Carbohydrate Polymers 87 (2012) 1026 1037

Contents lists available at ScienceDirect

Carbohydrate Polymers

j ourna l ho me pag e: www.elsev ier .com

Surface area and porosity of acid hydrolyzed cellucellulo s

Jing Guoa Intercollege G y Parkb Department o Parkc Center for Na

a r t i c l

Article history:Received 16 AReceived in reAccepted 26 JuAvailable onlin

Keywords:Cellulose nanoAcid hydrolysiBacterial celluSurface areaPorosity

uch atain

celluese murfacenus) ad pors pr

ting mydrol

porosity compared to plant CNWs. Cellulose synthesized by G. xylinus ATCC 700178 from agitated culturesalso exhibited less surface area and porosity than those from static cultures.

2011 Elsevier Ltd. All rights reserved.

1. Introdu

Celluloseglucopyranand aggreglose producstructure. Iis determinenvironmenJacob-Wilk,a rosette (Bof individubril. Multilarger bundin the amoalong the ceof elementaof porosity lose had powidth (Agg

Corresponogy, The PennTel.: +1 814 86

E-mail add

0144-8617/$ doi:10.1016/j.ction

is a high molecular weight polymer comprised of d-osyl units but is complicated due to different packingation of cellulose chains, which varies among cellu-ing organisms, yielding a heterogeneous and poroust is believed that the shape of an elementary briled by the cellulose synthase as well as by the localt in which the cellulose is produced (Doblin, Kurek,

& Delmer, 2002). In plants, cellulose is produced byrown & Montezinos, 1976), which produces a numberal glucan chains which crystallize into an elementaryple elementary brils will subsequently aggregate intoles. It has been suggested that pores may be presentrphous regions, which may exist as periodic defectsllulose bril (Nishiyama et al., 2003). The aggregationry brils into bundles has also been cited as a source(Delmer & Amor, 1995). It was found that cotton cellu-res 45 nm in diameter, with a crystallite of 45 nm inebrandt & Samuelson, 1964; Klemm, Heublein, Fink, &

ding author at: Intercollege Graduate Degree Program in Plant Biol-sylvania State University, University Park, PA 16802, USA.3 0414; fax: +1 814 863 1031.ress: jcatchmark@engr.psu.edu (J.M. Catchmark).

Bohn, 2005; Morosoff, 1974), while wood cellulose crystallites havebeen determined to have widths of 3.54 nm (Ioelovitch, 1992).Compared to plant cellulose, in which cellulose I is the majorcomponent, bacterial cellulose (e.g., bacterium Gluconacetobacterxylinus, a well characterized cellulose producer) is predominantlycellulose I (Attala & Vanderhart, 1984). Bacterial cellulose hasa unique structure, which is almost pure without any lignin orhemicellulose. Moreover, bacterial cellulose possesses higher sur-face area compared to plant cellulose (Krieger, 1990; Sakairi,Asano, Ogawa, Nishi, & Tokura, 1998) mainly due to the reducedber diameter and open network structure. It has been proposedthat bacterial cellulose has crystallinity of 75% (Kulshreshtha &Dweltz, 1973), and its crystallites are 56 nm in width (Klemmet al., 2005). It is known that different origins and treatmentsare responsible for the complexity and variability of cellulosestructures. Knowledge of surface area and porosity are impor-tant in understanding the structure, formation and degradationof cellulose from different origins and treatments. Understand-ing the surface structure and porosity of cellulose is fundamentalto its use as a feedstock for biofuels. For example, previous worksuggested that the surface area of cellulose was of particular sig-nicance in affecting enzymatic cellulose hydrolysis (Fan, Lee, &Beardmore, 1981) because the direct physical contact between cel-lulases and the surface of cellulose determined the accessibility ofthe cellulases to cellulose, which was a prerequisite to hydrolysis(Thompson, Chen, & Grethlein, 1992). High surface area and pore

see front matter 2011 Elsevier Ltd. All rights reserved.carbpol.2011.07.060se produced by Gluconacetobacter xylinua,c, Jeffrey M. Catchmarka,b,c,

