultsachieved. Contributions exceeding the standard length of 10 sheets (001including summary, tables, fig~ and references) by more than 50%may IS a rule Dot be accepted for publication. It is !herefore advisableto divide longer manuscriptS into tWO IX" - continuing series. Papersof _Iban IS p8&e5 may ooIy be submitted after prilX" consent of the

editor.-- ",--"~~--~" --- -~--" ~ ".-.

DiIk.a..P1eaae ~ a1ao an elecuonic file. Please submit your IeXt 01\diskettes fonnatted for Windows '9S. Indicate the operating system andthe programme Dame on !he disk label.Order secbOOS ICCording to the format of the journal.UK linale tab stops to indenLUse single carriage returnl between paragraphs.Save araphics and tables in separate files.Avoid using multiple 5p8CC characters to format IX" ~ 1eOIeDCes.Avoid the positioning of graphics and tables in the IeXLAvoid automatic numbering of pages aDd footnotes, manual pagebreBes.In short: the - simply you format your manuscri~ the easier it willbe to refomlat it fIX" typesening with a minimum of emIR!

Composition of papersAvoiding a lengthy introduction. papers -- report of tnIe IcieDtificor tedIak8I p~ the major dIapters being cbaracterized byappropriate headings. Paragraphs of less importanCe should be markedfor setting in 'pedt'. c

The tint page of the type-scri~ should pesent: running title; full title;name(s) of the autbor(s); name aDd addras of the laboratory where lhcwCX"k was carried out; Summary (awrox. 1/2 &beet DIN A4) contain-ing project, procedure, results and conclusing (concise in abstraci form)and Keywords.For the benefit of the author and the reader it is advisable to publi"hclear-cut resear\:h results in the form of Short No.. (but no provisi()IUIIor intennediate reports), consisting of an extended abstract of 001 mllr~than four double-spICed typewritten pages, in addition to the nece"!ia~

SulMlt-rlplklfl Inlorn.alktnHOL.7.F<)RSCIIUNfi i. publi~ bi-nMJIIdIly (6 i- per yar).Price of Yt)lumc ~~ 12IMII): l)M I.~.-. t~ 868.17; Iff l,sll.-;tiS 12.3'J~.-USA oml Co'~: lIS S I~).~Sin~1e i~: I)M 292.-; f:: 149.JO: 5" ~I.-: as 2,132.-lJSA and Cllnlklll: 1IJ1t1rt)A. US S IMI.-

Urdcr,; fUf IIUh!it:fipti, nlll)' Ite paM-oed wllh any tK1okIeller or with the

publistk"rlo.

R. Snell el al.: Surface Morphology of TMP Fibers mHi5!:i(~~~'i ;,,;,~55 (2001) 5".-520 ;

Characterizing the SUrlace Roughness of ThermomechanicalPulp Fibers with Atomic Force Microscopyl)By Rebecca SneUl, Leslie H. Groom2 and TimothyG. Rials21 The BioComposite Centre, University of Wales, Bangor, Wales, UK2 U.S. Forest Service, Southern Research Station, Pineville, LA, U.S.A.

Keywords Summary

Atomic force micn)!icoPY Loblolly pine, separated into Ii1ature and juvenile portions, was refined at various pressures (4, 8 and 12(AFM) bar). Fiber surfaces were investigated using a Scanning Electron Microscope (SEM) and an Atomi.(: F~

Scanning elcclnm microscopy Microsco~ (AFM). Refiner pressure had a significant effect on the fiber surfaces. SEM images ~owed(SEM) an apparent increase in surface roughness with increased refiner pressure. This was shown quantitative-

ThemMJmcchllnieul pulp (TMP) 1y with data from the AFM that was analyzed using 5, 2.5 and 1.25 ~ scan sizes. A scan size of 2.5 ~PitX'r .urfllcc loorphology was found to be the most informative in terms of quantifying the effect of the different treatments on theSurf~t ruuBMcss two fiber types. The calculated surface roughness was greatest at 8 bar for both wood types. JuvenileItctinct pn.'55Un: fibers in general had higher surface roughness values than mature fibers. The results suggest that refin-l~ftlcllIl~ ing pressure may influence the failure mechanism of juvenile and mature wood differently.

charged at atmospheric pressure into a cyclone where thesteam is removed. or blown through a blow pipe for minimalremoval of steam from the system (5uchland and Woodson1986). Changes in temperature, moisture and pressure, inconjunction with mechanical abrasion, affect the manner inwhich fibers become separated at the cellular level. This inturn may effect the surface structure of the fibers generated

during refining.Hypotheses have been developed to address what occurs

during breakdown of wood chips during the refining pr0-cess. Pearson (1983) believed that compression and expan-sionbetween the bars of the refiner plates together with

..exGCSS:.,:steam-'.cause--cleav.~.c.S2 :, Jay.er~- SI) nt:L!!i)lm(1993) proposed several steps including separation of thecompound middle lamella to produce fines, followed bydelamination and fibrillation of the fiber wall. Kamis(1993) however, suggested that the primary and 51 layerswere peeled away exposing the 52 layer.

Goring (1971) showed that lignin becomes softened atelevated temperatures. Therefore, at high temperatures themiddle lamella becomes a zone of weakness allowing foreasy cleavage of adjacent fibers. Goring (1971) found thatthe temperature at which the plasticization of lignin oc-curred was determined by water content, tl1e transition tem-perature being reduced with increased water content. Thiseffect can be seen during refining where there is a reductionof power input required to fiberize wood chips at tempera-tures greater than 150°C. As such, refining at varying tem-peratures and pressures may not only affect how fibersbecome separated, and which cell layer is the primary sur-

Introduction

Wood composite materials are fonned from fibers that havebeen generated either chemically or mechanically and arethen brought back together to fonD a complex matrix. Matrixconstruction may be achieved by hydrogen bonding eitherbetween fibers as, for example. in paper production, orbetween the fibers and resin as occurs in the fonnation ofmedium density fiberboard. Due to the reliance on hydrogenbonding and secondary interactions, fiber surface chemistryand morphology is considered to have a direct effect on thequality of the bonding mechanism between adjacent fibersin the matriX and this ultimatelydeterri1ines~ucrper-

---"C ..~~ -,--~~"CC-

fonnance (Nissan and Stemstein 1964; Page 1990).The analysis and design of wood fiber-based composites

is complex due to the fact that the primary component isitself a bio-based composite. Wood cells are comprised of athin primary wall and a three-layer secondary wall differingin thickness and varying in the chemical constituents cellu-lose, hemicellulose and lignin (Panshin and deZeeuw 1980).The cells are bound together by a lignin-rich middle lamella.Cellulose is present in the form of helically wound micro-fibrils encased in a matrix of lignin and coupled with hemi-cellulose. The angle and orientation of the microfibrils variesdepending on the position in the cell wall {Harada and COte,1985; Abe et al. 1991; Megraw 1999).

There is much interest in the mechanisms involved infiber separation during the mechanical breakdown of woodin the refining process. During the production of thenno-mechanical pulp (TMP), wood chips are subjected to extremeenvironmental conditions. The wood chips are softened atIIn elevated temperature and pressure in a saturated steamchamber, then forced at high pressure though a rapidly,~)j"ljng disk to become defiberized. The pulp is then dis-

I) This anicle waswrinen and prepared by a United States employ-

ee, on official time, and is tbereforein the public domain and notsubject to copyright.

Ili.ltti.r~hung 1 Vol. 55/.20011 No.5..~~~

2 4 6 8 10 12 14

Refiner Pressure (bars)

Fig. I. Energy consumption during refining at 4,8, and 12 bar.