raduate Degree Program in Plant Biology, The Pennsylvania State University, Universitf Agricultural and Biological Engineering, The Pennsylvania State University, UniversitynoCellulosics, The Pennsylvania State University, University Park, PA 16802, USA

e i n f o

pril 2011vised form 21 July 2011ly 2011e 10 August 2011

whiskersslose

a b s t r a c t

The physical parameters of cellulose sof cellulose composites which may conof acid hydrolyzed nano-dimensionalyet the surface area and porosity of thof this work was to characterize the ston/bacterium Gluconacetobacter xylias well as to compare surface area ancultures. Our results showed that CNWface area and porosity relative to starhydrolyzed CNWs but not in H2SO4 h/ locate /carbpol

lose nanowhiskers and

, PA 16802, USA, PA 16802, USA

s surface area and porosity are important in the developmentvaluable additives which bind to cellulose. In this area, the uselose nanowhiskers (CNWs) has attracted signicant interest,aterials have not been explored experimentally. The objective

area and porosity of CNWs from different origins (plant cot-nd different acid treatments (H2SO4/HCl) by N2 adsorption;osity of bacterial cellulose synthesized by static and agitatedoduced from H2SO4/HCl exhibited signicantly increased sur-aterial cotton ber CF11. Micropores were generated in HCl

yzed CNWs. Bacterial CNWs exhibited larger surface area and

J. Guo, J.M. Catchmark / Carbohydrate Polymers 87 (2012) 1026 1037 1027

volume guarantee sufcient adsorption of the cellulases to the cel-lulose surface. In this area, cellulose nanowhiskers (CNWs) are asubject of intense investigation due to their exceptional physicaland mechanical properties, size, availability, and ability to be func-tionalized wchemical pr& Rojas, 20a comprehehydrolyzed

Numeroacterizing sprobes (e.gsolute excluface area offrom that plase adsorpThis resultsaffected notby their spexclusion msured pore size, i.e., thnot be evalthe externasolution, cethe brils, wment; and competition1989). Howapproximatby most oth

Several cellulose mdiscrepanciGkaraveli, NAhlgren, 19Mackie, Cladependenceand drying porosity chusing the msample prepsamples treable to be cohydrolysis a

2. Materia

2.1. Cellulo

Avicel Ppulp, was plulose powInternationduced fromw/w) accorMathew, an1959; Marcadded into lowed by swashed cellw/w), the hstirring for 9washing wiized by dial

after sonication. The HCl treated CNWs were generated from eitherWhatmanTM CF11 (cotton) or G. xylinus cellulose by hydrochlo-ric acid (15%, w/w). The hydrolysis parameters were taken fromAraki, Wada, Kuga, and Okano (1998). The HCl treated CNW sus-

n wal at 8ntrifydroper frlowe

the er fse lly oror 7 dsed f

defo 2001

to aanos

at stalsdrieructureme

2 ads

N2 aeleraetri

es t amal ovriumsorpents e, an

& Shc surBET)t of ing t) stanto 3in thcropion oableg moo- an

develvin eHorvly ap

all tracteed ato the

ray d

detetion ANaluKith specic surface chemistries through a variety ofocesses (De Souza Lima & Borsali, 2004; Habibi, Lucia,10; Samir, Alloin, & Dufresne, 2005). To date, however,nsive study of the surface area and porosity of acid

CNWs has not been performed experimentally.us methods have been developed and applied for char-urface area and porosity of cellulose, e.g., N2 adsorption,., bovine serum albumin (BSA), peroxidase, etc.) andsion methods. However, by using probes, a specic sur-

0.08 m2/g of Avicel PH101 has been shown, differentredicted by the N2 adsorption (2.07 m2/g) or by cellu-tion method (2.5 m2/g) (Gama, Teixeira, & Mota, 1994).