Juvenile macerated Juvenile 4-bar

Juvenile 8-bar Juvenile 1 2-bart... - ~ltM """fl'l'f""'" III JI,vrllilt' tl~f" ,,1"lwing ~ufface morphology after maceration and pressure refined at 4,8, and 12 bar.

~~~!~g-;.-.: f\\d,.,'\ f 1JKi.\ IN

513

~ Detailed investigations of fiber ~urfuces are now po~~ible

using the atomic force microsco~ (AFM). The principlesof the AFM are covered elscwhere in the literature (Beland1996) and will not be detailed in thi~ paper. The AFMenables surface topography to be recorded in three dimen-sions by detecting the vertical deflection of an oscillatingtip that rasters across the sample surface. Quantitativesurface measuren\ents such as roughness can be calculatedfrom height data, and direct comparisons can be made be-tween samples. The AFM has been used to study surfacetopography of different wood fibers (Hanley and Gray1994; BOds and Gatenholm 1999). However ,little work hasmade use of quantitative data that is recorded during imagecapture. Pesacreta (1997) has used this technique exten-sively to quantify the effect of chemical treatments on cellwall roughness in cotton. In this work we study the effect ofdifferent refining treatments on fiber surfaces qualitativelywith an SEM and quantitatively with an AFM.

-~tined att~ ~'iure.'i (4. 8. and 12 bar) in a ~. pres-liurized. liinglc-disc n:liner al the BioComposites Centre. Univer':!lily of WaleN. A 30 cm plate was used with fiber retention timeof approximately 4 minutC!i. The plate gaps were set at SO. 250.and 50 ~m for the 4. 8. and 12 bar ~ures. Fibers were dried inII na!\h drier 80 melerli in length to a moisture content of approx-imately 12 pen:ent. The surface morpbology of the treated tiherliwas investigated using a JEOL scanning electron micffi!iC~(SEM) and a Nanoscope ilia AFM (DigitallnstnJments. Santa Bnr-bara. Calif. USA) mounted on a pneumatic isolation Ilible wilh unacoustic l1oOO. Fibers were mounted on stubs and sputter 1:(l;Ilc(1with ~ximately 15-nanometer thick layer of gold and OOo'iCrvcd

at 15 to 20 KV at a distance of 25 to 33 rom.Fiber preparation for the AFM was minimal. Individual libcr!l

were carefully attached to AFM stubs using carbon ta~. Thrc(.'5 ~m scans. located in the middle and qualterpoints of 10 individ-ual fibers. were collected for each treatment. Unrefined liberli wcrl:prepared by maceration in a solution of acetic acidnlydrogcn J1I.'r-ox,jde for a period of awroximately 48 hours. dlus pro~din~ itbaseline for comperison. Fibers were orientated with the I(mg axi~parallel to the raster scan direction. All AFMmicrographs wen: luk.:nat a resolution of 5 12 X 512 pixels.lrna&eI were obcained in inlcr.mittent-contKt mode. also refened to as Tappina mode ". in uir.The probes were standard tapping-mode tips comprised of eIchcdsilicon. The probes had a nominal tip radius of curvature of 5 1(120 nanometers and a ~t frequency of about 300 kHz. A .~:ln

Materials and Methods

Loblolly pine chips, separated into mature (growth rings 2S andbeyond) and juvenile (pith to growth ring 10) portions. were

Mature macerated Mature 4-bar

, Mature 1 2-barMature 8-barFig. 3. SEM micrographs of mature fibeR showing surface morphology afler m;\CcrOiliu" and pressure refined at 4,8, and 12 bar.

Holzfonchung / Vol. 55 /2001/ No.5

c --