from the limitation of using probes, which are highly only by the surface characteristics of cellulose but alsoecic adsorption afnities. On the other hand, soluteethod also exhibits several disadvantages: (1) the mea-volume and area are largely dependent on the probee pores not accessible to probes of certain sizes mayuated; only the internal pore surface can be estimated,l surface cannot be determined; (2) when measured inllulose may swell due to the penetration of water intohich may overestimate the porosity by pore enlarge-

(3) an inaccurate estimation may also rise from the of water to solute probes (Burns, Ooshima, & Converse,ever, the pore size determined by N2 adsorption is fromely 0.4 nm to 300 nm (Allen, 1997), a range not covereder methods.studies have been performed on the surface area ofaterials using the N2 adsorption method, which showes in the range of 0.58200 m2/g (Papadopoulos, Hill,talos, & Karastergiou, 2003; Stone, Scallan, Donefer, &69; Thode, Swanson, & Becher, 1958; Wong, Deverell,rk, & Donaldson, 1988). These discrepancies suggest a

of the surface area on sample selection, pretreatmentmethod. In this work, we evaluated the surface area andaracteristics of a number of relevant celluloses puriedost commonly implemented methods. By consistentaration and N2 adsorption procedure, various celluloseated by different acids and from different origins werempared. The scope of this study is to describe how acidnd origins inuence the microstructure of cellulose.

ls and methods

se samples

H101 microcrystalline cellulose, extracted from woodurchased from SigmaAldrich. WhatmanTM CF11 cel-der (cotton origin) was purchased from Whatmanal Ltd. The cellulose nanowhiskers (CNWs) were pro-

WhatmanTM CF11 cellulose by sulfuric acid (63.5%,ding to the procedures implemented by Bondeson,d Oksman (2006; Marchessault, Morehead, & Walter,hessault, Morehead, & Koch, 1961). 55 g cellulose was1 L of 0.1 M NaOH solution with stirring for 12 h, fol-everal washing steps to eliminate residual NaOH. Theulose was then added into 500 ml of sulfuric acid (63.5%,ydrolysis was carried out under 45 C with continuous0 min and further hydrolysis was stopped by additionalth deionized (DI) water, after which the pH was neutral-ysis with DI water and CNW suspension was obtained

pensio4 N HCeral ceAfter hthe upin the tion ofthe lowCellulostatica30 C fwas uits the(Galle,will beKarathfrozenice cryfreeze-nanostmeasu

2.2. N

Thean AccMicromquantiing themateriequilibThe adsuremvolumLowellspeciTeller (ear parAccord(IUPACsied ipores are materizatare capbindinof mesmodelthe Ketional is mainappliedity chadegassprior t

2.3. X-

To diffracon a Pwith Cs prepared by treating 5 g of cellulose with 175 ml of0 C for 225 min. The sediment was collected after sev-ugations, and then puried by dialysis and sonication.lysis, the solution contained both cellulose particles inaction of suspension and sediments which accumulatedr fraction of the container. In this work, the upper frac-solution refers to the smaller suspended CNWs whileraction of the solution refers to the larger sediments.lms were generated from G. xylinus strain ATCC 700178

under agitation in HestrinSchramm (HS) medium atays (Hestrin & Schramm, 1954). Freeze-drying methodor cellulose sample preparation as this process lim-rmation of the polymer during removal of the solvent; Ratti, 2001). Especially, the shrinkage of pore structure

large degree prevented at low temperature (Krokida,, & Maroulis, 1998). All the cellulose samples were rst80 C. This low temperature prevents the growth of big

that form in larger pores. After freezing, samples wered (40 C, 0.035 mbar, under such pressure cellulosere will not be affected) as powders for N2 adsorptionnts.