rate ofl Hz was employed to generale a lotal of 240 scans. Threedata channels. heighl. amplitude and pha.'ie shifl were moniloredduring image acquisilion. From each image a repre.~ntative areaof 2.5 by 2.5 11m and 1.25 by 1.25 11m were selected using thezoom feature and used for further analysis. A flatten order of I wasapplied to the height data. thus correcting for any 2-dimensionaldeviations about a flat plane while maintaining the integrity of SEM observations .. ,individual features. The images were evaluated and statisticallyanalyzed using the heighl data to quantifY surface roughness SEM images show that, in general, fibers refined at 4 bar .(RMS). total power (TP) and fractal dimensions (FRAC). have a much smoother surface than those refined at 12 bar. .'

, .The 12 bar fibers have surfaces that are fragmented, a result

R ults d D' , of intercellular delaminations. This is more noticeable in thees an ISCUSSlon .juvem1e than the mature fibers, There appears to be a pro-

Fiber throughput for the various fiber types was approxi- gressive increase in the surface roughness with increasedmately 25, 35, and 40 kg/hour for the 4, 8, and 12 bar sel- refiner pressure; juvenile fibers are shown in Figure 2 with .

tings, respectively. The various refining levels resulted in mature fibers represented in Figure 3. The surface of thelarge shifts in both energy consumption during refining and juvenile fibers refined at 4 bar looks to be much smootherresulting fiber appearance. Figure I shows the energy con- than the mature fibers. Delamination is evident in somesumption during refining drops by 80 percent from 4 bar places on the fiber surface for both maturity fiber types .

to 8 bar. This drop is indicative of the glass transition tem- refined at 8 bar. Surface layers have peeled and flaked off,perature of lignin residing in this range. Increasing the pres- as described by Karnis (1993). Surfaces for both juvenilesure to 12 bar increased refiner energy slightly from 0.13 to and mature fibers refined at 12 bar' look extremely rough,0.16 kWh/kg. with large globular deposits and a broken crust-like appear-

Visual differences between the different pressure treat- ance. The macerated fibers have an extremely clean surface,ments were immediately observed. Th~ was a distinct interrupted mainly by natural features such as pits. It can be --~

-~-

pressure. Fibers refined at 4 bar were pale tan while 12 barfibers were a deep rusty brown. There were no obvious dif-ferences between mature and juvenile fibers refined at thesame pressure. ,

Jw~nile mocerated Juvenile 4-bar

Juvenile 8-bar Juvenile 12-barFig. 4. Five-micron AFM scans of juvenile fibers showing surface morphology of macerated and pressure refined fibers at 4,8. and 12 bar.

Hnl7f~h"nlJ I Vnl ~~ 11M! I Nn ~

R. Sncll ('/ (t/.: SurlilccMllrphllhl~Y I!fTMP l"i1ll.:r" S1~

~~-- !;urfucc morphology vuriej; greutly .llong the libcr uxij; .Ind

between libcn;.

--- --- '- c~ -(m~v~'(\ could be Jignin lhal had been softened8rid re- ---,

c. . , "dcp()I'i.edas lhe glass lrdnsitionlemperature of native lignin .(148..9°)is between 4 bur and 8 bar.'The microfibrillarstruc-

AFM b . lure may'also be obscured by low molecular weight extrac- '"0 servatlons . h h . ed f . .

IIt.vest ;It ave mlgrat rom the mtenorwa ..The AFM images for juvenile and mature fibers are shown Figure 6 shows how the 2..5 an,<! 1..25 J1fil scans werein Figures 4 and 5. res~tively. Although the AFM images obtained using the zoom function on the AFM.. Statisticalwere difficult to interpret visually at such!! high n1agnifi. calculations using numbers obtained from off-line analysiscation. an overall trend could be seen.. Macerated fiber sur- using height. data, showed that there were differences in "faces looked extremely different from the refined fibers.. the surface roughness (RMS). total power (TP) and fractal ..