orption measurements

dsorption measurements were carried out at 77 K usingted Surface Area and Porosimetry Analyzer (ASAP 2020;tics Instrument Corp.). The N2 adsorption techniquehe surface area and porosity characteristics by measur-ount of N2 adsorbed and desorbed onto a porous solider a wide range of relative pressures (P/P0, where P is the

pressure and P0 is the saturation pressure) of 1081.0.tion isotherms obtained from these adsorption mea-allow for the determination of the surface area, pored pore size distributions (PSDs) (Gregg & Sing, 1982;ields, 1991; Rouquerol, Rouquerol, & Sing, 1999). Theface area is calculated based on Brunauer, Emmett and

theory (Brunauer, Emmett, & Teller, 1938) from the lin-the adsorption isotherm, at pressures 0.05 < P/P0 < 0.30.o the International Union of Pure and Applied Chemistryndard (Sing et al., 1985), porous materials can be clas-

categories: pores 50 nmores. Pores can be open or closed. In this work, charac-f the open pores is of specic interest, since open pores

of accommodating cellulases for cellulose hydrolysis, orlecules useful for cellulose nanocomposites. The PSDsd macropores are calculated using classical pore sizeloped by Barret, Joyner and Halenda (BJH), based onquation (Barret, Joyner, & Halenda, 1951). The conven-athKawazoe (HK) model (Horvath & Kawazoe, 1983)plied for micropore size calculations. In this work, wehree methods for evaluating the surface area and poros-ristics of various forms of cellulose. All the samples were

303.15 K for 240480 min under vacuum at 10 m Hg measurements.

iffraction measurements

rmine the crystallinity and crystallite size, the X-ray(XRD) patterns of the cellulose samples were collectedytical XPert Pro MPD (multi-purpose diffractometer)

radiation generation. The diffraction intensity was

1028 J. Guo, J.M. Catchmark / Carbohydrate Polymers 87 (2012) 1026 1037

Table 1Crystallinity and crystallite size of cellulose samples.

Sample Crystallinity (%) Crystallite size (nm)

CF11 CNW-H2SO4CNW-HClBC-CNW-HCBC-staticBC-agitated

Only the uppecellulose fromproduced by Gwere obtainedquantitatively(2 0 0) diffract

measured bData, Inc., Lterns and tcrystallite sintensity (F(Zhang, Xi,

hkl = Lhklwhere is plane in outaking the rays ( = 0.1crystallite l

2.4. Scannin

Celluloseonto an aland then sp(GEMINI) operating aphology of cellulose m

2.5. Atomic

A dropleand the samdust cover.used for CNin air with cwere obtainresonance fThe oscillat

3. Results

3.1. Cellulo

XRD waof celluloseand crystaluated via tOur XRD (TH2SO4 or Hhave compConversely,hydrolyzedcantly redu78.2 0.4 7.4 0.190.5 1.4 6.8 0.189.3 1.6 6.6 0.1

l 94.5 1.9 5.8 0.082.2 0.3 6.2 0.071.5 1.3 5.8 0.1

r fractions of CNWs were measured by XRD. BC represents bacterial G. xylinus ATCC 700178. BC static and agitated represent the cellulose. xylinus with static and agitated cultures, respectively. The values

from at least triplicates. The crystallite size of cellulose samples was evaluated via the width at half-height of maximum intensity of theion peak.

etween 2 of 560. MDI Jade 9 software (Materialsivermore, CA) was used to process the diffraction pat-o calculate the crystallinity of cellulose samples. Theize was estimated from the full width at half-maximumWHM) of the reection (2 0 0) using Scherrer equationMo, & Jin, 2003).

K

cos hkl

the breadth of the peak of a specic phase (hkl, (2 0 0)r case), K is a constant that varies with the method ofbreadth (K = 0.94), is the wavelength of incident X-5418 nm), is the center angle of the peak, and L is theength (size).

g electron microscopy (SEM)

samples were freeze-dried as powders or lms, xeduminum slab with two-sided adhesive carbon tape,utter-coated with gold (thickness 1 nm). A LEO 1530eld emission scanning electron microscope (FESEM)t 2 kV (low voltage was applied to make sure the mor-cellulose samples would not be affected) was used fororphology observation.

force microscopy (AFM)

t of diluted CNWs was dispensed onto a mica surfaceple was dried at room temperature overnight with a

A NanoScope III atomic force microscope (AFM) wasW morphology observation. All scans were performedommercial Si nanoprobe tips. Height and phase imagesed simultaneously in tapping mode at the fundamentalrequency of the cantilever with a scan rate of 0.50.8 Hz.ing amplitude was 0.5 V.

and discussionse crystallinity and crystallite size determined by XRD

s used to investigate the changes of phase structures before and after acid treatments. The crystallinitylite size of cellulose samples were quantitatively eval-he relative intensity of the (2 0 0) diffraction peak.able 1) results suggest that the CNWs produced fromCl hydrolysis (from both plant and bacterial origins)arable crystallinity, both higher than that of CF11.

decreased crystallite size is observed in H2SO4/HCl CNWs. This indicates that H2SO4/HCl digestion signi-ces the amorphous content and/or CNW surface glucan

Fig. 1. N2 adsorption and desorption isotherms of cellulose samples: (a) untreatedcotton CF11 and both upper and lower fractions of CNWs prepared from CF11 byH2SO4 and HCl hydrolysis; (b) upper and lower fractions of CNWs generated fromCF11 and G. xylinus cellulose; (c) cellulose by G. xylinus from static and agitatedcultures. Adsorption isotherm is depicted in solid line and desorption is in dashedline. STP stands for standard conditions of temperature and pressure.

J. Guo, J.M. Catchmark / Carbohydrate Polymers 87 (2012) 1026 1037 1029

chains delamination may give rise to smaller, more highly crys-talline cellulose nanocrystals. On the other hand, differences areobserved in the structure of bacterial cellulose crystals producedin different culture conditions. In agitated culture, the crystallinityand crystallthan pelliclvious reporsize were obby the G. xy

3.2. Validatarea

Avicel Ptially used measuremeof 1.32.2 m1981; Sinittions of thewas conside

3.3. N2 adsdistribution

CNWs ption due tmechanicaland differeN2 adsorptsamples froFig. 1 (Supp(PSDs) werFig. 2. Accomeso- and macropore (DBJH) and mples are sushow a typewhich is aclose (IUPACsample canAccording tuntreated cwhich the athe desorptThe adsorpsharply asceven at P/Pcellulose CFration preswith aggrepores (Lowis predictedallel bundlethe isothermcompared wHCl and H2adsorptionsacid hydrolexhibit highlikely that tmore amorpthat this ocas the size generation tion (Gregg

model to describe the lling behavior in individual pores, there is acorrelation between pore radius (r) and pore condensation pressure(P/P0),( )

=

P isses,

of ls theppen

the sultsted C

P/P0monpeakmacrisapp

and smacicauppetion

is sl hete

necps shysis tary

re bea slits as itminarestins oors. e sha/PSD

It is sncreation

CNWwith

pore it is

CNWcomfracts of ed. Ome s

mighen elated outeysis. inatioby acerenysis eW sus a hithe el hydy, duite size of the cellulose pellicles are found to be smalleres formed in static culture, which is consistent with pre-t that a comparable smaller crystallinity and crystalliteserved in agitated culture as compared to static culturelinus strain NQ-5 (Czaja, Romanovicz, & Brown, 2004).

ion of N2 adsorption for quantifying cellulose surface

H101, commercial microcrystalline cellulose, was ini-to test the validity and accuracy of the surface areant by N2 adsorption. Previous results showed a range2/g (Ardizzone et al., 1999; Burns et al., 1989; Fan et al.,syn, Gusakov, & Vlasenko, 1991), due to minor varia-

Avicel samples purchased, our result of 2.4 0.7 m2/gred to be acceptable.