Microfibrils were evident over the entire surface for both dimensions (FRAC)..the juvenile and mature macerated fibers.. The microfibrils RMS is the root mean squared of standard deviation ofwere oriented in a rather random pattern indicating that the the height (z) data and gives an overall indication of surfaceprimary cell wall layer had been exposed.. The microfibrii roughness.. TP is roughness amplitude squared and definesangle for the fibers refined at 4 bar was shown to be approx- which surface features have the greatest influence on theimately 30 degrees to the fiber axis suggesting that the S2 surface topography.. FRAC is the log of the dimensions of

": layer was predominant on the fiber surface.. This is more the entire surface of each scan versus the scan size; it (JefineS ,.'.f, evident for the juvenile fibers than the ma~Microfibril the total surface area, the l~er the fractal number the more

angle was difficult to see for the 8 and i2 bar pressure treat- complex the surface.ments.. The surfaces at these pressures appeared tobecov- Figures 7 and 8 show the RMS value frequency distribu-ered with a smooth layer that had pQssibly been either tionof the juvenile and mature fibers. res~tively. for thesmeared or re-deposited during refining at the higher pres- different refiner pressures a( a scan size of 5 J1fil scans.sures.. The temperatures corresponding to the refining pres- There area wide range ofRMS values for fibers within thesures from steam tables are 144 °C. 170°C and 1~8°C for sameb'eabnenLEvenforthe same fiber.RMS values were --

. 4.8. and 12 bar pressures (Weast. 1975).. e sm ayer ou to vary conSl era y. c=-

Mature 4-oorMature~erated

;E~fi,;c co""

Mature 8-bar Mature 12-barFig. S. Five-micronAFM SC8nsof mature fibers showing surface morphology of macerated and pressure refined fibers at 4,8, and 12 bar.

J.I"I..f"rcrh""o I ~I ~~ I 700 I 'No.5

$1.6 R. SI1~ll /.'J III.: S~d~~MI)rcR~~ ~!rI~~J?J:l!.1l:r,;

Horizontal distance (microns)

(0)

,

-Horizontal distance (microns)

(~

~

0

~ 0~;EClI'u:I:

.2,00"~~ 0..5 1

HooiDiJfal~ (mi'aiiiiiJ

.1;$

(C)

Fig. 6. Surface plot ofjuvenilc tihcrr\:li.nctJ al II bar !ih(IWing (a) 5 (b}2.5 and (c) 1.25 IJJn scan size (zoom location shown by white box),with corresponding surfa~-e line Ir.11:t: (j'kl!iililln !ihuwnby diagonal white line).

~,

.20

>.g ISU~

[~ 10

.30 40 so 60 70 IIJ(),"'.

RMS Roughness (nm:

..g. 7. I;rcqucncy dislribulion of juvenile fiber roughness macerated and refined al 4. 8. ,llItl 12 I~

t( Sltt:11 ,./ttl: Surf'act: Morphology ofTMP Fi~rs Sf7

~ -- st:M im;l~c~ ~howf:d that the surface is not homogeneous.

"111(" vilrilllillll i~ reflected in the standard deviation of themC1l1i RMS ill'"uble 1. Analysis of variance of the full RMSdlll;1 ~ct ~howed that refining pressure does have a signifi-~'lIlIt ("11'("1."1 on the surface roughness (p = 0.0001), this is

III~, I 11l1'cl."ted by the maturity of the fibers (p = 0.0499).t.l!iing Fisher's LSD tests run on the analysis of variance

I IIIIIJVU). each parameter RMS, TP and FRAC prOduce val-\ll'~ Ihut enable differentiation between the different treat-mcllts to a certain extent (Table 2). RMS values prove to belIte most useful. Fibers refined at 4 and 8 bar have signifi-

-- ---~--