orptiondesorption isotherms and pore sizes (PSDs)

roduced by acid treatments have received much atten-o their high surface area and unique physical and

properties. To study the effects of acid hydrolysisnt origins on the pore structure of cellulose, typicaliondesorption isotherms at 77 K for all the cellulosem different acid treatments and origins are shown inlementary data). Corresponding pore size distributionse calculated using the BJH method and are shown inrdingly, the calculated BET specic surface area, totalmacropore (1.7300 nm) area (ABJH), total meso- andvolume (VBJH), average meso- and macropore widthicropore (0.41.7 nm) volume (VHK) of cellulose sam-

mmarized in Table 2. Generally, these N2 isotherms IV adsorption isotherm behavior with hysteresis loop,companied by capillary condensation in porous cellu-

classication). The geometry of each porous cellulose be predicted based on the shape of the hysteresis loop.o IUPAC classication, the N2 adsorption isotherm ofotton CF11 reveals a type H3 hysteresis loop (Fig. 1a) indsorption branch is steep at saturation pressure whileion branch is steep at intermediate relative pressure.tion branch rises gradually until P/P0 = 0.9, and thenends over P/P0 = 0.9, and does not reach stable status0 = 1, showing that N2 vapor conned in the pores of11condenses at a relative pressure lower than its satu-sure. This type H3 hysteresis loop is usually observedgates of plate-like particles giving rise to slit-shapedell, Shields, Thomas, & Thommes, 2004). Therefore, it

that slit pores exist in cotton ber CF11 between par-s of cellulose elementary brils. After acid hydrolysis,s of CNWs produced by H2SO4 and HCl hydrolysis areith untreated CF11, as shown in Fig. 1a. In general, bothSO4 generated CNWs have signicantly increased N2

and wider hysteresis loops than CF11, indicating thatysis increases the surface area and porosity. Since CNWser crystallinity in comparison to CF11 (Table 1), it is nothe increased surface area comes from the formation ofhous regions in cellulose. Therefore, it is hypothesizedcurs because of the increase of external surface areaof the cellulose particle decreases, as well as new poreduring acid treatments. According to the Kelvin equa-

& Sing, 1982) with an assumption of cylindrical/slit pore

lnP

P0

wherecondentensionN2, R iwill hameansOur regenerahigheralso demajor these they druptedpeak to

Spe(both adsorpbranchwith anarrowsis loohydrolelementhe potively, widenhas eli

Intefractiobehavithat thradius2002).with iadsorptreatedbined similarCNWs,tion offrom inupper bundleretainethat sowhichbetwespeculphous hydroldelamCNWs

Diffhydrolthe CNthere icating the HCtionall2rRT

the pressure under which N2 in a pore of radius (r) P0 is the saturation pressure for N2, is the surfaceiquid N2, is the molar volume of the condensed liquid

gas constant and T is the temperature. Condensation at lower P/P0 in small pores than in large pores, whichhysteresis loop will start at lower P/P0 for small pores.

suggest that CF11 has larger pores than H2SO4/HClNWs because the hysteresis loop of CF11 starts at a(0.65 vs. 0.4 for H2SO4/HCl generated CNWs). This isstrated by the PSDs shown in Fig. 2a and b in which a

around 100 nm is observed in CF11. It is predicted thatopores must exist between aggregates of brils sinceear after acid treatments when the aggregates are dis-

no longer joined, as suggested by the shift of the majorller sized pores in PSDs of CNWs.lly, the hysteresis loops of H2SO4 generated CNWsr and lower fractions) are of type H4, in which thebranch is steep at saturation pressure, and desorptionoping. This type H4 hysteresis loop is always associatedrogeneous assembly of capillaries with wide bodies andks (Allen, 1997). After H2SO4 hydrolysis, the hystere-ifted from type H3 to H4, which suggests that H2SO4may open up disordered regions within the bundle of

brils and thus make double tapered capillaries (wherecomes wider at the center) from slit-like pores. Alterna--like pore may become wedge-shaped where the pore

penetrates into the bundle, where continued hydrolysisted amorphous cellulose contained inside.ngly, H2SO4 generated CNWs where the differentf the CNW suspension exhibit varied N2 adsorptionAn earlier study on ordered mesoporous silica showedpe and the width of hysteresis loop depended on pore

at a given temperature (Thommes, Koehn, & Froeba,uggested that the width of the hysteresis loop increasessed pore size and/or PSD. In our case, increased N2and wider hysteresis loop of the lower fraction of H2SO4s as compared to the upper fraction are observed. Com-

the PSD result from Fig. 2a and b, which suggests a size of the upper and lower fractions of H2SO4 treated

likely that the wider hysteresis loop of the lower frac-s results from broad PSD. In addition, larger aggregatesplete hydrolysis exist in the lower fraction but not in theion of the CNWs. These larger aggregates contain intactlementary brils in which some porous regions are stilln the other hand, PSDs from Fig. 2a and b also suggestmall pores 1.7 nm are generated in the upper fractiont be explained by hydrolysis of amorphous celluloseementary brils. Preston and Cronshaw (1958) havethat an elementary cellulose bril contains an amor-r shell, which would presumably be susceptible to acidAlternatively, these small pores may arise due to then of the cellulose sheets on the surface and ends of theid hydrolysis, as discussed in more detail in Section 3.5.t from H2SO4 generated CNWs, CNWs made from HClxhibit varied hysteresis loops from different fractions ofspension. As it is shown in Fig. 1a, for the upper fraction,gh adsorption at low relative pressure (P/P0 < 0.4), indi-xistence of large amount of micropores (0.41.7 nm) inrolyzed CNWs obtained from the upper fraction. Addi-e to a large number of small mesopores of 1.72 nm

1030 J. Guo, J.M. Catchmark / Carbohydrate Polymers 87 (2012) 1026 1037

Fig. 2. BJH PSDs from N2 adsorption isotherms. PSDs are plotted as pore volume per unit pore width (cm3/g A) and pore area per unit pore width (m2/g A), respectively. (a,b) CNWs generated by H2SO4 and HCl hydrolysis (plant origin); (c, d) plant cotton and bacterial CNWs generated by HCl hydrolysis; (e, f) bacterial cellulose produced fromstatic and agitated cultures.

J. Guo, J.M. Catchmark / Carbohydrate Polymers 87 (2012) 1026 1037 1031

Fig. 3. FESEM and AFM images of cellulose produced under different conditions: (a) cotton CF11; (b) CNWs generated from cotton CF11 by H2SO4 hydrolysis; (c) CNWsgenerated from cotton CF11 by HCl hydrolysis; (d) CNWs produced from G. xylinus ATCC 700178 cellulose by HCl hydrolysis; (e) cellulose from G. xylinus ATCC 700178produced statically; (f) cellulose from G. xylinus ATCC 700178 produced during agitation. Scale bars are shown in the gures. Cellulose samples were freeze-dried as powders(ad) or lms (e, f), xed onto an aluminum slab with two-sided adhesive carbon tape, and sputter-coated with gold (thickness 1 nm) for FESEM observation. (g) AFM phaseimage of cotton CNWs produced by H2SO4 hydrolysis; (h) AFM phase image of cotton CNWs produced by HCl hydrolysis.

1032 J. Guo, J.M. Catchmark / Carbohydrate Polymers 87 (2012) 1026 1037

Table 2Specic surface area, macro-, meso- and microporosity of cellulose samples by BET, BJH and HK methods.

Sample Origin Treatment BET surface (>0.4 nm) Macro-, mesopore (1.7300 nm) Micropore (0.41.7 nm)

A (m2/g) A (m2/g) V JH (cm3/g) Pore size (nm) V (cm3/g) Median pore

CF11 005 CNW U1 007 CNW L1 015 CNW U2 014 CNW L2 015 CNW U3 042 CNW L3 054 BC S 220 BC A 148

U, L represent o CNWrefers to the C e bacagitated cultur the c

generated umes are cohydrolyzedpore size (lower fractithe lower frsimilar to Htaken from and H4 typboth ink-bothe desorpthydrolyzedpores, ink-bnected porein more isois nearly coreporting sitypical widton the ordeThis suggeselementaryIt may be pbetween eledle, which mor within bsuch a bundwhich exhiloop whichFig. 1a, i.e., aviously in acomplete hytals. In thiswould be gthe cellulosmicropores

In gener(plant origiexistence oamount of ticular, theR-HClupper