cantly different surface roughnes:;;tOeaChother for the 5-pmscans. This is reflected in the frequency data in the respec-'tive juvenile and mature fiber!' in Figure!' 7 and 8 where the4 bar fibers have a greater frequency of !'mall RMS valueswhile 8 bar fibers have a larger proportion of high RMSvalues. This trend i:; also found when comparing 4 bar with12 bar treatments for juvenile but not mature fibers. RMSvalues for 8 bar fibers are not significantly different to 12 barfibers for either juvenile or mature fibers. This again can beseen from the frequency distribution curves in Figures 7 and8. Evaluation of the data using anova for the defined 2.5 ~m

2S

20

>-5 15

6-~ 10

~

0Ie iliO" 10 M ~80 --- ~

RMS Roughness (nm1

Fig. 8. Frequency distribution of mature fiber roughness macerated and refined at 4, 8, and 12 bar.

.-~8~~--v()

§...tI)

.-"0

-;()

.:5 250;>:

-250

0 2 4 ,6

Horizontal distance (microns)

Fig. 9. Surface line traces for juvenile fibers shown in Figure 4 taken perpendicular to the microfibril angle~a~1fllx (ICln II! libcr axiJi).

I hllzl"'r~,hlllll! !V"'. ~~J 21M)11 No.5

518- R. Snell eciaL: Surface Morphology ofTMP Fibers

.-{/j~Q,)

"Q)

e

~c--Q,)

~"in.-"0-~Q

"2

~

250

:2SD

0 :2 4 6

Horizontal distance (microns)Fig. 10. Surface line traces for ma~fibef$ shown.inFigureSlaken perpendicular 10 Ihe microfibril angle (approx 60 0 to fiber axis).

Table 1. Mean and standard deVia~_J)(sf4)OtRMS yaJ.q~~for5,2.5and 1.2Sllm~~clln5 fur juvcnilc: (J) and mature <M)fi~~(mac) or refined at 4, 8,or 12 bar, ~irtyimages(3 scans.fromrOti~) were taken ftlr el\ch treotment- -~-~ ~-~-~---

J-mac J-4 bar J-8b8r J-12bar M-mllt: M-4 bur M.8 bar I- c"'::: -d21;O =!:J6A ~ Ico! .~; ~~c~~"--- :.: -~ ~

~.12 bar

17--1~~.4

23.r~

J3;L~

5 IJInscan mean-std- ~

~.I~

68.7~

II.JW

2.5 IIIn scan meanstd

12;412.4

28..910.5

55.!22..4

42.8

~O.S.13.712.8

27.913.3

38.722.8

38.320.4

1.25 ~m scan meanstd

711.2

12555

22.910.3

18.78.2

6.3

5.0

14.08..

18.3IO.S

16.79.5

region of the 5 I.IIn scan show si~ificant differences for allthe treatments (p= 0.0001) and fiber maturity (p= 0.0106).Fisher's LSD tests (Table 2) confirmed that all treatmentsfor both juvenile and mature fibers were si~ificantly dif-ferent except for mature fibers refined at 8 or 12 bar. Thedifferences for fiber maturitY a~ascan size of II.IIn can nolonger be distinguished (p = 0..1957). However, the effectsof the treatments are still si~ificant (p =0.0001). Table 2shows that although the effect on juvenile fibers is signifi-cant for the 4 and 8 bar treatments, no differences wereapparent for mature fibers. The results suggest that 2.5 I.IInscans reveal the most useful information enabling differen-tiation between the refining treatments.

Total power and fractals were Dot such useful parame-ters for analyzing the data. Total power was more effectivein distinguishing differences between the juvenile fibersthan the mature fibers especially for the smaller scan size(Table 2). The surface features of the juvenile fibers for alltreatments were distinctly different at 2.5 ~ and 1.25 I.IDl.whereas DO statistical differences were found for the maturefibers. Fractals showed differences only for juvenile fibersbetween the 4 and 8 bar pressure treatments for all the scansizes. -

Surface profiles were examined for each of the AFM'images in Figures 4 and 5 by conducting a surface line traceperpendicular to the microfibril angle, approximately 60

R. Snell et at: Surface MorphologyofTMP Fibers 519

~ - - -~~-~

Table 1. Treatment comparisons using Fisher's LSD test run on analysis of variance of surface roughness (RMS), total flOWer (TP) an"fractal (FRAC) data generated from the 5, 2.5 and 1.25 ~m scans. Mature and juvenile fibers were analyzed separately. Asterisks signifytreatments that have significant p-values (p < 0.05). NS = not significant

- --RMS TP FRAC

5~m4 bar

4 bar

Sbar

8 bar12 bar12 bar

..NS

*

NS

NS

.

.

NS

.NSNS

.NSNS

NSNSNS

2.5~4 bar

4 bar

8 bar

8 bar12 b8r12 bar

**

NS

NSNSNS

.NSNS

NSNSNS

1.25 IJn1

4 bar

4 bar

8 bar

8 bar12 bar12 bar

.

.NS

NSNSNS~

NSNSNS

.

NSNS

NSNSNS

degrees to the fiber axis, and are shown in Fi~s 9 and10, respectively. It is possible to see from the surface pro-filesthcdJ'ff~~-ill sw:i~~c -topo~itphy uf tIre-area-scanned. Macerated fibers have a relatively flat surfacewhilst the treated fibers get progressively more undulating.The change in surface roughness increases up to refinerpressures of 8 bar for both fiber types; the frequency ofoscillations increasing with increased pressure. The sur-faces of fibers refined at 12 bar have a much larger variationin height, but there are fewer surface oscillations. Althoughevident in mature fibers, these surface features are espe-cially notable for the juvenile fibers. This was to be expectedas the SEM images and AFM 5 micron scans showed thatthe surfaces of fibers refined at 12 bar looked smooth butbulbous. Surface line traces are also shown in Figure 6 fora 5 ~ scan of juvenile 8 bar fibers and the subsequent 2.5aDd 1-.2.)~ images obtained using the zoom~function. Itcan be seen that with increased magnification the surfacetrace becomes flatter. This may explain the more statisti-cally significant results for the 2.5 ~m scan size for rough-ness. Surface features below 2.5 ~ may no longer be smallenough to be recorded.

the fiber. This would account for the high variability foundfor the quantitative measurements. SEM images show fibersrefilled at 12 barwerevclY 1._glucuted witt. luts uf breaksand lumps on the surface, whereas the 8 bar fibers showedmainly delamination.

One of the limitations of the AFM is that the maximumvertical deflection of the cantilever is 6 J1m. Thus, someareas randomly chosen for AFM -analysis had to be avoideddue to excessively large features. Debris on the fiber sur-faces also proved to be problematic occasionally due tovibrational interruptions of the cantilever. Thus, areas withloose debris also had to be avoided. Also if any loose mate-rial causes vibrations, a bad signal is obtained, so areas withdelaminated debris could not be sampled. This is probablywhy the re~ults showed that 8 bar fibers were rougher than12 bar fibers.

Conclusion

The results from this study show that refining at differentpressures has a significant effect on the surface morphol-ogy of wood fibers. Raising the pressure, in general, wasfound to increase the surface roughness. Most of the surfacechanges occur at refiner pressures between 4 and 8 bar.Fibers could be statistically distiguished best scanned at5 ~ and analyzed at 2.5~. Further study is now requiredto see how the increased roughness affects the bondingstrength in structural fiberboards. A subsequent study iscurrently underway utilizing the phase images of the 5 ~scans in an attempt to discover the usefulness of these dataas indicators of the chemical groups that are present on thesurface of the treated fibers.

General discussion

The combination of the two microscopy methods werefound to be complementary. Although the SEM requiresthat the sample is desiccated and coated with gold, it allowsgroups of fibers to be seen in detail and a larger surface areacan be observed. Images obtained from the AFM weredifficult to interpret visually and are, for practical reasons,restricted to individual fibers and scan sizes of approxi-mlilely 25 ~m or less. SEM images therefore gave an indi-t'ulillll of the overall effect of refining on the fiber surface,\\'hil~llht AFM allows for quantitative measurements of thefit"" -'Ir"u~'l' III :I much higher magnification. SEM images,...~ Ihnt lilt. xlirlilCC clln vary greatly along the length of

Acknowledgement

The authors wish to thank the Composite Panel Association fortheir financial support.

Holzforschung / Vol. 55/2001/ No. S

520 R. Snell el al.: Surface Morphology ofTMP Fibers

~' ,~

Panshin. AJ. and C. deZeeuw. 1980. Textbook ofW<K>d Technol-ogy (4th Edition). McGraw-Hili, New York. NY. 722 p.

Pearson, AJ. 1983. Towards a unified theory of mechanical pulp-ing and refining. In: Proceedings of the international mecjl1ni-cal pulping conference, Washington. pp. 131-138.

Pesacreta, T.C., L.C. Carlson and B.A. Triplett. 1997. Atomic forcemicroscopy of cotton fiber Cell wall surfaces in air and water:quantitative and qualitative aspects. Planta 202, 435-442.

Suchland, O. and G.E. W<K>dson. 1986. Fiberboard manufacturingpractices in the United States. United States Department ofAgriculture. Forest Service. Agricultural Handbook No. 640.263 p.

Sundholm, J. 1993. Can we reduce energy consumption in mechan-ical pulping? In: EUCEPA proceedings of the 18m interna-tional mechanical pulping conference, June 15-17, 1993, Oslo.pp. 133-142.

Weast, R.C, (Ed.). 1975. CRC Handbook of Chemistry andPhysics, 561h Edition. CRC Press, Cleveland, Ohio.

I

=~=-~"'~~~

References

Abe H., J. Ohtani and K. Fukuzawa K. 1991. FE-SEM observa-tions on the microfibrillar orientation in die secondary wall oftracheids. IAWA Bull. N.S. !2(4), 431-438.

Beland, M.-C. 1996. CLSM and AFM applied in pulp and paperresearch - A literature review. IOF-Technical Report TR 312,Institute of Optical Research, Stockholm, Sweden. 18 p.

Boras L. and P. Gatenholm. 1999. Surface composition and mor-phology of CTMP Fibers. Holzforschung 53,188-194.

Goring, D.A.I. 1971. Polymer properties of lignin and lignin deriv-atives. In: Lignins, Occurrence, Formation, Structure, andReactions. Eds. Sarkantem, K. V. and Ludwig, C.H. Wiley-Interscience, New York, NY. pp. 695-761.

Hanley, S.J. and D.G. Gray. 1994. Atomic force microscope imagesof black spruce wood sections and pulp fibres. Holzforschung48, 29-34.

Harada, H. Y. andW. A. COte. 1985. Structure of wood. ,": Biosyn-dlesis and Biodegradation of Wood Components. Ed. T. Higu-chi. Academic Press, Orlando, FL. pp. 1-42.

Kamis, A. 1993. The mechanism of fiber _elopment in mechan- Received May 18111 2<XX>ical pulping. In: EUCEPA proceedings of die 18111 internationalmechanical pulping conference, June 15-17, 1993, Oslo, pp.268-293. Rebecca Snell

Megraw, R.A., D. Bremer, G. Leaf and 1.. Roers. 1999. Stiffness in The Bi4JComposite Cen~loblolly as a function of ring~ition and height, and its rela- University of Wales!ionship to mi~rofibril angle and spec!fic gravity. In: ~- . Bangor, Wales, LL57 2UWing~ n( th~ Third Wn~~hn!':Ct:lllllectionBetween Silviculture YK - ---

and Wood Quality through MOdeling Approaches and Simula-tion Software.IUFRO working party S5:01-04, La Londe-Les-Maures, France, September 5-12, pp. 341-349. Leslie H. Groom and Timothy G. Rials

Nissan, A.H. and S.S. Sternstein. 1964. Cellulose-fiber bonding. U.S. Forat ServiceTappi 47(1), 1-6. Southern Research Station

Page, D.H. I 9!xJ. The structure and properties of paper. Part I - The 2500 Shreveport Hwystructure of paper. In: Wet-End Operations Seminar Notes. Pineville, LA, 71360Tappi,Atlanta, GA. May 8-12, 1989. pp. 7-12. U.S.A.